A Path Model of Relationships Between Bullying Victimization and Suicidal Thinking, Suicide Planning, and Suicide Attempts in Adolescents

Research has shown an increasing prevalence of suicide among adolescents.  The bullying of adolescents on school grounds (traditional bullying) has been shown to be significantly related with adolescent suicide. Today’s adolescents are confronted with additional and novel stressors as a result of cyberbullying.  Cyberbullying has been shown to be associated with suicidal ideation and attempts.  Previous research has also suggested relationships exist between both traditional and cyber bullying victimization and depression, violent behavior, and adolescent substance abuse.  The goal of this study was to test a comprehensive path model for the relationships between age, gender, traditional bullying and cyber bullying victimization, and violent behavior, substance abuse, depression, suicidal ideation and suicide attempts in adolescents.  Based on results of prior research, it was hypothesized that depression would mediate the relationships between both traditional and cyber bullying victimization and suicidal ideation; and that violent behavior and substance abuse would mediate the relationships between both traditional and cyber bullying victimization and suicidal ideation and suicidal behaviors.

The hypothesized path model was fit to data from the 2011 Youth Risk Behavior System Survey (YRBSS) on a nationally representative sample of 15,425 high-school students from across the United States. The path model was estimated using the robust weighted least squares method in Mplus, and the complex sampling used in the 2011 YRBSS was taken into account in analyses.  Missing data were handled using multiple imputation. A multiple groups approach was used to test the invariance of the path model across male and female adolescents.

The final model fit indices for the model were, Chi-squared (31) = 49.0, p = .02; RMSEA = .009 (90% CI, .003, .013), probability (RMSEA ≤ .05) = 1.0; CFI = .999, and TLI = .999.  In terms of effect sizes, the 99.9% confidence intervals for the R-squared for suicide attempts for females was, .76 - .86; and for males, .77 - .87. Results suggested that the effects of both traditional and cyber bullying victimization on suicidal thinking were partially mediated by depression.  Results further suggested that the relationships between both traditional and cyber bullying victimization and suicidal behaviors were mediated by both substance abuse and violent behavior. 

Results indicated statistically significant indirect paths from both traditional and cyber bullying victimization to suicide attempts without the involvement of depression, suicidal thinking, or suicide planning. These findings suggested that both traditional and cyber bullying victimization may be associated with spontaneous suicide attempts by non-depressed adolescents.  While the configuration of the model was invariant across males and females, there were a few differences in path coefficients for male and female adolescents.

The results have implications for social workers in school and mental health settings.  For example, adolescents who report being bullied, but no depression or suicidal thinking, should be considered at risk for a spontaneous suicide attempt.  Under these circumstances social workers should look for events in the adolescent’s life that might serve as a triggering event, such as a cyberbullying victimization incident, substance abuse, or a recent physical fight.