An identification of low back pain groups using biobehavioral variables.

A multivariate predictive model of low back pain (LBP) was developed. Following a semi-structured interview, 73 participants were assigned to dysfunctional chronic low back pain (DCLBP), functional chronic low back pain (FCLBP), acute low back pain (ALBP), and healthy control (HC) groups. All participants underwent a comprehensive physical, psychophysiological, and psychological evaluation. Multivariate analyses indicated no psychophysiological, few physical, and many psychological differences among the groups. The DCLBP group was found to be most impaired in flexion (p<.001), and the HC group performed the most total work (ft-lb) in extension (p<.001). Psychologically, the DCLBP group displayed greater levels of emotional distress and characterological disturbances and were more functionally impaired (p<.001). Few differences between FCLBP and HC were found. A classification analysis using physical and psychological variables correctly classified 83.3% of DCLBP patients, and it was found that the ALBP group was heterogeneous with some patients having a dysfunctional profile and other patients having a functional profile. The psychological variables were more potent predictors of group membership than were the physical variables. These findings indicate that potential DCLBP and FCLBP patients can be identified shortly following an injury, suggesting important implications for assessment and treatment of low back pain in general, and more specifically, for reducing health care costs and human suffering.