Diverging rate and absolute burden patterns of liver cancer attributable to NASH: global evidence and China as a sentinel case

BACKGROUND: The Global Burden of Disease Study (GBD) category "liver cancer due to non-alcoholic steatohepatitis" is modeled, rather than clinically adjudicated. We quantified this burden and tested whether declining age-standardized rates could coexist with rising absolute caseloads in China. METHODS: We analyzed the GBD 2023 estimates from 204 countries and territories for 1990-2023. Etiologic shares were reported without intervals because the matched posterior draws were unavailable. We used full-period log-linear trends, segmented trends with a 2019 cutoff sensitivity analysis, Shapley decomposition with parametric uncertainty, and Bayesian local-linear-trend projections to 2045. RESULTS: Global incident cases increased from 15,752 (95% uncertainty interval, 12,069-19,989) to 45,669 (34,794-57,442); the age-standardized incidence EAPC was 0.54% (95% confidence interval, 0.46-0.62). The modeled etiologic share rose from 5.8% to 7.9%. In China, the age-standardized incidence rate declined (EAPC, -0.68%), but cases increased from 4660 to 10,201. Population growth, aging, and age-specific rate change contributed +1413, +5382, and -1254 cases, respectively; their 95% simulation intervals were +1157 to +1729, +4401 to +6,597, and -3134 to +448. Full-data and fit-through-2019 incidence projections for 2045 were similar. CONCLUSIONS: NASH-attributable liver cancer remains a minority of all liver cancer but is increasing. Absolute caseloads should accompany age-standardized rates in planning. Decomposition does not identify causal mechanisms.