of third metacarpal bones demonstrated significant variation in the maximum extent of deformation (peak strain magnitude) between different animals. Peak strains measured in young, exercise naive horses were 1.5–2 times higher than those recorded in older animals, which had been in training for a prolonged period [36]. The difference was accounted for by a significant variation in the geometric properties of the bones between the two groups [37].
Figure 2.16 Eccentric compressive loading on bones of the appendicular skeleton results in bending forces, with one cortex in tension and the opposite in compression. Stress magnitude increases linearly with distance from the axis about which the object bends (the neutral axis).
Most bones of the appendicular skeleton of quadrupeds are loaded in axial compression due to gravity. However, the principal load is frequently applied eccentrically, resulting in a bending moment, which is often exaggerated by the pull of musculature. Loads due to bending result in a stress gradient across the bone, with one side loaded in compression and the other tension (Figure 2.16). The further the mass of the structure is from its neutral axis (the axis about which it bends), the greater ‘leverage’ the material has to resist the loads and the stronger and stiffer the bone is in that plane. In a situation where loads are unpredictable and an object may be subject to bending forces in any plane, a hollow cylinder provides the most mechanically effective distribution of mass. If one plane is loaded more heavily and more frequently than another, then eccentric distribution of mass around the circumference of the cylinder will offer optimum resistance to the predominant loads while also providing support in other planes (Figure 2.17).
Load is transmitted between bones at joints. Typically, bones flare at their ends, providing a wider surface area at the articulation and reducing stress on the load‐bearing structures: hyaline articular cartilage, mineralized articular cartilage, subchondral bone and deeper metaphyseal bone. A framework of cancellous bone supports the entire joint surface and transmits the load to the diaphyseal cortex. The relatively fine trabeculae forming the cancellous network provide a compliant structure more capable of absorbing impact loads than dense cortical bone. Fat and other soft tissues in the spaces between trabeculae may also play an important role in damping these loads. In moderation, localized fracture of individual trabeculae may be another physiological mechanism of mitigating loads that could be more damaging to other tissues.
Figure 2.17 Line drawing illustrating the stress intensity (colour density) either side of the neutral axis at the mid‐diaphysis of a bone which is loaded in bending. Bone modelling can alter the geometric properties of the bone to place the tissue at the optimal location to resist bending stresses.
Adaptation
There are many examples in mammalian biology of specific architectural features of bones that are ideally suited to their mechanical environment. It is generally accepted that while the basic form of each bone is genetically pre‐programmed, the shape, mass and fine structural features of bones of the appendicular skeleton are ultimately determined by a proactive response of bone cells to their mechanical environment [38, 39]. Increase in cyclical bone strains due to raised levels of activity encourages bone formation. Conversely, disuse results in net resorption. In addition, a change of activity that results in altered loading on the bone (a change in principal strain direction) stimulates a modelling response that redistributes the mass of bone about its central axis to achieve a structure that is better suited to resist the new strains. For instance, exercise at a fast gallop in the horse is associated with a shift in the neutral axis of the third metacarpal bone [36] and, presumably, this is the drive for the modelling response that occurs in this bone while young horses first adapt to fast work. Stress itself is not directly measurable, so the consequence of loading, strain or effects associated with strain, such as fluid flow or change in electrical fields throughout the tissue, must be at the root of any mechano‐sensitive mechanism. The interconnected network of osteocytes and bone surface lining cells are well placed to sense deformation, fluid flux, etc., and there is good evidence that these cells are the drivers for adaptive modelling [40]. Numerous experiments have been undertaken in an attempt to determine the precise mechanical stimulus that activates a modelling response and the biological objective of that response [40]. Intuitively, maintenance of peak strain magnitude within certain limits during routine activity would be a sensible objective, and there is evidence to support this. The concept of a thermostat‐like mechanism ‘The Mechanostat’, switching either bone resorption or formation on or off in response to decreased or increased peak bone strains, was championed by Harold Frost [41]. In all probability, there are likely to be a number of strain‐related drivers that impact the balance of cellular activity [42].
While the physiological response to a change in the mechanical environment is rapid [43] and effected after relatively few cycles of loading [44], the architectural response may take weeks or months to complete. Consequently, in the event of a sudden increase in physical activity, high bone strains associated with that activity persist for some time. This indicates the importance of ‘training’ in developing bone structure that suits needs: much as the cardiovascular system can be prepared for optimal athletic performance, then so can the skeleton. Ideal training of racehorses involves exercise programmes that are designed to stimulate an appropriate remodelling response without causing excessive damage [45].
An adaptive mechanism can only respond to strains that the bone can detect. It cannot empower the bone to resist extraordinarily high strains (monotonic overload) associated with an accident. The risk of failure due to supraphysiological loads can be mitigated by the inclusion of a margin of safety to any response: greater bone mass means stronger bones, less likely to fracture. However, the formation of bone, and, more critically, carrying an unnecessarily heavy skeleton, is metabolically costly and in its own right potentially disadvantageous, particularly in animals that rely on flight for survival. The safety factor built into any adaptive response of bone is, presumably, under genetic control. Horses ‘coded’ with light bones will have a potential speed advantage but may be more prone to accumulate fatigue damage and consequently suffer fractures. This area of skeletal physiology will inevitably receive increasing attention in the future [46, 47].
There is good evidence that cancellous bone is also modelled in response to changes in loading. Experimentally, the volume fraction of subchondral bone in the palmar aspect of the third metacarpal bone increases in response to raised levels of treadmill work [48]. There is some evidence that an increased density of subchondral bone in animals subject to increased physical activity may occur as a consequence of a response to fracture of trabecular struts [49]. It is arguable whether this can be termed ‘adaptive’ or, in fact, represents a healing response to pathology, whatever the initiating cause. Increased density is associated with reduced compliance, and this may have negative consequences for tissues, such as hyaline and calcified articular cartilage, sandwiched between the subchondral bone and point of load [49–51].
Stress Protection
While new bone is deposited to strengthen bones in response to increased
