Early Life Predictors of Adolescent Suicidality

Abstract

Background: Suicidal thoughts affect 12% of Canadian adolescents. Previous research has linked many factors to suicidality but has not considered how these factors may act together or their effect on non-mental health outcomes.

Methods: I used the National Longitudinal Survey of Children and Youth to construct predictive models for suicidal thinking. Recursive partitioning models were constructed and the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for suicidal thoughts and secondary outcomes was calculated. I tested the models in the Avon Longitudinal Survey of Parents and Children.

Results: Predictive model sensitivity was 24.2%, specificity was 89.8%, PPV was 24.7%, and NPV was 89.5% and had similar accuracy in the second dataset. The models were better at predicting other adverse outcomes compared to suicidal ideas.

Conclusion: Exposure to multiple risk factors is predictive of several poor outcomes in adolescence including suicidal thoughts.