The effect of vitamin D supplementation on the risk of type 2 diabetes mellitus (T2DM) remains controversial because most randomized controlled trials (RCTs) have been small or have reported low doses of vitamin D.

To conduct a meta-analysis of RCTs testing vitamin D supplementation in the prevention of T2DM.

Database search of PubMed/MEDLINE, EMBASE, and the Cochrane Library was performed by 2 reviewers from inception through September 15, 2019.

We included RCTs that reported the effect of vitamin D supplementation for at least 1 year on T2DM prevention.

Two independent reviewers extracted the data. The risk ratios (RRs) and 95% confidence intervals (CIs) were reported. Primary outcome of the meta-analysis was the incidence of T2DM.

Nine RCTs were included (43 559 participants). The mean age (standard deviation) was 63.5 (6.7) years. The RR for vitamin D compared with placebo was 0.96 (95% CI, 0.90-1.03); P = 0.30. In trials testing moderate to high doses of supplementation (≥1000 IU/day), all conducted among participants with prediabetes, the RR for vitamin D compared with placebo was 0.88 (95% CI, 0.79-0.99). In contrast, the trials testing lower doses, which were conducted in general population samples, showed no risk reduction (RR, 1.02; 95% CI, 0.94-1.10; P, interaction by dose = 0.04).

In patients with prediabetes, vitamin D supplementation at moderate to high doses (≥1000 IU/day), significantly reduced the incidence risk of T2DM, compared with placebo.