sign in sign up
<<
Home , Glucose tolerance test

S8 Care Volume 38, Supplement 1, January 2015

2. Classification and Diagnosis of American Diabetes Association Diabetes Diabetes Care 2015;38(Suppl. 1):S8–S16 | DOI: 10.2337/dc15-S005

CLASSIFICATION Diabetes can be classified into the following general categories:

1. (due to b-cell destruction, usually leading to absolute deficiency) 2. (due to a progressive insulin secretory defect on the background of ) 3. mellitus (GDM) (diabetes diagnosed in the second or third trimester of that is not clearly overt diabetes) 4. Specific types of diabetes due to other causes, e.g., monogenic diabetes syndromes (such as and maturity-onset diabetes of the young [MODY]), dis- eases of the exocrine pancreas (such as cystic fibrosis), and drug- or chemical-induced diabetes (such as in the treatment of HIV/AIDS or after organ transplantation)

This section reviews most common forms of diabetes but is not comprehensive. POSITION STATEMENT For additional information, see the American Diabetes Association (ADA) position statement “Diagnosis and Classification of Diabetes Mellitus” (1). Assigning a type of diabetes to an individual often depends on the circumstances present at the time of diagnosis, with individuals not necessarily fitting clearly into a single category. For example, some patients cannot be clearly classified as having type 1 or type 2 diabetes. Clinical presentation and disease progression may vary considerably in both types of diabetes. The traditional paradigms of type 2 diabetes occurring only in adults and type 1 diabetes only in children are no longer accurate, as both diseases occur in both cohorts. Occasionally, patients with type 2 diabetes may present with (DKA). Children with type 1 diabetes typically present with the hallmark symptoms of polyuria/polydipsia and occasionally with DKA. The onset of type 1 diabetes may be variable in adults and may not present with the classic symptoms seen in children. However, difficulties in diagnosis may occur in children, adolescents, and adults, with the true diagnosis becoming more obvious over time.

DIAGNOSTIC TESTS FOR DIABETES Diabetes may be diagnosed based on A1C criteria or plasma criteria, either the fasting plasma glucose (FPG) or the 2-h plasma glucose (2-h PG) value after a 75-g oral glucose tolerance test (OGTT) (1,2) (Table 2.1). The same tests are used to both screen for and diagnose diabetes. Diabetes may be identified anywhere along the spectrum of clinical scenarios: in seemingly low- risk individuals who happen to have glucose testing, in symptomatic patients, and in higher-risk individuals whom the provider tests because of a suspicion of diabetes. The same tests will also detect individuals with .

A1C The A1C test should be performed using a method that is certified by the NGSP and standardized or traceable to the Diabetes Control and Complications Trial (DCCT) Suggested citation: American Diabetes Association. reference assay. Although point-of-care (POC) A1C assays may be NGSP certified, Classification and diagnosis of diabetes. Sec. 2. d proficiency testing is not mandated for performing the test, so use of POC assays for In Standards of Medical Care in Diabetes 2015. Diabetes Care 2015;38(Suppl. 1):S8–S16 diagnostic purposes may be problematic and is not recommended. © 2015 by the American Diabetes Association. The A1C has several advantages to the FPG and OGTT, including greater conve- Readers may use this article as long as the work nience (fasting not required), greater preanalytical stability, and less day-to-day is properly cited, the use is educational and not perturbations during stress and illness. These advantages must be balanced by for profit, and the work is not altered. care.diabetesjournals.org Position Statement S9

Hemoglobinopathies/Anemias , Table 2.1—Criteria for the diagnosis FPG ( 126 mg/dL [7.0 mmol/L]), that of diabetes Interpreting A1C levels in the presence of person should nevertheless be consid- A1C $6.5%. The test should be performed certain hemoglobinopathies and anemia ered to have diabetes. in a laboratory using a method that is may be problematic. For patients with an Since all the tests have preanalytic and NGSP certified and standardized to the abnormal hemoglobin but normal red cell analytic variability, it is possible that an ab- DCCT assay.* turnover, such as those with the sickle cell normal result (i.e., above the diagnostic OR trait, an A1C assay without interference threshold), when repeated, will produce FPG $126 mg/dL (7.0 mmol/L). Fasting is from abnormal hemoglobins should be a value below the diagnostic cut point. fi de ned as no caloric intake for at least used. An updated list of interferences is This scenario is least likely for A1C, more 8h.* available at www.ngsp.org/interf.asp. In likely for FPG, and most likely for the 2-h OR conditions associated with increased red PG, especially if the glucose samples are $ 2-h PG 200 mg/dL (11.1 mmol/L) during cell turnover, such as pregnancy (second collected at room temperature and not an OGTT. The test should be performed and third trimesters), recent loss as described by the WHO, using centrifuged promptly. Barring labora- a glucose load containing the or transfusion, erythropoietin therapy, tory error, such patients will likely equivalent of 75 g anhydrous glucose or hemolysis, only blood glucose criteria have test results near the margins of dissolved in water.* should be used to diagnose diabetes. the diagnostic threshold. The health OR care professional should follow the In a patient with classic symptoms of Fasting and 2-Hour Plasma Glucose patient closely and repeat the test in or hyperglycemic crisis, In addition to the A1C test, the FPG and 3–6months. a random plasma glucose $200 mg/dL 2-h PG may also be used to diagnose diabe- (11.1 mmol/L). tes (Table 2.1). The concordance between CATEGORIES OF INCREASED RISK *In the absence of unequivocal the FPG and 2-h PG tests is imperfect, as fi FOR DIABETES (PREDIABETES) hyperglycemia, results should be con rmed is the concordance between A1C and ei- by repeat testing. ther glucose-based test. National Health Recommendations and Nutrition Examination Survey c Testing to assess risk for future di- (NHANES) data indicate that an A1C cut abetes in asymptomatic people point of $6.5% identifies one-third greater cost, the limited availability of should be considered in adults of fewer cases of undiagnosed diabetes A1C testing in certain regions of the any age who are overweight or than a fasting glucose cut point of developing world, and the incomplete obese (BMI $25 kg/m2 or $23 $126 mg/dL (7.0 mmol/L) (9). Numer- correlation between A1C and average kg/m2 in Asian Americans) and ous studies have confirmed that, com- glucose in certain individuals. who have one or more additional pared with these A1C and FPG cut It is important to take age, race/ risk factors for diabetes. For all points, the 2-h PG value diagnoses ethnicity, and anemia/hemoglobinopathies patients, particularly those who more people with diabetes. Of note, into consideration when using the A1C to are overweight or obese, testing the lower sensitivity of A1C at the desig- diagnose diabetes. should begin at age 45 years. B nated cut point may be offset by the c If tests are normal, repeat testing Age test’s ease of use and facilitation of carried out at a minimum of 3- The epidemiological studies that formed more widespread testing. year intervals is reasonable. C the framework for recommending A1C Unless there is a clear clinical diagno- c To test for prediabetes, the A1C, to diagnose diabetes only included adult sis (e.g., a patient in a hyperglycemic FPG, and 2-h PG after 75-g OGTT populations. Therefore, it remains un- crisis or with classic symptoms of hyper- are appropriate. B clear if A1C and the same A1C cut point glycemia and a random plasma glucose c In patients with prediabetes, iden- should be used to diagnose diabetes in $200 mg/dL), it is recommended that – tify and, if appropriate, treat other children and adolescents (3 5). the same test be repeated immediately cardiovascular disease (CVD) risk Race/Ethnicity using a new blood sample for confirma- factors. B A1C levels may vary with patients’ race/ tion because there will be a greater like- c Testing to detect prediabetes ethnicity (6,7). For example, African lihood of concurrence. For example, if should be considered in children Americans may have higher A1C levels the A1C is 7.0% and a repeat result is and adolescents who are over- than non-Hispanic whites despite simi- 6.8%, the diagnosis of diabetes is con- weight or obese and who have lar fasting and postglucose load glucose firmed. If two different tests (such as two or more additional risk factors levels. A recent epidemiological study A1C and FPG) are both above the diagnos- for diabetes. E found that, when matched for FPG, tic threshold, this also confirms the diag- African Americans (with and without di- nosis. On the other hand, if a patient has abetes) had higher A1C levels than non- discordant results from two different Description Hispanic whites, but also had higher levels tests, then the test result that is above In 1997 and 2003, the Expert Commit- of and glycated albumin the diagnostic cut point should be re- teeonDiagnosisandClassification of and lower levels of 1,5-anhydroglucitol, peated. The diagnosis is made on the ba- Diabetes Mellitus (10,11) recognized a suggesting that their glycemic burden sis of the confirmed test. For example, if a group of individuals whose glucose lev- (particularly postprandially) may be patient meets the diabetes criterion of els did not meet the criteria for diabetes higher (8). the A1C (two results $6.5%), but not but were too high to be considered S10 Position Statement Diabetes Care Volume 38, Supplement 1, January 2015

Table 2.2—Criteria for testing for diabetes or prediabetes in asymptomatic adults TYPE 1 DIABETES 1. Testing should be considered in all adults who are overweight (BMI $25 kg/m2 or $23 kg/m2 in Asian Americans) and have additional risk factors: Recommendation c physical inactivity c Inform the relatives of patients with c first-degree relative with diabetes type 1 diabetes of the opportunity c high-risk race/ethnicity (e.g., African American, Latino, Native American, Asian to be tested for type 1 diabetes risk, American, Pacific Islander) but only in the setting of a clinical c . women who delivered a baby weighing 9 lb or were diagnosed with GDM research study. E c hypertension ($140/90 mmHg or on therapy for hypertension) c HDL cholesterol level ,35 mg/dL (0.90 mmol/L) and/or a triglyceride level .250 mg/dL Immune-Mediated Diabetes (2.82 mmol/L) “ c women with polycystic ovary syndrome This form, previously called insulin- c A1C $5.7%, IGT, or IFG on previous testing dependent diabetes” or “juvenile-onset c other clinical conditions associated with insulin resistance (e.g., severe obesity, diabetes,” accounts for 5–10% of diabetes acanthosis nigricans) and is due to cellular-mediated autoimmune c history of CVD destruction of the pancreatic b-cells. 2. For all patients, particularly those who are overweight or obese, testing should Autoimmune markers include islet cell begin at age 45 years. autoantibodies, autoantibodies to insu- 3. If results are normal, testing should be repeated at a minimum of 3-year intervals, with lin, autoantibodies to GAD (GAD65), consideration of more frequent testing depending on initial results (e.g., those with autoantibodies to the tyrosine phospha- prediabetes should be tested yearly) and risk status. tases IA-2 and IA-2b, and autoantibodies to zinc transporter 8 (ZnT8). Type 1 di- abetes is defined by the presence of one normal. “Prediabetes” is the term used 9to25%).AnA1Crangeof6.0–6.5% or more of these autoimmune markers. for individuals with impaired fasting had a 5-year risk of developing diabe- The disease has strong HLA associations, glucose (IFG) and/or impaired glucose tes between 25–50% and a relative risk with linkage to the DQA and DQB genes. tolerance (IGT) and indicates an in- 20 times higher compared with an A1C These HLA-DR/DQ alleles can be either creased risk for the future develop- of 5.0% (12). In a community-based predisposing or protective. ment of diabetes. IFG and IGT should study of African American and non- Therateofb-cell destruction is quite not be viewed as clinical entities in Hispanic white adults without diabetes, variable, being rapid in some individuals their own right but rather risk factors baseline A1C was a stronger predictor (mainly infants and children) and slow in for diabetes (Table 2.2)andCVD.IFG of subsequent diabetes and cardiovas- others (mainly adults). Children and and IGT are associated with obesity cular events than fasting glucose (13). adolescents may present with ketoaci- (especially abdominal or visceral obe- Other analyses suggest that an A1C of dosis as the first manifestation of the sity), dyslipidemia with high triglycer- 5.7% is associated with a diabetes risk disease. Others have modest fasting hy- ides and/or low HDL cholesterol, and similar to that of the high-risk partici- perglycemia that can rapidly change to hypertension. pants in the Diabetes Prevention Pro- severe hyperglycemia and/or ketoacido- gram (DPP) (14). sis with infection or other stress. Adults Diagnosis Hence, it is reasonable to consider an may retain sufficient b-cell function to In 1997 and 2003, the Expert Commit- A1C range of 5.7–6.4% as identifying in- prevent ketoacidosis for many years; tee on Diagnosis and Classification of dividuals with prediabetes. As with those such individuals eventually become de- Diabetes Mellitus (10,11) defined IFG with IFG and/or IGT, individuals with an pendent on insulin for survival and are as FPG levels 100–125 mg/dL (5.6–6.9 A1C of 5.7–6.4% should be informed of at risk for ketoacidosis. At this latter mmol/L)andIGTas2-hPGafter75-g their increased risk for diabetes and CVD stage of the disease, there is little or OGTT levels 140–199 mg/dL (7.8–11.0 and counseled about effective strategies no insulin secretion, as manifested by mmol/L). It should be noted that the to lower their risks (see Section 5. Preven- low or undetectable levels of plasma WorldHealthOrganization(WHO)and tion or Delay of Type 2 Diabetes). Similar C-peptide. Immune-mediated diabetes numerous diabetes organizations de- to glucose measurements, the continuum fine the IFG cutoff at 110 mg/dL (6.1 of risk is curvilinear, so as A1C rises, the mmol/L). diabetes risk rises disproportionately Table 2.3—Categories of increased risk As with the glucose measures, sev- (12). Aggressive interventions and vigilant for diabetes (prediabetes)* eral prospective studies that used A1C follow-up should be pursued for those FPG 100 mg/dL (5.6 mmol/L) to 125 mg/dL to predict the progression to diabetes considered at very high risk (e.g., those (6.9 mmol/L) (IFG) demonstrated a strong, continuous with A1C .6.0%). OR association between A1C and sub- Table 2.3 summarizes the categories 2-h PG in the 75-g OGTT 140 mg/dL (7.8 sequent diabetes. In a systematic re- of prediabetes. For recommendations mmol/L) to 199 mg/dL (11.0 mmol/L) (IGT) view of 44,203 individuals from 16 regarding risk factors and screening for OR cohort studies with a follow-up interval prediabetes, see p. S12 (“Testing for A1C 5.7–6.4% averaging 5.6 years (range 2.8–12 Type 2 Diabetes and Prediabetes in *For all three tests, risk is continuous, years), those with an A1C between Asymptomatic Adults” and “Testing for extending below the lower limit of the range 5.5–6.0% had a substantially increased Type 2 Diabetes and Prediabetes in Chil- and becoming disproportionately greater at higher ends of the range. risk of diabetes (5-year incidence from dren and Adolescents”). care.diabetesjournals.org Position Statement S11

commonly occurs in childhood and ado- children who developed more than two c Testing to detect type 2 diabetes lescence, but it can occur at any age, autoantibodies, nearly 70% developed should be considered in children even in the 8th and 9th decades of life. type 1 diabetes within 10 years and 84% andadolescentswhoareover- b fi Autoimmune destruction of -cells has within 15 years (16,18). These ndings are weight or obese and who have fi multiple genetic predispositions and is highly signi cant because, while the Ger- two or more additional risk factors also related to environmental factors man group was recruited from offspring for diabetes. E that are still poorly defined. Although pa- of parents with type 1 diabetes, the Finn- tients are not typically obese when they ish and American groups were recruited present with type 1 diabetes, obesity from the general population. Remark- Description should not preclude the diagnosis. These ably, the findings in all three groups This form, previously referred to as “non- patients are also prone to other autoim- were the same, suggesting that the insulin-dependent diabetes” or “adult- mune disorders such as Graves’ disease, same sequence of events led to clinical onset diabetes,” accounts for ;90–95% Hashimoto’s thyroiditis, Addison’sdis- disease in both “sporadic” and genetic of all diabetes. Type 2 diabetes encom- ease, vitiligo, celiac disease, autoimmune cases of type 1 diabetes. passes individuals who have insulin resis- hepatitis, myasthenia gravis, and perni- While there is currently a lack of tance and usually relative (rather than cious anemia. accepted screening programs, one absolute) insulin deficiency. At least ini- should consider referring relatives of tially, and often throughout their lifetime, Idiopathic Diabetes those with type 1 diabetes for antibody these individuals may not need insulin Some forms of type 1 diabetes have no testing for risk assessment in the treatment to survive. known etiologies. These patients have per- setting of a clinical research study There are various causes of type 2 di- manent insulinopenia and are prone to (http://www2.diabetestrialnet.org). abetes. Although the specificetiologies ketoacidosis, but have no evidence of au- Widespread clinical testing of asymptom- are not known, autoimmune destruc- toimmunity. Although only a minority of atic low-risk individuals is not currently tion of b-cells does not occur, and pa- patients with type 1 diabetes fall into this recommended due to lack of approved tients do not have any of the other category, of those who do, most are of therapeutic interventions. Higher-risk in- known causes of diabetes. Most, but African or Asian ancestry. Individuals dividuals may be tested, but only in the not all, patients with type 2 diabetes with this form of diabetes suffer from ep- context of a clinical research setting. In- are obese. Obesity itself causes some isodic ketoacidosis and exhibit varying dividuals who test positive will be coun- degree of insulin resistance. Patients fi degrees of insulin de ciency between epi- seled about the risk of developing who are not obese by traditional weight sodes. This form of diabetes is strongly diabetes, diabetes symptoms, and DKA criteria may have an increased percent- inherited, lacks immunological evidence prevention. Numerous clinical studies age of body fat distributed predomi- b for -cell autoimmunity, and is not HLA are being conducted to test various meth- nantly in the abdominal region. associated. An absolute requirement for ods of preventing type 1 diabetes in Ketoacidosis seldom occurs sponta- insulin replacement therapy in affected those with evidence of autoimmunity neously in type 2 diabetes; when seen, patients may come and go. (www.clinicaltrials.gov). it usually arises in association with Testing for Type 1 Diabetes the stress of another illness such as in- The incidence and prevalence of type 1 TYPE 2 DIABETES fection. Type 2 diabetes frequently goes diabetes is increasing (15). Type 1 dia- undiagnosed for many years because hy- betic patients often present with acute Recommendations perglycemia develops gradually and at symptoms of diabetes and markedly ele- c Testing to detect type 2 diabetes earlier stages is often not severe enough vated blood glucose levels, and some are in asymptomatic people should for the patient to notice the classic di- diagnosed with life-threatening keto- be considered in adults of any abetes symptoms. Nevertheless, such acidosis. Several studies suggest that mea- age who are overweight or patients are at an increased risk of obese (BMI $25 kg/m2 or $23 suring islet autoantibodies in relatives of 2 developing macrovascular and micro- those with type 1 diabetes may identify kg/m in Asian Americans) and vascular complications. individuals who are at risk for developing who have one or more addi- Whereaspatientswithtype2diabetes type 1 diabetes. Such testing, coupled tional risk factors for diabetes. may have insulin levels that appear nor- with education about diabetes symptoms For all patients, particularly mal or elevated, the higher blood glucose and close follow-up in an observational those who are overweight or levels in these patients would be expected clinical study, may enable earlier identifi- obese, testing should begin at to result in even higher insulin values had cation of type 1 diabetes onset. There is age 45 years. B their b-cell function been normal. Thus, evidence to suggest that early diagnosis c If tests are normal, repeat testing insulin secretion is defective in these pa- may limit acute complications (16) and carried out at a minimum of 3-year tients and insufficient to compensate for extend long-term endogenous insulin intervals is reasonable. C insulin resistance. Insulin resistance may production (17). c To test for diabetes, the A1C, FPG, improve with weight reduction and/or A recent study reported the risk of pro- and 2-h PG after 75-g OGTT are pharmacological treatment of hyper- gression to type 1 diabetes from the time appropriate. B glycemia but is seldom restored to normal. of seroconversion to autoantibody posi- c In patients with diabetes, identify The risk of developing type 2 diabetes tivity in three pediatric cohorts from Fin- and, if appropriate, treat other increases with age, obesity, and lack of land, Germany, and the U.S. Of the 585 CVD risk factors. B physical activity. It occurs more frequently S12 Position Statement Diabetes Care Volume 38, Supplement 1, January 2015

in women with prior GDM, in those with beginning at age 30 or 45 years and Diagnostic Tests hypertension or dyslipidemia, and in cer- independent of risk factors, may be The A1C, FPG, and 2-h PG after 75-g OGTT tain racial/ethnic subgroups (African cost-effective (,$11,000 per quality- are appropriate for testing. It should be American, American Indian, Hispanic/ adjusted life-year gained) (20). noted that the tests do not necessarily Latino, and Asian American). It is often Additional considerations regarding detect diabetes in the same individuals. associated with a strong genetic predis- testing for type 2 diabetes and predia- The efficacy of interventions for primary position, more so than type 1 diabetes. betes in asymptomatic patients include prevention of type 2 diabetes (26–32) has However, the genetics of type 2 diabetes the following: primarily been demonstrated among in- dividuals with IGT, not for individuals with is poorly understood. Age isolated IFG or for those with prediabetes Testing recommendations for diabetes in defined by A1C criteria. Testing for Type 2 Diabetes and asymptomatic adults are listed in Table Prediabetes in Asymptomatic Adults 2.2. Age is a major risk factor for diabetes. Testing Interval Prediabetes and diabetes meet criteria for Testing should begin at age 45 years for all The appropriate interval between tests is conditions in which early detection is ap- patients, particularly those who are over- not known (33). The rationale for the propriate. Both conditions are common weight or obese. 3-year interval is that with this interval, and impose significant clinical and public the number of false-positive tests that re- BMI and Ethnicity health burdens. There is often a long pre- quire confirmatory testing will be reduced Testingshouldbeconsideredinadults symptomatic phase before the diagnosis and individuals with false-negative tests of any age with BMI $25 kg/m2 and one of type 2 diabetes. Simple tests to detect will be retested before substantial time or more additional risk factors for dia- preclinical disease are readily available. elapses and complications develop (33). The duration of glycemic burden is a strong betes. However, recent data (21) and predictor of adverse outcomes. There are evidence from the ADA position state- Community Screening “ effective interventions that prevent pro- ment BMI Cut Points to Identify At-Risk Ideally, testing should be carried out gression from prediabetes to diabetes Asian Americans for Type 2 Diabetes within a health care setting because of ” (see Section 5. Prevention or Delay of Screening (22)suggestthattheBMI the need for follow-up and treatment. Type 2 Diabetes) and reduce the risk cut point should be lower for the Asian Community testing outside a health care of diabetes complications (see Section American population. For diabetes setting is not recommended because peo- 8. Cardiovascular Disease and Risk screening purposes, the BMI cut points ple with positive tests may not seek, or – 2 Management and Section 9. Microvas- fall consistently between 23 24 kg/m have access to, appropriate follow-up test- cular Complications and Foot Care). (sensitivity of 80%) for nearly all Asian ing and care. Community testing may also Approximately one-quarter of people American subgroups (with levels slightly be poorly targeted; i.e., it may fail to reach with diabetes in the U.S. are undiag- lower for Japanese Americans). This the groups most at risk and inappropri- 2 nosed. Although screening of asymptom- makes a rounded cut point of 23 kg/m ately test those at very low risk or even atic individuals to identify those with practical. In determining a single BMI those who have already been diagnosed. prediabetes or diabetes might seem rea- cut point, it is important to balance sen- fi sonable, rigorous clinical trials to prove sitivity and speci city so as to provide a Testing for Type 2 Diabetes and the effectiveness of such screening have valuable screening tool without numer- Prediabetes in Children and not been conducted and are unlikely to ous false positives. An argument can be Adolescents occur. A large European randomized con- made to push the BMI cut point to lower 2 In the last decade, the incidence and prev- trolled trial compared the impact of than 23 kg/m in favor of increased alence of type 2 diabetes in adolescents has screening for diabetes and intensive sensitivity; however, this would lead increased dramatically, especially in ethnic fi multifactorial intervention with that of to an unacceptably low speci city populations (15). Recent studies question screening and routine care (19). General (13.1%). Data from the WHO also sug- the validity of A1C in the pediatric popula- $ 2 practice patients between the ages of gest that a BMI 23 kg/m should be tion, especially among certain ethnicities, fi 40–69 years were screened for diabetes used to de ne increased risk in Asian and suggest OGTT or FPG as more suitable and randomized by practice to intensive Americans (23). diagnostic tests (34). However, many of treatment of multiple risk factors or rou- Evidence also suggests that other these studies do not recognize that diabe- fi tine diabetes care. After 5.3 years of populations may bene tfromlower tes diagnostic criteria are based on long- follow-up, CVD risk factors were modestly BMI cut points. For example, in a large term health outcomes, and validations are but significantly improved with intensive multiethnic cohort study, for an equiv- not currently available in the pediatric pop- treatment compared with routine care, alent incidence rate of diabetes, a BMI 2 ulation (35). The ADA acknowledges the but the incidence of first CVD events or of 30 kg/m in non-Hispanic whites was limited data supporting A1C for diagnosing 2 mortality was not significantly different equivalent to a BMI of 26 kg/m in Afri- diabetes in children and adolescents. How- between the groups (19). The excellent can Americans (24). ever, aside from rare instances, such as cys- care provided to patients in the routine Medications tic fibrosis and hemoglobinopathies, the care group and the lack of an unscreened Certain medications, such as glucocorti- ADA continues to recommend A1C in this control arm limit our ability to prove coids, thiazide , and atypical anti- cohort (36,37). The modified recommenda- that screening and early intensive treat- psychotics (25), are known to increase the tions of the ADA consensus report “Type 2 ment impact outcomes. Mathematical risk of diabetes and should be considered Diabetes in Children and Adolescents” are modeling studies suggest that screening, when ascertaining a diagnosis. summarized in Table 2.4. care.diabetesjournals.org Position Statement S13

Table 2.4—Testing for type 2 diabetes or The ongoing epidemic of obesity and points for GDM as the average glucose prediabetes in asymptomatic children* diabetes has led to more type 2 diabetes values (fasting, 1-h, and 2-h PG) in the Criteria in women of childbearing age, resulting in HAPO study at which odds for adverse c Overweight (BMI .85th percentile an increase in the number of pregnant outcomes reached 1.75 times the esti- for age and sex, weight for height women with undiagnosed type 2 diabetes mated odds of these outcomes at the . . 85th percentile, or weight 120% (38). Because of the number of pregnant mean glucose levels of the study popu- of ideal for height) women with undiagnosed type 2 diabe- lation. This one-step strategy was an- Plus any two of the following risk factors: fi c tes, it is reasonable to test women with ticipated to signi cantly increase the Family history of type 2 diabetes in – ; – first- or second-degree relative risk factors for type 2 diabetes (Table 2.2) incidence of GDM (from 5 6% to 15 c Race/ethnicity (Native American, at their initial prenatal visit, using stan- 20%), primarily because only one abnormal African American, Latino, Asian dard diagnostic criteria (Table 2.1). value, not two, became sufficient to make American, Pacific Islander) Women with diabetes in the first trimes- the diagnosis. The ADA recognized that the c Signs of insulin resistance or ter would be classified as having type 2 anticipated increase in the incidence of conditions associated with insulin diabetes. GDM is diabetes diagnosed in GDM would have significant impact on resistance (acanthosis nigricans, the second or third trimester of preg- the costs, medical infrastructure capacity, hypertension, dyslipidemia, “ polycystic ovary syndrome, or small- nancy that is not clearly overt diabetes. and potential for increased medicaliza- for-gestational-age birth weight) tion” of previously catego- c Maternal or GDM Diagnosis rized as normal, but recommended these during the child’s gestation GDM carries risks for the mother and diagnostic criteria changes in the context Age of initiation: age 10 years or at onset neonate. Not all adverse outcomes are of worrisome worldwide increases in obe- of puberty, if puberty occurs at a of equal clinical importance. The Hyper- sity and diabetes rates with the intent of younger age glycemia and Adverse Pregnancy Out- optimizing gestational outcomes for Frequency: every 3 years come (HAPO) study (39), a large-scale women and their offspring. *Persons aged #18 years. (;25,000 pregnant women) multina- The expected benefits to these preg- tional cohort study, demonstrated that nancies and offspring are inferred from risk of adverse maternal, fetal, and neo- intervention trials that focused on GESTATIONAL DIABETES MELLITUS natal outcomes continuously increased women with lower levels of hyperglyce- as a function of maternal glycemia at mia than identified using older GDM di- Recommendations 24–28 weeks, even within ranges previ- agnostic criteria and that found modest c Test for undiagnosed type 2 diabe- ously considered normal for pregnancy. benefits including reduced rates of large- tes at the first prenatal visit in For most complications, there was no for-gestational-age births and preeclamp- those with risk factors, using stan- threshold for risk. These results have sia (42,43). It is important to note that dard diagnostic criteria. B led to careful reconsideration of the di- 80–90% of women being treated for c Test for GDM at 24–28 weeks of ges- agnostic criteria for GDM. GDM diagno- mild GDM in two randomized controlled tation in pregnant women not pre- sis (Table 2.5) can be accomplished with trials (whose glucose values overlapped viously known to have diabetes. A either of two strategies: with the thresholds recommended by c Screen women with GDM for per- the IADPSG) could be managed with life- sistent diabetes at 6–12 weeks 1. “One-step” 75-g OGTT or style therapy alone. Data are lacking on postpartum, using the OGTT and 2. “Two-step” approach with a 50-g how the treatment of lower levels of hy- clinically appropriate nonpreg- (nonfasting) screen followed by a perglycemia affects a mother’s risk for nancy diagnostic criteria. E 100-g OGTT for those who screen the development of type 2 diabetes in c WomenwithahistoryofGDM positive the future and her offspring’sriskfor should have lifelong screening for obesity, diabetes, and other metabolic the development of diabetes or Different diagnostic criteria will identify dysfunction. Additional well-designed prediabetes at least every 3 years. B different degrees of maternal hypergly- clinical studies are needed to determine c WomenwithahistoryofGDM cemia and maternal/fetal risk, leading the optimal intensity of monitoring and found to have prediabetes should some experts to debate, and disagree treatment of women with GDM diag- receive lifestyle interventions or on, optimal strategies for the diagnosis nosed by the one-step strategy. metformin to prevent diabetes. A of GDM. Two-Step Strategy Definition One-Step Strategy In 2013, the National Institutes of Health For many years, GDM was defined as In the 2011 Standards of Care (40), the (NIH) convened a consensus develop- any degree of glucose intolerance that ADA for the first time recommended ment conference on diagnosing GDM. was first recognized during pregnancy that all pregnant women not known to The 15-member panel had representatives (10), regardless of whether the condi- have prior diabetes undergo a 75-g from /gynecology, maternal- tion may have predated the pregnancy OGTT at 24–28 weeks of gestation, fetal medicine, pediatrics, diabetes re- or persisted after the pregnancy. This basedonarecommendationoftheIn- search, biostatistics, and other related definition facilitated a uniform strategy ternational Association of the Diabetes fields to consider diagnostic criteria (44). for detection and classification of GDM, and Pregnancy Study Groups (IADPSG) The panel recommended the two-step but it was limited by imprecision. (41). The IADPSG defined diagnostic cut approach of screening with a 1-h 50-g S14 Position Statement Diabetes Care Volume 38, Supplement 1, January 2015

Table 2.5—Screening for and diagnosis of GDM of diabetes are frequently characterized One-step strategy by onset of hyperglycemia at an early age Perform a 75-g OGTT, with plasma glucose measurement when patient is fasting and at 1 and (generally before age 25 years). 2h,at24–28 weeks of gestation in women not previously diagnosed with overt diabetes. Neonatal Diabetes The OGTT should be performed in the morning after an overnight fast of at least 8 h. Diabetes diagnosed in the first 6 months of The diagnosis of GDM is made when any of the following plasma glucose values are met or exceeded: life has been shown not to be typical au- c Fasting: 92 mg/dL (5.1 mmol/L) toimmune type 1 diabetes. This so-called c 1 h: 180 mg/dL (10.0 mmol/L) neonatal diabetes can either be transient c 2 h: 153 mg/dL (8.5 mmol/L) or permanent. The most common genetic Two-step strategy defect causing transient disease is a defect Step 1: Perform a 50-g GLT (nonfasting), with plasma glucose measurement at 1 h, at 24–28 on ZAC/HYAMI imprinting, whereas weeks of gestation in women not previously diagnosed with overt diabetes. permanent neonatal diabetes is most If the plasma glucose level measured 1 h after the load is $140 mg/dL* (7.8 mmol/L), proceed commonly a defect in the gene encoding to a 100-g OGTT. the Kir6.2 subunit of the b-cell KATP chan- Step 2: The 100-g OGTT should be performed when the patient is fasting. nel. Diagnosing the latter has implications, The diagnosis of GDM is made if at least two of the following four plasma glucose levels since such children can be well managed (measured fasting and 1 h, 2 h, 3 h after the OGTT) are met or exceeded: with sulfonylureas. Carpenter/Coustan (56) or NDDG (57)

c Fasting 95 mg/dL (5.3 mmol/L) 105 mg/dL (5.8 mmol/L) Maturity-Onset Diabetes of the Young c 1 h 180 mg/dL (10.0 mmol/L) 190 mg/dL (10.6 mmol/L) MODY is characterized by impaired insulin c 2 h 155 mg/dL (8.6 mmol/L) 165 mg/dL (9.2 mmol/L) secretion with minimal or no defects in in- c 3 h 140 mg/dL (7.8 mmol/L) 145 mg/dL (8.0 mmol/L) sulin action. It is inherited in an autosomal dominant pattern. Abnormalities at six ge- NDDG, National Diabetes Data Group. *The ACOG recommends a lower threshold of 135 mg/dL (7.5 mmol/L) in high-risk ethnic netic loci on different chromosomes have populations with higher prevalence of GDM; some experts also recommend 130 mg/dL been identified to date. The most common (7.2 mmol/L). form is associated with mutations on chro- mosome 12 in a hepatic transcription factor referred to as hepatocyte nuclear factor glucose load test (GLT) followed by a 3-h implement must therefore be made based (HNF)-1a. A second form is associated 100-g OGTT for those who screen on the relative values placed on factors with mutations in the glucokinase gene positive, a strategy commonly used in thathaveyettobemeasured(e.g.,cost- on chromosome 7p and results in a defec- fi the U.S. bene t estimation, willingness to change tive glucokinase molecule. Glucokinase ’ Key factors reported in the NIH panel s practice based on correlation studies converts glucose to glucose-6-phosphate, decision-making process were the lack of rather than clinical intervention trial the metabolism of which, in turn, stimu- clinical trial interventions demonstrating results, relative role of cost consid- lates insulin secretion by the b-cell. The fi the bene ts of the one-step strategy erations, and available infrastructure lo- less common forms of MODY result from and the potential negative consequences cally, nationally, and internationally). mutations in other transcription factors, in- of identifying a large new group of As the IADPSG criteria have been cluding HNF-4a,HNF-1b, insulin promoter women with GDM, including medicaliza- adopted internationally, further evi- factor (IPF)-1, and NeuroD1. tion of pregnancy with increased inter- dence has emerged to support im- ventions and costs. Moreover, screening proved pregnancy outcomes with cost Diagnosis with a 50-g GLT does not require fast- savings (47) and may be the preferred Readily available commercial genetic ing and is therefore easier to accomplish approach. In addition, pregnancies testing now enables a true genetic diag- for many women. Treatment of higher complicated by GDM per IADPSG crite- nosis. It is important to correctly diag- threshold maternal hyperglycemia, as ria, but not recognized as such, have nose one of the monogenic forms of identified by the two-step approach, re- comparable outcomes to pregnancies di- diabetes because these children may duces rates of neonatal macrosomia, agnosed as GDM by the more stringent be incorrectly diagnosed with type 1 or large-for-gestational-age births, and two-step criteria (48). There remains type 2 diabetes, leading to suboptimal shoulder dystocia, without increasing strong consensus that establishing a uni- treatment regimens and delays in diag- small-for-gestational-age births (45). form approach to diagnosing GDM will nosing other family members (49). The American College of Obstetricians benefit patients, caregivers, and policy- The diagnosis of monogenic diabetes and Gynecologists (ACOG) updated its makers. Longer-term outcome studies are should be considered in children with guidelines in 2013 and supported the currently underway. the following findings: two-step approach (46). MONOGENIC DIABETES ○ Diabetes diagnosed within the first 6 Future Considerations SYNDROMES months of life The conflicting recommendations from Monogenic defects that cause b-cell dys- ○ Strong family history of diabetes but expert groups underscore the fact that function, such as neonatal diabetes and without typical features of type 2 di- there are data to support each strategy. MODY, represent a small fraction of pa- abetes (nonobese, low-risk ethnic The decision of which strategy to tients with diabetes (,5%). These forms group) care.diabetesjournals.org Position Statement S15

○ Mild fasting hyperglycemia (100–150 with and without diabetes and have elim- Hemoglobin A1c versus oral glucose tolerance mg/dL [5.5–8.5 mmol/L]), especially if inatedthesexdifferenceinmortality(52). test in postpartum diabetes screening. Diabetes – young and nonobese Recent trials comparing insulin with oral Care 2012;35:1648 1653 10. The Expert Committee on the Diagnosis and ○ fi Diabetes with negative autoantibod- repaglinide showed no signi cant differ- Classification of Diabetes Mellitus. Report of the ies and without signs of obesity or in- ence between the groups. However, an- Expert Committee on the Diagnosis and Classi- sulin resistance other study compared three different fication of Diabetes Mellitus. Diabetes Care groups: premeal insulin aspart, repagli- 1997;20:1183–1197 – nide, or oral placebo in cystic fibrosis pa- 11. Genuth S, Alberti KG, Bennett P, et al.; Ex- CYSTIC FIBROSIS RELATED pert Committee on the Diagnosis and Classifica- DIABETES tients with abnormal glucose tolerance. tion of Diabetes Mellitus. Follow-up report on Patients all had weight loss; however, in the diagnosis of diabetes mellitus. Diabetes Recommendations the insulin-treated group, this pattern was Care 2003;26:3160–3167 c Annual screening for cystic fibrosis– reversed, and they gained 0.39 (6 0.21) 12. Zhang X, Gregg EW, Williamson DF, et al. A1C related diabetes (CFRD) with OGTT BMI units (P 5 0.02). Patients in the level and future risk of diabetes: a systematic re- – should begin by age 10 years in all repaglinide-treated group had initial weight view. Diabetes Care 2010;33:1665 1673 patients with cystic fibrosis who 13.SelvinE,SteffesMW,ZhuH,etal.Glycated gain, but this was not sustained by 6 hemoglobin, diabetes, and cardiovascular risk in do not have CFRD. B A1C as a months. The placebo group continued to nondiabetic adults. N Engl J Med 2010;362:800–811 screening test for CFRD is not rec- lose weight (53). Insulin remains the most 14. Ackermann RT, Cheng YJ, Williamson DF, ommended. B widely used therapy for CFRD (54). Gregg EW. Identifying adults at high risk for di- c Patients with CFRD should be Recommendations for the clinical man- abetes and cardiovascular disease using hemo- treated with insulin to attain in- globin A1c National Health and Nutrition agement of CFRD can be found in the ADA Examination Survey 2005-2006. Am J Prev dividualized glycemic goals. A position statement “Clinical Care Guide- Med 2011;40:11–17 c In patients with cystic fibrosis and lines for Cystic Fibrosis–Related Diabetes: 15. Dabelea D, Mayer-Davis EJ, Saydah S, et al.; IGT without confirmed diabetes, A Position Statement of the American SEARCH for Diabetes in Youth Study. Prevalence prandial insulin therapy should be Diabetes Association and a Clinical Prac- of type 1 and type 2 diabetes among children and adolescents from 2001 to 2009. JAMA 2014; considered to maintain weight. B tice Guideline of the Cystic Fibrosis 311:1778–1786 c Annual monitoring for complica- Foundation, Endorsed by the Pediatric 16. Ziegler AG, Rewers M, Simell O, et al. Sero- tions of diabetes is recommended, Endocrine Society” (55). conversion to multiple islet autoantibodies and beginning 5 years after the diagno- risk of progression to diabetes in children. JAMA sis of CFRD. E 2013;309:2473–2479 References 17. Sorensen JS, Johannesen J, Pociot F, et al. 1. American Diabetes Association. Diagnosis Residual b-cell function 3-6 years after onset of fi and classi cation of diabetes mellitus. Diabetes type 1 diabetes reduces risk of severe hypogly- CFRD is the most common comorbidity – Care 2014;37(Suppl. 1):S81 S90 cemiainchildrenandadolescents.Diabetes fi in people with cystic brosis, occurring in 2. The International Expert Committee. Inter- Care 2013;36:3454–3459 about 20% of adolescents and 40–50% of national Expert Committee report on the role 18. Sosenko JM, Skyler JS, Palmer JP, et al.; Type of the A1C assay in the diagnosis of diabetes. 1 Diabetes TrialNet Study Group; Diabetes Pre- adults. Diabetes in this population is as- – Diabetes Care 2009;32:1327 1334 vention Trial-Type 1 Study Group. The prediction sociated with worse nutritional status, 3. Nowicka P, Santoro N, Liu H, et al. Utility of of type 1 diabetes by multiple autoantibody levels more severe inflammatory lung disease, hemoglobin A for diagnosing prediabetes and 1c and their incorporation into an autoantibody risk diabetes in obese children and adolescents. Di- and greater mortality from respiratory score in relatives of type 1 diabetic patients. Di- abetes Care 2011;34:1306–1311 failure. Insulin insufficiency related to abetes Care 2013;36:2615–2620 4. Garc´ıa de Guadiana Romualdo L, Gonzalez´ fi 19. Griffin SJ, Borch-Johnsen K, Davies MJ, et al. partial brotic destruction of the islet Morales M, Albaladejo Oton´ MD, et al. [The Effect of early intensive multifactorial therapy mass is the primary defect in CFRD. Ge- value of hemoglobin A1c for diagnosis of diabe- on 5-year cardiovascular outcomes in individu- netically determined function of the re- tes mellitus and other changes in als with type 2 diabetes detected by screening maining b-cells and insulin resistance metabolism in women with recent gestational diabetes mellitus] Endocrinol Nutr 2012;59: (ADDITION-Europe): a cluster-randomised trial. associated with infection and inflamma- – 362–366 [in Spanish] Lancet 2011;378:156 167 tion may also play a role. While screening 5. Cowie CC, Rust KF, Byrd-Holt DD, et al. Prev- 20. Kahn R, Alperin P, Eddy D, et al. Age at ini- for diabetes before the age of 10 years alence of diabetes and high risk for diabetes tiation and frequency of screening to detect using A1C criteria in the U.S. population in type 2 diabetes: a cost-effectiveness analysis. can identify risk for progression to CFRD – 1988-2006. Diabetes Care 2010;33:562–568 Lancet 2010;375:1365 1374 in those with abnormal glucose tolerance, 21. Araneta MR, Gandinetti A, Chang HK. Opti- fi 6. Ziemer DC, Kolm P, Weintraub WS, et al. there appears to be no bene twithre- Glucose-independent, black-white differences in mum BMI cut points to screen Asian Americans spect to weight, height, BMI, or lung func- hemoglobin A1c levels: a cross-sectional analysis for type 2 diabetes: the UCSD Filipino Health tion compared with those with normal of 2 studies. Ann Intern Med 2010;152:770–777 Study and the North Kohala Study. Diabetes glucose tolerance ,10 years of age. The 7. Kumar PR, Bhansali A, Ravikiran M, et al. Util- 2014;63(Suppl. 1):A20 22. Hsu WC, Araneta MR, Kanaya AM, Chiang use of continuous glucose monitoring ity of in diagnosing type 2 diabetes mellitus: a community-based study. J JL, Fujimoto W. BMI cut points to identify at-risk may be more sensitive than OGTT to de- Clin Endocrinol Metab 2010;95:2832–2835 Asian Americans for type 2 diabetes screening. tect risk for progression to CFRD, but this 8. Selvin E, Steffes MW, Ballantyne CM, Diabetes Care 2015;38:150–158 likely needs more evidence. Hoogeveen RC, Coresh J, Brancati FL. Racial dif- 23. WHO Expert Consultation. Appropriate Encouraging data suggest that im- ferences in glycemic markers: a cross-sectional body-mass index for Asian populations and its analysis of community-based data. Ann Intern implications for policy and intervention strate- proved screening (50,51) and aggressive Med 2011;154:303–309 gies. Lancet 2004;363:157–163 insulin therapy have narrowed the gap in 9. Picon´ MJ, Murri M, Munoz~ A, Fernandez-´ 24. Chiu M, Austin PC, Manuel DG, Shah BR, Tu mortality between cystic fibrosis patients Garc´ıa JC, Gomez-Huelgas R, Tinahones FJ. JV. Deriving ethnic-specific BMI cutoff points for S16 Position Statement Diabetes Care Volume 38, Supplement 1, January 2015

assessing diabetes risk. Diabetes Care 2011;34: 35. Kapadia C, Zeitler P; Drugs and Therapeutics 47. Duran A, Saenz´ S, Torrejon´ MJ, et al. In- 1741–1748 Committee of the Pediatric Endocrine Society. troduction of IADPSG criteria for the screen- 25. Erickson SC, Le L, Zakharyan A, et al. New- Hemoglobin A1c measurement for the diagnosis ing and diagnosis of gestational diabetes onset treatment-dependent diabetes mellitus of type 2 diabetes in children. Int J Pediatr En- mellitus results in improved pregnancy out- and hyperlipidemia associated with atypical an- docrinol 2012;2012:31 comes at a lower cost in a large cohort of tipsychotic use in older adults without schizo- 36. Kester LM, Hey H, Hannon TS. Using hemoglo- pregnant women: the St. Carlos Gestational phrenia or bipolar disorder. J Am Geriatr Soc bin A1c for prediabetes and diabetes diagnosis in Diabetes Study. Diabetes Care 2014;37: 2012;60:474–479 adolescents: can adult recommendations be upheld 2442–2450 26. Knowler WC, Barrett-Connor E, Fowler SE, for pediatric use? J Adolesc Health 2012;50:321–323 48. Ethridge JK Jr, Catalano PM, Waters TP. et al.; Diabetes Prevention Program Research 37. Wu E-L, Kazzi NG, Lee JM. Cost-effectiveness Perinatal outcomes associated with the diagno- Group. Reduction in the incidence of type 2 di- of screening strategies for identifying pediatric sis of gestational diabetes made by the Inter- abetes with lifestyle intervention or metformin. diabetes mellitus and dysglycemia. JAMA Pe- national Association of the Diabetes and N Engl J Med 2002;346:393–403 diatr 2013;167:32–39 Pregnancy Study Groups criteria. Obstet Gyne- 27. Tuomilehto J, Lindstrom¨ J, Eriksson JG, 38. Lawrence JM, Contreras R, Chen W, Sacks col 2014;124:571–578 et al.; Finnish Diabetes Prevention Study Group. DA. Trends in the prevalence of preexisting 49. Hattersley A, Bruining J, Shield J, Njolstad Prevention of type 2 diabetes mellitus by diabetes and gestational diabetes mellitus P, Donaghue KC. The diagnosis and manage- changes in lifestyle among subjects with im- among a racially/ethnically diverse population ment of monogenic diabetes in children and paired glucose tolerance. N Engl J Med 2001; of pregnant women, 1999-2005. Diabetes Care adolescents. Pediatr Diabetes 2009;10(Suppl. 344:1343–1350 2008;31:899–904 12):33–42 28. Pan X-R, Li G-W, Hu Y-H, et al. Effects of diet 39. Metzger BE, Lowe LP, Dyer AR, et al.; HAPO 50. Kern AS, Prestridge AL. Improving screening and exercise in preventing NIDDM in people with Study Cooperative Research Group. Hyperglyce- for cystic fibrosis-related diabetes at a pediatric impaired glucose tolerance: the Da Qing IGT and mia and adverse pregnancy outcomes. N Engl J cystic fibrosis program. Pediatrics 2013;132: Diabetes Study. Diabetes Care 1997;20:537–544 Med 2008;358:1991–2002 e512–e518 29. Buchanan TA, Xiang AH, Peters RK, et al. 40. American Diabetes Association. Standards 51. Waugh N, Royle P, Craigie I, et al. Screen- Preservation of pancreatic b-cell function and of medical care in diabetes–2011. Diabetes ing for cystic fibrosis-related diabetes: a sys- prevention of type 2 diabetes by pharmacolog- Care 2011;34(Suppl. 1):S11–S61 tematic review. Health Technol Assess 2012; ical treatment of insulin resistance in high-risk 41. Metzger BE, Gabbe SG, Persson B, et al.; 16:iii–iv, 1–179 Hispanic women. Diabetes 2002;51:2796–2803 International Association of Diabetes and Preg- 52. Moran A, Dunitz J, Nathan B, Saeed A, 30. Chiasson J-L, Josse RG, Gomis R, Hanefeld nancy Study Groups Consensus Panel. Interna- Holme B, Thomas W. Cystic fibrosis-related di- M, Karasik A, Laakso M; STOP-NIDDM Trial Re- tional Association of abetes: current trends in prevalence, incidence, search Group. Acarbose for prevention of type 2 Study Groups recommendations on the diagno- and mortality. Diabetes Care 2009;32:1626– diabetes mellitus: the STOP-NIDDM randomised sis and classification of hyperglycemia in preg- 1631 trial. Lancet 2002;359:2072–2077 nancy. Diabetes Care 2010;33:676–682 53.MoranA,PekowP,GroverP,etal.;Cystic 31. Gerstein HC, Yusuf S, Bosch J, et al.; DREAM 42. Landon MB, Spong CY, Thom E, et al.; Eu- Fibrosis Related Diabetes Therapy Study Group. (Diabetes REduction Assessment with ramipril nice Kennedy Shriver National Institute of Insulin therapy to improve BMI in cystic fibrosis- and rosiglitazone Medication) Trial Investigators. Child Health and Human Development Ma- related diabetes without fasting hyperglycemia: Effect of rosiglitazone on the frequency of diabe- ternal-Fetal Medicine Units Network. A multi- results of the cystic fibrosis related diabetes tes in patients with impaired glucose tolerance or center, randomized trial of treatment for mild therapy trial. Diabetes Care 2009;32:1783– : a randomised con- gestational diabetes. N Engl J Med 2009;361: 1788 trolled trial. Lancet 2006;368:1096–1105 1339–1348 54. Onady GM, Stolfi A. Insulin and oral 32. Ramachandran A, Snehalatha C, Mary S, 43. Crowther CA, Hiller JE, Moss JR, McPhee agents for managing cystic fibrosis-related dia- Mukesh B, Bhaskar AD, Vijay V; Indian Diabetes AJ, Jeffries WS, Robinson JS; Australian Carbo- betes. Cochrane Database Syst Rev 2013;7: Prevention Programme (IDPP). The Indian Dia- hydrate Intolerance Study in Pregnant CD004730 betes Prevention Programme shows that life- Women (ACHOIS) Trial Group. Effect of treat- 55. Moran A, Brunzell C, Cohen RC, et al.; CFRD style modification and metformin prevent type ment of gestational diabetes mellitus on preg- Guidelines Committee. Clinical care guidelines 2 diabetes in Asian Indian subjects with im- nancy outcomes. N Engl J Med 2005;352: for cystic fibrosis-related diabetes: a position paired glucose tolerance (IDPP-1). Diabetologia 2477–2486 statement of the American Diabetes Associa- 2006;49:289–297 44. Vandorsten JP, Dodson WC, Espeland MA, tion and a clinical practice guideline of the Cystic 33.JohnsonSL,TabaeiBP,HermanWH.The et al. NIH consensus development conference: Fibrosis Foundation, endorsed by the Pediatric efficacy and cost of alternative strategies for diagnosing gestational diabetes mellitus. NIH Endocrine Society. Diabetes Care 2010;33: systematic screening for type 2 diabetes in the Consens State Sci Statements 2013;29:1–31 2697–2708 U.S. population 45-74 years of age. Diabetes 45. Horvath K, Koch K, Jeitler K, et al. Effects of 56. Carpenter MW, Coustan DR. Criteria for Care 2005;28:307–311 treatment in women with gestational diabetes screening tests for gestational diabetes. Am J 34. Buse JB, Kaufman FR, Linder B, Hirst K, El mellitus: systematic review and meta-analysis. Obstet Gynecol 1982;144:768–773 Ghormli L, Willi S; HEALTHY Study Group. Diabe- BMJ 2010;340:c1395 57. National Diabetes Data Group. Classifica- tes screening with hemoglobin A(1c) versus fast- 46. Committee on Practice Bulletins–Obstetrics. tion and diagnosis of diabetes mellitus and ing plasma glucose in a multiethnic middle-school Practice Bulletin No. 137: gestational diabetes other categories of glucose intolerance. Diabe- cohort. Diabetes Care 2013;36:429–435 mellitus. Obstet Gynecol 2013;122:406–416 tes 1979;28:1039–1057