Predictive ability of heel quantitative ultrasound for incident fractures: an individual-level meta-analysis

• Published in: Osteoporosis international 2015-07, Vol.26 (7), p.1979-1987
• Main Authors: McCloskey, E. V., Kanis, J. A., Odén, A., Harvey, N. C., Bauer, D., González-Macias, J., Hans, D., Kaptoge, S., Krieg, M. A., Kwok, T., Marin, F., Moayyeri, A., Orwoll, E., Gluёr, C., Johansson, H.
• Format: Article
• Language: English
• Subjects: Age Factors; Bones; Calcaneus - diagnostic imaging; Endocrinology; Follow-Up Studies; Fractures; Hip Fractures - etiology; Humans; Medicine; Medicine & Public Health; Meta-analysis; Original Article; Orthopedics; Osteoporosis; Osteoporosis - complications; Osteoporosis - diagnostic imaging; Osteoporotic Fractures - etiology; Predictions; Predictive Value of Tests; Rheumatology; Risk Assessment - methods; Ultrasonic imaging; Ultrasonography
• ISSN: 0937-941X; 1433-2965
• DOI: 10.1007/s00198-015-3072-7

Summary The relationship between bone quantitative ultrasound (QUS) and fracture risk was estimated in an individual level data meta-analysis of 9 prospective studies of 46,124 individuals and 3018 incident fractures. Low QUS is associated with an increase in fracture risk, including hip fracture. The association with osteoporotic fracture decreases with time. Introduction The aim of this meta-analysis was to investigate the association between parameters of QUS and risk of fracture. Methods In an individual-level analysis, we studied participants in nine prospective cohorts from Asia, Europe and North America. Heel broadband ultrasonic attenuation (BUA dB/MHz) and speed of sound (SOS m/s) were measured at baseline. Fractures during follow-up were collected by self-report and in some cohorts confirmed by radiography. An extension of Poisson regression was used to examine the gradient of risk (GR, hazard ratio per 1 SD decrease) between QUS and fracture risk adjusted for age and time since baseline in each cohort. Interactions between QUS and age and time since baseline were explored. Results Baseline measurements were available in 46,124 men and women, mean age 70 years (range 20–100). Three thousand and eighteen osteoporotic fractures (787 hip fractures) occurred during follow-up of 214,000 person-years. The summary GR for osteoporotic fracture was similar for both BUA (1.45, 95 % confidence intervals (CI) 1.40–1.51) and SOS (1.42, 95 % CI 1.36–1.47). For hip fracture, the respective GRs were 1.69 (95 % CI, 1.56–1.82) and 1.60 (95 % CI, 1.48–1.72). However, the GR was significantly higher for both fracture outcomes at lower baseline BUA and SOS ( p   0.20), but the predictive value of both BUA and SOS for osteoporotic fracture decreased with time ( p  = 0.018 and p  = 0.010, respectively). For example, the GR of BUA for osteoporotic fracture, adjusted for age, was 1.51 (95 % CI 1.42–1.61) at 1 year after baseline, but at 5 years, it was 1.36 (95 % CI 1.27–1.46). Conclusions Our results confirm that quantitative ultrasound is an independent predictor of fracture for men and women particularly at low QUS values.