NIH Press Release
National Cancer Institute

Tuesday, Nov. 5, 1996
4:00 PM Eastern Time

NCI Press Office
(301) 496-6641

Cigarette Smoke or Alcohol Consumption May Enhance Adverse Effects
of Beta Carotene in Vitamin Prevention Trials

Further analyses of two large cancer prevention trials suggest that smoking and alcohol consumption may contribute to the unexpected adverse effects of beta carotene (a vitamin A precursor) first seen in the trials. The new analyses of the trials, funded by the National Cancer Institute (NCI), as well as an editorial appear in the November 6 issue of the Journal of the National Cancer Institute (JNCI).

Both trials -- the Alpha-Tocopherol, Beta-Carotene Cancer (ATBC) Prevention Trial and the Beta Carotene and Retinol Efficacy Trial (CARET) -- were testing whether vitamin supplements would prevent lung cancer among people at high risk for the disease, and unexpectedly found a greater incidence of lung cancer among participants taking high-dose beta carotene supplements. The current reports look more closely at the subgroups of participants in an effort to explain the adverse effects of the beta carotene.

"The public health message of these new analyses remains unchanged -- beta carotene supplements do not lower the risk of lung cancer in a population of long-term cigarette smokers, or asbestos-exposed populations, and may even cause harm," said one of the lead investigators of the ATBC Trial, Demetrius Albanes, M.D., from the Division of Cancer Prevention and Control at NCI. "The best way for smokers to decrease their risk of getting lung cancer is still to stop smoking."

The ATBC Study, conducted by NCI in collaboration with the National Public Health Institute of Finland, took place between l985 and l993 and involved 29,133 Finnish male smokers, who had, on average, smoked a pack of cigarettes every day for 36 years. The participants took a pill containing one of the following: 50 milligrams (mg) of alpha tocopherol (a form of vitamin E), 20 mg of beta carotene (a precursor of vitamin A), both, or a placebo (a pill that looked like the supplements) daily for five to eight years.

In spite of extensive experimental and epidemiological studies suggesting protective effects of these supplements against cancer, particularly lung cancer, lung cancer incidence was increased by 16 percent among participants supplemented with beta carotene. No significant protective or harmful effects on lung cancer were found in the participants taking vitamin E.

In the current analysis of the ATBC Study1 of cigarette smokers by Albanes, the authors show that the adverse effects of beta carotene appear stronger in men with a relatively modest alcohol intake (more than 11 grams per day -- 15 grams of alcohol is equivalent to one drink) compared with those having a lower intake, and stronger in those smoking at least 20 cigarettes daily compared to those smoking less.

Albanes noted that this alcohol finding is new. "Although there's no clearly established mechanism to explain how the interaction of beta carotene and alcohol might promote lung cancer, it is a lead worth pursuing through basic research, and the finding should be evaluated or confirmed in other human studies," he said.

But the observation reported by Albanes that lung cancer risk was elevated only in heavy smokers, and not in smokers of fewer than 20 cigarettes per day, is consistent with the CARET results published this year in the May 2 issue of The New England Journal of Medicine (NEJM). These results also showed that in the group taking a supplement of beta carotene and vitamin A, current heavy smokers were at much greater risk for lung cancer than former heavy smokers.

In the CARET study, 18,314 U.S. men and women at elevated risk for lung cancer -- smokers, former smokers, and workers exposed to asbestos -- took daily supplements of 30 mg of beta carotene and 25,000 international units of vitamin A (retinyl palmitate) or placebo, for an average of four years. The intervention was stopped in January 1996, 21 months earlier than originally planned, because of interim results showing a 28 percent higher incidence of lung cancer in participants taking the supplements. Thus, CARET confirmed the adverse effects in the ATBC Cancer Prevention Study, published in 1994. Both CARET and ATBC participants will be followed for five more years to determine the long-term effects of the intervention.

The second report2 by CARET's lead investigator, Gilbert S. Omenn, M.D., Ph.D., of the Fred Hutchinson Cancer Research Center and the University of Washington School of Public Health in Seattle is a follow-up to the May 2 NEJM article. The authors found that in the group taking supplements, current smokers are at much greater risk for lung cancer than former smokers and raise the possibility that vitamin supplements in former smokers may have a protective effect.

Omenn also found a greater incidence of lung cancer in participants with the highest alcohol intake (greater than 50 grams/day or more than three drinks/day), but there was no consistent dose-response effect.

Omenn commented, "CARET and ATBC provide definite evidence of no benefit and substantial evidence of harm from beta carotene, both for lung cancer incidence and for cancer, cardiovascular, and total mortality. CARET used beta carotene and retinyl palmitate together; beta carotene is the shared exposure between ATBC and CARET. Now we must investigate the many actions of beta carotene which may account for its apparent carcinogenic effect, including the potential difference in response in current smokers versus former smokers."

He continues, "It would be a terrific reinforcement for smoking cessation if beta carotene or retinoids or other agents could be shown to reduce lung cancer risks in former smokers, since former smokers retain their accumulated high risk for lung cancer."

In an accompanying editorial in JNCI, Susan Taylor Mayne, Ph.D., Yale University School of Medicine, New Haven, CT., and her colleagues propose a molecular mechanism to account for the current findings. Focusing on the link between heavy smoking, beta carotene supplements and lung cancer, Mayne hypothesizes a direct interaction between cigarette smoke and beta carotene in lung tissue. She speculates that the oxidative gases in cigarette smoke may degrade beta-carotene in the blood, producing harmful tumor-promoting beta carotene by-products in the lung. This oxidative degradation of beta carotene may, in turn, reduce its ability to carry out its normally protective antioxidant functions.

Mayne also addresses the findings from another prevention trial -- the Physicians' Health Study (PHS) -- in which more than 22,000 U.S. male doctors took 50 mg of beta carotene or placebo every other day for an average of 12 years. No adverse or beneficial effects were observed, even for current or former smokers.

Mayne proposes that this may be because the blood levels of beta carotene achieved by the intervention in the Physicians' Study were considerably lower (1.2 micrograms per millilter - ug/ml) compared with those in the ATBC and CARET studies (3.0 ug/ml and 2.1 ug/ml, respectively.) Blood levels of beta carotene in most of the U.S. population are between 0.05 ug/ml and 0.5 ug/ml. These lower levels in PHS would presumably decrease the interaction between the cigarette smoke and the beta carotene among the 11 percent current smokers who participated in PHS.

Another finding common to both ATBC and CARET was that participants with higher blood beta carotene concentrations on entering the trials, had fewer subsequent lung cancers -- results consistent with studies done prior to these trials. Clearly, lung cancer risk changes with blood levels.

"Lung cancer risk may be lowered by eating more foods with beta carotene, on the one hand," said Albanes. "But when the concentration of beta carotene in the blood is very high -- 10 to 20 times higher than normal -- lung cancer risk may increase, especially for smokers."

Finally, both Albanes and Omenn raise the possibility that beta carotene itself may not be protective -- there may be other substances in the same fruits and vegetables that contain beta carotene that are protective by themselves, or in combination with beta carotene. Also, factors that are also associated with high fruit and vegetable intake, such as physical activity, low red meat consumption or other risk factors, may be important.

The new findings, however, do not alter the conclusions of E. Robert Greenberg, M.D., and Michael B. Sporn, M.D., both at Dartmouth Medical School in Lebanon, NH, who wrote an editorial in the May 2 NEJM.

"The clinical trials ought to be seen as representing a triumph of the scientific process rather than a failure of therapy. We now know that beta carotene supplements are not an effective means of lowering the risk of cancer. The results send a clear message to the public -- the tens of millions of dollars spent annually on beta carotene supplements should now be diverted to more useful purposes."

On the other hand, "No one should discount the importance of the epidemiologic studies of diet and chronic disease. Persons who eat a relatively large quantity of vegetables, fruits, and grains have a profoundly lower risk of death, particularly from cardiovascular disease and cancer."

The Cancer Information Service provides a nationwide telephone service for cancer patients and their families, the public, and health care professionals. The toll-free number is 1-800-4-CANCER (1-800-422-6237); services provided in English and Spanish. People with TTY equipment may call 1-800-332-8615.

This document is available through the NCI's CancerNet services on the Web ( and through Cancer Fax (dial 301-402-5874 from the handset on your fax machine).

1The title of the study is "Alpha-Tocopherol and Beta-Carotene Supplements and Lung Cancer Incidence in the ATBC Study: Effects of Baseline Characteristics and Study Compliance." The authors are Demetrius Albanes, Olli P. Heinonen, Philip R. Taylor, Jarmo Virtamo, Brenda K. Edwards, Matti Rautalahti, Anne M. Hartman, Juni Palmgren, Laurence S. Freedman, Jaason Haapakoski, Michael, J. Barrett, Pirjo Pietinen, Nea Malila, Eero Tala, Kari Liippo, Eija-Riitta Salomaa, Joseph A. Tangrea, Lyly Teppo, Frederic B. Askin, Eero Taskinen, Yener Erozan, Peter Greenwald, and Jussi K. Huttunen.

2The title of the study is "Risk Factors for Lung Cancer and For Intervention Effects in CARET, the Beta-Carotene and Retinol Efficacy Trial." The authors are Gilbert S. Omenn, Gary E. Goodman, Mark D.Thornquist, John Balmes, Mark R. Cullen, Andrew Glass, James P. Keogh, Frank L. Meyskens, Jr., Barbara Valanis, James H. Williams, Jr., Scott Barnhart, Martin G. Cherniack, Carol A. Brodkin, and Samuel Hammar.