Background: Although early reports have suggested an association between circumcision and prostate cancer (PCa) development, results of subsequent epidemiological studies have been conflicting. Here we examine published articles that explore this association.
Methods: We searched MEDLINE through PubMed and Embase for articles reporting on the association between PCa and circumcision, and performed a meta-analysis of qualifying studies.
Results: On the basis of seven reports of case-control studies published from 1971 to 2014, overall findings showed nonsignificant reduced risk (odds ratio (OR) 0.88, P=0.19) of PCa in circumcised men compared with uncircumcised men, obtained under heterogeneous conditions (I(2)=65%). Heterogeneity and nonsignificance were erased when the overall effect was subjected to outlier treatment and three studies omitted (OR 0.90, P=0.04, I(2)=0%). Furthermore, subgroup analysis showed significantly reduced risks in the following subgroups: (i) post-PSA testing publications (OR 0.88, P=0.01), (ii) population-based studies (OR 0.84, P=0.05), (iii) studies that collected data by personal interview (OR 0.83, P=0.03) and (iv) studies in black race (OR 0.59, P=0.02). The strengths of these summary effects lie in the robustness revealed by sensitivity analysis.
Conclusions: Stability of the reduced risks observed in key subgroups suggests that the protective feature of circumcision status against PCa is best seen in the context of the post-PSA testing and population-based studies as well as in the black race subgroup.