Networks can be evaluated at multiple levels in bone. Functional (i.e., compensatory) relationships among bone traits give rise to a network of variable sets of trait interactions that play a critical role in determining normal mechanical function. However, little is known about the nature of these interactions in the context of genetic variants affecting skeletal size and quality, which are key determinants of bone functionality and fracture susceptibility. Biological co-adaptation of morphologic and compositional bone traits has been shown to explain fracture susceptibility in long bones. However, most fractures typically occur in metaphyseal regions such as the distal radius, proximal femur, and vertebral body. It is not known whether similar co-adaptive mechanisms exist in more complex corticocancellous sites. Corticocancellous structures also can be regarded as a connected network of cortical and trabecular elements capable of transferring mechanical loads. Although traditional measures of cortical and trabecular bone adequately characterize the phenotypes of each bone type, knowledge of the trait values themselves does not fully reveal how function is established for each genotype. We developed novel traits based on percolation theory, which is an analytical tool that is widely used to quantify the topology of complex networks such as neurons, to quantify the architecture of the vertebral body in an integrative way.

There is growing evidence indicating that sets of interdependent morphologic and compositional traits, in addition to BMD, provide more accurate predictions of fracture risk. Improved understanding of the biological processes that contribute to these functional interactions may provide a novel approach to identifying sets of traits that better predict fracture risk. The goal of this study was to test whether trabecular, cortical, and compositional bone traits are functionally related, as this would imply there is a strong biological process in bone that co-adapts traits in corticocancellous structures. We postulate that the cortical and trabecular bone traits will co-vary with matrix composition contributing to bone stiffness and strength.