Pore network microarchitecture influences human cortical bone elasticity during growth and aging

Cortical porosity is a major determinant of bone strength. Haversian and Volkmann׳s canals are׳seen' as pores in 2D cross-section but fashion a dynamic network of interconnected channels in 3D, a quantifiable footprint of intracortical remodeling. Given the changes in bone remodeling across life, we hypothesized that the 3D microarchitecture of the cortical pore network influences its stiffness during growth and ageing. Cubes of cortical bone of 2 mm side-length were harvested in the distal 1/3 of the fibula in 13 growing children (mean age±SD: 13±4 yrs) and 16 adults (age: 75±13 yrs). The cubes were imaged using desktop micro-CT (8.14µm isotropic voxel size). Pores were segmented as a solid to assess pore volume fraction, number, diameter, separation, connectivity and structure model index. Elastic coefficients were derived from measurements of ultrasonic bulk compression and shear wave velocities and apparent mass density. The pore volume fraction did not significantly differ between children and adults but originates from different microarchitectural patterns. Compared to children, adults had 42% (p=0.033) higher pore number that were more connected (Connective Density: +205%, p=0.001) with a 18% (p=0.007) lower pore separation. After accounting for the contribution of pore volume fraction, axial elasticity in traction-compression mode was significantly correlated with better connectivity in growing children and with pore separation among adults. The changes in intracortical remodeling across life alter the distribution, size and connectedness of the channels from which cortical void fraction originates. These alterations in pore network microarchitecture participate in changes in compressive and shear mechanical behavior, partly in a porosity-independent manner. The assessment of pore volume fraction (i.e., porosity) provides only a limited understanding of the role of cortical void volume fraction in its mechanical properties.