Abstract
Prospective evidence linking n-3 polyunsaturated fatty acid (PUFA) supplementation to non-alcoholic fatty liver disease (NAFLD) risk remains limited. In this cohort study of 488 888 UK Biobank participants, we evaluated associations of fish oil supplementation (a key source of n-3 PUFAs) with severe NAFLD incidence using Cox proportional hazards models. Mediation analysis evaluated the roles of serum metabolic and inflammatory biomarkers. Genetic interaction analyses included polymorphisms linked to n-3 PUFA metabolism. Over 12.3 years of follow-up, 5671 participants developed severe NAFLD. Fish oil supplementation was associated with a 7% lower risk of severe NAFLD (HR 0.93, 95% CI 0.87-0.99; P = 0.022). Mediation analyses revealed that C-reactive protein (CRP) (8.8% mediated), high-density lipoprotein cholesterol (HDL-C) (7.5%), and cystatin C (10.0%) partially explained this association. No significant interactions were observed between supplementation and genetic variants (all P > 0.05). These findings support the use of n-3 PUFA supplements in the primary prevention of severe NAFLD.