Predictors of serum selenium in cigarette smokers and the lack of association with lung and prostate cancer risk

Epidemiological studies have suggested that low levels of selenium are asso ciated with a higher incidence of both lung and prostate cancer. We analyze d the selenium serum concentration in 356 Carotene and Retinol Efficacy Tri al (CARET) participants who later developed lung cancer and 356 matched con trols and in 235 prostate cancer cases and 456 matched controls. Serum samp les were obtained a mean of 4.7 years before diagnosis for both tumor types . Controls were matched to cases by year of randomization, age, smoking sta tus, treatment arm, exposure population (asbestos workers or cigarette smok ers), and year of blood draw. In the control population (it = 820), signifi cant predictors of low serum selenium concentration were current smoking st atus and East Coast locations of the study center. Overall, there was no si gnificant difference in mean serum selenium in lung cancer cases versus con trols (11.91 mug/dl versus 11.77 mug/dl) or prostate cancer cases versus co ntrols (11.48 mug/dl versus 11.43 mug/dl). No statistically significant tre nd in odds ratio was seen across quartiles of serum selenium for lung cance r (P = 0.49) or prostate cancer (P = 0.69). In a subpopulation of 174 prost ate cancer patients who had clinical and pathological staging material revi ewed, there was no association between serum selenium and Gleason score or clinical or pathological stage. In the CARET population of current and form er smokers consuming an ad libitum diet, the serum concentration of seleniu m was not a risk factor for either lung cancer or prostate cancer.