Sensationalism and Flawed Analysis Bias Lancet Report on Antioxidant Safety

Sensationalism and Flawed Analysis Bias Lancet Report on Antioxidant Safety

A Healthnotes Newswire Opinion

Healthnotes Newswire (October 7, 2004)—Regrettably, the publishers of The Lancet have sensationalized a review published in this week’s issue (2004;364:1219–28) in such a way as to discourage antioxidant use, even though the review methodology contains several flaws that call the authors’ conclusions into question. Based on a review of research studies, the article concluded that taking antioxidant supplements does not prevent cancers of the gastrointestinal tract and may even cause a small but statistically significant increase in mortality, though the article itself states the mortality analysis is not yet complete.

Furthermore, by prominently presenting an excerpt of commentary the journal appears to reflect an anti-nutritional-supplement bias that is not in keeping with its history of fair and objective analysis. The authors of the commentary who first raise the idea that “the prospect that vitamin pills may not only do no good but also kill their consumers is a scary speculation” conclude by saying the article “does not offer convincing proof of hazard” with respect to mortality. The fact that The Lancet featured the scary speculation, rather than the editorialists’ actual conclusion that debunked it was highly biased and unbecoming of a major medical journal.

For their review, the authors pooled the results from 14 randomized clinical trials published over the past 19 years that included a total of more than 170,000 participants. The studies used various doses of antioxidants, either alone or in combination, including beta-carotene, vitamin A, vitamin C, vitamin E, and selenium.

Of the 14 reviewed clinical trials, 9 presented data on mortality. In the 7 trials that the reviewers judged to be of “high quality” the average death rate was 6% higher among antioxidant users than among those given a placebo. The other two trials, which were considered to be of “low quality,” showed a 6% decrease in mortality associated with antioxidant use. Although the authors focused mainly on the so-called high-quality studies, it is not at all obvious that the other studies should not have been given equal attention. Study quality was judged by such factors as the reliability of the randomization procedure and the certainty of whether participants truly didn’t know to which treatment group they were assigned. While good randomization and blinding methods are always preferable in studies that assess subjective outcomes such as fatigue and depression, it is difficult to believe that such subtle nuances in study design could influence an outcome as nonsubjective as dying. For that reason, the results of all the studies should have been pooled. Combining all of the studies would weaken the conclusion that antioxidant supplements increase mortality.

Perhaps the biggest flaw in the new study is its attempt to draw general conclusions about antioxidants by combining the results of studies that varied widely in the types of supplements used and in the types of people studied. For example, synthetic beta-carotene supplements have clearly been shown to be detrimental to smokers, increasing the risk of lung cancer by 18 to 28%, and increasing overall mortality by 17%. Preliminary research also suggests that synthetic beta-carotene may be harmful to people who consume excessive amounts of alcohol. The apparent negative interaction between synthetic beta-carotene, smoking, and alcohol consumption is not new information, and doctors knowledgeable about nutrition have been advising smokers and drinkers for quite some time to avoid synthetic beta-carotene supplements. Two studies of synthetic beta-carotene supplementation in smokers were among the 14 reviewed in The Lancet report, and the inclusion of these studies accounted in large part for the adverse findings with respect to mortality.

Synthetic beta-carotene contains only one of the several forms of beta-carotene that occur naturally in food, and the effects of the synthetic and natural forms of this nutrient differ to some extent. While beta-carotene supplements from natural sources have not been studied for cancer prevention, foods rich in beta-carotene have repeatedly been shown to be protective. Consequently, most nutritionists recommend that beta-carotene and related compounds (carotenoids) be obtained from foods, rather than from supplements.

The results observed with beta-carotene in smokers have little or no relevance to the effects of vitamin E, vitamin C, selenium, or other nutrients on either cancer risk or mortality. There is reasonably good scientific evidence that selenium supplements can help prevent cancer and possibly reduce the overall death rate. Studies on the effects of vitamin C and vitamin E are conflicting, but there is some evidence that they can help prevent heart disease and some other chronic illnesses, even if they are not magic bullets against cancer.

Lastly, in regard to whether the antioxidants show any benefit in preventing gastrointestinal cancers (e.g., cancers of the esophagus, stomach, colon, pancreas, and liver), when the results from all of the studies were pooled, antioxidant supplementation reduced risk by 10% according to one method of analyzing the data, and by 4% according to another method. These risk reductions were not statistically significant, so the authors concluded that antioxidants did not reduce cancer risk. That conclusion is open to question, however, because failure to find statistical significance is not the same as proving that a treatment didn’t work. The correct interpretation of the results is that antioxidant supplements reduced the risk of gastrointestinal cancers by 4 to 10%, but there is more than a 5% probability that the results were due to chance.

While the new report may raise concern in some circles, it does not provide convincing evidence that people should stop taking antioxidant supplements. With the exception of synthetic beta-carotene, antioxidant supplements still have great potential for preventing and treating a wide range of health conditions.

Alan R. Gaby, MD, an expert in nutritional therapies, testified to the White House Commission on CAM upon request in December 2001. Dr. Gaby served as a member of the Ad-Hoc Advisory Panel of the National Institutes of Health Office of Alternative Medicine. He is the author of Preventing and Reversing Osteoporosis (Prima, 1994), and co-author of The Natural Pharmacy, 2nd Edition (Healthnotes, Three Rivers Press, 1999), the A–Z Guide to Drug-Herb-Vitamin Interactions (Healthnotes, Three Rivers Press, 1999), Clinical Essentials Volume 1 and 2 (Healthnotes, 2000), and The Patient’s Book of Natural Healing (Prima, 1999). A former professor at Bastyr University of Natural Health Sciences, in Kenmore, WA, where he served as the Endowed Professor of Nutrition, Dr. Gaby is the Chief Medical Editor for Healthnotes, Inc.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.