A personal construct investigation of posttraumatic stress and the healing nightmare phenomenon in military veterans

The study sought to address potential relationships between aspects of personal construct systems and Posttraumatic Stress Disorder (PTSD) symptom severity and to explore the “healing nightmare” phenomenon (Wilmer, 1986, 1996). Wilmer observed that PTSD symptoms were most severe in veterans whose dreams contained veridical replays of the trauma; symptoms decreased as dreams incorporated non-veridical elements of trauma, with metaphorical/bizarre dreams associated with the least amount of symptomatology. The study attempted to elaborate this phenomenon from a personal construct perspective (Kelly, 1955/1991a, 1955/1991b). Twenty-five male veterans associated with the Malcom Randall VA Medical Center (Gainesville, Florida) completed a Role Construct Repertory Grid; the Posttraumatic Diagnostic Scale (PDS; Foa et al., 1997); and a measure eliciting details about the most recent trauma-related dream. Participants also completed the Combat Exposure Scale (Keane et al., 1989), the Life Events Checklist (Gray et al., 2001), and a Social Support Questionnaire (Sarason et al., 1987). Current self-ideal similarity, degree of negativity of construing, and self-reported construct meaningfulness were correlated with severity of PTSD symptoms at a statistically significant level; degree of negativity and current self-ideal similarity remained statistically significant in multiple regression analyses controlling for combat exposure, social support, and overall trauma. Construct system integration (assessed by ordination and chi-square measures) and valence of emergent constructs were not associated with symptom severity. Data partially supported the “healing nightmare”; symptom severity was associated with self-report measures of content and emotional similarity between the reported dream and the trauma. Symptom severity was not significantly associated with a blind coding measure of dream content. Relationships between symptom severity and self-report measures were not statistically significant when controlling for combat exposure, social support, and overall trauma. Components of personal construct systems were generally not associated with dream content although negativity of construing was associated with self-report measures of content and emotional similarity. These relationships were not statistically significant when controlling for combat exposure, social support, and overall trauma. Discussion includes theoretical and clinical implications of the study, including how findings may inform working with posttraumatic stress from a personal construct perspective. Methodological limitations and directions for future research are elaborated.