Send rapid response to journal:
Re: Natural or Synthetic?

I would like to know the type of vitamin E used in the trial by Miller, et al. It is very important to know in how many studies the Vitamin E used was the Natural or synthetic form (d- or dl- tocopherols [also known as all-rac- -tocopherol], respectively), or other forms (beta, gamma, or mixed). The results of the study, especially as a meta-analysis could make a big difference in your findings. I am very concerned that the results may be confounded if we identify the types of vitamin E were used.

According to the article by Miller, et al, “the biological activities of vitamin E compounds are reported relative to all-rac- -tocopherol acetate on the basis of in vivo assays.” Also, the authors converted the vitamin E dosages of the studies included in the meta-analysis to IU/d relative to all-rac- -tocopherol acetate for standardization across studies. I am concerned about the Publication bias based on different sources.

Also, the study by Miller, et al, could not evaluate the generalizability of the findings to healthy adult populations taking high- dosage of vitamin E. Although it was mentions in the limitations section, I am curious what will the result be among healthy adults and not among patients with various chronic diseases.

Conflict of Interest:

Send rapid response to journal:
Re: Re: Vitamin E & all-cause mortality

Until recently, recommended dietary patterns included up to 10 percent of calories from polyunsaturated oils, an unusually large quantity of these easily oxidized fats. However, during production of common polyunsaturated oils, there may be loses of about 30 percent or more of naturally-occurring vitamin E,(1) predominantly gamma-tocopherol. Additional vitamin E is destroyed in food processing, preparation and storage.(2) Although vitamin E supplementation could theoretically correct potential deficits, large doses greater than 400 mg could also exaggerate an imbalance of alpha tocopherol vs other natural vitamin E compounds, perhaps accounting for the disappointing results of the Miller et al analysis.(3)

It is possible, however, that smaller doses of vitamin E could be beneficial. In fact, in the few trials of 150 mg or less included in this meta-analysis, mortality risk was lower. Therefore, a trial of lower dose vitamin E, ideally a blend of RRR-tocopherol compounds that more closely approximates that found in natural sources, seems warranted.

1. Jung MY, Yoon SH, Min DB. Effects of processing steps on the contents of minor compounds and oxidation of soybean oil. JAOCS 1989;66:118-120.

2. Bauernfiend JB. Tocopherols in foods. In: Machlin LJ, editor. Vitamin E, a comprehensive treatise. New York: Marcel Dekker Inc; 1980. p.104-17.

3. Edgar R. Miller, III, Roberto Pastor-Barriuso, Darshan Dalal, Rudolph A. Riemersma, Lawrence J. Appel, and Eliseo Guallar.Meta-Analysis: High-Dosage Vitamin E Supplementation May Increase All-Cause Mortality. Ann Intern Med 2004; 0: 0000605-200501040-00110-53.

Conflict of Interest:

Send rapid response to journal:
Re: Vitamin E & all-cause mortality

In the recent meta-analysis that associates vitamin E (E) supplementation with all-cause mortality, Miller et al. 1 reported that supplemental intake of E at doses greater than 150 IU/d progressively increases all-cause mortality. There are several problems with the analysis and interpretation of the data. The harmful effect above 150 IU/d is an artifact of the model they chose to fit. By fitting a quadratic- linear spline, the result must be a quadratic-linear spline whether or not the data behave that way. Another reasonable model is that the favorable response to E is constant until a certain dose above which the change in all-cause mortality risk difference is linear. The constant response to E persists up to a dose of 330 IU/d and the risk difference favors E until the dose is 400 IU/d. The sum of weighted squared differences between observed and predicted all-cause mortality risk differences shows that this model fits the data better than does the quadratic-linear spline.

The modeling issue is further supported by the following re-analyses of the data presented in Figure 2 of Miller et al. 1. We grouped the data into several intermediate dose ranges of E and performed meta-analyses using a random effects model and obtained the following results:

200-500 IU 7 studies 35,595 patients RR=0.98 (0.88 - 1.09) 330-500 IU 6 studies 32,184 patients RR=0.99 (0.88 - 1.11) 400-500 IU 4 studies 16,355 patients RR=1.00 (0.80 - 1.25).

None of the subgroups that included studies with E dosage of 500 IU/d or below suggests any harmful effects. Thus, it doesn't seem plausible that increased risk occurs until at least the daily dose of 400-500 IU has been exceeded.

We also question the applicability of the data to the general population as subjects in most studies with higher doses included in the meta-analysis had cardiovascular diseases. Furthermore, we are surprised at the lack of emphasis on the benefits of reduction in all cause mortality by doses of E below 400 IU, an effect, which reached statistical significance if E alone was considered. Related to this, in a double- blind, placebo-controlled trial of E supplementation (200 IU/day for 1 year) in nursing home subjects, we observed no statistically significant difference in all cause mortality between placebo and E groups (12.5% and 14.4% in E and placebo groups, respectively). However, significantly fewer subjects in the E group acquired respiratory tract infections, an important public health problem in this age group 2.

Simin Nikbin Meydani,*§ DVM, PhD Joseph Lau,+ MD Gerard E. Dallal,* PhD Mohsen Meydani,* DVM, PhD

*Jean Mayer USDA Human Nutrition Research Center on Aging, Boston, Massachusetts, §Department of Pathology, Sackler Graduate School of Biochemical Sciences, Tufts University, Boston, Massachusetts, +Institute for Clinical Research and Health Policy Studies, Tufts-New England Medical Center, Boston, Massachusetts,

References

1. Miller ER, 3rd, Pastor-Barriuso R, Dalal D, Riemersma RA, Appel LJ, Guallar E. Meta-Analysis: High-Dosage Vitamin E Supplementation May Increase All-Cause Mortality. Ann Intern Med 2004. 2. Meydani SN, Leka LS, Fine BC, et al. Vitamin E and respiratory tract infections in elderly nursing home residents: a randomized controlled trial. JAMA 2004; 292:828-36.

Conflict of Interest:

Send rapid response to journal:
Re: High dose vitamin E supplementation: time to return to the drawing board....

The work of Miller et al(1) highlights the danger of assuming the safety of high dose vitamin E in the absence of definitive long-term safety data. Its impact, however, may be mitigated, by methodological concerns.

The first issue is the restrictive inclusion criteria stipulating that a trial have at least 10 deaths to be included in the meta-analysis, apparently because the authors “anticipated that many small trials did not collect mortality data”. This exclusion contradicts the raison d’être of meta-analysis, which involves the statistical pooling of multiple trials that individually have inadequate statistical power. The exclusion of at least 3 reasonably large well-conducted trials(2-4) of high dose vitamin E in which fewer than 10 deaths occurred, while including only trials meeting this arbitrary cut-off, would spuriously increase the power of the meta-analysis. We would also be interested in the funnel plot analysis to determine whether publication bias affected the study results.

Although the authors attempted to adjust for average follow-up in their analysis, a more robust statistical treatment of the variance in follow-up periods across included trials would be to express the summary statistic of pooled death risk as the number of deaths per 10,000 person- year (as opposed to per 10,000 persons).

Heterogeneity in the study populations may not have been fully accounted for despite the use of the random effects model and dosage differentiation. In particular, cardiovascular disease (CVD) may constitute a select group at distinct risk from the effects of high dose vitamin E. Seven of the eight high-dosage trials showing harmful effects of vitamin E involve subjects with vascular risk factors or who had established CVD. In contrast, the DATATOP and ADCS studies utilised mega- doses of vitamin E (2000 IU/day) in individuals with neurodegenerative disorders, rather than CVD, but did not reveal safety concerns with vitamin E. A separate meta-analysis looking solely at the group of neurodegenerative diseases (including a recent study employing 5000 mg/day of vitamin E(5)) may be warranted.

Although the focus of the Miller et al study may be on safety, the data ultimately challenges the advocates of high dose vitamin E to re- examine the evidence for its benefit. Efficacy from controlled trials ranges from minimal to modest, in contrast to the more positive results of observational studies. It is time clinicians return to the drawing board and review both the safety and efficacy data for vitamin E supplementation to determine their practice.

References: 1. Miller ER 3rd, Pastor-Barriuso R, Dalal D, Riemersma RA, Appel LA, Guallar E. Meta-analysis: High-dosage vitamin E supplementation may increase all-cause mortality. Ann Intern Med. 2005;142:37-46. 2. Berson EL, Rosner B, Sandberg MA, Hayes KC, Nicholson BW, Weigel- DiFranco C, et al. A randomized trial of vitamin A and vitamin E supplementation for retinitis pigmentosa. Arch Ophthalmol. 1993;111:761- 72. 3. de Waart FG, Kok FJ, Smilde TJ, Hijmans A, Wollersheim H, Stalenhoef AF. Effect of gluthathione S-transferase M1 genotype on progression of atherosclerosis in lifelong male smokers. Atherosclerosis. 2001;158:227- 31. 4. Wluka AE, Stuckey S, Brand C, Cicuttini FM. Supplementary vitamin E does not affect the loss cartilage volume in knee osteoarthritis: a 2 year double blind randomized placebo controlled study. J Rheumatol. 2002;29:2585-91. 5. Graf M, Ecker D, Horowski R, Kramer B, Riederer P, Gerlach M, et al. High dose vitamin E therapy in amyotrophic lateral sclerosis as add-on therapy to riluzole: results of a placebo-controlled double-blind study. The German vitamin E/ALS Study Group. J Neural Transm 2004; Oct 27 (Epub ahead of print).

Conflict of Interest:

Send rapid response to journal:
Re: Not All Vitamin E Should Be Condemned : A Different Perspective

In light of this controversial vitamin E meta-analysis publication and the so many response by experts to it in the media, perhaps, one should look at vitamin E supplementation from a different perspective - that vitamin E should be taken as a wholesome mixture of d-mixed tocopherols and d-mixed tocotrienols.

Taking a single form of vitamin E (ie : alpha-tocopherol alone) denies the very fact that nature put seven (7) other forms on vitamin E (ie : gamma-tocopherol, beta-tocopherol, delta-tocopherol, alpha- tocotrienol, beta-tocotrienol, gamma-tocotrienol and delta-tocotrienol) out there for a reason.

As a matter of fact, we have seen this before - in 1996 with the beta -carotene debacle (The ATBC and CARET studies). These two studies provide evidence that taking beta-carotene alone rather than a multi-carotenoids (beta-carotene, alpha-carotene, gamma-carotene, lycopene, lutein - as produce in nature), may increase the cancer risks among smokers. This may be because all these carotenoids work synergistically as a team - recharging and supporting each other to confer the health benefits.

It goes to show that a single nutrient vitamin E (ie : alpha- tocopherol - synthetic or natural) is not the panacea. It is against conventional wisdom to take mega-doses of one nutrient without considering the potential side effects.

Similarly, high dosage of alpha-tocopherol alone has been shown to deplete the body's gamma-tocopherol. Despite alpha tocopherol's action as an antioxidant, gamma tocopherol is required to effectively remove the harmful peroxynitrite-derived nitrating species. Because large doses of dietary alpha tocopherol displace gamma tocopherol in plasma and other tissues, the current wisdom of vitamin E supplementation with primarily alpha tocopherol should be reconsidered. Other forms of vitamin E - gamma- tocopherol, delta-tocopherol and certainly tocotrienols have been proven to have unique health properties as well.

Tocotrienols on the other hand have been proven to be beneficial to the cardiovascular system. It is a potent antioxidant (40-60 times more potent) and has been proven to lower total blood cholesterol as well as reverse arterial blockage in Carotid Stenosis patients. In a recent NIH- funded study in collaboration with Ohio States University Medical Center and Carotech Inc (Tocomin®), it was found that tocotrienols especially alpha-tocotrienol are much more potent than tocopherol in protecting the neurons from glutamate-induced neuro-degeneration.

In many foods, alpha- and gamma-tocopherol account for most of the vitamin E activity. While tocopherols are generally present in common vegetable oils (i.e. soy, canola, wheat germ, sunflower), tocotrienols, on the other hand, are concentrated in cereal grains (i.e. oat, barley, and rye, rice bran), with the highest level found in crude palm oil. It is unfortunate that not many consumers are aware of tocotrienols due to the low level in the Western diets.

Looking at the whole realm of vitamin E research, it is prudent to take the wholesome full spectrum vitamin E : d mixed tocopherols + d-mixed tocotrienols - as what is produce and found in nature.

Mimicking nature is the best way for supplementation. Like the carotenoids, all these different forms of vitamin E work synergistically and depend on each other for optimum functionality.

Natural phytonutrients just don't work well in isolation from each other. I sincerely believe (from scientific evidence) that most people would benefit from taking a full spectrum Vitamin E supplement that consists of d-mixed tocopherols + d-mixed tocotrienols. And it would be safer than just the alpha-tocopherol alone.

Conflict of Interest:

Send rapid response to journal:
Re: ARBITRARY DEFINITION OF HIGH DOSE

In a recent vitamin E meta-analysis, authors from Johns Hopkins evaluated the results from 19 studies on vitamin E and examined whether vitamin E had an effect on total mortality. In the overall analysis, the found no effect on mortality, and in the dose-response analysis, they found a significant effect on total mortality only at intakes above 900 IU, a finding entirely lost in the furor over their third analysis.

For the main meta-analysis, the authors selected 400 IU as the dividing line between “low dose” and “high dose” trials and asserted that levels of 400 IU or more were associated with increased mortality. This is an arbitrary figure which happens to coincide with the most commonly marketed level of vitamin E. There are 2 studies at 330 IU, including the large and important GISSI study which found a decrease in total mortality. Inclusion of these 2 studies in the “high dose” category would have been logical and would have drawn the effect toward null.

There are only 2 studies at 400 IU. The largest of those is the HOPE study which found absolutely no effect on total mortality, and the other is the AREDS study which found no significant increase. The one study at 440 IU found a statistically significant decrease in total mortality in the vitamin E group. Thus, there is no evidence of an increased mortality risk in the 3 studies that actually gave 400 or 440 IU of vitamin E to more than 7500 people for a period of years.

The authors’ finding of a slightly increased risk of total mortality is driven entirely by the 8 studies using dosages of 500 IU or more. There is only one large study in this group, the MRC trial which gave 660 IU and found a slight but nonsignificant increase in risk. The other trials are all small, and five have wide confidence intervals running off the chart in one or both directions. These studies are a poor basis for the authors’ sweeping conclusions. Millions of consumers are now confused because the researchers failed to take their own good advice to use caution in generalizing the results of this meta-analysis.

Conflict of Interest:

Send rapid response to journal:
Re: Study is too simple

This study fails to differentiate between:

- Types of Vitamin E used: artificial (dl) or natural(d).

- Age (old versus young) of the populations.

- Other items taken: drugs or vitamins/minerals.

- Levels of health/illness as they influence the study participant’s choices. The failure to differentiate leads me to consider: 1. The principals were too naïve to attempt this type of study; or, 2. This study has been planned in order to support (later) legal attempts to reduce Vitamin E to a prescription status.

Conflict of Interest:

Send rapid response to journal:
Re: Eat an apple a day...and take your vitamins!

The meta-analysis by Miller et al has become part of the public landscape, because of the sensational headlines in newspapers around the world: "High Doses of Vitamin E Deadly," "Vitamin E's Fatal Flaw," "Lethal Consequences of Vitamin E Overdose." The wall-to-wall press makes it seem like one of the most important health warnings of our era, but this meta-analyses is nothing more than a tempest in a teapot brewed from a statistical study that many epidemiologists would not give much credence to.

For physicians (or their patients) unable to read past the alarming headlines, here's the story: The authors combined the findings in 19 previously published studies on vitamin E over the last 12 years. Virtually all of them failed to show any statistically significant harm--much less any increase in deaths. However, by combining the 19 old studies, the authors' believe they have found a statistically significant increase in deaths from all causes of mortality (ACM).

I believe the authors' bias is clear when they cite an apparent 0.4% increase in ACM at doses over 400 IU vitamin E while providing almost no comment on hundreds of excellent studies that show no increase in ACM--while also demonstrating benefits of vitamin E supplementation. Certainly as physicians assess the risk of vitamin E, its documented benefits muct be factored into the equation. Yet the authors fail entirely to acknowledge the benefits of vitamin E that were reported in the very same studies that they included in their own meta-analysis: Reduction in the risk of Alzheimer's disease, heart and blood vessel disease, age-related macular degeneration and several forms of cancer.

How narrowly informed is this meta-analysis? I believe the overwhelming fact is that the increase in mortality was just 39 out of 10,000 for those taking 400 IU of vitamin E or more. That's less than a one half of one percent increase in ACM, which may include floods, famine, earthquakes, homicides, suicides and accidents!

The reason why many epidemiologists do not give much credence to these types of meta-analyses is because statisticians are forced to ignore the nuances in each of the studies they have been examined. Interestingly, the data in the present statistical leap of faith suggest 400 IU vitamin E is more lethal in 1.4 - 8 years than smoking 2 packs of cigarettes for the same period of time. Clearly, this is not tenable. As a matter of fact, the Institute of Medicine has concluded that vitamin E is safe at levels as high as 1000 IU per day for the synthetic form and 1500 IU per day for the natural form.

The findings reported in this sensationalized study are really not news at all, and the headline hype would be just silly if it wasn't also a major distraction from genuinely important health news of the moment: tens of thousands of patients may have died as a result of taking Vioxx, because certain pharmaceutical firms may not have been completely honest with doctors or the public.

Read the newspapers or watch TV and anyone can see how much the drug industry influences the media with valuable advertising dollars. Drug ads have also become the economic lifeline for medical journals, which generally contain more drug advertisement pages than pages with scientific studies. Dr. Marcia Angell, the former editor of The New England Journal of Medicine, has written a book entitled, "The Truth About the Drug Companies: How They Deceive Us and What to Do About It," detailing how the companies are vast marketing empires with unprecedented control and influence over medical research, education, and how doctors do their jobs.

So in addition to scientific, evidence-based medical care, what's a sane, rational prescription for good health? Here's my "Wellness Solution:" healthy nutrition, weight management, drinking enough water, regular exercise, stress management, including being a good person, no smoking, no substance abuse, including alcohol, and taking a well-balanced vitamin and mineral supplement containing 400 IU of vitamin E every day--which is what I take myself.

Send rapid response to journal:
Re: Simple Statistics support paper's findings

If the clinical trials are considered to be independent of each other and if the clinical trials have been selected with enough care so that the trials are representative and the administration of extra doses or not of vitamin E is the only significant variable, than it is a reasonable statistical inference that the administration of large doses of vitamin E is harmfull to patients, and by reasonable extension to the healthy as well.

Of the 11 clinical trials which administered 400 or more units of vitamin E, 8 showed "excess" mortality associated with "excess" vitamin E.

If vitamin E is considered beneficial there is only a 11% chance that this can happen by chance (3 or less adverse events in 11 trials).

While such a result is in excess of the time honored criteria of reducing chance to a 5% or 1% level,that is only a convenient rule not an absolute one.

Reference:Probability and its Engineering Uses,1928,Thornton Fry

Conflict of Interest:

Send rapid response to journal:
Re: Incorrect Statistics Used in Vitamin E Report

The report on high dosage Vitamin E supplementation combined studies of patients with various degrees of medical conditions from healthy to seriously ill. Most of the seriously ill took high doses of Vitamin E. These combinations are statistically incompatible and the results are not reliable. Since the high doses of Vitamin E were taken mostly by seriously ill patients, the study would seem to be biased toward higher mortality rates for higher Vitamin E doses. The study does not show that high doses of Vitamin E are a risk for healthy individuals and this is admitted by the researchers. In their report the researchers write on page 9 (internet copy) that "we could not evaluate the generalizability of our findings to healthy adult populations."

High doses of Vitamin E might be harmful but this study does not provide proof.

Conflict of Interest:

Send rapid response to journal:
Re: Interaction of drugs with vitamin E

Since most of the subjects in these studies were older people with chronic diseases, many of them were probably taking multiple prescription drugs. Perhaps some of the mortality associated with vitamin E could be explained by an interaction of vitamin E with these drugs similar to the known interaction of an antioxidant cocktail containing vitamin E with statins and niacin that reduces the increase in HDL cholesterol.

Conflict of Interest:

Send rapid response to journal:
Re: Safety of Vitamin E Supplementation in Humans

Send rapid response to journal:
Re: Lancet was correct!

This meta-analysis combined 19 individual studies, eighteen of which showed no statistically significant increase in mortality. This can be verified by reviewing the data in Figure 2 with a simple calculator. One study alone accounted for the entire 5% negative death rate (MRC/BHF HPS 2002 [49]).

Interestingly this Lancet study made an entirely different conclusion from your above study. Combining of 19 clinical trials into a single large cohort gave greater statistical power than the Lancet study itself concluded. And, thus, can not be justified.

Lancet reported: "There were no significant differences in all-cause mortality (1446 [14.1%] vitamin-allocated vs 1389 [13.5%] placebo- allocated), or in deaths due to vascular (878 [8.6%] vs 840 [8.2%]) or non -vascular (568 [5.5%] vs 549 [5.3%]) causes." ... "Among the high-risk individuals that were studied, these antioxidant vitamins appeared to be safe."

And, the answers that everyone has been requesting are: alpha- tocopherol along with vitamin C and beta-carotene. Their use of beta- carotene taints the whole Lancet study in my opinion.

The only thing that that the Lancet study showed was that alpha- tocopherol was not a miracle cure for gravely ill individuals with coronary disease, occlusive arterial disease, or diabetes. While not a glowing recommendation for the use of Vitamin E, it certainly did not conclude that its use increased the risk of all cause mortality.

[49]MRC/BHF Heart Protection Study of antioxidant vitamin supplementation in 20,536 high-risk individuals: a randomised placebo- controlled trial. Lancet. 2002;360:23-33. [PMID: 12114037]

Conflict of Interest:

Send rapid response to journal:
Re: this article

As a long-time user and occasional prescriber of high-dose vitamin E, I read with interest the meta-analysis of studies evaluating use of high- dose Vitamin E supplements. I would like to see a study or even a meta- analysis of studies of HEALTHY patients using these supplements. I agree that unmonitored usage of OTC supplements can potentially do more harm than good, especially in acutely or chronically ill elderly patients, but what about the rest of us? Unfortunately, funding for such a study would probably be difficult to obtain. Any ideas?

Conflict of Interest:

Send rapid response to journal:
Re: question re: study design

It seems questionable that a study that was designed to determine the effects of vit E on mortality would exclude studies that did not have any mortalities. Omitting clinical trials that collected mortality data and had fewer than 10 deaths, seems like a large design flaw. Clearly, trials that did not collect mortality data would be excluded. However,in the methods of design, it was stated that of 36 identified trials to consider studying, 12 of them were excluded as they reported fewer than 10 deaths. I question the bias this creates as 30% of the available data on mortality, or lack there of, was systematically exluded. I would have to look at this study as designed and conducted with a preferred outcome. In addition, I agree, the form of Vit E tested would be helpful information.

Conflict of Interest:

Send rapid response to journal:
Re: Confounders falsify conclusion

I read with great interest the very useful paper by Miller et. al. as well as the rapid responses. I checked the type of vitamin E used in most of the studies reviewed by use of the fine review; Antioxidants and atherosclerosis: don't throw out the baby with the bath water. Jialal I, Devaraj S. PMID: 12600900 and by checking the information available on Medline for others. The use of synthetic E was not a confounder except for to possibly extremely small extent.

Beta-carotene was.

Adjustment for the concurrent use of beta-carotene with E in smokers eliminates the excess mortality entirely even after also adjusting for the positive effect of fish oil in the GISSI study.

Analysis of the relationship between mortality and the ratio of vitamin C to E given in the studies shows a strong significant trend with no effect or a small detriment possible at a ratio of less than unity and culminating in the significant 47% reduction of mortality seen at the doses of 440 for E and one gram for C in the PPS study. This relationship is consistent with in vitro and epidemiologic literature and the very corroborative HATS paper which showed a near complete stop to progression of plaque in coronary arteries with supplementation of C and E at a ration of over two.

By the way the mortality values listed for the CHAOS study in the text figure 2 seem to be incorrect. The original CHAOS study gives 36 and 27 for E and placebo.

Sinc.,

Thomas Carter

Conflict of Interest:

Send rapid response to journal:
Re: The benefit and harm of vitamin E supplementation

In their important meta-analysis focusing on the potential harmful effects of vitamin E supplementation, Miller et al. (1) assumed that there may be a precise threshold level, so that larger intakes of the vitamin would progressively increase the risk of harm. However, it is possible that there is biological heterogeneity between population groups, so that the characteristics of persons would determine whether vitamin E supplementation causes net benefit or harm.

In the ATBC Study, the effect of vitamin E on the risk of pneumonia was significantly modified by the age of smoking initiation (P = 0.0007)(2). Vitamin E increased pneumonia risk in those who initiated smoking at 20 years or earlier (RR = 1.14; 95% CI: 0.98-1.32), whereas the vitamin reduced pneumonia risk in participants that initiated smoking at later ages (RR = 0.65; 95% CI: 0.49-0.86). Furthermore, in the latter subgroup the benefit was greater among those who smoked less or quit smoking during the trial. Thus, less exposure to cigarette smoking was associated with greater benefit of vitamin E.

In the ATBC Study, the vitamin E dose was 50 IU/day (50 mg/day dl- alpha-tocopheryl acetate) which is substantially less than the threshold of 150 IU/day estimated by Miller et al.(1). However, concluding from the subgroup differences described, it seems probable that some population groups suffer ill effects at this low dosage also, but the same low dose seems beneficial to other persons. Thus, assuming that there is biological heterogeneity between population groups, further studies should characterize people that obtain benefit or harm from vitamin E, instead of just estimating a uniform threshold level for harm with a presumption that the single threshold is valid for the entire population.

1. Miller ER, Pastor-Barriuso R, Dalal D, Riemersma RA, Appel LJ, Guallar E. Meta-analysis: high-dosage vitamin E supplementation may increase all-cause mortality. Ann Intern Med 2005;142(1):in press

2. Hemilä H, Virtamo J, Albanes D, Kaprio J. Vitamin E and beta- carotene supplementation and hospital-treated pneumonia incidence in male smokers. Chest 2004;125:557-65.

Conflict of Interest:

Send rapid response to journal:
Re: Potential bias in methods may negate statistical significance of the high dose vitamin E analysis

19 Nov 04

Editor, Annals Internal Medicine

Sir or Ma’am:

We read with interest the recent article by Millar and colleagues (1). However, we question their conclusions regarding the high dose vitamin E.

First, it is not clear why the authors chose hierarchal logistic regression rather than traditional meta-analytic approaches. We reanalyzed their data from the 11 high dose trials in Figure 2 using two standard methods (Wolfe inverse variance and Mantel Haenszel). Both yielded the same point estimates as the authors, but with non-significant results (RR: 1.04, 95% CI: 0.99-1.10). Secondly, we are concerned that these results are heterogeneous. Both are borderline statistically heterogeneous (Q=16.1, d.f. = 10, p = 0.097), and there is visual evidence of heterogeneity on our Galbraith plots. Better examination of the sources of this heterogeneity is needed.

In addition, we found a suggestion of publication bias among the 11 high dose trials with Begg’s test (p=0.073) (2), which we confirmed with the trim and fill method (3). As the authors note, studies showing benefit from vitamin E are unlikely to be missing from the literature. However, small studies that demonstrate no effect could well be unpublished. It is also possible that the authors did not include these small trials because they did not search EMBASE, which may have excluded European trials. They also excluded trials with less than 10 deaths (which seems arbitrary) that, in particular, would tend to bias the results towards a finding of harm.

The authors searched for the influence of each trial and determined that “none seemed to be driving the results”. In our reanalysis, exclusion of the largest trial (MRC/BHF HPS) resulted in a wider confidence interval (95% CI: 0.96-1.13). We would like to know if the results became non-significant when the authors excluded this or other trials.

Finally, the authors did not account for study quality as a possible explanation of the results. This has been previously shown to effect the results of randomized controlled trials (4) and is recommended by the QUOROM statement (5).

In summary, we have trouble accepting the conclusions regarding high dose vitamin E due to the statistical methods used and the lack of controlling for study quality and publication or selection bias. We contend that correction of any one of these factors could negate the marginally significant results. With these problems and multiple other studies suggesting vitamin E has no effect of mortality, telling our patients that it may be harmful seems premature.

Disclaimer: The opinions or assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting the views of the Department of the Army or the Department of Defense.

References:

1. Millar ER, Pastor-Barriuso R, Dalal D, Riemersma RA, Appel LJ, Guallar E. Meta-Analysis: High-Dosage Vitamin E Supplementation May Increase All- Cause Mortality. Ann Intern Med. 2004:142.

2. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994 Dec;50(4):1088 101

3. Duval S, Tweedie R. Trim and fill: A simple funnel-plot-based method of testing and adjusting for publication bias in meta-analysis. Biometrics. 2000 Jun;56(2):455-63

4. Schulz KF, Chalmers I, Hayes RJ, Altman DG. Empirical evidence of bias. Dimensions of methodological quality associated with estimates of treatment effects in controlled trials. JAMA. 1995 Feb 1;273(5):408-12

5. Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF. Improving the quality of reports of meta-analysis of randomized controlled trials: the QUOROM statement. Lancet 1999 Nov 27;354(9193):1896-900.

Conflict of Interest:

Send rapid response to journal:
Re: VITAMIN E: Which isoform and how much?

There are several flaws in the meta-analysis by Miller et al. (1) that include erroneous interpretation of the pooled trials of alpha- tocopherol and lack of clarity in vitamin E nomenclature. The analysis includes trials from many time periods, with different trial designs, doses and combinations and end-points that make comparisons difficult and fallacious. Participants in several of the trials had significant medical conditions like coronary artery disease, end-stage renal disease, diabetes mellitus, Parkinson’s disease and Alzheimer’s disease. Given this heterogeneity in the participant pool, we would consider it presumptuous to draw solid conclusions of this magnitude even with complex statistical tools like meta-analysis and extend the observations to normal healthy people. In addition to clinical heterogeneity, this analysis also suffered from heterogeneity in test nutrients. In many of the trials vitamin E was used alone and in combination with another nutrient such as beta-carotene. In these cases, the authors combined the data for vitamin E alone with the data for vitamin E plus another nutrient even when the data for the other nutrient indicated that it was statistically associated with increased mortality. Moreover, those studies in which less than 10 deaths occurred were excluded from the meta-analyses giving an artificial weight to those studies in which a higher mortality occurred, i.e., those in which the subjects had serious illness as opposed to studies with healthy individuals. A close-look at the odds-ratios depicted in figure 2 does not suggest harmful effects at a dose of 400 IU of alpha-tocopherol and yet the authors conclude that harmful effects begin at a dose of 150 IU. Vitamin E is not a single compound and exists in eight different isoforms in nature (four tocopherols and four tocotrienols) and have distinct biopotencies, biokinetics and cancer preventive properties. Food sources vary in the content of the vitamin E isoforms. Gamma-tocopherol, the primary source of dietary vitamin E, is abundant in plant seeds (corn, soybean, sesame), vegetable oils and nuts (walnuts, pecans and peanuts).. It is not appropriate to “lump” all the different forms of vitamin E into a single basket and call them “vitamin E”. Natural vitamin E forms differ in their properties from synthetic vitamin E. Most of the trials cited in this analysis used synthetic alpha-tocopherol. This should have been emphasized. Dietary and supplemental sources of vitamin E isoforms possess unique properties that can influence critical pathways involved in cancer, inflammation, cardiovascular disease and neurodegenerative disease. For example, mechanistic differences between the alpha- and gamma-tocopherols and their metabolites provide a molecular basis for the superiority of gamma-tocopherol (2)(3)(4)(5). Although alpha-tocopherol has a high concentration in supplements, the primary form of vitamin E in the diet is gamma-tocopherol, which is present at 2-4 times higher concentration than alpha-tocopherol. A high intake of synthetic alpha- tocopherol can lower plasma and tissue levels of gamma-tocopherol. We believe that carefully conducted randomized studies, with long follow-up periods and well defined endpoints are required to address the potential clinical efficacy of the different isoforms of vitamin E. We also have to be clear on the terminology that is used when we discuss the properties and effects of the different forms of vitamin E.

Reference List

1. Miller ER, et al. Meta-analysis: High-dosage vitamin E supplementation may increase all-cause mortality. Ann Intern Med 2005; 142 (1): 1-11 2. Wagner KH, Kamal-Eldin A, Elmadfa I. Gamma-tocopherol-an underestimated vitamin? Ann Nutr Metab 2004; 48(3):169-188 3. Hensley K, Benaksas EJ, Bolli R, Comp P, Grammas P, Hamdheydari L, Mou S, Pye QN, Stoddard MF, Wallis G, Williamson KS, West M, Wechter WJ, Floyd RA. New perspectives on vitamin E: gamma-tocopherol and carboxyelthylhydroxychroman metabolites in biology and medicine. Free Radic Biol Med. 2004 Jan 1;36(1):1-15 4. Stone, W.L, Krishnan, K, Campbell, S.E., Qui, M, Whaley, S.G, Yang, H: Tocopherols and the Treatment of Cancer. Ann. N.Y. Acad. Sci. 2004;1031: 1-11 5. Campbell S, Stone W.L, Whaley S, Krishnan K. Development of gamma- tocopherol as a colorectal cancer chemopreventive. Critical Reviews in Oncology/Hematology, 2003; 47: 249-259

Conflict of Interest:

Send rapid response to journal:
Re: Dose-Response Analysis Reconsidered

Given the results of the dose-response analysis summarized in their Figure 3, the authors observe that the 95% confidence band for the risk difference does not include 0 for dosages greater than 900 IU/day, and go on (in their Table 2) to derive pooled risk differences and ratios therefrom that suggest a statistically significant relationship between vitamin E supplementation dosage and all-cause mortality.

The authors are to be commended for listing the data they used, thereby facilitating the exploration of alternative analyses: the ability to do so is especially important here because the statistical significance of the observed risk differences is not overwhelming.

If alternative analyses are just about as appropriate for the data at hand, and are generally as reasonable as those that the authors have adopted, then one might well expect that they should produce conclusions that are consistent with the authors'. This, unfortunately, is not the case for the dose-response analysis, as the following steps illustrate.

First, instead of focussing on risk difference, base the dose-response analysis on relative risk (measured by the logarithm of the odds ratio for all-cause mortality: for example, log((100*258)/(106*261))=-0.0698 for MIN.VIT.AOX in Figure 2), which rests on ample precedent and solid rationale [1].

Second, instead of the authors' spline, employ a non-parametric, locally quadratic, weighted regression [2] to model the relationship between relative risk and the logarithm of dosage.

In these circumstances, one finds that the 95% confidence band for this relationship straddles the value 0 (which corresponds to no difference in relative risk) throughout the whole range of dosages represented in the data: thus indicating that, from this viewpoint, and despite a suggestive trend, there is yet insufficient evidence incontrovertibly to establish a statistically significant difference in relative risk between the different groups that received vitamin E supplementation in different dosages.

Details of the computations aforementioned, performed using the data in the authors' Figure 2 (and also in Figure 4, with the same conclusion), using the "locfit" package for R [3], and the alternative to Figure 3 that summarizes the local regression analysis, will be produced upon request since this letter cannot accommodate them.

REFERENCES

[1] Breslow, NE, Day, NE. Statistical Methods in Cancer Research, Volume 1: The Analysis of Case-Control Studies. Lyon, France: International Agency for Research on Cancer; 1980.

[2] Loader, C. Local Regression and Likelihood. New York, NY: Springer-Verlag; 1999.

[3] R Development Core Team. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing (http://www.R-project.org); 2004.

Conflict of Interest:

Send rapid response to journal:
Re: Vitamin E toxicity ?

It has long been known that reducing agents/antioxidants such as vitamins C and E may have proxidant properties, e.g., by interacting with transition-series metals in Fenton's reactions to produce active oxygen species. Whether a reducing agent is a pro-or antioxidant depends on a complex state of circumstances such as its concnetration, that of oxygen, and so forth.

There has been at least one previous circumstance that may shed light on the possible mechanism of vitamin-E toxicity. We have found (1) that vitamin-E administered before adrimycin ameliorated intestinal toxicity of this drugs, but that continued vitamin-E treatment abolished this protective effect.

Two possible explainations of this odd finding are that "loading" a tissue in its proper tissue compartment with vitamin-E protects that tissue, but that circulating vitamin-E interacts with Adriamycin to make it more toxic (e.g., by donating electrons for redox cycling). Another possibility is some sort of hormesis. That is, vitamin-E exerts a mild oxidative stress that causes induction of further defenses against redox drugs like adriamycin.

Peter H Proctor, PhD, MD

1)McGinness JE, Grossie B Jr, Proctor PH, Benjamin RS, Gulati OP, Hokanson JA., Effect of dose schedule of vitamin E and hydroxethylruticide on intestinal toxicity induced by adriamycin. Physiol Chem Phys Med NMR. 1986;18(1):17-24. Pubmed: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=3774891

Conflict of Interest:

Send rapid response to journal:
Re: Re: Natural or Synthetic?

I have just read an article by Aileen Burford-Mason PhD, to "E" or not to "E" which appeared on the web site CTV.ca. She makes a strong arguement that because the studies did not specify whether natural or synthetic vitamin E was used the conclusions are not valid. Most people take the natural form of the vitamin but it seems that the synthetic form may have been used in the studies. It has long been known that the natural form is the preferred form. Separate trials should be done one using synthetic E and the other natural E. Only then will we be able to conclude harm or benefit and which form of the vitamin provides which.

Conflict of Interest:

Send rapid response to journal:
Re: Full Disclosure regarding which form of Vitamin E

Send rapid response to journal:
Re: Natural or Synthetic E

I am not a researcher or medical practitioner, but as an interested party and vitamin E user, I was not able to detect from the research results discussion whether it made a difference if one used natural or synthetic vitamin E. Was there a difference in the research results?

Thank You for your response.

Conflict of Interest:

Send rapid response to journal:
Re: Vitamin E study not convincing

The Meta-Analysis: High-Dosage Vitamin E Supplementation May Increase All-Cause Mortality, as published in ANNALS, does reveal a very slight increase in mortality at increasingly higher doses of supplemental vitamin E, which often points to a causal relationship. However, the slightly- increased mortality could also be explained by the fact that patients with cardiovascular disease take higher doses of vitamin E. A recent study, the Vitamins and Lifestyle (VITAL) study, conducted among 45,748 participants ranging in age from 50 to 75 years, revealed that persons with cardiovascular disease are more likely to be taking vitamin E supplements, presumably at the higher doses (>400 IU) attributed to the elevated mortality. (1)

Bill Sardi, Knowledge of Health, Inc., San Dimas, California.

Send rapid response to journal:
Re: this article

With interest I read the online publication in the Annals of Internal Medicine concerning your meta-analysis of the dose-response relationship between vitamin E supplementation and total mortality. You describe the pooled risk difference in the high-dosage vitamin E trials with 39 per 10.000 persons, referring to figure 2. But when I add the numbers of deaths and participants in the high-dosage vitamin E and the control groups I get a risk difference of 48 per 10.000 and I can’t explain this difference in results. May I therefore ask you to explain how you got the number of 39 per 10.000 persons.

Yours sincerely

S. SCHENK (Physician) Editorial staff arznei-telegramm

A.T.I. Arzneimittelinformation Berlin GmbH Bergstr. 38A, Wasserturm, D-12169 Berlin Fax: (0 30) 79 49 02 20 Internet: http://www.arznei-telegramm.de E-Mail: ati@berlin.snafu.de

Conflict of Interest:

Send rapid response to journal:
Re: Gamma (not alpha) Tocopherol should be studied

To the editor:

There is a tragic flaw in Dr. Miller's meta-analysis on Vitamin E.(1) In all of the studies he referenced, Vitamin E was used alone in the "alpha-tocopherol" form. Though alpha-tocopherol may be the most common supplement form of Vitamin E, the gamma form is the one that is found in greatest concentration in the U.S. diet.(2) The "gamma" (but not alpha) form of Vitamin E is the one that protect against cancer and blocks the COX-2 enzyme that is involved in heart disease.(3,4) I prescribe Vitamin E in the "mixed tocopherols" form for these reasons. We will not know the true effects of Vitamin E unless studies are conducted properly. Until then, we should be cautious against perptuating the bias against natural medicine.(5)

1. Miller ER, et al. Meta-Analysis: High-Dosage Vitamin E Supplementation May Increase All-Cause Mortality. Ann Intern Med. 2005;142(1).

2. Jiang Q, et al. Gamma-tocopherol, the major form of vitamin E in the US diet, deserves more attention. Am J Clin Nutr 2001;74(6): 714-22.

3. Helzlsouer KJ, et al. Association between alpha-tocopherol, gamma-tocopherol, selenium, and subsequent prostate cancer. J Natl Cancer Inst. 2000;92(24): 2018-23.

4. Jiang Q, et al. Gamma-tocopherol and its major metabolite, in contrast to alpha-tocopherol, inhibit cyclooxygenase activity in macrophages and epithelial cells. Proc Natl Acad Sci U S A. 2000;97(21): 11494-9.

5. Goodwin JS. Battling quackery: attitudes about micronutrient supplements in American academic medicine. Arch Intern Med. 1998;158(20): 2187-91.

Conflict of Interest:

Send rapid response to journal:
Re: Vitamin E Warning Premature

Is Vitamin E dangerous in high doses?

In some circumstances, it may be. But these are limited conditions and there are certainly some easy ways to get around them. Let me explain.

I am writing this in response to an article published this week in Annals of Internal Medicine. Researchers at Johns Hopkins have subjected some Vitamin E studies to a meta-analysis, which is where previously published studies are re-examined and compared in order to try to mine more information out of them. These meta-analyses are often highly speculative, because the protocol for each study is different enough that it becomes a stretch to link the results and imply a common thread runs through them despite the different variables. For example, the supplement tested may be given in different forms or doses, the patient group may have significant differences, the length of time taking the supplement may vary considerably or may even include previous users and new users, etc. There are also differences between natural and synthetic Vitamin E, with most studies using only the synthetic forms that are composed of different -shaped molecules and only half as effective as natural Vitamin E. Natural Vitamin E is called d-alpha tocopherol and synthetic Vitamin E is called dl-alpha tocopherol.

Official U.S. dietary guidelines set an upper tolerable intake limit of up to 1,500 IU per day, based on the scientific record.

While these meta-analyses are academically useful to point to potential new problems or solutions, by no means are they definitive proof of anything, due to the lack of uniform protocols and patient groups. But that won’t stop the medical lobby from trying to use these results to limit potencies of vitamins to everyone “for our own good”. The doctors and researchers may be well-meaning, but results of this kind of preliminary study can be publicized and take on a life of their own, with a new role as “proof” of the dangers of taking vitamins.

One thing that many of the patients in this analysis of previous studies had in common was being elderly and sick. So the first caution issued by the report’s authors is that their speculative findings would not even apply to healthy or younger people. That’s one BIG argument against using this analysis to set general restrictions on Vitamin E dosages. We already have some evidence that taking one antioxidant, rather than antioxidant formulas or multiple vitamins, may increase the cancer risks for aged smokers. This may be because antioxidants need to recharge and support each other or else some can actually transform into pro- oxidants that can increase the damage.

I always caution against taking mega-doses of one nutrient without considering potential side effects. Taking only one antioxidant may seem to deplete others because of the way they interact, with one antioxidant chemically supporting others. A surplus of one nutrient may increase a person's need for one or more other nutrients, creating an apparent functional deficiency.

There is a recent example that illustrates my point. Some years ago an antioxidant study in Finland was halted early, with a widely reported increase in cancer rates among male smokers taking beta-carotene. Headlines associated vitamins with cancer risk. Despite objections that the study was flawed, vitamin use dropped.

Fast-forward to this year. A new analysis published in July takes another look at that same Finnish smokers' study, but now takes into account their total antioxidant intake, and clears up that whole controversy. Their risk of getting lung cancer was closely associated with total antioxidants in the diet, with more antioxidants meaning less cancer.

A composite antioxidant index was generated for each of the 27,000 men over 14 years. The calculated amounts of carotenoids, flavonoids, Vitamin E, selenium and Vitamin C were compared to actual lung cancer rates, with a clear result: a combination of antioxidants lowers lung cancer risk in male smokers.

What does this all mean? I think we are in for another round of attacks on vitamins based on this crude analysis of Vitamin E, with some medical experts calling for restrictions on vitamin potency. That would be a mistake, both scientifically and from a public health viewpoint. The message should be that people should not try to take a high dose of one supplement without considering that it may increase our need for other nutrients. Elderly, sick people need a more holistic approach rather than using a single nutrient in high doses, as if it were a drug. Nutrients just don’t work well in isolation from each other. Vitamins are essential to health and life, but the average American gets only 1/3 of the recommended daily intake of Vitamin E that would prevent serious illnesses. Most people would benefit from taking a multiple vitamin and a Vitamin E supplement, and it would be safer than just the Vitamin E alone.

That shouldn’t be hard for the smart folks at Johns Hopkins to understand. But the media reporting this analysis got the facts wrong.

By Neil E. Levin, CCN, DANLA

REFERENCES:

1. July 2004 American Journal of Epidemiology

2. Poster session abstract: High-dose vitamin E supplementation may increase all-cause mortality, a dose response meta-analysis of randomized trials; Ernest N. Morial Convention Center, Exhibit Hall 1-2.) Annals of Internal Medicine, Nov. 10, 2004

Conflict of Interest:

Send rapid response to journal:
Re: Natural or Synthetic?

It has been shown that synthetic Vitamin A can be quite harmful whereas natural A is very beneficial. The same is likely true with Vitamin E. Before you aggregate results across studies you must know what what types people take. Most often they themselves do not know!

Conflict of Interest:

Send rapid response to journal:
Re: Type of vitamin E

Please report which type of vitamin E was used in the trial by Miller, et al. Whether or not d- or dl- tocopherols,beta, gamma, mixed, etc. were used could potentially be very important,indeed, make the difference in your findings altogether. I am concerned that this may be a very important study, but highly confounded if synthetic and/or various types of vitamin E were used.

Thank you Cynthia Gran

Conflict of Interest: