Background: A quarter of fractures needing admission happen in men, but few data are available that show the value of bone measures for prediction of fracture risk in men. We aimed to assess quantitative ultrasound of the calcaneum and fracture incidence in a prospective observational population study.
Methods: Calcaneum broadband ultrasound attenuation (BUA) was measured in men and women in the Norfolk cohort of the European Prospective Investigation into Cancer (EPIC-Norfolk) between 1997 and 2000. Incident fractures were ascertained by hospital record linkage.
Findings: In 14824 men and women aged 42-82 years, during mean follow-up of 1.9 years (SD 0.7), there were 121 incident fractures that needed admission, including 31 hip fractures. Men and women in the lowest 10% of the calcaneum BUA distribution had a relative risk of fracture of 4.44 (95% CI 2.24-8.89, p<0.0001) compared with those in the upper 30% of the distribution. A fall of about 1 SD in BUA (20 db/MHz) was associated with a relative risk of fracture of 1.95 (95% CI 1.50-2.52, p<0.0001), independent of age, sex, weight, height, cigarette smoking habit, and past history of fracture. BUA predicted fractures with consistent magnitude in subgroups stratified by sex, age 65 years or older and younger than 65 years, smoking habit, past history of fracture, and hip and non-hip fractures separately. The sex difference in fracture risk was largely accounted for by differences in BUA.
Interpretation: Quantitative calcaneum ultrasound predicts total and hip fracture risk in men and women in a continuous relation. The challenge now is to identify interventions to improve bone health in the whole population.