PURPOSE:

To evaluate the relation between consumption of coffee and other methylxanthine-containing beverages and liver cirrhosis.

METHODS:

A hospital-based case-control study of digestive tract and liver diseases was conducted in Greater Milan, Italy, including 101 cases with liver cirrhosis and 1538 controls.

RESULTS:

Compared with coffee non-drinkers, the multivariate odds ratio (OR) was 0.77 for one cup of coffee per day, 0.57 for two, and 0.29 for three or more. The OR for 40 years of coffee consumption or more was 0.45. Trends in risk were significant for both number of cups and duration of coffee drinking. No significant association was observed with decaffeinated coffee, tea and cola-containing beverages. The relation between coffee consumption and liver cirrhosis was not attributable to confounding and was observed across strata of tobacco, alcohol, and other major covariates of interest. In particular, an inverse relation was observed also in subjects reporting moderate alcohol drinking.

CONCLUSIONS:

The present study confirms, and further quantifies, the existence of an inverse association between coffee consumption and liver cirrhosis. However, the metabolism of caffeine is impaired in fasting subjects with liver cirrhosis, and the association could be due to a reduction of coffee drinking in subjects with liver cirrhosis.