Multi-omics Mendelian Randomization Reveals the Causal Association and Mediating Mechanisms of PM2.5 with Depression: A Systematic Study Integrating Transcriptomics, Metabolomics, and Radiomics

Background PM _2.5 exposure has been associated with major depressive disorder (MDD), but causal evidence remains inconclusive, and the underlying multilevel mechanisms are poorly understood. Objective To investigate the causal effect of PM _2.5 on MDD and to systematically explore potential mediators across multiple biological layers, including gene expression in blood and brain regions, peripheral and central metabolites, and brain imaging traits, using a multi‑omics Mendelian randomization (MR) framework. Methods We first confirmed a significant genetic overlap between PM _2.5 and MDD using LD score regression, and taking this as a premise, we then conducted a multi‑omics MR study to assess the causal effect of genetically predicted PM _2.5 on MDD. A two‑step MR approach was used to evaluate mediation effects through blood and brain‑region gene expression, blood and cerebrospinal fluid metabolites, and brain imaging‑derived phenotypes. Results LD score regression revealed significant genetic overlap (r _g = 0.3338, p = 1.92E‑11). Genetically predicted PM _2.5 causally increased MDD risk (OR = 1.27, 95% CI: 1.07–1.51, p = 5.95E‑03). Mediation analyses identified five blood genes (enriched in antigen presentation and xenobiotic metabolism), six brain‑region genes (involved in autophagy, proteostasis, and immune response), and two blood fatty acid ratio metabolites as significant mediators. No significant mediation was found for cerebrospinal fluid metabolites or brain imaging traits. Significance: This study provides robust genetic evidence for a causal role of PM _2.5 in MDD. By integrating multi‑omics data, it delineates a multilevel mechanistic landscape—spanning peripheral immunometabolic disturbances and region‑specific autophagy, ER stress, and adaptive immunity—thereby advancing our understanding of how air pollution contributes to depression and offering potential molecular targets for future prevention and intervention.