Abstract
PURPOSE: This study aims to assess the prognostic value of the fatty liver index (FLI), a noninvasive tool for hepatic steatosis assessment, in predicting all-cause and disease-specific mortality. METHODS: We linked data from the National Health and Nutrition Examination Survey III (1988-1994) with Public-Use Mortality Files, forming a cohort of 11 297 participants with a median follow-up period of 26.25 years. Cox proportional hazards models were used to evaluate the association between FLI and all-cause mortality, while Fine and Gray's models assessed the relationship between FLI and disease-specific mortality. RESULTS: The FLI ≥ 60 was independently associated with an increased risk of all-cause mortality (hazard ratio = 1.24, P < 0.001), as well as mortality from malignant neoplasms (hazard ratio = 1.18, P = 0.048), diabetes (hazard ratio = 2.62, P = 0.001), and cardiovascular diseases (CVDs) (hazard ratio = 1.18, P = 0.018), compared to FLI < 30. No significant associations were found with Alzheimer's disease, influenza and pneumonia, chronic lower respiratory diseases, or renal disorders. Subgroup analyses indicated that individuals who were females aged 40-60 (hazard ratio = 1.67, P = 0.003), non-overweight (hazard ratio = 1.75, P = 0.007), or without abdominal obesity (hazard ratio = 1.75, P = 0.007) exhibited a stronger association between FLI ≥ 60 and all-cause mortality. CONCLUSION: These findings support the prognostic value of the FLI for predicting mortality from all causes, malignant neoplasms, diabetes, and CVDs. Targeted attention is needed in postmenopausal women, non-overweight, and non-abdominally obese populations.