Comparisons of different vitamin D supplementation for prevention of osteoporotic fractures: a Bayesian network meta-analysis and meta-regression of randomised controlled trials.

Previous randomised controlled trials have shown the controversial effectiveness of oral vitamin D supplementation in preventing osteoporotic fractures. PubMed, EMBASE and Cochrane Library electronic databases were searched. Pairwise meta-analysis, Bayesian network meta-analysis and meta-regression were applied. A total of 33 studies containing 83,083 participants were included. Oral vitamin D supplementation showed no statistically significant on reducing the risk of total fractures (RR = 0.96, 95%CI = 0.87-1.05 p = 0.389). Vitamin D3 (700-800IU/d) plus calcium showed statistical significance in reducing the incidence of total, hip and non-vertebral fractures in the pairwise meta-analysis. Significant reductions were specifically identified in female in total and hip fractures. However, we did not observe any above significant results using Bayesian network meta-analyses. Strikingly, a meta-regression analysis identified an inverse association between the efficacy of fracture prevention and increased body mass index. Thus, we recommended that the vitamin D dose should be adjusted according to BMI based on further confirmation.