Background and purpose: Although a variety of epidemiological studies have suggested a U-shaped association between alcohol and cardiovascular disease, controversy still surrounds the role of atherogenesis in the mediation of alcohol effects.
Methods: Carotid atherosclerosis was measured with a sensitive and reproducible B-mode score in a random sample of 460 men drawn from the Bruneck Study (baseline examination in 1990).
Results: The age-adjusted relation between alcohol and carotid artery disease was U shaped, with light drinkers facing a lower atherosclerosis risk (odds ratio, 0.44; 95% confidence interval, 0.23 to 0.85; P = .01) than either abstainers (odds ratio, 1.00) or heavy drinkers (odds ratio, 2.78; 95% confidence interval, 1.32 to 5.84; P < .01). The association was not explained by the lifestyle of alcohol consumers (smoking) or inclusion of former (heavy) drinkers in the reference group. The effect of alcohol was modified by drinking behavior (type of beverage). Approximately a quarter of the atherosclerosis risk caused by severe alcohol consumption was mediated by the risk profile associated with drinking, whereas the apparent beneficial effect of low alcohol intake emerged independent of conventional risk attributes.
Conclusions: Our results support the hypothesis that adverse and beneficial effects of alcohol on cerebrovascular disease are mediated in part by analogous atherogenic and antiatherogenic properties.