Self‐injurious thoughts and behaviors that differentiate soldiers who attempt suicide from those with recent suicide ideation

• Published in: Depression and anxiety Vol. 37; no. 8; pp. 738 - 746
• Main Authors: Naifeh, James A., Nock, Matthew K., Dempsey, Catherine L., Georg, Matthew W., Bartolanzo, Danielle, Ng, Tsz Hin Hinz, Aliaga, Pablo A., Dinh, Hieu M., Fullerton, Carol S., Mash, Holly B. Herberman, Kao, Tzu‐Cheg, Sampson, Nancy A., Wynn, Gary H., Zaslavsky, Alan M., Stein, Murray B., Kessler, Ronald C., Ursano, Robert J.
• Format: Journal Article
• Language: English
• Published: United States Hindawi Limited 01-08-2020
• Subjects: Gender; Humans; Mental disorders; military; Military Personnel; nonsuicidal self‐injury; Post traumatic stress disorder; Risk Factors; Self destructive behavior; Self-Injurious Behavior - epidemiology; Suicidal Ideation; Suicide; suicide attempt; suicide ideation; suicide plan; Suicide, Attempted; Suicides & suicide attempts; Veterans; suicide ideation; nonsuicidal self-injury; suicide attempt; military; suicide plan
• ISSN: 1091-4269; 1520-6394
• DOI: 10.1002/da.23016

Background
Risk for suicide attempt (SA) versus suicide ideation (SI) is clinically important and difficult to differentiate. We examined whether a history of self‐injurious thoughts and behaviors (SITBs) differentiates soldiers with a recent SA from nonattempting soldiers with current/recent SI.
Methods
Using a unique case‐control design, we administered the same questionnaire (assessing the history of SITBs and psychosocial variables) to representative U.S. Army soldiers recently hospitalized for SA (n = 132) and soldiers from the same Army installations who reported 30‐day SI but did not make an attempt (n = 125). Logistic regression analyses examined whether SITBs differentiated attempters and ideators after controlling for previously identified covariates.
Results
In separate models that weighted for systematic nonresponse and controlled for gender, education, posttraumatic stress disorder, and intermittent explosive disorder, SA was positively and significantly associated with the history of suicide plan and/or intention to act (odds ratio [OR] = 12.1 [95% confidence interval {CI} = 3.6–40.4]), difficulty controlling suicidal thoughts during the worst week of ideation (OR = 3.5 [95% CI = 1.1–11.3]), and nonsuicidal self‐injury (NSSI) (OR = 4.9 [95% CI = 1.3–18.0]). Area under the curve was 0.87 in a full model that combined these SITBs and covariates. The top ventile based on predicted risk had a sensitivity of 24.7%, specificity of 99.8%, and positive predictive value of 97.5%.
Conclusions
History of suicide plan/intention, difficult to control ideation, and NSSI differentiate soldiers with recent SA from those with current/recent SI independent of sociodemographic characteristics and mental disorders. Longitudinal research is needed to determine whether these factors are prospectively associated with the short‐term transition from SI to SA.