The Milne Lecture at the American Association of Equine Practitioners (AAEP) Convention is also known as the State of the Art Lecture because each Milne Lecture, regardless of topic, is selected for its groundbreaking qualities and potential to change the paradigms by which veterinarians and researchers understand that topic in the horse. This year’s Milne Lecture (which is named for Frank J. Milne, an AAEP past president and distinguished life member) was an outstanding example of those ideals, combining biology research and biomechanics principles into an insightful, landmark report of more than 20 years of bone and fracture research by David M. Nunamaker, VMD, Dipl. ACVS, Jacques Jenny Orthopedic Research Chair at the University of Pennsylvania’s New Bolton Center.

Nunamaker began his two-part presentation with a discourse on his work on bucked shins in Thoroughbred racehorses, noting that an incidence of 70% has been reported in young Thoroughbreds in race training. The annual cost of this problem in the United States has been estimated by some at $10 million per year in lost training and racing days.

“This is a fatigue injury of bone and usually occurs in 2-year-old horses during the first six months of their training and may be seen bilaterally (in both left and right limbs),” he explained. “If the condition does occur bilaterally, the left limb is usually involved before the right,” due to the counterclockwise direction of racing in the United States and thus the greater forces on the left lead limb in the turns.

“Clinically, the condition is diagnosed by physical examination using palpation of MC3 (the cannon bone) to reveal heat, pain, and tenderness, with or without swelling over the dorsal (front) or dorsomedial (front inside) surface of MC3,” he continued. “Affected animals tend to be short-strided, uncomfortable at exercise, or lame. Radiographic diagnosis may be delayed from the clinical onset of signs, but is evidenced by periosteal new bone formation over the dorsal or dorsomedial aspect of this bone (see image here).

“With few exceptions, once this condition occurs and resolves, the animals do not experience this problem again,” he explained. “This observation has been used in the past to intentionally buck horses’ shins to get past the problem (at a more convenient time than during a racing campaign). One downside to this method, besides the lost training and racing days, is the risk that the horse will go on to develop a stress or saucer facture of MC3 that could lead to catastrophic fracture. These fractures seem to only occur in horses that have previously bucked their shins.”

Nunamaker added that while this problem is normally seen in 2-year-olds entering race training, it might also be seen in older horses when they enter race training for the first time. He noted that trained European racehorses which were trained on turf tracks might buck their shins when they race on North America’s harder dirt tracks, and that approximately 12% of horses which buck their shins will develop a radiographically visible stress or saucer fracture on the dorsolateral (front outside) surface of MC3 up to one year after the original bucked shin injury.

“Catastrophic complete midshaft fractures of MC3 can occur when these horses are exercised at speed or raced,” he warned. These fractures comprise about 10% of the fatal catastrophic musculoskeletal injuries seen on North American racetracks annually.

Standardbreds also have bucked shin problems, he said, but it’s far less common--about 10% of Standardbreds. The slower racing speed of the Standardbred (48 km/h) and different gait (trot or pace) compared to the Thoroughbred (64 km/h gallop/run), along with the different weight distribution of the harness vs. jockey, likely accounts for the difference.

Pathogenesis of Bucked Shins

Traditionally, Nunamaker said, the etiology of bucked shins was thought to be one of subperiosteal hematoma (bruising beneath the periosteum, the membrane that covers bone) with microfractures as a result of trauma. The microfractures were thought to elicit a healing response via callus formation, in which primary bone is initially laid down to capture two bone ends that are rubbing together; this mass gradually calcifies into a fully healed bone repair (secondary fracture healing).

Since there are no bone ends rubbing together in bucked shins, however, Nunamaker said that this description of the injury and healing process was not correct. Instead, he said that primary healing or direct bone remodeling, without callus formation, would be expected for this injury. His hypothesis was that bucked shins was a repetitive motion injury associated with high-strain cyclic fatigue, rather than a microfracture/bruising injury.

Understanding Bone Strength

To quantify the strength of racehorses' cannon bones, Nunamaker and colleagues evaluated cannon bone samples from Thoroughbred and Standardbred horses of various ages. Their experiment yielded data for Imin (measurement of dorsopalmar, or front-to-back, bending) and Imax (lateromedial, or side-to-side, bending).

“There are two types of deformation--plastic and elastic deformation,” he explained. “Plastic deformation is like when you bend a paper clip far enough to change its shape. Elastic deformation is like when you bend it just a little, and then it returns to its original shape. This is more like what bone does as the horse runs. Steel has an endurance limit, which is about 30% of the absolute limit (of bending without fracture), below which it can bend back and forth forever and it won’t break (bone does not seem to have an endurance limit). Stress as load per unit area we can only calculate, but strain deformation we can measure in vivo (in the live horse) and in vitro (on samples in the lab).

“You can have all these numbers, but keep in mind that they are not specific to this patient,” he cautioned. “Each horse is different.”

Nunamaker found that in Thoroughbreds, all section properties were much lower (weaker) for yearlings than for any other age group. “The most significant changes in the bone occurred at the midsection between one and two years of age, but continued change took place until age four,” he related. “No observable changes took place after four years of age. The Thoroughbred changed this property (Imin) to a greater extent than did the Standardbred during the first two to four years of life, just when the animal is at risk for bucked shins.”

In Vivo Strain Testing

Armed with a better understanding of in vitro bone characteristics, the researchers set out to measure these strains in live animals. Four 2-year-old racehorses were purchased and trained for six months for this portion of the study, and a 12-year-old racing veteran was also used. The horses had rosette strain gauges mounted on their cannon bones under general anesthesia, then the area was locally anesthetized and the horses were exercised on a dry dirt track. Measurements and sounds (to correlate jockey verbal commands with performance changes) were recorded in equipment carried by the jockey in a backpack weighing about 30 pounds (see image here).

They found that when horses work slowly, the front of the cannon bone is under tension--in other words, the front of the bone is stressed by stretching force rather than compressing force--and compressive forces occur at an angle to the bone surface. At higher speeds, the forces become more compressive. Also, the older horse had lower measures of strain than the younger horses at the same speeds, though the strains were in the same directions from the bone. Nunamaker noted that the peak strains in horses are higher than previously reported in any animal species.

Nunamaker concluded that younger horses develop higher strains on MC3 while running than older horses, and that this difference is due to not yet having remodeled the bone to meet the demands of racing. The older horse’s bones had remodeled to handle the stress of racing by thickening at the dorsal/dorsomedial aspect and thus becoming more resistant to bending (more on that later).

Treadmill exercise was also evaluated by running the same horses on the treadmill in the afternoon at the same speed as they had worked in the morning. The horses had decreased bone loads, and Nunamaker observed that the differences could be due to no rider and no wind resistance on the treadmill. He reported that the strain values changed to 30-50% of those measured on the racetrack, and concluded that for this reason, “Treadmill training is not indicated to train young horses. It might be useful for cardiovascular maintenance in a horse that is sore but already developed the correct MC3 shape.”

Shape and Strength of the Cannon Bone

We might think of bones as being shaped at birth and only changing in size as we (and horses) grow. That’s not necessarily the case, although shape changes aren’t extreme.

Eight untrained 2-year-old Thoroughbreds were purchased for this part of the study. Classical training, which involves mostly galloping work with speed work done about once every seven to 10 days, was used for groups 1 and 2. The only difference was training on a dirt track for group 1, whereas group 2 trained on a wood chip track. Group 3 consisted of horses trained only to halter and pastured for the study. Group 4 horses were trained with a modified program that included more frequent speed work on a dirt track.

The untrained group 3 horses exhibited a symmetrical cannon bone in cross section, with a large marrow cavity that has a “lacy” appearance at its edges. In comparison, group 1 and 2 horses showed a reduction in the “lacy” bone marrow cavity edge and a relatively smaller marrow cavity due to significant bone addition on the medial aspect (and to a lesser degree on the lateral aspect). The group 1 horses showed more complete remodeling than group 2, suggesting that the harder dirt surface stimulated remodeling more quickly. Groups 3 and 4 horses showed remodeling in the dorsal and dorsolateral areas that was not seen in the other groups.

“Classical training methods applied to horses training on a hard or soft track did not effectively change the inertial properties (Imin) that influence bending of MC3 in a dorsopalmar direction,” Nunamaker reported. “Exercise regimens (group 4) that stressed MC3 in compression on its dorsal surface did change Imin in a significant manner, consistent with adult racehorses evaluated previously that were no longer at risk for bucked shins.

“The results of this study supported the concept that exercise could be designed to optimize the shape of MC3. This, in turn, should influence (decrease) the incidence of bucked shins in this Thoroughbred racehorse model, and therefore, the problem within the industry.”

Testing Training Regimens

“The new bone formation (of bucked shins) is not disease per se, but an appropriate response of bone to high-strain repetitive motion injury of MC3,” Nunamaker stated. “Therefore, it is this injury that needs to be addressed. Because exercise is the problem, a change in the pattern of exercise may also be the solution.”

A larger study was required to test the efficacy of different training regimens, this one using five commercial training stables. Two training stables (numbers two and five) used the modified training program developed in the previously discussed study. The other three used classical training methods. In summary, 11 years of training data from 226 2-year-old Thoroughbreds was compiled, including daily records of accurate distances of jogging, galloping, and breezing. Physical exams were also recorded, and data collection stopped when horses reached the end of the one-year observation period, bucked their shins, were sold, or stopped training for unrelated reasons. Fifty-six of the horses bucked their shins, and the other 170 completed the observation period or were sold while in training.

Stable two, with the highest breezing rate, had the lowest incidence of bucked shins (9.3%), while stables one and four had the highest galloping rates and the highest incidence of bucked shins (41.3% for stable one). Thus, breezing was found to be protective against bucked shins, while galloping increased the likelihood of the problem.

Nunamaker warned that the hazard ratios of this study can only be applied to this breezing rate, which was still far lower than the galloping rate, and that increasing the breezing rate will change these numbers. Long-distance breeze rates are detrimental, he added.

“The high incidence of bucked shins in Thoroughbreds suggested that loading to produce such peak strains and concomitant adaptive remodeling did not occur in a large number of Thoroughbreds in classical training programs before racing,” said Nunamaker. “Increasing the number of short distance works (breezes) from once every 10 days to three times a week produced large changes in the modeling, remodeling, and inertial property measurements of MC3. Classical training produced little progressive change in the inertial properties of MC3, seemingly no better than no training at all; whereas the new modified training program showed inertial property (Imin and Imax) development that equaled or surpassed that observed in established older Thoroughbreds, those horses apparently no longer susceptible to bucked shins.”

The Modified Training Program

The training program developed by Nunamaker’s team (with John Fisher, VMD, a trainer in Fairhill, Md.) assumes that the young horse is broken to ride in fall and is able to gallop a mile by the end of December, and starts training in January. The guiding principle is that the young horse’s bones need to see the strain environment of racing as soon as possible so that bone modeling and remodeling can begin in a timely manner, he explained. The horses are worked six days a week, walked to and from the track, walked 0.5 mile on the track, and jogged 0.5 mile to warm up, then galloped a mile.

Stage 1--Horses finish the gallops two times a week (Tuesdays and Saturdays) with the last one-eighth mile completed in an open gallop in 15 seconds. This stage lasts for five full weeks.

Stage 2--The Tuesday and Saturday open gallops increase to one-quarter mile in 30 seconds, and this is includes in the one-mile total gallop. This stage also lasts five full weeks.

Stage 3--This stage adds speed work once a week (Saturday), breezing a quarter-mile in about 26 seconds, for four weeks. The daily gallops lengthen to 1 1/4 miles twice a week, and after the fourth week the one-quarter-mile breeze is continued with a strong “gallop out” for another furlong (about 40 seconds total for the breeze). This is done for three weeks, making stage 3 seven weeks long.

“After stage 3, the horses have effectively established their MC3’s shape and architecture for the longer high-speed workouts necessary for racing,” Nunamaker said. “They can now go on to four- to six-furlong works as needed to develop their other body systems to complete fitness for racing.”

He noted an incidence of bucked shins of less than 5% with this program.

The program takes 119-147 days, he added, depending on race availability and not including down time for illness or injury. Gate work begins early and often, and mental development is reportedly very good because of the relaxed walking to and from the track along with the familiarity with speed work.

If a trainer has to work with a horse with bucked shins, Nunamaker recommended a physical examination and training program re-evaluation, with a shift toward the modified program described above. In the event that a horse has to take time off because of illness, at 10 days off, he suggested a one-month backup in the training schedule, as this time off might be enough to activate bone remodeling. The initial phase of remodeling is, of course, resorption of bone and thus initial weakening before strengthening.

The Bottom Line

“It seems that horses are not born with the right bone structure for racing,” Nunamaker concluded. “They must develop it. Bone can only develop based on its own experience. Training adapts bone to training, and training that mimics racing adapts bone to racing. Bucked shins do not have to be a part of normal training for racing of young Thoroughbred racehorses.”


Near the end of the first part of his presentation, Nunamaker tantalized the audience with a short discussion of shoe type’s effect on forces on the hoof (and thus the leg). Toe grabs have been shown to correlate with increased rates of injury, and Nunamaker’s group took this concept a step further by measuring accelerations of the hoof with toe-grab shoes. A measurement package was glued to the hoof for evaluation of toe-grab shoes attached with nails and attached by glue.

They found that the highest accelerations were on take-off, reaching as high as 460 Gs (gravitational force equivalents) with nailed-on toe-grab shoes. Landing deceleration forces reached around 250 Gs. The glued-on toe-grab shoe yielded about 260 Gs of take-off force, and about 160 Gs at landing.

Nunamaker then previewed the results of similar testing of a new shoe he designed. The entire ground surface of this shoe is reminiscent of a serrated knife with low-profile teeth hooking rearward, yielded take-off forces of less than 200 Gs, and landing forces of less than 100 Gs. Nunamaker theorized that this design allowed the hoof to slide forward a bit on landing, thus avoiding the snatch of a “grabbier” shoe, while providing plenty of take-off traction--all without changing the angle of the foot as a toe grab does.

The University of Pennsylvania is currently in the process of negotiating a license agreement with a manufacturing company for production and distribution of this shoe design.

About the Author

Christy M. West

Christy West has a BS in Equine Science from the University of Kentucky, and an MS in Agricultural Journalism from the University of Wisconsin-Madison.

Stay on top of the most recent Horse Health news with FREE weekly newsletters from Learn More

Free Newsletters

Sign up for the latest in:

From our partners