We propose a novel approach for identifying skeletal trait patterns that predict fracture risk in older men and women, independent of areal bone mineral density (aBMD). The breakthrough was finding that aBMD loss results from different relative proportions of decline in bone mineral content (BMC) and increase in bone area based on baseline external bone size. aBMD is often used as a surrogate of strength, and loss in aBMD is often assumed to reflect loss in strength. However, this association has been predicated on an implicit assumption that the proportion of BMC-declines to area-increases is similar among individuals which our data contradict. In the proposed work, we will test the global hypothesis that external bone size is associated with variation in strength-decline trajectory and fracture risk, independent of aBMD. This hypothesis is supported by preliminary cadaveric studies showing different strength-age regressions for narrow and wide subgroups. Moreover, we show that low aBMD may explain fracture risk within subgroups stratified by external bone size but did not explain the two-fold increase in fracture risk between subgroups. Thus, we have identified limitations in the uniform application of aBMD to predict fracture risk, suggesting that recognizing population heterogeneity is necessary for advancing fracture risk prediction. To address these limitations, we will identify the structural traits that explain fracture risk within subgroups and test the hypothesis that the structural traits predicting fracture risk differ among the subgroups. We first leverage high resolution imaging and direct mechanical tests available for cadaveric femurs to test the hypothesis that the structural trait patterns determining low femur strength vary with external femur size (Aim 1). To systematically identify structural patterns associated with variation in fracture risk and strength-decline trajectories, we will use statistical shape and trait modeling within a novel computational framework to identify the specific structural patterns that best predict experimental strength and then test how these associations differ with external size, sex, and race. We address how these structure-function relationships change over time by leveraging existing longitudinal hip DXA data for elderly White men enrolled in the Osteoporotic Fractures in Men Study (MrOS) and elderly White women enrolled in Health ABC (Aim 2) and elderly Black men and women enrolled in Health ABC (Aim 3). We will test the hypothesis that strength-decline trajectories differ between baseline FN area tertiles for older men and women, and that the strength-trajectory is associated with risk of fracture within tertiles. Successful completion of these aims will provide evidence of multiple biomechanical pathways leading to reduced proximal femur strength in older men and women and will identify sets of traits and trait interactions that best predict fracture risk for subgroups of individuals. This...