Multiplicities in the assessment of multiple vitamins: Is it too soon to tell men that vitamins prevent cancer?

摘要

IN THIS ISSUE OF JAMA, GAZIANO AND COLLEAGUES1 RE-port one of several analyses of data from the Physicians' Health Study II (PHS II) randomized controlled trial. Among more than 14 000 male physicians who participated in this study, the authors report a marginally statistically significant (P= .04) inverse relationship between taking a daily multivitamin (Centrum Silver) and the occurrence of cancer.1 The authors did not find that the supplement prevented any particular cancer preferentially, and there was no evidence of an association between adherence and the protective effect. The vitamin supplement was not demonstrated to reduce overall or cancer-specific mortality, although die authors observed nonsignificant reductions in bodi.