Genetically predicted body mass index and prostate cancer risk

Abstract Objectives We employ network Mendelian randomization (MR) to investigate the causal relationship between body mass index (BMI) and prostate cancer, while concurrently examining the potential mediating role of both bioavailable and total testosterone. Methods The PRACTICAL consortium is used for prostate cancer. Data from GIANT consortium are used for BMI. The Inverse variance weighted (IVW), MR Egger, Weighted median, and Weighted mode methods are used to evaluate evidence for a causal relationship. The BMI-prostate cancer, BMI-testosterone, and testosterone-prostate cancer associations are analyzed. Results In the IVW method, it shows that BMI will decrease prostate cancer risk (odds ratio, 0.822; 95% confidence interval, 0.738–0.915; P < 0.001), proving that genetically predicted BMI has a negative causality with prostate cancer. BMI is causally negatively associated with bioavailable testosterone (in the IVW method, Beta, -0.146; SE, 0.041; P < 0.001) and total testosterone (in the IVW method, Beta, -0.210; SE, 0.047; P < 0.001). Yet, only higher bioavailable testosterone level is positively associated with an increased prostate cancer risk. Conclusion Our results support evidence for an inverse association between genetic predicted BMI and prostate cancer. Our study also suggests the hypothesis that bioavailable testosterone but not total testosterone may mediate the inverse association from BMI to prostate cancer.