Sociodemographic Factors and Alcohol Use Symptoms in the HELIUS study: A Bayesian Analysis of Networks

Background: The development of problematic alcohol use is argued to result from a bidirectional interplay between a variety of biopsychosocial factors--commonly termed network perspective. Epidemiological research has highlighted the crucial role of sociodemographic variables in the onset of alcohol use and alcohol dependence; furthermore, the network perspective has provided insights into underlying mechanisms leading to alcohol dependence. What remains unclear is how sociodemographic variables tie into the mechanisms of alcohol use disorder symptoms. Method: We conducted an exploratory network analysis in a Bayesian framework on individuals consuming alcohol in a multiethnic, urban sample in the Netherlands, the HELIUS study (full-sample: N = 22,164; drinkers: N = 10,877). The Bayesian approach allows for improved inference by providing a principled way to quantify the uncertainty associated with the estimated networks. Results: Results showed that sociodemographic factors were differentially associated with alcohol use disorder symptoms amongst drinkers, with religion being negatively associated with binge drinking and alcohol-related adverse events (such as guilt). Furthermore, employment was strongly interconnected with AUD symptoms for males but not for females. Different ethnic groups showed differential importance of socio-demographic determinants for alcohol use disorder symptoms; however, despite the exceptionally large sample size, the networks still showed insufficient evidence to robustly interpret edge presence and absence. Conclusion: The current study highlights the need to apply the Bayesian framework in network inference to determine robust inferences. Moreover, considering the trend towards increasingly diverse urban areas, our results point towards socio-demographic factors that may contribute to specific alcohol use disorder symptoms.