To assess age-related changes in the component of hip fracture risk associated with femoral strength, the mechanical properties of old (mean age 73) and young (mean age 32) cadaveric femurs were compared using a loading configuration simulating a fall on the hip. The effects of loading rate were also assessed by testing one femur from each pair at a low displacement rate and the other at a high rate to better simulate an actual fall. Older femurs had half the fracture load and absorbed one-third as much energy as the younger femurs. A 50-fold increase in displacement rate resulted in a 20% increase in fracture load but no increase in energy absorption due to an increase in stiffness. There is a need for methods to identify persons who would benefit most from

strategies to prevent hip fracture. One piece of information that is needed is a non-invasive estimate of the fracture load of an individual hip. However, the best in vivo measurement has not been determined. In addition, it is not known whether three-dimensional measures are required or whether more easily obtained two-dimensional measures are adequate. Three techniques for obtaining densitometric and geometric measures were compared:​ normal resolution, anteroposterior DXA (dual energy x-ray absorptiometry);​ high resolution, multiple angle DXA;​ and three-dimensional reconstructions of QCT scans. The values obtained from each technique were compared with the femoral fracture loads. There were very strong correlations between measures made by the two DXA-based techniques. Bone mineral density and cross-sectional area of the femoral neck measured using the normal resolution DXA technique (2D) correlated strongly with fracture load. QCT results were comparable to those reported in earlier studies for the older femurs, but correlations were weak or nonexistent for the younger femurs. Our results revealed some limitations of current QCT techniques which have implications for clinical research.

Age-related changes in the material properties of bone were also explored. Experimental evidence has shown that a damage process, rather than plasticity, is responsible for the non-linear behavior of cortical bone. However, the exact nature or location of the damage has not been made clear. Old and young human cortical bone specimens were tested to determine the effects of age on post-yield behavior. Both old and young specimens showed a 35% reduction in modulus after being strained to 1% strain, but the old specimens had a lower yield stress and strain. Test specimens and matched controls were then examined microscopically and damage was quantified. Crack length was not significantly different between the groups. At 200X, old test specimens had the most cracks, 50% more than the old controls. Young test specimens did not have more cracks than the young controls (even though the modulus drop was the same for both old and young specimens), suggesting that damage might not coalesce as easily to form sharp cracks in young bone. Finally, a mathematical model was developed for damage in cortical bone which assumes a damage mechanism of debonding along the cement lines. Results of mechanical tests and optical analysis suggest that the model is a reasonable approach.