Abstract
OBJECTIVE: To describe long-term trajectories of back pain. DESIGN: Monthly data collection for 6 months at 7-year follow-up of participants in a prospective cohort study. SETTING: Primary care practices in Staffordshire, UK. PARTICIPANTS: 228 people consulting their general practitioners with back pain, on whom information on 6-month back pain trajectories had been collected during 2001-2003, and who had valid consent and contact details in 2009-2010, were contacted. 155 participants (68% of those contacted) responded and provided sufficient data for primary analyses. OUTCOME MEASURES: Trajectories based on patients' self-reports of back pain were identified using longitudinal latent class analysis. Trajectories were characterised using information on disability, psychological status and presence of other symptoms. RESULTS: Four clusters with different back pain trajectories at follow-up were identified: (1) no or occasional pain, (2) persistent mild pain, (3) fluctuating pain and (4) persistent severe pain. Trajectory clusters differed significantly from each other in terms of disability, psychological status and other symptoms. Most participants remained in a similar trajectory as 7 years previously (weighted κ 0.54; 95% CI 0.42 to 0.65). CONCLUSIONS: Most people with back pain appear to follow a particular pain trajectory over long time periods, and do not have frequently recurring or widely fluctuating patterns. The results are limited by lack of information about the time between data collection periods and by loss to follow-up. However, findings do raise questions about standard divisions into acute and chronic back pain. A new framework for understanding the course of back pain is proposed.