Multidimensional prognostic factors for chronic low back pain-related disability: a longitudinal study in a Saudi population

BACKGROUND CONTEXT: Chronic low back pain (CLBP) is a major health problem. Identifying prognostic factors is essential for identifying people at risk of developing CLBP-related disability. PURPOSE: To examine associations between CLBP-related disability at 12-month follow-up and individual, psychosocial and physical factors at baseline, as well as treatment-related factors between baseline and 12-month follow-up among a Saudi population. Additionally, associations between pain intensity and general perceived efficacy (GPE) at 12 months were examined with the aforementioned factors. DESIGN: A prospective cohort study. PARTICIPANTS: One hundred Saudi participants over 18 years with a history of LBP greater than 3 months' duration. MAIN OUTCOME MEASUREMENTS: The primary outcome variable was CLBP-related disability measured by the Arabic Oswestry disability index. Secondary outcome measures were pain intensity over the prior week measured by the VAS and the participant's global perceptions of recovery (general perceived efficacy [GPE]) at 12 months. METHODS: At baseline (n=115), participants completed questionnaires covering demographics, disability, pain intensity, back beliefs, fear avoidance, psychological distress, and physical activity. They performed standardized physical performance tests, including assessment of pain behaviors using a pain behavior scale. After 12 months, participants (n=100) completed questionnaires on disability, pain intensity, GPE and provided treatment-related information during the previous year. Predictors of disability, pain, and GPE were explored using univariate and multivariate regression analyses. RESULTS: The prognostic model for moderate-severe CLBP-related disability at 12 months explained 53.0% of the variance. Higher pain intensity, higher fear-avoidance work, and older age predicted higher disability. Having no additional somatic symptoms predicted lower disability. Pain intensity at 12-month follow-up was explained by higher disability at baseline, while not being in paid employment appeared protective (25.7% of variance explained). As univariate associations were weak between predictor variables and GPE, multivariate analysis was not conducted. CONCLUSION: The study results supported the multifactorial nature of CLBP and reported an important prognostic model in the Saudi population.