Abstract
BACKGROUND: Posttraumatic stress disorder (PTSD) is a heterogeneous condition with diverse symptom presentations and emotional experiences. While fear is traditionally viewed as central, growing evidence highlights the role of non-fear-based emotions, such as sadness, guilt, and shame-collectively termed emotional pain. This study aimed to identify fear- and emotional pain-based PTSD symptom profiles and their neural correlates across 2 independent samples. METHODS: In study 1 (N = 838), trauma-exposed individuals with probable PTSD completed the PTSD Checklist for DSM-5 and subjective ratings of fear and emotional pain. Item-level network analysis was conducted to identify central symptoms and relationships. In study 2 (N = 162), recent trauma survivors with high PTSD symptoms underwent resting-state and task-based functional magnetic resonance imaging scans 1 month after trauma and completed follow-up clinical assessment at 14 months after trauma. Connectome-based predictive modeling (CPM) was used to predict chronic symptom severity for fear- and emotional pain-based profiles, identified in study 1. RESULTS: Emotional pain was rated as more impairing than fear by most participants (69%). Symptom networks showed distinct patterns: Fear was associated with flashbacks, nightmares, distressing memories, exaggerated startle, and external avoidance; emotional pain was linked to anhedonia, negative beliefs, negative emotions, sleep disturbance, and emotional reactivity. CPM predicted chronic fear-based symptom severity (ρ = 0.228, p < .001), but not emotional pain (ρ = 0.167, p = .055). Predictive features included connections across anterior default mode, central executive, salience, motor-sensory, and subcortical networks. CONCLUSIONS: Emotional pain and fear may represent distinct PTSD dimensions. Disentangling their neural signatures may improve diagnostic precision and guide personalized, mechanism-based interventions for trauma-related psychopathology.