Journal Article Ahead of print
PubMed: 42374333 DOI: 10.1186/s12891-026-09893-9
Association of trabecular texture and paraspinal muscle characteristics with prevalent vertebral fractures - QCT results from a subcohort of the AGES population.
BMC Musculoskelet Disord
Authors: Tobias Stumpf, Oliver Chaudry, Jana Hummel, Sandra Freitag-Wolf, Eren Yilmaz, Stefan Bartenschlager, Sigurdur Sigurdsson, Vilmundur Gudnason, Nicolai R Krekiehn, Claus C Glüer, Klaus Engelke
Affiliations
- Institute of Radiology, Friedrich-Alexander-University Erlangen-Nürnberg and Universitätsklinikum Erlangen, Erlangen, Germany.
- Department of Medicine 3, Friedrich-Alexander-University Erlangen-Nürnberg and Universitätsklinikum Erlangen, Erlangen, Germany.
- Institute of Medical Informatics and Statistics, Kiel University and University Hospital Schleswig-Holstein, Kiel, Germany.
- Section Biomedical Imaging, Department of Radiology and Neuroradiology, Kiel University and University Hospital Schleswig-Holstein, Kiel, Germany.
- Icelandic Heart Association Research Institute, Kopavogur, Iceland.
- Department of Medicine 3, Friedrich-Alexander-University Erlangen-Nürnberg and Universitätsklinikum Erlangen, Erlangen, Germany.
Abstract
BACKGROUND
Prediction of osteoporotic fractures remains suboptimal, leaving many high-risk individuals untreated while others fracture without prior diagnosis. Muscle characteristics are not part of prediction algorithms although studies showed that paraspinal muscle size, density, and fat infiltration are associated with vertebral fractures (VF). However, an analysis of the prospective AGES-Reykjavik study did not show improved prediction of the first incident VF when combining muscle parameters with BMD.
OBJECTIVE
To determine whether paraspinal muscle characteristics are associated with prevalent fractures using the prospective AGES-Reykjavik study.
METHODS
In the current study associations of muscle parameters with prevalent VF were analyzed in the same cohort of 486 women and 340 men with 96 and 78 prevalent VF, respectively. 50 parameters from CT scans of the L1 and L2 vertebrae, divided into a BMD, a trabecular texture and a muscle subset were used as discriminators. Each subset also included age and BMI. The number of parameters was reduced using stepwise logistic regression to create multivariable fracture discrimination models. Model accuracy was assessed using the likelihood ratio test (LRT) and the area under the curve (AUC). Bootstrap analyses were performed to assess stability of the model selection process.
RESULTS
23 parameters significantly discriminated prevalent VFs univariately in women and 5 in men. In women multivariable bone and muscle models showed significantly better fracture discrimination (p < 0.01) than the combination of age and BMI. Compared to the BMD model, LRT showed a significantly improved VF discrimination of the combinations of BMD with texture or with muscle models (p < 0.001). In men the BMD model (AUC 0.64) did not significantly improve VF discrimination compared to age and BMI.
CONCLUSIONS
In older women, but not men, paraspinal muscle characteristics significantly enhance the discrimination of prevalent VFs beyond BMD alone. Muscle parameters were not predictive of incident VFs in prior analyses of the same cohort, suggesting that muscle deterioration may occur concurrently or follow vertebral fracture.
MeSH
Muscles, Paraspinal Muscles