Associations between hoof shape and the position of the frontal plane ground reaction force vector in walking horses

AIMS: To determine the frontal plane position of the ground reaction force vector at its centre of pressure under the hoof of walking horses, and its projection through the distal limb joints, and to relate this to hoof geometric measurements.

METHODS: Reflective markers were glued to the forelimb hooves and skin of 26 horses, over palpable landmarks representing centres of the coffin, fetlock and carpal joints, and the dorsal toe at its most distal point. A 4-camera kinematic system recorded the position of these markers as the horse walked in hand across a force platform, to generate a frontal plane representation of the ground reaction force vector passing between the markers at the joints. The position of the vector was calculated as the relative distance between the lateral (0%) and medial (100%) markers at each joint. Digital photos were taken of the hoof in frontal and sagittal views to determine hoof geometric measurements. Associations between these and the position of the force vector at each joint were examined using Pearson correlation coefficients.

RESULTS: Mean vector position for both forelimbs at the toe, coffin, fetlock and carpal joint was 50.1 (SD 8.9), 53.0 (SD 9.2), 54.6 (SD 11.4) and 50.5 (SD17.3)%, respectively, of the distance between the lateral and medial sides of the joint in the frontal plane. Across all four joints, the vector position was slightly more medial (2–4%) for the right than left limb (p>0.05). Medial hoof wall angle was correlated (p<0.05) with force vector position at the fetlock (r=?0.402) and carpal (r=?0.317) joints; lateral hoof wall angle with vector position at the toe (r=0.288) and carpal (r=?0.34) joint, and medial hoof wall height with vector position at the fetlock (r=?0.306) and carpal (r=?0.303) joints.

CONCLUSION: The position of the two-dimensional frontal plane ground reaction force vector at the toe, and at the fetlock and carpal joints was associated with hoof shape. Mediolateral hoof balance has been shown in vitro to affect articular forces, which may be a factor in development of joint disease. The effect of hoof shape needs to be evaluated at faster gaits to determine the potential for joint injury in the presence of larger forces.     

Vertical ground reaction force-time histories of sound Warmblood horses trotting on a treadmill

The objective of this study was to establish representative treadmill ground reaction force (GRF) and interlimb co-ordination time data of clinically sound horses at the trot. It was anticipated that these normative standards would provide a reference data base against which lame horses could be compared. GRF-time histories were collected from 30 Warmblood riding horses with easy, wide natural gaits. Data were recorded of all four limbs simultaneously by the use of an instrumented treadmill. A total of 912 stride cycles per limb were analysed for force, time and spatial parameters and were averaged. The shape and amplitude of the treadmill force curves were very similar to force traces recorded with a stationary force plate. The horses showed a high degree of symmetry in all investigated parameters (95% reference interval of left-right asymmetry +/-1.8-6.8%). No significant differences were found between left and right mean values. Intra-individual coefficients of variance of the various parameters did not exceed 2.7%. Inter-individual coefficients of variance were 2.5-3.5 times larger than the respective intra-individual coefficients. An instrumented treadmill provides a number of decisive advantages, such as time-efficient data acquisition of all four feet simultaneously over successive strides, or the high regularity of the horse's gait pattern at controlled velocities, which allow the clinical assessment of locomotor performance of horses.     

The hindlimb in walking horses: 1. Kinematics and ground reaction forces

The objective was to study associations between kinematics and ground reaction forces in the hindlimb of walking horses. Video (60 Hz) and force (2000 Hz) data were gathered for 8 strides from each of 5 sound horses during the walk. Sagittal plane kinematics were measured concurrently with the vertical and longitudinal ground reaction forces. The hindlimb showed rapid loading and braking in the initial 10% stride. The stifle, tarsal and coffin joints flexed and the fetlock joint extended during this period of rapid loading. The vertical ground reaction force showed 2 peaks separated by a dip; this pattern was similar to the fetlock joint angle-time graph. Peaks in the longitudinal ground reaction force did not appear to correspond with kinematic events. Total braking impulse was equal to total propulsive impulse over the entire stride. Flexion and extension of the hip were responsible for protraction and retraction of the entire limb. Maximal protraction occurred shortly before the end of swing and maximal retraction occurred during breakover. During the middle part of stance the tarsal joint extended slowly, while the stifle began to flex when the limb was retracted beyond the midstance position at 28% stride. Flexion cycles of the stifle and tarsal joints were well coordinated during the swing phase to raise the distal limb as it was protracted. The results demonstrate a relationship between limb kinematics and vertical limb loading in the hindlimbs of sound horses. Future studies will elucidate the alterations in response to lameness.     

Effects of ground surface deformability, trimming, and shoeing on quasistatic hoof loading patterns in horses

OBJECTIVE: To determine whether solar load distribution pattern on a solid nondeformable ground surface is the product of contact erosion and is the mirror image of load distribution on a deformable surface in horses. ANIMALS: 30 clinically normal horses. PROCEDURES: Solar load distribution was compared among 25 clinically normal horses during quasistatic loading on a solid nondeformable surface and on a highly deformable surface. Changes in solar load distribution patterns were evaluated in 5 previously pasture-maintained horses housed on a flat nondeformable surface. Changes in solar load distribution created by traditional trimming and shoeing were recorded. RESULTS: Unshod untrimmed horses had a 4-point (12/25, 48%) or a 3-point (13/25, 52%) wall load distribution pattern on a flat solid surface. Load distribution on a deformable ground surface was principally solar and located transversely across the central region of the foot. Ground surface contact areas on solid (24.2 +/- 8.62 cm2) and deformable (69.4 +/- 22.55 cm2) surfaces were significantly different. Maintaining unshod horses on a flat nondeformable surface resulted in a loss of the 3- and 4-point loading pattern and an increase in ground surface contact area (17.9 +/- 2.77 to 39.9 +/- 12.77 cm2). Trimming increased ground surface contact area (24.2 +/- 8.60 to 45.7 +/- 14.89 cm2). CONCLUSION AND CLINICAL RELEVANCE: In horses, the solar surface is the primary weight-loading surface, and deformability of ground surface may have a role in foot expansion during loading. Increased surface area induced by loading on deformable surfaces, trimming, and shoeing protects the foot.     

Habituation of healthy dogs to treadmill trotting: Repeatability assessment of vertical ground reaction force

To assess the repeatability of kinetic gait analysis with a treadmill, 28 sound adult dogs were made to trot on an instrumented system. Vertical ground reaction force variables (Peak PFz and Impulse IFz) were collected by 10-s recordings, once a week, 4 weeks in succession. Data were analysed using a repeated-measures two-way ANOVA to investigate habituation to treadmill locomotion. Recorded data were stabilized from the end of the first and second sessions for IFz and PFz, respectively. The percentages of variance attributable to dogs, weeks, minutes and repetitions were, respectively, 72%, 10%, 7%, and 11% for PFz and 84%, 7%, 3%, and 6% for IFz. Habituation occurred after a single training session. Good repeatability was determined by a low coefficient of variation (3.4-4.7%). Adding a treadmill to kinetic gait analysis deserves consideration: reliable data are easily recorded using appropriate habituation and statistical model.     

Ground reaction force patterns of Dutch Warmblood horses at normal walk

The ground reaction force patterns from 20 clinically sound Dutch Warmblood horses (Group A) were recorded at the normal walk. The data from four to 10 stance phases of each limb were computer averaged after normalisation to the animal's body mass and to the stance time. This analysis method allowed comparison of data from left and right fore- and hindlimbs within and between horses. The left-to-right symmetry in the reaction force peaks of contralateral limbs of one horse exceeded 90 per cent. The time in the stance phase at which the peaks occur were even more symmetrically distributed. A characteristic force-force diagram was constructed by plotting the longitudinal horizontal and the vertical ground reaction forces against each other; in this way the symmetry of loading of contralateral fore- and hindlimbs could be interpreted easily. Force plate tracings were obtained from eight horses (Group B) in three successive years. The similarity of the tracings from a sound, well-trained horse over that period was better than the differences between horses of the same breed.     

Relationships between hoof-acceleration patterns of galloping horses and dynamic properties of the track

OBJECTIVE: To define relationships between hoof-acceleration patterns of galloping horses and dynamic properties of the track. ANIMALS: 8 Thoroughbred horses without lameness. PROCEDURE: Acceleration-time curves were recorded by use of accelerometers attached to each hoof as each horse galloped over the track straightaway. Four sessions were conducted for each horse, with the track surface modified by sequentially adding water before each session. These acceleration-time curves were analyzed to determine peak accelerations during the support phase of the stride. Track dynamic properties (hardness, rebound, deceleration rate, rebound rate, and penetration) were recorded with a track-testing device. Moisture content and dry density were measured from soil samples. Stepwise multiple regression was used to identify relationships between hoof-acceleration variables and track dynamic properties. RESULTS: Track rebound rate was most consistently related to hoof variables, especially through an inverse relationship with negative acceleration peaks for all hooves. Also, rebound rate was related to initial acceleration peak during propulsion of the hooves of the forelimb and the nonlead hind limb as well as to the second acceleration peak during propulsion of the lead hooves of the hind limb and nonlead forelimb. CONCLUSIONS AND CLINICAL RELEVANCE: The inverse relationship between track rebound rate and negative acceleration peaks for all hooves reflects the most important dynamic property of a track. Any factor that reduces negative acceleration of the hooves will increase stride efficiency by allowing smoother transition from retardation to propulsion and therefore may be important in determining the safety of racing surfaces.     

Use of an instrument sandwiched between the hoof and shoe to measure vertical ground reaction forces and three-dimensional acceleration at the walk, trot, and canter in horses

OBJECTIVE: To develop an instrument that could be sandwiched between the hoof and shoe of horses and that would reliably measure vertical ground reaction forces and three-dimensional acceleration at the walk, trot, and canter. ANIMALS: 5 clinically sound Thoroughbreds. PROCEDURES: The recording instrument (weight, 350 g) consisted of 2 metal plates, 2 bolts, 4 load cells, and 3 accelerometers. It was mounted to the hoof with a glue-on shoe and devised to support as much load exerted by a limb as possible. The load cells and accelerometers were wired to a 16-channel transmitter, and transmitted signals were received and amplified with a telemetry receiver. RESULTS: The recording instrument could measure in real time the 4 components of the ground reaction force or their resultant force along with acceleration in 3 dimensions as horses walked, trotted, or cantered on a treadmill. Patterns of force-time curves recorded for consecutive strides were similar to each other and to those previously reported, using a force plate. CONCLUSIONS AND CLINICAL RELEVANCE: The recording instrument developed for use in the present study allowed us to record vertical ground reaction force and acceleration in 3 dimensions in horses at the walk, trot, and canter.     

Effect of walking velocity on ground reaction force variables in the hind limb of clinically normal horses

OBJECTIVE: To measure the effect of subject velocity on hind limb ground reaction force variables at the walk and to use the data to predict the force variables at different walking velocities in horses. ANIMALS: 5 clinically normal horses. PROCEDURE: Kinematic and force data were collected simultaneously. Each horse was led over a force plate at a range of walking velocities. Stance duration and force data were recorded for the right hind limb. To avoid the effect of horse size on the outcome variables, the 8 force variables were standardized to body mass and height at the shoulders. Velocity was standardized to height at the shoulders and expressed as velocity in dimensionless units (VDU). Stance duration was also expressed in dimensionless units (SDU). Simple regression analysis was performed, using stance duration and force variables as dependent variables and VDU as the independent variable. RESULTS: Fifty-six trials were recorded with velocities ranging from 0.24 to 0.45 VDU (0.90 to 1.72 m/s). Simple regression models between measured variables and VDU were significant (R2 > 0.69) for SDU, first peak of vertical force, dip between the 2 vertical force peaks, vertical impulse, and timing of second peak of vertical force. CONCLUSION AND CLINICAL RELEVANCE: Subject velocity affects vertical force components only. In the future, differences between the forces measured in lame horses and the expected forces calculated for the same velocity will be studied to determine whether the equations can be used as diagnostic criteria.     

The relationship between digital perfusion pressure and hoof lamellar blood flow in isoflurane-anesthetized horses

Digital perfusion pressure (DPP) equals mean arterial pressure (MAP) at the hoof coronet minus digital interstitial pressure (DIP) within the hoof. To test whether lamellar blood flow (LBF) changes proportionately to DPP, anesthesia was induced and maintained with isoflurane in six horses to target a MAP of 60 mmHg. Arterial, venous, and hoof interstitial pressures were measured in each pelvic limb. LBF was measured using fluorescent microspheres during dobutamine infusions targeting either 60 (low), 80 (medium), or 100 (high) mmHg MAP. Following euthanasia, hoof lamina was collected for microsphere isolation. To reduce intra-individual variability, medium and high pressures and flows were divided by their respective low pressure and flow baseline values, yielding indexed variables of ΔLBF and ΔDPP. The ΔLBF correlated negatively with the ΔDPP. We conclude that LBF was not solely determined by passive pressure-flow relationships and that systemic hypertension may not effectively increase dermal LBF in horses.     

Instrumented treadmill for measuring vertical ground reaction forces in horses

OBJECTIVE: To develop and validate a novel instrumented treadmill capable of determining vertical ground reaction forces of all 4 limbs simultaneously in horses. SAMPLE POPULATION: Data obtained while a horse was walking and trotting on the treadmill. PROCEDURE: 18 piezo-electric force transducers were mounted between the treadmill frame and supporting steel platform to measure the actual forces at the corresponding bearing points. Each of the 18 sensor forces is equal to the sum of the unknown hoof forces weighted with the transfer coefficients of the corresponding force application points. The 4 force traces were calculated, solving at each time point the resulting equation system, using the Gaussian least-squares method. System validation comprised the following tests: determination of the survey accuracy of the positioning system, determination of the natural frequencies of the system, linearity test of the force transfer to the individual sensors, determination of superimposed forces with the treadmill-integrated force measuring system (TiF) in a static configuration, and comparison of vertical ground reaction forces determined simultaneously by use of TiF and force shoes mounted on the forelimbs of a horse. RESULTS: Comparison between static test loads and TiF-calculated forces revealed deviations of < 1.4%. Force traces of TiF-calculated values and those recorded by use of the force shoes were highly correlated (r > or = 0.998). CONCLUSIONS AND CLINICAL RELEVANCE: This instrumented treadmill allows a reliable assessment of load distribution and interlimb coordination in a short period and, therefore, is suitable for use in experimental and clinical investigations.     

The use of H(orse) INDEX: a method of analysing the ground reaction force patterns of lame and normal gaited horses at the walk

The amplitudes, impulses and times of occurrence of a number of selected peaks in the ground reaction force tracings of 17 horses with various clinical histories were compared with those of 20 horses used to establish values for the 'standard' Dutch Warmblood horse. The resulting factors were combined, using different weighting factors, into indices characterising each limb. The symmetry between the loading of the forelimbs and the hindlimbs was used to calculate amplitude and peak-time symmetry indices. Limb and symmetry indices were combined in one H(orse) INDEX. This method of quantifying the ground reaction force pattern, together with appropriate graphic display of the tracings, was useful in clinical evaluation of force plate measurements.     

Kinematics and ground reaction forces in horses with superficial digital flexor tendinitis

OBJECTIVE: To measure and correlate kinematic and ground reaction force (GRF) data in horses with superficial digital flexor tendinitis. ANIMALS: 6 sound horses. PROCEDURE: Horses were evaluated before (sound evaluation) and after (lame evaluation) induction of superficial digital flexor tendinitis in 1 forelimb (randomized) by injection of collagenase. As each horse trotted, kinematic data were collected by use of an optoelectronic system, and GRF data were measured by use of a force plate. Three-dimensional kinematic and GRF data were projected onto a 2-dimensional sagittal plane. RESULTS: Lame limbs had significantly lower peak vertical GRF, less flexion of the distal interphalangeal joint, and less extension of the metacarpophalangeal joint, compared with compensating limbs. Carpal joint kinematics did not change. Compensating limbs had a more protracted orientation throughout the stance phase and higher braking longitudinal force and impulse; however, total range of rotation from ground contact to lift off did not change. Transfer of body weight from lame to compensating limbs was smooth, without elevation of the body mass into a suspension phase. Propulsive components of longitudinal GRF did not differ between limbs. CONCLUSIONS AND CLINICAL RELEVANCE: In horses with experimentally induced superficial digital flexor tendinitis, changes in vertical GRF were reflected in angular excursions of the distal interphalangeal and metacarpophalangeal joints, whereas changes in longitudinal GRF were associated with alterations in the protraction-retraction angle of the entire limb.     

The forelimb in walking horses: 1. Kinematics and ground reaction forces

Video (60 Hz) and force (2000 Hz) data were collected from 5 sound horses during walking. Forelimb data were analysed for 8 strides (4 left, 4 right) per horse to determine sagittal plane kinematics and ground reaction forces (GRFs). The results suggested that brachial rotation was responsible for protraction and retraction of the limb as a whole, while rotations of the scapula and antebrachium elevated the distal limb during breakover and early swing then lowered it in preparation for ground contact. The coffin joint was flexed maximally at the time of peak longitudinal braking force, which occurred during breakover of the contralateral forelimb. The metacarpus was vertical at 28% stride. This was considerably earlier than the change from a braking to a propulsive longitudinal force (34% stride), which coincided with maximal extension of the fetlock joint. The longitudinal propulsive force peaked just after contact of the contralateral forelimb. During the swing phase the joints distal to the shoulder showed a single flexion cycle that peaked at 76% stride at the carpus, 81% stride at the fetlock and 84% stride at the elbow and coffin joints. The coffin and shoulder joints began to extend in the terminal swing phase and continued to extend through ground contact and early stance. The results provide normative data that will be applied in detecting changes in kinematics and ground reaction forces that are associated with specific lamenesses.     

External characteristics of the lateral aspect of the hoof differ between non-lame and lame horses

During a pre-purchase examination (PPE) there is always a debate about how clinical findings of the hoof different from ideal should be interpreted in relation to future lameness risk and/or unsuitability of the horse for the potential purchaser. The objectives of this study were to describe and compare external angular measurements, linear ratios and hoof capsule characteristics of non-lame and lame feet. Photographs of feet from 300 horses with foot pain and 25 non-lame horses were analysed. Hoof wall, heel and coronary band angles and hoof wall length and height, weight-bearing length, coronary band length and height of the coronary band at dorsal and palmar locations were measured and expressed as linear ratios.Mean hoof wall, heel and coronary band angles were larger in lame compared with non-lame feet; only the ratio of dorsal to palmar coronary band heights and the shape of the coronary band were significantly different between lame and non-lame horses. Growth rings were divergent and horn tubules were non-parallel in lameness of >3 months. At a PPE, a larger ratio of dorsal to palmar coronary band heights in one limb may be indicative of previous lameness in that foot. Changes in coronary band shape and divergent growth rings and horn tubules would suggest a longer duration.