摘要：Background : The inverse relationship between low to moderate alcohol consumption and several favorable health outcomes has been well established in many epidemiological studies and meta-analyses. However, several questions still remain controversial. Aims : To discuss a number of open questions relating to the healthy effect of a moderate intake of alcohol (especially wine) on cardiovascular disease and total mortality. This will be based on findings from the literature, with a particular emphasis on meta-analyses. Results and Conclusion : The role of different alcoholic beverages, age and sex, confounding, former drinkers and study design has been discussed. Whether wine is better than beer or spirits, though suggestive, remains to be established. Cardiovascular morbidity and total mortality is significantly reduced both in men and women who are regular drinkers of low amounts of alcohol; however, the predicted protection in women disappears at lower doses than in men. The primary protection of alcohol decreases after adjustment for known variables, thus confirming the importance of confounding in assessing drinking effects, but it remains significant and of undoubted public health value. As the cardiovascular protection by moderate alcohol consumption might have been unduly overestimated by inclusion in control groups of former drinkers, we compared studies that used as a reference group the category of no alcohol intake and/or formally excluded former drinkers with studies which did not: the protection was indeed somewhat lower in the former than in the latter studies, but was still statistically significant. We conclude that the dose-response relationship between alcohol intake and cardiovascular risk or total mortality, consistently described by J-shaped curves, can be reasonably attributed to a combination of both real beneficial (at lower doses) and harmful (at higher doses) effects of alcohol or wine consumption.
