Although osteoporotic fracture is a common and expensive healthcare problem, the factors responsible for patient’s susceptibility to fracture remain incompletely understood. Bone mineral density is strongly associated with bone strength, but its use to assess fracture risk in all patients has been limited. The potential addition of geometric and material properties, at multiple hierarchical levels, may improve fracture risk predictions significantly.

During the past decades, the mouse has evolved to the most important model organism in the field of modern biomedical research. This is due to two principal reasons:​ on the one hand, the successful application of various transgenic technologies, and on the other hand, the strong homology between mice and men. With the efficient techniques given to geneticists, they are nowadays able to identify locations on chromosomes, which play a major role in the process of modeling and remodeling of bone. Nevertheless, the geneticists need quantitative endpoints to accomplish that work. Micro-computed tomography (μCT) has proven to be well suited for assessing and quantifying a variety of three-dimensional microstructural bone phenotypes, also referred to as bone morphometry.

The gold standard to determine bone strength is a functional, mechanical test. Such a test gives several measures of bone competence, like stiffness, ultimate force, ultimate displacement and work to failure. Due to their small size these tests are challenging in murine bone;​ compared to data from larger animals the precision is often lower. Accurate mechanical and material data on bone strength are still lacking for the mouse.

For this reason, this thesis focused on improved prediction of bone mechanical properties. The main aims of this study were to experimentally assess bone mechanical properties, such as absolute strength, bone stiffness and post-yield behavior, and to relate these mechanical properties to bone overall bone geometry as well as to bone micro- and ultrastructure. To fulfill these objectives, an integrative approach for hierarchical investigation of bone was developed, working at different scales of resolution, ranging from the whole bone to its ultrastructure.

At the organ level, the influence of boundary conditions on the testing of whole murine femora was evaluated and the optimization of sample preparation and alignment increased significantly the reproducibility of the mechanical tests. Axial compression and three-point bending tests preformed on two murine inbred strains, C57BL/6 (B6) and C3H/He (C3H), revealed that C3H mechanical bone phenotype is less variable than in B6. Comparing the mechanical parameters with bone morphometry, it was shown that whole bone morphometrical indices, such as bone volume, cortical thickness and cross-sectional area predicted almost 80% of bone stiffness, strength and yield force. For the femoral neck, cortical thickness explained 83% of bone strength when femoral head was axially loaded. Our investigation showed that bone post-yield behavior could not be explained by morphometry.

Accurate mechanical parameters, compared to microstructural finite element (µFE), showed that the classical Euler-Bernoulli beam theory underestimated significantly the bone Young’s modulus, when estimated from three-point bending tests. Stiffness was measured mechanically and Young’s modulus was calculated based on the μFE analyses. The low aspect ratio of murine femora excludes the use of classical beam theory, without applying correcting factors. It was also shown that reported murine inbred strain-specific differences in tissue modulus as derived from three-point bending tests are largely an effect of geometric differences, not accounted for by beam theory.

Direct mechanical testing provides detailed information on overall bone mechanical and material properties, but fails in revealing local properties such as local deformations and local contribution of bone micro- and ultrastructure on bone failure. Therefore, we incorporated several imaging methods in our mechanical setups in order to get a better insight into bone deformation and failure characteristics. Whole bone testing was recorded with highresolution and high-speed cameras. The movies showed fracture initiation and bone whitening at high stress regions. This also helped to remove erroneous samples and with that to increase, though, the power and the relevance of the experiments. At a microstructural scale, imageguided failure assessment (IGFA) uncovered, in a three-dimensional manner, fracture initiation and propagation as well as local deformation of cancellous bone. It also showed that femoral neck of B6 and C3H fractured differently when the femoral head was loaded axially.

The poor prediction of bone post-yield behavior obtained with classical bone morphometry, together with the whitening processes observed before bone fracture, motivated the investigation of bone microcrack initiation and propagation at an ultrastructural scale. Synchrotron radiation based micro-computed tomography (SRμCT) permits visualization of bone porosity features with a resolution of 1 µm and below. We successfully imaged osteocyte lacunae, canals, uncracked ligament bridges and mineralized collagen fibers directly in 3D. Furthermore, the contribution of these ultrastructural entities on the microcrack initiation and propagation was showed. Canal voids were larger in C3H than in B6. In C3H, they were therefore able to initiate and determine the orientation of microcracks. Osteocytic lacunae did not inititate microcracks, but their distribution provided guidance for the propagation of microdamage. Microdamage accumulated more and faster in C3H than in B6. This difference at a submicron scale can explain the higher brittleness and the lower toughness of C3H observed at the macroscopic level.

In conclusion, a new hierarchical approach for functional investigation of bone was developed. It allowed relating bone structural properties at different scales, ranging from whole bone to bone ultrastructure, to the overall bone competence. In the future, this might help the development of better tools and measures for clinically predicting bone strength.