Level 1: Likely reliable evidence
AND MAY INCREASE MORTALITY
Level 2: Mid-level evidence
A systematic review evaluated 232,606 adults in 68 randomized trials testing antioxidant supplements (beta carotene, vitamin A, vitamin C, vitamin E, or selenium) for primary or secondary prevention of any condition (JAMA. 2007;297:842-857). Forty-seven trials (69%) were considered high quality based on adequate randomization, allocation concealment, blinding, and follow-up.
Doses of supplements varied by study, ranging from 1.2-50 mg for beta carotene, 2,500-200,000 units for vitamin A, 60-2,000 mg for vitamin C, 16.5-5,000 units for vitamin E, and 25-200 µg for selenium. There was no significant effect on mortality in an overall meta-analysis.
When the overall meta-analysis was stratified by specific antioxidant, the only effect on mortality that was statistically significant was a possible reduction in a meta-analysis of all 21 trials using selenium (relative risk 0.91). In meta-analyses limited to high-quality trials, there was a slightly increased risk for mortality overall (relative risk 1.05); increased mortality was found with beta carotene, vitamin A, and vitamin E, in analysis excluding selenium trials. A meta-analysis of low-quality trials suggested decreased mortality with borderline statistical significance.