The choice of indicators influences conclusions about the educational gradient of sex-specific alcohol consumption

Abstract

There has been considerable public interest in reports on harmful alcohol consumption of higher educated females. This study assesses the robustness of this finding with representative German data using ten different indicators of alcohol consumption. This cross-sectional study used data of the Epidemiological Survey on Substance Abuse from 2012. 4,225 females and 3,239 males represent the German population aged 18-64. It presents ten indicators of alcohol consumption by sex and education and provides group specific means and 95 %-confidence intervals. The main results are: (1) Higher educated males and females are drinking alcohol more frequently than lower educated males and females. (2) When drinking, higher educated males and females tend to drink less alcohol than lower educated males and females. (3) Only when using an indicator for hazardous alcohol consumption with different thresholds for males and females, the results indicate a pattern that significantly exposes hazardous alcohol consumption in the group of higher educated females. Concerning the choice of indicators, this study shows that sex-specific threshold-based indicators of alcohol consumption may lead to different conclusions as the majority of other indicators.