Abstract: Gambling and Polysubstance Use Behavior Patterns: Using Latent Class Analysis to Examine the Syndromal Model of Addiction (Society for Prevention Research 24th Annual Meeting)

270 Gambling and Polysubstance Use Behavior Patterns: Using Latent Class Analysis to Examine the Syndromal Model of Addiction

Schedule:
Wednesday, June 1, 2016
Pacific D/L (Hyatt Regency San Francisco)
* noted as presenting author
Bethany C. Bray, PhD, Research Assistant Professor, The Pennsylvania State University, University Park, PA
The “syndromal model of addiction” posits a single underling condition that leads to gambling disorder and to substance use disorders. A central tenant of this theory is that individuals with the condition have unique patterns of interacting behavioral expressions across both domains. In turn, these behavior patterns lead to a variety of direct consequences including gambling disorder, substance use disorders, and negative health conditions. Latent class analysis (LCA) is a methodological technique ideally suited to identifying subgroups of individuals with unique behavior patterns across multiple domains. Recent work on LCA has improved approaches to linking behavior patterns identified with LCA to later consequences (i.e., distal outcomes), making this methodological technique a natural way to investigate addiction syndrome empirically. To date, however, cross-cutting LCA models of gambling and polysubstance use have not been examined; this study examines comprehensive models of these behaviors in order to characterize early potential manifestations of addiction syndrome in a national sample of emerging adults.

Analyses included 1892 participants (mean=22 years; 40% female; 80% White) from the Add Health public use sample who provided data at Wave 3. Based on the theoretical model, indicators included past 30 day use of alcohol, cigarettes, chewing tobacco, marijuana, cocaine, methamphetamine, injection drugs, and other drugs; past 14 day binge drinking; and lifetime gambling expenditures and problems with gambling. LCA was used to identify underlying subgroups of individuals characterized by unique behavior patterns of gambling and polysubstance use. Preliminary results identified 7 latent classes: Alcohol Users (6% prevalence), Binge Drinkers (23%), Cigarette Smokers (29%), Marijuana Smokers (19%), Chewing Tobacco Users (8%), Multi-Substance Users (8%), and Gamblers (6%). Grouping variables for gender and race/ethnicity were included to examine health disparities in subgroup membership, and membership was linked to several adult health outcomes at Wave 4, including gambling disorder, financial problems, and substance use disorders.

This study is important for prevention research because it tests empirically new theory on addiction etiology, which may lead to more effective, targeted programs. Determining whether and for whom gambling co-occurs with substance use is critical for designing programs that can address both behaviors simultaneously. Results suggested considerable specialization in behavior; only 1 latent class was characterized by use of more than one substance, and only 1 latent class was characterized by gambling, whose members tended not to use substances. Detailed results will be presented and implications for prevention discussed.