A Longitudinal Investigation of Lower-Risk Gambling Limits in the Canadian National Study

Recent years have seen increased attention given to identifying and describing the levels of gambling participation that confer a risk of harm in order to generate public health advice regarding lower-risk gambling. However, most of the existing literature has failed to explicitly assess these limits in a prospective manner. The purpose of this study is to employ a methodology consistent with prior investigations to evaluate the level of gambling participation associated with an increased risk of future gambling-related harm. Using data from the Alberta Gambling Research Institute's National Project Online Panel Survey, risk ratios and Receiver Operating Characteristic (ROC) analyses were used to determine the relative risk of gambling-related harm associated with participating in a greater number of gambling formats, gambling more days per month, and spending a greater proportion of income gambling. Prospective lower-risk limits were largely consistent with those identified in previous cross-sectional analyses (e.g., no more than two gambling formats, no more than once a week), with the exception that higher limits were found for the percent of household income spent gambling (3.4-6.4% vs. 1%). We advise that future research on lower-risk gambling limits consider the use of more granular assessment instruments and prospective methods to more closely evaluate the association between gambling participation and gambling harm.

Are There Riskier Types of Gambling?

Gambling-related harm is a public health issue requiring market regulation and efforts aimed at prevention and treatment. An important consideration for the regulation of gambling is whether certain types of gambling are intrinsically more harmful than others. The present study was a comprehensive investigation of this issue in a nationwide sample of 10,199 Canadian adult gamblers that included 1346 individuals with problem gambling. We investigated (a) the univariate cross-sectional association between individual types of gambling and problem gambling; (b) the cross-sectional association between individual gambling types and problem gambling when controlling for breadth of gambling involvement; (c) the prospective/lagged relationship between participation in different gambling types and future problem gambling; and (d) the self-reports of people with gambling problems concerning the types and modalities they consider to be most problematic. Our collective results indicate that breadth of gambling involvement is a stronger predictor of gambling problems than involvement in any particular type, but that involvement in certain types (electronic gambling machines in particular, and casino table games and online gambling to a lesser extent) does confer additional risk.