Rather than resisting microscopic damage, human bone tissue is adapted to disperse energy through temporary plastic deformation, which it later repairs. Under low strains, bone cells continually turn over aging bone as a stochastic maintenance operation. Under high strains, stochastic remodeling is repressed, but microscopically damaged regions are resorbed and replaced through targeted remodeling. The triggering of bone remodeling by two opposing strain environments has long confounded attempts to link bone microstructure consistently to mechanical stimuli. Additionally, bone cells become uncoupled from mechanical control as they age, and begin eroding the cortex through more extensive and unrepaired bone resorption. Intracortical porosity is often treated as a consequence of aging, when it is accumulated enough to impact bone strength and fracture risk. Yet because cortical pores are produced by both stochastic and targeted remodeling activity, they are constantly forming in all strain environments and over the lifespan. This raises the question of whether pores in high-strain environments are morphologically “optimized” or resistant to the high risk of microcrack initiation and propagation in those regions. Moreover, does porosity increase fracture risk with age not only because it is increasing porosity, but because it is reshaping it morphologically?

This study fundamentally asked whether the three-dimensional geometry of pore networks is morphologically optimized to resist local mechanical strain. The hypothesis of structure-strain pore morphotypes in the human right-side femoral neck, a common site of osteoporotic fracture, and the human right-side midshaft fourth rib, a relatively unloaded control, for one male and one female per age decade from 20s to 80s. Extracted regions of each bone are visualized with high-resolution micro-CT imaging to reconstruct complete three-dimensional pore networks from 10 mm thick cross-sections of bone. These images are processed with custom routines that automatically extract and characterize porosity by bone type, pore type, and cross-sectional region.

Intraskeletally, the femoral neck and rib do vary significantly in pore morphometry, but not as expected. The more highly strained femoral neck is significantly more porous than the rib throughout the lifespan, apparently due to more permissive and uniform resorption at its endosteum, which the rib suppresses in its pleural cortex. However, regional comparions along the increasing strain gradient in the femoral neck can significantly confirm that high strain regions are distinct in pore morphometry. High- strain pores are significantly less densely populated, produce lower percentages of open and total porosity, are less convergent with other systems, and are more longitudinally oriented. Such isolated systems would be less vulnerable to the initiation and propagation of microdamage than the dense, broad, and widely convergent pore network permitted in lower strain regions. The pleural cortex of the rib, which has been hypothesized to experience relatively higher strain in some anatomical models, embodies a majority of these high-strain porosity markers. However, reduced mechanical control of the rib increases its sensitivity to physiological co-variates such as age and sex. While the pace of age-related effects is highly individualized, aging consistently and significantly reduces pore separation through convergence. This would increase vulnerability to microcrack initiation and propagation in high-strain regions.