DOI: 10.1097/brs.0000000000005881 ISSN: 0362-2436

Association of CT-Based Body Composition Phenotypes With Lumbar Degenerative Disease Characteristics

Qiao Xu, Xiaohong Wang, Mengchao Zhang, Liang Chen
Show PDF Cite

Study Design.

A two-center retrospective cross-sectional study.

Objective.

To investigate the association between CT-derived body composition phenotypes and lumbar degenerative disease (LDD) characteristics.

Summary of Background Data.

While individual measures of fat distribution and muscle status are known to be associated with LDD, their combined synergistic or antagonistic effects remain largely uncharacterized. Furthermore, conventional body mass index (BMI) often fails to accurately represent the complexity of individual body composition.

Methods.

A total of 262 patients with chronic low back pain were retrospectively included. Subcutaneous adipose tissue area (SAT), visceral adipose tissue area (VAT), abdominal muscle fat area (AMF), paraspinal muscle fat area (PMF), abdominal muscle area (AMA), and paraspinal muscle area (PMA) were quantified on axial CT images at the L3–L4 vertebral levels. After variable selection using Spearman correlation analysis ( P< 0.05) and variance inflation factor (VIF) diagnostics (VIF <5), K-means clustering was performed to derive distinct body composition phenotypes. Three BMI groups based on Chinese standards were used as the comparative classification. The primary outcome was the lumbar composite degeneration grade (grades I–IV). Trend tests were performed using the Jonckheere–Terpstra and Cochran–Armitage tests. Age- and sex-adjusted ordinal logistic regression was used to evaluate the association between phenotypes and degenerative grade.

Results.

VAT, AMF, PMF, AMA, and PMA were identified as key clustering variables. K-means clustering identified three distinct body composition phenotypes: high-fat muscle-rich (n=95), low-fat muscle-preserved (n=112), and fat-infiltrated muscle-atrophic (n=55). The three phenotypes exhibited significant gradient trends in degeneration grade, anterior osteophytes, posterior osteophytes, intervertebral space narrowing, and facet joint changes (all trend P< 0.01). After adjustment for age and sex, the fat-infiltrated muscle-atrophic phenotype was associated with higher degenerative grade (ordered logistic regression, OR=3.02, 95% CI 1.45–6.31, P =0.003). The Oswestry Disability Index (ODI) differed across phenotypes (trend P= 0.008), with the lowest adjusted ODI in the low-fat muscle-preserved phenotype (0.300±0.020) and similar adjusted ODI values in the high-fat muscle-rich (0.327±0.025) and fat-infiltrated muscle-atrophic phenotypes (0.333±0.031). In contrast, the conventional BMI groups showed no significant trend for any degenerative outcome (all P> 0.05). No significant differences were observed in supine sagittal parameters under either classification system.

Conclusion.

Body composition-based phenotypes are significantly associated with lumbar degenerative features, reflecting the synergistic and antagonistic interplay between fat distribution and muscle status. These phenotypes demonstrate superior stratification performance for lumbar degenerative disease compared with conventional BMI.

More from our Archive