Pesticides and prostate cancer: a review of epidemiologic studies with specific agricultural exposure information.

Prostate cancer is the most commonly diagnosed cancer in US men, and the second most commonly diagnosed cancer among men worldwide. Although pesticides have been implicated in studies of prostate cancer among farmers, meta-analyses have found heterogeneity across studies, and a number of exposures and lifestyle factors may be unique to farmers. The purpose of this paper is to review the epidemiologic literature to evaluate the hypothesis that agricultural exposure to pesticides is causally associated with prostate cancer risk. We analyzed the eight cohort studies and five case-control studies that quantified and/or evaluated agricultural exposure to particular pesticide classes or chemicals. Despite sporadic positive findings, these studies did not show consistently increased risks to support a causal association between agricultural pesticide use and prostate cancer. Studies using an 'external' comparison group must be interpreted in the context of confounding by differences in prostate-specific antigen screening intensity. Furthermore, most studies did not adjust for potential confounders other than age and time period. It is clearly not possible to exonerate any particular pesticide as a putative cause of prostate cancer - to do so would require an inverse empirical association with an upper confidence limit below the null value. Existing evidence does not point to any pesticide as satisfying widely used guidelines for establishing causation: a strong, exposure-dependent and demonstrably unconfounded, unbiased association, documented in several studies.