Abstract
Evidence from epidemiological literature on the association of circulating micronutrients with risk of nonalcoholic fatty liver disease (NAFLD) is inconsistent. We aimed to elucidate the causal relationships using Mendelian randomization (MR). Single-nucleotide polymorphisms associated with 14 circulating micronutrients (β-carotene, calcium, copper, folate, iron, magnesium, phosphorus, selenium, vitamin B6, B12, C, D, K1 and zinc) were employed as instrumental variables. Summary level data for NAFLD were obtained from a genome-wide association study (GWAS) meta-analysis of 8434 cases and 770,180 controls (discovery stage) and another two datasets including 1483 NAFLD cases and 17,781 controls (replication stage 1) and 2134 NAFLD cases and 33,433 controls (replication stage 2). Inverse variance-weighted method (IVW) was used as primary analysis, supplemented with a series of sensitivity analysis. Genetically predicted higher β‑carotene levels were suggestively associated with reduced NAFLD risk [odds ratio (OR) 0.81, 95% confidence interval (CI) 0.66-0.99; P = 0.047], whereas the association did not survive the false discovery rates (FDR) correction (P(FDR) = 0.164). Genetically predicted circulating iron (OR 1.16, 95% CI 1.05-1.29; P = 0.006, P(FDR) = 0.028), selenium (OR 1.11, 95% CI 1.03-1.20; P = 0.005, P(FDR) = 0.028) and vitamin B12 (OR 1.08, 95% CI 1.03-1.13; P = 0.002, P(FDR) = 0.028) were significantly associated with increased risk of NAFLD. Moreover, the findings were consistent in individual datasets (P(heterogeneity) > 0.05) and confirmed in sensitivity analysis. Our study provided evidence that circulating iron, selenium and vitamin B12 might be causally linked to the risk of NAFLD, which deserves further exploration of the potential biological mechanism.