[Influence of cannabis use on suicidal ideations among 491 high-school students]

The aim of the study was to evaluate the relative contributions of cannabis use and depressive symptoms in the prediction of suicidal ideation.

A random sample of 491 high-school students from the department of Haute-Garonne, France (290 girls, mean age=16.6+/-1.4; 201 boys, mean age=17.3 years+/-1.6) completed a questionnaire assessing cannabis use, the CES-D (center for epidemiological studies-depression scale) completed by the three-item sub-scale measuring suicidal ideation proposed by Garrison et al. (J Am Acad Child Adolesc Psychiatry 30 (1991) 636-641).

In the total sample, 15% of girls (n=44) and 23% of boys (n=46) reported having used cannabis at least once a month during the last three months. The mean CES-D score for girls was significantly higher than for boys (20.6+/-11 versus 18.4+/-10.9, t=2.4, p=0.01). According to the cut-off score of 24, 22% of boys and 36% of girls exhibited a moderate to severe depressive symptomatology (p=0.02). The mean suicidal ideation score was not significantly higher in girls than in boys (1.3+/-2 versus 0.95+/-1.9, t=1.7, p=0.10). At least, the occasional wish to kill oneself was reported by 17% of boys and 20% of girls (p=0.40). The contribution of cannabis use and depressive symptomatology in the prediction of suicidal ideation were tested with a hierarchical multiple regression analysis, to determine whether cannabis use improved prediction of suicidal ideation beyond that afforded by CES-D scores. In the first step, including age and sex as covariates, the model accounted for a significant and important part of the variance of suicidal ideation (41%). Age and sex were not significant predictors. The second step, with cannabis use added to the prediction of suicidal ideation, afforded an increase of 1% in predictable variance. Cannabis use (beta=0.08, t=2.2, p=0.03) was significantly but weakly related to suicidal ideation, while depressive symptoms (beta=0.63, t=17.8, p<0.001) remained the main predictor. In a second regression analysis conducted among cannabis users, frequency of use was almost a significant predictor (beta=0.11, t=1.84, p=0.07).

This result suggests an association between suicidal ideation, depressive symptoms and cannabis use in adolescents. Cannabis use appeared to be an independent predictor of suicidal ideation after controlling for depressive symptoms. However, the increment of variance accounted for by cannabis use was small. This suggests that cannabis use contributes to suicidal ideation independently from depressive symptoms, but weakly. A limitation to the study was the absence of control for other potential confounding variables. The association between cannabis use and suicidal ideation may be linked to common risk factors such as borderline personality disorder traits, sociodemographic and family factors. In this study, cannabis use does not appear as an important risk factor for suicidal ideation in adolescents.