Conclusions
In these women with average baseline VitD in sufficiency range, VitD compounds did not show clear beneficial associations with glucose homeostasis measures.
Methods
The analysis included 912 women. Inclusion criteria: <20 weeks gestation, body mass index ≥29 kg/m2 and information on exposure and outcome variables at baseline. Measurements: A 75 g OGTT at <20, 24−28 and 35−37 weeks gestation (except if previous diabetes diagnosis). Exposure variables: 25OHD2, 25OHD3 and C3-epimer. Outcome variables: fasting and post-challenge insulin sensitivity and secretion indices, corresponding disposition indices (DI), plasma glucose at fasting and 1 and 2 h, hyperglycemia in pregnancy (HiP). Statistics: Multivariate regression analyses with adjustment.
Results
Baseline VitD sufficiency was 66.3%. Overall, VitD compounds did not show strong associations with any glucose homeostasis measures. 25OHD3 showed direct significant associations with: FPG at <20 and 24−28 weeks (standardized β coefficient (β) 0.124, p = 0.030 and 0.111, p = 0.026 respectively), 2 h plasma glucose at 24−28 weeks (β 0.120, p = 0.018), and insulin sensitivity (1/HOMA-IR, β 0.127, p = 0.027) at 35−37 weeks; it showed an inverse association with fasting DI (QUCKI*HOMA-β) at <20 and 24−28 weeks (β −0.124, p = 0.045 and β −0.148, p = 0.004 respectively). 25OHD2 showed direct associations with post-challenge insulin sensitivity (Matsuda, β 0.149, p = 0.048) at 24−28 weeks) and post-challenge DI (Matsuda*Stumvoll phase 1) at 24−28 and 35−37 weeks (β 0.168, p = 0.030, β 0.239, p = 0.006). No significant association with C3-epimer was observed at any time period. Conclusions: In these women with average baseline VitD in sufficiency range, VitD compounds did not show clear beneficial associations with glucose homeostasis measures.