Fatty infiltration and cross-sectional area as indicators of muscle mass in osteoporosis: a meta-regression study.

Background: Paraspinal muscle morphology, including cross-sectional area (CSA) and fatty infiltration (FI), has been increasingly recognized as a potential imaging-based indicator of osteoporosis. However, the extent to which these muscle parameters differ across osteoporosis, osteopenia, and healthy populations remains unclear. Methods: A systematic meta-analysis was conducted based on 14 studies published from inception to January 25, 2025, comprising 125 effect size estimates related to CSA and FI across key paraspinal muscles. Pooled standardized mean differences (SMDs) were calculated using a random-effects model. Subgroup analyses were stratified by muscle group and diagnostic comparison. Three-level meta-regression models were implemented to examine the influence of study-level moderators, including age, sex, measurement level and comparison category. Results: A significant decrease in CSA was observed only in osteoporotic patients compared with controls, and multiple muscle groups were evaluated. In contrast, CSA differences in osteopenia were less consistent and appeared to vary by muscle type. FI demonstrated greater sensitivity across diagnostic comparisons, with significant increases observed in both osteopenia and osteo- porotic groups relative to controls, especially in the multifidus and erector spinae. Meta-regression identified age as a significant moderator, indicating that morphological differences diminish with increasing age. Both CSA and FI are associated with musculoskeletal deterioration in osteoporosis, with FI suggested to be relatively more sensitive and potentially capable of detecting early pathological changes during the osteopenia stage. However, when examined across specific measurement approaches for CSA and FI, the apparent advantage of FI was attenuated, and no clear difference in sensitivity was identified. The psoas major showed inconsistent findings across studies, likely due to its lower baseline fat content and lower responsiveness to aging. Conclusion: Both CSA and FI are associated with musculoskeletal deterioration in osteoporosis, with FI emerging as a more sensitive marker, potentially capable of detecting early pathological changes during the osteopenia stage. These findings highlight the value of paraspinal muscle assessments in osteoporosis research and clinical evaluation. Further studies are warranted to standardize measurement protocols and evaluate the integration of muscle morphology into imaging-based risk prediction models. Systematic Review Registration: https://www.crd.york.ac.uk/PROSPERO/view/CRD420251026322, identifier CRD420251026322.