(PS9-A16) Typologies of Psychiatric Diagnoses in Inpatient Suicide Attempters

Presence of a psychiatric diagnosis and multimorbidity are well-documented suicide risk factors. However, historical emphasis on variable-centered statistical approaches has precluded the identification of patterns of comorbidity that are uniquely associated with increased suicide risk. Extending prior work (e.g., Ginley & Bagge, 2017), the current study used latent class analysis (LCA) to examine distinct typologies of psychiatric heterogeneity based on the presence of current psychiatric diagnoses and features. We also explored associations between class membership and suicide history characteristics (i.e., suicidal ideation intensity, number of lifetime suicide attempts) as well as distal clinical correlates, including demographic characteristics, cognitive and emotional risk factors (i.e., suicide cognitions, depression, anxiety, stress, hopelessness), and behavioral risk factors (i.e., non-suicidal self-injury, physical aggression, impulsivity). Participants were 213 adult inpatients with a suicide attempt within the past two years (M age = 33.04 [SD = 12.67]; 50.2% female; 62.4% White; 23.9% Hispanic/Latino) who were administered a semi-structured diagnostic interview and suicide risk assessment, and completed a battery of self-report measures. LCA revealed that a three-class solution fit the data best. The Depressive-High Comorbidity class (N = 71) was characterized by the presence of a depressive disorder, high rates of comorbidity, especially with an anxiety or trauma/stressor-related disorder and borderline personality disorder, and the highest rate of substance use disorder. The Depressive-Low Comorbidity class (N = 85) contained the largest proportion of the sample, and involved a high probability of a depressive disorder without psychotic features and low rates of comorbid diagnoses. The Bipolar class (N = 57) was distinguished by its containment of most individuals with a bipolar and related disorder and the highest probability of psychosis. The Bipolar class also reflected high probability of having borderline personality disorder. The Depressive-High Comorbidity and Bipolar classes endorsed more severe past month suicidal ideation and a greater number of lifetime actual and interrupted suicide attempts than did the Depressive-Low Comorbidity class. These two classes also consistently reported significantly higher levels of cognitive, emotional, and behavioral risk factors than the Depressive-Low Comorbidity class. Additionally, a higher proportion of the Depressive-High Comorbidity class (35.94%) was homeless as compared to the other two classes. In the context of suicide risk, person-centered approaches can illuminate clusters of suicide attempters with idiosyncratic clinical needs, which is an essential step toward tailoring interventions to maximally address risk for re-attempts. Our findings emphasize the uniquely high risk for suicide conferred by bipolar, psychotic, and mixed internalizing-externalizing clinical presentations. These results also highlight the need to attend to environmental stressors, such as homelessness, as part of suicide prevention.