J Gastroenterol Hepatol. 2026 Oct 25. doi: 10.1111/jgh.70845. Online ahead of print.
ABSTRACT
BACKGROUND AND AIM: The prognosis of lean metabolic dysfunction-associated steatotic liver disease (MASLD) remains controversial. We investigated the long-term prognosis of lean MASLD and its time-dependent associations with all-cause and cardiovascular mortality.
METHODS: We analyzed 3568 participants with MASLD from the Third National Health and Nutrition Examination Survey. Cox proportional hazards and Fine-Gray competing-risk models assessed all-cause and cardiovascular mortality. Because the proportional hazards assumption was violated for all-cause mortality, 10-year landmark and time-varying coefficient Cox analyses were performed. Restricted cubic splines (RCS) evaluated the association between body mass index (BMI) and early all-cause mortality, and subgroup analyses identified high-risk populations.
RESULTS: During a median 27.7-year follow-up, lean MASLD was associated with higher all-cause mortality within 0-10 years (adjusted hazard ratio [HR], 1.68; 95% confidence interval [CI], 1.29-2.20), but not thereafter (adjusted HR, 1.04; 95% CI, 0.87-1.25). No significant association between lean MASLD and cardiovascular mortality was observed during either pre-landmark (adjusted subdistribution HR [sHR], 1.45; 95% CI, 0.95-2.22) or post-landmark (adjusted sHR, 0.97; 95% CI, 0.69-1.36) in the overall population. Time-specific HR demonstrated a gradual attenuation of the excess mortality association over follow-up. RCS demonstrated a nonlinear U-shaped association between BMI and early all-cause mortality. Excess early mortality risks were particularly evident among males aged ≥ 60 years and participants with baseline FIB-4 ≥ 1.30.
CONCLUSIONS: Lean MASLD conferred a time-dependent excess risk of all-cause mortality, predominantly within the first decade of follow-up, but was not robustly associated with cardiovascular mortality in the overall population. Older males and patients with elevated FIB-4 scores represented high-risk phenotypes.
PMID:42927930 | DOI:10.1111/jgh.70845