Clinical Question:
Do antioxidant supplements reduce all-cause mortality for adults?

Bottom Line:
Current evidence suggests that regular supplementation with the antioxidants beta carotene, vitamin A, and vitamin E increases mortality risk in adults. This report found no evidence of benefit or harm from supplementation with vitamin C and selenium. (LOE = 1a-)

Reference:
Bjelakovic G, Nikolova D, Gluud LL, Simonetti RG, Gluud C. Mortality in randomized trials of antioxidant supplements for primary and secondary prevention. Systematic review and meta-analysis. JAMA 2007;297:842-857.  [PMID:17327526]

Study Design:
Meta-analysis (randomized controlled trials)

Setting:
Various (meta-analysis)

Synopsis:
These investigators analyzed the effects of antioxidant supplements (beta carotene, vitamins A, C, and E, and selenium) on all-cause mortality. They performed a thorough search of multiple databases including the Cochrane Registry, Science Citation Index, MEDLINE, and relevant references for randomized controlled trials evaluating these supplements, either singly or in combination. Two authors independently reviewed individual trials for quality. Disagreements were resolved through consensus discussion with a third individual. Assignment of individual trial quality scores occurred using standard methods. Overall methodological quality of the individual trials was good. From an initial list of 1201 references reporting approximately 815 individual trials, 68 met study inclusion criteria (n = 232,606). The mean age of participants was 62 years (18 years - 103 years); follow-up occurred for a mean of 3.3 years (1 month - 14 years). When all the trials were combined, there was no significant effect of antioxidant supplements on mortality. An analysis of outcomes from only high-quality trials showed a significantly increased risk of mortality with beta carotene, vitamin A, and vitamin E, either singularly or combined. Selenium and vitamin C had no significant effect on overall mortality. Results of the individual trials did not meet standard criteria for heterogeneity (meaning that the findings from individual trials were generally consistent). No formal discussion of publication bias was reported. The majority of studies were funded by commercial sources (thus preferring not to publish negative trials), so it is likely that these summary results underestimate the true increased risks.

Copyright © 2013 John Wiley & Sons, Inc.