The effect of vitamin D supplementation on glucose metabolism in type 2 diabetes mellitus: A systematic review and meta-analysis of intervention studies.

Systematic review and meta-analysis of 29 intervention studies in adults with type 2 diabetes; 22 articles provided sufficient data for meta-analysis, with HbA1c analyzed in 19 studies and fasting blood glucose in 16 studies. Studies were conducted across multiple regions and included mixed baseline vitamin D status, including 15 studies enrolling participants with vitamin D deficiency.

Oral vitamin D supplementation, mainly cholecalciferol, was evaluated across intervention studies versus placebo or other controls. Regimens varied widely, from 400 IU daily to 450,000 IU once, given most commonly daily but sometimes weekly or as a one-time bolus, for 8 weeks to 5.5 years (median 12 weeks).

Vitamin D produced a modest improvement in HbA1c but did not improve fasting blood glucose overall. Pooled HbA1c fell by 0.32% versus placebo (WMD -0.32% [-0.53 to -0.10], P=0.022; I2=91.9%), while fasting blood glucose showed no overall difference (WMD -2.33 mg/dl [-6.62 to 1.95], P=0.542; I2=59.2%). Higher-dose and longer-duration subgroup results were inconsistent, and sensitivity analyses in vitamin D-deficient participants or low-risk-of-bias studies were not clearly significant. Overall, the authors conclude that evidence is not robust that vitamin D reliably improves glycemic control in adults with type 2 diabetes.

Substantial heterogeneity was present, especially for HbA1c (I2=91.9%), and regimens varied markedly by dose, duration, and formulation. Only 22 of 29 studies contributed to the meta-analysis, and 7 were excluded because outcome data were not reported in a usable form. Findings were inconsistent in subgroup and sensitivity analyses, limiting confidence in generalizability and in any effect among vitamin D-deficient participants.