PURPOSE: : This study was designed to empirically derive latent classes based on PG criteria and to assess the association between nongambling psychiatric disorders and specific classes. METHODS: : A total of 8138 community-based middle-aged men were surveyed, and 2720 were assessed for Diagnostic Interview Schedule, Version 3, Revised (DIS-III-R) pathologic gambling (PG). Latent class analysis (LCA) was applied to Diagnostic and Statistical Manual Version 3, Revised (DSM-III-R) criteria to identify gambling classes. χ and logistic regression models evaluated the association between gambling classes and lifetime psychiatric disorders. RESULTS: : The final model included 4 classes: class 0 (ie, 5418 individuals who never gambled 25 or more times per year) and classes 1-3 (identified by the LCA and comprising 2720 respondents assessed for PG). For the 9 individual criteria of PG, endorsement percentages ranged from 2%-6%, 4%-58%, and 53%-100% for classes 1-3, respectively. Nongambling psychiatric disorders were differentially associated with the 4 gambling classes, and psychopathology was more common in groups more frequently acknowledging PG criteria. CONCLUSIONS: : Empirical support is provided for distinct classes of gambling behaviors demonstrating differential associations with individual PG criteria and nongambling psychiatric disorders. The data-driven categorization of gambling behaviors provides direction for research on defining, preventing, and treating syndromal and subsyndromal PG.
PURPOSE: : This study was designed to empirically derive latent classes based on PG criteria and to assess the association between nongambling psychiatric disorders and specific classes. METHODS: : A total of 8138 community-based middle-aged men were surveyed, and 2720 were assessed for Diagnostic Interview Schedule, Version 3, Revised (DIS-III-R) pathologic gambling (PG). Latent class analysis (LCA) was applied to Diagnostic and Statistical Manual Version 3, Revised (DSM-III-R) criteria to identify gambling classes. χ and logistic regression models evaluated the association between gambling classes and lifetime psychiatric disorders. RESULTS: : The final model included 4 classes: class 0 (ie, 5418 individuals who never gambled 25 or more times per year) and classes 1-3 (identified by the LCA and comprising 2720 respondents assessed for PG). For the 9 individual criteria of PG, endorsement percentages ranged from 2%-6%, 4%-58%, and 53%-100% for classes 1-3, respectively. Nongambling psychiatric disorders were differentially associated with the 4 gambling classes, and psychopathology was more common in groups more frequently acknowledging PG criteria. CONCLUSIONS: : Empirical support is provided for distinct classes of gambling behaviors demonstrating differential associations with individual PG criteria and nongambling psychiatric disorders. The data-driven categorization of gambling behaviors provides direction for research on defining, preventing, and treating syndromal and subsyndromal PG.
Authors: Grace Kong; Jack Tsai; Suchitra Krishnan-Sarin; Dana A Cavallo; Rani A Hoff; Marvin A Steinberg; Loreen Rugle; Marc N Potenza Journal: J Addict Med Date: 2014 Nov-Dec Impact factor: 3.702