Abstract
The disputed relation between recent alcohol consumption and stroke was examined in a community case control study. One hundred and twenty five incident first time stroke patients and 198 controls, aged 35 to 74 years, were recruited over two years from a general practice population. The age and sex adjusted relative risks for stroke by recent weekly "drinks" of alcohol were; 0-1.0 (reference), 1 to 14-0.57, 15 to 29-0.63, and > 29-0.99. Among the controls it was noted that non-drinkers were more likely than light/moderate drinkers (1 to 29 drinks per week) to have the following characteristics; history of obesity (p < 0.001), not a recent walker (p < 0.05), and no vigorous exercise in early adulthood (p < 0.01). The apparent association of light and moderate alcohol consumption with decreased stroke risk disappeared when these variables were included in the multiple risk factor adjusted analysis; 0-1.0, 1 to 14-0.88, 15 to 29-1.11, and > 29-1.23. The pattern for proved cerebral infarction (n = 81) was similar. The results of this study do not support the idea that recent heavy alcohol consumption is an important cause of either overall stroke or cerebral infarction. The association of non-drinking with a history of overweight and inactivity may explain the apparent protective effect of lighter alcohol consumption on the risks of both stroke and coronary heart disease.