Nine RCTs (n=118,765) were included in the review. Seven trials were double-blinded. Follow-up (where reported as means or medians) ranged from 3.6 years to 10.1 years (mean) and 1.4 years to 7.6 years (median); completeness of follow-up (where reported) ranged from 92.3% to 100%. Intention-to-treat analysis was used in eight trials.
There was no statistically significant effect of Vitamin E supplementation on the risk for total stroke (seven RCTs, no statistical heterogeneity). Statistically significant effects were noted for an increased risk of haemorrhagic stoke (RR 1.22, 95% CI 1.00 to 1.48; five RCTs, no statistical heterogeneity) and a decreased risk of ischaemic stroke (RR 0.90, 95% CI 0.82 to 0.99; five RCTs, no statistical heterogeneity). Results were similar using both fixed-effect and random-effects models and were unaffected in the meta-regression and sensitivity analyses. There was no evidence of publication bias.
In absolute terms, there was likely to be one haemorrhagic stroke for every 1,250 patients (22% increase in risk) and one ischaemic stroke prevented for every 476 patients (10% reduction in risk) who took Vitamin E supplementation.