Introduction & Objectives: All Diabetes Prevention Program (DPP) participants had prediabetes and received the same interventions (lifestyle, metformin, or placebo), but racial/ethnic differences in diabetes incidence emerged in follow-up, which we try to explain by differences in beta-cell function.
Methods: In 1671 self-identified white (W), 618 black (B), and 487 Hispanic (H) participants in DPP repository data, we analyzed insulin sensitivity index (ISI) and response to oral glucose (30-minute corrected insulin response, CIR) during 17 years of follow-up.
Results: Incidence was 7.5 cases/100 person-years in B, 6.0 in H, and 5.4 in W. B and H had higher rates than W, adjusted for age, sex, and intervention (B: HR=1.35, 95%CI=1.19-1.53; H: HR=1.07, 0.93-1.24) and when further adjusted for BMI, ISI, and CIR (B: HR=1.48, 1.30-1.68; H: HR=1.15, 0.99-1.33). Lower baseline BMI and higher ISI and CIR were protective but did not explain the race/ethnicity effects. Longitudinal relationships of ISI and CIR differed by race/ethnicity (Figure); paradoxically B had the highest CIR which would predict lower diabetes incidence.
Conclusion: Higher diabetes incidence in B was not explained by differences in BMI or beta-cell function. We hypothesize that B with prediabetes compensate for other, unknown, risk factors by increasing CIR which, in many, is insufficient to prevent diabetes.
E. Vazquez Arreola: None. R.L. Hanson: None. W.C. Knowler: None.
Intramural Research Program of the National Institute of Diabetes and Digestive and Kidney Diseases
