Longitudinal relations between insomnia and suicide attempts in adolescence and young adulthood.

Introduction: Insomnia has been identified as a risk factor for adolescent suicidal behavior, often via clinical and cross-sectional studies. Longitudinal associations between insomnia and suicidal behavior have not been well-established in adolescent community samples, and many studies fail to control for the effect of depressive symptoms and their overlapwith both insomnia and suicidal behavior. Objectives: We modeled longitudinal associations between insomnia and suicide attempts in a diverse community-based cohort of US adolescents interviewed at Wave 1 (n=4175; 11-17yo), W2 (+1yr; n=3134), and W3 (+5yrs; n=2503). Methods: We estimated the association of W1 insomnia with W2 and W3 12-month suicide attempts, controlling for W1 attempts. We then estimated the inverse association between W1 attempts and W2 and W3 insomnia, controlling for W1 insomnia. Analyses used weighted survey logistic regression models, controlling forW1 depressive symptoms (excluding insomnia/suicidality) and sociodemographics. Results: Weighted prevalence of any insomnia symptom ranged from 22.2%-26.8% across waves; insomnia with impairment ranged from 6.1%-8.1%. Any insomnia symptom at W1 was not associated with W2 attempts but significantly increased the odds of W3 attempts (OR=3.82). In contrast, insomnia with impairment at W1 was associated with W2 attempts (OR=3.14) but not W3. The inverse analysis showed no significant associations between W1 attempts and W2/W3 insomnia. Conclusions: Presence of any insomnia symptom in adolescence may represent a distal risk factor for later suicide attempts, while insomnia with impairment may be a short-term risk factor. Preventive interventions to improve adolescent sleep, combined with clinical interventions for sleep disorders, may be an important multilevel suicide prevention strategy. [ABSTRACT FROM AUTHOR]