Assessment of a new quantitative ultrasound calcaneus measurement: Precision and discrimination of hip fractures in elderly women compared with dual X-ray absorptiometry

The incidence of osteoporotic hip fracture increases in postmenopausal wome n with low hip bone mineral density (BMD). Dual X-ray absorptiometry (DXA) is the most commonly used technique fur the assessment of bone status and p rovides good measurement precision. However, DXA affords little information about bone architecture. Quantitative ultrasound (QUS systems have been de veloped to evaluate bone status for assessment of fracture risk. Our study was designed to assess a new QUS system from Hologic, the Sahara; to compar e it with a previous model, the Walker-Sonix UBA 575+; and to investigate w hether it is able to discriminate between women with and without fracture. Using both ultrasound devices, the measurements were performed at the heels of 33 postmenopausal women who had recently sustained hip fracture. A cont rol group of 35 age-matched postmenopausal women was recruited for comparis on. The total, neck and trochanter femoral BMD values were assessed using D XA for both groups. QUS and DXA measurements were significantly lower in fr actured patients (p<0.005) than in the control group. The short-term, mid-t erm and standardized shortterm precisions were used to evaluate the reprodu cibility of the two QUS systems. The Sahara showed a better standardized co efficient of variation for broadband ultrasound attenuation (BUA) than did the UBA 575+ (p<0.001). The correlation of BUA and speed of sound (SOS) bet ween the two QUS devices was highly significant, with an r value of 0.92 fo r BUA and 0.91 for SOS. However, the correlation between DXA and ultrasound parameters ranged from 0.28 to 0.44. We found that ultrasound measurements at the heel were significant discriminators of hip fractures with odds rat ios (OR) ranging from 2.7 to 3.2. Even after adjusting the logistic regress ions for total, neck or trochanter femoral BMD, QUS variables were still si gnificant independent discriminators of hip fracture. The areas under the R OC curves of each ultrasound parameter ranged from 0.75 to 0.78, and compar ed very well with femoral neck BMD (p>0.05). in conclusion, our study indic ated that the calcaneal QUS variables, as measured by the Sahara system can discriminate hip fracture patients equally as well as hip DXA.