Background context: Although the effect of lumbar spinal stenosis (LSS) on the lower extremities is well documented, limited research exists on the effect of spinal stenosis on the posterior paraspinal musculature (PPM). Similar to neurogenic claudication, moderate to severe spinal canal compression can also interfere with the innervation of the PPM, which may result in atrophy and increased fatty infiltration (FI).

Purpose: This study aims to assess the association between LSS and atrophy of the PPM.

Study design: Retrospective cross-sectional study.

Patient sample: Patients undergoing MRI scans at a tertiary orthopedic center for low back pain or as part of a preoperative evaluation.

Outcome measures: The functional cross-sectional area (fCSA) and percent fatty infiltration (FI) of the PPM at L4.

Methods: Lumbar MRIs of patients at a tertiary orthopedic center indicated due to lower back pain (LBP) or as a presurgical workup were analyzed. Patients with previous spinal fusion surgery or scoliosis were excluded. LSS was assessed according to the Schizas classification at all lumbar levels. The cross-sectional area of the PPM was measured on a T2-weighted MRI sequence at the upper endplate of L4. The fCSA and fatty infiltration (FI) were calculated using custom software. Crude differences in FI and fCSA between patients with no stenosis and at least mild stenosis were tested with the Wilcoxon signed-rank test. To account for possible confounders, a multivariable linear regression model was used to adjust for age, sex, body mass index (BMI), and disc degeneration. A subgroup analysis according to MRI indication was performed.

Results: A total of 522 (55.7% female) patients were included. The median age was 61 years (IQR: 51 – 71). The greatest degree of moderate and severe stenosis was found at L4/5, 15.7%, and 9.2%, respectively. Stenosis was the least severe at L5/S1 and was found to be 2% for moderate and 0.2% for severe stenosis. The Wilcoxon test showed significantly increased FI of the PPM with stenosis at any lumbar level (p < 0.001), although no significant decrease in fCSA was observed. The multivariable regression model showed a significant increase in FI with increased LSS at L1/2, L2/3, and L3/4 (p = 0.013, p <0.01 and p =0.003). The severity of LSS at L4/5 showed a positive association with the fCSA (p = 0.019). The subgroup analysis showed, the effect of LSS was more pronounced in non-surgical patients than in patients undergoing surgery.

Conclusions: In this study, we demonstrated a significant and independent association between LSS and the composition of the PPM, which was dependent on the level of LSS relative to the PPM. In addition to neurogenic claudication, patients with LSS might be especially susceptible to axial muscle wasting, which could worsen LSS due to increased spinal instability, leading to a positive feedback loop.