Are antioxidant supplements really killing people?

There was seemingly bad news for the supplement-takers recently on the publication of a study in the Lancet which found that supplementing with ‘antioxidant’ nutrients (such beta-carotene, vitamins A, C and E and selenium) appeared not to reduce the risk of cancer in the digestive tract. However, worse was to come when the study authors also claimed that taking antioxidants might actually speed our demise. The Lancet emblazoned across its front page a quote from inside the journal stating: The prospect that vitamin pills may not only do no good but can also kill their consumers is a scary speculation given the vast quantities that are used in certain communities. Not surprisingly, I’ve has fair few requests to respond to what looks to be a very bitter pill to swallow indeed.

The Lancet study was what is known as a meta-analysis – where the results of several studies are pooled to get a more accurate assessment of a treatment than can be achieved by examining individual studies alone. The meta-analysis looked at a total of 14 trials, of which only seven were appropriate for appraising the effect of supplementation on risk of death. Close inspection of these seven studies reveals that six of them showed no significant increase in mortality. The overall results seem to have been somewhat skewed by one single study, the results of which were out of keeping with the others.

This observation was actually highlighted in an statistician-written editorial that accompanied the study. The authors of this editorial also pointed out that the statistical analysis used to churn out the apparent death risk was not the most appropriate one for this particular study. When a more relevant statistical test (something know as a ‘random effects meta-analysis’) was applied, the apparent increase in mortality disappeared.

The study that did show increased mortality was one in which supplementation with synthetic beta-carotene was found to increase death due to lung cancer in smokers and asbestos workers. This is not the only study that has found a link between the taking of chemically synthesised beta-carotene and increased risk of lung cancer in smokers. For this reason, my advice is for smokers to avoid supplementation with beta-carotene. There does, however, appear to be no risk from beta-carotene derived from the diet.

The apparent lack of benefit from taking antioxidants was another seeming surprise to come out of the Lancet study. However, the meta-analysis included studies that varied in terms of specific nutrients and dosages used. With such variability in study method, it is difficult for an overall picture to emerge. Also, the duration of these studies may simply not have been long enough to see benefits. It also possible, of course, that because Lancet study looked at digestive tract cancers only, it may therefore have missed other benefits with regard to cancer as a whole or conditions such as heart disease and stroke.

The deficiencies in this study’s scope and methodology led the authors of the accompanying editorial to conclude that it did not offer convincing proof of hazard�. Oddly, the Lancet’s fearsome front cover splash was actually lifted (out of context) from the very same editorial. One can only imagine what compelled its editors to employ such tabloid tactics. Whatever the motive, I reckon it’s a sad day for science when one of the World’s most respected medical journals chooses not to let the facts get in the way of a good story.

No comments yet.

Leave a Reply