Cerebrospinal fluid metabolites and their association with female reproductive disorders: insights from mendelian randomization

In recent years, the fertility rate has been declining. This is because of a drop in the desire for fertility among young people and the incidence of various degrees of reproductive dysfunction. Theories of 'kidney-generating marrow' and 'kidney-governing reproduction' exist in traditional Chinese medicine, and modern science has confirmed the existence of the hypothalamus–pituitary–gonadal axis as a regulatory mechanism. Cerebrospinal fluid (CSF), a naturally occurring fluid in the brain, has the potential to reflect the physiological and pathological states of the brain. Furthermore, its metabolites may play a crucial role in the early detection of ovarian diseases. In this study, the Mendelian randomization (MR) method was used to investigate the potential influence of CSF metabolites on four prevalent ovarian diseases affecting women of childbearing age and their relationship with female infertility. The most recent genome-wide association study was used to synthesize the statistical data, and a comprehensive two-sample MR analysis was used to assess the causal relationship between 338 CSF metabolites and four ovarian-related diseases (including polycystic ovary syndrome, ovarian dysfunction, benign ovarian tumors, and salpingitis), as well as female infertility. The two-sample MR analysis was conducted using the inverse-variance weighted (IVW) method as the primary approach and complemented by pleiotropy and heterogeneity tests. To ensure the reliability of the findings, a leave-one-out sensitivity analysis was performed. Multiple testing was controlled using false discovery rate (FDR) correction applied to the primary IVW P values (FDR < 0.05). The reliability of the results was further validated by incorporating the Bayesian Weighted Mendelian Randomization (BWMR) method. Finally, the robustness of the results was validated through a reverse MR analysis. After IVW-based FDR control, seven CSF metabolite–outcome associations remained robust. For ovarian dysfunction, each 1-SD increase in genetically predicted 2-oxoadipate levels was associated with an 11.1% higher risk (OR = 1.111, 95% CI 1.062–1.163; FDR = 0.002). In contrast, each 1-SD increase in genetically predicted N6-succinyladenosine levels was associated with a 17.8% lower risk (OR = 0.822, 95% CI 0.738–0.915; FDR = 0.039), and each 1-SD increase in genetically predicted urea levels was associated with a 30.3% lower risk (OR = 0.697, 95% CI 0.572–0.848; FDR = 0.039). For polycystic ovary syndrome, each 1-SD increase in genetically predicted 5,6-dihydrothymine levels was associated with a 51.0% higher risk (OR = 1.510, 95% CI 1.212–1.882; FDR = 0.041). For female infertility, each 1-SD increase in genetically predicted 2-hydroxy-3-methylvalerate levels was associated with a 22.7% higher risk (OR = 1.227, 95% CI 1.117–1.348; FDR = 0.003), whereas each 1-SD increase in genetically predicted 2,3-dihydroxyisovalerate levels was associated with a 3.4% lower risk (OR = 0.966, 95% CI 0.951–0.981; FDR = 0.003). Sensitivity analyses and BWMR supported robustness. Reverse MR showed no evidence of outcome-to-metabolite effects. CSF metabolites may be associated with ovarian disease and female infertility, opening new perspectives for the diagnosis and treatment of female reproductive health from a metabolic perspective.