Causal effects linking polycystic ovary syndrome and testosterone levels to obstetric outcomes: a two-sample Mendelian randomization study

Objective: Numerous observational studies indicate that polycystic ovary syndrome (PCOS) affects pregnancy progression and outcomes, yet causal evidence remains limited. This study investigated the genetic causal relationships of PCOS and testosterone with pregnancy-related diseases and outcomes. Methods: Two-sample Mendelian randomization (MR) analysis was used to estimate the causal relationship between exposures and outcomes using genetic data from two independent populations. Exposures included PCOS and androgen-related factors. The outcomes were gestational diabetes, pre-eclampsia, preterm labor, single delivery by forceps or vacuum extraction, prolonged labor and prolonged second stage of labor. Summary data were obtained from large-scale genome-wide association studies with sample sizes ranging from 1,902 to 463,010 European individuals. Causal estimates were primarily generated using the inverse-variance weighted (IVW) method, while the MR-Egger, weighted median, simple mode, and weighted mode methods served as complementary approaches. Results: MR analysis revealed distinct causal associations between exposures and obstetric outcomes. Specifically, PCOS was not associated with preterm labor (odds ratio [OR]=1.008, P =0.790) but showed a significant positive association with pre-eclampsia (OR=1.129, P =0.008). Among the androgen-related factors, bioavailable testosterone (BTE) levels had significant causal effects on multiple outcomes: a positive association with preterm labor (OR=1.317, P =0.013), a negative association with prolonged labor (OR=0.998, P =0.010), and an increased likelihood of operative vaginal delivery (forceps or vacuum extraction) in singleton pregnancies (OR=1.560, P =0.010). In contrast, total testosterone (TTE) and sex hormone-binding globulin (SHBG) levels showed no effects on prolonged labor (OR=1.000, P =0.468 and OR=1.000, P =0.914, respectively). Similarly, BTE levels were not causally associated with prolonged second-stage labor (OR=0.999, P =0.307). Notably, higher SHBG levels were associated with a reduced risk of pre-eclampsia (OR=0.773, P =0.013). Conclusions: Our findings suggest that PCOS increases the risk of pre-eclampsia. The high incidence of preterm labor observed among PCOS patients may be related to BTE levels. Furthermore, BTE appears to exert a modest genetic-level influence on prolonged labor and increase the likelihood of forceps and vacuum extractor use in singleton deliveries. These results highlight a potential role of androgen-related pathways in obstetric complications.