Correspondence
Antioxidant Vitamins, Cancer, and Cardiovascular Disease
N Engl J Med 1996; 335:1065-1069October 3, 1996
- Article
To the Editor:
Hennekens et al. and Omenn et al. (May 2 issue)1,2 report further failures to find a positive correlation between dosing with “beta carotene” and a reduced incidence of malignant neoplasms. Since the main stimulus for this and other prospective studies came from the highly suggestive results of retrospective studies of vegetarians and other nearly herbivorous human beings, it was reasonably concluded that some compound or compounds unique to vegetables should be selected for the first cast of the net. How a particular beta carotene came to be selected for worldwide testing is neither hard to understand nor easy to forgive.
Although chemically explicit identification of the specific compound tested was not provided in either report, it was almost certainly solely or predominantly all-trans-β-carotene, the only 1 among 272 configurational isomers, or stereoisomers, of beta carotene that is synthesized commercially in tonnage lots (BASF, Hoffmann–LaRoche, and Sumitomo) and is available by purchase. But for this convenience, it might have been equally justifiable to have selected any of the several mono- and di-cis-β-carotenes identified in the beta carotene fraction of vegetables or 15- or 13-cis-β-carotene, two stereoisomers that are produced3 slowly from all-trans-β-carotene at 37°C and are among those (including 9-cis-β-carotene) that have been identified in human tissue.4
The present findings confirm earlier fears that devoting exclusive attention to a single stereoisomer carried a clear and large risk of missing the active agent (or agents), if it were indeed to be found among the beta carotenes.3 A further, necessarily large-scale, experimental and epidemiologic investigation is now much more problematic. In keeping with the notion of “the better being the enemy of the good,” a study of a mixture of the relatively small number of stereoisomers identified in edible plants promises a more definitive answer. Such apposite mixtures can be obtained by heating all-trans-β-carotene for a brief period, as in the process of cooking; collecting stereoisomers, rich in 9-cis-β-carotene, produced and isolated from cultures of dunaliella; or extracting them from vegetables such as spinach.
The potential benefit of identifying the hypothetical components that lower the incidence of cancer among vegetarians continues to be great, as Greenberg and Sporn suggest in their editorial.5 The medical community would be well advised to respond to the ineffectiveness of studies using one stereoisomer, all-trans-β-carotene, not by abandoning the search but by strongly encouraging the performance of studies using a chemically more rational approach.
5 ReferencesWilliam Doering, Ph.D.
Harvard University, Cambridge, MA 02138-29021
Hennekens CH ,Buring JE ,Manson JE , et al. Lack of effect of long-term supplementation with beta carotene on the incidence of malignant neoplasms and cardiovascular disease. N Engl J Med 1996;334:1145-1149
Full Text | Web of Science | Medline2
Omenn GS ,Goodman GE ,Thornquist MD , et al. Effects of a combination of beta carotene and vitamin A on lung cancer and cardiovascular disease. N Engl J Med 1996;334:1150-1155
Full Text | Web of Science | Medline3
von E . Doering W, Sotiriou-Leventis C, Roth WR. Thermal interconversions among 15-cis-, 13-cis-, and all-trans-β-carotene. J Am Chem Soc 1995;117:2747-2757
CrossRef4
Stahl W ,Schwarz W ,Sundquist AR ,Sies H . cis-trans Isomers of lycopene and β-carotene in human serum and tissues. Arch Biochem Biophys 1992;294:173-177
CrossRef | Web of Science | Medline5
Greenberg ER ,Sporn MB . Antioxidant vitamins, cancer, and cardiovascular disease. N Engl J Med 1996;334:1189-1190
Full Text | Web of Science | Medline
To the Editor:
Beta carotene supplementation failed to reduce the incidence of lung cancer in the Beta-Carotene and Retinol Efficacy Trial (CARET), which used 30 mg of beta carotene per day plus 25,000 IU of retinol, and in the Physicians' Health Study, which used 50 mg of beta carotene every other day. A comparison of these protocols with the previously available evidence indicates possible reasons for the disappointing but different results of the two trials.1,2
Expectations were based largely on the results of dietary studies in which the intake of 2 to 6 mg of beta carotene per day and the corresponding plasma level of 0.5 μmol per liter or more were associated with a lower risk of lung cancer. Although there was no evidence of a pharmacologic effect of beta carotene over and above its physiologic antioxidant and provitamin A activity, the doses were at least 5 to 10 times higher than this dietary intake and achieved the high plasma levels of 3.9 and 2.2 μmol per liter in the two trials, respectively.
In the epidemiologic studies, beta carotene was also acting as a marker of increased fruit and vegetable consumption and thus of all components that have cancer-preventing potential (vitamin C, folic acid, and other carotenoids, antioxidants, vitamins, and so forth). Nevertheless, in the trials beta carotene was administered alone or only with its metabolic derivative, retinol.
Metabolically, nutrients act in combination at physiologic doses, whereas pharmacologic doses of a single nutrient (e.g., amino acids) may provoke a nutritional imbalance. Pharmacologic doses of a single carotenoid, such as beta carotene, can result in competition at the receptor with the other 50 naturally occurring dietary carotenoids and thus inhibition of their protective effects.
This hypothesis may also explain why the incidence of lung cancer was increased in the CARET study but not in the Physicians' Health Study. The blood levels of beta carotene in the Physicians' Health Study were considerably lower than those in the CARET study, and the two-day interval between doses (the absorption window) permitted adequate absorption of dietary carotenoids. If the blood levels of other carotenoids (e.g., lutein) were lower in the treatment group than in the placebo group in the CARET study, this would support the hypothesis.
The available data justify the conclusion that a balanced mixture of antioxidants, as found in fruits and vegetables, provides the best protection against cancer and that excessive intake of a single antioxidant may impair that protection. Population groups with low levels of fruit and vegetable consumption may be able to obtain similar protection by taking moderate amounts of a combination of several antioxidants (e.g., various carotenoids and vitamin C), provided that the consumption and doses of the combination conform to a physiologic pattern.
2 ReferencesKlaus Pietrzik, M.D.
University of Bonn, 53115 Bonn, Germany1
Omenn GS ,Goodman GE ,Thornquist MD , et al. Effects of a combination of beta carotene and vitamin A on lung cancer and cardiovascular disease. N Engl J Med 1996;334:1150-1155
Full Text | Web of Science | Medline2
Hennekens CH ,Buring JE ,Manson JE , et al. Lack of effect of long-term supplementation with beta carotene on the incidence of malignant neoplasms and cardiovascular disease. N Engl J Med 1996;334:1145-1149
Full Text | Web of Science | Medline
To the Editor:
Kushi et al. (May 2 issue)1 report a correlation between vitamin E intake and death from coronary artery disease. However, the unadjusted death rate in the highest quintile of intake was almost identical to that in the lowest quintile (51 and 52 deaths, respectively), suggesting that there are strong confounding variables correlated with vitamin E intake. The benefit of vitamin E is seen when the data are adjusted for age and total energy intake. The average age is similar in each quintile, but the total energy intake for each quintile is not reported. Eight other variables are reported in Table 1 of the article and discussed in the text. It appears that energy intake is a crucial confounding variable that negates any benefit of vitamin E, a point avoided by the authors.
Another unusual finding of the study is that a benefit was seen only from dietary vitamin E, not from supplements. Among other arguments, the authors propose that the lack of an effect with supplements could be due to the small number of subjects in the group taking the highest dose of supplements, if a benefit required higher doses. However, the benefit from dietary vitamin E is seen at a calculated intake of 7.63 to 9.63 U per day, and the group taking the lowest dose of supplements received 1 to 25 U per day. It is illogical to propose that vitamin E is beneficial only in high doses from supplements but is beneficial in low doses from diet.
The authors consider the possibility that a different beneficial factor is correlated with vitamin E. Let me propose another hypothesis. Since vitamin E is found in energy-rich foods and a benefit of eating such foods was seen only when total energy intake was not increased, the benefit may be due to the lower consumption of energy-rich food not containing vitamin E. For example, in two groups of people on diets that have the same number of calories but different ratios of animal fat to vegetable oils, a lower coronary risk may be correlated with a higher intake of vitamin E but may be due to a lower intake of animal fat. In this case, no similar association would be seen with vitamin E taken as a supplement.
I make this hypothesis simply to demonstrate the great difficulty in drawing conclusions from an observational study of a multifactorial process, particularly when the data are based on responses to a single questionnaire. I can assure you that I am not advising my patients to eat more mayonnaise.
1 ReferencesDavid DeGrand, M.D., Ph.D.
52 Richmond Ave., Worcester, MA 016021
Kushi LH ,Folsom AR ,Prineas RJ , et al. Dietary antioxidant vitamins and death from coronary heart disease in postmenopausal women. N Engl J Med 1996;334:1156-1162
Full Text | Web of Science | Medline
To the Editor:
Contrary to the editorial comments of Greenberg and Sporn, we believe that strong mechanistic data are available that would predict a beneficial effect of beta carotene and other provitamin A carotenoids in the prevention of cancer. Although we agree with their conclusion that beta carotene probably does not have a role as an antioxidant in cardiovascular disease, other mechanisms based on the metabolism of provitamin A carotenoids are still important. We do not understand the basis for Greenberg and Sporn's statement that “beta carotene is poorly converted to retinol.” When the National Research Council recommended the consumption of diets rich in fruits and vegetables, it was well aware that almost 50 percent of the public's daily intake of vitamin A came from the provitamin A carotenoids in these fruits and vegetables. Furthermore, we have known for over 25 years that beta carotene can be converted by central cleavage to retinal in vitro,1 and more recently, we have been able to document the conversion of beta carotene to retinoic acid by an independent eccentric cleavage mechanism both in vitro and in an animal model.2 In addition to retinoic acid, eccentric cleavage produces a variety of carbonyl and acid metabolites of beta carotene, but we know very little about their biologic activity. What we do know is that many tissues can carry out the conversion of beta carotene to retinoic acid and that this may be one means of regulating retinoic acid production.
In the Physicians' Health Study and the CARET study, as well as the Alpha-Tocopherol, Beta Carotene (ATBC) Cancer Prevention Study,3 only 9 percent of the subjects were women. From an investigative point of view, this is unfortunate, for we now know that breast tissue is particularly sensitive to the action of retinoic acid, since treatment of normal mammary epithelial cells induces the expression of retinoic acid receptor–β2, which can act as a suppressor of transformation.4 What remains unanswered is whether beta carotene metabolism in breast tissue can result in increased levels of retinoids that may modulate cell proliferation and have an effect on the incidence of breast cancer.
4 ReferencesNorman I. Krinsky, Ph.D.
Tufts University School of Medicine, Boston, MA 02111Monica Peacocke, M.D.
Columbia University College of Physicians and Surgeons, New York, NY 10032Robert M. Russell, M.D.
Department of Agriculture Human Nutrition Research Center on Aging at Tufts University, Boston, MA 021111
Goodman DS ,Olson JA . The conversion of all-Trans β-carotene into retinal. Methods Enzymol 1969;15:462-475
CrossRef2
Wang X-D ,Krinsky NI ,Benotti PN ,Russell RM . Biosynthesis of 9-cis-retinoic acid from 9-cis-β-carotene in human intestinal mucosa in vitro. Arch Biochem Biophys 1994;313:150-155
CrossRef | Web of Science | Medline3
The Alpha-Tocopherol, Beta Carotene Cancer Prevention Study Group
Full Text | Web of Science | Medline4
Lee X ,Si SP ,Tsou HC ,Peacocke M . Cellular aging and transformation suppression: a role for retinoic acid receptor β2. Exp Cell Res 1995;218:296-304
CrossRef | Web of Science | Medline
Author/Editor Response
The authors reply:
To the Editor: Few would disagree with the comments by Drs. Doering and Pietrzik about the likely wisdom of eating a diet rich in fruits and vegetables, which may well reduce the risk of cancer or cardiovascular disease through mechanisms that have little or nothing to do with beta carotene. As regards beta carotene supplementation, however, the evidence now includes data from several completed large-scale randomized trials1-3 and indicates that well-nourished populations accrue no overall benefit over the periods studied to date. This does not, of course, mean that other antioxidant vitamins, such as vitamin E or C, or even very prolonged beta carotene supplementation may not eventually be beneficial in reducing the risk of cancer, especially among poorly nourished persons with low blood levels of beta carotene,4 or in the secondary prevention of cardiovascular disease.5
Whether there is a true hazard associated with beta carotene supplementation among current cigarette smokers, a question raised by the ATBC study3 and the CARET study,2 will be evaluated by a collaborative overview of the postpublication results of continued follow-up in all the completed trials. If the apparent excess risk observed in those trials is in fact real, it should persist with longer follow-up. Conversely, if, as is still possible, there is eventually some benefit from beta carotene supplementation, then it may emerge with longer follow-up. A common protocol for a collaborative analysis of the postpublication results of all the beta carotene trials is therefore being developed, with the first analyses planned for sometime around the year 2000. This analysis will provide the most reliable data on whether the long-term effects are favorable, unfavorable, or (as our trial suggests) null.
5 ReferencesCharles H. Hennekens, M.D.
Julie E. Buring, Sc.D.
Brigham and Women's Hospital, Boston, MA 02115Richard Peto, F.R.S.
University of Oxford, Oxford OX2 6HE, United Kingdom1
Hennekens CH ,Buring JE ,Manson JE , et al. Lack of effect of long-term supplementation with beta carotene on the incidence of malignant neoplasms and cardiovascular disease. N Engl J Med 1996;334:1145-1149
Full Text | Web of Science | Medline2
Omenn GS ,Goodman GE ,Thornquist MD , et al. Effects of a combination of beta carotene and vitamin A on lung cancer and cardiovascular disease. N Engl J Med 1996;334:1150-1155
Full Text | Web of Science | Medline3
The Alpha-Tocopherol, Beta Carotene Cancer Prevention Study Group
Full Text | Web of Science | Medline4
Blot WJ ,Li J-Y ,Taylor PR , et al. Nutrition intervention trials in Linxian, China: supplementation with specific vitamin/mineral combinations, cancer incidence, and disease-specific mortality in the general population. J Natl Cancer Inst 1993;85:1483-1492
CrossRef | Web of Science | Medline5
Manson JE ,Gaziano JM ,Spelsberg A , et al. A secondary prevention trial of antioxidant vitamins and cardiovascular disease in women: rationale, design, and methods. Ann Epidemiol 1995;5:261-269
CrossRef | Medline
Author/Editor Response
Despite disappointing and adverse results in trials with beta carotene, in both Finland1 and the United States, the prevention of common cancers remains a crucial goal, especially for those that respond poorly to treatment, such as lung cancers. Beta carotene and retinoids were by far the leading candidates for prevention in the early 1980s,2 when the portfolio of trials was launched using beta carotene alone (in the Physicians' Health Study), beta carotene with or without alpha-tocopherol (in the ATBC study in Finland), and beta carotene with retinol (in the CARET study) as pharmacologic agents. There are many other potential chemopreventive agents in fruits and vegetables, including folic acid.3 Dr. Doering is correct in stating that agents already approved by the Food and Drug Administration and available for use in humans have an advantage in proposals for initial studies and potential large-scale trials. His proposals regarding inconstant mixtures or extracts would need systematic testing, including an assessment of bioisomerization.4
Dr. Pietrzik refers to “cancer-preventing potential,” then concludes that a balanced mixture of antioxidants “provides the best protection.” Such optimistic views, whether from scientists or from cereal or vitamin-supplement manufacturers, must be subjected to testing for both efficacy and safety. The three big beta carotene trials tested pharmacologic hypotheses, using the most promising agents in the 1980s. In the Finnish trial,1 the serum beta carotene levels were even higher than in the CARET study, despite the lower dose. The main conclusion is that all three trials produced marked elevations in serum beta carotene levels, as intended. Despite an initial claim to the contrary, there is now overwhelming evidence that serum vitamin E levels are not lowered. Gaziano et al. have found a reduction in the level of lycopene but not of other carotenoids4; analyses of several carotenoids, including lutein, during and after treatment in the CARET study are in progress. Although there is extensive molecular-receptor information available for the retinoids, I am unaware of the basis for Dr. Pietrzik's statement about inhibition among carotenoids “at the receptor.” Finally, the absence of an adverse effect on the risk of lung cancer in the Physicians' Health Study is not reassuring, given the low risk status of the participants and the limited numbers of lung cancers.
4 ReferencesGilbert S. Omenn, M.D., Ph.D.
Fred Hutchinson Cancer Research Center, Seattle, WA 981041
The Alpha-Tocopherol, Beta Carotene Cancer Prevention Study Group
Full Text | Web of Science | Medline2
Peto R ,Doll R ,Buckley JD ,Sporn MB . Can dietary beta-carotene materially reduce human cancer rates? Nature 1981;290:201-208
CrossRef | Web of Science | Medline3
Boushey CJ ,Beresford SA ,Omenn GS ,Motulsky AG . A quantitative assessment of plasma homocysteine as a risk factor for vascular disease: probable benefits of increasing folic acid intakes. JAMA 1995;274:1049-1057
CrossRef | Web of Science | Medline4
Gaziano JM ,Johnson EJ ,Russell RM , et al. Discrimination in absorption or transport of beta-carotene isomers after oral supplementation with either all-trans or 9-cis-beta-carotene. Am J Clin Nutr 1995;61:1248-1252
Web of Science | Medline
Author/Editor Response
We agree with Dr. DeGrand that in our study, the association between dietary vitamin E intake and mortality from coronary heart disease is confounded by total energy intake. In our analyses, we included total energy intake to account for this confounding effect. As shown in Table 1Table 1
Relative Risk of Death from Coronary Heart Disease (CHD), According to the Quintile of Vitamin E Intake from Food among Postmenopausal Women Who Did Not Take Supplemental Vitamin E, 1986 –1992., adjustment for age alone did not provide evidence of an influence of vitamin E on mortality from coronary heart disease. The analysis and meaning of total energy intake in nutritional epidemiology are the subject of ongoing debate.2 In most cases, a proper inference about the effect of dietary factors on disease must take total energy intake into account.It is somewhat unusual to observe an association between mortality and vitamin E from foods but not from supplements. The results of our study contrast with those that found an association only for supplemental intake.3,4 We suggested that the higher supplemental doses, along with a longer duration of use, in other studies3,4 may have permitted the detection of an effect of supplemental vitamin E. We did not suggest that our findings negate any role of supplemental vitamin E in the prevention of coronary disease; a recent study provides further evidence in support of such a role.5
Our study has not definitively ruled out the influence of other dietary factors, whether beneficial or detrimental with respect to the risk of coronary heart disease. It is precisely for the reasons presented by Dr. DeGrand that we conducted analyses adjusting for other factors, including meat and other foods and food groups, as we note in our article. However, the results were not altered substantially in these analyses; analyses adjusted for meat intake are also shown in Table 1. This does not eliminate the possibility that other, unmeasured factors may be important. In addition, if vitamin E intake is measured more precisely than other relevant dietary factors, its association with mortality from coronary heart disease will appear to be most important.
It would be inappropriate to recommend that people eat more mayonnaise on the basis of the results of our study. A more prudent message would be that characteristics of a diet high in vitamin E may decrease the risk of death from coronary heart disease. This is an important message in a society that has emphasized decreasing total fat intake, regardless of the source and possible benefit. Since vegetable oils are particularly rich in vitamin E, a reduction in dietary fat should probably be focused on animal food sources.
5 ReferencesLawrence H. Kushi, Sc.D.
Pamela J. Mink, M.P.H.
Aaron R. Folsom, M.D.
University of Minnesota School of Public Health, Minneapolis, MN 55454-10151
Kushi LH ,Folsom AR ,Prineas RJ , et al. Dietary antioxidant vitamins and death from coronary heart disease in postmenopausal women. N Engl J Med 1996;334:1156-1162
Full Text | Web of Science | Medline2
Willett W. Nutritional epidemiology. New York: Oxford University Press, 1990.
3
Stampfer MJ ,Hennekens CH ,Manson JE ,Colditz GA ,Rosner B ,Willett WC . Vitamin E consumption and the risk of coronary disease in women. N Engl J Med 1993;328:1444-1449
Full Text | Web of Science | Medline4
Rimm EB ,Stampfer MJ ,Ascherio A ,Giovannucci E ,Colditz GA ,Willett WC . Vitamin E consumption and the risk of coronary heart disease in men. N Engl J Med 1993;328:1450-1456
Full Text | Web of Science | Medline5
Losonczy KG ,Harris TB ,Havlik RJ . Vitamin E and vitamin C supplement use and risk of all-cause and coronary heart disease mortality in older persons: the Established Populations for Epidemiologic Studies of the Elderly. Am J Clin Nutr 1996;64:190-196
Web of Science | Medline
Author/Editor Response
Drs. Krinsky, Peacocke, and Russell assert that carotenoid supplements may prevent cancer through their conversion to retinol or retinoic acid. Clinical trials have consistently shown, however, that in the absence of a vitamin A deficiency, beta carotene supplementation does not measurably increase blood concentrations of retinol.1,2 There is also no evidence from studies in humans that beta carotene supplementation increases the concentration of retinol or retinoic acid in any tissues, including the breast. In fact, a recent report of a clinical study from Dr. Russell's laboratory indicated that a large oral dose of beta carotene had essentially no effect on concentrations of retinol or retinoic acid in plasma or adipose tissue. The authors concluded that “there may be little tissue metabolism of beta carotene to retinoids” and “the anti-carcinogenic effect of beta carotene, if any, may be due to properties of the molecule itself.”3 Thus, if it is desirable to alter gene expression with retinoids, it is better to use retinoic acid or one of its analogues than beta carotene. Finally, we question the implication that beta carotene shows promise in preventing breast cancer. The results of epidemiologic studies of beta carotene intake and the risk of breast cancer have been less impressive than the results of studies of beta carotene and the risk of lung cancer,4 a condition that is clearly not prevented by beta carotene supplementation.
4 ReferencesE. Robert Greenberg, M.D.
Michael B. Sporn, M.D.
Dartmouth Medical School, Lebanon, NH 037561
Willett WC ,Stampfer MJ ,Underwood BA ,Taylor JO ,Hennekens CH . Vitamins A, E, and carotene: effects of supplementation on their plasma levels. Am J Clin Nutr 1983;38:559-566
Web of Science | Medline2
Costantino JP ,Kuller LH ,Begg L ,Redmond CK ,Bates MW . Serum level changes after administration of a pharmacologic dose of β-carotene. Am J Clin Nutr 1988;48:1277-1283
Web of Science | Medline3
Johnson EJ ,Suter PM ,Sahyoun N ,Ribayo-Mercado JD ,Russell RM . Relation between β-carotene intake and plasma and adipose tissue concentrations of carotenoids and retinoids. Am J Clin Nutr 1995;62:598-603
Web of Science | Medline4
van Poppel G . Carotenoids and cancer: an update with emphasis on human intervention studies. Eur J Cancer 1993;29:1335-1344
CrossRef | Web of Science
- Citing Articles (5)
Citing Articles
1
Steven E Lipshultz, Stacy D Fisher, Wyman W Lai, Tracie L Miller. (2003) Cardiovascular risk factors, monitoring, and therapy for HIV-infected patients. AIDS 17, S96-S122
CrossRef2
P. Greenwald. (2003) -Carotene and Lung Cancer: A Lesson for Future Chemoprevention Investigations?. JNCI Journal of the National Cancer Institute 95:1, 1E-1E
CrossRef3
STEVEN E. LIPSHULTZ, STACY D. FISHER, WYMAN W. LAI, TRACIE L. MILLER. (2001) Cardiovascular Monitoring and Therapy for HIV-Infected Patients. Annals of the New York Academy of Sciences 946:1, 236-273
CrossRef4
Kathryn A Carolin, Helen A Pass. (2000) Prevention of breast cancer. Critical Reviews in Oncology/Hematology 33:3, 221-238
CrossRef5
A. Tavani, C. La Vecchia. (1999) β-Carotene and risk of coronary heart disease. A review of observational and intervention studies. Biomedicine & Pharmacotherapy 53:9, 409-416
CrossRef
