We propose to analyze baseline blood samples for beta-carotene, alpha- tocopherol and retinol among cases and controls of non-melanoma skin cancer (NMSC) that were ascertained in the Physicians' Health Study (PHS), a randomized trial of aspirin and beta-carotene among 22,071 U.S. male physicians aged 40-84 years. The trial begin in 1982, and the randomized period ended in December, 1995, for an average of 13 years of randomization, mostly in 1982, from 14,916 trial participants who were free from prior MI, stroke or cancer (excluding non-melanoma skin cancer). These have been maintained and stored at -82 degrees Celsius. During the trial period, a total of 3,613 physicians reported developing incident NMSC. Of these, 2,541 have provided baseline plasma specimens. Using a nested case-control design, we will analyze the specimens from a sample of 1300 cases and 1300 controls matched on age and smoking status. We will examine the hypothesis that the benefit of supplementation with randomized beta-carotene may depend on baseline levels of plasma beta-carotene, and may even be limited to those with the lowest baseline levels. In addition, we will determine whether risk of NMSC is inversely related to baseline plasma levels of beta-carotene, alpha-tocopherol, or retinol among those randomized to beta-carotene placebo. The nested case-control approaches provides an efficient means of testing hypotheses using blood samples that were collected at considerable expense and stored with great cre in the PHS.
