Pain and clinical findings in the low back: a study of industrial employees with 5-, 10-, and 28-year follow-ups

Abstract

Little is known about the relationships of clinical findings in the low back with low back pain (LBP) in the normal working population. We studied whether physiotherapist's findings in the low back were associated with local and radiating LBP among a cohort (n=902) of employees in the engineering industry. A systematic non-proportional sample was drawn in strata by age, gender, and occupational class. The non-proportionality aimed at increasing sample size in smaller strata. Physiotherapists performed the straight-leg raising test (SRL), and made assessments of the fingertip-to-floor distance and pain in palpation of the lumbar interspinous spaces. The variables on pain at the interspinous spaces and the SRL tests were entered in cluster analysis. Three clusters emerged: no, minor, and severe clinical findings. In logistic regression analysis at baseline, limited forward flexion and the clinical findings cluster variable were associated with local and, particularly, radiating LBP. Follow-ups of the occurrence of local and radiating LBP at 5, 10, and 28 years from baseline were made. At the 5-year follow-up among subjects with no radiating LBP at baseline, the OR of radiating LBP for the clusters of minor and severe clinical findings compared to no findings were 2.7 (95% CI 1.4-5.1) and 3.8 (2.0-6.9), respectively, adjusted for age, gender, and occupational class. At the 10-year follow-up, the latter cluster predicted new reports of radiating LBP (1.9; 1.0-3.1) and of local LBP (4.1; 1.9-9.0, among subjects with no local LBP at baseline), similarly adjusted. No associations between limited forward flexion and new cases of LBP were observed. Thus, membership in clusters with different levels of findings in simple clinical measurements predicted new reports of radiating LBP, in particular, among employees. The generalizability of the results may be limited due to the characteristics of the sampling.