Twelve RCTs (16,958 participants) were included.
Eight studies were double-blinded; the other four were open.
Statistical tests showed no significant heterogeneity between the studies and no evidence of publication bias.
All studies showed a net reduction in homocysteine levels with folic acid (from -1.5 to -26 micromoles/L).
There was no significant effect on any clinical outcome between treatment and control groups: the CVD was 0.95 (95% CI: 0.88, 1.03; 10 trials) for CVD, 1.04 (95% CI: 0.92, 1.17; 11 trials) for CHD, 0.86 (95% CI: 0.71, 1.04; 8 trials) for stroke and 0.96 (95% CI: 0.88, 1.04; 10 trials) for all-cause mortality.
In the sensitivity analysis, the removal of one large trial showed a significant protective effect of folic acid on stroke in pooled data from the remaining studies (RR=0.76, 95% CI: 0.63, 0.93). No other sensitivity analysis revealed any alterations in the overall result.
In subgroup analyses, stratified by existing disease (CVD or ESRF) and type of control used (placebo or usual care), there was no significant effect of treatment in any of the analyses.