Modeling alcohol use disorder as a set of interconnected symptoms – Assessing differences between clinical and population samples and across external factors. (February 2022)

• Record Type: Journal Article
• Authors: Huth, K.B.S.; Luigjes, J.; Marsman, M.; Goudriaan, A.E.; van Holst, R.J.
• Is Part Of: Addictive behaviors. Volume 125(2022)
• Publication Date: 2022-02
• Subjects: Alcohol use disorder -- Network analysis -- Clinical and population sample -- Loss of control -- Measurement invariance -- Bayesian analysis; Substance abuse -- Periodicals; Alcoholism -- Periodicals; Drug addiction -- Periodicals; Nicotine addiction -- Periodicals; Smoking -- Periodicals; Gambling -- Psychological aspects -- Periodicals; Electronic journals; 362.29
• DOI: 10.1016/j.addbeh.2021.107128
• Languages: English
• ISSNs: 0306-4603

Abstract

Highlights: Bayesian network analysis of alcohol use disorder in a clinical and population sample. In the population sample the time spent on alcohol is most strongly connected, in clinical sample rather the attempt to set limits. The clinical network suggests a sparser network structure, however, estimates are too uncertain to conclude the sparsity of the network. Strength of symptom associations depend on external factors like age, gender, ethnicity, and income in the population sample. Abstract: Alcohol use disorder is argued to be a highly complex disorder influenced by a multitude of factors on different levels. Common research approaches fail to capture this breadth of interconnecting symptoms. To address this gap in theoretical assumptions and methodological approaches, we used a network analysis to assess the interplay of alcohol use disorder symptoms. We applied the exploratory analysis to two US-datasets, a population sample with 23, 591 individuals and a clinical sample with 483 individuals seeking treatment for alcohol use disorder. Using a Bayesian framework, we first investigated differences between the clinical and population sample looking at the symptom interactions and underlying structure space. In the population sample the time spent drinking alcohol was most strongly connected, whereas in the clinical sample loss of control showed most connections. Furthermore, the clinical sample demonstrated less connections, however, estimates were too unstable to conclude the sparsity of the network. Second, for the population sample we assessed whether the network was measurement invariant across external factors like age, gender, ethnicity and income. The network differed across all factors, especially for age subgroups, indicating that subgroup specific networks should be considered when deriving implications for theory building or intervention planning. Our findings corroborate known theories of alcohol use disorder stating loss of control as a central symptom in alcohol dependent individuals.