[Patients with alcoholic liver disease hospitalized in gastroenterology. A national multicenter study]

To describe the characteristics of in-patients with alcoholic liver disease in Hepatogastroenterology and to evaluate whether geographic location was a risk factor for cirrhosis.

A French, national, multicenter, prospective investigation was performed in the last quarter of 1997. To be included in the study, patients had to have drunk at least 50 g of alcohol per day for the past year or to have cirrhosis.

Seventeen centers included 802 patients, 20% had histologically proven cirrhosis or probable cirrhosis. Thirty-five percent had undergone liver biopsy. Twenty five percent of these patients had cirrhosis without acute alcoholic hepatitis and 37% had cirrhosis with acute alcoholic hepatitis. After dividing France along a Bordeaux-Strasbourg axis, there was more histologically proven or probable cirrhosis in the North (46%) than in the South (36%) (P<0.005) while daily alcohol intake was greater the South (150 +/- 6 g) than in the North (129 +/- 4 g) (P<0.0001). When the six variables (age, sex, daily consumption of alcohol over the past 5 years, presence of hepatitis B surface antigen and antibodies to hepatitis C virus, total duration of alcohol abuse) were considered together in stepwise logistic regression analysis, geographic location changed the prediction of cirrhosis. The odds ratio for cirrhosis in patients living to the North of the Bordeaux-Strasbourg axis was 1.9 (95% confidence interval range 1.1-3.2) (P<0.02), suggesting the role of nutritional factors.