Latent emotion profiles of PTSD and specific emotions predicting differential therapy outcomes in a dismantling study of cognitive processing therapy

Posttraumatic stress disorder (PTSD) is theoretically maintained by avoidance of emotions elicited from trauma-related beliefs. Whether PTSD symptom profiles and specific emotions predict treatment response is unknown. This secondary data analysis examined: a) whether individuals with PTSD can be sub-classified based on symptom clusters and specific emotions, and b) if these subgroups predict differential responses to cognitive versus exposure-based PTSD interventions. Women with physical or sexual assault-related PTSD were random-ized to CPT (cognitive processing therapy elements only), CPT with written accounts (CPT+A), or written ac-counts (WA) only (n = 150). Participants completed baseline measures of PTSD, state anxiety, internalized anger, externalized anger, shame, and guilt, and weekly PTSD measures during and 6 months after treatment. Latent profile analyses revealed four subgroups: low symptoms and emotions; moderate-high reexperiencing, low internalized emotions (i.e., moderate-high reexperiencing, moderate avoidance/hyperarousal/guilt, low shame/internalized anger/anxiety); low reexperiencing, moderate emotions (i.e., low re-experiencing, moderate avoidance/hyper-arousal/guilt, moderate other emotions); and high symptoms and emotions (high symptoms and emotions except moderate externalized anger). The high symptom and emotion subgroup experienced greater PTSD symptom improvements in cognitive conditions than WA. Other groups did not exhibit differential change across condi-tions. Cognitive interventions may be well-suited for severe PTSD with high self-directed emotions. ClinicalTrials.gov identifier: NCT00245232