0881 Cognitive-Behavioral Therapy for Insomnia in PTSD: Differential Relationships with Symptom Clusters

Abstract

Introduction: Insomnia is highly prevalent in posttraumatic stress disorder (PTSD), and is associated with greater severity and decreased likelihood of remission. Sleep problems often persist after treatment for PTSD, and treating insomnia has been shown to ameliorate both PTSD and insomnia symptoms. Evidence also suggests that PTSD symptom clusters may exhibit differential relationships with sleep problems. In particular, hyperarousal is thought to impact sleep through increased physiological arousal. However, less is known about potential differential effects of cognitive-behavioral therapy for insomnia (CBT-I) on PTSD symptoms. Thus, we investigated the effect of CBT-I on PTSD DSM-IV symptom clusters. As we were also interested in differential effects of PTSD symptoms on treatment, we examined the effect of baseline symptoms on reduction in insomnia severity following intervention. Methods: Women aged 18-64 with PTSD and insomnia were included as part of a larger randomized controlled trial of CBT-I in individuals with PTSD following interpersonal violence exposure (N=81, M age=37.15, SD=10.93). Participants completed an initial screening and were randomized to receive either CBT-I or Attention Control. Participants were administered the Clinician Administered PTSD Scale (CAPS), the Insomnia Severity Index (ISI), and the Hamilton Rating Scale for Depression. Sleep items were excluded from the CAPS criterion D score. A series of ANOVA models were used to examine change in CAPS cluster scores between groups. Hierarchical linear regression was used to examine relationship between baseline symptoms and change in ISI. Results: Re-experiencing (F(1, 79)= 9.325, p= .003) and avoidance/ numbing (F(1, 79)= 3.996, p= .049) symptoms significantly improved in the CBT-I group as compared to the control group. CBT-I did not lead to improvements in hyperarousal symptoms (F(1, 77)= 1.873, p= .175). Baseline avoidance/numbing symptoms were significant in predicting improvements in insomnia following CBT-I (F(4, 29)= 5.466, p= .026). Conclusion: Findings suggest that the effect of CBT-I on PTSD may be due to reduction in re-experiencing and avoidance symptoms. Surprisingly, CBT-I did not improve hyperarousal symptoms. Findings highlight the importance of examining potentially different mechanisms underlying the relationship between insomnia and PTSD symptoms.