Convergent results of two methods of estimating the prevalence of problem drinking

Two studies were conducted using different methods of estimating the prevalence of problem drinking among the Canadian Armed Forces. In one case, data on alcohol sales from base outlets were used to calculate mean per capita consumption, which in turn was applied to a theoretical and, to some extent, empirically-validated plot of the distribution of consumption often referred to as the Ledermann curve. Criteria for different levels of problem drinking were then applied to this distribution. In the other case, estimates of problem drinking and presumptive problem drinking were derived from self-reported data on the correlates and sequelae of excessive drinking. The results derived from the two methods were highly convergent. As in the case of studies from the United States, the estimates were disturbingly high, both placing approximately 18% of the respective populations in he hazardous drinking category. The preventive and treatment-related implications of these findings are discussed.