Heavy drinking is a composite phenomenon being operationalized in vari
ous and at times misleading ways. With the ultimate goal of comparing
the magnitude of the association between the studied patterns of heavy
drinking and alcohol-related problems as they apply to public health
policy, this project involved the comparison of two criteria used to d
efine 'heavy drinking' in a general health survey that was carried out
in the Province of Quebec (Canada). A weekly volume of 29 drinks or m
ore (heavy weekly volume), an intake of 8 drinks or more within the sa
me day (binge drinking) and an 'alcohol-related problems' measure base
d on the CAGE questionnaire and some of the items of the DSM-III (prob
lem drinking) were used as indicators. The major finding of this study
was that, controlling for sex, education, psychological distress, pre
vious hospitalisation for alcohol abuse and previous suicide attempts,
the weekly volume of alcohol consumed is a better predictor of alcoho
l-related problems than is daily consumption. The results are discusse
d with respect to the implication on policy and prevention programs.