Use of quantitative ultrasound for identifying low bone density in older people

Objective: This study examined criterion, convergent, and discriminant validities of quantitative ultrasound (QUS) for identifying low bone density among people aged 55 years and older in Taiwan. Methods: We recruited 453 community-dwelling volunteers and 30 patients with lower extremity fractures. Bone density was assessed using both calcaneal QUS and femoral neck dual-energy x-ray absorptiometry (DXA). Two QUS parameters, speed of sound (SOS) and broadband ultrasound attenuation (BUA), were also used to estimate heel bone mineral density (HBMD). Results: Using DXA as the criterion for identifying low bone density (DXA T score of 1.0 or lower), likelihood ratios for BUA and SOS at the 50th percentile and HBMD for men were 1.50, 1.75, and 1.28, respectively; the counterparts for women were 1.54, 2.13, and 1.29. As for identifying osteoporosis (DXA T score of -2.5 or lower), higher likelihood ratios of the 3 QUS parameters were gained. For convergent validity, Pearson correlation coefficients for DXA with BUA, SOS, and HBMD ranged from 0.40 to 0.43 for men and from 0.48 to 0.53 for women. For the ability to discriminate men and women with lower extremity fractures from those without, no significant differences in the area under the receiver operating characteristic curve were detected between BUA, SOS, and HBMD and DXA after adjusting for age, body mass index, fall history, and current smoking. Conclusions: Although having very good convergent and discriminant validities and fair criterion validity, calcaneal QUS may be a screening tool for identifying low bone density.