A population-based study of the association between pathological gambling and attempted suicide.

The association between pathological gambling and attempted suicide was examined using data from a prevalence study conducted in Edmonton, Alberta, Canada. The sample size was 7,214, the questionnaire was the Diagnostic Interview Schedule, and diagnoses were made on a lifetime basis according to DSM-III criteria. Logistic regression analysis was performed, with attempted suicide as the dependent variable. The odds ratio for pathological gambling was statistically significant (odds ratio = 4.91; 95% confidence interval = [1.41,17.1]) when major depression was the only comorbid mental disorder in the model. As terms for additional mental disorders were included, pathological gambling ceased to be statistically significant. It was concluded that a history of pathological gambling is associated with previous attempted suicide, and that the association may be due to a common factor--"mental illness."