Predictors of Response to Prolonged Exposure, Sertraline, and Their Combination for the Treatment of Military PTSD

Objective: The current study is an analysis of predictors of posttraumatic stress disorder (PTSD) treatment response in a clinical trial comparing (1) prolonged exposure plus placebo (PE + PLB), (2) PE + sertraline (PE + SERT), and (3) sertraline + enhanced medication management (SERT + EMM) with predictors including time since trauma (TST), self-report of pain, alcohol use, baseline symptoms, and demographics. Methods: Participants (N = 196) were veterans with combat-related PTSD (DSM-IV-TR) of at least 3 months' duration recruited between 2012 and 2016 from 4 sites in the 24-week PROlonGed ExpoSure and Sertraline (PROGrESS) clinical trial (assessments at weeks 0 [intake], 6, 12, 24, 36, and 52). Results: Across treatment conditions, (1) longer TST was predictive of greater week 24 PTSD symptom improvement (beta = 1.72, P = .01) after adjusting for baseline, (2) higher baseline pain severity was predictive of smaller symptom improvement (beta = -2.96, P = .003), and (3) Hispanic patients showed greater improvement than non-Hispanic patients (beta = 12.33, P = .03). No other baseline characteristics, including alcohol consumption, were significantly predictive of week 24 improvement. Comparison of TST by treatment condition revealed a significant relationship only in those randomized to the PE + SERT condition (beta = 2.53, P = .03). Longitudinal analyses showed similar results. Conclusions: The finding that longer TST shows larger symptom reductions is promising for PTSD patients who might not seek help for years following trauma. Higher baseline pain severity robustly predicted attenuated and slower response to all treatment conditions, suggesting a common neuropathologic substrate. Finally, in the current study, alcohol use did not impede the effectiveness of pharmacotherapy for PTSD.