Screening for Diabetes and Pre-Diabetes With Proposed A1C-Based Diagnostic Criteria

DARIN E. OLSON PHD MARY K. RHEE KIRSTEN HERRICK MSC DAVID C. ZIEMER JENNIFER G. TWOMBLY PHD LAWRENCE S. PHILLIPS E p i d e m i o l o g y / H e a l t h OBJECTIVE - An International Expert Committee (IEC) and the American Diabetes Association (ADA) proposed diagnostic criteria for diabetes and pre-diabetes based on A1C levels. We hypothesized that screening for diabetes and pre-diabetes with A1C measurements would differ from using oral glucose tolerance tests (OGTT). RESEARCH DESIGN AND METHODS - We compared pre-diabetes, dysglycemia (diabetes or pre-diabetes), and diabetes identified by the proposed criteria (A1C 6.5% for diabetes and 6.0 - 6.4% [IEC] or 5.7- 6.4% [ADA] for high risk/pre-diabetes) with standard OGTT diagnoses in three datasets. Non-Hispanic white or black adults without known diabetes who had A1C and 75-g OGTT measurements were included from the prospective Screening for Impaired Glucose Tolerance study (n 1,581), and from the National Health and Nutrition Examination Survey (NHANES) III (n 2014), and NHANES 2005-2006 (n 1,111). RESULTS - OGTTs revealed pre-diabetes in 35.8% and diabetes in 5.2% of combined study subjects. A1C provided receiver operating characteristic (ROC) curve areas for diabetes of 0.79 - 0.83, but ROC curve areas were 0.70 for dysglycemia or pre-diabetes. The proposed criteria missed 70% of individuals with diabetes, 71- 84% with dysglycemia, and 82-94% with prediabetes. Compared with the IEC criteria, the ADA criteria for pre-diabetes resulted in fewer false-negative and more false-positive result. There were also racial differences, with falsepositive results being more common in black subjects and false-negative results being more common in white subjects. With use of NHANES 2005-2006 data, 5.9 million non-Hispanic U.S. adults with unrecognized diabetes and 43-52 million with pre-diabetes would be missed by screening with A1C. CONCLUSIONS - The proposed A1C diagnostic criteria are insensitive and racially discrepant for screening, missing most Americans with undiagnosed diabetes and pre-diabetes. - D can adults (1,2), with a lifetime risk iabetes affects 21 million Ameriranging from 20 to 50%, depending on sex and race (3). Identification of diabetes and its precursor, pre-diabetes, can permit management to prevent complications or delay progression from prediabetes to diabetes. Because most U.S. health care systems do not have systematic screening programs, many Americans have undiagnosed diabetes and prediabetes, and, therefore, these individuals are not initiating programs targeted at prevention (2). An International Expert Committee (IEC) recently proposed new diagnostic criteria based on measurement of A1C, with A1C 6.5% for diabetes and 6.0 6.4% for high risk of progression to diabetes (4). The American Diabetes Association (ADA) subsequently proposed A1C 6.5% for the diagnosis of diabetes and 5.7 6.4% for the highest risk to progress to diabetes (5). Because A1C testing is readily available in the U.S., is relatively well standardized, exhibits low intraindividual variation, and does not require fasting or restriction to certain times of the day (6), many clinicians might wish to use A1C measurements to screen for diabetes and pre-diabetes. However, the proposed diagnostic criteria were based largely on identification of diabetic retinopathy, and use of the proposed criteria as a screening test is not understood. The IEC A1C criteria have recently been compared with testing with fasting glucose or oral glucose tolerance tests (OGTTs) in various populations to diagnose diabetes (713) and high-risk/pre-diabetes (10,11,13), but the ADA A1C criteria have not been studied. We hypothesized that A1C diagnostic criteria would fail to identify many subjects with unrecognized diabetes or prediabetes. We evaluated the proposed criteria as screening tests in three populations, compared with the OGTT as a gold standard used for identification of diabetes and pre-diabetes around the world (14). RESEARCH DESIGN AND METHODS We examined three datasets in which non-Hispanic white and black adult subjects without known diabetes had both an OGTT and A1C measured (15). In the Screening for Impaired Glucose Tolerance (SIGT) study (16), health care system employees and community members in Atlanta were eligible if they were aged 18 or above, were nonHispanic white or black race, had no prior diagnosis of diabetes, were not pregnant or breastfeeding, were not taking glucocorticoids, and were well enough to work. A total of 1,581 subjects completed the protocol. The National Health and Nutrition Examination Surveys (NHANES) assessed adults and children across the U.S. In NHANES III (17), 2,057 non-Hispanic black or white subjects (to match the SIGT study population) aged 40 years Table 1Subjects identified by OGTT compared with A1C diagnoses with no known history of diabetes had an OGTT, meeting the inclusion criteria. In NHANES 20052006 (18), 1,154 nonHispanic adult subjects aged 18 years met the inclusion criteria. Age, BMI, blood pressure, lipids, and family history were categorized using conventional criteria. After subjects with missing data were excluded, there were 2,014 subjects in NHANES III and 1,111 subjects in NHANES 20052006, as described previously (15). Glucose and A1C measurements Plasma glucose and A1C measurements have been described previously (16 18). A1C measurements used NGSP certified systems (supplementary material, available in an online appendix at http://care. diabetesjournals.org/cgi/content/full/dc100433/DC1). Classification of glucose tolerance Glucose tolerance was classified by ADA criteria on the basis of glucose levels in a single 75-g OGTT. Subjects were grouped as normal glucose tolerance (fasting plasma glucose [FPG] 100 mg/dl, with 2-h plasma glucose 140 mg/dl), prediabetes (impaired fasting glucose with FPG 100 125 mg/dl and 2-h plasma glucose 200 mg/dl and/or impaired glucose tolerance with FPG 126 mg/dl and 2-h plasma glucose 140 199 mg/dl), and diabetes (FPG 126 mg/dl or 2-h plasma glucose 200 mg/dl). The additional IEC criteria identified subjects as normal (A1C 6%), high risk for diabetes (A1C 6.0 6.4%), and diabetes (A1C 6.5%), whereas new ADA criteria identified subjects as normal (A1C 5.7%), high risk (A1C 5.7 6.4%), and diabetes (A1C 6.5%). Additional evaluations used FPG 110 mg/dl for normal glucose tolerance. Dysglycemia includes prediabetes or diabetes (OGTT) or high risk or diabetes (A1C). Statistical analysis Means and frequencies were determined in aggregate and by subgroup analysis of the different glucose tolerance categories. The discriminative effectiveness of screening was evaluated by the area under receiver operating characteristic (ROC) curves using SAS 9.2 (SAS Institute, Cary, NC) for the NHANES data and SPSS 15.0 (SPSS, Chicago, IL) for SIGT study data. All NHANES analyses were conducted using SAS 9.2 and SUDAAN (...truncated)