Leg Muscle Assessment by HR-pQCT and DXA: Associations Between Imaging Measures and Muscle Function in Adults with Obesity – A Pilot Study

Abstract Dual energy X-ray absorptiometry (DXA) and high-resolution peripheral quantitative computed tomography (HR-pQCT) are used for bone assessment but also image adjacent soft tissue. This pilot study examined association between HR-pQCT-measured leg muscle cross-sectional area (CSA) and density vs. DXA-derived lean mass, and explored associations with muscle strength, power, and physical performance in adults with obesity. Patients eligible for bariatric surgery (BMI≥30 kg/m2; ages ≥30) received baseline HR-pQCT (XtremeCT II) and DXA (Hologic) scans. HR-pQCT scans of the non-dominant leg were acquired at 30% of tibial length and processed using a custom Python script for muscle segmentation, followed by manual editing. Whole-body DXA scans were analyzed to extract lean mass for the total leg and subregions analogous to the HR-pQCT. Functional measures included gait speed (400-m walk), muscle power (12-step stair climb), and knee extensor strength (isokinetic dynamometer). Pearson correlations assessed relationships between imaging modalities and functional measures. Among 28 adults (ages 47.7±8.6; BMI 43.8±4.6 kg/m2), HR-pQCT muscle CSA and density averaged 36.0±4.8 cm2 and 60.1±8.1 HU, respectively. DXA lean mass was 2.23±0.40 kg for the total leg and 0.107±0.019 kg for the 30% site. HR-pQCT muscle CSA correlated moderately with DXA lean mass (r=0.57-0.69, p=0.0017 to <0.0001), whereas HR-pQCT muscle density showed negligible correlations with DXA lean mass (r=-0.01-0.04, p = 0.84-0.96). Significant positive correlations were observed for DXA total leg lean mass with absolute muscle power (r=0.40) and HR-pQCT muscle CSA with knee strength (r=0.41; both p<0.05). In adults with obesity, HR-pQCT and DXA may provide complementary measures of leg muscle quantity, with moderate association for CSA and lean mass, but limited associations with muscle density. Body weight and whole-body lean mass showed stronger correlations with stair climb power (r = 0.42–0.70) than localized leg muscle measures (r = 0.17–0.40). Muscle density and muscle quantity may capture different aspects of musculoskeletal health, warranting further investigation in larger samples.