Main MR analysis
For the main MR analysis, we used the inverse-variance weighted (IVW) method 23. This method assumes that all instruments are valid, although it can still provide unbiased estimates in some scenarios where the IV assumptions are violated (e.g., balanced horizontal pleiotropy). In this study, our focus is on effect size and precision, and we interpret p-values as indicators of the strength of evidence 24,25. Since we examined four sex hormones (estradiol, TT, BT and SHBG) in the MR analysis, we considered a p-value less than 0.0125 (0.05/4) as strong evidence for MR estimates, and a p-value between 0.0125 and 0.05 as suggestive causal evidence.