The incidence of hepatocellular carcinoma was reported to be increasing in the United States. However, alternate explanations were diagnostic or reclassification bias and changes in the demographic features of the general population.

To examine the temporal trends in the incidence of hepatocellular carcinoma.

Retrospective cohort study.

Information collected by population-based registries of the Surveillance, Epidemiology, and End Results (SEER) program.

Persons given a diagnosis of hepatocellular carcinoma between 1975 and 1998.

Linear Poisson multivariate regression model, controlling for differences in age, sex, race or ethnicity, and geographic region among patients with hepatocellular carcinoma and in the underlying population.

The overall age-adjusted incidence rates of hepatocellular carcinoma increased from 1.4 per 100 000 in 1975 to 1977 to 3.0 per 100 000 in 1996 to 1998. There was a 25% increase during the last 3 years of the study compared with the preceding 3 years (1993 to 1995). The increase affected most age groups above 40 years, with the greatest increase in the 45- to 49-year-old age group. White men had the greatest increase (31%) in the last time period (1996 to 1998) compared with 1993 to 1995. The Poisson regression model confirmed an almost 2-fold increase in the incidence rate ratio for hepatocellular carcinoma between 1975 to 1978 and 1996 to 1998.

The incidence of hepatocellular carcinoma continues to increase rapidly in the United States, with rates increasing the fastest in white men 45 to 54 years of age. These findings are consistent with a true increase and could be explained by consequences of hepatitis C virus acquired during the 1960s and 1970s.