IDDF2023-ABS-0277 Circulating fatty acids (FAs) and non-alcoholic fatty liver disease (NAFLD): a two-sample mendelian randomization analysis

BackgroundChronic liver disease has become a major public health problem worldwide. It has been discovered that circulating fatty acids (FAs) have different effects on chronic diseases: polyunsaturated FAs (PUFAs) show a protective effect whereas saturated FAs (SFAs) have a pro-inflammatory effect. Therefore, we aimed to test the causality of circulating FAs on NAFLD using Mendelian randomization (MR) analysis, which incorporates SNP as instrumental variables (IVs).MethodsCirculating fatty acids were Omega-3, Omega-6, PUFA, monounsaturated FA (MUFA), SFA and total FA. Genetic summary statistics of circulating FAs and the NAFLD status were retrieved from UK Biobank (n=114,999) and the FinnGen Biobank (n=342,499), respectively. All participants were of European ancestry. SNPs with a genome-wide significance (p−8) were selected and pruned by linkage disequilibrium (r2≥0.0001, >100,000 kb). Inverse Variance Weighted (IVW) was used as the primary MR analysis method. The strength of instrumental variables was tested using F-statistic. Heterogeneity was tested using Cochrane’s Q test. Horizontal pleiotropy was tested using the MR-Egger intercept. Sensitivity analysis was performed using leave-one-out analysis, MR Egger, weighted median, simple mode, and weighted mode methods.ResultsThe F-statistic showed a small magnitude of the bias of the IVs. MR-Egger intercept had p-values >0.05, indicating a lower possibility of being biased by horizontal pleiotropy. IVW revealed insignificant associations between circulating FAs and the status of NAFLD: 22 SNPs were identified for Omega-3 fatty acids (r=-0.16, p=0.523); 24 SNPs for Omega-6 fatty acids (r=-0.031, p=0.933); 16 SNPs for Omega-6:3 ratio (r=0.086, p=0.741); 25 SNPs for PUFA (r=-0.044, p=0.893); 24 SNPs for MUFA (r=0.23, p=0.293); 26 SNPs for PUFA: MUFA ratio (r=0.0008, p=0.997); 27 SNPs for SFA (r=-0.07, p=0.818); 30 SNPs for total FA (r=0.007, p=0.980). Sensitivity analyses indicated that the results are consistent with IVW.ConclusionsGenetically predicted circulating FAs may not be causally associated with NAFLD.