Alcoholic and non-alcoholic liver disease in relation to alcohol consumption in Scotland, 1978-84. Part I: Epidemiology of liver diseases.

A common strategy in epidemiological studies linking alcohol consumption in the general population with liver cirrhosis mortality is to use non-specific cirrhosis mortality rates in which alcoholic and non-alcoholic causes of death are not distinguished. Evidence is presented from Scottish mortality data for 1979 to 1984 that the two forms of cirrhosis have quite different epidemiological profiles. Similar findings emerge from morbidity data. The two forms of disease should be distinguished in future studies in which liver cirrhosis is used as a proxy for consumption, despite the manifest shortcomings of currently available data.