Postural stability is achieved through precise sensory or neural integration of proprioceptive, vestibular, and visual information.1 Effective postural control also relies on motor components of the musculoskeletal system to maintain balance, which include joint stability and neuromuscular control.2 Muscles are responsible not only for inducing joint motion, but also for stabilizing joints during limb loading and the maintenance of stance,3 thereby providing crucial contributions to balance control. Diminished sensory function and altered sensorimotor processing from articular structures substantially compromise the ability to maintain postural stability.4 Joint injury causes damaged and altered articular tissues, which ultimately impairs both the sensory and motor components required for postural stability.4 Therefore, slowing or preventing the progression of osteoarthritis is of particular interest because it affects the integrity of articular cartilage, exacerbates synovial membrane inflammation, and induces joint capsule fibrosis and periarticular muscle weakness.5,6 These disease processes directly alter the neurophysiologic properties of joint mechanoreceptors that aid in the control of tonic muscular activity responsible for joint stabilization during quiet standing.4
Joint mechanoreceptors provide continuous sensory input to the CNS regarding mechanical conditions in and around articulations. Joint mechanoreceptors are responsible for signaling joint position and movement, aid in controlling both the timing and direction of joint movement, contribute to the initiation of reflexive muscular responses that maintain joint stability, and have a primary role in joint nociception.6 Mechanoreceptors function as both proprioceptors and modifiers of muscle activity, thereby monitoring joint stability and protecting articular structures from abnormal or excessive loading.7 Under normal circumstances, stretching of the joint capsule and surrounding ligaments causes increased firing of the joint mechanoreceptors; the mechanoreceptors synapse onto gamma motor neurons within the ventral horn of the spinal cord and induce fine adjustments in muscle tension that influence joint stability and balance.7 Musculoskeletal pain induces reflex mechanisms mediated by joint mechanoreceptors, and an injured joint is protected from further damage either by inhibition of muscles or the stimulation of muscle guarding.8 Persistent joint pain and inflammation cause afferent excitatory stimulation of 1b interneurons within the ventral horn of the spinal cord, which induces inhibition of the muscles that act across the affected joint.9,10 Altered afferent articular information also decreases gamma motor neuron activation, which in turn alters muscle spindle sensitivity and ultimately impairs proprioception.11 Several studies9,10,12,13 in humans have revealed inhibitory muscle responses as a result of joint effusion, pain, and osteoarthritis. Increased knowledge of the neural and patterned muscle responses to joint effusion and osteoarthritis provides a deeper understanding of compensatory mechanisms involved in the maintenance of postural control and the prevention of additional musculoskeletal injuries.
Finely tuned adjustments in muscle tension are required to maintain positioning of the body's COM during an upright stance. While humans stand quietly on a force platform, these slight adjustments in muscle tension induce movement that can be objectively measured as movement of the COM in both mediolateral and craniocaudal directions.14 Force platforms are used to record changes in the COP, which is defined as the point of application of the ground reaction forces under the feet.15,16 The trajectory (ie, amplitude and pattern) of the COP provides insights into the process of balance control because changes in the COP are directly related to motion of an individual's COM.15 Postural sway is defined from measures of COP movement in craniocaudal or mediolateral directions. Excessive postural sway in a mediolateral direction is often used to predict whether humans are at risk for falling.17 Individuals with proprioceptive deficits will have a large COP area because of altered signaling pathways that diminish the ability of neuromuscular responses to maintain postural stabilization during stance.18 In humans, osteoarthritis can induce subclinical musculoskeletal dysfunction, which can be quantitatively measured by means of postural sway analysis during quiet standing.2,6,19,20 Humans with osteoarthritis of the knee and hip joints have significantly greater postural sway disturbances, compared with findings for age-matched healthy individuals, which indicates that osteoarthritis negatively affects balance and postural stabilty.19,20 In addition, electromyography of the quadriceps muscle group in humans with knee joint osteoarthritis often reveals a decrease in muscle strength and delayed activation.2 Abnormal timing or amplitudes of quadriceps muscle group activation can cause balance deficits and a reduction in the ability to maintain, achieve, or restore postural stability.6 Increased COP displacement in either craniocaudal or mediolateral directions helps to identify altered neuromuscular response mechanisms responsible for maintaining stance in humans during postural sway analysis.
Postural sway analysis in humans is often assessed in both a double- and single-leg stance position17 and with subjects tested with their eyes open and closed.21 A single-leg stance position requires substantially more motor control and strength than does a double-leg stance position. Assessments performed with a subject's eyes closed are often undertaken to assess the contribution of visual pathways to postural stability. Changes in stance position and the removal of visual cues provide valuable information regarding the severity of balance deficits present in humans with osteoarthritis and are effective in assessing the degree of improvement in postural sway variables following exercise interventions.19,21–24
The subjective assessment of a horse's response to balance-related manipulative tests (ie, when horses are placed in a crossed-limb stance or are blindfolded) are routinely incorporated into clinical neurologic examinations. However, the relationship between changes in stance conditions and the associated balance deficits in horses remains inconclusive. Recent research has demonstrated that force platform analysis provides a reliable assessment of postural stability in horses.25,26 The quantification of balance control by means of postural sway analysis can be a valuable tool for the diagnosis of subclinical neuromuscular and musculoskeletal conditions and in the monitoring of the response to rehabilitation. Horses with abnormal sensory or afferent signaling attributable to joint pain, inflammation, or altered muscle activation are likely to have impaired motor control. Therefore, identifying alterations in static postural stability and the effects of impaired motor control on balance required during dynamic locomotor tasks (eg, half pass, leg yield, piaffe, and pirouette) will aid in the development of targeted diagnostic and treatment strategies.
Physical rehabilitation has become an effective treatment option for primary musculoskeletal injuries and for reducing or limiting harmful compensatory gait abnormalities in humans.6 Rehabilitation programs that address osteoarthritis and musculoskeletal injuries often incorporate some form of aquatic exercise. Exercising in water provides an effective medium for increasing joint mobility, increasing muscle activation, promoting normal motor patterns, and reducing the incidence of secondary musculoskeletal injuries.24 Humans with osteoarthritis of the joints of the lower extremities have a significant reduction in the amplitudes of postural sway following aquatic exercise.17 The improved muscle strength and function associated with aquatic exercise significantly improve proprioception and motor control and reduce the abnormal postural sway characteristics typically reported in osteoarthritic adults.19 The application of water exercise may enhance motor learning through multisensory impulses, thereby improving posture and stability.
Underwater treadmill exercise is widely used in the rehabilitation of horses with osteoarthritis and other musculoskeletal injuries. To our knowledge, no studies to date have specifically assessed the efficacy of underwater treadmill exercise with regard to horses. Therefore, the overall goal of the study reported here was to evaluate the effects of underwater treadmill exercise on static postural sway in horses with experimentally induced carpal joint osteoarthritis under various stance conditions. Characterization of altered postural stability associated with joint pain and osteoarthritis is critical for developing targeted, clinically relevant diagnostic and treatment strategies to minimize the progression of joint degeneration and compensatory changes that may lead to additional musculoskeletal injuries. Therefore, the study was designed to assess static postural sway of the horses under conditions of normal square stance, base-narrow stance (ie, base-narrow placement of the thoracic limbs), and normal square stance with removal of visual cues (ie, blindfolded) and to evaluate the effects of underwater treadmill exercise on static postural sway. Our hypotheses were that the base-narrow and blindfolded stance conditions would diminish postural stability and that horses with experimentally induced carpal joint osteoarthritis exposed to underwater treadmill exercise would have improved postural sway characteristics, compared with horses with experimentally induced carpal joint osteoarthritis that underwent overground treadmill exercise, which simulates controlled hand walking.
Materials and Methods
Project design—The horses of this study were not permanently disabled by experimental induction of carpal joint osteoarthritis, considering that they were euthanized at the conclusion of the study as part of a larger project investigating the gross morphological, histologic, and biochemical outcomes associated with the progression of osteoarthritis and quantifying the disease-modifying effects of underwater treadmill exercise. Nonlame horses were not used in the study because the displacement in the COP measurements used to determine the effects of underwater treadmill exercise was not expected to be of sufficient magnitude to identify any changes in a clinically normal horse population.
Horses—Sixteen skeletally mature horses ranging in age from 2 to 4 years (mean ± SD body weight, 385 ± 40 kg; mean height measured from the ground to the most dorsal aspect of the shoulders [ie, withers], 1.46 ± 0.04 m) were included in the study. Prior to inclusion in the study, all horses underwent body condition assessment, a lameness examination, range of motion (flexion) testing of the carpal joints, evaluation for the presence of middle carpal joint effusion, and radiography of the carpal joints; only horses without abnormal findings were permitted into the study. A minimum 14-day acclimatization period was provided prior to baseline data collection to introduce new environmental factors and to train the horses to safely enter and use the overground and underwater treadmills. The Colorado State University Institutional Animal Care and Use Committee approved the study protocol.
Experimental induction of osteoarthritis—Each horse was anesthetized and routinely prepared for surgery. Each horse underwent bilateral arthroscopic surgery of the middle carpal joints (day 0) to ensure that there were no preexisting articular lesions. During this procedure, an osteochondral fragment was created in 1 randomly selected middle carpal joint of each horse. An 8-mm-wide fragment was generated with an 8-mm curved osteotome directed perpendicular to the articular cartilage surface of the radial carpal bone at the level of the medial synovial plica.27 The dorsal surface of the osteochondral fragment was allowed to remain adhered to the joint capsule proximally. A motorized burr was used to debride the exposed subchondral bone between the fragment and the parent bone to create a 15-mm-wide defect bed. The debris was not actively lavaged from the joint. Diagnostic arthroscopy was performed on the contralateral middle carpal joint to ensure absence of any joint disease. The arthroscopic portals were closed with 2–0 nylon suture in a simple interrupted pattern, and cyanoacrylate gluea was applied to the skin incision. The thoracic limbs were bandaged, and the horse was allowed to recover from anesthesia and surgery. Each horse received a single dose of ceftiofurb (4.0 mg/kg, IM) just prior to surgery and 2 g of phenylbutazonec orally once a day for 5 days after surgery. Bandages were changed every 3 to 5 days and maintained until suture removal was performed 10 days after surgery. Animal-care personnel assessed horses daily for apparent comfort, movement, and respiratory character throughout the study. At the end of the study, all horses were euthanized by IV administration of pentobarbital sodium.
Exercise—Horses were individually housed in 3.7 × 3.7-m stalls. Beginning on day 15, all horses were exercised on a high-speed overground treadmilld 5 days each week until the end of the study to promote the development of osteoarthritis. Each day, the horses were trotted at 4.4 m/s for 2 minutes, galloped at 8.8 m/s for 2 minutes, and trotted again at 4.4 m/s as part of a standard exercise protocol used to aid in the induction and progression of carpal joint osteoarthritis.
Treatment groups—One week after surgery (day 7), the horses were ranked by lameness scores and randomly assigned to 1 of 2 groups,28 with 8 horses in each group. The aquatic exercise group was exposed to underwater treadmill exercisee starting on day 15. The underwater treadmill protocol consisted of exercise at a brisk walk (2.1 m/s) for 5 minutes, once a day, for 5 days. The water level in the underwater treadmill was maintained at the point of the shoulder for the duration of the study, which reduced weight bearing by approximately 60%.29 At weekly intervals, the duration of underwater treadmill exercise was increased by 5 minutes, until a maximum of 20 min/session was reached. Horses then continued to be exercised for 20 minutes once a day, 5 days a week, for the remainder of the 10-week study (ie, 8 weeks of exercise). The control group was exercised on an overground treadmill (without water) in a manner similar to that of the underwater treadmill protocol for the intervention group (ie, same speed, frequency, and duration of exercise), also starting at day 15. The overground treadmill exercise planned for the control group was designed to simulate conventional hand walking and light exercise used for rehabilitation after arthroscopic surgery.
Kinetic data collection—Displacement of the COM of each horse's body was determined from COP measurements recorded with 2 strain-gauge-based force platformsf (60 × 90 cm) mounted in series in a concrete base at the center of a 25-m runway. The COP accuracy for each force platform was ± 2 mm, and the static resolution for the system was ± 1 N. Postural sway data were collected from each horse at 4 time points during the study: prior to induction of the osteochondral fragment (day −7), after induction of the osteochondral fragment and 1 day prior to initiating treadmill exercise (day 14), 4 weeks after initiating treadmill exercise (day 42), and at study conclusion (day 70).
Stance conditions—Postural control was assessed in 3 stance conditions: normal square stance with vertical placement and even weight distribution on all 4 limbs, base-narrow placement of the thoracic limbs, and normal square stance with removal of visual cues (ie, blindfolding). Each horse was positioned with their thoracic limb hooves squarely placed on a force platform and their pelvic limb hooves placed squarely on the second force platform. The horse handler was instructed to stand in front of the horse and to hold the lead shank loosely with no direct contact while an assistant stood on either side of the horse near the pelvic limbs to encourage a static stance. The placement of each hoof was marked with chalk on the surface of the force platform, and the distance between each outlined hoof and the lateral edges of the force platform was recorded to provide consistent hoof placement across all trials (ie, the distance between the lateral aspect of the hoof wall and the lateral edge of the force platform perpendicular to the hoof and the distance from the toe of the hoof to the cranial edge of the force platform were measured). Each horse was initially positioned on the force platforms in a normal stance position with the head and neck maintained in a relaxed, neutral position. For the second stance condition, the thoracic limbs were placed with the medial aspect of each hoof in contact with each other in a base-narrow stance position and the toes at equal distance from the cranial aspect of the force platform. The third stance condition involved returning the thoracic limbs to a normal square stance position and placement of a blinder hood over the horse's head that completely blocked vision in both eyes (Figure 1). For each stance condition, each horse was required to stand quietly for 3 consecutive, 10-second trials. The mean of the 3 trial results in each condition was calculated and used for analysis.
Representative photograph of a horse standing in a balanced normal square stance position on 2 serial force platforms with a blinder hood applied to remove visual cues during postural sway analysis. Retroreflective markers were placed over the thoracic limb centers of joint rotation, hoof, and coronary band as well as the transverse process of C1, midline of the forehead, and the right side of the cervical region for kinematic analysis. Electromyography surface electrodes were placed over the thoracic limb muscles of interest and cables attached to a telemetry unit adhered to the surcingle worn by the horse. The kinematic and electromyography data were collected as part of a larger study
Citation: American Journal of Veterinary Research 74, 7; 10.2460/ajvr.74.7.971
Kinematic data collection—Kinematic data were used to determine the 2-D position of hoof placement on the force platforms. For each horse, cyanoacrylate glue was used to attach 2.5-cm, round, reflective markers to the mid-dorsum of each hoof (3 cm distal to the coronary band) and on the midline of the forehead between the supraorbital foramina. Trials with head and neck displacements > 20 cm in any direction were excluded because movements of the head and neck are known to cause substantial changes in COP measurements. Kinematic data were collected at 200 Hz with a motion analysis systemg and 8 high-speed infrared camerash distributed equally around the periphery of the force platforms. The capture volume over the region of the force platforms was calibrated with a customized calibration frame and wand with an accuracy of 0.71 mm.
Data processing—The COP data were collected at a sampling rate of 3,000 Hz and low-pass filtered with a fourth-order, recursive Butterworth filter at 15 Hz to remove noise. The COPs in both the craniocaudal and mediolateral directions were determined initially from each force platform and later combined to calculate a net COP in craniocaudal (y) and mediolateral (x) directions from the following equation16:
where COPi is the COP recorded in either the craniocaudal (i = y) or mediolateral (i = x) direction measured independently from force platform 1 or 2, and Fz plate 1 and Fz plate 2 are vertical ground reaction forces recorded from the respective force platforms. Stabilograms were created to illustrate the movement of the net COP in craniocaudal (y) and mediolateral (x) directions for the 3 stance conditions.
Postural sway variables—Postural sway variables were calculated from the net COP for each stance condition.26 Craniocaudal sway (mm) was calculated as the range of COP displacements in a craniocaudal direction (COPy) relative to the horse, where positive values indicate cranial movements. Mediolateral sway (mm) was calculated as the range of COP displacements in a mediolateral direction (COPx), where positive values indicate movements toward the right. The COP area (mm2) was calculated as the extreme bounds of the COP movement in both craniocaudal and mediolateral directions. The COP velocity (mm/s) is defined as the direction and amplitude of COM displacement immediately following time instant. The average velocity was calculated by summing the resultant distance between 2 consecutive COP data points divided by time.26 The COP radius (mm) was calculated as the mean displacement of each COP data point (xi, yi) relative to the geometric center (ie, mean location of the COP = x0, y0) of each stance condition.
Kinematic stance position variables—Because the hooves did not move during each trial, the 3-D transformed orthogonal coordinates of the reflective markers placed on the dorsal aspect of the hoof wall were used to determine the hoof positions at time 0.26 Craniocaudal length (m) was calculated as the distance between the midpoint of the 2 thoracic limb hoof markers and the midpoint of the 2 pelvic limb hoof markers. Mediolateral width (m) was calculated as the distance between the calculated midpoint of the left thoracic and left pelvic limb hoof markers and the calculated midpoint of the right thoracic and right pelvic limb hoof markers. Base of support (m2) was calculated as the total area contained within the 4 hoof marker locations. Head height (m) was calculated as the mean vertical distance between the reflective markers located on the forehead and the left thoracic limb hoof.
Normalization procedure—Prior to statistical analysis, the postural sway variables (Table 1) were normalized to the corresponding kinematic measurements26 and reported as a normalized percentage value (eg, postural sway variable divided by stance position variable times 100%).26 Normalization was used to remove the dependence of the postural sway variables on each individual horse's morphometric features. The craniocaudal sway was normalized to the craniocaudal length (Table 2).26 Similarly, mediolateral sway was normalized to the mediolateral width, COP area was normalized to the base of support area, and head height was normalized to the height at the withers for each horse.26
Mean ± SEM postural sway and kinematic values (prior to normalization) obtained from horses when they were in a normal square stance position (ie, with vertical placement and even weight distribution on all 4 limbs) before (day −7) and after surgical creation of an osteochondral fragment in 1 randomly selected carpal joint (day 0) and before (day 14) and after (days 42 and 70) subsequent commencement (day 15) of underwater treadmill exercise (UWT; n = 8) or overground treadmill exercise (control; 8).
| Variable | Treatment | Day −7 | Day 14 | Day 42 | Day 70 |
|---|---|---|---|---|---|
| COP | |||||
| Craniocaudal sway (mm) | Control | 11.5 ± 2.1 | 15.4 ± 2.1 | 17.2 ± 2.1 | 16.7 ± 2.1 |
| UWT | 13.8 ± 2.1 | 14.0 ± 2.1 | 12.7 ± 2.1 | 11.9 ± 2.1 | |
| Mediolateral sway (mm) | Control | 11.2 ± 1.9 | 16.0 ± 1.9 | 16.1 ± 1.9 | 16.5 ± 1.9 |
| UWT | 11.9 ± 1.9 | 10.5 ± 1.9 | 9.6 ± 1.9 | 8.9 ± 1.9 | |
| COP area (mm2) | Control | 141.0 ± 70.0 | 252.4 ± 70.0 | 301.1 ± 70.0 | 321.1 ± 70.0 |
| UWT | 191.7 ± 70.0 | 156.5 ± 70.0 | 124.8 ± 70.0 | 126.4 ± 70.0 | |
| COP velocity (mm/s) | Control | 12.2 ± 1.7 | 15.6 ± 1.7 | 14.5 ± 1.7 | 15.7 ± 1.7 |
| UWT | 13.8 ± 1.7 | 14.0 ± 1.7 | 11.4 ± 1.7 | 12.0 ± 1.7 | |
| COP radius (mm) | Control | 4.0 ± 0.6 | 5.1 ± 0.6 | 5.3 ± 0.6 | 6.1 ± 0.7 |
| UWT | 4.4 ± 0.6 | 3.9 ± 0.6 | 3.7 ± 0.6 | 3.4 ± 0.6 | |
| Kinematic | |||||
| Craniocaudal length (m) | Control | 1.09 ± 0.03 | 1.05 ± 0.03 | 1.07 ± 0.03 | 1.06 ± 0.03 |
| UWT | 1.03 ± 0.03 | 1.02 ± 0.03 | 1.09 ± 0.03 | 1.02 ± 0.03 | |
| Mediolateral width (m) | Control | 0.28 ± 0.02 | 0.26 ± 0.02 | 0.25 ± 0.02 | 0.24 ± 0.02 |
| UWT | 0.25 ± 0.02 | 0.25 ± 0.02 | 0.26 ± 0.02 | 0.24 ± 0.02 | |
| Base of support (m2) | Control | 0.30 ± 0.02 | 0.28 ± 0.02 | 0.27 ± 0.02 | 0.25 ± 0.02 |
| UWT | 0.26 ± 0.02 | 0.25 ± 0.02 | 0.28 ± 0.02 | 0.24 ± 0.02 | |
| Head height (m) | Control | 1.37 ± 0.05 | 1.35 ± 0.05 | 1.34 ± 0.05 | 1.30 ± 0.05 |
| UWT | 1.39 ± 0.05 | 1.38 ± 0.05 | 1.33 ± 0.05 | 1.30 ± 0.05 | |
Craniocaudal sway was calculated as the range of COP displacements in a craniocaudal direction (COPy) relative to the horse, where positive values indicate cranial movements. Mediolateral sway was calculated as the range of COP displacements in a mediolateral direction (COPx), where positive values indicate movements toward the right. The COP area was calculated as the extreme bounds of the COP movement in both craniocaudal and mediolateral directions. The average velocity was calculated by summing the resultant distance between 2 consecutive COP data points and dividing by time. The COP radius was calculated as the mean displacement of each COP data point (xi, yi) relative to the geometric center (ie, mean location of the COP = x0, y0) of each stance condition. Craniocaudal length was calculated as the distance between the midpoint of 2 thoracic limb hoof markers and the midpoint of 2 pelvic limb hoof markers. Mediolateral width was calculated as the distance between the calculated midpoint of the left thoracic and left pelvic limb hoof markers and the calculated midpoint of the right thoracic and right pelvic limb hoof markers. Base of support was calculated as the total area contained within the 4 hoof marker locations. Head height was calculated as the mean vertical distance between the reflective markers located on the forehead and the left thoracic limb hoof.
Mean ± SEM normalized postural sway variables obtained from the horses in Table 1 when they were in a normal square stance position before (day −7) and after surgical creation of an osteochondral fragment in 1 randomly selected carpal joint (day 0) and before (day 14) and after (days 42 and 70) subsequent commencement (day 15) of underwater treadmill exercise (UWT; n = 8) or overground treadmill exercise (control; 8).
| Normalized COP variables (%) | Treatment | Day −7 | Day 14 | Day 42 | Day 70 |
|---|---|---|---|---|---|
| Craniocaudal sway | Control | 1.6 ± 0.2 | 1.5 ± 0.2 | 1.6 ± 0.2 | 1.6 ± 0.2 |
| UWT | 1.4 ± 0.2 | 1.4 ± 0.2 | 1.2 ± 0.2 | 1.2 ± 0.2 | |
| Mediolateral sway | Control | 4.1 ± 0.9 | 6.3 ± 0.9 | 6.3 ± 0.9 | 6.8 ± 0.9 |
| UWT | 4.9 ± 0.9 | 4.3 ± 0.9 | 3.8 ± 0.9 | 3.7 ± 0.9 | |
| COP area | Control | 45.4 ± 28.4 | 90.0 ± 28.4 | 110.1 ± 28.4 | 127.7 ± 28.4 |
| UWT | 77.6 ± 28.4 | 63.6 ± 28.4 | 46.7 ± 28.4 | 52.7 ± 28.4 | |
| Head height | Control | 94.0 ± 3.7 | 92.5 ± 3.7 | 92.0 ± 3.7 | 89.2 ± 3.7 |
| UWT | 91.6 ± 3.7 | 91.0 ± 3.7 | 87.0 ± 3.7 | 85.3 ± 3.7 |
Assessment of thoracic limb loading patterns—The combined COPs for the thoracic limbs in both the craniocaudal and mediolateral directions were measured from a single force platform (plate 2). The direction of COP movement relative to thoracic limb hoof placement was determined with customized programming. The direction of COP movement toward or away from the limb with the surgically created osteochondral fragment in the carpal joint was determined for each of the postural sway variables at the 4 time points during all stance conditions.
Clinical outcomes—For each horse, clinical examinations of both thoracic limbs were performed weekly from day −7 (baseline; before surgery) throughout the study period. Lameness was graded by an experienced equine orthopedic surgeon, who was unaware of the treatment group assignments, using a standardized American Association of Equine Practitioners lameness scale30 (scale of 0 to 5, where 0 represented normal gait and 5 represented severe lameness) with the horse trotting on hard ground.
Data analysis—The COP data measured sequentially at different time points over the course of the study were evaluated with a commercially available statistical software programi with an ANOVA or ANCOVA framework depending on the presence or absence of a covariate, respectively. An ANOVA was also used to determine the influence of stance positions on the dependent postural sway variables. The ANOVA design involved assessing the COP variables obtained at days −7 and 14 (pretreatment analysis) to demonstrate that no treatment group effect was detectable prior to the initiation of underwater treadmill exercise. The ANCOVA was designed to assess the posttreatment postural sway variables (ie, data obtained on days 42 and 70), with day 14 as a covariate to adjust for any influences that existed at that time point. The ANOVA tables were used to determine significant main effects and interactions between main effect variables. When individual comparisons were made, least squares means were used. Values were reported as mean ± SEM, and values of P < 0.05 were considered significant.
Results
Craniocaudal sway—Craniocaudal sway differed significantly (P = 0.04) depending on the stance positions;when blindfolded horses had the largest amplitudes of craniocaudal sway, compared with the base-narrow and normal square stance positions (Table 3). Pretreatment analysis (data from days −7 and 14) revealed that craniocaudal sway did not differ significantly depending on whether the horse was in the aquatic exercise or control group (ie, no main effect of treatment group). However, posttreatment analysis (data from days 42 and 70) revealed that craniocaudal sway was significantly (P < 0.001) influenced by the presence or absence of aquatic exercise. Horses exercised on the underwater treadmill had significantly decreased craniocaudal sway, compared with that of the control group (Table 4). When specific individual comparisons were conducted, craniocaudal sway in the blindfolded normal square stance position was significantly decreased at days 42 (P = 0.004) and 70 (P = 0.03) in the aquatic exercise group, compared with findings in the control groups (Figure 2). Similarly, horses exercised on the underwater treadmill had significantly decreased craniocaudal sway when standing in a base-narrow stance position on both days 42 (mean ± SEM, 1.1 ± 0.2%; P = 0.03) and 70 (1.0 ± 0.2%; P = 0.009), compared with the control horses (1.7 ± 0.2% and 1.8 ± 0.2%, respectively). No significant differences were detected in the normal square stance position between the 2 groups across either of the posttreatment study days.
Mean ± SEM normalized amplitudes for craniocaudal COP movement during the blindfolded normal square stance position (ie, vertical placement and even weight distribution on all 4 limbs with blindfolding) in horses with a surgically created osteochondral fragment in 1 randomly selected carpal joint that subsequently underwent underwater treadmill exercise (UWT; n = 8) or overground treadmill exercise (control; 8) as determined on days 42 and 70 (osteochondral fragmentation was induced on day 0, and treadmill exercise commenced on day 15). Amplitudes for this variable are reported as a normalized percentage value (eg, postural sway variable divided by stance position variable times 100%).26 a–cValues with different letters are significantly (P < 0.05) different between the treatment groups within and across the study days.
Citation: American Journal of Veterinary Research 74, 7; 10.2460/ajvr.74.7.971
Mean ± SEM normalized postural sway variables obtained from all 16 horses in Table 1 when they were in a normal square stance position, a base-narrow stance position, and a blindfolded normal square stance position, controlling for the effects of time and treatment group.
| Stance position | |||
|---|---|---|---|
| Normalized COP variable | Normal | Base-narrow | Blindfolded |
| Craniocaudal sway (%) | 1.3 ± 0.1b | 1.4 ± 0.1b | 1.6 ± 0.1a |
| Mediolateral sway (%) | 4.8 ± 0.5b | 6.5 ± 0.5a | 6.0 ± 0.5a |
| COP area (%) | 69.0 ± 17.7b | 112.0 ± 17.7a | 126.0 ± 17.7a |
| COP velocity (mm/s) | 13.9 ± 1.1b | 14.8 ± 1.1a | 16.1 ± 1.1a |
| COP radius (mm) | 4.6 ± 0.3 | 4.5 ± 0.3 | 4.7 ± 0.3 |
Within a row, different superscript letters indicate significant (P < 0.05) differences between the stance positions.
See Tables 1 and 2 for remainder of key.
Mean ± SEM postural sway variables obtained from the 2 groups of horses in Table 1 at 28 and 56 days after commencement of treadmill exercise (ie, days 42 and 70 after surgical creation of an osteochondral fragment in 1 randomly selected carpal joint), controlling for the effect of stance position and study day.
| Treatment | ||
|---|---|---|
| COP variable | Control | UWT |
| Craniocaudal sway (%) | 1.7 ± 0.1a | 1.1 ± 0.1b |
| Mediolateral sway (%) | 6.6 ± 0.3a | 4.4 ± 0.3b |
| COP area (%) | 135.2 ± 11.0a | 56.4 ± 11.0b |
| COP velocity (mm/s) | 15.1 ± 0.8a | 12.5 ± 0.8b |
| COP radius (mm) | 5.5 ± 0.2a | 3.6 ± 0.2b |
See Tables 1, 2, and 3 for key.
The magnitude of the thoracic limb craniocaudal COP movement toward or away from the limb with the surgically created osteochondral fragment in the carpal joint was significantly (P = 0.02) influenced by the presence of the osteochondral fragment, depending on the stance position and applied treatment. When individual comparisons were conducted, the craniocaudal COP location shifted significantly (P < 0.001) toward the unaltered thoracic limb in the blindfolded normal square stance position within the control group (2.2 ± 0.1%), compared with findings for the limb with the surgically created osteochondral fragment (1.5 ± 0.1%). In addition, the blindfolded normal square stance position in the control group resulted in a significantly (P < 0.001) greater shift in the location of the craniocaudal COP toward the unaltered thoracic limb (2.2 ± 0.1%), compared with findings for the unaltered thoracic limb in horses exercised on the underwater treadmill (1.3 ± 0.1%). Similarly, the craniocaudal COP displacement during the base-narrow stance position had a significantly (P = 0.007) greater shift toward the unaltered thoracic limb in the control group (1.7 ± 0.1%), compared with findings for the unaltered thoracic limb in horses exercised on the underwater treadmill (1.2 ± 0.1%). Under both stance conditions, horses in the control group had a significant shift of the craniocaudal COP away from the limb with the surgically created osteochondral fragment, thereby unloading the injured and presumably painful joint.
Mediolateral sway—Mediolateral sway differed significantly (P = 0.01) depending on the stance position. Significantly increased amplitudes of mediolateral sway were evident in the base-narrow and blindfolded normal square stance positions, compared with that in the normal square stance position (Table 3). No significant differences were detected in mediolateral sway between the base-narrow and blindfolded normal square stance positions. Pretreatment analysis (data from days −7 and 14) revealed that mediolateral sway did not differ significantly between the aquatic exercise and control groups. However, posttreatment analysis (data from days 42 and 70) revealed that horses undergoing underwater treadmill exercise had significantly (P < 0.001) reduced mediolateral sway, compared with that of control horses (Table 4). Results of least squares mean comparisons indicated that mediolateral sway on posttreatment days 42 and 70 was significantly different depending on stance position and the presence or absence of aquatic exercise. Horses exercised on the underwater treadmill had significantly (P = 0.02) decreased mediolateral sway when blindfolded in the normal square stance position on day 42 (4.8 ± 0.7%), compared with that of the horses exercised on the overground treadmill (7.1 ± 0.7%). Similarly, the aquatic exercise group had significantly (P = 0.01) decreased mediolateral sway when standing in a base-narrow stance position on day 42 (4.8 ± 0.7%), compared with that of the control group (8.2 ± 0.7%). Lastly, horses undergoing aquatic exercise had significantly reduced mediolateral sway in the normal square stance position on both days 42 (P = 0.01) and 70 (P = 0.04), compared with findings for horses in the control group (Figure 3).
Mean ± SEM normalized amplitudes of the mediolateral COP movement during the normal square stance position (ie, vertical placement and even weight distribution on all 4 limbs) in horses with a surgically created osteochondral fragment in 1 randomly selected carpal joint that subsequently underwent underwater treadmill exercise (UWT; n = 8) or overground treadmill exercise (control; 8) as determined on days 42 and 70 (osteochondral fragmentation was induced on day 0 and treadmill exercise commenced on day 15). See Figure 2 for remainder of key.
Citation: American Journal of Veterinary Research 74, 7; 10.2460/ajvr.74.7.971
COP area—The removal of visual cues by blindfolding in a normal square stance position caused a significantly (P = 0.02) larger area of COP movement, compared with that in a nonblindfolded normal square stance position (Figure 4). Similarly, the base-narrow stance position had a significantly larger COP area, compared with that in a normal square stance position. As expected, the pretreatment analysis (data from days −7 and 14) revealed that COP area did not differ significantly depending on whether the horse was undergoing underwater or overground treadmill exercise. In contrast, posttreatment analysis (data from days 42 and 70) revealed significantly (P < 0.001) smaller COP areas for horses exercised on the underwater treadmill, compared with horses exercised on the overground treadmill (Table 4).
Stabilograms (depicting the movement of the COP in mediolateral [x-axis] and craniocaudal [y-axis] directions) obtained during 10-second data collection periods on day −7 from a single horse with a surgically created osteochondral fragment in a carpal joint in the blindfolded normal square stance position (A), the nonblindfolded normal square stance position (B), and the base-narrow stance position (C). In the blindfolded normal square stance and the base-narrow stance positions, the area of COP movement is increased, compared with that in the normal square stance position. In the normal square stance position, changes in the amplitudes of mediolateral and craniocaudal COP are small.
Citation: American Journal of Veterinary Research 74, 7; 10.2460/ajvr.74.7.971
COP velocity—Changing stance position alone was effective in influencing the COP velocity. A lack of visual stimulation (ie, blindfolding) in the normal square stance position caused a significantly (P = 0.007) greater velocity in the COM displacement, compared with that in a nonblindfolded normal square stance position (Table 3). The COP velocity was not significantly different between the blindfolded normal square and base-narrow stance position or between the base-narrow and normal square stance positions. Pretreatment analysis (data from days −7 and 14) revealed that COP velocity did not differ significantly depending on whether the horse was undergoing underwater or overground treadmill exercise. The presence or absence of aquatic exercise significantly (P = 0.05) influenced the COP velocity in the posttreatment analysis (data from days 42 and 70). Specifically, the aquatic exercise group maintained a slower COP velocity and thus better balance restraint, compared with findings for the control group (Table 4).
COP radius—Alterations in stance position did not significantly influence the COP radius (Table 3), and pretreatment analysis (data from days −7 and 14) revealed no significant difference between the aquatic exercise and control groups. However, the COP radius was significantly influenced by the presence or absence of aquatic exercise in the posttreatment analysis (data from days 42 and 70; Table 4). The horses undergoing underwater treadmill exercise had a significantly (P < 0.001) smaller COP radius than that of the horses undergoing overground treadmill exercise across all 3 stance positions.
Craniocaudal length—The craniocaudal length was significantly (P = 0.006) influenced by the 3 stance positions. Specifically, the normal square stance position had a significantly smaller craniocaudal length (1.06 ± 0.01 m), compared with lengths for both the blindfolded normal square stance position (1.07 ± 0.01 m; P = 0.02) and base-narrow stance position (1.08 ± 0.01 m; P = 0.002). No significant differences were detected between the craniocaudal lengths for the base-narrow and blindfolded normal square stance positions. Study day and the presence or absence of aquatic exercise did not significantly influence craniocaudal length.
Mediolateral width—The base-narrow stance position had a significantly (P < 0.001) smaller mediolateral width (0.210 ± 0.005 m), compared with widths for both the blindfolded (0.250 ± 0.005 m) and the normal square stance positions (0.250 ± 0.005 m). No significant differences were detected between the normal and blindfolded mediolateral width. Study day and the presence or absence of aquatic exercise did not significantly influence mediolateral width.
Head height—The height of the head was significantly (P < 0.001) lower in the base-narrow stance position (88.0 ± 2.0%), compared with the height of the head in the blindfolded normal square stance position (92.0 ± 2.0%). Head height did not significantly differ between the normal square (90.0 ± 2.0%) and base-narrow stance positions; likewise, head height did not differ between the blindfolded and nonblindfolded normal square stance positions. Height of the head did not significantly change among study days for each stance position, and a significant difference was not detected between the 2 groups of horses.
Clinical examinations—Mild lameness (mean score, 1.81 ± 0.05) was evident in all limbs of horses in which an osteochondral fragment was surgically created. This mean lameness score differed significantly (P < 0.001) from the baseline value determined before osteoarthritis was induced (0.03 ± 0.09), and the difference persisted throughout the study period (Figure 5). The mean lameness score for the limbs with a surgically created osteochondral fragment in the carpal joint in horses that underwent aquatic exercise (2.5 ± 0.2) was significantly (P = 0.04) higher at day 21 (7 days after the initiation of aquatic exercise), compared with the score for the limbs with a surgically created osteochondral fragment in the carpal joint of horses in the control group (2.0 ± 0.2). However, the mean lameness scores for the limbs with a surgically created osteochondral fragment in the carpal joint in horses that underwent aquatic exercise at day 42 (1.4 ± 0.2) and day 49 (1.9 ± 0.2), which was 28 and 35 days after the initiation of aquatic exercise, were significantly (P = 0.04) lower, compared with the scores for the limbs with a surgically created osteochondral fragment in the carpal joint of horses in the control group (1.9 ± 0.2 and 2.4 ± 0.2, respectively). Lameness scores for the horses that underwent underwater treadmill exercise decreased beginning on day 28 and continued for the remainder of the study, although significant differences between the 2 groups were only evident on days 42 and 49. Overall, there was no significant difference in the subjective lameness scores with respect to treatment.
Mean ± SEM lameness scores in the thoracic limbs with (red lines) and without (blue lines) a surgically created osteochondral fragment in 1 randomly selected carpal joint of horses that subsequently underwent underwater treadmill exercise (n = 8 [squares]) or overground treadmill exercise (8 [triangles]). Osteochondral fragmentation was induced on day 0, and treadmill exercise commenced on day 15 (arrow). Notice the significant increase in lameness (attributable to the osteochondral fragment) from day 0, which continued throughout the study period. *At a given time point, the value for limbs with a surgically created osteochondral fragment in the carpal joint of horses undergoing aquatic exercise is significantly (P < 0.005) different from the value for those altered limbs in horses undergoing overground treadmill exercise.
Citation: American Journal of Veterinary Research 74, 7; 10.2460/ajvr.74.7.971
Discussion
In humans, evaluation of balance control through postural sway analysis is a reliable and valid approach to determining static stability under various conditions.16,21,23 Results of the present study indicated that underwater treadmill exercise significantly improves static balance control in horses with carpal joint osteoarthritis, which is pivotal to initiating evidence-based support for the use of aquatic exercise in the management of joint disease in this species. This study has also provided the first evidence that variations in stance positions have profound effects on the mechanics of standing balance in horses. Direct measurement of changes in the COP movements confirmed that a narrow stance width and removal of visual stimuli significantly decreased postural stability.
Balance control is a dynamic task in which sensorimotor systems must interact with the external environment to maintain stability.31 One of the most striking findings of the present study was that when the horses’ thoracic limbs were placed in a base-narrow stance position, there was a significant increase in mediolateral sway but no change in craniocaudal sway. This may be explained by the mechanical differences in joint structure and function that contribute to motion or stability within the 2 planes of movement. Under normal circumstances, mediolateral stability of thoracic limb articulations is provided by robust, periarticular collateral ligaments.23 However, joint motion within a sagittal plane is normally achieved by the flexion-extension of the highly mobile hinge joints within the thoracic and pelvic limbs.23 In humans with mediolateral specific instability, a wider stance width is often adopted to reduce the lateral COP displacement. The increase in stance width reduces the lateral rotation about the ankle joint, thereby decreasing angular motion and improving mediolateral stabilization.23 The ankle, knee, and hip joints move independently in the sagittal plane, allowing for increased control of the COP movement in a craniocaudal direction during postural sway analysis.23 However, lateral joint movement in the frontal plane cannot occur independently: a change in one joint angle leads to predictable angle changes in the other joints.23 In a feet-together stance position, people often have increased mediolateral COP displacement because of increased lateral rotation about the ankle joint, which leads to enhanced proprioceptive signaling and increased recruitment of the pelvis and hip musculature to change the position of the hip joint to improve balance control.23 Evidence of a similar neuromuscular adaption would be expected during postural sway analysis in horses. A base-narrow stance position may also alter the direction of the ground reaction force vectors, thereby changing the force distribution across the joint surfaces.31 In particular, by reducing the size and configuration of the base of support, the horizontal force vectors become more laterally directed, which leads to excessive joint loading and instability.31 The altered force vector orientation may cause a counterproductive secondary increase in muscle activation, which results in increased levels of tonic muscle activity and large fluctuations in the COP.23,31 The base-narrow stance data obtained in the present study suggested that hoof position is an important determinant of standing balance in horses and should be considered when altering hoof position through trimming and shoeing applications or when certain terrain characteristics force a horse to move or stand in a base-narrow stance position.
The visual system is a major contributor to balance control, providing both spatial and temporal information.21 The removal of visual cues requires an increased reliance on proprioceptive and vestibular sensory cues to maintain balance, which signal body movement relative to the foot and head. Reduced afferent information results in less precise balance control because the CNS has less information with which to accurately estimate and control the COM position and velocity.18 In the present study, the blindfolded normal square stance position resulted in larger and more rapid displacements of the COP, which indicates a high reliance of postural control on visual cues in horses. This finding is consistent with the well-documented phenomenon of impaired balance control in humans who are standing with their eyes closed.19,21,32 The loss of visual cues also causes an increase in muscle stiffness and alters postural sway, thereby impacting reflexive responses to alterations in COM movement.18 In human studies,21,32 the compensatory effects of vision under conditions of muscular fatigue have been identified. In those studies, the destabilizing effects of muscular fatigue on balance control were significantly compensated for when individuals stood with their eyes open versus eyes closed. In addition, elderly humans typically have progressive loss of their peripheral vision,18 which causes reduced regulation of COM movements, subsequently increasing the risk of falling.18 The present study's findings have indicated that the removal of visual cues is an important determinant of standing balance control in horses and should strongly be considered when visual input is restricted for the purposes of various athletic activities (eg, application of blinders). However, a limitation of the present study was that horses were only assessed in a completely blindfolded and static condition; future studies need to address various forms of partial to complete blockade of visual stimuli in both static and dynamic settings.
The present study has provided the evidence that underwater treadmill exercise improves static balance control in horses with carpal joint osteoarthritis. The increased resistance and buoyancy inherent in aquatic exercise minimize joint instability and weight-bearing stresses applied to limbs.33 Humans with osteoarthritis who are involved in aquatic treatment programs often have a significant reduction in pain, increase in muscle strength, and improvement in motor control.34 Aquatic exercise reduces postural sway in women with osteoarthritis of the lower extremities, as demonstrated in 2-legged stance tests with the eyes closed.17 The most notable differences in postural stability of the horses in the present study were associated with the blindfolded state, when affected horses were deprived of visual cues and had to rely on vestibular and somatosensory input to maintain balance. Underwater treadmill exercise improved balance control in both the base-narrow and blindfolded normal square stance positions, whereas the control group maintained balance only when placed in a normal square stance position. Underwater treadmill exercise may have increased the afferent excitation of the motor neuron pool for the muscles responsible for stabilizing the thoracic and pelvic limbs. Therefore, sensory receptors (eg, muscle spindles, Golgi tendon organs, and joint mechanoreceptors), afferent transmission, and efferent motor control may have been positively affected through aquatic exercise, which resulted in improved postural control in those horses. Underwater treadmill exercise increases joint mobility, reduces signs of pain and inflammation associated with impact loading, promotes normal motor patterns, and increases proprioceptive acuity and postural control.6,35 Horses in the control group had a greater disturbance in postural stability, specifically when challenged with conditions requiring increased proprioceptive input and motor control for maintaining equilibrium. Increased variation in postural sway characteristics within the control group may be associated with increased stimulation of nociceptive pathways and altered afferent signaling from the joint mechanoreceptors as a result of joint pain and inflammation. Altered sensory feedback caused by synovial effusion and inflammation leads to altered muscle activity,13 which may ultimately impair a horse's ability to make motor output adjustments efficiently during functional mobility and dynamic control of balance.
During dynamic gait analysis, the decreased lameness scores for the horses undergoing underwater treadmill exercise corresponded with an improvement in postural sway variables at similar time points. Although underwater treadmill exercise did not have a beneficial effect on lameness scores at all time points, the improvements in balance control under various stance positions in those horses that underwent underwater treadmill exercise may translate into improved functional locomotion. Humans that undertook a 12-week aquatic exercise program had improvements in static postural control and dynamic balance test results.36 Aquatic exercise affects both balance and strength, which translates into improved dynamic control on land. The enhanced afferent input used to maintain the position of the base of support in those horses exercised on the underwater treadmill in the present study establishes a foundation for precise dynamic neuromuscular control.
For the horses of the present study, surgical creation of an osteochondral fragment in a carpal joint influenced the magnitude of craniocaudal COP displacement in both the base-narrow and blindfolded normal square stance positions. Under both of these conditions, horses in the control group were more likely to shift the craniocaudal COP away from the limb with a surgically created osteochondral fragment in the carpal joint, thereby unloading the injured or painful joint. This shift in the COP may be attributable to the creation of the osteochondral fragment, which remains adhered to the joint capsule proximally and may subsequently influence the sensitivity of the joint capsule mechanoreceptors. Altered afferent mechanoreceptor signaling as a result of joint pain and inflammation may have influenced the shift of the COP away from the limb with a surgically created osteochondral fragment in the carpal joint during assessment of the craniocaudal sway. The disruption of joint homeostasis creates altered neural activity, resulting in disrupted motor control and movement patterns. The reduced amplitudes of craniocaudal COP displacement in both limbs with and without a surgically created osteochondral fragment in the carpal joint of horses that were undergoing underwater treadmill exercise provided additional evidence for the therapeutic potential of aquatic exercise. Future studies are needed to determine the influence of altered static postural control on dynamic balance characteristics and performance capabilities in horses.
Although the exact mechanism for improved postural control after aquatic exercise could not be determined in the horses of the present study, postural sway analysis did detect positive changes in balance control in horses with carpal joint osteoarthritis. Postural sway analysis is a sensitive diagnostic technique that may aid in identifying potential proprioceptive or balance deficits associated with neuromuscular impairments associated with joint pain or inflammation. Postural sway analysis can also provide objective outcome variables with which to monitor the effects that specific rehabilitation programs may have on postural stability. Results of the present study have provided clinical insights into the effects of altered visual input and stance positions on a horse's ability to maintain postural control. The destabilizing effects of altered vision and base-narrow stance positions during quiet standing should be considered when evaluating performance and compensatory or adaptation strategies.
ABBREVIATIONS
| COM | Center of mass |
| COP | Center of pressure |
Super Glue (cyanoacrylate adhesive), Pacer Technology, Rancho Cucamonga, Calif.
Naxcel, Pfizer Inc, New York, NY.
Equi-Phar, Schering-Plough Animal Health Corp, Union, NJ.
Equigym, Equigym LLC, Lexington, Ky.
Equine underwater treadmill, Ferno Veterinary Systems, Wilmington, Ohio.
Model FP6090-15, Bertec Corp, Columbus, Ohio.
Motus Software, version 9.1, Vicon, Centennial, Colo.
200-Hz cameras, Vicon, Centennial, Colo.
Proc GLIMMIX, SAS, version 9.3, SAS Institute Inc, Cary, NC.
References
1. De Carli P, Patrizi M, Pepe L, et al. Postural control and risk of falling in bipodalic and monopodalic stabilometric tests of healthy subjects before, after visuo-proprioceptive vestibulo-postural rehabilitation and at 3 months thereafter: role of the proprioceptive system. Acta Otorhinolaryngol Ital 2010; 30: 182–189.
2. Lyytinen T, Liikavainio T, Bragge T, et al. Postural control and thigh muscle activity in men with knee osteoarthritis. J Electromyogr Kinesiol 2010; 20: 1066–1074.
3. Riemann B, Lephart S. The sensorimotor system, part 1: the physiologic basis of functional joint stability. J Athl Train 2002; 37: 71–79.
4. Boucher P, Descarreaux M, Normand M. Postural control in people with osteoarthritis of the cervical spine. J Manipulative Physiol Ther 2008; 31: 184–190.
5. McIlwraith CW, Vachon A. Review of pathogenesis and treatment of degenerative joint disease. Equine Vet J Suppl 1988;(6):3–11.
6. Hurley MV. The effects of joint damage on muscle function, proprioception and rehabilitation. Man Ther 1997; 2: 11–17.
7. Salo P. The role of joint innervation in the pathogenesis of arthritis. Can J Surg 1999; 42: 91–107.
8. Iles J, Stokes M, Young A. Reflex actions of knee joint afferents during contraction of the human quadriceps. Clin Physiol 1990; 10: 489–500.
9. Hopkins J, Ingersoll C, Edwards J, et al. Changes in soleus motoneuron pool excitability after artifical knee joint effusion. Arch Phys Med Rehabil 2000; 81: 1199–1203.
10. Hopkins J, Ingersoll C, Krause B, et al. Effect of knee joint effusion on quadriceps and soleus motoneuron pool excitability. Med Sci Sports Exerc 2001; 33: 123–126.
11. Newton RA. Joint receptor contributions to reflexive and kinesthetic responses. Phys Ther 1982; 62: 22–29.
12. Palmieri R, Ingersoll C, Hoffman M, et al. Arthrogenic muscle response to a simulated ankle joint effusion. Br J Sports Med 2004; 38: 26–30.
13. Palmieri R, Tom J, Edwards J, et al. Arthrogenic muscle response induced by an experimental knee joint effusion is mediated by pre- and post-synaptic spinal mechanisms. J Electromyogr Kinesiol 2004; 14: 631–640.
14. Peterka R. Sensorimotor Integration in human postural control. J Neurophysiol 2002; 88: 1097–1118.
15. Chiari L, Rocchi L, Cappello A. Stabilometric parameters are affected by anthropometry and foot placement. Clin Biomech 2002; 17: 666–677.
16. Lafond D, Corriveau H, Hebert R, et al. Intrasession reliability of center of pressure measures of postural steadiness in healthy elderly people. Arch Phys Med Rehabil 2004; 85: 896–901.
17. Suomi R, Koceja D. Postural sway characteristics in women with lower extremity arthritis before and after an aquatic exercise intervention. Arch Phys Med Rehabil 2000; 81: 780–785.
18. Jancová J. Measuring the balance control system—review. Acta Medica (Hradec Kralove) 2008; 51: 129–137.
19. Messier S, Royer T, Craven T, et al. Long-term exercise and its effect on balance in older, osteoarthritic adults: results from the fitness, arthritis, and seniors trial (FAST). J Am Geriatr Soc 2000; 48: 131–138.
20. Rasch A, Dalén N, Berg H. Muscle strength, gait, and balance in 20 patients with hip osteoarthritis followed 2 years after THA. Acta Orthop 2010; 81: 183–188.
21. Nagano A, Yoshioka S, Hay D, et al. Influence of vision and static stretch of the calf muscles on postural sway during quiet standing. Human Mov Sci 2006; 25: 422–434.
22. Bisson E, McEwen D, Lajoie Y, et al. Effects of ankle and hip muscle fatigue on postural sway and attentional demands during unipedal stance. Gait Posture 2011; 33: 83–87.
23. Day B, Steiger M, Thompson P, et al. Effect of vision and stance width on human body motion when standing: implications for afferent control on lateral sway. J Physiol 1993; 469: 479–499.
24. Prins J, Cutner D. Aquatic therapy in the rehabilitation of athletic injuries. Clin Sports Med 1999; 18: 447–460.
25. Bialski D, Lanovaz J, Bohart G, et al. Effect of detomidine on postural sway in horses. Equine Comp Exerc Physiol 2004; 1: 45–50.
26. Clayton HM, Bialski DE, Lanovaz JL, et al. Assessment of the reliability of a technique to measure postural sway in horses. Am J Vet Res 2003; 64: 1354–1359.
27. Frisbie D, Ghivizzani C, Robbins P, et al. Treatment of experimental equine osteoarthritis by in vivo delivery of the equine interleukin-1 receptor antagonist gene. Gene Ther 2002; 9: 12–20.
28. Frisbie DD, McIlwraith CW, Kawcak CE, et al. Evaluation of topically administered diclofenac liposomal cream for treatment of horses with experimentally induced osteoarthritis. Am J Vet Res 2009; 70: 210–215.
29. McClintock SA, Hutchins DR, Brownlow MA. Determination of weight reduction in horses in flotation tanks. Equine Vet J 1987; 19: 70–71.
30. American Association of Equine Practitioners. Definition and classification of lameness. In: Kester W, ed. Guide for veterinary services and judging veterinary events. 4th ed. Lexington, Ky: American Association of Equine Practitioners 1991;19.
31. Fung J, Macpherson J. Determinants of postural orientation in quadrupedal stance. J Neurosci 1995; 15: 1121–1131.
32. Vuillerme N, Nougier V, Prieur JM. Can vision compensate for a lower limbs muscular fatigue for controlling posture in humans? Neurosci Lett 2001; 308: 103–106.
33. Evans B, Cureton K, Purvis J. Metabolic and circulatory responses to walking and jogging in water. Res Q 1978; 49: 442–449.
34. Giaquinto S, Ciotola E, Margutti F, et al. Gait during hydrokinesitherapy following total hip arthroplasty. Dis Rehabil 2007; 29: 743–749.
35. Hassan B, Doherty S, Mockett S, et al. Effect of pain reduction on postural sway, proprioception, and quadriceps strength in subjects with knee osteoarthritis. Ann Rheum Dis 2002; 61: 422–428.
36. Waters D, Hale L. Do aqua-aerobics improve gait and balance in older people? Int J Ther Rehab 2007; 14: 538–542.
