To examine the association between structural abnormalities recorded on magnetic resonance imaging (MRI) and outcome after evidence-based conservative treatment in patients with chronic nonspecific low back pain (LBP).
Summary of Background Data.
In most guidelines for the management of LBP, MRI is not recommended unless the diagnostic triage suggests serious spinal pathology or nerve root involvement for which surgical treatment is foreseen. This is because many structural changes seen on MRI appear to be as common in asymptomatic individuals as in people with LBP and are, therefore, considered of little value in either explaining the cause of pain or deciding the subsequent course of management. However, to our knowledge, no studies have assessed whether the presence of such MRI abnormalities influences the outcome of the conservative treatment that patients with chronic nonspecific LBP typically receive.
T2-weighted, 4-mm spin-echo MRI sequences of the lumbar spine were obtained from 53 patients with chronic nonspecific LBP before a 3-month program of exercise therapy. Disc degeneration, disc bulging, high intensity zones, and endplate/bone marrow changes were assessed for each lumbar segment. Back pain (average and worst) and disability (Roland Morris score) were assessed before and after therapy, and 12 months later, and the improvements were examined in relation to the presence or absence of baseline MRI “abnormalities.”
Eighty-nine percent of patients had severe disc degeneration (grade 4 or 5), 74% had disc bulging, 60% had high intensity zones, and 62% had endplate/bone marrow changes in at least 1 lumbar segment. Only 11% patients had none of these changes at any level. The MRI abnormalities showed only minimal association with baseline symptoms. In multivariate regression analyses, in which age, gender, and baseline symptoms were controlled for, only 1 significant association between the MRI variables and outcome was observed: the presence of a high intensity zone in any vertebral segment was associated with lower average pain at the 12-month follow-up (standardized β −0.376, P = 0.006, 16.5% variance accounted for).
In the patient group examined, the presence of common “structural abnormalities” on MRI had no significant negative influence on the outcome after therapy.