Not too long ago, Time Magazine published an article with the fairly shocking headline "Why Do Heavy Drinkers Outlive Nondrinkers?" One of the first questions which should come to mind when one sees this is whether there is a problem with the definition of heavy drinking, or, are the limits which distinguish moderate drinking from heavy drinking set too low? Another recent article in the Daily Mail has the headline "Heavy Drinking 'Kills You Quicker than Smoking' and Poses Greater Risk to Women than Men." Rather than asking which statement is true, we should rather be asking whether these two articles are using the term "heavy drinking" to refer to the same amount of alcohol consumption. In point of fact, the definitions of heavy drinking used in these two articles are extremely different, with the latter article referring to DSM IV Alcohol Dependence as "heavy drinking."
The study which found that heavy drinkers live longer than abstainers was conducted by Charles J. Holahan et al. and published in the journal Alcoholism: Clinical and Experimental Research. In this study, moderate drinking was defined as less than 21 US standard drinks per week (294 grams of ethanol) for both men and women regardless of gender. Heavy drinking was defined as 21 or more US standard drinks (294 grams of ethanol) per week. It should be noted that this definition of the cut off point for moderate drinking is far higher than the US government's definitions of moderate drinking limits. The NIAAA defines heavy drinking for men as more than 14 US standard drinks (98 grams of ethanol) per week and for women heavy drinking is defined as more than 7 standard drinks (196 grams of ethanol) per week. The cut off level between moderate and heavy drinking in the Holahan article was 3 times greater than the U.S. government definition of the cut off level for women, and one and a half times greater than that for men. Yet in spite of these far more liberal definitions of moderation and heavy drinking, the heavy drinkers in the Holahan study still outlived abstainers.
Some critics of studies which show that heavy drinkers outlive abstainers have pointed out a possible confound in that some of the members of the abstainer category may be ex-alcoholics who have quit drinking entirely. However, the Holahan study controlled for this factor by excluding ex-alcoholics and running the data for only lifelong abstainers. The result was that heavy drinkers still significantly outlived lifelong abstainers even when ex-alcoholics were excluded from the data set.
A study on the protective effect of alcohol on coronary heart disease published by Giovanni Corrao et al. sheds further light on the question of where the cut off between moderate and heavy drinking may be found. This mea-analysis found that the maximum protective effect of alcohol on the heart occurred at a rate of consumption of about 10 US standard drinks per week (140 grams per week; 20 grams per day; 1.43 drinks per day). The crossover point where alcohol consumption became equally as harmful for the heart as alcohol abstinence occurred at the point of 40 US standard drinks per week (560 grams per week; 80 grams per day; 5.7 drinks per day). However, since this study covered only effects on the heart and not other causes of mortality, this number is probably too high for the cut off point between moderate and heavy drinking when all possible effects of alcohol are taken into consideration.
The article referenced in the Daily Mail was actually about Alcohol Dependence rather than heavy drinking per se, and did not mention any specific drink numbers. This article, by Ulrich John et al. is titled "Excess Mortality of Alcohol-Dependent Individuals After 14 Years and Mortality Predictors Based on Treatment Participation and Severity of Alcohol Dependence" and was published in Alcoholism: Clinical and Experimental Research. This article found that people who met the DSM IV criteria for Alcohol Dependence had lifespans an average of 20 years shorter than those who did not. Alcoholism treatment had no apparent protective effect in preventing premature death. No drink numbers were discussed in this article, severity was estimated using the number of dependence criteria met.
Although the John article did not mention drink numbers in conjunction with Alcohol Dependence, a study by A. Jarque-Lopez et al. titled "Prevalence and Mortality of Heavy Drinkers in a General Medical Hospital Unit" and published in the journal Alcohol and Alcoholism quantified dependent drinking as an average of 63 US standard drinks per week (9 per day) for a woman and 77 US standard drinks per week (11 per day) for a man, in both cases over a period of about 20 years. The Jarque-Lopez study found that this level of consumption also resulted in death approximately 20 years sooner than in the non-heavy drinking comparison group.
So at what level does alcohol consumption become equally as dangerous as alcohol abstinence? It appears that the cut off point is somewhere between 20 and 40 US standard drinks per week. We will split the difference and say that it probably lies at around 30 US standard drinks (420 grams of ethanol) per week, a far cry from the puritanical US government limits of 7 for women and 14 for men. Current government limits may have far more to do with the politics of the addiction treatment lobby than any relation to scientific evidence.