Testing the effects of expressive writing on PTSD symptom reduction : examining the role of event centrality

Advisors: Holly K. Orcutt.Committee members: Anne Britt; Michelle Lilly; Patricia Wallace; Katja Wiemer; Kevin Wu.Includes illustrations.Includes bibliographical references.Many existing treatments for posttraumatic stress disorder (PTSD) have clients focus on a worst-identified traumatic experience, which can be challenging for individuals with repeated trauma exposure. It is assumed that, by focusing on the worst-identified trauma, individuals will develop skills that improve their ability to cope with other, less emotionally-provocative incidents. Yet to be determined, however, is whether an event's personal significance affects treatment outcomes. The present study aimed to evaluate how individual differences in event centrality (i.e., the extent to which an event impacts one's identity) influenced the effectiveness of a 3-session expressive writing (EW) intervention for posttraumatic stress symptoms (PTSS). EW has been shown to reduce PTSS severity and negative trauma-related appraisals are believed to mediate this relationship. Due to high attrition at the 3-week follow-up assessment (71.6%), main study hypotheses were examined in a modified intent-to-treat sample (N = 159) using data from baseline to post-treatment. Contrary to expectation, no within- or between-group differences in negative appraisals across sessions were observed. Changes in negative appraisals did not mediate the relationship between written disclosure condition (EW vs. control) and changes in PTSS severity, although negative appraisals were significantly related to PTSS severity. Sensitivity analyses with study completers yielded similar findings. Event centrality did not moderate those direct or indirect effects. A supplementary analysis using follow-up data from study completers (N = 38) revealed no within- or between-group differences in negative appraisals or PTSS severity across time. Post hoc survival analyses showed similar rates of attrition across conditions and attrition was not predicted by demographic characteristics or reactions to study participation. Implications and study limitations are discussed.Ph.D. (Doctor of Philosophy