Professional Documents
Culture Documents
Standards of Medical Care in Diabetesd2021: 2. Classi Fication and Diagnosis of Diabetes
Standards of Medical Care in Diabetesd2021: 2. Classi Fication and Diagnosis of Diabetes
CLASSIFICATION
Diabetes can be classified into the following general categories:
This section reviews most common forms of diabetes but is not comprehensive. For
additional information, see the American Diabetes Association (ADA) position
statement “Diagnosis and Classification of Diabetes Mellitus” (1).
Suggested citation: American Diabetes Associa-
Type 1 diabetes and type 2 diabetes are heterogeneous diseases in which clinical tion. 2. Classification and diagnosis of diabetes:
presentation and disease progression may vary considerably. Classification is Standards of Medical Care in Diabetesd2021.
important for determining therapy, but some individuals cannot be clearly classified Diabetes Care 2021;44(Suppl. 1):S152S33
as having type 1 or type 2 diabetes at the time of diagnosis. The traditional paradigms © 2020 by the American Diabetes Association.
of type 2 diabetes occurring only in adults and type 1 diabetes only in children are no Readers may use this article as long as the work is
properly cited, the use is educational and not for
longer accurate, as both diseases occur in both age-groups. Children with type 1 profit, and the work is not altered. More infor-
diabetes typically present with the hallmark symptoms of polyuria/polydipsia, and mation is available at https://www.diabetesjournals
approximately one-third present with diabetic ketoacidosis (DKA) (2). The onset of .org/content/license.
S16 Classification and Diagnosis of Diabetes Diabetes Care Volume 44, Supplement 1, January 2021
type 1 diabetes may be more variable in and A1C levels rise well before the clin- tolerance (IGT) with or without elevated
adults; they may not present with the ical onset of diabetes, making diagnosis fasting glucose, not for individuals with
classic symptoms seen in children and feasible well before the onset of DKA. isolated impaired fasting glucose (IFG)
may experience temporary remission Three distinct stages of type 1 diabetes or for those with prediabetes defined by
from the need for insulin (3–5). Occa- can be identified (Table 2.1) and serve A1C criteria.
sionally, patients with type 2 diabetes as a framework for future research and The same tests may be used to screen
may present with DKA (6), particularly regulatory decision-making (8,10). There for and diagnose diabetes and to detect
ethnic and racial minorities (7). It is is debate as to whether slowly progres- individuals with prediabetes (Table 2.2
important for the provider to realize sive autoimmune diabetes with an adult and Table 2.5) (19). Diabetes may be
that classification of diabetes type is not onset should be termed latent autoim- identified anywhere along the spectrum
always straightforward at presentation mune diabetes in adults (LADA) or type 1 of clinical scenariosdin seemingly low-
and that misdiagnosis is common (e.g., diabetes. The clinical priority is aware- risk individuals who happen to have glu-
adults with type 1 diabetes misdiag- ness that slow autoimmune b-cell de- cose testing, in individuals tested based
Even in the absence of hemoglobin or in two different test samples, this also
and may serve as an indication for
variants, A1C levels may vary with race/ confirms the diagnosis. On the other
intervention in the setting of a
ethnicity independently of glycemia (28–30). hand, if a patient has discordant results
clinical trial. B
For example, African Americans may from two different tests, then the test
have higher A1C levels than non-Hispanic result that is above the diagnostic cut
Whites with similar fasting and postglu- point should be repeated, with careful Immune-Mediated Diabetes
cose load glucose levels (31). Though consideration of the possibility of A1C This form, previously called “insulin-
conflicting data exists, African Ameri- assay interference. The diagnosis is made dependent diabetes” or “juvenile-onset
cans may also have higher levels of on the basis of the confirmed test. For diabetes,” accounts for 5–10% of diabetes
fructosamine and glycated albumin example, if a patient meets the diabetes and is due to cellular-mediated autoim-
and lower levels of 1,5-anhydroglucitol, criterion of the A1C (two results $6.5% mune destruction of the pancreatic
suggesting that their glycemic burden [48 mmol/mol]) but not FPG (,126 mg/ b-cells. Autoimmune markers include islet
(particularly postprandially) may be dL [7.0 mmol/L]), that person should cell autoantibodies and autoantibodies
Table 2.3—Criteria for testing for diabetes or prediabetes in asymptomatic adults of developing diabetes between 25% and
1. Testing should be considered in adults with overweight or obesity (BMI $25 kg/m2 or $23 50% and a relative risk 20 times higher
kg/m2 in Asian Americans) who have one or more of the following risk factors: compared with A1C of 5.0% (31 mmol/
c First-degree relative with diabetes mol) (60). In a community-based study
c High-risk race/ethnicity (e.g., African American, Latino, Native American, Asian American, of African American and non-Hispanic
Pacific Islander) White adults without diabetes, baseline
c History of CVD
c Hypertension ($140/90 mmHg or on therapy for hypertension)
A1C was a stronger predictor of sub-
c HDL cholesterol level ,35 mg/dL (0.90 mmol/L) and/or a triglyceride level .250 mg/dL
sequent diabetes and cardiovascular
(2.82 mmol/L) events than fasting glucose (61). Other
c Women with polycystic ovary syndrome analyses suggest that A1C of 5.7%
c Physical inactivity (39 mmol/mol) or higher is associated
c Other clinical conditions associated with insulin resistance (e.g., severe obesity, acanthosis with a diabetes risk similar to that of the
nigricans)
high-risk participants in the Diabetes
excellent care provided to patients in the and later treatment after clinical di- type 2 diabetes (79); moreover, screen-
routine care group and the lack of an agnoses. Computer simulation model- ing, beginning at age 30 or 45 years and
unscreened control arm limited the au- ing studies suggest that major benefits independent of risk factors, may be
thors’ ability to determine whether are likely to accrue from the early di- cost-effective (,$11,000 per quality-
screening and early treatment improved agnosis and treatment of hyperglyce- adjusted life year gainedd2010 mod-
outcomes compared with no screening mia and cardiovascular risk factors in eling data) (80). Cost-effectiveness of
care.diabetesjournals.org Classification and Diagnosis of Diabetes S23
screening has been reinforced in cohort recommended for diagnosis and may i.e., it may fail to reach the groups most
studies (81,82). present challenges for monitoring (24). at risk and inappropriately test those at
Additional considerations regarding In those with prediabetes, weight loss very low risk or even those who have
testing for type 2 diabetes and predia- through healthy nutrition and physical already been diagnosed (93).
betes in asymptomatic patients include activity may reduce the progression to-
Screening in Dental Practices
the following. ward diabetes. Among patients with HIV
Because periodontal disease is associ-
Age
and diabetes, preventive health care
ated with diabetes, the utility of screen-
Age is a major risk factor for diabetes. using an approach used in patients with-
ing in a dental setting and referral to
Testing should begin at no later than age out HIV is critical to reduce the risks of
primary care as a means to improve the
45 years for all patients. Screening should microvascular and macrovascular com-
diagnosis of prediabetes and diabetes
be considered in adults of any age with plications. Diabetes risk is increased with
has been explored (94–96), with one
overweight or obesity and one or more certain PIs and NRTIs. New-onset diabe-
study estimating that 30% of patients
tes is estimated to occur in more than
interference are appropriate for chil- study) would detect more than 90% Diabetes’s 2014 clinical practice consen-
dren with hemoglobinopathies, the of cases and reduce patient screening sus guidelines (102).
ADA continues to recommend A1C burden (103,104). Ongoing studies are
for diagnosis of type 2 diabetes in underway to validate this approach. Re-
this cohort to decrease barriers to POSTTRANSPLANTATION
gardless of age, weight loss or failure of
DIABETES MELLITUS
screening (100,101). expected weight gain is a risk for CFRD
and should prompt screening (103,104). Recommendations
The Cystic Fibrosis Foundation Patient 2.19 Patients should be screened af-
CYSTIC FIBROSIS–RELATED
Registry (105) evaluated 3,553 cystic ter organ transplantation for
DIABETES
fibrosis patients and diagnosed 445 hyperglycemia, with a formal
Recommendations (13%) with CFRD. Early diagnosis and diagnosis of posttransplanta-
2.15 Annual screening for cystic treatment of CFRD was associated with tion diabetes mellitus being best
fibrosis–related diabetes (CFRD) preservation of lung function. The Eu- made once a patient is stable on
family history of diabetes, etc.) as well reported that metformin was safe to of age, whereas autoimmune type 1 di-
as transplant-specific factors, such as use in renal transplant recipients (128), abetes rarely occurs before 6 months
use of immunosuppressant agents (119). but its safety has not been determined of age. Neonatal diabetes can either be
Whereas posttransplantation hypergly- in other types of organ transplant. Thia- transient or permanent. Transient dia-
cemia is an important risk factor for zolidinediones have been used success- betes is most often due to overexpres-
subsequent PTDM, a formal diagnosis fully in patients with liver and kidney sion of genes on chromosome 6q24, is
of PTDM is optimally made once the transplants, but side effects include fluid recurrent in about half of cases, and may
patient is stable on maintenance immu- retention, heart failure, and osteopenia be treatable with medications other than
nosuppression and in the absence of (129, 130). Dipeptidyl peptidase 4 inhib- insulin. Permanent neonatal diabetes is
acute infection (117–120). In a recent itors do not interact with immunosup- most commonly due to autosomal dom-
study of 152 heart transplant recipients, pressant drugs and have demonstrated inant mutations in the genes encoding
38% had PTDM at 1 year. Risk factors for safety in small clinical trials (131,132). the Kir6.2 subunit (KCNJ11) and SUR1
PTDM included elevated BMI, discharge Well-designed intervention trials exam- subunit (ABCC8) of the b-cell KATP chan-
and uterine malformations (renal cysts autoantibodies for type 1 diabetes pre- mutations where multiple studies have
and diabetes [RCAD] syndrome). Other cludes further testing for monogenic shown that no complications ensue in
extremely rare forms of MODY have been diabetes, but the presence of auto- the absence of glucose-lowering therapy
reported to involve other transcription antibodies in patients with mono- (150). Genetic counseling is recommen-
factor genes including PDX1 (IPF1) and genic diabetes has been reported (146). ded to ensure that affected individuals
NEUROD1. Individuals in whom monogenic diabetes understand the patterns of inheri-
is suspected should be referred to a tance and the importance of a correct
Diagnosis of Monogenic Diabetes specialist for further evaluation if avail- diagnosis.
A diagnosis of one of the three most able, and consultation is available from The diagnosis of monogenic diabe-
common forms of MODY, including GCK- several centers. Readily available com- tes should be considered in children
MODY, HNF1A-MODY, and HNF4A-MODY, mercial genetic testing following the and adults diagnosed with diabetes in
allows for more cost-effective therapy criteria listed below now enables a early adulthood with the following
(no therapy for GCK-MODY; sulfonylur- cost-effective (147), often cost-saving, findings:
eas as first-line therapy for HNF1A-MODY genetic diagnosis that is increasingly
and HNF4A-MODY). Additionally, diag- supported by health insurance. A bio- c Diabetes diagnosed within the first 6
nosis can lead to identification of other marker screening pathway such as months of life (with occasional cases
affected family members. Genetic screen- the combination of urinary C-peptide/ presenting later, mostly INS and ABCC8
ing is increasingly available and cost- creatinine ratio and antibody screening mutations) (134,151)
effective (138,140). may aid in determining who should get c Diabetes without typical features of
A diagnosis of MODY should be genetic testing for MODY (148). It is type 1 or type 2 diabetes (negative
considered in individuals who have critical to correctly diagnose one of diabetes-associated autoantibodies,
atypical diabetes and multiple family the monogenic forms of diabetes be- nonobese, lacking other metabolic
members with diabetes not characteris- cause these patients may be incorrectly features, especially with strong fam-
tic of type 1 or type 2 diabetes, although diagnosed with type 1 or type 2 diabetes, ily history of diabetes)
admittedly “atypical diabetes” is becom- leading to suboptimal, even potentially c Stable, mild fasting hyperglycemia
ing increasingly difficult to precisely de- harmful, treatment regimens and delays (100–150 mg/dL [5.5–8.5 mmol/L]),
fine in the absence of a definitive set of in diagnosing other family members stable A1C between 5.6% and 7.6%
tests for either type of diabetes (135–137, (149). The correct diagnosis is espe- (between 38 and 60 mmol/mol), es-
139–145). In most cases, the presence of cially critical for those with GCK-MODY pecially if nonobese
care.diabetesjournals.org Classification and Diagnosis of Diabetes S27
positive, based on the work of Car- “medicalize” pregnancies previously cat- GDM in these two randomized controlled
penter and Coustan’s interpretation egorized as normal. A recent follow-up trials could be managed with lifestyle
of the older OʼSullivan (179) criteria. study of women participating in a blinded therapy alone. The OGTT glucose cutoffs
study of pregnancy OGTTs found that in these two trials overlapped with the
Different diagnostic criteria will iden- 11 years after their pregnancies, women thresholds recommended by the IADPSG,
tify different degrees of maternal hyper- who would have been diagnosed with and in one trial (185), the 2-h PG thresh-
glycemia and maternal/fetal risk, leading GDM by the one-step approach, as old (140 mg/dL [7.8 mmol/L]) was lower
some experts to debate, and disagree on, compared with those without, were at than the cutoff recommended by the
optimal strategies for the diagnosis of 3.4-fold higher risk of developing pre- IADPSG (153 mg/dL [8.5 mmol/L]). No
GDM. diabetes and type 2 diabetes and had randomized controlled trials of treating
children with a higher risk of obesity and versus not treating GDM diagnosed by
One-Step Strategy increased body fat, suggesting that the the IADPSG criteria but not the Carpenter-
The IADPSG defined diagnostic cut points larger group of women identified by the Coustan criteria have been published
the benefits of the one-step strategy and infrastructure, and importance of cost application to the discrimination between type 1
the potential negative consequences of considerations). and type 2 diabetes in young adults. Diagn Progn
Res 2020;4:6
identifying a large group of women with As the IADPSG criteria (“one-step 10. Insel RA, Dunne JL, Atkinson MA, et al.
GDM, including medicalization of preg- strategy”) have been adopted interna- Staging presymptomatic type 1 diabetes: a sci-
nancy with increased health care utiliza- tionally, further evidence has emerged to entific statement of JDRF, the Endocrine Society,
tion and costs. Moreover, screening with support improved pregnancy outcomes and the American Diabetes Association. Diabetes
a 50-g GLT does not require fasting and is with cost savings (197), and IADPSG may Care 2015;38:1964–1974
11. Zhu Y, Qian L, Liu Q, et al. Glutamic acid
therefore easier to accomplish for many be the preferred approach. Data com- decarboxylase autoantibody detection by elec-
women. Treatment of higher-threshold paring population-wide outcomes with trochemiluminescence assay identifies latent
maternal hyperglycemia, as identi- one-step versus two-step approaches autoimmune diabetes in adults with poor islet
fied by the two-step approach, reduces have been inconsistent to date (198,199). function. Diabetes Metab J 2020;44:260–266
12. Lynam A, McDonald T, Hill A, et al. De-
rates of neonatal macrosomia, large-for- In addition, pregnancies complicated by
velopment and validation of multivariable clin-
gestational-age births (192), and shoulder GDM per the IADPSG criteria, but not
26. Lacy ME, Wellenius GA, Sumner AE, et al. 41. Megia A, Näf S, Herranz L, et al. The type 1 diabetes in the Diabetes Prevention Trial-
Association of sickle cell trait with hemoglobin usefulness of HbA1c in postpartum reclassifi- Type 1. Diabetes Care 2009;32:2269–2274
A1c in African Americans. JAMA 2017;317:507– cation of gestational diabetes. BJOG 2012;119: 56. Jacobsen LM, Larsson HE, Tamura RN, et al.;
515 891–894 TEDDY Study Group. Predicting progression to
27. Wheeler E, Leong A, Liu C-T, et al.; EPIC-CVD 42. Welsh KJ, Kirkman MS, Sacks DB. Role of type 1 diabetes from ages 3 to 6 in islet auto-
Consortium; EPIC-InterAct Consortium; Lifelines glycated proteins in the diagnosis and man- antibody positive TEDDY children. Pediatr Di-
Cohort Study. Impact of common genetic deter- agement of diabetes: research gaps and fu- abetes 2019;20:263–270
minants of hemoglobin A1c on type 2 diabetes ture directions. Diabetes Care 2016;39:1299– 57. American Diabetes Association. Diagnosis
risk and diagnosis in ancestrally diverse popula- 1306 and classification of diabetes mellitus. Diabetes
tions: a transethnic genome-wide meta-analysis. 43. Kim C, Bullard KM, Herman WH, Beckles GL. Care 2011;34(Suppl. 1):S62–S69
PLoS Med 2017;14:e1002383 Association between iron deficiency and A1C 58. Genuth S, Alberti KGMM, Bennett P, et al.;
28. Ziemer DC, Kolm P, Weintraub WS, et al. Levels among adults without diabetes in the Expert Committee on the Diagnosis and Classi-
Glucose-independent, black-white differences in National Health and Nutrition Examination Sur- fication of Diabetes Mellitus. Follow-up report on
hemoglobin A1c levels: a cross-sectional analysis vey, 1999–2006. Diabetes Care 2010;33:780–785 the diagnosis of diabetes mellitus. Diabetes Care
of 2 studies. Ann Intern Med 2010;152:770–777 44. Selvin E, Wang D, Matsushita K, Grams ME, 2003;26:3160–3167
71. Yoshino M, Kayser BD, Yoshino J, et al. Effects intervention strategies. Lancet 2004;363:157– 101. Wu E-L, Kazzi NG, Lee JM. Cost-effective-
of diet versus gastric bypass on metabolic func- 163 ness of screening strategies for identifying pe-
tion in diabetes. N Engl J Med 2020;383:721–732 86. Menke A, Casagrande S, Geiss L, Cowie CC. diatric diabetes mellitus and dysglycemia. JAMA
72. Cresci B, Cosentino C, Monami M, Mannucci Prevalence of and trends in diabetes among Pediatr 2013;167:32–39
E. Metabolic surgery for the treatment of type 2 adults in the United States, 1988-2012. JAMA 102. Moran A, Pillay K, Becker D, Granados A,
diabetes: a network meta-analysis of randomized 2015;314:1021–1029 Hameed S, Acerini CL. ISPAD Clinical Practice
controlled trials. Diabetes Obes Metab 2020;22: 87. Centers for Disease Control and Prevention. Consensus Guidelines 2018: management of cystic
1378–1387 National Diabetes Statistics Report: estimates of fibrosis-related diabetes in children and adoles-
73. Palladino R, Tabak AG, Khunti K, et al. As- diabetes and its burden in the United States, cents. Pediatr Diabetes 2018;19(Suppl. 27):64–74
sociation between pre-diabetes and microvas- 2020. Accessed 28 October 2020. Available from 103. Gilmour JA. Response to the Letter to the
cular and macrovascular disease in newly https://www.cdc.gov/diabetes/data/statistics/ Editor From Dr. Boudreau et al, “Validation of a
diagnosed type 2 diabetes. BMJ Open Diabetes statistics-report.html Stepwise Approach Using Glycated Hemoglobin
Res Care 2020;8:e001061 88. Chiu M, Austin PC, Manuel DG, Shah BR, Tu Levels to Reduce the Number of Required Oral
74. Perreault L, Pan Q, Aroda VR, et al.; Di- JV. Deriving ethnic-specific BMI cutoff points for Glucose Tolerance Tests to Screen for Cystic
abetes Prevention Program Research Group. assessing diabetes risk. Diabetes Care 2011;34: Fibrosis-Related Diabetes in Adults”. Can J Di-
116. Ramirez SC, Maaske J, Kim Y, et al. The 133. Carmody D, Støy J, Greeley SA, Bell GI, 148. Shields BM, Shepherd M, Hudson M, et al.;
association between glycemic control and clinical Philipson LH. A clinical guide to monogenic di- UNITED study team. Population-based assess-
outcomes after kidney transplantation. Endocr abetes. In Genetic Diagnosis of Endocrine Dis- ment of a biomarker-based screening pathway to
Pract 2014;20:894–900 orders. 2nd ed. Weiss RE, Refetoff S, Eds. aid diagnosis of monogenic diabetes in young-
117. Thomas MC, Moran J, Mathew TH, Russ GR, Philadelphia, PA, Elsevier, 2016 onset patients. Diabetes Care 2017;40:1017–
Rao MM. Early peri-operative hyperglycaemia 134. De Franco E, Flanagan SE, Houghton JAL, 1025
and renal allograft rejection in patients without et al. The effect of early, comprehensive genomic 149. Hattersley A, Bruining J, Shield J, Njolstad P,
diabetes. BMC Nephrol 2000;1:1 testing on clinical care in neonatal diabetes: an Donaghue KC. The diagnosis and management of
118. Chakkera HA, Weil EJ, Castro J, et al. Hy- international cohort study. Lancet 2015;386: monogenic diabetes in children and adolescents.
perglycemia during the immediate period after 957–963 Pediatr Diabetes 2009;10(Suppl. 12):33–42
kidney transplantation. Clin J Am Soc Nephrol 135. Sanyoura M, Letourneau L, Knight Johnson 150. Rubio-Cabezas O, Hattersley AT, Njølstad
2009;4:853–859 AE, et al. GCK-MODY in the US Monogenic Di- PR, et al.; International Society for Pediatric and
119. Wallia A, Illuri V, Molitch ME. Diabetes care after abetes Registry: description of 27 unpublished Adolescent Diabetes. ISPAD Clinical Practice
transplant: definitions, risk factors, and clinical man- variants. Diabetes Res Clin Pract 2019;151:231– Consensus Guidelines 2014. The diagnosis and
agement. Med Clin North Am 2016;100:535–550 236 management of monogenic diabetes in children
164. Rosenn B, Miodovnik M, Combs CA, Khoury prevent gestational diabetes mellitus and im- 190. Donovan L, Hartling L, Muise M, Guthrie A,
J, Siddiqi TA. Glycemic thresholds for spontane- prove pregnancy outcome in overweight and Vandermeer B, Dryden DM. Screening tests for
ous abortion and congenital malformations in obese pregnant women. Am J Obstet Gynecol gestational diabetes: a systematic review for the
insulin-dependent diabetes mellitus. Obstet Gy- 2017;216:340–351 U.S. Preventive Services Task Force. Ann Intern
necol 1994;84:515–520 178. Metzger BE, Lowe LP, Dyer AR, et al.; HAPO Med 2013;159:115–122
165. Schaefer UM, Songster G, Xiang A, Study Cooperative Research Group. Hyperglyce- 191. Khalafallah A, Phuah E, Al-Barazan AM,
Berkowitz K, Buchanan TA, Kjos SL. Congenital mia and adverse pregnancy outcomes. N Engl J et al. Glycosylated haemoglobin for screening
malformations in offspring of women with hy- Med 2008;358:1991–2002 and diagnosis of gestational diabetes mellitus.
perglycemia first detected during pregnancy. Am 179. O’Sullivan JB, Mahan CM. Criteria for the BMJ Open 2016;6:e011059
J Obstet Gynecol 1997;177:1165–1171 oral glucose tolerance test in pregnancy. Diabe- 192. Horvath K, Koch K, Jeitler K, et al. Effects of
166. Poltavskiy E, Kim DJ, Bang H. Comparison of tes 1964;13:278–285 treatment in women with gestational diabetes
screening scores for diabetes and prediabetes. 180. Sacks DA, Hadden DR, Maresh M, et al.; mellitus: systematic review and meta-analysis.
Diabetes Res Clin Pract 2016;118:146–153 HAPO Study Cooperative Research Group. Fre- BMJ 2010;340:c1395
167. Feig DS, Hwee J, Shah BR, Booth GL, quency of gestational diabetes mellitus at col- 193. Carpenter MW, Coustan DR. Criteria for
Bierman AS, Lipscombe LL. Trends in incidence laborating centers based on IADPSG consensus screening tests for gestational diabetes. Am J