Below you can read a scientific explanation about why you should not be too concerned about taking Vit E. Be sure to take a complete antioxidant along with the Vit E that includes mixed tocopherols.

 

 

Commentary on Recent Meta-Analysis of High-Dosage Vitamin E Effects

 

by Anthony L. Rosner, Ph.D. F.C.E.R.

December 15, 2004

Synopsis

A recent PubMed literature search complemented by a search of the Cochrane Clinical Trials Database has yielded a total of 19 clinical trials involving 135,967 participants taking vitamin E at doses from 16.5 to 2000 IU/d. By selecting a cutoff value of 400 IU/d, the authors concluded in a statistical pooling of results [meta-analysis] that [a] for doses <400 IU/d, the pooled risk difference of mortality with 8 trials was -16 per 10,000 persons [CI -41 to 10 per 10,000 persons, risk ratio = 0.98] and [b] for doses >400 IU/d in the remaining 11 trials, the pooled risk difference was 39 per 10,000 persons [CI 3 to 74 per 10,000 persons, risk ratio = 1.04]. In a dose-response analysis, however, the risk difference appeared to increase as the vitamin E dosage rose over 150 IU/d. The conclusion was that high-dosage vitamin E supplements could increase overall mortality and should thus be avoided.1

Critique:

The authors' assertion that a true prospective dose-response trial of vitamin E to assess its possible effects on mortality is not feasible is clearly well-taken; instead, a retrospective statistical analysis such as the one reported here would be appropriate. However, there are numerous flaws and caveats-several by the authors' own admission--which appear in the text such that it must be interpreted with extreme caution:

1. Biological heterogeneity: The most obvious is the fact that chronically ill rather than healthy subjects have been chosen. Not only is generalization of these results to asymptomatic subjects problematical under these circumstances, but also to sick patients as well since the report does not distinguish types of ailments. In broad strokes, the fact that most of the seriously ill patients took high doses of vitamin E should make one think twice as to whether their increased mortality truly could be laid specifically at the feet of vitamin E. What about interactions with other potent prescription drugs which most of these patients presumably have taken? Furthermore, as cited by the authors of this study themselves, Alzheimer's patients may benefit from vitamin E supplementation because it has been suspected to delay the progression of this disease.2,3

2. Nature of the supplement: Eight forms of vitamin E [alpha-tocopherol, gamma-tocopherol, beta-tocopherol, delta-tocopherol, alpha-tocotrienol, beta-tocotrienol, gamma-tocotrienol, and delta-tocotrienol] have been reported to occur naturally.4 In all 19 trials combined in this report, however, there is no indication as to which species of vitamin E has been administered. Presumably the predominating fraction in the meta-analysis was alpha-tocopherol, high enough doses of which have been shown to deplete supplies of gamma-tocopherol.4 But until the vitamin E chemistry has been defined in these trials and until there is a clear indication as to whether natural or synthetic forms were used, speculation as to the toxicity of "high doses" of vitamin E may be premature.

3. Arbitrary cutoff values: To the casual observer, the risk difference of 1.04 with a confidence interval that very nearly reaches zero used to suggest that mortality levels of vitamin E doses above 400 IU/d seems to barely exceed neutrality. Worse, two studies at 330 IU/d, including the large and highly visible GISSI study with over 6100 patients, found a decrease in total mortality; had these two studies been included in the "high dose" category [arbitrarily set at 400 IU/d], they would have drawn this very tiny effect even closer to null. Finally, another large study [HOPE] involving over 5000 patients at a 400 IU/d dosage level displayed absolutely no disposition toward mortality while yet another investigation at 440 IU/d revealed a statistically significant decrease in total mortality in the vitamin E group. This leaves 8 studies with dosages of 500 IU or more driving the slightly increased risk of total mortality with 6 showing confidence intervals which straddle the line of demarcation between increased and decreased risk of mortality.5 From all these indications, the authors would appear to have overstated their case, and not once has the dose-response analysis suggesting problems over 150 IU/d been brought up in the discussion to suggest toxicity at that level. In short, the minuscule presumed elevation of mortality shown in this meta-analysis appears to be primarily the result of statistical maneuvering--pooling the results of 8 trials rather than any outstanding result yielded in any of the individual investigations, themselves.

4. Failure to report benefits: The presumed small harmful effect reported seems even more spurious in light of the fact that the benefits of vitamin E in the investigations pooled have been given virtually no mention. Other than the aforementioned beneficial effect for Alzheimer's patients, reductions in the risks of heart and blood vessel disease, age-related macular degeneration, and several forms of cancer5-7 need to be considered if a true risk-benefit assessment of vitamin E is to be made.

5. The data is at variance with other reports and needs to be reconciled with them: For example, the Institute of Medicine has concluded that vitamin E is safe at levels as high as 1000 IU/d for the synthetic and 1500 IU/day for the natural form.8

6. Potential bias in methods: The question has been raised why the authors chose a hierarchal logistic regression rather than traditional meta-analytic approaches. When data from the high dose trials is reanalyzed using two standard methods [Wolfe inverse variance and Mantel Haenszel], nonsignificant results are obtained. The heterogeneity of the results pooled has also come under fire. Furthermore, a statistical test has suggested a publication bias among the high dose trials sampled, such that [a] small studies showing no effect could be unpublished, [b] trials with less than 10 deaths have been excluded which would bias the results toward harm, and [c] European trials may have been excluded since the authors did not search EMBASE. Finally, there was no accounting for the quality of the different studies pooled.9 Under all these circumstances, the playing field has been clearly and spuriously tilted against the benefits of this nutritional supplement.

7. Variations in trials combined: Other variations in the trials combined in this meta-analysis which were not mentioned previously include the length of time the supplement was taken or whether or not it was ever taken previously.

Until a study which explicitly defines the different isoforms of vitamin E with clearly defined populations is undertaken, it would be premature to cast a cloud of suspicion upon doses of vitamin E at 400 IU/d or higher. It does seem odd that the usual over-the-counter marketing value of 400 IU/d has been pounced upon by this study with no substantial data to indicate any possible problems until levels of 600 IU/d or more are reached. At this stage, one cannot accept with any reasonable degree of confidence the results presented in Miller's study.

REFERENCES

1Miller ER, Barrius RP, Dalal D, Rimersma RA, Appel LJ, Guallar E. Meta-analysis: High-dosage vitamin E supplementation may increase all-cause mortality. Annals of Internal Medicine 2005; 142(1): 1-11.

2Rabins P, Blacker D, Bland W, Bright-Long L, Cohen E, Katz I et al. Practice guideline for the treatmment of patinets with Alzheimer's Disease and other dementias of late life. Arlington, VA: American Psychiatric Association, 1997. Accessed at www.psych/org/psych_pract/treatg/pg_dementia_32701.cfm? on 1 October 2004.

3Doody RS, Stevens JC, Beck C, Dubinsky RM, Kaye JA, Gwyther L et al. Practice parameter: Management of dementia [an evidence-based review]. Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 2001; 56: 1154-1166.

4Leong W. Not all vitamin Es should be condemned: A different perspective. www.annals.org/cgi/eletters/0000605-200501040-00110v, 13 December 2004.

5Dickinson A. Arbitrary definition of high dose. www.annals.org/cgi/eletters/0000605-200501040-00110v, 8 December 2004.

6Jialal I, Devaraj S. Antioxidants and atherosclerosis: Don't throw the baby with the bath water. Circulation 2003; 107(7): 926-928.

7Jialal I, Devaraj S. Scientific evidence to support a vitamin E and heart disease health claim: Research needs. Journal of Nutrition 2005; In press.

8Taub EA. East an apple a day...and take your vitamins! www.annals.org/cgi/eletters/0000605-200501040-00110v, 30 November 2004.

9Shimeall W, Douglas K, Jackson JL. Potential bias in methods may negate statistical significance of the high dose vitamin E analysis. www.annals.org/cgi/eletters/0000605-200501040-00110v, 19 November 2004.

 

Home Page