Author(s): Greenberg ER, Baron JA, Stukel TA, Stevens MM, Mandel JS,
Abstract Share this page
Abstract BACKGROUND: Beta carotene has been associated with a decreased risk of human cancer in many studies employing dietary questionnaires or blood measurements, and it has had protective effects in some animal models of carcinogenesis. METHODS: We tested the possible cancer-preventing effects of beta carotene by randomly assigning 1805 patients who had had a recent nonmelanoma skin cancer to receive either 50 mg of beta carotene or placebo per day and by conducting annual skin examinations to determine the occurrence of new nonmelanoma skin cancer. RESULTS: Adherence to the prescribed treatment was good, and after one year the actively treated group's median plasma beta carotene level (3021 nmol per liter) was much higher than that of the control group (354 nmol per liter). After five years of follow-up, however, there was no difference between the groups in the rate of occurrence of the first new nonmelanoma skin cancer (relative rate, 1.05; 95 percent confidence interval, 0.91 to 1.22). In subgroup analyses, active treatment showed no efficacy either in the patients whose initial plasma beta carotene level was in the lowest quartile or in those who currently smoked. There was also no significant difference between treated and control groups in the mean number of new nonmelanoma skin cancers per patient-year. CONCLUSIONS: In persons with a previous nonmelanoma skin cancer, treatment with beta carotene does not reduce the occurrence of new skin cancers over a five-year period of treatment and observation.
This article was published in N Engl J Med
and referenced in Journal of Biometrics & Biostatistics