Reducing Suicide among Military Personnel

Suicide is the leading cause of death for U.S. soldiers, who are almost nine times more likely to die by suicide as they are in combat. University of Denver Graduate School of Social Work Associate Professor Anthony Fulginiti is working to change that by improving the effectiveness of suicide prevention interventions.

In a new paper, “Suicide-Related Disclosure Intention Among Active-Duty Servicemembers: A Generalized Linear Mixed-Model Tree Analysis,” published by the Journal of the Society for Social Work and Research, Fulginiti identifies individual- and relationship-level characteristics associated with suicide-related disclosure intention among active-duty military personnel.

Fulginiti explains that the disclosure of one’s suicidal thoughts to others is a key component of suicide prevention efforts, but there has been little research among active-duty servicemembers. He says, “If a servicemember discloses suicidal thoughts, they can be prioritized to receive preventive interventions. However, a recent Department of Defense Suicide Event Report indicates that only 28% of military personnel who died by suicide had told someone that they were going to kill themselves. We need to understand the determinants of disclosure among active-duty military personnel to design suicide prevention strategies that promote the communication of suicidal thoughts.”

The study sample included 241 active-duty U.S. Army personnel recruited from a single battalion stationed in the United States. The research team used survey and social network research methods to collect individual- and relationship-level data. Participants were asked to list up to 10 individuals with whom they had interacted in the last month. Participants then answered questions about their relationships with and the characteristics of those network members, including whether they would communicate their suicide-related thoughts to each member of their social network.

Researchers then applied an artificial intelligence machine learning approach, generalized linear mixed-model (GLMM) trees, to predict disclosure intention. The decision tree approach accounts for multi-level data structures, allowing researchers to test for interactions between individual- and relationship-level characteristics that may influence suicide-related disclosure intention. The authors write, “This unique multi-level treatment of disclosure could help to more accurately capture the complexity of self-disclosure determinants and identify novel prevention targets among military personnel.”

Fulginiti found that two individual-level factors (military unit cohesion and number of siblings) and four relationship-level factors (network member emotional support, informational support, family disclosures and emotion regulation) contributed to the GLMM tree. Emotional support, informational support and high unit cohesion together were associated with the highest probability of disclosure intention. Notably, Fulginiti says, unit cohesion (how mutually supportive unit members are of one another) was an important military service factor associated with disclosure.

According to Fulginiti, identifying that relationship factors were more central to disclosure intention than individual factors underscores the potential value of social network interventions. The authors write, “Identifying ‘disclosure profiles’ can help to design and set priorities for military suicide prevention program planning.”

Fulginiti adds, “Relationship factors are significant for understanding how individuals manage and communicate private and potentially stigmatizing health and mental health information. We found that the most prominent relationship factors related to disclosure intention included whether social network members were sources of emotional and informational support. Social network members who are reliable sounding boards and provide good advice may be viewed as frontline candidates for support during a crisis.”

Funded by the Army Research Office (Principal Investigator: Professor Eric Rice, University of Southern California), the research holds important implications for suicide prevention program planning and implementation in the military. For example, Fulginiti says, “If we want to encourage soldiers to share their suicidal thoughts, then we need to develop and strengthen policies and programs that build peer support capacities, such as gatekeeper suicide prevention programs, and break down siloes between clinical and case management services. Such efforts should prioritize the active involvement of soldiers to ensure risks and benefits of disclosure are carefully balanced.”