Bone Abstracts

The objective of this study was to evaluate the ability of various quantitative ultrasound variables to discriminate osteoporotic female corticosteroid users. Thirty-two women aged between 26 and 79 years who have been receiving corticosteroids equivalent to a minimum of 5 mg daily prednisolone for more than 3 months underwent bone mineral density (BMD) measurements at the lumbar spine, hip, and one-third radius (33% radius) using dual-energy X-ray absorptiometry (DXA). Acoustic parameters of bone were also obtained using a calcaneal quantitative ultrasound (QUS) (Quantitative Ultrasound Index (QUI), Broadband Ultrasound Attenuation (BUA), and Speed of Sound (SOS)) and a multi-site QUS device (radial and tibial SOS). Osteoporosis was defined according to the WHO criteria (a T-score≤−2.5) at the regions of interest (ROIs) as measured using DXA in postmenopausal women. Whereas in premenopausal women, Z-scores of ≤−2.0 were used for defining osteoporosis or BMD below the expected range for age. For predicting osteoporosis or BMD below the expected range for age at the lumbar spine, only radial speed of sound (SOS) showed a fair area under the receiver operating characteristic curve (AUC) (0.754, P=0.028), while AUCs for other QUS parameters represented poor or failing accuracy ranging from 0.534 to 0.638. As for discriminating subjects with femoral neck, total hip, and radius osteoporosis or BMD below the expected range for age from those without, all calcaneal QUS variables could be considered to be good with AUCs ranging from 0.848 to 0.875 (P values ranging from 0.017 to 0.026) for the first ROI and AUCs calcaneal QUS variables ranging from 0.747 to 0.769 (for QUI: 0.769; P=0.043, for BUA:0.750; P=0.060, and for SOS: 0.747; P=0.063) for the second ROI, and AUCs for calcaneal QUS variables (for QUI: 0.814; P=0.005, for BUA: 0.816; P=0.005, and for SOS: 0.789; P=0.010) for the third ROI showed fair to good accuracy. However, AUCs for multi-site QUS variables which ranged from 0.526 to 0.709 (P values ranging from 0.847 to 0.061) revealed failing, poor, or marginally fair accuracy. The preliminary results of this study with a very small sample size have implications that various calcaneal QUS variables, but not multi-site QUS variables, may be useful in discriminating corticosteroid-induced osteoporosis at the hip and the radius.