risk of death from all causes was found with increasing alcohol consumption (Andreasson et al., 1988).

Although alcohol consumption is unlikely to reduce total mortality in people under 45 (Doll, 1997), the optimal duration of moderate alcohol consumption is not known in terms of reducing risk of cardiovascular disease mortality in older people. Furthermore, even though some of the benefits of alcohol are the result of long-term, habitual consumption (Jackson et al., 1992), many of the important effects of ethanol on high-density lipoproteins and clotting components are acute; thus, it is likely that alcohol consumption beginning in middle age would suffice while avoiding much of the risk of injuries and other external causes of death (although not necessarily of cancers or cirrhosis of the liver).

Optimal alcohol consumption differs for men and women for several reasons. Women metabolize alcohol less efficiently than men do (making women more prone to some health problems than are men who drink the same amount), and because women have less body water than men (making them more prone to intoxication than men after drinking the same amount of alcohol) (USDHHS, 2000). Women also have lower age-specific risks of cardiovascular disease and greater susceptibility to liver damage than men, and women are prone to a relatively high risk of breast cancer, which appears to increase with consumption of any amount of alcohol (Smith-Warner et al., 1998). Although men might be at risk for alcohol-related problems if they consume more than 14 drinks per week or more than 4 drinks per occasion, women could be at risk if they consume more than 7 drinks per week or more than 3 per occasion (USDHHS, 1995, 2000).

The problem of alcohol consumption is frequently one of maldistribution, with many abstaining and many consuming at a hazardous level (Holman and English, 1996). There seems to be no precedent for a public health campaign that simultaneously seeks to “pull in” both tails of a risk factor distribution, in this case reduction of both the prevalence of abstention and of heavy drinking (Holman and English, 1996). There are risks in promoting a population wide alcohol policy that discourages abstention, even if the policy encourages only light-to-moderate regular consumption. First, there is no evidence that moderate drinking is harmless. Second, a public health recommendation that encourages even light drinking over abstention could increase the number of heavy drinkers in a population since it has been noted that population distributions of risk factors tend to shift, either downward or upward, as an entity (Rose, 1992).

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement