Care rejection and aggression among veterans with dementia with and without posttraumatic stress disorder: A multi-group analysis.

Background: In residential long-term care, military service veterans with co-occurring posttraumatic stress disorder (PTSD) and dementia encounter a range of physical and social stimuli that may serve as triggers of trauma-related distress that manifests as care rejection or aggression. Yet, PTSD rarely has been examined in research to understand behaviors of care rejection and aggression in veterans with dementia.

Objectives: Guided by the need-driven dementia-compromised behavior model, we examined the moderation effect of PTSD on pathways from background factors and interpersonal triggers to rejection of care and aggression among veterans with dementia with and without co-occurring PTSD.

Design: Secondary data analysis of the Staff Training in Assisted Living Residences-Veterans Health Administration (STAR-VA) intervention evaluation by the U.S. Veterans Health Administration healthcare system.

Setting: 76 Veterans Health Administration-operated nursing homes.

Participants: 315 veterans with dementia who participated in STAR-VA.

Methods: We converted text data on the occurrence of care rejection and aggression to binary variables, combined them with data on sociodemographic and PTSD status obtained using medical chart review, and measured anxiety, cognition, depression, and function using validated instruments. A multi-group structural equation modeling analysis was then conducted to test the moderating effect of PTSD on rejection of care and aggression.

Results: Although multi-group structural equation modeling did not support the hypothesis of overall moderation by PTSD, distinct patterns between the two groups were observed with respect to how background factors and interpersonal triggers related to care rejection and aggression. The magnitude of the direct effects of interpersonal triggers on rejection of care was greater in veterans with PTSD (β = 0.42, p = .014 compared to those without β = 0.29, p = .008). Depression had a statistically significant indirect effect on rejection of care via interpersonal triggers only in veterans with PTSD (β = 0.09, p = .009). Functional status had a statistically significant direct effect on aggression only in the PTSD group (β = 0.28, p = .044).

Conclusions: Our study identified similar and distinct patterns of relationships among background factors, interpersonal triggers, and rejection of care and aggression between veterans with dementia with and without PTSD. The indirect effect of depression on care rejection via interpersonal triggers has implications for developing targeted interventions that focus on interpersonal triggers for veterans with dementia with PTSD who have greater depressive symptoms. This study underscores the importance of an enhanced focus on trauma-informed care for veterans with dementia and PTSD.