Does Vitamin E Reduce Or Increase Mortality?
By Christine Doyle
In case you read the recent scare stories about vitamin E causing death and are worried that the supplement you are taking is doing you more harm than good, I’d like to explain that this isn’t what the recent review of seven studies, published in the Annals of Internal Medicine, actually found.
Before explaining the results, there are two factors you need to be aware of—the people involved and the statistical significance. Firstly, this study of studies involved only very sick people. One in ten participants were dead within five years. This is important because, if you are reasonably healthy then the results of this trial may have very little relevance for you.
Regarding statistics, scientists like to see a result with a statistical significance where .P>0.001.. This means that if you ran the experiment 100 times you’d get the same result 99 times. You can be very confident about such a level of statistical significance. A much weaker, although still statistically significant, effect has a .P>0.05.. Anything less than this is not statistically significant.
Now, the results. Overall, vitamin E supplementation did not affect all-cause mortality. I’m quoting the authors of this study here. However, they then split the group of people studied into high-dose vitamin E (above 400iu) and low-dose (below 400iu) and looked at the pooled results of these patients. Those on high-dose vitamin E had a weak but significant increased number of deaths (P=0.022), while those on the low-dose vitamin E had a weak but significant decreased number of deaths (P=0.021). In actual numbers what this meant was that, of those taking high dose vitamin E, 34 more people in 10,000 died. While, of those taking low dose vitamin E, 33 less people died. This might sound like a lot of people but remember, in the “control group” (i.e. a similar group not taking any vitamin E at all but used for comparison) 1,000 in 10,000 died.
Now you know the results, you might wonder why the media reports were so negative. Of course, the media love bad news, but they are not all to blame for the bias. So too are the researchers, not because of their study, but because of the conclusions they drew from it. They conclude by saying “High-dose (above 400iu per day) supplements may increase all-cause mortality and should be avoided.” I can’t fault them on that. But what about concluding “Low-dose vitamin E supplements may decrease mortality and should be recommended.” This is equally true, but not stated. So, can you really blame the media for emphasising the negative?
By the way, I recommend 200iu as a basic supplemental level, which is what I take and will continue to do so, strengthened by the evidence of this review.
But why, you rightly ask, would large doses of vitamin E slightly increase risk of death in seriously sick people? The authors propose that it’s because vitamin E, and indeed any anti-oxidant, can also be a pro-oxidant. To understand this, when an antioxidant sacrifices itself to disarm an oxidant, e.g. fried food, it becomes an oxidant. It then has to be disarmed itself by another antioxidant. That’s why you need the whole bomb disposal team, which includes vitamin E, vitamin C, glutathione (or cysteine), anthocyandins (berry extracts), beta-carotene and lipoic acid. That’s what I take in my daily antioxidant.
The trouble is that the trials analysed didn’t do this. They gave, in most cases, just high dose vitamin E on its own and it’s quite possible that this isn’t a good idea. Unfortunately, we don’t know because none of these trials actually measured oxidation. This is a big failure in trials to date because it means we simply don’t know if large doses of single vitamins can have a pro-oxidant effect. We can only guess.
However, my guess is that this isn’t the main reason why high-dose vitamin E, given to sick people, caused slightly more deaths. I’m going to guess it’s for another reason. The vast majority of the people in these studies were taking prescription medication, the most common of which was probably aspirin, since most had high cardiovascular risk. Aspirin thins the blood. So too does vitamin E. This anti-coagulant effect is quite marked with 1000iu of vitamin E. The combination of both may be too much and cause a bleed in the brain, hence increased risk of death. There is some evidence to support this view. For example, one of the studies found that the vitamin E group had a 75% reduction in heart attack compared to placebo but no reduction in mortality. Anti-coagulants can decrease risk of heart attack, but increase risk of stroke. The authors of this study didn’t take this into account. They could, by analysing the type of deaths and by seeing whether those taking vitamin E but no anti-coagulant drug, had less deaths than those taking vitamin E and anti-coagulant drugs. I’m only guessing, but this theory is possibly more plausible than the pro-oxidant theory. If it’s true, then the take-home message would be “Don't give high dose (above 400iu) vitamin E to anyone on anti-coagulant drugs” or “Don't give anti-coagulant drugs to anyone on high dose vitamin E.”
Another possible factor in need of research is whether giving alpha-tocopherol (vitamin E) somehow interferes with other families of tocopherols, such as gamma-tocopherol. Food sources of vitamin E, mainly fish, nuts and seeds, provide a broad base of tocopherols. (Some supplements also provide vitamin E in a broad base of tocopherols.)
In summary, this study supports the view that we should be supplementing vitamin E, probably in the range of 100 to 400iu, and that higher doses of vitamin E, on its own, might not be wise to give to very sick people, for reasons as yet unknown.
In the Optimum Nutrition Bible, my current average estimated Optimum Daily Allowance is 300mg/450iu. A healthy diet providing fish, nuts and seeds can supply 50mg/75iu, leaving a recommended supplement intake of 250mg/375iu. The range given in my book for supplementation is between 150mg/225iu (for healthy people and the level at which I personally supplement) and 400mg/595iu (for those with symptoms or lifestyles suggestive of increased need). These recommendations are broadly consistent with the results of the study reviewed here, which concluded the level of vitamin E most positively associated with reduced mortality was 221mg/330iu.
Please note: 100iu = 67mg, 400iu = 268mg, 1000iu = 670mg.