Mendelian randomization analyses identified bioavailable testosterone mediates the effect of sex hormone-binding globulin on prostate cancer

ObjectiveA better knowledge of the hormonal etiology of prostate cancer is essential for its prevention and treatment. The goal of this study was to provide causal estimates of the connection between sex hormone-binding globulin and prostate cancer and investigate the possible mediating function of other modifiable risk indicators. MethodsWe used two-step, two-sample multivariable Mendelian randomization using single-nucleotide polymorphisms as instrumental variables for exposure and mediators. Single-nucleotide polymorphisms associated with sex hormone-binding globulin and bioavailable testosterone were screened via a genome-wide association study enrolling European-descent adult male individuals. Summary-level data for prostate cancer (79,148 cases and 61,106 controls) were extracted from the PRACTICAL consortium. The total effect of sex hormone-binding globulin on prostate cancer risk was decomposed into direct and indirect effects through the mediator, bioavailable testosterone. An inverse-variance-weighted method was the primary Mendelian randomization analysis method. Sensitivity analyses were performed via Mendelian randomization-Egger regression, heterogeneity test, pleiotropy test, and leave-one-out test. The directionality that exposure causes the outcome was verified using Mendelian randomization-Steiger test. ResultsIn the univariable Mendelian randomization analysis, genetically predicted higher sex hormone-binding globulin levels had a causal association with lower prostate cancer risk (odds ratio = 0.944, 95% confidence interval = 0.897-0.993, p = 0.027) and an inverse association with bioavailable testosterone level (odds ratio = 0.945, 95% confidence interval = 0.926-0.965, p = 1.62E-07) without controlling for other factors. Moreover, an increase of one standard deviation (59.5 pmol/L) in genetically predicted bioavailable testosterone level was significantly associated with a 22.0% increase in the overall prostate cancer risk (odds ratio = 1.220, 95% confidence interval = 1.064-1.398, p = 0.004) after adjusting for sex hormone-binding globulin level. The effect size ratio of bioavailable testosterone-mediated sex hormone-binding globulin to prostate cancer was further analyzed to clarify the importance of the mediating effect. Notably, the mediator bioavailable testosterone explained 19.28% (95% confidence interval = 10.76%, 73.78%) of the total effect of sex hormone-binding globulin level on prostate cancer risk. ConclusionThe results support the potentially protective causal effect of genetically predicted higher sex hormone-binding globulin levels against prostate cancer with mediation by the modifiable risk factor, bioavailable testosterone. More research is needed to determine how this possible sex hormone-binding globulin-bioavailable testosterone-prostate cancer link works. Targeting sex hormone-binding globulin and bioavailable testosterone traits may be a valuable strategy for preventing prostate cancer.