Stress-fractures of bone are similar to fatigue fractures in engineering materials as both are caused by microdamage accumulation under repetitive loading. However, stress-fractures have a biologic component, since bone tissue is capable of repairing damage and/or adding mass in response to damaging loading. The cellular damage repair processes cause transient porosities to develop in bone tissue, which temporarily reduce modulus and increase the damage accumulation rate if damaging loading is continued. Consequently, the stress fracture process is an interaction between bone’s mechanical response to highlevels of loading (damage accumulation), and the biological responses of repair and hypertrophy that change tissue modulus.

This work focuses on stress-reactions, bone tissue changes that precede a stress-fracture, in athletes. Specifically, racehorse proximal sesamoid bone (PSB) fracture is used as a naturally occurring model of an osteochondral stress fracture. The PSBs are a pair of bones in the metacarpophalangeal (fetlock) joint of the distal forelimb and PSB fracture is one of the most common fatal musculoskeletal injuries associated with race training. In this work, morphologic tissue properties between PSB from racehorses with (Case) and without (Control) a unilateral biaxial PSB are compared. Both Case and Control horses were in race-training at the time of death and their training histories were known. The observed tissue properties are related to exercise. In addition, a compartment model of bone’s turnover cycle is introduced and used to explain the associations among morphologic variables and exercise. The collected morphologic data is used to solve for the model’s steady-state rate constants. Finally, the relationship between PSB kinematics and how PSB motion may impact PSB fracture risk is explored.

Our primary finding was that horses in higher intensity training develop a subchondral stressreaction (lesion). A subchondral bone lesion was consistently found in the abaxial aspect of medial PSB from both Case fractured and Case intact contralateral PSB. This lesion was not found in Control PSBs. The lesion was characterized by low bone volume fraction, low tissue mineral density, and higher microdamage compared to surrounding tissue. The bilaterally of this lesion in Case horses coupled with the observed tissue properties make the lesion consistent with stress-reactions in subchondral tissue that develop before a complete stress-fracture. Within the subchondral lesion, bone volume fraction was negatively associated with exercise intensity and microcrack areal density was positively associated with exercise intensity. These findings imply the lesion was more severe in horses with more intense training. Generally, Controls had a less intense training program compared to Cases. Additionally, the bone volume fraction was higher in the internal trabecular bone of Case proximal sesamoid bones compared to Controls. At this internal site, microdamage was not observed and bone volume fraction increased with exercise intensity. Steady-state model rate constants were determined based on the observed morphologic tissue properties at these two locations. Modeling results suggest that the different relationships between exercise and tissue properties within the subchondral and internal sites will occur if model rate constants depend on exercise intensity in a location-specific manner. We hypothesize that the different relationships are due to a strain difference between the two regions that causes a damagerepair response in the subchondral region and an adaptive-response in the internal region. In vitro axial limb loading indicated that the PSBs may experience an articular incongruity at high-speed gallop loads. Further, kinematic analysis indicated that the medial PSB experienced external rotation at high-speed gallop loads, which may cause the stress-reaction to form on the abaxial surface of the medial PSB.

The presented work is clinically relevant because the identified abaxial subchondral site can be examined for a lesion in vivo to test whether a horse has or is developing a stress reaction that puts it at risk of fracture. Also, the lesion was found to develop 1- 8 months after an increase in exercise intensity, making this time-frame an important period for clinical examinations. Finally, the calibrated compartment model could be incorporated into a predictive dynamic simulation to predict the effects of exercise on PSB morphology and lesion development.