We developed a machine learning model to predict the first onset of suicidal thoughts and/or behaviours (STB) in adoelscence – a develomental stage most vulnerable for suicidality. Our model predicted first-onset STB over four years with moderate accuracy (AUCs ≈0.63–0.73 across comparisons with healthy and clinical controls. The most important predictors in this population included female sex, sleep disturbance, and adverse home and school environments.
Open Access: Child & Adolescent Psychiatry
