DOI: 10.1080/16501970510040344

Objective: To investigate whether personal and work-related factors, physical performance and back-specific questionnaires predict return to work. Design: A prospective study identifying prognostic factors for return to work. Subjects: Ninety-three patients sick-listed for 8–12 weeks for non-specific sub-acute low back pain included in a randomized controlled trial. Methods: Patients were examined with regard to demographic variables, a battery of back-specific questionnaires and physical tests before entering a randomized controlled trial. A stepwise backward Cox regression model was established to identify the most powerful predictors. Results: During follow-up 78. 5% of the patients have returned to full-time work. Fear-avoidance beliefs for work (relative risk (RR) for 1 SD change 0. 49; 95% confidence interval (CI) 0. 38–0. 64), disability (RR 1. 39, 95% CI 1. 02–1. 88) and cardiovascular fitness (RR 1. 42, 95% CI 1. 12–1. 79) were identified as the best predictors for return to work. The prevalence of correct predictions was 69. 3%. Conclusion: The predictors identified in the present study may reflect personal risk factors in a patient who gets acute low back pain. On the other hand, they may support that fear of pain and injury may be more disabling than pain itself, and that deconditioning is a result of altered behaviour reflecting attitudes towards low back pain in society, and information and advice given in primary healthcare.