Effects of Pain Reprocessing Therapy on Attributed Causes of Chronic Back Pain

Abstract

Fear of pain is a modifiable risk factor that predicts the development and persistence of chronic pain. It is due in part to patients' beliefs that pain reflects injury or tissue damage—though this is often not the case. We recently developed a novel psychological treatment, Pain Reprocessing Therapy (PRT), aiming to help patients reattribute their pain as due to mind/brain rather than peripheral tissue causes. Here, we tested whether such changes in pain attributions, as revealed by natural language measures of pain attributions, may serve as a psychological mechanism of PRT. We randomized chronic back pain patients (N = 151) to PRT, open-label placebo, or usual care, with large benefits observed for PRT vs. both comparison groups (Ashar et al. 2021). In this secondary analysis, blinded raters coded natural language attributions provided at pre- and post-treatment as either peripheral damage attributions (e.g., "disc") or mind/brain processes (e.g., "fear") attributions. Pre-treatment, only 10% of attributions were coded as mind/brain. PRT led to large increases in mind/brain attributions compared to both placebo, g = 1.74, and usual care, g = 1.64, both p < .001, with pre-to-post-treatment increases in mind/brain attributions correlating with pain intensity reductions within the PRT group, r(42) = -.34, p = .02. These increases in mind/brain attributions also partially mediated the effect of PRT vs. control on pain intensity at 1-year follow-up, βpath-ab = -0.37, p < .05. Within the PRT group, we observed marked decreases in the prevalence of peripheral-tissue related words (e.g., "activity", "posture") and increases in mind/brain-related words (e.g., "fear", "anxiety"). Overall, results suggest that patients largely attributed their pain to peripheral tissue causes at pre-treatment, that reattributing pain to mind/brain causes may be a mechanism underlying successful PRT treatment, and that natural language-based approaches may offer valuable insight into relevant pain beliefs. This study was funded by NIH grant R01 DA035484 (T.D.W.), R01 MH076136 (T.D.W.), NIH/NCATS Grant # TL1-TR-002386 (Y.K.A.), RSNA Presidents Circle Research Resident Grant, 2017 (T.F.F.), German Research Foundation (DFG) GE 2774/1-1 (S.G.), the Psychophysiologic Disorders Association, the Foundation for the Study of the Therapeutic Encounter, and community donations.