Myth of Vitamin E Benefits Perpetuated in Peer-Reviewed Journals


IOANNINA, Greece, Dec. 5 -- Although randomized clinical trials have discredited any purported cardiovascular benefits of vitamin E, peer-reviewed medical journals keep citing old disproved epidemiologic findings, found researchers here.
Action Points --->
Explain to interested patients that this study found that articles in the literature continue to cite the protective benefits of vitamin E, beta-carotene, and estrogen for cardiovascular disease, cancer, and Alzheimer's disease, respectively, even though strong evidence from clinical trials has found otherwise.
That emerged from an analysis of citation patterns in journal articles published before, soon after, and several years after the vitamin E claims were contradicted, John P.A. Ioannidis, M.D., of the University of Ioannina School of Medicine, and colleagues, reported in the Dec. 5 issue of the Journal of the American Medical Association.
They said the same pattern -- citing observational findings from epidemiologic studies that have been solidly countered by randomized trials -- also continued for studies of beta-carotene and estrogen.
This trend wastes time, effort, and money and can "also perpetuate a literature of false impressions that can affect future scientific efforts by providing a misleading starting base," said Dr. Ioannidis and colleagues. They did not cite the offending journals by name.
Two articles touting the benefits of vitamin E were published in 1993, but the HOPE trial, which was published in 2000, found that vitamin E did not reduce the risk of cardiovascular events.
A favorable finding for vitamin E was less likely in papers published in 2005 than in 2001 or 1997 (P<0.001), yet half of the 46 vitamin E studies published in 2005 cited the two 1993 papers as evidence of the benefit of vitamin E, Dr. Ioannidis and colleagues said.
Moreover, when researchers analyzed 29 articles published in 2005 that cited the HOPE trial, 20.7% of those were still favorable to vitamin E's role in cardiovascular protection, 37.9% were equivocal, and 41.4% were unfavorable.
The researchers found an even higher rate of erroneous claims for articles on beta-carotene and estrogen.
Clinical trials published in 1994 and 1996 found no evidence that beta-carotene supplements prevent cancer, but 10 of 16 articles published in 2006 cited a 1981 review of epidemiology research, which suggested the antioxidant may help reduce the risk of developing the disease. The remaining six articles were evenly divided between equivocal and unfavorable.
For estrogen, 61.7% of 47 articles published in 2006 indicated the hormone may protect against Alzheimer's disease on the basis of a 1996 observational study, even though a clinical trial in 2004 found a trend for increased risk of dementia in postmenopausal women taking estrogen. Of the remaining articles, 29.8% were equivocal, and 8.5% were unfavorable.
Articles that espoused beta-carotene's cancer preventive effect did not consider contradicting evidence, but many of the studies touting the positive benefits of vitamin E or estrogen included counterarguments for why the clinical trial findings did not hold, including information biases or differences in study participants, interventions, co-interventions, and outcomes measured.
The possibility that some of these counterarguments are valid cannot be fully excluded and poses one limitation for the study, Dr. Ioannidis said. However, he added, the more likely scenario is that these researchers have a "wish bias" and are unwilling to let go of their hypotheses, even when the contradicting evidence is overwhelming.
One way of improving this situation is for researchers who specialize in very narrow fields, such as specific vitamins or hormones, to communicate more with those in clinical science and evidence-based medicine, Ioannidis said. "This communication may improve the rational design of translational research efforts to find effective treatments," he said.
Christian Gluud, M.D., of Copenhagen University Hospital in Denmark, said he thinks this pattern of not accepting the findings of randomized clinical trials when they refute epidemiological studies happens in other areas beyond vitamins and hormones.
"I don't think they've identified a rare behavior," Dr. Gluud said in an interview. "I think what they've identified is the norm." Dr. Gluud was not involved in the study.
In addition to spurring research that is based on false premises, the practice could also harm the public, Dr. Gluud said. If the general population picks up on the erroneous results, people could jump the gun and start taking vitamin supplements for diseases for which they have not been shown to be effective or safe, he said.
The researchers reported no financial conflicts of interest and no funding source was reported.
Primary source: Journal of the American Medical AssociationSource reference:Tatsioni A, et al "Persistence of contradicted claims in the literature" JAMA 2007; 298: 2517-2526.