Although we appreciate Tony George's Public Health piece on problem drinking,1 there is an underlying assumption in the article that most of the disease burden of alcohol consumption is associated with alcohol use disorders, and this is not the case. In 2002, the most common cause of deaths attributed to alcohol consumption in Canada2,3 and globally4 was drinking while engaging in another activity, such as driving (38.1% of alcohol-attributable deaths in Canada).2 Liver cirrhosis, a condition that is often associated with alcohol use disorders, ranked third (15.4%); alcohol-attributable cancer ranked second (22.6%).2 In fact, many alcohol-attributable deaths from cancer occur in people who do not have an alcohol use disorder; for instance, the risk of developing breast cancer increases with alcohol consumption levels as low as 1 drink per day.5
It should be noted that these detrimental effects of alcohol consumption are far more significant than its cardioprotective or other beneficial effects.3 Alcohol also has important effects on people other than the drinker, such as newborns whose mothers drank while pregnant or bystanders who are struck by drunk drivers.
To reduce the disease burden of alcohol consumption, the most effective and cost-effective measures are not individual interventions, but population-level policy measures such as increasing the taxes levied on alcohol purchases or lowering the blood-alcohol concentration legally permitted for driving to 0.05%6; a recent national working group on alcohol in Canada has called for both of these measures.7 Brief individual interventions cost significantly more than taxation initiatives to generate the same reduction in alcohol-attributable harm.6 Therefore, although targeting problem drinkers who show up in their physician's office or a hospital emergency department is critical, changes in legislation may be more useful in terms of lives (and dollars) saved.