Osteoporotic hip fracture is a significant medical problem among the elderly, accounting for the majority of osteoporosis-related costs and associated morbidity. An improved understanding of the factors contributing to bone strength would provide enhanced diagnostic tools for measures of hip fracture risk as well as evaluation of the efficacy of treatment and preventative therapies. This dissertation examined the role of trabecular bone distribution and architecture—the distribution of trabecular bone within the femoral neck and the microstructural distribution (anisotropy) of trabecular bone throughout the proximal femur—in determining the strength of the proximal femur. Theoretical analyses, experimental testing, medical imaging and computational modeling were used to improve understanding of the role of geometric and architectural bone parameters in determining the strength of the proximal femur.

It was demonstrated here that the asymmetric distribution of trabecular bone in the femoral neck may help maintain the structural integrity of the femoral neck during normal daily activities in the face of age-related decreases in trabecular bone density and decreasing cortical thickness, but this form of functional adaptation increases the risk of fracture during a fall to the side of the hip. By including bone-specific trabecular anisotropy and a multiaxial failure criterion developed specifically for trabecular bone in continuum-level finite element models of the proximal femur, these analyses showed that the multiaxial failure criterion contributes significantly to prediction of whole bone failure loads under both stance and fall loading conditions. These results also showed that trabecular bone elastic anisotropy, independent of bone density, decreases the stiffness and strength of the proximal femur. However, the small differences observed in the degree of anisotropy between hip fracture and control patients are unlikely to have biomechanically important effects and, for these linear analyses, orthotropy did not improve predictions of experimental failure load. This implies that trabecular anisotropy can be neglected in future models of the proximal femur used for predicting fracture load, and that simpler isotropic models can be implemented with sufficient fidelity clinically. However, the multixial failure criterion remains a valuable tool for continued development of robust finite element models necessary for future exploration of the mechanistic behavior of the proximal femur under complex loading conditions and for improving fracture risk predictions.