Differences in the Relationship Between Psychosocial Distress and Self-Reported Disability in Patients with Chronic Low Back Pain in Six Pain Rehabilitation Centers in the Netherlands

Abstract

A cross sectional multicenter study in six outpatient Rehabilitation Centers (RCs) in the Netherlands. This study aims to confirm or refute the finding that a strong relationship exists between psychosocial distress and self-reported disability in patients with nonspecific chronic low back pain (CLBP) by analyzing this relationship in patients with CLBP admitted for treatment in six RCs. A strong relationship between psychosocial distress and self-reported disability in patients with CLBP is suggested. However, in former research weak relationships were found in two of the RCs participating in this study. Total study sample consisted of 293 patients (30-66 per RC) with CLBP, admitted for outpatient multidisciplinary rehabilitation in one of the six participating RCs. Psychosocial distress was measured with the Symptom Checklist-90-Revised (SCL-90-R), self-reported disability with the Roland Morris Disability Questionnaire (RMDQ). Pearson correlation coefficients between psychosocial distress and self-reported disability were calculated. Multivariate regression analysis was performed to analyze the relationship between SCL-90-R and VAS pain (independent variables) and RMDQ (dependent variable) for the total group and for each RC separately. A multivariate regression analysis was performed to analyze the relationship between all baseline characteristics and RMDQ in the total group. Correlation coefficient between the SCL-90-R and RMDQ was r = 0.38 for the total sample, indicating a significant (P < 0.05), but weak relationship. For the six individual RCs, correlation coefficients ranged between r = 0.22 and 0.67 (three of the six correlation coefficients were significant). The explained variance (r) of the regression models (SCL-90 and pain intensity as predictors of RMDQ) was 29% for the total sample, and varied between the RCs from 17% to 52%. Results of the multivariate regression analysis of all baseline characteristics of the total group revealed that the model explained 36% of the total variance observed in RMDQ score. Overall, the contributions of psychosocial distress to the models were smaller and more variable compared with pain intensity. The overall relationship between psychosocial distress and self reported disability was weak, and differences between RCs were considerable. This indicates that the relationship between psychosocial distress and disability in patients with CLBP is not uniform.