Predictors of problem gambling remission in adults: A Canadian longitudinal study.

Abstract

Objective: Remission from problem gambling (PG) continues to be a priority of clinicians and researchers. Data from cross-sectional studies indicate that some correlates are more predictive of PG, and existing longitudinal studies have exclusively examined risk factors that predict emergence of PG. This study’s objective is to fill in the remaining pieces of the puzzle by identifying factors that might facilitate remission from PG. Method: A stratified sample of 10,199 Canadian adult gamblers were recruited from an online panel. Respondents who screened positively for PG at baseline and completed a follow-up assessment 1 year later (n = 468) were assessed on a series of modifiable gambling, psychosocial, mental health, and substance use variables. A forward stepwise logistic regression was conducted to identify the strongest predictors of remission from PG at follow-up. A Least Absolute Shrinkage and Selection Operator regression was also conducted to confirm the most relevant predictors. Results: Out of 75 candidate variables, 10 were retained by the regression model. Two were related to cessation of specific gambling activities, two were related to gambling motivations, two were psychosocial in nature, two were related to substance use while gambling, and one was related to remission from a mental health disorder. The final and strongest predictor was PG severity at baseline. Conclusions: Although PG remission predictors were mostly gambling-related, psychosocial aspects may also be targeted by stakeholders aiming to reduce PG. Ceasing to use tobacco while gambling and diversifying leisure activities may be promising targets. Other mental health and substance use predictors may still possibly be relevant, but only for a subset of people with PG.