(PS11-C74) Examining Between- and Within-session Extinction as Treatment Mechanisms of Prolonged Exposure

Background: Between-session and within-session extinction, clinically referred to as habituation, are two proposed mechanisms of prolonged exposure (PE) for posttraumatic stress disorder (PTSD; Foa et al., 2006). Empirical work has found mixed evidence for this, with stronger evidence for between-session extinction as a predictor of recovery (Cooper et al., 2017). One paper (N = 14) using multilevel modeling, examined between- and within-session processes via imaginal exposure subjective units of distress (SUDs) across 10-12 sessions of PE (Sripada & Rauch, 2015). This analytic technique is useful for this context given that it does not assume an equal number of observations, allowing for inclusion of participants with differing lengths of exposures and numbers of sessions. They found that symptom change and treatment response predicted the slope of SUDS between sessions, but not within sessions, showing evidence of only between-session extinction as a potential mechanism of PE. The current study sought to replicate this study with a larger sample, combined across two doubly randomized preference trials. We hypothesized that pre- to post-treatment decreases in interviewer-rated PTSD scores and treatment response would predict the slopes of between-session SUDS ratings, but not within-session SUDS ratings.

Method: Participants (N = 96) were in the PE arm of a clinical trial who had completed at least one session involving imaginal exposure (age M = 35.77, SD = 11.45; 80.4% women; 68.0% White). Over the course of treatment, participants completed 1 – 8 sessions that included imaginal exposure (session M = 6.55, SD = 2.45), depending on dropout. During exposures participants rated SUDS (0 – 100) every 5 min for a max. of 60 min (exposure length M = 40.30, SD = 10.97). PTSD severity (PSS-I; Foa et al., 1993) and clinical global impression (CGI; Guy, 1976) were assessed via interview at baseline and post-treatment; responder status was defined as a post-treatment PSS-I score of < 24 and a global improvement score < 4. Mixed effects modeling was conducted, with treatment response and change in PSS-I scores included as predictors in separate models. Additional data from the second trial (N = 149) will be combined and used for cross validation.  

Results &

Discussion: Consistent with emotional processing theory (Foa et al., 2006), reductions in PTSD symptoms (t = -16.21, p < .001) and treatment response (t = -4.39, p < .001) predicted the slope of SUDS across sessions. Contrary to Sripada and Rauch’s (2015) findings, PTSD symptom change (t = -2.75, p = .006), but not treatment response, also predicted the slope of SUDs within sessions.

Implications:  Between-session extinction may be a key process of recovery from PTSD during PE, with mixed support for within-session extinction as a necessary component of recovery. Clinicians may want to normalize a lack of distress reduction within a given imaginal exposure, while looking to more global across-session decreases in SUDs as a more consistent marker of recovery. Additional research taking an idiographic approach to examining indicators of extinction (e.g., SUDS, cortisol levels) will enable identification of potential latent profiles of extinction, potentially helping to identify diverse pathways to recovery.