Previous research linked vitamin D deficiency in pregnancy to adverse pregnancy outcomes. Update a 2017 systematic review and meta- analysis of randomized controlled trials (RCTs) on the effect of vitamin D supple mentation during pregnancy, identify sources of heterogeneity between trials, and describe evidence gaps precluding a clinical recommendation. The MEDLINE, PubMed, Europe PMC, Scopus, Cochrane Database of Systematic Reviews, Web of Science, and CINAHL databases were searched. Articles were included that reported on RCTs that included pregnant women given vitamin D supplements as compared with placebo, no intervention, or active control (600IU d–1). Risk ratios (RRs) and mean differences were pooled for 38 maternal, birth, and infant outcomes, using random effects models. Subgroup analyses examined effect heterogeneity. The Cochrane risk of bias tool was used. Included articles reported on a total of 66 trials (n ¼ 17 276 participants). The median vitamin D supplementation dose was 2000IU d–1 (range: 400-60 000); 37 trials used placebo. Antenatal vitamin D supplementation had no effect on the risk of preeclampsia (RR, 0.81 [95% CI, 0.43-1.53]; n ¼ 6 trials and 1483 participants), potentially protected against gestational diabetes mellitus (RR, 0.65 [95% CI, 0.49-0.86; n ¼ 12 trials and 1992 participants), and increased infant birth weight by 53 g (95% CI, 16-90; n ¼ 40 trials and 9954 participants). No effect of vitamin D on the risk of preterm birth, small-for- gestational age, or low birth weight infants was found. A total of 25 trials had at least 1 domain at high risk of bias. Additional studies among the general pregnant population are not needed, given the many existing trials. Instead, high-quality RCTs among populations with low vitamin D status or at greater risk of key outcomes are needed. Benefits of supplementation in pregnancy remain uncertain because current evidence has high heterogeneity, including variation in study context, baseline and achieved end-line 25-hydroxyvitamin D level, and studies with high risk of bias.
