Association of Premilitary Mental Health with Suicide Attempts during US Army Service

James A. Naifeh, Robert J. Ursano*, Murray B. Stein, Holly B.Herberman Mash, Pablo A. Aliaga, Carol S. Fullerton, Hieu M. Dinh, Tzu Cheg Kao, Nancy A. Sampson, Ronald C. Kessler

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

4 Scopus citations


Importance: Approximately one-third of US soldiers who attempt suicide have not received a mental health diagnosis (MH-Dx) before their suicide attempt (SA), yet little is known about risk factors for SA in those with no MH-Dx. Objective: To examine whether premilitary mental health is associated with medically documented SA among US Army soldiers who do not receive an MH-Dx before their SA. Design, Setting, and Participants: This cohort study used data from a representative survey of soldiers in the US Army entering basic combat training from April 1, 2011, to November 30, 2012, who were followed up via administrative records for the first 48 months of service. Analyses were conducted from April 5, 2021, to January 21, 2022. Regular Army enlisted soldiers (n = 21772) recruited from 3 US Army installations during the first week of service who agreed to have their administrative records linked to their survey responses were included. Exposures: Preenlistment lifetime history of mental disorder, suicide ideation, SA, and nonsuicidal self-injury (NSSI) as reported during the baseline survey. Service-acquired MH-Dx and sociodemographic and service-related variables were identified using administrative records. Main Outcomes and Measures: Documented SAs were identified using administrative medical records. Using a discrete-time survival framework, linear splines examined the pattern of SA risk over the first 48 months of service. Logistic regression analysis examined associations of lifetime baseline survey variables with subsequent, medically documented SA among soldiers who did vs did not receive an MH-Dx during service. Models were adjusted for time in service and sociodemographic and service-related variables. Results: Of the 21722 respondents (86.2% male, 20.4% Black, 61.8% White non-Hispanic), 253 made an SA in the first 48 months of service (male [75.4%]; Black [22.7%], White non-Hispanic [59.9%], or other race or ethnicity [17.4%]). Risk of SA peaked toward the end of the first year of service for both those who did and did not receive an MH-Dx during service. Of the 42.3% of individuals reporting at least 1 of the 4 baseline risk factors, 50.2% received an administrative MH-Dx during service vs 41.5% of those with none, and 1.6% had a documented SA vs 1.0% of those with none. Among individuals with no MH-Dx, medically documented SAs were associated with suicide ideation (odds ratio [OR], 2.2; 95% CI, 1.1-4.4), SA (OR, 11.3; 95% CI, 4.3-29.2), and NSSI (OR, 3.0; 95% CI, 1.3-6.8). For those who received an MH-Dx, medically documented SAs were associated with mental disorder (OR, 1.4; 95% CI, 1.0-1.9), SA (OR, 3.4; 95% CI, 2.1-5.6), and NSSI (OR, 1.8; 95% CI, 1.1-2.8). Interactions indicated the only explanatory variable that differed based on history of MH-Dx was preenlistment SA (χ21= 4.7; P =.03), which had a larger OR among soldiers with no MH-Dx than among those with an MH-Dx. Conclusions and Relevance: In this study, the period of greatest SA risk and baseline risk factors for SA were similar in soldiers with and without an MH-Dx. This finding suggests that knowledge of the time course and preenlistment mental health factors can equally aid in identifying SA risk in soldiers who do and do not receive an MH-Dx.

Original languageEnglish
Pages (from-to)E2214771
JournalJAMA Network Open
Issue number6
StatePublished - 10 Jun 2022
Externally publishedYes


Dive into the research topics of 'Association of Premilitary Mental Health with Suicide Attempts during US Army Service'. Together they form a unique fingerprint.

Cite this