[Relationship between depressive symptoms, hopelessness and suicidal ideation among 1547 high school students].

OBJECTIVES: The aim of the study was to evaluate the incidence of depressive symptoms and suicidal ideation, and to test the mediating role of hopelessness between depressive symptoms and the wish to kill oneself.

METHOD: A random sample of 1547 high school students from the department of Haute-Garonne, France, (854 girls, mean age=16.9+/-1.5; 693 boys, mean age=17.4+/-1.5) completed a questionnaire assessing cannabis use, the Center for Epidemiological Studies-Depression Scale (CES-D) completed by the three items subscale measuring suicidal ideation proposed by Garrison et al. (1991) ("I felt life was not worth living"; "I felt like hurting myself"; "I felt like killing myself"). The measure of hopelessness was based on a single item, "I felt life was not worth living".

RESULTS: At least occasional wish to kill oneself were reported by 13% of boys and 14% of girls (NS). The mean CES-D score for girls was significantly higher than for boys (20.3+/-10.7 versus 16.7+/-9.9; p<0.01). According to the cut-off score of 24, 19% of boys and 34% of girls had a moderate to severe depressive symptomatology (p<0.0001). The mean suicidal ideation score was significantly higher in participants scoring 24 or above on the CES-D than participants scoring less than 24 (2.4+/-2.7 versus 0.3+/-0.9; p<0.0001). Among participants with CES-D greater or equal to 24, 34% reported at least occasional wish to kill oneself versus 6% of participants with CES-D less than 24 (p<0.0001). CES-D scores and suicidal ideation scores were moderately correlated in girls (Pearson's r=0.59) and boys (r=0.61) in the total sample. To explore the role of hopelessness as mediator between depressive symptoms and the wish to kill oneself, multiple regression analyses were performed separately by gender. To establish mediation, three regression equations should be estimated and the four following conditions must hold: First, the independent variable (CES-D scores) must affect the mediator in the first equation; second, the independent variable must affect the dependent variable (the wish to kill oneself) in the second equation; third, the mediator must affect the dependent variable in the third equation regressing the dependent variable on both the independent variable and on the mediator; fourth, the effect of the independent variable on the dependent variable must be less in the third equation than in the second. Among girls, in the first equation, CES-D score explained 35% of the variance of hopelessness (beta=0.59, t=21.5, p<0.001). In the second equation, CES-D score explained 16% of the variance in the wish to kill oneself (beta=0.40, t=12.7, p<0.001). In the third equation, CES-D and hopelessness scores explained 32% of the variance in the wish to kill oneself. Hopelessness was the main predictor (beta=0.50, t=14, p<0.001) while the effect of CES-D was markedly reduced (beta=0.10, t=2.9, p<0.01). Among boys, in the first equation, CES-D score explained 38% of the variance of hopelessness (beta=0.62, t=20.7, p<0.001). In the second equation, CES-D score explained 25% of the variance in the wish to kill oneself (beta=0.50, t=15.1, p<0.001). In the third equation, CES-D and hopelessness scores explained 47% of the variance in the wish to kill oneself. Hopelessness was the main predictor (beta=0.60, t=17, p<0.001) while the effect of CES-D was substantially weakened (beta=0.13, t=3.6, p<0.001).

DISCUSSION: These results showed a strong association between depressive symptoms and suicidal ideation in this nonclinical sample of adolescents. According to Beck's assumption, hopelessness appeared to be a mediator between depressive symptoms and the wish to kill oneself both in boys and girls. These findings are relevant for prevention and therapy. They suggest that targeting hopelessness may be as important in adolescents as in adults to reduce suicidal ideation and prevent suicidal attempts.