14.

9 mm 187 x 235 mm

Clinical Dilemmas in

Clinical Dilemmas in Diabetes

Diabetes
Edited by

Adrian Vella MD is the Earl and Annette R. McDonough Professor of Medicine, and the Research
Clinical Dilemmas in

Diabetes
Chair of the Division of Diabetes, Endocrinology & Metabolism in the Department of Medicine,
Mayo Clinic College of Medicine, USA.

Clinical Dilemmas in Diabetes answers the clinical questions commonly encountered when
­diagnosing, treating, and managing patients with diabetes and its associated complications.
­Designed to support informed, evidence-based care, this authoritative clinical guide includes
­contributions from leading endocrinologists and diabetes researchers that discuss a diverse range
of recent developments. Concise and focused chapters cover prediabetes, diabetes diagnosis, initial
Second Edition
evaluation and management, disease complications, and cardiovascular disease and risk factors.
Edited by Adrian Vella
Now in its second edition, Clinical Dilemmas in Diabetes contains extensively reviewed and
­revised information throughout. New and updated chapters examine prediction, diagnosis, and
­management of early Type 1 diabetes, ophthalmic complications, screening asymptomatic patients
for cardiovascular disease, new agents for treatment of dyslipidemia, closed loop systems in Type 1
diabetes, upper gastrointestinal manifestations, managing hyperglycemia in critically ill patients,
and more. Edited by Dr. Vella at the Mayo Clinic, this highly practical resource:

Vella
• Encourages evidence-based clinical decision-making, rather than algorithm-based approaches

• P rovides clear guidance on common problematic areas, especially in cases where conflicts in
­treatment for the disease and the complications occur

• E mphasizes the importance of translating the results of clinical trials to individual care and
­management of diabetes

• C
 ontains effective learning and revision tools, including learning points, chapter introductions

Edition
Second
and summaries, tables, figures, charts, and full references

Part of the popular Clinical Dilemmas series, Clinical Dilemmas in Diabetes is a must-have guide
for anyone involved in the treatment of patients with diabetes, particularly endocrinologists,
diabetes specialists and consultants, cardiologists, residents, fellows, specialist nurses, and general
­practitioners with an interest in diabetes.

Cover Design: Wiley

Cover Images: © dboystudio/Shutterstock, SCIEPRO/SCIENCE
PHOTO LIBRARY/Getty Images, MIRIAM MASLO/Science Source,
DR P. MARAZZI/Science Source, JUAN GAERTNER/SCIENCE
PHOTO LIBRARY/Getty Images, Scott Camazine/Getty Images

www.wiley.com
Clinical Dilemmas in

Diabetes
Clinical Dilemmas in

Diabetes
Second Edition

EDITED BY

Adrian Vella MD FRCP(Edin.)

Division of Diabetes, Endocrinology & Metabolism
Earl and Annette R. McDonough Professor
Mayo Clinic
Rochester, MN, USA
This edition first published 2022
© 2022 John Wiley & Sons Ltd
Edition History
Blackwell Publishing Ltd (1e, 2011)
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or
otherwise, except as permitted by law. Advice on how to obtain permission to reuse material
from this title is available at http://www.wiley.com/go/permissions.
The right of Adrian Vella to be identified as the author of the editorial material in this work has
been asserted in accordance with law.
Registered Offices
John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA
John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK
Editorial Office
9600 Garsington Road, Oxford, OX4 2DQ, UK
For details of our global editorial offices, customer services, and more information about Wiley
products visit us at www.wiley.com.
Wiley also publishes its books in a variety of electronic formats and by print-­on-­demand. Some
content that appears in standard print versions of this book may not be available in other formats.
Limit of Liability/Disclaimer of Warranty
The contents of this work are intended to further general scientific research, understanding, and
discussion only and are not intended and should not be relied upon as recommending or
promoting scientific method, diagnosis, or treatment by physicians for any particular patient. In
view of ongoing research, equipment modifications, changes in governmental regulations, and
the constant flow of information relating to the use of medicines, equipment, and devices, the
reader is urged to review and evaluate the information provided in the package insert or
instructions for each medicine, equipment, or device for, among other things, any changes in the
instructions or indication of usage and for added warnings and precautions. While the publisher
and authors have used their best efforts in preparing this work, they make no representations or
warranties with respect to the accuracy or completeness of the contents of this work and
specifically disclaim all warranties, including without limitation any implied warranties of
merchantability or fitness for a particular purpose. No warranty may be created or extended by
sales representatives, written sales materials or promotional statements for this work. The fact
that an organization, website, or product is referred to in this work as a citation and/or potential
source of further information does not mean that the publisher and authors endorse the
information or services the organization, website, or product may provide or recommendations it
may make. This work is sold with the understanding that the publisher is not engaged in
rendering professional services. The advice and strategies contained herein may not be suitable
for your situation. You should consult with a specialist where appropriate. Further, readers should
be aware that websites listed in this work may have changed or disappeared between when this
work was written and when it is read. Neither the publisher nor authors shall be liable for any loss
of profit or any other commercial damages, including but not limited to special, incidental,
consequential, or other damages.
Library of Congress Cataloging-­in-­Publication Data
Names: Vella, Adrian, editor.
Title: Clinical dilemmas in diabetes / edited by Adrian Vella.
Description: Second edition. | Hoboken, NJ : Wiley-Blackwell, 2022. |
Includes bibliographical references and index.
Identifiers: LCCN 2021025568 (print) | LCCN 2021025569 (ebook) | ISBN
9781119603160 (paperback) | ISBN 9781119603177 (adobe pdf) | ISBN
9781119603184 (epub)
Subjects: MESH: Diabetes Mellitus–therapy | Diabetes Mellitus–diagnosis |
Evidence-Based Medicine–methods
Classification: LCC RC660.4 (print) | LCC RC660.4 (ebook) | NLM WK 810 |
DDC 616.4/62–dc23
LC record available at https://lccn.loc.gov/2021025568
LC ebook record available at https://lccn.loc.gov/2021025569
Cover Design: Wiley
Cover Images: © dboystudio/Shutterstock, SCIEPRO/SCIENCE PHOTO LIBRARY/Getty
Images, MIRIAM MASLO/Science Source, DR P. MARAZZI/Science Source, JUAN
GAERTNER/SCIENCE PHOTO LIBRARY/Getty Images, Scott Camazine/Getty Images
Set in 8.75/12pt Minion by Straive, Pondicherry, India

10 9 8 7 6 5 4 3 2 1
 Contents
List of Contributors,  vii
Preface, x
Part I  Prediabetes and the Diagnosis
of Diabetes
1 “Is Prediabetes a Risk Factor or Is It a Disease?”  3
Jacob Kohlenberg and Adrian Vella
2 Early Diagnosis of Type 1 Diabetes – Useful or a Pyrrhic
Victory? 17
Paolo Pozzilli and Silvia Pieralice
3 Reclassifying or Declassifying Diabetes? Can Clinical
Characteristics Guide Classification and
Treatment? 37
Adrian Vella
4 How Should Secondary Causes of Diabetes
Be Excluded? 45
Tomás P. Griffin, Aonghus O’Loughlin, and Sean F.
Dinneen
5 How to Screen Appropriately for Monogenic Diabetes  68
Adrian Vella
Part II  Initial Evaluation and Management
of Diabetes
6 Managing Gestational Diabetes During and After
Pregnancy 75
Aoife M. Egan and Fidelma P. Dunne
7 What Is the Role of Self‐Monitoring in Diabetes? Is There
a Role for Postprandial Glucose Monitoring? How Does
Continuous Glucose Monitoring Integrate into Clinical
Practice? 87
Rami Almokayyad and Robert Cuddihy
8 Does HbA1c Remain the Most Important Therapeutic
Target in Outpatient Management of Diabetes?  101
Kristen Gonzales and Steve A. Smith
9 Technology Issues: Continuous Glucose Monitoring,
Insulin Pumps, and Closed Loop Control for Patients
with Diabetes 115
Ravinder Jeet Kaur, Shafaq R. Rizvi and Yogish C. Kudva
10 Optimizing Diet in Patients with Diabetes  124
Meera Shah
11 Are Insulin Sensitizers Useful Additions to Insulin
Therapy? 133
John W. Richard, III and Philip Raskin
12 Incretin‐Based Therapy for the Management of Type 2
Diabetes 146
Kristin Gonzales and Adrian Vella
Part III  Diagnosis and Management
of Cardiovascular Risk Factors
and Cardiovascular Disease
13 Screening Patients with Prediabetes and Diabetes
for Cardiovascular Disease  163
Sabreen Ahmed, Saritha Tirumalasetty and Vivian Fonseca
14 Choosing Medications for Type 2 Diabetes – What
Weighting Should Be Given to Cardiovascular Risk
Reduction? 174
Adrian Vella
15 Choosing Medications for Weight Loss in Type 2 Diabetes
Mellitus 187
Shubhada Jagasia, Chase Dean Hendrickson and
Alexander J. Williams
16 Are Statins the Optimal Therapy for Cardiovascular Risk
in Patients with Diabetes? What Newer Agents Are There
for the Treatment for Dyslipidemia in Diabetes? Are
Triglycerides an Important Risk Factor for Diabetes?  198
Recie Davern and Timothy O’Brien
v
 vi Contents

17 New Agents for Treatment of Dyslipidemia  211 21 Diagnosis and Management of Ophthalmic

Vinaya Simha Complications of Diabetes  252
18 The Role of Bariatric Surgery in Obese Patients
Andrew J. Barkmeier and Konstantin Astafurov
with Diabetes: Primary or Rescue Therapy?  220 22 Upper Gastrointestinal Manifestations of Diabetes  269
Praveena Gandikota and Blandine Laferrère Michael Camilleri
19 Treatment Strategies in Patients with Diabetes Mellitus
and Ischemic Heart Disease  228 Index 280
Madeline K. Mahowald, Chiam Leker Locker, Mandeep
Singh and Robert L. Frye

Part IV  Management of Disease

Complications
20 Diabetes Management in Patients with Critical and
Non-Critical Illness  241
Kalpana Muthusamy, Kristin Grdinovac and
John M. Miles
 List of Contributors

Sabreen Ahmed Robert Cuddihy Aoife M. Egan

Section of Endocrinology International Diabetes Center Division of Endocrinology, Diabetes,
Tulane University School of Medicine World Health Organization Collaborating Metabolism and nutrition
New Orleans, LA Center for Diabetes Education, Mayo Clinic
USA Translation and Computer Rochester, MN
Technology USA
Rami Almokayyad Minneapolis, MN
Division of Endocrinology, Department USA
Vivian Fonseca
of Medicine Section of Endocrinology
University of Minnesota Medical School Recie Davern Tulane University School of Medicine
University of Minnesota Department of Medicine and New Orleans, LA
Minneapolis, MN ­Endocrinology/Diabetes Mellitus USA
USA University College Hospital Galway and
National University of Ireland Robert L. Frye
Konstantin Astafurov Galway Department of Cardiovascular Disease
NJRetina/Rutgers Robert Wood Johnson Ireland Mayo Clinic
Medical School Rochester, MN
New Brunswick, NJ USA
Sean F. Dinneen
USA
Discipline of Medicine
NUI Galway Praveena Gandikota
Andrew J. Barkmeier Galway Endocrine, Diabetes and Nutrition Division
Department of Ophthalmology Ireland Department of Medicine
Mayo Clinic and St Luke’s Roosevelt Hospital
Rochester, MN Centre for Diabetes, Endocrinology and New York, NY
USA Metabolism USA
Galway University Hospitals
Galway
Michael Camilleri Kristen Gonzales
Ireland
Clinical Enteric Neuroscience Translational Department of Internal Medicine
and Epidemiological Research Program, Division of Endocrinology, Diabetes, and
Division of Gastroenterology and Fidelma P. Dunne Metabolism
Hepatology Galway Diabetes Research Centre University of New Mexico Health Sciences
Mayo Clinic National University of Ireland Center
Rochester, MN Galway Albuquerque, NM
USA Ireland USA

vii
 viii List of Contributors

Kristin Grdinovac Blandine Laferrère Silvia Pieralice

University of Kansas Division of Endocrinology, Diabetes Department of Endocrinology & Diabetes
Kansas City, KS and Nutrition Obesity Research Center Campus Bio-­Medico University of Rome
USA Department of Medicine Rome
St Luke’s Roosevelt Hospital Center Italy
Columbia University College of Physicians
Tomás P. Griffin
and Surgeons Paolo Pozzilli
Discipline of Medicine
New York, NY Department of Endocrinology & Diabetes
NUI Galway
USA Campus Bio-­Medico University of Rome
Galway
Rome
Ireland
Italy
and Chiam Leker Locker
Centre for Diabetes, Endocrinology and Department of Cardiovascular Disease
Philip Raskin
Metabolism Mayo Clinic
Clifton and Betsy Robinson Chair in
Galway University Hospitals Rochester, MN
B­iomedical Research
Galway USA
University of Texas Southwestern Medical
Ireland
Center at Dallas
Madeline K. Mahowald Dallas, TX
Chase Dean Hendrickson, Department of Cardiovascular Disease USA
Division of Diabetes, Endocrinology, and Mayo Clinic
Metabolism Rochester, MN John W. Richard III
Vanderbilt University Medical Center USA Division of Endocrinology, Diabetes,
Nashville, TN N­utrition and Metabolism
USA John M. Miles University of Texas Southwestern Medical
University of Kansas Center at Dallas
Kansas City, KS Dallas, TX
Shubhada Jagasia USA
Division of Diabetes, Endocrinology and USA
Metabolism Kalpana Muthusamy
Vanderbilt University Medical Center Shafaq R. Rizvi
Division of Endocrinology
Nashville, TN Division of Endocrinology, Diabetes,
Olmsted Medical Center
USA M­etabolism and Nutrition
Rochester, MN
Mayo Clinic
USA
Rochester, MN
Ravinder Jeet Kaur USA
Division of Endocrinology, Diabetes,
Timothy O’Brien
­Metabolism and Nutrition
Department of Medicine and Meera Shah
Mayo Clinic
­Endocrinology/Diabetes Mellitus Division of Endocrinology, Diabetes,
Rochester, MN
University College Hospital Galway and M­etabolism, and Nutrition
USA
National University of Ireland Mayo Clinic
Galway Rochester, MN
Jacob Kohlenberg Ireland USA
Division of Endocrinology, Diabetes,
­Metabolism, and Nutrition Aonghus O’Loughlin Vinaya Simha
Mayo Clinic Discipline of Medicine Division of Endocrinology, Diabetes,
Rochester, MN NUI Galway Metabolism and nutrition
USA Galway Mayo Clinic
Ireland Rochester, MN
Yogish C. Kudva and USA
Division of Endocrinology, Diabetes, Centre for Diabetes, Endocrinology and
­Metabolism and Nutrition Metabolism Mandeep Singh
Mayo Clinic Galway University Hospitals Department of Cardiovascular Disease
Rochester, MN Galway Mayo Clinic
USA Ireland Rochester, MN
USA
 List of Contributors ix

Steve A. Smith Adrian Vella Alexander J. Williams

Division of Endocrinology, Diabetes, Division of Endocrinology, Diabetes, Division of Diabetes, Endocrinology, and
M­etabolism, and Nutrition M­etabolism, and Nutrition Metabolism
Mayo Clinic Mayo Clinic Vanderbilt University Medical Center
Rochester, MN Rochester, MN Nashville, TN
USA USA USA

Saritha Tirumalasetty
Section of Endocrinology
Tulane University School of
Medicine
New Orleans, LA
USA
 Preface
While a week may be a long time in politics, the decade that
has elapsed since the original publication of Clinical
Dilemmas in Diabetes could be considered to be a very
(very) long time given the developments in clinical practice
that occurred. New therapeutics for the treatment of dys-
lipidemia and diabetes have appeared. Progress in the
development of an artificial pancreas and in sensor tech-
nology has improved the care of type 1 diabetes and some
diabetes medications may directly alter cardiovascular out-
comes. However, the original motivations behind the
book—­translating trial data into clinical practice—­remain
very relevant. The astute clinician will need to balance risks
and benefits, convenience and cost with changing clinical
needs, comorbidities, and patients’ social structures.
As before, the topics covered deserve discussion and
debate to avoid making every square peg fit into the
x
round hole of an algorithm or guideline. There has been
an expansion in the areas covered, reflecting my initial
desire for a book that reflects the growing facets of meta-
bolic care.
I would like to express my gratitude to Dr. Robert A.
Rizza MD, who served as a co-­editor for the first edition of
the book and as always serves as a sounding board for
many of my ideas. None of this would have been possible
without him. Of course, the other important characters in
this saga would be my parents—­especially my father, who
has now retired from General Practice—­ but whose
approach to problems I have always tried to emulate. Last,
but not least, I am very grateful to my wife Elsa, who hung
on gamely while I edited proofs of the first edition during
our honeymoon, and to our children Katie and Lucy, who
are always along for the ride.
PART I
Prediabetes and the Diagnosis
of Diabetes
 1 “Is Prediabetes a Risk Factor or Is It
a Disease?”
Jacob Kohlenberg1 and Adrian Vella2
1
Fellow and Instructor in Medicine, Division of Endocrinology, Diabetes, Metabolism, and Nutrition, Mayo Clinic,
Rochester, MN, USA
2
Professor of Medicine, Division of Endocrinology, Diabetes, Metabolism, and Nutrition, Mayo Clinic, Rochester,
MN, USA

LE A R NI NG P OI NT S between 140–199 mg/dL [2]. Unlike the ADA, the WHO

⚫⚫ Prediabetes is a heterogeneous condition with variable
risk of progression to type 2 diabetes
⚫⚫ In addition to diabetes risk, it is associated with an
increased risk of vascular disease.
To date, lifestyle modification is the single most
⚫⚫
important tool for altering the natural history of
prediabetes and progression to type 2 diabetes.
What is prediabetes?
Prediabetes is defined as an elevated fasting plasma glu­
cose (FPG), and/or an elevated 2‐hour plasma glucose
(2‐h PG) during a 75‐gram (g) oral glucose tolerance test
(OGTT), and/or an elevated Hemoglobin A1c (HbA1c),
without meeting diagnostic criteria for overt diabetes
mellitus (DM) [1]. The 2020 American Diabetes
Association (ADA) Guidelines define prediabetes as
impaired fasting glucose (IFG) with a FPG of 100–125
mg/dL, and/or impaired glucose tolerance (IGT) with a
2‐h PG during a 75‐g OGTT of 140–199 mg/dL, and/or a
HbA1c of 5.7–6.4% [1]. In contrast to the ADA, the 2016
World Health Organization (WHO) Guidelines define
intermediate hyperglycemia as IFG between 110–125 mg/
dL and/or IGT with a 2‐h PG during a 75‐g OGTT
does not include HbA1c as a diagnostic criterion for
prediabetes.
The definitions of both prediabetes and DM have
evolved in recent decades. The WHO first defined the
“borderline state” in 1965 as a 2‐h PG during a 50 or 100 g
OGTT between 110–129 mg/dL [3]. The ADA has long
recognized IGT, and its definition has undergone little
change since its inception. First adopted by the ADA in
1997 and WHO in 1999, the term IFG was originally
defined as FPG 110–125 mg/dL [4]. However, in 2003, the
ADA revised the criteria for IFG to 100–125 mg/dL based
on data from multiple studies showing that the risk of DM
increases markedly at a FPG concentration > 100 mg/dL
[5]. In 2010, the ADA added HbA1c as a diagnostic crite­
rion for prediabetes because the relationship between
HbA1c and the risk of retinopathy was similar to corre­
sponding FPG and 2‐h PG thresholds [6].
Rationale for the diagnostic criteria
for diabetes mellitus and prediabetes
Individuals with prediabetes have abnormal glucose regu­
lation and increased risk for developing DM type 2 (DM2)
and its complications [6]. The diagnostic thresholds for
DM are based on [1] the bimodal distribution of FPG and
2‐h PG during an OGTT and [2] the glycemic thresholds

Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.

© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.

3
 4 Prediabetes and the Diagnosis of Diabetes

Bimodal distribution of two-hour glucose levels

30

25

20
Percent

15

10

0
70 100 130 160 190 220 250 280 310 340 370 400 430
Two-hour plasma glucose mg/dL

FIG 1.1  Histogram with superimposed composite and component curves to describe the distribution of two‐hour plasma glucose
levels following an oral glucose load. Glucose concentrations and frequencies were arbitrarily chosen to illustrate a bimodal
distribution. The bimodal glucose distribution can be used to separate individuals into two groups, those with normoglycemia and
those with hyperglycemia. Intermediate glucose concentrations between normoglycemia and hyperglycemia helped define the
diagnostic thresholds for prediabetes.

for the development of microvascular complications, spe­ Epidemiology of prediabetes

cifically retinopathy (Figure 1.1). The bimodal distribution
and glycemic thresholds for the development of microvas­
cular complications have been demonstrated in many pop­
ulations including the Pima, Nauruans, South Africans,
Americans, Chinese, and Egyptians [7–15]. The bimodal
distribution of glucose has been used to separate individu­
als into two groups: those with normoglycemia and those
with hyperglycemia. IFG and IGT were defined as interme­
diates between normoglycemia and hyperglycemia. The
diagnostic thresholds for FPG, 2‐h PG during a 75‐g
OGTT, and HbA1c are relatively concordant in dis­
criminating between the two components of a bimodal
frequency distribution and their associations with
microvascular complications [4]. It is important to
note that defining a lower limit of an intermediate cat­
egory of FPG, 2‐h PG during a 75‐g OGTT, and HbA1c
is somewhat arbitrary because the risk of developing
DM is a continuum that extends into the normoglyce­
mic range [6].
The 2020 National Diabetes Statistics Report published by
the Centers for Disease Control and Prevention estimated
the prevalence of prediabetes (defined by 2020 ADA crite­
ria) to be 38.0% (95% confidence interval (CI) 35.2–40.8)
among adults in the United States (U.S.) [1, 16]. Overall,
the prevalence of prediabetes has not changed significantly
from 2005–2016. However, the number of U.S. adults who
are aware that they have prediabetes increased from 6.5%
(95% CI 5.3–7.9) in 2005–2008 to 13.3% (95% CI 11.0–16.0)
in 2013–2016. Further, the prevalence of prediabetes
increases with age: 29.1% (95% CI 25.2–33.3) of adults 18–44
years of age; 46.3% (95% CI 43.5–49.1) of adults 45–64 years
of age; and 51.0% (95% CI 46.5–55.5) of adults ≥ 65 years of
age. Prediabetes is also more common in men, whose preva­
lence is 42.3% (95% CI 38.1–46.5), compared to women,
whose prevalence is 33.7% (95% CI 30.7–36.8). The preva­
lence of prediabetes is similar among racial/ethnic groups
and among individuals of different education levels [16].
 “Is Prediabetes a Risk Factor or Is It a Disease?”
Pathogenesis of impaired fasting glucose
and impaired glucose tolerance
In epidemiologic studies, isolated IGT consistently has a
higher prevalence than isolated IFG [17]. The prevalence
of IFG and IGT increases with age [17]. In adults less than
55 years of age, IGT is more common in women and IFG is
more common in men [17]. This suggests that these two
states have different pathophysiologic mechanisms.
Genetics and lifestyle influence the pathogenesis of DM
[18, 19]. Although more than 400 genetic signals have been
identified as influencing risk for DM2, single polymor­
phisms add only small degrees of risk [20]. Polymorphisms
in the Transcription factor 7‐like 2 (TCF7L2) locus have
the largest‐known effect on risk for DM [20, 21]. Compared
with non‐carriers, heterozygous and homozygous carriers
of the at‐risk TCF7L2 variants have relative risks of devel­
oping DM of 1.45 and 2.41, respectively [20, 21].
The relationship between genetics and lifestyle on the
pathogenesis of DM was emphasized by a study comparing
the prevalence of DM2 in the Pima population in Mexico
versus the Pima population in the U.S. [22]. The prevalence
of DM2 in the Pima population in Mexico was 6.9%, com­
pared to 38% in the Pima population in the U.S. The preva­
lence of obesity in the Pima population in Mexico was
significantly lower than that in the Pima population in the
U.S. By comparison, the latter group also had significantly
lower physical activity levels.
Overall, islet cell function is the primary regulator of
glucose metabolism, but multiple additional factors con­
tribute. Postprandial and fasting glucose concentrations
are determined by insulin secretion, hepatic extraction,
insulin action, glucagon suppression, glucose effectiveness,
and the rate of gastric emptying (Figure 1.2) [23–29].
Studies using the minimal model to quantitate insulin
secretion and insulin action demonstrated that both indi­
ces are lower in subjects with NFG/IGT and IFG/IGT than
in subjects with NFG/NGT [25, 26]. However, there was no
significant difference in insulin secretion and insulin
action between subjects with IFG/NGT and those with
NFG/NGT. This implies that insulin secretion declines in
concert with insulin action across the spectrum of predia­
betes; however, subjects with isolated IFG may have an
altered glucose threshold for insulin secretion without
other intrinsic defects in β‐cell function.
5
Using human pancreatic tissue obtained at autopsy,
obese individuals with IFG have a 40% deficit in relative
beta‐cell volume compared to obese individuals with NFG
[30]. Additionally, the diabetes‐associated genetic varia­
tion in TCF7L2 is associated with impaired insulin secre­
tion [31–33]. The inability of insulin secretion to
compensate for a decline in insulin action results in
hyperglycemia.
It is also essential to consider the role of hepatic extrac­
tion in the pathophysiology of prediabetes and DM [26].
The peripheral insulin concentration reflects portal insulin
secretion, hepatic extraction, distribution, and degradation
[34]. Hepatic extraction of secreted insulin has a direct
relationship with disposition index. Reduced hepatic
extraction in prediabetes and DM may be a compensatory
measure to increase circulating insulin. Interestingly, insu­
lin pulse characteristics influence hepatic extraction of
insulin, and the diabetes‐associated genetic variation in
TCF7L2 is associated with abnormal insulin pulse charac­
teristics [33].
Furthermore, lack of glucagon suppression contributes
to hyperglycemia in subjects with impaired insulin secre­
tion [24]. The diabetes‐associated genetic variation in
TCF7L2 is associated with impaired glucagon suppression
[31, 35]. In a longitudinal study, defects in α‐cell function
with elevated fasting glucagon concentrations were associ­
ated with a subsequent decline in β‐cell function [36].
Screening recommendations
for prediabetes
The 2020 ADA Guidelines recommend screening asymp­
tomatic overweight or obese adults for DM or prediabetes
if at least one of the following risk factors is present: first‐
degree relative with DM; high‐risk race/ethnicity (e.g.
African American, Latino, Native American, Asian
American, Pacific Islander); history of cardiovascular dis­
ease; hypertension; high‐density lipoprotein cholesterol <
35 mg/dL; triglyceride level > 250 mg/dL; presence of poly­
cystic ovary syndrome; physical inactivity; or other clinical
considerations associated with reduced insulin action such
as acanthosis nigricans [1]. The 2020 ADA Guidelines fur­
ther recommend that patients with prediabetes be tested
yearly and that women with a history of gestational diabetes
be tested lifelong at least every three years. For all other
 6 Prediabetes and the Diagnosis of Diabetes
Peripheral
tissue
Glucose
Stomach
effectiveness
Glucose
Absorption effectiveness
Glucose
EGP
Glucagon
suppression
FIG 1.2  Postprandial and fasting glucose concentrations are determined by insulin secretion, hepatic insulin extraction, insulin
action, glucagon suppression, glucose effectiveness, endogenous glucose production, and the rate of gastric emptying [23–29].
EGP = Endogenous Glucose Production.
patients, screening should begin at age 45, and if results are
normal, should be repeated at least every three years.
Reproducibility, sensitivity, and specificity
of glycemic measurements and the role
of oral glucose tolerance testing in clinical
practice
A study that analyzed data from the Second Examination
of the Third National Health and Nutrition Examination
Survey (NHANES) found that the within‐person coeffi­
cient of variation was 16.7% (95% CI 15.0–18.3) for 2‐h
PG, 5.7% (95% CI 5.3–6.1) for FPG, and only 3.6% (95% CI
3.2–4.0) for HbA1c [37]. A study of non‐diabetic adults
reported a reproducibility rate of 65.6% for a 75‐g OGTT
repeated twice over a six‐week period [38].
HbA1c, FPG, and 2‐h PG during a 75‐g OGTT have
different sensitivities and specificities in the diagnosis of
Systematic appearance of portal
insulin/peripheral insulin action
Liver &
splanchnic
tissues Hepatic insulin
extraction/hepatic
insulin action
Portal insulin secretion
Glucagon
suppression
Pancreas
prediabetes and DM [39]. An analysis of NHANES data
from 2005–2010 examined adults without self‐reported
DM at baseline with available measurements for HbA1c,
FPG, and 2‐h PG [40]. Prediabetes and DM were defined
according to the current ADA Guidelines [1]. Using HbA1c
thresholds of ≥ 6.5% for DM and ≥ 5.7% for prediabetes
resulted in low sensitivity (24.9% and 35.4%, respectively)
and high specificity in identifying patients diagnosed with
FPG and 2‐h PG. When FPG and HbA1c were used
together for diagnosis, the false‐negative rate was 45.7% for
DM and 9.2% for prediabetes.
Similar findings were reported in the Insulin Resistance
Atherosclerosis Study (IRAS), which examined 855 sub­
jects and defined DM and prediabetes in accordance with
the 2020 ADA Guidelines (Tables  1.1 and 1.2) [1, 41].
HbA1c ≥ 6.5%, FPG ≥ 126 mg/dL, and 2‐h PG ≥ 200 mg/dL
identified 32.3%, 44.8%, and 86.8% of individuals with
 “Is Prediabetes a Risk Factor or Is It a Disease?”
TABLE 1.1  The sensitivity of various indices of glycemia used
to diagnose type 2 diabetes mellitus in the Insulin Resistance
Atherosclerosis Study. Diabetes mellitus was defined according
to the 2020 American Diabetes Association criteria. This data
illustrates that omitting an oral glucose tolerance test leads
to under‐diagnosis of diabetes mellitus, even when both fasting
plasma glucose and hemoglobin A1c are measured concur-
rently. HbA1c = Hemoglobin A1c. FPG = Fasting Plasma
Glucose. 2‐h PG = 2‐hour Plasma Glucose [41].
Sensitivity
HbA1c ≥ 6.5% 32.3%
FPG ≥ 126 mg/dL 44.8%
2‐h PG ≥ 200 mg/dL 86.8%
HbA1c ≥ 6.5% and/or FPG ≥ 126 mg/dL 52.2%
FPG ≥ 126 mg/dL and/or 2‐h PG ≥ 200 mg/dL 97.1%
TABLE 1.2  The sensitivity of various indices of glycemia used
to diagnose prediabetes in the Insulin Resistance Atherosclerosis
Study. Prediabetes was defined according to the 2020
American Diabetes Association criteria. This data illustrates that
omitting an oral glucose tolerance test leads to under‐diagnosis
of prediabetes, even when both fasting plasma glucose
and hemoglobin A1c are measured concurrently. HbA1c =
Hemoglobin A1c. FPG = Fasting Plasma Glucose. 2‐h PG = 2‐hour
Plasma Glucose [41].
Sensitivity
HbA1c 5.7–6.4% 23.6%
FPG 100–125 mg/dL 69.1%
2‐h PG 140–199 mg/dL 59.5%
HbA1c 5.7–6.4% and/or FPG 100–125 mg/dL 75.6%
FPG 100–125 mg/dL and/or 2‐h PG 140–199 95.8%
mg/dL
DM, respectively. The combination of HbA1c ≥ 6.5% and/
or FPG ≥ 126 mg/dL detected 52.2% of all individuals with
DM. HbA1c between 5.7–6.4%, FPG between 100–125
mg/dL, and 2‐h PG between 140–199 mg/dL identified
23.6%, 69.1%, and 59.5% of all non‐diabetic subjects with
prediabetes. The combination of HbA1c 5.7–6.4% and/or
FPG 100–125 mg/dL detected 75.6% of all non‐diabetic
subjects with prediabetes.
Despite the limitations of the OGTT – including lower
reproducibility and reduced patient convenience compared
to FPG and HbA1c – the aforementioned studies suggest
7
that the OGTT continues to have a role in clinical practice.
An individual eating three daily meals will be in a post­
prandial state for 6–9 hours per day [42]. Therefore, it is
logical that the knowledge gained from a standardized glu­
cose load is clinically informative. The 2020 ADA
Guidelines state that FPG, 2‐h PG during a 75‐g OGTT,
and HbA1c are equally appropriate to test for prediabetes
and DM [1]. Many clinicians screen with both a FPG and
HbA1c. However, the omission of an OGTT leads to a sig­
nificant under‐diagnosis of both DM and prediabetes.
Risk of progression from prediabetes
to diabetes mellitus
The transition from prediabetes to DM2 is variable and
influenced by heredity and lifestyle [18, 19, 43]. Multiple
variables have been used to estimate the risk of progression
to DM2, including glycemic indices, anthropometric data,
comorbidities, metabolomics, and genetic variants.
Numerous prediction models have been created based on
these variables [43–51].
Baseline FPG is a significant predictor of an individual’s
risk for developing DM [43]. In nondiabetic adults residing
in Minnesota, baseline FPG levels < 100 mg/dL, 100–109
mg/dL, and 110–125 mg/dL were associated with a 7%,
19%, and 39% risk, respectively, with progression to DM
over a median of 9 years. A discrete gradient of risk for
progression to DM was also observed among subjects with
a baseline FPG < 100 mg/dL.
During a 75‐g OGTT, 30‐minute, 60‐minute, and 120‐
minute PG concentrations were all significant predictors
for future risk of DM [49, 52]. One‐hour plasma glucose
during a 75‐g OGTT has been shown to be a better pre­
dictor of future DM than FPG, HbA1c, and PG at 30 min­
ute and 120 minutes during a 75‐g OGTT [50]. In one
study, the hazard ratio for the development of DM was 9.5
(7.90–11.43) for those with combined IFG‐IGT at base­
line, 4.5 (4.03–5.02) for those with isolated IGT at base­
line, and 3.98 (3.16–5.02) for those with isolated IFG at
baseline [45].
Baseline HbA1c is also a strong predictor for risk of
future DM, and the risk of incident DM significantly
increased across the HbA1c range of 5.0–6.5% [46]. A sys­
tematic review found that the 5‐year incidence of DM was
 8 Prediabetes and the Diagnosis of Diabetes
25–50%, 9–25%, and 5–9% among those with a baseline
HbA1c of 6.0–6.5%, 5.5–6.0%, and 5.0–5.5%, respectively.
Numerous studies have shown that demographic data,
anthropometric data, and comorbidities are associated
with progression to DM. Factors positively associated with
progression to DM include: family history of diabetes, for­
mer or active smoking, higher BMI, abdominal obesity,
increased waist circumference, hypertension, elevated tri­
glycerides, low HDL cholesterol, and elevated high sensi­
tivity C‐reactive protein [45, 53, 54]. Increasing age has
also been shown to be a predictor of DM in some studies.
Metabolomics profiles have been investigated as a tool
to estimate the risk of developing DM. Numerous metabo­
lites have been found to be positively and negatively associ­
ated with progression to DM [51]. More than 400 distinct
genetic signals that affect the risk of developing DM2 have
been identified [20, 44]. Polygenic scores have been used to
estimate the combined genetic risk for the development
of DM.
Microvascular complications associated
with prediabetes
Retinopathy
In a large cohort of Pima individuals, the frequency of
retinopathy – defined as the presence of at least one hem­
orrhage, one microaneurysm, or proliferative retinopa­
thy – was directly related to baseline FPG and 2‐h PG [55].
Beginning at a baseline FPG threshold of 6.0–6.4 mmol/L
and 2‐h PG threshold of 9.0–10.6 mmol/L, there was a sig­
nificant increase in period prevalence (10‐year interval) of
retinopathy. Additional data from the Pima individuals
illustrated that beyond a HbA1c threshold of 6.2% there
was a significant increase in the prevalence of retinopathy
[12]. Similar glycemic thresholds for an increase in the
prevalence of retinopathy were observed in a cross‐sec­
tional study of Egyptians and data from the Third National
Health and Nutrition Examination Survey [13, 14].
Additionally, in the Diabetes Prevention Program (DPP),
diabetic retinopathy was detected in 7.9% of the subjects
with IGT and in 12.6% of the subjects who developed DM
[56]. The prevalence of retinopathy is significantly higher
in subjects with DM, but retinopathy can develop in sub­
jects with prediabetic range dysglycemia.
Pooled data analysis of nine studies from five countries
examined glycemic thresholds for diabetes‐specific retin­
opathy (defined as moderate or more severe retinopathy)
in 44 623 subjects [57]. A curvilinear relationship was
found to exist for FPG and HbA1c when diabetic retinopa­
thy was plotted against continuous glycemic measures.
Diabetes‐specific retinopathy began to increase from a
FPG of 6.0–6.4 mmol/L and from a HbA1c of 6.0–6.4%.
Based on vigintile distributions, glycemic thresholds for
diabetes‐specific retinopathy were observed over the range
of 6.4–6.8 mmol/L for FPG, 9.8–10.6 mmol/L for 2‐h PG,
and 6.3–6.7% for HbA1C. Compared with the first vigin­
tile, the odds ratios for diabetes‐specific retinopathy for the
above vigintile distributions were 2.5 (95% CI 1.2–5.2) for
FPG, 10.1 (95% CI 1.3–79.4) for 2 h PG, and 4.5 (95% CI
1.4–15.2) for HbA1c.
A major limitation of many studies that examine the
prevalence of retinopathy is the diagnostic criteria used to
define diabetes‐specific retinopathy [57]. The use of an
overly broad definition for retinopathy leads to the inclu­
sion of subjects with mild retinopathy, which may have eti­
ologies other than hyperglycemia. Overall, the rate of
retinopathy increases with the degree and duration of
hyperglycemia [56, 57]. Subjects with prediabetes have a
higher prevalence of retinopathy than subjects with NFG/
NGT, although the prevalence remains relatively low. In
many populations, the glycemic threshold for retinopathy
occurs in the prediabetic range of dysglycemia [57, 58].
Nephropathy
Analysis of NHANES data from 1999–2006 revealed that
chronic kidney disease (CKD), defined as either reduced
kidney function or elevated albuminuria (urinary albu­
min‐creatinine ratio ≥ 30 mg/g), was present in 17.1% of
prediabetics compared to 11.8% of individuals with nor­
moglycemia [59]. A multiethnic study found that 16.1% of
subjects with IGT had microalbuminuria, compared to 4%
of subjects with NGT [60]. After adjusting for multiple
variables, glycemic status was found to be the most signifi­
cant determinant of urinary albumin concentration. A sys­
tematic review and meta‐analysis including 9 studies with
185 452 subjects reported that prediabetes was modestly
associated with an increased risk of CKD [61]. After adjust­
ing for established risk factors, the relative risk of CKD was
 “Is Prediabetes a Risk Factor or Is It a Disease?”
1.11 (95% CI 1.02–1.21) for subjects with FPG between
6.1–6.9 mmol/L.
The prevalence and five‐year incidence of nephropathy
increases as FPG, 2‐h PG, and HbA1c rise [12, 58]. Of note,
the association of glycemia with nephropathy is weaker
than the association between glycemia and retinopathy.
When plotting prevalence of microalbuminuria against
FPG, 2‐h PG, and HbA1c, there is a visible inflection point
and subsequent increase in microalbuminuria prevalence
beyond a FPG of 5.5 mmol/L, 2‐h PG of 5.5 mmol/L (and
again at 9.3 mmol/L), and HbA1c of 5.8%. In summary,
multiple studies suggest that prediabetic‐range hyperglyce­
mia is associated with higher rates of nephropathy.
Neuropathy
Although neural dysfunction is associated with hypergly­
cemia, clinicians should be mindful that neurologic deficits
can be attributable to non‐glycemic causes in individuals
with prediabetes and DM [62]. Subjects with isolated IGT
were shown to have subclinical neural dysfunction that was
generally asymptomatic and characterized by small‐fiber
neuropathy and mild impairment of cardiovascular auto­
nomic function [63]. Erectile dysfunction has also been
shown to be independently associated with IFG, with an
odds ratio of 1.26 (95% confidence interval 1.08–1.46)
[64]. Prediabetic‐range hyperglycemia is also associated
with chronic idiopathic axonal polyneuropathy (CIAP)
[65]. In a study of 100 subjects with CIAP, 36 individuals
were found to have IFG, 3 had FPG ≥ 126 mg/dL, 38 had
IGT, and 24 had 2‐h PG ≥ 200 mg/dL. Overall, the preva­
lence of dysglycemia in this cohort was approximately 2‐
fold higher than in an age‐matched general population
group.
Macrovascular complications
and mortality associated
with prediabetes
There is an increased prevalence of cardiovascular disease
(CVD) in individuals with prediabetes, but this relation­
ship is confounded by common‐risk factors present in
CVD and prediabetes [66–68]. However, after accounting
for non‐glycemic cardiovascular risk factors, both IFG and
IGT are still associated with a modestly increased risk of
9
developing CVD. It is possible that much of this risk is due
to the increased risk of ultimately progressing to DM.
Approximately 25% of first myocardial infarctions (MIs)
are unrecognized, which are predictive of future major car­
diovascular events including death [69]. In a multi‐ethnic
population‐based cohort study adjusted for cardiovascular
risk factors, it was shown that subjects with IFG have a
higher prevalence of unrecognized MIs than those with
NFG, with an odds ratio of 1.60 (95% CI 1.01–2.48).
In the Whitehall Study, after 5–10 years of follow‐up,
survival by baseline glucose tolerance status diverged
among the groups, and median survival differed by approx­
imately 4 years between the normoglycemic and glucose
intolerant groups [70]. Overall, all‐cause mortality, cardio­
vascular mortality, and respiratory mortality were higher
among participants with glucose intolerance. The hazard of
coronary mortality rose beginning at a 2‐h PG of 83 mg/dL;
however, the graded relationship diminished after adjust­
ing for multiple variables including baseline CVD.
A systematic review and meta‐analysis incorporated 53
prospective studies with 1 611 339 subjects who were fol­
lowed for a median of 9.5 years for cardiovascular and
mortality outcomes [68]. IFG and IGT diagnostic criteria
were in accordance with 2020 ADA guidelines [1].
Compared to individuals with normoglycemia, those with
IGT or IFG had an increased risk of composite CVD (rela­
tive risk (RR) 1.13 for IFG and 1.30 for IGT), coronary
heart disease (RR 1.10 for IFG and 1.20 for IGT), stroke
(RR 1.06 for IFG and 1.20 for IGT), and all‐cause mortality
(RR 1.13 for IFG and 1.32 for IGT). One limitation of this
systematic review and meta‐analysis is that some of the
included studies did not adjust for progression to DM dur­
ing the follow‐up period.
A separate meta‐analysis examined 102 prospective
studies with 698 782 subjects and showed that FPG was
modestly and non‐linearly associated with vascular dis­
ease, with hazard ratios for coronary heart disease of 1.11
for FPG of 5.6–6.09 mmol/L and 1.17 for FPG of 6.1–6.99
mmol/L [71]. Another meta‐analysis that included 97 pro­
spective studies with 820 900 subjects calculated hazard
ratios for cause‐specific death according to baseline FPG
[72]. After adjusting for multiple variables and excluding
subjects with known CVD at baseline, FPG was found to be
nonlinearly related to risk of death. Compared with
 10 Prediabetes and the Diagnosis of Diabetes
subjects with NFG, subjects with IFG had hazard ratios of
1.13 for cancer deaths, 1.17 for vascular deaths, and 1.12
for non‐cancer and non‐vascular deaths.
Management of prediabetes
The goals of prediabetes management include preserving
β‐cell function, delaying or preventing the onset of DM,
delaying or preventing the developing of microvascular
and macrovascular complications associated with hyper­
glycemia, and reducing the cost of diabetes care.
Diabetes prevention program
The Diabetes Prevention Program (DPP) enrolled 3234
nondiabetic adult subjects at 27 centers in the U.S. [18].
Eligibility criteria included FPG of 95–125 mg/dL and a
2‐h PG during a 75‐g OGTT 140–199 mg/dL. Subjects
were assigned to one of three groups: (1) intensive lifestyle
modification (goal ≥ 7% weight loss of initial body weight
and ≥ 150 minutes of moderate intensity physical activity/
week); (2) metformin 850 mg twice daily plus standard
lifestyle recommendations; or (3) placebo plus standard
lifestyle recommendations. Standard lifestyle recom­­men­
dations were provided in writing and through annual brief
individual sessions [73]. In contrast, the intensive lifestyle
modification provided comprehensive instruction in a
structured 16‐lesson curriculum.
The original DPP results were published after an aver­
age follow‐up of 2.8 years. The estimated cumulative inci­
dence of DM at three years was significantly different
among all groups: 28.9% in the placebo group; 21.7% in the
metformin group; and 14.4% in the intensive lifestyle
group. Weight loss was the main predictor of reduced DM
incidence, with a hazard ratio per five‐kilogram (kg)
weight loss of 0.42 (95% CI 0.35–0.51). Further, for every
one kg of weight loss, there was a 16% reduction in risk of
progression to DM [74].
Following randomization, DPPOS followed subjects for
15 years and metformin continued to be provided to the
group originally assigned to it [75]. Over 15 years of fol­
low‐up, the cumulative incidence of DM was 62% in the
placebo group, 56% in the metformin group, and 55% in
the intensive lifestyle group. At the end of DPPOS, the
aggregate prevalence of microvascular outcomes  –  which
included nephropathy, retinopathy, and neuropathy – did
not differ among the 3 treatment groups. However, for
women, intensive lifestyle intervention significantly
reduced aggregate microvascular disease at 15 years com­
pared to metformin and compared to placebo. Additionally,
for those subjects with a baseline BMI ≥ 35 kg/m2, the RR
for the development of aggregate microvascular disease
was significantly lower in the intensive lifestyle interven­
tion group compared to the placebo group. Among partici­
pants whose most recent HbA1c was ≥ 6.5%, the intensive
lifestyle intervention group showed statistically significant
reductions in the aggregate microvascular outcome, retin­
opathy, and neuropathy compared with placebo and
metformin.
Other benefits of intensive lifestyle changes were seen in
DPP subjects [76]. From baseline to year three after rand­
omization, hypertension increased in the placebo and met­
formin groups but decreased in the intensive lifestyle
group. From baseline to year three, dyslipidemia pro­
gressed in all three groups but the progression was less in
the intensive lifestyle group compared to the metformin
and placebo groups. After a mean follow‐up of 3.2 years in
the DPP, there were significant improvements in quality of
life measures for the intensive lifestyle group, but not for
the other two groups [77]. From a payer perspective, 10
years after randomization in DPP, intensive lifestyle
changes were cost‐effective, and metformin was marginally
cost‐saving compared to placebo [78].
Da Qing study
In 1986, a population‐based survey identified subjects in
Da Qing, China with IGT [79]. These subjects were then
randomized into four groups: control group, diet only,
exercise only, and diet plus exercise. At six years post rand­
omization, the mean rate of DM was significantly higher in
the control group at 66%, compared to 47% in the diet
group, 45% in the exercise group, and 44% in the diet plus
exercise group.
The original Da Qing participants were followed for up
to 30 years after randomization to assess the effects of
intervention of DM incidence, microvascular and macro­
vascular complications, and mortality [80]. Active inter­
vention occurred for the first six years after randomization
until 1992, after which subjects were informed of the study
 “Is Prediabetes a Risk Factor or Is It a Disease?”
results and asked to continue with normal medical care. No
specific interventions were offered after the initial six years,
and the three intervention groups were combined into one
group for analysis purposes.
Over the 30‐year follow‐up period, the intervention
group had a median delay in DM onset by four years (NNT
10) compared to the control group and a significantly lower
cumulative incidence of DM onset (HR 0.61) [80]. At 30
years, there were 26% fewer CVD events in the interven­
tion group compared to the control group. The difference
between the two groups continued to increase over time.
At 30 years, the incidence of retinopathy was 40% lower
in the intervention group than in the control group, and
incidence of nephropathy and neuropathy were numeri­
cally lower in the intervention group but not significantly
different [80]. The median delay of composite microvascu­
lar disease outcome was 5.2 years in the intervention group
(NNT 10). Cardiovascular and all‐cause mortality were
also significantly lower in the intervention group (25.6%
and 35.2%, respectively) than in the control group (45.5% and
56.3%, respectively). The median delay in CVD death and
all‐cause mortality in the intervention group were 7.3 years
and 4.8 years, respectively, with NNT of 10 for both
outcomes.
Pharmacologic therapy for the prevention or
delay of diabetes mellitus
A systematic review that included 20 randomized con­
trolled trials examined the efficacy of metformin for the
prevention or delay of DM [81]. The overall conclusion was
that for at‐risk subjects, metformin compared with placebo
reduced or delayed the risk of progression to DM. The inci­
dence of DM was not significantly different when compar­
ing metformin plus intensive diet and exercise and identical
intensive diet and exercise alone.
In addition to metformin, other medications have
proven efficacy in reducing progression to DM. Once‐daily
subcutaneous liraglutide 3.0 mg as adjunct therapy to life­
style modifications reduced progression to DM compared
to placebo after 160 weeks [82]. In adults with elevated
CVD risk and prediabetes or newly established DM, a
once‐daily insulin glargine injection reduced progression
to DM compared to placebo [83]. Several thiazolidinedi­
ones also have proven efficacy in reducing the progression
11
to DM, including pioglitazone, rosiglitazone, and troglita­
zone [84–86]. Additionally, combined hormone replace­
ment therapy in post‐menopausal women, glipizide,
valsartan, orlistat, and acarbose also significantly reduced
the progression to DM compared to placebo [87–92].
Interestingly, during the Study to Prevent Non‐Insulin
Dependent Diabetes Mellitus (STOP‐NIDDM), acarbose
was associated with a 49% relative risk reduction in cardio­
vascular events compared to placebo with a HR 0.51 (95%
CI 0.28–0.95). However, the methodology of STOP‐
NIDDM has been heavily criticized and the validity of the
results has been questioned [93]. Many pharmacologic
therapies have also been studied and shown to be ineffec­
tive in preventing progression to DM, including vitamin D,
nateglinide, glimepiride, and ramipril [94–97].
Bariatric surgery for the prevention or delay
of diabetes mellitus
A Swedish study enrolled a large cohort of adult nondia­
betic subjects who had already chosen to undergo various
bariatric surgeries including banding (19%), vertical
banded gastroplasty (69%), or gastric bypass (12%), and
matched them with a nonrandomized control group [98].
At the time of enrollment, 17.2% of the control group had
IFG, compared to 16% of the surgical group. After a median
follow‐up time of 10 years, bariatric surgery compared
with usual care reduced progression to DM by 87% among
subjects with IFG. Another study reported that 98.6% of
subjects with prediabetic range dysglycemia had normal
FPG following gastric bypass [99].
Exercise and diet for the prevention or delay
of diabetes mellitus
A systematic review and meta‐analysis included 28 studies
with 1 261 991 subjects and investigated the role of physical
activity in reducing progression to DM [100]. Overall, for
those who achieved 11.25 metabolic equivalent of task
(MET) hours/week (equivalent to 150 minutes/week of
moderate activity) there was a risk reduction of 26% for
DM. Further risk reductions occurred at greater MET
hours/week. The Health Professionals Follow‐up Study
showed that engaging in weight training or aerobic exercise
for at least 150 minutes/week was independently associated
with a lower risk of DM of 34% and 52%, respectively
 12 Prediabetes and the Diagnosis of Diabetes
[101]. The greatest reduction in DM risk was seen in men
who engaged in both aerobic exercise and weight training
for at least 150 minutes/week, with a risk reduction of 59%.
The Prevención con Dieta Mediterránea (PREDIMED)
study enrolled older adult subjects at high risk for but with­
out baseline CVD and randomized participants to either a
Mediterranean diet supplemented with extra‐virgin olive
oil, a Mediterranean diet supplemented with mixed nuts,
or a control diet based on general advice to reduce fat
intake [102]. Compared to the control group, the
Mediterranean diet supplemented with extra‐virgin olive
oil group had significantly lower progression to DM with a
HR of 0.60 (95% CI 0.54–0.85). The beneficial effect seen
in this study was thought to be due to dietary composition
itself and not attributed to calorie restriction, increased
physical activity, or weight loss.
American diabetes association 2020 guideline
recommendations for the management
of prediabetes
The 2020 ADA guidelines recommend that patients with
prediabetes participate in an intensive behavioral lifestyle
intervention program modeled on the DPP to achieve and
maintain 7% weight loss and to achieve ≥ 150 minutes/
week of moderate intensity physical activity (such as brisk
walking) [103]. Additionally, the ADA recommends that a
variety of eating patterns are acceptable for patients with
prediabetes including the Mediterranean diet and a low‐
calorie, low‐fat diet. Regarding pharmacologic therapy, the
2020 ADA guidelines state that metformin for prevention
of DM should be considered in patients with prediabetes,
especially those with BMI ≥ 35 kg/m2, those aged < 60
years, and women with prior GDM.
Conclusion
In summary, patients with prediabetes have abnormal glu­
cose regulation mediated by impaired insulin secretion and
reduced insulin action, as well as an increased risk of pro­
gression to DM [18, 25, 26]. Many clinicians screen for dia­
betes and prediabetes using the combination of FPG and
HbA1c. However, even when used together, the combined
sensitivity for diagnosing DM and prediabetes is poor [41].
Therefore, omission of a 75‐g OGTT results in both
misclassification and under‐diagnosis of DM and predia­
betes. The following question then arises: are there clinical
consequences to misclassifying diabetes as only prediabe­
tes or prediabetes as normal glucose metabolism?
We conclude that the answer is yes. There is a clear asso­
ciation of increasing microvascular complications with ris­
ing glucose concentrations, and patients with prediabetes
have an increased prevalence of retinopathy, nephropathy,
and neuropathy compared to individuals with normal glu­
cose metabolism [13, 14, 55, 57, 59–61, 63, 104–106].
There is also an increased prevalence of CVD, cardiovascu­
lar mortality, and all‐cause mortality in patients with pre­
diabetes compared to those with normal glucose
metabolism [66–68, 70].
Therefore, recognition and treatment of prediabetes is
essential to minimize morbidity and mortality. Intensive
lifestyle changes emphasizing weight loss and physical
activity have been shown to prevent or delay progression to
DM, potentially decrease microvascular complications,
potentially reduce macrovascular complications, improve
comorbidities, improve quality of life, reduce medical
costs, and decrease mortality [18, 75–80]. In addition to
lifestyle changes, numerous medications are also effective
in preventing or delaying progression to DM including
metformin, glipizide, liraglutide, insulin glargine, several
thiazolidinediones, orlistat, acarbose, valsartan, and estro­
gen/progestin [18, 82–85, 87–92, 107].
The totality of available evidence suggests that prediabe­
tes is not only a risk factor, it is in fact a disease. Perhaps the
term “prediabetes” should be changed to “early diabetes”
and managed as such clinically.
References
1. American Diabetes Association, 2. Classification and
diagnosis of diabetes: standards of medical care in
diabetes – 2020. Diabetes Care. 2020;43(Suppl 1):S14–S31.
2. World Health Organization, Global Report on Diabetes. 2016.
3. Diabetes Mellitus. Report of a WHO expert committee.
World Health Organ Tech Rep Ser. 1965;310:1–44.
4. Report of the Expert Committee on the Diagnosis and
Classification of Diabetes Mellitus. Diabetes Care.
1997;20(7):1183–1197.
5. Genuth S et al. Follow‐up report on the diagnosis of diabetes
mellitus. Diabetes Care. 2003;26(11):3160–3167.
 “Is Prediabetes a Risk Factor or Is It a Disease?”
6. American Diabetes Association, Standards of medical care
in diabetes – 2010. Diabetes Care. 2010;33(Suppl 1):S11–61.
7. Bennett PH et al. Epidemiologic studies of diabetes in the
Pima Indians. Recent Prog Horm Res. 1976;32:333–376.
8. Zimmet P and Whitehouse S. Bimodality of fasting and
two‐hour glucose tolerance distributions in a Micronesian
population. Diabetes. 1978;27(8):793–800.
9. Rosenthal M et al. Evidence of bimodality of two hour
plasma glucose concentrations in Mexican Americans:
results from the San Antonio Heart study. J Chronic Dis.
1985;38(1):5–16.
10. Omar MA et  al. South African Indians show a high preva­
lence of NIDDM and bimodality in plasma glucose distribu­
tion patterns. Diabetes Care. 1994;17(1):70–73.
11. Fan J et al. Bimodality of 2‐h plasma glucose distributions in
whites: the Rancho Bernardo study. Diabetes Care. 2005;28(6):
1451–1456.
12. McCance DR et al. Comparison of tests for glycated haemo­
globin and fasting and two hour plasma glucose concentra­
tions as diagnostic methods for diabetes. BMJ. 1994;308(6940):
1323–1328.
13. Davidson MB et  al. Relationship between fasting plasma
glucose and glycosylated hemoglobin: potential for false‐
positive diagnoses of type 2 diabetes using new diagnostic
criteria. JAMA. 1999;281(13):1203–1210.
14. Engelgau MM et al. Comparison of fasting and 2‐hour glu­
cose and HbA1c levels for diagnosing diabetes. Diagnostic
criteria and performance revisited. Diabetes Care. 1997;20(5):
785–791.
15. Zhang R et  al. The association of retinopathy and plasma
glucose and HbA1c: a validation of diabetes diagnostic crite­
ria in a Chinese population. J Diabetes Res. 2016;2016:
4034129.
16. Centers for Disease Control and Prevention. National
Diabetes Statistics Report, 2020. 2020.
17. Abdul‐Ghani MA, Tripathy D and DeFronzo RA.
Contributions of beta‐cell dysfunction and insulin resist­
ance to the pathogenesis of impaired glucose tolerance and
impaired fasting glucose. Diabetes Care. 2006;29(5):
1130–1139.
18. Knowler WC et al. Reduction in the incidence of type 2 dia­
betes with lifestyle intervention or metformin. N Engl J Med.
2002;346(6):393–403.
19. Talmud PJ et al. Sixty‐five common genetic variants and pre­
diction of type 2 diabetes. Diabetes. 2015;64(5):1830–1840.
20. Mahajan A et al. Fine‐mapping type 2 diabetes loci to single‐
variant resolution using high‐density imputation and islet‐­
specific epigenome maps. Nat Genet. 2018;50(11):1505–1513.
13
21. Grant SF et  al. Variant of transcription factor 7‐like
2 (TCF7L2) gene confers risk of type 2 diabetes. Nat Genet.
2006;38(3):320–323.
22. Schulz LO et al. Effects of traditional and western environ­
ments on prevalence of type 2 diabetes in Pima Indians in
Mexico and the U.S. Diabetes Care. 2006;29(8):1866–1871.
23. Bergman RN, Phillips LS and Cobelli C. Physiologic evalua­
tion of factors controlling glucose tolerance in man: meas­
urement of insulin sensitivity and beta‐cell glucose sensitivity
from the response to intravenous glucose. J Clin Invest.
1981;68(6):1456–1467.
24. Shah P et al. Lack of suppression of glucagon contributes to
postprandial hyperglycemia in subjects with type 2 diabetes
mellitus. J Clin Endocrinol Metab. 2000;85(11):4053–4059.
25. Bock G et al. Pathogenesis of pre‐diabetes: mechanisms of
fasting and postprandial hyperglycemia in people with
impaired fasting glucose and/or impaired glucose tolerance.
Diabetes. 2006;55(12):3536–3549.
26. Sathananthan A et al. A concerted decline in insulin secre­
tion and action occurs across the spectrum of fasting and
postchallenge glucose concentrations. Clin Endocrinol (Oxf).
2012;76(2):212–219.
27. Basu A et al. Impaired basal glucose effectiveness in NIDDM:
contribution of defects in glucose disappearance and pro­
duction, measured using an optimized minimal model inde­
pendent protocol. Diabetes. 1997;46(3):421–432.
28. Camilleri M. Peripheral mechanisms in appetite regulation.
Gastroenterology. 2015;148(6):1219–1233.
29. Camilleri M and Shin A. Novel and validated approaches for
gastric emptying scintigraphy in patients with suspected
gastroparesis. Dig Dis Sci. 2013;58(7):1813–1815.
30. Butler AE et  al. Beta‐cell deficit and increased beta‐cell
apoptosis in humans with type 2 diabetes. Diabetes. 2003;
52(1):102–110.
31. Shah M et al. TCF7L2 Genotype and alpha‐Cell Function in
Humans Without Diabetes. Diabetes. 2016;65(2):371–380.
32. Adams JD and Vella A. What can diabetes – associated genetic
variation in TCF7L2 teach us about the pathogenesis of Type 2
Diabetes? Metab Syndr Relat Disord. 2018;16(8):383–389.
33. Laurenti MC et al. Diabetes‐associated genetic variation in
TCF7L2 alters pulsatile insulin secretion in humans. JCI
Insight. 2020;5(7).
34. Meier JJ, Veldhuis JD, and Butler PC. Pulsatile insulin secretion
dictates systemic insulin delivery by regulating hepatic insulin
extraction in humans. Diabetes. 2005;54(6):1649–1656.
35. Smushkin G et al. Diabetes‐associated common genetic varia­
tion and its association with GLP‐1 concentrations and response
to exogenous GLP‐1. Diabetes. 2012;61(5):1082–1089.
 14 Prediabetes and the Diagnosis of Diabetes
36. Adams JD et al. Fasting glucagon concentrations are associ­
ated with longitudinal decline of beta‐cell function in non‐
diabetic humans. Metabolism. 2020;105:154175.
37. Selvin E et al. Short‐term variability in measures of glycemia
and implications for the classification of diabetes. Arch Intern
Med. 2007;167(14):1545–1551.
38. Ko GT et al. The reproducibility and usefulness of the oral
glucose tolerance test in screening for diabetes and other
cardiovascular risk factors. Ann Clin Biochem. 1998;35
(Pt 1):62–67.
39. Meijnikman AS et al. Not performing an OGTT results in
significant underdiagnosis of (pre)diabetes in a high risk
adult Caucasian population. Int J Obes (Lond). 2017;41(11):
1615–1620.
40. Guo F, Moellering DR, and Garvey WT. Use of HbA1c for
diagnoses of diabetes and prediabetes: comparison with
diagnoses based on fasting and 2‐hr glucose values and
effects of gender, race, and age. Metab Syndr Relat Disord.
2014;12(5):258–268.
41. Lorenzo C et al. A1C between 5.7 and 6.4% as a marker for
identifying pre‐diabetes, insulin sensitivity and secretion,
and cardiovascular risk factors: the Insulin Resistance
Atherosclerosis Study (IRAS). Diabetes Care. 2010;33(9):
2104–2109.
42. Tuomilehto J. Point: a glucose tolerance test is important for
clinical practice. Diabetes Care. 2002;25(10):1880–1882.
43. Dinneen SF et al. Effects of changing diagnostic criteria on
the risk of developing diabetes. Diabetes Care. 1998;21(9):
1408–1413.
44. Udler MS et al. Genetic risk scores for diabetes diagnosis and
precision medicine. Endocr Rev. 2019;40(6):1500–1520.
45. Han SJ et  al. Incidence and predictors of type 2 diabetes
among Koreans: a 12‐year follow up of the Korean Genome
and Epidemiology Study. Diabetes Res Clin Pract. 2017;123:
173–180.
46. Zhang X et al. A1C level and future risk of diabetes: a sys­
tematic review. Diabetes Care. 2010;33(7):1665–1673.
47. Hulman A et al. Glucose patterns during an oral glucose tol­
erance test and associations with future diabetes, cardiovas­
cular disease and all‐cause mortality rate. Diabetologia.
2018;61(1):101–107.
48. Abdul‐Ghani MA et  al. One‐hour plasma glucose concen­
tration and the metabolic syndrome identify subjects at high
risk for future type 2 diabetes. Diabetes Care.
2008;31(8):1650–1655.
49. Abdul‐Ghani MA et al. Fasting versus postload plasma glucose
concentration and the risk for future type 2 diabetes: results
from the Botnia Study. Diabetes Care. 2009;32(2):281–286.
50. Peddinti G et  al. 1‐hour post‐OGTT glucose improves the
early prediction of type 2 diabetes by clinical and metabolic
markers. J Clin Endocrinol Metab. 2019;104(4):1131–1140.
51. Peddinti G et al. Early metabolic markers identify potential
targets for the prevention of type 2 diabetes. Diabetologia.
2017;60(9):1740–1750.
52. Priya M et  al. 1‐hour venous plasma glucose and incident
prediabetes and diabetes in Asian Indians. Diabetes Technol
Ther. 2013;15(6):497–502.
53. Wu S et al. Transition from pre‐diabetes to diabetes and pre­
dictors of risk in Mexican‐Americans. Diabetes Metab Syndr
Obes. 2017;10:491–503.
54. Abdul‐Ghani MA et al. What is the best predictor of future
type 2 diabetes? Diabetes Care. 2007;30(6):1544–1548.
55. Gabir MM et al. Plasma glucose and prediction of microvas­
cular disease and mortality: evaluation of 1997 American
Diabetes Association and 1999 World Health Organization
criteria for diagnosis of diabetes. Diabetes Care.
2000;23(8):1113–1118.
56. Diabetes Prevention Program Research Group. The preva­
lence of retinopathy in impaired glucose tolerance and
recent‐onset diabetes in the Diabetes Prevention Program.
Diabet Med. 2007;24(2):137–144.
57. Colagiuri S et al. Glycemic thresholds for diabetes‐specific
retinopathy: implications for diagnostic criteria for diabetes.
Diabetes Care. 2011;34(1):145–150.
58. Tapp RJ et al. Diagnostic thresholds for diabetes: the associa­
tion of retinopathy and albuminuria with glycaemia.
Diabetes Res Clin Pract. 2006;73(3):315–321.
59. Plantinga LC et al. Prevalence of chronic kidney disease in
US adults with undiagnosed diabetes or prediabetes. Clin J
Am Soc Nephrol. 2010;5(4):673–682.
60. Metcalf PA et al. Microalbuminuria in a middle‐aged work­
force. Effect of hyperglycemia and ethnicity. Diabetes Care.
1993;16(11):1485–1493.
61. Echouffo‐Tcheugui JB et al. Association between prediabe­
tes and risk of chronic kidney disease: a systematic review
and meta‐analysis. Diabet Med. 2016;33(12):1615–1624.
62. Dyck PJ et al. The prevalence by staged severity of various
types of diabetic neuropathy, retinopathy, and nephropathy
in a population‐based cohort: the Rochester Diabetic
Neuropathy Study. Neurology. 1993;43(4):817–824.
63. Putz Z et al. Noninvasive evaluation of neural impairment in
subjects with impaired glucose tolerance. Diabetes Care.
2009;32(1):181–183.
64. Grover SA et al. The prevalence of erectile dysfunction in the
primary care setting: importance of risk factors for diabetes
and vascular disease. Arch Intern Med. 2006;166(2):213–219.
 “Is Prediabetes a Risk Factor or Is It a Disease?”
65. Hoffman‐Snyder C et al. Value of the oral glucose tolerance
test in the evaluation of chronic idiopathic axonal polyneu­
ropathy. Arch Neurol. 2006;63(8):1075–1079.
66. Nathan DM et  al. Impaired fasting glucose and impaired
glucose tolerance: implications for care. Diabetes Care. 2007;
30(3):753–759.
67. Barr EL et al. Risk of cardiovascular and all‐cause mortality
in individuals with diabetes mellitus, impaired fasting glu­
cose, and impaired glucose tolerance: the Australian
Diabetes, Obesity, and Lifestyle Study (AusDiab). Circulation.
2007;116(2):151–157.
68. Huang Y et al. Association between prediabetes and risk of
cardiovascular disease and all cause mortality: systematic
review and meta‐analysis. BMJ. 2016;355:i5953.
69. Stacey RB et al. Prediabetes and the association with unrec­
ognized myocardial infarction in the multi‐ethnic study of
atherosclerosis. Am Heart J. 2015;170(5):923–928.
70. Brunner EJ et al. Relation between blood glucose and coro­
nary mortality over 33 years in the Whitehall Study. Diabetes
Care. 2006;29(1):26–31.
71. Emerging Risk Factors Collaboration. Diabetes mellitus,
fasting blood glucose concentration, and risk of vascular dis­
ease: a collaborative meta‐analysis of 102 prospective stud­
ies. Lancet. 2010;375(9733):2215–2222.
72. Rao Kondapally Seshasai S et  al. Diabetes mellitus, fasting
glucose, and risk of cause‐specific death. N Engl J Med.
2011;364(9):829–841.
73. Diabetes Prevention Program Research Group. The Diabetes
Prevention Program (DPP): description of lifestyle interven­
tion. Diabetes Care. 2002;25(12):2165–2171.
74. Hamman RF et al. Effect of weight loss with lifestyle interven­
tion on risk of diabetes. Diabetes Care. 2006;29(9):2102–2107.
75. Diabetes Prevention Program Research Group. Long‐term
effects of lifestyle intervention or metformin on diabetes
development and microvascular complications over 15‐year
follow‐up: the Diabetes Prevention Program Outcomes
Study. Lancet Diabetes Endocrinol. 2015;3(11):866–875.
76. Ratner R et al. Impact of intensive lifestyle and metformin
therapy on cardiovascular disease risk factors in the diabetes
prevention program. Diabetes Care. 2005;28(4):888–894.
77. Florez H et  al. Impact of lifestyle intervention and met­
formin on health‐related quality of life: the diabetes preven­
tion program randomized trial. J Gen Intern Med.
2012;27(12):1594–1601.
78. Diabetes Prevention Program Research Group. The 10‐year
cost‐effectiveness of lifestyle intervention or metformin for
diabetes prevention: an intent‐to‐treat analysis of the DPP/
DPPOS. Diabetes Care. 2012;35(4):723–730.
15
79. Pan XR et  al. Effects of diet and exercise in preventing
NIDDM in people with impaired glucose tolerance. The Da
Qing IGT and Diabetes Study. Diabetes Care. 1997;
20(4):537–544.
80. Gong Q et al. Morbidity and mortality after lifestyle inter­
vention for people with impaired glucose tolerance: 30‐year
results of the Da Qing Diabetes Prevention Outcome Study.
Lancet Diabetes Endocrinol. 2019;7(6):452–461.
81. Madsen KS et al. Metformin for prevention or delay of type 2
diabetes mellitus and its associated complications in persons
at increased risk for the development of type 2 diabetes mel­
litus. Cochrane Database Syst Rev. 2019;12:CD008558.
82. le Roux CW et  al. 3 years of liraglutide versus placebo for
type 2 diabetes risk reduction and weight management in
individuals with prediabetes: a randomised, double‐blind
trial. Lancet. 2017;389(10077):1399–1409.
83. ORIGIN Trial Investigators. Basal insulin and cardiovascu­
lar and other outcomes in dysglycemia. N Engl J Med.
2012;367(4):319–328.
84. DeFronzo RA et al. Pioglitazone for diabetes prevention in
impaired glucose tolerance. N Engl J Med. 2011;364(12):
1104–1115.
85. The DREAM (Diabetes REduction Assessment with ramipril
and rosiglitazone Medication) Trial Investigators. Effect of
rosiglitazone on the frequency of diabetes in patients with
impaired glucose tolerance or impaired fasting glucose: a ran­
domised controlled trial. Lancet. 2006;368(9541):1096–1105.
86. Knowler WC et al. Prevention of type 2 diabetes with trogl­
itazone in the Diabetes Prevention Program. Diabetes.
2005;54(4):1150–1156.
87. Torgerson JS et al. XENical in the prevention of diabetes in
obese subjects (XENDOS) study: a randomized study of orl­
istat as an adjunct to lifestyle changes for the prevention of
type 2 diabetes in obese patients. Diabetes Care. 2004;27(1):
155–161.
88. Chiasson JL et al. Acarbose for prevention of type 2 diabetes
mellitus: the STOP‐NIDDM randomised trial. Lancet.
2002;359(9323):2072–2077.
89. Group NS et al. Effect of valsartan on the incidence of diabe­
tes and cardiovascular events. N Engl J Med. 2010;362(16):
1477–1490.
90. Eriksson JG et  al. Long‐term beneficial effects of glipizide
treatment on glucose tolerance in subjects with impaired
glucose tolerance. J Intern Med. 2006;259(6):553–560.
91. Kanaya AM et al. Glycemic effects of postmenopausal hor­
mone therapy: the Heart and Estrogen/progestin Replacement
Study. A randomized, double‐blind, placebo‐controlled trial.
Ann Intern Med. 2003;138(1):1–9.
 16 Prediabetes and the Diagnosis of Diabetes
92. Margolis KL et al. Effect of oestrogen plus progestin on the
incidence of diabetes in postmenopausal women: results
from the Women’s Health Initiative Hormone Trial. Dia­
betologia. 2004;47(7):1175–1187.
93. Sawicki PT and Kaiser T. Response to Chiasson et  al.:
Acarbose for the prevention of Type 2 diabetes, hypertension
and cardiovascular disease in subjects with impaired glucose
tolerance: facts and interpretations concerning the critical
analysis of the STOP‐NIDDM Trial data. Diabetologia.
2004;47(6):976–977.
94. Lindblad U et  al. Can sulphonylurea addition to life­style
changes help to delay diabetes development in subjects
with impaired fasting glucose? The Nepi ANti­
diabetes StudY (NANSY). Diabetes Obes Metab. 2011;13(2):
185–188.
95. Pittas AG et al. Vitamin D supplementation and prevention
of type 2 diabetes. N Engl J Med. 2019;381(6):520–530.
96. NAVIGATOR Study Group et al. Effect of nateglinide on the
incidence of diabetes and cardiovascular events. N Engl J
Med. 2010;362(16):1463–1476.
97. The DREAM Trial Investigators. Effect of ramipril on the
incidence of diabetes. N Engl J Med. 2006;355(15):
1551–1562.
98. Carlsson LM et al. Bariatric surgery and prevention of type 2
diabetes in Swedish obese subjects. N Engl J Med.
2012;367(8):695–704.
99. Pories WJ et  al. Is type II diabetes mellitus (NIDDM) a
surgical disease? Ann Surg. 1992;215(6):633–642; discus­
sion 643.
100. Smith AD et al. Physical activity and incident type 2 diabe­
tes mellitus: a systematic review and dose‐response meta‐
analysis of prospective cohort studies. Diabetologia. 2016;
59(12):2527–2545.
101. Grontved A et  al. A prospective study of weight training
and risk of type 2 diabetes mellitus in men. Arch Intern
Med. 2012;172(17):1306–1312.
102. Salas‐Salvado J et  al. Prevention of diabetes with Medi­
terranean diets: a subgroup analysis of a randomized trial.
Ann Intern Med. 2014;160(1):1–10.
103. American Diabetes Association. 3. Prevention or delay of
type 2 diabetes: standards of medical care in diabe­
tes – 2020. Diabetes Care. 2020;43(Suppl 1):S32–S36.
104. Hu W et al. Association of elevated glycosylated hemoglobin
A1c with hyperfiltration in a middle‐aged and elderly Chinese
population with prediabetes or newly diagnosed diabetes: a
cross‐sectional study. BMC Endocr Disord. 2015;15:47.
105. Kannan MA et  al. Prevalence of neuropathy in patients
with impaired glucose tolerance using various electrophys­
iological tests. Neurol India. 2014;62(6):656–661.
106. Wu JS et  al. Epidemiological evidence of altered cardiac
autonomic function in subjects with impaired glucose tol­
erance but not isolated impaired fasting glucose. J Clin
Endocrinol Metab. 2007;92(10):3885–3889.
107. Holman RR et al. Effects of acarbose on cardiovascular and
diabetes outcomes in patients with coronary heart disease and
impaired glucose tolerance (ACE): a randomised, double‐
blind, placebo‐controlled trial. Lancet Diabetes Endocrinol.
2017;5(11):877–886.
 2 Early Diagnosis of Type 1
Diabetes – Useful or a Pyrrhic Victory?
Paolo Pozzilli and Silvia Pieralice
Department of Endocrinology & Diabetes, Campus Bio-­Medico University of Rome, Rome, Italy
LE A R NI NG P OI NT S
●● Making an early diagnosis of type 1 diabetes can
permit intervention with agents that could modify the
course of disease.
●● Numerous clinical trials have attempted to change the
course of disease or prevent its onset. Of these, early
immune intervention holds the most promise.
Introduction
Type 1 diabetes (T1D) is one of the most widespread
chronic disease occurring in children and young adults. In
2001, people with diabetes (all types included) were
177  million worldwide, in 2010 285  million, in 2019
approximately 463  million and it is predicted that some
700  million people worldwide will live with diabetes in
2045 [1].
As the prevalence of diabetes continues to rise world-
wide, disease-­related morbidity and mortality are emerg-
ing as major healthcare problems. Epidemiologic data
show that diabetes-­related long-­term complications begin
early in the natural history of the disease  [2]–[4]. These
findings indicate that early identification and management
of individuals at increased risk of T1D has the potential to
reduce both the clinical onset of the disease and its related
complications.
The global incidence of T1D in children and adolescents
is rising with an estimated overall annual increase of
approximately 3%. These recent epidemiologic trends in
Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
T1D have been shown in countries having both high and
low prevalence figures, with an indication of a steeper
increase in some of the low-­prevalence countries. T1D
accounts for about 10% of all cases of diabetes, occurs most
commonly in people of European descent and affects
2  million people in Europe and North America  [1]. The
lowest incidence has been found in Asia and Oceania, the
highest in Europe.
As far as its pathogenesis is concerned, T1D results from
the autoimmune destruction of pancreatic insulin-­
secreting β-­cells. Genetic, metabolic, and environmental
factors act together to precipitate the onset of the disease.
The excess mortality associated with diabetes-­related com-
plications and the increasing prevalence of the disease
among the youth, emphasize the importance of novel ther-
apeutic strategies to prevent or slow down the autoimmune
process.
Pathogenesis of T1D: An Update in View
of Defining Preventive Tools
There are three main categories of factors involved in the
pathogenesis of T1D. These are genetic, immunological,
and environmental factors (Figure 2.1).
Like other organ-­specific autoimmune diseases, T1D
shows specific human leukocyte antigen (HLA) associa-
tions. The HLA complex on chromosome 6 comprises the
first gene shown to be associated with the disease and it is
considered to contribute to almost half of the familial basis
of T1D. Two combinations of HLA haplotypes are of par-
ticular importance. They are DR4-­DQ8 and DR3-­DQ2,
17
 18 Prediabetes and the Diagnosis of Diabetes
Immune Environmental
dysregulation triggers and
regulators
IAA
GADA ICA512A ICA
Interactions
Loss of first phase
between
β-cell mass
genes
imparting Variable insulitis
susceptibility β-cell sensitivity
and resistance to injury
Pre-diabetes
Time
FIG 2.1  Pathogenesis and natural history of Type 1 diabetes. Atkinson MA and Eisenbarth GS Type 1 diabetes: new perspectives on
disease pathogenesis and treatment. Lancet 2001;358(9277):221–229.
which are present in 90% of children with T1D [5]. A third
haplotype, DR15-­DQ6, is found in less than 1% of children
with T1D, compared to more than 20% of the general pop-
ulation, and it is considered to be protective. The genotype
combining the two susceptibility haplotypes (DR4-­DQ8/
DR3-­DQ2) contributes to the greatest risk of the disease
and it appears frequently in children with an earlier onset.
First-­degree relatives of these children are themselves at
greater risk of T1D compared to those of children in whom
the disease develops later.
Candidate gene studies also identified the insulin gene
on chromosome 11 as another important genetic suscepti-
bility factor, contributing 10% of the genetic susceptibility
to T1D [6]. Similarly, an allele of the gene acting as a nega-
tive regulator of T-­cell activation, cytotoxic T lymphocyte
antigen 4 (CTLA-­4), found on chromosome 2q33, is con-
sidered to be another susceptibility gene for T1D  [7]. A
variant of PTPN22, the gene encoding Lymphoid
Phosphatase (LYP), which is a suppressor of T-­cell activa-
tion, has been deemed as another susceptibility gene  [8].
Similarly, variation in IL2RA which encodes the α-­chain of
the IL-­2 receptor is also associated with T1D  [9]. The
observation that these susceptibility genes for T1D all play
important roles in antigen presentation to T-­ cells,
insulin response
(IVGTT)
Glucose intolerance
Absence of C-peptide
Overt diabetes
e­ mphasizes the potential importance of current therapeu-
tic strategies targeting this interaction [10].
Genetic studies have highlighted the importance of
large, well-­characterized populations in the identification
of susceptibility genes for T1D. Recruitment of increas-
ingly large populations of patients with T1D and their
families is required to provide statistically powerful cohorts
in which to identify other disease-­associated genes. Some
genes have a relatively minor individual impact on suscep-
tibility to disease but could nevertheless provide more clues
to future preventive therapies.
The presence of autoantibodies to β-­cells is the hallmark
of T1D (Figure  2.1). Abnormal activation of the T-­cell-­
mediated immune system in susceptible individuals leads
to an inflammatory response within the islets as well as to a
humoral response with production of antibodies to β-­cell
antigens. Islet-­cell antibodies (ICA) were the first ones
described, followed by more specific autoantibodies to
insulin (IAA), glutamic acid decarboxylase (GAD), the pro-
tein tyrosine phosphatase (IA-­2), and most recently Zinc
Transporter 8 (ZnT8A), all of which can be easily detected
by sensitive radioimmunoassay to identify subjects at risk
of developing T1D [11]. These autoantibodies are common
in both childhood and adult onset T1D, with many subjects
 Early Diagnosis of Type 1 Diabetes – Useful or a Pyrrhic Victory?
being positive for multiple autoantibodies. The type of
immune response is age-­dependent, but seroconversion to
multiple autoantibody positivity usually occurs tightly clus-
tered in time and is associated with genetic risk.
The presence of one or more type of antibodies can pre-
cede the clinical onset of T1D by years or even decades.
These autoantibodies are usually persistent, although a
small group of individuals may revert to being seronegative
without progressing to clinical diabetes. The presence and
persistence of positivity to multiple antibodies increases
the likelihood of progression to clinical disease.
On that note, recent evidence shows that antibodies spe-
cific to oxidative post-­ translationally modified insulin
(oxPTM-­INS) are present in the majority of newly diag-
nosed individuals with T1D being significantly more abun-
dant than autoantibodies to native insulin (NT-­INS) [12].
Furthermore, subsequent analysis found that oxPTM-­INS
auto-­reactivity is present before diabetes diagnosis in over
90% of individuals, suggesting a potential role for oxPTM-­
INS-­Ab as a predictive biomarker of T1D [13].
The progressive reduction of insulin-­secretory reserve
leads primarily to the loss of the first phase insulin secre-
tion in response to an intravenous glucose tolerance test,
and therefore to a state of absolute insulin deficiency.
Regarding the role of environmental factors, it should be
underlined that the increase in incidence of T1D is too
rapid to be caused by alterations in the genetic background
and is likely to be the result of environmental changes.
What Are the Environmental Factors Triggering
Type 1 Diabetes?
Certain viral infections may play a role in the pathogenesis
of human T1D. Congenital rubella is the classical example
of virus-­induced diabetes in human beings, but effective
immunization programs have eliminated congenital rubella
in most Western countries. Currently, the main candidate
for a viral trigger of human diabetes is members of the
group of Enterovirus  [14]. They are small non-­enveloped
RNA viruses, which belong to the Picornavirus family.
They consist of more than 60 different serotypes, with the
Polioviruses being their best-­ known representatives.
Enterovirus infections are frequent among children and
adolescents causing aseptic meningitis, myocarditis, rash,
hand-­food-­and-­mouth disease, paralysis, respiratory infec-
19
tions, and severe systemic infections in newborn infants.
Most infections, however, are subclinical or manifest with
mild respiratory symptoms. The primary replication of the
virus occurs in the lymphoid tissues of the pharynx and
small intestine, and during the following viremic phase the
virus can spread to various organs including the β-­cells.
Theoretically, Enterovirus could cause β-­cell damage by
two main mechanisms. They may infect β-­cells and destroy
them directly or they may induce an autoimmune response
against β-­cells. Direct virus-­induced damage has been sup-
ported by studies showing that Enterovirus are present in β-­
cells in patients who have died from severe systemic
Enterovirus infection and that the islet-­cells of these patients
are damaged. Enterovirus can also infect and damage β-­cells
in vitro and induce the expression of interferon-­alpha and
HLA-­class I molecules in β-­cells thus mimicking the situa-
tion observed in the pancreas of patients affected by T1D.
The first reports connecting Enterovirus infections to T1D
were published more than 30 years ago, showing that the
seasonal variation in the onset of T1D follows that of
Enterovirus infections. At the same time antibodies against
Coxsackievirus B serotypes were found to be more frequent
in patients with newly diagnosed T1D than in control sub-
jects [14]. Enterovirus have also been isolated from patients
with newly diagnosed T1D. In one case report Coxsackievirus
B4 was isolated from the pancreas of a child who had died
from diabetic ketoacidosis, and this virus caused diabetes
when transferred to a susceptible mouse strain. The β-­cells
of diabetic patients also express interferon-­alpha, a cytokine
that is induced during viral infections, suggesting the pres-
ence of some virus in the β-­cells. Prospective studies are par-
ticularly valuable in the evaluation of viral triggers because
they cover all stages of the β-­cell damaging process.
Enterovirus are not the only viruses that have been con-
nected to the pathogenesis of T1D. Mumps, measles, cyto-
megalovirus, and retroviruses also have been found to be
associated with T1D, but the evidence is less convincing
than that for Enterovirus.
The Role of Cow’s Milk
There is evidence that cow’s milk proteins can act as trig-
gers for the autoimmune process of β-­cell destruction
based on studies indicating bottle feeding as a triggering
factor for an autoimmune response to β-­cell [15].
 20 Prediabetes and the Diagnosis of Diabetes
There are several arguments for the milk hypothesis in
T1D, including the following:
●● Epidemiological studies show increased risk for T1D if
the breast-­feeding period is short and cow’s milk is
introduced before 3–4 months of age.
●● Skim milk powder can be "diabetogenic" in diabetes-­
prone BB rats.
●● Patients with T1D have increased levels of antibodies
against cow’s milk constituents.
●● Milk albumin and β-­casein have some structural similarity
to the islet autoantigen ICA69 and GLUT-­2, respectively.
A number of hypotheses have been postulated to explain
the pathogenic role of cow’s milk. One of the most convinc-
ing ones is that immature gut mucosa allows the passage of
high molecular weight, potentially antigenic proteins
which share some molecular mimicry with pancreatic β-­
cells. Among diabetogenic proteins in cow’s milk, β-­casein,
β-­lactoglobulin, and albumin have been implicated as
sources of potential antigens.
Casein represents the major protein in cow’s milk.
Human and bovine β-­casein are approximately 70% homol-
ogous and 30% identical. There are several reasons why it is
thought that β-­casein is a good candidate to explain the
observed association between cow’s milk consumption and
T1D: (1) it has several structural differences from the
homologous human protein; (2) casein is probably the milk
fraction promoting diabetes in the NOD mouse, since a
protein-­free diet prevents the disease while a diet contain-
ing casein as the sole source of protein produces diabetes in
the same animals; (3) several sequence homologies exist
between bovine β-­casein and β-­cell autoantigens; (4) spe-
cific cellular and humoral immune responses toward
bovine β-­casein are detectable in most T1D patients at the
time of diagnosis, highly suggestive that this protein may
participate in the immune events triggering the disease; (5)
casein hydrolysate was shown to be non-­diabetogenic in
the BB rat and NOD mouse models, therefore it was
thought that this dietary intervention might be beneficial in
humans as well for disease prevention.
The rationale behind the use of cow’s milk hydrolysate
for primary prevention of T1D is based on several epide-
miological and in vitro studies, indicating that intact cow’s
milk, if given before three months of age, may induce an
immune response towards β-­cells. The TRIGR trial (see
section on primary prevention of T1D) investigated
whether the administration of a cow’s milk hydrolysate
could prevent or delay the onset of T1D.
The Role of Vitamin D Deficiency
Several epidemiological studies have described an intrigu-
ing correlation between geographical latitude and the inci-
dence of T1D and an inverse correlation between monthly
hours of sunshine and the incidence of diabetes. A seasonal
pattern of disease onset has also been described for T1D,
once again suggesting an inverse correlation between sun-
light and the disease [16]. Vitamin D is an obvious candi-
date as a mediator of this sunshine effect.
Dietary vitamin D supplementation is often recom-
mended in pregnant women and in children to prevent
vitamin D deficiency. Cod liver oil taken during the first
year of life reportedly reduced the risk of childhood-­onset
T1D and a multicenter case-­control study also showed an
association between vitamin D supplementation in infancy
and a decreased risk of T1D. Two further meta-­analyses of
retrospective studies showed that the risk of T1D was lower
in children who were supplemented with calcitriol com-
pared with those who were not supplemented  [17].
Nonetheless, it remains to be determined whether these
observations are the result of supplementation of vitamin
D to supraphysiological levels or are simply the result of the
prevention of vitamin D deficiency. On that note, other
clinical studies reported no effect of vitamin D supplemen-
tation on β-­cell function in recent-­onset T1D  [18]. In
­summary, despite continuing interest in vitamin D supple-
mentation as a potential intervention to prevent islet auto-
immunity and T1D, there is still little supporting evidence
from prospective birth cohort studies.
Prediction of T1D as the Basis for Disease
Prevention
There are different approaches for the identification of indi-
viduals at risk for T1D. These approaches are based on family
history of T1D, genetic disease markers, autoimmune mark-
ers, or metabolic markers of T1D (Figure 2.2). These alterna-
tives may also be combined in various ways to improve the
predictive characteristics of the screening s­trategy. The
 Early Diagnosis of Type 1 Diabetes – Useful or a Pyrrhic Victory?
years
5 10
Secondary prevention
Genetic analysis
Primary prevention
FIG 2.2  Strategies to preserve β-­cell mass in T1D. Modified from Reimann M, Bonifacio E, Solimena M et al. An update on
preventive and regenerative therapies in diabetes mellitus. Pharmacol Ther 121(3): 317–331, 2009.
importance of understanding the natural history of immune
mediated pre-­diabetes lies in the development of prevention
strategies. Several randomized clinical intervention trials
have been concluded and the next generation of such trials
will rely upon improved and simplified identification of indi-
viduals who are at high risk of progression to T1D. This is
essential to ensure that trials have sufficient statistical power
to detect a given effect of the intervention within the time
available for the study. Such understanding is also needed to
avoid exposing those who will not develop T1D to the risk of
adverse effects of the intervention.
Prevention of T1D: Current Status
Although the process by which pancreatic β-­cells are
destroyed is not well understood, several risk factors and
immune-­related markers are known to accurately identify
first-­degree relatives of patients with T1D who may develop
the disease. Since we now can predict the development of
T1D, investigators have begun to explore the use of inter-
vention therapy to halt or even prevent β-­cell destruction
in such individuals. The autoimmune pathogenesis of T1D
determines the efforts to prevent it. Susceptible individuals
are identified by searching for evidence of autoimmune
21
Diabetes 100
Beta cell mass (%)
80
60
40
20
15 20
Tertiary prevention
activity directed against β-­cells. While direct evaluation of
T-­cell activity might be preferable, antibody determina-
tions are generally used for screening because these assays
are more robust. Antibody titers are often used in combi-
nation with an assessment of the genetic susceptibility, pri-
marily evaluated by HLA typing. In the near future, testing
for oxPTM-­INS-­Ab may help to identify children progress-
ing to overt diabetes.
Interventions are generally designed to delay or prevent
T1D by impacting some phases of the immune pathogen-
esis of the disease. As discussed below, current trials are
attempting to modify the course of disease progress at
many points along the presumed pathogenic pathway.
Most prevention trials include only relatives of T1D
patients, a group in which risk prediction strategies are
most established. Trials in genetically at-­risk infants evalu-
ate whether avoiding one of the putative environmental
triggers for T1D can delay or prevent its onset.
Primary Prevention
Primary prevention identifies and attempts to protect indi-
viduals at risk from developing T1D. It can therefore reduce
both the need for diabetes care and the need to treat
diabetes-­related complications (Figure 2.2).
 22 Prediabetes and the Diagnosis of Diabetes

T1D is relatively easy to prevent in animal models of the TABLE 2.1  Prevention in T1D
disease and an array of therapies is effective. However, the
Study
mechanism of prevention is usually poorly defined, and
there is a lack of surrogate assays of the immune response to Primary prevention DAISY:
define which therapies are likely to prevent diabetes in cow’s milk intake and IA development
humans. Inability to define surrogate assays probably results TRIGR:
from a fine balance of the immune system, so that even with Casein hydrolysate vs cow’s milk formula
inbred strains of animals, only a subset progress to diabetes, Babydiet:
Delayed introduction of dietary gluten
and thus relatively small changes in immune function may
TEDDY:
prevent disease. These observations have led to the hypoth-
Timed introduction of gluten-­containing
esis that identifying children at a very high genetic risk for cereals
diabetes, prior to development of measurable β-­cell autoim- DIPP:
munity, and treating them at that point may be a more effec- Intranasal insulin
tive means of diabetes prevention. Studies for the primary Secondary ENDIT: Nicotinamide
prevention of T1D, i.e., prior to the expression of islet prevention DPT-­1: Insulin/oral insulin
autoantibodies, are currently being designed and imple- DIPP: Intranasal insulin
INIT: Intranasal Insulin Trial
mented. These studies target young children at a very high
Tertiary prevention Cyclosporine
genetic risk for T1D and propose treatments that are very
Nicotinamide
safe. These studies require large-­scale screening to identify Vitamin D
high-­risk subjects and a follow-­up over a long period of Insulin
time to observe the outcome of anti-­islet autoimmunity as a Anti-­CD3
surrogate marker for the disease and onset of hyperglycemia Anti-­CD20
Anti TNF alfa
as the main end point (Table 2.1).
CTLA4-­Ig
As mentioned above, various nutritional components GAD-­ Alum
have been suggested to modulate the risk of T1D. Among Anti IL-­1
them, several epidemiological and in vitro studies indicat- Anti IL-­6
ing that intact cow’s milk, if given before 3 months of age, Anti IL-­8
CXCR1/2 inhibitor
may induce an immune response towards β-­cells.
In a prospective study called DAISY, which followed
children at increased T1D risk for IA and T1D develop-
ment, cow’s milk protein intake was associated with genetic disease susceptibility (diabetogenic HLA alleles and
increased IA risk in children with low/moderate risk first-­degree relatives with T1D) and whether the use of a
HLA-­DR genotypes [hazard ratio (HR): 1.41, 95% confi- cow’s milk hydrolysate could protect from the disease. The
dence interval (CI): 1.08–1.84], but not in children with recruitment was carried out over a 5-­year period in nine
high risk HLA-­DR genotypes. Furthermore, cow’s milk European countries, six major centers in the USA, 12 cent-
protein intake was associated with progression to T1D ers in Canada, and three centers in Australia. Due to statis-
(HR: 1.59, CI: 1.13–2.25) in children with IA [15]. tical considerations, the frequency of the high-­risk HLA
TRIGR is a large international randomized double-­blind genotype, consent and drop-­out rates, the trial required
intervention study intended to provide information on the initial access to 8000 pregnancies, which ultimately yielded
incidence of predictive islet-­ cell autoantibodies vs the 5156 infants necessary for randomization. 1081 were rand-
actual occurrence of clinical diabetes in two treatment omized to be weaned to the extensively hydrolyzed casein
groups [19]. The aim of this trial was to investigate whether formula and 1078 to a conventional adapted cow’s milk for-
early exposure to complex dietary proteins of cow’s milk mula supplemented with 20% of the casein hydrolysate. The
could increase the risk of T1D in new-­born infants with participants were observed for a median of 11.5 years.
 Early Diagnosis of Type 1 Diabetes – Useful or a Pyrrhic Victory?
The strength of this trial was the larger number of par-
ticipants, which provides substantially greater statistical
power in a more heterogeneous study population compared
with previous studies and, therefore, provides a more defin-
itive answer to whether weaning to an extensively hydro-
lyzed formula might be protective of diabetes. This trial
showed that among infants at risk for T1D, weaning to a
hydrolyzed formula compared with a conventional formula
did not reduce the cumulative incidence of T1D. Thus, the
absolute risk of T1D was 8.4% among participants rand-
omized to the casein hydrolysate vs 7.6% among those ran-
domized to the conventional formula; the hazard ratio for
T1D adjusted for HLA risk group, duration of breastfeed-
ing, duration of study formula consumption, sex, and
region while treating study center as a random effect was
1.1 (95% CI, 0.8 to 1.5; P = 0.46). The median age at diagno-
sis of type 1 diabetes was similar in the 2 groups (6.0 years
[Q1–Q3, 3.1–8.9] vs 5.8 years [Q1–Q3, 2.6–9.1]) [19].
In conclusion, studies investigating cow’s milk as an
environmental factor show inconsistent results and do not
currently support the need to revise the dietary recommen-
dations for infants at risk for T1D.
Another outstanding question concerns the association
between timing of gluten introduction and IA develop-
ment. The BABYDIET study was an intervention study
aimed to determine whether primary intervention through
delayed introduction of dietary gluten was feasible and
could reduce the incidence of IA in high-­risk first degree
relatives of patients with T1D [20]. The study was based on
the premise that introduction of foods containing gluten or
cereal before the age of 3 months was associated with an
increased risk of IA in childhood. New-­born children were
eligible if they were younger than 3 months, were offspring
or siblings of patients with T1D and had HLA genotypes
that confer a high T1D risk. In this study, authors did not
find a benefit in delaying gluten exposure with respect to
autoimmunity associated with diabetes and celiac disease
at 3 years of age.
The follow-­up findings of the BABYDIET study con-
firm previous results, failing to demonstrate that an inter-
vention based on delayed gluten introduction will reduce
the risk of developing autoimmunity related to T1D [21].
Conversely, in The Environmental Determinants of
Diabetes in the Young (TEDDY) study, which prospectively
23
followed 8676 children with increased genetic risk of T1D
in the U.S., Finland, Germany, and Sweden, authors
observed that later introduction of gluten-­containing cere-
als was associated with increased risk of any IA using Cox
regression models  [22]. In this study, the HRs for every
1-­month delay in gluten introduction were 1.05 (95% CI
1.01, 1.10; P = 0.02) and 1.08 (95% CI 1.00, 1.16; P = 0.04),
respectively. The risk of IA associated with introducing
gluten before 4 months of age was lower (HR 0.68; 95% CI
0.47, 0.99), whereas the risk of IA associated with introduc-
ing it after 9 months of age was higher (HR 1.57; 95% CI
1.07, 2.31) than introduction between 4 and 9 months of
age. Interestingly, another sub-­ analysis performed on
TEDDY participants shows that administration of probiot-
ics during the first 27 days of life reduced the risk for a first-­
appearing β-­ cell autoantibody in children with the
HLA-­DQ2/8 genotype [23]. However, once again, authors
acknowledged the need to confirm these results in further
studies before any recommendation of dietary intake or
probiotics use is made.
Also noteworthy are some observational studies sug-
gesting a protective role of vitamin D and long-­chain n-­3
fatty acid against T1D by modulating the immune system.
Nonetheless, clinical trials have been inconclusive so
far [14, 24].
Finally, the Diabetes Prediction and Prevention Project
(DIPP Study) (Clinical trial NCT03269084; www.
clinicaltrials.gov) was a longitudinal study on T1D predic-
tion and prevention carried out in the university hospitals
of Turku, Tampere, and Oulu (Finland). The aim of the
study was to investigate longitudinally the dietary factors
in relation to the development of diabetic autoantibodies
and clinical T1D. The diet of children was followed up by a
structured questionnaire and by 3-­day dietary records at
various ages. A food frequency questionnaire was applied
for studying the dietary intake of pregnant and lactating
mothers.
The aims of this project were: (1) to identify infants at
increased genetic risk for T1D from the general popula-
tion at birth; (2) to monitor such children for the
appearance of diabetes-­ associated autoantibodies, to
identify those at high risk to develop clinical disease and
to characterize the natural course of T1D; (3) to identify
the environmental factors inducing the seroconversion
 24 Prediabetes and the Diagnosis of Diabetes
to autoantibody-­ positivity in children at increased
genetic risk; and (4) to evaluate whether it is possible to
delay or prevent progression to clinical T1D by daily
administration of intranasal insulin. Whereas points
1–4 have been fulfilled and useful information has been
obtained, the trial with intranasal insulin started soon
after detection of autoantibodies, and did not show any
beneficial effect of this treatment in preventing or delay-
ing the disease [25].
In conclusion, since the failure of ENDIT and DPT1 tri-
als (see secondary prevention) in preventing the onset of
T1D in subjects who are β-­ cell autoantibody positive,
interest has switched to prevention trials starting before
islet-­cell autoimmunity has developed. These primary pre-
vention trials of T1D offer an exciting view of how our
knowledge of the pathogenesis of this disease can lead to
the possibility of intervening at birth. There is still a long
way to go; however, the rationale is sound and the pros-
pects seem good.
Secondary Prevention
Secondary prevention of T1D aims to reduce the inci-
dence of the disease by stopping progression of β-­cell
destruction in individuals with signs of such a process.
This approach may represent a viable alternative to an
actual cure by blocking the autoimmune response while
β-­cell mass is still functional. For this reason, in the last
few decades, prevention studies have started to focus on
subjects at risk of developing the disease, in which the
autoimmune process has begun but the β-­cell mass is still
preserved. The most-­used approach at this stage of T1D is
as an antigen-­specific therapy, based on insulin adminis-
tration to establish tolerance. Generally, mechanistic
long-­term effects of these treatments rely on effector
T-­cell depletion with expansion of T regulatory cells (T-­
reg) (Figure 2.2).
Three large multicenter trials of diabetes prevention in
autoantibody-­positive relatives have been completed. The
European Nicotinamide Diabetes Intervention Trial
(ENDIT), the Diabetes Prevention Trial-­Type 1 (DPT-­1)
and the Type 1 Diabetes Prediction and Prevention Study
(DIPP). However, no significant benefits of these treat-
ments on the prevention of clinical onset of T1D were
found (Table 2.1).
European Nicotinamide Diabetes Intervention
Trial (ENDIT)
The ENDIT study, conducted predominantly in Europe,
examined whether nicotinamide could lead to a reduction
in the rate of progression to T1D in at risk relatives of T1D
probands. Over 40,000 first-­ degree relatives aged 5–40
years were screened in centers in Europe and North
America. The study was designed to recruit at least 422
subjects with ICA titers ≥ 20 JDF units to be randomized to
either a nicotinamide-­or a placebo-­treated group. With an
expected rate of progression to diabetes of 40% in the pla-
cebo arm, the proposed 5-­year observation period should
have allowed a 90% power to observe a 35% reduction in
the incidence of disease [26, 27].
Nicotinamide treatment at the doses used did not show
any significant effect on the primary outcome – progres-
sion to T1D. A total of 159 participants developed the dis-
ease within 5 years of randomization to treatment, 82
(30%) in the active treatment group and 77 (28%) in the
placebo group. The unadjusted Cox proportional hazard
estimate showed no difference between the placebo and
nicotinamide groups on an intention to treat basis. Nor was
any difference was found between groups after adjustment
for age at baseline, glucose concentrations at 2-­h glucose in
the OGTT, and number of islet autoantibodies. The pro-
portion of relatives who developed diabetes within 5 years
was almost identical in those treated with nicotinamide
and those treated with placebo, and there was no sugges-
tion of a treatment effect in any of the subgroups defined
by well established markers of additional risk.
A useful message of this trial has been that large-­scale
collaborations were essential to move things forward and
that the place for single-­center trials was limited.
DPT-­1 Trials
The Diabetes Prevention Trial -­Type 1 (DPT-­1) consisted
of two clinical trials that sought to delay or prevent T1D.
Nine medical centers and more than 350 clinics in the
United States and Canada took part in the two trials of the
DPT-­1 [28].
Individuals who were eligible for testing were identified as
follows: age 3 to 45 years, with a brother or sister, child or
parent with T1D; age 3 to 20 years, with a cousin, uncle or
aunt, nephew or niece, grandparent, or half-­sibling with T1D.
 Early Diagnosis of Type 1 Diabetes – Useful or a Pyrrhic Victory?
Those who met these criteria had ICA antibodies measured.
To be eligible, a subject had to be positive for ICAs.
Animal research and small studies indicated that small,
regular doses of insulin could prevent or delay T1D in sub-
jects at risk. One DPT-­1 trial tested whether low-­dose insu-
lin injections could prevent or delay the development of
T1D in people at high risk for developing T1D within 5
years.
First-­degree relatives, 3 to 45 years of age, and second-­
degree relatives, 3 to 20 years of age, of patients with T1D
were screened for islet-­cell antibodies. Those with an islet-­
cell antibody titer of 10 JDF units or higher were offered
staging evaluations.
Subjects identified as having a high risk of T1D were eli-
gible for random assignment to the experimental interven-
tion (parenteral insulin therapy) or to a control group that
underwent close observation.
The results demonstrated that insulin, in small doses,
can indeed be administered safely to persons who are at
risk for T1D. The increase in presumed and definite hypo-
glycemia among the subjects in the intervention group did
not adversely affect cognitive function.
In high-­risk relatives of patients with diabetes, the
insulin regimen did not delay or prevent the development
of T1D [27]. There are several potential explanations for
the lack of effect observed so far. One is that the interven-
tion took place too late in the disease process to slow
down the progression of disease. Studies conducted ear-
lier in the disease process may be more successful.
Moreover, the low dose insulin used in the trial may have
failed to achieve such an effect on β-­cells, but the dose
was limited by the risk of hypoglycemia. With a different
dosing scheme or a different regimen, insulin or insulin-­
like peptides might alter the course of development of
diabetes.
The other study was an oral insulin trial that sought to
prevent T1D in subjects with a moderate risk for develop-
ing diabetes [28].
First-­degree (ages 3–45 years) and second-­degree (ages
3–20 years) relatives of patients with T1D were screened
for ICAs. Those with ICA titer ≥10 JDF units were invited
to undergo staging evaluations.
Staging confirmed ICA positivity, measured insulin
autoantibody (IAA) status, assessed first-­ phase insulin
25
response (FPIR) to intravenous glucose, assessed oral glu-
cose tolerance (OGT), and determined presence or absence
of HLADQA1* 0102/DQB1*0602 (a protective haplotype
that excluded subjects from participation).
The study was a double-­masked, placebo-­controlled,
randomized clinical trial, in which participants were
assigned to receive capsules of either oral insulin, 7.5 mg of
recombinant human insulin crystals (Eli Lilly, Indianapolis,
IN), or matched placebo. Subjects consumed the capsule
(insulin or placebo) as a single daily dose before breakfast
each day, either by taking the capsule or, if the subject could
not swallow capsules, sprinkling its contents in juice or on
food.
In the primary analysis of relatives selected and rand-
omized in DPT-­1, oral insulin did not delay or prevent
development of diabetes. There was greater variability in
the IAA assay for values 39–79  nU/ml than for values
≥80 nU/ml, particularly in confirmation of a positive result
(98.7% overall confirmation for values ≥80  nU/ml com-
pared with 70.6% for values 39–79 nU/ml). This prompted
comparison of the rate of evolution of diabetes by entry
IAA level. The cohort with confirmed IAA ≥80 nU/ml (the
original entry IAA criterion) progressed to diabetes at a
faster rate than those subjects who did not have confirmed
IAA ≥80  nU/ml. In addition, those with confirmed IAA
≥ 80  nU/ml had other risk characteristics that suggested
more rapid evolution to diabetes, including younger age,
greater likelihood of having other antibodies, and greater
loss of β-­cell function [28].
The effect of intervention in each of these two sub-
groups was further evaluated.
The group with confirmed IAA ≥ 80 nU/ml showed a
beneficial effect of oral insulin, whereas the group who did
not have confirmed IAA ≥ 80 nU/ml showed a trend sug-
gesting a detrimental effect of oral insulin [28]. This group
also had a much lower overall rate of development of
diabetes.
Furthermore, the rate of progression seemed to increase
when oral insulin therapy was stopped, suggesting that the
therapy was probably effective but required ongoing
administration  [29]. This observation has prompted a
larger and justified follow-­up study with oral insulin to con-
firm these preliminary studies (Clinical trial NCT00419562;
www.clinicaltrials.gov).
 26 Prediabetes and the Diagnosis of Diabetes
In conclusion, neither low-­dose insulin injections in
subjects at high risk for developing T1D nor insulin cap-
sules taken orally by those at moderate risk for T1D were
successful at preventing or delaying the disease.
Type 1 Diabetes Prediction and Prevention
Study (DIPP)
The DIPP study was a randomized double-­ blind trial
investigating whether nasal insulin could reduce the inci-
dence of T1D in children with HLA genotypes and autoan-
tibodies conferring increased risk of disease  [25]. Daily
doses of intranasal insulin were administered; however,
after 1.8 years of observation, no differences were found in
the rate of progression to T1D.
Similar results have been obtained in the Intranasal
Insulin Trial (INIT I). This pilot study, based in Australia
and New Zealand, treated autoantibody-­positive subjects
with intranasal insulin, showing that intranasal insulin did
not prevent T1D onset. However, investigators found that
intranasal insulin administration induced immune changes
consistent with mucosal tolerance to insulin, justifying a
formal trial to determine if intranasal insulin is immuno-
therapeutic and retards progression to clinical diabe-
tes [30]. The INIT II study is still ongoing and will expand
the number of enrolled subjects. Thus, clinical trials evalu-
ating insulin administration for disease prevention have
demonstrated to date limited success in preventing the dis-
ease’s progression.
Tertiary Prevention
Tertiary prevention is aimed at delaying or preventing the
development of complications in subjects who already have
T1D. A landmark trial investigating patients with T1D
showed that good glycemic control [31] as well as low gly-
cemic variability [32] can reduce the likelihood of micro-
vascular complications leading to blindness or kidney
disease, but the trend toward a decrease in macrovascular
disease was not statistically significant. Diabetes education
of health care professionals and those affected by diabetes
plays a key role in the tertiary prevention of the disease.
Tertiary prevention is identified by the maintenance of the
residual β-­cell function present at disease onset and can be
realized by immune suppression or immune modulation
since the time of clinical diagnosis of T1D (Figure 2.2).
The best results in this field were obtained 30 years ago
with the use of cyclosporine, subsequently abandoned
because of transient benefits and undesired adverse
effects [33].
In the following years none of the several treatments
that have been proposed has obtained appreciable results
but for nicotinamide [34] (Table 2.1).
Over the last few decades, there has been growing inter-
est in vitamin D and its active metabolites in relation to
T1D and its immune pathogenesis. Vitamin D metabolites
have been shown to exert several immunomodulatory
effects and 1,25-­ dihydroxyvitamin D3 [1,25-­ (OH)2D3]
can either prevent or suppress autoimmune encephalomy-
elitis, inflammatory bowel disease, and other autoimmune
diseases. Based on this rationale, several interventional and
randomized controlled trials evaluated the role for vitamin
D in the treatment of T1D, with mixed results.
Previous data in humans have demonstrated that reduc-
tion in vitamin D supplementation is associated with a
higher risk of the disease, whereas its supplementation is
associated with a decreased frequency of T1D [35]. Other
authors observed a significant increase in T-­reg cells in
T1D patients supplemented with cholecalciferol at differ-
ent dosages  [36, 37]. Similar effects were reported by
Treiber et al., who administrated 70 IU/Kg of cholecalcif-
erol daily for 12  months, demonstrating not only the
enhancement of the T-­reg cells, but also an increased T-­reg
cell suppressive capacity among the supplemented
group [38].
Different trials showed a better preservation of residual
pancreatic β-­cell function in T1D patients supplemented
with different forms of vitamin D (cholecalciferol, alfacal-
cidiol or calcitriol), as proven by the significantly higher
level of fasting C-­peptide and/or lower needed daily insulin
dose observed in supplemented groups [37, 39]. However,
there are also studies that indicate no significant role for
vitamin D in the treatment of T1D. The IMDIAB XI trial
was an open-­label randomized trial designed to determine
whether supplementation with the active form of vitamin
D (calcitriol) at diagnosis of T1D could improve parame-
ters of glycemic control [40]. The secretion of C-­peptide as
an index of residual pancreatic β-­cell function was the pri-
mary end point, with HbA1c and insulin requirement as
secondary end points. The aim of this study was to
 Early Diagnosis of Type 1 Diabetes – Useful or a Pyrrhic Victory?
i­nvestigate whether supplementation with the active form
of vitamin D (calcitriol) in subjects with recent-­onset T1D,
protects residual pancreatic β-­cell function and improves
glycemic control (HbA1c and insulin requirement). In this
open-­label randomized trial, 70 subjects with recent-­onset
T1D, mean age 13.6 ± 7.6 years, were randomized to calci-
triol (0.25  microg on alternate days) or nicotinamide
(25 mg/kg daily) and were followed up for 1 year. Intensive
insulin therapy was implemented with three daily injec-
tions of regular insulin + NPH insulin at bedtime. No sig-
nificant differences were observed between calcitriol and
nicotinamide groups in respect of baseline/stimulated
C-­peptide or HbA1c 1 year after diagnosis, but the insulin
dose at 3 and 6  months was significantly reduced in the
calcitriol group. In conclusion, at the dosage used, calcitriol
had a modest effect on residual pancreatic β-­cell function
and only temporarily reduced the insulin dose [40]. These
results were confirmed later by the same group (IMDIAB
XIII Trial), who found no effect of calcitriol in protecting
β-­cell function in subjects with recent-­onset T1D and high
C-­ peptide at diagnosis followed-­ up for 2 years  [41].
Currently, a pilot study (POSEIDON) is investigating the
safety and efficacy of a regimen that combines Omega-­3
fatty acids and cholecalciferol in subjects at T1D onset
(NCT03406897). Thus, omega-­3 fatty acids have effects on
several immunotypes and are known to increase T-­reg cell
differentiation. The recruitment is still open.
Immune Intervention Therapies at Diagnosis of T1D
Other strategies for prevention of β-­cells damage with
immune intervention at onset of the disease are based on
immunotolerance (monoclonal antibodies, antigen-­based
treatments, pro-­inflammatory cytokine-­based treatments)
(Figure 2.3, Tables 2.2 and 2.3).
In the last decades, experience obtained with the use of
the antiCD3 monoclonal antibody in two studies (one in the
USA and the other in Europe) has revitalized the interest in
these types of intervention [42, 43]. The drug, a modified
form of anti-­CD3 antibody that minimizes first-­dose side
effects, was studied by comparing 12 subjects aged 7 to
30 who were treated with the antibody to an equal number
of patients in a control group who did not receive the drug.
One year after treatment with anti-­ CD3, the treated
patients produced more insulin and needed less insulin
27
therapy than the untreated patients. Those who received
the antibody treatment also had better HbA1c levels. The
anti-­CD3  was designed to act on the immune system’s
T-­cells in a more specific manner than previous attempts at
immune intervention in early diabetes.
Most recently, a phase 2, randomized, placebo-­
controlled, double-­blind trial showed that a single 14-­day
course of teplizumab (an Fc receptor–nonbinding anti-
­CD3 monoclonal antibody) significantly slowed progres-
sion to clinical T1D in high-­risk, nondiabetic relatives of
patients with diabetes who had at least two autoantibod-
ies. At the conclusion of the trial, the percentage of
diabetes-­free persons in the teplizumab group (57%) was
double that in the placebo group (28%). Median delay in
the diagnosis of diabetes was 2 years. However, the cohort
was relatively small, and the estimated power limited.
Furthermore, authors did not assess the potential devel-
opment of antibodies to teplizumab, which would be a
concern [44].
In the DEFEND-­1 and DEFEND-­2 phase III trials, anti-
­CD3 antibody Otelixizumab revealed a narrow therapeutic
window. At a low dose, there was no preservation of β-­
mass  [45]. This was indeed observed at 18  months, but
with significant adverse effects. Otelixizumab is a chimeric
monoclonal antibody that targets the CD3/T-­cell receptor,
which is genetically modified by removing the glycosyla-
tion site in the Fc domain, thus affecting binding of com-
plement or Fc receptors. This reduces secondary reactions
due to cytokine release. Otelixizumab downregulates path-
ogenic T-­cells and upregulates T-­reg cells, thus halting the
autoimmune process in T1D.
Only one trial has focused on β-­cell function in T1D
after treatment with Rituximab, which produces B cell
depletion [46]. Although T1D is a T-­cell mediated auto-
immune disorder, B lymphocytes play a pathogenetic
role by acting as antigen-­presenting cells (APC) and
modulating the islet environment. In recent-­onset T1D
subjects, administration of Rituximab reduced HbA1c
levels and exogenous insulin demand due to the preser-
vation of C-­peptide levels over 1 year. Interestingly, after
2 years’ follow-­up, Rituximab delayed the decrease in
C-­peptide levels but did not influence insulin dose, sug-
gesting B cell deletion is not sufficient to restore β-­cell
tolerance.
 Thymocyte Depletion + Immunomobilization
Delay & AbATE– ATG + G-CSF –2016
Teplizumab ATG + G-CSF + Rapamycin + Islet Transplant – 2018
Alemtuzumab + Anakinra + Etanercept + Liraglutide +
DEFEND 1 & DEFEND 2– Plerixafor – 2021
1-alpha\ ATG + G-CSF + Low dose IL-2 + Etanercept +
otelixizumab Exenatide – 2021
hydroxyvitamin D3 Cytokines
DILT1D–IL-2
Anti-IL-6 (Siltuximab) – 2017
1alpha, 25- REPAIR- Anti-IL-6 Receptor (Tocilizumab) – 2018
Autologous Canakinumab Autologous Anti-IL12/IL-23 (Ustekinumab) – 2017
dihydroxyvitamin D3 T1D–
Hematopoietic CD3+CD4+CD25+CD127– Anti-IL12/IL-23 (Ustekinumab) + INGAP – 2017
Stem Cell AIDA– Sitagliptin + Anti-IL21 (NNC0114–0006) + Liraglutide - 2019
polyclonal Treg cell
Autologous Transplant Anakinra Lansoprazole Aldesleukin (IL-2) – 2016
therapy
Non-myeloablative ChAglyCD3 IL-2–2017
Hematopoietic Stem Low dose rhIL-2–2018
Anakinra
Cell Transplant with Low dose Proleukin (IL-2) – 2023
Protégé– Anti-TNFα (Golimumab) – 2019 + 2021
ATG + Teplizumab Autologous Autologous
Teplizumab Bone marrow T cells
Cyclophosphamide Umbilical Cord IMCY-0098–2018
Conditioning Stem Cells Blood Otelixizumab – 2018
Exenatide +/– START – ATG
Teplizumab – 2022
Daclizumab DIATOR– Wharton′s Tregs
Mycophenolate Atorvastatin Autologous Cord
jelly-derived ATG + G-CSF CLBSO3 (Autologous Ex Vivo Expanded Polyclonal
mofetil + Blood + oral
Mesenchymal Regulatory T-cells) – 2020
Daclizumab BCG Docosahexaenoic] CD4+CD127lo/-CD25+ Polyclonal Tregs + IL-2 – 2021
Stem cells G-CSF
Vaccination acid + Vitamin D Umbilical Cord Blood Tregs + Liraglitude – 2019
Umbilical Cord Blood Tregs – 2019
Tregs + anti-CD20 antibody – N/A
T cells Costimulation
+/– T&B T cell- Stem Hygiene Stem Increti Thymocytes +/– Abatacept (CTLA4-lg) – 2018
Stem Cells Vitamin D T cells IL-1 T cells UCB IL-1 Treg
Incretin cells APC cells Hypothesis cells n& H2+ Immunomobilizatio

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Ongoing

Global Immune T cell co- Innate Vitamin Innate

B cells Cytokines Suppression UCB DCs Vitamin D Treg +/– β cell Immunity Stem cells Immunity Innate Immunity
stimulation D
Alpha-1 Proteinase Inhibitor – 2017
PGD2 receptor antagonist MK-1092 – 2018
Autologous Immunoregulatory Densritic Cells – 2019
Stem Cells
Allogeneic mesenchymal Stem Cells – 2017
Aplha-1 Aplha-1 Mesenchymal Stem Cells – 2017
Autologous antitrypsin
Tolerogenic antitrypsin Allogeneic Adipose Mesenchymal Stem Cells with
ENBREL Bone Marrow Mononuclear Cells - 2018
(Etanercept) Dendritic Stem Cell Educator Therapy – 2019
cells TIDAL – Alefacept aLD-SCs / CB-MSCs + G-CSF – 2019
Cholecalciferol Chemotherapeutics
Oral IFN-α
Imatinib – 2019
Abatacept Neuropeptide
Rapamycin Substance P – 2017
+ IL-2 CXCR1/2 Inhibitor
Autologous Immunoablation (ATG + Cyclophosphamide) + Ladarixin – 2018
Umbilical Incretin
DF-IL2– Autologous Hematopoietic Stem Cell Transplant Sitagliptin – 2017
Cord Blood IL-2
Transfusion + G-CSF Exenatide – 2017
Saxagliptin – 2017
Albiglutide – 2017
Rituximab Umbilical Cord Mesenchymal Stromal Cell + Vildagliptin – 2017
Cholecalciferol Autologous Bone marrow Mononuclear Cell Liraglitude – 2018 + N/A
Stem Cell Transplant Rapamycin + Vildagliptin – 2018
Calcitriol 𝛃 cell Stress
Alpha1-Antitrypsin – 2017
Adipose-derived Insulin-secreting Mesenchymal Tauroursodeoxycholic Acid (TUDCA) – 2018
Cyclosporin + Verapamil – 2019
Methotrexate Stromal Cells + Bone marrow-derived
Alpha difluoromethylornithine (DFMO) – 2019
Hematopoietic Stem Cell Infusion
Vitamin D
Ergocalciferol – 2020
Stem Cell Educator Therapy
Environmental Factors
Prebiotic Fiber – 2017
Autologous Mesenchymal Stromal Cells Probiotics – 2018
GNbAC1 – 2018
Enterovirus vaccination – 2022
BCG vaccination – 2023

FIG 2.3  T1D clinical research over the past 10 year. Atkinson MA, Roep BO, Posgai A et al. The challenge of modulating β-­cell autoimmunity in type 1 diabetes.
Lancet Diabetes Endocrinol 2019;7(1):52–64.
 Early Diagnosis of Type 1 Diabetes – Useful or a Pyrrhic Victory?
TABLE 2.2  Current TrialNet studies. (Source: www.trialnet.net)
Study Status
Teplizumab Prevention Study Completed
Oral Insulin Prevention Study Completed
Hydroxychloroquine (HCQ) Currently Enrolling
ATG/GCSF New Onset Study Completed
Abatacept Prevention Study Currently Not Enrolling
Pathway to Prevention Currently Enrolling
The use of immunosuppressive drugs has also been pro-
posed as a possible approach to decelerate the evolution of
the disease.
In a multicenter, double-­ masked, randomized con-
trolled trial, Abatacept (CTLA4-­Ig) has been administered
on days 1, 14, 28, and then every 28 days through a 30 min
intravenous infusion at a dose of 10  mg/kg  [47]. Results
from this study showed that Abatacept was more efficient
compared to placebo in preserving the β-­cell mass, as evi-
denced by stimulated C-­peptide secretion. However, the
effect diminished with time; therefore, further investiga-
tion will be necessary in order to unravel whether the ben-
eficial effect persists after cessation of infusions. Indeed,
Abatacept is a potential candidate to be used in tertiary
prevention trials, and is a candidate for use in combination
therapies for recent-­onset T1D patients.
GAD65 (the 65 kDa isoform of glutamic acid decarbox-
ylase) is a human enzyme that has an important role in the
nervous system and in several nervous system diseases, e.g.
Parkinson’s disease and chronic pain.
GAD65 is also found in the insulin producing β-­cells of
the pancreas, although its function at this site is not yet
29
Description
This study tested the drug teplizumab to see if it could delay
or prevent progression of early stage T1D (stage 2) and
prevent clinical diagnosis (stage 3).
This study tested the drug oral insulin to see if it can delay or
prevent T1D (stage 1) from progressing to stage 2 and
ultimately prevent clinical diagnosis (stage 3).
This study tests the drug hydroxychloroquine (HCQ) to see if
it can delay or prevent early stage T1D (stage 1) from
progressing to abnormal glucose tolerance (stage 2) and
ultimately prevent clinical diagnosis (stage 3).
This study was designed to build on prior findings of a pilot
study suggesting thymoglobulin (ATG) combined with
pegylated granulocyte colony stimulating factor (GCSF)
preserved insulin production for more than 1 year after
treatment in people who had type 1 diabetes for 4 months
to 2 years.
This study tests the drug abatacept to see if it can delay or
prevent progression of early stage T1D (stage 1 or stage 2),
and ultimately prevent clinical diagnosis (stage 3).
This study screens and observes relatives of people with type
1 diabetes to learn more about how the disease occurs
fully established. It is however clear that GAD65 is one of
the most important targets when the immune system
attacks the insulin producing β-­cells in autoimmune diabe-
tes. Thus, treatment with rhGAD65 is thought to induce
tolerance to GAD65, thereby intervening in the autoim-
mune attack and preserving the capacity to produce insulin
in patients with autoimmune diabetes.
Although ongoing studies are investigating whether
rhGAD65 can preserve β-­cell function in recently diag-
nosed individuals with T1D, trials carried out to date do
not demonstrate a significant reduction in the loss of stim-
ulated C-­peptide in recently diagnosed children and young
adults (10–20 years) with T1D over a 15-­month period [48].
GAD65  was also targeted in NOD mice in order to
reduce the number of GAD65-­specific T effector cells [49],
achieving normoglycemia in 70% of NOD mice. Based on
this findings, antigen-­based immunotherapy therapy with
subcutaneous GAD-­alum have been tested to slow down
the course of loss of insulin in patients with recently diag-
nosed T1D. Recently, the prevention trial DIAPREV-­IT has
showed that GAD-­Alum as a subcutaneous prime and
boost injection was safe in prediabetic young children but
 30 Prediabetes and the Diagnosis of Diabetes

TABLE 2.3  Immunotherapy trials that are awaiting, or currently recruiting, participants. (Source: www.clinicaltrials.gov)

ClinicalTrials.gov
identifier Title Intervention Primary outcome

NCT02307695 The Effect of Saxagliptin on Glucose Fluctuation and Saxagliptin MAGE at 24 weeks
Immune Regulation in Patients with Type 1 Diabetes
NCT01559025 Evaluation of Vildagliptin (Galvus®) as add-­on to Vildagliptin MMTT C-­peptide at
Insulin in Residual β-­cell Function and Inflammatory 3, 6, 9 and
Markers in New-­onset Type 1 Diabetes Mellitus 12 month
NCT02442544 Effect of Prebiotic Fibre on Gut Microbiota, Intestinal Prebiotic 1:1 HbA1C at 3 months
Permeability and Glycaemic Control in Children with oligofructose:inulin
Type 1 Diabetes: A Pilot Randomized, Double Blind,
Placebo Controlled Study
NCT02820558 A Phase I Study of Safety and Pharmacological Activity Substance P Safety at
of Substance P in the Reversal of Recent-­Onset Type 1 20–27 days
Diabetes
NCT02505893 A Monocentric, Open-­label Pilot Study to Assess the ATG + pG-­CSF + rapamycin Safety and MMTT
Safety and Efficacy of Minimal Islet Transplantation in + human pancreatic islet C-­peptide at
Patients with New-­onset Type 1 Diabetes 12 months
NCT02940418 Use of Stem Cells in Diabetes Mellitus Type 1 AD-­MSCs + BM-­MNCs Safety at 6 months
NCT02293837 EXTEND Tocilizumab MMTT C-­peptide at
12 months
NCT02617654 A Randomized, Double-­blinded Placebo-­controlled, Liraglutide MMTT C-­peptide at
Paralleled Designed, Investigator Sponsored Study of 12 months
the Effect of the GLP-­1 Receptor Agonist Liraglutide
on Β-­cell Function in C-­peptide Positive Type 1
Diabetic Patients
NCT03170544 A Single Ascending Dose Clinical Trial to Study the MK-­1092 Safety and maximal
Safety, Tolerability, Pharmacokinetics, and glucose infusion
Pharmacodynamics of MK-­1092 in Healthy Subjects rate at 33 days
and in Subjects With Type 1 Diabetes Mellitus
NCT02814838 A Phase 2, Multicentre, Randomized, Double-­blind, Ladarixin MMTT C-­peptide at
Placebo-­controlled Study in Patients with New-­onset 13±1 weeks
Type 1 Diabetes
NCT02411253 DIABIL-­2 rhIL-­2 MMTT C-­peptide at
12 months
NCT02803892 MONORAPA Rapamycin /+ Vildagliptin MMTT C-­peptide at
4±1, 12±2 weeks
NCT02218619 Clinical Investigation of Efficacy of Tauroursodeoxycholic Acid MMTT C-­peptide at
Tauroursodeoxycholic Acid (TUDCA) to Enhance (TUDCA) 6, 12, and
Pancreatic Β-­cell Survival in Type 1 Diabetes by 18 months
Reducing Endoplasmic Reticulum Stress
NCT03272269 A Phase I Placebo-­controlled, Double-­blind, Dose IMCY-­0098 Safety at 24 weeks
Escalation Clinical Trial to Evaluate the Safety and
Immune Responses of Imcyse’s IMCY-­0098 in Patients
With Recent Onset Type 1 Diabetes
NCT03032354 Effect of Lactobacillus Rhamnosus GG and Probiotics MMTT C-­peptide at
Bifidobacterium Lactis BB 12 on Beta-­cell Function in 6 and 12 months
Children With Newly Diagnosed Type 1 Diabetes -­a
Randomized Controlled Trial
NCT02644759 Transplantation of Autologous Stem Cells for the aLD-­SCs/CB-­MSCs + G-­CSF Insulin requirement
Treatment of Type 1 Diabetes Mellitus at 1 month
 Early Diagnosis of Type 1 Diabetes – Useful or a Pyrrhic Victory? 31

TABLE 2.3  (Continued )

ClinicalTrials.gov
identifier Title Intervention Primary outcome

NCT02354911 A Randomized, Double-­Blind, Placebo-­Controlled, Immunoregulatory Dendritic MMTT C-­peptide at

Cross-­Over Study of the Safety and Efficacy of Cells 12 and 24 months
Autologous Immunoregulatory Dendritic Cells in
Patients With Type 1 Diabetes
NCT02624804 A Pilot Study of the Therapeutic Potential of Stem Cell Stem Cell Educator Therapy Safety at
Educator Therapy in Type 1 Diabetes 12 months
NCT02846545 T1GER Golimumab MMTT C-­peptide at
12 months
NCT03011021 Safety and Efficacy of Umbilical Cord Blood UCB -­Tregs + Liraglutide Safety at
Regulatory T Cells Plus Liraglutide on Autoimmune 24 months
Diabetes
NCT02932826 Safety Study and Therapeutic Effects of Umbilical UCB -­ Tregs Safety at
Cord Blood Treg Cells on Autoimmune Diabetes 24 months
NCT02384889 Targeting Polyamines Using DFMO in Persons with Difluoromethylornithine Safety with dose
Type 1 Diabetes: A Randomized, Double-­Masked, escalation at
Placebo-­Controlled Phase I Study to Evaluate the 6 months
Safety, Tolerability, and Initial Pharmacodynamics of
Multiple Ascending Doses
NCT03046927 Vitamin D and Residual Beta-­Cell Function in Type 1 Ergocalciferol MMTT C-­peptide at
Diabetes 12 months
NCT01773707 CTLA4-­Ig (Abatacept)for Prevention of Abnormal Abatacept Abnormal glucose
Glucose Tolerance and Diabetes in Relatives At -­Risk tolerance
for Type 1
NCT03298542 A Phase 1b Study to Evaluate SIMPONI (Golimumab) Golimumab Safety at 26, 52,
Therapy in Children, Adolescents and Young Adults and 78 weeks
With Pre-­Symptomatic Type 1 Diabetes
NCT03182426 Autologous Hematopoietic Stem Cell Mobilization Alemtuzumab + anakinra + Safety and MMTT
(Plerixafor) and Immunologic Reset in New Onset Type etanercept + liraglutide + C-­peptide at 3, 6,
1 Diabetes Mellitus plerixafor 9, 12, 18, and
24 months
NCT02772679 TILT Tregs + IL-­2 Safety and Treg
proportion at 3
years
NCT02804165 Gene-­virus Interactions Implicated in Type 1 Diabetes Enterovirus vaccination T1D diagnosis
NCT03243058 A Randomized, Double Blind, Phase I/II Trial of Proleukin (IL-­2) MMTT C-­peptide at
Low-­Dose Interlekin-­2 Immunotherapy in Established 12 months
Type 1 Diabetes
NCT02081326 Repeat BCG Vaccinations for the Treatment of BCG HbA1C at 1,2,3,4
Established Type 1 Diabetes and 5 years
EudraCT: TregVac2.0 Tregs + anti-­CD20 antibody (Phase II Trial
2014-­004319-­35 assessing expanded
polytTregs in
patients with recent
onset Type 1
diabetes mellitus)
EudraCT Incretin-­based therapy in non-­symptomatic, early Liraglutide MMTT C-­peptide at
2014-­004760-­37 diagnosed Type 1 Diabetics 3, 6, 9, and
12 months
 32 Prediabetes and the Diagnosis of Diabetes
did not affect progression to T1D [50]. The beneficial for
prevention of T1D of GAD-­alum should be tested in future
prevention studies.
Pro-­inflammatory cytokine-­based treatments have
proven to be safe and effective for treatment of various
autoimmune diseases. Thus, inhibition of expression of
those molecules can induce important changes in pancre-
atic β-­ cells induce important changes in pancreatic
β-­cells [51].
The aim of using the Anti-­Interleukin-­1 in newly diag-
nosed T1D subjects is to test the feasibility, safety/tolerabil-
ity and potential efficacy of anti-­IL-­1 therapy in maintaining
or enhancing β-­cell function in people with new onset
T1D. Anti-­IL-­1 administration for rheumatoid arthritis
has been proven to be well tolerated in patients  [52, 53].
IL-­1 is also involved in T1D progression by activating
T-­helper cells and improving the number of circulating
memory T-­cells [54]. The active substance is interleukin-­1
receptor antagonist, a blocker of an immune-­signal mole-
cule named interleukin-­ 1. Two randomized placebo-­
controlled trials aimed to assess whether canakinumab, a
human monoclonal anti-­interleukin-­1 antibody, or anak-
inra, a human interleukin-­1 receptor antagonist, improved
β-­cell function in recent-­onset T1D, but their effectiveness
was not demonstrated [54, 55].
More recently, the ongoing clinical trial EXTEND
(Clinical trial NCT02293837; www.clinicaltrials.gov) is
currently examining whether the blockade of IL-­6 signal-
ing through tocilizumab, an anti-­IL-­6 receptor antibody,
can induce a protection of β-­cell function in T1D patients
(ages 6 to 17 years) (Table 2.3).
Interleukin-­8 appears to be another important mediator
in the progression of T1D. Circulating levels of IL-­8 are
elevated in children with T1D compared to non-­diabetic
controls. Furthermore, levels of IL-­8 correlate with glyce-
mic control, higher level being associated to poorer glucose
control. As a result, the modulation or inhibition of IL8
activity may be a valid target for the development of novel
treatments aimed to control the progression of T1D.
A multicenter, randomized, double-­ blind, placebo-­
controlled phase 2 trial of CXCR1/2 IL-­ 8  inhibitor
(Ladarixin) has just presented its results at the American
Diabetes Association’s (ADA) 80th Scientific Sessions
(Clinical trial NCT02814838; www.clinicaltrials.gov). The
trial involved 76 patients with new-­onset T1D, randomly
(2:1) assigned to receive either Ladarixin treatment
(400 mg b.i.d. for 3 cycles of 14 days on/14 days off – treat-
ment group) or placebo (control group). Although results
indicated no statistically significant differences in stimu-
lated C-­peptide at weeks 13 and 26, investigators noted
76.6% of patients receiving Ladarixin had an HbA1c below
7% and a daily insulin requirement of less than 0.50 IU/kg
compared to just 45.8% of patients receiving placebo.
Furthermore, in a prespecified subgroup analysis of
patients with fasting C-­peptide below the median value of
the trial population at baseline, MMTT AUC of C-­peptide
trended at week 13 and reached statistical significance at
week 26.
Incretin-­based Therapies
Recent knowledge regarding the heterogeneity in the
extent of the β-­cell impairment and pancreatic lesions as
well as the differences in circulating T-­cell and autoanti-
body immune signatures underscore the potential applica-
tions for incretin treatments, which improve capacity for
insulin production by residual β-­cells and suppress gluca-
gon secretion, as well as the need for therapeutics to reduce
β-­cell stress co-­administered with immunomodulatory
therapy to reverse autoimmunity in symptomatic T1D.
Hence, therapies once considered only applicable to those
with T2D may be of potential benefit for those with T1D.
In this regard, ongoing T1D immunotherapy trials are
investigating the potential benefits on β-­cell function in
C-­peptide positive early diagnosed T1D patients.
In a post hoc analysis of pooled data from five rand-
omized, placebo-­controlled studies, the dipeptidyl pepti-
dase 4 (DPP-­ 4) inhibitor saxagliptin improved β-­cell
function as assessed by HOMA2 of β-­cell function and
post-­prandial C-­peptide from baseline in patients with
Latent Autoimmune Diabetes in the Adult (LADA)  [56].
Another small study found that sitagliptin, as an add-­on
treatment to insulin, had a beneficial effect on C-­peptide
decline compared with insulin alone. Moreover, a recent
trial evaluated the effect of saxagliptin in combination with
vitamin D3  in subjects with LADA with promising
results [57]. As far as T1D is concerned, the effect of saxa-
gliptin on immune regulation have been investigated in a
phase IV trial (NCT02307695). Similarly, a phase III
 Early Diagnosis of Type 1 Diabetes – Useful or a Pyrrhic Victory?
r­ andomized controlled trial is evaluating the action of vild-
agliptin in the prevention of progressive β-­cell dysfunction
in patients with newly diagnose of T1D (NCT01559025).
In another ongoing study, researchers are testing the effi-
cacy of 4 weeks rapamycin treatment and 4 weeks rapamy-
cin treatment plus 3 months vildagliptin treatment versus
placebo in increasing endogenous insulin production and
correcting glycemic lability (NCT02803892). Glucagon-­
like peptide (GLP-­1) analogues have been tested in large-­
scale clinical trial to prove their various benefits for β-­cell
and glucolipid metabolism in T2D and obesity patients.
This has led to concerns regarding the potential applica-
tions in T1D patients. A small phase II randomized con-
trolled trial (NCT02617654) is currently investigating the
effect of 52 weeks of treatment with liraglutide 1.8 mg/day,
compared to placebo, on stimulated C-­peptide concentra-
tions in patients with long-­standing type 1 diabetes and
residual insulin production (primary outcome: MMTT
C-­peptide at 12 months).
Other studies on preservation of β-­cell function using
incretin-­based therapies are currently active but not
recruiting participants (NCT02443155; NCT02127047).
Results from these studies are warranted to prove that
incretin-­based therapy might preserve C-­peptide secretion
(Table 2.3).
Conclusions
Today, one of the therapeutic goals in T1D is the preserva-
tion of the residual C-­peptide secretion that is detected in a
significant percentage of patients at diagnosis and which
potentially may influence the clinical course of the
disease.
Several studies have been demonstrated that residual
C-­peptide secretion, after T1D diagnosis, depends on
genetic factors, the patient’s age at the diabetes diagnosis,
the number of anti-­ islet antibodies, and the residual
C-­peptide secretion. In the same way, intensive insulin
therapy and immunomodulators drugs may be useful in
this direction.
The ultimate goal of any therapeutic intervention is to
prevent or reverse T1D by abrogation of pathogenic auto-
reactivity and by preservation or restoration of the β-­cell
mass and function to physiologically sufficient levels to
33
maintain stable glucose control. Early diagnosis of T1D is
crucial if we want to restore and to save β-­cell mass.
Different trials using antigen-­ specific or non-­ specific
interventions have shown some benefit in modulation of
the autoimmune process and in preventing the loss of insu-
lin secretion in the short term after early diagnosis of T1D.
Unfortunately, there are still limitations to current strat-
egies, including a lack of suitable markers to predict and
monitor the success of interventions, uncertainty about the
long-­term adverse effects or the duration of treatment
effect and the feasibility of restoration of β-­cell mass.
Moreover, we should remember that T1D is a heterogene-
ous disease with an age at onset spanning from childhood
to adult age.
Ideally, the interventions would be specific for T1D, free
of adverse effects, and effective prior to disease onset, with
long-­term and clinically meaningful improvements over
standard therapies. The success of these approaches will
eventually be evaluated by their impact on glycemic con-
trol as this is the definitive determinant of long-­term out-
come of the disease.
In conclusion we can affirm that early diagnosis of T1D
is very valuable and is not a Pyrrhic victory. Hopefully a
better comparison would be with the encounter of David
and Goliath, where early immune intervention dramati-
cally changes the clinical course of T1D.
Acknowledgments
We would like to thank Juvenile Diabetes Research
Foundation (JDRF), National Institute of Health (NIH)
Consortia, Centro Internazionale Studi Diabete (CISD),
Diabete e Metabolismo (DEM) Foundation and University
Campus Bio-­Medico that support clinical research on T1D
in our University Hospital.
References
  1. Nam Han Cho. IDF Diabetes Atlas, 9th edn. Brussels: 2019.
2017.
  2. de Ferranti SD, de Boer IH, Fonseca V et al. Type 1 diabetes
mellitus and cardiovascular disease: a scientific statement
from the American Heart Association Andamerican
Diabetes Association. Circulation, 2014;130:1110–1130.
 34 Prediabetes and the Diagnosis of Diabetes
  3. Dabelea D, Stafford JM, Mayer-­Davis EJ et al. Association of
type 1 diabetes vs type 2 diabetes diagnosed during child-
hood and adolescence with complications during teenage
years and young adulthood. JAMA. 2017;317:825–835.
  4. Hamman RF, Bell RA, Dabelea D et  al. The SEARCH for
diabetes in youth study: Rationale, findings, and future
directions. Diabetes Care. 2014;37:3336–3344.
  5. Cerolsaletti K, Hao W, and Greenbaum CJ. Genetics coming
of age in type 1 diabetes. Diabetes Care. 2019;42:189–191.
  6. Lee HS and Hwang JS. Genetic aspects of type 1 diabetes.
Annals of Pediatric Endocrinology and Metabolism.
2019;24:143–148.
  7. Forgetta V, Manousaki D, Istomine R et al. Rare genetic vari-
ants of large effect influence risk of type 1 diabetes. Diabetes.
2020;69:784–795.
  8. Haider MZ, Rasoul MA, Al-­Mahdi M et al. Association of
protein tyrosine phosphatase non-­ receptor type 22 gene
functional variant c1858t, HLA-­ dq/dr genotypes and
autoantibodies with susceptibility to type-­1 diabetes mellitus
in Kuwaiti Arabs. PLoS One. 2018;13:e0198652.
  9. Vella A, Cooper JD, Lowe CE et al. Localization of a type 1
diabetes locus in the IL2RA/CD25 region by use of tag
single-­nucleotide polymorphisms. Am J Hum Genet
2005;76(5):773–779.
10. Todd JA, Walker NM, Cooper JD et al. Robust associations
of four new chromosome regions from genome-­wide analy-
ses of type 1 diabetes. Nat Genet 2007;39(7):857–864.
11. Regnell SE and Lernmark Å. Early prediction of autoim-
mune (type 1) diabetes. Diabetologia. 2017;60:1370–1381.
12. Strollo R, Vinci C, Arshad MH et  al. Antibodies to post-­
translationally modified insulin in type 1 diabetes.
Diabetologia. 2015;58:2851–2860.
13. Strollo R, Vinci C, Napoli N et  al. Antibodies to post-­
translationally modified insulin as a novel biomarker for
prediction of type 1 diabetes in children. Diabetologia.
2017;60:1467–1474.
14. Rewers M and Ludvigsson J. Environmental risk factors for
type 1 diabetes. The Lancet. 2016;387:2340–2348.
15. Lamb MM, Miller M, Seifert JA et al. The effect of childhood
cow’s milk intake and HLA-­DR genotype on risk of islet auto-
immunity and type 1 diabetes: The Diabetes Autoimmunity
Study in the Young. Pediatr. Diabetes. 2015;16:31–38.
16. Schramm S, Lahner H, Jöckel KH et al. Impact of season and
different vitamin D thresholds on prevalence of vitamin D
deficiency in epidemiological cohorts  – a note of caution.
Endocrine. 2017;56:658–666.
17. Chowdhury R, Kunutsor S, Vitezova A et al. Vitamin D and
risk of cause specific death: Systematic review and meta-­
analysis of observational cohort and randomised interven-
tion studies. BMJ. 2014;348:g1903.
18. Napoli N, Strollo R, Pitocco D et  al. Effect of calcitriol on
bone turnover and osteocalcin in recent-­onset type 1 diabe-
tes. PLoS One. 2013;8:e56488.
19. Knip M, Akerblom HK, Al Taji E et al. Effect of hydrolyzed
infant formula vs conventional formula on risk of type 1 dia-
betes the TRIGR randomized clinical trial. JAMA.
2018;319:38–48.
20. Hummel S, Pflüger M, Hummel M et  al. Primary dietary
intervention study to reduce the risk of islet autoimmunity
in children at increased risk for type 1 diabetes. Diabetes
Care. 2011;34:1301–1305.
21. Beyerlein A, Chmiel R, Hummel S et  al. Timing of gluten
introduction and islet autoimmunity in young children:
Updated results from the Babydiet study. Diabetes Care.
2014;37:e194–195.
22. Uusitalo U, Lee HS, Andren Aronsson C et al. Early infant
diet and islet autoimmunity in the TEDDY study. Diabetes
Care. 2018;41:522–530.
23. Uusitalo U, Liu X, Yang J et al. Association of early exposure
of probiotics and islet autoimmunity in the TEDDY study.
JAMA Pediatr. 2016;170:20–28.
24. Sørensen IM, Joner G, Jenum PA et al. Serum long chain n-­3
fatty acids (EPA and DHA) in the pregnant mother are inde-
pendent of risk of type 1 diabetes in the offspring. Diabetes.
Metab. Res. Rev. 2012;28:431–438.
25. Näntö-­Salonen K, Kupila A, Simell S et al. Nasal insulin to
prevent type 1 diabetes in children with HLA genotypes and
autoantibodies conferring increased risk of disease: a
double-­ blind, randomised controlled trial. Lancet.
2008;372:1746–1755.
26. Bingley PJ and Gale EAM. Progression to type 1 diabetes in
islet cell antibody-­ positive relatives in the European
Nicotinamide Diabetes Intervention Trial: The role of addi-
tional immune, genetic and metabolic markers of risk.
Diabetologia. 2006;49:881–890.
27. Beik P, Ciesielska M, Kucza M et  al. Prevention of type 1
diabetes: Past experiences and future opportunities. J. Clin.
Med. 2020;9.
28. Skyler JS, Krischer JP, Wolfsdorf J. Effects of oral insulin in
relatives of patients with type 1 diabetes: The diabetes pre-
vention trial-­type 1. Diabetes Care. 2005;28:1068–1076.
29. Vehik K, Cuthbertson D, Ruhlig H et al. Long-­term outcome of indi-
viduals treated with oral insulin: Diabetes Prevention Trial-­Type 1
(DPT-­1) oral insulin trial. Diabetes Care. 2011;34:1585–1590.
30. Ryhänen SJ, Harkonen T, Siljander H et al. Impact of intra-
nasal insulin on insulin antibody affinity and isotypes in
 Early Diagnosis of Type 1 Diabetes – Useful or a Pyrrhic Victory?
young children with HLA-­conferred susceptibility to type 1
diabetes. Diabetes Care. 2011;34:1383–1388.
31. Gubitosi-­Klug RA, Lachin JM, Backlund JYC et al. Intensive
diabetes treatment and cardiovascular outcomes in type 1
diabetes: The DCCT/EDIC study 30-­ year follow-­ up.
Diabetes Care. 2016;39(5):686–693.
32. Lachin JM, Bebu I, Bergenstal RM et al. Association of glyce-
mic variability in type 1 diabetes with progression of micro-
vascular outcomes in the diabetes control and complications
trial. Diabetes Care. 2017;40:777–783.
33. Bone RN and Evans-­Molina C. Combination immunotherapy
for type 1 diabetes. Current Diabetes Reports. 2017;17:50 .
34. Atkinson MA, Von Herrath M, Powers AC, and Clare-­
Salzler M. Current concepts on the pathogenesis of type 1
diabetes-­considerations for attempts to prevent and reverse
the disease. Diabetes Care. 2015;38:979–988.
35. Rak K and Bronkowska M. Immunomodulatory effect of
vitamin D and its potential role in the prevention and treat-
ment of type 1 diabetes mellitus: A narrative review.
Molecules. 2018;24:24.
36. Gabbay MAL, Sato MN, Finazzo C et al. Effect of cholecal-
ciferol as adjunctive therapy with insulin on protective
immunologic profile and decline of residual β-­cell function
in new-­onset type 1 diabetes mellitus. Arch. Pediatr. Adolesc.
Med. 2012;166:601–607.
37. Bogdanou D, Penna-­ Martinez M, Filmann N et  al.
T-­lymphocyte and glycemic status after vitamin D treatment
in type 1 diabetes: A randomized controlled trial with
sequential crossover. Diabetes. Metab. Res. Rev. 2017;33.
38. Treiber G, Prietl B, Frohlich-­Reiterer E et al. Cholecalciferol
supplementation improves suppressive capacity of regula-
tory T-­cells in young patients with new-­onset type 1 diabetes
mellitus: A randomized clinical trial. Clin. Immunol.
2015;161:217–224.
39. Ataie-­Jafari A, Loke SC, Rahmat AB et  al. A randomized
placebo-­controlled trial of alphacalcidol on the preservation
of beta cell function in children with recent onset type 1 dia-
betes. Clin. Nutr. 2013;32:911–917.
40. Pitocco D, Crino A, Di Stasio E et al. The effects of calcitriol
and nicotinamide on residual pancreatic β-­cell function in
patients with recent-­onset Type 1 diabetes (IMDIAB XI).
Diabet. Med. 2006;23:920–923.
41. Bizzarri C, Pitocco D, Napoli N et al. No protective effect of
calcitriol on β-­cell function in recent-­onset type 1 diabetes:
The IMDIAB XIII trial. Diabetes Care. 2010;33:1962–1963.
42. Herold KC, Hagopian W, Auger JA et al. Anti-­CD3 mono-
clonal antibody in new-­onset type 1 diabetes mellitus. N.
Engl. J. Med. 2002;346:1692–1698.
35
43. Keymeulen B, Vandemeulebroucke E, Ziegler AG et  al.
Insulin needs after CD3-­antibody therapy in new-­onset type
1 diabetes. N. Engl. J. Med. 2005;352:2598–2608.
44. Herold KC, Bundy BN, Long SA et  al. An anti-­CD3 anti-
body, teplizumab, in relatives at risk for type 1 diabetes. N.
Engl. J. Med. 2019;381:603–613.
45. Aronson R, Gottlieb PA, Christiansen JS et al. Low-­dose ote-
lixizumab anti-­ CD3  monoclonal antibody DEFEND-­ 1
study: Results of the randomized phase III study in recent-­
onset human type 1 diabetes. Diabetes Care 2014;37:
2746–2754.
46. Pescovitz MD, Greenbaum CJ, Bundy B et al. B-­lymphocyte
depletion with rituximab and β-­ cell function: Two-­ year
results. Diabetes Care. 2014;37:453–459.
47. Orban T, Bundy B, Becker DJ et al. Co-­stimulation modula-
tion with abatacept in patients with recent-­onset type 1 dia-
betes: A randomised double-­masked controlled trial. Lancet.
2011;378:412–419.
48. Ludvigsson J, Krisky D, Casas R et al. GAD65 antigen ther-
apy in recently diagnosed type 1 diabetes mellitus. N. Engl. J.
Med. 2012;366:433–442.
49. Ludvigsson J. Autoantigen treatment in type 1 diabetes:
Unsolved questions on how to select autoantigen and
administration route. International Journal of Molecular
Sciences. 2020;21.
50. Elding Larsson H, Lundgren M, Jonsdottir B et al. Safety and
efficacy of autoantigen-­ specific therapy with 2 doses of
alum-­formulated glutamate decarboxylase in children with
multiple islet autoantibodies and risk for type 1 diabetes: A
randomized clinical trial. Pediatr. Diabetes. 2018;19:
410–419.
51. Lopes M, Kutlu B, Miani M et  al. Temporal profiling of
cytokine-­induced genes in pancreatic β-­ cells by meta-­
analysis and network inference. Genomics. 2014;103:
264–275.
52. Sota J, Vitale A, Insalaco A et al. Safety profile of the inter-
leukin-­1 inhibitors anakinra and canakinumab in real-­life
clinical practice: a nationwide multicenter retrospective
observational study. Clin. Rheumatol. 2018;37:
2233–2240.
53. Cabello-­Olmo M, Araña M, Radichev I et al. New insights
into immunotherapy strategies for treating autoimmune
diabetes. International Journal of Molecular Sciences.
2019;20.
54. Moran A, Bundy B, Becker DJ et  al. Interleukin-­1 antago-
nism in type 1 diabetes of recent onset: Two multicentre,
randomised, double-­blind, placebo-­controlled trials. Lancet.
2013;381:1905–1915.
 36 Prediabetes and the Diagnosis of Diabetes

55. Bundy BN and Krischer JP. A quantitative measure of treat- 57. Zhang Z, Yan X, Wu C et al. Adding vitamin D3 to the dipep-
ment response in recent-­onset type 1 diabetes. Endocrinol. tidyl peptidase-­4 inhibitor saxagliptin has the potential to pro-
Diabetes Metab. 2020;3:e00143. tect β-­cell function in LADA patients: A 1-­year pilot study.
56. Buzzetti R, Pozzilli P, Frederich R et al. Saxagliptin improves Diabetes. Metab. Res. Rev. 2020;36:e3298.
glycaemic control and C-­peptide secretion in latent autoim-
mune diabetes in adults (LADA). Diabetes. Metab. Res. Rev.
2016;32:289–296.
 3 Reclassifying or Declassifying Diabetes?
Can Clinical Characteristics Guide
Classification and Treatment?
Adrian Vella
Professor of Medicine, Division of Endocrinology, Diabetes, Metabolism, and Nutrition, Mayo Clinic, Rochester, MN, USA
LE A R NI NG P OI NT S
●● Type 2 diabetes is differentiated from type 1 diabetes
by what it is not i.e. a lack of evidence of immune‐
mediated destruction of insulin‐secreting cells. This
oversimplification may miss some of the heterogeneity
present in type 2 diabetes.
●● Current methods of differentiating type 1 from type 2
diabetes have significant limitations and lack sensitivity
and specificity.
●● Glucose‐ and lipo‐toxicity can adversely affect insulin
secretion in the short term although this state of affairs
is not necessarily permanent.
●● Attention to a possible diagnosis of monogenic
diabetes and achieving glycemic control safely and
effectively are more likely to be clinically relevant than
current attempts to sub‐classify type 2 diabetes.
The pathophysiology of hyperglycemia –
type 1 vs type 2 diabetes
It is important to appreciate that the classification of type 1
and type 2 diabetes was originally somewhat arbitrary and
has now evolved into a “positive” diagnosis of type 1 diabe­
tes – based on the presence of autoantibodies – and a diag­
nosis of type 2 diabetes “by exclusion” i.e. no evidence of
islet autoimmunity. However, this approach has significant
limitations driven in part by the nature of the tests used
and by the heterogeneity of type 2 diabetes [1]. Of course,
the ultimate demonstration of an immune‐mediated cause
Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
of diabetes would require documenting an immune infil­
trate within islets [2]. This is extremely invasive, has a sig­
nificant risk of serious complications and for these reasons
is almost never undertaken (nor should it).
Several authors have argued persuasively that if the ulti­
mate goal of diabetes management is to achieve glycemic
control safely and effectively, then the underlying diagnosis
may be less important  [1]. In the absence of therapeutic
choices this may be less important but given the prolifera­
tion of therapeutic classes in diabetes pharmacotherapy a
greater understanding of the pathophysiologic abnormali­
ties resulting in hyperglycemia may help optimize therapy.
This is perhaps best illustrated by the demonstration that
several forms of monogenic diabetes respond to sulfonylu­
reas and do not necessarily need insulin therapy [3]. Early
detection and treatment of hemochromatosis may prevent
loss of pancreatic islet function and diabetes associated
with generalized lipodystrophy responding to leptin ther­
apy serve as other examples.
How does hyperglycemia arise? In the postprandial state
glucose concentrations are the net result of stimulation of
insulin secretion, suppression of glucagon secretion, the
ability of insulin (insulin action) and glucose (glucose
effectiveness) to suppress endogenous glucose production
and stimulate glucose uptake by the tissues, and the rate of
gastric emptying. In type 1 diabetes, all defects are ulti­
mately secondary to insulin deficiency whereas in type 2
diabetes the relative contribution of these parameters is
more variable [4, 5].
37
 38 Prediabetes and the Diagnosis of Diabetes
Defective and/or delayed insulin secretion is the hall­
mark of all forms of diabetes and the presence of hypergly­
cemia implies that the degree of insulin secretion is
inadequate for the prevailing insulin action [6]. Increasing
secretory demands may overwhelm the ability of β‐cells to
assemble insulin leading to protein misfolding and a cas­
cade of mechanisms, including the unfolded protein
response, that result in dysfunction and if unchecked the
death of the β‐cell [7, 8]. Common genetic variation asso­
ciated with type 2 diabetes has helped to identify multiple
pathways associated with the synthesis and secretion of
insulin. It is important to note, however, that the genetic
predisposition to fasting hyperglycemia differs from that
predisposing to glucose intolerance  [9, 10]. This is con­
cordant with the observation that impaired fasting glucose
can occur independently of impaired glucose tolerance and
vice‐versa in prediabetes [11].
Abnormalities of glucagon suppression are observed in
both type 1 and type 2 diabetes and were initially attributed
to insulin deficiency within the islet [12]. However, insulin
restraint of α‐cell secretion may not be as important as pre­
viously thought [13, 14]. Certainly there are other parac­
rine regulators of glucagon secretion  [15]. Abnormal
glucagon secretion arises early in prediabetes and occurs
independently of defects in insulin secretion  [14].
Underlining its importance in the pathogenesis of diabetes
is the observation that people with diabetes‐associated
genetic variation exhibit defects of α‐cell function [16, 17].
The factors altering the ability of insulin and of glucose
itself to suppress endogenous glucose production (and
release into the circulation) as well as stimulate uptake are
less well understood although weight, adiposity and physi­
cal activity all influence these parameters. Genetic predis­
position to defects in insulin action, for example, is less
well characterized. Certain syndromes such as polycystic
ovarian syndrome are associated with diabetes through
effects on insulin resistance [18, 19].
The final variable affecting glycemic control is upper
gastrointestinal function. The stomach and proximal small
bowel function in unison to dampen fluctuations in the
rate of appearance of ingested calories into the duodenum
and jejunum after meals. Multiple neural and hormonal
inputs regulate gastric volume and wall tension (allowing
accommodation of ingested food), pyloric tone and the
rate of gastric emptying  [20]. Although this integrated
system regulates satiety and, to a lesser extent, caloric
intake, there has been no clear association with predisposi­
tion to diabetes. For example, most of the benefits of bari­
atric surgery occur through changes in caloric intake and
weight loss [21] and are likely independent of the rate of
gastric emptying – outcomes after sleeve gastrectomy and
Roux‐en‐Y‐Gastric Bypass (RYGB) are broadly compara­
ble – despite a far higher rate of gastric (pouch) emptying
after RYGB [22].
Genetic and environmental influences
on disease presentation
Having somewhat downplayed the notion that there are
two discrete forms of diabetes, one must acknowledge that
the extremes of the disease spectrum are fairly easy to rec­
ognize and seem to differ in their genetic predisposition.
Both type 1 and type 2 exhibit genetic predisposition to the
disease but the environmental contribution to type 2 dia­
betes is much more prominent [23].
Common genetic variation in more than 200  loci has
been associated with type 2 diabetes. The risk conferred by
most of these variants is small – indeed knowledge of
genetic variation at the 18 loci with greatest effect on dis­
ease risk did not appreciably alter the performance of a pre­
diction model utilizing anthropometric information and
family history  [24]. Although later models incorporating
genetic information from additional loci improved their
predictive performance, especially in younger adults, it
remains apparent that genetic information is unhelpful in
predicting type 2 diabetes risk at an individual level [25].
In contrast, genetic variation in the human leukocyte
antigen (HLA) confers more than half of the genetic risk of
type 1 diabetes. HLA binds and processes antigen‐presen­
tation to the immune system. A handful of other loci
involved in immune response pathways confer significant
additional risk. Other variants (~50) also contribute
smaller effects. Most of the loci associated with type 1 dia­
betes alter immune regulation [26–28].
The environmental events, if any, that trigger the cas­
cade of immune‐mediated destruction in type 1 diabetes
are not well defined. Often hyperglycemia is detected for
the first time in the context of a febrile illness, but it is well
accepted that significant islet destruction has occurred by
the time hyperglycemia develops. In contrast, excess caloric

intake and physical inactivity are associated with type 2
diabetes. However, this is not uniform, so that of all patients
coming to bariatric surgery in the United States (BMI >
35 kg/m2) only a third have diabetes. The reasons for this
heterogeneity are unclear at present [29]. Another interest­
ing observation is the association of shift work and circa­
dian misalignment with type 2 diabetes – this may be
mediated through effects on insulin synthesis and secre­
tion. Genetic variation in the melatonin receptor
(MTNR1B) – a key part of circadian signaling – is associ­
ated with type 2 diabetes [30, 31].
Measuring insulin secretion
Insulin is secreted by the β‐cell in response to various stim­
uli. However, peripheral insulin concentrations may be an
imperfect measure of insulin secretion because they repre­
sent the net sum of two opposing processes. The first is
actual insulin secretion into the portal vein while the sec­
ond is hepatic extraction of insulin that occurs across the
liver as insulin reaches the systemic circulation. It is uncer­
tain whether hepatic extraction is an active or a passive
process, but it is likely proportional to the magnitude of
insulin secretion  [32] and declines as β‐cell function
declines [33].
C‐peptide arises from the post‐translational processing
of insulin as preproinsulin which folds upon itself to form
specific disulfide bonds, resulting in a dimeric structure
after cleavage of the connecting (C‐)peptide. This peptide
is secreted in a 1:1 ratio with insulin but does not undergo
hepatic extraction. Therefore, in theory, C‐peptide concen­
trations serve as a better measure of insulin secretion than
do insulin concentrations themselves. However, C‐peptide
which is cleared by the kidney (and therefore cannot be
used reliably in renal dysfunction or failure) has a half‐life
of ~30 minutes and therefore accumulates in the circula­
tion compared to insulin (half‐life of < 5minutes).
Deconvolution of insulin secretion rates from C‐peptide
concentrations requires knowledge of the kinetics under­
ling C‐peptide clearance. This can be estimated from
anthropometric criteria thanks to the work of Van Cauter
et al. so that insulin secretion can be measured accurately
in humans with intact renal function [34, 35].
The other factor when assessing insulin secretion is the
nature of the stimulus – a physiological stimulus mimicking
Reclassifying or Declassifying Diabetes? 39
situations similar to everyday life e.g. an oral glucose toler­
ance test or a mixed meal tolerance test may be better at
detecting subtle defects in insulin secretion as compared to
supraphysiologic stimuli. The latter use stimuli intended to
produce submaximal insulin secretion e.g. glucagon or
arginine stimulation tests  [36–38]. Similarly, an intrave­
nous bolus of glucose that can change peripheral concen­
trations of glucose very rapidly (as part of an intravenous
glucose tolerance test – IVGTT) is also a supraphysiologic
stimulus to β‐cell secretion. Intravenous glucose chal­
lenges produce a characteristic biphasic pattern of insu­
lin secretion in healthy subjects that is not observed in
people with early type 1 diabetes or in people with type 2
diabetes [39].
While isolated measurement of insulin secretion in
response to a standardized stimulus might provide a quali­
tative measurement of β‐cell function, truly quantitative
measures require expressing insulin secretion as a function
of insulin action. Such indices generated through mathe­
matical modelling can predict risk of progression to diabe­
tes or quantify response to an intervention. However, at
present they have limited utility on an individual basis due
to lack of sufficient normative data  [6]. Nevertheless,
absent or near absent C‐peptide concentrations in the fast­
ing state or in response to stimuli may help demonstrate an
absence of endogenous insulin secretion.
Autoantibodies
Several autoantibodies can be used in clinical practice to
document the presence of islet autoimmunity. These are
antibodies directed against Glutamic Acid Decarboxylase
(GAD‐65), insulin, tyrosine phosphatases, insulinoma‐
associated protein 2 (IA‐2 and IA‐2β), and the zinc trans­
porter (ZnT8). The greater the number of autoantibodies
that are positive and the higher the titer of these antibodies,
the more likely that islet autoimmunity is present  [40].
However, people with a clinical presentation compatible
with type 1 diabetes are often antibody negative. The oppo­
site sometimes applies. Indeed, in a cohort of Belgian
patients with type 1 diabetes presenting before age 40, 24%
did not have autoantibodies [41]. In the Botnia study, 4.4%
of healthy individuals had GAD antibodies [42].
Latent autoimmune diabetes of adults (LADA) likely
represents a “forme fruste” of type 1 diabetes where a lower
 40 Prediabetes and the Diagnosis of Diabetes
genetic burden of predisposition results in a later age at
presentation, decreased amount and degree of autoanti­
body positivity, and a longer persistence of endogenous
insulin secretion. Patients with LADA tend to have a lower
BMI than patients with type 2 diabetes. Of note, antibodies
against insulin and IA‐2 are rarely present in late‐onset dia­
betes. Islet autoantibodies are most useful in screening
family members of patients with type 1 diabetes for risk of
developing the disease [43].
“Atypical” diabetes, glucose toxicity
and the honeymoon period
Atypical diabetes or “ketosis‐prone type 2 diabetes” were
monikers used to describe obese minority (typically
African‐American) patients presenting with diabetic
ketoacidosis as their first manifestation of diabetes  [44].
Typically, after acute metabolic control is achieved, such
patients are able to maintain glycemic control over long
periods of time without insulin (often with diet alone). The
concept of glucose toxicity where hyperglycemia per se
impairs insulin secretion and action has been put forward
to explain these observations  [45]. Indeed, during recov­
ery, these indices do not differ from those of age‐ and
weight‐matched controls  [46]. Similar short‐term expo­
sure to lipotoxicity also has little effect on β‐cell
function [47].
Short‐term intensive insulin therapy in a cohort of
Chinese patients with a first presentation of diabetes pro­
duced a “remission” of diabetes with preservation of insu­
lin secretion at one year when compared to conventionally
treated patients  [48]. Diabetes remission, independent of
weight‐loss, is also reported after bariatric surgery but – at
least in the short‐term is likely explained by caloric restric­
tion and β‐cell “rest” [49]. This is in keeping with the con­
cept that insulin over‐production may produce
endoplasmic reticulum stress and if unchecked lead to β‐
cell death [50].
These factors probably explain the resumption of
endogenous insulin secretion and decreased or absent
exogenous insulin requirements during the honeymoon
period in type 1 diabetes. The term refers to a variable
interval after presentation with hyperglycemia and subse­
quent restoration of metabolic and glycemic control. Some
patients have a prolonged remission and indeed tight
glycemic control seems to prolong the persistence of
endogenous insulin secretion in affected patients.
Cystic fibrosis‐related diabetes
Diabetes is the most common extra‐pulmonary complica­
tion of cystic fibrosis and rises with age so that 50% of
affected patients over age 30 have diabetes [51]. There is a
recognized abnormality in insulin secretion which is likely
explained by the actions of the Cystic Fibrosis
Transmembrane Regulator (CFTR) on β‐cell electrophysi­
ology. Drugs that improve CFTR function also have salu­
tary effects on insulin secretion [52]. Exocrine pancreatic
insufficiency may decrease intraluminal release of nutri­
ents that stimulate incretin hormone secretion and delay
gastric emptying thereby contributing to postprandial
hyperglycemia. Endocrine dysfunction is also strongly cor­
related with destruction of the exocrine pancreas.
Nutritional deficiency and also the systemic inflamma­
tion associated with cystic fibrosis likely impair insulin
action which is also a feature of Cystic Fibrosis‐related
Diabetes. Macrovascular complications are uncommon
and microvascular complications occur at lower rates than
they do in “classical” type 1 and type 2 diabetes [53].
Novel approaches to reclassifying
diabetes
Ahlqvist et al. undertook a data‐driven cluster analysis in
patients with newly diagnosed diabetes from a Swedish
cohort. They utilized six variables measured at the time of
diagnosis – age at diagnosis, BMI, HbA1c, GAD antibodies,
and homoeostatic model assessment (HOMA) estimates of
β‐cell function and insulin resistance. The investigators
identified five clusters based on these variables – one clus­
ter was antibody positive, and relatively lean with poor
metabolic control. The second cluster differed from the
first in that antibodies were absent, while the third cluster
had an elevated BMI and insulin resistance. Cluster 4 was
also obese but defects in insulin action as measured by
HOMA were not as severe as in the 3rd cluster. Finally,
patients in the 5th cluster were older but, like cluster 4, had
less severe metabolic abnormalities. These clusters were
subsequently validated in 3 independent cohorts. Over sub­
sequent follow‐up, the incidence of diabetic ketoacidosis

and of non‐alcoholic fatty liver disease were highest in the
first and third cluster, respectively. This latter cluster also
had the highest risk of chronic kidney disease  [54].
HOMA measures have significant limitations and poor
predictive value for progression to diabetes  [55] –
although in fairness the investigators used C‐peptide
rather than insulin to estimate β‐cell function. Whether
this will help guide therapeutic interventions and improve
outcomes in individual patients remains to be
ascertained.
Another approach to characterizing different sub‐types
of diabetes is based on common genetic variation and its
influence on quantitative traits such as glucose tolerance in
response to a standardized challenge. For example, Dimas
et  al. identified 5 clusters of genetic risk loci that altered
glycemic traits [56]. In this study examining data from ~58
000 non‐diabetic subjects, one cluster altered insulin sensi­
tivity. A second cluster altered fasting glucose while a third
cluster altered the ratio of proinsulin to insulin in the fast­
ing state. Another cluster was primarily characterized by
defects in post‐challenge insulin secretion. The final clus­
ter of risk loci did not alter glycemic traits. While this exer­
cise has certainly helped understand the effect of genotype
on phenotype, the effect size of each risk allele on a given
phenotypic trait is so small as be unhelpful in terms of pre­
dicting individual clinical behavior.
Pharmacogenomics has also held promise as a way of
classifying diabetes. This hope was reinforced by the discov­
ery that the target for thiazolidinediones and sulfonylureas
are risk loci for type 2 diabetes. Unfortunately, their effect
on response to therapy is difficult to discern at an individ­
ual  level and genotype at PPARG and KCNJ11 should not
alter therapeutic choices in type 2 diabetes [57, 58].
Conclusions
Multiple roads lead to the pathophysiologic defects that
cause hyperglycemia. While autoimmune diabetes can be
characterized quite readily in many patients, the remaining
patients are far more heterogeneous. Efforts to reinvent the
classification, at least to date, have not appreciably changed,
or guided, the management of individual patients – the
notable exception being monogenic forms of diabetes.
Nevertheless, appreciating the underlying defects as well as
the limitations of autoantibody and β‐cell function testing
Reclassifying or Declassifying Diabetes? 41
should help improve the care of patients with unclear clas­
sification and with associated comorbidities.
Bibliography
1. Gale EA. Declassifying diabetes. Diabetologia.
2006;49:1989–1995.
2. Kaestner KH, Powers AC, Naji A, Consortium H, Atkinson
MA. NIH Initiative to Improve Understanding of the
Pancreas, Islet, and Autoimmunity in Type 1 Diabetes:
The Human Pancreas Analysis Program (HPAP). Diabetes.
2019;68:1394–1402.
3. Hattersley AT, Pearson ER. Minireview: pharmacogenetics
and beyond: the interaction of therapeutic response,
beta‐cell physiology, and genetics in diabetes.
Endocrinology. 2006;147:2657–2663.
4. Smushkin G, Vella A. What is type 2 diabetes? Medicine
(Baltimore). 2010;38:597–601.
5. Vella A, Camilleri M, Rizza RA. The gastrointestinal tract
and glucose tolerance. Curr Opin Clin Nutr Metab Care.
2004;7:479–484.
6. Cobelli C, Dalla Man C, Toffolo G, Basu R, Vella A, Rizza
R. The oral minimal model method. Diabetes.
2014;63:1203–1213.
7. Haataja L, Manickam N, Soliman A, Tsai B, Liu M, Arvan P.
Disulfide mispairing during proinsulin folding in the
endoplasmic reticulum. Diabetes. 2016;.
8. Sun J, Cui J, He Q, Chen Z, Arvan P, Liu M. Proinsulin
misfolding and endoplasmic reticulum stress during the
development and progression of diabetes. Mol Aspects Med
2015;42:105–118.
9. Smushkin G, Vella A. Genetics of type 2 diabetes. Curr
Opin Clin Nutr Metab Care. 2010;13:471–477.
10. Saxena R, Hivert MF, Langenberg C, Tanaka T, Pankow JS,
Vollenweider P, Lyssenko V, Bouatia‐Naji N, Dupuis J,
Jackson AU, Kao WH, Li M, Glazer NL, Manning AK, Luan
J, Stringham HM, Prokopenko I, Johnson T, Grarup N,
Boesgaard TW, Lecoeur C, Shrader P, O’Connell J, Ingelsson
E, Couper DJ, Rice K, Song K, Andreasen CH, Dina C,
Kottgen A, Le Bacquer O, Pattou F, Taneera J, Steinthorsdottir
V, Rybin D, Ardlie K, Sampson M, Qi L, van Hoek M,
Weedon MN, Aulchenko YS, Voight BF, Grallert H, Balkau
B, Bergman RN, Bielinski SJ, Bonnefond A, Bonnycastle LL,
Borch‐Johnsen K, Bottcher Y, Brunner E, Buchanan TA,
Bumpstead SJ, Cavalcanti‐Proenca C, Charpentier G, Chen
YD, Chines PS, Collins FS, Cornelis M, G JC, Delplanque J,
Doney A, Egan JM, Erdos MR, Firmann M, Forouhi NG,
Fox CS, Goodarzi MO, Graessler J, Hingorani A, Isomaa B,
Jorgensen T, Kivimaki M, Kovacs P, Krohn K, Kumari M,
 42 Prediabetes and the Diagnosis of Diabetes
Lauritzen T, Levy‐Marchal C, Mayor V, McAteer JB, Meyre
D, Mitchell BD, Mohlke KL, Morken MA, Narisu N, Palmer
CN, Pakyz R, Pascoe L, Payne F, Pearson D, Rathmann W,
Sandbaek A, Sayer AA, Scott LJ, Sharp SJ, Sijbrands E,
Singleton A, Siscovick DS, Smith NL, Sparso T, Swift AJ,
Syddall H, Thorleifsson G, Tonjes A, Tuomi T, Tuomilehto J,
Valle TT, Waeber G, Walley A, Waterworth DM, Zeggini E,
Zhao JH, Illig T, Wichmann HE, Wilson JF, van Duijn C, Hu
FB, Morris AD, Frayling TM, Hattersley AT, Thorsteinsdottir
U, Stefansson K, Nilsson P, Syvanen AC, Shuldiner AR,
Walker M, Bornstein SR, Schwarz P, Williams GH, Nathan
DM, Kuusisto J, Laakso M, Cooper C, Marmot M, Ferrucci
L, Mooser V, Stumvoll M, Loos RJ, Altshuler D, Psaty BM,
Rotter JI, Boerwinkle E, Hansen T, Pedersen O, Florez JC,
McCarthy MI, Boehnke M, Barroso I, Sladek R, Froguel P,
Meigs JB, Groop L, Wareham NJ, Watanabe RM. Genetic vari­
ation in GIPR influences the glucose and insulin responses to
an oral glucose challenge. Nat Genet. 2010;42:142–148.
11. Bock G, Dalla Man C, Campioni M, Chittilapilly E, Basu R,
Toffolo G, Cobelli C, Rizza R. Pathogenesis of pre‐diabetes:
mechanisms of fasting and postprandial hyperglycemia in
people with impaired fasting glucose and/or impaired glu­
cose tolerance. Diabetes. 2006;55:3536–3549.
12. Unger RH, Orci L. The essential role of glucagon in the
pathogenesis of diabetes mellitus. Lancet. 1975;1:14–16.
13. Faerch K, Vistisen D, Pacini G, Torekov SS, Johansen NB,
Witte DR, Jonsson A, Pedersen O, Hansen T, Lauritzen T,
Jorgensen ME, Ahren B, Holst JJ. Insulin resistance is
accompanied by increased fasting glucagon and delayed
glucagon suppression in individuals with normal and
impaired glucose regulation. Diabetes. 2016;65:3473–3481.
14. Sharma A, Varghese RT, Shah M, Man CD, Cobelli C, Rizza
RA, Bailey KR, Vella A. Impaired insulin action is associated
with increased glucagon concentrations in non‐diabetic
humans. J Clin Endocrinol Metab. 2018;103:314–319.
15. Ashcroft FM, Rorsman P. K(ATP) channels and islet hor­
mone secretion: new insights and controversies. Nature
Reviews. 2013;9:660–669.
16. Karanth S, Adams JD, Serrano MLA, Quittner‐Strom EB,
Simcox J, Villanueva CJ, Ozcan L, Holland WL, Yost HJ,
Vella A, Schlegel A. A hepatocyte FOXN3‐alpha cell gluca­
gon axis regulates fasting glucose. Cell Rep. 2018;24:
312–319.
17. Shah M, Varghese RT, Miles JM, Piccinini F, Dalla Man C,
Cobelli C, Bailey KR, Rizza RA, Vella A. TCF7L2 Genotype
and alpha‐cell function in humans without diabetes.
Diabetes. 2016;65:371–380.
18. Corbould A, Kim YB, Youngren JF, Pender C, Kahn BB, Lee
A, Dunaif A. Insulin resistance in the skeletal muscle of
women with PCOS involves intrinsic and acquired defects in
insulin signaling. American Journal of Physiology Endocrinology
and Metabolism. 2005;288:E1047–1054.
19. Urbanek M, Legro RS, Driscoll D, Strauss JF, 3rd, Dunaif A,
Spielman RS. Searching for the polycystic ovary syndrome
genes. J Pediatr Endocrinol Metab. 2000;13 Suppl 5:1311–1313.
20. Vella A, Camilleri M. The gastrointestinal tract as an inte­
grator of mechanical and hormonal response to nutrient
ingestion. Diabetes. 2017;66:2729–2737.
21. Ikramuddin S, Korner J, Lee WJ, Thomas AJ, Connett JE,
Bantle JP, Leslie DB, Wang Q, Inabnet WB, 3rd, Jeffery RW,
Chong K, Chuang LM, Jensen MD, Vella A, Ahmed L, Belani
K, Billington CJ. Lifestyle intervention and medical manage­
ment with vs without Roux‐en‐Y Gastric Bypass and control
of hemoglobin A1c, LDL cholesterol, and systolic blood
pressure at 5 years in the Diabetes Surgery Study. JAMA.
2018;319:266–278.
22. Schauer PR, Bhatt DL, Kirwan JP, Wolski K, Brethauer SA,
Navaneethan SD, Aminian A, Pothier CE, Kim ES, Nissen
SE, Kashyap SR. Bariatric surgery versus intensive medical
therapy for diabetes – 3‐year outcomes. The N Engl J Med.
2014;370:2002–2013.
23. Florez JC, Hirschhorn J, Altshuler D. The inherited basis of
diabetes mellitus: implications for the genetic analysis of com­
plex traits. Annu Rev Genomics Hum Genet. 2003;4:257–291.
24. Meigs JB, Shrader P, Sullivan LM, McAteer JB, Fox CS,
Dupuis J, Manning AK, Florez JC, Wilson PW, D’Agostino
RB, Sr., Cupples LA. Genotype score in addition to common
risk factors for prediction of type 2 diabetes. N Engl J Med.
2008;359:2208–2219.
25. Khera AV, Chaffin M, Aragam KG, Haas ME, Roselli C, Choi
SH, Natarajan P, Lander ES, Lubitz SA, Ellinor PT, Kathiresan
S. Genome‐wide polygenic scores for common diseases
identify individuals with risk equivalent to monogenic
mutations. Nat Genet. 2018;50:1219–1224.
26. Lowe CE, Cooper JD, Brusko T, Walker NM, Smyth DJ,
Bailey R, Bourget K, Plagnol V, Field S, Atkinson M, Clayton
DG, Wicker LS, Todd JA. Large‐scale genetic fine mapping
and genotype‐phenotype associations implicate polymor­
phism in the IL2RA region in type 1 diabetes. Nat Genet.
2007;39:1074–1082.
27. Todd JA, Walker NM, Cooper JD, Smyth DJ, Downes K,
Plagnol V, Bailey R, Nejentsev S, Field SF, Payne F, Lowe CE,
Szeszko JS, Hafler JP, Zeitels L, Yang JH, Vella A, Nutland S,
Stevens HE, Schuilenburg H, Coleman G, Maisuria M,

Meadows W, Smink LJ, Healy B, Burren OS, Lam AA,
Ovington NR, Allen J, Adlem E, Leung HT, Wallace C,
Howson JM, Guja C, Ionescu‐Tirgoviste C, Simmonds MJ,
Heward JM, Gough SC, Dunger DB, Wicker LS, Clayton
DG. Robust associations of four new chromosome regions
from genome‐wide analyses of type 1 diabetes. Nat Genet.
2007;39:857–864.
28. Smyth DJ, Howson JM, Payne F, Maier LM, Bailey R,
Holland K, Lowe CE, Cooper JD, Hulme JS, Vella A,
Dahlman I, Lam AC, Nutland S, Walker NM, Twells RC,
Todd JA. Analysis of polymorphisms in 16 genes in type 1
diabetes that have been associated with other immune‐
mediated diseases. BMC Med Genet. 2006;7:20.
29. Ikramuddin S, Billington CJ, Lee WJ, Bantle JP, Thomas AJ,
Connett JE, Leslie DB, Inabnet WB, 3rd, Jeffery RW, Chong
K, Chuang LM, Sarr MG, Jensen MD, Vella A, Ahmed L,
Belani K, Schone JL, Olofson AE, Bainbridge HA, Laqua PS,
Wang Q, Korner J. Roux‐en‐Y gastric bypass for diabetes
(the Diabetes Surgery Study): 2‐year outcomes of a 5‐year,
randomised, controlled trial. Lancet Diabetes Endocrinol.
2015;3:413–422.
30. Sharma A, Laurenti MC, Dalla Man C, Varghese RT, Cobelli
C, Rizza RA, Matveyenko A, Vella A. Glucose metabolism
during rotational shift‐work in healthcare workers.
Diabetologia. 2017;60:1483–1490.
31. Bouatia‐Naji N, Bonnefond A, Cavalcanti‐Proenca C,
Sparso T, Holmkvist J, Marchand M, Delplanque J, Lobbens
S, Rocheleau G, Durand E, De Graeve F, Chevre JC, Borch‐
Johnsen K, Hartikainen AL, Ruokonen A, Tichet J, Marre M,
Weill J, Heude B, Tauber M, Lemaire K, Schuit F, Elliott P,
Jorgensen T, Charpentier G, Hadjadj S, Cauchi S, Vaxillaire
M, Sladek R, Visvikis‐Siest S, Balkau B, Levy‐Marchal C,
Pattou F, Meyre D, Blakemore AI, Jarvelin MR, Walley AJ,
Hansen T, Dina C, Pedersen O, Froguel P. A variant near
MTNR1B is associated with increased fasting plasma glu­
cose levels and type 2 diabetes risk. Nat Genet. 2009;41:
89–94.
32. Meier JJ, Veldhuis JD, Butler PC. Pulsatile insulin secretion
dictates systemic insulin delivery by regulating hepatic insu­
lin extraction in humans. Diabetes. 2005;54:1649–1656.
33. Sathananthan A, Man CD, Zinsmeister AR, Camilleri M,
Rodeheffer RJ, Toffolo G, Cobelli C, Rizza RA, Vella A. A
concerted decline in insulin secretion and action occurs
across the spectrum of fasting and postchallenge glucose
concentrations. Clin Endocrinol (Oxf). 2012;76:212–219.
34. Varghese RT, Dalla Man C, Laurenti MC, Piccinini F, Sharma
A, Shah M, Bailey KR, Rizza RA, Cobelli C, Vella A.
Reclassifying or Declassifying Diabetes? 43
Performance of individually measured vs population‐based
C‐peptide kinetics to assess beta‐cell function in the pres­
ence and absence of acute insulin resistance. Diabetes Obes
Metab. 2018;20:549–555.
35. Van Cauter E, Mestrez F, Sturis J, Polonsky KS. Estimation of
insulin secretion rates from C‐peptide levels. Comparison of
individual and standard kinetic parameters for C‐peptide
clearance. Diabetes. 1992;41:368–377.
36. Robertson RP, Bogachus LD, Oseid E, Parazzoli S, Patti ME,
Rickels MR, Schuetz C, Dunn T, Pruett T, Balamurugan AN,
Sutherland DE, Beilman G, Bellin MD. Assessment of beta‐
cell mass and alpha‐ and beta‐cell survival and function by
arginine stimulation in human autologous islet recipients.
Diabetes. 2015;64:565–572.
37. Robertson RP, Raymond RH, Lee DS, Calle RA, Ghosh A,
Savage PJ, Shankar SS, Vassileva MT, Weir GC, Fryburg DA,
Beta cell project team of the foundation for the NIHBC.
Arginine is preferred to glucagon for stimulation testing of
beta‐cell function. Am J Physiol Endocrinol Metab.
2014;307:E720–727.
38. Shankar SS, Vella A, Raymond RH, Staten MA, Calle RA,
Bergman RN, Cao C, Chen D, Cobelli C, Dalla Man C, Deeg
M, Dong JQ, Lee DS, Polidori D, Robertson RP, Ruetten H,
Stefanovski D, Vassileva MT, Weir GC, Fryburg DA,
Foundation for the National Institutes of Health beta‐Cell
Project T. Standardized mixed‐meal tolerance and arginine
stimulation tests provide reproducible and complementary
measures of beta‐cell function: results from the foundation
for the National Institutes of Health Biomarkers Consortium
Investigative Series. Diabetes Care. 2016;39:1602–1613.
39. Bergman RN. Lilly lecture 1989. Toward physiological
understanding of glucose tolerance. Minimal‐model
approach. Diabetes. 1989;38:1512–1527.
40. Skyler JS, Bakris GL, Bonifacio E, Darsow T, Eckel RH,
Groop L, Groop PH, Handelsman Y, Insel RA, Mathieu C,
McElvaine AT, Palmer JP, Pugliese A, Schatz DA, Sosenko
JM, Wilding JP, Ratner RE. Differentiation of diabetes by
pathophysiolog, natural history, and prognosis. Diabetes.
2017;66:241–255.
41. Decochez K, Tits J, Coolens JL, Van Gaal L, Krzentowski G,
Winnock F, Anckaert E, Weets I, Pipeleers DG, Gorus FK.
High frequency of persisting or increasing islet‐specific
autoantibody levels after diagnosis of type 1 diabetes pre­
senting before 40 years of age. The Belgian Diabetes Registry.
Diabetes Care. 2000;23:838–844.
42. Tuomi T, Carlsson A, Li H, Isomaa B, Miettinen A, Nilsson
A, Nissen M, Ehrnstrom BO, Forsen B, Snickars B, Lahti K,
 44 Prediabetes and the Diagnosis of Diabetes
Forsblom C, Saloranta C, Taskinen MR, Groop LC. Clinical
and genetic characteristics of type 2 diabetes with and with­
out GAD antibodies. Diabetes. 1999;48:150–157.
43. Hummel M, Bonifacio E, Schmid S, Walter M, Knopff A,
Ziegler AG. Brief communication: early appearance of islet
autoantibodies predicts childhood type 1 diabetes in offspring
of diabetic parents. Ann Intern Med. 2004;140:882–886.
44. Umpierrez GE, Smiley D, Kitabchi AE. Narrative review:
ketosis‐prone type 2 diabetes mellitus. Annals of Internal
Medicine. 2006;144:350–357.
45. Poitout V, Robertson RP. Glucolipotoxicity: fuel excess and
beta‐cell dysfunction. Endocr Rev. 2008;29:351–366.
46. Gosmanov AR, Smiley D, Robalino G, Siqueira JM, Peng L,
Kitabchi AE, Umpierrez GE. Effects of intravenous glucose
load on insulin secretion in patients with ketosis‐prone dia­
betes during near‐normoglycemia remission. Diabetes Care.
2010;33:854–860.
47. Umpierrez GE, Smiley D, Robalino G, Peng L, Gosmanov
AR, Kitabchi AE. Lack of lipotoxicity effect on {beta}‐cell
dysfunction in ketosis‐prone type 2 diabetes. Diabetes Care.
2010;33:626–631.
48. Weng J, Li Y, Xu W, Shi L, Zhang Q, Zhu D, Hu Y, Zhou Z,
Yan X, Tian H, Ran X, Luo Z, Xian J, Yan L, Li F, Zeng L,
Chen Y, Yang L, Yan S, Liu J, Li M, Fu Z, Cheng H. Effect of
intensive insulin therapy on beta‐cell function and glycae­
mic control in patients with newly diagnosed type 2 diabe­
tes: a multicentre randomised parallel‐group trial. Lancet.
2008;371:1753–1760.
49. Laedtke T, Kjems L, Porksen N, Schmitz O, Veldhuis J, Kao
PC, Butler PC. Overnight inhibition of insulin secretion
restores pulsatility and proinsulin/insulin ratio in type 2 dia­
betes. Am J Physiol Endocrinol Metab. 2000;279:E520–528.
50. Arunagiri A, Haataja L, Cunningham CN, Shrestha N, Tsai
B, Qi L, Liu M, Arvan P. Misfolded proinsulin in the endo­
plasmic reticulum during development of beta cell failure in
diabetes. Ann N Y Acad Sci. 2018;.
51. Moran A, Dunitz J, Nathan B, Saeed A, Holme B, Thomas W.
Cystic fibrosis‐related diabetes: current trends in pre­
valence, incidence, and mortality. Diabetes Care. 2009;32:
1626–1631.
52. Bellin MD, Laguna T, Leschyshyn J, Regelmann W, Dunitz J,
Billings J, Moran A. Insulin secretion improves in cystic
fibrosis following ivacaftor correction of CFTR. a small pilot
study. Pediatr Diabetes. 2013;14:417–421.
53. Moheet A, Moran A. CF‐related diabetes: containing the
metabolic miscreant of cystic fibrosis. Pediatr Pulmonol
2017;52:S37–S43.
54. Ahlqvist E, Storm P, Karajamaki A, Martinell M, Dorkhan
M, Carlsson A, Vikman P, Prasad RB, Aly DM, Almgren P,
Wessman Y, Shaat N, Spegel P, Mulder H, Lindholm E,
Melander O, Hansson O, Malmqvist U, Lernmark A, Lahti
K, Forsen T, Tuomi T, Rosengren AH, Groop L. Novel sub­
groups of adult‐onset diabetes and their association with
outcomes: a data‐driven cluster analysis of six variables.
Lancet Diabetes Endocrinol. 2018;6:361–369.
55. Xiang AH, Watanabe RM, Buchanan TA. HOMA and
Matsuda indices of insulin sensitivity: poor correlation with
minimal model‐based estimates of insulin sensitivity in lon­
gitudinal settings. Diabetologia. 2014;57:334–338.
56. Dimas AS, Lagou V, Barker A, Knowles JW, Magi R, Hivert
MF, Benazzo A, Rybin D, Jackson AU, Stringham HM, Song
C, Fischer‐Rosinsky A, Boesgaard TW, Grarup N, Abbasi
FA, Assimes TL, Hao K, Yang X, Lecoeur C, Barroso I,
Bonnycastle LL, Bottcher Y, Bumpstead S, Chines PS, Erdos
MR, Graessler J, Kovacs P, Morken MA, Narisu N, Payne F,
Stancakova A, Swift AJ, Tonjes A, Bornstein SR, Cauchi S,
Froguel P, Meyre D, Schwarz PE, Haring HU, Smith U,
Boehnke M, Bergman RN, Collins FS, Mohlke KL,
Tuomilehto J, Quertemous T, Lind L, Hansen T, Pedersen O,
Walker M, Pfeiffer AF, Spranger J, Stumvoll M, Meigs JB,
Wareham NJ, Kuusisto J, Laakso M, Langenberg C, Dupuis J,
Watanabe RM, Florez JC, Ingelsson E, McCarthy MI,
Prokopenko I, Investigators M. Impact of type 2 diabetes
susceptibility variants on quantitative glycemic traits
reveals mechanistic heterogeneity. Diabetes. 2014;63:
2158–2171.
57. Florez JC, Jablonski KA, Kahn SE, Franks PW, Dabelea D,
Hamman RF, Knowler WC, Nathan DM, Altshuler D. Type 2
diabetes‐associated missense polymorphisms KCNJ11 E23K
and ABCC8 A1369S influence progression to diabetes and
response to interventions in the Diabetes Prevention
Program. Diabetes. 2007;56:531–536.
58. Florez JC, Jablonski KA, Sun MW, Bayley N, Kahn SE,
Shamoon H, Hamman RF, Knowler WC, Nathan DM,
Altshuler D. Effects of the type 2 diabetes‐associated PPARG
P12A polymorphism on progression to diabetes and
response to troglitazone. J Clin Endocrinol Metab.
2007;92:1502–1509.
 4 How Should Secondary Causes
of Diabetes Be Excluded?
Tomás P. Griffin, Aonghus O’Loughlin, and Sean F. Dinneen
Discipline of Medicine, NUI Galway, Galway, Ireland and Centre for Diabetes, Endocrinology and Metabolism, Galway
University Hospitals, Galway, Ireland
LE A R NI NG P OI NT S
●● Current diabetes clinical practice guidelines do not
address the issue of population‐wide screening for
specific types of diabetes due to secondary causes.
●● Case finding (or opportunistic screening) should be
undertaken among patients with diabetes and clinical
features of the condition under consideration (on a
case by case basis).
●● Serum ferritin and transferrin saturation should be
used to assess iron stores when screening for
hereditary haemochromatosis.
●● When screening for Cushing’s syndrome the following
tests should be considered two 24‐hour urinary free
cortisol, a 1 mg overnight dexamethasone suppression
test, or two late‐night salivary cortisol.
●● Annual oral glucose tolerance testing is recommended
in adolescents and adults with cystic fibrosis.
●● Baseline measurement of height, weight (with
calculation of body mass index), measurement of FPG
and fasting lipid profile should be undertaken before
commencing atypical antipsychotic drugs.
Introduction
In the early 1990s the terminology used to classify diabetes
mellitus was based mainly on the approach used to treat the
condition; “insulin dependent” or “non‐insulin‐dependent”
diabetes. The term “other specific types” was used to
Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
describe a variety of forms of diabetes for which a cause
was clearly identified. Problems arose with this approach
because many patients with “non‐insulin‐dependent”
diabetes ended up being treated with insulin and having
the very cumbersome label of “insulin‐treated non‐insu­
lin‐dependent” diabetes. Current American Diabetes
Association (ADA) Guidelines recommend categorizing
diabetes into four groups: Type 1 diabetes (autoimmune,
insulin deficient), Type 2 diabetes (insulin resistance,
progressive loss of insulin secretory β‐cell function), ges­
tational diabetes (diagnosis of diabetes in an individual
who had no known diagnosis of diabetes prior to preg­
nancy), and specific types of diabetes due to other
causes [3] (Table 4.1). Currently, approximately 463 mil­
lion adults worldwide aged between 20 and 79 years have
diabetes (9.3% of individuals in this age bracket) [4]. This
is expected to rise to 578.4 million by 2030 and 700.2 mil­
lion by 2045 [4]. Approximately 90% have type 2 diabetes,
5–10% have type 1 diabetes and the remainder are spe­
cific types of diabetes due to other causes [5].
In this chapter, we explore the specific types of diabetes
due to other causes (often referred to as secondary diabe­
tes). Recognition of these specific types of diabetes is
important as diagnosis impacts on clinical care decisions.
Correct diagnosis reduces the risk of complications such as
diabetic ketoacidosis (DKA), hyperosmolar hyperglycemic
syndrome (HHS), or the unnecessary initiation of insulin
therapy (which can have a profound impact on quality of
45
 46 Prediabetes and the Diagnosis of Diabetes

TABLE 4.1  Types of diabetes.

Type of diabetes

Type 1 Diabetes Mellitus Immune‐mediated Endocrinopathies Acromegaly

Idiopathic Cushing’s syndrome

Type 2 Diabetes Mellitus Hyperthyroidism
Gestational Diabetes Mellitus Glucagonoma
Specific types of diabetes due to Phaeochromocytoma
other causes
Genetic defects of β‐cell function MODY3 (HNF‐1α) Somatostatinoma
MODY1 (HNF‐4α) Hyperaldosteronism
MODY2 (GCK) Others
MODY5 (HNF‐1β) Drug or chemical‐induced Immune checkpoint
inhibitors
MODY4 (IPF‐1) Glucocorticoids
MODY6 (ND1) Thyroid hormone
Other rarer forms of MODY Diazoxide
Transient neonatal diabetes Statin
Permanent neonatal diabetes Β‐adrenergic agonists
Mitochondrial diabetes Others
Other Infections Congenital rubella
Genetic defects in insulin action Type A insulin resistance Cytomegalovirus
Leprechaunism Other
Rabson–Mendenhall syndrome Other genetic syndromes Down’s syndrome
Lipoatrophic diabetes Klinefelter syndrome
Others Turner syndrome
Diseases of the exocrine pancreas Pancreatitis Wolfram syndrome
Trauma/pancreatectomy Friedreich ataxia
Neoplasia Myotonic dystrophy
Cystic fibrosis Other
Haemochromatosis
Other

life) particularly those with some types of monogenic dia­
betes. There is a lack of consensus among clinicians on
when and how to test for other causes of diabetes. As well
as traditional forms of secondary diabetes, we will discuss
newer genetic syndromes that have been elucidated in
recent years and new therapies (immune checkpoint inhib­
itors (ICIs)) which can contribute to the onset of diabetes.
Although these conditions may not be curable their
r­ ecognition can have a profound effect (therapeutic or oth­
erwise) on an individual patient and/or their family
members.
Specific types of diabetes due to other causes can be
identified in many ways. In some cases, patients present
with a new diagnosis of diabetes and the challenge is to find
the underlying cause. To delve beyond the most common
diagnoses of type 1 and type 2 diabetes can be challenging
 How Should Secondary Causes of Diabetes Be Excluded?
for a physician due to the vast number of other causes and
the fact that some are exceedingly rare. On the other hand,
patients can present with a diagnosis which is known to be
associated with (cystic fibrosis (CF)) or cause diabetes
(post‐pancreatectomy) where the challenge for the physi­
cian is to diagnose or screen for diabetes. It is worth con­
trasting the challenge of searching for secondary forms of
hypertension with that of searching for other causes of dia­
betes. In a hypertensive cohort, there are certain “red flags”
that most clinicians are alert to which may indicate an
underlying secondary cause. These include a young age at
onset of hypertension, difficult to control blood pressure
(e.g. requiring ≥ 3 antihypertensive drugs), the presence of
certain clinical (e.g. paroxysmal symptoms) or biochemical
(e.g. hypokalemia) markers at first presentation. Unfortuna­
tely, these “red flags” do not easily translate into the diabe­
tes setting. Onset of diabetes at a young age is not unusual
and is likely to be classified as type 1 diabetes if the patient
is lean or as type 2 if the patient is overweight. The latter is
increasingly recognized as a feature of the obesity epidemic
and is especially common among ethnic minority groups.
Similarly, diabetes that is difficult to control with tablets
would not typically suggest an underlying cause but rather
it would lead to the earlier introduction of insulin. The
“failure” of oral anti‐hyperglycemic agents is not unusual
(albeit due to poor compliance or relative insulin defi­
ciency) and does not represent a “red flag” for secondary
causes. It is in the clinical domain that the similarities
between approaches to identifying secondary hypertension
and secondary diabetes hold up. Indeed, several endo­
crinopathies such as Cushing’s syndrome or acromegaly
can lead to both diabetes and hypertension. Because these
conditions are so rare, clinical acumen is the most impor­
tant factor in identifying other causes of diabetes. A poten­
tial portent for the onset of diabetes is the commencement
of new therapies such as glucocorticoids or ICIs. In this
chapter we will address the question “should we be under­
taking more systematic case finding among our patients
with diabetes?”
Screening for a disease at population level is a complex
and expensive endeavor. Public Health England (PHE) for
example provide guidance on conditions for which screen­
ing programs should or should not be developed. The
criteria by which the PHE appraises the viability, effec­
tiveness and appropriateness of a screening program are
47
c­ omprehensive [6]. They include questions relating to the
condition (the condition should be an important public
health problem, does the natural history of the condition
include progression from latent to declared disease?), the
test (is the test simple, safe, validated and precise? Are there
appropriate cut‐off levels for the test?), the intervention (is
there evidence that early treatment in those who have been
screened leads to better outcomes than for those who
receive usual care?) and the screening program itself (is
there randomized controlled trial evidence that the screen­
ing program itself is cost effective and reduces morbidity or
mortality?)  [6]. At present, only 5 of 39 adult conditions
have been considered by the UK National Screening
Committee (NSC) undergo widespread screening (breast,
bowel and cervical cancer, diabetic retinopathy, and
abdominal aortic aneurysm  [7]). Diabetes itself has not
been recommended for population‐wide screening by the
UK NSC, but the United States Preventive Services Task
Force (USPSTF) recommends screening for “abnormal
blood glucose as part of cardiovascular risk assessment in
adults aged 40–70 years who are overweight or obese” [8].
Although the exercise of searching for other forms of dia­
betes is not, strictly speaking, screening (which is searching
for disease in a healthy population) it is a very similar pro­
cess. Table  4.2 summarizes the main secondary forms of
diabetes mellitus and screening methods used to detect
them. Prior to discussing the identification of other forms
of diabetes, it is worth reminding ourselves of the current
ADA criteria for diagnosis of diabetes [3]:
●● Fasting (no caloric intake for ≥ 8 h) plasma glucose ≥
7.0 mmol/L.
●● 2‐h plasma glucose ≥ 11.1 mmol/L during an oral glu­
cose tolerance test (OGTT) performed in accordance
with WHO criteria.
●● HbA1C ≥ 48  mmol/mol (carried out in a laboratory
using a method that is NGSP certified and standardized
to the Diabetes Control and Complications Trial
(DCCT) assay).
●● In a patient with classic symptoms of hyperglycemia or
hyperglycemic crisis, a random plasma glucose ≥
11.1 mmol/L.
In the absence of classical clinical symptoms such as
hyperglycemic crisis, classic symptoms of hyperglyce­
mia or random plasma glucose ≥ 11.1 mmol/L diagnosis
 48 Prediabetes and the Diagnosis of Diabetes

TABLE 4.2  Overview of secondary diabetes due to pancreatic disease and hormone excess.

Evidence for benefit Screening for DM

Screening for condition from population‐wide among individuals
Condition among people with DM screening with the condition

Pancreatic Hereditary Serum ferritin ++ OGTT/FG

disease haemochromatosis Transferrin saturation
Genotyping
CFRD N/A + Annual OGTT
Pancreatic cancer Abdominal CT scan +/‐ EUS − OGTT

Hormone Cushing’s 24hour UFC − OGTT

excess 1 mg DST
Late night salivary cortisol
Acromegaly IGF‐1 +/‐ OGTT with growth − OGTT
hormone (70)
Phaeochromocytoma Plasma metanephrines sampled − OGTT
following 30 minutes supine
rest (142‐144)

EUS: Endoscopic Ultrasound; DST: Dexamethasone Suppression Test; OGTT: Oral Glucose Tolerance Test; IGF‐1: Insulin‐like Growth
Factor; FG: Fasting Glucose
Evidence for benefit from population screening: ++ Very Strong Evidence and improvement of dysglycemia with treatment; +
Strong Evidence; − Little Evidence

requires two abnormal test results from the same sample
or in two separate test samples [3].
Hereditary haemochromatosis
Hereditary haemochromatosis (HH) is a genetic disorder
characterized by iron overload  [9]. It is associated with
increased and unregulated absorption of iron by entero­
cytes in the gastrointestinal tract due a deficiency of hepci­
din [9]. HH is typically inherited as an autosomal recessive
trait due to mutations such as C282Y, H63D and S65C in
the HFE gene. Affected individuals may develop excessive
deposition of iron in the parenchymal cells of certain
organs including liver, pancreas, skin and heart. Clinical

Clini c a l Vi g n e tt e :
A 43‐year‐old Caucasian male presented with a 3‐month history of polydipsia, polyuria, 4 kg unintentional weight loss, fatigue,
and lethargy. Random blood glucose was 16 mmol/L. He was diagnosed with haemochromatosis 12 years prior, at which time
his serum ferritin was > 4000 ng/mL and hepatic fibrosis was found on liver biopsy. Regular phlebotomy was initiated but
attendance was erratic in recent years. Medication included methotrexate and folic acid for haemochromatosis‐related arthritis.
The patient drank 24 units of alcohol per week. Normal libido and erectile function with no cardiac symptoms were noted.
On examination he had a bronzed appearance and no stigmata of chronic liver disease. Body mass index was 26 kg/m2
and blood pressure was 135/87 mmHg. Cardiovascular, respiratory, and gastrointestinal system examinations were normal.
Ankles were swollen over the lateral malleoli with scars from previous arthroscopies. Normal peripheral pulses and no loss of
vibration sensation were noted. Biochemical investigations revealed a HbA1c of 133 mmol/mol, normal liver chemistries,
normal alpha‐fetoprotein, serum ferritin 1340 ng/mL, transferrin saturation 97% and normal serum testosterone and
gonadotrophins. He was instructed in self‐monitoring of blood glucose and reported home readings that were consistently
> 15 mmol/L. A multiple daily injection regimen with insulin glargine and aspart was commenced. By attending for regular
venesection, he maintained his serum ferritin < 50 ng/mL
 How Should Secondary Causes of Diabetes Be Excluded?
manifestations include deranged liver enzymes, hepatic
fibrosis/cirrhosis, cardiomyopathy, hyperglycemia, diabe­
tes mellitus and a bronzed appearance on the skin. Joint
pains, fatigue, reduced libido, and erectile dysfunction
are   other potential problems. This vignette illustrates a
typical  patient who developed diabetes secondary to
haemochromatosis.
Natural history of haemochromatosis
Historically, HH often presented with end‐organ damage
due to increased levels of iron deposition in tissue. Although
the genetic defect is present from birth HH rarely presents
< 40 years [10]. In women, signs and symptoms present later
as menstruation can reduce iron accumulation prior to men­
opause. Factors such as excessive alcohol consumption lead
to earlier presentation. Prior to the identification of the HFE
gene, confirmation of the diagnosis was undertaken by doc­
umenting excessive iron stores on biochemistry and measur­
ing hepatic or bone marrow iron levels by quantitative
methods. Initial evaluation of suspected cases currently
involves laboratory measurement of serum iron, ferritin and
total iron binding capacity and estimation of the transferrin
saturation. Due to ease of access to these laboratory tests,
patients are typically asymptomatic at presentation.
Confirmation of the diagnosis involves genetic testing for
the commonest mutations (C282Y, H63D, and often S65C).
The frequency of homozygosity for the C282Y mutation
(commonest mutation) is 4.4 per 1000 in Caucasian popula­
tions in the US. It is less common in Hispanic (0.27 per
1000), African American (0.14 per 1000) and Asian
American (< 0.001 per 1000) populations [11]. Not everyone
homozygous for the C282Y mutation develops the full‐
blown clinical syndrome  [12]. Among those with C282Y
homozygosity 38–50% develop iron overload and only
10–33% develop HH associated morbidity [13].
Historically, untreated HH was a cause of premature mor­
bidity and mortality (due to liver cirrhosis, diabetes, conges­
tive cardiac failure.  .  .). With the advent of phlebotomy,
outcomes depend primarily on degree of iron overload and
sustained aggressive treatment if indicated. In patients with
diagnosed HH (C282Y homozygotes), only those with a
serum ferritin> 2000 ug/L had increased mortality (primar­
ily as a result of liver disease) [14]. The goal of phlebotomy is
to achieve a serum ferritin level < 50 ng/ml. This may reverse
49
some of the end‐organ damage particularly if phlebotomy is
initiated early in the course of the disease. Referred to as
“bronze diabetes”, the prevalence of diabetes in patients with
HH is estimated at between 13–23% [15].
Diabetes in HH results from β‐cell dysfunction (iron
deposition and hepcidin expression in pancreatic islets)
and decreased insulin secretion  [16]. In patients with
C282Y HH, there is an association between the diagnosis
of diabetes and a higher fibrosis stage (independent of
gender, alcohol, steatosis and hepatic iron overload) [15].
Interestingly, diabetes itself does not improve with serial
phlebotomy [17]. There is a lack of evidence on the effect
of phlebotomy on preventing the onset of diabetes [17].
The case for screening
HH appears to be a perfect condition for screening: insidi­
ous onset over many years, good biochemical and genetic
tests available for case detection, a safe treatment associ­
ated with benefit. Despite these considerations,
USPSTF [13, 18], UK NSC [19], European Association for
the Study of the Liver (EASL)  [20] and American
Association for the Study of Liver Diseases [21] do not rec­
ommend population‐wide screening for HH. The USPSTF
based this decision on:
1 Full‐blown clinical disease is rare in the general
population.
2 Disease penetrance is low among those with a high‐risk
genotype.
3 Lack of evidence that early phlebotomy in the screen‐
detected patient provides additional benefit over phle­
botomy in clinically detected patients.
4 Potential for harm (anxiety, denial of life insurance, dis­
crimination through health insurance) through labeling
of homozygous individuals as “diseased” or “high risk”
despite the possibility that they may never develop clini­
cal disease [22].
The UK NSC currently recommends further research to
determine which of those with a genetic susceptibility for
HH with normal iron stores and which of those with
genetic susceptibility for HH with iron overload but no evi­
dence of end organ damage will eventually develop the full‐
blown syndrome. Screening of family members (such as
first‐degree relatives) of patients with a diagnosis of HH is
 50 Prediabetes and the Diagnosis of Diabetes
recommended  [23]. Studies have shown that often
homozygous relatives have biochemical and clinical evi­
dence of disease [24, 25].
The evidence for screening for HH among individuals
with pre‐existing diabetes is lacking. There are conflicting
reports as to whether there is an association between dia­
betes and a genetic susceptibility to HH. This is due to the
fact that diabetes is prevalent in the general population and
genetic susceptibility to HH does not always lead to iron
accumulation [17]. In our practice (Ireland has the highest
genotype prevalence reported), we measure iron stores in
patients with newly diagnosed diabetes. This is not a popu­
lation‐wide approach to screening (not all patients with
newly diagnosed diabetes will be referred to hospital). It is
common for iron stores to be measured in patients with
diabetes and another potential manifestation of HH (such
as abnormal liver function tests or a bronzed appearance
on the skin). There is no current formal recommendation
to screen for HH in patients with newly diagnosed diabe­
tes [17] and further research is required. When measuring
iron stores, it is important to remember that serum ferritin
may be elevated as an acute phase reactant (especially in
patients with newly diagnosed diabetes) and may not
reflect iron overload. A combination of serum ferritin and
transferrin saturation is the preferred method of biochemi­
cal screening.
Screening HH patients for diabetes is complicated;
HbA1c is likely to be underestimated following venesec­
tion. In those undergoing active venesection, we recom­
mend the use of fasting plasma glucose (FPG)/OGTT to
diagnosis diabetes. Considering the prevalence of diabetes
among those with HH, it is sensible that all patients with
HH are screened for diabetes at time of diagnosis.
Thereafter, provided they have good HH control, the crite­
ria for testing for diabetes or prediabetes in asymptomatic
adults outlined by the ADA could be adhered to [28].
So where does all of this leave the practicing diabetes
clinician? In the absence of guidance from the ADA, the
Clinical Practice Guideline of the American College of
Gastroenterology (ACG) provides sensible guidance on
screening for HH in routine clinical practice  [29]. It
emphasizes the measurement of serum iron, transferrin
saturation and serum ferritin as a first step and targets rela­
tives of individuals known to have the genetic disorder, as
well as patients with symptoms or manifestations linked to
the clinical disorder. A modified ACG algorithm is repro­
duced in Figure 4.1.
Cystic fibrosis‐related diabetes
Cystic fibrosis (CF) is the most common life shortening
autosomal recessive disorder seen in Caucasian popula­
tions. The prevalence of CF among those with Northern
European ancestry is 1 in 2500–3000 live births with a car­
rier frequency of 1 in 25 [30]. It results from a defect in the
CF gene which encodes a protein called cystic fibrosis
transmembrane regulator (CFTR). The genetic defect leads
to disordered ion transport which in turn affects the func­
tion of sweat glands, the exocrine pancreas, the gastrointes­
tinal and respiratory tracts. The classical clinical picture is
recurrent pulmonary infections, exocrine pancreatic insuf­
ficiency, malnutrition, and premature death. In the past
2–3 decades, improvements in the pulmonary, antimicro­
bial, and nutritional management means that patients with
CF are now living into their 3rd, 4th, or 5th decades. Cystic
fibrosis‐related diabetes (CFRD) is the commonest comor­
bidity (20% adolescents; 40–50% adults) [31].
Natural history of CFRD
The primary pathogenetic defect in CFRD is insulin defi­
ciency resulting from pancreatic exocrine damage and
pancreatic fibrosis. Glucagon deficiency also occurs and
may explain why ketoacidosis is seldom seen in CFRD. It is
uncommon for CFRD to appear in childhood. Varying
degrees of insulin insufficiency occur as a result of acute
and chronic illness  [32, 33]. A single‐center North
American study that undertakes annual OGTT in its
patients with CF reported CFRD rates of 2% in children,
19% in adolescents, 40% in individuals in their 20s, and
45–50% in those aged ≥ 30 years [31]. In the 30–39 years
age group, women with CFRD outnumbered men but typi­
cally there was no gender difference in prevalence. CFRD is
associated with reduced survival and poorer prognosis in
females compared to males [33].
A comparison with other forms of diabetes
Although CFRD shares many similarities with Type 1 and
Type 2 diabetes, it represents a distinct form of diabetes.
 How Should Secondary Causes of Diabetes Be Excluded? 51

Target Population

Adult 1st degree

Step 1 symptomatic asymptomatic relative of HH

Fasting transferrin saturation

and serum ferritin

TS <45% and TS >45% and ferritin

normal ferritin elevated

No further iron Genotype

Step 2
evaluation

C282Y/C282Y
Compound heterozygote
C282Y/H63D
Heterozygote C282Y or Age < 40 years Age > 40 years
Ferritin<1000 and/or Ferritin
non-C282Y
and Normal >1000
or Elevated
ALT/AST ALT/AST
Exclude other liver or
Step 3 hematologic disease
Therapeutic Liver biopsy for HIC
± Liver biopsy Phlebotomy and histopathology

FIG 4.1  Proposed algorithm for screening and treatment of haemochromatosis in patients with diabetes [23, 29]. TS: Transferrin
Saturation; HIC: Hepatic Iron Concentration.

The presence of CF does not out rule the possibility of type
1 diabetes rather than CFRD. DKA is uncommon in
patients with CFRD and the presence of DKA should
prompt a diabetes autoantibody screen for type 1 diabe­
tes  [32]. Table  4.3 compares features of CFRD with the
2 major forms of diabetes seen in the population at large.
Although insulin deficiency is the primary defect in
patients with CFRD, insulin resistance is becoming more
common as median age of survival increases. This is espe­
cially true during periods of intercurrent illness (e.g. res­
piratory infections, steroid therapy). Autoimmunity does
not appear to be a factor in the pathogenesis of CFRD. Low
body mass index is associated with premature mortality in
CF [34] and this can lead to conflict in the dietary manage­
ment of patients with CFRD. While calorie restriction is
usually recommended in the management of patients with
type 2 diabetes, this is not the case in CFRD. Instead a more
liberal individualized diet is recommended and the glu­
cose‐lowering therapy (usually insulin) is adjusted to
achieve near normal plasma glucose levels  [32]. Current
dogma is that patients with CFRD should be treated with
insulin therapy and that oral hypoglycemic agents may not
be as effective in optimizing nutritional and metabolic out­
comes  [32]. As well as achieving and maintaining good
glycemic control the anabolic effects of insulin are also
believed to be advantageous to patients with CFRD. While
the primary concerns are the pulmonary and nutritional
consequences [32], similar to patients with type 1 or type 2
diabetes, patients should undergo annual monitoring for
the microvascular complications of diabetes. This should
begin 5 years after initial diagnosis or, if the date of diagno­
sis is not known, when fasting hyperglycemia was first
 52 Prediabetes and the Diagnosis of Diabetes

TABLE 4.3  Clinical features of T1DM, T2DM and CFRD. Adapted from [49].

T1DM T2DM CFRD

BMI Usually normal Usually increased Usually decreased

Presentation Acute Sub‐clinical Sub‐clinical
Age of onset Childhood and Adulthood Increasing incidence with age
adolescence
Autoantibodies Yes No Probably No
β cell function Severely reduced/absent Reduced Severely reduced
Insulin action Usually normal Severely impaired Somewhat impaired*
Treatment Insulin Diet, glucose lowering Insulin and nutrition
medication including insulin
Microvascular Yes Yes Reduced
complications
Macrovascular Yes Yes No
complications
Cause of death Cardiovascular disease Cardiovascular disease Pulmonary disease
and nephropathy

BMI: body mass index; CFRD: Cystic fibrosis‐related diabetes; T1DM: Type 1 Diabetes Mellitus; T2DM: Type 2 Diabetes Mellitus
*
 Insulin sensitivity reduced during acute illness

observed  [32]. Microvascular complications do not typi­
cally occur within the first 5 years of diagnosis but compli­
cations such as microalbuminuria have been observed in
4–21%  [35–37] and diabetic retinopathy in 10–27%  [35–
38]. Macrovascular complications are infrequent and while
most patients with diabetes die as a result of macrovascular
disease, patients with CFRD are most likely to succumb to
pulmonary complications of their CF [39].
The case for screening
Patients with CFRD often have no clear symptoms of dia­
betes. Early diagnosis and management of CFRD is essen­
tial as the presence of CFRD is associated with severity of
lung disease  [40], more difficulty in maintaining weight
and increased mortality. Screening for CFRD is not a diag­
nostic exercise but rather an exercise in case finding, initi­
ating earlier treatment and (ideally) achieving improved
outcomes. Patients with a new diagnosis of diabetes are not
screened for CF; it is extremely unlikely that patients with
CF develop diabetes as the first manifestation of their dis­
ease. Case finding of patients with CF in a general diabetes
clinic will not be discussed. Instead, we will examine the
case for screening patients with CF for diabetes. Few
patients with CF have truly normal glucose status  [41].
Glucose tolerance status varies across individuals with CF
and within an individual patient. For example, at times of
severe pulmonary infection or during use of corticosteroid
therapy, glucose intolerance can develop but may not be
permanent. CFRD is part of a spectrum of glucose toler­
ance in patients with CF:
●● Normal glucose tolerance (FPG < 5.6  mmol/l; 2‐hour
OGTT glucose < 7.8 mmol/l)
●● Indeterminate glucose tolerance (INDET) (FPG <
5.6  mmol/l; 1‐hour OGTT glucose ≥ 11.1  mmol/l; 2‐
hour OGTT glucose < 7.8 mmol/l)
●● Impaired glucose tolerance (FPG < 5.6 mmol/l; 2‐hour
OGTT glucose 7.8–11.1 mmol/l)
●● CFRD FH− (FPG < 7.0 mmol/l; 2‐hour OGTT glucose ≥
11.1 mmol/l)
●● CFRD FH+ (FPG ≥ 7.0 mmol/l; 2‐hour OGTT glucose ≥
11.1 mmol/l)
●● Impaired fasting glucose (FPG 5.6–6.9 mmol/l; 2‐hour
OGTT glucose N/A) [42].
The ADA and the CF foundation recommend the 2‐
hour 75 g OGTT as the screening test of choice [32]. This
should be performed when the patient is well (≥ 6 weeks
after an acute exacerbation) [32] and fasting for ≥ 8 hours
having consumed at least 600 kcal or 150 g of carbohydrate
per day for the preceding 3  days  [41]. OGTTs should be
 How Should Secondary Causes of Diabetes Be Excluded? 53

performed on 2 separate days in the absence of classical Pancreatic‐cancer‐associated diabetes

symptoms of hyperglycemia such as polyuria and polydip­
Diabetes is a risk factor for pancreatic cancer [50]. Up to
sia. An abnormal FPG or elevated HbA1c can be used to
80% of patients with a new diagnosis of pancreatic cancer
confirm the diagnosis, but if these are normal the OGTT
have either hyperglycemia or diabetes [51]. Pancreatic can­
must be repeated [41]. Annual screening for CFRD is rec­
cer can present as a new diagnosis of diabetes [52]. These
ommended from 10 years in all patients without a known
associations between pancreatic cancer and diabetes have
diagnosis of CFRD [3, 32]. HbA1c is not recommended as
long been recognized but whether the pancreatic cancer
a screening tool as it is not sensitive enough [32].
causes diabetes or vice versa has been difficult to ascer­
During acute illness or if requiring glucocorticoid ther­
tain [53]. The association is likely bi‐directional. Diabetes
apy, glucose levels should be monitored both fasting and 2
due to pancreatic cancer is likely mediated through
hours postprandially for at least the first 48 hours of treat­
humoral factors produced by the tumor, insulin resistance
ment. CFRD can manifest during initiation of continuous
or due to destruction of β‐cells  [54]. Interestingly, in a
enteral feeling or gastrostomy tube feeding (diagnosis
group of patients with pancreatic cancer who underwent
based on mid or immediate post feeding values of ≥
pancreatico‐duodenectomy, 17/30 (57%) of new‐onset dia­
11.1  mmol/l confirmed on two separate nights) and it is
betes resolved after tumor resection  [54]. It is worth
important to check for CFRD at these times.
remembering that in patients who do not have diabetes
Insulin therapy is a safe and effective therapy in reversing
pre‐operatively there is a risk of developing de‐novo
chronic weight loss in patients with CF without fasting
diabetes post pancreatectomy. Among patients undergoing
hyperglycemia [44]. Initiation of insulin glargine is associ­
distal pancreatectomy between 14–39% developed a de‐
ated with decreased progression of lung disease in patients
novo diagnosis of diabetes post‐operatively [55].
with CF and fasting hyperglycemia  [45]. In patients with
Diabetes is associated with an almost 2‐fold increased
CFRD, insulin improved nutritional status, temporarily
risk of pancreatic cancer  [56, 57]. Approximately 1% of
improved lung function  [46] and improved protein
patients with diabetes diagnosed at ≥ 50 years of age will
synthesis [47]. Although current clinical practice guidelines
have a diagnosis of pancreatic cancer within the first 3
recommend insulin and suggest that oral hypoglycemics
years of meeting the diagnostic criteria for diabetes [58].
may not be the optimal therapy, a systematic review has sug­
Diabetes itself is prevalent in the age range where pancre­
gested that there is no clear difference between the two [43].
atic cancer is most common [59]. As the 5‐year survival of
Randomized controlled trials are required to evaluate this.
pancreatic cancer is so poor [60], new‐onset diabetes may
More intensive monitoring of blood glucose through
be a marker of early stage cancer  [51] and could poten­
use of continuous glucose sensors should be considered if
tially lead to earlier recognition with a greater chance of
hyperglycemic symptoms develop or at times of acute pul­
cure. In the diabetes literature it has been suggested that
monary infection. It is unlikely that a clinical trial compar­
age ≥ 50 years, a lean body habitus and lack of a family
ing treatment versus no treatment of screen detected CFRD
history of type 2 diabetes [61] are clinical indicators of the
will be undertaken. In the absence of such a trial the next
potential for underlying pancreatic cancer. However, this
best level of evidence comes from well described case
clinical impression has not been borne out by careful
series. One such case series was recently reported from the
comparison of the characteristics of patients with pancre­
University of Minnesota, a unit that has experience in the
atic‐cancer‐associated diabetes and type 2 diabetes. A
management of CFRD. The report highlights a temporal
group at the Mayo Clinic has shown that a similar per­
trend over the past 3 decades towards improved outcomes
centage of patients in both groups are overweight and
and greater longevity in patients with CFRD  [31]. The
obese and have a positive family history for diabetes. In
authors attribute this to universal screening (which they
fact, obesity itself is a risk factor for pancreatic cancer [62];
undertake from age 6), earlier recognition of CFRD and
while effective treatment of diabetes is associated with a
more aggressive treatment with insulin and nutritional
reduced risk [53].
strategies.
 54 Prediabetes and the Diagnosis of Diabetes
For new onset diabetes mellitus to be useful as a poten­
tial indicator of underlying pancreatic cancer, it will need
to be combined with a tumor biomarker. As yet, no such
biomarker has been identified. Adrenomedullin has been
associated with increased aggressiveness of pancreatic can­
cer [63] and may contribute to pancreatic associated diabe­
tes [64]. Studies are ongoing to determine its potential as a
marker of pancreatic cancer in patients with new‐onset
diabetes [65]. The presence of symptoms such as jaundice,
abdominal pain, anorexia, and weight loss is problematic
because by the time the patient develops these symptoms
the tumor is almost always unresectable. Until a suitable
biomarker is identified, widespread use of abdominal
imaging among older patients with new onset diabetes is
not recommended as a method of searching for occult pan­
creatic cancer. The USPSTF does not recommend screen­
ing for pancreatic cancer among asymptomatic patients in
the general population. This advice also applies to individ­
uals who have risk factors such as new‐onset diabetes and
pre‐existing diabetes [66].
Diabetes due to hormone excess
Several disorders resulting in hormone excess can cause
diabetes including Cushing’s syndrome (glucocorticoid
excess), acromegaly (growth hormone excess), hyperaldo­
steronism (aldosterone excess), hyperthyroidism, pheo­
chromocytoma/ paraganglioma (PPGL) (catecholamine
excess), glucagonoma and somatostatinoma. Diabetes can
be a feature of autoimmune polyglandular syndrome (more
common in type 2 than type 1) and POEMS syndrome
(polyneuropathy, organomegaly, endocrinopathy, mono­
clonal gammopathy, skin changes). The benefit of early
identification and management of these conditions in a
patient with concomitant diabetes is that treatment of the
endocrinopathy may lead to cure or amelioration of the
glucose intolerance. Furthermore, many of these endo­
crinopathies are associated with other metabolic complica­
tions such as hypertension [67] which can lead to worsening
of the micro‐ and macro‐ vascular complications of diabe­
tes. In the case of acromegaly and pheochromocytoma rou­
tine screening among individuals with diabetes is not
justified since both conditions are extremely rare (acro­
megaly annual incidence range from 0.2–1.1 cases/100 000
people  [68]; PPGL annual incidence 0.8/100 000 person
years [69]). Screening could be associated with high rates
of false positive test results with resultant increased distress
and cost from unnecessary testing. Instead case finding
should be undertaken when additional clinical features
consistent with the endocrinopathy are present.
In the case of acromegaly, these would include enlarge­
ment of the hands and feet, characteristic changes in facial
appearance, headache, visual disturbance, and hyperhidro­
sis. These changes are often subtle and take many years to
manifest, they may not be apparent to the patient or even to
family members. A comparison of the patient’s current
appearance with that obtained from old photographs and
rings or shoes getting too small can be informative. The
Acromegaly: Endocrine Society Clinical Practice Guideline
recommends measuring IGF‐1 in patients without the typ­
ical manifestations of acromegaly but who have several
linked conditions such as obstructive sleep apnea syn­
drome, type 2 diabetes, hypertension, hyperhidrosis, carpal
tunnel syndrome or severe arthritis [70]. All patients who
present with acromegaly should be evaluated for diabetes.
Diabetes is found in 11–38% of patients with acromegaly
– excess growth hormone leads to insulin resistance [70].
The classical features of PPGL include paroxysmal
hypertension, palpitations, headaches, pallor, perspiration,
and pain. Table  4.3 includes initial testing that can be
undertaken if the clinical suspicion of growth hormone or
catecholamine excess is sufficiently high to justify a case
finding exercise. One caveat with acromegaly screening is
that poorly controlled diabetes is associated with elevated
levels of growth hormone  [71] and normal or low
IGF‐1 [72].
Cushing’s syndrome merits mention because it is the
commonest endocrine disorder associated with diabetes.
This is due to the widespread use of exogenous glucocorti­
coids (iatrogenic Cushing’s syndrome) in the treatment of a
variety of inflammatory and malignant diseases. A detailed
medication history should highlight the use of exogenous
steroids. Glucocorticoids may be contained in topical, over
the counter or herbal preparations or may be taken sur­
reptitiously. Furthermore, use of some steroids (such as
medroxyprogesterone acetate [73]) that are not primarily
glucocorticoids has been associated with Cushing’s syn­
drome. The mechanism whereby glucocorticoid excess
 How Should Secondary Causes of Diabetes Be Excluded?
causes diabetes is multifactorial involving increased insulin
resistance and beta‐cell dysfunction. If Cushing’s syndrome
from exogenous glucocorticoid use is excluded, then
(endogenous) Cushing’s syndrome is rare (3 new cases of
endogenous Cushing’s syndrome per million per year [74]).
However, among populations of patients with diabetes who
have undergone screening for Cushing’s syndrome preva­
lence as high as 3–5% has been reported  [75, 76]. In a
cross‐sectional prospective study of patients attending 24
diabetes clinics across Italy Terzolo et al. found no evidence
to support screening patients with type 2 diabetes for
Cushing’s syndrome. However, this group concluded that
Cushing’s syndrome was more common than previously
thought (5/813 had a definitive diagnosis of Cushing’s syn­
drome). These results support case‐finding in those
patients with poorly controlled diabetes and hypertension
despite best medical therapy [77]. Budyal et al. conducted a
single‐center study to determine if patients with type 2 dia­
betes should be screened for subclinical Cushing’s syn­
drome (unregulated cortisol production without the
classical signs and symptoms of Cushing’s syndrome).
They found that out of a cohort of 993 participants no one
had subclinical Cushing’s and thus screening should only
take place if mandated by clinical suspicion  [78]. Some
authors have previously advocated screening all patients
with newly diagnosed diabetes (in a hospital setting) [75]
while others have suggested that screening for Cushing’s
should be limited to patients with poorly controlled diabe­
tes. While diabetes is prevalent among patients with
Cushing’s syndrome, Cushing’s syndrome remains rela­
tively rare among patients with diabetes. Recent Endocrine
Society Guidelines support testing for Cushing’s syndrome
following a detailed drug history to exclude exogenous
steroids among groups of patients with certain “overlap
disorders”, i.e. conditions which although common in the
community can also be a feature of occult Cushing’s syn­
drome  [79]. These include osteoporosis, hypertension,
polycystic ovary syndrome and diabetes mellitus. The
guidelines state that Cushing’s is more likely when these
conditions occur at a young age. Another way of increasing
the pre‐test probability of Cushing’s is to limit screening to
those patients with diabetes and one or more of the clinical
features associated with tissue glucocorticoid excess. These
include (but are not limited to) facial plethora, easy bruis­
55
ing, proximal myopathy, and striae. The tests recom­
mended by the Endocrine Society are listed in Table  4.3.
The choice of which test(s) to use depends on several fac­
tors including concomitant medication use, local labora­
tory expertise as well as the index of clinical suspicion for
severe Cushing’s. Diagnostic criteria that suggest Cushing’s
syndrome are two or more 24‐hour urinary free cortisol
results above the upper limit of normal for the assay, a posi­
tive overnight 1‐mg dexamethasone suppression test (i.e.
an 08:00 cortisol > 50 nmol/L after administration of dexa­
methasone 1  mg at midnight) or two or more late‐night
salivary cortisol values above 4  nmol/L. Some authors
advocate using higher cut‐off levels as a way of reducing
the number of false positives. Current best evidence sug­
gests that screening for Cushing’s syndrome should take
place on a case by case basis with signs such as easy bruis­
ing, facial plethora, proximal myopathy, and striae seen as
“red flags”. All patients with Cushing’s syndrome should be
regularly screened for diabetes.
Post‐transplantation diabetes mellitus
(PTDM)
Post‐transplantation diabetes mellitus (PTDM) refers to
the recognition of diabetes in post‐transplant recipients.
This finding does not depend on the timing of the diagno­
sis or whether it was present but undiagnosed prior to
transplantation. New‐onset diabetes mellitus after trans­
plantation (NODAT) is the term used to describe those
patients who prior to transplantation did not have diabetes
and who develop new‐onset diabetes post transplantation.
NODAT does not include patients who likely had diabetes
pre‐transplant but were not diagnosed. Other terminology
used in this area includes “transplant associated hypergly­
cemia”. There are wide variations in the reported preva­
lence of PTDM due in part to varying definitions of PTDM,
follow‐up strategies over time, type of transplant, immuno­
suppressive regimen used and donor and recipient charac­
teristics  [80]. Hyperglycemia can occur early in the
post‐operative period (when the dose of immunosuppres­
sant therapy is high). This can regress to normoglycemia or
result in persistent long‐term hyperglycemia [81]. PTDM
has been reported in 4–25% of renal, 2.5–25% of liver,
4–40% of heart [82] and 25–40% of lung transplants [83].
 56 Prediabetes and the Diagnosis of Diabetes
Risk factors include increasing age, ethnicity, pre‐trans­
plant overweight or obesity, post‐transplant weight gain,
adult polycystic kidney disease, a strong family history of
diabetes, previous glucose intolerance (should be assessed
in all patients prior to transplantation) and a history of
cytomegalovirus or hepatitis C infection [84]. Hepatitis C
infection itself is associated with an increased risk of type 2
diabetes mellitus [85]. Eradication of hepatitis C infection
with direct‐acting antiviral agents has been shown to
improve diabetes control [86]. Many immunosuppressant
drugs including calcineurin inhibitors (tacrolimus > cyclo­
sporin) as well as corticosteroids are diabetogenic. The lat­
ter causes diabetes predominantly through increasing
insulin resistance, enhancing gluconeogenesis and lipolysis
and decreasing glucose uptake in skeletal muscle in patients
whose beta‐cell response is compromised [81]. Calcineurin
inhibitors likely contribute to diabetes through a combina­
tion of β‐cell apoptosis, impaired insulin secretion and
insulin resistance although further study is required to
fully understand the mechanism [87].
Clinical features
Weight gain is a common occurrence after transplanta­
tion [81] and the risk of developing PTDM increases by a
factor of 1.4 for every 10  kg increase in weight over
60 kg [88]. Contributors to weight gain are likely multifac­
torial and include steroid therapy and fewer dietary restric­
tions (for example no longer requiring a strict renal
diet)  [81]. Taken with the increasing insulin resistance
associated with the immunosuppressant medication, the
clinical features of PTDM resemble those of type 2
diabetes.
The case for screening
PTDM predicts graft failure/function  [89], mortality  [90,
91], infection and micro‐[92]/macro‐vascular [93] compli­
cations of diabetes. Early recognition of PTDM with the
potential to prevent the associated morbidity and mortality
has been advocated. The ADA advise that patients should
be screened when stable on immunosuppressive therapy
and no active infection [94]. There are no formal recom­
mendations regarding the exact frequency of screening for
PTDM. Some authors recommend measurement of FPG
weekly for the first 4  weeks, at 3, 5 and 12  months and
annually thereafter [95]. The OGTT is the diagnostic test
of choice  [94]. HbA1c can be difficult to interpret in the
immediate post‐transplant period especially in those with
renal disease as HbA1c readings are impacted by blood
loss, blood transfusion and iron replacement [81].
In the absence of randomized control trial evidence,
management of PTDM is like that of type 2 diabetes –
involving structured diabetes care and the use of oral anti‐
hyperglycemic medication and insulin. As for patients with
type 2 diabetes, drug‐specific and patient factors need to be
considered when selecting appropriate therapy including
the degree of dysglycemia, risk of hypoglycemia, weight
change, cardiovascular and renal effects, cost, and route of
administration. The ADA advise that irrespective of the
risk of NODAT, the immunosuppressive therapy associated
with the best outcomes and graft survival for the patient
should be used [94]. In addition, patients post transplanta­
tion may have renal or hepatic impairment, the severity of
which may be a contraindication or require dose adjust­
ment for certain medications (such as biguanides) and
consideration should be given to the interaction of immu­
nosuppressants and oral anti‐hyperglycemic agents  [96].
Modification of the immunosuppressant regimen can be
considered in certain cases (under appropriate supervision
and guidance, for example changing from tacrolimus to
ciclosporin or mycophenolate mofetil plus azathio­
prine [97]) must be balanced with the risk of graft rejec­
tion. Furthermore, the benefits of reducing corticosteroid
dosage and reducing or substituting calcineurin inhibitors
have not been proven fully in clinical trials. A screening
program should involve pre‐ and post‐transplant testing of
glucose tolerance. A knowledge of pre‐transplant risk fac­
tors is essential to characterize the subsequent risk of
PTDM.
Diabetes associated with atypical
antipsychotic drug use
Antipsychotic drugs are licensed for use in schizophrenia
and related psychotic disorders. Irrespective of antipsy­
chotic therapy, schizophrenia itself is associated with an
increased risk of developing diabetes, the cause of which is
multifactorial including both environmental and biological
factors such as sedentary lifestyle and poor diet [98] as well
 How Should Secondary Causes of Diabetes Be Excluded?
as disease and treatment specific effects [99]. First‐genera­
tion antipsychotics were useful for alleviating the positive
symptoms of psychosis such as hallucinations and delu­
sions. They are less efficacious at managing the negative
symptoms of psychosis which include withdrawal, poverty
of speech and cognitive problems. The introduction of sec­
ond‐generation (or atypical) antipsychotics was greeted
with enthusiasm as they were efficacious against both posi­
tive and negative psychotic symptoms. These agents have
less debilitating extrapyramidal side‐effects than the first‐
generation agents but have a stronger diabetogenic
effect  [100]. Second‐generation antipsychotics including
aripiprazole, asenapine, clozapine, iloperidone, olanzapine,
quetiapine, and risperidone have transformed the lives of
many patients with schizophrenia and have enabled many
individuals to live in the community as opposed to being
managed as inpatients in psychiatric institutions. These
medications have become popular with psychiatrists and
general practitioners and are now prescribed for many psy­
chiatric conditions other than those for which they are
licensed (attention‐deficit hyperactivity disorder, depres­
sion, eating disorders, insomnia, obsessive compulsive dis­
order. . .). Among nursing home residents for example (in
the 2004 National Nursing Home Survey), 86% of second‐
generation antipsychotics prescribed were for off‐label
indications  [101]. Side‐effects include agranulocytosis
associated with clozapine (very rare; regular monitoring of
the patient’s full blood count is mandatory) to the much
commoner metabolic disorders which we will discuss in
detail.
Natural history
Psychiatric disorders for which atypical antipsychotics are
prescribed are a risk factor for or a complication of diabe­
tes [102]. The underlying disorders for which the atypical
antipsychotic drugs are used (schizophrenia, other psy­
chotic disorders, major depression, etc.) are often associ­
ated with a sedentary lifestyle and poor dietary habits (less
fruit and vegetables consumption [103]). In patients with
severe mental illness, modification of risk factors associ­
ated with diabetes such as dietary and exercise interven­
tions can be difficult due to low mood, impaired judgement,
and/or disordered thinking. Drug‐naïve and drug‐free
patients with schizophrenia are more likely to have central
57
obesity than age and sex matched controls [104], a risk fac­
tor for the onset of type 2 diabetes  [105]. Patients with
schizophrenia who are prescribed second‐generation
antipsychotics compared to first generation antipsychotics
have a relative risk of having diabetes of 1.32 (95%CI 1.15–
1.51) [106]. A systematic review indicated that there is an
increased risk of type 2 diabetes among children and youth
(aged 6–24 years) prescribed antipsychotic drugs that
increased with cumulative dose [107].
Although high‐quality cohort studies have been diffi­
cult to undertake, it is generally accepted that these disor­
ders are associated with an increased risk of obesity,
dyslipidemia, diabetes, and resultant increased cardiovas­
cular morbidity. Because of the effect of the underlying dis­
ease, it is difficult to tease out the exact contribution of an
additional drug effect on top of this. Clinical experience
does suggest that many of the atypical antipsychotics are
associated with weight gain and metabolic abnormalities.
Between 15–72% of patients on antipsychotics have weight
gain during the acute or maintenance phase of treatment of
schizophrenia and affective disorders [99]. Atypical antip­
sychotics except for ziprasidone also appear to be associ­
ated with weight gain (clozapine followed by olanzapine
and quetiapine have the highest weight gain) [108]. Use of
these agents has been associated with profound metabolic
disturbance including DKA. In some case reports presenta­
tion with DKA is followed by subsequent recovery of β‐cell
function akin to the so‐called “Flatbush” or “ketosis‐prone
type 2 diabetes”  [109]. More commonly the metabolic
derangement is less profound and associated with weight
gain and dyslipidemia. As well as causing new diabetes
among individuals not previously recognized as having the
disease these drugs can also lead to worsening of pre‐exist­
ing diabetes and a need for intensification of anti‐hypergly­
cemic therapy. Co‐ordinated shared care plans should be
established by the psychiatry team and diabetes centers to
optimize patient care [99].
The case for screening
In consultation with the European Association for the
Study of Diabetes (EASD) and the European Society of
Cardiology (ESC), the European Psychiatric Association
(EPA) published a position statement outlining how
patients with serious mental illness have an increased risk
 58 Prediabetes and the Diagnosis of Diabetes
of cardiovascular disease and diabetes  [99]. Patients with
severe mental illnesses have worse physical health,
decreased life expectancy and a 2‐ to 3‐fold increased mor­
tality rate compared to the general population [99]. There
is a gradation of metabolic risk across drugs in the
class [110] with clozapine and olanzapine carrying a higher
risk of metabolic side‐effects than quetiapine and risperi­
done [110]. For example a metanalysis found that clozap­
ine (4.45 kg) and olanzapine (4.15 kg) had greater risk of
weight gain than quetiapine and risperidone (2.1 kg) while
drugs like aripiprazole, amisulpride, and ziprasidone had
limited effect on weight [111]. Causes for this weight gain
are multifactorial and poorly understood (increased appe­
tite and modified energy expenditure)  [99]. A recent
review concluded that it is unclear if antipsychotic medica­
tion exacerbates a predisposition to new onset of type 2
diabetes or whether there is a direct effect [113].
The EPA position statement called for increased coop­
eration and shared cared among the relevant healthcare
professionals to improve care for those who have a serious
mental illness  [99]. Management is often complicated by
lack of insight into the underlying conditions (both psychi­
atric and medical) and the patients’ inability to follow a
treatment program. All patients at initial presentation
should have a detailed medical history (including past
medical history, family history, smoking, exercise, dietary).
Clinical examination should include recording of height
and weight (and calculation of body mass index), waist cir­
cumference and blood pressure as well as biochemical
measurement of fasting glucose, HbA1c and lipid levels.
Although these are measured routinely in diabetes clinics
around the world they are not undertaken routinely in
Departments of Psychiatry. However, if we are to stem the
rise in metabolic disorders that would appear to be hap­
pening with increasing use of these drugs then this is an
important first step. The ADA recommends annual
screening for patients who are prescribed atypical antipsy­
chotic medications for prediabetes or diabetes in accord­
ance with ADA guidelines  [28]. In addition, among
treatment naïve patients with serious mental illness who
do not have diabetes at baseline the EPA recommends
repeat fasting glucose and lipids at week 6 and 12 post ini­
tiation of therapy [99].
Diabetes associated with HIV infection
and its treatment
With the advent of safe and effective antiretroviral therapy
(ART) over the past 40 years human immunodeficiency
virus (HIV) infection in the developed world has seen a
transformation from a disease associated with opportunis­
tic infection, AIDS and premature death to a chronic dis­
ease associated with good long‐term survival. Improved
life expectancy of patients with HIV and ART, has seen the
development of chronic disease complications in particular
cardiometabolic diseases such as diabetes and increased
cardiovascular event rates [115]. In the Multicenter AIDS
cohort study, the prevalence of diabetes was 14% among
411 HIV‐infected men using ART (4.7 cases per 100‐per­
son years) compared to 5% among 711 HIV seronegative
men (1.4 cases per 100‐person years) [116]. Major reported
risk factors for the development of dysglycemia in those
with HIV include family history of diabetes, increasing age,
hepatitis C infection, elevated BMI, fat redistribution and
medications such as indinavir and lopinavir/ritona­
vir [117]. The long‐term complications of diabetes can be
accentuated among patients with diabetes and HIV [117].
Similar to atypical antipsychotics, clinicians prescribing
these drugs need to be aware that the risk of metabolic
derangement varies between the different agents [118]. In
those with dysglycemia, it is important to consider if alter­
native ARTs with a less adverse glycemic profile could be
used while still effectively and safely managing HIV. That
said considerable controversy exists on the reported links
between ART and diabetes [119–121], which may be due
to variable standards/methods of screening for diabetes, to
heterogeneity between studies used in systematic
reviews  and meta‐analyses and to lack of long‐term
follow‐up [120].
For patients with HIV, the ADA recommends screening
for diabetes and prediabetes with a fasting glucose test
before starting antiretroviral therapy, at the time of switch­
ing antiretroviral therapy, and 3–6 months after starting or
switching antiretroviral therapy. If initial screening results
are normal, fasting glucose should be checked annu­
ally  [28]. HbA1c is not an appropriate screening tool in
patients with HIV as it underestimates the degree of dys­
glycemia by 1.6±0.2 mmol/L [122]. Kim et al. found that
 How Should Secondary Causes of Diabetes Be Excluded?
HbA1c is particularly impacted in those on nucleotide
analog reverse transcriptase inhibitors particularly patients
with macrocytosis or on abacavir [122].
Diabetes secondary to immunotherapy
The development of immunotherapy as a treatment for a
variety of cancers has revolutionized clinical outcomes
(improved prognosis) for these conditions  [134].
CTLA‐4  inhibitors such as Ipilimumab, PD‐1  inhibitors
such as Pembrolizumab and Nivolumab and PD‐L1 inhibi­
tors such as Atezolizumab, Avelumab, and Durvalumab in
addition to combination therapies (CTLA‐4 and PD‐1 inhi­
bition) such as Ipilimumab/Nivolumab are approved for
treatment in a broad spectrum of conditions from mela­
noma to renal cell cancer to Hodgkin’s lymphoma  [134].
The number of cancers which have been shown to benefit
from immunotherapy are increasing [135].
While other endocrinopathies such as hypothyroidism,
hyperthyroidism, and hypophysitis are more common, dia­
betes secondary to ICIs represents a rare but important
complication of immunotherapy; the first presentation can
be with life‐threatening DKA  [136]. Insulin‐dependent
diabetes occurs in up to 1% of patients who receive either a
PD‐1 or PD‐L1 inhibitors [137]. Reports suggest that time
to onset of diabetes from the first dose of the ICI varies
from 5–790 days (median 116 days) [138]. The number of
doses before the onset of diabetes varies from 1–24
(median 3) [138].
FPG should be measured before commencing ICIs and
at the start of each course of treatment for the first
6 months, every two courses for the next 6 months and as
clinical signs and symptoms dictate [139]. The Society for
Endocrinology recommends the measurement of FPG in
all unwell patients who are on ICIs [135, 140]. Furthermore,
they advise the management of DKA and HHS in accord­
ance with local guidelines and measurement of autoanti­
bodies and C‐peptide [135] and ICIs should be held until
the patient has recovered  [141]. Patients on ICIs often
require high‐dose steroids which in itself will make the
patient prone to developing steroid‐induced diabetes [135].
HbA1c is not suitable for diagnosing diabetes in the acute
setting, due to the often‐rapid onset of diabetes but in
59
addition to FPG should be used to monitor patients with
known type 1 or 2 diabetes on ICIs.
Monogenic forms of diabetes
C ase V ign et t e:
A 23‐year‐old Caucasian male was admitted to hospital
with acute urinary retention and renal failure. He
complained of progressive thirst, polyuria and nocturia
over the last 18 months. His past history was remarkable
for diabetes mellitus diagnosed 10 years previously. This
was managed with insulin aspart and insulin glargine. He
had poor eyesight since childhood which was attributed to
Leber’s optic atrophy. Family history: mother – type 2
diabetes mellitus (on insulin therapy), gestational diabetes;
father – hypercholesterolemia; grandmother – diabetes
mellitus, died aged 73 years; three sisters alive and well
with no history of diabetes.
On examination he was very hard of hearing. Blood
pressure was 170/95 mmHg. Body mass index was
elevated at 29.8 kg/m2. He had markedly decreased visual
acuity. Dilated fundoscopy revealed pale optic disks and
scattered retinal hemorrhages. There was no evidence of
peripheral neuropathy. The remainder of the physical
examination was normal. An ultrasound revealed bilateral
hydronephrosis. His acute kidney injury resolved with
insertion of a suprapubic catheter. An audiogram revealed
bilateral sensorineural deafness. A water deprivation test
confirmed cranial diabetes insipidus.
Genetic testing revealed a complex heterozygote state
with c.506G and c.1611_1624del14 mutations in the
WFS1 gene on chromosome 4 confirming the clinical
impression of Wolfram’s or DIDMOAD (Diabetes Insipidus,
Diabetes Mellitus, Optic Atrophy and Deafness) syndrome.
Both parents were confirmed to be carriers and the family
received genetic counselling on the inheritance of this
autosomal recessive disorder.
One of the main justifications for identifying specific
types of diabetes due to other causes is the potential for
cure through treatment of the underlying condition.
However, at present this does not apply to monogenic
forms of diabetes. Instead justification comes from:
1 Potential to adapt the treatment regimen to the type
of monogenic diabetes (a step towards precision
medicine).
 60 Prediabetes and the Diagnosis of Diabetes
2 Ability to provide an individual patient and their family
with a more accurate prognosis (genetic counselling).
3 Potential to advance our understanding of diabetes in
general through elucidation of rare molecular and cel­
lular defects causing diabetes [124].
Monogenic diabetes represents a diverse group of con­
ditions with varied presentations and underlying genetic
causes [125]. In the case described above, the family was
not aware of the progressive nature of this rare condition
(1  in 770 000  in the UK  [126]) Wolfram syndrome with
almost inevitable death before age 50 (median age of death
39 (range 25–49) years) [126]; most families appreciate the
additional knowledge gained through genetic testing and
genetic counselling.
Examples where identification of monogenic forms of
diabetes result in a better fit between the diabetes syn­
drome and the treatment regimen come from both the
Maturity Onset Diabetes of the Young (MODY) and neo­
natal diabetes syndromes. MODY is inherited in an autoso­
mal dominant pattern, is often underdiagnosed and
accounts for > 1% of all cases of diabetes [123]. Several dif­
ferent gene defects have been recognized in MODY. MODY
3, due to a defect in the hepatocyte nuclear factor 1‐alpha
(HNF1α) gene, is the commonest form of MODY (30–65%
of MODY patients). Other common forms of MODY
include MODY 2 (due to GCK mutation; 30–50% of
MODY patients), MODY 1 (due to HNF4α mutation;
5–10% of MODY patients) and MODY 5 (due to HNF1β
mutation; < 5% of MODY patients) [127]. Several clinical
clues may suggest a diagnosis of MODY and the mutation
contributing to the diagnosis for example lower renal
threshold for glucose (HNF1α)  [128], large birthweight
(HNF4α) [129], transient or persistent neonatal hypogly­
cemia (HNF4α) [130], and renal disease (HNF‐1β) [131].
Of import, patients with MODY 2 often require no treat­
ment (stable fasting hyperglycemia), while patients with
MODY 3 and MODY 1 can be managed effectively with
sulphonylureas as a first‐line agent. Because diabetes due
to MODY can be diagnosed in childhood or adolescence,
many of these patients are started on insulin on the
assumption that they have type 1 diabetes. The ability to
recognize previously undiagnosed MODY and potentially
remove the need for lifelong insulin is a huge benefit to the
patient and allows patients to be treated with a more cost‐
effective therapy.
This phenomenon (of stopping insulin after many years
of use) has also been reported in patients with permanent
neonatal diabetes due to mutations in the Kir6.2 subunit of
the ATP sensitive potassium channels in the beta‐cell
(KCNJ11  mutation) (inherited spontaneously (80%) or
autosomal dominant)  [125, 132]. Cellular mechanism
relates to the mutated channels remaining open resulting
in the beta‐cell membrane remaining in a hyperpolarized
state thereby preventing insulin from being secreted.
Sulfonylureas leads to closure of the channel thus enabling
insulin secretion. Diabetes occurring before 6  months of
age is typically (80–85% [133]) due to monogenic causes,
which can occur transiently or permanently. Again, early
identification is essential as some forms of neonatal diabe­
tes can be treated effectively with sulphonylureas while
others require intensive insulin strategies.
Patients with other monogenic forms of diabetes are
often misdiagnosed as having type 1 diabetes (if childhood
onset) or type 2 diabetes (if adult onset). These monogenic
forms include that associated with the m.3243A>G muta­
tion in mitochondrial DNA as well as lipodystrophic forms
of diabetes due to mutations in the LMNA gene causing
familial partial lipodystrophy. The phenotypic characteris­
tics that should alert the clinician to the possibility of a
monogenic form of diabetes include (1) young age of onset,
(2) family history demonstrating either a pattern of autoso­
mal dominant inheritance (e.g. MODY kindreds) or of
maternal transmission (e.g. mitochondrial diabetes), (3)
association with deafness (e.g. mitochondrial diabetes), or
(4) neurological or neuromuscular disorders (e.g. Frederich’s
ataxia).
The ADA  [3] recommends screening for monogenic
diabetes syndromes in the following circumstances:
1 Diagnosis of diabetes < 6 months (neonatal diabetes).
2 Diagnosis of diabetes in children and early adulthood
with a pattern of inheritance suggestive of Autosomal
Dominance (MODY).
In addition, the ADA recommends consultation with a
center that has expertise in the management of monogenic
diabetes to interpret the mutations, optimize patient care,
and facilitate genetic counselling [3].
 How Should Secondary Causes of Diabetes Be Excluded?
References
1. Report of the Expert Committee on the Diagnosis and
Classification of Diabetes Mellitus. Diabetes Care.
1997;20:1183–1197.
2. Gale EA. Declassifying diabetes. Diabetologia.
2006;49:1989–1995. DOI: 10.1007/s00125‐006‐0348‐7.
3. Classification and diagnosis of diabetes: standards of
medical care in diabetes – 2020. Diabetes Care.
2020;43:S14–S31. DOI: 10.2337/dc20‐S002.
4. International Diabetes Federation. IDF Diabetes Atlas, 9th
edn. Brussels, Belgium, 2019. Available at www.diabetes
atlas.org.
5. McCulloch DK. Classification of Diabetes Mellitus and
Genetic Diabetic Syndromes. Waltham, MA. UpToDate.
2020.
6. Public Health England. Guidance. Criteria for Appraising
the Viability, Effectiveness and Appropriateness of a
Screening Programme. 2015.
7. UK National Screening Committee (UK NSC). Current UK
NSC recommendations.
8. U.S. Preventive Services Task Force. Abnormal Blood
Glucose and Type 2 Diabetes Mellitus: Screening. 2015.
9. Brissot P, Pietrangelo A, Adams PC, de Graaff B, McLaren
CE, Loreal O. Haemochromatosis. Nat Rev Dis Primers.
2018;4:18016. DOI: 10.1038/nrdp.2018.16.
10. Crownover BK, Covey CJ. Hereditary hemochromatosis.
Am Fam Physician. 2013;87:183–190.
11. Adams PC, Reboussin DM, Barton JC, McLaren CE, Eckfeldt
JH, McLaren GD et al. Hemochromatosis and iron‐overload
screening in a racially diverse population. N Engl J Med.
2005;352:1769–1778. DOI: 10.1056/NEJMoa041534.
12. Rossi E, Jeffrey GP. Clinical penetrance of C282Y homozy­
gous HFE haemochromatosis. Clin Biochem Rev. 2004;25:
183–190.
13. Whitlock EP, Garlitz BA, Harris EL, Beil TL, Smith PR.
Screening for hereditary hemochromatosis: a systematic
review for the U.S. Preventive Services Task Force. Ann
Intern Med. 2006;145:209–223. DOI: 145/3/209 [pii].
14. Bardou‐Jacquet E, Morcet J, Manet G, Laine F, Perrin M,
Jouanolle AM et al. Decreased cardiovascular and extrahe­
patic cancer‐related mortality in treated patients with mild
HFE hemochromatosis. J Hepatol. 2015;62:682–689. DOI:
10.1016/j.jhep.2014.10.025.
15. Wood MJ, Powell LW, Dixon JL, Ramm GA. Clinical cofac­
tors and hepatic fibrosis in hereditary hemochromatosis: the
role of diabetes mellitus. Hepatology. 2012;56:904–911. DOI:
10.1002/hep.25720.
61
16. Barton JC, Acton RT. Diabetes in HFE Hemochromatosis. J
Diabetes Res. 2017;2017:9826930. DOI: 10.1155/2017/982
6930.
17. Utzschneider KM, Kowdley KV. Hereditary hemochromato­
sis and diabetes mellitus: implications for clinical practice. Nat
Rev Endocrinol. 2010;6:26–33. DOI: 10.1038/nrendo.2009.241.
18. Screening for hemochromatosis: recommendation statement.
Ann Intern Med. 2006;145:204–208. DOI: 145/3/204 [pii].
19. UK National Screening Committee. Screening for Haemo­
chromatosis. 2015.
20. European Association For The Study Of The L. EASL clini­
cal practice guidelines for HFE hemochromatosis. J Hepatol.
2010;53:3–22. DOI: 10.1016/j.jhep.2010.03.001.
21. Bacon BR, Adams PC, Kowdley KV, Powell LW, Tavill AS,
American Association for the Study of Liver D. Diagnosis and
management of hemochromatosis: 2011 practice guideline by
the American Association for the Study of Liver Diseases.
Hepatology. 2011;54:328–343. DOI: 10.1002/hep.24330.
22. Force USPST. Screening for hemochromatosis: recommen­
dation statement. Ann Intern Med. 2006;145:204–208. DOI:
10.7326/0003‐4819‐145‐3‐200608010‐00008.
23. Kowdley KV, Brown KE, Ahn J, Sundaram V. ACG Clinical
Guideline: hereditary hemochromatosis. Am J Gastro­
enterol. 2019;114:1202–1218. DOI: 10.14309/ajg.00000000000
00315.
24. Bulaj ZJ, Ajioka RS, Phillips JD, LaSalle BA, Jorde LB, Griffen
LM et al. Disease‐related conditions in relatives of patients
with hemochromatosis. N Engl J Med. 2000;343:1529–1535.
DOI: 10.1056/NEJM200011233432104.
25. Jacobs EM, Hendriks JC, Marx JJ, van Deursen CT,
Kreeftenberg HG, de Vries RA et al. Morbidity and mortality
in first‐degree relatives of C282Y homozygous probands
with clinically detected haemochromatosis compared with
the general population: the HEmochromatosis FAmily Study
(HEFAS). Neth J Med. 2007;65:425–433.
26. O’Brien T, Barrett B, Murray DM, Dinneen S, O’Sullivan DJ.
Usefulness of biochemical screening of diabetic patients for
hemochromatosis. Diabetes Care. 1990;13:532–534.
27. Pilling LC, Tamosauskaite J, Jones G, Wood AR, Jones L,
Kuo CL et al. Common conditions associated with heredi­
tary haemochromatosis genetic variants: cohort study in UK
Biobank. BMJ. 2019;364:k5222. DOI: 10.1136/bmj.k5222.
28. American Diabetes A. 2. Classification and diagnosis of dia­
betes: standards of medical care in diabetes – 2020. Diabetes
Care. 2020;43:S14–S31. DOI: 10.2337/dc20‐S002.
29. Tavill AS. Diagnosis and management of hemochromatosis.
Hepatology. 2001;33:1321–1328. DOI:  S0270913901325302
[pii] 10.1053/jhep.2001.24783.
 62 Prediabetes and the Diagnosis of Diabetes
30. Ioannou L, McClaren BJ, Massie J, Lewis S, Metcalfe SA,
Forrest L et al. Population‐based carrier screening for cystic
fibrosis: a systematic review of 23 years of research. Genet
Med. 2014;16:207–216. DOI: 10.1038/gim.2013.125.
31. Moran A, Dunitz J, Nathan B, Saeed A, Holme B, Thomas W.
Cystic fibrosis‐related diabetes: current trends in prevalence,
incidence, and mortality. Diabetes Care. 2009;32:1626–1631.
DOI: dc09‐0586 [pii] 10.2337/dc09‐0586.
32. Moran A, Brunzell C, Cohen RC, Katz M, Marshall BC,
Onady  G et  al. Clinical care guidelines for cystic fibrosis‐
related diabetes: a position statement of the American
Diabetes Association and a clinical practice guideline of the
Cystic Fibrosis Foundation, endorsed by the Pediatric
Endocrine Society. Diabetes Care. 2010;33:2697–2708. DOI:
10.2337/dc10‐1768.
33. Milla CE, Billings J, Moran A. Diabetes is associated with
dramatically decreased survival in female but not male sub­
jects with cystic fibrosis. Diabetes Care. 2005;28:2141–2144.
DOI: 10.2337/diacare.28.9.2141.
34. Chamnan P, Shine BS, Haworth CS, Bilton D, Adler AI.
Diabetes as a determinant of mortality in cystic fibrosis.
Diabetes Care. 2010;33:311–316. DOI: 10.2337/dc09‐1215.
35. Schwarzenberg SJ, Thomas W, Olsen TW, Grover T, Walk D,
Milla C et al. Microvascular complications in cystic fibrosis‐
related diabetes. Diabetes Care. 2007;30:1056–1061. DOI:
10.2337/dc06‐1576.
36. Andersen HU, Lanng S, Pressler T, Laugesen CS, Mathiesen
ER. Cystic fibrosis‐related diabetes: the presence of micro­
vascular diabetes complications. Diabetes Care. 2006;29:
2660–2663. DOI: 10.2337/dc06‐0654.
37. van den Berg JM, Morton AM, Kok SW, Pijl H, Conway SP,
Heijerman HG. Microvascular complications in patients
with cystic fibrosis‐related diabetes (CFRD). J Cyst Fibros.
2008;7:515–519. DOI: 10.1016/j.jcf.2008.05.008.
38. Yung B, Landers A, Mathalone B, Gyi KM, Hodson ME.
Diabetic retinopathy in adult patients with cystic fibrosis‐
related diabetes. Respir Med. 1998;92:871–872. DOI: 10.1016/
s0954‐6111(98)90390‐0.
39. Litvin M, Nwachukwu S. Cystic Fibrosis Related Diabetes: a
Unique Challenge in Diabetes Care. Mo Med. 2016;113:
384–389.
40. Koch C, Rainisio M, Madessani U, Harms HK, Hodson ME,
Mastella G et al. Presence of cystic fibrosis‐related diabetes
mellitus is tightly linked to poor lung function in patients
with cystic fibrosis: data from the European Epidemiologic
Registry of Cystic Fibrosis. Pediatr Pulmonol. 2001;32:
343–350. DOI: 10.1002/ppul.1142.
41. Moran A, Becker D, Casella SJ, Gottlieb PA, Kirkman MS,
Marshall BC et  al. Epidemiology, pathophysiology, and
prognostic implications of cystic fibrosis‐related diabetes: a
technical review. Diabetes Care. 2010;33:2677–2683. DOI:
10.2337/dc10–1279.
42. Ode KL, Frohnert B, Laguna T, Phillips J, Holme B,
Regelmann W et  al. Oral glucose tolerance testing in chil­
dren with cystic fibrosis. Pediatr Diabetes. 2010;11:487–492.
DOI: 10.1111/j.1399‐5448.2009.00632.x.
43. Onady GM, Stolfi A. Insulin and oral agents for managing
cystic fibrosis‐related diabetes. Cochrane Database Syst Rev.
2016;4:CD004730. DOI: 10.1002/14651858.CD004730.pub4.
44. Moran A, Pekow P, Grover P, Zorn M, Slovis B, Pilewski J
et  al. Insulin therapy to improve BMI in cystic fibrosis‐
related diabetes without fasting hyperglycemia: results of the
cystic fibrosis related diabetes therapy trial. Diabetes Care.
2009;32:1783–1788. DOI: 10.2337/dc09‐0585.
45. Mozzillo E, Franzese A, Valerio G, Sepe A, De Simone I,
Mazzarella G et al. One‐year glargine treatment can improve
the course of lung disease in children and adolescents with
cystic fibrosis and early glucose derangements. Pediatr
Diabetes. 2009;10:162–167. DOI: 10.1111/j.1399‐5448.2008.
00451.x.
46. Mohan K, Israel KL, Miller H, Grainger R, Ledson MJ,
Walshaw MJ. Long‐term effect of insulin treatment in cystic
fibrosis‐related diabetes. Respiration. 2008;76:181–186.
DOI: 10.1159/000110206.
47. Rafii M, Chapman K, Stewart C, Kelly E, Hanna A, Wilson
DC et al. Changes in response to insulin and the effects of
varying glucose tolerance on whole‐body protein metabo­
lism in patients with cystic fibrosis. Am J Clin Nutr.
2005;81:421–426. DOI: 10.1093/ajcn.81.2.421.
48. Moran A, Hardin D, Rodman D, Allen HF, Beall RJ, Borowitz
D et  al. Diagnosis, screening and management of cystic
fibrosis related diabetes mellitus: a consensus conference
report. Diabetes Res Clin Pract. 1999;45:61–73. DOI:
10.1016/s0168‐8227(99)00058‐3.
49. O’Riordan SM, Robinson PD, Donaghue KC, Moran A,
Consensus ICP. Management of cystic fibrosis‐related diabetes.
Pediatr Diabetes. 2008;9:338–344. DOI: 10.1111/j.1399‐5448
.2008.00437.x.
50. Tan J, You Y, Guo F, Xu J, Dai H, Bie P. Association of ele­
vated risk of pancreatic cancer in diabetic patients: a system­
atic review and meta‐analysis. Oncol Lett. 2017;13:1247–1255.
DOI: 10.3892/ol.2017.5586.
51. Pannala R, Basu A, Petersen GM, Chari ST. New‐onset dia­
betes: a potential clue to the early diagnosis of pancreatic
 How Should Secondary Causes of Diabetes Be Excluded?
cancer. Lancet Oncol. 2009;10:88–95. DOI: 10.1016/S1470‐
2045(08)70337‐1.
52. Patel R, Ede J, Collins J, Willens D. Pancreatic cancer pre­
senting as new‐onset diabetes. Case Rep Oncol. 2014;7:
171–174. DOI: 10.1159/000360812.
53. Magruder JT, Elahi D, Andersen DK. Diabetes and pancre­
atic cancer: chicken or egg? Pancreas. 2011;40:339–351.
DOI: 10.1097/MPA.0b013e318209e05d.
54. Pannala R, Leirness JB, Bamlet WR, Basu A, Petersen GM,
Chari ST. Prevalence and clinical profile of pancreatic can­
cer‐associated diabetes mellitus. Gastroenterology. 2008;134:
981–987. DOI: 10.1053/j.gastro.2008.01.039.
55. De Bruijn KM, van Eijck CH. New‐onset diabetes after distal
pancreatectomy: a systematic review. Ann Surg. 2015;261:
854–861. DOI: 10.1097/SLA.0000000000000819.
56. Ben Q, Xu M, Ning X, Liu J, Hong S, Huang W et al. Diabetes
mellitus and risk of pancreatic cancer: a meta‐analysis of
cohort studies. Eur J Cancer. 2011;47:1928–1937. DOI:
10.1016/j.ejca.2011.03.003.
57. Li D, Tang H, Hassan MM, Holly EA, Bracci PM, Silverman
DT. Diabetes and risk of pancreatic cancer: a pooled analysis
of three large case‐control studies. Cancer Causes Control.
2011;22:189–197. DOI: 10.1007/s10552‐010‐9686‐3.
58. Chari ST, Leibson CL, Rabe KG, Ransom J, de Andrade M,
Petersen GM. Probability of pancreatic cancer following dia­
betes: a population‐based study. Gastroenterology. 2005;129:
504–511. DOI: 10.1016/j.gastro.2005.05.007.
59. Fernandez‐del Castillo C. Clinical manifestations, diagnosis,
and staging of exocrine pancreatic cancer. Waltham, MA.
UpToDate. 2020.
60. McGuigan A, Kelly P, Turkington RC, Jones C, Coleman
HG, McCain RS. Pancreatic cancer: a review of clinical diag­
nosis, epidemiology, treatment and outcomes. World J
Gastroenterol. 2018;24:4846–4861. DOI: 10.3748/wjg.v24.
i43.4846.
61. Noy A, Bilezikian JP. Clinical review 63: diabetes and pan­
creatic cancer: clues to the early diagnosis of pancreatic
malignancy. J Clin Endocrinol Metab. 1994;79:1223–1231.
DOI: 10.1210/jcem.79.5.7962312.
62. Xu M, Jung X, Hines OJ, Eibl G, Chen Y. Obesity and pancre­
atic cancer: overview of epidemiology and potential preven­
tion by weight loss. Pancreas. 2018;47:158–162. DOI:
10.1097/MPA.0000000000000974.
63. Ramachandran V, Arumugam T, Hwang RF, Greenson JK,
Simeone DM, Logsdon CD. Adrenomedullin is expressed in
pancreatic cancer and stimulates cell proliferation and inva­
sion in an autocrine manner via the adrenomedullin receptor,
63
ADMR. Cancer Res. 2007;67:2666–2675. DOI: 10.1158/0008‐
5472.CAN‐06‐3362.
64. Aggarwal G, Ramachandran V, Javeed N, Arumugam T,
Dutta S, Klee GG et al. Adrenomedullin is up‐regulated in
patients with pancreatic cancer and causes insulin resistance
in beta cells and mice. Gastroenterology. 2012;143:1510–
1517 e1511. DOI: 10.1053/j.gastro.2012.08.044.
65. Antolino L, Rocca M, Todde F, Catarinozzi E, Aurello P,
Bollanti L et al. Can pancreatic cancer be detected by adre­
nomedullin in patients with new‐onset diabetes? The
PaCANOD cohort study protocol. Tumori. 2018;104:312–
314. DOI: 10.5301/tj.5000693.
66. Force USPST, Owens DK, Davidson KW, Krist AH, Barry
MJ, Cabana M et  al. Screening for pancreatic cancer: US
Preventive Services Task Force Reaffirmation Recommen­
dation Statement. JAMA. 2019;322:438–444. DOI: 10.1001/
jama.2019.10232.
67. O’Shea PM, Griffin TP, Fitzgibbon M. Hypertension: the
role of biochemistry in the diagnosis and management. Clin
Chim Acta. 2017;465:131–143. DOI: 10.1016/j.cca.2016.
12.014.
68. Lavrentaki A, Paluzzi A, Wass JAH, Karavitaki N.
Epidemiology of acromegaly: review of population studies.
Pituitary. 2017;20:4–9. DOI: 10.1007/s11102‐016‐0754‐x.
69. Beard CM, Sheps SG, Kurland LT, Carney JA, Lie JT.
Occurrence of pheochromocytoma in Rochester, Minnesota,
1950 through 1979. Mayo Clin Proc. 1983;58:802–804.
70. Katznelson L, Laws ER, Jr., Melmed S, Molitch ME, Murad
MH, Utz A et al. Acromegaly: an endocrine society clinical
practice guideline. J Clin Endocrinol Metab. 2014;99:3933–
3951. DOI: 10.1210/jc.2014‐2700.
71. Vigneri R, Squatrito S, Pezzino V, Filetti S, Branca S, Polosa
P. Growth hormone levels in diabetes. Correlation with the
clinical control of the disease. Diabetes. 1976;25:167–172.
DOI: 10.2337/diab.25.3.167.
72. Lim DJ, Kwon HS, Cho JH, Kim SH, Choi YH, Yoon KH
et  al. Acromegaly associated with type 2 diabetes showing
normal IGF‐1  levels under poorly controlled glycemia.
Endocr J. 2007;54:537–541. DOI: 10.1507/endocrj.k06‐083.
73. Siminoski K, Goss P, Drucker DJ. The Cushing Syndrome
induced by medroxyprogesterone acetate. Annals of Internal
Medicine. 1989;111:758–760. DOI: 10.7326/0003‐4819‐111‐
9‐758.
74. Wengander S, Trimpou P, Papakokkinou E, Ragnarsson O.
The incidence of endogenous Cushing’s syndrome in the
modern era. Clin Endocrinol (Oxf). 2019;91:263–270. DOI:
10.1111/cen.14014.
 64 Prediabetes and the Diagnosis of Diabetes
75. Reimondo G, Pia A, Allasino B, Tassone F, Bovio S, Borretta
G et  al. Screening of Cushing’s syndrome in adult patients
with newly diagnosed diabetes mellitus. Clin Endocrinol
(Oxf). 2007;67:225–229. DOI: 10.1111/j.1365‐2265.2007.
02865.x.
76. Boscaro M, Arnaldi G. Approach to the patient with possible
Cushing’s syndrome. J Clin Endocrinol Metab. 2009;94:3121–
3131. DOI: 94/9/3121 [pii] 10.1210/jc.2009‐0612.
77. Terzolo M, Reimondo G, Chiodini I, Castello R, Giordano R,
Ciccarelli E et al. Screening of Cushing’s syndrome in outpa­
tients with type 2 diabetes: results of a prospective multicen­
tric study in Italy. J Clin Endocrinol Metab. 2012;97:
3467–3475. DOI: 10.1210/jc.2012‐1323.
78. Budyal S, Jadhav SS, Kasaliwal R, Patt H, Khare S, Shivane V
et al. Is it worthwhile to screen patients with type 2 diabetes
mellitus for subclinical Cushing’s syndrome? Endocrine
Connections. 2015;4:242–248. DOI: 10.1530/EC‐15‐0078.
79. Nieman L, Biller B, Findling J, Newell‐Price J, Savage M,
Stewart P et  al. The diagnosis of Cushing’s syndrome: an
Endocrine Society Clinical Practice Guideline. Eur J
Endocrinol. 2009. DOI: EJE‐09‐0695 [pii] 10.1530/EJE‐09
‐0695.
80. Kuypers DR, Claes K, Bammens B, Evenepoel P,
Vanrenterghem Y. Early clinical assessment of glucose
metabolism in renal allograft recipients: diagnosis and pre­
diction of post‐transplant diabetes mellitus (PTDM).
Nephrol Dial Transplant. 2008;23:2033–2042. DOI: 10.1093/
ndt/gfm875.
81. Chowdhury TA. Post‐transplant diabetes mellitus. Clin
Med  (Lond). 2019;19:392–395. DOI: 10.7861/clinmed.
2019‐0195.
82. Munshi VN, Saghafian S, Cook CB, Werner KT, Chakkera
HA. Comparison of post‐transplantation diabetes mellitus
incidence and risk factors between kidney and liver trans­
plantation patients. PLoS One. 2020;15:e0226873. DOI:
10.1371/journal.pone.0226873.
83. Hackman KL, Snell GI, Bach LA. Prevalence and predictors
of diabetes after lung transplantation: a prospective, longitu­
dinal study. Diabetes Care. 2014;37:2919–2925. DOI: 10.2337/
dc14‐0663.
84. Rodrigo E, Fernandez‐Fresnedo G, Valero R, Ruiz JC, Pinera
C, Palomar R et al. New‐onset diabetes after kidney trans­
plantation: risk factors. J Am Soc Nephrol. 2006;17:S291–S295.
DOI: 10.1681/ASN.2006080929.
85. Lecube A, Hernandez C, Genesca J, Simo R. Glucose abnor­
malities in patients with hepatitis C virus infection: epidemi­
ology and pathogenesis. Diabetes Care. 2006;29:1140–1149.
DOI: 10.2337/diacare.2951140.
86. Hum J, Jou JH, Green PK, Berry K, Lundblad J, Hettinger BD
et  al. Improvement in glycemic control of type 2 diabetes
after successful treatment of hepatitis C virus. Diabetes Care.
2017;40:1173–1180. DOI: 10.2337/dc17‐0485.
87. Chakkera HA, Mandarino LJ. Calcineurin inhibition and
new‐onset diabetes mellitus after transplantation. Trans­
plan­­ta­tion. 2013;95:647–652. DOI: 10.1097/TP.0b013e31826
e592e.
88. Cosio FG, Pesavento TE, Kim S, Osei K, Henry M, Ferguson
RM. Patient survival after renal transplantation: IV. Impact
of post‐transplant diabetes. Kidney Int. 2002;62:1440–1446.
DOI: kid582 [pii] 10.1111/j.1523–1755.2002.kid582.x.
89. Elmagd MM, Bakr MA, Metwally AH, Wahab AM. Clinico­
epidemiologic study of posttransplant diabetes after living‐
donor renal transplant. Exp Clin Transplant. 2008;6:42–47.
90. Kasiske BL, Snyder JJ, Gilbertson D, Matas AJ. Diabetes mel­
litus after kidney transplantation in the United States. Am J
Trans­plant. 2003;3:178–185. DOI: 10.1034/j.1600‐6143.2003.
00010.x.
91. Miles AM, Sumrani N, Horowitz R, Homel P, Maursky V,
Markell MS et al. Diabetes mellitus after renal transplanta­
tion: as deleterious as non‐transplant‐associated diabetes?
Transplantation. 1998;65:380–384. DOI: 10.1097/00007890‐
199802150‐00014.
92. Londero TM, Giaretta LS, Farenzena LP, Manfro RC, Canani
LH, Lavinsky D et al. Microvascular complications of post­
transplant diabetes mellitus in kidney transplant recipients:
a longitudinal study. J Clin Endocrinol Metab. 2019;104:557–
567. DOI: 10.1210/jc.2018‐01521.
93. Sharif A, Baboolal K. Complications associated with new‐
onset diabetes after kidney transplantation. Nat Rev Nephrol.
2011;8:34–42. DOI: 10.1038/nrneph.2011.174.
94. American Diabetes A. 2. Classification and diagnosis of dia­
betes: standards of medical care in diabetes – 2019. Diabetes
Care. 2019;42:S13–S28. DOI: 10.2337/dc19‐S002.
95. Bodziak KA, Hricik DE. New‐onset diabetes mellitus after
solid organ transplantation. Transpl Int. 2009;22:519–530.
DOI: TRI800 [pii] 10.1111/j.1432‐2277.2008.00800.x.
96. Vanhove T, Remijsen Q, Kuypers D, Gillard P. Drug‐drug
interactions between immunosuppressants and antidiabetic
drugs in the treatment of post‐transplant diabetes mellitus.
Transplantation Reviews (Orlando, Fla). 2017;31:69–77. DOI: 10.
1016/j.trre.2016.09.001.
97. Oberholzer J, Thielke J, Hatipoglu B, Testa G, Sankary HN,
Benedetti E. Immediate conversion from tacrolimus to
cyclosporine in the treatment of posttransplantation diabetes
mellitus. Transplant Proc. 2005;37:999–1000. DOI: 10.1016/
j.transproceed.2004.12.085.
 How Should Secondary Causes of Diabetes Be Excluded?
  98. Rouillon F, Sorbara F. Schizophrenia and diabetes: epide­
miological data. European Psychiatry. 2005;20:S345–S348.
DOI: https://doi.org/10.1016/S0924‐9338(05)80189‐0.
  99. De Hert M, Dekker JM, Wood D, Kahl KG, Holt RI, Moller
HJ. Cardiovascular disease and diabetes in people with
severe mental illness position statement from the European
Psychiatric Association (EPA), supported by the European
Association for the Study of Diabetes (EASD) and the
European Society of Cardiology (ESC). Eur Psychiatry.
2009;24:412–424. DOI: 10.1016/j.eurpsy.2009.01.005.
100. Cohen D, Stolk RP, Grobbee DE, Gispen‐de Wied CC.
Hyperglycemia and diabetes in patients with schizophrenia
or schizoaffective disorders. Diabetes Care. 2006;29:786–791.
DOI: 10.2337/diacare.29.04.06.dc05‐1261.
101. Kamble P, Sherer J, Chen H, Aparasu R. Off‐label use of
second‐generation antipsychotic agents among elderly
nursing home residents. Psychiatric Services. 2010;61:130–
136. DOI: 10.1176/ps.2010.61.2.130.
102. Llorente MD, Urrutia V. Diabetes, Psychiatric disorders,
and the metabolic effects of antipsychotic medications.
Clinical Diabetes. 2006;24:18–24. DOI: 10.2337/diaclin.
24.1.18.
103. McCreadie R, Macdonald E, Blacklock C, Tilak‐Singh D,
Wiles D, Halliday J et  al. Dietary intake of schizophrenic
patients in Nithsdale, Scotland: case‐control study. BMJ.
1998;317:784–785. DOI: 10.1136/bmj.317.7161.784.
104. Thakore JH, Mann JN, Vlahos I, Martin A, Reznek R.
Increased visceral fat distribution in drug‐naive and drug‐
free patients with schizophrenia. Int J Obes Relat Metab
Disord. 2002;26:137–141. DOI: 10.1038/sj.ijo.0801840.
105. Freemantle N, Holmes J, Hockey A, Kumar S. How strong
is the association between abdominal obesity and the inci­
dence of type 2 diabetes? International Journal of Clinical
Practice. 2008;62:1391–1396. DOI: 10.1111/j.1742–1241.
2008.01805.x.
106. Smith M, Hopkins D, Peveler RC, Holt RI, Woodward M,
Ismail K. First‐ v. second‐generation antipsychotics and risk
for diabetes in schizophrenia: systematic review and meta‐
analysis. Br J Psychiatry. 2008;192:406–411. DOI: 10.1192/
bjp.bp.107.037184.
107. Bobo WV, Cooper WO, Stein CM, Olfson M, Graham D,
Daugherty J et al. Antipsychotics and the risk of type 2 dia­
betes mellitus in children and youth. JAMA Psychiatry.
2013;70:1067–1075. DOI: 10.1001/jamapsychiatry.2013.2053.
108. Taylor DM, McAskill R. Atypical antipsychotics and
weight gain – a systematic review. Acta Psychiatrica
Scandinavica. 2000;101:416–432. DOI: 10.1034/j.1600‐0447
.2000.101006416.x.
65
109. Hui Fang CE, Rafey MF, Cunningham A, Dinneen SF,
Finucane FM. Risperidone‐induced type 2 diabetes pre­
senting with diabetic ketoacidosis. Endocrinol Diabetes
Metab Case Rep. 2018;May 10. DOI: 10.1530/EDM‐18‐0031.
110. Consensus development conference on antipsychotic
drugs and obesity and diabetes. Diabetes Care. 2004;27:
w596–601.
111. Citrome L. Risk‐benefit analysis of available treatments for
schizophrenia. Psychiatric Times. 2007;24:27–30.
112. Ader M, Kim SP, Catalano KJ, Ionut V, Hucking K, Richey
JM et  al. Metabolic dysregulation with atypical antipsy­
chotics occurs in the absence of underlying disease: a pla­
cebo‐controlled study of olanzapine and risperidone in
dogs. Diabetes. 2005;54:862–871. DOI: 54/3/862 [pii].
113. Whicher CA, Price HC, Holt RIG. Mechanisms in endocri­
nology: antipsychotic medication and type 2 diabetes and
impaired glucose regulation. Eur J Endocrinol.
2018;178:R245–R258. DOI: 10.1530/EJE‐18‐0022.
114. Churchward S OS, Olotu VO, Thalitaya MD. Setting stand­
ards for physical health monitoring in patients on antipsy­
chotics. Psychiatric Bulletin. 2009;33:451–454.
115. Morse CG, Kovacs JA. Metabolic and skeletal complica­
tions of HIV infection: the price of success. JAMA.
2006;296:844–854. DOI: 296/7/844 [pii] 10.1001/
jama.296.7.844.
116. Brown TT, Cole SR, Li X, Kingsley LA, Palella FJ, Riddler
SA et  al. Antiretroviral therapy and the prevalence and
incidence of diabetes mellitus in the multicenter AIDS
cohort study. Arch Intern Med. 2005;165:1179–1184. DOI:
10.1001/archinte.165.10.1179.
117. Wohl DA, McComsey G, Tebas P, Brown TT, Glesby MJ,
Reeds D et al. Current concepts in the diagnosis and man­
agement of metabolic complications of HIV infection and
its therapy. Clin Infect Dis. 2006;43:645–653. DOI: 10.1086/
507333.
118. De Wit S, Sabin CA, Weber R, Worm SW, Reiss P, Cazanave
C et al. Incidence and risk factors for new‐onset diabetes in
HIV‐infected patients: the Data Collection on Adverse
Events of Anti‐HIV Drugs (D:A:D) study. Diabetes Care.
2008;31:1224–1229. DOI: dc07‐2013 [pii] 10.2337/dc07
‐2013.
119. Nduka CU, Stranges S, Kimani PK, Sarki AM, Uthman
OA. Is there sufficient evidence for a causal association
between antiretroviral therapy and diabetes in HIV‐
infected patients? A meta‐analysis. Diabetes Metab Res
Rev. 2017;33. DOI: 10.1002/dmrr.2902.
120. Echecopar‐Sabogal J, D’Angelo‐Piaggio L, Chanamé‐Baca
DM, Ugarte‐Gil C. Association between the use of protease
 66 Prediabetes and the Diagnosis of Diabetes
inhibitors in highly active antiretroviral therapy and inci­
dence of diabetes mellitus and/or metabolic syndrome in
HIV‐infected patients: a systematic review and meta‐analysis.
Int J STD AIDS. 2018;29:443–452. DOI: 10.1177/0956462417
732226.
121. Prioreschi A, Munthali RJ, Soepnel L, Goldstein JA,
Micklesfield LK, Aronoff DM et  al. Incidence and preva­
lence of type 2 diabetes mellitus with HIV infection in
Africa: a systematic review and meta‐analysis. BMJ Open.
2017;7:e013953. DOI: 10.1136/bmjopen‐2016‐013953.
122. Kim PS, Woods C, Georgoff P, Crum D, Rosenberg A,
Smith M et al. A1C underestimates glycemia in HIV infec­
tion. Diabetes Care. 2009;32:1591–1593. DOI: 10.2337/
dc09‐0177.
123. Kleinberger JW, Pollin TI. Undiagnosed MODY. Time for
Action. Curr Diab Rep. 2015;15:110. DOI: 10.1007/s11892‐
015‐0681‐7.
124. O’Rahilly S. Human genetics illuminates the paths to meta­
bolic disease. Nature. 2009;462:307–314. DOI: nature08532
[pii] 10.1038/nature08532.
125. Philipson LH, Murphy R, Ellard S, Hattersley AT, Stoy J,
Greeley SAW et al. Chapter 2 – genetic testing in diabetes
mellitus: a clinical guide to monogenic diabetes. In: Weiss
RE, Refetoff S, eds. Genetic Diagnosis of Endocrine
Disorders. (First Edition). San Diego: Academic Press;
2010:17–26.
126. Barrett TG, Bundey SE, Macleod AF. Neurodegeneration and
diabetes: UK nationwide study of Wolfram (DIDMOAD)
syndrome. Lancet. 1995;346:1458–1463. DOI: 10.1016/
s0140‐6736(95)92473‐6.
127. Naylor R, Knight Johnson A, del Gaudio D. Maturity‐Onset
Diabetes of the Young Overview. 2018 May 24. In: Adam
MP, Ardinger HH, Pagon RA et al. editors. GeneReviews®
[Internet]. Seattle (WA): University of Washington, Seattle;
1993–2020. Available from: https://www.ncbi.nlm.nih.gov/
books/NBK500456/.
128. McDonald TJ, Ellard S. Maturity onset diabetes of the
young: identification and diagnosis. Ann Clin Biochem.
2013;50:403–415. DOI: 10.1177/0004563213483458.
129. Dickens LT, Naylor RN. Clinical management of women
with monogenic diabetes during pregnancy. Curr Diab Rep.
2018;18:12. DOI: 10.1007/s11892‐018‐0982‐8.
130. Kapoor RR, Locke J, Colclough K, Wales J, Conn JJ,
Hattersley AT et al. Persistent hyperinsulinemic hypoglyce­
mia and maturity‐onset diabetes of the young due to
heterozygous HNF4A mutations. Diabetes. 2008;57:1659–
1663. DOI: 10.2337/db07‐1657.
131. Malecki MT, Skupien J, Gorczynska‐Kosiorz S, Klupa T,
Nazim J, Moczulski DK et al. Renal malformations may be
linked to mutations in the hepatocyte nuclear factor‐1alpha
(MODY3) gene. Diabetes Care. 2005;28:2774–2776. DOI:
10.2337/diacare.28.11.2774.
132. Hattersley A, Bruining J, Shield J, Njolstad P, Donaghue
KC. The diagnosis and management of monogenic diabetes
in children and adolescents. Pediatr Diabetes. 2009;10
Suppl 12:33–42. DOI: PDI571 [pii] 10.1111/j.1399‐5448.2009.
00571.x.
133. De Franco E, Flanagan SE, Houghton JA, Lango Allen H,
Mackay DJ, Temple IK et al. The effect of early, comprehen­
sive genomic testing on clinical care in neonatal diabetes:
an international cohort study. Lancet. 2015;386:957–963.
DOI: 10.1016/S0140‐6736(15)60098‐8.
134. Esfahani K, Roudaia L, Buhlaiga N, Del Rincon SV, Papneja
N, Miller WH, Jr. A review of cancer immunotherapy: from
the past, to the present, to the future. Curr Oncol.
2020;27:S87–S97. DOI: 10.3747/co.27.5223.
135. Higham CE, Olsson‐Brown A, Carroll P, Cooksley T,
Larkin J, Lorigan P et  al. Society for Endocrinology
Endocrine Emergency Guidance. Acute management of
the endocrine complications of checkpoint inhibitor ther­
apy. Endocr Connect. 2018;7:G1–G7. DOI: 10.1530/EC‐
18‐0068.
136. de Filette JMK, Pen JJ, Decoster L, Vissers T, Bravenboer B,
Van der Auwera BJ et al. Immune checkpoint inhibitors and
type 1 diabetes mellitus: a case report and systematic review.
Eur J Endocrinol. 2019;181:363–374. DOI: 10.1530/EJE‐
19‐0291.
137. Stamatouli AM, Quandt Z, Perdigoto AL, Clark PL, Kluger
H, Weiss SA et  al. Collateral damage: insulin‐dependent
diabetes induced with checkpoint inhibitors. Diabetes. 2018;
67:1471–1480. DOI: 10.2337/dbi18‐0002.
138. Wright JJ, Salem JE, Johnson DB, Lebrun‐Vignes B,
Stamatouli A, Thomas JW et  al. Increased reporting of
immune checkpoint inhibitor‐associated diabetes. Diabetes
Care. 2018;41:e150–e151. DOI: 10.2337/dc18‐1465.
139. Castinetti F, Borson‐Chazot F. [Immunotherapy‐induced
endocrinopathies: insights from the 2018 French Endocrine
Society Guidelines]. Bull Cancer. 2019;106:492–496. DOI:
10.1016/j.bulcan.2019.02.003.
140. Barroso‐Sousa R, Barry WT, Garrido‐Castro AC, Hodi FS,
Min L, Krop IE et al. Incidence of endocrine dysfunction fol­
lowing the use of different immune checkpoint inhibitor regi­
mens: a systematic review and meta‐analysis. JAMA Oncol.
2018;4:173–182. DOI: 10.1001/jamaoncol.2017.3064.
141. Puzanov I, Diab A, Abdallah K, Bingham CO, 3rd, Brogdon
C, Dadu R et  al. Managing toxicities associated with
immune checkpoint inhibitors: consensus recommenda­
tions from the Society for Immunotherapy of Cancer
 How Should Secondary Causes of Diabetes Be Excluded?
(SITC) Toxicity Management Working Group. J Immunother
Cancer. 2017;5:95. DOI: 10.1186/s40425‐017‐0300‐z.
142. Griffin TP, Bogdanet D, Navin P, Callagy G, O’Shea PM,
Bell M. The importance of standardisation of measurement
and reference intervals for detection of phaeochromocy­
toma and paraganglioma (PPGL). Ir J Med Sci. 2018;187:
993–998.DOI: 10.1007/s11845‐018‐1756‐7.
143. Casey R, Griffin TP, Wall D, Dennedy MC, Bell M, O’Shea
PM. Screening for phaeochromocytoma and paragangli­
67
oma: impact of using supine reference intervals for plasma
metanephrines with samples collected from fasted/seated
patients. Ann Clin Biochem. 2017;54:170–173. DOI:
10.1177/0004563216646395.
144. Griffin TP, Casey R, Wall D, Bell M, O’Shea PM.
Evaluating the optimum rest period prior to blood col­
lection for fractionated plasma free metanephrines
analysis. Pract Lab Med. 2016;5:39–46. DOI: 10.1016/j.
plabm.2016.05.001.
 5 How to Screen Appropriately
for Monogenic Diabetes
Adrian Vella
Professor of Medicine, Division of Endocrinology, Diabetes, Metabolism, and Nutrition, Mayo Clinic, Rochester, MN, USA
LE A R N I N G P OI NT S
●● The term “Maturity Onset Diabetes of the Young” has
been superseded by the term “Monogenic Diabetes”
for reasons related to the underlying pathophysiology
and heterogeneity of the condition.
●● The commonest forms of monogenic diabetes are those
caused by mutations in GCK, HNF1A and HNF4A.
●● Extra‐pancreatic manifestations may help raise the
clinical suspicion of monogenic diabetes and suggest a
genetic diagnosis.
●● A logistic regression model using demographic data
and disease history can be used to predict whether
genetic testing is indicated.
Maturity onset diabetes of the young
(MODY)
Diabetes is a clinical syndrome characterized by hypergly-
cemia [1]. The causation of diabetes in a given individual
may be heterogeneous but is broadly categorized into insu-
lin‐mediated diabetes (type 1 diabetes) and non‐insulin
mediated diabetes (type 2 diabetes) [2]. However, it is read-
ily apparent that a small minority of patients have a mono-
genic disorder underlying their diabetes  [3]. These
disorders have been collectively termed “Maturity‐Onset
Diabetes of the Young (MODY)”. The challenge to recog-
nizing such patients is balancing expensive genetic screen-
ing against the potential harm of misdiagnosis and
consequent failure to optimize therapy. Making a correct
Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
68
diagnosis may also facilitate genetic counselling to affected
individuals and their families.
Affected individuals were initially characterized on the
basis of a family history suggesting an autosomal mode of
inheritance and the presence of residual insulin secretion –
hence the term “Maturity‐Onset” implying that the disease
behaves like type 2 diabetes albeit presenting at a younger
age (at least one member of an affected kindred diagnosed
before age 25). Mutations in at least seven different genes
have been identified thereby accounting for the significant
heterogeneity in clinical presentation with different age at
presentation, differences in the pattern of hyperglycemia
(due in part to the severity of insulin deficiency) and the
presence or absence of extra‐pancreatic manifestations [4].
Accordingly, monogenic diabetes has become the preferred
term for describing disease(s) previously characterized as
MODY [5].
Differentiating monogenic diabetes
from type 1 diabetes
The diagnosis of neonatal diabetes (< 6 months of age) is
beyond the scope of this chapter. However, such forms of
diabetes often have a monogenic cause and have may have
associated (“syndromic”) clinical features. Diabetes may be
transient in some monogenic forms of neonatal diabetes
(e.g. mutations in KCNJ11)  [6]. Type 1 diabetes rarely
occurs before 6 months of age and, in fact, patients with per-
manent diabetes presenting before 6  months of age do not
have increased genetic susceptibility to type 1 diabetes  [7].
 How to screen appropriately for Monogenic Diabetes
In  contrast, in diabetes presenting after 6  months of age,
type 1 diabetes is the commonest underlying cause.
As discussed elsewhere, the presence of detectable C‐pep-
tide – a marker of endogenous insulin secretion – is unhelp-
ful in making a diagnosis of type 1 diabetes at the time of
presentation with hyperglycemia and/or ketosis. The pres-
ence of autoantibodies associated with an autoimmune pro-
cess in the pancreatic islets is certainly helpful in making a
diagnosis. However, in their absence, a genetic cause should
be considered especially if there is a two‐ or three‐genera-
tion family history of diabetes. Persistence of endogenous
insulin secretion for more than 5 years after initial presenta-
tion should also question a diagnosis of type 1 diabetes [8].
The decreasing costs of sequencing large swathes of the
genome has resulted in the development of genetic panels
that can simultaneously test for multiple sequence changes
associated with different forms of monogenic diabetes. This
has decreased the need for a priori selection of genes for
sequencing most likely to explain such presentations.
Nevertheless, mutations in HNF1A are most likely to explain
a presentation similar to that of type 1 diabetes [5].
Differentiating monogenic diabetes
from type 2 diabetes
Type 2 diabetes is an imprecisely defined disease where age at
presentation has decreased because of the increased preva-
lence of obesity in the population  [1]. There are numerous
reports of presentation in adolescents and young adults with a
subsequent relatively rapid progression to severe β‐cell dys-
function [9]. The absence of obesity, the presence of a multi-
generational family history of diabetes and no evidence of
insulin resistance such as acanthosis nigricans may suggest a
diagnosis other than that of type 2 diabetes. Disorders of tri-
glyceride or lipid metabolism do not exclude monogenic dia-
betes since these dyslipidemias can be observed in diabetes
due to GCK or HNF4A mutations respectively [5, 10].
When to screen for monogenic diabetes?
In the absence of blanket testing, the challenge is to identify
the 1–2% of the population with diabetes who may have an
underlying monogenic form of diabetes. The falling cost of
gene­tic characterization using a gene panel approach decreases
the need to define the genetic etiology prior to testing.
69
However, applying such testing to a group of patients that
are less rigorously selected decreases the prior probability
of monogenic diabetes being present and increases the
likelihood of false positive findings. Also, benign polymor-
phisms may be reported as disease‐causing variants in the
absence of an understanding of the structural conse-
quences of a coding change [11].
Traditional clinical expertise has focused on features of
the history (and family history), a physical examination
and specialized testing to identify features associated with
the syndromes of monogenic diabetes. C‐peptide and islet
antibody testing has significant limitations and family his-
tory may not be well characterized [12, 13].
Using research databases where type 1 diabetes was
defined by the use of insulin within 6 months of diagnosis
and where a genetic diagnosis was present (mutation test-
ing in GCK, HNF1A and HNF4A) Shields et al. developed a
clinical prediction model to identify patients with a high
likelihood of having monogenic diabetes. This model
improves the sensitivity and specificity of the diagnosis
compared to using the prior criteria of having an affected
parent and an age < 25 at the time of diagnosis [14].
The factors that discriminated monogenic diabetes from
type 1 diabetes were an older age at diagnosis, better glycemic
control (lower HbA1c), female sex and a parent with diabetes.
To discriminate from type 2 diabetes, the model identified
body mass index (lower), female sex, better glycemic control
(lower HbA1c), the absence of treatment with insulin or oral
hypoglycemic agents, and a parent with diabetes. A version of
the logistic regression models that identified these associa-
tions is now available online (www.diabetesgenes.org) and as
Diabetes Diagnostic application for mobile platforms.
This probalistic assessment should be combined with
islet autoantibody testing and C‐peptide analysis in
patients treated with insulin. The presence of autoantibod-
ies or a C‐peptide < 200pmol/l makes a diagnosis of mono-
genic diabetes unlikely [15].
Common forms of monogenic diabetes
GCK mutations
Glucokinase catalyzes the phosphorylation of glucose
which in practice is the rate‐limiting step of glycolysis and
of glucose uptake in the liver, skeletal muscle and in the
β‐cell. It acts as a glucose sensor in the islet. Heterozygous
 70 Prediabetes and the Diagnosis of Diabetes
inactivating mutations of this enzyme produce an altered
set point, increasing the threshold for insulin secretion.
Consequently, affected individuals typically have mild
hyperglycemia and are asymptomatic. Since this is a set‐
point abnormality, the β‐cell response to an oral challenge
is intact so that HbA1c rarely exceed 7.5% and microvascu-
lar complications are rare.
Unfortunately, a subset of patients with prediabetes have
impaired fasting glucose and normal glucose tolerance – in
effect behaving like people with GCK mutations  [16].
Recent evidence suggests that these individuals progress to
type 2 diabetes at the same rate as do those with impaired
glucose tolerance so that a genetic test may be necessary to
counsel patients appropriately [17]. Typically, patients with
GCK mutations do not require glucose‐lowering therapy.
However, pregnant women with this mutation run the risk
of fetal macrosomia if their child does not inherit this
mutation and secretes excess insulin (stimulated by mater-
nal hyperglycemia) which promotes fetal growth [18].
HNF1A mutations
Affected patients typically develop diabetes in their teens
and experience progressive β‐cell failure. HNF1A is a tran-
scription factor necessary for pancreatic islet development
and functioning. The risk of microvascular complications
in affected individuals is related to poor glycemic control as
observed in other forms of diabetes. The insulin secretory
apparatus is intact in affected β‐cells and therefore sulfony-
lureas are able to stimulate insulin secretion even when
patients are treated with insulin.
HNF4A mutations
The clinical features of affected patients overlap signifi-
cantly with those affected by HNF1A mutations and are
characterized by progressive β‐cell dysfunction which is
responsive to sulfonylurea treatment. Dyslipidemia with
elevated LDL‐cholesterol and decreased HDL‐cholesterol
is a common accompanying feature.
Around 10% of families with autosomal dominant β‐cell
dysfunction do not have an identified genetic diagnosis
implying the presence of undiscovered genetic mutations.
Rare forms of monogenic diabetes include those caused by
mutations in NEUROD1 and PAX1 – both genes encoding
transcription factors important in early β‐cell develop-
ment. Wolfram syndrome is caused by mutations in WFS1
and is characterized by diabetes, diabetes insipidus and
optic atrophy. Common variants in the same gene are asso-
ciated with type 2 diabetes [19]. Wolfram syndrome is an
autosomal recessive disorder.
HNF1B mutations are associated with the presence of renal
disease that may range from cysts to renal dysplasia and other
forms of developmental renal abnormalities. Genitourinary
tract abnormalities may also be present. Some common vari-
ants in HNF1B alter predisposition to type 2 diabetes and to
prostate cancer [19]. The degree of pancreatic endocrine dys-
function that is usually present means that sulfonylureas are
often ineffective in these patients.
Conclusions
Monogenic diabetes is often overlooked as a diagnostic
possibility but should be considered when patients present
with diabetes at an “atypical” age for either type 1 or type 2
diabetes, have a multigenerational family history, no evi-
dence of autoantibodies and evidence of residual insulin
secretion. The commonest causes of monogenic diabetes
encountered in clinical practice are those caused by muta-
tions in GCK, HNF1A, and HNF4A. In such circumstances
use of a probalistic model [14] may help determine whether
it is reasonable to screen for these disorders. In other situa-
tions, appropriate clinical suspicion coupled with targeted
detection of syndromic extra‐pancreatic abnormalities
may help guide the decision to perform genetic testing and
identify other less common forms of monogenic diabetes.
References
1. Smushkin G, Vella A. What is type 2 diabetes? Medicine
(Baltimore). 2010;38:597–601.
2. Gale EA. Declassifying diabetes. Diabetologia.
2006;49:1989–1995.
3. Fajans SS, Bell GI, Polonsky KS. Molecular mechanisms and
clinical pathophysiology of maturity‐onset diabetes of the
young. N Engl J Med. 2001;345:971–980.
4. Hattersley AT, Pearson ER. Minireview: pharmacogenetics
and beyond: the interaction of therapeutic response, beta‐
cell physiology, and genetics in diabetes. Endocrinology.
2006;147:2657–2663.
5. Murphy R, Ellard S, Hattersley AT. Clinical implications of a
molecular genetic classification of monogenic beta‐cell
diabetes. Nat Clin Pract Endocrinol Metab. 2008;4:200–213.
 How to screen appropriately for Monogenic Diabetes
6. Stanley CA. Perspective on the Genetics and Diagnosis of
Congenital Hyperinsulinism Disorders. J Clin Endocrinol
Metab. 2016;101:815–826.
7. Flanagan SE, Patch AM, Mackay DJ, Edghill EL, Gloyn AL,
Robinson D, Shield JP, Temple K, Ellard S, Hattersley AT.
Mutations in ATP‐sensitive K+ channel genes cause
transient neonatal diabetes and permanent diabetes in
childhood or adulthood. Diabetes. 2007;56:1930–1937.
8. Effect of intensive therapy on residual beta‐cell function in
patients with type 1 diabetes in the diabetes control and
complications trial. A randomized, controlled trial. The
Diabetes Control and Complications Trial Research Group.
Ann Intern Med. 1998;128:517–523.
9. Mechanick JI, Adams S, Davidson JA, Fergus IV, Galindo RJ,
McKinney KH, Petak SM, Sadhu AR, Samson SL, Vedanthan
R, Umpierrez GE. Transcultural Diabetes Care. in the United
States – a Position Statement by the American Association of
Clinical Endocrinologists. Endocr Pract. 2019;25:729–765.
10. Agius L. Targeting hepatic glucokinase in type 2 diabetes:
weighing the benefits and risks. Diabetes. 2009;58:18–20.
11. Awa WL, Thon A, Raile K, Grulich‐Henn J, Meissner T,
Schober E, Holl RW, DPV‐Wiss. Study Group. Genetic and
clinical characteristics of patients with HNF1A gene varia-
tions from the German‐Austrian DPV database. Eur J
Endocrinol. 2011;164:513–520.
12. Cousminer DL, Ahlqvist E, Mishra R, Andersen MK, Chesi
A, Hawa MI, Davis A, Hodge KM, Bradfield JP, Zhou K, Guy
VC, Akerlund M, Wod M, Fritsche LG, Vestergaard H,
Snyder J, Hojlund K, Linneberg A, Karajamaki A, Brandslund
I, Kim CE, Witte D, Sorgjerd EP, Brillon DJ, Pedersen O,
Beck‐Nielsen H, Grarup N, Pratley RE, Rickels MR, Vella A,
Ovalle F, Melander O, Harris RI, Varvel S, Grill VER, Bone
Mineral Density in Childhood S, Hakonarson H, Froguel P,
Lonsdale JT, Mauricio D, Schloot NC, Khunti K, Greenbaum
CJ, Asvold BO, Yderstraede KB, Pearson ER, Schwartz S,
71
Voight BF, Hansen T, Tuomi T, Boehm BO, Groop L, Leslie
RD, Grant SFA. First genome‐wide association study of
latent autoimmune diabetes in adults reveals novel insights
linking immune and metabolic diabetes. Diabetes Care.
2018;41:2396–2403.
13. Mishra R, Akerlund M, Cousminer DL, Ahlqvist E, Bradfield
JP, Chesi A, Hodge KM, Guy VC, Brillon DJ, Pratley RE,
Rickels MR, Vella A, Ovalle F, Harris RI, Melander O, Varvel
S, Hakonarson H, Froguel P, Lonsdale JT, Mauricio D,
Schloot NC, Khunti K, Greenbaum CJ, Yderstraede KB,
Tuomi T, Voight BF, Schwartz S, Boehm BO, Groop L, Leslie
RD, Grant SFA. Genetic discrimination between LADA and
childhood‐onset type 1 diabetes within the MHC. Diabetes
Care. 2020;43:418–425.
14. Shields BM, McDonald TJ, Ellard S, Campbell MJ, Hyde C,
Hattersley AT. The development and validation of a clinical
prediction model to determine the probability of MODY in
patients with young‐onset diabetes. Diabetologia.
2012;55:1265–1272.
15. Hattersley AT, Patel KA. Precision diabetes: learning from
monogenic diabetes. Diabetologia. 2017;60:769–777.
16. Bock G, Chittilapilly E, Basu R, Toffolo G, Cobelli C,
Chandramouli V, Landau BR, Rizza RA. Contribution of
hepatic and extrahepatic insulin resistance to the pathogen-
esis of impaired fasting glucose: role of increased rates of
gluconeogenesis. Diabetes. 2007;56:1703–1711.
17. Han SJ, Kim HJ, Kim DJ, Lee KW, Cho NH. Incidence and
predictors of type 2 diabetes among Koreans: a 12‐year fol-
low up of the Korean Genome and Epidemiology Study.
Diabetes Res Clin Pract. 2017;123:173–180.
18. Spyer G, Hattersley AT, Sykes JE, Sturley RH, MacLeod KM.
Influence of maternal and fetal glucokinase mutations in ges-
tational diabetes. Am J Obstet Gynecol. 2001;185:240–241.
19. Smushkin G, Vella A. Genetics of type 2 diabetes. Curr Opin
Clin Nutr Metab Care. 2010;13:471–477.
PART II
Initial Evaluation and Management
of Diabetes
 6 Managing Gestational Diabetes During
and After Pregnancy
Aoife M. Egan1 and Fidelma P. Dunne2
1
Division of Endocrinology, Diabetes, Metabolism and nutrition, Mayo Clinic, Rochester, MN, USA
2
Galway Diabetes Research Centre, National University of Ireland, Galway, Ireland
LE A R NI NG P OI NT S
●● Gestational Diabetes Mellitus (GDM) is a common
disorder associated with adverse pregnancy outcomes.
●● Although the ideal diagnostic strategy remains unclear,
it is typically identified using an oral glucose tolerance
test (OGTT) at 24–28 weeks’ gestation.
●● The majority of women can successfully treat GDM
with lifestyle changes.
●● Women with GDM have a high future risk of type 2
diabetes and require lifelong surveillance.
Introduction
Gestational diabetes mellitus (GDM) is defined as glucose
intolerance with onset or first recognition during preg­
nancy, excluding those with diabetes likely to represent
overt diabetes mellitus  [1]. It affects approximately 13%
pregnancies and is associated with adverse pregnancy out­
comes  [2]. Risk factors for GDM reflect those for type 2
diabetes and include: overweight or obese body mass
index, GDM in a prior pregnancy (or previous delivery of a
macrosomic baby), increasing age, non‐European ethnic­
ity, family history of Type 2 diabetes and conditions associ­
ated with insulin resistance such as polycystic ovarian
syndrome [3].
Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
In order to preserve carbohydrate for the growing fetus,
insulin resistance increases in the second and early third
trimesters of pregnancy [4]. GDM will develop if the com­
pensatory pancreatic response with increased insulin
secretion is inadequate. Due to the transfer of glucose
across the placenta down a concentration gradient, mater­
nal hyperglycemia will lead to fetal hyperglycemia with
subsequent fetal hyperinsulinemia, which drives the devel­
opment of pregnancy complications [5]. The largest impact
of GDM involves increasing birth weight and the propor­
tion of infants that are classified as large for gestational age
(LGA) [6]. Beyond this, a diagnosis of GDM also confers
an increased risk of multiple additional complications
including pre‐eclampsia  [7], perineal trauma  [8], and
cesarean delivery  [9] for the mother; and neonatal birth
injury, neonatal hypoglycemia  [10], jaundice, and even
stillbirth [11] for the infant. GDM is a strong risk factor for
future type 2 diabetes in the mother (relative risk (RR)
compared to women without GDM 7.43, 95% confidence
interval (CI) 4.79–11.51)  [12], and offspring of women
with GDM are also at high risk of metabolic complications
including obesity and diabetes [13, 14].
GDM diagnosis
One might argue that an accurate diagnosis is critical for
optimal management of any medical condition. However,
75
 76 Initial Evaluation and Management of Diabetes

in the case of GDM this is fraught with difficulty. Central to
this is the strong, continuous association between increas­
ing maternal glucose levels and key adverse outcomes such
as increased birth weight, neonatal hypoglycemia and pri­
mary cesarean delivery [10]. The lack of an obvious thresh­
old at which risk increases makes it challenging to choose
an appropriate diagnostic cut‐off, especially once addi­
tional factors such as clinical benefit and cost‐effectiveness
of treatments are considered. This has resulted in the use of
countless strategies for screening and diagnosing GDM
that vary from country to country and even center to
center [15, 16]. Table 6.1 outlines the more commonly used
approaches internationally. One should note that the one‐
step, International Association of Diabetes and Pregnancy
Study Groups (IADPSG)/World Health Organization
(WHO) criteria were chosen primarily based on an odds
ratio for adverse outcomes of at least 1.75 compared to
women with mean glucose levels at 24–28  weeks in the
Hyperglycemia and Neonatal Outcomes (HAPO)
study [10, 17, 18]. These criteria based on pregnancy out­
comes are also endorsed by a number of additional organi­
zations including the European Board and College of
Obstetrics and Gynaecology (EBCOG)  [19], the
International Federation of Gynecology and Obstetrics
(FIGO)  [20] and the European Association of Perinatal
Medicine (EAPM)  [21]. The American Diabetes
Association (ADA) also endorses this approach but state
that alternative, two‐step approaches are also accepta­
ble  [22]. The diagnostic cut‐offs used in the two‐step
approaches are largely based on thresholds that optimize
predictive ability for subsequent development of maternal
diabetes, rather than pregnancy outcomes [23].
The majority of clinical centers conduct GDM screening
at 24–28 weeks gestation, but it is known that accelerated

TABLE 6.1  Options for GDM screening and diagnosis.

Number of abnormal Glucose cut‐offs

values required Oral glucose load mg/dL (mmol/L)

Two‐step strategies:
Non‐fasting glucose challenge test 1 50 g ≥ 130, 135 or 140
(7.2, 7.5 or 7.8)
Followed by either of the options below:
1. Carpenter and Coustan ≥2a 100 g Fasting ≥ 95 (5.3)
1 hour ≥ 180 (10.0)
2 hour ≥ 155 (8.6)
3 hour ≥ 140 (7.8)
2. NDDG ≥2a 100 g Fasting ≥ 105 (5.8)
1 hour ≥ 190 (10.6)
2 hour ≥ 165 (9.2)
3 hour ≥ 145 (8.0)
One‐step strategies:
2010 IADPSG/2013 WHO ≥1 75 g Fasting ≥ 92 (5.1)
1 hour ≥ 180 (10.0)
2 hour ≥ 153 (8.5)
2015 NICE ≥1 75 g Fasting ≥ 5.6 (100)
2 hour ≥ 7.8 (140)
EASD ≥1 75 g Fasting ≥ 6.0 (108)
2 hour ≥ 9.0 (162)

GDM: Gestational diabetes mellitus; IADPSG: International Association of Diabetes in Pregnancy Study Groups; NICE: National
Institute for Health and Care Excellence; OGTT: oral glucose tolerance test; WHO: World Health Organization; EASD: European
Association for the Study of Diabetes.
a
 ACOG notes that one elevated value may be used for diagnosis.
 Managing Gestational Diabetes During and After Pregnancy
fetal growth may occur earlier in pregnancy in women with
GDM [24]. A significant amount of work including a large
randomized controlled trial is underway to examine
whether diagnosing and treating GDM in early pregnancy
improves pregnancy outcomes [25]. In the meantime, the
American College of Obstetricians and Gynecologists
(ACOG) accept the 24‐week criteria applied to a 100 g oral
glucose tolerance test (OGTT) in early pregnancy, but
there is minimal evidence to support this practice and the
IADPSG have indicated that their criteria are not suitable
for use in early pregnancy  [26, 27]. What is clear, is that
women with risk factors for type 2 diabetes should be tested
for undiagnosed pre‐gestational diabetes using standard
diagnostic criteria at their initial prenatal visit [22, 26].
GDM management
Glycemic goals
The following glycemic goals are recommended for women
with GDM: fasting glucose ≤ 95  mg/dL (5.3  mmol/L), 1
hour after eating ≤ 140 mg/dL (7.8 mmol/L) and 2 hours
after eating ≤ 120 mg/dL (6.7 mmol/L) [22, 26, 28].
Lifestyle modification
Over 70% women with GDM diagnosed using a two‐step
approach can be treated with lifestyle modifications
alone [22], and use of the more stringent IADPSG criteria
increases this proportion [9]. Women should meet with a
registered dietician or registered dietician nutritionist who
is familiar with the management of GDM  [22]. An indi­
vidualized plan should be developed, taking into account
baseline BMI, cultural, and other dietary preferences. A
minimum of 175 g of carbohydrates, 71 g of protein and 28
g fiber is recommended based on the Dietary Reference
Intakes for pregnant women  [22]. Although there is no
clear guidance on the ideal total calorie intake, limiting
carbohydrates to 34–45% of total calories per day is recom­
mended by the Endocrine Society [29]. Added sugars and
refined carbohydrates should be excluded from the diet
with a focus on carbohydrate sources such as vegetables,
lower glycemic index fruits and whole grains [30].
Even light postprandial exercise decreases postprandial
glucose excursions in women with GDM [31] and physical
activity will maintain or improve physical fitness and assist
77
with weight management. For most women, an exercise
program with a goal of moderate‐intensity exercise of
20–30  minutes per day is reasonable  [32]. Observational
studies suggest that in routine clinical practice, the intro­
duction of lifestyle modifications and remaining within
recommended guidelines for gestational weight gain are
associated with improved clinical outcomes including a
reduction in large‐for‐gestational age infants  [33, 34]. A
meta‐analysis of trials involving lifestyle interventions also
demonstrated a reduction in large‐for‐gestational age neo­
nates and neonatal fat mass in those randomized to the
intervention [35].
Fasting and postprandial self‐monitoring of blood glu­
cose is recommended in women diagnosed with GDM [22].
Intensification of therapy should occur within 1–2 weeks if
glycemic targets are not met following lifestyle interven­
tion – this is generally defined as > 15–20% glucose read­
ings above goal. In addition, it is common practice for
pharmacological therapy to be introduced at diagnosis if
the fasting glucose level is ≥ 126  mg/dL (7.0  mmol/L) as
this makes the presence of type 2 diabetes highly likely [36].
Insulin therapy
Insulin is the most commonly used agent for treatment of
GDM and reassuringly, it has not been demonstrated to
cross the placenta [37–39]. Traditionally isophane insulin
(neutral protamine Hagedorn, NPH) was used as interme­
diate‐acting insulin (two‐three times per day) but the long‐
acting insulin analogues detemir and glargine are
increasingly used and appear safe [40, 41]. Although more
extensively studied in the setting of pregestational diabetes,
aspart and lispro are also safe during pregnancy and are
frequently used to treat GDM [42, 43]. A randomized con­
trolled, open‐label trial is currently underway comparing
aspart with faster‐acting aspart insulin (Fiasp) in women
with pre‐existing diabetes with completion estimated in
2022 [44]. In addition, a randomized trial is underway to
evaluate insulin degludec versus insulin detemir in the
treatment of pregnant women with type 1 diabetes and is
expected to be complete in 2021 [45]. Results from these
trials will be useful when considering these newer insulin
options for women with GDM.
Insulin regimens are typically tailored according to the
pattern of maternal hyperglycemia – for example, if the
 78 Initial Evaluation and Management of Diabetes
primary concern is fasting hyperglycemia, intermediate or
long‐acting insulin is initiated initially, with rapid‐acting
insulin introduced to manage postprandial hyperglycemia.
If a combination of long‐acting/intermediate‐acting and
short‐acting insulin is required at the outset, ACOG sug­
gest a starting total daily dosage of 0.7–1.0units/kg which
may be divided with a regimen of multiple injections [26].
Women requiring basal‐bolus insulin therapy should moni­
tor fasting, pre meal and postprandial glucose levels to allow
for better adjustment of insulin  [22]. As pregnancy pro­
gresses, postprandial glucose control can be impaired by
slower glucose disposal [46], and the time‐to‐peak prandial
insulin concentration is increasingly delayed [47]. Bolusing
meal insulin at least 15–30 minutes before eating may there­
fore ameliorate postprandial glycemic excursions.
Insulin therapy is very effective for treating hyperglyce­
mia and in “real‐world” clinical settings, it appears to
reduce rates of large‐for‐gestational age towards that of
women without GDM [48]. The main clinical concern with
insulin use is risk of maternal hypoglycemia and additional
instruction is necessary to educate women on self‐injection
and recognition and treatment of hypoglycemia.
Oral glucose‐lowering agents
Although not approved for use in pregnancy, metformin
and glyburide (glibenclamide) are commonly used for
treatment of GDM [49]. Metformin is typically introduced
at a dose of 500 mg with the evening meal and titrated up
to 1000 mg twice daily. It is a biguanide and exhibits its glu­
cose‐lowering effect by suppressing gluconeogenesis and
enhancing insulin suppression of endogenous glucose pro­
duction in the liver. In addition, it possibly improves glu­
cose uptake and utilization by peripheral tissues and
reduces intestinal glucose absorption  [50]. Slow upward
titrations in the dose of metformin can reduce the risk of
gastrointestinal disturbance. It has been used since the late
1970s for women with type 2 diabetes and GDM, particu­
larly in South Africa; and is commonly used in women
with polycystic ovarian syndrome pre‐conceptually and
during pregnancy [51, 52].
Metformin does cross the placenta in significant quanti­
ties, and this has raised concerns in relation to both long‐ and
short‐term offspring effects  [53]. A large population‐
based case‐control study (using controls with congenital
malformations) did not find any evidence for an increased
risk of all non‐genetic congenital anomalies combined fol­
lowing first trimester metformin exposure, however, the
risk of pulmonary valve atresia was increased among met­
formin exposed babies (aOR 3.54, CI 1.05 – 12.00 com­
pared with controls with non‐genetic anomalies) [54]. The
Metformin in Gestational Diabetes (MiG) trial randomized
751 women to metformin or insulin and there was no sig­
nificant difference in the composite fetal outcome (neona­
tal hypoglycemia, respiratory distress, need for
phototherapy, birth trauma, 5‐minute Apgar < 7 or prema­
turity) between the two groups. However, there was a
higher rate of preterm birth among the metformin‐exposed
infants (RR 1.6, CI 1.02–2.52, p=0.04) [55]. Another point
to note is that within the metformin group, 46.3% required
supplemental insulin, but the women preferred metformin
to insulin treatment.
A follow‐up study (MiG trial participants) observed
that children exposed to metformin (or a combination of
insulin and metformin) in utero had similar total and
abdominal body fat percent and metabolic markers at 7–9
years compared to those exposed to insulin only. However,
at 9 years, metformin exposed offspring were larger by
measures of weight, arm and waist circumferences [56]. A
meta‐analysis examining outcomes of GDM‐affected preg­
nancies randomized to treatment with metformin versus
insulin (28 studies, 3976 participants) found that met­
formin exposed children had a lower average birth weight
(mean difference −107.7 g, 95% confidence interval −182.3
to −32.7, p = 0.005) but appeared to experience accelerated
postnatal growth resulting in higher BMI by mid‐child­
hood (mean difference 0.78  kg/m2, 95% confidence
interval 0.23 to 1.33, p=0.005) [57]. Whether these dif­
ferences will result in long‐term cardio‐metabolic out­
comes remains unknown. We await the results of the
first placebo‐controlled metformin trial in GDM, which
is currently recruiting with estimated completion in
2022 [58].
Glyburide is a sulfonylurea and exerts its glucose‐lower­
ing effect by stimulating insulin secretion from the pancre­
atic beta cell. The primary mechanism of action is through
closure of the APT‐sensitive potassium channels in the
 Managing Gestational Diabetes During and After Pregnancy
plasma membrane of the beta cell. This initiates a chain of
events culminating in insulin release  [59]. Glyburide is
generally used at total daily doses of 2.5–20.0mg, although
a dose‐response relationship in pregnancy is not clear [60].
In 2000, the first randomized controlled trial compared
glyburide to insulin in 404 women with GDM [61]. Four
mothers in the glyburide group required supplemental
insulin and there were no significant difference in multiple
adverse outcomes including large‐for‐gestational age,
hypoglycemia and admission to a neonatal intensive care
unit. The study also reported that glyburide was not
detected in the cord serum of any infant in the glyburide
group, however, subsequent work has demonstrated that
glyburide does cross the placenta [62]. Likely based on the
reassuring results of the aforementioned randomized con­
trolled trial, glyburide use for treatment of GDM increased
from 7.4% to 64.5% between 2000 and 2011 in the United
States, replacing insulin as the most common pharmaco­
therapy [49]. More recently, concern has been raised about
safety of glyburide in pregnancy  [60]. A meta‐analysis
including 2509 randomized‐controlled trial participants
reported that glyburide resulted in higher birth weight
and more macrosomia and neonatal hypoglycemia com­
pared with insulin therapy [63]. Interestingly, the same
study reported that metformin was associated with less
maternal weight gain, macrosomia and lower birth‐
weight but increased risk of preterm birth compared to
Excessive
Maternal
gestational weight
hyperinsulinemia
gain
Maternal
Maternal obesity
hyperglycemia
Excessive
fetal
growth
79
glyburide [63]. There are no data with respect to long‐
term outcomes in infants exposed to glyburide in utero.
Guidelines and providers vary in terms of their approach to
pharmacological intervention for GDM treatment. The ADA
recommends insulin as first‐line therapy but do not express a
preference of metformin over glyburide in situations where
insulin may not be feasible. Although metformin is likely
associated with a reduction in hypertensive disorders of preg­
nancy  [64], it is not recommended for women with estab­
lished preeclampsia, hypertension or those at risk of
intrauterine growth restriction [22]. ACOG also recommend
insulin as first‐line therapy but suggest metformin (or rarely
glyburide) as a reasonable alternative [26]. On the other hand,
the National Institute for Health and Care Excellence (NICE)
in the United Kingdom recommend metformin as first‐line
therapy with insulin as the next step [65].
Gestational weight gain
Despite what appears to be excellent glycemic control,
women with diabetes can still have a fetus that is mac­
rosomic [66]. This may be due to the presence of hypergly­
cemia overlooked on standard glucose testing. However, in
addition to hyperglycemia, prepregnancy BMI and exces­
sive gestational weight gain play key roles in fetal growth
(Figure 6.1). This is demonstrated in study findings where
overweight women with well controlled GDM have a 50%
greater risk of having a macrosomic baby than women of
FIGURE 6.1  Key factors contributing to
excessive fetal growth.
 80 Initial Evaluation and Management of Diabetes
TABLE 6.2  Guidelines for gestational weight gain based
on prepregnancy BMI.
Recommended total
BMI (kg/m2) weight gain (kg)
< 18.5 12.5 – 18.0
18.5–24.9 11.5 – 16.0
25.0–29.9 7.0 – 11.5
≥ 30.0 5.0 – 9.0
BMI: body mass index
normal weight with well‐controlled GDM [67]. Increasing
evidence points towards factors other than glucose as play­
ing a key role in modulating fetal overgrowth – including
maternal lipids [68]. While it is not considered appropriate
to advise weight loss once pregnancy is established, a
restriction on gestational weight gain may help mitigate
risk of fetal overgrowth, particularly in women who are
overweight or obese at baseline. In women with GDM, ges­
tational weight gain that remains within Institute of
Medicine Guidelines  [69] (Table  6.2) is associated with
lower risk of having a large‐for‐gestational age infant and
gestational hypertension [33].
Fetal monitoring
Antenatal fetal testing in women with poorly controlled or
medication‐requiring GDM without other complications is
typically initiated at 28–32  weeks’ gestation  [26, 65].
Guidance does not specify the approach to monitoring
women with GDM who are well controlled with lifestyle
interventions. However, antenatal fetal testing is generally
started at some point from 28 weeks’ gestation [26, 65]. The
optimal delivery time for women with GDM is not clear. A
recent trial randomized 425 women with GDM to induc­
tion at 38 weeks or expectant management up to 41 weeks’
gestation. Although the study was underpowered, no dif­
ference was detected in birth outcomes between
groups  [70]. An older systematic review (including one
randomized trial and four observational studies), found
that active rather than expectant management of labor at
term for women with GDM may reduce rates of large‐for‐
gestational age infants. However, heterogeneity of included
studies was a significant issue in the review [71]. A reason­
able approach in the setting of uncomplicated GDM would
be to manage expectantly until term with induction taking
placing between 39 + 0 and 40 + 6  weeks’ gestation. For
women requiring pharmacological intervention, delivery
between 39 + 0 and 39 + 6  weeks is generally accepted.
Earlier delivery should be considered on a case‐by‐case basis
for women with additional complications. For those with an
estimated fetal weight of ≥ 4500 g on ultrasound, the option
of elective cesarean delivery should be discussed [26]. Third
trimester anesthesiology consultation is helpful for many
women to discuss the birth plan and discuss options for pain
management during labor and delivery. It should be particu­
larly considered in women with additional complications
including morbid obesity. Women with diabetes should give
birth in a location where advanced neonatal resuscitation
skills are available continuously [65].
Glucocorticoid therapy
Glucocorticoids are commonly used to accelerate fetal lung
maturation when early delivery is anticipated, but they
have the propensity to significantly exacerbate hyperglyce­
mia in women with GDM. The hyperglycemia usually
starts about 6 hours after initial glucocorticoid dosing but
remains for at least 24 hours after the last dose. Additional
glucose monitoring is advised following steroid adminis­
tration and supplemental insulin is typically administered
intravenously with infusion rates adjusted hourly, targeting
a glucose of 70–126  mg/dL (3.9–7.0  mmol/L)  [72]. It is
important to monitor infants carefully for neonatal hypo­
glycemia, particularly if the glucocorticoids were adminis­
tered close to delivery.
Intrapartum management
Intrapartum hyperglycemia has been linked with neonatal
hypoglycemia and most local policies specify a glycemic
goal range that falls between 70–126  mg/dL (3.9–
7.0 mmol/L) [73]. IV insulin is frequently used in women
who required insulin during their pregnancy and insulin
should also be initiated in insulin‐naïve women who are
outside of this glycemic range. A randomized controlled
trial published in 2019 randomized 76 women with GDM
to tight and liberalized glucose goal ranges during labor.
“Tight control” constituted glucose measurements hourly
and treatment for maternal glucose levels lower than
 Managing Gestational Diabetes During and After Pregnancy
60  mg/dL (3.3  mmol/L) or > 100  mg/dL (5.6  mmol/L).
“Liberalized control” involved glucose measurements
every 4 hours and treatment for maternal glucose levels <
60 mg/dL (3.3 mmol/L) or > 120 mg/dL (6.7 mmol/L). The
primary outcome of neonatal blood glucose levels was sim­
ilar between both groups [53 mg/dL (2.9 mmol/L) versus
58  mg/dL (3.2  mmol/L), mean difference −4.18, 95% CI
−12.66 to 4.29]. However, mean neonatal glucose level
within the first 24 hours of life was lower in the tight con­
trol group [54  mg/dL (3.0  mmol/L) vs 58  mg/dL
(3.2  mmol/L), mean difference −3.39, 95% CI −7.07 to
0.29] [74].
Postpartum neonatal care
Ideally, infants will remain with the mother on delivery and
feeding should be encouraged as soon as possible after
birth. Capillary blood glucose testing should take place 2–4
hours after birth and if capillary plasma glucose values are
< 36  mg/dL (2  mmol/L) on two consecutive occasions
despite feeding, or if there are abnormal clinical signs,
additional supports such as intravenous glucose or tube
feeding will be required [65].
A Canadian‐based retrospective cohort study identified
that GDM was associated with lower rates of breastfeeding
in hospital and on discharge  [75]. With this in mind,
women with GDM should receive extra lactation support
postpartum, as breastfeeding is associated with nutritional
and immunological benefits for the baby and a reduced
risk of postpartum dysglycemia (8.2% vs 18.4%, p< 0.001 in
a prospective cohort study) [76]. Benefits of breastfeeding
were also demonstrated in the 30‐year CARDIA study
where lactation duration showed an inverse association
with diabetes incidence – adjusted HR for > 0 to 6 months:
0.75 (95% CI, 0.51–1.09); > 6 months to < 12 months: 0.52
(95% CI, 0.31–0.87); and ≥ 12 months 0.53 (0.29–0.98) vs
no breastfeeding [77].
Postpartum maternal care
Postpartum, insulin resistance decreases dramatically and
any glucose lowering therapies should be discontinued. It is
recommended that women with GDM should have blood
glucose testing prior to discharge from hospital to exclude
persistent hyperglycemia [65]. Formal screening for a per­
sistent glucose disorder typically takes place between 4 and
81
12  weeks postpartum. This should include a 75 g OGTT
with standard criteria used to diagnose diabetes or pre‐dia­
betes [26]. Women with persistent diabetes should receive
appropriate treatment and metformin should be consid­
ered in women with prediabetes as it may delay progres­
sion to type 2 diabetes in this situation [78]. Attendance at
postpartum glucose testing is often poor. However, pro­
grams with a dedicated coordinator who makes verbal and
written contact can achieve up to a 75% recall rate [79].
At a median follow‐up of 11.4 years, 52.2% women with
GDM developed a disorder of glucose metabolism versus
20.1% without GDM (OR 3.44, 95% CI 2.85 – 4.14).
Women with GDM should therefore have lifelong, regular
assessment of glucose status at 1‐ to 3‐year intervals [22].
Inter‐pregnancy weight gain is a risk for adverse obstetric
outcomes including GDM, pre‐eclampsia and still­
birth [80]. The postpartum visit is an ideal opportunity to
discuss future pregnancy planning, review contraceptive
options, and provide assistance with weight loss if
indicated.
Future directions
As the prevalence of GDM continues to rise in line with the
worldwide epidemic of type 2 diabetes, a large number of
high‐quality trials have attempted to develop effective
interventions for GDM prevention [6]. A 2017 Cochrane
meta‐analysis of trials evaluating the combination of diet
and exercise interventions for GDM prevention concluded
that there was a possible reduced risk of GDM in the inter­
vention groups (RR 0.85, 95% CI 0.71–1.01; 6633 women;
19 RCTs) but study heterogeneity limited ability to inform
practice  [81]. One observation is that women at risk of
GDM are often reluctant to engage in lifestyle changes and
find it difficult to adhere to recommended interventions
even within a trial setting [6]. Metformin has been exam­
ined for GDM prevention in two randomized controlled
trials and it was ineffective in achieving this outcome [82,
83]. Indeed, GDM is a heterogenous disorder, in terms of
its time of onset and the underlying pathophysiology that
may involve predominantly insulin resistance, impaired
insulin secretion or both. This makes it difficult to envisage
a “one size fits all approach”, and successful prevention and
treatment strategies will likely require personalized pro­
grams depending on individual phenotype.
 82 Initial Evaluation and Management of Diabetes
The optimal strategy for diagnosis of GDM continues to
be debated. Each approach will identify different degrees of
hyperglycemia and risk for adverse outcomes  [60].
Identifying GDM in early pregnancy is particularly chal­
lenging as early pregnancy hyperglycemia may have
resolved by late second trimester [84], and as already men­
tioned, the clinical benefit of treating women from early
pregnancy is not clear. Numerous studies have evaluated
alternative biomarkers for detection of GDM in maternal
plasma and urine  [85]. One example is plasma glycated
CD59, which has high sensitivity and specificity for identi­
fying the risk for LGA when measured at 24–28 weeks’ ges­
tation [86]. Another diagnostic option under investigation
involves the use of machine‐learning to predict GDM. This
has the potential to provide a cost‐effective screening
approach that avoids the need for a cumbersome
OGTT [87].
While it is clear that much work is needed to develop
novel approaches for GDM treatment and evaluate long‐
term effects of current treatment strategies, future efforts
should also address the risk of type 2 diabetes. Our under­
standing of the factors driving β‐cell failure or recovery
after GDM is limited [88], and further mechanistic studies
are needed to evaluate islet cell function postpartum. For
example, prior work indicates that compared to women
without GDM, women with GDM have elevated fasting
proinsulin levels during pregnancy and 8–14 weeks post­
partum  [89, 90]. An increase in the fasting proinsulin to
insulin ratio has been used as a marker of impaired β‐cell
function and is associated with rapid conversion to type 2
diabetes in people with prediabetes [91]. With this in mind,
measurement of how prandial proinsulin and insulin
secretion change relative to one another, and changes in the
proinsulin to insulin ratio over time may provide insights
into β‐cell function postpartum. Ultimately, identifying
accurate markers of postpartum β‐cell health will generate
data to assist in developing effective type 2 diabetes preven­
tion programs for women with GDM.
Conclusion
The prevalence of GDM is high and represents a significant
public health concern. Women with GDM benefit from a
multidisciplinary approach to care in order to support
normal fetal growth and reduce maternal complications.
Postpartum efforts should focus on screening for persistent
diabetes and reducing risk of future type 2 diabetes. It is
hoped that ongoing work will help identify the best diag­
nostic and treatment strategies for GDM.
References
1. Egan AM, Vellinga A, Harreiter J et al. Epidemiology of
gestational diabetes mellitus according to IADPSG/WHO
2013 criteria among obese pregnant women in Europe.
Diabetologia. 2017;60(10):1913–1921.
2. International Diabetes Federation. IDF Atlas. Brussels,
Belgium 2013.
3. Avalos GE, Owens LA, Dunne F, Collaborators AD.
Applying current screening tools for gestational diabetes
mellitus to a European population: is it time for change?
Diabetes Care. 2013;36(10):3040–3044.
4. Catalano P, Huston L, Amini S, Kalhan S. Longitudinal
changes in glucose metabolism during pregnancy in obese
women with normal glucose tolerance and gestational
diabetes. Am J Obstst Gyneol. 1999;180(4):903–916.
5. Pedersen J. Diabetes and pregnancy: blood sugar of newborn
infants (PhD Thesis). Copenhagen, Danish Science Press; 1952.
6. Egan AM, Simmons D. Lessons learned from lifestyle
prevention trials in gestational diabetes mellitus. Diabet
Med. 2019;36(2):142–150.
7. Yogev, Chen, Hod et al. Hyperglycemia and Adverse
Pregnancy Outcome (HAPO) study: preeclampsia. Am J
Obstet Gynecol. 2010;202(3):255.e251–257.
8. Jastrow N, Roberge S, Gauthier RJ et al. Effect of birth weight
on adverse obstetric outcomes in vaginal birth after cesarean
delivery. Obstet Gynecol. 2010;115(2 Pt 1):338–343.
9. O’Sullivan EP, Avalos G, O’Reilly M et al. Atlantic Diabetes
in Pregnancy (DIP): the prevalence and outcomes of
gestational diabetes mellitus using new diagnostic criteria.
Diabetologia. 2011;54(7):1670–1675.
10. Metzger BE, Lowe LP, Dyer AR et  al. Hyperglycemia and
adverse pregnancy outcomes. N Engl J Med. 2008;358(19):
1991–2002.
11. Sweeting AN, Ross GP, Hyett J et  al. Gestational diabetes
mellitus in early pregnancy: evidence for poor pregnancy
outcomes despite treatment. Diabetes Care. 2016;39(1):
75–81.
12. Bellamy L, Casas JP, Hingorani AD, Williams D. Type 2 dia­
betes mellitus after gestational diabetes: a systematic review
and meta‐analysis. Lancet. 2009;373(9677):1773–1779.
 Managing Gestational Diabetes During and After Pregnancy
13. Guerrero‐Romero F, Aradillas‐García C, Simental‐Mendia
LE, Monreal‐Escalante E, de la Cruz Mendoza E, Rodríguez‐
Moran M. Birth weight, family history of diabetes, and met­
abolic syndrome in children and adolescents. J Pediatr.
2010;156(5):719–723, 723.e711.
14. Whincup PH, Kaye SJ, Owen CG et al. Birth weight and risk
of type 2 diabetes: a systematic review. JAMA.
2008;300(24):2886–2897.
15. Benhalima K, Mathieu C, Van Assche A et al. Survey by the
European Board and College of Obstetrics and Gynaecology
on screening for gestational diabetes in Europe. Eur J Obstet
Gynecol Reprod Biol. 2016;201:197–202.
16. Bilous R. Diagnosis of gestational diabetes, defining the net,
refining the catch. Diabetologia. 2015;58(9):1965–1968.
17. Metzger BE, Gabbe SG, Persson B et al. International asso­
ciation of diabetes and pregnancy study groups recommen­
dations on the diagnosis and classification of hyperglycemia
in pregnancy. Diabetes Care. 2010;33(3):676–682.
18. World Health Organization. Diagnostic criteria and classifica­
tion of hyperglycaemia first detected in pregnancy: a World
Health Organization guideline. 2013. Available at: https://
apps.who.int/iris/bitstream/handle/10665/85975/WHO_
NMH_MND_13.2_eng.pdf. Accessed 27 March, 2020.
19. Benhalima K, Mathieu C, Damm P et al. A proposal for the
use of uniform diagnostic criteria for gestational diabetes in
Europe: an opinion paper by the European Board & College
of Obstetrics and Gynaecology (EBCOG). Diabetologia.
2015;58(7):1422–1429.
20. Hod M, Kapur A, Sacks DA et  al. The International
Federation of Gynecology and Obstetrics (FIGO) Initiative
on gestational diabetes mellitus: a pragmatic guide for diag­
nosis, management, and care. Int J Gynaecol Obstet. 2015;131
Suppl 3:S173–211.
21. Hod M, Pretty M, Mahmood T, FIGO, EAPMand EBCOG.
Joint position statement on universal screening for GDM in
Europe by FIGO, EBCOG and EAPM. Eur J Obstet Gynecol
Reprod Biol. 2018;228:329–330.
22. American Diabetes Association. 14. Management of diabe­
tes in pregnancy: standards of medical care in diabetes –
2020. Diabetes Care. 2020;43(Suppl 1):S183–S192.
23. Ferrara A, Hedderson M, Quesenberry C, Selby J. Prevalence
of gestational diabetes mellitus detected by the National
Diabetes Data Group or the Carpenter and Coustan plasma
glucose thresholds. Diabetes Care. 2020;25(9):1625–1630.
24. Sovio U, Murphy HR, Smith GC. Accelerated fetal growth
prior to diagnosis of gestational diabetes mellitus: a prospec­
tive cohort study of nulliparous women. Diabetes Care.
2016;39(6):982–987.
83
25. Simmons D, Hague WM, Teede HJ et al. Hyperglycaemia in
early pregnancy: the Treatment of Booking Gestational dia­
betes Mellitus (TOBOGM) study. A randomised controlled
trial. Med J Aust. 2018;209(9):405–406.
26. ACOG Practice Bulletin No. 190 Summary: gestational dia­
betes mellitus. Obstet Gynecol. 2018;131(2):406–408.
27. McIntyre HD, Sacks DA, Barbour LA et al. Issues with the
diagnosis and classification of hyperglycemia in early preg­
nancy. Diabetes Care. 2016;39(1):53–54.
28. Metzger BE, Buchanan TA, Coustan DR et al. Summary and
recommendations of the Fifth International Workshop‐
Conference on Gestational Diabetes Mellitus. Diabetes Care.
2007;30 Suppl 2:S251–260.
29. Blumer I, Hadar E, Hadden DR et  al. Diabetes and preg­
nancy: an endocrine society clinical practice guideline.
J Clin Endocrinol Metab. 2013;98(11):4227–4249.
30. Castorino K, Jovanovič L. Pregnancy and diabetes manage­
ment: advances and controversies. Clin Chem. 2011;57(2):
221–230.
31. García‐Patterson A, Martín E, Ubeda J, María MA, de
Leiva A, Corcoy R. Evaluation of light exercise in the treat­
ment of gestational diabetes. Diabetes Care. 2001;24(11):
2006–2007.
32. Gynecologists TACoOa. Committee Opinion Number 650:
Physical Activity and Exercise During Pregnancy and the
Postpartum Period. 2015.
33. Egan AM, Dennedy MC, Al‐Ramli W, Heerey A, Avalos G,
Dunne F. ATLANTIC‐DIP: excessive gestational weight gain
and pregnancy outcomes in women with gestational or
pregestational diabetes mellitus. J Clin Endocrinol Metab.
2014;99(1):212–219.
34. Kgosidialwa O, Egan AM, Carmody L, Kirwan B, Gunning P,
Dunne FP. Treatment With Diet and Exercise for Women
With Gestational Diabetes Mellitus Diagnosed Using
IADPSG Criteria. J Clin Endocrinol Metab. 2015;100(12):
4629–4636.
35. Brown J, Alwan NA, West J et al. Lifestyle interventions for
the treatment of women with gestational diabetes. Cochrane
Database Syst Rev. 2017;5:CD011970.
36. Egan AM, Dunne FP. Optimal management of gestational
diabetes. Br Med Bull. 2019;131(1):97–108.
37. Suffecool K, Rosenn B, Niederkofler EE et al. Insulin detemir
does not cross the human placenta. Diabetes Care. 2015;38(2):
e20–21.
38. McCance DR, Damm P, Mathiesen ER et  al. Evaluation of
insulin antibodies and placental transfer of insulin aspart in
pregnant women with type 1 diabetes mellitus. Diabetologia.
2008;51(11):2141–2143.
 84 Initial Evaluation and Management of Diabetes
39. Boskovic R, Feig DS, Derewlany L, Knie B, Portnoi G, Koren G.
Transfer of insulin lispro across the human placenta: in vitro;
perfusion studies. Diabetes Care. 2003;26(5):1390–1394.
40. Callesen NF, Damm J, Mathiesen JM, Ringholm L, Damm P,
Mathiesen ER. Treatment with the long‐acting insulin ana­
logues detemir or glargine during pregnancy in women with
type 1 diabetes: comparison of glycaemic control and preg­
nancy outcome. J Matern Fetal Neonatal Med. 2013;26(6):
588–592.
41. Mathiesen ER, Hod M, Ivanisevic M et al. Maternal efficacy
and safety outcomes in a randomized, controlled trial com­
paring insulin detemir with NPH insulin in 310 pregnant
women with type 1 diabetes. Diabetes Care. 2012;35(10):
2012–2017.
42. Mathiesen ER, Kinsley B, Amiel SA et al. Maternal glycemic
control and hypoglycemia in type 1 diabetic pregnancy: a
randomized trial of insulin aspart versus human insulin in
322 pregnant women. Diabetes Care. 2007;30(4):771–776.
43. Garg SK, Frias JP, Anil S, Gottlieb PA, MacKenzie T, Jackson
WE. Insulin lispro therapy in pregnancies complicated by
type 1 diabetes: glycemic control and maternal and fetal out­
comes. Endocr Pract. 2003;9(3):187–193.
44. Insulin Fiasp vs. Insulin Novorapid during pregnancy and lac­
tation in women with pre‐existing diabetes. ClinicalTrials.gov.
Available at: https://clinicaltrials.gov/ct2/show/NCT03770767.
Accessed 27 March, 2020.
45. Research Study Comparing Insulin Degludec to Insulin
Detemir, Together With Insulin Aspart, in Pregnant Women
With Type 1 Diabetes (EXPECT). clinicaltrials.gov. Available
at: https://clinicaltrials.gov/ct2/show/NCT03377699. Accessed
27 March, 2020.
46. Murphy HR, Elleri D, Allen JM et  al. Pathophysiology of
postprandial hyperglycaemia in women with type 1 diabetes
during pregnancy. Diabetologia. 2012;55(2):282–293.
47. Goudie RJ, Lunn D, Hovorka R, Murphy HR. Pharma­
cokinetics of insulin aspart in pregnant women with type 1
diabetes: every day is different. Diabetes Care. 2014;37(6):
e121–122.
48. Bogdanet D, Egan AM, Reddin C et  al. ATLANTIC DIP.
Insulin Therapy for Women With IADPSG‐Diagnosed
Gestational Diabetes Mellitus. Does It Work? J Clin Endocrinol
Metab. 2017;102(3):849–857.
49. Camelo Castillo W, Boggess K, Stürmer T, Brookhart MA,
Benjamin DK, Jonsson Funk M. Trends in glyburide com­
pared with insulin use for gestational diabetes treatment in
the United States, 2000–2011. Obstet Gynecol. 2014;123(6):
1177–1184.
50. Foretz M, Guigas B, Bertrand L, Pollak M, Viollet B.
Metformin: from mechanisms of action to therapies. Cell
Metab. 2014;20(6):953–966.
51. Coetzee EJ, Jackson WP. Metformin in management of preg­
nant insulin‐independent diabetics. Diabetologia. 1979;
16(4):241–245.
52. Feig DS, Moses RG. Metformin therapy during pregnancy:
good for the goose and good for the gosling too? Diabetes
Care. 2011;34(10):2329–2330.
53. Vanky E, Zahlsen K, Spigset O, Carlsen SM. Placental pas­
sage of metformin in women with polycystic ovary syn­
drome. Fertil Steril. 2005;83(5):1575–1578.
54. Given JE, Loane M, Garne E et al. Metformin exposure in
first trimester of pregnancy and risk of all or specific con­
genital anomalies: exploratory case‐control study. BMJ.
2018;361:k2477.
55. Rowan JA, Hague WM, Gao W, Battin MR, Moore MP,
Investigators MT. Metformin versus insulin for the treat­
ment of gestational diabetes. N Engl J Med. 2008;358(19):
2003–2015.
56. Rowan JA, Rush EC, Plank LD et  al. Metformin in gesta­
tional diabetes: the offspring follow‐up (MiG TOFU): body
composition and metabolic outcomes at 7–9 years of age.
BMJ Open Diabetes Res Care. 2018;6(1):e000456.
57. Tarry‐Adkins JL, Aiken CE, Ozanne SE. Neonatal, infant,
and childhood growth following metformin versus insulin
treatment for gestational diabetes: a systematic review and
meta‐analysis. PLoS Med. 2019;16(8):e1002848.
58. Dunne F. A Clinical Trial of the Effectiveness of Metformin
in Addition to Usual Care in the Reduction of Gestational
Diabetes Mellitus (EMERGE). clinicaltrials.gov Available at:
https://clinicaltrials.gov/ct2/show/NCT02980276. Accessed
27 March, 2020.
59. Ashcroft FM. Mechanisms of the glycaemic effects of sulfo­
nylureas. Horm Metab Res. 1996;28(9):456–463.
60. Wexler DJ, Powe CE, Barbour LA et al. Research gaps in ges­
tational diabetes mellitus: executive summary of a national
institute of diabetes and digestive and kidney diseases work­
shop. Obstet Gynecol. 2018;132(2):496–505.
61. Langer O, Conway DL, Berkus MD, Xenakis EM, Gonzales
O. A comparison of glyburide and insulin in women with
gestational diabetes mellitus. N Engl J Med. 2000;343(16):
1134–1138.
62. Hebert MF, Ma X, Naraharisetti SB et al. Are we optimizing
gestational diabetes treatment with glyburide? The pharma­
cologic basis for better clinical practice. Clin Pharmacol
Ther. 2009;85(6):607–614.
 Managing Gestational Diabetes During and After Pregnancy
63. Balsells M, García‐Patterson A, Solà I, Roqué M, Gich I,
Corcoy R. Glibenclamide, metformin, and insulin for the
treatment of gestational diabetes: a systematic review and
meta‐analysis. BMJ. 2015;350:h102.
64. Kalafat E, Sukur YE, Abdi A, Thilaganathan B, Khalil A.
Metformin for prevention of hypertensive disorders of preg­
nancy in women with gestational diabetes or obesity: sys­
tematic review and meta‐analysis of randomized trials.
Ultra­sound Obstet Gynecol. 2018;52(6):706–714.
65. Excellence NIfHaC. Diabetes in pregnancy: management
from preconception to the postnatal period. In:2015.
66. Evers IM, de Valk HW, Mol BW, ter Braak EW, Visser GH.
Macrosomia despite good glycaemic control in Type I dia­
betic pregnancy; results of a nationwide study in The Nether­
lands. Diabetologia. 2002;45(11):1484–1489.
67. Langer O, Yogev Y, Xenakis EM, Brustman L. Overweight
and obese in gestational diabetes: the impact on pregnancy
outcome. Am J Obstet Gynecol. 2005;192(6):1768–1776.
68. Catalano PM, Hauguel‐De Mouzon S. Is it time to revisit the
Pedersen hypothesis in the face of the obesity epidemic? Am
J Obstet Gynecol. 2011;204(6):479–487.
69. Institute of Medicine (US) and National Research Council
(US) Committee to Reexamine IOM Pregnancy Weight
Guidelines. In: Rasmussen KM, Yaktine AL, ed. Weight Gain
During Pregnancy: Reexamining the Guidelines. Washington,
DC. National Acadamies Press (US); 2009.
70. Alberico S, Erenbourg A, Hod M et al. Immediate delivery
or expectant management in gestational diabetes at term: the
GINEXMAL randomised controlled trial. BJOG. 2017;124(4):
669–677.
71. Witkop CT, Neale D, Wilson LM, Bass EB, Nicholson WK.
Active compared with expectant delivery management in
women with gestational diabetes: a systematic review. Obstet
Gynecol. 2009;113(1):206–217.
72. Egan AM, Murphy HR, Dunne FP. The management of
type 1 and type 2 diabetes in pregnancy. QJM. 2015;108(12):
923–927.
73. Ryan EA, Al‐Agha R. Glucose control during labor and
delivery. Curr Diab Rep. 2014;14(1):450.
74. Hamel MS, Kanno LM, Has P, Beninati MJ, Rouse DJ,
Werner EF. Intrapartum glucose management in women
with gestational diabetes mellitus: a randomized controlled
trial. Obstet Gynecol. 2019.
75. Finkelstein SA, Keely E, Feig DS, Tu X, Yasseen AS, Walker
M. Breastfeeding in women with diabetes: lower rates
despite greater rewards. A population‐based study. Diabet
Med. 2013;30(9):1094–1101.
85
76. O’Reilly MW, Avalos G, Dennedy MC, O’Sullivan EP, Dunne
F. Atlantic DIP. high prevalence of abnormal glucose toler­
ance post partum is reduced by breast‐feeding in women
with prior gestational diabetes mellitus. Eur J Endocrinol.
2011;165(6):953–959.
77. Gunderson EP, Lewis CE, Lin Y et al. Lactation duration and
progression to diabetes in women across the childbearing
years: the 30‐year CARDIA study. JAMA Intern Med.
2018;178(3):328–337.
78. Ratner RE, Christophi CA, Metzger BE et al. Prevention of
diabetes in women with a history of gestational diabetes:
effects of metformin and lifestyle interventions. J Clin
Endocrinol Metab. 2008;93(12):4774–4779.
79. Carmody L, Egan AM, Dunne FP. Postpartum glucose test­
ing for women with gestational diabetes mellitus: improving
regional recall rates. Diabetes Res Clin Pract. 2015.
80. Villamor E, Cnattingius S. Interpregnancy weight change
and risk of adverse pregnancy outcomes: a population‐based
study. Lancet. 2006;368(9542):1164–1170.
81. Shepherd E, Gomersall JC, Tieu J, Han S, Crowther CA,
Middleton P. Combined diet and exercise interventions for
preventing gestational diabetes mellitus. Cochrane Database
Syst Rev. 2017;11:CD010443.
82. Chiswick C, Reynolds RM, Denison F et al. Effect of met­
formin on maternal and fetal outcomes in obese pregnant
women (EMPOWaR): a randomised, double‐blind, placebo‐
controlled trial. Lancet Diabetes Endocrinol. 2015;3(10):
778–786.
83. Syngelaki A, Nicolaides KH, Balani J et al. Metformin versus
placebo in obese pregnant women without diabetes mellitus.
N Engl J Med. 2016;374(5):434–443.
84. Zhu WW, Yang HX, Wei YM et al. Evaluation of the value of
fasting plasma glucose in the first prenatal visit to diagnose
gestational diabetes mellitus in china. Diabetes Care.
2013;36(3):586–590.
85. Lorenzo‐Almorós A, Hang T, Peiró C et  al. Predictive and
diagnostic biomarkers for gestational diabetes and its associ­
ated metabolic and cardiovascular diseases. Cardiovasc
Diabetol. 2019;18(1):140.
86. Ghosh P, Luque‐Fernandez MA, Vaidya A et al. Plasma gly­
cated CD59, a novel biomarker for detection of pregnancy‐
induced glucose intolerance. Diabetes Care. 2017;40(7):
981–984.
87. Artzi NS, Shilo S, Hadar E et  al. Prediction of gestational
diabetes based on nationwide electronic health records. Nat
Med. 2020;26(1):71–76.
 86 Initial Evaluation and Management of Diabetes
88. Xiang AH, Peters RK, Kjos SL et al. Effect of pioglitazone on
pancreatic beta‐cell function and diabetes risk in Hispanic
women with prior gestational diabetes. Diabetes. 2006;55(2):
517–522.
89. Genova M, Todorova‐Ananieva K, Atanasova Bea.
Proinsulin in healthy pregnancy, pregnancy with gestational
diabetes and after delivery. Acta Medica Bulgarica. 2014(41):
13–21.
90. Kautzky‐Willer A, Thomaseth K, Ludvik B et  al. Elevated
islet amyloid pancreatic polypeptide and proinsulin in lean
gestational diabetes. Diabetes. 1997;46(4):607–614.
91. Haffner SM, Gonzalez C, Mykkänen L, Stern M. Total
immunoreactive proinsulin, immunoreactive insulin and
specific insulin in relation to conversion to NIDDM. the
Mexico City Diabetes Study. Diabetologia. 1997;40(7):
830–837.
 7 What Is the Role of Self‐Monitoring
in Diabetes? Is There a Role
for Postprandial Glucose Monitoring?
How Does Continuous Glucose
Monitoring Integrate into Clinical
Practice?
Rami Almokayyad1 and Robert Cuddihy2
1
 ndocrine Fellow Division of Endocrinology, Department of Medicine, University of Minnesota Medical School,
E
University of Minnesota, Minneapolis, MN, USA
2
 edical Director, International Diabetes Center, World Health Organization Collaborating Center for Diabetes
M
Education, Translation and Computer Technology, Minneapolis, MN, USA

LE A R NI NG P OI NT S Self‐monitoring of blood glucose

●● There is general consensus that the use of SMBG in
most individuals with T1DM or those with T2DM
treated with insulin is of benefit.
●● The benefit of SMBG in individuals with T2DM, not
on insulin therapy, remains inconclusive and
controversial.
●● The role of postprandial glucose monitoring has
previously been firmly established in gestational
diabetes but remains controversial in T1DM or
T2DM.
●● Continuous Glucose Monitoring (CGM) devices have
documented benefit in adults with T1DM on an insulin
pump therapy who use the data it provides to modify
their therapy.
●● There is a clear need for well‐controlled clinical trials to
assess the value of SMBG in T2DM for those on other
than insulin therapy.
●● The lack of a simple universal output of SMBG and
CGM data are an impediment to their appropriate
uptake and use in primary care.
Self‐monitoring of blood glucose (SMBG) can be consid-
ered one of the bigger advances in diabetes management
after the discovery of insulin and oral agents. Bedside
measurement of plasma glucose was introduced in the
mid‐1920s as a necessity, after the discovery and use of the
first insulin. At that time, the only modality to monitor dia-
betes control was qualitative urine glucose measurements,
which only gave patients an estimate of their levels over the
proceeding 1–2 days, revealing hyperglycemia only when
serum glucose had exceeded the renal threshold (∼200 mg/
dl) resulting in glycosuria. The procedure of measuring
glucose would typically take 20  minutes, as described by
Maclean in his book Modern Methods in the Diagnosis and
Treatment of Glycosuria and Diabetes, published in
1924 [1]. The use of this method was limited to use by the
clinician.
By the early 1940s, semiquantitative urine glucose
measurement kits became commercially available for home

Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.

© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.

87
 88 Initial Evaluation and Management of Diabetes
use, and they became the major modality for glucose self‐
monitoring at home.
In the mid‐1960s, reagent strips were introduced as a
semiquantitative method of measuring capillary whole
blood glucose; a drop of blood was added to these reagents
that utilized a glucose oxidase reaction to effect a colori-
metric change. The resultant color from the reaction could
then be compared with a standard chart and this would
give a rough estimate of the blood glucose level. These
strips were initially limited to office and hospital use, with
limited patient access for home use.
By the late 1960s reflectance meters were introduced
that utilized the same type of reagent strips where a light
source was reflected on the strip (instead of reading them
manually) enabling the reflected light to be read by a pho-
toelectric cell, which in turn gave a readout of the estimated
glucose (using a swinging needle at that time). In the late
1970s meters were developed for patient use.
With advances in technology, electrochemical meters
were introduced, and as the technology evolved, improve-
ments were made to reduce the size of these meters, reduce
their reaction and reading times, simplify blood sampling
techniques, and to reduce the discomfort associated with
blood sampling  [2]. Today, glucose meters have a central
role in diabetes management, with nearly 70% of all
patients with diabetes performing some SMBG [3].
Long‐term follow‐up of participants in the Diabetes
Complication and Control Trial (DCCT) as part of the
Epidemiology of Diabetes Interventions and Complications
(EDIC) in type 1 diabetes mellitus (T1DM), and the 10‐
year follow‐up of the United Kingdom Prospective
Diabetes Study (UKPDS) in type 2 diabetes mellitus
(T2DM) provide clear evidence that the early and intensive
control of hyperglycemia reduces the long‐term risks of
microvascular and macrovascular complications in diabe-
tes. The effect of good glycemic control seems to persist for
years, to some extent independently of subsequent glyce-
mic control. This phenomenon is referred to as “metabolic
memory” or “the legacy effect” [4–6]. However, tight glyce-
mic control does come at a cost, which is a threefold
increased risk of severe hypoglycemia seen in the DCCT
trial (T1DM), as well as the Action to Control
Cardiovascular Risks in Diabetes (ACCORD) trial in
T2DM [7, 8]. ACCORD also suggested an increased mor-
tality risk with assignment to a strategy aiming to achieve
near normoglycemia. However, somewhat surprisingly,
this risk was inversely proportional to the rate of fall in
HbA1c, being lowest in those individuals who rapidly cor-
rected their HbA1c and highest in those who lowered their
HbA1c slowly, or not at all, and the highest mortality in
both the standard and intensive glycemic control arms of
ACCORD occurred in those patients with higher HbA1c
(publicly presented ACCORD data in  [9]). The available
evidence would suggest that in most patients with diabetes
one should maintain “tight” glycemic control as early in the
course of the disease as possible and to continue to main-
tain this level of control as tightly as possible [10], generally
targeting an A1c level < 7.0% by ADA guidelines (or < 6.5%
by IDF or AACE guidelines), if this can be accomplished
safely without exposing the patient to undue risk of severe
hypoglycemia.
SMBG therefore can play a pivotal role, when used
properly, as an adjunctive tool to enhance patient self‐man-
agement of their diabetes. SMBG provides day‐to‐day data
on glycemic control; it provides an immediate feedback
about the effect of nutrition, physical activity, and medica-
tions on blood glucose. It allows prompt determination
of hypoglycemia or hyperglycemia that not only can
improve patient safety, but also can motivate individuals
to make appropriate changes in diet, exercise, and
medications.
By shaping meal and activity patterns and providing
feedback on medication dosing or titration while simulta-
neously providing useful information on the potential
occurrence or risks of hypoglycemia, the information
gleaned from SMBG monitoring can help optimize the
delicate balance between the benefits and risks of tight gly-
cemic control.
In this section, we will discuss the utility of SMBG in
diabetes management, and we will discuss the evidence
behind its use in different patient groups.
SMGB in type 1 DM and insulin‐requiring
T2DM patients
The use of SMBG in patients with type 1 diabetes and in
those with type 2 diabetes treated with intensive insulin
therapies or multiple daily insulin injections (MDI) appears
 What Is the Role of Self-Monitoring in Diabetes?
logical for several reasons, is supported by clinical trial
data, and thus is not controversial.
First, SMGB will provide day‐to‐day information that
can help adjust insulin dose to optimize the overall glyce-
mic control. Patients on bolus (“prandial or meal‐related”)
insulin use the premeal SMBG data to adjust the dose of
their prandial insulin, correcting it for the amount of car-
bohydrate consumed, as well for any variance of the pre-
meal glucose above or below the desired target range
(so‐called corrective dose insulin). While various insulin
adjustment methods may differ in their specific titration
schemes, they all utilize premeal SMBG values as an
actionable item aiding in appropriately adjusting the insu-
lin dosing before a meal so as to achieve tighter glycemic
control.
One of the earliest studies that demonstrated the effec-
tiveness of SMBG in improving glycemic control was con-
ducted in 1982 [11]. This concept was emphasized in the
Diabetes Complication and Control Trial, in which inten-
sive insulin therapy has shown to improve glycemic con-
trol and reduce microvascular complications, and this was
achieved by intensifying insulin regimens utilizing
SMBG [7].
Also, SMBG is an important safety tool in insulin treated
patients to detect hypoglycemia, especially in patients with
hypoglycemia unawareness.
The recommendation from the DCCT was to perform
SMBG at least four times per day; however, observational
studies have shown that most patients with type 1 diabetes
on intensive (physiological) insulin regimen fall short of
these recommendations [12].
SMBG in non‐insulin‐requiring T2DM
Although SMBG is now widely accepted as a part of the
management of patients with non‐insulin‐treated type 2
diabetes, its efficacy and rationale are more controversial.
Most of the existing data for the efficacy of SMBG in these
subjects come from cross‐sectional and retrospective
studies, which slightly favored more frequent SMBG
[13–15].
Many randomized clinical trials (RCTs) have been car-
ried out in small groups of patients. Participants were not
recruited from representative populations in the commu-
89
nity and the strategies for use of the results from SMBG
were not clearly defined. The relative benefit in terms of
HbA1c reduction typically was modest, even in those ran-
domized clinical trials which did show benefit, in the range
from one quarter to two thirds of 1%, and most meta‐anal-
ysis of various combinations of these RCTs favor SMBG.
Epidemiologic studies utilizing large patient databases
have generally noted improved glycemic control in indi-
viduals with diabetes who monitor more often, but such
studies can only implicate an association between more
frequent SMBG and improved glycemic control, but not a
causal relationship. Selection biases and other confounding
variables may also affect these results  [16]. Other cross‐
sectional studies, such as the Fremantle Diabetes Study, do
not show benefit in terms of glycemic control with
SMBG [17].
The Cochrane Collaborative best‐evidence‐based
review of SMBG use in patients with T2DM on non‐insulin
therapies also appears to favor SMBG but noted that more
evidence is needed [18].
Farmer and colleagues conducted a randomized con-
trolled trial (Diabetes Glycaemic Education and
Monitoring [DiGEM] study) that aimed to test whether
SMBG, used with or without instruction in incorporating
findings into self‐care, can improve glycemic control in
non‐insulin‐treated diabetes patients compared with
standardized usual care [19]. A total of 453 patients were
individually randomized to one of three groups: (1)
standardized usual care with 3‐monthly HbA1c (control);
(2) blood glucose self‐testing with patient training
focused on clinician interpretation of results in addition
to usual care (less intensive self‐monitoring); or (3)
SMBG with additional training of patients in interpreta-
tion and application of the results to enhance motivation
and maintain adherence to a healthy lifestyle (more inten-
sive self‐monitoring).
There was no evidence of glycemic benefit between the
three groups at the end of 12 months (no difference in the
primary outcome; Hemoglobin A1C). In addition, there
was no evidence of a significantly different impact of self‐
monitoring on glycemic control when comparing sub-
groups of patients defined by duration of diabetes,
therapy, and diabetes‐related complications. Patients who
were in the more intensive SMBG arm detected more
 90 Initial Evaluation and Management of Diabetes
hypoglycemia. The economic analysis suggested that
SMBG resulted in extra health care costs and was unlikely
to be cost‐effective if used routinely. There was an initial
negative impact associated with more frequent use of
SMBG on the quality of life [20].
The potential clinical ramifications from this study have
been huge and called into question the utility, cost‐effec-
tiveness, and effect on quality of life of SMBG individuals
with T2DM who are not on insulin therapy. In an era of
tightening financial resources, an epidemic of T2DM, and
attempts to curb health care expenditures in general, this
and subsequent reports from the DiGEM study had lead to
a wide reappraisal of benefits or SMBG in individuals who
are not yet treated with insulin. Reappraisal of the need to
cover SMBG testing supplies for individuals with T2DM
not treated with insulin by payers and national groups such
as the National Health Service in the U.K. and the Centers
for Medicare and Medicaid Services (CMS) in the United
States have reportedly taken place. Thus, a careful look at
some of the potential concerns or criticism of the DiGEM
study is warranted.
This study enrolled individuals with relatively recent
onset of diabetes (median duration of 3 years) treated with
diet or oral agents with reasonably good glycemic control
(mean HbA1c of 7.5%), and specifically selected individu-
als who were either not monitoring SMBG at all or moni-
toring no more than a single one‐time SMBG per week.
Thus, the study may have inadvertently selected a biased
population less geared toward, or less compliant with,
SMBG monitoring and with potentially less to gain from
improvements in glucose control. Of concern is the case that
it was in the intensive SMBG cohort that more individuals
quit SMBG monitoring than the less intensive cohort. Also,
for those who were to utilize the SMBG data to modify their
lifestyle or medication, a delineation of a specified action
plan in response to the SMBG data is lacking and not deline-
ated. While there was a minimal decline in HbA1c by 0.17 %
in this group, it was not statistically significant. Of potential
concern was the issue that the reduction in HbA1c in this
study was far less than those reported from the majority of
other RCT evaluating the effect of SMBG.
Criticisms aside, this study was a careful attempt to get
at the issue of the value of SMBG in terms of improvements
in glycemic control in such a population of individuals and
at least raises the stakes for proponents of SMBG to prove
its worth in non‐insulin‐treated populations. Several
groups have formed to attempt to outline the necessary
components of a large scale RCT to better evaluate the role
and utility of SMBG in T2DM [16, 21].
From a philosophical standpoint it is of vital importance
to understand the use of SMBG as a useful diagnostic tool
to enhance patient self‐management of diabetes rather
than as a direct therapeutic intervention targeting glucose
levels. As such, there are multiple aspects of the use of
SMBG that must be in place for the data it generates to be
accurate, beneficial to glycemic control and, most impor-
tantly, used by the patient. Important potential barriers to
appropriate SMBG include proper technique, correct cod-
ing of glucose meters to match the testing strips, correct
setting of the time and date of the meter to aid in SMBG
review of downloaded meter data and, most importantly,
appropriate patient education and understanding of the
timing of SMBG and the use of the data derived from it to
modify the patient’s self‐management. The data can then
become an actionable item leading to modification in ther-
apy (whether leading to changes in diet or activity through
behavioral change, or adjustment in medication). Ideally,
to maximize the impact of SMBG, the patients themselves
should be educated on how to appropriately use SMBG.
Such a tool could be used at various time points to optimize
management throughout the day, uncovering needed
behavioral modifications in diet and activity, providing
feedback for potential problematic periods of marked
hyperglycemia or hypoglycemia, and providing the patient
an early detection of worsening overall glycemic control
due either to situational factors such as intercurrent illness
or steroid usage, or the progressive nature of T2DM itself.
Such information would indicate the need to titrate ther-
apy to reestablish target levels of glycemic control in a
timely manner.
There is a great need for specific algorithms instructing
patients what do in terms of altering their management or
therapy in response to their SMBG data. Some early pre-
liminary data suggest that patient‐driven algorithms based
on SMBG can be more effective in helping them reach tar-
get than management that relies on the patient’s health care
 What Is the Role of Self-Monitoring in Diabetes?
provider or clinician to review the data and recommend
changes in therapy. For example, in the commonly seen
patient with good control of AM fasting blood glucose
(BG), but HbA1c’s above target, it is likely that BG values
are higher at other time points throughout the day.
Bergenstal et al. showed that utilizing an algorithm for
titrating premeal insulin solely based on preprandial
SMBG values, rather than utilizing strict carbohydrate
counting with matching insulin to carbohydrate dosing,
was equally effective in controlling glycemic levels in insu-
lin‐treated individuals with T2DM [22].
The technical aspects of glucose meters that serve as
barriers are the easiest to correct and, in fact, meters
that do not require coding of the meter to match the
testing strips are beginning to enter the market. Time
and date stamping of SMBG values will become auto-
mated. Meal markers are available on some units to aid
in interpreting fasting or premeal glucose patterns from
postprandial patterns. Many meters can inform the
patient if the blood sample is inadequate for an accurate
test result.
The lack of patient education and training in diabetes
self‐management, including proper use of SMBG, remains
a significant barrier that requires more effort to correct. In
clinical practice, one frequently encounters patients treated
with lifestyle modification and Metformin monotherapy
who have been instructed to perform SMBG and dutifully
obtain one fasting blood glucose value each morning. No
alternative testing schema is offered if the AM fasting glu-
coses are within goal but the HbA1c remains above goal.
Clearly, rather than obtain seven “normal fasting readings”
per week the utility of SMBG would be markedly enhanced
if such patients were educated to use these same seven
weekly readings in a more dispersed fashion, sampling
once daily, but at alternating times each day. Perhaps
obtaining AM fasting blood glucose on Monday, 2‐hour
postprandial BG after breakfast on Tuesday, prelunch BG
on Wednesday, 2‐hour post‐lunch BG on Thursday,
predinner BG on Friday, post‐dinner BG on Saturday, and
bedtime BG on Sunday would provide a wealth of action-
able information? For instance, a significant postprandial
rise in BG following a specific meal could signal the patient
to make modifications in the timing or quantity of carbo-
91
hydrate consumed during the day or to add more physical
activity prior to this meal. If unable to correct this issue
with dietary or activity changes, such findings could lead
the patient to engage their physician to consider the addi-
tion of another pharmacologic agent targeting postpran-
dial BGs. The individual could “learn” from the response of
SMBG just how various meals and meal composition
affect their BGs and what the affect of various activities
are on their BGs, acting as a useful tool and reinforce-
ment for beneficial behavioral modification. If the post-
prandial BG rise was more pronounced following one
specific meal, then the pattern of SMBG monitoring
might temporarily change while the individual “works on
that particular problem area,” perhaps using many of
their weekly SMBG determinations before and after that
particular meal to assess the result of various attempted
interventions, returning to a widespread surveillance pat-
tern of SMBG once the “problem is solved” or corrected.
This surveillance SMBG would then inform the individ-
ual if and where the next issue in fine‐tuning glycemic
control should occur.
If individuals are on agents that can cause hypoglyce-
mia, such as a sulfonylurea (SU), then surveillance
SMBG can indicate if there is a problematic period of
increased risk of hypoglycemia during the day, such as
the late afternoon or predinner period or overnight,
which should ideally lead to corrective intervention to
reduce this risk.
Utilizing SMBG for continuous surveillance and quality
improvement of their diabetes self‐management can help
counteract the clinical inertia currently seen in our health
systems, which are not properly designed to manage non‐
acute chronic diseases such as diabetes. Brown et al. [23]
have demonstrated how this inertia can result in patients’
encountering 8–10 years of exposure to significant chronic
hyperglycemia, with the increased risk for complications
that this entails while their medical regimen is very slowly
progressed through the different available therapies.
Patients should be empowered to contact their health care
provider as soon as they encounter problematic hypergly-
cemia that has not responded to their attempts at correc-
tion, as medication may need to be advanced. Rather than
await their next regularly scheduled 3‐ to 6‐month appoint-
 92 Initial Evaluation and Management of Diabetes

ment before advancement in therapy is undertaken, the
therapies could be quickly optimized until the glycemic
goals or treatment targets are achieved.
Another barrier to the optimal use of patient derived
SMBG data in diabetes management occurs in physician
offices. In today’s environment where health care providers
have less and less time in which to see their patients, they
are often forced to complete the entire visit in 15 to 20 min-
utes. In such a scenario, the use of the glucose logbook to
look for patterns from which to make therapeutic recom-
mendations is problematic. It is difficult to expect busy
providers to visually scan often messy, hand‐scribbled col-
umns of individual glucose values and try and make some
sense of any emerging pattern after flipping through sev-
eral pages in a standard glucose logbook (Figure 7.1). The
ability to download glucose meter data and present verified
aggregate glucose data in an organized fashion, with basic
statistical summaries, is certainly an improvement.
However, the multitude of differing proprietary software
programs needed to download the data from each compa-
ny’s meter and the slightly different presentation format of
each of these software programs severely inhibit the broad
generalizability of this important tool and thus limits its
uptake in offices, especially in primary care settings. As
opposed to another diagnostic tool, the electrocardiogram
(ECG), which has the same standardized universal output,
no such common format and universal output exists for
SMBG data. Thus, while the standard 12‐lead ECG tracing
enjoys widespread use in clinical practice, SMBG use
remains most widely relegated to the use of a handwritten
glucose logbook. This is a major gap in our ability to teach
glycemic pattern recognition to our patients and fellow cli-
nicians, severely impeding the great potential of SMBG to
help shape therapeutic interventions and improve overall

BREAKFAST LUNCH DINNER BEDTIME

SMBG values
Date Night BG Med BG BG Med BG BG Med BG BG Med Notes/Ketones tranlated in
BG
mg/dL mg/dL mg/dL mmol/L

2/24 179 199 227 9.9-11.0-15.1

2/25 198 177 189 Shopping 11.0-9.8-10.5

2/26 165 188 182 9.2-10.4-10.1

2/27 195 209 248 10.8-11.6-13.8

2/28 187 225 10.4-12.5

3/1 153 167 288 20 min walk 8.5-9.3-16.0

3/2 182 159 219 10.1-8.9-12.1

3/3 197 155 203 10.9-8.6-11.3

3/4 189 182 199 20 min walk 10.5-10.1-11.1

3/5 210 153 262 20 min walk 11.7-8.5-14.6

3/6 173 9.6

3/7

FIG 7.1  Typical patient glucose logbook with self‐reported data.

 What Is the Role of Self-Monitoring in Diabetes? 93

220 Overall target

200
180
160
140
Glucose

120
100
80
60
40
20
0

0 AM 0 AM 0 AM 0 PM 0 PM 0 PM 0 PM 0 PM 0 PM 0 AM 0 AM 0 AM 0 AM
6:0 8:0 10:0 12:0 2:0 4:0 6:0 8:0 10:0 12:0 2:0 4:0 6:0

Time of day
Statistics

Glucose average: 127 Target type: Personal

% Within target: 50 Before meal target: 70–110
# of glucose readings: 16 After meal target: 90–140
# of hypo. readings: 1 Hypoglycemic: 67
Standard deviation: 48

FIG 7.2  Modal Day presentation of downloaded meter with SMBG data.

levels of control. The representation of all SMBG data
expressed as a single day over 24 hours was an attempt to
have some more common output from the multitude of
meters, many of which do allow data to also be expressed in
this form (Figure 7.2). Unfortunately, the widespread use of
such a modal day output never became commonplace,
especially in primary care practice where they may be most
useful.
The many barriers to the proper use of SMBG and the
controversy still remaining around the utility of SMBG in
patients with T2DM who are not on insulin have given rise
to an interesting question. Given that many available
therapies including dietary and activity modification,
or pharmacologic agents such as metformin, dipeptidyl
peptidase‐IV (DDP‐4) inhibitors, glucagon‐like peptide‐1
(GLP‐1) receptor agonists, and thiazolidinediones (TZDs)
all can improve overall glycemic control and do not result
in increased risks for hypoglycemia, could many patients
use these therapies singly or in combination without the
need for any SMBG?
These patients could be followed by period HbA1c
measurement to be sure that they are achieving their over-
all glycemic targets. Might the cost savings based on forgo-
ing the need for expensive glucose testing strips one or
many times daily, as well as the ease of daily self‐manage-
ment of their diabetes and potential perceived improve-
ment in quality of life, justify such an approach?
Thus, there remains an urgent and ongoing need for
properly designed, well‐controlled RCTs to evaluate just
what is the benefit of SMBG in those individuals with
T2DM treated with non‐insulin therapies. This popula-
tion, given their greater numbers, likely makes up the
lion’s share of the SMBG market and thus this question
has enormously important ramifications for general pub-
 94 Initial Evaluation and Management of Diabetes
lic health care policy in dealing with the growing diabetes
epidemic.
Postprandial SMBG
The role of postprandial SMBG is firmly established in
women with gestational diabetes mellitus (GDM).
Adjusting insulin therapy in mothers with GDM, based
on postprandial BG, resulted in improved HbA1c, lower
birth weights (i.e., less macrosomia), less neonatal hypo-
glycemia and less need for a cesarean section at
delivery [24].
Meal‐based SMBG is a valuable tool for improving out-
comes in pregnancy complicated by diabetes and has been
shown to improve fetal perinatal outcomes  [25]. This is
more evident in insulin‐treated patients, and less clear with
diet‐controlled patients  [26], although such monitoring
may be useful in providing feedback for behavioral modifi-
cation and surveillance as to the adequacy of glucose con-
trol and the need to intensify therapies if not within target.
Also, it encourages patients to actively participate in their
own care.
The role of postprandial SMBG in T1DM and especially
in T2DM is controversial. The landmark trials, such as the
DCCT‐EDIC and UKPDS and its 10‐year follow‐up trials,
clearly showed that improving overall glycemic control,
thus reducing chronic exposure to hyperglycemia as meas-
ured by the HbA1c, reduced the risks of microvascular and
macrovascular complications [4–7]. Thus, the current rec-
ommendations by the ADA are to target an HbA1c of < 7%
in most individuals with diabetes [27]. These trials did not
typically require postprandial SMBG, nor did they com-
pare the effects of targeting fasting and premeal SMBG ver-
sus postprandial SMBG in reducing complications.
In T1DM it is not uncommon for individuals to monitor
postprandial SMBG regularly to assess the adequacy of
control exerted by their premeal short‐ or rapid‐acting
insulin on an MDI program or their bolus, if they are on
insulin‐pump continuous subcutaneous insulin infusion
(CSII) therapy. Monitoring postprandial SMBG provides
these individuals with feedback both on insulin dosage and
self‐management behavior such as carbohydrate counting
and their insulin to carbohydrate ratio. Nonetheless, it
remains a common clinical practice for many individuals
with either T1DM or T2DM to use their next or following
premeal SMBG, rather than a 2‐hour PPG, to gauge the
efficacy of their previous premeal insulin dose or other
therapy directed toward glycemic control. This assumes
that if the next premeal SMBG is within target, then the
postprandial glucose values were likely “acceptable”. While
this method does provide a crude, indirect estimate of gly-
cemic control over this period, it cannot guarantee glucose
excursions, or overall glycemic exposure, are well con-
trolled. Thus, while this method is less labor‐intensive and
may be an appropriate compromise in terms of “quality of
life” for the individual, more intensive SMBG monitoring,
including PPG, should be considered if HbA1c is not
within the desired target range.
In T2DM the role of postprandial SMBG is even less
clear. T2DM usually has a more indolent onset, slowly pro-
gressing through phases of glucose intolerance to frank
diabetes, and is closely associated with the metabolic syn-
drome and the increased cardiovascular risk that this
entails. There have been several studies, such as the
DECODE, and others, that have found correlations with
elevated postprandial glucoses and cardiovascular disease
risk extending from impaired glucose tolerance right down
into the normal range for glucose, and in fact indicate
greater association with risks of cardiovascular disease for
elevations in postprandial glucose (PPG) than fasting
plasma glucose (FPG) [28, 29].
Again, most of the trials correlated improvement in
chronic glucose exposure as measured by HbA1c with
decreased risk of long‐term complications. Some studies
suggest a closer correlation of postprandial or post‐chal-
lenge glucose with HbA1c and mortality than fasting glu-
cose [30], but this remains controversial and is not seen in
yet other studies. Therapies that specifically target PPG
such as Acarbose have been shown to reduce CV disease
and all‐cause mortality in the Stop‐NIDDM trial [31, 32].
Studies by Ceriello, and others, have provided indirect evi-
dence that glycemic variability or excursions (typically
most marked in the immediate postprandial period)
directly contribute to diabetic complications through oxi-
dative stress, and the generation of free radical formation,
activation of the polyol pathways, and generation of PKC β
and advanced glycosylation end products (AGES) [33, 34].
The role PPG plays in the development of long‐term dia-
 What Is the Role of Self-Monitoring in Diabetes?
betic complications remains very controversial. One often‐
referenced paper, used to support this evidence that PPG
plays a role in complications in T1DM, noted that in the
DCCT trial those individuals in the intensive glycemic
treatment arm suffered from less diabetes‐related compli-
cations than did their counterparts in the standard arm,
even when matched for HbA1c, which is actually incor-
rect [35]. This was actually a hypothesized extrapolation or
modeling of the data and not directly representative of the
clinical data itself [36]. While the controversy continues to
rage, some groups like the International Diabetes
Federation (IDF) have published guidelines for the control
of PPG [37].
Given the relative lack of strong RCT data comparing
the benefits of postprandial SMBG versus fasting and
immediate premeal SMBG measurement, why might post-
prandial SMBG measurement be of importance in indi-
viduals with T2DM who are not treated with insulin? It
may be necessary to target PPG to get more of these indi-
viduals to their glycemic goal. Monnier et al. have shown in
a population of individuals with T2DM not treated with
insulin that the contribution of PPG to overall glycemic
exposure rises progressively the closer one approaches a
target HbA1c < 7% [38]. At HbA1c < 7.6% the contribution
of PPG to overall glycemic exposure totals approximately
70%. This, in turn, may help to explain why in several RCTs
that targeted interventions aimed at achieving an AM fast-
ing blood sugar within a specific goal range (typically less
than 100 mg/dL), many individuals who have achieved this
target still have HbA1c that remains above 7%. With nearly
40% of the U.S. population not in target in terms of HbA1c,
more focus on PPG through proper postprandial SMBG
monitoring and patient education on what action to take if
they are not in target may be necessary! As noted, the IDF
has published a guideline of the targeting of postprandial
SMBG [37], and the ADA and AACE guidelines give rec-
ommended postprandial target glucose levels.
A reasonable approach may be to recommend targeted
PPG or intensification of SMBG in individuals who are
either not within their goal range for HbA1c on their stabi-
lized maintenance diabetes regimen, or in newly treated
patients with initially high HbA1c’s, as their HbA1c’s are
lowered by initial therapies and are approaching 7%. The
key is to use SMBG when it serves a discreet purpose either
95
as an actionable item to potentially modify therapy or in
surveillance to ensure adequate glycemic control is being
maintained.
Continuous glucose monitoring systems
The relatively recent advent of commercially available sub-
cutaneous continuous glucose monitoring (CGM) systems
that continuously measure interstitial fluid glucose has
added to the armamentarium of tools potentially useful for
the self‐management of diabetes. Devices from three man-
ufacturers are currently available and licensed as an adjunc-
tive tool for diabetes management, most commonly in
T1DM for persons on MDI or CSII insulin regimens. The
FDA has mandated that these devices not be used for
directly calculating an insulin dose based on the most cur-
rent CGM reading, as their accuracy in comparison to
those from approved glucose meters measuring capillary
venous whole blood has not been firmly established.
Rather, the data generated from CGM can be useful in
guiding necessary SMBG testing by delineating the real‐
time trending in sequential glucose values, either indicat-
ing worsening hyperglycemia that may result in the need to
confirm with SMBG and potentially take an added supple-
mental dose of insulin; or in indicating rapid glucose low-
ering that may result in eventual hypoglycemia, warranting
treatment. These systems also make available alarms that
can call one’s attention to developing hyperglycemia or
hypoglycemia prompting corrective action.
Most studies of CGM to date have utilized this tool in
patients with T1DM on MDI or CSII insulin therapies, tak-
ing advantage of the real‐time glucose trending data to
fine‐tune insulin therapy. Just as with SMBG, CGM is but a
tool to help guide one’s diabetes self‐management, and not
an antihyperglycemic intervention in and of itself. The
effects of CGMonHbA1c have been relatively mild but real.
The recent JDRF CGM study [39] showed that the use of
CGM in patients with T1DM on CSII did improve HbA1c
in adult patients (over 25 years of age), but not in children
and teenagers who had much higher rates of non‐adher-
ence with proper CGM use (i.e., wearing a sensor) or using
the provided data for decision making.
The Star 1 trial  [40] did not demonstrate dramatic
HbA1c lowering in individuals with T1DM using CSII who
 96 Initial Evaluation and Management of Diabetes
used CGM. It did suggest that in those who used the CGM,
the increased usage was associated with an increased prob-
ability of HbA1c lowering. Many other moderate‐sized tri-
als (n of 100–200) assessing CGM use have come to similar
conclusions, i.e., increased use of the CGM devices and the
data they provide are associated with higher probability of
HbA1c lowering. The Star 3 trial compared the use of sen-
sor‐augmented pump (SAP) therapy with a conventional
MDI program [41]. The trial demonstrated a ~0.8% decrease
in HbA1c compared to a 0.2% decrease in the control group.
Of course, by design, the study could not distinguish between
the effect of CGM and CSII on the outcome.
Like almost any intervention, there has been a “down-
side” occasionally noted in some individuals with T1DM
who use a CGM. In some individuals there is a tendency to
“overbolus” in response to continuing high readings viewed
on the device. This occurs when individuals feel compelled
to frequently re‐bolus with short‐ or rapid‐acting insulin
when high glucose readings are viewed on the CGM device,
most of which provide a new reading every 1–5 minutes.
This impatience, upon frequently viewing high glucose fol-
lowing an intervention that has not yet had adequate time
to work, can lead to “insulin stacking”. This phenomenon
occurs where the effect of the most recent insulin dose is in
addition to (“on top of ”) the ongoing insulin effect from
the residual insulin on board from the previous injection,
resulting in hypoglycemia. Proper education and training
of individuals as well as careful patient selection is required
in choosing individuals who will benefit from CGM.
Another effect that has been seen is “sensor burnout”,
where it becomes difficult to maintain the ongoing effort
required to respond to the wealth of data provided by a
device in “real time”, which may or may not require an
action on the individual’s part. In some studies, utilizing
CGM the HbA1c can be seen to improve over the first
3–9  months only to regress toward the mean after more
prolonged follow‐up because the information is no longer
being used with the same intensity.
The past decade has seen significant advances towards
the adoption of pump therapy controlled by the continuous
data derived from a CGM device. The devices currently
available and the ongoing work in this field are reviewed in
a separate chapter.
The potential future growth of CGM may well depend
largely on its intermittent use for collection and summary
of large quantities of glucose data for “glycemic pattern rec-
ognition” in individuals with T2DM (or T1DM). The use of
periodic CGM in this way would function as a “diagnostic
biopsy” providing much more detailed and potentially
more useful information than an HbA1c. This glycemic
pattern data would aid in targeting appropriate therapeutic
changes as well as delineating the optimal and appropriate
timing of SMBG once the short period of CGM monitoring
is concluded.
However, just as with SMBG, similar barriers are arising
with CGM, as each company promotes its own proprietary
software for data downloading. With no standardized uni-
versal output and at least three competing and differing
software outputs, the widespread adoption of CGM tools in
T2DM, especially in primary care, will face similar barriers
as the proper use of SMBG. Its use will inevitably be
retarded by the increased effort, logistical issues, and more
complicated training and education the widespread use of
such varying outputs in the CGM will require of busy phy-
sician practices. These issues are made even more complex
by the sheer amount of data generated by CGM devices (up
to 15K glucose values over a 2‐week period) making even
organizing the data as modal day challenging to make clin-
ically relevant sense of (Figures 7.3–7.5).
Flash glucose monitoring systems
A flash glucose monitoring system is akin to an intermit-
tently scanned CGM and has been available for clinical use
since 2017. Large, randomized controlled trials suggest
superiority over conventional SMBG in avoiding hypogly-
cemia and decreasing glycemic variability [42, 43]. The fre-
quency of scanning and therefore the amount of
information available to guide management is patient
dependent and optimal benefit requires careful patient
selection as well as education in how to use the data avail-
able  [44]. In such circumstances, this system provides
many of the benefits of CGM with less complexity.
However, it cannot fulfil the sentinel functions of most
conventional CGM – alerting patients when glucose con-
centrations fall.
 400

300
Guardian RT

Glucose-mg/dL
200

100

3:00 AM 6:00 AM 9:00 AM 12:00 PM 3:00 PM 6:00 PM 9:00 PM

Modal Day
DexCom Seven 420
390
360
330
300
270
240

mg/dl
210
180
150
120
90
60
30

2.00 AM 4.00 AM 6.00 AM 8.00 AM 10.00 AM 12.00 PM 2.00 PM 4.00 PM 6.00 PM 8.00 PM 10.00 PM
Tue 2/13 Wed 2/14 Thu 2/15 Sat 2/17 Sun 2/18 Mon 2/19

FreeStyle Navigator 550

500
450
Glucose (mg/dL)

400
350
300
250
200
150
100
50

6:00 AM 12:00 PM 6:00 PM 12:00 AM

Time

FIG 7.3  Varying outputs from commercially available CGM devices.

350
Modal Day
300

250

200

150

100

50

0
12:00 AM

2:00 AM

4:00 AM

6:00 AM

8:00 AM

10:00 AM

12:00 PM

2:00 PM

4:00 PM

6:00 PM

8:00 PM

100:00 PM

120:00 AM

FIG 7.4  CGM device output as a Modal Day.

 98 Initial Evaluation and Management of Diabetes

350
AGP (without diabetes)
300
AGP (patient with normal
glucose tolerance)
250
Glucose (mg/dL)

200

150

100

50

0
12:00 AM

2:00 AM

4:00 AM

6:00 AM

8:00 AM

10:00 AM

12:00 PM

2:00 PM

4:00 PM

6:00 PM

8:00 PM

100:00 PM

120:00 AM
350
AGP (with T2DM)
300

250
90th
Glucose (mg/dL)

200
75th
150 median
25th
100 10th

50

0
12:00 AM

2:00 AM

4:00 AM

6:00 AM

8:00 AM

10:00 AM

12:00 PM

2:00 PM

4:00 PM

6:00 PM

8:00 PM

100:00 PM

120:00 AM

N TARGETS AB O V E W IT H IN BELO W ME N SD M AX M IN AUC DAY

3632 140 70 43.0 53.1 4.0 137.3 44.2 279.0 39.0 3213.5 2372
Days Basal Bolus HbA1c 10th 25th 50th 75th 90th IQ Range NRM NORM
30 40 LA 45 RA 7.20% 87.0 107.9 134.0 164.0 198.1 56.0 133.89583 139.52941

FIG 7.5  Universal CGM output as ambulatory glucose profile (AGP). A. Subject without diabetes. B. Subject with T2DM.

It is important to keep in mind that both SMBG and the benefits of SMBG or CGM, it must be used with appro-
CGM are but tools to aid in diabetes self‐management and to priate understanding of how the information gleaned from
provide actionable data for patients and their caregivers. these tools is to be used to monitor, and if need be alter, one’s
They are not a therapeutic intervention expected to improve current therapies. These tools need to be viewed in the larger
overall glycemic control in and of themselves. To optimize context of the patient’s knowledge, abilities, financial
 What Is the Role of Self-Monitoring in Diabetes?
resources, and desire to utilize these tools appropriately.
Current data, especially on the use of SMBG in patients with
T2DM not utilizing insulin therapy remains fraught with
controversy. Opinions are driven by contradictory conclu-
sions from multiple small, sometimes poorly designed
clinical trials, as well as various meta‐analysis. There
remains a real need for well‐controlled, thoughtfully
designed randomized clinically controlled trials to help
answer these questions. Groups are now attempting to
define those important characteristics to be included in
such trials [16, 21].
References
1. MacLean H. Modern Methods in the Diagnosis and Treatment
of Glycosuria and Diabetes. London: Constable, 1924.
2. Dufaitre‐Patouraux L, Vague P, Lassmann‐Vague V. History,
accuracy and precision of SMBG devices. Diabetes Metab.
2003;29(2 Pt 2):S7–14.
3. Centers for Disease Control and Prevention. MMWR self‐
monitoring of blood glucose among adults with diabetes –
United States, 1997–2006. Morb Mortal Wkly Rep. 2007;
56(43):1133–1137.
4. The Diabetes Control and Complications Trial/
Epidemiology of Diabetes Interventions and Complications
(DCCT/EDIC) Study Research Group. Intensive diabetes
treatment and cardiovascular disease in patients with type 1
diabetes. N Engl J Med. 2005;353:2643–2653.
5. The Diabetes Control and Complications Trial/Epidemiology
of Diabetes Interventions and Complications Research
Group. Sustained effect of intensive treatment of type 1
diabetes mellitus on development and progression of
diabetic nephropathy. JAMA. 2003;290: 2159–2167.
6. Holman RR, Paul SK, Bethel MA et al. 10‐year follow‐up of
intensive glucose control in type 2 diabetes. N Engl J Med.
2008;359:1577–1589.
7. The Diabetes Control and Complications Trial Research
Group. The effect of intensive treatment of diabetes on the
development and progression of long‐term complications in
insulin‐dependent diabetes mellitus. N Engl J Med.
1993;329:977–986.
8. The Action to Control Cardiovascular Risk in Diabetes Study
Group. Effects of intensive glucose lowering in type 2
diabetes. N Engl J Med. 2008;358:2545–2559.
9. Riddle MC, Ambrosius WT, Brillon DJ, Buse JB, Byington
RP, Cohen RM, Goff DC Jr, Malozowski S, Margolis KL,
Probstfield JL, Schnall A, Seaquist ER, and for the Action to
99
Control Cardiovascular Risk in Diabetes (ACCORD)
Investigators. Diabetes Care. 2010;33(5):983–990.
10. Del Prato S, LaSalle J, Matthaei S, Bailey CJ, on behalf of the
Global Partnership for Effective Diabetes Management.
Tailoring treatment to the individual in type 2 diabetes prac-
tical guidance from the Global Partnership for Effective
Diabetes Management. Int J Clin Pract. 2010;64:295–304.
11. Schiffrin A, Belmonte M. Multiple daily self‐glucose moni-
toring: its essential role in long‐term glucose control in insu-
lin‐dependent diabetic patients treated with pump and
multiple subcutaneous injections. Diabetes Care. 1982;5:
479–484.
12. Hansen MV, Pedersen‐Bjergaard U, Heller SR et  al.
Frequency and motives of blood glucose self‐monitoring in
type 1 diabetes. Diabetes Res Clin Pract. 2009;85(2):183–
188. Epub 2009 Jun 3.
13. Evans JM, Newton RW, Ruta DA et al. Frequency of blood
glucose monitoring in relation to glycaemic control: obser-
vational study with diabetes database. BMJ. 1999;319(7202):
83–86.
14. Davidson PC, Bode BW, Steed RD, Hebblewhite HR. A
cause‐and‐effect‐based mathematical curvilinear model that
predicts the effects of self‐monitoring of blood glucose fre-
quency on hemoglobin A1c and is suitable for statistical cor-
relations. J Diabetes Sci Technol. 2007;1(6):850–856.
15. Murata GH, Shah JH, Hoffman RM et al. Intensified blood
glucose monitoring improves glycemic control in stable,
insulin‐treated veterans with type 2 diabetes: the Diabetes
Outcomes in Veterans Study (DOVES). Diabetes Care.
2003;26(6):1759–1763.
16. Klonoff DC, Bergenstal R, Blonde R et al. Consensus report
of the coalition for clinical research – self‐monitoring of
blood glucose. J Diabetes Sci Technol. 2008;2(6):
1030–1053.
17. Davis WA, Bruce DG, Davis TME. Does self‐monitor of
blood glucose improve outcome in type 2 diabetes? The
Fremantle Study. Diabetologia. 2007;50:510–515.
18. Welschen LMC, Bloemendal E, Nijpels G et al. Self‐monitor-
ing of blood glucose in patients with type 2 diabetes mellitus
who are not using insulin. Cochrane Database Syst Rev.
2005;2 (Art. No.: CD005060). DOI: 10.1002/14651858.
CD005060.pub2.
19. Farmer A, Wade A, Goyder E et al. Impact of self monitoring
of blood glucose in the management of patients with non‐
insulin treated diabetes: open parallel group randomized
trial. BMJ. 2007;335(7611):132.
20. Farmer A, Wade A, French DP et al. Blood glucose self‐mon-
itoring in type 2 diabetes: a randomised controlled trial
Health Technol Assess. 2009;13(15).
 100 Initial Evaluation and Management of Diabetes
21. Hirsch IB, Bode BW, Childs BP et  al. Self‐Monitoring of
Blood Glucose (SMBG) in insulin‐ and non‐insulinusing
adults with diabetes: consensus recommendations for
improving SMBG accuracy, utilization, and research.
Diabetes Technol Ther. 2008;10(6):419–439.
22. Bergenstal R, Johnson M, Powers M et al. Adjust to target in
type 2 diabetes. Diabetes Care. 2008;31(7):1305–1310.
23. Brown LB, Nichols GA, Perry A. The burden of treatment
failure in type 2 diabetes. Diabetes Care. 2004;27:
1535–1540.
24. DeVeciana M, Major C, Morgan M et al. Postprandial versus
preprandial blood glucose monitoring in women with gesta-
tional diabetes mellitus requiring insulin therapy. N Engl J
Med. 1995;333(19):1237–1241.
25. Wecher DJ, Kaufmann RC, Amankwah KS et al. Prevention
of neonatal macrosomia in gestational diabetes by the use of
intensive dietary therapy and home glucose monitoring. Am
J Perinatol. 1991;8:131–134.
26. Homko CJ, Sivan E, Reece EA. The impact of self‐monitor-
ing of blood glucose on self‐efficacy and pregnancy out-
comes in women with diet‐controlled gestational diabetes.
Diabetes Educ. 2002;28(3):435–443.
27. 2010 ADA Clinical Practice Recommendations.
28. The DECODE Study Group. Glucose tolerance and mortal-
ity: comparison of WHO and American Diabetes
Association diagnostic criteria. The DECODE study group.
European Diabetes Epidemiology Group. Diabetes epidemi-
ology: collaborative analysis of diagnostic criteria in Europe.
Lancet. 1999;354:617–621.
29. Ohkubo Y, Kishikawa H, Araki E et  al. Intensive insulin
therapy prevents the progression of diabetic microvascular
complications in Japanese patients with non‐insulin‐
dependent diabetes mellitus: a randomized prospective 6‐
year study. Diabetes Res Clin Pract. 1995;28:103–117.
30. Avignon A, Radauceanu A, Monnier L. Nonfasting plasma glu-
cose is a better marker of diabetic control than fasting plasma
glucose in type 2 diabetes. Diabetes Care. 1997;20: 1822–1826.
31. Chiasson JL, Josse RG, Gomis R et  al. Acarbose treatment
and the risk of cardiovascular disease and hypertension in
patients with impaired glucose tolerance: the STOP‐NIDDM
trial. JAMA. 2003;290:486–494.
32. Hanefeld M, Cagatay M, Petrowitsch T et  al. Acarbose
reduces the risk for myocardial infarction in type 2 diabetic
patients: meta‐analysis of seven long‐term studies. Eur
Heart J. 2004;25(1):10–16.
33. Ceriello A, Ihnat M, Thorpe J. Clinical review 2: the “meta-
bolic memory”: is more than just tight glucose control nec-
essary to prevent diabetic complications? Journal Of Clinical
Endocrinology And Metabolism. 94(2):410–415, 2009.
34. Brownlee M, Hirsch I. Glycemic variability: a hemoglobin
A1c‐independent risk factor for diabetic complications.
JAMA. 295(14):1707–1708, 2006.
35. DCCT Study Group. The relationship of glycemic exposure
(HbA1c) to the risk of development and progression of
retinopathy in the diabetes control and complications trial.
Diabetes. 1995;44:968–983.
36. Lachin JM, Genuth S, Nathan DM, Zinman B, Rutledge BN
and for the DCCT/EDIC Research Group. Effect of glycemic
exposure on the risk of microvascular complications in the
diabetes control and complications trial – Revisited.
Diabetes. 2008;57:995–1001.
37. Guideline for Management of Postmeal Glucose.
International Diabetes Federation, 2007; www.idf.org.
38. Monnier L, Lapinski H, Colette C. Contributions of fasting
and postprandial plasma glucose increments to the overall
diurnal hyperglycemia of type 2 diabetic patients: variations
with increasing levels of HbA1c. Diabetes Care. 2003;26:
881–885.
39. Juvenile Diabetes Research Foundation Continuous Glucose
Monitoring Study Group. Continuous glucose monitoring
and intensive treatment of type 1 diabetes. N Engl J Med.
2008;359:1464–1476. DOI: 10.1056/NEJMoa0805017.
40. Hirsch IB, Abelseth J, Bode BW et  al. Sensor‐augmented
insulin pump therapy: results of the first randomized treat‐
to‐target study. Diabetes Technol Ther. 2008;10:377–383.
41. Bergenstal et al. Effectiveness of sensor‐augmented insulin‐
pump therapy in type 1 diabetes. N Engl J Med. 2010;363:
311–320.
42. Bolinder et al. Novel glucose‐sensing technology and hypo-
glycaemia in type 1 diabetes: a multicentre, non‐masked,
randomised controlled trial. Lancet. 2016;388(10057):
2254–2263.
43. Haak et  al. Flash glucose‐sensing technology as a replace-
ment for blood glucose monitoring for the management of
insulin‐treated type 2 diabetes: a multicenter, open‐label
randomized controlled trial. Diabetes Ther. 2017;8(1):
55–73.
44. Kudva et al. Approach to using trend arrows in the FreeStyle
Libre flash glucose monitoring systems in adults. J Endocr
Soc. 2018;2(12):1320–1337.
 8 Does HbA1c Remain the Most Important
Therapeutic Target in Outpatient
Management of Diabetes?
Kristen Gonzales1 and Steve A. Smith2
1
Department of Internal Medicine, Division of Endocrinology, Diabetes, and Metabolism, University of New Mexico
Health Sciences Center, Albuquerque, NM, USA
2
Professor of Medicine, Division of Endocrinology, Diabetes, Metabolism, and Nutrition, Mayo Clinic, Rochester, MN, USA
LE A R NI NG P OI NT S
●● HbA1c is a measure for not only characterizing
diabetes, but also in relaying predictive information
regarding the effects of chronic hyperglycemia
●● Various clinical states can interfere with accurate
interpretation of HbA1c.
●● Going beyond HbA1c as a measure implies the
utilization of novel glycemic measures to guide more
comprehensive assessments that place the patient at
the front‐lines of management.
The glucose hypothesis and vascular risk
Hyperglycemia is strongly associated with both micro‐ and
macrovascular events in people with diabetes. The Diabetes
Control and Complication (DCCT) and the Kumamoto
trials were the first to demonstrate the direct effects of tar-
geted improvements in glycemic control in both Type 1
(DM1) and Type 2 (DM2) diabetes, respectively  [1, 2].
Since those early studies, many randomized control trials
have demonstrated favorable impacts on cardiovascular
and renal outcomes with improved glycemic control, fur-
ther confirming the notion that glycemic control is strongly
linked to vascular risk [3–8].
A number of hypotheses have been proposed for the
development of vascular complications as a result of
Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
intermittent and/or sustained hyperglycemia  [9]. These
have included impaired endothelial cell dysfunction [10],
endothelium‐dependent relaxation, endothelial cell apop-
tosis [11], altered nitric oxide and aldose reductase, oxygen
derived free radicals  [12, 13], protein kinase C, Na+K+‐
ATPase, and advanced glycosylation end products  [14].
Some of the most highly‐researched mechanisms for
microvascular disruption related to hyperglycemia have
been the aldose reductase theory, advanced glycation end‐
product theory, reactive oxygen intermediate pathway, and
the protein kinase C pathway. However, clinical evidence is
lacking for these proposed mechanisms, making develop-
ment of targeted therapies challenging [9]. Thus, minimi-
zation and prevention of vascular complications through
glycemic control is critical, and accurate measures of
hyperglycemia are imperative to this goal.
Ideal clinical measures of hyperglycemia should inform
the patient and clinician in their management and be pre-
dictive of both short‐ and long‐term vascular complica-
tions. In this chapter, we will review the historical and
future assessments of glycemic control in the care of people
with diabetes. This requires precise methods for measuring
(1) chronic and intermittent hyperglycemia, (2) hypoglyce-
mia, and (3) glycemic variability. We will address measures
for each of these and further evaluate the evidence sup-
porting standard or more novel measures of glycemic
control.
101
 102 Initial Evaluation and Management of Diabetes
Measuring glycemic control in the clinical
management of diabetes
Hemoglobin A1c: Discovery and integration
into clinical practice
Discussion of the hemoglobin A1c (HbA1c) and its role in
diabetes management is key to the examination of optimal
modalities for measuring glycemic control. HbA1c was
first discovered in 1968 and introduced into clinical prac-
tice in 1977  [15, 16]. It has continued to be used as the
main outcome for diabetes management in both clinical
and research settings given its value in predicting risk asso-
ciated with long‐term glycemic exposure [17]. Historically,
the American Diabetes Association (ADA) has maintained
that the primary measures of glycemic control should include
both self‐monitored blood glucose values (SMBG) and
HbA1c. However, with the advent of continuous glucose
monitoring (CGM) and the recognition that interstitial glu-
cose as measured through CGM is well‐correlated with blood
glucose, the ADA officially incorporated CGM as a standard
measure of glycemic control in 2008 [18]. CGM, as will be dis-
cussed, has paved the way for enhanced precision in glycemic
measures which have been associated with improved out-
comes. We will discuss each of these measures in the subse-
quent sections, beginning with an overview of the role for
HbA1c in both past and present‐day diabetes care.
HbA1c is a hemoglobin A moiety that has undergone
valine glycosylation at the N‐terminus of its beta‐chain
through a non‐enzymatic ketoamine linkage. This ketoam-
ine linkage creates a stable HbA1c complex which forms
slowly and continuously throughout the life span of the red
blood cell, thus reflecting hemoglobin glycosylation over a
2‐ to 3‐month span. It is therefore an indirect measure of
average glycemic exposure. Rhabar and Trivelli first
described its importance in diabetes by noting a two‐fold
increase in this moiety among individuals with diabetes
compared to those without [16, 19]. Trivelli and colleagues
also described a relationship between HbA1c, mean blood
glucose, and long‐term complications of diabetes, thus
establishing HbA1c as a measure not only for characteriz-
ing diabetes, but also in relaying predictive information
regarding the effects of chronic hyperglycemia  [19].
However, validation and standardization of the HbA1c
assay took several decades, as the methods used for isola-
tion were laborious and inconsistent, thus prolonging its
acceptance into routine clinical practice [15].
The impetus for standardization of the HbA1c assay was
borne out of the landmark findings of the DCCT. In this
study, HbA1c was used as a primary measurement of glyce-
mic control, and measurements were obtained through a
centralized laboratory using high‐performance liquid
chromatography. With the demonstration of marked ben-
efit in microvascular outcomes through improved glyce-
mic control, the HbA1c quickly became recognized as a
key target for diabetes care. As a result, national programs
in the United States, Japan, and Sweden were established to
standardize the assay such that techniques and results
would be directly comparable to those of the DCCT.
Ultimately, the International Federation of Clinical
Chemistry and Laboratory Medicine (IFCC) Working
Group on Standardization of HbA1c established an inter-
national HbA1c assay employing either HPLC mass spec-
trometry or capillary electrophoresis. The IFCC also
established a reporting system using units of either mmol/
dL or percent glycosylation  [15]. In total, these efforts
allowed for great strides to be made in the minimization of
inter‐laboratory variability and propelled the HbA1c for-
ward as a meaningful population measure of glycemic
control.
HbA1c: Role in predicting micro‐ and
macrovascular complications
Multiple prospective trials have employed HbA1c as a pri-
mary measure for glucose control when assessing out-
comes associated with long‐standing hyperglycemia. The
DCCT (1982–1993), was a prospective clinical trial which
randomized patients with DM1 to receive either intensive
insulin therapy, where the average HbA1c achieved was
7.1%, or conventional insulin therapy, where the average
HbA1c achieved was 9.1%. Primary outcomes were the
development and progression of retinopathy, nephropathy,
and neuropathy. The study demonstrated a strong correla-
tion between both prevention and progression of retinopa-
thy with HbA1c reduction in the intensive treatment
group, and progression was likewise linked exponentially
with rising HbA1c levels [1]. In a long‐term observational
follow‐up of the DCCT cohort, the Epidemiology of
Diabetes Interventions and Complications (EDIC) trial,
 Is HbA1c the Most Important Therapeutic Target in the Outpatient Management of Diabetes?
HbA1c values converged between both groups. Yet, there
remained a marked risk reduction in microvascular disease
in the intensive group compared to the conventional group
over 2 decades of follow‐up. There were also significant
reductions in fatal and non‐fatal myocardial infarctions and
strokes. These data suggested that early, intensive lowering
of HbA1c portended long‐term vascular benefits in people
with DM1, a phenomenon known as “metabolic mem-
ory.” [20] Similar findings regarding micro‐ and macrovas-
cular risk reduction among people with DM2 who achieved
early HbA1c lowering were demonstrated in the United
Kingdom Prospective Diabetes Study (UKPDS) [21, 22].
Notably, neither the DCCT nor the UKPDS were
designed to establish HbA1c thresholds beyond which
microvascular disease progression was likely to occur.
Rather, insight into these HbA1c cutoffs was gleaned from
the 6‐year Kumamoto Study, a randomized control trial
involving individuals with DM2 randomized to receive
either intensive or conventional insulin therapy. As
expected, rates of microvascular complications were
reduced in those randomized to the intensive insulin ther-
apy group, and the threshold for microvascular progression
was correlated with an average HbA1c > 6.5%, a FPG <
110 mg/dL, and a 2‐hour postprandial glucose > 180 mg/
dL  [2]. Similarly, the Stockholm Diabetes Intervention
Study, a smaller 7.5‐year prospective trial, demonstrated
that individuals with mild retinopathy at study entry did
not develop severe retinopathy, nephropathy, or progres-
sive eye disease if mean HbA1c was maintained < 7% for
the duration of the study [23]. Finally, a compilation of epi-
demiologic data has shown that HbA1c is useful for pre-
dicting development of retinopathy, and an HbA1c of ≥
6.5% represents an inflection point beyond which retinop-
athy risk substantially increases, thus justifying this as a
threshold for the diagnosis of diabetes [24].
With identification of HbA1c thresholds below which
vascular risks are significantly reduced, a new challenge in
glycemic control and diabetes management emerged: that
of hypoglycemia. Perhaps the most robust trials to high-
light this risk with intensive HbA1c lowering were the
VADT, ADVANCE and ACCORD trials, each involving
individuals with DM2. These trials targeted HbA1c goals at
least less than 7% in the intensive treatment arms and dem-
onstrated significantly higher rates of hypoglycemia in
103
these groups [25–27]. Thus, the benefits of HbA1c lower-
ing based on established cutoffs for development of micro-
vascular complications must be counterbalanced by the
risks of hypoglycemia, particularly in select populations,
thus broadening the scope of glycemic control.
Clinical utility of HbA1c
There are several reasons for why the HbA1c has so deeply
permeated general understanding of chronic glycemic con-
trol and the complications associated with diabetes.
HbA1c, being a measure of hemoglobin glycosylation, pro-
vides insight into the average degree of glycemic exposure
over a 2‐ to 3‐month span [28]. It is widely accepted that
reducing blood glucose levels corresponds to a reduction
in microvascular and macrovascular complications, at least
when using HbA1c and SMBG values as surrogate meas-
ures, and these generally serve as good predictors for devel-
opment of diabetes complications [1, 2, 20, 21, 25].
In 2009, the International Expert Committee on diabe-
tes officially incorporated HbA1c into the diagnostic crite-
ria for diabetes, establishing a cutoff of ≥ 6.5% as being
consistent with the diagnosis and correlating with a point
in the disease beyond which microvascular risk markedly
increased  [24]. Thus, based on the available evidence to
that point, HbA1c was and continues to be a valuable
measure of glycemic exposure. From a clinical perspective,
performance of the HbA1c assay is convenient, timely,
does not require fasting, portends a high degree of pre‐ana-
lytical stability, and is less impacted by daily perturbations
in other metabolic factors. From a population health stand-
point, the HbA1c is perhaps the best available tool for
establishing glycemic goals which can then be further indi-
vidualized according to patient‐specific factors. The use of
HbA1c to set individualized therapeutic targets, however,
is limited. As we will explore in the subsequent sections, a
focus on HbA1c targets alone may be insufficient for opti-
mizing diabetes management and preventing related vas-
cular outcomes.
Limitations of HbA1c measurement
Due to the fact that HbA1c measures glycosylation of
hemoglobin, there are various clinical states that can inter-
fere with accurate interpretation of HbA1c. Conditions
such as sickle cell anemia, pregnancy and post‐partum
 104 Initial Evaluation and Management of Diabetes
states, glucose‐6‐phosphate dehydrogenase deficiency,
HIV, hemodialysis, recent blood loss or transfusion, eryth-
ropoietin therapy, and B12, iron, or folate deficiencies may
significantly impact the reported values [29]. Thus, alter-
nate measures should be employed for management deci-
sions in such cases. In clinical practice, fasting plasma
glucose concentrations, and/or other measures of glycemia
which will be discussed should be utilized.
Additionally, HbA1c does not reflect rapid excursions in
glucose levels and, and it provides limited value in guiding
daily self‐management. As daily self‐management entails
use of data derived from either SMBG measurements or
CGM, correlation of these values with HbA1c is critical for
disease management.
Traditional measures of blood glucose
and glucose exposure
Hyperglycemia can be detected through both clinical and bio-
chemical assessments. Polyuria, polydipsia, and nocturia are
perhaps the most common symptoms. Depending upon the
duration and severity of hyperglycemia, other clinical features
may include fatigue, lethargy, blurred vision, weight loss, keto-
sis, and infections. However, reliance on clinical symptoms is
insufficient for the assessment of glycemic control. [30]
Prior to the mid‐1970s, glucose control was measured pri-
marily through urine testing. While random or 24‐hour urine
glucose collections provide some insight into glycemic bur-
den, they are relatively inaccurate and fail to adequately guide
targeted treatment goals. Renal glucose thresholds are highly
variable between individuals, and lack of glucosuria provides
little information regarding the presence of hypoglycemia,
euglycemia, or even mild hyperglycemia [31]. Thus, there is
poor correlation between urinary and blood glucose values.
The development of capillary glucose testing largely
replaced home urinary measurements by the 1980s and
revolutionized management of diabetes by allowing for tar-
geted goals for blood glucose  [31]. SMBG with capillary
glucose measurements remains a mainstay of measuring
glycemic control and provides day‐to‐day guidance for
management decisions. The evidence for SMBG will be
discussed in the subsequent section.
In the 1990s, discovery of glycated proteins, namely fruc-
tosamine, provided an alternative measurement of glycemic
control. Given the protein half life in circulation is roughly
14 days, fructosamine provides a semi‐chronic assessment
of glucose exposure  [31]. Finally, 1,5‐anhydroglucitol is
another measure that provides some insight into hypergly-
cemia, traditionally declining in the serum with progres-
sive hyperglycemia and rising with improved glycemic
control [32]. While these latter measures are reasonable to
consider in the presence of medical conditions which may
interfere with accurate HbA1c measurements, their prog-
nostic value for diabetes complications is not well‐estab-
lished, thus limiting their utility as ideal measures of
glycemic control.
Self‐monitored blood glucose (SMBG)
SMBG, as measured through hand‐held glucometers, has
largely replaced traditional measures of glycemia in the
day‐to‐day care of individuals with diabetes [33, 34]. Use of
SMBG, particularly when used with increased frequency,
provides details regarding daily glycemic patterns, thus
enabling patients and clinicians to tailor therapies toward
specific glucose goals. These glycemic “snapshots” have
been deeply embedded into clinical management of diabe-
tes, and their use is strongly correlated with reductions in
HbA1c.
Currently, the ADA recommends that patients on an
intensive insulin regimen perform SMBG measurements
before meals and snacks, fasting, before bedtime, if needed
postprandial, before exercise, when experiencing a low,
until the low is resolved, and before driving. It is recog-
nized that frequent SMBG measures can guide medical
nutrition therapy, physical activity, insulin dose adjust-
ments, and prevention/treatment of hypoglycemia [35].
In the Diabetes Control and Complications Trial
(DCCT), participants randomized to the intensive treat-
ment group performed SMBG measurements at least 4
times daily, in addition to a 3 a.m. check once weekly, and
7 times daily once every 3 months [1, 36]. In contrast the
conventional treatment group performed SMBG measure-
ments twice daily. While SMBG frequency was not corre-
lated directly to the primary outcomes of the trial, the
intensive treatment group overall demonstrated delayed
progression and development of microvascular complica-
tions  [1]. This data would suggest that SMBG frequency
may in fact be a significant component of improved glyce-
mic control.
 Is HbA1c the Most Important Therapeutic Target in the Outpatient Management of Diabetes?
Several large registry studies have since demonstrated
that SMBG frequency is strongly correlated with a decrease
in HbA1c for all age groups among both insulin pump and
multiple daily injection (MDI) users. In one study using
the Type 1 Diabetes Exchange Registry, a repository of 20
555 children and adults with type 1 diabetes, the individu-
als who tested 3–4 times daily had a mean HbA1c of 8.6%,
whereas those who tested ≥ 10 times daily had a mean
HbA1c of 7.6%. The HbA1c lowering correlated with
increasing frequency of SMBG measurements and reached
a plateau at 10–12 checks daily  [37]. In a similar study
using an Austrian database of both patients with type 1 and
2 diabetes, HbA1c reduction was again correlated with
SMBG frequency. The HbA1c reduction in this study was
shown to be 0.26% for every one additional SMBG check/
day, p < 0.0001. Notably, this effect was only seen in patients
using insulin therapy. Indeed, there was a negative correla-
tion of SMBG monitoring frequency and glycemic control
noted among patients with type 2 diabetes treated with oral
agents [38].
While SMBG frequency has been correlated with
improvement in glycemic control, this alone is not the sole
factor in HbA1c improvement. Rather, accurate interpreta-
tion of the readings and integration of the data obtained
from SMBG into daily management is an integral compo-
nent of diabetes management. Additional factors to con-
sider regarding SMBG as a measure of glycemic control are
patient technique when testing, device accuracy, and the
patient burden of frequent monitoring. With this back-
ground, how is the information obtained through
SMBG measurement linked to overall outcomes in diabe-
tes, and is this information valuable in influencing vascular
outcomes?
Additional therapeutic targets to consider
for hyperglycemia in the outpatient
management of diabetes
The term “beyond A1c” has recently been introduced into
the diabetes lexicon as a concept describing the advance-
ment of glycemic measures toward more comprehensive
assessments that place the patient at the front‐lines of man-
agement. SMBG has paved the way for insights into these
more nuanced measures, providing insight into valuable
105
metrics such as fasting, pre‐ and postprandial levels. As
these measures are further examined in relation to diabetes
outcomes, it becomes apparent that a focus on HbA1c
alone may obscure other meaningful therapeutic targets.
Post prandial glucose (PPG)
Population studies have proposed a possible relationship
between risk of cardiovascular events and postprandial
hyperglycemia (PPG)  [39, 40]. Initial studies designed to
investigate this possible relationship have been limited to
cross sectional observations and associations of elevated
values at 2 hours on oral glucose tolerance testing (OGTT)
as opposed to mixed‐meal or real‐life glucose expo-
sure  [41–43]. Studies that have compared outcomes for
individuals that emphasize PPG; meglitinides
(NAVIGATOR) [44], alpha glucosidase inhibitors (STOP‐
NIDDM)  [45], and basal insulin only (HEART2D) have
been limited based on technical challenges in assessing
duration of glycemic exposure  [42]. In the San Luigi
Gonzaga Diabetes Study with 14 years of follow‐up, HbA1c
and glucose levels 2 hours after lunch were predictive of
first cardiovascular events and all‐cause mortality,  [46]
while Hopper and colleagues have completed a meta analy-
sis that included 23 152 patients with impaired glucose
intolerance from 10 controlled trials randomized to diet,
exercise, or hypoglycemic agent. After an average of 3.75 of
years of follow‐up, there was no reduction in all‐cause or
cardiovascular mortality [47].
Just as the Kumamoto trial demonstrated a reduced pro-
gression of microvascular complications in patients who
maintained an average HbA1c of < 6.5% during the 6‐year
follow‐up period, it also demonstrated that this reduction
was correlated with 2‐hour postprandial levels < 180 mg/
dL. Rising PPG levels were associated with worsening
retinopathy and nephropathy [2].
PPG, like HbA1c, offers little insight into the assessment
of glycemic variability and, therefore, provides little addi-
tional information beyond HbA1c. Because of the observa-
tion that the best predictor of postprandial glucose control
is preprandial levels, it has been argued that the best
treatment approach for hyperglycemia would be to limit
the incidence of hypoglycemia by lowering preprandial
fasting hyperglycemia prior to emphasis of postprandial
hyperglycemia [48].
 106 Initial Evaluation and Management of Diabetes
Continuous glucose monitoring (CGM) and
glycemic control
While SMBG has been crucial in diabetes care and fre-
quency of use linked to enhanced glycemic control, many
individuals on insulin are unable to achieve glycemic tar-
gets even with standard 4‐times‐daily SMBG monitoring.
One reason for this is the rarity with which postprandial
and overnight glucose profiles are able to be examined with
standard SMBG measures and home glucometers. Over the
last few decades, CGM has been developed in an effort to
overcome the monitoring limitations of conventional
SMBG.
CGM first came into use in 1999 and, since that time,
technological advances have led to marked improvements
in both accuracy and ease of use  [49]. There has subse-
quently been an increasing real‐time use of CGM in the
management of glucose control. More recently, intermit-
tently‐scanned CGM devices, which minimize the burden
of frequent fingerstick checks, have been introduced into
clinical practice. These devices have all allowed for
enhanced utilization of glycemic measurements while
minimizing the daily burden of disease on individuals with
diabetes.
CGM mathematically derives estimates of plasma glu-
cose from the measurement of interstitial glucose (which
correlates well with plasma glucose), providing up to 288
glucose measurements daily. Sensors are calibrated with
SMBG measurements and, as of 2013, the latter is recom-
mended for making acute adjustments in insulin. Several
studies of CGM use have been conducted in patients with
DM1, although more robust data is needed in those with
DM2. Using insulin pumps with CGM has demonstrated
benefit in achieving glycemic control, as well. Still, fewer
studies have looked at the use of CGM in patients with
DM1 using MDI. This is an important issue given the
majority of patients with DM1 use MDI rather than pump
therapy.
The DIAMOND trial sought to specifically determine
the role of CGM in patients with DM1 using MDI. It found
that patients using CGM had a 1% reduction in in HbA1c
after 6 months compared to 0.4% reduction in the control
group (MDI without CGM) or SMBG checks 4 times daily.
Notably, this trial also measured many secondary out-
comes including glucose variability (CV), time per day in
range (70–180  mg/dL), time spent per day in hypoglyce-
mia, and time spent in hyperglycemia [50]. The magnitude
of HbA1c benefit over 6  months in the CGM vs. control
group was similar to that seen in the SWITCH study. In the
latter, there was a 0.43% mean difference in HbA1c between
patients with insulin pumps using CGM vs. those not using
CGM. Secondary endpoints in SWITCH included changes
in glycemic patterns, and time spent hypoglycemia, hyper-
glycemia, and normoglycemia [51]. As with any intensive
method of achieving improved glycemic control and
HbA1c reduction, the counterbalance of hypoglycemia is
of paramount importance.
CGM and hypoglycemia
Several major trials examining the effectiveness and safety
of CGM have either reported no increase or significant
reduction in hypoglycemia events compared to non‐CGM
users despite universal HbA1c‐lowering in the studies
among CGM users. The Juvenile Diabetes Research
Foundation (JDRF) study was perhaps one of the first to
design a randomized trial in which individuals with DM1
and HbA1c values < 7.0% at study entry were recruited to
either CGM‐use or standard monitoring. The study dem-
onstrated a decrease in time spent with a blood glucose ≤
70  mg/dL among CGM‐users (54 vs. 91  mins/day),
although the difference did not reach statistical signifi-
cance  [52]. The SWITCH trial documented an overall
improvement in HbA1c and no difference in hypoglycemia
events among CGM‐users vs. non‐users [51]. Similarly, the
(STAR) 3 trial examined sensor‐augmented pump (SAP)
therapy vs. standard MDI therapy in participants with
DM1 and found no significant differences in hypoglyce-
mia, but overall greater reduction in HbA1c among the
SAP group [53]. However, the ASPIRE trial included DM1
patients randomized to SAP vs. standard pump therapy
without CGM. In this study the interventional group had
31.8% reduced frequency of nocturnal hypoglycemia over
3‐month use with no difference in HbA1c compared to
standard group over that time [54].
One crucial element of modern CGMs that has been
shown to reduce hypoglycemia is that of automated low
glucose suspend (LGS). LGS is an automated feature of
SAP therapy in which insulin delivery is automatically sus-
pended when interstitial glucose values fall below a certain
 Is HbA1c the Most Important Therapeutic Target in the Outpatient Management of Diabetes? 107

threshold. Indeed, the ASPIRE study utilized this LGS involving CGM has been the hybrid closed‐loop pump, a
feature  [53]. Studies involving pediatric populations system in which basal rates are determined and adjusted
with DM1  have also demonstrated promising results automatically by an individualized algorithm derived from
regarding reduced time spent in hypoglycemia (< CGM. Use of this pump has not only shown improve-
70 mg/dl, 101 ± 68 versus 58 ± 33 min/day; p < .002), as ment in HbA1c, but also in many of the newly‐standard-
well as reduced number of hypoglycemic excursions ized CGM metrics: Time in target range increased from
during the day (< 70 mg/dl, 1.27 ± 0.75 vs. 0.95 ± 0.49, p 64% to > 70%; time below target range (≤ 70  mg/dL)
< .01; < 40  mg/dl, 0.28 ± 0.18 versus 0.13 ± 0.14, p < decreased by 30%; and coefficient of variation decreased
.005) [55]. from 33% to 30% [57]. Thus, familiarity with how these
metrics relate to glycemic control is critical as diabetes
Utility of CGM metrics in the management of care transitions beyond HbA1c. We will discuss a few of
diabetes the major metrics and the evidence for their use in dia-
CGM use has been strongly linked to improved glycemic betes care below.
control as measured using the primary outcome of HbA1c.
However, these benefits stem from the detailed informa- Metric 1: Duration of CGM use
tion obtained from these devices, as they allow for immedi- One caveat to the benefits seen in studies involving
ate detection of glycemic excursions and daily patterns not CGM was that sufficient use of the CGM device was
otherwise attainable through standard SMBG using finger- necessary. Among CGM‐users in the SWITCH trial,
stick checks. Using CGM data, a series of metrics have been mean sensor use was 80% of the required time, and
developed which have proven correlation with glycemic those with < 70% use demonstrated smaller decreases in
control. Some of the more common ones include time HbA1c [51]. In the JDRF study, the group that achieved
within/above/below range, glycemic variability as meas- the greatest HbA1c benefit with CGM were those ≥ 25
ured through a coefficient of variation or standard devia- years of age, and roughly 83% of the participants in this
tion (CV or SD), and duration of CGM use. Table  8.1 group utilized the CGM at least 6  days per week. Use
provides a comprehensive list of standardized metrics for was significantly less in the younger age groups (8–24
CGM reporting [56]. Notably, the most recent technology years old), and there were no differences in glycemic
control noted between CGM‐users and non‐users (con-
trol) in these groups [58].
Duration of CGM use, in addition to having proven
TABLE 8.1  International consensus on use of CGM.
benefit on glycemic control during the study time‐frames,
Mean glucose eA1c has also been correlated with long‐term glycemic control.
Xing et al. showed that CGM use between 12–15 days was
% time level 2 TBR Glucose metrics
well‐correlated (R2 range 0.66–0.75) with 3‐month glyce-
< 54 mgm% (3 mmol/L) Sleep, Wake, 24 hours
% time level 1 TBR Data sufficiency (2 weeks) mic control as measured by mean glucose, percentage of
< 70‐54 mgm% (3.9–3.0 mmol/L) values between 71–180 mg/dL and < 70 mg/dL, and glyce-
% time TIR Data sufficiency mic variability [59]. More recently, data using newer CGM
70‐180mgm% (3.9–10 mmol/L) 70–80% CGM data over technology has demonstrated an even greater correlation
2 weeks
(R2 range 0.84–0.86) of 14‐day CGM use with 3‐month
% time level 1 TAR Hypoglycemia standard
> 180mgm% (10 mmol//L)1 definition measures of glycemic control [60]. Beyond 14 days of use,
Hyperglycemia standard there was little incremental change noted in correlation
definition with long‐term outcome measures. Thus, 10‐ to 14‐day
% time level 2 TAR Hypoglycemia index sensor use is a reasonable target from which the available
> 250mgm% (13.9 mmol/L)
data can be used to derive estimations regarding average
Glycemic variability Hyperglycemia index
glucose and HbA1c.
 108 Initial Evaluation and Management of Diabetes
Metric 2: Time in range (TIR)
TIR is a key metric that has been identified as being per-
haps the most feasible therapeutic target in day‐to‐day
management of diabetes with use of CGM. It typically
refers to glucose values between 70 – 180 mg/dL. Although
the metrics listed in Table  8.1 address not only TIR, but
also time above range (TAR) and time below range (TBR),
it is the TIR metric that has been most heavily analyzed in
recent studies. However, in clinical practice, achievement
of TIR targets practically involves first a foremost an assess-
ment of TBR in order to prevent serious hypoglycemia, fol-
lowed then with a focus on reducing TAR. These
approaches, in turn, lead to increased TIR.
Perhaps the most robust data for TIR in patients with
DM1 comes from a post‐hoc analysis of the DCCT trial data
in which 7‐point blood glucose profiles collected quarterly
were analyzed to develop Cox proportional hazard models
for assessing the relationship between TIR and microvascu-
lar outcomes. Indeed, the hazard ratios for retinopathy and
microalbuminuria progression increased by 64% and 30%,
respectively, for each 10% reduction in TIR  [36]. Similar,
although less robust, data has been collected for patients
with DM2  in an observational trial which demonstrated a
correlation between diabetic retinopathy severity, reduced
TIR, and increased variability [61].
Observational studies using CGM metrics in ambula-
tory patients have found that pre‐ and postprandial hyper-
glycemia and glycemic variation contribute equally to
HbA1c levels and presumed vascular risk  [62]. Despite
these observational data, however, secondary analysis of
glycemic measures from DCCT/EDIC trials have not
shown an independent effect for cardiovascular risk
beyond mean glucose and HbA1c [63]. With the develop-
ment of CGM and the extent of more complete capture of
clinical glycemia, it is anticipated that “big data” analysis
methods will provide quantification and familiarity with
predictive methods and tools that go beyond HbA1c.
In clinical practice, the ability to provide targeted
goals for TAR, TIR, and TBR for patients is crucial in
optimizing the data collected with CGM. Targets should
be safe, achievable, and individualized according to
patient‐specific factors. Age, comorbidities, and preg-
nancy will all affect target glycemic goals for TIR. Using
CGM data, TIR recommendations are now endorsed by
the ADA and European Association for the Study of
Diabetes (EASD).
Metric 3: Glycemic variability
Glycemic goals in diabetes care are to maintain normal glu-
cose patterns while avoiding hypoglycemia [64]. Table 8.2
lists 14 different measures describing glycemic variability.
While variability as a concept is intuitive, the calculation
and interpretation of these measures have not been trans-
parent or intuitive to clinicians and patients unfamiliar
with the metrics. To date, they have not been universally
used to provide guidance to patients or clinicians in their
glycemic management [65].
Mathematical models attempting to quantify glycemic
variability have been focused on the statistical middle, and
range of glycemic measures have used non‐parametric
analyses due to lack of normal distribution of the data [66].
The graphic presentation of trends over time have helped
TABLE 8.2  Measures used to quantify glycemic variation
(modified from [64]).
Lability index mean +3 SD of the urinary glucose in gm/hr
Mean glucose Arithmetic mean within a time frame
Glucose index Low/High glucose index
Risk index Blood glucose risk index
Rate of change Blood glucose rate of change
M value Stability of glucose compared to ideal
MAGE Each blood glucose decreases from peak to
nadir
MODD or Across The absolute value of the difference
Day Variation between glucose values at the same time on
(ADV) consecutive days
J index J = 0.001 (MBG + SD)2 for BG
measurements in mg/dl, J = 0.324(MBG +
SD)2 for BG measurements in mmol/l.
VGA Variable Grid Analysis (VGA) – minimum/
maximum plot over a certain period of
observation
CONGA n or Continuous Overall Net Glycemic Action
Within Day (CONGA and CONGA n) or Within Day
Variation (WDV) Variation (WDV) of the standard deviation of
the difference between two time points
(e.g. work, school, meals, other)
TIR Time In Range 4 to 12 mmol/l 7‐180 mgm%
TBR Time Below Range < 4 mmol/l < 70 mgm%
(TAR) Time Above Range > 12 mmol/l >
180 mgm%
 Is HbA1c the Most Important Therapeutic Target in the Outpatient Management of Diabetes? 109

with interpretation and identification of corrective action
in decision making. For example: Case Study 1a is an
example of variability due to frequent reactive across day
insulin dosing (ADV) while Case Study 1b demonstrates
the within day variability (WDV) associated with elevated
cortisol levels from Cushing’s Syndrome. Variability Grid
Analysis (VGA) or Poincare Plot recognizes the impor-
tance of assessing quantitative and qualitative patterns to
assist the clinician and patient in management decisions.
While glucose variation has been stated to have two com-
ponents, amplitude and timing, that are predictive of hyper
and hypoglycemia, these parameters have been mostly
directed towards hyperglycemia and have not been used
extensively in the clinical management of diabetes [65]. Mean
of Daily Difference (MODD) or Across Day Variation
(MODD or ADV) and Continuous Overlapping Net Glycemic
Action (CONGAn) or Within Day Variation (CONGn or
WDV) can be used not only as a continued effort in the quan-
titative assessment of within day and across day variation, but
they graphically offer the opportunity for qualitative assess-
ment. Most importantly, they can serve as clinical tools for
review and discussion with individual patients and their man-
agement behavior. For example, Figure 8.1a is a patient that
dominantly uses sliding scale versus preventive insulin dose
adjustment. Figure 8.1b demonstrates a pattern of significant
postprandial hyperglycemia (WDV) related to steroids.

FIG 8.1A  Glycemic variation with frequent insulin supplements. 40‐year‐old woman with Type 1 diabetes on MDI. She is measuring
pre and postprandial glucose levels and notes significant variation at the same time each day despite similar diet and activity. Her
reactive supplemental insulin should be considered as a cause of her Across Day Variation (ADV).
 110 Initial Evaluation and Management of Diabetes
400
350
300
250
mg/dL
200
150
100
50
12:00 3:00 6:00 9:00
FIG 8.1B  Glycemic variation due to within day variation (WDV) from steroids.
A 30‐year‐old male with type 2 diabetes on a single
long‐acting daily insulin. He is undergoing an evaluation
for the question of Cushing’s Syndrome from an adrenal
adenoma. He is observed to have within day variation
(WDV) that includes a progressive rise in postprandial
reflectance meter glucoses with highest values at the end of
the day and progressive fall nocturnally that places him at
risk for early morning hypoglycemia.
Recognizing that there are many causes of glycemic
variation to include people and medication factors, the
Beyond A1c Working Group has offered an attempt to
quantitate a more predictive measure of glycemic control
they call the Comprehensive Glucose Pentagon and the
glycemic risk parameter (GRP). The GRP reflects the five
main axes of glucose control: time out of range, coefficient
of variation, severity of hyperglycemia, severity of hypogly-
cemia, and mean glucose [67]. HbA1c is not considered as
part of the main axes of their model, and validation will
be necessary to determine whether it has a prognostic
value greater than HbA1c. Currently the authors provide
three case examples to describe its use. Sufficiently pow-
ered long‐term study will be necessary in which assess-
ment of these measures maybe be possible due to the
wide range of outcomes of information collected from
CGMs and pattern analysis.
12:00 3:00 6:00 9:00 12:00
Summary
The evidence for HbA1c as a measure of long‐term glyce-
mic control and development of vascular outcomes is
robust, and it remains a valuable metric in the manage-
ment of both DM1 and DM2. However, as diabetes tech-
nology has advanced with development of CGM and
closed‐loop systems of insulin delivery, the limitations of
HbA1c have become even more pronounced. CGM has
allowed for a real‐time view into actionable metrics that
permit clinicians, patients, and researchers to tailor man-
agement toward specific goals based on patient‐specific
factors. This focused approach using metrics such as TIR,
glucose variability, and magnitude of excursions has led
to reductions in hyperglycemia while minimizing risks of
hypoglycemia, both cornerstones of glycemic control.
Preliminary studies have demonstrated the role for met-
rics such as variability and TIR in reducing risk of at least
microvascular complications, and future studies are
needed to further identify their role in other diabetes‐
related outcomes such as macrovascular disease. Thus,
standardized CGM metrics are appropriate complements
to the traditional measure of HbA1c and provide valuable
tools for both patients and clinicians in the day‐to‐day
management of diabetes.
 Is HbA1c the Most Important Therapeutic Target in the Outpatient Management of Diabetes?
References
1. Diabetes C, Complications Trial Research G, Nathan DM,
Genuth S, Lachin J, Cleary P, Crofford O, Davis M, Rand L,
Siebert C. The effect of intensive treatment of diabetes on the
development and progression of long‐term complications in
insulin‐dependent diabetes mellitus. N Engl J Med.
1993;329(14):977–986. Epub 1993/09/30. doi: 10.1056/
NEJM199309303291401. PubMed PMID: 8366922.
2. Ohkubo Y, Kishikawa H, Araki E, Miyata T, Isami S,
Motoyoshi S, Kojima Y, Furuyoshi N, Shichiri M. Intensive
insulin therapy prevents the progression of diabetic
microvascular complications in Japanese patients with non‐
insulin‐dependent diabetes‐mellitus – a randomized
prospective 6‐year study. Diabetes Res Clin Pr.
1995;28(2):103–117. doi: 10.1016/0168‐8227(95)01064‐K.
PubMed PMID: WOS:A1995RR63400004.
3. Mann JFE, Orsted DD, Brown‐Frandsen K, Marso SP,
Poulter NR, Rasmussen S, Tornoe K, Zinman B, Buse JB,
LEADER Steering Committee and Investigators. Liraglutide
and renal outcomes in type 2 diabetes. N Engl J Med.
377(9):839–848. PubMed PMID: 28854085.
4. Marso SP, Daniels GH, Brown‐Frandsen K, Kristensen P, Mann
JF, Nauck MA, Nissen SE, Pocock S, Poulter NR, Ravn LS,
Steinberg WM, Stockner M, Zinman B, Bergenstal RM, Buse JB,
LEADER Steering Committee, LEADER Trial Investigators.
Liraglutide and Cardiovascular Outcomes in Type 2 Diabetes.
N Engl J Med. 375(4):311–322. PubMed PMID: 27295427.
5. Tamborlane WV, Barrientos‐Perez M, Fainberg U, Frimer‐
Larsen H, Hafez M, Hale PM, Jalaludin MY, Kovarenko M,
Libman I, Lynch JL, Rao P, Shehadeh N, Turan S, Weghuber
D, Barrett T, Ellipse Trial Investigators. Liraglutide in
children and adolescents with type 2 diabetes. N Engl J Med.
381(7):637–646. PubMed PMID: 31034184.
6. Verma S, Bhatt DL, Bain SC, Buse JB, Mann JFE, Marso SP,
Nauck MA, Poulter NR, Pratley RE, Zinman B, Michelsen
MM, Monk Fries T, Rasmussen S, Leiter LA, LEADER
Publication Committee on behalf of the LEADER Trial
Investigators. Effect of liraglutide on cardiovascular events in
patients with type 2 diabetes mellitus and polyvascular
disease: results of the LEADER trial. Circulation.
137(20):2179–2183. PubMed PMID: 29760228.
7. Zinman B, Wanner C, Lachin JM, Fitchett D, Bluhmki E,
Hantel S, Mattheus M, Devins T, Johansen OE, Woerle HJ,
Broedl UC, Inzucchi SE for the EPMPA‐REG OUTCOME
Investigators. Empagliflozin, cardiovascular outcomes, and
mortality in type 2 diabetes. N Engl J Med. 2015;Nov
26;373(22):2117–2128
111
8. Zinman B, Marso SP, Christiansen E, Calanna S, Rasmussen
S, Buse JB, Investigators LEADER Publication Committee on
behalf of the LEADER Trial Investigators. Hypoglycemia,
cardiovascular outcomes, and death: the LEADER
experience. Diabetes Care. 41(8):1783–1791. PubMed PMID:
29903847.
9. Sheetz MJ, King GL. Molecular understanding of hyperglyce-
mia’s adverse effects for diabetic complications. JAMA.
2002;288(20):2579–2588. Epub 2002/11/28. doi: 10.1001/
jama.288.20.2579. PubMed PMID: 12444865.
10. Cohen R. Dysfunction of vascular endothelium in diabetes
mellitus. Circulation. 1993;87(Suppl V):V67–V76.
11. Risso A, Mercuri F, Quagliaro L, Damante G, Ceriello A.
Intermittent high glucose enhances apoptosis in human
umbilical vein endothelial cells in culture. American Journal
of Physiology – Endocrinology & Metabolism. 281(5):E924–
930. PubMed PMID: 11595647.
12. Nishikawa T, Edelstein D, Du XL, Yamagishi S, Matsumura
T, Kaneda Y, Yorek MA, Beebe D, Oates PJ, Hammes HP,
Giardino I, Brownlee M. Normalizing mitochondrial super-
oxide production blocks three pathways of hyperglycaemic
damage. Nature. 404(6779):787–790. PubMed PMID:
10783895.
13. Piconi L, Quagliaro L, Assaloni R, Da Ros R, Maier A,
Zuodar G, Ceriello A. Constant and intermittent high glu-
cose enhances endothelial cell apoptosis through mitochon-
drial superoxide overproduction. Diabetes/Metabolism
Research Reviews. 22(3):198–203. PubMed PMID: 16453381.
14. Quagliaro L, Piconi L, Assaloni R, Martinelli L, Motz E,
Ceriello A. Intermittent high glucose enhances apoptosis
related to oxidative stress in human umbilical vein endothe-
lial cells: the role of protein kinase C and NAD(P)H‐oxidase
activation. Diabetes. 52(11):2795–2804. PubMed PMID:
14578299.
15. John WG, Mosca A, Weykamp C, Goodall I. HbA1c stand-
ardisation: history, science and politics. Clin Biochem Rev.
2007;28(4):163–168. Epub 2008/04/09. PubMed PMID:
18392123; PMCID. PMC2282401.
16. Rahbar S. An abnormal hemoglobin in red cells of diabetics.
Clin Chim Acta. 1968;22(2):296–298. Epub 1968/10/01.
PubMed PMID: 5687098.
17. Smith SA. HbA1c: Is it the most important therapeutic tar-
get in outpatient management of diabetes? In: Vella A and
Rizza RA (eds) Clinical Dilemmas in Diabetes. Blackwell
Publishing; 2011.
18. American Diabetes A. Standards of medical care in diabetes –
2008. Diabetes Care. 2008;31(Suppl 1):S12–54. Epub 2008/
01/10. doi: 10.2337/dc08‐S012. PubMed PMID: 18165335.
 112 Initial Evaluation and Management of Diabetes
19. Trivelli LA, Ranney HM, Lai HT. Hemoglobin components
in patients with diabetes mellitus. N Engl J Med.
1971;284(7):353–357. Epub 1971/02/18. doi: 10.1056/
NEJM197102182840703. PubMed PMID: 5539916.
20. Diabetes Control and Complications Trial/Epidemiology of
Diabetes Interventions and Complications Research Group,
Lachin JM, Genuth S, Cleary P, Davis MD, Nathan DM.
Retinopathy and nephropathy in patients with type 1 diabe-
tes four years after a trial of intensive therapy. N Engl J Med.
2000;342(6):381–389. Epub 2000/02/10. doi: 10.1056/
NEJM200002103420603. PubMed PMID: 10666428;
PMCID. PMC2630213.
21. Effect of intensive blood‐glucose control with metformin on
complications in overweight patients with type 2 diabetes
(UKPDS 34). UK Prospective Diabetes Study (UKPDS)
Group. Lancet. 1998;352(9131):854–865. Epub 1998/09/22.
PubMed PMID: 9742977.
22. Intensive blood‐glucose control with sulphonylureas or
insulin compared with conventional treatment and risk of
complications in patients with type 2 diabetes (UKPDS 33).
UK Prospective Diabetes Study (UKPDS) Group. Lancet.
1998;352(9131):837–853. Epub 1998/09/22. PubMed PMID:
9742976.
23. Reichard P. Are there any glycemic thresholds for the serious
microvascular diabetic complications. J Diabetes Complicat.
1995;9(1):25–30. doi: 10.1016/1056‐8727(94)00008‐C.
PubMed PMID: WOS:A1995RK84900004.
24. International Expert Committee. International Expert
Committee report on the role of the A1C assay in the diag-
nosis of diabetes. Diabetes Care. 2009;32(7):1327–1334.
Epub 2009/06/09. doi: 10.2337/dc09–9033. PubMed PMID:
19502545; PMCID. PMC2699715.
25. Patel A, MacMahon S, Chalmers J, Neal B, Billot L,
Woodward M, Marre M, Cooper M, Glasziou P, Grobbee D,
Hamet P, Harrap S, Heller S, Liu LS, Mancia G, Mogensen
CE, Pan CY, Poulter N, Rodgers A, Williams B, Bompoint S,
de Galan BE, Joshi R, Travert F, Grp AC. Intensive blood glu-
cose control and vascular outcomes in patients with type 2
diabetes. New Eng J Med. 2008;358(24):2560–2572. PubMed
PMID: WOS:000256593600004.
26. Ismail‐Beigi F, Craven T, Banerji MA, Basile J, Calles J,
Cohen RM, Cuddihy R, Cushman WC, Genuth S, Grimm
RH, Hamilton BP, Hoogwerf B, Karl D, Katz L, Krikorian A,
O’Connor P, Pop‐Busui R, Schubart U, Simmons D, Taylor
H, Thomas A, Weiss D, Hramiak I, for the ACCORD trial
group. Effect of intensive treatment of hyperglycaemia on
microvascular outcomes in type 2 diabetes: an analysis of the
ACCORD randomised trial. Lancet. 2010;376(9739):419–
430. doi: 10.1016/S0140‐6736(10)60576–60574. PubMed
PMID: WOS:000281012900027.
27. Duckworth W, Abraira C, Moritz T, Reda D, Emanuele N,
Reaven PD, Zieve FJ, Marks J, Davis SN, Hayward R, Warren
SR, Goldman S, McCarren M, Vitek ME, Henderson WG,
Huang GD, VADT Investigators. Glucose control and vascu-
lar complications in veterans with type 2 diabetes. New Eng J
Med. 2009;360(2):129–39. doi: 10.1056/NEJMoa0808431.
PubMed PMID: WOS:000262258300006.
28. Sacks DB, Bruns DE, Goldstein DE, Maclaren NK,
McDonald JM, Parrott M. Guidelines and recommendations
for laboratory analysis in the diagnosis and management of
diabetes mellitus. Clin Chem. 2002;48(3):436–472. Epub
2002/02/28. PubMed PMID: 11861436.
29. Beck RW, Connor CG, Mullen DM, Wesley DM, Bergenstal
RM. The fallacy of average: how using HbA1c alone to assess
glycemic control can be misleading. Diabetes Care.
2017;40(8):994–999. Epub 2017/07/25. doi: 10.2337/
dc17‐0636. PubMed PMID: 28733374; PMCID.
PMC5521971.
30. Nathan DM, Singer DE, Hurxthal K, Goodson JD. The
clinical information value of the glycosylated hemoglobin
assay. N Engl J Med. 1984;310(6):341–346. Epub 1984/
02/09. doi: 10.1056/NEJM198402093100602. PubMed PMID:
6690962.
31. Goldstein DE, Little RR, Lorenz RA, Malone JI, Nathan D,
Peterson CM. Tests of glycemia in diabetes. Diabetes Care.
1995;18(6):896–909. Epub 1995/06/01. doi: 10.2337/diac-
are.18.6.896. PubMed PMID: 7555528.
32. Kishimoto M, Yamasaki Y, Kubota M, Arai K, Morishima T,
Kawamori R, Kamada T. 1,5‐Anhydro‐D‐glucitol evaluates
daily glycemic excursions in well‐controlled NIDDM.
Diabetes Care. 1995;18(8):1156–1159. Epub 1995/08/01. doi:
10.2337/diacare.18.8.1156. PubMed PMID: 7587851.
33. Judd SL, Sonksen PH. Teaching diabetics about self‐man-
agement. Diabetes Care. 1980;3(1):134–139.
34. Sonksen PH, Judd SL, Lowy C. Home monitoring of blood‐
glucose. Method for improving diabetic control. Lancet.
1978;1(8067):729–732. PubMed PMID: 76745.
35. American Diabetes A. 6. Glycemic targets: standards of medi-
cal care in diabetes – 2019. Diabetes Care. 2019;42(Suppl 1):
S61–S70. Epub 2018/12/19. doi: 10.2337/dc19‐S006. PubMed
PMID: 30559232.
36. Beck RW, Bergenstal RM, Riddlesworth TD, Kollman C, Li Z,
Brown AS, Close KL. Validation of time in range as an out-
come measure for diabetes clinical trials. Diabetes Care.
2019;42(3):400–405. Epub 2018/10/26. doi: 10.2337/dc18‐
1444. PubMed PMID: 30352896.
 Is HbA1c the Most Important Therapeutic Target in the Outpatient Management of Diabetes?
37. Miller KM, Beck RW, Bergenstal RM, Goland RS, Haller MJ,
McGill JB, Rodriguez H, Simmons JH, Hirsch IB, T1D
Exchange Clinic Network. Evidence of a strong association
between frequency of self‐monitoring of blood glucose and
hemoglobin A(1c) levels in T1D Exchange Clinic Registry
participants. Diabetes Care. 2013;36(7):2009–2014. doi:
10.2337/dc12‐1770. PubMed PMID: WOS:000321472700034.
38. Schutt M, Kern W, Krause U, Busch P, Dapp A, Grziwotz R,
Mayer I, Rosenbauer J, Wagner C, Zimmermann A, Kerner
W, Holl RW, DPV Initiative. Is the frequency of self‐moni-
toring of blood glucose related to long‐term metabolic con-
trol? Multicenter analysis including 24,500 patients from
191 centers in Germany and Austria. Exp Clin Endocrinol
Diabetes. 2006;114(7):384–388. Epub 2006/08/18. doi: 10.
1055/s‐2006‐924152. PubMed PMID: 16915542.
39. Brownlee M. The pathobiology of diabetic complications: a
unifying mechanism. Diabetes. 54(6):1615–1625. PubMed
PMID: 15919781.
40. Hirsch IB, Brownlee M. Should minimal blood glucose vari-
ability become the gold standard of glycemic control?
Journal of Diabetes & its Complications. 19(3):178–181. Pub
Med PMID: 15866065.
41. Ceriello A. Postprandial hyperglycemia and diabetes com-
plications: is it time to treat? Diabetes. 54(1):1–7. PubMed
PMID: 15616004.
42. Ceriello A. Point: postprandial glucose levels are a clinically
important treatment target. Diabetes Care. 2010;33(8):
1905–1907.
43. Meier JJ, Baller B, Menge BA, Gallwitz B, Schmidt WE,
Nauck MA. Excess glycaemic excursions after an oral glu-
cose tolerance test compared with a mixed meal challenge
and self‐measured home glucose profiles: is the OGTT a
valid predictor of postprandial hyperglycaemia and vice
versa? Diabetes, Obesity & Metabolism. 11(3):213–222. Pub
Med PMID: 18564177.
44. NAVIGATOR Study Group, Holman RR, Haffner SM,
McMurray JJ, Bethel MA, Holzhauer B, Hua TA, Belenkov Y,
Boolell M, Buse JB, Buckley BM, Chacra AR, Chiang FT,
Charbonnel B, Chow CC, Davies MJ, Deedwania P, Diem P,
Einhorn D, Fonseca V, Fulcher GR, Gaciong Z, Gaztambide
S, Giles T, Horton E, Ilkova H, Jenssen T, Kahn SE, Krum H,
Laakso M, Leiter LA, Levitt NS, Mareev V, Martinez F,
Masson C, Mazzone T, Meaney E, Nesto R, Pan C, Prager R,
Raptis SA, Rutten GE, Sandstroem H, Schaper F, Scheen A,
Schmitz O, Sinay I, Soska V, Stender S, Tamas G, Tognoni G,
Tuomilehto J, Villamil AS, Vozar J, Califf RM. Effect of nat-
eglinide on the incidence of diabetes and cardiovascular
events. New Eng J Med. 362(16):1463–1476
113
45. Chiasson JL, Gomis R, Hanefeld M, Josse RG, Karasik A,
Laakso M. The STOP‐NIDDM Trial: an international study
on the efficacy of an alpha‐glucosidase inhibitor to prevent
type 2 diabetes in a population with impaired glucose toler-
ance: rationale, design, and preliminary screening data.
Study to prevent non‐insulin‐dependent diabetes mellitus.
Diabetes Care. 21(10):1720–1725. PubMed PMID: 9773737.
46. Cavalot F, Pagliarino A, Valle M, Di Martino L, Bonomo K,
Massucco P, Anfossi G, Trovati M. Postprandial blood glu-
cose predicts cardiovascular events and all‐cause mortality
in type 2 diabetes in a 14‐year follow‐up: lessons from the
San Luigi Gonzaga Diabetes Study. Diabetes Care.
34(10):2237–2243. PubMed PMID: 21949221.
47. Hopper I, Billah B, Skiba M, Krum H. Prevention of diabetes
and reduction in major cardiovascular events in studies of
subjects with prediabetes: meta‐analysis of randomised con-
trolled clinical trials. European Journal of Cardiovascular
Prevention & Rehabilitation. 18(6):813–823. PubMed PMID:
21878448.
48. Davidson MB. Counterpoint: postprandial glucose levels are
not a clinically important treatment target. Diabetes Care.
2010;33:1908–1910.
49. Liebl A, Henrichs HR, Heinemann L, Freckmann G,
Biermann E, Thomas A, Continuous Glucose Monitoring
Working Group of the Working Group Diabetes Technology
of the German Diabetes A. Continuous glucose monitoring:
evidence and consensus statement for clinical use. J Diabetes
Sci Technol. 2013;7(2):500–519. Epub 2013/04/10. doi:
10.1177/193229681300700227. PubMed PMID: 23567009;
PMCID. PMC3737652.
50. Beck RW, Riddlesworth T, Ruedy K, Ahmann A, Bergenstal
R, Haller S, Kollman C, Kruger D, McGill JB, Polonsky W,
Toschi E, Wolpert H, Price D, DIAMOND Study Group.
Effect of continuous glucose monitoring on glycemic control
in adults with type 1 diabetes using insulin injections. The
DIAMOND Randomized Clinical Trial. JAMA.
2017;317(4):371–378. doi: 10.1001/jama.2016.19975.
PubMed PMID: WOS:000392509600019.
51. Battelino T, Conget I, Olsen B, Schutz‐Fuhrmann I, Hommel
E, Hoogma R, Schierloh U, Sulli N, Bolinder J, SWITCH
Study Group. The use and efficacy of continuous glucose
monitoring in type 1 diabetes treated with insulin pump
therapy: a randomised controlled trial. Diabetologia.
2012;55(12):3155–3162. doi: 10.1007/s00125‐012‐2708‐9.
PubMed PMID: WOS:000310381800003.
52. Juvenile Diabetes Research Foundation Continuous Glucose
Monitoring Study G, Beck RW, Hirsch IB, Laffel L,
Tamborlane WV, Bode BW, Buckingham B, Chase P,
 114 Initial Evaluation and Management of Diabetes
Clemons R, Fiallo‐Scharer R, Fox LA, Gilliam LK, Huang
ES, Kollman C, Kowalski AJ, Lawrence JM, Lee J, Mauras N,
O’Grady M, Ruedy KJ, Tansey M, Tsalikian E, Weinzimer
SA, Wilson DM, Wolpert H, Wysocki T, Xing D. The effect
of continuous glucose monitoring in well‐controlled type 1
diabetes. Diabetes Care. 2009;32(8):1378–1383. Epub 2009/
05/12. doi: 10.2337/dc09‐0108. PubMed PMID: 19429875;
PMCID. PMC2713649.
53. Bergenstal RM, Tamborlane WV, Ahmann A, Buse JB,
Dailey G, Davis SN, Joyce C, Peoples T, Perkins BA, Welsh
JB, Willi SM, Wood MA, STAR3 Study Group. Effectiveness
of sensor‐augmented insulin‐pump therapy in type 1 diabe-
tes. N Engl J Med. 2010;363(4):311–320. Epub 2010/07/01.
doi: 10.1056/NEJMoa1002853. PubMed PMID: 20587585.
54. Bergenstal RM, Klonoff DC, Garg SK, Bode BW, Meredith
M, Slover RH, Ahmann AJ, Welsh JB, Lee SW, Kaufman FR,
ASPIRE In‐Home Study Group. Threshold‐based insulin‐
pump interruption for reduction of hypoglycemia. New Eng
J Med. 2013;369(3):224–232. doi: 10.1056/NEJMoa1303576.
PubMed PMID: WOS:000321945800008.
55. Danne T, Kordonouri O, Holder M, Haberland H,
Golembowski S, Remus K, Blasig S, Wadien T, Zierow S,
Hartmann R, Thomas A. Prevention of hypoglycemia by
using low glucose suspend function in sensor‐augmented
pump therapy. Diabetes Technol Ther. 2011;13(11):1129–1134.
Epub 2011/08/11. doi: 10.1089/dia.2011.0084. PubMed PMID:
21827318.
56. Danne T, Nimri R, Battelino T, Bergenstal RM, Close KL,
DeVries JH, Garg S, Heinemann L, Hirsch I, Amiel SA, Beck
R, Bosi E, Buckingham B, Cobelli C, Dassau E, Doyle FJ, 3rd,
Heller S, Hovorka R, Jia W, Jones T, Kordonouri O, Kovatchev
B, Kowalski A, Laffel L, Maahs D, Murphy HR, Norgaard K,
Parkin CG, Renard E, Saboo B, Scharf M, Tamborlane WV,
Weinzimer SA, Phillip M. International consensus on use of
continuous glucose monitoring. Diabetes Care. 2017;40(12):
1631–1640. Epub 2017/11/23. doi: 10.2337/dc17‐1600. Pub
Med PMID: 29162583; PMCID. PMC6467165.
57. Vigersky RA. Going beyond HbA1c to understand the ben-
efits of advanced diabetes therapies. Journal of Diabetes. 2019;
11(1):23–31. doi: 10.1111/1753‐0407.12846. PubMed PMID:
WOS:000451868200005.
58. Tamborlane WV, Beck RW, Bode BW, Buckingham B, Chase
HP, Clemons R, Fiallo‐Scharer R, Fox LA, Gilliam LK,
Hirsch IB, Huang ES, Kollman C, Kowalski AJ, Laffel L,
Lawrence JM, Lee J, Mauras N, O’Grady M, Ruedy KJ,
Tansey M, Tsalikian E, Weinzimer S, Wilson DM, Wolpert
H, Wysocki T, Xing DY, Continuous JDRF. continuous glu-
cose monitoring and intensive treatment of type 1 diabetes.
New Eng J Med. 2008;359(14):1464–U65. PubMed PMID:
WOS:000259631700007.
59. Xing D, Kollman C, Beck RW, Tamborlane WV, Laffel L,
Buckingham BA, Wilson DM, Weinzimer S, Fiallo‐Scharer
R, Ruedy KJ, Juvenile Diabetes Research Foundation
Continuous Glucose Monitoring Study Group. Optimal
sampling intervals to assess long‐term glycemic control
using continuous glucose monitoring. Diabetes Technol
Ther. 2011;13(3):351–358. Epub 2011/02/09. doi: 10.1089/
dia.2010.0156. PubMed PMID: 21299401; PMCID.
PMC6468940.
60. Riddlesworth TD, Beck RW, Gal RL, Connor CG, Bergenstal
RM, Lee S, Willi SM. Optimal Sampling duration for continu-
ous glucose monitoring to determine long‐term glycemic con-
trol. Diabetes Technol Ther. 2018;20(4):314–316. Epub 2018/
03/23. doi: 10.1089/dia.2017.0455. PubMed PMID: 29565197.
61. Lu J, Ma X, Zhou J, Zhang L, Mo Y, Ying L, Lu W, Zhu W, Bao
Y, Vigersky RA, Jia W. Association of time in range, as
assessed by continuous glucose monitoring, with diabetic
retinopathy in type 2 diabetes. Diabetes Care. 2018;41(11):
2370–2376. Epub 2018/09/12. doi: 10.2337/dc18‐1131. Pub
Med PMID: 30201847.
62. Faerch K, Alssema M, Mela DJ, Borg R, Vistisen D. Relative
contributions of preprandial and postprandial glucose expo-
sures, glycemic variability, and non‐glycemic factors to
HbA1c in individuals with and without diabetes. Nutrition &
Diabetes. 2018;8(1):38. PubMed PMID: 29855488.
63. Siegelaar SE, Kerr L, Jacober SJ, Devries JH. A decrease in
glucose variability does not reduce cardiovascular event
rates in type 2 diabetic patients after acute myocardial
infarction: a reanalysis of the HEART2D study. Diabetes
Care. 2011;34(4):855–857. PubMed PMID: 21447661.
64. McDonnell CM, Donath SM, Vidmar SI, Werther GA,
Cameron FJ. A novel approach to continuous glucose analy-
sis utilizing glycemic variation. Diabetes Technology &
Therapeutics. 2005;7(2):253–263. PubMed PMID: 15857227.
65. Cobelli C, Facchinetti A. Yet another glucose variability
index: time for a paradigm change? Diabetes Technology &
Therapeutics. 2018;20(1):1–3. PubMed PMID: 29320256.
66. Zhou T, Knopp JL, Chase G. The state of variability: a vision
for descriptors of glycaemia. Annual Reviews in Control.
2019;48:472–484. doi.org/10.1016/j.arcontrol.2019.06.004.
67. Thomas A, Heinemann L. Prediction of the risk to develop
diabetes‐related late complications by means of the glucose
pentagon model: analysis of data from the Juvenile Diabetes
Research Foundation continuous glucose monitoring study.
Journal of Diabetes Science & Technology. 2012;6(3):572–580.
PubMed PMID: 22768888.
 9 Technology Issues: Continuous Glucose
Monitoring, Insulin Pumps, and Closed
Loop Control for Patients with Diabetes
Ravinder Jeet Kaur, Shafaq R. Rizvi and Yogish C. Kudva
Division of Endocrinology, Diabetes, Metabolism and Nutrition, Mayo Clinic, Rochester, MN
LE A R NI NG P OI NT S
●● There is enough evidence to indicate that optimal
management of type 1 diabetes requires a personal
continuous glucose monitor.
●● The best current treatment for patients with type 1
diabetes and severe insulin deficiency is a closed loop
glucose insulin control (CLC) system.
●● Currently, the closed loop system has only been tested
and approved in patients with severe or stage III type 1
diabetes.
●● So far CLC has not been tested in stage I and II type
1diabetes.
●● CLC is being tested in specific subgroups and contexts
such as pregnancy with type 1 diabetes at present.
●● CLC is expensive and may only be available in certain
countries at this time.
●● Technology use for type 2 diabetes mellitus has to be
studied further.
Introduction
Type 1 diabetes mellitus (T1DM) is a disease characterized
by immune-­ mediated destruction of insulin producing
pancreatic β-­cells leading to insulin deficiency and hyper-
glycemia [1, 2], thus requiring lifelong insulin therapy to
achieve glucose control. Despite complex insulin therapy,
T1DM is characterized by high glucose variability (GV).
GV occurs due to severe deficiency in insulin secretion and
the difficulty of replacing insulin without frequent glucose
measurements that enable prompt change in insulin
Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
d­elivery. Based on deficiency in insulin secretion, three
stages occur in the natural history of T1D, normal glucose
in stage 1, dysglycemia (hyperglycemia) diagnosed with
sophisticated laboratory testing in stage 2 and symptomatic
hyperglycemia in stage 3 [3, 4]. For stage 2 disease onwards,
insulin therapy can be delivered by multiple daily injec-
tions (MDI), continuous subcutaneous insulin infusion
(CSII) or insulin pump with insulin dosing adjusted fre-
quently based on glucose patterns. T1DM patients have
used MDI (basal-­bolus) for more than 40 years to ade-
quately maintain their blood glucose (BG) guided by self-­
monitoring of blood glucose (SMBG) four or more times
daily (before meals and at bedtime) using capillary blood
sampling. The association of such management with fre-
quent hypoglycemia, hypoglycemia unawareness and
severe hypoglycemia accelerated the development of con-
tinuous glucose monitors (CGM). Insulin pumps to deliver
insulin continuously were developed in the 1970s. The
maturation of these device technologies and the develop-
ment of control algorithms that could automatically adjust
insulin delivery have led to the testing and approval of
Closed loop control (CLC) in the last decade [5]. Recently,
international committees have developed consensus about
managing diabetes to achieve maximal time in target (TIR)
on CGM. The consensus opinion is that CGM is used to
guide optimization of therapy to increase TIR (70–180mg/
dL) and minimize hypoglycemia (< 70mg/dL) and hyper-
glycemia (> 250mg/dL) in patients with T1DM.
115
 116 Initial Evaluation and Management of Diabetes
Type 2 diabetes mellitus (T2DM) is a disease with differ-
ent pathophysiology and therefore the role of technology in
the management of the disease is different. Also, technol-
ogy testing for its role in the treatment of the disease is at
an early stage and therefore the amount of information
provided in this chapter regarding the role of technology in
the treatment of T2DM is limited.
We restricted our content to ambulatory care and have
not discussed inpatient use of technologies for diabetes
management.
Continuous Glucose Monitoring (CGM)
for Diabetes Care
Currently, CGM provides real-­time data every 5–15 min-
utes depending on the device. CGM allows for real-­time
assessment of glucose concentrations and trends by meas-
uring interstitial glucose concentrations. Interstitial glu-
cose concentrations are usually close to blood glucose
concentrations when glucose concentrations are stable and
lag by just a few minutes, but this difference is greater dur-
ing rapid changes in blood glucose such as after food intake
or during exercise. Current CGM systems may provide
data on demand (Libre) or continuously (all other sys-
tems). These systems are placed subcutaneously and
replaced every 7–14  days or may be implanted for
3–6  months (Eversense). The subcutaneous systems may
be factory-­calibrated or calibrated daily.
CGM data can be used to prevent and protect against
extremes of glucose control and improve glucose control
either with MDI or CSII or with a CLC system. The alerts
associated with hypo-­and hyperglycemia prevent and
decrease severe hypo-­and hyperglycemia.
Impact of CGM in Patients on MDI or CSII
The impact of CGM on diabetes management has been
observed in both T1D and T2D and in select populations
with diabetes. Real-­time CGM has been available since
2005 and the major CGM systems were used in a multiple-­
center randomized control trail (RCT) sponsored by the
JDRF, called the JDRF CGM trial and published in 2008 [6].
This trial showed reduction in HbA1c and hypoglycemia
in adults (≥25 years of age) from baseline to 26 weeks of the
study with the benefit strongly related to age (Table 9.1) [6].
At 26-­weeks of study, reduction in HbA1c with HbA1c <
7.0%, occurred in a significantly larger proportion in the
younger group (8–14 yr, 27 vs 12%) and adults (34 vs 9%).
Another RCT conducted in the USA with 158 T1D sub-
jects on MDI and HbA1c 7.5–9.9% randomized subjects
2:1 to CGM (Dexcom G4, n=105) and usual care (controls,
53) [7]. During 24 weeks of the study, a greater decrease in
HbA1c level was seen with increased TIR, decreased time
below 70mg/dL decreased glucose variability and high
degree of participant satisfaction for CGM use (Table 9.1).
A similar study performed in 161 adults with T1D using
MDI with HbA1c of at least 7.5% in Sweden also showed
reduction in HbA1c (7.92% vs 8.35%) with CGM use
(Dexcom G4 PLATINUM) compared to usual care [8].
Impact in a Higher Risk Population: Patients
with T1D and Impaired Hypoglycemia
Awareness
In 2018, another RCT (n=149), was conducted using CGM
(Libre, n=75) in T1D adults with impaired hypoglycemia
awareness or severe hypoglycemia treated with MDI
(HypoDE) to observe the effect on clinical hypoglycemic
events (glucose ≤3.0 mmol/L for ≥20 min) compared to the
control group (n=74)  [9]. Significant reductions in mean
hypoglycemic events (glucose values of 3.0  mmol/l
(≤54  mg/dL) or lower for at least 20  min, preceded by a
minimum of 30  min with glucose concentration greater
than 3.0 mmol/l (> 54 mg/dL)) were observed in rtCGM
group (10.8 vs 3.5 events from baseline to follow-­up phase)
compared with SMBG group (14.4 vs 13.7 events).
Additionally, rtCGM use decreased glycemic variability
with slight improvement in HbA1c (Table 9.1).
Efficacy of CGM in the Elderly: The WISDM
Study
Most recently, a RCT was performed on 203 older adults at
least 60 years of age with T1DM  [10]. Participants were
randomly assigned in a 1:1 ratio to use CGM (103) and
standard BGM (100). Results from this study showed a
small but significant improvement in hypoglycemia (<
70mg/dL) over 6  months in the CGM group (Table  9.1)
along with decrease in mean HbA1c (7.6% at baseline to
7.2% at 6 months). There was also significant improvement
in TIR (56% to 63%) along with reduction in time in hyper-
glycemia with CGM use.
 Closed Loop Systems in Type 1 Diabetes 117

TABLE 9.1  Continuous glucose monitoring (CGM) impact in diabetes (randomized control trials)

Hypoglycemia
Number (≤70mg/dL)
of Insulin (control vs
Studies subjects delivery CGM Groups Control CGM group) HbA1C

JDRF [6] 322 Pump ad MDI Dexcom 8–14 yr 8–14 yr Decreased (9% Decreased
users Seven, 15–24 yr 15–24 yr vs 34%) (age (0.02±0.45% vs
(Predominately Medtronic ≥25 yr ≥25 yr group ≥25 yr) −0.50±0.56%)
pump) Paradigm, (CGM (SMBG users (age group ≥25
Abbott users=165) =157) yr)
FreeStyle
Navigator
WISDM [10] 203 Pump and MDI Dexcom CGM BGM Decreased Decreased
G5 (N=103) (N=100) (BGM: 4.9% vs (BGM: 7.4% vs
CGM –2.7%) CGM – 7.2%)
Baseline: Baseline:
5.1% – CGM, 7.65 – CGM,
4.7% – BGM 7.5% – BGM
HypoDE [9] 149 100%MDI Dexcom rtCGM SMBG Decreased Improved
G5 mobile (N=75) (N=74) (6.4% vs 1.6%) Slightly but not
system (N=66 in significant
follow up (from 7.4 to
phase) 7.3% (rtCGM)
vs from 7.6 to
7.4%(SMBG))
DIAMOND [7] 158 100% MDI Dexcom CGM users Usual care with Decreased Decreased
G4 with with MDI MDI (home (minutes per (8.2 vs 7.7%)
software blood glucose day < 70mg/
505 monitoring at dL – median of
least 4 times 80 vs 43)
daily – Bayer
Contour Next
USB meter)

BGM – Blood glucose monitoring, MDI – Multiple daily injections, CGM – Continuous glucose monitoring.

CGM in Pregnancy of 11.1% for reference samples > 4.2mmol/L (75mg/dL). As

CGM has been tested in special contexts such as pregnancy
with preexisting T1D (Medtronic iPRO2) and showed
reduction in HbA1c, improved TIR, less time in hypergly-
cemia, and improved neonatal outcomes [11].
Implantable CGM
A multinational, multicenter pivotal trial was performed
on 71 participants with T1DM and T2DM was conducted
with the Eversense CGM (implanted under the skin for
180 days) [12]. Results showed a median sensor lifetime of
149 days, overall mean absolute relative difference (MARD)
expected, performance in the hypoglycemic range (≤
75mg/dL) was less accurate, with a MARD of 21.7 % com-
pared to 11.6 % when glucose was 40–400mg/dL, and
HbA1c significantly improved from 7.5% to 7.2%. CGM
accuracy significantly decreased in the last month of use,
but overall showed safety and accurate performance over
the full sensor life. The implantable Eversense XL CGM
was approved in Europe (180-­day version) in 2017 and in
the USA (90-­day version) in 2018.
CGM has improved greatly over the past decade with
longer sensor life (7 days – 6 months), factory calibration
 118 Initial Evaluation and Management of Diabetes

(Dexcom G6 and Abbott FreeStyle Libre) or twice a day
calibration (Medtronic Guardian Sensor and Senseonics
Eversense), improved alarms and alerts, easy insertion and
ability to share glucose values in real time. This has
improved quality of life  [13] immensely in T1D patients
with a decrease in hypoglycemia (< 70mg/dL) [6, 7, 10] and
improvement in A1c [13]. Over the last decade, CGM use
increased from 7% (2010–2012) to 30% (2016–2018) as per
the T1D Exchange registry [14], especially in children < 12
years old (> 10 fold) [12].
The evidence base for CGM use in T2DM is significantly
less. Studies in T2DM using older CGM devices have been
published  [15]. Personal or intermittent clinic-­requested
RT CGM has been shown to improve A1c by 0.3 % (meta-­
analysis of 4 RCT) with benefits to diet, lifestyle, and ther-
apy adjustment [16]. Although the use of CGM in T2DM
would be intuitively beneficial, more investigation is
needed.
Currently Approved CLC for T1DM
The optimal current treatment for patients with T1DM and
severe insulin deficiency is a CLC system. So far CLC has
not been tested in clinical stages 1 and 2 of T1DM.
Currently, two hybrid CLC systems are approved for use in
the USA by the FDA. Both CLC were tested and approved
in patients with stage III T1DM and are in commercial use.
These systems have predictive low-­glucose insulin-­suspend
functionality and increase basal insulin delivery rates using
an algorithm in response to predicted hyperglycemia or
hyperglycemia signals received from real-­time CGM. They
are called hybrid because of patient engagement at meal-
time that is similar to SAP therapy or CSII.
MiniMed 670G
MiniMed 670G (Medtronic, Northridge, CA, USA) was the
first hybrid CLC system available in the USA after 3 months
of a single-­arm safety trial completed on 124 T1DM par-
ticipants [17]. Patients had to periodically calibrate CGM
in this study and dose mealtime insulin using SMBG. The
Control Algorithm utilized is the Proportional Integrated
Derivative (PID). Participants spent around 87.2% time in
closed loop mode. Patients improved A1c and every CGM
metric (Table 9.2). There were no episodes of severe hypo-
glycemia and DKA during this study. The lack of a control
group is a major limitation of the study. The approval of the
device was contingent on the conduct of an RCT compar-
ing hybrid-­closed loop to three different treatments: MDI,
CSII, and SAP. This trial is scheduled to be completed by
December 2021 (NCT02748018).
Most recently, a 6-­month International Diabetes Closed
loop (iDCL) RCT enrolled 168 T1DM patients at seven US

TABLE 9.2  Pivotal studies of closed loop system.

Baseline versus hybrid closed loop

Type of > 180 > 250 > 300

Studies study Geography T1R < 70mg/dL mg/dL mg/dL mg/dL HbA1c (%)

Medtronic Non RCT USA, Israel 66.7% vs 5.9% vs 3.3% 27.4% vs N/A 2.3% vs 7.4% vs
670G [17] (n=124) 72.2% 24.5% 1.7% 6.9%
Tandem RCT USA 61% vs 71% 3.6% vs 1.6% 36% vs N/A N/A 7.4% vs
Control (n=168) (HCL) 2.8% vs 2.2% 27% 7.1%
IQ [18] 59% vs 59% 38% vs 7.4% vs
(control) 38% 7.4%
FLAIR RCT USA, 57% vs 67% 2.3% vs 2.1% 41% vs 13% vs N/A 7.9% vs
Trail [19] (n=113) Germany, (AHCL) 2.3% vs 2.1% 31% 9% 7.4%
Israel, 575 vs 63% 41% vs 13% vs 7.9% vs
Slovenia (670G) 34% 10% 7.6%
Medtronic RCT USA 68.8% vs 3.3% vs 2.3% 27.9% vs 6.2% vs N/A 7.5% vs
780G [20] (n=157) 74.5% 23.1% 4.6% 7.0%

TIR – Time in Range (70–180mg/dL), glycated hemoglobin (HbA1c) %.


centers to compare safety and efficacy of an advanced
hybrid closed-­loop system (Control IQ, Tandem Diabetes
Care) SAP with both groups using Dexcom G6 CGM [18].
The algorithm used in this system was a Model Predictive
Control (MPC) with hypoglycemia safety module, auto-
mated correction boluses, and overnight intensification of
basal insulin delivery to keep glucose within normal lim-
its – the algorithm was constrained by a fixed active insulin
time. All CGM metrics improved in the CLC cohort
(Table  9.2). No serious hypoglycemia events occurred in
either group. One episode of diabetic ketoacidosis occurred
in the CLC group because of a pump infusion set failure.
FLAIR trial
Recently, the Medtronic MiniMedTM 780G Advanced
Hybrid Closed Loop (AHCL) automated insulin delivery
system was compared to 670G in adolescents and young
adults  [19]. This RCT was associated with better glucose
control and patient satisfaction. One hundred and thirteen
patients aged 14 to 29 years were randomized 1:1 to receive
either a commercially approved 670G system (n=57) or an
AHCL system (n=56) at seven sites in four countries with
12  weeks in each arm. CGM metrics improved in the
AHCL (Table 9.2). Patients without CGM or insulin pump
prior to study experienced the greatest improvement in
TIR. A single patient experienced severe hypoglycemia
(SH) and two patients experienced hyperglycemia or keto-
sis in the AHCL arm related to insulin pump use compared
to zero patients with SH, three had hyperglycemia or keto-
sis related to pump and one with hyperglycemia or ketosis
unrelated to insulin pump use in the 670G arm. Fewer sys-
tem alerts, more time spent in Auto Mode and reduced
Auto Mode exits occurred in the AHCL cohort.
780G pivotal trial
Medtronic Diabetes conducted a 90-­day at-­home trial with
the MiniMedTM 780G Advanced Hybrid Closed Loop
(AHCL) and showed exciting results with no severe hypo-
glycemia and DKA [20]. Participants stayed in closed loop
(SmartGuardTM) 95% of the time with improved CGM
metrics. This study showed average A1C of 7% with high
user satisfaction. This system features a default target of
100mg/dL with the option of 120mg/dL and two to eight
hours of programmable insulin action time (AIT). With
Closed Loop Systems in Type 1 Diabetes 119
100mg/dL target and 2–3 hours of AIT, TIR increased to
76% (74% daytime and 84% overnight) and even more
(79%) with 2hours of AIT (76% daytime and 87% over-
night). Data from more challenging patients including
those with less controlled diabetes and a younger popula-
tion from New Zealand also showed improved outcomes
compared to predictive low-­ glucose management algo-
rithm  [21]. There was overall improvement in TIR and
decrease in time below range with participants staying in
closed loop 95% of the time.
T:slim X2 pump users in Canada will have the option to
add Control-­IQ technology to their existing pump begin-
ning in March 2021 via a remote software update and the
780G has received approval in select European countries
Post-­marketing Studies
SmartGuardTM Auto-­mode enabled MiniMedTM
670G System
After it became available in the USA in 2017, CareLinkTM
system data that were uploaded by real-­world patients were
analyzed [22] and published retrospectively using data for
the first 3  months from 3141 670G users. These subjects
showed adherent satisfied patients with results sufficiently
close to the pivotal 670G study. Percentage time in manual
mode and subsequent auto mode in real world cohort were
TIR: 66 to 73%, < 70mg/dL=2.7 to 2.1%, > 180mg/dL =31.4
to 24.6% and > 250mg/dL =8.1 to 5.4%. These data confirm
the robustness of the system in subjects that uploaded to
the manufacturer’s website.
The t:slim X2TM Insulin Pump with Control IQ
Technology
After becoming available in the US in early 2020, real-­world
outcomes were reported in 1435 study participants at time
point 1 (at least 3 weeks after starting control-­IQ technol-
ogy) versus time point 2 (24 weeks from T1) [23]. The results
showed improved TIR over from T1 to T2 (76.7±11.6% vs
77.2±12.3%), reduction in time below 70mg/dL (1.3(T1) vs
1.2(T2) – not significant), reduction in time above 180mg/
dL (19.8 (T1) vs 19.0 (T2)) and time above 250mg/dL (2.9
(T1) vs 2.5 (T2)). Participants expressed high device-­related
satisfaction and reduced d­ iabetes impact. This system
 120 Initial Evaluation and Management of Diabetes
showed improvement in psychosocial outcomes and
improvement in TIR glycemic outcomes in people with
T1DM. Currently this system is approved in 18 countries
worldwide (United States, Australia, Canada, UK, Denmark,
Nederland, Czech Republic, France, Italy, New Zealand,
South Africa, Finland, Sweden, Norway, Spain, Belgium,
Germany, and Luxembourg).
Ongoing Pivotal Studies/Anticipated
Closed Loop System
HorizonTM Pivotal study (Omnipod HorizonTM)
Omnipod recently showed data on a hybrid, Omnipod®
5automated insulin delivery system. Two hundred and thirty-­
five participants have been enrolled in a single arm, multi-­
center, prospective clinical study. Study design includes a
14-­day outpatient, standard therapy phase, followed by 13-­
week (94-­day) HCL phase in an outpatient setting. After com-
pleting HCL phase, subjects have the option to continue for an
additional 6 months. During the HCL phase, a subset of sub-
jects will participate in 5-­days of supervised meal and exercise
challenges. Primary outcome measures are Incidence rate of
SH, DKA, A1c, and TIR (70–180mg/dl). The estimated study
completion date is December 2021.
Patch Pump Series: Diabeloop Generation 1
(DBLG1) Hybrid Closed Loop
A recent RCT done on 71 patients (65 completed) using
DBLG1 (Cellnovo Generation 1; Cellnovo, Paris, France)
hybrid CLC system (tubeless patch pump) or SAP over
12 weeks of free living [24], followed by 8-­week washout
period and then cross-­over to the other intervention for
12 weeks showed higher TIR (68.5%) in comparison with
SAP (59.4%). 2% time below 70mg/dL in DBLG1 is com-
pared to 4.3% in SAP group. There were no DKA noted in
each group but nine severe hyperglycemic events (capillary
blood glucose > 20mmol/L) in CLC and five SH events in
closed loop, compared to three SH events in SAP.
Closed Loop Control Initiation at Onset
of T1D
Hybrid CLC was attempted for 72 hours in one study
within 7  days of onset of T1D to test its effectiveness in
preservation of islet function at 1 year [25]. CLC was used
for 3 days in patients, followed by randomization to SAP vs
MDI. A1c achieved was 7.3 % and Area Under the Curve
(AUC) of Mixed Meal (MM) stimulated C-­peptide was did
not differ between groups. A larger, more definitive study is
currently being performed (see below).
Currently, a randomized parallel assignment RCT
(NCT02871089) is being conducted to assess the effect of CLC
from onset in T1D (CLOuD). The hypothesis is to improve
glucose control with the anticipated improvements in AUC
MM stimulated C-­peptide secretion. Participants will be ran-
domized to CLC (Florence M or CamAPS FX) or MDI, with
each treatment lasting 24 months. All participants that com-
plete the 24-­month study period will be invited to continue in
an optional extension phase with the treatment allocated at
randomization for a further 24 months. Estimated 24-­month
study completion date is June 2021.
CLC is Being Tested in Specific Subgroups
with Type 1 Diabetes at Present
T1D Pregnancy
Two CLC studies have been conducted in pregnancy
occurring in pre-­existing T1D in the UK. The first study
utilized CLC overnight improving with TIR 15 % higher
(74.7 vs 59.5%). The second study used CLC for 24 hours
and showed a non-­significant change in TIR (62.3 vs 60.1
%) [26, 27]. Currently, three CLC clinical trials are being
conducted in pregnancy with preexisting T1D:
OO NCT04492566: In this trial, CLC is being evaluated in 21
pregnant adult subjects at three US centers for a 48–60
hour closed loop session in a supervised outpatient envi-
ronment using a CLC with glucose targets similar to
pregnancy specific targets.
OO NCT03774186: In these two US centers trial, up to
47 women would be enrolled at ≤ 9 weeks’ gestation to
compare 670G with SAP therapy throughout most of
gestation and the first 6  weeks of the post-­ partum
period.
OO NCT04520971: In this trial starting in Europe (2021), up
to 92 women would be recruited at up to 12 weeks gesta-
tion and randomized to the 780G pump or standard of
care and followed up till delivery. The same CGM sys-
tem (Guardian™ Sensor 3) will be used to account for
differences in type of CGM in both groups for control
 Closed Loop Systems in Type 1 Diabetes 121

group and data are masked with collection at at least to pay threshold of $50,000 per QALY gained  [28]. The
four different time points in pregnancy: at 14–17 weeks, technologies significantly improve daily quality of life, with
20–23 weeks, 26–29 weeks, and 33–36 weeks. burden of self-­care decreasing iteratively.

Closed Loop in Adults Prone to Hypoglycemia

(DCLP2)
NCT04266379: A clinical trial is currently being conducted
in France testing Tandem Control-­IQ driven by CGM
(Dexcom G6) through an algorithm, in free-­living condi-
tions vs SAP for 3  months, with a 3-­month extension
phase. Estimated enrollment is of 72 participants and the
estimated study completion date is January 2021.
Cost of Diabetes Technologies
These technologies are expensive and have been used more
in Western Europe, USA, and Australia/New Zealand.
There is a lag between approval of technologies and availa-
bility in each country. A detailed cost benefit analysis done
by Pease et  al. reported that HCL were associated with a
mean of $37,767 per QALY gained and 86% of stimulations
predicted the HCLC to be cost-­effective at the willingness
Insulin Pump Use for People With T2DM
Though studies of insulin pump use in T2DM have been
published, they pre-­date pivotal studies of newer medica-
tions such as GLP1 Receptor Agonists and SGLT2 inhibitors
that have improved therapy for T2DM. Therapeutic technol-
ogy use for T2DM needs to be addressed in future studies.
Conclusion
In conclusion, technologies to treat diabetes have signifi-
cantly advanced in the last decade. The next decade is
expected to provide further advances. Please refer to
Table 9.3, which describes the technologies currently avail-
able and being developed further. These include advances
in glucose sensing, utilization of additional sensors, deliv-
ery of additional hormones in addition to insulin, espe-
cially glucagon and more advanced control algorithms.

TABLE 9.3  Brief synopsis of current and future technologies in closed loop.

Sensors Glucose CGM Implantable Eversense

Non-­implantable Dexcom G6, Medtronic Guardian
3, Abbott FreeStyle Libre
Non-­glucose Ketone Detects: β-­hydroxybutyrate (HB)
single
Non-­glucose CKM Detects: HB + glucose or HB/lactate
dual
Hormonal Single hormone Insulin
delivery Dual hormone Insulin + glucagon
Insulin + GLP-­1/SGLT2 inhibitors
Triple hormone Insulin + glucagon + pramlintide
Control Proportional integral derivative Medtronic 670G and Medtronic 780G
algorithm (PID)
Model predictive control (MPC) Tandem Control-­IQ
Fuzzy logic (FL) Omnipod
Target Single set point
glucose Single zone set point
Varying set points Overnight and daytime targets differ
Varying Day time set points Day time set points may differ between post-­
prandial and post-­absorptive

CGM – Continuous Glucose Monitoring, CKM – Continuous Ketone bodies monitoring

 122 Initial Evaluation and Management of Diabetes
References
 1. Atkinson MA and Maclaren NK. The pathogenesis of
insulin-­dependent diabetes mellitus. N Engl J Med.
1994;331(21):1428–36.
  2. Atkinson MA, Eisenbarth GS, and Michels AW. Type 1 dia-
betes. Lancet. 2014;383(9911):69–82.
  3. Eisenbarth GS. Type I diabetes mellitus. A chronic autoim-
mune disease. N Engl J Med. 1986;314(21):1360–8.
  4. Insel RA, Dunne JL, Atkinson MA et al. Staging presympto-
matic type 1 diabetes: a scientific statement of JDRF, the
Endocrine Society, and the American Diabetes Association.
Diabetes Care. 2015;38(10):1964–74.
  5. Kovatchev B. The artificial pancreas in 2017: The year of
transition from research to clinical practice. Nat Rev
Endocrinol. 2018;14(2):74–6.
  6. Tamborlane WV, Beck RW, Bode BW et al. Continuous glu-
cose monitoring and intensive treatment of type 1 diabetes.
N Engl J Med. 2008;359(14):1464–76.
  7. Beck RW, Riddlesworth T, Ruedy K et al. Effect of continu-
ous glucose monitoring on glycemic control in adults with
type 1 diabetes using insulin injections: The DIAMOND
randomized clinical trial. JAMA. 2017;317(4):371–8.
  8. Lind M, Polonsky W, Hirsch IB et  al. Continuous glucose
monitoring vs conventional therapy for glycemic control in
adults with type 1 diabetes treated with multiple daily insu-
lin injections: The GOLD randomized clinical trial. JAMA.
2017;317(4):379–87.
  9. Heinemann L, Freckmann G, Ehrmann D et  al. Real-­time
continuous glucose monitoring in adults with type 1 diabe-
tes and impaired hypoglycaemia awareness or severe hypo-
glycaemia treated with multiple daily insulin injections
(HypoDE): a multicentre, randomised controlled trial.
Lancet. 2018;391(10128):1367–77.
10. Pratley RE, Kanapka LG, Rickels MR et al. Effect of continu-
ous glucose monitoring on hypoglycemia in older adults
with type 1 diabetes: A randomized clinical trial. JAMA.
2020;323(23):2397–406.
11. Feig DS, Donovan LE, Corcoy R et al. Continuous glucose
monitoring in pregnant women with type 1 diabetes
(CONCEPTT): A multicentre international randomised
controlled trial. Lancet. 2017;390(10110):2347–59.
12. Kropff J, Choudhary P, Neupane S et al. Accuracy and lon-
gevity of an implantable continuous glucose sensor in the
PRECISE study: A 180-­day, prospective, multicenter, pivotal
trial. Diabetes Care. 2017;40(1):63–8.
13. Gilbert T, Noar A, Blalock O, and Polonsky W. Change in
HbA1c and quality of life with real-­time CGM use by people
with insulin-­treated diabetes in the landmark study. Diabetes
Technol Ther. 2021.
14. Foster NC, Beck RW, Miller KM et al. State of type 1 diabetes
management and outcomes from the T1D exchange in
2016–2018. Diabetes Technol Ther. 2019;21(2):66–72.
15. Carlson AL, Mullen DM, and Bergenstal RM. Clinical use of
continuous glucose monitoring in adults with type 2 diabe-
tes. Diabetes Technol Ther. 2017;19(S2):S4-­s11.
16. Poolsup N, Suksomboon N, and Kyaw AM. Systematic
review and meta-­analysis of the effectiveness of continuous
glucose monitoring (CGM) on glucose control in diabetes.
Diabetol Metab Syndr. 2013;5:39.
17. Bergenstal RM, Garg S, Weinzimer SA et  al. Safety of a
hybrid closed-­loop insulin delivery system in patients with
type 1 diabetes. JAMA. 2016;316(13):1407–8.
18. Brown SA, Kovatchev BP, Raghinaru D et al. Six-­month ran-
domized, multicenter trial of closed-­loop control in type 1
diabetes. N Engl J Med. 2019;381(18):1707–17.
19. Bergenstal RM, Nimri R, Beck RW et al. A comparison of
two hybrid closed-­loop systems in adolescents and young
adults with type 1 diabetes (FLAIR): a multicentre, ran-
domised, crossover trial. Lancet.
2021;397(10270):208–19.
20. Carlson AL, Bode BW, Brazg RL et al. 97-­LB: Safety and gly-
cemic outcomes of the Minimed advanced hybrid closed-­
loop (AHCL) system in subjects with T1D. Diabetes.
2020;June;69(Supplement 1).
21. Collyns O, Meier R, Betts Z et al. 199-­OR: Improved glyce-
mic outcomes with Medtronic Minimed advanced hybrid
closed-­loop delivery: Results from a randomized crossover
trial comparing automated insulin delivery with predictive
low glucose suspend in people with type 1 diabetes. Diabetes.
2020;June;69 (Supplement 1).
22. Stone MP, Agrawal P, Chen X et al. Retrospective analysis of
3-­month real-­world glucose data after the MiniMed 670G
system commercial launch. Diabetes Technol Ther.
2018;20(10):689–92.
23. Pinsker JE, Müller L, Constantin A et al. Real-­world patient
reported outcomes and glycemic results with initiation of
Control-­IQ technology. Diabetes Technol Ther.
2020;23(2):120–7.
24. Benhamou PY, Franc S, Reznik Y et al. Closed-­loop insulin
delivery in adults with type 1 diabetes in real-­life conditions:
a 12-­week multicentre, open-­label randomised controlled
crossover trial. Lancet Digit Health. 2019;1(1):e17-­e25.
25. Buckingham B, Beck RW, Ruedy KJ et  al. Effectiveness of
early intensive therapy on β-­cell preservation in type 1 dia-
betes. Diabetes Care. 2013;36(12):4030–5.
 Closed Loop Systems in Type 1 Diabetes 123

26. Stewart ZA, Wilinska ME, Hartnell S et al. Closed-­loop insu- 28. Pease A, Zomer E, Liew D, Earnest A et al. Cost-­effectiveness
lin delivery during pregnancy in women with type 1 diabe- analysis of a hybrid closed-­loop system versus multiple daily
tes. N Engl J Med. 2016;375(7):644–54. injections and capillary glucose testing for adults with type 1
27. Stewart ZA, Wilinska ME, Hartnell S et  al. Day-­and-­night diabetes. Diabetes Technol Ther. 2020;22(11):812–21.
closed-­loop insulin delivery in a broad population of preg-
nant women with type 1 diabetes: A randomized controlled
crossover trial. Diabetes Care. 2018;41(7):1391–9.
 10 Optimizing Diet in Patients with Diabetes
Meera Shah
Division of Endocrinology, Diabetes, Metabolism, and Nutrition, Mayo Clinic, Rochester, MN, USA
LE A R N I N G P OI NT S
●● A plant‐based diet can lower the incidence of type 2
diabetes in an average risk population.
●● A program of modest calorie restriction, resulting in
5–10% weight loss, should be the initial goal for
patients with glucose intolerance and an elevated BMI.
Time‐restricted feeding, which naturally aligns with the
biological circadian rhythm, may also be beneficial, at
least in the short term.
●● Aggressive calorie restriction in early type 2 diabetes
may be beneficial in the remission of type 2 diabetes;
however, this approach may only be suitable for a
select few. Overall, emphasis on healthy eating
patterns, and avoidance of ultra‐processed foods,
added sugars and refined grains should be the focus of
a medical nutrition therapy plan for type 2 diabetes.
●● Patients with type 2 diabetes have similar macronutri-
ent needs to the general population. Emphasis should
be on quality of the individual macronutrient, rather
than quantity. Sugar substitutes may have a role in
reducing overall carbohydrate and calorie intake in
patients with type 2 diabetes.
●● Patients with type 1 diabetes should be counseled on a
personalized dietary plan that accounts for other
comorbidities that may impact food intake.
Type 2 diabetes mellitus was once called a “disease of
affluence” because of its initial rise in societies that acquired
material wealth. Now, type 2 diabetes affects populations at
all levels of the socio‐economic strata worldwide and in
fact disproportionately affects populations in lower‐
income countries [1]. Across all cultures and countries, the
common theme has been changes in traditional lifestyles,
Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
124
particularly to food intake and activity, resulting in more
sedentary lifestyles and a rise in the prevalence of over-
weight and obesity.
Medical nutrition therapy is a cornerstone in the man-
agement of type 2 diabetes, and has been shown to be at
least as efficacious as anti‐hyperglycemic drugs in lowering
hemoglobin A1c [2]. Patients obtain dietary advice for the
management of type 2 diabetes from various sources; how-
ever, not all are reliable or individualized for that particular
patient. A common question from a patient with diabetes
for the practicing clinician is “What should I eat?”, and the
answer can vary based on where the individual is in terms
of glycemic disturbance, what their body weight is, and
what their own personal and cultural preferences are.
A 40‐year‐old man with no family history of type 2 diabetes
asks you about how he can prevent the onset of diabetes
through a dietary plan. He is of normal body weight and has
normal fasting glucose.
The Nurses Health Study  [3] was an early reporter on
the impact of lifestyle on the risk of developing type 2 dia-
betes. Being overweight or having obesity was the single
best predictor of developing diabetes in this study.
However, when controlled for other variables, diet quality
had a measurable impact on the risk of developing the dis-
ease. Specifically, the group with the highest intake of
cereal fiber reduced the risk of the disease by approxi-
mately 40%; while the group that had the highest consump-
tion of trans fats increased their risk of developing the
disease by approximately 40%.

Plant‐based diets have been extensively studied. This
type of diet is variably described across studies, but com-
mon elements are:
●● Foods derived from plant sources with lower consump-
tion of or exclusion of animal products
●● May exclude potentially beneficial yogurt and fish
●● May include consumption of refined grains, starches,
and sugars
The Adventist Health Study enrolled approximately 40
000 Adventist Church members in the United States and
Canada in the early 2000s [4]. Over a two‐year period, the
incidence of diabetes was higher in members not following
a vegetarian diet compared to members following a semi‐
vegetarian or vegetarian diet. Similar findings were seen in
a combined analysis of participants from the Nurses’
Health Study, Nurses’ Health Study II and Health
Professionals Follow‐up Study totaling approximately 200
000 participants. Subjects who adhered to a plant‐based
diet or healthy plant‐based diet reduced their risk of devel-
oping type 2 diabetes by 50%. Interestingly, participants
scoring high on the unhealthy plant‐based index (examples
include consuming fruit juices, refined grains, potato prod-
ucts, and sweets and desserts) had a higher risk of develop-
ing type 2 diabetes even after adjustment for BMI  [5]. A
recent meta‐analysis and systematic review combining
nine large prospective studies showed similar results, con-
firming both the benefit a healthy plant‐based diet which
minimizes refined carbohydrates, and the detrimental
effects of consuming an unhealthy plant‐based diet [6].
A popular version of the plant‐based diet is the
Mediterranean diet. This food plan is also variably
described across studies, but common elements include:
●● primarily plant‐based
●● high consumption of foods of vegetable origin, such as
cereals and whole grains, fruits, vegetables, legumes, nuts
●● olive oil as the principal source of fat
●● fish and poultry consumed in low to moderate amounts
●● relatively low consumption of red meat
●● moderate consumption of wine, normally with meals.
The impact of the Mediterranean diet on the develop-
ment type 2 diabetes in the general European population
was studied by the InterAct Consortium [7]. Almost half
Optimizing Diet in Patients with Diabetes 125
million people were followed over approximately two dec-
ades and the incidence of diabetes was recorded. Compared
to low adherence to the Mediterranean diet plan, high
adherence to the Mediterranean diet plan reduced the risk
of developing type 2 diabetes by 12%. The association was
less strong in subjects with obesity, or those younger than
50 years of age, however.
In another European population with traditional cardi-
ovascular risk factors, adhering to a Mediterranean style
diet reduced the risk of developing type 2 diabetes by 50%
when compared to a low‐fat diet, an effect that was inde-
pendent of changes in weight or physical activity [8].
Cultural norms, food availability, native produce and
food preferences are some of the factors that determine
how different populations eat. In a large meta‐ analysis
combining studies from across the globe (Asia, Europe,
North American) a 20% reduced risk of developing diabe-
tes was observed in populations with greatest adherence to
a plant‐based diet, independent of weight changes  [9].
Hence, any form of a healthy plant‐based food plan,
adjusted to an individual’s personal and cultural prefer-
ences, can be beneficial in reducing the overall risk of
developing type 2 diabetes.
Several mechanisms have been proposed to explain the
benefit of a plant‐based diet in the risk reduction of type 2
diabetes. Components commonly present in fruits and
vegetables such as fiber, vitamins and minerals, antioxi-
dants, phenolic compounds, and unsaturated fatty acids
improve insulin sensitivity and blood pressure, reduce
long‐term weight gain, and ameliorate systemic inflamma-
tion. There is also the de‐emphasis or avoidance of red and
processed meats, which can adversely affect risk of type 2
diabetes, possibly owing to high heme iron or dietary cho-
lesterol content  [10]. The interaction of animal‐derived
compounds (e.g. choline and l‐carnitine) with the gut
microbiome and the production of trimethylamine‐N‐
oxide have been implicated in the risk of developing gesta-
tional diabetes and type 2 diabetes, although the current
evidence remains inconclusive [11, 12].
C lin ical pear l
A plant‐based diet can lower the incidence of type 2
diabetes in an average risk population.
 126 Initial Evaluation and Management of Diabetes
A 45‐year‐old woman with impaired fasting glucose asks you
if she can reduce her risk of developing diabetes with a diet
plan alone. She has a BMI of 31 kg/m2.
Being overweight or having obesity is the most impor-
tant risk factor for the development of type 2 diabetes.
Patients with type 2 diabetes who have an elevated BMI (>
than 25  kg/m2 or > 23  kg/m2 in the Asian population)
should initially be counseled on weight loss through life-
style change. A modest reduction in weight of approxi-
mately 5–10% has been shown to provide many health
benefits with improvement in metabolic parameters  [13,
14]. Calorie restriction is a core component of any weight
loss program. However, no one food plan has been shown
to be superior to another when it comes to efficacy or sus-
tainability of weight loss. Hence, a patient’s personal prefer-
ence should be the most important determinant of what
dietary plan is advised.
Intermittent fasting is one popular way of managing calo-
rie consumption. The health benefits of intermittent fasting
result from its impact on cellular adaptations that promote
metabolic flexibility, i.e. the ability of the body to switch
from using glucose as a primary fuel source to fatty acids and
ketone bodies; increase stress reduction and reduce inflam-
mation [15]. The concept relies on the total absence of calo-
rie consumption over a consecutive period of time, typically
at least 16 hours, during which these advantageous cellular
adaptations occur [16, 17]. An observational study in mem-
bers of the Church of Latter Day Saints, who routinely
observe one 24‐hour fast a month as part of their religious
obligations, reported the risk of developing type 2 diabetes
was half that of non‐fasters [18].
A corollary to intermittent fasting is time‐restricted eat-
ing (TRE), which is based on the concept that endogenous
metabolic rhythms (glucose, lipids and energy metabo-
lism) have circadian patterns  [19, 20]. A small but rigor-
ously conducted study on eight men with pre‐diabetes
showed an improvement in insulin sensitivity and ß cell
function after TRE, when compared to an isocaloric, tradi-
tional “three meals a day” eating pattern suggesting that the
beneficial impact of TRE is independent of weight loss [21].
Data is not as strong in patients with established type 2 dia-
betes, thus no firm conclusions can be made about the use
of intermittent fasting or TRE in the management of type 2
diabetes. An important consideration before recommend-
ing such an eating pattern will be to ensure that the patient
is proficient in adjusting anti‐diabetic medications, espe-
cially insulin, because of the risk of precipitating
hypoglycemia [22].
C lin ical pear l
A program of modest calorie restriction, resulting in
5–10% weight loss, should be the initial goal for patients
with glucose intolerance and an elevated BMI. Time‐
restricted feeding, which naturally aligns with the
biological circadian rhythm, may also be beneficial, at
least in the short term.
A 55‐year‐old man with recently diagnosed type 2 diabetes
asks if he can “reverse” diabetes with dietary change alone.
He has an A1c of 7.5% and is on a maximally tolerated dose
of metformin.
The Look AHEAD study enrolled over 5000 partici-
pants with type 2 diabetes and randomized them to inten-
sive lifestyle modification or standard medical care  [23].
Participants in the intervention group were provided a
calorie‐restricted meal plan with the use of meal replace-
ment products in the first year of the intervention. At the
end of an almost 10‐year follow‐up period, participants in
the intensive lifestyle intervention group lost more weight
and had better improvements in hemoglobin A1c  [23].
However, the primary outcome of composite of death from
cardiovascular causes, non‐fatal MI, non‐fatal stroke, or
hospitalization for angina was not different between the
two groups. Nevertheless, other clinically important out-
comes as a result of the intensive lifestyle intervention pro-
gram, such as an improvement in glycemic control with the
subsequent reduction in the need for insulin, improve-
ments in physical fitness, and overall improvement in
mobility and quality of life should be acknowledged.
Significant calorie restriction in the early phases of type
2 diabetes may be able to achieve disease remission, as
demonstrated by the DiRECT trial [24]. This study enrolled
approximately 300 participants with type 2 diabetes diag-
nosed in the last six years, with variable glycemic control
(hemoglobin A1c 6% to 12%) and not on insulin. All
patients were at least in the overweight category of body

mass index and randomized to an intervention arm or con-
trol arm. Patients enrolled in the intervention arm were
taken off all diabetes medications and placed on a very low
calorie diet (approximately 850 kilocalories) in the form of
meal replacement products over a period of 12 to 20 weeks.
Following the initial weight loss phase, patients were tran-
sitioned to a stepped food re‐introduction program over
the next 2–8  weeks with a goal of weight maintenance.
Patients who gained weight during the weight maintenance
phase were offered products to attenuate this.
Approximately 25% of the intervention group achieved at
least 15  kg weight loss. Diabetes remission, defined as
hemoglobin A1c of less than 6.5% after at least two months
off all antidiabetic medications, was achieved in 46% of the
intervention group, and sustained for one year. At the two‐
year follow‐up mark, this proportion was approximately
36%. In the study population as a whole, the likelihood of
diabetes remission was related to the degree of weight loss
achieved. In the same vein, sustained remission of diabetes
was more likely in participants with sustained weight loss.
Of note, there was a 22% drop‐out rate after 24 months in
the intervention arm (versus 6% in the control group),
which speaks to the difficulty in adhering to such an inten-
sive program.
The Mediterranean diet plan has been studied in the
management of early type 2 diabetes. Patients with early
type 2 diabetes and on no medications were randomized to
either an iso‐caloric low‐carbohydrate Mediterranean diet
or a low‐fat diet. Participants in the low‐carbohydrate
Mediterranean diet group were able to postpone the initia-
tion of an anti‐diabetes medication by about two years,
when compared to the low‐fat group, largely independent
of weight loss [25].
In a one year, prospective randomized trial consisting of
approximately 250 participants with class I obesity, the
low‐carbohydrate Mediterranean diet (less than 35% low
glycemic index carbohydrates) and standard Mediterranean
diet were superior to the 2003 American Diabetes
Association (50–55% carbohydrate, 30% fat and 20% pro-
tein) diet in terms of improvements in A1c. This result
occurred despite no significant between‐group differences
in calorie consumption or achieved weight loss [26].
Interest in the ketogenic or very low carbohydrate diet
for the management of type 2 diabetes remains strong. The
Optimizing Diet in Patients with Diabetes 127
benefit of inducing ketonemia was first described by Drs
Wilder and Winter at the Mayo Clinic in the 1920s, in the
context of treating epilepsy. Numerous studies since then
have focused on its role in the management of type 2 diabe-
tes. Ketogenic diets are described differently in the litera-
ture; however, common components include a reduction in
carbohydrates (usually to less than 50 g/day) and a relative
increase in the proportions of protein and fat. Others have
also variably described an intake between 50–150 g/day, or
less than 30% of calories from carbohydrates as “low carb
intake”. There are many commercial diet programs that uti-
lize the concept of a low carbohydrate/ketogenic diet for
weight loss, each with its own macronutrient composition
and other proprietary products. This does make the
research space heterogeneous and complex and hence firm
conclusions cannot be drawn about ketogenic diets in the
management of type 2 diabetes or impaired glucose
tolerance.
A common confounder in these studies is what role
weight loss alone plays in explaining the beneficial effects
seen by following the diet plan. Additionally, these diets
can be difficult to adhere to in the long term. A recent
meta‐analysis showed that the metabolic advantages seen
with the low‐carbohydrate diet (versus a high carbohydrate
diet) at 6 months were lost at 12 months [27]. Because of
this well recognized phenomenon of difficulty with long‐
term adherence, individuals interested in this diet plan
should be regularly assessed and their diet plans individu-
alized as needed. Currently, the American Diabetes
Association dietary guidelines recommend avoiding the
ketogenic diet plan in patients with type 2 diabetes who are
pregnant or lactating; have disordered eating; have renal
disease; or are taking an SGLT‐2 inhibitor because of the
potential risk of ketoacidosis [28].
On the opposite side of the coin, ultra‐processed foods
may have a detrimental effect on glycemic control. Ultra‐
processed foods have been defined as “formulations mostly
of cheap industrial sources of dietary energy and nutrients
plus additives, using a series of processes” and containing
minimal whole foods [29]. Ultra‐processed foods are typi-
cally ready‐to‐eat formulations with five or more ingredi-
ents, often including flavor‐enhancing additives, dyes, or
stabilizers (NOVA classification  [30]) and can include
foods undergoing physical and chemical processes such as
 128 Initial Evaluation and Management of Diabetes

extruding, molding, pre‐frying, and hydrogenation. Ultra‐

processed foods currently constitute the majority of calo-
ries consumed in America [31]. In a large population‐based
prospective study, comprising over 100 000 participants,
consuming a higher proportion of ultra‐processed foods in
the diet was associated with a higher risk of developing
type 2 diabetes. This was true even when adjusting for the
consumption of unprocessed or minimally processed foods
concurrently [32]. Several theories have been put forward
to explain this observation, including the propensity to
consume more calories when eating a diet high in ultra‐
processed foods  [33]. However, no definitive mechanism
has been identified.
The American Diabetes Association in its most recent
statement on medical nutrition therapy in the management
of type 2 diabetes, emphasizes that there are no comparison
studies between eating patterns  [28]. However, certain
commonalities between seemingly beneficial eating pat-
terns such as an emphasis on fruits, non‐starchy vegetables,
whole grains and minimizing refined grains and added
sugar should form the basis of any dietary intervention in
the management of type 2 diabetes.
Glycemic Index
C lin ical pear l
Aggressive calorie restriction in early type 2 diabetes may
be beneficial in inducing remission of type 2 diabetes;
however, this approach may only be suitable for a select
few. Overall, emphasis on healthy eating patterns, and
avoidance of ultra‐processed foods, added sugars and
refined grains should be the focus of a medical nutrition
therapy plan for type 2 diabetes.
A 60‐year‐old man with type 2 diabetes wonders if he should
limit carbohydrates or include a certain amount of protein or
fiber in his diet to optimize glycemic control. He also wonders
about the role of artificial sweeteners, as he has read a lot
about their detrimental health effects.
Overall, consensus guidelines for macronutrient intake
in people with type 2 diabetes are similar to the general
population  [28]. Carbohydrate quality, rather than quan-
tity, should be emphasized when providing dietary advice,
as quality of carbohydrate has the greatest impact on glyce-
mic response  [34] (Figure  10.1). Optimal health benefits
are seen when carbohydrate intake is from whole grains,

Calorie density

FIG 10.1  Schematic representation of the relationship between caloric density and glycemic index of typical foods. It is important to
remember that degree of ripeness, texture and cooking method can affect the glycemic index of some foods. Source: Adapted
from The University of Sydney, from: http://www.glycemicindex.com/index.php

fruits, and vegetables. Some patients with diabetes count
carbohydrates to manage post‐meal glycemic excursions
and for others, prandial insulin dosing. There are two
assumptions when carbohydrate counting is utilized: one,
that carbohydrate is the main macronutrient affecting gly-
cemic response and two, that carbohydrates are quickly
converted to glucose after a meal. Carbohydrate counting
involves detailed education, consistent practice and regular
continued visits with a registered dietitian to ensure ongo-
ing benefit with this practice.
Based on current evidence, protein needs in patients
with type 2 diabetes are no different to that in the general
population, assuming normal renal function. Protein
intake should represent approximately 20% of calories con-
sumed, and adjusted to renal function.
Similar to the general population, guidelines for dietary
fat intake in people with diabetes is approximately 20–35%
of total calorie intake. As is the case with carbohydrates,
quality of fat matters more than quantity, and certain
higher‐fat dietary patterns have been shown to improve
cardiovascular risk factors and glycemia in people with
type 2 diabetes [35]. There is clear evidence that consump-
tion of trans‐fat is detrimental to health and should be
avoided  [36, 37]. An important consideration in people
with diabetes who are also trying to manage weight is the
energy density associated with fat. These patients should
be mindful of total caloric consumption, even if from
“healthier fats” so as not to promote weight gain.
Regular dietary fiber intake is associated with lower all‐
cause mortality in people with type 2 diabetes  [34, 38].
Current dietary guidelines recommend 14 g of fiber for
every 1000  kilocalories consumed, with no distinction
made between soluble and non‐soluble fiber  [37]. Fiber
supplementation may improve glycemic control. When
overweight and obese volunteers were placed on an iso‐
caloric, weight‐maintaining diet that differed only by the
amount of dietary fiber from whole grains consumed for
6 weeks (28 g versus 17.8 g), the group on the whole grain
diet had a significant improvement in HOMA‐IR and
other insulin sensitivity indices as measured by the eugly-
cemic‐hyperinsulinemic clamp, compared to individuals
on the refined grain diet  [39]. The American Diabetes
Association recommends that fiber be incorporated
through foods that are naturally high in fiber, as opposed to
Optimizing Diet in Patients with Diabetes 129
supplements, to reap other benefits such as micronutrients
and phytochemicals naturally found in foods.
Sugar substitutes, which include non‐nutritive sweeten-
ers, high‐intensity sweeteners, low‐calorie sweeteners, and
artificial sweeteners, have been deemed safe for consump-
tion by humans. When sugar substitutes are used to replace
sugar or other caloric sweeteners, they may have a role in
reducing overall carbohydrate and caloric intake [40]. To
date, there is no conclusive evidence that sugar substitutes
are associated with any health risk to humans [40]. Sugar
alcohols such as sorbitol or xylitol are associated with lower
postprandial glucose rise and lower calorie content com-
pared to sucrose. Because sugar alcohol is less sweet, a
greater quantity is often required to achieve an equivalent
level of sweetness, thus negating any calorie benefit when
compared to sugar [41, 42]. Additionally, large amounts of
sugar alcohols (30–49 g) can be associated with osmotic
diarrhea; hence use of sugar alcohols should be balanced
with gastrointestinal side effects.
C lin ical pear l
Patients with type 2 diabetes have similar macronutrient
needs to the general population. Emphasis should be on
quality of the individual macronutrient, rather than
quantity. Sugar substitutes may have a role in reducing
overall carbohydrate and calorie intake in patients with
type 2 diabetes.
A 20‐year‐old lady with Type 1 diabetes asks if there is a
specific dietary plan she should follow to better manage dia-
betes. She has no complications from diabetes. Her BMI is
30 kg/m2.
Current dietary guidelines published by the American
Diabetes Association are applicable to all people living with
diabetes. Some groups have suggested that patients with
type 1 diabetes, who tend to have greater glycemic variabil-
ity, could benefit from a low‐carbohydrate diet  [43, 44].
Unfortunately, studies of dietary patterns in patients with
type 1 diabetes are often of short duration and in small
numbers of participants, and hence firm conclusions can-
not be drawn. Patients with type 1 diabetes may also be
living with other conditions that alter dietary intake, for
example celiac disease or gastroparesis. In recent times, the
 130 Initial Evaluation and Management of Diabetes
proportion of patients with type 1 diabetes and concurrent
obesity has increased, and it is currently estimated at
50% [45]. These patients would benefit from a weight loss
program to both improve glycemic control as well as other
aspects of health [46–48].
Clinical pearl
Patients with type 1 diabetes should be counseled on a
personalized dietary plan that accounts for other
comorbidities that may impact food intake.
Summary
Medical nutrition therapy remains the foundation in the
management of patients with diabetes mellitus. Dietary
guidelines favor healthy eating plans over specific macro-
nutrients. Ultimately, the dietary plan that may best benefit
the individual patient should take into account available
evidence, and the patient’s own personal and cultural
preferences.
References
1. Roglic G, World Health Organization. Global Report on
Diabetes. Geneva, Switzerland: World Health Organization;
2016.
2. Evert AB, Franz MJ, American Diabetes Association.
American Diabetes Association Guide to Nutrition Therapy
for Diabetes. Third edition. Arlington: American Diabetes
Association; 2017. xvii.
3. Hu FB, Manson JE, Stampfer MJ, Colditz G, Liu S, Solomon
CG et al. Diet, lifestyle, and the risk of type 2 diabetes
mellitus in women. N Engl J Med. 2001;345(11):790–797.
4. Tonstad S, Stewart K, Oda K, Batech M, Herring RP,
Fraser GE. Vegetarian diets and incidence of diabetes in
the Adventist Health Study‐2. Nutr Metab Cardiovasc Dis.
2013;23(4):292–299.
5. Satija A, Bhupathiraju SN, Rimm EB, Spiegelman D, Chiuve
SE, Borgi L et al. Plant‐based dietary patterns and incidence
of type 2 diabetes in us men and women: results from three
prospective cohort studies. PLoS Med. 2016;13(6):
e1002039.
6. Qian F, Liu G, Hu FB, Bhupathiraju SN, Sun Q. Association
between plant‐based dietary patterns and risk of type 2
diabetes: a systematic review and meta‐analysis. JAMA
Intern Med. 2019.
7. InterAct C, Romaguera D, Guevara M, Norat T, Langenberg C,
Forouhi NG et  al. Mediterranean diet and type 2 diabetes
risk in the European Prospective Investigation into Cancer
and Nutrition (EPIC) study: the InterAct project. Diabetes
Care. 2011;34(9):1913–1918.
8. Salas‐Salvado J, Bullo M, Babio N, Martinez‐Gonzalez MA,
Ibarrola‐Jurado N, Basora J et al. Reduction in the incidence
of type 2 diabetes with the Mediterranean diet: results of the
PREDIMED‐Reus nutrition intervention randomized trial.
Diabetes Care. 2011;34(1):14–19.
9. Esposito K, Chiodini P, Maiorino MI, Bellastella G,
Panagiotakos D, Giugliano D. Which diet for prevention of
type 2 diabetes? A meta‐analysis of prospective studies.
Endocrine. 2014;47(1):107–116.
10. Micha R, Michas G, Mozaffarian D. Unprocessed red and
processed meats and risk of coronary artery disease and type
2 diabetes – an updated review of the evidence. Curr
Atheroscler Rep. 2012;14(6):515–524.
11. Li P, Zhong C, Li S, Sun T, Huang H, Chen X et al. Plasma
concentration of trimethylamine‐N‐oxide and risk of gesta-
tional diabetes mellitus. Am J Clin Nutr. 2018;108(3):
603–610.
12. Svingen GF, Schartum‐Hansen H, Pedersen ER, Ueland PM,
Tell GS, Mellgren G et  al. Prospective associations of sys-
temic and urinary choline metabolites with incident type 2
diabetes. Clin Chem. 2016;62(5):755–765.
13. Pan XR, Li GW, Hu YH, Wang JX, Yang WY, An ZX et al.
Effects of diet and exercise in preventing NIDDM in people
with impaired glucose tolerance. The Da Qing IGT and
Diabetes Study. Diabetes Care. 1997;20(4):537–544.
14. Diabetes Prevention Program Research G. The Diabetes
Prevention Program (DPP): description of lifestyle interven-
tion. Diabetes Care. 2002;25(12):2165–2171.
15. Mattson MP, Moehl K, Ghena N, Schmaedick M, Cheng A.
Intermittent metabolic switching, neuroplasticity and brain
health. Nat Rev Neurosci. 2018;19(2):63–80.
16. Panda S. Circadian physiology of metabolism. Science.
2016;354(6315):1008–1015.
17. Di Francesco A, Di Germanio C, Bernier M, de Cabo R. A
time to fast. Science. 2018;362(6416):770–775.
18. Horne BD, Muhlestein JB, May HT, Carlquist JF, Lappe DL,
Bair TL et al. Relation of routine, periodic fasting to risk of
diabetes mellitus, and coronary artery disease in patients
undergoing coronary angiography. Am J Cardiol. 2012;109(11):
1558–1562.
19. Poggiogalle E, Jamshed H, Peterson CM. Circadian regula-
tion of glucose, lipid, and energy metabolism in humans.
Metabolism. 2018;84:11–27.

20. Scheer FA, Hilton MF, Mantzoros CS, Shea SA. Adverse meta-
bolic and cardiovascular consequences of circadian misalign-
ment. Proc Natl Acad Sci USA. 2009;106(11):4453–4458.
21. Sutton EF, Beyl R, Early KS, Cefalu WT, Ravussin E, Peterson
CM. Early time‐restricted feeding improves insulin sensitiv-
ity, blood pressure, and oxidative stress even without weight
loss in men with prediabetes. Cell Metab. 2018;27(6):
1212–1221 e3.
22. Carter S, Clifton PM, Keogh JB. The effect of intermittent
compared with continuous energy restriction on glycaemic
control in patients with type 2 diabetes: 24‐month follow‐up
of a randomised noninferiority trial. Diabetes Res Clin Pract.
2019;151:11–19.
23. Cardiovascular Effects of Intensive Lifestyle Intervention in
Type 2 Diabetes. New Eng J Med. 2013;369(2):145–154.
24. Lean ME, Leslie WS, Barnes AC, Brosnahan N, Thom G,
McCombie L et al. Primary care‐led weight management for
remission of type 2 diabetes (DiRECT): an open‐label, clus-
ter‐randomised trial. Lancet. 2018;391(10120):541–551.
25. Esposito K, Maiorino MI, Petrizzo M, Bellastella G,
Giugliano D. The effects of a Mediterranean diet on the need
for diabetes drugs and remission of newly diagnosed type 2
diabetes: follow‐up of a randomized trial. Diabetes Care.
2014;37(7):1824–1830.
26. Elhayany A, Lustman A, Abel R, Attal‐Singer J, Vinker S. A
low carbohydrate Mediterranean diet improves cardiovascu-
lar risk factors and diabetes control among overweight
patients with type 2 diabetes mellitus: a 1‐year prospective
randomized intervention study. Diabetes Obes Metab.
2010;12(3):204–209.
27. Korsmo‐Haugen HK, Brurberg KG, Mann J, Aas AM.
Carbohydrate quantity in the dietary management of type 2
diabetes: a systematic review and meta‐analysis. Diabetes
Obes Metab. 2019;21(1):15–27.
28. Evert AB, Dennison M, Gardner CD, Garvey WT, Lau KHK,
MacLeod J et al. Nutrition therapy for adults with diabetes or
prediabetes: a consensus report. Diabetes Care. 2019;42(5):
731–754.
29. Monteiro CA, Cannon G, Moubarac JC, Levy RB, Louzada
MLC, Jaime PC. The UN Decade of Nutrition, the NOVA
food classification and the trouble with ultra‐processing.
Public Health Nutr. 2018;21(1):5–17.
30. Monteiro CA CG, Levy RB, Moubarac J‐C, Jaime P, Martins AP,
Canella D, Louzada ML, Parra D; with Ricardo C, Calixto G,
Machado P, Martins C, Martinez E, Baraldi L,
Garzillo J, Sattamini I. NOVA. The star shines bright [Food
classification. Public health]. World Nutrition. 2016:1–3, 28–38.
31. Martinez Steele E, Raubenheimer D, Simpson SJ, Baraldi LG,
Monteiro CA. Ultra‐processed foods, protein leverage and
Optimizing Diet in Patients with Diabetes 131
energy intake in the USA. Public Health Nutr.
2018;21(1):114–124.
32. Srour B, Fezeu LK, Kesse‐Guyot E, Alles B, Debras C,
Druesne‐Pecollo N et al. Ultraprocessed food consumption
and risk of type 2 diabetes among participants of the
NutriNet‐Santé prospective cohort. JAMA Intern Med.
2020;180(2):283–291.
33. Hall KD, Ayuketah A, Brychta R, Cai H, Cassimatis T, Chen
KY et  al. Ultra‐processed diets cause excess calorie intake
and weight gain: an inpatient randomized controlled trial of
ad libitum food intake. Cell Metab. 2019;30(1):226.
34. He M, van Dam RM, Rimm E, Hu FB, Qi L. Whole‐grain,
cereal fiber, bran, and germ intake and the risks of all‐cause
and cardiovascular disease‐specific mortality among women
with type 2 diabetes mellitus. Circulation. 2010;121(20):
2162–2168.
35. Qian F, Korat AA, Malik V, Hu FB. Metabolic effects of mon-
ounsaturated fatty acid‐enriched diets compared with car-
bohydrate or polyunsaturated fatty acid‐enriched diets in
patients with type 2 diabetes: a systematic review and meta‐
analysis of randomized controlled trials. Diabetes Care.
2016;39(8):1448–1457.
36. Aronis KN, Khan SM, Mantzoros CS. Effects of trans fatty
acids on glucose homeostasis: a meta‐analysis of rand-
omized, placebo‐controlled clinical trials. Am J Clin Nutr.
2012;96(5):1093–1099.
37. U.S. Department of Health and Human Services and U.S.
Department of Agriculture. 2015–2020 Dietary Guidelines
for Americans. December 2015.
38. Burger KN, Beulens JW, van der Schouw YT, Sluijs I,
Spijkerman AM, Sluik D et  al. Dietary fiber, carbohydrate
quality and quantity, and mortality risk of individuals with
diabetes mellitus. PLoS One. 2012;7(8):e43127.
39. Pereira MA, Jacobs DR, Jr., Pins JJ, Raatz SK, Gross MD,
Slavin JL et al. Effect of whole grains on insulin sensitivity in
overweight hyperinsulinemic adults. Am J Clin Nutr.
2002;75(5):848–855.
40. Gardner C, Wylie‐Rosett J, Gidding SS, Steffen LM, Johnson
RK, Reader D et  al. Nonnutritive sweeteners: current use
and health perspectives: a scientific statement from the
American Heart Association and the American Diabetes
Association. Diabetes Care. 2012;35(8):1798–1808.
41. Wiebe N, Padwal R, Field C, Marks S, Jacobs R, Tonelli M. A
systematic review on the effect of sweeteners on glycemic
response and clinically relevant outcomes. BMC Med.
2011;9:123.
42. Fitch C, Keim KS, Academy of N, Dietetics. Position of the
Academy of Nutrition and Dietetics: use of nutritive and non-
nutritive sweeteners. J Acad Nutr Diet. 2012;112(5):739–758.
 132 Initial Evaluation and Management of Diabetes
43. Nielsen JV, Gando C, Joensson E, Paulsson C. Low carbohy-
drate diet in type 1 diabetes, long‐term improvement and
adherence: a clinical audit. Diabetol Metab Syndr. 2012;
4(1):23.
44. Ranjan A, Schmidt S, Damm‐Frydenberg C, Holst JJ, Madsbad
S, Norgaard K. Short‐term effects of a low carbohydrate diet on
glycaemic variables and cardiovascular risk markers in patients
with type 1 diabetes: a randomized open‐label crossover trial.
Diabetes Obes Metab. 2017;19(10):1479–1484.
45. Conway B, Miller RG, Costacou T, Fried L, Kelsey S, Evans
RW et  al. Temporal patterns in overweight and obesity in
Type 1 diabetes. Diabet Med. 2010;27(4):398–404.
46. Chillaron JJ, Benaiges D, Mane L, Pedro‐Botet J, Flores Le‐
Roux JA. Obesity and type 1 diabetes mellitus management.
Minerva Endocrinol. 2015;40(1):53–60.
47. de Ferranti SD, de Boer IH, Fonseca V, Fox CS, Golden SH,
Lavie CJ et al. Type 1 diabetes mellitus and cardiovascular
disease: a scientific statement from the American Heart
Association and American Diabetes Association. Circulation.
2014;130(13):1110–1130.
48. Corbin KD, Driscoll KA, Pratley RE, Smith SR, Maahs DM,
Mayer‐Davis EJ et al. Obesity in type 1 diabetes: pathophysiol-
ogy, clinical impact, and mechanisms. Endocr Rev. 2018;39(5):
629–663.
 11 Are Insulin Sensitizers Useful Additions
to Insulin Therapy?
John W. Richard, III1 and Philip Raskin2
1
Endocrinology Fellow, Division of Endocrinology, Diabetes, Nutrition and Metabolism, University of Texas
Southwestern Medical Center at Dallas, Dallas, TX, USA
2
Professor of Medicine, Clifton and Betsy Robinson Chair in Biomedical Research, University of Texas Southwestern
Medical Center at Dallas, Dallas, TX, USA

tion characterized by decreased insulin sensitivity, high

LE A R NI NG P OI NT S
●● Patients with type 2 diabetes often need higher doses
of insulin because of obesity and decreased insulin
sensitivity.
●● Insulin sensitizing agents act as useful additions to
insulin therapy to combat worsening insulin sensitivity.
Metformin suppresses hepatic glucose production
●●
insulin levels, high triglycerides, and low HDL cholesterol
leading to type 2 diabetes and cardiovascular disease risk.
Considering the role of insulin resistance in the develop-
ment of these disease processes, insulin action at the tar-
get tissue level is an attractive therapeutic target in type 2
diabetes  [1]. The biguanides (i.e., metformin) and the
thiazolidinediones (TZDs) act directly to improve insulin

(gluconeogenesis) and increases peripheral tissue

insulin sensitivity.
●● Thiazolidinediones increase insulin‐stimulated glucose
uptake in peripheral tissues, hepatic insulin sensitivity,
and insulin sensitivity in adipose tissue through
suppression of fatty acid production.
●● Insulin sensitizing agents have proven effectiveness at
improving glucose control and lipid profiles, and
reducing insulin dose requirements; however, these
agents are not without side effects.
Insulin sensitizers
Insulin resistance is not a new phenomenon. In the 1930s
British physician Harold Percival Himsworth coined the
term “insulin insensitivity” to describe patients he observed
with resistance to injectable insulin. Fifty years later, Dr.
Gerald Reaven described metabolic syndrome X, a condi-
sensitivity, and so are regarded as insulin sensitizing
drugs.
Biguanides
In Europe, Galega officinalis (goat’s rue or French lilac) was
used for centuries to treat diabetes. This guanidine‐rich
substance later led to the development of several glucose‐
lowering guanidine derivatives in the 1920s. With the dis-
covery of insulin these antidiabetic agents were forgotten
until the 1950s when the biguanides: metformin, phen-
formin and buformin were reintroduced into diabetes
treatment [1]. Phenformin was withdrawn in many coun-
tries and from the U.S. market in 1975 because of a high
incidence of lactic acidosis. Buformin received only limited
use in a few countries, making metformin the principal
biguanide drug used in pharmacotherapy worldwide, espe-
cially with its improved safety profile and lower cost [1].

Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.

© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.

133
 134 Initial Evaluation and Management of Diabetes

Mechanism of action in the splanchnic bed also increases through insulin‐

The mechanism of action of biguanides (i.e., metformin) is
still not fully understood, but this agent is distinctly differ-
ent from oral sulfonylureas. Biguanides do not stimulate
pancreatic insulin secretion and only partial suppression of
gluconeogenesis occurs in the liver, therefore, hypoglyce-
mia with monotherapy is rare [2, 3]. Metformin has a vari-
ety of metabolic effects, some of which extend beyond
glucose lowering (Table 11.1). At a cellular level, metformin
improves insulin sensitivity through a mediated modifica-
tion of post‐receptor signaling in the insulin pathway.
Recent data have suggested that a protein, adenosine 5′‐
monophosphate protein kinase (AMPK), has been identi-
fied as a possible target of metformin [1]. The mainstay of
action of metformin can be attributed to its hepatic effects.
Hepatic sensitivity to insulin is increased by metformin,
thereby reducing gluconeogenesis (Figure 11.1) as well as
glycogenolysis, which contribute to the postprandial
plasma glucose lowering effects. Skeletal muscle and adi-
pocytes undergo upregulation of the insulin‐sensitive
GLUT‐4 and GLUT‐1 transporters to the cell membranes,
thereby increasing glucose uptake [2]. Glucose metabolism
dependent mechanisms. Additional metabolic effects
include the suppression of fatty acid oxidation (Figure 11.2)
as well as a reduction in triglyceride levels in patients with
hypertriglyceridemia [1, 2]. Combined, the cellular effects
of metformin counter insulin resistance and reduce the
toxic metabolic effects of glucose toxicity and lipotoxicity
in type 2 diabetes.
TABLE 11.1  Direct and indirect effects of metformin therapy.
Decreases hyperglycemia
Improves diastolic function
Decreases total cholesterol levels
Decreases very low‐density lipoprotein cholesterol levels
Decreases low‐density lipoprotein cholesterol levels
Increases high‐density lipoprotein cholesterol levels
Decreases oxidative stress
Improves vascular relaxation
Decreases plasminogen activator inhibitor‐1 levels
Increases tissue plasminogen activator activity
Decreases von Willebrand factor levels
Decreases platelet aggregation and adhesion

Metformin Metformin
Lactate
(–)
Glycerol (–) Pyruvate
Ca++
Amino (+) (–)
(–)
acids CO2 ATP
Ca++ ATP+Pi
(–)
OAA

Glucose PEP

Metformin

(+) TK
Increased glycogen Alanine
GLUT 4 storage
FIG 11.1  Hypothesized mechanism of
Glucose Lactate
action of metformin on hepatic
glucose production.
 Are Insulin Sensitizers Useful Additions to Insulin Therapy? 135

Liver (–)
PDH

Acyl CoA Krebs Citrate Pyruvate

β- cycle
oxidation
PFK
FA-CoA (–)
G6P
OAA
CPT Glycolysis
(–) PEP Glucose
Insulin/IGF-1
Glucose receptor

Metformin PO4
n
Glucose
tio
ca
transport GLUT 4
slo

Glucose
n
Tra

Adipose
tissues (–) (+) P85 IRS-1
(–) Pl-3
Kinase P110
(–)
FFA

Muscle Cell

FIG 11.2  Effects of metformin on fatty acid metabolism.

Intriguingly, a relatively recent study intended to explore
the observation that intravenous administration of
metformin seemed less efficacious, utilized a formulation
of metformin designed to target the ileum where met-
formin absorption is low. Despite much lower bioavailabil-
ity, therapeutic efficacy was similar to the equivalent oral
dose of conventional metformin formulations. This would
suggest that actions within the gut e.g. altering entero‐
endocrine secretion are key to metformin’s mechanism(s)
of action and deserve further study [4].
Pharmacokinetics
In normal subjects, studies demonstrate that metformin is
excreted unchanged in the urine and does not undergo
hepatic metabolism or biliary excretion. Because
­ etformin is not metabolized, there is no interference
m
with the metabolism of coadministered drugs. Renal
clearance is approximately 3.5 times greater than creati-
nine clearance, which suggests that tubular secretion is
the main route of metformin elimination. With an oral
administration of metformin, patients with normal
renal function have a plasma half‐life of 2–5 hours, and
almost 90% of an absorbed dose is eliminated within 12
hours [1].
Efficacy of combination therapy
Many people with type 2 diabetes are overweight and insu-
lin‐resistant, making high doses of insulin often necessary
to achieve adequate glucose control. However, insulin ther-
apy is associated with weight gain, which could impede any
 136 Initial Evaluation and Management of Diabetes

Case
A 52‐year‐old woman with obesity and a 9‐year history of type 2 diabetes presents with complaints of fatigue and difficulty
losing weight. She denies polyuria, polydipsia, polyphagia, blurred vision, or vaginal infections.
She states that she has gained an enormous amount of weight since being placed on insulin 6 years ago. Her weight has
continued to increase over the past 5 years, and she is presently at the highest weight she has ever been. She states that
every time she tries to cut down on her eating, she has symptoms of shakiness, diaphoresis, and increased hunger. She
does not follow any specific diet and has been so fearful of hypoglycemia that she often eats extra snacks.
Her health care practitioners have repeatedly advised weight loss and exercise to improve her health status. She
complains that the pain in her knees and ankles makes it difficult to do any exercise.
Her blood glucose values on capillary blood glucose testing have been 170–200 mg/d1 before breakfast. Before supper
and bedtime values range from 150 mg/dl to > 300 mg/dl. Her current insulin regimen is 45 U of NPH plus 10 U of regular
insulin before breakfast and 35 U of NPH plus 20 U of regular insulin before supper. This dose was recently increased after
her HbA1c, was found to be 8.9% (goal < 7.0 %).
Past medical history is remarkable for hypertension, hypertriglyceridemia, and arthritis. Current medications include only
insulin, lisinopril (Prinivil), and hydrochlorthiazide (Dyazide) with triamterene.
On physical exam, her height is 5′ 1½” and her weight is 265 lb. Her blood pressure is 160/88 mmHg. The remainder of
the physical exam is unremarkable.
On laboratory testing, chemistries, BUN, creatinine, and liver function tests are normal. Thyroid function tests and urine
microalburnin are also normal.
After an explanation that the increasing insulin doses were contributing to her weight gain and that she would need to
decrease her insulin dose along with her food intake to prevent hypoglycemia, the patient agreed to follow a restricted‐
calorie diet and to decrease her insulin to 30 U of NPH and 10 U of regular insulin twice daily. As she had no contraindica-
tions to metformin (Glucophage), she was also started on 500 mg orally and it was increased to twice daily after one week.
She returned to the clinic 3 months later, still on the same dose of insulin. She continued to complain of fear of
hypoglycemia in the middle of the night and was overeating at night. Despite this she had lost 7 lb. Her blood glucose
values were still elevated in a range of 120–275 mg/dl before meals.
She was reassured that further insulin reduction would prevent hypoglycemia. Her insulin dosage was decreased to 25 U
of NPH and 5 U of regular insulin twice daily and metformin was increased to 500 mg three times daily. Two months later,
she returned to the clinic with an average blood glucose level of 160 mg/dl. Her weight was now 246 lb, and her HbA1c
was 7.5%. She was feeling much more energetic and was able to start a walking program.

Questions
1. Are insulin sensitizers useful additions to insulin therapy?

progress in achieving glucose control that would normally
be expected. In a double‐blind, placebo‐controlled study
where patients were randomly assigned to receive placebo
or metformin in combination with insulin for 24 weeks,
the data showed that adding metformin to an intensified
insulin regimen resulted in an 11% reduction in hemo-
globin A1c compared to insulin therapy alone. The study
also reported improved glucose control using 29% less
insulin, and a less complicated insulin regimen with no
increase in the occurrence of hypoglycemia or weight
gain [5].
Further studies have verified these findings, like this
randomized trial comparing insulin monotherapy to com-
bined therapy with insulin and metformin, which showed
clear benefit to using metformin with insulin compared to
insulin alone. In this study, insulin as a monotherapy
resulted in a reduction in HbA1c 8.7 ± 1.6 to 7.0 ± 1.0%
with approximately 69% more insulin needed from base-
line to achieve this effect. Patients in this arm of the study
also required a more complicated insulin regimen in
approximately 25% of the cases that became time consum-
ing for both the patient and providers. These patients also
 Are Insulin Sensitizers Useful Additions to Insulin Therapy?
gained 4.4  kg of weight. In the second arm of this study
using insulin in combination with metformin resulted in a
reduction in HbA1c of 8.8 ± 1.2 to 7.1± 1.0%, which was
comparable to the monotherapy arm, however, these
results were achieved without an increase in the total daily
dose of insulin. In this arm, complexity of the insulin regi-
men was not affected, and there was virtually no weight
gain or hypoglycemia reported. The only drawbacks were
that two‐thirds of patients experienced gastrointestinal
side effects, but they were mild and transient [6].
Because studies have shown that decreasing the required
daily dose of exogenous insulin is associated with a
decreased risk of cardiovascular disease, another placebo‐
controlled, randomized, double‐blind trial was designed to
look at effects on glucose control and insulin requirements
when metformin was added in patients with type 2 diabe-
tes intensively treated with insulin. After 16 weeks of com-
bination therapy, the data showed that when metformin
was compared with placebo use, there was a statistically
significant improvement in glucose control shown by a
hemoglobin A1c of 6.9 versus 7.6%. The data also showed
that the use of metformin was associated with less weight
gain (−1.6  kg compared with placebo) and with a small
decrease in LDL cholesterol (– 0.19 mmol/l) [7].
Taken together, these data suggest that metformin can
be an effective adjunct to insulin therapy. Studies have
shown that adding metformin to insulin therapy reduces
hemoglobin A1c, total daily dose of insulin, and reduces or
prevents the weight gain associated with intensive insulin
treatment. However, metformin has gastrointestinal side
effects that can hinder titration to its maximum effective
dose, but these tend to be mild and transient.
Additional indications
Polycystic ovary syndrome
We now understand that the main feature of polycystic
ovary syndrome (PCOS), as it relates to the use of insulin
sensitizers, is the decreased insulin sensitivity and com-
pensatory hyperinsulinemia that is associated with the
condition. Women suffering from PCOS have an increased
risk of diabetes later in life, and an increased incidence of
gestational diabetes if they conceive. The pathophysiology
of the disorder is not fully understood, but the use of an
insulin sensitizer appears to have a profound effect on
137
many of the symptoms. In clinical studies, women with
PCOS being treated with insulin sensitizers like metformin
saw a reduction in circulating insulin levels, a reduction in
androgen levels, improvements in ovarian function, and
improved lipid profiles and severity of hirsutism [8].
Contraindications
Metformin therapy is contraindicated in patients with liver
failure, alcoholism, and active moderate to severe infec-
tion  [1]; these conditions predispose to development of
lactic acidosis, either by increased production or decreased
metabolism of lactic acid [1]. Metformin is also contraindi-
cated in people with kidney disorders (creatinine levels
above 1.4 mg/dl in women and 1.5 mg/dl in men, depend-
ing on lean body mass [1], according to the package insert),
and lung disease. Heart failure has long been considered a
contraindication for metformin use, although a 2007 sys-
tematic review showed metformin to be the only antidia-
betic drug not associated with harm in people with heart
failure [9].
Current recommendations suggest that metformin should
be temporarily discontinued before any radiographic study
involving iodinated contrast, as contrast dye may tempo-
rarily impair kidney function, indirectly leading to lactic
acidosis by causing retention of metformin in the body [10].
Once metformin is withheld, hydration should be main-
tained until preserved kidney function is documented at 24
and 48 hours after the intervention [1]. General anesthesia
should be used cautiously to prevent hypotension, which
leads to renal hypoperfusion and peripheral tissue hypoxia
with subsequent lactate accumulation  [1]. Metformin
should also be used cautiously in the elderly, whose
decreased lean body mass leads to reduced serum creati-
nine concentrations that often mask impaired glomerular
filtration rates [1].
Adverse effects
Lactic acidosis
Lactic acidosis is a life‐threatening complication of
biguanide therapy that carries a mortality rate of
30–50%  [1]. The estimated incidence of metformin‐
associated lactic acidosis is 0.03 cases per 1000 patient‐
years [1], which is 10 to 20 times lower than that seen with
phenformin therapy [1], which was withdrawn because of
 138 Initial Evaluation and Management of Diabetes

an increased risk of lactic acidosis (up to 60 cases per mil-
lion patient‐years). Development of lactic acidosis is almost
always related to coexistent hypoxic conditions that are
probably responsible for the associated high mortality rate.
In one report, 91% of patients who developed lactic acido-
sis while being treated with metformin had a predisposing
condition, such as congestive heart failure, renal insuffi-
ciency, chronic lung disease with hypoxia, or age older than
80 years [1]. Therefore, patients with impaired renal func-
tion or coexistent hypoxic conditions should not be given
metformin. Excessive alcohol consumption may also
potentiate the effect of metformin on lactate metabolism. A
careful history of alcohol use is therefore important before
starting metformin therapy [1].
Gastrointestinal
Side effects of metformin are mostly limited to digestive
tract symptoms, such as diarrhea, flatulence, and abdomi-
nal discomfort [1, 11]. Approximately 5% of patients can-
not tolerate treatment because of gastrointestinal side
effects  [1]. The mechanisms of these side effects remain
unclear but probably are related to accumulation of high
amounts of metformin in the intestinal tissue  [2], with
resultant elevation of local lactate production. These symp-
toms are dose dependent and can usually be avoided by
slow titration and, in some cases, reduction of the dose [1].
Metformin therapy should be initiated with a single dose of
medication (usually 500 mg) taken with the patient’s larg-
est meal to prevent gastrointestinal symptoms. These
symptoms generally disappear within 2  weeks of treat-
ment [12]. Medication doses should be increased in 500‐
mg increments given after meals every 1 to 2 weeks until a
desirable blood glucose level or the maximal effective daily
metformin dose of 2000 mg is reached [13].
Thiazolidinediones
The peroxisome‐proliferator–activated receptors (PPARs)
are a subfamily of super‐receptors that regulate gene
expression in response to ligand binding  [14]. Three
PPARs, identified as PPARα, PPARδ (also known as
PPARβ), and PPARγ, have been discovered to date. PPARs
regulate gene transcription by two mechanisms: transacti-
vation, which is DNA‐dependent and transrepression,
which is DNA‐independent and may explain the anti‐
inflammatory actions of PPARs [14].
PPARα is expressed mainly in the liver, heart, and mus-
cle, as well as in the vascular wall [2]. Fibrates (i.e., fenofi-
brate) act as full or partial PPARα agonists. In general, the
activation of PPARα enhances free fatty acid oxidation,
controls how multiple genes regulate lipoprotein concen-
trations, and has anti‐inflammatory effects (Table  11.2).
Studies have shown that PPARα agonists can retard or even
prevent atherosclerosis in both mice and humans [14].
PPARδ is expressed in several tissue types, with the most
expression in the skin, brain, and adipose tissue. PPARγ is

TABLE 11.2  Molecular targets of PPARγ and PPARα action.

Liver Skeletal muscle Adipose tissue Vascular wall

PPARγ • Decreased • Increased • Increased fatty acid transport protein‐1 • Decreased intercellular
C‐reactive GLUT4 • Increased acyl–coenzyme A synthetase adhesion molecule‐1
protein • Increased • Increased adiponectin • Decreased vascular‐cell
phosphatidyl • Increased LPL adhesion molecule‐1
3‐kinase • Increased phosphatidyl 3‐kinase • Decreased iNOS
• Decreased • Increased GLUT4 • Decreased interleukin‐6
PDK‐4
PPARα • Decreased • Decreased vascular‐cell
C‐reactive adhesion molecule‐1
protein • Decreased cyclooxygenase‐2
• Decreased • Decreased TNFα
fibrinogen B • Decreased interleukin‐6
• Decreased tissue factor
 Are Insulin Sensitizers Useful Additions to Insulin Therapy?
mainly found in adipose tissue but can also be found in
pancreatic beta cells, vascular endothelium, and mac-
rophages [14]. Its expression is low in tissues that express
predominantly PPARα, such as the liver, the heart, and
skeletal muscle. PPARγ was discovered as a target for thia-
zolidinediones only after testing many agents during mul-
tiple large clinical trials.
The first thiazolidinedione, troglitazone, was approved in
1997 for use in treating patients with type 2 diabetes in the
United States. This glucose‐lowering agent was subsequently
withdrawn from the market, in March 2000, because of cases
of hepatotoxicity. In 1999, two γ agonists, rosiglitazone and
pioglitazone, were approved in the United States for use in
treating hyperglycemia in type 2 diabetes.
Mechanism of action
The simulation of PPARγ is considered a major mechanism
through which thiazolidinediones enhance insulin sensi-
tivity. As mentioned previously, PPARγ is mainly found in
adipose tissue and less concentrated in skeletal muscle and
liver. PPARγ is essential for normal adipocyte differentia-
tion and proliferation as well as fatty acid uptake and stor-
age  [14]. Because PPARγ has such a profound effect on
adipose tissue it has been postulated that thiazolidinedi-
ones achieve their insulin‐sensitizing actions through
direct means or indirectly, through altered adipokine
release affecting insulin sensitivity beyond the adipose tis-
sue itself. Basically, thiazolidinediones act directly by pro-
moting fatty acid uptake and storage in adipose tissue
(Figure 11.3). Through this process, there is an increase in
adipose‐tissue mass and other insulin‐sensitive tissues, like
skeletal muscle, the liver, and possibly pancreatic beta cells,
are spared from the harmful effects of high concentrations
of free fatty acids [14]. Therefore, thiazolidinediones keep
fat in its proper place. Data from studies of PPARγ knock-
out mouse models supports the notion that thiazolidinedi-
ones act mainly on adipose tissue as long as normal
amounts of adipose tissue exist [14].
Although thiazolidinediones keep fat in its proper place
(in adipose tissue), there are still indirect processes that
may be responsible for enhancing insulin sensitivity.
Studies have shown that select thiazolidinediones (i.e.,
pioglitazone) regulate the expression of more than 100
genes [9] and some of the established PPARγ target genes
139
Thiazolidinediones
Skeletal muscle Adipose tissue Liver
↑ Glucose uptake Adipogenesis ↓ Gluconeogenesis
↑ Fatty acid uptake
↑ Lipogenesis
↑ Glucose uptake
↓ Plasma free fatty acids
↓ Hyperglycemia
FIG 11.3  Effects of thiazolidinediones.
found in human adipose tissue are listed in Table  11.2.
Several adipokines are regulated by PPARγ but we will
focus mainly on adiponectin, an adiopocytokine produced
exclusively by adipose tissue, because it has been shown to
increase insulin sensitivity in rodents  [14]. Whether
increases in adiponectin increase hepatic insulin sensitivity
in humans is still not clear, only rodents have shown this
effect of PPARγ on adiponectin levels.
Pharmacokinetics
The thiazolidinediones are absorbed completely and rap-
idly, with peak concentrations achieved within 1–2 hours,
but this may be slightly delayed when taken with food.
Both rosiglitazone and pioglitazone are extensively metab-
olized by the liver. The metabolites of rosiglitazone are
weakly active and are excreted mainly in the urine.
Pioglitazone produces metabolites that are more active and
are excreted mostly in the bile. Both thiazolidinediones are
metabolized by the cytochrome P450 system and no clini-
cally significant reduction in plasma concentrations of
other drugs has been reported [16].
Efficacy of combination therapy
Studies have shown that thiazolidinediones used in combi-
nation with insulin therapy result in improved blood glu-
 140 Initial Evaluation and Management of Diabetes
cose control, reduced total daily insulin dose, and reduced
triglyceride levels. In a randomized, double‐blind, placebo‐
controlled study of more than 300  insulin‐treated type 2
diabetic patients, the addition of troglitazone at doses of
200 mg and 600 mg daily resulted in a reduction in HbA1c
level of 0.8 and 1.4%, while insulin dose was reduced by 11
and 29% respectively. On the other hand, subjects taking
insulin and placebo experienced a decrease of 0.1% in
HbAlc level and a 1% increase in insulin dose [17]. Subjects
randomized to placebo therapy gained 1.5 kg, while those
taking the study drug in 200 mg and 600 mg doses in com-
bination with insulin gained 1.9 and 3.6 kg. Total choles-
terol, LDL, and HDL cholesterol levels were found to be
increased with troglitazone use at 600  mg/day. An open‐
label study showed that troglitazone had insulin‐sparing
effects that persisted up to 24  months  [18]. In a rand-
omized, double‐blind, placebo‐controlled study of more
than 200 type 2 diabetic patients who were given troglita-
zone 400 mg daily in combination with their baseline insu-
lin dose, the results showed a significantly greater reduction
in total daily dose of insulin and HbAlc level compared
with those patients using insulin and placebo [19]. In this
study, end points were defined as a 50% reduction in
injected insulin or either a reduction in blood glucose by >
15%, and only 7% of patients on insulin and placebo
achieved these goals, while 22% of patients taking insulin
and troglitazone 400 mg daily achieved these goals.
Studies using rosiglitazone or pioglitazone in combina-
tion with insulin therapy have yielded similar results. In a
randomized control study of inadequately controlled insu-
lin‐treated type 2 diabetes patients, those receiving
26 weeks of treatment with rosiglitazone 4 and 8 mg expe-
rienced a reduction in HbAlc level of 0.6 and 1.2% respec-
tively, as compared with virtually no change in the group
taking insulin and placebo [20]. In this same study, insulin
doses also decreased by 4.8 and 9.4% in the groups taking
insulin plus study drug compared to a decrease of 0.6% in
patients taking insulin and placebo. Total cholesterol,
HDL, and LDL cholesterol levels were found to increase
significantly on treatment with rosiglitazone. Unfortunately,
significant weight gain occurred in the placebo, rosiglita-
zone 4mg, and rosiglitazone 8mg groups with correspond-
ing values of 0.9 kg, 4.0 kg, and 5.3 kg. In a randomized,
placebo‐controlled study in patients receiving stable insu-
lin therapy, when patients assigned to pioglitazone 15 or
30 mg daily in combination with their baseline doses of insu-
lin for 16 weeks, HbAlc levels were reduced by 1.0 and 1.3%
respectively, compared to a reduction of 0.3% in patients on
insulin and placebo  [14]. Those patients on insulin and
pioglitazone gained 2.3 and 3.7 kg, compared with virtually
no change in weight in those using insulin and placebo.
This data suggests that thiazolidinediones can be effec-
tive adjuncts to insulin therapy. A prospective, randomized
trial of cardiovascular outcomes, called Prospective
Pioglitazone Clinical Trial in Macrovascular Events
(PROACTIVE) has looked at coronary and peripheral vas-
cular events as a primary outcome and myocardial infarc-
tion, stroke, and death as secondary outcomes and the data
support pioglitazone as having a favorable effect, especially
on lipids, particularly triglycerides, more so than rosiglita-
zone  [21]. In contrast to the previous study, the
Rosiglitazone Evaluated for Cardiac Outcomes and
Regulation of Glycemia in Diabetes (RECORD) trial, a
long‐term, multicenter, randomized, open‐label study,
found conflicting results when looking at cardiovascular
outcomes in patients with type 2 diabetes treated with
rosiglitazone plus metformin or sulfonylurea, as compared
with the combination of metformin and sulfonylurea. The
data showed that the rate of primary end points (hospitali-
zation or death from cardiovascular causes) was low at
3.1% per year, while secondary end points like acute myo-
cardial infarction, death from cardiovascular causes or any
cause, or the composite of cardiovascular death, myocar-
dial infarction, and stroke showed no statistically signifi-
cant difference between the rosiglitazone group and the
control group [22]. However, the fact remains that TZDs
are still responsible for significant weight gain and
increased peripheral edema.
Additional indications
Nonalcoholic fatty liver disease
Type 2 diabetes has a strong association with nonalcoholic
fatty liver disease (NAFLD), which has a spectrum of liver
damage that ranges from simple fatty liver (steatosis) to
irreversible, advanced scarring of the liver (cirrhosis) [14].
There are an estimated 6.4  million adults in the United
States diagnosed with NAFLD, which is the most common
cause of elevated levels of liver enzymes [14]. Elevated lev-
 Are Insulin Sensitizers Useful Additions to Insulin Therapy?
els of alanine amino‐transferase (ALT) have been shown to
predict type 2 diabetes independently of obesity [14]. Fatty
liver disease is associated with decreased hepatic insulin
sensitivity and correlates with insulin requirements during
insulin therapy in patients with type 2 diabetes [14].
Several recent studies have shown that thiazolidinedi-
ones actually reduce fat accumulation in the liver in
patients with type 2 diabetes as well as in patients with lipo-
dystrophy associated with the use of highly active antiret-
roviral therapy (HAART). Studies have also shown that
liver enzymes actually decrease rather than increase during
treatment with pioglitazone and rosiglitazone [14].
Polycystic ovary syndrome
The polycystic ovary syndrome is a disorder that affects
approximately 4% of women of reproductive age  [14].
Women with PCOS frequently develop insulin resistance
and hence have an increased risk for type 2 diabetes [14].
Hyperinsulinemia, which accompanies insulin resistance,
is thought to contribute to the hyperandrogenism that is
seen in patients with PCOS  [14]. Interventions with the
role of reducing insulin levels, like weight loss and medica-
tions (i.e., metformin), can decrease hyperandrogenism
and reduce insulin resistance [14]. A large‐scale placebo‐
controlled trial of 410 women showed that the use of trogl-
itazone showed significant improvements in ovulatory
function, hirsutism, hyperandrogenism, and insulin resist-
ance  [14]. A more recent, small placebo‐controlled study
that randomized women to either rosiglitazone and pla-
cebo or to rosiglitazone and clomiphene has shown similar
results. This study demonstrated that 56% of women previ-
ously resistant to clomiphene were able to ovulate  [14].
Although metformin is considered safe for women who
become pregnant, rosiglitazone and pioglitazone are classi-
fied as pregnancy category C, which indicates toxic effects
in studies in animal models, but the results in human stud-
ies are inadequate. If these agents are used during preg-
nancy, then the potential benefit must justify the potential
risk to the fetus. Polycystic ovary syndrome is currently not
an approved indication for the use of TZDs.
Lipid lowering
Studies have shown that low‐density lipoprotein (LDL)
cholesterol levels have remained unchanged when mono-
141
therapy with pioglitazone or combination therapy with
pioglitazone and sulfonylurea, metformin, or insulin has
been used. However, some studies have shown increases in
LDL cholesterol levels, between 8 to 16% higher with
rosiglitazone use. Studies have also shown an approximate
10% increase in high‐density lipoprotein (HDL) choles-
terol levels with the use of both drugs. The effects of TZDs
on triglycerides have been more variable, with decreases in
triglyceride levels having been observed more often with
pioglitazone than with rosiglitazone [14]. In a direct com-
parison of rosiglitazone and pioglitazone, one study of 127
patients previously treated with troglitazone, supports the
notion that both drugs have similar effects on glucose lev-
els and body weight [14]. This same study supported the
notion that pioglitazone is more effective than rosiglita-
zone at lowering LDL cholesterol and serum triglyceride
levels. The difference in efficacy of these two drugs on
lipids cannot be attributed to the effect they have on serum
free fatty acid concentrations, which decreases by approxi-
mately 20 to 30% in both [14]. Pioglitazone appears to be a
partial agonist of PPARα in  vitro;, and rosiglitazone
appears to be a pure PPARγ agonist [14]. So far, however,
the data on mechanisms underlying the effects of the TZDs
on lipids in humans is quite limited.
Lipodystrophies
The most common form of lipodystrophy is that associated
with highly active antiretroviral therapy use in patients
with human immunodeficiency virus (HIV) disease. After
only 12 to 18  months of therapy with HAART, approxi-
mately half of patients develop a lipodystrophy‐related
symptom like facial lipoatrophy  [14]. Facial lipoatrophy
can be disfiguring and stigmatizing, especially since there
is no pharmacologic therapy for this condition, which is
usually accompanied by marked insulin resistance.
Thiazolidinediones would seem to be a wonderful solution
to insulin resistance and lipoatrophy caused by HAART
because these drugs increase both insulin sensitivity and
subcutaneous fat mass. Unfortunately, there has been only
one placebo‐controlled trial in which patients with
HAART‐associated lipodystrophy were treated with rosigl-
itazone 8  mg per day for six months, and there was no
increase in adipose tissue or body weight, in contrast to
studies in patients with type 2 diabetes [14].
 142 Initial Evaluation and Management of Diabetes
Carotid intima‐media thickness
Carotid IMT is a well‐established surrogate marker for car-
diovascular risk. A thickened carotid intima‐media layer
not only correlates with increased cardiovascular risk but
also with the risk of future macrovascular events like myo-
cardial infarction and stroke  [23]. Several studies using
multiple agents (i.e., ACE inhibitors, calcium channel
blockers, β‐blockers, and statins) have shown a reduction
or even regression of carotid IMT in patients without dia-
betes [23]. However, carotid IMT appears to be more sig-
nificant in patients with type 2 diabetes reflecting a more
dramatic cardiovascular risk in this patient popula-
tion [23]. Limited data currently exists about the effect of
intervention in type 2 diabetes on carotid IMT. However, a
recent randomized control study of 192 patients showed
that treatment with pioglitazone for 24 weeks led to a sig-
nificant decrease in carotid IMT in patients with type 2
diabetes, and this was found to be independent of glucose
control [23].
Contraindications
In patients with diabetes, hypertension and coronary artery
disease occur frequently. These conditions are risk factors
for the development of congestive heart failure (CHF) [24].
Diabetes can affect cardiac structure and systolic or dias-
tolic function, independent of other established risk factors
for CHF because of diabetic cardiomyopathy  [24]. This
phenomenon makes diabetes an independent risk factor
for CHF, which was supported by an analysis of 9591 peo-
ple with type 2 diabetes in the Kaiser Permanente
Northwest Division that demonstrated that 11.8% of dia-
betic subjects had CHF at baseline, and an additional 7.7%
developed CHF during a 30‐month follow‐up period [24].
This suggests that CHF may be present prior to physicians
beginning therapy with TZDs or that this condition may
develop during the course of treatment.
TZDs can still be used in patients with underlying
asymptomatic heart disease, although its safety has not
been fully established. The package inserts for both rosigli-
tazone and pioglitazone indicate that patients with more
advanced heart disease (NYHA class III or IV) were
excluded in pre‐marketing clinical trials, and hence, these
drugs are not recommended in such patients. Currently,
there are no guidelines on the use of TZDs in patients with
diabetes who have any degree of heart disease or for those
already on a TZD who develop CHF. What makes this clin-
ical dilemma more perplexing is the more common side
effect of peripheral edema associated with TZDs, which
makes the origin of the development of edema or weight
gain more difficult to decipher.
Physicians should be cautious before prescribing TZDs
to patients with diabetes who have been previously diag-
nosed with a condition that increases risk of bone fractures
(i.e., osteoporosis, hyperparathyroidism, Paget’s disease,
etc.). Recent studies have shown a correlation between
TZD use and an increased risk of bone fractures. In fact,
according to data taken from the ADOPT study group and
others, the relative risk of fractures with thiazolidinediones
remained consistently elevated irrespective of age or meno-
pausal status of women  [25]. In addition, a recent meta‐
analysis showed that the long‐term use of thiazolidinediones
(rosiglitazone, pioglitazone or troglitazone) doubles the
risk of fractures among women with type 2 diabetes, and
the overall use of thiazolidinediones significantly increased
the risk of fractures among patients with type 2 diabetes.
Also identified in this meta‐analysis was that thiazolidine-
dione use was also associated with significant changes in
bone mineral density at the lumbar spine and the hip [25].
Adverse effects
Multiple studies have correlated weight gain and glyco-
sylated hemoglobin with thiazolidinedione use – it appears
that TZDs lead to an increase in body weight of 2 to 3 kg for
every 1% decrease in glycosylated hemoglobin values. This
increase in body weight is the same irrespective of TZD use
as a monotherapy or in combination with insulin or met-
formin in patients with type 2 diabetes. One of the mecha-
nisms proposed to explain this phenomenon is that the
increase in body weight is attributed to expansion of the
subcutaneous fat depot, and in some patients to edema,
whereas the mass of visceral fat remains unchanged or even
decreases [14].
The use of TZDs is not only associated with weight gain,
but some patients experience fluid retention and plasma
volume expansion, which lead to the development of
peripheral edema. The development of peripheral edema
 Are Insulin Sensitizers Useful Additions to Insulin Therapy?
has been reported in 4 to 6% of patients undergoing treat-
ment with TZDs as compared with 1 to 2% of those receiv-
ing placebo or other hypoglycemic therapies. Edema
development appears to occur most when either of the
TZDs is used in combination with insulin. Studies have
shown that the use of rosiglitazone 4 or 8  mg per day in
combination with insulin was associated with a 13.1% and
16.2% incidence of edema, respectively, compared with
4.7% in those taking insulin alone [20], and with pioglita-
zone used at 15  mg or 30  mg daily in combination with
insulin resulted in a combined 15.3% incidence of edema,
compared with 7.0% for insulin alone [24]. Edema appears
to occur at a higher incidence when either of the TZDs is
combined with insulin, as opposed to when TZDs are used
in combination therapy with other oral hypoglycemic
agents. This increase in body weight and edema has been
associated with an increase in the incidence of congestive
heart failure in patients treated with TZDs and insulin.
As a result, the Food and Drug Administration added a
warning in the drug packet information for those
patients taking rosiglitazone and pioglitazone. The
European Agency for the Evaluation of Medicinal
Products actually considers insulin therapy a contraindi-
cation to the use of TZDs. According to the data this
agency presents, the frequency of CHF was 2.5 times
greater in combination therapy of insulin and thiazoli-
dinediones than with insulin alone. However, the cause
for this result remains unclear.
In a meta‐analysis of 42 trials containing more than 27
000 patients, over 15 000 of them on treatment with rosigli-
tazone, the data showed an overall odds ratio of 1.43 for
myocardial infarction (MI) and 1.64 for death from cardio-
vascular cause. From this meta‐analysis, compared with
placebo or with other hypoglycemic agents, treatment
with rosiglitazone was associated with a significant increase
in the risk of MI and with an increase in the risk of death
from  cardiovascular causes that was of borderline
significance [15].
A recent study suggests an elevated risk of fracture asso-
ciated with TZDs compared with other oral hypoglycemic
agents. The risk of fracture with TZD use was present at
multiple anatomic sites and these results were potentially
clinically significant [26]. This suggests that caution should
143
be taken when considering treatment of type 2 diabetic
patients with TZDs who have an increased risk for
fractures.
Discussion
So the question remains, “Are insulin sensitizers useful
additions to insulin therapy?” Currently, the data suggest
that metformin is a feasible option in all patients with type
2 diabetes and should be initiated at the onset of insulin
therapy unless a specific contraindication to metformin
use exists. The benefits of metformin therapy in these
patients is an improvement in hemoglobin A1c, a reduc-
tion in total daily dose of insulin, and most importantly, a
reduction in weight gain despite intensive insulin regi-
mens [5–7]. The only potential disadvantage to metformin
therapy is its side effect profile. This medication is con-
traindicated in patients with impaired renal function due
to risk of lactic acidosis, but renal impairment can be a
common condition in those with a prolonged diabetes
course [1, 10]. The other disadvantage to metformin ther-
apy relates to its gastrointestinal side effects which may
impede titration to its maximal effective dose. However,
these side effects are often mild and transient, and usually
can be avoided by slow titration, and in some cases, reduc-
tion of the dose [1]. Therefore, metformin appears to be an
appropriate choice when considering adding an inexpen-
sive insulin sensitizer with few side effects to all patients
with type 2 diabetes currently on insulin therapy or plan-
ning its initiation.
Similarly, studies have shown that thiazolidinediones
used in combination with insulin therapy result in
improved blood glucose control, reduced total daily insulin
dose, and reduced triglyceride levels [17–20]. However, the
side effects of these agents can prove to significantly reduce
any benefit that may have been gained from their use. In
some studies, rosiglitazone has been associated with a sig-
nificant increase in the risk of MI and with an increase in
the risk of death from cardiovascular causes  [15].
Rosiglitazone has also been found to significantly increase
total cholesterol, HDL, and LDL cholesterol levels  [14].
This data suggests that extreme caution should be used
when treating patients with this medication, and likely
 144 Initial Evaluation and Management of Diabetes
even avoided. In addition, recent studies have suggested
that TZDs have an elevated risk of bone fracture at multiple
anatomic sites compared with other oral hypoglycemic
agents [26]. Studies of pioglitazone have shown more favora-
ble outcomes than rosiglitazone with less risk of myocardial
infarction and other cardiovascular‐related deaths, as well as
a better lipid profile response, especially with triglycer-
ides  [15, 21]. However, both medications cause significant
weight gain and peripheral edema [14, 20].
With dual therapy including insulin plus metformin or
insulin plus thiazolidinediones, significant reductions have
been shown in hemoglobin A1c, total daily dose of insulin,
and weight gain. These results would suggest that triple
therapy including insulin plus metformin and TZDs would
produce an even more profound effect, and select studies
have considered these combinations. For example, in this
randomized study of 28 type 2 patients with diabetes using
insulin monotherapy, 4 months of triple therapy was initi-
ated by adding metformin to insulin, then troglitazone
(TGZ) versus adding TGZ to insulin, then metformin.
Researchers found that hemoglobin A1c decreased with
dual therapy but improved more during triple therapy
(insulin + metformin 7.0%, insulin + TGZ 6.2%; insulin +
metformin, adding TGZ 6.1%; insulin + TGZ, adding met-
formin 5.8%). Total daily dose of insulin was significantly
reduced in the insulin + TGZ group (−14.1 units), insulin
+ TGZ, adding metformin (−13.7 units), and the insulin +
metformin, adding TGZ (−17.3 units), but not in the insulin
+ metformin group (−3.2 units). However, patients in the
insulin + TGZ group experienced significant weight gain
(4.4  kg), and those in the insulin + metformin, insulin +
metformin, adding TGZ, and insulin + TGZ; adding met-
formin saw no weight gain. Thus, the order in which met-
formin and TZDs are added to the treatment regimen may
be important, especially as it relates to weight gain. Adding a
TZD to a patient who is already on metformin seems to pro-
tect the patient from TZD‐induced weight gain [27].
Ultimately, metformin is an excellent choice for all indi-
viduals with type 2 diabetes seeking to reach glucose and
weight targets, and perhaps should be used in combination
with insulin unless there is some specific contraindication
to its use. Thiazolidinediones, however, appear to be more
problematic.
References
1. Krentz AJ, Bailey CJ. Oral antidiabetic agents current role in
type 2 diabetes mellitus. Drugs. 2005;65.
2. Klip A, Leiter LA. Cellular mechanism of action of met-
formin. Diabetes Care. 1990;13(6):696–704.
3. Kirpichnikov D, McFarlane SI, Sowers JR. Metformin: an
update. Ann Intern Med. 2002;137:25–33.
4. Buse J B. et al. The primary glucose‐lowering effect of met-
formin resides in the gut, not the circulation: results from
short‐term pharmacokinetic and 12‐week dose‐ranging
studies. Diabetes Care. 2016;39(2):198–205.
5. Aviles‐Santa ML, Sinding J, Raskin P. Effects of metformin
in patients with poorly controlled, insulin‐treated type 2 dia-
betes mellitus. Ann Intern Med. 1999;131:182–188.
6. Strowig SM, Aviles‐Santa ML, Raskin P. Comparison of
insulin monotherapy and combination therapy with insulin
and metformin or insulin and troglitazone in type 2 diabe-
tes. Diabetes Care. 2002;25:1691–1698.
7. Strowig SM, Raskin P. Combination therapy using met-
formin or thiazolidinediones and insulin in the treatment of
diabetes mellitus. Diabetes Obes Metabol. 2005;7:633–641.
8. Muth S, Norman J, Sattar N, Fleming R. Women with poly-
cystic ovary syndrome (PCOS) often undergo protracted
treatment with metformin and are disinclined to stop: indi-
cations for a change in licensing arrangements? Hum Reprod.
2004;19:2718–2720.
9. Eurich DT, McAlister FA, Blackburn DF et al. Benefits and
harms of antidiabetic agents in patients with diabetes and
heart failure: systematic review. BMJ. 2007;335:497.
10. Weir J. Guidelines with regard to metformin‐induced lactic
acidosis and X‐ray contrast medium agents. R Coll Radiol.
1999;54:29–33.
11. Garber AJ, Duncan TG, Goodman AM, Mills DJ, Rohlf JL.
Efficacy of metformin in type II diabetes: results of a double‐
blind, placebo‐controlled, dose response trial. Am J Med.
1997;103:491–497.
12. Haupt E, Knick B, Koschinsky T, Liebermeister H, Schneider J,
Hirche H. Oral antidiabetic combination therapy with sul-
phonylureas and metformin. Diabetes Metab. 1991;17:
224–231.
13. DeFronzo RA, Goodman AM. Efficacy of metformin in
patients with noninsulin‐dependent diabetes mellitus. The
Multicenter Metformin Study Group. NEJM. 1995;333:
541–549.
14. Yki‐Jarvinen H. Thiazolidinediones. NEJM. 2004;351:
1106–1118.
 Are Insulin Sensitizers Useful Additions to Insulin Therapy?
15. Nissen SE, Nissen SE, Wolski K. Effect of rosiglitazone on
the risk of myocardial infarction and death from cardiovas-
cular causes. NEJM. 2007;356:2457–2471.
16. Baldwin SJ, Clarke SE, Chenery RJ. Characterisation of the
cytochrome P450 enzymes involved in the in vitro; metabo-
lism of rosiglitazone. Br J Clin Pharmacol. 1999;48: 424–432.
17. Schwartz S, Raskin P, Fonseca V, Graveline JF. Effect of tro-
glitazone in insulin‐treated patients with type II diabetes
mellitus. NEJM. 1998;338:861–866.
18. Fonseca V, Foyt HL, Shen K, Whitcomb R. Long‐term effects
of troglitazone: open‐label extension studies in type 2 dia-
betic patients. Diabetes Care. 2000;23:354–359.
19. Buse JB, Gumbiner B, Mathias NP, Nelson DM, Faja BW,
Whitcomb RW. Troglitazone use in insulin‐treated type 2
diabetic patients. The Troglitazone Insulin Study Group.
Diabetes Care. 1998;21:1455–1461.
20. Raskin P, Rendell M, Riddle MC, Dole JF, Freed MI,
Rosenstock J. A randomized trial of rosiglitazone therapy in
patients with inadequately controlled insulin‐treated type 2
diabetes. Diabetes Care. 2001;24:1226–1232.
21. Dormandy JA, Charbonnel B, Eckland DJ et al. Secondary
prevention of macrovascular events in patients with type 2
diabetes in the PROactive Study (PROspective pioglitAzone
145
Clinical Trial In macroVascular Events): a randomized con-
trolled trial. Lancet. 2005;366:1279–1289.
22. Home PD, Pocock SJ, Beck‐Nielson H et  al. Rosiglitazone
evaluated for cardiovascular outcomes – an interim analysis.
NEJM. 2007;357:28–38.
23. Langenfeld MR, Forst T, Hohberg C et  al. Pioglitazone
decreases carotid intima‐media thickness independently of
glycemic control in patients with type 2 diabetes mellitus:
results from a controlled randomized study. Circulation.
2005;111:2525–2531.
24. Nesto RW, Bell D, Bonow RO et al. Thiazolidinedione use,
fluid retention, and congestive heart failure. Diabetes Care.
2004;27(1):256–263.
25. Loke YK, Singh S, Furberg CD. Long‐term use of thiazoli-
dinediones and fractures in type 2 diabetes: a meta‐analysis.
CMAJ. 2009;180:32–39.
26. Solomon DH, Cadarette SM, Choudhry NK, Canning C,
Levin R, Sturmer T. A cohort study of thiazolidinediones
and fractures in older adults with diabetes. JCEM.
2009;94:2792–2798.
27. Strowig SM, Aviles‐Santa ML, Raskin P. Improved glycemic
control without weight gain using triple therapy in type 2
diabetes. Diabetes Care. 2004;27:1577–1583.
 12 Incretin‐Based Therapy for the
Management of Type 2 Diabetes
Kristin Gonzales1 and Adrian Vella2
1
Department of Internal Medicine, Division of Endocrinology, Diabetes, and Metabolism, University of New Mexico
Health Sciences Center, Albuquerque, NM, USA
2
Professor of Medicine, Division of Endocrinology, Diabetes, Metabolism, and Nutrition, Mayo Clinic, Rochester, MN, USA

LE A R N I N G P OI NT S production (EGP). In the fasting state, reduced plasma glu­

●● Incretin‐based therapy comprises Dipeptidyl
Peptidase‐4 inhibitors (DPP‐4i) which raise concentra-
tions of endogenous incretin hormones or Glucagon‐
Like Peptide‐1 (GLP‐1) Receptor Agonists which
activate the endogenous GLP‐1 receptor.
●● While DPP‐4i are weight‐neutral, GLP‐1 receptor
agonists induce (some) weight loss.
●● GLP‐1 receptor agonists have also been associated
with cardiovascular benefit.
Introduction
Diabetes is characterized by an absolute or relative insulin
deficiency. Type 2 diabetes is a heterogeneous disorder
resulting from dysregulation of multiple processes involved
in glucose homeostasis. Under normal conditions, glucose
homeostasis is tightly maintained through a combination
of nutrient sensing and hormonal signaling to balance
peripheral glucose utilization and production. In the fed
state, elevated plasma glucose concentrations, which are
factors of gastric emptying and systemic glucose appear­
ance, suppress glucagon secretion, stimulate insulin release,
and directly stimulate hepatic glucose uptake for storage as
glycogen. Insulin serves as the main regulatory hormone in
this state to inhibit hepatic gluconeogenesis and glycogen­
olysis, the main processes involved in endogenous glucose
cose concentrations signal the release of glucagon, inhibit
insulin secretion, and promote hepatic EGP. In order for
glucose concentrations to be maintained within a strict
physiologic range during each of these states, the rate of
EGP must match that of glucose utilization by peripheral
tissues. Thus, normal glucose tolerance is dependent upon
five key factors: (1) insulin secretion, (2) glucagon suppres­
sion, (3) insulin action i.e. the ability of insulin to stimulate
glucose uptake and suppress EGP, (4) glucose effectiveness
i.e. the ability of glucose to stimulate its own uptake, and
suppress EGP, and (5) gastric emptying.
In diabetes or states of impaired glucose tolerance,
defects in some or all of these parameters can result in dys­
regulation of glucose homeostasis. This leads to chronic
elevations in plasma glucose concentrations and predis­
poses to systemic vascular complications [1].
The incretin effect
Under states of normal glucose tolerance, the insulin secre­
tory response to an oral glucose load is robust, and the pat­
tern of insulin secretion differs substantially from that
observed following intravenous glucose infusions. During
intravenous glucose challenges, insulin release is biphasic,
with an initial rapid release followed by a more steady
release over several hours [2]. This biphasic response has
been observed to be blunted in patients with diabetes dur­
ing intravenous glucose challenges [3].

Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.

© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.

146
 Incretin-Based Therapy for the Management of Type 2 Diabetes
In contrast, during an oral glucose challenge, patients
with normal glucose tolerance demonstrate a substantially
greater insulin response compared to those seen during
intravenous glucose challenges, despite similar plasma glu­
cose concentrations  [4]. This differential in insulin
responses is due to a phenomenon known as the incretin
effect, a process mediated by both hormonal and neural gut
stimuli. The hormonal stimulus for this phenomenon orig­
inates from endogenous incretins – enteroendocrine hor­
mones secreted by K and L cells of the intestine in response
to meal ingestion. They exert direct regulatory actions on
insulin and glucagon secretion, as well as on gastric empty­
ing and satiety. Glucagon‐like peptide‐1 (GLP‐1) and glu­
cose‐dependent insulinotropic polypeptide (GIP) are the
main incretin hormones implicated in these processes and
exert their insulinotropic and glucagon‐suppressive actions
in a glucose‐dependent fashion [5, 6]. Among patients with
impaired glucose tolerance and diabetes, the insulin‐secre­
tory and glucagon‐suppressive effects of incretins are nota­
bly impaired [7].
Several hypotheses have been proposed regarding the
mechanisms for these diminished incretin effects, as well
as the role of incretins in development of diabetes.
Although some studies suggest alterations in secretion of
GIP and GLP‐1 in response to glucose loads, there remains
to be conclusive evidence implicating changes to endoge­
nous incretin hormone secretion in the pathogenesis of
type 2 diabetes [8, 9].
Due to the insulinotropic and glucagon‐lowering effects
of endogenous GIP and GLP‐1, research efforts have
sought to determine their relevance in the clinical manage­
ment of diabetes. Pharmacologic doses of GIP have dem­
onstrated little effect on diabetic islets with regard to
insulin secretion and glucagon suppression, although
pharmacologic GLP‐1  has shown substantial promise in
patients with type 2 diabetes [10]. In 1993, Nauck and col­
leagues demonstrated that pharmacologic doses of syn­
thetic GLP‐1 (7‐36 amide), when administered via
continuous infusion, lowered plasma glucose and glucagon
concentrations, and raised insulin concentrations in a glu­
cose‐mediated fashion, among patients with type 2 diabe­
tes. Similar to native GLP‐1, however, these early synthetic
forms of GLP‐1 underwent rapid degradation by dipepti­
dyl peptidase‐4 (DPP‐4). This enzyme cleaves the N‐ter­
147
minal dipeptides His7‐Ala8 of GLP‐1, thus rendering it
inactive. The result of this rapid degradation is a biological
half‐life of 2  minutes, a limiting factor when considering
GLP‐1 for use in clinical pharmacology [11].
Early development of GLP‐1 RA therapy
The limitations regarding use of early GLP‐1 analogs were
circumvented following the discovery of Exendin‐4, or
exenatide. Exenatide is a 39‐amino acid polypeptide dis­
covered in nature as having 53% sequence homology with
GLP‐1 and demonstrating similar binding capability to the
GLP‐1 receptor. However, due to alterations in its amino
acid chain with substitution of an alanine residue for gly­
cine at position 8, exenatide is not susceptible to DPP‐4
degradation. This results in a relatively prolonged half‐life
of 3.3 to 4 hours and enables intermittent rather than con­
tinuous administration [12].
Early studies in diabetic mice demonstrated that once
daily administration of intravenous exenatide resulted in a
significant insulinotropic and glucagon‐lowering effect, as
did twice daily subcutaneous injections of exenatide.
Similar findings were observed in humans with type 2 dia­
betes who received intravenous exenatide during hyper­
glycemic insulin clamps [13–16]. Between 2003 and 2005,
randomized controlled trials (RCTs) found that exenatide
could be feasibly and effectively administered as a twice‐
daily subcutaneous injection to patients with type 2, and it
could be safely added to metformin for glucose‐lowering
and weight reduction  [17, 18]. Thus, immediate‐release
exenatide was the first GLP‐1 receptor agonist (GLP‐1
RA) to be FDA‐approved for use in type 2 diabetes in
2005. In 2015, an extended‐release formulation requiring
only once weekly subcutaneous administration was
approved [19].
To date, multiple other GLP‐1 RAs have been approved
for use in type 2 diabetes. These include liraglutide, lixi­
senatide, dulaglutide, and semaglutide. Albiglutide has
received FDA approval but is not currently marketed for
use. These agents share sequence homology ranging from
53–97% with native GLP‐1. Each is administered as subcu­
taneous injections and available in pre‐filled pens  [20].
Oral semaglutide received FDA‐approval in 2019, making
it the first oral GLP‐1 RA available in its class [21].
 148 Initial Evaluation and Management of Diabetes
Clinical benefits and side effects of
GLP‐1 RAs
GLP‐1 RAs share common features unique to this class
including HbA1c‐lowering and weight loss. GLP‐1 RAs are
known to suppress appetite and reduce gastric emptying,
although there is tachyphylaxis to the gastric‐emptying effects,
suggesting that weight loss is independent of this process.
Table  12..1 delineates the ranges of weight loss and HbA1c
reduction observed in head‐to‐head trials of GLP‐1 RAs.
In addition to the insulinotropic and glucagon‐lowering
effects of GLP‐1 RAs, cell and animal models have impli­
cated these agents in the trophic regulation of beta cells.
Exenatide has been shown to augment beta cell mass, pro­
mote beta cell differentiation, and to inhibit beta cell apop­
tosis. These data would collectively imply a potential role
for incretin mimetics in attenuation of disease progression.
However, human studies have yet to demonstrate any sig­
nificant effect in this regard [12, 22, 23].
Although data is limited for use of GLP‐1 RAs in the
setting of mild to moderate chronic kidney disease (CKD),
several RCTs have shown them to be safe and effective in
these settings [24]. The majority of data pertaining to use
of these agents in CKD is derived from large cardiovascular
outcome trials (CVOTs) wherein renal outcomes were
examined as secondary endpoints. In these trials, signifi­
cant improvements in composite renal endpoints were
noted with liraglutide, semaglutide, dulaglutide, extended‐
release exenatide, and lixisenatide; however, these improve­
ments were largely driven by reductions in new‐onset
macroalbuminuria [25–29]. The extent of impact on other
renal outcomes such as progression to end‐stage renal dis­
ease (ESRD), progressive microalbuminuria, or change in
estimated glomerular filtration rate (eGFR) remains
unclear, as many of these trial did not demonstrate sub­
stantial improvement in these endpoints. There have been
some reports of acute renal impairment in association with
GLP‐1 RAs, although these findings are likely confounded
by underlying conditions such as dehydration due to the
gastrointestinal (GI) side effects of GLP‐1 RAs [30].
The most common adverse events associated with
GLP‐1 RAs are nausea and vomiting, both effects of their
actions on gastric emptying and the paraventricular
nucleus of the hypothalamus. Unlike insulin and insulin
secretagogues, GLP‐1 RAs rarely cause hypoglycemia, as
their influence on beta cell insulin release is largely glu­
cose‐dependent. However, increased incidence of hypogly­
cemia has been noted in trials employing combined use of
these agents with sulfonylureas. Thus, caution is recom­
mended when initiating GLP‐1 RAs in the setting of con­
current insulin or sulfonylurea use.
In addition to the common GI side effects of GLP‐1
RAs, preclinical safety studies in rats have implicated these
agents in the development of c‐cell hyperplasia, c‐cell ade­
nomas, and medullary thyroid carcinoma. Although the
relevance of these findings to humans remains unclear, the
FDA has issued a black‐box warning on all GLP‐1 RAs cau­
tioning against use in patients who have a personal or fam­
ily history of medullary thyroid cancer or multiple
endocrine neoplasia 2 (MEN 2)  [23]. Notably, in large
RCTs involving GLP‐1 RAs in humans, there have been no
documented cases of medullary thyroid cancer related to
pharmacological use of these agents.
Post‐marketing reports have also implicated incretin
mimetics in the development of acute pancreatitis, prompt­
ing the FDA to issue warning labels regarding this potential
risk [23]. Significant controversy has also emerged regard­
ing the risk for dysplastic pancreatic lesions in patients
treated with GLP‐1 RAs. However, diabetes, itself, is asso­
ciated with a 2‐fold increased risk of acute pancreatitis and
a higher baseline risk of pancreatic cancer compared to the
general population. These factors preclude causal associa­
tions from being made between clinically‐significant pan­
creatic pathology and GLP‐1 RAs. Furthermore, adverse
event reporting in large‐scale RCTs have not reported sig­
nificant increases in these outcomes among patients treated
with GLP‐1 RAs or DPP‐4  inhibitors. Yet, these associa­
tions should also be factored into clinical decision‐making
when considering use of GLP‐1 RAs [23, 25–29, 31–34].
The evidence for GLP‐1 ras in
management of type 2 diabetes
Exenatide
Exenatide is available in both immediate and extended‐
release formulations, allowing for either twice daily or once
weekly administration. Dosing options for immediate‐
release exenatide are 5μg and 10 μg, and extended‐release
 Incretin-Based Therapy for the Management of Type 2 Diabetes 149

exenatide is available as a 2mg dose. In RCTs with sulfony­
lureas, metformin, thiazolinediones (TZDs), and insulin as
comparators, immediate‐release exenatide was shown to
lower HbA1c by approximately 0.4–1.0% and to reduce
weight by 1.6–2.8  kg among patients with type 2 diabe­
tes [17, 35–38]. Extended‐release exenatide, however, dem­
onstrated a greater impact on HbA1c‐lowering of 1.4% vs.
1.0% when compared in a head‐to‐head fashion with
immediate‐release exenatide. Weight loss was similar
between groups receiving the two drugs, although GI side
effects were less common with extended‐release exenatide.
The only reported events of hypoglycemia occurred in
patients receiving GLP‐1 RA therapy in combination with
sulfonylureas [39].
Exenatide has also been studied as add‐on therapy to
basal insulin. One RCT of immediate‐release exenatide
added to basal insulin, with or without metformin and/or
pioglitazone, demonstrated an average HbA1c reduction of
1.74% in the exenatide‐treated group compared to a 1.04%
reduction in the group managed without exenatide.
Average weight loss was 1.8  kg in the exenatide group,
whereas a weight gain of 1.0 kg occurred in the comparator
group. There were no significant differences in hypoglyce­
mia  [40]. Similarly, an RCT examining the efficacy of
extended‐release exenatide added to insulin glargine with
or without metformin and/or a sulfonylurea demonstrated
an average decline in HbA1c of 0.96% vs. 0.23%. Weight
also declined significantly in the exenatide‐treated group
by 1.04  kg vs. 0.48  kg. However, more individuals in the
weekly exenatide group experienced GI‐related adverse
events prompting withdrawal from the study (3.9% vs.
1.7%, respectively). Major hypoglycemic events did not
differ between groups, although more minor episodes of
hypoglycemia occurred in the placebo [41]. These findings
indicated a pertinent role for GLP‐1 RAs as adjunct ther­
apy to basal insulin in patients with suboptimal glycemic
control.
Liraglutide
Liraglutide was the second FDA‐approved GLP‐1 RA and
is available in three dosing regimens: 0.6 mg daily, 1.2 mg
daily, and 1.8 mg daily. Standard practice is to initiate ther­
apy at the lowest dose and gradually escalate to the maxi­
mally‐tolerated dose. In a series of clinical trials conducted
to examine the efficacy of liraglutide either as mono‐ or
combination therapy with sulfonylureas, TZDs, or met­
formin, liraglutide reduced HbA1c on average by 0.8–1.5%.
When compared directly to immediate‐release exenatide,
liraglutide 1.8mg daily lowered HbA1c by a mean of 1.12%
vs. 0.79%. Weight loss was similar between the two
groups [42]. Similarly, when compared head‐to‐head with
extended‐release exenatide, liraglutide demonstrated sig­
nificantly greater HbA1c and weight reduction (1.48% vs.
1.28% and 3.57 kg vs. 2.68 kg, respectively) [43].
Several studies have also examined the utility of liraglu­
tide as add‐on therapy to basal insulin. One such RCT ana­
lyzed the effect of liraglutide added to insulin degludec.
The results noted an HbA1c reduction of 1.9% in the group
treated with liraglutide vs. 0.9% in the comparator group,
along with significant reductions in overall 9‐point self‐
monitored blood glucose values. Average weight loss of
2.7  kg occurred in the liraglutide‐treated group vs. no

TABLE 12.1  GLP‐1 RAs: effects on glucose and weight in head‐to‐head trials.

Immediate‐ Extended‐
release release
exenatide [42, exenatide [43, Liraglutide Dulaglutide Semaglutide
Drug 45, 48, 82, 83] 52, 82, 83] [42, 43, 49, 84] Lixisenatide [45] [48, 49, 53] [52, 53]

Average A1c 0.79–1.5 0.9–1.9 0.99–1.48 0.79 1.42–1.51 1.5–1.8

reduction (%)*
Average weight 1.24–3.8 1.9–3.7 2.91–3.6 2.8 1.3–2.9 5.6–6.5
loss (kg)*

 Average HbA1c and weight reductions reported with maximum doses of respective GLP‐1 RAs
*
 150 Initial Evaluation and Management of Diabetes
weight loss in the comparator group [44]. These findings
paved the way for use of long‐acting GLP‐1 RAs such as
liraglutide in combination with basal insulin.
Lixisenatide
Lixisenatide is available as a once‐daily injection in doses
of 10 μg and 20 μg. When compared head‐to‐head with
immediate‐release exenatide, lixisenatide proved to be
non‐inferior with regard to reduction in HbA1c  [45].
When compared to liraglutide as add‐on therapy to met­
formin, liraglutide lowered HbA1c significantly by 1.8%
compared to 1.2% with lixisenatide. Weight loss effects of
each were not statistically different [46].
One RCT compared lixisenatide to rapid‐acting insulin
glusiline, in combination with basal insulin glargine. This
study demonstrated a significant reduction in 2‐hour post­
prandial glucose concentrations in the lixisenatide group
compared to that receiving insulin glusiline three times
daily. These findings further supported earlier data dem­
onstrating that GLP‐1 RAs could be effectively added to
basal insulin regimens as a safe and viable replacement for
bolus meal‐time insulin in select patient populations [47].
Dulaglutide
Dulaglutide is available as a weekly GLP‐1 RA in doses of
0.75 mg or 1.5mg. When compared to immediate‐release
exenatide with metformin and pioglitazone, HbA1c
decreased by 1.51% and 1.3% with 1.5  mg and 0.75  mg,
respectively vs. 0.99% with exenatide. Weight reductions
were similar between groups in this study [48]. In a non‐
inferiority RCT comparing dulaglutide with liraglutide,
HbA1c reductions were comparable (1.42% vs. 1.36 %), but
significantly greater weight reduction was noted with lira­
glutide vs. dulaglutide (3.6 kg vs. 2.9 kg) [49].
Dulaglutide 1.5mg weekly was studied as add‐on ther­
apy to basal insulin glargine alone in a prospective RCT to
determine its effect on HbA1c longitudinally over
28 weeks. During this time, HbA1c declined by 1.44% in
the dulaglutide group vs. 0.67% in the comparator group.
Weight declined by 1.91 kg with dulaglutide vs. a gain of
0.5  kg in the comparator group. Additionally, hypoglyce­
mia rates were similar between the two, although more
patients treated with dulaglutide discontinued the study for
GI‐related adverse events [50].
Semaglutide
Semaglutide is the newest GLP‐1 RA to be FDA‐approved
for use in type 2 diabetes. It is currently available as a once
weekly injection at doses of 0.5 mg and 1.0 mg. Structurally,
it is similar to liraglutide, although has a longer half‐life of
approximately 7 days.
In the first RCT to examine the effects of this GLP‐1 RA,
semaglutide was compared to diet and exercise in patients
with type 2 diabetes who were treatment naïve and who
had failed to achieve glycemic control. Average HbA1c
reductions in the semaglutide groups were 1.45% and
1.55% with 0.5mg and 1.0mg, respectively vs. 0.02% in the
placebo group. Weight decreased by 3.73  kg and 4.54  kg
with 0.5mg and 1.0mg of semaglutide, respectively vs.
0.98  kg with placebo  [51]. Semaglutide has also demon­
strated greater weight loss when compared head‐to‐head
with extended‐release exenatide and dulaglutide [52, 53].
Thus, to date, semaglutide portends the greatest weight loss
effect of all available GLP‐1RAs.
Like other GLP‐1 RAs when examined as add‐on ther­
apy to basal insulin, semaglutide plus basal insulin showed
substantial HbA1c‐lowering of 1.4% and 1.8% with 0.5mg
and 1.0mg semaglutide weekly. This was in contrast to a
0.1% reduction in the group managed without semaglu­
tide. Reductions were greater with higher baseline HbA1c
values. Significantly more patients in the semaglutide
groups achieved HbA1c levels < 7% without an increase in
hypoglycemia. Overall reductions in 7‐point self‐moni­
tored blood glucose values were greater in the semaglutide
groups, as well. Weight reductions were also greater in the
semaglutide group (3.7  kg and 6.4  kg with semaglutide
0.5 mg and 1.0 mg, respectively vs. 1.4 kg in the comparator
group). However, there were a higher number of hypogly­
cemic events noted in the semaglutide groups among
patients whose baseline HbA1c levels were < 8.0%. No dif­
ferences in rates of hypoglycemia were noted between
groups among individuals whose baseline HbA1c values
were > 8.0% [54].
A notable finding that has emerged from studies of
semaglutide is its association with progressive diabetic
retinopathy. In a major CVOT of semaglutide compared
with placebo, there were significantly higher rates of dia­
betic retinopathy complications (HR 1.76) among patients
treated with semaglutide. Post‐hoc analysis demonstrated
 Incretin-Based Therapy for the Management of Type 2 Diabetes
that rates were highest among those with pre‐existing dia­
betic retinopathy at study entry, and the trial was not
designed to adequately assess progression of retinal
changes over time. While signals of early retinopathy pro­
gression were noted in trials involving both liraglutide and
exenatide, these findings did not reach statistical signifi­
cance, and the trials were similarly not designed to ade­
quately examine this outcome. Despite these limitations,
there is substantial evidence to suggest that rapid glucose
lowering with agents such as insulin and GLP‐1 RAs does
result in an “early worsening” phenomenon of diabetic
retinopathy. This is related to both the magnitude and
rapidity of glucose lowering. Thus, pre‐existing diabetic
retinopathy should be carefully evaluated and managed
prior to initiation of GLP‐1 RAs, and risks of progression
should be thoroughly discussed with patients prior to com­
mencement of therapy [26, 55].
With glycemic control and weight reduction being the
cornerstones of type 2 diabetes care, GLP‐1 RAs demon­
strate efficacy in each of these realms. Although direct
comparisons cannot be made between many of the trials
examining utility of these agents individually, general con­
clusions about this class can be made based on the available
evidence: (1) reductions in HbA1c and weight occur along
a spectrum within the class of GLP‐1 RAs, with longer‐act­
ing agents demonstrating greater overall HbA1c reduction
compared to shorter‐acting agents, and semaglutide exhib­
iting the most potent effect on weight loss; (2) GLP‐1RAs
are useful as both monotherapy and in combination with
basal insulin or other oral antidiabetic agents, and (3) their
relative safety, efficacy, and ease of use allow for glycemic
optimization while minimizing risks of hypoglycemia and
overall burden of disease.
The role of DPP‐4 inhibitors in type 2
diabetes management
Endogenous DPP‐4 functions to degrade GLP‐1, thus
reducing its biological half‐life. The class of incretin
mimetics can, therefore, be expanded to include drugs spe­
cifically designed for inhibiting this enzymatic degradation
and augmenting the endogenous effects of native GLP‐1:
DPP‐4 inhibitors. Rationale for the use of DPP‐4 inhibitors
stems from studies of DPP‐4 knockout mice which, com­
151
pared to wild type mice, exhibited higher plasma GLP‐1
and insulin levels, as well as lower plasma glucose levels, in
response to a glucose challenge. Administration of a
DPP‐4 inhibitor to wild type mice resulted in similar out­
comes to those of DPP‐4  knockout mice. This suggested
that the absence of DPP‐4 augments both the plasma con­
centrations and insulinotropic effects of endogenous
GLP‐1 [56].
The GLP‐1‐augmenting effect of DPP‐4 inhibitors has
been well‐established, with several studies demonstrating a
doubling of active GLP‐1 in the postprandial state [57–59].
In association with this elevation is a lowering of postpran­
dial glucose and glucagon concentrations, although the
direct effect on insulin secretion is controversial. Some
studies report an increase in insulin levels, while others
show no change with DPP‐4  inhibition  [57, 59, 60].
Additionally, gastric emptying does not seem to be signifi­
cantly impaired by DPP‐4  inhibition, suggesting that
delayed meal appearance is not a suitable explanation for
the glucose‐lowering effect of these agents [57].
All FDA‐approved DPP‐4 inhibitors interact with vari­
ous sites of the DPP‐4 enzyme to exert > 80% inhibi­
tion  [61]. As demonstrated in early human studies
regarding mechanisms of action, this degree of inhibition
is sufficient to raise active GLP‐1  levels to produce both
insulinotropic and glucagon‐lowering effects, yet insuffi­
cient to alter gastric emptying [57]. Collectively, this data
suggests that the glucose lowering effects of DPP‐4 inhibi­
tors are likely due to actions on postprandial beta cell func­
tion rather than on mediation of gastric emptying. This is
unique from GLP‐1 RAs and may explain the glucose‐low­
ering differential observed between these two drug classes.
A multitude of DPP‐4  inhibitors have been FDA‐
approved for use in type 2 diabetes. Currently available
DPP‐4 inhibitors include sitagliptin, saxagliptin, linaglipi­
tin, vildagliptin, and alogliptin, all of which can be admin­
istered orally. When compared to long‐acting GLP‐1 RAs,
they exert a lesser effect on glycemic control and are con­
sidered to be weight‐neutral. Yet, despite these differences
in efficacy, both GLP‐1 RAs and DPP‐4 inhibitors share a
low risk for hypoglycemia, with the highest propensity for
hypoglycemia being in the setting of combined use with
insulin or sulfonylureas. In accordance with their lack of
effect on gastric emptying, DPP‐4  inhibitors are also not
 152 Initial Evaluation and Management of Diabetes
known to cause significant GI side effects and are generally
well‐tolerated [61].
DPP‐4 inhibitors have been shown to be safe for use in
the setting of renal impairment or ESRD. As sitagliptin,
saxagliptin, vildagliptin, and alogliptin are renally‐
excreted, doses should be adjusted when used in these set­
tings. Linaglitptin does not undergo renal excretion but is
rather hepatically‐metabolized. Thus, dose adjustments are
not necessary when this agent is initiated in the setting of
CKD. Both RCTs and meta‐analyses have implicated
DPP‐4  inhibitors and GLP‐1 RAs as having potential
nephroprotective effects in patients with type 2 diabetes,
although the extent of this remains unclear. While several
RCTs have demonstrated improvements in development of
macroalbuminuria with use of these agents, there remains
to be consistent data demonstrating overall improvement
in progression to ESRD or microalbuminuria [62]. Thus,
further research is necessary to better define the role of
incretin‐based therapies on clinically‐important renal out­
comes in diabetes.
Similar to GLP‐1 RAs, considerable debate has ensued
regarding the potential risk for acute pancreatitis in asso­
ciation with DPP‐4 inhibitor use. While robust data has not
substantiated these concerns, the FDA has issued a warn­
ing label on all incretin‐based therapies regarding risk of
acute pancreatitis  [23, 63]. Unlike GLP‐1 RAs, however,
there have been no signals indicating a heightened risk for
c‐cell hyperplasia or medullary thyroid cancers with use of
DPP‐4 inhibitors.
Traditionally, oral diabetic agents such as sulfonylureas
and TZDs have been advocated for as adjuncts to met­
formin therapy if glycemic targets were not met. However,
these agents are associated with inherent weight gain and,
in the case of sulfonylureas, significant risk of hypoglyce­
mia. Thus, given their beneficial effect on hyperglycemia,
low risk of hypoglycemia, weight neutrality, and ease of
administration, DPP‐4 inhibitors have become viable con­
tenders as second or third‐line agents in the management
of type 2 diabetes. In a series of RCTs, DPP‐4  inhibitors
were shown to be non‐inferior to sulfonylureas with regard
to HbA1c‐lowering and demonstrated statistically signifi­
cant benefits over sulfonylureas with regard to hypoglyce­
mia and weight loss  [63–69]. One of these trials also
suggested that DPP‐4  inhibitors were associated with
enhanced beta cell responsiveness, a function that was not
preserved with use of sulfonylureas [67]. Several RCTs have
also compared the glycemic effectiveness of DPP‐4 inhibi­
tors to TZDs, namely pioglitazone, and demonstrated simi­
lar if not enhanced efficacy in glucose‐lowering with
DPP‐4 inhibitors. However, TZDs continue to have greater
impact on insulin resistance through their actions on
PPARℽ in adipocytes and hepatocytes, thus distinguishing
them as a unique therapeutic class to be considered in
select cases [70–72].
Although decisions regarding second and third‐line
therapies for type 2 diabetes are dependent upon a balance
of both risks and benefits of various drug classes,
DPP‐4  inhibitors have demonstrated substantial efficacy
for their intended use. They are associated with minimal
hypoglycemia and can be administered in such a way that
limits the daily burden of disease for patients. Given the
robust data available, it is reasonable to consider
DPP‐4  inhibitors in advance of sulfonylureas as add‐on
therapy, although cost of these medications continues to
limit more widespread use.
Cardiovascular outcomes with incretin‐
based therapies for type 2 diabetes
Diabetes confers significant morbidity and mortality, with
atherosclerotic cardiovascular disease (ASCVD) being the
leading cause of death among those affected. Patients with
diabetes possess approximately a 2‐fold increased risk of
coronary artery disease, stroke, and death from other vas­
cular causes [73]. In 2008, amid concerns for increased car­
diovascular risk with rosiglitazone, the FDA mandated that
cardiovascular safety trials be conducted for all agents
approved and marketed for type 2 diabetes [74]. The major
criterion for safety was achievement of a minimum hazard
ratio (HR) of 1.30  with composite endpoints of major
adverse cardiovascular events (MACE) including non‐fatal
myocardial infarction, non‐fatal stroke, and death from
cardiovascular disease  [74]. Subsequently, this led to a
robust influx of CVOTs involving incretin‐based therapies,
as well a novel class of drugs targeting renal sodium‐glu­
cose transporters, or SGLT‐2  inhibitors. These CVOTs
were designed to demonstrate non‐inferiority to standard
of care. The emergence of these trials dramatically trans­
 Incretin-Based Therapy for the Management of Type 2 Diabetes
formed the landscape of diabetes management and opened
the door to novel therapies portending both glycemic and
cardiovascular benefit.
To date, CVOTs have been published for all FDA‐
approved GLP‐1 RAs. These CVOTs are prospective,
multi‐center RCTs designed to demonstrate non‐inferior­
ity to standard of care in patients with type 2 diabetes. All
have achieved the FDA‐mandated criteria for safety in the
primary specified cardiovascular outcomes. Standard dia­
betes care in these trials has generally consisted of a combi­
nation of oral hypoglycemic agents, excluding incretin‐based
therapies, and/or basal insulin. Inclusion criteria have
encompassed patients with cardiovascular risk factors and/
or established ASCVD.
Of the GLP‐1 RA CVOTs published to date, LEADER,
SUSTAIN‐6, HARMONY, and REWIND have reported
liraglutide, semaglutide, albiglutide, and dulaglutide,
respectively, to have statistically significant reductions in
cardiovascular events compared to placebo  [25–27, 31].
Trials involving sustained‐release exenatide (EXSCEL) and
lixisenatide (ELIXA) have demonstrated non‐inferiority
for these cardiovascular endpoints  [28, 29]. Clinical sub­
group analyses have also provided insights regarding use of
these agents in clinical practice. In a meta‐analysis of the
major GLP‐1 RA CVOTs, it was suggested that patients
with baseline characteristics of BMI > 30 kg/m2, male gen­
der, and pre‐existing ASCVD were most likely to benefit
from GLP‐1 RAs, and specifically those that have been
shown to significantly improve MACE outcomes  [75].
Thus, based on these trials, the ADA recommends that
GLP‐1 RAs demonstrating cardiovascular benefit be initi­
ated in patients with type 2 diabetes who have established
ASCVD or multiple risk factors for such [76].
The CVOTs for DPP‐4 inhibitors have established non‐
inferiority to standard care with regard to MACE out­
comes, although superiority has not been demonstrated [33,
34, 77, 78]. As a result, DPP‐4 inhibitors are considered to
have neutral effects on cardiovascular disease in patients
with type 2 diabetes. One signal for increased heart failure
hospitalizations was noted in the SAVOR‐TIMI 53 trial, a
CVOT that compared saxagliptin to standard care with
placebo. In this study, saxagliptin was associated with a
heart failure hospitalization incidence of 3.5% compared to
2.4% in the placebo group (HR 1.27).(78) However, similar
153
signals were not observed in CVOTs involving sitagliptin
(TECOS), alogliptin (EXAMINE), or linagliptin
(CARMELINA) [33, 34, 77]. Similarly, there have been no
demonstrable associations with heart failure among the
GLP‐1 RAs, suggesting that this is unlikely to be a class
effect among incretin‐based therapies.
In the last decade, there has been increased recognition
of the effects of incretin‐based therapy beyond glycemic
control. Based on their proven impact on weight and car­
diovascular disease, incretin‐based therapies have been
firmly integrated into the mainstream of diabetes care.
Incorporating incretin‐based therapy into
clinical practice
The approach to management of type 2 diabetes has tradi­
tionally followed a hierarchical algorithm in accordance
with disease severity and rate of progression. As insulin
resistance is strongly correlated with obesity, weight reduc­
tion through lifestyle modifications is an early intervention
which should be maintained throughout the disease
course [79]. However, given the inevitability of disease pro­
gression despite patients’ best efforts, pharmacologic ther­
apy is ultimately crucial in optimizing diabetes care.
Unless contraindicated or poorly‐tolerated, metformin
remains the first‐line pharmacologic agent of choice for
type 2 diabetes. In addition to modulating insulin resist­
ance and improving glycemic control, metformin has also
been shown to have modest cardiovascular benefit, an
advantage which has not been reported with sulfonylu­
reas [80, 81].
If glycemic goals remain unmet with metformin mono­
therapy, initiation of additional agents is then based on fac­
tors pertaining to patient preference, comorbidities, disease
severity, cost‐effectiveness, and risk of associated adverse
events. Insulin should be initiated early if hyperglycemia is
severe (plasma glucose > 300 mg/dL or HbA1c > 10%), or
if signs of hyperglycemia or increased catabolism are pre­
sent. In the absence of these findings, oral agents can be
considered as add‐on therapy to metformin. Sulfonylureas
and TZDs have traditionally been the recommended sec­
ond and third‐line therapies in these settings. How­
ever, while generally well‐tolerated and effective for
glu­cose‐lowering, these agents portend undesirable risks
 154 Initial Evaluation and Management of Diabetes
including weight gain and, in the case of sulfonylureas,
hypoglycemia [81].
In light of the recent data that has emerged for incretin‐
based therapies, institutional guidelines regarding stand­
ard pharmacotherapy for type 2 diabetes have shifted
substantially. Current ADA guidelines recommend consid­
eration of GLP‐1 RAs for patients not adequately con­
trolled with metformin who (1) are at increased risk for
ASCVD, (2) have established ASCVD, (3) are in need of
therapies that minimize hypoglycemia, or (4) would bene­
fit from weight reduction or minimization of weight
gain [81]. For patients with heightened risk of ASCVD, or
for those with established ASCVD, GLP‐1 RAs with proven
cardiovascular benefit such as liraglutide, dulaglutide, and
semaglutide are recommended. Although albiglutide has
proven cardiovascular benefit in CVOTs, it is not currently
marketed for use in the United States. Worth mentioning,
although outside the scope of this chapter, are
SGLT‐2  inhibitors. These agents have gained significant
traction in diabetes care, having proven cardiovascular and
renal efficacy in patients with type 2 diabetes. Therefore,
they have also been incorporated into current guidelines
for consideration in patients with or at risk for ASCVD,
heart failure, and chronic kidney disease [81].
In the absence of ASCVD or significant risk thereof,
DPP‐4 inhibitors can be considered as second‐line therapy.
These agents may also be added when GLP‐1 RAs are not
concurrently used  [81]. Due to its association with
increased heart failure hospitalizations, saxagliptin should
be avoided in the setting of heart failure [78]. However, all
other DPP‐4  inhibitors, given their neutral status in the
role of cardiovascular disease, are considered safe for
patients at higher risk for or with established ASCVD. Of
note, combined use of GLP‐1 RAs and DPP‐4 inhibitors is
not recommended due to lack of expected benefit with
additive therapy and paucity of data regarding their com­
bined use.
Although incretin‐based therapies and SGLT‐2 inhibi­
tors now occupy a dominant role in type 2 diabetes care, it
is important for clinicians to remain familiar with the role
of sulfonylureas and TZDs in select cases. According to
current ADA guidelines, these medications should be con­
sidered as second‐line therapy to metformin when cost is a
major consideration for patients. While lacking in compa­
rable efficacy for major cardiovascular outcomes and being
associated with weight gain and hypoglycemia, sulfonylu­
reas remain potent glucose‐lowering medications and can
be carefully titrated to maximize both safety and efficacy.
TZDs are also associated with a variety of adverse events
including heart failure, weight gain, fractures, and bladder
cancer. Yet, pioglitazone, the only available TZD, remains
effective for hyperglycemia and can be considered as a
means of cost‐effective treatment in select clinical scenar­
ios.(81)
Conclusions
Type 2 diabetes is characterized by chronic hyperglycemia
resulting from impaired beta cell function, insulin resist­
ance, and diminished endogenous incretin effects. Incretin
hormones are integral to glycemic control and thus have
been the targets of considerable research for their role in
both treatment and attenuation of disease progression.
High‐quality evidence has shown GLP‐1 RAs to be effec­
tive in multiple aspects of type 2 diabetes, from glucose‐
lowering and weight reduction to cardiovascular outcome
improvement. DPP‐4 inhibitors have demonstrated mod­
est efficacy in glycemic control while minimizing weight
gain, and both carry a low risk for hypoglycemia. Indeed,
many questions remain unanswered regarding the dura­
tion of glycemic control that can be expected with incretin‐
based therapies, as well as their role in mitigating
progression of disease. However, available evidence sug­
gests that these agents exhibit substantial promise in reduc­
ing diabetes‐related morbidity, and they have dramatically
modified the approach to care for patients with diabetes.
Key points
●● The incretin effect mediates insulin release and gluca­
gon suppression in response to meal ingestion. This
effect is disrupted in diabetes.
●● Incretin mimetics enhance islet insulinotropic and
glucagon suppressive responses in a glucose‐mediated
fashion.
●● Hypoglycemia is rare with GLP‐1 RAs and DPP‐4 inhib­
itors, although may occur when used in combination
with insulin or sulfonylureas.
 Incretin-Based Therapy for the Management of Type 2 Diabetes
●● Weight loss is expected with GLP‐1 RAs, with semaglu­
tide being most effective for weight reduction.
●● DPP‐4  inhibitors are weight‐neutral and less effective
for glycemic control compared to GLP‐1 RAs.
●● Liraglutide, semaglutide, albiglutide, and dulaglutide
have been shown to significantly reduce cardiovascular
events in major cardiovascular outcome trials.
●● DPP‐4 inhibitors have neutral effects on cardiovascular
outcomes. Saxagliptin should be avoided in the setting
of heart failure.
References
1. Sharabi K, Tavares CD, Rines AK, Puigserver P. Molecular
pathophysiology of hepatic glucose production. Mol Aspects
Med. 2015;46:21–33. Epub 2015/11/10. doi: 10.1016/j.mam
.2015.09.003. PubMed PMID: 26549348; PMCID. PMC46
74831.
2. Brunzell JD, Robertson RP, Lerner RL, Hazzard WR, Ensinck
JW, Bierman EL, Porte D. Relationships between fasting plasma
glucose levels and insulin‐secretion during intravenous glu­
cose‐tolerance tests. J Clin Endocr Metab. 1976;42(2):
222–229. doi:. DOI 10.1210/jcem‐42‐2‐222. PubMed PMID:
WOS:A1976BG96800003.
3. Del Prato S, Marchetti P, Bonadonna RC. Phasic insulin
release and metabolic regulation in type 2 diabetes. Diabetes.
2002;51 Suppl 1:S109–116. Epub 2002/01/30. doi: 10.2337/
diabetes.51.2007.s109. PubMed PMID: 11815468.
4. Elrick H, Stimmler L, Hlad CJ, Jr., Arai Y. Plasma insulin
response to oral and intravenous glucose administration.
J Clin Endocrinol Metab. 1964;24:1076–1082. Epub 1964/
10/01. doi: 10.1210/jcem‐24‐10‐1076. PubMed PMID:
14228531.
5. Toft‐Nielsen MB, Damholt MB, Madsbad S, Hilsted LM,
Hughes TE, Michelsen BK, Holst JJ. Determinants of the
impaired secretion of glucagon‐like peptide‐1 in type 2 dia­
betic patients. J Clin Endocrinol Metab. 2001;86(8):
3717–3723. Epub 2001/08/15. doi: 10.1210/jcem.86.8.7750.
PubMed PMID: 11502801.
6. Nauck M, Stockmann F, Ebert R, Creutzfeldt W. Reduced
incretin effect in type 2 (non‐insulin‐dependent) diabetes.
Diabetologia. 1986;29(1):46–52. Epub 1986/01/01. doi:
10.1007/bf02427280. PubMed PMID: 3514343.
7. Nauck MA, Meier JJ. The incretin effect in healthy individuals
and those with type 2 diabetes: physiology, pathophysiology, and
response to therapeutic inter­ventions. Lancet Diabetes Endo.
155
2016;4(6):525–536. doi: 10.1016/S2213‐8587(15)00482‐9. Pub
Med PMID: WOS:000376462800018.
8. Calanna S, Christensen M, Holst JJ, Laferrere B, Gluud LL,
Vilsboll T, Knop FK. Secretion of glucagon‐like peptide‐1 in
patients with type 2 diabetes mellitus: systematic review and
meta‐analyses of clinical studies. Diabetologia.
2013;56(5):965–972. Epub 2013/02/05. doi: 10.1007/s00125‐
013‐2841‐0. PubMed PMID: 23377698; PMCID. PMC3
687347.
9. Calanna S, Christensen M, Holst JJ, Laferrere B, Gluud LL,
Vilsboll T, Knop FK. Secretion of glucose‐dependent insuli­
notropic polypeptide in patients with type 2 diabetes: sys­
tematic review and meta‐analysis of clinical studies. Diabetes
Care. 2013;36(10):3346–3352. Epub 2013/09/26. doi:
10.2337/dc13‐0465. PubMed PMID: 24065842; PMCID.
PMC3781498.
10. Nauck MA, Heimesaat MM, Orskov C, Holst JJ, Ebert R,
Creutzfeldt W. Preserved incretin activity of glucagon‐like
peptide 1 [7‐36 amide] but not of synthetic human gastric
inhibitory polypeptide in patients with type‐2 diabetes mel­
litus. J Clin Invest. 1993;91(1):301–307. Epub 1993/01/01.
doi: 10.1172/JCI116186. PubMed PMID: 8423228; PMCID.
PMC330027.
11. Nauck MA, Kleine N, Orskov C, Holst JJ, Willms B,
Creutzfeldt W. Normalization of fasting hyperglycaemia by
exogenous glucagon‐like peptide 1 (7‐36 amide) in type 2
(non‐insulin‐dependent) diabetic patients. Diabetologia.
1993;36(8):741–744. Epub 1993/08/01. doi: 10.1007/
bf00401145. PubMed PMID: 8405741.
12. Chia CW, Egan JM. Incretin‐based therapies in type 2 diabe­
tes mellitus. J Clin Endocrinol Metab. 2008;93(10):3703–
3716. Epub 2008/07/17. doi: 10.1210/jc.2007–2109. PubMed
PMID: 18628530; PMCID. PMC2579648.
13. Greig NH, Holloway HW, De Ore KA, Jani D, Wang Y, Zhou
J, Garant MJ, Egan JM. Once daily injection of exendin‐4 to
diabetic mice achieves long‐term beneficial effects on blood
glucose concentrations. Diabetologia. 1999;42(1):45–50.
Epub 1999/02/23. doi: 10.1007/s001250051111. PubMed
PMID: 10027577.
14. Young AA, Gedulin BR, Bhavsar S, Bodkin N, Jodka C,
Hansen B, Denaro M. Glucose‐lowering and insulin‐sensi­
tizing actions of exendin‐4: studies in obese diabetic (ob/ob,
db/db) mice, diabetic fatty Zucker rats, and diabetic rhesus
monkeys (Macaca mulatta). Diabetes. 1999;48(5):1026–
1034. Epub 1999/05/20. doi: 10.2337/diabetes.48.5.1026.
PubMed PMID: 10331407.
15. Egan JM, Clocquet AR, Elahi D. The insulinotropic effect of
acute exendin‐4 administered to humans: comparison of
 156 Initial Evaluation and Management of Diabetes
nondiabetic state to type 2 diabetes. J Clin Endocrinol Metab.
2002;87(3):1282–1290. Epub 2002/03/13. doi: 10.1210/jcem.87.
3.8337. PubMed PMID: 11889200.
16. Fehse F, Trautmann M, Holst JJ, Halseth AE, Nanayakkara
N, Nielsen LL, Fineman MS, Kim DD, Nauck MA. Exenatide
augments first‐ and second‐phase insulin secretion in
response to intravenous glucose in subjects with type 2 diabe­
tes. J Clin Endocrinol Metab. 2005;90(11):5991–5997. Epub
2005/09/08. doi: 10.1210/jc.2005‐1093. PubMed PMID:
16144950.
17. DeFronzo RA, Ratner RE, Han J, Kim DD, Fineman MS,
Baron AD. Effects of exenatide (exendin‐4) on glycemic
control and weight over 30  weeks in metformin‐treated
patients with type 2 diabetes. Diabetes Care. 2005;28(5):1092–
1100. Epub 2005/04/28. doi: 10.2337/diacare.28.5.1092. Pub
Med PMID: 15855572.
18. Kolterman OG, Buse JB, Fineman MS, Gaines E, Heintz S,
Bicsak TA, Taylor K, Kim D, Aisporna M, Wang Y, Baron
AD. Synthetic exendin‐4 (Exenatide) significantly reduces
postprandial and fasting plasma glucose in subjects with
type 2 diabetes. J Clin Endocr Metab. 2003;88(7):3082–3089.
doi: 10.1210/jc.2002‐021545. PubMed PMID: WOS:0001
83926300020.
19. Knop FK, Bronden A, Vilsboll T. Exenatide: pharmacokinet­
ics, clinical use, and future directions. Expert Opin
Pharmaco. 2017;18(6):555–571. doi: 10.1080/14656566.
2017.1282463. PubMed PMID: WOS:000399620000003.
20. Aroda VR. A review of GLP‐1 receptor agonists: evolution
and advancement, through the lens of randomised con­
trolled trials. Diabetes Obes Metab. 2018;20 Suppl 1:22–33.
Epub 2018/01/25. doi: 10.1111/dom.13162. PubMed PMID:
29364586.
21. Anderson SL, Beutel TR, Trujillo JM. Oral semaglutide in
type 2 diabetes. J Diabetes Complicat. 2020:107520. Epub
2020/01/19. doi: 10.1016/j.jdiacomp.2019.107520. PubMed
PMID: 31952996.
22. Lee YS, Lee C, Choung JS, Jung HS, Jun HS. Glucagon‐like
peptide 1  increases beta‐cell regeneration by promoting
alpha‐ to beta‐cell transdifferentiation. Diabetes. 2018;67(12):
2601–2614. Epub 2018/09/28. doi: 10.2337/db18‐0155. Pub
Med PMID: 30257975.
23. Tasyurek HM, Altunbas HA, Balci MK, Sanlioglu S.
Incretins: their physiology and application in the treatment
of diabetes mellitus. Diabetes Metab Res Rev. 2014;30(5):354–
371. Epub 2014/07/06. doi: 10.1002/dmrr.2501. PubMed
PMID: 24989141.
24. De Cosmo S, Viazzi F, Piscitelli P, Leoncini G, Mirijello A,
Bonino B, Pontremoli R. Impact of CVOTs in primary and
secondary prevention of kidney disease. Diabetes Res Clin
Pract. 2019;157:107907. Epub 2019/11/05. doi: 10.1016/j.dia
bres.2019.107907. PubMed PMID: 31676332.
25. Marso SP, Daniels GH, Brown‐Frandsen K, Kristensen P,
Mann JF, Nauck MA, Nissen SE, Pocock S, Poulter NR, Ravn
LS, Steinberg WM, Stockner M, Zinman B, Bergenstal RM,
Buse JB, for the LEADER Steering Committee on behalf of
the LEADER Trial Investigators. Liraglutide and cardiovas­
cular outcomes in type 2 diabetes. N Engl J Med. 2016;375(4):
311–322. Epub 2016/06/14. doi: 10.1056/NEJMoa1603827.
PubMed PMID: 27295427; PMCID. PMC4985288.
26. Marso SP, Bain SC, Consoli A, Eliaschewitz FG, Jodar E,
Leiter LA, Lingvay I, Rosenstock J, Seufert J, Warren ML,
Woo V, Hansen O, Holst AG, Pettersson J, Vilsboll T, for the
SUSTAIN‐6  Investigators. Semaglutide and cardiovascular
outcomes in patients with type 2 diabetes. N Engl J Med.
2016;375(19):1834–1844. Epub 2016/09/17. doi: 10.1056/NEJ
Moa1607141. PubMed PMID: 27633186.
27. Gerstein HC, Colhoun HM, Dagenais GR, Diaz R,
Lakshmanan M, Pais P, Probstfield J, Riesmeyer JS, Riddle
MC, Ryden L, Xavier D, Atisso CM, Dyal L, Hall S, Rao‐
Melacini P, Wong G, Avezum A, Basile J, Chung N, Conget I,
Cushman WC, Franek E, Hancu N, Hanefeld M, Holt S,
Jansky P, Keltai M, Lanas F, Leiter LA, Lopez‐Jaramillo P,
Cardona Munoz EG, Pirags V, Pogosova N, Raubenheimer
PJ, Shaw JE, Sheu WH, Temelkova‐Kurktschiev T, for the
REWIND Investigators. Dulaglutide and cardiovascular out­
comes in type 2 diabetes (REWIND): a double‐blind, ran­
domised placebo‐controlled trial. Lancet. 2019;394(10193):
121–130. Epub 2019/06/14. doi: 10.1016/S0140‐6736(19)
31149‐3. PubMed PMID: 31189511.
28. Holman RR, Bethel MA, Mentz RJ, Thompson VP,
Lokhnygina Y, Buse JB, Chan JC, Choi J, Gustavson SM,
Iqbal N, Maggioni AP, Marso SP, Ohman P, Pagidipati NJ,
Poulter N, Ramachandran A, Zinman B, Hernandez AF, for
the EXSCEL Study Group. Effects of once‐weekly exenatide
on cardiovascular outcomes in type 2 diabetes. N Engl J Med.
2017;377(13):1228–1239. Epub 2017/09/15. doi: 10.1056/
NEJMoa1612917. PubMed PMID: 28910237.
30. Scheen AJ. Pharmacokinetics and clinical use of incretin‐
based therapies in patients with chronic kidney disease and
type 2 diabetes. Clin Pharmacokinet. 2015;54(1):1–21. doi:
10.1007/s40262‐014‐0198‐2. PubMed PMID: WOS:000347147
300001.
31. Hernandez AF, Green JB, Janmohamed S, D’Agostino RB,
Sr., Granger CB, Jones NP, Leiter LA, Rosenberg AE, Sigmon
KN, Somerville MC, Thorpe KM, McMurray JJV, Del Prato
S, for the Harmony Outcomes committees and investigators.
 Incretin-Based Therapy for the Management of Type 2 Diabetes
Albiglutide and cardiovascular outcomes in patients with
type 2 diabetes and cardiovascular disease (Harmony
Outcomes): a double‐blind, randomised placebo‐controlled
trial. Lancet. 2018;392(10157):1519–1529. Epub 2018/10/07.
doi: 10.1016/S0140‐6736(18)32261‐X. PubMed PMID: 302
91013.
32. Husain M, Birkenfeld AL, Donsmark M, Dungan K,
Eliaschewitz FG, Franco DR, Jeppesen OK, Lingvay I,
Mosenzon O, Pedersen SD, Tack CJ, Thomsen M, Vilsboll
T, Warren ML, Bain SC, for the PIONEER 6 Investigators.
Oral semaglutide and cardiovascular outcomes in patients
with type 2 diabetes. N Engl J Med. 2019. Epub 2019/06/12.
doi: 10.1056/NEJMoa1901118. PubMed PMID:
31185157.
33. White WB, Cannon CP, Heller SR, Nissen SE, Bergenstal
RM, Bakris GL, Perez AT, Fleck PR, Mehta CR, Kupfer S,
Wilson C, Cushman WC, Zannad F, for the EXAMINE
Investigators. Alogliptin after acute coronary syndrome in
patients with type 2 diabetes. N Engl J Med. 2013;369(14):
1327–1335. Epub 2013/09/03. doi: 10.1056/NEJMoa1305889.
PubMed PMID: 23992602.
34. Green JB, Bethel MA, Armstrong PW, Buse JB, Engel SS,
Garg J, Josse R, Kaufman KD, Koglin J, Korn S, Lachin JM,
McGuire DK, Pencina MJ, Standl E, Stein PP, Suryawanshi S,
Van de Werf F, Peterson ED, Holman RR, for the TECOS
Study Group. Effect of sitagliptin on cardiovascular outcomes
in type 2 diabetes. N Engl J Med. 2015;373(3):232–242. Epub
2015/06/09. doi: 10.1056/NEJMoa1501352. PubMed PMID:
26052984.
35. Buse JB, Henry RR, Han J, Kim DD, Fineman MS, Baron
AD, for the Exenatide‐113 Clinical Study Group. Effects of
exenatide (exendin‐4) on glycemic control over 30 weeks in
sulfonylurea‐treated patients with type 2 diabetes. Diabetes
Care. 2004;27(11):2628–2635. Epub 2004/10/27. doi: 10.2337/
diacare.27.11.2628. PubMed PMID: 15504997.
36. Kendall DM, Riddle MC, Rosenstock J, Zhuang D, Kim DD,
Fineman MS, Baron AD. Effects of exenatide (exendin‐4) on
glycemic control over 30 weeks in patients with type 2 diabe­
tes treated with metformin and a sulfonylurea. Diabetes
Care. 2005;28(5):1083–1091. Epub 2005/04/28. doi: 10.2337/
diacare.28.5.1083. PubMed PMID: 15855571.
37. Zinman B, Hoogwerf BJ, Duran Garcia S, Milton DR,
Giaconia JM, Kim DD, Trautmann ME, Brodows RG. The
effect of adding exenatide to a thiazolidinedione in subopti­
mally controlled type 2 diabetes: a randomized trial. Ann
Intern Med. 2007;146(7):477–485. Epub 2007/04/04. doi: 10.
7326/0003‐4819‐146‐7‐200704030‐00003. PubMed PMID:
17404349.
157
38. Nauck MA, Duran S, Kim D, Johns D, Northrup J, Festa A,
Brodows R, Trautmann M. A comparison of twice‐daily
exenatide and biphasic insulin aspart in patients with type 2
diabetes who were suboptimally controlled with sulfonylu­
rea and metformin: a non‐inferiority study. Diabetologia.
2007;50(2):259–267. Epub 2006/12/13. doi: 10.1007/s00125
‐006‐0510‐2. PubMed PMID: 17160407.
39. Wysham CH, Rosenstock J, Vetter ML, Dong F, Ohman P,
Iqbal N. Efficacy and tolerability of the new autoinjected
suspension of exenatide once weekly versus exenatide
twice daily in patients with type 2 diabetes. Diabetes Obes
Metab. 2018;20(1):165–172. Epub 2017/07/08. doi:
10.1111/dom.13056. PubMed PMID: 28685973; PMCID.
PMC5724491.
40. Buse JB, Bergenstal RM, Glass LC, Heilmann CR, Lewis
MS, Kwan AYM, Hoogwerf BJ, Rosenstock J. Use of twice‐
daily exenatide in basal insulin‐treated patients with type 2
diabetes a randomized, controlled trial. Annals of Internal
Medicine. 2011;154(2):103‐+. doi: 10.7326/0003‐4819‐
154‐2‐201101180‐00300. PubMed PMID: WOS:000286
223500004.
41. Guja C, Frias JP, Somogyi A, Jabbour S, Wang H, Hardy E,
Rosenstock J. Effect of exenatide QW or placebo, both added
to titrated insulin glargine, in uncontrolled type 2 diabetes:
the DURATION‐7 randomized study. Diabetes Obes Metab.
2018;20(7):1602–1614. Epub 2018/02/24. doi: 10.1111/
dom.13266. PubMed PMID: 29473704; PMCID. PMC6
032936.
42. Buse JB, Rosenstock J, Sesti G, Schmidt WE, Montanya E,
Brett JH, Zychma M, Blonde L, Group L‐S. Liraglutide once
a day versus exenatide twice a day for type 2 diabetes: a 26‐
week randomised, parallel‐group, multinational, open‐label
trial (LEAD‐6). Lancet. 2009;374(9683):39–47. Epub 2009/
06/12. doi: 10.1016/S0140‐6736(09)60659‐0. PubMed PMID:
19515413.
43. Buse JB, Nauck M, Forst T, Sheu WH, Shenouda SK,
Heilmann CR, Hoogwerf BJ, Gao A, Boardman MK,
Fineman M, Porter L, Schernthaner G. Exenatide once weekly
versus liraglutide once daily in patients with type 2 diabetes
(DURATION‐6): a  randomised, open‐label study. Lancet.
2013;381(9861):117–124. Epub 2012/11/13. doi: 10.1016/
S0140‐6736(12)61267‐7. PubMed PMID: 23141817.
44. Buse JB, Vilsboll T, Thurman J, Blevins TC, Langbakke IH,
Bottcher SG, Rodbard HW, Investigators NNT. Contribution
of liraglutide in the fixed‐ratio combination of insulin deglu­
dec and liraglutide (IDegLira). Diabetes Care. 2014;37(11):
2926–2933. Epub 2014/08/13. doi: 10.2337/dc14‐0785.
PubMed PMID: 25114296.
 158 Initial Evaluation and Management of Diabetes
45. Rosenstock J, Raccah D, Koranyi L, Maffei L, Boka G,
Miossec P, Gerich JE. Efficacy and safety of lixisenatide once
daily versus exenatide twice daily in type 2 diabetes inade­
quately controlled on metformin: a 24‐week, randomized,
open‐label, active‐controlled study (GetGoal‐X). Diabetes
Care. 2013;36(10):2945–2951. Epub 2013/05/24. doi: 10.2337/
dc12‐2709. PubMed PMID: 23698396; PMCID. PMC
3781502.
46. Nauck M, Rizzo M, Johnson A, Bosch‐Traberg H, Madsen J,
Cariou B. Once‐daily liraglutide versus lixisenatide as add‐
on to metformin in type 2 diabetes: a 26‐week randomized
controlled clinical trial. Diabetes Care. 2016;39(9):1501–1509.
doi: 10.2337/dc15‐2479. PubMed PMID: WOS:000383663
000017.
47. Rosenstock J, Guerci B, Hanefeld M, Gentile S, Aronson R,
Tinahones FJ, Roy‐Duval C, Souhami E, Wardecki M, Ye J,
Perfetti R, Heller S, on behalf of the GetGoal Duo‐2 Trial
Investigators. Prandial options to advance basal insulin glar­
gine therapy: testing lixisenatide plus basal insulin versus
insulin glulisine either as basal‐plus or basal‐bolus in type 2
diabetes: the GetGoal Duo‐2 Trial. Diabetes Care. 2016;39(8):
1318–1328. Epub 2016/05/26. doi: 10.2337/dc16‐0014. PubMed
PMID: 27222510.
48. Wysham C, Blevins T, Arakaki R, Colon G, Garcia P, Atisso C,
Kuhstoss D, Lakshmanan M. Efficacy and safety of dulaglutide
added onto pioglitazone and metformin versus exenatide in
type 2 diabetes in a randomized controlled trial (AWARD‐1).
Diabetes Care. 2014;37(8):2159–2167. Epub 2014/06/01. doi:
10.2337/dc13‐2760. PubMed PMID: 24879836.
49. Dungan KM, Povedano ST, Forst T, Gonzalez JG, Atisso C,
Sealls W, Fahrbach JL. Once‐weekly dulaglutide versus once‐
daily liraglutide in metformin‐treated patients with type 2 dia­
betes (AWARD‐6): a randomised, open‐label, phase 3, non‐
inferiority trial. Lancet. 2014;384(9951):1349–1357. Epub 2014/
07/16. doi: 10.1016/S0140‐6736(14)60976‐4. PubMed PMID:
25018121.
50. Pozzilli P, Norwood P, Jodar E, Davies MJ, Ivanyi T, Jiang
HH, Woodward DB, Milicevic Z. Placebo‐controlled, rand­
omized trial of the addition of once‐weekly glucagon‐like
peptide‐1 receptor agonist dulaglutide to titrated daily insu­
lin glargine in patients with type 2 diabetes (AWARD‐9).
Diabetes Obesity & Metabolism. 2017;19(7):1024–1031. doi:
10.1111/dom.12937. PubMed PMID: WOS:0004032366
00013.
51. Sorli C, Harashima SI, Tsoukas GM, Unger J, Karsbol JD,
Hansen T, Bain SC. Efficacy and safety of once‐weekly sema­
glutide monotherapy versus placebo in patients with type 2
diabetes (SUSTAIN 1): a double‐blind, randomised, placebo‐
controlled, parallel‐group, multinational, multicentre phase
3a trial. Lancet Diabetes Endocrinol. 2017;5(4):251–260. Epub
2017/01/24. doi: 10.1016/S2213‐8587(17)30013‐X. PubMed
PMID: 28110911.
52. Ahmann AJ, Capehorn M, Charpentier G, Dotta F, Henkel
E, Lingvay I, Holst AG, Annett MP, Aroda VR. Efficacy and
safety of once‐weekly semaglutide versus exenatide ER in
subjects with type 2 diabetes (SUSTAIN 3): a 56‐week, open‐
label, randomized clinical trial. Diabetes Care. 2018;41(2):258–
266. Epub 2017/12/17. doi: 10.2337/dc17‐0417. PubMed
PMID: 29246950.
53. Pratley RE, Aroda VR, Lingvay I, Ludemann J, Andreassen
C, Navarria A, Viljoen A, investigators S. Semaglutide versus
dulaglutide once weekly in patients with type 2 diabetes
(SUSTAIN 7): a randomised, open‐label, phase 3b trial.
Lancet Diabetes Endocrinol. 2018;6(4):275–286. Epub 2018/
02/06. doi: 10.1016/S2213‐8587(18)30024‐X. PubMed PMID:
29397376.
54. Rodbard HW, Lingvay I, Reed J, de la Rosa R, Rose L,
Sugimoto D, Araki E, Chu PL, Wijayasinghe N, Norwood P.
Semaglutide added to basal insulin in type 2 diabetes
(SUSTAIN 5): a randomized, controlled trial. J Clin
Endocrinol Metab. 2018;103(6):2291–2301. Epub 2018/04/25.
doi: 10.1210/jc.2018‐00070. PubMed PMID: 29688502; PMCID.
PMC5991220.
55. Vilsboll T, Bain SC, Leiter LA, Lingvay I, Matthews D, Simo
R, Helmark IC, Wijayasinghe N, Larsen M. Semaglutide,
reduction in glycated haemoglobin and the risk of diabetic
retinopathy. Diabetes Obes Metab. 2018;20(4):889–897. Epub
2017/11/28. doi: 10.1111/dom.13172. PubMed PMID: 2917
8519; PMCID. PMC5888154.
56. Marguet D, Baggio L, Kobayashi T, Bernard AM, Pierres M,
Nielsen PF, Ribel U, Watanabe T, Drucker DJ, Wagtmann N.
Enhanced insulin secretion and improved glucose tolerance
in mice lacking CD26. Proc Natl Acad Sci USA. 2000;97(12):
6874–6879. Epub 2000/05/24. doi: 10.1073/pnas.120069197.
PubMed PMID: 10823914; PMCID. PMC18768.
57. Vella A, Bock G, Giesler PD, Burton DB, Serra DB, Saylan
ML, Dunning BE, Foley JE, Rizza RA, Camilleri M. Effects
of dipeptidyl peptidase‐4  inhibition on gastrointestinal
function, meal appearance, and glucose metabolism in
type 2 diabetes. Diabetes. 2007;56(5):1475–1480. Epub
2007/02/17. doi: 10.2337/db07‐0136. PubMed PMID:
17303799.
58. Bergman AJ, Stevens C, Zhou Y, Yi B, Laethem M, De Smet
M, Snyder K, Hilliard D, Tanaka W, Zeng W, Tanen M, Wang
 Incretin-Based Therapy for the Management of Type 2 Diabetes
AQ, Chen L, Winchell G, Davies MJ, Ramael S, Wagner JA,
Herman GA. Pharmacokinetic and pharmacodynamic
properties of multiple oral doses of sitagliptin, a dipeptidyl
peptidase‐IV inhibitor: a double‐blind, randomized, pla­
cebo‐controlled study in healthy male volunteers. Clin Ther.
2006;28(1):55–72. Epub 2006/02/24. doi: 10.1016/j.
clinthera.2006.01.015. PubMed PMID: 16490580.
59. Ahren B, Landin‐Olsson M, Jansson PA, Svensson M,
Holmes D, Schweizer A. Inhibition of dipeptidyl peptidase‐4
reduces glycemia, sustains insulin levels, and reduces gluca­
gon levels in type 2 diabetes. J Clin Endocrinol Metab.
2004;89(5):2078–2084. Epub 2004/05/06. doi: 10.1210/
jc.2003‐031907. PubMed PMID: 15126524.
60. Herman GA, Bergman A, Stevens C, Kotey P, Yi B, Zhao P,
Dietrich B, Golor G, Schrodter A, Keymeulen B, Lasseter
KC, Kipnes MS, Snyder K, Hilliard D, Tanen M, Cilissen C,
De Smet M, de Lepeleire I, Van Dyck K, Wang AQ, Zeng W,
Davies MJ, Tanaka W, Holst JJ, Deacon CF, Gottesdiener
KM, Wagner JA. Effect of single oral doses of sitagliptin, a
dipeptidyl peptidase‐4 inhibitor, on incretin and plasma glu­
cose levels after an oral glucose tolerance test in patients
with type 2 diabetes. J Clin Endocrinol Metab.
2006;91(11):4612–4619. Epub 2006/08/17. doi: 10.1210/
jc.2006‐1009. PubMed PMID: 16912128.
61. Nauck M. Incretin therapies: highlighting common features
and differences in the modes of action of glucagon‐like pep­
tide‐1 receptor agonists and dipeptidyl peptidase‐4  inhibi­
tors. Diabetes Obes Metab. 2016;18(3):203–216. Epub
2015/10/23. doi: 10.1111/dom.12591. PubMed PMID:
26489970; PMCID. PMC4785614.
62. Scheen AJ. Effects of glucose‐lowering agents on surrogate
endpoints and hard clinical renal outcomes in patients with
type 2 diabetes. Diabetes Metab. 2019;45(2):110–121. Epub
2018/11/28. doi: 10.1016/j.diabet.2018.10.003. PubMed
PMID: 30477733.
63. Deacon CF, Holst JJ. Dipeptidyl peptidase‐4  inhibitors for
the treatment of type 2 diabetes: comparison, efficacy and
safety. Expert Opin Pharmacother. 2013;14(15):2047–2058.
Epub 2013/08/08. doi: 10.1517/14656566.2013.824966.
PubMed PMID: 23919507.
64. Gallwitz B, Rosenstock J, Rauch T, Bhattacharya S, Patel S,
von Eynatten M, Dugi KA, Woerle HJ. 2‐year efficacy and
safety of linagliptin compared with glimepiride in patients
with type 2 diabetes inadequately controlled on metformin:
a randomised, double‐blind, non‐inferiority trial. Lancet.
2012;380(9840):475–483. Epub 2012/07/04. doi: 10.1016/
S0140‐6736(12)60691‐6. PubMed PMID: 22748821.
159
65. Del Prato S, Camisasca R, Wilson C, Fleck P. Durability of
the efficacy and safety of alogliptin compared with glipizide
in type 2 diabetes mellitus: a 2‐year study. Diabetes Obes
Metab. 2014;16(12):1239–1246. Epub 2014/08/19. doi:
10.1111/dom.12377. PubMed PMID: 25132212.
66. Goke B, Gallwitz B, Eriksson JG, Hellqvist A, Gause‐Nilsson
I. Saxagliptin vs. glipizide as add‐on therapy in patients with
type 2 diabetes mellitus inadequately controlled on met­
formin alone: long‐term (52‐week) extension of a 52‐week
randomised controlled trial. Int J Clin Pract. 2013;67(4):307–
316. Epub 2013/05/03. doi: 10.1111/ijcp.12119. PubMed
PMID: 23638466.
67. Seck T, Nauck M, Sheng D, Sunga S, Davies MJ, Stein PP,
Kaufman KD, Amatruda JM, Sitagliptin Study G. Safety and
efficacy of treatment with sitagliptin or glipizide in patients
with type 2 diabetes inadequately controlled on metformin:
a 2‐year study. Int J Clin Pract. 2010;64(5):562–576. Epub
2010/05/12. doi: 10.1111/j.1742‐1241.2010.02353.x.
PubMed PMID: 20456211.
68. Filozof C, Gautier JF. A comparison of efficacy and safety of
vildagliptin and gliclazide in combination with metformin
in patients with Type 2 diabetes inadequately controlled
with metformin alone: a 52‐week, randomized study. Diabet
Med. 2010;27(3):318–326. Epub 2010/06/12. doi:
10.1111/j.1464‐5491.2010.02938.x. PubMed PMID:
20536495.
69. Matthews DR, Dejager S, Ahren B, Fonseca V, Ferrannini E,
Couturier A, Foley JE, Zinman B. Vildagliptin add‐on to
metformin produces similar efficacy and reduced hypogly­
caemic risk compared with glimepiride, with no weight gain:
results from a 2‐year study. Diabetes Obes Metab.
2010;12(9):780–789. Epub 2010/07/24. doi:
10.1111/j.1463‐1326.2010.01233.x. PubMed PMID:
20649630.
70. Liu SC, Chien KL, Wang CH, Chen WC, Leung CH. Efficacy
and safety of adding pioglitazone or sitagliptin to patients
with type 2 diabetes insufficiently controlled with met­
formin and a sulfonylurea. Endocr Pract. 2013;19(6):980–
988. Epub 2013/06/29. doi: 10.4158/EP13148.OR. PubMed
PMID: 23807528.
71. Takihata M, Nakamura A, Tajima K, Inazumi T, Komatsu Y,
Tamura H, Yamazaki S, Kondo Y, Yamada M, Kimura M,
Terauchi Y. Comparative study of sitagliptin with pioglita­
zone in Japanese type 2 diabetic patients: the COMPASS
randomized controlled trial. Diabetes Obes Metab.
2013;15(5):455–462. Epub 2013/01/03. doi: 10.1111/
dom.12055. PubMed PMID: 23279373.
 160 Initial Evaluation and Management of Diabetes
72. Bae J, Kim G, Lee YH, Lee BW, Kang ES, Cha BS. Differential
effects of thiazolidinediones and dipeptidyl pepti­
dase‐4 inhibitors on insulin resistance and beta‐cell function
in type 2 diabetes mellitus: a propensity score‐matched anal­
ysis. Diabetes Ther. 2019;10(1):149–158. Epub 2018/12/07.
doi: 10.1007/s13300‐018‐0541‐y. PubMed PMID: 30506494;
PMCID. PMC6349276.
73. Emerging Risk Factors C, Sarwar N, Gao P, Seshasai SR,
Gobin R, Kaptoge S, Di Angelantonio E, Ingelsson E, Lawlor
DA, Selvin E, Stampfer M, Stehouwer CD, Lewington S,
Pennells L, Thompson A, Sattar N, White IR, Ray KK,
Danesh J. Diabetes mellitus, fasting blood glucose concen­
tration, and risk of vascular disease: a collaborative meta‐
analysis of 102 prospective studies. Lancet. 2010;375(9733):
2215–2222. Epub 2010/07/09. doi: 10.1016/S0140‐6736(10)
60484‐9. PubMed PMID: 20609967; PMCID. PMC2904878.
74. Home P. Cardiovascular outcome trials of glucose‐lowering
medications: an update. Diabetologia. 2019;62(3):357–369.
Epub 2019/01/05. doi: 10.1007/s00125‐018‐4801‐1. PubMed
PMID: 30607467.
75. Wang Q, Liu L, Gao L, Li Q. Cardiovascular safety of GLP‐1
receptor agonists for diabetes patients with high cardiovas­
cular risk: a meta‐analysis of cardiovascular outcomes trials.
Diabetes Res Clin Pract. 2018;143:34–42. Epub 2018/06/24.
doi: 10.1016/j.diabres.2018.06.009. PubMed PMID: 29935211.
76. American Diabetes A. 10. Cardiovascular disease and risk
management: standards of medical care in diabetes – 2020.
Diabetes Care. 2020;43(Suppl 1):S111–S34. Epub 2019/12/22.
doi: 10.2337/dc20‐S010. PubMed PMID: 31862753.
77. Rosenstock J, Perkovic V, Johansen OE, Cooper ME, Kahn
SE, Marx N, Alexander JH, Pencina M, Toto RD, Wanner C,
Zinman B, Woerle HJ, Baanstra D, Pfarr E, Schnaidt S,
Meinicke T, George JT, von Eynatten M, McGuire DK,
Investigators C. Effect of linagliptin vs placebo on major car­
diovascular events in adults with type 2 diabetes and high
cardiovascular and renal risk: the CARMELINA rand­
omized clinical trial. JAMA. 2019;321(1):69–79. Epub 2018/
11/13. doi: 10.1001/jama.2018.18269. PubMed PMID: 3041
8475; PMCID. PMC6583576.
78. Scirica BM, Bhatt DL, Braunwald E, Steg PG, Davidson J,
Hirshberg B, Ohman P, Frederich R, Wiviott SD, Hoffman
EB, Cavender MA, Udell JA, Desai NR, Mosenzon O,
McGuire DK, Ray KK, Leiter LA, Raz I, for the SAVOR‐
TIMI 53 Steering Committee and Investigators. Saxagliptin
and cardiovascular outcomes in patients with type 2 diabetes
mellitus. N Engl J Med. 2013;369(14):1317–1326. Epub 2013/
09/03. doi: 10.1056/NEJMoa1307684. PubMed PMID: 239
92601.
79. Schnabel CA, Wintle M, Kolterman O. Metabolic effects of
the incretin mimetic exenatide in the treatment of type 2
diabetes. Vasc Health Risk Manag. 2006;2(1):69–77. Epub
2007/02/27. doi: 10.2147/vhrm.2006.2.1.69. PubMed PMID:
17319471; PMCID. PMC1993968.
80. Holman RR, Paul SK, Bethel MA, Matthews DR, Neil HA.
10‐year follow‐up of intensive glucose control in type 2 dia­
betes. N Engl J Med. 2008;359(15):1577–1589. Epub 2008/
09/12. doi: 10.1056/NEJMoa0806470. PubMed PMID: 1878
4090.
81. American Diabetes Association. 9. Pharmacologic
approaches to glycemic treatment: standards of medical care
in diabetes – 2020. Diabetes Care. 2020;43(Suppl 1):S98–S110.
Epub 2019/12/22. doi: 10.2337/dc20‐S009. PubMed PMID:
31862752.
82. Drucker DJ, Buse JB, Taylor K, Kendall DM, Trautmann M,
Zhuang D, Porter L, for the DURATION‐1 Study Group.
Exenatide once weekly versus twice daily for the treatment
of type 2 diabetes: a randomised, open‐label, non‐inferiority
study. Lancet. 2008;372(9645):1240–1250. Epub 2008/09/11.
doi: 10.1016/S0140‐6736(08)61206‐4. PubMed PMID: 18
782641.
83. Blevins T, Pullman J, Malloy J, Yan P, Taylor K, Schulteis C,
Trautmann M, Porter L. DURATION‐5: exenatide once
weekly resulted in greater improvements in glycemic control
compared with exenatide twice daily in patients with type 2
diabetes. J Clin Endocrinol Metab. 2011;96(5):1301–1310.
Epub 2011/02/11. doi: 10.1210/jc.2010‐2081. PubMed PMID:
21307137.
84. Pratley RE, Nauck MA, Barnett AH, Feinglos MN, Ovalle F,
Harman‐Boehm I, Ye J, Scott R, Johnson S, Stewart M,
Rosenstock J, group Hs. Once‐weekly albiglutide versus
once‐daily liraglutide in patients with type 2 diabetes inad­
equately controlled on oral drugs (HARMONY 7): a ran­
domised, open‐label, multicentre, non‐inferiority phase 3
study. Lancet Diabetes Endocrinol. 2014;2(4):289–297. Epub
2014/04/08. doi: 10.1016/S2213‐8587(13)70214‐6. PubMed
PMID: 24703047.
PART III
Diagnosis and Management
of Cardiovascular Risk Factors
and Cardiovascular Disease
 13 Screening Patients with Prediabetes
and Diabetes for Cardiovascular Disease
Sabreen Ahmed, Saritha Tirumalasetty and Vivian Fonseca
Section of Endocrinology, Tulane University School of Medicine, New Orleans, LA, USA
LE A R NI NG P OI NT S
●● Patients with dysglycemia are at risk of developing
cardiovascular disease before and after the develop-
ment of overt diabetes.
●● Risk factor assessment will help stratify risk and guide
treatment.
●● Appropriate screening likely requires a combination of
risk factor assessment in combination with appropriate
functional or imaging studies intended to assess
cardiovascular risk.
●● Screening tests are not recommended for people with
very high risk or very low risk. However, they may be
useful for people at intermediate risk allowing
re-stratification and appropriate therapeutic approaches.
Introduction
Cardiovascular disease (CVD) continues to be the leading
cause of death accounting for up to 31% (17.9 million) of
global deaths [1]. The majority of this mortality is due to
myocardial infarction and strokes. The increasing preva-
lence correlates to a higher amount of procedural care. In
the United States alone, the economic burden has doubled
from 1996–1997 to 2014–2015 at which time direct costs
were estimated to be up to $213.8 billion [2]. People with
diabetes mellitus (DM) have been shown to have twice the
risk of CVD as the general population  [3]. The CDC’s
National Health and Nutrition Examination Survey
(NHANES) has shown an increase in the prevalence of
diagnosed and undiagnosed DM in both men and women
from 1988–1994 to 2013 to 2016 which corresponds to the
changes in diagnostic testing for DM namely the lowered
Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
threshold of fasting glucose to ≥ 126 mg/dL in 1997 and the
addition of hemoglobin A1c (HbA1c) ≥ 6.5% in 2010 [2, 4, 5].
CVD continues to be a significant cause of mortality
worldwide and in those with DM [6].
Patients with prediabetes in the San Antonio Heart
Study had atherogenic risk factors suggesting that risk for
CVD may start before the diagnosis of diabetes  [7]. In a
study of Finnish patients with and without diabetes fol-
lowed for seven years, the risk of a recurrent event was
greatest in diabetic patients with a known history of
CVD [8]. The risk was similar in patients with diabetes and
no history of CVD and patients without diabetes who had
a history of known CVD. This gave rise to the notion of
“cardiac risk equivalent” and the need to evaluate and treat
patients with diabetes in the same way as those with pre‐
existing heart disease. However, more recent data suggest
that risk varies depending on several factors including age
and duration and control of diabetes. Therefore, predicting
risk of events or screening for CVD in asymptomatic
patients may be an important strategy in appropriate pre-
vention of CV events.
The aim of this review is to discuss optimal identifica-
tion of patients at risk, current recommendations for such
identification, recent developments, and possible future
directions in such efforts.
Guidelines
Within guidelines published for the United States there is
variability in screening recommendations with the
American Diabetes Association (ADA) at this time not rec-
163
 164 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
ommending CVD screening in diabetics without symp-
toms and the American College of Cardiology/American
Heart Association (ACC/AHA) as well as ACC‐Imaging
Council proclaiming possible usefulness in screening  [9,
10, 11]. The ADA guidance is based on findings of the
Detection of Ischemia in Asymptomatic Diabetics (DIAD)
study, which describes resolution to ischemia on follow‐up
perfusion imaging. DIAD randomized asymptomatic
patients into “no imaging” or “cardiac perfusion imaging”
groups. Of all the patients who underwent perfusion imag-
ing, 22% had abnormal testing. Of these, 40% had less than
two risk factors for CVD. Three years later when the testing
was repeated, of those that had a negative test 10% now had
a positive test [12]. A more striking finding was the resolu-
tion of inducible ischemia in 79% of those that had
ischemia initially and no intervening revascularization
procedure. On follow‐up, 4.8 years after the original study,
the incidence of CV events was low. There was a significant
increase in primary prevention efforts in both groups and
there was no difference in mortality between those that had
myocardial perfusion testing and those that did not [13]. It
is important to recognize that the DIAD cohort had very
good control of CV risk factors, perhaps better than the
general population with diabetes.
Aggressive risk factor control is supported by the
Courage trial, which randomized patients into percutane-
ous coronary intervention (PCI) with medical therapy vs
medical therapy alone in order to gauge changes in future
cardiovascular event rates. At 5‐year follow‐up both study
groups met similar targets in pressure control < 130/85,
low‐density lipoprotein (LDL) < 85, and HbA1c goal of 7%.
Rates of primary outcomes including death from any cause
and non‐fatal myocardial infarctions (MI) were about
equal [14]. It is arguable that screening may lead to invasive
intervention which may be unnecessary in those with good
medical treatment. Even further secondary outcomes of
event rates were also similar despite the treatment group.
The BARI 2D study group produced similar findings with
no significant differences in all‐cause mortality or major
cardiac events even within different DM management
styles except for patients within the CABG arm [15]. Those
who underwent coronary artery bypass surgery (CABG)
had fewer major cardiovascular events when DM manage-
ment included insulin sensitization.
In contrast, possible utility in screening for cardiovascu-
lar disease may influence prevention treatments and deter-
mining risk for CVD events. The PREDICT and MESA
studies in particular showed a direct correlation between
high coronary artery calcium (CAC) scores and CVD risk
for both non‐diabetic and diabetic patients. Both studies
also may help with revealing low risk patients with very low
CAC scores requiring avoidance of invasive therapies.
These trials will be discussed in further depth in the CAC
section.
Hyperglycemia and risks of CVD
Glucose is a continuous variable in the population with an
almost linear association with CVD. This is seen even in
glucose concentrations below that used to diagnose diabe-
tes. DECODE, an epidemiologic study in Europe, looked at
over 25 000 subjects and demonstrated that mortality was
increased not just in patients with diabetes but also in those
with impaired glucose tolerance (IGT). The study also
demonstrated that IGT carries a greater risk than impaired
fasting glucose ((IFG), a finding of practical importance as
glucose tolerance testing is often not carried out in patients
at possible risk [16]. Norhammer et al. demonstrated that
in patients with established coronary artery disease (CAD),
IGT is common and may be missed [17]. This has led to
recommendations for more widespread use of GTTs.
Therefore, patients at risk for dysglycemia should be
screened early and have risk factor management early and
those with CVD should be screened for dysglycemia.
In patients with known diabetes, treatment of hypergly-
cemia reduces the risk of CVD Early prospective studies
such as DCCT and UKPDS [18, 19] showed improvement
in microvascular complications in patients with type 1 and
type 2 diabetes, respectively. When these patients were fol-
lowed, a decrease in macrovascular complications was also
demonstrated [20, 21]. The EDIC trial [20] describes long‐
term follow‐up of the patients in DCCT trial. In EDIC,
lower HbA1c while the patients were actively enrolled in
the DCCT trial appeared to correlate with improved car-
diovascular outcomes. It is important to recognize that the
DCCT and UKPDS trials enrolled relatively young patients
with short duration of DM. In addition, after the original
DCCT and UKPDS trials were terminated, the patients no
 Screening Asymptomatic Patients with Prediabetes and Diabetes for Cardiovascular Disease
longer maintained their randomization. Despite this, there
appeared to be an improvement in CVD outcomes in
intensively treated patients during long‐term follow‐up. In
three large studies (ACCORD, ADVANCE and VADT)
where the patients studied were older than those in UKPDS
and had DM for a longer duration tight blood sugar control
did not improve CV outcomes in the short term. Those
with long‐standing DM may have already developed irre-
versible and progressive atherosclerosis and management
of non‐glycemic factors takes on more importance.
The evidence supports the notion that elevated blood
sugars do increase risk for CVD and control of hyperglyce-
mia does improve that risk. The control of other risk fac-
tors also appears to be important, especially in those with
long‐standing type 2 diabetes.
Traditional risk factors
Several risk factors have been associated with CVD. The
INTERHEART study evaluated a number of these risks
and tried to assess the contribution of these risk factors to
the presence of CVD [22]. Nine commonly recognized risk
factors (smoking, abnormal lipids (apolipoprotein B1/
apolipoprotein A1 ratio), hypertension (self‐reported),
diabetes, abdominal obesity, psychosocial factors) were
evaluated. The percent attributable risk (PAR) for each risk
factor was calculated. In other words, these risks accounted
for 90% of the risk for CVD. The vast majority of risk is
accounted for by what have come to be termed “tradi-
tional” risk factors.
The COURAGE and BARI 2D trials as described previ-
ously showed that there was no significant difference in
death, non‐fatal MI and other cardiovascular event rates
between intensive medical therapy alone and medical ther-
apy plus PCI, although the patient populations had estab-
lished CAD  [14, 15]. When it comes to intensive or
conventional medical therapy, the Steno‐2 trial suggests
that intensive treatment of multiple risk factors reduced the
risk of vascular events [23]. 160 patients with diabetes and
microalbuminuria received conventional or intensive ther-
apy, which consisted of targets of HbA1c < 6.5%, blood
pressure < 140/85 mmHg and later < 130/80 mmHg, tri-
glycerides < 150 mg/dl, total cholesterol < 190 mg/dl, angi-
otensin‐converting enzyme inhibitor (ACE‐I) and aspirin.
165
Intensive treatment patients had a 20% absolute reduction
of CV death, nonfatal MI, CABG, PCI, peripheral vascular
disease, cerebrovascular accident (CVA) and amputations
after a mean of 7.8 years as well as a significant reduction in
microvascular complications. In a follow‐up study after 5.5
years where all patients were told to follow intensive treat-
ment regimen, the initial intensive therapy group had
fewer CV deaths and showed the importance of early risk
factor reduction.
Diabetes‐specific clinical risk predictors of
CVD
Individuals with diabetes have increased risk of cardiovas-
cular disease even when hyperglycemia is controlled [24].
A recent large observational cohort study evaluated 32611
patients with diabetes type 1 over 10.4 years and 17 risk
factors for MI, CVA, heart failure (HF) and death [25]. At
the end of the study, the strongest predictors of these end
points were albuminuria, glycohemoglobin, duration of
diabetes, systolic blood pressure (SBP), and LDL choles-
terol. Estimated glomerular filtration rate was also one of
the most important risk factors of HF hospitalization.
Lower HbA1c, SBP, and LDL levels had a linear correlation
with the outcomes other than HbA1c and all‐cause mortal-
ity, which raises the question of whether with tighter con-
trol the current guidelines might be beneficial.
Patients with diabetes are screened for nephropathy
with albumin‐to‐creatinine ratio. There are many trials
that demonstrate moderate albuminuria ≥ 30 mg/g of cre-
atinine is a risk factor for CVD, although the mechanism is
not well understood. As seen in the Heart Outcomes
Prevention Evaluation (HOPE) trial, moderate albuminu-
ria was associated with MI, CV death, CVA and HF hospi-
talizations even in patients without diabetes  [26]. Of the
1140 patients with diabetes and microalbuminuria, the use
of an ACE‐I reduced albuminuria and a 21% decrease in
CV outcomes according to the Microalbuminuria,
Cardiovascular, and Renal Outcomes (MICRO)‐HOPE
substudy [27]. This held true after adjusting for changes in
blood pressure.
Peripheral artery disease (PAD) is an indication that
there might be atherosclerosis in other areas of the body
and diabetes mellitus is a known risk factor. The UKPDS
 166 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
showed the prevalence correlated with longer diabetes
duration, increased diabetes severity and female, African
American, and Hispanic patients with diabetes  [19].
Theoretically, identifying PAD in asymptomatic patients
with diabetes would benefit from aggressive risk factor
modification. The ADA currently recommends that a
screening ankle brachial index be done if patients with dia-
betes over the age of 50 every 5 years. However, a rand-
omized clinical trial is needed to determine which patients
with diabetes to screen and if screening in these asympto-
matic patients would be clinically beneficial.
Nontraditional risk factors
Although the majority of CVD appears to be explained by
traditional risk factors, there is residual risk that may be
due to the presence of other factors. These have been
termed “non‐traditional” risk factors. These may be useful
in determining risk in patients who by history would be at
high risk for CVD but have normal values for the tradi-
tional risk factors. As our understanding of the mecha-
nisms of CVD in diabetes has improved, the role of
inflammation in development of both DM and CVD has
been recognized. In addition, abnormal fibrinolysis,
endothelial function, alteration in adipokine release, and
vascular wall abnormalities may all play a role in vascular
dysfunction in patients with diabetes [28]. As these mecha-
nisms are investigated, novel risk factors have been identi-
fied. A number of interventions we institute for traditional
risk factors such as statins and ACE‐I appear to improve
these nontraditional risk factors.
Although promising, the nontraditional factors are not
ready for routine use in all patients. There are multiple reasons
for this, such as lack of standardized commercially available
tests, lack of data on CVD event reduction following risk
improvement etc. Of all the novel risks, high‐sensitivity C‐
reactive protein (hsCRP) and coronary calcium scores may be
the only ones that are currently clinically applicable.
Screening tools
Coronary artery calcium score
As described in the ACC/AHA and ACC‐Cardiac Imaging
guidelines, CAC may be useful in ascribing cardiovascular
event risk and developing treatment plans for said patients.
In the 2004 Raggi et  al. study referred > 10 000 patients
(903 of them with DM) for EBT to determine CAC scores.
In a 5‐year follow‐up all‐cause mortality was higher in the
DM population, which also had higher CAC scores  [29].
Of significance, the DM population in Raggi’s study also
had older patients with more hypertension and tobacco use
than those without DM.
In the 2008 PREDICT study 589 patients with DM type
2 and no history of CVD from Central and West London
had CAC scores measured and were followed up annually
for 4 years. Black African patients were excluded in this
study due to low coronary heart disease (CHD) rates at the
time of this study. In analysis of CAC scores and primary
endpoints, there is a logarithmic association between
increasing CAC scores and first CHD or stroke event in
these patients [30]. Furthermore when taking into account
risk calculator models, CAC scores seem to further enhance
the prediction models of Framingham and UKPDS.
More recently the Multi‐Ethnic Study of Atherosclerosis
(MESA) study reviewed a large population of varying eth-
nicities including Caucasian, African American, Hispanic,
and Chinese from 6 communities across the United States.
Over 6600 people with metabolic syndrome (MetS) with-
out DM, DM, and individuals without MetS or DM were
assessed for CAC (using EBT vs CT) and Carotid Intimal
Medial Thickness (CIMT) using ultrasound. After about a
6.4‐year follow‐up CHD and CVD endpoints were
assessed. They had a median of 6.4‐year follow‐up. CAC
with traditional risk factors were found to significantly be
predictive of CHD and CVD events at all CAC levels with
increased events at higher CAC scores  [31]. Although
CIMT did show an increase in events over time this
increase was not statistically significant except for in CHD
events in the 4th quartile for patients without MetS or DM.
Of note in diabetic patients a doubling of events occurred
when CAC scores reached above 400. Further follow‐up
after 11–12 years was done within the MESA study. Patients
with DM and MetS had significantly higher rates of CHD
and ASCVD events compared to DM alone, MetS alone,
and those with neither MetS nor DM [32]. It is notable that
those with DM alone had more of these events than MetS
alone; those with neither condition had the lowest event
rate. While specifically looking at the range of CAC scores,
 Screening Asymptomatic Patients with Prediabetes and Diabetes for Cardiovascular Disease
those with higher scores specifically CAC > 400 showed
the same trend in event rates. Upon extrapolation of the
Diabetic population data with CAC scores > 400, those
with higher CHD/ASCVD event rates were found in
patients with long‐term insulin use and DM duration of
over 10 years. Furthermore DM patients with a CAC score
of 0 had less CHD and ASCVD events compared to CAC of
> 0  with Hazard Ratios of 0.35 and 0.43 respectively.
Interestingly DM patients with CAC scores of > 400 and
tight control of HbA1c to < 7 were found to have higher
CHD/ASCVD rates.
Thus, it appears that CAC is a highly predictive for coro-
nary atherosclerosis, and has emerged as the primary rec-
ommended test to evaluate risk for future atherosclerotic
cardiovascular disease, as it is an easy to perform, cost
effective and reproducible means of quantifying plaque in
the coronary arteries. As discussed above, the ACC and
AHA cholesterol and prevention guidelines now include
CAC as part of the clinical algorithm in asymptomatic peo-
ple for planning primary prevention interventions such as
statins and aspirin. On the other hand for people with dia-
betes such therapy is appropriate anyway and the test may
be less imperative. Nevertheless, when done it is important
for clinicians and patients alike to be aware that a calcium
score of zero indicates virtually no risk, and the recom-
mendation is to consider no statin, with re‐assessment in
5–10 years. A CAC score of 1–99 favors statin (especially
after age 55) and a CAC of 100+ or > 75th percentile is an
indication to initiate statin therapy.
Coronary computed tomography angiography
The development of high‐resolution multidetector coro-
nary computed tomography angiography (CCTA) is able to
provide coronary anatomy noninvasively and determine
the severity of atherosclerosis  [33]. The CORE 64 study
demonstrated that CCTA and invasive coronary angiogra-
phy have a high correlation of determining CAD severity
and extent  [34]. In the FACTOR‐64 study, a randomized
clinical trial, 900 patients with DM type 1 or 2 of at least
3–5 years’ duration and without CAD were randomized to
either CCTA screening or standard of care (SBP <
130 mmHg, LDL < 100 mg/dL and HbA1c < 7.0%) [33]. Of
the patients who received CCTA, 31% had no luminal dis-
ease, 58% had mild, 12% had moderate, and 11% were
167
found to have severe stenosis. All patients with ≥ 10% ste-
nosis received aggressive therapy (SBP < 120 mmHg, LDL
< 70 mg/dL, HDL > 50 mg/dL, triglyceride < 150 mg/dL,
and HbA1c < 6.0%). Patients with severe stenosis ≥ 70%
had diagnostic coronary angiography and patients with
moderate stenosis ≥ 50% received stress cardiac imaging
and then angiography if needed. This resulted in 8% of
patients who had CCTA undergoing angiography and 5.8%
needing revascularization. After a mean follow‐up of 4 ±
1.7 years, there was no statistically significant difference in
primary event outcomes (all‐cause mortality, nonfatal
myocardial infarction and hospitalization for unstable
angina) in the CCTA arm vs. standard care, 6.2% and 7.6%
respectively.
CCTA may be valuable in providing long‐term prognos-
tic information like CAC for asymptomatic patients with
diabetes as demonstrated in several studies. An observa-
tional study of 591 patients with diabetes and without signs
or symptoms of CVD studied the event‐free survival at the
end of a median of 5.3 years, defined as no cardiac death,
nonfatal MI, unstable angina, or late revascularization [35].
After a CCTA, patients were categorized into no luminal
stenosis (28%), nonobstructive CAD (40%) and obstruc-
tive CAD ≥ 50% (32%). The event‐free survivals were 99.3
± 0.7%, 96.7 ± 1.2% and 86.2 ± 3.0% (p < 0.001), respec-
tively. Similarly, in a prospective study, 390 patients with
type 1 and 2 diabetes with suspected CAD (new chest pain,
abnormal stress test, multiple cardiovascular risk factors)
had CCTA and were followed up after 62 ± 9 months for
end points of cardiac death, nonfatal MI and unstable
angina [36]. It also showed patients without any CAD had
an excellent prognosis and can help predict cardiac events.
The event‐free survival was 100% for normal coronary
arteries, 78% for nonobstructive CAD, and 60% for
obstructive CAD. These conclusions are supported by
other studies [37, 38].
In a prospective multicenter observational cohort study
(400 of the patients had diabetes and were asymptomatic
for CVD), CCTA findings were classified into per‐patient
maximal stenosis, number of vessels with ≥ 50% stenosis
and segment stenosis score  [39]. It was determined that
increases in these were associated with incremental risk
estimates of MACE after adjusting for CAD risk factors
and CACS, improving risk classification. Although these
 168 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
studies suggest that CCTA is predictive of adverse events, it
does not address whether there is a benefit of invasive
revascularization in the obstructive group.
Stress testing – Myocardial perfusion imaging
The DIAD study noted no significant differences in CV
events and mortality rates in patients receiving myocardial
perfusion screening versus no imaging [12, 13]. Similarly,
in the Do You Need to Assess Myocardial Ischemia in
Type‐2 diabetes (DYNAMIT) trial of 631 asymptomatic
patients with diabetes with at least two additional CV risk
factors, 21.5% of the patients screened with symptom‐lim-
ited bicycle exercise and radionuclide myocardial perfu-
sion imaging (rMPI) had silent ischemia and after mean
follow‐up of 3.5 years, there was no significant difference
in death from all‐cause, nonfatal MI, nonfatal stroke, or HF
requiring intervention [40]. The BARDOT trial took 400
asymptomatic type 2 diabetic patients and performed myo-
cardial perfusion single‐photon emission computed
tomography (MPS) at the beginning of the study and after
two years [41]. A substudy of patients with abnormal MPS
were randomly assigned to medical versus invasive and
medical treatment. As expected, intervention on a stenosed
lesion resulted in reduction of subclinical progression to
silent CAD and scintigraphic ischemia, though not rate or
MACE, cardiac death, and MI.
Stress testing – Echocardiogram for CAD
An open label randomized pilot study of 141 people with
type 2 diabetes without CVD either were in the control arm
or had screening with an exercise electrocardiogram and
dipyridamole stress echocardiography (DSE) with subse-
quent coronary angiography if either testing was abnor-
mal  [42]. 21.1% of patients had a positive DSE. After a
mean of 53.5  month, the absolute risk reduction in the
screened group was 15.8% and the study concluded that a
proportion of all cardiac events was significantly lower in
this group (P = 0.018).
One randomized study compared the positive predictive
value of dobutamine stress echocardiography (DE) to
stress myocardial scintigraphy (SPECT)  [43]. It consisted
of 204 individuals with diabetes type 2 without symptoms
but at high risk of CVD. Of the patients that underwent
SPECT, 13% had silent MI and 4% had significant CAD
while in the DE group, it was 11% and 5%, respectively.
These findings were not significant. Furthermore, the posi-
tive predictive value of detecting significant CAD between
the two also was not significant. After 3 years, cardiovascu-
lar death and MI rates were comparable. Of note, 4% of all
the participants had significant CAD revealed on
angiography.
Though promising, currently it is not recommended to
screen with rMPI or stress echo. Many studies have shown
prevalence and diagnosis of CHD in diabetes patients.
However, most prognostic studies were in patients who
were symptomatic and more studies are needed in asymp-
tomatic patients and of the prognosis after invention.
Carotid ultrasound and endothelial function
studies
CIMT was originally proposed as an alternative method to
determine cardiovascular risk without ionizing radiation.
It also had the advantage of being closely related to insulin
resistance  [44]. However, concerns over reproducibility,
and difficulty in obtaining high quality imaging in practice
have resulted in this test to a Class III (no benefit) recom-
mendation from the ACC and AHA for general screening
of CVD risk. CIMT can therefore be used to assess cardio-
vascular risk in populations where risk estimation with tra-
ditional risk factors does not apply, such as populations
with specific genetic mutations in lipid metabolism.
Younger patients (where CAC testing is not useful, due to
lack of time to develop calcification) with familial hyper-
cholesterolemia (FH) represent a population in which
CIMT has proven to be a valuable tool.
Similarly, methods for evaluating endothelial function
(brachial artery reactivity) and tonometry etc. are useful in
research studies but less useful in clinical practice [45].
Screening for HF
Patients with diabetes have over two times increased risk
of HF compared to patients without DM [46]. In a cross‐
sectional study in the Netherlands of 581 patients with
type 2 diabetes and over 60 years old, 27.7% were found to
have HF that was not previously diagnosed  [47]. 22.9%
had preserved ejection fraction and 4.8% had reduced
ejection fraction. However, the patients were not
asymptomatic.
 Screening Asymptomatic Patients with Prediabetes and Diabetes for Cardiovascular Disease
Currently, there are not any recommendations to screen
for HF in patients with diabetes. In a recent cluster analysis
based on baseline echocardiography variables, data from
two large prospective cohorts were used of 745 subjects
with type 2 diabetes without heart disease [48]. Cluster 1
patients had the lowest hypertrophy, high LVEF and myo-
cardial strain. They were predominantly male with the low-
est comorbidity rates including obesity and HTN. Cluster
2 had the worst diastolic dysfunction, the patient popula-
tion were mostly elderly and female with the highest BP,
BMI, and heart rate. Finally, cluster 3 had the lowest hyper-
trophic systolic dysfunction and strain where patients were
largely male and similar to cluster 1 in terms of age, HTN
and obesity. After a median follow‐up of 67  months, CV
mortality and hospitalization (arrhythmia, HF, CAD) were
increased in clusters 2 and 3  with similar hazard ratios.
Based on these results, a patient’s clinical features are not
helpful in risk stratifying patients.
Studies have suggested that microalbuminuria may pos-
sibly be used as a prescreening tool for echocardiography
(ECHO)  [49]. The Strong Heart Study divided 1748
American Indian patients with type 2 diabetes into no
albuminuria, microalbuminuria, and macroalbuminuria.
After adjusting for age, sex, SBP, BMI, diabetes duration,
CAD, and LV mass, results showed that the degree of dias-
tolic dysfunction was proportional to the level of microal-
buminuria. The HOPE study mentioned earlier
demonstrated a correlation between albuminuria and
CVD, including HF [27].
Another proposed tool is N‐terminal pro‐B‐type natriu-
retic peptide (NT‐proBNP). 300 patients with diabetes
type 2 and without heart disease but NT‐proBNP > 125 pg/
ml were randomized into a control group or intervention
group in the NT‐proBNP selected prevention of cardiac
events in a population of diabetic patients without a history
of cardiac disease (PONTIAC) trial  [50]. The intervened
group had beta blockers and renin‐angiotensin system
(RAS) antagonists titrated up until the maximum recom-
mended or tolerated dose was reached or NT‐proBNP con-
centrations were below normal or decreased by 50%. At the
end of one year, there was no significant decrease in NT‐
proBNP concentrations between the two but a significant
decrease in HF hospitalization, CV hospitalization and
CV death.
169
Further studies are needed to recommend if NT‐
proBNP and microalbuminuria can identity patients with
diabetes who may benefit for HF screening
Risk calculators
Risk calculators have been used to stratify patients with
CVD risk factors. Previous prediction tools include those
based on the Framingham Heart Study and UKPDS.
Limitations to these calculators included small diabetic
population size and short length of disease diagnosis for
each study, respectively leading to difficulty in assessing
risk. With more recent studies, newer calculators have been
developed.
In 2013 the ACC/AHA guidelines developed a new
model for risk assessment based on pooled data from the
following studies: ARIC (Atherosclerosis Risk in
Communities) study, the CHS (Cardiovascular Health
Study), and CARDIA (Coronary Artery Risk Development
in Young Adults) study, combined with applicable data
from the Framingham Original and Offspring Study.
Together these studies consisted of 9098  non‐Hispanic
white men, 1647 African‐American men, 11 240  non‐
Hispanic white women, and 2641 African‐American
women between 40 and 79 years of age who had no previ-
ous history of an MI, stroke, CHF, PCI, CABG, or afib [51].
Key components to the ACC/AHA calculator include the
addition of race and diabetes which were not included in
the Framingham risk score  [52]. Furthermore this 2013
ASCVD evaluation tool includes stroke outcome in risk
prediction has led to improved identification of at‐risk
female patients.
The large population of the MESA trial, patients of vary-
ing ethnicities from across the US were followed for an
average of 10 years to cardiovascular mortality endpoints.
From this follow‐up and integration of CAC scores algo-
rithm, The Mesa Risk score, was developed. The algorithm
tool takes into account traditional Framingham risk factors
(age, sex, smoking, total cholesterol, HDL cholesterol, sys-
tolic blood pressure, and antihypertensive medication use)
and includes diabetes, family history of MI, race, and CAC
score. Notably CAC score integration with traditional risk
factors led to improvement in risk prediction (Harrell’s C‐
statistic 0.8 vs 0.75; p‐value < 0.0001) [53]. This improve-
 170 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
ment to prediction was evidenced in two external
validations studies Heinz Nixdorf Recall Study (HNR) and
Dallas Heart Study (DHS). For the validation study > 3500
Caucasian patients from Germany and > 1000 patients of
varying ethnicities from Dallas County in Texas, US
respectively were included [54–56]. The DHS cohort repre-
sents the diversity of the MESA population more than the
HNR cohort. However, the prediction results still improved
in both patient populations. Use of The MESA Risk score
on HNR and DHS studies resulted in good discrimination
and calibration with Harrell’s C‐statistic scores of 0.779 and
0.816 respectively.
Based on another large population derivation and vali-
dation study, the PREDICT risk estimator enrolls a large
New Zealand’s primary care population (over 400 000 par-
ticipants) in order to estimate cardiovascular risk [57]. This
population included the Maori population as well as Asian,
Pacific, and Caucasian New Zealanders. Patients with
known cardiovascular disease, renal disease, and conges-
tive HF were excluded. Apart from other risk calculators,
this estimator includes the traditional CHD, MI, and stroke
outcomes as well as congestive HF.
Specifically in patients with diabetes, risk models devel-
oped in the past are derived from data from the United
Kingdom Prospective Diabetes Study (UKPDS) which
dates back to populations of the UK in the 1970s. At that
time around 80% of the UK’s population was Caucasian [19].
The 2018 BRAVO study developed an updated risk estima-
tion tool based on data from the ACCORD trial which
derived data from the US consisting of over 10 000 partici-
pants with diabetes who were followed for about 4
years [58]. The BRAVO calculator includes the importance
of racial/ethnic background and medical history such as
history of MI, stroke, angina, CHF, and previous revascu-
larization for estimations in a variety of outcomes.
Endpoint measurements of this study included stroke,
CHF, non‐fatal MI, angina end‐stage renal disease (ESRD),
blindness, revascularization surgery, severe pressure sensa-
tion loss, all‐cause mortality, and CVD death. Notably a
history of severe hypoglycemia was included for predictors
of certain outcomes. Internal validation for this study
showed close fit Kaplan Meier curves within 95% confi-
dence intervals for microvascular and macrovascular com-
plications. The external validation process was conducted
against the ASPEN, ADVANCE, and CARDS trials. This
process when plotted had an estimated R2 of 0.86. With F‐
test application of the fitted line to the line of 100% accu-
racy, there was no statistically significant difference (p =
0.82).
Another interesting risk prediction tool new in 2018 is
the Astro‐CHARM calculator which was developed to esti-
mate CV risk for astronauts. When building this estimator,
patient populations were pooled from the MESA trial,
DHS, Prospective Army Coronary Calcium project
(PACC) and aimed to calculate risk in a middle‐aged popu-
lation (ages 40–65 years). All pooled cohorts measured
CAC; some participants also had high sensitivity c‐reactive
protein (hs‐CRP) measured. The prediction tool for the
first time incorporates CAC scores and hs‐CRP measure-
ments, along with various risk factors similar to the ones
mentioned for other calculators, to develop a novel estima-
tor [59]. In internal validation against the risk factor model
only, c‐statistic scores were 0.784, 0.720, 0.813 (P=0.0002),
and 0.817 (P ≤ 0.0001) for the risk factor model, CAC
model, risk factor model+CAC, and full Astro‐CHARM
model respectively. A modified model without hs‐CRP was
found to have a c‐statistic of 0.826. In external validation
against the FHS cohorts demonstrated c‐statistics of 0.78
and 0.79 for models with and without hs‐CRP
respectively.
Summary
Recommendations for screening for CVD have undergone
many changes in the last decade, and guidelines vary
among medical organizations. Better quality evidence to
support some recommendations have emerged that help us
individualize risk prediction to guide therapy. For most
people with diabetes, screening is still of limited value, as
screening imaging modalities may lead to both false posi-
tives and negatives and high‐risk patients may be better
served by aggressive risk factor modification as suggested
by a number of guidelines.
Studies have shown that the sodium‐glucose co‐trans-
porter 2 (SGLT‐2) inhibitors and some glucagon‐like‐pep-
tide‐1 (GLP‐1) receptor agonists, specifically liraglutide,
dulaglutide and semaglutide, decreased the risk of cardiovas-
cular death, nonfatal MI and nonfatal stroke. SGLT‐2 inhib-
 Screening Asymptomatic Patients with Prediabetes and Diabetes for Cardiovascular Disease
itor dapagliflozin reduced HF hospitalizations. However,
most of the patients in these studies already had established
cardiovascular disease or multiple CV risk factors. There is
not enough data to suggest that these medications are also
beneficial for primary prevention in patients with asymp-
tomatic type 2 diabetes, with fewer risk factors. In such
patients, screening may determine the appropriateness of
deferring such therapy or other risk lowering therapy if
patients so prefer. For example, a negative CAC may allow
a clinician to defer aggressive therapy for a few years. For
patients with atypical symptoms screening for CVD may
help determine the need for more invasive testing and risk
factor reduction.
References
1. World Health Organization. Cardiovascular Disease (CVDs).
https://www.who.int/news‐room/fact‐sheets/detail/
cardiovascular‐diseases‐(cvds) (accessed 29 September 2020).
2. Heart disease and stroke statistics‐2019 update: a report
from the American Heart Association. Circulation.
2019;139(10):e56–e528.
3. Buse JB, Ginsberg HN, Bakris GL et al. Primary prevention
of cardiovascular diseases in people with diabetes mellitus: a
scientific statement from the American Heart Association
and the American Diabetes Association. Diabetes Care.
2007;30(1):162–172.
4. Expert Committee on the Diagnosis and Classification of
Diabetes Mellitus. Report of the Expert Committee on the
Diagnosis and Classification of Diabetes Mellitus. Diabetes
Care. 1997;20(7):1183–1197.
5. American Diabetes Association. Standards of medical care in
diabetes – 2010 [published correction appears in Diabetes
Care. 2010;33(3):692]. Diabetes Care. 2010;33(Suppl
1):S11–S61.
6. Pasquel FJ, Gregg EW, Ali MK. The evolving epidemiology
of atherosclerotic cardiovascular disease in people with
diabetes. Endocinol Metab Clin N Am. 2018;47(1):1–32.
7. Haffner SM, Stern MP, Hazuda HP et al. Cardiovascular risk
factors in confirmed prediabetic individuals: does the clock
for coronary heart disease start ticking before onset of
clinical diabetes? JAMA. 1990;263(21):2893–2898.
8. Haffner SM, Lehto S, Ronnemaa T et al. Mortality from
coronary heart disease in subjects 0with type 2 diabetes
mellitus and in nondiabetic subjects with and without prior
myocardial infarction. N Engl J Med. 1998;339(4):229–234.
171
9. American Diabetes Association. 10. Cardiovascular disease
and risk management: standards of medical care in diabetes
– 2020. Diabetes Care. 2020;43(Suppl 1):
S111–S134.
10. Arnett DK, Blumenthal RS, Albert MA et  al. 2019 ACC/
AHA Guideline on the primary prevention of cardiovascular
disease: Executive Summary: a report of the American
College of Cardiology/American Heart Association Task
Force on clinical practice guidelines. J Am Coll Cardiol.
2019;74(10):1376–1414.
11. Budoff MJ, Raggi P, Beller GA et al. Noninvasive cardiovas-
cular risk assessment of the asymptomatic diabetic patient:
the Imaging Council of the American College of Cardiology.
JACC Cardiovasc Imaging. 2016;9(2):176–192.
12. Wackers FJ, Chyun DA, Young LH et  al. Resolution of
asymptomatic myocardial ischemia in patients with type 2
diabetes in the Detection of Ischemia in Asymptomatic
Diabetics (DIAD) study. Diabetes Care. 2007;30(11):
2892–2898.
13. Young LH, Wackers FJ, Chyun DA et al. Cardiac outcomes
after screening for asymptomatic coronary artery disease in
patients with type 2 diabetes: the DIAD study: a randomized
controlled trial. JAMA. 2009;301(15),1547–1555. 2009.
14. Boden WE, O’Rourke RA, Teo KK et  al. COURAGE Trial
Research Group. Optimal medical therapy with or without
PCI for stable coronary disease. N Engl J Med.
2007;356:1503–1516.
15. BARI 2D Study Group, Frye RL, August P et  al. A rand-
omized trial of therapies for type 2 diabetes and coronary
artery disease. N Engl J Med. 2009;360:2503–2515.
16. The Decode Study Group. Is the current definition for dia-
betes relevant to mortality risk from all causes and cardio-
vascular and noncardiovascular disease. Diabetes Care.
2003;26(3):688–696.
17. Norhammer A, Tenerz A, Nilsson G et al. Glucose metabo-
lism in patients with acute myocardial infarction and no pre-
vious diagnosis of diabetes mellitus: a prospective study.
Lancet. 2002;359(9324):2140–2144.
18. The Diabetes Control and Complications Trial Research Group.
The effect of intensive treatment of diabetes on the develop-
ment and progression of long‐term complications in insulin‐
dependent diabetes mellitus. N Engl J Med. 1993;329(14):
977–986.
19. UK Prospective Diabetes Study (UKPDS) Group. Intensive
blood sugar control with sulphonylureas or insulin com-
pared with conventional treatment and risk of complications
in patients with type 2 diabetes (UKPDS 33). Lancet.
1998;352:837–853.
 172 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
20. The Diabetes Control and Complications Trial/Epidemi­ology
of Diabetes Interventions and Complications (DCCT/EDIC)
Study Research Group. Intensive diabetes treatment and car-
diovascular disease in patients with type 1 diabetes. N Engl J
Med. 2005;353(25):2643–2653.
21. Holman RR, Paul SK, Bethel AM et al. 10‐year follow‐up of
intensive glucose control in type 2 diabetes. N Engl J Med.
2008;359(15):1577–1589.
22. Yusuf S, Hawken S, Ounpuu S et  al. Effect of potentially
modifiable risk factors associated with myocardial infarction
in 52 countries (the interheart study):case‐control study.
Lancet. 2004;364(9438):937–952.
23. Gaede P, Vedel P, Larsen N et al. Multifactorial intervention
and cardiovascular disease in patients with type 2 diabetes.
N Engl J Med. 2003;348(5):383–393.
24. Lind M, Svensson AM, Kosiborod M et al. Glycemic control
and excess mortality in type 1 diabetes. N Engl J Med. 2014;
371:1972–1982.
25. Rawshani A, Rawshani A, Sattar N et al. Relative prognostic
importance and optimal levels of risk factors for mortality
and cardiovascular outcomes in type 1 diabetes mellitus.
Circulation. 2019;139(16):1900–1912.
26. Gerstein HC, Mann JF, Yi Q et al. Albuminuria and risk of
cardiovascular events, death, and heart failure in diabetic
and nondiabetic individuals. JAMA. 2001;286(4):421–426.
27. Heart Outcomes Prevention Evaluation Study Investigators.
Effects of ramipril on cardiovascular and microvascular out-
comes in people with diabetes mellitus: results of the HOPE
study and MICRO‐HOPE substudy. Lancet. 2000;355:
253–259.
28. Fonseca V, Desouza C, Asnani S et  al. Nontraditional risk
factors for cardiovascular disease in diabetes. Endocrine
Reviews. 2004;25(1):153–175.
29. Raggi, P, Shaw LJ, Berman DS et al. Prognostic value of coro-
nary artery calcium screening in subjects with and without
diabetes. J Am Coll Cardiol. 2004;43(9):1663–1669.
30. Elkeles RS, Godsland IF, Feher MD et al. Coronary calcium
measurement improves prediction of cardiovascular events
in asymptomatic patients with type 2 diabetes: the PREDICT
study. Eur Heart J. 2008;29(18):2244–2251.
31. Malik, S, Budoff MJ, Katz R et al. Impact of subclinical ath-
erosclerosis on cardiovascular disease events in individuals
with metabolic syndrome and diabetes: the multi‐ethnic
study of atherosclerosis. Diabetes Care. 2011;34(10):
2285–2290.
32. Malik, S, Zhao Y, Budoff M et  al. Coronary artery calcium
score for long‐term risk classification in individuals with type
2 diabetes and metabolic syndrome from the multi‐ethnic
study of atherosclerosis. JAMA Cardiol. 2017;2(12):
1332–1340.
33. Muhlestein JB, Lappé DL, Lima JAC et al. Effect of screening
for coronary artery disease using ct angiography on mortal-
ity and cardiac events in high‐risk patients with diabetes: the
FACTOR‐64 randomized clinical trial. JAMA. 2014;312(21):
2234–2243.
34. Miller JM, Rochitte CE, Dewey M et al. Diagnostic perfor-
mance of coronary angiography by 64‐row CT. N Engl J Med.
2008;359:2324–2336.
35. Kang SH, Park G‐M, Lee S‐W et al. Long‐term prognostic
value of coronary ct angiography in asymptomatic type 2
diabetes mellitus. JACC. Cardiovascular Imaging. 2016;9(11):
1292–1300.
36. Andreini D, Pontone G, Mushtaq S et al. Prognostic value of
multidetector computed tomography coronary angiography
in diabetes. Diabetes Care. 2013;36(7):1834–1841.
37. Blanke P, Naoum C, Ahmadi A et al. Long‐term prognostic
utility of coronary ct angiography in stable patients with dia-
betes mellitus. JACC. Cardiovascular Imaging. 2016;9(11):
1289–1291.
38. Werkhoven JMV, Cademartiri F, Seitun S et  al. Diabetes:
prognostic value of CT coronary angiography – comparison
with a nondiabetic population. Radiology. 2010;256(1):
83–92.
39. Min, JK, Labounty TM, Gomez MJ et al. Incremental prog-
nostic value of coronary computed tomographic angiogra-
phy over coronary artery calcium score for risk prediction of
major adverse cardiac events in asymptomatic diabetic indi-
viduals. Atherosclerosis. 2014;232(2):298–304.
40. Lièvre, MM, Moulin P, Thivolet C et al. Detection of silent
myocardial ischemia in asymptomatic patients with diabe-
tes: results of a randomized trial and meta‐analysis assessing
the effectiveness of systematic screening. Trials. 2011;12:23.
41. Zellweger MJ, Maraun M et al. Progression to overt or silent
CAD in asymptomatic patients with diabetes mellitus at high
coronary risk: main findings of the prospective multicenter
BARDOT trial with a pilot randomized treatment substudy.
JACC Cardiovasc Imaging. 2014;7(10):1001–1010.
42. Jacqueminet S, Barthelemy O, Rouzet F et al. A randomized
study comparing isotope and echocardiography stress test-
ing in the screening of silent myocardial ischaemia in type 2
diabetic patients. Diabetes Metab. 2010;36(6):463–469.
43. Faglia E, Manuela M, Antonella Q et  al. Risk reduction of
cardiac events by screening of unknown asymptomatic coro-
nary artery disease in subjects with type 2 diabetes mellitus
 Screening Asymptomatic Patients with Prediabetes and Diabetes for Cardiovascular Disease
at high cardiovascular risk: an open‐label randomized pilot
study. Am Heart J. 2005;149(2):1–6.
44. Fonseca V, Desouza C, Asnani S et  al. Nontraditional risk
factors for cardiovascular disease in diabetes. Endocrine
Reviews. 2004;25(1):153–175.
45. DeSouza CA, Shapiro LF, Clevenger CM et al. Regular aero-
bic exercise prevents and restores age‐related declines in
endothelium‐dependent vasodilation in healthy men.
Circulation. 2000;102(12):1351–1357.
46. Dei Cas A, Khan SS, Butler J et  al. Impact of diabetes on
epidemiology, treatment, and outcomes of patients with
heart failure. JACC Heart Fail. 2014;3(2):126–145.
47. Boonman‐de Winter LJ, Rutten FH, Cramer MJ et al. High
prevalence of previously unknown heart failure and left ven-
tricular dysfunction in patients with type 2 diabetes.
Diabetologia. 2012;55(8):2154–2162.
48. Ernande L, Audureau E, Jellis CL et al. Clinical implications
of echocardiographic phenotypes of patients with diabetes
mellitus. J Am Coll Cardiol. 2017;70(14):1704–1716.
49. Liu JE, Robbins DC, Palmieri V et al. Association of albumi-
nuria with systolic and diastolic left ventricular dysfunction
in type 2 diabetes: the Strong Heart Study. J Am Coll Cardiol.
2003;41(11):2022–2028.
50. Huelsmann M, Neuhold S, Resl M et  al. PONTIAC (NT‐
proBNP selected prevention of cardiac events in a popula-
tion of diabetic patients without a history of cardiac disease):
a prospective randomized controlled trial. J Am Coll Cardiol.
2013;62(15):1365–1372.
51. Karmali KN, Goff DC, Ning H et al. A systematic examina-
tion of the 2013 ACC/AHA pooled cohort risk assessment
tool for atherosclerotic cardiovascular disease. J Am Coll
Cardiol. 2014;64(10):959–968.
52. Goff DC Jr, Lloyd‐Jones DM, Bennett G et  al. 2013 ACC/
AHA guideline on the assessment of cardiovascular risk: a
report of the American College of Cardiology/American
Heart Association Task Force on Practice Guidelines
173
[published correction appears in J Am Coll Cardiol. 2014;Jul 1;
63(25 Pt B):3026]. J Am Coll Cardiol. 2014;63(25 Pt
B):2935–2959.
53. McClelland RL, Jorgensen NW, Budoff M et al. 10‐year coro-
nary heart disease risk prediction using coronary artery cal-
cium and traditional risk factors: derivation in the MESA
(Multi‐Ethnic Study of Atherosclerosis) with validation in
the HNR (Heinz Nixdorf Recall) Study and the DHS (Dallas
Heart Study). J Am Coll Cardiol. 2015;66(15):1643–1653.
54. Schmermund A, Möhlenkamp S, Stang A et al. Assessment
of clinically silent atherosclerotic disease and established
and novel risk factors for predicting myocardial infarction
and cardiac death in healthy middle‐aged subjects: rationale
and design of the Heinz Nixdorf RECALL Study. Am Heart
J. 2002;144 (2):212–218.
55. Stang A, Moebus S, Dragano N et al. Baseline recruitment
and analyses of nonresponse of the Heinz Nixdorf Recall
Study: identifiability of phone numbers as the major deter-
minant of response. Eur J Epidemiol. 2005;20(6):489–496.
56. Victor RG, Haley RW, Willett DL et  al. The Dallas Heart
Study: a population‐based probability sample for the multi-
disciplinary study of ethnic differences in cardiovascular
health. Am J Cardiol. 2004;93(12):1473–1480.
57. Pylypchuk R, Wells S, Kerr A et al. Cardiovascular disease
risk prediction equations in 400 000 primary care patients in
New Zealand: a derivation and validation study. The Lancet.
2018;391(10133):1897–1907.
58. Shao H, Fonseca V, Stoecker C et al. Novel risk engine for
diabetes progression and mortality in USA. Building,
Relating, Assessing, and Validating Outcomes (BRAVO).
PharmacoEconomics. 2018;36:1125–1134.
59. Khera A, Budoff MJ, O’Donnell CJ et  al. Astronaut
Cardiovascular Health and Risk Modification (Astro‐
CHARM) coronary calcium atherosclerotic cardiovascular
disease risk calculator. Circulation. 2018;138(17):1819–1827.
doi:10.1161/CIRCULATIONAHA.118.033505.
 14 Choosing Medications for Type 2 Diabetes –
What Weighting Should Be Given
to Cardiovascular Risk Reduction?
Adrian Vella
Professor of Medicine, Division of Endocrinology, Diabetes, Metabolism, and Nutrition, Mayo Clinic, Rochester, MN, USA
LE A R N I N G P OI NT S
●● Cardiovascular risk reduction is an integral part of
diabetes management.
●● This should be achieved with lipid‐lowering and
antihypertensive agents in addition to pharmacother-
apy intended to achieve glucose lowering.
●● Some agents may have direct effects on cardiovascular
outcomes over and above their effects on glycemic
control.
Diabetes and cardiovascular risk:
how does it affect therapeutic goals?
Diabetes mellitus causes morbidity, mortality and health-
care costs in excess of ~$170 billion/year  [1]. Chronic
hyperglycemia leads to the micro‐ and macrovascular
complications that frequently accompany diabetes. The
Diabetes Control and Complications Trial (DCCT) was
able to show that “tight” glucose control i.e. glucose con-
centrations closer to the normal physiologic range (as
measured by Hemoglobin A1c) compared to “conventional”
control prevented or delayed retinopathy and nephropathy
in people with type 1 diabetes  [2]. Since that landmark
study the goal of diabetes management has been to safely
achieve good glycemic control while avoiding hypoglyce-
mia. Although it is reasonable to extrapolate the conclu-
sions of the DCCT to the management of type 2 diabetes,
Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
174
the effects of glycemic control in type 2 diabetes on cardio-
vascular risk are less certain. In fairness, similar findings
have been reported by the United Kingdom Prospective
Diabetes Study group (UKPDS) for type 2 diabetes [3, 4].
Since hypoglycemia can also produce acute and chronic,
potentially catastrophic, complications, the goal of diabetes
therapy has been to safely achieve glucose concentrations
as close to the normal physiologic range as possible while
avoiding hypoglycemia.
Cardiovascular disease (CVD) is the leading cause of
morbidity and mortality in the world regardless of race,
ethnicity, or gender  [5], and costs related to CVD in the
United States surpassed $300 billion in 2010 [6]. In spite of
success in reducing the mortality from acute complications
of heart disease, the incidence of atherosclerosis and its
burden has continued to grow. The estimated annual inci-
dence of a myocardial infarction is 785 000  while 470
000  will suffer a recurrent acute coronary syndrome.
Depending on their gender and clinical outcome, individu-
als who survive the acute stage of a myocardial infarction
have 1.5 to 15 times higher probability of illness and death
compared to the general population.
Patients with T2DM are at an increased risk of adverse
cardiovascular events compared to those without diabe-
tes  [7], and despite efforts at risk reduction including
smoking cessation and blood pressure and cholesterol opti-
mization, the majority of people with T2DM continue to
die from cardiovascular disease [8]. Observational studies
have shown a significant association between glycemic

control and CVD  [9–12]. However, results from rand-
omized clinical trials have been more controversial. The
UK Prospective Diabetes Study (UKPDS) showed that
HbA1c, was an independent predictor of cardiovascular
outcomes [13]; the ADVANCE study showed that intensive
glucose control (HbA1c < 6.5 %) was associated with a
reduction in major microvascular events, but not in major
macrovascular events  [14]; the VADT study showed that
intensive glucose control had no significant effect on rates
of major cardiovascular events, death, or microvascular
complications  [15] while the ACCORD trial showed that
aggressive glycemic control targeting normal HbA1c was
associated with an increased mortality but did not reduce
major cardiovascular events [16].
There are multiple caveats to these observations. Over
the past two decades non‐diabetes therapies that success-
fully mitigate cardiovascular risk by targeting cardiovascu-
lar risk factors such as hypertension and dyslipidemia as
well as the proliferation of diabetes therapies have improved
glycemic control and cardiovascular risk to the point that
demonstrating benefit of other interventions has paradoxi-
cally made it more difficult to demonstrate benefit. It is
also important to appreciate that the prediabetic state is
associated with high cardiovascular risk conferred by dys-
lipidemia, obesity, and hypertension despite comparatively
minimal hyperglycemia [17].
The logical conclusions one can make from these obser-
vations is that diabetes is associated with adverse cardio-
vascular risk and ameliorating cardiovascular risk is a
central goal of therapy. Glycemic control is also important
and should not be neglected. Some agents used to treat dia-
betes may affect cardiovascular risk beyond their ability to
lower glucose concentrations (see Figure  14.1). The evi-
dence for each therapeutic class of “non‐insulin” therapy
will be discussed in the relevant sections of this chapter.
Metformin
Metformin is the only biguanide approved for use in the
United States and seems to improve insulin action in
hepatic and peripheral tissues. The mechanism of action of
this compound remains unknown. However, as would be
anticipated from its known efficacy of improving insulin
action in patients with type 2 diabetes, fasting and
Choosing Medications for Type 2 Diabetes 175
­ ostprandial insulin secretion decrease with metformin
p
therapy. Hypoglycemia in patients treated with metformin
monotherapy is unlikely and should prompt a search for
other causes of hypoglycemia. Because of its ease of use and
general safety, metformin is usually considered to be first‐
line pharmacotherapy in the treatment of type 2 diabetes.
Its use has also been advocated in people with
prediabetes [18].
The primacy of this compound in the treatment of type
2 diabetes arises out of the UK Prospective Diabetes Study
(UKPDS) which demonstrated that it decreased the risk of
microvascular complications with less associated weight
gain and fewer episodes of hypoglycemia compared to
insulin or sulfonylureas [4]. It may be the first‐line phar-
macological therapy of choice in these patients. In this
study, subjects in the intensive treatment arm had less fatal
and non‐fatal myocardial infarction after 10 years of fol-
low‐up although no significant differences attributable to
treatment modality could be ascertained [13]. Other direct
comparisons with sulfonylureas or other compounds
have  shown no clear evidence of superiority for
metformin [19].
Sulfonylureas
Sulfonylureas bind to the potassium channel complex
(encoded by KCJN11/ABCC8) causing decreased potas-
sium influx, depolarization of the β‐cell membrane and a
subsequent increase of calcium flux into the cell. This
results in exocytosis of insulin from secretory gran-
ules [20]. Unlike incretin‐based therapy, discussed later in
this chapter, this effect is not glucose‐dependent and all
patients treated with sulfonylureas are theoretically at risk
of hypoglycemia. Good initial response to therapy followed
by subsequent failure is often referred to as secondary fail-
ure. This may be due to patient‐related factors (poor com-
pliance with therapy/lifestyle, co‐morbidities that preclude
sulfonylurea use) as well disease‐progression. Although a
similar rate of secondary failure with metformin has been
observed in the UKPDS [21], durability of effect was infe-
rior to metformin or thiazolidinedione therapy in the
ADOPT study [22].
The safety of this class of compounds was first ques-
tioned by the University Group Diabetes Program trial
 176 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease

SU Metformin GLP-1RA
Magnitude of Glycemic lowering

TZDs

SGLT-2i
DPP-4i
Bromocriptine
Acarbose

Pramlintide Colesevelam

Evidence of direct Cardiovascular benefit*

FIG 14.1  Therapeutic classes clustered by therapeutic effect versus the level of evidence supporting a direct cardio­vascular benefit.
SU = sulfonylurea; TZDs = Thiazolidinediones; GLP‐1RA = Glucagon-Like Peptide‐1 Receptor Agonists; SGLT‐2i = Sodium‐
Glucose cotransporter‐2 inhibitors; DPP‐4i = DiPeptidyl Peptidase‐4 inhibitors. *Position on the axis represents the strength of the
current evidence not the magnitude of cardiovascularbenefit.

where tolbutamide was associated with increased mortal-
ity [23]. Tolbutamide is no longer used in clinical practice.
However, in the UKPDS outcomes were similar to those of
metformin [3]. In a meta‐analysis of 116 randomized clini-
cal trials, Monami et al. reported no effect of major cardio-
vascular events and a trend towards benefit in myocardial
infarction [24]. On the other hand, glimepiride increased
the risk of stroke. More notable are that these observations
depend on multiple relatively underpowered studies.
While comparisons versus placebo or no therapy are
favorable, this was not the case against other comparators
such as dipeptidyl peptidase‐4  inhibitors. Nevertheless,
these compounds are associated with a risk of hypoglyce-
mia and they should be used cautiously in patients with
established, active heart disease. Well‐powered cardiovas-
cular outcome trials are still needed to establish the safety
of this class of compounds.
Thiazolidinediones
The peroxisome proliferator activated receptor (PPAR) is
activated by fatty acids and fatty acid‐derived eicosanoids.
It exists in three isoforms with different tissue specificity.
When activated, PPAR acts as a transcriptional enhancer of
genes with PPAR‐response elements  [25]. The PPAR‐α
isoform is expressed in the liver, heart, and brown fat where
active fatty acid catabolism occurs. PPAR‐γ is expressed in
adipose tissue although it is also expressed in the vascula-
ture. PPAR‐δ is more generally expressed. PPAR‐α regu-
lates the transcription of genes important in fatty acid
uptake and oxidation. Fibrates used to treat dyslipidemia,
act as PPAR‐α agonists [26].
Thiazolidinediones are PPAR‐γ agonists and stimulate
adipocyte differentiation as well as the expression of glu-
cose transporters (GLUT‐1 and GLUT‐4), increasing
hepatic and peripheral glucose uptake [27]. The first agent
in this class, troglitazone, was withdrawn after reports of
hepatotoxicity  [28]. This seems to be unique to troglita-
zone. Clinical use of these agents has been associated with
weight gain. This is in part due to stimulation of adipocyte
differentiation as well as fluid retention which can be sig-
nificant and may be to a degree that results in hemodilu-
tion  [25]. For this reason, thiazolidinediones are
contraindicated in congestive cardiac failure [29].

The discontinuation of troglitazone allowed randomiza-
tion of previously treated patients to either rosiglitazone or
pioglitazone [30]. This study reported that in the absence
of differences in glycemic control pioglitazone lowered
total and LDL‐cholesterol whereas rosiglitazone produced
a slight increase in these parameters. Nissen et  al. subse-
quently reported that a meta‐analysis of published studies
suggested that rosiglitazone increased risk of myocardial
infarction and cardiovascular mortality  [31]. Because of
this and other data the Food and Drug Administration
(FDA) placed significant restrictions on rosiglitazone use.
These restrictions were only rescinded after and open‐label
non‐inferiority study specifically designed to assess cardio-
vascular outcomes [32] showed no increased risk of mor-
tality or heart attack. The BARI 2D trial also reported no
increase in risk of myocardial infarction associated with
rosiglitazone [33].
There is significant in vitro; and rodent data to support
a favorable effect of thiazolidinediones on vascular biol-
ogy  [25]. Indeed, in humans, pioglitazone slows progres-
sion of carotid intima‐media thickness [34] and decreased
coronary atheroma [35]. A secondary cardiovascular pre-
vention study utilizing pioglitazone failed to show benefit
for a primary composite cardiovascular endpoint. However,
the study was able to fulfil its secondary endpoint which
was a composite of death, heart attack, and cerebrovascular
events compared to placebo  [36]. Post hoc analyses in
patients with a previous heart attack and stroke show a
decrease in recurrent events with pioglitazone  [37, 38].
This was borne out by the Insulin Resistance Intervention
after Stroke (IRIS) trial [39].
Against these benefits one must balance two potential
adverse events associated with thiazolidinedione use. The
first is that of bladder cancer which seems to be increased
after more than 2 years of drug exposure [40]. Bladder can-
cer is so uncommon, however, that randomized controlled
trials are insufficiently powered to detect these effects. For
example, the PROactive study cited above suggested an
association that did not persist after 6 years of follow‐
up  [41]. In the interests of framing the prospective risk
appropriately, an individual’s annual risk of bladder cancer
goes from 0.043% to 0.049% when on pioglitazone or more
than 20 000 would need to be treated to cause 1 new case of
bladder cancer [25, 42].
Choosing Medications for Type 2 Diabetes 177
The other potential adverse event associated with this
therapeutic class is that of fracture risk. The ADOPT (A
Diabetes Outcome Progression Trial) study demonstrated
an increased risk of limb fractures in pre‐ and postmeno-
pausal women [22]. At present it appears that the benefits
of decreased cardiovascular mortality outweigh the mor-
tality attributed to osteoporotic fracture [43].
Acarbose
Acarbose is an α‐glycosidase inhibitor which prevents the
luminal digestion of polysaccharides and, therefore, delays
the absorption of monosaccharides such as glucose from
the gut. It has also been argued that part of its effect arises
from the increased delivery of undigested carbohydrate to
the terminal ileum thereby stimulating Glucagon‐like pep-
tide‐1 (GLP‐1) secretion by the intestinal L‐cells  [44].
Since acarbose does not directly interfere with insulin
secretion, hypoglycemia does not occur in diabetic patients
using this medication as monotherapy. However, if hypo-
glycemia (due to another medication e.g. sulfonylurea)
does occur, pure glucose should be used to treat this, as
absorption of disaccharides or more complex carbohy-
drates will be delayed [45].
The Study to Prevent Non‐Insulin Dependent Diabetes
Mellitus (STOP‐NIDDM Trial) randomized ~1400 sub-
jects with impaired fasting glucose or diabetes (≥ 100 mg/
dL (5.6 mmol/L) but ≤ 140 mgd/L (7.8 mmol/L) and with
impaired glucose tolerance to placebo or acarbose taken
with each meal (three times a day). The investigators sub-
sequently reported that acarbose decreased the incidence
of diabetes, cardiovascular disease and hypertension [46].
The incidence of pre‐specified) 2° endpoints of myocardial
infarction (1 vs 12) and other cardiovascular events (15 vs.
32) was decreased. These results have been the subject of
some controversy [47].
Subsequently, the Acarbose Cardiovascular Evaluation
(ACE) study in ~6500 patients with established coronary
heart disease and impaired glucose tolerance demonstrated
no effect of acarbose on the prevention of cardiovascular
events [48]. In this cohort of people with impaired glucose
tolerance, acarbose decreased the incidence of diabetes and
improved glucose tolerance. There is no evidence to sug-
gest that this benefit persists after cessation More recently,
 178 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease

acarbose was compared to sulfonylureas as adjunctive ther- of pramlintide differ between type 1 and type 2 diabetes.
apy for people with type 2 diabetes [49]. The investigators Currently it is approved as adjunctive therapy, adminis-
demonstrated a decrease in major atherosclerotic events tered pre‐prandially with prandial insulin. Its use is associ-
and decreased hospitalization for hypoglycemia but no ated with satiety and a mild decrease in weight  [57, 58].
effect on all‐cause mortality. There are no published cardiovascular outcomes data for
this compound. However, some relatively small studies
suggest that there are no direct effects on cardiovascular
Colesevelam
risk factors such as blood pressure and lipoprotein
Bile‐acid sequestrants such as colesevelam and cholesty- concentrations [59].
ramine increase fecal excretion of bile acids by interrupting
the enterohepatic circulation. This leads to a decrease in Bromocriptine
the bile acid pool resulting in diversion of hepatic choles-
A quick‐release formulation of bromocriptine (a central
terol synthesis to the production of bile acids with concom-
Dopamine D2 receptor agonist used to treat prolactinoma)
itant upregulation of hepatic (low‐density lipoprotein)
is approved for the treatment of type 2 diabetes. The mecha-
LDL‐cholesterol receptors  [50]. These actions result in a
nism by which bromocriptine produces glucose lowering is
small, but significant, lowering of cholesterol concentra-
uncertain. Gaziano et  al. randomized 3095 patients with
tions. Intriguingly, their use in people with type 2 diabetes
type 2 diabetes (BMI of < 43 kg/m2, and an A1C concentra-
has resulted in ~0.3% decrease in HbA1c  [51–53].
tion ≤ 10.0%) to bromocriptine or placebo as an adjunct to
Consequently, colesevelam has been approved as adjunctive
the patient’s usual diabetes therapy  [60]. The drug caused
therapy for type 2 diabetes. In humans it lowers fasting
nausea but did not alter weight. However, a composite end-
glucose concentrations likely through an insulin‐sensitizing
point consisting of stroke, coronary revascularization, myo-
effect on the liver. Postprandial hepatic glucose uptake is
cardial infarction, and hospitalization for angina or heart
also enhanced [54].
failure was significantly decreased by bromocriptine therapy.
Although there is general consensus as regards the cho-
Intriguingly, in this study, bromocriptine resulted in a 0.1%
lesterol‐lowering properties of these compounds, evidence
increase in HbA1c (versus a 0.2% increase in the placebo
that this results in improved outcomes is relatively sparse.
arm) and a significant decrease (5%) in triglycerides. Other
For example, analysis of the healthcare insurance claims
studies have reported a ~0.6% decrease in HbA1c [61].
data of a third‐party payer compared a composite cardio-
The mechanism of action whereby bromocriptine use
vascular event score which included myocardial infarction,
results in glucose‐lowering is somewhat speculative. It is
stroke, angina, or revascularization in patients taking
suggested by the observation in hibernating mammals that
ezetimibe or colesevelam. After 12  months of treatment,
the transition to an insulin‐resistant state during hiberna-
subjects taking colesevelam had a lower event score (odds
tion is governed by dopamine concentrations in hypotha-
ratio (OR) 0.54, 95 % confidence interval (CI) 0.30–0.97)
lamic nuclei. In humans who do not experience such
compared to those taking ezetimibe [55].
seasonal changes in their metabolic state it is speculated
that stimulation of central dopamine signaling may restore
Pramlintide circadian misalignment [62]. Small, short‐term studies in
humans have suggested an effect on insulin sensitivity but
Pramlintide is a synthetic derivative of amylin – a product this has not been consistently observed [63].
of the β‐cell that is co‐secreted with insulin  [56].
Concentrations are elevated in people with type 2 diabetes
Glucagon‐like peptide‐1‐based therapy
and low or undetectable in people with type 1 diabetes [57].
Pharmacologic concentrations of amylin and its analogue The physiological basis for Glucagon‐like peptide‐1
inhibit gastric emptying through a central vagally‐mediated (GLP‐1)‐based therapy is outside the scope of this chapter
mechanism. Intriguingly, the dose‐response characteristics [64]. Nevertheless, it is important to recognize the basis of

the two therapeutic classes in this category [65]. GLP‐1 has
a short half‐life in the circulation (secondary to its rapid
inactivation by dipeptidyl peptidase 4 – DPP‐4 – an
enzyme which is widely distributed and cleaves all peptides
whose second N‐terminal amino acid is a proline or ala-
nine). Accordingly, inhibition of DPP‐4 impairs clearance
of all DPP‐4 substrates. From a glycemic point of view the
most important is GLP‐1. The rise in active GLP‐1 seen
with DPP‐4 inhibition is transient and there are no associ-
ated gastrointestinal side‐effects or weight loss [66, 67].
In contrast, GLP‐1 receptor agonists comprise modified
forms of GLP‐1 that render them resistant to the actions of
DPP‐4 or analogues of GLP‐1 capable of binding and stimu-
lating the GLP‐1 receptor. For example, exenatide is a natu-
rally occurring analogue of GLP‐1 that is found in the saliva of
large reptiles such as Heloderma horridum. Its second amino
acid is a glycine, making it resistant to degradation by DPP‐4,
conferring a half‐life of ~2 hours as opposed to the 2–5 min-
utes of native GLP‐1. Detectable concentrations are present
for up to 10 hours after administration. GLP‐1 receptor ago-
nists have more powerful glucose‐lowering effects than do
DPP‐4 inhibitors (a decrease of HbA1c of ~1% vs. ~0.5% asso-
ciated with DPP‐4i monotherapy). They produce weight loss
although there is a significant incidence of gastrointestinal
side‐effects such as nausea and vomiting [68, 69].
DPP‐4  inhibition can affect circulating concentrations
of other peptide mediators that are potentially substrates
for DPP‐4  [70]. Indeed, this might explain the (uncom-
mon) interaction between angiotensin‐converting enzyme
inhibitors and DPP‐4  inhibitors to produce angi-
oedema [71]. On the other hand, differences in selectivity
for other members of the dipeptidyl peptidase family were
originally thought to explain potential differences in side‐
effect profiles of these therapeutic agents [72]. In practice,
significant side‐effects attributable to this class of agents
has been hard to document. However, it is possible that (a
lack of) dipeptidyl peptidase might explain between‐com-
pound differences in cardiovascular outcomes trials.
For example, in the SAVOR‐TIMI study the primary
end point was a composite of death from cardiovascular
causes, myocardial infarction, or ischemic stroke.
Saxagliptin did not have any effect on this endpoint. The
secondary endpoint, a composite of death from cardiovas-
cular causes, myocardial infarction, ischemic stroke, hospi-
Choosing Medications for Type 2 Diabetes 179
talization for unstable angina, coronary revascularization,
or heart failure was similarly unaffected  [73]. However,
more patients in the saxagliptin group than in the placebo
group were hospitalized for heart failure (3.5% vs. 2.8%,
according to 2‐year Kaplan–Meier estimates; hazard ratio,
1.27; 95% CI, 1.07 to 1.51; p = 0.007). Non‐inferiority for
safety compared to placebo was also reported for aloglip-
tin [74] and sitagliptin [75] with no significant effects on
cardiovascular endpoints. No signal in terms of heart fail-
ure was observed in these trials.
The mechanism underlying the observation of heart
failure associated with saxagliptin remains unexplained. Of
note, a study of 254 patients with a left ventricular ejection
fraction of ~30% were randomized to placebo or vildaglip-
tin. After a year of therapy vildagliptin slightly, but signifi-
cantly increased left ventricular end‐diastolic and
end‐systolic volume. Ejection fraction was unchanged [76].
Overall, to date, the data would support cardiovascular
safety but little direct benefit of DPP‐4 inhibitors on car-
diovascular disease. Recent data has suggested that the
increased risk of hospitalization for heart failure observed
in SAVOR‐TIMI could be explained by deteriorating renal
function [77].
Marso et al. reported that liraglutide decreased the inci-
dence of a composite endpoint (cardiovascular death, non-
fatal myocardial infarction or nonfatal stroke). There was
also a lower risk of all‐cause mortality and microvascular
complications than were observed in the placebo group.
The authors estimated that 66 people would need to be
treated for 3 years to prevent a cardiovascular death [78].
This has not been observed with lixisenatide [79] or exena-
tide  [80], which, while non‐inferior to placebo, had no
effects on cardiovascular endpoints.
In addition to liraglutide, semaglutide studied in
patients with cardiovascular disease (~80% of the study
population) or with multiple risk factors, successfully
decreased the primary composite endpoint (cardiovascular
death, nonfatal myocardial infarction or nonfatal stroke).
This was driven by a decrease in cerebrovascular events
and a trend towards decreased myocardial infarction [81].
Albiglutide also decreased the identical composite primary
endpoint – this time driven by a decrease in myocardial
infarction  [82]. Dulaglutide also produced similar
effects [83].
 180 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
Although the tolerability of GLP‐1 receptor agonists has
improved over time, gastrointestinal distress remains the
most prominent side‐effect that leads to discontinuation.
Cost may also be a barrier to their use. The other potential
consideration is patient resistance to injectable therapy,
although the advent of once‐weekly injectable therapy may
make this less of an obstacle.
SGLT‐2i
Sodium glucose cotransporter 2 inhibitors (SGLT‐2i) are a
new class of anti‐hyperglycemic agents that function to
concomitantly inhibit the reabsorption of glucose and
sodium in the renal proximal convoluting tubule  [84].
These compounds produce glycosuria and natriuresis,
which translates into approximately a 0.7–1.0 % reduction
in circulating Hemoglobin A1c, approximately a 5/2 mm Hg
blood pressure reduction, and a 2–3 kg loss in body weight,
a 30–40% reduction in albuminuria via a reduction in
intra‐glomerular pressure, and other favorable metabolic
effects  [85]. By promoting glycosuria, these compounds
also induce calorie wastage and induce a mild degree of
weight loss. Nevertheless, the use of an SGLT‐2i may result
in potential risk, with increased risk for genitourinary tract
infections, diabetic ketoacidosis, acute kidney injury, frac-
tures, and atraumatic below‐knee lower extremity amputa-
tion [86, 87].
It is notable that the actions of SGLT‐2i are independent
of the integrity of the insulin secretory apparatus, and can
theoretically lower glucose in people with type 1 diabetes.
Unexpectedly, in addition to an improvement in HbA1c,
treatment with SGLT‐2i decreases cardiovascular out-
comes. Trials in patients with type 2 diabetes mellitus and
high CVD risk have recently reported reductions in major
adverse cardiovascular events, specifically the composite of
CV mortality, nonfatal myocardial infarction, and nonfatal
stroke, and particular benefit in reducing hospitalization
for heart failure, namely the EMPA‐REG OUTCOME trial,
and the CANVAS Program (CANVAS and
CANVASS‐R)  [86, 88, 89]. The mechanisms of these
actions are uncertain at the present time but are the subject
of active investigation  [90–92]. Intriguingly, the EMPA‐
HEART CardioLink‐6 Randomized Clinical Trial demon-
strated that empagliflozin decreased left ventricular mass
in diabetic patients with heart failure, 6 months after initia-
tion of treatment [93].
The CREDENCE (Canagliflozin and Renal Events in
Diabetes With Established Nephropathy Clinical
Evaluation) study randomized 4401 participants with type
2 diabetes mellitus and chronic kidney disease to canagli-
flozin or placebo and demonstrated improved cardiovas-
cular endpoints regardless of whether it was used in
patients without or with established cardiovascular dis-
ease  [94]. The DECLARE‐TIMI (Dapagliflozin and
Cardiovascular outcomes in Type 2 Diabetes) studied
patients with cardiovascular risk factors or established car-
diovascular disease (~40%) of the study population over a
4‐year period. It failed to reduce the composite of stroke,
myocardial infarction, or cardiovascular death. However, a
composite of cardiovascular death and hospitalization for
heart failure was significantly reduced. This was a pre‐
specified endpoint where the decrease was driven by
decreased heart failure [95].
The findings from these studies need to be kept in per-
spective. The cardiovascular benefit of these compounds,
while impressive, may have been driven by the relatively
high‐risk populations (established CVD at baseline) stud-
ied. In the above referenced studies, the difference in glyce-
mia between the treatment groups was minimal, suggesting
that extra‐glycemic effects of the drugs are responsible for
the reduction in CVD events  [96]. Indeed, the Dapa‐HF
trial, designed to study the effect of dapagliflozin on car-
diovascular mortality and worsening heart failure, demon-
strated benefit in patients without (and with) type 2
diabetes [97]. The other notable observation for this class
of compounds was that the side‐effects observed were rela-
tively minor and did not lead to discontinuation of
therapy.
Conclusions
It is important to appreciate that modern diabetes therapy
is multi‐faceted and the clinician has multiple options that
may allow better individualization of pharmacotherapy for
a given individual. Recently, the emergence of data to sug-
gest that specific compounds may have direct cardiovascu-
lar benefit has re‐emphasized the importance of
cardiovascular risk reduction in the management of type 2
 Choosing Medications for Type 2 Diabetes 181

diabetes (Figure 14.1). Accordingly, it is also important to References

remember the importance of lifestyle modification and
1. Albright AL, Gregg EW. Preventing type 2 diabetes in
appropriate management of dyslipidemia and
communities across the U.S.: the National Diabetes
hypertension. Prevention Program. American Journal of Preventive
Glycemic control per se has established benefits on Medicine. 2013;44:S346–351.
micro‐ and macrovascular disease although the benefits on 2. Diabetes C, Complications Trial Research G, Nathan DM,
the latter are less clear in type 2 diabetes than they are in Genuth S, Lachin J, Cleary P, Crofford O, Davis M, Rand L,
type 1 diabetes. While recent data suggests direct benefits Siebert C. The effect of intensive treatment of diabetes on the
of SGLT‐2i and GLP‐1 receptor agonists it is important to development and progression of long‐term complications in
remember that there is evidence of direct benefit for other insulin‐dependent diabetes mellitus. N Engl J Med.
compounds and also that there are some limitations to the 1993;329:977–986.
3. UK Prospective Diabetes Study (UKPDS) Group. Intensive
generalizability of these findings – especially so, given the
blood‐glucose control with sulphonylureas or insulin
reliance on composite endpoints.
compared with conventional treatment and risk of
With these caveats, however, it may be reasonable to
complications in patients with type 2 diabetes (UKPDS 33).
give preferential consideration to therapeutic classes Lancet. 1998;352:837–853.
depending on the clinical context (see Table  14.1). For 4. UK Prospective Diabetes Study (UKPDS) Group. Effect of
example, the presence of weight gain or abnormal eating intensive blood‐glucose control with metformin on
patterns, the presence of heart failure or renal disease or complications in overweight patients with type 2 diabetes
lipodystrophy [98, 99] may all suggest a therapeutic choice (UKPDS 34). Lancet. 1998;352:854–865.
in addition to standard first‐line therapy. 5. Roger VL, Go AS, Lloyd‐Jones DM, Benjamin EJ, Berry JD,
Borden WB, Bravata DM, Dai S, Ford ES, Fox CS, Fullerton
TABLE 14.1  Patient characteristics that might affect prescribing. HJ, Gillespie C, Hailpern SM, Heit JA, Howard VJ, Kissela
BM, Kittner SJ, Lackland DT, Lichtman JH, Lisabeth LD,
Patient factors that Therapeutic class Makuc DM, Marcus GM, Marelli A, Matchar DB, Moy CS,
might affect
To avoid To use Mozaffarian D, Mussolino ME, Nichol G, Paynter NP,
prescribing
Soliman EZ, Sorlie PD, Sotoodehnia N, Turan TN, Virani SS,
Hypoglycemia risk Sulfonylureas DPP‐4i Wong ND, Woo D, Turner MB, and on behalf of the
Acarbose* GLP‐1RA American Heart Association Statistics Committee and Stroke
SGLT‐2i Statistics Subcommittee. Executive summary: heart disease
Lipodystrophy TZDs and stroke statistics – 2012 update: a report from the
Heart failure TZDs SGLT‐2i American Heart Association. Circulation. 2012;125:188–197.
Obesity TZDs† GLP‐1RA
6. Lloyd‐Jones D, Adams RJ, Brown TM, Carnethon M, Dai S,
Nephropathy Sulfonylureas SGLT‐2i
De Simone G, Ferguson TB, Ford E, Furie K, Gillespie C, Go
Active cardiac disease Sulfonylureas TZDs
GLP‐1RA A, Greenlund K, Haase N, Hailpern S, Ho PM, Howard V,
SGLT‐2i Kissela B, Kittner S, Lackland D, Lisabeth L, Marelli A,
Dyslipidemia despite Colesevelam McDermott MM, Meigs J, Mozaffarian D, Mussolino M,
statin use Nichol G, Roger VL, Rosamond W, Sacco R, Sorlie P, Stafford
Multiple daily Pramlintide R, Thom T, Wasserthiel‐Smoller S, Wong ND, Wylie‐Rosett
injections of insulin J, and on behalf of the American Heart Association Statistics
Committee and Stroke Statistics Subcommittee. Executive
TZDs = Thiazolidinediones; GLP‐1RA = Glucagon‐Like Peptide‐1
summary: heart disease and stroke statistics – 2010 update: a
Receptor Agonists; SGLT‐2i = Sodium‐Glucose cotrans-
report from the American Heart Association. Circulation.
porter‐2 inhibitors; DPP‐4i = DiPeptidyl Peptidase‐4 inhibitors.
*
 Acarbose might alter the ability to treat hypoglycemia not 2010;121:948–954.
cause hypoglycemia. 7. Sarwar N, Gao P, Seshasai SR, Gobin R, Kaptoge S, Di
†
 TZDs may exacerbate weight gain but still improve insulin Angelantonio E, Ingelsson E, Lawlor DA, Selvin E, Stampfer
action and glycemic control overall. M, Stehouwer CD, Lewington S, Pennells L, Thompson A,
 182 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
Sattar N, White IR, Ray KK, Danesh J. Diabetes mellitus,
fasting blood glucose concentration, and risk of vascular
disease: a collaborative meta‐analysis of 102 prospective
studies. Lancet. 2010;375:2215–2222.
8. Gu K, Cowie CC, Harris MI. Mortality in adults with and
without diabetes in a national cohort of the U.S. population,
1971–1993. Diabetes Care. 1998;21:1138–1145.
9. Khaw KT, Wareham N, Bingham S, Luben R, Welch A, Day
N. Association of hemoglobin A1c with cardiovascular
disease and mortality in adults: the European prospective
investigation into cancer in Norfolk. Ann Intern Med.
2004;141:413–420.
10. Selvin E, Marinopoulos S, Berkenblit G, Rami T, Brancati
FL, Powe NR, Golden SH. Meta‐analysis: glycosylated
hemoglobin and cardiovascular disease in diabetes mellitus.
Ann Intern Med. 2004;141:421–431.
11. Lawes CM, Parag V, Bennett DA, Suh I, Lam TH, Whitlock G,
Barzi F, Woodward M, Asia Pacific Cohort Studies
Collaboration. Blood glucose and risk of cardiovascular disease
in the Asia Pacific region. Diabetes Care. 2004;27:2836–2842.
12. Tancredi M, Rosengren A, Svensson AM, Kosiborod M,
Pivodic A, Gudbjornsdottir S, Wedel H, Clements M,
Dahlqvist S, Lind M. Excess mortality among persons with
type 2 diabetes. N Engl J Med. 2015;373:1720–1732.
13. Chalmers J, Cooper ME. UKPDS and the legacy effect. N
Engl J Med. 2008;359:1618–1620.
14. Group AC, Patel A, MacMahon S, Chalmers J, Neal B, Billot L,
Woodward M, Marre M, Cooper M, Glasziou P, Grobbee D,
Hamet P, Harrap S, Heller S, Liu L, Mancia G, Mogensen CE,
Pan C, Poulter N, Rodgers A, Williams B, Bompoint S, de
Galan BE, Joshi R, Travert F. Intensive blood glucose control
and vascular outcomes in patients with type 2 diabetes. N
Engl J Med. 2008;358:2560–2572.
15. Duckworth W, Abraira C, Moritz T, Reda D, Emanuele N,
Reaven PD, Zieve FJ, Marks J, Davis SN, Hayward R, Warren
SR, Goldman S, McCarren M, Vitek ME, Henderson WG,
Huang GD, VADT Investigators. Glucose control and vascu-
lar complications in veterans with type 2 diabetes. N Engl
J Med. 2009;360:129–139.
16. Gerstein HC, Miller ME, Byington RP, Goff DC, Jr., Bigger
JT, Buse JB, Cushman WC, Genuth S, Ismail‐Beigi F, Grimm
RH, Jr., Probstfield JL, Simons‐Morton DG, Friedewald WT.
Effects of intensive glucose lowering in type 2 diabetes.
N Engl J Med. 2008;358:2545–2559.
17. Meigs JB, Nathan DM, D’Agostino RB, Sr., Wilson PW.
Fasting and postchallenge glycemia and cardiovascular dis-
ease risk: the Framingham Offspring Study. Diabetes Care.
2002;25:1845–1850.
18. Knowler WC, Hamman RF, Edelstein SL, Barrett‐Connor
E, Ehrmann DA, Walker EA, Fowler SE, Nathan DM, Kahn
SE. Prevention of type 2 diabetes with troglitazone in
the  Diabetes Prevention Program. Diabetes 2005;54:
1150–1156.
19. Gnesin F, Thuesen ACB, Kahler LKA, Madsbad S,
Hemmingsen B. Metformin monotherapy for adults with
type 2 diabetes mellitus. Cochrane Database Syst Rev 2020;6:
CD012906.
20. Siconolfi‐Baez L, Banerji MA, Lebovitz HE. Characterization
and significance of sulfonylurea receptors. Diabetes Care.
1990;13 Suppl 3:2–8.
21. United Kingdom Prospective Diabetes Study Group. United
Kingdom Prospective Diabetes Study 24: a 6‐year, rand-
omized, controlled trial comparing sulfonylurea, insulin,
and metformin therapy in patients with newly diagnosed
type 2 diabetes that could not be controlled with diet ther-
apy. Ann Intern Med. 1998;128:165–175.
22. Kahn SE, Haffner SM, Heise MA, Herman WH, Holman
RR, Jones NP, Kravitz BG, Lachin JM, O’Neill MC, Zinman
B, Viberti G. Glycemic durability of rosiglitazone, met-
formin, or glyburide monotherapy. N Engl J Med. 2006;
355:2427–2443.
23. The University Group Diabetes Program. A study of the
effects of hypoglycemic agents on vascular complications in
patients with adult‐onset diabetes. V. Evaluation of pheni-
formin therapy. Diabetes. 1975;24(Suppl 1):65–184.
24. Monami M, Genovese S, Mannucci E. Cardiovascular safety
of sulfonylureas: a meta‐analysis of randomized clinical tri-
als. Diabetes Obes Metab. 2013;15:938–953.
25. Soccio RE, Chen ER, Lazar MA. Thiazolidinediones and the
promise of insulin sensitization in type 2 diabetes. Cell
Metab. 2014;20:573–591.
26. Staels B, Fruchart JC. Therapeutic roles of peroxisome pro-
liferator‐activated receptor agonists. Diabetes. 2005;54:
2460–2470.
27. Inzucchi SE, Maggs DG, Spollett GR, Page SL, Rife FS,
Walton V, Shulman GI. Efficacy and metabolic effects of
metformin and troglitazone in type II diabetes mellitus. N
Engl J Med. 1998;338:867–872.
28. Vella A, de Groen PC, Dinneen SF. Fatal hepatotoxicity asso-
ciated with troglitazone. Ann Intern Med. 1998;129:1080.
29. Nesto RW, Bell D, Bonow RO, Fonseca V, Grundy SM,
Horton ES, Le Winter M, Porte D, Semenkovich CF, Smith S,
Young LH, Kahn R. Thiazolidinedione use, fluid retention,
and congestive heart failure: a consensus statement from the
American Heart Association and American Diabetes
Association. Diabetes Care. 2004;27:256–263.

30. Khan MA, St Peter JV, Xue JL. A prospective, randomized
comparison of the metabolic effects of pioglitazone or
rosiglitazone in patients with type 2 diabetes who were pre-
viously treated with troglitazone. Diabetes Care. 2002;25:
708–711.
31. Nissen SE, Wolski K. Effect of rosiglitazone on the risk of
myocardial infarction and death from cardiovascular causes.
N Engl J Med. 2007;356:2457–2471.
32. Home PD, Pocock SJ, Beck‐Nielsen H, Curtis PS, Gomis R,
Hanefeld M, Jones NP, Komajda M, McMurray JJ, Team RS.
Rosiglitazone evaluated for cardiovascular outcomes in oral
agent combination therapy for type 2 diabetes (RECORD): a
multicentre, randomised, open‐label trial. Lancet. 2009;373:
2125–2135.
33. Bach RG, Brooks MM, Lombardero M, Genuth S, Donner
TW, Garber A, Kennedy L, Monrad ES, Pop‐Busui R, Kelsey
SF, Frye RL, Investigators BD. Rosiglitazone and outcomes for
patients with diabetes mellitus and coronary artery disease in
the Bypass Angioplasty Revascularization Investigation
2  Diabetes (BARI 2D) trial. Circulation. 2013;128:
785–794.
34. Mazzone T, Meyer PM, Feinstein SB, Davidson MH, Kondos
GT, D’Agostino RB, Sr., Perez A, Provost JC, Haffner SM.
Effect of pioglitazone compared with glimepiride on carotid
intima‐media thickness in type 2 diabetes: a randomized
trial. JAMA. 2006;296:2572–2581.
35. Nissen SE, Nicholls SJ, Wolski K, Nesto R, Kupfer S, Perez A,
Jure H, De Larochelliere R, Staniloae CS, Mavromatis K, Saw
J, Hu B, Lincoff AM, Tuzcu EM, Investigators P. Comparison
of pioglitazone vs glimepiride on progression of coronary
atherosclerosis in patients with type 2 diabetes: the
PERISCOPE randomized controlled trial. JAMA. 2008;299:
1561–1573.
36. Dormandy JA, Charbonnel B, Eckland DJ, Erdmann E,
Massi‐Benedetti M, Moules IK, Skene AM, Tan MH,
Lefebvre PJ, Murray GD, Standl E, Wilcox RG, Wilhelmsen
L, Betteridge J, Birkeland K, Golay A, Heine RJ, Koranyi L,
Laakso M, Mokan M, Norkus A, Pirags V, Podar T, Scheen
A, Scherbaum W, Schernthaner G, Schmitz O, Skrha J, Smith
U, Taton J. Secondary prevention of macrovascular events in
patients with type 2 diabetes in the PROactive Study
(PROspective pioglitAzone Clinical Trial In macroVascular
Events): a randomised controlled trial. Lancet. 2005;366:
1279–1289.
37. Erdmann E, Dormandy JA, Charbonnel B, Massi‐Benedetti
M, Moules IK, Skene AM, Investigators PR. The effect of
pioglitazone on recurrent myocardial infarction in 2,445
Choosing Medications for Type 2 Diabetes 183
patients with type 2 diabetes and previous myocardial
infarction: results from the PROactive (PROactive 05) Study.
J Am Coll Cardiol. 2007;49:1772–1780.
38. Wilcox R, Bousser MG, Betteridge DJ, Schernthaner G,
Pirags V, Kupfer S, Dormandy J, Investigators PR. Effects of
pioglitazone in patients with type 2 diabetes with or without
previous stroke: results from PROactive (PROspective piogl-
itAzone Clinical Trial In macroVascular Events 04). Stroke.
2007;38:865–873.
39. Kernan WN, Viscoli CM, Furie KL, Young LH, Inzucchi SE,
Gorman M, Guarino PD, Lovejoy AM, Peduzzi PN, Conwit
R, Brass LM, Schwartz GG, Adams HP, Jr., Berger L, Carolei
A, Clark W, Coull B, Ford GA, Kleindorfer D, O’Leary JR,
Parsons MW, Ringleb P, Sen S, Spence JD, Tanne D, Wang D,
Winder TR, for the IRIS Trial Investigators. Pioglitazone
after ischemic stroke or transient ischemic attack. N Engl
J Med. 2016;374:1321–1331.
40. Lewis JD, Ferrara A, Peng T, Hedderson M, Bilker WB,
Quesenberry CP, Jr., Vaughn DJ, Nessel L, Selby J, Strom BL.
Risk of bladder cancer among diabetic patients treated with
pioglitazone: interim report of a longitudinal cohort study.
Diabetes Care. 2011;34:916–922.
41. Erdmann E, Song E, Spanheimer R, van Troostenburg de
Bruyn AR, Perez A. Observational follow‐up of the PROactive
study: a 6‐year update. Diabetes Obes Metab. 2014;16:
63–74.
42. Ferwana M, Firwana B, Hasan R, Al‐Mallah MH, Kim S,
Montori VM, Murad MH. Pioglitazone and risk of bladder
cancer: a meta‐analysis of controlled studies. Diabet Med.
2013;30:1026–1032.
43. Brauer CA, Coca‐Perraillon M, Cutler DM, Rosen AB.
Incidence and mortality of hip fractures in the United States.
JAMA. 2009;302:1573–1579.
44. Seifarth C, Bergmann J, Holst JJ, Ritzel R, Schmiegel W,
Nauck MA. Prolonged and enhanced secretion of glucagon‐
like peptide 1 (7‐36 amide) after oral sucrose due to alpha‐
glucosidase inhibition (acarbose) in Type 2 diabetic patients.
Diabet Med. 1998;15:485–491.
45. DeFronzo RA. Pharmacologic therapy for type 2 diabetes
mellitus. Ann Intern Med. 1999;131:281–303.
46. Chiasson JL, Josse RG, Gomis R, Hanefeld M, Karasik A,
Laakso M, STOP‐NIDDM Trial Research Group. Acarbose
treatment and the risk of cardiovascular disease and hyper-
tension in patients with impaired glucose tolerance: the
STOP‐NIDDM trial. JAMA. 2003;290:486–494.
47. Chiasson JL, Josse RG, Gomis R, Hanefeld M, Karasik A,
Laakso M. Acarbose for the prevention of Type 2 diabetes,
 184 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
hypertension and cardiovascular disease in subjects with
impaired glucose tolerance: facts and interpretations con-
cerning the critical analysis of the STOP‐NIDDM Trial data.
Diabetologia. 2004;47:969–975; discussion 976–967.
48. Holman RR, Coleman RL, Chan JCN, Chiasson JL, Feng H,
Ge J, Gerstein HC, Gray R, Huo Y, Lang Z, McMurray JJ,
Ryden L, Schroder S, Sun Y, Theodorakis MJ, Tendera M,
Tucker L, Tuomilehto J, Wei Y, Yang W, Wang D, Hu D, Pan
C, ACE Study Group. Effects of acarbose on cardiovascular
and diabetes outcomes in patients with coronary heart dis-
ease and impaired glucose tolerance (ACE): a randomised,
double‐blind, placebo‐controlled trial. Lancet Diabetes
Endocrinol. 2017;5:877–886.
49. Hsu PF, Sung SH, Cheng HM, Shin SJ, Lin KD, Chong K, Yen
FS, Yu BH, Huang CT, Hsu CC. Cardiovascular benefits of
acarbose vs sulfonylureas in patients with type 2 diabetes
treated with metformin. Journal of Clinical Endocrinology &
Metabolism. 103:3611–3619.
50. Staels B, Fonseca VA. Bile acids and metabolic regulation:
mechanisms and clinical responses to bile acid sequestra-
tion. Diabetes Care. 2009;32 Suppl 2:S237–245.
51. Zieve FJ, Kalin MF, Schwartz SL, Jones MR, Bailey WL.
Results of the glucose‐lowering effect of WelChol study
(GLOWS): a randomized, double‐blind, placebo‐controlled
pilot study evaluating the effect of colesevelam hydrochlo-
ride on glycemic control in subjects with type 2 diabetes.
Clin Ther. 2007;29:74–83.
52. Fonseca VA, Rosenstock J, Wang AC, Truitt KE, Jones MR.
Colesevelam HCl improves glycemic control and reduces
LDL cholesterol in patients with inadequately controlled
type 2 diabetes on sulfonylurea‐based therapy. Diabetes
Care. 2008;31:1479–1484.
53. Beysen C, Murphy EJ, Deines K, Chan M, Tsang E, Glass A,
Turner SM, Protasio J, Riiff T, Hellerstein MK. Effect of bile acid
sequestrants on glucose metabolism, hepatic de novo lipogen-
esis, and cholesterol and bile acid kinetics in type 2 diabetes: a
randomised controlled study. Diabetologia. 2012;55:432–442.
54. Smushkin G, Sathananthan M, Piccinini F, Dalla Man C,
Law JH, Cobelli C, Zinsmeister AR, Rizza RA, Vella A. The
effect of a bile acid sequestrant on glucose metabolism in
subjects with type 2 diabetes. Diabetes. 2013;62:1094–1101.
55. Schwab P, Louder A, Li Y, Mallick R, Bays H. Cholesterol
treatment patterns and cardiovascular clinical outcomes
associated with colesevelam HCl and ezetimibe. Drugs Aging
2014;31:683–694.
56. Ryan GJ, Jobe LJ, Martin R. Pramlintide in the treatment of type
1 and type 2 diabetes mellitus. Clin Ther. 2005;27:1500–1512.
57. Vella A, Lee JS, Camilleri M, Szarka LA, Burton DD,
Zinsmeister AR, Rizza RA, Klein PD. Effects of pramlintide,
an amylin analogue, on gastric emptying in type 1 and 2 dia-
betes mellitus. Neurogastroenterol Motil. 2002;14:123–131.
58. Samsom M, Szarka LA, Camilleri M, Vella A, Zinsmeister
AR, Rizza RA. Pramlintide, an amylin analog, selectively
delays gastric emptying: potential role of vagal inhibition.
Am J Physiol Gastrointest Liver Physiol. 2000;278:G946–951.
59. Hoogwerf BJ, Doshi KB, Diab D. Pramlintide, the synthetic
analogue of amylin: physiology, pathophysiology, and effects on
glycemic control, body weight, and selected biomarkers of
vascular risk. Vascular Health & Risk Management. 4:355–362.
60. Gaziano JM, Cincotta AH, O’Connor CM, Ezrokhi M, Rutty
D, Ma ZJ, Scranton RE. Randomized clinical trial of quick‐
release bromocriptine among patients with type 2 diabetes
on overall safety and cardiovascular outcomes. Diabetes
Care. 2010;33:1503–1508.
61. Cincotta AH, Meier AH, Cincotta M, Jr. Bromocriptine
improves glycaemic control and serum lipid profile in obese
Type 2 diabetic subjects: a new approach in the treatment of
diabetes. Expert Opin Investig Drugs. 1999;8:1683–1707.
62. DeFronzo RA. Bromocriptine: a sympatholytic, d2‐dopa-
mine agonist for the treatment of type 2 diabetes. Diabetes
Care. 2011;34:789–794.
63. Pijl H, Ohashi S, Matsuda M, Miyazaki Y, Mahankali A,
Kumar V, Pipek R, Iozzo P, Lancaster JL, Cincotta AH,
DeFronzo RA. Bromocriptine: a novel approach to the treat-
ment of type 2 diabetes. Diabetes Care. 2000;23:1154–1161.
64. Kieffer TJ. Gastro‐intestinal hormones GIP and GLP‐1. Ann
Endocrinol (Paris). 2004;65:13–21.
65. Drucker DJ. Evolving concepts and translational relevance
of enteroendocrine cell biology. Journal of Clinical
Endocrinology and Metabolism. 2016;101(3):778–786.
66. Vella A, Bock G, Giesler PD, Burton DB, Serra DB, Saylan
ML, Deacon CF, Foley JE, Rizza RA, Camilleri M. The effect
of dipeptidyl peptidase‐4 inhibition on gastric volume, satia-
tion and enteroendocrine secretion in type 2 diabetes: a double‐
blind, placebo‐controlled crossover study. Clin Endocrinol
(Oxf). 2008;69:737–744.
67. Vella A, Bock G, Giesler PD, Burton DB, Serra DB, Saylan
ML, Dunning BE, Foley JE, Rizza RA, Camilleri M. Effects of
dipeptidyl peptidase‐4  inhibition on gastrointestinal func-
tion, meal appearance, and glucose metabolism in type 2
diabetes. Diabetes. 2007;56:1475–1480.
68. Holst JJ. Implementation of GLP‐1 based therapy of type 2
diabetes mellitus using DPP‐IV inhibitors. Adv Exp Med
Biol. 2003;524:263–279.

69. Kieffer TJ, Habener JF. The glucagon‐like peptides. Endocr
Rev. 1999;20:876–913.
70. Fadini GP, Avogaro A. Cardiovascular effects of DPP‐4  79. Pfeffer MA, Claggett B, Diaz R, Dickstein K, Gerstein HC,
inhibition: beyond GLP‐1. Vascul Pharmacol. 2011;55:
10–16.
71. Brown NJ, Byiers S, Carr D, Maldonado M, Warner BA.
Dipeptidyl peptidase‐IV inhibitor use associated with
increased risk of ACE inhibitor‐associated angioedema.
Hypertension. 2009;54:516–523.
72. Baetta R, Corsini A. Pharmacology of dipeptidyl pepti-
dase‐4  inhibitors: similarities and differences. Drugs.
2011;71:1441–1467.
73. Scirica BM, Bhatt DL, Braunwald E, Steg PG, Davidson J,
Hirshberg B, Ohman P, Frederich R, Wiviott SD, Hoffman
EB, Cavender MA, Udell JA, Desai NR, Mosenzon O,
McGuire DK, Ray KK, Leiter LA, Raz I, for the SAVOR‐
TIMI 53 Steering Committee and Investigators. Saxagliptin
and cardiovascular outcomes in patients with type 2 diabetes
mellitus. N Engl J Med. 2013;369:1317–1326.
74. White WB, Cannon CP, Heller SR, Nissen SE, Bergenstal
RM, Bakris GL, Perez AT, Fleck PR, Mehta CR, Kupfer S,
Wilson C, Cushman WC, Zannad F, for the EXAMINE
Investigators. Alogliptin after acute coronary syndrome in
patients with type 2 diabetes. N Engl J Med. 2013;369:
1327–1335.
75. Green JB, Bethel MA, Armstrong PW, Buse JB, Engel SS,
Garg J, Josse R, Kaufman KD, Koglin J, Korn S, Lachin JM,
McGuire DK, Pencina MJ, Standl E, Stein PP, Suryawanshi S,
Van de Werf F, Peterson ED, Holman RR, for the TECOS
Study Group. Effect of sitagliptin on cardiovascular out-
comes in type 2 diabetes. N Engl J Med. 2015;373:232–242.
76. McMurray JJV, Ponikowski P, Bolli GB, Lukashevich V,
Kozlovski P, Kothny W, Lewsey JD, Krum H, VIVIDD Trial
Committees and Investigators. Effects of vildagliptin on ven-
tricular function in patients with type 2 diabetes mellitus
and heart failure: a randomized placebo‐controlled trial.
JACC Heart Fail. 2018;6:8–17.
77. Scirica BM, Mosenzon O, Bhatt DL, Udell JA, Steg PG,
McGuire DK, Im K, Kanevsky E, Stahre C, Sjostrand M, Raz
I, Braunwald E. Cardiovascular outcomes according to uri-
nary albumin and kidney disease in patients with type 2 dia-
betes at high cardiovascular risk: observations from the
SAVOR‐TIMI 53 trial. JAMA Cardiol. 2018;3:155–163.
78. Marso SP, Daniels GH, Brown‐Frandsen K, Kristensen P, Mann
JF, Nauck MA, Nissen SE, Pocock S, Poulter NR, Ravn LS,
Steinberg WM, Stockner M, Zinman B, Bergenstal RM, Buse
JB, for the LEADER Steering Committee on behalf of the
Choosing Medications for Type 2 Diabetes 185
LEADER Trial Investigators. Liraglutide and cardiovascular
outcomes in type 2 diabetes. N Engl J Med. 2016;375:311–322.
Kober LV, Lawson FC, Ping L, Wei X, Lewis EF, Maggioni
AP, McMurray JJ, Probstfield JL, Riddle MC, Solomon SD,
Tardif JC, for the ELIXA Investigators. Lixisenatide in patients
with type 2 diabetes and acute coronary syndrome. N Engl J
Med. 2015;373:2247–2257.
80. Holman RR, Bethel MA, Mentz RJ, Thompson VP,
Lokhnygina Y, Buse JB, Chan JC, Choi J, Gustavson SM,
Iqbal N, Maggioni AP, Marso SP, Ohman P, Pagidipati NJ,
Poulter N, Ramachandran A, Zinman B, Hernandez AF, for
the EXSCEL Study Group. Effects of once‐weekly exenatide
on cardiovascular outcomes in type 2 diabetes. N Engl J Med.
2017;377:1228–1239.
81. Marso SP, Bain SC, Consoli A, Eliaschewitz FG, Jodar E,
Leiter LA, Lingvay I, Rosenstock J, Seufert J, Warren ML,
Woo V, Hansen O, Holst AG, Pettersson J, Vilsboll T, for the
SUSTAIN‐6  Investigators. Semaglutide and cardiovascular
outcomes in patients with type 2 diabetes. N Engl J Med.
2016;375:1834–1844.
82. Hernandez AF, Green JB, Janmohamed S, D’Agostino RB,
Sr., Granger CB, Jones NP, Leiter LA, Rosenberg AE, Sigmon
KN, Somerville MC, Thorpe KM, McMurray JJV, Del Prato
S, Harmony Outcomes committees and investigators.
Albiglutide and cardiovascular outcomes in patients with
type 2 diabetes and cardiovascular disease (Harmony
Outcomes): a double‐blind, randomised placebo‐controlled
trial. Lancet. 2018;392:1519–1529.
83. Gerstein HC, Colhoun HM, Dagenais GR, Diaz R,
Lakshmanan M, Pais P, Probstfield J, Riesmeyer JS, Riddle
MC, Ryden L, Xavier D, Atisso CM, Dyal L, Hall S, Rao‐
Melacini P, Wong G, Avezum A, Basile J, Chung N, Conget I,
Cushman WC, Franek E, Hancu N, Hanefeld M, Holt S,
Jansky P, Keltai M, Lanas F, Leiter LA, Lopez‐Jaramillo P,
Cardona Munoz EG, Pirags V, Pogosova N, Raubenheimer
PJ, Shaw JE, Sheu WH, Temelkova‐Kurktschiev T, for the
REWIND Investigators. Dulaglutide and cardiovascular
outcomes in type 2 diabetes (REWIND): a double‐blind,
randomised placebo‐controlled trial. Lancet.
2019;394:121–130.
84. Perkins BA, Udell JA, Cherney DZ. No need to sugarcoat the
message: is cardiovascular risk reduction from SGLT2 inhi-
bition related to natriuresis? Am J Kidney Dis. 2016;68:
349–352.
85. Heerspink HJ, Perkins BA, Fitchett DH, Husain M, Cherney
DZ. Sodium glucose cotransporter 2 inhibitors in the treat-
 186 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
ment of diabetes mellitus: cardiovascular and kidney effects,
potential mechanisms, and clinical applications. Circulation.
2016;134:752–772.
86. Neal B, Perkovic V, Matthews DR. Canagliflozin and cardio-
vascular and renal events in type 2 diabetes. N Engl J Med.
2017;377:2099.
87. Fralick M, Schneeweiss S, Patorno E. Risk of diabetic ketoac-
idosis after initiation of an SGLT2 inhibitor. N Engl J Med.
2017;376:2300–2302.
88. Zinman B, Lachin JM, Inzucchi SE. Empagliflozin, cardio-
vascular outcomes, and mortality in type 2 diabetes. N Engl J
Med. 2016;374:1094.
89. Mahaffey KW, Neal B, Perkovic V, de Zeeuw D, Fulcher G,
Erondu N, Shaw W, Fabbrini E, Sun T, Li Q, Desai M,
Matthews DR, CANVAS Program Collaborative Group.
Canagliflozin for primary and secondary prevention of car-
diovascular events: results from the CANVAS Program
(Canagliflozin Cardiovascular Assessment Study). Circulation.
2018;137:323–334.
90. Mudaliar S, Alloju S, Henry RR. Can a shift in fuel energetics
explain the beneficial cardiorenal outcomes in the EMPA‐
REG OUTCOME Study? A unifying hypothesis. Diabetes
Care. 2016;39:1115–1122.
91. Ferrannini E, Baldi S, Frascerra S, Astiarraga B, Heise T,
Bizzotto R, Mari A, Pieber TR, Muscelli E. Shift to fatty sub-
strate utilization in response to sodium‐glucose cotrans-
porter 2 inhibition in subjects without diabetes and patients
with type 2 diabetes. Diabetes. 2016;65:1190–1195.
92. Ferrannini E, Muscelli E, Frascerra S, Baldi S, Mari A, Heise
T, Broedl UC, Woerle H‐J. Metabolic response to sodium‐
glucose cotransporter 2 inhibition in type 2 diabetic patients.
Journal of Clinical Investigation. 2014;124:499–508.
93. Verma S, Mazer CD, Yan AT, Mason T, Garg V, Teoh H, Zuo
F, Quan A, Farkouh ME, Fitchett DH, Goodman SG,
Goldenberg RM, Al‐Omran M, Gilbert RE, Bhatt DL, Leiter
LA, Juni P, Zinman B, Connelly KA. Effect of empagliflozin
on left ventricular mass in patients with type 2 diabetes mel-
litus and coronary artery disease: the EMPA‐HEART
CardioLink‐6 randomized clinical trial. Circulation.
2019;140:1693–1702.
94. Mahaffey KW, Jardine MJ, Bompoint S, Cannon CP, Neal B,
Heerspink HJL, Charytan DM, Edwards R, Agarwal R,
Bakris G, Bull S, Capuano G, de Zeeuw D, Greene T, Levin
A, Pollock C, Sun T, Wheeler DC, Yavin Y, Zhang H, Zinman
B, Rosenthal N, Brenner BM, Perkovic V. Canagliflozin and
cardiovascular and renal outcomes in type 2 diabetes melli-
tus and chronic kidney disease in primary and secondary
cardiovascular prevention groups. Circulation. 2019;140:
739–750.
95. Wiviott SD, Raz I, Bonaca MP, Mosenzon O, Kato ET, Cahn
A, Silverman MG, Zelniker TA, Kuder JF, Murphy SA, Bhatt
DL, Leiter LA, McGuire DK, Wilding JPH, Ruff CT, Gause‐
Nilsson IAM, Fredriksson M, Johansson PA, Langkilde AM,
Sabatine MS, for the DECLARE–TIMI 58  Investigators.
Dapagliflozin and cardiovascular outcomes in type 2 diabe-
tes. N Engl J Med. 2019;380:347–357.
96. Scheen AJ. Cardiovascular effects of new oral glucose‐lower-
ing agents: DPP‐4 and SGLT‐2  inhibitors. Circ Res.
2018;122:1439–1459.
97. McMurray JJV, Solomon SD, Inzucchi SE, Kober L,
Kosiborod MN, Martinez FA, Ponikowski P, Sabatine MS,
Anand IS, Belohlavek J, Bohm M, Chiang CE, Chopra VK,
de Boer RA, Desai AS, Diez M, Drozdz J, Dukat A, Ge J,
Howlett JG, Katova T, Kitakaze M, Ljungman CEA, Merkely
B, Nicolau JC, O’Meara E, Petrie MC, Vinh PN, Schou M,
Tereshchenko S, Verma S, Held C, DeMets DL, Docherty KF,
Jhund PS, Bengtsson O, Sjostrand M, Langkilde AM, for the
DAPA‐HF Trial Committees and Investigators. Dapagliflozin
in patients with heart failure and reduced ejection fraction.
N Engl J Med. 2019;381:1995–2008.
98. Arioglu E, Duncan‐Morin J, Sebring N, Rother KI, Gottlieb
N, Lieberman J, Herion D, Kleiner DE, Reynolds J,
Premkumar A, Sumner AE, Hoofnagle J, Reitman ML,
Taylor SI. Efficacy and safety of troglitazone in the treatment
of lipodystrophy syndromes. Ann Intern Med. 2000;133:
263–274.
99. Hegele RA, Joy TR, Al‐Attar SA, Rutt BK. Thematic review
series: adipocyte biology. Lipodystrophies: windows on adi-
pose biology and metabolism. J Lipid Res. 2007;48:
1433–1444.
 15 Choosing Medications for Weight Loss
in Type 2 Diabetes Mellitus
Shubhada Jagasia1, Chase Dean Hendrickson2 and Alexander J. Williams3
1
 rofessor of Medicine, Division of Diabetes, Endocrinology and Metabolism, Vanderbilt University Medical Center,
P
Nashville, TN, USA
2
 ssistant Professor of Medicine, Division of Diabetes, Endocrinology, and Metabolism, Vanderbilt University Medical
A
Center, Nashville, TN, USA
3
 ellow, Division of Diabetes, Endocrinology, and Metabolism, Vanderbilt University Medical Center, Nashville,
F
TN, USA

classified as obesity class 1, BMI 35.0–39.9 is classified as

LE A R NI NG P OI NT S
●● Intensive lifestyle interventions in people with type 2
diabetes result in weight loss and are associated with
significant improvements in glycemic control and
cardiovascular disease risk factors.
●● Therapy for type 2 diabetes and improved glycemic
control can result in weight gain.
●● Newer therapies for type 2 diabetes can decrease
caloric intake or increase caloric wastage causing
weight loss or at minimum prevent weight gain while
improving glycemic control
Introduction: Prevalence and relationship
of obesity and type 2 diabetes
Type 2 diabetes develops with increasing frequency at
higher weight and body mass index (BMI). Even when
compared to other known risk factors such as lack of exer­
cise, smoking and poor diet, elevated BMI as a marker for
excess body fat remains the most important risk factor for
developing type 2 diabetes [1].
The National Heart, Lung, and Blood Institute bases
obesity definitions on BMI: BMI < 18.5 is classified as
underweight, BMI 18.5–24.9 is classified as normal weight,
BMI 25.0–29.9 is classified as overweight, BMI 30.0–34.9 is
obesity class 2, and BMI of 40.0 or more is classified as obe­
sity class 3  [2]. In the National Health and Nutrition
Examination Survey (NHANES), obesity is defined as a
BMI of 30 or more, and severe obesity is defined as a BMI
of 40 or more. In adults 20 years old or greater, the preva­
lence of obesity increased from 33.7% in 2007–2008 to
39.6% in 2015–2016, while the prevalence of severe obesity
also increased, from 5.7% in 2007–2008 to 7.7% in
2015–2016 [3]. Abdominal obesity has been described as a
stronger independent predictor for type 2 diabetes [4]. The
prevalence of type 2 diabetes in people with abdominal
obesity increased from 16.0% in 1999–2000 to 20.1% in
2013–2014, compared to 6.2% to 6.3% in non‐obese adults
45–64 years old [5].
The 20‐year Nurses’ Health Study followed 84 941
female nurses from 1980–1996 and found that the risk of
developing diabetes was 38.8 times greater if BMI was 35.0
or more and 20.1 times greater if BMI was 30.0–34.9, com­
pared to a BMI less than 23.0  [1]. The data from the
NHANES study between 1999–2006 also showed the prev­
alence of diabetes increased with BMI: 8% with normal
BMI, 15% in overweight, 23% in obesity class 1, 33% in
obesity class 2, and 43% in obesity class 3 [6]. These studies
show that the prevalence of both obesity and type 2 diabe­
tes are continuing to increase and that the risk of deve­
loping diabetes is strongly tied to excess body weight.

Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.

© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.

187
 188 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
The  Look AHEAD (Action for Health in Diabetes) trial
demonstrated that after one year, intensive lifestyle inter­
ventions in people with type 2 diabetes resulted in weight
loss (8.6% of their initial weight) and was associated with
significant improvements in glycemic control and risk fac­
tors for cardiovascular disease (CVD). However, the differ­
ences in cardiovascular morbidity and mortality between
the intervention and control group were not significant at
10 years. The patients in the intervention group that lost
weight were prescribed fewer glucose‐lowering medica­
tions and had a lower glycated hemoglobin (A1c). In the
intervention group about one‐third of patients were unable
to attain the goal of 5% weight loss after 1 year, and of all
the patients that lost about 5% of their body weight about
half regained some or all of their weight by year 8 [7].
Given this relationship between diabetes and obesity, it
is important to consider the effects of glucose‐lowering
therapy on weight when choosing medications to treat
diabetes.
Therapeutic classes for the treatment
of type 2 Diabetes and their effects
on weight
Understanding the intimate relationship between weight
gain and type 2 diabetes led to an appropriate focus on
weight loss to help prevent and treat type 2 diabetes. The
American Diabetes Association (ADA) recommends that
patients with prediabetes engage in lifestyle changes to
achieve and maintain 7–10% loss of initial body weight to
reduce the risk of conversion to type 2 diabetes  [8]. The
first intervention for most overweight/obese patients with
diabetes is to improve their diet, increase physical activity,
and ultimately reduce caloric intake. However, diet‐
induced weight loss has been shown to increase orexigenic
hormones such as ghrelin and gastric inhibitory polypep­
tide (GIP) and decrease anorexigenic hormones such as
leptin, amylin, cholecystokinin (CCK), and glucagon‐like
peptide 1 (GLP‐1) [9]. Elucidating this hormonal dysregu­
lation, coupled with the fact that some of the earliest glu­
cose‐lowering medications, including insulin, cause weight
gain, led to changes in the suggested pharmaco‐therapeutic
algorithms for type 2 diabetes. In the UK  Pros­pec­tive
Diabetes Study (UKPDS), sulfonylureas, biguanides,
thiazolidinediones (TZDs), and insulin were all shown to
reduce the risk of microvascular and macrovascular com­
plications but also gave insight into the effects of different
medications classes on body weight in people with diabe­
tes [10]. In the ACCORD trial, the intensive arm gained an
average of 3.5 kg compared to 0.4 kg in the conventional
group, thought to be mostly due to insulin and TZDs [11].
Table 15.1 summarizes some of the pertinent literature on
the effects of pharmacotherapy for type 2 diabetes on
weight and can be used as a reference to accompany the
discussion below.
Insulin
The weight gain associated with insulin use has been
explained by various mechanisms, primarily by the increase
in glucose utilization in peripheral tissues and glycemic
levels that fall below the renal threshold for excretion, lead­
ing to increased conservation of ingested calories  [12].
Insulin has also been shown to inhibit protein catabolism
and promote lipogenesis [13]. Another suggested explana­
tion is that insulin resistance in the central nervous system
causes dysregulation of catabolic and anorexic neuronal
responses  [14]. Patients’ fears of hypoglycemia may also
contribute to weight gain via increased intake of
calories [15].
In the UKPDS, patients treated with insulin monother­
apy gained an average of 4.0 kg during the 10‐year study
period, with most of the weight gain occurring in the first
12 months [16]. Henry et al. found that patients with type
2 diabetes treated with an intensive insulin regimen to
obtain tight glycemic control over a 6‐month study period,
the average weight gained was approximately 9  kg  [17].
Yki‐Jarvinen showed that weight gain was more substantial
in obese people with type 2 diabetes treated with insulin,
with an average 5.1  kg increase in the 12‐month study
period [18].
Sulfonylureas
Sulfonylureas work by stimulating pancreatic beta‐cells to
secrete insulin, and, therefore, the mechanisms by which
they cause weight gain are thought to be similar to exoge­
nous insulin administration [19]. In the UKPDS, patients
given chlorpropamide or glibenclamide had an average of
2.15 kg (1.7 for glibenclamide and 2.6 for chlorpropamide)
TABLE 15.1 

Study (reference Average net Duration of Average A1C Primary side

No.) weight change (kg) treatment change (%) effects Comparator

Insulin
  UKPDS (16) +4.0 10 years ‐0.9 Hypoglycemia, compared to placebo/ lifestyle
  Henry RR (17) +9.0 6 months NR weight gain compared to placebo/ lifestyle
  Yki (18) +5.1 1 year ‐2.1 compared to placebo/ lifestyle
Sulfonylureas
  UKPDS (16) +2.2 10 years ‐0.9 Hypoglycemia, compared to placebo
  Advance (20) +0.7 5 years NR weight gain compared to placebo
  Bautista (21) +4.4 3 months ‐1.1 compared to placebo
  Simonson (22) ~0 16 weeks ‐1.5 compared to placebo
Metformin
  DPPOS (26) ‐1.5 10 years ‐0.2 Diarrhea, compared to placebo
  DeFronzo (25) ‐3.8 29 weeks ‐1.5 nausea compared to glyburide and
placebo
  Goldstein (27) ‐2.1 18 weeks ‐0.1 compared to glipizide
Thiazolidinediones
  Miyazaki (29) +3 16 weeks ‐1.4 Edema, compared to placebo
  Basu (30) +2.6 12 weeks +0.4 weight gain compared to glipizide
Alpha‐glucosidase
inhibitors
  Feinbock (34) ‐1.5 20 weeks ‐1.8 Flatulence compared to glimepiride
  Lindstrom (35) ~0 24 weeks ‐0.9 compared to placebo
  Wolever (36) ‐1.0 1 year ‐1.4 compared to placebo
  Phillips (33) ‐0.9 24 weeks ‐0.2 compared to placebo
Amylin
  Riddle (39) ‐2.3 16 weeks ‐0.7 Nausea compared to placebo
  Thompson (38) ‐0.9 4 weeks ‐0.6 compared to placebo
  Hollander (40) ‐2.1 52 weeks ‐0.7 compared to placebo
  Hollander (41) ‐1.8 26 weeks ‐0.6 compared to placebo
GLP‐1 receptor agonists
  Buse (43) ‐1.6 30 weeks ‐0.8 Nausea, compared to placebo
  DeFronzo (44) ‐2.8 30 weeks ‐1.0 vomiting, compared to placebo
  Kendall (45) ‐0.5 30 weeks ‐0.8 diarrhea compared to placebo
  Heine (46) ‐4.1 26 weeks ‐1.1 compared to insulin glargine
  Zinman (47) ‐2.0 26 weeks ‐1.6 compared to placebo
  Garber (48) ‐3.3 52 weeks ‐0.8 compared to glimepiride
  Kim (49) ‐3.8 15 weeks ‐1.4 compared to placebo
  Davies (50) ‐6.4 26 weeks ‐1.9 compared to placebo
  Henry (51) ‐1.2 26 weeks ‐2.8 compared to placebo
DPP‐4 inhibitors
  Ristic (53) ‐0.1 4 weeks ‐0.5 Nausea compared to placebo
  Goldstein (55) +1.0 24 weeks ‐2.9 compared to placebo
  Bolli (54) ~0 24 weeks ‐0.9 compared to placebo
  Ahren (56) ‐2.0 104 weeks ‐0.4 compared to glimepiride
SGLT‐2 inhibitors
  Bolinder (58) ‐4.5 102 weeks ‐0.3 Urinary tract compared to placebo
  Sakai (61) ‐2.6 52 weeks ‐0.7 infections, compared to placebo
  Leiter (60) ‐4.4 104 weeks ‐0.6 genital compared to glimepiride
  Lavelle‐Gonzalez (59) ‐2.5 52 weeks ‐0.8 infections compared to sitagliptin
  Kawata (62) ‐2.6 24 weeks ‐0.8 compared to placebo
  Zinman (63) ‐3.0 3.1 years ‐0.5 compared to placebo
  Cherney (64) ‐2.1 24 weeks ‐0.9 compared to placebo
 190 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
weight gain over the 10‐year study period  [16]. The
ADVANCE trial saw a smaller weight gain of 0.71 kg (gli­
clazide) over a 5‐year period  [20]. Bautista et  al. demon­
strated that patients with type 2 diabetes taking glimepiride
gained 2.3 kg compared to a 2.1 kg decrease in the placebo
group after 3  months  [21]. Conversely, Simonsen et  al.
reported that extended‐release glipizide was not found to
have any significant effect on weight over a 16‐week period,
potentially related to altered pharmacokinetics that pro­
duce smaller fluctuations in insulin levels [22].
Biguanides/Metformin
Metformin has been in use since the 1950s although its
mechanism of action is uncertain. It has been shown to
decrease hepatic glucose production and increase insulin
sensitivity. In the UKPDS and in subsequent studies, met­
formin has been shown to have weight loss effects in people
with type 2 diabetes [23]. DeFronzo et al. compared met­
formin to glyburide and placebo and found that patients
with type 2 diabetes taking metformin lost 3.8  kg com­
pared to those taking glyburide, who experienced no sig­
nificant change in weight. There was a 0.4 kg weight gain
observed in the metformin‐glyburide group over
29  weeks  [24]. In the Diabetes Prevention Program
Outcomes Study (DPPOS), patients taking metformin lost
an average of 2.5 kg compared to a loss of 1 kg in the pla­
cebo group over 10 years, with the placebo group subse­
quently regaining the weight while the metformin group
maintained their weight loss [25]. Goldstein et al. demon­
strated in a multicenter, double‐masked, parallel‐group,
active controlled study in which patients with type 2 diabe­
tes were randomized to receive glipizide, metformin, or
glipizide‐metformin that patients in the metformin group
lost an average of 2.7 kg compared to a loss of 0.4 kg in the
glipizide and 0.3 kg in the metformin‐glipizide groups [26].
Thiazolidinediones
Thiazolidinediones (TZDs) act as insulin sensitizers by
increasing the expression of perioxisome‐proliferator‐acti­
vated receptor gamma (PPARγ). This is a nuclear receptor
which binds, and is activated by, numerous naturally
occurring substances such as arachidonic acid metabolites
and polyunsaturated fatty acids. The improvement in insu­
lin sensitivity is also accompanied by weight gain, thought
to be at least partially due to plasma volume expansion and
peripheral edema [27]. Miyazaki et al. reported that after
16 weeks of pioglitazone patients gained 3.0 kg due to an
increase in fat mass compared to placebo in patients
already taking a sulfonylurea  [28]. These findings were
similar to those reported by Basu et al., where pioglitazone
increased body weight by 3.1 kg, 2.4 kg of which was total
body water compared to 0.5 kg in the glipizide group over
12  weeks  [29]. Despite the effects of weight gain, TZDs
generally have neutral effects on or even decrease the
amount of visceral fat, which may have a positive effect on
cardiovascular disease [30].
Alpha‐glucosidase Inhibitors
Alpha‐glucosidase inhibitors slow the rise in postprandial
glucose and insulin levels by delaying glucose absorp­
tion [31]. Phillips et al. demonstrated that patients taking
acarbose lost 1.32 kg compared to 0.43 kg in the placebo
group in patients already taking metformin over a 24‐week
study period [32]. Feinbock et al. showed that patients with
type 2 diabetes taking acarbose lost an average of 1.9  kg
compared to 0.4  kg in the glimepiride group over
20  weeks  [33]. Conversely, Lindstrom et  al. showed that
patients taking acarbose had no significant weight change
compared to placebo after 24 weeks [34]. Finally, Wolever
et al. demonstrated that after 1 year of treatment patients
with diabetes (not on insulin) lost 0.46 kg compared to a
0.33 kg weight gain in the placebo group [35].
Amylin Mimetics
Amylin is a neuroendocrine hormone secreted by pancre­
atic beta cells along with insulin and suppresses postpran­
dial glucagon secretion and slows gastric emptying  [36].
Pramlintide is an amylin analog and reduces postprandial
glucose levels. Thompson et al. showed a modest reduction
in weight of pramlintide (0.89 kg decrease) after 4 weeks in
people with type 2 diabetes [37]. While Riddle et al. showed
that people with type 2 diabetes taking pramlintide lost an
average of 1.6  kg after 16  weeks compared to placebo,
which had a 0.7  kg weight increase  [38]. Hollander et  al.
found a 1.4 kg weight loss in people with type 2 diabetes
started on pramlintide compared to a 0.7 kg weight increase
after 52 weeks [39]. In a post hoc analysis of overweight/
obese patients with type 2 diabetes, Hollander et  al.
 Choosing Medications for Weight Loss in Type 2 Diabetes Mellitus
reported that pramlintide given for 26‐weeks caused 9% of
patients to lose more than 5% of their body weight com­
pared to 3% of patients in the placebo group. This trans­
lated to an average of 1.8 kg weight loss and was greatest in
patients with a BMI over 40 (3.2 kg) and in patients also
taking metformin (2.5 kg) [40].
GLP‐1 Receptor agonists
Glucagon‐like peptide (GLP)‐1 is an incretin hormone
released from the intestinal L‐cell in response to nutrients
and inhibits glucagon secretion and stimulates insulin exo­
cytosis in a glucose‐dependent manner. It also reduces the
gastric‐emptying rate and stimulates satiety via actions at
the hypothalamus [41]. Buse et al. reported that exenatide,
the first studied GLP‐1 receptor agonist, decreased weight
by 1.6  kg after 30  weeks compared to placebo in people
with type 2 diabetes already on a sulfonylurea  [42].
DeFronzo et  al. found that the addition of exenatide to
metformin in people with type 2 diabetes caused a dose‐
dependent decrease in weight: 2.8 kg in the group receiving
10 mcg twice a day, compared to 1.6 kg in the group receiv­
ing 5 mcg twice a day over a 30 week period [43]. Kendall
et al. showed that exenatide use in people with type 2 dia­
betes on both metformin and sulfonylureas resulted in a
1.6 kg weight loss in both dosage strengths compared to a
0.9 kg weight loss in the placebo group after 30 weeks [44].
Heine et al. compared exenatide to insulin glargine in peo­
ple with type 2 diabetes, and found that in the exenatide
arm there was a 2.3  kg weight decrease compared to a
1.8 kg weight increase in the arm treated with insulin over
26 weeks of treatment [45]. Zinman et al. studied liraglu­
tide in a 26‐week double‐blind, placebo‐controlled trial in
people with type 2 diabetes already on metformin and
rosiglitazone, and reported that the patients receiving lira­
glutide lost 1.0 and 2.0 kg in the 1.2 mg and 1.8 mg daily
regimens, respectively [46]. When Garber et al. compared
liraglutide to glimepiride in people with type 2 diabetes not
on other medications, the liraglutide group lost 1.9  kg
(1.2 mg daily) and 2.1 kg (1.8 mg daily), whereas the glime­
piride group gained 1.2 kg [47]. Kim et al. showed that the
long‐acting formulation of exenatide (2 mg once a week)
caused a 3.8 kg weight loss in people with type 2 diabetes
uncontrolled on metformin compared to placebo at
15 weeks [48]. Since all prior GLP‐1 receptor agonists had
191
been administered subcutaneously, Davies et al. investigated
oral semaglutide compared to subcutaneous delivery and to
placebo. Oral semaglutide had up to a 6.9 kg weight reduction
at 40‐mg a day compared to a 6.4 kg weight reduction in the
subcutaneous form at 1.0 mg once a week [49]. Henry et al.
showed that an osmotic mini‐pump that continuously deliv­
ers exenatide subcutaneously reduced weight in people with
uncontrolled type 2 diabetes by 1.2 kg after 26 weeks [50].
DPP‐4 inhibitors
GLP‐1 and glucose‐dependent insulinotropic peptide
(GIP) are incretins that are inactivated by dipeptidyl pepti­
dase IV (DPP‐4). In 2003 DPP‐4  inhibitors were formu­
lated as a therapy to increase endogenous incretin
activity [51]. Ristic et al. found that vildagliptin had mini­
mal effects on body weight compared to placebo in people
with type 2 diabetes, with a loss of 0.07 kg after 4 weeks [52].
Bolli et al. also showed that vildagliptin had no effects on
body weight when added to metformin  [53]. Similarly,
Goldstein et  al. did not show any meaningful effects on
body weight in people with type 2 diabetes after 24 weeks
with sitagliptin; a loss of 0.1  kg compared to a 1.1  kg
decrease in the placebo group. In a separate trial, sitagliptin
again showed no changes in weight [54]. Ahren et al. com­
pared the addition of albiglutide, sitagliptin, glimepiride
and metformin in people with type 2 diabetes already on
metformin for 104‐weeks in the HARMONY‐3 Trial. This
showed that sitagliptin reduced weight by 0.86  kg, com­
pared to a loss of 1.21 kg in the albiglutide group and a gain
of 1.17 kg in the glimepiride group [55].
SGLT‐2 inhibitors
Sodium‐glucose cotransporters (SGLT) are a family of pro­
teins that actively transport glucose across cell membranes.
SGLT‐2 is primarily expressed in the proximal tubule of
the kidney, but is also present at low levels in the liver and
brain. The role of this protein is to reabsorb glucose from
the urine, thus inhibition lowers the renal threshold for
glucose reabsorption from the collecting system resulting
in glycosuria [56]. Bolinder et al. reported that dapagliflo­
zin reduced body weight by 4.54 kg over 102 weeks in peo­
ple with type 2 diabetes uncontrolled on metformin [57].
Lavelle‐Gonzalez compared canagliflozin to sitagliptin in
people with type 2 diabetes and found that canagliflozin
 192 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
resulted in a 3.3 (100 mg) and 3.7 (300 mg) weight reduc­
tion, whereas sitagliptin caused only a 1.2 kg weight reduc­
tion after 52 weeks [58]. Leiter et al. found that canagliflozin
in addition to metformin in people with type 2 diabetes
resulted in a 3.6 kg weight reduction in both the 100 mg
and 300 mg groups compared to a 0.8 kg weight increase in
the glimepiride group at 104  weeks  [59]. Sakai et  al.
reported that people with type 2 diabetes started on luse­
ogliflozin lost an average of 2.56 kg after 52 weeks, with the
largest weight reductions in the group with a BMI of 30 or
above  [60]. Kawata et  al. demonstrated that ipragliflozin
caused a 2.6  kg weight reduction in patients with type 2
diabetes after 24  weeks  [61]. Zinman et  al. showed that
people with type 2 diabetes started on empagliflozin lost
2.0  kg (10  mg) and 3.0  kg (25  mg) compared to placebo
after 3.1 years in the EMPA‐REG outcomes study  [62].
Cherney et al. stated that in patients with reduced glomer­
ular filtration rates (< 30 ml/min), the degree of weight loss
is attenuated (1.4 kg) compared with normal GFR (> 90 ml/
min) (2.1 kg) but was still greater than placebo [63]. The
proposed mechanism by which weight is lost is through
direct osmotic water loss but more importantly loss of
body fat. This is thought to be accomplished by stimulation
of lipolysis and lipid oxidation due to the decreased levels
of plasma glucose and insulin [64].
Use of diabetic medications for weight
loss in patients without diabetes
Obesity is not only associated with a risk for the develop­
ment of type 2 diabetes but is also a risk factor for stroke,
heart disease, sleep apnea, osteoarthritis, and hyperlipi­
demia [65]. With the prevalence of obesity continuing to
increase [3], there has been an increased emphasis placed
on medical therapies to help induce weight loss. This has
led to the use of medications for type 2 diabetes for indi­
viduals without diabetes to help induce weight loss and
decrease the incidence of comorbid conditions. For exam­
ple, a 5% reduction in body weight in people with a BMI of
34 decreased the incidence of type 2 diabetes by 58% [66].
Metformin
In the DPP study, Knowler et al. initially showed that met­
formin not only reduced the incidence of type 2 diabetes in
obese patients but also reduced weight by 2.1 kg, though
the lifestyle‐intervention group lost more weight (5.6  kg
after 4 years)  [66]. Seifarth et  al. found that metformin
therapy in obese patients caused an average of 5.8  kg
weight loss compared to a 0.8 kg weight gain in the pla­
cebo group, which resulted in 47.4% of the 154  met­
formin‐treated patients achieving a loss of 5% of body
weight at the end of 6 months [67]. Wu et al. used met­
formin to prevent weight gain associated with antipsy­
chotic medications and reported a 4.7 kg weight reduction
in the metformin group compared to a 3.1 weight gain in
the placebo group [68].
GLP‐1 Receptor agonists
Astrup et al. compared orlistat to liraglutide in obese patients
without diabetes and found that liraglutide caused up to a
7.2  kg weight reduction compared to 4.1  kg with orlistat
(lipase inhibitor) and 2.8 kg with placebo after 20 weeks. A
subsequent analysis compared the number of subjects who
lost more than 5% of their body weight; 76% of people on
liraglutide lost more than 5% body weight, compared to 44%
of those treated with orlistat and 30% of those receiving pla­
cebo  [69]. During a 2‐year extension the investigators
reported that liraglutide‐treated patients sustained a 7.8 kg
weight loss and the incidence of metabolic syndrome and
prediabetes decreased by 59% and 52%, respectively [70].
Wadden et  al. reported similar results: in obese patients
without diabetes treated with liraglutide a 6.0  kg weight
loss was achieved. This compared favorably to a 0.1  kg
weight loss in the placebo group after 56  weeks of treat­
ment [71]. O’Neil et al. compared semaglutide to liraglu­
tide and found semaglutide caused a 17.4  kg weight
reduction, whereas liraglutide caused an 8.5  kg weight
reduction [72].
SGLT‐2 inhibitors
Bays et al. reported that canagliflozin reduced body weight
by 2.4  kg in obese patients without diabetes after
12 weeks [73]. Ramirez‐Rodriguez et al. reported a 3.0 kg
weight reduction in obese patients with prediabetes after
12 weeks compared to placebo [74]. Hollander et al. inves­
tigated the combination of canagliflozin with phentermine
in obese patients without diabetes and found up to a 7.3 kg
reduction in body weight [75].
 Choosing Medications for Weight Loss in Type 2 Diabetes Mellitus
A Practical Guide to Choosing medication
for type 2 diabetes with weight loss
in mind
It has been shown that a 5% weight reduction by diet and
exercise can decrease the risk of transitioning from pre‐
diabetes to diabetes by ~50% [25]. While the same degree
of weight reduction in people with type 2 diabetes results in
small changes in A1c [76], this degree of weight reduction
reduces mortality by 25%  [77]. Thus, lifestyle changes
remain an important component in the treatment of type 2
diabetes. The development of multiple, novel therapeutic
classes for the treatment of type 2 diabetes over the past two
decades has given clinicians more options to choose from
when deciding which pharmacological approach is optimal
for a given patient. Metformin continues to be considered
the universal first‐line medication for type 2 diabetes due to
its low cost, effectiveness, and potential for weight loss [78].
When considering second‐line therapy or add‐on therapy,
the ADA recommends a patient‐centric approach, taking
into account cardiovascular risk, avoidance of side effects,
and cost [78]. The cost of each medication can vary based on
insurance coverage and payer source.
In 2014 The Institute for Clinical and Economic
Review (ICER) reported that the combination of met­
formin and a GLP‐1 receptor agonist is superior to met­
formin and a sulfonylurea, and that the combination of
metformin, a GLP‐1 receptor agonist, and a sulfonylurea
is superior to the combination of metformin, a GLP‐1
receptor agonist, and NPH insulin. These recommenda­
tions were based on the clinical benefits of weight reduc­
tion, risk of hypoglycemia, and cost to patients. They
did not find adequate evidence to recommend
DPP‐4 inhibitors over sulfonylureas due to lower clini­
cal benefits with higher cost  [79]. In 2019 the ICER
updated their recommendations comparing oral sema­
glutide to other second line therapies for type 2 diabetes
and reported that there was inconclusive evidence to
recommend oral semaglutide or empagliflozin over each
other but there was high certainty that semaglutide has
at least a small health benefit over sitagliptin. The net
price per year was also compared: oral semaglutide
$6520.02, sitagliptin $1505.07, empagliflozin $2088.13,
liraglutide $5341.90, metformin $917.19, and sulfonylu­
193
reas $578.54. However, the authors also concluded that
the overall cost‐effectiveness (cost per major adverse
cardiovascular events avoided) was higher in oral sema­
glutide compared to sitagliptin and liraglutide but not
empagliflozin [80].
While medication costs will be variable over time, the
evidence to date suggests that GLP‐1 receptor agonists and
SGLT‐2  inhibitors provide clinically relevant weight loss
and provide positive overall health benefits compared to
sulfonylureas. DPP‐4 inhibitors have not shown the same
beneficial health effects, have lower weight‐loss potential,
and are expensive compared to sulfonylureas. Alpha‐glu­
coside inhibitors and amylin mimetics provide weight loss
but are not considered second‐line therapy due to clear
benefits in decreasing cardiovascular endpoints. TZDs are
associated with weight gain  [78] therefore limiting their
use as first‐line agents.
References
1. Hu FB, Manson JE, Stampfer MJ, Colditz G, Liu S, Solomon
CG et al. Diet, lifestyle, and the risk of type 2 diabetes mellitus
in women. N Engl J Med. 2001;345(11):790–797.
2. NHLBI Obesity Education Initiative Expert Panel on the
Identification, Evaluation, and Treatment of Obesity in Adults.
Clinical Guidelines on the Identification, Evaluation, and
Treatment of Overweight and Obesity in Adults. Bethesda,
MI: NHLBI, 1998.
3. Hales CM, Fryar CD, Carroll MD, Freedman DS, Ogden CL.
Trends in obesity and severe obesity prevalence in us youth
and adults by sex and age, 2007–2008 to 2015–2016. JAMA.
2018;319(16):1723–1725.
4. Mamtani M, Kulkarni H, Dyer TD, Almasy L, Mahaney MC,
Duggirala R et al. Waist circumference independently associates
with the risk of insulin resistance and type 2 diabetes in Mexican
American families. PLoS One. 2013;8(3):e59153.
5. Caspard H, Jabbour S, Hammar N, Fenici P, Sheehan JJ,
Kosiborod M. Recent trends in the prevalence of type 2 diabe­
tes and the association with abdominal obesity lead to grow­
ing health disparities in the USA. An analysis of the NHANES
surveys from 1999 to 2014. Diabetes Obes Metab. 2018;20(3):
667–671.
6. Nguyen NT, Nguyen XM, Lane J, Wang P. Relationship
between obesity and diabetes in a US adult population: find­
ings from the National Health and Nutrition Examination
Survey, 1999–2006. Obes Surg. 2011;21(3):351–355.
 194 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
7. Wing RR, Bolin P, Brancati FL, Bray GA, Clark JM, Coday M
et  al. Cardiovascular effects of intensive lifestyle interven­
tion in type 2 diabetes. N Engl J Med. 2013;369(2):145–154.
8. American Diabetes Association. Standards of medical care
in diabetes – 2020. Diabetes Care. 2020;43(Suppl 1).
9. Sumithran P, Prendergast LA, Delbridge E, Purcell K,
Shulkes A, Kriketos A et al. Long‐term persistence of hor­
monal adaptations to weight loss. N Engl J Med.
2011;365(17):1597–1604.
10. United Kingdom Prospective Diabetes Study (UKPDS). 13:
Relative efficacy of randomly allocated diet, sulphonylurea,
insulin, or metformin in patients with newly diagnosed non‐
insulin dependent diabetes followed for three years. BMJ.
1995;310(6972):83–88.
11. Fonseca V, McDuffie R, Calles J, Cohen RM, Feeney P,
Feinglos M et al. Determinants of weight gain in the action
to control cardiovascular risk in diabetes trial. Diabetes Care.
2013;36(8):2162–2168.
12. Carlson MG, Campbell PJ. Intensive insulin therapy and
weight gain in IDDM. Diabetes. 1993;42(12):1700–1707.
13. Umpleby AM, Boroujerdi MA, Brown PM, Carson ER,
Sonksen PH. The effect of metabolic control on leucine
metabolism in type 1 (insulin‐dependent) diabetic patients.
Diabetologia. 1986;29(3):131–141.
14. Schwartz MW, Niswender KD. Adiposity signaling and bio­
logical defense against weight gain: absence of protection or
central hormone resistance? J Clin Endocrinol Metab. 2004;
89(12):5889–5897.
15. Weight gain associated with intensive therapy in the diabetes
control and complications trial. The DCCT Research Group.
Diabetes Care. 1988;11(7):567–573.
16. Intensive blood‐glucose control with sulphonylureas or
insulin compared with conventional treatment and risk of
complications in patients with type 2 diabetes (UKPDS 33).
UK Prospective Diabetes Study (UKPDS) Group. Lancet.
1998;352(9131):837–853.
17. Henry RR, Gumbiner B, Ditzler T, Wallace P, Lyon R,
Glauber HS. Intensive conventional insulin therapy for type
II diabetes. Metabolic effects during a 6‐mo outpatient trial.
Diabetes Care. 1993;16(1):21–31.
18. Yki‐Jarvinen H, Ryysy L, Kauppila M, Kujansuu E, Lahti J,
Marjanen T et al. Effect of obesity on the response to insulin
therapy in noninsulin‐dependent diabetes mellitus. J Clin
Endocrinol Metab. 1997;82(12):4037–4043.
19. Ashcroft FM. Mechanisms of the glycaemic effects of sulfo­
nylureas. Horm Metab Res. 1996;28(9):456–463.
20. van Dieren S, Czernichow S, Chalmers J, Kengne AP,
de  Galan BE, Poulter N et  al. Weight changes and their
predictors amongst 11 140 patients with type 2 diabetes in
the ADVANCE trial. Diabetes Obes Metab. 2012;14(5):
464–469.
21. Luis Bautista J, Bugos C, Dirnberger G, Atherton T. Efficacy
and safety profile of glimepiride in Mexican American
Patients with type 2 diabetes mellitus: a randomized, placebo‐
controlled study. Clin Ther. 2003;25(1):194–209.
22. Simonson DC, Kourides IA, Feinglos M, Shamoon H,
Fischette CT. Efficacy, safety, and dose‐response characteris­
tics of glipizide gastrointestinal therapeutic system on glyce­
mic control and insulin secretion in NIDDM. Results of two
multicenter, randomized, placebo‐controlled clinical trials.
The Glipizide Gastrointestinal Therapeutic System Study
Group. Diabetes Care. 1997;20(4):597–606.
23. Effect of intensive blood‐glucose control with metformin on
complications in overweight patients with type 2 diabetes
(UKPDS 34). UK Prospective Diabetes Study (UKPDS)
Group. Lancet. 1998;352(9131):854–865.
24. DeFronzo RA, Goodman AM. Efficacy of metformin in
patients with non‐insulin‐dependent diabetes mellitus. The
Multicenter Metformin Study Group. N Engl J Med. 1995;
333(9):541–549.
25. Knowler WC, Fowler SE, Hamman RF, Christophi CA,
Hoffman HJ, Brenneman AT et al. 10‐year follow‐up of dia­
betes incidence and weight loss in the Diabetes Preven­
tion Program Outcomes Study. Lancet. 2009;374(9702):
1677–1686.
26. Goldstein BJ, Pans M, Rubin CJ. Multicenter, randomized,
double‐masked, parallel‐group assessment of simultaneous
glipizide/metformin as second‐line pharmacologic treat­
ment for patients with type 2 diabetes mellitus that is inade­
quately controlled by a sulfonylurea. Clin Ther. 2003;25(3):
890–903.
27. Yki‐Jarvinen H. Thiazolidinediones. N Engl J Med. 2004;
351(11):1106–1118.
28. Miyazaki Y, Mahankali A, Matsuda M, Mahankali S, Hardies
J, Cusi K et al. Effect of pioglitazone on abdominal fat distri­
bution and insulin sensitivity in type 2 diabetic patients.
J Clin Endocrinol Metab. 2002;87(6):2784–2791.
29. Basu A, Jensen MD, McCann F, Mukhopadhyay D, Joyner
MJ, Rizza RA. Effects of pioglitazone versus glipizide on
body fat distribution, body water content, and hemodynam­
ics in type 2 diabetes. Diabetes Care. 2006;29(3):510–514.
30. Wilding J. Thiazolidinediones, insulin resistance and obe­
sity: finding a balance. Int J Clin Pract. 2006;60(10):
1272–1280.
31. Bischoff H. Pharmacology of alpha‐glucosidase inhibition.
Eur J Clin Invest. 1994;24 Suppl 3:3–10.
 Choosing Medications for Weight Loss in Type 2 Diabetes Mellitus
32. Phillips P, Karrasch J, Scott R, Wilson D, Moses R. Acarbose
improves glycemic control in overweight type 2 diabetic
patients insufficiently treated with metformin. Diabetes
Care. 2003;26(2):269–273.
33. Feinbock C, Luger A, Klingler A, Egger T, Bielesz GK,
Winkler F et al. Prospective multicentre trial comparing the
efficacy of, and compliance with, glimepiride or acarbose
treatment in patients with type 2 diabetes not controlled
with diet alone. Diabetes Nutr Metab. 2003;16(4):214–221.
34. Lindstrom J, Tuomilehto J, Spengler M. Acarbose treatment
does not change the habitual diet of patients with type 2 dia­
betes mellitus. The Finnish Acargbos Study Group. Diabet
Med. 2000;17(1):20–25.
35. Wolever TM, Chiasson JL, Josse RG, Hunt JA, Palmason
C, Rodger NW et al. Small weight loss on long‐term acar­
bose therapy with no change in dietary pattern or nutri­
ent intake of individuals with non‐insulin‐dependent
diabetes. Int J Obes Relat Metab Disord. 1997;21(9):
756–763.
36. Weyer C, Maggs DG, Young AA, Kolterman OG. Amylin
replacement with pramlintide as an adjunct to insulin ther­
apy in type 1 and type 2 diabetes mellitus: a physiological
approach toward improved metabolic control. Curr Pharm
Des. 2001;7(14):1353–1373.
37. Thompson RG, Pearson L, Schoenfeld SL, Kolterman OG.
Pramlintide, a synthetic analog of human amylin, improves
the metabolic profile of patients with type 2 diabetes using
insulin. The Pramlintide in Type 2 Diabetes Group. Diabetes
Care. 1998;21(6):987–993.
38. Riddle M, Frias J, Zhang B, Maier H, Brown C, Lutz K et al.
Pramlintide improved glycemic control and reduced weight
in patients with type 2 diabetes using basal insulin. Diabetes
Care. 2007;30(11):2794–2799.
39. Hollander PA, Levy P, Fineman MS, Maggs DG, Shen LZ,
Strobel SA et al. Pramlintide as an adjunct to insulin therapy
improves long‐term glycemic and weight control in patients
with type 2 diabetes: a 1‐year randomized controlled trial.
Diabetes Care. 2003;26(3):784–790.
40. Hollander P, Maggs DG, Ruggles JA, Fineman M, Shen L,
Kolterman OG et al. Effect of pramlintide on weight in over­
weight and obese insulin‐treated type 2 diabetes patients.
Obes Res. 2004;12(4):661–668.
41. Deacon CF. Therapeutic strategies based on glucagon‐like
peptide 1. Diabetes. 2004;53(9):2181–2189.
42. Buse JB, Henry RR, Han J, Kim DD, Fineman MS, Baron
AD. Effects of exenatide (exendin‐4) on glycemic control
over 30  weeks in sulfonylurea‐treated patients with type 2
diabetes. Diabetes Care. 2004;27(11):2628–2635.
195
43. DeFronzo RA, Ratner RE, Han J, Kim DD, Fineman MS,
Baron AD. Effects of exenatide (exendin‐4) on glycemic
control and weight over 30  weeks in metformin‐treated
patients with type 2 diabetes. Diabetes Care.
2005;28(5):1092–1100.
44. Kendall DM, Riddle MC, Rosenstock J, Zhuang D, Kim DD,
Fineman MS et al. Effects of exenatide (exendin‐4) on glyce­
mic control over 30 weeks in patients with type 2 diabetes
treated with metformin and a sulfonylurea. Diabetes Care.
2005;28(5):1083–1091.
45. Heine RJ, Van Gaal LF, Johns D, Mihm MJ, Widel MH,
Brodows RG. Exenatide versus insulin glargine in patients
with suboptimally controlled type 2 diabetes: a randomized
trial. Ann Intern Med. 2005;143(8):559–569.
46. Zinman B, Gerich J, Buse JB, Lewin A, Schwartz S, Raskin P
et  al. Efficacy and safety of the human glucagon‐like pep­
tide‐1 analog liraglutide in combination with metformin
and thiazolidinedione in patients with type 2 diabetes
(LEAD‐4  Met+TZD). Diabetes Care.
2009;32(7):1224–1230.
47. Garber A, Henry R, Ratner R, Garcia‐Hernandez PA,
Rodriguez‐Pattzi H, Olvera‐Alvarez I et al. Liraglutide ver­
sus glimepiride monotherapy for type 2 diabetes
(LEAD‐3 Mono): a randomised, 52‐week, phase III, double‐
blind, parallel‐treatment trial. Lancet.
2009;373(9662):473–481.
48. Kim D, MacConell L, Zhuang D, Kothare PA, Trautmann M,
Fineman M et  al. Effects of once‐weekly dosing of a long‐
acting release formulation of exenatide on glucose control
and body weight in subjects with type 2 diabetes. Diabetes
Care. 2007;30(6):1487–1493.
49. Davies M, Pieber TR, Hartoft‐Nielsen ML, Hansen OKH,
Jabbour S, Rosenstock J. Effect of oral semaglutide com­
pared with placebo and subcutaneous semaglutide on glyce­
mic control in patients with type 2 diabetes: a randomized
clinical trial. JAMA. 2017;318(15):1460–1470.
50. Henry RR, Rosenstock J, Denham DS, Prabhakar P, Kjems L,
Baron MA. Clinical impact of ITCA 650, a novel drug‐
device GLP‐1 receptor agonist, in uncontrolled type 2 diabe­
tes and very high baseline HbA1c: the FREEDOM‐1 HBL
(High Baseline) Study. Diabetes Care. 2018;41(3):613–619.
51. Drucker DJ. Therapeutic potential of dipeptidyl peptidase
IV inhibitors for the treatment of type 2 diabetes. Expert
Opin Investig Drugs. 2003;12(1):87–100.
52. Ristic S, Byiers S, Foley J, Holmes D. Improved glycaemic
control with dipeptidyl peptidase‐4  inhibition in patients
with type 2 diabetes: vildagliptin (LAF237) dose response.
Diabetes Obes Metab. 2005;7(6):692–698.
 196 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
53. Bolli G, Dotta F, Rochotte E, Cohen SE. Efficacy and tolera­
bility of vildagliptin vs. pioglitazone when added to metformin:
a 24‐week, randomized, double‐blind study. Diabetes Obes
Metab. 2008;10(1):82–90.
54. Goldstein BJ, Feinglos MN, Lunceford JK, Johnson J,
Williams‐Herman DE. Effect of initial combination therapy
with sitagliptin, a dipeptidyl peptidase‐4 inhibitor, and met­
formin on glycemic control in patients with type 2 diabetes.
Diabetes Care. 2007;30(8):1979–1987.
55. Ahren B, Johnson SL, Stewart M, Cirkel DT, Yang F, Perry C
et  al. HARMONY 3: 104‐week randomized, double‐blind,
placebo‐ and active‐controlled trial assessing the efficacy
and safety of albiglutide compared with placebo, sitagliptin,
and glimepiride in patients with type 2 diabetes taking met­
formin. Diabetes Care. 2014;37(8):2141–2148.
56. Abdul‐Ghani MA, Norton L, DeFronzo RA. Role of sodium‐
glucose cotransporter 2 (SGLT 2) inhibitors in the treatment
of type 2 diabetes. Endocr Rev. 2011;32(4):515–531.
57. Bolinder J, Ljunggren O, Johansson L, Wilding J, Langkilde
AM, Sjostrom CD et al. Dapagliflozin maintains glycaemic
control while reducing weight and body fat mass over 2 years
in patients with type 2 diabetes mellitus inadequately con­
trolled on metformin. Diabetes Obes Metab. 2014;16(2):
159–169.
58. Lavalle‐Gonzalez FJ, Januszewicz A, Davidson J, Tong C,
Qiu R, Canovatchel W et al. Efficacy and safety of canagliflo­
zin compared with placebo and sitagliptin in patients with
type 2 diabetes on background metformin monotherapy: a
randomised trial. Diabetologia. 2013;56(12):2582–2592.
59. Leiter LA, Yoon KH, Arias P, Langslet G, Xie J, Balis DA et al.
Canagliflozin provides durable glycemic improvements and
body weight reduction over 104  weeks versus glimepiride in
patients with type 2 diabetes on metformin: a randomized, dou­
ble‐blind, phase 3 study. Diabetes Care. 2015;38(3):355–364.
60. Sakai S, Kaku K, Seino Y, Inagaki N, Haneda M, Sasaki T
et al. Efficacy and safety of the SGLT2 inhibitor luseogliflo­
zin in Japanese patients with type 2 diabetes mellitus strati­
fied according to baseline body mass index: pooled analysis
of data from 52‐week phase iii trials. Clin Ther. 2016;38(4):
843–862.e9.
61. Kawata T, Iizuka T, Iemitsu K, Takihata M, Takai M,
Nakajima S et  al. Ipragliflozin improves glycemic control
and decreases body fat in patients with type 2 diabetes mel­
litus. J Clin Med Res. 2017;9(7):586–595.
62. Zinman B, Wanner C, Lachin JM, Fitchett D, Bluhmki E,
Hantel S et al. Empagliflozin, cardiovascular outcomes, and
mortality in type 2 diabetes. N Engl J Med. 2015;373(22):
2117–2128.
63. Cherney DZI, Cooper ME, Tikkanen I, Pfarr E, Johansen
OE, Woerle HJ et al. Pooled analysis of Phase III trials indi­
cate contrasting influences of renal function on blood pres­
sure, body weight, and HbA1c reductions with empagliflozin.
Kidney Int. 2018;93(1):231–244.
64. Ferrannini E, Baldi S, Frascerra S, Astiarraga B, Heise T,
Bizzotto R et al. Shift to fatty substrate utilization in response
to sodium‐glucose cotransporter 2  inhibition in subjects
without diabetes and patients with type 2 diabetes. Diabetes.
2016;65(5):1190–1195.
65. Li Z, Maglione M, Tu W, Mojica W, Arterburn D, Shugarman
LR et al. Meta‐analysis: pharmacologic treatment of obesity.
Ann Intern Med. 2005;142(7):532–546.
66. Knowler WC, Barrett‐Connor E, Fowler SE, Hamman RF,
Lachin JM, Walker EA et al. Reduction in the incidence of
type 2 diabetes with lifestyle intervention or metformin. N
Engl J Med. 2002;346(6):393–403.
67. Seifarth C, Schehler B, Schneider HJ. Effectiveness of
metformin on weight loss in non‐diabetic individuals
with obesity. Exp Clin Endocrinol Diabetes. 2013;121(1):
27–31.
68. Wu RR, Zhao JP, Jin H, Shao P, Fang MS, Guo XF et  al.
Lifestyle intervention and metformin for treatment of antip­
sychotic‐induced weight gain: a randomized controlled trial.
JAMA. 2008;299(2):185–193.
69. Astrup A, Rossner S, Van Gaal L, Rissanen A, Niskanen L,
Al Hakim M et al. Effects of liraglutide in the treatment of
obesity: a randomised, double‐blind, placebo‐controlled study.
Lancet. 2009;374(9701):1606–1616.
70. Astrup A, Carraro R, Finer N, Harper A, Kunesova M, Lean
ME et al. Safety, tolerability and sustained weight loss over 2
years with the once‐daily human GLP‐1 analog, liraglutide.
Int J Obes (Lond). 2012;36(6):843–854.
71. Wadden TA, Hollander P, Klein S, Niswender K, Woo V,
Hale PM et  al. Weight maintenance and additional weight
loss with liraglutide after low‐calorie‐diet‐induced weight
loss: the SCALE Maintenance randomized study. Int J Obes
(Lond). 2013;37(11):1443–1451.
72. O’Neil PM, Birkenfeld AL, McGowan B, Mosenzon O,
Pedersen SD, Wharton S et al. Efficacy and safety of sema­
glutide compared with liraglutide and placebo for weight
loss in patients with obesity: a randomised, double‐blind,
placebo and active controlled, dose‐ranging, phase 2 trial.
Lancet. 2018;392(10148):637–649.
 Choosing Medications for Weight Loss in Type 2 Diabetes Mellitus
73. Bays HE, Weinstein R, Law G, Canovatchel W. Canagliflozin:
effects in overweight and obese subjects without diabe­
tes mellitus. Obesity (Silver Spring). 2014;22(4):1042–1049.
74. Ramirez‐Rodriguez AM, Gonzalez‐Ortiz M, Martinez‐
Abundis E. Effect of dapagliflozin on insulin secretion and
insulin sensitivity in patients with prediabetes. Exp Clin
Endocrinol Diabetes. 2018.
75. Hollander P, Bays HE, Rosenstock J, Frustaci ME, Fung A,
Vercruysse F et  al. Coadministration of canagliflozin and
phentermine for weight management in overweight and
obese individuals without diabetes: a randomized clinical
trial. Diabetes Care. 2017;40(5):632–639.
76. Pastors JG, Franz MJ, Warshaw H, Daly A, Arnold MS. How
effective is medical nutrition therapy in diabetes care? J Am
Diet Assoc. 2003;103(7):827–831.
197
77. Gregg EW, Gerzoff RB, Thompson TJ, Williamson DF.
Trying to lose weight, losing weight, and 9‐year mortality in
overweight U.S. adults with diabetes. Diabetes Care. 2004;
27(3):657–662.
78. American Diabetes Association. 9. Pharmacologic
Approaches to Glycemic Treatment: Standards of Medical
Care in Diabetes – 2019. 2019.
79. Alzaid AA, Dinneen SF, Turk DJ, Caumo A, Cobelli C, Rizza
RA. Assessment of insulin action and glucose effectiveness
in diabetic and nondiabetic humans. J Clin Invest.
1994;94(6):2341–2348.
80. O’Brien TD, Rizza RA, Carney JA, Butler PC. Islet amyloido­
sis in a patient with chronic massive insulin resistance due to
antiinsulin receptor antibodies. J Clin Endocrinol Metab.
1994;79(1):290–292.
 16 Are Statins the Optimal Therapy
for Cardiovascular Risk in Patients
with Diabetes? What Newer Agents Are
There for the Treatment for Dyslipidemia
in Diabetes? Are Triglycerides
an Important Risk Factor for Diabetes?
Recie Davern and Timothy O’Brien
Department of Medicine and Endocrinology/Diabetes Mellitus, University College Hospital Galway and National
University of Ireland, Galway, Ireland

33 kg/m2 [20–25]. Lipid profile shows a total cholesterol of

LE A R N I N G P OI NT S
●● Treating dyslipidemia is central to the prevention of
cardiovascular disease in diabetes.
●● There is strong epidemiological evidence to date that
hypertriglyceridemia is an independent risk factor for
ischemic heart disease.
●● Combination therapy may be necessary to adequately
address dyslipidemia in diabetes.
Introduction
A 54‐year‐old man with a four‐year history of Type 2
Diabetes Mellitus is reviewed in the outpatient department.
Past medical history is otherwise significant for hyperten­
sion and dyslipidemia. His medications are metformin
500  mg tds, ezetimibe/simvastatin 10/20  mg od, and tel­
misartan 80 mg od. Glycosylated hemoglobin (HbA1C) is
5.4% (4.0–6.0). Blood pressure is well‐controlled at
110/78  mmHg. Body mass index (BMI) is elevated at
4.4 mmol/l (< 4.6 mmol/l), LDL 1.4 mmol/l (< 2.6 mmol/l),
HDL of 0.78 mmol/l (> 1.0 mmol/l) and fasting triglycer­
ides of 4.4 mmol/l (< 1.7 mmol/l). His weekly alcohol con­
sumption is 16–20 units per week.
The above clinical scenario is very commonly encoun­
tered in outpatient diabetes care. Although evidence‐based
targets such as HbA1C, blood pressure and LDL choles­
terol are all within desired limits, uncertainty arises over
whether treatment should be tailored to account for ele­
vated fasting plasma triglycerides and low HDL choles­
terol, which are the two most common lipid derangements
in Type 2 diabetes mellitus. In this chapter we will discuss
the pathophysiology of dyslipidemia in Type 2 diabetes
mellitus, review the arguments for use of statins in both
primary and secondary prevention of cardiovascular dis­
ease (CVD) in diabetes mellitus, discuss the evidence base
for elevated triglycerides as an independent risk factor for
CVD in diabetes mellitus and discuss new emerging thera­
pies for the treatment of dyslipidemia.

Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.

© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.

198
 Are Statins the Optimal Therapy for Cardiovascular Risk in Patients with Diabetes? 199

Dyslipidemia in type 2 diabetes mellitus Insulin resistance also promotes the conversion LDL to
smaller lipoproteins termed small dense LDL  [8]. These
Dyslipidemia is one of the major risk factors for cardiovas­
small dense particles have been shown to have atherogenic
cular disease (CVD) in type 2 diabetes mellitus. Despite
properties  [9]. Small dense LDL particles have reduced
mounting data from clinical trials, the management of dys­
affinity for the LDL B/E receptor and are preferentially
lipidemia other than lowering low‐density lipoprotein (LDL)
taken up by macrophages, through the scavenger receptor,
cholesterol continues to be debated. While the term “dyslipi­
leading to the formation of foam cells. Small dense LDL
demia” describes a vast array of lipoprotein abnormalities,
particles have higher affinity for intimal proteoglycans
diabetic dyslipidemia is a type of secondary dyslipidemia.
than large LDL particles which may favor the penetration
Both quantitative and qualitative lipoprotein abnormali­
of LDL particles into the arterial wall [10]. This promotes
ties are observed in type 2 diabetic patients [1]. The quanti­
atherosclerosis and contributes to their increased cardio­
tative abnormalities are hypertriglyceridemia and low levels
vascular risk.
of cardioprotective high‐density lipoprotein (HDL) choles­
To fully understand the role of the lipid abnormalities
terol levels. The qualitative abnormalities include the pres­
seen in type 2 diabetes mellitus we must review the meta­
ence of large very low‐density lipoproteins (VLDL) with
bolic syndrome. The metabolic syndrome refers to a group
relatively high levels of triglycerides and small dense LDL.
of metabolic abnormalities that are frequently seen in type
Insulin plays a central role in the regulation of lipid
2 diabetes mellitus that increase a person’s risk of CVD,
metabolism. In adipose tissue, insulin inhibits hormone
both individually and collectively [11]. There are different
sensitive lipase, therefore, it has anti‐lipolytic action. It pro­
sets of diagnostic criteria for metabolic syndrome that have
motes storage of triglyceride (TG) in the adipocytes and
been published by the WHO, the European Group for the
reduces secretion in the circulation of free fatty acids (FFA)
Study of Insulin Resistance (EGIR), the National
from adipose tissue.
Cholesterol Education Program (NCEP) Adult Treatment
The hypertriglyceridemia seen in type 2 diabetic
Panel III (ATP III), and the International Diabetes
patients is a product of insulin resistance. Insulin resistance
Foundation (IDF). The NCEP ATP III is the most com­
leads to excessive amounts of FFA, many of which build up
monly used and we have summarized those criteria in
in the liver as TG [2]. The liver has three different handling
Box 16.1.
mechanisms for the excess, it either: stores it, burns it
through β‐oxidation in the mitochondria or exports it by
Treatment of dyslipidemia in type 2
synthesizing VLDL particles, which are TG rich.
diabetes mellitus
The decrease in HDL cholesterol is caused by increased
catabolism due to insulin resistance. This is due to an increased There is a large body of evidence supporting treatment of
pool of TG rich VLDL. This increase leads to the transfer of dyslipidemia in type 2 diabetes mellitus for primary and
TG from TG rich lipoproteins to HDL leading to the forma­ secondary prevention of CVD [12, 13].
tion of TG‐rich HDL particles [3]. Cholesteryl ester transfer
protein (CETP) is responsible for these transfers to HDL and Box 16.1 NCEP guidelines for diagnosis
of the metabolic syndrome
CETP inhibition has been studied as a mechanism to increase
HDL and reduce CVD risk. Unfortunately, many clinical tri­ Presence of three of the following five risk factors:
als  [4, 5, 6] have not shown that CETP inhibition reduces OO Abdominal obesity (waist circumference > 40 inches in
men, > 35 inches in women)
CVD risk. Though a recent trial [7] of anacetrapib showed a
OO Plasma triglycerides ≥ 150 mg/dl (1.7 mmol/l)
modest reduction in CVD risk after a long follow‐up period. OO Plasma HDL cholesterol ≤ 40 mg/dl in men
Hepatic lipase whose activity is increased in insulin resistance (1.02 mmol/l) and ≤ 50 mg/dl in women (1.27 mmol/l)
and type 2 diabetes leads to increased catabolism of these TG OO Blood pressure > 130/85 mmHg
rich HDL particles. OO Fasting plasma glucose > 110 mg/dl (6.1 mmol/l)
 200 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
Statins are the mostly extensively studied treatment for
dyslipidemia. Two studies have examined statin use for
the primary prevention of CVD exclusively in those with
diabetes mellitus. The Collaborative Atorvastatin
Diabetes Study (CARDS) involved 2838 subjects aged
40–75 years from centers around the UK and Ireland who
were randomized to either 10 mg atorvastatin or placebo.
Participants had type 2 diabetes mellitus with no history
of CVD, but had at least one other cardiovascular risk
factor. The study was terminated 2 years early because the
pre‐specified stopping rule for efficacy in reducing car­
diovascular events had been achieved. Subjects allocated
to atorvastatin had a 37% reduction in the primary end­
point of acute coronary heart disease (CHD) events
(myocardial infarction, unstable angina, acute CHD
death, resuscitated cardiac arrest), coronary revasculari­
zation or stroke. All‐cause mortality was numerically
lower in the atorvastatin group but was not statistically
significantly reduced [14].
The Atorvastatin Study for Prevention of CHD
Endpoints in Non‐insulin‐dependent diabetes mellitus
(ASPEN) study included 2410 subjects with type two dia­
betes mellitus who were also assigned to receive 10  mg
atorvastatin or placebo. This international study was origi­
nally intended to look at secondary prevention with a fol­
low‐up of 4 years; however, due to changing guidelines
impairing recruitment, the protocol was changed and a
primary prevention cohort were also recruited. For the pri­
mary prevention cohort a significant 30% reduction in
LDL levels was observed, but no beneficial effect was found
on the primary outcome which was a composite of cardio­
vascular death, non‐fatal myocardial infarction, non‐fatal
stroke, recanalization, coronary artery bypass surgery,
resuscitated cardiac arrest, and hospitalization for unstable
angina. Similar results were found in the small secondary
prevention cohort. There are multiple limitations of this
study including the change in the study design, the multi­
ple endpoints, the low risk of events in the primary preven­
tion cohort and the extensive use of non‐study lipid‐lowering
therapy in both groups [15].
Several studies have investigated statin use for primary
prevention with substantial subgroups of subjects with dia­
betes mellitus. The Heart Protection Study (HPS) recruited
20 536 patients with CHD, other occlusive vascular disease
or diabetes mellitus, and included 5963 subjects with diabe­
tes mellitus. Simvastatin 40 mg or placebo was given over the
course of 5 years. 2912 of the participants with diabetes mel­
litus had no prior vascular disease (49%) (primary preven­
tion) and in this subgroup the rate of the defined endpoint
for subcategories which was major vascular events (cardio­
vascular death, non‐fatal myocardial  infarction, stroke,
revascularization) was reduced signi­ficantly from 13% to 9%
in the group receiving simvastatin [16].
The Anglo‐Scandinavian Cardiac Outcomes Trial
(ASCOT) was primarily aimed at investigating treatment
of hypertension in 19 342 hypertensive patients, but half of
the study group (n = 10 305), including 2532 with type two
diabetes mellitus, were allocated to a lipid sub‐study with
atorvastatin 10 mg or placebo. Overall results for the lipid‐
lowering trial showed a significant reduction in the pri­
mary endpoint of non‐fatal myocardial infarction and fatal
CHD with atorvastatin. No statistically significant reduc­
tion in this endpoint was found when looking at the sub­
group with diabetes mellitus. However, a subsequent
pre‐specified analysis of an expanded composite of total
cardiovascular outcomes (major cardiovascular events plus
procedures) demonstrated a significant reduction for the
subgroup with diabetes mellitus [17].
The Management of Elevated Cholesterol in the Primary
Prevention Group of Adult Japanese study (MEGA)
recruited 7832 subjects with dyslipidemia and assigned
them to diet modification only or diet plus open‐label low‐
dose pravastatin (10–20  mg). 1632 subjects (21%) were
reported by their physicians to have diabetes mellitus at
baseline. Overall, the rate of CHD events (fatal or non‐fatal
myocardial infarction, angina, revascularization, sudden
cardiac death) was significantly reduced in the pravastatin‐
treated group compared with diet only. Subgroup analysis
did not show significant interaction in any subgroup,
including the diagnosis of diabetes mellitus or not. A post‐
hoc analysis looked at the subgroup with diabetes mellitus
and categorized them according to reductions in LDL cho­
lesterol and increases in HDL cholesterol. The greatest
reduction in risk of CVD was found where LDL cholesterol
decreased more than 15% and HDL cholesterol increased
by more than 5% [18].
 Are Statins the Optimal Therapy for Cardiovascular Risk in Patients with Diabetes?
The Antihypertensive and Lipid‐Lowering treatment to
prevent Heart Attack Trial (ALLHAT) [19] looked at treat­
ment of hypertension in subjects aged 55 years or older.
Within this study there was a Lipid‐Lowering Trial compo­
nent (ALLHAT‐LLT) which randomized 10 355 subjects,
(of which 3638 (35%) had type two diabetes mellitus), to
open label pravastatin 40 mg treatment or usual care. No
significant difference was found in all‐cause mortality (pri­
mary outcome) or CHD outcomes (secondary outcome of
fatal CHD, non‐fatal myocardial infarction) in the whole
study group or in the subgroup with diabetes mellitus. It is
felt that one of the reasons for the negative results seen in
this trial may be the relatively high rates of statin use in the
usual care group, reaching nearly 25%, as intensive lipid
lowering became a routine part of clinical care following
the publication of other statin studies.
When it comes to secondary prevention there are also
many landmark studies. The Scandinavian Simvastatin
Survival Study (4S) recruited 4444 patients with a prior his­
tory of CHD and hypercholesterolemia who were rand­
omized to treatment with 20–40 mg simvastatin or placebo.
Over a median duration of 5.4 years of follow‐up, those
randomized to simvastatin had a significant reduction in
all‐cause mortality and major CHD events, and this
included a significant reduction in CHD events in the
small subgroup of 202 with diabetes mellitus [20]. A later
post‐hoc analysis was performed using newer diagnostic
criteria for type two diabetes mellitus and confirmed a sig­
nificant reduction in CHD events in 483 subjects with
baseline diabetes mellitus [21].
The Long‐term Intervention with Pravastatin in
Ischemic Disease (LIPID) trial involved a larger cohort of
9014 patients with CHD, with 1077 (12%) diagnosed as
having diabetes mellitus at baseline. Overall, pravastatin
40 mg significantly reduced the primary outcome of mor­
tality from CHD, and in a pre‐specified subgroup analysis,
pravastatin reduced the occurrence of major CHD events
and major coronary events (CHD death or non‐fatal myo­
cardial infarction) in the subgroup with diabetes melli­
tus  [22]. The effects shown on triglyceride levels in this
study will be discussed later in the chapter.
The Cholesterol and Recurrent Events (CARE) [23] trial
was a secondary prevention study comparing pravastatin
40 mg with placebo in 4159 patients with a history of CHD
201
and average cholesterol levels. The incidence of the pri­
mary endpoint of a fatal coronary event or non‐fatal myo­
cardial infarction was significantly reduced by 24% in the
treatment group overall, demonstrating benefit of statin
use in secondary prevention even in the absence of raised
LDL levels. In the 586 subjects with diabetes mellitus
(14%), the reduction in the primary endpoint was not sig­
nificant but there was a 25% significant reduction in an
expanded coronary endpoint.
There is also evidence comparing high‐dose versus low‐
dose statin in diabetes mellitus. The Treating to New
Targets (TNT) study  [24] compared atorvastatin 80  mg
versus atorvastatin 10  mg in 10 001 people with stable
CHD. 1501 (15%) of the subjects had diabetes mellitus and
subgroup analysis showed high‐dose atorvastatin was ben­
eficial in significantly reducing the risk of major cardiovas­
cular events (CHD death, myocardial infarction, stroke,
resuscitation after cardiac arrest) by 25%. Safety analysis
demonstrated a significantly greater number of adverse
events in the intensively treated group, particularly an
increase in abnormal liver function tests, with no signifi­
cant differences between the rates of reported myalgia and
rhabdomyolysis in both groups.
The Pravastatin or Atorvastatin Evaluation and
Infection Therapy (PROVE IT) Thrombolysis in
Myocardial Infarction (TIMI) 22 trial  [25] compared
40 mg pravastatin versus 80 mg atorvastatin in people with
recent acute coronary syndromes. High intensity statin
treatment significantly reduced the primary extended car­
diovascular composite end point by 16% in the general
study population. Nearly one quarter of the subjects in
PROVE IT had diabetes mellitus and, although subgroup
analysis showed a reduction in primary end points in the
high intensity group, this did not reach statistical signifi­
cance. The authors attributed this to the sub study being
underpowered. A higher number of people in the intensive
arm had abnormal liver function tests, but the rates of dis­
continuation of statin therapy due to side effects was not
significantly different between groups.
There is also evidence of benefit in non‐statin therapies
for dyslipidemia. Ezetimibe reduces cholesterol absorption
by inhibiting Niemann‐Pick C1‐Like 1 (NPC1L1) protein
in the small intestine and hepatocytes. In the IMPROVE‐
IT (Improved Reduction of outcomes: Vytorin Efficacy
 202 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
International Trial) [26], the first trial to show an improve­
ment in cardiovascular (CV) outcomes with the addition of
a non‐statin drug to a statin, 18 144 patients who had been
hospitalized for an acute coronary syndrome (ACS) within
the preceding 10  days were randomized to simvastatin–
ezetimibe combination therapy or simvastatin monother­
apy. Participants with diabetes mellitus was a prespecified
subgroup in this trial. It had a median follow‐up of 6 years,
the addition of ezetimibe to simvastatin reduced LDL‐C by
16  mg/dL and resulted in a 6.4% reduction (32.7% vs.
34.7%; hazard ratio (HR) 0.936; 95% confidence interval
(CI) 0.89–0.99; p = 0.016) in the primary endpoint, which
was a composite of CV death, MI, unstable angina requir­
ing hospitalization, coronary revascularization, or stroke,
compared to simvastatin monotherapy.
Fibrates are agonists of peroxisome proliferator‐acti­
vated receptor alpha (PPARα), mediating transcription
factors that control lipoprotein metabolism. They improve
the lipid profile of diabetic dyslipidemia by decreasing TG
level and increasing HDL‐C level, but recent trials have
failed to show a benefit in outcomes, both with monother­
apy [27] and in addition to a statin [28]. In the ACCORD
(Action to Control Cardiovascular Risk in Diabetes) lipid
trial, the addition of fenofibrate to simvastatin in high‐risk
patients with type 2 diabetes mellitus did not reduce the
primary endpoint, which was a composite of MI, stroke, or
CV death, with a mean follow‐up of 4.7 years. However, in
a prespecified subgroup analysis, a possible benefit for
patients with both a high baseline TG level ≥ 204 mg/dL
and a low baseline level of HDL‐C ≤ 34  mg/dL was sug­
gested, with similar post hoc subgroup analysis in the
FIELD (Fenofibrate Intervention and Event Lowering in
Diabetes) study [29].
Evidence that PSCK 9 is associated with
increased CVD risk
The newest agents in the treatment of dyslipidemia are the
Proprotein Convertase Subtilisin/Kexin Type 9 (PCSK9)
Inhibitors. In 2003, Abifadel et al. were the first to identify
mutations in the genes encoding for PCSK9 as a cause of
autosomal familial hypercholesterolemia [30] Investigating
the role of PCSK9 showed its association with LDL recep­
tor intracellular degradation [31]. Plasma LDL cholesterol
binds to LDL receptors expressed on the surface of hepato­
cytes and is internalized by endocytosis. LDL receptors are
usually recycled to the cell surface, but when PCSK9 binds
with LDL receptors, LDL receptors are delivered to lys­
osomes for degradation, resulting in lower expression of
LDL receptors and an increase in LDL cholesterol. The
relationship between LDL cholesterol levels and athero­
sclerosis has been extensively studied. Higher LDL choles­
terol levels are strongly correlated with atherosclerosis and
higher cardiovascular comorbidities and mortality  [32].
Whereas gain‐of‐function mutations were associated with
increased levels of LDL cholesterol and early onset of ath­
erosclerosis, loss‐of‐function mutations on the other hand
was linked to a lower LDL cholesterol and a subsequent
decrease in cardiovascular risk [33]
The FOURIER [34] trial included 27 564 patients with
ASCVD and LDL ≥ 70  mg/dL who were receiving statin
therapy. Patients were randomly assigned to evolocumab
(140 mg every 2 weeks or 420 mg monthly) or placebo, and
at 48 weeks, evolocumab reduced LDL cholesterol by 59%
compared to placebo, from a median baseline value of
92  mg/dL to 30  mg/dL. With a median follow‐up of 2.2
years, evolocumab significantly reduced the primary end­
point, which was a composite of CV death, MI, stroke, hos­
pitalization for unstable angina, or coronary
revascularization, by 15%, and the key secondary endpoint,
which was a composite of CV death, MI, or stroke, by 20%.
Diabetes mellitus was present at baseline in 11 031 (40%)
patients, and in a prespecified secondary analysis, similar
efficacy in the primary and key secondary endpoints were
observed in patients with and without diabetes mellitus.
However, since patients with diabetes mellitus had a higher
baseline risk, they seemed to have a greater absolute risk
reduction in the primary endpoint over 3 years. This ben­
efit was driven largely by a greater absolute risk reduction
in coronary revascularization, and there was no difference
in the absolute risk reduction for the key secondary
endpoint.
ODYSSEY Outcomes [35] included 18 924 patients who
had been hospitalized for an ACS 1 to 12 months prior to
randomization. After a run‐in period of 2 to 16 weeks on
high‐intensity statins, patients with LDL cholesterol ≥
70 mg/dL, non‐HDL cholesterol ≥ 100 mg/dL, or apolipo­
protein B ≥ 80  mg/dL were randomized to alirocumab
 Are Statins the Optimal Therapy for Cardiovascular Risk in Patients with Diabetes?
(75 mg every 2 weeks) or placebo. A target LDL cholesterol
level of 25 to 50 mg/dL was specified, with up‐titration of
alirocumab to 150 mg every 2 weeks in patients with LDL
≥ 50  mg/dL and a blinded switch to placebo in patients
who consistently had LDL < 15 mg/dL. In the on‐treatment
analysis, which excluded LDL cholesterol values after pre­
mature treatment discontinuation or blinded switch to pla­
cebo, alirocumab reduced LDL by 61% from a mean LDL
of 96.4 mg/dL to 42.3 mg/dL at 1 year, and by 54.7% from a
mean LDL of 101.4 mg/dL to 53.3 mg/dL at 4 years. With a
median follow‐up of 2.8 years, alirocumab significantly
reduced the primary endpoint, which was a composite of
CHD death, MI, ischemic stroke, or unstable angina
requiring hospitalization, by 15%, and the secondary com­
posite endpoint of all‐cause death, MI, or ischemic stroke
by 14%. Although all‐cause death was significantly lower
with alirocumab, there were no significant differences in
CHD death and CV death. Roughly 30% of patients had
diabetes mellitus. This led to the ODYSSEY DM –
DYSLIPIDEMIA and ODYSSEY DM – INSULIN [36] tri­
als being undertaken. In ODYSSEY DM – DYSLIPIDAEMIA
participants with type 2 diabetes mellitus and mixed dys­
lipidemia (defined as non‐ HDL‐C > 100 mg/dl) were ran­
domized to open label alirocumab 75 mg every two weeks
or usual care for 24 weeks. 142 participants were included
for analysis and 95.1% of these participants had evidence of
coronary heart disease. In the treatment arm, 64.6%
achieved non‐HDL‐C of < 100 mg/dl compared with 23.8%
in the usual treatment arm.
In ODYSSEY DM – INSULIN, participants had type 2
diabetes mellitus treated with insulin and had an LDL>
70 mg/dl. They were randomized in a double‐blind fashion
to alirocumab 75  mg every two weeks or placebo for
24 weeks. The analysis included 177 participants and 86.4%
of them had evidence of coronary heart disease. The results
showed 65.4% of participants achieved non‐HDL‐C of <
100 mg/dl compared with 14.9% in the placebo group.
Recently, a trial looking at the use of inclisiran, an inter­
fering RNA shown to inhibit the hepatic synthesis of
PCSK9, in patients with heterozygous familial hypercho­
lesterolemia has been published [37]. This study contained
482 patients, of which only 48 had diabetes mellitus. The
treatment group showed a 38.1% reduction in LDL choles­
terol levels. This is another interesting pathway for lower­
203
ing LDL cholesterol levels. Also, as it is only a twice‐yearly
injection it offers a more practical alternative to the mono­
clonal antibody therapies targeting PCSK9.
Evidence linking hypertriglyceridemia to
CVD risk in diabetes mellitus
While there is substantial evidence that raised LDL choles­
terol levels are a risk factor for CVD in patients with diabe­
tes mellitus, whether hypertriglyceridemia independently
is a risk factor for CVD is a matter for debate. This is prob­
ably due to conflicting results in the studies performed.
In most studies assessing the role of hypertriglyceri­
demia in CVD development in diabetes mellitus patients
the triglyceride levels are obtained after an 8–12 hour fast.
It is important to note this when assessing triglyceride lev­
els in our patients as many present with triglyceride results
from non‐fasting samples. However, there is some evi­
dence that non‐fasting triglyceride levels may have stronger
associations with CVD risk than fasting levels [38]. In the
Women’s Health Study, non‐fasting triglyceride levels were
found to be independently associated with CVD events
while fasting triglycerides were not [39].
The combination of raised triglyceride levels, low HDL
levels and the presence of VLDL particles has been shown
to be associated with increased CVD risk [40] but the inde­
pendent role of triglyceride levels has proved
controversial.
On review of the available evidence, when differences in
HDL and LDL cholesterol levels are taken into account the
association between triglycerides and CVD risk is less
impressive. For example, in patients being treated with
statins, triglyceride level was not associated with CVD risk
in the Air Force/Texas Coronary Atherosclerosis
Prevention Study (AFCAPS/TexCAPS)  [41]. Triglyceride
levels were also not shown to have an association with
CVD risk in the Department of Veterans Affairs High‐
Density Lipoprotein Intervention Trial (VA‐HIT) [42].
It has also been shown that the risk of CVD does not
appear to be elevated in patients with inherited forms of
severe hypertriglyceridemia, unless the inherited form is
associated with increased apolipoprotein production or
associated with an increase in triglyceride‐rich remnant
particles [43].
 204 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
However, more recent data increasingly shows an inde­
pendent role of triglycerides in the development of CVD.
For example, a 2016 study that controlled for body mass
index showed that an elevated triglyceride level was associ­
ated with increased coronary plaque development in
patients whose LDL‐ Cholesterol was well controlled with
lipid‐lowering therapy [44].
There are also studies showing that triglyceride levels
are an independent marker for recurrent disease after MI
in those with controlled LDL [45]. It has also been shown
that non fasting triglycerides were associated with an
increased risk of CVD of 2.8 times for each 1  mmol/L
(89 mg/dL) increase in triglyceride levels [46].
The Hoorne study [47] was a population‐based cohort
study with 869  men and 948  women participants which
showed hypertriglyceridemia to be an independent risk
factor for CVD in those with abnormal blood glucose.
After 10 years of follow‐up, they found the age‐ and sex‐
adjusted hazard ratios for cardiovascular disease were 1.35
(1.11–1.64) and 1.71 (1.40–2.08) for high triglycerides and
high non‐HDL‐cholesterol respectively after mutual
adjustment. After stratification for glucose metabolism sta­
tus, the hazard ratios for cardiovascular disease for non‐
HDL‐cholesterol were 1.70 (1.31–2.21) in normal glucose
metabolism and 1.56 (1.12–2.18) in abnormal glucose
metabolism. Triglycerides were not a risk factor in subjects
with normal glucose metabolism, with a hazard ratio of
0.94 (0.73–1.22), but in subjects with abnormal glucose
metabolism, the hazard ratio for cardiovascular disease
was 1.54 (1.07–2.22). In subjects with abnormal glucose
metabolism, the hazard ratio for the combined presence of
high triglycerides and non‐HDL‐cholesterol was 2.12
(1.35–3.34).
In 2001 the influence of low high‐density lipoprotein
cholesterol and elevated triglyceride on coronary heart dis­
ease events and response to simvastatin therapy in 4S (the
Scandinavian Simvastatin Survival Study) reported that
patients with elevated LDL cholesterol, low HDL choles­
terol, and elevated triglycerides were more likely than
patients with isolated LDL cholesterol elevation to have
other characteristics of the metabolic syndrome, had
increased risk for CHD events on placebo, and received
greater benefit with simvastatin therapy. This was a sub
study analyses of the Helsinki Heart Study and the
Bezafibrate Infarction Prevention Study and included both
diabetic and non‐diabetic patients. The increased risk for
CHD related to elevated triglycerides was seen in both dia­
betic and non‐diabetic subgroups.
Nordestgaard et  al. examined the hypothesis that very
high levels of non‐fasting triglycerides were predictors of
myocardial infarction, ischemic heart disease and
death  [48]. This was a prospective cohort study of
7587 women and 6394 men from the general Danish popu­
lation followed from 1976–2004. Baseline non‐fasting tri­
glycerides were stratified into categories of increasing
severity of hypertriglyceridemia, and compared to triglyc­
eride levels of less than 88.5mg/dl (1 mmol/l). Non‐fasting
triglyceride levels of 442.5mg/dl (≥ 5 mmol/l) or more were
found to be predictive of CHD. This study also supported
the concept that non‐fasting triglyceride levels more
strongly predict CHD risk than levels measured after a
12–14 hour fast. Postprandial lipoproteins are generally tri­
glyceride‐rich, and if an individual has a predisposition to
small, dense LDL or has insulin‐resistance, then clearance
of these lipoprotein particles can be delayed for up to 12
hours. The authors thus postulate that prolonged exposure
of the endothelium to triglyceride‐rich, atherogenic rem­
nant particles, or the associated states in which atherogenic
lipoprotein particles occur, such as obesity or the metabolic
syndrome, may explain why non‐fasting triglycerides are
more predictive of CHD risk.
In summary, there is strong epidemiological evidence to
date that hypertriglyceridemia is an independent risk fac­
tor for CHD. Prospective randomized controlled trials
examining this link are difficult given the large number of
patients with diabetes mellitus who are already receiving
statin treatment (which modestly lowers triglycerides) and
due to the presence of other lipoprotein and metabolic
derangements which may confound results.
Evidence that treating
hypertriglyceridemia will decrease CVD
risk in DM
Now that we have reviewed the evidence linking elevated
triglyceride levels to increased cardiovascular risk in diabe­
tes mellitus patients, we next need to establish if there is
evidence that treating this lipid abnormality decreases this
 Are Statins the Optimal Therapy for Cardiovascular Risk in Patients with Diabetes?
risk. Fenofibrates, niacin and omega 3 fish oils are some
of the pharmacological interventions for treating
hyper­triglyceridemia.
In the Bezafibrate Infarction Prevention (BIP), 3090
patients with a previous myocardial infarction or stable
angina, total cholesterol of 180 to 250 mg/dl, HDL‐choles­
terol < 45 mg/dl, triglycerides < 300 mg/dl, and LDL‐cho­
lesterol < 180  mg/dl were randomized to receive either
400 mg of bezafibrate daily or placebo, and followed for a
mean of 6.2 years [49]. Bezafibrate reduced triglycerides by
21% and raised HDL cholesterol by 18%. No difference was
apparent in the all‐cause and cardiac mortality between the
bezafibrate and placebo groups. However, on post‐hoc
analysis, there was a significant reduction in the primary
end point in 459 patients with high baseline triglycerides
(≥ 200 mg/dl or 2.26 mmol/l). The reduction in the cumu­
lative probability of the primary endpoint by bezafibrate
was 39.5% (P=0.02). Bezafibrate may thus have a promi­
nent role in the management of dyslipidemia and CHD
when targeted to a subgroup of patients with CHD. This
supports the evidence from meta‐analyses and epidemio­
logical studies that triglycerides are indeed an independent
risk factor for CHD in patients with both normal and
abnormal glucose metabolism.
The authors subsequently evaluated the effect of bezafi­
brate on the incidence of MI in patients enrolled in the BIP
study who met the criteria for the metabolic syndrome [50].
The study sample for this post hoc subgroup analysis com­
prised 1470 patients aged 42 to 74 years who were randomly
assigned to receive bezafibrate 400 mg daily (740 patients) or
placebo (730 patients). The follow‐up period was 6.2 years
for events and 8.1 years for mortality data. New myocardial
infarction was recorded in 82 patients from the bezafibrate
group (11.1%) and 111 patients (15.2%) from the placebo
group (P=0.02). Bezafibrate was associated with a reduced
risk of any MI and non‐fatal MI with hazard ratios of 0.71
(95% CI, 0.54–0.95) and 0.67 (95% CI, 0.49–0.91) respec­
tively. The cardiac mortality risk tended to be lower in
patients taking bezafibrate (HR, 0.74; 95% CI, 0.54–1.03).
However, posthoc analysis on the effect on MI risk of specifi­
cally lowering triglycerides within this group with metabolic
syndrome was not performed.
Regarding fenofibrate, HDL, and cardiovascular disease
in Type‐2 diabetes, the DAIS trial [51] screened 731 type 2
205
diabetes mellitus patients, of which 418 were randomized
to either receive fenofibrate 200 mg od or placebo for three
years. This was both a primary and secondary prevention
trial as one‐half of the participants had a history of CVD.
Total plasma cholesterol, HDL‐cholesterol, LDL‐choles­
terol, and triglyceride concentrations all changed signifi­
cantly more than baseline in the fenofibrate group (n=207)
than in the placebo group (n=211). The fenofibrate group
showed a significantly smaller increase in percentage
diameter stenosis on coronary angiogram than the placebo
group (mean 2.11 vs 3.65, P=0.02). Although the trial was
not powered to examine clinical endpoints, there were
fewer in the fenofibrate than placebo group (38 vs 50).
Although there was angiographic benefit in the setting of
reduction of hypertriglyceridemia, as in most trials on tri­
glycerides it is difficult to deduce if this is a direct triglycer­
ide‐lowering effect or the result of correction of other
lipoprotein abnormalities.
Niacin or nicotinic acid has been shown to reduce tri­
glyceride levels and increase HDL by up to 30% when
doses of 1.5–3 g daily are used [52]. In older niacin trials
for secondary cardiovascular prevention before the advent
of statin therapy, absolute mortality reductions of 6.2% and
7.8% were demonstrated, compared to the best absolute
mortality reduction of 3.5% with a statin in the
Scandinavian Simvastatin Survival Study [53].
However, there is no evidence to suggest outcomes are
improved with niacin for those already established on statin
therapy. The HPS2‐THRIVE (Heart Protection Study
2‐Treatment of HDL to Reduce the Incidence of Vascular
Events) study of 25 673 patients, which examined the impact
of niacin 2 g daily plus laropiprant (to reduce flushing) ver­
sus double placebo [54], was conducted on background sim­
vastatin ± ezetimibe therapy. After a median follow‐up of 3.9
years, the primary endpoint of first major vascular event was
not significantly lower in niacin‐treated patients. Patients on
niacin experienced more serious adverse events including
myopathy, gastrointestinal effect, rash, and an increase in
infections and bleeding. This study also demonstrated unfa­
vorable glycemic effects, with an increase in both new cases
of diabetes mellitus and significant worsening of glycemic
control in patients with established disease.
Omega‐3 fatty acid preparations containing eicosapen­
taenoic acid (EPA) and/or docosahexaenoic acid (DHA)
 206 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
reduce fasting and postprandial triglycerides through sup­
pression of hepatic VLDL production  [55]. They are fre­
quently used as an adjutant to other therapies. Doses of 3–4
g per day of EPA ± DHA are required to reduce triglycerides
by up to 45% [56]. Outcomes trials yielded mixed results for
cardiovascular benefit with the use of omega‐3 fatty
acids  [57, 58]. However, doses used in initial studies were
much lower than currently recommended for hypertriglyc­
eridemia (1 g instead of 3–4 g), and these low doses were
examined on a background statin therapy. A subsequent
landmark study in Japan utilized a moderate dose of EPA
ethyl esters (1.8 g) in conjunction with low‐dose statin ther­
apy versus statin therapy alone, and observed a 19% reduc­
tion in major coronary events in those receiving EPA [59].
Discussion
There are exciting new developments in the treatment of
dyslipidemia in diabetes mellitus patients with the arrival
of the PCSK 9  inhibitors. These agents are changing the
landscape of lipid management but cost remains a barrier
to widespread adoption. As we have shown above, hyper­
triglyceridemia is associated with the metabolic syndrome.
While LDL is an established independent risk factor for
CVD in type 2 diabetes mellitus, there are still controver­
sies surrounding the evidence that suggests aggressively
lowering triglycerides with pharmacotherapy is of benefit
in the diabetic population.
We would recommend statin treatment for all patients
with type 2 diabetes mellitus to lower LDL, thus reducing
cardiovascular risk. The ADA recommends initiating a sta­
tin for primary prevention when LDL is > 130  mg/dL
(3.36 mmol/L) with the goal of reducing LDL to 100 mg/dl
(< 2.57  mmol/l). Target LDL in current guidelines is <
2.6  mmol/l and < 1.8  mmol/L for secondary prevention.
We recommend the lowest possible LDL for secondary pre­
vention based on available trial data. We would suggest that
a target of LDL cholesterol < 70 mg/dl (1.8 mmol/l) may be
reasonable in patients with type two diabetes mellitus and
CVD. This, however, must be taken in the context of
increased risk of adverse events related to statin‐lowering
therapy on higher doses of these agents. Optimal HDL lev­
els are > 40 mg/dL (1.02 mmol/L) in men; a higher target
may be more desirable in women (> 50  mg/dL or
1.28 mmol/L) based on physiologically higher HDL levels
in the female population. Increasing HDL cholesterol levels
pharmacologically in patients with diabetes mellitus is dif­
ficult since the most effective agent is nicotinic acid, which
is relatively contraindicated in diabetes mellitus. We agree
with current ADA guidelines that triglycerides are a recog­
nized target for intervention in diabetes and that optimal
triglyceride levels are < 150 mg/dL (1.7 mmol/L).
Now to relate these guidelines to common clinical sce­
narios encountered in our practice. Firstly, we will look at
the scenario described at the beginning of the chapter
where our patient’s LDL was at target but they had an iso­
lated hypertriglyceridemia. The initial therapeutic inter­
vention for hypertriglyceridemia is behavioral modification
with weight loss, increased physical activity and modera­
tion of alcohol consumption. Though not required in our
scenario, should the patient have poor glycemic control we
would advise intensification of their diabetic treatment
regimen.
Another clinical scenario to consider is the patient with
mixed dyslipidemia as is commonly seen in the type 2 dia­
betes mellitus population. If the patient is not receiving
treatment, we would advise the behavioral modifications
outlined above and improvements in glycemic control. If
this fails to achieve the required targets, we would advise
statin as first‐line therapy. The decision as to whether to
start high‐dose or low‐dose statin therapy, we would argue,
needs to be individualized to each patient after reviewing a
number of factors. Initially, we need to look for the pres­
ence of other CVD risk factors and, if present, assess how
well are these risk factors are controlled. The greater the
CVD risk the more likely we would be to initiate high‐dose
statin therapy, especially if the modifiable risk factors are
poorly controlled. The trials looking at high‐dose statins
versus low‐dose statins were undertaken in participants
with established CVD and so these are the individuals that
should be selected for high‐dose therapy in our patient
cohort. Next, we would consider the degree of the derange­
ment in LDL and TG. Those with only modest increase
with a low CVD risk and no evidence of CVD should be
started on low‐dose statin therapy. Lastly, we need to con­
sider if this patient is at risk from any of the described sta­
tin side effects and consider whether gradual dose titration
would be more successful in this population.
 Are Statins the Optimal Therapy for Cardiovascular Risk in Patients with Diabetes?
The third scenario is the patient on statin therapy who
has an LDL cholesterol at or below target, but triglycerides
are elevated and HDL suboptimal. The physician may con­
sider adding a fibrate or fish oil therapy. We would not rec­
ommend niacin be used routinely in the diabetic mellitus
population. Unfortunately, the combination of a statin and
fibrate increases the risk of side effects such as myopathy.
Therefore, for more modest hypertriglyceridemia (1.7–
3.0 mmol/l), we would recommend the addition of fish oil
therapy to help achieve targets. A Cochrane review of 23
randomized control trials involving 1075 participants with
type 2 diabetes mellitus showed that supplementation with
fish oil (average dose of 3.5 g/day for an average of
8.9 weeks) lowered TG levels by 8.1 mg/dl and VLDL by
1.26  mg/dl  [60]. LDL levels increased by 1.98  mg/dl and
there was no change in HDL levels. There is a lack of strong
evidence to guide management in this scenario. It would be
reasonable to monitor the TG levels and only intervene if
approaching levels where pancreatitis could develop (TG >
5  mmol/L) or in case of existing CAD (although no evi­
dence to support this). Features consistent with familial
combined hyperlipidemia (varying phenotype over time or
family history) might also prompt intervention. However,
should the above patient have an LDL and TG level above
target then the risk/benefit balance could be in favor of
combination statin and fibrate therapy. Fibrates such as
fenofibrate, gemfibrozil, or bezafibrate can be considered
as an adjunct to statin therapy.
These are our current recommendations for managing
dyslipidemia in type 2 diabetes mellitus. In the future,
when further long‐term studies emerge relating the PCSK9
agents, we may find that our LDL targets have been too
conservative and more aggressive targets may become
standard practice.
References
1. Taskinen MR. Diabetic dyslipidaemia: from basic research to
clinical practice. Diabetologia. 2003;46:733–749.
2. Savage DB, Petersen KF, Shulman GI. Disordered lipid
metabolism and the pathogenesis of insulin resistance.
Physiol Rev. 2007;87:507–520.
3. C Caste, Kuiper S, Blake W, Paigen B, Marotti K, Melchior G.
Remodeling of the HDL in NIDDM has a fundamental role
207
for cholesteryl ester transfer protein. Am J Physiol. 1998;274:
E1091–1098.
4. Barter PJ, Caulfield M, Eriksson M et  al.; ILLUMINATE
Investigators. Effects of torcetrapib in patients at high risk
for coronary events. N Engl J Med. 2007;357:2109–2122.
5. Kastelein JJ, van Leuven SI, Burgess L et  al.; RADIANCE
1 Investigators. Effect of torcetrapib on carotid atherosclero­
sis in familial hypercholesterolemia. N Engl J Med.
2007;356:1620–1630.
6. Lincoff AM, Nicholls SJ, Riesmeyer JS et al.; ACCELERATE
Investigators. Evacetrapib and cardiovascular outcomes
in high‐risk vascular disease. N Engl J Med. 2017;376:
1933–1942.
7. Bowman L, Hopewell JC, Chen F et  al.; HPS3/TIMI55–
REVEAL Collaborative Group. Effects of anacetrapib in
patients with atherosclerotic vascular disease. N Engl J Med.
2017;377:1217–1227pmid:28847206.
8. Krauss RM. Heterogeneity of plasma low‐density lipopro­
teins and atherosclerosis risk. Curr Opin Lipidol. 1994;5:
339–349.
9. Lamarche B, Chernof A, Moorjani S et al. Small, dense low‐
density lipoprotein particles as a predictor of the risk of
ischemic heart disease in men. Prospective results from the
Quebec Cardiovascular Study. Circulation. 1997;95:69–75.
10. Chapman MJ, Guerin M, Bruckert E. Atherogenic, dense
low‐density lipoproteins. Pathophysiology and new thera­
peutic approaches. Eur Heart J. 1998;19(Suppl A): A24–30.
11. Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH,
Franklin BA, Gordon DJ, Krauss RM, Savage PJ, Smith SC Jr,
Spertus JA, Costa F, American Heart Association, National
Heart, Lung, and Blood Institute. Diagnosis and manage­
ment of the metabolic syndrome: an American Heart
Association/National Heart, Lung, and Blood Institute
Scientific Statement. Circulation. 2005;Oct 25;112(17):
2735–2752.
12. Giugliano, RP, Cannon CP, Blazing MA, Nicolau JC,
Corbalan R, Spinar J, Park JG, White JA, Bohula E,
Braunwald E. Benefit of adding ezetimibe to statin therapy
on cardiovascular outcomes and safety in patients with vs.
without diabetes: results from IMPROVE‐IT. Circulation.
2018;Apr 10;137(15):1571–1582.
13. Goff DC, Gerstein HC, Ginsberg HN, Cushman WC,
Margolis KL, Byington RP, Buse JB, Genuth S, Probstfield JL,
Simons‐Morton DG. Prevention of cardiovascular disease in
persons with type 2 diabetes mellitus: current knowledge
and rationale for the Action to Control Cardiovascular Risk
in Diabetes (ACCORD) trial. Am J Cardiol. 2007;99: 4i–20i.
 208 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
14. Colhoun HM, Betteridge DJ, Durrington PN et al. on behalf
of the CARDS Investigators. Primary prevention of cardio­
vascular disease with atorvastatin in type 2 diabetes in the
Collaborative Atorvastatin Diabetes Study (CARDS): multi­
centre randomised placebo‐controlled trial. Lancet. 2004;
364:685–696.
15. Knopp RH, d’Emden M, Smilde JG, Pocock SJ, on behalf of
the ASPEN Study Group. Efficacy and safety of atorvastatin
in the prevention of cardiovascular end points in subjects
with type 2 diabetes: the Atorvastatin Study for Prevention
of Coronary Heart Disease Endpoints in Non‐insulin‐
dependent diabetes mellitus (ASPEN). Diabetes Care. 2006;
29:1478–1485.
16. Heart Protection Study Collaborative Group. MRC/BHF
Heart Protection Study of cholesterol lowering with simvas­
tatin in 5963 people with diabetes: a randomised placebo‐
controlled trial. Lancet. 2003;361:2005–2016.
17. Sever PS, Poulter NR, Dahlöf B et  al. for the ASCOT
Investigators. Reduction in cardiovascular events with ator­
vastatin in 2,532 patients with type 2 diabetes. Anglo‐
Scandinavian Cardiac Outcomes Trial‐Lipid Lowering Arm
(ASCOT‐LLA). Diabetes Care. 2005;28:1151–1157.
18. Nakamura H, Arakawa K, Itakura H et  al. for the MEGA
Study Group. Primary prevention of cardiovascular disease
with pravastatin in Japan (MEGA Study): a prospective ran­
domised controlled trial. Lancet. 2006;368:1155–1163.
19. ALLHAT Collaborative Research Group. Major outcomes in
moderately hypercholesterolemic, hypertensive patients
randomized to pravastatin vs usual care. The
Antihypertensive and Lipid‐Lowering Treatment to Prevent
Heart Attack Trial (ALLHAT‐LLA). JAMA. 2002;288:
2998–3007.
20. Pyörälä K, Pedersen TR, Kjekshus J, Faergeman O, Olsson
AG, Thorgeirsson G. Cholesterol lowering with simvastatin
improves prognosis of diabetic patients with coronary heart
disease: a subgroup analysis of the Scandinavian Simvastatin
Survival Study (4S). Diabetes Care. 1997;20:614–620.
21. Haffner SM, Alexander CM, Cook TJ et  al. for the
Scandinavian Simvastatin Survival Study Group. Reduced
coronary events in simvastatin‐treated patients with coro­
nary heart disease and diabetes or impaired fasting glucose
levels. Arch Intern Med. 1999;159:2661–2667.
22. The Long‐term Intervention with Pravastatin in Ischaemic
Disease (LIPID) Study Group. Prevention of cardiovascular
events and death with pravastatin in patients with coronary
heart disease and a broad range of initial cholesterol levels. N
Engl J Med. 1998;339:1349–1357.
23. Sacks FM, Pfeffer MA, Moye LA et  al. for the Cholesterol
And Recurrent Events Trial Investigators. The effect of
pravastatin on coronary events after myocardial infarction
in patients with average cholesterol levels. N Engl J Med.
1996;335:1001–1009.
24. LaRosa JC, Grundy SM, Waters DD et al. for the Treating to
New Targets (TNT) Investigators. Intensive lipid lowering
with atorvastatin in patients with stable coronary disease. N
Engl J Med. 2005;352:1425–1435.
25. Cannon CP, Braunwald E, McCabe CH et  al. for the
Pravastatin or Atorvastatin Evaluation and Infection
Therapy‐Thrombolysis in Myocardial Infarction 22 Investi­
gators. Intensive versus conventional lipid lowering with
statins after acute coronary syndromes. N Engl J Med. 2004;
350:1495–1504.
26. Cannon CP, Blazing MA, Giugliano RP, McCagg A, White
JA, Theroux P, Darius H, Lewis BS, Ophuis TO, Jukema JW
et al. Ezetimibe added to statin therapy after acute coronary
syndromes. N Engl J Med. 2015;372:2387–2397.
27. Keech A, Simes RJ, Barter P, Best J, Scott R, Taskinen MR,
Forder P, Pillai A, Davis T, Glasziou P et al. Effects of long‐
term fenofibrate therapy on cardiovascular events in 9795
people with type 2 diabetes mellitus (the FIELD study):
Randomised controlled trial. Lancet. 2005;366:1849–1861.
28. Ginsberg HN, Elam MB, Lovato LC, Crouse JR, 3rd, Leiter
LA, Linz P, Friedewald WT, Buse JB, Gerstein HC,
Probstfield J et  al. Effects of combination lipid therapy in
type 2 diabetes mellitus. N Engl J Med. 2010;362:
1563–1574.
29. Ginsberg HN, Elam MB, Lovato LC, Crouse JR, 3rd, Leiter
LA, Linz P, Friedewald WT, Buse JB, Gerstein HC,
Probstfield J et  al. Effects of combination lipid therapy in
type 2 diabetes mellitus. N Engl J Med. 2010;362:
1563–1574.
30. Abifadel M, Varret M, Rabès JP, Allard D, Ouguerram K,
Devillers M, Cruaud C, Benjannet S, Wickham L, Erlich D,
Derré A, Villéger L, Farnier M, Beucler I, Bruckert E,
Chambaz J, Chanu B, Lecerf JM, Luc G, Moulin P,
Weissenbach J, Prat A, Krempf M, Junien C, Seidah NG,
Boileau C. Mutations in PCSK9 cause autosomal dominant
hypercholesterolemia. Nat Genet. 2003;Jun;34(2):154–156.
31. Abifadel M, Varret M, Rabes JP, Allard D, Ouguerram K,
Devillers M, Cruaud C, Benjannet S, Wickham L, Erlich D
et al. Mutations in PCSK9 cause autosomal dominant hyper­
cholesterolemia. Nat Genet. 2003;34:154–156.
32. Maiolino G, Rossitto G, Caielli P, Bisogni V, Rossi GP, Calò
LA. The role of oxidized low‐density lipoproteins in athero­
 Are Statins the Optimal Therapy for Cardiovascular Risk in Patients with Diabetes?
sclerosis: the myths and the facts. Mediators Inflamm.
2013;2013:714653.
33. Cohen JC, Boerwinkle E, Mosley TH, Jr, Hobbs HH.
Sequence variations in PCSK9, low LDL, and protection
against coronary heart disease. N Engl J Med. 2006;354(12):
1264–1272.
34. Sabatine MS, Giugliano RP, Keech AC, Honarpour N,
Wiviott SD, Murphy SA, Kuder JF, Wang H, Liu T,
Wasserman SM et al. Evolocumab and clinical outcomes in
patients with cardiovascular disease. N Engl J Med. 2017;376:
1713–1722.
35. Steg PG. Evaluation of cardiovascular outcomes after an
acute coronary syndrome during treatment with alirocumab
– ODYSSEY OUTCOMES. In Proceedings of the American
College of Cardiology Annual Scientific Session (ACC
2018), Orlando, FL, USA, 10–12 March 2018.
36. Ray KK et al. Alirocumab therapy in individuals with type 2
diabetes mellitus and atherosclerotic cardiovascular disease:
analysis of the ODYSSEY DM‐DYSLIPIDEMIA and DM‐
INSULIN studies. Cardiovasc Diabetol. 2019;18:149 https://
doi.org/10.1186/s12933‐019‐0951‐9.
37. Raal, FJ, Kallend DR, Kausik K, Turner T, Koenig W, Wright
R Scott, Wijngaard, PLJ, Curcio D, Jaros MJ, Leiter LA,
Kastelein JJP. Inclisiran for the treatment of heterozygous
familial hypercholesterolemia, 2020/03/18, N Engl J Med.
1520–1530, 382, 16, 10.1056/NEJMoa1913805.
38. Di Angelantonio E, Sarwar N, Perry P, Kaptoge S, Ray KK,
Thompson A, Wood AM, Lewington S, Sattar N, Packard CJ,
Collins R, Thompson SG, Danesh J. Emerging risk factors
collaboration – major lipids, apolipoproteins, and risk of
vascular disease. JAMA. 2009;302(18):1993–2000.
39. Nordestgaard BG, Benn M, Schnohr P, Tybjaerg‐Hansen A.
Nonfasting triglycerides and risk of myocardial infarction,
ischemic heart disease, and death in men and women.
JAMA. 2007;298(3):299–308.
40. Musunuru K. Atherogenic dyslipidemia: cardiovascular risk
and dietary intervention. Lipids. 2010;Oct;45(10):907–914.
41. Gotto AMJ Jr, Whitney E, Stein EA, Shapiro DR, Clearfield
M, Weis S, Jou JY, Langendörfer A, Beere PA, Watson DJ,
Downs JR, de Cani JS. Relation between baseline and on‐
treatment lipid parameters and first acute major coronary
events in the Air Force/Texas Coronary Atherosclerosis
Prevention Study (AFCAPS/TexCAPS). Circulation. 2000;
101(5):477–484.
42. Robins SJ, Collins D, Wittes JT, Papademetriou V, Deedwania
PC, Schaefer EJ, McNamara JR, Kashyap ML, Hershman JM,
Wexler LF, Rubins HB, VA‐HIT Study Group. Veterans
209
Affairs High‐Density Lipoprotein Intervention Trial.
Relation of gemfibrozil treatment and lipid levels with major
coronary events: VA‐HIT: a randomized controlled trial.
JAMA. 2001;Mar 28;285(12):1585–1591.
43. Miller M, Stone NJ, Ballantyne C, Bittner V, Criqui MH,
Ginsberg HN, Goldberg AC, Howard WJ, Jacobson MS, Kris‐
Etherton PM, Lennie TA, Levi M, Mazzone T, Pennathur S,
American Heart Association Clinical Lipido­logy, Thrombosis,
and Prevention Committee of the Council on Nutrition,
Physical Activity, and Metabolism, Council on Arteriosclerosis,
Thrombosis and Vascular Biology, Council on Cardiovascular
Nursing, Council on the Kidney in Cardiovascular Disease.
Triglycerides and cardiovascular disease: a scientific statement
from the American Heart Association. Circulation. 2011;May
24;123(20):2292–2333.
44. Puri R, Nissen SE, Shao M, Elshazly MB, Kataoka Y, Kapadia
SR, Tuzcu EM, Nicholls SJ. Non‐HDL Cholesterol and tri­
glycerides: implications for coronary atheroma progression
and clinical events. Arterioscler Thromb Vasc Biol.
2016;Nov;36(11):2220–2228.
45. Miller M, Cannon CP, Murphy SA, Qin J, Ray KK, Braunwald
E, PROVE IT‐TIMI 22 Investigators. Impact of triglyceride
levels beyond low‐density lipoprotein cholesterol after acute
coronary syndrome in the PROVE IT‐TIMI 22 trial. J Am
Coll Cardiol. 2008;Feb 19;51(7):724–730.
46. Watts GF, Karpe F. Triglycerides and atherogenic dyslipidae­
mia: extending treatment beyond statins in the high‐risk
cardiovascular patient. Heart. 2011;97(5):350–356.
47. Bos G, Dekker JM, Nijpels G, de Vegt FA, Diamant M,
Stehouwer CDA, Bouter LM, Heine RJ. A combination of
high concentrations of serum triglyceride and non‐HDL
cholesterol is a risk factor for cardiovascular disease in sub­
jects with abnormal glucose metabolism: The Hoorn Study.
Diabetologia. 2003;46:910–916.
48. Nordestgaard BG, Benn M, Schnohr P et al. Nonfasting triglyc­
erides and risk of myocardial infarction, ischemic heart disease
and death in men and women. JAMA. 2007;298(3):299–308.
49. The Bezafibrate Infarction Prevention (BIP) study.
Secondary prevention by raising HDL cholesterol and
reducing triglycerides in patients with coronary artery dis­
ease. Circulation. 2000;102:21–27.
50. Tenenbaum A, Motro M, Enrique Z et al. Bezafibrate for the
secondary prevention of myocardial infarction in patients
with metabolic syndrome. Arch Intern Med. 2005;165:
1154–1160.
51. Steiner G et  al. Effect of fenofibrate on progression of
coronary‐­artery disease in type 2 diabetes: the Diabetes
 210 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
Atherosclerosis Intervention Study (DAIS), a randomised
study. Lancet. 2001;357:905–910.
52. Chapman MJ, Redfern JS, McGovern ME, Giral P. Niacin
and fibrates in atherogenic dyslipidemia: pharmacotherapy
to reduce cardiovascular risk. Pharmacol Ther.
2010;126(3):314–345.
53. Superko HR, Zhao XQ, Hodis HN, Guyton JR. Niacin and
heart disease prevention: engraving its tombstone is a mis­
take. J Clin Lipidol. 2017;11(6):1309–1317.
54. Group HTC, Landray MJ, Haynes R, Hopewell JC, Parish S,
Aung T et  al. Effects of extended‐release niacin with laro­
piprant in high‐risk patients. N Engl J Med. 2014;371(3):
203–212.
55. Nestel PJ, Connor WE, Reardon MF, Connor S, Wong S,
Boston R. Suppression by diets rich in fish oil of very low
density lipoprotein production in man. J Clin Invest.
1984;74(1):82–89.
56. Harris WS, Ginsberg HN, Arunakul N, Shachter NS,
Windsor SL, Adams M et al. Safety and efficacy of Omacor
in severe hypertriglyceridemia. J Cardiovasc Risk.
1997;4(5–6):385–391.
57. The ORIGIN Trial Investigators, Bosch J, Gerstein HC,
Dagenais GR, Diaz R, Dyal L et al. n‐3 fatty acids and cardio­
vascular outcomes in patients with dysglycemia. N Engl J
Med. 2012;367(4):309–318.
58. Risk and Prevention Study Collaborative Group, Roncaglioni
MC, Tombesi M, Avanzini F, Barlera S et al. n‐3 fatty acids in
patients with multiple cardiovascular risk factors. N Engl J
Med. 2013;368(19):1800–1808.
59. Yokoyama M, Origasa H, Matsuzaki M, Matsuzawa Y, Saito
Y, Ishikawa Y et al. Effects of eicosapentaenoic acid on major
coronary events in hypercholesterolaemic patients (JELIS): a
randomised open‐label, blinded endpoint analysis. Lancet.
2007;369(9567):1090–1098.
60. Hartweg J, Perera R, Montori V, Dinneen S, Neil HA, Farmer
A. Omega‐3 polyunsaturated fatty acids (PUFA) for type 2
diabetes mellitus. Cochrane Database Syst Rev. 2008;(1):
CD003205.
 17 New Agents for Treatment
of Dyslipidemia
Vinaya Simha
Associate Professor, Division of Endocrinology, Diabetes, Metabolism and nutrition, Mayo Clinic, Rochester, MN, USA

LE A R NI NG P OI NT S

●● Diabetes mellitus is associated with elevated risk for atherosclerotic cardiovascular disease (ASCVD), and statin therapy has
been shown to reduce this risk.
●● Despite optimal statin therapy, there is considerable residual risk, and additional lipid‐lowering therapies may help
ameliorate it.
●● Proprotein convertase subtilisin/kexin type 9 (PCSK9) is a secreted protein which modulates the concentration of low‐den-
sity lipoprotein (LDL)‐cholesterol by promoting lysosomal degradation of LDL receptors. Gain‐of‐function mutations in
PCSK9 is a cause of familial hypercholesterolemia, while loss‐of‐function mutations are associated with low LDL‐cholesterol
and reduced ASCVD risk.
●● Monoclonal antibodies to PCSK9 have been developed (evolocumab and alirocumab), and shown to reduce LDL‐choles-
terol by 50–60% when administered as monotherapy or added to statin therapy.
●● Cardiovascular outcome trails have shown a significant risk reduction (HR = 0.85, ARR = 1.6–2) with both evolocumab and
alirocumab when added on to statin therapy. These benefits were similar in patients with and without diabetes, and were
not associated with significant adverse effects.
●● Current guidelines recommend the addition of PCSK9 inhibitors to maximally tolerated statin therapy in: 1. patients with
familial hypercholesterolemia, 2. patients with established ASCVD whose LDL‐cholesterol ≥ 70 mg/dL on statin +
ezetimibe.
●● PCSK9 inhibitor therapy should be considered in patients with diabetes and ASCVD whose non‐HDL cholesterol ≥ 100 mg/
dL on maximally tolerated statin therapy.
●● Novel triglyceride‐lowering therapies which are being developed also have the potential for lowering ASCVD risk when
added to statin therapy, but need further investigation.

Introduction
Diabetes mellitus increases the risk for atherosclerotic car­
diovascular disease (ASCVD) at least two‐fold, with
increasing risk as the duration of diabetes increases  [1].
Indeed ASCVD is the leading cause for morbidity and
mortality in patients with type 2 diabetes  [2], and the
recent heartening trend of decreasing ASCVD mortality in
the overall population  [3] is unfortunately not seen in
patients with diabetes  [4]. It is therefore imperative to
aggressively address cardiovascular risk in diabetes. Stain
therapy has been consistently shown to reduce ASCVD
risk in patients with diabetes [5], and has been strongly rec­
ommended in most patients with diabetes  [6]. Together
with appropriate lifestyle changes and glucose‐lowering
therapies, this will significantly reduce the burden of
ASCVD. However, considerable residual risk persists even
after adequate statin therapy, with an estimated 1  in 7
patients experiencing an ASCVD event within 5 years

Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.

© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.

211
 212 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
despite ongoing statin therapy [7]. This has prompted an
intense search for additional and novel therapies that
reduce the levels of atherogenic lipoproteins and overall
ASCVD risk.
Patients with type 2 diabetes have characteristic lipid
abnormalities marked by hypertriglyceridemia, low HDL‐
cholesterol and elevated levels of small, dense LDL‐choles­
terol [8, 9]. Since hypertriglyceridemia is the most common
lipid abnormality in type 2 diabetes, considerable attention
has focused on primary triglyceride lowering therapies
such as fibrates, omega‐3 fatty acid and niacin. Except for
the recent REDUCE‐IT trial showing the benefit of 4 g
eicosapent‐ethyl [10], other studies have not shown a sig­
nificant benefit in statin treated patients, though sub‐group
analysis has suggested some risk reduction with fibrate
therapy  [11]. Additional LDL‐cholesterol lowering using
ezetimibe has been shown to reduce risk in post‐Acute
Coronary Syndrome (ACS) patients  [12]. However, the
most promising development in recent years has been the
introduction of a novel class of lipid‐lowering medications,
Proprotein convertase subtilisin/kexin type 9 (PCSK9)
inhibitors. This review briefly summarizes the role of
PCSK9 in regulation of circulating cholesterol levels, clini­
cal trials of PCSK9  inhibitors, and their optimal use in
clinical practice, followed by a brief discussion on emerg­
ing triglyceride‐lowering therapies.
PCSK9 and regulation of LDL‐Cholesterol
levels
The role of PCSK9 in regulation of circulating cholesterol
levels was first appreciated around the turn of the century,
when Abifadel and colleagues identified gain‐of‐function
mutations in PCSK9 as a cause for autosomal dominant
familial hypercholesterolemia  [13]. Since then, there has
been considerable progress in understanding the physio­
logical role of this protein  [14, 15] which has also led to
rapid therapeutic advances [16].
PCSK9 is a circulating serine protease which binds to
the LDL receptor, leading to their lysosomal degradation.
Interestingly, the expression of both LDL receptors and
PCSK9 are regulated by the same transcription factor,
sterol regulatory element binding protein (SREBP), and is
increased in states of cellular cholesterol deficiency. The
LDL receptor participates in receptor‐mediated endocyto­
sis of circulating LDL particles, thus helping to restore cel­
lular cholesterol levels, while simultaneously decreasing
LDL particles in the circulation. After internalization
through clathrin‐coated pits on the hepatocyte membrane,
the receptor dissociates from the LDL particle and recycles
back to the surface to continue this process nearly 100–150
times during its lifecycle. However, when the LDL receptor
is bound to PCSK9, it is unable to dissociate from the LDL
particle, and undergoes degradation along with the LDL
particle and PCSK9 within the lysosome (Figure 17.1). The
net effect of PCSK9 is reduced recycling of LDL receptor to
the hepatocyte membrane and resultant decrease in clear­
ance of circulating LDL particles. Indeed, overexpression
of PCSK9 in mice is associated with high circulating cho­
lesterol concentrations [17], while PCSK9 deficiency leads
to low cholesterol levels [18]. These experiments were fol­
lowed by epidemiological studies which clearly showed
decreased LDL‐cholesterol and ASCVD risk in individuals
harboring deleterious sequence variations in PCSK9 [19–
21]. This has been the basis for development of pharmaco­
logical approaches to decrease PCSK9 activity and thus
circulating LDL‐cholesterol levels.
PCSK9 inhibitors: Efficacy and safety
Several monoclonal antibodies and a small inhibiting RNA
(siRNA) have been developed for inhibition of PCSK9
activity. Evolocumab and alirocumab are two fully human
monoclonal antibodies currently approved for clinical use.
They are administered by subcutaneous injection, usually
every two weeks (evolocumab 140 mg and alirocumab 75
or 150 mg). They can also be administered every 4 weeks,
evolocumab at a dose of 420 mg, and alirocumab at a dose
of 300 mg, respectively.
Several phase III trials have been conducted in a variety
of patient populations including those with familial hyper­
cholesterolemia, statin intolerance and moderate hyper­
cholesterolemia, both as monotherapy and in combination
with statin therapy. In general, the reduction in LDL‐cho­
lesterol was about 60%, and was accompanied by a roughly
50% decrease in Apo B levels and 25% decrease in Lp (a)
levels. There was a modest 15% reduction in serum triglyc­
erides with negligible effect on HDL‐cholesterol [22, 23].
 New Agents for Treatment of Dyslipidemia 213

L
LDL L P PCSK9 Ab
L
L
P
PCSK9
L P L P

LDL-R

HM

L Endosome L P

Lysosome

L L P

FIG 17.1  Role of PCSK9 in hepatic uptake of LDL particles. Low density lipoprotein (LDL) particles bind to LDL receptors (LDL‐R) on
the hepatocyte membrane (HM), and the receptor‐ligand complex is internalized to form an endosome. LDL‐R dissociates from the
complex and recycles back to the surface while the LDL particle undergoes lysosomal degradation. PCSK9 is a secreted protein
which also binds to the LDL‐R, and prevents dissociation of the LDL/LDL‐R complex leading to lysosomal degradation of both the
receptor and LDL particle. Also shown are PCSK9 antibodies (Ab) which can bind to circulating PCSK9 and thereby decrease LDL‐R
degradation.

No effect on highly sensitive C‐reactive protein was noted
when added on to statin therapy  [24]. As expected, the
lipid‐lowering effects are much less robust in patients with
homozygous familial hypercholesterolemia, and depend
on the genotype and number of LDL receptor‐null
alleles [25].
Very few consistent adverse effects beyond injection site
reactions have been observed in these trials. Specifically,
there has been no evidence of increased incidence of liver
function abnormalities, myopathy or new onset diabetes. A
substantial number of patients achieved very low LDL‐
cholesterol levels (< 25 mg/dL), but did not appear to expe­
rience higher rates of adverse effects, except for a slight
increase in the incidence of cataracts [26]. The initial con­
cerns about increased risk for neurocognitive adverse
events have been largely allayed by a dedicated neurocog­
nitive sub‐study involving nearly 2000 patients which did
not show any effect on cognitive functions [27].
PCSK9 inhibitors: Cardiovascular outcome
trails
Both evolocumab and alirocumab have been evaluated for
cardiovascular outcomes in large, randomized, double‐
blind, placebo‐controlled trials, namely FOURIER  [28]
and ODYSSEY Outcomes Trial [29], respectively.
The FOURIER trial was a secondary prevention trial
which enrolled 27 564 patients with established ASCVD
(myocardial infarction, non‐hemorrhagic stroke or periph­
eral vascular disease) who had LDL‐cholesterol ≥ 70 mg/
dL or non HDL‐cholesterol ≥ 100 mg/dL on stable, opti­
mized lipid‐lowering therapy (moderate – high intensity
statin ± ezetimibe). The baseline plasma LDL‐cholesterol
was 91 mg/dL and nearly 70% of the patients were on high
intensity statin therapy. The study was initially planned for
4 years, but ended after a mean follow‐up of 2.2 years as
there was a faster than expected accrual of events.
 214 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
Evolocumab therapy decreased LDL‐cholesterol by about
59%, and the mean LDL‐cholesterol in the treatment group
at the end of the study was about 30 mg/dL. There was a
15% relative risk reduction of the primary composite car­
diovascular end point, with an absolute risk reduction of
2%, which was projected to be 3.3% if the trial were to be
extended to 5 years. There was a significant reduction in
risk for myocardial infarction and stroke, but no difference
in overall mortality or cardiovascular deaths. Except for
injection‐site reactions, there was no difference in the inci­
dence of any adverse events.
The ODYSSEY Outcomes Trial was also a similar sec­
ondary prevention trial in 18 924 patients who had a recent
(1–12 months) Acute Coronary Syndrome, and had LDL‐
cholesterol ≥ 70 mg/dL, non HDL‐cholesterol ≥ 100 mg/dL
or Apo B ≥ 80 mg/dL on high intensity or maximally toler­
ated statin therapy. The baseline plasma LDL‐cholesterol
was 87 mg/dL and nearly 90% of the patients were on high
intensity statin therapy. The mean follow‐up duration was
2.8 years. Alirocumab dose was adjusted during the study
to target an LDL‐cholesterol concentration of 25–50  mg/
dL. There was a significant 15% relative risk reduction of
the primary composite cardiovascular end point, with an
absolute risk reduction of 1.6%. Reduction in all‐cause
mortality was also noted (HR = 0.85, 95% CI = 0.73–0.98),
but was not considered statistically significant as it was
tested after two non‐significant outcomes (CHD death and
CVD death) in the testing hierarchy. Similar to the evo­
locumab trial, there was no difference in the incidence of
adverse events except for local injection‐site reactions.
PCSK9 inhibitor use in patients with
diabetes
The use of PCSK9 inhibitors has been found to be safe and
efficacious in patients with diabetes. Sub‐group analysis of
multiple phase 3 trials of both evolocumab and alirocumab
have shown that lipid lowering in patients with diabetes is
similar to those without diabetes, and was not associated
with worsening of diabetes control or other adverse
effects [30]. Similarly, the reduction in LDL‐cholesterol did
not differ in patients with pre‐diabetes, impaired fasting
glucose or metabolic syndrome compared to normoglyce­
mic individuals  [31, 32]. Interestingly, the incidence of
local injection‐site adverse effects was noted to be less in
patients with diabetes compared to those without diabe­
tes [33, 34].
In addition, there have been a few phase 3 clinical stud­
ies of PCSK9 inhibitors exclusively in patients with diabe­
tes which have also demonstrated safety and
efficacy  [35–37]. The ODYSSEY DM‐INSULIN rand­
omized trial studied the effect of alirocumab in insulin‐
treated patients with both type 1 and type 2 diabetes, and
showed about 50% reduction in LDL‐cholesterol without
raising any concerns about co‐administration of insulin
and alirocumab [35]. The ODYSSEY DM‐DYSLIPIDEMIA
trial demonstrated significant reduction in non HDL‐cho­
lesterol in patients with type 2 diabetes and mixed hyper­
lipidemia who were on maximally tolerated statin therapy.
Compared to “usual care” which included the option to add
either fenofibrate, omega‐3 fatty acids or niacin, treatment
with alirocumab led to a mean 38% reduction in non‐HDL
cholesterol in this 24‐week study  [36]. Similarly evel­
ocumab has been shown to reduce LDL‐cholesterol by
about 65% and non HDL‐cholesterol by about 56%, besides
other atherogenic lipid markers in patients with type 2 dia­
betes and dyslipidemia on background statin therapy [37].
In all these studies, there was no change in measures of gly­
cemic control.
While there are no outcome trials exclusively in patients
with diabetes, a pre‐specified sub‐group analysis of the
FOURIER trial showed that evolocumab reduced cardio­
vascular risk equally in patients with and without diabetes,
without increasing the risk for new onset diabetes or dete­
rioration in glucose control  [38]. Of the 27 564 study
patients, 40% had diabetes at baseline, and the hazard
ratios for the primary end point were 0.83 (95% CI, 0.75–
0.93) for patients with diabetes and 0.87 (95% CI, 0.79–
0.96) for patients without diabetes. There was no change in
hemoglobin A1c or fasting plasma glucose levels between
the evolocumab and placebo groups over time in patients
with diabetes, prediabetes, or normoglycemia. These data
suggest that evelocumab use is safe and efficacious in
patients with ASCVD and diabetes.
PCSK9 inhibitors: Clinical use and future
perspectives
The available data clearly show the ability of this class of
lipid‐lowering drugs to safely and effectively reduce the

level of LDL‐cholesterol and other atherogenic lipoproteins
leading to reduced incidence of major vascular events.
However, this has to be tempered with other factors includ­
ing the current high cost of these drugs, and the lack of
more long‐term safety studies, before recommending a
broader use. The current guidelines recommend using
them in patients who are likely to see the greatest absolute
risk reduction. These would be patients with extremely
high cholesterol, such as those with familial hypercholes­
terolemia, or patients with established ASCVD and sub‐
optimally treated dyslipidemia who remain at very high
risk for future ASCVD events. The ODYSSEY ESCAPE
study in patients with heterozygous familial hypercholes­
terolemia showed that alirocumab therapy was effective in
substantially decreasing or even eliminating the need for
LDL apheresis [39]. Indeed the number of patients requir­
ing LDL apheresis for severe hypercholesterolemia has
been steadily declining since the advent of PCSK9 inhibitor
therapy.
In patients with diabetes, those with concomitant
ASCVD who have non HDL‐cholesterol ≥ 100 mg/dL on
maximally tolerated statin therapy should be considered
for PCSK9  inhibitor therapy. It is quite likely that other
patients with long duration of diabetes, and other risk fac­
tors, will also benefit from earlier initiation of such therapy,
even for primary prevention. It is well recognized that both
the magnitude of hyperlipidemia, and its duration are the
key determinants of atherogenesis. Initiation of effective,
safe lipid‐lowering therapy at a younger age has the poten­
tial to greatly decrease the incidence of ASCVD in the
population, but its cost effectiveness (under current pric­
ing), feasibility and long‐term safety needs to be deter­
mined. Many patients on PCSK9  inhibitors achieve
extremely low LDL‐cholesterol levels. In the FOURIER
trial, patients with baseline LDL‐cholesterol < 70  mg/dL
achieved a mean LDL‐cholesterol of 21 mg/dL which was
associated with a 30% relative risk reduction. Secondary
analysis of data from this trial also show a monotonic rela­
tionship between major ASCVD events and achieved LDL‐
cholesterol concentrations down to < 10  mg/dL without
any concomitant increase in adverse effects [40]. This has
prompted suggestions for more aggressive treatment
thresholds, even below 20 mg/dL [16]. However, the mean
follow‐up in the FOURIER trial was 2.2 years, and we
clearly need more long‐term studies to establish the safety
New Agents for Treatment of Dyslipidemia 215
of such aggressive measures. Similarly the effect of
PCSK9 inhibitors on new onset diabetes also needs to be
examined in longer term studies, especially since
Mendelian randomization studies do show an association
between PCSK9 variants which cause low cholesterol and
increased risk for diabetes [41]. If long‐term safety is estab­
lished, and cost reduction is achieved, PCSK9 inhibitors do
have the potential to significantly reduce the burden of
ASCVD in both patients with and without diabetes. Future
advances including the use of siRNA to decrease PCSK9
activity, immunization to develop antibodies against native
PCSK9, and CRISPR‐Cas9 gene editing of PCSK9 may fur­
ther facilitate this goal. While the siRNA, Inclisiran, has
already shown promising results in phase III trials with a
50% reduction in LDL‐cholesterol when administered
every 6 months [42], the other options are yet to be tested
in humans [43, 44].
Other novel lipid‐lowering therapies
Besides ezetimibe and PCSK9 inhibitors, bempedoic acid
can be used to decrease LDL‐cholesterol further in statin
treated (or intolerant) individuals. It inhibits the enzyme
ATP‐citrate lyase in the cholesterol biosynthetic pathway,
and has been shown to reduce LDL‐cholesterol by about
21% [45]. It was recently approved for clinical use by the
US Food and Drug Administration (FDA) for patients with
heterozygous familial hypercholesterolemia or those with
established ASCVD who need additional LDL‐cholesterol
reduction. It is administered orally at a dose of 180  mg
daily.
Recently, there has been growing interest in novel tri­
glyceride lowering therapies which has focused on increas­
ing lipoprotein lipase (LPL) mediated clearance of
triglyceride‐rich‐lipoproteins (TGRL). Figure 17.2 depicts
the metabolism of chylomicrons and very low‐density lipo­
proteins (VLDL), the two principal TGRLs. After initial
hydrolysis by LPL, the remnants of TGRL are either taken
up by the liver or further hydrolyzed by the hepatic lipase
to generate LDL particles. However, the remnant particles
can also permeate the vascular endothelium, akin to LDL
particles, and promote inflammation and atherogenesis by
a variety of mechanisms including abnormal endothelial
secretion of inflammatory cytokines and impaired flow‐
mediated dilatation  [46–48]. This assumes great signifi­
 216 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease

B-48
B-100

E TG A-V
E TG A-V

C
C
CE
DL
CE CH
VL YL
O

Apo C-II
Insulin

FFA HSL LPL

FFA
P
LR FFA
HL Apo C-III
ANGPTL3
ANGPTL4
LDL
TG
CE CE
TG
CE
Vas TG
c TG
infla ular CE
mm CE
atio
n

RLP: CMR. VLDLr, lDL

FIG 17.2  Metabolism of triglyceride‐rich lipoproteins (TGRL). The two principal triglyceride‐rich lipoproteins, Chylomicrons (Chylo)
and Very Low Density Lipoproteins (VLDL) are secreted by the intestine and liver respectively. They undergo hydrolysis by the
Lipoprotein lipase (LPL) predominantly expressed in adipose tissue and skeletal muscle releasing Free Fatty Acids (FFA) for these
tissues and multiple remnant lipoproteins (RLP) including Chylomicron Remnants (CR), VLDL remnants (VLDLr), and Intermediate
Density Lipoproteins (IDL). These are either cleared by the liver through LDL receptor related proteins (LRP), or undergo further
hydrolysis by the Hepatic Lipase (HL) leading to generation of Low Density Lipoprotein (LDL) particles. Similar to LDL, RLP can also
be taken up into the vessel wall and promote vascular inflammation and atherogenesis. Also shown in the Figure are positive and
negative influencers of LPL activity, and the role of insulin in suppressing adipose tissue Hormone Sensitive Lipase (HSL). Other
abbreviations: TG, triglyceride; CE, Cholesteryl ester; B‐100, Apo B‐100; B‐48, Apo B‐48; E, Apo E; A‐V, Apo A‐V; C, Apo C.

cance in patients with type 2 diabetes who often have
hypertriglyceridemia due to excess hepatic VLDL synthesis
and secretion, driven by enhanced delivery of free fatty
acids released by adipose tissue hormone sensitive lipase.
Adiposity, insulin resistance and insulin deficiency, which
are the hallmarks of type 2 diabetes, serve to amplify this
process. Excess TGRL generation also results in excess
remnant lipoprotein concentrations which are
primarily responsible for the increased risk for ASCVD
despite normal LDL cholesterol concentrations in patients
with dia­ betes and hypertriglyceridemia. The novel
triglyceride‐lowering medications work by decreasing the
activity of proteins that inhibit LPL such as Apo C3 and
ANGPTL 3/4. Volanesorsen is an antisense oligonucleotide

which inhibits Apo C3 and has been shown in Phase II tri­
als to reduce Apo C3 levels by 40–80% and serum triglyc­
erides by 31–71% in a dose dependent effect  [49]. It was
recently approved for clinical use in Europe for patients
with severe hypertriglyceridemia due to Familial
Chylomicronemia Syndrome, but was not approved for
clinical use by the US Food and Drug Administration due
to risk of thrombocytopenia. An n‐acetyl galactosamine
conjugated version of this drug has been developed and
shown to significantly reduce triglyceride levels in healthy
volunteers [50]. Both a monoclonal antibody (Evinacumab)
and an anti‐sense oligonucleotide to ANGPTL3 have also
been developed, and are awaiting clinical trials. While
these drugs are primarily being studied for their triglycer­
ide‐lowering effect, they do have the potential to reduce
ASCVD in patients with hypertriglyceridemia and elevated
postprandial remnant lipoprotein cholesterol levels, as sug­
gested by genetic studies linking Apo C3 and ANGPTL4
variations with serum triglycerides and ASCVD [51, 52].
Patients with diabetes and hypertriglyceridemia do have
elevated remnant lipoproteins, and future studies will help
understand the role of these therapies as an adjunct to LDL
cholesterol‐lowering therapies to further reduce cardiovas­
cular morbidity and mortality.
Conclusions
Patients with diabetes, both type 1 and type 2, are at high
risk for ASCVD. Besides healthy lifestyle changes and good
glucose control, addition of statin therapy has been shown
to reduce the risk, and should be initiated in most patients
with diabetes. To further reduce the residual risk after sta­
tin therapy, additional LDL‐cholesterol lowering therapies
such as ezetimibe, bempedoic acid, and PCSK9 inhibitors
can be considered, especially for patients with established
ASCVD. PCSK9  inhibitors have been shown to robustly
reduce LDL‐cholesterol by about 60% with minimal
adverse events. They have also been shown to reduce car­
diovascular outcomes in patients with established ASCVD,
both with and without diabetes. Novel triglyceride‐lower­
ing therapies targeting LPL‐mediated clearance of triglyc­
eride‐rich‐lipoproteins are being developed which have the
potential for additional ASCVD risk reduction and are also
deserving of further study.
New Agents for Treatment of Dyslipidemia 217
References
1. Hu FB, Stampfer MJ, Solomon CG, Liu S, Willett WC,
Speizer FE et al. The impact of diabetes mellitus on mortality
from all causes and coronary heart disease in women: 20
years of follow‐up. Arch Intern Med. 2001;161(14):
1717–1723.
2. Fox CS, Golden SH, Anderson C, Bray GA, Burke LE, de
Boer IH et al. Update on prevention of cardiovascular dis­
ease in adults with type 2 diabetes mellitus in light of recent
evidence: a scientific statement from the American Heart
Association and the American Diabetes Association.
Circulation. 2015;132(8):691–718.
3. Yeh RW, Sidney S, Chandra M, Sorel M, Selby JV, Go AS.
Population trends in the incidence and outcomes of acute
myocardial infarction. N Engl J Med. 2010;362(23):
2155–2165.
4. Rawshani A, Rawshani A, Franzen S, Eliasson B, Svensson
AM, Miftaraj M et al. Mortality and cardiovascular disease
in type 1 and type 2 diabetes. N Engl J Med. 2017;376(15):
1407–1418.
5. Cholesterol Treatment Trialists C, Kearney PM, Blackwell L,
Collins R, Keech A, Simes J et al. Efficacy of cholesterol‐low­
ering therapy in 18,686 people with diabetes in 14 ran­
domised trials of statins: a meta‐analysis. Lancet.
2008;371(9607):117–125.
6. Grundy SM, Stone NJ, Bailey AL, Beam C, Birtcher KK,
Blumenthal RS et  al. 2018 AHA/ACC/AACVPR/AAPA/
A B C / AC P M / A DA / AG S / A Ph A / ASP C / N L A / P C NA
Guideline on the management of blood cholesterol.
Circulation. 2018:CIR0000000000000625.
7. Cholesterol Treatment Trialists C, Fulcher J, O’Connell R,
Voysey M, Emberson J, Blackwell L et al. Efficacy and safety
of LDL‐lowering therapy among men and women: meta‐
analysis of individual data from 174,000 participants in 27
randomised trials. Lancet. 2015;385(9976):1397–1405.
8. Jacobs MJ, Kleisli T, Pio JR, Malik S, L’Italien GJ, Chen RS
et al. Prevalence and control of dyslipidemia among persons
with diabetes in the United States. Diabetes Res Clin Pract.
2005;70(3):263–269.
9. Hirano T. Pathophysiology of diabetic dyslipidemia. J
Atheroscler Thromb. 2018;25(9):771–782.
10. Bhatt DL, Steg PG, Miller M, Brinton EA, Jacobson TA,
Ketchum SB et al. Cardiovascular risk reduction with icosa­
pent ethyl for hypertriglyceridemia. N Engl J Med.
2019;380(1):11–22.
 218 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
11. Sacks FM, Carey VJ, Fruchart JC. Combination lipid therapy
in type 2 diabetes. N Engl J Med. 2010;363(7):692–694;
author reply 4–5.
12. Cannon CP, Blazing MA, Giugliano RP, McCagg A, White
JA, Theroux P et al. Ezetimibe added to statin therapy after
acute coronary syndromes. N Engl J Med. 2015;372(25):
2387–2397.
13. Abifadel M, Varret M, Rabes JP, Allard D, Ouguerram K,
Devillers M et  al. Mutations in PCSK9 cause autosomal
dominant hypercholesterolemia. Nat Genet. 2003;34(2):
154–156.
14. Horton JD, Cohen JC, Hobbs HH. PCSK9: a convertase that
coordinates LDL catabolism. J Lipid Res. 2009;50(Suppl):
S172–177.
15. Tavori H, Rashid S, Fazio S. On the function and homeosta­
sis of PCSK9: reciprocal interaction with LDLR and addi­
tional lipid effects. Atherosclerosis. 2015;238(2):264–270.
16. Sabatine MS. PCSK9 inhibitors: clinical evidence and imple­
mentation. Nat Rev Cardiol. 2019;16(3):155–165.
17. Maxwell KN, Breslow JL. Adenoviral‐mediated expression
of Pcsk9 in mice results in a low‐density lipoprotein receptor
knockout phenotype. Proc Natl Acad Sci USA. 2004;101(18):
7100–7105.
18. Rashid S, Curtis DE, Garuti R, Anderson NN, Bashmakov Y,
Ho YK et al. Decreased plasma cholesterol and hypersensi­
tivity to statins in mice lacking Pcsk9. Proc Natl Acad Sci
USA. 2005;102(15):5374–5379.
19. Cohen J, Pertsemlidis A, Kotowski IK, Graham R, Garcia
CK, Hobbs HH. Low LDL cholesterol in individuals of
African descent resulting from frequent nonsense mutations
in PCSK9. Nat Genet. 2005;37(2):161–165.
20. Cohen JC, Boerwinkle E, Mosley TH, Jr., Hobbs HH.
Sequence variations in PCSK9, low LDL, and protection
against coronary heart disease. N Engl J Med. 2006;354(12):
1264–1272.
21. Kathiresan S, Myocardial Infarction Genetics C. A PCSK9  aemic patients stratified by glycaemic status and metabolic
missense variant associated with a reduced risk of early‐
onset myocardial infarction. N Engl J Med. 2008;358(21):
2299–2300.
22. Sabatine MS, Giugliano RP, Wiviott SD, Raal FJ, Blom DJ,
Robinson J et al. Efficacy and safety of evolocumab in reducing
lipids and cardiovascular events. N Engl J Med. 2015;372(16):
1500–1509.
23. Robinson JG, Farnier M, Krempf M, Bergeron J, Luc G,
Averna M et al. Efficacy and safety of alirocumab in reduc­
ing lipids and cardiovascular events. N Engl J Med.
2015;372(16):1489–1499.
24. Bohula EA, Giugliano RP, Leiter LA, Verma S, Park JG, Sever
PS et al. Inflammatory and Cholesterol Risk in the FOURIER
Trial. Circulation. 2018;138(2):131–140.
25. Raal FJ, Hovingh GK, Blom D, Santos RD, Harada‐Shiba M,
Bruckert E et  al. Long‐term treatment with evolocumab
added to conventional drug therapy, with or without apher­
esis, in patients with homozygous familial hypercholesterol­
aemia: an interim subset analysis of the open‐label TAUSSIG
study. Lancet Diabetes Endocrinol. 2017;5(4):280–290.
26. Robinson JG, Rosenson RS, Farnier M, Chaudhari U, Sasiela
WJ, Merlet L et al. Safety of very low low‐density lipoprotein
cholesterol levels with alirocumab: pooled data from rand­
omized trials. J Am Coll Cardiol. 2017;69(5):471–482.
27. Giugliano RP, Mach F, Zavitz K, Kurtz C, Im K, Kanevsky E
et  al. Cognitive function in a randomized trial of evo­
locumab. N Engl J Med. 2017;377(7):633–643.
28. Sabatine MS, Giugliano RP, Keech AC, Honarpour N,
Wiviott SD, Murphy SA et al. Evolocumab and clinical out­
comes in patients with cardiovascular disease. N Engl J Med.
2017;376(18):1713–1722.
29. Schwartz GG, Steg PG, Szarek M, Bhatt DL, Bittner VA, Diaz R
et  al. Alirocumab and cardiovascular outcomes after
acute coronary syndrome. N Engl J Med. 2018;379(22):
2097–2107.
30. Handelsman Y, Lepor NE. PCSK9 Inhibitors in lipid man­
agement of patients with diabetes mellitus and high cardio­
vascular risk: a review. J Am Heart Assoc. 2018;7(13):e008953.
31. Leiter LA, Muller‐Wieland D, Baccara‐Dinet MT, Letierce
A, Samuel R, Cariou B. Efficacy and safety of alirocumab in
people with prediabetes vs those with normoglycaemia at
baseline: a pooled analysis of 10 phase III ODYSSEY clinical
trials. Diabet Med. 2018;35(1):121–130.
32. Blom DJ, Koren MJ, Roth E, Monsalvo ML, Djedjos CS,
Nelson P et al. Evaluation of the efficacy, safety and glycae­
mic effects of evolocumab (AMG 145) in hypercholesterol­
syndrome. Diabetes Obes Metab. 2017;19(1):98–107.
33. Sattar N, Preiss D, Robinson JG, Djedjos CS, Elliott M,
Somaratne R et al. Lipid‐lowering efficacy of the PCSK9 inhi­
bitor evolocumab (AMG 145) in patients with type 2 diabetes:
a  meta‐analysis of individual patient data. Lancet Diabetes
Endocrinol. 2016;4(5):403–410.
34. Leiter LA, Zamorano JL, Bujas‐Bobanovic M, Louie MJ,
Lecorps G, Cannon CP et  al. Lipid‐lowering efficacy and
safety of alirocumab in patients with or without diabetes: A
sub‐analysis of ODYSSEY COMBO II. Diabetes Obes Metab.
2017;19(7):989–996.

35. Leiter LA, Cariou B, Muller‐Wieland D, Colhoun HM, Del
Prato S, Tinahones FJ et al. Efficacy and safety of alirocumab
in insulin‐treated individuals with type 1 or type 2 diabetes
and high cardiovascular risk: The ODYSSEY DM‐INSULIN
randomized trial. Diabetes Obes Metab. 2017;19(12):
1781–1792.
36. Ray KK, Leiter LA, Muller‐Wieland D, Cariou B, Colhoun
HM, Henry RR et  al. Alirocumab vs usual lipid‐lowering
care as add‐on to statin therapy in individuals with type 2
diabetes and mixed dyslipidaemia: The ODYSSEY DM‐
DYSLIPIDEMIA randomized trial. Diabetes Obes Metab. 2018;
20(6):1479–1489.
37. Lorenzatti AJ, Eliaschewitz FG, Chen Y, Lu J, Baass A,
Monsalvo ML et  al. Randomised study of evolocumab in
patients with type 2 diabetes and dyslipidaemia on back­
ground statin: Primary results of the BERSON clinical trial.
Diabetes Obes Metab. 2019;21(6):1455–1463.
38. Sabatine MS, Leiter LA, Wiviott SD, Giugliano RP,
Deedwania P, De Ferrari GM et al. Cardiovascular safety and
efficacy of the PCSK9 inhibitor evolocumab in patients with
and without diabetes and the effect of evolocumab on gly­
caemia and risk of new‐onset diabetes: a prespecified analy­
sis of the FOURIER randomised controlled trial. Lancet
Diabetes Endocrinol. 2017;5(12):941–950.
39. Moriarty PM, Parhofer KG, Babirak SP, Cornier MA, Duell
PB, Hohenstein B et  al. Alirocumab in patients with hete­
rozygous familial hypercholesterolaemia undergoing lipo­
protein apheresis: the ODYSSEY ESCAPE trial. Eur Heart J.
2016;37(48):3588–3595.
40. Giugliano RP, Pedersen TR, Park JG, De Ferrari GM, Gaciong
ZA, Ceska R et al. Clinical efficacy and safety of achieving very
low LDL‐cholesterol concentrations with the PCSK9 inhibi­
tor evolocumab: a prespecified secondary analysis of the
FOURIER trial. Lancet. 2017;390(10106):1962–1971.
41. Ference BA, Robinson JG, Brook RD, Catapano AL,
Chapman MJ, Neff DR et  al. Variation in PCSK9 and
HMGCR and risk of cardiovascular disease and diabetes.
N Engl J Med. 2016;375(22):2144–2153.
42. Ray KK, Wright RS, Kallend D, Koenig W, Leiter LA, Raal FJ
et al. Two phase 3 trials of inclisiran in patients with elevated
LDL cholesterol. N Engl J Med. 2020;382(16):1507–1519.
New Agents for Treatment of Dyslipidemia 219
43. Landlinger C, Pouwer MG, Juno C, van der Hoorn JWA,
Pieterman EJ, Jukema JW et al. The AT04A vaccine against
proprotein convertase subtilisin/kexin type 9 reduces total
cholesterol, vascular inflammation, and atherosclerosis in
APOE*3Leiden.CETP mice. Eur Heart J. 2017;38(32):
2499–2507.
44. Ding Q, Strong A, Patel KM, Ng SL, Gosis BS,
Regan SN et al. Permanent alteration of PCSK9 with in vivo;
CRISPR‐Cas9 genome editing. Circ Res. 2014;115(5):
488–492.
45. Laufs U, Banach M, Mancini GBJ, Gaudet D, Bloedon LT,
Sterling LR et  al. Efficacy and safety of bempedoic acid in
patients with hypercholesterolemia and statin intolerance.
J Am Heart Assoc. 2019;8(7):e011662.
46. Doi H, Kugiyama K, Oka H, Sugiyama S, Ogata N, Koide SI
et al. Remnant lipoproteins induce proatherothrombogenic
molecules in endothelial cells through a redox‐sensitive
mechanism. Circulation. 2000;102(6):670–676.
47. Chait A, Ginsberg HN, Vaisar T, Heinecke JW, Goldberg IJ,
Bornfeldt KE. Remnants of the triglyceride‐rich lipoproteins,
diabetes, and cardiovascular disease. Diabetes. 2020;69(4):
508–516.
48. Zheng XY, Liu L. Remnant‐like lipoprotein particles impair
endothelial function: direct and indirect effects on nitric
oxide synthase. J Lipid Res. 2007;48(8):1673–1680.
49. Gaudet D, Alexander VJ, Baker BF, Brisson D, Tremblay K,
Singleton W et al. Antisense inhibition of apolipoprotein c‐
iii in patients with hypertriglyceridemia. N Engl J Med.
2015;373(5):438–447.
50. Alexander VJ, Xia S, Hurh E, Hughes SG, O’Dea L, Geary RS
et  al. N‐acetyl galactosamine‐conjugated antisense drug to
APOC3  mRNA, triglycerides and atherogenic lipoprotein
levels. Eur Heart J. 2019;40(33):2785–2796.
51. The TG and HDL Working Group of the Exome Sequencing
Project, National Heart, Lung, and Blood Institute, Crosby J,
Peloso GM, Auer PL et  al. Loss‐of‐function mutations in
APOC3, triglycerides, and coronary disease. N Engl J Med.
2014;371(1):22–31.
52. Dewey FE, Gromada J, Shuldiner AR. Variants in ANGPTL4
and the risk of coronary artery disease. N Engl J Med.
2016;375(23):2305–2306.
 18 The Role of Bariatric Surgery in Obese
Patients with Diabetes: Primary or Rescue
Therapy?
Praveena Gandikota1 and Blandine Laferrère2
1
Endocrine Fellow, Endocrine, Diabetes and Nutrition Division, Department of Medicine, St Luke’s Roosevelt Hospital,
New York, NY, USA
2
Assistant Professor of Medicine, Division of Endocrinology, Diabetes and Nutrition Obesity Research Center
Department of Medicine, St Luke’s Roosevelt Hospital Center, Columbia University College of Physicians and Surgeons,
New York, NY, USA

c­ onsiderable public health burden. Approximately half of

LE A R N I N G P OI NT S
●● Obesity and type 2 diabetes mellitus (T2DM) are major
public health issues that are closely related. Lifestyle
modifications (LSM) with diet and exercise leading to
weight loss are cornerstones of T2DM management
along with pharmacologic therapy. Limitations are cost,
effectiveness, and short-­term effect for LSM and
compliance, cost, and side effects for medications.
●● Newer medications may help preserve beta-­cell
function over time.
●● Bariatric surgery is increasingly becoming popular for
management of morbid obesity and its related
comorbidities including T2DM. Diabetes control
improves for most patients after bariatric surgery, with
full remission in over 50% of patients.
●● Although bariatric surgery is approved for diabetic
patients with BMI ≥ 35 kg/m2, the use of surgery to
treat diabetes in less obese individuals is currently
proposed. Limitations to this approach are lack of
high-­quality data on short and long term complications
of the surgery; and long-­term data on diabetes
remission and cost analysis.
Introduction
The prevalence of Type 2 diabetes mellitus (T2DM) and
obesity is increasing worldwide and represents a
patients with diabetes in the United States are obese.
Medical therapy for patients with diabetes combines costly
and often inefficient lifestyle modifications (LSM), with
various medications, some of them inducing unwanted
weight gain and/or hypoglycemia. However, in the last few
years, management and prognosis of T2DM have changed
considerably. Novel pharmacotherapies have emerged,
both for the management of diabetes as well as for the pre-
vention and treatment of its cardiometabolic complica-
tions. Glycemic remission of diabetes can be achieved after
bariatric surgery, although this is affected by many factors –
indeed, the likelihood of remission can be predicted on the
basis of several characteristics such as age, duration of dia-
betes, medications used to control diabetes etc. [1]. What is
increasingly clear is that bariatric surgery does not alter
beta-­cell mass but reverses glucose toxicity (in the short-­
term) and improves insulin action  – the response to oral
and intravenous secretagogues does not change  [2, 3]. In
view of the spectacular effect of bariatric surgery on T2DM,
it is understandable that bariatric surgery be proposed as
treatment of T2DM, independently of the beneficial weight
loss effect, i.e., in patients with body mass index (BMI)
< 35 kg/m2. This could be a particularly attractive alterna-
tive in some ethnic groups, like Asians, who have a high
risk of T2DM at lower BMI compared to Caucasians.

Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.

© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.

220
 The Role of Bariatric Surgery in Obese Patients with Diabetes: Primary or Rescue Therapy? 221

The objective of this chapter is to discuss whether bari- oral glucose compared to intravenous glucose. The incre-
atric surgery should be offered as first-­line therapy, or as tins play a key role in postprandial glucose-­mediated insu-
rescue therapy, when medical options have failed to control lin secretion. In addition, GLP-­ 1 decreases glucagon
diabetes. In addition, hypotheses on mechanisms of action induces weight loss and, in animal models, decreased
of bariatric surgery on glucose homeostasis, the role of apoptosis and increased beta-­cell mass. The incretin effect
bariatric surgery in treatment of patients with T2DM with is blunted in T2DM. Both incretin levels and effect increase
BMI < 35  kg/m2, risks and benefits of different types of after GBP [8, 9], but not after diet [10] or purely restrictive
bariatric surgery, and future techniques will be discussed. procedures  [11]. Levels of GLP-­ 1  increase as early as
2 days [12] and persist up to 20 years [13]. The mechanisms
of increased incretin levels after GBP are not fully under-
Possible Mechanisms by Which Bariatric
stood. The rapid delivery of nutrients to the lower intestine
Surgery Improves T2DM
after bypass surgeries or after vertical sleeve gastrectomy
Effect of Weight Loss (VSG) [14] may increase the production of GLP-­1 by direct
T2DM is a very complex disorder with insulin resistance in nutrient exposure of the L cells. This hypothesis has been
the muscle and the liver, impaired beta-­and alpha-­cell fueled by results of ileal transposition in rodents [15]. The
function, impaired incretin effect, and alteration of glucose foregut exclusion hypothesis emphasizes the role of exclu-
homeostasis in the kidney and the brain. A significant sion of the proximal part of the gut from nutrients  [16],
degree of beta-­cell failure and vascular complications are suggesting the presence of a would-­be anti-­incretin factor
often present at the time of diagnosis of T2DM. Reduced secreted by the proximal small intestine. This hypothesis
calorie intake and increased physical activity are essential was the basis for trials with endoluminal sleeves and duo-
for the management of T2DM. denal mucosal resurfacing (see below).
Multiple studies have shown the effectiveness of LSM in Other possible mechanisms of successful weight loss and
preventing T2DM, with a risk reduction up to 58%. metabolic improvement after bariatric surgery include
Similarly, weight loss and LSM in patients with established changes in ghrelin and PYY [11, 17], bile acids [18], changes
T2DM are beneficial [4]. A modest 10% weight loss is asso- in taste patterns, inflammatory markers, gastric emptying
ciated with significant improvement in diabetes and other and intestinal transit time [9], and possibly in gut flora [19].
obesity-­related comorbidities such as hypertension and
dyslipidemia. Contrary to LSM that result in weight loss of
Role of Bariatric Surgery in T2DM:
small magnitude and short duration [5] bariatric surgery
First-­line Therapy or Rescue Therapy
results in weight loss of greater magnitude, 40–50% excess
When Medical Options Fail?
weight loss (EWL), sustained over time up to 14 years [6,
7]. Weight loss is a key factor in the improvement of T2DM Pharmacological agents used to treat T2DM  [20] are not
after bariatric surgery. without risks and side effects. Weight gain, and/or the risk
of hypoglycemia, especially with sulfonylureas and insulin,
Effect of Incretins are a major hindrance in the management of the already
Resolution of T2DM after bypass procedures has been obese patients. Additionally, challenges occur in patients
described within days of the surgery, suggesting that fac- with associated conditions like renal failure or heart failure.
tors other than weight loss could be responsible for the dia- However, in spite of the various excellent pharmacological
betes improvement. The change in incretins may explain agents available in the United States, 43% of patients with
part of diabetes remission after GBP. The two main incre- T2DM fail medical therapy and do not achieve the HbA1C
tins are glucagon-­ like peptide 1 (GLP1) and glucose-­ target of < 7% [21]. In a study by Holman et al, only 31.9–
dependent insulinotropic peptide (GIP), produced by the L 44.7% of patients achieved an HbA1c < 6.5% at 3 years with
and K cells, respectively. Together, they are responsible for different insulin regimen [22]. None of the medical clinical
the “incretin effect” or the greater insulin response after trials report diabetes remission.
 222 Surgery for Weight Management
Nearly 30% of patients who undergo bariatric surgery
have T2DM [6], although about a third of these patients are
undiagnosed [23]. In the Buchwald meta-­analysis [24], cri-
teria used to define diabetes resolution was fasting blood
glucose < 100 mg/dl and/or HbA1c < 6%, off diabetes med-
ications. The percentage of diabetes remission varies
according to surgery: 95% after BPD/duodenal switch, 80%
after RYGBPP, 80% with gastroplasty and 57–73% after
gastric banding (GB). The greater the malabsorption
(BPD) and weight loss, the more likely the patient will
achieve diabetes remission. Significant improvement in
T2DM has also been shown in the Swedish Obesity Study
(SOS), a long-­term observational prospective study that
compared surgical versus nonsurgical groups [7]. The sur-
gical group fared better at both 2 years’ and 10 years’ fol-
low-­up with respect to decrease in glucose and insulin
levels, incidence as well as recovery from T2DM. In a ran-
domized controlled trial (RCT), Dixon et al. showed that
GB resulted in a 73% diabetes remission rate at two years
compared to conventional medical interventions
(13%)  [25]. The benefit of bariatric surgery include not
only a ∼40% weight loss, but the resolution and/or signifi-
cant improvement of diabetes and most obesity-­related
comorbidities like hyperlipidemia (70%), hypertension
(61.7%), and of obstructive sleep apnea (85.7%)  [24].
Recent studies have shown a decreased mortality [26], par-
ticularly the mortality related to diabetes, after bariatric
surgery.
Would patients failing medical treatment, who tend to
be older, with longer duration of diabetes, with less optimal
access to care, respond to bariatric surgery? Which bariat-
ric surgery is more likely to help patients achieve their
blood glucose target and/or go into remission? Should
obese patients with BMI ≥ 35 kg/m2 with T2DM be offered
bariatric surgery as a first-­line therapy? It is difficult to
apply the current data to individual patients in terms of
balancing risks versus benefits. Available surgical studies
on diabetes remission do not always report data on diabe-
tes control and/or duration [27]. Diabetes duration, when
provided, is short  [25, 28] and few patients, 0.5%  [25] to
39%  [29], are insulin-­ treated. The study by Schauer
et al. [28] is the only one that provides outcome data ana-
lyzed according to preoperative diabetes status and dura-
tion. Patients with less severe disease had significantly
better fasting plasma glucose and HbA1c post surgery
compared with insulin-­requiring patients with more severe
disease. Additionally, the longer the duration of diabetes,
the more likely patients were to remain on medications
and/or insulin. However, the preoperative duration of
T2DM did not result in significant differences in postop-
erative glucose control among the groups  [29]. Another
issue in interpreting published data is the definition used
for diabetes remission. As diabetes remission is indeed a
novel concept [30], no uniform definition was used in pre-
vious surgical studies [24, 25]. The recent consensus state-
ment proposed to define complete remission of T2DM as
“normal glycemic measures (HbA1c in the normal range,
fasting glucose < 100 mg/dl) for at least 1 year duration in
the absence of active pharmacologic therapy or ongoing
procedures” [30]. Whether this clinical definition of diabe-
tes remission is accompanied by reversal of the pathophysi-
ological defects seen with diabetes remains to be
demonstrated.
According to the current guidelines, bariatric surgery is
approved for patients with BMI ≥ 40  kg/m2 or BMI
≥ 35 kg/m2 in the presence of any comorbidity. Less than
3% of the patients who qualify for bariatric surgery
undergo this treatment of their obesity and over 200,000
surgeries are performed yearly in the United States. As
morbidly obese patients gain significant health and quality
of life benefits from bariatric surgery, an effort to increase
the number of surgeries on patients with higher BMI (>
45 kg/m2) is a reasonable approach [31]. According to the
latest ADA 2010 guidelines [20], bariatric surgery should
be considered for adults with BMI ≥ 35 kg/m2 and T2DM,
especially if diabetes or other comorbidities are difficult to
control with lifestyle and pharmacologic therapy. What
defines “difficult to control” and for how long a patient has
to be failing medical therapy prior to being offered bariat-
ric surgery is unclear. Identifying preoperative predictors
of diabetes resolution is critical not only for patient selec-
tion but also for determining which type of surgery would
be the best. The amount of weight loss is clearly a major
determinant. This is particularly true after GB when dia-
betes remission is directly proportional to the degree of
weight loss [25]. The duration of diabetes and/or preop-
erative insulin use decreases the chances of diabetes remis-
sion [28, 29]. This implies that the less the d
­ eterioration of
 The Role of Bariatric Surgery in Obese Patients with Diabetes: Primary or Rescue Therapy?
beta-­cell function at the time of surgery, the higher the
chances of diabetes remission. However, even if patients
do not go into remission after bariatric surgery, most of
them will experience significant improvement of their dis-
ease and achieve glucose control on less medications [28].
There are no data from RCT comparing aggressive medi-
cal treatment with bariatric surgery in morbidly obese
persons. Even with the best medical scenario, however,
results of bariatric surgery will be likely hard to match, at
least in the first 5–10 years after surgery.
Gastric Banding or Gastric Bypass?
There are three main categories of bariatric surgery.
Restrictive procedures, with or without gastrectomy, aim
to reduce gastric volume in order to limit food intake and
induce weight loss. These procedures include laparo-
scopic adjustable gastric banding (GB), vertical banded
gastroplasty (VBG), rarely performed now, and vertical
sleeve gastrectomy (VSG). The Roux-­en-­Y gastric bypass
(RYGBP) surgery combines gastric restriction with some
malabsorption and is the most commonly performed
procedure in the United States. Finally, biliopancreatic
diversion (BPD) is a procedure with significant malab-
sorption with (Figure 18.1) or without gastric restriction.
Although surgeons empirically tend to perform malab-
sorptive procedures in patients with BMI > 50  kg/m2,
rather than GB, there are no data and/or guidelines deter-
mining the best surgery for a particular patient. The
choice of the type of surgery should not be based on sur-
geons’ preference and/or on early 6–12  months weight
loss outcome, but rather on short-­and long-­term data
generated by RCT comparing various surgical proce-
dures. Collecting long-­ term data is essential to learn
about the effect of bariatric surgery on weight loss, reso-
lution of comorbidities, as well as complications such as
nutritional deficiencies. Data about nutritional deficien-
cies, weight regain, comorbidities resolution, and overall
morbidity are essentially poorly known, as loss to follow-
­up is notably high for this patient population. In a recent
review [32] only five studies were identified with data on
nutritional deficiencies with at least 12 month follow-­up
and the data were suboptimal.
Available data suggests reasonably comparable out-
comes of VSG and RYGB at 1 year [33] and slight, but sig-
223
nificant, benefits for RYGB at 3 years, at least in terms of
glycemic control [34].
Morbid obesity is associated with an increased mortal-
ity. The mortality rate associated with bariatric surgery has
been reported to be 0.1–1.1% [24]. In a recent prospective,
multicenter observational study [31], the 30-­day death rate
was 0% after laparoscopic GB, 0.2% after laparoscopic
RYGBP and 2.1% after open RYGBP. These rates are com-
parable to the 90-­day mortality rate of 1% seen after chol-
ecystectomy. Both prospective (e.g., SOS)  [26] and
retrospective studies  [35] have shown decreased overall
mortality after bariatric surgery. Specifically, cause-­specific
mortality in the RYGBP surgery group decreased by 92%
for diabetes [35]. Bariatric surgery is associated with early
postoperative complications such as thromboembolism,
wound complications like infection/dehiscence, pulmo-
nary complications like atelectasis, and late postoperative
complications like anastomotic stricture, intestinal
obstruction, incisional hernias, and gastrogastric fistula. In
addition, dumping syndrome can be disabling. Long-­term
metabolic and nutritional complications related to altered
micronutrient and vitamin absorption are frequently
observed after malabsorptive surgeries [32] and need to be
aggressively treated, often for a lifetime. Vitamin D defi-
ciency may require weekly doses of ergocalciferol 50,000
units weekly or a few times a week and iron and B12 often
are needed parenterally. Deficiencies need to be carefully
monitored at least twice yearly, for a lifetime. At present,
there are no established guidelines for mineral and vitamin
supplementation after bariatric surgery.
Hyperinsulinemic hypoglycemia associated is a rela-
tively uncommon complication of RYGB and has variable
severity. Dietary management with or without glucosidase
inhibitors, octreotide, and other pharmacological treat-
ment should always be tried first prior to performing par-
tial pancreatectomy [36].
Bariatric surgery is the treatment of choice for morbid
obesity, with the additional benefit of diabetes improve-
ment and/or remission in over 80% of cases. As all morbid
obese individuals will greatly benefit from bariatric sur-
gery, long duration, and/or poor control of diabetes should
not be contraindications to the surgery. Long-­term, high-­
quality data (> 10 years) on weight loss, safety, and nutri-
tional complications are essential. These long-­term studies
 224 Surgery for Weight Management

(b)
(a) Bypassed
segment

Adjustable
band

Port

(c) New (d)

stomach
pouch
(Gastric sleeve)

Part of
stomach Stomach
removed removed

Bypassed
segment

FIG 18.1  Bariatric surgeries. (A) Laparoscopic adjustable gastric banding (GB); (B) Roux-­en-­Y gastric bypass (RYGBP);
(C) biliopancreatic diversion (BPD) with duodenal switch; (D) vertical sleeve gastrectomy (VSG). www.mayoclinic.org/
bariatric-­surgery/bariatric-­procedures.html. Figure used by permission of Mayo Foundation for Medical Education
and Research. All rights reserved.

will help establish predictors of success, delineate the ric care will grow exponentially. Primary care physicians
choice of surgery, and help implement guidelines for vita- and/or endocrinologists need to be trained appropriately to
min supplementation. As the number of bariatric surgeries optimize patient management both preoperatively and
increases, the number of patients in need of lifetime bariat- postoperatively.
 The Role of Bariatric Surgery in Obese Patients with Diabetes: Primary or Rescue Therapy?
Role of Weight Loss Surgery in the Treatment
of Patients with T2DM and BMI < 35 kg/m2
There are currently few studies of bariatric surgery on
patients with T2DM and BMI < 35 kg/m2. In the RCT by
Dixon et al. [25], patients with well-­controlled T2DM, of
< 2 years’ duration, mean BMI of 37.1  kg/m2 went into
diabetes remission in 73% of cases after GB at 2 years (the
dropout rate was low) versus 13% in the diet group [25].
The studies from De Paula et al. with complicated surger-
ies, including ileal transposition, report high remission
rate, but a non-­negligible mortality, at least in the initial
report  [37]. A study reports remission of diabetes and
improvement of insulin sensitivity 18 months after BPD
in five patients with BMI < 35 kg/m2 [38]. A recent non-
randomized study by Shah et al. in Asian Indians reported
100% remission rate in 15 patients 3  months after GBP
surgery. Preoperatively, 80% of the patients were insulin
treated, diabetes control was poor with HbA1C ∼10%,
and diabetes duration was 9 years  [39]. The RCT by
Dixon et al. provides the best evidence-­based data to sug-
gest the use of GB in patients with diabetes and lower
BMI [25]. The diabetes remission was directly related to
the amount of weight loss and the preoperative HbA1C
levels. Participants had fairly mild diabetes. Whether
these results can be achieved in a more severe diabetic
population, different ethnic groups, and with a different
health care system, is unknown. There is currently not
enough evidence to suggest that bariatric surgery, par-
ticularly bypass, should be indicated for patients with
BMI < 35 kg/m2. As for more invasive surgeries (bypass ±
ileal transposition), it is unclear whether the risk they
represent is worth their use in clinical practice outside of
very well controlled research RCT.
Indeed, the diabetes surgery study randomized 120
adult participants, with BMI 30.0-­39.9 kg/m2 and HbA1c
>/=8.0% to either RYGB + intensive medical management
of diabetes, hypertension, and dyslipidemia or to medical
management alone. None of the subjects managed medi-
cally experienced diabetes remission at 36  months while
17% of gastric bypass patients had full remission and 19%
had partial remission. However, RYGB was associated with
more significant adverse events. In addition, the effect on
glycemic control decreased with time [40].
225
New Techniques that can be an Option
in the Future
Endoluminal duodenal sleeve either in lean diabetic
rats [17] or in rats with diet-­induced obesity [41] improves
glucose metabolism and induces weight loss. The human
trials of endoluminal treatments of obesity are limited to
restrictive interventions such as intragastric balloons, tran-
soral gastroplasty, and endoluminal vertical gastroplasty.
Trials with duodenojejunal bypass sleeves have been disap-
pointing both in terms of their efficacy and also because of
associated complications such as bolus obstruction and
hepatic abscess associated with their use [42].
Duodenal mucosal resurfacing employs thermal energy
to ablate the duodenal mucosa – theoretically to alter hor-
monal and mucosal signaling. A multicenter study with 46
participants reported effects on glycemic control despite
modest weight changes. These effects were sustained at
12  months. Further mechanistic studies are required to
better understand the mechanism(s) of action and long-­
term benefit, if any, on glycemic control [43].
In conclusion, bariatric surgery should be the treatment
of choice for patients with morbid obesity complicated by
T2DM and with high metabolic and cardiovascular risks.
The surgery has been proven to not only improve quality of
life but also prolong life in these high-­risk patients. There is
not enough evidence to perform invasive bariatric surger-
ies and/or to experiment with new surgical procedures in
patients with lower BMI and/or with milder diabetes, out-
side of well-­controlled RCT. Optimizing chronic care after
GB, a noninvasive procedure, to ensure proper weight loss
and diabetes remission is important. Multidisciplinary
teams including surgeons, endocrinologists, gastroenter-
ologists, psychiatrists, nutritionists, and primary care are
needed to follow up these patients. Bariatric surgery offers
a unique model to understand the complexity of T2DM
and help develop new treatments for this chronic disease.
References
  1. Still CD, Wood GC, Benotti P et al. Preoperative prediction
of type 2 diabetes remission after Roux-­en-­Y gastric bypass
surgery: a retrospective cohort study. Lancet 2(1): 38–45,
2014.
 226 Surgery for Weight Management
  2. Nguyen KT, Billington CJ, Vella A et  al. Preserved insulin
secretory capacity and weight loss are the predominant pre-
dictors of glycemic control in patients with type 2 diabetes
randomized to Roux-­en-­Y gastric bypass. Diabetes 64(9):
3104–3110, 2015.
 3. Dirksen C, Bojsen-­ Møller KN, Jørgensen NB et  al.
Exaggerated release and preserved insulinotropic action of
glucagon-­like peptide-­1 underlie insulin hypersecretion in
glucose-­tolerant individuals after Roux-­en-­Y gastric bypass.
Diabetologia 56(12): 2679–2687, 2013.
  4. Pi-­Sunyer X, Blackburn G, Brancati FL et  al. Reduction in
weight and cardiovascular disease risk factors in individuals
with type 2 diabetes: one-­year results of the look AHEAD
trial. Diabetes Care. 30(6):1374–1383, 2007.
  5. Wadden TA, Sternberg JA, Letizia KA et  al. Treatment of
obesity by very low calorie diet, behavior therapy, and their
combination: a five-­year perspective. Int J Obes. 13(Suppl
2):39–46, 1989.
  6. Pories WJ, Swanson MS, MacDonald KG et al. Who would
have thought it? An operation proves to be the most effective
therapy for adult-­ onset diabetes mellitus. Ann Surg.
222(3):339–350, 1995.
  7. Sjostrom L, Lindroos AK, Peltonen M et al. Lifestyle, diabe-
tes, and cardiovascular risk factors 10 years after bariatric
surgery. N Engl J Med. 351(26):2683–2693, 2004.
  8.. Laferrère B, Heshka S, Wang K et al. Incretin levels and effect
are markedly enhanced 1  month after Roux-­en-­Y gastric
bypass surgery in obese patients with type 2 diabetes.
Diabetes Care. 30(7):1709–1716, 2007.
  9. Morinigo R, Moize V, Musri M et al. Glucagon-­like peptidel,
peptide YY, hunger, and satiety after gastric bypass surgery
in morbidly obese subjects. J Clin Endocrinol Metab.
91(5):1735–1740, 2006.
10. Laferrère B, Teixeira J, McGinty J et al. Effect of weight loss
by gastric bypass surgery versus hypocaloric diet on glucose
and incretin levels in patients with type 2 diabetes. J Clin
Endocrinol Metab. 93(7):2479–2485, 2008.
11. Bose M, Machineni S, Olivan B et al. Superior appetite hor-
mone profile after equivalent weight loss by gastric bypass
compared to gastric banding. Obesity (Silver Spring).
18(6):1085–1091, 2010.
12. le Roux CW, Welbourn R, Werling M et al. Gut hormones as
mediators of appetite and weight loss after Roux-­en-­Y gas-
tric bypass. Ann Surg. 246(5):780–785, 2007.
13. Naslund E, Gryback P, Hellstrom PM et al. Gastrointestinal
hormones and gastric emptying 20 years after jejunoileal
bypass for massive obesity. Int J Obes Relat Metab Disord.
21(5):387–392, 1997.
14. Vidal J, Ibarzabal A, Nicolau J et  al. Short-­term effects of
sleeve gastrectomy on type 2 diabetes mellitus in severely
obese subjects. Obes Surg. 17(8):1069–1074, 2007.
15. Strader AD, Vahl TP, Jandacek RJ et al. Weight loss through
ileal transposition is accompanied by increased ileal hor-
mone secretion and synthesis in rats. Am J Physiol Endocrinol
Metab. 288(2):E447–E453, 2005.
16. Rubino F, Forgione A, Cummings DE et al. The mechanism
of diabetes control after gastrointestinal bypass surgery
reveals a role of the proximal small intestine in the patho-
physiology of type 2 diabetes. Ann Surg. 244(5):741–749,
2006.
17. Olivan B, Teixeira J, Bose M et al. Effect of weight loss by diet
or gastric bypass surgery on peptide YY3-­36  levels. Ann
Surg. 249(6):948–953, 2009.
18. Patti ME, Houten SM, Bianco AC et al. Serum bile acids are
higher in humans with prior gastric bypass: potential contri-
bution to improved glucose and lipid metabolism. Obesity
(Silver Spring). 17(9):1671–1677, 2009.
19. Zhang H, DiBaise J, Zuccolo A et al. Human gut microbiota
in obesity and after gastric bypass. PNAS. 106(7):2364–2370,
2009.
20. American Diabetes Association. Diabetes Care. 33(Suppl 1):
S1–S2, 2010.
21. Cheung BM, Ong KL, Cherny SS et al. Diabetes prevalence
and therapeutic target achievement in the United States,
1999 to 2006. Am J Med. 122(5):443–453, 2009.
22. Holman RR, Farmer AJ, Davies MJ et al. Three-­year efficacy
of complex insulin regimens in type 2 diabetes. NEngl J Med.
361(18):1736–1747, 2009.
23. Residori L, Garcia-­Lorda P, Flancbaum L et al. Prevalence of
co-­morbidities in obese patients before bariatric surgery:
effect of race. Obes Surg. 13(3):333–340, 2003.
24. Buchwald H, Avidor Y, Braunwald E et al. Bariatric surgery:
a systematic review and meta-­analysis. JAMA. 292(14):1724–
1737, 2004.
25. Dixon JB, O’Brien PE, Playfair J et  al. Adjustable gastric
banding and conventional therapy for type 2 diabetes: a ran-
domized controlled trial. JAMA. 299(3):316–323, 2008.
26. Sjostrom L, Narbro K, Sjostrom CD et al. Effects of bariatric
surgery on mortality in Swedish obese subjects. NEngl J Med.
357(8):741–752, 2007.
27. Vetter ML, Cardillo S, Rickels MR, and Iqbal N. Narrative
review: effect of bariatric surgery on type 2 diabetes mellitus.
Ann Intern Med. 150(2):94–103, 2009.
28. Schauer PR, Burguera B, Ikramuddin S et al. Effect of lapa-
roscopic Roux-­en Y gastric bypass on type 2 diabetes melli-
tus. Ann Surg. 238(4):467–784, 2003.
 The Role of Bariatric Surgery in Obese Patients with Diabetes: Primary or Rescue Therapy?
29. Sugerman HJ, Wolfe LG, Sica DA, and Clore JN. Diabetes
and hypertension in severe obesity and effects of gastric
bypass-­induced weight loss. Ann Surg. 237(6):751–756,
2003.
30. Buse JB, Caprio S, Cefalu WT et al. How do we define cure
of diabetes? Diabetes Care. 32(11):2133–2135, 2009.
31. Purnell JQ and Flum DR. Bariatric surgery and diabetes:
who should be offered the option of remission? JAMA.
301(15):1593–1595, 2009.
32. Shah M, Simha V, and Garg A. Review: long-­term impact of
bariatric surgery on body weight, comorbidities, and nutri-
tional status. J Clin Endocrinol Metab. 91(11):4223–4231,
2006.
33. Schauer PR, Kashyap SR, Wolski K et  al. Bariatric surgery
versus intensive medical therapy in obese patients with dia-
betes. N Engl J Med 366(17): 1567–1576, 2012.
34. Schauer PR, Bhatt BL, Kirwan JP et al. Bariatric surgery ver-
sus intensive medical therapy for diabetes  – 3-­year out-
comes. N Engl J Med 370(21): 2002–2013, 2014.
35. Adams TD, Gress RE, Smith SC et al. Long-­term mortality
after gastric bypass surgery. N Engl J Med. 357(8):753–761,
2007.
36. Salehi, M, Vella A, McLaughlin T et al. Hypoglycemia after
gastric bypass surgery: current concepts and controversies. J
Clin Endocrinol Metab 103(8): 2815–2826, 2018.
37. Depaula AL, Macedo AL, Mota BR, and Schraibman V.
Laparoscopic ileal interposition associated to a diverted
227
sleeve gastrectomy is an effective operation for the treatment
of type 2 diabetes mellitus patients with BMI 21-­29. Surg
Endosc. 23(6):1313–1320, 2009.
38. Chiellini C, Rubino F, Castagneto M et al. The effect of bilio-­
pancreatic diversion on type 2 diabetes in patients with BMI
<35 kg/m2. Diabetologia. 52(6):1027–1030, 2009.
39. Shah SS, Todkar JS, Shah PS, and Cummings DE. Diabetes
remission and reduced cardiovascular risk after gastric
bypass in Asian Indians with body mass index <35 kg/m2.
Surg Obes Relat Dis. 6(4):332–338, 2010.
40. Ikramuddin S, Korner J, Lee W-­J et al. Durability of addition
of Roux-­en-­Y gastric bypass to lifestyle intervention and
medical management in achieving primary treatment goals
for uncontrolled type 2 diabetes in mild-­to-­moderate obe-
sity: a randomized control trial. Diabetes Care 39(9): 1510–
1518, 2016.
41. Aguirre V, Stylopoulos N, Grinbaum R, and Kaplan LM. An
endoluminal sleeve induces substantial weight loss and nor-
malizes glucose homeostasis in rats with diet-­induced obe-
sity. Obesity (Silver Spring). 16(12):2585–2592, 2008.
42. Egan AM and Vella A. Endoscopic treatments for obesity:
the good, the bad, and the ugly. Endocrinol Metab Clin North
Am 49(2): 315–328, 2020.
43. van Baar ACG et al. Endoscopic duodenal mucosal resurfac-
ing for the treatment of type 2 diabetes mellitus: one year
results from the first international, open-­label, prospective,
multicentre study. Gut 69(2): 295–303, 2020.
 19 Treatment Strategies in Patients
with Diabetes Mellitus and Ischemic Heart
Disease
Madeline K. Mahowald1, Chiam Leker Locker2, Mandeep Singh3 and
Robert L. Frye4
1
Fellow, Cardiovascular Disease, Mayo Clinic
2
Assistant Professor, Cardiovascular Surgery, Mayo Clinic
3
Professor, Cardiovascular Disease, Mayo Clinic
4
Professor, Cardiovascular Disease, Mayo Clinic

Originally considered to be a disease of disordered lipid

LE A R N I N G P OI NT S
●● Medical therapy is an important part of the revasculari-
zation strategy in any patient with diabetes and
ischemic heart disease.
●● Revascularization therapy may be indicated for
symptom control as well as improving prognosis and
preventing subsequent myocardial infarction.
●● In the absence of contraindications, coronary artery
bypass grafting seems to be a superior choice to
percutaneous therapy in people with diabetes.
Introduction
The synergistic relationship between the altered metabo­
lism and underlying inflammatory state in patients with
diabetes mellitus (DM) accounts for the high coincidence
of cardiovascular disease in this population [1, 2], which
remains the leading cause of death among patients with
DM in the United States [3]. Understanding the appropri­
ate management of both diseases is imperative to improv­
ing clinical outcomes, especially as the global prevalence of
diabetes and related conditions rises [4, 5].
metabolism and storage, the role of inflammation in coro­
nary artery disease (CAD) has become clearer in recent dec­
ades [6, 7]. Intracellular hyperglycemia has direct effects on
the vasculature by decreasing endothelial‐dependent vaso­
dilation [8] and also indirect effects via triggering high levels
of reactive oxygen species. Oxidative stress, in turn, stimu­
lates changes in lipid metabolism, altered mitochondrial
function, and the increased production of inflammatory
compounds [9]. These reactions affect platelet function,
fibrinolysis, and coagulation cascades. Once present, athero­
sclerosis further impairs endothelial function.
As the prevalence of type 2 DM has grown, the contem­
porary patient with coronary artery disease (CAD) and
DM has other inflammatory conditions, especially obesity,
that contribute to the hostile metabolic milieu. On a cellu­
lar level, adipose cells express inflammatory compounds,
such as TNF‐alpha, that affect insulin resistance and cellu­
lar uptake of glucose [10]. Specific therapies targeting
inflammation’s role in atherosclerosis and acute coronary
syndromes (ACS) have shown mixed results and are the
focus of much current research [6, 7, 11, 12].
The pattern of atherosclerosis that develops in this met­
abolic environment often differs from that in patients with­

Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.

© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.

228
 Revascularization Strategies in Patients with Diabetes Mellitus and Ischemic Heart Disease
out DM. The volume and extent of atheroma is greater in
patients with DM, and vascular remodeling is impaired,
such that diseased vessels are diffusely and extensively
involved, revealing narrow lumens on coronary angiogra­
phy [13]. This pattern of involvement has significant impli­
cations for revascularization, as patients with DM may not
have optimal distal vessel targets for bypass graft anasto­
mosis and may not have clear distal “landing zones” for
stents. Additionally, endothelial dysfunction leads to a
diminished ability to form collateral circulation [14]. The
Bypass Angioplasty Revascularization Investigation
(BARI) trial showed that patients with DM had a larger
amount of jeopardized myocardium during acute MI than
patients without DM [15]. Perhaps because of these differ­
ent disease patterns, patients with DM have worse out­
comes after myocardial infarction (MI) and in stable CAD
with or without revascularization [16–19].
To guide clinicians on the various treatment options
available for diabetic patients with CAD, this chapter will
focus on:
1 The importance of aggressive medical management
regardless of revascularization strategy selected
2 The indications for coronary revascularization
3 Factors influencing the selection of PCI or CABG if
revascularization is pursued.
Role of medical therapy and glucose
control
Optimal medical therapy (OMT) is imperative to clinical
outcomes for all patients with CAD, perhaps in patients
with DM more than those without. This additional benefit
may be due to the way in which medications target specific
pathways affected by metabolic derangements and oxida­
tive stress, such as glycoprotein IIb/IIIa inhibitors, which
reduce the impact of hyperglycemia on promoting platelet
aggregation [20].
Briefly, medical therapy is generally considered to
include antiplatelet agents, statins, beta‐blockers, and angi­
otensin‐converting enzyme inhibitors or angiotensin
receptor blockers. OMT is under‐prescribed among
patients with DM and CAD [21], despite being indepen­
dently associated with improved survival and bestowing a
greater treatment benefit than the method of revasculariza­
229
tion [22]. The precise classes of agents included in OMT
may change in the near future with the widespread use of
PCSK9  inhibitors, SGLT2  inhibitors, and glucagon‐like
peptide‐1 agonists, all of which show promise in patients
with DM and cardiovascular disease [23, 24].
Antiplatelet therapy
The role, components, and duration of antiplatelet therapy
after revascularization are evolving and situation‐ and
patient‐specific [25, 26] due to research that quantifies the
benefits of aspirin on preventing cardiovascular death, MI,
stroke, or TIA at the cost of more major bleeding [27, 28].
At present, the use of daily low‐dose aspirin for primary
prevention of cardiovascular disease in patients with dia­
betes is not routinely recommended [29].
For patients with known CAD, indefinite low‐dose daily
aspirin is recommended regardless of DM status unless con­
traindicated. The addition of a P2Y12 inhibitor for 12 months
following an ACS is recommended, but the risks of bleeding
generally outweigh the benefits if dual antiplatelet therapy
(DAPT) is extended beyond that time period.
Recommendations after elective revascularization in
stable ischemic heart disease are more nuanced and are
generally dictated by procedural and patient details,
including susceptibility to bleeding as well as ischemia.
After PCI, the duration of DAPT can be as brief as 1 month
in patients with high bleeding risk [30]. There is no dedi­
cated randomized trial to guide the duration of DAPT after
elective CABG in the absence of recent ACS or PCI, but it
may be considered for 12  months following surgery to
improve graft patency [31]. Procedural techniques are
important considerations, and a stronger argument for
DAPT may be made after off‐pump CABG [32].
Management of hyperglycemia
After patients with DM are diagnosed with obstructive
CAD, abundant evidence points to improved outcomes in
patients with tight glycemic control. In patients admitted
with ACS or for elective surgical revascularization, the
degrees of hyperglycemia and glycated hemoglobin eleva­
tion at the time of presentation are predictive of long‐term
mortality [33, 34]. Intensive glucose control after acute
myocardial infarction (MI) can mitigate the poorer out­
comes for patients with DM.
 230 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
Several studies have investigated the effects of different
strategies for treatment of hyperglycemia in the setting of
myocardial revascularization. The BARI 2D trial found
that therapy enhancing insulin sensitization was preferable
over therapy based on insulin provision for patients with
type 2 DM [35]. A subgroup analysis of the FREEDOM
trial also showed worse outcomes in patients who were
treated with insulin than those who were treated with other
hypoglycemic medications [36]. Even with second‐genera­
tion drug‐eluting stents (DES), patients treated with insu­
lin have more target lesion failure [37]. However, in many
of these trials (with the notable exception of BARI 2D),
patients were not randomized to insulin provision versus
insulin sensitization; patients on insulin therapy may
therefore represent more advanced diabetics with a pro­
pensity for worse outcomes. These trials will be discussed
in more detail later in this chapter.
Coronary revascularization
Guidelines from The American College of Cardiology/
American Heart Association (ACC/AHA) in 2014 [38]
(and expanded with appropriate use criteria in 2017 [39])
are essential references. The 2014 document includes the
following Class 1 recommendations to improve survival in
patients with multivessel CAD and DM: (1) a Heart Team
approach should be utilized in decision‐making regarding
revascularization in patients with DM and CAD, and (2)
CABG is generally recommended in preference to PCI in
those for whom revascularization is likely to improve sur­
vival (namely, those with the most extensive disease), par­
ticularly if an internal thoracic artery (ITA) can be
anastomosed to the left anterior descending (LAD) artery
and the patient is a good surgical candidate. However, PCI
has a role in patients with focal disease who warrant revas­
cularization for symptom relief. Additionally, the
European Society of Cardiology (ESC) released overlap­
ping guidelines in 2014 and, in conjunction with the
European Association for the Study of Diabetes, recently
collaborated to release task force guidelines specific to car­
diovascular disease in patients with diabetes and pre‐dia­
betes. These documents provide guidance on making the
diagnosis of DM, risk factor modification, medical man­
agement, and revascularization [40]. Specific ESC and
TABLE 19.1  Class I indications for myocardial revascularization
in stable ischemic heart disease per ACC/AHA and ESC
guidelines.
To improve symptoms
• For patients with ≥ 1 significant coronary artery stenoses
amenable to revascularization and unacceptable angina
despite medical therapy
To improve survival
• For survivors of sudden cardiac death with presumed
ischemia‐mediated ventricular tachycardia caused by
significant stenosis in a major coronary artery (ACC/AHA only)
• For patients with significant (≥ 50% lumen diameter) left
main coronary artery stenosis
• For patients with significant stenoses in 3 major coronary
arteries (with or without involvement of the proximal LAD
artery) or in the proximal LAD artery plus 1 other major
coronary artery
• For patients with multivessel disease and impaired LV
function, a large area of ischemia, or a single remaining
patent coronary artery with a significant stenosis (ESC only)
Unique recommendations for diabetic patients
• A Heart Team approach to revascularization is
recommended in patients with diabetes mellitus and
complex multi­vessel CAD
• CABG is generally recommended in preference to PCI to
improve survival in patients with diabetes mellitus and
multivessel CAD for which revascularization is likely to
improve survival (3‐vessel CAD or complex 2‐vessel CAD
involving the proximal LAD), particularly if an ITA graft can be
anastomosed to the LAD artery and provided the patient is a
surgical candidate
ACC/AHA indications for revascularization are summa­
rized in Table 19.1 [38, 41].
The specific method of revascularization must be tai­
lored to the individual patient and situation. In some
patients, the severity, extent, and precise anatomic location
of obstructive atherosclerotic lesions will dictate the prefer­
able technique. In other cases, relative equipoise allows for
consideration of additional factors, including the presence
of symptoms, documented ischemia, myocardial dysfunc­
tion, as well as the patient’s clinical characteristics, comor­
bidities, and preferences.
Here, we will discuss the three distinct situations in which
revascularization is indicated and the evidence relevant to
each one. Regardless of the presence of DM, the two basic
 Revascularization Strategies in Patients with Diabetes Mellitus and Ischemic Heart Disease 231

Algorithm to Guide Diagnosis and Management

Clinical evaluation of
patient with DM and
suspected CAD

Acute coronary
syndrome or severe Mild or no ischemic
symptoms of ischemia symptoms

Risk stratification with

Invasive coronary stress test (exercise
angiography preferred), high-sensitivity
Troponin-T, NT-proBNP

Left main disease Large area of ischemia

Focal disease
Severe multivessel disease or significantly Low-risk stress test,
without No obstructive CAD
LV dysfunction elevated biomarkers normal biomarkers
high risk features
SYNTAX score >22

PCI Consider CT coronary

CABG with arterial Aggressive risk Consider CT
or angiogram for young
grafts factor reduction coronary calcium
CABG patients

FIG 19.1  Algorithm to guide diagnosis and management.

reasons to consider elective invasive revascularization are
(1) control of symptoms from myocardial ischemia despite
optimal medical therapy, and (2) to reduce mortality and
myocardial infarction in high‐risk patients. Many patients
may benefit for both reasons. The third situation, ACS, is
discussed separately. Figure 19.1 provides an algorithm for
managing patients with DM and CAD.
Symptom control for stable patients despite
medical therapy
Randomized trials have demonstrated benefits in symptoms
and in quality of life with elective revascularization. The
increasing use of chest computed tomography has increased
lead time for the diagnosis of CAD, but determining whether
nonspecific symptoms are indeed related to anatomic find­
ings can be difficult. This challenge is common in patients
with DM, who often present with atypical symptoms, such as
dyspnea. Functional or physiologic testing, such as determi­
nation of fractional flow reserve, instantaneous wave‐free
ratio, or intravascular ultrasound, should be routinely
employed to assess lesions of borderline significance.
Documentation of ischemia makes symptom improvement
more likely after intervention.
PCI and CABG both have roles in symptom control.
PCI is often the first choice for patients with DM and sin­
gle‐vessel disease if the lesion is appropriate for stenting.
CABG appears to have a more durable effect, but major
surgery is often delayed until there is multivessel involve­
ment. For patients with left main or three‐vessel disease,
CABG has remained the standard of care [22].
Prolonging life and preventing subsequent
MI in stable patients
Since the early days of CABG, trials comparing surgical
revascularization with OMT alone demonstrated that the
greatest benefit was derived by patients with the most severe
disease and those with decreased left ventricular systolic
function. Several landmark trials are reviewed below.
 232 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
Trials for patients with diabetes and cardiovascular
disease
BARI 2D was designed to test treatment strategies in stable,
mildly symptomatic or asymptomatic patients with type 2
DM and angiographically confirmed CAD to determine if
prompt, elective revascularization would reduce mortality
or subsequent MI compared to OMT. A second layer of
randomization compared therapy with insulin‐sensitizing
agents to an insulin‐provision strategy to achieve target
glycemic control. Randomization was stratified according
to optimal revascularization strategy determined by each
patient’s treating physicians. Neither treatment strategy
demonstrated differences in mortality, MI, or stroke.
However, among the patients with the most extensive ath­
erosclerosis, there was a significant reduction in the com­
bined end point of death, MI, and stroke in patients that
underwent prompt CABG. This trial is notable among
studies of revascularization for demonstrating the preven­
tion of subsequent nonfatal MI (whereas many trials rely
on composite endpoints including repeat revasculariza­
tion). Updated BARI 2D analysis from 2019 reported that
the combination of insulin‐sensitization therapy and
CABG offered the lowest risk of cardiovascular death,
ACS, stroke, transient ischemic attack, repeat coronary
revascularization, and lower extremity revascularization or
amputation [42]. In addition to lower levels of hyperglyce­
mia, the patients in the sensitization group had lower levels
of the inflammatory markers C‐reactive protein and
fibrinogen.
Since BARI 2D, three additional trials have looked at
survival in patients with DM and multivessel disease rand­
omized to CABG or PCI that warrant mention. By far, the
largest was the Future Revascularization Evaluation in
Patients with Diabetes Mellitus: Optimal Management of
Multivessel Disease trial (FREEDOM); the two smaller
studies are the Coronary Artery Revascularization in
Diabetics trial (CARDia) and the Randomized Comparison
of Coronary Artery Bypass Surgery and Everolimus‐
Eluting Stent Implantation in the Treatment of Patients
with Multivessel Coronary Artery Disease (BEST) trial.
The FREEDOM trial included 1900 patients with DM
and multivessel disease at 140  international centers who
were randomized to CABG vs PCI and followed for a
median of 3.8 years. The primary outcome was a composite
of death from any cause, nonfatal MI, and nonfatal stroke.
The CABG arm was superior and significantly reduced
rates of death and MI, although at a higher risk of stroke.
Of note, the benefits of CABG in this study were driven by
reductions in all‐cause mortality and MI and extended to
all patients regardless of categorization by disease com­
plexity [43]. In analyses of patients treated with insulin
compared with oral agents, CABG was superior to PCI in
both groups [36]. Subsequent publication of the FREEDOM
Follow‐On study confirmed that the mortality benefits of
CABG over PCI with DES persisted over a median of 7.5
years [44].
CARDia included 510 symptomatic patients with a
median follow‐up of 365  days. At the trial’s conclusion,
there was no significant difference in the rate of composite
outcome (death, MI, stroke) but a significant higher rate of
repeat revascularization in the PCI arm [45]. The more
recent BEST trial was stopped early due to slow enrollment,
but was on track to support these results [17].
Common findings among these trials that may help
explain the difference in outcomes is the achievement of
more complete revascularization in the surgical patients
compared with those undergoing PCI; more grafts per
patient were utilized than stents per patient. Interestingly,
the benefit of CABG over PCI has not extended to left main
disease [18, 46, 47].
Subgroup analysis of patients with diabetes
Like the BARI trial, the Synergy between Percutaneous
Coronary Intervention with Taxus and Cardiac Surgery
(SYNTAX) trial evaluated intervention for stable ischemic
heart disease. Patients were randomized to PCI or CABG;
the primary end point was a composite of death from any
cause, stroke, MI, or repeat revascularization. Publication
of results at 12  months after randomization showed that
CABG was superior to PCI; these results were primarily
driven by repeat revascularization occurring in patients
with the highest SYNTAX scores (representing more
extensive and complex disease) who underwent PCI [48].
At study conclusion, there was no significant difference
between groups except for the subgroup of patients with
treated DM, who benefited from CABG [49].
Most recently, results of the International Study of
Comparative Health Effectiveness with Medical and
 Revascularization Strategies in Patients with Diabetes Mellitus and Ischemic Heart Disease
Invasive Approaches (ISCHEMIA) trial showed no benefit
of elective revascularization over OMT on all‐cause or car­
diovascular mortality in a sample of 5179 patients over a
median of 3.2 years [50]. ISCHEMIA trial data for the sub­
group of patients with DM have not been released at the
time of publication, but similar earlier trials have not con­
sistently shown benefit [19].
One must proceed cautiously in drawing conclusions
from subgroup analyses. However, one large meta‐analysis
pooling individual patient‐level data from 11 trials pub­
lished from 2001 to 2016 demonstrated a survival benefit
with CABG in patients with diabetes and multivessel dis­
ease [46]. PCI has not demonstrated survival benefit over
medical therapy.
Biomarkers for risk stratification
Identifying patients with low symptom burden but high‐
risk anatomy remains a clinical challenge in daily practice.
Measurement of biomarkers has been proposed as a
method of risk stratification in patients at high risk for
CAD but whose coronary anatomy is unknown. For exam­
ple, routine measurement of high‐sensitivity troponin‐T
levels was recorded among the 2285 patients with DM2
and stable ischemic heart disease who enrolled in the BARI
2D trial [51]. Among the 39% of patients who had abnor­
mal troponin‐T concentrations at baseline, the hazard ratio
for 5‐year event rate of cardiovascular death, MI, or stroke
was 1.85 compared to those with normal baseline tro­
ponin‐T. However, immediate revascularization in this
high risk population showed no benefit compared to OMT.
Elevated levels of plasma N‐terminal pro‐brain natriuretic
peptide (NT‐proBNP) have also been associated with car­
diovascular mortality [52], but how to incorporate this into
clinical practice is less clear.
Acute coronary syndromes
ACS include ST‐elevation MI (STEMI), non‐ST‐elevation
MI (NSTEMI), and unstable angina (UA). Emergent PCI is
the recommended revascularization strategy for all patients
with STEMI as there is clear benefit in restoring flow as
soon as possible. Importantly, patients with DM often have
delayed presentations for ACS compared with patients
without DM [53]. In‐hospital mortality has improved dra­
matically in recent decades for patients with DM and
233
STEMI. However, women with DM have not seen the same
mortality benefit as men and remain undertreated with the
cornerstones of medical therapy in ACS [54].
Revascularization within 24 hours is recommended for
patients with NSTEMI and UA and reduces death and MI
in all comers. In fact, according to at least one trial, the
benefits of early invasive therapy in patients with UA were
more pronounced in patients with DM than in those with­
out [55]. Emergent PCI is often the preferred strategy for
patients with NSTEMI complicated by hemodynamic
instability, malignant arrhythmia, or refractory ischemic
chest pain.
Coronary artery bypass grafting
For some clinicians, the FREEDOM trial sufficiently con­
cluded debate between revascularization strategies among
patients with DM as CABG demonstrated clear benefits in
subsequent cardiac events and survival regardless of dis­
ease extent and complexity. Additional studies from the
surgical literature suggest that details of the operative strat­
egy may have significant consequences. Perhaps most
importantly, the use of multiple arterial grafts may confer
improved long‐term survival, lower risk of adverse cardiac
events, and a higher rate of patency without a difference in
perioperative mortality compared to the conventional
CABG strategy with the use of the left internal thoracic
artery (ITA) to the LAD and vein grafts to the non‐LAD
vessels [56–58]. It has been shown that graft patency was
not inferior in diabetics over 20 years follow‐up and the
patency rate of BITA remained greater than 90%, outper­
forming the SVG patency rate of 40%, and those rates were
similar both in diabetics and non‐diabetics [59, 60].
In a large clinical trial randomizing patients to a surgical
strategy of single ITA or bilateral internal thoracic artery
(BITA) grafting, there were non‐significant decreases in
death and in a composite outcome of death, MI or stroke at
10 years favoring the BITA group. However, results of this
study were confounded by a high crossover rate and a
higher than expected use of the radial artery. In the “as
treated cohort,” multiarterial grafting showed significant
advantage compared with single ITA group [61]. Although
the use of BITA has been controversial in patients with DM
due to historically higher rates of deep sternal wound
infections, this technique has demonstrated better survival,
 234 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
less frequent need for revascularization, and longer free­
dom from major adverse cardiovascular events in patients
with and without diabetes [62, 63].
The increased incidence of deep sternal wound infec­
tion seen in patients with DM (especially women) can be
minimized – if not altogether eliminated – by harvesting
ITA in a skeletonized rather than pedicled manner. The
importance of the skeletonized technique over pedicle
grafts is explicitly stated in clinical practice guidelines for
arterial graft harvest [64]. For patients with DM, skeletoni­
zation of BITA lowers the risk of sternal infection to that of
pedicled single graft regardless of treatment with insulin
[63, 65–67].
Percutaneous coronary intervention
DES are currently the standard of care for both diabetic
and non‐diabetic patients; metal struts are typically coated
with anti‐proliferative mTOR inhibitors. Although there is
interest in stents comprised of absorbable polymers, out­
comes have not demonstrated significant benefits and their
use is not widespread. Despite advances in stent technol­
ogy and anti‐thrombotic medications, target lesion failure
remains more common in patients with DM due to in‐
stent restenosis and stent thrombosis [68, 69]. Subsequent
generations of DES seek to limit the restenosis caused by
intimal hyperplasia, which is particularly exuberant in
patients with diabetes [70], and it remains an area of active
research.
Why is coronary artery bypass grafting more
effective?
The benefits of surgical over percutaneous revasculariza­
tion are likely multifactorial and secondary to both struc­
tural and biochemical effects on the subtended myocardium
[71]. Proposed mechanisms include the following:
1 Reperfusion of the grafted coronary axis with additional
protection against the progression of native vessel dis­
ease [72, 73];
2 Restoration of endothelial function of the grafted and
downstream vessels due to the production of nitric oxide
of arterial segments [74]. This improvement in regional
endothelial function is in contrast to the changes
induced by stenting, which induced both acute and
chronic inflammatory changes [75]; and
3 A newly perfused distal bed receptive to the vascular
growth factors secreted by ITAs and SVGs that promote
formation of collateral circulation [76].
Conclusions
Despite the ongoing challenges for patients with DM and
CAD, evidence suggests that current medical and revascu­
larization strategies have significantly improved outcomes
[77]. Patients with non‐severe CAD and mild symptoms
may be safely managed conservatively for many years.
However, regardless of the presence of symptoms, patients
with DM and severe coronary artery disease can derive sig­
nificant benefit regarding symptoms, quality of life, mor­
tality, and freedom from revascularization by undergoing
early CABG. Patients with less extensive disease but
severe symptoms may benefit from PCI, but there is no
clear evidence that these patients have improved survival
or freedom from MI or revascularization. Aggressive
medical management and risk factor modification is
imperative to improving the outcomes of patients with
DM and any degree of CAD, and insulin sensitization
therapy may confer additional benefits as one compo­
nent of OMT.
References
1. Libby P, Theroux P. Pathophysiology of coronary artery
disease. Circulation. 2005;111(25):3481–3488.
2. Reardon CA et al. Obesity and insulin resistance promote
atherosclerosis through an IFNγ‐regulated macrophage
protein network. Cell Reports. 2018;23(10):3021–3030.
3. Rodriguez F et al. Diabetes‐attributable mortality in the
United States from 2003 to 2016 using a multiple‐cause‐of‐
death approach. Diabetes Research and Clinical Practice.
2019;148:169–178.
4. Bhupathiraju SN, Hu FB. Epidemiology of obesity and
diabetes and their cardiovascular complications. Circulation
Research. 2016;118(11):1723–1735.
5. Cho N et al. IDF Diabetes Atlas: Global estimates of diabetes
prevalence for 2017 and projections for 2045. Diabetes
Research and Clinical Practice. 2018;138:271–281.
6. Tardif J‐C. et al. Efficacy and safety of low‐dose colchicine
after myocardial infarction. New Engl J Med.
2019;381(26):2497–2505.
 Revascularization Strategies in Patients with Diabetes Mellitus and Ischemic Heart Disease
7. Ridker PM et al. Rosuvastatin to prevent vascular events in
men and women with elevated C‐reactive protein. New Engl
J Med. 2008;359(21):2195–2207.
8. Beckman JA et al. Inhibition of protein kinase Cβ prevents
impaired endothelium‐dependent vasodilation caused by
hyperglycemia in humans. Circulation Research.
2002;90(1):107–111.
9. Shah MS, Brownlee M. Molecular and cellular mechanisms
of cardiovascular disorders in diabetes. Circulation Research.
2016;118(11):1808–1829.
10. Hotamisligil GS, Shargill NS, B.M. Spiegelman, Adipose
expression of tumor necrosis factor‐alpha: direct role in obe­
sity‐linked insulin resistance. Science. 1993;259(5091):87–91.
11. Ridker PM et  al. Antiinflammatory therapy with canaki­
numab for atherosclerotic disease. N Engl J Med. 2017;377(12):
1119–1131.
12. Ridker PM et al. Low‐dose methotrexate for the prevention
of atherosclerotic events. N Engl J Med. 2019;380(8):
752–762.
13. Nicholls SJ et al. Effect of diabetes on progression of coro­
nary atherosclerosis and arterial remodeling: a pooled anal­
ysis of 5  intravascular ultrasound trials. Journal of the
American College of Cardiology. 2008;52(4):255–262.
14. Abaci A et al. Effect of diabetes mellitus on formation of coro­
nary collateral vessels. Circulation. 1999;99(17):2239–2242.
15. Kip KE et al. Differential influence of diabetes mellitus on
increased jeopardized myocardium after initial angioplasty
or bypass surgery: bypass angioplasty revascularization
investigation. Circulation. 2002;105(16):1914–1920.
16. Bangalore S et  al. Everolimus‐eluting stents or bypass sur­
gery for multivessel coronary disease. N Engl J Med. 2015;
372(13):1213–1222.
17. Park S‐J. et  al. Trial of everolimus‐eluting stents or bypass
surgery for coronary disease. N Engl J Med. 2015;372(13):
1204–1212.
18. Milojevic M et  al. Bypass surgery or stenting for left main
coronary artery disease in patients with diabetes. Journal of
the American College of Cardiology. 2019;73(13):1616–1628.
19. Maron DJ et al. Impact of metabolic syndrome and diabetes
on prognosis and outcomes with early percutaneous coro­
nary intervention in the COURAGE (Clinical Outcomes
Utilizing Revascularization and Aggressive Drug Evaluation)
trial. Journal of the American College of Cardiology. 2011;58(2):
131–137.
20. Roffi M et al. Platelet glycoprotein IIb/IIIa inhibitors reduce
mortality in diabetic patients with non‐ST‐segment‐eleva­
tion acute coronary syndromes. Circulation. 2001;104(23):
2767–2771.
235
21. Farkouh ME et  al. Risk factor control for coronary artery
disease secondary prevention in large randomized trials.
Journal of the American College of Cardiology. 2013;61(15):
1607–1615.
22. Iqbal J et al. Optimal medical therapy improves clinical out­
comes in patients undergoing revascularization with percu­
taneous coronary intervention or coronary artery bypass
grafting: insights from the Synergy Between Percutaneous
Coronary Intervention with TAXUS and Cardiac Surgery
(SYNTAX) trial at the 5‐year follow‐up. Circulation.
2015;131(14):1269–1277.
23. Husain M et  al. Oral semaglutide and cardiovascular out­
comes in patients with type 2 diabetes. N Engl J Med.
2019;381(9):841–851.
24. Perkovic V et al. Canagliflozin and renal outcomes in type 2
diabetes and nephropathy. N Engl J Med. 2019;380(24):
2295–2306.
25. Valgimigli M et al. 2017 ESC focused update on dual anti­
platelet therapy in coronary artery disease developed in col­
laboration with EACTS. European Journal of Cardio‐Thoracic
Surgery. 2018;53(1):34–78.
26. Levine GN et al. 2016 ACC/AHA guideline focused update
on duration of dual antiplatelet therapy in patients with cor­
onary artery disease: a report of the American College of
Cardiology/American Heart Association Task Force on
Clinical Practice Guidelines: an update of the 2011 ACCF/
AHA/SCAI guideline for percutaneous coronary interven­
tion, 2011 ACCF/AHA guideline for coronary artery bypass
graft surgery, 2012 ACC/AHA/ACP/AATS/PCNA/SCAI/
STS guideline for the diagnosis and management of patients
with stable ischemic heart disease, 2013 ACCF/AHA guide­
line for the management of ST‐elevation myocardial infarc­
tion, 2014 AHA/ACC guideline for the management of
patients with non–ST‐elevation acute coronary syndromes,
and 2014 ACC/AHA guideline on perioperative cardiovas­
cular evaluation and management of patients undergoing
noncardiac surgery. Circulation. 2016;134(10):e123–e155.
27. Steg PG et al. Ticagrelor in patients with stable coronary dis­
ease and diabetes. N Engl J Med. 2019;381(14):1309–1320.
28. The ASCEND Study Collaborative Group. Effects of aspirin
for primary prevention in persons with diabetes mellitus. N
Engl J Med. 2018;379(16):1529–1539.
29. Arnett DK et al. 2019 ACC/AHA guideline on the primary
prevention of cardiovascular disease: a report of the
American College of Cardiology/American Heart Association
Task Force on Clinical Practice Guidelines. Journal of
the  American College of Cardiology. 2019;74(10):
e177–e232.
 236 Diagnosis and Management of Cardiovascular Risk Factors and Cardiovascular Disease
30. Windecker S et al. Polymer‐based or polymer‐free stents in
patients at high bleeding risk. N Engl J Med. 2020;382:
1208–1218.
31. van Diepen S et al. Dual antiplatelet therapy versus aspirin
monotherapy in diabetics with multivessel disease undergo­
ing CABG: FREEDOM insights. Journal of the American
College of Cardiology. 2017;69(2):119–127.
32. Deo SV et al. Dual anti‐platelet therapy after coronary artery
bypass grafting: is there any benefit? A systematic review
and meta‐analysis. Journal of Cardiac Surgery. 2013;28(2):
109–116.
33. Halkos ME et  al. Elevated preoperative hemoglobin A1c
level is predictive of adverse events after coronary artery
bypass surgery. The Journal of Thoracic and Cardiovascular
Surgery. 2008;136(3):631–640.
34. Robich MP et al. Intensity of glycemic control affects long‐
term survival after coronary artery bypass graft surgery. The
Annals of Thoracic Surgery. 2019;107(2):477–484.
35. Group BDS. A randomized trial of therapies for type 2 dia­
betes and coronary artery disease. N Engl J Med. 2009;360(24):
2503–2515.
36. Dangas GD et al. Long‐term outcome of PCI versus CABG
in insulin and non–insulin‐treated diabetic patients: results
from the FREEDOM trial. Journal of the American College of
Cardiology. 2014;64(12):1189–1197.
37. Pepe M et  al. Impact of insulin‐treated and noninsulin‐
treated diabetes mellitus in all‐comer patients undergoing
percutaneous coronary interventions with polymer‐free
biolimus‐eluting stent (from the RUDI‐FREE Registry). The
American Journal of Cardiology. 2019;124(10):1518–1527.
38. Fihn SD et  al. 2014 ACC/AHA/AATS/PCNA/SCAI/STS
focused update of the guideline for the diagnosis and man­
agement of patients with stable ischemic heart disease: a
report of the American College of Cardiology/American
Heart Association Task Force on Practice Guidelines, and
the American Association for Thoracic Surgery, Preventive
Cardiovascular Nurses Association, Society for
Cardiovascular Angiography and Interventions, and Society
of Thoracic Surgeons. Journal of the American College of
Cardiology. 2014;64(18):1929–1949.
39. Patel MR et al. ACC/AATS/AHA/ASE/ASNC/SCAI/SCCT/
STS 2017 appropriate use criteria for coronary revasculari­
zation in patients with stable ischemic heart disease: a report
of the American College of Cardiology appropriate use crite­
ria task force, American Association for Thoracic Surgery,
American Heart Association, American Society of
Echocardiography, American Society of Nuclear Cardiology,
Society for Cardiovascular Angiography and Interventions,
Society of Cardiovascular Computed Tomography, and
Society of Thoracic Surgeons. Journal of the American
College of Cardiology. 2017;69(17):2212–2241.
40. Cosentino F et al. 2019 ESC guidelines on diabetes, pre‐dia­
betes, and cardiovascular diseases developed in collabora­
tion with the EASD: the task force for diabetes, pre‐diabetes,
and cardiovascular diseases of the European Society of car­
diology (ESC) and the European association for the study of
diabetes (EASD). European Heart Journal. 2020;41(2):
255–323.
41. Windecker S et  al. 2014 ESC/EACTS guidelines on myo­
cardial revascularization. EuroIntervention. 2015;10(9):
1024–1094.
42. Genuth SM et al. BARI 2D: a reanalysis focusing on cardio­
vascular events. Mayo Clinic Proceedings. 2019; Elsevier.
43. Farkouh ME et  al. Strategies for multivessel revasculariza­
tion in patients with diabetes. N Engl J Med. 2012;367(25):
2375–2384.
44. Farkouh ME et al. Long‐term survival following multivessel
revascularization in patients with diabetes: the FREEDOM fol­
low‐on study. Journal of the American College of Cardiology.
2019;73(6):629–638.
45. Kapur A et  al. Randomized comparison of percutaneous
coronary intervention with coronary artery bypass grafting
in diabetic patients: 1‐year results of the CARDia (Coronary
Artery Revascularization in Diabetes) trial. Journal of the
American College of Cardiology. 2010;55(5):432–440.
46. Head SJ et al. Mortality after coronary artery bypass grafting
versus percutaneous coronary intervention with stenting for
coronary artery disease: a pooled analysis of individual
patient data. Lancet. 2018;391(10124):939–948.
47. Giacoppo D et  al. Percutaneous coronary intervention vs
coronary artery bypass grafting in patients with left main
coronary artery stenosis: a systematic review and meta‐anal­
ysis. JAMA Cardiology. 2017;2(10):1079–1088.
48. Serruys PW et al. Percutaneous coronary intervention ver­
sus coronary‐artery bypass grafting for severe coronary
artery disease. N Engl J Med. 2009;360(10):961–972.
49. Kappetein AP et al. Treatment of complex coronary artery
disease in patients with diabetes: 5‐year results comparing
outcomes of bypass surgery and percutaneous coronary
intervention in the SYNTAX trial. European Journal of
Cardio‐Thoracic Surgery. 2013;43(5):1006–1013.
50. Maron DJ et al. Initial invasive or conservative strategy for
stable coronary disease. N Engl J Med. 2020;382:1395–1407.
51. Everett BM et  al. Troponin and cardiac events in stable
ischemic heart disease and diabetes. N Engl J Med.
2015;373(7):610–620.
 Revascularization Strategies in Patients with Diabetes Mellitus and Ischemic Heart Disease
52. Bhalla MA et  al. Prognostic role of B‐type natriuretic pep­
tide levels in patients with type 2 diabetes mellitus. Journal of
the American College of Cardiology. 2004;44(5):1047–1052.
53. The BARI Investigators. Influence of diabetes on 5‐year
mortality and mobidity in a randomized trial comparing
CABG and PTCA in patients with multivessel disease. The
Bypass Angioplasty Revascularization Investigation (BARI).
Circulation. 1997;96:1761–1769.
54. Roffi M et  al. Gender‐related mortality trends among dia­
betic patients with ST‐segment elevation myocardial infarc­
tion: insights from a nationwide registry 1997–2010.
European Heart Journal: Acute Cardiovascular Care. 2013;2(4):
342–349.
55. Invasive compared with non‐invasive treatment in unstable
coronary‐artery disease: FRISC II prospective randomised
multicentre study. FRagmin and Fast Revascularisation dur­
ing InStability in Coronary artery disease Investigators.
Lancet. 1999;354:708–715.
56. Locker C et al. Multiple arterial grafts improve late survival
of patients undergoing coronary artery bypass graft surgery:
analysis of 8622 patients with multivessel disease. Circulation.
2012;126(9):1023–1030.
57. Schwann TA et al. Incremental value of increasing number
of arterial grafts: the effect of diabetes mellitus. The Annals
of Thoracic Surgery. 2018;105(6):1737–1744.
58. Gaudino M et al. Radial‐artery or saphenous‐vein grafts in
coronary‐artery bypass surgery. N Engl J Med. 2018;378(22):
2069–2077.
59. Raza S et  al. Influence of diabetes on long‐term coronary
artery bypass graft patency. Journal of the American College
of Cardiology. 2017;70(5):515–524.
60. Schwartz L et al. Coronary bypass graft patency in patients
with diabetes in the Bypass Angioplasty Revascularization
Investigation (BARI). Circulation. 2002;106(21):2652–2658.
61. Taggart DP et  al. Bilateral versus single internal‐thoracic‐
artery grafts at 10 years. N Engl J Med. 2019;380(5):
437–446.
62. Dorman MJ et al. Bilateral internal mammary artery graft­
ing enhances survival in diabetic patients: a 30‐year follow‐
up of propensity score–matched cohorts. Circulation.
2012;126(25):2935–2942.
63. Zhou P et al. Is the era of bilateral internal thoracic artery
grafting coming for diabetic patients? An updated meta‐
analysis. 2019. The Journal of Thoracic and Cardiovascular
Surgery. 158(6):1559–1570.
64. Aldea GS et  al. The Society of Thoracic Surgeons clinical
practice guidelines on arterial conduits for coronary artery
bypass grafting. The Annals of Thoracic Surgery. 2016;101(2):
801–809.
237
65. Kieser TM et  al. Toward zero: deep sternal wound infection
after 1001 consecutive coronary artery bypass procedures using
arterial grafts: implications for diabetic patients. The Journal of
Thoracic and Cardiovascular Surgery. 2014;148(5):1887–1895.
66. Di Mauro M et  al. Bilateral internal mammary artery for
multi‐territory myocardial revascularization: long‐term fol­
low‐up of pedicled versus skeletonized conduits. European
Journal of Cardio‐Thoracic Surgery. 2015;47(4):698–702.
67. Deo SV et al. Bilateral internal thoracic artery harvest and
deep sternal wound infection in diabetic patients. The
Annals of Thoracic Surgery. 2013;95(3):862–869.
68. Kereiakes DJ et  al. Outcomes in diabetic and nondiabetic
patients treated with everolimus‐or paclitaxel‐eluting stents:
results from the SPIRIT IV clinical trial (Clinical Evaluation
of the XIENCE V Everolimus Eluting Coronary Stent
System). Journal of the American College of Cardiology.
2010;56(25):2084–2089.
69. Konigstein M et  al. Outcomes among diabetic patients
undergoing percutaneous coronary intervention with con­
temporary drug‐eluting stents: analysis from the BIONICS
randomized trial. JACC: Cardiovascular Interventions. 2018;
11(24):2467–2476.
70. Kornowski R et al. Increased restenosis in diabetes mellitus
after coronary interventions is due to exaggerated inti­
mal  hyperplasia: a serial intravascular ultrasound study.
Circulation. 1997;95(6):1366–1369.
71. Ruel M et al. The SYNTAX score according to diabetic sta­
tus: What does it mean for the patient requiring myocardial
revascularization? The Journal of Thoracic and Cardiovascular
Surgery. 2019;159(3):857–860.
72. Dimitrova KR et al. Arterial grafts protect the native coro­
nary vessels from atherosclerotic disease progression. The
Annals of Thoracic Surgery. 2012;94(2):475–481.
73. Zhang M et al. Left internal mammary artery versus coro­
nary stents: impact on downstream coronary stenoses and
conduit patency. Journal of the American Heart Association.
2016;5(9):e003568.
74. Kinoshita T et  al. Off‐pump bilateral versus single skele­
tonized  internal thoracic artery grafting in patients with diabe­
tes. The Annals of Thoracic Surgery. 2010;90(4):1173–1179.
75. Gomes WJ et al. Coronary artery and myocardial inflamma­
tory reaction induced by intracoronary stent. The Annals of
Thoracic Surgery. 2003;76(5):1528–1532.
76. Couffinhal T et al. Vascular endothelial growth factor/vas­
cular permeability factor (VEGF/VPF) in normal and ath­
erosclerotic human arteries. Am J Pathol. 1997;150(5):1673.
77. Gregg EW et al. Changes in diabetes‐related complications
in the United States, 1990–2010. N Engl J Med. 2014;370(16):
1514–1523.
PART IV
Management of Disease
Complications
 20 Diabetes Management in Patients with
Critical and Non-Critical Illness
Kalpana Muthusamy1, Kristin Grdinovac2 and John M. Miles3
1
Division of Endocrinology, Olmsted Medical Center, Rochester, MN, USA
2
Assistant Professor of Medicine, University of Kansas, Kansas City, KS, USA
3
Professor of Medicine, University of Kansas, Kansas City, KS, USA
LE A RNI NG P OI NT S
OO To discuss evidence for and against the use of insulin
infusion to control hyperglycemia in the critically ill.
OO To consider the role of hypoglycemia in the disparate
results of large clinical trials investigating the use of
intensive insulin therapy in the critically ill.
OO To review the use of nutrition and its potential
influence on insulin requirements and outcomes in the
critically ill.
OO To discuss mechanisms by which insulin infusion may
modulate outcomes when used in critically ill patients.
Introduction
Hyperglycemia is a frequent finding in acute care hospitals,
with a prevalence rate in excess of one-­third of all inpa-
tients [1, 2]. Among patients with hyperglycemia (admis-
sion glucose > 7 mmol/L or random glucose > 11.1 mmol/L),
only two-­thirds are known to have diabetes; the remainder
either have newly diagnosed diabetes or temporary
(“stress”) hyperglycemia. Hyperglycemia has been identi-
fied as an independent predictor for higher infection
rates [3, 4] and increased mortality [5, 6] in critically ill
and postoperative patients. A number of large clinical tri-
als have been conducted over the past 15 years to address
the question of whether intravenous infusion of insulin
could improve morbidity and mortality in the critically
Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
ill. One of these trials [7] led to a consensus conference
convened in 2004 by the American College of
Endocrinology and cosponsored by the American
Diabetes Association, the American Heart Association,
the American Society of Anesthesiologists, the Endocrine
Society, the Society of Critical Care Medicine, and the
Society of Thoracic and Cardiovascular Surgeons [8]. The
consensus conference concluded, again on the basis of a
single clinical trial  [7], that blood glucoses should be
maintained < 6.1 mmol/L in critically ill patients, and that
insulin infusion was the means to achieve that goal. The
conference also recommended that a preprandial glucose
target < 6.1 mmol/L in patients who are eating on general
medical-­surgical wards should be adopted. This latter
recommendation, which in our view had and has very lit-
tle evidence to support it, is not the subject of this chapter
and will not be discussed further. Instead, we will address
mechanisms of hyperglycemia-­ related morbidity and
mortality in the critically ill and will undertake a detailed
review of the evidence supporting the use of insulin infu-
sion in the ICU, with emphasis on ten large clinical trials.
We will also discuss the importance of hypoglycemia,
nutritional support, and imprecision in blood glucose
testing in interpreting the results of these trials. We con-
clude that there is strong evidence to support the use of
insulin infusion to control hyperglycemia in the critically
ill, but very little evidence to indicate what the blood glu-
cose target should be.
241
 242 Management of Hyperglycemia in Critical Illness

Mechanisms Mediating Adverse Effects not proving) a role for reduced inflammation in improved
of Hyperglycemia During Critical Illness outcomes [15]. Modulation of nitric oxide synthase expres-
sion, and activity by insulin therapy mitigates the increased
During acute illness, hyperglycemia is exacerbated in peo-
nitric oxide levels seen during stress, thus protecting
ple with diabetes as a consequence of the release of counter-­
against endothelial damage  [16]. Endothelial adhesion
regulatory hormones and cytokines  [9]. The same
molecules such as ICAM-­1 and E-­selectin, despite being
mechanisms, along with subtle defects in insulin secretion,
comparably elevated at admission are lower during inten-
contribute to stress, or temporary, hyperglycemia in some
sive insulin therapy, indicating reduced endothelial activa-
nondiabetic individuals. The exact mechanisms by which
tion [16]. These factors in combination likely help maintain
hyperglycemia mediates adverse outcomes such as infec-
the integrity of the microvasculature, leading to less organ
tion and increased mortality are not fully understood,
damage and providing survival benefits.
however, nor are the mechanisms responsible for improved
There is also evidence that insulin therapy may favora-
outcomes resulting from correction of hyperglycemia.
bly influence outcomes by improving altered lipid homeo-
Several interesting hypotheses have been proposed involv-
stasis. Plasma-­free fatty acids (FFA) are elevated in the
ing both metabolic and non-­ metabolic pathways  [10].
critically ill, in spite of hyperinsulinemia due to resistance
Hyperglycemia can cause direct cellular damage as well as
to insulin’s antilipolytic effects in adipocytes [17]. This is of
indirect effects via increased generation of reactive oxygen
potential importance because elevated FFA contribute to
species. Although the normal protective response of cells
acute endothelial dysfunction  [18], have pressor
to hyperglycemia is to down regulate glucose transport,
effects [19], and can contribute to the dyslipidemia of criti-
expression and localization of GLUT-­1 and GLUT-­3 in sev-
cal illness by driving hepatic VLDL production  [20].
eral cell types has been shown to be increased by elevated
Mesotten et al., in a post hoc analysis of a large clinical trial,
levels of cytokines, growth factors, and hypoxia, which
found that intensive insulin therapy in the critically ill pro-
often occur together in the critically ill, potentially leading
duced substantially lower triglyceride levels  [21]. There
to cellular glucose overload [11]. Increased substrate avail-
was a four-­to fivefold increase in mortality over a range of
ability for glycolysis results in excess superoxide formation
triglyceride levels, although in a multivariate analysis HDL
that can overwhelm the normal detoxification process,
cholesterol had the strongest association with mortality.
leading to mitochondrial dysfunction [11]. Intensive insu-
The mechanism for increased morbidity and mortality
lin therapy has been shown to reverse or prevent hepatic
related to dyslipidemia is not known, but could involve
mitochondrial abnormalities noted with critical ill-
impairment of reticuloendothelial function, as discussed
ness [12]. There is also evidence showing improvement in
elsewhere [22]. Other proposed mechanisms of benefit are
the phagocytic and bactericidal activities of polymorpho-
insulin-­induced increase in expression of adiponectin [23]
nuclear leukocytes with improved glycemic control in dia-
and a potential role in enhancing innate immunity  [24].
betic subjects, thus enhancing immunity [13]. Glucotoxicity
Collectively, the evidence supports a role for nonglycemic
can also impair β-­cell secretory function, potentially initi-
effects of insulin therapy in improving outcomes in critical
ating a vicious cycle of relative insulin insufficiency and
illness, although the relative contribution of these factors is
further increases in glucose concentrations [14].
an important area for future research.
The pleiotrophic effects of insulin, independent of its
glucose-­lowering effect, have attracted considerable atten-
Large Intervention Studies
tion as a contributor to clinical benefits. Insulin has been
shown to have anti-­inflammatory properties, producing a Background
decrease in concentration of acute phase reactants. As recently as a decade ago, at a time when hyperglycemia
Reduced mortality and organ failure in critically ill patients was recognized as a predictor of adverse outcomes, ele-
treated with insulin infusion is associated with decreased vated blood glucose in the critically ill was often ignored.
high sensitivity C-­reactive protein levels, suggesting (but However, since that time the results of large clinical trials
 Diabetes Management in Patients with Critical and Non-Critical Illness
have forced a reconsideration of this attitude of indiffer-
ence. These results have produced wide swings in the pen-
dulum of expert opinion. Initially, the changes were subtle,
proposed glucose targets were quite liberal and, in fact,
were based more on associations than on harder evidence;
blood glucose levels of < 12.2 mmol/L were considered by
some to be optimal, as they were associated with improved
postoperative infection rates [25].
In 2004, an expert committee declared that a glucose
target of < 6.1 mmol/L was appropriate for most critically
ill patients  [8]. In January, 2009, the ADA stated in a
­position paper that “Critically ill surgical patients’ blood
glucose levels should be kept as close to 110  mg/dl
(6.1 mmol/L) as possible. . .” [26]. Four months later, in a
joint statement with the American Association of Clinical
Endocrinologists, the ADA declared that “a glucose range
of 140 to 180 mg/dl (7.8 to 10.0 mmol/L) is recommended
for the majority of critically ill patients.” [27]. What is the
explanation for these wide shifts in prevailing views?
In 2001, Van den Berghe and colleagues published a
landmark study (hereafter referred to as “Leuven 1”) of
more than 1500 patients in a surgical ICU. In this study,
they aimed for (and achieved) a target blood glucose of
4.4–6.1  mmol/L in an intensive control group and
10–11.1  mmol/L in a conventional treatment group  [7].
The in-­hospital mortality was significantly lower in the
intensive group at 7.2% compared to 10.9% in the conven-
tional arm. They also reported a reduction in mortality
during intensive care from 8% in the conventional group to
4.6% in the intensive group (42% relative reduction). Most
of these benefits were observed in the patients who stayed
in the intensive care unit (ICU) longer than 5 days. There
was significant improvement in morbidity related to fewer
days on mechanical ventilation, shorter ICU stay, reduced
infection, acute renal failure, critical-­illness polyneuropa-
thy, and fewer blood transfusions. These overwhelmingly
positive results were the sole basis for the target recom-
mended in the 2004 consensus statement and obviously
the driving force behind the January, 2009 ADA recom-
mendation. A survey revealed that the majority of inten-
sivists prefer a target blood glucose of 6.1  mmol/L  [28].
Although there was little published disagreement with the
guidelines, it is not clear that they penetrated to the local
level; a 2008–2009 informal poll of over 30 medium-­sized
243
hospitals in the United States, most of which did not have
a critical care service, revealed that whereas there was
a  general recognition of the value of insulin infusion
in  the  critically ill, virtually none were strictly adhering
to  the  6.1  mmol/L benchmark (JMM, unpublished
observations).
In March 2009 the results of the NICE-­SUGAR trial
were published, demonstrating that excessively tight con-
trol (target 6.0 mmol/L, average glucose 6.4 mmol/L) might
actually be harmful compared to a less stringent target
(10 mmol/L, average glucose 8.0 mmol/L) [29]. The odds
ratio for death in the intensive arm of this study was 1.14 (p
= 0.02). In the subset of patients with traumatic brain-­
injury a later paper concluded that despite a higher inci-
dence of hypoglycemia in the intensively treated group,
these patients exhibited no clinically-­significant difference
in outcomes compared to the conventionally treated group.
Finfer S et  al. Intensive Care Medicine. 41(6):1037-­47,
2015 JunThe ADA-­AACE consensus statement relaxing the
guidelines was published less than two months later.
Where does this leave us? The results of meta-­
analyses  [31, 32] conclude that intensive insulin therapy
does not reduce mortality in critically ill patients and
increases the risk of hypoglycemia, although one
report [32] acknowledges it might benefit surgical patients.
Taken at face value, it could be argued that these studies
provide a rationale for abandoning intensive treatment
with insulin infusion altogether. The fact that the authors
of the ADA-­AACE consensus statement still endorse inten-
sive treatment, albeit in a less aggressive form  [26], indi-
cates that they must see merit in the several studies that
have reported benefit from this treatment. For this chapter,
we will review in detail the results of ten large clinical trials
that have produced disparate, difficult-­ to-­
reconcile
results  [7, 29, 33–40]. We included all trials of > 500
patients in which glycemic targets were part of the study
design, but we excluded studies in which glucose infusion
was employed in one group but not the other for the pur-
pose of allowing infusion of insulin at higher rates than
would otherwise be necessary to control hyperglyce-
mia [41]. Our purpose is to attempt to identify features of
the studies that might explain divergent findings leading to
conflicting conclusions and recommendations. Specifically,
we hoped to determine whether variables such as acuity of
 244 Management of Hyperglycemia in Critical Illness
illness in the study population, hypoglycemia, blood glu-
cose testing methodology, and nutritional support might
explain differences among studies.
General Findings (Tables 20.1 and 20.2)
Table 20.1 shows the 10 studies in chronological order and
provides details on trial design and study populations,
together with glucose targets and levels achieved. Eight of
the studies were randomized, controlled trials; two used
historical controls. Diabetes mellitus was an entry criterion
for two of the studies; in the other eight, the majority of
patients did not have diabetes. Patients were receiving
nutritional support in eight of the studies but did not
receive artificial nutrition in two [33, 34]. Table 20.2 pro-
vides information on severity of illness (APACHE II scores)
and mortality.
We believe that the first five studies should be consid-
ered to indicate benefit from intensive insulin therapy,
whereas the second five studies should be considered
negative. Certainly, this interpretation could be ques-
tioned. The DIGAMI study was likely underpowered, and
showed benefit in the whole cohort only with long-­term
follow-­up. However, in-­ hospital mortality was signifi-
cantly improved with intensive therapy in the diabetic
patients who were lower risk and not previously treated
with insulin [33]. The Portland and Stamford studies uti-
lized historical controls, an inherently weaker design, but
both had the strength of a large sample size. The Leuven 2
study has been interpreted as unequivocally negative by
some authors because there was no difference in the pri-
mary end point, i.e., mortality in the aggregate cohort [36].
However, in the subset of individuals who were in the
ICU for ≥3 days, mortality was significantly reduced with
intensive therapy. This delay in apparent benefit was also
observed to a lesser extent in the surgical ICU study from
the same investigators [7]. The benefit in the long-­stayers
in ICU in Leuven 2 was least apparent in patients in the
highest quartile of APACHE II scores [36]. The mortality
data in Table 20.2 clearly indicate the Leuven 2 patients to
be the sickest of those in any of the ten studies. It has been
pointed out that a failure of benefit with short-­term ther-
apy could dilute to insignificance an effect in patients
with longer ICU stays [42].
The last five of the studies found no benefit of intensive
insulin therapy, and the NICE-­ SUGAR study actually
reported higher mortality in intensively treated patients [29].
Three of the five studies were smaller and perhaps under-
powered. Furthermore, in two of the five studies the glu-
cose target range in the control group was lower than the
Leuven trials. The threshold for initiating insulin infusion
(a higher value than the actual target range in many stud-
ies) in the control group was lower than in the Leuven
studies in two of the trials [29, 38] and not stated in two
others [39, 40]. It is therefore not surprising that the aver-
age glucose in the control group of four of the five negative
studies was lower or equal to the average glucose in the
control group of all of the positive studies. Mesotten and
van den Berghe suggest that this is due to the strong influ-
ence of the positive studies on the design of the subsequent
negative studies, and could have minimized differences in
mortality between groups [43].
In a pediatric ICU setting 1369 patients under the age of
16 at 13 centers in England were randomized to tight glyce-
mic control (694) and to conventional glycemic control
(675); 60% had undergone cardiac surgery. Participants
were expected to require mechanical ventilation and vaso-
pressor support for at least 12 hours. As seen in adult stud-
ies, tight glycemic control had no effect on major clinical
outcomes, although the incidence of hypoglycemia was
higher in this group Allen E et al. N Engl J Med. 370(2):107–
118, 2014 Jan 09.
Role of Hypoglycemia in Interpretation
of Study Results (Table 20.3)
Hypoglycemia (defined as a blood glucose ≤ 2.2 mmol/L, or
< 40 mg/dl) was reported in the landmark van den Berghe
study of surgical patients as more common in patients
receiving intensive treatment than in the control group, but
not leading to hemodynamic deterioration or convulsions.
It was not discussed further. The frequency of hypoglycemia
was remarkably low in the Krinsley study of medical ICU
patients, especially considering that a more liberal definition
of hypoglycemia (< 3.3 mmol/L) was employed. The Leuven
2 study was the first of the trials to indicate that hypoglyce-
mia might be responsible for mortality in some patients. In
three of the five subsequent negative studies there was an
TABLE 20.1  Trial design, patient characteristics.

Mean blood glucose Mean blood glucose

target (mmoI/L) achieved (mmoI/L)

Type of Patient Number of Diabetes

Study study population patients (%) intensive conventional intensive conventional

DIGAMI, RCT, Coronary care 620 100 7.0–10.9 Usual care 9.6 11.7
1997 [33] multicenter unit
Leuven-­1, RCT, Surgical ICU 1548 13 4.4–6.1 10–11.1 5.7 8.5
2001 [7] single-­center
Furnary, Historical CABG 3554 100 variable < 11.1 9.8 11.9
2003 [34] controls,
single-­center
Krinsley, Historical Medical-­surgical 1600 17 < 7.8 Usual care 7.3 8.4
2004 [35] controls, ICU
single-­center
Leuven-­2, RCT, Medical-­surgical 1200 17 4.4–6.1 10–11.1 5.8 8.6
2006 [36] single-­center ICU
De La Rosa, RCT, Medical-­surgical 504 29–32 4.4–6.1 10–11.1 6.5 8.2
2008 [37] single-­center ICU
VISEP, RCT, Medical-­surgical 537 30 4.4–6.1 10–11.1 6.2 8.4
2008 [38] multicenter ICU
Arabi, RCT, Medical-­surgical 523 32–48 4.4–6.1 10-­11.1 6.4 9.5
2008 [39] single-­center ICU
NICE-­ RCT, Medical-­surgical 6104 20 4.5–6.1 < 10 6.4 8.0
SUGAR, multicenter ICU
2009 [29]
Glucontrol, RCT, Medical-­surgical 1078 16–21 4.4–6.1 7.8–10 6.5 8.0
2009 [40] multicenter ICU

TABLE 20.2  APACHE scores and mortality data.

Mortality (%)

APACHE Relative reduction in Basis for difference in

Study score Intensive Conventional mortality (%) mortality vs controls

DIGAMI, 1997 [33] NR 19 (at 1 year 26 30 NR

follow-­up)
Leuven-­1, 2001 [7] 9 (median) 7.2 (in-­hospital) 10.9 34 ↓ sepsis, MOF
Furnary, 2003 [34] NR 2.5 (in-­hospital) 5.3 47 Cardiac related
Krinsley, 2004 [35] 15 (median) 14.8 20.9 20.9 ↓ in subpopulation with
septic shock, neurological
or surgical diagnosis
Leuven-­2, 23 (mean) 37.3 (43 for > 3 40 (52.5 for > 3 ICU NS (18,P< 0.05* ) Multiple
2006 [36] ICU days) days)
(in-­hospital)
De La Rosa 16 (mean) 36.6 (in-­hospital) 32.4 NS –
2008 [37]
VISEP, 2008 [38] 20 (mean) 39.7 (90 day) 35.4 NS –
Arabi, 2008 [39] 23 (mean) 13.5 17.1 NS –
NICE-­SUGAR, 21 (mean) 27.5 24.9 Increased by 14 Increases in cardiovascular
2009 [29] death
Glucontrol, 15 (mean) 15.3 17.2 NS –
2009 [40]
 246 Management of Hyperglycemia in Critical Illness

TABLE 20.3  Hypoglycemia data.

Incidence of hypoglycemia (%) Association between

Definition of hypoglycemia and
Study hypoglycemia (mmoI/L) Intensive Conventional mortality?

DIGAMI, 1997 [33] < 3.0 15.0 0.0 NR

Leuven-­1, 2001 [7] ≤2.2 5.1 0.8 No
Furnary, 2003 [34] NR NR NR NR
Krinsley, 2004 [35] < 3.3 1.4 0.9 No
Leuven-­2, 2006 [36] ≤2.2 18.7 3.1 Yes
De La Rosa, 2008 [37] ≤2.2 8.3 0.8 NR
VISEP, 2008 [38] ≤2.2 12.1 2.1 Yes
Arabi, 2008 [39] ≤2.2 28.6 3.1 Yes
NICE-­SUGAR, ≤2.2 6.8 0.5 No
2009 [29]
Glucontrol, 2009 [40] ≤2.2 8.7 2.7 Yes

association between hypoglycemia and mortality  [38–40].
An association between hypoglycemia and mortality was
confirmed in the NICE-­SUGAR study, and the authors
speculated that hypoglycemia could be partly responsible
for adverse effects observed with intensive therapy  [29].
Thus, there is substantial evidence that hypoglycemia could
have material harmful effects on outcomes in patients
receiving intensive treatment with insulin infusion.
It is noteworthy that hypoglycemia is conventionally
defined in the critical care literature as a value
≤2.2 mmol/L [7, 29, 36–40]. It appears that this definition
is used because of a prevailing view that less severe hypo-
glycemia is not clinically important. In fact, a recent survey
of adult intensivists showed that over 40% thought hyper-
glycemia was more dangerous than hypoglycemia, and a
majority preferred a target of 6.1 mmol/L [28]. Ironically,
in the same survey the median value for blood glucose that
was considered to be the hypoglycemic threshold was
3.33  mmol/L. In fact, severe symptoms of hypoglycemia
can occur at glucose concentrations much higher than
2.2 mmol/L, and the counterregulatory response, including
sympathoadrenal activation, routinely occurs as glucose
levels descend through the 3.6–3.8  mmol/L range  [45].
Moreover, defective counter-­regulation  [42] and lack of
clinical cues of hypoglycemia may make hypoglycemia
more difficult to identify, but not necessarily less threaten-
ing. For this reason, hypoglycemia experts have argued that
the threshold for clinical important hypoglycemia lies in
the 3.5 mmol/L [46] to 3.9 mmol/L [47] range. It is there-
fore simply not reasonable to dismiss glucose levels of 2.3–
3.8 mmol/L as clinically unimportant in patients who are
intubated  [7], receiving propofol or other sedatives  [37,
39], or both. Most of the studies reviewed here are encum-
bered by the limitation that clinically important hypoglyce-
mia may have been excluded from consideration by an
inappropriately narrow definition of hypoglycemia. A
rather high incidence of hypoglycemia (15%) was reported
in the intensive arm of the DIGAMI study [33], but this is
likely because a liberal definition (< 3.0 mmol/L) of hypo-
glycemia was used. The incidence of “severe” (<
2.2  mmol/L) hypoglycemia in the DIGAMI study is not
known.
In summary, it is clear that hypoglycemia is a potentially
serious limitation to the pursuit of near-­euglycemia with
intensive insulin therapy in the critically ill. It is possible,
even likely, that hypoglycemia mitigated beneficial effects
of intensive insulin therapy in negative  [29, 36–40] and
equivocal [36] studies. Failure to consider less severe (2.3–
3.8 mmol/L) hypoglycemia is a serious limitation of these
studies. A recent retrospective analysis of hypoglycemia in
1109 patients from two Australian ICUs demonstrated a
step-­wise and significant increase in mortality associated
with increasingly severe hypoglycemia, including patients
with glucose values of 4.0–4.5 mmol/L [48]. This indicates
that even very mild hypoglycemia may have significant
effects on survival in the critically ill.
 Diabetes Management in Patients with Critical and Non-Critical Illness
Methodology for Glucose Testing
The observation that glucose values (4.0–4.5 mmol/L) not
considered to be in the hypoglycemic range by even the
most liberal definition of hypoglycemia could be associated
with increased mortality is difficult to understand. In peo-
ple with diabetes, there may be a shift upward in the glyce-
mic threshold for the sympathoadrenal response  [49].
Another possible explanation would apply to nondiabetic
as well as diabetic patients, and relates to the methods used
to measure glucose in the critically ill. Point-­of care (POC)
blood glucose testing was used in a number of the studies,
and it has been shown to be very inaccurate and imprecise,
especially during hypoglycemia. CLIA defines inaccuracy
with POC glucose devices as values > 20% different from
the laboratory value except in the case of hypoglycemia,
where values > 0.83 mmol/L different are considered inac-
curate. Thus, a true glucose level of 4.0 mmol/L could be
reported as 3.2  mmol/L, or a true level of 2.8 could be
reported as 3.6 mmol/L, and both values would be consid-
ered to be accurate. Critchell et  al. reported that POC
devices, when used by trained laboratory technicians, were
inaccurate in 19% of samples  [50]. Kanji et  al. found the
agreement of POC measurements with a central laboratory
to be 56%, and only 26% in hypoglycemia samples [51]. The
potential for error when POC devices are used is thus read-
ily apparent. POC devices may also systematically overesti-
mate true glucose  [50]. Among the large clinical trials
reviewed here, a bedside meter was used exclusively in two
of the negative studies [37, 39]. NICE-­SUGAR used both a
blood gas machine, which has excellent precision and accu-
racy [52], and a POC device. When a POC device was used
in NICE-­SUGAR, laboratory confirmation was obtained
only 60% of the time. Even when hypoglycemic values were
confirmed, it should be recognized that the antecedent glu-
cose level was unconfirmed and thus could have dictated an
inappropriate decision about insulin infusion rate that led
to the subsequent hypoglycemic episode. The Leuven 1
study used a blood gas instrument, and Leuven 2 used the
Hemocue, a POC device that is designed for clinical use,
not for self-­monitoring. The VISEP study utilized both a
blood gas device and the Hemocue  [38]. Thus, the tech-
niques used for blood glucose measurement in several of
the studies may have contributed to hypoglycemia, failed to
identify hypoglycemia, and even overcalled hypoglycemia.
247
The Role of Nutritional Support in Results
of Trials
A physician, so the story goes, once gave 1 mg of thyroxine
daily to an elderly, frail woman who complained of fatigue.
When the patient developed acute atrial fibrillation and
died, the physician concluded that death was to be expected
from thyroxine administration, and that it should never be
used again. This kind of reasoning has been attributed to
the “Chagrin Factor,” which derives from aversion to a
course of action that has produced an unexpected poor
outcome [53].
Nutrition support is often used in the care of the criti-
cally ill, affecting glucose levels and insulin requirements.
In two of the ten large trials [33, 34], nutrition was given to
very few patients [22]. Enteral nutrition was used almost
exclusively in the study of Krinsley (J Krinsley, personal
communication to JMM). Enteral nutrition was used pri-
marily in four studies [29, 37, 39, 40], whereas enteral and
parenteral nutrition were used equally in the VISEP
study [38]. In the two Leuven studies, the majority of nutri-
tional support was parenteral, at least for the first 7 days in
the ICU. This has an impact on interpretation of the data
from the Leuven 1 study, since parenteral nutrition is
known to increase infection rates [54], and differences in
infection, especially sepsis, accounted for most of the dif-
ferences in mortality between the intensive and control
groups in Leuven 1  [22, 55]. Enteral feeding does not
appear to confer the same risk of infection [55], although
control data on this issue are not available. In contrast, the
differences in mortality in the NICE-­SUGAR study were
based primarily on cardiovascular outcomes, which may be
influenced adversely by hypoglycemia [56].
In addition to the route of nutrition, the amount of nutri-
tion given may influence the results of studies of glycemic
control in the critically ill. We calculated energy supply in
the Leuven 1 trial in relation to estimated basal energy
expenditure, and thus basal energy requirements, using the
Harris–Benedict equation  [22]. This can be done for any
study in which the calories given are reported and height
and weight are provided. In the absence of height and weight
data, the calculation can be made from body mass index and
estimated height [22]. We therefore calculated energy given
in relation to energy requirements in five of the reports
where adequate information was available  [7, 29, 36–39],
 248 Management of Hyperglycemia in Critical Illness

140

130

Energy supply (% BEE)

120

110

100

90

80

70
Leuven 1 Leuven 2 De La VISEP NICE-
Rosa SUGAR

FIG 20.1  Energy supply expressed as a percentage of estimated basal energy expenditure (BEE, calculated from the Harris–Benedict
equation) in five studies.

150
LEUVEN 1

130
Energy supply (% BEE)

DE LA ROSA LEUVEN 2

110

NICE-SUGAR y = 1.4169x + 34.752

90 R2= 0.923
VISEP

70
30 40 50 60 70 80
Insulin infusion rate (U/day)

FIG 20.2  Energy supply versus insulin infusion rate in five studies.

using data on the intensively treated arm of the studies.
Height was estimated from online information (e.g., http://
forums.interbasket.net/f10/average-­m ale-­h eight-­b y-­
country-­updated-­9287/ and http://en.wikipedia.org/wiki/
Human height). Including protein in the energy calculation
(estimated to be 1.0 g/kg per day when not provided), the
values used for daily feeding rate were 27 kcal/kg for Leuven
1 [22], 25 kcal/kg for Leuven 2 [36], 25.5 kcal/kg for De La
Rosa  [37], 1236  kcal/day for VISEP  [38] [from online
appendix], and 1624  kcal for NICE-­SUGAR  [29] [from
online appendix]. In several of the studies  [7, 29, 38] the
average energy intake on days 3–5 was used. Energy intake
 Diabetes Management in Patients with Critical and Non-Critical Illness
in the Arabi study was provided only as an average given
over the first 7  days, and was much lower than expected
based on the method for feeding described by the authors –
using the Harris–Benedict equation and adjusting for stress
factors  [39]; for this reason, we excluded the Arabi study
from the analysis. Energy intake expressed as a percentage of
estimated basal energy requirements is shown in Figure 20.1.
The feeding rates among the various studies are surprisingly
different. The amount of nutrition given in the Leuven stud-
ies, although within published guidelines, is greater than
probable energy requirements in the critically ill  [57] and
may account for the rather high insulin requirements in the
intensively treated groups in those studies, especially note-
worthy considering that very few of the participants had
diabetes. In contrast, studies that relied primarily on enteral
feedings such as the NICE-­SUGAR study provided energy
more in line with probably energy requirements [57]. This
may be in part because tube feeding intolerance tends to
limit infusion rate of enteral formulae. The relationship
between amount of feeding and insulin infusion rate for the
intensive groups of five studies is shown in Figure  20.2. As
can be seen, there was a strong correlation between feeding
rate and insulin requirement, particularly impressive since
the modes of feeding were heterogenous among the studies,
and when energy supply is controlled for, insulin require-
ments related to intravenous feeding are greater than
requirements during enteral feeding [58]. The data shown in
the figures are admittedly estimates, but do serve to illustrate
how variable approaches to nutrition support can impact
insulin requirements in the critically ill. This is an important
point, as it is common to invoke “stress” as the cause of tem-
porary hyperglycemia in critically ill nondiabetic patients,
without sufficient attention to the role of overfeeding [57].
A recently published meta-­analysis concluded that there
is no evidence to support the use of intensive insulin ther-
apy in general medical-­ surgical patients who are fed
according to current guidelines [59]. The study was unu-
sual in that it attempted to assess the role of nutritional
support in the outcomes of some of the same clinical trials
reviewed here. It found that there was a high incidence of
hypoglycemia and increased risk of death in patients on
intensive insulin therapy not receiving parenteral nutri-
tion  [29, 37, 39, 40]. Unfortunately, the authors did not
consider the impact of how hypoglycemia is defined or the
249
methodology used for blood glucose testing on frequency
of hypoglycemia, and did not consider the possibility that
(1) hypoglycemia could be avoided, that (2) selection of
more reasonable glucose targets and use of better glucose
testing techniques could facilitate avoidance of hypoglyce-
mia, and (like the physician who gave up on thyroxine
rather than consider a different dose) that (3) successful
avoidance of hypoglycemia might completely change study
outcomes and conclusions.
Conclusions
Available data concerning intensive therapy of hyperglyce-
mia in the critically ill is inconsistent and confusing.
However, much of the contradictory data is likely related to
differences in experimental design that result in different
frequency and severity of hypoglycemia. In most of the
studies reviewed here, hypoglycemia was poorly defined
and insufficiently respected. It should be emphasized that
hypoglycemia may exert adverse effects of its own, masking
beneficial effects of glycemic control. Problems with the
accuracy and precision of available blood glucose testing
methodology impose considerable limitations on interpre-
tation of data and contribute to safety concerns. Nutritional
support may also confound the interpretation of large clini-
cal trials, although avoidance of overfeeding  [57] should
minimize concerns in that regard. Based on available data,
we believe that intensive insulin therapy has a role in the
management of the critically ill, although it is unclear what
the optimal target range should be. Pending further
research and in the interest of minimizing hypoglycemia,
we believe that the current ADA/AACE guidelines for
treatment of hyperglycemia on the critically ill, which rec-
ommend insulin infusion with a blood glucose target of
7.8–10  mmol/L, are appropriate. Robust and carefully
standardized methods for blood glucose testing should be
used, staff should be well trained, and hypoglycemia should
be appropriately defined and assiduously avoided.
Acknowledgments
This work was supported in part by the USPHS (HL67933)
and the Mayo Foundation. We thank P. E. Cryer for helpful
comments on the manuscript.
 250 Management of Hyperglycemia in Critical Illness
References
  1. Umpierrez GE, Isaacs SD, Bazargan N et al. Hyperglycemia:
an independent marker of in-­hospital mortality in patients
with undiagnosed diabetes. J Clin Endocrinol Metab 87:978–
982, 2002.
  2. Cook CB, Kongable GL, Potter DJ et  al. Inpatient glucose
control: a glycemic survey of 126 U.S. hospitals. J Hosp Med.
4:E7–E14, 2009.
  3. Golden SH, Peart-­Vigilance C et al. Perioperative glycemic
control and the risk of infectious complications in a cohort
of adults with diabetes. Diabetes Care 22:1408–1414, 1999.
  4. Zerr KJ, Furnary AP, Grunkemeier GL et al. Glucose control
lowers the risk of wound infection in diabetics after open
heart operations. Ann Thorac Surg 63:356–361, 1997.
  5. Weir CJ, Murray GD, Dyker AG, and Lees KR. Is hypergly-
caemia an independent predictor of poor outcome after
acute stroke? Results of a long-­term follow up study. BMJ.
314:1303–1306, 1997.
 6. Krinsley JS. Association between hyperglycemia and
increased hospital mortality in a heterogeneous population
of critically ill patients. Mayo Clin Proc. 78:1471–1478, 2003.
 7. van den Berghe G, Wouters P, Weekers F et  al. Intensive
insulin therapy in critically ill patients. N Engl J Med.
345:1359–1367, 2001.
  8. Garber AJ, Moghissi ES, Bransome ED, Jr. et al. ACE posi-
tion statement. Endocr Pract 10:5–9, 2004.
  9. McCowen KC, Malhotra A, and Bistrian BR. Stress-­induced
hyperglycemia. Crit Care Clin. 17:107–124, 2001.
10. Andreelli F, Jacquier D, and Troy S. Molecular aspects of
insulin therapy in critically ill patients. Curr Opin Clin Nutr
Metab Care. 9:124–130, 2006.
11. van den Berghe G. How does blood glucose control with
insulin save lives in intensive care? J Clin Invest. 114:1187–
1195, 2004.
12. Vanhorebeek I, De Vos R, Mesotten D et  al. Protection of
hepatocyte mitochondrial ultrastructure and function by
strict blood glucose control with insulin in critically ill
patients. Lancet. 365:53–59, 2005.
13. Bagdade JD, Stewart M, and Walters E. Impaired granulo-
cyte adherence. A reversible defect in host defense in
patients with poorly controlled diabetes. Diabetes. 27:677–
681, 1978.
14. Poitout V and Robertson RP. Glucolipotoxicity: fuel excess
and beta-­cell dysfunction. Endocrine Rev. 29:351–366, 2008.
15. Hansen TK, Thiel S, Wouters PJ et al. Intensive insulin ther-
apy exerts antiinflammatory effects in critically ill patients
and counteracts the adverse effect of low mannose-­binding
lectin levels. J Clin Endocrinol Metab. 88:1082–1088, 2003.
16. Langouche L, Vanhorebeek I, Vlasselaers D et al. Intensive
insulin therapy protects the endothelium of critically ill
patients. J Clin Invest. 115:2277–2286, 2005.
17. Stoner HB, Little RA, Frayn KN et al. The effect of sepsis on
the oxidation of carbohydrate and fat. Br J Surg. 70:32–35,
1983.
18. Steinberg HO, Tarshoby M, Monestel R et al. Elevated circu-
lating free fatty acid levels impair endothelium-­dependent
vasodilation. J Clin Invest. 100:1230–1239, 1997.
19. Stojiljkovic MP, Zhang D, Lopes HF et  al. Hemodynamic
effects of lipids in humans. Am J Physiol. 280:R1674–R1679,
2001.
20. Lewis GF, Uffelman KD, Szeto LW et al. Interaction between
free fatty acids and insulin in the acute control of very low
density lipoprotein production in humans. J Clin Invest.
95:158–166, 1995.
21. Mesotten D, Swinnen JV, Vanderhoydonc F et  al.
Contribution of circulating lipids to the improved outcome
of critical illness by glycemic control with intensive insulin
therapy. J Clin Endocrinol Metab. 89:219–226, 2004.
22. Miles JM, McMagon MM, and Isley WL. For debate: no, the
glycemic target in the critically ill should not be <6.1 mmol/L.
Diabetologia. 51:916–920, 2008.
23. Langouche L, Vander Perre S, Wouters PJ et  al. Effect of
intensive insulin therapy on insulin sensitivity in the criti-
cally ill. J Clin Endocrinol Metab. 92:3890–3897, 2007.
24. Weekers F, Van Herck E, Coopmans W et al. A novel in vivo
rabbit model of hypercatabolic critical illness reveals a
biphasic neuroendocrine stress response. Endocrinology.
143:764–774, 2002.
25. Pomposelli JJ, Baxter JK, 3rd, Babineau TJ et al. Early post-
operative glucose control predicts nosocomial infection rate
in diabetic patients. JPEN: J Parenter Enteral Nutr. 22:77–81,
1998.
26. ADA: Standards of Medical Care in Diabetes  – 2009.
Diabetes Care 32:S41, 2009.
27. Moghissi ES, Korytkowski MT, DiNardo M et al. American
Association of Clinical Endocrinologists and American
Diabetes Association consensus statement on inpatient gly-
cemic control. Diabetes Care. 32:1119–1131, 2009.
28. Hirshberg E, Lacroix J, Sward K et al. Blood glucose control
in critically ill adults and children: a survey on stated prac-
tice. Chest 133:1328–1335, 2008.
29. Investigators N-­SS, Finfer S, Chittock DR et al. Intensive ver-
sus conventional glucose control in critically ill patients.
NEngl J Med. 360:1283–1297, 2009.
30. Finfer S, Chittock D, Li Y et al. Intensive versus conventional
glucose control in critically ill patients with traumatic brain
injury: long-­term follow-­up of a subgroup of patients from
 Diabetes Management in Patients with Critical and Non-Critical Illness
the NICE-­ SUGAR study. Intensive Care Medicine.
41(6):1037–1047, 2015.
31. Wiener RS, Wiener DC, and Larson RJ. Benefits and risks of
tight glucose control in critically ill adults: a meta-­analysis.
[Erratum appeared in JAMA. 2009  Mar4;301(9):936].
JAMA. 300:933–944, 2008.
32. Griesdale DEG, de Souza RJ, van Dam RM et al. Intensive
insulin therapy and mortality among critically ill patients: a
meta-­analysis including NICE-­SUGAR study data. CMAJ.
180:821–827, 2009.
33. Malmberg K. Prospective randomised study of intensive insulin
treatment on long term survival after acute myocardial infarc-
tion in patients with diabetes mellitus. DIGAMI (Diabetes
Mellitus, Insulin Glucose Infusion in Acute Myocardial
Infarction) Study Group. BMJ. 314:1512–1515, 1997.
34. Furnary AP, Gao G, Grunkemeier GL et  al. Continuous
insulin infusion reduces mortality in patients with diabetes
undergoing coronary artery bypass grafting. J Thorac
Cardiovasc Surg. 125:1007–1021, 2003.
35. Krinsley JS. Effect of an intensive glucose management pro-
tocol on the mortality of critically ill adult patients. Mayo
Clin Proc. 79:992–1000, 2004.
36. Van den Berghe G, Wilmer A, Hermans G et al. Intensive insulin
therapy in the medical ICU. N Engl J Med. 354:449–461, 2006.
37. De La Rosa GDC, Donado JH, Restrepo AH et al. Strict gly-
caemic control in patients hospitalised in a mixed medical
and surgical intensive care unit: a randomised clinical trial.
Crit Care (London, England). 12:R120, 2008.
38. Brunkhorst FM, Engel C, Bloos F et  al. Intensive insulin
therapy and pentastarch resuscitation in severe sepsis. N
Engl J Med. 358:125–139, 2008.
39. Arabi YM, Dabbagh OC, Tamim HM et al. Intensive versus
conventional insulin therapy: a randomized controlled trial
in medical and surgical critically ill patients. Crit Care Med.
36:3190–3197, 2008.
40. Preiser J-­C, Devos P, Ruiz-­Santana S et al. A prospective ran-
domised multi-­centre controlled trial on tight glucose control
by intensive insulin therapy in adult intensive care units: the
Glucontrol study. Intensive Care Med. 35:1738–1748, 2009.
41. Gray CS, Hildreth AJ, Sandercock PA et  al. Glucose-­
potassium-­insulin infusions in the management of post-­
stroke hyperglycaemia: the UK Glucose Insulin in Stroke
Trial (GIST-­UK). Lancet Neurol. 6:397–406, 2007.
42. Cryer PE. Hypoglycaemia: the limiting factor in the glycae-
mic management of the critically ill? Diabetologia. 49:1722–
1725, 2006.
43. Mesotten D, Van den Berghe G. Clinical benefits of tight gly-
caemic control: focus on the intensive care unit. Best Practice
& Research Clinical Anaesthesiology. 23:421–429, 2009.
251
44. Macrae D, Grieve R, Allen E et  al. A randomized trial of
hyperglycemic control in pediatric intensive care. N Engl J
Med. 370(2):107–118, 2014 Jan 09
45. Cryer PE. Hierarchy of physiological responses to hypogly-
cemia: relevance to clinical hypoglycemia in type I (insulin
dependent) diabetes mellitus. Horm Metab Res. 29:92–96,
1997.
46. Frier BM. Defining hypoglycaemia: what level has clinical
relevance? Diabetologia. 52:31–34, 2009.
47. Cryer PE. Preventing hypoglycaemia: what is the appropri-
ate glucose alert value? Diabetologia. 52:35–37, 2009.
48. Egi M, Bellomo R, Stachowski E et  al. Hypoglycemia and
outcome in critically ill patients. Mayo Clin Proc.
85:217–224.
49. Boyle PJ, Schwartz NS, Shah SD et al. Plasma glucose con-
centrations at the onset of hypoglycemic symptoms in
patients with poorly controlled diabetes and in nondiabetics.
N Engl J Med. 318:1487–1492, 1988.
50. Critchell CD, Savarese V, Callahan A et al. Accuracy of bed-
side capillary blood glucose measurements in critically ill
patients. Intens Care Med. 33:2079–2084, 2007.
51. Kanji S, Buffie J, Hutton B et al. Reliability of point-­of-­care
testing for glucose measurement in critically ill adults. Crit
Care Med. 33:2778–2785, 2005.
52. Beneteau-­Burnat B, Bocque M-­C, Lorin A et al. Evaluation
of the blood gas analyzer Gem PREMIER 3000. Clin Chem
Lab Med. 42:96–101, 2004.
53. Feinstein AR. The ‘chagrin factor’ and quantitative decision
analysis. Arch Int Med. 145:1257–1259, 1985.
54. Perioperative total parenteral nutrition in surgical patients.
The Veterans Affairs Total Parenteral Nutrition Cooperative
Study Group. N Engl J Med. 325:525–532, 1991.
55. Braunschweig CL, Levy P, Sheean PM, and Wang X. Enteral
compared with parenteral nutrition: a meta-­analysis. Am J
Clin Nutr. 74:534–542, 2001.
56. Adler GK, Bonyhay I, Failing H et al. Antecedent hypoglyce-
mia impairs autonomic cardiovascular function: implica-
tions for rigorous glycemic control. Diabetes. 58:360–366,
2009.
57. Miles JM. Energy expenditure in hospitalized patients:
implications for nutritional support. Mayo Clin Proc.
81:809–816, 2006.
58. van den Berghe G, Wouters PJ, Bouillon R et  al. Outcome
benefit of intensive insulin therapy in the critically ill:
Insulin dose versus glycemic control. Crit Care Med. 31:359–
366, 2003.
59. Marik PE and Preiser J-­C. Toward understanding tight gly-
cemic control in the ICU: a systematic review and metaanal-
ysis. Chest. 137:544–551, 2010.
 21 Diagnosis and Management
of Ophthalmic Complications of Diabetes
Andrew J. Barkmeier1 and Konstantin Astafurov2
1
Associate Professor of Ophthalmology, Department of Ophthalmology, Mayo Clinic, Rochester, MN, USA
2
Konstantin Astafurov, MD PhD, Fellow in Vitreoretinal Surgery, NJRetina/Rutgers Robert Wood Johnson Medical
School, New Brunswick, NJ, USA

currently available therapies for ocular complications of

LE A R N I N G P OI NT S
diabetes (Table 21.1) and briefly touch upon future poten­
●● The most significant threat for vision loss in diabetes is
tial therapies.
due to retinal disease.
Glycemic and metabolic control helps prevent ocular
Risk Factors for Diabetic Retinopathy
●●

microvascular complications of diabetes.

●● In combination with laser photo‐coagulation,
anti‐VEGF therapy improves outcomes in diabetic
macular edema and proliferative retinopathy.
Introduction
Long‐standing DM may affect many parts of the eye
including the lens, cornea, and orbit; however, the most
significant threat for vision loss in DM is due to its effects
on the retina. Diabetic retinopathy (DR) is a microvascular
complication of diabetes that develops in substantial pro­
portion of diabetic patients over time. Epidemiologic stud­
ies have estimated that one in three patients with diabetes
will develop DR [1]. If untreated, advanced stages of DR,
such as proliferative DR (PDR) and diabetic macular
edema (DME) may lead to irreversible blindness. In fact,
DR is the primary cause of visual loss in patients between
25 and 74 years of age [2]. If detected early and treated in a
timely manner, however, the most of visual impairment
from diabetes‐related ocular complications could be mini­
mized and even prevented. In this chapter we discuss
The prevalence of diabetic retinopathy increases with the
duration of diabetes in patients with both Type 1 and Type
2 diabetes. Early epidemiologic studies have revealed that
after 20 years of DM, nearly 97% of patients with Type 1
DM and between 50% and 80% of patients with Type 2 DM
exhibit some degree of DR  [3]. Two other major factors
associated with increased risk of development and progres­
sion of DR are poor glycemic control and uncontrolled
hypertension. Additional risk factors include the presence
of other microvascular complications, such as diabetes‐
related nephropathy and neuropathy, as well as dyslipi­
demia, obstructive sleep apnea, and pregnancy [4–6].
In recent decades, as insulin therapy has become more
widespread, the prevalence of DR has decreased  [7]. It is
noteworthy that intensive insulin therapy may lead to
worsening of retinopathy during the first year following
initiation, yet the long‐term benefits of intensive insulin
treatment significantly outweigh the risks  [8]. This para­
doxical phenomenon of early diabetic retinopathy acceler­
ation following rapid improvement of glycemic control is
well‐established, but does not yet have a conclusive patho­
physiologic explanation. Several theories have been pro­

Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.

© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.

252
 Diagnosis and Management of Ophthalmic Complications of Diabetes 253

TABLE 21.1  Ocular complications of diabetes and treatment options.

Ocular complication Treatment options

Diabetic retinopathy
• Non‐proliferative Observation; control of risk factors*
• Proliferative Pan‐retinal photocoagulation and/or intravitreal
• Diabetic macular edema • Anti‐VEGF injections
• Intravitreal anti‐VEGF injections
• Focal laser
• Intravitreal steroid injections and implants
• Pars plana vitrectomy
• Macular ischemia No established treatment available
Complications of proliferative diabetic retinopathy
• Vitreous hemorrhage Observation, pars plana vitrectomy
• Tractional retinal detachment Pars plana vitrectomy if threatening the macula
• Neovascular glaucoma intravitreal anti‐VEGF injections and/or PRP combined with topical glaucoma
medications; commonly requires incisional glaucoma surgery
Cataract Observation, cataract surgery
Corneal complications
• Dry eye syndrome Lubricating agents, topical cyclosporine A, punctal plugs, autologous serum
tears
• Recurrent epithelial erosions Anterior stromal puncture, phototherapeutic keratectomy
• Corneal edema Hypertonic saline drops
Orbital cellulitis/Mucormycosis Aggressive antifungal therapy with tissue debridement; orbital exenteration

*
 Recommended for all ocular diabetic complications.

posed including changes in intravascular osmotic pressure
leading to small vessel damage, insulin‐induced produc­
tion of reactive oxygen species causing an increase in
VEGF levels, or other potential etiologies for rapid changes
in local cytokine productions, but all hypotheses remain
tentative at present and lack substantial evidence [9].
of undiagnosed metabolic derangement. Up to one‐fifth of
individuals with Type 2 DM already have retinopathy at the
time of diagnosis [12]. Following the initial screening exami­
nation, recommended screening intervals have been out­
lined in several national and international guidelines and
depend on retinopathy severity [13, 14].

Screening TABLE 21.2  Initial screening recommendations for DR in

patients with diabetes. Adapted from [11].
Current guidelines for DR screening differentiate between
the types of diabetes and incorporate several risk factors. Examination by
Studies have shown that in patients with Type 1 DM, vision ophthalmologist or
threatening retinopathy rarely develops in the first 5 years Classification optometrist
following diabetes diagnosis, or before puberty [10]. Thus,
Type 1 diabetes Within 5 years of diagnosis
current screening guidelines recommend an initial compre­ Type 2 diabetes At time of diagnosis
hensive eye exam with dilated funduscopic examination by Women with preexisting Before pregnancy or
an ophthalmologist or optometrist within 5 years of diagno­ diabetes planning pregnancy or during first trimester
sis [11] (Table 21.2). On the other hand, patients with Type 2 who have become pregnant
Women who develop No screening is necessary
DM should have an initial ophthalmic screening exam at the
gestational diabetes
time of diagnosis [11], as these patients may have had years
 254 Management of Disease Complications
Progression of retinopathy is seen in a substantial propor­
tion of diabetic women who become pregnant – between
16% and 85%, depending on the initial stage of DR [5].
The Diabetes in Early Pregnancy Study (DEPS) has
found that both poor glycemic control and increasing
severity of pre‐existing retinopathy elevate the risk of DR
progression [15]. In addition, elevated systolic blood pres­
sure and longer duration of diabetes also appear to contrib­
ute to the risk of worsening retinopathy during
pregnancy [5]. The current guidelines recommend for all
women with diabetes who plan to become pregnant to have
an ophthalmic examination either prior to pregnancy or
during the first trimester [11]. The follow‐up examination
timing is then based upon retinopathy severity. Women
without pre‐existing diabetes who develop gestational dia­
betes mellitus do not have an increased risk of developing
DR and do not require ophthalmologic screening [16].
Prevention
Glycemic Control
Strict glycemic control has remained the mainstay for pre­
vention and slowing progression of diabetic eye complica­
tions. The Diabetes Control and Complications Trial
(DCCT) demonstrated substantially reduced risks of both
DR development and progression in patients with Type 1
DM with tight glycemic control vs conventional glycemic
control (average hemoglobin A1c (HbA1c) of approxi­
mately 7% vs 9%, respectively)  [17]. The EDIC study
(extension of DCCT) has further highlighted the long‐
term benefits of maintaining good glycemic control, with
HbA1c levels being the most important predictors for
development of PDR, diabetic macular edema, and pro­
gression to ocular surgery for diabetic eye complica­
tions  [18]. Similar beneficial outcomes of good glycemic
control on DR were seen in Type 2 DM patients as revealed
by The United Kingdom Prospective Diabetes Study
(UKPDS)  [19] and a more recent Action to Control
Cardiovascular Risk in Diabetes Eye Study (ACCORD) [20].
In most patients, an HbA1c level of 7% is a recommended
target according to current guidelines [11]. However, this
target may not be appropriate for certain patients, and
patient‐specific characteristics such as comorbidities and
life expectancy need to be taken into account.
Blood Pressure Control
In diabetic patients with concomitant hypertension, good
control of blood pressure (BP) is associated with lower risk
of retinopathy. The UKPDS data suggested that maintaining
BP pressure under 150/85 results in 34% and 47% reduction
in worsening of retinopathy and visual acuity, respectively,
when compared to a goal of blood pressure of lower than
180/105 [21]. Subsequent trials, such as Action in Diabetes
and Vascular Disease: Preterax (Perindopril/indapamide)
and Diamicron‐MR (gliclazide) Controlled Evaluation
(ADVANCE) and the ACCORD trial did not demonstrate
the benefit of strict blood pressure control on progression of
DR when comparing systolic blood pressure targets of
120 mmHg vs. 140 mmHg [20, 22]. Current guidelines rec­
ommend optimization of BP in all patients with diabe­
tes  [11], although there is no current level 1 evidence
supporting a blood pressure target lower than 140  with
respect to slowing DR progression.
Lipid‐lowering Strategies
Hyperlipidemia is associated with an increased risk
of  microvascular complications of diabetes. Several
studies have specifically evaluated the role of hyperlipi­
demia and lipid‐lowering medications on diabetic
retinopathy.
Fenofibrate
The results of the Fenofibrate Intervention and Event
Lowering in Diabetes (FIELD) study demonstrated that
fenofibrate (200  mg daily) versus placebo is effective in
reducing the need for laser photocoagulation [23]. A sub­
study of the FIELD participants with fundus photographs
demonstrated a beneficial effect of fenofibrate on slowing
progression of DR in patients with retinopathy at baseline,
as well as on the development of macular edema. Similarly,
the ACCORD Eye Study revealed a reduction in progres­
sion of retinopathy in patients assigned to intensive ther­
apy with fenofibrate versus placebo (6.5% versus 10.2%) in
addition to a statin medication [24]. However, fenofibrate
has not been shown to reduce the risk of cardiovascular
mortality (at least in patients with triglycerides < 400 mg/
dl) resulting in reluctance of general medical practitioners
to prescribe it for patients with DR [11].
 Diagnosis and Management of Ophthalmic Complications of Diabetes 255

Statins low‐up compared to those on a regular diet [30]. Similarly,

In a small prospective trial of 50 diabetic patients who were
randomly assigned to simvastatin daily or to placebo over
6 months, none of the patients in the simvastatin group, but
a substantial proportion of patients in the placebo group
experienced worsening visual acuity [25]. The fluorescein
angiography findings and fundus photographs showed
worsening of DR in the placebo group as compared to the
statin group. In addition, there is evidence that use of
statins in diabetic patients may lower the risk of developing
DME, while hypertriglyceridemia increases such risk [26].
Many patients with DM have lipid abnormalities and are
prescribed a statin to decrease the risk of cardiovascular
disease. Although the role of fenofibrate in slowing DR has
been suggested by two large clinical trials, it is not com­
monly used clinically for this purpose due to the lack of
effect on cardiovascular outcomes. Further research in this
area and collaborations between the ophthalmologists and
general medical physicians to consider this treatment in
patients with DR is encouraged.
another recent study concluded that consumption of ω‐3
polyunsaturated fatty acids (which have previously been
shown to be protective in animal models of diabetic retin­
opathy [31]) decreased the risk of incident sight‐threaten­
ing DR  [32]. Overall, there is a growing, yet limited,
evidence base regarding the impact of diet on DR develop­
ment, and additional research in this area is warranted.
Ocular Complications
Diabetic Retinopathy
Diabetic retinopathy is a primarily microvascular retinal
pathology and is classified based on clinical features
(Table  21.3). The pathogenesis of DR is complex and
involves microvascular damage to pericytes followed be
endothelial cells loss due to hyperglycemia, formation of
free radicals and advanced glycosylation end products as
well as activation of pro‐angiogenic pathways and factors,

Aspirin and anticoagulation TABLE 21.3  International classification of diabetic retinopathy

and diabetic macular edema. Adapted from [13].
The ETDRS trial assessed potential benefits and side‐
effects of high‐dose daily aspirin (650 mg) in patients with Findings observable on dilated
diabetic retinopathy (mild to severe NPDR or early PDR) Disease ophthalmoscopy
versus placebo. The results indicated no increase in the rate
Diabetic retinopathy (DR)
of development of high‐risk PDR or decrease in the risk of
No visible DR No abnormalities
visual loss with aspirin versus placebo, and the rate of vitre­
ous hemorrhage was similar between groups [27]. Similar Mild NPDR Microaneurysms only
findings were observed in a recent systematic review of Moderate NPDR Microaneurysms and other signs (e.g.,
clinical trials looking at high‐dose aspirin therapy in dia­ dot and blot hemorrhages, hard
betic patients [28]. Anticoagulation also did not appear to exudates, cotton wool spots), but less
alter the rate of peri‐operative vitreous hemorrhage in than severe NPDR
patients undergoing vitrectomy for diabetic eye complica­ PDR Severe NPDR plus neovascularization
tions as reported by a recent retrospective study [29].
Diabetic macular edema (DME)

Diet No DME No retinal thickening or hard

There is some early evidence suggesting that diet modifica­ exudates in the macula
tions may be beneficial in preventing DR. Specifically, a Non center‐involving Macular edema that does not involve
nutritional trial assessing effects of a Mediterranean diet DME the central 1 mm subfield
(MedDiet) in patients at high risk for cardiovascular dis­ Center‐involving DME Macular edema involving the central
ease reported that the subset of diabetic patients receiving 1 mm subfield
MedDiet combined with extra virgin olive had an approxi­ NPDR: non proliferative diabetic retinopathy; PDR: proliferative
mately 50% reduced risk of developing DR at 6 year fol­ diabetic retinopathy.
 256 Management of Disease Complications
including a prominent role of vascular endothelial growth
factor (VEGF) [33, 34]. In addition, there is a strong evi­
dence that diabetes leads to stimulation of the supportive
glial cells in the retina, upregulation of pro‐inflammatory
cytokines (such as interleukin‐8 and tumor necrosis fac­
tor‐alpha among others), and activation of the complement
system all of which lead to the establishment of chronic low
grade neuro‐inflammatory state that contributes to the
microvascular pathology [35]. The early stages of DR are
referred to as non‐proliferative DR (NPDR) and are char­
acterized by a number of intraretinal vascular changes.
NPDR is subdivided into mild, moderate, and severe based
on the extent and pattern of retinal changes identified on
examination. Retinal ischemia due to capillary obstruction
may eventually lead to formation of new extraretinal1
blood  vessels, typically accompanied by fibrous tissue
(referred to as neovascularization or fibrovascular prolif­
eration), which is the hallmark of the more severe, vision‐
threatening stage of DR, proliferative DR (PDR). PDR is
further subdivided in early, high‐risk, and advanced.
Fluorescein angiography (FA) is a useful adjunct to clinical
examination in characterizing DR as it allows for more
careful assessment of retinal ischemia and degree and loca­
tion of neovascularization and its regression in response to
the treatment.
The following abnormalities cause vision decline in DR:
●● vascular leakage (macular edema)
●● capillary obstruction (macular ischemia)
●● complications of neovascularization (vitreous hemorrhage,
tractional retinal detachment, neovascular glaucoma).
Importantly, visual decline in DR may not occur until
very advanced stages of retinopathy, thus highlighting the
importance of screening of all diabetic patients. Diabetic
macular edema (discussed later in this chapter) can develop
at any stage of retinopathy, but is more frequently seen in
more severe stages; the development of visually significant
DME is usually what causes patients to seek the initial oph­
thalmologic evaluation. PDR, when left untreated, often
leads to significant visual deterioration.
Macular ischemia is a less common complication of DR
and is usually seen in patients with long‐standing disease.
Its features are decreased capillary density in the fovea (the
center of the macula responsible for the sharpest vision)
with enlargement of the foveal avascular zone and may lead
to vision loss even in the absence of DME or other pathol­
ogy [36]. Currently, there are no specific treatment known
to reverse macular ischemia; management primarily
focuses on the control of DR risk factors.
Treatment of Diabetic Retinopathy
Non‐proliferative Diabetic Retinopathy
NPDR often does not significantly impact vision unless it is
accompanied by diabetic macular edema (see below) or
macular ischemia. Thus, current management of isolated
NPDR primarily involves controlling of systemic risk fac­
tors to prevent progression to more advanced stages of
retinopathy. Some patients with severe NPDR and features
suggesting an increased risk of progression to PDR may
benefit from early treatment with laser panretinal photoco­
agulation (PRP). The decision for PRP treatment in these
cases is based on consideration of various clinical factors
including fellow eye status (especially if fellow eye vision
loss related to sequelae of PDR), inability to or risk for
being lost to follow‐up, poorly controlled diabetes, and
presence of other comorbid conditions among others.
Proliferative Diabetic Retinopathy
Development of PDR places patients at a much higher risk
for visual loss (Figure 21.1). Thus, treatment is initiated for
most patients whose retinopathy has progressed to this
stage. In selected patients with early PDR and low risk of
progression, close observation may be considered if the
patient is motivated to defer treatment and adhere to strict
follow‐up.
Laser Photocoagulation
Pan‐retinal laser photocoagulation has remained the cor­
nerstone of treatment for advanced stages of DR for many
decades. The procedure involves placing 1200 to 1600 laser
burns to the peripheral retina, outside the macula, to
induce thermal damage to ischemic parts of the retina in
order to decrease angiogenic drive (by reducing intravitreal
VEGF levels  [34]) and prevent complications of uncon­
trolled neovascularization.
The Diabetic Retinopathy Study (DRS), completed in
early 1980s, established that eyes receiving treatment with
 Diagnosis and Management of Ophthalmic Complications of Diabetes 257

800

700

600
Retina Thickness [μm]

500

400

300

200

100

FIG 21.1  Diabetic retinopathy and diabetic macular edema. A. Fundus photograph of the left eye of a diabetic patient showing
features of proliferative diabetic retinopathy and diabetic macular edema. There is significant neovascularization of the disc (NVD)
as well as a very large front of abnormal fibrovascular tissue inferotemporally to the disc (neovascularization elsewhere, NVE) and a
small patch of NVE in the superotemporal macula. The patient received prompt pan‐retinal photocoagulation and multiple laser
burns are visible outside the macula. Patient also had center‐involving DME on presentation with numerous hard exudates and
central macular thickening. Features of non‐proliferative DR could also be appreciated with dot‐blot hemorrhages and microaneu-
rysms (arrow heads) scattered throughout the fundus. B. Fluorescein angiography imaging of the same eye shows significant dye
leakage from the sites of NVD and NVE. In addition, areas of retinal ischemia due to capillary non‐perfusion are clearly seen. It can
also be appreciated that neovascularization develops close to the boundaries of ischemic and non‐ischemic areas. C. Optical
coherence tomography (OCT) generated thickness map of the macula of the same patient shows abnormal retinal thickening
which is especially pronounced temporally to the fovea. D. Cross‐sectional optical coherence tomography scan of the macula
through the fovea reveals retinal thickening due to the presence of intra‐retinal fluid (seen as hypo‐reflective cystic areas). Exudates
are also visible as focal areas of signal hyper‐reflectivity. E. Anterior‐segment photograph showing iris neovascularization (NVI) due
to advanced PDR.

PRP experienced 50% lower rate of severe vision loss (SVL)
compared to the untreated eyes at 5‐year follow‐up  [37].
The Early Treatment Diabetic Retinopathy Study (ETDRS)
subsequently provided evidence‐based guidelines for the
timing for PRP initiation (also termed scatter laser). The
ETDRS found that eyes with mild and moderate NPDR
should be observed, while PRP may be considered in eyes
with more severe non‐proliferative diabetic retinopathy
and should not be delayed if retinopathy has advanced to
the proliferative stage with high‐risk features related to the
 258 Management of Disease Complications
presence, location (optic disc versus elsewhere), and sur­
face area of neovascularization, and whether associated
vitreous hemorrhage is present [38].
The PRP procedure is most commonly performed in an
office setting with either topical or local (peri‐ or retrobul­
bar) anesthesia, and may take place over two or more ses­
sions. PRP is associated with a number of potential adverse
effects including pain during the procedure, worsening or
development of macular edema with associated reduction
in visual acuity (usually temporary), decreased peripheral
vision, decrease in dark adaptation, and, rarely, subretinal
neovascularization or vascular occlusions. PRP induces
regression of neovascular activity and often contraction of
fibrovascular tissue. Although this effect is desirable, it
occasionally leads to vitreous hemorrhage and/or develop­
ment or worsening tractional retinal detachment within
the first 1–2  months following treatment. Despite those
adverse effects, clinical evidence demonstrates that the
long‐term, durable effects of PRP in preventing progres­
sion of retinopathy and severe vision loss significantly out­
weigh the associated procedural risks.
Anti‐VEGF
The use of anti‐VEGF agents in diabetic eye disease was
initially focused on treating DME (discussed later in this
chapter). More recently, several clinical trials have revealed
that intravitreal injections of anti‐VEGF halt the progres­
sion of advanced stage of DR with efficiency similar to that
of PRP  [39, 40]. However, for most patients with active
PDR, anti‐VEGF treatment requires multiple injections,
initially with a monthly frequency, which is associated with
significant economic and healthcare system burden. In
addition, interruptions of anti‐VEGF therapy could lead to
rapid worsening of proliferative retinopathy and significant
vision loss [41]. In comparison, PRP often provides long‐
term control of the disease with lower costs and much less
frequent visits for most patients. Each intravitreal injection
also poses a small risk of endophthalmitis, vitreous hemor­
rhage, and retinal detachment (see more detailed discus­
sion later in the section on DME). A particular advantage
of anti‐VEGF therapy over PRP includes less frequent rate
of DME. The initial benefit of greater peripheral visual
field preservation with anti‐VEGF therapies versus PRP
was not sustained through 5‐year follow‐up [40].
Combination Treatment
Given the advantages and disadvantages of both therapies,
treatment of PDR should be individualized based on
patient‐specific factors including the presence of DME, the
overall patient’s health and ability to follow up, and the sta­
tus of the fellow eye among others. At present, the majority
of providers prefer a combination approach utilizing both
PRP and anti‐VEGF in most PDR patients [42]. This prac­
tice does have a clinical trial evidence basis [43]. A careful
explanation of the options and discussion with the patients
regarding their preferences is essential.
Surgical Treatment/Vitrectomy
The current indications for pars plana vitrectomy in eyes
with DR include the presence of vitreous hemorrhage
(VH) that has not cleared over a meaningful period of
observation, progressive tractional retinal detachment
(TRD) that involves or threatens the macula, combined
tractional‐rhegmatogenous retinal detachment, and pro­
gressive neovascularization unresponsive to PRP.
VH arises due to bleeding from the sites of neovascu­
larization as those immature vessels lack endothelial tight
junctions, which predispose them to spontaneous bleed­
ing. The fibrous component that accompanies those fragile
vessels may also put additional stress on them due to its
contraction  [44]. TRDs develop when fibrovascular epi­
centers coalesce into large vitreoretinal tractional bands
that contract and elevate the retina. TRDs may progress to
extensive detachment of the retina, but this is often a slow
process which may be halted with prompt PRP; TRDs that
progress despite medical therapy and encroach on the
macula and the fovea require semi‐urgent vitrectomy to
preserve central vision. Tractional forces from pre‐retinal
fibrous membranes may also cause breaks in the retina
which could result in combined tractional‐rhegmatoge­
nous retinal detachment, which require urgent surgical
intervention due to their threat of rapid progression.
The Diabetic Retinopathy Vitrectomy Study (DRVS)
completed in 1985 found improved visual outcomes in
Type I DM patients with severe vitreous hemorrhage who
were randomized to either early or delayed (1 year) vitrec­
tomy [37]. No clear benefit was seen in Type II diabetics.
However, it is unclear whether those findings remain appli­
cable to contemporary patients due to significant interval
 Diagnosis and Management of Ophthalmic Complications of Diabetes
changes in surgical technologies (e.g. endolaser) and tech­
niques, as well as anti‐VEGF pharmacotherapy. Docu­
mented outcomes of vitrectomy in the ETDRS study
published in 1992 were significant improvement in visual
acuity in diabetic eyes with either vitreous hemorrhage or
retinal detachment; the total rate of eyes requiring this pro­
cedure was around 5% among all recruited diabetic
patients [45].
More recent data shows that with modern vitrectomy
tools and techniques, the vision could be stabilized in
about 80% of patients even with advanced diabetic TRDs
and combined tractional‐rhegmatogenous detach­
ments  [46]. Complications of vitrectomy include forma­
tion or progression of cataract, recurrent VH, retinal
detachment, and endophthalmitis.
Retinal Vein Occlusion
While DM mostly causes retinal microangiopathy, studies
have also revealed an association of DM with intermediate
size retinal vascular pathology, specifically, a higher inci­
dence of branch retinal vein occlusion (BRVO) in diabetic
patients [47]; no such association was seen for central reti­
nal vein occlusion. BRVO can lead to further retinal
ischemia and neovascularization; it can also lead to vision
loss from macular edema, either due to the BRVO itself or
due to exacerbation of pre‐existing DME. The treatment of
neovascularization and macular edema in such eyes is sim­
ilar to that described for PDR and DME.
Neovascular Glaucoma
PDR and other conditions that result in significant ocular
ischemia (such as retinal venous or arterial occlusions,
ocular ischemic syndrome, and radiation retinopathy)
could lead to a fearsome complication of neovascular glau­
coma (NVG). Some earlier studies have estimated that up
to 43% of patient with PDR may develop NVG [48]; how­
ever, the incidence of this disease has substantially
decreased after the advent of intravitreal anti‐VEGF ther­
apy [49]. The hallmark of this type of glaucoma is neovas­
cularization of the iris and of the outflow pathways of the
eye (the iridocorneal angle), resulting in their blockage and
a subsequent severe increase in intraocular pressure. Visual
259
prognosis of NVG is especially poor as severe intraocular
pressure elevations may cause irreversible vision loss at the
time of initial presentation, and because pressure control
may be labile or refractory even with prompt surgical glau­
coma interventions, leading to irreversible visual decline in
many patients [50]. Hence, early detection of neovasculari­
zation of the anterior eye structures and the outflow path­
ways is a crucial part of each diabetic eye exam. Current
management of NVG includes lowering of intraocular
pressure and aggressive control of neovascularization with
a combination of anti‐VEGF agents, PRP, and potentially
incisional glaucoma surgery, all of which are much more
effective at early stages of the disease.
Diabetic Macular Edema
Diabetic macular edema (DME) is one of the main causes
of visual impartment in patients with diabetes.
Epidemiologic studies estimate the overall prevalence of
DME in all patients with diabetes around 7% [1]. DME is
characterized by the presence of intraretinal fluid (edema)
with associated retinal thickening within the macula. The
macula is the central anatomic zone of the retina that is
responsible for high resolution vision within approximately
the 20 central degrees of visual field. The ETDRS trial con­
ducted by NIH in 1990s established the first widely‐used
classification of DME based on clinical funduscopic exami­
nation [51]. It defined the term clinically significant macu­
lar edema (CSME) which was based on the presence and
area of retinal thickening, or hard exudates associated with
thickening, as well as the proximity to the center of the
macula. The term CSME is still used, but has less clinical
relevance due to advancements in both imaging technol­
ogy and pharmacologic treatments. Optical coherence
tomography (OCT) is a non‐invasive test with micrometer
resolution that provides a high‐resolution “optical biopsy”
of the retina in an office setting within a few minutes.
Modern guidelines recommend classifying DME as either
center‐involving (retinal thickening that affects 1mm cen­
tral subfield zone) and non‐center involving (retinal thick­
ening outside that area) DME based on clinical and OCT
criteria [13].
Fluorescein angiography (FA) is another useful test for
evaluating DME. It involves intravenous injection of fluo­
 260 Management of Disease Complications
rescein dye followed by a sequence of fundus photographs
depicting flow of the dye through the retinal vasculature.
FA facilitates assessment of vascular and microvascular reti­
nal abnormalities and visualization of vascular leakage in
areas of blood‐retinal barrier breakdown. The pattern of
leakage seen on FA may help distinguish DME from other
etiologies of macular edema. Leakage on FA, however, does
not always indicate retinal swelling, and correlation of FA
with examination findings and OCT is needed. FA can be
helpful in characterizing whether areas of retinal thickening
are predominantly caused by leaking microaneurysms
(which may benefit from focal laser treatment), or whether
there is more diffuse breakdown of the blood/retinal barrier
that may be more effectively treated pharmacologically.
Treatment of DME
Macular Laser
Focal laser therapy as a treatment for CSME was first evalu­
ated by the ETDRS trial. It demonstrated a 50% or greater
reduction in the risk of moderate vision loss in eyes with
CSME that were treated with focal laser versus eyes that
were observed [51]. However, approximately 40% of eyes
with CSME in this trial had visual acuity of 20/20 or bet­
ter [52]. In the era of current pharmacotherapeutics, new
evidence suggests that observation until vision loss occurs
may be a reasonable option now that better rescue therapy
exists [53].
The current protocol for focal laser treatment of DME is
based on modified ETDRS guidelines and involves applica­
tion of laser burns directly to microaneurysms and micro­
vascular lesions in the center of the rings of hard exudates,
at least 500 microns from the fovea. The procedure usually
takes only a few minutes with topical anesthesia and causes
minimal discomfort. In most cases, the reduction in retinal
edema takes place over weeks or months  [54]. Compli­
cations of focal laser photocoagulation are rare and include
laser scars that increase in size over time causing a blind
spot near the center of the vision (paracentral scotoma),
inadvertent application of laser to the fovea (which may
decrease visual acuity), and choroidal neovascular­iza­tion.
An alternative to focal laser treatment, mild macular
grid (MMG) laser was evaluated in a randomized, prospec­
tive trial [55]. This technique involves application of laser
throughout the macula and not to microaneurysms. While
it was not inferior to the focal laser with respect to visual
acuity outcomes at 12‐month interval, the reduction of
retinal edema was less favorable  [55]. Subthreshold laser
technique (with lower energy delivered via micropulses
such that laser spots are not visible after the application)
has also been evaluated in treating DME and has shown
promising results [56, 57]; however, the data on its efficacy
is currently limited.
Corticosteroid
Corticosteroids can produce rapid improvement in retinal
thickening in patients with DME. However, this effect
comes with the downsides of increased risk of glaucoma
and cataract development. Moreover, reduction in edema
is usually transient and repeated steroid injections are nec­
essary which increases the risk of adverse effects.
Corticosteroid therapies for DME have been evaluated
by multiple comparative trials. Current evidence suggests
that steroid monotherapy is less efficacious than anti‐
VEGF pharmacotherapy when factoring in complications
of treatment. A two‐year trial of intravitreal triamcinolone
acetonide versus focal laser therapy demonstrated better
mean visual acuity in the laser treatment group at two
years [58] and patients receiving triamcinolone had higher
rate of glaucoma and cataract.
Several intravitreal biodegradable implants have been
designed offering sustained low‐dose release of a steroid
medication. These implants are injected intravitreally in an
office setting similar to other intravitreal injections.
Dexamethasone 0.7mg implant (Ozurdex) provides a 3‐ to
4‐month interval of sustained dexamethasone release. In
the 3‐year Macular Edema: Assessment of Implantable
Dexamethasone in Diabetes (MEAD) trial, repeated injec­
tions of this implant were associated with modestly
improved visual acuity as compared to sham injections
(22% of patients gained greater than 15  letters in dexa­
methasone group versus 12% in sham group)  [59].
However, a greater proportion of patients developed cata­
ract and elevated IOP in the dexamethasone implant group.
In another study, addition of Ozurdex to anti‐VEGF ther­
apy in patients with persistent DME did not produce any
additional visual improvement, although anatomical
reduction of retinal edema was enhanced  [60]. Direct
 Diagnosis and Management of Ophthalmic Complications of Diabetes
head‐to‐head comparison of Ozurdex to one of the availa­
ble anti‐VEGF agents (bevacizumab) showed similar visual
gains between the groups at 24 months, with significantly
reduced injection burden in the Ozurdex group  [61].
Similarly to other studies, patients in the Ozurdex group
required cataract surgery more frequently and had higher
IOP.
Fluocinolone acetonide (FA) 0.19 mg implant (Iluvien)
was designed to deliver a low‐dose of FA over extended
period of time, up to 36  months. The Fluocinolone
Acetonide in Diabetic Macular Edema (FAME) study that
evaluated this implant in patients with refractory DME
who had undergone at least one prior macular laser treat­
ment found that a significantly larger proportion of
patients gained more than 15 letters of vision versus sham
implant at 36 months [62]. In current practice, monother­
apy with corticosteroids for DME is uncommonly
employed as initial therapy. It is most commonly utilized
for patients whose DME is refractory to serial anti‐VEGF
injections, or as an adjunctive treatment for incomplete
responders.
Anti‐VEGF Therapy
Pharmacologic inhibition of VEGF via intravitreal injec­
tion of anti‐VEGF agents has rapidly become standard of
care therapy for DME since the 2010 DRCR network
Protocol I trial which demonstrated excellent long‐term
visual outcomes in patients with DME treated with anti‐
VEGF agents as compared to macular laser photocoagula­
tion or intravitreal corticosteroid therapy  [63]. Soon
thereafter, practice patterns shifted toward a significant
increase in intravitreal anti‐VEGF therapy for DME [64].
Current guidelines now recommend anti‐VEGF pharma­
cotherapy as the first line of treatment for DME [13].
Anti‐VEGF inhibitors are usually administered on a
monthly basis for the first few injections. Following that,
one of several clinical extension protocols is usually
employed. In the treat‐and‐extend protocol, as the disease
activity is controlled, patients are evaluated and treated at
increased intervals; if disease activity increases, the treat­
ment intervals are shortened. In the as‐needed or pro re
nata (PRN) protocol, after retinal edema has subsided,
patients are monitored without treatment and injections
are re‐instituted if edema worsens. Recent data demon­
261
strate that a substantial proportion of patients with DME
(up to 50%) may not need additional anti‐VEGF treat­
ments after 3 years of injections [65].
Choice of the agent
Currently, there are four anti‐VEGF agents on the market
that have been studied and used for ocular disease: bevaci­
zumab (Avastin; Roche, Switzerland), ranibizumab
(Lucentis; Novartis, Switzerland), aflibercept (Eylea; Bayer,
Germany), and brolucizumab (Beovu; Novartis,
Switzerland). Ranibizumab and aflibercept are FDA‐
approved for the treatment of DME and DR  [66, 67].
Bevacizumab does not carry an FDA label for treatment of
ocular diseases, yet it is commonly used off‐label by oph­
thalmologists for a variety of conditions including DME
due to its lower cost; when used intravitreally, bevacizumab
is repackaged in a compounding pharmacy into aliquots
containing approximately 1/500th of the dose used in sys­
temic cancer therapy. Brolicuzimab was only recently
approved by the FDA as a treatment of exudative age‐
related macular degeneration (AMD) and has not yet been
studied for DME.
There is a growing literature guiding the choice of
anti‐VEGF agent for treatment of DME. The DRCR
Protocol T clinical trial recently released 2‐year results
comparing efficacy of three anti‐VEGF agents for this
indication  [68]. All three medications offered signifi­
cant improvement in visual acuity at 1 and 2 year time
points. The study suggested that at 2‐year follow‐up all
three agents have similar visual outcomes in eyes with
mild visual impairment (20/32 to 20/40 vision).
Aflibercept showed better visual outcomes over bevaci­
zumab, but not over ranibizumab in eyes with worse ini­
tial visual acuity (20/50 or worse) at 2 year time point.
Bevacizumab therapy was found to be not as effective in
reducing retinal thickening vs two other agents at 2
years. The median number of injections was similar
among the three drugs.
Given a substantial cost difference between FDA‐
approved and off‐label anti‐VEGF agents clinicians are
required to make difficult decisions in selecting a medica­
tion for each patient. Such factors as the cost of care for the
patient or society as a whole, the stage or severity of pathol­
ogy, prior ophthalmic history and the state of the fellow
 262 Management of Disease Complications
eye, as well as published comparative results of treatment
efficacies all play a role.
It is noteworthy, however, that poor response to a par­
ticular anti‐VEGF agent does not indicate a therapeutic
failure of the whole class of the medications and that an
individual patient may have an improved response to a sec­
ond or third drug. Therefore, many clinicians use bevaci­
zumab as an initial treatment (after a discussion of its off‐label
status with the patient) and then switch to more expensive
FDA‐approved agents if there is suboptimal response.
Efficacy over other treatments
There is growing evidence that anti‐VEGF agents are supe­
rior to other forms of treatment of DME. Two head‐to‐
head randomized comparison trials of aflibercept versus
focal laser demonstrated significantly better visual and
anatomic outcomes in the aflibercept group at
52 weeks [67]. In a recent meta‐analysis of 13 clinical trials
comparing aflibercept and ranibizumab with macular laser
treatment, both of these anti‐VEGF agents showed supe­
rior visual outcomes at 12 month timepoint [69]. Patients
who have DME and only mild visual impairment (visual
acuity 20/25 or better) may not require treatment for
extended periods of time and could be monitored. As a
recent prospective randomized trial has suggested, such
patients do equally well with initial observation and rescue
aflibercept therapy if vision decreases when compared to
initial treatment with anti‐VEGF injections or laser ther­
apy over the mean follow‐up of 2 years [53].
Side effects
Regardless of the drug used, intra‐vitreal injections carry a
risk of infectious endophthalmitis (approximately 1  in
3000–5000 injections), vitreous hemorrhage, lens damage,
and retinal detachment. In addition, there is risk of sterile
inflammation with all anti‐VEGF agents that appears to be
slightly higher for aflibercept versus other agents.
A recent study has also brought to light a concern for
potentially increased risk for death and cerebrovascular acci­
dents in those patients with the highest anti‐VEGF injections
exposure (treated with monthly injections for 24 months) [70].
On the other hand, meta‐analysis of trials comparing anti‐
VEGF drugs with either sham or laser photocoagulation for
DME did not show a significant difference regarding the risk
for arterial thromboembolic events or overall mortality [71].
Similarly, a recent retrospective cohort study using a large
U.S. insurance database did not reveal a higher rate of sys­
temic adverse effects (cerebrovascular disease, myocardial
infarction, major bleeding) in patients treated with anti‐
VEGF agents versus those treated with macular laser for
DME [72]. Further investigations are needed to assess long‐
term safety of these medications as some of the patients may
receive ongoing treatment for well over a decade.
Surgery
Pars plana vitrectomy (PPV) as an alternative to other treat­
ments for DME has been evaluated in several studies. DRCR
Protocol D assessed eyes with at least moderate vision loss
from DME and vitreomacular traction [73]. The study was
designed as a prospective cohort study with 87 patients
undergoing PPV as well as additional procedures based on
surgeon’s usual routine (those included epiretinal mem­
brane (ERM) peeling, internal limiting membrane (ILM)
peeling, panretinal photocoagulation and steroid injection
at the end of the procedure). At 6‐month follow‐up central
retinal thickening was decreased by 50% in 68% of patients
with improvement in visual acuity by greater than 10 letters
in 38%. At the same time, visual acuity decreased by more
than 10  letters in 22% of patients. Subsequent subgroup
analysis of the study results revealed that eyes with worse
baseline visual acuity and those that underwent ERM peel­
ing experienced greater visual acuity improvement [74].
However, the results of other trials that examined the
role of vitrectomy for DME were variable with some show­
ing advantage of PPV over macular laser [75], while others
failing to determine significant visual acuity
improvement [76].
A recent systematic review of vitrectomy trials for DME
comparing vitrectomy surgery versus observation or focal
photocoagulation has suggested that PPV does not provide
significant functional advantage over laser; the safety of
PPV for DME was comparable to that of PPV for other
indications [77]. At present, vitrectomy for DME is primar­
ily to address vitreoretinal interface abnormalities that may
exacerbate DME, or less commonly as a last resort for
patients who have failed all medical therapies.
 Diagnosis and Management of Ophthalmic Complications of Diabetes
Future Directions
Sustained intraocular delivery of VEGF‐neutralizing
agents could substantially reduce both the treatment and
financial burden for patients and the health care system.
Current therapies require frequent office visits with associ­
ated imaging, pharmaceutical, professional, and travel
costs. Several recent Phase 2 and Phase 3 trials have exam­
ined various methodologies for such sustained anti‐VEGF
release including port delivery system (a permanent, surgi­
cally implanted, office refillable reservoir for holding a
large pool of anti‐VEGF agent) [78] and gene therapy (with
adeno‐associated virus vector encoding aflibercept)  [79]
among others. Phase 3 trials of some of those systems are
currently underway.
Another method of inhibiting VEGF production is
through the use of small interfering RNA (siRNA). This
approach has shown promising results in a mouse model
when siRNA was delivered to the RPE cells to inhibit
expression of VEFG receptor  [80]. Yet another avenue
explored for targeting VEGF pathway is through sequester­
ing this peptide within the cell. Flt23k intraceptor can bind
to VEGF and prevent it from leaving the endoplasmic
reticulum resulting in effective inhibition. Experiments in
primate and murine models have demonstrated efficient
VEGF pathway inhibition with the delivery of recombinant
Flt23K intraceptor plasmid with a single intravenous injec­
tion [81]. However, these two latter technologies are still in
the research stage and have not yet been translated to
human studies.
There is also ongoing research in non‐molecular treat­
ment approaches for DR and DME. Photobiomodulation
(PBM) has recently shown promising results in treating
DME in preclinical trials [82, 83]. PBM involves applica­
tion of far‐read or near‐infrared light (NIR) (in the 600–
1000 μm range) which allows deep tissue penetration and
may work by inhibiting production of ROS and attenuat­
ing cell death as well as by increasing energy production
by mitochondria  [84]. Currently, the DRCR Retina
Network is recruiting patients with DME for a pilot trial
with a PBM device that used twice per day to deliver 90
seconds of NIR light through closed eyelid. If found effi­
cacious, PBM may become be the first at‐home therapy to
treat DME.
263
DME – Summary
A significant proportion of patients with DM will develop
DME. Current evidence suggests that only center‐involv­
ing DME necessitates treatment. Among patients with CI‐
DME, some patients may retain normal or nearly normal
vision, and new data suggests that such patients could be
observed for at least 2 years as long as visual acuity remains
very good (20/25 or better). For patients with significant
visual decline from DME, intravitreal injections of anti‐
VEGF agents are usually the preferred initial treatment.
Injections are started on a monthly basis with frequent
(usually monthly) follow‐ups; however, following the first
few months those are typically extended. Many patients
may not need treatment after 3 years. A choice of anti‐
VEGF agent is determined by patient‐specific characteris­
tics and cost of care.
A proportion of patients may not have a full response to
the first anti‐VEGF agent, in such cases a trial of a second
and third agent should be considered. Patients with incom­
plete response to anti‐VEGF therapy could be candidates
for focal laser treatment or intra‐vitreal steroid injections
(or intra‐vitreal steroid implants) as adjuncts. Laser treat­
ment can also be used as primary therapy in patients who
are not good candidates for repeat intravitreal injections
due to comorbidities or social factors.
Surgical options for DME include pars plana vitrectomy,
but it is used only rarely and is typically reserved for a select
number of refractory cases. Overall, with the currently
available treatment options, many patients who develop
DME can retain good vision. Future therapies currently
being developed may reduce the patient and healthcare
burden associated with frequent treatments typically
required needed for DME.
Cornea
The cornea is affected by diabetes in a variety of ways col­
lectively known as diabetic keratopathy. Studies have esti­
mated that up 70% of patients with DM may have diabetes
–associated corneal complications  [85]. Diabetic eyes
exhibit weakening of corneal epithelial barrier which may
lead to higher risks of corneal infections. This risk may be
exacerbated by diminished corneal peripheral sensation
 264 Management of Disease Complications
that is also seen in patients with DM [86] and is thought to
arise from reduced small fiber nerve density in diabetic
cornea  [87]. Decreased corneal sensitivity also leads to
higher risk of dry eye syndrome [88]. Other studies have
revealed corneal endothelial dysfunction in DM [89] which
may lead to increase in corneal thickness and edema [90].
Diabetes may also diminish the strength of adhesion of the
epithelium to the underling basement membrane which
may lead to prolonged healing and recurrent corneal ero­
sions and defects [91].
Treatment
The treatment of dry eye syndrome consists of daily appli­
cation of lubricating eye drops and ointments. Additional
therapies include punctal occlusions, topical use of low
doses of immunosuppressive drugs (Cyclosporine A) that
reduce inflammation in the tear glands and improve tear
secretion. Refractory cases may require autologous serum
drops.
Corneal edema can be managed by hypertonic agents,
such as sodium chloride 5% solution eye drops and oint­
ment. They act by drawing the water out of the cornea pre­
venting or reducing the buildup of edema.
Recurrent corneal erosions could be alleviated by ante­
rior stromal puncture technique [92]. In this office‐based
procedure, a small needle is used to create several micro‐
punctures through the epithelium just below the Bowman’s
layer which leads to a strengthened bond between these
corneal layers reducing the risk of erosions. Similar results
could be achieved using laser phototherapeutic keratec­
tomy, a more invasive surgical procedure which utilizes
excimer laser [93].
Cataract
Cataract is an important cause of vision impairment in dia­
betics. In a population‐based study the risk of cataract was
approximately two‐fold higher than in general popula­
tion  [94]; this risk increases with duration of diabetes,
severity of retinopathy and higher HbA1c levels [94, 95].
Visual recovery after uncomplicated cataract extraction
(CE) and intra‐ocular lens implantation surgery in general
depends on presence of DME and stage of DR. A recent
study looking at the development of “treatment‐requiring
DME” after CE over a 2‐year period reported a rate of 1%
for eyes with no DR versus 13% for eyes with severe NPDR
at baseline; the risk of DME development peaked at
3–6 months after surgery [96].
The effect of CE on DR progression is not clear with
some studies suggesting worsening of DR [97], while oth­
ers showing no such association [98, 99]. Universally, stabi­
lization of retinopathy and DME should be accomplished
prior to CE.
Orbital Disease
Patients with uncontrolled diabetes are in general more sus­
ceptible to infections and in particular are at a higher risk for
acute orbital cellulitis. Most common pathogens in orbital
cellulitis are typical sinus bacteria such as Streptococcus
pneumoniae and Haemophilus influenzae. However,
patients with uncontrolled DM, especially those in diabetic
ketoacidosis, are prone to mucormycosis – a rapidly spread­
ing fungal infection of the orbit and sinuses caused by sapro­
phytic fungi of the order Mucorales [100]. Unless diagnosed
early and treated aggressively mucormycosis may lead to
death in a few days. The management of this severe infection
usually requires a combination of aggressive systemic anti­
fungal therapy (such as Amphotericin B) as well as surgical
debridement. In cases of significant orbital invasion, orbital
exenteration (removal of the orbital content) may be
required to control the disease spread. Given its high mortal­
ity, mucormycosis needs to be on the differential list in all
cases of orbital cellulitis in diabetics, especially in those with
significant metabolic derangements.
Note
1. In this context meaning projecting from the retina into the
vitreous cavity.
References
1. Yau JW et al. Global prevalence and major risk factors of
diabetic retinopathy. Diabetes Care. 2012;35(3):556–564.
2. Cheung N, Mitchell P, Wong TY. Diabetic retinopathy.
Lancet. 2010;376(9735):124–136.
3. Klein R et al. The Wisconsin epidemiologic study of diabetic
retinopathy: XXII the twenty‐five‐year progression of
retinopathy in persons with type 1 diabetes. Ophthalmology.
2008;115(11):1859–1868.
 Diagnosis and Management of Ophthalmic Complications of Diabetes
4. Altaf QA et al. Obstructive sleep apnea and retinopathy in
patients with type 2 diabetes. a longitudinal study. Am J
Respir Crit Care Med. 2017;196(7):892–900.
5. Axer‐Siegel R et al. Diabetic retinopathy during pregnancy.
Ophthalmology. 1996;103(11):1815–1819.
6. Zander E et al. Increased prevalence of proliferative retinopathy
and cardiovascular autonomic dysfunction in IDDM patients
with proteinuria. Exp Clin Endocrinol. 1992;99(2):102–107.
7. Lecaire T et al. Lower‐than‐expected prevalence and sever­
ity of retinopathy in an incident cohort followed during the
first 4–14 years of type 1 diabetes: the Wisconsin Diabetes
Registry Study. Am J Epidemiol. 2006;164(2):143–150.
8. Early worsening of diabetic retinopathy in the Diabetes
Control and Complications Trial. Arch Ophthalmol.
1998;116(7):874–886.
9. Bain SC et al. Worsening of diabetic retinopathy with rapid
improvement in systemic glucose control: A review. Diabetes
Obes Metab. 2019;21(3):454–466.
10. Klein R et al. The Wisconsin epidemiologic study of diabetic
retinopathy. II. Prevalence and risk of diabetic retinopathy
when age at diagnosis is less than 30 years. Arch Ophthalmol.
1984;102(4):520–526.
11. Solomon SD et al. Diabetic retinopathy: a position statement
by the American Diabetes Association. Diabetes Care. 2017;
40(3):412–418.
12. Klein R et al. The Wisconsin epidemiologic study of diabetic
retinopathy. III. Prevalence and risk of diabetic retinopathy
when age at diagnosis is 30 or more years. Arch Ophthalmol.
1984;102(4):527–532.
13. Wong TY et  al. Guidelines on Diabetic Eye Care: The
International Council of Ophthalmology Recommenda­tions
for Screening, Follow‐up, Referral, and Treatment Based on
Resource Settings. Ophthalmology. 2018;125(10):1608–1622.
14. International Council of Ophthalmology. Updated 2017
ICO Guidelines for Diabetic Eye Care. 2017. San Francisco,
CA: ICO.
15. Chew EY et al. Metabolic control and progression of retin­
opathy. The Diabetes in Early Pregnancy Study. National
Institute of Child Health and Human Development Diabetes
in Early Pregnancy Study. Diabetes Care. 1995;18(5):
631–637.
16. Gunderson EP et  al. A 20‐year prospective study of child­
bearing and incidence of diabetes in young women, control­
ling for glycemia before conception: the Coronary Artery
Risk Development in Young Adults (CARDIA) Study.
Diabetes. 2007;56(12):2990–2996.
17. Diabetes C et al. The effect of intensive treatment of diabetes
on the development and progression of long‐term complica­
265
tions in insulin‐dependent diabetes mellitus. N Engl J Med.
1993;329(14):977–986.
18. Hainsworth DP et al. Risk factors for retinopathy in type 1
diabetes: the DCCT/EDIC Study. Diabetes Care. 2019;42(5):
875–882.
19. Stratton IM et al. Association of glycaemia with macrovas­
cular and microvascular complications of type 2 diabetes
(UKPDS 35): prospective observational study. BMJ. 2000;
321(7258):405–412.
20. Group AS et al. Effects of medical therapies on retinopathy
progression in type 2 diabetes. N Engl J Med. 2010;
363(3):233–244.
21. UK Prospective Diabetes Study Group. Tight blood pressure
control and risk of macrovascular and microvascular com­
plications in type 2 diabetes: UKPDS 38. BMJ. 1998;317(7160):
703–713.
22. Beulens JW et  al. Effects of blood pressure lowering and
intensive glucose control on the incidence and progression
of retinopathy in patients with type 2 diabetes mellitus: a
randomised controlled trial. Diabetologia. 2009;52(10):
2027–2036.
23. Keech AC et  al. Effect of fenofibrate on the need for laser
treatment for diabetic retinopathy (FIELD study): a
randomised controlled trial. Lancet. 2007;370(9600):
­
1687–1697.
24. Chew EY et al. The effects of medical management on the
progression of diabetic retinopathy in persons with type 2
diabetes: the Action to Control Cardiovascular Risk in
Diabetes (ACCORD) Eye Study. Ophthalmology. 2014;
121(12):2443–2451.
25. Sen K et al. Simvastatin retards progression of retinopathy in
diabetic patients with hypercholesterolemia. Diabetes Res
Clin Pract. 2002;56(1):1–11.
26. Chung YR et al. Association of statin use and hypertriglyc­
eridemia with diabetic macular edema in patients with type
2 diabetes and diabetic retinopathy. Cardiovasc Diabetol.
2017;16(1):4.
27. Early Treatment Diabetic Retinopathy Study Research Group.
Effects of aspirin treatment on diabetic retinopathy. ETDRS
report number 8. Ophthalmology. 1991;98(5 Suppl):757–765.
28. Bergerhoff K, Clar C, Richter B. Aspirin in diabetic retinopa­
thy. A systematic review. Endocrinol Metab Clin North Am.
2002;31(3):779–793.
29. Brown JS, Mahmoud TH, Anticoagulation and clinically sig­
nificant postoperative vitreous hemorrhage in diabetic vit­
rectomy. Retina. 2011;31(10):1983–1987.
30. Diaz‐Lopez A et  al. Mediterranean diet, retinopathy,
nephropathy, and microvascular diabetes complications: a
 266 Management of Disease Complications
post hoc analysis of a randomized trial. Diabetes Care.
2015;38(11):2134–2141.
31. Sapieha P et  al. Omega‐3 polyunsaturated fatty acids pre­
serve retinal function in type 2 diabetic mice. Nutr Diabetes.
2012;2:e36.
32. Sala‐Vila A et  al. Dietary marine omega‐3 fatty acids and
incident sight‐threatening retinopathy in middle‐aged and
older individuals with type 2 diabetes: prospective investiga­
tion from the PREDIMED trial. JAMA Ophthalmol. 2016;
134(10):1142–1149.
33. Wang W, Lo ACY. Diabetic retinopathy: pathophysiology
and treatments. Int J Mol Sci. 2018;19(6).
34. Aiello LP et al. Vascular endothelial growth factor in ocular
fluid of patients with diabetic retinopathy and other retinal
disorders. N Engl J Med. 1994;331(22):1480–1487.
35. Rubsam A, Parikh S, Fort PE. Role of inflammation in dia­
betic retinopathy. Int J Mol Sci. 2018;19(4).
36. Dupas B et al. Association between vessel density and visual
acuity in patients with diabetic retinopathy and poorly con­
trolled type 1 diabetes. JAMA Ophthalmol. 2018;136(7):
721–728.
37. The Diabetic Retinopathy Study Research Group.
Photocoagulation treatment of proliferative diabetic retin­
opathy. Clinical application of Diabetic Retinopathy Study
(DRS) findings, DRS Report Number 8. Ophthalmology.
1981;88(7):583–600.
38. Early Treatment Diabetic Retinopathy Study Research
Group. Early photocoagulation for diabetic retinopathy.
ETDRS report number 9. Ophthalmology. 1991;98(5 Suppl):
766–785.
39. Sivaprasad S et al. Clinical efficacy of intravitreal aflibercept
versus panretinal photocoagulation for best corrected visual
acuity in patients with proliferative diabetic retinopathy at
52  weeks (CLARITY): a multicentre, single‐blinded, ran­
domised, controlled, phase 2b, non‐inferiority trial. Lancet.
2017;389(10085):2193–2203.
40. Gross JG et al. Five‐year outcomes of panretinal photocoag­
ulation vs intravitreous ranibizumab for proliferative dia­
betic retinopathy: a randomized clinical trial. JAMA
Ophthalmol. 2018;136(10):1138–1148.
41. Wubben TJ, Johnson MW, Anti‐VEGF Treatment
Interruption Study Group. Anti, anti‐vascular endothelial
growth factor therapy for diabetic retinopathy: conse­
quences of inadvertent treatment interruptions. Am J
Ophthalmol. 2019;204:13–18.
42. American Society of Retina Specialists. Preferences and
Trends Survey 2019. 2019.
43. Figueira J et al. Ranibizumab plus panretinal photocoagula­
tion versus panretinal photocoagulation alone for high‐risk
proliferative diabetic retinopathy (PROTEUS Study).
Ophthal­mology. 2018;125(5):691–700.
44. Garner A, Histopathology of diabetic retinopathy in man.
Eye (Lond). 1993;7(Pt 2):250–253.
45. Flynn HW, Jr. et  al, The Early Treatment Diabetic
Retinopathy Study Research Group. Pars plana vitrec­
tomy in the Early Treatment Diabetic Retinopathy Study.
ETDRS report number 17. Ophthalmology. 1992;99(9):
1351–1357.
46. Storey PP et al. Visual and anatomical outcomes after dia­
betic traction and traction‐rhegmatogenous retinal detach­
ment repair. Retina. 2018;38(10):1913–1919.
47. Klein R et al. The epidemiology of retinal vein occlusion: the
Beaver Dam Eye Study. Trans Am Ophthalmol Soc.
2000;98:133–141; discussion 141–143.
48. Madsen PH. Rubeosis of the iris and haemorrhagic glau­
coma in patients with proliferative diabetic retinopathy. Br J
Ophthalmol. 1971;55(6):368–371.
49. Al‐Bahlal A et  al. Changing epidemiology of neovascular
glaucoma from 2002 to 2012 at King Khaled Eye Specialist
Hospital, Saudi Arabia. Indian J Ophthalmol. 2017;65(10):
969–973.
50. Shazly TA, Latina MA. Neovascular glaucoma: etiology,
diagnosis and prognosis. Semin Ophthalmol. 2009;24(2):
113–121.
51. Early Treatment Diabetic Retinopathy Study Research
Group. Photocoagulation for diabetic macular edema. Early
Treatment Diabetic Retinopathy Study report number 1.
Arch Ophthalmol. 1985;103(12):1796–1806.
52. Early Treatment Diabetic Retinopathy Study design and
baseline patient characteristics. ETDRS report number 7.
Ophthalmology. 1991;98(5 Suppl):741–756.
53. Baker CW et  al. Effect of initial management with
Aflibercept vs laser photocoagulation vs observation on
vision loss among patients with diabetic macular edema
involving the center of the macula and good visual acuity:
a randomized clinical trial. JAMA. 2019;321(19):1880–
1894.
54. Diabetic Retinopathy Clinical Research Network. The
course of response to focal/grid photocoagulation for dia­
betic macular edema. Retina. 2009;29(10):1436–1443.
55. Writing Committee for the Diabetic Retinopathy Clinical
Research Network et al. Comparison of the modified Early
Treatment Diabetic Retinopathy Study and mild macular
grid laser photocoagulation strategies for diabetic macular
edema. Arch Ophthalmol. 2007;125(4):469–480.
56. Friberg TR, Infrared micropulsed laser treatment for dia­
betic macular edema – subthreshold versus threshold
lesions. Semin Ophthalmol. 2001;16(1):19–24.
 Diagnosis and Management of Ophthalmic Complications of Diabetes
57. Vujosevic S et  al. Microperimetry and fundus autofluores­
cence in diabetic macular edema: subthreshold micropulse
diode laser versus modified early treatment diabetic retin­
opathy study laser photocoagulation. Retina. 2010;30(6):
908–916.
58. Diabetic Retinopathy Clinical Research Network. A rand­
omized trial comparing intravitreal triamcinolone acetonide
and focal/grid photocoagulation for diabetic macular edema.
Ophthalmology. 2008;115(9):1447–1449, 1449 e1–10.
59. Boyer DS et  al. Three‐year, randomized, sham‐controlled
trial of dexamethasone intravitreal implant in patients with
diabetic macular edema. Ophthalmology. 2014;121(10):
1904–1914.
60. Maturi RK et al. Effect of adding dexamethasone to contin­
ued ranibizumab treatment in patients with persistent dia­
betic macular edema: A DRCR network phase 2 randomized
clinical trial. JAMA Ophthalmol. 2018;136(1):29–38.
61. Fraser‐Bell S et al. Bevacizumab or dexamethasone implants
for DME: 2‐year results (the BEVORDEX study).
Ophthalmology. 2016;123(6):1399–1401.
62. Campochiaro PA et al. Long‐term benefit of sustained‐deliv­
ery fluocinolone acetonide vitreous inserts for diabetic mac­
ular edema. Ophthalmology. 2011;118(4):626–635 e2.
63. Diabetic Retinopathy Clinical Research Network et  al.
Randomized trial evaluating ranibizumab plus prompt or
deferred laser or triamcinolone plus prompt laser for dia­
betic macular edema. Ophthalmology. 2010;117(6):1064–
1077 e35.
64. Parikh R et  al. Trends of anti‐vascular endothelial growth
factor use in ophthalmology among privately insured and
medicare advantage patients. Ophthalmology. 2017;124(3):
352–358.
65. Elman MJ et al. Intravitreal ranibizumab for diabetic macu­
lar edema with prompt versus deferred laser treatment: 5‐
year randomized trial results. Ophthalmology. 2015;122(2):
375–381.
66. Brown DM et al. Long‐term outcomes of ranibizumab ther­
apy for diabetic macular edema: the 36‐month results from
two phase III trials: RISE and RIDE. Ophthalmology.
2013;120(10):2013–2022.
67. Korobelnik JF et al. Intravitreal aflibercept for diabetic mac­
ular edema. Ophthalmology. 2014;121(11):2247–2254.
68. Wells JA et al. Aflibercept, bevacizumab, or ranibizumab for
diabetic macular edema: two‐year results from a compara­
tive effectiveness randomized clinical trial. Ophthalmology.
2016;123(6):1351–1359.
69. Muston D et  al. An efficacy comparison of anti‐vascular
growth factor agents and laser photocoagulation in diabetic
macular edema: a network meta‐analysis incorporat­
267
ing  individual patient‐level data. BMC Ophthalmol. 2018;
18(1):340.
70. Avery RL, Gordon GM. Systemic safety of prolonged
monthly anti‐vascular endothelial growth factor therapy for
diabetic macular edema: a systematic review and meta‐anal­
ysis. JAMA Ophthalmol. 2016;134(1):21–29.
71. Virgili G et  al. Anti‐vascular endothelial growth factor for
diabetic macular oedema. Cochrane Database Syst Rev.
2014(10):CD007419.
72. Maloney MH et al. Risk of systemic adverse events associ­
ated with intravitreal anti‐VEGF therapy for diabetic macu­
lar edema in routine clinical practice. Ophthalmology.
2019;126(7):1007–1015.
73. Diabetic Retinopathy Clinical Research Network Writing
Committee et al. Vitrectomy outcomes in eyes with diabetic
macular edema and vitreomacular traction. Ophthalmology.
2010;117(6):1087–1093.e3.
74. Flaxel CJ et  al. Factors associated with visual acuity out­
comes after vitrectomy for diabetic macular edema: diabetic
retinopathy clinical research network. Retina. 2010;30(9):
1488–1495.
75. Yanyali A et  al. Modified grid laser photocoagulation versus
pars plana vitrectomy with internal limiting membrane removal
in diabetic macular edema. Am J Ophthalmol. 2005;139(5):
795–801.
76. Cho M, D’Amico DJ. Transconjunctival 25‐gauge pars
plana vitrectomy and internal limiting membrane peeling
for chronic macular edema. Clin Ophthalmol. 2012;6:
981–989.
77. Jackson TL et al. Pars plana vitrectomy for diabetic macular
edema: a systematic review, meta‐analysis, and synthesis of
safety literature. Retina. 2017;37(5):886–895.
78. Campochiaro PA et al. The port delivery system with ranibi­
zumab for neovascular age‐related macular degeneration:
results from the randomized phase 2  ladder clinical trial.
Ophthalmology. 2019;126(8):1141–1154.
79. Grishanin R et  al. Preclinical evaluation of ADVM‐022, a
novel gene therapy approach to treating wet age‐related
macular degeneration. Mol Ther. 2019;27(1):118–129.
80. Liu HA et  al. A lipid nanoparticle system improves
siRNA efficacy in RPE cells and a laser‐induced murine
CNV model. Invest Ophthalmol Vis Sci. 2011;52(7):4789–
4794.
81. Luo L et  al. Targeted intraceptor nanoparticle therapy
reduces angiogenesis and fibrosis in primate and murine
macular degeneration. ACS Nano. 2013;7(4):3264–3275.
82. Eells J et  al. 670  nm photobiomodulation as a therapy for
diabetic macular edema: a pilot study. Investigative
Ophthalmology & Visual Science. 2017;58(8):932–932.
 268 Management of Disease Complications
83. Tang J, Herda AA, Kern TS. Photobiomodulation in the
treatment of patients with non‐center‐involving diabetic
macular oedema. Br J Ophthalmol. 2014;98(8):1013–1015.
84. Geneva, II. Photobiomodulation for the treatment of reti­
nal diseases: a review. Int J Ophthalmol. 2016;9(1):
145–152.
85. Herse PR. A review of manifestations of diabetes mellitus in
the anterior eye and cornea. Am J Optom Physiol Opt. 1988;
65(3):224–230.
86. Dogru M, Katakami C, Inoue M. Tear function and ocular
surface changes in noninsulin‐dependent diabetes mellitus.
Ophthalmology. 2001;108(3):586–592.
87. Messmer EM et al. in vivo; confocal microscopy of corneal
small fiber damage in diabetes mellitus. Graefes Arch Clin
Exp Ophthalmol. 2010;248(9):1307–1312.
88. Zhang X et al. Dry eye syndrome in patients with diabetes
mellitus: prevalence, etiology, and clinical characteristics. J
Ophthalmol. 2016; 2016:8201053.
89. Saini JS, Mittal S. in  vivo; assessment of corneal endothelial
function in diabetes mellitus. Arch Ophthalmol. 1996;114(6):
649–653.
90. Su DH et  al. Diabetes, hyperglycemia, and central corneal
thickness: the Singapore Malay Eye Study. Ophthalmology.
2008;115(6):964–968 e1.
91. Foulks GN et al. Factors related to corneal epithelial compli­
cations after closed vitrectomy in diabetics. Arch Ophthalmol.
1979;97(6):1076–1078.
92. Cormier G et  al. Anterior stromal punctures for bullous
keratopathy. Arch Ophthalmol. 1996;114(6):654–658.
  93. O’Brart DP, Muir MG, Marshall J. Phototherapeutic kera­
tectomy for recurrent corneal erosions. Eye (Lond). 1994;
8(Pt 4):378–383.
  94. Becker C et al. Cataract in patients with diabetes mellitus‐
incidence rates in the UK and risk factors. Eye (Lond).
2018;32(6):1028–1035.
  95. Klein BE, Klein R, Moss SE. Prevalence of cataracts in a
population‐based study of persons with diabetes mellitus.
Ophthalmology. 1985;92(9):1191–1196.
 96. Denniston AK et  al. The UK Diabetic Retinopathy
Electronic Medical Record (UK DR EMR) Users Group,
Report 2: real‐world data for the impact of cataract sur­
gery on diabetic  macular oedema. Br J Ophthalmol.
2017;101(12):1673–1678.
  97. Chew EY et al. Results after lens extraction in patients with
diabetic retinopathy: early treatment diabetic retinopathy
study report number 25. Arch Ophthalmol. 1999;117(12):
1600–1606.
  98. Squirrell D et al. A prospective, case controlled study of the
natural history of diabetic retinopathy and maculopathy
after uncomplicated phacoemulsification cataract surgery
in patients with type 2 diabetes. Br J Ophthalmol.
2002;86(5):565–571.
  99. Wagner T et  al. Influence of cataract surgery on the dia­
betic eye: a prospective study. Ger J Ophthalmol. 1996;5(2):
79–83.
100. Dokmetas HS et al. Diabetic ketoacidosis and rhino‐orbital
mucormycosis. Diabetes Res Clin Pract. 2002;57(2):
139–142.
 22 Upper Gastrointestinal Manifestations
of Diabetes
Michael Camilleri
Clinical Enteric Neuroscience Translational and Epidemiological Research Program, Division of Gastroenterology and
Hepatology, Mayo Clinic, Rochester, MN, USA
LE A R NI NG P OI NT S
●● Gastrointestinal manifestations of diabetes typically
accompany other microvascular complications.
●● Gastroparesis is associated with increased mortality in
diabetes.
Nutritional support is a key part of the management
●●
of gastroparesis.
Introduction
Although diabetes can affect the entire gastrointestinal
tract, most of the impact results from upper gastrointesti­
nal dysfunction, which may affect quality of life, impair
nutrition, and affect glycemic control.
Epidemiology
Studies in selected patient groups, often from tertiary refer­
ral centers, suggest that gastrointestinal symptoms are com­
mon in diabetes mellitus [1, 2]. In an epidemiological study
from Olmsted County, Minnesota, [3] the prevalence of gas­
trointestinal symptoms (i.e., nausea and/or vomiting, dys­
pepsia, heartburn, irritable bowel syndrome, constipation,
and fecal incontinence) was not significantly different
between individuals with either type 1 (T1DM) or type 2
(T2DM) diabetes mellitus and age‐matched controls in
whom it is known that there is a high background popula­
tion prevalence (~20%) of symptoms suggestive of gastroin­
Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
testinal dysfunction  [4]; however, people with T2DM and
men with T1DM used laxatives more frequently than con­
trols. Moreover, that study and a Finnish population‐based
study reported that people with T1DM had a lower preva­
lence of heartburn [5].
In contrast, a study from Australia reported the preva­
lence of several upper and lower gastrointestinal symptoms
was higher in 423 patients with predominantly (95%)
T2DM than in controls [6].
In Olmsted County, Minnesota, the age‐adjusted inci­
dence per 100  000 person‐years of definite gastroparesis
for the years 1996–2006 was 2.4 (95% confidence interval
(CI) 1.2–3.8) for men and 9.8 (95% CI 7.5–12.1) for
women. The age‐adjusted prevalence of definite gastropa­
resis per 100 000 persons on January 1, 2007, was 9.6 (95%
CI 1.8–17.4) for men and 37.8 (95% CI 23.3–52.4) for
women [7].
Diabetic gastroparesis may cause severe symptoms and
result in nutritional compromise, impaired glucose control
and a poor quality of life, independently of other factors
such as age, tobacco use, alcohol use or type of diabetes [8].
Studies of the natural history of gastroparesis have been
limited by relatively small numbers of patients, potential
referral bias or short follow‐up periods. The data suggest
that gastric emptying and its symptoms are generally stable
during 12–25 years of follow‐up or more [9, 10]. In a study
of 86 patients with diabetes who were followed for at least 9
years, gastroparesis was not associated with mortality after
adjustment for other disorders [11]. Among patients with
gastroparesis, overall survival was significantly lower than
269
 270 Management of Disease Complications
the age‐ and sex‐specific expected survival computed from
the Minnesota white population  [7]. The most common
causes of death were cardiovascular disease (24.6%), res­
piratory failure (23.2%), malignancy (15.9%), chronic renal
failure in 11 (15.9%), cerebrovascular accident (10.1%),
and other causes (10.1%).
Delayed radionuclide gastric emptying predicts morbid­
ity in patients with diabetes with gastroparesis symptoms,
including hospitalizations and days of hospitalizations,
emergency department and office visits  [12]. Over one
decade (1996–2006), gastroparesis was associated with
increased mortality  [7], although the cause of death was
not gastrointestinal and the gastroparesis was simply a
marker of the severity of the diabetes and the associated
complications, the classical triopathy of nephropathy, neu­
ropathy and retinopathy [13]. More recent studies confirm
that delayed gastric emptying is associated with retinopa­
thy and a number of complications of diabetes; however,
with so many patients having type 2 diabetes rather than
insulin‐dependent and ketosis‐prone type 1 diabetes, 39%
of diabetic patients with gastroparesis in the NIH
Gastroparesis Consortium database did not have any dia­
betic complications [14].
Mechanisms and pathophysiology
underpinning upper gastrointestinal
symptoms in diabetes mellitus
Gastrointestinal dysmotility in diabetes is caused by extrin­
sic (i.e., sympathetic and parasympathetic) neural dysfunc­
tion, intrinsic or enteric nervous system (excitatory and
inhibitory neurons) or pacemaker dysfunction (interstitial
cells of Cajal or PDGFRα‐positive fibroblast‐like cells),
and hyperglycemia  [15]. Neural dysfunction has been
attributed to several mechanisms (e.g., oxidative stress) [16]
due to loss of anti‐inflammatory macrophages and
increased expression of genes associated with pro‐inflam­
matory macrophages that have been reported in full‐thick­
ness gastric biopsies from patients with gastroparesis [17].
Pathophysiology of diabetic gastroparesis:
insights from animal studies
In animal models, extrinsic neural dysfunction has been
primarily implicated to a loss of myelinated and
unmyelinated fibers, without much neuronal loss [18, 19].
The loss of nerve fibers is often multifocal, suggestive of
ischemic injury. Within the enteric nervous system,
reduced neuronal staining and, to a lesser extent, neuronal
loss, particularly inhibitory neurons expressing nitric oxide
synthase (NOS), have been described in several animal
models of diabetes [20]. In theory, this reduction in nitrer­
gic inhibitory functions may contribute to impaired gastric
accommodation and accelerated intestinal transit in diabe­
tes. Loss of NOS may impair pyloric relaxation and thereby
retard gastric emptying. Loss of intestinal cells of Cajal
(ICCs), documented in several animal models and case
reports of diabetes, may also contribute to gut dysmotil­
ity [15, 20].
Several mechanisms, including apoptosis, oxidative
stress (possibly associated with reduced CD 206 positive
M2  macrophages), advanced glycation end products and
neuroimmune mechanisms may be responsible for neu­
ronal loss and gut dysmotility  [16]. The loss of ICCs has
been attributed to a reduction in heme‐oxygenase
(HO‐1)  and other protective mechanisms against hyper­
glycemia [15].
Pathophysiology of gastric dysfunctions
in humans with diabetes
Diabetes is associated with accelerated or delayed gastric
emptying, increased and reduced gastric sensation, and
impaired gastric accommodation (Figure  22.1). A vagal
neuropathy can cause antral hypomotility and/or pyloros­
pasm, which may delay gastric emptying [21]. Among 108
patients with diabetes who underwent tests to evaluate
both gastric emptying of solids and gastric accommodation
using SPECT imaging, there were roughly equal propor­
tions of patients with abnormal gastric emptying, abnor­
mal gastric accommodation, and both abnormal or neither
abnormal; the latter presumably reflecting gastroduodenal
hypersensitivity [22]. Gastric emptying may be accelerated
in diabetic patients with upper gastrointestinal symp­
toms [23]. The pathophysiology of rapid gastric emptying
in diabetes may result from impaired gastric accommoda­
tion resulting from a vagal neuropathy  [24], thereby
increasing gastric pressure and accelerating gastric empty­
ing of liquids. Some patients with diabetes and gastropare­
sis also have small intestinal dysmotility, more frequently
 Liquids or solids in stomach (% total consumed)
100

Gastroparesis
Gastric emptying
of solids
Gastric
emptying of
50 homogenized
solids

Gastric
emptying
of liquids

0 1 2 3 4 6
Hours

0h 2h 4h
normal
0% 57% 100% emptied

slow 0h 2h 4h
0% 20% 71% emptied

0h 1h 2h
fast
0% 52% 98% emptied

FIG 22.1  Assessment of gastric emptying by scintigraphy. Normally liquids are emptied in an exponential manner while solid
emptying is characterized by an initial lag phase, followed by a more rapid, linear emptying (left panel). The lag phase corresponds
to the time required for antral trituration and gastric accommodation, during which approximately 10% of solids are emptied.
Representative examples of normal, delayed and accelerated gastric emptying respectively of an egg meal, labeled with 99mTc, in
patients with diabetes are shown in the right panel. At time 0 (i.e. first image in each panel), the entire meal was in the stomach.
Thereafter, the normal ranges for gastric emptying are 4.4–35% at 1 hour, 25–78.5% at 2 hours and 76.2–100% at 4 hours.
 272 Management of Disease Complications
characterized by reduced than by increased motility [25].
Small bowel dysmotility may also contribute to gastric
stasis.
Whereas acute hyperglycemia delays gastric emptying
in healthy subjects and in patients with T1DM  [26–29],
the clinical significance of hyperglycemia in terms of
symptoms and morbidity, or the reversibility of
delayed emptying by optimizing control of glycemia are
the topic  of continuing research without unequivocal
answers [30, 31].
Upper gastrointestinal manifestations
associated with diabetes
Dysphagia and heartburn
Clinical features
Esophageal dysmotility is common, may cause dysphagia,
and may be related to cardiovascular autonomic neuropa­
thy in diabetes mellitus [32]. The amplitude of peristaltic
contractions and basal lower esophageal sphincter pres­
sures are generally normal, and dysphagia is mostly associ­
ated with incoordinated contractions. Symptoms of
gastroesophageal reflux are also common, particularly in
patients with impaired gastric emptying who have vomit­
ing. Rarely, recurrent vomiting may lead to Mallory–Weiss
tears and bleeding.
Diagnosis
Dysphagia and heartburn should prompt upper gastroin­
testinal endoscopy to exclude reflux and other incidental
mucosal diseases, such as candidiasis and neoplasms.
Because of the high prevalence of coronary atherosclerosis
in diabetes, testing for coronary artery disease should be
considered when necessary in patients with chest pain.
Gastroparesis and other upper gastrointestinal
symptoms
Clinical features
Although gastroparesis refers to a syndrome characterized
by symptoms of nausea, vomiting, early satiation after
meals, impaired nutrition and objective evidence of mark­
edly delayed gastric emptying, gastric retention may be
asymptomatic  [14], perhaps because of the afferent dys­
function associated with vagal denervation [33].
Whereas postprandial symptoms without delayed gas­
tric emptying were previously called diabetic dyspepsia in
which vomiting was less prevalent, appreciation that the
pathophysiology includes delayed gastric emptying,
impaired gastric accommodation and abnormal sensation,
the term “gastroparesis” has been extended to include
patients with normal gastric emptying and a significant
component of upper abdominal pain, which is associated
with opioid use and impaired quality of life  [34–37].
Severely delayed gastric emptying of less than 65% at 4
hours reflects a significant delay often associated with
nutritional consequences, the need for nutritional supple­
mentation, jejunal feeding or gastric decompression [38].
Patients with diabetic gastroparesis frequently have
long‐standing T1DM with other microvascular complica­
tions including retinopathy, nephropathy, peripheral neu­
ropathy and other forms of autonomic dysfunction  [13].
These can present as abnormal pupillary responses, anhi­
drosis, gustatory sweating, orthostatic hypotension, impo­
tence, retrograde ejaculation and dysfunction of the
urinary bladder. In patients with diabetes, delayed gastric
emptying is associated with retinopathy and a number of
complications of diabetes [14].
Clinical clues to the presence of gastroparesis include
vomiting of undigested food eaten several hours previ­
ously, weight loss, gastric succussion splash or features of
an autonomic neuropathy.
Diagnosis
In patients with upper gastrointestinal symptoms, an upper
gastrointestinal endoscopy is necessary to exclude peptic
ulcer disease and neoplasms, either of which can cause gas­
tric outlet obstruction. Upper endoscopy may reveal gas­
tric bezoars, which suggest antral hypomotility. Metabolic
derangements such as diabetic ketoacidosis or uremia, and
medications, particularly opiates, calcium‐channel block­
ers and anticholinergic agents, may contribute to dysmotil­
ity. Rarely, patients with gastroparesis present with
retrosternal or epigastric pain, and cardiac, biliary or pan­
creatic disease may be considered.
Upper gastrointestinal endoscopy or barium meal is
essential to exclude mechanical obstruction of the stomach
outflow. Barium x‐rays of the small intestine or enterogra­
phy with computed tomography or MRI should be consid­
 Upper Gastrointestinal Manifestations of Diabetes
ered only when the clinical features raise the possibility of
small intestinal obstruction.
Whereas, autonomic function testing is useful  [39] to
indicate the underlying severity of neuropathy, delayed
gastric emptying can be documented by scintigraphy or
stable isotope breath tests for solids, by ultrasonography for
liquids, and by the presence of a large amount of retained
food in the stomach. The proportion of radioisotope
retained in the stomach at 2 and 4 hours distinguishes nor­
mal function from delayed gastric emptying with a sensi­
tivity of 90% and a specificity of 70% [40]. The importance
of obtaining scans at 4 hours after a meal cannot be over­
emphasized. Because gastric emptying is slow initially, it is
not accurate to extrapolate emptying from scans taken for
a shorter duration. Another useful test for measuring solid
phase gastric emptying utilizes a standardized meal with
biscuit enriched with 13C, a substrate containing the stable
isotope. When metabolized, the proteins, carbohydrates
and lipids of the S. platensis or the medium chain triglycer­
ide, octanoate, give rise to respiratory CO2 that is enriched
in 13C. Measurement of 13CO2 breath content (a reflection
of the amount of biscuit remaining in the stomach) by iso­
tope ratio mass spectrometry allows an estimation of gas­
tric emptying t1/2  [41]. Antropyloroduodenal manometry
and EndoFLIP are specialized techniques that assess pres­
sure profiles in the stomach and small bowel, as well as the
dimensions and distensibility of the pylorus. They may
play a greater role in the future to guide management, such
as selection of prokinetics versus performance of G‐POEM
procedure.
Gastric accommodation in response to meal ingestion
may be impaired in diabetes [24]. This may contribute to
the gastrointestinal symptoms of nausea, bloating and early
satiety. Imaging of the stomach wall using SPECT and i.v.
99m
Tc pertechnetate allows measurement of gastric volume
after meal ingestion. Other centers use an intragastric
barostat or intraluminal high resolution manometry to
measure gastric accommodation  [42, 43]. However, a
nutrient drink test is a surrogate for gastric accommoda­
tion and increased sensation [44]; thus, a maximum toler­
ated volume of less than about 750  kilocalories is
significantly correlated with impairment of gastric accom­
modation measured by an intragastric balloon and could
serve as a biomarker for accommodation and sensation.
273
Abdominal Pain
Patients with diabetes are obviously susceptible to the usual
causes of abdominal pain seen in the general population.
There is an increased prevalence of gallstones because of
altered gallbladder contractility and of mesenteric ischemia
caused by generalized atherosclerosis; however, there is lit­
tle evidence that altered gallbladder contractility per se
(i.e., in the absence of gallstones) causes symptoms.
Thoracolumbar radiculopathy may result in pain in a gir­
dle‐like distribution which does not cross the midline.
Specific tests are indicated in the clinical features of pain
suggest these disorders. It is essential to elicit a careful his­
tory. Among 583 patients in the NIH Gastroparesis
Consortium database, 41% were taking opioids which
included 33% on potent morphine‐like agents, and the
indication was abdominal pain for 61% of patients taking
opioids. These patients have higher symptoms scores,
greater levels of gastric retention, worse quality of life,
increased hospitalization, and increased use of antiemetic
and pain modulator medications compared with nonusers
or those using weaker opioids (e.g., tramadol, tapentadol,
codeine, or propoxyphene) [34].
Management of gastroparesis
and dyspepsia
The principles of management are to address fluid and
nutritional requirements, improve glycemic control and
treat symptoms (Table 22.1 [45]).
Nutritional support
Because both liquids and homogenized solids are more
readily emptied from the stomach than solids, food that is
liquid or blenderized is better tolerated. A randomized,
controlled trial has shown that small particle size meals are
associated with reduced symptoms in gastroparesis  [46].
While nutritional requirements and symptoms can be
addressed to a variable extent in patients with mild and
compensated gastroparesis, patients with severe gastropa­
resis often require hospitalization for one or more of the
following measures: intravenous hydration and correction
of metabolic derangements (ketoacidosis, uremia, hypo‐/
hyperglycemia), nasoenteric decompression, and/or
enteral nutrition to manage vomiting and nutritional
 TABLE 22.1  Treatment strategies for patients with gastroparesis (reproduced from [45]).

Mild gastroparesis (10–15% Moderate gastroparesis (15–35% Severe gastroparesis (> 35%

Management strategies 4 h gastric retention) 4 h gastric retention) 4 h gastric retention)

General measures Review and eliminate medications

inhibiting motility and optimize
glycaemic control in patients with
diabetes
Diet • Small, frequent meals • Small, frequent meals • Blenderized diet
• Low‐fat low‐fibre diet • Low‐fat low‐fibre diet • Routine use of liquid nutrient
• Small‐particle diet when • Small‐particle diet when supplements
symptomatic symptomatic
Nutritional support Rarely needed • Caloric liquids • Caloric liquids
• Orally • Orally
• Rarely requires nutrition by PEJ • May require nutrition by PEJ
feeding tube feeding tube
Pharmacological: prokinetic Metoclopramide Metoclopramide • Metoclopramide or domperidone
• Erythromycin
• Prucalopride
Pharmacological: anti‐emetic Promethazine or prochlorperazine • Promethazine or prochlorperazine • Ondansetron
• Ondansetron • Aprepitant or mirtazapine
Pharmacological: symptom Not needed Not needed Nortriptyline
modulators
Non‐pharmacological Not needed Not needed • Gastrostomy tube decompression
• Laparoscopic and endoscopic
interventions

c22.indd 274 09/24/2021 11:55:48

 Upper Gastrointestinal Manifestations of Diabetes
requirements [38]. Parenteral nutrition may become neces­
sary in cases of malnutrition. Bezoars may be mechanically
disrupted during endoscopy, followed by gastric decom­
pression to drain residual non‐digestible particles.
For patients with severe gastroparesis, it may be neces­
sary to bypass the stomach with a jejunal feeding tube. This
procedure should be preceded by a trial of nasojejunal
feeding for a few days with infusion rates of at least 60 mL
iso‐osmolar nutrient per hour. It is preferable to place jeju­
nal feeding tubes directly into the jejunum either by endos­
copy or, if necessary, by laparoscopy, rather than via
percutaneous endoscopic gastrostomy tubes. Such tubes
allow restoration of normal nutritional status, but they are
associated with adverse effects. There is no evidence to
suggest that gastrectomy relieves symptoms or enhances
quality of life. Patients with gastroparesis often have con­
comitant small intestinal denervation which is likely to
cause persistent symptoms after gastrectomy  [24, 47] or
difficulties with tolerance of jejunal feeding.
Prokinetics
1 Erythromycin at a dose of 3 mg/kg body weight intrave­
nously every 8 hours can accelerate gastric emptying [48,
49]. When oral intake is resumed, treatment with oral
250 mg erythromycin t.i.d. for 1–2 weeks is worthwhile.
Thereafter, the prokinetic effects of erythromycin are
limited by tachyphylaxis. Anecdotal findings suggest
that erythromycin may be effective, if courses are sepa­
rated by a drug‐free period (e.g., lasting 2  weeks).
Hyperglycemia interferes with the prokinetic effect of
intravenous erythromycin on gastric emptying in
healthy subjects and in patients with diabetes [50]. If the
patient remains symptomatic, other prokinetic agents
may be considered as adjuncts.
2 Metoclopramide is the only available medication in the
USA; it is a peripheral cholinergic and antidopaminergic
agent. During acute administration, it initially enhances
gastric emptying of liquids in patients with diabetic gas­
troparesis, but its symptomatic efficacy is probably
related to its central antiemetic effects. Its long‐term use
is restricted by a decline in efficacy and by a troubling
incidence of central nervous system side effects.
Therefore, the author prefers to prescribe a dose of
275
5–10 mg t.i.d., administered 30 minutes before meals as
a liquid formulation, for a short duration; the higher
dose is 20 mg t.i.d. A systematic review of trials con­
cluded there was limited evidence to support the use of
domperidone, which is another dopaminergic antagonist
not approved for use in the USA [51].
3 Experimental prokinetics in development Potentially,
effective pharmacological approaches are directed at the
"traditional" pathways, specifically, 5‐HT4, dopamine D2
and D3, and NK1 receptors, or novel mechanisms (dis­
cussed in the next section).
i  5‐HT4 receptor agonists In a randomized, placebo‐con­
trolled, cross‐over study of 4 weeks duration with a 2‐
week washout, prucalopride improved symptom
control as well as gastric emptying in 28 patients with
idiopathic and 6 patients with diabetic gastropare­
sis  [52]. Similarly, velusetrag, was efficacious in the
treatment of diabetic and idiopathic gastropare­
sis [53]. Among "new generation" 5‐HT4 agonists, pru­
calopride, velusetrag, and naronapride are selective for
5‐HT4 receptors without hERG effects [54, 55]. Results
of the pharmacodynamics effects of a new specific 5‐
HT4 receptor agonist, TAK‐954 (TD‐8954), are keenly
awaited (NCT03281577). Velusetrag and TAK‐954 had
no significant effects on canine, porcine, and human
coronary artery tone, human platelet aggregation,
hERG potassium channel conductance, or off‐target
actions [56].
ii  Targeting sensations with D2/D3 and NK1 antagonists
For increased gastric sensation, the dopamine D2/3
antagonist, TAK‐506, significantly increased the vol­
ume to fullness compared to baseline with 1 week of
treatment [57]. The NK1 receptor antagonist, aprepi­
tant, improved multiple symptoms of gastroparesis
including nausea  [58], and it has been shown to
enhance gastric accommodation rather than affect
gastric emptying in healthy controls  [59]. Similarly,
tradipitant, a novel NK1 receptor antagonist, improved
nausea and other symptoms of gastroparesis in a 4‐
week, randomized, controlled trial [60].
iii  Ghrelin receptor agonist Ghrelin is a 28‐amino acid
orexigenic hormone found mainly in the stomach.
Administration of a pharmacological dose of ghrelin
 276 Management of Disease Complications
induced premature phase III of the migrating motor
complex, increased proximal gastric tone through
central and peripheral sites of action  [61, 62], and
accelerated gastric emptying in some studies of
patients with gastroparesis (reviewed in  [63]).
Relamorelin, a pentapeptide ghrelin receptor agonist,
has potent prokinetic effects estimated to be 15‐ to
130‐fold more potent than natural ghrelin  [64].
Relamorelin, 100 mg subcutaneous (s.c.), accelerated
gastric emptying of solids in patients with prior docu­
mentation of delayed gastric emptying and either type
1 or type 2 diabetes mellitus [65, 66], and it increased
the frequency of distal antral contractions without
inhibiting gastric accommodation or inducing satia­
tion in healthy volunteers  [67]. Relamorelin has
proven clinical efficacy and safety in phase 2A and 2B,
randomized, controlled trials in patients with diabetic
gastroparesis [68–70].
Endoscopic and device therapies
Endoscopic injection of botulinum toxin into the pylorus was
not effective in controlled studies primarily of patients with
idiopathic gastroparesis [71, 72], There have been multiple,
open‐label trials of intra‐pyloric botulinum toxin injec­
tions attesting to the efficacy of this approach [73–75], but
there have been only two randomized, controlled trials [71,
72], both of which failed to show any improvement in
symptoms. However, at least in the trial performed by
Friedenberg and colleagues  [71], botulinum neurotoxin
improved gastric emptying compared to placebo, suggest­
ing that the intervention can result in the desired physio­
logical effect, but not the clinical outcome.
Gastric electrical stimulation is approved as a humanitar­
ian use device for refractory gastroparesis with efficacy
possibly greater in diabetic gastroparesis; its use for this
indication is controversial. Two recent studies argue in
favor of beneficial effects of gastric electrical stimulation.
First, in a study from the Gastroparesis Clinical Research
Consortium (GpCRC), 238 patients without gastric electri­
cal stimulation and 81  with gastric electrical stimulation
showed improvement by ≥ 1 point and change from enroll­
ment with gastric electrical stimulation compared to the
group without the device. However, when adjusting for
patient characteristics, only the symptom of nausea
remained significant [76]. Second, in a large, multicenter,
randomized, double‐blind trial with cross‐over from
France performed in 172 patients (66% women; mean age
45±12 years; 133  with gastroparesis) with chronic
(>  12  months) refractory vomiting  [77], the vomiting
scores improved significantly when the device was on in
patients with delayed or normal gastric emptying. However,
gastric emptying and quality of life were not different.
Gastric per‐oral endoscopic myotomy (G‐POEM): Pyloric
dysfunction was first described in diabetic gastroparesis in
1986 [78]. Multiple studies have demonstrated efficacy in
relief of symptoms and improvement in gastric emptying
in patients with gastroparesis who undergo G‐POEM, as
summarized in a recent systematic review and meta‐
analysis [79].
Conclusion
Diabetic gastroparesis is a significant problem in
patients with other complications of diabetes including
retinopathy and triopathy. Patient management requires
a multi‐disciplinary approach, including coordination
between diabetologist and gastroenterologist, as well as
support from nutrition experts. There are promising
new medications for the treatment of upper gastrointes­
tinal symptoms in patients with diabetes, and potential
endoscopic treatment with pyloromyotomy. However,
clinical trials are still required for the promising new
medications, and sham‐controlled trials for pyloromy­
otomy. Future research is required to identify optimal,
individualized therapy in patients with diabetes and
gastroparesis.
References
1. Feldman M, Schiller ER. Disorders of gastrointestinal
motility associated with diabetes mellitus. Ann Intern Med.
1983;98:378–384.
2. Clouse RE, Lustman PJ. Gastrointestinal symptoms in
diabetic patients: lack of association with neuropathy. Am J
Gastroenterol. 1989;84:868–872.
3. Dyck PJ, Karnes JL, O’Brien PC et al. The Rochester
Diabetic Neuropathy Study: reassessment of tests and
criteria for diagnosis and staged severity. Neurology.
1992;42:1164–1170.
 Upper Gastrointestinal Manifestations of Diabetes
4. Maleki D, Locke GR 3rd, Camilleri M et al. Gastrointestinal
tract symptoms among persons with diabetes mellitus in the
community. Arch Intern Med. 2000;160:2808–2816.
5. Janatuinen E, Pikkarainen P, Laakso M et al. Gastrointestinal
symptoms in middle‐aged diabetic patients. Scand J
Gastroenterol. 1993;28:427–432.
6. Bytzer P, Talley NJ, Leemon M et al. Prevalence of gastroin­
testinal symptoms associated with diabetes mellitus: a popu­
lation‐based survey of 15000 adults. Arch Intern Med.
2001;161:1989–1996.
7. Jung HK, Choung RS, Locke GR 3rd et  al. The incidence,
prevalence, and outcomes of patients with gastroparesis in
Olmsted County, Minnesota, from 1996 to 2006.
Gastroenterology. 2009;136:1225–1233.
8. Talley NJ, Young L, Bytzer P et al. Impact of chronic gastro­
intestinal symptoms in diabetes mellitus on health‐related
quality of life. Am J Gastroenterol. 2001;96:71–76.
9. Jones KL, Russo A, Berry MK et al. A longitudinal study of
gastric emptying and upper gastrointestinal symptoms in
patients with diabetes mellitus [see Comment]. Am J Med.
2002;113:449–455.
10. Chang J, Russo A, Bound M et  al. A 25‐year longitudinal
evaluation of gastric emptying in diabetes. Diabetes Care.
2012;35:2594–2596.
11. Kong MF, Horowitz M, Jones KL et al. Natural history of dia­
betic gastroparesis. Diabetes Care. 1999;22:503–507.
12. Hyett B, Martinez FJ, Gill BM et al. Delayed radionucleotide
gastric emptying studies predict morbidity in diabetics with
symptoms of gastroparesis. Gastroenterology. 2009;137:
445–452.
13. Kassander P. Asymptomatic gastric retention in diabetics
(gastroparesis diabeticorum). Ann Intern Med. 1958;48:
797–812.
14. Parkman HP, Wilson LA, Farrugia G et al. Delayed gastric
emptying associates with diabetic complications in diabetic
patients with symptoms of gastroparesis. Am J Gastroenterol.
2019;114:1778–1794.
15. Ordog T. Interstitial cells of Cajal in diabetic gastroenteropa­
thy. Neurogastroenterol Motil. 2008;20:8–18.
16. Chandrasekharan B, Srinivasan S. Diabetes and the enteric
nervous system. Neurogastroenterol Motil. 2007;19:951–960.
17. Grover M, Bernard CE, Pasricha PJ et al. Diabetic and idio­
pathic gastroparesis is associated with loss of CD206‐posi­
tive macrophages in the gastric antrum. Neurogastroenterol
Motil. 2017;29(6). doi: 10.1111/nmo.13018.
18. Schmidt RE. Neuropathology and pathogenesis of dia­
betic autonomic neuropathy. Int Rev Neurobiol. 2002;50:
257–292.
277
19. Sinnreich M, Taylor BV, Dyck PJ. Diabetic neuropathies:
classification, clinical features, and pathophysiological basis.
Neurologist. 2005;11:63–79.
20. Vittal H, Farrugia G, Gomez G et al. Mechanisms of disease:
the pathological basis of gastroparesis – a review of experi­
mental and clinical studies. Nat Clin Pract Gastroenterol
Hepatol. 2007;4:336–346.
21. Feldman M, Corbett DB, Ramsey EJ et al. Abnormal gastric
function in lonstanding, insulin‐dependent diabetic patients.
Gastroenterology. 1979;77:12–17.
22. Chedid V, Brandler J, Vijayvargiya P et al. Characterization
of upper gastrointestinal symptoms, gastric motor functions
and associations in patients with diabetes at a referral center.
Am J Gastroenterol. 2019;114:143–154.
23. Bredenoord AJ, Chial HJ, Camilleri M et al. Gastric accom­
modation and emptying in evaluation of patients with upper
gastrointestinal symptoms. Clin Gastroenterol Hepatol.
2003;1:264–272.
24. Samsom M, Salet GA, Roelofs JM et al. Compliance of the
proximal stomach and dyspeptic symptoms in patients with
type 1 diabetes mellitus. Dig Dis Sci. 1995;40:2037–2042.
25. Camilleri M, Malagelada JR. Abnormal intestinal motility in
diabetics with gastroparesis. Eur J Clin Invest. 1984;14:
420–427.
26. MacGregor IL, Gueller R, Watts HD et al. The effect of acute
hyperglycemia on gastric emptying in man. Gastroenterology.
1976;70:190–196.
27. Fraser RJ, Horowitz M, Maddox AF et  al. Hyperglycaemia
slows gastric emptying in type 1 (insulin‐dependent) diabe­
tes mellitus. Diabetologia. 1990;33:675–680.
28. Oster‐Jorgensen E, Pedersen SA, Larsen ML. The influence
of induced hyperglycaemia on gastric emptying rate in
healthy humans. Scand J Clin Lab Invest. 1990;50:831–836.
29. Schvarcz E, Palmer M, Aman J et al. Physiological hypergly­
cemia slows gastric emptying in normal subjects and patients
with insulin‐dependent diabetes mellitus. Gastroenterology.
1997;113:60–66.
30. Bharucha AE, Kudva Y, Basu A et  al. Relationship between
glycemic control and gastric emptying in poorly controlled
type 2 diabetes. Clin Gastroenterol Hepatol. 2015;13:466–476.
31. Parthasarathy G, Kudva YC, Low PA et  al. Relationship
between gastric emptying and diurnal glycemic control in
type 1 diabetes mellitus: A randomized trial. J Clin Endocrinol
Metab. 2017;102:398–406.
32. Ascaso JF, Herreros B, Sanchiz V et al. Oesophageal motility
disorders in type 1 diabetes mellitus and their relation to car­
diovascular autonomic neuropathy. Neurogastroenterol Motil.
2006;18:813–822.
 278 Management of Disease Complications
33. Rathmann W, Enck P, Frieling T et al. Visceral afferent neu­
ropathy in diabetic gastroparesis. Diabetes Care. 1991;14:
1086–1089.
34. Hasler WL, Wilson LA, Nguyen LA et  al. Opioid use and
potency are associated with clinical features, quality of life,
and use of resources in patients with gastroparesis. Clin
Gastroenterol Hepatol. 2019;17:1285–1294.
35. Parkman HP, Wilson LA, Hasler WL et al. Abdominal pain
in patients with gastroparesis: associations with gastropare­
sis symptoms, etiology of gastroparesis, gastric emptying,
somatization, and quality of life. Dig Dis Sci. 2019;64:
2242–2255.
36. Mearin F, Malagelada JR. Gastroparesis and dyspepsia in
patients with diabetes mellitus. Eur J Gastroenterol Hepatol.
1995;7:717–723.
37. Kumar A, Attaluri A, Hashmi S et al. Visceral hypersensitiv­
ity and impaired accommodation in refractory diabetic gas­
troparesis. Neurogastroenterol Motil. 2008;20:635–642.
38. Camilleri M. Clinical practice: diabetic gastroparesis. N Engl
J Med. 2007;356:820–829 [erratum appears in N Engl J Med.
2007;357:427].
39. Bharucha AE, Camilleri M, Low PA et al. Autonomic dys­
function in gastrointestinal motility disorders. Gut. 1993;34(3):
397–401.
40. Camilleri M, Zinsmeister AR, Greydanus MP et al. Towards
a less costly but accurate test of gastric emptying and small
bowel transit. Dig Dis Sci. 1991;36:609–615.
41. Szarka LA, Camilleri M, Vella A et  al. A. stable isotope
breath test with a standard meal for abnormal gastric empty­
ing solids in the clinic and in research. Clin Gastroenterol
Hepatol. 2008;6:635–643.
42. Tack J, Piessevaux H, Coulie B et al. Role of impaired gastric
accommodation to a meal in functional dyspepsia.
Gastroenterology. 1998;115:1346–1352.
43. Carbone F, Tack J, Hoffman I. The intragastric pressure
measurement: a novel method to assess gastric accommoda­
tion in functional dyspepsia children. J Pediatr Gastroenterol
Nutr. 2017;64:918–924.
44. Tack J, Caenepeel P, Piessevaux H et al. Assessment of meal
induced gastric accommodation by a satiety drinking test in
health and in severe functional dyspepsia. Gut. 2003;52:
1271–1277.
45. Camilleri M, Chedid V, Ford AC et al. Gastroparesis. Nat Rev
Dis Primers. 2018;4:41.
46. Olausson EA, Störsrud S, Grundin H et al. A small particle
size diet reduces upper gastrointestinal symptoms in patients
with diabetic gastroparesis: a randomized controlled trial.
Am J Gastroenterol. 2014;109:375–385.
47. Jones MP, Maganti K. A systematic review of surgical therapy
for gastroparesis. Am J Gastroenterol. 2003;98:2122–2129.
48. Annese V, Janssens J, Vantrappen G et  al. Erythromycin
accelerates gastric emptying by inducing antral contractions
and improved gastroduodenal coordination. Gastroenterology.
1992;102:823–828.
49. Janssens J, Peeters TL, Vantrappen G et al. Improvement of
gastric emptying in diabetic gastroparesis by erythromycin:
preliminary studies [see Comment]. N Engl J Med. 1990;322:
1028–1031.
50. Samsom M, Akkermans LM, Jebbink RJ et al. Gastrointestinal
motor mechanisms in hyperglycaemia induced delayed gas­
tric emptying in type 1 diabetes mellitus. Gut. 1997;40:
641–646.
51. Sugumar A, Singh A, Pasricha PJ. A systematic review of the
efficacy of domperidone for the treatment of diabetic gastro­
paresis. Clin Gastroenterol Hepatol. 2008;6:726–733.
52. Carbone F, Van den Houte K, Clevers E et al. Prucalopride in
gastroparesis: a randomized placebo‐controlled crossover
study. Am J Gastroenterol. 2019;114:1265–1274.
53. Abell T, Kuo B, Esfandyari T et al. Velusetrag improves gastro­
paresis both in symptoms and gastric emptying in patients with
diabetic or idiopathic gastroparesis in a 12‐week global phase
2B study. Gastroenterology. 2019;156(6):S164.
54. De Maeyer JH, Prins NH, Schuurkes JA et  al. Differential
effects of 5‐hydroxytryptamine4 receptor agonists at gastric
versus cardiac receptors: an operational framework to
explain and quantify organ specific behavior. J Pharmacol
Exp Ther. 2006;317:955–964.
55. Tack J, Camilleri M, Chang L et al. Systematic review: car­
diovascular safety profile of 5‐HT4 agonists developed for
GI disorders. Aliment Pharmacol Ther. 2012;35:745–767.
56. Beattie DT, Higgins DL, Ero MP et al. An in vitro; investiga­
tion of the cardiovascular effects of the 5‐HT(4) receptor
selective agonists, velusetrag and TD‐8954. Vascul Pharmacol.
2013;58:150–156.
57. Dukes GE, Scimia C, Kuo B et  al. Safety, tolerability and
pharmacodynamics of tak‐906, a dopamine 2,3 antagonist,
in patients with diabetic or idiopathic gastroparesis.
Neurogastroenterol Motil. 2019;31(Suppl 3):e13657.
58. Pasricha PJ, Yates KP, Sarosiek I et al. Aprepitant has mixed
effects on nausea and reduces other symptoms in patients
with gastroparesis and related disorders. Gastroenterology.
2018;154:65–76.e11.
 Upper Gastrointestinal Manifestations of Diabetes
59. Jacob D, Busciglio I, Burton D et al. Effects of NK1 receptors
on gastric motor functions and satiation in healthy humans:
results from a controlled trial with the NK1 antagonist,
aprepitant. Am J Physiol Gastrointest Liver Physiol. 2017;313:
G505–G510.
60. Carlin JL, Lieberman VR, Dahal A et  al. Tradipitant, a
novel NK‐1 receptor antagonist, significantly improved
nausea and other symptoms of gastroparesis: results of a
multicenter, randomized, double‐blind, placebo‐controlled
phase II trial. Gastroenterology. 2019;156(Suppl 1):
S1510–S1511.
61. Tack J, Depoortere I, Bisschops R et al. Influence of ghrelin
on interdigestive gastrointestinal motility in humans. Gut.
2006;55:327–333.
62. Peeters TL. Central and peripheral mechanisms by which
ghrelin regulates gut motility. J Physiol Pharmacol. 2003;54:
95–103.
63. Camilleri M, Papathanasopoulos A, Odunsi ST. Actions and
therapeutic pathways of ghrelin for gastrointestinal disor­
ders. Nat Rev Gastroenterol Hepatol. 2009;6:343–352.
64. Van der Ploeg L, Laken H, Sharma S et al. Preclinical gastro­
intestinal prokinetic efficacy and endocrine effects of the
ghrelin mimetic RM‐131. Life Sci. 2014;109:20–29.
65. Shin A, Camilleri M, Busciglio I et  al. Randomized con­
trolled phase Ib study of ghrelin agonist, RM‐131, in type 2
diabetic women with delayed gastric emptying: pharmacoki­
netics and pharmacodynamics. Diabetes Care. 2013;36:
41–48.
66. Shin A, Camilleri M, Busciglio I et  al. The ghrelin agonist
RM‐131 accelerates gastric emptying of solids and reduces
symptoms in patients with type 1 diabetes mellitus. Clin
Gastroenterol Hepatol. 2013;11:1453–1459.
67. Nelson AD, Camilleri M, Acosta A et  al. Effects of ghrelin
receptor agonist, relamorelin, on gastric motor functions and
satiation in healthy volunteers. Neurogastroenterol Motil.
2016;28:1705–1713.
68. Lembo A, Camilleri M, McCallum R et  al. Relamorelin
reduces vomiting frequency and severity and accelerates
gastric emptying in adults with diabetic gastroparesis.
Gastroenterology. 2016;151:87–96.e6.
279
69. Camilleri M, McCallum RW, Tack J et al. Efficacy and safety
of relamorelin in diabetes with symptoms of gastroparesis:
a randomized, placebo‐controlled study. Gastroenterology.
2017;153:1240–1250.e2.
70. Camilleri M, Lembo A, McCallum R et al. Overall safety and
tolerability of relamorelin in adults with diabetic gastropare­
sis: analysis of phase 2a and phase 2b trial data. Neurogastro­
enterol Motil. 2019;31(S4):S15.
71. Friedenberg FK, Palit A, Parkman HP et al. Botulinum toxin
A for the treatment of delayed gastric emptying [see
Comment]. Am J Gastroenterol. 2008;103:416–423.
72. Arts J, Holvoet L, Caenepeel P et  al. Clinical trial: a rand­
omized‐controlled crossover study of intrapyloric injection
of botulinum toxin in gastroparesis. Aliment Pharmacol Ther.
2007;26:1251–1258.
73. Bai Y, Xu MJ, Yang X et al. A systematic review on intrapy­
loric botulinum toxin injection for gastroparesis. Digestion.
2010;81:27–34.
74. Pasricha TS, Pasricha PJ. Botulinum toxin injection for
treatment of gastroparesis. Gastrointest Endosc Clin N Am.
2019;29:97–106.
75. Thomas A, de Souza Ribeiro B, Malespin M et al. Botulinum
toxin as a treatment for refractory gastroparesis: a literature
review. Curr Treat Options Gastroenterol. 2018;16:479–488.
76. Abell TL, Yamada G, McCallum RW et al. Effectiveness of
gastric electrical stimulation in gastroparesis: Results from a
large prospectively collected database of national gastropa­
resis registries. Neurogastroenterol Motil. 2019;31:e13714.
77. Ducrotte P, Coffin B, Bonaz B et al. Gastric electrical stimu­
lation reduces refractory vomiting in a randomized cross‐
over trial. Gastroenterology. 2019;Oct 21:ii. S0016–5085(19)
41463–41467. doi: 10.1053/j.gastro.2019.10.018. [Epub ahead
of print].
78. Mearin F, Camilleri M, Malagelada JR. Pyloric dysfunction
in diabetics with recurrent nausea and vomiting. Gastro­
enterology. 1986;90:1919–1925.
79. Aghaie Meybodi M, Qumseya BJ, Shakoor D et al. Efficacy
and feasibility of G‐POEM in management of patients with
refractory gastroparesis: a systematic review and meta‐analysis.
Endosc Int Open. 2019;7:E322–E329.
 Index

abdominal pain, 273 incretins effect, 221

acarbose, 177–178 new techniques, 225
A Diabetes Outcome Progression Trial (ADOPT), 177 rescue therapy, 221–223
advanced glycosylation end products (AGES), 94 surgery, weight loss, 225
Advanced Hybrid Closed Loop (AHCL), 119 type 2 diabetes mellitus (T2DM), 221
aggressive risk factor control, 164 vertical sleeve gastrectomy (VSG), 221
alanine amino‐transferase (ALT), 141 weight loss, 221
alpha‐glucosidase inhibitors, 190 Bezafibrate Infarction Prevention (BIP), 205
American College of Gastroenterology (ACG), 50 biguanides/metformin, 133, 190
American College of Obstetricians and Gynecologists combination therapy, 135–137
(ACOG), 77 contraindications, 137
American Diabetes Association (ADA), 12, 45, 102, 128 gastrointestinal, 138
amylin mimetics, 190–191 lactic acidosis, 137–138
Anglo‐Scandinavian Cardiac Outcomes Trial mechanism of action, 134–135
(ASCOT), 200 pharmacokinetics, 135
Antihypertensive and Lipid‐Lowering treatment to polycystic ovary syndrome (PCOS), 137
prevent Heart Attack Trial (ALLHAT), 201 bilateral internal thoracic artery (BITA), 233
antipsychotic drug blood glucose (BG), 91
natural history, 57 bromocriptine, 178
schizophrenia, 56
screening, 57–58 caloric sweeteners, 129
second‐generation antipsychotics, 57 Canagliflozin and Renal Events in Diabetes With
atherosclerotic cardiovascular disease (ASCVD), 152 Established Nephropathy Clinical Evaluation
(CREDENCE), 180
bariatric surgery, 11 candidate gene studies, 18
first‐line therapy, 221–222 carbohydrate, 129
gastric banding, 223–224 cardiovascular disease (CVD)
gastric bypass, 223–224 aggressive risk factor control, 164

Clinical Dilemmas in Diabetes, Second Edition. Edited by Adrian Vella.

© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.

280
 Index 281

carotid ultrasound, 168 Centers for Medicare and Medicaid Services (CMS), 90
coronary artery calcium score, 166–167 Cholesteryl ester transfer protein (CETP), 199
coronary artery disease (CAD), echocardiogram chronic kidney disease (CKD), 148
for, 168 closed loop control (CLC)
coronary computed tomography angiography Auto‐mode enabled MiniMedTM, 119
(CCTA), 167–168 continuous glucose monitoring (CGM), 116
diabetes‐specific clinical risk predictors, 165–166 continuous subcutaneous insulin infusion (CSII), 116
endothelial function studies, 168 control IQ technology, 119–120
guidelines, 163–164 cost of diabetes technologies, 121
heart failure (HF), screening for, 168–169 FLAIR trial, 119
hyperglycemia, 164–165 future technologies, 121
hypertriglyceridemia, 203–204 impaired hypoglycemia awareness, 116
myocardial perfusion imaging, 168 insulin pump, 119–121
National Health and Nutrition Examination Survey MiniMed 670G, 118–119
(NHANES), 163 multiple daily injections (MDI), 116
nontraditional risk factors, 166 ongoing pivotal studies, 120
percutaneous coronary intervention (PCI), 164 post‐marketing studies, 119–121
proprotein convertase subtilisin/kexin type 9 SmartGuardTM, 119
(PCSK9), 202–203 T1D, 120
risk calculators, 169–170 T1DM, currently approved CLC for, 118
screening tools, 166–170 t:slim X2TM, 119–120
stress testing, 168 Cochrane Collaborative, 89
traditional risk factors, 165 colesevelam, 178
cardiovascular risk reduction Collaborative Atorvastatin Diabetes Study (CARDS), 200
acarbose, 177–178 Congenital rubella, 19
bromocriptine, 178 continuous glucose monitoring (CGM), 95–96, 102
colesevelam, 178 diabetes care, 116
diabetes, 174–175 diabetes management, 107
glucagon‐like peptide‐1 (GLP‐1)‐based therapy, duration of, 107
178–180 glycemic control, 106
metformin, 175 glycemic variability, 108–110
pramlintide, 178 hypoglycemia, 106–107
sodium glucose cotransporter 2 inhibitors (SGLT‐2i), implantable, 117–118
180 pregnancy, 117
sulfonylureas, 175–176 time in range (TIR), 108
thiazolidinediones, 176–177 WISDM Study, 116
Carotid Intimal Medial Thickness (CIMT), 166 continuous subcutaneous insulin infusion (CSII)
carotid intima‐media thickness, 142 therapy, 94
carotid ultrasound, 168 cornea, 263–264
cataract, 264 treatment, 264
 282 Index

coronary artery bypass surgery (CABG), 164 diagnostic criteria, 3–4

coronary artery calcium score, 166–167 diet, 11–12
coronary artery disease (CAD), echocardiogram exercise, 11–12
for, 168 glucose control, 229
coronary computed tomography angiography (CCTA), haemochromatosis, natural history of, 49
167–168 hereditary haemochromatosis (HH), 48–49
cow’s milk, 19–20 hormone excess, 54–55
Coxsackievirus B4, 19 human immunodeficiency virus (HIV) infection,
C‐peptide, 39, 69 58–59
Cushing’s syndrome, 47, 54–55 hyperglycemia, 229
cystic fibrosis‐related diabetes (CFRD) immunotherapy, 59–60
clinical features of, 52 medical therapy, 229
definition, 50 optimal medical therapy (OMT), 229
natural history, 50 oxidative stress, 228
other forms of diabetes, 50–52 pancreatic‐cancer‐associated diabetes, 53–54
screening, 52–53 patients, subgroup analysis of, 232–233
cystic fibrosis transmembrane regulator (CFTR), 40 percutaneous coronary intervention, 234
cytotoxic T lymphocyte antigen 4 (CTLA‐4), 18 post‐transplantation diabetes mellitus (PTDM), 55–56
prediabetes, 7–8, 11
Da Qing studies, 10–11 prevention, 11
Detection of Ischemia in Asymptomatic Diabetics reclassifying, novel approaches to, 40–41
(DIAD), 164 risk stratification, biomarkers for, 233
Diabeloop Generation 1 (DBLG1) Hybrid Closed symptom control for stable patients, 231
Loop, 120 types of , 45–46
Diabetes Complication and Control Trial (DCCT), 88, diabetes prediction and prevention study (DIPP), 26
89, 174 diabetes prevention trial‐type 1 (DPT‐1), 24–26
diabetes mellitus (DM) diabetes‐specific clinical risk predictors, 165–166
acute coronary syndromes (ACS), 228, 233 diabetic gastroparesis, 269, 270
antiplatelet therapy, 229 diabetic macular edema (DME), 259–260
antipsychotic drug use, 56–57 anti‐VEGF therapy, 261–262
bariatric surgery, 11 corticosteroid, 260–261
Bypass Angioplasty Revascularization Investigation future directions, 263
(BARI), 229 macular laser, 260
cardiovascular disease (CVD), 232 surgery, 262
cardiovascular risk reduction, 174–175 diabetic retinopathy (DR)
coronary artery bypass grafting, 233–234 anti‐VEGF, 258
coronary artery disease (CAD), 228 blood pressure control, 254
coronary revascularization, 230–231 cataract, 264
cystic fibrosis‐related diabetes, 50 combination treatment, 258
delay of, 11 cornea, 263–264
 Index 283

diabetic macular edema (DME) see diabetic macular type 2 diabetes (T2D), 199–202
edema (DME) dyspepsia, 273
glycemic control, 254 dysphagia, 272
laser photocoagulation, 256–258
lipid‐lowering strategies, 254–255 eicosapen taenoic acid (EPA), 205–206
neovascular glaucoma, 259 electrocardiogram (ECG), 92
non‐proliferative diabetic retinopathy, 256 endogenous glucose production (EGP), 146
ocular complications, 255–256 endothelial function studies, 168
orbital disease, 264 end‐stage renal disease (ESRD), 148, 170
prevention, 254–255 Enterovirus, 19
proliferative diabetic retinopathy, 256 Epidemiology of Diabetes Interventions and
retinal vein occlusion, 259 Complications (EDIC), 88, 102
risk factors for, 252–253 erythromycin, 275
screening, 253–254 estimated glomerular filtration rate (eGFR), 148
surgical treatment/vitrectomy, 258–259 European Association for the Study of Diabetes
treatment of, 256–259 (EASD), 57
diet, 11–12 European nicotinamide diabetes intervention trial
American Diabetes Association, 128 (ENDIT), 24
caloric sweeteners, 129 European Psychiatric Association (EPA), 57
carbohydrate, 129 exenatide, 148–149
cultural norms, 125 exercise, 11–12
food availability, 125
intermittent fasting, 126 fasting plasma glucose (FPG), 50, 94
ketogenic diets, 127 fetal monitoring, 80
medical nutrition therapy, 124 first‐phase insulin response (FPIR), 25
Mediterranean diet plan, 127 5‐HT 4 receptor agonists, 275
plant‐based diets, 125 flash glucose monitoring systems, 96–99
regular dietary fiber intake, 129 food availability, 125
sugar substitutes, 129
time‐restricted eating (TRE), 126 Galega officinalis, 133
type 1 diabetes, 129 gastric dysfunctions, 270–272
type 2 diabetes, 124, 129 gastroparesis, 272, 273
dipeptidyl peptidase‐IV (DDP‐4) inhibitors, 93, 191 gestational diabetes mellitus (GDM)
drug‐eluting stents (DES), 230 definition, 75
dulaglutide, 150 diagnosis, 75–77
dyslipidemia fetal monitoring, 80
atherosclerotic cardiovascular disease (ASCVD), future perspectives, 81–82
211–212 gestational weight gain, 79–80
PCSK9 see proprotein convertase subtilisin/kexin glucocorticoid therapy, 80
type 9 (PCSK9) glycemic goals, 77
 284 Index

gestational diabetes mellitus (GDM) (cont’d) micro‐ and macrovascular complications, 102–103
insulin therapy, 77–78 post prandial glucose (PPG), 105
intrapartum management, 80–81 self‐monitored blood glucose (SMBG), 104–105
lifestyle modification, 77 vascular risk, 101
oral glucose‐lowering agents, 78–79 hepatitis C infection, 56
postpartum maternal care, 81 hereditary haemochromatosis (HH)
postpartum neonatal care, 81 natural history of, 49
gestational weight gain, 79–80 screening case, 49–50
Ghrelin receptor agonist, 275–276 high‐density lipoprotein (HDL), 141
glucagon‐like peptide‐1 (GLP‐1), 93, 178–180, 191 highly active antiretroviral therapy (HAART), 141
clinical benefits, 147 HNF1A mutations, 70
dulaglutide, 150 HNF4A mutations, 70
early development of, 147 HorizonTM Pivotal study, 120
exenatide, 148–149 human immunodeficiency virus (HIV) infection, 58–59
liraglutide, 149–150 human leukocyte antigen (HLA), 17–18, 38
lixisenatide, 150 hyperglycemia, 69, 105, 164–165
semaglutide, 150 critical illness, 242
side effects, 147 glucose testing, 247
glucagon suppression, 38 interpretation, 244–246
glucocorticoids, 54, 80 mechanisms, 242
glucokinase mutations, 69–70 nutritional support, 247–249
glucose‐dependent insulinotropic polypeptide hypertriglyceridemia, 199
(GIP), 147
glucose variability (GV), 115 immune intervention therapies, 27–32
glutamic acid decarboxylase (GAD‐65), 18, 29, 39 immunotherapy trials, 30–31
glyburide, 78 impaired fasting glucose (IFG), 3,
glycemic goals, 77 5, 9–11, 164
glycemic measurements, 6–7 impaired glucose tolerance (IGT), 3, 5, 164
incretin‐based therapy, 32–33
heartburn, 272 cardiovascular outcomes, 152–153
heart failure (HF), screening for, 168–169 clinical practice, 153–154
Heart Outcomes Prevention Evaluation (HOPE), 165 dipeptidyl peptidase‐4 inhibitors (DPP‐4i) see
hemoglobin A1c (HbA1c) dipeptidyl peptidase‐4 inhibitors (DPP‐4i)
clinical practice, integration into, 102 glucagon‐like peptide‐1 (GLP‐1) see glucagon‐like
clinical utility of, 103 peptide‐1 (GLP‐1)
continuous glucose monitoring (CGM) see incretin effect, 146–147
continuous glucose monitoring (CGM) insulin, 18, 188
discovery, 102 insulin autoantibody (IAA), 25
glucose hypothesis, 101 insulin sensitizers
hyperglycemia, 105 biguanides see biguanides
limitations of, 103–104 thiazolidinediones see thiazolidinediones (TZDs)
 Index 285

insulin therapy, 53, 77–78 retinopathy, 8

intermittent fasting, 126 Model Predictive Control (MPC), 119
internal thoracic artery (ITA), 230 monogenic diabetes
International Diabetes Federation (IDF), 95 glucokinase mutations, 69–70
intestinal cells of Cajal (ICCs), 270 HNF1A mutations, 70
intrapartum management, 80–81 HNF4A mutations, 70
islet‐cell antibodies (ICA), 18 maturity onset diabetes of the young (MODY), 68
screen for , 69
ketogenic diets, 127 type 1 diabetes, 68–69
ketosis, 69 type 2 diabetes, 69
mortality, 9–10
lactic acidosis, 138 Multi‐Ethnic Study of Atherosclerosis (MESA), 166
large for gestational age (LGA), 75 myocardial infarction (MI), 143, 229, 231
latent autoimmune diabetes of adults (LADA), 39–40 myocardial perfusion imaging, 168
LDL‐cholesterol levels, 212
lipid lowering, 141 National Health and Nutrition Examination Survey
anticoagulation, 255 (NHANES), 6, 8, 163, 187
aspirin, 255 National Screening Committee (NSC), UK, 47
diet, 255 nephropathy, 8–9
fenofibrate, 254 neuropathy, 9
statins, 255 Niacin/nicotinic acid, 205
lipodystrophies, 141 Nicotinamide treatment, 24
liraglutide, 149–150 nitric oxide synthase (NOS), 270
lixisenatide, 150 nonalcoholic fatty liver disease (NAFLD), 140–141
Long‐term Intervention with Pravastatin in Ischemic nontraditional risk factors, 166
Disease (LIPID), 201
low‐density lipoprotein (LDL), 141, 164 omega‐3 fatty acid, 205
lymphoid phosphatase (LYP), 18 oral glucose‐lowering agents, 78–79
oral glucose tolerance test (OGTT), 3, 6–7, 77
maturity‐onset, 68 orbital disease, 264
maturity onset diabetes of the young (MODY), 60, 68 oxidative post‐translationally modified insulin
mean absolute relative difference (MARD), 117 (oxPTM‐ INS), 19
medical nutrition therapy, 124
Mediterranean diet plan, 127 pancreatic‐cancer‐associated diabetes, 53–54
metabolic memory, 88 Patch Pump Series, 120
metformin, 78, 175, 192 see biguanides percutaneous coronary intervention (PCI), 164
metoclopramide, 275 peripheral artery disease (PAD), 165
microvascular complications, prediabetes peroxisome proliferator activated receptor (PPAR),
mortality, 9–10 138–139, 176
nephropathy, 8–9 mechanism, 139
neuropathy, 9 phenformin, 133
 286 Index

pheochromocytoma/paraganglioma (PPGL), 54 diabetes, patients with, 214

pioglitazone, 140 efficacy, 212–213
plant‐based diets, 125 future perspectives, 214–215
polycystic ovary syndrome (PCOS), 137 LDL‐cholesterol levels, 212
postpartum maternal care, 81 novel lipid‐lowering therapies, 215–217
postpartum neonatal care, 81 safety, 212–213
post prandial glucose (PPG), 94, 105 Prospective Pioglitazone Clinical Trial in Macrovascular
post‐transplantation diabetes mellitus (PTDM) Events (PROACTIVE), 140
clinical features, 56 Public Health England (PHE), 47
screening, 56
pramlintide, 178 randomized clinical trials (RCTs), 89, 147
prediabetes randomized control trail (RCT), 116
American diabetes association 2020, 12 regular dietary fiber intake, 129
bariatric surgery, 11 retinopathy, 8
Da Qing studies, 10–11 rosiglitazone, 140
definition, 3 Roux‐en‐Y‐Gastric Bypass (RYGB), 38
diabetes mellitus (DM), 3–4, 7–8, 11–12
diabetes prevention program, 10 secondary prevention, 24
diagnostic criteria, 3–4 self‐monitoring of blood glucose (SMBG), 104–105
diet, 11–12 continuous glucose monitoring systems, 95–96
epidemiology of, 4 definition, 87–88
exercise, 11–12 flash glucose monitoring systems, 96–99
glycemic measurements, 6–7 postprandial, 94–95
impaired fasting glucose, 5 type 1 diabetes (T1D), 88–89
impaired glucose tolerance (IGT), 3, 5 type 2 diabetes (T2D), 89–94
management of, 10 semaglutide, 150
microvascular complications see microvascular sensitivity, 6–7
complications, prediabetes 780G pivotal trial, 119
oral glucose tolerance test (OGTT), 3, 6–7 short‐term intensive insulin therapy, 40
pharmacologic therapy, 11 670G System, 119
progression, risk of, 7–8 small inhibiting RNA (siRNA), 212
reproducibility, 6–7 sodium glucose cotransporter 2 ­inhibitors (SGLT‐2i), 180
screening recommendations, 5–6 sodium‐glucose cotransporters (SGLT), 191–192
sensitivity, 6–7 sterol regulatory element binding protein (SREBP), 212
primary prevention, 21–24 stress testing, 168
prokinetics, 275–276 sugar substitutes, 129
Proportional Integrated Derivative (PID), 118 sulfonylurea (SU), 91
proprotein convertase subtilisin/kexin type 9 (PCSK9) sulfonylureas, 175–176, 188–190
cardiovascular, 213–214 Synergy between Percutaneous Coronary Intervention
clinical use, 214–215 with Taxus and Cardiac Surgery (SYNTAX), 232
 Index 287

tertiary prevention, 26–27 primary prevention, 21–24

The Environmental Determinants of Diabetes in the secondary prevention, 24
Young (TEDDY) study, 23 self‐monitoring of blood glucose (SMBG), 88–89
thiazolidinediones (TZDs), 138–139, 176–177, 190 tertiary prevention, 26–27
adverse effects, 142–143 vitamin D deficiency, 20
carotid intima‐media thickness, 142 type 2 diabetes (T2D), 88, 116
combination therapy, 139–140 autoantibodies, 39–40
contraindications, 142 clinical features of, 52
lipid lowering, 141 cystic fibrosis‐related diabetes, 40
lipodystrophies, 141 diet, 124
mechanism of action, 138–139 environmental influences, disease presentation,
nonalcoholic fatty liver disease (NAFLD), 140–141 38–39
pharmacokinetics, 139 first‐line therapy, 221–222
polycystic ovary syndrome, 141 gastric banding, 223–224
time‐restricted eating (TRE), 126 gastric bypass, 223–224
Treating to New Targets (TNT), 201 genetic influences, disease presentation, 38–39
troglitazone (TGZ), 144 glucose toxicity, 40
type 1 diabetes (T1D) hyperglycemia, 37–38
autoantibodies, 39–40 incretin‐based therapy see ­incretin‐based therapy
clinical features of, 52 incretins effect, 221
cow’s milk, 19–20 insulin secretion, 39
cystic fibrosis‐related diabetes, 40 monogenic diabetes, 69
diabetes prediction and prevention study (DIPP), 26 new techniques, 225
diabetes prevention trial‐type 1 (DPT‐1), 24–26 pathophysiology, 37–38
diet, 129 rescue therapy, 221–223
environmental factors, 19 self‐monitoring of blood glucose (SMBG), 89–94
environmental influences, disease presentation, 38–39 surgery, weight loss, 225
European nicotinamide diabetes intervention vertical sleeve gastrectomy (VSG), 221
trial (ENDIT), 24 weight loss, 187–188, 221
GAD65, 29 United Kingdom Prospective Diabetes Study (UKPDS),
genetic influences, disease presentation, 38–39 88, 103, 175
glucose toxicity, 40 upper gastrointestinal manifestations
hyperglycemia, 37–38 abdominal pain, 273
immune intervention therapies, 27–32 diabetic gastroparesis, 270
immunotherapy trials, 30–31 diagnosis, 272–273
incretin‐based therapies, 32–33 dyspepsia, 273
insulin secretion, 39 dysphagia, 272
monogenic diabetes, 68–69 endoscopic and device therapies, 276
pathogenesis of, 17–19 epidemiology, 269–270
prevention of, 21–22 gastric dysfunctions, 270–272
 288 Index

upper gastrointestinal manifestations (cont’d) diabetic medications for, 192

gastroparesis, 272, 273 DPP‐4 inhibitors, 191
heartburn, 272 GLP‐1 receptor agonists, 192
mechanisms, 269 glucagon‐like peptide (GLP)‐1, 191
nutritional support, 273–275 guide to, 193
pathophysiology, 269 insulin, 188
prokinetics, 275–276 metformin, 192
obesity, 187–188
vitamin D deficiency, 20, 26–27, 223 sodium‐glucose cotransporters (SGLT), 191–192
sulfonylureas, 188–190
weight loss, medication for thiazolidinediones (TZDs), 190
alpha‐glucosidase inhibitors, 190 type 2 diabetes, 187–188
amylin mimetics, 190–191
biguanides/metformin, 190 Zinc Transporter 8 (ZnT8), 18, 39
 WILEY END USER LICENSE AGREEMENT
Go to www.wiley.com/go/eula to access Wiley’s ebook EULA.