An inverse or reduced risk association between moderate alcohol consumption and total mortality has been reported in most prospective epidemiologic studies, even after adjustments for all known potential confounders such as educational level, job and health.