The American Journal of Kidney Diseases (ISSN 0272-6386) is
published monthly by Elsevier Inc., 230 Park Avenue, Suite
800, New York, NY 10169-0901, USA. Periodicals postage
paid at New York, NY and additional mailing offices.
POSTMASTER: Send address changes to The American
Journal of Kidney Diseases, Elsevier, Journal Returns, 1799
Highway 50 East, Linn, MO 65051, USA.
Editorial correspondence should be addressed to Harold I.
Feldman, MD, MSCE, Editor-in-Chief, AJKD, Perelman School of
Medicine at the University of Pennsylvania, Blockley Hall, Room
824, 423 Guardian Drive, Philadelphia, PA 19104-6021.
Business correspondence (orders, claims, online, change of
address): Please visit our Support Hub page https://service.
elsevier.com for assistance or contact: United States and Canada:
Telephone (800) 654-2452; E-mail: JournalsCustomer
Service-usa@elsevier.com (for print support); journalsonline
support-usa@elsevier.com (for online support). Europe, Africa,
Middle East: Telephone 44 (0) 1865 843434; E-mail: Journals
CustomerServiceEMEA@elsevier.com. Asia and Australia: Tele-
phone 65 6349-0222; E-mail: asiainfo@elsevier.com.
Change of address notices, including both the old and new addresses of
the subscriber, should be sent at least one month in advance.
Customer Service: Telephone 1-800-654-2452; outside the
United States and Canada, 314-447-8871; fax 314-447-8029.
Yearly subscription rates: United States and possessions: individual,
$735.00. All other countries: individual, $947.00. For all areas
outside the United States and possessions, there is no additional
charge for surface delivery. Student and resident: United States and
possessions, $367.00; all other countries, $488.00. To receive
student/resident rate, orders must be accompanied by name of
affiliated institution, date of term, and the signature of program/
residency coordinator on institution letterhead. Orders will be
billed at individual rate until proof of status is received. Single
issues, both current and back, exist in limited quantities and are
offered for sale subject to availability. Current prices are in effect for
back volumes and back issues. Back issues sold in conjunction with
a subscription are on a prorated basis. Prices are subject to change
without notice. Checks should be made payable to Elsevier and sent
to American Journal of Kidney Diseases, Elsevier Health Sciences
Division, Subscription Customer Service, 3251 Riverport Lane,
Maryland Heights, MO 63043.
Copyright Ó 2021, National Kidney Foundation. All rights
reserved. No part of this publication may be reproduced or
transmitted in any form or by any means now or hereafter
known, electronic or mechanical, including photocopy,
recording, or any information storage and retrieval system,
without permission in writing from the publisher. Printed in
the United States of America.
Permission may be sought directly from Elsevier’s Global
Rights Department in Oxford, UK: telephone 215-239-3804 or
+44 (0)1865 843830, fax +44 (0)1865 853333. Requests may
also be completed online via the Elsevier site (http://www.
elsevier.com/authors/obtaining-permission-to-re-use-elsevier-
material).
The appearance of the code at the bottom of the first page of an
article in this journal indicates the copyright owner’s consent that
copies of the article may be made for personal or internal use, or
for the personal or internal use of specific clients. This consent
is given on the condition, however, that the copier pay the
stated per-copy fee through the Copyright Clearance Center, Inc
(222 Rosewood Dr, Danvers, MA 01923; (978) 750-8400) for
copying beyond that permitted by Sections 107 or 108 of the US
Copyright Law. This consent does not extend to other kinds of
copying, such as copying for general distribution, for advertising or
promotional purposes, for creating new collective works, or for
resale. Absence of the code indicates that the material may not
be processed through the Copyright Clearance Center, Inc. Reprints
of single articles available online may be obtained by purchasing
Pay-Per-View access for $30 per article on the journal web site,
http://www.ajkd.org.
Reprints: To order 100 or more reprints for educational,
commercial, or promotional use, contact Derrick Imasa at 212-633-
3874, Elsevier Inc., 230 Park Ave, Ste 800, New York, NY 10169-
0901, USA. Fax: 212-462-1935; email: reprints@elsevier.com.
Advertising representative: Jane Liss, Triple Threat Media,
630 Madison Avenue, Manalapan, NJ 07726. Telephone 1-732-
890-9812.
The ideas and opinions expressed in the American Journal of
Kidney Diseases do not necessarily reflect those of the National
Kidney Foundation, the Editor, or the Publisher. Publication of
an advertisement or other product mentioned in the American
Journal of Kidney Diseases should not be construed as an
endorsement of the product or the manufacturer’s claims.
Readers are encouraged to contact the manufacturer with any
questions about the features or limitations of the products
mentioned. Practitioners and researchers must always rely
on their own experience and knowledge in evaluating and
using any information, methods, compounds or experiments
described herein. Because of rapid advances in the medical
sciences, in particular, independent verification of diagnoses
and drug dosages should be made. To the fullest extent of the
law, no responsibility is assumed by the publisher for any
injury and/or damage to persons or property as a matter of
products liability, negligence or otherwise, or from any use
or operation of any methods, products, instructions or ideas
contained in the material herein.
The contents of the American Journal of Kidney Diseases
are indexed in Index Medicus and MEDLINE, Current Contents/
Clinical Medicine, Science Citation Index, EMBASE/Excerpta
Medica, ISI, and CINAHL.
The American Journal of Kidney Diseases home page is located
at http://www.ajkd.org.
EDITORIAL BOARD

VOL 77 - NO 6 - JUNE 2021

EDITOR-IN-CHIEF FEATURE EDITORS EDITORIAL INTERNS

Harold I. Feldman Agnes B. Fogo Krishna Agarwal
University of Pennsylvania Atlas of Renal Pathology Beth Israel Deaconess Medical Center
Philadelphia, PA Vanderbilt University Harvard Medical School
Nashville, TN Boston, MA
DEPUTY EDITORS
Linda Fried Debbie Chen
Jeffrey S. Berns In Practice University of California
University of Pennsylvania San Francisco, CA
University of Pittsburgh
Philadelphia, PA Pittsburgh, PA
Racquel Holmes
Laura M. Dember Duke University
University of Pennsylvania Asghar Rastegar Durham, NC
Philadelphia, PA Core Curriculum
Yale University Arun Rajasekaran
New Haven, CT
ENGAGEMENT EDITOR University of Alabama at Birmingham
Birmingham, AL
Holly Kramer Dena E. Rifkin
Loyola Medical Center In a Few Words
Maywood, IL University of California EDITORIAL OFFICE
San Diego, La Jolla, CA
ASSOCIATE EDITORS Nijsje Dorman
Managing Editor
Daniel E. Weiner
Roy D. Bloom
Policy Forum Alina Foo
University of Pennsylvania
Tufts Medical Center Associate Managing Editor
Philadelphia, PA
Boston, MA
Tamara Isakova Noah B. Skaroff
Northwestern University Timothy Yau Editorial Assistant
Chicago, IL Social Media
Washington University Andrew Fleming
James S. Kaufman St. Louis, MO Editorial Coordinator
VA NY Harbor Healthcare System
New York, NY Barbara Saper
Editorial Office Associate
Charmaine E. Lok
University of Toronto STATISTICS/METHODS EDITORS
Toronto, Canada Jesse Yenchih Hsu EDITORS EMERITI
University of Pennsylvania
Sankar D. Navaneethan George Porter
Philadelphia, PA
Baylor College of Medicine 1981-1986
Michael E. DeBakey VA Medical Center
Houston, TX Christopher H. Schmid
Brown University Robert G. Luke
Providence, RI 1987-1991
Peter P. Reese
University of Pennsylvania
Philadelphia, PA Saulo Klahr
Wei (Peter) Yang
1992-1996
University of Pennsylvania
INTERNATIONAL EDITORS Philadelphia, PA
Neil A. Kurtzman
Richard Haynes 1997-2001
University of Oxford
Oxford, United Kingdom PATHOLOGY EDITOR Bertram L. Kasiske
2002-2006
Magdalena Madero Matthew B. Palmer
National Heart Institute University of Pennsylvania Andrew S. Levey
Mexico City, Mexico Philadelphia, PA 2007-2016
Vol 77 - No 6 - June 2021

EDITORIAL BOARD

Charles E. Alpers Amit X. Garg Paul Muntner Tariq Shafi

Seattle, WA London, Canada Birmingham, AL Jackson, MD

Nisha Bansal Morgan Grams Devika Nair Michael G. Shlipak

EDITORIAL BOARD
Seattle, WA Baltimore, MD Nashville, TN San Francisco, CA

Yoshio Hall Behzad Najafian Manoocher Soleimani

Andrew S. Bomback
Seattle, WA Seattle, WA Cincinnati, OH
New York, NY
Ann O’Hare Manish Sood
L. Ebony Boulware Meera Harhay
Seattle, WA Ottawa, Canada
Durham, NC Philadelphia, PA
Ikechi Okpechi
Swapnil Hiremath Matthew A. Sparks
Ursula C. Brewster Cape Town, South Africa
Ottawa, Canada Durham, NC
New Haven, CT Paul M. Palevsky
Jeff Hodgin Pittsburgh, PA Wendy L. St. Peter
Maurice Alan Brookhart
Ann Arbor, MI Minneapolis, MN
Chapel Hill, NC Biff F. Palmer
Lesley A. Inker Dallas, TX Natalie Staplin
Steven M. Brunelli
Boston, MA Oxford, United Kingdom
Needham, MA Chirag Parikh
Vivekanand Jha Baltimore, MD Maarten W. Taal
Emmanuel A. Burdmann New Delhi, India Nottingham, United Kingdom
São Paulo, Brazil Martha Pavlakis
Kenar D. Jhaveri Boston, MA Navdeep Tangri
Anna Burgner Great Neck, NY Winnipeg, Canada
Nashville, TN Roberto Pecoits-Filho
Kirsten Johansen Curitiba, Brazil Gregory E. Tasian
Kerri Cavanaugh Minneapolis, MN Philadelphia, PA
Nashville, TN Carmen A. Peralta
David Johnson San Francisco, CA Marcello Tonelli
Glenn Chertow Brisbane, Australia Calgary, Canada
Mark Perazella
Palo Alto, CA
Kamyar Kalantar-Zadeh New Haven, CT Allison Tong
Michael J. Choi Orange, CA Sydney, Australia
Jeffrey Perl
Baltimore, MD Toronto, Canada
Pascale Khairallah Joel M. Topf
Beatrice Concepcion Houston, TX Paul Phelan Detroit, MI
Nashville, TN Anna Köttgen Edinburgh, United Kingdom Robert Toto
Freiburg, Germany Laura Plantinga Dallas, TX
Mirela Dobre
Cleveland, OH Csaba P. Kovesdy Atlanta, GA Ifeoma Ulasi
Memphis, TN Kevan R. Polkinghorne Enugu, Nigeria
Paul Drawz
Melbourne, Australia
Minneapolis, MN Abhijit Kshirsagar Mark Unruh
Chapel Hill, NC Neil R. Powe Albuquerque, NM
Kevin Erickson San Francisco, CA
Houston, TX Eduardo Lacson, Jr Bradley A. Warady
Boston, MA Giuseppe Remuzzi Kansas City, MO
Samira Farouk Bergamo, Italy
New York, NY James P. Lash Angela Webster
Chicago, IL Helmut G. Rennke Sydney, Australia
Steven Fishbane Boston, MA
Great Neck, NY Edgar V. Lerma Francis Weng
Berwyn, IL Ana Ricardo Livingston, NJ
Jennifer E. Flythe Chicago, IL
Kathleen Liu Francis Wilson
Chapel Hill, NC Bruce M. Robinson
San Francisco, CA New Haven, CT
Ann Arbor, MI
Robert Foley Claudia M. Lora Wolfgang Winkelmayer
Minneapolis, MN Michael V. Rocco Houston, TX
Chicago, IL
Winston-Salem, NC
Linda Fried Mark A. Lusco Jerry Yee
Pittsburgh, PA Roger A. Rodby
Nashville, TN Detroit, MI
Chicago, IL
Masafumi Fukagawa Laura Mariani Xueqing Yu
Rajiv Saran
Isehara, Japan Ann Arbor, MI Guangzhou, China
Ann Arbor, MI
Susan Furth Kunihiro Matsushita Deirdre Sawinski Luxia Zhang
Philadelphia, PA Baltimore, MD Philadelphia, PA Beijing, China

Ron T. Gansevoort Mara McAdams-DeMarco Jane O. Schell Carmine Zoccali

Groningen, the Netherlands Baltimore, MD Pittsburgh, PA Reggio Calabria, Italy
 VOL 77 - NO 6 - JUNE 2021

ON THE COVER When ripe, winter cherries blaze

like bright flames within skeletal outer “lanterns”.
The immature green berries are less striking and
have an unpleasant taste, but would be familiar to
readers of Thomas Culpeper’s The English Physitian of
1652 as a preventive for kidney stones. Presumably
this property derived from the berries’ atropinic compounds, which
could aid in the passage of ureteric stones by relaxing the smooth muscle
of the walls of the ureter. Less desirable would have been the other
possible consequences of ingesting the unripe fruit, including dry
mouth, hallucinations, and death. In this month’s issue of AJKD, Scales et
al describe a more refined approach to the prevention of urinary stones,
reporting the design and rationale of the PUSH clinical trial, which
involves use of a smart water bottle to encourage hydration. See p. 898.
The photograph “Physalis alkekengi L. Rosaceae” is by Dr Henry Oakeley. From the Wellcome Collection,
available under CC BY license.

CONTENTS
QUIZ
Recurrent Nephrolithiasis in a Patient With Hypercalcemia and Normal to Mildly
Elevated Parathyroid Hormone
Alexander Ritter, Rosa Vargas-Poussou, Nilufar Mohebbi, and Harald Seeger

KDOQI COMMENTARY

833 KDOQI US Commentary on the 2020 KDIGO Clinical Practice Guideline on the
Evaluation and Management of Candidates for Kidney Transplantation
Chethan M. Puttarajappa, Carrie A. Schinstock, Christine M. Wu, Nicolae Leca,
Vineeta Kumar, Brahm S. Vasudev, and Sundaram Hariharan

EDITORIALS

857 APOL1, Black Race, and Kidney Disease: Turning Attention to Structural Racism
Jessica P. Cerdeña, Jennifer Tsai, and Vanessa Grubbs

861 Familial Aggregation of CKD: Gene or Environment?

Lucrezia Carlassara, Francesca Zanoni, and Ali G. Gharavi
Related Article, p. 869

863 APOL1 and Preeclampsia: Intriguing Links, Uncertain Causality, Troubling

Implications
John R. Sedor, Leslie A. Bruggeman, and John F. O’Toole
Related Article, p. 879
 Vol 77 - No 6 - June 2021

866 One Step Closer to Developing a National Dialysis Registry in China

Samaya Javed Anumudu, Vishnu Parvathareddy, and Kevin F. Erickson
Related Article, p. 889

ORIGINAL INVESTIGATIONS

869 Familial Aggregation of CKD and Heritability of Kidney Biomarkers in the General
Population: The Lifelines Cohort Study
Jia Zhang, Chris H.L. Thio, Ron T. Gansevoort, and Harold Snieder
Editorial, p. 861

879 Joint Associations of Maternal-Fetal APOL1 Genotypes and Maternal Country of

Origin With Preeclampsia Risk
Xiumei Hong, Avi Z. Rosenberg, Boyang Zhang, Elizabeth Binns-Roemer, Victor David,
Yiming Lv, Rebecca C. Hjorten, Kimberly J. Reidy, Teresa K. Chen, Guoying Wang, Yuelong Ji,
Claire L. Simpson, Robert L. Davis, Jeffrey B. Kopp, Xiaobin Wang, and Cheryl A. Winkler
Editorial, p. 863

889 Estimation of Prevalence of Kidney Disease Treated With Dialysis in China: A Study
of Insurance Claims Data
Chao Yang, Zhao Yang, Jinwei Wang, Huai-Yu Wang, Zaiming Su, Rui Chen, Xiaoyu Sun,
Bixia Gao, Fang Wang, Luxia Zhang, Bin Jiang, and Ming-Hui Zhao
Editorial, p. 866

898 Prevention of Urinary Stones With Hydration (PUSH): Design and Rationale of a
CONTENTS

Clinical Trial
Charles D. Scales Jr, Alana C. Desai, Jonathan D. Harper, H. Henry Lai, Naim M. Maalouf,
Peter P. Reese, Gregory E. Tasian, Hussein R. Al-Khalidi, Ziya Kirkali, and Hunter Wessells,
on behalf of the Urinary Stone Disease Research Network

907 Change in Cardiac Biomarkers and Risk of Incident Heart Failure and Atrial
Fibrillation in CKD: The Chronic Renal Insufficiency Cohort (CRIC) Study
Nisha Bansal, Leila R. Zelnick, Elsayed Z. Soliman, Amanda Anderson, Robert Christenson,
Christopher DeFilippi, Rajat Deo, Harold I. Feldman, Jiang He, Bonnie Ky, John Kusek,
James Lash, Stephen Seliger, Tariq Shafi, Myles Wolf, Alan S. Go, and Michael G. Shlipak,
on behalf of the CRIC Study Investigators

920 Design and Rationale of HiLo: A Pragmatic, Randomized Trial of Phosphate

Management for Patients Receiving Maintenance Hemodialysis
Daniel L. Edmonston, Tamara Isakova, Laura M. Dember, Steven Brunelli, Amy Young,
Rebecca Brosch, Srinivasan Beddhu, Hrishikesh Chakraborty, and Myles Wolf

931 Psychosocial Factors, Intentions to Pursue Arteriovenous Dialysis Access, and Access
Outcomes: A Cohort Study
Jace Ming Xuan Chia, Zhong Sheng Goh, Pei Shing Seow, Terina Ying-Ying Seow,
Jason Chon Jun Choo, Marjorie Wai-Yin Foo, Stanton Newman, and Konstadina Griva
Vol 77 - No 6 - June 2021

941 Mobile Health (mHealth) Technology: Assessment of Availability, Acceptability, and

Use in CKD
Sarah J. Schrauben, Lawrence Appel, Eleanor Rivera, Claudia M. Lora, James P. Lash,
Jing Chen, L. Lee Hamm, Jeffrey C. Fink, Alan S. Go, Raymond R. Townsend, Rajat Deo,
Laura M. Dember, Harold I. Feldman, and Clarissa J. Diamantidis, on behalf of the CRIC Study
Investigators

POLICY FORUM
Perspective
951 Racism and Kidney Health: Turning Equity Into a Reality
Dinushika Mohottige, Clarissa J. Diamantidis, Keith C. Norris, and L. Ebony Boulware

Editorial
963 Reducing the Shortage of Transplant Kidneys: A Lost Opportunity for the US Health
Resources and Services Administration (HRSA)
Frank McCormick, Philip J. Held, Glenn M. Chertow, Thomas G. Peters, and John P. Roberts

Commentary

967 The True Meaning of Financial Neutrality in Organ Donation

Gabriel M. Danovitch, Alexander M. Capron, and Francis L. Delmonico

CORE CURRICULUM IN NEPHROLOGY

CONTENTS
969 Reducing Kidney Function Decline in Patients With CKD: Core Curriculum 2021
Teresa K. Chen, Christopher J. Sperati, Sumeska Thavarajah, and Morgan E. Grams

IN THE LITERATURE EDITORIAL

984 Belimumab in Lupus Nephritis: New Trial Results Arrive During an Exciting Time
for Therapeutics
Nestor Oliva-Damaso and Andrew S. Bomback

CORRESPONDENCE
Research Letters
988 Temporal Relationship of Glycemia With Cardiac Arrhythmias in Patients With
Type 2 Diabetes and CKD
Nisha Bansal, Leila R. Zelnick, Irl B. Hirsch, Dace Trence, Nazem Akoum, and Ian H. de Boer

990 Acute Kidney Injury After the CAR-T Therapy Tisagenlecleucel

Meghan D. Lee, Ian A. Strohbehn, Harish S. Seethapathy, Nifasha Rusibamayila,
Keagan S. Casey, Shruti Gupta, David E. Leaf, Matthew J. Frigault, and Meghan E. Sise

992 Tubular Secretion of Creatinine and Risk of Kidney Failure: The Modification of Diet
in Renal Disease (MDRD) Study
Pranav S. Garimella, Hocine Tighiouart, Mark J. Sarnak, Andrew S. Levey, and Joachim H. Ix
 Vol 77 - No 6 - June 2021

NOTICE

995 Erratum Regarding “Comparability of Plasma Iohexol Clearance Across Population-

Based Cohorts” (Am J Kidney Dis. 2020;76[1]:54-62)

AJKDblog.org

facebook.com/AJKDonline

instagram.com/AJKDonline

twitter.com/AJKDonline
CONTENTS
Quiz
Recurrent Nephrolithiasis in a Patient With
Hypercalcemia and Normal to Mildly Elevated
Parathyroid Hormone
Alexander Ritter, Rosa Vargas-Poussou, Nilufar Mohebbi, and Harald Seeger
Clinical Presentation
A 42-year-old man was referred to our stone clinic for
recurrent episodes of symptomatic nephrolithiasis.
Spontaneous stone passages occurred at the ages of 25,
40, and 42 years with one need for endoscopic stone
removal. Calculi were composed of 100% calcium ox-
alate monohydrate. The patient had not taken any
medication or supplements. Increased fluid intake was
the only intervention for stone prevention. Metabolic
work-up at presentation revealed mild hypercalcemia
and normal to mildly elevated intact parathyroid hor-
mone, even after repletion with oral 25-
hydroxyvitamin D. 1,25-Dihydroxyvitamin D level
and 24-hour urinary calcium excretion were within
normal ranges (Table 1). Physical examination was
unremarkable. Family history was negative for neph-
rolithiasis or hypercalcemia.
• What are the main differential diagnoses of hy-
percalcemia with an elevated PTH level in this
patient?
• What tests help to distinguish between these
two entities?
• Which additional studies help to verify the
diagnosis?
Discussion
What are the main differential diagnoses of
hypercalcemia with a normal to mildly elevated
PTH in this patient?
Hypercalcemia and high-normal or elevated PTH levels
in the context of recurrent nephrolithiasis are strongly
indicative of primary hyperparathyroidism (PHPT),
which is a well-recognized cause of calcium oxalate
stones. Urinary calcium excretion can be normal or
elevated. Parathyroid surgery, which was considered in
this case, is usually the treatment of choice for symp-
tomatic patients. Ultrasound and nuclear imaging did
not show evidence for parathyroid adenoma, which,
however, does not rule out PHPT.
Familial hypocalciuric hypercalcemia (FHH) is a rare
autosomal dominant disease that resembles PHPT, but is
usually benign and not curable by parathyroidectomy. It
is typically characterized by mild hypercalcemia, inap-
propriately normal or elevated PTH (in roughly 20% of
patients),1 low urinary calcium, and a positive family
history. FHH is genetically heterogeneous. It is caused
by inactivating mutations in CASR, the gene encoding
for the calcium-sensing receptor (CaSR) or in genes
encoding two CaSR intracellular partner proteins. CaSR
plays a major role in maintaining extracellular calcium
homeostasis (Fig 1).
We did not measure PTH-related protein in our pa-
tient, as there was no clinical suspicion of malignancy.
Furthermore, if the hypercalcemia were due to malig-
nancy, PTH would typically be suppressed, unlike the
situation in this case.
What laboratory tests help to distinguish
between these two entities?
The calcium-creatinine clearance ratio (CCCR) helps to
discriminate between PHPT and FHH in many cases,
given that 24-hour urinary calcium excretion is
often <100 mg/d in FHH and >200-300 mg/d in
PHPT. In our patient, CCCR confirmed significant
hypocalciuria with a ratio <0.01 (0.0096 at initial
presentation, 0.0044 a month and a half later), sug-
gestive of FHH. In PHPT, the ratio is usually >0.02.
However, particularly in patients with PHPT and
concomitant 25-hydroxyvitamin D deficiency or
extremely low calcium intake, urinary calcium may be
low-normal or low (CCCR ≤ 0.02; urinary calcium
excretion < 200-300 mg/d). Of note, atypical FHH
cases with CCCR > 0.01 (or even hypercalciuria) have
been reported.
A recent cross-sectional study compared a large
cohort of FHH to PHPT patients and found considerable
overlap in the phenotypes, making the distinction based
on laboratory data alone difficult.1
What additional studies help to verify the
diagnosis?
Family screening can often establish the diagnosis of
QUIZ
FHH owing to its high penetrance. Since FHH is usually
asymptomatic, an active investigation for abnormalities
of serum and urinary calcium should be performed if
feasible, even if the family history is negative. Such an
analysis could not be conducted in our patient’s family.
Genetic testing is recommended in patients with atyp-
ical presentation (Fig 1), like our patient (neph-
rolithiasis, negative family history). Sequencing of CASR
disclosed the patient was heterozygous for a previously
described inactivating mutation (cytosine-to-thymine
 Quiz
Table 1. Laboratory Test Results
Laboratory Findings Initial Presentation + 1.5 Months + 10 Months Reference Range
Blood
Calcium, mg/dL 10.7 11.0 11.0 8.4-10.2
Calcium ion, mEq/L 2.6 2.7 2.7 2.2-2.6
Magnesium, mEq/L 1.8 1.9 ND 1.3-2.1
Phosphorus (inorganic), mg/dL 2.2 2.7 3.2 2.7-4.5
Creatinine, mg/dL 1.0 1.0 1.0 0.7-1.2
iPTH, pg/mL 64 63 117 15-65
25-Hydroxyvitamin D, ng/mL 17 ND 29 >20
1,25-Dihydroxyvitamin D, pg/mL 48 ND ND >26-95
Urine
Volume, mL/d 2,150 1,950 ND
pH 6.0 6.0 ND 5.0-7.5
Citrate, mg/d 430 626 ND 307-865
Oxalate, mg/d 20 32 ND <45
Phosphorus, mg/d 839 859 ND 400-1,300
Calcium, mg/d 160 81 ND <300
Magnesium, mEq/d 4.9 4.1 ND 6.0-10.0
Creatinine, g/d 1.55 1.69 ND 0.81-2.01
Conversion factors for units: calcium in mg/dL to mmol/L, ×0.2495; calcium ion in mEq/L to mmol/L, ×0.5; citrate in mg/dL to μmol/L, ×52.05; creatinine in mg/dL
to μmol/L, ×88.4; magnesium in mEq/L to mmol/L, ×0.5; oxalate in mg/L to μmol/L, ×11.1; phosphorus (inorganic) in mg/dL to mmol/L, ×0.3229; 1,25-
dihydroxyvitamin D in pg/mL to pmol/L, ×2.6; 25-hydroxyvitamin D in ng/mL to nmol/L, ×2.496.
Abbreviations: iPTH, intact PTH; ND, not determined.

substitution at nucleotide 788 of the coding sequence,
predicted to lead to a threonine-to-methionine substi-
tution at amino acid 263 [c.788C>T p.Thr263Met]).2
It has recently been appreciated that FHH is not only
a differential diagnosis of PHPT in patients with hy-
percalcemia and abnormally elevated PTH, but is also
associated with kidney stones.1,3 In contrast to PHPT,
literature about nephrolithiasis in the context of FHH is
sparse and the underlying pathophysiology is not well
Different types of FHH:
• FHH1: CASR (calcium-sensing receptor)
• FHH2: GNA11 (G-protein subunit α11)
• FHH3: AP2S1 (adaptor-related protein
complex 2, sigma subunit) Gα11
AP2
Recommendation for genetic testing Renal calcium absorption
(most relevant examples): PTH secretion
• CCCR between 0.01 and 0.02 (without vitamin D deficiency)
• FHH phenotype, negative family screening
• Atypical presentation, family screening not feasible
• Familial form of hyperparathyroidism suspected
QUIZ
understood, especially given the presence of hypo-
calciuria. In a French cohort, 21.7% of PHPT patients
had a history of nephrolithiasis compared to 7.7% of
patients with CASR mutations. Notably, all FHH patients
with a history of nephrolithiasis were genetically
identified as FHH type 1.1 In a study from Italy in which
patients were selected for genetic testing based on
clinical and biochemical compatibility with FHH, a
history of nephrolithiasis was present in up to 67% of
patients with CASR mutations.3 Furthermore, several
studies, including a large genome-wide association
study from Iceland, showed polymorphisms of the CASR
gene regulatory region leading to reduced CaSR
expression are associated with nephrolithiasis.4,5
There is increasing evidence that the CaSR plays a
complex role in contributing to the urinary equilibrium
between inhibitors and promotors of calcium oxalate and
phosphate precipitation. Downregulated tubular CaSR
expression may impair dilution capacity, distal urinary
acidification, and citrate production, while also increasing
phosphaturia, thus resulting in calcium stone formation.

Figure 1. Overview of familial hypocalciuric hypercalcemia
(FHH). Each type is caused by an inactivating mutation in
the indicated gene. The calcium-sensing receptor (CaSR)
mediates the negative feedback mechanism of extracellular
calcium on renal calcium absorption and parathyroid hor-
mone (PTH) excretion. Gα11 (encoded by GNA11) is
involved in its signaling, while adaptor-related protein complex
2 (AP2) is essential for clathrin-mediated endocytosis. Some
authors recommend genetic testing of the more common
types first (FHH1 > FHH3 > FHH2). Abbreviation: CCCR,
calcium-creatinine clearance ratio.
In particular, decreased water excretion in the collecting
duct and reduced citrate excretion in the proximal tubule
could contribute to calcium oxalate precipitation.5 Future
studies will have to further dissect the specific functions of
the CaSR in different tubular segments that protect the
kidney against negative effects of calcium salts.
This case explicitly illustrates that misdiagnosing
FHH as PHPT as a cause for nephrolithiasis might have
severe consequences, such as unnecessary para-
thyroidectomy. In stone formers with hypercalcemia
and inappropriately normal or mildly elevated PTH,
Quiz
determination of CCCR may serve as a useful tool to
differentiate between FHH and PHPT. If the diagnosis is
in doubt, genetic analysis is recommended.
Final Diagnosis
Familial hypocalciuric hypercalcemia (FHH) with
recurrent nephrolithiasis.
Article Information
Authors’ Full Names and Academic Degrees: Alexander Ritter,
MD, Rosa Vargas-Poussou, MD, PhD, Nilufar Mohebbi, MD, and
Harald Seeger, MD.
Authors’ Affiliations: Division of Nephrology, University Hospital
ˇ
Zurich, Zurich, Switzerland (AR, NM, HS), and Hopital Europ
een
Georges Pompidou, D
epartement de G
en
etique, INSERM UMR
970, Paris, France (RV-P).
Address for Correspondence: Alexander Ritter, MD, Division of
Nephrology, University Hospital Zurich, R€ amistrasse 100, 8091
Zurich, Switzerland. Email: alexander.ritter@usz.ch
Support: There was no funding support for this work.
Financial Disclosure: The authors declare that they have no
relevant financial interests.
Acknowledgements: We thank Andrew Hall, MD, PhD (Institute
of Anatomy, University of Zurich) for his valuable comments.
Patient Protections: The authors declare that they have
obtained consent from the patient reported in this article for
publication of the information about him that appears within
this Quiz.
Peer Review: Received May 22, 2020. Direct editorial input from
a Deputy Editor. Accepted in revised form September 2, 2020.
Publication Information: © 2021 by the National Kidney Foun-
dation, Inc. doi 10.1053/j.ajkd.2020.09.022
References
1. Vargas-Poussou R, Mansour-Hendili L, Baron S, et al. Fa-
milial hypocalciuric hypercalcemia types 1 and 3 and pri-
mary hyperparathyroidism: similarities and differences.
J Clin Endocrinol Metab. 2016;101(5):2185-2195.
2. Guarnieri V, Canaff L, Yun FH, et al. Calcium-
sensing receptor (CASR) mutations in hypercalcemic
states: studies from a single endocrine clinic over
three years. J Clin Endocrinol Metab. 2010;95(4):
1819-1829.
3. Stratta P, Merlotti G, Musetti C, et al. Calcium-sensing-
related gene mutations in hypercalcaemic hypocalciuric
patients as differential diagnosis from primary hyperpara-
thyroidism: detection of two novel inactivating mutations in
an Italian population. Nephrol Dial Transplant. 2014;29(10):
1902-1909.
4. Oddsson A, Sulem P, Helgason H, et al. Common and rare
variants associated with kidney stones and biochemical
traits. Nat Commun. 2015;6:7975.
5. Vezzoli G, Macrina L, Magni G, Arcidiacono T. Calcium-
sensing receptor: evidence and hypothesis for its role in
nephrolithiasis. Urolithiasis. 2019;47(1):23-33.
QUIZ
KDOQI Commentary

KDOQI US Commentary on the 2020 KDIGO Clinical

Practice Guideline on the Evaluation and Management of
Candidates for Kidney Transplantation
Chethan M. Puttarajappa, Carrie A. Schinstock, Christine M. Wu, Nicolae Leca, Vineeta Kumar,
Brahm S. Vasudev, and Sundaram Hariharan

Evaluation of patients for kidney transplant candidacy is a comprehensive process that involves a Complete author and article
detailed assessment of medical and surgical issues, psychosocial factors, and patients’ physical and information provided before
references.
cognitive abilities with an aim of balancing the benefits of transplantation and potential risks of surgery
and long-term immunosuppression. There is considerable variability among transplant centers in their Am J Kidney Dis.
approach to evaluation and decision-making regarding transplant candidacy. The 2020 KDIGO 77(6):833-856. Published
online March 18, 2021.
(Kidney Disease: Improving Guidelines Outcome) clinical practice guideline on the evaluation and
management of candidates for kidney transplantation provides practice recommendations that can doi: 10.1053/
serve as a useful reference guide to transplant professionals. The guideline, covering a broad range of j.ajkd.2020.11.017
topics, was developed by an international group of experts from transplant and nephrology through a © 2020 by the National
review of literature published until May 2019. A work group of US transplant nephrologists convened Kidney Foundation, Inc.
by NKF-KDOQI (National Kidney Foundation–Kidney Disease Quality Initiative) chose key topics for
this commentary with a goal of presenting a broad discussion to the US transplant community. Each
section of this article has a summary of the key KDIGO guideline recommendations, followed by a brief
commentary on the recommendations, their clinical utility, and potential implementation challenges.
The KDOQI work group agrees broadly with the KDIGO recommendations but also recognizes and
highlights the decision-making challenges that arise from lack of high-quality evidence and the need to
balance equity with utility of organ transplantation.

of transplantation may be smaller and come with higher

Because they are designed to reflect the views and rec-
ommendations of the responsible KDOQI Commentary
work group and they are reviewed and approved by KDOQI
and NKF leadership, KDOQI Commentaries are not peer
reviewed by AJKD. This article was prepared by a KDOQI
Commentary work group comprising the authors and
chaired by Dr Sundaram Hariharan. It was reviewed and
approved by the NKF Scientific Advisory Board and the
KDOQI Chair and Vice Chairs.
Introduction
Kidney transplantation is the preferred treatment option
for patients with kidney failure. It offers the best chance of
improving both quality and quantity of life but requires
careful consideration of the risks and benefits. Patients with
kidney failure often have serious medical comorbidities
and challenging financial and psychosocial issues that in-
crease the risk of complications and poor outcomes
following transplantation. Key among these complications
are the risk of infection, malignancy, postoperative car-
diovascular complications, and complications arising from
nonadherence to therapy after transplantation. Benefits of
kidney transplantation have been demonstrated even
among patients considered as “high-risk” such as those
who are elderly, those with obesity, and those with
longstanding diabetes with vascular and coronary artery
disease. However, for such high-risk groups, the benefits
risk of complications. Long-term success of transplantation
depends on reducing postsurgical complications, carefully
balancing the risks and benefits of immunosuppression
over time, and developing strategies that ensure long-term
adherence to transplant follow-up and medications. The
comprehensive pretransplant evaluation of a potential
transplant candidate has several objectives, but chief
among them are: 1) identify absolute contraindications for
transplantation; 2) recognize medical, surgical, psychoso-
cial, and social risk factors that will require optimization
prior to transplantation; 3) identify and update recom-
mended screening and health maintenance procedures; 4)
assess immunological risk; and 5) offer education, guid-
ance, and counseling to patients about several aspects of
kidney transplantation. Implementing this in an evidence-
based manner requires incorporating and adopting new
research findings from several disciplines of medicine.
Additionally, the evaluation and decision-making pro-
cesses needs to adapt to changes in organ transplant pol-
icies instituted at a national level.
The KDIGO (Kidney Disease: Improving Guidelines
Outcome) clinical practice guideline on the evaluation and
management of candidates for kidney transplantation
published in April 2020 provides suggestions and rec-
ommendations that cover all major areas of pretransplant
evaluation of kidney transplant candidates.1 While many of
the guidelines are backed by good-quality evidence, there
was a lack of high-quality studies in some areas, which

AJKD Vol 77 | Iss 6 | June 2021 833


necessitated recommendations based on expert opinion.
In this article, a work group from NKF-KDOQI (National
Kidney Foundation–Kidney Disease Quality Initiative) pro-
vides an overview of a select group of guideline recommen-
dations, commenting on their clinical utility and applicability
to kidney transplantation in the United States.
KDOQI Commentary Process
The KDOQI Steering Committee selected the commentary
chair, who was given the mandate to assemble a work
group comprising transplant nephrologists from a diverse
group of academic transplant centers in the United States.
The work group was asked to identify the important
clinical areas to be addressed in this commentary. For each
topic, this commentary presents the relevant KDIGO
guideline statements (reproduced with permission of
KDIGO), a short commentary, discussion of clinical utility,
and consideration of implementation and its associated
challenges. This content of this commentary was deter-
mined by discussion among the KDOQI work group, and
all work group members reviewed and approved the
commentary after reaching consensus. The article was also
reviewed and approved by the NKF Scientific Advisory
Board and KDOQI leadership.
Guideline Statements and Commentary
Transplant Recipients of Older Age
2.1: Consider age in the context of other comorbidities,
including frailty, that may impact outcome when deciding
about suitability for kidney transplantation (Not Graded).
2.1.1: We recommend not excluding patients from kidney
transplantation because of age alone (1A).
Commentary
The KDIGO guideline highlights the need to evaluate the
transplant candidacy of older individuals in the context of
their physiologic reserve, mental health, and medical
comorbidities rather than an age-based exclusion. Life
expectancy in the United States for individuals age 65 is
approximately 20 years, and has improved significantly
over the last decade.2 Individuals >75 years of age account
for nearly 25% of those beginning kidney replacement
therapy for kidney failure, and the elderly make up the
fastest growing segment of the transplant population.3
However, the likelihood of being waitlisted and success-
fully getting a transplant is substantially lower among
older patients than among younger patients. While the US
Renal Data System Annual Data Report shows a survival
advantage of kidney transplantation over remaining on the
waiting list in all age groups, this advantage diminishes
with age, and many in the transplant and nephrology
community still struggle with the decision of when to
offer transplant to the elderly. Large cohort studies
834
Puttarajappa et al
demonstrate that the time to offset the increased risk of
early posttransplant mortality is longer for older patients
and varies with donor type and recipient risk factors.4,5 For
moderate- to high-risk older patients receiving a deceased-
donor kidney, the offset does not occur until around 1 year
after transplantation and is the longest for recipients of
extended criteria kidneys.5 Thus, assessment of comor-
bidity, including frailty (discussed later in this commen-
tary), remains critical in the risk analysis and decision to
transplant older individuals. The current approach to pre-
transplant evaluation remains focused on discrete comor-
bidities with an emphasis on cancer screening and
cardiovascular evaluation for older individuals. There
is increasing recognition of the need for more compre-
hensive assessments, and various scoring systems and co-
morbidity indices have been developed to assess
combinations of factors as opposed to age alone as a single
comorbidity. Attempts to quantify and assess frailty (walk
speed, sit-to-stand test, grip strength), cognitive function,
and social support are also gaining traction. Given the long
waiting times for kidney transplantation, it is important to
assess these at each evaluation to monitor for trends. The
KDIGO guideline recommends not excluding patients
solely based on age but does not provide specific guidance
for accepting older recipients for transplantation.
It is equally important to recognize the impact of se-
lective and timely acceptance of organ offers, balancing the
benefits of transplantation with the use of an expanded
donor pool, including high Kidney Donor Profile Index
(KDPI), hepatitis C virus nuclear acid amplification
test–positive (HCV NAAT+), donation after cardiac death
(DCD), and Public Health Services (PHS) increased-risk
organs. Finally, the KDIGO guideline does not address
quality of life changes from transplantation, which, for
some elderly patients, may be the primary reason for
seeking transplantation.
Clinical Utility
As both life expectancy and, more importantly, active life
expectancy has increased over the past decade, it is
important to not exclude patients from kidney trans-
plantation solely for chronologic age. A growing body of
literature supports the benefits of kidney transplantation in
the elderly. The assessment of frailty, cognitive function,
and other comorbidities beyond age alone are important to
consider during the evaluation of elderly recipients.
Implementation and Challenges
The first challenge begins with referral to transplantation;
awareness among community nephrologists regarding the
need for preemptive referral will increase timely evaluation
of eligible candidates. Given the long waiting time for
deceased-donor kidney transplantation, general nephrolo-
gists and the transplant team should educate and counsel
elderly patients regarding the benefits of exploring living-
donor kidney transplantation as an option.
AJKD Vol 77 | Iss 6 | June 2021
Puttarajappa et al
In 2019, the Scientific Report on Transplant Registry
(SRTR) added the deceased-donor kidney transplantation
rate and waitlist mortality to their list of transplant
center metrics. While these metrics are adjusted for age,
they may not fully account for frailty and cognitive is-
sues that are common yet vary among the elderly. This
may also have the unintended consequence of decreasing
access to listing and transplantation for older candidates.
Some transplant centers may not be aggressive in listing
older patients due to concerns for inferior outcomes,
both on waiting list and after transplantation. A potential
solution may be for SRTR to consider reporting trans-
plant center program–specific reports separately for pa-
tients >70 years.
The initial evaluation of comorbidity, frailty, cognitive
function, and social support are necessary before listing
and require ongoing surveillance while on the transplant
wait list. Assessment of frailty using gait speed, the sit-to-
stand test, and hand grip strength require additional
training and time for clinic personnel but may improve
assessment of transplant candidacy. Educating patients
regarding risks and benefits for nonstandard donor offers
(high KDPI, DCD, PHS increased risk, HCV NAAT+) will
also require additional time but may help reduce time to
transplantation. Consideration of human leukocyte anti-
gen (HLA)-A2 to -B transplants should also be made for
appropriate patients. Calculators such as SRTR decision
tool (www.srtr.org/reports-tools/kidney-transplant-
decision-tool/) and the ones developed by Johns Hop-
kins University (www.transplantmodels.com/) may be
used to estimate probabilities of clinical outcomes on the
waiting list and after transplantation. These could help
with shared decision-making. However, it is important
to note that these models do not consider all comor-
bidity, social support, frailty, and cognitive issues. Also,
even though the estimated survival benefit may vary
based on age and the KDPI of transplanted kidney, the
calculators always yield a survival benefit for trans-
plantation (even with a marginal organ such as KDPI
>85) compared to staying on the wait list or on dialysis.
Hence, decisions regarding candidacy will still need to
be individualized.
New optional care models introduced by the Centers for
Medicare & Medicaid Services (Kidney Care First and
Comprehensive Kidney Care Contracting Graduated, Pro-
fessional, and Global Models) which incentivize cost-
effective care of chronic kidney disease (CKD) patients,
including those with kidney failure, will likely lead to
additional scrutiny of the cost of transplant and post-
transplant care in older recipients.6,7
Like the general population, the population with
kidney failure is aging. Thus, there is a growing need to
improve early referral and to address the medical, social,
and psychological barriers to transplantation. This will
lead to the timely and comprehensive evaluation of
AJKD Vol 77 | Iss 6 | June 2021
older patients to ensure their access to transplantation
and will improve the quality and quantity of life that
transplant offers.
Psychosocial Assessment
4.1: We suggest performing a psychosocial assessment in all
candidates (2D).
4.1.1: Refer candidates to a health care professional
experienced in the psychosocial aspects of kidney
transplantation (eg, social worker, psychologist,
psychiatrist, psychiatric nurse/nurse practitioner) to
perform this assessment (Not Graded).
4.1.2: Use measurement tools completed by the patient
and/or evaluating clinician to supplement the
assessment (Not Graded).
4.1.2.1: We suggest not using measurement tools
in isolation to determine transplant candi-
dacy (2D).
4.1.3: Refer candidates with a diagnosable psychiatric or
psychological condition, substance use disorder or
nonadherence for pre-transplant counseling and
services to enhance the likelihood of a favorable
post-transplant outcome (Not Graded).
4.2: We recommend not transplanting patients with an un-
stable psychiatric disorder that affects decision-making or
puts the candidate at an unacceptable level of post-
transplant risk (1C).
4.3: We recommend not transplanting patients with ongoing
substance use disorder that affects decision-making or
puts the candidate at an unacceptable level of post-
transplant risk (1C).
4.4: We suggest that patients without current social support
be considered for kidney transplantation if they are able to
care for themselves and have an identified support plan in
place prior to transplantation (2D).
Commentary
Transplant success is highly dependent on patients’ ability
to adhere to clinic follow-up, laboratory tests, and medi-
cation intake. Presence of serious psychiatric or psycho-
social barriers therefore pose a significant threat to
successful posttransplant outcomes. Hence, the KDOQI
work group agrees that patients with unstable psychiatric
disorders and significant substance use disorders are not
suitable for transplantation. Identification of underlying
psychosocial issues is complex and will thus require
assessment from a trained health care professional. The
KDIGO guideline highlights the lack of evidence for using
psychometric measurement tools as the sole decision-
making criteria for transplantation but acknowledges that
they can complement the assessment of a trained profes-
sional. Guideline statement 4.4 addresses the key issue of
social or caregiver support, and recommends favoring
transplantation for patients capable of self-management
even if they lack current social support, provided they
835

have some pretransplant support plan. This statement is
vague and does not provide clarity on the type of support
system necessary. Psychosocial challenges among trans-
plant candidates can be profound and often difficult to
reverse completely, thus hindering the evaluation process,
time to listing, and success after transplantation. Evaluating
social support should include considering how the support
system can complement the patient’s ability for self-care
and management of chronic health issues. Kidney disease
in the United States disproportionately affects minorities
and patients from lower socioeconomic strata. Thus, to
avoid worsening inequities, it is imperative to consider
patients’ ability to manage posttransplant issues in situa-
tions with minimal or marginal social support. Notably,
the KDIGO guideline does not address specific substance
use disorders. Given the expansion of legalization of
medical and recreational marijuana use, guidelines on
approaching patients with marijuana use may have been
useful. The legalization of marijuana in certain geographic
areas makes it difficult to make a statement that is
acceptable to all. Transplant centers vary in their approach
to offering kidney transplantation for patients with
ongoing marijuana use.8 While concern has been raised
regarding the use of marijuana and posttransplant adher-
ence, interaction with immunosuppressants, posttransplant
psychosocial issues, and clinical outcomes,8-12 the KDOQI
work group acknowledges that the evidence continues to
evolve and further research is necessary before major
recommendations can be made. The guideline also leaves
out issues related to monitoring and psychosocial evalua-
tions when patients are on the waiting list.
Clinical Utility
Psychosocial assessment of transplant candidates is a key
component of pretransplant assessment and is performed
routinely in all US transplant programs. There is high
prevalence of cognitive impairment and psychiatric and
substance use disorders among potential candidates as well
as among those on the transplant wait list.13 It is critical to
adequately stabilize mood disorders pretransplant, partic-
ularly depression, given its association with inferior post-
transplant outcomes.14 Evidence, however, is mixed for
the association between social support and poor outcomes
following transplantation.15-17 The studies reporting
impact of social support may suffer from selection bias
since patients with major issues are often not referred for
transplantation or excluded at the time of evaluation.
Implementation and Challenges
It is accepted that psychosocial factors that may nega-
tively affect transplant outcomes should be carefully
screened for prior to transplantation. Programs may have
different structural components to this assessment but
generally use social workers to perform the initial
836
Puttarajappa et al
psychosocial assessment, followed by referral of select
patients for behavioral health evaluations. The thresholds
at which psychiatric disorders are considered a contra-
indication for transplantation vary among transplant
centers. Hence, the KDIGO guideline makes a broad but
clear recommendation regarding the importance of
assessing the impact of patient’s psychiatric disorder on
decision-making capabilities and risk of poor outcomes
posttransplant. Though tools that assess transplant read-
iness are available, such as the Psychosocial Assessment
of Candidates for Transplant (PACT), Stanford Integrated
Psychosocial Assessment for Transplant (SIPAT), and
Transplant Evaluation Rating Scale (TERS), it is unclear if
they can be used as a standalone assessment.18,19 The
quantitative and predictive aspects of these scales have
not been thoroughly evaluated and validated. Thus,
while a threshold score on these scales cannot be used
for deciding transplant candidacy, they can help stan-
dardize psychosocial evaluations. Additional tools such as
Patient Health Questionnaire 9 (PHQ-9) for depression
and the Montreal Cognitive Assessment (MOCA) can be
used to screen select patients identified as high-risk on
routine psychosocial assessment.
Transplant teams should also corroborate medical and
psychosocial history and issues with other health care
providers, including dialysis unit staff and social workers,
and should review prior psychosocial evaluations. This will
provide an additional layer of confirmation regarding the
patient’s ability to adhere to posttransplant regimens. A
period of stability with regard to serious mood or psy-
chotic disorders and without hospitalization for psychiatric
illness is reasonable prior to active listing for
transplantation.
Lack of social support accounts for 10%-20% of denials
for kidney transplant listing.20 There are no standardized
criteria for defining adequate social support, and transplant
programs and clinicians differ in their approach to use of
social support criteria for determining transplant candi-
dacy.20 This lack of standardized criteria may accentuate
disparities in access to transplantation. While evidence
linking social support and transplant outcomes is not
strong, findings suggest that social support does play an
important role in facilitating good posttransplant out-
comes.15-17 Adequate social support is especially vital for
patients with complex medical issues, older age, physical
or mental impairments, and those with a language barrier.
Thus, while programs may not all have a clear policy to-
ward social support as a criterion for transplant candidacy,
evaluating social support in the context of the overall
psychosocial assessment will allow for an individualized
approach. This will enable improving transplant access to
some patients who may have marginal support systems but
can otherwise demonstrate good ability to adhere to
posttransplant follow-up.
AJKD Vol 77 | Iss 6 | June 2021
Puttarajappa et al
Nonadherence to Therapy
5.1: Assess adherence and adherence barriers pre-
transplantation to allow for appropriate education, coun-
seling and post-transplant surveillance (Not Graded).
5.2: Refer candidates with a history of health-compromising
nonadherent behavior or identified adherence barriers
for adherence-based education and counseling pre-
transplant (Not Graded).
5.3: We suggest that candidates with a history of graft loss
due to nonadherence undergo adherence-based coun-
seling prior to re-transplantation (2D).
5.4: We recommend that candidates with a history of non-
adherence be considered for transplantation unless there
is ongoing, health-compromising nonadherent behavior
(1D).
Commentary
Nonadherence following transplantation is highly preva-
lent, increases with time after transplant, and is a major
cause of rejection and graft loss. This may manifest in the
domain of medication taking and/or missed laboratory
tests and clinic appointments. The KDIGO guideline pro-
vides recommendations for a comprehensive pretransplant
evaluation for nonadherence risk factors, suggests using
adherence counseling and education for at-risk patients,
and recommends avoiding transplantation in the presence
of serious ongoing nonadherence behavior. These rec-
ommendations are in line with prior recommendations
from other groups.21,22 The guideline does not provide
recommendations on specific methods to use for assessing
nonadherence risk given the lack of reliable and validated
tools. Consistent with other recommendations, the
guideline recommends not excluding patients with prior
allograft loss due to nonadherence provided there is no
ongoing nonadherence behavior. The KDOQI work group
agrees with this recommendation, particularly when prior
nonadherence was secondary to adolescence or financial
issues, provided patients demonstrate improved under-
standing and maturity and all financial barriers have been
resolved satisfactorily. However, it should be emphasized
that prior nonadherence does predict future nonadherence,
and every effort at education, counseling, and rectifying
barriers to nonadherence should be addressed prior to
listing for transplantation.
Clinical Utility
Nonadherence, particularly to immunosuppressive medi-
cations, contributes to premature allograft loss with asso-
ciated morbidity and mortality. Additionally, increased
sensitization limits future transplant potential. Premature
transplant failure is a major loss to society given the limited
number of organs available for transplantation. Thus,
transplant teams have a key role in assessing nonadherence
risk, making decisions regarding transplantation, and
providing appropriate tools for rectifying nonadherence.
AJKD Vol 77 | Iss 6 | June 2021
Implementation and Challenges
Several factors contribute to nonadherence: poor health
literacy, certain health beliefs, low socioeconomic status,
cognitive impairment, limited social support, psychiatric
disorders, and substance abuse. It is thus important to
identify all potential barriers for posttransplant non-
adherence prior to listing for transplantation. All efforts
should be made to assess overall adherence patterns as
they relate to dialysis, other medical visits, and
medication-taking. While nonadherence on the waiting
list can be gauged by assessing the number of inap-
propriately missed or shortened dialysis treatments, large
interdialytic weight gain, and poor phosphorus, hyper-
tension, and diabetes control, it is important to recog-
nize that nonadherence to some aspects of dialysis
regimen may not be always be associated with post-
transplant nonadherence.23,24
The KDIGO guideline recommends adherence-based
education and counseling but does not provide specifics
about how such interventions should be implemented.
Outside of the teaching provided as part of the transplant
evaluation process, there is no specific adherence-based
counseling that is available for clinical practice. Addi-
tionally, it is unclear if pretransplant adherence education
will have long-lasting effects. There is evidence from
nonkidney solid organ transplantation that techniques
such as motivational interviewing may provide a longer
duration of improvement in adherence.25 Similar studies
in kidney transplantation, however, are lacking, and
further investigation is required. In addition, access to
trained professionals providing adherence-based coun-
seling may be limited. Programs should at least consider
augmenting education and counseling at repeated in-
tervals for specific patient groups with certain high-risk
characteristics such as adolescents and those with
limited support system and financial limitations to reduce
risk of posttransplant nonadherence. Decisions regarding
retransplantation for those with prior nonadherence are
often challenging. Generally, in accordance with all
guidelines, retransplantation for such patients is routinely
practiced in the United States provided patients can
demonstrate improved understanding and insight into
posttransplant adherence needs and all contributing
financial and social barriers are eliminated or reduced to
an acceptable degree.
In summary, given the contribution of nonadherence to
inferior outcomes and the numerous risk factors, it is
crucial that transplant programs carefully evaluate and
mitigate nonadherence risk through education and finan-
cial and social support to allow successful transplantation.
Even with adequate pretransplant consideration, post-
transplant nonadherence will remain an issue, though
there is hope that innovative technological solutions
(electronic health records and smartphone-based technol-
ogies) and behavioral interventions will help improve
837

adherence. Increasing research funding for studies evalu-
ating nonadherence, including interventions, will pave a
path toward improving long-term transplant outcomes.
This will be particularly important in the ongoing social
and economic crises resulting from the coronavirus disease
2019 (COVID-19) pandemic.
Transplant Recipients With Obesity
7.1: We recommend candidates to have their body habitus
examined by a transplant surgeon at the time of evaluation
and while on the waiting list (1B).
7.1.1: We suggest that candidates not be excluded from
transplantation because of obesity (as defined by
body mass index or waist-to-hip ratio) (2B).
7.1.2: We suggest weight loss interventions be offered to
candidates with obesity prior to transplantation
(2D).
Commentary
Obesity-related guidelines are deliberately broad regarding
consideration for transplantation in recognition of the di-
chotomy between survival advantage conferred by trans-
plantation over dialysis for patients with obesity and the
increased risks for suboptimal outcomes post–kidney
transplant experienced by patients with versus without
obesity.26,27 In addition, a specific body mass index (BMI)
cutoff is difficult to assign, as the experience and tech-
niques of individual transplant centers and surgeons may
influence acceptance criteria. The KDIGO guideline is
generally in agreement with prior guidelines21,22 but dif-
fers in not making recommendations based on specific
obesity cutoffs for transplantation. The complexity of
transplant surgical intervention and the increased risk of
wound complications with obesity is reflected in the
recommendation to have the patient’s body habitus
examined by a transplant surgeon prior to transplantation.
The guideline also suggests offering weight-loss in-
terventions for transplant candidates with BMI >35 kg/m2
to reduce complications.
Although in overall agreement with the recommenda-
tions, the KDOQI work group feels that the KDIGO
guideline should have gone a step further and offered
recommendations regarding a BMI cutoff (of >40 kg/m2)
at which transplant candidacy needs to be evaluated and
weight-loss interventions such as bariatric surgery should
be considered. This is because patients with BMI >40 kg/
m2 are at high risk of staying inactive on the waiting list28
and, even when transplanted, are at higher risk of peri-
operative complications, with survival benefit from trans-
plantation being smaller.27 Additionally, in the general
population, there are recommendations for patients with
diabetes mellitus and BMI >40 kg/m2 to consider bariatric
surgery as a potential weight-loss intervention.29
838
Puttarajappa et al
Finally, the guideline recommendations on obesity are
centered around the risk for postsurgical issues and do not
stress sufficiently the contribution of obesity to long-term
metabolic complications, including elevated risk for new-
onset diabetes after transplantation (NODAT).
Clinical Utility
Obesity prevalence is substantial, with 6.7% patients in the
United States having class III obesity (BMI >40 kg/m2).
Based on data from the REGARDS (Reasons for Geographic
and Racial Differences in Stroke) study, the association of
obesity with kidney failure appears to be mediated by
metabolic syndrome.30 While obesity is associated with
lower risk of death among patients on maintenance dialysis
(despite the independent association with cardiovascular
disease), the impact of obesity on transplantation out-
comes is complex. Despite a survival advantage of trans-
plantation when compared with remaining on dialysis,
obesity also correlates with poorer outcomes posttrans-
plant. Patients with obesity have a higher risk of death,
delayed graft function, rejection, wound complications,
and prolonged hospitalization stays compared with those
without obesity.31 Thus, obesity remains an important
issue for potential transplant candidates.
Implementation and Challenges
The exact underlying mechanisms for the paradoxical
survival benefit conferred by obesity for dialysis patients
are unclear, but it is hypothesized that patients with high
BMI may have lower chronic inflammation, better nutri-
tional status, and higher physical activity.32–34 However,
similar outcomes have not been shown after trans-
plantation, and studies consistently have shown an
increased risk of complications with higher BMI.31 Most of
the transplant programs in the United States have BMI
cutoffs above which patients need to demonstrate sus-
tained weight loss to be considered eligible for trans-
plantation. It is well known that patients inactive on the
waiting list due to obesity have a significant challenge in
achieving the necessary weight loss to become eligible for
transplantation.28
Programs should consider obesity in the context of
coexisting comorbidities, body habitus, and functional
status. Such an approach may allow for higher BMI
cutoffs in active patients with no major cardiovascular or
age-related comorbidities. Additionally, since some
muscular patients may be disadvantaged by use of BMI
criteria alone, careful attention and a case-based
approach will be necessary to minimize this bias. Also,
since patients on peritoneal dialysis can gain excessive
weight, it may be prudent for transplant physicians to
discuss with the referring nephrologists the potential for
a switch to hemodialysis in select patients; this will need
careful consideration of the benefits that come with
home dialysis and dialysis access issues. As
AJKD Vol 77 | Iss 6 | June 2021
Puttarajappa et al
recommended by a guideline developed by KDOQI and
the Academy of Nutrition and Dietetics, patients may
also benefit from medical nutrition therapy, which in-
volves working with a registered dietitian nutritionist to
optimize nutritional status and minimize risks due to
nutrition-related comorbidity.35,36
The evolution of bariatric surgery techniques has been
promising, and they offer an option for some patients to
achieve the desired weight loss. While there are risks with
surgical approaches to obesity, these risks do not appear to
be significantly higher among carefully selected dialysis
patients compared to the general population.37 Bariatric
surgery approaches to correct obesity may offer additional
benefits far beyond better access to transplantation.38
Single-center studies of transplant centers offering bariat-
ric surgery options to transplant candidates with obesity
have shown encouraging outcomes.38,39 Finally, select
centers are offering robotic transplantation for individuals
with severe obesity, and this can be considered as an
alternative option.26
Obesity continues to represent a substantial barrier to
transplantation that requires a multidisciplinary approach
for evaluation and potential interventions to control its risk
associations, which may include bariatric surgical options.
Solving the obesity challenge for potential transplant re-
cipients remains a daunting task, at least in the near future.
Transplant Recipients With Frailty
7.2: We suggest that candidates be assessed for frailty at the
time of evaluation and while on the waitlist to inform post-
transplant risk and enable optimization strategies, such as
pre-operative rehabilitation (2C).
Commentary
The KDIGO guideline recognizes the importance of frailty
as an important negative prognostic factor in trans-
plantation. Several studies have shown that frailty is highly
prevalent and correlates with poor outcomes during wai-
tlisting as well as following transplantation.40–42 The
guideline recognizes that frailty is currently not evaluated
routinely and suggests incorporating this formally into the
initial transplant evaluation. Given preliminary evidence
that some patients may have modifiable frailty compo-
nents, the guideline suggests considering strategies to
optimize certain patients with frailty.43
The guideline refrains from making any specific rec-
ommendations regarding frailty assessment or the type of
intervention for frail patients. Understandably, the guide-
line does not provide specific targets or metrics due to
paucity of data. Though not included as a guideline
statement, the KDIGO work group recommends against
using frailty as a contraindication for transplantation, given
that a significant proportion of frail patients also benefit
AJKD Vol 77 | Iss 6 | June 2021
from transplantation and have improvement in frailty
scores posttransplant.
The KDOQI work group agrees with the recommen-
dations and that the focus should be on identifying frailty
with the aim toward risk modification. A formal approach
to quantify frailty may inadvertently reduce access to some
frail patients who may otherwise benefit from trans-
plantation, unless it is shown that interventions can
effectively modify frailty. In addition, frailty should not be
included as a sole factor; instead, it should be evaluated
with other concomitant comorbid conditions such as older
age, diabetes, and obesity.
Clinical Utility
Frailty, which, by definition, is different than comorbidity
burden, is a state of reduced physiological reserve and is
related to age and associated comorbidities. It is particu-
larly common in the context of kidney transplantation,
with a prevalence of 10%-20% among patients on the
waitlist and transplant recipients.40,42 Frailty increases with
age and dialysis vintage; in addition, it is more prevalent in
patients who are female (3.3-fold higher) and kidney
failure patients who are African American.42 Frailty confers
an increased risk of mortality and morbidity among pa-
tients while on the waitlist and after transplantation.40,41
Post–kidney transplant frailty is associated with graft loss
and mortality in addition to higher rates of delayed graft
function and longer hospitalization. Hence, attempts to
quantify frailty may prove useful by identifying patients
who may benefit from strategies aimed at improving frailty
scores.
Implementation and Challenges
There is widespread acceptance among the transplant
community regarding the need to utilize frailty measure-
ment of transplant candidates. In a recent survey conducted
by the American Society of Transplantation (AST), 99% of
transplant professionals (of all organ groups) felt frailty
assessment is useful among transplant candidates.44 Addi-
tionally, 24% reported utilizing frailty measurements
routinely, while another 44% reported measuring frailty at
least on some occasions.44 Despite the widespread support
for implementing routine frailty assessment, difficulties
may arise regarding choice of frailty assessment tool given
the plethora available.45 Nevertheless, many respondents
to the AST survey favored using tools that incorporated
measures of physical strength (gait speed, grip strength,
and sit-to-stand test) along with weight loss, suggesting
that achieving some uniformity and standardization may
not be very difficult.
The KDIGO guideline recommends not using frailty as a
reason to exclude patients from transplantation. However,
close to 70% of the survey respondents in the AST survey
felt that results of the frailty assessment should be used to
839

make transplant candidacy decisions.45 While this is not the
intention of the guideline statement regarding frailty
assessment, it is going to be difficult to separate the 2 as-
pects. In a large prospective study of patients from 3 US
transplant centers, even without physicians knowing the
results of a formal frailty assessment, frail patients were less
likely to be waitlisted and less likely to receive a transplant,
even after adjusting for other factors.42 This suggests that
physicians are able to identify frailty-related phenotypes
even without a formal frailty assessment protocol. Adding a
formal frailty assessment without an interventional strategy
may thus run the risk of reducing access to transplantation
for frail patients. Finally, transplant centers may not have
adequate resources for long-term frailty management, and
this needs further investigation.
Recurrent Glomerular Diseases: Focal Segmental
Glomerulosclerosis
9.2.1: We recommend not excluding candidates with primary
FSGS from kidney transplantation; however, the risk of
recurrence should be considered and discussed with
the candidate (1B).
9.2.1.1: Loss of a prior graft due to recurrent FSGS
indicates a high risk of recurrence upon sub-
sequent transplantation and this factor should
be a major consideration in determining can-
didacy (Not Graded).
9.2.2: We suggest genetic testing (eg, for podocin and
nephrin gene mutations, among others) be performed in
children and young adults with a clinical course
consistent with genetic FSGS to inform the risk of
recurrence (2C).
9.2.3: We suggest avoiding routine use of pre-transplant
plasma exchange or rituximab to reduce the risk of
recurrent FSGS (2D).
Commentary
Primary focal segmental glomerulosclerosis (FSGS) has a
high risk of recurrence in the kidney allograft (30%-50%
with first transplant) and high rates of irreversible graft
loss. The risk of recurrence is higher, up to 80%, in can-
didates who have previously lost a transplant due to FSGS
recurrence.46 Pathogenesis of recurrent FSGS has not been
fully explored, and, hence, risk factors associated with
recurrence remain poorly defined. The identification of a
hypothesized circulating factor has been elusive, and, thus,
there are no confirmatory tests available for the diagnosis
of recurrent FSGS. This also makes it difficult to appro-
priately risk-stratify candidates prior to kidney trans-
plantation. Finally, specific therapeutic interventions
before transplantation have failed to consistently reduce
risk of recurrence.
Clinical Utility
The guideline recommendation about discussing the risk
of recurrence with the patient is reasonable and
840
Puttarajappa et al
clinically useful considering the lack of robust advances
in the understanding of pathogenesis in recurrent FSGS
and the high rate of recurrence among recipients with
prior graft loss from recurrent FSGS.47,48 The recom-
mendation to order genetic tests for the young is also
appropriate given the lower recurrence risk with many
genetic forms of FSGS.47,48 Furthermore, the recom-
mendation to not avoid retransplant in patients with
known prior recurrence despite higher rate of graft loss
is balanced by poor median survival on long-term
dialysis, especially in very young candidates. Currently,
the lack of proven pretransplant interventions to prevent
recurrence in patients with known prior recurrence of
FSGS makes it difficult to advocate for treatment before
transplantation.
Implementation and Challenges
The guideline recommendations are appropriate but
generally limited by the lack of progress identifying the
circulating factor and the continued ill-defined pathogen-
esis of recurrent FSGS in the clinical setting. Advances in
genetic testing and the progressive reduction in the costs
associated with next-generation and whole-exome
sequencing has made genetic testing easier than before.
Thus, implementing genetic testing for transplant candi-
dates with a diagnosis of FSGS, particularly among young
adults and those with a family history of FSGS, is reason-
able and should be considered by transplant programs.
This will allow for better risk stratification for recurrence
after transplantation and could also potentially help in the
evaluation of related kidney donor candidates. Since ge-
netic variants of FSGS recur very rarely after trans-
plantation, a positive genetic test could provide comfort to
patients and confidence to the transplant team about pro-
ceeding with transplantation. Despite availability of better
genetic tests, emphasis remains on pretransplant clinical
risk assessment and counseling potential candidates
regarding recurrence of FSGS.
Although the important negative impact of recurrent
FSGS on graft survival was recognized decades ago, lack of
progress in the elucidation of the pathogenesis in FSGS has
left KDIGO with guideline recommendations that are
similar to current clinical practice and easy to implement.
Transplant programs do not consider FSGS an absolute
contraindication for the first transplant, but many are
rightfully hesitant to retransplant a patient with prior
allograft failure due to FSGS recurrence. Additionally,
given that recurrence occurs in <50% of the patients, is
difficult to predict pretransplant, and efficacy of preemp-
tive therapies remain unproven, it should be easy to
implement KDIGO recommendations to not use preemp-
tive plasmapheresis and rituximab to prevent FSGS recur-
rence. These recommendations may change over time as
further exploration of the pathogenesis of recurrent FSGS
paves the path for future diagnostic and therapeutic
interventions.
AJKD Vol 77 | Iss 6 | June 2021
Puttarajappa et al
Recurrent Glomerular Diseases: Membranous
Nephropathy
9.3.1: We recommend not excluding candidates with MN from
kidney transplantation; however, the risk of recurrence
should be considered and discussed with the candidate
(1B).
9.3.1.1: We suggest not excluding candidates with
prior graft loss due to MN; however, the risk of
recurrence should be considered and dis-
cussed with the candidate (2D).
9.3.2: We suggest that autoantibodies to phospholipase A2
receptor (PLA2R) be measured pre-transplant to inform
the risk of recurrence in patients with MN (2C).
9.3.3: We suggest not routinely using rituximab or alkylating
agents to reduce the risk of recurrent MN (2D).
Commentary
Membranous nephropathy (MN) is the most common
cause of adult nephrotic syndrome. Clinical recurrence of
MN is estimated to be around 10% and can go up to 50%
with longer follow-up. Multiple autoantibodies have been
implicated in the pathogenesis of MN; primary among
these are autoantibodies to the phospholipase A2 receptor
(PLA2R), which are seen in 60%-80% of patients with
primary/idiopathic MN.49 The KDOQI work group agrees
with the KDIGO guideline recommendations for testing
PLA2R before transplantation to assess the risk for recurrent
MN. The higher rate of remission associated with ritux-
imab and cyclosporine in the treatment of primary MN is
well established,50 but similar data do not exist for the
prevention and treatment of recurrent MN for transplant
recipients.
Clinical Utility
Graft loss associated with recurrent MN occurs in ~10%-
13% of cases. Therefore, counseling patients about disease
recurrence is a valuable consideration. Additionally,
measuring disease activity by assaying for anti-PLA2R an-
tibodies before transplantation may help assess risk of
recurrent MN. Other antibodies have been identified
recently among patients with PLA2R-negative MN (eg,
THSD7A [thrombospondin type 1 domain-containing 7A],
NELL-1 [neural epidermal growth factor-like 1], and NEP
[neutral endopeptidase]), and their measurement should
be considered in specific cases. The recommendation
against routine pretransplant treatment with rituximab or
other alkylating agents is appropriate due to lack of suffi-
cient supporting evidence.
Implementation and Challenges
The risk of allograft loss due to recurrent MN is not a
major issue, and, thus, transplantation remains a treatment
option for kidney failure patients with MN. PLA2R testing
before transplantation should be easy to implement since
AJKD Vol 77 | Iss 6 | June 2021
standardized tests are readily available. The guideline does
not specify how and when these antibodies should be
measured pretransplantation. Since studies demonstrate an
association between presence and strength of anti-PLA2R
antibodies and posttransplant recurrence, it may be
reasonable to monitor the levels in those who are positive
for anti-PLA2R antibodies.51 However, it is currently un-
clear if patients with high or increasing levels of anti-
PLA2R antibodies should have a period of waiting time
prior to kidney transplantation. Additional studies are
required to integrate the testing of these antibodies in the
management of recurrent MN. Prospective controlled
studies on recurrent MN with special reference to pre-
vention and treatment are lacking due to inherent diffi-
culties in performing a randomized study for a rare
disease. Thus, treatment for recurrent MN will be based on
extrapolating results from MN in native kidneys. The risk-
benefit balance of rituximab for recurrent MN with
reference to infectious complications in an already
immunocompromised population and its long-term effi-
cacy need further evaluation.
Recurrent Glomerular Diseases:
Membranoproliferative Glomerulonephritis
9.6. Immune complex-mediated membranoproliferative
glomerulonephritis (IC-MPGN) and C3 glomerul-
opathy (C3G)
9.6.1 IC-MPGN
9.6.1.1: We recommend not excluding candidates
with IC-MPGN from kidney trans-
plantation; however, the risk of recurrence
should be considered and discussed with
the candidate (1B).
9.6.1.2: We recommend investigation for an infec-
tive, autoimmune, or paraprotein-mediated
cause of IC-MPGN prior to transplantation
to guide treatment and inform risk of
recurrence (1C).
9.6.1.3: We suggest that, when possible, the cause
of the IC-MPGN be treated prior to
transplantation (2C).
9.6.2 C3G, including dense deposit disease (DDD)
and C3 glomerulonephritis (C3GN)
9.6.2.1: We recommend not excluding candidates
with C3G from kidney transplantation;
however, the risk of recurrence should be
considered and discussed with the candi-
date (1B).
9.6.2.2: We suggest that candidates with C3G be
screened for genetic or acquired causes
for the dysregulation of the complement
alternative pathway to guide treatment and
inform risk of recurrence (2C).
9.6.2.3: Loss of a prior graft due to recurrent C3G
indicates a high risk of recurrence upon
subsequent transplantation and this factor
should be a major consideration in deter-
mining candidacy (Not Graded).
841

Commentary
Kidney transplant remains a viable treatment for patients
with membranoproliferative glomerulonephritis (MPGN).
The classification schema for MPGN has been updated to
reflect pathophysiology and is now classified based on
immunofluorescence findings into immune complex (IC)
MPGN, complement-mediated MPGN (C3 glomerulop-
athy), and MPGN without IC or complement activation. IC-
MPGN can further be classified based on polyclonal antibody
(infections, autoimmune disorders) or monoclonal anti-
body (gammopathies), while C3 glomerulopathy has 2
subtypes based on structural characteristics observed on
electron microscopy (dense deposit disease [DDD] and C3
glomerulonephritis [C3GN]). The pathogenesis of C3 glo-
merulopathies is secondary to unopposed activation of the
alternative complement pathway from either a loss of
function of one of the complement regulatory proteins
(factor H [CFH] or factor I [CFI]) or from gain-of-function
mutations in C3 that lead to resistance to regulation by CFH.
It is critical to classify patients prior to transplantation at risk
for recurrence, as the risk is variable according to the type of
MPGN. Chances of recurrent MPGN are low among patients
with polyclonal IC-MPGN and higher among those with
monoclonal MPGN or C3 glomerulopathies.50 In over 70%
of cases, there can be recurrence of C3GN, which is often
aggressive, seen early after transplantation, and associated
with high rates of graft failure.52
Clinical Utility
It is imperative to subcategorize MPGN according to IC-
MPGN, monoclonal antibody–related MPGN, and C3
glomerulopathy, as the outcome after transplant is
different for each category. Subclassification also helps
clinicians to approach specific treatment strategies to pre-
vent recurrent MPGN. For example, management of sys-
temic infections should remain the focus for the
prevention of IC-MPGN, targeted therapy against mono-
clonal antibodies for monoclonal MPGN, and anti-
complement therapy for C3 glomerulopathies.
Implementation and Challenges
Kidney failure patients with MPGN are not routinely
investigated to characterize the type of MPGN based on its
pathogenesis. It is important for the transplant team to take
the lead in investigating the type of MPGN and convey the
chances of recurrence to patients. While testing for a
monoclonal paraprotein and autoimmune antibodies are
relatively easy to accomplish, testing for complement-
mediated C3 glomerulopathies may be challenging since
such tests are only available at select laboratories. This
arena continues to evolve rapidly, and newer therapies will
change the landscape in the prevention and treatment of
posttransplant MPGN including C3GN.
842
Puttarajappa et al
Recurrent Glomerular Diseases: Hemolytic Uremic
Syndrome
9.11.1: We recommend not excluding candidates with HUS
due to infection with a Shiga-toxin producing organ-
ism, usually E. coli (STEC-HUS), from kidney trans-
plantation (1A).
9.11.2: We recommend assessment of candidates with sus-
pected atypical HUS (aHUS) for a genetic or acquired
defect in complement regulation or other genetic
causes of aHUS to inform risk of recurrence (1B).
9.11.3: We recommend not excluding candidates with aHUS
from kidney transplantation; however, the risk of
recurrence should be considered and discussed with
the candidate (1B).
9.11.3.1: We recommend that if the candidate has an
abnormality in complement regulation
placing them at high risk of recurrence, kid-
ney transplantation should not proceed un-
less a complement inhibitor can be
administered or combined liver-kidney
transplant can be performed (1B).
Commentary
Hemolytic uremic syndrome (HUS) secondary to Shiga
toxin from Escherichia coli is a rare cause of kidney failure
without any recurrence after transplantation. However,
atypical HUS (aHUS), although rare, has high rates of
recurrence after kidney transplantation.53 The pathogenesis
of the disease is a defect in the complement system leading
to unopposed activation of complements, resulting in
microangiopathic lesions in the kidney. The recurrence
rate is as high as 80% among those with a defect in CFH or
CFI and with a prior history of recurrence.53 Candidates
with no identified genetic mutation are presumed to have
an intermediate risk of recurrence. Disease penetrance even
with identification of a pathogenic variant is approximately
50%, suggesting that an environmental factor like infec-
tion, pregnancy, or transplantation may be a necessary
trigger54; however, a precipitating factor in many cases
may not be apparent. Before the approval of monoclonal
anti-C5 complement inhibitor, the recurrence risk was
50%, with graft loss occurring in 80%-90% of the cases.53
Availability of effective complement inhibitors has revo-
lutionized the prognosis in patients with aHUS. Transplant
candidates with mutations affecting proteins primarily
synthesized in the liver (CFH, CFI, C3, and CFB) serve as a
basis for recommendation of a simultaneous liver and
kidney transplantation for a subset of patients with aHUS.
Clinical Utility
Recurrent aHUS can be a catastrophic event with a high
rate of graft loss. The outcome, though, can vary
depending on the complement abnormality. It is critical to
AJKD Vol 77 | Iss 6 | June 2021
Puttarajappa et al
identify the defect in the complement system, and,
accordingly, patients should be counseled for recurrence
and given appropriate therapy to prevent recurrence of
aHUS.
Implementation and Challenges
Lack of easy access to genetic testing, its prohibitive cost,
and variation in test results and interpretation from one
laboratory to another makes it difficult for patients as well
the transplant center to establish a clear protocol for aHUS.
When testing is undertaken, a substantial portion can
reveal no mutations or have a genetic variant of unknown
significance. The heterogeneity in the mechanisms leading
to dysregulation of the alternate pathway and epigenetics
related to disease recurrence make counseling a patient
challenging and underscore a need for further under-
standing of the pathogenesis. Currently available anti-
complement inhibitors, while very effective, are very
expensive, making it prohibitive for many patients and
transplant centers to continue long-term therapy. A trun-
cated form of therapy (continuing for a few years) has
been attempted with varied success. Lowering the infusion
burden with longer-acting anticomplement agents, oral
inhibitors, and more selective alternative complement in-
hibitors that are in the pipeline perhaps may prove to be
more useful options in the future.
Recurrent disease remains an important problem for
young transplant recipients who do not have diabetes or
autosomal dominant polycystic kidney disease. Many pa-
tients present at a late stage of CKD with atrophic kidneys
and thus lack a definite histologic diagnosis for the kidney
disease. Hence, it is prudent that all such kidney transplant
recipients be monitored very closely for proteinuria after
transplantation with prompt renal histological diagnosis to
characterize the type of recurrent disease.
Recipient Vaccination
10.7.1: We recommend that the vaccination series be
commenced using an accelerated schedule, if neces-
sary, prior to kidney transplantation for any inactivated
vaccines (Table 12) (1B).
10.7.1.1: We suggest not excluding candidates who
do not complete an inactivated vaccine se-
ries prior to kidney transplantation (2D).
10.7.2: We recommend that the vaccination series be
completed prior to kidney transplantation for any live
attenuated vaccines (Table 12) (1B).
10.7.2.1: We recommend a 4-week delay in kidney
transplantation if a live vaccine is adminis-
tered (eg, MMR, VZV, shingles, yellow fever,
oral typhoid, oral polio vaccine) (1B).
Commentary
The KDIGO guideline appropriately places emphasis on the
importance of pretransplant vaccination for transplant
AJKD Vol 77 | Iss 6 | June 2021
candidates. The guideline recommends completing all
necessary live attenuated vaccines prior to transplantation
and waiting at least 4 weeks prior to moving forward with
transplantation. While inactivated vaccines can be admin-
istered posttransplant, it is preferable to administer these
pretransplant since responses to immunization are sub-
optimal posttransplant. Hence, the guideline recommends
using an accelerated vaccination schedule for inactivated
vaccines if necessary so that patients receive all necessary
vaccines prior to transplantation. Given the long wait times
to transplantation and high mortality and morbidity
associated with maintenance dialysis, KDIGO recommends
not to delay transplantation just to complete the vaccina-
tion series. However, as the risk for posttransplant varicella
infection—although small—is real, exceptions could be
considered in situations where patients have no demon-
strable antibodies against varicella zoster virus. This aspect
has to be discussed with the patient prior to transplantation
without vaccination.55,56
There are relatively few contraindications to vaccina-
tion. Since infection is the second leading cause of death
among transplant patients, prevention of infectious com-
plications remains prudent, and vaccination is a very
effective way to lower morbidity and mortality from
certain infections.57 Thus, the KDOQI work group agrees
with the listed KDIGO recommendations including waiting
4 weeks after a live-virus vaccination before trans-
plantation. To reduce the risk of infections from encap-
sulated bacteria, asplenic kidney transplant candidates
should receive pretransplant pneumococcal, Haemophilus,
and meningococcal vaccination, and patients who may
need complement inhibitors in the peri- or posttransplant
period should get meningococcal vaccination to reduce the
risk of meningitis. Overall, these recommendations are
similar to the 2019 guidelines from the AST Infectious
Diseases Community of Practice.58
Clinical Utility
Vaccinations for vaccine-preventable illnesses are a highly
cost-effective way to reduce morbidity and mortality.
Kidney transplant recipients are higher risk for serious
illnesses and invasive or disseminated disease, and, hence,
appropriate and timely vaccination is paramount. The
guideline statements thus offer clear and useful recom-
mendations on vaccination schedules that can be followed
both pre- and posttransplant as appropriate.
Implementation and Challenges
Obtaining an accurate vaccination history and checking
serologies (such as for hepatitis viruses and varicella) at
transplant evaluation and coordinating with dialysis cen-
ters and primary care physicians will ensure that vaccina-
tion is completed well in advance of transplantation.
Hepatitis B vaccination is routinely provided to dialysis
patients. For patients not on dialysis and for those with a
need for expanded vaccination due to risk factors such as
splenectomy or need for terminal complement inhibitors,
843

consultation with a transplant infectious disease specialist
will help with timely vaccination. This will allow for better
communication between transplant teams to ensure
transplantation is delayed for at least 4 weeks after
administration of a live attenuated virus vaccine.
The widespread availability of vaccines, statewide im-
munization registries, and transplant infectious
disease–trained physicians at most transplant centers allow
pre- and posttransplant options for getting the indicated
vaccinations. Modern vaccines are well tolerated, and, even
though adverse effects have been reported with all vac-
cines, these events are very rare. Diverse beliefs about
vaccination in the United States, lack of easy access and
universal insurance coverage for vaccination, and less than
optimal attention toward giving pretransplant vaccinations
may contribute to lower pretransplantation vaccination
rates among kidney transplant recipients.59 Thus, patient
education is very important, and transplant centers can
refer patients to the Centers for Disease Control and Pre-
vention website, which maintains a vaccine information
statement for every vaccine and for specific patient
populations.60
Vaccination is a relatively simple way to decrease the
burden of posttransplant infections, but there are chal-
lenges in implementing this for kidney failure patients and
transplant recipients. Since the review of vaccination his-
tory mainly requires a review of records, e-consults or
teleconsults by an infectious disease specialist as soon as
the patient is listed for kidney transplantation may help
improve the rate of vaccinations.61 The Pfizer-BioNTech
and Moderna mRNA vaccines for COVID-19 have recently
been approved for use.62 Although immunosuppressed
patients were excluded from the clinical trials of these
vaccines, it is hoped that they will still offer reasonable
protection against COVID-19 to transplant recipients and
for those with advanced kidney disease.
Cancer Screening
11.1.1: We recommend candidates undergo routine cancer
screening, as per local guidelines for the general
population (Table 13) (1D).
11.1.1.2: We suggest chest CT for current or former
heavy tobacco users (≥ 30 pack-years) as
per local guidelines, and chest radiograph
for other candidates (2C). (Same as rec.
12.2.1)
Commentary
The KDIGO guideline allows for variations in cancer
screening practices “per local guidelines.” With advances
in screening technology and shifting emphasis toward
cost-effectiveness and increasing recognition of the unin-
tended consequences of false-positive testing and over-
diagnosis, screening guidelines for cancers continue to
evolve. In addition, guidelines in the United States are is-
sued by many agencies, and, while generally consistent,
844
Puttarajappa et al
often include subtle variations.63,64 Thus, the optimal
screening strategy for some cancers remains incompletely
defined. The guideline also specifically acknowledges the
growing evidence regarding the benefit of lung cancer
screening in high-risk populations and recommends
screening chest computed tomography for patients with
heavy smoking history.65
Clinical Utility
Given the increased risk of cancer in the setting of
immunosuppression and the need to avoid transplanting
patients with underlying cancers where early detection is
possible, appropriate cancer screening is vital in the
assessment of potential kidney transplant recipients. There
are notable differences in US recommendations when
compared to European best practices.66 For example, ta-
ble 13 of the KDIGO guideline lists biennial fecal immu-
nochemical testing as the method of choice for colon
cancer screening, followed by flexible sigmoidoscopy.1
However, colonoscopy remains the first-line screening
tool recommended by many groups in the United States
due to its high sensitivity and the ability to remove pre-
cancerous lesions at the time of screening.67,68 Flexible
sigmoidoscopy is used less often in the United States and
will miss right-sided colonic cancers. Similarly, while the
KDIGO guideline suggests that women aged 40-49 years
should have the choice to start annual breast cancer
screening, the US Preventive Services Task Force guidelines
give a lower level of recommendation for this age group.69
Implementation and Challenges
Communication and agreement between the transplant
center, insurance companies, and referring physician is
necessary but remains a challenge given evolving guide-
lines, increasing wait times, and the inevitable lag period
in updating transplant center as well as insurance autho-
rization screening protocols. There is a need to simplify
and individualize screening without increasing risk for
patients.
Potential Kidney Transplant Candidates With Prior
Cancer
11.2.1: We recommend that candidates with active malig-
nancy be excluded from kidney transplantation except
for those with indolent and low-grade cancers such as
prostate cancer (Gleason score ≤ 6), superficial non-
melanoma skin cancer, and incidentally detected renal
tumors (≤ 1 cm in maximum diameter) (1B).
11.2.2: Timing of kidney transplantation after potentially cura-
tive treatment for cancer is dependent on the cancer
type and stage at initial diagnosis (Not Graded).
Commentary
Oncology has witnessed significant improvements in the
management of various cancers, a better understanding of
AJKD Vol 77 | Iss 6 | June 2021
Puttarajappa et al
the natural history of certain malignancies, and marked
improvements in survival of patients with certain cancers.
This improvement is not reflected in the KDIGO recom-
mended waiting times between cancer remission and
kidney transplantation. This is likely due to the fact that
much data on cancer outcomes are derived from the
general population and not transplant patients receiving
immunosuppression. For example, the KDIGO guideline
recommends a waiting time of at least 2 years for in-
dividuals with Gleason score 7 prostate cancer and at least
2 years for stage 2 thyroid disease. Analyses of large
databases in the general population have demonstrated
nearly 100% 5-year survival for all but distant metastatic
prostate and thyroid malignancies.70,71 Similarly, 5-year
survival rates for patients with localized renal cell cancer
(stage IB or <7 cm) are now >90%.72 Patients with these
tumors, particularly if low-grade, could potentially be
considered for transplant without additional waiting time,
during which they would accumulate dialysis-associated
morbidity and mortality risk. In addition, while some
cancers may not be fully cured, new maintenance thera-
pies (for example, for multiple myeloma) are changing
previously deadly illness into chronic disease, not unlike
the evolution that has taken place in HIV management.
The prediction of cancer survival has begun to move away
from pathologic and radiologic diagnosis to novel blood
and tissue tumor markers as well as genetic analysis.
While improvements in the management of kidney failure
have remained incremental, understanding in cancer sci-
ence has surged forward, and the risk of death associated
with maintenance dialysis now exceeds that of many
cancers.73 If evidence accumulates that advances in cancer
treatment options leading to better survival is true for the
transplant population as well, it should in the future
translate to shorter cancer-free intervals prior to
transplant.
Clinical Utility
Transplant candidates with a history of prior malignancy is a
common clinical scenario. Since cancer management con-
tinues to improve and dialysis-attributable mortality re-
mains high, sometimes surpassing that of certain cancers,
the recommended guidelines could be used as a practical
starting point, and assessment of mortality from cancer
versus remaining on maintenance dialysis can be individu-
alized when considering a patient for transplantation.
Implementation and Challenges
The time course and influence of immunosuppression on
cancer disease recurrence need to be balanced against the
risk of poor outcomes on dialysis. One large challenge is
that cancer survival data are largely gathered from the
general, immunocompetent population; risks may be
higher in the setting of immunosuppression and kidney
failure. Oncologists, nephrologists, and transplant
AJKD Vol 77 | Iss 6 | June 2021
physicians offer different insight in the management of
kidney failure and transplant patients with cancer. Dia-
logue between transplant centers and insurance providers
must also take place to override the dogma of previously
published cancer-free wait times. Prolonged wait time for
transplant and improving outcomes with certain cancers
has provoked many transplant physicians to accept patients
with cancer for kidney transplantation.
The balance of risks and benefits of transplantation,
dialysis, and cancer will continue to challenge transplant
physicians. Advances in cancer diagnosis and treatment, as
well as the outcome, will require transplant providers to
continue to review their practices to optimize the care of
patients with pre- and posttransplant cancer alike.
Coronary Artery Disease
13.3: We suggest that asymptomatic candidates at high risk
for coronary artery disease (CAD) (eg, diabetes, previ-
ous CAD) or with poor functional capacity undergo non-
invasive CAD screening (2C).
13.3.1: We recommend that asymptomatic candidates
with known CAD not be revascularized exclu-
sively to reduce perioperative cardiac events
(1B).
13.3.2: We suggest that patients with asymptomatic,
advanced triple vessel coronary disease be
excluded from kidney transplantation unless
they have an estimated survival which is
acceptable according to national standards
(2D).
13.7: We suggest that patients with uncorrectable, symp-
tomatic New York Heart Association (NYHA) Functional
Class III/IV heart disease [severe CAD; left ventricular
dysfunction (ejection fraction <30%); severe valvular
disease] be excluded from kidney transplantation unless
there are mitigating factors that give the patient an
estimated survival which is acceptable according to
national standards (2D).
13.7.1: Patients with severe heart failure (NYHA III/IV)
who are otherwise suitable for kidney trans-
plantation should be assessed by a cardiologist
and considered for combined/simultaneous
heart and kidney transplantation (Not Graded).
Commentary
CAD is highly prevalent among kidney failure patients
referred for kidney transplantation. Nearly all patients
are routinely screened for CAD prior to listing and while
on the transplant waiting list. Conclusive evidence that
such a strategy improves patient outcomes, including
postoperative cardiac events, is lacking. Nevertheless,
several major society guidelines have recommended
screening for CAD in at-risk patients undergoing elective
surgery.21,22,74–77 While cardiac and anesthesia guide-
lines recommend CAD screening only in those with
845

impaired functional status (defined as functional capacity
of <4 metabolic equivalents), transplant-specific guide-
lines have recommended screening for CAD in at-risk
patients irrespective of functional status. Similarly,
KDIGO guideline statement 13.3 recommends screening
for CAD in asymptomatic patients with poor functional
status and for asymptomatic patients who have high risk
of CAD (prior CAD, diabetes mellitus) irrespective of
their functional status. When interpreting results and
developing a treatment strategy for coronary disease,
clinicians should, however, be aware that lack of coro-
nary symptoms in patients with kidney failure is com-
mon and is possibly related to limited mobility, poor
functional capacity, and concurrent autonomic
neuropathy.
The KDIGO guideline does not recommend coronary
revascularization for the sole purpose of reducing peri-
operative cardiac events (recommendation 13.3.1). The
KDOQI work group agrees with this recommendation,
given lack of data for reduction in perioperative cardiac
events with preoperative coronary revascularization.78
This recommendation, however, can be challenging to
interpret, given that CAD screening will uncover
ischemia and unknown coronary lesions. Also, though
elective revascularization does not reduce perioperative
cardiac events, patients with risk factors (ischemia, dia-
betes, decreased kidney function) are at higher risk for
these events.79 The KDIGO guideline should have
included recommendations about when revascularization
may be reasonably appropriate in transplant candidates
found to have CAD. These can be found in the 2014
ACC/AHA (American College of Cardiology/American
Heart Association) report and the criteria for appropriate
use of coronary revascularization in patients with stable
ischemic heart disease, released jointly in 2017 by
several organizations.80 KDIGO guideline statement
13.3.2 suggests that patients with advanced 3-vessel
disease not be considered for transplantation except
when their survival is considered acceptable. The KDOQI
work group felt that this recommendation is somewhat
vague and leaves individual transplant programs to
decide on the candidacy of such patients based on their
overall estimated risk of major adverse cardiovascular
events over time and overall survival. There is a risk that
this will lead to significant variation in how centers
approach patients with 3-vessel disease.
KDIGO guideline statement 13.7 recommends not
transplanting patients with moderate to severe symptoms
(defined as NYHA class III or IV) and coexisting uncor-
rected valvular disease, CAD, or ejection fraction <30%.
The guideline recommends that such patients should be
evaluated for the possibility of a combined heart/kidney
transplantation. The KDOQI work group agrees with these
recommendations. Among patients with ejection frac-
tion <30%, it may be reasonable to consider trans-
plantation if patients have well compensated heart failure,
846
Puttarajappa et al
have good functional status, and do not have significant
and overt CAD or valvular lesions.
Clinical Utility
CAD remains the primary cause for early posttransplant
mortality and for death with a functioning graft following
transplantation.81 Risk factors for CAD are highly preva-
lent among kidney transplant candidates, particularly
among patients with hypertension, diabetes mellitus, and
obesity. As a result, screening for CAD, even if patients are
asymptomatic, is one of the most common tests per-
formed in transplant candidates prior to and while on the
waitlist. This will, on some occasions, uncover ischemia
in asymptomatic patients that will necessitate further
testing and invasive coronary evaluations.82–84 However,
evidence that such a strategy improves outcomes peri-
operatively and in the long term is lacking. While the
primary reason for screening asymptomatic patients is to
reduce perioperative coronary events, a second reason is
to identify patients who may have severe underlying
uncorrectable 3-vessel disease in whom transplantation is
not advisable; this latter scenario, however, is uncommon,
with a high number needed to screen.
Implementation and Challenges
The KDIGO guideline suggests that it may be reasonable
to screen for CAD among asymptomatic patients with
risk factors and those with poor functional capacity.
However, the guideline strongly recommends against
revascularization solely for the purposes of reducing
perioperative cardiac events. This implies that revascu-
larization should take into consideration other features
of CAD that would make a decision to revascularize
meet the AHA/ACC appropriate use criteria.80 To add to
the data showing no benefit with elective coronary
revascularization in the general population, a recent
large randomized study among patients with CKD
showed no benefit with coronary intervention for
moderate to severe ischemia when compared to optimal
medical management.85 However, once ischemia is
noted on CAD screening and/or a significant coronary
lesion is identified on angiography, transplant physi-
cians, surgeons, and anesthesiologists are reluctant to
proceed with transplant surgery without revasculariza-
tion. This is particularly challenging if the lesion does
not meet criteria for revascularization and/or if revas-
cularization is not possible. Since exclusion of such pa-
tients may be inappropriate, consideration should be
given to the extent of ischemia, location and degree of
CAD lesions, ejection fraction, and patient characteristics
(such as functional status, age, and comorbidities) to
evaluate if transplantation can be pursued with optimal
medical management. Finally, if ischemia is identified
that requires a coronary angiogram but the patient is not
on dialysis, then decisions will need to take into account
AJKD Vol 77 | Iss 6 | June 2021
Puttarajappa et al
the availability of living donors, extent of ischemia, time
to deceased-donor transplantation, and symptoms to
decide on the timing of the angiogram. Consultation
with cardiology teams will be highly valuable in these
situations.
An additional downside of aggressive CAD screening is
the need for invasive procedures and the potential for
some patients to be deemed ineligible for transplantation
despite being asymptomatic and in good functional status.
This restricts the opportunity for successful transplantation
for such recipients. A recent cost-benefit analysis suggested
that no screening for CAD on the waiting list may be more
cost-effective than routine screening of asymptomatic
waitlisted kidney transplant candidates.86 So, how should
programs achieve a balance between avoiding unnecessary
testing and also reducing risk of transplantation in the
presence of severe untreated CAD? A reasonable starting
point may be to consider the following goals while pur-
suing CAD testing: 1) to identify severe 3-vessel or left-
main CAD that may benefit from revascularization ac-
cording to the appropriate use criteria, 2) to identify severe
advanced CAD and thus avoid transplantation in such pa-
tients, and 3) to not pursue coronary revascularization for
minor levels of ischemia and nonsignificant coronary le-
sions and also not exclude such patients from
transplantation.
Thus, the KDIGO guideline tackles only severe forms of
disease such as 3-vessel disease and those that restrict life
expectancy. The KDIGO work group was unable to make a
strong recommendation in favor of or against testing
asymptomatic individuals. This leaves testing, interpreta-
tion, intervention, and management for CAD to transplant
centers. The issue surrounding CAD for kidney failure
patients prior to transplant and while on the waitlist re-
mains a complex issue and requires careful risk-benefit
assessment of transplantation.
Pulmonary Hypertension
13.4: We suggest that asymptomatic candidates who have
been on dialysis for at least two years or have risk factors
for pulmonary hypertension (eg, portal hypertension,
connective tissue disease, congenital heart disease,
chronic obstructive pulmonary disease) undergo echo-
cardiography (2D).
13.5: Patients with an estimated pulmonary systolic pressure
greater than 45 mm Hg by echocardiographic criteria
should be assessed by a cardiologist (Not Graded).
13.5.1: We recommend not excluding candidates with
uncorrectable pulmonary artery systolic pres-
sure greater than 60 mm Hg (obtained from
right heart catherization) from kidney trans-
plantation; however, the risks of sudden deteri-
oration or progression after transplantation
should be a key consideration and the patient
should have an estimated survival which is
acceptable according to national standards
(1C).
AJKD Vol 77 | Iss 6 | June 2021
Commentary
Pulmonary hypertension (PH) is highly prevalent among
CKD patients, including those on dialysis, and is associated
with worse patient and transplant outcomes after kidney
transplantation.87–90 Hence, the KDOQI work group
agrees with the recommendation to screen patients who
have been on dialysis for >2 years for the presence of PH.
The guideline also suggests obtaining cardiology evalua-
tion for echocardiographic evidence of moderate PH
(pulmonary arterial systolic pressure [PASP] >45 mm Hg).
These guidelines are in line with the 2012 AHA/ACC
guidelines on cardiac evaluation for kidney transplant
candidates.77 However, the KDIGO guideline leaves out
specifics on right heart catheterization for confirmation
and subsequent management of PH. The KDIGO guideline
does make a strong statement (level 1 recommendation)
about not excluding from transplantation patients with
severe PH (ie, PASP >60 mm Hg). The KDOQI work
group, however, feels that the strong evidence for good
outcomes among patients with severe PH is lacking, and a
more guarded approach is prudent for patients with un-
correctable severe PH. Otherwise, the KDOQI work group
broadly agrees with the listed recommendations but notes
the paucity of data clearly characterizing posttransplant
outcomes for different subtypes of PH and for patients
with severe PH.
Clinical Utility
PH prevalence increases with CKD stage and is highest
among hemodialysis patients, with some studies reporting a
prevalence of up to 70%.87 Several factors contribute to this
increased prevalence, but, most commonly, it is related to
increased intravascular volume and elevated left sided heart
pressures. Additionally, patients on dialysis have additional
coexisting risk factors such as obesity, sleep apnea, con-
nective tissue disease, shunting of blood through dialysis
access, cardiac disease, and fluctuating volume status.
Diagnostic and management strategies are challenging given
that the etiology of PH among kidney transplant candidates
varies, with patients falling in several of the World Health
Organization PH types.91 Approved pharmacological treat-
ments have mostly been directed at group 1 PH characterized
by elevated pulmonary vascular resistance (>3 Woods units)
and pulmonary capillary wedge pressure <15 mm Hg.
Though PH is common, the prevalence of severe PH
(defined as PASP >60 mm Hg) is uncommon but often in-
dicates presence of causes other than just volume overload
and left heart disease. Hence, the KDOQI work group agrees
with the KDIGO guideline recommendation to obtain
evaluation by cardiology or PH specialists.
Patients with PH have increased risk of delayed graft
function, inferior graft function, and higher mortality.88,89
However, studies have also shown PH improves after
successful kidney transplantation, along with improvement
in patient-reported symptoms.92,93 This, combined with
the poor outcomes for patients with PH who are on
847

dialysis, suggest that transplant should still be the favored
strategy for PH patients with kidney failure. Thus, the
KDOQI work group in general agrees with the KDIGO
recommendation to not exclude patients from trans-
plantation purely based on PASP criteria. However, a
guarded approach is prudent, as these patients are at risk
for poor outcome.
Implementation and Challenges
Obtaining a baseline echocardiography for PH screening
in at-risk patients does not require additional resources,
since the majority of transplant candidates undergo
echocardiography at evaluation as part of assessment of
left ventricular (LV) function and valvular issues. How-
ever, challenges may arise in ensuring adequate “dry
weight” status at the time of echocardiography, partic-
ularly when volume issues are suspected to be contrib-
uting to PH. Performing echocardiography immediately
postdialysis may help reduce need for right heart cath-
eterization evaluation, but this is often challenging due
to logistical barriers arising from dialysis schedules,
caregiver support, and transportation issues. Addition-
ally, the criteria for obtaining additional investigations
and cardiology input are based on PASP criteria, which
may not always correlate with right heart catheterization
measurements.94 However, use of PASP along with
tricuspid jet peak velocity and other echocardiography
features of PH such as right ventricular size, function,
and thickness increases accuracy of PH recognition by
echocardiography.95
Implementing the recommendation to not exclude pa-
tients with PASP >60 mm Hg may be more challenging,
since the guideline does not provide specifics regarding
evaluating such patients. How should programs consider
patients with persistent elevation of PASP to >60 mm Hg?
Patients with severe PH who have a precapillary compo-
nent to PH should be trialed with pharmacological therapy
under a PH expert to evaluate if pressures can be improved
to <50 mm Hg. Additionally, the candidacy for kidney
transplantation will have to consider the patient’s overall
functional status, right and left ventricular function,
presence of concomitant cardiac conditions (such as CAD,
valvular disease), systemic hypotension, and other
comorbidities. Such patients with persistent severe PH
despite volume management and pharmacological therapy
are likely not transplant candidates at the majority of US
centers.
Thus, prolonged waiting times and a complex interplay
of risk factors determine the severity and persistence of PH,
which poses problems for potential transplant recipients.
Future research should focus on patients with persistent
moderate PH to identify clinical characteristics that will
help identify characteristics favorable to successful trans-
plant outcomes.
848
Puttarajappa et al
Immunological Assessment
19.1: Communicate all sensitizing events (eg, blood product
transfusion, including platelets, pregnancy or miscar-
riage) or clinical events that can impact panel reactive
antibody (PRA) (eg, vaccination, withdrawal of immuno-
suppression, transplant nephrectomy, significant infec-
tion) to the human leukocyte antigen (HLA) laboratory
(Not Graded).
19.2: Perform HLA antibody testing at transplant evaluation, at
regular intervals prior to transplantation and after a
sensitizing event or a clinical event that can impact PRA
(Not Graded).
19.3: We recommend that HLA antibody testing be performed
using solid phase assays (1B).
19.4: We recommend HLA typing of candidates at evaluation
using molecular methods, optimally at all loci (1D).
19.5: We suggest not routinely testing candidates for non-
HLA antibodies (2C).
19.6: We suggest not routinely testing candidates for
complement-binding HLA antibodies (2C).
19.7: We suggest informing candidates about their access to
transplantation based on blood type and histocompati-
bility testing results (2C).
19.7.1: We recommend offering candidates with
immunologically-reduced access to transplant
access to a larger deceased donor pool, kidney
exchange programs, and/or desensitization (1C).
19.7.2: We suggest that antibody avoidance (eg, kidney
exchange programs or deceased donor
acceptable mismatch allocation) be considered
before desensitization (2C).
Commentary
The first recommendation, to communicate all sensitiza-
tion events to HLA laboratory directors, is important
because of the limitations of HLA antibody testing. Solid-
phase HLA antibody tests have inherent assay limitations
and do not provide any measure of immunologic mem-
ory.96 Currently, there are no readily available tests of
immunologic memory outside of the research setting, and
obtaining important sensitization history from the candi-
date provides some meaningful assessment of the potential
for memory response. This key point was highlighted by a
2017 meeting report from the AST’s STAR (Sensitization in
Transplantation: Assessment of Risk) work group.96
The KDIGO guideline also contains recommendations
about obtaining candidate serum at regular intervals to test
new HLA antibodies. Lack of consensus regarding the of
frequency of pretransplant testing remains an issue. The
intent is to keep information on unacceptable antigens
current to avoid unexpected positive crossmatches. This
practice also ensures that HLA laboratories have fresh
recipient serum available for a timely physical crossmatch
when a deceased-donor organ is available for
transplantation.
AJKD Vol 77 | Iss 6 | June 2021
Puttarajappa et al
The KDOQI work group agrees with the recommen-
dation that HLA antibody testing be done with solid-phase
assays and that routine non-HLA antibody and comple-
ment binding antibodies be avoided, but the KDIGO
guideline lacks important information about the role of
adjunctive testing including flow cytometric and/or
complement-dependent cytotoxicity crossmatches.97,98
There is also no mention of the increasingly important
role of the virtual crossmatch in an era of widespread HLA
typing with molecular methods and sensitive single anti-
gen bead HLA antibody testing.99 The virtual crossmatch is
very reliable among kidney transplant candidates who have
no sensitization history and negative single antigen bead
HLA antibody tests, but may be less reliable among
sensitized candidates with a high calculated PRA (cPRA).
These highly sensitized candidates may benefit from
physical crossmatch testing and solid-phase assays with
diluted serum to assess for factors such as prozone and
bead saturation.100,101
The KDOQI work group also agrees that sensitized
candidates should be offered access to larger deceased
donor pools and kidney paired donation (KPD) programs
if they have a potential living donor, but it must be
acknowledged that patients with a very high cPRA
(>99.9%) in the United States have disproportionately
long wait times,102,103 and transplantation in the context
of HLA donor-specific antibody (DSA) and or positive
crossmatch may provide patient survival benefit.104
Lastly, the term “desensitization” needs to be used
carefully. Not all patients who undergo a transplant in the
context of preformed DSA and/or positive crossmatch
undergo therapy to remove antibody (desensitization);
thus, HLA-incompatible transplant is not synonymous
with desensitization.98 There are also inadequate data to
inform desensitization practices. Randomized controlled
trials have not shown that desensitization is associated
with improved long-term allograft survival. However,
several retrospective studies have shown that kidney
transplantation can be safely performed with a low degree
of DSA and/or positive flow cytometric crossmatch
without specific desensitization therapy and lead to
acceptable short-term results among patients with limited
options.98 Moreover, it is important to acknowledge that,
before sensitive single antigen bead and flow cytometric
methods, kidney transplants were routinely (if inadver-
tently) performed in the context of low levels of
alloantibody.
The KDIGO guideline does not contain recommenda-
tions regarding emphasizing patient education concerning
sensitization events and the importance of immunosup-
pression adherence in avoiding sensitization. Recommen-
dations regarding immunosuppression withdrawal among
patients with a failing allograft are also missing; these
patients are at high risk of developing HLA antibodies after
immunosuppression withdrawal. An approach that takes
into consideration residual kidney function, patient age,
AJKD Vol 77 | Iss 6 | June 2021
risk of side effects, and chances of a repeat transplant
within a reasonable time period is necessary to tailor
immunosuppression withdrawal. The KDOQI work group
strongly believes that all transplant centers should focus
heavily on educating transplant recipients about sensitizing
events and the role of nonadherence in HLA sensitization.
This will improve chances of a repeat transplant should the
need arise.
Clinical Utility
The recommendations regarding obtaining an immuno-
logical assessment at the time of transplant evaluation and
at regular intervals thereafter are prudent, but there are
challenges to implementation. Routine anti-HLA testing is
costly, and it can be difficult to contact patients to ensure
that testing is performed regularly.
Implementation and Challenges
In the United States, there are disparities in access to KPD
programs and transplant center expertise with HLA-
incompatible transplantation (transplant in the context of
DSA and/or positive crossmatch). Inadequate information
exists to inform providers about the appropriate time to
move forward with an HLA-incompatible transplant versus
seeking KPD or remaining on dialysis. Solely focusing on
the avoidance of DSA is overly simplistic. The compati-
bility between a donor and recipient should also consider
characteristics including cytomegalovirus and Epstein–Barr
virus serostatus, age and/or size mismatch, and the ex-
pected survival of a particular living-donor transplant.
Factors such as exhausted dialysis access or the ability to
get a preemptive transplant play into decisions to move
forward with an HLA-incompatible transplant.
KPD programs in the United States also remain frag-
mented. Transplant programs with low volumes are likely
to be disadvantaged. Small programs with internal KPD
programs often have a small pool of candidates, and large
multicenter programs such as the National Kidney Registry
are based on donor-allocation schemes that tend to favor
large programs with access to nondirected donors. The
intent of these donor allocation schemes is to keep the
donor pool large, but it has a potential to disadvantage
recipients from certain centers.
A paucity of data exists to inform how to form chains to
increase the longevity of allografts in general. The National
Kidney Registry primarily focuses on avoiding HLA
mismatch, which is certainly important, but it must be
acknowledged that factors beyond HLA mismatching in-
fluence long-term allograft survival. Great benefit would
come from a high-functioning national KPD program with
an independent governing body focused on the ethical
tenets of utility, justice, and respect for persons while
maximizing the longevity of kidney transplants and
avoiding disparities.105 Lastly, certain centers have not
uniformly adopted the virtual crossmatch, as it might
decrease the financial reimbursement to HLA laboratories.
849

Development of HLA antibody before and after trans-
plantation will continue to hinder transplant outcome. This
barrier must be broken through immunological interven-
tion, both by prevention and treatment of antibody-
mediated rejection. Further scientific advances centered
around HLA antibodies are needed to inform guidelines in
this area.
Plasma Cell Dyscrasias, Monoclonal Gammopathy,
and Multiple Myeloma
9.13.1 Multiple myeloma
9.13.1.1: We suggest that candidates with multiple
myeloma be excluded from kidney trans-
plantation unless they have received a
potentially curative treatment regimen and
are in stable remission (2D).
9.13.2 Monoclonal immunoglobulin deposition disease
(MIDD)
9.13.2.1: We suggest that candidates with light chain
deposition disease (LCDD) be excluded from
kidney transplantation unless they have
received a potentially curative treatment
regimen and are in stable remission (2D).
9.13.2.2: We suggest that candidates with heavy chain
deposition disease (HCDD) be excluded
from kidney transplantation unless they have
received a potentially curative treatment
regimen and are in stable remission (2D).
9.13.2.3: We suggest that candidates with light and
heavy chain deposition disease (LHCDD) be
excluded from kidney transplantation unless
they have received a potentially curative
treatment regimen and are in stable remis-
sion (2D).
9.13.3 AL amyloidosis
9.13.3.1: We suggest that candidates with AL
amyloidosis be excluded from kidney trans-
plantation unless they have minimal extra-
renal disease (eg, cardiac amyloid), have
received a potentially curative treatment
regimen and are in stable remission (2D).
17.6 Monoclonal gammopathy of undetermined sig-
nificance (MGUS)
17.6.1: We suggest not excluding candidates with
MGUS from kidney transplantation; however, a
higher risk of post-transplant lymphoprolifer-
ative disease and other hematological malig-
nancies should be considered and discussed
with candidates (2D).
17.6.2: We suggest not excluding candidates with
smoldering multiple myeloma from kidney
transplantation; however, a significant risk of
transformation into multiple myeloma should be
considered and discussed with candidates
(2D).
17.6.3: We recommend careful evaluation of candi-
dates with MGUS for other types of plasma cell
disorders prior to kidney transplantation (1D).
850
Puttarajappa et al
Commentary
Plasma cell dyscrasias (PCDs) have a substantial risk of
recurrence and poor outcomes after kidney trans-
plantation. Hence, the KDIGO work group recommends
excluding patients with PCD from consideration for
transplantation unless there has been a potentially curative
treatment with achievement of stable remission. It is
important to note that the quality of data and the strength
of this recommendation are not strong (2D). Furthermore,
no subclassification is considered to stratify the risk for
recurrence. The KDIGO work group emphasizes the need
for a multidisciplinary treatment approach involving he-
matologists and nephrologists. Good outcomes have been
reported for patients with multiple myeloma and kidney
failure. However, no clear recommendations are made
regarding wait time between induction of remission for
multiple myeloma and transplantation. The combination
of living-donor kidney transplantation followed by autol-
ogous stem cell transplantation has been reported, but is
sparingly used.106
LCDD, LHCDD, and HCDD can manifest with kidney
disease and are frequently associated with MGUS (20%)
and multiple myeloma (60%). Recent data suggest that
reasonable outcomes can be obtained with kidney trans-
plantation among patients with monoclonal immuno-
globulin deposition disease using chemotherapy and
autologous stem cell transplantation.107 Similarly, good
kidney transplant outcomes have been demonstrated for
patients with AL amyloidosis.108,109 However, there re-
mains a risk of recurrence, which is particularly high if
patients have not achieved a complete response to AL
amyloidosis therapy; in a single-center study, the
recurrence-free survival at 5 years was 0% for patients with
partial response to AL amyloidosis therapy and 53.3% for
those with very good partial response.109 The risks of
MGUS and smoldering myeloma progressing to overt
multiple myeloma are low and are estimated to be 1% and
3% per year, respectively.110 In general, no hematological
treatment is recommended prior to progression to multi-
ple myeloma.111 The impact of transplantation despite low
risk for of myeloma transformation remains controversial.
Increased risk of posttransplant lymphoproliferative disease
has been noted and should be kept in mind.112
Clinical Utility
The clinical utility of these KDIGO guideline statements is
limited by lack of good data and heterogeneity of disease
subtypes and spectrum of pathological renal manifesta-
tions. As highlighted by the KDIGO work group, it is
imperative to consider a multidisciplinary approach that
includes a collaboration between hematology and
nephrology in these cases. The curative approach implied
in the KDIGO guideline should not be taken literally, as
very few true curative approaches exist for PCD.
AJKD Vol 77 | Iss 6 | June 2021
Puttarajappa et al
Nevertheless, very notable therapeutic advances have been
made that can induce clinical remission for a prolonged
period of time and allow a successful kidney trans-
plantation with better longevity and good quality of life.
Although the combination of a kidney transplantation and
stem cell transplantation is attractive given the potential for
minimizing immunosuppression and achieving remission
of PCD, it is rarely achievable on clinical grounds due to
limited availability of donors and the differences in criteria
for stem cell transplantation compared to those for kidney
transplantation.
Substantial improvement in survival has also been noted in
patients with AL amyloidosis in whom hematological
remission was obtained prior to transplantation. The dis-
cussion on kidney transplantation for patients with MGUS
and smoldering myeloma is based on concern about the risk
of overt myeloma transformation, which is not more than
3% per year. By contrast, the annual mortality rate for all
patients on the kidney transplant waitlist is around 5%-6%.
Implementation and Challenges
The KDIGO guideline is restrictive with regard to trans-
plantation for the listed hematological conditions. Imple-
mentation considerations will become relevant only if
transplantation for these clinical entities is permitted based
on their pathologic findings, cause of renal manifestation,
response to therapy, and risk of recurrence posttransplant.
In addition, as newer therapies become available, their
ability to suppress the hematological disease will also be
relevant. Continuation of therapy for hematological dis-
eases after transplantation may not always be curative but
could potentially allow long-term suppression of clinical
disease sufficient to allow good outcomes. Due to this
complexity and lack of data to guide best practice, it is
imperative that a multidisciplinary approach is taken in
these cases. Finally, it is important to recognize that pa-
tients presenting for transplant evaluation with an un-
known cause of kidney failure may have an unrecognized
monoclonal gammopathy of renal significance (MGRS).
Evaluating for a monoclonal protein among such patients,
particularly when older than 40 years of age, may be
reasonable. However, since MGRS is diagnosed by kidney
biopsy, further evaluation to diagnose MGRS will be
limited in patients with kidney failure. Such patients
should be carefully monitored for signs of MGRS and
consider kidney biopsy posttransplant.
It should be acknowledged that kidney transplant out-
comes remain inferior for this subset of recipients. How-
ever, since kidney transplantation represents a better
treatment option than long-term dialysis, this should
remain an option, at least for select patients with favorable
characteristics and a reasonable life expectancy. Recog-
nizing that more data are needed, a more permissive
approach regarding access to transplantation may be war-
ranted for the subset of patients with stable PCDs, which
may enhance overall patient survival as compared to long-
term dialysis.
AJKD Vol 77 | Iss 6 | June 2021
Combined Liver and Kidney Transplant
16.7.3: We recommend that candidates with cirrhosis or sus-
pected cirrhosis be referred to a specialist with
expertise in combined liver-kidney transplantation for
evaluation (1B).
16.7.3.1: We recommend that patients undergo iso-
lated kidney transplantation if deemed to
have compensated cirrhosis after specialist
evaluation (1B).
Commentary
The KDIGO guideline recommendation that patients can
undergo isolated kidney transplant if deemed to have
compensated cirrhosis is perhaps overly simplified. This
guideline does not consider ongoing inciting factors such
as untreated hepatitis C or alcoholism that must be
addressed to avoid progression of liver disease. The
importance of recognizing portal hypertension with high
hepatic venous pressure gradients was also not consid-
ered. It is well known those with high hepatic venous
pressure gradient (>10 mm Hg) can progress to
decompensated cirrhosis with poor outcomes following
kidney transplantation.113 The recommendation about
specialty consultation is important, but it is critical to
acknowledge the differences in the nephrologist versus
hepatologist perspective. The guideline appears to be
written with a patient with known kidney disease and
concomitant liver failure in mind. The KDIGO work
group offers minimal guidance about when it is appro-
priate to consider a simultaneous kidney transplant
among patients with end-stage liver failure. Decom-
pensated liver failure with hepatorenal syndrome often
presents as moderate to severe kidney dysfunction and
many require dialysis. With timely liver transplantation,
hepatorenal syndrome is often reversible. The prolonged
acute kidney injury from hepatorenal syndrome, espe-
cially in patients who are older or who have diabetes or
nonalcoholic steatohepatitis, can lead to permanent kid-
ney damage. Moreover, other factors, including under-
lying CKD, infections with hemodynamic instability, and
medications play a role in the development of kidney
failure requiring kidney replacement therapy among pa-
tients with concomitant liver disease.
Clinical Utility
Optimizing organ utility is a must. The facts that liver
transplant alone in the context of kidney failure is associ-
ated with inferior outcomes114–116 and that unnecessary
combined liver and kidney transplants contribute to
improper kidney organ utilization must be balanced.117
Currently, per the Organ Procurement and Transplant
Network/United Network of Organ Sharing (OPTN/
UNOS) allocation policy, patients can be eligible for
combined liver and kidney transplant if they have known
CKD for ≥90 consecutive days with glomerular filtration
rate (GFR) <35 mL/min, GFR <25 mL/min (including the
851

need for sustained dialysis) for 6 consecutive weeks, or in
rare cases when a candidate has rare metabolic disease such
as primary hyperoxaluria.118 However, it should be
remembered that meeting the GFR criteria alone should
not automatically lead to a decision to perform combined
liver and kidney transplantation without evaluating the
potential for reversibility among patients with a high
likelihood of having hepatorenal syndrome as the under-
lying etiology for kidney dysfunction. While native kidney
biopsy has been investigated as a potential means of
identifying patients in need of a combined liver and kidney
transplant rather than a liver-alone transplant, data on this
are limited.119–121 It should be considered in select pa-
tients at low risk of kidney biopsy–related complications to
make a decision regarding combined liver and kidney
transplant versus liver-alone transplant.
Implementation and Challenges
OPTN/UNOS recognizes the issues that surround combined
liver and kidney transplantation, including organ scarcity,
fairness, and the occasional uncertainty about the medical
need for a combined transplant. In response, OPTN/UNOS
implemented a new “safety net” kidney allocation policy.
Patients who are listed for kidney transplant within 60-365
days of a liver transplant are given additional priority for
local kidney allocation if they meet specific medical re-
quirements.118 Critics of this new allocation system argue
that it is too liberal and allows for transplantation among
patients with acute kidney injury likely to recover.117 Others
argue that that this policy may lead to inferior outcomes
among patients who would have been better served
receiving organs simultaneously rather than liver followed
by kidney transplantation. These recommendations stem
from lack of data informing clinical practice and lack of
availability of combined organ transplantation expertise at
many transplant centers. Important issues regarding organ
availability must always be considered. There is some evi-
dence that survival with kidney-after-liver transplantation is
similar to combined liver and kidney transplant and that
kidney outcomes are better for the former.122,123 This may
serve as an impetus to increase the use of the safety net
provision, since, despite the implementation of the new
policy, the number of combined liver and kidney transplants
has remained the same.
Lastly, in the United States, it is also important to
recognize that there are certain patients in need of a
simultaneous liver and kidney transplant who may never
have access to organs regardless of their listing, particularly
those patients with a low MELD (Model for End-Stage Liver
Disease) score but obvious liver dysfunction. Examples
include the patient with liver disease on dialysis who has
frequent ascites or hepatic encephalopathy, but with low
International Normalized Ratio and bilirubin.
In summary, candidates with combined kidney and
liver disease are complex medically, and collaboration with
an experienced transplant hepatologist is essential. When
considering combined liver and kidney transplantation,
852
Puttarajappa et al
careful attention must be paid to issues surrounding organ
scarcity and utilization.
Conclusion
In summary, this commentary highlights the clinical utility
and implementation challenges of the KDIGO guideline for
key areas pertaining to the evaluation and management of
potential kidney transplant candidates. The KDIGO
guideline statements and the strength of these recom-
mendations are centered around evidence-based medicine.
However, many diagnostic and therapeutic approaches
employed in the evaluation and management of kidney
transplant candidates have remained outside the reach of
evidence-based medicine. Thus, while the KDIGO guide-
line is comprehensive, it is limited by the available evi-
dence in many areas, because of which several guideline
statements come as only suggestions or ungraded recom-
mendations. Transplant professionals applying these
guidelines should pay close attention to the strength of
recommendation and the quality of evidence, and use their
clinical judgement to make decisions that serve the best
interest of transplant patients while also balancing utility
and equity of kidney transplantation.
Article Information
Authors’ Full Names and Academic Degrees: Chethan
Puttarajappa, MD, MS, Carrie A. Schinstock, MD, Christine M.
Wu, MD, Nicolae Leca, MD, Vineeta Kumar, MD, Brahm S.
Vasudev, MD, and Sundaram Hariharan, MD.
Authors’ Affiliations: Renal-Electrolyte Division, Department of
Medicine, University of Pittsburgh Medical Center, Pittsburgh, PA
(CMP, CMW, SH); Division of Nephrology & Hypertension,
Department of Medicine, Mayo Clinic, Rochester, MN (CAS);
Division of Nephrology, Department of Medicine, University of
Washington, Seattle, WA (NL); Division of Nephrology,
Department of Medicine, University of Alabama at Birmingham,
Birmingham, AL (VK); and Division of Nephrology, Department of
Medicine, Medical College of Wisconsin, Milwaukee, WI (BSV).
Address for Correspondence: Sundaram Hariharan, MD, Division
of Nephrology, University of Pittsburgh Medical Center, 3459 Fifth
Ave, 7S, Pittsburgh, PA 15213. Email: hariharans@upmc.edu
Support: No financial support was required for the development of
this commentary.
Financial Disclosure: The authors declare that they have no
relevant financial interests.
Acknowledgements: Guideline recommendations included in this
article originally were published in Transplantation, are © 2020
KDIGO, and were reproduced with permission from KDIGO.
Peer Review: Received November 12, 2020, following review and
approval by the NKF Scientific Advisory Board (membership listed
at kidney.org/about/sab; as AJKD Editor-in-Chief, Dr Feldman was
recused) and KDOQI Chair and Vice Chairs (listed at kidney.org/
professionals/guidelines/leadership). Accepted November 17,
2020, after editorial review by a Deputy Editor.
References
1. Chadban SJ, Ahn C, Axelrod DA, et al. KDIGO Clinical Prac-
tice Guideline on the Evaluation and Management of
AJKD Vol 77 | Iss 6 | June 2021
Puttarajappa et al
Candidates for Kidney Transplantation. Transplantation.
2020;104(4)(suppl 1):S11-S103.
2. Freedman VA, Spillman BC. Active life expectancy in the older
US population, 1982–2011: differences between Blacks and
Whites persisted. Health Affairs. 2016;35(8):1351-1358.
3. Saran R, Robinson B, Abbott KC, et al. US Renal Data System
2018 Annual Data Report: epidemiology of kidney disease in the
United States. Am J Kidney Dis. 2019;73(3)(suppl 1):A7-A8.
4. Legeai C, Andrianasolo RM, Moranne O, et al. Benefits of
kidney transplantation for a national cohort of patients aged 70
years and older starting renal replacement therapy. Am J
Transplant. 2018;18(11):2695-2707.
5. Gill JS, Schaeffner E, Chadban S, et al. Quantification of the
early risk of death in elderly kidney transplant recipients. Am J
Transplant. 2013;13(2):427-432.
6. Axelrod DA, Schnitzler MA, Xiao H, et al. An economic
assessment of contemporary kidney transplant practice. Am J
Transplant. 2018;18(5):1168-1176.
7. Heldal K, Midtvedt K, Lønning K, et al. Kidney transplantation:
an attractive and cost-effective alternative for older patients? A
cost-utility study. Clin Kidney J. 2019;12(6):888-894.
8. Alhamad T, Koraishy FM, Lam NN, et al. Cannabis dependence
or abuse in kidney transplantation: implications for posttrans-
plant outcomes. Transplantation. 2019;103(11).
9. Marks WH, Florence L, Lieberman J, et al. Successfully treated
invasive pulmonary aspergillosis associated with smoking
marijuana in a renal transplant recipient. Transplantation.
1996;61(12):1771-1774.
10. Fabbri KR, Anderson-Haag TL, Spenningsby AM, Israni A,
Nygaard RM, Stahler PA. Marijuana use should not preclude
consideration for kidney transplantation. Clin Transplant.
2019;33(10):e13706.
11. Leino AD, Emoto C, Fukuda T, Privitera M, Vinks AA,
Alloway RR. Evidence of a clinically significant drug-drug
interaction between cannabidiol and tacrolimus. Am J Trans-
plant. 2019;19(10):2944-2948.
12. Greenan G, Ahmad SB, Anders MG, Leeser A, Bromberg JS,
Niederhaus SV. Recreational marijuana use is not associated
with worse outcomes after renal transplantation. Clin Trans-
plant. 2016;30(10):1340-1346.
13. Jesse MT, Eshelman A, Christian T, et al. Psychiatric profile of
patients currently listed for kidney transplantation: evidence of
the need for more thorough pretransplant psychiatric evalua-
tions. Transplant Proc. 2019;51(10):3227-3233.
14. Dew MA, Rosenberger EM, Myaskovsky L, et al. Depression
and anxiety as risk factors for morbidity and mortality after or-
gan transplantation: a systematic review and meta-analysis.
Transplantation. 2015;100(5):988-1003.
15. Dobbels F, Vanhaecke J, Dupont L, et al. Pretransplant pre-
dictors of posttransplant adherence and clinical outcome: an
evidence base for pretransplant psychosocial screening.
Transplantation. 2009;87(10):1497-1504.
16. Ladin K, Daniels A, Osani M, Bannuru RR. Is social support
associated with post-transplant medication adherence and
outcomes? A systematic review and meta-analysis. Transplant
Rev. 2018;32(1):16-28.
17. Dew MA, DiMartini AF, De Vito Dabbs A, et al. Rates and risk
factors for nonadherence to the medical regimen after adult solid
organ transplantation. Transplantation. 2007;83(7):858-873.
18. Nghiem DM, Gomez J, Gloston GF, Torres DS, Marek RJ.
Psychological assessment instruments for use in liver and kid-
ney transplant evaluations: scarcity of evidence and recom-
mendations. J Pers Assess. 2020;102(2):183-195.
19. Chen G, Bell CS, Loughhead P, et al. Exploration of the
Stanford Integrated Psychosocial Assessment for
AJKD Vol 77 | Iss 6 | June 2021
transplantation with psychosocial and medical outcomes in
kidney and kidney-pancreas transplant recipients. Prog Trans-
plant. 2019;29(3):230-238.
20. Ladin K, Emerson J, Berry K, et al. Excluding patients from
transplant due to social support: results from a national survey
of transplant providers. Am J Transplant. 2019;19(1):193-203.
21. Campbell S, Pilmore H, Gracey D, Mulley W, Russell C,
McTaggart S. KHA-CARI guideline: recipient assessment for
transplantation. Nephrology (Carlton). 2013;18(6):455-462.
22. Knoll G, Cockfield S, Blydt-Hansen T, et al. Canadian Society
of Transplantation consensus guidelines on eligibility for kidney
transplantation. CMAJ. 2005;173(10):1181-1184.
23. Hucker A, Lawrence C, Sharma S, Farrington K. Adherence
behavior in subjects on hemodialysis is not a clear predictor of
posttransplantation adherence. Kidney Int Rep. 2019;4(8):
1122-1130.
24. Kitamura M, Mochizuki Y, Kitamura S, et al. Prediction of non-
adherence and renal prognosis by pre-transplantation serum
phosphate levels. Ann Transplant. 2019;24:260-267.
25. Dobbels F, De Bleser L, Berben L, et al. Efficacy of a medica-
tion adherence enhancing intervention in transplantation: the
MAESTRO-Tx trial. J Heart Lung Transplant. 2017;36(5):
499-508.
26. Garcia-Roca R, Garcia-Aroz S, Tzvetanov I, Jeon H,
Oberholzer J, Benedetti E. Single center experience with ro-
botic kidney transplantation for recipients with BMI of 40 kg/
m2 or greater: a comparison with the UNOS Registry. Trans-
plantation. 2017;101(1):191-196.
27. Gill JS, Lan J, Dong J, et al. The survival benefit of kidney
transplantation in obese patients. Am J Transplant. 2013;13(8):
2083-2090.
28. Huang E, Shye M, Elashoff D, Mehrnia A, Bunnapradist S.
Incidence of conversion to active waitlist status among
temporarily inactive obese renal transplant candidates. Trans-
plantation. 2014;98(2):177-186.
29. Rubino F, Nathan DM, Eckel RH, et al. Metabolic surgery in the
treatment algorithm for type 2 diabetes: a joint statement by
international diabetes organizations. Diabetes Care.
2016;39(6):861-877.
30. Panwar B, Hanks LJ, Tanner RM, et al. Obesity, metabolic
health, and the risk of end-stage renal disease. Kidney Int.
2015;87(6):1216-1222.
31. Lafranca JA, IJermans JN, Betjes MG, Dor FJ. Body mass index
and outcome in renal transplant recipients: a systematic review
and meta-analysis. BMC Med. 2015;13:111.
32. Fleischmann E, Teal N, Dudley J, May W, Bower JD,
Salahudeen AK. Influence of excess weight on mortality and
hospital stay in 1346 hemodialysis patients. Kidney Int.
1999;55(4):1560-1567.
33. Vashistha T, Mehrotra R, Park J, et al. Effect of age and dialysis
vintage on obesity paradox in long-term hemodialysis patients.
Am J Kidney Dis. 2014;63(4):612-622.
34. Noori N, Kopple JD, Kovesdy CP, et al. Mid-arm muscle
circumference and quality of life and survival in maintenance
hemodialysis patients. Clin J Am Soc Nephrol. 2010;5(12):
2258-2268.
35. Academy of Nutrition and Dietetics. CKD: Medical Nutrition
Therapy: Executive Summary of Recommendations (2020).
Accessed October 24, 2020, https://www.andeal.org/topic.
cfm?menu=5303&cat=5974
36. Ikizler TA, Burrowes JD, Byham-Gray LD, et al. KDOQI Clinical
Practice Guideline for Nutrition in CKD: 2020 update. Am J
Kidney Dis. 2020;76(3)(suppl 1):S1-S107.
37. Mozer AB, Pender JRT, Chapman WH, Sippey ME, Pories WJ,
Spaniolas K. Bariatric surgery in patients with dialysis-
853

dependent renal failure. Obesity Surg. 2015;25(11):2088-
2092.
38. Kassam AF, Mirza A, Kim Y, et al. Long-term outcomes in pa-
tients with obesity and renal disease after sleeve gastrectomy.
Am J Transplant. 2020;20(2):422-429.
39. Al-Bahri S, Fakhry TK, Gonzalvo JP, Murr MM. Bariatric surgery
as a bridge to renal transplantation in patients with end-stage
renal disease. Obesity Surg. 2017;27(11):2951-2955.
40. McAdams-DeMarco MA, Ying H, Olorundare I, et al. Individual
frailty components and mortality in kidney transplant recipients.
Transplantation. 2017;101(9):2126-2132.
41. McAdams-DeMarco MA, Ying H, Thomas AG, et al. Frailty, in-
flammatory markers, and waitlist mortality among patients with
end-stage renal disease in a prospective cohort study. Trans-
plantation. 2018;102(10):1740-1746.
42. Haugen CE, Chu NM, Ying H, et al. Frailty and access to
kidney transplantation. Clin J Am Soc Nephrol. 2019;14(4):
576-582.
43. McAdams-DeMarco MA, Ying H, Van Pilsum Rasmussen S,
et al. Prehabilitation prior to kidney transplantation: results from
a pilot study. Clin Transplant. 2019;33(1):e13450.
44. Kobashigawa J, Dadhania D, Bhorade S, et al. Report from the
American Society of Transplantation on frailty in solid organ
transplantation. Am J Transplant. 2019;19(4):984-994.
45. Buta BJ, Walston JD, Godino JG, et al. Frailty assessment
instruments: Systematic characterization of the uses and con-
texts of highly-cited instruments. Ageing Res Rev. 2016;26:
53-61.
46. Cosio FG, Cattran DC. Recent advances in our understanding
of recurrent primary glomerulonephritis after kidney trans-
plantation. Kidney Int. 2017;91(2):304-314.
47. Vincenti F, Ghiggeri GM. New insights into the pathogenesis
and the therapy of recurrent focal glomerulosclerosis. Am J
Transplant. 2005;5(6):1179-1185.
48. Lee BT, Kumar V, Williams TA, et al. The APOL1 genotype of
African American kidney transplant recipients does not impact
5-year allograft survival. Am J Transplant. 2012;12(7):1924-
1928.
49. Kattah A, Ayalon R, Beck LH Jr, et al. Anti-phospholipase A₂
receptor antibodies in recurrent membranous nephropathy. Am
J Transplant. 2015;15(5):1349-1359.
50. Fervenza FC, Appel GB, Barbour SJ, et al. Rituximab or
cyclosporine in the treatment of membranous nephropathy.
N Engl J Med. 2019;381(1):36-46.
51. Quintana LF, Blasco M, Seras M, et al. Antiphospholipase A2
receptor antibody levels predict the risk of posttransplantation
recurrence of membranous nephropathy. Transplantation.
2015;99(8):1709-1714.
52. Zand L, Lorenz EC, Cosio FG, et al. Clinical findings, pathology,
and outcomes of C3GN after kidney transplantation. J Am Soc
Nephrol. 2014;25(5):1110-1117.
53. Zuber J, Le Quintrec M, Sberro-Soussan R, Loirat C,
Fremeaux-Bacchi V, Legendre C. New insights into postrenal
transplant hemolytic uremic syndrome. Nature Rev Nephrol.
2011;7(1):23-35.
54. Rodríguez de C ordoba S, Hidalgo MS, Pinto S, Tortajada A.
Genetics of atypical hemolytic uremic syndrome (aHUS).
Semin Thromb Hemost. 2014;40(4):422-430.
55. Geel AL, Landman TS, Kal JA, van Doomum GJ, Weimar W.
Varicella zoster virus serostatus before and after kidney trans-
plantation, and vaccination of adult kidney transplant candi-
dates. Transplant Proc. 2006;38(10):3418-3419.
56. Pergam SA, Limaye AP; AST Infectious Diseases Community
of Practice. Varicella zoster virus in solid organ transplantation.
Am J Transplant. 2013;13(suppl 4):138-146.
854
Puttarajappa et al
57. Kumar D. Immunizations following solid-organ transplantation.
Curr Opin Infect Dis. 2014;27(4):329-335.
58. Danziger-Isakov L, Kumar D. Vaccination of solid organ trans-
plant candidates and recipients: Guidelines from the American
Society of Transplantation infectious diseases community of
practice. Clin Transplant. 2019;33(9):e13563.
59. Lee DH, Boyle SM, Malat G, Sharma A, Bias T, Doyle AM. Low
rates of vaccination in listed kidney transplant candidates.
Transpl Infect Dis. 2016;18(1):155-159.
60. Centers for Disease Control and Prevention. Vaccines & im-
munizations. Accessed October 27, 2020, https://www.cdc.
gov/vaccines/index.html
61. Kasper AK, Pallotta AM, Kovacs CS, Spinner ML. Infectious
diseases consult improves vaccination adherence in kidney
transplant candidates. Vaccine. 2018;36(34):5112-5115.
62. Polack FP, Thomas SJ, Kitchin N, et al; C4591001 Clinical Trial
Group. Safety and efficacy of the BNT162b2 mRNA Covid-19
Vaccine. N Engl J Med. 2020;383:2603-2615.
63. Wilt TJ, Harris RP, Qaseem A. Screening for cancer: advice for
high-value care from the American College of Physicians. Ann
Intern Med. 2015;162(10):718-725.
64. Smith RA, Andrews KS, Brooks D, et al. Cancer screening in
the United States, 2019: a review of current American Cancer
Society guidelines and current issues in cancer screening. CA.
Cancer J Clin. 2019;69(3):184-210.
65. Mazzone PJ, Silvestri GA, Patel S, et al. Screening for Lung
Cancer: CHEST guideline and expert panel report. Chest.
2018;153(4):954-985.
66. Benard F, Barkun AN, Martel M, von Renteln D. Systematic
review of colorectal cancer screening guidelines for average-
risk adults: summarizing the current global recommendations.
World J Gastroenterol. 2018;24(1):124-138.
67. Rex DK, Boland CR, Dominitz JA, et al. Colorectal cancer
screening: recommendations for physicians and patients from
the U.S. Multi-Society Task Force on Colorectal Cancer. Am J
Gastroenterol. 2017;112(7):1016-1030.
68. Wolf AMD, Fontham ETH, Church TR, et al. Colorectal cancer
screening for average-risk adults: 2018 guideline update from
the American Cancer Society. CA Cancer J Clin. 2018;68(4):
250-281.
69. U.S. Preventive Services Task Force. Breast cancer screening.
Accessed October 27, 2020, https://www.
uspreventiveservicestaskforce.org/uspstf/recommendation/
breast-cancer-screening
70. American Cancer Society. Survival Rates for Prostate Cancer.
Accessed July 15, 2020, https://www.cancer.org/cancer/
prostate-cancer/detection-diagnosis-staging/survival-rates.
html
71. American Cancer Society. Thyroid cancer survival rates, by
type and stage. Accessed July 15, 2020, https://www.cancer.
org/cancer/thyroid-cancer/detection-diagnosis-staging/survival-
rates.html
72. American Cancer Society. Survival Rates for Kidney Cancer.
Accessed July 15, 2020, https://www.cancer.org/cancer/
kidney-cancer/detection-diagnosis-staging/survival-rates.html
73. Naylor KL, Kim SJ, McArthur E, Garg AX, McCallum MK,
Knoll GA. Mortality in incident maintenance dialysis patients
versus incident solid organ cancer patients: a population-based
cohort. Am J Kidney Dis. 2019;73(6):765-776.
74. Wolk MJ, Bailey SR, Doherty JU, et al. ACCF/AHA/ASE/
ASNC/HFSA/HRS/SCAI/SCCT/SCMR/STS 2013 multi-
modality appropriate use criteria for the detection and risk
assessment of stable ischemic heart disease: a report of the
American College of Cardiology Foundation Appropriate Use
Criteria Task Force, American Heart Association, American
AJKD Vol 77 | Iss 6 | June 2021
Puttarajappa et al
Society of Echocardiography, American Society of Nuclear
Cardiology, Heart Failure Society of America, Heart Rhythm
Society, Society for Cardiovascular Angiography and In-
terventions, Society of Cardiovascular Computed Tomography,
Society for Cardiovascular Magnetic Resonance, and Society of
Thoracic Surgeons. J Am Coll Cardiol. 2014;63(4):380-406.
75. Kristensen SD, Knuuti J, Saraste A, et al. 2014 ESC/ESA
guidelines on non-cardiac surgery: cardiovascular assessment
and management: the joint task force on non-cardiac surgery:
cardiovascular assessment and management of the European
Society of Cardiology (ESC) and the European Society of
Anaesthesiology (ESA). Eur J Anaesthesiol. 2014;31(10):
517-573.
76. Fleisher LA, Fleischmann KE, Auerbach AD, et al. 2014 ACC/
AHA guideline on perioperative cardiovascular evaluation and
management of patients undergoing noncardiac surgery: a
report of the American College of Cardiology/American Heart
Association Task Force on Practice Guidelines. Circulation.
2014;130(24):e278-333.
77. Lentine KL, Costa SP, Weir MR, et al. Cardiac disease evalu-
ation and management among kidney and liver transplantation
candidates: a scientific statement from the American Heart
Association and the American College of Cardiology Founda-
tion. J Am Coll Cardiol. 2012;60(5):434-480.
78. McFalls EO, Ward HB, Moritz TE, et al. Coronary-artery
revascularization before elective major vascular surgery. N Engl
J Med. 2004;351(27):2795-2804.
79. Garcia S, Moritz TE, Goldman S, et al. Perioperative compli-
cations after vascular surgery are predicted by the revised
cardiac risk index but are not reduced in high-risk subsets with
preoperative revascularization. Circ Cardiovasc Qual Out-
comes. 2009;2(2):73-77.
80. Patel MR, Calhoon JH, Dehmer GJ, et al. ACC/AATS/AHA/
ASE/ASNC/SCAI/SCCT/STS 2017 Appropriate use criteria
for coronary revascularization in patients with stable ischemic
heart disease: a report of the American College of Cardiology
Appropriate Use Criteria Task Force, American Association for
Thoracic Surgery, American Heart Association, American So-
ciety of Echocardiography, American Society of Nuclear Car-
diology, Society for Cardiovascular Angiography and
Interventions, Society of Cardiovascular Computed Tomogra-
phy, and Society of Thoracic Surgeons. J Nucl Cardiol.
2017;24(5):1759-1792.
81. Ojo AO, Hanson JA, Wolfe RA, Leichtman AB, Agodoa LY,
Port FK. Long-term survival in renal transplant recipients with
graft function. Kidney Int. 2000;57(1):307-313.
82. Gowdak LH, de Paula FJ, Cesar LA, et al. Screening for sig-
nificant coronary artery disease in high-risk renal transplant
candidates. Coron Artery Dis. 2007;18(7):553-558.
83. Patel RK, Mark PB, Johnston N, et al. Prognostic value of
cardiovascular screening in potential renal transplant re-
cipients: a single-center prospective observational study. Am J
Transplant. 2008;8(8):1673-1683.
84. De Lima JJ, Wolff Gowdak LH, de Paula FJ, Ianhez LE, Franchini
Ramires JA, Krieger EM. Validation of a strategy to diagnose
coronary artery disease and predict cardiac events in high-risk
renal transplant candidates. Coron Artery Dis. 2010;21(3):
164-167.
85. Bangalore S, Maron DJ, O’Brien SM, et al. Management of
coronary disease in patients with advanced kidney disease.
N Engl J Med. 2020;382(17):1608-1618.
86. Ying T, Tran A, Webster AC, et al. Screening for asymptomatic
coronary artery disease in waitlisted kidney transplant candi-
dates: a cost-utility analysis. Am J Kidney Dis. 2020;75(5):
693-704.
AJKD Vol 77 | Iss 6 | June 2021
87. Schoenberg NC, Argula RG, Klings ES, Wilson KC,
Farber HW. Prevalence and mortality of pulmonary hyperten-
sion in ESRD: a systematic review and meta-analysis. Lung.
2020;198(3):535-545.
88. Issa N, Krowka MJ, Griffin MD, Hickson LJ, Stegall MD,
Cosio FG. Pulmonary hypertension is associated with reduced
patient survival after kidney transplantation. Transplantation.
2008;86(10):1384-1388.
89. Wang SC, Garcia R, Torosoff M, et al. Influence of mildly and
moderately elevated pulmonary artery systolic pressure on
post-renal transplantation survival and graft function. Echocar-
diography. 2019;36(1):22-27.
90. Caughey MC, Detwiler RK, Sivak JA, Rose-Jones LJ,
Kshirsagar AV, Hinderliter AL. Five-year outcomes of pulmonary
hypertension with and without elevated left atrial pressure in
patients evaluated for kidney transplantation. Transplantation.
2020;104(10):2113-2119.
91. Walther CP, Nambi V, Hanania NA, Navaneethan SD. Diag-
nosis and management of pulmonary hypertension in patients
with CKD. Am J Kidney Dis. 2020;75(6):935-945.
92. Reddy YN, Lunawat D, Abraham G, et al. Progressive pulmo-
nary hypertension: another criterion for expeditious renal
transplantation. Saudi Journal Kidney Dis Transpl. 2013;24(5):
925-929.
93. Casas-Aparicio G, Castillo-Martínez L, Orea-Tejeda A, Abasta-
Jim
enez M, Keirns-Davies C, Rebollar-Gonz alez V. The effect of
successful kidney transplantation on ventricular dysfunction
and pulmonary hypertension. Transplant Proc. 2010;42(9):
3524-3528.
94. Finkelhor RS, Lewis SA, Pillai D. Limitations and strengths of
Doppler/echo pulmonary artery systolic pressure-right heart
catheterization correlations: a systematic literature review.
Echocardiography. 2015;32(1):10-18.
95. Gali
e N, Humbert M, Vachiery JL, et al. 2015 ESC/ERS
Guidelines for the diagnosis and treatment of pulmonary hy-
pertension: the Joint Task Force for the Diagnosis and Treat-
ment of Pulmonary Hypertension of the European Society of
Cardiology (ESC) and the European Respiratory Society
(ERS): Endorsed by: Association for European Paediatric and
Congenital Cardiology (AEPC), International Society for Heart
and Lung Transplantation (ISHLT). Eur Heart J. 2016;37(1):
67-119.
96. Tambur AR, Campbell P, Claas FH, et al. Sensitization in Trans-
plantation: Assessment of Risk (STAR) 2017 Working Group
meeting report. Am J Transplant. 2018;18(7):1604-1614.
97. Bray RA, Lebeck LK, Gebel HM. The flow cytometric cross-
match. Dual-color analysis of T cell and B cell reactivities.
Transplantation. 1989;48(5):834-840.
98. Schinstock CA, Gandhi M, Cheungpasitporn W, et al. Kidney
transplant with low levels of DSA or low positive b-flow cross-
match: an underappreciated option for highly sensitized
transplant candidates. Transplantation. 2017;101(10):2429-
2439.
99. Zachary AA, Sholander JT, Houp JA, Leffell MS. Using real
data for a virtual crossmatch. Hum Immunol. 2009;70(8):574-
579.
100. Tambur AR, Herrera ND, Haarberg KM, et al. Assessing anti-
body strength: comparison of MFI, C1q, and titer information.
Am J Transplant. 2015;15(9):2421-2430.
101. Schinstock CA, Gandhi MJ, Stegall MD. Interpreting anti-HLA
antibody testing data: a practical guide for physicians. Trans-
plantation. 2016;100(8):1619-1628.
102. Jackson KR, Covarrubias K, Holscher CM, et al. The national
landscape of deceased donor kidney transplantation for the
highly sensitized: transplant rates, waitlist mortality, and
855

posttransplant survival under KAS. Am J Transplant.
2019;19(4):1129-1138.
103. Schinstock CA, Smith BH, Montgomery RA, et al. Managing
highly sensitized renal transplant candidates in the era of kidney
paired donation and the new kidney allocation system: Is there still
a role for desensitization? Clin Transplant. 2019;33(12):e13751.
104. Orandi BJ, Luo X, Massie AB, et al. Survival benefit with kidney
transplants from HLA-incompatible live donors. N Engl J Med.
2016;374(10):940-950.
105. Kulkarni S, Ladin K. Ethical principles governing organ trans-
plantation apply to paired exchange programs. Am J Trans-
plant. 2020;20(5):1223-1224.
106. Leung N, Griffin MD, Dispenzieri A, et al. Living donor kidney
and autologous stem cell transplantation for primary systemic
amyloidosis (AL) with predominant renal involvement. Am J
Transplant. 2005;5(7):1660-1670.
107. Joly F, Cohen C, Javaugue V, et al. Randall-type monoclonal
immunoglobulin deposition disease: novel insights from a
nationwide cohort study. Blood. 2019;133(6):576-587.
108. Angel-Korman A, Stern L, Sarosiek S, et al. Long-term outcome
of kidney transplantation in AL amyloidosis. Kidney Int.
2019;95(2):405-411.
109. Heybeli C, Bentall A, Wen J, et al. A study from The Mayo Clinic
evaluated long-term outcomes of kidney transplantation in pa-
tients with immunoglobulin light chain amyloidosis. Kidney Int.
Published online July 23, 2020;99(3):707-715.
110. International Myeloma Working Group. Criteria for the classifi-
cation of monoclonal gammopathies, multiple myeloma and
related disorders: a report of the International Myeloma Work-
ing Group. Br J Haematol. 2003;121(5):749-757.
111. Rajkumar SV, Dimopoulos MA, Palumbo A, et al. International
Myeloma Working Group updated criteria for the diagnosis of
multiple myeloma. Lancet Oncol. 2014;15(12):e538-548.
112. Alfano G, Fontana F, Colaci E, et al. Monoclonal gammopathy of
undetermined significance after kidney transplantation: single-
center experience. Transplantation. 2017;101(11):e337-e342.
113. Ripoll C, Groszmann R, Garcia-Tsao G, et al. Hepatic venous
pressure gradient predicts clinical decompensation in patients
with compensated cirrhosis. Gastroenterology. 2007;133(2):
481-488.
856
Puttarajappa et al
114. Watt KD, Pedersen RA, Kremers WK, Heimbach JK,
Charlton MR. Evolution of causes and risk factors for mortality
post-liver transplant: results of the NIDDK long-term follow-up
study. Am J Transplant. 2010;10(6):1420-1427.
115. Sharma P, Shu X, Schaubel DE, Sung RS, Magee JC. Propensity
score-based survival benefit of simultaneous liver-kidney trans-
plant over liver transplant alone for recipients with pretransplant
renal dysfunction. Liver Transpl. 2016;22(1):71-79.
116. Al Riyami D, Alam A, Badovinac K, Ivis F, Trpeski L,
Cantarovich M. Decreased survival in liver transplant patients
requiring chronic dialysis: a Canadian experience. Trans-
plantation. 2008;85(9):1277-1280.
117. Asch WS, Bia MJ. New organ allocation system for combined
liver-kidney transplants and the availability of kidneys for
transplant to patients with stage 4-5 CKD. Clin J Am Soc
Nephrol. 2017;12(5):848-852.
118. Boyle GJ. OPTN/UNOS Kidney Transplantation Committee.
Simultaneous liver kidney (SLK) allocation policy. Accessed
June 29, 2020, http://optn.transplant.hrsa.gov/media/1192/
0815-12_SLK_Allocation.pdf
119. Pichler RH, Huskey J, Kowalewska J, et al. Kidney biopsies may
help predict renal function after liver transplantation. Trans-
plantation. 2016;100(10):2122-2128.
120. Tanriover B, Mejia A, Weinstein J, et al. Analysis of kidney
function and biopsy results in liver failure patients with
renal dysfunction: a new look to combined liver kidney alloca-
tion in the post-MELD era. Transplantation. 2008;86(11):
1548-1553.
121. Wadei HM, Geiger XJ, Cortese C, et al. Kidney allocation to
liver transplant candidates with renal failure of undetermined
etiology: role of percutaneous renal biopsy. Am J Transplant.
2008;8(12):2618-2626.
122. Cullaro G, Verna EC, Emond JC, Orandi BJ, Mohan S, Lai JC.
Early kidney allograft failure after simultaneous liver-kidney
transplantation (SLKT): evidence for utilization of the safety
net? Transplantation. 2021;105(4):816-823.
123. Jay CL, Washburn WK, Rogers J, Harriman D, Heimbach J,
Stratta RJ. Difference in survival in early kidney after liver
transplantation compared with simultaneous liver-kidney
transplantation: evaluating the potential of the “safety net.
J Am Coll Surg. 2020;230(4):463-473.
AJKD Vol 77 | Iss 6 | June 2021
Editorial
APOL1, Black Race, and Kidney Disease: Turning
Attention to Structural Racism
~a, Jennifer Tsai, and Vanessa Grubbs
Jessica P. Cerden
T he incidence of kidney failure is nearly 3 times higher
in Black relative to White Americans.1 A significant
body of biomedical literature links this racial inequity to
polymorphic variation in the gene encoding apolipopro-
tein L1 (APOL1). In fact, in a PubMed search of “APOL1”
and “kidney disease,” 86% of results feature abstracts
mentioning African ancestry or African American, Black, or
non-White race. Multiple recent articles state that APOL1
genetic variants are found “exclusively” in people, or
chromosomes, of “African” origin.2-4 Further, an ongoing
clinical trial—the APOL1 Long-term Kidney Transplantation
Outcomes Network (APOLLO) study—seeks to examine
the relationship between APOL1 genotype and racial dis-
parities in allograft survival and in kidney failure among
living kidney donors.5 Such tight coupling of APOL1 vari-
ation and Black race lends the impression that racial dis-
parities in kidney disease are genetically determined and
homogeneously distributed across persons labeled as
“Black,” a misguided hypothesis that can skew investment
toward interventions with limited effect. In this editorial,
we summarize the current literature on APOL1, discuss the
scientific flaws of linking particular APOL1 variants with
Black race, and emphasize the role of structural racism as a
determinant of racial disparities in kidney disease.
Background on APOL1
Genetic variation on chromosome 22q has been associated
with increased susceptibility to focal segmental glomer-
ulosclerosis, HIV-associated nephropathy, and
hypertension-attributed kidney failure. In particular,
strong signals have been identified in association with 2
nonsynonymous coding variants (for amino acids 342 and
384) of APOL1, termed G1, and with a 6-base-pair deletion
that removes 2 amino acids (388-389), termed G2.6
Patterns of linkage disequilibrium in the region of G1
and G2 provide strong evidence for positive selection. In
particular, because the APOL1 variants evade activity of the
serum resistance–associated protein of T. b. rhodiense—one
of the vectors of African sleeping sickness—it is hypoth-
esized that heterozygosity for these alleles confers a se-
lective advantage against the disease.7 Furthermore, the
presence of these variants in Yoruban populations from
Nigeria and in some Black Americans, and their relative
scarcity in European and Asian populations, evince the
likelihood of a recent selection event in West Africa.6,8
Together, these genetic data have led clinical in-
vestigators to focus on APOL1 polymorphisms as a key,
rather than a contributory, explanatory variable in racial
disparities in kidney disease. For instance, in a recent
AJKD Vol 77 | Iss 6 | June 2021
editorial, Kopp and Winkler9 name “variants in APOL1 as
major driver[s] of kidney disease in Blacks” and accord-
ingly discuss recommendations for clinical testing of risk
genotypes. This approach, however, is based upon a
misapplication of population genetics and exacerbated by
conflation of genetic ancestry and race.
The Error and Harm of Linking APOL1 Variants to
Black Race
Race is a social and political construct perpetuated by those
in power and should not be used as a substitute for
thoughtful, hypothesis-driven genetic analyses.10 Human
genetic and phenotypic variants, including superficial
physical differences like skin color and hair texture, are
clinally distributed across geographic space and do not map
neatly to races or continents.11,12 Researchers may lean on
terms like “ancestry,” which imply evolutionary history and
may carry biological significance at an individual level of
analysis13,14; however, when paired with a wide
geographical modifier (ie, “African”), such phrases lose
meaning and become little more than a substitute for race.
Genetic conclusions cannot be made about the populations
of entire continents or entire races; accordingly, discussing
APOL1-associated kidney disease as a concern of—and
exclusive to—people of “African descent” reflects racist
practice in medicine and research,15 even if and often done
unconsciously. In this section, we identify specific flaws
inherent in the logic of associating APOL1 G1 and G2 alleles
with Black race and outline potential harms to patients.
Sweeping associations between certain APOL1 poly-
morphisms and Black race risk overlooking the impact of
APOL1-associated nephropathy on other populations.
Although APOL1 G1 and G2 are most often seen in people
of West African descent, other groups—including Euro-
pean American, Pakistani, and Latin American patient-
s—also carry these polymorphisms.16-18 Even though this
is likely owing to admixture, these findings suggest that
limiting clinical inquiries about APOL1-associated kidney
disease to “Black” patients inappropriately excludes pa-
tients of other racial or ethnic backgrounds. The APOLLO
study, for instance, only recruits Black kidney donors,
restricting insights into the role of the APOL1 genotype on
clinical outcomes to an artificially smaller population.
Researchers may be reluctant to generalize their findings to
White, Asian, and Latinx patients who carry G1 and G2
alleles who would otherwise benefit.
Moreover, the frequency of the G1 and G2 alleles varies
widely across the African continent. For instance, the fre-
quency of G1 among the Igbo in Nigeria is 30.2% but 0%
857

among the Lemande of Cameroon.17 Yet, patients from
both of these ethnic groups would be racialized as “Black”
in the United States. Additionally, even among genetically
similar ethnic groups, significant variation in allelic fre-
quency is evident. Genetic research comparing the geno-
types of the Yoruba of Nigeria and the Luhya of Kenya, 2
ethnic groups of high genetic similarity, found that no
other sites were as differentiated between the 2 groups as
G1. This large variation cannot be explained by genetic
drift and instead points to a selection event occurring
within a narrowly circumscribed population in West Af-
rica.7 Therefore, conclusions about APOL1 polymorphisms
cannot be generalized to all residents of the African
continent, nor to all individuals of African descent.
Because many enslaved Africans were captured from the
West African coast during the trans-Atlantic slave trade,
Black Americans descended from slaves may be expected to
exhibit higher frequencies of G1 and G2. Indeed, current
research suggests that over 20% of Black Americans carry 1
of these variants and 13% carry 2.19 However, recent
patterns of Black African migration reveal an influx from
Northern and Eastern Africa, regions where the G1 and G2
alleles are far less prevalent.20 Accordingly, given the
continued inflow of Black migrants from diverse regions of
Africa—and ongoing admixture between descendants of
the enslaved with genetically distinct populations—wide
heterogeneity in frequency of APOL1 variants among in-
dividuals racialized as “Black” may be expected, depending
on population sampling. In summary, sweeping attribu-
tions of kidney disease risk among Black Americans to
particular APOL1 alleles overlook the increasing genetic
diversity within this community and others.
Incautious conflation of APOL1 genotype and Black race
can promote misinformed conclusions that all Black patients
experience genetic predisposition toward kidney disease.21
This is reflected in the Kidney Donor Risk Index, a tool that
includes Black race in its calculation of allograft longevity for
donated kidneys. Substitution of APOL1 genotype for race in
the equation would improve the Kidney Donor Risk Index
for 85% to 90% of Black donors, likely reducing discard of
high-demand kidneys.22 Thus, generalizations about Black
race and APOL1 G1 and G2 alleles are scientifically inap-
propriate and contribute to health care waste.
Finally, the link between APOL1 polymorphisms and
kidney disease is not as straightforward as some research
implies. Although possessing 2 “risk” alleles yields an odds
ratio of 10.5 and 7.3 for focal segmental glomerulosclerosis
and hypertension-attributed kidney failure, respectively,6
the overall lifetime risk is just 4% and 12%. Additionally,
possessing 2 “risk” alleles explains just 37%, 18%, and 7%
of the variance in risk for HIV-associated nephropathy, focal
segmental glomerulosclerosis, and hypertension-attributed
kidney failure, respectively,19 while most reports find no
association with diabetic nephropathy, the leading cause of
kidney disease.17 APOL1 variants account for roughly 30% of
the excess nondiabetic kidney disease in Black Americans,23
858
~a et al
Cerden
highlighting the importance of environmental factors un-
derpinning racial disparities in kidney disease.
Turning Attention to Structural Racism as a
Determinant of Kidney Disease
Racial groups are not clearly demarcated by patterns of
genetic variation; however, in the United States, histories
of inequitable distribution of resources along racial lines
have systematically created unequal environments and life
opportunities between Black and White communities,
which differentially impact their health. Structural racism
refers to “the totality of ways in which societies foster
racial discrimination through mutually reinforcing sys-
tems.”24 Disparities in kidney disease derive from unequal
conditions in wealth, employment, residence, toxic envi-
ronmental exposures, nutrition, education, health care
access, and psychosocial stress.25-29 Mediators of kidney
failure—such as diabetes, hypertension, and HIV—can be
traced to racialization of the US food system,30 mass
incarceration,31 and poverty.32 Structural determinants of
health thereby deserve at least as many dedicated resources
as those afforded to genetic inquiry, yet research and
funding streams consistently favor investigation into bio-
logical and genetic underpinnings of racial disparities in
health over issues of racism or racialization.33,34 In fact, a
simple Boolean search of the terms “kidney disease” and
“racism” in the National Institutes of Health Research
Portfolio Online Reporting Tools Database for 2019 to
2020 yields just 5 results, whereas pairing “kidney dis-
ease” with “genetic” yields 754 results.33
Racism itself can induce multiple physiologic changes
involved in the pathogenesis of kidney disease. Chronic
stress from interpersonal and structural racism can initiate
long-term activation of the hypothalamic-pituitary-adrenal
axis, which can constitutively increase vascular tension.
Such stress can also lead to failure to downregulate the
inflammatory response and increased production of reac-
tive oxygen species, which can promote interstitial
fibrosis.25 Further, epigenetic modifications—such as
those resulting through fetal programming when racial
discrimination increases prenatal cortisol exposure—may
contribute to the excess risk for kidney disease among
Black Americans.35 Thus, although race is a poor proxy for
population genetics, adverse social conditions experienced
by marginalized populations are consistent and powerful
contributors to disease risk. Overemphasis on racialized
genetics—rather than racism—may result in insufficient
examination of structural inequities as major determinants
of racial disparities in kidney disease.
Future research should examine how racist—and anti-
racist—policies affect kidney health. Studies can assess the
effects of policies like Medicaid expansion,36,37 decrimi-
nalization,38 or, in the future, reparations39 on kidney
health. Racism should be identified and carefully analyzed:
Research that merely examines race as a “risk factor”—or
provides a genetic basis for racial differences in health
AJKD Vol 77 | Iss 6 | June 2021
 ~a et al
Cerden
outcomes—should be scrupulously critiqued.40 Under-
standing the effects of policy on health can inform both
public and private action, dismantling the oppressive
structures that contribute to racial disparities.
Conclusion
Carrying 2 APOL1 “risk” alleles significantly increases an
individual’s odds of developing kidney disease; however,
even with a high-risk genotype, overall lifetime risk re-
mains relatively low and much of the variance in risk is
explained by other factors, including structural racism.
Although some Black Americans carry high levels of West
African ancestry owing to the shameful history of chattel
slavery, wide variation in APOL1 “risk” allele frequency is
expected among all individuals racialized as “Black.” It is
therefore inappropriate to conclude that APOL1 mutations
increase risk for kidney disease among all Black Americans
or to limit research or testing of APOL1 genotypes to Black
individuals: Such determinations may constrain clinical
decision making and misdirect resources to the detriment
of patients. Decisions regarding APOL1 testing in clinical
care should engage key stakeholders, including a diverse
group of patients, nephrologists, primary care providers,
researchers, and bioethicists, considering opportunities for
patient counseling and the need to develop targeted
treatments.41 Finally, since APOL1 polymorphisms cannot
fully explain the racial disparities in kidney disease,
research that examines how structural racism mediates
impairments in kidney function deserves closer attention.
Article Information
Authors’ Full Names and Academic Degrees: Jessica P. Cerde~na,
MPhil, Jennifer Tsai, MD, MEd, and Vanessa Grubbs, MD, MPH.
Authors’ Affiliations: Yale School of Medicine (JPC), Department of
Anthropology, Yale University (JPC), and Department of Emergency
Medicine, Yale School of Medicine (JT), New Haven, CT; and
Division of Nephrology, Department of Medicine, University of
California, San Francisco School of Medicine, San Francisco, CA
(VG).
Address for Correspondence: Jessica P. Cerde~ na, MPhil, 10
Sachem St, New Haven, CT 06511. Email: jessica.cerdena@yale.
edu
Support: Ms Cerde~ na is funded by the Robert Wood Johnson
Foundation Health Policy Research Scholars program and the
National Institute of Health Medical Scientist Training Program
Grant T32GM136651.
Financial Disclosure: The authors declare that they have no
relevant financial interests.
Peer Review: Received August 6, 2020. Evaluated by 2 external
peer reviewers, with direct editorial input from an Associate Editor
and a Deputy Editor. Accepted in revised form November 17, 2020.
Publication Information: © 2021 by the National Kidney Founda-
tion, Inc. Published online January 22, 2021 with doi 10.1053/
j.ajkd.2020.11.029
References
1. Saran R, Robinson B, Abbott KC, et al. US Renal Data System
2018 Annual Data Report: epidemiology of kidney disease in the
United States. Am J Kidney Dis. 2019;73 (3)(suppl 1):A7-A8.
AJKD Vol 77 | Iss 6 | June 2021
2. Kopp JB, Winkler CA. Genetics, genomics, and precision
medicine in end-stage kidney disease. Semin Nephrol.
2018;38(4):317-324.
3. Cheatham AM, Davis SE, Khatua AK, Popik W. Blocking the 50
splice site of exon 4 by a morpholino oligomer triggers APOL1
protein isoform switch. Sci Rep. 2018;8(1):8739.
4. Hughson MD, Hoy WE, Mott SA, Bertram JF, Winkler CA,
Kopp JB. APOL1 risk variants independently associated with
early cardiovascular disease death. Kidney Int Rep. 2018;3(1):
89-98.
5. Freedman BI, Moxey-Mims MM, Alexander AA, et al. APOL1
Long-term Kidney Transplantation Outcomes Network
(APOLLO): design and rationale. Kidney Int Rep. 2020;5(3):
278-288.
6. Genovese G, Friedman DJ, Ross MD, et al. Association of
trypanolytic ApoL1 variants with kidney disease in African
Americans. Science. 2010;329(5993):841-845.
7. Thomson R, Genovese G, Canon C, et al. Evolution of the
primate trypanolytic factor APOL1. PNAS. 2014;111(20):
E2130-E2139.
8. Simino J, Rao DC, Freedman BI. Novel findings and future di-
rections on the genetics of hypertension. Curr Opin Nephrol
Hypertens. 2012;21(5):500-507.
9. Kopp JB, Winkler CA. Genetic testing for APOL1 genetic
variants in clinical practice: finally starting to arrive. CJASN.
2020;15(1):126-128.
10. Tsai J, Cerde~ na JP, Khazanchi R, et al. There is no ‘African
American Physiology’: the fallacy of racial essentialism. J Intern
Med. 2020;288(3):368-370.
11. American Association of Physical Anthropologists. AAPA
statement on race & racism. March 27, 2019. Accessed March
25, 2020. https://physanth.org/about/position-statements/aapa-
statement-race-and-racism-2019
12. The American Society of Human Genetics. ASHG denounces
attempts to link genetics and racial supremacy. Am J Hum
Genet. 2018;103(5):636.
13. Yudell M, Roberts D, DeSalle R, Tishkoff S. Taking race out of
human genetics. Science. 2016;351(6273):564-565.
14. Bolnick D. Individual ancestry inference and the reification of
race as a biological phenomenon. In: Lee S, Koenig B,
Richardson SS, eds. Revisiting Race in a Genomic Age.
Rutgers University Press; 2008:70-85.
15. Cerde~ na JP, Plaisime MV, Tsai J. From race-based to race-
conscious medicine: how anti-racist uprisings call us to act.
Lancet. 2020;396(10257):1125-1128.
16. Kopp JB, Nelson GW, Sampath K, et al. APOL1 genetic
variants in focal segmental glomerulosclerosis and HIV-
associated nephropathy. J Am Soc Nephrol. 2011;22(11):
2129-2137.
17. Limou S, Nelson GW, Kopp JB, Winkler CA. APOL1 kidney risk
alleles: population genetics and disease associations. Adv
Chronic Kidney Dis. 2014;21(5):426-433.
18. Nadkarni GN, Gignoux CR, Sorokin EP, et al. Worldwide fre-
quencies of APOL1 renal risk variants. N Engl J Med.
2018;379(26):2571-2572.
19. Dummer PD, Limou S, Rosenberg AZ, et al. APOL1 kidney
disease risk variants: an evolving landscape. Semin Nephrol.
2015;35(3):222-236.
20. Capps R, McCabe K, Fix M. Diverse Streams: Black African
Migration to the United States. Migration Policy Institute; 2012:
1-23. https://www.fcd-us.org/assets/2016/04/African-Migration-
to-the-United-States.pdf
21. Gordon EJ, Am ortegui D, Blancas I, Wicklund C, Friedewald J,
Sharp RR. A focus group study on African American living
donors’ treatment preferences, sociocultural factors, and health
859

beliefs about apolipoprotein L1 genetic testing. Prog Transpl.
2019;29(3):239-247.
22. Julian BA, Gaston RS, Brown WM, et al. Effect of replacing
race with apolipoprotein L1 genotype in calculation of kidney
donor risk index. Am J Transplant. 2017;17(6):1540-1548.
23. Kaufman JS, Rushani D, Cooper RS. Nature versus nurture in
the explanations for racial/ethnic health disparities. In: Suzuki K,
Von Vacano DA, eds. Reconsidering Race: Social Science
Perspectives on Racial Categories in the Age of Genomics.
Oxford University Press; 2018:120.
24. Bailey ZD, Krieger N, Ag enor M, Graves J, Linos N, Bassett MT.
Structural racism and health inequities in the USA: evidence
and interventions. Lancet. 2017;389(10077):1453-1463.
25. Nicholas SB, Kalantar-Zadeh K, Norris KC. Racial disparities in
kidney disease outcomes. Semin Nephrol. 2013;33(5):409-
415.
26. Laster M, Shen JI, Norris KC. Kidney disease among African
Americans: a population perspective. Am J Kidney Dis.
2018;72(5):S3-S7.
27. Young BA. The Interaction of Race, Poverty, and CKD. Am J
Kidney Dis. 2010;55(6):977-980.
28. Norris K, Mehrotra R, Nissenson AR. Racial differences in
mortality and end-stage renal disease. Am J Kidney Dis.
2008;52(2):205-208.
29. Crews DC, Purnell TS. COVID-19, racism, and racial dispar-
ities in kidney disease: galvanizing the kidney community
response. JASN. 2020;31(8):1-3.
30. Ayazi H, Elsheikh E. The US Farm Bill: Corporate Power and
Structural Racialization in the US Food System. UC Berkeley
Othering & Belonging Institute. October 28, 2015. Accessed
July 22, 2020. https://escholarship.org/uc/item/55v6q06x
31. Wang EA, Pletcher M, Lin F, et al. Incarceration, incident hy-
pertension, and access to health care: findings from the Cor-
onary Artery Risk Development in Young Adults (CARDIA)
Study. Arch Intern Med. 2009;169(7):687-693.
860
~a et al
Cerden
32. Arnold EA, Rebchook GM, Kegeles SM. “Triply cursed”: racism,
homophobia and HIV-related stigma are barriers to regular HIV
testing, treatment adherence and disclosure among young
Black gay men. Cult Health Sex. 2014;16(6):710-722.
33. Hardeman RR, Karbeah J. Examining racism in health services
research: a disciplinary self-critique. Health Serv Res.
2020;55(S2):777-780.
34. Krieger N. Stormy weather: race, gene expression, and the
science of health disparities. Am J Public Health. 2005;95(12):
2155-2160.
35. Kuzawa CW, Sweet E. Epigenetics and the embodiment of
race: developmental origins of US racial disparities in cardio-
vascular health. Am J Hum Biol. 2009;21(1):2-15.
36. Sommers BD, Blendon RJ, Orav EJ, Epstein AM. Changes in
utilization and health among low-income adults after Medicaid
expansion or expanded private insurance. JAMA Intern Med.
2016;176(10):1501-1509.
37. Kurella-Tamura M, Goldstein BA, Hall YN, Mitani AA,
Winkelmayer WC. State Medicaid coverage, ESRD incidence,
and access to care. JASN. 2014;25(6):1321-1329.
38. Grucza RA, Vuolo M, Krauss MJ, et al. Cannabis decriminal-
ization: a study of recent policy change in five U.S. states. Int J
Drug Policy. 2018;59:67-75.
39. Vigdor N. North Carolina City Approves Reparations for Black
Residents. The New York Times. July 16, 2020; section A:15.
Accessed July 22, 2020. https://www.nytimes.com/2020/
07/16/us/reparations-asheville-nc.html
40. Boyd RW, Lindo EG, Weeks LD, McLemore MR. On Racism: A
New Standard For Publishing On Racial Health Inequities.
Health Affairs: Health Affairs Blog. July 2, 2020. Accessed July
23, 2020. https://www.healthaffairs.org/do/10.1377/hblog202
00630.939347/full/
41. Young BA, Blacksher E, Cavanaugh KL, et al. Apolipoprotein
L1 testing in African Americans: involving the community in
policy discussions. Am J Nephrol. 2019;50(4):303-311.
AJKD Vol 77 | Iss 6 | June 2021
Editorial
Familial Aggregation of CKD: Gene or
Environment?
Lucrezia Carlassara, Francesca Zanoni, and Ali G. Gharavi
N ephrologists often see patients who have a family
history of chronic kidney disease (CKD), but the
clinical utility of this information is not always clear.
Alongside comorbid conditions (such as diabetes, hyper-
Related Article, p. 869
tension, and hypercholesterolemia) and constitutive pa-
rameters (such as age, ethnicity, and sex), a positive family
history of CKD has gained greater attention as a major risk
factor for CKD and is predictive of a positive genetic
diagnosis.1 In recent years, nephrologists also increasingly
use genetic testing in medical care, and patients frequently
inquire about genetic kidney disease and the risk of disease
transmission to their offspring. The identification of ge-
netic risk factors for CKD has the potential to improve early
detection and also increase our understanding of the
pathogenesis of disease. However, several questions
remain unanswered, such as, what is the magnitude of risk
imparted by a positive family history of CKD? To what
extent can family history of CKD be explained by shared
genetic factors versus environmental triggers? How can we
best use the family history information to improve risk
stratification of CKD and its sequelae? In this issue of AJKD,
Zhang et al2 begin to explore several aspects of these
matters.
A positive family history of CKD could reflect shared
genetic susceptibility as well as shared socioeconomic and
environmental factors within a family. Genetic factors
include both Mendelian (monogenic) disorders and
polygenic risk. Recent clinical sequencing studies have
detected diagnostic variants in 10% to 25% of patients of
various CKD populations, suggesting a high burden of
Mendelian disorders among patients with nephropathy.1
In addition to monogenic disorders, polygenic risk—the
cumulative effect of common genetic variants with low
effect size conferring risk of disease—has been demon-
strated for a number of glomerulonephritides, such as IgA
nephropathy (over 15 risk loci described to date),3
membranous nephropathy,4 and nephrotic syndrome
(APOL1 loci).5 Moreover, recent genome-wide association
studies (GWAS) discovered nearly 300 common genetic
variants associated with CKD, kidney function, and albu-
minuria.6,7 On the other side of the spectrum, many
environmental susceptibility factors, such as diet, lifestyle,
toxicant exposure (eg, secondhand smoke), or infectious
diseases, may be shared within the same family and pro-
mote CKD. Nowadays, there are emerging methods to
systematically analyze potential environmental exposures
contributing to risk of disease8.
AJKD Vol 77 | Iss 6 | June 2021
In the current report, Zhang et al2 explore the familial
aggregation of CKD, and estimate the heritability of kidney
disease and its related traits, in a cross-sectional 3-
generation family study within the northern Netherlands
general population (155,911 participants, predominantly
European ancestry). Zhang et al reported a CKD prevalence
of 1.19% in the studied cohort as a whole, and 5.8% in
families where at least 1 member is affected by CKD. By
including household and spousal effects in the models, the
authors were also able to estimate the influence of envi-
ronmental effects on disease variance. Individuals with a
first-degree relative affected by CKD had a lower age-
specific mean value of estimated glomerular filtration rate
(eGFR) and a 3-fold increased recurrence risk ratio for CKD
(3.04 [95% CI, 2.26-4.09]) compared with the risk in the
entire cohort. The authors did not identify any clear pattern
by kinship type or sex accounting for familial aggregation of
CKD. Importantly, they also demonstrated that spouses of
individuals with CKD also had higher CKD recurrence risk
ratio compared with the risk in the entire population (1.56
[95% CI, 1.20-2.00]), suggesting that shared environ-
mental factors (or potentially, assortative mating) have a
significant role in the determination of familial aggregation
of CKD. In addition, the authors explored the heritability
estimates of several parameters related to kidney function
and CKD. Heritability aims to quantify the influences of
genetic factors on the phenotypic variance of a specific trait.
After accounting for household or spousal effects, they
detected significant heritability for uric acid level; eGFR;
concentrations of serum creatinine, urea, serum potassium,
calcium, and sodium; 24-hour urinary albumin excretion;
and urinary albumin-creatinine ratio. These heritability es-
timates are consistent with results of prior studies.9
How should we interpret these results? The first major
finding is that individuals with a first-degree family
member affected by CKD have a 3-fold higher risk of CKD,
even in its early stages, and independently from other
known major risk factors for CKD. These findings are
consistent with what has been previously studied in the
context of kidney failure.10 CKD is often a silent disease,
and many individuals with an unrecognized kidney disease
may come to medical attention at its late or end stage. The
findings provided by Zhang et al2 therefore suggest that
the knowledge of familial aggregation of CKD could be an
additional useful metric for risk stratification for early
CKD. Currently, other known CKD risk factors explain only
50%-70% of CKD risk.11 The family history data may
therefore be incorporated into a risk calculator in the
assessment of future risk of kidney disease. A positive
family history of CKD may also provide additional
861

predictive information. For example, a recent study sug-
gested that familial forms of IgA nephropathy are associ-
ated with increased lifetime risk of kidney failure.12 Future
studies will need to assess whether a positive family history
of CKD also predicts complications of CKD such as pro-
gression or risk of cardiovascular disease. Moreover, it
would be interesting to determine whether the risk
imparted varies across different CKD categories such as
glomerular diseases or diabetic nephropathy.
The second major finding by Zhang et al is confirmation
of the high heritability of eGFR and other CKD-related
parameters. Recent large-scale GWAS have discovered
hundreds of common variants associated with CKD risk
and kidney function metrics,6,7 but these loci account for a
small fraction of the heritability estimated by population
studies such as the Zhang et al2 study. This “heritability
gap” could be related to a limitation of the GWAS
approach, which does not identify the contribution of rare
variants, structural variations,13 or mitochondrial DNA
variation.14 Other studies have also shown that estimation
of heritability improves significantly by calculation of
polygenic risk scores, which combine the weighted sums
of all risk variants across the genome.15 Polygenic risk
scores are being contemplated for clinical risk stratification
for some complex disorders such as obesity and coronary
heart disease and currently may be useful for some kidney
disorders such as membranous nephropathy, where the
genetic risk is well defined. These findings emphasize the
need for high-resolution sequencing of large CKD pop-
ulations, such as whole-genome sequencing, to compre-
hensively assess common and rare single-nucleotide and
structural variants associated with CKD. Finally, the “her-
itability gap” and the finding that spouses are also at higher
risk of CKD point to the need for better assessment of
environmental risk factors for kidney disease.
In conclusion, the article by Zhang et al2 provides
additional elements supporting both genetic and envi-
ronmental contributions to CKD. These results highlight
the need for further investigation of genetic factors that
contribute to CKD to move daily practice towards a
precision-medicine approach. The early identification of
individuals at risk for CKD and its potential contributors
could ultimately guide the clinician towards better
assessment of potentially actionable kidney disease risk
factors and improve patient management and outcomes.
Article Information
Authors’ Full Names and Academic Degrees: Lucrezia
Carlassara, MD, Francesca Zanoni, MD, Ali G. Gharavi, MD.
Authors’ Affiliations: Division of Nephrology, Department of
Medicine, Columbia University, New York, New York (LC, FZ,
AGG); and Division of Nephrology and Dialysis, Hospital of
Belluno, Belluno, Italy (LC).
Address for Correspondence: Ali G. Gharavi, MD, Division of
Nephrology, Department of Medicine, Columbia University, 1150
862
Carlassara et al
St Nicholas Ave, Russ Berrie Pavilion #412, New York, NY 10032.
Email: ag2239@cumc.columbia.edu
Support: None.
Financial Disclosure: The authors declare that they have no
relevant financial interests.
Peer Review: Received December 30, 2020 in response to an
invitation from the journal. Accepted December 31, 2020 after
editorial review by an Associate Editor and a Deputy Editor.
Publication Information: © 2021 by the National Kidney Founda-
tion, Inc. Published online February 11, 2021 with doi 10.1053/
j.ajkd.2020.12.010
References
1. Groopman EE, Marasa M, Cameron-Christie S, et al. Diag-
nostic utility of exome sequencing for kidney disease. N Engl J
Med. 2019;380(2):142-151.
2. Zhang J, Thio CH, Gansevoort R, Snieder H. Familial aggre-
gation of CKD and heritability of kidney biomarkers in the
general population: the Lifelines Cohort Study. Am J Kidney
Dis. 2021;77(6):869-878.
3. Kiryluk K, Li Y, Scolari F, et al. Discovery of new risk loci for
IgA nephropathy implicates genes involved in immunity
against intestinal pathogens. Nat Genet. 2014;46(11):1187-
1196.
4. Xie J, Liu L, Mladkova N, et al. The genetic architecture of
membranous nephropathy and its potential to improve non-
invasive diagnosis. Nat Commun. 2020;11(1):1600.
5. Debiec H, Dossier C, Letouz e L, et al. Transethnic, genome-
wide analysis reveals immune-related risk alleles and phenotypic
correlates in pediatric steroid-sensitive nephrotic syndrome.
J Am Soc Nephrol. 2018;29(7):2000-2013.
6. Wuttke M, Li Y, Sieber KB, et al. A catalog of genetic loci
associated with kidney function from analyses of a million in-
dividuals. Nat Genet. 2019;51(6):957-972.
7. Teumer A, Li Y, Ghasemi S, et al. Genome-wide association
meta-analyses and fine-mapping elucidate pathways influ-
encing albuminuria. Nat Commun. 2019;10(1):4130.
8. Dupre TV, Schnellmann RG, Miller GW. Using the exposome to
address gene–environment interactions in kidney disease. Nat
Rev Nephrol. 2020;16(11):621-622.
9. Govindaraju DR, Cupples AL, Kannel WB, et al. Genetics of
the Framingham Heart Study population. Adv Genet. 2008;62:
33-65.
10. Skrunes R, Svarstad E, Reisæter AV, Vikse BE. Familial clus-
tering of ESRD in the Norwegian population. Clin J Am Soc
Nephrol. 2014;9(10):1692-1700.
11. Echouffo-Tcheugui JB, Kengne AP. Risk models to predict
chronic kidney disease and its progression: a systematic re-
view. PLoS Med. 2012;9(11):e1001344.
12. Shi M, Yu S, Ouyang Y, et al. Increased lifetime risk of end-
stage renal disease in familial IgA nephropathy. Kidney Int Rep.
2020;6(1):91-100.
13. Verbitsky M, Westland R, Perez A, et al. The copy number
variation landscape of congenital anomalies of the kidney and
urinary tract. Nat Genet. 2019;51:117-127.
14. Govers LP, Toka HR, Hariri A, Walsh SB, Bockenhauer D.
Mitochondrial DNA mutations in renal disease: an overview.
Pediatr Nephrol. 2021;36(1):9-17.
15. Wray NR, Lin T, Austin J, et al. From basic science to clinical
application of polygenic risk scores: a primer. JAMA Psychiatry.
2021;78(1):101-109.
AJKD Vol 77 | Iss 6 | June 2021
Editorial
APOL1 and Preeclampsia: Intriguing Links, Uncertain
Causality, Troubling Implications
John R. Sedor, Leslie A. Bruggeman, and John F. O’Toole
P reeclampsia is a common, serious, multisystem dis-
order of pregnancy characterized by hypertension
onset after 20 weeks of gestation (or worsening of hy-
pertension in a patient with chronic hypertension) and
Related Article, p. 879
onset or worsening of proteinuria.1 Data from the National
Inpatient Sample showed approximately 177,000 of 3.8
million deliveries (4.7%) have a diagnosis of preeclamp-
sia/eclampsia.2 This disorder is of interest to nephrolo-
gists, since women with chronic kidney disease (CKD) are
at increased risk to develop preeclampsia and kidney
physicians are often consulted to help manage their blood
pressure and kidney manifestations. In addition, many but
not all studies have suggested women with preeclampsia
and normal kidney function have increased risk for later
CKD and kidney failure, although absolute risk for kidney
failure is under 1% within 20 years.3 Like CKD, pre-
eclampsia is a health disparity. For Black women, the rate
of preeclampsia is 60% higher than for White women,2
and the former have greater risk of maternal and fetal
complications in the peripartum period.2,4
Coding variants (called G1 and G2) in APOL1, the gene
encoding apolipoprotein L1, associate with large effect
sizes with a spectrum of nondiabetic kidney diseases in
Black individuals. Several candidate gene association
studies now have tested the hypothesis that the kidney risk
variants (KRV) of APOL1 have pleiotropic effects and also
contribute to the risk for preeclampsia (Table 1). Reidy
and her colleagues5 first reported that fetal but not
maternal APOL1 KRV associated with preeclampsia in Black
women from 2 study locations. A subsequent study in
South African women showed an association of the
maternal G1 but not the G2 allele with early-onset, but not
late-onset, preeclampsia.6 In Black women from a single
center in Ohio, infant APOL1 genotype was significantly
associated with preeclampsia.7 In this issue of AJKD, Hong
et al8 have published the latest association study of APOL1
KRV in Black mother-infant pairs with and without pre-
eclampsia. Using a nested case-control design and APOL1
alleles determined by direct genotyping, the study showed
that neither infant nor maternal APOL1 KRV genotype
associated with preeclampsia. When the cohort was strat-
ified by maternal country of origin, fetal and maternal
APOL1 KRV genotypes significantly associated with
increased risk of preeclampsia in African American but not
Haitian women. In addition, preeclampsia risk was higher
with maternal-fetal APOL1 genotype discordance, an effect
driven by the African American mother-infant pairs. An
AJKD Vol 77 | Iss 6 | June 2021
earlier analysis of publicly available data from this cohort
had failed to associate maternal APOL1 KRV genotypes with
preeclampsia risk,9 but this prior study had design limi-
tations that Hong et al have addressed.
Preeclampsia risk is complex to dissect and is not only
influenced by a mother’s demographic characteristics and
comorbid conditions but also impacted by both the
maternal and fetal genomes. In aggregate, these studies
hint that both maternal and fetal APOL1 KRV may increase
preeclampsia risk. However, beyond finding a significant
odds ratio in at least 1 analysis in each study, the studies
have significant inconsistencies in conclusions about the
contribution of maternal APOL1 KRV alleles to preeclampsia
risk and the genetic mode of inheritance for the APOL1 KRV
association with preeclampsia risk. Sample sizes available
to these investigative groups may have been limiting; the
effect size of APOL1 KRV for preeclampsia risk appears to be
similar to those described for other common traits rather
than the large effect association of APOL1 KRV with kidney
diseases. Given the potential clinical impact of a definitive
association of APOL1 KRV with preeclampsia, investigators
working in this area should consider collaboration for a
meta-analysis. In addition, we know from empirical data
that extended haplotypes appear to be important in
mediating some APOL1 effects,10 and a more comprehen-
sive analysis of APOL1 genetic variation and its association
with preeclampsia risk could be informative.
Hong et al have added several new twists to this
evolving story. Their data suggest that the association of
APOL1 KRV may be unique to African American Black in-
dividuals, at least when compared to risk in Haitians.
Reasons for this discrepancy are not clear. Both Haitians
and African Americans have predominantly West African
ancestry and ancestry principal component analysis did not
show significant differences in population structure by
country of origin. Country of origin can be a surrogate for
different cultural norms and traditions between groups,
which presumably result in unique gene-environment
interactions that increase preeclampsia risk only in Afri-
can American women with APOL1 KRV. In another new
finding, Hong et al have identified a discordance in
maternal-fetal APOL1 genotypes associated with pre-
eclampsia risk in African American but not Haitian
women. Intriguingly, an agnostic screen using bacterial
display peptides identified APOL1 autoantibodies that have
been reported to associate with preeclampsia.11 Several
studies have reported that other interactions between
maternal and fetal genotypes increase preeclampsia risk.12
The variance in maternal contribution to preeclampsia risk
between studies may reflect inadequate power to identify a
863
 Sedor et al

Table 1. Studies Reporting Association of APOL1 Kidney Risk Variants With Preeclampsia
Study Site, Design,
and Participants Sample Size, n Mode of Inheritance OR (95% CI)
NYC: case only5
Infant 121 cases Recessive 1.84 (1.11-2.93)
Mom 24 cases Recessive 0.72 (0.11-2.49)
Tennessee: case-control5
Infant 73 cases, 666 controls Recessive 1.92 (1.03-3.42)
Mom 93 cases, 293 controls Recessive 0.54 (0.21-1.17)
South Africa: case-control6,a
Infant Not applicable
Mom 119 cases, 98 controls Dominant 1.88 (1.02-3.45)
Ohio: case-control7
Infant 395 cases, 282 controls Dominant 1.41 (1.29-9.74)
Mom Not applicable
Boston: nested case-control8
Infant (African American) 134 cases, 133 controls Recessive 3.55 (1.29-9.74)
Infant (Haitian) 75 cases, 75 controls Recessive 0.57 (0.18-1.74)
Mom (African American) 134 cases, 133 controls Additive 1.53 (1.02-2.28)
Recessive 1.76 (1.01-3.06)
Mom (Haitian) 76 cases, 74 controls Additive 0.72 (0.42-1.23)
Recessive 0.30 (0.08-1.09)
Results shown for G1 allele only and for early-onset preeclampsia (clinical signs prior to 33 weeks ± days of gestation).
a

modest effect size. In addition, mother and fetus share half
their genomes, and variants associated with preeclampsia
risk only in the fetal genome would have half the effect
size in the mother and vice versa.13 None of the published
studies assessed if the mother’s APOL1 genotype conferred
preeclampsia risk independently of the fetal APOL1 geno-
type, although in African Americans both the maternal and
fetal APOL1 KRV appear to increase risk if an APOL1 ge-
notype discordance exists.
Experimental data support the hypothesis that APOL1
KRV mediate preeclampsia. APOL1 levels, APOL1-derived
peptides, and APOL1 autoantibodies have been linked to
preeclampsia. Transgenic mice that expressed APOL1 using
the nephrin (Nphs1) promoter developed a pregnancy-
associated phenotype characterized by hypertension,
proteinuria, and seizures, which was more severe in
transgenic animals with an APOL1 KRV transgene compared
to mice transgenic for reference APOL1.14 However this
phenotype has not been described in other APOL1 trans-
genic mouse models driven by the endogenous promoter.
Since nephrin is expressed in the placenta, the Nphs1:APOL1
mouse would recapitulate human placental APOL1
expression, which may have disrupted placentation from
its cytotoxic effects. However, contrary to expectation,
APOL1 KRV failed to associate with preeclamptic histo-
pathologic placental phenotypes.5,14
In closing, we would like to revisit the surprising
discordance between African Americans and Haitians in
APOL1 KRV association with preeclampsia risk. Known risk
factors for preeclampsia were similar between groups, and
a subset of participants in this study had comparable ge-
netic ancestry. Society would label all these woman as
Black, ostensibly eliminating race as a predictor of
nongenetic factors regulating preeclampsia risk. However,
one difference between these groups is the majority of
Haitian woman studied by Hong et al have been in the
United States less than 10 years, consonant with immi-
gration patterns for all Haitians.15 Although Haitian
Americans in 1 study have higher perceived stress than
African Americans,16 these Haitian women have had a
shorter exposure to the policies and social conditions, the
structural racism, and the environmental contexts that
through allostatic load and epigenetic pathways can man-
ifest in biological responses and disease.17
Preeclampsia is more prevalent in African Americans
and their case fatality is 3 times higher than rates in White
women.18 Genetics alone is unlikely to explain disparities
in preeclampsia risk, but identification of genetic de-
terminants has been hampered by the limited numbers of
diverse patients in populations used for genome-wide as-
sociation studies. To address this stark health inequality,
we as a community, along with our Ob-Gyn colleagues,
need to determine if APOL1 KRV unambiguously associates
with preeclampsia risk in African American women and
assess the clinical utility of this observation. Perhaps more
importantly, we also need to advocate for strategies that
address the unconscionable, root social factors responsible
for biases in care of Black women.
Article Information
Authors’ Full Names and Academic Degrees: John R. Sedor, MD,
Leslie A. Bruggeman, PhD, and John F. O’Toole, MD.
Authors’ Affiliations: Glickman Urology and Kidney Institute (JRS,
LAB, JFO) and Lerner Research Institute (JRS, LAB, JFO),

864 AJKD Vol 77 | Iss 6 | June 2021

Sedor et al
Cleveland Clinic; and Departments of Molecular Medicine (JRS,
LAB, JFO) and Physiology and Biophysics (JRS), Case Western
Reserve University, Cleveland, OH.
Address for Correspondence: John R. Sedor, MD, Glickman
Urology and Kidney Institute, 9500 Euclid Ave, Q7, Cleveland, OH
44195. Email: sedorj@ccf.org
Support: The authors are supported by NIH grants DK097836,
DK108329, DK127638, and AI135434.
Financial Disclosure: The authors developed the mouse model
mentioned in this editorial; they and Case Western Reserve
University have received licensing fees for this model.
Peer Review: Received January 27, 2021 in response to an
invitation from the journal. Accepted January 31, 2021 after
editorial review by an Associate Editor and a Deputy Editor.
Publication Information: © 2021 by the National Kidney Founda-
tion, Inc. Published online April 17, 2021 with doi 10.1053/
j.ajkd.2021.01.013
References
1. Gestational hypertension and preeclampsia: ACOG Practice
Bulletin No. 222. Obstet Gynecol. 2020;135(6):e237-260.
2. Fingar KR, Mabry-Hernandez I, Ngo-Metzger Q, Wolff T,
Steiner CA, Elixhauser A. Delivery Hospitalizations Involving
Preeclampsia and Eclampsia, 2005-2014: Statistical Brief
#222, Healthcare Cost and Utilization Project (HCUP) Sta-
tistical Briefs. Rockville (MD): Agency for Healthcare Research
and Quality (US); 2006. PMID: 28722848.
3. Ferreira RC, Fragoso MBT, Dos Santos Tenorio MC, et al. Pre-
eclampsia is associated with later kidney chronic disease and
end-stage renal disease: Systematic review and meta-analysis of
observational studies. Pregnancy Hypertens. 2020;22:71-85.
4. Shahul S, Tung A, Minhaj M, et al. Racial disparities in comor-
bidities, complications, and maternal and fetal outcomes in
women with preeclampsia/eclampsia. Hypertens Pregnancy.
2015;34(4):506-515.
5. Reidy KJ, Hjorten RC, Kaskel FJ, Winkler CA, Kopp JB. Fetal –
not maternal – APOL1 genotype associated with risk for pre-
eclampsia in those with African ancestry. Am J Hum Genet.
2018;103(3):367-376.
6. Thakoordeen-Reddy S, Winkler C, Moodley J, et al. Maternal
variants within the apolipoprotein L1 gene are associated with
preeclampsia in a South African cohort of African ancestry. Eur
J Obstet Gynecol Reprod Biol. 2020;246:129-133.
AJKD Vol 77 | Iss 6 | June 2021
7. Miller AK, Azhibekov T, O’Toole JF, et al. Association of pre-
eclampsia with infant APOL1 genotype in African Americans.
BMC Med Genet. 2020;21(1):110.
8. Hong X, Rosenberg AZ, Zhang B, et al. Joint associations
of maternal-fetal APOL1 genotypes and maternal country of
origin with preeclampsia risk. Am J Kidney Dis. 2021;77(6):
879-888.
9. Robertson CC, Gillies CE, Putler RKB, et al. An investigation of
APOL1 risk genotypes and preterm birth in African American
population cohorts. Nephrol Dial Transplant. 2017;32(12):
2051-2058.
10. Lannon H, Shah SS, Dias L, et al. Apolipoprotein L1 (APOL1)
risk variant toxicity depends on the haplotype background.
Kidney Int. 2019;96(6):1303-1307.
11. Elliott SE, Parchim NF, Liu C, et al. Characterization of antibody
specificities associated with preeclampsia. Hypertension.
2014;63(5):1086-1093.
12. Del Gobbo GF, Konwar C, Robinson WP. The significance of
the placental genome and methylome in fetal and maternal
health. Hum Genet. 2020;139(9):1183-1196.
13. Steinthorsdottir V, McGinnis R, Williams NO, et al. Genetic
predisposition to hypertension is associated with preeclampsia
in European and Central Asian women. Nat Commun.
2020;11(1):5976.
14. Bruggeman LA, Wu Z, Luo L, et al. APOL1-G0 or
APOL1-G2 transgenic models develop preeclampsia but
not kidney disease. J Am Soc Nephrol. 2016;27(12):
3600-3610.
15. Olsen-Medina K, Batalova J. Haitian Immigrants in the United
States, Migration Information Source. Washington, DC:
Migration Policy Institute; 2020. Accessed January 27, 2021.
https://www.migrationpolicy.org/article/haitian-immigrants-united-
states-2018
16. Fatma HG, Joan VA, Ajabshir S, Gustavo ZG, Exebio J, Dixon Z.
Perceived stress and self-rated health of Haitian and African
Americans with and without Type 2 diabetes. J Res Med Sci.
2013;18(3):198-204.
17. Boulware LE, Mohottige D. The seen and the unseen: race
and social inequities affecting kidney care. Clin J Am Soc
Nephrol. 2021;CJN.12630820. https://doi.org/10.2215/CJN.
12630820.
18. US Preventive Services Task Force; Bibbins-Domingo K,
Grossman DC, Curry SJ, et al. Screening for Preeclampsia: US
Preventive Services Task Force Recommendation Statement.
JAMA. 2017;317(16):1661-1667.
865

One Step Closer to Developing a National Dialysis
Registry in China
Samaya Javed Anumudu, Vishnu Parvathareddy, and Kevin F. Erickson
n this issue of AJKD, Yang et al1 report findings from an
Ieffort to expand a national surveillance system for kidney
care in China. To fully appreciate their findings, it is
important to first review changes and challenges in dialysis
Related Article, p. 889
care delivery in China, and how administrative data may
help guide future dialysis policy. The role of “big data” in
shaping US dialysis policy is particularly instructive.2
The burden of chronic kidney disease (CKD) is
increasing in the developing world, largely owing to
increasing incidences of obesity, diabetes, and hyperten-
sion.3 As life expectancies increase and populations age,
the number of patients with CKD who progress to kidney
failure and require kidney replacement therapy is also
rising. Meanwhile, economic growth in many transitional
countries is generating the resources necessary to finance
dialysis care.4 These trends have been occurring for several
decades in China,5 where an estimated 120 million people
have CKD, representing approximately 11% of the adult
population.6 Between 1999 and 2012, the prevalence of
kidney failure requiring dialysis in China increased 7-
fold.7
Health policy initiatives by the Chinese government
have facilitated increases in dialysis by expanding access to
care. In 2003, the creation of the New Cooperative Medical
Scheme (NCMS) expanded access to health insurance for
patients living in rural areas. Reforms enacted in 2009-
2011 expanded 2 major health insurance schemes—the
urban employee-based basic medical insurance (UEBMI)
and urban resident-based basic medical insurance
(URBMI)—leading to near-universal coverage among ur-
ban residents. Because the Chinese government defines
CKD as a major chronic disease, all 3 major medical in-
surance schemes cover some dialysis therapy. More
recently, the 2015 Critical Illness Insurance Program made
dialysis more affordable for some patients with cata-
strophic medical expenses.8
In many ways, rapid growth in the incidence of kidney
failure and dialysis in China resembles the early experience
of the US end-stage renal disease (ESRD) program. In the
decade following enactment of a federal ESRD program in
1972, annual growth in the US prevalence of kidney
failure was 22%, comparable to 15% in China in 1999-
2012.9 The rapid, and unexpected, rise in dialysis in the
1970s and 1980s in the United States presented challenges
for policymakers. Namely, because dialysis therapy is
expensive, efforts by US policymakers to ensure broad
access to high-quality dialysis care at an affordable cost
866
Editorial
began to conflict with one another. Balancing these
competing goals soon became a central theme of US
dialysis policy.
Recognizing the unique challenges associated with na-
tional policy for patients with kidney failure, in 1986 the
US Congress called for the development of a “National
ESRD Registry” that would facilitate efforts to study eco-
nomic and epidemiologic developments in the care of
patients with kidney failure. Two years later, development
of the US Renal Data System (USRDS) registry made
Medicare’s comprehensive set of administrative dialysis
data accessible to the broad research community. Findings
from the USRDS registry have been instrumental in efforts
to monitor and improve US dialysis care, leading to
numerous successful clinical and policy initiatives, such as
the Fistula First initiative, ESRD Network quality initiatives,
and expansion of “bundled” payments for dialysis care.2 In
the setting of rapidly growing dialysis care, policymakers
in China may soon be confronted with a similar set of
challenges.
In 2014, the China Kidney Disease Network (CK-NET)
initiative began working with hospitals and health insurers
to integrate administrative data from multiple sources at a
national level in order to create a comprehensive CKD and
dialysis surveillance system for China, similar to the
USRDS. The first CK-NET annual data report, published in
2017, demonstrated how the extraction of cross-sectional
information from electronic discharge summaries at
select hospitals and the application of big data analytics
could successfully identify benchmark characteristics of
hospitalized patients with CKD.10 In 2019, a second data
report expanded the registry by including information
about patients receiving dialysis from public and com-
mercial claims.11 In what the authors refer to as the CK-
NET Surveillance System, Yang et al1 now report on their
addition of 4 years of longitudinal follow-up from dialysis
claims at urban hospitals across mainland China.
The authors used a 2-stage sampling design to obtain
information about hospitalizations between January 2013
and December 2017. Administrative claims came from the
UEBMI and URBMI programs. For each hospitalization,
information about admission, discharge, prescription
drugs administered, medical services rendered, and di-
agnoses were extracted from claims. Claims for dialysis
services, combined with International Classification of
Diseases, Tenth Revision diagnosis codes, were used to
identify patients receiving dialysis for treatment of kidney
failure, along with their dialysis modality. These data were
then combined with demographic information from a
2010 census and previously observed hospital rates among
AJKD Vol 77 | Iss 6 | June 2021
Anumudu et al
the insured population to generate crude and age- and sex-
standardized estimates of disease prevalence. The authors
also fit estimated prevalence curves to predict future rates
of kidney failure.
The study reports continued rapid growth in the prev-
alence of kidney failure in China in 2013-2017. In 2013,
there were an estimated 346,178 patients receiving dialysis
nationwide, or 255 patients per million population (pmp).
By the end of 2017, this had increased to 581,273 pa-
tients, or 419 patients pmp. Based on this trend, the au-
thors predict that dialysis prevalence would rise to 630
pmp by 2025. By contrast, estimated prevalences of dial-
ysis in 2010 were 1,839 pmp in North America, 719 pmp
in Europe, and >2,000 pmp in Taiwan and Japan.12 The
crude prevalence of dialysis in China was the highest
among people aged 65-74, underscoring the important
role of an aging population on the prevalence of kidney
failure. Approximately 92% of patients with kidney failure
received hemodialysis, and nearly all of the increase in
dialysis during the study period was in the use of hemo-
dialysis. In contrast, peritoneal dialysis (PD) did not in-
crease markedly over the observation period.
These findings highlight the potential for a national
dialysis registry based on administrative data to inform
clinical and policy interventions in China, similar to the
ways in which USRDS data has informed US policy.
Referring to the observed growth in the use of dialysis, the
authors discuss how findings from the CK-NET Surveillance
System emphasize the national importance of finding ways
to prevent major noncommunicable chronic diseases in
order to “reduce the escalating burden of kidney failure.”1
Additionally, only about 8% of patients with kidney
failure who were included in the CK-NET Surveillance
System study received PD. Unlike Hong Kong, where a
“PD first” policy has led to much higher use of this mo-
dality,13 urban areas in mainland China appear to be
expanding dialysis care delivery with a focus on hemodi-
alysis. This is similar to what occurred in the United States,
and is a development that the United States has found very
difficult to reverse. Yet, for many patients, PD may be
preferable and less costly. Findings from Yang et al high-
light an opportunity for policymakers to encourage PD use
while China’s dialysis program is still in a phase of rapid
expansion and, perhaps, before investments in hemodial-
ysis infrastructure limit future efforts to encourage
alternatives.
Key limitations from the study point to additional op-
portunities for a national dialysis registry to help shape the
future of dialysis care delivery. A major limitation of the
study derives from the data sources. The registry does not
include information about rural residents covered by the
NCMS. Yet, many important clinical and policy questions
relate to dialysis care among rural residents. For example,
access to care may be more limited in rural areas owing to
an absence of dialysis providers and among rural residents
who have recently migrated to cities but who are not
eligible for the urban programs. Likewise, growth in
AJKD Vol 77 | Iss 6 | June 2021
community dialysis centers cannot be monitored, as they
are not included in the current registry. Important questions
about the quality of twice-weekly dialysis, which is com-
mon among patients living in rural areas,14 will also be
difficult to address without information outside of the urban
hospitals. A recent proposal to collect comprehensive data
on patients with CKD in the Yinzhou district exemplifies the
ways in which the data registry could be expanded.15
In the United States, prior to the creation of the USRDS,
only a small number of researchers and institutions had both
access to Medicare’s administrative dialysis data and the
technical capacity to analyze them. These researchers
showed how Medicare data could be used to examine the
quality and cost of dialysis care, but the number and scope
of these studies was limited. Calls for a national US dialysis
registry by lawmakers and the research community stem-
med from an interest in making Medicare’s data more
widely accessible to researchers and policymakers. Many
of the clinical and policy advancements in US dialysis
care resulting from analyses of USRDS data came from
collaborative efforts between independent researchers and
USRDS coordinating centers. Without efforts by USRDS
coordinating centers to clean administrative data and make
them accessible to researchers, many important advance-
ments may not have occurred. If the true potential of the
CK-NET initiative is to be realized, similar efforts will
need to make data widely accessible to a community of
researchers.
In summary, Yang and colleagues observed rapid in-
creases in the number of patients receiving dialysis in
China that are predicted to continue into the near future.
Their analysis demonstrates the potential for aggregated
longitudinal administrative claims data to provide valuable,
policy-relevant insights into dialysis care at a national level
in China. In the United States, national administrative data
have been instrumental in guiding policy initiatives and
addressing challenges related to cost, quality, and access to
dialysis care. As China’s economy and dialysis population
grow, policymakers are likely to face similar challenges.
Future efforts to expand the scope of the CK-NET Sur-
veillance System project and to make the dialysis registry
accessible to a broad community of researchers could help
Chinese policymakers to promote access to affordable,
high-quality dialysis care.
Article Information
Authors’ Full Names and Academic Degrees: Samaya Javed
Anumudu, MD, FASN, Vishnu Parvathareddy, MD, CPE, and Kevin
F. Erickson, MD, MS.
Authors’ Affiliations: Selzman Institute for Kidney Health and
Section of Nephrology, Baylor College of Medicine, Houston, TX
(SJA, VP, KFE), and Baker Institute for Public Policy, Rice
University, Houston, TX (KFE).
Address for Correspondence: Kevin Erickson, MD, MS, 2002
Holcombe Blvd, Mail Code 152, Houston, TX 77030. Email: kevin.
erickson@bcm.edu
Support: None.
867

Financial Disclosure: The authors declare that they have no
relevant financial interests.
Peer Review: Received January 14, 2021 in response to an
invitation from the journal. Accepted January 18, 2021 after
editorial review by an Associate Editor and a Deputy Editor.
Publication Information: © 2021 by the National Kidney Founda-
tion, Inc. Published online April 1, 2021 with doi 10.1053/
j.ajkd.2021.01.012
References
1. Yang C, Yang Z, Wang J, et al. on behalf of China Kidney
Disease Network (CK-NET) Work Group. Estimation of preva-
lence of kidney disease treated with dialysis in China: A study
of insurance claims data. Am J Kidney Dis. 2021;77(6):
889-897.
2. Erickson KF, Qureshi S, Winkelmayer WC. The role of big data
in the development and evaluation of US dialysis care. Am J
Kidney Dis. 2018;72:560-568.
3. Foreman KJ, Marquez N, Dolgert A, et al. Forecasting life ex-
pectancy, years of life lost, and all-cause and cause-specific
mortality for 250 causes of death: reference and alternative
scenarios for 2016-40 for 195 countries and territories. Lancet.
2018;392:2052-2090.
4. Jha V, Garcia-Garcia G, Iseki K, et al. Chronic kidney disease:
global dimension and perspectives. Lancet. 2013;382:260-272.
5. Liu ZH. Nephrology in China. Nat Rev Nephrol. 2013;9:523-
528.
6. Zhang L, Wang F, Wang L, et al. Prevalence of chronic kidney
disease in China: a cross-sectional survey. Lancet. 2012;379:
815-822.
868
Anumudu et al
7. Zhang L, Zuo L. Current burden of end-stage kidney
disease and its future trend in China. Clin Nephrol. 2016;86:
27-28.
8. Fang P, Pan Z, Zhang X, Bai X, Gong Y, Yin X. The effect of
critical illness insurance in China. Medicine. 2018;97:
e11362.
9. Eggers PW, Connerton R, McMullan M. The Medicare expe-
rience with end-stage renal disease: trends in incidence,
prevalence, and survival. Health Care Financ Rev. 1984;5:69-
88.
10. Saran R, Steffick D, Bragg-Gresham J. The China Kidney
Disease Network (CK-NET): “Big Data-Big Dreams.”. Am J
Kidney Dis. 2017;69:713-716.
11. Wang F, Yang C, Long J, et al. Executive summary for the 2015
Annual Data Report of the China Kidney Disease Network (CK-
NET). Kidney Int. 2019;95:501-505.
12. Liyanage T, Ninomiya T, Jha V, et al. Worldwide access to
treatment for end-stage kidney disease: a systematic review.
Lancet. 2015;385:1975-1982.
13. Choy AS, Li PK. Sustainability of the peritoneal dialysis-first
policy in Hong Kong. Blood Purif. 2015;40:320-325.
14. Bieber B, Qian J, Anand S, et al. Two-times weekly hemodial-
ysis in China: frequency, associated patient and treatment
characteristics and Quality of Life in the China Dialysis Out-
comes and Practice Patterns study. Nephrol Dial Transplant.
2014;29:1770-1777.
15. Wang J, Bao B, Shen P, et al. Using electronic health re-
cord data to establish a chronic kidney disease surveillance
system in China: protocol for the China Kidney Disease
Network (CK-NET)-Yinzhou Study. BMJ Open. 2019;9:
e030102.
AJKD Vol 77 | Iss 6 | June 2021
Original Investigation

Familial Aggregation of CKD and Heritability of Kidney

Biomarkers in the General Population: The Lifelines
Cohort Study
Jia Zhang,* Chris H.L. Thio,* Ron T. Gansevoort, and Harold Snieder

Rationale & Objective: Chronic kidney disease
(CKD) has a heritable component. We aimed to
quantify familial aggregation of CKD in the gen-
eral population and assess the extent to which
kidney traits could be explained by genetic and
environmental factors.
Study Design: Cross-sectional 3-generation
family study.
Setting & Participants: Data were collected at
entry into the Lifelines Cohort Study from a
sample of the general population of the northern
Netherlands, composed predominantly of in-
dividuals of European ancestry.
Exposure: Family history of CKD.
Outcomes: The primary outcome was CKD,
defined as estimated glomerular filtration rate
(eGFR) <60 mL/min/1.73 m2, where GFR was
estimated using the CKD Epidemiology Collab-
oration creatinine equation. Among a subsample
for which urinary albumin concentration was
available (n = 59,943), urinary albumin excretion
was expressed as the rate of urinary albumin
excretion (UAE) per 24 hours or urinary albumin-
creatinine ratio (UACR).
Analytical Approach: Familial aggregation of
CKD was assessed by calculating the recurrence
risk ratio (RRR), using adapted Cox proportional
hazards models. Heritability of continuous kidney-
related traits was estimated using linear mixed
models and defined as the ratio of the additive Complete author and article
genetic variance to total phenotypic variance. information provided before
references.
All models were adjusted for age, sex, and
known risk factors for kidney disease. Correspondence to
H. Snieder (h.snieder@
Results: Among 155,911 participants with umcg.nl)
available eGFR data, the prevalence of CKD *J.Z. and C.H.L.T.
was 1.19% (1,862 cases per 155,911). The risk contributed equally to this
of CKD in those with an affected first-degree work.
relative was 3 times higher than the risk in the Am J Kidney Dis.
total sample (RRR, 3.04 [95% CI, 2.26-4.09). 77(6):869-878. Published
In those with an affected spouse, risk of CKD online December 22, 2020.
was also higher (RRR, 1.56 [95% CI, 1.20- doi: 10.1053/
1.96]), indicative of shared environmental j.ajkd.2020.11.012
factors and/or assortative mating. Heritability
© 2020 The Authors.
estimates of eGFR, UAE, and UACR were Published by Elsevier Inc.
44%, 20%, and 18%, respectively. For serum on behalf of the National
urea, creatinine, and uric acid, estimates were Kidney Foundation, Inc. This
31%, 37%, and 48%, respectively, whereas is an open access article
under the CC BY license
estimates for serum electrolytes ranged from
(http://creativecommons.
22% to 28%.
org/licenses/by/4.0/).
Limitations: Use of estimated rather than
measured GFR. UAE data only available in a
subsample.
Conclusions: In this large population-based
family study, a positive family history was
strongly associated with increased risk of CKD.
We observed moderate to high heritability of
kidney traits and related biomarkers. These
results indicate an important role of genetic
factors in CKD risk.

C hronic kidney disease (CKD) is recognized as a global

public health problem1 with prevalence ranging be-
tween 3.3% and 17.3% in adult European populations.2 CKD
is defined by decreased estimated glomerular filtration rate
(eGFR) and/or increased albuminuria, and is associated with
Editorial, p. 861
an increased risk of cardiovascular disease (CVD) and pro-
gression to kidney failure.3-6
Established risk factors for CKD, such as hypertension
and diabetes, explain 50%-70% of cases and are the main
targets of current risk prediction models for CKD.7 Familial
clustering of CKD and kidney-related markers suggest that
genetic factors or shared environmental factors are also
important in the pathogenesis of this disease.8-12 Indirect
support for a genetic component to CKD comes from
recent genome-wide association studies (GWAS) of
eGFR13 and albuminuria,14 in which a large number of
genetic loci have been reported to associate with the CKD-
defining traits eGFR and albuminuria.
Several knowledge gaps exist with regard to the genetic
contribution to CKD susceptibility in the general population.
For example, most familial aggregation studies of CKD focused
on its later stages (ie, kidney failure) using medical records
and registry data.8,9,12,15 Focusing on early-stage CKD rather
than kidney failure may have added value for risk stratification.
In addition to assessing familial aggregation, one can
estimate the heritability of disease traits. Heritability quan-
tifies the relative importance of genetic and environmental
factors in explaining the distribution of a trait or disease
within a population.16 Both kidney function and related
blood biomarkers have been shown to be heritable.11,17,18
Related blood biomarkers are, for example, alternative
kidney function markers of creatinine (eg, serum urea, uric
acid) or biomarkers for which homeostasis is regulated
partly through kidney function (eg, serum electrolytes). To
date, the heritability of kidney traits has been estimated in

AJKD Vol 77 | Iss 6 | June 2021 869


twin studies,17,19-22 in family studies with relatively small
sample size,11,18 and in studies in isolated founder or dis-
ease populations.23,24 Due to the differences in family
design (ie, twins vs pedigrees), relatively small sample sizes,
and specific populations, random sampling error may have
occurred,16 thus, generalizability of those studies may be
uncertain. Therefore, heritability estimates from a large,
representative sample of the general population are needed.
This study aimed to quantify the familial aggregation of
CKD and to obtain heritability estimates of kidney traits
and related biomarkers in the general population.
Methods
Study Design and Population
In this cross-sectional family study, we used baseline data
from the Lifelines Cohort Study and Biobank, a multi-
disciplinary prospective population-based cohort study of
the northern Netherlands with a unique 3-generation
design that included 167,548 participants, of
whom >95% were of European ancestry. It uses a broad
range of investigative procedures in assessing the socio-
demographic, biomedical, physical, behavioral, and psy-
chological factors that contribute to the health and disease
of the general population, with a special focus on mul-
timorbidity and complex genetics. The overall design and
rationale of this study have been described in detail
elsewhere.25,26 Recruitment of participants and, subse-
quently, their families is detailed in Item S1. Briefly,
kinship was derived from questionnaires and validated in
those with genetic data (available in ~50,000 partici-
pants).27 After signing informed consent, participants
received a baseline questionnaire and an invitation to a
health assessment at one of the Lifelines research sites.
The Lifelines Cohort Study was conducted according to
principles of the Declaration of Helsinki and in accor-
dance with the research code of University Medical Center
Groningen and was approved by its medical ethical
committee. All participants gave written informed
consent.
Measurements
Kidney Outcomes
Participants (≥8 years of age) were invited to 1 of 12 local
research sites in northern Netherlands for their physical
examinations. The baseline assessment consisted of 2 visits.
During the first visit, a trained research nurse performed
physical examinations and provided containers and oral
and written instruction for collection of a 24-hour urine
sample (for those ≥18 years of age). Two weeks later,
during the second visit, a fasting blood sample (for those
≥8 years of age) and the 24-hour urine sample were
collected.
Measurements of serum creatinine (Scr) were per-
formed by using an isotope-diluted mass spectrometry–
traceable enzymatic method on a modular analyzer, using
870
Zhang et al
reagents and calibrators from Roche Diagnostics, with
intra- and interassay coefficients of variation of 0.9% and
2.9%, respectively. Urinary albumin concentration was
measured by nephelometry, with a lower threshold of
detection of 2.3 mg/L and intra- and interassay coefficients
of variation of 2.2% and 2.6%, respectively (Dade Behring
Diagnostic), and multiplied by urine volume to obtain a
value of urinary albumin excretion (UAE) in milligrams
per 24 hours. Urinary albumin-creatinine ratio (UACR)
was determined from spot urine (for those ≥8 years of
age). After addition of a constant of 1 to handle zero
values, UAE and UACR were natural log–transformed to
approximate a normal distribution prior to statistical
analyses.
CKD was defined in the primary analysis as
eGFR <60 mL/min/1.73 m2 (CKDScr) in the complete
sample. In secondary analyses in a subsample where uri-
nary albumin was available (in ~60,000 participants), we
applied 2 additional definitions of CKD that incorporated
albuminuria according to KDIGO (Kidney Disease:
Improving Global Outcomes) guidelines.28,29 The first
additional definition took into account UAE (CKDScr+UAE:
eGFR <60 mL/min/1.73 m2 and/or UAE ≥30 mg/d); and
the second definition took into account UACR rather than
UAE (CKDScr+UACR: eGFR <60 mL/min/1.73 m2 and/or
UACR ≥30 mg/g). In addition, in this subsample, we
repeated our analyses using moderately increased albu-
minuria alone (CKDUAE: UAE ≥30 mg/d; CKDUACR:
UACR ≥30 mg/g).
We calculated eGFR using the 2012 CKD Epidemiology
Collaboration (CKD-EPI) creatinine equation30 for adults
and the bedside Schwartz equation for juveniles (age <18
years).31
Kidney-Related Biomarkers
Kidney-related biomarkers were determined using stan-
dard methods for serum uric acid, an enzymatic colori-
metric assay; for serum urea an ultraviolet kinetic assay on
a Roche Modular analyzer; and for serum electrolytes
(calcium, potassium, and sodium) a Roche Modular P
chemistry analyzer (Roche).
Covariates
Known CKD risk factors (body mass index [BMI], hyper-
tension, diabetes mellitus, hypercholesterolemia, smoking
status, and history of CVD) were included as covariates
(details in Item S1).
Statistical Analysis
Baseline Characteristics
Baseline characteristics were examined for the total pop-
ulation and separately for adults and in juveniles.
Recurrence Risk Ratio
Mean eGFR and prevalence of CKD were calculated for
the general population and for individuals with affected
AJKD Vol 77 | Iss 6 | June 2021
Zhang et al
first-degree relatives. Recurrence risk ratio (RRR) of CKD was
defined as the adjusted prevalence ratio between first-degree
relatives of an affected individual and the general population
according to the Risch definition.32 We used Cox proportional
hazards models, adapted according to Breslow,33 to estimate
prevalence ratios in a cross-sectional study by applying an
equal follow-up time for all participants. This method pro-
duces consistent estimates for prevalence ratios close to true
limits.34,35 A marginal proportional hazards model was used
to handle correlated observations due to familial clustering.
This model estimates the mean population hazard function
and uses a robust sandwich method to estimate confidence
intervals (CI).36,37 This approach has been applied and vali-
dated in previous studies of other diseases.38-40
For each CKD definition (CKDScr, CKDUAE, CKDUACR,
CKDScr+UAE, CKDScr+UACR), we calculated RRR for in-
dividuals with an affected first-degree relative of any
kinship. Models based on type of kinship and sex of
affected relatives (eg, parents, siblings, offspring, and their
sex) were explored. Additionally, we estimated RRR for
individuals with an affected spouse to assess effects of
shared environment and/or assortative mating. We
adjusted for age, age squared, sex, and CKD risk factors.
Heritability Estimates
For all continuous traits, we estimated narrow-sense
heritability, defined as the ratio of the additive genetic
variance, which reflects transmissible resemblance be-
tween relatives, to the total phenotypic variance. We
used the residual maximum likelihood-based variance
decomposition method41 implemented in ASReml soft-
ware (VSNi).42 With this method, the overall pheno-
typic variance is decomposed into genetic and
environmental components. We also included household
or spousal effects in the model to estimate the influence
of shared environment by using family identifier or
spouse identifier as a proxy. This allowed us to distin-
guish between shared genes and shared environment as
potential sources of familial resemblance. In addition,
we calculated spousal correlations for all continuous
traits. P values for heritability were derived from
likelihood ratio tests, which compare the likelihood
of a heritability model to that of a model in which
heritability is constrained to zero. Age, age squared, sex,
and CKD risk factors were included as covariates,
regardless of their statistical significance. We report the
percentage of variance explained by these covariates
(PVC).
To ascertain bias due to missingness, we examined the
consistency of RRR estimates for CKDScr, as well as heri-
tability estimates for eGFR, Scr, and serum potassium
within the full sample and the subsample.
All analyses were performed using ASReml 4.1 soft-
ware42 and R version 3.3.1 software.43 A two-sided sig-
nificance level for analyses was set at α = 0.05.
AJKD Vol 77 | Iss 6 | June 2021
Results
Baseline Characteristics
From the 167,548 Lifelines participants at baseline, we
included 155,911 participants (within 29,703 family
clusters; 39,836 singletons) with available eGFR data
during the baseline visit (Fig 1). In a subsample of
59,938 (including 743 juveniles), both eGFR and UACR
were available, whereas in a subsample of 59,145 par-
ticipants (only adults), both eGFR and UAE were avail-
able (Fig S1-S2). Table S1 provides more details on
family structure.
In the full sample (N = 155,911; 58.1% female
participants; mean age, 43.1 ± 14.7 [SD] years) par-
ticipants had a mean eGFR of 97.2 ± 15.7 mL/min/
1.73 m2. In the subsample with albuminuria measure-
ments, a median UAE of 3.86 (interquartile range
[IQR], 2.33-6.92) mg/d and a median UACR of 2.72
(IQR, 1.58-7.33) mg/g were observed (Table 1). Male
participants had a slightly less favorable kidney risk
profile than female participants (ie, higher prevalence
of smoking, hypertension, diabetes, and high choles-
terol) but similar distributions in CKD risk and kidney
markers (ie, eGFR and UAE) (Table S2). Participants
with a family history of CKDScr had a less favorable
kidney profile than those without a family history
(Table S3). Distributions of age, sex, and covariates in
the subsample were similar to those in the full sample
(Table S4).
We identified 1,862 CKDScr cases, translating to a crude
prevalence of 1.19% (Table 1). In 1,725 of 29,703 fam-
ilies (5.8% of family clusters), there was at least 1 CKDScr
case. A total of 2,211 individuals had at least 1 first-degree
relative with CKDScr: 1,680 with at least 1 affected parent,
56 with at least 1 affected offspring, and 499 with at least
1 affected sibling.
There was a dramatically greater prevalence of CKDScr in
those older than 60. Mean eGFR was lower at a higher age,
and age-specific mean values of eGFR were lower among
individuals with affected first-degree relatives than among
the general population (Fig 2A). Accordingly, the age-
specific prevalence rates were significantly higher in
those with a first-degree relative affected with CKDScr (Fig
2B). In the subsample, the crude prevalence rates of
CKDScr+UAE and CKDScr+UACR were 5.5% and 6.8%,
respectively (Table 1).
Recurrence Risk Ratio for CKD Scr
Estimates of RRR for CKDScr are shown in Fig 3. Generally,
having a first-degree relative affected with CKDScr was
associated with an RRR of 3.04 (95% CI, 2.26-4.09). Fa-
milial recurrence showed no clear pattern by kinship type
or sex of the affected family member. Spouses of affected
individuals had higher risk than the general population
(RRR, 1.56 [95% CI, 1.20-2.00]).
871
 Zhang et al

Lifelines participants for the

baseline visit=167,548
No serum creatinine data for 11,612
No blood draw (n=11,523)
No blood assessment in children aged<8 y (n=5,069)
Did not agree to participate in blood draw in children 8-17 y (n=1,520)
Did not agree to participate in blood draw in adults >=18 y (n=4,934)
With blood draw but without serum creatinine data (n=89)
No serum creatinine data in children 8-17 y (n=32)
No serum creatinine data in adults >=18 y (n=57)
155,936 with serum
creatinine data
Calculating eGFR according to
appropriate formula
155,936 with eGFR data
25 participants with extreme values of eGFR (21 extremely
low <mean – 5SD; 4 extremely high > mean + 5SD)
155,911 for final analysis

Figure 1. Flow chart of eGFR analysis. There were 21 participants with extremely low eGFR values (1 <−5 SD from the mean) who were
considered CKD patients and retained in analyses of CKD. Abbreviations: eGFR, estimated glomerular filtration rate; SD, standard deviation.

Recurrence Risk Ratio for CKD Scr+UAE and CKD UAE CKDScr+UAE was more prevalent in those with an affected first-
In the subsample (~60,000 participants), prevalence of degree relative (Fig S4), although the trend was less pro-
CKDScr+UAE was higher than that of CKDScr (Fig S3). nounced compared to that for CKDScr. In the subsample, the

Table 1. Baseline Characteristics of Adult and Juvenile Participants

Adults (Age ≥18) Juveniles (Age 8-17) Total
Complete sample
No. of participants 147,715 8,196 155,911
Age, y 44.83 ± 13.12 12.21 ± 2.76 43.11 ± 14.71
Male sex 61,512 (41.64%) 3,891 (47.47%) 65,403 (41.95%)
BMI, kg/m2 26.07 ± 4.33 18.91 ± 3.18 25.69 ± 4.56
Current smoker 31,156 (21.38%) NAa NAa
Hypertension 38,605 (26.13%) NAa NAa
Diabetes 5,673 (3.84%) NAa NAa
Hypercholesterolemia 29,888 (20.23%) NAa NAa
Serum potassium, mEq/L 3.86 ± 0.30 3.84 ± 0.28 3.86 ± 0.3
Serum creatinine, mg/dL 0.83 ± 0.14 0.61 ± 0.13 0.82 ± 0.15
eGFR, mL/min/1.73 m2 96.41 ± 15.29 111.46 ± 16.81 97.2 ± 15.74
CKDScrb 1,858 (1.26%) 4 (0.05%) 1,862 (1.19%)
Subsample
No. of participants 59,195 748 59,943
Serum calcium, mg/dL 9.14 ± 0.32 9.50 ± 0.28 9.14 ± 0.32
Serum sodium, mmol/L 141.74 ± 1.84 141.63 ± 1.68 141.73 ± 1.84
Uric acid, mg/dL 4.88 ± 1.18 4.37 ± 1.01 4.88 ± 1.18
Serum urea, mg/dL 14.48 ± 3.56 12.41 ± 2.75 14.45 ± 3.56
UACR, mg/g 2.72 [1.57-5.05] 2.93 [1.84-4.84] 2.72 [1.58-7.33]
UAE, mg/d 3.86 [2.33-6.92] NAa NAa
UACR ≥30 mg/g 1,622 (2.73%) 20 (2.68%) 1,642 (2.73%)
UAE ≥30 mg/d 2,431 (4.10%) NAa NAa
CKDScr+UACR 3,338 (5.52%) 24 (3.20%) 3,362 (5.49%)
CKDScr+UAE 4,127 (6.83%) NAa NAa
Data for continuous variables given as mean ± SD or median [interquartile range], and for categorical variables, as number (%). Conversion factors for calcium in mg/dL to
mmol/L, ×0.2495; creatinine in mg/dL to μmol/L, ×88.4; uric acid in mg/dL to μmol/L, ×59.48; serum urea in mg/dL to mmol/L, ×0.357.
Abbreviations: BMI, body mass index; CKD, chronic kidney disease; CKDScr, CKD defined by eGFR <60 mL/min/1.73 m2; CKDScr+UACR, CKD defined by eGFR <60 mL/
min/1.73 m2 and/or UACR ≥30 mg/g; CKDScr+UAE, CKD defined by eGFR <60 mL/min/1.73 m2 and/or UAE ≥30 mg/d; eGFR, estimated glomerular filtration rate; NA,
not available; Scr, serum creatinine; UACR, urinary albumin-creatinine ratio; UAE, urinary albumin excretion.
a
Data were not available for juveniles.
b
These included 21 participants with extremely low eGFR values (less than −5 SD from the mean) who were considered CKD patients and were retained in analyses of
CKD recurrence but were excluded from heritability analyses.

872 AJKD Vol 77 | Iss 6 | June 2021

Zhang et al

150

eGFR (mL/min/1.73 m2)

100

50 Mean value of eGFR

First-degree relative affected with CKDSCr
General population

0
10

3
-1

-2

-2

-3

-3

-4

-4

-5

-5

-6

-6

-7

-7

-8

-8

-9
8-

11

16

21

26

31

36

41

46

51

56

61

66

71

76

81

86
Age group

B
100
Age-specific prevalence of CKDSCr
90
First-degree relative affected with CKDSCr
80 General population

70
Prevalence (%)

60

50

40

30

20

10

0
10

3
-1

-2

-2

-3

-3

-4

-4

-5

-5

-6

-6

-7

-7

-8

-8

-9
8-

11

16

21

26

31

36

41

46

51

56

61

66

71

76

81

86

Age group

Figure 2. Comparisons of (A) age-specific mean values of eGFR and (B) age-specific prevalence of CKDScr between individuals with
affected first-degree relatives and the general population. Error bars indicate 95% CI. Abbreviation: CKDScr, CKD defined by
eGFR<60 mL/min/1.73 m2.

RRR for CKDScr+UAE was 1.34 (95% CI, 1.14-1.58), whereas
for CKDScr this was 2.35 (95% CI, 1.74-3.17) (Fig S5). Use
of UACR instead of UAE as a measurement of albuminuria
yielded highly similar results (Figs S4-S8). Sensitivity analysis
excluding juveniles yielded similar results (Fig S9).
We repeated our analyses using CKDUAE. On average,
those with a first-degree relative affected by CKDUAE had
higher UAE (Fig S6). Higher prevalence of CKDUAE was
observed among those with an affected first-degree rela-
tive (Fig S7). The RRR for CKDUAE was 1.60 (95% CI,
1.26-2.03). Risk was elevated mainly in those with an
affected mother or sibling. No elevated spousal risk was
observed (Fig S8). Results for CKDUACR were highly
similar.
Heritability Estimates
We report heritability estimates of the CKD-defining traits
eGFR (44%), UAE (20%), and UACR (19%), for the kidney
biomarkers serum urea (31%), Scr (37%), and uric acid
(48%), and for the serum electrolytes potassium (28%),
calcium (27%), and sodium (22%) (Table 2). In the sub-
sample, heritability estimates of eGFR, Scr, and serum po-
tassium were consistent with the estimates in the full sample
but less precise (Table S5), indicating low risk of bias due to
missingness. Heritability estimates did not change when ac-
counting for household or spousal effects. These effects
explained <0.1% of the variance in each outcome, which was
corroborated by the modest spousal correlations (Table S6).
Inclusion of additional covariates did not substantially change

AJKD Vol 77 | Iss 6 | June 2021 873

 Zhang et al

First degree relatives with CKD 3.04 (2.26-4.09)

Parent with CKD 3.09 (2.13-4.48)
Father with CKD 3.31 (1.63-6.73)
Mother with CKD 2.80 (1.71-4.60)
Offspring with CKD 2.78 (1.73-4.48)
Son with CKD 4.75 (2.30-9.82)
Daughter with CKD 2.21 (1.27-3.86)
Sibling with CKD 3.63 (2.28-5.78)
Brother with CKD 2.39 (1.18-4.82)
Sister with CKD 4.52 (2.73-7.47)
Spouse with CKD 1.56 (1.20-1.96)
Husband with CKD 1.57 (1.23-2.02)
Wife with CKD 1.55 (1.20-2.00)

0.1 1 10
Decreased risk Increased risk

Figure 3. Recurrence risk ratios for CKDScr, adjusted for age, age2, sex, BMI, hypertension, diabetes, high cholesterol, history of car-
diovascular disease, and smoking status. Error bars indicate 95% CI. Abbreviations: BMI, body mass index; CKDScr, CKD defined by
eGFR <60 mL/min/1.73 m2.

the estimates of heritability. Age and sex explained 42% of
the phenotypic variance of eGFR and only 2.6% of UAE and
8.3% of UACR. Inclusion of additional covariates increased
the PVC for most traits only slightly, with the exception of
uric acid, which showed a substantial increase in PVC of
10.3 percentage points. The PVC for model 2 ranged from
4.6% for potassium to 44.1% for uric acid (Table 2).
Sensitivity analysis excluding juveniles yielded slightly higher
heritability estimates for eGFR and serum creatinine
(Table S7).
Discussion
In this large population-based family study, we investi-
gated the familial aggregation of CKD by comparing the
risk of CKD in individuals with an affected first-degree
relative to that in the general population. Participants
with an affected first-degree relative were observed to have
a 3-fold higher risk of CKD than that in the general pop-
ulation, independent of BMI, hypertension, diabetes, hy-
percholesterolemia, history of CVD, and smoking status.
We observed a 1.56-fold higher risk in those with an
affected spouse, suggesting that shared environmental
factors and/or assortative mating play a role. Heritability
of eGFR was considerable (44%), whereas heritability of
UAE was moderate (20%). Heritability of kidney-related
markers and serum electrolytes ranged between 20% and
50%. These results indicate an important role for genetic
factors in modulating susceptibility to kidney disease in the
general population.
In this study, participants with an affected first-degree
relative were observed to have a 3-fold higher risk of

Table 2. Heritability of Kidney Traits and Related Biomarkers

Model 1 Model 2
Traits N h2 ± SE PVC h2 ± SE PVC
eGFR 155,911 0.435 ± 0.007 0.420 0.436 ± 0.007 0.423
ln(UAE)a 59,145 0.199 ± 0.014 0.026 0.193 ± 0.014 0.048
ln(UACR)a 59,938 0.185 ± 0.014 0.083 0.178 ± 0.014 0.103
Uric acida 58,519 0.481 ± 0.013 0.338 0.498 ± 0.013 0.442
Serum creatinine 155,911 0.373 ± 0.007 0.374 0.379 ± 0.007 0.377
Serum ureaa 58,481 0.307 ± 0.013 0.218 0.308 ± 0.013 0.219
Serum potassium 155,842 0.279 ± 0.007 0.041 0.278 ± 0.007 0.050
Serum calciuma 58,488 0.268 ± 0.013 0.059 0.267 ± 0.013 0.079
Serum sodiuma 58,444 0.217 ± 0.013 0.066 0.221 ± 0.013 0.074
Model 1, adjusted for age, age2, sex. Model 2, adjusted for age, age2, sex, body mass index, diabetes, hypertension, high cholesterol, history of cardiovascular disease, and
smoking status. P < 0.001 for all heritability estimates. UAE and UACR were natural log (ln)-transformed; eGFR and other markers were normally distributed.
Abbreviations: eGFR, estimated glomerular filtration rate; h2, heritability; PVC, proportion of variance due to covariates; SE, standard error; UACR, urinary albumin-
creatinine ratio; UAE, urinary albumin excretion.
a
Data were available only for a subsample of adult participants.

874 AJKD Vol 77 | Iss 6 | June 2021

Zhang et al
CKD. Previous studies that examined familial aggregation
of CKD focused on its end stage (ie, kidney failure).
Recurrence risk of kidney failure in the case of an affected
first-degree relative has been estimated in African Ameri-
cans (9-fold higher risk),44 Taiwanese Han-Chinese (2.5-
fold higher risk),12 and in a multiancestry (African and
European) US population-based case-control study (1.3- to
10-fold higher risk).8 Relative risk in a large Norwegian
registry-based study was 7.2.15 Among dialysis patients in
End-Stage Renal Disease Network 6 in the United States,
23% have close relatives receiving kidney replacement
therapy for kidney failure,45 and individuals with a family
history of kidney replacement therapy are at increased risk
for CKD.46
In registry data, early stages of CKD remain unrecog-
nized. The present study is unique in that it is based on
objective laboratory measurements of eGFR and UAE
spanning data from 3 generations, and is therefore more
sensitive to nonsymptomatic, early-stage CKD. Data for
familial recurrence of CKD may guide clinical decision
making with regard to CKD diagnosis and prevention.
Further study is warranted to assess the added value of
family history in CKD risk stratification, and to investigate
the potential impact of targeting families of CKD patients
in screening and prevention.
The RRR of CKDScr+UAE was significant, although
considerably lower than that of CKDScr. This may be
explained by our observation that familial patterns of
CKDUAE were less pronounced than those of CKDScr.
Furthermore, we observed only moderate heritability
(~20%) for measures of albuminuria (ie, genetic factors
contribute relatively little to between-individual variation
in urinary albumin excretion, whereas eGFR was highly
heritable (~44%). The comparatively higher RRR of
CKDScr is therefore expected.
Spouses of those affected by CKD were at a 1.56-fold
higher risk of CKD; in addition, kidney traits showed
weak but significant positive correlations between spouses.
As spouses are unrelated, the increased risk of CKD in
spouses and spousal correlations of kidney traits may
reflect effects of shared environmental factors or assortative
mating. To further assess the effects of shared environment
on kidney traits, we examined family and spouse effects as
variance components in our heritability models. These
effects were negligible, therefore the elevated risk in
spouses seems more related to assortative mating (ie,
partner selection based on phenotypes that convey higher
risk of CKD). Strong evidence of assortment exists for
factors such as smoking,47 height, BMI, and educational
attainment,48 each of which is a potential determinant of
CKD risk. In the present study, however, spousal correla-
tions of eGFR and UAE did not diminish after adjustment
for renal risk factors (including BMI and smoking status).
Thus, assortment likely occurs on factors other than
currently known CKD determinants, which may be
explored in future study in spousal pairs.
AJKD Vol 77 | Iss 6 | June 2021
Between-study comparisons of heritability estimates are
not straightforward, as phenotypic variance and contri-
bution of genetic factors depend on population, ethnicity,
environment, measurement methods, and sampling error.
Some variability in estimates can therefore be expected.
Previously, the heritability of eGFR was described in
several twin studies and a few community-based studies.
In the present large-scale study, we observed a heritability
of 44% for eGFR, corroborating estimates from previous,
relatively small-scale population-based studies, such as
from Switzerland (46%)18 and from South Tyrol, Italy
(39%).13 In pedigree data from the population-based
Framingham Heart Study, the heritability estimate for
eGFR was lower (33%),49 similar to that in Zuni Indians
(33%).24 The lower estimates are possibly due to popu-
lation differences, random sampling error, or use of older,
less precise GFR estimating methods.50 As generally
observed for most traits, twin studies (50%-67.3%)21,22,51
yielded somewhat higher heritability estimates for eGFR
than family-based studies (33%-46%).13,18,24,52
The heritability estimate of urinary albumin (ie, UAE
and UACR) in the present study (20%) was similar to that
in a previous Swiss population-based study (23%)18 that
also collected 24-hour urine samples. The heritability of
UACR was 21% in Pima Indians,45 and 25% in Zuni In-
dians.24 Heritability of UACR in European ancestry twins
was 45%.53 Previous studies in patients with diabetes re-
ported highly variable heritability estimates (21%-46%) of
albuminuria measured in spot urine samples.54-58 Finally,
we observed 22%-28% heritability for the serum electro-
lytes potassium, calcium, and sodium, confirming the
potential for identifying genetic variants involved in elec-
trolyte homeostasis in the general population.
Across all traits studied in twins, heritability is on
average ~49%.59 Heritability for kidney traits is compa-
rable with those estimated for, for example, blood pressure
traits (h2: 17-52%).60
The heritability estimates in the present study provide
an upper bound to the amount of phenotypic variance that
can be attributed to genetic factors. Large-scale GWAS thus
far identified 306 common single-nucleotide poly-
morphisms (SNPs) for Scr-based eGFR, explaining 7.1% of
phenotypic variance,13 whereas the present study estimates
the heritability of Scr-based eGFR to be 44%. Similarly, the
59 SNPs thus far identified in GWAS on UACR explain
0.7%, which is modest compared to our heritability esti-
mate of 20%.14 Future genetic study of rare variants may
potentially explain this “missing” heritability.61,62
For the present work, CKD was defined according to an
age-independent eGFR cutoff value of <60 mL/min/1.73
m2 as described in international guidelines.29,63 It has been
argued that an age-dependent eGFR cutoff for CKD is also
appropriate, as the current definition has been suggested to
lead to overdiagnosis of CKD in elderly64 and misclassifica-
tion of CKD. Future study may involve analyzing heritability
using alternative definitions of CKD.
875

We present the largest family-based study of kidney
traits that uses laboratory-defined CKD and the first study
to quantify the familial clustering of CKD including early
(ie, non–kidney failure) stages of CKD. Furthermore, the
Lifelines Cohort Study is representative of the general
population of northern Netherlands,65 facilitating precise
heritability estimation. Additionally, albuminuria was
determined in both spot urine samples and in 24-hour
urine collections, the latter being considered the gold
standard to assess albuminuria and available to few large-
scale epidemiological studies.
Several limitations need to be addressed. First,
although the gold standard 24-hour albuminuria mea-
surements were available, this was only true for a sub-
sample of approximately 60,000 participants. However,
age, sex, and covariate distributions were highly similar
between the subsample and the full sample, as were
heritability estimates of eGFR, Scr, and serum potassium.
Thus, missingness was likely random and unlikely to
have seriously biased our results. Second, GFR was not
measured but was estimated from Scr. Therefore, bias is
possible due to creatinine metabolism. In addition, GFR
estimating equations are known to be less precise in the
higher range (>60 mL/min/1.73 m2),66-68 possibly
causing downward bias in our heritability estimates.
Third, no kidney biopsy data were available, nor could
we exclude Mendelian forms of inherited kidney disease;
we therefore could not distinguish between the different
etiologies of CKD. Fourth, potential preferential miss-
ingness of data from nonparticipating affected family
members may have caused underestimation of recur-
rence risk ratios. Fifth, the low prevalence of CKD in our
study population (CKDScr: 1.19%) may have inflated
relative risk estimates. Finally, >95% of Lifelines Cohort
Study participants were of European ancestry69; there-
fore, we cannot generalize our results to other ancestries.
In summary, we demonstrate that CKD clusters in
families in the general population, given that risk of CKD
was strongly elevated in those with an affected relative.
Considerable heritability (20-50%) of kidney traits was
observed. Therefore, much of the familial clustering may
be attributed to genetic factors. The data presented in this
study inform future work on risk stratification based on
family history, and provide a step forward in disentangling
genetic and environmental risk factors in CKD.
Supplementary Material
Supplementary File (PDF)
Figure S1: Flow chart of eGFR and UACR analysis.
Figure S2: Flow chart of eGFR and UAE analysis.
Figure S3: Age-specific prevalence of CKD in the general popula-
tion diagnosed by different criteria.
Figure S4: Comparisons of age-specific prevalence of CKD be-
tween individuals with affected first-degree relatives and the general
population.
876
Zhang et al
Figure S5: RRRs for CKD in individuals with affected first-degree
relatives or spouse.
Figure S6: Comparisons of age-specific geometric means of UAE
and UACR.
Figure S7: Comparisons of age-specific prevalence of
UAE ≥ 30 mg/d and UACR ≥ 30 mg/g.
Figure S8: RRRs for UAE ≥ 30 mg/d and UACR ≥ 30 mg/g.
Figure S9: Recurrence risk of CKDScr in adult participants only.
Item S1: Detailed methods.
Table S1: Family structure.
Table S2: Baseline characteristics of adult and juvenile participants
stratified by sex and age.
Table S3: Baseline characteristics of participants with and without
CKDScr family history.
Table S4: Baseline characteristics of adult and juvenile participants
in the subsample, stratified by sex and age.
Table S5: Heritability of eGFR, Scr, and potassium in subsample
participants only.
Table S6: Spousal correlations.
Table S7: Heritability of eGFR, Scr, and potassium in adult partici-
pants only.
Article Information
Authors’ Full Names and Academic Degrees: Jia Zhang, PhD,
Chris H.L. Thio, PhD, Ron T. Gansevoort, MD, PhD, and Harold
Snieder, PhD.
Authors’ Affiliations: Department of Epidemiology, University
Medical Center Groningen, University of Groningen, Groningen,
the Netherlands (JZ, CHLT, HS); Division of Nephrology,
Department of Internal Medicine, University Medical Center
Groningen, University of Groningen, Groningen, the Netherlands
(RTG); Shenzhen Center for Chronic Disease Control, Shenzhen,
Guangdong, China (JZ); and Department of Epidemiology and
Statistics, Institute of Basic Medical Sciences, Chinese Academy
of Medical Sciences, and School of Basic Medicine, Peking
Union, Medical College, Beijing, China (JZ).
Address for Correspondence: Harold Snieder, PhD, Department of
Epidemiology (HPC FA40), University Medical Center Groningen,
Hanzeplein 1, PO Box 30001, 9700 RB, Groningen, the
Netherlands. Email: h.snieder@umcg.nl
Authors’ Contributions: Study conception and design: HS, CHLT,
JZ; data analysis: JZ, CHLT; data interpretation and literature
review: CHLT, JZ, HS, RTG; research mentorship and supervision:
HS, RTG. Each author contributed important intellectual content
during manuscript drafting or revision and agrees to be personally
accountable for the individual’s own contributions and to ensure
that questions pertaining to the accuracy or integrity of any portion
of the work, even one in which the author was not directly
involved, are appropriately investigated and resolved, including
with documentation in the literature if appropriate.
Support: The Lifelines Cohort Study was supported by the
Netherlands Organization of Scientific Research NWO grant
175.010.2007.006; the Economic Structure Enhancing Fund (FES)
of the Dutch Government; the Ministry of Economic Affairs; the
Ministry of Education, Culture and Science; the Ministry for Health,
Welfare and Sports; the Northern Netherlands Collaboration of
Provinces (SNN); the Province of Groningen; the University
Medical Center Groningen; the University of Groningen; the Dutch
Kidney Foundation; and the Dutch Diabetes Research Foundation.
Dr Zhang was awarded a personal grant by China Scholarship
Council (CSC) during his study period in the Netherlands. The
AJKD Vol 77 | Iss 6 | June 2021
Zhang et al
funders had no role in study design, data collection, analysis,
reporting, or the decision to submit for publication.
Financial Disclosure: The authors declare that they have no
relevant financial interests.
Acknowledgements: We acknowledge the services of the Lifelines
Cohort Study, the contributing research centers delivering data to
Lifelines, and all study participants. We also thank Dr Chang Fu
Kuo for modeling the recurrence risk ratio of CKD in the general
population, and Arthur Gilmour for technical support with ASReml
software.
Peer Review: Received March 1, 2020. Evaluated by 2 external peer
reviewers, with direct editorial input from a Statistics/Methods
Editor, an Associate Editor, and the Editor-in-Chief. Accepted in
revised form November 6, 2020.
References
1. Jha V, Garcia-Garcia G, Iseki K, et al. Chronic kidney disease:
Global dimension and perspectives. Lancet. 2013;382(9888):
260-272.
2. Bruck K, Stel VS, Gambaro G, et al. CKD prevalence varies
across the European general population. J Am Soc Nephrol.
2016;27(7):2135-2147.
3. Go AS, Chertow GM, Fan D, McCulloch CE, Hsu CY. Chronic
kidney disease and the risks of death, cardiovascular events,
and hospitalization. N Engl J Med. 2004;351(13):1296-1305.
4. Kahn MR, Robbins MJ, Kim MC, Fuster V. Management of
cardiovascular disease in patients with kidney disease. Nat Rev
Cardiol. 2013;10(5):261.
5. Gansevoort RT, Correa-Rotter R, Hemmelgarn BR, et al. Chronic
kidney disease and cardiovascular risk: epidemiology, mecha-
nisms, and prevention. Lancet. 2013;382(9889):339-352.
6. Nitsch D, Grams M, Sang Y, et al. Associations of esti-
mated glomerular filtration rate and albuminuria with mor-
tality and renal failure by sex: a meta-analysis. BMJ.
2013;346:f324.
7. Echouffo-Tcheugui JB, Kengne AP. Risk models to predict
chronic kidney disease and its progression: a systematic re-
view. PLoS Med. 2012;9(11):e1001344.
8. Lei HH, Perneger TV, Klag MJ, Whelton PK, Coresh J. Familial
aggregation of renal disease in a population-based case-con-
trol study. J Am Soc Nephrol. 1998;9(7):1270-1276.
9. Queiroz Madeira EP, da Rosa Santos O, Ferreira Santos SF,
Alonso da Silva L, MacIntyre Innocenzi A, Santoro-Lopes G.
Familial aggregation of end-stage kidney disease in Brazil.
Nephron. 2002;91(4):666-670.
10. Satko SG, Freedman BI. The familial clustering of renal disease
and related phenotypes. Med Clin North Am. 2005;89(3):447-
456.
11. Fava C, Montagnana M, Burri P, et al. Determinants of kidney
function in Swedish families: role of heritable factors.
J Hypertens. 2008;26(9):1773-1779.
12. Wu HH, Kuo CF, Li IJ, et al. Family aggregation and heritability
of ESRD in Taiwan: a population-based study. Am J Kidney
Dis. 2017;70(5):619-626.
13. Wuttke M, Li Y, Li M, et al. A catalog of genetic loci associated
with kidney function from analyses of a million individuals. Nat
Genet. 2019;51(6):957-972.
14. Teumer A, Li Y, Ghasemi S, et al. Genome-wide association
meta-analyses and fine-mapping elucidate pathways influ-
encing albuminuria. Nat Commun. 2019;10(1):1-19.
15. Skrunes R, Svarstad E, Reisaeter AV, Vikse BE. Familial clus-
tering of ESRD in the Norwegian population. Clin J Am Soc
Nephrol. 2014;9(10):1692-1700.
AJKD Vol 77 | Iss 6 | June 2021
16. Visscher PM, Hill WG, Wray NR. Heritability in the genomics
era - concepts and misconceptions. Nat Rev Genet.
2008;9(4):255-266.
17. Hunter DJ, Lange M, Snieder H, et al. Genetic contribution to
renal function and electrolyte balance: a twin study. Clin Sci
(Lond). 2002;103(3):259-265.
18. Moulin F, Ponte B, Pruijm M, et al. A population-based
approach to assess the heritability and distribution of renal
handling of electrolytes. Kidney Int. 2017;92(6):1536-1543.
19. Whitfield J, Martin N. The effects of inheritance on constituents
of plasma: a twin study on some biochemical variables. Ann
Clin Biochem. 1984;21(3):176-183.
20. Bathum L, Fagnani C, Christiansen L, Christensen K. Herita-
bility of biochemical kidney markers and relation to survival in
the elderly—results from a Danish population-based twin study.
Clinica Chimica Acta. 2004;349(1-2):143-150.
21. Raggi P, Su S, Karohl C, Veledar E, Rojas-Campos E,
Vaccarino V. Heritability of renal function and inflammatory
markers in adult male twins. Am J Nephrol. 2010;32(4):317-
323.
22. Arpegard J, Viktorin A, Chang Z, de Faire U, Magnusson PK,
Svensson P. Comparison of heritability of cystatin C- and
creatinine-based estimates of kidney function and their relation
to heritability of cardiovascular disease. J Am Heart Assoc.
2015;4(1):e001467.
23. Pilia G, Chen W, Scuteri A, et al. Heritability of cardiovascular
and personality traits in 6,148 Sardinians. PloS Genet.
2006;2(8):e132.
24. MacCluer JW, Scavini M, Shah VO, et al. Heritability of mea-
sures of kidney disease among Zuni indians: the Zuni kidney
project. Am J Kidney Dis. 2010;56(2):289-302.
25. Stolk RP, Rosmalen JG, Postma DS, et al. Universal risk factors
for multifactorial diseases. Eur J Epidemiol. 2008;23(1):67-74.
26. Scholtens S, Smidt N, Swertz MA, et al. Cohort profile: Life-
Lines, a 3-generation cohort study and biobank. Int J Epi-
demiol. 2014;44(4):1172-1180.
27. Maya L, Esteban A, van der Graaf A, et al. Lack of association
between genetic variants at ACE2 and TMPRSS2 genes
involved in SARS-CoV-2 infection and human quantitative
phenotypes. Front Genet. 2020;11:613.
28. Levey AS, de Jong PE, Coresh J, et al. The definition, classifi-
cation, and prognosis of chronic kidney disease: a KDIGO
controversies conference report. Kidney Int. 2011;80(1):17-28.
29. Levin A, Stevens PE, Bilous RW, et al. Kidney disease:
Improving global outcomes (KDIGO) CKD work group. KDIGO
2012 clinical practice guideline for the evaluation and man-
agement of chronic kidney disease. Kidney International Sup-
plements. 2013;3(1):1-150.
30. Levey AS, Stevens LA, Schmid CH, et al. A new equation to
estimate glomerular filtration rate. Ann Intern Med.
2009;150(9):604-612.
31. Schwartz GJ, Munoz A, Schneider MF, et al. New equations to
estimate GFR in children with CKD. J Am Soc Nephrol.
2009;20(3):629-637.
32. Risch N. Linkage strategies for genetically complex traits. I.
multilocus models. Am J Hum Genet. 1990;46(2):222-228.
33. Breslow N. Covariance analysis of censored survival data.
Biometrics. 1974:89-99.
34. Skov T, Deddens J, Petersen MR, Endahl L. Prevalence pro-
portion ratios: Estimation and hypothesis testing. Int J Epi-
demiol. 1998;27(1):91-95.
35. Barros AJ, Hirakata VN. Alternatives for logistic regression in
cross-sectional studies: an empirical comparison of models
that directly estimate the prevalence ratio. BMC Med Res
Methodol. 2003;3(1):21.
877

36. Lin D. Cox regression analysis of multivariate failure time data:
The marginal approach. Stat Med. 1994;13(21):2233-2247.
37. Biswas A, Datta S, Fine JP, Segal MR. Statistical Advances in
the Biomedical Science. Wiley Online Library; 2007. Accessed
January 27, 2021. https://onlinelibrary.wiley.com/
38. Lichtenstein P, Yip BH, Bj€ ork C, et al. Common genetic de-
terminants of schizophrenia and bipolar disorder in Swedish
families: a population-based study. Lancet. 2009;373(9659):
234-239.
39. Kuo C, Grainge MJ, Valdes AM, et al. Familial risk of Sj€ ogren’s
syndrome and co-aggregation of autoimmune diseases in
affected families: a nationwide population study. Arthritis
Rheumatol. 2015;67(7):1904-1912.
40. Kuo C, Grainge MJ, Valdes AM, et al. Familial aggregation of
systemic lupus erythematosus and coaggregation of autoim-
mune diseases in affected families. JAMA Intern Med.
2015;175(9):1518-1526.
41. Gilmour AR, Thompson R, Cullis BR. Average information
REML: An efficient algorithm for variance parameter estimation
in linear mixed models. Biometrics. 1995:1440-1450.
42. Gilmour A, Gogel B, Cullis B, Welham S, Thompson R. ASReml
User Guide Release 4.1 Structural Specification. Hemel
Hempstead, UK: VSN International Ltd; 2015. Accessed
January 27, 2021. https://asreml.kb.vsni.co.uk/wp-content/
uploads/sites/3/2018/02/ASReml-4.1-Functional-Specification.
pdf
43. R Core Team. R: A Language and Environment for Statistical
Computing. Vienna, Austria: R foundation for statistical
computing; 2014. Accessed January 27, 2021. http://www.R-
project.org/
44. Freedman BI, Spray BJ, Tuttle AB, Buckalew VM Jr. The familial
risk of end-stage renal disease in African Americans. Am J
Kidney Dis. 1993;21(4):387-393.
45. Freedman BI, Volkova NV, Satko SG, et al. Population-based
Screening for family history of end-stage renal disease among
incident dialysis patients. Am J Nephrol. 2005;25(6):529-
535.
46. McClellan WM, Satko SG, Gladstone E, Krisher JO, Narva AS,
Freedman BI. Individuals with a family history of ESRD are a
high-risk population for CKD: Implications for targeted surveil-
lance and intervention activities. Am J Kidney Dis. 2009;53(3):
S100-S106.
47. Agrawal A, Heath AC, Grant JD, et al. Assortative mating for
cigarette smoking and for alcohol consumption in female
Australian twins and their spouses. Behav Genet. 2006;36(4):
553-566.
48. Robinson MR, Kleinman A, Graff M, et al. Genetic evidence of
assortative mating in humans. Nat Hum Behav. 2017;1(1):
0016.
49. Fox CS, Yang Q, Cupples LA, et al. Genomewide linkage
analysis to serum creatinine, GFR, and creatinine clearance in a
community-based population: the Framingham Heart Study.
J Am Soc Nephrol. 2004;15(9):2457-2461.
50. Michels WM, Grootendorst DC, Verduijn M, Elliott EG,
Dekker FW, Krediet RT. Performance of the Cockcroft-Gault,
MDRD, and new CKD epidemiology formulas in relation to
GFR, age, and body size. Clin J Am Soc Nephrol. 2010;5(6):
1003-1009.
51. Pasha DN, Davis JT, Rao F, et al. Heritable influence of DBH on
adrenergic and renal function: twin and disease studies. PloS
One. 2013;8(12):e82956.
878
Zhang et al
52. Bochud M, Elston RC, Maillard M, et al. Heritability of
renal function in hypertensive families of African descent
in the Seychelles (Indian Ocean). Kidney Int. 2005;67(1):
61-69.
53. Rao F, Wessel J, Wen G, et al. Renal albumin excretion: Twin
studies identify influences of heredity, environment, and adrenergic
pathway polymorphism. Hypertension. 2007;49(5):1015-1031.
54. Forsblom CM, Kanninen T, Lehtovirta M, Saloranta C,
Groop LC. Heritability of albumin excretion rate in families of
patients with type II diabetes. Diabetologia. 1999;42(11):
1359-1366.
55. Fogarty DG, Rich SS, Hanna L, Warram JH, Krolewski AS.
Urinary albumin excretion in families with type 2 diabetes is
heritable and genetically correlated to blood pressure. Kidney
Int. 2000;57(1):250-257.
56. Imperatore G, Knowler WC, Pettitt DJ, Kobes S, Bennett PH,
Hanson RL. Segregation analysis of diabetic nephropathy in
Pima Indians. Diabetes. 2000;49(6):1049-1056.
57. Langefeld CD, Beck SR, Bowden DW, Rich SS,
Wagenknecht LE, Freedman BI. Heritability of GFR and albu-
minuria in Caucasians with type 2 diabetes mellitus. Am J
Kidney Dis. 2004;43(5):796-800.
58. Krolewski AS, Poznik G, Placha G, et al. A genome-wide link-
age scan for genes controlling variation in urinary albumin
excretion in type II diabetes. Kidney Int. 2006;69(1):129-136.
59. Polderman TJ, Benyamin B, De Leeuw CA, et al. Meta-analysis
of the heritability of human traits based on fifty years of twin
studies. Nat Genet. 2015;47(7):702-709.
60. Kolifarhood G, Daneshpour M, Hadaegh F, et al. Heritability of
blood pressure traits in diverse populations: a systematic re-
view and meta-analysis. J Hum Hypertens. 2019:1-11.
61. Nolte IM, Tropf FC, Snieder H. Missing Heritability Of Complex
Traits and Diseases. eLS. Wiley Online Library; 2019.
Accessed January 27, 2021. https://doi.org/10.1002/978047
0015902.a0028223
62. Wainschtein P, Jain DP, Yengo L, et al. Recovery of Trait Heri-
tability from Whole Genome Sequence Data. Accessed March
1, 2020. https://www.biorxiv.org/. 2019:588020
63. Coresh J, Gansevoort RT. CKD prognosis consortium. Levin A,
Jadoul M, KDIGO. Current CKD definition takes into account
both relative and absolute risk. J Am Soc Nephrol. 2020;31(2):
447-448.
64. Delanaye P, Jager KJ, Bokenkamp A, et al. CKD: A call for an
age-adapted definition. J Am Soc Nephrol. 2019;30(10):1785-
1805.
65. Klijs B, Scholtens S, Mandemakers JJ, Snieder H, Stolk RP,
Smidt N. Representativeness of the LifeLines cohort study.
PloS One. 2015;10(9):e0137203.
66. Stevens LA, Coresh J, Greene T, Levey AS. Assessing kidney
function—measured and estimated glomerular filtration rate.
N Engl J Med. 2006;354(23):2473-2483.
67. Earley A, Miskulin D, Lamb EJ, Levey AS, Uhlig K. Estimating
equations for glomerular filtration rate in the era of creatinine
standardization: A systematic review. Ann Intern Med.
2012;156(11):785-795.
68. Levey AS, Inker LA, Coresh J. GFR estimation: From physiology
to public health. Am J Kidney Dis. 2014;63(5):820-834.
69. van der Ende, Yldau M, Hartman MH, Hagemeijer Y, et al. The
LifeLines cohort study: prevalence and treatment of cardio-
vascular disease and risk factors. Int J Cardiol. 2017;228:495-
500.
AJKD Vol 77 | Iss 6 | June 2021
Original Investigation

Joint Associations of Maternal-Fetal APOL1 Genotypes

and Maternal Country of Origin With Preeclampsia Risk
Xiumei Hong,* Avi Z. Rosenberg,* Boyang Zhang, Elizabeth Binns-Roemer, Victor David, Yiming Lv,
Rebecca C. Hjorten, Kimberly J. Reidy, Teresa K. Chen, Guoying Wang, Yuelong Ji, Claire L. Simpson,
Robert L. Davis, Jeffrey B. Kopp, Xiaobin Wang, and Cheryl A. Winkler

Visual Abstract online

Rationale & Objectives: Preeclampsia, which
disproportionately affects Black women, is a
leading cause of preterm delivery and risk for
future hypertension and chronic kidney disease
(CKD). Apolipoprotein L1 (APOL1) kidney risk
alleles, common among Black individuals,
contribute substantially to CKD disparities. Given
the strong link between preeclampsia and CKD,
we investigated whether maternal and fetal
APOL1 risk alleles can jointly influence pre-
eclampsia risk, and explored potential modifiers of
the association between APOL1 and
preeclampsia.
Study Design: Nested case-control study.
Setting & Participants: 426 Black mother-infant
pairs (275 African Americans and 151 Haitians)
from the Boston Birth Cohort.
Exposure: Maternal and fetal APOL1 risk alleles.
Outcomes: Preeclampsia.
Analytical Approach: Logistic regression models
with adjustment for demographic characteristics
were applied to analyze associations between
fetal and maternal APOL1 risk alleles and risk of
preeclampsia and to investigate the effects of
modification by maternal country of origin.
Results: Fetal APOL1 risk alleles tended to be
associated with an increased risk of preeclamp-
sia, which was not statistically significant in the
total genotyped population. However, this asso-
ciation was modified by maternal country of origin Complete author and article
(P < 0.05 for interaction tests): fetal APOL1 risk information provided before
references.
alleles were significantly associated with an
increased risk of preeclampsia among African Correspondence to
Americans under recessive (odds ratio [OR], 3.6 X. Wang (xwang82@jhu.
edu) or C.A. Winkler
[95% CI, 1.3-9.7]; P = 0.01) and additive (OR,
(winklerc@mail.nih.gov)
1.7 [95% CI, 1.1-2.6]; P = 0.01) genetic models
but not in Haitian Americans. Also, maternal-fetal *X.H. and A.Z.R. contributed
equally to this work.
genotype discordance at the APOL1 locus was
associated with a 2.6-fold higher risk of Am J Kidney Dis.
preeclampsia (P < 0.001) in African Americans. 77(6):879-888. Published
online December 22, 2020.
Limitations: Limited sample size in stratified an- doi: 10.1053/
alyses; self-reported maternal country of origin; j.ajkd.2020.10.020
pre-pregnancy estimated glomerular filtration
© 2020 by the National
rate (eGFR) and proteinuria data in mothers
Kidney Foundation, Inc.
were not collected; unmeasured confounding
social and/or environmental factors; no
replication study.
Conclusions: This study supports the hypothe-
sis that fetal APOL1 kidney risk alleles are
associated with increased risk for preeclampsia
in a recessive mode of inheritance in African
Americans and suggests that maternal-fetal
genotype discordance is also associated with
this risk. These conclusions underscore the
need to better understand maternal-fetal
interaction and their genetic and environmental
factors as contributors to ethnic disparities in
preeclampsia.

preeclampsia13,14 and preeclampsia increases the risk for

P reeclampsia, principally characterized by the onset of
hypertension after the 20th week of pregnancy, is a
leading cause of medically indicated preterm delivery,
future risk of cardiovascular diseases,1-3 and chronic kid-
Editorial, p. 863
ney disease (CKD)4-6 in mothers and offspring.7 Pre-
eclampsia disproportionately affects Black women
worldwide.8 Despite decades of research, the cause of
preeclampsia remains unclear, and there are no established
genetic or environmental factors that account for the
higher risk of preeclampsia in Black women. Genetic var-
iants in the apolipoprotein L1 (APOL1) gene account for
much of the excess risk of kidney disease in Black in-
dividuals.9-12 It is estimated that 12%-14% of African
Americans carry 2 APOL1 kidney risk alleles. Given that pre-
existing CKD increases the risk of developing
subsequent CKD, 4-6 it is likely that APOL1 variants play a
role in preeclampsia susceptibility. Bruggeman et al15 re-
ported that pregnant transgenic mice expressing the APOL1
variant G0 or G2 allele in the placenta developed a
preeclampsia-like phenotype. Although the preeclampsia
phenotype was observed in mice, whether they were
transgenic for APOL1 G0 or G2, the latter exhibited more
severe disease. Preeclampsia was dependent on the APOL1
genotype of the pup rather than the dam, suggesting
transgenic APOL1 expressed by the fetus or placenta was
responsible for this phenotype. Consistent with this, Reidy
et al16 found that fetal but not maternal APOL1 high-risk
genotypes were associated with a 2-fold increased risk of
preeclampsia in 2 African American cohorts.
Previous studies have indicated that both maternal and
fetal genetic backgrounds influence preeclampsia risk.17
Discordance between maternal and fetal genotypes may

AJKD Vol 77 | Iss 6 | June 2021 879


PLAIN-LANGUAGE SUMMARY
Preeclampsia, characterized by increased blood pressure
after 20 weeks of pregnancy, as well as other abnor-
malities (eg, protein in the urine), is dangerous to
mothers and their infants. Previous studies found that
individuals with African ancestry may carry APOL1 ge-
netic variants that increase risk for chronic kidney dis-
ease. This study found that fetal high-risk APOL1
genotypes and maternal-fetal APOL1 genotype discor-
dance independently contribute to preeclampsia risk in
African American mothers. This association was not
observed in Haitian mother-infant pairs possibly
because of different environmental exposures and cul-
tural milieu. Additional studies are required to under-
stand why APOL1 associations with preeclampsia differ
by maternal country of origin and to improve man-
agement of mothers at risk for preeclampsia.
lead to altered risk of preeclampsia through the maternal
immune system.18 For example, a study of family triads
showed that particular combinations of mother-child HLA-
G genotypes influenced the risk for preeclampsia.19 We
therefore planned to test the novel hypothesis that
maternal-fetal genotype combinations at the APOL1 locus
influence risk of preeclampsia.
Within the US Black population, there is a significant
variation in country of origin (eg, maternal self-identified
country of identity) and nativity (birthplace), which may
affect disease profiles through factors such as different
proportions of African descent, distinct within-group
cultural norms, and health behaviors.20-23 Given that the
frequencies of APOL1 risk alleles may vary among distinct
ethnic populations in Africa and among the African dias-
pora,24,25 further studies are needed to investigate whether
country of origin and nativity may modify the associations
between APOL1 genotypes and risk of preeclampsia.
We genotyped APOL1 genetic variants in 426 Black
mother-infant pairs enrolled in the Boston Birth Cohort
(BBC), a predominantly urban, low-income population at
high risk for preeclampsia. We tested maternal and fetal
APOL1 kidney risk allele-preeclampsia associations sepa-
rately and explored whether these associations were
affected by 1) maternal country of origin (African Amer-
icans versus Haitians); and 2) maternal birthplace (US-
born vs non–US-born). We further tested whether
maternal-fetal genotype discordance (or mismatch) of the
APOL1 genotypes affected preeclampsia risk.
Methods
Study Design, Setting, and Participants
Using a nested case-control study design, we studied 213
Black mother-infant pairs with preeclampsia (cases) and
213 Black pairs without preeclampsia (controls) from the
880
Hong et al
Exclusion:
Figure 1. Flow chart of the study participants.
BBC. The BBC study began in 1998 at the Boston Medical
Center, as previously reported elsewhere.26,27 The BBC
study population mirrors the patient population of the
Boston Medical Center and contains a relatively high
prevalence of preterm births. Mothers who delivered
singleton live births were invited to participate in the BBC
study within 24-72 hours after delivery. Pregnancies that
involved multiple gestations, fetal chromosomal abnor-
malities, or major birth defects or were the result of
in vitro fertilization were excluded. After each mother
provided written informed consent, she was interviewed
using a standardized questionnaire to gather sociodemo-
graphic and other epidemiologic data. Their electronic
medical records were abstracted at delivery. The study
protocol was approved by the Institutional Review Boards
of Boston University Medical Center and Johns Hopkins
Bloomberg School of Public Health.
Selection of cases and controls for the present study is pre-
sented in Figure 1. Briefly, we included all available Black
mother-infant pairs with preeclampsia (cases, n = 213), and
we sought to enroll the same number of Black controls. For
control selection (Figure 1), we first performed individual
matching on 3 variables including maternal age (+/- 5 years),
parity (defined as the number of times that the mother has given
birth to a fetus with a gestational age of 24 weeks or more) and
infant’s sex. For those cases without matching controls, we
selected controls that matched with the cases on either 2 of the 3
variables. After selection, cases and controls were largely
comparable as to the 3 matching variables (Table 1).
Definition of Outcome
Physician diagnoses of preeclampsia were extracted from the
maternal electronic medical records.28 Preeclampsia was
defined according to the report of the National High Blood
Pressure Education Program Working Group on High Blood
Pressure in Pregnancy, as systolic blood pressure
of ≥140 mm Hg or diastolic blood pressure of ≥90 mm Hg
on at least 2 occasions, and proteinuria ≥1+ after 20 weeks
of gestation.29 For women with preexisting hypertension,
AJKD Vol 77 | Iss 6 | June 2021
Hong et al

Table 1. Characteristics of 426 Black Mother-Infant Pairs With and Without Preeclampsia
Enrolled Pairsa
Variables Cases Controls Eligible Controls in the BBCb
No. of pairs 213 213 1,691
Maternal age, y 29.6 ± 7.0 29.2 ± 6.9 28.2 ± 6.6
Marital status
Married 75 (35.2%) 70 (32.9%) 584 (34.5%)
Others 136 (63.9%) 137 (64.3%) 1,076 (63.6%)
Missing 2 (0.9%) 6 (2.8%) 31 (1.8%)
Maternal prepregnancy BMI
18.5-24.9 kg/m2 63 (29.6%) 106 (49.8%) 783 (46.3%)
25.0-29.9 kg/m2 59 (27.7%)c 49 (23.0%) 455 (26.9%)
≥ 30 kg/m2 80 (37.6%)d 40 (18.8%) 351 (20.8%)
Missing 11 (5.2%) 18 (8.5%) 102 (6.0%)
Haitian maternal country-of-origin 75 (35.2%) 76 (35.7%) 538 (31.8%)
Maternal nativity or birthplace
Non-US born 118 (56.2%) 138 (64.8%) 1,008 (59.6%)
US born 92 (43.8%)e 75 (35.2%) 674 (39.9%)
Missing 3 (1.4%) 0 9 (0.5%)
Years stayed in the United States
<5.0 years 40 (18.8%) 53 (24.9%) 390 (23.1%)
5.0-9.9 years 29 (13.6%) 31 (14.6%) 245 (14.5%)
≥10.0 years 34 (16.0%) 36 (16.9%) 225 (13.3%)
Born in the United States 92 (43.2%) 75 (35.2%) 674 (39.9%)
Missing 18 (8.4%) 18 (8.4%) 157 (9.3%)
Maternal highest education level
≤High school 137 (64.3%) 127 (59.6%) 1,005 (59.5%)
College or above 75 (35.2%) 83 (39.0%) 662 (39.1%)
Missing 1 (0.5%) 3 (1.4%) 24 (1.4%)
Nulliparity 112 (52.6%) 105 (49.3%) 723 (42.8%)
Maternal smoking during pregnancy
Never 178 (83.5%) 187 (87.8%) 1,402(82.9%)
Quitter 20 (9.4%) 12 (5.6%) 127 (7.5%)
Current smoker 14 (6.6%) 10 (4.7%) 144 (8.5%)
Missing 1 (0.5%) 4 (1.9%) 18 (1.1%)
Alcohol drinking during pregnancy
No 180 (84.5%) 192 (90.1%) 1,488 (88.0%)
Yes 22 (10.3%) 13 (6.1%) 126 (7.5%)
Missing 11 (5.2%) 8 (3.8%) 77 (4.5%)
Chronic hypertension
No 153 (71.8%) 209 (98.1%) 1,630 (96.4%)
Yes 59 (27.7%)d 4 (1.9%) 55 (3.3%)
Missing 1 (0.5%) 0 6 (0.4%)
Preterm birth 116 (54.6%)d 0 0
Cesarean delivery 116 (54.6%)d 47 (22.1%) 487 (28.9%)b,e
Female sex of newborn 115 (54.0%) 109 (51.2%) 841 (49.7%)
Small for gestational age 43 (20.2%)d 18 (8.5%) 193 (11.4%)
Values are mean ± SD for maternal age and count (%) for other variables. Black mother-infant pairs with preeclampsia are cases; and matched mother-infant pairs without
pre-eclampsia are controls. Eligible controls are all eligible Black mother-infant pairs without preeclampsia in the parent Boston Birth Cohort.
Abbreviations: BBC, Boston Birth Cohort; BMI, body mass index.
a
The difference of population characteristics between preeclampsia cases and controls was tested based on univariate logistic regression models.
b
The differences in population characteristics between the enrolled 213 mother-infant pairs without preeclampsia and all eligible Black mother-infant pairs without pre-
eclampsia in the Boston Birth Cohort by t-test for continuous variables and χ2 test for categorical variables.
c
P < 0.01.
d
P < 0.001.
e
P < 0.05.

evidence of worsening hypertension (a systolic blood pres- during pregnancy (HELLP) syndrome was considered pre-
sure ≥160 mm Hg or a diastolic blood pressure of ≥110 mm sent when a physician made this diagnosis contemporane-
Hg) was required for a diagnosis of preeclampsia. Hemo- ously; the 213 cases included 210 mothers with
lysis, elevated liver enzymes, and low platelets developing preeclampsia and 3 with HELLP.

AJKD Vol 77 | Iss 6 | June 2021 881


Data Measurement
Using a standard questionnaire interview, maternal
epidemiological factors were collected, including self-
reported ethnicity defined by maternal country of
origin (African Americans vs Haitian), birthplace (US-
born vs non–US-born), age at delivery, highest educa-
tion level, parity, smoking, and alcohol consumption
during pregnancy. Non–US-born mothers were asked
about how long they had resided in the United States,
which was converted into a categorical variable (<5.0,
5.0-9.9, ≥10.0 years). Maternal pre-pregnancy body
mass index, calculated as self-reported weight in kilo-
grams divided by height in meters squared, was divided
into 4 groups: normal (<25.0 kg/m2), overweight (25-
29.9 kg/m2), obese (≥30 kg/m2), and unknown. Clin-
ical complications before and during pregnancy,
including chronic hypertension and pre-existing or
gestational diabetes, were extracted from electronic
medical records. The missing rate of each covariate is
presented in Table 1. In data analyses, we replaced
missing data with the most frequent values (for cate-
gorical variables with missingness in ≤5 participants) or
by using a missing indicator/category (for categorical
variables with missingness in >5 participants).
APOL1 Genotyping, and Genetic Ancestry
Proportion
Maternal DNA was isolated from blood within 1-3 days of
delivery, and fetal DNA was isolated from umbilical cord
blood. Three APOL1 variants (G1 allele [rs73885319 A>G
and rs60910145 T>G] and G2 allele [rs71785313 6-base
pair deletion]) were genotyped using customized Taq-
Man assays (Thermo Fisher Scientific) in maternal and fetal
DNA samples.30 The APOL1 alleles, including variants G1,
G2, and G0 (the last being neither the G1 nor the G2
allele), were coded accordingly. The number of APOL1
kidney risk alleles carried by each participant was coded as
0 for the G0/G0 genotype; 1 for the G0G1 or G0G2 ge-
notype; or 2 for the G1G1, G1G2, or G2G2 genotypes. The
APOL1 high-risk genotype was defined as carriage of any
combination of 2 G1 and/or G2 risk alleles and the low-
risk genotype was defined as carriage of 0 or 1 risk al-
leles. The APOL1 risk alleles were analyzed for associations
with preeclampsia under 3 genetic modes of inheritance:
recessive (2 vs 1 or 0 risk alleles; or high-risk vs low-risk
genotype); dominant (1 or 2 risk alleles vs 0 risk alleles);
and additive (coded as 0, 1, or 2 risk alleles as a continuous
variable).
A random subset of 161 mother-infant pairs (82 with
preeclampsia and 79 without preeclampsia) were geno-
typed using the Infinium Quality Control array (Infinium
QC Array-24), containing 15,949 ancestry-informative
markers for ancestry estimation. A principal component
analysis (“EIGENSTRAT” function in the R “AssocTests”
application; R Foundation software) was performed to
calculate eigenvectors for each participant. The first 2
882
Hong et al
eigenvectors, PC1 and PC2, representing the estimated
genetic ancestry for each participant, were then plotted by
maternal country of origin and birthplace to indicate
whether genetic ancestry varied by these 2 variables.
Statistical Analyses
Maternal and infant case characteristics were compared
with control characteristics using univariate logistic
regression models. Maternal and fetal APOL1 genotypes in
the controls were examined separately for agreement with
Hardy-Weinberg equilibrium (HWE) expectations using
χ 2 tests. Logistic regression models were applied to analyze
the association between each maternal or fetal APOL1 risk
allele and risk of preeclampsia, adjusting for conventional
factors associated with preeclampsia (pre-pregnancy body
mass index, smoking status, alcohol consumption, chronic
hypertension), and country of origin, and factors varied
between cases and controls in our cohort (years the
mothers resided in the United States). Point estimates and
95% confidence intervals (CI) of odds ratios (OR) were
calculated. Recessive, additive, and dominant modes of
inheritance were tested. Interaction effects between APOL1
genotypes and maternal country of origin (African
Americans vs Haitians) and maternal birthplace (US-born
vs non–US-born) were tested by adding those 2 variables
and their interaction term into the same logistic model.
Stratified analyses by maternal country of origin and
birthplace were also performed.
To investigate whether maternal and fetal APOL1 geno-
type discordance affected the risk of preeclampsia, we
generated a binary variable (“MFG_index” application) to
represent the APOL1 maternal-fetal genotype discordance
state, which was coded as “yes” if there were any differ-
ences between maternal and fetal APOL1 genotypes, or
“no” if maternal and fetal APOL1 genotypes were the same.
The association between the APOL1 maternal-fetal geno-
type discordance and the risk of preeclampsia was analyzed
using the logistic regression model, with adjustment of
covariates as described above, as well as with additional
adjustment for the main effect of maternal and/or fetal
APOL1 genotypes.
Results
Participant Characteristics
As shown in Table 1, the prevalence of pre-pregnancy
overweight/obesity, chronic hypertension, and US-born
nativity was higher in mothers with preeclampsia (cases,
n = 213) than in mothers without preeclampsia (controls,
n = 213); all are well-established risk factors for pre-
eclampsia. Preeclamptic mothers were more likely to un-
dergo cesarean delivery and have a small-for-gestational
age infant for the preeclamptic pregnancy (all P < 0.05).
Participant characteristics among the 213 controls enrolled
in this study and all the available Black controls
(n = 1,691) in the BBC (Fig 1) are shown in Table 1.
AJKD Vol 77 | Iss 6 | June 2021
Hong et al

Table 2. Association of Fetal APOL1 Risk Alleles With Risk of Preeclampsia

APOL1 Risk Allele Controls Cases OR (95% CI)a Pa
Total sample
0 110 (52.9%) 91 (43.5%) 1.00 (reference)
1 79 (38.0%) 90 (43.1%) 1.17 (0.73-1.87) 0.5
2b 19 (9.1%) 28 (13.4%) 1.79 (0.87-3.70) 0.1
Recessive 19 (9.1%) 28 (13.4%) 1.67 (0.83-3.34) 0.2
Dominant 98 (47.1%) 118 (56.5%) 1.28 (0.83-1.99) 0.3
Additive – – 1.28 (0.93-1.77) 0.1
APOL1 × maternal country-of-origin interactione
Recessive model 0.04
Dominant model 0.2
Additive model 0.05
Stratified by maternal country-of-origin
African Americans
0 74 (55.6%) 54 (40.3%) 1.00 (reference)
1 52 (39.1%) 59 (44.0%) 1.35 (0.75-2.44) 0.3
2c 7 (5.3%) 21 (15.7%) 4.06 (1.43-11.6) 0.009
Recessive 7 (5.3%) 21 (15.7%) 3.55 (1.29-9.74) 0.01
Dominant 59 (44.4%) 80 (59.7%) 1.64 (0.94-2.87) 0.08
Additive – – 1.72 (1.11-2.64) 0.01
Haitian
0 36 (48.0%) 37 (49.3%) 1.00 (reference)
1 27 (36.0%) 31 (41.3%) 0.93 (0.42-2.07) 0.9
2d 12 (16.0%) 7 (9.3%) 0.55 (0.17-1.78) 0.3
Recessive 12 (16.0%) 7 (9.3%) 0.57 (0.18-1.74) 0.3
Dominant 39 (52.0%) 38 (50.6%) 0.81 (0.39-1.71) 0.6
Additive 0.79 (0.46-1.35) 0.4
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio.
a
Adjusted for pre-pregnancy BMI category, smoking status during pregnancy, alcohol drinking during pregnancy, chronic hypertension, country-of-origin (for the analyses in
the total sample only), and years the mothers stayed in the United States.
b,c,d
The 2-degree of freedom test was performed for APOL1 risk allele, with bP = 0.3 in the total sample, cP = 0.02 in African Americans, and dP = 0.6 in Haitians.
e
In the interaction test, maternal country-of-origin was treated as a binary variable with the African American category as the reference group. APOL1 risk allele was
considered in a recessive, dominant, or additive model as shown.

These variables were comparable between those 2 control
groups, except that cesarean delivery was less frequent in
the 213 controls (Table 1).
Distribution of APOL1 Kidney Risk Alleles and
Associations with Preeclampsia
In mothers without preeclampsia (controls), the minor
allele frequencies for rs73885319, rs60910145, and
rs71785313 were approximately 19%, 19%, and 9%,
respectively, which were comparable to minor allele fre-
quencies in their infants (Table S1). Genotypic distribution
of rs71785313 obeyed HWE expectations in both mothers
and infants. Genotypic distribution of rs73885319 and
rs60910145 (in nearly complete linkage disequilibrium
with each other, R2 ~0.99) in the mothers whose condi-
tions varied slightly from HWE expectations (P < 0.05),
with an excess of the rs73885319 GG variant genotype and
the rs60910145 GG variant genotype (Table S1). This
variation was no longer significant when the HWE test was
performed separately in Haitian and African American
mothers.
Of the 421 successfully genotyped mothers, 214
(51%), 155 (37%), and 52 (12%) carried 0, 1, or 2 APOL1
risk alleles, respectively. Among the 417 genotyped in-
fants, 201 (48%), 169 (41%), and 47 (11%) carried 0, 1,
and 2 risk alleles, respectively. Among all genotyped par-
ticipants, neither fetal (Table 2) nor maternal (Table 3)
APOL1 risk alleles, under the 3 genetic models, had sig-
nificant associations with risk of preeclampsia, although
fetal APOL1 high-risk genotypes showed a trend toward an
increased risk of preeclampsia.
Modification Effects by Maternal Country-of-Origin
and Birthplace
To test whether the associations between APOL1 genotypes
and risk of preeclampsia varied by maternal country of
origin (Haitian vs African Americans), we performed
interaction tests and found a borderline interaction be-
tween fetal APOL1 risk alleles and maternal country of
origin on risk of preeclampsia, with P < 0.05 for the
interaction term when APOL1 alleles were analyzed using
the recessive or additive mode of inheritance (Table 2). In

AJKD Vol 77 | Iss 6 | June 2021 883

 Hong et al

Table 3. Associations of Maternal APOL1 Risk Alleles With Preeclampsia

APOL1 Risk Allele Controls Cases OR (95% CI)a Pa
Total Sample
0 115 (55.0%) 98 (46.4%) 1.00 (reference)
1 69 (33.0%) 86 (40.8%) 1.50 (0.93-2.40) 0.09
2b 25 (12.0%) 27 (12.8%) 1.11 (0.55-2.23) 0.8
Recessive 25 (12.0%) 27 (12.8%) 0.95 (0.49-1.86) 0.9
Dominant 94 (45.0%) 113 (53.6%) 1.39 (0.90-2.14) 0.1
Additive – – 1.17 (0.86-1.60) 0.3
APOL1 × maternal country-of-origin interactione
Recessive model 0.02
Dominant model 0.1
Additive model 0.03
Stratified by maternal country-of-origin
African-Americans
0 75 (56.4%) 56 (40.9%) 1.00 (reference)
1 45 (33.8%) 60 (43.8%) 1.63 (0.89-2.98) 0.1
2c 13 (9.8%) 21 (15.3%) 2.20 (0.91-5.34) 0.08
Recessive 13 (9.8%) 21 (15.3%) 1.81 (0.77-4.23) 0.2
Dominant 58 (43.6%) 81 (59.1%) 1.76 (1.01-3.06) 0.05
Additive 1.53 (1.02-2.28) 0.04
Haitian
0 40 (52.6%) 42 (56.8%) 1.00 (reference)
1 24 (31.6%) 26 (35.1%) 1.15 (0.51-2.56) 0.7
2d 12 (15.8%) 6 (8.1%) 0.31 (0.08-1.17) 0.08
Recessive 12 (15.8%) 6 (8.1%) 0.30 (0.08-1.09) 0.07
Dominant 36 (47.4%) 32 (43.2%) 0.83 (0.40-1.74) 0.6
Additive 0.72 (0.42-1.23) 0.2
a
Adjusted for pre-pregnancy BMI category, smoking status during pregnancy, Alcohol drinking during pregnancy, chronic hypertension, country-of-origin (for the analyses
in the total sample only), and years the mothers stayed in the United States.
b,c,d
The 2-degree of freedom test was performed for maternal APOL1 risk allele, with bP = 0.2 in the total sample, cP = 0.1 in African Americans, and dP = 0.1 in Haitian
subset.
e
In the interaction test, maternal country-of-origin is treated as a binary variable with the African American category as the reference group. APOL1 risk allele was
considered in a recessive, dominant, or additive model as shown.

stratified analyses, we found that, among the African
American subset, fetal APOL1 risk alleles were signifi-
cantly associated with an increased risk of preeclampsia
under recessive (OR, 3.6 [95% CI, 1.3-9.7]; P = 0.01) and
additive (OR, 1.7 [95% CI, 1.1-2.6]; P = 0.01) modes of
inheritance. No such associations were observed in Hai-
tians (Table 2). In comparison, no significant interactions
were observed between fetal APOL1 risk alleles and
maternal birthplace, although fetal APOL1 risk alleles were
significantly associated with a higher risk of preeclampsia
under an additive (OR, 2.2 [95% CI, 1.2-4.0]; P = 0.01)
and a dominant (OR, 2.1 [95% CI, 1.0-4.4]; P = 0.04)
model in US-born but not in non–US-born subsets
(Table S2).
Similar interactions were also observed between
maternal APOL1 risk alleles and country of origin on risk of
preeclampsia, in recessive (P = 0.02) and additive
(P = 0.03) modes of inheritance, respectively (Table 3).
With stratified analyses and in an additive mode of in-
heritance, maternal APOL1 risk alleles were borderline
significantly associated with a higher risk of preeclampsia
(OR, 1.5 [95% CI, 1.0-2.3]; P = 0.04) only in the African
American subset, and not the Haitian subset. A similar
borderline association was observed when maternal APOL1
risk alleles were analyzed in a dominant mode (Table 3).
No significant effect modification of maternal birthplace
was observed for the relationships between maternal APOL1
risk alleles and risk of preeclampsia.
In a subset of the mother-infant pairs, we performed
ancestry principal component analyses, which did not
reveal significant differences in population substructure by
maternal country-of-origin or birthplace (Fig S1).
APOL1 Mother-Fetal Genotype Discordance and
Risk of Preeclampsia
Among the 411 pairs with available genotypic data in both
mothers and infants, 198 pairs had a discordance between
maternal and fetal APOL1 genotypes. Preeclampsia risk was
significantly higher in pairs with APOL1 maternal-fetal
genotype discordance compared with those without such
discordance (P = 0.04). Again, we found that this associ-
ation was significantly modified by maternal country-of-
origin (with P = 0.006 for interaction) (Table 4): APOL1
maternal-fetal genotype discordance was associated with

884 AJKD Vol 77 | Iss 6 | June 2021

Hong et al

Table 4. Associations Between APOL1 MFG Discordance and Risk of Preeclampsia

Model 1b Model 2c Model 3d
Controls Cases OR (95% CI) P OR (95% CI) P OR (95% CI) P
Total Sample
MFG 113 (55.4%) 100 (48.3%) 1.00 (reference) 1.00 (reference) 1.00 (reference)
discordance (N)a
MFG 91 (44.6%) 107 (51.7%) 1.58 (1.02-2.44) 0.04 1.53 (0.95-2.45) 0.08 1.46 (0.89-2.38) 0.1
discordance (Y)a
MFG 0.006 0.006 0.004
discordance ×
maternal
country-of-origin
interaction
African American mother-infant pairs
MFG 80 (62.0%) 59 (44.4%) 1.00 (reference) 1.00 (reference) 1.00 (reference)
discordance (N)a
MFG 49 (38.0%) 74 (55.6%) 2.61 (1.48-4.60) <0.001 2.34 (1.29-4.26) 0.005 2.19 (1.19-4.03) 0.01
discordance (Y)a
Haitian mother-infant pairs
MFG 33 (44.0%) 41 (55.4%) 1.00 (reference) 1.00 (reference) 1.00 (reference)
discordance (N)a
MFG 42 (56.0%) 33 (44.6%) 0.60 (0.29-1.27) 0.2 0.68 (0.30-1.57) 0.4 0.68 (0.27-1.68) 0.4
discordance (Y)a
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio.
a
Maternal-fetal genotype (MFG) discordance was coded as “Y” if there were any differences between maternal and fetal APOL1 genotypes or “N” if there were no
differences between maternal and fetal APOL1 genotypes.
b
Model 1: associations were tested using the logistic regression model, adjusted for pre-pregnancy BMI category, smoking status during pregnancy, alcohol drinking
during pregnancy, chronic hypertension, country of origin (for the analyses in the total sample only), and years mothers stayed in the United States.
c
Model 2 = Model 1 + maternal APOL1 allele status.
d
Model 3 = Model 1+ maternal APOL1 allele status + fetal APOL1 allele status.

~2.6-fold higher risk of preeclampsia (95% CI, 1.5-4.6;
P < 0.001) in African Americans but not in Haitians. These
associations remained largely unchanged after adjusting
for the main effect of maternal and/or fetal APOL1 risk
alleles (Table 4).
Discussion
This study confirmed the adverse impact of fetal APOL1 risk
alleles on preeclampsia risk among African Americans but
not among Haitians, which suggests that other environ-
mental or genetic factors may modify APOL1 risk allele
penetrance for preeclampsia. Importantly, we found that
APOL1 maternal-fetal genotype discordance was associated
with a 2.6-fold higher risk of preeclampsia in African
Americans, which was not confounded by maternal or fetal
APOL1 genotypes. This finding suggests that both maternal
and fetal APOL1 high-risk genotypes contribute to the
pathogenesis of preeclampsia.
Although the pathophysiological mechanisms by which
APOL1 risk variants damage the kidney are not yet fully
defined, current evidence suggests multiple mechanisms
may contribute, including plasma membrane cation channel
activity, endosomal trafficking, mitochondrial function, pro-
inflammatory cytokine expression, activation of inflamma-
somes and protein kinase R, and interactions with
APOL3.31–35 In the placenta, APOL1 is expressed in tropho-
blasts, Hoffbauer cells, and endothelial cells. Previously, we
showed very low levels of placental growth factor expression
that, together with high APOL1 expression in trophoblasts,
raised the possibility that an APOL1-mediated defect in
placentation contributed to preeclampsia. Consistently,
circulating antibodies against APOL1 are found in women
with preeclampsia, and serum APOL1 levels are higher in
women with preeclampsia.36,37
Very few studies have investigated the role of APOL1 risk
alleles in preeclampsia.16,38 In 2 US study cohorts, Reidy
et al16 reported that fetal but not maternal APOL1 high-risk
genotype was associated with an approximately 2-fold
higher risk of preeclampsia. Miller et al38, in 395 pre-
eclamptic cases and 282 controls, confirmed that infant
APOL1 genotype was associated with preeclampsia but in a
dominant inheritance model (Fig S2).38 Neither study
investigated potential factors that might modify APOL1-pre-
eclampsia associations, nor did they test the impact of
maternal-fetal genotype discordance. Robertson et al39
analyzed data from 2 genome-wide association studies of
prematurity, including 960 mothers from our BBC study and
519 mothers and 867 infants from GENEVA (Gene Envi-
ronment Association Studies). The study found no significant
associations between APOL1 high-risk genotypes and pre-
eclampsia. The 2 genome-wide association studies were
originally designed for genetic studies of prematurity, which
might have led to participant selection bias when pre-
eclampsia was analyzed as the outcome of interest, because
preeclampsia is a cause of prematurity. Also, fetal genome-

AJKD Vol 77 | Iss 6 | June 2021 885


wide association studies were not available from the BBC,
and determination of the APOL1 G2 variant in the mothers
was based on imputation. Finally, in GENEVA, preeclampsia
and chronic hypertension were assessed as a pooled binary
trait. Using chronic hypertension to define whether the
mother had preeclampsia could lead to phenotype misclas-
sification. Also, only ~50% of the study population had data
for this binary trait, limiting statistical power. In the African
American subset from the present study, we replicated the
association between fetal APOL1 risk alleles and preeclampsia
under the recessive mode of inheritance as reported by Reidy
et al16 in Black individuals from New York City, New York
and from Memphis, Tennessee. Although the present results
indicated that this association remained significant in an
additive mode of inheritance, we note that the effect size for
carrying 1 risk allele (OR, 1.4) was modest compared to the
effect size of carrying 2 risk alleles (OR, 4.1). Therefore, we
propose that the impact of fetal APOL1 genotypes on the risk
of preeclampsia is best explained by the recessive model, but
this awaits further validation in a larger cohort.
The present study suggests that maternal country-of-origin
modifies the relationship between fetal APOL1 high-risk ge-
notypes and risk of preeclampsia. One possible explanation is
that there are differences in regional African or European
genetic contributions between Haitians and African Ameri-
cans. Although our admixture analyses indicate that African
American and Haitian subsets had comparable genetic
ancestry, the ancestry-informative markers genotyped in this
study were designed to differentiate between European and
African ancestry and may not be able to distinguish regional
African or European contributions that may differ between
Haitians and African Americans. Another possibility is that
there are distinct environmental/social exposures or other
unmeasured genetic variants between these 2 subsets that
modify the APOL1-preeclampsia associations.
We observed a borderline-significant positive relation-
ship between maternal APOL1 risk allele and preeclampsia
risk in African Americans under both the additive and
dominant modes of inheritance compared to a null asso-
ciation as reported by Reidy et al.16 In comparison, the
study by Thakoordeen-Reddy et al40 reported that the
maternal APOL1 G1 risk allele was associated with 2.2 times
higher risk of having early-onset preeclampsia in 428
South African women of African ancestry. Maternal APOL1
risk alleles may affect the risk of preeclampsia directly or
through transmission of the risk allele to the infant,
because mothers would be obligate carriers of at least 1
risk allele for a fetus carrying 2 risk alleles and have a 50%
chance of being the source of the 1 risk allele in the het-
erozygous fetus. We further demonstrated that APOL1
maternal-fetal genotype discordance was a risk factor for
preeclampsia in African Americans. This finding, if vali-
dated in other studies, suggests a potential joint mother-
fetal genetic contribution to preeclampsia risk, high-
lighting the need to evaluate combined maternal-fetal ge-
notypes when assessing pregnancy complications as the
outcomes. Although the underlying biological mechanism
886
Hong et al
for the impact of maternal-fetal genotype combination is
unknown, a possible explanation is suggested by the
immunological maladaptation theory, which proposes that
the maternal immune system does not adapt fully to the
semi-allogenic fetus. Although the mother is usually
tolerant of discordant fetal antigens, there are exceptions.
These include paternal antigens that are discordant from
those of the mother and are absent from semen. Further
studies are needed to validate our findings on the joint
effects of mother-fetal APOL1 genotypes.
This study is to our knowledge the first to systematically
investigate the impact of the maternal APOL1 genotype,
fetal APOL1 genotype, and their combination effects on
preeclampsia in Black women, as well as to explore the
modification effects by maternal country-of-origin. Several
limitations should be acknowledged. First, statistical power
was relatively limited, especially for stratified analyses.
Second, maternal country-of-origin was self-reported and
may have been subject to reporting bias. Other population
characteristics were collected (within 1-3 days after de-
livery) after the diagnosis of presence or absence of pre-
eclampsia and thus might have led to recall bias. However,
none of the participants knew their APOL1 genotypes
during their interview, so recall bias, if it existed, will not
substantially influence the identified APOL1-preeclampsia
associations. Third, pre-pregnancy estimated glomerular
filtration rate and proteinuria data in mothers were not
collected in this study; therefore, their impact on the
APOL1-preeclampsia association cannot be analyzed.
Fourth, we acknowledge the possibility that the associa-
tions between fetal APOL1 risk allele (or maternal-fetal
genotype discordance) and risk of preeclampsia in Afri-
can Americans may be driven by some unmeasured ge-
netic/environmental variables and that the APOL1 risk allele
may actually serve as the surrogate for the unmeasured
variable. Fifth, in the analyses for APOL1 maternal-fetal
genotype discordance, we applied the most robust cod-
ing scheme for the genotypic discordance between the
mothers and the infants, as reported.18 However, it is
likely that other discordance models may be more pre-
dictive, and further studies with case-parent triads are
needed to validate and extend our findings. Sixth, the data
reported here have been collected since 1998, well before
the introduction of the 2019 definition of preeclampsia.
The new definition, in which proteinuria is no longer
necessary for diagnosis, will not significantly change our
classification of the case versus control group, and thus,
will not significantly change the associations direction as
we reported.
In conclusion, this study suggests that the effects of
both maternal and fetal APOL1 risk alleles are risk factors for
preeclampsia but APOL1 penetrance may be modified by
maternal country of origin, which may reflect differences
in as yet undetected environmental and/or genetic expo-
sures. Future research addressing these knowledge gaps
would advance understanding of preeclampsia, enable
more effective prevention and treatment of preeclampsia
AJKD Vol 77 | Iss 6 | June 2021
Hong et al
and might have implications for the prevention and
treatment of APOL1-mediated CKD across generations.
Supplementary Material
Supplementary File (PDF)
Figure S1: Plot for the principal component analysis of ancestry-
informative genotypes in mother-infant pairs, colored by country of
origin and birthplace.
Figure S2: Forest plot showing previously reported associations
between fetal APOL1 risk alleles and risk of preeclampsia.
Table S1: MAF and HWE tests of the three genotyped variants in
the APOL1 gene in Black mothers and their infants from the BBC.
Table S2: Characteristics of the non-Haitian Black mothers stratified
by maternal birthplace.
Article Information
Authors’ Full Names and Academic Degrees: Xiumei Hong, MD,
PhD, Avi Z. Rosenberg, MD, PhD, Boyang Zhang, BS, Elizabeth
Binns-Roemer, BS, Victor David, MS, Yiming Lv, BS, Rebecca C.
Hjorten, MD, Kimberly J. Reidy, MD, Teresa K. Chen, MD, MHS,
Guoying Wang, MD, PhD, Yuelong Ji, PhD, Claire L. Simpson,
PhD, Robert L. Davis, MD, MPH, Jeffrey B. Kopp, MD, Xiaobin
Wang, MD, MPH, ScD, and Cheryl A. Winkler, PhD.
Authors’ Affiliations: Department of Population, Family and
Reproductive Health, Center for the Early Life Origins of Disease,
Johns Hopkins University Bloomberg School of Public Health,
Baltimore, MD (XH, YL, GW, YJ, XW); Department of Pathology,
Johns Hopkins University, Baltimore, MD (AZR); Department of
Biostatistics, Johns Hopkins University Bloomberg School of
Public Health, Baltimore, MD (BZ); Molecular Genetic
Epidemiology Section, Frederick National Laboratory for Cancer
Research in the Basic Research Laboratory, National Cancer
Institute, Frederick, MD (EB-R, CAW); Basic Research Laboratory,
Center for Cancer Research, National Cancer Institute, Frederick,
MD (VD); Department of Medicine, Division of Nephrology,
University of WA, Seattle, WA (RCH); Department of Pediatrics,
Division of Pediatric Nephrology, Children’s Hospital at Montefiore,
Albert Einstein College of Medicine, Bronx, NY (KJR); Department
of Medicine, Division of Nephrology and Welch Center for
Prevention, Epidemiology, and Clinical Research, Johns Hopkins
Medical Institutions, Baltimore, MD (TKC); Department of
Genetics, Genomics, and Informatics, University of TN Health
Science Center, Memphis, TN (CLS); Center for Biomedical
Informatics, Department of Pediatrics, University of TN Health
Science Center, Memphis, TN (RLD); Kidney Diseases Section,
National Institute of Diabetes and Digestive and Kidney Disease,
National Institutes of Health, Bethesda, MD (JBK); and Division of
General Pediatrics and Adolescent Medicine, Department of
Pediatrics, Johns Hopkins University School of Medicine,
Baltimore, MD (XW).
Address for Correspondence: Cheryl A. Winkler, PhD, Frederick
National Laboratory for Cancer Research, 8560 Progress Drive,
Frederick, MD 21702 (email: winklerc@mail.nih.gov) or Xiaobin
Wang, MD, MPH, ScD, 615 N. Wolfe Street, E4132, Baltimore, MD
21205 (email: xwang82@jhu.edu).
Authors’ Contributions: Contributed to the study design: AZR, XH,
XW, CAW, JBK; oversaw the genotyping performed in this study and
genetic analyses: CAW, VD; performed data cleaning and data
analyses: XH, BZ, YL; provided critical inputs on study design,
data analyses, interpretation of the study findings: all authors;
founded and principal investigator of the Boston Birth Cohort (the
parent study): XW; oversaw participant recruitment and data
collection for the parent study: XW; contributed to participant
AJKD Vol 77 | Iss 6 | June 2021
recruitment and data collection for the parent study: XH, GW, YJ.
Each author contributed important intellectual content during
manuscript drafting or revision and agrees to be personally
accountable for the individual’s own contributions and to ensure
that questions pertaining to the accuracy or integrity of any portion
of the work, even one in which the author was not directly
involved, are appropriately investigated and resolved, including
with documentation in the literature if appropriate.
Support: The Boston Birth Cohort (the parent study) was supported
by National Insitutes of Health (NIH) grants R03HD096136,
R21HD085556, 2R01HD041702, R01HD086013,
R01HD098232, and R21AI154233. This project was supported
by National Institute of Diabetes and Digestive and Kidney
Diseases (NIDDK) Intramural Research Program grant ZO1 DK-
043308 (to JBK), an NIH Bench to Bedside grant supplementing
ZO1 DK-043308 and 3R01HD098232-02S1, and National
Cancer Institute of NIH Intramural Research Program (to Dr
Winkler) and under contract HHSN26120080001E. Dr Hong is
supported by the Johns Hopkins Population Center (Eunice
Kennedy Shriver National Institute of Child Health and Human
Development grant R24HD042854). Dr Chen was supported by
the Extramural Grant Program by Satellite Healthcare, a not-for-
profit renal care provider. Dr Reidy receives support from a
catalytic seed grant (NIH Clinical and Translational Science Award
1 UL1 TR001073 [Einstein/Montefiore]), NIDDK, and the
Preeclampsia Foundation. The funders were not involved in study
design, data collection, analysis, reporting, or the decision to
submit for publication.
Financial Disclosure: The authors declare that they have no
relevant financial interests.
Acknowledgements: The authors thank study participants in the
Boston Birth Cohort study for their support. The authors also
thank for the dedication and hard work of the field team at the
Department of Pediatrics, Boston University School of Medicine,
and for the support of the obstetric nursing staff at Boston
Medical Center. The authors also thank Linda Rosen, Boston
University Clinical Data Warehouse, for assistance in obtaining
relevant clinical information. The Clinical Data Warehouse service
is supported by Boston University Clinical and Translational
Institute and the NIH Clinical and Translational Science Award
U54-TR001012.
Disclaimer: Research reported in this publication was supported
through the NIH Bench-to-Bedside award made possible by
Office of Research on Women’s Health (ORWH). The content of
this publication does not necessarily reflect the view or policy of
the Department of Health and Human Services, nor does mention
of trade names, commercial products, or organizations imply
endorsement by the government.
Peer Review: Received April 23, 2020. Evaluated by 2 external peer
reviewers, with direct editorial input from a Statistics/Methods
Editor, an Associate Editor, and the Editor-in-Chief. Accepted in
revised form October 24, 2020.
References
1. Bergen NE, Schalekamp-Timmermans S, et al. Hypertensive
disorders of pregnancy and subsequent maternal cardiovas-
cular health. Eur J Epidemiol. 2018;33(8):763-771.
2. Veerbeek JH, Hermes W, Breimer AY, et al. Cardiovascular
disease risk factors after early-onset preeclampsia, late-onset
preeclampsia, and pregnancy-induced hypertension. Hyper-
tension. 2015;65(3):600-606.
3. Hammad IA, Meeks H, Fraser A, et al. Risks of cause-specific
mortality in offspring of pregnancies complicated by hyperten-
sive disease of pregnancy. Am J Obstet Gynecol.
2020;222(1):75 e71-75 e79.
887

4. Kristensen JH, Basit S, Wohlfahrt J, Damholt MB, Boyd HA.
Pre-eclampsia and risk of later kidney disease nationwide
cohort study. BMJ. 2019;365:l1516.
5. Khashan AS, Evans M, Kublickas M, et al. Preeclampsia and
risk of end stage kidney disease: A Swedish nationwide cohort
study. PLoS Med. 2019;16(7):e1002875.
6. Barrett PM, McCarthy FP, Kublickiene K, et al. Adverse preg-
nancy outcomes and long-term maternal kidney disease: a
systematic review and meta-analysis. JAMA Netw Open.
2020;3(2):e1920964.
7. Bokslag A, van Weissenbruch M, Mol BW, de Groot CJ. Pre-
eclampsia: short and long-term consequences for mother and
neonate. Early Hum Dev. 2016;102:47-50.
8. Urquia ML, Glazier RH, Gagnon AJ, et al. Disparities in pre-
eclampsia and eclampsia among immigrant women giving birth
in six industrialised countries. BJOG. 2014;121(12):1492-1500.
9. Reidy KJ, Hjorten R, Parekh RS. Genetic risk of APOL1 and
kidney disease in children and young adults of African ancestry.
Curr Opin Pediatr. 2018;30(2):252-259.
10. Genovese G, Friedman DJ, Ross Maryland, et al. Association of
trypanolytic ApoL1 variants with kidney disease in African
Americans. Science. 2010;329(5993):841-845.
11. Parsa A, Kao WH, Xie D, et al. APOL1 risk variants, race, and
progression of chronic kidney disease. N Engl J Med.
2013;369(23):2183-2196.
12. Williams WW, Pollak MR. Health disparities in kidney disease—
emerging data from the human genome. N Engl J Med.
2013;369(23):2260-2261.
13. Masuyama H, Nobumoto E, Okimoto N, Inoue S, Segawa T,
Hiramatsu Y. Superimposed preeclampsia in women with
chronic kidney disease. Gynecol Obstet Invest. 2012;74(4):
274-281.
14. Bramham K, Briley AL, Seed PT, Poston L, Shennan AH,
Chappell LC. Pregnancy outcome in women with chronic kid-
ney disease: a prospective cohort study. Reprod Sci.
2011;18(7):623-630.
15. Bruggeman LA, Wu Z, Luo L, et al. APOL1-G0 or APOL1-G2
transgenic models develop preeclampsia but not kidney dis-
ease. J Am Soc Nephrol. 2016;27(12):3600-3610.
16. Reidy KJ, Hjorten RC, Simpson CL, et al. Fetal-not maternal-
APOL1 genotype associated with risk for preeclampsia in
those with African ancestry. Am J Hum Genet. 2018;103(3):
367-376.
17. Gray KJ, Saxena R, Karumanchi SA. Genetic predisposition to
preeclampsia is conferred by fetal DNA variants near FLT1, a
gene involved in the regulation of angiogenesis. Am J Obstet
Gynecol. 2018;218(2):211-218.
18. Parimi N, Tromp G, Kuivaniemi H, et al. Analytical approaches to
detect maternal/fetal genotype incompatibilities that increase
risk of pre-eclampsia. BMC Med Genet. 2008;9:60.
19. Hylenius S, Andersen AM, Melbye M, Hviid TV. Association
between HLA-G genotype and risk of pre-eclampsia: a case-
control study using family triads. Mol Hum Reprod.
2004;10(4):237-246.
20. Wartko PD, Wong EY, Enquobahrie DA. Maternal birthplace is
associated with low birth weight within racial/ethnic groups.
Matern Child Health J. 2017;21(6):1358-1366.
21. Hammond WP, Mohottige D, Chantala K, Hastings JF,
Neighbors HW, Snowden L. Determinants of usual source of
care disparities among African American and Caribbean black
men: findings from the National Survey of American Life.
J Health Care Poor Underserved. 2011;22(1):157-175.
888
Hong et al
22. Huffman FG, De La Cera M, Vaccaro JA, et al. Healthy eating
index and alternate healthy eating index among Haitian Ameri-
cans and African Americans with and without type 2 diabetes.
J Nutr Metab. 2011;2011:398324.
23. Messiah SE, Atem F, Lebron C, et al. Comparison of early life
obesity-related risk and protective factors in non-Hispanic black
subgroups. Matern Child Health J. 2020;24(9):1130-1137.
24. Nadkarni GN, Gignoux CR, Sorokin EP, et al. Worldwide fre-
quencies of APOL1 renal risk variants. N Engl J Med.
2018;379(26):2571-2572.
25. Limou S, Nelson GW, Kopp JB, Winkler CA. APOL1 kidney risk
alleles: population genetics and disease associations. Adv
Chronic Kidney Dis. 2014;21(5):426-433.
26. Wang G, Divall S, Radovick S, et al. Preterm birth and random
plasma insulin levels at birth and in early childhood. JAMA.
2014;311(6):587-596.
27. Wang X, Zuckerman B, Pearson C, et al. Maternal cigarette
smoking, metabolic gene polymorphism, and infant birth weight.
JAMA. 2002;287(2):195-202.
28. Bustamante Helfrich B, Chilukuri N, He H, et al. Maternal
vascular malperfusion of the placental bed associated with
hypertensive disorders in the Boston Birth Cohort. Placenta.
2017;52:106-113.
29. Report of the National High Blood Pressure Education Pro-
gram Working Group on High Blood Pressure in Pregnancy.
Am J Obstet Gynecol. 2000;183(1):S1-S22.
30. David VA, Binns-Roemer EA, Winkler CA. Taqman assay for
genotyping CKD-associated APOL1 SNP rs60910145: a
cautionary note. Kidney Int Rep. 2019;4(1):184-185.
31. Ma L, Chou JW, Snipes JA, et al. APOL1 renal-risk variants
induce mitochondrial dysfunction. J Am Soc Nephrol.
2017;28(4):1093-1105.
32. Shah SS, Lannon H, Dias L, et al. APOL1 kidney risk variants
induce cell death through mitochondrial translocation and
opening of the mitochondrial permeability transition pore. J Am
Soc Nephrol. 2019;30(12):2355-2368.
33. Jha A, Kumar V, Haque S, et al. Alterations in plasma membrane
ion channel structures stimulate NLRP3 inflammasome activation
in APOL1 risk milieu. FEBS J. 2020;287(10):2000-2022.
34. O’Toole JF, Schilling W, Kunze D, et al. ApoL1 overexpression
drives variant-independent cytotoxicity. J Am Soc Nephrol.
2018;29(3):869-879.
35. Okamoto K, Rausch JW, Wakashin H, et al. APOL1 risk allele
RNA contributes to renal toxicity by activating protein kinase R.
Commun Biol. 2018;1:188.
36. Wen Q, Liu LY, Yang T, et al. Peptidomic identification of serum
peptides diagnosing preeclampsia. PLoS One. 2013;8(6):
e65571.
37. Elliott SE, Parchim NF, Liu C, et al. Characterization of antibody
specificities associated with preeclampsia. Hypertension.
2014;63(5):1086-1093.
38. Miller AK, Azhibekov T, O’Toole JF, et al. Association of pre-
eclampsia with infant APOL1 genotype in African Americans.
BMC Med Genet. 2020;21(1):110.
39. Robertson CC, Gillies CE, Putler RKB, et al. An investigation of
APOL1 risk genotypes and preterm birth in African American
population cohorts. Nephrol Dial Transplant. 2017;32(12):
2051-2058.
40. Thakoordeen-Reddy S, Winkler C, Moodley J, et al. Maternal
variants within the apolipoprotein L1 gene are associated with
preeclampsia in a South African cohort of African ancestry. Eur
J Obstet Gynecol Reprod Biol. 2020;246:129-133.
AJKD Vol 77 | Iss 6 | June 2021
Hong et al

Joint Associations of Maternal-Fetal APOL1 Genotypes and

Maternal Country-of-Origin With Preeclampsia Risk
Study Design Data Analyses Results: Preeclampsia Risk
Nested Case-Control Study APOL1 African-American Haitian
Boston Birth Cohort

APOL1 fetal G0 allele (REF)

vs
APOL1 fetal G1/G2 allele P = 0.04 for APOL1 country-of-origin interaction

Mother-fetal
APOL1 genotype concordance
vs
Mother-fetal
APOL1 genotype discordance P = 0.006 for discordance country-of-origin interaction

CONCLUSION: APOL1

Xiumei Hong, Avi Z. Rosenberg, Boyang Zhang, et al (2020)

@AJKDonline | DOI: 10.1053/j.ajkd.2020.10.020

AJKD Vol 77 | Iss 6 | June 2021 888.e1

Original Investigation

Estimation of Prevalence of Kidney Disease Treated With

Dialysis in China: A Study of Insurance Claims Data
Chao Yang, Zhao Yang, Jinwei Wang, Huai-Yu Wang, Zaiming Su, Rui Chen, Xiaoyu Sun, Bixia Gao, Fang Wang,
Luxia Zhang, Bin Jiang, and Ming-Hui Zhao, on behalf of the China Kidney Disease Network Work Group

Visual Abstract online

Rationale & Objective: The national prevalence
of dialysis in China has not been well studied. We
aimed to estimate the prevalence of kidney dis-
ease treated with dialysis and predict the trend
using claims data in order to provide evidence for
developing prevention strategies.
Study Design: Cross-sectional study of insur-
ance claims.
Setting & Participants: Medical claims data from
January 1, 2013, to December 31, 2017, were
extracted from a large claims database by using a
2-stage sampling design to obtain a national
sample covered by the urban basic medical
insurance, the most predominant insurance
program in China.
Exposure: Patients receiving maintenance dial-
ysis, including hemodialysis (HD) and peritoneal
dialysis (PD), were identified according to medical
billing data and International Classification of Dis-
eases, Tenth Revision (ICD-10) codes.
Outcomes: The age- and sex-standardized
population prevalence of kidney disease treated
with dialysis was estimated by year and
treatment modality.
Analytical Approach: Crude and age- and sex-
standardized prevalence of kidney disease
treated with dialysis were calculated stratified
by year and treatment modality. The gray Complete author and article
Verhulst model was used to predict dialysis information provided before
references.
prevalence from 2018 to 2025.
Correspondence to F. Wang
Results: The age-and sex-standardized
(wangfang@bjmu.edu.cn)
prevalence of dialysis patients increased from
or L. Zhang (zhanglx@bjmu.
255.11 per million population (pmp) in 2013 to edu.cn)
419.39 pmp in 2017. The age- and sex-
standardized prevalence of HD and PD in 2017 Am J Kidney Dis.
77(6):889-897. Published
were 384.41 pmp and 34.98 pmp, respectively,
online January 7, 2021.
and the total number of dialysis patients in China
doi: 10.1053/
was estimated to be 581,273. The prevalence of
j.ajkd.2020.11.021
dialysis was predicted to rise above 2017 levels,
© 2021 by the National
with a predicted prevalence of 534.60 pmp in
Kidney Foundation, Inc.
2020 and 629.67 pmp in 2025, corresponding
to 744,817 and 874,373 patients, respectively.
Limitations: Claims data have potential errors in
classification of patients, and population selec-
tion bias may have limited inferences to the entire
Chinese population.
Conclusions: The prevalence of kidney disease
treated with dialysis has risen between 2013
and 2017 in China and is predicted to increase
further through 2025. These findings highlight
the importance of prevention and control stra-
tegies to reduce the escalating burden of kidney
failure.

K idney failure has been recognized as a leading cause of

morbidity and mortality worldwide.1,2 When kidney
disease progresses to kidney failure, patients need kidney
replacement therapy (KRT)—including hemodialysis
(HD), peritoneal dialysis (PD), and kidney trans-
plantation—or conservative kidney management to
Editorial, p. 866
maintain life. The prevalence of patients receiving KRT
globally was reported to be 759 per million population
(pmp) according to the Global Kidney Health Atlas sur-
vey.2 However, no more than 50% of patients needing
KRT have access to it, with the largest gaps in low-income
countries, especially in Asia and Africa.1
China has the largest population in the world, and
chronic kidney disease (CKD) is prevalent in the country.3
Far-reaching societal and environmental change has shifted
the spectrum of CKD to mirror that of developed countries,
which may lead to a growing burden of kidney failure in
China.4,5 According to the National Report on the Services,
Quality and Safety in Medical Care System, the number of
HD and PD patients reported by the Chinese National Renal
Data System in 2017 was 510,101 and 86,264, respec-
tively, and increasing annually.6 The regional Beijing He-
modialysis Registry and Shanghai Renal Registry have
reported that the prevalence of HD was 524.6 pmp in
Beijing7 and 544.7 pmp in Shanghai8 in 2011, substan-
tially higher than national rates but still lower than those in
Japan (2,233 pmp in 2007).9 However, there has been no
well-functioning national surveillance system for kidney
diseases in China. Traditionally, dialysis registries have
been the only way to understand the epidemiology of
kidney failure, but this approach still has some disadvan-
tages, such as a complicated reporting process and missing
information, due to the large population and limited re-
sources in China. Therefore, the national prevalence of
dialysis patients in China has not been well studied.
Currently, more than 95% of China’s population are
covered by the Urban Employee Basic Medical Insurance
(UEBMI), the Urban Residents Basic Medical Insurance
(URBMI), and the New Co-operative Medical Care Scheme
(NCMS). The UEBMI and URBMI are the predominant
insurance programs in urban areas, whereas rural residents

AJKD Vol 77 | Iss 6 | June 2021 889


PLAIN-LANGUAGE SUMMARY
The national prevalence of dialysis in China has not
been well studied due to its large population and
limited resources. Insurance claims data provide a
unique opportunity to understand the burden of kidney
failure and have been used to characterize dialysis pa-
tients in the United States. Using a large nationwide
claims database, the age- and sex-standardized preva-
lence of kidney disease treated with dialysis in China
was estimated between 2013 and 2017. In addition, we
predicted the prevalence trend over time to estimate a
rising prevalence through 2025. Kidney failure has
become a major public health problem in China. It is
imperative to develop prevention and control strategies
to reduce the escalating burden of advanced kidney
disease.
are covered by the NCMS. By the end of 2018, the number
of people covered by the UEBMI and URBMI reached 317
and 897 million, respectively.10 As data volumes grow,
medical claims data have started to be used to identify and
monitor dialysis patients, functioning as an important data
source similar to the US Renal Data System (USRDS). Most
USRDS analyses use claims-based data.11 The China Kidney
Disease Network (CK-NET), initiated in 2014,12 has also
provided detailed data for the epidemiology of dialysis
patients based on government-funded and commercial
claims data.13
In this study, we have established a national surveillance
system for dialysis patients based on claims data, with a
strict data cleaning, analysis, and management process, the
CK-NET Surveillance System Based on Claims Data for
Dialysis (CK-NET SSCD). The prevalence of kidney disease
treated with dialysis in China was estimated, and the trend
was predicted, which could provide evidence for under-
standing the burden of kidney failure in China and
developing population-based prevention and control
strategies.
Methods
Source of Claims Data
The China Health Insurance Research Association (CHIRA)
database is a national claims database initiated in 2007,
covering information on diagnosis and treatment of out-
patients and inpatients at primary, secondary, and tertiary
hospitals. A 2-stage sampling design was used to extract a
national sample insured by the UEBMI and URBMI in 22
provinces, 5 autonomous regions, and 4 municipalities
directly under the central government (excluding Hong
Kong, Macao, and Taiwan). Details of the CHIRA database
have been described previously.13 Briefly, in the first stage,
convenience sampling was conducted in 4 municipalities
directly under the Central Government (Beijing, Shanghai,
890
Yang et al
Tianjin, and Chongqing), 27 provincial capital cities, and a
number of prefecture cities. In the second stage, a sys-
tematic random sampling sorted by age was used to extract
~2% of the insured population from the municipalities/
provincial capital cities, and ~5% from the selected
prefecture-level cities. The insured population was
resampled every year, with their whole-year claims data
recorded.
For each hospital visit, detailed claim records, including
any prescription drugs, surgical procedures or other
medical services were included in the CHIRA database
without missing values. Demographics, admission date,
primary diagnoses, and 2 secondary diagnoses as well as
medical expenditures were also recorded. The proportion
of missing data for diagnoses was 16.9% as the CHIRA
database was primarily designed to investigate the medical
service use of the insured population.
Medical claims data from January 1, 2013, to
December 31, 2017, were extracted into the CK-NET
SSCD under the authorization of management de-
partments of the CHIRA database. The number sampled
of the insured population was 6,555,217 (2013),
5,230,785 (2014), 7,131,626 (2015), 8,516,679
(2016), and 9,765,615 (2017). This study was approved
by the ethics committee of Peking University First Hos-
pital and the informed consent requirement was waived
by the ethics committee.
Identification of Prevalent Dialysis Patients
Patients receiving maintenance dialysis (including chil-
dren, adolescents, and adults) were identified according to
the service items in medical billing. HD patients were
identified by claims records of HD, including hemodia-
lyzer and related operations, and PD patients were selected
by having claims records of PD fluid. Text-based diagnosis
and International Classification of Diseases, Tenth revision
(ICD-10) codes were used to ascertain dialysis-related di-
agnoses, which are provided in Table S1. Patients with
ICD-10 codes of acute kidney injury were excluded
(Table S2). Comorbidities, including diabetes, hyperten-
sion, cardiovascular disease, dyslipidemia, anemia, and
CKD–mineral and bone disorder were identified mainly
based on the claims records of therapeutic drugs, surgical
procedures, and related medical services. The identification
strategies including claim records and ICD-10 codes were
manually reviewed by 3 senior nephrologists. The flow
chart of study population selection is shown in Fig 1. A
total of 64,372 maintenance dialysis patients were
included in the final analysis.
All personal information including name, identity card
number, medical insurance number, telephone number,
and home address were anonymized and de-identified
prior to analysis for privacy protection reasons.
Definitions of Proportion and Prevalence
The proportion was defined as the percentage of dialysis
patients among all the sampled, insured population in the
AJKD Vol 77 | Iss 6 | June 2021
Yang et al
China Health Insurance Research
Association (CHIRA) database from Jan 1,
2013 to Dec 31, 2017, n=37,199,922
Dialysis patients based on medical
billings, text-based diagnosis, and ICD-10
codes, n=80,921
Excluding patients
with AKI, n=16,549
Maintenance dialysis patients in the final
analysis, n=64,372
(n=57,780 for HD, n=6,592 for PD)
Figure 1. Flow chart of study population selection. Abbrevia-
tions: HD, hemodialysis; ICD-10, International Classification of
Diseases, 10th Revision; PD, peritoneal dialysis.
CHIRA database. Considering that insured persons without
any disease or records of a hospital visit could not be
captured by the CHIRA database, we used the total pop-
ulation in the CHIRA database and the rate of hospital visits
to calculate the denominator for prevalence estimates. The
rate of hospital visits was derived from the China Statistics
Yearbook and Statistical Communique of the People’s Re-
public of China on the National Economic and Social
Development.14,15 The crude prevalence was estimated by
multiplying the proportion and the rate of hospital visits of
the total insured population. The age- and sex-standardized
prevalence rates were calculated by the direct method using
the crude prevalence and the age and sex distribution based
on 2010 national census data. The following proportions,
rate of hospital visits, crude and age- and sex-standardized
prevalence rates were calculated as follows.
1) Proportion (%):
Prevalent dialysis patients in CHIRA database at the
end of a given year
× 100
Insured population visiting hospitals in CHIRA
database in the same year
2) Rate of hospital visit (%):
Insured population visiting hospitals in a given year
× 100
All insured population in the same year
3) Crude prevalence (pmp):
Prevalent dialysis patients at the end of a given year
× 1; 000; 000 dividuals and 114.29 pmp in female individuals), and the
All insured population in the same year
= Proportion × Visiting rate × 1; 000; 000 men and 835.32 pmp in women). The standardized
4) Age- and sex-standardized prevalence (pmp):
P
Crude prevalence in a given age and sex group × no: of
P
individuals in the corresponding group from the 2010 census
No: of individuals from the 2010 census × 1; 000; 000
AJKD Vol 77 | Iss 6 | June 2021
Statistical Analysis
Continuous data were presented as mean ± SD, and cate-
gorical variables were presented as proportions. Crude and
age- and sex-standardized prevalence of kidney disease
treated with dialysis were calculated, stratified by year and
treatment modality. The estimated total number of dialysis
patients was obtained by multiplying the standardized
prevalence with the total population in China in a given year.
The gray Verhulst model is one of the methods of gray
system theory that has been widely applied in various
fields such as social and medical research because of its
advantages in studying uncertainty problems and fore-
casting trends with poor information.16-20 We applied the
Verhulst model to fit the prevalence curve (2013-2017)
and to predict the prevalence and corresponding number
of patients (2018-2025). The detailed methods of the
Verhulst model and its time response formula have been
reported previously (Table S3).17,19 The post-test ratio (C
value) and small error probability (p value) were com-
bined to evaluate the fitting degree of the Verhulst model,
which was based on the standardized prevalence from
2013 to 2017 (Table S3, Table S4). The C value reflects the
concentration degree of the difference between predicted
values and actual values.
All data cleaning and statistical analyses were performed
using Excel version 2016 software (Microsoft) and R
version 3.6.3 software (R Foundation for Statistical
Computing).
Results
Characteristics and Overall Prevalence
The mean age of 19,923 dialysis patients in 2017 was
55.2 ± 16.2 years; 58.51% were male; and the largest
number of patients were in the age group of 45-64 years
(45.07%). HD patients accounted for 92.12% of all pa-
tients. The top 3 comorbidities were cardiovascular disease
(69.43%), anemia (67.25%), and hypertension (63.31%).
The proportion of patients with diabetes was 21.42%
(Table 1). The age- and sex-standardized prevalence of
dialysis patients increased from 255.11 pmp in 2013 to
419.39 pmp in 2017. The total number of dialysis patients
in China was estimated to be 346,178 in 2013 and
581,273 in 2017 (Table 2).
Prevalence by Age and Sex
The crude prevalence of dialysis in 2017 was the lowest in
the age group of 0-19 years (142.51 pmp in male in-
highest among people 65-74 years of age (956.70 pmp in
prevalence was higher in male than in female individuals.
For male individuals, the standardized prevalence was
301.62 pmp in 2013 and 472.03 pmp in 2017, respec-
tively, yielding a 1.6-fold increase. For female individuals,
the standardized prevalence rose from 206.31 pmp to
891
 Yang et al

Table 1. Characteristics of Dialysis Patients in China Based on Medical Claims Data, 2013-2017
2013 2014 2015 2016 2017
No. of patients in CHIRA database 8,173 6,364 11,829 18,083 19,923
Male sex 59.18% 57.98% 56.58% 57.73% 58.51%
Age, y 54.5 ± 15.8 56.4 ± 15.7 54.9 ± 15.3 55.6 ± 16.3 55.2 ± 16.2
Age group
0-19 y 0.51% 1.02% 0.52% 1.47% 1.98%
20-44 y 27.43% 22.22% 25.50% 23.17% 22.30%
45-64 y 43.00% 43.68% 45.21% 43.29% 45.07%
65-74 y 16.86% 19.61% 17.91% 19.06% 18.85%
≥75 y 12.20% 13.47% 10.87% 12.91% 11.62%
Unknown 0 0 0 0.11% 0.18%
Hemodialysis 82.71% 82.97% 90.98% 91.94% 92.12%
Comorbidity
Diabetes 26.33% 30.11% 18.62% 24.18% 21.42%
Hypertension 75.06% 73.62% 56.89% 62.79% 63.31%
Cardiovascular disease 79.54% 80.22% 61.97% 69.47% 69.43%
Dyslipidemia 21.27% 20.10% 14.84% 23.18% 21.94%
Anemia 71.70% 72.63% 71.97% 66.39% 67.25%
CKD-MBD 56.93% 57.10% 42.13% 44.27% 48.19%
Proportion receiving dialysisa 0.13% 0.12% 0.17% 0.21% 0.20%
Abbreviations: CHIRA, China Health Insurance Research Association; CKD-MBD, chronic kidney disease–mineral and bone disorder.
a
The proportion was defined as the percentage of dialysis patients among all the sampled insured population in the CHIRA database.

364.17 pmp over the same period, a 1.8-fold increase
(Table 3).
Prevalence by Treatment Modality
From 2013 to 2017, the age- and sex-standardized prev-
alence of HD patients was generally on the rise, whereas
the prevalence of PD patients was relatively stable (Fig 2).
The prevalence and number of HD patients almost doubled
in 5 years. The age- and sex-standardized prevalence of HD
and PD patients in 2017 was 384.41 pmp and 34.98 pmp,
respectively. Accordingly, it was estimated that the total
number of HD and PD patients was 532,791 and 48,482,
respectively.
Prediction of Prevalence
The predicted prevalence and number of patients receiving
dialysis had an increasing trend overall. The prevalence
was predicted to be 534.60 pmp in 2020 and 629.67 pmp
in 2025, with the corresponding numbers of patients be-
ing 744,817 and 874,373, respectively (Fig 3). The time
response formula of the Verhulst model was expressed as
x(1)(k + 1) = (0.001485 + 0.002435 × exp
[−0.263137k])−1, with an average relative error of
11.37%. The prediction accuracy was at level II
(C = 0.402; p = 1.000), which meant that the fitting model
was good.
Discussion
To the best of our knowledge, this is the first study to
assess the prevalence of dialysis patients in China by using
national claims data. The estimated number of dialysis
patients from CK-NET-SSCD is close to the data reported by
the Chinese National Renal Data System (581,273 vs
596,365 in 2017, respectively).6 Therefore, CK-NET-SSCD
could be used as an important supplementary means to
monitor the burden of kidney failure and help inform
medical practices in China. Considering the increasing
number of dialysis patients in the future, the development
of prevention and control strategies to reduce the esca-
lating burden of kidney failure is a priority.
Findings from an International Society of Nephrology
survey showed global variations in the prevalence of KRT,

Table 2. Crude and Age- and Sex-Standardized Prevalence of Dialysis Patients in China Based on Medical Claims Data, 2013-2017
2013 2014 2015 2016 2017
Crude prevalence, pmp 304.74 307.49 418.77 553.83 546.42
Age- and sex-standardized prevalence, pmpa 255.11 236.85 312.04 419.34 419.39
Estimated total no. of patientsb 346,178 323,068 427,806 578,273 581,273
Abbreviation: pmp, per million population.
a
The age- and sex-standardized prevalence was calculated by the direct method using 2010 national population census data.
b
The estimated total number of dialysis patients was obtained by multiplying the standardized prevalence by the total population in China: 1.357 billion (2013), 1.364 billion
(2014), 1.371 billion (2015), 1.379 billion (2016), and 1.386 billion (2017).

892 AJKD Vol 77 | Iss 6 | June 2021

Yang et al

Table 3. Age- and Sex-Specific Prevalence of Dialysis Patients in China Based on Medical Claims Data, 2013-2017
2013 2014 2015 2016 2017
Male Female Male Female Male Female Male Female Male Female
Age group
0-19 y 21.35 14.63 43.73 36.06 35.52 35.85 133.47 83.25 142.51 114.29
20-44 y 257.78 145.11 219.73 129.54 321.06 213.80 396.21 259.78 400.67 260.76
45-64 y 549.11 361.57 488.83 322.77 583.46 496.24 767.72 623.07 781.79 612.44
65-74 y 665.67 553.77 694.45 534.91 687.98 660.93 995.44 901.94 956.70 835.32
≥75 y 620.47 538.83 598.03 545.29 594.83 595.39 954.47 854.45 872.35 742.43
Totals
Crude prevalence 370.28 242.50 370.03 249.34 461.01 374.11 623.23 480.72 621.20 467.13
Standardized prevalencea 301.62 206.31 277.50 194.21 340.70 281.97 468.99 367.26 472.03 364.17
Prevalence is given per million population.
a
The 2010 national census was used as the reference population.

from 4 pmp in Rwanda to 3,392 pmp in Taiwan.2 The
prevalence of patients with kidney failure in the United
States has continued to rise and reached 746,557 cases
(2,203.5 pmp) in 2017, representing an increase of 2.6%
since 2016.11 But the standardized incidence rate in the
United States has declined slightly since 2006, which likely
reflects improvements in the prevention or postponement
of kidney failure.11 Overall, the prevalence of dialysis in
China reported by the present study is lower than that in
the United States, but our average annual growth rate is
significantly higher. In the past 2 decades, China has
witnessed a rapid surge of dialysis patients that is out of
proportion to its population growth.21 It was reported by
the Chinese Society of Nephrology in 1999 that the
prevalence and incidence of HD in China was 33.2 pmp
and 15.3 pmp, respectively; however, the response rates to
data requests in regions other than Beijing and Shanghai
were less than 50%.22 The crude prevalence of dialysis in
adults in 2015 was estimated in the CK-NET Annual Data
500
HD
Standardized prevalence (PMP)
Report to be 442.13 pmp,13 showing a rapidly increasing
trend in dialysis patients in China.
In this study, we calculated the age- and sex-
standardized prevalence from 2013 to 2017. The preva-
lence rate was obviously lower than that in Taiwan2 and
slightly lower than data reported from Beijing and
Shanghai.7,8 It is reported that the prevalence of CKD in
Taiwan (11.9%) was similar to that in mainland China
(10.8%).3,23 The high prevalence of patients receiving KRT
in Taiwan can be attributed to the broad coverage of the
National Health Insurance program, where patients can
receive dialysis almost free of charge.24 The ease of
accessibility to high-quality medical services also increases
patient survival in Taiwan.24 In contrast, insurance reim-
bursement policies for dialysis patients have been devel-
oped relatively recently in mainland China. Substantial
geographical and socioeconomic differences also exist in
the accessibility and affordability of kidney care.22,25
Moreover, the present data also show that PD still has
great potential application among the insured population.
The present findings about the sex-related disparities in
the prevalence of kidney failure are similar to those from
previous studies.26,27 There was a male preponderance

PD 384.38 384.41 among patients starting KRT, and female patients were less
400 likely to experience kidney failure.26,27 Potential explana-
283.39
tions include unhealthier lifestyles in men and greater
300 propensity to kidney failure, protective effects of endog-
210.14 enous estrogens and/or the damaging effects of testos-
196.34
terone, and other factors related to access to care.26,28
200
Additionally, elderly women are more likely to choose
conservative care instead of KRT.28
100 44.97 Accurate prediction of disease trends may help to
40.51 28.65 34.97 34.98
develop prevention and control strategies, but it is difficult
0 to make predictions with limited information, especially
2013 2014 2015 2016 2017 for time series and machine learning models.16 Therefore,
the gray prediction model is the best choice, and the
Figure 2. Age- and sex-standardized prevalence of HD and PD Verhulst model was used to predict the prevalence of pa-
patients in China based on medical claims data, 2013-2017. tients on dialysis, taking into account the S-shaped increase
Calculation was by the direct method using 2010 national pop- in prevalence observed in our study. The gray system
ulation census data. Abbreviations: HD, hemodialysis; PD, peri-
theory was proposed by Deng29 in 1982 and has shown
toneal dialysis; pmp, per million population.
great capability for studying real-world uncertainty

AJKD Vol 77 | Iss 6 | June 2021 893


800
Actual age- and sex-standardized prevalence
Predicted prevalence by Verhulst model
600
Stanadrdized prevalence (PMP)
503.33
467.74
428.34
386.03
400 342.07
297.94
255.11
200
0 0
2013 2014 2015 2016 2017 2018 2019 2020
Figure 3. Prediction curve of prevalence and number of dialysis patients in China from 2013 to 2025. Calculation was by the direct
method using 2010 national population census data. The numbers and columns in the figure show predicted prevalence and pre-
dicted number of patients, respectively, using the gray Verhulst model. Abbreviation: pmp, per million population.
problems with limited data and small sample size. The
present analysis showed that the number of dialysis pa-
tients was expected to increase to nearly 0.9 million by
2025. If kidney transplant recipients and the dispropor-
tionate costs were taken into consideration, the burden of
patients with kidney failure would undoubtedly be enor-
mous. In a systematic review, the worldwide use of KRT
will more than double from 2.62 million in 2010 to 5.44
million in 2030, with the most growth in Asia.1 A recent
study found that the prevalence of dialysis patients in
Nanjing, China, would increase by 1.95% annually, pre-
dicting prevalence rates of 1,373 pmp in 2020 and 1,505
pmp in 2025 and reflecting a growth rate similar to that in
the present study.30
The increase in the prevalence of dialysis patients may
be driven by 2 factors. First, China has experienced a
dramatic shift in diet and lifestyle in recent decades. These
changes often bring with them rapid increases in the
prevalence of obesity, diabetes, and hypertension, which
substantially affect the prevalence and pattern of CKD.4 The
spectrum of CKD in China has been evolving to mirror that
of developed countries.4 Similar patterns of diabetes as the
major cause have been reported among incident dialysis
patients in China, which may partly contribute to the in-
crease of patients with kidney failure.5 Presently, most
patients with diabetes, especially in rural areas, are not
894
Yang et al
1200000
629.67
617.55
602.46
583.90
561.40
900000
534.60
Predicted number of patients
600000
300000
2021 2022 2023 2024 2025
referred to nephrologists until they have reached kidney
failure.31,32 In this study, we based our diagnoses and
comorbidities on claims for treatment, which may lead to
underestimation of the prevalence of diabetes. This is
because kidney failure may result in delayed insulin
clearance and mitigate the need for any treatment in pa-
tients with type 2 diabetes.33 Moreover, our study found
that the burden of other comorbidities of dialysis patients
was also substantial. These findings suggest the importance
of early prevention, timely referral, and optimal manage-
ment for patients with CKD. Decisions regarding the
initiation of dialysis should also be based on careful clinical
evaluation, considering that the optimal dialysis initiation
time still remains unknown.34
In addition, health system reform in China may be
another driver. The lower rates of dialysis in the past
largely stem from unaffordable health care associated with
out-of-pocket expenses, and inequalities in access to
medical resources.25 However, universal health insurance
coverage, proposed by the China’s health system reforms
in 2009,35 has greatly improved the affordability and
accessibility of dialysis. The critical illness insurance pro-
gram was initially launched in 2012 and implemented
nationally in 2015.35 Since then, kidney failure has
become a so-called special disease that can be highly
reimbursed, with a reimbursement rate of up to 90%.22
AJKD Vol 77 | Iss 6 | June 2021
Yang et al
This also drives claims to some extent. The policy of
allowing independent HD centers (ie, not affiliated with a
public hospital) was launched in 2016, which improved
the accessibility of KRT in less well-developed areas.5
Therefore, to minimize the impact of the improved
accessibility, the focus should be on the more recent time
frame during our study period to understand the actual
burden of kidney failure in China.
To control the rising burden of diabetes, hypertension,
and CKD, the State Council issued the “Healthy China
2030” planning outline in 2016,36 and the “Medium- to
Long-Term Plan (2017-2025) on the Prevention and
Treatment of Chronic Diseases” in 2017.37 Concrete
measures covering various aspects of prevention of major
noncommunicable diseases, including kidney failure,
were proposed to raise the emphasis on non-
communicable diseases to the national level. Given the
rapid upward trend of kidney failure and its impact on
the health care system, the challenges regarding surveil-
lance and management of CKD need to be addressed
urgently. For countries with limited resources, capturing
data from health information systems or existing multiple
sources is cost-effective.38
Our study has the advantages of using a large national
claims database and predicting trends based on limited
information. However, there are certain limitations that
deserve mention. First, claims data have potential errors in
classification of patients. Despite this, we used claims re-
cords to ascertain dialysis-related diagnoses. Second, the
crude proportions of comorbidities identified based on
claims records may be an imprecise reflection of the
treatment and management of these comorbidities and
interpretation of these trends should be cautious. We did
not evaluate the distribution of causes of kidney failure
because this may lead to an inaccurate understanding of
the etiology in the case of incomplete diagnostic data.
Third, population selection bias caused by the sampling
design of the CHIRA database and estimates of visiting
rates of the insured population may lead to the underes-
timation of prevalence rates. Given the geographical dis-
parities in CKD prevalence in China,3 the first stage of the
sampling procedure may also have an impact on the
generalizability of study findings. The prevalence in
different geographical areas was not calculated because the
sampling framework might bring bias to regional preva-
lence estimates. Fourth, our study did not include the
population insured by NCMS. Despite the wide coverage of
UEBMI and URBMI in China, the extrapolation of study
findings may be affected by the population representa-
tiveness, because the proportion of dialysis use among
rural patients may be relatively low due to the limited
access to dialysis care.25 Finally, although the gray model
showed an advantage in time series predictions with small
sample size, only 5-year data were collected to develop the
prediction model and other potential confounding factors
were not included. Additionally, we did not consider
AJKD Vol 77 | Iss 6 | June 2021
future demographic changes and the trends in risk factors
in the general population, such as obesity, diabetes, and
aging, which might also result in an underestimation of
prevalence rates.
Kidney failure has become an important public
health problem in China that merits attention. With
acceleration of population aging and rapidly rising
diabetes and hypertension, which are known to be
drivers of CKD, the prevalence of dialysis patients will
increase in the future. The CK-NET SSCD has made an
attempt to establish a national surveillance system for
dialysis by leveraging big data in medical claims sys-
tems, which could give insights into the development
of surveillance strategies for kidney diseases. Future
studies should include more variables and data sources,
and further verify the accuracy and applicability of CK-
NET SSCD.
Supplementary Material
Supplementary File (PDF)
Table S1: ICD-10 codes for identifying dialysis patients.
Table S2: ICD-10 codes for identifying patients with acute kidney
injury.
Table S3: Formulas and parameters of the gray Verhulst model.
Table S4: Accuracy evaluation criteria of gray Verhulst model.
Article Information
China Kidney Disease Network (CK-NET) Work Group Members:
Hong Chu, MD, Xinwei Deng, MD, Lanxia Gan, MS, Yifang Jiang,
MD, Lili Liu, MD, Jianyan Long, MS, Ying Shi, MS, Wen Tang, MD,
Song Wang, MD, Dongliang Zhang, MD, Xinju Zhao, MD, Liren
Zheng, MD, and Zhiye Zhou, MS.
Authors’ Full Names and Academic Degrees: Chao Yang, MS,
Zhao Yang, MS, Jinwei Wang, PhD, Huai-Yu Wang, MD, PhD,
Zaiming Su, MS, Rui Chen, MD, Xiaoyu Sun, MS, Bixia Gao, MD,
Fang Wang, MD, Luxia Zhang, MD, MPH, Bin Jiang, PhD, and
Ming-Hui Zhao, MD.
Authors’ Affiliations: Renal Division, Department of Medicine,
Peking University First Hospital, Peking University Institute of
Nephrology, Beijing, China (CY, JW, RC, BG, FW, LZM-HZ);
Scientific Research Department, Peking University First Hospital,
Beijing, China (ZY); National Institute of Health Data Science at
Peking University, Beijing, China (H-YW, ZS, XS, LZ); School of
Pharmaceutical Sciences, Peking University, Beijing, China (BJ);
and Peking-Tsinghua Center for Life Science, Beijing, China (M-HZ).
Address for Correspondence: Fang Wang, MD, Renal Division,
Department of Medicine, Peking University First Hospital, Peking
University Institute of Nephrology, 8 Xishiku Street, Xicheng
District, Beijing 100034, China (email: wangfang@bjmu.edu.cn) or
Luxia Zhang, MD, MPH, Renal Division, Department of Medicine,
Peking University First Hospital, Peking University Institute of
Nephrology, 8 Xishiku Street, Xicheng District, Beijing 100034, China
(email: zhanglx@bjmu.edu.cn).
Authors’ Contributions: Research idea and study design: CY, FW,
LZ; data acquisition: LZ; data analysis/interpretation: CY, ZY, JW,
H-YW, ZS, XS, BG; statistical analysis: CY, RC; supervision or
mentorship: FW, LZ, BJ, M-HZ. Each author contributed important
intellectual content during manuscript drafting or revision and
agrees to be personally accountable for the individual’s own
895

contributions and to ensure that questions pertaining to the
accuracy or integrity of any portion of the work, even one in which
the author was not directly involved, are appropriately investigated
and resolved, including with documentation in the literature if
appropriate.
Support: This study was supported by National Natural Science
Foundation of China grants 91846101, 81771938, 81301296,
and 81900665; Beijing Nova Programme Interdisciplinary
Cooperation Project grant Z191100001119008; National Key
Research and Development Program of the Ministry of Science
and Technology of China grants 2016YFC1305405 and
2019YFC2005000; University of Michigan Health System–Peking
University Health Science Center Joint Institute for Translational
and Clinical Research grants BMU20160466, BMU2018JI012,
and BMU2019JI005; PKU-Baidu Fund award 2019BD017; and
Peking University grants BMU2018MX020 and PKU2017LCX05.
The funders of this study had no role in the design of this study;
collection, analysis, or interpretation of data; writing the report; or
the decision to submit this report for publication.
Financial Disclosure: Dr Zhang has received research funding from
AstraZeneca. The remaining authors declare that they have no
relevant financial interests.
Acknowledgements: The authors thank the China Health Insurance
Research Association for the support of this study.
Peer Review: Received May 14, 2020. Evaluated by 2 external peer
reviewers, with direct editorial input from a Statistics/Methods
Editor, an Associate Editor, and the Editor-in-Chief. Accepted in
revised form November 14, 2020.
References
1. Liyanage T, Ninomiya T, Jha V, et al. Worldwide access to
treatment for end-stage kidney disease: a systematic review.
Lancet. 2015;385(9981):1975-1982.
2. Bello AK, Levin A, Lunney M, et al. Status of care for end stage
kidney disease in countries and regions worldwide: interna-
tional cross sectional survey. BMJ. 2019;367:l5873.
3. Zhang L, Wang F, Wang L, et al. Prevalence of chronic kidney
disease in China: a cross–sectional survey. Lancet.
2012;379(9818):815-822.
4. Zhang L, Long J, Jiang W, et al. Trends in chronic kidney dis-
ease in China. N Engl J Med. 2016;375(9):905-906.
5. Yang C, Wang H, Zhao X, et al. CKD in China: evolving
spectrum and public health implications. Am J Kidney Dis.
2020;76(2):258-264.
6. National Health Commission. National report on the services,
quality and safety in medical care system, 2018. Beijing:
Scientific and Technical Documentation Press; 2019.
7. Beijing Hemodialysis Quality Control and Improvement Center.
The Annual Data Report of Beijing Hemodialysis Registry,
2012. Chin J Blood Purif. 2012;11(z1):42.
8. Zhang W, Qian J. Current status of dialysis therapy in Shanghai
(results from Shanghai Renal Registry, 2011). Chin J Blood
Purif. 2012;11(5):233-236.
9. Yamagata K, Yagisawa T, Nakai S, et al. Prevalence and inci-
dence of chronic kidney disease stage G5 in Japan. Clin Exp
Nephrol. 2015;19(1):54-64.
10. National Bureau of Statistics of China. Statistical Communiqu e
of the People’s Republic of China on the 2018 National Eco-
nomic and Social Development. 2019. Accessed April 22,
2020. http://www.stats.gov.cn/english/PressRelease/201
902/t20190228_1651335.html
11. Saran R, Robinson B, Abbott KC, et al. US renal data system
2019 annual data report: epidemiology of kidney disease in the
United States. Am J Kidney Dis. 2020;75(1)(suppl 1):A6-A7.
896
Yang et al
12. Zhang L, Wang H, Long J, et al. China kidney disease network
(CK-NET) 2014 annual data report. Am J Kidney Dis.
2017;69(6s2):S1-S149.
13. Zhang L, Zhao M-H, Zuo L, et al. China kidney disease network
(CK-NET) 2015 annual data report. Kidney Int Suppl.
2019;9(1):e1-e81.
14. National Bureau of Statistics of China. China Statistics Year-
book 2018. 2018. Accessed March 19, 2020. http://www.
stats.gov.cn/tjsj/ndsj/2018/indexeh.htm
15. National Bureau of Statistics of China. Statistical Communique
of the People’s Republic of China on the 2017 National
Economic and Social Development. 2018. Accessed March
19, 2020. http://www.stats.gov.cn/english/PressRelease/2018
02/t20180228_1585666.html
16. Lu Y, Yan H, Zhang L, Liu J. A Comparative study on the pre-
diction of occupational diseases in China with hybrid algorithm
combing models. Comput Math Methods Med. 2019;2019:
8159506.
17. Feng ZX, Lu SS, Zhang WH, Zhang NN. Combined prediction
model of death toll for road traffic accidents based on inde-
pendent and dependent variables. Comput Intell Neurosci.
2014;2014:103196.
18. Shen X, Ou L, Chen X, Zhang X, Tan X. The application of the
gray disaster model to forecast epidemic peaks of typhoid and
paratyphoid fever in China. PLoS One. 2013;8(4):e60601.
19. Zhao YF, Shou MH, Wang ZX. Prediction of the number of
patients infected with COVID-19 based on rolling gray Verhulst
models. Int J Environ Res Public Health. 2020;17(12):4582.
20. Zhang X, Zhang L, Zhang Y, Liao Z, Song J. Predicting trend of
early childhood caries in mainland China: a combined meta-
analytic and mathematical modelling approach based on
epidemiological surveys. Sci Rep. 2017;7(1):6507.
21. Tang SCW, Yu X, Chen HC, et al. Dialysis care and dialysis
funding in Asia. Am J Kidney Dis. 2020;75(5):772-781.
22. Zhang L, Zuo L. Current burden of end-stage kidney disease
and its future trend in China. Clin Nephrol. 2016;86(13):27-28.
2016.
23. Wen CP, Cheng TY, Tsai MK, et al. All-cause mortality attrib-
utable to chronic kidney disease: a prospective cohort study
based on 462 293 adults in Taiwan. Lancet. 2008;371(9631):
2173-2182.
24. Chen CF, Chen FA, Lee TL, et al. Current status of dialysis and
vascular access in Taiwan. J Vasc Access. 2019;20(4):368-373.
25. Liu ZH. Nephrology in China. Nat Rev Nephrol. 2013;9(9):523-
528.
26. Ricardo AC, Yang W, Sha D, et al. Sex-related disparities in
CKD progression. J Am Soc Nephrol. 2019;30(1):137-146.
27. Menon V, Wang X, Sarnak MJ, et al. Long-term outcomes in
nondiabetic chronic kidney disease. Kidney Int. 2008;73(11):
1310-1315.
28. Carrero JJ, Hecking M, Chesnaye NC, Jager KJ. Sex and
gender disparities in the epidemiology and outcomes of chronic
kidney disease. Nat Rev Nephrol. 2018;14(3):151-164.
29. Deng J. Introduction to gray system theory. J Gray Syst.
1989;1:1-24.
30. Sun L, Zou LX, Han YC, et al. Forecast of the incidence,
prevalence and burden of end-stage renal disease in Nanjing,
China to the Year 2025. BMC Nephrol. 2016;17(1):60.
31. Yang YZ, Wang JW, Wang F, et al. Incidence, development, and
prognosis of diabetic kidney disease in China: design and
methods. Chin Med J. 2017;130(2):199-202.
32. Zhang W, Gong Z, Peng X, Tang S, Bi M, Huang W. Clinical
characteristics and outcomes of rural patients with ESRD in
Guangxi, China: one dialysis center experience. Int Urol
Nephrol. 2010;42(1):195-204.
AJKD Vol 77 | Iss 6 | June 2021
Yang et al
33. Pina AF, Borges DO, Meneses MJ, et al. Insulin: trigger and
target of renal functions. Front Cell Dev Biol. 2020;8:519.
34. Lin ZH, Zuo L. When to initiate renal replacement therapy: the
trend of dialysis initiation. World J Nephrol. 2015;4(5):521-
527.
35. Fang H, Eggleston K, Hanson K, Wu M. Enhancing financial
protection under China’s social health insurance to achieve
universal health coverage. BMJ. 2019;365:l2378.
36. State Council. State Council policies aim to improve people’s
lives. 2016. Accessed April 23, 2020. http://english.www.gov.
AJKD Vol 77 | Iss 6 | June 2021
cn/policies/policy_watch/2016/10/28/content_281475477184
957.htm
37. State Council. State Council issues plan to prevent chronic dis-
eases. 2017. Accessed April 23, 2020. http://english.www.gov.cn/
policies/latest_releases/2017/02/14/content_281475567482818.
htm
38. Pecoits-Filho R, Okpechi IG, Donner JA, et al. Capturing and
monitoring global differences in untreated and treated end-
stage kidney disease, kidney replacement therapy modality,
and outcomes. Kidney Int Suppl. 2020;10(1):e3-e9.
897
 Yang et al

897.e1 AJKD Vol 77 | Iss 6 | June 2021

 Original Investigation

Prevention of Urinary Stones With Hydration (PUSH):

Design and Rationale of a Clinical Trial
Charles D. Scales Jr, Alana C. Desai, Jonathan D. Harper, H. Henry Lai, Naim M. Maalouf, Peter P. Reese,
Gregory E. Tasian, Hussein R. Al-Khalidi, Ziya Kirkali, and Hunter Wessells, on behalf of the Urinary Stone
Disease Research Network

Visual Abstract online

Rationale & Objective: Although maintaining
high fluid intake is an effective low-risk
intervention for the secondary prevention of
urinary stone disease, many patients with stones
do not increase their fluid intake.
Study Design: We describe the rationale and
design of the Prevention of Urinary Stones With
Hydration (PUSH) Study, a randomized trial of a
multicomponent behavioral intervention program
to increase and maintain high fluid intake. Par-
ticipants are randomly assigned (1:1 ratio) to the
intervention or control arm. The target sample
size is 1,642 participants.
Setting & Participants: Adults and adolescents
12 years and older with a symptomatic stone
history and low urine volume are eligible. Exclu-
sion criteria include infectious or monogenic
causes of urinary stone disease and comorbid
conditions precluding increased fluid intake.
Interventions: All participants receive usual care
and a smart water bottle with smartphone appli-
cation. Participants in the intervention arm receive
a fluid intake prescription and an adaptive pro-
gram of behavioral interventions, including finan- Complete author and article
cial incentives, structured problem solving, and information (including a list
of the members of the
other automated adherence interventions. Con-
Urinary Stone Disease
trol arm participants receive guideline-based fluid Research Network)
instructions. provided before references.
Outcomes: The primary end point is recurrence Correspondence to
of a symptomatic stone during 24 months of C.D. Scales (chuck.scales@
duke.edu)
follow-up. Secondary end points include changes
in radiographic stone burden, 24-hour urine Am J Kidney Dis.
output, and urinary symptoms. 77(6):898-906. Published
online November 16, 2020.
Limitations: Periodic 24-hour urine volumes may doi: 10.1053/
not fully reflect daily behavior. j.ajkd.2020.09.016
© 2020 by the National
Conclusions: With its highly novel features, the Kidney Foundation, Inc.
PUSH Study will address an important health Published by Elsevier Inc.
care problem. All rights reserved.
Funding: National Institute of Diabetes and
Digestive and Kidney Diseases.
Trial Registration: Registered at Clinical-
Trials.gov with study number NCT03244189.

U rinary stone disease (USD) is an important medical,

scientific, and public health problem. The prevalence
of USD in the United States has nearly doubled in the past
15 years,1-3 and health care use for treating these patients
has increased in parallel.4,5 USD affects 1 in 11 people in
the United States, a prevalence similar to that of dia-
betes.1,2 Estimates from the Urologic Diseases in America
project suggest that USD is one of the most expensive
urologic conditions.4
High fluid intake is a leading intervention for the sec-
ondary prevention of USD because it is low risk, inex-
pensive, and effective. Current guidelines agree on one
fundamental lifestyle change for USD prevention: main-
taining oral fluid intake sufficient to cause urine output
(UOP) of at least 2 to 2.5 L per day.6,7 Among adults with
USD who receive counseling in a stone clinic, the mean
increase in 24-hour UOP is only 300 mL per day,8 sug-
gesting that adherence to increased fluid intake recom-
mendations is low even when the consequences of
nonadherence are severe. Barriers to maintaining fluid
intake for patients with USD are common, including dis-
counting the probability of a future stone event and
forgetfulness. In addition, higher fluid intake may pose
specific challenges to patients with lower urinary tract
symptoms or overactive bladder, restricted access to fluids,
or limited bathroom access. These patients may require
tailored solutions to overcome their specific barriers to
increasing fluid intake.9,10
Substantial evidence exists for the efficacy of adherence
interventions grounded in behavioral economics. Appro-
priately structured financial incentives can help overcome
the present bias by helping patients maintain a mental
connection between a current behavior, such as fluid
intake, and a future event with uncertain probability, such
as a recurrent stone episode. Financial incentives have been
shown to improve smoking cessation,11-14 exercise,15-18
substance abuse management,19-21 and medication
adherence.22-24 The Prevention of Urinary Stones With
Hydration (PUSH) Study is a randomized controlled trial
that will test a theory-based program of behavioral in-
terventions, including financial incentives, directed toward
increasing and maintaining a high fluid intake to reduce
USD recurrence.
Methods
Study Design
PUSH (ClinicalTrials.gov identifier NCT03244189) is a 2-
arm, multicenter, randomized controlled clinical trial that
incorporates pragmatic features, an adaptable behavioral

898 AJKD Vol 77 | Iss 6 | June 2021

Scales et al
PLAIN-LANGUAGE SUMMARY
Increasing fluid intake in urinary stone formers with
low urine volume is essential to prevent stone recur-
rence but difficult to achieve in clinical practice. The
Prevention of Urinary Stones With Hydration (PUSH)
Study is a randomized clinical trial that uses a smart
water bottle to track fluid intake and examines whether
a combination of behavioral interventions (including a
fluid prescription, financial incentives, and coaching)
improves adherence to higher fluid intake and reduces
stone recurrence. We describe the rationale and design
of this large multicenter randomized trial currently
enrolling 1,642 stone formers, with outcomes
including symptomatic stone recurrence, asymptomatic
(radiographic) stone disease progression, and urinary
parameters associated with disease.
intervention, patient choice, individualized fluid pre-
scriptions, and automated monitoring of fluid intake
through a commercially available wireless smart water
bottle that communicates with a smartphone application
(Hidrate Spark; Hidrate, Inc). The study will enroll 1,642
participants to be followed up for 24 months. Both
intervention and control arm participants receive a smart
water bottle capable of tracking fluid consumption; usual
care, including guideline-based recommendations of
adequate fluid intake; periodic 24-hour urine collections;
and baseline and end-of-study imaging (Fig 1; Table 1).
After enrollment, most study procedures and interactions
between study participants and research personnel will be
conducted remotely, with in-person visits occurring only
at the beginning and end of the study. Participants will
undergo imaging (low-dose computed tomography for
adults and ultrasonography for adolescents) at baseline and
24 months to assess changes in stone burden.
Aims and Hypotheses
The primary aim of PUSH is to assess the effect of a
behavioral intervention program on the primary end point
of symptomatic stone recurrence at 2 years (Table 2). We
hypothesize that a multicomponent behavioral interven-
tion program will lead to sustained high fluid intake and
UOP, resulting in decreased symptomatic stone re-
currences compared with usual care.
Secondary end points include the formation of asymp-
tomatic new stones on imaging, growth of stones present
at baseline (by ≥2 mm in any dimension) on imaging,
change in 24-hour UOP versus baseline, and change in
lower urinary tract symptoms. A composite outcome of
symptomatic stone recurrence, formation of new asymp-
tomatic stones, or stone growth on imaging will also be
compared between study arms. Finally, costs of study in-
terventions and treatments for USD during the PUSH
follow-up period will be assessed.
AJKD Vol 77 | Iss 6 | June 2021
Measures
Daily fluid intake is monitored through the use of a
wireless-enabled smart water bottle and data transmitted
daily to the web-based platform for implementation of
behavioral economic components at Way to Health
(University of Pennsylvania, Philadelphia, PA). Way to
Health provides a secure user-friendly approach to moni-
toring participation, delivering feedback/communications
(email, text, or telephone messages using interactive voice
response), and administering financial incentives.25
Study participants undergo baseline and end-of-study
imaging (at the conclusion of a 24-month follow-up;
low-dose computed tomography for participants aged ≥18
years and renal ultrasonography for those aged <18 years at
enrollment). In addition, imaging that occurs for clinical
care during the follow-up period may also be reviewed for
outcome events, such as formation of a new stone. A central
committee, blinded to the randomization allocation, will
adjudicate the primary and secondary (imaging) end points.
UOP is assessed using 24-hour urine collections at
baseline and every 6 months during the 24-month study
period. Participants are queried every 3 months electron-
ically or by telephone regarding the occurrence of stone
events (primary outcome), such as stone passage or pro-
cedural intervention for a stone. Supporting details from a
review of medical records are obtained using standard
procedures in the protocol. Lower urinary tract symptoms
in adult participants are assessed at baseline and every 6
months thereafter using the Comprehensive Assessment of
Self-reported Urinary Symptoms (CASUS) questionnaire.26
Setting and Participants
The study is supported by the National Institute of Diabetes
and Digestive and Kidney Diseases (NIDDK) and con-
ducted at Urinary Stone Disease Research Network clinical
centers consisting of the University of Pennsylvania Health
System and Children’s Hospital of Philadelphia, University
of Texas Southwestern Medical Center, University of
Washington, and Washington University in St. Louis. The
Cleveland Clinic Foundation and the Mayo Clinic were
added as recruiting centers in 2019. Duke University serves
as the Scientific Data Research Center. The institutional
review boards of each institution reviewed and approved
the study; participants provide informed consent/assent to
participate. Participants receive usual care for their USD
clinical centers or at referring institutions.
Eligibility criteria include age 12 years or older, a
symptomatic urinary stone (stone passage or surgery)
within 3 years before inclusion or a symptomatic stone
event within 5 years if a new stone has been detected on
imaging, low 24-hour UOP, and access to a device
(smartphone or tablet) compatible with the smart water
bottle. For this study, low UOP is defined as <2.0 L per day
for participants 18 years and older or those aged 12 to 17
years weighing at least 75 kg. For participants aged 12
to 17 years weighing less than 75 kg, the low UOP
899
 Scales et al

Study Population

Low-dose CT scan (ultrasound if age < 18)

RANDOMIZATION

Control Intervention

Wireless Bottle
Usual Urological Care

Fluid “prescription”

24-h urine collection Q6 months (by mail)

Questionnaires Q3 months (online)

Financial Incentive (loss aversion, $1.50/day)

Months 1-6
Succeed Fail Booster (may repeat x1)
Induction

Financial Incentive Months 7-12

Financial Incentive Fail
+ Fluid Coach (SPS) Consolidation

Tapering financial incentive Months 13-18

+ Low-touch intervention Tapering

Months 19-24
Low-touch intervention
Maintenance

Low-dose CT scan (ultrasound if age < 18)

Figure 1. Prevention of Urinary Stones With Hydration (PUSH) Study schema. Abbreviations: CT, computed tomographic; SPS,
structured problem solving.

requirement is <25 mL per kilogram of body weight per video regarding the rationale for increased fluid intake as
day. Exclusion criteria and rationale are listed in Table 3. a preventive measure in USD. Participants then receive an
individualized fluid prescription, which is the additional
Randomization and Blinding
volume of fluid intake needed to maintain the target UOP.
The UOP target is ≥2.5 L per day for participants aged 18
Study participants are randomly assigned in a 1:1 ratio to 1
years and older and those aged 12 to 17 years weighing at
of 2 parallel study groups, designated as intervention or
least 75 kg. For participants aged 12 to 17 years weigh-
control. Randomization is stratified within clinical centers
ing less than 75 kg, the UOP target is ≥30 mL/kg per day.
by adult (aged ≥18 years) versus adolescent (aged 12-17
The prescription is determined using the following equa-
years) and first-time versus recurrent stone former.
tion: initial fluid prescription = 1.2 × (2.5 L − median
Randomization is conducted at the Clinical Center level in
UOP on all 24-hour urine collections in the 12 months
blocks of variable size to minimize potential confounders
before enrollment). The 20% factor was intended to
related to differences in site-based care practices.
mitigate, at least partially, insensible losses.27 Participants
Randomization is performed using an internet-accessible
are instructed to consume the prescribed additional fluid
integrated randomization module in the trial electronic
from the smart water bottle. In addition to the fluid pre-
data capture system to provide concealment of allocation.
scription, they are instructed to drink the same volume of
It was not feasible to blind participants or study co-
fluids they consumed before enrollment using other fluid
ordinators; however, all investigators, clinicians, and
containers. The fluid prescription is modified as needed
outcome adjudicators are blinded to randomization allo-
based on the 24-hour urine volume measurement obtained
cation. In addition, the statistical team responsible for
approximately 2 weeks after randomization. To reduce the
analysis of trial end points will be blinded to allocation.
risk for unblinding treatment providers, adjustments to the
fluid prescription are communicated by the study coordi-
Intervention Arm nator. Participants also receive standard clinical care ac-
Immediately after randomization, participants in the cording to American Urological Association guidelines,
intervention arm receive education through a web-based including general guidance regarding dietary modification.

900 AJKD Vol 77 | Iss 6 | June 2021

Scales et al

Table 1. Overview of Intervention and Control Arms in the PUSH Study

Control Intervention
Platform • Participants provided a wireless- • Participants provided a wireless-enabled smart water
enabled smart water bottle bottle and taught how to use it
• Participants are allowed but not • Participants must sync water bottle with their mobile
required to use or sync the water device daily
bottle with their mobile device
Fluid goal Usual care (sufficient fluid intake to 120% of baseline UOP deficit by the water bottle, with
achieve UOP ≥ 2.5 L/d or 30 mL/kg/ modification possible at wk 2
d if <18 y and <75 kg)
UOP 24-h urine volume check at 2 wk and 24-h urine volume check at 2 wk and 6, 12, 18, and 24
6, 12, 18, and 24 mo mo
Behavioral None Months 1-6 (induction):
Intervention • Daily financial incentive for success at meeting fluid
Program intake goal via water bottle (incentive possible 100% of
days)
• SPS triggered if nonadherence (ie, fluid goal ach-
ieved <80% of days over 14 da for two 14-d
assessment periods)
• SPS “booster” (intensified counseling) triggered if
nonadherence (ie, fluid goal achieved <80% of days
over 14 da for two 14-d assessment periods after initial
SPS)
Months 7-12 (consolidation):
• Daily financial incentive for success at meeting fluid
intake goal via water bottle (incentive possible 80% of
days in 1-mo period)
• SPS triggered if nonadherence (ie, fluid goal was
achieved <80% of days over 14 da for two 14-d
assessment periods)
• SPS “booster” (intensified counseling) triggered if
nonadherence (ie, fluid goal was achieved <80% of
days over 14 da for two 14-d assessment periods
following initial SPS)
Months 13-18 (tapering):
• Gradual tapering of daily financial incentive (75% of
days at mo 13 and 15% of days at mo 18)
• Self-selected/low-touch intervention(s): text/telephone
reminders, commitment contract, social intervention
(eg, team)
Months 19-24 (maintenance):
• Self-selected/low-touch intervention(s): text/telephone
reminders, commitment contract, social intervention
(eg, team), feedback on adherence
Stone-directed Per primary stone treating provider Per primary stone treating provider
pharmacologic
or urologic care
Abbreviations: PUSH, Prevention of Urinary Stones with Hydration; SPS, structured problem solving; UOP, urine output.
a
Greater than 3 days in which fluid goal was not met during 14-day period.

Adaptable Behavioral Intervention
The intervention arm also includes a behavioral program
that consists of financial incentives and, if this intervention
is unsuccessful, structured problem solving (SPS) to
maintain the recommended fluid intake (Fig 1), reflect the
responsive nature of clinical care, and help participants
change behavior and then sustain the change. The inter-
vention period is divided into 6-month phases of induc-
tion, consolidation, tapering, and maintenance. The
induction phase consists of a daily financial incentive
($1.50) for meeting the fluid intake prescription. This
daily financial incentive is delivered when the recom-
mended volume of fluid intake is consumed through and
recorded by the smart water bottle. During induction, the
intensity of the intervention can be increased for partici-
pants who do not consistently meet daily fluid intake goals
(defined as at least 80% of days over a 2-week period) by
adding SPS to help participants develop feasible solutions
to overcoming personal barriers to maintaining the pre-
scribed fluid intake. The consolidation phase also includes
financial incentives and SPS (when indicated by the pro-
tocol), but the frequency of reward is slightly lower
(Table 1). The tapering phase comprises months 13 to 18,
during which SPS is withdrawn, financial incentives are
sequentially tapered, and “low-touch” and low-cost in-
terventions (support partners and regular feedback about
adherence) selected by the participant are started to help
sustain new habits of fluid consumption. In the final 6-

AJKD Vol 77 | Iss 6 | June 2021 901


Table 2. Primary and Secondary End Points in PUSH Study
End Point
Primary end point
Symptomatic stone recurrence
Secondary end points
1. Asymptomatic formation of new stone
2. Increased stone size by ≥2 mm (any dimension)
3. Composite of primary end point or secondary end
points 1 and 2
4. 24-h urine total volume
5. Presence of lower urinary tract symptoms
Abbreviations: CASUS, Comprehensive Assessment of Self-reported Urinary Symptoms; PUSH, Prevention of Urinary Stones with Hydration.
month maintenance phase, participants have access to only
low-touch interventions.
Structured Problem Solving
SPS is an intervention facilitated by a health coach in which
the solutions to maintaining high fluid intake are identified
by and personalized for each participant. Participants
randomly assigned to the intervention arm who meet
criteria for receiving SPS initially meet with the health
coach. The introductory session starts with education
about kidneys and USD, the importance of fluid intake,
and the prescribed fluid intake regimen. The rest of the
introductory session is devoted to identifying barriers and
feasible solutions to maintaining high fluid intake (Item
S1). Based on smart water bottle data, participants also
receive daily automated feedback of their fluid intake
through Way to Health.
During months 1 to 12, participants who have received
SPS but still have fluid adherence less than 80% of days
over two 14-day assessment periods (>3 days in which
fluid goal was not met during 14-day period) receive up to
2 “booster” sessions, which consist of root cause analysis
to identify reasons for failure of the initial approach chosen
to maintain high fluid intake. A new solution is chosen and
developed by the participants by repeating the SPS process.
Low-Touch Interventions
If after 2 SPS booster sessions adherence to daily fluid goal
is still less than 80% for subsequent assessment periods,
the participant is offered the opportunity to select 1 or
more low-touch interventions to continue throughout the
rest of the study period. Examples include activating
reminder features on the smart water bottle application,
receiving a weekly summary of intake, engaging a support
partner, and the addition of gamification elements (levels
to denote status by consistent adherence). The rationale is
that application of an intensive intervention that has not
resulted in behavior change is unlikely to do so with
continued application and may decrease participant
retention. Regardless of whether the participant in the
intervention arm has received SPS, they continue to be
902
Scales et al
Definition/Assessment
Stone passage; procedure for symptomatic stone;
urologic procedure for removal of asymptomatic stone
Detected on imaging
Detected on imaging
Not applicable
Home 24-h urine collection
CASUS questionnaire
eligible for financial incentives according to the study
schedule.
Control Arm
Participants in the control arm receive standard care in-
structions according to American Urological Association
guidelines to increase overall fluid consumption to achieve
UOP of ≥2.5 L per day7 or the weight-based fluid-intake
goal if they weigh less than 75 kg, as well as general di-
etary modification. They also receive the same wireless-
enabled smart water bottle with the capability to self-
monitor fluid intake as do the intervention arm partici-
pants, although use of the bottle is optional in the control
arm. Approximately 2 weeks after randomization, a 24-
hour urine volume measurement is obtained to provide
control arm participants with initial feedback as to whether
they are meeting the UOP goal. Participants in the control
arm do not receive an individualized fluid prescription or
financial incentives but receive monetary compensation for
completion of study activities (urine collections, study-
related imaging, and study completion).
Statistical Analysis
The primary analysis will be conducted using intention-to-
treat principles. The primary end point will be analyzed
using time-to-event methodology, and survival curves will
be compared using log-rank test. For participants who are
lost to follow-up or those who drop out of the study
without experiencing the primary end point, their time-to-
event measure will be censored on their last contact date.
Safety data will be collected on all randomly assigned
participants. All statistical analyses will be tested at a 2-
sided nominal significance level of 0.05. If missing data
constitute <2% of observations, the analysis will use single
imputation (mean for continuous variables and mode for
categorical variables). If missing data are >2% of obser-
vations, the approach to missing data will depend on
checks of the missing-at-random assumption.
The primary outcome of a symptomatic stone event is
defined as stone passage, symptomatic stone requiring
procedural intervention, or any asymptomatic stone
AJKD Vol 77 | Iss 6 | June 2021
Scales et al
Table 3. Study Exclusion Criteria and Rationale
Criterion
Spinal cord injury
Undergoing active cancer treatment
Known infectious or monogenic cause
History or presence of hyponatremia
Comorbid conditions precluding increased
fluid intake
Comorbid condition with GI fluid loss
Pregnant or planning pregnancy
Kidney transplant recipient
History of recurrent UTI (>3/y)
Medication use increasing stone risk
Less than 2-y life expectancy
Non–English speaker
History of SIADH
Anatomic urologic abnormality (eg, ileal
conduit)
Psychiatric condition impairing study
adherence
Vulnerable population
Unable to participate per investigator
Abbreviations: GI, gastrointestinal; SIADH, syndrome of inappropriate antidiuretic hormone secretion; UTI, urinary tract infection.
undergoing procedural intervention. A prespecified sensi-
tivity analysis of the definition of symptomatic stone event
will include stone passage, symptomatic stones requiring
procedural intervention, and asymptomatic stones under-
going procedural intervention that meet at least 1 of the
following criteria: stone size at least 4 mm in any
dimension, mobile stone, hematuria, or recurrent urinary
tract infection. An additional prespecified sensitivity anal-
ysis will exclude any symptomatic stone events that occur
within 30 days of randomization.
Secondary end points include asymptomatic stone for-
mation, stone growth on imaging, urinary symptom
burden, and a composite outcome of symptomatic stone
event, asymptomatic stone formation, or stone growth
assessed at study conclusion. These outcomes will be
analyzed using a logistic regression model because
asymptomatic stone formation or stone growth is subject
to interval censoring (the exact time of occurrence will be
unknown).
Other secondary outcomes, including 24-hour urine
volume, osmolality, supersaturation of stone-forming
salts, and lower urinary tract symptoms, will be
compared between the intervention and control arms us-
ing t test or Wilcoxon rank sum test based on the
assumption of normality of the data. The proportion of
participants developing symptomatic hyponatremia will be
compared in the intervention and control arms using
Pearson χ 2 test or Fisher exact test, as appropriate. All
statistical analyses will be conducted using SAS, version 9.4
or higher (SAS Institute, Inc).
AJKD Vol 77 | Iss 6 | June 2021
Rationale
Increased stone risk, impaired sensation
Potential limited life expectancy
Therapies other than high fluid intake may significantly alter disease course
Minimize risk for symptomatic hyponatremia from higher fluid intake
Physiologic barrier to adherence
Pathophysiologic barrier to adherence, difficulty achieving goal urine volume with
usual fluid intake
Higher fluid intake recommended during pregnancy and breastfeeding; bioethical
consideration
Altered renal colic sensation
Risk for infection-based stones
Stone-provoking medication is typically stopped; potential lack of hydration
effectiveness
Potential inability to complete follow-up period
Potential to affect ability to use smart water bottle app
Increased risk for symptomatic hyponatremia
Impaired/altered sensation
Potential poor adherence to follow-up
Bioethical consideration
Investigator’s discretion for protection of safety and welfare of study participants
Sample Size and Power
Power calculations were based on symptomatic recurrence
rates reported by Rule et al.28 Power calculations assumed
a 15% event rate (symptomatic stone recurrence) in the
control arm and a 30% relative risk reduction (hazard ra-
tio, 0.70) for the intervention arm. The attrition rate was
estimated at 20% with a 24-month follow-up period.
Based on these conservative estimations (see Item S1) and
using a 2-sided alpha of 0.05 and a log-rank test for time
to first symptomatic stone recurrence, 821 participants per
arm (a total of 1,642) provide 80% power to detect the
prespecified relative risk reduction. Sample size calcula-
tions were performed using PROC POWER in SAS, version
9.4. There is no planned interim analysis for efficacy or
futility.
Discussion
We designed a multicenter randomized clinical trial with a
pragmatic adaptable behavioral intervention to increase
sustained fluid intake as the primary adherence behavior.
This study is expected to provide evidence for the effec-
tiveness of a multicomponent behavioral intervention to
decrease stone recurrence. Importantly, the primary
outcome of symptomatic stone recurrence is a clinically
relevant and patient-centered outcome. The study is also
expected to contribute evidence to support the role of
increased fluid intake with subsequent increased UOP to
prevent recurrent USD, as well as radiographic disease
progression.
903

The multicomponent intervention offers an important
extension of the use of financial incentives to motivate
behavior change. As implemented in the PUSH trial, a loss-
framed financial incentive acts to counter present bias,
wherein an individual at risk for recurrent USD discounts
future risk and does not engage in preventive behavior,
such as increasing fluid intake. The loss-framed incentive
moves the adverse consequence of nonadherence into the
present, thereby counteracting an individual’s tendency to
overlook future consequences. Prior randomized trials
leveraging this principle support the efficacy of financial
incentives for smoking cessation, increased physical ac-
tivity, and weight loss.13,29,30 However, not all partici-
pants are able to achieve the desired behavior change, and
other trials of financial incentive structures have not
demonstrated efficacy.31 These variable results suggest that
adaptability of an intervention, particularly if financial
incentives do not appear to result in the desired behavior
change, may be important to achieving success.
As implemented in the PUSH trial, the financial in-
centives are intended to focus the study participant’s
attention on fluid intake. We note that the initial financial
incentives rely on an individual’s insight and self-efficacy
in identifying and overcoming specific barriers to fluid
intake. However, some participants may not achieve fluid
intake goals as a result of financial incentives alone. PUSH
was therefore designed to offer participants not able to
adopt new behaviors based on financial incentives an
additional supportive intervention, namely health behavior
coaching in the form of SPS. The SPS intervention is
designed to provide guidance in identifying each patient’s
specific barriers and solutions to increased fluid intake.
The final part of the multicomponent intervention is
grounded in the assumption that financial incentives and
adaptable implementation of SPS will permit formation of
new habits, which ideally will be sustained through the
nudge of low-touch interventions. Thus, the PUSH trial
will offer evidence regarding whether an adaptable inter-
vention built on the foundation of financial incentives can
be effective in supporting behavior change.
A second key feature of the PUSH Study is the clinically
important and patient-centered end point of symptomatic
stone recurrence. Many behavioral intervention trials focus
on surrogate end points, such as increasing step counts as a
measure of physical activity, or medication adherence,
which have validity as health behaviors but are not health
events per se. In addition, stone prevention likely requires
durable behavior change over a relatively long period
(years). In the PUSH trial, the initial interventions
(financial incentives and SPS) will taper or end well before
the end of the 24-month study follow-up period. The
PUSH trial will add evidence to help answer the important
question of durability of behavior change, as well as
whether a multicomponent program to support behavior
change can result in a reduction of a clinically relevant
health event.
904
Scales et al
Finally, the PUSH trial attempts to address another
important unanswered question in the science of
behavior change, namely that of cessation of financial
incentives. Very little evidence exists to guide cessation or
tapering of a behavioral incentive, although sudden
cessation appears deleterious.32 The financial incentive in
the PUSH trial is tapered over time, along with the
introduction of either social incentives (support partner)
or low-cost interventions (text message reminders).
Observing how adherence to increased fluid intake
changes as the nature of the incentive changes over time
will provide valuable insight into this question, particu-
larly in the context of a substantial lag between the end of
the financial incentive and SPS components and the final
outcome assessment.
If the behavioral interventions described decrease the
symptomatic stone recurrence rate in the intervention
arm, the study findings will have important implications
for the secondary prevention of a highly painful con-
dition associated with a number of long-term negative
health effects.33 The multicomponent intervention could
readily be adapted for remote delivery or in generalist
care settings. In a value-based care setting, health sys-
tems may see a strong economic rationale to reduce
recurrent USD and thus could consider investing re-
sources into a financial incentive or health coaching
program. Although some patients with specific causes of
stones would still likely require specialist care, an
effective prevention program readily delivered to the
general population of idiopathic stone formers could
substantially reduce the burden of this very common
disease.
The PUSH trial has some design limitations. Incor-
poration of 24-hour urine volume as a secondary
outcome reflects current clinical practice yet may not be
reflective of daily urine volume if participants purposely
increase fluid intake only during the 24-hour collection
period. Although it is possible for participants to meet
daily fluid goals by consuming fluids without bottle use,
they are instructed to use the bottle for their fluid goal
and for self-monitoring; they also do not receive
financial incentives if not using the bottle. Likewise, the
study relies on participants to supplement their usual
fluid intake with the prescribed amount, although the
risk for participants decreasing other sources of fluid
intake exists.
The PUSH trial evaluates a multicomponent behav-
ioral intervention to increase adherence to fluid intake as
a means to prevent recurrent USD. The study design is
anticipated to contribute to clinical knowledge regarding
lifestyle prevention of USD. In addition, this study is
expected to make an important contribution to under-
standing the impact of behavioral interventions by
evaluating not only the change in the targeted behavior,
but also the associated and patient-centered health
outcome of recurrent symptomatic stones.
AJKD Vol 77 | Iss 6 | June 2021
Scales et al
Supplementary Material
Supplementary File (PDF)
Item S1: Methods supplement.
Article Information
Urinary Stone Disease Research Network: Aside from the authors,
the following individuals were instrumental in the planning and
conduct of the PUSH Study at each of the participating institutions.
Clinical Centers: University of Pennsylvania/Children’s Hospital of
Pennsylvania, Philadelphia, PA: Co-Investigators Sandra Amaral MD,
MHS, and Janet Audrain-McGovern, PhD, and Study Coordinators
Emily Funsten, Brittney Henderson, Kristen Koepsell, and Adam
Mussell; University of Texas Southwestern Medical Center, Dallas,
TX: Co-Investigators Jodi A. Antonelli, MD, and Linda A. Baker, MD,
and Study Coordinators Joyce Obiaro, Cynthia Rangel, Martinez Hill,
and Madeline Worsham; University of Washington, Seattle, WA: Co-
Investigators Fionnuala Cormack, MD, Mathew Sorensen, MD,
Karyn Yonekawa, MD, and Study Coordinators Holly Covert, Tristan
Baxter, and Elsa Ayala; Washington University in St. Louis, St. Louis,
MO: Co-Investigators: Vincent Mellnick, MD, and Douglas Coplen,
MD, and Study Coordinators Juanita Taylor, Aleksandra Klim, and
Deborah Ksiazek. Recruiting Centers: Cleveland Clinic Foundation,
Cleveland, OH: Principal Investigator Sri Sivalingam, MD, MSc,
FRCSC, Co-Investigators Katherine Dell, MD, and Juan Calle, MD,
and Study Coordinators Paige Gotwald and Marina Markovic; Mayo
Clinic Foundation, Rochester, MN: Principal Investigator John
Lieske, MD, Co-Investigators Andrew Rule, MD, Stephen Erickson,
MD, Aaron Potrezke, MD, Andrea Ferrero, PhD, and David Sas, DO,
and Study Coordinators Angela Waits and Courtney Lenort;
Scientific Data Research Center: Duke Clinical Research Institute,
Duke University, Durham, NC: Co-Investigators Kevin Weinfurt, PhD,
and Hayden Bosworth, MD, Statistician Honqiu Yang, PhD, Project
Lead Laura Johnson, Lead CRA Sharon Settles, CRA Angela
Venetta, and Data Manager Omar Thompson.
Authors’ Full Names and Academic Degrees: Charles D. Scales
Jr, MD, MSHS, Alana C. Desai, MD, Jonathan D. Harper, MD, H.
Henry Lai, MD, Naim M. Maalouf, MD, Peter P. Reese, MD,
MSCE, Gregory E. Tasian, MD, MSc, MSCE, Hussein R. Al-
Khalidi, PhD, Ziya Kirkali, MD, and Hunter Wessells, MD.
Author’s Affiliations: Urologic Surgery and Population Health
Science, Duke Surgical Center for Outcomes Research, Duke
Clinical Research Institute, Duke University School of Medicine,
Durham, NC (CDS); Division of Urologic Surgery, Department of
Surgery, Washington University School of Medicine, St Louis, MO
(ACD, HHL); Department of Urology, University of Washington
School of Medicine, Seattle, WA (JDH, HW); Department of
Internal Medicine and Charles and Jane Pak Center for Mineral
Metabolism and Clinical Research, University of Texas
Southwestern Medical Center, Dallas, TX (NMM); Renal-Electrolyte
and Hypertension Division (PPR) and Department of Biostatistics,
Epidemiology, and Informatics (PPR, GET), Perelman School of
Medicine at the University of Pennsylvania; Division of Pediatric
Urology, Department of Surgery, The Children’s Hospital of
Philadelphia, Philadelphia, PA (GET); Department of Biostatistics
and Bioinformatics, Duke Clinical Research Institute, Duke
University School of Medicine, Durham, NC (HRA-K); and National
Institute of Diabetes and Digestive and Kidney Diseases,
Bethesda, MD (ZK).
Address for Correspondence: Charles D. Scales Jr, MD, MSHS,
Urologic Surgery and Population Health Science, Duke Clinical
Research Institute, Duke University School of Medicine, Durham,
NC 27710. Email: chuck.scales@duke.edu
Authors’ Contributions: Research idea and study design: CDS,
ACD, JDH, HHL, NMM, PPR, GET, HRA, ZK, HW; data
AJKD Vol 77 | Iss 6 | June 2021
acquisition: ACD, JDH, HHL, NMM, PPR, GET, HW; data
analysis/interpretation: CDS, ACD, JDH, NMM, PPR, GET, HRA,
ZK, HW; statistical analysis: HRA. Each author contributed
important intellectual content during manuscript drafting or revision
and agrees to be personally accountable for the individual’s own
contributions and to ensure that questions pertaining to the
accuracy or integrity of any portion of the work, even one in which
the author was not directly involved, are appropriately investigated
and resolved, including with documentation in the literature if
appropriate.
Support: The PUSH Study was supported by NIDDK DK110986
(Washington University in St. Louis); DK110961 (Children’s
Hospital of Philadelphia); DK110954 (University of Washington);
DK110995 (University of Texas Southwestern Medical Center);
and DK110988 (Drs Scales and Al-Khalidi; Duke University). This
is a cooperative agreement; that means there is substantial federal
scientific or programmatic involvement in the research activities.
The NIDDK Project Scientist (ZK) is involved in the design and
development of the clinical protocol, preparation of questionnaires
and other data recording forms, coordination of research,
statistical evaluations and analyses of data, and the publication of
results. The program is overseen by an independent NIDDK
Program Official.
Financial Disclosure: Dr Scales reports receipt of funding from
Allena Pharmaceuticals for a scientific study. Dr Lai reports receipt
of research support from the National Institutes of Health and
serving as a consultant for Medtronic, Neuspera, Allergan, and
Teva. Dr Reese reports serving as an Associate Editor for AJKD,
and receipt of investigator-initiated grants from Merck and AbbVie
to the University of Pennsylvania to support research on
transplantation of hepatitis C virus–infected organs into uninfected
recipients, followed by antiviral treatment; investigator-initiated
grants from CVS Caremark and Merck to the University of
Pennsylvania to support research on medication adherence (focus:
statins); consulting with Collaborative Healthcare Research & Data
Analytics (COHRDATA) on epidemiology of medications to
improve laboratory parameters including potassium among dialysis
patients. Dr Tasian reports receipt of research funding from
Lumenis and consulting for Dicerna Pharmaceuticals, Allena
Pharmaceuticals, and Novome Biotechnologies. The remaining
authors declare that they have no relevant financial interests.
Acknowledgements: Elizabeth Cook (Duke Clinical Research
Institute) assisted with manuscript preparation and submission.
She received no additional compensation beyond her employment
at Duke University.
Peer Review: Received January 21, 2020. Evaluated by 3 external
peer reviewers and a statistician, with editorial input from an
Acting Editor-in-Chief (Editorial Board Member Kathleen Liu, MD,
PhD). Accepted in revised form September 15, 2020. The
involvement of an Acting Editor-in-Chief to handle the peer-review
and decision-making processes was to comply with AJKD’s
procedures for potential conflicts of interest for editors, described
in the Information for Authors & Journal Policies.
References
1. Scales CD Jr, Smith AC, Hanley JM, Saigal CS. Prevalence of
kidney stones in the United States. Eur Urol. 2012;62(1):160-
165.
2. Tasian GE, Ross ME, Song L, et al. Annual incidence of
nephrolithiasis among children and adults in South Carolina
from 1997 to 2012. Clin J Am Soc Nephrol. 2016;11(3):488-
496.
3. Kittanamongkolchai W, Vaughan LE, Enders FT, et al. The
changing incidence and presentation of urinary stones over 3
decades. Mayo Clin Proc. 2018;93:291-299.
905

4. Litwin MS, Saigal C, eds. Urologic Diseases in America. US
Dept of Health and Human Service. US Government Printing
Office; 2012.
5. Fwu CW, Eggers PW, Kimmel PL, Kusek JW, Kirkali Z. Emer-
gency department visits, use of imaging, and drugs for uro-
lithiasis have increased in the United States. Kidney Int.
2013;83:479-486.
6. Qaseem A, Dallas P, Forciea MA, Starkey M, Denberg TD;
Clinical Guidelines Committee of the American College of
Physicians. Dietary and pharmacologic management to prevent
recurrent nephrolithiasis in adults: a clinical practice guideline
from the American College of Physicians. Ann Intern Med.
2014;161:659-567.
7. Pearle MS, Goldfarb DS, Assimos DG, et al. Medical man-
agement of kidney stones: AUA guideline. J Urol. 2014;192:
316-324.
8. Parks JH, Goldfischer ER, Coe FL. Changes in urine volume
accomplished by physicians treating nephrolithiasis. J Urol.
2003;169:863-866.
9. Tarplin S, Monga M, Stern KL, McCauley LR, Sarkissian C,
Nguyen MM. Predictors of reporting success with increased
fluid intake among kidney stone patients. Urology. 2016;88:49-
56.
10. McCauley LR, Dyer AJ, Stern K, Hicks T, Nguyen MM. Factors
influencing fluid intake behavior among kidney stone formers.
J Urol. 2012;187:1282-1286.
11. Volpp KG, Gurmankin Levy A, Asch DA, et al. A randomized
controlled trial of financial incentives for smoking cessation.
Cancer Epidemiol Biomarkers Prev. 2006;15:12-18.
12. Volpp KG, Troxel AB, Pauly MV, et al. A randomized, controlled
trial of financial incentives for smoking cessation. N Engl J Med.
2009;360:699-709.
13. Halpern SD, French B, Small DS, et al. Randomized trial of four
financial-incentive programs for smoking cessation. N Engl J
Med. 2015;372:2108-2117.
14. Higgins ST, Bernstein IM, Washio Y, et al. Effects of smoking
cessation with voucher-based contingency management on
birth outcomes. Addiction. 2010;105:2023-2030.
15. Volpp KG, John LK, Troxel AB, Norton L, Fassbender J,
Loewenstein G. Financial incentive-based approaches for
weight loss: a randomized trial. JAMA. 2008;300:2631-2637.
16. John LK, Loewenstein G, Troxel AB, Norton L, Fassbender JE,
Volpp KG. Financial incentives for extended weight loss: a
randomized, controlled trial. J Gen Intern Med. 2011;26:621-
626.
17. Forster JL, Jeffery RW, Sullivan S, Snell MK. A work-site weight
control program using financial incentives collected through
payroll deduction. J Occup Med. 1985;27:804-808.
18. Jeffery RW, Forster JL, Snell MK. Promoting weight control at
the worksite: a pilot program of self-motivation using payroll-
based incentives. Prev Med. 1985;14:187-194.
19. Vandrey R, Bigelow GE, Stitzer ML. Contingency management
in cocaine abusers: a dose-effect comparison of goods-based
906
Scales et al
versus cash-based incentives. Exp Clin Psychopharmacol.
2007;15:338-343.
20. Higgins ST, Budney AJ, Bickel WK, Foerg FE, Donham R,
Badger GJ. Incentives improve outcome in outpatient behav-
ioral treatment of cocaine dependence. Arch Gen Psychiatry.
1994;51:568-576.
21. Silverman K, Wong CJ, Higgins ST, et al. Increasing opiate
abstinence through voucher-based reinforcement therapy.
Drug Alcohol Depend. 1996;41:157-165.
22. Volpp KG, Loewenstein G, Troxel AB, et al. A test of financial
incentives to improve warfarin adherence. BMC Health Serv
Res. 2008;8:272.
23. Kimmel SE, Troxel AB, Loewenstein G, et al. Randomized trial
of lottery-based incentives to improve warfarin adherence. Am
Heart J. 2012;164:268-274.
24. Priebe S, Bremner SA, Lauber C, Henderson C, Burns T.
Financial incentives to improve adherence to antipsychotic
maintenance medication in non-adherent patients: a cluster
randomised controlled trial. Health Technol Assess. 2016;20:
1-122.
25. Asch DA, Volpp KG. On the Way to Health. Leonard Davis
Institute Issue Brief July/August 17, 2012. Accessed
August 4, 2018. https://ldi.upenn.edu/sites/default/files/pdf/
IssueBrief17_9.pdf
26. Weinfurt KP, Griffith JW, Flynn KE, et al. The comprehensive
assessment of self-reported urinary symptoms: a new tool for
research on subtypes of patients with lower urinary tract
symptoms. J Urol. 2019;201:1177-1183.
27. Bernard J, Song L, Henderson B, Tasian GE. Association be-
tween daily water intake and 24-hour urine volume among
adolescents with kidney stones. Urology. 2020;140:150-154.
28. Rule AD, Lieske JC, Li X, Melton LJ 3rd, Krambeck AE,
Bergstralh EJ. The ROKS nomogram for predicting a second
symptomatic stone episode. J Am Soc Nephrol. 2014;25:
2878-2886.
29. Patel MS, Asch DA, Rosin R, et al. Individual versus team-
based financial incentives to increase physical activity: a
randomized, controlled trial. J Gen Intern Med. 2016;31:746-
754.
30. Kullgren JT, Troxel AB, Loewenstein G, et al. Individual- versus
group-based financial incentives for weight loss: a randomized,
controlled trial. Ann Intern Med. 2013;158:505-514.
31. Patel MS, Asch DA, Troxel AB, et al. Premium-based financial
incentives did not promote workplace weight loss in a 2013-15
study. Health Aff (Millwood). 2016;35:71-79.
32. Wong CA, Miller VA, Murphy K, et al. Effect of financial in-
centives on glucose monitoring adherence and glycemic control
among adolescents and young adults with type 1 diabetes: a
randomized clinical trial. JAMA Pediatr. 2017;171:1176-1183.
33. Scales CD Jr, Tasian GE, Schwaderer AL, Goldfarb DS,
Star RA, Kirkali Z. Urinary stone disease: advancing knowledge,
patient care, and population health. Clin J Am Soc Nephrol.
2016;11:1305-1312.
AJKD Vol 77 | Iss 6 | June 2021
Scales et al

Prevention of Urinary Stones with Hydration (PUSH):

Design and Rationale of a Clinical Trial
Design
Desig Interventions Outcomes
R
a Control Arm Primary Outcome:
n
d • Symptomatic stone event
1,642
42 participan
participants o Usual care smart water bottle
≥12 years of age m
i
z
e
d Intervention Arm Secondary Outcomes:
• New stone formation
Symptomatic
mptomatic sstone
t history
h Usual care smart water bottle • Increase in stone size
and low urine volume Multi-component
ulti-compo t behavio
behavioral intervention: • 24-hour urine volume
• Lower urinary tract symptoms

Smartt water bottle cap

capable of Fluid Financial Behavioral
tracking fluid consumption prescription incentive coach

CONCLUSION: Using novel features, the randomized PUSH trial tests if behavioral
CON
interventions to increase adherence to higher fluid intake prevent stone recurrence.
Charles D. Scales, Alana C. Desai, Jonathan D. Harper, et al (2020)
@AJKDonline | DOI: 10.1053/j.ajkd.2020.09.016

AJKD Vol 77 | Iss 6 | June 2021 906.e1

Original Investigation

Change in Cardiac Biomarkers and Risk of Incident Heart

Failure and Atrial Fibrillation in CKD: The Chronic Renal
Insufficiency Cohort (CRIC) Study
Nisha Bansal, Leila R. Zelnick, Elsayed Z. Soliman, Amanda Anderson, Robert Christenson,
Christopher DeFilippi, Rajat Deo, Harold I. Feldman, Jiang He, Bonnie Ky, John Kusek, James Lash,
Stephen Seliger, Tariq Shafi, Myles Wolf, Alan S. Go, and Michael G. Shlipak, on behalf of the CRIC Study
Investigators

Rationale & Objective: Circulating cardiac bio-
markers may signal potential mechanistic path-
ways involved in heart failure (HF) and atrial
fibrillation (AF). Single measures of circulating
cardiac biomarkers are strongly associated with
incident HF and AF in chronic kidney disease
(CKD). We tested the associations of longitudinal
changes in the N-terminal fragment of the pro-
hormone brain natriuretic peptide (NT-proBNP),
high-sensitivity troponin T (hsTnT), galectin-3,
growth differentiation factor 15 (GDF-15), and
soluble ST-2 (sST-2) with incident HF and AF in
patients with CKD.
Study Design: Observational, case-cohort study
design.
Setting & Participants: Adults with CKD
enrolled in the Chronic Renal Insufficiency
Cohort study.
Exposures: Biomarkers were measured at
baseline and 2 years later among those without
kidney failure. We created 3 categories of abso-
lute change in each biomarker: the lowest quar-
tile, the middle 2 quartiles, and the top quartile.
Outcomes: The primary outcomes were incident
HF and AF.
Analytical Approach: Cox proportional hazards
regression models were used to test the associ-
ations of the change categories of each cardiac
biomarker with each outcome (with the middle 2
quartiles of change as the referent group), Complete author and article
adjusting for potential confounders and baseline information provided before
references.
concentrations of each biomarker.
Correspondence to
Results: The incident HF analysis included 789 N. Bansal (nbansal@
participants (which included 138 incident HF nephrology.washington.
cases), and the incident AF analysis included 774 edu)
participants (123 incident AF cases). In multivari- Am J Kidney Dis.
able models, the top quartile of NT-proBNP change 77(6):907-919. Published
(>232 pg/mL over 2 years) was associated with online December 9, 2020.
increased risk of incident HF (HR, 1.79 [95% CI, doi: 10.1053/
1.06-3.04]) and AF (HR, 2.32 [95% CI, 1.37- j.ajkd.2020.09.021
3.93]) compared with the referent group. © 2020 by the National
Participants in the top quartile of sST2 change Kidney Foundation, Inc.
(>3.37 ng/mL over 2 years) had significantly
greater risk of incident HF (HR, 1.89 [95% CI,
1.13-3.16]), whereas those in the bottom quartile
(≤−3.78 ng/mL over 2 years) had greater risk of
incident AF (HR, 2.43 [95% CI, 1.39-4.22])
compared with the 2 middle quartiles. There was
no association of changes in hsTnT, galectin-3, or
GDF-15 with incident HF or AF.
Limitations: Observational study.
Conclusions: In CKD, increases in NT-
proBNP were significantly associated with
greater risk of incident HF and AF, and
increases in sST2 were associated with HF.
Further studies should investigate whether
these markers of subclinical cardiovascular
disease can be modified to reduce the risk of
cardiovascular disease in CKD.

H eart failure (HF) and atrial fibrillation (AF) are leading

causes of cardiovascular disease (CVD) among pa-
tients with chronic kidney disease (CKD) and are associ-
ated with greater risks of death and other poor clinical
outcomes.1-9 Previous work by our group and others has
shown that higher concentrations of the cardiac bio-
markers N-terminal fragment of the prohormone brain
natriuretic peptide (NT-proBNP), high-sensitivity
troponin T (hsTnT), galectin-3, growth differentiation
factor 15 (GDF-15), and soluble ST-2 (sST-2) are associ-
ated with increased risk of incident HF and AF and could
signal potential mechanistic pathways that contribute to
HF in patients with CKD.10-39 More specifically, NT-
proBNP is secreted from cardiac myocytes in response to
myocardial stretch from pressure or volume overload.40-43
Concentrations of hsTnT increase in response to myocar-
dial injury or myocardial remodeling.44,45 Galectin-3 be-
longs to the β-galactoside–binding protein family and is
proinflammatory and profibrotic in cardiomyocytes.46,47
GDF-15 is a member of the transforming growth factor
ß cytokine family48,49 and plays a role in cardiomyocyte
repair.50 ST-2 is a member of the interleukin 1 receptor
family. It has 2 forms, soluble ST2 (sST-2) and trans-
membrane ST-2. ST-2 is a marker of cardiac stress that is
upregulated with myocyte stretch.
Prior studies have observed associations of one-time
measures of these cardiac biomarkers with HF and AF
incidence. In select non-CKD populations such as those
with prevalent HF or cohorts of elderly participants,
studies have suggested that trajectories of cardiac

AJKD Vol 77 | Iss 6 | June 2021 907


PLAIN-LANGUAGE SUMMARY
Patients with chronic kidney disease are at higher risk of
developing heart failure and atrial fibrillation for rea-
sons that are not completely understood. Cardiac bio-
markers measured in the blood may provide insight
into possible mechanisms that contribute to the higher
risk of heart failure and atrial fibrillation. In this study,
we examined changes in cardiac biomarkers over time
and their association with development of heart failure
and atrial fibrillation. We found that increases over a
period of 2 years in 2 of these cardiac biomarkers (NT-
proBNP and sST2) were significantly associated with
higher risk of heart failure, and increases in sST2 were
associated with higher risk of atrial fibrillation.
biomarkers (increases or decreases) may also be associated
with incident HF or AF; thus, repeated measurement of
biomarkers may potentially provide prognostic informa-
tion beyond single baseline measures.15,51-59 However, it
is not known whether changes in cardiac biomarkers
indicate dynamic risk for the development of incident HF
and AF in persons with CKD. If potential associations are
identified, they might identify opportunities for risk sur-
veillance and for targeted primary prevention of HF and
AF. Therefore, we performed a prospective study of a well
characterized CKD cohort to describe changes in 2 cardiac
biomarkers over 2 years and to test for independent as-
sociations of these changes in each cardiac biomarker with
risk of incident HF and AF.
Methods
Study Population
We studied adults with CKD at entry in the Chronic Renal
Insufficiency Cohort (CRIC) study. A total of 3,939 par-
ticipants were enrolled into the CRIC study between June
2003 and August 2008 at 7 clinical centers across the
United States (Ann Arbor/Detroit, MI; Baltimore, MD;
Chicago, IL; Cleveland, OH; New Orleans, LA; Philadel-
phia, PA; and Oakland, CA). Details on study design and
baseline characteristics of the participants were previously
published.60,61 All study participants provided written
informed consent, and the study protocol was approved by
institutional review boards at each of the participating
sites. Inclusion and exclusion criteria have been previously
described.60
The present analysis was an ancillary study to the main
CRIC study, in which we conducted a case-cohort study
designed to evaluate the association of change in cardiac
biomarkers measured at baseline and year 2 with devel-
opment of incident HF and AF after year 2 among par-
ticipants without kidney failure. We randomly sampled
and measured cardiac biomarkers in a subcohort of 700
908
Bansal et al
participants at baseline after excluding 144 with prevalent
HF or AF at study entry (by self-report) or development of
HF or AF between baseline and year 2 (by self report,
electrocardiography, or occurrence of an adjudicated HF or
AF hospitalization) and 34 in whom kidney failure
developed by year 2 (the subcohort size was determined
based on power for the analyses and available resources).
In addition, we measured cardiac biomarkers in 135
additional participants in whom incident HF or AF
developed after year 2, but who were free of kidney failure
(added cases). Therefore, for the incident HF analysis, we
included 700 participants in the subcohort (of whom 49
experienced incident HF) and an additional 89 HF cases,
for a total study population of 789 (including 138 incident
HF cases). For the incident AF analysis, we included 700
participants in the subcohort (of whom 49 experienced
incident AF) and an additional 74 AF cases, for a total
study population of 774 (including 123 incident AF
cases).
Cardiac Biomarkers
GDF-15, sST2, and galectin-3 were measured from EDTA
plasma stored at −70 C from samples at baseline and at
year 2 in a random subcohort in batch at the University of
Pennsylvania CRIC Central Laboratory. All assays were
measured in duplicate. GDF-15 and sST2 were measured
using ELISA (R&D Systems). For GDF-15, the quantitative
range was 23.4-1,500 pg/mL with a lower limit of
detection of <2.0 pg/mL. At a concentration of 98.8 pg/
mL, the coefficient of variation (CV) was 7.2%; at a con-
centration of 624 pg/mL, the CV was 4.5%. For sST2, the
quantitative range was 0.63-40 ng/mL with a limit of
detection of <0.1 ng/mL. At a concentration of 2.6 ng/
mL, the CV was 11.2%; at 0.94 ng/mL, the CV was 8.5%.
For galectin-3, the quantitative range was 0.31-10.0 ng/
mL with a lower limit of detection of <0.016 ng/mL. At
an average concentration of 0.78 ng/mL, the CV was
4.0%; at 5.36 ng/mL, the CV was 4.2%.
hsTnT and NT-proBNP were measured at baseline in
2008 from EDTA plasma stored at −70 C by using a
chemiluminescent microparticle immunoassay (Roche
Diagnostics) on an Elecsys 2010 analyzer at the University
of Maryland. hs-TnT was measured by using a high-
sensitivity assay with a range of values from 3 to
10,000 ng/L.62 The limit of blank was 3 ng/L and limit of
detection was <5 ng/L. For hsTnT, the CVs were 3% at a
concentration of 30 ng/L and 5.8% at 2,213 ng/L. The
value at the 99th-percentile cutoff from a healthy reference
population was 13 ng/L for hsTnT, with a 10% CV.62 The
range of values for NT-proBNP was from 114 to
5,900 ng/L, and the CVs were 4.25% at a concentration of
132 ng/L and 5.3% at 4,640 ng/L.
In 2017, we added year-2 measures of NT-proBNP and
hsTnT and remeasured a subset of baseline samples at the
same testing laboratory to calibrate the measures. The new
measurements in 2017 were performed on the Roche
AJKD Vol 77 | Iss 6 | June 2021
Bansal et al
E601 analyzer. We remeasured NT-proBNP in 100 random
samples from baseline and all of the year-2 samples. We
developed and applied a Deming regression63 to calibrate
the 2008 baseline NT-proBNP measures with the 2017
NT-proBNP measures. The goodness of fit with this cali-
bration had an R2 of 0.991507 for NT-proBNP. The re-
sidual plots for the original and recalibrated measures were
similar (Fig S1).
Similarly, for hsTnT, we remeasured any baseline hsTnT
measure with a value <5 ng/L by using the newer Roche
E601 instrument, which had a limit of blank of 2.5 ng/L
and limit of detection of <3 ng/L. With the E601 instru-
ment, at a concentration of 13.5 ng/L, the CV was 1.9%; at
4,831 ng/L, the CV was 0.8%. We also measured a
random subset of 100 samples at baseline and all samples
at the year-2 visit. We developed and applied a Deming
regression (similar methods as used for NT-proBNP) to
calibrate the 2008 baseline hsTnT measures with the 2017
hsTnT measures. The goodness of fit with this calibration
had an R2 of 0.970514. The residual plots for the original
and recalibrated measures were similar (Fig S2).
Incident HF
The primary outcome was incident HF that occurred be-
tween year 2 of study entry through May 2014 (the last
date that HF events were adjudicated). HF was identified
by asking study participants biannually if they were hos-
pitalized and by reviewing electronic health records from
selected hospitals or health care delivery systems. The first
30 discharge codes were identified for all hospitalizations,
and codes relevant to HF resulted in retrieval of medical
records by study personnel for centralized adjudicated
review. At least 2 study physicians reviewed all possible HF
events and deaths by using medical records and guidelines
on clinical symptoms, radiographic evidence of pulmonary
congestion, physical examination of the heart and lungs,
and, when available, central venous hemodynamic moni-
toring data and echocardiographic imaging. HF was
confirmed when both reviewers agreed upon a “probable”
or “definite” occurrence of HF based on modified clinical
Framingham criteria.64 Deaths were identified from report
from next of kin, retrieval of death certificates or obitu-
aries, or review of hospital records, and through the Social
Security Death Master File and National Death Index.
Incident AF
Incident AF was defined as a hospitalization for AF and
confirmed by dual physician adjudication.19 Every 6
months, participants were asked if they had visited an
emergency department or had been hospitalized. Medical
records from corresponding hospitals or health care systems
were queried for qualifying encounters. Diagnostic codes
for AF (International Classification of Diseases, Ninth Revision, Clinical
Modification 427.31 or 427.32) prompted retrieval of medical
records and centralized review for the ascertainment of
incident AF. Final adjudication of events was done after at
AJKD Vol 77 | Iss 6 | June 2021
least 2 study physicians reviewed all possible AF events by
manual review of relevant medical records. Electrocardio-
grams of hospitalized participants (when available) were
reviewed and were part of the adjudication process.
Covariates
At the baseline visit and each annual visit, participants
provided information on their sociodemographic char-
acteristics, medical history, medication use, and lifestyle
behaviors. Race/ethnicity was categorized as non-
Hispanic white, non-Hispanic black, Hispanic, and
other. History of CVD was determined by self-report.
Current tobacco use was determined by self-report.
Diabetes mellitus was defined as a fasting glucose lev-
el >126 mg/dL, a nonfasting glucose level >200 mg/dL,
or use of insulin or other antidiabetic medication.
Anthropometric measurements and blood pressure were
assessed by using standard protocols.65 Body mass index
was derived as weight in kilograms divided by the
square of height in meters. Serum creatinine was
measured by an enzymatic method on an Ortho Vitros
950 analyzer at the CRIC Central Laboratory and stan-
dardized to isotope-dilution mass spectrometry–traceable
values,66,67 and the CKD-EPI (CKD Epidemiology
Collaboration) creatinine equation was used to estimate
GFR.68 Additional assays measured included serum
phosphorus, 24-hour urine total protein, glucose, low-
density lipoprotein cholesterol, high-density lipoprotein
cholesterol, and total parathyroid hormone. The afore-
mentioned assays were performed at the central labora-
tory with the exception of parathyroid hormone
(measured at Scantibodies Laboratory Inc).
Statistical Approach
There are no established cutoff values to define clinically
meaningful changes in each of the cardiac biomarkers of
interest. Therefore, to define change, we subtracted the
concentration at year 2 from the baseline concentration,
meaning a negative number represented a longitudinal
decrease in the biomarker level, whereas a positive number
represented an increase in the biomarker level. We then
examined the distribution of absolute change from base-
line to year 2 of each cardiac biomarker among the sub-
cohort and created 3 categories of change: the lowest
quartile of absolute change in each biomarker, the middle
2 quartiles of absolute change, and the top quartile of
absolute change. The middle 2 quartiles were designated as
the referent group, as these were likely the participants
with the most stable levels of each biomarker over the
course of 2 years.
We examined characteristics of the study population in
the subcohort overall and across these categories of abso-
lute change in each biomarker. Among the subcohort, we
used proportional odds regression to test the association of
participant characteristics with odds of a higher category of
absolute change for each cardiac biomarker.
909

To account for the case-cohort study design, we used
Cox models with robust variance estimation to examine
the association of change in cardiac biomarkers (modeled
in the categories described earlier) with incident HF and
AF by using the Prentice weighting method.69,70 Partici-
pants were considered at risk from the date of the year-2
visit until the first occurrence of definite or probable HF
or AF or until they were censored as a result of death,
dropout, progression to kidney failure, or loss to follow-
up. We used unadjusted and adjusted Cox models,
adjusting for covariates at year 2 that included de-
mographic characteristics (age, sex, race, study site), dia-
betes, CVD (defined as history of myocardial infarction
[MI], stroke, and HF), smoking, measures of kidney
function (estimated GFR [eGFR] and 24-hour urine pro-
tein), systolic blood pressure, low- and high-density li-
poprotein cholesterol, and baseline levels of the biomarker.
We performed several sensitivity analyses. In the first, we
adjusted for the change in eGFR between baseline and year
2 to evaluate whether concurrent change in kidney function
may have accounted for some of the observed associations
between change in cardiac biomarkers and risk of HF and
AF. In the second, we adjusted for use of angiotensin-
converting enzyme inhibitors/angiotensin receptor
blockers, diuretic agents, and β-blockers at baseline and
year 2 to also evaluate whether differential use of cardio-
vascular medications, which may affect cardiovascular
structure/function and circulating levels of the biomarkers,
was a possible confounder in the observed associations. In
the third sensitivity analysis, to account for possible effects
of interval cardiovascular events on the observed associa-
tions, we adjusted for interim, time-updated adjudicated
hospitalizations for AF and MI for the HF outcome and for
HF and MI hospitalizations for the AF outcome. In the
fourth sensitivity analysis, we examined the association of
relative changes in each cardiac biomarker over 2 years with
subsequent HF and AF. Similar to the approach for the
primary analysis, we examined the distribution of relative
change from baseline to year 2 of each cardiac biomarker
among the subcohort and created 3 categories of relative
change: the lowest quartile of relative change in each
biomarker, the middle 2 quartiles of relative change, and
the top quartile of relative change. The middle 2 quartiles of
relative change were designated as the referent group, as
these were likely the participants with the most stable levels
of each biomarker over 2 years.
In a secondary analysis, we also modeled the association
of baseline levels of the biomarker with risk of incident HF
and AF. In another secondary analysis, we examined the
ability of each cardiac biomarker to predict HF and AF and
evaluated the discriminatory ability via the 10-fold cross-
validated Harrell’s C-index and the integrated discrimina-
tion index with accompanying 95% confidence intervals
(CIs).71,72 We compared several models to predict AF or
HF: clinical model (age, sex, race/ethnicity, site, diabetes,
CVD, smoking, log-transformed 24-hour urine protein,
eGFR, systolic blood pressure, body mass index, low- and
910
Bansal et al
high-density lipoprotein cholesterol), clinical model plus
baseline cardiac biomarker, and clinical model plus base-
line cardiac biomarker plus change in cardiac biomarker.
There were small (<3%) amounts of missingness in the
variables used in these sensitivity analyses; we accounted
for this by using multiple imputation with chained equa-
tions, combining the resulting estimates with Rubin’s rules
to account for the variability in the imputation
procedure.73
A nominal P value of <0.05 was taken as evidence of
statistical significance in all analyses. All analyses were
conducted using the R 3.4.3 computing environment (R
Foundation for Statistical Computing).
Results
Change in Cardiac Biomarkers Over 2 Years in
Study Population
Among participants in the subcohort, mean age was 59.4
years, 44% were women, and 36% were Black. Mean
eGFR was 46.8 mL/min/1.73 m2, and 43% had a history
of diabetes (Table 1). Among the subcohort, the median
absolute change in NT-proBNP was 50.0 (interquartile
range [IQR], –11.9 to 190.7) pg/mL; in hsTnT, it was
2.8 (IQR, –0.2 to 9.4) ng/mL; in GDF-15, 110.0 (IQR,
–96.0 to 524.0) pg/mL; in galectin-3, 2.8 (IQR, –0.5 to
6.4) ng/mL; and in sST2, –0.5 (IQR, –3.7 to 3.3) ng/mL
(Fig S3 and Table 1). Median relative change in NT-
proBNP was 85.0% (IQR, –10% to –330%); in hsTnT,
it was 23% (IQR, –2% to –70%); in GDF-15, 11% (IQR,
–9% to 41%); in galectin-3, 22% (IQR, –4% to 52%); and
in sST2, –4% (IQR, –24% to 27%). Participants with the
greatest absolute elevation in NT-proBNP had lower
baseline eGFR and were more likely have diabetes and
higher baseline systolic blood pressure and less likely to
have used diuretic agents at baseline (Table S1). The
participants with the greatest absolute increase in hsTnT
were more likely to be men, to be Black, to have diabetes,
to have AF, and to have higher systolic blood pressure and
lower eGFR (Table S2). Participants with the greatest
absolute increase in GDF-15 over 2 years were more
likely to be Black, have diabetes, smoke, and have higher
systolic blood pressure and lower baseline eGFR
(Table S3). Participants with the greatest change in
galectin-3 over 2 years were more likely to have diabetes,
smoke, and to have higher systolic blood pressure and
lower eGFR at baseline (Table S4). Participants with the
greatest change in sST2 over 2 years were more likely to
be Black, have higher systolic blood pressure, and have a
greater decrease in eGFR over 2 years (Table S5).
In multivariable models, lower eGFR was associated
with greater odds of a higher category of absolute change
for all the cardiac biomarkers of interest with the exception
of hsTnT and sST2 (Table 2). Higher systolic blood pres-
sure was associated with greater odds of change in NT-
proBNP. Black race and diabetes were associated with
greater odds of change in GDF-15 (Table 2).
AJKD Vol 77 | Iss 6 | June 2021
Bansal et al

Association of Changes in NT-proBNP and hsTnT Table 1. Demographic Characteristics of Subcohort

With Incident HF and AF Parameter Value
The mean follow-up time from the second biomarker NT-proBNP
measurement among the cohort eligible for the HF analysis Baseline, pg/mL 96 [37, 216]
was 5.3 ± 2.4 (SD) years. In unadjusted models, the Absolute change, pg/mL 50 [−12, 191]
highest and lowest categories of change in NT-proBNP Relative change, % 85 [−10, 330]
over 2 years were significantly associated with increased hsTnT
risk of incident HF. However, with adjustment for po- Baseline, ng/mL 12.4 [7.9, 19.8]
tential confounders in model 1, only the highest quartile Absolute change, ng/mL 2.8 [−0.2, 9.4]
of change remained significantly associated with risk of Relative change, % 23 [−2, 70]
incident HF. This association remained statistically signif- GDF-15
icant when adjusted for baseline levels of NT-proBNP (HR, Baseline, pg/mL 1318 [906, 1,874]
1.79 [95% CI, 1.06-3.04]; Table 3 and Table S6). In Absolute change, pg/mL 110 [−96, 524]
sensitivity analyses, with adjustment of change in eGFR Relative change, % 11 [−9, 41]
between baseline and year 2, use of cardiovascular medi- Galectin-3
cations at baseline and/or year 2, and development of Baseline, ng/mL 13.5 [9.7, 17.9]
interval AF or MI, the observed association remained sta- Absolute change, ng/mL 2.8 [−0.5, 6.4]
tistically significant (Table S7). When we evaluated relative Relative change, % 22 [−4, 52]
change in NT-proBNP over 2 years, there was not a sig- sST-2
nificant association with incident HF (Table S9). Baseline, ng/mL 14.7 [10.5, 20.1]
The mean follow-up time from the second biomarker Absolute change, ng/mL −0.5 [−3.7, 3.3]
measurement among the cohort eligible for the AF analysis Relative change, % −4 [−24, 27]
Age, y 59.4 ± 11.2
was 5.3 ± 2.4 (standard deviation) years. The highest
Female sex 310 (44%)
category of change in NT-proBNP was significantly asso-
Race/ethnicity
ciated with greater risk of incident AF in models adjusted
Non-Hispanic White 332 (47%)
for potential confounders, including kidney function and
Non-Hispanic Black 249 (36%)
cardiovascular risk factors. This association remained sta-
Hispanic 86 (12%)
tistically significant when adjusting for baseline levels of Other 33 (5%)
NT-proBNP (HR, 2.32 [95% CI, 1.37-3.93]; Table 4 and eGFR
Table S6). This association remained robust with adjust- Baseline, mL/min/1.73m2 46.8 ± 14.1
ment for change in eGFR, use of cardiovascular medica- Relative change, % −8.2 ± 20.9
tions, and development of interim HF or MI (Table S7). 24-h urine protein, g/d 0.1 [0.1, 0.5]
There was no association between the lowest category of Diabetes 302 (43%)
change in NT-proBNP with incident AF. When we evalu- History of CVD
ated relative change in NT-proBNP over 2 years, there was Never 522 (75%)
not a significant association with incident AF (Table S10). At year 2 only 21 (3%)
In unadjusted models, there was a strong association of the At baseline and year 2 157 (22%)
highest quartile of change in hsTnT with risk of incident HF Current smoker 70 (10%)
(HR, 1.85 [95% CI, 1.20-2.85]); however, this association BMI, kg/m2 31.5 ± 7.4
was attenuated and no longer statistically significant with SBP, mm Hg 125.1 ± 20.4
adjustment for potential confounders and mediators (Table 3 DBP, mm Hg 69.9 ± 12.4
and Table S7). There was no association of high or low cat- LDL cholesterol, mg/dL 100.7 ± 33.6
egories of change in hsTnT with risk of incident AF (Table 4). HDL cholesterol, mg/dL 48.1 ± 15.5
Similarly, there was no association of relative changes in ACEi/ARBs
hsTnT with risk of incident HF or AF (Tables S9 and S10). Never 160 (23%)
At year 2 only 64 (9%)
Association of Change in GDF-15, Galectin-3, and At baseline and year 2 426 (61%)
sST2 With Incident HF and AF Diuretics
In unadjusted models, the highest category of change in Never 262 (37%)
GDF-15 was associated with a 2-fold greater risk of incident At year 2 only 62 (9%)
HF (HR, 2.56 [95% CI, 1.67-3.94]; Table 3). However, this At baseline and year 2 297 (42%)
β-Blockers
association was attenuated and no longer statistically sig-
Never 358 (51%)
nificant with adjustment for potential confounders,
including baseline GDF-15 level. There was no association (Continued)

AJKD Vol 77 | Iss 6 | June 2021 911


Table 1 (Cont'd). Demographic Characteristics of Subcohort
Parameter Value
At year 2 only 62 (9%)
At baseline and year 2 239 (34%)
PTH, median 46.5 [31.0, 76.0]
Note: N = 675 with 2 measures of each biomarker. Values for continuous variables
given as mean ± SD or median [interquartile range].
Abbreviations: ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin
receptor blocker; BMI, body mass index; CVD, cardiovascular disease; DBP, dia-
stolic blood pressure; eGFR, estimated glomerular filtration rate; GDF-15, growth
differentiation factor 15; HDL, high-density lipoprotein; hsTnT, high-sensitivity
troponin T; LDL, low-density lipoprotein; NT-proBNP, prohormone brain natri-
uretic peptide; PTH, parathyroid hormone; SBP, systolic blood pressure; sST-2,
soluble ST-2.
of change in GDF-15 with incident AF (Table 4). Similarly,
there was no association of relative changes in GDF-15 with
risk of incident HF or AF (Tables S9 and S10).
In unadjusted models, the highest category of change in
galectin-3 was associated with a greater risk of incident HF
(HR, 1.72 [95% CI, 1.13-2.64]; Table 3). However, this
association was completely attenuated and no longer sta-
tistically significant with adjustment for potential con-
founders. There was no association of change in galectin-3
with incident AF (Table 4). Similarly, there was no asso-
ciation of relative changes in galectin-3 with risk of inci-
dent HF or AF (Tables S9 and S10).
In unadjusted models, the lowest and highest categories
of change in sST2 were significantly associated with greater
risk of incident HF. With adjustment for potential con-
founders and baseline sST2 level, the quartile of greatest
change remained significantly associated with risk of inci-
dent HF (HR, 1.89 [95% CI, 1.13-3.16]) and remained
significant with adjustment for change in eGFR, use of
cardiovascular medications, and development of interim AF
and MI (Table 3 and Table S3). The association of higher
relative change in sST2 with incident HF was also significant
(Table S9). There was also a significant association between
the lowest category of absolute change in sST2 with higher
risk of incident AF in the fully adjusted model (HR, 2.43
[95% CI, 1.39-4.22]; Table 4); this association persisted
when adjusting for change in eGFR, use of cardiovascular
medications, and development of interim AF and MI
(Table S7). Similar associations were observed for relative
change in sST2 over 2 years as well (Table S10).
Discrimination of Cardiac Biomarkers to Predict
Incident HF and AF
Addition of the baseline level or change in each cardiac
biomarker did not improve the ability to predict incident
HF or AF beyond the use of only clinical variables
(Tables S11 and S12).
Discussion
In this study of a well-characterized CKD cohort free of HF
and AF, we examined associations of longitudinal changes
in NT-proBNP, hsTnT, GDF-15, galectin-3, and sST2 levels
over a period of 2 years with the risk of incident HF and
AF. Participants with the greatest level of change in NT-
912
Bansal et al
proBNP (>232-pg/mL increase over 2 years) had a 79%
higher risk of incident HF and a 2-fold higher risk of
incident AF. In addition, participants with the greatest
change in sST2 (>3.37-ng/mL increase over 2 years) had
an 89% greater risk of incident HF, and those with the
greatest decreases in sST2 (≥3.78-ng/mL decrease over
2 years) had a greater risk of incident AF. There were no
associations between changes in hsTnT, galectin-3, or
GDF-15 with incident HF or AF.
Our study found that there was a significant association
of absolute increases in NT-proBNP over 2 years with risk
of subsequent HF and AF in this CKD cohort, independent
of important confounders such as CVD and kidney func-
tion. Prior studies in CKD have evaluated one-time mea-
sures of NT-proBNP,26,74 but not longitudinal changes,
with the risk of HF. However, studies in other populations
have reported that changes in NT-proBNP are significantly
associated with HF and HF outcomes.75-79 For example, a
study of the Valsartan Heart Failure Therapy trial showed
that temporal trends over a period of 4 months in NT-
proBNP superseded the prognostic value of a single mea-
sure.79 In a study of 190 patients with HF, changes from
normal to increased NT-proBNP levels were significantly
associated with death and hospitalization.80 Prior studies
have also reported strong associations of one-time mea-
sures of NT-proBNP with AF in various populations
(including CKD)32,81-83; however, there are limited prior
data on changes in NT-proBNP with risk of AF in patients
with CKD. It is plausible that increases in NT-proBNP
reflect progression of underlying subclinical cardiovascu-
lar pathophysiology, such as volume overload and cardiac
stretch. If so, NT-proBNP may be a good surrogate marker
to reflect physiologic abnormalities in patients with CKD
that contribute to the risk of HF and AF but may not be a
strong biomarker to predict HF or AF beyond other clinical
variables.
We did not see a similar association when evaluating
relative increases in NT-proBNP, suggesting that absolute
increases in NT-proBNP capture the risk of incident HF and
AF better than relative change does. The reasons to explain
this finding are not clear from the present study. However,
it is interesting that the median relative changes were quite
large compared with the absolute changes; it is possible
that large relative changes are actually capturing healthy
participants who start with lower levels of NT-proBNP.
Future interventional studies may help explain this finding.
We also did not find that absolute or relative decreases
in NT-proBNP were associated with lower risk of HF and
AF in participants with CKD. This differs from prior studies
of patients with known HF, which have demonstrated that
decreases in NT-proBNP are associated with reduced risk
of HF. For example, an analysis of 172 patients with HF
found that reductions in NT-proBNP level were associated
with reduced risk of death.84 In a study of older adults,
those with a 25% increase in NT-proBNP had a 2-fold
greater risk of HF, and those who had a 25% decrease
had a 40% decreased risk of HF.85 It is possible that
AJKD Vol 77 | Iss 6 | June 2021
Bansal et al

Table 2. Baseline Characteristics Associated With Category of 2-Year Change in Cardiac Biomarker
Characteristic NT-proBNP hsTnT GDF-15 Galectin-3 sST-2
Age (per 10-y increment) 0.94 (0.79-1.12) 0.84 (0.72-0.99) 0.98 (0.83-1.15) 0.92 (0.78-1.09) 1.04 (0.88-1.23)
Female sex 1.64 (1.12-2.39)a 1.37 (0.95-1.97) 0.97 (0.67-1.40) 1.43 (1.00-2.05)a 0.71 (0.50-1.01)
Race/ethnicity
Non-Hispanic Black 0.89 (0.61-1.30) 1.12 (0.78-1.62) 1.45 (1.01-2.09)a 1.08 (0.75-1.54) 1.19 (0.83-1.71)
Hispanic 1.95 (1.18-3.24)a 1.02 (0.62-1.68) 1.26 (0.75-2.10) 0.92 (0.56-1.53) 1.27 (0.78-2.07)
Other 0.78 (0.36-1.66) 0.76 (0.36-1.58) 0.72 (0.32-1.61) 0.66 (0.31-1.42) 0.73 (0.35-1.56)
eGFR (per 15-mL/min/ 1.21 (1.01-1.46)a 1.16 (0.98-1.38) 1.25 (1.04-1.49)a 1.19 (1.01-1.41)a 0.93 (0.78-1.10)
1.73m2 decrement)
24-h urine protein 1.11 (0.95-1.30) 0.98 (0.84-1.14) 1.03 (0.88-1.20) 1.18 (0.99-1.39) 1.14 (0.98-1.32)
(per 1-g/d increment)
Diabetes 0.99 (0.68-1.45) 0.75 (0.51-1.10) 1.44 (1.00-2.09)a 1.28 (0.90-1.84) 1.05 (0.73-1.50)
CVD 1.31 (0.91-1.90) 0.72 (0.49-1.05) 1.25 (0.86-1.81) 1.30 (0.91-1.85) 0.93 (0.65-1.35)
Smoking 1.10 (0.66-1.82) 1.24 (0.75-2.07) 1.49 (0.88-2.51) 1.27 (0.77-2.09) 1.11 (0.68-1.80)
BMI (per 5-kg/m2 0.99 (0.88-1.12) 0.98 (0.88-1.10) 1.00 (0.89-1.13) 0.94 (0.84-1.05) 1.07 (0.96-1.20)
increment)
SBP (per 10-mm Hg 1.13 (1.02-1.25)a 1.06 (0.96-1.17) 1.07 (0.97-1.19) 0.96 (0.86-1.06) 1.03 (0.94-1.14)
increment)
DBP (per 5-mm Hg 0.88 (0.81-0.97) 0.99 (0.91-1.08) 0.95 (0.87-1.03) 1.00 (0.92-1.09) 0.99 (0.91-1.08)
increment)
LDL (per 10-mg/dL 1.03 (0.98-1.09) 0.96 (0.91-1.01) 0.99 (0.94-1.04) 0.95 (0.90-1.00) 0.94 (0.89-0.99)
increment)
HDL (per 10-mg/dL 0.90 (0.79-1.03) 0.97 (0.86-1.10) 0.98 (0.87-1.10) 0.93 (0.83-1.05) 1.06 (0.95-1.19)
increment)
ACEi/ARB 0.68 (0.48-0.97) 1.08 (0.76-1.55) 0.84 (0.58-1.20) 0.98 (0.69-1.39) 0.99 (0.70-1.40)
Diuretics 0.87 (0.62-1.23) 1.10 (0.79-1.53) 0.92 (0.65-1.28) 1.01 (0.73-1.40) 1.17 (0.84-1.63)
β-Blockers 1.29 (0.92-1.80) 1.11 (0.80-1.54) 1.16 (0.83-1.61) 1.15 (0.83-1.59) 0.78 (0.56-1.08)
BL NT-proBNP (per 1- 1.09 (0.91-1.31) 1.00 (0.83-1.21) 0.89 (0.69-1.15) 0.98 (0.80-1.20) 1.20 (1.01-1.43)a
SD increment)
BL hsTnT (per 1-SD 1.07 (0.90-1.29) 1.79 (1.42-2.25)a 1.03 (0.87-1.22) 1.16 (0.98-1.37) 1.08 (0.92-1.28)
increment)
BL GDF-15 (per 1-SD 1.21 (0.98-1.48) 1.01 (0.82-1.24) 1.21 (1.01-1.46)a 0.96 (0.79-1.17) 1.02 (0.84-1.23)
increment)
BL galectin-3 (per 1-SD 0.81 (0.68-0.96) 1.03 (0.88-1.22) 1.02 (0.87-1.21) 1.14 (0.98-1.33) 1.01 (0.86-1.20)
increment)
BL sST-2 (per 1-SD 1.10 (0.94-1.30) 1.21 (1.03-1.43)a 1.04 (0.88-1.23) 0.97 (0.83-1.15) 1.15 (1.00-1.33)a
increment)
Note: Values indicate odds of a higher category of absolute change for each cardiac biomarker. Categories are quartile 1, quartiles 2 and 3, and quartile 4 of 2-year
biomarker change; the quartile representing the smallest increase of the biomarker serves as the referent. Entries are odds ratios (95% CI) from a proportional-odds
model category of biomarker change, estimated among subcohort only, in a model that includes all listed covariates.
Abbreviations: ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; BL, baseline; BMI, body mass index; CVD, cardiovascular disease; DBP,
diastolic blood pressure; eGFR, estimated glomerular filtration rate; GDF-15, growth differentiation factor 15; HDL, high-density lipoprotein; hsTnT, high-sensitivity troponin
T; LDL, low-density lipoprotein; NT-proBNP, prohormone brain natriuretic peptide; SBP, systolic blood pressure; SD, standard deviation; sST-2, soluble ST-2.
a
P < 0.05.

differences in the study populations may explain the
disparate findings.
We also noted a statistically significant association be-
tween absolute and relative increases in sST2 with risk of
incident HF. However, sST2 measured at baseline or seri-
ally did not improve discrimination of HF in this CKD
population. In 2 studies of patients hospitalized with HF,
the change in sST2 level from admission to discharge was
significantly associated with mortality.58,59 In a study of
151 patients with chronic HF, serial sST2 measurement
added prognostic information to the baseline concentra-
tions and predicted change in left ventricular function.86
An international consensus panel concluded that, even
though there are strong data supporting single
measurements of sST2, further data are needed to test the
merits of serial testing of sST2.87 Our study provides
support that serial sST2 measures may provide important
mechanistic data in patients with CKD at risk for HF. It is
surprising that we also noted an association between
decreasing absolute and relative sST2 levels with increased
risk of incident AF. The reasons for this observation are not
clear, but may be related to unmeasured confounders such
as treatment effects (eg, confounding by indication). In
addition, this suggests that changes in sST2 are not causally
linked to risk of AF.
After adjustment for potential confounders, we did not
find statistically significant associations of absolute or
relative changes in hsTnT, galectin-3, and GDF-15 with

AJKD Vol 77 | Iss 6 | June 2021 913

 Bansal et al

Table 3. Associations of Absolute Change in Cardiac Biomarkers Over 2 Years With Risk of Incident Heart Failure
No. Unadjusted Model 1 Model 2
Cardiac Biomarker At Risk Events HR (95% CI) P HR (95% CI) P HR (95% CI) P
ΔNT-proBNP
≤−11.4 pg/mL 196 41 1.91 (1.22-2.99) 0.005 1.68 (0.99-2.85) 0.05 1.51 (0.88-2.59) 0.1
−11.3 to 232 pg/mL 382 47 1.00 (reference) – 1.00 (reference) – 1.00 (reference) –
>232 pg/mL 185 47 2.61 (1.69-4.05) <0.001 1.94 (1.15-3.27) 0.01 1.79 (1.06-3.04) 0.03
BL NT-proBNP, per – – NA NA 1.19 (1.03-1.37) 0.02
1-SD greater
ΔhsTnT
≤−0.201 ng/mL 195 35 1.33 (0.85-2.09) 0.2 1.12 (0.66-1.92) 0.7 1.04 (0.60-1.80) 0.9
−0.20 to 10.1 ng/mL 391 57 1.00 (reference) – 1.00 (reference) – 1.00 (reference) –
>10.1 ng/mL 191 43 1.85 (1.20-2.85) 0.005 1.42 (0.86-2.34) 0.2 1.28 (0.76-2.15) 0.4
BL hsTnT, per 1-SD – – NA NA 1.17 (0.99,1.38) 0.06
greater
ΔGDF-15
≤−96.8 pg/mL 182 33 1.33 (0.83-2.12) 0.2 1.15 (0.66-2.00) 0.6 0.95 (0.52-1.72) 0.9
−96.7 to 532 pg/mL 357 49 1.00 (reference) – 1.00 (reference) – 1.00 (reference) –
>532 pg/mL 184 49 2.56 (1.67-3.94) <0.001 1.27 (0.74-2.17) 0.4 1.26 (0.74-2.15) 0.4
BL GDF-15, per 1-SD – – NA NA 1.28 (1.01-1.61) 0.04
greater
ΔGalectin 3
≤−0.405 ng/mL 194 32 1.03 (0.66-1.63) 0.9 0.84 (0.49-1.45) 0.5 0.66 (0.36-1.24) 0.2
−0.404 to 405 64 1.00 (reference) – 1.00 (reference) – 1.00 (reference) –
6.95 ng/mL
>6.95 ng/mL 187 42 1.72 (1.13-2.64) 0.01 0.90 (0.53-1.53) 0.7 0.94 (0.55-1.59) 0.8
BL galectin-3, per – – NA NA 1.25 (0.97-1.61) 0.08
1-SD greater
ΔsST-2
≤−3.78 ng/mL 196 40 1.85 (1.18-2.90) 0.007 1.93 (1.11-3.35) 0.02 1.66 (0.91-3.04) 0.1
−3.77 to 3.37 ng/mL 373 49 1.00 (reference) – 1.00 (reference) – 1.00 (reference) –
>3.37 ng/mL 203 48 2.35 (1.52-3.62) <0.001 1.87 (1.11-3.13) 0.02 1.89 (1.13-3.16) 0.02
BL sST2, per 1-SD – – NA NA 1.14 (0.94-1.39) 0.2
greater
Note: Model 1: year-2 age, sex, race, site, diabetes, CVD, smoking, 24-h urinary protein, eGFR, SBP, BMI, LDL, HDL. Model 2: model 1 + baseline biomarker.
Abbreviations: BL, baseline; BMI, body mass index; CVD, cardiovascular disease; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; GDF-15,
growth differentiation factor 15; HDL, high-density lipoprotein; hsTnT, high-sensitivity troponin T; LDL, low-density lipoprotein; NT-proBNP, prohormone brain natriuretic
peptide; SBP, systolic blood pressure; SD, standard deviation; sST-2, soluble ST-2.

risk of incident HF or AF. This differs from previous
studies of patients with HF, older adults, and patients
undergoing dialysis that found associations of changes in
troponin T with risk of events such as HF and
death.75,76,78,88 For example, in a study of older adults,
decreases in troponin T were associated with reduced risk
of incident HF, whereas increases in troponin T over 2 or 3
years were associated with increased risk.76 Also, in the
same population of older adults, increases in hsTnT >
10.97 ng/mL were associated with a 90% increased risk of
incident AF.55 In addition, a few studies have evaluated
longitudinal measures of GDF-1515,56 and galectin-
3,37,38,89-91 with conflicting findings. For example, in a
study of participants in the PREVEND study, the highest
quartile of change in galectin-3 over 4 years was signifi-
cantly associated with a composite outcome of new-onset
HF, cardiovascular mortality, all-cause mortality, and new-
onset AF.56 In contrast, prior studies reported no
association of even baseline measures of galectin-3 with
incident AF.38,83 In patients with HF, increases in GDF-15
were significantly associated with greater risk of mortality
in one study.15 However, another study of patients with
HF did not report a significant association.86 Thus, our
data in a CKD population do not suggest that serial mea-
sures of hsTnT, galectin-3, and GDF-15 add prognostic
information.
The findings from the present study have potentially
important implications. Despite the disproportionate risk
for HF and AF, the mechanisms that contribute to these
diseases in CKD remain incompletely understood. These
circulating biomarkers may help identify pathways that
could be targets for therapies. Measurement of NT-proBNP
has the advantage of being widely available in clinical
practice. In addition, the use of NT-proBNP measurements
to guide use of HF therapies in patients with prevalent HF
has been controversial, with some trials suggesting a

914 AJKD Vol 77 | Iss 6 | June 2021

Bansal et al

Table 4. Associations of Absolute Change in Cardiac Biomarkers Over 2 Years With Incident Atrial Fibrillation
No. Unadjusted Model 1 Model 2
Cardiac Biomarker At Risk Events HR (95% CI) P HR (95% CI) P HR (95% CI) P
ΔNT-proBNP
≤−11.4 pg/mL 191 27 1.34 (0.80-2.23) 0.3 1.25 (0.71-2.20) 0.4 1.19 (0.67-2.09) 0.6
−11.3 to 373 44 1.00 (reference) – 1.00 (reference) – 1.00 (reference) –
232 pg/mL
>232 pg/mL 183 47 2.87 (1.83-4.49) <0.001 2.43 (1.44-4.09) <0.001 2.32 (1.37-3.93) 0.002
BL NT-proBNP, per – – NA NA 1.14 (0.96-1.36) 0.1
1-SD greater
ΔhsTnT
≤−0.201 ng/mL 186 23 0.79 (0.48-1.30) 0.4 0.88 (0.51-1.51) 0.6 0.79 (0.45-1.38) 0.4
−0.20 to 10.1 392 66 1.00 (reference) – 1.00 (reference) – 1.00 (reference) –
ng/mL
>10.1 ng/mL 183 29 1.13 (0.71-1.80) 0.6 1.09 (0.64-1.86) 0.8 0.98 (0.56-1.71) 0.9
BL hsTnT, per 1-SD – – NA NA 1.21 (1.00-1.47) 0.05
greater
ΔGDF-15
≤−96.8 pg/mL 183 36 1.42 (0.89-2.25) 0.1 1.15 (0.69-1.93) 0.6 1.06 (0.61-1.83) 0.8
−96.7 to 532 350 50 1.00 (reference) – 1.00 (reference) – 1.00 (reference) –
pg/mL
>532 pg/mL 174 29 1.48 (0.91-2.41) 0.1 0.96 (0.55-1.68) 0.9 0.95 (0.55-1.66) 0.9
BL GDF-15, per – – NA NA 1.15 (0.89-1.48) 0.3
1-SD greater
ΔGalectin 3
≤−0.405 ng/mL 193 24 0.71 (0.43-1.16) 0.2 0.74 (0.43-1.29) 0.3 0.55 (0.29-1.05) 0.07
−0.404 to 402 68 1.00 (reference) – 1.00 (reference) – 1.00 (reference) –
6.95 ng/mL
>6.95 ng/mL 176 31 1.23 (0.78-1.95) 0.4 1.18 (0.69-1.99) 0.6 1.22 (0.72-2.07) 0.5
BL galectin 3, per – – NA NA 1.30 (1.30-1.69) 0.05
1-SD greater
ΔsST-2
≤−3.78 ng/mL 201 46 2.22 (1.43-3.43) <0.001 2.55 (1.53-4.26) <0.001 2.43 (1.39-4.22) 0.002
−3.77 to 3.37 ng/ 372 48 1.00 (reference) – 1.00 (reference) – 1.00 (reference) –
mL
>3.37 ng/mL 185 29 1.42 (0.87-2.33) 0.2 1.14 (0.65-1.99) 0.7 1.14 (0.65-2.00) 0.7
BL sST2, per 1-SD – – NA NA 1.05 (0.86-1.29) 0.6
greater
Note: Model 1: year-2 age, sex, race, site, diabetes, CVD, smoking, 24-h urinary protein, eGFR, SBP, BMI, LDL, HDL. Model 2: model 1 + baseline biomarker.
Abbreviations: BL, baseline; BMI, body mass index; CVD, cardiovascular disease; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; GDF-15,
growth differentiation factor 15; HDL, high-density lipoprotein; hsTnT, high-sensitivity troponin T; LDL, low-density lipoprotein; NT-proBNP, prohormone brain natriuretic
peptide; SBP, systolic blood pressure; SD, standard deviation; sST-2, soluble ST-2.

benefit in reducing recurrent HF hospitalizations51-54 and
others suggesting none.92 The findings from prior trials
appear to differ based on the study population of interest.
Based on our findings, it is possible that serial measures of
NT-proBNP and sST2 may successfully guide use of car-
diovascular therapies to reduce the risk of incident HF and
AF in select patients with CKD. Further investigation of
biomarker-guided interventions and other surrogate
markers are needed in CKD to reduce the risk of CVD.
Our study had several strengths. We studied a large,
multicenter, well-characterized CKD cohort with longi-
tudinal follow-up. Our study population was diverse. We
had serial measures of all 5 cardiac biomarkers measured
2 years apart. HF and AF were ascertained through a
rigorous physician-adjudication process. We were able to
adjust for a broad range of potential confounders. We
recognize a few limitations as well. Cardiac biomarkers
were measured only twice, 2 years apart; this may have
been too short an interval for meaningful changes, even
though large differences were observed. There may have
been day-to-day biological variability in the biomarker
concentrations that may have made our assessment of
change less precise. NT-proBNP and hsTnT at baseline and
year 2 were not measured in batch; however, the excel-
lent goodness of fit suggests that it is not likely that the
timing of the measurements would have influenced our
findings. We adjusted for potential confounders and
mediators, but the possibility of residual confounding
remains. This was an observational study, so we cannot
determine causality. Finally, this was a study of research
volunteers with CKD, so findings may not be generaliz-
able to all populations.

AJKD Vol 77 | Iss 6 | June 2021 915


In conclusion, in this longitudinal cohort of patients
with CKD without HF, we found that increases in NT-
proBNP were significantly associated with greater risk
of incident HF and AF and that increases in sST2 were
associated with incident HF. Further studies are needed
to test whether interventions to modify markers of
subclinical CVD can decrease the risk of HF and AF in
CKD.
Supplementary Material
Supplementary File (PDF)
Figure S1: Residual plots for original and retested NT-proBNP.
Figure S2: Residual plots for original and retested hsTnT.
Figure S3: Histograms of absolute change in each biomarker
(N = 700).
Table S1: Demographic data based on absolute change over 2
years in NT-proBNP (n = 675).
Table S2: Demographic data based on absolute change over 2
years in hsTnT (n = 689).
Table S3: Demographic data based on absolute change over 2
years in GDF-15 (n = 638).
Table S4: Demographic data based on absolute change over 2
years in galectin-3 (n = 697).
Table S5: Demographic data based on absolute change over 2
years in sST-2 (n = 684).
Table S6: Association of baseline cardiac biomarkers with risk of
incident HF and AF.
Table S7: Associations of absolute change in biomarkers over 2
years with risk of incident HF, adjusting for medication use and
interim cardiovascular events.
Table S8: Associations of absolute change in biomarkers over 2
years with risk of incident AF, adjusting for medication use and
interim cardiovascular events.
Table S9: Associations of relative change over 2 years in cardiac
biomarkers with incident HF.
Table S10: Associations of relative change over 2 years in cardiac
biomarkers with incident AF.
Table S11: Discrimination of cardiac biomarkers for predicting
incident HF in CKD.
Table S12: Discrimination of cardiac biomarkers for predicting
incident AF in CKD.
Article Information
CRIC Study Investigators: Lawrence J. Appel, MD, MPH,
Panduranga S. Rao, MD, Mahboob Rahman, MD, Raymond R.
Townsend, MD.
Authors’ Full Names and Academic Degrees: Nisha Bansal, MD,
Leila R. Zelnick, PhD, Elsayed Soliman, MD, Amanda Anderson,
PhD, Robert Christenson, PhD, Christopher DeFilippi, MD, Rajat
Deo, MD, Harold I. Feldman, MD, Jiang He, MD, PhD, Bonnie Ky,
MD, John Kusek, PhD, James Lash, MD, Stephen Seliger, MD,
Tariq Shafi, MD, Myles Wolf, MD, Alan S. Go, MD, and Michael G.
Shlipak, MD.
Authors’ Affiliations: From the Kidney Research Institute, University
of Washington, Seattle, WA (NB); Division of Nephrology, University
of Washington, Seattle, WA (LRZ); Departments of Epidemiology
and Medicine, Wake Forest University, Winston-Salem, NC (ES);
916
Bansal et al
School of Public Health & Tropical Medicine (AA) and Department
of Epidemiology (JH), Tulane University, New Orleans, LA;
Department of Pathology (RC) and Division of Nephrology (SS),
University of Maryland, Baltimore, MD; Division of Cardiology,
Inova Health System, Falls Church, VA (CD); Division of
Cardiology (RD, BK) and Department of Epidemiology (JK),
University of Pennsylvania, Philadelphia, PA; Department of
Biostatistics, Epidemiology and Informatics, University of
Pennsylvania Perelman School of Medicine, Philadelphia, PA (HIF);
Renal, Electrolyte, and Hypertension Division, Department of
Medicine, University of Pennsylvania Perelman School of Medicine,
Philadelphia, PA (HIF); Center for Clinical Epidemiology and
Biostatistics, University of Pennsylvania Perelman School of
Medicine, Philadelphia, PA (HIF); Division in Nephrology, University
of Chicago, Chicago, IL (JL); Division of Nephrology, University of
Mississippi, Oxford, MS (TS); Division of Nephrology, Duke
University, Durham, NC (MW); Division of Research, Kaiser
Permanente of Northern California, Oakland, CA (ASG); and
Department of Medicine, University of California, San Francisco,
San Francisco, CA (MGS).
Address for Correspondence: Nisha Bansal, MD, MAS, Kidney
Research Institute, University of Washington, 908 Jefferson St, 3rd
floor, Seattle, WA 98104. Email: nbansal@nephrology.washington.
edu
Authors’ Contributions: Research idea and study design: NB,
MGS; data acquisition: NB, AA, RC, HIF, JH, JK, JL, ASG, MGS;
data analysis and interpretation: NB, LRZ, EZS, AA, RC, CD, RD,
HIF, JH, BK, JK, JL, SS, TS, MW, ASG, MGS; statistical analysis:
LRZ; supervision or mentorship: NB, MGS. Each author
contributed important intellectual content during manuscript
drafting or revision and agrees to be personally accountable for
the individual’s own contributions and to ensure that questions
pertaining to the accuracy or integrity of any portion of the work,
even one in which the author was not directly involved, are
appropriately investigated and resolved, including with
documentation in the literature if appropriate.
Support: This study was supported by NIH grants R01 DK103612 (NB)
and R01 DK104730 (AA). Roche Diagnostics provided partial funding for
the NT-proBNP and hsTnT assays. Funding for the CRIC Study was
obtained under a cooperative agreement from National Institute of
Diabetes and Digestive and Kidney Diseases (U01DK060990,
U01DK060984, U01DK061022, U01DK061021, U01DK061028,
U01DK060980, U01DK060963, and U01DK060902). In addition,
this work was supported in part by the Perelman School of Medicine at
the University of Pennsylvania Clinical and Translational Science Award
National Institutes of Health (NIH)/National Center for Advancing
Translational Sciences (NCATS) UL1TR000003, Johns Hopkins
University UL1 TR-000424, University of Maryland General Clinical
Research Center M01 RR-16500, Clinical and Translational Science
Collaborative of Cleveland, UL1TR000439 from the NCATS
component of NIH and NIH roadmap for Medical Research, Michigan
Institute for Clinical and Health Research UL1TR000433, University of
Illinois at Chicago Clinical and Translational Science Award
UL1RR029879, Tulane Center of Biomedical Research Excellence for
Clinical and Translational Research in Cardiometabolic Diseases P20
GM109036, Kaiser Permanente NIH/National Center for Research
Resources UCSF-CTSI UL1 RR-024131, and an unrestricted fund
from the Northwest Kidney Centers. The funders had no role in study
design, data collection, analysis, reporting, or the decision to submit for
publication.
Financial Disclosure: The authors declare that they have no
relevant financial interests.
Peer Review: Received March 30, 2020. Evaluated by 3 external
peer reviewers and a statistician, with editorial input from an
Acting Editor-in-Chief (Editorial Board Member Wendy L. St. Peter,
PharmD). Accepted in revised form September 26, 2020. The
AJKD Vol 77 | Iss 6 | June 2021
Bansal et al
involvement of an Acting Editor-in-Chief to handle the peer-review
and decision-making processes was to comply with AJKD’s
procedures for potential conflicts of interest for editors, described
in the Information for Authors & Journal Policies.
References
1. Kottgen A, Russell SD, Loehr LR, et al. Reduced kidney func-
tion as a risk factor for incident heart failure: the atherosclerosis
risk in communities (ARIC) study. J Am Soc Nephrol.
2007;18(4):1307-1315.
2. Go AS, Chertow GM, Fan D, McCulloch CE, Hsu CY. Chronic
kidney disease and the risks of death, cardiovascular events,
and hospitalization. N Engl J Med. 2004;351(13):1296-1305.
3. Bansal N, Katz R, Robinson-Cohen C, et al. Absolute rates of
heart failure, coronary heart disease, and stroke in chronic
kidney disease: an analysis of 3 community-based cohort
studies. JAMA Cardiol. 2017;2(3):314-318.
4. Saran R, Robinson B, Abbott KC, et al. US Renal Data System
2016 annual data report: epidemiology of kidney disease in the
United States. Am J Kidney Dis. 2017;69(3)(suppl 1):S1-
S434.
5. Porter AC, Lash JP, Xie D, et al. Predictors and outcomes of
health-related quality of life in adults with CKD. Clin J Am Soc
Nephrol. 2016;11(7):1154-1162.
6. Harel Z, Wald R, McArthur E, et al. Rehospitalizations and
emergency department visits after hospital discharge in pa-
tients receiving maintenance hemodialysis. J Am Soc Nephrol.
2015;26(12):3141-3150.
7. Bansal N, Fan D, Hsu CY, Ordonez JD, Go AS. Incident atrial
fibrillation and risk of death in adults with chronic kidney dis-
ease. J Am Heart Assoc. 2014;3(5):e001303.
8. Bansal N, Fan D, Hsu CY, Ordonez JD, Marcus GM, Go AS.
Incident atrial fibrillation and risk of end-stage renal disease in
adults with chronic kidney disease. Circulation. 2013;127(5):
569-574.
9. Bansal N, Xie D, Tao K, et al. Atrial fibrillation and risk of ESRD
in adults with CKD. Clin J Am Soc Nephrol. 2016;11(7):1189-
1196.
10. Wang TJ, Wollert KC, Larson MG, et al. Prognostic utility of
novel biomarkers of cardiovascular stress: the Framingham
Heart Study. Circulation. 2012;126(13):1596-1604.
11. Rohatgi A, Patel P, Das SR, et al. Association of growth dif-
ferentiation factor-15 with coronary atherosclerosis and mor-
tality in a young, multiethnic population: observations from the
Dallas Heart Study. Clin Chem. 2012;58(1):172-182.
12. Kempf T, Horn-Wichmann R, Brabant G, et al. Circulating
concentrations of growth-differentiation factor 15 in apparently
healthy elderly individuals and patients with chronic heart failure
as assessed by a new immunoradiometric sandwich assay.
Clin Chem. 2007;53(2):284-291.
13. Bonaca MP, Morrow DA, Braunwald E, et al. Growth differenti-
ation factor-15 and risk of recurrent events in patients stabilized
after acute coronary syndrome: observations from PROVE IT-
TIMI 22. Arterioscler Thromb Vasc Biol. 2011;31(1):203-210.
14. Kempf T, von Haehling S, Peter T, et al. Prognostic utility of
growth differentiation factor-15 in patients with chronic heart
failure. J Am Coll Cardiol. 2007;50(11):1054-1060.
15. Anand IS, Kempf T, Rector TS, et al. Serial measurement of
growth-differentiation factor-15 in heart failure: relation to dis-
ease severity and prognosis in the Valsartan Heart Failure Trial.
Circulation. 2010;122(14):1387-1395.
16. Xanthakis V, Larson MG, Wollert KC, et al. Association of novel
biomarkers of cardiovascular stress with left ventricular
AJKD Vol 77 | Iss 6 | June 2021
hypertrophy and dysfunction: implications for screening. J Am
Heart Assoc. 2013;2(6):e000399.
17. Lind L, Wallentin L, Kempf T, et al. Growth-differentiation factor-
15 is an independent marker of cardiovascular dysfunction and
disease in the elderly: results from the Prospective Investigation
of the Vasculature in Uppsala Seniors (PIVUS) Study. Eur
Heart J. 2009;30(19):2346-2353.
18. Kempf T, Bjorklund E, Olofsson S, et al. Growth-differentiation
factor-15 improves risk stratification in ST-segment elevation
myocardial infarction. Eur Heart J. 2007;28(23):2858-2865.
19. Bansal N, Xie D, Sha D, et al. Cardiovascular events after new-
onset atrial fibrillation in dults with CKD: results from the
Chronic Renal Insufficiency Cohort (CRIC) Study. J Am Soc
Nephrol. 2018;29(12):2859-2869.
20. Wollert KC, Kempf T, Lagerqvist B, et al. Growth differentiation
factor 15 for risk stratification and selection of an invasive
treatment strategy in non ST-elevation acute coronary syn-
drome. Circulation. 2007;116(14):1540-1548.
21. Wollert KC, Kempf T, Peter T, et al. Prognostic value of growth-
differentiation factor-15 in patients with non-ST-elevation acute
coronary syndrome. Circulation. 2007;115(8):962-971.
22. Wollert KC, Kempf T, Wallentin L. Growth differentiation factor
15 as a biomarker in cardiovascular disease. Clin Chem.
2017;63(1):140-151.
23. Gleissner CA, Erbel C, Linden F, et al. Galectin-3 binding
protein, coronary artery disease and cardiovascular mortality:
insights from the LURIC study. Atherosclerosis. 2017;260:
121-129.
24. O Miro, Gonz alez de la Presa B, Herrero P, et al. The GALA
study: relationship between galectin-3 serum levels and short-
and long-term outcomes of patients with acute heart failure.
Biomarkers. 2017;22(8):731-739.
25. Drechsler C, Delgado G, Wanner C, et al. Galectin-3,
renal function, and clinical outcomes: results from the
LURIC and 4D studies. J Am Soc Nephrol. 2015;26(9):
2213-2221.
26. Bansal N, Hyre Anderson A, Yang W, et al. High-sensitivity
troponin T and N-terminal pro-B-type natriuretic peptide (NT-
proBNP) and risk of incident heart failure in patients with CKD:
the Chronic Renal Insufficiency Cohort (CRIC) Study. J Am
Soc Nephrol. 2015;26(4):946-956.
27. Tuegel C, Katz R, Alam M, et al. GDF-15, Galectin 3, soluble
ST2, and risk of mortality and cardiovascular events in CKD.
Am J Kidney Dis. 2018;72(4):519-528.
28. Czaya B, Seeherunvong W, Singh S, et al. Cardioprotective
effects of paricalcitol alone and in combination with FGF23
receptor inhibition in chronic renal failure: experimental and
clinical studies. Am J Hypertens. 2019;32(1):34-44.
29. Faul C. Cardiac actions of fibroblast growth factor 23. Bone.
2017;100:69-79.
30. Xie J, Yoon J, An SW, Kuro-o M, Huang CL. Soluble klotho
protects against uremic cardiomyopathy independently of
fibroblast growth factor 23 and phosphate. J Am Soc Nephrol.
2015;26(5):1150-1160.
31. Faul C, Amaral AP, Oskouei B, et al. FGF23 induces left ven-
tricular hypertrophy. J Clin Invest. 2011;121(11):4393-4408.
32. Fan J, Cao H, Su L, et al. NT-proBNP, but not ANP and C-
reactive protein, is predictive of paroxysmal atrial fibrillation in
patients undergoing pulmonary vein isolation. J Interv Card
Electrophysiol. 2012;33(1):93-100.
33. Kornej J, Schmidl J, Ueberham L, et al. Galectin-3 in patients
with atrial fibrillation undergoing radiofrequency catheter abla-
tion. PLoS One. 2015;10(4):e0123574.
34. Sinner MF, Stepas KA, Moser CB, et al. B-type natriuretic
peptide and C-reactive protein in the prediction of atrial
917

fibrillation risk: the CHARGE-AF Consortium of community-
based cohort studies. Europace. 2014;16(10):1426-1433.
35. Wallentin L, Hijazi Z, Andersson U, et al; ARISTOTLE In-
vestigators. Growth differentiation factor 15, a marker of oxida-
tive stress and inflammation, for risk assessment in patients with
atrial fibrillation: insights from the Apixaban for Reduction in
Stroke and Other Thromboembolic Events in Atrial Fibrillation
(ARISTOTLE) trial. Circulation. 2014;130(21):1847-1858.
36. Wu XY, Li SN, Wen SN, et al. Plasma galectin-3 predicts
clinical outcomes after catheter ablation in persistent atrial
fibrillation patients without structural heart disease. Europace.
2015;17(10):1541-1547.
37. Yalcin MU, Gurses KM, Kocyigit D, et al. The association of serum
galectin-3 levels with atrial electrical and structural remodeling.
J Cardiovasc Electrophysiol. 2015;26(6):635-640.
38. Fashanu OE, Norby FL, Aguilar D, et al. Galectin-3 and inci-
dence of atrial fibrillation: the Atherosclerosis Risk in Commu-
nities (ARIC) study. Am Heart J. 2017;192:19-25.
39. Nortamo S, Ukkola O, Lepojarvi S, et al. Association of sST2
and hs-CRP levels with new-onset atrial fibrillation in coronary
artery disease. Int J Cardiol. 2017;248:173-178.
40. Yasue H, Yoshimura M, Sumida H, et al. Localization and
mechanism of secretion of B-type natriuretic peptide in com-
parison with those of A-type natriuretic peptide in normal
subjects and patients with heart failure. Circulation.
1994;90(1):195-203.
41. Vickery S, Price CP, John RI, et al. B-type natriuretic peptide
(BNP) and amino-terminal proBNP in patients with CKD:
relationship to renal function and left ventricular hypertrophy.
Am J Kidney Dis. 2005;46(4):610-620.
42. Mishra RK, Li Y, Ricardo AC, et al. Association of N-terminal pro-
B-type natriuretic peptide with left ventricular structure and
function in chronic kidney disease (from the Chronic Renal
Insufficiency Cohort [CRIC]). Am J Cardiol. 2013;111(3):432-
438.
43. DeFilippi CR, Fink JC, Nass CM, Chen H, Christenson R. N-
terminal pro-B-type natriuretic peptide for predicting coronary
disease and left ventricular hypertrophy in asymptomatic CKD
not requiring dialysis. Am J Kidney Dis. 2005;46(1):35-44.
44. de Lemos JA, Drazner MH, Omland T, et al. Association of
troponin T detected with a highly sensitive assay and cardiac
structure and mortality risk in the general population. JAMA.
2010;304(22):2503-2512.
45. Mishra RK, Li Y, DeFilippi C, et al; CRIC Study Investigators.
Association of cardiac troponin T with left ventricular struc-
ture and function in CKD. Am J Kidney Dis. 2013;61(5):701-
709.
46. Calvier L, Miana M, Reboul P, et al. Galectin-3 mediates
aldosterone-induced vascular fibrosis. Arterioscler Thromb
Vasc Biol. 2013;33(1):67-75.
47. Vergaro G, Prud’homme M, Fazal L, et al. Inhibition of galectin-3
pathway prevents isoproterenol-induced left ventricular
dysfunction and fibrosis in mice. Hypertension. 2016;67(3):606-
612.
48. Unsicker K, Spittau B, Krieglstein K. The multiple facets of the
TGF-beta family cytokine growth/differentiation factor-15/
macrophage inhibitory cytokine-1. Cytokine Growth Factor
Rev. 2013;24(4):373-384.
49. Nair V, Robinson-Cohen C, Smith MR, et al. Growth differen-
tiation factor-15 and risk of CKD progression. J Am Soc
Nephrol. 2017;28(7):2233-2240.
50. Kempf T, Eden M, Strelau J, et al. The transforming growth
factor-beta superfamily member growth-differentiation factor-
15 protects the heart from ischemia/reperfusion injury. Circ
Res. 2006;98(3):351-360.
918
Bansal et al
51. Pfisterer M, Buser P, Rickli H, et al. BNP-guided vs symptom-
guided heart failure therapy: the Trial of Intensified vs Standard
Medical Therapy in Elderly Patients With Congestive Heart Failure
(TIME-CHF) randomized trial. JAMA. 2009;301(4):383-392.
52. Jourdain P, Jondeau G, Funck F, et al. Plasma brain natriuretic
peptide-guided therapy to improve outcome in heart failure: the
STARS-BNP Multicenter Study. J Am Coll Cardiol.
2007;49(16):1733-1739.
53. Ledwidge M, Gallagher J, Conlon C, et al. Natriuretic peptide-
based screening and collaborative care for heart failure: the
STOP-HF randomized trial. JAMA. 2013;310(1):66-74.
54. Felker GM, Hasselblad V, Hernandez AF, O’Connor CM.
Biomarker-guided therapy in chronic heart failure: a meta-
analysis of randomized controlled trials. Am Heart J.
2009;158(3):422-430.
55. Hussein AA, Bartz TM, Gottdiener JS, et al. Serial measures of
cardiac troponin T levels by a highly sensitive assay and inci-
dent atrial fibrillation in a prospective cohort of ambulatory older
adults. Heart Rhythm. 2015;12(5):879-885.
56. van der Velde AR, Meijers WC, Ho JE, et al. Serial galectin-3
and future cardiovascular disease in the general population.
Heart. 2016;102(14):1134-1141.
57. Eggers KM, Kempf T, Wallentin L, Wollert KC, Lind L. Change in
growth differentiation factor 15 concentrations over time
independently predicts mortality in community-dwelling elderly
individuals. Clin Chem. 2013;59(7):1091-1098.
58. Boisot S, Beede J, Isakson S, et al. Serial sampling of ST2
predicts 90-day mortality following destabilized heart failure.
J Card Fail. 2008;14(9):732-738.
59. Manzano-Fernandez S, Januzzi JL, Pastor-Perez FJ, et al. Serial
monitoring of soluble interleukin family member ST2 in patients
with acutely decompensated heart failure. Cardiology.
2012;122(3):158-166.
60. Feldman HI, Appel LJ, Chertow GM, et al. The Chronic Renal
Insufficiency Cohort (CRIC) study: design and methods. J Am
Soc Nephrol. 2003;14(7)(suppl 2):S148-S153.
61. Lash JP, Go AS, Appel LJ, et al. Chronic Renal Insufficiency
Cohort (CRIC) study: baseline characteristics and associations
with kidney function. Clin J Am Soc Nephrol. 2009;4(8):1302-
1311.
62. Giannitsis E, Kurz K, Hallermayer K, Jarausch J, Jaffe AS,
Katus HA. Analytical validation of a high-sensitivity cardiac
troponin T assay. Clin Chem. 2010;56(2):254-261.
63. Linnet K. Necessary sample size for method comparison
studies based on regression analysis. Clin Chem. 1999;45(6
pt 1):882-894.
64. Ho KK, Anderson KM, Kannel WB, Grossman W, Levy D.
Survival after the onset of congestive heart failure in Fra-
mingham Heart Study subjects. Circulation. 1993;88(1):107-
115.
65. National Center for Health Statistics (NCHS). National Health
and Nutrition Examination Survey Anthropometry Procedures
Manual. Centers for Disease Control and Prevention [serial
online]; 2000. Accessed August 15, 2020. https://www.cdc.
gov/nchs/data/nhanes/nhanes_07_08/manual_an
66. Joffe M, Hsu CY, Feldman HI, et al. Variability of creatinine
measurements in clinical laboratories: results from the CRIC
study. Am J Nephrol. 2010;31(5):426-434.
67. Levey AS, Coresh J, Greene T, et al. Expressing the Modifica-
tion of Diet in Renal Disease Study equation for estimating
glomerular filtration rate with standardized serum creatinine
values. Clin Chem. 2007;53(4):766-772.
68. Levey AS, Stevens LA, Schmid CH, et al. A new equation to
estimate glomerular filtration rate. Ann Intern Med.
2009;150(9):604-612.
AJKD Vol 77 | Iss 6 | June 2021
Bansal et al
69. Therneau TM, Li H. Computing the Cox model for case cohort
designs. Lifetime Data Anal. 1999;5(2):99-112.
70. Barlow WE. Robust variance estimation for the case-cohort
design. Biometrics. 1994;50(4):1064-1072.
71. Harrell FE, Lee KL, Mark DB. Multivariable prognostic models:
issues in developing models, evaluating assumptions and ad-
equacy, and measuring and reducing errors. Stat Med.
1996;15(4):361-387.
72. Uno H, Tian L, Cai T, Kohane IS, Wei LJ. A unified inference
procedure for a class of measures to assess improvement in
risk prediction systems with survival data. Stat Med.
2013;32(14):2430-2442.
73. Rubin DB. Multiple Imputation for Nonresponse in Surveys.
Wiley; 1987.
74. Bansal N, Zelnick L, Go A, et al. Cardiac biomarkers and risk of
incident heart failure in chronic kidney disease: the CRIC
(Chronic Renal Insufficiency Cohort) study. J Am Heart Assoc.
2019;8(21):e012336.
75. Miller WL, Hartman KA, Burritt MF, Grill DE, Jaffe AS. Profiles of
serial changes in cardiac troponin T concentrations and
outcome in ambulatory patients with chronic heart failure. J Am
Coll Cardiol. 2009;54(18):1715-1721.
76. deFilippi CR, de Lemos JA, Christenson RH, et al. Association
of serial measures of cardiac troponin T using a sensitive assay
with incident heart failure and cardiovascular mortality in older
adults. JAMA. 2010;304(22):2494-2502.
77. Eggers KM, Venge P, Lind L. Prognostic usefulness of the
change in N-terminal pro B-type natriuretic peptide levels to
predict mortality in a single community cohort aged >/= 70
years. Am J Cardiol. 2013;111(1):131-136.
78. Masson S, Anand I, Favero C, et al. Serial measurement of
cardiac troponin T using a highly sensitive assay in patients with
chronic heart failure: data from 2 large randomized clinical tri-
als. Circulation. 2012;125(2):280-288.
79. Masson S, Latini R, Anand IS, et al. Prognostic value of
changes in N-terminal pro-brain natriuretic peptide in Val-HeFT
(Valsartan Heart Failure Trial). J Am Coll Cardiol. 2008;52(12):
997-1003.
80. Miller WL, Hartman KA, Burritt MF, et al. Serial biomarker
measurements in ambulatory patients with chronic heart failure:
the importance of change over time. Circulation. 2007;116(3):
249-257.
81. Patton KK, Ellinor PT, Heckbert SR, et al. N-terminal pro-B-type
natriuretic peptide is a major predictor of the development of
atrial fibrillation: the Cardiovascular Health Study. Circulation.
2009;120(18):1768-1774.
82. Patton KK, Heckbert SR, Alonso A, et al. N-terminal pro-B-type
natriuretic peptide as a predictor of incident atrial fibrillation in
AJKD Vol 77 | Iss 6 | June 2021
the Multi-Ethnic Study of Atherosclerosis: the effects of age,
sex and ethnicity. Heart. 2013;99(24):1832-1836.
83. Lamprea-Montealegre JA, Zelnick LR, Shlipak MG, et al; CRIC
Study Investigators. Cardiac biomarkers and risk of atrial
fibrillation in chronic kidney disease: the CRIC study. J Am
Heart Assoc. 2019;8(15):e012200.
84. Miller WL, Hartman KA, Grill DE, Burnett JC Jr, Jaffe AS. Only
large reductions in concentrations of natriuretic peptides (BNP
and NT-proBNP) are associated with improved outcome in
ambulatory patients with chronic heart failure. Clin Chem.
2009;55(1):78-84.
85. deFilippi CR, Christenson RH, Gottdiener JS, Kop WJ,
Seliger SL. Dynamic cardiovascular risk assessment in
elderly people. The role of repeated N-terminal pro-B-type
natriuretic peptide testing. J Am Coll Cardiol. 2010;55(5):
441-450.
86. Gaggin HK, Szymonifka J, Bhardwaj A, et al. Head-to-head
comparison of serial soluble ST2, growth differentiation
factor-15, and highly-sensitive troponin T measurements in
patients with chronic heart failure. JACC Heart Fail.
2014;2(1):65-72.
87. Januzzi JL, Pascual-Figal D, Daniels LB. ST2 testing for
chronic heart failure therapy monitoring: the International ST2
Consensus Panel. Am J Cardiol. 2015;115(7)(suppl):70b-
75b.
88. Sandoval Y, Herzog CA, Love SA, et al. Prognostic value of
serial changes in high-sensitivity cardiac troponin I and T over 3
months using reference change values in hemodialysis pa-
tients. Clin Chem. 2016;62(4):631-638.
89. Anand IS, Rector TS, Kuskowski M, Adourian A, Muntendam P,
Cohn JN. Baseline and serial measurements of galectin-3 in
patients with heart failure: relationship to prognosis and effect
of treatment with valsartan in the Val-HeFT. Eur J Heart Fail.
2013;15(5):511-518.
90. Motiwala SR, Szymonifka J, Belcher A, et al. Serial measure-
ment of galectin-3 in patients with chronic heart failure: results
from the ProBNP Outpatient Tailored Chronic Heart Failure
Therapy (PROTECT) study. Eur J Heart Fail. 2013;15(10):
1157-1163.
91. van der Velde AR, Gullestad L, Ueland T, et al. Prognostic value
of changes in galectin-3 levels over time in patients with heart
failure: data from CORONA and COACH. Circ Heart Fail.
2013;6(2):219-226.
92. Felker GM, Anstrom KJ, Adams KF, et al. Effect of natriuretic
peptide–guided therapy on hospitalization or cardiovascular
mortality in high-risk patients with heart failure and reduced
ejection fraction: a randomized clinical trial. JAMA.
2017;318(8):713-720.
919
 Original Investigations

Design and Rationale of HiLo: A Pragmatic, Randomized

Trial of Phosphate Management for Patients Receiving
Maintenance Hemodialysis
Daniel L. Edmonston, Tamara Isakova, Laura M. Dember, Steven Brunelli, Amy Young, Rebecca Brosch,
Srinivasan Beddhu, Hrishikesh Chakraborty, and Myles Wolf

Visual Abstract online

Rationale & Objective: Hyperphosphatemia is a
risk factor for poor clinical outcomes in patients
with kidney failure receiving maintenance dialysis.
Opinion-based clinical practice guidelines
recommend the use of phosphate binders and
dietary phosphate restriction to lower serum
phosphate levels toward the normal range in
patients receiving maintenance dialysis, but the
benefits of these approaches and the optimal
serum phosphate target have not been tested in
randomized trials. It is also unknown if
aggressive treatment that achieves
unnecessarily low serum phosphate levels
worsens outcomes.
Study Design: Multicenter, pragmatic, cluster-
randomized clinical trial.
Setting & Participants: HiLo will randomize 80-
120 dialysis facilities operated by DaVita Inc and
the University of Utah to enroll 4,400 patients un-
dergoing 3-times-weekly, in-center hemodialysis.
Intervention: Phosphate binder prescriptions
and dietary recommendations to achieve the “Hi”
serum phosphate target (≥6.5 mg/dL) or the “Lo”
serum phosphate target (<5.5 mg/dL).
Outcomes: Primary outcome: Hierarchical com-
posite outcome of all-cause mortality and all- Complete author and article
cause hospitalization. Main secondary information provided before
references.
outcomes: Individual components of the primary
outcome. Correspondence to M. Wolf
(myles.wolf@duke.edu)
Results: The trial is currently enrolling. Am J Kidney Dis.
Limitations: HiLo will not adjudicate causes of 77(6):920-930. Published
online December 3, 2020.
hospitalizations or mortality and does not pro-
tocolize use of specific phosphate binder classes. doi: 10.1053/
j.ajkd.2020.10.008
Conclusions: HiLo aims to address an important © 2020 The Authors.
clinical question while more generally advancing Published by Elsevier Inc.
methods for pragmatic clinical trials in nephrology on behalf of the National
by introducing multiple innovative features including Kidney Foundation, Inc. This
stakeholder engagement in the study design, liberal is an open access article
eligibility criteria, use of electronic informed con- under the CC BY-NC-ND
license (http://
sent, engagement of dietitians to implement the creativecommons.org/
interventions in real-world practice, leveraging licenses/by-nc-nd/4.0/).
electronic health records to eliminate dedicated
study visits, remote monitoring of serum
phosphate separation between trial arms, and use
of a novel hierarchical composite outcome.
Trial Registration: Registered at Clinical-
Trials.gov with study number NCT04095039.

Scientific Basis of the HiLo Study
Dialysis outcomes have improved somewhat in recent
years, but hospitalizations and mortality rates are still
unacceptably elevated.1 These poor outcomes mainly
arise from an increased risk of cardiovascular disease,
but most interventions targeting traditional cardiovascu-
lar risk factors that improve clinical outcomes in the
general population have not been found to be effective
when evaluated in randomized trials of patients with
kidney failure.2-4 The lack of benefit of traditional car-
diovascular interventions has led the nephrology com-
munity to invoke and target putative kidney
failure–specific risk factors for cardiovascular disease
and death.
Hyperphosphatemia
Hyperphosphatemia is a ubiquitous complication of
kidney failure that is associated with increased risks of
cardiovascular disease and death in observational
studies.5-9 Experimental data suggest that hyper-
phosphatemia contributes to arterial calcification, which
itself is associated with increased risk of cardiovascular
events and death.10-13 Arterial stiffness due to calcifica-
tion promotes left ventricular hypertrophy, which is
associated with heart failure, arrhythmia, and death.14,15
Hyperphosphatemia also exacerbates kidney
failure–associated increases in circulating levels of para-
thyroid hormone and fibroblast growth factor 23, each
of which are associated with left ventricular hypertro-
phy, heart failure, arrhythmias, and infectious mortality
in patients with kidney failure.6,16-22
The Contemporary Standard of Care
Based on these data, the nephrology community has
embraced opinion-based practice guidelines that recom-
mend aggressive treatment of hyperphosphatemia to less
than 5.5 mg/dL or “toward the normal range” in patients
with kidney failure undergoing hemodialysis.23-25 Because
standard hemodialysis removes an insufficient amount of
phosphate, most patients must restrict their dietary phos-
phate intake, and more than 80% require treatment with
phosphate binders to achieve recommended serum phos-
phate targets.26,27

920 AJKD Vol 77 | Iss 6 | June 2021

Edmonston et al
PLAIN-LANGUAGE SUMMARY
Citing observational and preclinical studies that link
hyperphosphatemia to adverse clinical outcomes, cur-
rent clinical practice guidelines recommend reduction
of serum phosphate “toward the normal range” in pa-
tients with kidney failure undergoing hemodialysis.
However, no randomized clinical trials have tested
whether lowering serum phosphate levels improves
clinical outcomes. HiLo is a pragmatic cluster-
randomized clinical trial that will test the effects of
targeting a “Hi” or a “Lo” serum phosphate level (≥6.5
vs <5.5 mg/dL) on the hierarchical composite outcome
of all-cause mortality and all-cause hospitalization. HiLo
incorporates multiple pragmatic features including lib-
eral eligibility criteria, cluster randomization, electronic
informed consent, dietitian-implemented interventions,
remote study monitoring, real-world data collection
from existing electronic health records, and a novel
hierarchical composite outcome.
Phosphate binders have been the cornerstone of
hyperphosphatemia management for patients under-
going maintenance hemodialysis for decades, but all
contemporary binders received approval from the US
Food and Drug Administration on the basis of
lowering serum phosphate levels in patients with
Figure 1. Equipoise for HiLo. With multiple factors in favor of and against more aggressive reduction of serum phosphate levels,
there is clinical equipoise to conduct a randomized trial of strict versus liberal management of hyperphosphatemia in patients
with kidney failure undergoing hemodialysis. Abbreviation: CVD, cardiovascular disease. Created with BioRender.com.
AJKD Vol 77 | Iss 6 | June 2021
hyperphosphatemia and kidney failure; no placebo-
controlled randomized trial of any phosphate binder
demonstrated beneficial effects on hard clinical out-
comes. The single outcomes trial that compared
calcium-based versus polymer-based binders yielded
equivocal results.28 Thus, major questions in the field
that affect daily clinical practice remain unanswered.
Do phosphate binders, as currently deployed, improve
survival or other clinical outcomes in patients with
kidney failure undergoing hemodialysis? More funda-
mentally, does aggressive lowering of serum phosphate
levels toward the normal range improve clinical out-
comes in such patients?
Equipoise
Clinical practice guidelines are predicated on the assump-
tion that tight phosphate control will improve clinical
outcomes by attenuating adverse effects of hyper-
phosphatemia,23,24 but it is also possible that patients
who are prescribed strict phosphate control may incur
significant risks that have eluded detection precisely because
of the lack of randomized outcomes trials (Fig 1).29-40
The uncertainty of whether strict phosphate control
improves, worsens, or has no effect on clinical outcomes
in patients with kidney failure undergoing hemodialysis
provides clinical equipoise to perform a randomized out-
comes trial to compare different strategies to manage
hyperphosphatemia.
921

Design of the HiLo Study
Overview of HiLo
HiLo is a pragmatic, multicenter, open-label, cluster-ran-
domized trial that will compare the effects of 2 phosphate
management strategies on the hierarchical composite
outcome of all-cause mortality and all-cause hospitaliza-
tion. HiLo will enroll 4,400 adults with kidney failure
undergoing standard 3-times-weekly maintenance hemo-
dialysis in facilities managed by DaVita Inc and the Uni-
versity of Utah, each of which contributed to the design of
HiLo. Acting as the central institutional review board, the
Duke University Institutional Review Board approved the
study.
HiLo: A Pragmatic Clinical Trial
In contrast to explanatory trials, pragmatic trials are based
in real-world practice and rely on real-world data to
answer real-world clinical questions. Many aspects of he-
modialysis care are ideally suited to large-scale pragmatic
clinical trials.41 Patients are treated 3 times weekly for
several hours during which research activities can occur.
Results of standardized laboratory testing and incident
deaths and hospitalizations are reported in robust elec-
tronic health records that can be the source of trial data.
Routine use of clinical protocols by dialysis personnel fa-
cilitates the weaving of research protocols into the fabric of
Figure 2. Differences between pragmatic and traditional explanatory clinical trials. Pragmatic clinical trials differ from traditional
explanatory trials across multiple domains. Here, we demonstrate how the current pragmatic design of HiLo would differ if it were
conducted as a traditional explanatory trial. Created with BioRender.com.
922
Edmonston et al
daily clinical practice while also providing a mechanism
for rapid translation of trial results into future clinical
practice. These attributes supported execution of the
pragmatic Time to Reduce Mortality in ESRD (TiME) trial
in more than 7,000 patients treated at 266 dialysis facilities
across the United States.42 Like the TiME trial, HiLo in-
corporates many pragmatic features (Fig 2). Unlike a
traditional explanatory trial that might incur $50-$100
million in costs to compare the effects of 2 different
phosphate management strategies on clinical outcomes,
HiLo will be executed for 5%-10% of the cost.
Cluster Randomization
HiLo randomizes dialysis facilities to simplify trial opera-
tions and to ensure that only a single study-specific
phosphate target is implemented in each facility. To
minimize imbalances between the “Hi” and “Lo” arms,
the cluster randomization is stratified by dialysis provider
organization and facility size (Fig 3).
Liberal Eligibility Criteria
For a specific dialysis facility to be eligible to participate in
HiLo, its medical director, dietitian, and lead administrator
must be willing to have their facility randomized to either
the Hi or Lo arm. All adult patients aged 18 years or older
are eligible if they have received maintenance, in-center
hemodialysis 3 times weekly for at least 3 months.
AJKD Vol 77 | Iss 6 | June 2021
Edmonston et al

Figure 3. Cluster versus individual randomization in HiLo. To maintain the fidelity of the intervention using a pragmatic approach, HiLo
employs cluster randomization in which entire dialysis facilities are randomized instead of randomizing patients individually. To balance
the sizes of the Hi and Lo arms, HiLo will stratify the cluster randomization by facility type (Davita Inc or University of Utah) and by
facility size (less or more than the median; latter represented as outlined groups). Created with BioRender.com.

Intervention what degree of hyperphosphatemia would be accepted

HiLo is randomizing dialysis facilities to one of 2 arms.
The Lo arm will have a serum phosphate target of <5.5
mg/dL, which is the current standard of care. Patients in
the Lo arm should experience no change in their man-
agement. The Hi arm will have a serum phosphate target
of ≥6.5 mg/dL, which is the novel intervention to be
tested. Patients in the Hi arm should experience liberalized
diets and less intensive phosphate-binder regimens.
The target serum phosphate level in the Hi arm was
chosen based on results of epidemiologic studies of
phosphate and clinical outcomes, preliminary data
from previous pilot clinical trials, and consideration of
by patients, practitioners, and regulatory bodies.5,43-48
HiLo expects the mean serum phosphate levels to
range between 5.0 and 5.4 mg/dL in the Lo arm and
6.5 and 6.9 mg/dL in the Hi arm. This will result in
achieving HiLo’s goal of a time-averaged separation of
at least 1 mg/dL between arms. Because the vast
majority of patients undergoing hemodialysis exhibit
maintained serum phosphate levels of 4-7 mg/dL,49
sustaining ≥1 mg/dL separation between randomized
treatment arms would represent a time-averaged dif-
ference in serum phosphate exposure of at least 33%
of the modifiable range.

AJKD Vol 77 | Iss 6 | June 2021 923


To mirror usual clinical practice, phosphate-binder
prescriptions and dietary recommendations are left
to the discretion of local care teams. Any phosphate
binder may be used, and there are no HiLo-specific
titration protocols. Likewise, local care teams manage
all other aspects of dialysis independently of guidance
from HiLo.
Outcomes
The primary outcome of HiLo is a hierarchical composite
outcome that prioritizes time to all-cause mortality and
secondarily considers all-cause hospitalization rate when
patients cannot be compared in terms of mortality (details
described in the following). The main prespecified sec-
ondary outcomes are the individual components of the
composite outcome: time to all-cause mortality and total
hospitalization events per total follow-up time.
Limitations and Potential Challenges
HiLo will not account for use of calcium- versus
non–calcium-based phosphate binders, which may influ-
ence clinical outcomes.50 Because hospitalizations are
reported by dialysis facilities when patients miss sessions,
short hospitalizations that do not result in a missed
outpatient dialysis session may escape detection. Further,
HiLo will not adjudicate causes of hospitalizations. Because
of the cluster-randomized design, outbreaks of coronavirus
disease 2019 (COVID-19) within specific dialysis facilities
could reduce power by increasing within-cluster correla-
tions of mortality and hospitalizations. Although the
pragmatic, multicenter design of HiLo will enhance
generalizability, the results may not generalize to mainte-
nance dialysis populations outside the United States. Also,
the trial may fail to achieve ≥1 mg/dL phosphate separa-
tion in the intervention groups.
Novel Aspects of HiLo
Electronic Consent
Informed consent is often waived in pragmatic trials that
test strategies to enhance implementation of proven ther-
apies.51 Because HiLo will test a “more than minimal risk”
intervention that differs from the current opinion–based
standard of care, HiLo must obtain individual-level
informed consent.25 To obtain consent from 4,400 pa-
tients without on-site study staff, HiLo distributes tablet
devices that connect to secure web-based electronic con-
sent (eConsent) modules, one each for the Hi and Lo arms,
on the HiLo web site (hilostudy.org). The eConsent videos
are co-narrated by a nephrologist and a patient with kidney
failure and are accompanied by a suite of concise educa-
tional videos about clinical research participation, phos-
phate and its management in kidney failure, and the “nuts
and bolts” of HiLo. As critical stakeholders, patients
participated in the HiLo Patient Advisory Group, which
was convened by the American Association of Kidney Pa-
tients and provided input on all videos and the design of
924
Edmonston et al
HiLo. To backstop the eConsent, a central team of HiLo
nephrologists is available to answer potential participants’
questions by telephone.
Rolling Recruitment Strategy
HiLo is performing brief, time-limited enrollment of
prevalent patients; for a given facility, 1 week of
enrollment will be allotted per 10 eligible patients. This
“get-in, get-out” approach will speed enrollment by
ensuring that more enrollment time is spent in facilities
with large numbers of eligible patients rather than
maintaining active enrollment in previously enrolled
facilities that slowly accrue smaller numbers of new
patients (typically, less than 10%-20% of the facility’s
census per year). Reducing the time until facilities can
return to usual clinical practice, free from enrollment,
should ease their operational burdens. Rolling recruit-
ment will also enable an ongoing assessment of the
enrollment rate overall and in each arm, which HiLo
will use to guide the number of facilities that ultimately
need to be activated.
Engaging Dietitians
Dialysis dietitians helped design HiLo and are represented
on its steering committee. Dietitians are highly motivated,
scientifically inquisitive caregivers who are ideally posi-
tioned to be the on-the-ground personnel who support
execution of HiLo. Dietitians are employed by dialysis
organizations, are present in all facilities, establish long-
term relationships with their patients through at least
monthly contacts, and often serve as primary decision
makers for titration of phosphate-related treatments. The
strong rapport between dietitians and patients may also
facilitate adherence. Operationally, dietitians approach
patients with the HiLo tablets that are preloaded with all
study materials, including the eConsent modules. The di-
etitians keep a master list that contains the names, eligi-
bility status, approach status, and consent status of each
patient. The Duke Clinical Research Institute site-
management team runs weekly reports with REDCap
software and reconciles with a deidentified version of the
master list to ensure that dietitians are approaching patients
and using the consent module properly. Although di-
etitians facilitate the consent process, they do not obtain
consent and thus are not considered “engaged” in research
from a regulatory standpoint.52 Dietitians help enrolled
patients achieve their serum phosphate targets by using a
combination of phosphate binder and dietary recommen-
dations. Dietitians are not asked to alter any other aspect of
their care, nor are they asked to engage in study-specific
data collection.
Eliminating Case-Report Forms and Traditional
Adverse Event Reporting
The HiLo bioinformatics team created information-
technology portals to receive monthly data transfers from
AJKD Vol 77 | Iss 6 | June 2021
Edmonston et al
DaVita Inc and the University of Utah and collate them into
a single HiLo-specific database that contains demographic
data; laboratory test results; dates of hospitalizations,
transfers to other facilities, or other kidney-replacement
modalities; and death. This bioinformatic solution elimi-
nated the need for case-report forms.
Because most clinically important, serious adverse
events in patients with kidney failure treated by mainte-
nance hemodialysis culminate in hospitalizations, which
are captured in the primary outcome, HiLo is not per-
forming traditional adverse event reporting. Instead, HiLo
monitors laboratory parameters, including serum phos-
phate, calcium, and parathyroid hormone levels. Just as in
usual practice, local care teams are empowered to reduce
or discontinue phosphate binders, for example, in the
event of hypercalcemia, gastrointestinal symptoms, hypo-
phosphatemia, or patient preference.
Remote Monitoring to Maintain Phosphate
Separation
The success of HiLo hinges on maintaining sufficient
separation of serum phosphate concentrations between
arms over time. This is especially challenging in a prag-
matic trial that lacks on-site study personnel or monitors.
To meet this challenge, HiLo delivers monthly reports to
participating dietitians that provide details on achieved
serum phosphate levels in the overall facility and within
each individual study participant, along with local and
study-wide enrollment data. This pragmatic approach will
significantly reduce monitoring costs by directing reme-
diation strategies only to specific facilities and patients that
deviate excessively from targets.
Some patients assigned to the Lo arm will be unable to
maintain their serum phosphate concentration less than
5.5 mg/dL, and, in some patients assigned to the Hi arm,
serum phosphate level will decrease to less than 6.5 mg/
dL. This potential limitation, which is analogous to
incomplete adherence in other trials, will tend to reduce
the overall separation in serum phosphate level between
treatment arms. Like drug intervention trials, HiLo is tar-
geting an average of at least 75% “adherence” with targets
in each arm throughout the study and will conduct sec-
ondary “as-treated” analyses in patients in whom the target
serum phosphate level is achieved in at least 75% of
months in the study.53,54 In addition, the sample size for
HiLo offers 85% rather than 80% power (discussed further
below). This cushion will help offset some loss of power
due to failure to reach serum phosphate level targets in a
proportion of patients in each arm.
Primary End Point and Analysis
Conundrum
All-cause mortality and all-cause hospitalization are plau-
sibly related to and potentially modified by phosphate
control. Avoiding hospitalization is of paramount impor-
tance to all stakeholders in care of patients receiving
AJKD Vol 77 | Iss 6 | June 2021
maintenance hemodialysis, including payers, dialysis
provider organizations, clinicians, and, most importantly,
patients and their families. Indeed, for many patients with
kidney failure, it is more important to enhance quality of
life by avoiding hospitalizations than to prolong survival.55
Despite its clinical importance, constructing a primary
outcome incorporating hospitalization presents several
challenges. If all-cause hospitalization alone was the pri-
mary outcome, death would be a censoring and competing
event. Because some patients with kidney failure die
without prior hospitalizations, censoring deaths in a trial
of phosphate management could distort the results.
Alternatively, hospitalization and mortality could be
combined in a time-to-event composite. However, given
the high rates of hospitalization, often for reasons unre-
lated to phosphate, this approach would yield a trial with
seemingly high power on account of a high event rate,
but, in actuality, power would be decreased by many
noninformative events. A different approach was needed to
address the challenge of combining into a single primary
endpoint time to all-cause mortality and all-cause hospi-
talization rate.
Solution: Hierarchical Composite Outcome
Finkelstein and Schoenfeld developed a method to
accommodate analyses of composite outcomes that include
individual components that are measured on different
scales and can be prioritized differently.56 For HiLo, the
hierarchical endpoint prioritizes all-cause mortality as
the most important endpoint; all-cause hospitalization
will be considered only when patients cannot be compared
in terms of mortality. This method was employed to
secure Food and Drug Administration approval for a
treatment for a rare cardiovascular disease,57 but, to our
knowledge, few, if any, nephrology trials have used this
novel approach.
Figure 4 conceptualizes the analysis of the hierarchical
composite outcome of HiLo, which is based on the
concept of the “win ratio.”57 Each patient in the Hi arm is
compared with each patient in the Lo arm. Follow-up time
for each pairwise comparison is restricted to the amount of
time shared by both patients, which is equivalent to the
shorter follow-up time between the pair. In each pair, the
patients are compared first in terms of mortality. Those
who survive longer are assigned a “win,” and their score is
increased by 1; a “loss” is attributed to the comparators
who died sooner, and their score is reduced by 1. If
neither patient in a pair died during their shared follow-up
time, the comparison shifts to hospitalizations during the
shared follow-up period. In each paired comparison, pa-
tients with more hospitalizations are assigned a loss; the
comparator patients are assigned a win. If the frequency of
hospitalizations is equal, a “tie” results and the matched
pair’s scores remain unchanged. This process is continued
until all permutations of all possible comparisons between
all patients in the Hi and Lo arms are concluded. (Of note,
925
 Edmonston et al

Figure 4. Analysis of the primary hierarchical composite outcome of HiLo. The primary outcome of the HiLo trial is a hierarchical
composite outcome of time to all-cause mortality and number of hospitalizations. In a pairwise manner, the outcomes for each patient
randomized to the Lo and Hi groups are compared. Follow-up duration for each pairwise comparison is defined by the earliest occur-
rence of death or a censoring event (eg, loss to follow-up or end of study) in either comparator patient. The pairs are first compared
on the basis of survival to determine the “winner” who survived longer. If both patients in a pairwise comparison survive throughout
the duration of the shared follow-up period, the winner is the patient who had fewer hospitalizations during follow-up. Note that the
shared follow-up period is the only time relevant to a given pairwise comparison. As a corollary, an individual patient’s follow-up time
will vary across pairwise comparisons depending on the duration of follow-up in their comparators. The final 2 scenarios denote the
continuation of pairwise comparisons for Lo patient 1 with all n remaining Hi patients (second from bottom) and all n remaining Lo
patients with all n Hi patients (bottom) until all comparisons of all Hi versus all Lo patients are complete. After all pairwise compar-
isons are complete, the total scores in the 2 arms are tallied and tested for significant differences. Created with BioRender.com.

926 AJKD Vol 77 | Iss 6 | June 2021

Edmonston et al

individual patients’ outcomes vary across different paired
comparisons because of variable follow-up time in their
comparators [Fig 4].)
After all comparisons are complete, patients’ final scores
describe their overall outcomes compared with the com-
plete population in the other arm of the trial; higher positive
scores indicate better outcomes, and more negative scores
indicate worse outcomes.58 A statistical test is used to test
the null hypothesis that there is no difference in the dis-
tributions of the scores between treatment groups.
Sample Size Determination
Because standard power calculation methods cannot be
used for hierarchical composite endpoints, HiLo per-
formed data simulations to determine power and sample
size. We generated the simulated trial data by using the
Moran algorithm59 based on internal data from DaVita Inc
and the TiME trial.42 We generated the time-to-event data
assuming proportional hazards. We also assumed uniform
accrual to simulate administrative censoring. Other as-
sumptions and parameters are shown in Fig 5. The HiLo
biostatistical team generated 5,000 independently simu-
lated datasets for each of 2 different enrollment patterns:
80 clusters of 55 patients each and 120 clusters of 36
patients each. After calculating the scores for the hierar-
chical composite endpoint using the Finkelstein and
Schoenfeld method, 2 methods were used to analyze the
scores: a parametric method based on repeated-measures
mixed modeling60 and a nonparametric adaptation of the
Wilcoxon rank-sum method for clustering.61 To estimate
power, the number of iterations that had a significant
result (P < 0.05) was divided by the total of 5,000 itera-
tions. As shown in Figure 5, the parametric and
nonparametric methods each yielded ≥85% power.
Statistical Analysis
We will use the clustered Wilcoxon test to analyze the pri-
mary hierarchical outcome in the intention-to-treat popula-
tion.61 For the secondary analysis of all-cause mortality, we
will use a Cox proportional-hazards model with censoring at
study withdrawal, loss to follow-up, and end of follow-up;
we will account for clustering by using the sandwich esti-
mator. For the secondary analysis of all-cause hospitalization,
we will reject the null hypothesis of no difference between
groups if the 95% confidence interval around the estimated
mean difference in hospitalization rates between the arms
excludes zero; to account for clustering, we will inflate the
variance by using the analysis of variance method.62
Limitations of the Hierarchical Composite
Outcome
Potential limitations of the hierarchical endpoint include
how to interpret the results of the analysis of rank scores in
clinical terms. However, the primary analysis of the scores

Figure 5. Sample size and power simulations for HiLo. Because conventional power calculation methods cannot be applied to hi-
erarchical composite outcomes, HiLo used simulated datasets based on data from DaVita Inc and the TiME trial to determine power
and sample size. The figure lists the key assumptions that were used to randomly generate 5,000 iterations of simulated study da-
tabases for each of 2 different study compositions: 80 clusters of 55 patients each and 120 clusters of 36 patients each. Then, 2
analytic approaches were used to analyze the simulated datasets: one parametric (mixed model) and one nonparametric (Wilcoxon
rank-sum method). All 4 approaches yielded power of 85% or higher. Abbreviation: ICC, intraclass correlation coefficient. Created
with BioRender.com.

AJKD Vol 77 | Iss 6 | June 2021 927


should be viewed as the test of whether one arm was su-
perior to the other, whereas the prespecified secondary
analyses will quantify the clinically meaningful magnitude
of effect on each component of the primary outcome.
Indeed, HiLo will have more than 80% power to detect a
hazard ratio of 0.85 for mortality between the Hi and Lo
arms and more than 90% power to detect a hazard ratio of
0.80. Power will be even greater for the secondary analysis
of all-cause hospitalization rate.
Although it would be desirable to capture causes of hos-
pitalization and death, this would require extensive, non-
pragmatic ascertainment and adjudication of data not currently
collected by dialysis facilities. This could be addressed in the
future by merging HiLo data with Medicare claims data, which
are available for the majority of the dialysis population.
Conclusion
Although considered dogma, the currently recommended
serum phosphate target level in patients with kidney failure
treated by maintenance hemodialysis is built on a shaky
foundation. HiLo is using a pragmatic yet rigorous design
that was vetted by key stakeholders—patients, dietitians,
nephrologists, and large dialysis organizations—to compare
clinical outcomes in response to titrating to higher versus
lower serum phosphate target levels. Enrollment began in
2020, and follow-up will extend to 2024. For the study’s
primary objective, HiLo will be a success if it yields evidence
that the Hi or Lo target is superior to the other. What must
be avoided is a noninformative result caused by insufficient
enrollment or insufficient separation of the serum phos-
phate curves. Beyond phosphate management, HiLo will
also be a success if it enhances the clinical trial culture in
nephrology by ushering in wider use of stakeholder
engagement, eConsent, electronic health record data,
remote study monitoring, and innovations in primary
outcomes. By retaining rigor at reduced cost, these advances
could help the nephrology community answer many open
questions with clinical trial–grade evidence.
Article Information
Authors’ Full Names and Academic Degrees: Daniel L.
Edmonston, MD, Tamara Isakova, MD, MMSc, Laura M. Dember,
MD, Steven Brunelli, MD, Amy Young, PhD, Rebecca Brosch, RD,
Srinivasan Beddhu, MD, Hrishikesh Chakraborty, DrPH, and Myles
Wolf, MD, MMSc.
Authors’ Affiliations: Division of Nephrology, Department of
Medicine (DLE, MW), and Duke Clinical Research Institute (DLE,
MW, HC), Duke University School of Medicine, Durham, NC;
Division of Nephrology, Department of Medicine, and Center for
Translational Metabolism and Health, Institute of Public Health and
Medicine, Northwestern University Feinberg School of Medicine,
Chicago, IL (TI); Renal, Electrolyte and Hypertension Division,
Department of Medicine, and Center for Clinical Epidemiology and
Biostatistics, University of Pennsylvania Perelman School of
Medicine, Philadelphia, PA (LMD); DaVita Clinical Research,
DaVita Inc, Minneapolis, MN (SBrunelli, SY, RB); and Division of
Nephrology, Department of Medicine, University of Utah School of
Medicine, Salt Lake City, UT (SBeddhu).
928
Edmonston et al
Address for Correspondence: Myles Wolf, MD, MMSc, 2 Genome
Ct, Suite 1009, Durham, NC 27710. Email: myles.wolf@duke.edu
Authors’ Contributions: Research idea and study design: MW, TI,
DLE, LMD, SBrunelli, AY, RB, SBeddhu, HC, DLE; statistical
analysis: HC; supervision or mentorship: TI, MW. Each author
contributed important intellectual content during manuscript
drafting or revision and agrees to be personally accountable for
the individual’s own contributions and to ensure that questions
pertaining to the accuracy or integrity of any portion of the work,
even one in which the author was not directly involved, are
appropriately investigated and resolved, including with
documentation in the literature if appropriate.
Support: Research reported in this publication was supported by
cooperative agreement UG3/UH3DK118748 from the National
Institute of Diabetes and Digestive and Kidney Diseases (NIDDK).
This work also received logistical and technical support from the
National Institute of Health (NIH) Collaboratory Coordinating
Center through cooperative agreement U24AT009676 with
cofunding by NIDDK. Representatives from the funding agency
contributed to the design of this study.
Financial Disclosure: Dr Wolf reports honoraria from Akebia and
Ardelyx. The remaining authors declare that they have no relevant
financial interests.
Acknowledgements: The Duke Clinical Research Institute is home
to the Data Coordinating Center of HiLo. The HiLo Steering
Committee includes all authors of this report. Additional members
of the HiLo team at the Duke Clinical Research Institute include
Laura Johnson, Davy Andersen, MHA, and Yashika Johnson, MS
(Clinical Operations); Andrew MacKelfresh, MBA, and James
Topping (Informatics); and Peter Merrill, PhD (Biostatistics). We
thank the leadership of the American Association of Kidney
Patients, including Richard Knight, MBA, President, and Diana
Clynes, Executive Director, for their support for HiLo and for
convening the members of the HiLo Patient Advisory Group: Malie
Robb (deceased), Scott Burton, and Dale Rogers.
Disclaimer: The content is solely the responsibility of the authors
and does not necessarily represent the official views of the NIH.
Peer Review: Received May 12, 2020. Evaluated by 3 external peer
reviewers, with editorial input from a Statistics/Methods Editor and
an Acting Editor-in-Chief (Editorial Board Member Chirag R.
Parikh, MD, PhD). Accepted in revised form October 1, 2020. The
involvement of an Acting Editor-in-Chief to handle the peer-review
and decision-making processes was to comply with AJKD’s
procedures for potential conflicts of interest for editors, described
in the Information for Authors & Journal Policies.
References
1. Saran R, Robinson B, Abbott KC, et al. US Renal Data System
2018 Annual Data Report: epidemiology of kidney disease in the
United States. Am J Kidney Dis. 2019;73(3)(suppl 1):A7-A8.
2. Wanner C, Krane V, Marz W, et al. Atorvastatin in patients with
type 2 diabetes mellitus undergoing hemodialysis. N Engl J
Med. 2005;353(3):238-248.
3. Fellstrom BC, Jardine AG, Schmieder RE, et al. Rosuvastatin
and cardiovascular events in patients undergoing hemodialysis.
N Engl J Med. 2009;360(14):1395-1407.
4. Baigent C, Landray MJ, Reith C, et al. The effects of lowering
LDL cholesterol with simvastatin plus ezetimibe in patients with
chronic kidney disease (Study of Heart and Renal Protection):
a randomised placebo-controlled trial. Lancet. 2011;377(9784):
2181-2192.
5. Block GA, Hulbert-Shearon TE, Levin NW, Port FK. Association
of serum phosphorus and calcium x phosphate product with
AJKD Vol 77 | Iss 6 | June 2021
Edmonston et al
mortality risk in chronic hemodialysis patients: a national study.
Am J Kidney Dis. 1998;31(4):607-617.
6. Block GA, Klassen PS, Lazarus JM, Ofsthun N, Lowrie EG,
Chertow GM. Mineral metabolism, mortality, and morbidity in
maintenance hemodialysis. J Am Soc Nephrol. 2004;15(8):
2208-2218.
7. Stevens LA, Djurdjev O, Cardew S, Cameron EC, Levin A.
Calcium, phosphate, and parathyroid hormone levels in com-
bination and as a function of dialysis duration predict mortality:
evidence for the complexity of the association between mineral
metabolism and outcomes. J Am Soc Nephrol. 2004;15(3):
770-779.
8. Rodriguez-Benot A, Martin-Malo A, Alvarez-Lara MA,
Rodriguez M, Aljama P. Mild hyperphosphatemia and mortality
in hemodialysis patients. Am J Kidney Dis. 2005;46(1):68-77.
9. Kalantar-Zadeh K, Kuwae N, Regidor DL, et al. Survival pre-
dictability of time-varying indicators of bone disease in main-
tenance hemodialysis patients. Kidney Int. 2006;70(4):771-
780.
10. Jono S, McKee MD, Murry CE, et al. Phosphate regulation of
vascular smooth muscle cell calcification. Circ Res.
2000;87(7):e10-e17.
11. Moe SM, Chen NX. Mechanisms of vascular calcification in
chronic kidney disease. J Am Soc Nephrol. 2008;19(2):213-
216.
12. Goodman WG, Goldin J, Kuizon BD, et al. Coronary-artery
calcification in young adults with end-stage renal disease who are
undergoing dialysis. N Engl J Med. 2000;342(20):1478-1483.
13. Blacher J, Guerin AP, Pannier B, Marchais SJ, London GM.
Arterial calcifications, arterial stiffness, and cardiovascular risk
in end-stage renal disease. Hypertension. 2001;38(4):938-
942.
14. Guerin AP, Pannier B, Metivier F, Marchais SJ, London GM.
Assessment and significance of arterial stiffness in patients
with chronic kidney disease. Current Opin Nephrol Hypertens.
2008;17(6):635-641.
15. Silberberg JS, Barre PE, Prichard SS, Sniderman AD. Impact
of left ventricular hypertrophy on survival in end-stage renal
disease. Kidney Int. 1989;36(2):286-290.
16. Gutierrez OM, Mannstadt M, Isakova T, et al. Fibroblast growth
factor 23 and mortality among patients undergoing hemodial-
ysis. N Engl J Med. 2008;359(6):584-592.
17. Isakova T, Xie H, Yang W, et al. Fibroblast growth factor 23
and risks of mortality and end-stage renal disease in patients
with chronic kidney disease. JAMA. 2011;305(23):2432-
2439.
18. Faul C, Amaral AP, Oskouei B, et al. FGF23 induces left ven-
tricular hypertrophy. J Clin Invest. 2011;121(11):4393-4408.
19. Scialla JJ, Xie H, Rahman M, et al. Fibroblast growth factor-23
and cardiovascular events in CKD. J Am Soc Nephrol.
2014;25(2):349-360.
20. Mehta R, Cai X, Lee J, et al. Association of fibroblast growth
factor 23 with atrial fibrillation in chronic kidney disease, from
the Chronic Renal Insufficiency Cohort Study. JAMA Cardiol.
2016;1(5):548-556.
21. Strozecki P, Adamowicz A, Nartowicz E, Odrowaz-Sypniewska G,
Wlodarczyk Z, Manitius J. Parathormon, calcium, phosphorus, and
left ventricular structure and function in normotensive hemodial-
ysis patients. Ren Fail. 2001;23(1):115-126.
22. Bogin E, Massry SG, Harary I. Effect of parathyroid hormone on
rat heart cells. J Clin Invest. 1981;67(4):1215-1227.
23. National Kidney Foundation. K/DOQI clinical practice guide-
lines for bone metabolism and disease in chronic kidney dis-
ease. Am J Kidney Dis. 2003;42(4)(suppl 3):S1-S201.
AJKD Vol 77 | Iss 6 | June 2021
24. Kidney Disease: Improving Global Outcomes (KDIGO) CKD-
MBD Work Group. KDIGO Clinical Practice Guideline for
the Diagnosis, Evaluation, Prevention, and Treatment of
Chronic Kidney Disease-Mineral and Bone Disorders (CKD-
MBD). Kidney Int Suppl. 2009;113:S1-S130.
25. Kidney Disease: Improving Global Outcomes (KDIGO) CKD-
MBD Work Group. KDIGO 2017 Clinical Practice Guideline
Update for the Diagnosis, Evaluation, Prevention, and Treat-
ment of Chronic Kidney Disease-Mineral and Bone Disorder
(CKD-MBD). Kidney Int Suppl (2011). 2017;7(1):1-59.
26. Tonelli M, Pannu N, Manns B. Oral phosphate binders in pa-
tients with kidney failure. N Engl J Med. 2010;362(14):1312-
1324.
27. Lopes AA, Tong L, Thumma J, et al. Phosphate binder use and
mortality among hemodialysis patients in the Dialysis Outcomes
and Practice Patterns Study (DOPPS): evaluation of possible
confounding by nutritional status. Am J Kidney Dis.
2012;60(1):90-101.
28. Suki WN, Zabaneh R, Cangiano JL, et al. Effects of sevelamer
and calcium-based phosphate binders on mortality in hemodi-
alysis patients. Kidney Int. 2007;72(9):1130-1137.
29. Block GA, Raggi P, Bellasi A, Kooienga L, Spiegel DM. Mor-
tality effect of coronary calcification and phosphate binder
choice in incident hemodialysis patients. Kidney Int.
2007;71(5):438-441.
30. Jamal SA, Vandermeer B, Raggi P, et al. Effect of calcium-
based versus non-calcium-based phosphate binders on
mortality in patients with chronic kidney disease: an updated
systematic review and meta-analysis. Lancet. 2013;382(9900):
1268-1277.
31. Nakanishi T, Hasuike Y, Nanami M, Yahiro M, Kuragano T. Novel
iron-containing phosphate binders and anemia treatment in
CKD: oral iron intake revisited. Nephrol Dial Transplant.
2016;31(10):1588-1594.
32. Hutchison AJ, Wilson RJ, Garafola S, Copley JB. Lanthanum
carbonate: safety data after 10 years. Nephrology (Carlton).
2016;21(12):987-994.
33. Shinaberger CS, Greenland S, Kopple JD, et al. Is controlling
phosphorus by decreasing dietary protein intake beneficial or
harmful in persons with chronic kidney disease? Am J Clin
Nutr. 2008;88(6):1511-1518.
34. Ruospo M, Palmer SC, Natale P, et al. Phosphate binders for
preventing and treating chronic kidney disease-mineral and
bone disorder (CKD-MBD). Cochrane Database Syst Rev.
2018;8:CD006023.
35. Chiu YW, Teitelbaum I, Misra M, de Leon EM, Adzize T,
Mehrotra R. Pill burden, adherence, hyperphosphatemia, and
quality of life in maintenance dialysis patients. Clin J Am Soc
Nephrol. 2009;4(6):1089-1096.
36. Fissell RB, Karaboyas A, Bieber BA, et al. Phosphate binder pill
burden, patient-reported non-adherence, and mineral bone
disorder markers: Findings from the DOPPS. Hemodial Int.
2016;20(1):38-49.
37. Qureshi AR, Alvestrand A, Divino-Filho JC, et al. Inflammation,
malnutrition, and cardiac disease as predictors of mortality in
hemodialysis patients. J Am Soc Nephrol. 2002;13(suppl 1):
S28-S36.
38. Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC,
Barre PE. Hypoalbuminemia, cardiac morbidity, and mortality in
end-stage renal disease. J Am Soc Nephrol. 1996;7(5):728-
736.
39. Ghimire S, Castelino RL, Lioufas NM, Peterson GM, Zaidi ST.
Nonadherence to medication therapy in haemodialysis patients:
a systematic review. PLoS One. 2015;10(12):e0144119.
929

40. How PP, Fischer JH, Arruda JA, Lau AH. Effects of lanthanum
carbonate on the absorption and oral bioavailability of cipro-
floxacin. Clin J Am Soc Nephrol. 2007;2(6):1235-1240.
41. Dember LM, Archdeacon P, Krishnan M, et al. Pragmatic trials
in maintenance dialysis: perspectives from the Kidney Health
Initiative. J Am Soc Nephrol. 2016;27(10):2955-2963.
42. Dember LM, Lacson E Jr, Brunelli SM, et al. The TiME trial: a fully
embedded, cluster-randomized, pragmatic trial of hemodialysis
session duration. J Am Soc Nephrol. 2019;30(5):890-903.
43. Ganesh SK, Stack AG, Levin NW, Hulbert-Shearon T, Port FK.
Association of elevated serum PO(4), Ca x PO(4) product, and
parathyroid hormone with cardiac mortality risk in chronic he-
modialysis patients. J Am Soc Nephrol. 2001;12(10):2131-
2138.
44. Slinin Y, Foley RN, Collins AJ. Calcium, phosphorus, para-
thyroid hormone, and cardiovascular disease in hemodialysis
patients: the USRDS waves 1, 3, and 4 study. J Am Soc
Nephrol. 2005;16(6):1788-1793.
45. Naves-Diaz M, Passlick-Deetjen J, Guinsburg A, et al. Calcium,
phosphorus, PTH and death rates in a large sample of dialysis
patients from Latin America. The CORES Study. Nephrol Dial
Transplant. 2011;26(6):1938-1947.
46. Floege J, Kim J, Ireland E, et al. Serum iPTH, calcium and
phosphate, and the risk of mortality in a European haemodial-
ysis population. Nephrol Dial Transplant. 2011;26(6):1948-
1955.
47. Wald R, Rabbat CG, Girard L, et al. Two phosphAte taRGets in
End-stage renal disease Trial (TARGET): a randomized
controlled trial. Clin J Am Soc Nephrol. 2017;12(6):965-973.
48. Bhargava R, Kalra PA, Hann M, Brenchley P, Hurst H,
Hutchison AJ. A randomized controlled trial of different serum
phosphate ranges in subjects on hemodialysis. BMC Nephrol.
2019;20(1):37.
49. Rivara MB, Ravel V, Kalantar-Zadeh K, et al. Uncorrected and
albumin-corrected calcium, phosphorus, and mortality in pa-
tients undergoing maintenance dialysis. J Am Soc Nephrol.
2015;26(7):1671-1681.
50. Patel L, Bernard LM, Elder GJ. Sevelamer versus calcium-
based binders for treatment of hyperphosphatemia in CKD: A
meta-analysis of randomized controlled trials. Clin J Am Soc
Nephrol. 2016;11(2):232-244.
930
Edmonston et al
51. Anderson ML, Califf RM. Sugarman J; participants in the NIH
Health Care Systems Research Collaboratory Cluster Ran-
domized Trial Workshop. Ethical and regulatory issues of
pragmatic cluster randomized trials in contemporary health
systems. Clin Trials. 2015;12(3):276-286.
52. US Department of Health and Human Services. Federal Policy
for the Protection of Human Subjects. Fed Regist.
2017;802015:53933-54061. To be codified at 45 CFR x46.
53. Erkens PM, ten Cate H, Buller HR, Prins MH. Benchmark for
time in therapeutic range in venous thromboembolism: a sys-
tematic review and meta-analysis. PLoS One. 2012;7(9):
e42269.
54. White HD, Gruber M, Feyzi J, et al. Comparison of outcomes
among patients randomized to warfarin therapy according to
anticoagulant control: results from SPORTIF III and V. Arch
Intern Med. 2007;167(3):239-245.
55. Evangelidis N, Tong A, Manns B, et al; Standardized Out-
comes in Nephrology–Hemodialysis (SONG-HD) Initiative.
Developing a set of core outcomes for trials in hemodialysis:
an international Delphi survey. Am J Kidney Dis. 2017;70(4):
464-475.
56. Finkelstein DM, Schoenfeld DA. Combining mortality and lon-
gitudinal measures in clinical trials. Stat Med. 1999;18(11):
1341-1354.
57. Maurer MS, Schwartz JH, Gundapaneni B, et al. Tafamidis
treatment for patients with transthyretin amyloid cardiomyopa-
thy. N Engl J Med. 2018;379(11):1007-1016.
58. Pocock SJ, Ariti CA, Collier TJ, Wang D. The win ratio: a new
approach to the analysis of composite endpoints in clinical trials
based on clinical priorities. Eur Heart J. 2012;33(2):176-182.
59. Moran PA. Testing for correlation between non-negative vari-
ates. Biometrika. 1967;54(3):385-394.
60. Li Yuelin, Baron J. Linear mixed-effects models in cluster-
randomized studies. In: Yuelin Li, Baron J, eds. Behavioral
Research Data Analysis with R. Springer; 2012:177-204.
61. Rosner B, Glynn RJ, Lee ML. Incorporation of clustering effects
for the Wilcoxon rank sum test: a large-sample approach.
Biometrics. 2003;59(4):1089-1098.
62. Ukoumunne OC. A comparison of confidence interval methods
for the intraclass correlation coefficient in cluster randomized
trials. Stat Med. 2002;21(24):3757-3774.
AJKD Vol 77 | Iss 6 | June 2021
Edmonston et al

HiLo: A Pragmatic, Randomized Trial of Phosphate Management

for Patients on Maintenance Hemodialysis
Setting & Participants Intervention Novel Design Features

‘Hi’ phosphate target Extensive stakeholder engagement

(≥6.5 mg/dl) with patients, dietitians, nephrologists
w

Pragmatic, vs Hierarchical composite outcome of

H
cluster-randomized trial ‘Lo’ phosphate target all-cause mortality & hospitalizations
al
a
(<5.5 mg/dl)
Pragmatic trial with liberal eligibility
P
Pr
criteria
cr
Follow-up: 27-45 months
_ Electronic informed consent
El
4,400 patients receiving Interventions to reach phosphate ((eConsent)
thrice-weekly hemodialysis targets at the discretion of the
in 80-120 dialysis facilities dietitians & providers Real-world data collection from EHR
R

CONCLUSION: HiLo will address the question of what serum phosphate

h h target to use in
CON
hemodialysis while advancing methods for pragmatic clinical trials in nephrology.
Daniel L. Edmonston, Tamara Isakova, Laura M. Dember, et al (2020)
@AJKDonline | DOI: 10.1053/j.ajkd.2020.10.008

AJKD Vol 77 | Iss 6 | June 2021 930.e1

Original Investigation

Psychosocial Factors, Intentions to Pursue

Arteriovenous Dialysis Access, and Access Outcomes: A
Cohort Study
Jace Ming Xuan Chia, Zhong Sheng Goh, Pei Shing Seow, Terina Ying-Ying Seow, Jason Chon Jun Choo,
Marjorie Wai-Yin Foo, Stanton Newman, and Konstadina Griva

Rationale & Objective: Suboptimal dialysis
preparation of patients with chronic kidney dis-
ease (CKD) is common, but little is known about
its relationship to psychosocial factors. This study
aimed to assess patients’ attitudes about access
creation and to identify factors associated with
patients’ intentions regarding dialysis access
creation and outcomes.
Study Design: Prospective cohort study.
Setting & Participants: 190 patients with stage
4/5 CKD not receiving dialysis treated at 2 hos-
pitals in Singapore and 128 of their family
members.
Predictors: Self-reported measures of illness
perception, health-related quality of life, and
attitudes toward access creation.
Sociodemographic and clinical measures were
also obtained.
Outcome: Intention to create an arteriovenous
fistula (AVF; ie, proceed with access vs wait and
see) and time to creation of a functional AVF.
Analytical Approach: Exploratory factor analysis
(EFA) was undertaken to construct internally
consistent subscales for a newly developed
questionnaire about attitudes toward access
creation. Logistic regression and cause-specific Complete author and article
hazards models were conducted to identify information provided before
references.
psychosocial factors associated with patients’
access creation intentions and access Correspondence to K. Griva
outcomes, respectively. (konstadina.griva@ntu.edu.
sg)
Results: EFA (explained 50.1% variance)
Am J Kidney Dis.
revealed 4 domains: access and dialysis con- 77(6):931-940. Published
cerns, need for dialysis, worry about cost, and online December 3, 2020.
value of access. A high risk of intention to delay doi: 10.1053/
access creation (51.1%) was found among pa- j.ajkd.2020.09.019
tients despite early referral and education. Multi-
© 2020 by the National
variable analysis (R2 = 0.45) showed that the
Kidney Foundation, Inc.
intention to proceed with access creation was Published by Elsevier Inc.
associated with greater perceived value from All rights reserved.
access (odds ratio, 2.61; 95% CI, 1.46-4.65;
P < 0.001).
Limitations: Limited generalization, as only those
already receiving nephrology care were studied.
Conclusions: Approximately half of the patients
studied planned to delay access creation. The
questionnaire developed to evaluate attitudes
about access creation may help identify in-
dividuals for whom decision-support programs
would be useful. These findings highlight the
need to understand and address patients’
concerns about access creation.

A growing concern in chronic kidney disease (CKD)

care is patient delay in decision-making about kidney
replacement therapy (KRT) modalities and urgent initia-
tion of hemodialysis without a permanent access.1,2 The
2006 NKF-KDOQI (National Kidney Foundation–Kidney
Disease Outcomes Quality Initiative) clinical practice
guideline for vascular access recommended that patients
engage in early preparation of an arteriovenous fistula
(AVF) as permanent vascular access for hemodialysis.3 AVF
use as opposed to central venous catheter (CVC) use is
particularly important given its associations with better
clinical outcomes and reduced health costs.4 However,
despite the evidence in support of timely access creation
and ensuing improvements in early referral and accessi-
bility of predialysis care, rates of suboptimal hemodialysis
initiation (ie, unplanned initiation with a CVC) remain
alarmingly high worldwide, such as 57% in the United
Kingdom and 80% in the United States.5,6
Individual- and system-level factors determine when
and how patients make renal care decisions. System-level
factors such as late referral to renal care or scheduling
delay for surgical review and access creation are important
barriers. Yet, these alone cannot fully explain the high rates
of suboptimal initiation for those already undergoing renal
care and exposed to predialysis education.7,8 Individual
beliefs and attitudes as outlined in theoretical frameworks
(ie, health belief model9 and the common sense model of
illness representations10) appear to play a key role in
treatment decision-making.
Decisions related to dialysis (ie, opting for dialysis,
choosing between modalities, and access preparation) are
inherently difficult and characterized by decisional con-
flict.11 Qualitative studies have identified attitudes toward
CKD and treatment potentially implicated in decision-
making. These include perceived effectiveness of dialysis,
fear of dialysis, concerns with fistulas, and attribution of
symptoms to non-CKD causes.12-14 The importance of
these cognitive and affective factors warrants greater
attention to elucidate their potential contributions to CKD
decision-making.
Although qualitative retrospective work has identified
important constructs related to patients’ personal

AJKD Vol 77 | Iss 6 | June 2021 931


PLAIN-LANGUAGE SUMMARY
Many patients start hemodialysis with temporary
vascular access despite regular kidney care and pre-
dialysis education. Delay is often related to patient
choice, but research on patients’ perspectives is limited
and no measure of attitudes toward hemodialysis
preparation presently exists. In this study, we surveyed
predialysis patients and their family members about
their perceptions of chronic kidney disease and their
intentions to undergo access creation. We also report on
a new survey instrument to measure attitudes toward
hemodialysis preparation. Domains covered in the in-
strument included perceptions about the value of
vascular access as well as concerns, including those
about the need for dialysis and costs. Beliefs about value
of vascular access predicted patients’ intentions to pre-
pare for hemodialysis as well as their access outcomes.
understanding of disease and treatment, limitations
remain.13-15 Qualitative findings do not examine the
strength of associations between variables and health be-
haviors; retrospective data are open to recall bias and may
not generalize to patients with CKD at the time decisions
about dialysis/access are made. Additionally, no validated
tool is presently available to assess attitudes toward dialysis
and access despite evidence in other patient populations
supporting the role of treatment-specific attitudes in pa-
tient behavior.16
Understanding the key psychosocial factors associated
with dialysis access creation behaviors and constructing a
measure to quantify them is important, as these are
modifiable targets for intervention. To address these gaps,
this study aims to (1) develop a tool to measure attitudes
toward dialysis access creation and (2) identify factors that
explain patients’ access creation intentions and outcomes
among a set of theory- and evidence-informed parameters.
Methods
Participants
Eligible patients and their family members were recruited
between 2016 and 2019 from outpatient renal clinics at 2
hospitals in Singapore based on several criteria: (1) pa-
tients or family members of patients with stage 4/5 CKD
not receiving KRT, (2) patients receiving nephrology care
and attended at least 1 renal counseling session for dialysis
preparation in the preceding 3 months, and (3) patients
aged at least 21 years and (4) literate in English, Mandarin,
or Malay. Exclusion criteria included (1) patients (or
family members of patients) already undergoing KRT or
who opted for conservative management and (2) inability
to provide consent and complete assessments for cognitive
or language reasons. Renal health care providers identified
932
Chia et al
potential participants for recruitment. The study was
approved by the National Healthcare Group Domain Spe-
cific Review Board (ref. 2015/01225) and SingHealth
Centralized Institutional Review Board (ref. 2016/2979).
All participants provided written informed consent.
Study Data
Dialysis Access Attitudes Questionnaire
Drawing on the Patient-Reported Outcomes Measurement
Information System (PROMIS) framework,17 we devel-
oped and evaluated the Dialysis Access Attitudes Ques-
tionnaire (DAAQ) in a 6-part iterative process. Steps 1-4
were undertaken before this study (2016-2018) and
informed the development of the initial DAAQ item pool:
(1) literature review, (2) qualitative interviews, (3) in-
strument review and item screening by experts, and (4)
cognitive pretest with patients and family members. The
literature review (including validated survey tools and
published works) was conducted to identify existing
questionnaires and key issues related to access creation,
dialysis initiation, and decision-making for KRT using the
following keywords: access, creation, vascular, fistula,
barriers, facilitators, beliefs, perceptions, patient(s), treat-
ment delay, hemodialysis, renal replacement therapies,
decision making, survey, and questionnaire. Vascular ac-
cess instruments identified focused on the impact of access
already in place (ie, symptoms, quality of life [QoL]) or
access self-care and were intended for use with patients
already undergoing hemodialysis with an AVF or CVC.18-21
No instrument measuring dialysis access attitudes relevant
to access-related decision-making behaviors before KRT
initiation was found. Items were compiled based on this
review and dialysis access surveys used in prior studies.
Qualitative data from interviews with 96 patients with
CKD, family members, and renal health care providers
were used to expand the item pool to k = 35 for testing.22
Screening for relevance, clarity, and redundancy by 3 renal
experts, 1 researcher, and 1 patient representative elimi-
nated 11 items. Step 4 comprised a cognitive pretest with 5
patients and 2 family members to assess item compre-
hension (ie, ease of understanding, wording ambiguity,
and face validity). Three additional items were dropped,
resulting in a preliminary 21-item version of the DAAQ.
Content was modified to achieve a target grade 6 reading
level.
DAAQ items assessed symptom attribution, perceived
necessity of dialysis and access creation, and perceived
benefits and barriers related to access. Example items
include “I worry about the effects of having the fistula/PD
[peritoneal dialysis] catheter on my everyday life” and “I
see several advantages/benefits of preparing the fistula
early.” Items presented to family members were suitably
modified with “I” and “my” changed to “the patient” and
“the patient’s,” respectively (eg, “I worry about the effects
of having the fistula/PD catheter on the patient’s everyday
life”) to elicit their views on access. The items assessed
AJKD Vol 77 | Iss 6 | June 2021
Chia et al
family members’ own perspectives rather than their per-
spectives as a proxy for the patient. All items were rated on
a 5-point Likert scale (1, strongly agree; 2, agree; 3,
neutral; 4, disagree; 5, strongly disagree). Step 5 involved
survey administration of the preliminary 21-item DAAQ,
followed by factor analysis for the final DAAQ refinement
and dimensionality. Step 6 involved evaluation of criterion
validity of this final DAAQ item set against study outcomes.
Clinical and Sociodemographic Variables
Data on sociodemographic and clinical characteristics were
obtained. Participants reported on gender, age, ethnicity,
relationship status, employment status, and household
income levels. Family members’ relation to the patient and
whether the family member was the primary caregiver
were noted. Patients’ medical records were reviewed to
collate information on laboratory values and estimated
glomerular filtration rate (eGFR) using the isotope-dilution
mass spectrometry–traceable 4-variable Modification of
Diet in Renal Disease (MDRD) Study equation23 and details
of nephrology care (ie, referral date, number of KRT
counseling sessions with a renal coordinator). Comorbid
burden was measured with the age-adjusted Charlson co-
morbidity index,24 in which higher scores indicate greater
burden.
Illness Perception
The Brief Illness Perception Questionnaire (BIPQ)25 was
used to measure CKD illness perceptions: consequences,
timeline, personal control, treatment control, identity,
concern, coherence, and emotional representation. Each
item was scored on a scale with a range of 1 to 10. The
BIPQ has been shown to have good psychometric prop-
erties and used with renal patients.25,26 As recommended
by authors and widely implemented in prior applications,
items were made disease-specific by replacing the term
“illness” with “kidney disease.”25,26
Health-Related QoL
Health-related QoL was measured by using the Kidney
Disease Quality of Life Short Form, which comprises 5
kidney disease-specific domains (ie, symptom burden,
effects of kidney disease, burden of kidney disease, patient
satisfaction, and staff encouragement), and the 12-Item
Short Form Health Survey (ie, general health, physical
functioning, physical role limitation, emotional role lim-
itation, emotional well-being, bodily pain, vitality, and
social functioning). Standard scoring methods27,28 were
used to compute subscale scores and the derived composite
scores (ie, physical component summary and mental
component summary) ranging from 0 to 100. Higher
scores indicated better QoL.
Study Outcomes
Primary outcomes were intention for and timely creation
of a functional AVF. Intention was measured with a single
AJKD Vol 77 | Iss 6 | June 2021
dichotomous item whereby respondents self-reported
their intention regarding access creation (proceed with
access vs wait and see). Access creation events (ie, AVF
creation, type of access and modality on KRT initiation),
censoring events (ie, loss to follow-up and end of follow-
up), and competing events (ie, hemodialysis initiation
with CVC, modality change, and death) were extracted
from medical records during the study observation win-
dow (ie, study entry to March 2020). Hemodialysis
initiation with CVC was considered a censoring event, as it
signaled suboptimal initiation of hemodialysis. Patients
with low eGFR (<10 mL/min/1.73 m2) at the time of AVF
surgery and whose AVF creation shortly preceded hemo-
dialysis initiation with a CVC were considered to not have
had a functional AVF created in time. Access outcomes for
patients related to the family-member participants were
not recorded unless patients too had enrolled and
consented.
Statistical Analyses
A quantitative evaluation of the newly developed DAAQ was
conducted to establish reliability and factorial integrity.
Exploratory factor analysis (EFA) using Promax rotation was
performed to derive factor structure of the DAAQ based on
an eigenvalue >1 cutoff and scree plot. EFA was based on
complete data cases, with no missing data imputation. Data
were screened for factorability based on the Kaiser-Meyer-
Olkin measure, Bartlett’s test of sphericity, and anti-image
correlational matrix. Mean subscale scores were computed
based on final factor loadings. Internal consistency was
assessed using Cronbach’s α. Items with unclear factor
loading patterns (ie, primary loading <0.4 with cross-
loading >0.3) and low communalities were eliminated.
Item-total correlations were computed within each subscale,
and items were dropped if they had low values and their
elimination led to a moderate increase in Cronbach’s α. EFA
was run on the combined patient and family member
dataset to ensure that it could be administered to both
groups. This was guided by the qualitative work for DAAQ
on the role of family input on access-related decisions.
Confirmatory factor analyses were run on patient data alone
to confirm the factorial structure of DAAQ.
Unadjusted and adjusted logistic regressions were
conducted to identify predictors of patients’ intentions
toward access creation. Unadjusted and adjusted odds ra-
tios (ORs), 95% CIs, and Nagelkerke pseudo-R2 statistics
were reported. Nagelkerke R2 indicates the goodness of fit
of the model to the data, with higher values indicating
better fit.29 ORs of 1.5, 2.5, and 4 indicated small, me-
dium, and large effect sizes, respectively.30 Tolerance was
computed, whereby a value <0.2 suggests no issue with
multicollinearity.31 Time to event was measured from
baseline assessment to one of the following: outcome of
interest (ie, functional AVF creation/hemodialysis with
AVF), death, initiation of PD, initiation of hemodialysis
with a CVC, loss to follow-up, or end of follow-up (March
933
 Chia et al

Table 1. Participant Sociodemographic Data and Clinical Profile

Characteristics Patients (n = 190) Family Members (n = 128)
Age, y 62.8 ± 10.8 48.1 ± 14.2
Female sex 41.1% 65.6%
Ethnicity
Chinese 60.0% 63.3%
Malay 27.9% 27.3%
Indian 8.4% 5.5%
Others 3.7% 3.9%
Employment status
Employed 37.9% 71.9%
Retired 40.0% 8.6%
Unemployed but job-seeking 11.1% 4.7%
Others (including homemaker) 1.1% 12.5%
Work abilitya
Able to work full time 34.7% 78.9%
Able to work part time 17.9% 7.0%
Unable to work 47.4% 12.5%
Missing data – 1.6%
Monthly household incomeb
<$2,000 38.4% 19.5%
$2,000-$4,999 21.6% 31.3%
$5,000-$9,999 11.1% 20.3%
$10,000-$17,999 6.8% 7.8%
$18,000 and above 2.1% 3.9%
Do not know/do not wish to answer 15.3% 17.2%
Missing data 2.6% –
Relationship status
Married or engaged 71.1% 70.3%
Divorced or widowed 20.8% 5.5%
Single 7.9% 24.2%
Relation to patient
Spouse – 30.5%
Child – 61.7%
Sibling/parent – 3.9%
Others – 3.9%
Primary caregiver – 69.5%
Patient-family member dyads 58 (30.5%) 58 (45.3%)
Clinical markers
eGFR 10.79 ± 4.72 –
Charlson comorbidity index 6.84 ± 1.86 –
No. of KRT counseling sessions 2.12 ± 0.97 –
Access creation outcome at follow-up
Functional AVF prepared for/used at HD initiation 45 (23.7%) –
Started HD with temporary accessc 66 (34.7%) –
No directive to prepare for AVF creation yet 12 (6.3%) –
Initiated PD/planning for PD 44 (23.2%) –
Died/lost to follow-up/conservative care 23 (12.1%) –
Values for continuous variables given as mean ± standard deviation; for categorical variables, as percentage or count (percentage).
Abbreviations: AVF, arteriovenous fistula; eGFR, estimated glomerular filtration rate; HD, hemodialysis; KRT, kidney replacement therapy; PD, peritoneal dialysis.
a
Missing data, n = 2.
b
Missing data, n = 4.
c
Includes those with nonfunctional AVF.

2020). Cause-specific hazards models, in which competing and clinical variables in block 1 to control for known and
events were treated as censoring, were used to estimate potential confounds,7,8 followed by selected psychosocial
hazard ratios (HRs) and their 95% CIs. Multivariable variables in block 2 (using likelihood-ratio tests with sig-
models were constructed as follows: sociodemographic nificance level of P < 0.05 in each unadjusted model).

934 AJKD Vol 77 | Iss 6 | June 2021

Chia et al

DAAQ domains were included in block 2 to assess their
contribution to prognostication of AVF creation outcomes.
In sensitivity analyses, the univariate associations be-
tween family members’ DAAQ results and their own in-
tentions, as well as patient study outcomes (intentions and
access creation), were examined. For patient outcomes,
sensitivity analyses were possible only for consenting
dyads. Multivariable analysis was not conducted as a result
of small numbers of consenting dyads. All analyses were
conducted using SPSS software (version 25; IBM Corp)
with an α level of 0.05.
Results
Study Sample
Of the 335 patients and 165 family members approached,
190 patients (56.7%) and 128 family members (77.6%)
consented to participate (318 of 500 [63.8%] overall).
Those who declined most often cited lack of interest or
willingness to discuss dialysis. A small proportion (n = 45)
were deemed ineligible as a result of inability to consent
because of confusion, limited comprehension or language,
frail health, or planning for conservative management or
kidney transplantation. Fifty-eight patients (30.5%) had
another family member in the study (Table 1). Mean age
of patients was 62.8 ± 10.8 (standard deviation) years,
higher than family members’ mean age of 48.1 ± 14.2
years. Most family members were the patient’s children or
spouse and identified themselves as the primary caregiver.
Psychometric Evaluation and DAAQ Dimensionality
The preliminary k = 21 items were administered to pa-
tients and family members. Three patient-specific items, 1
PD-related item, and 1 item that was phrased differently in
patient and family-member questionnaires were excluded
from the EFA. The remaining k = 16 items that assessed
perceptions of hemodialysis and were presented to patients
and family members were identified for the EFA. Sixty-
three patients and 29 family members with missing data
on DAAQ items were not included in the EFA. This mainly
comprised those who only considered PD.
Complete DAAQ datasets of 127 patients and 99 family
members were used in the EFA to derive the final DAAQ
factor structure. Diagnostic criteria for factorability were
met. EFA identified 4 factors accounting for 50.1% of the
total variance (Table 2 provides final factor loadings and
other statistics). Four extraneous items with low item-total
correlations ≤0.28 were eliminated to obtain a final set of

Table 2. Factor Loadings for Exploratory Factor Analysis With Promax Rotation of the Dialysis Access Attitudes Questionnaire
(N = 226)
Factor
Access/Dialysis Worry Need for Value of
Concerns About Cost Dialysis Access
Mean score 3.77 ± 0.92 2.92 ± 0.89 4.30 ± 0.89 3.60 ± 0.86
Cronbach α 0.81 0.76 0.61 0.64
Variance explained 25.7% 12.5% 6.8% 5.1%
Factor loadings
Having to undergo the operation for dialysis 0.82 – – –
access (fistula/PD catheter) worries me
I worry about the effects of having the fistula/PD 0.86 – – –
catheter on my everyday life.
I am worried that the fistula will fail to work 0.68 – – –
I will worry about having to start dialysis earlier 0.63 – – –
than usual once the fistula is in place
I am afraid of dialysis because of the bad things 0.44 – – –
people say about dialysis
I am concerned about the cost of the fistula/PD – 0.58 – –
catheter surgery
I worry about the cost of dialysis – 0.91 – –
I do not see any value in early fistula creation as – – 0.62 –
I may never need dialysis
I do not need dialysis as I feel that I am fine (no – – 0.56 –
or minimum symptoms)
My symptoms are not due to my kidney – – 0.64 –
condition
I see several advantages/benefits of preparing – – – 0.77
the fistula early
Early fistula creation will make transition to – – – 0.64
dialysis more smooth in the future
Note: Values for score given as mean ± standard deviation. All factor loadings > |.40|.
Abbreviation: PD, peritoneal dialysis.

AJKD Vol 77 | Iss 6 | June 2021 935

 Chia et al

Table 3. Unadjusted and Adjusted Logistic Regression Explaining Patients’ Access Intentions
OR (95% CI) for Access Intention (vs Delay Intention)
Unadjusted Model Adjusted Model (n = 136)
Variable (n = 152-190) Block 1 Block 2
Sociodemographic/clinical variable
Age, per 1 y older 1.01 (0.98-1.04) 1.01 (0.97-1.06) 1.04 (0.99-1.10)
Female sex 0.69 (0.39-1.24) 0.41a (0.19-0.89) 0.52 (0.20-1.35)
Chinese ethnicityb 0.78 (0.44-1.40) 1.09 (0.51-2.32) 1.22 (0.48-3.09)
Married or cohabitingc 0.80 (0.43-1.51) 0.90 (0.41-1.97) 0.87 (0.34-2.25)
eGFR, per 1 mL/min/1.73 m2 greater 0.95 (0.89-1.01) 0.86d (0.78-0.95) 0.84d (0.75-0.95)
Charlson comorbidity index 1.00 (0.86-1.16) 1.05 (0.83-1.32) 1.04 (0.78-1.39)
No. of KRT counseling sessions, per 1 session greater 0.93 (0.69-1.25) 1.02 (0.69-1.50) 0.96 (0.61-1.52)
DAAQ, per 1-point greater score
Value of access 2.46e (1.57-3.86) – 2.61d (1.46-4.65)
Need for dialysis 1.74d (1.20-2.51) – –
Access and dialysis concerns 0.73a (0.54-0.98) – –
Worry about cost 0.94 (0.69-1.28) – –
BIPQ, per 1-point greater score
Consequences 1.14d (1.04-1.25) – –
Timeline 1.17d (1.06-1.30) – –
Emotional representation 1.10a (1.01-1.20) – 1.18a (1.02-1.36)
KDQOL-SF, per 1-point greater score
Mental component summary 0.97a (0.94-0.99) – –
Burden of kidney disease 0.99a (0.98-1.00) – –
Patient satisfaction 1.02d (1.01-1.04) – 1.03d (1.01-1.05)
Role-physical 0.99a (0.99-1.00) – –
Role-emotional 0.99a (0.99-1.00) – –
General health 0.98d (0.96-0.99) – 0.96e (0.93-0.98)
Note: R2 = 0.45 for multivariable model. Criterion variable coded as 1 = proceed with access, 0 = wait and see.
Abbreviations: BIPQ, Brief Illness Perception Questionnaire; DAAQ, Dialysis Access Attitudes Questionnaire; eGFR, estimated glomerular filtration rate; KDQOL-SF,
Kidney Disease Quality of Life Short Form; KRT, kidney replacement therapy; OR, odds ratio.
a
P < 0.05.
b
Versus Malay, Indian, or others.
c
Versus not married or in a cohabiting relationship.
d
P < 0.01.
e
P < 0.001.

k = 12 items. Factor 1 (access and dialysis concerns)
comprised 5 items that measured concerns regarding the
procedure, side effects, and outcome of access creation.
Two items that loaded onto factor 2 (worry about cost)
assessed financial concerns surrounding access creation and
dialysis. Three items that loaded onto factor 3 (need for
dialysis) probed respondents’ perceived necessity of dial-
ysis treatment. Factor 4 (value of access) comprised 2
items that assessed perceived benefits of early access cre-
ation. All factors had acceptable internal consistency. Mean
scores ranged from 1 to 5, with higher scores denoting
greater concerns about access and financial costs, stronger
beliefs in the need for dialysis, and higher perceived value
of access creation. A 4-factor confirmatory factor analysis
conducted on the patient dataset alone yielded similar
loading patterns.
Factors Associated with Patients’ Access Intentions
We focused on patients’ access intentions because access
creation procedures cannot be initiated without patient
consent. Therefore, only patients’ data (n = 190) are
included in all subsequent analyses. Ninety-three patients
(48.9%) indicated an intention to proceed (“access
intention”), whereas 97 (51.1%) indicated an intention to
delay.
Table 3 presents results from the unadjusted and
adjusted models with access intention as the outcome.
Univariate analyses indicated significant associations be-
tween access intentions and 1-point greater scores for need
for dialysis (OR, 1.74; 95% CI, 1.20-2.51), value of access
(OR, 2.46; 95% CI, 1.57-3.86), access and dialysis con-
cerns (OR, 0.73; 95% CI, 0.54-0.98), timeline (OR, 1.17;
95% CI, 1.06-1.30), consequences (OR, 1.14; 95% CI,
1.04-1.25), emotional representation (OR, 1.10; 95% CI,
1.01-1.20), and several QoL domains. No sociodemo-
graphic or clinical variables emerged as statistically
significant.
Multivariable analyses revealed that eGFR, value of ac-
cess, emotional representation, general health, and patient
satisfaction were independent predictors of access inten-
tion. The final model was significant compared with the
null model: χ 2(11) = 54.06, P < 0.001. No issues with

936 AJKD Vol 77 | Iss 6 | June 2021

Chia et al

Table 4. Association Between AVF Creation and Sociodemographic, Clinical, and Psychosocial Variables Using Cox Proportional
Hazards Model
HR (95% CI) for AVF Creation (vs No AVF Creation)
Unadjusted Model Adjusted Model (n = 136)
Variable (n = 152-190) Block 1 Block 2
Sociodemographic/clinical variable
Age, per 1 y older 0.99 (0.96-1.02) 1.00 (0.97-1.04) 1.04a (1.00-1.08)
Female sex 0.78 (0.43-1.43) 0.53 (0.27-1.05) 0.71 (0.34-1.47)
Chinese ethnicityb 0.68 (0.38-1.23) 0.77 (0.42-1.43) 1.01 (0.52-1.97)
Married or cohabitingc 0.68 (0.38-1.24) 0.75 (0.39-1.42) 0.74 (0.36-1.52)
eGFR, per 1 mL/min/1.73 m2 greater 0.90d (0.84-0.97) 0.91a (0.83-0.99) 0.87d (0.79-0.96)
Charlson comorbidity index, per 1 point greater 0.90 (0.77-1.06) 0.95 (0.76-1.18) 0.92 (0.73-1.15)
No. of KRT counseling sessions, per 1 session greater 1.13 (0.86-1.49) 1.11 (0.83-1.48) 0.90 (0.64-1.27)
DAAQ, per 1-point greater score
Value of access 1.70d (1.18-2.45) – 1.60a (1.06-2.42)
Need for dialysis 1.20 (0.83-1.74) – –
Access and dialysis concerns 0.92 (0.70-1.22) – –
Worry about cost 1.00 (0.75-1.35) – –
BIPQ, per 1-point greater score
Consequences 1.21e (1.09-1.34) – 1.25d (1.08-1.45)
Timeline 1.14a (1.02-1.28) – –
Identity 1.16d (1.04-1.29) – –
Concern 1.13a (1.00-1.28) – –
KDQOL-SF, per 1-point greater score
Physical component summary 0.95e (0.92-0.98) – 0.96a (0.92-0.99)
Effects of kidney disease 0.98e (0.97-0.99) – 0.98a (0.97-1.00)
Burden of kidney disease 0.98d (0.97-0.99) – –
Note: Secondary analyses only on patients with AVF/hemodialysis outcomes (ie, excluding those with other events, loss to follow-up, or no outcomes) replicated effects for
value of access and consequences. Criterion variable coded as 1 = AVF created, 0 = no AVF created.
Abbreviations: AVF, arteriovenous fistula; BIPQ, Brief Illness Perception Questionnaire; CVC, central venous catheter; DAAQ, Dialysis Access Attitudes Questionnaire;
eGFR, estimated glomerular filtration rate; HR, hazard ratio; KDQOL-SF, Kidney Disease Quality of Life Short Form; KRT, kidney replacement therapy.
a
P < 0.05.
b
Versus Malay, Indian, or others.
c
Versus not married or in a cohabiting relationship.
d
P < 0.01.
e
P < 0.001.

multicollinearity were observed. Higher odds of access
intention were reported in patients with lower eGFR (OR
per 1 mL/min/1.73 m2 greater, 0.84; 95% CI, 0.75-
0.95), higher value of access (OR per 1-point higher score,
2.61; 95% CI, 1.46-4.65), lower general health (OR per 1-
point higher score, 0.96; 95% CI, 0.96–0.99), greater
emotional representation (OR per 1-point higher score,
1.18; 95% CI, 1.02-1.36), and higher patient satisfaction
(OR per 1-point higher score, 1.03; 95% CI, 1.01-1.05).
Sensitivity analyses on family members’ DAAQ data
showed similar associations: need for dialysis and value of
access had significant associations with self-reported own
access intentions in unadjusted models. Family members’
perceived need for dialysis was also associated with
patient-reported access intentions (Table S1).
Factors Associated with Patients’ AVF Creation
Within the study observation window (median duration,
19 [IQR, 17–22] mo; mean ± SD, 18.3 ± 3.1 mo), 45
patients had an AVF access placed (of whom 33 subse-
quently started hemodialysis with a functioning AVF). The
remaining competing events were censored as follows:
initiated hemodialysis with a CVC (n = 66; including 12
who had a fistula created with a low eGFR with subsequent
hemodialysis initiation at a mean of 1.25 ± 1.42 mo
[median, 1 mo]), started PD (n = 35), died before KRT
initiation (n = 10), were lost to follow-up (n = 11), or had
a pending outcome that included patients who opted for
PD (n = 9) or conservative nondialytic management
(n = 2) or were undecided or had stable renal markers and
hence no care directive for access creation (n = 12).
In univariate cause-specific hazards models, lower eGFR
(HR per 1 mL/min/1.73 m2 greater, 0.90; 95% CI, 0.84-
0.97), higher value of access (HR per 1-point greater
score, 1.70; 95% CI, 1.18-2.45), higher scores in 4 BIPQ
domains (HRs ranged from 1.13 to 1.21), and lower QoL
in 3 domains (HRs ranged from 0.95 to 0.98) were
associated with timely and functional AVF creation.
Table 4 shows the unadjusted and adjusted HRs. In the
multivariable model adjusting for sociodemographic and
clinical parameters (final model χ 2[11] = 45.10;
P < 0.001), value of access (HR per 1-point greater score,
1.60; 95% CI, 1.06-2.42) and consequences (HR per 1-
point greater score, 1.25; 95% CI, 1.08-1.45) remained

AJKD Vol 77 | Iss 6 | June 2021 937

 Chia et al

significant predictors of AVF creation, together with eGFR Value of access was observed to associate with access in-
(HR per 1 mL/min/1.73 m2 greater, 0.87; 95% CI, 0.79- tentions and timely AVF creation in multivariable analyses.
0.96). QoL parameters were not reliably associated with Perceptions of benefit have been highlighted in several
AVF outcomes as shown by 95% CIs (Table 4). No issues behavioral change models (ie, the transtheoretical model)
with multicollinearity were observed. as a key driver for developing and sustaining readiness to
progress from ambivalence to activation.38
Discussion Study limitations should be noted. Although observed
associations provide support for the criterion validity of
Significant efforts have been invested to address late
DAAQ, the study design precludes causal inferences.
referral to nephrology care and improve access to pre-
Findings are limited to those who attended nephrology
dialysis education. Nonetheless, the rates of suboptimal
care before KRT initiation and may not be generalizable to
initiation of KRT are stubbornly slow to decline. Intentions
those who defaulted on appointments or declined study
to delay were found to be dominant even among those
participation. It is possible that these groups have weaker
already undergoing nephrology care. Although patients in
inclinations toward dialysis preparation. Selection biases
our study had attended one or more counseling sessions
were likely to be minimal given a fairly good response rate
focused on KRT, the majority preferred to delay access
and given that ethnic and gender proportions of the patient
creation (51.1%) and initiated hemodialysis with a CVC,
sample closely resembled the national renal population.39
replicating trends from other reports.1
The specific health care system in which study findings
Although the role of patient beliefs in health-related de-
are embedded may influence patient decisions, limiting
cision-making has received strong theoretical and empirical
generalizability to other health care contexts.
support in other patient populations,32 these have not been
The development of the DAAQ to quantify attitudes
systematically examined in CKD, and no instrument specif-
toward dialysis access offers clinicians a tool to elicit key
ically assesses beliefs about dialysis access. To bridge this gap,
perceptions toward access creation and guide conversa-
the DAAQ was developed though a systematic process of
tions about dialysis access. The concerns identified could
literature review, content-expert review, qualitative research,
serve as individualized targets for predialysis education and
pilot testing, and psychometric testing. Factor analysis iden-
decision-support interventions, and clinicians may
tified 4 domains: access and dialysis concerns, need for
leverage identified benefits to bolster confidence in
dialysis, worry about cost, and value of access. The domains
decision-making. Still, we note that this is the first vali-
align to constructs in health behavior theories.9,33
dation study of DAAQ. Although findings support its
Importantly, DAAQ domains were shown to have pre-
relevance for patients and clinicians in our study, more
dictive value for access intention and creation outcomes in
research is needed to cross-validate the DAAQ in other
the ensuing 19 months. Univariate analyses indicate that
renal settings. More work is also needed to establish test-
need for dialysis and value of access were significantly
retest reliability, minimally important clinical differences
associated with access intentions. Treatment-related de-
or responsiveness of DAAQ domains to disease progres-
cisions typically involve balancing risks against perceived
sion, and aspects of predialysis education (ie, exposure,
necessity and anticipated gains.34 In the multivariable
content, or providers). Patients’ attitudes about access may
model, value of access remained a significant predictor.
be influenced by variation in effectiveness of providers in
Furthermore, high threat (indexed by low eGFR, worse
delivering the standard content of predialysis education,
general health perceptions, and more negative emotional
but this was not assessed or controlled for in this study.
response) coupled with higher patient satisfaction fosters
Much research on dialysis-related decision-making has
an intention to undergo access preparation. Given that
focused on the retrospective assessment of patients already
dialysis initiation is seen as a threatening prospect,15,35
undergoing dialysis. Less is known about patients not
emphasizing the positive outcomes afforded by timely
receiving KRT who are faced with the impending decision
access creation may help mitigate the emotional threat and
regarding dialysis. The present study built on existing
concerns. Without this emphasis, education efforts may
research through the development and preliminary vali-
become counterproductive, as patients may adopt un-
dation of a measure of attitudes toward dialysis access and
helpful coping strategies that interfere with treatment
assessed psychosocial factors predictive of dialysis access
activation such as denial or defensive avoidance.36
intentions and outcomes in patients approaching kidney
For AVF creation, significant prognostic associations for
failure. Several psychosocial risk and protective factors
perceived benefits of access and consequences suggest that
were identified. Study findings have important implica-
patients may feel more compelled to proceed with AVF
tions for the design of interventions to address concerns
creation when they recognize the negative impact of their
about access creation and for structuring patient-centered
CKD on their lives and the benefits of early access prepa-
KRT counseling sessions.
ration. Balancing negatively and positively framed infor-
mation (ie, consequences of delay vs benefits of timely
preparation) has been shown to influence and instill Supplementary Material
confidence in decision-making in other health contexts.37 Supplementary File (PDF)

938 AJKD Vol 77 | Iss 6 | June 2021

Chia et al
Table S1: Sensitivity analyses explaining family members’ and pa-
tients’ access intentions.
Article Information
Authors’ Full Names and Academic Degrees: Jace Ming Xuan
Chia, BSocSci, Zhong Sheng Goh, BSocSci, Pei Shing Seow,
BSocSci, Terina Ying-Ying Seow, MRCP, Jason Chon Jun Choo,
MRCP, Marjorie Wai-Yin Foo, FRCP, Stanton Newman, PhD, and
Konstadina Griva, PhD.
Authors’ Affiliations: Centre for Population Health Sciences, Lee
Kong Chian School of Medicine, Nanyang Technological University
Singapore (JMXC, ZSG, KG); Department of General Medicine,
Khoo Teck Puat Hospital (PSS); Department of General Medicine,
Sengkang General Hospital (TY-YS); Department of Renal
Medicine, Singapore General Hospital (JCJC, MW-YF), Singapore;
and School of Health Sciences, Division of Health Services
Research and Management, City University of London, London,
United Kingdom (SN).
Address for Correspondence: Konstadina Griva, PhD, Centre for
Population Health Sciences, Lee Kong Chian School of Medicine,
Nanyang Technological University Singapore, Clinical Sciences
Building, Singapore 308232, Singapore. Email: konstadina.griva@
ntu.edu.sg
Authors’ Contributions: Research conceptualization and study
design: KG, TY-YS; data acquisition: PSS, ZSG, JMXC; data
analyses/interpretation: KG, JMXC, ZSG; supervision or
mentorship: KG, TY-YS, JCJC, MW-YF, SN. Each author
contributed important intellectual content during manuscript
drafting or revision and agrees to be personally accountable for
the individual’s own contributions and to ensure that questions
pertaining to the accuracy or integrity of any portion of the work,
even one in which the author was not directly involved, are
appropriately investigated and resolved, including with
documentation in the literature if appropriate.
Support: This research is supported by the Singapore Ministry of
Health's National Medical Research Council under its Health
Services Research Grant (NMRC/HSRG/0058/2016). The
funding source had no role in the study design or intervention,
recruitment of patients, data collection, analysis, or interpretation
of the results, writing of the manuscript, or decision to submit the
manuscript for publication.
Financial Disclosure: The authors declare that they have no
relevant financial interests.
Acknowledgements: The authors thank health care professionals in
the participating sites, Tonia Griva, and all study participants for their
support in this study.
Data Sharing: The data generated and analyzed are not available
publicly but are available from the corresponding author upon
request.
Peer Review: Received November 25, 2019. Evaluated by 2
external peer reviewers, with direct editorial input from a Statistics/
Methods Editor, an Associate Editor, and the Editor-in-Chief.
Accepted in revised form September 17, 2020.
References
1. Mendelssohn DC, Curtis B, Yeates K, et al. Suboptimal initia-
tion of dialysis with and without early referral to a nephrologist.
Nephrol Dial Transplant. 2011;26(9):2959-2965.
2. Moist LM, Lok CE. Incident dialysis access in patients with end-
stage kidney disease: what needs to be improved. Semin
Nephrol. 2017;37(2):151-158.
AJKD Vol 77 | Iss 6 | June 2021
3. Vascular Access Work Group. Clinical Practice Guidelines for
Vascular Access. Am J Kidney Dis. 2006;48(suppl 1):S248-
S273.
4. Ortega T, Ortega F, Diaz-Corte C, Rebollo P, Baltar JM, Alvarez-
Grande J. The timely construction of arteriovenous fistulae: a
key to reducing morbidity and mortality and to improving cost
management. Nephrol Dial Transplant. 2005;20(3):598-603.
5. Fluck R, Pitcher D, Steenkamp R. Vascular Access Audit
Report 2012. UK Renal Registry and NHS Kidney Care; 2012.
6. Saran R, Robinson B, Abbott KC, et al. US Renal Data System
2018 Annual Data Report: epidemiology of kidney disease in
the United States. Am J Kidney Dis. 2019;73(3)(suppl 1):A7-
A8.
7. Marr on B, Ortiz A, de Sequera P, et al. Impact of end-stage
renal disease care in planned dialysis start and type of renal
replacement therapy–a Spanish multicentre experience.
Nephrol Dial Transplant. 2006;21(suppl 2):ii51-ii55.
8. Hassan R, Akbari A, Brown PA, Hiremath S, Brimble KS,
Molnar AO. Risk factors for unplanned dialysis initiation: a
systematic review of the literature. Can J Kidney Health Dis.
2019;6. 2054358119831684.
9. Rosenstock IM. Why people use health services. Milbank Q.
1966;44(3):94-127.
10. Leventhal H, Meyer D, Nerenz. The common sense represen-
tation of illness danger. In: Rachman SJ, ed. Contributions to
Medical Psychologyvol. 2. Pergamon; 1980:7-30.
11. Campbell-Crofts S, Stewart G. How perceived feelings of
“wellness” influence the decision-making of people with pre-
dialysis chronic kidney disease. J Clin Nurs. 2018;27(7-8):1561.
12. Murray MA, Brunier G, Chung JO, et al. A systematic review of
factors influencing decision-making in adults living with chronic
kidney disease. Patient Educ Couns. 2008;76(2):149-158.
13. Lovell S, Walker RJ, Schollum JBW, Marshall MR, McNoe BM,
Derrett S. To dialyse or delay: a qualitative study of older New
Zealanders’ perceptions and experiences of decision-making, with
stage 5 chronic kidney disease. BMJ Open. 2017;7(3):e014781.
14. Tonkin-Crine S, Okamoto I, Leydon GM, et al. Understanding
by older patients of dialysis and conservative management for
chronic kidney failure. Am J Kidney Dis. 2015;65(3):443-450.
15. Lin CC, Lee BO, Hicks FD. The phenomenology of deciding
about hemodialysis among Taiwanese. West J Nurs Res.
2005;27(7):915-929.
16. Tola HH, Davoud S, Azar T, et al. Psychological and educational
intervention to improve tuberculosis treatment adherence in
Ethiopia based on Health Belief Model: a cluster randomized
control trial. PLoS ONE. 2016;11(5):e0155147.
17. HealthMeasures. PROMIS® Instrument Development and
Validation Scientific Standards. Version 2.0. Northwestern
University; 2013.
18. Kosa SD, Bhola C, Lok CE. Measuring patient satisfaction with
vascular access: vascular access questionnaire development
and reliability testing. J Vasc Access. 2015;16(3):200-205.
19. Sousa CN, Apostolo JLA, Figueiredo MHJS, Dias VFF, Teles P,
Martins MM. Construction and validation of a scale of
assessment of self-care behaviors with arteriovenous fistula in
hemodialysis. Hemodial Int. 2015;19(2):306-313.
20. Quinn RR, Lamping DL, Lok CE, et al. The Vascular Access
Questionnaire: assessing patient-reported views of vascular
access. J Vasc Access. 2008;9(2):122-128.
21. Nordyke RJ, Nicholson G, Gage SM, et al. Vascular access-
specific health-related quality of life impacts among hemodial-
ysis patients: qualitative development of the hemodialysis
access-related quality of life (HARQ) instrument. BMC Neph-
rol. 2020;21(1):16.
939

22. Griva K, Seow PS, Seow TYY, Goh ZS, Choo JCJ, Foo M, et al.
Patient-Related Barriers to Timely Dialysis Access Preparation:
A Qualitative Study of the Perspectives of Patients, Family
Members, and Health Care Providers. Kidney Med. 2020;2(1):
29-41.
23. Levey AS, Coresh J, Greene T, et al. Expressing the modifica-
tion of diet in renal disease study equation for estimating
glomerular filtration rate with standardized serum creatinine
values. Clin Chem. 2007;53(4):766-772.
24. Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a
combined comorbidity index. J Clin Epidemiol. 1994;47(11):
1245-1251.
25. Broadbent E, Petrie KJ, Main J, Weinman J. The Brief Illness
Perception Questionnaire. J Psychosom Res. 2006;60(6):631-
637.
26. Broadbent E, Wilkes C, Koschwanez H, Weinman J, Norton S,
Petrie KJ. A systematic review and meta-analysis of the Brief
Illness Perception Questionnaire. Psychol Health.
2015;30(11):1361-1385.
27. Hays RD, Sherbourne CD, Mazel RM. The Rand 36-item health
survey 1.0. Health Econ. 1993;2(3):217-227.
28. Ware J, Snow K, Ma K, Bg G. SF36 Health Survey: Manual
and Interpretation Guide. New England Medical Center, The
Health Institute; 1993.
29. Peng CYJ, Lee KL, Ingersoll GM. An introduction to logistic
regression analysis and reporting. J Educ Res. 2002;96(1):3-
14.
940
Chia et al
30. Rosenthal JA. Qualitative descriptors of strength of association
and effect size. J Social Serv Res. 1996;21(4):37-59.
31. Chatterjee S, Yilmaz M. A review of regression diagnostics for
behavioral research. Appl Psychol Meas. 1992;16(3):209-227.
32. Hand C, Morley S, Adams M. Developing a questionnaire to
measure patients’ beliefs about inhaler treatment for asthma:
tests of validity and reliability. Prim Care Respir J. 2000;9:12-15.
33. Ajzen I. From intentions to actions: a theory of planned
behavior. In: Kuhl J, Beckmann J, eds. Action-Control: From
Cognition to Behavior. Springer; 1985:11-39.
34. West LM, Borg Theuma R, Cordina M. The ‘Neces-
sity–Concerns Framework’ as a means of understanding non-
adherence by applying polynomial regression in three chronic
conditions. Chronic Illness. 2020;16(4):253-265.
35. Krespi R, Bone M, Ahmad R, Worthington B, Salmon P. Hae-
modialysis patients’ beliefs about renal failure and its treatment.
Patient Educ Couns. 2004;53(2):189-196.
36. Brown S, Locker E. Defensive responses to an emotive anti-
alcohol message. Psychol Health. 2009;24(5):517-528.
37. Shamaskin AM, Mikels JA, Reed AE. Getting the message
across: age differences in the positive and negative framing of
health care messages. Psychol Aging. 2010;25(3):746-751.
38. Prochaska JO, Velicer WF. The transtheoretical model of health
behavior change. Am J Health Promot. 1997;12(1):38-48.
39. Health Promotion Board. Singapore Renal Registry Annual
Report 2018. National Registry of Diseases Office; January 20,
2020.
AJKD Vol 77 | Iss 6 | June 2021
Original Investigation

Mobile Health (mHealth) Technology: Assessment of

Availability, Acceptability, and Use in CKD
Sarah J. Schrauben, Lawrence Appel, Eleanor Rivera, Claudia M. Lora, James P. Lash, Jing Chen, L. Lee Hamm,
Jeffrey C. Fink, Alan S. Go, Raymond R. Townsend, Rajat Deo, Laura M. Dember, Harold I. Feldman, and
Clarissa J. Diamantidis, on behalf of the CRIC Study Investigators

Visual Abstract online

Rationale & Objective: Digital and mobile health
(mHealth) technologies improve patient-provider
communication and increase information
accessibility. We assessed the use of
technology, attitudes toward using mHealth
technologies, and proficiency in using mHealth
technologies among individuals with chronic
kidney disease (CKD).
Study Design: Cross-sectional survey with open
text responses.
Setting & Participants: Chronic Renal Insuffi-
ciency Cohort (CRIC) Study participants who
completed current use and interest in using
mHealth technologies questionnaires and the
eHealth literacy Survey (eHEALS).
Exposure: Participant characteristics.
Outcomes: Use of technology (ie, internet, email,
smartphone, and mHealth applications [apps]),
interest in future mHealth use, and proficiency in
using digital and mHealth technologies, or
eHealth literacy, determined by eHEALS score.
Analytical Approach: Poisson regression and a
qualitative content analysis of open-ended
responses.
Results: Study participants (n = 932) had a mean
age of 68 years old and an estimated glomerular
filtration rate (eGFR) of 54 mL/min/1.73 m2, and
59% were male. Approximately 70% reported
current use of internet, email, and smartphones,
and 35% used mHealth apps; only 27% had Complete author and article
adequate eHealth literacy (eHEALS score ≥ 32). information provided before
references.
Participants <65 years of age (vs. ≥65), with more
education, higher income, better cognition, and Correspondence to
adequate health literacy reported more use of S.J. Schrauben (Sarah.
technology, and greater interest in using tech- Schrauben@pennmedicine.
upenn.edu)
nologies. Participants of White (vs. non-White)
race reported more use of internet and email Am J Kidney Dis. 77(6):
941-950. Published online
but less interest in future use of mHealth.
December 9, 2020
Younger age, higher annual income, and greater
disease self-efficacy were associated with doi: 10.1053/
adequate eHealth literacy. Three themes j.ajkd.2020.10.013
regarding interest in using digital and mHealth © 2020 by the National
technologies emerged: willingness, concerns, Kidney Foundation, Inc.
and barriers.
Limitations: Residual confounding, ascertain-
ment bias.
Conclusions: Many individuals with CKD
currently use the internet and smartphones and
are interested in using mHealth in the future, but
few use mHealth apps or have adequate eHealth
literacy. mHealth technologies present an op-
portunity to engage individuals with CKD, espe-
cially members of racial or ethnic minority groups
because those groups reported greater interest
in using mHealth technology than the nonminority
population. Further research is needed to identify
strategies to overcome inadequate eHealth
literacy.

M obile health (mHealth) technology pertains to the use

of mobile and wireless digital technologies to support
health and wellness by extending education, communica-
tion, health interventions, and research beyond the reach
of conventional clinical care.1 In chronic disease pop-
ulations, mHealth has been shown to improve communi-
cation between patients and providers, enhance patient
autonomy, and empower patients to become active par-
ticipants in their own care,2,3 each of which increases the
patient-centeredness of care.
Individuals with chronic conditions such as chronic
kidney disease (CKD) often need to participate in
complicated disease management strategies but have low
levels of participation in self-management behaviors, a
characteristic that links to poor outcomes.4 mHealth
technology has the potential to improve the management
of CKD by facilitating patient education, supporting
behavior engagement (such as adherence to medication
regimens and dietary modification), and aiding in patient-
provider communication.5 mHealth is also an attractive
option to supplement the taxed resources of health care
providers, resulting, in part, from the increasing preva-
lence of CKD. In addition, the coronavirus 2019 (COVID-
19) pandemic has necessitated that nephrology care be
delivered remotely to improve social distancing and, thus,
has created the need to harness the potential of using
mHealth technologies.
Despite the potential of mHealth to support CKD
management, few mHealth strategies have been used to
assist individuals with this condition.5,6 To date, little is
known about the availability and use of mHealth tech-
nologies, attitudes about their use, or level of proficiency
in using digital and mHealth technologies for health pur-
poses (ie, eHealth literacy) among people with CKD.7 This
mixed methods study describes the use of mHealth tech-
nologies, as well as willingness to use mHealth

AJKD Vol 77 | Iss 6 | June 2021 941


PLAIN-LANGUAGE SUMMARY
Mobile health (mHealth) technology improves patient-
provider communication and increases access to infor-
mation. We wanted to explore how extensively patients
are using mHealth technology, what they think about
using it, and how proficient they are using it. Our study
focused on a group of 932 individuals with chronic
kidney disease (CKD). We found that 70% currently
used internet, email, and smartphones and that 35%
used health-specific applications. Less than one-third of
our patients reported proficient use of mHealth tech-
nology. Individuals who were under 65 years of age,
had more education, and higher income were more
likely to use mHealth technology and had a greater
interest in it. Black and Hispanic individuals used
technology less overall but were more interested in
using mHealth technology. mHealth technology pre-
sents an opportunity to engage individuals with CKD.
technologies and the level of eHealth literacy among in-
dividuals with mild-to-moderate CKD in the Chronic Renal
Insufficiency Cohort (CRIC) study.
Methods
Study Design and Study Population
This study used data from the CRIC registry, a multicenter,
prospective observational cohort of persons with mild to
moderate CKD in the United States.8,9 CRIC participants
were recruited from 7 clinical centers (University of
Pennsylvania, Johns Hopkins University, Case Western
Reserve University, University of Michigan, University of
Illinois at Chicago, Tulane University, and Kaiser Perma-
nente of Northern California) during 2 phases. Phase I ran
from 2003 to 2008 (3,939 participants), and Phase III ran
from 2013 to 2015 (1,560 participants), and participants
were prospectively followed annually. The full inclusion
and exclusion criteria have been previously reported.8,9
Briefly, Phase I included adults 21-74 years of age with
age-based estimated glomerular filtration rate (eGFR) 20-
70 mL/min/1.73 m2; and entry into Phase III required
eGFR 45-70 mL/min/1.73m2 and 45-79 years of age.
Exclusion criteria included the inability to consent and
certain severe conditions. For this study, data were used
from participants who completed at least 1 of 2 surveys
(mHealth/Technology Use Survey or e-Literacy Survey)
that contained questions related to the use of internet,
email, smartphones, or mHealth applications (apps),
future interest in using mHealth technologies, and eHealth
literacy. Potential survey participants consisted of a con-
venience sample of CRIC participants who attended annual
study visits between late 2016 and mid-2018. Research
coordinators offered the mHealth/Technology Use Survey
942
Schrauben et al
to the first 60 participants at their annual visit at all 7
clinical centers, and the e-Literacy Survey was offered at 4
of the 7 clinical centers (Fig 1). The surveys were self-
administered at in-person visits or administered by the
research coordinator if the visit was conducted over the
phone. The CRIC study protocol was approved by the
Institutional Review Boards of all participating centers and
is in accordance with the Declaration of Helsinki. All
participants provided written informed consent.
Measurements
The mHealth/Technology Use Survey and e-Literacy Sur-
vey were developed within the CRIC study in parallel for
distinct but complementary reasons (Item S1 provides
more information). Briefly, the mHealth/Technology Use
Survey consisted of 12 questions that were developed to
collect information from participants about their willing-
ness to use mHealth technologies, along with their current
technology availability and use in the effort to plan for
future research in the CRIC study. The survey questions
were drafted and vetted by CRIC investigators and then
piloted among CRIC research staff. The e-Literacy Survey
was developed for use in a CRIC ancillary study to learn
about participants’ current use of and proficiency in using
digital and mHealth technology and consisted of 19
questions, 11 of which were developed de novo by CRIC
ancillary investigators, and 8 corresponded to the validated
eHealth Literacy Scale (eHEALS).10,11
Study Outcomes
The primary study outcome was use of technology and
secondary outcomes were interest in future use of mHealth
technologies and level of eHealth literacy (Table S1). Use
of technology was categorized as current or not current use
of 3 types of technology: internet or email, smartphones,
and mHealth apps. Use of internet or email was collapsed
into 1 category because email requires the internet. Use of
the internet was assessed by responses to “do you use the
internet?” or by confirming the use of the internet of at
least once per month. Use of email was assessed by an-
swers to “do you use and read email?” or by confirming
the use of email when using the internet or a smartphone.
Use of the smartphone was assessed by answers to “do you
use a smartphone (a cell phone that can access the internet
such as an iPhone)?” or by selecting “smartphone” when
asked “what type of cell/mobile phone do you primarily
use?” Use of the mHealth app was determined by
endorsing a response to “which of the following types of
health apps have you used on your cell/mobile phone?”
Interest in future use of mHealth technologies was assessed
by answers to “are you willing to answer study ques-
tionnaires using the internet, email, or smartphone?” as
these study questions focus on the participants’ willingness
to use mHealth to report on their health status or by an
affirmative response to at least 1 statement related to the
development of a new internet or mHealth tool for the
AJKD Vol 77 | Iss 6 | June 2021
Schrauben et al
Figure 1. Flow diagram of study sample selection.
following purposes: “I’d like more help remembering to
take my medications,” “I’d like more help learning about
what my medications are for,” or “I’d like more help
learning about possible side effects from my medications.”
eHealth literacy is the skill set needed to effectively find,
appraise, and use health information from electronic or
digital materials, including mHealth technology. eHealth
literacy was assessed with eHEALS to measure the level of
an individual’s knowledge and perceived skills at finding,
evaluating, and applying digital health information to their
health problems.10,11 eHEALS consisted of 8 questions
with 5 response options on a Likert scale and was scored as
a cumulative score of the 8 questions (range 8-40, with a
score of ≥32 considered to be adequate) or individually
(range 1-5).10,11 In our analysis, a dichotomous cumula-
tive score of <32 or ≥32 was used for adequate eHealth
literacy.
Covariates
Covariates were chosen for their potential influence on
technology use, interest in digital and mHealth technologies,
and eHealth literacy.12-14 Age was determined by the year of
survey completion. Sex, race/ethnicity, education, income,
and history of comorbidity were collected through partici-
pants’ self-reports during the CRIC baseline assessment. Re-
ported comorbidities were used to construct a Charlson
comorbidity index.15 Data for health literacy, depressive
symptoms, and cognitive function were also collected at
baseline using validated questionnaires, the short version of
the Test of Functional Health Literacy in Adults (S-TOFHLA),
Beck Depression Inventory (BDI), and Modified Mini-Mental
AJKD Vol 77 | Iss 6 | June 2021
State Exam, respectively.16-18 Measurements of disease
management self-efficacy (Self-Efficacy for Management of
Chronic Disease) and social network (Lubben Social
Network Score) were available only at baseline from Phase III
cohort participants.19,20 Serum samples collected at the study
visit closest to survey completion were used to determine an
eGFR with a CRIC-specific equation using serum creatinine
and cystatin C concentrations.21 Location of recruitment site
was included in the models to account for potential
geographic differences.
Statistical Analysis
We describe the study population by using mean and
standard deviation or median and interquartile range (IQR)
for continuous variables and frequency and proportion for
categorical variables. Pearson χ 2 and Kruskal-Wallis tests
were used to compare categorical and continuous variables,
respectively. We report the prevalence of use of current
technology and interest in future use of digital and mHealth
technologies in the total study sample and by age group
(<65 and ≥65 years old) as age has been shown to have an
important influence on digital technology.22-26
We examined the association between sociodemographic
and clinical factors with current technology use and interest
in mHealth and eHealth literacy in multivariate-adjusted
Poisson (log-link) regression models with robust error
variance to model rate ratios to approximate prevalence
ratios (PR).27,28 Poisson regression was preferred over lo-
gistic regression since the odds ratio is known to overstate
the risk ratio versus the PR in cross-sectional studies with
common outcomes.29 Analyses were performed with
943
 Schrauben et al

Table 1. Demographic and Clinical Characteristics of the Study Sample and by Survey Type
e-Literacy survey mHealth/technology
Total sample subsample use survey subsample
(N = 932) (n = 633) (n = 424) Pa
Age, y 67.9 ± 9.3 67.6 ± 9.2 68.5 ± 9.2 0.06
Male sex 58.5% 60.5% 56.1% 0.08
Distribution of race/ethnicity <0.001
Non-Hispanic White 37.9% 31.9% 42.9%
Non-Hispanic Black 51.4% 54.8% 49.8%
Hispanic 7.6% 10.7% 3.5%
Other 3.1% 2.5% 3.8%
Annual household income 0.2
<$20,000 25.4% 26.2% 25.5%
$20,000-$50,000 25.1% 26.9% 23.6%
>$50,000 34.2% 31.4% 35.9%
Prefer not to answer 15.2% 15.5% 15.1%
Education <0.001
<High school 15.2% 18.5% 12.7%
High school/some college 47.0% 47.6% 47.2%
College or higher 37.8% 33.9% 40.1%
Clinical recruitment sitea <0.001
East Coast 50.6% 67.3% 28.5%
Midwest 35.9% 32.4% 42.2%
South 6.4% 0% 14.2%
West Coast 7.0% 0.3% 15.1%
Adequate health literacyb 89.0% 87.4% 90.8% 0.04
Cognition scorec 91.7 ± 8.0 91.1 ± 8.5 92.2 ± 6.9 0.02
Depressive symptom scored 7.6 ± 8.0 7.8 ± 8.3 7.3 ± 7.3 0.7
eGFR, mL/min/1.73 m2 54.4 ± 15.3 53.6 ± 15.7 54.4 ± 14.6 0.06
Prevalent CVD 25.5% 25.6% 26.4% 0.9
Diabetes 46.6% 49.3% 44.3% 0.04
Hypertension 85.2% 88.2% 81.6% <0.001
Charlson comorbidity score 5.1 ± 2.0 5.2 ± 2.0 5.1 ± 2.0 0.05
Disease self-efficacye 41.3 ± 8.7 41.3 ± 9.1 41.2 ± 7.9 0.3
Social support scoref,h 16.0 ± 5.9 15.7 ± 6.0 16.4 ± 5.9 0.3
Adequate eHealth literacyg – 27.2% – –
Note: Values for continuous variables are given as mean ± SD.
Abbreviations: CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate; eHEALS, eHealth Literacy Scale.
a
East Coast (Philadelphia, PA, Baltimore, MD), Midwest (Ann Arbor, MI; Cleveland, OH; Chicago, IL), South (New Orleans, LA); West Coast (San Francisco, CA; Oakland,
CA).
b
Assessed using the short version of the Test of Functional Health Literacy in Adults (S-TOFHLA) instrument (adequate score >16 of 36).16
c
Assessed with modified mini-Mental State Examination (score range 0-100).18
d
Assessed with Beck depression inventory (score range 0-63).17
e
Assessed with disease self-efficacy20 (score range 5-50); assessed only in Phase III cohort participants (n = 462).
f
Assessed with Lubben social network scale (score range 0-30)19; only assessed in Phase III cohort participants ≥65 years of age (n = 235).
g
Assessed with eHEALS survey (adequate score ≥32 of 40).10,11
χ tests were used to test differences across survey type subsamples for categorical variables and the Kruskal-Wallis test was used for continuous variables.
h 2

STATA 14.2. Adjusted PRs with 95% CI are reported. Sig-
nificance level was 0.05 for 2-sided tests.
We also conducted a qualitative content analysis of re-
sponses to the open-ended question included in the
mHealth/Technology Use Survey: “Do you have any
comments about what we’ve asked you in this question-
naire?” Given the lack of literature for this topic, we did
not begin with preconceived codes but developed them
inductively through examination of the data.30 Coding was
conducted independently by 2 members of the authors’
team with qualitative research training (S.J.S., E.R.) using
Excel software for MacIntosh version 16.32 (Microsoft).
Each member examined all open question responses and
created her own list of codes to categorize the responses
with no predetermined structure. They then shared those
codes and resolved any coding differences through itera-
tive discussion, finally arriving at a common final coding
scheme.
Results
Participant Characteristics
Among the total sample of 932 CRIC participants who re-
ported their use of technology, 633 completed the e-Literacy

944 AJKD Vol 77 | Iss 6 | June 2021

Schrauben et al

Table 2. Current Technology Use and Future Interest in Use of mHealth Technologies
Age
All <65 years ≥65 years Pa
Current technology useb
Internet or email 690 (75%) 265 (87%) 425 (68%) <0.001
Smartphone 648 (70%) 267 (88%) 831 (61%) <0.001
mHealth appsc 226 (36%) 107 (50%) 119 (29%) <0.001
If mHealth app was used, purpose was reported
as:
To access personal medical record 116 (51%) 60 (56%) 56 (47%)
To look up nutrition or diet information 110 (49%) 58 (54%) 52 (44%)
To learn about health conditions 109 (48%) 61 (57%) 48 (40%)
To track exercise activity or weight 103 (46%) 51 (48%) 52 (44%)
To keep track of medications 59 (26%) 30 (28%) 29 (24%)
Track eating 40 (18%) 23 (22%) 17 (14%)
Record blood pressure 24 (11%) 14 (13%) 10 (8%)
Record serum glucose 22 (10%) 11 (10%) 11 (9%)
Interest in future use of digital and mHealth
technologies
Internet/email/smartphone for kidney research 328 (77%) 114 (89%) 215 (75%) 0.001
purposesd
mHealth apps for medication knowledge and 501 (80%) 177 (82%) 324 (78%) 0.2
managementc
Table values are given as number (percentage).
a
P values were derived by χ2 testing of differences between age groups (<65 vs ≥65 years).
b
Sample sizes: All Ages (n = 922), <65 years (n = 302), ≥65 years (n = 620).
c
Available only in the e-Literacy Survey subsample: all ages, n = 630 (n = 215 aged < 65 years, n = 415 aged ≥ 65 years).
d
Available only in the mHealth/Technology Use Survey subsample: all Ages, n = 424 (n = 128 aged < 65 years, n = 296 aged ≥ 65 years).

Survey, and 424 responded to the mHealth/Technology Use
Survey. Table S2 provides details of participants who did and
did not respond to the surveys. There were 125 participants
(13.4%) who completed both sets of questions. The partic-
ipants who completed either survey were similar, except that
the e-Literacy Survey group had more participants from a
racial or ethnic minority group with hypertension, diabetes,
lower education, and more were categorized as having
inadequate health literacy and less representation from the
South and West Coast than the mHealth/Technology Use
Survey group (Table 1). Younger participants (<65 years of
age) reported more current use of internet and email,
smartphones, and mHealth apps, as well as more interest in
the future use of digital and mHealth technologies than older
participants. The reasons for using an mHealth app were
similar across age categories (Table 2).
Associations with Technology Use and Interest in
Using Digital and mHealth Technologies
Younger persons (ie, per 10 years), those with a high
school education or higher (vs less) and higher annual
household income (≥$20,000 vs <$20,000) were associ-
ated with more current use of technology (Table 3).
Adequate health literacy was also associated with more
internet, email, and smartphone use, and a similar trend
was observed for the use of mHealth apps. Non-Hispanic
Black participants reported less internet and email use
than Non-Hispanic White participants, and non-White
(Non-Hispanic Black and Hispanic) participants reported
more interest in future use of digital and mHealth tech-
nologies. Younger age was the only other characteristic
associated with more interest in future use of digital and
mHealth technologies.
Associations with eHealth Literacy
Younger age, an annual income of ≥$20,000 (vs <20,000),
and a higher disease self-efficacy score were associated with
adequate eHealth literacy (Table 3). There was no apparent
association between education or race/ethnicity and eHealth
literacy. Individual eHEALS response scores were associated
with participant characteristics in a manner similar to that of
the cumulative eHEALS scores (Table S3).
Patient Perspectives Regarding mHealth
Technologies
Of the 424 participants who completed the mHealth/
Technology Use Survey, 66 participants provided open text
responses. Three themes emerged from the coding of the
participant perspectives regarding mHealth technologies:
concerns, barriers, and willingness (Box 1). Reported con-
cerns generally related to apprehensions about the level of
intrusiveness of monitoring and losing face-to-face in-
teractions with clinic staff as illustrated by an 80-year-old
female participant who said, “I love coming to the clinic
and the interaction between the personal contact with all of
the staff,” as well as needing more details about the

AJKD Vol 77 | Iss 6 | June 2021 945

 Schrauben et al

Table 3. Associations of Participant Characteristics With Current Use and Interest in Future Use of mHealth Technologies and
Adequate eHealth Literacy
Interest in Adequate
Internet/E-mail Smartphone mHealth app mHealth eHealth
Characteristic usea usea useb technology usea literacyb
Age (per 10 y older) 0.87 (0.84-0.91) 0.84 (0.80-0.88) 0.71 (0.63-0.80) 0.92 (0.88-0.95) 0.74 (0.63-0.85)
Race/ethnicity groups
Non-Hispanic White 1.00 (reference) 1.00 (reference) 1.00 (reference) 1.00 (reference) 1.00 (reference)
Non-Hispanic Black 0.88 (0.81-0.95) 1.08 (0.98-1.19) 0.89 (0.71-1.10) 1.17 (1.08-1.27) 1.01 (0.75-1.36)
Hispanic 0.87 (0.71-1.07) 1.16 (0.97-1.38) 1.02 (0.64-1.63) 1.20 (1.03-1.41) 1.27 (0.76-2.12)
Other 0.90 (0.72-1.11) 1.00 (0.77-1.31) 1.09 (0.59-2.02) 1.01 (0.81-1.26) 1.04 (0.44-2.46)
Education level
<High school 1.00 (reference) 1.00 (reference) 1.00 (reference) 1.00 (reference) 1.00 (reference)
≥High school 1.56 (1.20-2.04) 1.27 (1.04-1.56) 2.62 (1.32-5.17) 1.05 (0.93-1.19) 1.90 (1.01-3.61)
Annual Income
<$20,000 1.00 (reference) 1.00 (reference) 1.00 (reference) 1.00 (reference) 1.00 (reference)
≥$20,000 1.32 (1.16-1.49) 1.13 (1.00-1.28) 1.46 (1.07-2.00) 1.06 (0.97-1.16) 1.90 (1.28-2.83)
No answer 1.22 (1.05-1.41) 1.20 (1.03-1.38) 1.31 (0.90-1.90) 1.05 (0.93-1.18) 1.78 (1.12-2.84)
Health literacy
Inadequate 1.00 (reference) 1.00 (reference) 1.00 (reference) 1.00 (reference) 1.00 (reference)
Adequate 1.61 (1.21-2.13) 1.25 (1.00-1.57) 1.92 (0.98-3.77) 1.05 (0.93-1.20) 2.07 (0.95-4.52)
Cognition score 1.09 (1.01-1.17) 1.05 (0.98-1.13) 1.04 (0.87-1.25) 1.00 (0.95-1.05) 1.10 (0.87-1.38)
(per 1-SD greater)
eGFR (per 1-mL/min/ 1.00 (0.99-1.00) 1.00 (0.99-1.01) 1.00 (0.99-1.01) 1.00 (0.99-1.00) 1.00 (1.00-1.01)
1.73 m2 greater)
Comorbidity score
<4 1.00 (reference) 1.00 (reference) 1.00 (reference) 1.00 (reference) 1.00 (reference)
4-6 1.05 (0.97-1.13) 1.00 (0.91-1.10) 1.15 (0.90-1.48) 1.08 (0.99-1.18) 1.32 (0.95-1.82)
>6 1.01 (0.90-1.14) 0.96 (0.84-1.10) 0.98 (0.69-1.39) 1.09 (0.98-1.22) 1.01 (0.64-1.61)
Depression symptom 0.99 (0.96-1.03) 1.01 (0.98-1.05) 1.09 (1.00-1.19) 1.03 (0.93-1.20) 1.01 (0.89-1.14)
score (per 1-SD greater)
Disease self-efficacy 1.05 (0.9-1.13) 1.03 (0.96-1.11) 1.11 (0.93-1.32) 1.01 (0.96-1.06) 1.26 (1.01-1.58)
(per 1-SD greater)c
Social supportd 0.99 (0.92-1.07) 1.10 (0.99-1.22) 1.11 (0.90-1.37) 0.99 (0.92-1.07) 1.04 (0.77-1.39)
(per 1-SD greater)
Note: Prevalence ratios (PR), and 95% CI are reported. Models were adjusted for age, sex, race, income, education, clinical recruitment site, health literacy, cognition,
depression, baseline eGFR, and baseline comorbidity score.
a
Total study sample without missing variables (n = 827).
b
e-Literacy Survey subsample without missing variables (n = 567).
c
Model adjusted for variables listed above plus self-efficacy: assessed only in Phase III participants (n = 403 for all models except eHealth literacy; n = 291 for eHealth
literacy).
d
Model were adjusted for variables listed above plus social support: assessed only in Phase III participants ≥65 years of age (n = 207 for all models except eHealth literacy;
n = 146 for eHealth literacy).

technologies. Concerns were predominantly reported by Discussion

those older than 65 years. Barriers to use of technology were
mainly related to physical limitations, as highlighted by a
76-year-old female participant who shared, “I had a stroke 3
years ago and now have problems trying to read on the
internet.” Some participants reported they would require
assistance from others or they lacked technology access. Of
the participants who reported barriers to use of technology,
fewer than half reported current use of smartphone tech-
nology, and only half used internet and email. In regard to
the willingness to use digital and mHealth technologies,
more participants were enthusiastic compared to those who
opposed it. Those who opposed it reported slightly lower
use of internet and email than those who were enthusiastic
and were more likely to be older than 65 years.
Among a cohort of individuals with CKD without kidney
replacement therapy, we identified patient characteristics
that likely impacted the current uptake of and future in-
terest in using digital and mHealth technologies, including
age, health literacy, annual income, and education level.
Interestingly, members of racial or ethnic minority groups
reported less use of internet and email but more interest in
future use of mHealth technologies than White in-
dividuals. Another important finding was that approxi-
mately three-fourths of our sample did not report an
adequate level of eHealth literacy. Prior to this study,
mHealth and use of digital technology among individuals
with kidney diseases in the United States had not been well
described, nor had the level of interest in using mHealth

946 AJKD Vol 77 | Iss 6 | June 2021

Schrauben et al
Box 1. Major and minor themes of participant perspectives of
future use of digital and mHealth technologies with exemplar
quotations.
Concerns about use
Intrusiveness
“I work full time, so answers are contingent on devices being
unobtrusive in a work environment” (Female, Age 55, Record
310)
“Do not want any kind of constant monitoring” (Female, Age
61, Record 9)
Losing face-to-face interaction
“I love coming to the clinic and the interaction between the
personal contact with all of the staff” (Female, Age 80, Record
44)
“I like coming here. Less I have to do at home; the better I like
it” (Male, Age 74, Record 40)
Needs more details
“More details needed” (Female, Age 79, Record 294)
“What are the side effects?” (Male, Age 68, Record 161)
Barriers to use
Physical limitations
“Had stroke 3 years ago and now have problems trying to read
on the internet” (Female, Age 76, Record 326)
“I am blind, but willing to use talking instruments” (Female, Age
58, Record 184)
Requires assistance
“Would need assistance from spouse for email/internet” (Male,
Age 69, Record 68)
“Good instruction, including video would be important for me
to make sure I was doing it correctly” (Male, Age 51, Record
291)
Lack of technology access
“Do not own a smartphone but is willing to learn” (Female, Age
66, Record 22)
“Do not have internet access” (Female, Age 59, Record 1)
Willingness to use
Enthusiasm
“It would be easier for me at home” (Female, Age 74, Record
269)
“Great idea!” (Male, Age 57, Record 29)
Opposition
“I am old school and would rather write” (Male, Age 64, Record
91)
“Do not use smartphone and have no desire to do so” (Male,
Age 68, Record 222)
technologies, or the level of eHealth literacy in this pop-
ulation, despite the growing interest and need to use
mHealth technologies for chronic disease management.
Estimates of the use of technology among those with
kidney disease have been reported to be lower than that in
the general US population,24,31,32 but the estimates were
based largely on individuals with kidney failure. In the
current study of individuals with CKD without kidney
AJKD Vol 77 | Iss 6 | June 2021
replacement therapy, a higher use of internet or email and
smartphones was found than in the prior reports. Greater
use of technology could be due to more recent timing of
data collection and potentially due to the exclusion of
individuals with kidney failure, who tend to be older and
have more comorbidities, which has been associated with
decreased use of technology.23,25,33
Despite the substantial use of digital and mHealth
technologies in the United States, a “digital divide” re-
mains, with an uneven distribution in access to, use of,
and potential benefits from these technologies,34 which is
driven by socioeconomic status, age, and racial minority
status.25,26,33,35,36 We also found that a digital divide is
present in the population with CKD, observing that older
age, membership in a racial or ethnic minority group, and
lower socioeconomic status (as measured by education
level and yearly income) had strong negative relationships
with use of all forms of technology. However, smartphone
technology may offer an opportunity to bridge the digital
divide as smartphones provide more internet access to
underserved communities than in-home internet.36-38
Additionally, we observed that Black and Hispanic in-
dividuals reported more interest in using mHealth tech-
nologies than White participants and that this interest was
not linked to education level and yearly income, further
supporting the fact that smartphone technology may help
overcome the digital divide to facilitate remote health
management.
As of 2014, more than half of smartphone users in the
United States (62%) reported using their phones to look
up information about a health condition,39 and 32%
specifically used medical-related apps.39-42 We also found
that approximately one-third of participants (35.9%) re-
ported using an mHealth app, primarily for learning about
health conditions, looking up nutrition or diet informa-
tion, or accessing their medical record, which is likely
driven by the types of mHealth apps available for those
with kidney disease that focus mainly on patient education
rather than tracking health information.43 mHealth apps
that focus on tracking health parameters, facilitating
remote monitoring, and supporting self-management be-
haviors could represent a major opportunity to improve
management of CKD, which are particularly relevant to the
remotely delivered nephrology care that has grown
exponentially during the coronavirus pandemic.44
Importantly, a recent systematic review found that
mHealth self-management interventions were highly
feasible and acceptable and have the potential to improve
management and health outcomes for CKD patients.45
However, the effectiveness of the interventions is still
inconclusive and warrants further research.
Among the present cohort of individuals with CKD
without kidney replacement therapy, several participants
opposed the future use of mHealth technologies. Those
opposed tended to be older and did not currently use the
947

internet or email and therefore might have had a low level
of “digital readiness,” which is the willingness to use
technology to navigate one’s environment, solve prob-
lems, or make decisions. A recent report has shown that
many US adults have low digital readiness.46 In order for
mHealth technologies to be effective, one must not only
have access to technology but also be willing to use it.
More research is needed to assess the readiness of using
mHealth technology among those with CKD.
A major concern reported by CRIC participants about
using digital and mHealth technologies included the fear
of losing face-to-face interaction with clinic staff. In prior
reports, individuals with chronic disease have reported that
they strongly prefer to receive health-related education and
advice from a professional47 and that most patients would
like mHealth in conjunction with visits to the doctors.48
Given that patients perceive quality in interactions with
providers, mHealth is likely to complement rather than
replace such interactions in the management of chronic
diseases.
Reported barriers to using mHealth technologies closely
linked to physical limitations, which likely relates to the
advancing age and growing number of comorbidities
among individuals with CKD. Generally, internet and
mHealth use has been shown to be inversely related to age
and number of medical conditions,25,33 with many older
adults reporting comorbid conditions, disabilities, or other
types of physical limitations that may prevent them from
fully using health-related digital technologies.23 More
research is needed to more clearly identify the reasons why
certain populations such as older adults do not use
mHealth technologies and if supporting the use of
mHealth technologies for CKD management can improve
outcomes.
It is important to target a population’s eHealth literacy
level to successfully leverage the capabilities of digital and
mHealth technologies for disease management. Assessing
eHealth literacy is especially vital for the CKD population
because many patients are already at risk for low use of
technology due to older age, lower socioeconomic status,
and inadequate health literacy. To date, eHealth literacy has
not been well characterized among individuals with CKD.
In this study, approximately three-fourths of the in-
dividuals had inadequate eHealth literacy, suggesting that
lack of adequate eHealth literacy will likely serve as a
barrier to fully leveraging the benefits of mHealth tech-
nologies in this population. Older age and less use of
technology were associated with inadequate eHealth lit-
eracy; both of which likely relate to the level of familiarity
and confidence in using technology to carry out tasks, as
demonstrated by findings from a 2017 Pew Research
Survey in which 34% of internet users 65 year of age and
older reported they had little to no confidence in their
ability to use electronic devices to perform online tasks,23
and our observation that confidence in managing one’s
chronic disease was associated with adequate eHealth lit-
eracy. In combination, these findings support the potential
948
Schrauben et al
role for perceived confidence and adequate eHealth literacy
in the CKD population.
This study’s strengths include representation of a large
number of well-characterized individuals from multiple
locations across the United States and inclusion of a wide
range of ages, including older adults, as well as several
racial and ethnic groups. Additionally, this is to our
knowledge the largest study to date to describe use of
technology and eHealth literacy in the population with
CKD in the United States. However, there are several
limitations. Due to the observational study design, the
study was unable to assess causality and is susceptible to
residual confounding, particularly by socioeconomic sta-
tus, as we were only able to adjust for education and
household income, and by location (urban vs rural),
which has been shown in other studies to influence
technology access.24 To best account for location, we
adjusted for recruitment site. Additionally, not all survey
questions were validated, and we had incomplete data for
response rates, both of which limit the external validity of
our findings. Questions assessing the interest in future use
of mHealth specifically focused on medication manage-
ment or participation in a research study, and this interest
may not apply to other mHealth applications. Further-
more, our qualitative data collection did not give the op-
portunity to expand on identified themes due to lack of
follow up or formal interview, which limits the depth of
exploration of individual experiences. Finally, ascertain-
ment bias is a concern because the study used a conve-
nience sampling strategy, whereas healthier and more
engaged individuals might have decided to complete the
surveys and participate in the CRIC study in general.
In summary, many individuals with CKD without
kidney replacement therapy currently use the internet,
email, and smartphones and are interested in using
mHealth in the future, but few use mHealth apps or
have adequate eHealth literacy. mHealth technologies
have the potential to allow for greater digital equity, as
smartphones can lower barriers to internet access for
underserved communities, racial minorities, and adults
with less education and those with lower incomes, and
also because minorities reported more interest in using
mHealth and digital technologies. Leveraging mHealth
represents a potential opportunity to engage individuals
with CKD, but further research needs to be conducted
regarding barriers to use, inadequate eHealth literacy, and
low digital readiness.
Supplementary Material
Supplementary File (PDF)
Item S1: Supplementary methods.
Table S1: Survey questions used to determine study outcomes.
Table S2: Characteristics of CRIC participants who did and did not
respond to the e-Literacy Survey, by question type.
Table S3: Associations of baseline characteristics with eHealth lit-
eracy domains.
AJKD Vol 77 | Iss 6 | June 2021
Schrauben et al
Article Information
CRIC Study Investigators: Jiang He, MD, PhD, Robert G. Nelson,
MD, PhD, MS, Panduranga S. Rao, MD, Mahboob Rahman, MD,
Vallabh O. Shah, PhD, MS, MD, and Mark L. Unruh, MD, MS.
Authors’ Full Names and Academic Degrees: Sarah J. Schrauben,
MD, MSCE, Lawrence Appel, MD, MPH, Eleanor Rivera, PhD, RN,
Claudia M. Lora, MD, James P. Lash, MD, Jing Chen, MD, MMSc,
L. Lee Hamm, MD, Jeffrey C. Fink, MD, Alan S. Go, MD, Raymond
R. Townsend, MD, Rajat Deo, MD, MTR, Laura M. Dember, MD,
Harold I. Feldman, MD, MSCE, and Clarissa J. Diamantidis, MD,
MHS.
Authors’ Affiliations: Department of Medicine, Perelman School of
Medicine at the University of Pennsylvania, Philadelphia, PA (SJS,
RRT, RD, LMD, HIF); Johns Hopkins University School of
Medicine and School of Public Health, Baltimore, MD (LA);
College of Nursing, University of Illinois Chicago, Chicago, IL (ER);
Department of Medicine, School of Medicine, University of Illinois
at Chicago, Chicago, IL (CML, JPL); Tulane University School of
Medicine, New Orleans, LA (JC, LLH); Department of Medicine,
School of Medicine, University of Maryland, Baltimore, MD (JCF);
Kaiser Permanentee Northern California, Oakland, CA (ASG);
Department of Medicine, Duke University School of Medicine,
Durham, NC (CJD); and Department of Biostatistics,
Epidemiology, and Informatics, Perelman School of Medicine at
the University of Pennsylvania, Philadelphia, PA (SJS, LMD, HIF)
Address for Correspondence: Sarah J. Schrauben, MD, MSCE,
930 Blockley Hall, 423 Guardian Drive, Philadelphia, PA 19104.
Email: Sarah.Schrauben@pennmedicine.upenn.edu
Author Contributions: Research idea and study design: SJS, CJD;
data acquisition: LA, JC, LLH, JCF, ASG, RRT, HIF, JPL, CJD; data
analysis/interpretation: SJS, ER, CML, JPL, JC, JCF, RD, LMD, HIF,
CJD; statistical analysis: SJS; supervision or mentorship: LMD, HIF,
CJD. Each author contributed important intellectual content during
manuscript drafting or revision and agrees to be personally
accountable for the individual’s own contributions and to ensure
that questions pertaining to the accuracy or integrity of any portion
of the work, even one in which the author was not directly
involved, are appropriately investigated and resolved, including
with documentation in the literature if appropriate.
Financial Disclosure: The authors declare that they have no
relevant financial interests.
Support: This work was supported by grants K23-DK099385-01A1
(CJD) and K23DK118198-01A1 (SJS) from the National Institute of
Diabetes and Digestive and Kidney Diseases (NIDDK). RRT has
received grants from US National Institutes of Health (NIH).
Funding for the CRIC study was obtained under a cooperative
agreement with NIDDK diseases (grants U01DK060990,
U01DK060984, U01DK061022, U01DK061021, U01DK061028,
U01DK060980, U01DK060963, U01DK060902, and
U24DK060990). This work was also supported by a Perelman
School of Medicine, University of Pennsylvania Clinical and
Translational Science Award (NIH/NCATS UL1TR000003); Johns
Hopkins University (UL1 TR-000424); University of Maryland
(GCRC M01 RR-16500); Clinical and Translational Science
Collaborative of Cleveland; National Center for Advancing
Translational Sciences (NCATS) component of the NIH and NIH
roadmap for Medical Research (award UL1TR000439); Michigan
Institute for Clinical and Health Research (UL1TR000433);
University of Illinois at Chicago Clinical and Translational Science
Award (UL1RR029879); Tulane Center of Biomedical Research
Excellence award for Clinical and Translational Research in
Cardiometabolic Diseases (P20 GM109036); Kaiser Permanentee
NIH/ National Center for Research Resources University of
California San Francisco award (CTSI UL1 RR-024131); and
Department of Internal Medicine, University of New Mexico School
AJKD Vol 77 | Iss 6 | June 2021
of Medicine, Albuquerque, NM (R01DK119199). The funders had
no role in study design, data collection, analysis, reporting, or the
decision to submit for publication.
Peer Review: Received March 1, 2020. Evaluated by 3 external peer
reviewers and a statistician, with editorial input from an Acting
Editor-in-Chief (Editorial Board Member Allison Tong, PhD).
Accepted in revised form October 8, 2020. The involvement of an
Acting Editor-in-Chief to handle the peer-review and decision-
making processes complied with AJKD’s procedures for potential
conflicts of interest for editors, described in the Information for
Authors & Journal Policies.
References
1. Kay M, Santos J, Takane M. mHealth: new horizons for health
through mobile technologies. World Health Organization
Global Observatory for e-Health Series Web site. 2011.
Accessed June 1, 2019. http://www.who.int/goe/publications/
goe_mhealth_web.pdf
2. Fjeldsoe BSMA, Miller YD. Behavior change interventions
delivered by mobile telephone short-messaging service. Am J
Prev Med. 2009;36(2):165-173.
3. Hurling RCM, Boni MD, Fairley BW, et al. Using internet and
mobile phone technology to deliver an automated physical
activity program: randomized controlled trial. J Med Internet
Res. 2007;9(e7). Author: update ref 3 to include complete
page range.
4. Schrauben S, Hsu J, Rosas S, et al. CKD self-management:
phenotypes and associations with clinical outcomes. Am J
Kidney Dis. 2018;72:360-370.
5. Ong S, Jassal S, Miller J, et al. Integrating a Smartphone-
based self-management system into usual care of
advanced CKD. Clin J Am Soc Nephrol. 2016;11:1054-
1062.
6. Becker S, Kribben A, Mesiter S, Diamantidis C, Unger N,
Mitchell A. User profiles of a Smartphone application to support
drug adherence-experiences from the iNephro project. PLoS
One. 2013;8(e78547).
7. Tuot D, Diamantidis C, Corbett C, et al. The last mile: trans-
lational research to improve CKD outcomes. Clin J Am Soc
Nephrol. 2014;9:1802-1805.
8. Feldman HI, Appel LJ, Chertow GM, et al. The Chronic Renal
Insufficiency Cohort (CRIC) study: design and methods. J Am
Soc Nephrol. 2003;14(Suppl 2):S148-S153.
9. Lash JP, Go AS, Appel LJ, et al. Chronic Renal Insufficiency
Cohort (CRIC) study: baseline characteristics and associations
with kidney function. Clin J Am Soc Nephrol. 2009;4:
1302-1311.
10. Nguyen J, Moorhouse M, Curbow B, Christie J, Walsh-
Childers K, Islam S. Construct validity of the eHealth Literacy
Scale (eHEALS) among two adult populations: a Rasch anal-
ysis. JMIR Public Health Surveill. 2016;2:e24.
11. Norman C, Skinner H. eHEALS: the eHealth Literacy Scale.
J Med Internet Res. 2006;8:e27.
12. Vart P, Gansevoort R, Joosten M, Bultmann U, Reijneveld S.
Socioeconomic disparities in chronic kidney disease: a sys-
tematic review and meta-analysis. Am J Prev Med. 2015;48:
580-592.
13. Butler J, Carter M, Hayden C, et al. Understanding adoption of
a personal health record in rural health clinics: revealing bar-
riers and facilitators of adoption including attributions about
potential patient portal users and self-reported characteristics
of early adopting users. AMIA Annu Symp Proc. 2013;2013:
152-161.
949

14. Berkman N, Seridan S, Donahue E, Halpern D, Crotty K. Low
health literacy and health outcomes: an updated systemic re-
view. Ann Intern Med. 2011;155(2):97-107.
15. Charlson M, Pompei P, Ales K, MacKenzie C. A new method of
classifying prognostic comorbidity in longitudinal studies:
development and validation. J Chronic Dis. 1987;40(5):373-
383.
16. Baker DW, Williams MV, Parker RM, Gazmararian JA, Nurss J.
Development of a brief test to measure functional health liter-
acy. Patient Educ Couns. 1999;38:33-42.
17. Craven JL, Rodin GM, Littlefield C. The Beck Depression
Inventory as a screening device for major depression in
renal dialysis patients. Int J Psychiatry Med. 1988;18:365-
374.
18. Teng EL, Chui HC. The Modified Mini-Mental State (3MS) ex-
amination. J Clin Psychiatry. 1987;48(8):314-318.
19. Lubben J, Blozik E, Gillmann G, et al. Performance of an
abbreviated version of the Lubben Social Network Scale
among three European community-dwelling older adult pop-
ulations. Gerontologist. 2006;46(4):503-513.
20. Lorig K, Sobel D, Ritter P, et al. Effect of a self-management
program for patients with chronic disease. Eff Clin Pract.
2001;4:256-262.
21. Anderson A, Yang W, Hsu C, et al. Estimating GFR among
participants in the Chronic Renal Insufficiency Cohort (CRIC)
study. Am J Kidney Dis. 2012;60:250-261.
22. Pew Research Center. Record shares of Americans now own
Smartphones, have home broadband. FACTANK: News in the
Numbers. January 12, 2017. Accessed May 17, 2019. www.
pewresearch.org/fact-tank/2017/01/12/evolution-of-technology/
23. Anderson M, Perrin A. Tech adoption climbs among older
adults. Pew Research Center. 2017. Accessed June 1, 2019.
https://www.pewinternet.org/2017/05/17/tech-adoption-climbs-
among-older-adults/
24. Pew Research Center. Internet & Technology. 2018. Accessed
June 1, 2019. https://www.pewinternet.org/fact-sheet/
25. Kontos E, Blake K, Chou W, Prestin A. Predictors of eHealth
usage: insights of the digital divide from the Health Information
National Trends Survey 2012. J Med Internet Res. 2014;16:
e172.
26. Sarkar U, Karter A, Liu J, et al. Social disparities in internet
patient portal use in diabetes: evidence that the digital divide
extends beyond access. J Am Med Inform Assoc. 2011;18:
318-321.
27. Zou G. A modified Poisson regression approach to prospective
studies with binary data. Am J Epidemiol. 2004;159:702-706.
28. Greenland S. Model-based estimation of relative risks and
other epidemiologic measures in studies of common outcomes
and in case-control studies. Am J Epidemiol. 2004;160:301-
305.
29. Behrens T, Taeger D, Wellmann J, Keil U. Different methods to
calculate effect estimates in cross-sectional studies. A com-
parison between prevalence odds ratio and prevalence ratio.
Methods Inf Med. 2004;43:505-509.
30. Bryant A, Charmaz K. The SAGE Handbook of Grounded
Theory. Thousand Oaks, CA: Sage Publisher; 2007.
31. McGillicuddy J, Weiland A, Frenzel R, et al. Patient attitudes
toward mobile phone-based health monitoring: questionnaire
950
Schrauben et al
study among kidney transplant recipients. J Med Internet Res.
2013;15(1):e6.
32. Bonner A, Gillespie K, Campbell K, et al. Evaluating the prev-
alence and opportunity for technology use in chronic kidney
disease patients: a cross-sectional study. BMC Nephrology.
2018;19:28.
33. Choi N, Dinnito D. The digital divide among low-income
homebound older adults: internet use patterns, eHealth liter-
acy, and attitudes toward computer/internet use. J Med
Internet Res. 2013;15(5):e93.
34. Lorence D, Park H, Fox S. Racial disparities in health informa-
tion access: resilience of the digital divide. J Med Syst.
2006;30(4):241-249.
35. Yamin C, Emani S, Williams D, et al. The digital divide in
adoption and use of a personal health record. Arch Intern Med.
2011;171:568-574.
36. Smith A. A Portrait of Smartphone Ownership. Pew Research
Institute. 2015. Accessed June 1, 2019. http://www.
pewinternet.org/2015/04/01/chapter-one-a-portrait-of-smartphone-
ownership/
37. Zickuhr K, Smith A. Digital differences. Pew Research Center;
2012. Accessed June 1, 2019. http://www.pewinternet.org/
files/old-media/Files/Reports/2012/PIP_Digital_differences_
041312.pdf
38. Pew Research Center: Mobile Access. 2010. Accessed June
5, 2019. http://www.pewinternet.org/Reports/2010/Mobile-
Access-2010.aspx
39. Smith A. Smartphone Use in 2015. Pew Research Center;
2015. Accessed June 1, 2019. http://www.pewinternet.org/2
015/04/01/us-smartphone-use-in-2015
40. Lella A. The U.S. Mobile App Report. Published 2014.
Accessed June 5, 2019. https://www.comscore.com/Insights/
Presentations-and-Whitepapers/2014/The-US-Mobile-App-
Report
41. Japsen B. Health app usage soars as consumers take charge.
http://www.forbes.com/sites/brucejapsen/2015/12/10/health-
app-usage-soars-as-prescription-for-consumer-medical-needs/
#20352b5859e
42. Fox S, Duggan M. Tracking for Health, 2013. Pew Research
Center. 2013. Accessed June 5, 2019. http://www.
pewinternet.org/2013/01/28/tracking-for-health
43. Singh K, Diamantidis C, Ramani S, et al. Patients’ and ne-
phrologists’ evaluation of patient-facing Smartphone apps for
CKD. Clin J Am Soc Nephrol. 2019;14:523-529.
44. Jain G, Ahamd M, Wallace E. Technology, telehealth and
nephrology: The time is now. Kidney360. 2020;1(8):834-836.
45. Shen H, van der Kleij R, van der Boog P, Chang X,
Chavannes N. Electronic health self-management interventions
for patients with chronic kidney disease: systematic review of
quantitative and qualitative evidence. J Med Internet Res.
2019;21(11):e12384.
46. Center PR. Digital Readiness Gaps. September. September
20, 2016 2016.
47. Kuehn B. Patients on online seeking support, practical advice
on health conditions. JAMA. 2011;305(16):1664-1665.
48. Rai A, Chen L, Pye J, Baird A. Understanding determinants of
consumer mobile health usage intentions, assimilation, and
channel preferences. J Med Internet Res. 2013;15(8):e149.
AJKD Vol 77 | Iss 6 | June 2021
Schrauben et al

mHealth Technology: Availability, Acceptability, and Use in CKD

Setting & Participants mHealth & eHealth Literacy Perspectives

p of mHealth

Internet
Int and smartphone use Q
Qualitative analysis of
was high (>70%), but mHealth
wa o
open-text responses,
app use was low (36%)
ap 3 themes emerged:
2016-2018
• Intrusive
Black and Hispanic
B Concerns
• No face-to-face
participants more likely to
pa
report interest in using
re • Physical limitations
932
2 participants Barriers • Assistance needed
mHealth compared to Whites
m
• Mean age 68 • Lack of access
• 41% Female • More participants
• 62% Black or Hispanic A
Adequate eHealth literacy was
Willingness were enthusiastic
• mean eGFR 54 ml/min/1.73m2 lo
low (27%) overall
than opposed

CONCLUSION: Many CKD patients use internet/smartphones and are interested in using
CON
mHealth technologies, but few use mHealth apps, and overall proficiency to use mHealth is low.
Sarah J. Schrauben, Lawrence Appel, Eleanor Rivera, et al (2020)
@AJKDonline | DOI: 10.1053/j.ajkd.2020.10.013

AJKD Vol 77 | Iss 6 | June 2021 950.e1

Policy Forum Perspective

Racism and Kidney Health: Turning Equity

Into a Reality
Dinushika Mohottige, Clarissa J. Diamantidis, Keith C. Norris, and L. Ebony Boulware

Kidney disease continues to manifest stark racial inequities in the United States, revealing the Complete author and article
entrenchment of racism and bias within multiple facets of society, including in our institutions, prac- information provided before
references.
tices, norms, and beliefs. In this perspective, we synthesize theory and evidence to describe why an
understanding of race and racism is integral to kidney care, providing examples of how kidney health Am J Kidney Dis.
disparities manifest interpersonal and structural racism. We then describe racialized medicine and 77(6):951-962. Published
online February 24, 2021.
“colorblind” approaches as well as their pitfalls, offering in their place suggestions to embed antiracism
and an “equity lens” into our practice. We propose examples of how we can enhance kidney health doi: 10.1053/
equity by enhancing our structural competency, using equity-focused race consciousness, and j.ajkd.2021.01.010
centering investigation and solutions around the needs of the most marginalized. To achieve equitable © 2021 Published by
outcomes for all, our medical institutions must embed antiracism and equity into all aspects of Elsevier Inc. on behalf of the
advocacy, policy, patient/community engagement, educational efforts, and clinical care processes. National Kidney Foundation,
Inc.
Organizations engaged in kidney care should commit to promoting structural equity and eliminating
potential sources of bias across referral practices, guidelines, research agendas, and clinical care.
Kidney care providers should reaffirm our commitment to structurally competent patient care and
educational endeavors in which empathy and continuous self-education about social drivers of health
and inequity, racism, and bias are integral. We envision a future in which kidney health equity is a reality
for all. Through bold collective and sustained investment, we can achieve this critical goal.

Introduction field of nephrology can chart a path toward FEATURE EDITOR:

kidney health equity. Daniel E. Weiner
To get to a point where race won’t make a
difference, we have to wrestle, first, with ADVISORY BOARD:
Understanding Race and Racism:
the difference that race makes. L. Ebony Boulware
Integral to US Nephrology Practice Kevin Erickson
Michael Eric Dyson1 Eduardo Lacson Jr
It is time to repair the effects of racism on The idea of race was in fact a deliberate Bruce M. Robinson
kidney health. In recent years, profound racial creation of an exploiting class which was Wolfgang Winkel-
and ethnic kidney health disparities have been seeking to maintain its privileges against mayer
among a litany of inequalities magnified by a what was profitably regarded as an inferior
social caste. Policy Forum high-
number of factors, including the coronavirus lights aspects of
disease 2019 (COVID-19) pandemic, in the Montague Francis Ashley-Montagu (born Israel nephrology relating to
United States.2,3 Longstanding racial and ethnic Ehrenberg)17 payment and social
disparities in the burden of kidney disease have policy, legislation,
been well described over the past decade,4-6 Race regulation, de-
mographics, politics,
providing overwhelming evidence of the The concept of race was originally imple- and ethics, contextu-
inextricable links between unequal resources, mented as a skin color–based classification alizing these issues
opportunities, and social contexts shaped by scheme and was introduced into scientific as they relate to the
racialization/racism and poor kidney health.7,8 writing by Linnaeus and others.18,19 Race lives and practices of
Black individuals who have borne a dispro- evolved into an implicitly and explicitly or- members of the kid-
ney community,
portionate burden of racism-driven inequality dered tool for the racialization (see Box 1 for including providers,
have greater prevalence of chronic kidney dis- definitions) and hierarchical categorization of payers, and patients.
ease (CKD) risk factors, are 2-4–fold more individuals. Once used to justify atrocities (eg,
likely than White individuals to have kidney chattel slavery, eugenics), race remains deeply
failure,9 and experience persistently worse ac- embedded within present-day policies,
cess to kidney transplants, compounded by norms, practices (eg, medical algo-
decreasing rates of access to living-donor kid- rithms),17,20 and other social structures.21 The
ney transplants.10-12 There is an urgent ethical Human Genome Project, the American Society
imperative for the nephrology community to of Physical Anthropology, and others have
disrupt this incessant trajectory of inequity. In categorically debunked dangerous assertions
this perspective, we reference transdisciplinary that race provides meaning about genetic
scholarship13-16 to advance a dialog on how the difference.17,22-25 Despite this, the

AJKD Vol 77 | Iss 6 | June 2021 951

 Mohottige et al

Box 1. Definitions for Terms Used Throughout Discussion

Antiracism: “An active and consistent process of change to eliminate individual, institutional, and systemic racism and redistribute
power and privilege.”120
Centering at the margins: Ensuring socially marginalized group perspectives are the “central axis around which discourse
revolves.”14,121
Discrimination: “Differential actions toward others based on race and/or other social domains.”26
Equality: “The state of being equal, especially in rights, status, advantages, etc.”122
Equity: “Assurance of the conditions for optimal health of all people” that “requires valuing all individuals and populations equally,
recognizing and rectifying historical and contemporary injustices, and providing resources according to need.”78,123
Health disparities: “Systematic…avoidable health differences according to race/ethnicity, skin color, religion, or nationality; so-
cioeconomic resources or position (reflected by, income, wealth, education, or occupation); gender, sexual orientation, gender
identity; age, geography, disability, illness, political or other affiliation; or other characteristics associated with discrimination or
marginalization. These categories reflect social advantage or disadvantage when they determine an individual’s or group’s po-
sition in a social hierarchy.”124
Personally mediated racism: Involves “differential assumptions about abilities, motives, and intent of others by ‘race,’ differential
actions based on those assumptions, and prejudice and discrimination.”26
Prejudice: “Differential assumptions about the abilities, motives, and intents of others by race and/or other social domains.”26
Race-based medicine: A system by which race may be characterized as “an essential biological variable” and subsequently
embedded into clinical practice and other guidelines and algorithms.20
Racism: “[A] system of structuring opportunity and assigning value based on the social interpretation of how one looks (ie, ‘race’),
that unfairly disadvantages some individuals and communities, unfairly advantages other individuals and communities, and saps
the strength of the whole society through the waste of human resources.”26
Structural racism: Macro-level systems, social forces, institutions, ideologies, and processes that interact with one another to
create and perpetuate inequities disadvantaging racial and ethnic groups and, when established, continue without the
requirement for individual actions or intent.125
Structural competency: “[Recognition] of the ways that institutions, neighborhood conditions, market forces, public policies, and
health care delivery systems shape symptoms and diseases.”83
Racialization: “Processes by which a group of people is defined by their ‘race,’ [which may] begin by attributing racial meaning to
people’s identity and, in particular, as they relate to social structures and institutional systems, such as housing, employment, and
education. In societies in which ‘White’ people have economic, political, and social power, processes of racialization have
emerged from the creation of a hierarchy in social structures and systems based on ‘race.’ The visible effects of processes of
racialization are the racial inequalities embedded within social structures and systems.”126
White supremacy: “A term used to characterize various belief systems central to which are one or more of the following key
tenets: 1) Whites should have dominance over people of other backgrounds, especially where they may coexist; 2) Whites
should live by themselves in a Whites-only society; 3) White people have their own ‘culture’ that is superior to other cultures; 4)
White people are genetically superior to other people.”127 These tenets underlie the laws, policies, and practices that govern and
operationalize all of the social determinants of health in the United States and many other nations.

fortification of race in our social bedrock is evidenced in Racism

many health arenas. This includes efforts to “adjust” or
account for race as a biological contributor to differential
health risks,26 often with a lack of appreciation for why
race imposes a health impact in the context of a racialized
society.27 With few exceptions, discoveries of racially
disparate kidney health outcomes reflect the byproducts of
interdependent vehicles through which racism (eg, racial
segregation of neighborhood resources via redlining/dif-
ferential housing lending/mortgages,28 wealth
29
inequality ) disadvantages some individuals while
affording privileges to others. Egregious kidney health
disparities impacting individuals racialized as Black in the
United States provide an important lens to examine con-
tributions of race- and racism-related mechanisms to kid-
ney health disparities across multiple domains of kidney
care (Table 1).
Racism is the primary mechanism through which socio-
political constructions exert their influences on individuals
who have been racialized.30 Racism operates at individual,
interpersonal, and systemic or structural levels. Personally
mediated racism, which can be overt or covert, intentional
or unintentional, is manifest primarily by prejudice and
discrimination.26 Described sometimes as “everyday
racism,”31 racial prejudice and discrimination occurring
within and outside of health care may contribute to poor
health and detrimental health-related behaviors.32,33 Lack
of respect and dignity (eg, providing suboptimal medical
options or pain management on the basis of race34),
“hidden” medical judgments (eg, presenting clinical cases
in which race is implicitly linked to negative characteris-
tics, including nonadherence35,36), suspicion (eg, having
security alerted to patients while in a health care facility37),

952 AJKD Vol 77 | Iss 6 | June 2021

Mohottige et al
Table 1. Domains of Kidney Care, Examples of Disparities Impacting Black Individuals in the United States, and Possible Links to
Various Levels of Racism
Domain Example of Disparate Outcomes
CKD • Greater CKD incidence in Black vs White
individuals128
• Greater mortality among young (<65 y) Black
vs White individuals with CKD129
• Delayed referral to nephrology care and less
overall receipt11 of predialysis kidney care in
Black vs White patients130,131
Kidney failure • Higher mortality rates among young (18-50 y)
and dialysis care Black vs White patients receiving dialysis135
• Poorer quality dialysis among Black vs White
patients135
• Less likelihood of receiving home dialysis
modalities among Black vs White patients136
Transplant • Lower LDKT incidence and preemptive kidney
transplant in Black patients,140 with growing
disparities141
• Less likelihood of Black individuals to be
identified as potential transplant candidates,
referred for evaluation, or listed for
DDKT8,49,142,143
Abbreviations: CKD, chronic kidney disease; DDKT, deceased-donor kidney transplantation; HTN, hypertension; LDKT, living-donor kidney transplantation; SES, socio-
economic status.
scapegoating (eg, blaming a group for a problem, including
use of terms like “the Chinese virus”38), and dehumaniza-
tion (eg, eugenics, police brutality39) are examples of
the many ways personally mediated racism operates in
health care.39 Implicit or subconscious bias, including racial
bias, permeates all individual beliefs and behavior, with
impacts on health care interactions, treatment decisions,
and health outcomes.40 Even when explicit beliefs about
minoritized patients (eg, that racial minorities have greater
nonadherence and more “risky behaviors” vs White in-
dividuals) are ostensibly modified, susceptibility to implicit
bias may be a threat, particularly when work conditions
are stressful or institutional culture reinforces bias.40
At the systemic and structural levels, racism is a
powerful construct through which the institutional codi-
fication of unequal power (eg, voter disenfranchisement41
and disproportionately low representation in leadership
AJKD Vol 77 | Iss 6 | June 2021
Evidence (Level of Racism)
• Greater exposure among Black (vs White) individuals to
discrimination and greater neighborhood-level stressors
(linked to higher HTN prevalence)132 (structural)
• 10% excess risk of CKD attributed to less use of a
usual source of care among insured and uninsured
Black individuals133 (structural)
• Racial residential segregation (higher percentage of
Black individuals) associated with decreased likelihood
of receiving predialysis care134 (structural)
• Higher mortality rates and poorer quality dialysis linked
to neighborhood composition137 (structural)
• Higher incidence of kidney risk in Black vs White in-
dividuals who live in areas characterized by poverty138
(structural)
• In young incident dialysis patients, greater mortality in
Black vs White patients, with differences worse in low-
SES neighborhoods50 (structural)
• Less offering of home dialysis in dialysis units primarily
serving racial minorities139 (interpersonal)
• Higher perceived racism and discrimination cited
among Black (vs White) transplant candidates144
(interpersonal)
• Low provider knowledge about the existence and cau-
ses of racial transplant inequity145 (interpersonal)
• Medical mistrust (due to historical and contemporary
mistreatment), experienced discrimination, and
perceived racism is associated with lower transplant
evaluation initiation117 (interpersonal)
• Black persons in poor predominantly Black neighbor-
hoods are less likely to be waitlisted for transplant than
counterparts in wealthy predominantly Black neighbor-
hoods and in poor predominantly White neighbor-
hoods146 (structural)
• Dialysis units in neighborhoods characterized by higher
proportions of Black individuals, lower educational
attainment, and more severe poverty are associated
with lower transplant rates147 (structural)
• High influence of unstandardized and contextually
influenced factors like “social support” and “adherence”
in determining transplant listing decisions148
(interpersonal)
structures, including US government and medical in-
stitutions42) and material conditions (eg, housing, health
care, employment, health-promoting environmental con-
ditions26) is realized. Structural racism refers to “mutually
reinforcing inequitable systems”43 (including policies)
that have created health-harming environments in
marginalized communities. Social structures including
health care institutions and policies influence kidney health
by creating “durable, patterned arrangements of resources,
opportunities, and norms,”44 which can produce and
maintain social and racial inequities.
Reductionist conceptualizations of kidney disease as a
product of purely biological or behavioral factors ignore
root structural causes, limiting opportunities to intervene
on root causes of disparities in kidney health.45 Consider
how the racial patterning of neighborhood demographics
(ie, via redlining policies and discriminatory mortgage/
953
 Mohottige et al

Box 2. Approaches to Apply Antiracist Principles in Kidney Care, Education, Research, Advocacy and Policy, and Clinical Care
Domains
Education
1. Promote structural competency as a core competency in nephrology training
• Educate about the links between sociocontextual factors and disparate kidney health outcomes (eg, transplant access,
predialysis care). (Antiracist principles: equity-focused race consciousness, structural competency)
2. Enhance education about race and racism with explicit consideration of racialized/race-based practices
• Define and explore the context and reason for including measures like race in study design, ensuring that race and ethnicity
(and other social variables) are not espoused as embodying biological truths. (Antiracist principle: equity-focused race
consciousness)
• Integrate rigorous genetic, physiologic, and social-science education into curriculum along with epidemiologic data regarding
existing disparities. (Antiracist principle: equity-focused race consciousness)
3. Apply an intersectional lens to kidney health equity education
• Integrate core concepts about how race, gender, immigration status, sexual orientation, and other social identities are linked to
experiences of marginalization, potentially impairing multiple facets along the continuum of kidney care. (Antiracist principles:
equity-focused race consciousness, structural competency)
4. Ensure that training environments are free of bias, harassment, and discrimination
• Implement antidiscrimination and supportive bystander intervention training across training programs, with a goal of promoting
equity and inclusivity. (Antiracist principle: equity-focused race consciousness)
5. Embed antiracism and equity into ongoing education for nephrology providers
• Consider CME opportunities and opportunities to integrate key equity concepts into professional certification. (Antiracist
principles: equity-focused race consciousness, structural competency)
Research
1. Define and contextualize social variables like race and ethnicity
• Distinguish between race, racism (and other variables including, sex, gender, etc) and counteract the erroneous conflation of
biological risk, race, and other social identity markers, recognizing that the risk is often conferred in the process of margin-
alization toward a group (eg, racism). (Antiracist principle: structural competency)
2. Promote the rigorous study of how racism and other forms of marginalization are associated with kidney health inequity
• Expand funding opportunities and training to ensure robust infrastructure for lines of research that address policy reform and
broad structural interventions targeting kidney disease disparities. (Antiracist principles: equity-focused race consciousness,
structural competency)
3. Center patient and community expertise in the process of identifying and solving the most pressing problems in nephrology
• Engage community stakeholders and patients throughout the research process, with careful attention to transparency and
aligning research priorities and processes. (Antiracist principle: centering at the margins)
4. Enhance trustworthiness of the research enterprise
• Earn trust, actively confront and dismantle barriers to trustworthiness (tailor studies to the needs of patients/communities who
are served, with a lens toward equity in participation) and engagement. (Antiracist principle: centering at the margins)
• Invest in sustainable and collaborative partnerships with kidney care–focused community organizations serving marginalized
individuals. (Antiracist principle: centering at the margins)
Advocacy and Policy
1. Amplify voices of marginalized communities and individuals when defining advocacy priorities and strategies
• Engage community stakeholders and patients throughout ASN/NKF/organizational processes when determining key policy
agendas and priorities. (Antiracist principle: centering at the margins)
2. Identify and address the role of organizations and institutions in perpetuating and creating inequities
• Apply an equity lens to the consideration of existing and proposed policies, recognizing potential and disproportionate harms
caused by existing financial structures/schemes. Establish a required periodic reevaluation of all policies to assess outcomes
and ensure no unintended consequences have occurred. (Antiracist principle: structural competency)
3. Address prevention and early CKD management as critical advocacy areas to disrupt racial kidney health inequities
• Invest in promoting insurance coverage for critical medications like SGLT2i and expansion of health care coverage, including
prevention. (Antiracist principle: structural competency)
4. Integrate a social justice– and equity-focused framework throughout all policy development and evaluation processes (eg,
ensure transparent evaluation of gains and losses, investments, and perverse unidimensional financial incentives)
• Promote cost neutrality for donors and recipients in LDKT. (Antiracist principle: structural competency)
• Promote lifelong coverage for immune suppression posttransplant regardless of payer source. (Antiracist principle: structural
competency)
• Consider alternate equity-focused schemes to increase equity in kidney transplantation (based on need, differential rates of
progression among groups). (Antiracist principle: structural competency)

(Continued)

954 AJKD Vol 77 | Iss 6 | June 2021

Mohottige et al

Box 2 (Cont'd). Approaches to Apply Antiracist Principles in Kidney Care, Education, Research, Advocacy and Policy, and Clinical
Care Domains
Clinical care
1. Embed an equity and antiracist/antibiased lens into clinical care processes
• Reconsider and dismantle potential sources of bias in kidney transplant candidacy evaluation (ie, restrictive policies around
social support, reconsideration of factors leading to “poor adherence,” including awareness of effects of structural racism).
(Antiracist principles: equity-focused race consciousness, structural competency)
2. Invest in structural interventions at the health system level to address social drivers of inequity
• Apply an equity lens to the consideration of existing and proposed policies (eg, dialysis payment, transplantation reim-
bursement), recognizing potential and disproportionate harms caused by existing financial structures/schemes. (Antiracist
principles: equity-focused race consciousness, structural competency)
3. Ensure clinical care environments are inclusive and equity-focused
• Educate providers and staff with a common language to understand bias, racism, and other forms of marginalization necessary
to create inclusive environments (eg, inclusive communication strategies, intake forms). (Antiracist principle: structural
competency)
4. Embed unbiased clinical practice alerts into care systems
• Develop electronic health tools that bypass potential provider biases (eg, trigger to consider discussions regarding SGLT2i,
transplant referral). (Antiracist principles: equity-focused race consciousness, structural competency)
5. Incorporate equity evaluations into interval outcome/quality assessments
• Analyze data regarding outcomes, referrals using an equity lens across race, ethnicity, and other social domains. (Antiracist
principle: equity-focused race consciousness)
• Consider equity when developing risk prediction models and waitlisting and other decisions (ie, dissemination of therapeutic
agents). (Antiracist principles: equity-focused race consciousness, structural competency)
6. Resist a deficit mindset and reframe disparities within the context of structural inequity
• Use a strengths-based approach to counteract narratives regarding individual or community “deficits” as a cause of health
problems; rather, contextualizing behaviors (ie, nonadherence) within lived experiences and structural barriers. (Antiracist
principles: equity-focused race consciousness, structural competency)
Abbreviations: ASN, American Society of Nephrology; CKD, chronic kidney disease; CME, continuing medical education; LDKT, living-donor kidney transplantation;
NKF, National Kidney Foundation; SGLT2i, sodium/glucose cotransporter 2 inhibitor.

lending policies46,47) and associated resources (including
food, health care access, housing, and education) are
associated with poor kidney health outcomes, including
longer time to transplant,48 lower rates of transplant
waitlisting,49 higher dialysis mortality rates,50 suboptimal
dialysis outcomes,19,51 and other poorer outcomes among
Black individuals compared with others (Table 1). Despite
our best efforts to individualize patients’ care, our efforts
may fail when we cannot rectify structural barriers. For
instance, provision of an evidence-based bundle of diabetic
kidney disease care (eg, sodium/glucose cotransporter 2
inhibitors) may fail to alter the course of individual and
population-level kidney outcomes if underlying structural
health barriers such as inadequate health care coverage and
access to these medications are not addressed.
It is important to recognize that well-established asso-
ciations between socioeconomic class status and racial
categories in the United States6,46 are rooted in historical
events (eg, slavery, Jim Crow–era laws, New Deal policies,
and US Federal Housing Authority policies, including
redlining47 and the GI Bill52,53) and are perpetuated by
present-day expressions of racism (including health care
segregation; education, housing, and employment
discrimination54; unequal access to savings/investment
vehicles29,55; disproportionate incarceration/criminaliza-
tion56; voter disenfranchisement41; and police violence57).
Hence, the assessment of socioeconomic status and race are
both important because a principal feature of racism in the
United States has been to prevent people with shared class
interests from acting in their common interest to achieve a
more equitable and just society.58 Further, it is important
to recognize that race and socioeconomic status are not
equivalent concepts, even though both represent impor-
tant domains through which societal resources, opportu-
nity, and health-impacting exposures are unequally
distributed. Because economic resources and social op-
portunities are often racialized, it is frequently difficult to
disentangle the effects of socioeconomic disenfranchise-
ment from racism in a racialized society.59 Despite this,
numerous studies have shown that, even after accounting
for socioeconomic status (eg, income), exposure to
structural and interpersonal racism, including racism-
mediated life-course experiences (eg, stress, poverty),
negatively influences health outcomes.60
Considering Impacts of Racialized Medicine
I feel that if we don’t take seriously the ways in which
racism is embedded in structures of institutions, if we
assume that there must be an identifiable racist who is
the perpetrator, then we won’t ever succeed in eradi-
cating racism.
Angela Y. Davis61

AJKD Vol 77 | Iss 6 | June 2021 955


Racialized medical practices embed the social construct of
race into medical practice without careful consideration of
how and why race influences health outcomes.22,62 Careful
interrogation of the implicit use of race in scientific
discourse, predictive algorithms (including artificial intel-
ligence), and clinical practice is important because race: 1)
lacks biological meaning, 2) is dynamic and contextually
sensitive in how and by whom it is defined, and 3) often
reinforces erroneous beliefs regarding the inferiority and
“otherness” of minoritized groups.20 Despite scientific
evidence that race is not biologically meaningful,6,22-26
examples of race-based medical practice are far-reaching
and are prominent in kidney care.63,64 This includes the
incorporation of a Black racial coefficient into kidney
function estimating equations. Although developed with
the intent of enhancing the precision of prior kidney
function estimating equations, a plausible biological
mechanism to explain observed differences in kidney
function on the basis of race is lacking. Further, emerging
evidence suggests that conferring individuals racialized
as Black with differential estimated glomerular filtration
rate via the Chronic Kidney Disease Epidemiology
Collaboration (CKD-EPI) equation may contribute to
harm and racial inequities impacting Black individuals
through delays in diagnoses and guideline-based care
kidney care, contributing to profound inequities in
kidney health outcomes.65-67
Race-based medical practice is also deeply ingrained
into the management of other chronic diseases, including
hypertension. Examples include: 1) Joint National Com-
mittee V guidelines for blood pressure management,
which reified the notion that Black patients did not
respond as well to angiotensin-converting enzyme in-
hibitors or β-blockers68; and 2) pharmaceutical market-
ing of BiDil (isosorbide dinitrate/hydralazine HCl) as a
race-specific drug for patients with heart failure despite
a paucity of evidence to suggest this.69 Sometimes
described as “medical racial profiling,” these examples
highlight how sometimes “race [is used] as a rationale for
differential treatment of patients on the basis of pre-
dictions regarding prevalence of complications and
response to specific therapies,”70 as well as for generating
profits. However, other examples demonstrate how race
could be incorporated into practice guidelines in align-
ment with a goal of reducing racial disparities and
increasing health equity. For instance, atherosclerotic risk
prediction models and American Heart Association lipid
management recommendations altered the threshold for
statin agent eligibility, thus considering racial disparities
in prescriptions for statin agents71 and the dispropor-
tionately higher risk of atherosclerotic cardiovascular
disease events in Black individuals.72 Finally, although
some have proposed using self-reported race, ethnicity,
or ancestry as determining factors to screen for high-risk
genetic polymorphisms like APOL1 variants, it is impor-
tant to recognize: 1) the lack of precision or consistency
in use of terms like race/ethnicity and their poor
956
Mohottige et al
correlation with genetic ancestry; 2) inaccuracies with
self-identified ancestry; 3) the need to integrate socio-
contextual factors like poverty/environmental stress into
studies examining associations between racialized life
experiences, genetics, and disease phenotype73; and 4)
that 90%-95% of genetic variation is found within and
not across populations/races.74,75 Furthermore, as
precision-medicine approaches are pursued in kidney
care (eg, APOL1 screening), ongoing attention is needed
to ensure equity and justice in opportunities for
screening, payment for testing and treatment, and
communication of risks and benefits, as well as to combat
stigmatization and the inappropriate racialization of dis-
ease processes.73,76
Reconciling with Race and Racism on a Path
Toward Kidney Health Equity
The beauty of antiracism is that you don’t have to
pretend to be free of racism to be antiracist. Antiracism
is the commitment to fight racism wherever you find it,
including in yourself. And it’s the only way forward.
Ijoema Oluo77
A bold path forward embracing structural competency, eq-
uity, and antiracism in nephrology should incorporate stra-
tegies that spark new insight and action across the field’s
clinical, research, policy, advocacy, and educational plat-
forms (Box 2). Antiracism initiatives have been adopted by
organizations such as the American Public Health Associa-
tion, which launched a scientifically grounded campaign
against racism under the direction of past president Camara P.
Jones.78 Similar efforts embracing principles of structural
competency, centering at the margins, the avoidance of so-
called “colorblind” approaches, and practicing racial con-
sciousness will help to guide the field as we work to rectify
harms experienced by marginalized individuals.
Structural Competency
Structural competency involves understanding and recog-
nizing how complex social institutions including health
insurance,79 immigration policy,80 and housing
agencies81,82 influence kidney health and health-
influencing behaviors.43 By striving for structural compe-
tency,83 we are better equipped to: 1) address the
numerous contextual factors influencing individual clinical
presentations (eg, the role of neighborhood-level envi-
ronmental toxins on kidney health),84 2) engage in
research and policy advocacy that addresses fundamental
causes of kidney health disparities (ie, poverty, poor access
to care, food insecurity, housing insecurity),85 and 3)
commit to sustained multidisciplinary collaboration and
equitable patient and community engagement.83 This
approach also requires accounting for the multidimen-
sional and intersectional experiences of individuals for
whom forms of marginalization occur across multiple
identity domains (eg, race, gender, sexual orientation).13
AJKD Vol 77 | Iss 6 | June 2021
Mohottige et al
Structural competency in practice may involve integrating
genetic studies of kidney disease with rigorous in-
vestigations of how socially (eg, racially) patterned expo-
sures to kidney-harming environments impact disease
expression.86 It will also require equity-focused policy
reform (eg, the Kidney Allocation System87) that enables
more universal health care coverage mechanisms that
recognize and rectify unequal social drivers of health.88
Centering at the Margins
Centering at the margins is an important part of antiracist
process whereby the perspective of a marginalized group
becomes “the central axis around which discourse oc-
curs.”14 This practice, which may occur in research, in-
dividual clinical interactions, and policy advocacy, ensures
attention to those with the least social advantages and
serves as a hallmark of a truly civilized and advanced so-
ciety. Centering at the margins requires creating nonhier-
archical opportunity for patient and community
engagement, thus yielding invaluable contextual data and
expertise. In research partnerships, for example, opportu-
nities for participation account for resources/equity, the
distribution of power is consistently adjusted toward eq-
uity, interest and value are bidirectional, and decision-
making and content/product ownership are shared.89-91
In an individual clinical interaction, centering at the mar-
gins requires a careful and empathetic attention to our
patients’ stories and lived experiences, including the context
in which their kidney disease manifests and is managed.
Centering at the margins also involves service, advocacy,
and the democratization of data that in ways that are per-
son-/community-centered.92 Examples of incorporating
these principles into research are extensive and include
ACT,93,94 SONG,95 PREPARED,96 RADIANT,97 policy-
changing work engaging undocumented immigrants,98
community-engaged studies involving the Indian Health
Service,99 APOL1,100 and kidney health screening (including
KEEP).101,102 Perspectives describing patient/stakeholder
engagement in PCORI103 and community engagement in
NIDDK research104 shed light on key benefits afforded by
approaches that seek to center at the margins.
Avoiding Colorblind Approaches
Colorblind approaches to race are based in the concept that
race is irrelevant and that treating everyone equally should
equate to equal outcomes/equality.105 Yet, evidence
abounds that these ideals of societal equity are far from
being realized.14,106,107 On the contrary, based in the
pernicious fallacy that race has “disappeared as a factor
shaping the life chances of people in the [United
States],”107 colorblind approaches have often reinforced
and/or masked longstanding health inequalities and
facilitated the rationalization and dismissal of structural
racism105,108 (eg, racist policies impacting housing,
incarceration, and education)43,109 and perpetuation of
health inequities. A contemporary example is found in the
AJKD Vol 77 | Iss 6 | June 2021
colorblind approach to early COVID-19 surveillance that
resulted in the omission of key race/ethnicity data; this
obscured stark inequalities and hindered resource alloca-
tion to marginalized communities disproportionately
burdened by the pandemic.110,111 Rectifying inequity
across racial and ethnic groups and other marginalized
communities (ie, immigrants, sexual and gender minor-
ities) requires understanding the diverse experiences of
individuals in different groups and using the best available
tools to measure differences across groups, which in-
cludes, for example, self-identified race and gender.112
Precisely defining what we mean and measure by race
and other categories, and accurately collecting and
analyzing data across groups, is essential for researchers,
clinicians, and policymakers. By contrast, the absence of
racial/ethnic data would eliminate the ability to rectify
kidney health disparities by obscuring underlying root
causes such as structural forms of racism, impeding
transparency and posing barriers to any meaningful solu-
tions, including the provision of key equity-enhancing
resources. We must avoid falling into the trap of pro-
moting equal treatment after years of creating inequity,
rather than just treatment. These white
supremacy–generated narratives and practices (that often
cite Dr Martin Luther King Jr) are designed to perpetuate
inequities in workforce diversity and patient outcomes
rather than to advance equity and justice.
Practicing Racial Consciousness
Racial consciousness, if operating through a white-
supremacy lens, can be an egregious mechanism to
perpetuate injustices and medical atrocities (eg, racial
profiling in criminalization, forced sterilization abuse).
However, racial consciousness, when operating through a
lens of justice and equity, is fundamental for truly anti-
racist practice, and it recognizes the persistence and
omnipresence of racism in US society and the importance
of dismantling racism. Equity, which stands in contradis-
tinction to equality, requires tailoring support, resources,
and other interventions to address the unique needs of
systematically marginalized groups with the goal of
achieving just outcomes. Striving for this equity-focused
consciousness allows for a more transparent and authentic
investigation of the ways in which racism may operate to
produce differential health outcomes. We should strive to
accurately identify, investigate, and systematically dismantle
mechanisms through which interpersonal and structural
racism contribute to kidney health inequities.
When race and other social variables are used in
research and predictive tools, we should: 1) carefully
define and reassess their dynamic meaning and intent for
use within the sociopolitical and historic context of lived
experience (ie, racism, poverty),113,114 2) be precise in
distinguishing them from genetic factors and ancestry,
3) examine variables in association with a health
outcome and structurally mediated factors (ie, interper-
sonal racism, exposure to harmful environmental toxins,
957

and other adversity) that may become embodied over
time in kidney health,115 and 4) ask the fundamental
question of who is harmed and who is helped by use of
these variables.67 Simply put, we must “move beyond
the misuse of social categories of race and ethnicity as a
proxy for genomic variation”116 in our research, clinical
care, education, and policy initiatives. We must instead
rise to the vital challenge of reckoning with a complex
interplay between biology, genetic ancestry, behavior,
and sociostructural forces like racism,63,114,116 through
which the multifactorial mechanisms of kidney health
inequities can be identified and addressed.
Numerous studies in nephrology exemplify equity-
focused race consciousness, including one evaluating the
role of racial discrimination and perceived racism on
initiation of transplant evaluations, which demonstrated
the potency of these social forces on Black Americans.117
Findings from such investigations compel us to identify
and mitigate personal biases while working to enhance the
trustworthiness of our care and our institutions/organi-
zations. Applied to other domains, an equity-focused,
race-conscious approach allows us to celebrate and value
one another, recognize our unique histories and life ex-
periences, and enhance transparency about equity while
holding ourselves, our organizations, and our leaders
accountable for just outcomes. When we move from
asking first what is wrong with a patient or a given
community (because of their race), but rather what we did
to them or that community (racism), we are starting to
practice race-conscious approaches and antiracism.
Ensuring Equity Now and in the Future
There must exist a paradigm, a practical model for
social change that includes an understanding of ways
to transform consciousness that are linked to efforts to
transform structures.
Bell Hooks118
Progress toward kidney health equity will require that we
undo harmful assumptions that are embedded in our
practice and often implicitly and explicitly influence our
clinical judgments. We should engage in the continuous
reflection and rigorous interrogation of what we have long
believed as truisms, seeking a more complex truth that
recognizes the nuance and historical underpinnings of
how even the most ostensibly “objective” facts have been
influenced by history, culture, politics, and social forces,
including racism. This may require learning new concepts
that have not been traditionally acknowledged in
nephrology but have value to our society and are clearly
aligned with the equity we seek. Equity is not a zero-sum
game; the pursuit of antiracist, justice-focused practices
will yield greater social equity for all of us while better
highlighting intervenable mechanisms resulting in dispa-
rate outcomes, curbing profound social and economic
958
Mohottige et al
costs associated with kidney disease, and addressing a
moral imperative. Current calls to reevaluate the use of
race-based clinical algorithms such as glomerular filtration
rate–estimating equations, which may erroneously esti-
mate clinical risk based on the faulty premise of race as a
biological construct (as race is, by design, a social
construct)22,63-65; calls to reform the harmful patchwork
system of emergency-only dialysis for undocumented
immigrants119; and calls to integrate social medicine
frameworks and structural competency into practice and
policy reforms43,45 are examples of ongoing critical work
on the path to equity. Yet, there is substantially more work
to be done. Antiracism, antioppression, and equity in
practice require humility, empathy, persistence, account-
ability, critical consciousness, and self-education, as well as
a commitment to translate beliefs about equity and justice
into meaningful actions and policy. In nephrology, we can
translate our aspirations for equity into action and out-
comes: we will all be better for not allowing the worst of
our history to be perpetuated.
Article Information
Authors’ Full Names and Academic Degrees: Dinushika
Mohottige, MD, MPH, Clarissa J. Diamantidis MD, MHS, Keith C.
Norris, MD, PhD, and L. Ebony Boulware, MD, MPH.
Authors’ Affiliations: Division of Nephrology, Department of
Medicine (DM, CJD); Center for Community and Population
Health Improvement, Clinical and Translational Science Institute
(DM, CJD, LEB); and Division of General Internal Medicine,
Department of Medicine (CJD, LEB), Duke University School of
Medicine, Durham, NC; and Divisions of Nephrology and General
Internal Medicine and Health Services Research, David Geffen
School of Medicine at UCLA, Los Angeles, CA (KCN).
Address for Correspondence: Dinushiika Mohottige, MD, MPH,
Duke University School of Medicine, 200 N Morris St, 3rd Floor,
Durham, NC 27701. Email: dm26@duke.edu
Support: Dr Mohottige was supported by the Duke Center for
Research to Advance Healthcare Equity (REACH Equity), which is
supported by the National Institute on Minority Health and Health
Disparities under award number U54MD012530.
Financial Disclosure: Dr Diamantidis reports consultancy for United
Health Group. The remaining authors declare that they have no
relevant financial interests.
Peer Review: Received October 1, 2020. Evaluated by 3 external
peer reviewers and a member of the Feature Advisory Board, with
direct editorial input from the Feature Editor and a Deputy Editor.
Accepted in revised form January 15, 2021.
References
1. Dyson ME. Tears We Cannot Stop: A Sermon to White
America. St: Martin’s Press; 2017.
2. Hirsch JS, Ng JH, Ross DW, et al. Acute kidney injury in pa-
tients hospitalized with COVID-19. Kidney Int. 2020;98(1):
209-218.
3. Hooper MW, N
apoles AM, P
erez-Stable EJ. COVID-19 and
racial/ethnic disparities. JAMA. 2020;323(24):2466-2467.
AJKD Vol 77 | Iss 6 | June 2021
Mohottige et al
4. Powe NR, Boulware LE. The uneven distribution of kidney
transplants: getting at the root causes and improving care. Am
J Kidney Dis. 2002;40(4):861-863.
5. Norris K, Nissenson AR. Race, Gender, and socioeconomic
disparities in CKD in the United States. J Am Soc Nephrol.
2008;19(7):1261-1270.
6. Norris KC, Agodoa LY. Unraveling the racial disparities asso-
ciated with kidney disease. Kidney Int. 2005;68(3):914-924.
7. Ayanian JZ, Cleary PD, Keogh JH, Noonan SJ, David-
Kasdan JA, Epstein AM. Physicians’ beliefs about racial differ-
ences in referral for renal transplantation. Am J Kidney Dis.
2004;43(2):350-357.
8. Patzer RE, Amaral S, Wasse H, Volkova N, Kleinbaum D,
McClellan WM. Neighborhood poverty and racial disparities in
kidney transplant waitlisting. J Am Soc Nephrol. 2009;20(6):
1333-1340.
9. Saran R, Robinson B, Abbott KC, et al. US Renal Data System
2019 Annual Data Report: epidemiology of kidney disease in
the United States. Am J Kidney Dis. 2020;75(1)(suppl 1):S1-
S64.
10. Purnell TS, Crews DC. Persistent disparities in preemptive
kidney transplantation. Clin J Am Soc Nephrol. 2019;14(10):
1430-1431.
11. Purnell TS, Bae S, Luo X, et al. National trends in the associ-
ation of race and ethnicity with predialysis nephrology care in
the United States From 2005 to 2015. JAMA Netw Open.
2020;3(8):e2015003-e2015003.
12. Purnell TS, Luo X, Cooper LA, et al. Association of race and
ethnicity with live donor kidney transplantation in the United
States from 1995 to 2014. JAMA. 2018;319(1):49-61.
13. Crenshaw K. Mapping the margins: intersectionality, identity
politics, and violence against women of color. Stanford Law
Rev. 1991;43(6):1241-1299.
14. Ford CL, Airhihenbuwa CO. Critical race theory, race equity,
and public health: toward antiracism praxis. Am J Public
Health. 2010;100(suppl 1):S30-S35.
15. Boulware LE, Cooper LA, Ratner LE, LaVeist TA, Powe NR.
Race and trust in the health care system. Public Health Rep.
2003;118(4):358-365.
16. Corbie-Smith G, Thomas SB, St George DM. Distrust, race,
and research. Arch Intern Med. 2002;162(21):2458-2463.
17. Montagu MFA. Man’s Most Dangerous Myth: The Fallacy of
Race. Columbia University Press; 1942.
18. Harawa NT, Ford CL. The foundation of modern racial cate-
gories and implications for research on black/white disparities
in health. Ethn Dis. 2009;19(2):209-217.
19. Norris KC, Williams SF, Rhee CM, et al. Hemodialysis dispar-
ities in African Americans: the deeply integrated concept of
race in the social fabric of our society. Semin Dial. 2017;30(3):
213-223.
20. Cerde~ na JP, Plaisime MV, Tsai J. From race-based to race-
conscious medicine: how anti-racist uprisings call us to act.
Lancet. 2020;396(10257):1125-1128.
21. Cooper R, David R. The biological concept of race and its
application to public health and epidemiology. J Health Polit
Policy Law. 1986;11(1):97-116.
22. Roberts D. Fatal invention: How Science, Politics, and Big
Business Re-Create Race in the Twenty-First Century. New
Press/ORIM; 2011.
23. Gravlee CC, Sweet E. Race, ethnicity, and racism in medical
anthropology, 1977–2002. Med Anthropol Q. 2008;22(1):27-
51.
24. Collins FS. What we do and don’t know about’race’,’ethnicity’,
genetics and health at the dawn of the genome era. Nat Genet.
2004;36(11)(suppl):S13-S15.
AJKD Vol 77 | Iss 6 | June 2021
25. Gould SJ, Gold SJ. The Mismeasure of Man. WW Norton &
Co; 1996.
26. Jones CP. Levels of racism: a theoretic framework and a gar-
dener’s tale. Am J Public Health. 2000;90(8):1212-1215.
27. Duster T. Race and Reification in Science. Science.
2005;307(5712):1050-1051.
28. Williams DR, Collins C. Racial residential segregation: a
fundamental cause of racial disparities in health. Public Health
Rep. 2001;116(5). 440-416.
29. Darity W Jr, Hamilton D, Paul M, et al. What We Get Wrong
About Closing the Racial Wealth Gap. Samuel DuBois Cook
Center on Social Equity and Insight Center for Community
Economic Development; 2018.
30. James A. Making sense of race and racial classification. Race
Soc. 2001;4(2):235-247.
31. Essed P. Understanding Everyday Racism: An Interdisciplinary
Theory. 2. Sage; 1991.
32. Mendez DD, Hogan VK, Culhane JF. Institutional racism,
neighborhood factors, stress, and preterm birth. Ethn Health.
2014;19(5):479-499.
33. Hammond WP. Psychosocial correlates of medical mistrust
among African American men. Am J Community Psychol.
2010;45(1-2):87-106.
34. Nelson A. Unequal treatment: confronting racial and ethnic
disparities in health care. J Natl Med Assoc. 2002;94(8):666-
668.
35. Tsai J, Ucik L, Baldwin N, Hasslinger C, George P. Race mat-
ters? Examining and rethinking race portrayal in preclinical
medical education. Acad Med. 2016;91(7):916-920.
36. Turbes S, Krebs E, Axtell S. The hidden curriculum in multi-
cultural medical education: the role of case examples. Acad
Med. 2002;77(3):209-216.
37. Feagin JR. The continuing significance of race: antiblack
discrimination in public places. Am Sociol Rev. 1991;56(6):
101-116.
38. Little B. Trump’s ‘Chinese’ virus is part of a long history of
blaming other countries for disease. Time. March. 2020;20.
39. Alang S, McAlpine D, McCreedy E, Hardeman R. Police bru-
tality and Black health: setting the agenda for public health
scholars. Am J Public Health. 2017;107(5):662-665.
40. Hall WJ, Chapman MV, Lee KM, et al. Implicit racial/ethnic bias
among health care professionals and its influence on health
care outcomes: a systematic review. Am J Public Health.
2015;105(12):e60-e76.
41. Haley DF, Edmonds A, Schoenbach VJ, et al. Associations
between county-level voter turnout, county-level felony voter
disenfranchisement, and sexually transmitted infections among
women in the Southern United States. Ann Epidemiol.
2019;29:67-73. e61.
42. Crews DC, Wesson DE. Persistent bias: a threat to diversity
among health care leaders. Clin J Am Soc Nephrol.
2018;13(11):1757-1759.
43. Bailey ZD, Krieger N, Ag
enor M, Graves J, Linos N, Bassett MT.
Structural racism and health inequities in the USA: evidence
and interventions. Lancet. 2017;389(10077):1453-1463.
44. Stonington SD, Holmes SM, Hansen H, et al. Case studies in
social medicine—attending to structural forces in clinical
practice. N Engl J Med. 2018;379(20):1958-1961.
45. Greene JA, Loscalzo J. Putting the patient back togeth-
er—social medicine, network medicine, and the limits of
reductionism. N Engl J Med. 2017;377(25):2493-2499.
46. Crews DC, Charles RF, Evans MK, Zonderman AB, Powe NR.
Poverty, race, and CKD in a racially and socioeconomically
diverse urban population. Am J Kidney Dis. 2010;55(6):992-
1000.
959

47. Acevedo-Garcia D, Lochner KA, Osypuk TL, Subramanian SV.
Future directions in residential segregation and health
research: a multilevel approach. Am J Public Health.
2003;93(2):215-221.
48. Rodriguez RA, Sen S, Mehta K, Moody-Ayers S, Bacchetti P,
O’Hare AM. Geography matters: relationships among urban
residential segregation, dialysis facilities, and patient outcomes.
Ann Intern Med. 2007;146(7):493-501.
49. Gander JC, Zhang X, Plantinga L, et al. Racial disparities in
preemptive referral for kidney transplantation in Georgia. Clin
Transplant. 2018;32(9):e13380.
50. Johns TS, Estrella MM, Crews DC, et al. Neighborhood so-
cioeconomic status, race, and mortality in young adult dialysis
patients. J Am Soc Nephrol. 2014;25(11):2649-2657.
51. Crews DC, Liu Y, Boulware LE. Disparities in the burden,
outcomes, and care of chronic kidney disease. Curr Opin
Nephrol Hypertens. 2014;23(3):298-305.
52. Krieger N, Chen JT, Coull BA, Beckfield J, Kiang MV,
Waterman PD. Jim Crow and premature mortality among the
US Black and White population, 1960-2009: an age-period-
cohort analysis. Epidemiology. 2014;25(4):494-504.
53. Quadagno JS. The Color of Welfare: How Racism Undermined
the War on Poverty. Oxford University Press; 1994.
54. Beyer KMM, Laud PW, Zhou Y, Nattinger AB. Housing
discrimination and racial cancer disparities among the 100
largest US metropolitan areas. Cancer. 2019;125(21):3818-
3827.
55. Hamilton D, Darity WA. The political economy of education,
financial literacy, and the racial wealth gap, Federal Reserve
Bank of St. Louis Review. St. Louis: Federal Reserve Bank of;
2017.
56. Wildeman C, Wang EA. Mass incarceration, public health, and
widening inequality in the USA. Lancet. 2017;389(10077):
1464-1474.
57. Bor J, Venkataramani AS, Williams DR, Tsai AC. Police killings
and their spillover effects on the mental health of black Amer-
icans: a population-based, quasi-experimental study. Lancet.
2018;392(10144):302-310.
58. Kawachi I, Daniels N, Robinson DE. Health disparities by race
and class: why both matter. Health Aff. 2005;24(2):343-352.
59. Hirschman D, Garbes L. Toward an economic sociology of
race. Socio-EconRev. SocArXiV. Preprint posted online
February 26, 2019. https://doi.org/10.31235/osf.io/jrktq
60. Williams DR, Priest N, Anderson NB. Understanding associa-
tions among race, socioeconomic status, and health: patterns
and prospects. Health Psychol. 2016;35(4):407-411.
61. Davis AY. Freedom Is a Constant Struggle: Ferguson,
Palestine, and the Foundations of a Movement. Haymarket
Books; 2016.
62. Tsai J, Ucik L, Baldwin N, Hasslinger C, George P. Race mat-
ters? Examining and rethinking race portrayal in preclinical
medical education. Acad Med. 2016;91(7):916-920.
63. Grubbs V. Precision in GFR reporting: let’s stop playing the
race card. Clin J Am Soc Nephrol. 2020;15(8):1201-1202.
64. Eneanya ND, Yang W, Reese PP. Reconsidering the conse-
quences of using race to estimate kidney function. JAMA.
2019;322(2):113-114.
65. Ahmed S, Nutt CT, Eneanya ND, et al. Examining the potential
impact of race multiplier utilization in estimated glomerular
filtration rate calculation on African-American care outcomes.
J Gen Intern Med. 2021;36(2):464-471.
66. Diao JA, Wu GJ, Taylor HA, et al. Clinical implications of
removing race from estimates of kidney function. JAMA.
2021;325(2):184-186.
960
Mohottige et al
67. Norris KC, Eneanya ND, Boulware LE. Removal of race from
estimates of kidney function: first, do no harm. JAMA.
2021;325:135-137.
68. The fifth report of the Joint National Committee on Detection,
Evaluation, and Treatment of High Blood Pressure (JNC V).
Arch Intern Med. 1993;153(2):154-183.
69. Sankar P, Kahn J. BiDil: race medicine or race marketing?
Health Aff (Millwood). 2005;24(suppl 1). W5-455-W455-464.
70. Hinkson LR. The right profile? An examination of race-based
pharmacological treatment of hypertension. Sociol Race Ethn
(Thousand Oaks). 2015;1(2):255-269.
71. Dorsch MP, Lester CA, Ding Y, Joseph M, Brook RD. Effects of
race on statin prescribing for primary prevention with high
atherosclerotic cardiovascular disease risk in a large healthcare
system. JAMA. 2019;8(22):e014709.
72. Stone NJ, Robinson JG, Lichtenstein AH, et al. 2013 ACC/
AHA guideline on the treatment of blood cholesterol to reduce
atherosclerotic cardiovascular risk in adults: a report of the
American College of Cardiology/American Heart Association
Task Force on Practice Guidelines. J Am Coll Cardiol.
2014;63(25 part B):2889-2934.
73. Via M, Ziv E, Burchard EG. Recent advances of genetic
ancestry testing in biomedical research and direct to consumer
testing. Clin Genet. 2009;76(3):225-235.
74. Braun L. Race, ethnicity, and health: can genetics explain dis-
parities? Perspect Biol Med. 2002;45(2):159-174.
75. Barbujani G, Magagni A, Minch E, Cavalli-Sforza LL. An
apportionment of human DNA diversity. Proc Natl Acade Sci.
1997;94(9):4516-4519.
76. Young BA, Fullerton SM, Wilson JG, et al. Clinical genetic
testing for APOL1: are we there yet? Semin Nephrol.
2017;37(6):552-557.
77. Oluo I. July 14, 2019. Accessed January 15, 2021. https://
twitter.com/ijeomaoluo/status/1150565193832943617
78. Jones CP. Toward the science and practice of anti-racism:
launching a national campaign against racism. Ethn Dis.
2018;28(suppl 1):231.
79. Harhay MN, McKenna RM, Boyle SM, et al. Association be-
tween Medicaid expansion under the Affordable Care Act and
preemptive listings for kidney transplantation. Clin J Am Soc
Nephrol. 2018;13(7):1069-1078.
80. Cervantes L, Tuot D, Raghavan R, et al. Association of
emergency-only vs standard hemodialysis with mortality and
health care use among undocumented immigrants with end-
stage renal disease. JAMA Intern Med. 2018;178(2):188-195.
81. Crews DC, Novick TK. Social determinants of CKD hotspots.
Semin Nephrol. 2019;39(3):256-262.
82. Novick TK, Gadegbeku CA, Crews DC. Dialysis for patients
with end-stage renal disease who are homeless. JAMA Intern
Med. 2018;178(12):1581-1582.
83. Metzl JM, Hansen H. Structural competency: theorizing a new
medical engagement with stigma and inequality. Soc Sci Med.
2014;103:126-133.
84. Stanifer JW, Stapleton HM, Souma T, Wittmer A, Zhao X,
Boulware LE. Perfluorinated chemicals as emerging environ-
mental threats to kidney health: a scoping review. Clin J Am
Soc Nephrol. 2018;13(10):1479-1492.
85. Link BG, Phelan J. Social conditions as fundamental causes of
disease. J Health Soc Behav. 1995:80-94. Spec No.
86. Kuzawa CW, Sweet E. Epigenetics and the embodiment of
race: developmental origins of US racial disparities in cardio-
vascular health. Am J Human Biol. 2009;21(1):2-15.
87. Melanson TA, Hockenberry JM, Plantinga L, et al. New kidney
allocation system associated with increased rates of
AJKD Vol 77 | Iss 6 | June 2021
Mohottige et al
transplants among black and hispanic patients. Health Aff.
2017;36(6):1078-1085.
88. Hall YN. Social Determinants of health: addressing unmet
needs in nephrology. Am J Kidney Dis. 2018;72(4):582-591.
89. Israel BA, Coombe CM, Cheezum RR, et al. Community-based
participatory research: a capacity-building approach for policy
advocacy aimed at eliminating health disparities. Am J Public
Health. 2010;100(11):2094-2102.
90. Cashman SB, Adeky S, Allen AJ 3rd, et al. The power and the
promise: working with communities to analyze data, interpret
findings, and get to outcomes. Am J Public Health.
2008;98(8):1407-1417.
91. Doll KM, Snyder CR, Ford CL. Endometrial cancer disparities: a
race-conscious critique of the literature. Am J Obstet Gynecol.
2018;218(5):474-482. e472.
92. Boulware LE, Lyn M. Who will drive the change? Democratizing
health data. Am J Public Health. 2019;109(4):547-548.
93. Ameling JM, Auguste P, Ephraim PL, et al. Development of a
decision aid to inform patients’ and families’ renal replacement
therapy selection decisions. BMC Med Informatics Decis Mak.
2012;12:140.
94. Ameling JM, Ephraim PL, Bone LR, et al. Adapting hypertension
self-management interventions to enhance their sustained
effectiveness among urban African Americans. Fam Commun
Health. 2014;37(2):119-133.
95. Tong A, Craig JC, Nagler EV, et al. Composing a new song for
trials: the Standardized Outcomes in Nephrology (SONG)
initiative. Nephrol Dial Transplant. 2017;32(12):1963-1966.
96. Ephraim PL, Powe NR, Rabb H, et al. The providing
resources to enhance African American patients’ readiness
to make decisions about kidney disease (PREPARED) study:
protocol of a randomized controlled trial. BMC Nephrol.
2012;13. 135-135.
97. Patzer RE, Paul S, Plantinga L, et al. A randomized trial to
reduce disparities in referral for transplant evaluation. J Am Soc
Nephrol. 2017;28(3):935-942.
98. Cervantes L, Fischer S, Berlinger N, et al. The illness experi-
ence of undocumented immigrants with end-stage renal dis-
ease. JAMA Intern Med. 2017;177(4):529-535.
99. Narva A. Population health for CKD and diabetes: lessons from
the Indian Health Service. Am J Kidney Dis. 2018;71(3):407-
411.
100. Young BA, Blacksher E, Cavanaugh KL, et al. Apolipoprotein
L1 testing in African Americans: involving the community in
policy discussions. Am J Nephrol. 2019;50(4):303-311.
101. Vargas RB, Jones L, Terry C, et al. Community-partnered ap-
proaches to enhance chronic kidney disease awareness, pre-
vention, and early intervention. Adv Chronic Kidney Dis.
2008;15(2):153-161.
102. Umeukeje EM, Wild MG, Maripuri S, et al. Black Americans’
perspectives of barriers and facilitators of community screening
for kidney disease. Clin J Am Soc Nephrol. 2018;13(4):
551-559.
103. Cukor D, Cohen LM, Cope EL, et al. Patient and other stake-
holder engagement in patient-centered outcomes research
institute funded studies of patients with kidney diseases. Clin J
Am Soc Nephrol. 2016;11(9):1703-1712.
104. Kimmel PL, Jefferson N, Norton JM, Star RA. How community
engagement is enhancing NIDDK research. Clin J Am Soc
Nephrol. 2019;14(5):768-770.
105. Neville HA, Gallardo ME, Sue DWE. The Myth of Racial Color
Blindness: Manifestations, Dynamics, and Impact. American
Psychological Association; 2016.
106. Collins PH. Science, critical race theory and colour-blindness.
Br J Sociol. 2015;66(1):46-52.
AJKD Vol 77 | Iss 6 | June 2021
107. Bonilla-Silva E. Racism Without Racists: Color-Blind Racism
and the Persistence of Racial Inequality in the United States.
Rowman & Littlefield Publishers; 2006.
108. Krishnan L, Ogunwole SM, Cooper LA. Historical insights on
coronavirus disease 2019 (COVID-19), the 1918 influenza
pandemic, and racial disparities: illuminating a path forward.
Ann Intern Med. 2020;173(6):474-481.
109. Williams DR, Lawrence JA, Davis BA, Vu C. Understanding how
discrimination can affect health. Health Serv Res.
2019;54(suppl 2):1374-1388.
110. Laurencin CT, McClinton A. The COVID-19 pandemic: a call to
action to identify and address racial and ethnic disparities.
J Racial Ethn Health Disparities. 2020;7(3):1-5.
111. Gross C, Essien U, Pasha S, Gross J, Wang S, Nunez-Smith M.
Racial and ethnic disparities in population level COVID-19
mortality. J Gen Intern Med. 2020;35(10):3097-3099.
112. Mohottige D, Lunn MR. Ensuring Gender-Affirming Care in
Nephrology: Improving Care for Transgender and Gender-
Expansive Individuals. Clin J Am Soc Nephrol. 2020;15(8):
1195-1197.
113. Williams DR, Collins C. Racial residential segregation: a
fundamental cause of racial disparities in health. Public Health
Rep. 2001;116(5):404-416.
114. Boyd RW, Lindo EG, Weeks LD, McLemore M. On racism: a
new standard for publishing on racial health inequities. Health
Affairs Blog. July 2, 2020.
115. Krieger N. Living and dying at the crossroads: racism,
embodiment, and why theory is essential for a public health of
consequence. Am J Public Health. 2016;106(5):832.
116. Bonham VL, Green ED, P
erez-Stable EJ. Examining how race,
ethnicity, and ancestry data are used in biomedical research.
JAMA. 2018;320(15):1533-1534.
117. Hamoda RE, McPherson LJ, Lipford K, et al. Association
of sociocultural factors with initiation of the kidney trans-
plant evaluation process. Am J Transplant. 2020;20(1):
190-203.
118. Hooks Bell. Killing Rage: Ending Racism. Henry Holt and Co.;
1995.
119. Rizzolo K, Novick TK, Cervantes L. Dialysis care for undocu-
mented immigrants with kidney failure in the COVID-19 era:
public health implications and policy recommendations. Am J
Kidney Dis. 2020;76(2):255-257.
120. Canadian Race Relations Foundation. Glossary of terms. 2020.
Accessed December 15, 2020. https://www.crrf-fcrr.ca/en/
resources/glossary-a-terms-en-gb-1/item/22793-anti-racism
121. Ford CL, Airhihenbuwa CO. The public health critical race
methodology: praxis for antiracism research. Soc Sci Med.
2010;71(8):1390-1398.
122. Merriam-Webster.com. Equality. 2011. Accessed Januarry 11,
2021. https://www.merriam-webster.com/dictionary/equality
123. Jones CP. Systems of power, axes of inequity: parallels,
intersections, braiding the strands. Med Care.
2014;52(10)(suppl 3):S71-S75.
124. Braveman PA, Kumanyika S, Fielding J, et al. Health disparities
and health equity: the issue is justice. Am J Public Health.
2011;101(suppl 1):S149-S155.
125. Gee GC, Ford CL. Structural racism and health inequities: old
issues, new directions. Du Bois Rev. 2011;8(1):115-132.
126. Schaefer RT, ed. Encyclopedia of Race, Ethnicity, and Society.
SAGE Publications; 2008.
127. Anti Defamation League. Glossary terms: white supremacy.
2020. Accessed December 14, 2020. https://www.adl.org/
resources/glossary-terms/white-supremacy
128. Tarver-Carr ME, Powe NR, Eberhardt MS, et al. Excess risk of
chronic kidney disease among African-American versus White
961

subjects in the United States: a population-based study of potential
explanatory factors. J Am Soc Nephrol. 2002;13(9):2363-2370.
129. Mehrotra R, Kermah D, Fried L, Adler S, Norris K. Racial dif-
ferences in mortality among those with CKD. J Am Soc
Nephrol. 2008;19(7):1403-1410.
130. Kinchen KS, Sadler J, Fink N, et al. The timing of specialist
evaluation in chronic kidney disease and mortality. Ann Intern
Med. 2002;137(6):479-486.
131. Navaneethan SD, Aloudat S, Singh S. A systematic review
of patient and health system characteristics associated
with late referral in chronic kidney disease. BMC Nephrol.
2008;9:3.
132. Mujahid MS, Diez Roux AV, Cooper RC, Shea S, Williams DR.
Neighborhood stressors and race/ethnic differences in hyper-
tension prevalence (the Multi-Ethnic Study of Atherosclerosis).
Am J Hypertens. 2011;24(2):187-193.
133. Evans K, Coresh J, Bash LD, et al. Race differences in
access to health care and disparities in incident chronic
kidney disease in the US. Nephrol Dial Transplant. 2011;26(3):
899-908.
134. Prakash S, Rodriguez RA, Austin PC, et al. Racial composition
of residential areas associates with access to pre-ESRD
nephrology care. J Am Soc Nephrol. 2010;21(7):1192-1199.
135. Kucirka LM, Grams ME, Lessler J, et al. Association of race and
age with survival among patients undergoing dialysis. JAMA.
2011;306(6):620-626.
136. Mehrotra R, Soohoo M, Rivara MB, et al. Racial and ethnic
disparities in use of and outcomes with home dialysis in the
United States. J Am Soc Nephrol. 2016;27(7):2123-2134.
137. Rodriguez RA, Hotchkiss JR, O’Hare AM. Geographic infor-
mation systems and chronic kidney disease: racial disparities,
rural residence and forecasting. J Nephrol. 2013;26(1):3-15.
138. Volkova N, McClellan W, Klein M, et al. Neighborhood poverty
and racial differences in ESRD incidence. J Am Soc Nephrol.
2008;19(2):356-364.
962
Mohottige et al
139. Hall YN, Xu P, Chertow GM, Himmelfarb J. Characteristics and
performance of minority-serving dialysis facilities. Health Serv
Res. 2014;49(3):971-991.
140. King KL, Husain SA, Jin Z, Brennan C, Mohan S. Trends in
disparities in preemptive kidney transplantation in the United
States. Clin J Am Soc Nephrol. 2019;14(10):1500-1511.
141. Purnell TS, Luo X, Cooper LA, et al. Association of race and
ethnicity with live donor kidney transplantation in the United
States from 1995 to 2014. JAMA. 2018;319(1):49-61.
142. Patzer RE, McPherson L, Wang Z, et al. Dialysis facility referral
and start of evaluation for kidney transplantation among pa-
tients treated with dialysis in the Southeastern United States.
Am J Transplant. 2020;20(8):2113-2125.
143. Ng YH, Pankratz VS, Leyva Y, et al. Does racial disparity in
kidney transplant waitlisting persist after accounting for social
determinants of health? Transplantation. 2020;104(7):1445-
1455.
144. Myaskovsky L, Almario Doebler D, Posluszny DM, et al.
Perceived discrimination predicts longer time to be accepted
for kidney transplant. Transplantation. 2012;93(4):423-429.
145. Kim JJ, Basu M, Plantinga L, et al. Awareness of racial
disparities in kidney transplantation among health care
providers in dialysis facilities. Clin J Am Soc Nephrol.
2018;13(5):772-781.
146. Saunders M, Cagney K, Ross L, Alexander GC. Neighborhood
poverty, racial composition and renal transplant waitlist. Am J
Transplant. 2010;10(8):1912-1917.
147. Plantinga L, Pastan S, Kramer M, McClellan A, Krisher J,
Patzer RE. Association of U.S. dialysis facility neighborhood
characteristics with facility-level kidney transplantation. Am J
Nephrol. 2014;40(2):164-173.
148. Ladin K, Emerson J, Butt Z, et al. How important is social
support in determining patients’ suitability for transplantation?
Results from a national survey of transplant clinicians. J Med
Ethics. 2018;44(10):666-674.
AJKD Vol 77 | Iss 6 | June 2021
Policy Forum Editorial
Reducing the Shortage of Transplant
Kidneys: A Lost Opportunity for the
US Health Resources and Services
Administration (HRSA)
Frank McCormick, Philip J. Held, Glenn M. Chertow, Thomas G. Peters, and John
P. Roberts
Note from Editors: Given the importance and controversial
nature of this topic, a diversity of viewpoints is critical. For this
reason, AJKD solicited a commentary on this editorial from
Danovitch, Capron, and Delmonico, whose counterpoint appears
in the accompanying Policy Forum Commentary (p. 967).
During the past 3 decades in the United
States, tens of thousands of patients with kidney
failure have died while awaiting transplantation
owing to a shortage of transplant kidneys. The
President’s July 2019 “Executive Order on
Advancing American Kidney Health” included
2 sections specifically aimed at increasing kid-
ney transplantation.1 Section 7 directed the
Secretary of Health and Human Services (HHS)
to propose regulations to increase the recovery
of deceased donor organs and reduce discards,
and section 8 directed the Secretary to reduce
financial barriers to living organ donation.
The 2 agencies of HHS that are responsible for
implementing these life-saving sections respon-
ded in very different ways (Box 1). The Centers
for Medicare and Medicaid Services (CMS) made
a serious attempt to comply with both the letter
and the spirit of section 7. However, the Health
Resources and Services Administration (HRSA)
made only a perfunctory response to the much
more important section 8. In its recently issued
final rule, HRSA did not go beyond the modest
changes suggested in the executive order, which
we estimate will increase kidney donations in the
neighborhood of only 2,000 per year. If HRSA
had taken a more expansive view of its man-
date—as CMS did—and modified its rule as we
suggest below, we estimate it could have
increased kidney donations in the neighborhood
of 15,000 per year. This means about 13,000
additional kidney failure patients a year could
have lived significantly longer and healthier lives.
Deceased Donors
In December 2019, CMS issued a detailed
report regarding section 7 that proposed
replacing 3 existing metrics for organ pro-
curement organization (OPO) performance in
AJKD Vol 77 | Iss 6 | June 2021
recovering the organs of deceased donors with
FEATURE EDITOR:
2 new measures: donation rate and organ Daniel E. Weiner
transplantation rate.2 Every 4 years, OPOs
would have to meet CMS requirements for ADVISORY BOARD:
both rates. Indeed, CMS proposed that all L. Ebony Boulware
OPOs perform at or above the current 75th Kevin Erickson
percentile. These new rules will apply much- Eduardo Lacson Jr
Bruce M. Robinson
needed regulatory pressure on OPOs to in- Wolfgang Winkel-
crease the number of organs recovered.3 mayer
However, the rules represent a challenge to
existing OPOs, and were met with extensive Policy Forum high-
concern by much of the transplant lights aspects of
community. nephrology relating to
payment and social
policy, legislation,
Financial Barriers to Living Donors regulation, de-
In contrast to the detailed CMS proposal on mographics, politics,
recovering organs from deceased donors, HRSA and ethics, contextu-
made only a perfunctory response to section 8 alizing these issues
as they relate to the
of the executive order concerning living do- lives and practices of
nors. In September 2020, HRSA issued a final members of the kid-
rule—”Removing Financial Disincentives to ney community,
Living Organ Donation”—that laid out the including providers,
additional donor expenses that the govern- payers, and patients.
ment will reimburse (beyond the travel and
lodging expenses that are already reim-
bursed).4 The list did not go beyond those
expenses specified in the original executive
order—lost wages, child care, and elder care.
Also in September, HRSA issued a final
notice, “Reimbursement of Travel and Subsis-
tence Expenses Toward Living Organ Donation
Program Eligibility Guidelines,” in which it
proposed increasing the ceiling on income that
donors and recipients can earn and still be
eligible to have the donor’s expenses reim-
bursed.5 HRSA raised the eligibility threshold
from 300% of the HHS poverty guidelines
to 350%. The latter number is far below the
500% recommended by both the National
Living Donor Assistance Center and the Advisory
Committee on Organ Transplantation (which
provides recommendations to the Secretary
of HHS on organ transplantation).
Taken together, we estimate these 2 re-
sponses of HRSA will only increase kidney
963

donations in the neighborhood of 2,000 per year (an in-
crease of only 9% of the 23,401 donations in 2019). This
result falls far short of the goal set by HHS Secretary Alex
Azar, who stated6: “On living donations, we’re going to
dramatically expand support for living kidney donors, so
that Americans who wish to be generous living donors
don’t face unnecessary financial barriers to doing so” and
“When an American wishes to become a living donor, we
don’t believe their financial situation should limit their
generosity.”
Box 1. Summary of the HRSA Response to Section 8 of the
Executive Order on Advancing American Kidney Health
The problem:
Tens of thousands of kidney failure patients are not receiving
kidney transplants owing to an insufficient number of available
organs
The solution proposed in section 8 of the Executive
Order:
Substantially expand government reimbursement of the ex-
penses of living kidney donors to increase kidney donations
and transplants
Specific directives in section 8:
• Include lost wages and dependent care in reimbursed
expenses
> HRSA’s response: Included only the expenses specified;
did not broaden to include additional expense categories
• Raise income limit on donors who are eligible to have ex-
penses reimbursed
> HRSA’s response: Raised income limit on eligibility to
only 350% of poverty guidelines (from 300%)
Projected result of HRSA’s response:
Living kidney donors would increase by ~2,000 per year
Our recommendations to HRSA:
• Additional disincentives that should be offset by the federal
government
> The very low risk of dying from the nephrectomy
> Pain and discomfort of the nephrectomy
> Long-term health consequences of kidney removal
> Concern that a relative or close friend may need a kidney
in the future
• Income limit on eligibility to have donor expenses reimbursed
should be further raised
> Raise income limit to at least 500% of the poverty
guidelines
Expected results with our recommendations:
• Living kidney donors could increase by ~15,000 per year
• Quality of life and longevity could be improved for ~15,000
patients per year with kidney failure
• Government would save money because of reduced
spending on dialysis therapy
964
McCormick et al
Thus, HRSA seems to have missed an opportunity to
markedly increase the supply of donor kidneys. About
13,000 kidney failure patients a year could have enjoyed
significantly longer and healthier lives if HRSA had
modified 2 key parts of its initial (December 2019) pro-
posal as we suggest below.
Removing Disincentives Would Save the
Government Money
In the “Summary of Impacts” section of its initial proposal,
HRSA stated: “While expanding the list of expenses
eligible for reimbursement for living organ donors will
increase the average amount of reimbursement, the federal
government can expect to save overall due to an increase in
additional organ transplants performed and the aversion of
dialysis.” We strongly support this view. The key point is
the government is already committed to paying for dialysis
therapy for most patients with kidney failure, and that
therapy is much more expensive than the cost of a trans-
plant.7 Therefore, if the government spends a relatively
small amount to remove financial barriers to kidney
donation, and that causes more transplants to occur, then
the government will spend less money on very expensive
dialysis therapy, so it will save money in the long run.
Although it is clear from the quote above that HRSA
well understands this reasoning, it surprisingly failed to
apply it when it raised the income limit on the donors who
are eligible for government reimbursement to only 350%
of the poverty guidelines. If it had followed its own logic,
it would have concluded that, from the viewpoint of the
government budget, the limit should be raised to 500% or
even higher (assuming wage reimbursement is capped at
$10,000, as advocated in our previous work8). This is
because the savings from stopping dialysis are so large that
they would exceed the cost of a transplant, even including
the slightly larger reimbursement costs of affluent donors.
(For the arithmetic supporting this statement, see the left-
hand column of Table 3 in Held et al.7) So the more
donors—rich or poor—whose expenses are reimbursed by
the government, the more money the government would
save in the long run.
Additional Barriers to Organ Donation
In the discussion section of its initial proposal, HRSA
invited comments on additional financial barriers to organ
donation that it might remove. In our earlier empirical
research, we detailed 7 disincentives facing living kidney
donors, the magnitude of which totaled almost $38,0008:
(1) cost of travel to, and lodging at, a transplant center; (2)
loss of income while recovering from surgery; (3) cost of
dependent care while recovering from surgery; (4) risk of
dying during surgery; (5) pain and discomfort of surgery;
(6) decrease in the long-term quality of life due to kidney
removal; (7) concern that a relative or friend might need
the donor’s kidney in the future.
HRSA has already removed the first disincentive, but
only for low-income donors and recipients, by
AJKD Vol 77 | Iss 6 | June 2021
McCormick et al
reimbursing donor expenses through the National Living
Donor Assistance Center. Now, in response to section 8 of
the President’s executive order, HRSA will do the same for
the second and third disincentives.
In our previous analysis,8 we suggested HRSA also
remove disincentives 4 through 7, and reimburse donors
of all income levels. Specifically, to remove disincentives 4
and 6, HRSA should provide an insurance policy covering
death, disability, and long-term health problems due to
donation. To remove disincentive 5, HRSA should give
donors a tax credit of $6,500. And to remove disincentive
7, HRSA should promise to provide a kidney for a specific
person in the future (ie, give them a voucher) in return for
the donor donating a kidney today.
If the government removes all of these disincentives, it
would not only be a major step toward economic fairness,
but it would also significantly increase the number of
living donors. The very limited evidence available on the
latter effect suggests that if the government removes all 7
disincentives, with a total magnitude of almost $38,000,
the number of kidney donations from living donors would
increase in the neighborhood of 15,000 per year. (It
should be emphasized there is considerable uncertainty
surrounding this estimate, since it is based on only a small
number of historical examples.) This would raise total
kidney donations (from both living and deceased donors)
from about 23,500 in 2019 to about 38,500 per year. In
contrast, we estimate HRSA’s final rule would increase
kidney donations in the neighborhood of only 2,000 per
year, raising total transplants to only about 25,500 per
year.8
Ethical Considerations
The primary ethical argument in favor of having the
government remove all 7 disincentives and substantially
raising the income limit on donor eligibility for reim-
bursement is that it would increase kidney donations from
living donors, enabling tens of thousands of kidney failure
patients to obtain transplants and live longer and healthier
lives. A second ethical consideration is economic fairness.
Donors should not be penalized for their generosity. They
should not suffer any economic loss because they have
donated a kidney; rather, they should be made whole
again.
However, this is not the dominant view of the trans-
plant community. That view is expressed in the Declaration
of Istanbul, which has been endorsed by more than 100
countries.9 The declaration states: “Organ donation should
be a financially neutral act.” Although the declaration does
not define financial neutrality, a group of experts have
undertaken that task.10 Their definition specifically ex-
cludes “pain and suffering” and “household chores” as
reimbursable expenses. Thus, if the US government were
to follow the advice of the declaration, it would not
remove disincentive 5 (pain and discomfort of nephrec-
tomy) and part of disincentive 3 (dependent care), but it
would remove the other 5 disincentives. Although doing
AJKD Vol 77 | Iss 6 | June 2021
so might not increase the number of kidney donations by
15,000 (the gain we estimate resulting from removing all
7 disincentives) it would still be much closer to that
number than the gain of 2,000 we estimate would result
from HRSA’s final rule.
Article Information
Authors’ Full Names and Academic Degrees: Frank McCormick,
PhD, Philip J. Held, PhD, Glenn M. Chertow, MD, MPH, Thomas
G. Peters, MD, and John P. Roberts, MD.
Authors’ Affiliations: Division of Nephrology, Stanford University
School of Medicine, Palo Alto, CA (PJH, GMC); Department of
Surgery, University of Florida, Jacksonville, FL (TGP); and
Department of Surgery, University of California San Francisco, San
Francisco, CA (JPR). Dr McCormick is retired Director of US
Economic and Financial Research at the Bank of America.
Address for Correspondence: Frank McCormick, PhD, 506
Monarch Ridge Dr, Walnut Creek, CA 94597. Email: fmccorm@
astound.net
Authors’ Contributions: FM and PJH contributed equally to this
editorial.
Support: There was no specific support for this editorial.
Financial Disclosure: Dr Chertow reports personal fees from
Satellite Healthcare, personal fees from Akebia, personal fees from
Amgen, personal fees and other from Ardelyx, personal fees from
Astra Zeneca, personal fees from Baxter, other from CloudCath,
personal fees from Cricket Health, personal fees from DiaMedica,
other from Durect, other from DxNow, personal fees from Gilead,
other from Outset, personal fees from Reata, personal fees from
Sanifit, personal fees from Vertex, personal fees from Angion,
personal fees from Bayer, and personal fees from ReCor. Dr
Roberts receives research support from Merck. The remaining
authors declare that they have no relevant financial interests.
Acknowledgements: The authors thank Robert Heller, PhD; Philip
Simon, JD; and Joshua Morrison, JD, for providing important
reviews and insight. The authors of this editorial are solely
responsible for any errors. The data and statistics reported here
have been supplied by the Scientific Registry of Transplant
Recipients and the US Renal Data System.
Disclaimer: The interpretation and reporting of these data are the
responsibility of the authors and in no way should be seen as an
official policy or interpretation of the US government.
Peer Review: Received June 29, 2020. Evaluated by 3 external peer
reviewers, with direct editorial input from the Feature Editor and a
Deputy Editor. Accepted in revised form October 13, 2020.
Publication Information: © 2020 by the National Kidney Founda-
tion, Inc. Published online November 30, 2020 with doi 10.1053/
j.ajkd.2020.10.007
References
1. Trump D. Executive Order on Advancing American Kidney
Health. July 10, 2019. Accessed June 12, 2020. https://www.
whitehouse.gov/presidential-actions/executive-order-advancing-
american-kidney-health/
2. Centers for Medicare and Medicaid Services. Revisions to
the outcome measure requirements for organ procurement
organization. December 2019. Accessed June 12, 2020. https://
www.federalregister.gov/documents/2019/12/23/2019-27418/
medicare-and-medicaid-programs-organ-procurement-organizations-
conditions-for-coverage-revisions-to
3. Snyder JJ, Musgrove D, Zaun D, et al. The Centers for
Medicare and Medicaid Services’ proposed metrics for
965

recertification of organ procurement organizations: evaluation
by the Scientific Registry of Transplant Recipients. Am J
Transplant. 2020;20(9):2466-2480.
4. Health Resources Services Administration. Removing
financial disincentives to living organ donation. September
2020. Accessed June 12, 2020. https://www.federalregister.
gov/documents/2020/09/22/2020-20804/removing-financial-
disincentives-to-living-organ-donation
5. Health Resources Services Administration. Reimbursement
of travel and subsistence expenses toward living organ
donation program eligibility guidelines. September 2020.
Accessed June 12, 2020. https://www.federalregister.gov/
documents/2020/09/22/2020-20805/reimbursement-of-travel-
and-subsistence-expenses-toward-living-organ-donation-
program-eligibility
6. Azar A. Trump administration proposes new rules to
increase accountability and availability of the organ supply. Press
966
McCormick et al
release of Department of Health and Human Services. December
17, 2019. Accessed June 12, 2020. https://www.hhs.gov/about/
news/2019/12/17/trump-administration-proposes-new-rules-
increase-accountability-availability-organ-supply.html
7. Held PJ, McCormick F, Ojo A, Roberts JP. A cost-benefit
analysis of government compensation of kidney donors. Am J
Transplant. 2016;16(3):877-885.
8. McCormick F, Held PJ, Chertow GM, Peters TG, Roberts JP.
Removing disincentives to kidney donation: a quantitative
analysis. J Am Soc Nephrol. 2019;30(8):1340-1357.
9. The Declaration of Istanbul on Organ Trafficking and Transplant
Tourism (2018 Edition). Accessed June 12, 2020. http://www.
declarationofistanbul.org/images/Policy_Documents/2018_Ed_
Do/2018_Edition_of_the_Declaration_of_Istanbul_Final.pdf
10. Hays R, Rodrigue JR, Cohen D, et al. Financial neutrality for
living organ donors: reasoning, rationale, definitions, and imple-
mentation strategies. Am J Transplant. 2016;16(7):1973-1981.
AJKD Vol 77 | Iss 6 | June 2021
Policy Forum Commentary
The True Meaning of Financial
Neutrality in Organ Donation
Gabriel M. Danovitch, Alexander M. Capron, and Francis L. Delmonico
Note from Editors: Given the importance and controversial
nature of this topic, a diversity of viewpoints is critical. For this
reason, AJKD solicited this commentary on a Policy Forum
Editorial from McCormick et al (p. 963).
With determination to stop exploitation of
vulnerable organ donors, the international
transplant community, including the United
States, reached a consensus regarding the
principle of financial neutrality as a necessity
for ethical organ donation.1-3 Financial
neutrality means that “donors and their fam-
ilies neither lose nor gain financially as a result
of donation.”1 Although McCormick et al
discuss this principle in the accompanying
Policy Forum Editorial,4 it is our opinion that
they distort the meaning of financial
neutrality.5 The authors expand the legitimate
costs of donation, including travel, lodging,
lost work, and other verifiable expenses, to
reach a $38,000 fixed payment to donors by
including a dollar value for wholly subjective
factors such as pain, fear, risk, and quality of
life. For example, “risk” is assigned a value of
$6,500. These factors are intrinsic to the
process of organ donation, and it is disin-
genuous to include them under the rubric of
financial neutrality.
The full explanation of how McCormick
et al derive the $38,000 sum and our response
to it have been published elsewhere.6,7,8 Their
proposal may be dangerous, particularly in
countries with rampant organ markets, where
those who profit from this trade will use the
position taken by McCormick and colleagues
to justify legalizing payments to donors. If
$6,500 is the value that risk is assigned in the
United States, this sum could surely be
doubled or tripled elsewhere, with a resulting
international auction of living donor kidneys.
Untethering payment to donors from their
actual out-of-pocket costs inherently makes
providing an organ more attractive to those in
poverty. This not only exploits these donors
but also risks harm to recipients. The temp-
tation for those in financial distress to obfus-
cate relevant medical information is not a
theoretical concern. Poor psychosocial out-
comes and increased risks of infections and
other complications have been observed
AJKD Vol 77 | Iss 6 | June 2021
repeatedly when the motive for donation be- FEATURE EDITOR:
comes monetary.9 Daniel E. Weiner
We clearly must do more to encourage both
living and deceased donation. In September ADVISORY BOARD:
2020, the Health Resources and Services L. Ebony Boulware
Kevin Erickson
Administration (HRSA) announced an updated
Eduardo Lacson Jr
rule, expanding the scope of qualified reim- Bruce M. Robinson
bursable expenses incurred by living organ Wolfgang Winkel-
donors to include lost wages as well as child- mayer
care and elder-care expenses.10 Regrettably,
the updated rule continued to cap reimburse- Policy Forum high-
lights aspects of
ment at $6,000, unchanged from the previ-
nephrology relating to
ously existing program that covered only travel, payment and social
lodging, meals, and incidental expenses policy, legislation,
incurred in donating an organ. Nonetheless, regulation, de-
elements of this update represent important mographics, politics,
and ethics, contextu-
steps toward making donation financially
alizing these issues
neutral for living donors. Critically, HRSA as they relate to the
explicitly rejects the suggestion that payment lives and practices of
be made for “undertaking a ‘risk,’ whether it be members of the kid-
a long-term health risk or surgical risk.”10 ney community,
including providers,
More donors will save more lives, and
payers, and patients.
ensuring financial neutrality is the right way to
treat living kidney donors. However, although
the introduction of financial incentives, covert
or otherwise, seems tempting, it is a “Trojan
horse”11 that will undermine the very fabric
of the evaluation and care to which living
kidney donors are entitled.
Article Information
Authors’ Full Names and Academic Degrees:
Gabriel M. Danovitch, MD, Alexander M. Capron,
LLB, and Francis L. Delmonico, MD.
Authors’ Affiliations: Department of Medicine,
University of California, Los Angeles, Los Angeles,
CA (GMD); Pacific Center for Health Policy and
Ethics, University of Southern California, Los
Angeles, CA (AMC); and Harvard University,
Cambridge, MA (FLD).
Address for Correspondence: Gabriel M. Danovitch,
MD, Department of Medicine, University of California,
Los Angeles, 10833 Le Conte Ave, Los Angeles, CA
90095. Email: gdanovitch@mednet.ucla.edu
Support: None.
Financial Disclosure: The authors declare that they
have no relevant financial interests.
Other Disclosures: Mr Capron is an Honorary
Member of the Declaration of Istanbul Custodian
Group (DICG) Board of Councillors. Drs Danovitch
967

and Delmonico previously held DICG roles, as Co-Chair and
Executive Secretary, respectively.
Peer Review: Received October 30, 2020, in response to an
invitation from the journal. Accepted November 5, 2020, after
editorial review by the Feature Editor and a Deputy Editor.
Publication Information: © 2020 by the National Kidney Founda-
tion, Inc. Published online November 30, 2020 with doi 10.1053/
j.ajkd.2020.11.006
References
1. The Declaration of Istanbul on Organ Trafficking and Transplant
Tourism (2018 Edition). Accessed October 29, 2020. https://
www.declarationofistanbul.org/images/Policy_Documents/2018_
Ed_Do/2018_Edition_of_the_Declaration_of_Istanbul_Final.pdf
2. Delmonico FL, Martin D, Domínguez-Gil B, et al. Living and
deceased organ donation should be financially neutral acts. Am
J Transplant. 2015;15(5):1187-1191.
3. Hays R, Rodrigue JR, Cohen D, et al. Financial neutrality for
living organ donors: reasoning, rationale, definitions, and
implementation strategies. Am J Transplant. 2016;16(7):
1973-1981.
4. McCormick F, Held PJ, Chertow GM, Peters TG, Roberts JP.
Reducing the shortage of transplant kidneys: a lost opportunity
968
Danovitch et al
for the US Health Resources and Services Administration
(HRSA). Am J Kidney Dis. 2021;77(6):963-966.
5. Held PJ, McCormick F, Chertow GM, Peters TG, Roberts JP.
Would government compensation of living kidney donors
exploit the poor? An empirical analysis. PLoS One.
2018;13(11):e0205655.
6. McCormick F, Held PJ, Chertow GM, Peters TG, Roberts JP.
Removing disincentives to kidney donation: a quantitative
analysis. J Am Soc Nephrol. 2019;30(8):1349-1357.
7. Capron AM, Delmonico FL, Danovitch GM. Financial
neutrality in organ donation. J Am Soc Nephrol. 2020;31(1):
229-230.
8. Held PJ, McCormick F, Ojo A, Roberts JP. A cost-benefit
analysis of government compensation of kidney donors. Am J
Transplant. 2016;16(3):877-885.
9. Danovitch GM. The high cost of organ transplant commer-
cialism. Kidney Int. 2014;85(2):248-250.
10. U.S. Department of Health and Human Services, Health Re-
sources and Services Administration. Removing financial dis-
incentives to living organ donation. Fed Regist. 2020;85:
59438-59445.
11. Danovitch GM, Leichtman AB. Kidney vending: the "Trojan
horse" of organ transplantation. Clin J Am Soc Nephrol.
2006;1(6):1133-1135.
AJKD Vol 77 | Iss 6 | June 2021
Core Curriculum in Nephrology

Reducing Kidney Function Decline in

Patients With CKD: Core Curriculum 2021
Teresa K. Chen, Christopher J. Sperati, Sumeska Thavarajah, and Morgan E. Grams

An estimated 8% to 16% of the world’s population has chronic kidney disease, defined by low glo- Complete author and article
merular filtration rate or albuminuria. Progression of chronic kidney disease is associated with adverse information provided at end
of article.
outcomes, including incident kidney failure with replacement therapy, accelerated cardiovascular
disease, disability, and mortality. Therefore, slowing kidney function decline is paramount in the man- Am J Kidney Dis.
agement of a patient with chronic kidney disease. Ascertaining the cause of kidney disease is an 77(6):969-983. Published
online April 21, 2021.
important first step and may compel specific therapies. Effective approaches that apply to the vast
majority of patients with chronic kidney disease include the optimization of blood pressure and doi: 10.1053/
blockade of the renin-angiotensin-aldosterone system, particularly if albuminuria is present. Recent j.ajkd.2020.12.022
studies suggest that sodium/glucose cotransporter 2 inhibitors are highly effective treatments in © 2021 by the National
patients with diabetes and/or albuminuria. For patients with type 2 diabetes, glycemic control is Kidney Foundation, Inc.
important in preventing the development of microvascular complications, and glucagon-like peptide 1
receptor agonists may help reduce albuminuria levels. Other strategies include correcting metabolic
acidosis, maintaining ideal body weight, following diets that are low in sodium and animal protein, and
avoiding potential nephrotoxins such as nonsteroidal anti-inflammatories, proton-pump inhibitors, and
iodinated contrast.

Introduction ➢ Saran R, Robinson B, Abbott KC, et al. US Renal

Chronic kidney disease (CKD) affects more than
697 million individuals worldwide and is asso-
ciated with increased morbidity and mortality. In
2017, 1.2 million deaths and 35.8 million
disability-adjusted life-years were attributed to
CKD. Among Medicare beneficiaries in the Uni-
ted States, annual spending for kidney failure
with replacement therapy (KFRT) and earlier
stages of CKD exceeded $120 billion. Different
causes of kidney disease may require specific
treatments such as immunosuppressive therapy.
However, some strategies to delay the pro-
gression of CKD to KFRT are applicable to most
patients. Early detection and treatment to slow
kidney function decline are paramount to
improving outcomes in patients with CKD.
Hallmarks of CKD management include control
of hypertension and hyperglycemia, inhibition
of the renin-angiotensin-aldosterone system
(RAAS), correction of metabolic acidosis, lifestyle
modification, and avoidance of nephrotoxins.
Two new classes of medications, sodium/glu-
cose cotransporter 2 (SGLT2) inhibitors and
glucagon-like peptide 1 (GLP-1) receptor ago-
nists, also improve kidney outcomes among
individuals with diabetes and/or albuminuria.
Additional Readings
➢ GBD Chronic Kidney Disease Collaboration. Global,
regional, and national burden of chronic kidney
disease, 1990-2017: a systematic analysis for the
Global Burden of Disease Study 2017. Lancet.
2020;395:709-733.
FEATURE EDITOR:
Data System 2019 annual data report: epidemi-
Asghar Rastegar
ology of kidney disease in the United States. Am J
Kidney Dis. 2020;75(1)(suppl 1):S1-S64.
ADVISORY BOARD:
Ursula C. Brewster
Blood Pressure Control Michael Choi
Ann O’Hare
Case 1: A 60-year-old man with CKD glomerular Biff F. Palmer
filtration rate category 3b (G3b) and albuminuria
category 2 (A2, corresponding to an urinary The Core Curriculum
albumin-creatinine ratio [UACR] of 30-300 mg/g), aims to give trainees
hypertension, and stable angina returns for a in nephrology a
follow-up visit. His estimated glomerular filtration strong knowledge
rate (eGFR) has declined from 57 to 44 mL/min/ base in core topics in
1.73 m2 over the past 13 years. His blood pres- the specialty by pro-
viding an overview of
sure (BP) averages 135/72 mm Hg on a regimen
the topic and citing
of valsartan at 320 mg daily, amlodipine at 5 mg key references,
daily, and indapamide at 1.25 mg daily. including the founda-
tional literature that
Question 1: Based on the results of led to current clinical
SPRINT, which one of the following state- approaches.
ments is most accurate regarding a systolic
BP goal of <120 versus <140 mm Hg?
a) All-cause mortality is reduced
b) CKD progresses more slowly at the lower
BP goal
c) Incidence of KFRT is higher at the lower BP goal
d) Incidence of kidney transplantation is lower
at the lower BP goal
Question 2: Which one of the following
patients would be most appropriate for a
lower BP goal to help slow the pro-
gression of CKD?
a) CKD G3aA1 with UACR of 10 mg/g
b) CKD G4A1 with critical bilateral renal artery
stenosis

AJKD Vol 77 | Iss 6 | June 2021 969

 Chen et al

c) CKD G3bA3 with UACR of 3,000 mg/g

d) CKD G3bA3 with UACR of 1,200 mg/g and
history of repeated falls
For the answers to the questions, see the fol-
lowing text.
The AHA/ACC recommend a goal BP <130/80 mm Hg
for all patients with CKD, whereas the KDIGO guidelines
recommend a target of ≤140/90 mm Hg when the UACR
is <30 mg/d and ≤130/80 mm Hg when the UACR
is ≥30 mg/d (Table 1). The KDIGO recommendations are
based, in part, on 2 landmark randomized controlled trials.
The AASK trial randomized participants without diabetes to
a mean arterial pressure (MAP) goal of ≤92 versus 102-
107 mm Hg. Although there was no difference in the rate
of eGFR decline or a composite clinical outcome (eGFR
decline, KFRT, or death) overall, participants with a
baseline urinary protein-creatinine ratio of >0.22 g/g
were 27% less likely to develop a doubling of serum cre-
atinine, KFRT, or death when randomized to intensive
versus standard BP control in the extended cohort phase.
The MDRD Study randomized participants to a MAP goal of
92 versus 107 mm Hg. Again, there were no differences
overall, but participants with proteinuria of ≥3 g/d had
lesser GFR decline in the intensive BP control group. These
and other trials of BP control are summarized in Table 2.
More recently, SPRINT randomized adults without
diabetes but at increased risk for cardiovascular events to a
systolic BP < 120 mm Hg versus < 140 mm Hg. Intensive
BP control was associated with a lower risk of myocardial
infarction, acute coronary syndrome, stroke, heart failure,
and cardiovascular death (hazard ratio [HR], 0.75 [95%
CI, 0.64-0.89]) and all-cause mortality (HR, 0.73 [95%
CI, 0.60-0.90]). The results were consistent among par-
ticipants with baseline CKD (n = 2,646). Intensive BP
control did not prevent adverse kidney outcomes (≥50%
eGFR decline or KFRT). Among participants without
baseline CKD (n = 6,677), intensive BP control resulted in
a 3.5-fold higher risk of ≥30% reduction in eGFR
to < 60 mL/min/1.73 m2, a finding that may reflect
hemodynamic changes rather than true kidney injury.
For Question 1, (a) reduced all-cause mortality is the
correct answer. A lower BP goal did not slow progression

Table 1. Summary of Guidelines for Slowing Kidney Function Decline in Patients With CKD
Guidelines BP Control RAAS Inhibition Glycemic Control Diet
KDIGO 2012 Goal BP: ≤140/90 mm Start ACEI or ARB if Goal HbA1c ~7.0%a; Goal < 2 g/d of sodium;
Hg (if UACR < 30 mg/ diabetes and UACR avoid HbA1c < 7.0% if protein intake < 1.3 g/
g) or ≤ 130/80 mm Hg 30-300 mg/g; start at risk of hypoglycemia; kg/d if at risk for CKD
(if UACR ≥ 30 mg/g) ACEI or ARB if allow HbA1c > 7.0% if progression; protein
UACR ≥ 300 mg/g comorbidities, limited intake 0.8 g/kg/d if
life expectancy, or at GFR < 30 mL/min/
risk of hypoglycemia 1.73 m2
AHA/ACC 2017 Goal BP: Start ACEI if HTN and NA Sodium reduction if
<130/80 mm Hg eGFR < 60 mL/min/ HTN
1.73 m2 or
UACR ≥ 300 mg/g; use
ARB if above
indications and ACEI
not tolerated
ADA and EASD 2018 NA NA Goal HbA1c ≤ 7.0% but NA
(for patients with type 2 should be
diabetes) individualized; consider
SGLT2i if at risk for
CKD progression;
consider GLP-1RA if
SGLT2i not tolerated or
contraindicated
ERBP 2015 (for Goal SBP: Start ACEI if diabetes Goal HbA1c ≤ 8.5% if NA
patients with CKD <140 mm Hg and cardiovascular comorbidities, limited
G3b+) indication life expectancy, or at
risk of hypoglycemia;
goal HbA1c ≤ 7.0-8.0%
otherwise
None of the guidelines included recommendations relating to uric acid. Based on KDIGO CKD Work Group 2013 (Kidney Int Suppl, https://doi.org/10.1038/kisup.2
012.77), KDIGO Diabetes Work Group 2020 (Kidney Int, https://doi.org/10.1016/j.kint.2020.06.019), Whelton et al 2018 (Hypertension, https://doi.org/10.1161/
hyp.0000000000000065), Davies et al 2018 (Diabetologia, https://doi.org/10.1007/s00125-018-4729-5), Bilo et al 2015 (Nephrol Dial Transplant, https://doi.org/1
0.1093/ndt/gfv100).
Abbreviations: ACC, American College of Cardiology; ACEI, angiotensin-converting enzyme inhibitor; ADA, American Diabetes Association; AHA, American Heart
Association; ARB, angiotensin receptor blocker; BP, blood pressure; CKD, chronic kidney disease; EASD, European Association for the Study of Diabetes; eGFR,
estimated glomerular filtration rate; ERBP, European Renal Best Practice; GFR, glomerular filtration rate; GLP-1RA , glucagon-like peptide 1 receptor agonist; HbA1c,
hemoglobin A1c; HTN, hypertension; KDIGO, Kidney Disease: Improving Global Outcomes; NA, not available; SBP, systolic blood pressure; SGLT2i, sodium/glucose
cotransporter 2 inhibitor; UACR, urinary albumin-creatinine ratio.
a
The 2020 KDIGO Diabetes in CKD guideline recommends an individualized HbA1c target of <6.5% to <8.0%.

970 AJKD Vol 77 | Iss 6 | June 2021

Chen et al

Table 2. Summary of Major Clinical Trials on Intensive Versus Standard BP Control and Kidney Function Decline
AASK Trial and Cohort SPRINT
(n = 1,094) MDRD Study (n = 840) REIN-II (n = 335) (n = 9,361)
Kidney-related GFR 20-65 mL/min/ Scr 1.2 (F) or 1.4 (M) to Proteinuria 1-3 g/d and eGFR 20-60 mL/min/
inclusion criteria 1.73 m2; 7.0 mg/dL or CLcr < 70 mL/ CLcr < 45 mL/min/1.73 m2 or 1.73 m2;
UPCR ≤ 2.5 g/d min/1.73 m2; proteinuria ≥ 3 g/d and proteinuria < 1 g/d
proteinuria < 10 g/d CLcr < 70 mL/min/1.73 m2
Follow-up Range: 8.8-12.2 y Mean: 2.2 y Median: 1.6 y Median: 3.3 y
% With diabetes 0 0 with “diabetes mellitus 0 with “type 1 diabetes 0
requiring insulin therapy” mellitus”
Intervention MAP ≤ 92 vs 102- MAP 92 vs 107 mm Hg BP < 130/80 vs SBP < 120
107 mm Hg DBP < 90 mm Hg vs < 140 mm Hg

GFR 25-55 GFR 13-24 eGFR 20-59 eGFR ≥60

UPCR UPCR mL/min/ mL/min/ Proteinuria Proteinuria mL/min/ mL/min/
Subgroups ≤0.22 g/g >0.22 g/g 1.73 m2 1.73 m2 1-3 g/d ≥3 g/d 1.73 m2 1.73 m2
Mean baseline eGFR, 51.1 40.0 38.6a 18.5a ~32.9-35.9 ~31.1-41.7 ~47.8-47.9 ~81.1-81.3
mL/min/1.73 m2
Baseline UPCR Median: Median: Median: Median: Mean: ~1.7- Mean: NA NA
or UPE 0.04 g/g 0.68 g/g 0.15 g/g 0.63 g/g 1.8 g/d ~4.9 g/d
Baseline UACR NA NA NA NA NA NA Mean: ~41.1-44.1 mg/g
Kidney function HR, 1.18 HR, 0.73 −1.6 (−0.8, −0.5 (−0.4, HR, 1.06 HR, 1.09 HR, 0.89 HR, 3.49
declinec (0.93-1.50) (0.58-0.93) 3.9) mL/min/ 1.4) mL/min/y (0.51-2.20) (0.55-2.19) (0.42-1.87) (2.44-5.10)
3y
P interaction 0.02 0.02b 0.01b NA NA
Based on information in Appel et al 2010 (N Engl J Med, https://doi.org/10.1056/nejmoa0910975), Klahr et al 1994 (N Engl J Med, https://doi.org/10.1056/nejm1994
03313301301), Ruggenenti et al 2005 (Lancet, https://doi.org/10.1016/s0140-6736(05)71082-5), SPRINT Research Group 2015 (N Engl J Med, https://doi.org/10.1
056/nejmoa1511939).
Abbreviations: AASK, African American Study of Kidney Disease and Hypertension; BP, blood pressure; CLcr, creatinine clearance; DBP, diastolic blood pressure; GFR,
glomerular filtration rate; eGFR, estimated glomerular filtration rate; F, female; HR, hazard ratio; KFRT, kidney failure with replacement therapy; M, male; MAP, mean arterial
pressure; MDRD, Modification of Diet in Renal Disease; NA, not available; REIN, Ramipril Efficacy in Nephropathy; SBP, systolic blood pressure; Scr, serum creatinine;
SPRINT, Systolic Blood Pressure Intervention Trial; UACR, urinary albumin-creatinine ratio; UPCR, urinary protein-creatinine ratio; UPE, urine protein excretion.
a
Denotes GFR.
b
For interaction with baseline proteinuria (<1 g/d vs 1 to <3 g/d vs ≥3 g/d).
c
Defined in each study as follows: AASK (HR for doubling of serum creatinine, KFRT, or death); MDRD Study (mean difference in rate of GFR decline); REIN-II (HR for
KFRT); SPRINT, baseline eGFR 20-59 mL/min/1.73 m2 (HR for first occurrence of ≥50% reduction in eGFR, maintenance dialysis, or kidney transplantation); SPRINT,
eGFR ≥ 60 mL/min/1.73 m2 (HR for ≥30% reduction in eGFR to a level of <60 mL/min/1.73 m2). Values in parentheses are 95% CI.

of CKD, and SPRINT was not powered to assess KFRT and
kidney transplantation events. For Question 2, (c) the
patient with CKD G3bA3 and a UACR of 3,000 mg/g
would most likely benefit from a lower BP goal based on
subgroup analysis from clinical trials. Patients with A1
albuminuria, critical bilateral renal artery stenosis, or
repeated falls are less likely to benefit from a lower BP goal
or may be at higher risk of treatment-related
complications.
Additional Readings
➢ Appel LJ, Wright JT, Greene T, et al. Intensive blood-pressure
control in hypertensive chronic kidney disease. N Engl J Med
2010;363(10):918-929. +ESSENTIAL READING
➢ Cheung AK, Rahman M, Reboussin DM, et al. Effects of intensive
BP control in CKD. J Am Soc Nephrol. 2017;28(9):2812-2823.
➢ Kidney Disease: Improving Global Outcomes (KDIGO) CKD
Work Group. KDIGO 2012 clinical practice guideline for the
evaluation and management of chronic kidney disease. Kidney Int
Suppl. 2013;3(1):1-150. +ESSENTIAL READING
➢ Klahr S, Levey AS, Beck GJ, et al. The effects of dietary protein
restriction and blood-pressure control on the progression of
chronic renal disease. N Engl J Med. 1994;330(13):877-884.
+ESSENTIAL READING
➢ Ruggenenti P, Perna A, Loriga G, et al. Blood-Pressure Control
for Renoprotection in Patients With Non-diabetic Chronic Renal
Disease (REIN-2): multicenter, randomized controlled trial. Lan-
cet.2005;365(9463):939-946.
➢ SPRINT Research Group. A randomized trial of intensive versus
standard blood-pressure control. N Engl J Med.
2015;373(22):2103-2116. +ESSENTIAL READING
➢ Whelton PK, Carey RM, Aronow WS, et al. 2017 ACC/AHA/
AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA
guideline for the prevention, detection, evaluation, and manage-
ment of high blood pressure in adults: a report of the American
College of Cardiology/American Heart Association Task
Force on Clinical Practice Guidelines. Hypertension. 2018;71(6):
e13-e115.
➢ Wright JT, Bakris G, Greene T, et al. Effect of blood pressure
lowering and antihypertensive drug class on progression of
hypertensive kidney disease: results from the AASK trial. JAMA.
2002;288(19):2421-2431.
RAAS Inhibition
Case 2: A 46-year-old woman with type 2 diabetes returns
for her second appointment. Her history is notable for retin-
opathy and CKD G3aA3 attributed to diabetic kidney dis-
ease. She denies having orthostatic symptoms or chest
discomfort. Her automated office BP is 118/75 mm Hg on
atenolol and chlorthalidone. Laboratory testing reveals stable
eGFR (at 55 mL/min/1.73 m2) with a UACR of 1,200 mg/g.

AJKD Vol 77 | Iss 6 | June 2021 971


Question 3: Which one of the following would be the
most appropriate antihypertensive therapy to help
slow CKD progression?
a) No change in therapy because her BP is controlled to goal
b) Change atenolol to an angiotensin receptor blocker (ARB)
c) Change chlorthalidone to an ARB
d) Add an ARB to the current 2-drug regimen
Case 3: A 56-year-old woman with CKD G3aA3 due to
biopsy-proven diabetic kidney disease has an average out-of-
office BP of 144/83 mm Hg on a regimen of lisinopril at
20 mg daily, chlorthalidone at 50 mg daily, and amlodipine at
10 mg daily. Her UACR is 800 mg/g.
Question 4: Which one of the following interventions
would be most appropriate to reduce the risk of CKD
progression?
a) Add an ARB to the current regimen
b) Change lisinopril to a mineralocorticoid receptor antago-
nist (MRA)
c) Increase the lisinopril dosage
d) Change chlorthalidone to indapamide
For the answers to the questions, see the following text.
The cornerstone of albuminuria management is RAAS
inhibition. The KDIGO guidelines recommend that all
adults with CKD, hypertension, and a UACR of >300 mg/
g be treated with an angiotensin-converting enzyme
inhibitor (ACEI) or ARB. Among those with diabetes and
UACR > 30 mg/g, ACEI or ARB use should be considered.
RAAS inhibition in patients with CKD and hypertension is
also supported by all major hypertension guidelines
(Table 1). Multiple trials have demonstrated that ACEI or
ARB therapy delays CKD progression among individuals
with albuminuria (Table 3). The REIN trial, which
randomized patients with CKD to ramipril versus placebo,
showed that mean GFR decline was significantly slower in
the ramipril group among participants with
proteinuria ≥ 3 g/d. In the RENAAL study, patients with
type 2 diabetes and CKD randomized to losartan treatment
had a 16% lower risk of developing a doubling of serum
creatinine, KFRT, or death compared with the placebo
group. Similarly, IDNT reported that irbesartan treatment
was associated with a lower risk of a doubling of serum
creatinine, serum creatinine ≥ 6.0 mg/dL, KFRT, or death
compared with amlodipine or placebo treatments among
patients with hypertension and CKD attributed to type 2
diabetes. Finally, in AASK, use of ramipril was independ-
ently associated with 22% and 38% lower risks of the
clinical composite outcome (GFR decline ≥50%
or ≥25 mL/min/1.73 m2 from baseline, KFRT, or death)
compared with metoprolol and amlodipine, respectively.
The current literature does not support the use of dual
blockade with an ACEI and ARB in diabetic kidney disease.
VA NEPHRON-D, which randomized veterans with type 2
972
Chen et al
diabetes and CKD G2-G3bA3 to losartan plus lisinopril or
losartan alone, was terminated early due to safety con-
cerns, with the combination therapy group having a
markedly higher risk of hyperkalemia (HR, 2.8 [95% CI,
1.8-4.3]) and acute kidney injury (HR, 1.7 [95% CI, 1.3-
2.2]) compared with the monotherapy group. Addition-
ally, there was no significant difference in risk of kidney
function decline between the 2 treatment groups, though
the follow-up period was short (Table 3).
Decreased sodium intake may enhance the renopro-
tective effects of RAAS inhibitors. A meta-analysis of 11
studies (23 cohorts with 516 participants) reported that
dietary sodium restriction (average decrease of 92 mmol/
d) was associated with a 32% lower urine albumin
excretion. The reduction in urine albumin excretion was
greater in the cohorts with concomitant RAAS blockade
therapy than in those without (pooled mean differences
of −41.9% and −17.2%, respectively; P = 0.01 for inter-
action), suggesting a synergistic effect of low sodium
intake with RAAS inhibition. In a post hoc analysis of 500
participants in the REIN and REIN II trials receiving ram-
ipril therapy, a diet with > 14 g/d of salt was associated
with 3.3-fold and 2.4-fold greater risks of KFRT compared
with diets of <7 g/d and 7 to 14 g/d of salt, respectively.
Importantly, the proteinuria-reducing effects of ramipril
were greatest in the low-sodium diet group. In another
post hoc analysis of the RENAAL study and IDNT
(n = 1,177), ARB therapy was associated with a 43% lower
risk of a renal event, defined as a doubling of serum cre-
atinine or KFRT, compared with non-RAAS inhibitor
therapy among participants in the lowest tertile of the
24-hour urinary sodium-creatinine ratio with no sig-
nificant difference in risk between the 2 treatment groups
for higher tertiles of sodium intake (P < 0.001 for inter-
action; Fig 1). Given these findings, patients on RAAS
inhibitors for treatment of albuminuria should be
encouraged to follow a low-sodium diet.
For patients intolerant of ACEI/ARB therapy, an MRA
can be considered. A recent meta-analysis of 31 random-
ized controlled trials evaluated the efficacy and safety of
MRAs (spironolactone, eplerenone, canrenone, or finer-
enone) compared with active control or placebo in
reducing albuminuria. In the 18 trials (n = 2,036) that
examined UACR as an outcome, proportional change in
UACR from baseline to end of treatment was 22% lower in
MRA treatment compared with active control and placebo.
The effect persisted when comparing MRAs to placebo
(n = 1,436 in 11 trials) in patients on ACEI/ARB therapy.
When comparing MRAs to renin-angiotensin blockers,
there was no significant difference in change in albu-
minuria (n = 201 in 2 trials), but the risk of incident
hyperkalemia was 70% higher (n = 855 in 5 trials).
Although reduction in albuminuria is not a universally
accepted surrogate end point for KFRT, the FIDELIO-DKD
trial of patients with type 2 diabetes and CKD (>98% on
concomitant ACEI or ARB therapy) reported that
AJKD Vol 77 | Iss 6 | June 2021
Chen et al

Table 3. Summary of Major Clinical Trials on ACEI and ARB Therapy on Kidney Function Decline
REIN, Stratum 2 RENAAL AASK Trial IDNT VA NEPHRON-D
(n = 166) (n = 1,513) (n = 1,094) (n = 1,715) (n = 1,448)
Kidney-related CLcr 20-70 mL/ Scr 1.3-3.0 mg/dL; GFR 20-65 mL/ Scr 1.0 (F) or eGFR 30-<90 mL/
inclusion criteria min/1.73 m2; UACR ≥ 300 mg/g min/1.73 m2; 1.2 (M) to 3.0 mg/dL; min/1.73 m2;
proteinuria ≥ 3 g/d UPCR ≤ 2.5 g/d proteinuria ≥ UACR ≥ 300 mg/g
900 mg/d
Follow-up Mean: ~1.3 y Mean: 3.4 y Range: 3.0-6.4 y Mean: 2.6 y Median: 2.2 y
% With diabetes 0a 100% 0 100% 100%
% With HTN 87% 93%b 100% 100% NA
Intervention Ramipril vs Losartan vs Ramipril vs Irbesartan vs placebo Losartan +
placebo placebo metoprolol vs vs amlodipine lisinopril vs
amlodipine losartan +
placebo
Mean baseline eGFR, GFR 37.4- Scr ~1.9 mg/dLc GFR 45.6 mL/min/ Scr ~1.7 mg/dLc eGFR ~53.6-
GFR, or Scr 40.2 mL/min/1.73 1.73 m2 53.7 mL/min/1.73
m2 m2
Baseline UPCR or Mean: 5.1-5.6 g/d NA Median: 0.08 g/g Median: ~2.9 g/d Median: ~1.6-2.1 g/
UPE g
Baseline UACR or NA Median: ~1,237- NA Median: ~1.9 g/d Median: 847 mg/g
UAE 1,261 mg/g
Kidney function 0.53 vs 0.88 mL/ HR, 0.84 Risk reduction for RR for irbesartan: HR, 0.88 (0.70-1.12)
declined min/mo; P = 0.03e (0.72-0.98) ramipril: 22% (1%- 0.81 (0.67-0.99) vs
38%) vs placebo, 0.76 (0.63-
metoprolol, 38% 0.92) vs amlodipine
(14%-56%) vs
amlodipine
Based on information in GISEN Group 1997 (Lancet, https://doi.org/10.1016/S0140-6736(96)11445-8), Brenner et al 2001 (N Engl J Med, https://doi.org/10.1056/
nejmoa011161), Wright et al 2002 (JAMA, https://doi.org/10.1001/jama.288.19.2421), Lewis et al 2001 (N Engl J Med, https://doi.org/10.1056/nejmoa011303), Fried
et al 2013 (N Engl J Med, https://doi.org/10.1056/nejmoa1303154).
Abbreviations: AASK, African American Study of Kidney Disease and Hypertension; ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; CLcr,
creatinine clearance; eGFR, estimated glomerular filtration rate; F, female; GFR, glomerular filtration rate; HTN, hypertension; HR, hazard ratio; IDNT, Irbesartan Diabetic
Nephropathy Trial; KFRT, kidney failure with replacement therapy; M, male; NA, not available; RENAAL, Reduction in End Points in Non–Insulin-Dependent Diabetes With the
Angiotensin II Antagonist Losartan; RR, relative risk; Scr, serum creatinine; REIN, Ramipril Efficacy in Nephropathy; UACR, urinary albumin-creatinine ratio; UAE, urine albumin
excretion; UPCR, urinary protein-creatinine ratio; UPE, urine protein excretion; VA NEPHRON-D, Veterans Affairs Nephropathy in Diabetes Study.
a
None with “insulin-dependent diabetes mellitus.”
b
Percentage receiving anti-HTN drugs at baseline.
c
Neither eGFR nor GFR is not available.
d
Defined in each study as follows: REIN (GFR decline per month); RENAAL (HR for Scr doubling, KFRT, or death); IDNT (RR for Scr doubling, Scr ≥ 6.0 mg/dL, KFRT, or
death); AASK (GFR decline ≥ 50% or ≥ 25 mL/min/1.73 m2 from baseline, KFRT, or death); VA NEPHRON-D (HR for first occurrence of absolute decline in eGFR ≥ 30 mL/
min/1.73 m2 if eGFR at randomization ≥ 60 mL/min/1.73 m2, relative decline in eGFR ≥ 50% if eGFR at randomization < 60 mL/min/1.73 m2; eGFR < 15 ml/min/1.73 m2;
KFRT; or death).
e
Analysis among 117 participants with at least 3 GFR measurements.

Figure 1. Kaplan-Meier curves for renal events by tertiles of 24-hour urinary sodium-creatinine ratio (<121 mmol/g; 121 to <153 m-
mol/g; ≥153 mmol/g) among RENAAL and IDNT trial participants on non–RAASi-based therapy and ARB therapy. Renal event
defined as a doubling of serum creatinine from baseline or KFRT (RENAAL and IDNT) or serum creatinine ≥ 6.0 mg/dL (IDNT
only). Abbreviations: ARB, angiotensin receptor blocker; IDNT, Irbesartan Diabetic Nephropathy Trial; KFRT, kidney failure with
replacement therapy; non-RAASi, non–renin-angiotensin-aldosterone system inhibitor; RENAAL, Reduction in End Points in
Non–Insulin-Dependent Diabetes With the Angiotensin II Antagonist Losartan. Adapted from Heerspink et al 2012 (Kidney Int,
https://doi.org/10.1038/ki.2012.74) with permission of the copyright holder. Original graphic © 2012 International Society of
Nephrology.
AJKD Vol 77 | Iss 6 | June 2021 973

finerenone conferred an 18% lower risk of composite
kidney outcome (sustained decline in eGFR by ≥40% or
to <15 mL/min/1.73 m2, KFRT, or death from kidney
causes) compared with placebo. Thus, MRAs reduce
albuminuria and may also slow CKD progression. These
benefits, however, must be balanced against the potential
risk of hyperkalemia.
In Question 3, (b) changing atenolol to an ARB is the
correct answer. ACEIs and ARBs have been shown to slow
the progression of CKD in patients with diabetes while
β-blockers have not. In the management of hypertension,
β-blockers are an add-on therapy after the use of first-line
agents such as ACEIor ARBs and thiazide diuretics. Adding
an ARB to the current regimen is less desirable, as this may
result in hypotension in a patient with BP already con-
trolled to goal.
In Question 4, (c) increasing the lisinopril dose is the
best answer. Combination ACEI and ARB therapy is asso-
ciated with an increased risk of adverse outcomes.
Although an MRA may reduce albuminuria when com-
bined with an ACEI or ARB, no randomized controlled
trials have been performed to support changing an ACEI to
an MRA with the intent of slowing progression to KFRT.
Exchanging chlorthalidone for indapamide is not antici-
pated to slow this progression.
Additional Readings
➢ Alexandrou ME, Papagianni A, Tsapas A, et al. Effects of miner-
alocorticoid receptor antagonists in proteinuric kidney disease: a
systematic review and meta-analysis of randomized controlled
trials. J Hypertens. 2019;37(12):2307-2324. +ESSENTIAL
READING
➢ Bakris GL, Agarwal R, Anker SD, et al. Effect of finerenone on
chronic kidney disease outcomes in type 2 diabetes. N Engl J
Med. 2020;383(23):2219-2229.
➢ Brenner BM, Cooper ME, De Zeeuw D, et al. Effects of losartan
on renal and cardiovascular outcomes in patients with type 2
diabetes and nephropathy. N Engl J Med. 2001;345(12):861-
869. +ESSENTIAL READING
➢ D’Elia L, Rossi G, Schiano di Cola M, et al. Meta-analysis of the
effect of dietary sodium restriction with or without concomitant
renin-angiotensin-aldosterone system-inhibiting treatment on
albuminuria. Clin J Am Soc Nephrol. 2015;10(9):1542-1552.
➢ Fried LF, Emanuele N, Zhang JH, et al. Combined angiotensin
inhibition for the treatment of diabetic nephropathy. N Engl J
Med. 2013;369(20):1892-1903. +ESSENTIAL READING
➢ GISEN Group (Gruppo Italiano di Studi Epidemiologici in
Nefrologia). Randomised placebo-controlled trial of effect of
ramipril on decline in glomerular filtration rate and risk of terminal
renal failure in proteinuria, non-diabetic nephropathy. Lancet.
1997;349(9069):1857-1863.
➢ Kidney Disease: Improving Global Outcomes (KDIGO) CKD
Work Group. KDIGO 2012 clinical practice guideline for the
evaluation and management of chronic kidney disease. Kidney Int
Suppl. 2013;3(1):1-150. +ESSENTIAL READING
➢ Lambers Heerspink HJ, Holtkamp FA, Parving HH, et al. Mod-
eration of dietary sodium potentiates the renal and cardiovascular
protective effects of angiotensin receptor blockers. Kidney Int.
2012;82(3):330-337.
➢ Lewis EJ, Hunsicker LG, Clark WR, et al. Renoprotective effect of
the angiotensin-receptor antagonist irbesartan in patients with
974
Chen et al
nephropathy due to type 2 diabetes. N Engl J Med.
2001;345(12):851-860.
➢ Vegter S, Perna A, Postma MJ, et al. Sodium intake, ACE
inhibition, and progression to ESRD. J Am Soc Neph-
rol.2012;23(1):165-173. +ESSENTIAL READING
➢ Wright JT, Bakris G, Greene T, et al. Effect of blood pressure
lowering and antihypertensive drug class on progression of
hypertensive kidney disease: results from the AASK trial. JAMA.
2002;288(19):2421-2431. +ESSENTIAL READING
Glycemic Control
Case 4: A 48-year-old man with type 2 diabetes is seen for a
routine follow-up visit. He has stable CKD (G3aA2 for 3
years) in addition to retinopathy, hypertension, and hyper-
lipidemia. He is currently taking losartan at 50 mg daily,
metoprolol at 75 mg twice daily, metformin at 500 mg twice
daily, and atorvastatin at 20 mg daily. His BP is 127/68 mm
Hg with a heart rate of 64 bpm. The remainder of the physical
examination is unremarkable. Laboratory evaluation demon-
strates his eGFR is 52 mL/min/1.73 m2, UACR 35 mg/g, and
hemoglobin A1c (HbA1c) 7.9%.
Question 5: Which one of the following interventions
would be most likely to slow the progression of his
CKD?
a) Increase losartan to reduce BP to < 120/80 mm Hg
b) Change metoprolol to a dihydropyridine calcium channel blocker
c) Increase metformin to target HbA1c < 7%
d) No changes, as the management of hypertension and
glycemic control are at goal
For the answer to the question, see the following text.
The 2020 KDIGO guidelines on diabetes in CKD rec-
ommend that HbA1c goals be individualized based on CKD
severity, comorbidities, and hypoglycemia risk, among
other factors (Table 4). Dosing adjustments or dis-
continuation of glucose-lowering agents are often neces-
sary as CKD progresses. In particular, insulins,
sulfonylureas, and meglitinides are more likely to cause
hypoglycemia in a patient with reduced kidney function.
Most major randomized controlled trials have suggested that
intensive glycemic control reduces albuminuria and possibly
also kidney function decline in patients with diabetes (Table 5).
In the ADVANCE-ON and DCCT/EDIC studies, intensive gly-
cemic control was associated with a 46% lower risk of KFRT or
death from kidney disease and 50% lower risk of incident
eGFR < 60 mL/min/1.73 m2, respectively. A large meta-
analysis of the ADVANCE, ACCORD, UKPDS, and VADT trials
showed that more intensive glycemic control was associated
with a 20% lower risk of developing a primary kidney event
(ie, incident eGFR < 30 mL/min/1.73 m2, UACR >
300 mg/g, KFRT, or death from kidney disease), mostly due
to a reduction in risk of albuminuria. Participants with more
intensive control were also more likely to have regression of
UACR from >300 to 30-300 mg/g (HR, 1.23 [95% CI,
1.03-1.48]), regression of UACR from 30-300 to <30 mg/g
AJKD Vol 77 | Iss 6 | June 2021
Chen et al

Table 4. Factors to Consider When Determining Hemoglobin ➢ UK Prospective Diabetes Study Group. Intensive blood-
A1c Targets in Patients With Diabetes and CKD glucose control with sulphonylureas or insulin compared with
conventional treatment and risk of complications in patients
Goal hemoglobin A1c with type 2 diabetes (UKPDS 33). Lancet. 1998;352(9131):
Factors <6.5% <8.0% 837-853.
CKD severity Lower (eg, Higher (eg, ➢ Wong MG, Perkovic V, Chalmers J, et al. for the ADVANCE-ON
G1) G5) Collaborative Group. Long-term benefits of intensive glucose
Macrovascular complications None or mild Severe control for preventing end-stage kidney disease: ADVANCE-ON.
Comorbidities None or few Many Diabetes Care. 2016;39(5):694-700.
➢ Zoungas S, Arima H, Gerstein HC, et al. Effects of intensive
Life expectancy Long Short
glucose control on microvascular outcomes in patients with type
Hypoglycemia awareness Good Diminished
2 diabetes: a meta-analysis of individual participant data from
Resources for hypoglycemia Available Limited randomized controlled trials. Lancet Diabetes Endocrinol.
management
2017;5(6):431-437. +ESSENTIAL READING
Likelihood of treatments Low High
causing hypoglycemia
Based on information in KDIGO Diabetes Work Group 2020 (Kidney Int, https://
doi.org/10.1016/j.kint.2020.06.019). SGLT2 Inhibitors
Abbreviations: CKD, chronic kidney disease; G, glomerular filtration rate category.
Case 4, cont’d: The patient returns for a follow-up visit. His
CKD has steadily worsened over the past 12 months. Lab-
(HR, 1.15 [95% CI, 1.03-1.28]), and maintenance of oratory studies reveal:
UACR < 30 mg/g (HR, 1.16 [95% CI, 1.08-1.25]). Parameter Present 6 mo prior 12 mo prior
For Question 5, (c) targeting a HbA1c of 7% or less is the Sodium, mEq/L 132 134 131
best option among those listed to help slow CKD pro- Potassium, mEq/L 5.3 5.2 5.4
gression. In patients with CKD G4-G5 or significant com- Glucose, mg/dL 315 280 210
peting comorbidities where risk of hypoglycemia is higher eGFR, mL/min/1.73 m2 46 51 52
and benefits of intense control less well established, the UACR, mg/g 1,000 400 35
HbA1c target should be individualized. The patient’s BP is HbA1c 9.4% — 7.9%
currently controlled to goal, and further reduction or sub-
stitution of a dihydropyridine calcium blocker for a
β-blocker has not been shown to slow CKD progression. Question 6: Which one of the following interventions
would be the next best step?
Additional Readings a) Start treatment with a SGLT2 inhibitor to reduce albuminuria
➢ Bilo H, Coentr~ao L, Couchoud C, et al. for the Guideline b) Start treatment with a SGLT2 inhibitor to achieve an
Development Group. Clinical practice guideline on management HbA1c <7%
of patients with diabetes and chronic kidney disease stage 3b or c) Do not start treatment with a SGLT2 inhibitor because of a
higher (eGFR<45 mL/min). Nephrol Dial Transplant. risk for worsening hyperkalemia
2015;30(suppl 2):ii1-ii142. d) Do not start treatment with a SGLT2 inhibitor because of a
➢ DCCT/EDIC Research Group. Intensive diabetes therapy and risk for worsening hyponatremia
glomerular filtration rate in type 1 diabetes. N Engl J Med.
2011;365(25):2366-2376. Case 5: A 70-year-old woman with moderate obesity (body
➢ Diabetes Control and Complications Trial Research Group. The mass index [BMI] of 32 kg/m2) and uncontrolled diabetes
effect of intensive treatment of diabetes on the development and mellitus (HbA1c 8.2%) returns to discuss the management of
progression of long-term complications in insulin-dependent her CKD G3aA2. Her history includes coronary artery disease
diabetes mellitus. N Engl J Med. 1993;329(14):977-986. and recurrent furunculosis requiring antibiotics several times
➢ Duckworth W, Abraira C, Mortiz T, et al. Glucose control and per year.
vascular complications in veterans with type 2 diabetes. N Engl J
Med. 2009;360(2):129-139.
Question 7: In counseling her on the potential benefits
➢ Epidemiology of Diabetes Interventions and Complications
and risks of therapy with an SGLT2 inhibitor, for which one
Research Group. Sustained effect of intensive treatment of type
of the following adverse effects is she at greatest risk?
1 diabetes mellitus on development and progression of diabetic
nephropathy. JAMA. 2003;290(16):2159-2167. a) Genitourinary fungal infections
➢ Ismail-Beigi F, Craven T, Banerji MA, et al. Effect of intensive b) Lower extremity amputation
treatment of hyperglycaemia on microvascular outcomes in type c) Severe hypoglycemia
2 diabetes: an analysis of the ACCORD randomized trial. Lancet. d) Acute kidney injury
2010;376(9739):419-430.
For the answers to the questions, see the following text.
➢ Kidney Disease: Improving Global Outcomes (KDIGO) Diabetes
Work Group. KDIGO 2020 clinical practice guideline for dia-
betes management in chronic kidney disease. Kidney Int.
In recent years, SGLT2 inhibitors have emerged as new,
2020;98(4S):S1-S115. +ESSENTIAL READING
➢ Perkovic V, Heerspink HL, Chalmers J, et al. Intensive glucose
exciting therapies for delaying CKD progression, partic-
control improves kidney outcomes in patients with type 2 dia- ularly among patients with type 2 diabetes and/or albu-
betes. Kidney Int. 2013;83(3):517-523. minuria. Current ADA and EASD guidelines recommend

AJKD Vol 77 | Iss 6 | June 2021 975

 Chen et al

Table 5. Summary of Major Clinical Trials on Intensive Versus Standard Glycemic Control on Kidney Outcomes
ADVANCE ACCORD UKPDS VADT DCCT EDIC
(n = 11,140) (n = 10,251) (n = 3,867) (n = 1,791) (n = 1,441) (n = 1,349)
Kidney- Not specified Scr ≤ 132.6 μmol/ Scr ≤ 175 μmol/L Scr ≤ 1.6 mg/dL Scr < 1.2 mg/dL Scr < 1.2 mg/dL or
related L or CLcr > 100 mL/ CLcr > 100 mL/
inclusion min/1.73 m2; min/1.73 m2;
criteria UAE < 40 mg/d UAE < 40 mg/d
Follow-up Median: 5 y Mean: ~3.5 y Median: ~10 y Median: 5.6 y Mean: 6.5 y Mean: ~8 y
Diabetes Type 2 Type 2 Type 2 Type 2 Type 1 Type 1
type
Diabetes Median: 7 y Median: ~10 y “Newly Mean: ~11.5 y Mean: ~2.6a and Mean: ~12 y
duration diagnosed” ~8.6-8.9b y
Intervention HbA1c ≤6.5% vs HbA1c <6.0% vs Median HbA1c Median HbA1c HbA1c <6.05% vs None (obs follow-
standarda 7.0%-7.9% ~7.0% vs ~7.9% ~6.9% vs 8.4% standard up of DCCT)
Mean ~7.5% ~8.1% 7.08% ~9.4% ~8.8%b and ~7.4%d and 9.1%e
baseline ~8.9%-9.0%c
HbA1c
Baseline Mean eGFR Median eGFR NA Mean Scr ~1.0 Mean CLcr ~127- Mean CLcr ~122
eGFR, CLcr, ~78.0-78.3 mL/ ~90 mL/min/1.73 mg/dL 128b and ~128- mL/min
or Scr min/1.73 m2 m2 130c mL/min
Baseline Median UACR Median UACR NA NA Mean UAE ~12b Median UAE 8.6d
UACR or ~14.9-15.0 μg/mg ~1.54 mg/mmol and ~19-21c mg/d and 10.1e mg/d
UAEf
Kidney UACR ≥ 30 μg/ UACR ≥ 30 mg/g: UACR ≥ 30 mg/g: Any ↑ in Risk reduction: Risk reduction:
outcomeg mg: HR, 0.91 HR, 0.79 (0.69- RR, 0.70 (0.46- albuminuria: 9.1% 39% (21%-52%) 59% (39%-73%)
(0.85-0.98); 0.90); 1.07); vs 13.8% for UAE ≥ 40 mg/ for UAE ≥ 40 mg/
UACR > 300 μg/ UACR ≥ 300 mg/ “Proteinuria”: RR, (P = 0.01); from d; 54% (19%- d; 84% (67%-
mg: HR, 0.70 g: HR, 0.69 (0.55- 0.58 (0.23-1.43); normal to 74%) for 92%) for UAE >
(0.57-0.85); KFRT: 0.85); KFRT or Pcr doubling: RR, UACR ≥ 30 mg/g: UAE ≥300 mg/d 300 mg/d
HR, 0.35 (0.15- SCr >291.72 1.25 (0.16-9.55) 10.0% vs 14.7% Risk reduction: 50% (18%, 69%)h for
0.83) μmol/L: HR, 0.95 (P = 0.03); Scr sustained eGFR < 60 mL/min/1.73 m2
(0.73-1.24) doubling: 8.8% vs
8.8% (P = 0.99)
Based on information in Perkovic et al 2013 (Kidney Int, https://doi.org/10.1038/ki.2012.401), Ismail-Beigi 2010 (Lancet, https://doi.org/10.1016/s0140-6736(10)
60576-4), UKPDS Group 1998 (Lancet, https://doi.org/10.1016/S0140-6736(98)07019-6), Duckworth et al 2009 (N Engl J Med, https://doi.org/10.1056/nejmoa08
08431), DCCT group 1993 (N Engl J Med, https://doi.org/10.1056/nejm199309303291401), EDIC group 2003 (JAMA, https://doi.org/10.1001/jama.290.16.215
9), DCCT/EDIC group 2011 (N Engl J Med, https://doi.org/10.1056/nejmoa1111732).
Abbreviations: ACCORD, Action to Control Cardiovascular Risk in Diabetes; ADVANCE, Action in Diabetes and Cardiovascular Disease: Preterax and Diamicron
Modified Release Controlled Evaluation; CLcr, creatinine clearance; DCCT, Diabetes Control and Complications Trial; EDIC, Epidemiology of Diabetes Interventions and
Complications; eGFR, estimated glomerular filtration rate; HbA1c, hemoglobin A1c; HR, hazard ratio; obs, observational; RR, relative risk; Scr, serum creatinine; UAE,
urine albumin excretion; UKPDS, UK Prospective Diabetes Study; UACR, urinary albumin-creatinine ratio; UPCR, urinary protein-creatinine ratio; VADT, Veterans Affairs
Diabetes Trial.
a
Based on local guidelines.
b
Primary prevention cohort of the DCCT.
c
Secondary intervention cohort of the DCCT.
d
Intensive group of original DCCT.
e
Conventional group of original DCCT.
f
Baseline UPCR information not available.
g
For UKPDS, from 0-15 years of follow-up with outcome assessed every 3 years.
h
Over median follow-up of 22 years (includes DCCT and EDIC years 1-16).

that SGLT2 inhibitors be considered in all patients with
type 2 diabetes at risk of CKD progression, regardless of
cardiovascular disease history (Table 1).
Initial reports of the potential kidney protective effects
of SGLT2 inhibitors came from cardiovascular outcome
trials. These studies, however, primarily included indi-
viduals with mild or no CKD and were limited by small
numbers of kidney events. In 2019, the results of the
landmark CREDENCE trial were published. This trial, the
first to examine the association of a SGLT2 inhibitor with a
primary kidney outcome, reported that among patients
with type 2 diabetes and CKD G2-G3bA3, randomization
to canagliflozin was associated with a 30% lower risk (95%
CI, 18%-41%) of developing a composite outcome (KFRT,
sustained eGFR < 15 ml/min/1.73 m2, doubling of serum
creatinine from baseline, death from kidney disease, or
death from cardiovascular disease) compared with pla-
cebo. Similar conclusions were obtained when considering
individual components of the composite outcome.
Importantly, all participants were on an ACEI or ARB, thus
suggesting that the benefits of canagliflozin extended
beyond standard pharmacologic therapy (ie, RAAS
inhibition).
Neuen et al performed a meta-analysis of four
randomized, controlled trials that investigated the effect of
SGLT2 inhibitors on major kidney outcomes among
patients with type 2 diabetes and eGFR ≥ 30 mL/min/
1.73 m2: CREDENCE, CANVAS Program, EMPA-REG
OUTCOME, and DECLARE-TIMI 58. The majority of
patients in the latter 3 trials did not have baseline CKD

976 AJKD Vol 77 | Iss 6 | June 2021

Chen et al

(Table 6). Among 38,723 participants, use of SGLT2
inhibitors was associated with a 33% lower risk of a
composite outcome of dialysis, transplantation, or death
from kidney disease compared with placebo. Importantly,
the benefits of SGLT2 inhibitors were statistically sig-
nificant in all subgroups of baseline eGFR (30 to <45, 45
to <60, 60 to <90, and ≥90 mL/min/1.73 m2). The eGFR
decline was also slower in the SGLT2 inhibitor group
versus placebo in CREDENCE (absolute difference, 2.74
[95% CI, 2.37-3.11] mL/min/1.73 m2 per year),
CANVAS Program (absolute difference, 1.18 [95% CI,
1.02-1.35] mL/min/1.73 m2 per year), and EMPA-REG
OUTCOME (absolute difference, 1.68 [95% CI, 1.02-
1.35] mL/min/1.73 m2 per year).
More recently, the DAPA-CKD trial demonstrated that
the renoprotective effects of SGLT2 inhibitors likely extend
beyond patients with type 2 diabetes. This trial, which
enrolled individuals with CKD (32.5% without type 2
diabetes), was stopped early because of clear efficacy of
dapagliflozin over placebo for the primary outcome

Table 6. Summary of Trials on SGLT2 Inhibitors With Kidney Outcomes in Patients With and Without Type 2 Diabetes
CANVAS EMPA-REG DECLARE-
Program OUTCOME TIMI 58 DAPA-CKD
Trial CREDENCE (n = 4,401) (n = 10,142) (n = 7,020) (n = 17,160) (n = 4,304)
Study or Participant Characteristics
Inclusion criteria eGFR 30-<90 mL/min/ eGFR ≥ 30 mL/ eGFR ≥ 30 mL/ CLcr ≥ 60 mL/ eGFR 25-75 mL/min/
1.73 m2; UACR > 300- min/1.73 m2 min/1.73 m2 min 1.73 m2; UACR 200-
5,000 mg/g 5,000 mg/g
SGLT2i Canagliflozin Canagliflozin Empagliflozin Dapagliflozin Dapagliflozin
Median follow-up, y 2.6 2.4 3.1 4.2 2.4
Baseline ACEI or ARB 4,395 (99.9%) 8,116 (80%) 5,666 (81%) 13,950 (81%) 1354 (31%)a; 2,870
use (67%)b
eGFR, mL/min/1.73 m2
≥90 0 (0) 2,476 (24%) 1,538 (22%) 8,162 (48%) 0 (0)
60-<90 1,809 (41%) 5,625 (55%) 3,661 (52%) 7,732 (45%) 454 (11%)
45-<60 1,279 (29%) 1,485 (15%) 1,249 (18%) 1,265 (7%) 1,328 (31%)
30-<45 1,313 (30%) 554 (5%) 570 (8%) NA 1,898 (44%)
<30 0 (0) 0 (0) 0 (0) 0 (0) 624 (14%)
Missing 0 (0) 2 (<0.1%) 2 (<0.1%) 1 (<0.1%) 0 (0)
UACR, mg/g
<30 0 (0) 7,007 (69%) 4,171 (59%) 11,644 (68%) NAc
30-300 0 (0) 2,266 (22%) 2,013 (29%) 4,030 (24%) NA
>300c 4,401 (100%) 760 (7%) 769 (11%) 1,169 (7%) NA
Missing 0 (0) 109 (1%) 67 (1%) 317 (2%) NA
RR or HR Ratio Comparing SGLT2i With Placebo
Dialysis, Tx, or death from 0.72 (0.54-0.97) 0.56 (0.23- 0.90 (0.30-2.67) 0.42 (0.20- NA
kidney disease 1.32) 0.87)
Dialysis, Tx, or sustained 0.68 (0.54-0.86) 0.77 (0.30- 0.60 (0.18-1.98) 0.31 (0.13- 0.64 (0.50-0.82)
eGFR < 15 mL/min/ 1.97) 0.79)
1.73 m2,d
Substantial loss of kidney 0.66 (0.53-0.81) 0.53 (0.33- 0.54 (0.40-0.75) 0.53 (0.43- 0.56 (0.45-0.68)
functione; dialysis, Tx, or 0.84) 0.66)
sustained
eGFR < 15 mL/min/
1.73 m2,d; or death from
kidney disease
All participants in CREDENCE, CANVAS Program, EMPA-REG OUTCOME, and DECLARE-TIMI 58 were with type 2 diabetes. In DAPA-CKD, 67.5% of participants
were with type 2 diabetes. Based on information in Neun et al 2019 (Lancet Diabetes Endocrinol, https://doi.org/10.1016/s2213-8587(19)30256-6), and Heerspink et al
2020 (N Engl J Med, https://doi.org/10.1056/nejmoa2024816).
Abbreviations: ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; CANVAS, Canagliflozin Cardiovascular Assessment Study; CLcr,
creatinine clearance; CREDENCE, Canagliflozin and Renal Events in Diabetes with Established Nephropathy Clinical Evaluation; DAPA-CKD, Dapagliflozin and Pre-
vention of Adverse Outcomes in Chronic Kidney Disease; DECLARE-TIMI 58, Dapagliflozin Effect on Cardiovascular Events–Thrombolysis in Myocardial Infarction 58;
eGFR, estimated glomerular filtration rate; EMPA-REG OUTCOME, BI 10773 (Empagliflozin) Cardiovascular Outcome Event Trial in Type 2 Diabetes Mellitus Patients;
HR, hazard ratio; KFRT, kidney failure with replacement therapy; NA, not available; RR, relative risk; SGLT2i, sodium/glucose cotransporter 2 inhibitor; Tx, transplantation;
UACR, urinary albumin-creatinine ratio.
a
On ACEI.
b
On ARB.
c
n = 2,079 (48%) with UACR of >1,000 mg.
e
Defined as a doubling of serum creatinine with the exception of DECLARE-TIMI 58 (sustained 40% decline in eGFR) and DAPA-CKD (sustained 50% decline in eGFR).
d
Except for the EMPA-REG OUTCOME trial, which did not include sustained eGFR < 15 mL/min/1.73 m2.

AJKD Vol 77 | Iss 6 | June 2021 977


(sustained eGFR decline ≥ 50% from baseline, KFRT, sus-
tained eGFR < 15 mL/min/1.73 m2, or death from kidney
or cardiovascular cause), with an HR of 0.61 (95% CI,
0.51-0.72). The benefits of dapagliflozin were consistent
in participants with and without type 2 diabetes. Safety
profiles were similar between the 2 treatment arms with
the exception of volume depletion (more common with
dapagliflozin) and major hypoglycemia (more common
with placebo).
Several mechanisms have been proposed for how SGLT2
inhibitors improve kidney outcomes. Blockade of SGLT2,
responsible for w90% of glucose reabsorption that
occurs in the proximal tubule, promotes urinary excretion
of glucose. However, SGLT2 inhibitors are associated with
only modest HbA1c reductions, suggesting that the kidney
effects are not driven by improved glycemic control.
Rather, increased distal delivery of sodium to the macula
densa (in tandem with glucose excretion) activates tubu-
loglomerular feedback, leading to vasoconstriction of the
afferent arteriole and ultimately a reduction in intra-
glomerular pressure. SGLT2 inhibitors also reduce systolic
and diastolic BP, likely due to osmotic diuresis (from
glucosuria), natriuresis, and possibly inhibition of sym-
pathetic nervous system activity. Other potential mecha-
nisms for the renoprotective effects of SGLT2 inhibitors
include weight loss, lowering of serum uric acid levels,
and reduction of albuminuria.
Although SGLT2 inhibitors are generally well tolerated,
some safety concerns warrant mentioning. Genitourinary
fungal infections, particularly in women, are the most
commonly reported adverse effect. Fournier gangrene,
which occurs much more rarely, is another potential
severe complication. The US Food and Drug Admin-
istration (FDA) issued a black box warning on canagli-
flozin regarding an increased risk of lower limb
amputations based on a nearly 2-fold higher risk of
amputations in the CANVAS Program. In contrast, there
was no heightened amputation risk in CREDENCE.
Still, it is prudent to perform regular foot examinations
in all patients on SGLT2 inhibitors, particularly those
with a history of neuropathy, peripheral vascular dis-
ease, and/or diabetic foot ulcers. An increased risk of
fracture with SGLT2 inhibitors was reported in the
CANVAS Program but not CREDENCE, EMPA-REG
OUTCOME, or DECLARE-TIMI 58. The role of SGLT2
inhibitors as a precipitant for acute kidney injury
remains controversial, with some published studies
reporting protective effects.
For Question 6, SGLT2 inhibitors are associated with (a) a
reduction in albuminuria and a 30% to 40% decreased risk of
CKD progression. This class of medications is not commonly
associated with hyponatremia or hyperkalemia and results in
only a small reduction in HbA1c.
For Question 7, although all of the choices have been
reported with the use of SGLT2 inhibitors, the most
common is (a) genitourinary fungal infection.
978
Chen et al
Additional Readings
➢ Davies MJ, D’Alessio DA, Fradkin J, et al. Management of
hyperglycaemia in type 2 diabetes, 2018. A consensus by the
American Diabetes Association (ADA) and the European
Association for the Study of Diabetes (EASD). Diabetologia.
2018;61:2461-2498.
➢ Heerspink HJL, Stef ansson BV, Correa-Rotter R, et al. Dapagli-
flozin in patients with chronic kidney disease. N Engl J Med.
2020;383(15):1436-1446.
➢ Neal B, Perkovic V, Mahaffey KW, et al. Canagliflozin and car-
diovascular and renal events in type 2 diabetes. N Engl J Med.
2017;377(7): 644-657.
➢ Neuen, BL, Young T, Heerspink HJL, et al. SGLT2 inhibitors for
the prevention of kidney failure in patients with type 2 diabetes: a
systematic review and meta-analysis. Lancet Diabetes Endo-
crinol. 2019;7(11):845-854. +ESSENTIAL READING
➢ Perkovic V, Jardine MJ, Neal B, et al. Canagliflozin and renal
outcomes in type 2 diabetes and nephropathy. N Engl J Med.
2019;380(24):2295-2306. +ESSENTIAL READING
➢ Van Bommel EJM, Muskiet MHA, Tonneijck L, et al. SGLT2
inhibition in the diabetic kidney—from mechanisms to clinical
outcome. Clin J Am Soc Nephrol. 2017;12(4):700-710.
GLP-1 Receptor Agonists
Case 6: A 60-year-old man with type 2 diabetes mellitus and
CKD G4A3 (eGFR 27 mL/min/1.73 m2) has stable UACR
(1,800 mg/g for 1 year). His BP is 126/70 mm Hg and
HbA1c is 9.0%. His medications include lisinopril at 40 mg
daily, diltiazem sustained release at 180 mg daily, and insulin
glargine.
Question 8: Which one of the following interventions
would be appropriate to manage risk factors asso-
ciated with CKD progression?
a) Addition of a GLP-1 receptor agonist to reduce HbA1c
but at an increased risk for hypoglycemia
b) Addition of metformin rather than a GLP-1 receptor ago-
nist due to the favorable side-effect profile of metformin
c) Addition of a GLP-1 receptor agonist to slow progression
to KFRT without change in albuminuria
d) Addition of a GLP-1 receptor agonist to reduce BP
to <120/80 mm Hg
For the answer to the question, see the following text.
The GLP-1 receptor agonists are another novel class of
diabetes medications that improve kidney outcomes. In the
AWARD-7 trial, dulaglutide 0.75 mg or 1.5 mg weekly
resulted in slower eGFR decline over 52 weeks compared
with daily insulin glargine among participants with type 2
diabetes, CKD G3-G4, and a UACR > 300 mg/g. Fur-
thermore, UACR reduction occurred with dulaglutide in a
dose-dependent manner. Among participants with a
baseline UACR of ≤300 mg/g, no significant differences in
eGFR decline were observed between the 3 treatment
arms. In a meta-analysis of 5 cardiovascular outcome trials
(ELIXA, LEADER, SUSTAIN-6, EXSCEL, and REWIND),
Kristensen et al reported that GLP-1 receptor agonists were
associated with a 17% lower risk of a composite kidney
AJKD Vol 77 | Iss 6 | June 2021
Chen et al
outcome (new-onset UACR of >300 mg/g, a doubling of
serum creatinine, and a ≥40% decline in eGFR, KFRT, or
death from kidney disease), with an HR of 0.83 (95% CI,
0.78-0.89). However, when considering the more
restrictive outcome of worsening kidney function, defined
by a doubling of serum creatinine or a ≥40% eGFR decline
(except for EXSCEL, which also included KFRT or death
from kidney disease), there was no statistically significant
protective effect of GLP-1 receptor agonists (HR, 0.87
[95% CI, 0.73-1.03]).
GLP-1 receptor agonists act by binding to the GLP-1
receptor, enhancing glucose-dependent insulin secretion,
delaying gastric emptying, and decreasing appetite. Modest
improvements in body weight, BP, and lipid parameters
have also been reported. Prior studies suggest that multiple
cell types (eg, glomerular, tubular, and vascular) within
the kidney have GLP-1 receptors, but the mechanisms by
which GLP-1 receptor agonists improve kidney outcomes
are less clear. Altered renal hemodynamics, increased
natriuresis, and reductions in inflammation and reactive
oxidative species have all been proposed.
The most frequent side effect of GLP-1 receptor agonists
is nausea, which usually resolves after 4 to 8 weeks of
continued therapy. Diarrhea, hypoglycemia (particularly if
used in combination with insulin therapy), tachycardia,
gallbladder disease, pancreatitis, and retinopathy may also
occur. Other major safety concerns include increased risks
of pancreatic and thyroid cancer. Although Kristensen et al
did not find an association of GLP-1 receptor agonist
therapy with severe hypoglycemia, pancreatitis, or pan-
creatic or thyroid cancer, the trials excluded individuals
with a personal or family history of medullary thyroid
Figure 2. Proposed algorithm for SGLT2 inhibitor and GLP-1 receptor agonist use in chronic kidney disease. Metabolic risks factors
include uncontrolled diabetes or obesity/weight gain. Abbreviations: ASCVD, atherosclerotic cardiovascular disease; eGFR, esti-
mated glomerular filtration rate; GLP-1, glucagon-like peptide 1; HF, heart failure; SGLT2, sodium/glucose cotransporter 2;
UACR, urinary albumin-creatinine ratio. Based on information in Li et al 2020 (Clin J Am Soc Nephrol, https://doi.org/10.2215/
CJN.02690320).
AJKD Vol 77 | Iss 6 | June 2021
carcinoma or multiple endocrine neoplasia type 2, and the
FDA label warns against the use of GLP-1 receptor agonists
in these patients.
Although no trial has directly compared SGLT2 inhib-
itors with GLP-1 receptor agonists, a meta-analysis of 8
cardiovascular trials found a 38% (HR, 0.62 [95% CI,
0.58-0.67]) and 18% (HR, 0.82 [95% CI, 0.75-0.89])
lower risk of new-onset UACR > 300 mg/g, doubling of
serum creatinine, a ≥40% decline in eGFR, KFRT, or death
from kidney disease for SGLT2 inhibitors and GLP-1
receptor agonists, respectively. When an incident UACR
of >300 mg/g was not included in the outcome, SGLT2
inhibitors were associated with a 45% lower risk (HR,
0.55 [95% CI, 0.48-0.64]) whereas no association was
observed for GLP-1 receptor agonists (HR, 0.92 [95% CI,
0.80-1.06]). Thus, SGLT2 inhibitors appear to be more
effective in slowing kidney disease progression and should
be considered before GLP-1 receptor agonists (Fig 2).
For Question 8, studies best support (a) a reduction in
UACR after addition of a GLP-1 receptor agonist. There is
an increased risk of hypoglycemia when used concurrently
with insulin, and a reduction in the rate of progression to
KFRT has not been shown. While GLP-1 receptor agonists
may result in a small reduction in BP, lowering to <120/
80 mm Hg has not been shown to slow CKD progression.
Initiating metformin would be inappropriate at this eGFR.
Additional Readings
➢ Kristensen S, Rørth R, Jhund PS, et al. Cardiovascular, mortality,
and kidney outcomes with GLP-1 receptor agonists in patients
with type 2 diabetes: a systematic review and meta-analysis of
cardiovascular outcome trials. Lancet Diabetes Endocrinol.
2019;7(10):776-785. +ESSENTIAL READING
979

➢ Li J, Albajrami O, Zhuo et al. Decision algorithm for prescribing
SGLT2 inhibitors and GLP-1 receptor agonists for diabetic kid-
ney disease. Clin J Am Soc Nephrol. 2020;15(11):1678-1688.
➢ Muskiet MHA, Tonneijck L, Smits MM, et al. GLP-1 and the kid-
ney: from physiology to pharmacology and outcomes in diabetes.
Nat Rev Nephrol. 2017;13:605-628.
➢ Tuttle KR, Lakshmanan MC, Rayner B, et al. Dulaglutide versus
insulin glargine in patients with type 2 diabetes and moderate-to-
severe chronic kidney disease (AWARD-7): a multicenter, open-
label, randomized trial. Lancet Diabetes Endocrinol.
2018;6(8):605-617.
➢ Zelniker TA, Wiviott SD, Raz I, et al. Comparison of the effects of
glucagon-like peptide receptor agonists and sodium-glucose
cotransporter 2 inhibitors for prevention of major adverse car-
diovascular and renal outcomes in type 2 diabetes mellitus.
Circulation. 2019;139(17):2022-2031. +ESSENTIAL READING
Chronic Metabolic Acidosis and Dietary Protein
Restriction
Case 7: A 63-year-old woman with CKD G4A2 and
osteopenia returns for a follow-up visit, having been last seen
4 months ago. She underwent left nephrectomy 30 years
ago after trauma. She reports no interval symptoms, and her
weight has been stable. Her eGFR has slowly declined from
33 to 28 mL/min/1.73 m2 over the past 2 years, with her last
2 total carbon dioxide values in the range of 19 to 21 mmol/
L. On physical examination, her BP is 118/65 mm Hg, her
lungs are clear, and she has trace pedal edema.
Question 9: Which one of the following is most accu-
rate in treating metabolic acidosis associated with
CKD?
a) Modest dietary protein restriction should decrease urine
ammoniagenesis
b) Dietary supplementation with sodium bicarbonate should
decrease bone mineral density
c) Modest dietary protein restriction should increase skeletal
muscle catabolism
Case 8: A 58-year-old man with IgA nephropathy has had
progressive CKD over the past 18 months. His eGFR is
29 mL/min/1.73 m2 with total carbon dioxide ranging
between 18 and 20 mmol/L.
Question 10: Which one of the following interventions
has the greatest efficacy in improving metabolic
acidosis?
a) Increase daily fruit intake to 4 servings
b) Add sodium bicarbonate as a 650-mg tablet once daily
c) Ensure 2 servings of pasta daily
d) Replace one serving of red meat with one serving of fish
daily
For the answers to the questions, see the following text.
Metabolic acidosis is a common complication of CKD
due to impairments in the kidney’s ability to excrete acid.
Dietary composition also influences acid-base balance,
with animal-derived proteins contributing primarily
hydrogen ions, and fruits and vegetables contributing
alkali. Thus, treatment of metabolic acidosis in patients
980
Chen et al
with CKD typically relies on 3 strategies: reduction of
dietary animal protein, increased consumption of fruits
and vegetables, and administration of oral alkali salts.
Metabolic acidosis is a risk factor for KFRT, decreased
bone mineralization, and sarcopenia. Correction of met-
abolic acidosis may slow CKD progression. A systematic
review of 13 small, primarily open-label clinical trials
suggested that both oral alkali supplementation and
dietary interventions slow GFR decline, with a meta-
analyzed effect on mean GFR decline of >3 mL/min/
1.73 m2 per year for both strategies. However, only 4 of
the 13 studies had durations more than 1 year, and 4 had
durations of 6 months or less. The largest randomized con-
trolled trial of dietary protein restriction to date (MDRD
Study) suggested a more modest effect size that failed to
demonstrate statistical significance. Among the participants
randomized to usual-protein, low-protein, or very-low-
protein diet (1.3 vs 0.58 vs 0.28 grams per kilogram of
body weight per day), the difference in mean GFR decline
over a mean follow-up of 2.2 years was 0.8 mL/min per
year for very-low-protein versus low-protein (P = 0.07)
and 0.4 mL/min per year for the low-protein versus usual-
protein diets (P = 0.30).
Despite little clinical trial evidence, the KDIGO guide-
lines suggest consideration of dietary protein restriction
to < 1.3 grams per kilogram of body weight per day for
patients with or at risk for CKD G3 and 0.8 grams per
kilogram of body weight per day for patients with CKD
G4-G5, given the theoretical benefit. Patients with CKD
and bicarbonate of <22 mmol/L should also be treated
with oral alkali therapy to maintain their bicarbonate
concentrations in the normal range, recognizing the risks
of increased BP and edema. Diets enriched in fruits and
vegetables may provide as much or more alkali than
bicarbonate supplementation and, in one small study, were
similarly effective in slowing eGFR decline (Fig 3).
Returning to Question 9, chronic metabolic acidosis is
associated with progression of CKD, stimulates increased
renal ammoniagenesis, increases bone resorption, and is
associated with the development of sarcopenia. Therefore,
the best answer is (a) as reducing protein intake will
decrease dietary acid production. For Question 10, the best
answer is (a) since 4 servings of fruits and vegetables
provide more alkali than low-dose sodium bicarbonate.
Carbohydrates and animal meats contribute to net acid
production, although replacing servings of red meat with
fish may have other health benefits.
Additional Readings
➢ Banerjee T, Crews DC, Wesson DE, et al. High dietary acid load
predicts ESRD among adults with CKD. J Am Soc Nephrol.
2015;26(7):1693-1700.
➢ Goraya N, Simoni J, Jo CH, Wesson DE. Treatment of metabolic
acidosis in patients with stage 3 chronic kidney disease with
fruits and vegetables or oral bicarbonate reduces urine angio-
tensinogen and preserves glomerular filtration rate. Kidney Int.
2014;86(5):1031-1038.
AJKD Vol 77 | Iss 6 | June 2021
Chen et al

Figure 3. Mean (± SE) estimated glomerular filtration rates among patients with CKD G3 randomized to usual care, sodium bicar-
bonate supplementation, or base-producing fruits and vegetables. Abbreviations: CKD, chronic kidney disease; crGFR, plasma
creatinine-based glomerular filtration rate; cysGFR, plasma cystatin C-based glomerular filtration rate; F+V, fruits and vegetables;
HCO3, sodium bicarbonate supplementation. Reproduced from Goraya et al 2014 (Kidney Int, https://doi.org/10.1038/ki.2
014.83) with permission of the copyright holder. Original graphic © 2014 International Society of Nephrology.

➢ Jain N, Reilly RF. Effects of dietary interventions on the incidence b) Amiodarone

and progression of CKD. Nat Rev Nephrol. 2014;10(12):712-
724.
➢ Kidney Disease: Improving Global Outcomes (KDIGO) CKD
Work Group. KDIGO 2012 clinical practice guideline for the
evaluation and management of chronic kidney disease. Kidney Int
Suppl. 2013;3:1-150. +ESSENTIAL READING
➢ Klahr S, Levey AS, Beck GJ, et al. The effects of dietary protein
restriction and blood pressure control on the progression of
chronic renal disease. New Engl J Med. 1994;330(13):877-884.
+ESSENTIAL READING
➢ Navaneethan SD, Shao J, Buysse J, Bushinsky DA. Effects of
treatment of metabolic acidosis in CKD: a systematic review and
meta-analysis. Clin J Am Nephrol. 2019;14(7):1011-1020.
➢ Raphael KL. Metabolic acidosis in CKD: Core Curriculum 2019.
Am J Kidney Dis. 2019;74(2):263-275.
Avoidance of Nephrotoxins
Case 9: A 62-year-old woman with recently diagnosed
ovarian cancer has presented to the emergency department
with a urinary tract infection and atrial fibrillation. Her medical
history is notable for CKD G4A1 in the setting of hyper-
tension and chronic hepatitis B. Her home medications
include lisinopril, tenofovir disoproxil fumarate, and multi-
vitamin. Computed tomography imaging of her abdomen/
pelvis demonstrates a large ovarian mass with peritoneal
carcinomatosis.
Question 11: Which one of the following medications
is safest to use in the setting of CKD G4?
a) Gentamicin
c) Tenofovir disoproxil fumarate
d) Cisplatin
Case 10: A 74-year-old man with CKD G4A2 in the setting
of diabetes mellitus and IgA nephropathy is hospitalized for a
nonhealing lower extremity ulceration. His eGFR is currently
23 mL/min/1.73 m2, as compared with 26 mL/min/1.73 m2
1 month before. You are consulted before the planned lower
extremity angiography for recommendations to reduce the
risk for contrast-associated acute kidney injury.
Question 12: Which one of the following interventions
is most appropriate before administration of intra-
arterial intravenous iodinated contrast in hospi-
talized patients with diabetes and CKD?
a) Oral N-acetylcysteine
b) 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA)
reductase inhibitor
c) Vitamin C
d) Normal saline hydration
For the answers to the questions, see the following text.
Nephrotoxins can contribute to CKD progression by
causing acute kidney injury, chronic interstitial nephritis,
tubular dysfunction, or glomerular changes. Avoidance of
nephrotoxins is not always possible, especially in the
hospital or acute care setting; thus, an individualized
approach that carefully weighs the risks versus benefits for

AJKD Vol 77 | Iss 6 | June 2021 981


each patient is necessary. Many chemotherapeutic (eg,
platinum-based agents, gemcitabine, immunotherapies)
and antimicrobial (eg, aminoglycosides, colistin, ampho-
tericin B, tenofovir disoproxil fumarate) agents require
special attention in CKD given their potential for harm to
the kidney and/or need for dose adjustments. Despite
known toxicities, alternative drug options may not be
appropriate due to susceptibility patterns or decreased
efficacy. In such cases, counseling patients on the potential
worsening of CKD, close monitoring of kidney function,
and dose adjustments as needed is a reasonable approach.
Other potential nephrotoxins include gastrointestinal
agents (eg, phosphate-containing bowel preparations,
proton-pump inhibitors), pain relievers (eg, nonsteroidal
anti-inflammatory agents), and herbal supplements or
remedies. Although proton-pump inhibitors do not nec-
essarily need to be stopped in patients with CKD, providers
should review need and candidacy for alternative therapy
(eg, H2 blockers) regularly.
Contrast-associated acute kidney injury remains a con-
cern among patients with CKD, particularly those with
eGFR < 30 mL/min/1.73 m2 or diabetes. Proposed
mechanisms of injury include vasoconstriction leading
to renal ischemia, direct tubular toxicity, and oxidative
stress from free radical generation. High-osmolar
(>1,200 mOsm/kg) ionic contrast agents are more
likely to be nephrotoxic than low-osmolar (700-
850 mOsm/kg) or iso-osmolar (w290 mOsm/kg) non-
ionic agents. The risk for acute kidney injury is thought
to be higher with arterial compared with venous
contrast administration. However, patients with CKD
should not be denied necessary tests that require contrast
for diagnosis and management. Fundamental risk reduc-
tion measures include (1) use of minimum dose of con-
trast necessary; (2) use of low-osmolar or iso-osmolar
agents; (3) expansion of intravascular volume as
tolerated with intravenous normal saline before, during,
and after the procedure; and 4) avoidance of concurrent
nephrotoxins.
Returning to Question 11, the best answer is (b)
because amiodarone does not require discontinuation or
dose adjustment in CKD G4. For Question 12, although all
the listed agents have been reported to reduce the risk of
contrast-associated kidney injury, the best answer is (d)
because periprocedural hydration with normal saline is the
most widely accepted prophylaxis.
Additional Readings
➢ Chen TK, Knicely DH, Grams ME. Chronic kidney disease
diagnosis and management: a review. JAMA.
2019;322(13):1294-1304. +ESSENTIAL READING
➢ McCullough PA, Choi JP, Feghali GA, et al. Contrast-induced
acute kidney injury. J Am Coll Cardiol. 2016;68(13):1465-1473.
➢ Perazella MA. Pharmacology behind common drug nephrotox-
icities. Clin J Am Soc Nephrol. 2018;13(12):1897-1908.
982
Chen et al
Uric Acid–Lowering Therapies
Prior studies have reported an association between elevated
serum uric acid levels and increased risk of CKD pro-
gression. Whether uric acid directly causes CKD pro-
gression or is an indirect marker of some other process is
unclear. Two recent trials, PERL and CKD-FIX, investigated
whether treatment with allopurinol slowed eGFR decline.
The PERL study enrolled patients with type 1 diabetes and
early diabetic kidney disease (mean GFR 68 mL/min/
1.73 m2 and median urine albumin excretion rate of
60 mg/d) and showed no difference in GFR slope over 3
years between the allopurinol and placebo groups. The
CKD-FIX study, which included individuals with more
advanced CKD (mean eGFR 32 mL/min/1.73 m2 and
median UACR 717 mg/g), also reported no significant
difference in eGFR change between allopurinol and pla-
cebo over 2 years (−3.33 and −3.23 mL/min/1.73 m2 per
year; mean difference, −0.10 [95% CI, −1.18 to 0.97]
mL/min/1.73 m2 per year). Of note, baseline uric acid
levels in both trials were not markedly elevated (PERL:
6.1 mg/dL; CKD-FIX: 8.2 mg/dL). Another randomized
clinical trial of febuxostat in patients with CKD G3 and
asymptomatic hyperuricemia (mean uric acid w7.8 mg/
dL) similarly found no difference in eGFR slopes compared
with placebo.
Additional Readings
➢ Badve SV, Pascoe EM, Tiku A, et al. for the CKD-FIX
Study Investigators. Effects of allopurinol on the progression of
chronic kidney disease. N Engl J Med. 2020;382(26):2504-
2513.
➢ Doria A, Galecki AT, Spino C, et al. for the PERL Study Group.
Serum urate lowering with allopurinol and kidney function in type
1 diabetes. N Engl J Med. 2020;382(26):2493-2503.
➢ Kimura K, Hosoya T, Uchida S, et al. Febuxostat therapy for
patients with stage 3 CKD and asymptomatic hyperuricemia: a
randomized trial. Am J Kidney Dis. 2018;72(6):798-810.
Weight Loss and Bariatric Surgery
Case 11: A 54-year-old man has CKD G4A1 in the setting
of diabetes mellitus and hypertension. His weight has been
stable for the past year after losing 15 pounds. His medi-
cations include lisinopril at 40 mg daily and a GLP-1
receptor agonist. His BP is 125/70 mm Hg and BMI is
32 kg/m2. His laboratory tests demonstrate an eGFR of
27 mL/min/1.73 m2, a UACR of 25 mg/g, HbA1c 7.2%, and
serum uric acid of 8.1 mg/dL. The kidney failure risk equation
predicts a 4.6% risk of kidney failure at 2 years.
Question 13: Which one of the following interventions
would be most appropriate to reduce his risk of pro-
gression to kidney failure?
a) Referral for bariatric surgery
b) Starting allopurinol treatment
AJKD Vol 77 | Iss 6 | June 2021
Chen et al
c) Starting treatment with a second antihypertensive agent
to lower his BP to <120/80 mm Hg
d) No additional therapy
For the answer to the question, see the following text.
In observational studies, higher BMI has been asso-
ciated with substantially greater risk of developing
hypertension and diabetes and a more modest risk for
CKD. Mechanisms for the latter association may be
through the development of hypertension/diabetes, or
there may independent effects through inflammation and
hemodynamic alterations in the glomerulus. Weight loss
through lifestyle modification or bariatric surgery may
improve kidney outcomes.
A meta-analysis of 4 small studies suggested a benefit of
weight loss achieved through nonsurgical interventions on
reduction in albuminuria; however, only 2 of the studies
were clinical trials, with 40 and 18 participants each. A post
hoc analysis of the Look AHEAD trial of people with type 2
diabetes who were overweight or had obesity suggested a
31% reduction in risk of developing very-high-risk CKD
(defined as G4, G3bA2-3, or G3aA3) associated with
intensive lifestyle intervention, which aimed to reduce
caloric consumption and increase physical activity. Inter-
estingly, the effect of the intervention was only partially
mediated by weight loss and reductions in HbA1c and
systolic BP. A propensity-matched study of 985 patients
who underwent bariatric surgery compared with 985
controls with obesity suggested that long-term GFR decline
was attenuated in the bariatric surgery group, with a 57%
lower risk of doubling of serum creatinine, an eGFR <15
mL/min/1.73 m2, or KFRT over a median follow-up period
of 4 years. Thus, it appears that weight loss may confer
benefits for both GFR decline and worsening albuminuria,
although clinical trial evidence is lacking.
In Question 13, of the options shown, response (d), or no
additional therapy, would be the most appropriate inter-
vention. Bariatric surgery at this BMI or a BP goal of <125/
70 mm Hg have not been shown to slow the progression to
KFRT. And, as discussed in the previous section, randomized
controlled trials have not shown a benefit of uric
acid–lowering therapy in preserving kidney function.
AJKD Vol 77 | Iss 6 | June 2021
Additional Readings
➢ Chang AR, Chen Y, Still C, et al. Bariatric surgery is associated
with improvement in kidney outcomes. Kidney Int.
2016;90(1):164-171.
➢ Chang AR, Grams ME, Ballew SH, et al. Adiposity and risk
of decline in glomerular filtration rate: meta-analysis of
individual participant data in a global consortium. BMJ.
2019;364:k5301.
➢ Look AHEAD Research Group. Effect of a long-term behavioural
weight loss intervention on nephropathy in overweight or obese
adults with type 2 diabetes: a secondary analysis of the Look
AHEAD randomised clinical trial. Lancet Diabetes Endocrinol.
2014;2(10):801-809. +ESSENTIAL READING
➢ Navaneethan SD, Yehnart H, Moustarah F, et al. Weight loss
interventions in chronic kidney disease: a systematic review
and meta-analysis. Clin J Am Soc Nephrol. 2009;4(10):
1565–1574.
Article Information
Authors’ Full Names and Academic Degrees: Teresa K. Chen,
MD, MHS, Christopher J. Sperati, MD, MHS, Sumeska
Thavarajah, MD, and Morgan E. Grams, MD, PhD.
Authors’ Affiliations: Division of Nephrology, Department of
Medicine, School of Medicine, Johns Hopkins University, Baltimore,
MD (TKC, CJS, ST, MEG); Welch Center for Prevention,
Epidemiology, and Clinical Research, Johns Hopkins Medical
Institutions, Baltimore, MD (TKC, MEG).
Address for Correspondence: Teresa K. Chen, MD, MHS, 1830 E
Monument St, Suite 416, Baltimore, MD 21287. Email: tchen39@
jhmi.edu
Support: This Core Curriculum was produced without any direct
financial support. Dr Chen is supported by NIH/NIDDK grant
K08DK117068 and a George M. O’Brien Center for Kidney
Research Pilot and Feasibility Grant from Yale University (under
award number NIH/NIDDK P30DK079310). Dr Grams is
supported by NIH/NIDDK R01DK115534.
Financial Disclosure: Dr Sperati is a member of the American
Board of Internal Medicine (ABIM) Longitudinal Assessment
Committee for Nephrology. No ABIM exam questions are shared
or otherwise disclosed in this article. The other authors declare
that they have no relevant financial interests.
Peer Review: Received July 28, 2020, in response to an invitation
from the journal. Evaluated by 2 external peer reviewers and a
member of the Feature Advisory Board, with direct editorial input
from the Feature Editor and a Deputy Editor. Accepted in revised
form December 5, 2020.
983

Belimumab in Lupus Nephritis: New Trial Results
Arrive During an Exciting Time for Therapeutics
Nestor Oliva-Damaso and Andrew S. Bomback
A t least 50% of all patients with systemic lupus ery-
thematosus (SLE) will manifest kidney involvement,
or lupus nephritis (LN), in the course of the disease.1
Long-term outcomes in LN have clearly improved.
Before 1980, kidney survival at 5 years was as low as 20%,
whereas modern treatments have improved this rate
to 80%.2,3 Current standard induction and maintenance
Commentary on Furie R, Rovin BH, Houssiau F, et al. Two-
year, randomized, controlled trial of belimumab in lupus
nephritis. N Engl J Med. 2020;383(12):1117-1128.
therapy regimens in LN are based on cyclophosphamide-
azathioprine or mycophenolate mofetil (MMF) in induc-
tion and maintenance,4,5 high-dose glucocorticoid pulses
during induction followed by oral prednisone tapers,
and nonimmunomodulatory agents such as hydroxy-
chloroquine and angiotensin-converting enzyme in-
hibitors or angiotensin receptor blockers.6,7 Although
long-term outcomes have improved and the preferential
use of MMF over cyclophosphamide has significantly
reduced treatment-related adverse events, the incidence of
kidney failure within 15 years of LN diagnosis remains at
22% for all patients with LN and 44% for patients with
class IV LN.8 Therefore, the search for new therapies that
can reduce morbidity and mortality in patients with LN
continues.9
Belimumab is a human monoclonal antibody that in-
hibits the soluble form of a B-cell survival factor known as
BLyS or BAFF. BLyS levels are elevated in some patients
with SLE and may play a role in the pathogenesis of the
disease.10 The first successful randomized controlled trial
of belimumab in SLE was the BLISS-52,11 followed in short
order by a phase 3 clinical trial, BLISS-76, which used a
prolonged observation period.12 The early BLISS trials
excluded patients with severe active LN but a substantial
proportion of participants had renal involvement and
proteinuria at the time of inclusion. A post hoc analysis
reported that belimumab lowered rates of renal flares and
reduced proteinuria.13 However, the effect of belimumab
as an add-on standard-of-care therapy in patients with
active LN was not formally evaluated until the BLISS-LN
trial, the results of which have been recently published.14
What Does This Important Study Show?
BLISS-LN was a phase 3, multicenter, international, 104-
week, randomized, double-blind, placebo-controlled trial
of belimumab (at a dose of 10 mg per kilogram of body
weight) versus matching placebo on top of standard
therapy (cyclophosphamide-azathioprine or MMF and,
984
In the Literature Editorial
in most patients, glucocorticoids, hydroxychloroquine,
angiotensin-converting enzyme inhibitors, or angiotensin
receptor blockers) in adults with active LN (biopsy-proven
class III or IV with or without coexisting class V or pure
class V). The standard therapy was based on Euro-Lupus
Nephritis Trial4 and ALMS5 protocols. Randomization
was stratified according to induction regimen (cyclo-
phosphamide or MMF) and race group (Black or non-
Black). The primary end point at week 104 was deemed
a “primary efficacy renal response,” defined as urinary
protein-creatinine ratio (UPCR) ≤ 0.7 g/g, deterioration in
estimated glomerular filtration rate (eGFR) of no more
than 20% compared with pre-flare value or at least
60 mL/min/1.73 m2, and no use of rescue therapy. The
secondary end point was complete renal response at week
104, defined as UPCR < 0.5 g/g, eGFR that was no worse
than 10% below pre-flare value or ≥90 mL/min/1.73 m2,
and no rescue therapy.
At study completion, of 448 patients enrolled in the
study, 43% treated with adjuvant belimumab achieved the
primary efficacy renal response compared with 32%
treated with only standard therapy (odds ratio [OR], 1.6;
P = 0.03), and 30% with added belimumab versus 20% of
patients receiving placebo achieved a complete renal
response (OR, 1.7; P = 0.02).14 In addition, patients who
received belimumab had less kidney-related events or
death (hazard ratio, 0.51 [95% CI, 0.34-0.77]). Impor-
tantly, in subgroup analysis, the beneficial effect of beli-
mumab was only apparent in those using MMF as
induction and maintenance therapy and was not seen in
patients using cyclophosphamide for induction and
azathioprine for maintenance (the use of MMF vs cyclo-
phosphamide was not randomly assigned but rather left to
physician discretion). Adverse events occurred with similar
frequency in both the belimumab and placebo groups.
The primary efficacy renal response end point is unique
to this trial and was changed by investigators 5 years after
the trial began. The original primary end point used
traditional parameters for response (complete remission,
partial remission, or no response) at week 104 using
urinary sediment plus proteinuria and creatinine clearance
from a 24-hour urine collection. The new end point used
eGFR based on serum creatinine level, omitted the urinary
sediment component, and loosened the proteinuria crite-
rion. This change was made, according to the study’s au-
thors, to harmonize with accumulating evidence on
predictors of long-term kidney outcomes.14 Many trials
use eGFR rather than 24-hour creatinine clearance, and
urinary sediment results can be dependent on observer
technique and confounded by nonglomerular bleeding.15
AJKD Vol 77 | Iss 6 | June 2021
Oliva-Damaso and Bomback

Table 1. Recent Positive Lupus Nephritis Trials Using Add-On Therapy With Standard of Care
BLISS-LN14 NOBILITY17,18 AURORA20
Drug vs placebo added to Belimumab Obinutuzumab Voclosporin
standard therapy
Total no. of participants 448 125 357
Mean eGFR at enrollment, 100.5 ± 40.2 >90 91 ± 30
mL/min/1.73 m2
Follow-up time, wk 104 76 52
Primary end point Primary efficacy renal response: Complete renal response: Renal response:
• UPCR ≤ 0.7 g/g • UPCR < 0.5 g/g • UPCR ≤ 0.5 g/g
• <20% ↓ in eGFR vs pre-flare value • Scr ≤ ULN • eGFR ≥ 60 mL/min/1.73 m2
• eGFR ≥ 60 mL/min/1.73 m2 • Scr ≤ 115% of pt BL • <20% ↓ in eGFR from BL
• No use of rescue therapy • <10 RBC/HPF without • No use of rescue therapy
RBC casts
Intervention vs placebo
Primary outcome +12.2%a +22% +18.3%
All adverse events +2% −2% +2.2%
Serious adverse events −4% −7% −0.5%
Infectious adverse events −2% +6% −1.1%
Abbreviations: BL, baseline; eGFR, estimated glomerular filtration rate; HPF, high-power field; pt, patient; RBC, red blood cell; ULN, upper limit of normal in laboratory;
UPCR, urinary protein-creatinine ratio.
a
Kidney outcomes of BLISS-LN section from participants treated with mycophenolate mofetil–based standard therapy (n = 329) and excluding cyclophosphamide-treated
patients.

Still, the novel proteinuria threshold of ≤ 0.7 g/g by spot
ratio used in this study may lead clinicians to put more
faith in the secondary end point of complete remission,
which showed similar superiority for belimumab versus
placebo as the primary end point, albeit at significantly
lower rates.
How Does This Study Compare With Other
Studies?
Remission rates with MMF in LN have been evaluated in
prospective trials. The best data remain those from ALMS,
which compared induction therapy with MMF versus
cyclophosphamide in 370 patients with LN. At week 24,
response rates with MMF were 56.2%. In the BLISS-LN
trial, which used different criteria than the ALMS defini-
tion of treatment response, patients with MMF-based
standard therapy (ie, randomly assigned to the placebo
arm) had a 34% response rate that augmented to 46% with
the addition of belimumab. One of the most notable
findings in ALMS was the significantly different response
rates with MMF (60.4%) versus cyclophosphamide
(38.5%) in patients grouped as “other,” which mostly
comprised Black and mixed-race participants. The BLISS-
LN trial enrolled only 63 Black participants (14% of total
study population), but nonetheless reported that Black
participants who received belimumab appeared more
likely to respond (both primary efficacy renal response and
complete renal response) than those receiving placebo.
However, in both the active and placebo arms, the per-
centage of Black participants who responded was lower
than response rates in the overall study population.
Although response rates overall and in some subgroups
were lower than might be expected in BLISS-LN, the
benefit of adding a novel agent on top of standardized
therapy has until recently not been an effective strategy in
LN trials, marking the BLISS-LN study as a unique “posi-
tive” trial in this space. For example, recent studies looking
at the addition of rituximab, abatacept, and laquinimod to
standard-of-care induction therapy have all failed to show
any benefit compared with placebo.16
However, the last year has produced positive results
in 2 LN trials besides BLISS-LN (Table 1). The
NOBILITY trial, a phase 2 randomized controlled trial
comparing the humanized type II anti-CD20 monoclonal
antibody obinutuzumab (n = 63) versus placebo
(n = 62) added to background corticosteroids and MMF,
reported that complete (P = 0.007) and overall renal
response (complete plus partial remissions, P = 0.02)
were both enhanced by 22% at 76 weeks.17,18 Voclo-
sporin, a next-generation calcineurin inhibitor, showed
in the phase 2 AURA-LV study superior remission rates
when combined with MMF and steroids at weeks 24
and 48 compared with placebo,19 and results from the
much larger phase 3 study, AURORA, showed similar
benefits of adding voclosporin to a standard induction
regimen.20 At week 52, renal response, defined as
UPCR ≤ 0.5 g/g and eGFR ≥ 60 mL/min/1.73 m2 or no
more than 20% decrease from baseline eGFR, was
augmented in the voclosporin group by 18.3%
compared with placebo (40.8% vs 22.5%; P < 0.001).
What Are the Implications for Nephrologists?
At 2 years, participants who took belimumab on top of
standard induction and maintenance therapies for LN
achieved better outcomes than those who took standard
therapy alone. This benefit was primarily observed in
participants treated with MMF rather than cyclophospha-
mide as induction therapy. The safety profile of adding

AJKD Vol 77 | Iss 6 | June 2021 985


belimumab to standard therapies was tolerable. These
findings are a “win” for the LN community, both patients
with active disease and nephrologists and rheumatologists
who care for these patients, but still leave the most
important question unanswered. Namely, how will the
BLISS-LN results influence treatment in the clinic?
Will clinicians plan to give a 100-week regimen of beli-
mumab, on top of MMF-based induction therapy, right
from the start, essentially mirroring the protocol of BLISS-
LN? Or will belimumab be reserved only for patients with
LN who do not respond adequately to MMF alone? The data
from BLISS-LN only support the former utilization strategy
and not the latter. The much smaller CALIBRATE study
(n = 43) of patients with recurrent or refractory LN did not
show any benefit of adding belimumab to cyclophospha-
mide and rituximab.21 Because the use of MMF versus
cyclophosphamide was not randomly assigned in BLISS-LN
but rather left to physician discretion, the apparent non-
benefit of belimumab in the cyclophosphamide subgroup
may be due to these participants having more severe forms
of LN, including longer disease duration and greater expo-
sure to prior treatments, than those prescribed MMF. With
Black participants receiving belimumab more likely to
respond than those given placebo, should belimumab be
considered a particularly promising therapy for this high-
risk patient population? The low number of Black partici-
pants in the BLISS-LN study (n = 63) makes such speculation
difficult. More robust data to address this question are ex-
pected from the EMBRACE study (ClinicalTrials.gov identi-
fier NCT01632241), which has enrolled more than 500
participants.
Finally, for clinicians who believe that add-on therapy is
required for their patients with LN, whether at the onset of
treatment or after treatment resistance emerges, how will
belimumab distinguish itself from obinutuzumab, voclo-
sporin, or any other new therapies that emerge with
positive results? The answers to these questions will only
come with time and specifically with dedicated studies of
how the drug is now used in real-world settings. The data
from the randomized controlled trial most likely are just
the first set of results to consume for those of us who treat
LN on a regular basis.
Article Information
Authors’ Full Names and Academic Degrees: Nestor Oliva-
Damaso, MD, and Andrew S. Bomback, MD, MPH.
Authors’ Affiliations: Division of Nephrology, Department of
Medicine, Hospital Costa del Sol, Marbella, Malaga, Spain (NO-D);
and Division of Nephrology, Department of Medicine, Columbia
University Irving Medical Center, New York, NY (ASB).
Address for Correspondence: Andrew S. Bomback, MD, MPH,
622 West 168th St, PH 4-124, New York, NY 10032. Email:
asb68@cumc.columbia.edu
Support: None.
Financial Disclosure: Dr Bomback was a co-investigator in the
BLISS-LN study at the Columbia University Medical Center
site. Dr Oliva-Damaso declares that he has no relevant financial
interest.
986
Oliva-Damaso and Bomback
Peer Review: Received October 7, 2020, in response to an
invitation from the journal. Direct editorial input from an Associate
Editor and a Deputy Editor. Accepted in revised form November
4, 2020.
Publication Information: © 2020 by the National Kidney Founda-
tion, Inc. Published online November 20, 2020 with doi 10.1053/
j.ajkd.2020.11.003
References
1. Danila MI, Pons-estel GJ, Zhang J, Vil
a LM, Reveille JD, Alarc
on GS.
Renal damage is the most important predictor of mortality within
the damage index: data from LUMINA LXIV, a multiethnic US
cohort. Rheumatology (Oxford). 2009;48(5):542-545.
2. Ward MM. Changes in the incidence of end-stage renal dis-
ease due to lupus nephritis, 1982-1995. Arch Intern Med.
2000;160(20):3136-3140.
3. Ward MM. Changes in the incidence of endstage renal disease
due to lupus nephritis in the United States, 1996-2004.
J Rheumatol. 2009;36(1):63-67.
4. Houssiau FA, Vasconcelos C, D’cruz D, et al. Immunosuppres-
sive therapy in lupus nephritis: the Euro-Lupus Nephritis Trial,
a randomized trial of low-dose versus high-dose intravenous
cyclophosphamide. Arthritis Rheum. 2002;46(8):2121-2131.
5. Dooley MA, Jayne D, Ginzler EM, et al. Mycophenolate versus
azathioprine as maintenance therapy for lupus nephritis. N Engl
J Med. 2011;365(20):1886-1895.
6. Houssiau FA, D’cruz D, Sangle S, et al. Azathioprine versus
mycophenolate mofetil for long-term immunosuppression in
lupus nephritis: results from the MAINTAIN Nephritis Trial. Ann
Rheum Dis. 2010;69(12):2083-2089.
7. Appel GB, Contreras G, Dooley MA, et al. Mycophenolate
mofetil versus cyclophosphamide for induction treatment of
lupus nephritis. J Am Soc Nephrol. 2009;20(5):1103-1112.
8. Tektonidou MG, Dasgupta A, Ward MM. Risk of end-stage
renal disease in patients with lupus nephritis, 1971-2015: a
systematic review and Bayesian meta-analysis. Arthritis Rheu-
matol. 2016;68(6):1432-1441.
9. Almaani S, Meara A, Rovin BH. Update on lupus nephritis. Clin
J Am Soc Nephrol. 2017;12(5):825-835.
10. Cancro MP, D’cruz DP, Khamashta MA. The role of B
lymphocyte stimulator (BLyS) in systemic lupus erythematosus.
J Clin Invest. 2009;119(5):1066-1073.
11. Navarra SV, Guzm an RM, Gallacher AE, et al. Efficacy and
safety of belimumab in patients with active systemic lupus er-
ythematosus: a randomised, placebo-controlled, phase 3 trial.
Lancet. 2011;377(9767):721-731.
12. Furie R, Petri M, Zamani O, et al. A phase III, randomized, placebo-
controlled study of belimumab, a monoclonal antibody that inhibits
B lymphocyte stimulator, in patients with systemic lupus erythe-
matosus. Arthritis Rheum. 2011;63(12):3918-3930.
13. Dooley MA, Houssiau F, Aranow C, et al. Effect of belimumab
treatment on renal outcomes: results from the phase 3 beli-
mumab clinical trials in patients with SLE. Lupus. 2013;22(1):
63-72.
14. Furie R, Rovin BH, Houssiau F, et al. Two-year, randomized,
controlled trial of belimumab in lupus nephritis. N Engl J Med.
2020;383(12):1117-1128.
15. Ward M, Tektonidou MG. Belimumab as add-on therapy in
lupus nephritis. N Engl J Med. 2020;383(12):1184-1185.
16. Narain S, Furie R. Update on clinical trials in systemic lupus
erythematosus. Curr Opin Rheumatol. 2016;28(5):477-487.
17. Furie R, Aroca G, Alvarez A, et al. A phase II randomized, double-
blind, placebo-controlled study to evaluate the efficacy and safety
of obinutuzumab or placebo in combination with mycophenolate
AJKD Vol 77 | Iss 6 | June 2021
Oliva-Damaso and Bomback
mofetil in patients with active class III or IV lupus nephritis [abstract].
Arthritis Rheumatol. 2019;71(suppl 10). Accessed November 5,
2020. https://acrabstracts.org/abstract/a-phase-ii-randomized-
double-blind-placebo-controlled-study-to-evaluate-the-efficacy-and-
safety-of-obinutuzumab-or-placebo-in-combination-with-mycophenolate-
mofetil-in-patients-with-active-class-iii/
18. Rovin B, Furie R, Aroca G, et al. B-Cell depletion and response
in a randomized, controlled trial of obinutuzumab for prolifera-
tive lupus nephritis [abstract]. Kidney Int Rep. 2020;5(3):
S352.
19. Rovin BH, Solomons N, Pendergraft WF, et al. A randomized,
controlled double-blind study comparing the efficacy and safety
AJKD Vol 77 | Iss 6 | June 2021
of dose-ranging voclosporin with placebo in achieving remis-
sion in patients with active lupus nephritis. Kidney Int.
2019;95(1):219-231.
20. Gibson KL, Parikh S, Saxena A, et al. AURORA phase 3 trial
demonstrates voclosporin statistical superiority over standard
of care in lupus nephritis (LN) [abstract]. Am J Kidney Dis.
2020;75(5):819.
21. Atisha-fregoso Y, Malkiel S, Harris KM, et al. CALI-
BRATE: a phase 2 randomized trial of rituximab plus
cyclophosphamide followed by belimumab for the treat-
ment of lupus nephritis. Arthritis Rheumatol. 2021;73(1):
121-131.
987

RESEARCH LETTERS
Temporal Relationship of Glycemia With
Cardiac Arrhythmias in Patients With Type
2 Diabetes and CKD
To the Editor:
Trials suggest that hypoglycemia contributes to higher
mortality in patients with type 2 diabetes mellitus
(T2DM).1 It is proposed that hypoglycemia-induced car-
diac arrhythmias may be one mechanism to explain this
relationship.2 Preclinical studies have shown that hypo-
glycemia leads to upregulation of sympathetic, para-
sympathetic, and inflammatory pathways,3,4 which
promote cardiac irritability. Definitive causal data in
humans are limited.
We demonstrated that hypoglycemia is common in
patients with chronic kidney disease (CKD) and T2DM5
and these patients have high rates of cardiac arrhyth-
mias.6 However, the association of glycemia with ar-
rhythmias in such patients remains poorly elucidated,
and even small relative differences in glucose levels may
be important. Continuous glucose monitoring (CGM)
and mobile cardiac telemetry enable characterization of
the temporality of glycemia with arrhythmia onset. In
this study, we explored the association of glycemia with
occurrence of cardiac arrhythmias in patients with CKD
and T2DM.
In the CANDY observational study of patients with CKD
with T2DM,5 38 participants underwent simultaneous
CGM and mobile cardiac telemetry for up to two 6-day
periods (detailed methods in Item S1).
Mean eGFR was 38 ± 13 (SD) mL/min/1.73 m2.
From CGM, the mean glucose level was 173 ± 42 mg/
dL, and the proportion of time below range (<70 mg/
dL) was 2.4% ± 2.9%, with lower glucose levels in the
morning (Table 1; Fig S1). Participants experiencing an
arrhythmia (n = 13 vs n = 25 who did not) were more
likely to have a history of CVD (69% vs 24%), had
similar rates of β-blocker use (62% vs 52%), had higher
HbA1c levels (8.2% vs 7.2%), and had higher GMI
(7.6% vs 7.2%).
We limited our analysis to the first arrhythmia event
during monitoring. In these analyses, arrhythmias
occurred more frequently in the morning (Fig S2).
There was no significant difference in mean glucose
levels at the time of an arrhythmia on “event” versus
“nonevent” days (Fig 1). However, mean glucose level
was lower during the 12 hours leading up to the event
compared with nonevent days, with a maximum dif-
ference 10 hours before the event of 26 (95% CI, 1-51)
mg/dL, though no mean glucose value was <70 mg/dL.
These differences in glucose levels before the arrhyth-
mias appeared to be greater when the arrhythmia
occurred between 10 PM and 10 AM (Fig S3). Further, the
pattern of lower mean glucose levels on event versus
988
Correspondence
nonevent days was most evident for atrial fibrillation
and ventricular arrhythmias (Fig S4).
Hypoglycemia (sustained glucose < 70 mg/dL) was
nominally more frequent during the 12 hours preceding
the event on event days versus nonevent days, though
this difference was not statistically significant (Table S1).
The glucose CV was lower in the 12 hours preceding
the event on event days versus nonevent days.
In summary, in this analysis of patients with T2DM and
CKD who underwent simultaneous CGM and mobile car-
diac telemetry, we found that glucose levels may be lower
during the 12 hours preceding an arrhythmic event
compared with the same period on a nonevent day within
the same participant. These data support the hypothesis
that lower relative blood glucose levels may promote
cardiac arrhythmias.
Mean blood glucose values were largely within the
normal glucose range without frank hypoglycemia, which
may suggest that smaller relative differences in glucose
levels may be important in risk for arrhythmias. Further,
we observed a long lag time between lower glucose levels
and onset of the arrhythmia. Possible explanations may be
delayed onset of proarrhythmogenic sympathetic, para-
sympathetic, and inflammatory responses in the absence of
frank hypoglycemia.
Clinical data evaluating the association of glycemia and
arrhythmia have been conflicting. In a study of 30 pa-
tients with T2DM who underwent 5 days of CGM and
ECG recording, patients with severe hypoglycemia had
a greater number of ventricular arrhythmias.7 This
finding was consistent in other analyses of patients with
T2DM.8,9 However, among 32 patients with T1DM
and no structural heart disease, no relationship between
hypoglycemia and cardiac arrhythmias were reported.10
Collectively, these studies are intriguing but interpreta-
tion is limited because the associations are susceptible
to confounding by a wide range of participant
characteristics.
Our study adds to the literature in several ways. First,
we studied patients with CKD, a population with a high
risk for cardiac arrhythmias. Further, unlike prior
studies, we performed within-person analyses, which
may reduce potential confounding and increase study
power. The data generated from our study are hypoth-
esis generating. Nonetheless, they suggest that blood
glucose levels hours before the onset of arrhythmic
events may be important and that comparisons within
individuals may offer a useful approach to test this
hypothesis in the future.
Nisha Bansal, Leila R. Zelnick, Irl B. Hirsch, Dace Trence,
Nazem Akoum, Ian H. de Boer
Supplementary Material
Supplementary File (PDF)
Figures S1-S4, Item S1, Table S1.
AJKD Vol 77 | Iss 6 | June 2021
Correspondence

Table 1. Characteristics of Patients With T2DM and CKD Table 1 (Cont'd). Characteristics of Patients With T2DM and
Included in the Study (N = 38) CKD Included in the Study (N = 38)
Value Value
Demographics Supraventricular 6 (6)
Age, y 67.9 ± 9.1 Ventricular 9 (5)
Male sex 25 (66%) Note: Entries are mean ± SD or n (%) or median [interquartile range], except as
noted.
Race/ethnicity Abbreviations: CV, coefficient of variation; CVD, cardiovascular disease.
White 32 (84%) a
Glucose Management Indicator (GMI) was calculated as: GMI (%) =
Black 4 (11%) 3.31 + 0.02392 × (mean glucose in mg/dL).
b
Values in parentheses are no. of unique participants; for atrial fibrillation, of the 4
Other 2 (5%) participants, 1 had a history of prior atrial fibrillation.
Hispanic ethnicity 6 (16%)
Health History
Current smoking 0 (0%) Article Information
History of CVD (myocardial infarction, 15 (39%) Authors’ Full Names and Academic Degrees: Nisha Bansal, MD,
heart failure, stroke, peripheral vascular MAS, Leila R. Zelnick, PhD, Irl B. Hirsch, MD, Dace Trence, MD,
disease) Nazem Akoum, MD, and Ian H. de Boer, MD.
History of atrial fibrillation 7 (18%) Authors’ Affiliation: Department of Medicine, University of
Duration of DM, y 20.1 ± 10.7 Washington, Seattle, WA.
Physical Characteristics Address for Correspondence: Nisha Bansal, MD, MAS, Kidney
BMI, kg/m2 34.0 ± 6.2 Research Institute, University of Washington, 908 Jefferson St, 3rd
Systolic blood pressure, mm Hg 133.5 ± 17.3 Fl, Seattle, WA 98104. Email: nbansal@nephrology.washington.edu
Diastolic blood pressure, mm Hg 70.0 ± 12.2 Authors’ Contributions: Research idea and study design: NB,
eGFR, mL/min/1.73 m2 38.3 ± 13.1 IHdB, IBH, DT, NA; data acquisition: NB, IHdB, IBH, DT, NA; data
Urinary albumin-creatinine ratio, mg/g 128.6 [33.1-600.7] analyses/interpretation: NB, IHdB, IBH, DT, NA; statistical
HbA1c, % 7.8 ± 1.5 analyses: LRZ; supervision: IHdB. Each author contributed
important intellectual content during manuscript drafting or revision
Serum albumin, mg/dL 3.6 ± 0.4
and agrees to be personally accountable for the individual’s own
Transferrin saturation, % 28.0 ± 11.1 contributions and to ensure that questions pertaining to the
Medication Use accuracy or integrity of any portion of the work, even one in which
Antihypertensive medications 36 (95%) the author was not directly involved, are appropriately investigated
ACEi/ARBs 30 (79%) and resolved, including with documentation in the literature if
appropriate.
β-Blockers 21 (55%)
Calcium channel blockers 21 (55%) Support: This work was supported by an American Diabetes
Association grant (PI: de Boer) and an unrestricted fund from the
DHP calcium channel blockers 19 (50%)
Northwest Kidney Centers. Equipment (including the SEEQ
Non-DHP calcium channel blockers 2 (5%) devices) and supplies for research were received by the institution
Diuretics 21 (55%) from Abbott and Medtronic for this study. The funders had no role
Statins 36 (95%) in study design; data collection, analysis, or reporting; or the
Lipid-lowering medications 36 (95%) decision to submit for publication.
Thyroid medications 4 (11%) Financial Disclosure: Dr de Boer reports consulting for
Any insulin 34 (89%) Boehringher-Ingelheim, Cyclerion Therapeutics, George Clinical,
Secretagogues 7 (18%) Goldfinch Bio, and Ironwood. Dr Hirsch reports receiving research
Other glucose-lowering agents 15 (39%)
CGM Metrics
170

Mean glucose, mg/dL 172.7 ± 42.7 Any Arrhythmia

Glucose Management Indicator, %a 7.4 ± 1.0

160
Glucose (mg/dL)

SD glucose, mg/dL 53.7 ± 14.1

CV glucose, % 31.2 ± 5.1
150

Proportion of time in glucose range 58.2% ± 23.0%

of 70-180 mg/dL
140

Proportion of time < 70 mg/dL 2.4% ± 2.9% Event days

Non−event days
Proportion of time < 54 mg/dL 0.4% ± 0.8%
130

* *
12 10 8 6 4 2 0
Proportion of time > 180 mg/dL 39.4% ± 24.2% Hours before arrhythmia event
Total No. of Cardiac Arrhythmia Episodesb
Any 33 (13) Figure 1. Mean glucose level in the 12 hours preceding the first
Atrial fibrillation 9 (4) arrhythmic event of any type during monitoring among patients
Conduction abnormalities 9 (3) with T2DM and CKD (within-person matching of “event” and
(Continued)
“nonevent” days).

AJKD Vol 77 | Iss 6 | June 2021 989


funds from Medtronic Diabetes and Insulet and consulting fees from
Abbott Diabetes Care, Bigfoot Biomedical, and Roche. Dr Trence is
a stockholder in Medtronic. The remaining authors declare that they
have no relevant financial interests.
Peer Review: Received May 8, 2020. Evaluated by 2 external peer
reviewers, with direct editorial input from a Statistics/Methods Editor
and an Associate Editor, who served as Acting Editor-in-Chief.
Accepted in revised form August 9, 2020. The involvement of an
Acting Editor-in-Chief was to comply with AJKD’s procedures for
potential conflicts of interest for editors, described in the
Information for Authors & Journal Policies.
Publication Information: © 2020 by the National Kidney Founda-
tion, Inc. Published online October 1, 2020 with doi 10.1053/
j.ajkd.2020.08.008
References
1. Gerstein HC, Miller ME, Byington RP, et al. Effects of intensive
glucose lowering in type 2 diabetes. N Engl J Med.
2008;358(24):2545-2559.
2. Lindstrom T, Jorfeldt L, Tegler L, Arnqvist HJ. Hypoglycaemia
and cardiac arrhythmias in patients with type 2 diabetes melli-
tus. Diabetic Med. 1992;9(6):536-541.
3. Reno CM, Bayles J, Huang Y, et al. Severe hypoglycemia-
induced fatal cardiac arrhythmias are mediated by the para-
sympathetic nervous system in rats. Diabetes. 2019;68(11):
2107-2119.
4. Reno CM, Daphna-Iken D, Chen YS, VanderWeele J, Jethi K,
Fisher SJ. Severe hypoglycemia-induced lethal cardiac ar-
rhythmias are mediated by sympathoadrenal activation. Dia-
betes. 2013;62(10):3570-3581.
5. Ahmad I, Zelnick LR, Batacchi Z, et al. Hypoglycemia in people
with type 2 diabetes and CKD. Clin J Am Soc Nephrol.
2019;14(6):844-853.
6. Akoum N, Zelnick LR, de Boer IH, et al. Rates of cardiac rhythm
abnormalities in patients with CKD and diabetes. Clin J Am
Soc Nephrol. 2019;14(4):549-556.
7. Stahn A, Pistrosch F, Ganz X, et al. Relationship between hy-
poglycemic episodes and ventricular arrhythmias in patients with
type 2 diabetes and cardiovascular diseases: silent hypoglyce-
mias and silent arrhythmias. Diabetes Care. 2014;37(2):516-
520.
8. Pistrosch F, Ganz X, Bornstein SR, Birkenfeld AL, Henkel E,
Hanefeld M. Risk of and risk factors for hypoglycemia and
associated arrhythmias in patients with type 2 diabetes and
cardiovascular disease: a cohort study under real-world con-
ditions. Acta Diabetol. 2015;52(5):889-895.
9. Chow E, Bernjak A, Williams S, et al. Risk of cardiac arrhythmias
during hypoglycemia in patients with type 2 diabetes and car-
diovascular risk. Diabetes. 2014;63(5):1738-1747.
10. Lainetti KR, Pimenta J, Vendramini MF. Can hypoglycemic ep-
isodes in type 1 diabetics trigger cardiac arrhythmias? Dia-
betes Res Clin Pract. 2019;158:107878.
Acute Kidney Injury After the CAR-T
Therapy Tisagenlecleucel
To the Editor:
Chimeric antigen receptor T-cell (CAR-T) therapies use
genetically engineered T cells to target tumor antigens.
After antigen recognition, CAR-T cells rapidly expand in
number and produce inflammatory cytokines leading to
990
Correspondence
cytokine release syndrome (CRS). CRS is a systemic in-
flammatory response, on a spectrum with macrophage
activation syndrome/hemophagocytic lymphohistiocytosis
(MAS/HLH), which can manifest with fever, tachycardia,
fatigue, and multiorgan system dysfunction, including
acute kidney injury (AKI). Two prior studies demonstrated
an AKI incidence of about 19%-30% after commercial
CAR-T.1,2 Most of these patients (>80%) received axi-
cabtagene ciloleucel (Yescarta), a CD19-targeting CAR-T
that has a CD28 costimulatory domain and is character-
ized by rapid T-cell expansion and robust inflammatory
cytokine secretion.1,2 However, use of tisagenlecleucel
(Kymriah), a CAR-T that targets the same epitope of CD19
but has a different costimulatory domain (4-1BB), is
increasing.3,4 Tisagenlecleucel is associated with a reduced
inflammatory profile and lower rates of toxicity, which
may lead to lower rates of CRS.4-6 Due to these important
differences, we hypothesized that AKI may be less com-
mon in patients receiving tisagenlecleucel.
We performed a retrospective review of adults 18 years
or older with diffuse large B-cell lymphoma (DLBCL)
treated with tisagenlecleucel at Massachusetts General
Hospital from January 2019 to April 2020. The primary
aim was to describe the incidence and clinical features of
AKI following tisagenlecleucel treatment. Baseline de-
mographics, laboratory data, and clinical outcomes were
obtained from electronic health records. AKI was defined
as a ≥1.5-fold increase in creatinine level from the
pre–CAR-T baseline (day 0). CRS was graded by the pa-
tient’s primary oncology team.7 We also collected baseline,
peak, and nadir values for electrolytes. The study was
approved by the Partners Healthcare Institutional Review
Board; the need for informed consent was waived.
Overall, 37 patients received tisagenlecleucel; mean age
was 60 years, 65% were men, and 86% were White. CRS
occurred in 20 patients (54%): 11 patients had grade 1; 4,
grade 2; and only 1, high-grade CRS (grade 5; Table 1).
Sixteen patients (42%) received immunosuppressive (IS)
therapy to treat CRS, including dexamethasone (35%),
tocilizumab (22%), and anakinra (22%). Electrolyte ab-
normalities including hyponatremia, hypokalemia, and
hypophosphatemia were common and occurred in the first
week after tisagenlecleucel treatment (Table 1).
Overall, 5 patients (14%) died within 30 days, all
attributable to disease progression and/or infectious
complications (Table 1). AKI occurred in only 2 patients
(5%), stage 3 for both. The first was an 80-year-old man
who had grade 2 CRS and neurotoxicity and aspirated the
day after the tisagenlecleucel infusion; septic shock led to
rapidly progressive nonoliguric AKI and death 4 days after
tisagenlecleucel treatment. The second was a 50-year-old
man with a new diagnosis of DLBCL. One week after
tisagenlecleucel treatment, he developed severe CRS with
clinical and laboratory findings of MAS/HLH-like syn-
drome in the setting of rhinocerebral mucormycosis
requiring hemi-maxillectomy, with fever, splenomegaly,
pancytopenia, coagulopathy, hyperferritinemia, and
AJKD Vol 77 | Iss 6 | June 2021

funds from Medtronic Diabetes and Insulet and consulting fees from
Abbott Diabetes Care, Bigfoot Biomedical, and Roche. Dr Trence is
a stockholder in Medtronic. The remaining authors declare that they
have no relevant financial interests.
Peer Review: Received May 8, 2020. Evaluated by 2 external peer
reviewers, with direct editorial input from a Statistics/Methods Editor
and an Associate Editor, who served as Acting Editor-in-Chief.
Accepted in revised form August 9, 2020. The involvement of an
Acting Editor-in-Chief was to comply with AJKD’s procedures for
potential conflicts of interest for editors, described in the
Information for Authors & Journal Policies.
Publication Information: © 2020 by the National Kidney Founda-
tion, Inc. Published online October 22, 2020 with doi 10.1053/
j.ajkd.2020.08.008
References
1. Gerstein HC, Miller ME, Byington RP, et al. Effects of intensive
glucose lowering in type 2 diabetes. N Engl J Med.
2008;358(24):2545-2559.
2. Lindstrom T, Jorfeldt L, Tegler L, Arnqvist HJ. Hypoglycaemia
and cardiac arrhythmias in patients with type 2 diabetes melli-
tus. Diabetic Med. 1992;9(6):536-541.
3. Reno CM, Bayles J, Huang Y, et al. Severe hypoglycemia-
induced fatal cardiac arrhythmias are mediated by the para-
sympathetic nervous system in rats. Diabetes. 2019;68(11):
2107-2119.
4. Reno CM, Daphna-Iken D, Chen YS, VanderWeele J, Jethi K,
Fisher SJ. Severe hypoglycemia-induced lethal cardiac ar-
rhythmias are mediated by sympathoadrenal activation. Dia-
betes. 2013;62(10):3570-3581.
5. Ahmad I, Zelnick LR, Batacchi Z, et al. Hypoglycemia in people
with type 2 diabetes and CKD. Clin J Am Soc Nephrol.
2019;14(6):844-853.
6. Akoum N, Zelnick LR, de Boer IH, et al. Rates of cardiac rhythm
abnormalities in patients with CKD and diabetes. Clin J Am
Soc Nephrol. 2019;14(4):549-556.
7. Stahn A, Pistrosch F, Ganz X, et al. Relationship between hy-
poglycemic episodes and ventricular arrhythmias in patients with
type 2 diabetes and cardiovascular diseases: silent hypoglyce-
mias and silent arrhythmias. Diabetes Care. 2014;37(2):516-
520.
8. Pistrosch F, Ganz X, Bornstein SR, Birkenfeld AL, Henkel E,
Hanefeld M. Risk of and risk factors for hypoglycemia and
associated arrhythmias in patients with type 2 diabetes and
cardiovascular disease: a cohort study under real-world con-
ditions. Acta Diabetol. 2015;52(5):889-895.
9. Chow E, Bernjak A, Williams S, et al. Risk of cardiac arrhythmias
during hypoglycemia in patients with type 2 diabetes and car-
diovascular risk. Diabetes. 2014;63(5):1738-1747.
10. Lainetti KR, Pimenta J, Vendramini MF. Can hypoglycemic ep-
isodes in type 1 diabetics trigger cardiac arrhythmias? Dia-
betes Res Clin Pract. 2019;158:107878.
Acute Kidney Injury After the CAR-T
Therapy Tisagenlecleucel
To the Editor:
Chimeric antigen receptor T-cell (CAR-T) therapies use
genetically engineered T cells to target tumor antigens.
After antigen recognition, CAR-T cells rapidly expand in
number and produce inflammatory cytokines leading to
990
Correspondence
cytokine release syndrome (CRS). CRS is a systemic in-
flammatory response, on a spectrum with macrophage
activation syndrome/hemophagocytic lymphohistiocytosis
(MAS/HLH), which can manifest with fever, tachycardia,
fatigue, and multiorgan system dysfunction, including
acute kidney injury (AKI). Two prior studies demonstrated
an AKI incidence of about 19%-30% after commercial
CAR-T.1,2 Most of these patients (>80%) received axi-
cabtagene ciloleucel (Yescarta), a CD19-targeting CAR-T
that has a CD28 costimulatory domain and is character-
ized by rapid T-cell expansion and robust inflammatory
cytokine secretion.1,2 However, use of tisagenlecleucel
(Kymriah), a CAR-T that targets the same epitope of CD19
but has a different costimulatory domain (4-1BB), is
increasing.3,4 Tisagenlecleucel is associated with a reduced
inflammatory profile and lower rates of toxicity, which
may lead to lower rates of CRS.4-6 Due to these important
differences, we hypothesized that AKI may be less com-
mon in patients receiving tisagenlecleucel.
We performed a retrospective review of adults 18 years
or older with diffuse large B-cell lymphoma (DLBCL)
treated with tisagenlecleucel at Massachusetts General
Hospital from January 2019 to April 2020. The primary
aim was to describe the incidence and clinical features of
AKI following tisagenlecleucel treatment. Baseline de-
mographics, laboratory data, and clinical outcomes were
obtained from electronic health records. AKI was defined
as a ≥1.5-fold increase in creatinine level from the
pre–CAR-T baseline (day 0). CRS was graded by the pa-
tient’s primary oncology team.7 We also collected baseline,
peak, and nadir values for electrolytes. The study was
approved by the Partners Healthcare Institutional Review
Board; the need for informed consent was waived.
Overall, 37 patients received tisagenlecleucel; mean age
was 60 years, 65% were men, and 86% were White. CRS
occurred in 20 patients (54%): 11 patients had grade 1; 4,
grade 2; and only 1, high-grade CRS (grade 5; Table 1).
Sixteen patients (42%) received immunosuppressive (IS)
therapy to treat CRS, including dexamethasone (35%),
tocilizumab (22%), and anakinra (22%). Electrolyte ab-
normalities including hyponatremia, hypokalemia, and
hypophosphatemia were common and occurred in the first
week after tisagenlecleucel treatment (Table 1).
Overall, 5 patients (14%) died within 30 days, all
attributable to disease progression and/or infectious
complications (Table 1). AKI occurred in only 2 patients
(5%), stage 3 for both. The first was an 80-year-old man
who had grade 2 CRS and neurotoxicity and aspirated the
day after the tisagenlecleucel infusion; septic shock led to
rapidly progressive nonoliguric AKI and death 4 days after
tisagenlecleucel treatment. The second was a 50-year-old
man with a new diagnosis of DLBCL. One week after
tisagenlecleucel treatment, he developed severe CRS with
clinical and laboratory findings of MAS/HLH-like syn-
drome in the setting of rhinocerebral mucormycosis
requiring hemi-maxillectomy, with fever, splenomegaly,
pancytopenia, coagulopathy, hyperferritinemia, and
AJKD Vol 77 | Iss 6 | June 2021
Correspondence

Table 1. Baseline Characteristics and Clinical Outcomes Table 2. Baseline and Laboratory Values in the Patient Who
Among Patients Receiving Tisagenlecleucel Developed MAS/HLH and Nephrotic Syndrome
Value Diagnosis of
Baseline characteristics Baseline MAS/HLH
Mean age ± SD, y 60 ± 18 Kidney function/markers of nephrotic syndromea
Male sex 24 (65%) Creatinine, mg/dL 0.54 1.36
White race 32 (86%) Serum urea nitrogen, mg/dL 19 44
Comorbid conditions Proteinuria Negative on 5.0 g/g
Diabetes 6 (16%) dipstick
Hypertension 9 (24%) Dipstick hematuria Negative 2+
Congestive heart failure 0 (0%) Albumin, g/dL 2.5 2.4
Cirrhosis 0 (0%) Triglycerides, mg/dL 370 629
CRS 20 (51%) Inflammatory markers
Grade 1 15 (41%) Ferritin, μg/L 5,010 298,610
Grade 2 4 (11%) C-Reactive protein, mg/L 138.7 9.3
Grade 3 or 4 0 (0%) Lactic acid dehydrogenase, U/L 274 17,642
Grade 5 1 (3%) Interleukin 2 receptor, pg/mL Not done 11,763
Baseline medication use Blood cell counts
Acyclovir 19 (51%) White blood cell count, ×103/μL 0.12 2.30
ACEi/ARB 5 (14%) Hemoglobin, g/dL 8.0 5.5
Bactrim 14 (38%) Platelets, ×103/μL 32 15
Proton pump inhibitor 16 (43%) Coagulation markers
H2 blocker 9 (22%) International normalized ratio 1.1 1.9
Statin 9 (24%) Partial thromboplastin time, s 27 48
Allopurinol 9 (24%) D-dimer, ng/mL Not done >10,000
a
Creatinine level continued to increase and was 1.82 mg/dL at death on day +28
Diuretic 3 (8%) after tisagenlecleucel treatment. Proteinuria test was repeated 6 days after diag-
Nonsteroidal anti-inflammatory drugs 2 (5%) nosis and was 6.0 g/g.
Use of IS to treat CRS or neurotoxicitya 16 (43%)
Dexamethasone 13 (35%)
Tocilizumab 8 (22%)
prior series, affecting ≥70% of patients.1 Hypo-
Anakinra 8 (22%)
phosphatemia likely results from rapid T-cell expansion
AKIa 2 (5%)
after infusion of either CAR-T product.
Electrolyte disorders
Hyponatremia: sodium < 130 mEq/L 7 (19%)
AKI after CAR-T is driven by cytokine-mediated vaso-
Hypokalemia: potassium < 3.0 mmol/L 5 (14%)
dilation and capillary leak leading to third spacing, intra-
Hypophosphatemia: phosphate < 2.0 mg/dL 26 (70%) vascular volume depletion, and reduced cardiac output,
30-d mortality 5 (14%) typically causing decreased kidney perfusion or acute
Abbreviations: ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin
tubular necrosis (ATN).8 However, in this series, we also
receptor blocker; SD, standard deviation.
a
identified the first case of new-onset proteinuria and AKI
Neither required dialysis.
occurring in a patient with CAR-T–triggered MAS/HLH.
MAS/HLH is a rare life-threatening complication of CAR-T
therapy that has been observed but not extensively re-
hypertriglyceridemia (Table 2). He developed nonoliguric ported.9 MAS/HLH is most commonly associated with
AKI and had spot protein-creatinine ratios > 5 g/g ATN, yet associated collapsing glomerulopathy has been
measured twice on nonconsecutive days. AKI progressed reported in other settings.9 Unfortunately, kidney biopsy
during administration of liposomal amphotericin B for was not performed in this case due to the patient’s critical
invasive mucormycosis and acyclovir for herpes simplex illness.
prophylaxis, and the patient died on day +28. Our study is limited by the relatively small cohort size
We found lower rates of AKI (5%) in patients receiving from a single center, as well as the retrospective ascer-
tisagenlecleucel compared with prior series evaluating AKI tainment of outcomes, which led to uncertainty that AKI
after CAR-T. Two historical series, predominantly evalu- rates would be 5% in a larger sample. Furthermore,
ating patients receiving axicabtagene ciloleucel, found because of the small number of patients receiving axi-
higher rates of overall CRS (~83%), severe CRS (13%), cabtagene ciloleucel at our center during the study period,
and AKI (23%).1,2 In the current series, we found lower we compare only with historical controls. However, this is
rates of hyponatremia and hypokalemia after tisagenle- the largest series describing the AKI rates in patients
cleucel treatment, which may also be due to lower rates of receiving tisagenlecleucel.
CRS. However, severe hypophosphatemia (phosphate < This series highlights that each CAR-T product must be
2.0 mg/dL) was extremely common in both this and a evaluated individually to define its unique toxicity profile.

AJKD Vol 77 | Iss 6 | June 2021 991


This is important for nephrologists to understand given the
intense interest in developing novel cellular-based thera-
pies for multiple cancer types.
Meghan D. Lee, Ian A. Strohbehn, Harish S. Seethapathy,
Nifasha Rusibamayila, Keagan S. Casey, Shruti Gupta, David
E. Leaf, Matthew J. Frigault and Meghan E. Sise
Article Information
Authors’ Full Names and Academic Degrees: Meghan D. Lee, BS,
Ian A. Strohbehn, BA, Harish S. Seethapathy, MBBS, Nifasha
Rusibamayila, MPH, Keagan S. Casey, BS, Shruti Gupta, MD,
MPH, David E. Leaf, MD, MMSc, Matthew J. Frigault, MD, and
Meghan E. Sise, MD, MS.
Authors’ Affiliations: Divisions of Nephrology (ML, IAS, HSS, NR,
MES) and Hematology and Oncology (KSC, MJF), Department of
Medicine, Massachusetts General Hospital; and Division of
Nephrology, Department of Medicine, Brigham and Women’s
Hospital, Boston, MA (SG, DEL).
Address for Correspondence: Meghan E. Sise, MD, MS, 165
Cambridge St, Ste 302, Boston, MA 02144. Email: msise@mgh.
harvard.edu
Authors’ Contributions: Research idea and study design: MDL, HS,
MES; data acquisition: MDL, IAS, NR, MES, KSC, MJF; data
analysis/interpretation: MDL, IAS, HS, NR, MJF, SG, DEL, MES;
statistical analysis: NR, MES; supervision or mentorship: MES;
joint senior authors: MJF and MES (equal contribution to this
work). Each author contributed important intellectual content
during manuscript drafting or revision, accepts personal
accountability for the author’s own contributions, and agrees to
ensure that questions pertaining to the accuracy or integrity of any
portion of the work are appropriately investigated and resolved,
including with documentation in the literature if appropriate.
Support: MES is supported by NIH K23 DK117014; MJF, by NIH
K12 CA087723. Funders did not have any role in study design;
data collection, analysis, or reporting; or the decision to submit for
publication.
Financial Disclosure: MJF receives research grant funding from
Novartis and Kite and has received consulting fees/honoraria from
Novartis, Kite, Celgene/BMS, and Arcellx. The remaining authors
declare that they have no relevant financial interests.
Peer Review: Received July 17, 2020. Evaluated by 3 external peer
reviewers, with direct editorial input from a Statistics/Methods Editor
and an Associate Editor, who served as Acting Editor-in-Chief.
Accepted in revised form August 29, 2020. The involvement of an
Acting Editor-in-Chief was to comply with AJKD’s procedures for
potential conflicts of interest for editors, described in the
Information for Authors & Journal Policies.
Publication Information: © 2020 by the National Kidney Founda-
tion, Inc. Published online October 22, 2020 with doi 10.1053/
j.ajkd.2020.08.017
References
1. Gupta S, Seethapathy H, Strohbehn IA, et al. Acute kidney
injury and electrolyte abnormalities after chimeric antigen re-
ceptor T-cell (CAR-T) therapy for diffuse large B-Cell lym-
phoma. Am J Kidney Dis. 2020;76(1):63-71.
2. Gutgarts V, Jain T, Zheng J, et al. Acute kidney injury after CAR-
T cell therapy: low incidence and rapid recovery. Biol Blood
Marrow Transplant. 2020;26(6):1071-1076.
3. Pasquini MC, Locke FL, Herrera AF, et al. Post-marketing use
outcomes of an anti-CD19 chimeric antigen receptor (CAR) T
992
Correspondence
cell therapy, axicabtagene ciloleucel (Axi-Cel), for the treatment
of large B cell lymphoma (LBCL) in the United States (US).
Blood. 2019;134:764.
4. Jaglowski S, Hu Z-H, Zhang Y, et al. Tisagenlecleucel chimeric
antigen receptor (CAR) T-cell therapy for adults with diffuse
large B-cell lymphoma (DLBCL): real world experience from
the Center for International Blood & Marrow Transplant
Research (CIBMTR) Cellular Therapy (CT) Registry. Blood.
2019;134:766.
5. Ying Z, He T, Wang X, et al. Parallel comparison of 4-1BB
or CD28 co-stimulated CD19-targeted CAR-T cells for B cell
non-Hodgkin’s lymphoma. Mol Ther Oncolytics. 2019;15:60-
68.
6. Kawalekar OU, O’Connor RS, Fraietta JA, et al. Distinct
signaling of coreceptors regulates specific metabolism path-
ways and impacts memory development in CAR T cells. Im-
munity. 2016;44(2):380-390.
7. Lee DW, Santomasso BD, Locke FL, et al. ASTCT consensus
grading for cytokine release syndrome and neurologic toxicity
associated with immune effector cells. Biol Blood Marrow
Transplant. 2019;25:625-638.
8. Perazella M, Shirali A. Nephrotoxicity of cancer immunother-
apies: past, present, and future. J Am Soc Nephrol.
2018;29(8):2039-2052.
9. Thaunat O, Delahousse M, Fakhouri F, et al. Nephrotic syn-
drome associated with hemophagocytic syndrome. Kidney Int.
2006;69(10):1892-1898.
Tubular Secretion of Creatinine and Risk
of Kidney Failure: The Modification of
Diet in Renal Disease (MDRD) Study
To the Editor:
Kidney function is most commonly monitored using
estimated glomerular filtration rate (eGFR) and albumin-
uria, which are markers of glomerular function.1 Tubular
secretion is an important nonglomerular kidney function
and is critical for toxin excretion and excretion of non-
filtered endogenous metabolites.2,3 Creatinine is filtered by
the glomerulus and secreted by the proximal tubule such
that creatinine clearance (CLcr) overestimates GFR by 10%
to 20%. Whether measuring secretory function provides
additional insights into kidney tubule health is uncertain.
The difference between measured CLcr (mCLcr) and
measured GFR (mGFR) represents the clearance of creati-
nine due to tubular secretion (TScr). To assess whether TScr
could be used to assess the impact of secretion on long-
term clinical outcomes, we performed a post hoc anal-
ysis of the MDRD Study.
During the baseline period of the MDRD Study before
randomization, mGFR was determined using 125I-iothala-
mate and mCLcr was determined using a 24-hour urine
collection performed at 2 time points (Item S1).4 Of 840
participants with mGFR data, 838 completed a baseline
urine collection. For participants with mCLcr and mGFR at
both time points (n = 718), the average of the 2 was
considered (Table S1). We evaluated the association of TScr
with time to initiation of kidney replacement therapy
AJKD Vol 77 | Iss 6 | June 2021

This is important for nephrologists to understand given the
intense interest in developing novel cellular-based thera-
pies for multiple cancer types.
Meghan D. Lee, Ian A. Strohbehn, Harish S. Seethapathy,
Nifasha Rusibamayila, Keagan S. Casey, Shruti Gupta, David
E. Leaf, Matthew J. Frigault and Meghan E. Sise
Article Information
Authors’ Full Names and Academic Degrees: Meghan D. Lee, BS,
Ian A. Strohbehn, BA, Harish S. Seethapathy, MBBS, Nifasha
Rusibamayila, MPH, Keagan S. Casey, BS, Shruti Gupta, MD,
MPH, David E. Leaf, MD, MMSc, Matthew J. Frigault, MD, and
Meghan E. Sise, MD, MS.
Authors’ Affiliations: Divisions of Nephrology (ML, IAS, HSS, NR,
MES) and Hematology and Oncology (KSC, MJF), Department of
Medicine, Massachusetts General Hospital; and Division of
Nephrology, Department of Medicine, Brigham and Women’s
Hospital, Boston, MA (SG, DEL).
Address for Correspondence: Meghan E. Sise, MD, MS, 165
Cambridge St, Ste 302, Boston, MA 02144. Email: msise@mgh.
harvard.edu
Authors’ Contributions: Research idea and study design: MDL, HS,
MES; data acquisition: MDL, IAS, NR, MES, KSC, MJF; data
analysis/interpretation: MDL, IAS, HS, NR, MJF, SG, DEL, MES;
statistical analysis: NR, MES; supervision or mentorship: MES;
joint senior authors: MJF and MES (equal contribution to this
work). Each author contributed important intellectual content
during manuscript drafting or revision, accepts personal
accountability for the author’s own contributions, and agrees to
ensure that questions pertaining to the accuracy or integrity of any
portion of the work are appropriately investigated and resolved,
including with documentation in the literature if appropriate.
Support: MES is supported by NIH K23 DK117014; MJF, by NIH
K12 CA087723. Funders did not have any role in study design;
data collection, analysis, or reporting; or the decision to submit for
publication.
Financial Disclosure: MJF receives research grant funding from
Novartis and Kite and has received consulting fees/honoraria from
Novartis, Kite, Celgene/BMS, and Arcellx. The remaining authors
declare that they have no relevant financial interests.
Peer Review: Received July 17, 2020. Evaluated by 3 external peer
reviewers, with direct editorial input from a Statistics/Methods Editor
and an Associate Editor, who served as Acting Editor-in-Chief.
Accepted in revised form August 29, 2020. The involvement of an
Acting Editor-in-Chief was to comply with AJKD’s procedures for
potential conflicts of interest for editors, described in the
Information for Authors & Journal Policies.
Publication Information: © 2020 by the National Kidney Founda-
tion, Inc. Published online November 20, 2020 with doi 10.1053/
j.ajkd.2020.08.017
References
1. Gupta S, Seethapathy H, Strohbehn IA, et al. Acute kidney
injury and electrolyte abnormalities after chimeric antigen re-
ceptor T-cell (CAR-T) therapy for diffuse large B-Cell lym-
phoma. Am J Kidney Dis. 2020;76(1):63-71.
2. Gutgarts V, Jain T, Zheng J, et al. Acute kidney injury after CAR-
T cell therapy: low incidence and rapid recovery. Biol Blood
Marrow Transplant. 2020;26(6):1071-1076.
3. Pasquini MC, Locke FL, Herrera AF, et al. Post-marketing use
outcomes of an anti-CD19 chimeric antigen receptor (CAR) T
992
Correspondence
cell therapy, axicabtagene ciloleucel (Axi-Cel), for the treatment
of large B cell lymphoma (LBCL) in the United States (US).
Blood. 2019;134:764.
4. Jaglowski S, Hu Z-H, Zhang Y, et al. Tisagenlecleucel chimeric
antigen receptor (CAR) T-cell therapy for adults with diffuse
large B-cell lymphoma (DLBCL): real world experience from
the Center for International Blood & Marrow Transplant
Research (CIBMTR) Cellular Therapy (CT) Registry. Blood.
2019;134:766.
5. Ying Z, He T, Wang X, et al. Parallel comparison of 4-1BB
or CD28 co-stimulated CD19-targeted CAR-T cells for B cell
non-Hodgkin’s lymphoma. Mol Ther Oncolytics. 2019;15:60-
68.
6. Kawalekar OU, O’Connor RS, Fraietta JA, et al. Distinct
signaling of coreceptors regulates specific metabolism path-
ways and impacts memory development in CAR T cells. Im-
munity. 2016;44(2):380-390.
7. Lee DW, Santomasso BD, Locke FL, et al. ASTCT consensus
grading for cytokine release syndrome and neurologic toxicity
associated with immune effector cells. Biol Blood Marrow
Transplant. 2019;25:625-638.
8. Perazella M, Shirali A. Nephrotoxicity of cancer immunother-
apies: past, present, and future. J Am Soc Nephrol.
2018;29(8):2039-2052.
9. Thaunat O, Delahousse M, Fakhouri F, et al. Nephrotic syn-
drome associated with hemophagocytic syndrome. Kidney Int.
2006;69(10):1892-1898.
Tubular Secretion of Creatinine and Risk
of Kidney Failure: The Modification of
Diet in Renal Disease (MDRD) Study
To the Editor:
Kidney function is most commonly monitored using
estimated glomerular filtration rate (eGFR) and albumin-
uria, which are markers of glomerular function.1 Tubular
secretion is an important nonglomerular kidney function
and is critical for toxin excretion and excretion of non-
filtered endogenous metabolites.2,3 Creatinine is filtered by
the glomerulus and secreted by the proximal tubule such
that creatinine clearance (CLcr) overestimates GFR by 10%
to 20%. Whether measuring secretory function provides
additional insights into kidney tubule health is uncertain.
The difference between measured CLcr (mCLcr) and
measured GFR (mGFR) represents the clearance of creati-
nine due to tubular secretion (TScr). To assess whether TScr
could be used to assess the impact of secretion on long-
term clinical outcomes, we performed a post hoc anal-
ysis of the MDRD Study.
During the baseline period of the MDRD Study before
randomization, mGFR was determined using 125I-iothala-
mate and mCLcr was determined using a 24-hour urine
collection performed at 2 time points (Item S1).4 Of 840
participants with mGFR data, 838 completed a baseline
urine collection. For participants with mCLcr and mGFR at
both time points (n = 718), the average of the 2 was
considered (Table S1). We evaluated the association of TScr
with time to initiation of kidney replacement therapy
AJKD Vol 77 | Iss 6 | June 2021
Correspondence

prevalent CVD was lower (Table S2). At any level of mGFR,

Figure 1. Scatter plot of creatinine secretion across the spec-
trum of mGFR at baseline among 838 MDRD Study partici-
pants, showing a correlation (Pearson correlation = 0.3)
between creatinine secretion and mGFR (P for significance <
0.001).
(KRT) for kidney failure (primary outcome) and all-cause
and cardiovascular disease (CVD) mortality (secondary
outcomes) through December 31, 2010. We used Cox
regression after adjusting for demographics, comorbid
conditions, lifestyle factors, and laboratory parameters
(including baseline proteinuria and mGFR). In sensitivity
analyses, we calculated the slope of mGFR between the 2
prerandomization visits and adjusted for it instead of
mGFR.
Mean age of 838 participants was 51.7 years and 60.5%
were men. At baseline, mean mGFR, mCLcr, and TScr were
33.0, 42.4, and 9.5 ± 7.3 mL/min/1.73 m2, respectively.
Compared with the highest TScr quartile, there were no
statistically significant differences in age, sex, race, smok-
ing status, or diabetes in participants with lower TScr,
whereas BMI was higher across quartiles of TScr and
there was a wide distribution of TScr across participants,
and TScr was positively and modestly correlated with
mGFR (Fig 1).
There were 626 cases of incident KRT during a me-
dian follow-up of 6.0 (interquartile range, 3.5-11.6)
years. Each 10-mL/min/1.73 m2 higher TScr was asso-
ciated with 25% lower multivariable-adjusted risk for
incident KRT; the lowest quartile of TScr had 1.6-fold
higher risk for incident KRT versus the top quartile
(P < 0.001; Table 1). Results of sensitivity analysis
adjusting for mGFR slope instead of mGFR were similar,
showing 27% lower risk for incident KRT (HR, 0.73
[95% CI, 0.65-0.82]) per 10-mL/min/1.73 m2 higher
TScr. There was no interaction between diet randomiza-
tion arm and TScr (P > 0.05). There were 444 cases of
all-cause mortality and 202 CVD-related deaths during
follow-up. Although the incidence rate of both outcomes
was lower with higher quartiles of TSCr, after multivari-
able adjustment, TSCr was not statistically significantly
associated with either outcome in multivariable-adjusted
models (Tables S3 and S4).
Our findings suggest that lower secretion of creatinine
may give insights about kidney health above and beyond
GFR and proteinuria. Prior studies have shown that a
number of small molecules are present in the plasma of
dialysis patients and are excreted primarily through
tubular secretion in healthy individuals.5 Low clearance of
some of these markers has also been associated with
mortality risk6 and kidney failure7 independent of eGFR
in persons with advanced chronic kidney disease (CKD).
However, measurement accuracy for these metabolites is
yet to be established. Our results support the same
concept but evaluate a precisely measured metabolite that
is well known to be secreted by the proximal tubule
(creatinine). Given that tubulointerstitial fibrosis is com-
mon in nearly all forms of kidney disease 8-10 and its
severity is the most reliable feature on biopsy to predict
progression to kidney failure,10 developing additional
markers of tubule health is likely to refine our ability to
detect and monitor global kidney health above and
beyond eGFR and albuminuria.

Table 1. Association of Tubular Secretion of Creatinine With Clinical Outcomes

No. of Incident KRT All-Cause Mortality CVD Mortality

Patients Events HR (95% CI)a Events HR (95% CI) Events HR (95% CI)
TScr, per 10 mL/min/1.73 m2 838 626 0.75 (0.67-0.85) 444 0.90 (0.78-1.04) 202 0.83 (0.67-1.02)
greater
TScr category
≤4.9 mL/min/1.73 m2 209 168 1.64 (1.28-2.10) 120 1.06 (0.79-1.42) 58 1.16 (0.75-1.79)
>4.9-9.0 mL/min/1.73 m2 210 170 1.46 (1.15-1.87) 114 1.22 (0.91-1.62) 50 1.27 (0.82-1.97)
>9.0-13.5 mL/min/1.73 m2 210 150 1.16 (0.91-1.48) 105 0.87 (0.65-1.15) 51 0.99 (0.65-1.51)
>13.5 mL/min/1.73 m2 209 138 1.00 (reference) 105 1.00 (reference) 43 1.00 (reference)
Abbreviation: HR, hazard ratio.
a
Multivariable adjusted (for age, sex, race, smoking status, MDRD [Modification of Diet in Renal Disease] Study A/B, blood pressure target, protein diet randomization,
cause of kidney disease, history of CVD, proteinuria, transferrin level, mean arterial pressure, lower serum high-density lipoprotein cholesterol level, and mGFR).

AJKD Vol 77 | Iss 6 | June 2021 993


Our study is limited by the use of only prerandomiza-
tion TScr; potential measurement errors in variables used to
calculate TScr; the lack of older participants and those with
insulin-dependent diabetes, and the limited number of
those of non-European ancestry; and the lack of detailed
smoking history. Strengths include the large number of
participants, long-term follow-up and large number of
events, detailed ascertainment of potentially confounding
variables, and concurrent availability of mGFR and mClcr at
baseline (twice in most participants).
In conclusion, lower TScr is associated with risk for
incident KRT independent of mGFR or proteinuria in a
large cohort of persons with mild to moderate predomi-
nantly nondiabetic CKD. Future studies should further
evaluate the role of tubule secretory markers, confirm their
association with adverse kidney outcomes, and determine
whether they may can be used for improving the dosing of
drugs that are primarily excreted by secretion.
Pranav S. Garimella, Hocine Tighiouart, Mark J. Sarnak,
Andrew S. Levey, Joachim H. Ix
Supplementary Material
Supplementary File (PDF)
Item S1, Tables S1-S4.
Article Information
Authors’ Full Names and Academic Degrees: Pranav S. Garimella,
MD, MPH, Hocine Tighiouart, MS, Mark J. Sarnak, MD, MS, Andrew
S. Levey, MD, and Joachim H. Ix, MD, MAS.
Authors’ Affiliations: Division of Nephrology and Hypertension,
Department of Medicine, University of California San Diego, San
Diego, CA (PSG, JHI); The Institute for Clinical Research and
Health Policy Studies, Tufts Medical Center (HT); Tufts Clinical
and Translational Science Institute, Tufts University (HT); Division
of Nephrology, Tufts Medical Center, Boston, MA (MJS, ASL);
Division of Preventive Medicine, Department of Family Medicine
and Public Health, University of California San Diego, San Diego
(JHI); and Nephrology Section, Veterans Affairs San Diego
Healthcare System, La Jolla, CA (JHI).
Address for Correspondence: Pranav S. Garimella, MD, MPH,
9452 Medical Center Dr, MC 7424, La Jolla, CA 92093. Email:
pgarimella@health.ucsd.edu
Authors’ Contributions: Research idea and study design: PSG, JHI,
MJS; data acquisition: MJS, ASL; data analysis/interpretation: HT,
PSG, JHI; supervision or mentorship: JHI, MJS. Each author
contributed important intellectual content during manuscript
drafting or revision and agrees to be personally accountable for
the individual’s own contributions and to ensure that questions
pertaining to the accuracy or integrity of any portion of the work,
even one in which the author was not directly involved, are
appropriately investigated and resolved, including with
documentation in the literature if appropriate.
994
Correspondence
Support: PSG is supported by NIDDK career development grant
K23 DK114556. JHI is supported by grants from NIDDK
(2R01DK098234 and K24DK110427) and the American Heart
Association (14EIA18560026). The funders had no role in the
design, analysis, interpretation, or reporting of these results.
Financial Disclosure: PSG has received clinical trial support from
Kadmon Inc and speaker fees from Otsuka. MJS serves on the
steering committee for Akebia, served on an advisory board for
Bayer, and is a consultant for Cardurian. JHI is principal investigator
of an investigator-initiated research project from Baxter Int. The
other authors declare that they have no relevant financial interests.
Peer Review: Received May 15, 2020. Evaluated by 2 external peer
reviewers and a statistician, with direct editorial input from an
Associate Editor, who served as Acting Editor-in-Chief. Accepted
in revised form September 18, 2020. The involvement of an Acting
Editor-in-Chief was to comply with AJKD’s procedures for
potential conflicts of interest for editors, described in the
Information for Authors & Journal Policies.
Publication Information: © 2020 by the National Kidney Founda-
tion, Inc. Published online November 20, 2020 with doi 10.1053/
j.ajkd.2020.09.017
References
1. Levey AS, Becker C, Inker LA. Glomerular filtration rate and
albuminuria for detection and staging of acute and chronic
kidney disease in adults: a systematic review. JAMA.
2015;313(8):837-846.
2. Nigam SK, Wu W, Bush KT, Hoenig MP, Blantz RC,
Bhatnagar V. Handling of drugs, metabolites, and uremic toxins
by kidney proximal tubule drug transporters. Clin J Am Soc
Nephrol. 2015;10(11):2039-2049.
3. Wang K, Kestenbaum B. Proximal tubular secretory clearance:
a neglected partner of kidney function. Clin J Am Soc Nephrol.
2018;13(8):1291-1296.
4. Klahr S, Levey AS, Beck GJ, et al. The effects of dietary protein
restriction and blood-pressure control on the progression of
chronic renal disease. Modification of Diet in Renal Disease
Study Group. N Engl J Med. 1994;330(13):877-884.
5. Sirich TL, Aronov PA, Plummer NS, Hostetter TH, Meyer TW.
Numerous protein-bound solutes are cleared by the kidney with
high efficiency. Kidney Int. 2013;84(3):585-590.
6. Suchy-Dicey AM, Laha T, Hoofnagle A, et al. Tubular
secretion in CKD. J Am Soc Nephrol. 2016;27(7):2148-
2155.
7. Chen Y, Zelnick LR, Wang K, et al. Kidney clearance of
secretory solutes is associated with progression of CKD: the
CRIC Study. J Am Soc Nephrol. 2020;31(4):817-827.
8. Nath KA. Tubulointerstitial changes as a major determinant in
the progression of renal damage. Am J Kidney Dis.
1992;20(1):1-17.
9. Ong AC, Fine LG. Loss of glomerular function and tubu-
lointerstitial fibrosis: cause or effect? Kidney Int. 1994;45(2):
345-351.
10. Rule AD, Amer H, Cornell LD, et al. The association between
age and nephrosclerosis on renal biopsy among healthy adults.
Ann Intern Med. 2010;152:561-567.
AJKD Vol 77 | Iss 6 | June 2021
Erratum

Erratum Regarding “Comparability of Plasma Iohexol

Clearance Across Population-Based Cohorts” (Am J Kidney
Dis. 2020;76[1]:54-62)
In the Original Investigation article entitled “Comparability of Plasma Iohexol Clearance Across Population-Based Co-
horts” that appeared in the July 2020 issue of AJKD (Eriksen et al, volume 76, issue 1, pages 54-62), the name of one of
the authors was spelled incorrectly. The name Olafur S. Indridasson should have appeared as Olafur S. Indridason. The
author list has been corrected in the HTML and PDF versions of the article as of April 20, 2021.

AJKD Vol 77 | Iss 6 | June 2021 995