Abstract
OBJECTIVE: We examined associations between body weight and plasma 25-hydroxyvitamin D concentration (25OHD) in prediabetes and sought to estimate the impact of adiposity on these associations. METHODS: The study was conducted in the placebo (n = 1082) and intensive lifestyle (n = 1079) groups of the Diabetes Prevention Program (DPP), a multicenter trial to prevent type 2 diabetes in adults with prediabetes. Weight and 25OHD were measured at baseline, month 6, years 1 and 2. In a subset (n = 584), visceral (VAT) and subcutaneous (SAT) adiposity were assessed by computed tomography at baseline and year 1. RESULTS: In cross-sectional analyses, baseline body weight, total fat, VAT, and SAT were inversely associated with plasma 25OHD concentration after multivariable adjustment. VAT accounted for 40 % [95 % CI 11, 69] of the association of body weight with plasma 25OHD concentration. There was no significant contribution by total fat or SAT. Two-year changes in plasma 25OHD concentration varied inversely with changes in body weight (p < 0.0001). One-year changes in total fat, VAT, or SAT were not significant mediators of the association between change in plasma 25OHD concentration and body weight. CONCLUSION: Our study found an inverse association between body weight and plasma 25OHD concentration at baseline and over a 2-year period in adults with prediabetes. These findings in the DPP, a weight loss intervention study, raise the possibility that weight loss increases plasma 25OHD concentration. Whether adiposity mediates this association remains inconclusive.