Alcohol Consumption and Ischemic Heart Disease: Some Evidence from Population Studies *

Summary Spatial correlations for a series of countries in each of three years, and for the states of the U.S.A. in one year, confirmed previous reports of a negative association between mortality from ischemic heart disease (IHD) and per capita alcohol consumption, and of a positive association between the latter and liver cirrhosis mortality. A temporal analysis of the same variable for each of 19 countries gave results consistent with the well‐established consumption‐cirrhosis relationship but inconsistent with the postulated ‘protective’ effect of alcohol consumption on IHD. For consumption and cirrhosis there were 18 highly significant positive coefficients, while for IHD and consumption, 13 were positive and 6 negative, and 8 out of 19 were not significant. These results did not appear to be attributable either to the absence of adjustment for the effect of the 8th revision of the International Classification of Diseases or to the lack of a time lag in the consumption data. The difference in the findings of the spatial and temporal analyses suggests that the association between IHD mortality and alcohol consumption in general populations is dependent upon variation in a third as yet unidentified factor. Such a factor would be expected to correlate with alcohol consumption cross‐nationally but remain more or less constant through time in any given country. Certain food habits appear to have this characteristic.