Purpose: Lumbar paraspinal intramuscular fat (IMF) has emerged as a biological factor in low back pain (LBP). Traditional assessments measure IMF across the entire muscle or at specific levels and may miss key information on the role of IMF in LBP. Despite known variations across the lumbar spine, the three-dimensional (3D) distribution of IMF has not been characterized across people. Here we develop a template-based spatial parametric mapping approach to explore the 3D spatial distribution of lumbar paraspinal IMF.
Methods: To generate a lumbar spine template, we used T2-weighted magnetic resonance imaging from 76 people who recently recovered from LBP. We spatially normalized fat probability maps from Gaussian mixture modeling to the template and then calculated group-level spatial parametric maps of IMF and the associations between IMF and age, BMI, and sex.
Results: The template had clear delineation of the neural, vertebral, and muscular structures. We observed muscle-specific and spatially varying 3D patterns of IMF across the lumbar spine along with age-, BMI-, and sex-related associations. For the lumbar multifidus and erector spinae, IMF increased inferiorly and was greatest at the anterior-medial muscle borders, while psoas major IMF was more evenly distributed. The associations between IMF and age, BMI, and sex varied spatially with both positive and negative associations present within an individual muscle.
Conclusion: The developed 3D spatial parametric mapping approach provides a comprehensive assessment of lumbar paraspinal IMF, potentially enhancing our understanding of lumbar spine function and pathology, treatment mechanisms, and the modifiable factors promoting recovery from LBP.
Keywords: Adipose tissue; Computer-assisted; Image processing; Low back pain; Magnetic resonance imaging; Muscles.
© 2024. The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.