Genome-Wide Meta-Analysis and Mendelian Randomization Identify Early Biomarkers of Non-Alcoholic Fatty Liver Disease

Background: The diagnosis of non-alcoholic fatty liver disease (NAFLD) is often challenging. Blood-based biomarkers which are causally influenced by NAFLD and that are not modulated by secondary non-causal pathways, are promising candidates for the identification of patients with NAFLD. Objectives: To identify blood metabolites and blood proteins that are causally impacted by the presence of NAFLD using Mendelian randomization (MR). Methods: We created a NAFLD genetic instrument through the identification of independent single-nucleotide polymorphisms associated with NAFLD in a meta-analysis of genome-wide association studies (GWAS) (6715 cases and 682,748 controls). Using inverse-variance weighted MR, we investigated the impact of NAFLD on 123 blood metabolites (in 24,925 participants from 10 European cohorts) and 3283 blood proteins (in 3301 participants from the INTERVAL cohort). Results: Our genetic instrument for genetically predicted NAFLD included 12 SNPs at the MTARC1, GCKR, LPL, TRIB1, LMO3, FTO, TM6SF2, APOE and PNPLA3 loci. After correction for false-discovery rate, we found a positive effect of NAFLD on blood tyrosine levels and on blood levels of eight proteins (encoded by the IDUA, ADH4, HMGCS1, GSTA1, ASL, POR, FBP1 and CTSZ genes). These association were robust to outliers and we found to evidence of horizontal pleiotropy. Conclusions: We report the existence of a potentially causal impact of the presence of NAFLD on tyrosine metabolism as well as on eight circulating proteins, which could potentially represent early biomarkers of NAFLD.