Why the Cochrane Review is Flawed The 2007 Cochrane ‘meta-analysis’ spawned headlines such as ‘ Vitamin Pills May Cause Early Death’. This study by Bjekaolvic and colleagues lumped together the data of 67 disparate studies the majority of which involved older people (average age 61) with life-threatening diseases, many of which were not set up to measure mortality and actually found no change in mortality overall. But, by excluding any studies in which people were given a multivitamin, or selenium, alongside other antioxidants, and studies they deemed called ‘high bias’ they were able to come up with a negative effect for antioxidants overall.
Many top scientists don’t agree with this kind of approach: “Blenderising these diverse trials to come up with one simple proclamation is just silly. There was no difference (in mortality) based on vitamin intake,” said Bernadine Healey, the first woman to head the National Institutes of Health, the world’s largest funder of health research. “It’s flawed results based on flawed data,” said top antioxidant expert Dr Balz Fry from Oregon State University. This meta-analysis was also criticised for how it selected the trials it included. To give you an example, one excellent trial (called the NIT trial) was eliminated because there wasn’t enough information on why those who dropped out had done so. Admittedly, if you give a drug with short-term side-effects and half your participants drop out this might be relevant, but there are no known or reported short-term side effects of antioxidants.
Anyway, leaving all these flaws aside, what is remarkable about this damning review is that if you exclude studies in which beta-carotene is given to smokers, and high dose vitamin E is given to cardiovascular patients, not only does any evidence of risk for antioxidants vanish, but you are left with more studies showing benefit. Here’s how it works: 26 trials left after exclusions (including those with a zero weighting, meaning those that reported either no, or no difference in deaths) of which.. 2 (ATBC and CARET) only show negative effect in smokers on beta-carotene and 7 give high dose vitamin E to cardiovascular patients presumably on statins. If we know that combining smoking and beta-carotene, or statins plus vitamin E, is a confounding variable, then these 9 studies should be excluded. That leaves 17 studies. Six of these show a positive effect and only two show a substantial negative effect, in terms of mortality.
The first was headed by Dr Pelayo Correa from Louisiana State University whose group tested whether antioxidant supplements would help gastric cancer. When I told him how his trial had been used in this review he was ”amazed”, he said, because his research, “far from being negative, had shown clear benefit from taking vitamins”. “It did not intend, and did not have the power, to study mortality and has no value to examine mortality of cancer.” The second, nick named REACT (Roche European American Cataract Trial), gave elderly people vitamins A, C and E to see if it would slow down the progression of cataracts. Again, a positive result. The conclusion reads “Daily use of the micronutrients for three years produced a small deceleration in progression of age-related cataracts.”
During this three year trial three people died in the placebo group and six people died in the supplement group. This was not significant. Of these 17 studies 10 reported a benefit from taking antioxidants for the change that their trial had set out to measure, for example reducing cataract or gastric cancer risk, while 7 showed no significant difference. One reported a negative effect, but only for smokers and drinkers. But even though many of these trials weren’t designed to examine mortality what Bjekaolvic and colleagues did was to add up the total number of deaths. (The vast majority of these trials were in old people with life-threatening diseases hence there were a number of deaths during these trials, most of which were conducted over years.) So they had a total of 46,410 people in the 17 studies who took supplement and 45,124 who got a placebo. Among those taking supplements 2678 died (5.8%) compared to 2521 (5.6%) in the ‘control’ group, What this shows is that there was no difference in risk of mortality from taking anti-oxidants and many positive effects.
That’s a far cry from the damning headlines in the papers. Of course, it would be nice if taking antioxidants on their own (excluding multivitamins and selenium) had shown a reduction in mortality but that’s quite a tall order in a group of mainly elderly people, the majority of which have life-threatening diseases. It may be a case of too little too late. Professor Bruce Ames, a biochemist at the Children’s Hospital Oakland Research Institute in California and one of the originators of the antioxidant theory of ageing, is highly skeptical of relying on this kind of pooled analyses of randomized controlled trials to test for the effects of vitamins.”They are extremely difficult to do properly,” he says. “They must be conducted for decades to detect effects on diseases like cancer and heart disease that take a long time to develop and it is very difficult to make sure that those in the placebo group don’t start taking a multivitamin. This is why they are not likely to yield definitive answers.”
In addition he goes on to ask is we really want to wait, perhaps decades, for results of these long-term randomised controlled trials, which anyway will almost certainly will not provide definitive evidence. Instead, he suggests, it makes more sense to run trials that can be done more quickly. So his research team have, instead, focused on the effects of antioxidants on the aging that is going on all the time in our cells. The single biggest source of oxidants are the ‘exhaust fumes’ we create every second as a by-product of turning food into energy. This is done inside nearly every cell in the energy factories, known as mitochondria. Ames has studied how to prevent the ageing and decay of mitochondria in animals by giving them different combinations of antioxidants. This work has shown that increasing the intake of antioxidants can produce better liver function, brain function, reduction in blood pressure, inflammation and protection against macular degeneration. This kind of evidence is just as important, if not more so, than large-scale trials that often don’t take into account important variables such as the synergistic effects of antioxidants or the way in which combined medication may distort or negate their benefit.
So, at the end of the day, you have to pick your way through the mire of confusing reports and make up your own mind as to the best way to protext your health. Although in recent years we have had a spate of negative findings about antioxidants, they haven’t actually shown us anything new about how they work. Can you imagine what would happen if you pooled all drug trials? How meaningless would be the conclusions? My personal advice is to start by eating a diet high in antioxidants and to keep taking an optimum nutrition-style multivitamin. Then if you are getting on in years and want to stay healthy, or if you are suffering from a condition, for example heart disease, where the evidence shows that reducing oxidation may help, then also supplement an all-round antioxidant supplement containing CoQ10, lipoic acid, glutathione or N-acetyl cysteine, resveratrol, anthocyandins, vitamins A, C, E and beta-carotene, plus zinc and selenium.