Lameness is one of the most important medical issues of horses1 and accounts for annual losses up to $1 billion for the US equine industry.2 Horses with subclinical or mild lameness often have suboptimal performance.3,4 Early detection of mild lameness is important for horses, particularly competition horses, in which suboptimal performance can be career limiting, so that appropriate measures can be taken to alleviate the lameness and improve the performance and quality of life of affected horses.
Lameness in horses is generally detected and monitored by means of a subjective lameness examination.5 The most commonly used lameness scoring systems have grades, or scores, that are defined by specific criteria; however, there can be substantial variability within a grade such that use of that system can be challenging for longitudinal monitoring of lameness in an individual horse, especially when the severity of lameness changes only minimally between examinations.6 Results of multiple studies6–9 suggest that subjective lameness scoring systems are not reliable for clinical use, particularly for horses with mild lameness. Additionally, in 1 study,6 observer bias was detected in the subjective lameness scores assigned to horses following administration of perineural anesthesia. Therefore, an accurate, objective method is needed to supplement the subjective lameness examination for the detection and monitoring of horses with mild lameness as well as to assess the response of those horses to perineural anesthesia.
Several studies5,10–15 have been conducted to investigate the efficacy of kinetics and kinematics for the evaluation of horses with mild lameness. In a study10 in which stationary force platform analysis was used as a kinetic method to detect horses with lameness, peak vertical force and impulse parameters were significantly decreased in horses with mild lameness (ie, grade, < 1.5/5), compared with those for nonlame horses. Unfortunately, the use of stationary force platform analysis for evaluation of lameness in horses is limited because of the lack of availability of equipment and experienced personnel to run the equipment, the increased time required for data collection and analysis, and expense, compared with the time and expense for a subjective lameness examination. In horses, optical methods can be used to detect alterations in distal limb kinematics such as stride length, step length, hoof height, and sagittal-plane joint angles after induction of lameness.11–13 Although alterations in those variables have been detected at both a walk and a trot, the alterations are more pronounced at a trot.11 Because optical kinematics suffer from many of the limitations of stationary force platform analyses, other kinematic analysis systems are currently being investigated to objectively characterize lameness in horses. These kinematic analysis systems use multiple microelectromechanical components, such as accelerometers, gyroscopes, and GPS tracking devices, which have wireless or telemetric components for data transmission.14–16 Results of a study5 indicate that the use of an inertial sensor system that monitors movement of the horse's head or pelvis during a trot detected unilateral forelimb or hind limb lameness earlier (ie, when lameness was less severe) than did 3 experienced equine veterinarians who used a subjective lameness scoring system. However, the use of an inertial sensor system to detect lameness in horses at a walk or for longitudinal assessment of lameness in an individual horse has not been evaluated.17 Because inertial sensors are becoming increasingly small and lightweight, it should be possible to attach them to the distal aspect of a limb of a horse without causing substantial alteration to the movement of that limb. In fact, the rigid attachment of an inertial sensor to the hoof is ideal because motion artifact is eliminated. In horses, the kinematics of the limbs change when 1 limb becomes lame11–13; therefore, measurement of hoof kinematics might be another method to diagnose and monitor lameness. Although hoof displacement or position has been investigated in lame horses,11–13 to our knowledge, no studies have been conducted to evaluate other linear and angular changes in the forelimb hooves of horses following induction of unilateral lameness. Optical methods remain the gold standard for collection of kinematic data, and intra- and interlimb comparisons of kinematic changes might be useful for the identification of horses with mild lameness. Linear and angular limb movement can also be measured by means of an IMU, which could be used in a horse-mounted method to evaluate lameness.
The objective of the study reported here was to use optical methods to characterize kinematic variables of the hoof for horses at a trot before and after induction of unilateral, weight-bearing forelimb lameness and following administration of perineural anesthesia to alleviate that lameness. Following lameness induction, we hypothesized that sagittal-plane kinematic variables for the lame limb would differ significantly from those of that limb prior to lameness induction (baseline) and those of the contralateral nonlame limb, and these differences would be detectable even at the mildest grade of lameness induced. We also hypothesized that following perineural anesthesia of the medial and lateral palmar nerves of the lame limb, the sagittal-plane kinematics of that limb would not differ from those at baseline.
Materials and Methods
Animals—Six Quarter Horses were used for the study reported here as well as a companion study,18 and the data for both studies were obtained concurrently. Each horse was determined to be clinically normal on the basis of results of a physical examination and was not perceptibly lame at a walk or trot (ie, had a subjective lameness score of 0/5 as determined with the lameness scale developed by the American Association of Equine Practitioners19). The age of the horses ranged from 2 to 9 years, and the mean ± SD body weight and wither height of the horses were 364 ± 19 kg and 1.46 ± 0.03 m, respectively. Prior to initiation of the study, all horses were acclimated to the laboratory where the gait analysis data were collected. The hooves of each horse were trimmed and balanced, and the forelimb hooves were shod. A steel keg shoe (weight, 324.8 ± 23.5 g) was applied to the hoof of the left forelimb, and a similar steel keg shoe (weight, 333.7 ± 25.6 g; Figure 1) with a nut welded to the inner edge of both the medial and lateral branches of the shoe between the third and fourth nail holes was applied to the hoof of the right forelimb. The nuts were welded to the shoe in a manner similar to that described in another study20 such that they were flush with the solar aspect of the shoe and did not contact the horse's sole during weight bearing. All study procedures were approved by the Colorado State University Institutional Animal Care and Use Committee.
Photographs of a modified steel keg shoe that was applied to the hoof of the right forelimb of each of 6 clinically normal Quarter Horses for induction of lameness by means of a sole-pressure model. A—A nut was welded to the inner edges of the medial and lateral branches of the shoe between the third and fourth nail holes such that it was flush with the solar aspect of the shoe. B—Lameness was induced by threading a 6-mm-diameter screw with a blunt or tapered end into each nut such that the head of the screw was flush with the ground when the horse was bearing weight on the limb. The length of the screws used to induce lameness varied and was positively associated with the grade, or severity, of lameness that was induced. Three grades of lameness were induced in the right forelimb in a sequential manner (ie, induction of grade 1 lameness, followed by grade 2, and then grade 3). Grade of lameness was subjectively determined at a trot. Grade 1 was defined as intermittent lameness; grade 2 was defined as consistent, mild lameness; and grade 3 was defined as consistent, moderate lameness, and none of the grades resulted in lameness at a walk.
Citation: American Journal of Veterinary Research 74, 9; 10.2460/ajvr.74.9.1183
Study design—For each horse, kinematic data for both forelimbs were obtained before (baseline) and after induction of each of 3 grades (grades 1, 2, and 3) of increasing lameness in the right forelimb as well as after the administration of perineural anesthesia to the right forelimb to alleviate the lameness. The lameness grades induced were subjectively defined and were modified slightly from the lameness scale developed by the American Association of Equine Practitioners.19 Briefly, grade 1 was defined as an intermittent lameness at a trot; grade 2 was defined as a consistent, mild lameness at a trot; and grade 3 was defined as a consistent, moderate lameness at a trot. None of the grades resulted in lameness at a walk. Horses were allowed to rest for several minutes between data collection periods to minimize the effect of fatigue.
Instrumentation of horses—An aluminum base plate (8.8 × 1.9 × 0.3 cm; weight, 14.2 g) was adhered to the hoof of each forelimb with hoof acrylic.a To this base plate, screws were used to attach another piece of aluminum (7 to 9 × 1.5 × 0.1 cm; weight, 3.4 g) that was conformed to the dorsal aspect of the hoof to provide additional support and surface area, to which a strain gauge was attached with adhesive. A cable connected each strain gauge to a data collection source that was mounted on the horse's back with a surcingle that encircled the thorax at approximately the level of the sixth and seventh ribs.
Two 4-mm screws were used to attach a marker triad to the lateral aspect of the base plate (Figure 2). The marker triad (15 × 13 cm; weight, 37.6 g) was composed of an aluminum frame stiffened with a uniaxial carbon sandwich structure with a balsa wood core (4.6 × 2.8 × 0.6 cm) and moved rigidly with the hoof. Three spherical retroreflective markers (diameter, 2.0 cm) were attached at the distal aspect of each arm of the triad with machine screws such that the markers were 10 to 11 cm apart. An IMUb (5.1 × 3.8 × 1.6 cm; weight, 58.6 g) was attached to the marker triad of the right forelimb, and a custom-machined piece of metal (3.6 × 3.1 × 1.2 cm; weight, 75.7 g) was attached to the marker triad of the left forelimb. A cable connected the IMU to a handheld computer that was mounted on the horse's back adjacent to the data collection source for the strain gauges with the same surcingle (combined weight of the surcingle, data collection source for strain gauges, and handheld computer, 9.5 kg). The cables for the IMU and strain gauges were loosely secured to each forelimb at the distal aspect of the metacarpus and distal aspect of the antebrachium with an elastic bandage.c The total weight of the marker triad attached to the right forelimb was 113.8 g, and the total weight of the marker triad attached to the left forelimb was 130.9 g; the weight of the marker triad attached to left forelimb was greater than that of the marker triad attached to the right forelimb because of the larger mass of the machined metal on the left marker triad; this difference was balanced by the additional weight of the cable that was attached to the IMU unit on the right forelimb.
Photograph of the marker triad that was attached to the lateral aspect of the hoof of the right forelimb of each of 6 clinically normal Quarter Horses to obtain optical kinematic data. The marker triad (15 × 13 cm; weight, 37.6 g) was positioned on the hoof such that it moved rigidly with the hoof and did not touch the metacarpophalangeal joint. It was composed of an aluminum frame stiffened with a uniaxial carbon sandwich structure with a balsa wood core (4.6 × 2.8 × 0.6 cm). Three spherical retroreflective markers (diameter, 2.0 cm) were attached at the distal aspect of each arm of the triad with machine screws such that the markers were 10 to 11 cm apart. An IMU (5.1 × 3.8 × 1.6 cm; weight, 58.6 g) was attached to the marker where the 3 arms intersected. A similar marker triad was attached to the lateral aspect of the hoof of the left forelimb of each horse, except the IMU was replaced with a custom-machined piece of metal (3.6 × 3.1 × 1.2 cm; weight, 75.7 g). Superimposed on the photograph are the virtual hoof marker (black circle), which served as the reference point for monitoring the motion of the hoof, and vectors (white arrows) to indicate the directions of the kinematic measurements that were obtained in the sagittal plane. The cranial-to-caudal (X) axis was positive cranially, the vertical (Z) axis was positive proximally, and the sagittal-plane rotation (θ) was positive in a counterclockwise rotation (ie, heel-down was positive).
Citation: American Journal of Veterinary Research 74, 9; 10.2460/ajvr.74.9.1183
Induction of lameness and perineural anesthesia—For each horse, 3 grades of lameness were induced in the right forelimb in a sequential manner (ie, induction of grade 1 lameness, followed by grade 2, and then grade 3). A 6-mm-diameter screw with either a blunt or 2-mm-diameter tapered end was threaded completely into both the medial and lateral nuts welded to the steel shoe such that the head of the screw was in contact with the ground when the horse was bearing weight on the right forelimb (Figure 1). The horse was then trotted briefly to subjectively determine the severity of lameness. This process was repeated with longer or shorter screws as necessary until each desired severity of lameness was achieved, at which time lameness trials were performed for data collection purposes. The screw length required to induce each grade of lameness was recorded for each horse. The weight (median, 7.8 g; range, 6.8 to 10.6 g) added to the shoe of the right forelimb by the 2 screws used for lameness induction varied with the screw length required to achieve the desired severity of lameness.
Following data collection for grade 3 lameness, perineural anesthesia was administered to the right forelimb of each horse. Briefly, 3 mL of a 2% mepivacaine solution was injected SC in the regions surrounding the medial and lateral palmar nerves. If the lameness in the right forelimb had not resolved by > 80% at 10 minutes after mepivacaine administration, an additional 1.5 mL of a 2% mepivacaine solution was injected SC in the regions surrounding the medial and lateral palmar nerves, and the horse was reassessed 5 minutes later.
Lameness trial facilities and protocol—Horses were walked or trotted on a runway (length, 24.8 m; width, 1.2 m) that consisted of an asphalt surface covered with a 9.3-mm-thick rubberized mat. The optical capture volume (ie, portion of the runway where kinematic data for the horse were obtained) had a length of 3.7 m, width of 1.3 m, and height of 2.4 m and was located near the midway point of the runway such that the horse could maintain a constant velocity when passing through it. Three-dimensional optical kinematic data were obtained with 8 infrared camerasd that were suspended from overhead beams; 4 cameras were suspended on each side of the optical capture volume. The cameras operated at 200 Hz and were connected to an optical kinematic systeme that was calibrated to yield coordinates to within 1.2 mm.
The velocity of each horse during each trial (ie, trip over the runway) was calculated by the use of 5 infrared timing gatesf that were spaced 1.5 m apart along the central portion of the runway, which included the entire optical capture volume. The timing gates were triggered to send a signal to the optical kinematic system by the horse as it traveled on the runway, and the time stamps of those signals were used to calculate the horse's mean velocity.
During collection of baseline data, the mean velocity at a walk and a trot was determined for each horse. Subsequently, for each horse during each data collection period, 4 to 5 acceptable trials were recorded for the right and left forelimbs at both a walk and trot. An acceptable trial was defined as a trial during which the horse traveled straight through the optical capture volume at a consistent velocity that was within ± 10% of its mean baseline velocity for the gait being evaluated. Because not all trials through the optical capture volume contained a full stance and swing for both forelimbs, a horse might have to perform up to 8 trials during each data collection period to ensure that 4 acceptable trials were obtained for each forelimb.
Data collection—Optical coordinate data were low-pass filtered at 15 Hz with a recursive fourth-order Butterworth filter. A virtual marker was created between the cranial and caudal retroreflective markers of the marker triad, and this served as a local origin to track the motion of the hoof. The linear movement of the hoof was tracked in the sagittal plane (cranial-to-caudal [X] and proximal-to-distal [Z] acceleration; Figure 2). Hoof events were determined by evaluation of the X and Z acceleration profiles of the stride. Briefly, hoof contact was defined as the last peak in the Z acceleration curve before a period of smaller accelerations. Heel-off was defined as the first peak in the Z acceleration curve after the period of smaller accelerations. Toe-off was defined as the second peak in the Z acceleration curve, which also corresponded to an inflection point in the X acceleration curve. These hoof events were used to divide the stride into total-stance (hoof-contact to toe-off), break-over (heel-off to toe-off), total-swing (toe-off to hoof-contact), initial-swing (toe-off to initial 25% of swing), and terminal-swing (75% of swing to hoof-contact) segments.
Toe-off was set as the global origin for the coordinate system; thus, translations of the hoof at all other events were relative to the location of the virtual marker at toe-off. To ensure that the coordinate system was in line with the direction of travel by the horse, the global x-axis was aligned with the virtual marker at the second hoof contact. Subsequently, the global x- and z-axes were positive cranially and proximally, respectively. Within the sagittal plane about the y-axis (medial to lateral) through the virtual marker, heel-down hoof orientation (ie, counterclockwise rotation from the lateral aspect; Figure 2) was positive. Hoof orientation was determined with the markers on the triad. Because the marker triad was not perfectly parallel to the ground, the orientation of the hoof during the middle of total stance (when the metacarpal bone was perpendicular to the ground as determined by visual assessment of the optical data) was used to adjust the sagittal orientation of the hoof such that 0° was level with the ground.
For each forelimb during each lameness trial, data collected included the instantaneous position, velocity, acceleration, and sagittal-plane orientation of the hoof at hoof-contact, heel-off, and toe-off. Additionally, the total range of motion for the hoof was determined for the total-swing, initial-swing, and terminal-swing segments of the stride.
Statistical analysis—For each forelimb hoof during each segment of the stride (hoof-contact, break-over, initial-swing, terminal-swing, and total swing), descriptive statistics were generated for variables such as duration and hoof orientation, acceleration, velocity, and range of motion in the sagittal plane. The data for each variable were plotted and visually examined for normality. When necessary, a logarithmic transformation was applied to the data to achieve a normal distribution prior to the performance of statistical comparisons. Each data collection period was considered a treatment, which was categorized as baseline (prior to lameness induction), grade 1, grade 2, grade 3, and after block (after administration of perineural anesthesia). For each variable (ie, outcome of interest), comparisons within and between forelimbs were made by use of mixed ANOVA for repeated measures, in which treatment and forelimb (lame or nonlame) were included as fixed effects, horse identification was included as a random effect, and horse velocity during the trial was included as a confounding variable. For within-limb comparisons, each respective treatment (grade 1, grade 2, grade 3, or after block) was compared with the baseline treatment. All analyses were performed with a commercially available statistical software program,g and values of P < 0.05 were considered significant.
Results
Animals—Lameness was successfully induced in all horses. For the first 2 horses evaluated, blunt-ended screws were used to induce lameness; however, in 1 of those horses, only grade 1 and grade 2 lameness could be induced because the horse developed decreased sensitivity to the sole pressure model, even with the use of the longest screws that were available at that time. Additionally, the longest blunt-ended screws tended to push the shoe away from the hoof instead of threading into the sole. Therefore, for the subsequent 4 horses that were evaluated, longer screws with tapered ends were used, which more readily induced the desired severity of lameness. Consequently, analyses for baseline, grade 1, and grade 2 treatments included data from all 6 horses, whereas analyses for grade 3 and after-block treatments included data from only 5 horses. Within 24 hours after lameness induction, none of the horses had perceptible lameness when trotting.
Intralimb kinematic changes—Select kinematic variables for the right (lame) and left (nonlame) forelimbs during total stance (Table 1) and swing (Table 2) at a trot were summarized. For the lame limb, significant kinematic changes from baseline were detected for grades 1, 2, and 3 and after-block treatments and were most frequently detected during the total stance (hoof-contact and break-over) phase of the stride. For the nonlame limb, significant kinematic changes from baseline were detected for grades 1, 2, and 3 and after-block treatments, and these changes were detected during both the total stance and swing phases of the stride.
Mean (SD) kinematic variables for the lame (right) and nonlame (left) forelimbs during the total stance (hoof contact to toe-off) phase of the stride at a trot for 6 clinically normal Quarter Horses before (baseline) and after induction of 3 grades (grades 1, 2, 3) of increasingly severe lameness in the right forelimb and following perineural anesthesia in the right forelimb to alleviate the lameness (after block).
| Treatment | ||||||
|---|---|---|---|---|---|---|
| Stance segment | Variable | Baseline | Grade 1 | Grade 2 | Grade 3 | After block |
| Hoof-contact | X acceleration (m/s2) | |||||
| Lame | −40.348 (10.071) | −45.114 (15.102)*† | −45.814 (12.104)*† | −46.679 (9.758)*† | −42.930 (10.966)*† | |
| Nonlame | −38.214 (10.219) | −41.073 (13.913) | −41.831 (12.231)† | −38.573 (13.385) | −38.872 (11.858)† | |
| Orientation (°) | ||||||
| Lame | −1.35 (2.35) | −1.28 (2.60)* | −1.19 (2.75)* | −0.36 (2.38)* | −0.47 (2.60) | |
| Nonlame | −0.78 (2.95) | 0.16 (2.22)† | 0.09 (2.59)† | 0.17 (2.69)† | −0.89 (2.31) | |
| Break-over | Duration (s) | |||||
| Lame | 0.055 (0.011) | 0.054 (0.009) | 0.053 (0.009)† | 0.053 (0.009) | 0.052 (0.006)† | |
| Nonlame | 0.054 (0.009) | 0.054 (0.009) | 0.054 (0.009) | 0.052 (0.008) | 0.054 (0.009) | |
| X acceleration (m/s2) | ||||||
| Maximum | ||||||
| Lame | 45.784 (12.707) | 48.726 (12.910)*† | 49.108 (16.059)*† | 51.771 (13.988)* | 48.776 (11.327)* | |
| Nonlame | 44.481 (14.262) | 47.835 (15.087) | 44.797 (13.593) | 42.909 (11.633) | 44.395 (13.451) | |
| Mean | ||||||
| Lame | 27.682 (6.512)* | 28.823 (6.021)* | 28.408 (7.468)*† | 28.598 (6.190)*† | 27.551 (4.423)* | |
| Nonlame | 24.836 (6.650) | 26.199 (6.549) | 25.262 (6.554)† | 25.139 (5.384)† | 23.467 (4.444)† | |
| Orientation (°) | ||||||
| Minimum | ||||||
| Lame | −45.73 (5.96) | −44.82 (5.24)* | −44.04 (3.72)* | −43.27 (3.84)* | −43.63 (5.46)* | |
| Nonlame | −46.98 (7.63) | −48.57 (6.89)† | −47.87 (5.82) | −47.12 (5.02) | −49.29 (5.40) | |
| Range of motion (°) | ||||||
| Lame | 42.09 (5.48) | 41.37 (5.11)* | 40.05 (3.49)*† | 39.75 (3.77)*† | 39.56 (5.42)*† | |
| Nonlame | 43.05 (7.44) | 44.66 (6.82)† | 43.97 (5.80) | 42.96 (4.92) | 44.79 (5.08) | |
Grade of lameness was subjectively determined at a trot. Grade 1 was defined as intermittent lameness; grade 2 was defined as consistent, mild lameness; and grade 3 was defined as consistent, moderate lameness, and none of the grades resulted in lameness at a walk. Each stride taken by a horse was divided into segments on the basis of 3 hoof events that were defined by the kinematic data curves. The hoof events included hoof contact, which was defined as the last peak in the Z acceleration curve before a period of smaller accelerations; heel-off, which was defined as the first peak in the Z acceleration curve after the period of smaller accelerations; and toe-off, which was defined as the second peak in the Z acceleration curve and corresponded to an inflection point in the X acceleration curve. The segments of the stride were total stance, which consisted of hoof contact and break-over (heel-off to toe-off); initial swing (toe-off to initial 25% of swing); terminal swing (75% of swing to hoof contact); and total swing (toe-off to hoof contact). The orientation of the hoof during the middle of total stance (when the metacarpal bone was perpendicular to the ground) was used to adjust the sagittal orientation of the hoof such that 0° was level with the ground. The cranial-to-caudal (X) variables were positive cranially, the vertical (Z) variables were positive proximally, and the sagittal-plane orientation was positive in a counterclockwise rotation (ie, heel-down was positive). Range of motion was calculated from the difference in the maximum and minimum orientations for each stride segment. Grade 3 lameness could not be induced in 1 horse; therefore, for grade 3 and after-block treatments, the mean (SD) represents data from only 5 horses, whereas the mean (SD) for the other treatments represents data from all 6 horses.
Within a treatment and variable, the value for the lame forelimb differs significantly (P < 0.05) from that for the nonlame forelimb.
Within a forelimb, value differs significantly (P < 0.05) from that at baseline.
Mean (SD) kinematic variables for the lame and nonlame forelimbs during the swing (toe-off to hoof-contact) phase of the stride at a trot for the horses of Table 1.
| Treatment | ||||||
|---|---|---|---|---|---|---|
| Swing segment | Variable | Baseline | Grade 1 | Grade 2 | Grade 3 | After block |
| Initial-swing | Orientation (°) | |||||
| Maximum | ||||||
| Lame | −45.03 (6.35) | −44.75 (5.27)* | −44.24 (3.82)* | −44.30 (5.93)* | −43.83 (4.48)* | |
| Nonlame | −47.18 (7.30) | −49.17 (6.90) | −46.82 (5.84) | −47.16 (4.75) | −48.59 (5.35) | |
| Minimum | ||||||
| Lame | −108.61 (5.71) | −109.95 (4.49)* | −109.96 (5.05) | −109.90 (5.19)* | −109.97 (4.11)* | |
| Nonlame | −109.44 (4.89) | −111.19 (4.83)† | −110.59 (5.93)† | −111.44 (3.80) | −112.05 (5.24)† | |
| Mean | ||||||
| Lame | −92.60 (5.21) | −93.35 (3.51)* | −93.10 (3.85)* | −93.28 (4.65)* | −93.09 (3.45)* | |
| Nonlame | −93.76 (4.94) | −95.71 (4.26)† | −94.48 (4.56) | −95.17 (3.36) | −95.79 (3.89) | |
| Range of motion (°) | ||||||
| Lame | 63.58 (8.40) | 65.20 (7.97)* | 65.71 (7.27)* | 65.60 (7.51) | 66.14 (6.18)* | |
| Nonlame | 62.26 (7.83) | 62.02 (8.35) | 63.77 (8.38) | 64.28 (5.79) | 3.45 (7.51) | |
| Terminal-swing | X velocity (m/s) | |||||
| Maximum | ||||||
| Lame | 6.441 (0.447) | 6.570 (0.543)* | 6.549 (0.311)* | 6.506 (0.301)* | 6.408 (0.375) | |
| Nonlame | 6.541 (0.461) | 6.798 (0.631)† | 6.776 (0.350)† | 6.630 (0.283)† | 6.554 (0.569) | |
| X acceleration (m/s2) | ||||||
| Minimum | ||||||
| Lame | −112.149 (19.935) | −111.768 (18.594)* | −111.204 (18.745)* | −108.434 (21.057)*† | −107.882 (18.833)*† | |
| Nonlame | −115.652 (21.334) | −117.054 (21.282) | −117.807 (18.521) | −115.521 (17.555) | −120.979 (21.790) | |
| Z position (m) | ||||||
| Maximum | ||||||
| Lame | 0.051 (0.011) | 0.048 (0.011) | 0.050 (0.010)* | 0.052 (0.011) | 0.049 (0.010)* | |
| Nonlame | 0.049 (0.007) | 0.048 (0.009) | 0.045 (0.007)† | 0.049 (0.011) | 0.043 (0.008)† | |
| Z velocity (m/s) | ||||||
| Maximum | ||||||
| Lame | 0.361 (0.422) | 0.349 (0.328)* | 0.393 (0.378) | 0.322 (0.333)* | 0.475 (0.349)* | |
| Nonlame | 0.411 (0.336) | 0.452 (0.346) | 0.481 (0.343) | 0.461 (0.339) | 0.540 (0.271) | |
| Total-swing | Duration (s) | |||||
| Lame | 0.384 (0.020) | 0.384 (0.015)* | 0.390 (0.018)* | 0.387 (0.024) | 0.385 (0.016) | |
| Nonlame | 0.382 (0.022) | 0.380 (0.017) | 0.379 (0.017) | 0.373 (0.015) | 0.385 (0.021) | |
| X position (m) | ||||||
| Mean | ||||||
| Lame | 0.968 (0.084) | 0.977 (0.073) | 0.997 (0.104)* | 0.989 (0.098) | 0.972 (0.073) | |
| Nonlame | 0.973 (0.097) | 0.995 (0.092) | 0.966 (0.080) | 0.958 (0.081) | 0.960 (0.073) | |
| Z position (m) | ||||||
| Maximum | ||||||
| Lame | 0.105 (0.022) | 0.105 (0.021) | 0.105 (0.021) | 0.112 (0.023)* | 0.110 (0.018)* | |
| Nonlame | 0.105 (0.021) | 0.102 (0.014) | 0.100 (0.016) | 0.104 (0.022) | 0.089 (0.017) | |
| Z acceleration (m/s2) | ||||||
| Minimum | ||||||
| Lame | −64.874 (20.016) | −64.342 (22.429) | −62.268 (19.104)* | −59.883 (17.749)* | −72.141 (28.065) | |
| Nonlame | −68.435 (19.209) | −68.599 (27.219) | −76.094 (26.389) | −70.488 (16.100) | −64.823 (15.505) | |
See Table 1 for key.
Interlimb kinematic changes—Among the treatments, 34 of 94 (36.2%) kinematic variables varied significantly between the lame and nonlame forelimbs. Of those variables, significant interlimb differences were detected for 14 of 36 cranial-to-caudal (X) variables, 17 of 35 vertical (Z) variables, 2 of 17 sagittal-plane orientation variables, and 1 of 6 temporal variables. Significant interlimb differences were detected during all segments of the stride and for all treatments.
Discussion
Results of the present study indicated that multiple sagittal-plane hoof kinematic variables were significantly altered at a trot following induction of unilateral, weight-bearing forelimb lameness in clinically normal horses. These kinematic alterations were identified during both the stance and swing phases of the stride at even the most mild (grade 1) severity of lameness induced. Because the characterization of kinematic alterations among the different grades of lameness was beyond the scope of this study, multiple comparisons were not performed to determine whether alterations in kinematics were associated with severity of lameness; however, as expected, the number of kinematic variables that varied significantly from baseline (prior to induction of lameness) and between the lame and nonlame limbs increased as the severity of lameness that was induced increased.
During the total stance phase of the stride for horses at a trot, multiple kinematic variables were significantly altered, even at grade 1 lameness. Following induction of grades 1, 2, and 3 lameness, the right (lame) forelimb had a significantly greater caudal (X) acceleration at hoof-contact (ie, beginning of total stance), compared with that at baseline or for the left (nonlame) forelimb, and after administration of the perineural anesthesia, caudal acceleration of the lame limb at hoof contact appeared to return to that at baseline. This change in the cranial-to-caudal acceleration of the lame limb suggested that horses slow the advance of the lame limb to a greater extent than that of the nonlame limb before maximum weight bearing, which occurs during the middle of stance.
Following induction of each grade of lameness in the present study, the hoof orientation at hoof-contact for the nonlame limb had a more positive angle, compared with that at baseline or for the lame limb, which indicated that the hoof of the nonlame limb was landing in a heel-first manner. During the break-over segment, the lame limb was more rapidly unloaded than was the nonlame limb, as evidenced by the increased maximum and mean cranial (X) acceleration and decreased minimum orientation angle and range of motion for the lame limb, compared with those for the nonlame limb. The minimum orientation during break-over represents the hoof orientation at toe-off; the fact that the minimum orientation for the lame limb was less than that for the nonlame limb after induction of each grade of lameness likely contributed to the decreased range of motion for the lame limb, compared with that for the nonlame limb. Because significant interlimb differences were detected for maximum cranial acceleration, minimum orientation, and range of motion during break-over at the mildest (grade 1) lameness induced in the present study, those variables might be sensitive indicators for the diagnosis of subclinical forelimb lameness in horses.
Results of other studies21,22 indicate that the duration of total stance increases in both lame and nonlame limbs as severity of lameness increases. Investigators of another study12 reported that for horses at a trot, the duration of the stance phase was increased for the lame diagonal pair of limbs, compared with that for the nonlame diagonal pair of limbs. Conversely, in the present study, following lameness induction, the duration of the stance phase did not increase from baseline for either the lame or nonlame limb and did not vary significantly between the lame and nonlame limbs, and those findings were consistent with results of a study conducted by Ishihara et al.10
Investigators of another study22 suggest that weight-bearing lameness has only a minimal effect on kinematics during the swing phases of the stride. In the present study, we identified several kinematic variables that were altered between the lame and nonlame limbs during the swing segments. For example, following induction of grade 1 and grade 2 lameness, the duration of total-swing was significantly longer for the lame limb, compared with that for the nonlame limb. Results of other studies12,22 indicate that the duration of swing for a lame limb decreased from baseline only after induction of moderate lameness (ie, mild lameness was detectable at a walk). For the horses of the present study, the most severe lameness induced (grade 3) did not result in perceptible lameness at a walk, which suggested that the lameness induced in this study was not as severe as that induced in those other studies12,22 and may explain why the duration of the swing for the lame limb was not shortened.
Even after the mildest grade of lameness was induced in the present study, several kinematic variables during the swing phase of the stride differed significantly between the lame and nonlame limbs. During the initial-swing segment, the range of motion for the lame limb was significantly greater, compared with that for the nonlame limb. Because the lame limb went through a smaller range of motion during break-over than did the nonlame limb, it is possible that this extra rotation during the swing phase was a compensatory change. During the terminal-swing segment, the nonlame limb had greater maximum cranial (X) and vertical (Z) velocity and decreased minimum cranial acceleration, compared with those for the lame limb. The cranial variable differences between the 2 limbs indicated that the nonlame limb began the terminal-swing segment sooner than did the lame limb; thus, more caudal acceleration was required to slow the hoof for impact for the nonlame limb than for the lame limb. As expected, the total-swing segment was longer for the lame limb, compared with that for the nonlame limb; therefore, the lame limb moved slower and had smaller accelerations through the swing phase than did the nonlame limb. For the forelimb of a horse, maximum vertical velocity typically occurs as the hoof is undergoing a final rotation to prepare for landing.16 For the horses of the present study, the hoof of the nonlame limb generally landed in a flat or heel-first manner and the vertical velocity of the hoof likely attributed to its rotation; however, the hoof of the lame limb generally landed with a toe-down orientation such that less proximal velocity was required to rotate the hoof into its final position for landing.
In the present study, the lame limb had a higher maximum vertical position during the swing phase, compared with that of the nonlame limb, as evidenced by the significant difference between the 2 limbs for this variable at grade 2 lameness during the terminal-swing segment and at grade 3 lameness during the total-swing segment. In another study,11 the vertical position of the nonlame limb during the swing phase was greater than that of the lame limb. The reason the findings of the present study appear to conflict with those of that other study11 was most likely associated with differences in how the position of the hoof was determined between the 2 studies. In the present study, the position of the hoof was determined by multiple markers and increased rotation of the hoof exaggerated the extent of vertical movement. Because the hoof of the lame limb had a more toe-down orientation than did the hoof of the nonlame limb, the lame limb would appear to have a greater maximum vertical position, compared with that of the nonlame limb.
Following administration of perineural anesthesia in the lame limb, only some of the kinematic variables returned to baseline values, whereas other kinematic variables remained significantly altered from baseline. For both the lame and nonlame limbs, the orientation of the hoof at hoof-contact and maximum cranial velocity during the terminal-swing segment did not vary significantly from baseline after perineural anesthesia. However, for the lame limb, cranial acceleration at hoof-contact, range of motion during the break-over segment, and cranial acceleration during the terminal-swing segment varied significantly from baseline after perineural anesthesia; for the nonlame limb, cranial acceleration at hoof-contact, mean cranial acceleration during the break-over segment, minimum orientation during the initial-swing segment, and vertical position during the terminal-swing segment varied significantly from baseline after perineural anesthesia. Perineural anesthesia alters the neural pathways in the limb; therefore, it was not totally unexpected that some of the kinematic variables for the lame and nonlame limbs did not return to values similar to those at baseline.
Some kinematic variables evaluated in the present study varied significantly between the forelimbs before induction of lameness, and because of this inherent asymmetry between the limbs, interlimb comparisons were not performed for those variables after the other treatments (grade 1, grade 2, grade 3, and after block). Interlimb asymmetry at baseline might have been caused by lameness that could not be detected by the subjective method used in this study. Results of other studies indicate that the use of a stationary force platform10 or an inertial sensor system5 had greater sensitivity than did subjective lameness examination for identification of horses with mild lameness. It is possible that the optical kinematics system used in the present study may similarly be more sensitive than subjective lameness examination for the detection of subclinically lame horses; however, evaluation of a larger number of horses is necessary to verify this. Another possible explanation for the forelimb asymmetry observed at baseline for some of the kinematic variables was the laterality, or handedness, of the horses evaluated. Many horses preferably use one forelimb more than the other during both grazing and ambulation, and this laterality begins at a young age.23,24 Additionally, investigators of another study25 reported that forelimb asymmetries exist in clinically normal horses during various phases of the stride and are likely associated with differences in hoof conformation between the 2 limbs. Within a horse, conformational asymmetry between the hooves of the forelimbs could contribute to significant interlimb differences in the kinematics of the distal portion of the limb. Measurement of hoof conformation during the various phases of the stride was beyond the scope of the present study, and we chose to make interlimb comparisons after lameness induction and perineural anesthesia only for those kinematic variables that did not differ significantly at baseline. Nevertheless, the fact that forelimb asymmetry was identified at baseline for some of the kinematic variables evaluated in the present study suggested that both intra- and interlimb kinematic comparisons should be performed for detection of lameness in horses.
The sole-pressure model used in the present study induced lameness consistently, and that lameness was rapidly reversible. Although most horses that are clinically lame do not have lameness caused by pressure to the sole, the kinematic changes associated with lameness induced by the sole-pressure model are believed to be similar to those associated with lameness induced by other causes.11 Consequently, the sole-pressure model has been accepted and used in many studies5,12–14,20 that assessed both kinetic and kinematic methods for the detection of lameness in horses. In 1 of the first 2 horses evaluated in the present study, we were unable to induce grade 3 lameness with blunt-ended screws; thus, we modified the sole-pressure model slightly and used screws with tapered ends instead of blunt ends so that all 3 grades of lameness could be consistently induced in the remaining 4 study horses. Because the primary objective of this study was to identify kinematic variables that were useful for the diagnosis of mild lameness in horses, we do not feel that the use of data from only 5 instead of 6 horses for the analyses for the grade 3 and after-block treatments was a major limitation.
The horizontal velocity of a horse affects the kinematics of the distal aspect of the limbs26; therefore, in the present study, we ensured that each horse traveled over the runway with a consistent velocity during each trial. During the collection of baseline data, each horse was allowed to trot at a comfortable velocity, the mean velocity was calculated for the baseline trials, and then a range of ± 10% of that mean velocity was calculated and used to delimit the range within which that horse's velocity had to be for all subsequent trials to be considered acceptable. This process mimics that of subjective clinical lameness examinations in which a horse is routinely examined at a speed dictated by that individual horse. Results of another study27 indicate that the most reproducible kinematic measurements are obtained when a horse is allowed to move at its own individual optimum speed because at that speed, the variation in motion is minimized. In a study28 in which the vertical head excursion of lame horses was evaluated, horses with mild lameness had no increase in the asymmetry of head motion when horizontal speed was increased. Thus, lameness in a horse can be adequately detected when it is moving at its own individual optimum speed, rather than at a predetermined speed that is used for the evaluation of all horses.
Investigators of other studies12,13 that evaluated the kinematics of the stride of horses divided the stride into only 2 phases, stance and swing. In the present study, we further subdivided the stance and swing phases on the basis of specific hoof events that could be identified on the X and Z acceleration curves. To our knowledge, this is the first study in which the stance phase of the stride was divided into hoof-contact and break-over segments and the swing phase was divided into initial-swing, terminal-swing, and total-swing segments. As expected, subtle kinematic alterations occurred during each segment of the stance and swing phases (eg, increase in maximum vertical velocity of the nonlame limb during terminal-swing and changes in hoof orientation during hoof-contact, break-over, and initial-swing) that would have been missed had those 2 phases not been subdivided into their respective segments.
In addition to the various segments of the stride, linear and angular kinematic data were also evaluated for hoof events (hoof-contact, heel-off, and toe-off). The kinematic variables determined at heel-off and toe-off were reflected in the data for the break-over and initial-swing segments; however, the kinematic variables determined at hoof-contact were not reflected well in the data for the terminal-swing segment. This was especially true for sagittal plane orientation of the hoof because the maximum sagittal plane orientation during the terminal-swing segment occurred immediately before hoof-contact when the hoof underwent a final counterclockwise rotation. Consequently, we decided to report the instantaneous kinetic data at hoof-contact in addition to the kinematic data for each segment of the stride.
Results of the present study indicated that even mild, weight-bearing lameness in a forelimb of a horse can result in altered kinematics for the distal portion of both the lame and nonlame limbs when the horse is trotting. Further study is necessary to determine whether these kinematic changes also occur in horses that are walking. For several kinematic variables following induction of lameness, a significant difference between the 2 forelimbs was often identified, whereas within the lame limb, a significant difference from baseline was not detected; therefore, data should be obtained from both forelimbs during kinematic analyses. Following perineural anesthesia, kinematic values for hoof orientation at hoof-contact and maximum cranial acceleration during the break-over segment did not differ from those at baseline; therefore, those values could be useful for the objective assessment of the effect of the perineural anesthesia. Additional studies with a larger number of horses with clinical lameness are necessary to determine whether the use of a hoof-mounted kinematic monitoring system is clinically useful for lameness diagnosis and monitoring in horses.
ABBREVIATION
| IMU | Inertial measurement unit |
Equi-Thane SuperFast, Vettec Inc, Oxnard, Calif.
H3-IMU, MemSense LLC, Rapid City, SD.
Vetrap, 3M Co, Saint Paul, Minn.
Volant, Peak Performance Technologies Inc, Centennial, Colo.
Vicon-Motus, version 9.2, Vicon Motion Systems Inc, Centennial, Colo.
MEK 92-PAD photoelectric control, Mekontrol Inc, Northboro, Mass.
STATA, version 11, Stata Corp LP, College Station, Tex.
References
1. Kaneene JB, Ross WA, Miller RA. The Michigan equine monitoring system. II. Frequencies and impact of selected health problems. Prev Vet Med 1997; 29:277–292.
2. USDA. National economic cost of equine lameness, colic, and equine protozoal myeloencephalitis in the United States. Information sheet. Fort Collins, Colo: USDA, APHIS, Veterinary Service, National Health Monitoring System, 2001.
3. Gaughan EM. Skeletal origins of exercise intolerance in horses. Vet Clin North Am Equine Pract 1996; 12:517–535.
4. Parente EJ, Russau AL, Birks EK. Effects of mild forelimb lameness on exercise performance. Equine Vet J Suppl 2002;(34):34–252.
5. McCracken MJ, Krammer J, Keegan KG, et al. Comparison of an inertial sensor system of lameness quantification with subjective lameness evaluation. Equine Vet J 2012; 44:652–656.
6. Arkell M, Archer RM, Guitian FJ, et al. Evidence of bias affecting the interpretation of the results of local anesthetic nerve blocks when assessing lameness in horses. Vet Rec 2006; 159:346–349.
7. Fuller CJ, Bladon BM, Driver AJ, et al. The intra- and inter-assessor reliability of measurement of functional outcome by lameness scoring in horses. Vet J 2006; 171:281–286.
8. Hewetson M, Christley RM, Hunt ID, et al. Investigations of the reliability of observational gait analysis for the assessment of lameness in horses. Vet Rec 2006; 158:852–858.
9. Keegan KG, Dent EV, Wilson DA, et al. Repeatability of subjective evaluation of lameness in horses. Equine Vet J 2010; 42:92–97.
10. Ishihara A, Bertone AL, Rajala-Schultz PJ. Association between subjective lameness grade and kinetic gait parameters in horses with experimentally induced forelimb lameness. Am J Vet Res 2005; 66:1805–1815.
11. Buchner HHF, Savelberg HH, Schamhardt HC, et al. Limb movement adaptations in horses with experimentally induced fore-or hindlimb lameness. Equine Vet J 1996; 28:63–70.
12. Galisteo AM, Cano MR, Morales JL, et al. Kinematics in horses at the trot before and after an induced forelimb supporting lameness. Equine Vet J Suppl 1997;(23):23–97.
13. Keegan KG, Wilson DA, Smith BK, et al. Changes in kinematic variables observed during pressure-induced forelimb lameness in adult horses trotting on a treadmill. Am J Vet Res 2000; 61:612–619.
14. Keegan KG, Yonezawa Y, Pai PF, et al. Evaluation of a sensor-based system of motion analysis for detection and quantification of forelimb and hind limb lameness in horses. Am J Vet Res 2004; 65:665–670.
15. Pfau T, Robilliard JJ, Weller R, et al. Assessment of mild hindlimb lameness during over ground locomotion using linear discriminant analysis of inertial sensor data. Equine Vet J 2007; 39:407–413.
16. Moorman VJ, Reiser RF II, McIlwraith CW, et al. Validation of an equine inertial measurement unit system in clinically normal horses during walking and trotting. Am J Vet Res 2012; 73:1160–1170.
17. Keegan KG, Kramer J, Yonezawa Y, et al. Assessment of repeatability of a wireless, inertial sensor-based lameness evaluation system for horses. Am J Vet Res 2011; 72:1156–1163.
18. Moorman VJ, Reiser RF, Peterson ML, et al. Effect of forelimb lameness on hoof kinematics of horses at a walk. Am J Vet Res 2013; 74:1192–1197.
19. American Association of Equine Practitioners. Guide for veterinary service and judging of equestrian events. 4th ed. Lexington, Ky: American Association of Equine Practitioners, 1991;19.
20. Merkens HW, Schamhardt HC. Evaluation of equine locomotion during different degrees of experimentally induced lameness. I: lameness model and quantification of ground reaction patterns of the limbs. Equine Vet J Suppl 1988;(6):6–99.
21. Weishaupt MA, Wiestner T, Hogg HP, et al. Compensatory load redistribution of horses with induced weight-bearing forelimb lameness trotting on a treadmill. Vet J 2006; 171:135–146.
22. Buchner HHF, Savelberg HHCM, Schamhardt HC, et al. Temporal stride parameters in horses with experimentally induced fore- or hindlimb lameness. Equine Vet J Suppl 1995;(18):18–161.
23. Murphy J, Sutherland A, Arkins S. Idiosyncratic motor laterality in the horse. Appl Anim Behav Sci 2005; 91:297–310.
24. van Heel MCV, Kroekenstoel AM, van Dierendonck MC, et al. Uneven feet in a foal may develop as a consequence of lateral grazing behavior induced by conformational traits. Equine Vet J 2006; 38:646–651.
25. Wilson GH, McDonald K, O'Connell MJ. Skeletal forelimb measurements and hoof spread in relation to asymmetry in the bilateral forelimb of horses. Equine Vet J 2009; 41:238–241.
26. Clayton HM. Comparison of the stride kinematics of the collected, working, medium and extended trot in horses. Equine Vet J 1994; 26:230–234.
27. Peham C, Licka T, Mayr A, et al. Speed dependency of motion pattern consistency. J Biomech 1998; 31:769–772.
28. Peham C, Licka T, Mayr A, et al. Individual speed dependency of forelimb lameness in trotting horses. Vet J 2000; 160:135–138.
