OBJECTIVESTo evaluate the relationship between Modic change and disc height together with lumbar hyperosteogeny and study the role of Modic change in lumbar degeneration.

METHODSThe imaging data of 150 elderly patients with chronic low back pain were analysed retrospectively. All patients underwent MRI and lumbar lateral X-ray examination. The lumbar disc from L1-L2 to L5-S1 were selected for this study, including 750 discs, vertebral and endplate close to disc in 150 patients. The incidence rate of lumbar endplate Modic change, disc height and the degree of vertebral bone hyperplasia were recorded. The ratio of disc height/lumbar intervertebral disc height < 50% was defined as disc collapse. The patients were divided into 4 groups in the basis of imaging changes. Group A1:disc collapse without severe lumbar hyperosteogeny; Group A2: disc collapse with severe lumbar hyperosteogeny; Group B1: Neither disc collapse nor severe lumbar hyperosteogeny; Group B2: severe lumbar hyperosteogeny without disc collapse. The incidence rates of Modic change were compared between the 4 groups by χ(2) test. Finally, the influence of disc height and vertebral bone hyperplasia on the incidence rate of Modic change was analysed.

RESULTSFour groups of patients observed a total of 750 discs. The number of intervertebral discs in the group A1 was 208, the incidence rate was 54.3%. The number of intervertebral discs in the group A2 was 135, the incidence rate of group A2 was 34.8%. The number of intervertebral discs in the B1 group was 225, the incidence rate of group B1 was 16.9%. The number of intervertebral discs in the B2 group was 182, the incidence rate of group B2 was 29.7%. There was significant difference of lumbar endplate Modic change incidence rate among the 4 groups(χ(2) = 69.565, P < 0.05). The results of post hoc test showed that the incidence rate of Modic change in group A1 was higher than group A2, B1 and B2 (χ(2) = 12.524, 66.701 and 24.102, P < 0.00714). There was significant difference of Modic change incidence rate between group A2 and B1(χ(2) = 15.032, P < 0.00714), but there was no significant difference of Modic change incidence rate between group A2 and B2 (χ(2) = 0.945, P > 0.00714) . There was significant difference of Modic change incidence rate between group B2 and group B1 (χ(2) = 9.395, P < 0.00714).

CONCLUSIONSThe incidence rate of Modic change with disc collapse but without severe lumbar hyperosteogeny is high in elderly patients with chronic low back pain. There is no significant difference of Modic change incidence between patients with both disc collapse and severe lumbar hyperosteogeny and patients with severe lumbar hyperosteogeny but without disc collapse.