193-LB: Refining Diabetes Risk for Individuals with Prediabetes

Rate of progression from prediabetes to diabetes varies greatly, and better risk stratification would help guide prevention efforts. We evaluated predictors of incident diabetes in the placebo arm (n=1212) of the vitamin D and type 2 diabetes (D2d) study. At baseline, participants met 2 or 3 ADA glycemic criteria for prediabetes (fasting glucose, FPG, 100-125mg/dL; 2-hour glucose during a 75-g OGTT, 2hPG, 140-199 mg/dL; HbA1c, 5.7-6.4%). Diabetes was diagnosed based on semiannual testing. After a median follow up of 2.5 years, 27% met criteria for diabetes. Among those who did and did not develop diabetes, mean BMI (±SD) was similar (32.1 kg/m2 ± 4.4), as well as other demographic factors including age, race, and ethnicity. More men than women developed diabetes (29% vs 23%, p=0.02). Incident diabetes rates differed by prediabetes subgroups (Figure); and those meeting all three prediabetes criteria at baseline (35.4%) were most likely to develop diabetes (log-rank p <0.001). There were no differences in the incident diabetes rate among those people meeting only 2 of the prediabetes criteria (Figure). Also, area under the receiver operator characteristic curves for incident diabetes at 36 months increased when all 3 criteria were used in the diagnosis (0.736) vs only FPG and HbA1c (0.686). We conclude that ‘prediabetes’ is heterogeneous, and the 2hPG may refine diabetes risk and guide intervention strategies in adults with high-risk prediabetes.