Vitamin E--Poisoning the Well

Saturday, March 15

While I was recently in Spain, I checked the internet for health stories, and was dismayed to see that another smear story about vitamins was hitting the papers back home, with lurid, misleading headlines like:

Vitamin E Supplements Could Cause Up to 27 Per Cent Increase in Lung Cancer

Vitamin E Linked to Lung Cancer

Vitamin E Warning

Taking Vitamin E Can Lead to Cancer

MY REACTION: These headlines are dumbed down versions of the actual study.

Typically, newspapers that generate press accounts from wire service reports have no analytical capability to contextualize the information. When you fail to process raw sewage, the results are predictable. Although I'm a respected nutrition expert, and available for quotes if journalists would take the time to do their due diligence, I almost never get called for my opinion before the papers frighten vitamin consumers out of their wits with stories like this.

After getting back to the States, I began doing some damage control on this study. First, let me state that I'm reluctant to go on the defensive about vitamins and supplements. Usually, if the negative stories are unsubstantiated, I don't even air them. Defending vitamin E after a story like this is a little like me saying "I don't beat my wife". What people remember is: "Did you hear that Dr. Hoffman was arrested on domestic violence charges?" But, with all the misinformation that's being promulgated, I feel I owe some accurate reporting to my listeners and patients.

So what did this study actually show? As reported in the American Journal of Respiratory and Critical Care, 364,418 Seattle area residents were mailed questionnaires asking them about various lifestyle factors, including the use of multivitamins, vitamin E, folic acid, or vitamin C.

Significantly, only 22% of questionnaire recipients returned their questionnaires (more about that later).

As might be predicted, a past or present history of smoking was the biggest predictor of lung cancer risk. So far so good.

The questionnaire recipients were asked if they had taken vitamins over the last ten years (more about that later, too).

When the researchers looked at vitamin C, folic acid, and multivitamins, there was NO relationship, either protective or detrimental, on the risk of lung cancer.

With regard to vitamin E, the researchers reported a relationship, but it was barely statistically significant (more on that later, too). They noted a paltry 5 per cent increase in the risk of lung cancer for each 100 mg (100 mg is approximately 100 IU) of vitamin E consumed. For example, for smokers who were daily consumers of a standard dose of 400 IU vitamin E, the risk of cancer was said to be increased by 20 per cent.

But there was NO increase in lung cancer for consumers of vitamin E who never smoked or were former smokers. The detrimental effect was only seen in current smokers.

What's the problem with this study and the alarmist reports it generated?

First, the press totally misrepresented the study, suggesting that vitamin E somehow "causes" lung cancer. Only in smokers was an adverse effect noted, and that was barely statistically significant.

What is "statistical significance"? Basically, it's a mathematical construct that's derived from the number of subjects studied in an experiment. For example, if you flip a coin 3 times, and it comes up heads all three times, you could not conclude that this coin delivers heads 100 per cent of the time. The sample is too small, and statistical analysis would show the results of your experiment are not statistically significant. Flip the coin 10,000 times, and when it comes up heads 5,017 times and tails 4,983 times, you can conclude with a high degree of statistical confidence that the likelihood of heads vs. tails is 50:50.

What about this study? Well, despite the large number of questionnaires sent out, only 77,000 responses were received. Sounds like a lot, but since lung cancer isn't all that common, only 521 lung cancers were reported by the respondents. That's a pretty small number for basing conclusions on vitamin E and cancer risk--barely statistically significant.

Now here's where it gets interesting. This study is what researchers call a "retrospective cohort" study. Retrospective studies look backwards in time, and rely on people's (sometimes faulty) memories. These studies are notoriously inaccurate, but unsophisticated newspaper journalists, and the general public, never took statistics classes, and don't know this.

The gold standard for drawing scientific conclusions is the "double blind placebo-controlled" study. But these take years to complete and are very expensive. Imagine recruiting 77,000 people and then feeding them vitamins or placebo pills and then waiting ten or twenty years to see who gets cancer?

What information gets lost when 79 per cent of questionnaire recipients don't get back to you? Do people actually remember if and when they took vitamins? Do they fudge what vitamins they took? Were they high quality or cheap discount vitamins? Does the fact that they got cancer change their recall? These are questions that the study--and the subsequent news reports--simply ignore.

Well known to statisticians is a phenomenon called "recall bias". Basically, the idea behind a form of recall bias is that when something bad happens to people, they tend to concentrate more on the minutiae of the antecedents leading up to their misfortune.

A good example is research performed a few years back for a popular book called "Dressed to Kill". The premise behind the book was that wearing bras causes breast cancer (the theory was that restrictive bras block lymphatic flow and allow "toxins" to build up in breast tissue).

A "study" was conducted by the author of the book that involved telephone polling. Women were asked about breast cancer history and bra-wearing (how 'bout that for a cold call!). The author claimed that there was a three-fold increase in breast cancer for women who regularly wore bras!

But recall bias offers an alternative explanation: women who got breast cancer said "Come to think of it, I WAS wearing a bra all those years leading to my breast cancer, maybe THAT's the reason I got it!" Recall bias highlights the intensely human need to perceive causality.

A safe conclusion from the present study on vitamin E is that neither vitamin E nor any of the other vitamins studied offers any protection from lung cancer IF YOU SMOKE.

I'm OK with that conclusion, because I continually reiterate that supplements can't combat the ravages of smoking. This would be like expecting a combo of vitamins to protect you completely from sunburn if you have light skin and broil in Acapulco for six hours continually without sunscreen; or, alternatively, to provide a total bulwark against radiation poisoning if you were to walk into the reactor at Chernobyl wearing a mosquito net.

A question not addressed by the study is what would have happened if the smokers had continually taken the RIGHT kind of vitamin E? Most vitamin E studies, especially the ones drawing negative conclusions, use the cheapest possible vitamin E, the dry or succinate form, or at best natural alpha tocopherol. These studies ignore the fact that vitamin E is present in nature in MIXED tocopherol form, rich in gamma tocopherol which appears to be decisive for vitamin E action.

In fact, studies now show that when you give alpha tocopherol in the absence of gamma tocopherol, tissue levels of vitamin E actually DECLINE. Gamma tocopherol appears to be essential for vitamin E bio-availability. Maybe that's why the addition of the wrong kind of vitamin E to the regimens of these smokers actually set them back, rather that conferring at least partial protection.

Finally, NO study has yet examined the effects of a truly state-of-the art, high quality supplement program on people facing oxidative stress, like smokers, or 9-11 Ground Zero rescue workers. Such a supplement program would have to include not just premium mixed tocopherol vitamin E rich in gamma tocopherol, but also natural mixed carotenoids, C, zinc, selenium, N-acetylcysteine, alpha lipoic acid, EGCG, sulforophane glucosinolate (SGS), intravenous glutathione, etc.

But after fiascos like the current vitamin E/cancer "study" and its attendant mis-reporting in the media, the likelihood of additional good antioxidant research being performed is zilch. And that is truly poisoning the well for conscientious Americans who take supplements to safeguard their health.

There, now I've said it, and it's off my chest. WHEW! And no, I don't beat my wife--I'm not even married!