Abstract:
As research expands into exploring the relationship between body composition and site-specific fractures, the relationship between increasing body mass index (BMI) and fracture risk is not as clear. BMI is contributed to by fat and lean mass and how either are related to fracture risk is not fully understood. The aim of this thesis is to identify the relationship between body composition, more specifically fat and lean mass, and site-specific fracture risk. The analysis also included measures of bone mineral density (BMD) to determine how these relationships may be contributed to by BMD but also, to identify the relationship between different measures of BMD and site-specific fracture risk. In tertiary analysis, the relationship between total falls and site-specific fracture risk was also explored. Data collected from the Auckland Calcium Study was used for analysis. A total of 1471 predominantly New Zealand European, postmenopausal women were included. Of the participants, 238 incurred at least one fracture with 346 total fractures recorded. Fracture types included were any, osteoporotic, lower arm, humerus, lower leg, pelvis, hip, vertebral bodies, rib and ankle. The primary endpoint was time to first fracture using Cox proportional hazards model and a secondary endpoint included number of fractures per person using Poisson regression. The study aimed to determine the influence of variables of body composition collected at baseline on each of the endpoints. All independent variables were explored alone. Analyses were repeated independently for each fracture type. In the time to first fracture analysis, increasing height in 5 cm increments was associated with an increased risk of any (HR, 95% CI: 1.17, 1.05 – 1.31), osteoporotic (HR, 95% CI: 1.19, 1.05 – 1.34) and vertebral fractures (HR, 95% CI: 1.33, 1.07 – 1.64). No relationships were observed between BMI, weight, and fat mass and fracture risk. Lean mass was associated with risk reduction for rib fractures only. Increasing BMD in 0.1 g/cm2 increments measured at total body, lumbar spine, femoral neck and total hip was associated with a reduction in risk of any, osteoporotic, lower leg and vertebral fractures. Increasing BMD at the four measured sites was inconsistently associated with fracture risk for lower arm, rib, hip and pelvis fractures. No association between any measure of BMD and ankle and humerus fractures was observed. In the time to first fracture analysis, increasing falls was associated with an 11-43% per fall increased risk of fracture at all fracture sites except for pelvic fractures. In conclusion, guidelines should continue to include bone mineral density and falls prevention to support the management of osteoporosis.