POI is a highly heterogeneous, multifactorial condition, and dysregulated lipid metabolism has been implicated in its inflammatory pathogenesis This study is the first to systematically investigate causal relationships between 179 lipid species, 91 inflammatory factors, and POI using Two-Sample Mendelian Randomization (TSMR) and Multivariable Mendelian Randomization (MVMR). By integrating lipidomics and inflammatories data with POI from Genome-wide association study (GWAS) and FinnGen, we identified 18 causally significant lipids, including risk-elevating phosphatidylcholines and sphingomyelins, and protective triglycerides. Methodologically, we innovatively applied Bayesian Weighted Mendelian Randomization (BWMR) to confirm the robustness of causal estimates, addressing limitations of conventional MR in pleiotropy-prone metabolic networks. Biologically, we discovered IL-10 mediates 7.02–9.03 % of the effects of sphingomyelin (d40:2) and (d42:2) on POI, reconciling lipid-driven inflammation with ovarian aging—a mechanism previously unreported. Sensitivity analyses confirmed no horizontal pleiotropy (p > 0.05). This work establishes three advances: (1) First MR evidence linking specific lipid subclasses (not just broad categories) to POI; (2) Identification of IL-10 as a novel inflammatory mediator bridging sphingolipids and POI pathogenesis; (3) A validated framework combining MVMR and mediation analysis to disentangle direct/indirect effects in reproductive aging. Our findings provide clinically actionable insights: IL-10 emerge as potential biomarkers, while triglycerides highlight dietary/therapeutic targets. This mechanistic clarity advances POI research beyond prior observational associations into causal biology.
扫码关注我们
求助内容:
应助结果提醒方式:
