Professional Documents
Culture Documents
n
Diabetes: Standards of Care in Florence
ce M. Brown, Dennis Bruemmer,
tio
Billy S. Collins, Marisa E. Hilliard,
Diabetes—2023 Diana
na Isaacs, Eric L. Johnson, Scott Kahan,
Kamlesh Khunti, Jose Leon, Sarah K. Lyons,
Diabetes Care 2023;46(Suppl. 1):S19–S40 | https://doi.org/10.2337/dc23-S002 Mary Lou Perry, Priya Prahalad,
c ia
Richard E. Pratley, Jane Jeffrie Seley,
Robert C. Stanton, and Robert A. Gabbay,
on behalf of the American Diabetes
Association
so
sA
provide the components of diabetes care, general treatment goals and guide-
lines, and tools to evaluate quality of care. Members of the ADA Professional
Practice Committee, a multidisciplinary expert committee, are responsible for up-
dating the Standards of Care annually, or more frequently as warranted. For a de-
te
tailed description of ADA standards, statements, and reports, as well as the
evidence-grading system for ADA’ss clinical practice recommendations and a full
listt of Professional Practice Committee members, please refer to Introduction
be
and Methodology.. Readers who wish to comment on the Standards of Care are
invited to do so at professional.diabetes.org/SOC.
ia
CLASSIFICATION
nD
diseases of the exocrine pancreas (such as cystic fibrosis and pancreatitis), and
drug- or chemical-induced diabetes (such as with glucocorticoid use, in the treat-
ment of HIV/AIDS, or after organ transplantation)
m
approximately half present with diabetic combining clinical, pathophysiological, progressive autoimmune b-cell destruc-
ketoacidosis (DKA) (2–4). The onset of and genetic characteristics to more pre- tion (16), thus accelerating insulin initiation
type 1 diabetes may be more variable cisely define the subsets of diabetes that prior to deterioration of glucose manage-
in adults; they may not present with are currently clustered into the type 1 ment or development of DKA (6,17).
the classic symptoms seen in children diabetes versus type 2 diabetes nomen- -cell demise and dys-
The paths to b-cell
and may experience temporary remis- clature with the goal of optimizing per- function are less well defined ned in type 2
n
sion from the need for insulin (5–7). The sonalized treatment approaches. Many diabetes, but deficient b-cell
-cell insulin se-
features most useful in discrimination of of these studies show great promise cretion, frequently in the setting of insulin
tio
type 1 diabetes include younger age at and may soon be incorporated into the resistance, appears to be the common de-
diagnosis (<35 years) with lower BMI diabetes classification system (13). nominator. Type 2 diabetes is associated
(<25 kg/m2), unintentional weight loss, Characterization of the underlying path- with insulin secretory defects related to
c ia
ketoacidosis, and glucose >360 mg/dL ophysiology is more precisely developed ammation, and metabolic
genetics, inflammation,
(20 mmol/L) at presentation (8). Occa- in type 1 diabetes than in type 2 diabe- classification schemes for
classifi
stress. Future classification
sionally, people with type 2 diabetes tes. It is now clear from prospective stud- diabetes will likely focus on the patho-
may present with DKA (9,10), particularly ies that the persistent presence of two or physiology of the underlying b-cell dys-
so
members of ethnic and racial minorities more islet autoantibodies is a near-certain (12,13,18
function (12,13,18–20).
(11). It is important for the health care predictor of clinical diabetes (14). The rate
professional to realize that classification of progression is dependent on the age DIAGNOSTIC TESTS FOR DIABETES
of diabetes type is not always straight- at first
rst detection of autoantibody, number
s
Diabetes may be diagnosed based on
forward at presentation and that mis- of autoantibodies, autoantibody specific-
specific-
specifi
diagnosis is common (e.g., adults with ity, and autoantibody titer. Glucose and plasma glucose criteria, either the fast-
sA
type 1 diabetes misdiagnosed as having A1C levels rise well before the clinical ing plasma glucose (FPG) value or the
type 2 diabetes, individuals with maturity- onset of diabetes, making diagnosis feasi- 2-h plasma glucose (2-h PG) value during
onset diabetes of the young [MODY] ble well before the onset of DKA. Three a 75-g oral glucose tolerance test (OGTT)
misdiagnosed as having type 1 diabe- distinct stages of type 1 diabetes can or A1C criteria (21) (Table 2.2).
te
tes). Although difficulties in distinguish- be identified (Table 2.1) and serve as
Table 2.1) Generally, FPG, 2-h PG during 75-g
ing diabetes type may occur in all age a framework for research and regula- OGTT, and A1C are equally appropriate
for diagnostic screening. It should be
be
groups at onset, the diagnosis becomes tory decision-making (12,15). There is
more obvious over time in people with debate as to whether slowly progressive noted that detection rates of different
b-cell deficiency as the degree of b-cell autoimmune diabetes with an adult on- screening tests vary in both populations
deficiency becomes clear. set should be termed latent autoimmune and individuals. Moreover, the efficacy
ia
In both type 1 and type 2 diabetes, diabetes in adults (LADA) or type 1 dia- of interventions for primary preven-
various genetic and environmental factors betes. The clinical priority with detection tion of type 2 diabetes has mainly been
can result in the progressive loss of b-cell
-cell of LADA is awareness that slow auto- demonstrated among individuals who
nD
mass and/or function that manifests clini- immune b-cell destruction can occur in have impaired glucose tolerance (IGT)
cally as hyperglycemia. Once hypergly- adults leading to a long duration of mar- with or without elevated fasting glucose,
cemia occurs, people with all forms of ginal insulin secretory capacity. For the not for individuals with isolated im-
diabetes are at risk for developing the purpose of this classification, all forms paired fasting glucose (IFG) or for those
ica
same chronic complications, although rates of diabetes mediated by autoimmune with prediabetes defined by A1C criteria
of progression may differ. The identification
identification
identifi b-cell destruction are included under (22,23).
of individualized therapies for diabetes the rubric of type 1 diabetes. Use of the The same tests may be used to screen
in the future will be informed by better term LADA is common and acceptable for and diagnose diabetes and to detect
er
characterization of the many paths to in clinical practice and has the practical individuals with prediabetes (Table 2.2 and
-cell demise or dysfunction (12). Across
b-cell impact of heightening awareness of Table 2.5) (24). Diabetes may be identified
the globe, many groups are working on a population of adults likely to have anywhere along the spectrum of clinical
m
Table 2.1
2.1—Staging
Stag of type 1 diabetes (12,16)
©A
FPG, fasting plasma glucose; IFG, impaired fasting glucose; IGT, impaired glucose tolerance; 2-h PG, 2-h plasma glucose.
diabetesjournals.org/care Classification and Diagnosis of Diabetes S21
n
dissolved in water.* Laboratory Improvement Amendments
OR (CLIA)-regulated and CLIA-waived settings.
tio
A1C $6.5% (48 mmol/mol). The test should be performed in a laboratory using a method
FDA-approved point-of-care A1C testing
that is NGSP certified and standardized to the DCCT assay.* can be used in laboratories or sites that
are CLIA certified, ed, are inspected, and
OR
c ia
meet the CLIA quality standards. These
In a patient with classic symptoms of hyperglycemia or hyperglycemic crisis, a random specified
standards include specifi
specified personnel re-
plasma glucose $200 mg/dL (11.1 mmol/L). quirements (including documented annual
DCCT, Diabetes Control and Complications Trial; FPG, fasting plasma glucose; OGTT, oral glu- competency assessments) and participa-
tion three times per year in an approved
so
cose tolerance test; NGSP, National Glycohemoglobin Standardization Program; WHO, World
Health Organization; 2-h PG, 2-h plasma glucose. *In the absence of unequivocal hyperglyce- ciency testing program (29–32).
proficiency (29 As
mia, diagnosis requires two abnormal test results from the same sample or in two separate discussed in Section 6, “Glycemic Targets,”
test samples.
point-of-care A1C assays may be more
s
generally applied for assessment of glyce-
scenarios—in seemingly low-risk individ- mic stability in the clinic.
sA
Food and Drug Administration–
uals who happen to have glucose testing, A1C has several advantages compared
approved devices at laborato-
in individuals screened based on diabetes with FPG and OGTT, including greater
ries proficient
cient in performing
risk assessment, and in symptomatic pa- convenience (fasting not required), greater
testing of moderate complex- preanalytical stability, and fewer day-to-day
tients. For additional details on the evi-
te
ity or higher by trained per-
dence used to establish the criteria for perturbations during stress, changes in
sonnel. B nutrition, or illness. However, these ad-
the diagnosis of diabetes, prediabetes,
2.2 Marked discordance between vantages may be offset by the lower
be
and abnormal glucose tolerance (IFG,
IGT), see the ADA position statement measured A1C and plasma glu- sensitivity of A1C at the designated cut
“Diagnosis and Classification of Diabetes cose levels should raise the point, greater cost, limited availability of
Mellitus” (1) and other reports (21,25,26). possibility of A1C assay interfer- A1C testing in certain regions of the de-
ence and consideration of using
ia
). The concor-
agnose diabetes (Table 2.2). to diagnose diabetes. B test, with a diagnostic threshold of $6.5%
dance between the FPG and 2-h PG tests 2.3 In conditions associated with (48 mmol/mol), diagnoses only 30% of the
is imperfect, as is the concordance be- an altered relationship between diabetes cases identified collectively using
tween A1C and either glucose-based test. A1C and glycemia, such as A1C, FPG, or 2-h PG, according to National
ica
Compared with FPG and A1C cut points, hemoglobinopathies including Health and Nutrition Examination Survey
the 2-h PG value diagnoses more people sickle cell disease, pregnancy (NHANES) data (33). Despite these limi-
with prediabetes and diabetes (27). In (second and third trimesters tations with A1C, in 2009, the Interna-
people in whom there is discordance and the postpartum period), tional Expert Committee added A1C to
between A1C values and glucose values, the diagnostic criteria with the goal of in-
er
glucose-6-phosphate dehydro-
FPG and 2-h PG are more accurate (28). genase deficiency, HIV, hemo- creased screening (21).
dialysis, recent blood loss or When using A1C to diagnose diabetes,
it is important to recognize that A1C is
m
diagnosis, the A1C test should diagnose diabetes. B into consideration that may impact he-
be performed using a method 2.4 Adequate carbohydrate intake moglobin glycation independently of gly-
certi
that is certified by the National (at least 150 g/day) should be cemia, such as hemodialysis, pregnancy,
assured for 3 days prior to HIV treatment (34,35), age, race/ethnicity,
Glycohemoglobin Standardiza-
oral glucose tolerance testing genetic background, and anemia/
tion Program (NGSP) and stan-
as a screen for diabetes. A hemoglobinopathies. (See OTHER CONDI-
dardized to the Diabetes Control
TIONS ALTERING THE RELATIONSHIP OF A 1C AND
and Complications Trial (DCCT)
GLYCEMIA below for more information.)
assay. B
The A1C test should be performed us-
2.1b Point-of-care A1C testing for
ing a method that is certified by the Age
diabetes screening and diagno-
NGSP (ngsp.org) and standardized or The epidemiologic studies that formed
sis should be restricted to U.S.
traceable to the Diabetes Control and the basis for recommending A1C to
S22 Classification and Diagnosis of Diabetes Diabetes Care Volume 46, Supplement 1, January 2023
diagnose diabetes included only adult measured with continuous glucose on the basis of the confirmatory screen-
populations (33). However, recent ADA monitoring (46). A recent report in ing test. For example, if a patient meets
clinical guidance concluded that A1C, Afro-Caribbean people demonstrated a the diabetes criterion of the A1C (two
FPG, or 2-h PG could be used to test lower A1C than predicted by glucose lev- 6.5% [48 mmol/mol]) but not
results $6.5%
for prediabetes or type 2 diabetes in els (47). Despite these and other reported FPG (<126 126 mg/dL [7.0 mmol/L]), that
children and adolescents (see SCREENING differences, the association of A1C with person should nevertheless be consid-
n
AND TESTING FOR PREDIABETES AND TYPE 2 DIABETES risk for complications appears to be ered to have diabetes.
below for addi- similar in African American and non- Each of the screening tests has prea-
tio
IN CHILDREN AND ADOLESCENTS
tional information) (36). Hispanic White populations (42,48). nalytic and analytic variability, so it is
In the Taiwanese population, age and possible that a test yielding an abnor-
Race/Ethnicity/Hemoglobinopathies sex have been reported to be associ- mal result (i.e., above the diagnostic
c ia
Hemoglobin variants can interfere with ated with increased A1C in men (49); threshold), when repeated, will produce
the measurement of A1C, although most the clinical implications of this finding a value below the diagnostic cut point.
assays in use in the U.S. are unaffected are unclear at this time. This scenario is likely for FPG and 2-h PG
by the most common variants. Marked if the glucose samples remain at room
discrepancies between measured A1C temperature and are not centrifuged
so
Other Conditions Altering the Relationship
and plasma glucose levels should prompt of A1C and Glycemia promptly. Because of the potential for
consideration that the A1C assay may In conditions associated with increased preanalytic variability, it is critical that
not be reliable for that individual. For red blood cell turnover, such as sickle samples for plasma glucose be spun and
s
individuals with a hemoglobin variant but cell disease, pregnancy (second and third separated immediately after they are
normal red blood cell turnover, such as trimesters), glucose-6-phosphate dehydro- drawn. If patients have test results near
sA
those with the sickle cell trait, an A1C as- genase deficiency
ciency (50,51), hemodialysis, the margins of the diagnostic threshold,
say without interference from hemoglo- recent blood loss or transfusion, or the health care professional should dis-
bin variants should be used. An updated erythropoietin therapy, only plasma blood cuss signs and symptoms with the pa-
list of A1C assays with interferences is glucose criteria should be used to diag- tient and repeat the test in 3–6 months.
te
available at ngsp.org/interf.asp. nose diabetes (52). A1C is less reliable People should consume a mixed diet
African American individuals heterozy- than blood glucose measurement in other with at least 150 g of carbohydrates on
gous for the common hemoglobin vari- conditions such as the postpartum state the 3 days prior to oral glucose tolerance
be
ant HbS may have, for any given level of (53–55),
55), HIV treated with certain protease testing (58–60). Fasting and carbohydrate
mean glycemia, lower A1C by about inhibitors (PIs) and nucleoside reverse restriction can falsely elevate glucose level
0.3% compared with those without the transcriptase inhibitors (NRTIs) (34), and with an oral glucose challenge.
ia
ican individuals, was associated with a Unless there is a clear clinical diagnosis surement of plasma glucose is sufficient
decrease in A1C of about 0.8% in homo- (e.g., patient in a hyperglycemic crisis or to diagnose diabetes (symptoms of hy-
zygous men and 0.7% in homozygous with classic symptoms of hyperglycemia perglycemia or hyperglycemic crisis plus
women compared with those without and a random plasma glucose $200 mg/dL a random plasma glucose $200 mg/dL
ica
the variant (38). For example, in Tanza- [11.1 mmol/L]), diagnosis requires two [11.1 mmol/L]). In these cases, knowing
nia, where there is a high likelihood of abnormal screening test results, either the plasma glucose level is critical because,
hemoglobinopathies in people with HIV, from the same sample (57) or in two in addition to confirming that symptoms
A1C may be lower than expected based separate test samples. If using two sepa- are due to diabetes, it will inform manage-
on glucose, limiting its usefulness for rate test samples, it is recommended that
er
als may have higher A1C levels than the diagnosis of diabetes is confirmed. If
TYPE 1 DIABETES
non-Hispanic White individuals with simi- two different tests (such as A1C and FPG)
lar fasting and post–glucose
post load glucose are both above the diagnostic threshold Recommendations
levels (43). Though conflicting
con data exist, when analyzed from the same sample or 2.5 Screening for presymptomatic
African American individuals may also in two different test samples, this also type 1 diabetes using screen-
have higher levels of fructosamine and confirms the diagnosis. On the other ing tests that detect autoanti-
glycated albumin and lower levels of hand, if a patient has discordant results bodies to insulin, glutamic acid
1,5-anhydroglucitol, suggesting that their from two different tests, then the test decarboxylase (GAD), islet anti-
glycemic burden (particularly postprandi- result that is above the diagnostic cut gen 2, or zinc transporter 8 is
ally) may be higher (44,45). Similarly, point should be repeated, with careful currently recommended in the
A1C levels may be higher for a given consideration of the possibility of A1C setting of a research study or
mean glucose concentration when assay interference. The diagnosis is made
diabetesjournals.org/care Classification and Diagnosis of Diabetes S23
can be considered an option dependent on insulin for survival and are checkpoint inhibitors (73). To date, the
for first-degree family mem- at risk for DKA (5–7,67,68). At this majority of immune checkpoint inhibitor–
bers of a proband with type 1 later stage of the disease, there is lit- related cases of type 1 diabetes occur in
tle or no insulin secretion, as manifested people with high-risk HLA-DR4 (present in
diabetes. B
by low or undetectable levels of plasma C- 76% of patients), whereas other high-risk
2.6 Development of and persistence
peptide. Immune-mediated diabetes is HLA alleles are not more common than
n
of multiple islet autoantibodies
the most common form of diabetes in those in the general population (73). To
is a risk factor for clinical di-
childhood and adolescence, but it can date, risk cannot be predicted by family
tio
abetes and may serve as an
occur at any age, even in the 8th and history or autoantibodies, so all health care
indication for intervention in
9th decades of life. professionals administering these medica-
the setting of a clinical trial or Autoimmune destruction of b-cells has tions should be mindful of this adverse ef-
screening for stage 2 type 1
c ia
multiple genetic factors and is also re- fect and educate patients appropriately.
diabetes. B lated to environmental factors that are
still poorly defined. Although individuals Idiopathic Type 1 Diabetes
Immune-Mediated Diabetes
do not typically have obesity when they Some forms of type 1 diabetes have no
present with type 1 diabetes, obesity is
so
This form, previously called “insulin- known etiologies. These individuals have
dependent diabetes” or “juvenile-onset increasingly common in the general pop- permanent insulinopenia and are prone
diabetes,” accounts for 5–10% of diabetes ulation; as such, obesity should not pre- to DKA but have no evidence of b-cell
and is due to cell-mediated autoimmune clude testing for type 1 diabetes. People autoimmunity. However, only a minority
s
with type 1 diabetes are also prone to of people with type 1 diabetes fall into this
destruction of the pancreatic b-cells. Au-
other autoimmune disorders such as
sA
toimmune markers include islet cell auto- category. Individuals with autoantibody-
Hashimoto thyroiditis, Graves disease, ce- negative type 1 diabetes of African or
antibodies and autoantibodies to GAD
liac disease, Addison disease, vitiligo, Asian ancestry may suffer from episodic
(glutamic acid decarboxylase, GAD65),
autoimmune hepatitis, myasthenia gra- DKA and exhibit varying degrees of insu-
insulin, the tyrosine phosphatases islet
vis, and pernicious anemia (see Section 4,
antigen 2 (IA-2) and IA-2b, and zinc trans-
te lin deficiency between episodes (possibly
Comprehensive Medical Evaluation and
“Comprehensive
porter 8. Numerous clinical studies are ketosis-prone diabetes) (74). This form of
Assessment of Comorbidities”).
Comorbidities”). Type 1
Comorbidities”
being conducted to test various methods diabetes is strongly inherited and is not
diabetes can be associated with mono-
be
of preventing type 1 diabetes in those HLA associated. An absolute requirement
genic polyglandular autoimmune syn-
with evidence of islet autoimmunity for insulin replacement therapy in affected
dromes, including immune dysregulation,
(trialnet.org/our-research/prevention- individuals may be intermittent. Future
polyendocrinopathy, enteropathy, and
studies) (14,17,61–64). Stage 1 of research is needed to determine the
ia
toimmune markers. The disease has the forkhead box protein 3 (FOXP3)(
ng HLA associations, with linkage
strong gene, and another caused by the auto- Screening for Type 1 Diabetes Risk
to the DQB1 and DRB1 haplotypes, and immune regulator (AIRE)
( gene mutation The incidence and prevalence of type 1
genetic screening has been used in some (69,70).
(69,70 As indicated by the names, diabetes are increasing (75). People with
research studies to identify high-risk type 1 diabetes often present with acute
ica
variable, being rapid in some individuals treatment has led to unexpected ad- suring islet autoantibodies in relatives of
(particularly but not exclusively in infants verse events, including immune system those with type 1 diabetes (15) or in chil-
and children) and slow in others (mainly activation precipitating autoimmune dis- dren from the general population (76,77)
m
but not exclusively adults) (65,66). Chil- ease. Fulminant onset of type 1 diabe- can effectively identify those who will de-
dren and adolescents often present with tes can develop, with DKA and low or velop type 1 diabetes. A study reported
the risk of progression to type 1 diabetes
©A
recruited from the general popula- 2.8 Testing for prediabetes and/ 2.14 Risk-based screening for predi-
tion. Remarkably, the findings in all or type 2 diabetes in asymp- abetes and/or type 2 diabetes
three groups were the same, suggesting tomatic people should be should be considered after
that the same sequence of events led to considered in adults of any age the onset of puberty or after
clinical disease in both “sporadic” and fa- with overweight or obesity 10 years of age, whichever
milial cases of type 1 diabetes. Indeed, (BMI $25 kg/m2 or $23 kg/m2
n
occurs earlier, in children and
the risk of type 1 diabetes increases as in Asian American individuals) adolescents with overweight
the number of relevant autoantibodies
tio
who have one or more risk (BMI $85th85th percentile) or
detected increases (63,78,79). In The En- factors (Table 2.3). B obesity (BMI $95th percentile)
vironmental Determinants of Diabetes in 2.9 For all people, screening should
the Young (TEDDY) study, type 1 diabetes and who have one or more
begin at age 35 years. B risk factors for diabetes. (See
c ia
developed in 21% of 363 subjects with 2.10 If tests are normal, repeat
at least one autoantibody at 3 years of Table 2.4 for evidence grad-
screening recommended at ing of risk factors.) B
age (80). Such testing, coupled with edu- a minimum of 3-year inter-
cation about diabetes symptoms and close 2.15 People with HIV should be
vals is reasonable, sooner with
follow-up, has been shown to enable ear- screened for diabetes and pre-
so
symptoms or change in risk
lier diagnosis and prevent DKA (81,82). diabetes with a fasting glucose
(i.e., weight gain). C
While widespread clinical screening test before starting antiretrovi-
2.11 To screen for prediabetes and
of asymptomatic low-risk individuals is ral therapy, at the time of
type 2 diabetes, fasting plasma
s
not currently recommended due to lack switching antiretroviral therapy,
glucose, 2-h plasma glucose
of approved therapeutic interventions, during 75-g oral glucose toler- and 3–6 months after starting
sA
several innovative research screening pro- ance test, and A1C are each or switching antiretroviral ther-
grams are available in Europe (e.g., Fr1da, appropriate (Table
Table 2.2 and apy. If initial screening results
gppad.org) and the U.S. (trialnet.org, Table 2.5). B are normal, fasting glucose
askhealth.org). Participation should be 2.12 When using oral glucose toler- should be checked annually. E
te
encouraged to accelerate development ance testing as a screen for di-
of evidence-based clinical guidelines for abetes, adequate carbohydrate
the general population and relatives of Prediabetes
be
intake (at least 150 g/day) “Prediabetes” is the term used for indi-
those with type 1 diabetes. Individuals should be assured for 3 days
who test positive should be counseled viduals whose glucose levels do not
prior to testing. A
about the risk of developing diabetes, meet the criteria for diabetes yet have
2.13 In people with prediabetes and
diabetes symptoms, and DKA preven- abnormal carbohydrate metabolism (48,85).
ia
of preventing and treating stage 2 type 1 and/or A1C 5.7–6.4% (39–47 mmol/mol)
diabetes in those with evidence of auto-
immunity with promising results (see
Table 2.3—Criteria for screening for diabetes or prediabetes in asymptomatic
clinicaltrials.gov and trialnet.org). Delay
adults
of overt diabetes development in stage 2
ica
1. Testing should be considered in adults with overweight or obesity (BMI $25 kg/m2 or
type 1 diabetes with the anti-CD3 anti- $23 kg/m2 in Asian American individuals) who have one or more of the following risk factors:
body teplizumab in relatives at risk for First-degree relative with diabetes
type 1 diabetes was reported in 2019, High-risk race/ethnicity (e.g., African American, Latino, Native American, Asian
with an extension of the randomized American, Pacific Islander)
er
(2.82 mmol/L)
or prevention of clinical type 1 diabetes Individuals with polycystic ovary syndrome
in at-risk individuals. Neither this agent Physical inactivity
nor others in this category are currently Other clinical conditions associated with insulin resistance (e.g., severe obesity,
©A
n
Family history of type 2 diabetes in first- or second-degree relative A
their increased risk for diabetes and CVD
Race/ethnicity (Native American, African American, Latino, Asian American, Pacific
tio
Islander) A and counseled about effective strategies
Signs of insulin resistance or conditions associated with insulin resistance (acanthosis nigricans, to lower their risks (see Section 3,
hypertension, dyslipidemia, polycystic ovary syndrome, or small-for-gestational-age birth Prevention or Delay of Type 2 Diabetes
“Prevention
weight) B Comorbidities Similar
and Associated Comorbidities”).
c ia
GDM, gestational diabetes mellitus. *After the onset of puberty or after 10 years of age, to glucose measurements, the continuum
whichever occurs earlier. If tests are normal, repeat testing at a minimum of 3-year intervals of risk is curvilinear, so as A1C rises, the
(or more frequently if BMI is increasing or risk factor profile deteriorating) is recommended. diabetes risk rises disproportionately (86).
Reports of type 2 diabetes before age 10 years exist, and this can be considered with nu-
Aggressive interventions and vigilant
so
merous risk factors.
low-up should be pursued for those
follow-up
considered at very high risk (e.g., those
(Table 2.5). Prediabetes should not be defined
ned by A1C criteria demonstrated a >6.0% [42 mmol/mol]).
with A1C >6.0%
s
viewed as a clinical entity in its own right strong, continuous association between Table 2.5 summarizes the categories
but rather as a risk factor for progression A1C and subsequent diabetes. In a sys- of prediabetes, and Table 2.3 outlines
sA
to diabetes and cardiovascular disease tematic review of 44,203 individuals from the criteria for screening for prediabe-
(CVD). Criteria for screening for diabetes 16 cohort studies with a follow-up interval tes. The ADA Diabetes Risk Test is an
or prediabetes in asymptomatic adults averaging 5.6 years (range 2.8–12
2.8––12 years),
2.8 additional option for assessment to
are outlined in Table 2.3. Prediabetes is those with A1C between 5.5% and 6.0% determine the appropriateness of screen-
te
associated with obesity (especially ab- (between 37 and 42 mmol/mol) had a ing for diabetes or prediabetes in asymp-
dominal or visceral obesity), dyslipidemia substantially increased risk of diabetes tomatic adults (Fig. 2.1) (diabetes.org/
socrisktest). For additional background
be
with high triglycerides and/or low HDL (5-year incidence from 9% to 25%). Those
cholesterol, and hypertension. The pres- with an A1C range of 6.0–6.5%
6.0 (42–48 regarding risk factors and screening
ence of prediabetes should prompt com- mmol/mol) had a 5-year risk of develop- for prediabetes, see SCREENING AND TESTING
prehensive screening for cardiovascular ing diabetes between 25% and 50% and FOR PREDIABETES AND TYPE 2 DIABETES IN ASYMPTOM-
ia
risk factors. a relative risk 20 times higher compared ATIC ADULTS and also SCREENING AND TESTING FOR
with A1C of 5.0% (31 mmol/mol) (86). In a PREDIABETES AND TYPE 2 DIABETES IN CHILDREN
Diagnosis community-based study of African American AND ADOLESCENTS below. For details re-
nD
7.8 to 11.0 mmol/L) (25). It should be (87). Other analyses suggest that A1C of of Type 2 Diabetes and Associated
noted that the World Health Organiza- 5.7% (39 mmol/mol) or higher is associ- Comorbidities.”
tion and numerous other diabetes organ- ated with a diabetes risk similar to that of
ne the IFG lower limit at
izations define the high-risk participants in the Diabetes Type 2 Diabetes
er
110 mg/dL (6.1 mmol/L). Prevention Program (DPP) (88), and A1C Type 2 diabetes, previously referred to
As with the glucose measures, several at baseline was a strong predictor of the as “non-insulin-dependent diabetes”
prospective studies that used A1C to development of glucose-defined diabetes or “adult-onset diabetes,” accounts for
m
predict the progression to diabetes as during the DPP and its follow-up (89). 90–95% of all diabetes. This form en-
compasses individuals who have relative
(rather than absolute) insulin deficiency
©A
Table 2.5—Criter
2.5—Criteria defining prediabetes*
FPG 100 mg/dL (5.6 mmol/L) to 125 mg/dL (6.9 mmol/L) (IFG)
and have peripheral insulin resistance.
At least initially, and often throughout
OR
their lifetime, these individuals may not
2-h PG during 75-g OGTT 140 mg/dL (7.8 mmol/L) to 199 mg/dL (11.0 mmol/L) (IGT) need insulin treatment to survive.
OR There are various causes of type 2 di-
A1C 5.7–6.4% (39–47 mmol/mol)
abetes. Although the specific etiologies
are not known, autoimmune destruction
FPG, fasting plasma glucose; IFG, impaired fasting glucose; IGT, impaired glucose tolerance; of b-cells does not occur, and patients
OGTT, oral glucose tolerance test; 2-h PG, 2-h plasma glucose. *For all three tests, risk is
do not have any of the other known
continuous, extending below the lower limit of the range and becoming disproportionately
greater at the higher end of the range. causes of diabetes. Most, but not all,
people with type 2 diabetes have
S26 Classification and Diagnosis of Diabetes Diabetes Care Volume 46, Supplement 1, January 2023
American
Diabetes
® Association®
n
Are you at risk for type 2 diabetes?
tio
WRITE YOUR SCORE
Diabetes Risk Test: IN THE BOX.
c ia
Height Weight (lbs.)
1. How old are you? ................................................... 4’ 10” 119–142 143–190 191+
Less than 40 years (0 points)
4’ 11” 124–147 148–197 198+
40–49 years (1 point)
5’ 0” 128–152 153–203 204+
so
50–59 years (2 points)
5’ 1” 132–157 158–210 211+
60 years or older (3 points)
5’ 2” 136–163 164–217 218+
2. Are you a man or a woman? ................................. 5’ 3” 141–168 169–224 225+
Man (1 point) Woman (0 points) 5’ 4” 145–173 174–231 232+
s
5’ 5” 150–179 180–239 240+
3. If you are a woman, have you ever been
sA
5’ 6” 155–185 186–246 247+
diagnosed with gestational diabetes?..................
5’ 7” 159–190 191–254 255+
Yes (1 point) No (0 points)
5’ 8” 164–196 197–261 262+
4. Do you have a mother, father, sister or brother 5’ 9” 169–202 203–269 270+
with diabetes? ........................................................
te 5’ 10” 174–208 209–277 278+
Yes (1 point) No (0 points) 5’ 11” 179–214 215–285 286+
5. Have you ever been diagnosed with high 6’ 0” 184–220 221–293 294+
be
blood pressure? ..................................................... 6’ 1” 189–226 227–301 302+
Yes (1 point) No (0 points) 6’ 2” 194–232 233–310 311+
6’ 3” 200–239 240–318 319+
6. Are you physically active? ....................................
ia
which blood glucose levels are higher than normal risk for type 2 diabetes. Small steps make
but not yet high enough to be diagnosed as diabetes. a big difference in helping you live a longer,
healthier life.
er
body weight than the rest of the general public (about 15 steps you can take to help lower your risk.
pounds lower).
overweight or obesity. Excess weight it- percentage of body fat distributed pre- illness such as infection or myocardial infarc-
self causes some degree of insulin re- dominantly in the abdominal region. tion or with the use of certain drugs (e.g.,
sistance. Individuals who do not have DKA seldom occurs spontaneously in corticosteroids, atypical antipsychotics, and
obesity or overweight by traditional type 2 diabetes; when seen, it usually arises sodium–glucose cotransporter 2 inhibitors)
weight criteria may have an increased in association with the stress of another (90,91). Type 2 diabetes frequently goes
diabetesjournals.org/care Classification and Diagnosis of Diabetes S27
undiagnosed for many years because guide health care professionals on (109). Employing a probabilistic model,
hyperglycemia develops gradually and, whether performing a diagnostic test Peterson et al. (110) demonstrated cost
at earlier stages, is often not severe (Table 2.2) is appropriate. Prediabetes and health benefits of preconception
enough for the patient to notice the and type 2 diabetes meet criteria for screening.
classic diabetes symptoms caused by hy- conditions in which early detection via A large European randomized con-
perglycemia, such as dehydration or un- screening is appropriate. Both conditions trolled trial compared the impact of
n
intentional weight loss. Nevertheless, are common and impose significant clin- screening for diabetes and intensive
even undiagnosed people with diabetes ical and public health burdens. There is multifactorial intervention with that of
tio
are at increased risk of developing macro- often a long presymptomatic phase be- screening and routine care (111). Gen-
vascular and microvascular complications. fore the diagnosis of type 2 diabetes. eral practice patients between the ages
People with type 2 diabetes may have Simple tests to detect preclinical disease of 40 and 69 years were screened for
c ia
insulin levels that appear normal or ele- are readily available (101). The duration diabetes and randomly assigned by prac-
vated, yet the failure to normalize blood of glycemic burden is a strong predictor tice to intensive treatment of multiple
glucose reflects a relative defect in glu- of adverse outcomes. There are effective risk factors or routine diabetes care.
cose-stimulated insulin secretion. Thus, interventions that prevent progression After 5.3 years of follow-up, CVD risk
insulin secretion is defective in these factors were modestly but significantly
signi
so
from prediabetes to diabetes. It is im-
individuals and insufficient to compensate portant to individualize risk/benefit of roved
improved
improve d with intensive treatment com-
for insulin resistance. Insulin resistance formal intervention for people with pre- pared with routine care, but the inci-
may improve with weight reduction, diabetes and consider patient-centered dence of first
fi CVD events or mortality
s
physical activity, and/or pharmacologic goals. Risk models have explored the was not significantly
signi different between
treatment of hyperglycemia but is sel- the groups (26). The excellent care pro-
sA
benefit, in general finding
nding higher ben-
dom restored to normal. Recent inter- efitt of intervention in those at highest vided to patients in the routine care
ventions with intensive diet and exercise risk (102) (see Section 3, “Prevention
Prevention or group and the lack of an unscreened
or surgical weight loss have led to diabe- Delay of Type 2 Diabetes and Associated control arm limited the authors’ ability
tes remission (92–98) (see Section 8, to determine whether screening and
te
Comorbidities”)) and reduce the risk
“Obesity and Weight Management for early treatment improved outcomes com-
of diabetes complications (103) (see
the Prevention and Treatment of Type 2 pared with no screening and later treat-
Section 10, “Cardiovascular
Cardiovascular Disease and
ment after clinical diagnoses. Computer
be
Diabetes”).
Risk Management,”
Management, Section 11, “Chronic
The risk of developing type 2 diabetes simulation modeling studies suggest that
Kidney Disease and Risk Management,”
Management,
increases with age, obesity, and lack of major benefits are likely to accrue from
and Section 12, “Retinopathy,
Retinopathy, Neuropathy,
physical activity (99,100). It occurs more the early diagnosis and treatment of
and Foot Care”).
Care”). In the most recent Na-
Care”
ia
GAD65 autoantibodies) to exclude the with diabetes are undiagnosed (106,107). 35 years for all people (116). Screening
diagnosis of type 1 diabetes (8). Although screening of asymptomatic indi- should be considered in adults of any
viduals to identify those with prediabetes age with overweight or obesity and one
Screening and Testing for Prediabetes or diabetes might seem reasonable, rigor- or more risk factors for diabetes.
and Type 2 Diabetes in Asymptomatic ous clinical trials to prove the effective-
Adults ness of such screening have not been BMI and Ethnicity
Screening for prediabetes and type 2 di- conducted and are unlikely to occur. In general, BMI $25 kg/m2 is a risk fac-
abetes risk through an informal assess- Clinical conditions, such as hyperten- tor for diabetes. However, data suggest
ment of risk factors (Table 2.3) or with sion, hypertensive pregnancy, and obe- that the BMI cut point should be lower
an assessment tool, such as the ADA sity, enhance risk (108). Based on a for the Asian American population
risk test (Fig. 2.1) (online at diabetes. population estimate, diabetes in people (117,118). The BMI cut points fall con-
org/socrisktest), is recommended to of childbearing age is underdiagnosed sistently between 23 and 24 kg/m2
S28 Classification and Diagnosis of Diabetes Diabetes Care Volume 46, Supplement 1, January 2023
(sensitivity of 80%) for nearly all Asian pancreatic b-cells. NRTIs also affect fat and 14.60% meeting criteria for pre-
American subgroups (with levels slightly distribution (both lipohypertrophy and diabetes and diabetes, respectively,
lower for Japanese American individ- lipoatrophy), which is associated with using random blood glucose. Further
uals). This makes a rounded cut point insulin resistance. For people with HIV research is needed to demonstrate
of 23 kg/m2 practical. An argument can and ARV-associated hyperglycemia, it may the feasibility, effectiveness, and cost-
be made to push the BMI cut point to be appropriate to consider discontinuing effectiveness of screening in this setting.
n
lower than 23 kg/m2 in favor of in- the problematic ARV agents if safe and
creased sensitivity; however, this would effective alternatives are available (124). Screening and Testing for Prediabetes
tio
lead to an unacceptably low specificity Before making ARV substitutions, care- and Type 2 Diabetes in Children and
(13.1%). Data from the World Health fully consider the possible effect on Adolescents
Organization also suggest that a BMI of HIV virological control and the poten- In the last decade, the incidence and
c ia
$23 kg/m2 should be used to define in- tial adverse effects of new ARV agents. prevalence of type 2 diabetes in chil-
creased risk in Asian American individu- In some cases, antihyperglycemic agents dren and adolescents has increased dra-
als (119). The finding that one-third to may still be necessary. matically, especially in racial and ethnic
one-half of diabetes in Asian American minority populations (75). See Table 2.4
people is undiagnosed suggests that for recommendations on risk-based
so
Testing Interval
testing is not occurring at lower BMI The appropriate interval between screen- screening for type 2 diabetes or pre-
thresholds (99,120). ing tests is not known (125). The rationale diabetes in asymptomatic children and
Evidence also suggests that other pop- for the 3-year interval is that with this in- adolescents in a clinical setting (36). See
s
ulations may benefit from lower BMI cut terval, the number of false-positive tests Table 2.2 and Table 2.5 for the criteria
points. For example, in a large multiethnic rmatory testing will
that require confirmatory for the diagnosis of diabetes and pre-
sA
cohort study, for an equivalent incidence be reduced, and individuals with false- diabetes, respectively, that apply to
rate of diabetes, a BMI of 30 kg/m2 in negative tests will be retested before children, adolescents, and adults. See
non-Hispanic White individuals was equiv- substantial time elapses and complica- Section 14, “Children and Adolescents,”
alent to a BMI of 26 kg/m2 in African tions develop (125). In especially high- for additional information on type 2 dia-
te
American individuals (121). risk individuals, particularly with weight betes in children and adolescents.
gain, shorter intervals between screen- Some studies question the validity of
ing may be useful. A1C in the pediatric population, espe-
be
Medications
Certain medications, such as glucocorti- cially among certain ethnicities, and
coids, thiazide diuretics, some HIV medi- Community Screening suggest OGTT or FPG as more suitable
cations (34), and atypical antipsychotics Ideally, screening should be carried out diagnostic tests (131). However, many
ia
(92), are known to increase the risk of within a health care setting because of of these studies do not recognize that
diabetes and should be considered when the need for follow-up and treatment. diabetes diagnostic criteria are based
deciding whether to screen. Community screening outside a health on long-term health outcomes, and vali-
nD
care setting is generally not recommended dations are not currently available in
HIV because people with positive tests may the pediatric population (132). The ADA
Individuals with HIV are at higher risk not seek, or have access to, appropriate acknowledges the limited data support-
for developing prediabetes and diabetes follow-up testing and care. However, in ing A1C for diagnosing type 2 diabetes
ica
on antiretroviral (ARV) therapies; a specific situations where an adequate re- in children and adolescents. Although
screening protocol is therefore recom- ferral system is established beforehand A1C is not recommended for diagnosis
mended (122). The A1C C test may underes- for positive tests, community screening of diabetes in children with cystic fibro-
timate glycemia in people with HIV; it is may be considered. Community screen- sis or symptoms suggestive of acute on-
not recommended for diagnosis and may ing may also be poorly targeted; i.e., it
er
not previously diagnosed associated with preservation of lung diabetes mellitus is best made
with cystic fibrosis–related dia- function. The European Cystic Fibrosis once the individual is stable on
betes. B Society Patient Registry reported an in- an immunosuppressive regi-
2.17 A1C is not recommended as crease in CFRD with age (increased 10% men and in the absence of
per decade), genotype, decreased lung
a screening test for cystic an acute infection. B
function, and female sex (140,141). Con-
n
fibrosis–related diabetes. B 2.21 The oral glucose tolerance test
tinuous glucose monitoring or HOMA of
2.18 People with cystic fibrosis– is the preferred test to make a
b-cell function (142) may be more sensi-
tio
related diabetes should be diagnosis of posttransplanta-
tive than OGTT to detect risk for progres-
treated with insulin to attain in- tion diabetes mellitus. B
sion to CFRD; however, evidence linking
dividualized glycemic goals. A 2.22 Immunosuppressive regimens
these results to long-term outcomes is
2.19 Beginning 5 years after the shown to provide the best out-
c ia
lacking, and these tests are not recom-
diagnosis of cystic fibrosis– mended for screening outside of the re- comes for patient and graft
related diabetes, annual mon- search setting (143). survival should be used, irre-
itoring for complications of di- CFRD mortality has significantly de- spective of posttransplantation
abetes is recommended. E creased over time, and the gap in mor- diabetes mellitus risk. E
so
tality between people with cystic fibrosis
Cystic fibrosis–related diabetes (CFRD) is with and without diabetes has consider- Several terms are used in the literature
the most common comorbidity in people ably narrowed (144). There are limited to describe the presence of diabetes
s
with cystic fibrosis, occurring in about clinical trial data on therapy for CFRD. following organ transplantation (147).
The largest study compared three regi-
sA
20% of adolescents and 40–50% of adults New-onset diabetes after transplantation
““New-onset transplantation”
mens: premeal insulin aspart, repagli-
(135). Diabetes in this population, com- (NODAT) is one such designation that
nide, or oral placebo in people with cystic
pared with individuals with type 1 or describes individuals who develop new-
brosis and diabetes or abnormal glucose
fibrosis
type 2 diabetes, is associated with worse onset diabetes following transplant.
tolerance. Participants all had weight loss
nutritional status, more severe inflamma-
te NODAT excludes people with pretrans-
in the year preceding treatment; however,
tory lung disease, and greater mortality. In- plant diabetes that was undiagnosed as
in the insulin-treated group, this pattern
sulin insufficiency is the primary defect in well as posttransplant hyperglycemia
was reversed, and participants gained 0.39
be
CFRD. Genetically determined b-cell func- that resolves by the time of discharge
(± 0.21) BMI units (P
(P = 0.02). The repagli-
tion and insulin resistance associated with (148). Another term, “posttransplantation
nide-treated group had initial weight
infection and inflammation may also diabetes mellitus” (PTDM) (148,149), de-
gain, but it was not sustained by 6
contribute to the development of CFRD. scribes the presence of diabetes in the
ia
at earlier ages than CFRD. Whether indi- The primary rationale for the use of insu- Hyperglycemia is very common dur-
viduals with IGT should be treated with lin in people with CFRD is to induce an ana- ing the early posttransplant period, with
insulin replacement has not currently bolic state while promoting macronutrient !90% of kidney allograft recipients ex-
been determined. Although screening retention and weight gain. hibiting hyperglycemia in the first few
for diabetes before the age of 10 years weeks following transplant (148–151).
ica
lung function. OGTT is the recommended American Diabetes Association and a pies is a major contributor to the devel-
screening test; however, recent publica- Clinical Practice Guideline of the Cystic opment of PTDM, the risks of transplant
tions suggest that an A1C cut point Fibrosis Foundation, Endorsed by the rejection outweigh the risks of PTDM,
m
threshold of 5.5% (5.8% in a second Pediatric Endocrine Society” (146) and and the role of the diabetes care health
study)
dy) would detect more than 90% of in the International Society for Pediatric care professional is to treat hyperglyce-
©A
cases and reduce patient screening bur- and Adolescent Diabetes 2018 clinical mia appropriately regardless of the type
den (136,137). Ongoing studies are un- practice consensus guidelines (135). of immunosuppression (148). Risk fac-
derway to validate this approach, and tors for PTDM include both general dia-
A1C is not recommended for screening betes risks (such as age, family history
POSTTRANSPLANTATION
(138). Regardless of age, weight loss or of diabetes, etc.) as well as transplant-
DIABETES MELLITUS
failure of expected weight gain is a risk specific factors, such as use of immuno-
for CFRD and should prompt screening Recommendations suppressant agents (153–155). Whereas
(136,137). The Cystic Fibrosis Foundation 2.20 After organ transplantation, posttransplantation hyperglycemia is an
Patient Registry (139) evaluated 3,553 screening for hyperglycemia important risk factor for subsequent
people with cystic fibrosis and diag- should be done. A formal di- PTDM, a formal diagnosis of PTDM is op-
nosed 445 (13%) with CFRD. Early di- agnosis of posttransplantation timally made once the patient is stable
agnosis and treatment of CFRD was on maintenance mmunosuppression and
S30 Classification and Diagnosis of Diabetes Diabetes Care Volume 46, Supplement 1, January 2023
in the absence of acute infection people with liver and kidney transplants, 6 months of age. Neonatal diabetes can
(151–153,156). In a recent study of 152 but side effects include fluid retention, either be transient or permanent. Tran-
heart transplant recipients, 38% had heart failure, and osteopenia (167,168). sient diabetes is most often due to over-
PTDM at 1 year. Risk factors for PTDM Dipeptidyl peptidase 4 inhibitors do not expression of genes on chromosome
included elevated BMI, discharge from interact with immunosuppressant drugs 6q24, is recurrent in about half of cases,
the hospital on insulin, and glucose val- and have demonstrated safety in small and may be treatable with medications
n
ues in the 24 h prior to hospital dis- clinical trials (169,170). Well-designed inter- other than insulin. Permanent neonatal
charge (157). In an Iranian cohort, 19% diabetes is most commonly due to auto-
tio
vention trials examining the efficacy and
had PTDM after heart and lung trans- safety of these and other antihyperglyce- somal dominant mutations in the genes
plant (158). The OGTT is considered mic agents in people with PTDM are encoding the Kir6.2 subunit (KCNJ11)(
the gold-standard test for the diagnosis needed. and SUR1 subunit (ABCC8
((ABCC8)
ABCC8)) of the b-cell
c ia
of PTDM (1 year posttransplant) (148, KATP channel. A recent report details a
149,159,160). Pretransplant elevation MONOGENIC DIABETES SYNDROMES de novo mutation in EIF2B1 affecting
in hs-CRP was associated with PTDM in eIF2 signaling associated with permanent
the setting of renal transplant (161,162). Recommendations neonatal diabetes and hepatic dys-
However, screening people with fasting Regardless of current age, all
so
2.23 function, similar to Wolcott-Ra
Wolcott-Rallison
glucose and/or A1C can identify high- people diagnosed with diabe- syndrome but with few severe comor-
risk individuals requiring further assess- tes in the first 6 months of bidities (176). The recent ADA-European
ment and may reduce the number of life should have immediate Association for the Study of Diabetes
s
overall OGTTs required. genetic testing for neonatal type 1 diabetes consensus report recom-
Few randomized controlled studies diabetes. A
sA
mends that regardless of current age, in-
have reported on the short- and long- 2.24 Children and young adults who dividuals diagnosed under 6 months of
term use of antihyperglycemic agents do not have typical characteris- age should have genetic testing (8). Cor-
in the setting of PTDM (153,163,164). tics of type 1 or type 2 diabetes rect diagnosis has critical implications be-
Most studies have reported that trans-
te
and who often have a family cause 30–50% of people with KATP-related
plant patients with hyperglycemia and history of diabetes in successive neonatal diabetes will exhibit improved
PTDM after transplantation have higher
generations (suggestive of an blood glucose levels when treated with
rates of rejection, infection, and rehospi-
be
autosomal dominant pattern high-dose oral sulfonylureas instead of in-
talization (151,153,165). Insulin therapy
of inheritance) should have sulin. Insulin gene (INS) mutations are the
is the agent of choice for the manage-
genetic testing for maturity- second most common cause of perma-
ment of hyperglycemia, PTDM, and pre-
onset diabetes of the young. A nent neonatal diabetes, and while inten-
ia
on their pretransplant regimen if they diabetes genetics is recom- important genetic counseling considera-
were in good control before transplanta- mended to understand the tions, as most of the mutations that cause
tion. Those with previously poor glycemic significance of genetic muta- diabetes are dominantly inherited.
stability or with persistent hyperglycemia tions and how best to approach
ica
should continue insulin with frequent further evaluation, treatment, Maturity-Onset Diabetes of the Young
home glucose monitoring to determine and genetic counseling. E MODY is frequently characterized by on-
when insulin dose reductions may be set of hyperglycemia at an early age
needed and when it may be appropriate (classically before age 25 years, although
Monogenic defects that cause b-cell diagnosis may occur at older ages).
to switch to noninsulin agents.
er
agent is usually made based on the side describes the most common causes of istent obesity). It is inherited in an autoso-
profile of the medication and
profi
effect profile monogenic diabetes. For a comprehen- mal dominant pattern with abnormalities
sive list of causes, see Genetic Diagnosis
©A
patient
possible interactions with the patient’s in at least 13 genes on different chromo-
immunosuppression regimen (153). Drug of Endocrine Disorders (171). somes identified to date (177). The most
dose adjustments may be required be- commonly reported forms are GCK-MODY
cause of decreases in the glomerular Neonatal Diabetes (MODY2), HNF1A-MODY (MODY3), and
filtration rate, a relatively common Diabetes occurring under 6 months of HNF4A-MODY (MODY1).
complication in transplant patients. A age is termed “neonatal” or “congenital” For individuals with MODY, the treat-
small short-term pilot study reported diabetes, and about 80–85% of cases ment implications are considerable and
that metformin was safe to use in re- can be found to have an underlying warrant genetic testing (178,179). Clini-
nal transplant recipients (166), but its monogenic cause (8,172–175). Neonatal cally, people with GCK-MODY exhibit
safety has not been determined in diabetes occurs much less often after mild, stable fasting hyperglycemia and
other types of organ transplant. Thiazolidi- 6 months of age, whereas autoimmune do not require antihyperglycemic ther-
nediones have been used successfully in type 1 diabetes rarely occurs before apy except commonly during pregnancy.
diabetesjournals.org/care Classification and Diagnosis of Diabetes S31
n
HNF4A AD HNF4A-MODY: progressive insulin secretory defect with presentation in adolescence or
early adulthood; may have large birth weight and transient neonatal hypoglycemia;
tio
sensitive to sulfonylureas
HNF1B AD HNF1B-MODY: developmental renal disease (typically cystic); genitourinary abnormalities;
atrophy of the pancreas; hyperuricemia; gout
GCK AD GCK-MODY: higher glucose threshold (set point) for glucose-stimulated insulin
c ia
secretion, causing stable, nonprogressive elevated fasting blood glucose; typically,
does not require treatment; microvascular complications are rare; small rise in
2-h PG level on OGTT (<54 mg/dL [3 mmol/L])
Neonatal diabetes KCNJ11 AD Permanent or transient: IUGR; possible developmental delay and seizures; responsive
to sulfonylureas
so
INS AD Permanent: IUGR; insulin requiring
ABCC8 AD Permanent or transient: IUGR; rarely developmental delay; responsive to sulfonylureas
6q24 (PLAGL1, AD for paternal Transient: IUGR; macroglossia; umbilical hernia; mechanisms include UPD6, paternal
HYMA1) duplications duplication, or maternal methylation defect; may be treatable with medications
s
other than insulin
GATA6 AD Permanent: pancreatic hypoplasia; cardiac malformations; pancreatic exocrine
sA
insufficiency; insulin requiring
EIF2AK3 AR Permanent: Wolcott-Rallison syndrome: epiphyseal dysplasia; pancreatic exocrine
insufficiency; insulin requiring
EIF2B1 AD Permanent diabetes: can be associated with fluctuating liver function (172)
X-linked Permanent: immunodysregulation, polyendocrinopathy, enteropathy X-linked (IPEX)
FOXP3
te
syndrome: autoimmune diabetes, autoimmune thyroid disease, exfoliative dermatitis;
insulin requiring
be
erine growth restriction; OGTT, oral glucose tolerance test; UPD6, uniparental
AD, autosomal dominant; AR, autosomal recessive; IUGR, intrauterine
disomy of chromosome 6; 2-h PG, 2-h plasma glucose.
ia
Individuals with HNF1A- or HNF4A-MODY with diabetes not characteristic of type 1 be incorrectly diagnosed with type 1
usually respond well to low doses of sul- or type 2 diabetes, although admit- or type 2 diabetes, leading to subopti-
fonylureas, which are considered first-line atypical diabetes”
tedly, “atypical diabetes is becoming mal, even potentially harmful, treatment
nD
cysts and diabetes [RCAD] syndrome). of autoantibodies for type 1 diabetes studies have shown that no complications
Other extremely rare forms of MODY precludes further testing for mono- ensue in the absence of glucose-lowering
have been reported to involve other genic diabetes, but the presence of auto- therapy (189). The risks of microvascular
antibodies in people with monogenic and macrovascular complications with
transcription factor genes, including PDX1
diabetes has been reported (185). Indi-
er
• Diabetes without typical features of GESTATIONAL DIABETES MELLITUS found to have prediabetes
type 1 or type 2 diabetes (negative should receive intensive life-
Recommendations
diabetes-associated autoantibodies, no style interventions and/or met-
2.26a In individuals who are plan-
obesity, lacking other metabolic fea- formin to prevent diabetes. A
ning pregnancy, screen those
tures, especially with strong family
with risk factors B and con-
history of diabetes)
n
sider testing all individuals of
• Stable, mild fasting hyperglycemia Definition
childbearing potential for un-
(100–150 mg/dL [5.5–8.5 mmol/L]),
tio
diagnosed diabetes. E For many years, GDM was defined de ned as
stable A1C between 5.6% and 7.6% any degree of glucose intolerance that
2.26b Before 15 weeks of gestation,
(between 38 and 60 mmol/mol), es- was firstrst recognized during pregnancy
test individuals with risk factors
pecially if no obesity (86), regardless of the degree of hyper-
B and consider testing all indi-
c ia
viduals E for undiagnosed dia- glycemia. This defi de finition facilitated a
definition
PANCREATIC DIABETES OR DIABETES betes at the first prenatal uniform strategy for detection and clas-
IN THE CONTEXT OF DISEASE OF visit using standard diagnos- sification
cation of GDM, but this de definition
THE EXOCRINE PANCREAS tic criteria if not screened has serious limitations (203). First, the
so
Pancreatic diabetes includes both struc- preconception. best available evidence reveals that many
tural and functional loss of glucose- 2.26c Individuals of childbearing cases of GDM represent preexisting
normalizing insulin secretion in the con- potential identified as having hyperglycemia that is detected by rou-
text of exocrine pancreatic dysfunction diabetes should be treated tine screening in pregnancy, as routine
s
and is commonly misdiagnosed as type 2 as such. A screening is not widely performed in
sA
diabetes. Hyperglycemia due to general 2.26d Before 15 weeks of gestation, nonpregnant individuals of reproduc-
pancreatic dysfunction has been called screen for abnormal glucose tive age. It is the severity of hypergly-
“type 3c diabetes,” and, more recently, metabolism to identify individu- cemia that is clinically important with
diabetes in the context of disease of als who are at higher risk of regard to both short- and long-term
the exocrine pancreas has been termed
te
adverse pregnancy and neona- maternal and fetal risks.
pancreoprivic diabetes (1). The diverse tal outcomes, are more likely The ongoing epidemic of obesity and
set of etiologies includes pancreatitis to need insulin, and are at diabetes has led to more type 2 diabe-
be
(acute and chronic), trauma or pancrea- high risk of a later gestational tes in people of reproductive age, with
tectomy, neoplasia, cystic fibrosis (ad- diabetes mellitus diagnosis. B an increase in the number of pregnant
dressed elsewhere in this chapter), Treatment may provide some individuals with undiagnosed type 2 dia-
hemochromatosis, fibrocalculous pan- benefit. E betes in early pregnancy (204–206). Ide-
ia
creatopathy, rare genetic disorders (193), 2.26e Screen for early abnormal glu- ally, undiagnosed diabetes should be
and idiopathic forms (1); as such, pancre- cose metabolism using fasting identified preconception in individuals
nD
atic diabetes is the preferred umbrella glucose of 110–125 mg/dL with risk factors or in high-risk popula-
terminology. (6.1 mmol/L) or A1C 5.9–6.4% tions (207–212), as the preconception
Pancreatitis, even a single bout, can (41–47 mmol/mol). B care of people with preexisting diabetes
lead to postpancreatitis diabetes melli- 2.27 Screen for gestational diabe- results in lower A1C and reduced risk of
tes mellitus at 24–28 weeks birth defects, preterm delivery, perinatal
ica
monoclonal fecal elastase 1 test or direct detected earlier in the current early screening at <15 weeks of gestation
function tests), pathological pancreatic pregnancy. A for undiagnosed diabetes may be consid-
imaging (endoscopic ultrasound, MRI, 2.28 Screen individuals with gesta- ered over selective screening (Table
m
computed tomography), and absence tional diabetes mellitus for pre- 2.3), particularly in populations with
of type 1 diabetes-associated autoim- diabetes or diabetes at 4–12 high prevalence of risk factors and un-
weeks postpartum, using the diagnosed diabetes in people of child-
©A
diabetes by the standard diagnostic crite- gestation or later. See Recommendation GDM diagnosis (Table 2.7) can be ac-
ria used outside of pregnancy should be 2.3 above. complished with either of two strategies:
classified as having diabetes complicat- GDM is often indicative of underlying
ing pregnancy (most often type 2 diabe- b-cell dysfunction (225), which confers 1. The “one-step” 75-g OGTT derived
tes, rarely type 1 diabetes or monogenic marked increased risk for later develop- from the IADPSG criteria, or
diabetes) and managed accordingly. ment of diabetes, generally but not al- 2. The older “two-step” approach with
n
Early abnormal glucose metabolism, ways type 2 diabetes, in the mother after a 50-g (nonfasting) screen followed
defined as fasting glucose threshold of delivery (226,227). As effective prevention by a 100-g OGTT for those who
tio
110 mg/dL (6.1 mmol/L) or an A1C of interventions are available (228,229), screen positive based on the work
5.9% (39 mmol/mol), may identify in- individuals diagnosed with GDM should of Carpenter-Coustan’s
Carpenter-Coustan s interpretation
dividuals who are at higher risk of ad- receive lifelong screening for prediabetes O Sullivan and Mahan
of the older O’Sullivan
c ia
verse pregnancy and neonatal outcomes to allow interventions to reduce diabetes (232) criteria.
(preeclampsia, macrosomia, shoulder dys- risk and for type 2 diabetes to allow
tocia, perinatal death), are more likely to treatment at the earliest possible time Different diagnostic criteria will identify dif-
need insulin treatment, and are at high (230). ferent degrees of maternal hyperglycemia
so
risk of a later GDM diagnosis (214–220). and maternal/fetal risk, leading some ex-
An A1C threshold of 5.7% has not been Diagnosis perts to debate, and disagree on, opti-
shown to be associated with adverse peri- GDM carries risks for the mother, fetus, mal strategies for the diagnosis of GDM.
natal outcomes (221,222). and neonate. The Hyperglycemia and
s
If early screening is negative, individ- Adverse Pregnancy Outcome (HAPO) One-Step Strategy
sA
uals should be rescreened for GDM be- study (231), a large-scale multinational The IADPSG de defined diagnostic cut points
tween 24 and 28 weeks of gestation cohort study completed by more than for GDM as the average fasting, 1-h, and
(see Section 15, “Management of 23,000 pregnant individuals, demon- 2-h PG values during a 75-g OGTT in in-
Diabetes in Pregnancy”). The Interna- strated that risk of adverse maternal, fe- dividuals at 24–28 weeks of gestation
tional Association of the Diabetes and
te
tal, and neonatal outcomes continuously who participated in the HAPO study at
Pregnancy Study Groups (IADPSG) GDM increased as a function of maternal gly- which odds for adverse outcomes
diagnostic criteria for the 75-g OGTT, as cemia at 24–28 28 weeks of gestation, reached 1.75 times the estimated odds
be
well as the GDM screening and diagnos- even within ranges previously consid- of these outcomes at the mean fasting,
tic criteria used in the two-step ap- ered normal for pregnancy. For most 1-h, and 2-h PG levels of the study pop-
proach, were not derived from data in complications, there was no threshold ulation. This one-step strategy was
the first half of pregnancy and should for risk. These results have led to careful anticipated to significantly increase the inci-
ia
not be used for early screening (223). To reconsideration of the diagnostic criteria dence of GDM (from 5–6% to 15–20%),
date, most randomized controlled trials for GDM. primarily because only one abnormal value,
nD
gestation. Step 1: Perform a 50-g GLT (nonfasting), with plasma glucose measurement at 1 h, at
24–28 weeks of gestation in individuals not previously diagnosed with diabetes.
Both the fasting glucose and A1C are
If the plasma glucose level measured 1 h after the load is $130, 135, or 140 mg/dL
low-cost tests. An advantage of the A1C
©A
not two, became sufficient to make the one-step method using IADPSG criteria ver- Two-Step Strategy
diagnosis (233). Many regional studies sus the two-step method using a 1-h 50-g In 2013, the NIH convened a consensus
have investigated the impact of adopting glucose loading test (GLT) and, if posi- development conference to consider
the IADPSG criteria on prevalence and tive, a 3-h OGTT by Carpenter-Coustan diagnostic criteria for diagnosing GDM
have seen a roughly one- to threefold in- criteria identified twice as many individu- (246). The 15-member panel had repre-
crease (234). The anticipated increase als with GDM using the one-step method sentatives from obstetrics and gynecol-
n
in the incidence of GDM could have a compared with the two-step method. ogy, maternal-fetal medicine, pediatrics,
substantial impact on costs and medical diabetes research, biostatistics, and other
tio
Despite treating more individuals for
infrastructure needs and has the po- related fields.
elds. The panel recommended a
GDM using the one-step method, there
tential to “medicalize” pregnancies pre- two-step approach to screening that used
was no difference in pregnancy and peri-
viously categorized as normal. A recent a 1-h 50-g GLT followed by a 3-h 100-g
natal complications (239). However, con-
c ia
follow-up study of individuals participating OGTT for those who screened positive.
cerns have been raised about sample
in a blinded study of pregnancy OGTTs The American College of Obstetricians
size estimates and unanticipated sub-
found that 11 years after their pregnan- and Gynecologists (ACOG) recommends
optimal engagement with the protocol any of the commonly used thresholds
cies, individuals who would have been
with regard to screening and treatment. of 130, 135, or 140 mg/dL for the 1-h
diagnosed with GDM by the one-step
so
For example, in the two-step group, 165 50-g GLT (247). Updated from 2014, a
approach, as compared with those
participants who did not get counted as 2021 U.S. Preventive Services Task Force
without, were at 3.4-fold higher risk of
developing prediabetes and type 2 di- having GDM were treated for isolated el- systematic review continues to conclude
s
abetes and had children with a higher evated fasting glucose >95 95 mg/dL (240). that one-step versus two-step screening
risk of obesity and increased body fat, The high prevalence of prediabetes in is associated with increased likelihood of
sA
suggesting that the larger group of indi- people of childbearing age may sup- GDM (11.5% vs. 4.9%) but without im-
viduals identified by the one-step ap- port the more inclusive IADPSG criteria. proved health outcomes. It reports that
proach would benefit from the increased NHANES data demonstrate a 21.5% preva- the oral glucose challenge test using 140
screening for diabetes and prediabetes lence of prediabetes in people of reproduc- or 135 mg/dL thresholds had sensitivities
te
that would accompany a history of tive age 20–44
44 years, which is comparable of 82% and 93% and specificities of 82%
GDM (235,236). The ADA recommends to or higher than the prevalence of GDM di- and 79%, respectively, against Carpenter-
the IADPSG diagnostic criteria with the agnosed by the one-step method (241). Coustan criteria. Fasting plasma glucose
be
intent of optimizing gestational outcomes The one-step method identifies
identi the cutoffs of 85 mg/dL or 90 mg/dL had
because these criteria are the only ones long-term risks of maternal prediabetes sensitivities of 88% and 81% and specif-
based on pregnancy outcomes rather and diabetes and offspring abnormal glu- icities of 73% and 82%, respectively,
ia
than end points such as prediction of cose metabolism and adiposity. Post hoc against Carpenter-Coustan criteria (248).
subsequent maternal diabetes. GDM in individuals diagnosed by the one- The use of A1C at 24–28 weeks of gesta-
The expected benefits ts of using IADPSG step method in the HAPO cohort was as- tion as a screening test for GDM does
nD
criteria to the offspring are inferred from sociated with higher prevalence of IGT; not function as well as the GLT (249).
intervention trials that focused on individ- higher 30-min, 1-h, and 2-h glucoses dur- Key factors cited by the NIH panel in
uals with lower levels of hyperglycemia ing the OGTT; and reduced insulin sensi- their decision-making process were the
than identified ed using older GDM diag- tivity and oral disposition index in their lack of clinical trial data demonstrating
ica
nostic criteria. Those trials found mod- offspring at 10–14 years of age compared the benefits of the one-step strategy
est benefits ts including reduced rates of with offspring of mothers without GDM. and the potential negative consequences
large-for-gestational-age births and pre- of identifying a large group of individu-
Associations of mother’s fasting, 1-h, and
eclampsia (237,238). It is important to als with GDM, including medicalization
2-h values on the 75-g OGTT were contin-
note that 80–90% 90% of participants being of pregnancy with increased health care
er
FUS) data demonstrate that neonatal adi- plish for many individuals. Treatment
The OGTT glucose cutoffs in these two
trials overlapped the thresholds recom- posity and fetal hyperinsulinemia (cord of higher-threshold maternal hyper-
C-peptide), both higher across the contin-
©A
mended by the IADPSG, and in one trial glycemia, as identified by the two-step
(238), the 2-h PG threshold (140 mg/dL uum of maternal hyperglycemia, are me- approach, reduces rates of neonatal mac-
[7.8 mmol/L]) was lower than the diators of childhood body fat (243). rosomia, large-for-gestational-age births
cutoff recommended by the IADPSG Data are lacking on how the treatment (250), and shoulder dystocia without
(153 mg/dL [8.5 mmol/L]). of mother’s hyperglycemia in pregnancy increasing small-for-gestational-age births.
No randomized controlled trials of affects her offspring’s risk for obesity, dia- ACOG currently supports the two-step ap-
treating versus not treating GDM diag- betes, and other metabolic disorders. Ad- proach but notes that one elevated value,
nosed by the IADPSG criteria but not ditional well-designed clinical studies are as opposed to two, may be used for the
the Carpenter-Coustan criteria have been needed to determine the optimal in- diagnosis of GDM (247). If this approach
published to date. However, a recent tensity of monitoring and treatment of is implemented, the incidence of GDM
randomized trial of testing for GDM individuals with GDM diagnosed by the by the two-step strategy will likely
at 24–28 weeks of gestation by the one-step strategy (244,245). increase markedly. ACOG recommends