(PS9-26) Prediction of Future Suicide Attempts Among Individuals with History of Exposure to Violence and Suicidal Ideation

Abstract Text: Exposure to violence (EV) can have both direct and indirect impacts on loss of life. Beyond the direct possibility of death at the hands of another, EV has been identified as a risk factor for suicide (e.g., Castellvi et al., 2016; MacIsaac et al., 2016). Recent research has increasingly highlighted the importance of differentiating those risk factors that are associated with suicidal ideation (SI) versus those that predict suicide attempt (SA) given the presence of SI (e.g., Nock, Kessler, & Franklin, 2016; Nock et al., 2018). It is unclear whether EV independently predicts SAs. Additionally, prior research has not addressed whether types of EV – witnessing versus being the victim – differentially predicts SAs.
To address these gaps in existing research on EV and suicide risk, we aimed to test whether types of EV were associated with both history of SAs and prediction of future SAs during a 6-month study window, among individuals with preexisting SI. We recruited a sample of individuals presenting to the hospital with SI: n=252 adults presenting to an emergency department and n=300 adolescents treated on a psychiatric inpatient unit. Participants completed a baseline assessment including SA and EV history. After hospital discharge, participants’ SAs were documented over the following 6 months through a combination of electronic health records data, daily surveys, and interaction with study staff. All analyses included age, gender, and racial identity as covariates.
Nearly 55% of the adults (n=141) and 65% of the adolescents (n=190) reported a history of one or more SAs, and 67% of adults (n=168) and 25% of adolescents (n=76) reported a history of EV. In the full sample, EV broadly was a strong predictor of future SA over the 6-month period following hospitalization when controlling for age, gender, and race, (p =.002). Prediction of future SAs did not remain significant when also controlling for prior history of SA. However, among adults but not adolescents, EV was significantly associated with history of SA (p =.011), and history of both witnessing violence (p =.011), and being a victim of violence (p =.007) predicted suicide attempt during the study, even when including history of prior attempt as a covariate.
Overall, our findings suggest two important takeaways. First, that exposure to violence is associated with both a history of SAs and future SAs, even among individuals with SI. Second, that both history of being a victim of and witnessing violence are useful predictors of future SAs. Across many of the nation’s largest cities, violence has been on the rise over the past decade, including an overall convergence of community violence, police violence and contact with the criminal system (Sharkey & Marsteller, 2022). Better understanding of the association between EV and risk of SA may inform risk assessment approaches and targeted intervention for suicide prevention. Advancing this knowledge can help open the door for bettering behavioral health treatment and community resilience to counteract the rise in violence being seen.