Moral transgression during the Vietnam War: a path analysis of the psychological impact of veterans' involvement in wartime atrocities

Abstract

Background and objectives: Involvement in wartime combat often conveys a number of deleterious outcomes, including posttraumatic stress disorder (PTSD), depression, hostility, aggression, and suicidal ideation. Less studied is the effect of engagement in wartime atrocities, including witnessing and perpetrating abusive violence.

Design and methods: This study employed path analysis to examine the direct effects of involvement in wartime atrocities on hostility, aggression, depression, and suicidal ideation independent of combat exposure, as well as the indirect effects via guilt and PTSD symptom severity among 603 help-seeking male Vietnam War veterans.

Results: Involvement in wartime atrocities was predictive of increased guilt, PTSD severity, hostility, aggression, depressive symptoms, and suicidal ideation after controlling for overall combat exposure. Combat-related guilt played a minor role in mediating the effect of atrocity involvement on depression and suicidal ideation. PTSD severity had a larger mediational effect. However, it still accounted for less than half of the total effect of involvement in wartime atrocities on hostility, aggression, and suicidal ideation.

Conclusions: These findings highlight the heightened risk conveyed by involvement in wartime atrocities and suggest that the psychological sequelae experienced following atrocity involvement may extend well beyond guilt and PTSD.

Keywords: Abusive violence; depression; guilt; moral injury; posttraumatic stress disorder; suicidal ideation.

Figures

Figure 1

Social-cognitive model of moral injury.

Figure 2

Path analysis model of influence of involvement in wartime atrocities on PTSD, aggression, and suicide. Solid lines represent direct effects after controlling for predictors entered earlier in the model. Dashed lines represent unadjusted direct effects (i.e., direct effects not statistically adjusting for potential mediators/subsequently entered predictors). Numerical values are standardized effects. *p < .05, **p < .01. N = 599.