There may be some inherent flaws in the studies used for the meta-analysis (ex. vitamin E was used in isolation, the synthetic form was used, patients were used who had pathologies, behaviours, diets, and other factors characterized by high oxidative stress).