Abstract

The current benchmark for the assessment of fracture risk is the status of osteoporosis based on the measurement of bone mineral density (BMD) by dual-energy X-ray absorptiometry (DXA). However, DXA-based BMD has been shown to lack predictive ability for individual fracture risk. More than half of the hip fractures occur among people who are not classified as having osteoporosis. Osteoporosis (i.e. reduced bone mass) is only one risk factor for a fracture. In addition to bone mass, the mechanical strength of a bone is influenced by material and structural factors. However, we have limited information about the combined effects of BMD and bone structural properties in the evaluation of fracture risk, with regard to different types of hip fractures in particular. Therefore, this study investigated the radiograph-based structural factors of the upper femur for the assessment of bone mechanical competence and cervical and trochanteric hip fracture risk.

The subjects of the clinical study comprised 74 postmenopausal women with non-pathologic cervical or trochanteric hip fracture and 40 age-matched controls. The impact of bone structure on the bone mechanical competence was studied using the experimental material of 140 cadaver femurs. The femora were mechanically tested in order to determine the failure load in a side impact configuration, simulating a sideways fall. In all study series, standard BMD measurements were performed, and the structural parameters of bone were determined from digitized plain radiographs.

The present study showed that the large variation in the mechanical competence of bone is associated with the geometrical and architectural variation of bone. Moreover, the results strongly suggested that the etiopathology of different types of hip fractures significantly differs, and that fracture risk prediction should thus be performed separately for the cervical and trochanteric hip fractures. Furthermore, the study implied that the current clinical procedure can better be used for the assessment of the risk of trochanteric fracture, whereas cervical fracture is more strongly affected by the geometrical factors than by BMD. Finally, radiograph-based structural parameters of trabecular bone and bone geometry predicted in vitro failure loads of the proximal femur with a similar accuracy as DXA, when appropriate image analysis technology was used. Thus, the technology may be suitable for further development and application in clinical fracture risk assessment.