The Leisure, Lifestyle, & Lifecycle Project (LLLP): A Longitudinal Study of Gambling in Alberta. Final Report for the Alberta Gambling Research Institute

Abstract

The Leisure, Lifestyle, and Lifecycle Project (LLLP) is a five-year prospective longitudinal study designed to collect data on the factors influencing change in gambling and problem gambling behavior over time. A sample of 1808 participants from four locations representing the diversity of the province of Alberta (Edmonton, Calgary, Lethbridge area, and Grand Prairie area) were recruited primarily through random digit dialing. In order to assess the development of gambling problems over the lifespan, five critical age ranges were targeted: 13-15, 18-20, 23-25, 43-45 and 63-65 year-olds. Individuals with relatively heavy involvement with gambling were over sampled. A broad array of psychosocial variables was assessed at baseline via telephone, face-to-face and computer self-completion interviews. The sample was weighted to match the population of Alberta according to age, gender, geographic location and the over sampling procedure. The three follow-up interviews of the cohort were completed by paper- or Internet-based surveys. Retention in the fourth and final assessment was 76.2% for the adult cohorts, 71.8% for the adolescent cohort, and 75.1% for the combined cohort. Three primary questions directed this project: 1. What are the normal patterns of continuity and discontinuity in gambling and problem gambling behaviour? 2. What biopsychosocial variables and behaviour patterns are most predictive of current and future problem gambling? 3. What etiological model of problem gambling is best supported by the longitudinal findings? This report provides analyses of the adult sample and focuses primarily on the first two of the primary research questions above - specifically, on identifying variables that are robust predictors of future problem gambling onset, the stability of gambling problems over time, and the development of a multivariate model that illustrates the interaction of gambling behaviour and problem gambling over time. A tentative etiological model is also presented to address the last research question. The LLLP sample problem gambler prevalence at wave 1 was 4.7% (weighted prevalence 3.2%). A similar longitudinal study was conducted during the same time period in Ontario, namely the Quinte Longitudinal Study. A set of parallel analyses was conducted on the QLS dataset to identify findings that were robustly supported in both studies. The collective findings of the 8 LLLP and QLS studies represent the most comprehensive longitudinal analysis of gambling and problem gambling currently in the literature.