STUDY DESIGN

A cross-sectional population-based study of paraspinal muscle asymmetry.

OBJECTIVE

To examine level- and side-to-side differences in paraspinal muscle areas on magnetic resonance images in a population-based sample of middle aged men without low back pain (LBP).

SUMMARY OF BACKGROUND DATA

Level- and side-specific multifidus muscle atrophy and fat infiltration have been suggested as possible markers for localized spinal pathology and LBP, but prior studies have limited generalizability due to small sample sizes, young age of samples and measurement issues.

METHODS

From a general population sample of 600 twin men, those reporting no LBP during the prior year, no previous spinal fractures, and no bed rest for at least 1 week in the last 12 months were included in the study. All subjects had T2-weighted axial images available for the three lowest lumbar levels. Both total and fat-free cross-sectional areas (CSAs) of the multifidus and erector spinae muscles at the mid-disc level were measured. Intrarater reliability ranged between 0.90 and 0.98 for area measurements and 0.86 and 0.92 for measurements of side-to-side differences. Data were analyzed using descriptive statistics and paired t tests.

RESULTS

Subjects consisted of 126 men whose mean multifidus total CSA measurements varied between 7.3 and 11.1 cm and between 6.9 and 10.8 cm for right and left sides, respectively, depending on the level. The corresponding mean areas for erector spinae were 9.4 to 19.6 cm2 for right side and 10.4 to 19.7 cm2 for left side. The multifidus was larger on the right side than on the left side in 65% to 68% of subjects, depending on spinal level (P < 0.001). The mean asymmetry at the three lowest lumbar levels was 10% to 13.2% and was smallest at L4-L5. Multifidus side-to-side asymmetry ranged from 0.1% to 44.3%. For erector spinae, the left-side measurements tended to be larger, reaching statistical significance (P < 0.0001) for the two lowest levels. The mean side-to-side asymmetry increased caudally for erector spinae, from 8.2% to 18.8% and was significantly different between adjacent levels (P < 0.01). The amount of intramuscular fat significantly increased caudally for both muscles.

CONCLUSION

Paraspinal muscle asymmetry greater than 10% was commonly found in men without a history of LBP. This suggests caution in using level- and side-specific paraspinal muscle asymmetry to identify subjects with LBP and spinal pathology.