Abstract
BACKGROUND: This study explored the causal connection among the plasma lipidome, nonalcoholic fatty liver disease (NAFLD), and potential metabolome mediators through Mendelian randomization (MR). METHODS: We obtained summary statistics for 179 plasma lipidome traits (N = 7,174), 1,400 plasma metabolome traits (N = 8,299), and one NAFLD trait from publicly available genome-wide association studies. A two-sample MR analysis was conducted to infer causality. Additionally, multiple sensitivity analyses were conducted to assess the heterogeneity, horizontal pleiotropy, and robustness of the MR findings. MetaboAnalyst 6.0 was used for the pathway analysis of the identified lipids and metabolites. Furthermore, we used mediation analysis to assess whether the effect of plasma lipidome on NAFLD was mediated by plasma metabolome. RESULTS: The MR analysis predicted a genetically determined causal relationship between plasma lipidomes and NAFLD. No compelling proof was found that genetically predicted NAFLD influenced the risk of the five plasma lipidomes mentioned earlier. Based on established causal relationships between lipids and metabolites, we identified that eight metabolic pathways are closely associated with NAFLD. Our mediation analysis revealed six mediation relationships, indicating the causal pathway from plasma lipids to NAFLD mediated by five specific metabolites. No potential pleiotropy was found in the sensitivity analysis. CONCLUSIONS: In summary, our study identified causal relationships between plasma lipidomes, plasma metabolomes, and NAFLD. Certainly, the impact of plasma lipidomes on NAFLD is not limited to plasma metabolomes, indicating a need to further investigate into other possible mediators. These identified factors may become new biomarkers of the NAFLD contributing to its prevention, diagnosis, and treatment.