RESUMO
OBJECTIVE: Individuals with existing vertebral fractures may not be aware that they are at high risk of subsequent fractures. We investigated if calcaneal quantitative ultrasonometry (QUS) and assessment of thoracic kyphosis could discriminate a group of older women with prevalent vertebral fracture from those without. METHODS: One hundred four women (mean age 71.3 +/- 5.8 yrs) underwent dual-energy x-ray absorptiometry (DEXA) bone mineral density (BMD; lumbar spine and hip), calcaneal QUS, and video rasterstereographic thoracic kyphosis measurements. They were dichotomized into a group with prevalent vertebral fracture (VF, n = 24) or without vertebral fracture (NVF, n = 80). RESULTS: Univariate variables associated with the VF group included broadband ultrasound attenuation (BUA; age-adjusted OR 1.96, 95% CI 1.12-3.42, p = 0.018); speed of sound (SOS; age-adjusted OR 2.01, 95% CI 1.09-3.70, p = 0.026); and thoracic kyphosis (age-adjusted OR 1.72, 95% CI 1.01-2.92, p = 0.049). A composite model (BUA and thoracic kyphosis) had higher area under the receiver-operating characteristic curve (AUC = 0.75) compared to lumbar spine DEXA BMD (AUC = 0.50, p = 0.0004) and total hip DEXA BMD (AUC = 0.60, p = 0.057). CONCLUSION: Reduced calcaneal QUS values and greater thoracic kyphosis were found to be significantly associated with the group of women with prevalent vertebral fractures. A composite risk score (BUA and thoracic kyphosis) had better discriminatory power than the individual risk factor of (low) DEXA BMD.