Association of protein signature of fine particulate matter with MASLD: the underlying pathways and Mendelian randomization analysis

INTRODUCTION: MASLD is a common chronic liver disease worldwide; identifying modifiable risk factors is crucial. Although fine particulate matter (PM2.5) has been epidemiologically linked to MASLD, its mechanism remains unclear. OBJECTIVES: To explore the mediating role of protein signature on the association between PM2.5 and metabolic dysfunction-associated steatotic liver disease (MASLD), and the underlying biological pathways. METHODS: We included 52,190 participants free of MASLD at baseline of a large cohort. Annual average PM2.5 exposure was estimated by bilinear interpolation, and time-weighted method. PM2.5-related protein signature was constructed by elastic net regression, and KEGG/GO enrichment analysis was performed to identify potential pathways. Cox models were used to evaluate the associations of PM2.5 and the protein signature with incident MASLD. Mendelian randomization (MR) analysis was applied to examine potential causality between proteins and MASLD. Structural equation modeling (SEM) was employed to the evaluate statistical mediation involving the protein and metabolic signatures. RESULTS: During a median follow-up of 11.8 years, 624 incident MASLD cases were identified. The PM2.5-related protein signature comprised 700 proteins, which were mainly enriched in pathways including NF-κB signaling pathway, and cytokine-cytokine receptor interaction. Both PM2.5 exposure, and protein signature were significantly associated with incident MASLD, with hazard ratios (95 % confidence intervals) of 1.19 (1.10, 1.30), and 1.19 (1.08, 1.31). MR analysis provided genetic evidence supporting that CD28, CSF3 and HAVCR2 were associated with MASLD. Mediation analysis indicated that PM2.5-related protein signature mediated 15 % (95 % CI: 2.7 %, 37.7 %) of the PM2.5-MASLD association. SEM further suggested that PM2.5-related protein signature might mediate the PM2.5-MASLD association by regulating metabolic signatures. CONCLUSIONS: The PM2.5-related protein signature was associated with MASLD, and showed statistical mediation of PM2.5-MASLD association potentially involving inflammatory and metabolic pathways.