Inverse Association Between Serum 25-Hydroxyvitamin D and Nonalcoholic Fatty Liver Disease

Abstract

Serum 25-hydroxyvitamin D [S-25(OH)D] and nonalcoholic fatty liver disease (NAFLD) are correlated in many observational studies, whereas the causality of this association is uncertain, especially in European populations. We conducted a bidirectional Mendelian randomization study to determine the association between S-25(OH)D and NAFLD. Seven and 6 independent genetic variants associated with S-25(OH)D and NAFLD at the genome-wide-significance level, respectively, were selected as instrumental variables. Summary-level data for S-25(OH)D were obtained from the Study of Underlying Genetic Determinants of Vitamin D and Highly Related Traits consortium including 79,366 individuals. Summary-level data for NAFLD were available from a genome-wide association meta-analysis (1483 cases and 17,781 controls), the FinnGen consortium (894 cases and 217,898 controls), and the UK Biobank study (275 cases and 360,919 controls). Summary-level data for 4 liver enzymes were obtained from the UK Biobank. There were genetic correlations of S-25(OH)D with NAFLD and certain liver enzymes. Genetically predicted higher levels of S-25(OH)D were consistently associated with a decreased risk of NAFLD in the 3 sources. For a 1-SD increase in genetically predicted S-25(OH)D levels, the combined odds ratio of NAFLD was 0.78 (95% confidence interval [CI], 0.69 to 0.89). Genetically predicted higher levels of S-25(OH)D showed a borderline association with aspartate aminotransferase levels (change -1.17; 95% CI, -1.36 to 0.01). Genetic predisposition to NAFLD was not associated with S-25(OH)D (change 0.13; 95% CI, -1.26 to 0.53). Our findings have clinical implications as they suggest that increased vitamin D levels may play a role in NAFLD prevention in European populations.