The interplay of alcohol use symptoms and sociodemographic factors in the HELIUS study: A network perspective

Read the full article

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Purpose: Research on alcohol use disorder has exclusively focused on either its symptom-level mechanisms--the network perspective-- or sociodemographic determinants--epidemiological research. Moreover, such research failed to stratify analyses for important person-level factors (e.g., sex or ethnicity). Here, we combine network and epidemiological research and stratify analyses by person-level factors. Method: Using Bayesian inference, we estimated (1) a logistic regression model predicting past-year alcohol consumption from various sociodemographic factors within a large, multiethnic, urban sample in the Netherlands (complete sample: N = 22,164), (2) a cross-sectional network model of alcohol use symptoms and sociodemographic factors among alcohol drinkers of the same sample (drinkers: N = 10,877), and (3) stratified networks at the sex- and ethnic- levels in the same drinkers subsample. Results: All of our examined sociodemographic factors predicted past-year alcohol consumption (in order of magnitude: religion, sex, education, employment, perceived ethnic discrimination, and age). Our Bayesian analysis of networks revealed three notable patterns. First, religion was uniquely and negatively related to adverse alcohol use problems (such as having an injury due to drinking). Second, socioeconomic proxies (education and employment) were positively related to binge drinking, but negatively related to its adverse effects (such as 'needing a drink in the morning'). Finally, employment and education were particularly negatively related to alcohol use symptoms within male and female networks, respectively. Conclusion: Our results suggest that alcohol use symptoms are differentially related to sociodemographic factors and that these effects are moderated by sex and ethnicity. Our highlighted network links and Bayesian methodologies could prove useful for future research and prevention and intervention efforts on alcohol use disorders.

Article activity feed