The pooled analysis, published in the Journal of the American Medical Association , could have significant negative effects on the sale of antioxidant dietary supplements, much like the effects of another meta-analysis, published in 2005, that reported a 10 per cent increased risk of mortality for people taking 400 International Units per day of vitamin E. The earlier meta-analysis, published in the Annals of Internal Medicine (142(1):37-46), has been highly criticised and discredited as flawed, but sales of vitamin E, particularly in the US, declined significantly in the wake of its publication.
The new pooled analysis has already drawn criticism from both inside and outside the dietary supplements industry concerning the methodology applied, and the conclusions drawn.
The most prevalent criticism of the study has come from the diverse nature of the studies pooled together, described by one industry expert as comparing "apples and oranges" .
Moreover, Meir Stampfer, a professor of nutrition and epidemiology at the Harvard School of Public Health who was not connected to the meta-analysis, told the Associated Press that the studies reviewed were too different to be able to pool them together.
"This study does not advance our understanding, and could easily lead to misinterpretation of the data," Stampfer told the AP.
A wealth of epidemiological evidence has linked dietary antioxidant intake to reduced risk of a wide-range of diseases, such as cancer and cardiovascular disease, but when such antioxidants have been extracted and purified, or synthesised and put into supplements, the antioxidants, according to randomized clinical trials (RCTs), do not always produce the same positive results.
"Our findings contradict the findings of observational studies, claiming that antioxidants improve health.
Considering that 10 per cent to 20 per cent of the adult population (80-160 million people) in North America and Europe may consume the assessed supplements, the public health consequences may be substantial," wrote Goran Bjelakovic, from Copenhagen University Hospital's Center for Clinical Intervention Research, and lead author of the new meta-analysis.
Bjelakovic and collaborators from the University of Nis in Serbia, and Ospedale V. Cervello in Palermo, followed the Cochrane Collaboration method for meta-analysis and started with 1201 references describing 815 trials.
Seven hundred and forty seven trials were excluded for several reasons, including no mortality in the study groups (405 trials), the studies were not randomised trials (69 trials), they did not fulfil inclusion criteria (245 trials), or the studies are ongoing (four).
The remaining 68 randomised trials used beta-carotene doses ranging from 1.2 to 50 milligrams, vitamin A from 1333 to 200 000 International Units (RDI 5000 IU, Upper Safe Limit 10,000 IU), vitamin C from 60 to 2000 mg (RDI 60 mg, UL 2000 mg), vitamin E from 10 to 5000 IU (RDI 30 IU, UL 900 IU), and selenium from 20 to 200 micrograms (RDI 65 micrograms, UL 450 micrograms).
"Most trials investigated the effects of supplements administered at higher doses than those commonly found in a balanced diet, and some of the trials used doses well above the recommended daily allowances and even above the tolerable upper intake level," said the reviewers.
Many of the studies included only subjects already classified as suffering from certain diseases.
Indeed, only two of the studies classified as "low bias" were carried out in the 'general population'.
When taken in their entirety, antioxidants were found to have no significant effect on mortality.
This result did not hold however when the reviewers looked at the nutrients singly and removed studies that they classed as having a 'high bias' in favour of the antioxidants.
In the 47 trials classified as low-bias (180,938 participants), overall antioxidant supplements were associated with a five per cent increased risk of mortality.
When the reviewers classified vitamin A intake as "given singly or in combination with other antioxidants supplements after exclusion of high-bias risk and selenium trials," use of the vitamin was associated with a 16 per cent increased risk of mortality.
Beta carotene and vitamin E was associated with seven and four per cent increased risk of mortality, respectively, whereas there was no increased mortality risk associated with vitamin C, and selenium use was associated with a nine per cent decreased risk of mortality.
"There are several possible explanations for the negative effect of antioxidant supplements on mortality.
Although oxidative stress has a hypothesized role in the pathogenesis of many chronic diseases, it may be the consequence of pathological conditions.
By eliminating free radicals from our organism, we interfere with some essential defensive mechanisms.
"Antioxidant supplements are synthetic and not subjected to the same rigorous toxicity studies as other pharmaceutical agents.
Better understanding of mechanisms and actions of antioxidants in relation to a potential disease is needed," concluded the researchers.
The study has been slammed by the US-based Council for Responsible Nutrition, who said the researchers " misuse[d] meta-analysis methods to create generalized conclusions that may inappropriately confuse and alarm consumers who can benefit from supplementing with antioxidants."
Moreover, Andrew Shao, Ph.D., CRN's vice president, scientific and regulatory affairs said that the combined studies were far too diverse and different in terms of dosage, duration, study population and nutrients tested that the results of the analysis were "compromised".
Additionally, Dr. Shao noted that most of the trials included in the meta-analysis tested for secondary prevention, looking at how a nutrient works in diseased populations, instead of primary prevention studies in healthy populations.
"Combining secondary prevention and primary prevention trials and then making conclusions for the entire population is an unsound scientific approach," said Shao.
"Additionally, many of the trials had limitations, including the expectation that a simple antioxidant vitamin could be expected to overturn serious illness, such as cancer or heart disease.
These trials likely statistically skewed the results."
Daniel Fabricant, vice president of scientific affairs for Natural Products Association (NPA), another US-based trade association, supported this view.
"Despite the authors' contention, this analysis is assessing mortality of at-risk and diseased populations - versus a healthy population - in prevention trials.
The risk of mortality must be attributed to the appropriate population studied, those with an existing health condition, which it isn't in this case.
Instead, those findings are generalized to a healthy general population, which is wrong on many levels," said Fabricant.
"But what's most troubling is that people who are safely and beneficially taking vitamins might stop, which may actually put their health at greater risk."
Fabricant also stated that if a true mortality risk had been apparent in any of these clinical studies using antioxidants, the study would have been halted, but none were.
"Nearly 160 million people in North America and Europe use the attributed supplements.
If antioxidants were a significant public health risk, as the authors speculate, it would have become clearly apparent by now," he said.
"This is obviously not the case: the negative effect they refer to has not been seen in the population."
Source: Journal of the American Medical Association February 28 2007, Volume 297, Pages 842-857 "Mortality in Randomized Trials of Antioxidant Supplements for Primary and Secondary Prevention:
Systematic Review and Meta-analysis" Authors: G. Bjelakovic; D. Nikolova; L. Lotte Gluud; R.G. Simonetti; C. Gluud