点击显示 收起
Department of Pathology (MC 847)
College of Medicine
University of Illinois at Chicago
840 South Wood Street, Room 130
Chicago, IL 60612
E-mail: mewright{at}uic.edu
Nutritional Epidemiology Branch
Division of Cancer Epidemiology and Genetics
National Cancer Institute
National Institutes of Health
Bethesda, MD
Dear Sir:
We appreciate the earlier editorial by Traber (1) and the current comments from Hemilä and Miller. In our study, we found that higher prerandomization serum concentrations of -tocopherol were associated with significantly lower total and cause-specific mortality in men participating in the Alpha-Tocopherol, Beta-Carotene Cancer Prevention (ATBC) Study (2). Only 10% of participants reported vitamin E supplement use before randomization, and the exclusion of these men from our analyses did not alter the observed relations. This indicates that pretrial serum vitamin E concentrations in the ATBC Study population were achieved primarily through dietary intakes and other host factors known to affect circulating vitamin E concentrations (eg, age, body mass index, and serum cholesterol) and not through vitamin E supplement use. It is important to note that neither the use of supplemental vitamin E before the trial nor the trial intervention itself (50 mg all-rac--tocopheryl acetate) was the focus of our report.
As Traber (1) pointed out in her editorial, we observed the lowest overall mortality at serum -tocopherol concentrations of 13 mg/L (14 mg/L for cardiovascular disease mortality; see Figure 2 in reference 2). It should be emphasized that mortality did not diminish further at higher concentrations: relative mortality estimates drifted back toward unity (relative risk = 1) as blood concentrations rose beyond 13–14 mg/L. The precise vitamin E intake required to achieve this "optimum" serum concentration cannot be inferred from our study, however. Even though men in the fourth quintile of serum vitamin E (ie, 12.2–13.5 mg/L) consumed an average of 13.3 mg -tocopherol/d (see Table 1 in reference 2), that mean value reflected a wide range of intakes (5.7–29.3 mg/d) within the specific serum quintile. This finding highlights the multifactorial determinants of serum -tocopherol concentrations, including dietary intake, absorption, lipoprotein concentrations, blood transport, tissue uptake, oxidative stress load, and the genotypic variants that likely affect these specific contributory phenotypes. Carefully controlled feeding studies can help shed light on the amounts of vitamin E that need to be ingested to achieve particular blood concentrations. In this regard, however, studies have made clear that a range of serum concentrations can result from any single daily dietary intake and, conversely, that a range of intakes can lead to a single target blood or tissue concentration. Finally, it should be reemphasized that any "optimal" serum -tocopherol value that we observed with respect to overall mortality among Finnish male smokers may not be applicable in other groups, including nonsmokers, women, and ethnically diverse populations. This question should be addressed in other studies.
Traber correctly highlights the possibility that dietary recommendations based on preventing overt deficiency symptoms—peroxide-dependent erythrocyte hemolysis, in the case of vitamin E—may differ from recommendations based on the prevention of chronic disease or death. As she notes, the Recommended Dietary Allowance (RDA) for vitamin E is 15 mg -tocopherol/d for men and women >18 y old, and this amount is based on experiments conducted almost a half-century ago in men who were experimentally vitamin E depleted (3). Overt vitamin E deficiency is extremely rare in the United States, despite the fact that most US men and women are not meeting the dietary recommendation for vitamin E. Again, additional research aimed at clarifying the optimal serum concentrations of vitamin E for chronic disease prevention in multiple populations, as well as the amount of dietary vitamin E required to achieve those concentrations, will be informative. As more data accumulate, the RDAs for vitamin E may need to be reevaluated with respect to important public health endpoints such as chronic disease risk and mortality and not only in relation to the avoidance of deficiency states.
We agree with Hemilä and Miller that populationwide vitamin E supplementation is not warranted at this time, according to the available research. This body of evidence includes both a demonstrated lack of efficacy for overall mortality in several supplementation trials and the elevated mortality suggested—but not universally accepted (4)—by a recent meta-analysis for high-dose vitamin E supplementation (5). We explicitly state in our report, "Because supplemental vitamin E has not been shown to reduce mortality in randomized trials, efforts to improve vitamin E status through dietary means (eg, through increasing consumption of foods rich in vitamin E, including nuts, seeds, whole grains, and dark-green leafy vegetables) may be warranted, particularly if future prospective studies show similar serum mortality associations in diverse populations, including nonsmokers" (2). Although, as Traber suggests, vitamin E–rich food sources have traditionally been of limited popularity in the American diet, we support dietary modification rather than supplementation at this time.
Results from well-designed prospective cohort studies have made, and will continue to make, substantial contributions to our knowledge regarding micronutrient-disease relations, even when the research findings appear to contradict those from controlled trials. A case in point is the diametric opposition of the conclusions of the original ATBC Study findings for ß-carotene and lung cancer (6) to the findings from most case-control and cohort studies available at the time (7). We must remain cognizant of the fact that observational studies and clinical trials often address different questions. For example, trials typically test the efficacy of single-nutrient supplements, at various dosages and administered over several years, whereas observational studies examine the associations between habitual dietary intake (or serum concentrations) of nutrients that are derived primarily from foods, which contain many other, potentially anticarcinogenic substances. Whereas it is true that observational studies are susceptible to confounding and measurement error and that trials are typically free from such biases, we believe that recommendations regarding supplement use should be based on the totality of evidence provided by basic experimental and epidemiologic studies, as well as by randomized controlled trials.
ACKNOWLEDGMENTS
None of the authors had any personal or financial conflict of interest.
REFERENCES