Suicide is the second leading cause of death among young people and suicide rates in youth have been steadily increasing over the past 10 years. Emerging demographic data also point to rising suicide rates in pre-teens and early adolescents, particularly among females and African Americans. Unfortunately, the underlying reasons for the rising rates of youth suicide and these alarming demographic trends are not yet understood. While no causal association has been established, the upward trend in social media usage over the past decade coincides with the rise in suicide rates. During the developmental periods of late childhood and adolescence, when individuals become acutely attuned to social status and feedback from peers, social media use may transform interpersonal experiences in both positive and negative ways, including more frequent opportunities for social support as well as peer victimization and social comparison. Notably, teens who report using social media sites greater than 2 hours a day are more likely to report poor mental health outcomes, such as psychological distress and suicidal ideation. Negative upward social comparison and cyberbullying have been associated with depression, anxiety, and suicidal ideation in adolescents. Social media use, 24/7 availability, may also contribute to sleep difficulties. The objective of this R21 application is to examine proximal antecedents of suicidal attempt in 100 youth between the ages of 10-17 with suicide attempts requiring medical hospitalization and compare them to an equal number of controls who were admitted for non-chronic illness or injury not related to suicide attempt. In addition to assessment of known risk factors, this study will examine patterns in the frequency, duration, timing and content of SM use prior to the index suicide attempt: objectively using innovative methods to collect time-stamped SM text and emojis, and subjectively from parent and child report. We will explore variations in psychopathology, SM use, positive and negative consequences of SM use, and suicidal ideation and behavior prior to attempt across the groups over the year prior to attempt. We will also explore differences by age [(pre-and early adolescents (age 10-13) vs. middle and late adolescents (age 14-17)], race, biological sex and self-reported gender The research proposed in this R21 application is the first step to address a major gap in the science by identifying digital risk factors and characterizing youth suicide attempters in domains that may contribute to increased rates of youth suicidal behavior. Findings from this study will provide feasibility and preliminary data for a future prospective R01 submission in a larger sample. The findings could add significantly to currently known clinical predictors of suicidal behavior and aid in earlier identification and management of suicidal risk.

This proposed R21 is the first step to address a major gap in the science by identifying digital risk factors and characterizing youth suicide attempters in domains that may be contributing to increased rates and changing demographics of youth suicidal behavior. This study will use innovative methods to collect social media data and will provide preliminary data for a future R01 submission to further develop and test, in a large sample of youth, the ability of NLP algorithms to detect suicidal risk in real time. The findings could add significantly to currently known clinical predictors of suicidal behavior and aid in earlier identification and management of suicidal risk.