A community case-control study of alcohol consumption in stroke.

This study examines the hypothesis that there may be a clinically important association between alcohol intake and stroke. Alcohol consumption and the haematological and biochemical markers of alcohol intake were studied in hospital admissions for stroke and compared with community based control subjects from an occupational screening survey. In males, moderate to heavy alcohol consumption (greater than 30 units per week) was associated with an increased relative risk of stroke. Light drinking (less than 30 units per week) was associated with reduced relative risk when compared to teetotallers. The relative risk of stroke in moderate and heavy consumers of alcohol compared with teetotallers was elevated 1.8 times. Similar patterns of risk were present for increasing levels of aspartate transaminase and uric acid. Relative risk was increased for all levels of gamma-glutamyl transferase above the lowest. There was a decrease in relative risk associated with increasing levels of mean erythrocyte cell volume though this did not achieve statistical significance. There were few heavy drinkers among the female cases or controls. We conclude that high alcohol intake may be a significant preventable risk factor particularly among male strokes.