Abstract
Using data from a Department of Veterans Affairs study of schizophrenia (SZ) and bipolar I disorder (BP), we evaluated lifetime risks for suicidal ideation and behavior. We were interested in the prevalence and correlates of these outcomes, in populations of patients with severe mental illness (SMI), who have not been compared directly in previous studies despite data on high risk in each group separately.
Data were collected on demographic factors, medical and psychiatric comorbidity, cognitive and functional status, and lifetime suicidal ideation or behavior in a study of veterans with SZ (N = 3942) or BP (N = 5414). In-person diagnosis and evaluation, including performance-based assessments of cognition and functional skills, make this study unique compared to studies of completed suicide. Multinomial logistic regression examined how risk factors, including major depression and negative symptoms in SCZ patients, correlated with ideation and behavior.
A lifetime history of suicidal ideation or behavior was reported by a majority of Veterans with SZ (69.9%) or BP (82.3%). Lower risk was found for patients with SZ vs. BP (odds ratio [OR] = 0.82 for ideation; OR = 0.81 for behavior). The highest risk was found for patients with multiple psychiatric comorbidities (OR = 2.61 for ideation; OR = 3.82 for behavior). Clinical factors (e.g., psychiatric comorbidity) contributed more of the variance in the predictive model than demographic factors.
A history of suicidal ideation or behavior is common among US Veterans with SZ or BP, and specific demographic and clinical attributes correlate with variation in risk. These findings underscore the need for continuous monitoring for suicidal ideation and behavior in veteran populations with SMI, as well as the development of better risk predictions, including genomic factors. Although PTSD is a major current focus in veteran populations, the large number of veterans with SZ or BP and their high suicide risk suggests a greater clinical focus may be warranted.
