Risk and Prognostic Factors of Low Back Pain: Repeated Population-based Cohort Study in Sweden

Abstract

Study Design. Prospective longitudinal cohort study. Objective. To determine the associations for workload and health-related factors with incident and recurrent low back pain (LBP), and to determine the mediating role of health-related factors in associations between physical workload factors and incident LBP. Summary of Background Data. It is not known whether the risk factors for the development of LBP are also prognostic factors for recurrence of LBP and whether the associations between physical workload and incident LBP are mediated by health-related factors. We used data from the Swedish Longitudinal Occupational Survey of Health study. Those responding to any two subsequent surveys in 2010 to 2016 were included for the main analyses (N = 17,962). Information on occupational lifting, working in twisted positions, weight/height, smoking, physical activity, depressive symptoms, and sleep problems were self-reported. Incident LBP was defined as pain limiting daily activities in the preceding three months in participants free from LBP at baseline. Recurrent LBP was defined as having LBP both at baseline and follow-up. For the mediation analyses, those responding to three subsequent surveys were included (N = 3516). Methods. Main associations were determined using generalized estimating equation models for repeated measures data. Mediation was examined with counterfactual mediation analysis. Results. All risk factors at baseline but smoking and physical activity were associated with incident LBP after adjustment for confounders. The strongest associations were observed for working in twisted positions (risk ratio = 1.52, 95% CI 1.37, 1.70) and occupational lifting (risk ratio = 1.52, 95% CI 1.32, 1.74). These associations were not mediated by health-related factors. The studied factors did not have meaningful effects on recurrent LBP. Conclusion. The findings suggest that workload and health-related factors have stronger effects on the development than on the recurrence or progression of LBP, and that health-related factors do not mediate associations between workload factors and incident LBP.