To the Editor: We have several concerns about the meta-analysis of randomized trials of antioxidant supplements by Dr Bjelakovic and colleagues.1 First, establishing causality requires considering temporal relationship, dose-response relationship, evidence of supplement use, effects that disappear after discontinuation of supplementation, lack of alternative explanations, and recurrence after restarting supplement use. It is difficult to establish a causal relationship between supplement use and risk of death when 2 of the criteria (response to rechallenge and response to discontinuation of use) cannot be applied to the outcome of death. Hence, greater reliance must be placed on plausible biological mechanisms and evidence that nutrients affect specific disorders. As the authors pointed out, it is likely that increased cancer and cardiovascular mortality are the main reasons for the increased all-cause mortality. The review would have been more convincing if it had also addressed cause-specific mortality.
Huang H, Teutsch S, Bass E. Antioxidant Supplements and Mortality. JAMA. 2007;298(4):400-403. doi:10.1001/jama.298.4.400-b