Integrated Treatment for Co-occurring Eating Disorders and Posttraumatic Stress Disorder

Standard ED treatment is characterized by partial response in many individuals, and the risk of relapse is high. This, in combination with significant rates of comorbid PTSD, evidence that trauma history can negatively impact ED treatment outcome, and the observation that PTSD may be a central ED maintaining factor in these individuals, has led us to develop a program of research aimed at demonstrating that the alleviation of PTSD symptoms concurrent with ED treatment facilitates a fuller and sustained ED recovery for these individuals. The specific aims of this initial randomized controlled trial (RCT) are to determine: (1) the immediate and short-term efficacy of concurrent evidence-based ED and PTSD treatment in reducing PTSD symptoms relative to standard ED treatment; and (2) estimates of the effect sizes of concurrent treatment on ED symptomatology relative to standard ED treatment at post-treatment and follow-up. Primary Hypothesis: 1. Concurrent CBT for ED and PTSD will result in significantly greater improvements in clinician-rated PTSD symptoms at post-treatment, as well as at 3- and 6-month follow-up compared with CBT for ED alone. Secondary Hypotheses: 2. Concurrent CBT for ED and PTSD will result in significantly greater improvements in self-reported PTSD symptoms at post-treatment, as well as at 3- and 6-month follow-up compared with CBT for ED alone. 3. Concurrent CBT for ED and PTSD will result in significantly greater improvements in concomitant symptoms (e.g., anxiety and depression) at post-treatment, as well as at 3- and 6-month follow-up compared with CBT for ED alone. Exploratory Research Questions 1. What proportion of participants experience reliable improvement in ED symptomatology and what proportion experience reliable deterioration in ED symptomatology over the course of each treatment? 2. What is the relative efficacy of concurrent CBT for ED and PTSD versus CBT for ED alone in reducing functional deficits at post-treatment, and 3- and 6-month follow-up? 3. What is the relative efficacy of concurrent CBT for ED and PTSD versus CBT for ED alone in improving other concomitant symptoms at post-treatment, and 3- and 6-month follow-up? METHOD: Following a course of intensive ED treatment in the inpatient or day hospital ED program at University Health Network (UHN), participants will be randomly assigned to receive either individually-administered: (1) standard CBT for ED alone or (2) concurrent CBT for ED and PTSD. Participants in both conditions will receive 16 sessions. Eleven of the 16 concurrent CBT for ED and PTSD sessions will be 90 minutes long in order to accommodate the administration of CBT for ED and PTSD. Participants in the CBT for ED alone condition will also receive eleven 90-minute sessions in order to control for time in therapy. In the CBT for ED condition, therapists will have more time to cover the ED-related content. Accordingly, both treatments will consist of eleven 90-minute sessions and five 50-minute sessions, with two sessions per week for the first 8 sessions, weekly sessions for the following 6 sessions, and biweekly for the final two sessions. Participants will be assessed pre-CBT for ED/PTSD and post-CBT for ED/PTSD, as well as 3 and 6 months after treatment. The assessors will be blind to treatment condition, and an intent-to-treat (ITT) approach to data collection and analysis will be adopted. ED and PTSD symptoms will also be assessed on a weekly basis during the study treatments. STATISTICS: 1. Primary hypothesis: Changes in clinician-rated PTSD symptoms will be analyzed using multilevel modeling, including a random effect for time and fixed effects for therapy condition, time, and the condition by time interaction. Change in PTSD symptoms over time and differences between the conditions over time will be examined. Empirically derived covariates that significantly reduce error variance will be used. Cohen's d effect sizes will be calculated. 2. Secondary hypotheses: Changes in self-reported PTSD symptoms, anxiety, and depresion will be analyzed using multilevel modeling, including a random effect for time and fixed effects for therapy condition, time, and the condition by time interaction. Change in PTSD symptoms, anxiety, and depression over time and differences between the conditions over time will be examined. Empirically derived covariates that significantly reduce error variance will be used. Cohen's d effect sizes will be calculated. 3. The proportion of participants who experience reliable improvement and the proportion who experience reliable deterioration in ED symptomatology in each therapy condition will be calculated using reliable change criteria (Jacobson \& Truax, 1991). 4. The proportion of participants meeting diagnostic criteria for PTSD and an eating disorder will be compared between conditions at post-treatment and follow-up. 5. Changes in other concomitant symptoms and functional deficits will be analyzed using multilevel modeling as described above.