brand logo

Am Fam Physician. 2005;72(5):895-940

Oxidative stress is implicated in the etiology of cancer. Several observational studies have associated diets with high levels of antioxidants from fruit and vegetables with reduced risk of cancer. However, randomized trials of diet supplements have given contradictory results. Bjelakovic and colleagues reviewed the benefits and harms of antioxidant supplementation to prevent gastrointestinal cancers.

They used registries of controlled trials, searches of electronic databases, reference lists, and contacts with manufacturers and others to identify published and unpublished clinical trials. The 681 references identified were verified and assessed for quality using protocols of the Cochrane Collaboration.

Fourteen placebo-controlled trials of oral antioxidant supplements involving 170,525 persons met criteria for inclusion in the meta-analysis. The substances used were selenium (50 to 228 mcg), beta-carotene (15 to 50 mg), vitamin A (1.5 to 15 mg), vitamin C (120 to 2,000 mg), and vitamin E (30 to 600 mg) daily or on alternate days. All but one trial reported the incidence of cancers of the esophagus, stomach, colon, rectum, pancreas, and liver. Nine trials reported mortality data.

In studies lasting between one and 12 years, the overall incidence of gastrointestinal cancers was not significantly altered by antioxidant supplementation. Selenium was associated with reduced risk of hepatocellular cancer in high-risk patients in three low-quality trials. Selenium, but not other agents, also was associated with significantly reduced incidence of gastrointestinal cancers in four trials (three of low quality).

In high-quality trials, use of antioxidants was associated with significantly higher total mortality than placebo. This effect was not apparent in the low-quality trials. Significant increases in mortality were associated with combinations of beta-carotene with vitamin A or E. Beta-carotene alone increased mortality. Adverse effects were common with antioxidant use, particularly beta-carotene.

The authors conclude that antioxidants do not appear to have a significant impact on the prevention of gastrointestinal cancer and could be associated with increased overall mortality. Only selenium showed possible reduction in gastrointestinal cancer, but the poor quality of the trials makes the results difficult to interpret. The authors speculate that antioxidants in fruits and vegetables could be effective because of their complex interactions with other constituents of food. Antioxidants could be protective at certain levels and potentially toxic at higher levels or in susceptible individuals.

editor’s note: Once again, the research leaves the family physician in a difficult situation. Observational studies held out the promise of reduced cancer and improved mortality from diet supplements—results quickly exploited by the “natural remedies” industry. Now comes evidence suggesting not only the futility of taking these substances but the potential hazard. Commenting on these unexpected results, an accompanying editorial offers the commentary, “the prospect that vitamin pills may not only do no good but also kill their consumers is a scary speculation.”1 Much more research must be done before physicians talk with confidence to patients about supplements. Studies have to sort out all the combinations of supplements and subtypes of patients involved as well as address the many methodologic problems with these studies. In the meantime, many patients, especially health-conscious older patients, take large quantities of vitamins. For example, many take supplements specifically for macular degeneration in addition to other daily vitamins and supplements. The totals can be surprisingly large. Perhaps the wisest course is to suggest that supplements be used in moderation and, as mother always said, “eat your vegetables.”—a.d.w.

Continue Reading

More in AFP

Copyright © 2005 by the American Academy of Family Physicians.

This content is owned by the AAFP. A person viewing it online may make one printout of the material and may use that printout only for his or her personal, non-commercial reference. This material may not otherwise be downloaded, copied, printed, stored, transmitted or reproduced in any medium, whether now known or later invented, except as authorized in writing by the AAFP.  See for copyright questions and/or permission requests.