Day-to-day dynamics of facial emotion expressions in posttraumatic stress disorder

Emotion-related posttraumatic stress disorder (PTSD) symptoms are highly heterogenous. Efforts to understand diverse and even opposing symptoms (e.g., heightened arousal, emotional numbing) have been limited by methods that fail to capture the complex, dynamic nature of emotion. In this study, we addressed these limitations by using data-driven methods to identify patterns of facial emotional expressions and PTSD symptoms that characterize the daily functioning of trauma-exposed individuals. We studied a sample of World Trade Center responders (N=112) with PTSD pathology who recorded a daily video diary and self-reported symptoms for 90 days. Facial expressions and head movements were detected from the video recordings with a validated facial emotion recognition model. We used idiographic network models to estimate each person’s day-to-day expressivity and symptom patterns. The model identified six dynamics that were significant for >10% of the sample: same-day associations between changes in symptoms and changes in (1) emotional arousal, (2) scared expressions, (3) neutral expressions, and (4) head movement, as well as the cross-lagged predictions of next-day symptoms from (5) angry and (6) neutral expressions. Each path showed substantial heterogeneity (positive for 4-7% and negative for 4-12% of participants). Our results suggest there are people whose PTSD symptoms manifest in increased expressivity and emotional dysregulation, others who become less expressive and suppress emotions, and some who expressed some and suppressed other emotions. Empirically derived expressivity profiles can suggest mechanisms that maintain PTSD, and also serve as objective targets for etiologic research and treatment development.