Heel bone ultrasound predicts non-spine fracture in Japanese men and women

Abstract

A number of prospective studies in the USA and Europe have demonstrated that quantitative ultrasound (QUS) measurements predict fracture risk. To our knowledge, there has been no such study in a Japanese population, and very few studies have measured the prognostic value of QUS measurements among men, even in the USA and Europe. We performed a three-center prospective study to investigate the relationship between baseline heel QUS measurements and non-spine fracture risk. There were 4,028 subjects (1,004 men and 3,024 women), 67.5+/-8.9 years [mean +/- standard deviation (SD)] of age), who underwent heel QUS (Achilles device) at three centers between 1993 and 2000. In 2002, the subjects were mailed a standardized questionnaire that asked about their history of fracture. The mean follow-up period was approximately 5 years. The Achilles measured speed of sound (SOS) and broadband ultrasound attenuation (BUA). We used Cox regression analysis to determine the hazard ratio (HR), using weighted coefficients. SOS, BUA, and stiffness index (SI) predicted self-reported hip, wrist, and total non-spine fractures. After we had adjusted for age, gender, and weight, the HRs of total non-spine fracture were 1.54 [95% confidence interval (CI) 1.39-1.69], 1.53 (1.37-1.70), and 1.80 (1.62-1.98) for 1 SD decrease in SOS, BUA, and SI, respectively. In men, SOS and SI also predicted total non-spine fractures with HRs similar to those in women. The HR of prediction for hip fracture by SOS and SI was better in the short term than in the long term, and the prediction for hip, wrist, and non-spine fracture remained significant between 5 to 10 years of follow-up. Measurements obtained from heel QUS predicted non-spine fracture in Japanese men and women, and the HRs of Japanese of both genders was similar to the risk ratio (RR) of Caucasian men and women. QUS parameters can predict hip, wrist, and non-spine fracture up to 10 years.