BACKGROUND & AIMS:: We performed a population-based study to assess factors that are independently associated with hepatocellular carcinoma (HCC)-related mortality.

METHODS:: We evaluated clinico-demographic, laboratory, and mortality data collected from 15,866 individuals in the Third National Health and Nutrition Examination Survey (NHANES III) from1988 to 1994. The etiology of chronic liver disease was determined using serologic tests to measure hepatitis C virus (HCV) RNA, hepatitis B surface antigen, and iron; excessive alcohol consumption and non-alcoholic fatty liver disease were determined. Cohorts were compared with controls using a stratum- specific chi(2) tests. The Cox proportional hazard model was used to identify independent predictors of HCC-related mortality.

RESULTS:: After a follow-up of 160 months, 14.55% of the individuals had died; 83 deaths were liver-related (25 HCC and 58 non-HCC liver related). Factors that independently predicted HCC-related mortality were age (hazard ratio [HR]=1.10; 1.04-1.16; P=0.0021), Hispanic ethnicity (HR=5.14;1.75-15.06; P=0.0036), and HCV infection (HR=18.12; 3.57-91.98; P=0.0008). Factors that independently predicted non-HCC liver-related mortality included age (HR=1.07; 1.04-1.10; P < 0.0001), male sex (HR=3.29; 1.15-9.42: P=0.0277), alcoholic liver disease (HR=10.81; 1.32-88.26: P= 0.0271, HCV (HR=27.00; 4.70-155.1; P=0.0004), iron overload (HR=6.18; 1.82-20.97); P=0.0043), or NAFLD (HR=11.56; 3.21-41.67; P=0.0004).

CONCLUSIONS:: This population-based study showed that HCV infection and Hispanic ethnicity independently increase the risk for HCC-related mortality. All liver diseases, including NAFLD, increase the risk for non-HCC liver-related mortality.