Alcohol consumption and blood pressure: analysis of data from the Canada Health Survey

Recent epidemiologic studies report a significant association between alcohol consumption and elevations in both systolic (SBP) and diastolic (DBP) blood pressures. To test this hypothesis, we conducted a multivariate analysis of physical examination and other data on 721 men and 697 women aged 20 or more collected during the Canada Health Survey in 1978-1979. SBP and DBP were considered as separate dependent variables in multiple regression models with the following independent variables: age, alcohol consumption (measured as a 7-day recall history and as an average frequency of consumption), serum cholesterol, plasma glucose, physical activity, Quetelet index, parental history of hypertension, cigarette consumption, income, education, and exogenous hormonal use in women. In both weighted and unweighted multiple regression analyses, we could not demonstrate for either sex, a significant association between alcohol consumption (as recorded and following quadratic and logarithmic transformations) and either SBP or DBP. For both sexes, only age and Quetelet index were highly significantly (P less than 0.0001) and consistently associated with both SBP and DBP. No other independent variables were consistently associated, for either sex, with SBP and DBP. Further, the dose-response patterns noted by other investigators suggesting either a positive and linear relationship or a curvilinear relationship were not found in either our univariate or multivariate analyses. Rather, the alcohol-blood pressure curves showed no consistent patterns of any kind in either sex. These findings do not support recent claims that alcohol consumption is a determinant of elevations in either SBP or DBP.