Fourteen RCTs were included (n= 21,268). Eleven RCTs were reported explicitly as double blinded. Where reported, the percentage of drop-outs ranged from four to 33.
Two RCTs evaluated vitamin D analogues (n=524).
Over all studies, vitamin D treatment was associated with a statistically significant reduction in the risk of falls compared to placebo, RR 0.94 (95% CI: 0.90, 0.99). Heterogeneity p =0.09.
Vitamin D analogues were associated with a statistically significant reduction in the risk of falls compared to native vitamin D: RR for analogues 0.79 (95% CI: 0.64, 0.96) versus RR for native vitamin D 0.95 (95% CI: 0.91, 1.00); ANOVA p =0.048. No significant heterogeneity was found for either analysis (analogues p=0.62 and native vitamin D p=0.17).
Differences between treatments were maintained when only double-blind RCTs were analysed (ANOVA p=0.049).
There was no evidence of publication bias.