Abstract
BACKGROUND: There is evidence for an association between the gut microbiome and endometriosis. However, their causal relationship and the mediating role of lipid metabolism remain unclear. METHODS: Using genome-wide association study (GWAS) data, we conducted a bidirectional Mendelian randomization (MR) analysis to investigate the causal relationships between gut microbiome and endometriosis. The inverse variance weighted (IVW) method was used as the primary model, with other MR models used for comparison. Sensitivity analysis based on different statistical assumptions was used to evaluate whether the results were robust. A two-step MR analysis was further conducted to explore the mediating effects of lipids, by integrating univariable MR and the multivariate MR method based on the Bayesian model averaging method (MR-BMA). RESULTS: We identified four possible intestinal bacteria genera associated with the risk of endometriosis through the IVW method, including Eubacterium ruminantium group (odds ratio [OR] = 0.881, 95% CI: 0.795-0.976, P = 0.015), Anaerotruncus (OR = 1.252, 95% CI: 1.028-1.525, P = 0.025), Olsenella (OR = 1.110, 95% CI: 1.007-1.223, P = 0.036), and Oscillospira (OR = 1.215, 95% CI: 1.014-1.456, P = 0.035). The further two-step MR analysis identified that the effect of Olsenella on endometriosis was mediated by triglycerides (proportion mediated: 3.3%; 95% CI = 1.5-5.1%). CONCLUSION: This MR study found evidence for specific gut microbiomes associated with the risk of endometriosis, which might partially be mediated by triglycerides.