About
Quantitative ultrasound (QUS) of the heel is a convenient and affordable method for assessment of osteoporosis and captures heel bone properties that predict fracture risk independently of major risk factors, including bone mineral density (BMD). Our recent genome-wide association study robustly identified genetic loci associated with heel QUS measures, of which some commonality was demonstrated to those previously found for hip or spine BMD. Our proposal aims to assess the cross-sectional correlates of heel QUS measures, the magnitude of independent association with fracture risk, and gene-environment interactions when data become available, to assess implications for osteoporosis prevention. Since measurement of heel QUS is a convenient and affordable method for assessment of osteoporosis, powerful studies of the physical, environmental, and lifestyle correlates of heel QUS measures and their potential interactions with genetic determinants, should have potential to inform risk assessment and interventions to mitigate the growing burden of osteoporosis and fractures in the population.
The extensive data available in UKBiobank on heel QUS measures and physical, environmental, and lifestyle variables provide a unique opportunity to powerfully assess the above associations.
This research is in the public health interest as it should reveal new pathways to modify osteoporosis risk. We will assess the cross-sectional correlates of heel QUS measures (BUA, VOS, estimated BMD, and stiffness index) and associations with fracture risk in the full cohort using regression modelling techniques that we have developed in large-scale collaborations. We will assess interactions between the correlates of QUS and known SNPs associated with QUS measures to test aetiological hypothesis and assess implications for fracture risk stratification.
This application requires access to data only (i.e. no samples), specifically data from the baseline visit (plus genetic data when available) and follow-up data on incident fractures from primary/secondary care records. We will require data from the full cohort of 500,000 participants to facilitate the most powerful analyses.