Grant: CIHR Fellowship

Suicide is the second leading cause of death among youth ages 10-24. Although most widely recognized suicide risk factors, such as hopelessness and impulsivity, are strong predictors for considering suicide, these do not accurately predict who will eventually attempt suicide. Developing prediction models that distinguish patients who will attempt suicide from patients with suicidal ideation but no attempts can save lives. Although individuals with depressed mood and impulsivity commonly present with cognitive processing difficulties (e.g., attention, memory), or "deficits," research findings are mixed regarding whether these factors predict suicide attempt. Furthermore, research shows enhanced cognitive processing abilities predict suicide behavior as well. For these reasons, I hypothesize that complex relationships exist, such that cognitive deficits alone may not predict attempt, but unique combinations of cognitive abilities and psychological symptoms (e.g., social withdrawal) will predict suicide attempt beyond single or traditional risk factors. The ability of computer-based algorithms to learn unique patterns that are difficult to detect with traditional methods is central to this investigation. To test this hypothesis, I am conducting the first investigation to my knowledge that predicts youth suicide attempt using a cognitive functioning and psychological symptom profile. The CAMH Toronto Adolescent and Youth Cohort Study is unprecedented in Canada in collecting data from psychodiagnostic assessments, multi-informant clinical measures, and cognitive testing on an annual basis for five years from 3000 at-risk youth ages 11 to 24. Computer-based algorithms will utilize this assessment data to predict future suicide attempts over the next year. Using information technology to identify patients at heightened suicide risk will have a critical impact on improving clinical decision making and suicide prevention efforts for high-risk patients.