Predictive value of MRI fat quantification parameters combined with 25-hydroxyvitamin D in assessing fracture risk in osteoporosis patients
摘要
This study aimed to evaluate the synergistic predictive value of lumbar IDEAL-IQ quantitative marrow fat fraction (FF) and circulating 25-hydroxyvitamin D [25(OH)D] concentrations in assessing the likelihood of developing osteoporotic spinal compression injuries (OSCI), thereby establishing innovative biomarkers for enhanced fracture risk assessment and management strategies.
MethodsWe conducted a retrospective analysis of a prospectively maintained database. One hundred osteoporosis patients who underwent both lumbar quantitative computed tomography (QCT) and magnetic resonance imaging (MRI) examinations at The Second Hospital of Longyan City between January 2024 and January 2025 were enrolled. Participants were stratified into two cohorts: those with documented fractures (n = 50) vs. non-fractured controls (n = 50). Comprehensive data collection encompassed demographic characteristics, physical parameters, lipid profiles, calcium metabolism markers, phosphorus levels, and additional clinical variables. Lumbar skeletal mineral density (BMD) assessment, circulating 25(OH)D quantification, and L1–L5 vertebral adipose fraction (determined via IDEAL-IQ methodology) were performed. Statistical analysis incorporated multivariable logistic modeling to identify fracture-associated risk determinants, while receiver operating characteristic analysis evaluated the diagnostic capabilities of FF and 25(OH)D markers both independently and in combination.
ResultsPatients with fractures were significantly older and demonstrated diminished BMD and 25(OH)D concentrations, alongside increased L2 vertebral and aggregate L1–5 adipose fractions relative to controls (P < 0.01). Correlation analysis demonstrated inverse associations between L2 adipose content and BMD (r = − 0.539, P < 0.001) while revealing direct relationships linking 25(OH)D concentrations with BMD values(r = 0.765, P < 0.001). Multivariate logistic regression identified age (OR = 1.081, 95% CI 1.012–1.155) and L₂ FF (OR = 1.259, 95% CI 1.146–1.485)) as independent risk factors for fractures, while BMD (OR = 0.259, 95% CI 0.112–0.597) and 25(OH)D (OR = 0.672, 95% CI 0.532–0.849) were protective factors. ROC results showed that the AUC values for L₂ FF, 25(OH)D, and their combined prediction model were 0.69, 0.74, and 0.86, respectively, with the combined model demonstrating the best predictive performance (P < 0.001).
ConclusionsElevated L₂ FF measured by lumbar IDEAL-IQ and reduced 25(OH)D levels are both independently associated with high fracture risk. Their combined model significantly improves the predictive accuracy for osteoporotic vertebral fractures. Early risk stratification assessment using combined biomarkers may provide a novel approach for clinical fracture risk management and preventive intervention, though further validation is required.