The suicide rate among active duty service members and veterans increased substantially following the onset of post-9/11 conflicts in Iraq and Afghanistan1. Accordingly, veteran suicide prevention has been identified as a national healthcare and research priority2. Psychosocial interventions for suicide risk vary substantially in dose and resource allocation. A single therapy session designed to evaluate risk factors and provide support resources (e.g., Enhanced Crisis Response Plans [ECRP]3) has been shown to reduce risk for future suicide attempts. Other interventions consisting of 10-12 outpatient sessions following inpatient discharge (e.g., Brief Cognitive Behavioral Therapy for suicide prevention [BCBT]4) have been shown to reduce suicide attempts by 50-60% relative to treatment as usual. Although both forms of intervention have been shown to reduce risk, interventions that vary in dose and resource allocation have yet to be directly compared, leaving two critical gaps in our ability to intervene most effectively. First, the assumption that more time- and resource-intensive 10-12 session interventions translate to greater suicide risk reduction has yet to be demonstrated. Second, it may be that less resource intensive interventions are adequate for some individuals whereas others require more intensive care. To date, there is no evidence to guide what interventions are indicated for specific clinical presentations. Pharmacological and brain stimulation interventions for suicide risk are extremely limited. This is due, in part, to an incomplete understanding of the neurobiological mechanisms of suicide risk. Although numerous studies have examined cross-sectional neuroimaging correlates of current suicide ideation or compared individuals with and without history of a suicide attempt, to date no studies have examined a) neurobiological predictors of future suicide attempts in high-risk samples, b) how changes in neurobiological markers over time relate to changes in suicide risk, or c) theoretically and mechanistically relevant neuroimaging procedures in a prospective design. Cross-sectional research examining neuroimaging markers of past or current self-injurious thoughts and behaviors (SITBs) has identified dysfunction in regions associated with emotion regulation, inhibitory control, and decision-making5,6, namely in cognitive control networks (CCN). On the other hand, dysfunction has also been observed in regions associated with negative affect and rumination such as limbic (LN) and default mode (DMN) networks. Despite these cross-sectional findings, identification of neuroimaging predictors of future suicide attempts, and neural markers of successful suicide risk intervention outcomes represents a completely novel, critical step to guiding optimal targeting of neurobiologically-informed interventions and translating neuroimaging of suicide into practice. Whether these potential neuroimaging predictors are identifiable during resting state, or whether more suicide-relevant cognitive tasks are required, such as death-related bias or inhibitory control, remains an open yet critical question. The purpose of our proposed study is to compare two evidence-based suicide risk interventions that vary in dose in order to a) directly test if a more intensive intervention produces greater risk reduction, b) identify veterans for whom a more intensive intervention is indicated, and c) identify resting-state and task- based neurobiological markers of future suicide attempts and examine how changes in these markers relate to changes in suicide risk over time. We will recruit and evenly randomize 136 male and female veterans hospitalized for suicide risk to ECRP or BCBT. We will collect neuroimaging data immediately upon discharge, post-treatment, and 12-months post-discharge and assess SITBs out to 12-months post-discharge.

Veteran suicide prevention has been identified as a national healthcare and research priority2. The aims of this study are to address three prominent gaps in the scientific literature. We will compare two evidence-based suicide risk interventions that vary in dose in order to a) directly test if a more intensive intervention produces greater risk reduction, b) identify veterans for whom a more intensive intervention is indicated, and c) identify resting-state and task-based neurobiological markers of future suicide attempts and examine how changes in these markers relate to changes in suicide risk over time. We will recruit and evenly randomize 136 male and female veterans hospitalized for suicide risk to an efficient or a more time-intensive intervention. We will collect neuroimaging data immediately upon discharge, post-treatment, and 12-months post-discharge and assess self-injurious thoughts and behaviors out to 12-months post-discharge.