To sign up for ALL of our updates, including the new Show Me the Science services, send a blank email to
NDannounce-subscribe@topica.com, or click the link below:
All Comparative Guide News
To sign up for only the Show Me the Science services news, please fill out this short contact form
Is what you see really what you get?
In the United States, between 20 to 40 percent of all dietary supplements on the market fail to meet labelling standards. What’s more, the burden of proof for unsafe or adulterated products, and false or misleading labelling, lies with the federal Food and Drug Administration rather than with the manufacturer. Ouch!
Recent Vitamin E Study Reaches Unfounded Conclusions
by Lyle MacWilliam, MSc, FP
Nineteenth century British statesman Benjamin Disraeli once mused: "There are three kinds of lies: lies, damned lies, and statistics." The recent announcement by John Hopkins University researchers Edgar Miller and coworkers that high vitamin E intake may increase the risk of early death is an exemplary case in point. The findings of the study, “Meta-Analysis: High Dosage Vitamin E Supplementation May Increase All-Cause Mortality,” presented at the American Heart Association meeting in New Orleans, November 10, 2004, have been exaggerated and sensationalized. The conclusions drawn are unsupported by the data, and the announcements made are irresponsible and alarmist.
The fall-out has been agonizingly predictable. Media outlets across Canada, the United States and elsewhere have taken the bit at a full run, trumpeting the message that high-dose vitamin E may be deadly, with no consideration of the study’s limitations or the wealth of scientific evidence to the contrary. Consumers, confused by the headlines and unable to decipher the details, have justifiably reacted with fear.
The study finds marginally significant evidence of harm from high-dose intake of vitamin E, which the authors, themselves, caution against generalizing to a healthy population. Unfortunately, the authors ignored their own guidance when breaking the news during the press conference, raising the specter of harm to the general public. The findings are at complete odds with the wealth of scientific evidence: past studies on vitamin E supplementation generally support the contention that high levels of vitamin E confer many protective benefits and that supplementation constitutes a sensible approach in the prevention and mitigation of chronic degenerative disease. A Medline key-word search of over 3300 scientific studies from 1964 to the present, conducted by this writer subsequent to the announcement, found scant evidence of reported harmful effects for vitamin E. The vast weight of scientific evidence is to the contrary.
Study Limitations and Flaws
It is important to understand that the Miller study was not a clinical trial but a meta-analysis, a statistical technique where researchers combine data from several available clinical trials to tease out a common thread, if indeed there is one. Such analyses are highly speculative because the scientific protocol and quality of each study is sufficiently different that it is extremely difficult to accurately link the results to reveal a common factor—in this case whether high-level supplementation with vitamin E affects mortality. A good meta-analysis is a useful tool that can put apparently contradictory evidence into perspective. The weakness in the approach, however, is that inclusion of poorly designed studies can lead to faulty conclusions: “garbage in, garbage out.” [1] Such is the case with this study.
In announcing their findings, the authors failed to provide appropriate limitations to their conclusions, such as the population to which the results pertain and the type of supplementation which should be cautioned against (in this case, the isolated use of high-dose synthetic vitamin E). As well, the study design has several flaws and weaknesses, which serve to skew the results:
- Many of the studies included in the analysis were conducted on elderly subjects who were already suffering from advanced chronic degenerative disease, rather than on healthy adults. For example, one study included 196 patients and 60 of them died. That means that 31 percent of the patients died during the study period —clearly, these patients had significant health challenges. Consequently, the results of the analysis simply cannot be extrapolated to the average healthy person. Nevertheless, in their rush to create headlines, the authors ignored their own scientific caution regarding the study’s limitations (which states that the results cannot be applied to a healthy population) and reached the remarkable conclusion that high vitamin E intake could be deadly—end of story.
- In selecting the studies from which to pool their data, the authors intentionally excluded several smaller studies and those in which there were fewer than 10 deaths reported in the trial. Exclusion studies with little or no mortality does not seem justified and further biases the results to favor the evidence of harm.
- A small risk difference in one subset of the data is reported as statistically significant. In fact, multiple testing such as this should be controlled with a standard statistical correction, the Bonferonni correction. When this correction is applied, the already marginal results are rendered “statistically non-significant.”
- While the authors state that high-level intake of vitamin E is dangerous, their analysis reveals that “overall, vitamin E supplementation did not affect all-cause mortality” and that, at the highest dosages, the risk of death was actually lower. Dr. David Freeman, a statistician at the University of California at Berkeley has stated, “As a statistician, I find this paper unpersuasive.” This sentiment is also reflected by Dr. James Robins, a statistician at Harvard University, who concludes that the results are “somewhat oversold statistically” and hardly amount to definitive evidence.
- In clinical studies, an ‘intention-to-treat’ analysis is often carried out. This type of analysis assesses the success of a treatment given that some patients forget to take their medication (in this case, vitamin E). This procedure underestimates the effect that a treatment has on individuals who do comply and in the case of the Miller study biases the results to favor harm. It may well be that some of the people who died in the high-dose vitamin E group were not following protocol and thus not actually receiving the supplement. Because non-compliance is not reported, there is no means by which to accurately assess this bias, which may be responsible for the small increased risk associated with Vitamin E.
- As noted by the authors, the high-dose studies used in the meta-analysis had much smaller sample sizes than the low-dose studies, which showed no harmful effect from vitamin E intake. Therefore, there is more confidence in the results of the low-dose group (which showed no harm) than there is in the high-dose group (which showed marginally significant harm).
- The conclusions drawn are also flawed in that the authors do not clearly state that their findings only pertain to the isolated use of high-dose vitamin E. Most of the studies included in the report used vitamin E as a stand-alone nutrient, rather than as part of a holistic treatment regime. It has long been known that nutrients work best when they work together; this axiom is particularly true about antioxidants, such as vitamin E, that work to replenish one another (antioxidants are nutrients known to reduce the level of oxidative damage in the cells of the body). Studying high doses of a single antioxidant in isolation may provide results that are markedly different than when the same antioxidant is provided as part of a complete nutritional approach. A good example of this is the alpha tocopherol and beta carotene cancer prevention study, conducted in Finland in 1994. The initial results of this study, when the effects of beta carotene (a type of antioxidant) were viewed in isolation, led to the erroneous conclusion that high levels of beta carotene increased the risk of cancer in male smokers. To be expected, media headlines associated an increased cancer risk with vitamin usage. Despite objections that the findings of the study were flawed, the charge stuck and vitamin use dropped. Fast-forward to July 2004 when the same data, using total antioxidant status, was reviewed in the American Journal of Epidemiology.[2] When total antioxidant intake was considered, the findings of the 1994 study were found to be in error. The new study, using the same data as the original study came to a remarkably different conclusion: beta carotene in combination with other dietary antioxidants significantly reduced cancer risk in the same male smokers. Two different approaches, two different conclusions—same data.
Type of Vitamin E Used
The study by Miller and coworkers does, however, raise a legitimate concern regarding the type of vitamin E used both in research studies and some nutritional supplements. There are actually several natural tocopherols that demonstrate biological activity, the most active being alpha tocopherol. There are also the beta, gamma, and delta tocopherols, along with a group of related compounds known as the tocotrienols.
To explain their findings, the authors suggest that high doses of alpha tocopherol, used in the reported studies, may displace the other forms of vitamin E, particularly the gamma form, which may have a harmful effect. This argument has some merit because we know that beta, gamma and delta tocopherol confer several important health benefits. They are anti-inflammatory, anticarcinogenic and cardioprotective. In a 1996 study, reported in the New England Journal of Medicine, researchers found that a high dietary intake of natural vitamin E (containing mixed tocopherols), but not of alpha tocopherol supplements, was associated with lower stroke and cardiovascular mortality.[3] Another study found that men with coronary heart disease were low in gamma tocopherol, not alpha tocopherol, suggesting that the gamma form is just as important as alpha tocopherol in preventing heart disease.[4] In addition, several studies have found that high blood levels of gamma tocopherol are correlated with a reduced incidence of cardiovascular disease and mortality.[5] Others report that blood levels of gamma tocopherol, not alpha tocopherol, are lower in cardiovascular patients than in healthy adults. While less effective as an antioxidant, gamma tocopherol is much better able to trap nitrogen-based free radicals. This protection against nitrogen free radicals is extremely important in the prevention of heart disease, degenerative brain disorders and cancer[6] (free radicals are molecules with unpaired electrons which are responsible for the oxidative damage in cells).
While alpha and gamma tocopherol are absorbed equally through the gut, the liver preferentially binds alpha tocopherol to lipoproteins, which then transport it through the body. It is suggested that the intake of high levels of alpha tocopherol, through supplementation, may selectively inhibit the uptake of gamma tocopherol, thereby depressing its presence in the body. Recent findings confirm that taking large doses of alpha tocopherol can deplete levels of gamma tocopherol, whereas supplementation with high levels of the gamma form have a sparing effect on the level of alpha tocopherol, presumably by reducing its degradation in the liver. While this does not mean that supplementation with alpha tocopherol only is harmful, it does indicate that high-dose supplementation of the alpha form of vitamin E without the attendant gamma form, found naturally in foods, is not optimal. This suggests a serious need to re-consider the common practice of supplementing with alpha tocopherol only.
Natural or Synthetic?
Another concern exposed by the Miller study is the use of synthetic vitamin. Each of the forms of vitamin E has a right-handed (d-) and left-handed (l-) version, which are mirror images of each other. These are known as isomers. The human body recognizes only the right-handed (d) versions and simply excretes the left-handed (l) versions. Synthetic vitamin E, made commercially, contains both the d- and l-isomers. On the other hand, natural vitamin E, found in plants, contains only the d-isomers of the various tocopherol forms. Less expensive nutritional supplements use synthetic vitamin E or d/l vitamin E, half of which is useless to the body; quality supplements will, instead, use the natural and more biologically active d-forms of the vitamin.
There is no evidence of adverse effects from the consumption of natural (d) vitamin E; however, the U.S. National Academies of Science reports evidence of adverse effects from the synthetic (d/l) forms of the vitamin, commonly found in many lower-priced supplements, which includes hemorrhagic toxicity. Consequently, in their recent redesign of the new Dietary Reference Intakes, the Food and Nutrition Board of the National Academies established a Tolerable Upper Limit of intake of 1500 International Units (IU) per day for synthetic vitamin E. Note that this cautionary limit, set by an international panel of experts, is much higher than the levels of daily intake that Miller and coworkers suggest is harmful. Many of the studies cited in the Miller paper used the synthetic form of vitamin E, rather than its natural form, throwing even more doubt on the validity of the results.
Conclusions
So, what does the Miller analysis actually reveal? Despite the authors’ sensationalization of their findings—delighting the pharmaceutical interests who helped fund the study—the results fail to make the case that high vitamin E intake was the cause of death among study participants, most of whom were already inflicted with life-threatening illnesses. The results, marginal at best, are further compromised because of poor study design; the conclusions rendered simply cannot be applied to the general population. Moreover, the conclusions are at odds with the cumulative scientific evidence which demonstrates the value of supplementation with vitamin E. One bad apple shouldn’t spoil the whole bunch.
However, the results do raise legitimate questions as to the type of vitamin E that should be used in nutritional supplement. This suggests that further research on the use of alpha tocopherol versus mixed tocopherols and the use of natural versus synthetic forms of the vitamin is warranted.
The results also highlight the folly of looking at high levels of intake of a single nutrient, rather than investigating the value of a broad-spectrum approach to nutritional intervention. It is unlikely these same results would have been found in patients supplementing with a balanced nutritional formulation containing a wide range of nutrients. The message here is that people should not take a high dose of one supplement without considering that it may increase the need for other nutrients. Nutritional supplementation requires a holistic approach rather than using a single nutrient in high doses as if it were a drug.
As noted U.S. humorist and author Mark Twain once said: “Facts are stubborn things, statistics are more pliable.” The Miller analysis showcases the dangers of overstating the case. Unfortunately, in doing so, Miller and his coworkers have spread needless anxiety and fear amongst the public and have provided fodder for the pharmaceutical interests—which do not profit from preventive, natural and holistic approaches to wellness.
Author, educator and biochemist, Lyle MacWilliam serves as a consultant and public advocate for the natural healthcare industry. Mr. MacWilliam is a former Canadian Member of Parliament and Member of the Legislative Assembly for British Columbia. He recently served at the behest of Canada’s federal Minister of Health to help develop a new regulatory framework for natural health products, ensuring Canadians access to safe, effective and high quality nutritional supplements. Mr. MacWilliam has also served in a consulting capacity for Environment Canada, Human Resources Development Canada and the British Columbia Science Council. He is the author of the Comparative Guide to Nutritional Supplements, which examines the role of supplementation as part of a healthy lifestyle. He can be reached at www.comparativeguide.com.
References
1. Levin N. Vitamin E Warning Premature. Ann Intern Med. Available at: http://www.annals.org/cgi/eletters/0000605-200501040-00110v1. Accessed Nov 11, 2004
2. Wright ME, Mayne ST, Stolzenberg-Solomon RZ, et al. Development of a comprehensive dietary antioxidant index and application to lung cnacer risk in a cohert of male amokers. Am J Epidemiol. Jul 2004;160(1):68-76.
3. Kushi LH, et al. Dietary antioxidant vitamins and deth from coronary heart disease in postmenopausal women. New Engl J Med. 1996;334:1156-62.
4. Koyiyama K, et al. Studies on the biological activity of tocotrienols. Chem Pharm Bull. 1989;37:1369-81.
5. Greenwell, I. Newly Discovered Benefits of Gamma Tocopherol. Life Extension. October 2002:39-42.
6. Ibid.