The recent study of Zhou et al showed negative effects of alcohol consumption on the risk for coronary artery disease. Despite a large number of observational studies showing protective effects of moderate alcohol consumption on heart disease, we consider the findings of Zhou et al to be valid. Apparent cardioprotective effects of moderate alcohol consumption may well be an artifact of uncontrolled confounding, similar to the misleading observation from epidemiologic studies, now known to be an aberration of methodologic bias, that female hormone replacement therapy is protective against heart disease. Moderate drinking has been found to be associated with clusters of factors (confounders), all closely associated with socioeconomic status, that are protective against coronary heart diseases.

China is a country very different from the developed world. Although the number of coronary arteriographic studies conducted in China has been increasing rapidly, the prevalence of coronary heart disease in the Chinese population is still much higher than the prevalence of coronary arteriography. This is partly due to the relatively high cost of the procedure, and in China, health insurance does not cover a substantial part of the population. In China, those who have the opportunity to undergo coronary arteriography probably belong to a relatively high socioeconomic class, although they may be biologically similar to the general population at risk for coronary heart disease.

The fact that the patients and controls included in Zhou et al’s study probably came from similar socioeconomic strata provides a great advantage in measuring the real effect of alcohol consumption on coronary artery disease, as it would largely eliminate the confounding effects of socioeconomic status.