Aims: The present study investigates the combined effect of average volume and binge drinking in predicting alcohol-related social problems and estimates the proportion of alcohol-related harms related to specific drinking patterns that could be prevented if transferred to a low-risk drinking group.
Methods: Data came from the 1997 and 2000 German Epidemiological Survey of Substance Abuse (ESA) (age: 18-59 years; response rate: 65% and 51%, respectively). The pooled sample consisted of 12,668 current drinkers. By using nine categories of average daily intake and three groups of binge drinking, individuals were grouped into 22 mutual exclusive groups. Social problems were defined as the occurrence of 'repeated family quarrels', 'concern of family members or friends', 'loss of partner or friend' or 'physical fight or injury' in relation to alcohol.
Results: The effect of average daily intake is modified by binge drinking frequency such that the association was strongest in those with four or more binge drinking occasions during the last 30 days. Within each binge drinking group, adjusted relative risks (aRR) increased with alcohol intake up to a certain threshold and decreased thereafter. Overall, compared to the reference group (<or=7 g ethanol/day-no binge), the population-attributable fraction (PAF) related to the other drinking groups was 71.4% (95% CI: 64.4-77.1%).
Conclusions: The frequency of binge drinking occasions seems to be a better predictor of alcohol-related social problems than volume. Alcohol-related social harms especially among drinkers with moderate volume per day may be reduced by targeting prevention strategies towards episodic heavy drinkers.