Résumé·s
Background and Aims: The Pathways Model of Problem Gambling suggests the existence of
three developmental pathways to problem gambling, each differentiated by a set of predisposing
biopsychosocial characteristics: Behaviorally Conditioned (BC), Emotionally Vulnerable (EV),
and Biologically Vulnerable (BV) gamblers. This study examined the empirical validity of the
Pathways Model among adolescents followed up to early adulthood. Design: A prospectivelongitudinal design was used, thus overcoming limitations of past studies that used concurrent or
retrospective designs. Setting: Two samples were used: a) a population sample of French-speaking
adolescents (N = 1,033) living in low socio-economic status (SES) neighborhoods from the Greater
Region of Montreal (Quebec, Canada), and b) a population sample of adolescents (N = 3,017),
representative of French-speaking students in Quebec. Participants: Only participants with at-risk
or problem gambling by mid-adolescence or early adulthood were included in the main analysis
(n = 180). Analyses: Latent Profile Analyses were conducted to identify the optimal number of
profiles, in accordance with participants’ scores on a set of variables prescribed by the Pathways
Model and measured during early adolescence: depression, anxiety, impulsivity, hyperactivity,
antisocial/aggressive behavior, and drug problems. Findings: A four-profile model fit the data
best. Three profiles differed from each other in ways consistent with the Pathways Model (i.e.,
BC, EV, and BV gamblers). A fourth profile emerged, resembling a combination of EV and BV
gamblers. Conclusions: Four profiles of at-risk and problem gamblers were identified. Three of
these profiles closely resemble those suggested by the Pathways Model.