Joanna Cosgrove03.01.07
The Real Scoop on Antioxidants and Mortality
The supplement industry scoffs at JAMA’s recent meta-analysis, which pinned antioxidants as killers.
By
Joanna Cosgrove
Online Editor
The Journal of the American Medical Association (JAMA) has a long and respected reputation, but it recently faced harsh criticism for its decision to publish a meta-analysis on antioxidants, which the supplement industry roundly panned as biased and inaccurate.
The article, “Mortality in Randomized Trials of Antioxidant Supplements for Primary and Secondary Prevention” (JAMA, February 28, 2007;297(8):842-857), set out to “assess the effect of antioxidant supplements on mortality in randomized primary and secondary prevention trials.” The team behind the study systematically reviewed 68 randomized trials from 385 publications that had a total of 232,606 participants.
The meta-analysis concluded that the antioxidants covered in this study—beta-carotene, vitamin A and vitamin E—“may increase mortality,” with vitamin C and selenium’s effect on mortality requiring further study. If these conclusions are to be believed, they could pose a huge problem for the supplement industry’s most widely accepted segment, worth approximately $3 billion (Nutrition Business Journal, "Supplement Business Report 2006").
But the supplement industry’s trade associations wasted no time dismissing the report. Dr. Daniel Fabricant, vice president of scientific affairs, Natural Products Association, Washington, D.C., was “disappointed” by the study and asserted that if antioxidants were truly harmful, it’s doubtful that the more than 160 million people taking them in one form or another would still be doing so.
He, along with Washington, D.C.-based Council for Responsible Nutrition’s (CRN) Dr. Andrew Shao, Ph.D., vice president, scientific and regulatory affairs, were quick to point out that the study was dogged by a number obvious flaws.
“The beauty of the meta-analysis is that it provides advantages by increasing your sample size, therefore generating greater statistical power, but you’ve got to combine studies that are similar in inclusion criteria, such as design, study population, dosage, duration, etc.,” said Dr. Shao. “This study combined all sorts of different clinical trials that were very heterogeneous, different in so many ways. How one could combine them and call them a legitimate meta-analysis is unanswerable.”
Dr. Shao continued on, explaining that when diseased patients with liver failure, heart disease and cancer are included in a clinical trial they can stay on their medications in addition to being treated with antioxidants. But, he points out, “Even if the methodology used was legitimate—which I question—it’s questionable whether the results are relevant to the average consumer who uses antioxidants simply to maintain health and avoid disease in the first place.” He said what we saw in this case was results that got communicated as if they were relevant to a healthy population.
Dr. Shao also red flagged the study’s use of “all cause mortality” terminology. “This meta-analysis doesn’t shed any light on the benefits or risks of using supplements because they couldn’t tell what the deaths were due to,” he said. “‘All cause mortality’ isn’t something you die from, it can mean a heart attack or even a car accident. Why would they use a ton of data from studies using diseased patients to try and send a message to the larger community in general? FDA doesn’t even review studies using diseased populations for health claims because health claims are for healthy people.”
Dr. Fabricant concurred, adding that the qualifications of the researchers behind the meta-analysis should also be questioned. “These authors were primarily from the oncology field so this may not be their area of expertise,” he said. “We need to incorporate the appropriate people—perhaps those with a nutrition or pharmacognosy background—to sidestep some of the pitfalls. You wouldn’t study an arthritis drug without a rheumatologist, would you?”
It should be noted that of the study authors, Prof. Goran Bjelakovic, MD, DrMed Sc, was contacted through the Medical Faculty at the University of Nis in Serbia and offered the opportunity to rebut the allegations raised by the supplement industry trade groups with regard to his team’s qualifications as well as the alleged bias in the studies used in the meta-analysis, but he did not respond to Nutraceuticals World’s inquiry.
Part of a Larger Problem?
Whenever a negative supplement story surfaces, it’s common for many news media outlets to turn the story into a salacious sound bite to arouse a greater public response. But what’s interesting to note, said Dr. Shao, is that although the JAMA article attacked a supplement taken by some 160 million people, manufacturers of branded antioxidant products have received very few calls from consumers concerned about how the study results might impact their own antioxidant consumption. While this general lack of concern might come as a relief to weary manufacturers, it could also be a questionable development as it relates to consumer confidence.
“It’s an indication that consumers are turning a blind eye to this, they’re becoming numb, which is not necessarily a good thing,” said Dr. Shao. “Consumers are tuning out the media’s attempts to sensationalize things but it’s a double-edged sword because they could tune out news that could really help them.”
“One silver lining to this cloud is that consumers have become somewhat skeptical of the headlines and the information they see in the media,” said Dr. Fabricant. “I think they really understand that the science may not always work in sound bites so a number of them seek sources of other information. There’s a void there that we as an industry can fill.
“Another silver lining is that these types of stories tend to spring the industry and manufacturers into action, but we really need to do more in education and fund more on the clinical side,” he continued. “Education and research are the cornerstones of any industry where you want to build consumer confidence in and demonstrate the integrity of products. We need to fund good clinical studies by quality researchers and top notch institutions to gain that level of credibility.”
“Consumers care about their health, more than ever, but a given story doesn’t go as far as it used to 10 years ago because consumers are getting their information from so many different sources now. Likewise, the media outlets have to differentiate themselves as well. Some may sensationalize stories like this, whereas others may dumb it down,” added Dr. Shao. “Fortunately some media outlets knew better and held back (on reporting on the JAMA article), after doing a little due diligence and checking the merits of the study with some independent researchers to clearly see that there were flaws.”
He argued that the level of responsibility that lies with JAMA for accepting and publishing a “flawed” meta-analysis on face value could potentially be far-reaching. “The ramification is basically that a doctor will not read through the whole paper and pick it apart like (the supplement community) would do? Of course not! They’re going to read the abstract and the conclusion, and then advise their patients accordingly,” he said. “That could have a negative effect on someone who could benefit from antioxidants who stopped using them. Obviously it’s because JAMA needs to sell subscriptions and ads. If it weren’t in there you wouldn’t see CNN and other media outlets picking it up.”
Dr. Shao concluded by reiterating that there is a long-standing body of evidence that overwhelmingly suggests regular use of antioxidant supplements is associated with a lower risk of chronic disease. “We can’t simply dismiss that as the latest study du jour attempts to do,” he said. “Researchers have a responsibility to evaluate this stuff and put it in the proper context and remind folks that a healthy lifestyle with antioxidants supplements has been show to have benefits.”