LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DateDuepes-p. 15

 

BIOMECHANICS OF THE EQUINE TARSAL JOINT
By

Siripom Khumsap

A DISSERTATION
Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Large Animal Clinical Sciences

2002

ABSTRACT
BIOMECHANICS OF THE EQUINE TARSAL JOINT
By

Siripom Khumsap

This study was performed to test whether two-dimensional (2D) and three-
dimensional (3D) kinematic, ground reaction force (GRF) and 2D inverse dynamic
analysis could identify differences between sound horses and horses with mild tarsal joint
lameness. Hypotheses tested were that kinematic and kinetic profiles of horses with distal
tarsal synovitis differ from those of sound horses, and that the tarsal joint complex
undergoes 3D motion. Four sound horses were selected on the basis of clinical
evaluation, radiography and nuclear scintigraphy. Gait analysis was performed for the
sound condition, after which synovitis was induced by injecting endotoxin into the right
distal intertarsal and tarsometatarsal joints. Gait analysis was repeated 24-30 hours later
for the lame condition. Independent t-tests were performed to identify differences
between the variables in the two conditions. In the sound condition, tarsal joint ranges of
motion were 12. 12° 1- 0.96 and 44.67° i 1.90 during stance and swing, respectively. The
peak vertical force on the hind limb was 9.05 i 0.40 N/kg. During stance, the tarsal joint
power profile showed two cycles of elastic energy storage and release on the extensor
aspect, which was a mechanism for minimizing energy expenditure at the trot. During
swing, the tarsal joint power profiles showed two bursts of energy generation, on the
flexor aspect in early swing then on the extensor aspect in late swing. After synovitis

induction, tarsal joint range of motion during stance decreased by 2.20° i 1.28. Peak

vertical force and vertical impulse decreased by 0.31 i- 0.18 N/kg and 0.05 i 0.03 Ns/kg.
The mechanical deficits at the tarsal joint included a trend toward a decrease in energy
absorption during early stance, and a decrease in energy generation during push off.
During swing, there was a trend toward a decrease in tarsal positive net joint energy.
There was no compensatory increase in energy generation from the other joints within the
lame hind limb. The contralateral hind limb showed a trend toward an increase in peak
vertical force of 0. 1 7 i 0.13 N/kg and increase in positive power at the coxofemoral joint
during early swing. Vertical impulse in the contralateral forelimb decreased by 0.05 i
0.02 Ns/kg. It was found that the tarsal joint complex showed 3D rotational and
translational motions. In the sound condition, the tarsal joint complex was flexed,
abducted and internally rotated, and translated in cranial, lateral and proximal directions
during stance. The motions during swing were similar to those during stance, except that
the joint externally rotated. Skin-based markers could estimate those motions fairly
accurately, except for intemal/external rotation and medial/lateral translation. Synovitis
resulted in decreases in range of tarsal joint flexion of 25° i 1.4, cranial translation of 4.2
i 2.5 mm during stance and proximal translation of 2.4 i 1.4 mm during swing. It is
concluded that the tarsal joint is an important source of mechanical energy generation
throughout the stride, and that mechanical deficiencies due to distal tarsal synovitis were
mainly present during stance. Compensation by the other limbs included load shifting to
the contralateral hind limb and unloading of the diagonal lame limb pair. The decrease in
tarsal joint translation during stance due to synovitis might lead to stationary articular
cartilage of the distal tarsal joints. High intensity loading on the cartilage may lead to

cartilage necrosis, which may explain pathogenesis of bone spavin.

ACKNOWLEDGMENTS

First, I would like to extend my appreciation for all the advice, support and kindly
understanding from my guidance committee: Dr. John Stick, Dr. Steven Arnoczky, Dr.
Frank Nickels and Dr. Diana Rosenstein. A sincerely grateful appreciation to my major
professor and advisor, Dr. Hilary Clayton. Being a graduate student is a hard work. It is
even harder to be an international graduate student. With her knowledge, generous
understanding and good humor, she is the best advisor ever and will be the goal standard
for me in the future.

I would like to thank Dr. Christopher Byron who help me performed the clinical
procedures in this study, all the graduate students in the Mary Anne McPhail Equine
Performance Center and Carissa Wickens for assistance during data collection. A special
thank to Joel Lanovaz for his expertise, friendship and advice.

This study supported by the Mary Anne McPhail endowment and the Department
of Large Animal Clinical Sciences. Thanks to those horses whom ‘volunteered’ to
participate in my study: Apollo, Homer, Pulga, Athena and Trouble.

Finally, I would like to thank my parents who always support my will for

education and my husband who always makes me smile and happy.

iv

TABLE OF CONTENTS

LIST OF TABLES ................................................................................. viii

LIST OF FIGURES ................................................................................ xiii

INTRODUCTION ................................................................................... l

Rationale ..................................................................................... 1

Problem statement ........................................................................... 2

Objectives .................................................................................... 5

Hypotheses ................................................................................... 5

Overview ..................................................................................... 6
CHAPTER 1

EQUINE GAIT ANALYSIS: REVIEW ON CLINICAL APPLICATION ................ 8

History of gait analysis in horses ........................................................ 8

Gait analysis in sound horses ............................................................. 1 l

Kinematic analysis ................................................................ 11

Kinetic analysis ................................................................... l4

Gait analysis in lame horses .............................................................. l7

Kinematic analysis in lame horses .............................................. 17

Kinetic analysis in lame horses ................................................. 22

Combined kinematic and kinetic analyses ..................................... 25

Effects of treatment regimes .................................................... 26

Inverse dynamic analysis: clinical application .......................................... 29

Inverse dynamics in sound horses .............................................. 31

Inverse dynamics in lame horses and effects of treatment .................. 33

Three-dimensional analysis in equine biomechanics .................................. 36

Conclusion .................................................................................. 38
CHAPTER 2

MATERIALS AND METHODS ................................................................. 39

Subject selection ........................................................................... 40

Lameness evaluation ............................................................. 40

Radiography ....................................................................... 41

Nuclear scintigraphy ............................................................. 41

Ground reaction force analysis .................................................. 42

Tarsal synovitis induction ................................................................ 43

Endotoxin preparation ............................................................ 43

lntraarticular injection of endotoxin ............................................ 43

Care after endotoxin injection ................................................... 44
Gait analysis ................................................................................ 44
Kinematic data ..................................................................... 46
Force plate data ................................................................... 53
Data collection procedures ....................................................... 54
Data processing and analysis ............................................................. 57
Two-dimensional variables ...................................................... 57
Three-dimensional kinematic variables for the tarsal joint .................. 65
Statistical analysis ......................................................................... 70
CHAPTER 3
SAGITTAL PLANE KINEMATICS AND KINETICS OF THE HIND LIMBS
IN SOUND TROTTING HORSES: RESULTS AND DISCUSSION ..................... 73
Horse selection ........................................................................... 73
Test for difference in velocity and symmetry ......................................... 74
Ground reaction force variables ........................................................ 74
Two-dimensional biomechanical profiles ............................................. 80
Coxofemoral joint ............................................................... 80
Femorotibial joint ................................................................ 80
Tarsal joint ........................................................................ 81
Metatarsophalangeal joint ...................................................... 82
Distal interphalangeal joint .................................................... 83
Discussion ................................................................................. 1 O4
Kinematic data ................................................................... 104
Force plate data .................................................................. 107
Inverse dynamic analysis ....................................................... 109
Coxofemoral joint ............................................................... 1 10
Femorotibial joint ............................................................... 1 ll
Tarsal joint ........................................................................ 1 l3
Metatarsophalangeal joint ...................................................... l 14
Distal interphalangeal joint .................................................... 115
Hind limb coordination ......................................................... 116
Conclusion ................................................................................. 1 19
CHAPTER 4

EFFECT OF UNILATERAL SYNOVITIS OF DISTAL INTERTARSAL AND
TARSOMETATARSAL JOINTS ON SAGITTAL PLANE KINEMATICS AND

KINETICS OF TROTTING HORSES: RESULTS AND DISCUSSION ................. 120
Lameness and physical examination ................................................... 120

Effect of synovitis on intra-limb coordination ....................................... 121

Test for differences in velocities .............................................. 121

Kinematic variables ............................................................. 122

Ground reaction forces ......................................................... 122

vi

Net joint moments, net joint powers and net joint energies ............... 123

Forelimb variables in the sound condition ............................................ 137
Inter-limb compensation for lameness of the right hind limb ...................... 142
Kinematic variables ............................................................. 142
Ground reaction forces ......................................................... 142
Net joint moments, net joint powers and net joint energies ............... 143
Discussion ................................................................................. 147
Lameness model .................................................................. 147
Endotoxin induced synovial membrane inflammation ..................... 147
Effects of synovitis on symmetry of motion and
vertical impulse of hind limbs ................................................. 148
Intra-limb coordination ......................................................... 151
Inter-limb compensation ........................................................ 155
Conclusion ................................................................................. l 60
CHAPTER 5
THREE-DIMENSIONAL KINEMATICS OF THE EQUINE TARSAL JOINT:
RESULTS AND DISCUSSION ................................................................ 161
Reference bone data ..................................................................... 161
Comparison between reference bone data and skin marker data .................. 167
Effects of tarsal lameness on 3D motion ............................................... 189
Discussion ................................................................................ l 92
Conclusion ................................................................................. 200
CONCLUSION .................................................................................... 201
APPENDIX A

THE DEVELOPMENT OF CORRECTIONS FOR SKIN DISPLACEMENT
ARTIFACTS AT THE EQUINE TARSAL JOINT AND APPLICATION

TO 3D JOINT KINEMATICS ................................................................... 204
APPENDIX B

BIOMECHANICAL VARIABLES IN SOUND AND LAME CONDITIONS .......... 234
REFERENCES .................................................................................... 244

vii

Table 3-1

Table 3—2

Table 3-3

Table 3-4

Table 3-5

Table 3-6

Table 3-7

Table 3-8

Table 3-9

Table 3-10

LIST OF TABLES

Examination results from lameness evaluation, radiography
and nuclear scintigraphy ........................................................ 73

Statistical analysis results for testing differences of velocity
(m/s) and velocity in dimensionless units between left and
right limbs. Values are mean and (SD) ....................................... 74

Symmetry values in sound condition of vertical force peaks

(N/kg) and vertical impulses (N s/kg) between left and right limbs.
Standard values represent lefi-right symmetry of vertical

force peaks and vertical impulses expressed as mean and (SD)

(Merkens, et al., 1993) .......................................................... 75

Ground reaction force (GRF) variables, expressed as mean
values for both hind limbs, in sound trotting horses. Values are
mean and (SD) ................................................................... 78

Vertical impulse (NS/kg) on all four limbs of each horse .................. 79

Joint angle peaks (degrees) averaged from both hind limbs in

sound trotting horses during the stance phase. Values are mean

and (SD) ........................................................................... 96
Net joint moment peaks (Nm/kg) averaged from both hind limbs

in sound trotting horses during the stance phase. Values are mean

and (SD) ........................................................................... 97

Net joint power peaks (W/kg) averaged from both hind limbs in
sound trotting horses during the stance phase. Values are mean
and (SD) ........................................................................... 98

Net joint energies (J/kg) calculated by time integration of the
corresponding net joint power peaks during the stance phase.
Values were averaged from both hind limbs in sound trotting
horses. Values are mean and (SD) ............................................. 99

Joint angle peaks (degrees) averaged from both hind limbs in

sound trotting horses during the swing phase. Values are mean
and (SD) ........................................................................... 100

viii

 

Table 3-11 Net joint moment peaks (Nm/kg) averaged from both hind limbs
in sound trotting horses during the swing phase. Values are mean
and (SD) ........................................................................... 101

Table 3-12 Net joint power peaks (W/kg) averaged from both hind limbs in
sound trotting horses during the swing phase. Values are mean
and (SD) ........................................................................... 102

Table 3-13 Net joint energies (J/kg) calculated by time integration of the
corresponding net joint power peaks during the swing phase.
Values were averaged from both hind limbs in sound trotting
horses. Values are mean and (SD) ............................................. 103

Table 3-14 Summary of means energy generation and absorption (J/kg)
at the joints of the hind limb during stance phase, swing phase
and the total stride ............................................................... 104

Table 4-1 Statistical analysis comparing velocity (m/s) and velocity in
dimensionless units of each limb between sound and lame
conditions. Values are mean and (SD) ....................................... 121

Table 4-2 Differences between sound and lame conditions in variables that
differed significantly or showed a trend toward a significant
difference during the stance phase of the right hind limb .................. 124

Table 4-3 Differences between sound and lame conditions in variables that
differed significantly or showed a trend toward a significant
difference during the swing phase of the right hind limb .................. 125

Table 4-4 Symmetry index as a quotient value of difference of tuber coxae
displacement between the hind limb pair from the sound and lame
conditions. Values greater than one mean the displacement range
is greater in the right hind limb than the left hind limb ..................... 145

Table 4-5 Symmetry index as a percent quotient of difference of vertical
impulses between the hind limb pair from the sound and lame
conditions. Negative values mean the impulse is greater in the left
hind limb than the right hind limb ............................................. 145

Table 4-6 Symmetry index as a percent quotient of difference of vertical
impulses between the sound and lame conditions in the same
hind limb. Negative values mean the impulse is greater in the
sound condition than the lame condition ..................................... 145

Table 4-7 Differences between sound and lame conditions in variables

ix

Table 4-8

Table 5-1

Table 5-2

Table 5-3

Table 5-4

Table 5-5

Table 5-6

Table A-1

Table A-2

Table A-3

that differed significantly or showed a trend toward a significant
difference during stance and swing phases of the left hind limb ......... 146

Differences between sound and lame conditions in variables
that differed significantly or showed a trend toward a significant
difference during stance phases of the left fore limb ....................... 146

Root mean square (RMS) errors of three rotational motions

during stance between reference bone data and corrected and

uncorrected skin data. Shape agreement is the assessment

between reference bone data and 3D corrected skin data .................. 169

Root mean square (RMS) errors of three translational motions

during stance between reference bone data and corrected and

uncorrected skin data. Shape agreement is the assessment

between reference bone data and 3D corrected skin data .................. 170

Root mean square (RMS) errors of three rotational motions

during swing between reference bone data and corrected and

uncorrected skin data. Shape agreement is the assessment

between reference bone data and 3D corrected skin data .................. 171

Root mean square (RMS) errors of three translational motions

during swing between reference bone data and corrected and

uncorrected skin data. Shape agreement is the assessment

between reference bone data and 3D corrected skin data .................. 17 2

Differences of three-dimensional variables between sound and
lame conditions of the right hind limb. Values are mean and (SD) ...... 191

Range of motion obtained from reference bone data during
stance and swing phases ........................................................ 192

Descriptive data from the subjects and the mean descriptive
data from the kinematic trials ......................................... 224

Mean (SD) of marker locations from the standing poses of

subjects 1-3. Data are expressed with respect to the local

bone coordinate systems and are given in percentage of

segment length .................................................................. 224

Descriptive statistics from the modeled skin displacements
for the X, Y and Z coordinate for each skin surface marker.
The order indicates the optimal number of harmonics in the

Fourier model and the po value is the mean offset from the
standing pose location. The RMS error is the difference

Table A-4

Table 8-1

Table 8-2

Table 8-3

Table 8-4

Table 8-5

Table B-6

between the actual displacement and the model. The RMS

amp and Peak amp are the RMS and Peak to Peak amplitudes

of the skin displacement calculated using the model.

All percent values refer to percent of the tibia or third

metatarsus segment length ..................................................... 225

RMS differences between bone-fixed and skin-based kinematics ........ 226

Ground reaction force (GRF) variables in the hind limbs.

Values from sound condition are averaged from both hind limbs.
Lame RH is the limb in which synovitis was induced.
Compensating LH is the contralateral hind limb. Values are

mean and (so). a P<0.05. b 0.05<P<0.1 .................................... 234

Joint angle (degrees) during stance in the hind limbs.

Values from sound condition are averaged from both hind limbs.
Lame RH is the limb in which synovitis was induced.
Compensating LH is the contralateral hind limb. Values are

mean and (so). a P<0.05. b 0.05<P<0.1 .................................... 235

Net joint moment peaks (Nm/kg) during stance in the hind limbs.
Values from sound condition are averaged from both hind limbs.
Lame RH is the limb in which synovitis was induced.
Compensating LH is the contralateral hind limb. Values are

mean and (so). 3 P<0.05. b 0.05<P<O.l .................................... 236

Net joint power peaks (W/kg) during stance in the hind limbs.
Values from sound condition are averaged from both hind limbs.
Lame RH is the limb in which synovitis was induced.
Compensating LH is the contralateral hind limb. Values are

mean and (so). a P<0.05. b 0.05<P<0.1 .................................... 237

Net joint energies (J/kg) during stance in the hind limbs.

Values from sound condition are averaged from both hind limbs.
Lame RH is the limb in which synovitis was induced.
Compensating LH is the contralateral hind limb. Values are

mean and (so). 3 P<0.05. b 0.05<P<0.1 .................................... 238

Joint angle (degrees) during swing in the hind limbs.

Values from sound condition are averaged from both hind limbs.
Lame RH is the limb in which synovitis was induced.
Compensating LH is the contralateral hind limb. Values are

mean and (SD). a P<0.05. b 0.05<P<0.1 .................................... 239

xi

Table B-7 Net joint moment peaks (Nm/kg) during swing in the hind limbs.
Values from sound condition are averaged from both hind limbs.
Lame RH is the limb in which synovitis was induced.
Compensating LH is the contralateral hind limb. Values are

mean and (SD). 3 P<0.05. b 0.05<P<0.1 .................................... 240

Table B-8 Net joint power peaks (W/kg) during swing in the hind limbs.
Values from sound condition are averaged from both hind limbs.
Lame RH is the limb in which synovitis was induced.
Compensating LH is the contralateral hind limb. Values are

mean and (so). 3 P<0.05. b 0.05<P<0.1 .................................... 241

Table B-9 Net joint energies (J/kg) during swing in the hind limbs.
Values from sound condition are averaged from both hind limbs.
Lame RH is the limb in which synovitis was induced.
Compensating LH is the contralateral hind limb. Values are

mean and (so). 3 P<0.05. b 0.05<P<O.1 .................................... 242

Table 8-10 Ground reaction force (GRF) variables in the forelimbs.
Values from sound condition are averaged from both forelimbs.
Lame RF and Lame LF are the compensating limb. Values are

mean and (so). 3 P<0.05. b 0.05<P<0.1 .................................... 243

xii

Figure 2-1
Figure 2-2
Figure 2-3
Figure 2—4
Figure 2-5
Figure 2-6
Figure 2-7

Figure 2-8

Figure 2-9

Figure 2-10
Figure 2-11
Figure 2-12
Figure 2-13

Figure 2—14

Figure 2-15

Figure 3-1

LIST OF FIGURES

Flow chart showing sequence of procedures ................................ 39
Flow chart for gait analysis procedures ...................................... 45
Motion analysis video system .................................................. 47
Position of a rigid L—shaped frame in the data collection field. . . . . . . . 48
Data collection volume after calibration ...................................... 49

Position of retro-reflective markers on the cranial part of the body. . 50
Position of retro-reflective markers on the caudal part of the body ...... 51

Stick figure of whole body marker setup displayed by the
computer monitor ............................................................... 53

Diagram showing the method used to identify the center

of a circular object ............................................................... 59
Lateral radiograph of the hoof and hoof coordinate system ............... 60
Lateral radio graph of the tarsus and tarsal coordinate system ............ 61
Lateral side of equine femur and hip coordinate system. . . . . . . . . . . . 62
Segmental coordinate system of tibial and metatarsal segments ......... 67

Three—dimensional angular motion of the metatarsal segment

relative to the tibial segment; flexion(-)/extension(+) (left),
abduction(-)/adduction(+) (middle) and intemal(+)/extemal(-)

rotation (right). Figure illustrates the right hind limb ...................... 69

Three-dimensional translational motion of the metatarsal

segment relative to the tibial segment; cranial(+)/caudal(-)

translation (left), medial(+)/lateral(-) translation (middle) and
proximal(+)/distal(-) translation (right). Figure illustrates the

right hind limb ................................................................... 70

Vertical (above), longitudinal (middle) and transverse (below)

components of the ground reaction forces from hind limbs of
sound trotting horses. Thick solid line indicates mean values

xiii

Figure 3-2

Figure 3-3

Figure 3-4

Figure 3-5

Figure 3-6

Figure 3-7

Figure 3-8

Figure 3-9

Figure 3-10

from both hind limb. Thin solid lines indicate one standard
deviation above and below the mean ......................................... 77

Load distribution of vertical impulse of 4 sound horses.
Each dot represents the value from one horse ............................... 79

Joint angle, net joint moment and net joint power of coxofemoral

joint during the stance phase. Thick solid line indicates mean

values from both hind limbs of sound trotting horses. Thin solid lines
indicate one standard deviation above and below the mean ............... 84

Joint angle, net joint moment and net joint power of coxofemoral

joint during the swing phase. Thick solid line indicates mean

values from both hind limbs of sound trotting horses. Thin solid lines
indicate one standard deviation above and below the mean ............... 85

Joint angle, net joint moment and net joint power of femorotibial

joint during the stance phase. Thick solid line indicates mean

values from both hind limbs of sound trotting horses. Thin solid lines
indicate one standard deviation above and below the mean ............... 86

Joint angle, net joint moment and net joint power of femorotibial

joint during the swing phase. Thick solid line indicates mean

values from both hind limbs of sound trotting horses. Thin solid lines
indicate one standard deviation above and below the mean ............... 87

Joint angle, net joint moment and net joint power of tarsal joint

during the stance phase. Thick solid line indicates mean values

from both hind limbs of sound trotting horses. Thin solid lines

indicate one standard deviation above and below the mean. . . . . . . . 88

Joint angle, net joint moment and net joint power of tarsal joint

during the swing phase. Thick solid line indicates mean values

from both hind limbs of sound trotting horses. Thin solid lines

indicate one standard deviation above and below the mean. ............. 89

Joint angle, net joint moment and net joint power of
metatarsophalangeal joint during the stance phase. Thick

solid line indicates mean values fi'om both hind limbs of

sound trotting horses. Thin solid lines indicate one standard

deviation above and below the mean ......................................... 90

Joint angle, net joint moment and net joint power of
metatarsophalangeal joint during the swing phase. Thick
solid line indicates mean values from both hind limbs of
sound trotting horses. Thin solid lines indicate one standard

xiv

 

deviation above and below the mean ......................................... 91

Figure 3-1 1 Joint angle, net joint moment and net joint power of distal
interphalangeal joint during the stance phase.
Thick solid line indicates mean values from both hind limbs
of sound trotting horses. Thin solid lines indicate one
standard deviation above and below the mean .............................. 92

Figure 3-12 Joint angle, net joint moment and net joint power of distal
interphalangeal joint during the swing phase. Thick solid line
indicates mean values from both hind limbs
of sound trotting horses. Thin solid lines indicate one
standard deviation above and below the mean .............................. 93

Figure 4-1 Vertical and longitudinal ground reaction force data from
the right hind limb. The thick black line and dashed lines
indicate the mean value and one standard deviation above and
below the mean of the sound condition. The thick gray line
indicates mean of the lame condition. Dot represents variable
that differs significantly between sound and lame conditions.

a P<0.05. b 0.05<P<O.1 ........................................................ 126

Figure 4-2 Kinematic and kinetic variables of the right coxofemoral joint
during stance. The thick black line and dashed lines indicate
the mean value and one standard deviation above and below
the mean of the sound condition. The thick gray line indicates
mean of the lame condition. Dot represents variable that differs
significantly between sound and lame conditions

3 P<0.05. b 0.05<P<0.1 ......................................................... 127

Figure 4-3 Kinematic and kinetic variables of the right coxofemoral joint
during swing. The thick black line and dashed lines indicate
the mean value and one standard deviation above and below
the mean of the sound condition. The thick gray line indicates
mean of the lame condition .................................................... 128

Figure 4-4 Kinematic and kinetic variables of the right femorotibial joint
during stance. The thick black line and dashed lines indicate
the mean value and one standard deviation above and below
the mean of the sound condition. The thick gray line indicates
mean of the lame condition. Dot represents variable that
differs significantly between sound and lame conditions.

a P<0.05. b o.os<1><o.1 ........................................................ 129

XV

Figure 4-5 Kinematic and kinetic variables of the right femorotibial joint
during swing. The thick black line and dashed lines indicate
the mean value and one standard deviation above and below
the mean of the sound condition. The thick gray line indicates
mean of the lame condition. Dot represents variable that
differs significantly between sound and lame conditions.

3 P<0.05. b 0.05<P<0.1 ......................................................... 130

Figure 4-6 Kinematic and kinetic variables of the right tarsal joint
during stance. The thick black line and dashed lines indicate
the mean value and one standard deviation above and below
the mean of the sound condition. The thick gray line indicates
mean of the lame condition. Dot represents variable that
differs significantly between sound and lame conditions.

3 P<0.05. b 0.05<P<0.1 ......................................................... 131

Figure 4-7 Kinematic and kinetic variables of the right tarsal joint
during swing. The thick black line and dashed lines indicate
the mean value and one standard deviation above and below
the mean of the sound condition. The thick gray line indicates
mean of the lame condition. Dot represents variable that
differs significantly between sound and lame conditions.

3 P<0.05. 0.05<P<O.l ......................................................... 132

Figure 4-8 Kinematic and kinetic variables of the right metatarsophalangeal
joint during stance. The thick black line and dashed lines indicate
the mean value and one standard deviation above and below the
mean of the sound condition. The thick gray line indicates mean
of the lame condition. Dot represents variable that differs
significantly between sound and lame conditions.

3 P<0.05. b 0.05<P<0.1 ........................................................ 133

Figure 4-9 Kinematic and kinetic variables of the right metatarsophalangeal
joint during swing. The thick black line and dashed lines indicate
the mean value and one standard deviation above and below the
mean of the sound condition. The thick gray line indicates mean
of the lame condition. Dot represents variable that differs
significantly between sound and lame conditions.

a P<0.0S. b 0.05<P<0.1 ......................................................... 134

Figure 4-10 Kinematic and kinetic variables of the right distal interphalangeal
joint during stance. The thick black line and dashed lines indicate
the mean value and one standard deviation above and below the
mean of the sound condition. The thick gray line indicates mean
of the lame condition. Dot represents variable that differs

xvi

significantly between sound and lame conditions.
3 P<0.05. b 0.05<P<0.l ........................................................ 135

Figure 4-1 1 Kinematic and kinetic variables of the right distal interphalangeal
joint during swing. The thick black line and dashed lines indicate
the mean value and one standard deviation above and below the
mean of the sound condition. The thick gray line indicates mean
of the lame condition. Dot represents variable that differs
significantly between sound and lame conditions.

a P<0.05. b 0.05<P<0.l ........................................................ 136

Figure 4-12 Net joint moment profiles at the scapulohumeral, cubital, carpal,
metacarpophalangeal and distal interphalangeal joints averaged
from both forelimbs of sound trotting horses during the stance
phase. The heavy line is the mean curve and the two lighter
lines indicate one standard deviation ......................................... 138

Figure 4-13 Net joint moment profiles at the scapulohumeral, cubital, carpal,
metacarpophalangeal and distal interphalangeal joints averaged
from both forelimbs of sound trotting horses during the swing
phase. The heavy line is the mean curve and the two lighter
lines indicate one standard deviation ......................................... 139

Figure 4-14 Net joint power profiles at the scapulohumeral, cubital, carpal,
metacarpophalangeal and distal interphalangeal joints averaged
from both forelimbs of sound trotting horses during the stance
phase. The heavy line is the mean curve and the two lighter
lines indicate one standard deviation ......................................... 140

Figure 4-15 Net joint power profiles at the scapulohumeral, cubital, carpal,
metacarpophalangeal and distal interphalangeal joints averaged
from both forelimbs of sound trotting horses during the swing
phase. The heavy line is the mean curve and the two lighter
lines indicate one standard deviation ......................................... 141

Figure 4-16 Load distribution of vertical impulse of 4 horses. The black
dot represents the value in the sound condition. The gray dot
represents the value in the lame condition. Arrow represents
the direction of load distribution changes in each horse ................... 144

Figure 5-1 Three-dimensional rotation of the tarsal joint from
reference bone data during stance: flexion(-)/extension(+)
angle (top), abduction(-)/adduction angle (middle) and
internal(+)/external(-) rotation angle (bottom). The thick
black line indicates mean value from 4 horses. Thin lines
indicate one standard deviation above and below the mean.

xvii

Figure 5-2

Figure 5-3

Figure 5-4

Figure 5-5

Figure 5-6

Figure 5-7

Figure 5-8

Zero indicates the impact value ............................................... 163

Three-dimensional translation of the tarsal joint from

reference bone data during stance: cranial(+)/caudal(-)

translation (top), medial(+)/lateral(-) translation (middle) and
proximal(+)/distal(-) translation (bottom). The thick

black line indicates mean value from 4 horses. Thin lines

indicate one standard deviation above and below the mean.

Zero indicates the impact value ............................................... 164

Three-dimensional rotation of the tarsal joint from

reference bone data during swing: flexion(-)/extension(+)

angle (top), abduction(-)/adduction angle (middle) and
intemal(+)/external(—) rotation angle (bottom). The thick

black line indicates mean value from 4 horses. Thin lines

indicate one standard deviation above and below the mean.

Zero indicates the impact value ............................................... 165

Three-dimensional translation of the tarsal joint from

reference bone data during swing: cranial(+)/caudal(-)

translation (top), medial(+)/lateral(-) translation (middle) and
proximal(+)/distal(-) translation (bottom). The thick black

line indicates mean value from 4 horses. Thin lines

indicate one standard deviation above and below the mean.

Zero indicates the impact value ............................................... 166

Flexion (-)/extension (+) angle of horses 1-4 (top to bottom)

during stance. The gray line indicates angle obtained from skin

markers. The black line indicates angle obtained after application

of 3D skin correction algorithm. The dotted line indicates angle

obtained from bone pins. Zero indicates the impact value ................ 173

Abduction (-)/adduction (+) angle of horses 1-4 (top to bottom)

during stance. The gray line indicates angle obtained from skin

markers. The black line indicates angle obtained after application

of 3D skin correction algorithm. The dotted line indicates angle

obtained fi'om bone pins. Zero indicates the impact value ................ 174

Internal (+)/extemal (-) rotation angle of horses 1-4 (top to bottom)
during stance. The gray line indicates angle obtained from skin

markers. The black line indicates angle obtained after application

of 3D skin correction algorithm. The dotted line indicates angle

obtained from bone pins. Zero indicates the impact value ................ 175

Cranial (+)/caudal (-) translation of horses 1-4 (top to bottom)
during stance. The gray line indicates angle obtained from skin

xviii

 

markers. The black line indicates angle obtained after application
of 3D skin correction algorithm. The dotted line indicates angle
obtained from bone pins. Zero indicates the impact value ................ 176

Figure 5-9 Medial (+)/lateral (-) translation of horses 1-4 (top to bottom)
during stance. The gray line indicates angle obtained from skin
markers. The black line indicates angle obtained after application
of 3D skin correction algorithm. The dotted line indicates angle
obtained from bone pins. Zero indicates the impact value ................ 177

Figure 5-10 Proximal (+)/distal (-) translation of horses 1-4 (top to bottom)
during stance. The gray line indicates angle obtained from skin
markers. The black line indicates angle obtained after application
of 3D skin correction algorithm. The dotted line indicates angle
obtained from bone pins. Zero indicates the impact value. ............... 178

Figure 5-1 1 Three-dimensional rotation of the tarsal joint obtained from
all horses during stance: flexion(-)/extension(+) angle (top),
abduction(-)/adduction angle (middle) and internal(+)/external(-)
rotation angle (bottom). The black line indicates mean value from
corrected skin data. The dotted line indicates mean value from
reference bone data. Zero indicates the impact value ...................... 179

Figure 5-12 Three-dimensional translation of the tarsal joint obtained from
all horses during stance: cranial(+)/caudal(-) translation (top),
medial(+)/lateral(-) translation (middle) and proximal(+)/distal(-)
translation (bottom). The black line indicates mean value from
corrected skin data. The dotted line indicates mean value fi'om
reference bone data. Zero indicates the impact value ...................... 180

Figure 5-13 F lexion (-)/extension (+) angle of horses 1-4 (top to bottom)
during swing. The gray line indicates angle obtained from skin
markers. The black line indicates angle obtained after application
of 3D skin correction algorithm. The dotted line indicates angle
obtained from bone pins. Zero indicates the impact value ................ 181

Figure 5-14 Abduction (-)/adduction (+) angle of horses 1-4 (top to bottom)
during swing. The gray line indicates angle obtained from skin
markers. The black line indicates angle obtained after application
of 3D skin correction algorithm. The dotted line indicates angle
obtained from bone pins. Zero indicates the impact value ................ 182

Figure 5-15 Internal (+)/external (-) rotation angle of horses 1-4 (top to bottom)
during swing. The gray line indicates angle obtained from skin
markers. The black line indicates angle obtained after application
of 3D skin correction algorithm. The dotted line indicates angle

xix

Figure 5-16

Figure 5-17

Figure 5-18

Figure 5-19

Figure 5-20

Figure A-l

Figure A-2

obtained from bone pins. Zero indicates the impact value ...............

Cranial (+)/caudal (-) translation of horses l-4 (top to bottom)
during swing. The gray line indicates angle obtained from skin
markers. The black line indicates angle obtained after application
of 3D skin correction algorithm. The dotted line indicates angle

obtained from bone pins. Zero indicates the impact value ...............

Medial (+)/lateral (-) translation of horses 1-4 (top to bottom)
during swing. The gray line indicates angle obtained from skin
markers. The black line indicates angle obtained after application
of 3D skin correction algorithm. The dotted line indicates angle

obtained from bone pins. Zero indicates the impact value ...............

Proximal (+)/distal (-) translation of horses 1-4 (top to bottom)
during swing. The gray line indicates angle obtained from skin
markers. The black line indicates angle obtained after application
of 3D skin correction algorithm. The dotted line indicates angle

obtained from bone pins. Zero indicates the impact value ...............

Three-dimensional rotation of the tarsal joint obtained

from all horses during swing: flexion(-)/extension(+) angle (top),
abduction(-)/adduction angle (middle) and internal(+)/extemal(-)
rotation angle (bottom). The black line indicates mean value from
corrected skin data. The dotted line indicates mean value from

reference bone data. Zero indicates the impact value .....................

Three-dimensional translation of the tarsal joint obtained from
all horses during swing: cranial(+)/caudal(-) translation (top),
medial(+)/lateral(-) translation (middle) and proximal(+)/distal(-)
translation (bottom). The black line indicates mean value from
corrected skin data. The dotted line indicates mean value from

reference bone data. Zero indicates the impact value .....................

Lateral view of the tibia (left) and third metatarsus (right)
of a right limb showing the locations of the skin surface

markers relative to the underlying bones ...................................

Skin displacements for each coordinate of each tibial marker.

The data are generated from the truncated Fourier series models.
The dashed and dotted lines are from models fit to the three
individual horses, while the thick solid line is from the model based
on the pool of 12 trials from subjects 1-3. The columns represent
the x, y and z coordinates respectively. Each row represents a

marker with the top row being TIB-A and bottom row being TIB-B. . ..

XX

.183

.184

.185

.186

.187

.188

.227

228

Figure A-3

Figure A-4

Figure A-S

Figure A-6

Figure A-7

Skin displacements for each coordinate of each third metatarsal

marker. The data are generated from the truncated Fourier series

models. The dashed and dotted lines are from models fit to the three
individual horses, while the thick solid line is from the model based

on the pool of 12 trials from subjects 1-3. The columns represent the

x, y and z coordinates respectively. Each row represents a marker

with the top row being MET-B and bottom row being MET-A .......... 229

Comparison of skin displacement models for the proximal and distal

ends of the tibia and third metatarsus. The solid lines are from the

models for the TIB-A, TIB-B, MET-A and MET-B markers from

the current study. The open dotted lines are from the models given in

van Weeren (van Weeren, et al., 1992) for the corresponding locations.
The first column is the displacement data for the x-axis and the

second column is the displacement data for the z-axis. The data

are given as a percentage of segment length ................................. 230

Angular kinematics of the tibia and third metatarsus with

respect to the global coordinate system. The solid lines are from

the bone-fixed kinematics and the dotted lines are from corrected

skin kinematics. Data are a mean of the 12 trials from subjects

1-3, normalized to percent of stride .......................................... 231

Kinematics of the tarsal joint (motion of the third metatarsus relative

to the tibia). The right hand column is the angular data (in degrees)

and the left hand column is the displacements (in mm). The solid

lines represents bone-fixed kinematics, the filled dotted lines

represents corrected skin kinematics, and the open dotted lines are
kinematics from uncorrected skin markers. Data are a mean of

the 12 trials from subjects 1-3 ................................................. 232

Application of correction models to kinematics of the tarsal joint

(motion of the third metatarsus relative to the tibia) in absolute

values and referenced to the start of the stride. The right hand

columns are the angular data (in degrees) and the left hand columns

are the displacements (in mm). The solid lines represents bone-fixed
kinematics from subjects 1-3, the filled dotted lines represents

corrected skin kinematics from subject 4, and the open dotted

lines are kinematics from corrected skin markers of subject 5. . . . . . . . 233

xxi

INTRODUCTION

RATIONALE

Horses have played an important role in the development of human civilization
(van Weeren, 2001). Unlike most of the other domesticated species, horses were used to
transport people, to work the land and as vehicles in wartime. These uses developed due
to the horse’s outstanding locomotor capabilies. Today, some horses are still used for
transportation but their primary use is in equestrian sports, such as racing, dressage,
jumping, polo and rodeo.

In sporting horses, lameness in either the forelimb or the hind limb will limit the
horse’s ability to perform to a greater or lesser degree. Lameness can be observed while
the horse is in motion. Characteristics of the stride, such as the stance and swing phases,
the arc of foot flight, how the foot lands, joint flexion angles and symmetry of head and
hip motion, are used for identifying the lame limb (Stashak, 1987). However, human eyes
have limited temporal resolution, subclinical lameness is difficult to detect because the
asymmetries are subtle and the evaluation is considered subjective. In an attempt to
enhance the ability to evaluate lameness in horses, gait analysis has been used to quantify
gait asymmetries.

Biomechanics is the application of mechanical laws to living structures. For
several decades, biomechanical principles have been applied in human locomotion, both
for studying clinical aspects of gait and for exploring methods of improving performance
in sport. Recently, the techniques for biomechanical analysis have been adapted to study

locomotion in horses. The technique of gait analysis includes analysis of movement

(kinematics) and the forces responsible for movement (kinetics). A further step in the
procedure, inverse dynamic analysis, combines kinematic and kinetic measurements with
inertial properties of the limb segment to analyze the mechanical work performed across
the joint. The advantage of inverse dynamic analysis is that it calculates variables that
cannot be measured directly, including net joint moments, net joint powers and net joint
energies. The net joint moment represents the net torque produced by the action of soft
tissues (muscles, tendons and ligaments) around the representative center of rotation of
the joint. Net joint power is the rate of net mechanical work done by the soft tissues
around the joint. Net joint energy is the amount of mechanical work done across a joint
during a period of time. Similar patterns of limb movement may be the result of different
combinations of muscular activity and these are demonstrated by inverse dynamic
analysis. It gives information describing the amount of energy each joint absorbs
(negative work), and the amount of energy each joint generates (positive work) to move
the limb and body. The ability to calculate kinetic variables around a joint enhances the
ability to detect abnormal patterns of these variables in response to pain and thus
facilitates our understanding of the underlying mechanisms of locomotor compensation

for lameness.

PROBLEM STATEMENT

Osteoarthritis of the distal intertarsal and tarsometatarsal joints (bone spavin) is
one of the common causes of hind limb lameness in the horse. Estimates of prevalence of
bone spavin range from 27% (Gabel, 1983) to more than one-third (Gough and Munroe,

1998) of cases with hind limb lameness. The prevalence of radiographic signs of bone

spavin in Icelandic horses was found to be 23% in Sweden (Eksell, et al., 1998) and
30.3% in Iceland (Bjomsdottir, et al., 2000). It is regarded as one of the most important
diseases in sport horses (Willms, etal., 1996) as well as in racehorses (Moyer, et al.,
1983). The problem is usually subtle and it may be difficult to identify the improvement
after diagnostic local anesthesia (Moyer, et al., 1993). There are different levels of
correlation between clinical evaluation and radiographic findings, from poor correlation
(Gough and Munroe, 1998) up to 73% agreement (Axelsson, et al., 1998). It would be
useful to have a more reliable modality to diagnose osteoarthritis in the distal intertarsal
and tarsometatarsal joints. High detail radiography has been used to detect subchondral
bone plate irregularities, joint margin changes and sclerosis in the distal tarsal joints of
weanling, young and adult horses. High detail radiography may enhance the ability to
find radiographic changes that could not be detected by clinical radiography, especially
subchondral bone plate irregularities and joint margin changes (Laverty, et al., 1991).
Nevertheless the evaluation should be supplemented with other ancillary aids. Nuclear
scintigraphy is probably the most accurate modality to detect active turnover of
subchondral bone. The ability of cartilage to attenuate and transfer loading forces to
subchondral bone alters with degeneration. As a result, subchondral bony structures are
subjected to more load, resulting in microfracture. In an attempt to repair the
microfi‘acture, metabolic activity increases shortly after injury, and long before the
sclerotic change detected by radiography. Therefore, nuclear scintigraphy is a promising
modality to evaluate the early stage of distal tarsal osteoarthritis. The effects of multiple
injections of radiopharrnaceutical agent on the horse’s long-term health have not yet been

identified, so repeated evaluation in horses suspected to have osteoarthritis may be a

concern. Gait analysis, which requires markers attached on the skin and force plate
evaluation, is a non-invasive technique and it can be used as frequently as needed. The
kinematic and kinetic analysis of the tarsal joint may be useful for detection early signs of
disease and for monitoring the response to treatment.

To date, there has been a limited number of publications describing net joint
moments, net joint powers and net joint energies in horses. In the hind limb, the only
reports of inverse dynamic analysis have described normal horses at the walk (Clayton, et
al., 2001; Khumsap, et al., 2001a). Before using this technique as a clinical tool, however,
kinematic and kinetic data describing the normal function of the tarsal joint must be
known. The gait used in lameness evaluation is the trot, which implies a need for
establishing normative profiles at the trot. Once the information from normal tarsal joints
has been obtained, it can be used as a basis for evaluating deviations from the normal
pattern in specific types of lameness, such as pain due to osteoarthritis of the distal
intertarsal and tarsometatarsal joints.

Movements of the limb joints distal to the shoulder and hip in normal horses are
primarily those of flexion and extension, and consequently, the published kinematic
studies have focused on two-dimensional analysis in the sagittal plane. However,
complex joints, such as the tarsal joint, may not function simply as a hinge joint.
Measurable amounts of motion from percutaneous pins inserted into the tibia and
metatarsus have shown that tarsal joint motion occurs in the frontal (abduction and
adduction) and transverse (axial rotation) planes. Motion also appears to occur outside

the tarsocrural joint (Lanovaz, et al., 2002). The reference bone data have been used to

develop skin correction algorithms (Appendix A) to increase the accuracy of skin marker

based data describing three-dimensional motion of the normal equine tarsal joint.

OBJECTIVES

A pre-requisite to the use of gait analysis as a clinical tool, normative values for
the kinematics, ground reaction forces, net joint moment, net joint power and net joint
energy profiles in sound horses should be established. Once the normative profiles have
been identified, they can be used as a basis to evaluate the effects of a specific type of
lameness. The establishment of non-invasive techniques to evaluate the motion of the
tarsal joint complex outside the sagittal plane will be useful in the evaluation of sport
horses. The objectives of the study were to:
1. Study the two-dimensional kinematic and kinetic analysis of the equine hind limb in
sound trotting horses.
2. Study the effect on locomotion of unilateral synovitis in the distal intertarsal and
tarsometatarsal joints, including compensations within the lame limb (intra-limb) and
between different limbs (inter-limb).
3. Study the three-dimensional kinematic analysis of the tarsal joint complex in sound
trotting horses and the effect of distal tarsal synovitis on three-dimensional kinematics of

the lame tarsal joint.

HYPOTHESES
1. Kinematic and kinetic variables of the hind limbs in sound horses have a similar

pattern in different horses after adjustment for the effect of subject velocity and body

mass.

2. Kinematic and kinetic profiles of horses affected with distal tarsal synovitis differ fiom
the profiles obtained from sound horses.

3. Tarsal joint complex has the motion outside the sagittal plane which can be identified

using skin markers.

OVERVIEW

In this dissertation, the primary interest was to firrther knowledge of tarsal joint
function in sound horses and in horses with pathological conditions of the distal
intertarsal and tarsometatarsal joints. Data were obtained from a group of normal horses
using retro-reflective markers attached to the skin. The effects of unilateral lameness
located at the distal intertarsal and tarsometatarsal joints were studied by the induction of
synovitis in these joints in one hind limb. Chapter 1 describes a literature review of the
clinical applications of gait analysis. Chapter 2 describes the materials and methods,
including the kinematic and force plate techniques, and the establishment of segmental
coordinate systems for three-dimensional analysis of the equine tarsal joint complex.
Chapter 3 addresses Hypothesis 1, by reporting the two-dimensional kinematics, ground
reaction forces, net joint moments, net joint powers and net joint energies of the equine
hind limbs in sound trotting horses. Chapter 4 addresses Hypothesis 2, by reporting the
biomechanical profiles in horses affected with distal intertarsal and tarsometatarsal
synovitis and identifies differences from sound horses. This chapter includes a
description of the compensatory effects found within the limb and between the limbs.

Chapter 5 addresses Hypothesis 3, by reporting three-dimensional kinematics of the

equine tarsal joint complex in sound horses. This chapter also includes an evaluation of a
3D skin marker set to obtain 3D tarsal joint motion and effect of the distal intertarsal and
tarsometatarsal synovitis on 3D motion. The last section of this dissertation is a

conclusion of the results.

 

CHAPTER 1

EQUINE GAIT ANALYSIS: REVIEW OF CLINICAL APPLICATIONS

This chapter includes five main sections of the literature review. The first section
is the brief history of gait analysis in horses, including the development of equipment and
techniques for gait analysis in horses. The second section is a review of gait analysis in
sound horses, including the characteristics of normal kinematic and ground reaction
forces in trotting horses. The third section reviews the application of gait analysis in
varying aspects, including the use of motion analysis and a force plate to evaluate
abnormalities in locomotion and to monitor the effect of treatment. The fourth section is a
review of the inverse dynamic technique and its application in normal and lame horses.

The last section reviews three-dimensional kinematic analysis in horses.

HISTORY OF GAIT ANALYSIS IN HORSES

The early work in this modern era on gait analysis in horses began in the late
18005. A major contribution was Muybridge’s work in 1877 using a battery of still
cameras to capture the horse in different gaits and activities. His work triggered the
interest of many researchers to further analyse locomotion using photographic
techniques. When Leach and Dagg (1983) reviewed the research on equine locomotion
and biomechanics, it was stated that the understanding of locomotion and structure and
function of the musculoskeletal system was far fiom complete. The techniques of

lameness diagnosis had not really improved since the 17th century, and the concepts of

 

how a horse adapted to lameness in one or more limbs was limited. They stated three
areas of concern to the veterinarian; 1. research must be directed at understanding and
distinguishing the factors influencing locomotion; 2. clinical evaluation of lameness must
be based on scientifically validated information; and 3. the effects of treatment on
locomotion must be more thoroughly documented. Leach and Crawford (1983) proposed
six major research areas for equine locomotion: 1. characteristics of normal locomotion;
2. factors influencing locomotion; 3. factors that can be used to predict performance and
thereby aid in the selection of a prospective athlete; 4. factors predisposing to injury; 5.
epidemiology and economics of lameness; and 6. clinical identification and management
of lameness. To facilitate gait studies, a standard terminology for analysis of equine
locomotion was suggested by Leach, et al. (1984) and revised in the First International
Workshop on Animal Locomotion (Leach, 1993).

A review by Dalin and Jeffcott (1985) stated the limitations of kinematic and
kinetic analysis as a routine clinical tool. The availability of high-speed cinematography
brought many advances in kinematic analysis, particularly in measuring temporal
variables. The delay from recording to analysis was a major limiting factor in the
practical application of this technique. Video analysis allowed immediate play back, but
it gave poor spatial and temporal resolution, which caused problems when measuring
small distances, small angular movements or short duration events. There have been
marked improvements in temporal resolution and some improvement in spatial resolution
of video systems in recent years but, high-speed video systems are expensive.

In the 19808, automatic motion analysis systems became available for kinematic

measurement and analysis (Leach, 1987). Most of these systems used infrared light,

 

either reflected from passive markers, or emitted by light emitting diodes attached to the
limbs. Due to their sensitivity to light, most of these systems could not be used in
daylight. Digital infi'ared cameras with excellent spatial and temporal resolution are now
available for incorporation into fully automated gait analysis systems that offer real time
digitizing and three-dimensional display of markers.

The requirement that markers be attached to the skin gives rise to measurement
error because of skin movement relative to the underlying bony structures (Leach, 1987).
In an attempt to reduce the errors due to skin displacement, van Weeren (1989)
developed models to correct for two-dimensional skin displacement in common marker
locations. His thesis reported skin correction algorithms for the fore and hind limbs at
walk and trot, which have been used in several studies to improve the accuracy of data
describing the motion of limb segments from skin markers.

Force plate analysis is the method of choice for rapid and accurate measurement
of ground reaction forces (GRFs). Data collection requires a single hoof to hit the force
plate at a time. The early force plates were small, and it many runs were usually required
to obtain one successful hoof hit at the trot, which greatly limited its use. Recently, larger
force plates have been marketed, enabling a reduction in the ratio of unsuccessful to
successful trials (Clayton, 1996). A good force plate hit was obtained every five passes at
the trot on a force plate measuring 60 x 60 cm (Merkens, et al., 1993).

The availability of video-based motion analysis and large force plates, in
combination with advanced computer technology has greatly facilitated progress in
equine gait analysis. Several studies have been performed to identify normal kinematic

and kinetic patterns, and to study pathological patterns obtained from lame horses. The

10

most ambitious aim of clinical biomechanics in horses is computer diagnosis of gait
deficits. Analysis of multiple kinematic or kinetic characteristics might allow early and

objective diagnosis of a lameness (Buchner, 2001a).

GAIT ANALYSIS IN SOUND HORSES

One of the major research areas for equine locomotion according to Leach and
Crawford (1983) is the characteristics of normal locomotion. In sound horses, gait
analysis techniques have been used to obtain quantitative measurements of kinematic and
kinetic data, as an aid to increase the understanding animal movement patterns and force
generation to move the limb. In a clinical setting, the goal is to use this information as

normative data for comparison it with data fi'om different types of lameness.

Kinematic analysis. Kinematic analysis is a tool for quantifying movements. The
output is in the form of temporal, linear, and angular measurements that describe the
movements of the body segments and joint angles (Clayton, 1996). Patterns of limb
motion have been quantified and described. For purposes of lameness evaluation, the trot
is the gait of choice, so this section will focus on the trot. In trotting horses, the forelimb
and hind limb rotate like pendulum around rotation point, at the proximal part of scapula
and at the acetabulum (Back, et al., 1995a; Back, et al., 1995b). During the stance phase,
the limb shortens as it accepts load. In the forelimb, the shortening in early stance was
due to hyperextension of the carpal joint and shortening of distance from the cubital joint

to the hoof (Hjerten and Drevemo, 1993). In the hind limb, shortening was due to the

11

 

flexion of the femorotibial, tarsal and distal interphalangeal joints and extension of the
metatarsophalangeal joint (Hjerten, et al., 1994).

Clinical evaluation is based on observation of abnormal or asymmetrical motion
between left and right forelimb or hind limb pairs at the trot. It is important, therefore, to
evaluate the degree of symmetry using kinematic analysis in trotting horses. In
assessment of locomotion symmetry (Pourcelot, et al., 1997 a; Pourcelot, et al., 1997b),
skin markers were attached over anatomical landmarks of the forelimb and hind limb
pairs of sound horses. Joint angle motion was found to be more symmetrical than vertical
displacement of the markers overlying the joint centers of rotation. F urtherrnore, vertical
displacement of the proximal markers at the trunk was less symmetrical than that of the
markers overlying the distal limb joints. Comparing the symmetry of proximal markers
between forelimbs and hind limbs, the hind limbs had a lower level of vertical
displacement symmetry than the forelimbs. The asymmetry in vertical displacement of
the tuber coxae can be detected in sound trotting horses (Buchner, et al., 19960). This is
probably due to the ability to move the vertebral column in flexion/extension, rotation
and lateral bending. Since the forces generated by muscles are used to accelerate the
body, the symmetry can be assessed by comparing acceleration of the point of interest on
the left and right sides of the body, such as the acceleration quotient of the tuber coxae
(Buchner, et al., 1993). Perfect symmetry in sound horses should yield a quotient of 1.
That study reported a range of acceleration quotients from 1.03 to 1.54 in sound horses.
This supported the idea that asymmetrical motion of the tuber coxae occurred in sound
trotting horses, which is due to the different swinging characteristics of the coxofemoral

joints and the flexibility of the vertebral column of the individual horses.

12

Movement of body center of mass (BCM) in sound horses has been studied
(Buchner, et al., 2000). BCM is defined as the centroid of all mass elements of an object.
The position of BCM changes according to the orientation of all body segments. A
knowledge of BCM position can be used to study the load distribution on the limbs and
the external work of locomotion. At the trot, observation from the side of the horse
showed that the BCM moves down and slightly cranially at the beginning of stance, then
caudally and down until the middle of stance. After that, it started to move upward and
cranially to the end of stance. When observed in a frontal view, the BCM has a slight
medic-lateral movement toward the weight bearing forelimb during the first half of
stance and then moves back toward the midline in terminal stance. The displacement
magnitude of the BCM is significantly smaller than the externally visible trunk
movements, which contributes to the high locomotor efficiency of horses.

The characteristics of hoof landing were studied and indicated differences
between the fore and hind hooves. F ore hooves had higher vertical hoof velocity and
acceleration at impact than hind hooves, while hind hooves had higher horizontal velocity
and acceleration than fore hooves (Back, et al., 1995c; Johnston, et al., 1996). Therefore,
the fore hoof bounces more at impact, whereas the hind hoof slides into the ground.

Several factors may influence the kinematic variables, notably subject velocity. At
the walk, increasing velocity was associated with increasing frequency and constant
stride length. At the trot, increasing velocity was associated with variable alterations in
stride length and stride frequency (Rooney, et al., 1991). As velocity increases, there are
increases in retraction-protraction range of motion, tarsal joint flexion and tarsal and

coxofemoral joint ranges of motion (Galisteo, et al., 1998). Training also affects the

13

kinematic variables. Elite trained horses showed different kinematics compared to
untrained horses, including a decrease in percent stance duration, and several differences
in forelimb joint angle variables, but minimal differences in hind limb joint angle
variables (Morales, etal., 1998). Kinematics also changes with farriery modifications.
Horses with acute angulation of the hind hooves had increases in breakover time,
overreach distance and overreach duration (Clayton, 1990). Each horse from a
homogenous population adopts a constant angular pattern that is very close to that of
other horses (Degueurce, et al., 1997). Variation can be reduced by comparing horses

from a homogeneous population under similar conditions.

Kinetic analysis. The force plate is an accurate tool to evaluate ground reaction
forces (GRFs). The ground reaction forces are recorded and resolved into three
orthogonal components: vertical, longitudinal and transverse. The point of force
application to the hoof (center of pressure) is also recorded. The vertical force component
represents the force of the body mass that loads the limb. The longitudinal component
may be positive or negative, representing propulsion and braking forces, respectively.
During the first half of stance, the negative longitudinal force decelerates the forward
motion. In the second half of stance, the positive longitudinal force acts to propel the
body forward. The transverse component represents the force that turns the body to left or
right. When the horses move in straight line, this force component is very small. The
variables obtained from force plate analysis are peak forces and times of their
occurrences. Force impulses, calculated by time integration of the areas underneath the

force curves, represent the total amount of force applied over a period of time.

14

The advantage of force plate analysis is the relatively easy instrumental set up and
data collection procedures. The limitation is the need for several trials to collect sufficient
good hits. Probably due to the relatively small force plate, which required a large number
of attempts to obtain one successive trial, the early studies that established the standard
GRF of normal horses was performed in walking horses (Merkens, et al., 1985; Merkens,
et al., 1988; Merkens and Schamhardt, 1988). At the walk, the magnitudes of vertical
peak forces were higher in the forelimbs than the hind limbs. The magnitudes of the
longitudinal force peaks and impulses were similar in the forelimbs and hind limbs,
indicating an equal contribution to deceleration and acceleration of each of the four limbs
during walking. At the trot, the mean vertical forces were highly correlated with the body
weight (Barr, et al., 1995). The peak longitudinal force is around 1/10 of the peak vertical
force and the peak transverse force is very small, about 1/ 100 of the peak vertical force
(Seeherman, et al., 1987). The magnitudes of peak vertical force, vertical impulse and
peak braking force were greater in the forelimbs than in the hind limbs, but the peak
propulsive force was higher in the hind limbs (Morris and Seeherman, 1987). In the
forelimb, peak force and impulse were higher for the braking component than the
propulsive component, whereas the opposite was found in the hind limbs (Merkens, et al.,
1993). This indicates the important role of the forelimb to decelerate the body, and the
hind limb to accelerate and propel the body forward at the trot. Transverse forces were
inconsistent and corresponding within-subject variation was considered too high for a
meaningful further analysis. The GRF variables are consistent and show little variation

within horses when data are collected on different days (Seeherman, et al., 1987).

15

In the analysis of equine locomotion, identification of the times of hoof impact
and lift off are require to calculate temporal stride variables, such as stance duration,
swing duration and stride duration. Force plate analysis proved to be an accurate method
to identify the time when the hoof was on the ground (Scharnhardt and Merkens, 1994).
In the laboratories that do not have a force plate, an alternative method is to use
accelerometer attached to the hoof. At impact, the sudden change of hoof acceleration is
fairly easy to detect but at the lift off hoof acceleration is less abrupt, so the
accelerometer may not determine lift off time correctly. Recently, a new method has been
reported (Peham, et al., 1999b) in which stance duration at the walk obtained from high
speed (240 frames/s) kinematic analysis was highly correlated with that obtained from a
force plate.

Subject velocity influences the magnitude of the GRF. When walking velocity
increases, there is an increase in peak vertical force but a decrease in vertical impulse on
the hind limb (Khumsap, et al., 2001b). In the forelimb, an increase in walking velocity
results in an increase in peak vertical force and decreases in vertical impulse and braking
impulse (Khumsap, et al., 2002). In the forelimbs, higher trotting speed may be
associated with greater vertical (Barr, et al., 1995) and braking (McLaughlin, etal., 1996)
forces. Therefore, when comparing information among different horses or assessing the
effect of different conditions, subject velocity is an important factor that needs to be

controlled.

l6

GAIT ANALYSIS IN LAME HORSES

One of the goals of studying gait analysis in horses is to apply the technique as a
research or clinical tool to evaluate lameness. After the normal patterns of gait have been
identified, they can be used as a standard for comparison with pathological patterns
obtained from lame horses. Comparison of normal and abnormal functions of selected
joints and limb segments enhances the understanding of gait pathology and offers a
method of monitoring the outcome of treatment. Gait analysis principles have been
applied to evaluate lameness in a variety of disciplines. The studies on lameness in horses
have been performed either on naturally occurring lameness or, more often, on induced-
lameness. The induced-lameness provides less variability of the site and degree of
lameness in the limb, which facilitates comparisons of gait variables between lame and
sound horses, and may be able to assess specific lame patterns or to compare between

lamenesses at the different sites in the limb.

Kinematic analysis in lame horses. Mild lameness at the trot may be difficult to
assess subjectively. A study comparing subjective evaluation and kinematic analysis of
mild lameness at the trot before and after palmar digital nerve block showed that
experienced clinicians had higher within-observer agreement of lameness improvement
than interns or residents, but they still had more difficulty detecting the true state of
lameness than kinematic analysis (Keegan, et al., 1998b). Another study indicated similar
findings in assessment of mild forelimb lameness. A clinician and a motion analysis

system assigned the lameness to the same limb in all cases but grading of the lameness

l7

 

differed in 6 out of 29 cases (Peham, et al., 1999a). Therefore, kinematic analysis can be
used as an informative tool supporting subjective veterinary judgment.

A kinematic symmetry index has been developed to evaluate the magnitude of
motion of the forelimb and hind limb pairs in horses. In lame horses, vertical
displacement indices of anatomical landmarks generally have lower symmetry but are
more sensitive to lameness than the joint angle indices (Pourcelot, et al., l997a). In a
group of horses with unilateral fore or hind limb lameness, the most sensitive lameness
indicators were the symmetry indices of the vertical displacement of the proximal
markers of the limbs (Audigie, et al., 2001 ). Asymmetry in proximal marker
displacement might reflect supporting limb lameness, while asymmetry in distal marker
displacement might reflect swinging limb lameness. A limitation to the use of symmetry
indices is that they do not determine the lame side and are unable to detect bilateral
lameness (Pourcelot, et al., l997a).

Each lame horse has an optimum trotting speed that reveals the maximum
asymmetrical movement (Peham, et a1, 1998). In the forelimb, lameness assessed by
asymmetry of vertical head motion increased significantly with trotting speed in
moderately lame horses, while there was no significant correlation between speed and
asymmetrical motion in mild or subclinical forelimb lameness (Peham, et al., 2000).
Therefore, moderate forelimb lameness measured as head motion asymmetry depends on
the speed at which the measurements are taken. Measurements taken at two trotting
speeds can be used to construct an individual regression calculation to standardize the

magnitude of head motion at any speed within the trotting range.

18

A group of horses with fore or hind limb lameness was observed kinematically
before and after diagnostic blocks to assess the compensatory movement patterns (Uhlir,
et al., 1997). Asymmetrical acceleration of the head and sacrum indicates that horses with
a unilateral supporting forelimb lameness showed a false supporting lameness in the
contralateral hind limb, while horses with a true hind limb lameness showed a false
lameness in the ipsilateral forelimb. It was suggested that a horse that appears to have
ipsilateral fore and hind limb lameness is likely to have a true hind limb lameness, and a
horse that appears to have a forelimb and contralateral hind limb lameness is likely to
have a true forelimb lameness. This may be useful for planning diagnostic blocks in
lameness evaluation.

Horses with induced fore or hind hoof lameness had compensatory patterns both
within and between the limbs (Buchner, et al., 1996b). When lameness presented at the
hoof, there were decreases in joint extension of the
metacarpophalangeal/metatarsophalangeal joint and in joint flexion of the distal
interphalangeal joint on the lame limb at the trot during stance. The contralateral limbs
had compensatory increases in metacarpophalangeal/metatarsophalangeal joint extension
and distal interphalangeal joint flexion. In the proximal joints, such as scapulohumeral
and tarsal joints, there were increases in joint flexion with increasing lameness degree,
indicating that these joints were acting as load dampers when there was a pain in the
hoof. Vertical motion of the trunk, represented by the withers for forelimb lameness and
the tuber sacrale/tuber coxae for hind limb lameness was smaller during stance of the

lame limb in the lame condition than in the sound condition (Buchner, et al., 1996c). The

19

trunk was supported earlier at the start of the lame limb stance phase to reduce peak
loading and save the horse energy to lift the body mass into swing.

Front hoof lameness induced by pressure on the frog area was associated with a
decrease in peak metacarpophalangeal joint extension during stance, and increases in
head vertical displacement and asymmetrical head movement at the trot. A similar
lameness model using pressure at the toe area resulted in similar changes in
metacarpophalangeal kinematics and head movement, together with the earlier
occurrence of peak metacarpophalangeal joint extension, maximum hoof height near mid
swing and an increase in maximum limb protraction (Keegan, et al., 2000). Kinematic
analysis may enable differentiation between lameness at the heel and toe.

Evaluation of horses at the trot with amphotericin-induced carpal synovitis at 9,
l6 and 23 days after induction, indicated that stride length, stance duration and forelimb
abduction were not changed. Carpal joint range of motion on the sound limb and
asymmetrical motion of the withers during stance phases of sound and lame forelimbs
significantly decreased only at 9 days after the induction. The metacarpophalangeal joint
range of motion in the lame limb and asymmetrical head excursion during stance phases
of sound and lame forelimbs significantly decreased at all measurement times compared
to baseline data before lameness induction (Peloso, et al., 1993b). Head excursion was
the most consistent variable for assessment of carpal lameness in the horse.

Another aspect of kinematic analysis assesses the compensatory movement
patterns of the body center of mass (BCM). Lameness induction by pressure-induced
mild or moderate pain on the sole of the fore hoof has been used to study the movement

of BCM (Buchner, et al., 2001). During the stance phase of a moderate forelimb

20

lameness, the magnitude of vertical BCM movement decreased by 34% and vertical
BCM acceleration decreased by 21% compared with the sound condition. During the
stance phase of the contralateral forelimb, vertical BCM movement increased by 9%. In
the medic-lateral direction, there was a mean shift of BCM motion toward the sound
forelimb. The caudal movement of the BCM during the lame diagonal stance was only 9
mm, representing a minor compensatory effect. Marked changes in vertical acceleration
of BCM during the lame diagonal stance resulted from adjustment of head and neck
movement, which creates a torque on the trunk that changes its momentum
(Vorstenbosch, et al., 1997). The effect is an increase in loading of the sound forelimb
and unloading of the diagonal hind limb. In the lame diagonal stance, the forelimb was
unloaded and the load shifted to the diagonal hind limb without changing the BCM
position.

Horses with various hind limb lamenesses of different degrees had increased
vertical displacement of the tuber coxae in the lame hind limb compared to the
contralateral hind limb, and this is probably the best criterion for the diagnosis of hind
limb lameness (May and Wyn-Jones, 1987). Since the same amount of vertical
displacement can be achieved with different vertical velocities, vertical acceleration of
the tuber coxae may represent the effects of lameness in a more suitable way (Buchner, et
al., 1993). The maximum and minimum peaks of the coxofemoral acceleration were used
to calculate the coxofemoral acceleration quotient. In the lame horses the acceleration
quotient ranged from 1.32 to 2.96, which overlapped the value of 1.03 to 1.54 in sound
horses. It has been stated that horses with an acceleration quotient greater than 1.61 can

be assessed as lame with 95% confidence (Buchner, et al., 1993).

21

The effects of unilateral lameness induction in the distal intertarsal and
tarsometatarsal joints, and the improvement after infra-articular anesthesia were evaluated
kinematically. After lameness induction, tarsal joint flexion during stance increased,
while metatarsophalangeal joint extension during stance and the limb protraction distance
decreased (Kramer, et al., 2000). Vertical excursion of the tuber coxae became more
asymmetric than in the sound condition. After intra-articular anesthesia, limb protraction
returned to pre-lameness values, and vertical excursion of tuber coxae became more
symmetric than in the lame condition. Hind limb protracting distance may be one of the
more sensitive indicators of the severity of tarsal joint lameness. Evaluation of limb

protraction is assessed when observing the horse from the side.

Kinetic analysis in lame horses. Force plate analysis is a non-invasive tool for
measuring GRFs, which can be used to aid in the localization of a subtle lameness to a
particular limb and to evaluate the effect of diagnostic regional anesthesia, particularly in
proximal limb lameness where regional anesthesia is rarely 100% effective and small
improvements in weight bearing are considered a positive response (Seeherman, et al.,
1987)

In a study comparing clinical examination with force plate analysis to evaluate
horses with mild lameness, in 15 out of 22 cases the same limb was identified but there
was no correlation between the degree of lameness assigned by the two methods
(W eishaupt, et al., 2001). Therefore, assessment of asymmetry in peaks force or impulses
in fore or hind limb pairs may complement the clinical evaluation in difficult or mild

C3868.

22

Perineural and intra-articular anesthesia are useful diagnostic aids to localize the
site of lameness within a limb. The method is based on the assumption that the technique
itself does not change the movement pattern. Force plate analysis of normal horses after
palmar digital nerve block, abaxial sesamoid nerve block and high palmar nerve block
did not change the clinical score or the GRF variables (Keg, et al., 1996). This confirms
that local anesthesia does not alter the loading pattern in healthy horses.

At the trot, horses with an induced full-thickness articular cartilage defect in the
carpal joint had decreases in peak vertical force, peak braking force and the
corresponding impulses in the lame forelimb compared to the values obtained before the
operation. The contralateral forelimb had no change in peak vertical force, but there were
increases in peak braking force and impulse. The peak vertical force, peak propulsive
force and corresponding impulses in the ipsilateral hind limb also decreased. The stance
time on all four limbs did not change (Morris and Seeherman, 1987). Changes in GRF
variables within the lame limb and redistribution of forces between the other limbs
indicates the potential for secondary injury resulting from compensation for a specific
type of lameness.

Measurement of the repeatable peaks of vertical and longitudinal forces and the
slope between the peaks has been used to assess subclinical tendon injury (Dow, et al.,
1991). GRF measurements collected from 24 Thoroughbred horses at the trot over a
period of two years were repeatable with low variation within horses, between sessions
within horses and between horses. A characteristic increase in first slope measurement,
which corresponded to a reduced rate of limb loading during early stance, became

apparent in the force patterns before the horse exhibited clinical lameness due to

23

tendinitis. Thus GRF analysis provided an objective means of detecting tendon injury at
an early stage.

The GRF abnormalities may be described by a single value, such as asymmetry in
peak vertical force during mid stance, or by an evaluation of characteristic changes in the
selected points on the force curve. Principal component analysis has been applied to
evaluate the characteristic points on the vertical GRF trace of the equine forelimbs at the
trot. Four points at the beginning of stance and four points at the end of stance on vertical
force curve were obtained (Williams, et al., 1999) in three groups of horses: sound,
horses with superficial digital flexor (SDF) tendinitis and horses with navicular disease.
Principal component analysis could identify differences in early stance between sound
and SDF tendinitis horses and in late stance between sound and navicular disease horses,
while asymmetry of peak vertical force could not be detected between the three groups.
Evaluation of peak vertical force alone may be misleading, whereas asymmetries at the
other specific points on the force curve may be more useful diagnostically. Further
analysis by principal component analysis between sound horses and horses affected with
navicular disease showed significant differences in component values during early and
late stance. After palmar digital nerve block, the locomotion pattern in late stance
returned to similar values as in sound horses, but the pattern during early stance did not
change (Williams, 2001). This indicated that horses with navicular disease have an
abnormal limb loading pattern during early stance, that is not due to pain in the navicular
area. This may have prognostic value in horses susceptible to development of navicular

disease.

24

Shoeing with a medial or lateral 3.7-degree wedge on the fore hooves resulted in
displacement of the point of force application by approximately 10 mm in the direction of
the wedge throughout stance. Even 24 hours after shoeing, the horses could not
compensate for an acute hoof imbalance by redistributing the load under the hoof
(Wilson, et al., 1998)

To assess the differentiation between flexor tendon lameness and suspensory
ligament lameness using high plantar nerve block, force plate analysis has been used as
an objective evaluation tool (Keg, et al., 1994). Tendinitis of the superficial or deep
digital flexor tendon, or desmitis of the suspensory ligament were induced by collagenase
injection of the hind limb. The affected horses showed decreases in peak force values and
impulses of the vertical, braking and propulsive forces. The braking impulse was highly
correlated with clinical lameness score. High plantar nerve block alleviated lameness due
to flexor tendinitis, with less consistent alleviation of lameness due to suspensory
desmitis. Since this block relieved lameness due to flexor tendinitis and suspensory

desmitis, it cannot distinguish between them.

Combined kinematic and kinetic analyses. No single method of gait analysis
can evaluate every parameter of locomotion. Kinematic analysis of lame horses describes
changes in movement patterns due to lameness but does not explain the relationship
between kinematic changes and the changes in force generation. Force plate analysis
alone provides meaningfiil information of changes in vertical, longitudinal and transverse
components of the GRF due to lameness, but it measures one limb at a time and does not

indicate how changes in forces affect movement profiles. Ideally, a method of gait

25

analysis should include both kinematic and kinetic measurements, but to date only a few
studies have reported complete sets of data.

Horses with induced SDF tendinitis in one forelimb were evaluated by
comparison between the lame forelimb and the contralateral compensating forelimb. The
lame limbs had lower peak vertical GRF, less flexion of the distal interphalangeal joint
and less extension of the metacarpophalangeal joint, compared with compensating limbs.
Compensating limbs had a more protracted orientation throughout the stance, and higher
peak braking force and impulse (Clayton, et al., 2000b). Protracted orientation of the
compensating limb at initial ground contact facilitated support of the trunk in a less
elevated position, allowing the horse to transfer its weight from the lame to the
compensating forelimb without raising the body mass into a suspension phase.

After immobilization of the normal metacarpophalangeal joint in a fiberglass cast
for 7 weeks, followed by cast removal, resulted in all horses being lame on the cast limb.
An exercise program, beginning with hand-walking and ending with treadmill exercise
for 8 weeks, slightly improved lameness but did not restore the normal joint function.
There were significant decreases in the metacarpophalangeal joint range of motion and
asymmetrical GRF peaks between the lame and contralateral forelimbs (Harreveld, et al.,

2002)

Effects of treatment regimes. Both kinematic and force plate analysis can be
used to monitor the outcome of specific treatment objectively.
The effect of electroacupuncture in horses suffering from chronic larninitis or

navicular disease (Steiss, et al., 1989) was evaluated using a force plate to measure load

26

 

distribution before and after four consecutive weeks of treatment. Even though the horses
seemed to improve clinically, the objective evaluation did not show significant
differences between control and treatment groups. The use of line firing and tendon
splitting to treat collagenase-induced SDF tendinitis showed progressive changes of GRF
variables up to 11 months post treatment, compared to the pre-induction variables. The
horses that were treated with rest alone showed a progressive return to the pre-induction
state (Goodship, et al., 1987).

The effect of a modified hyaluronan on amphotericin-induced carpal synovitis
was studied using computerized motion analysis to evaluate the horses’ motion before
lameness induction, after lameness induction but before the hyaluronan administration,
and after intra-articular hyaluronan administration. After lameness induction, there were
increases in asymmetrical head and withers excursions during the sound and lame
forelimb stance phases. The hyaluronan did not significantly improve the symmetry of
head and withers excursions. It was concluded that infra-articular administration of
hyaluronan was not effective in reducing lameness associated with amphotericin-induced
synovitis (Peloso, et al., 1993a).

In horses, synovectomy is used as palliative therapy for traumatic and septic
arthritis. Gait analysis was used for objective evaluation of gait changes associated with
carpal synovectomy and a subsequent exercise program (Clayton, et al., 1998). Video
recording was made in three sessions: before the operation, two days after the operation
and four months after the operation following two months rest and two months of a
progressive exercise program. There was no significant difference in clinical lameness

score, vertical head displacement, peak metacarpophalangeal joint extension, peak carpal

27

joint flexion and peak carpal joint extension between the surgical limb and contralateral
forelimb on any of the three evaluation sessions. Therefore, synovectomy in sound horses
did not cause lameness in the initial post operative period and horses can safely return to
work two months after carpal synovectomy.

The effect of oral administration of propentofylline on the treatment of navicular
disease was evaluated using force plate analysis (Kirker-Head, et al., 1993). The vertical
GRFs were measured for all four limbs to quantify the weight bearing profile prior to and
following treatment with 7.5 mg/kg propentofylline twice a day for 6 weeks. The
proportion of the total mass normalized force born by the forelimbs increased
significantly following treatment, although more of the horses returned to complete
soundness. The force plate analysis provided an objective means of assessing
effectiveness of treatment and was more sensitive in assessing drug efficacy than clinical
examination.

Several types of lameness are treated by alterations in hoof balance and
application of orthopedic shoes. In horses with navicular disease, kinematic studies
showed that application of either flat bar stock with rolled-toe or caudally extending
branches/egg—bar shoes for three weeks did not significantly reduce the asymmetrical
head movement. During the swing phase, carpal joint flexion and maximum hoof height
increased, but these changes were not solely the result of a reduction in the level of
lameness (Keegan, et al., 1998a). The balance and shoeing could be performed
concurrently with other treatments without interfering with the evaluation of the outcome

of the other treatments.

28

 

INVERSE DYNAMIC ANALYSIS: CLINICAL APPLICATION

In the early computer era, kinematic and GRF data were reported and analyzed
separately. The resulting information gave some useful criteria to quantify abnormalities
in movement and force patterns due to lameness, but was not useful for discriminating
between lameness at different sites within the limb. For example, lameness induced by
pressure on the hoof sole of the forelimb (Buchner, et al., 1996b; Galisteo, et al., 1997;
Keegan, et al., 2000) produced similar kinematic changes to an induced lameness of the
carpal joint (Back, et al., 1993; Peloso, et al., 1993b). These changes included a decrease
in maximal metacarpophalangeal joint extension during lame limb stance and an increase
in asymmetry of head excursion during left and right stance phases. Most horses also
showed a decrease in carpal joint range of motion. The use of a force plate to measure
changes in GRF due to carpal lameness (Morris and Seeherman, 1987) and SDF
tendinitis (Clayton, 2000b) reported similar findings, most notably a decrease in peak
vertical GRF, vertical impulse, peak braking GRF and braking impulse. If the objective is
to use gait analysis as a diagnostic tool, kinematic and kinetic data should be combined to
provide more information about the underlying mechanisms.

Inverse dynamic analysis is a technique that estimates the forces that cause an
observed motion. It offers a method of calculating variables that cannot be measured
directly. The technique treats the body as a system of rigid, linked segments connected at
the joints. The net joint moment, or joint torque, is calculated. It represents the action of
soft tissues around the joint. By combining net joint moment data with joint angular
velocity, net joint power can be calculated, which measures the rate of mechanical work

done around the joint. Integration of the area under the net joint power curve yields net

29

joint energy, which measures the amount of mechanical work done across a joint over a
period of time. Negative net joint power indicates negative work is being performed
(eccentric muscular contraction, or elastic stretching of ligaments or tendons) that absorbs
mechanical energy and controls joint movement. Positive net joint power indicates
positive work is being performed (concentric muscular contraction, or release of stored
elastic energy) to generate mechanical energy to move the joint. Inverse dynamic analysis
allows the evaluation of the sources of mechanical power for movement. In human
studies, it has been shown that mechanical power is the most informative variable for
evaluation of pathological gait (Winter, 1990).

The first reports of equine inverse dynamic analysis were based on the concept of
quasistatic equilibrium. The problem with this approach is that only the external force
taken into account is the GRF, which ignores the effect of inertial and gravitational forces
acting on individual limb segments. In humans, most of the joint moments in the lower
extremity during stance can reasonably be approximated by the quasistatic approach (Wu
and Ladin, 1996). Therefore, the early equine studies, which described joint moments at
the distal interphalangeal and metacarpophalangeal joints were reasonable and
acceptable. It was impossible to calculate net joint moments in the more proximal parts of
the equine limb with reasonable accuracy until the complete inertial properties of equine
body segments had been identified (Buchner, et al., 1997). Furthermore, inverse dynamic
analysis during the swing phase requires knowledge of the inertial properties.
Descriptions of inverse dynamics that take account of inertial properties of the limb
segments to analyze equine locomotion have been published, both during stance and

swing phases.

30

Inverse dynamics in sound horses. The technique has been used to identify the
kinetic pattern of the limb joints at walk and trot. Information has been reported for the
forelimb of normal horses at the walk (Colborne, et al., 1997a; Colborne, et al., 1998;
Clayton, et al., 2000a), and the mechanical function of each joint has been described.
During stance, the distal interphalangeal and scapulohumeral joints act as energy
dampers, as shown by a power profile dominated by energy absorption. The
metacarpophalangeal joint functions elastically to store and release strain energy. The
carpal joint acts like a strut as the body rolls over the grounded limb and does not appear
to play an important role in energy absorption or generation. The cubital joint is the only
joint that shows net joint energy generation at the walk. During the braking phase, energy
generated at the cubital joint maintains the extended position of the limb. Later in the
stance phase, energy generation at this joint provides forward propulsion. During the
swing phase, the cubital joint is, again, the only joint that provides energy generation to
protract and retract the entire limb, while the more distal joints are driven by inertial
forces, with their movements being controlled by negative work around each joint. The
main sources of energy absorption during swing are the scapulohumeral and carpal joints.

The net joint moment and power profiles of the normal forelimbs at the trot
showed consistent and repeatable patterns in each joint (Clayton, et al., 1998; Lanovaz, et
al., 1999). During stance, the distal interphalangeal and metacarpophalangeal joints
perform similar functions to the walk, except that the magnitudes of net joint energy
absorption and elastic energy release are greater at the trot than at the walk. The
metacarpophalangeal joint acts as an elastic spring, absorbing then releasing elastic

energy to bounce the limb upward into the swing phase. The net joint power profile at the

31

carpal joint shows two small cycles of elastic energy storage and release. The cubital and
scapulohumeral joints also appear to work elastically in the first half of stance. The
scapulohumeral joint plays an important role in energy generation to propel the body
forward during the second half of stance. The function of forelimb joints during the
swing phase of the trot is similar to their functions at the walk.

In the hind limb, inverse dynamic analysis has been applied to analyze net joint
mechanical work at the walk (Clayton, et al., 2001). During stance, the distal
interphalangeal and metatarsophalangeal joints perform similar firnctions to those joints
in the forelimbs. The tarsal joint performs mostly positive work to move the limb
throughout stance, whereas the femorotibial joint performs mostly negative work to
control the joint motion. The coxofemoral joint is the main source of energy generation to
propel the body forward. During swing, the coxofemoral joint is again the main source of
positive work, acting to protract and retract the entire limb. It is assisted by positive work
at the tarsal joint that raises and lowers the distal limb. The femorotibial joint performs
negative work to control flexion and extension of this joint. The metatarsophalangeal and
distal interphalangeal joints are driven by inertial forces under the influence of the
proximal limb joints’ motion.

The effects of an increase in walking velocity on net mechanical work have been
studied in the hind limb during stance phase (Khumsap, et al., 2001a). An increase in
walking velocity requires an increase in net joint energy generation at the coxofemoral
and tarsal joints, which confirms the important role of these two joints in performing
positive work during locomotion. The mechanical work patterns of the equine hind limb

at the trot have not yet been described.

32

Inverse dynamic analysis has been applied to study the effects of different types
of shoes. The application of normal shoes in sound horses did not effect the net joint
moments of the metacarpophalangeal or distal interphalangeal joints during the stance
phase. Shoeing improves the quality of the trot in sound horses because the increased
inertia of the lower limb produces an increase in swing phase retraction and animation of
the trot (Willemen, et al., 1997). Rocker-toed shoes are used to facilitate breakover, but
analysis of the net joint moment in the distal interphalangeal joint did not reveal any
benefits associated with this type of shoe (Willemen, et al., 1996). The duration of
breakover time and distal interphalangeal net joint moment did not differ between rocker-
, toed shoes and standard flat shoes in healthy Dutch warmblood horses.

A study that applied inverse dynamic analysis to study the landing in jumping
horses showed that the highest vertical GRF occurred in the trailing forelimb at landing.
The tendon forces calculated using inverse dynamics indicated the relative loading of the
tendons increased from deep digital flexor tendon to suspensory ligament, and was
highest in SDF tendon (Meershoek, et al., 2001). An increase in fence height fiom 0.80 m
to 1.20 m substantially increased superficial digital flexor tendon forces, whereas it
hardly influenced suspensory ligament forces and did not influence distal accessory

ligament strain.

Inverse dynamics in lame horses and effects of treatment. Inverse dynamic
analysis describing the effects of SDF tendinitis on forelimb locomotion (Clayton, et al.,
2000c) showed that the peak value of the net joint moments was lower in the lame limb

than in the contralateral forelimb at all joints crossed by the SDF muscle and tendon

33

(metacarpophalangeal, carpal and cubital joints). There were decreases in energy
absorption at the distal interphalangeal joint, in elastic energy absorption and release at
the metacarpophalangeal and carpal joints, and in energy generation at the cubital joint.
The decrease in net joint energy profiles in all joints indicated that gait was energetically
inefficient in horses with SDF tendinitis.

Inverse dynamic analysis has been used to evaluate the effects of desmotomy of
the accessory ligament of the deep digital flexor tendon on equine forelimb function
(Buchner, et al., 1996a). The net joint moment of the distal interphalangeal joint was
significantly decreased 10 days after desmotomy. There was no transfer of load from the
operated limb to the contralateral forelimb. The net joint moment at the
metacarpophalangeal joint did not change 10 days after operation. Instead, the
compensatory response was due to redistribution of force loading in the other structures
of the operated limb, with increases in force generation in the suspensory ligament and
SDF tendon to maintain the net joint moment at the metacarpophalangeal joint.

Long-term consequences of desmotomy have been observed (Savelberg, et al.,
1997; Becker, et al., 1998). Six months later, the peak net joint moment during the second
half of stance reached about 80% of its presurgical value, which indicated that function of
the deep digital flexor tendon and accessory ligament was not completely recovered after
6 months. There were no changes in joint angle, peak joint moment, or peak loading of
suspensory ligament and SDF tendon at the metacarpophalangeal joint. It was concluded
that desmotomy of the accessory ligament of the deep digital flexor tendon was not

associated with long-term changes that could be considered hazardous for the horses.

34

The effects of orthopedic shoes on locomotion have also been studied. Horses
with SDF tendinitis were shod with flat shoes and 6-degree heel wedges (Clayton, et al.,
2000d). After five days adaptation to the wedge shoe, there were decreases in peak net
joint moment and energy absorption at the distal interphalangeal joint. The different
profiles of the net joint moment between the lame limb and contralateral limb at the distal
interphalangeal joints indicated the deep digital flexor muscle in the lame limb was
contracting more strongly to compensate for the impaired SDF tendon. Significant
findings were confined to the hoof but no advantage could be identified in terms of a
decrease in the peak net joint moment at the metacarpophalangeal joint. Therefore,
application of heel wedges may not be an effective way to alleviate lameness due to SDF
tendinitis.

Horses suffering from navicular disease have been shod with two types of
orthopedic shoes. Heel wedges have been used to reduce the tension in the deep digital
flexor tendon and egg-bar shoes have been used to correct the weight distribution and
reduce the moment of force at the distal interphalangeal joint. Inverse dynamic analysis
has been used to evaluate the effect of these shoes on forces exerted on the navicular
bone (Willemen, et al., 1999). The maximal force on the navicular bone was reduced by
24% when the horses were shod with heel wedge shoes compared with flat shoes. Egg-
bar shoes did not reduce the force on the navicular bone. Therefore, horses suffering from
navicular disease may benefit from heel wedge shoes but it is doubtful that egg-bar shoes

will be beneficial.

35

THREE-DIMENSIONAL ANALYSIS IN EQUINE BIOMECHANICS

Most of the measurements of the limb joints in normal horses are primarily those
of flexion and extension, and most of the published kinematic studies have been restricted
to 2-D analysis in the sagittal plane. However, complex joints, such as the carpus and
tarsus, may also show measurable amounts of abduction-adduction and internal-external
rotation, which are important components of normal movement and which may change in
response to pathological conditions. There have been very few three-dimensional
analyses of equine locomotion, though multiple camera techniques have been used
synchronously to capture the horse’s motion in more than one plane. Images from two or
more cameras are combined into one picture, in which the depth of the fields of view and
location of skin markers are taken into account. True three-dimensional analysis
describes three rotational motions and three translational motions of the joint.

In horses, the concept of a joint coordinate system has been applied in vitro to
study the three-dimensional motion of the equine forelimb (Degueurce, et al., 2000).
Reflective markers were attached to the metacarpal bone and three phalanges, then the
limbs were loaded with a compressive force. During the neutral loading, the
metacarpophalangeal joint extended and the proximal phalanx externally rotated relative
to the third metacarpal bone, but there was no specific trend in abduction/adduction
(Chateau, et al., 2001). Addition of a lateral hoof wedge induced internal rotation and
abduction of the proximal phalanx. The opposite phenomenon was observed with a
medial wedge (Chateau, et al., 2001). A similar technique was used to study proximal
interphalangeal joint motion (Degueurce, et al., 2001). It was found that the joint had

low-amplitude of flexion, with the rate of flexion increasing exponentially with load. This

36

confirms that flexion of the proximal interphalangeal joint is only provoked by rather
high loads. The abduction/adduction and axial rotation remained constant during the
neutral loading.

Three-dimensional evaluation of limb joints in vivo have been performed on the
carpal joint of horses at walk (Nicodemus, 2000). A joint coordinate system was applied
to analyze the three-dimensional rotation of the carpal joint. The carpal joint flexed,
adducted and internally rotated during mid swing, then approached the standing joint
angles in preparation for the start of stance.

Recently, three-dimensional kinematic analysis has been performed on the tarsal
joint of sound horses at the trot (Lanovaz, et al., 2002). The motions of the third
metatarsal bone relative to tibia were observed via retro-reflective marker triads attached
Steinmann pins transcutaneously into the lateral aspects of those bones. It was found that
the tarsal joint underwent flexion, abduction and internal rotation during stance, and
underwent flexion, abduction and external rotation during swing.

Other studies have used intra-medullary pins driven to the dorsal spinous
processes to observe the three-dimensional motion of the vertebral columns of horses at
the walk (Faber, etal., 2000), trot (Faber, et al., 2001a) and canter (Faber, et al., 2001b).
In a symmetrical gait, such as walk and trot, the vertebral column showed two periods of
extension and flexion in each stride, but only one cycle of extension and flexion at canter.
At the walk, the lateral bending and axial rotation were characterized by one peak and
one trough, whereas one peak and one trough of lateral bending with two cycles of axial

rotation were found at the trot.

37

CONCLUSION

The gait analysis procedures in horses have been developed and improved
through time. Advance in computer technology have enhanced the ability to use gait
analysis as an aid in clinical evaluation of equine locomotion in variety of disciplines.
Techniques used in human locomotion, such as principal component analysis and inverse
dynamic analysis, have been adapted to analyze equine locomotion. The joint mechanical
energy profiles in horses have been described and it should be possible to apply inverse
dynamic analysis for evaluation the effects of specific type of lameness or assessment of

treatment regimes.

38

CHAPTER 2

MATERIALS AND METHODS

The procedures included in this chapter describe the selection of suitable horses,

the method of induction of synovitis, and the techniques used for gait analysis (Figure 2-

1). The data comprized kinematic variables, ground reaction force variables, net joint

moments, net joint powers and net joint energies. Statistical analyses were performed to

compare the data before and after lameness induction.

Figure 2-1 Flow chart showing sequence

of procedures.

 

Lameness Evaluation

 

I]

 

\
\
\‘e

 

l

\ ad
\‘ GA
\“‘ l

 

 

Tarsal Joint Radiography

 

“A
____2_ll_)1_19_1'l_11§l____. Exclude from study

 

 

 

 

l

 

I '
69E}”’ I I
I I
”9’ ’ I,

 

Tarsal Joint Nuclear Scintigraphy

 

” I

 

 

‘ l

 

Gait Analysis : Sound Condition

 

 

 

I

 

 

 

Synovitis Induction

 

 

 

 

f Endotoxin Preparation

 

 

l

 

Gait Analysis : Lame Condition

 

 

 

39

SUBJECT SELECTION

Four horses, aged 2-3 years old, were selected for this study on the basis of the
findings from clinical evaluation, radiography, nuclear scintigraphy and ground reaction
force (GRF) analysis. The horses’ weight and height at the withers ranged from 333-392

kg and 1.43-1.50 m, respectively.

Lameness evaluation. Each horse was evaluated for lameness using the grading
system of 0-5 (American Association of Equine Practitioners, 1991). The definition of
lameness is a deviation from the normal gait or posture due to pain or a mechanical

dysfunction. The classification is as follow,

Grade 1 : Difficult to observe; not consistently apparent regardless of

circumstances (i.e. weight-carrying, circling, inclines, hard surface, etc.)

- Grade 2 : Difficult to observe at a walk or trotting a straight line; consistently
apparent under certain circumstances (i.e. weight-carrying, circling, inclines,
hard surface, etc.)

- Grade 3 : Consistently observable at a trot under all circumstances

- Grade 4 : Obvious lameness; marked nodding, hitching or shortened stride

- Grade 5 : Minimal weight-bearing in motion and / or at rest; inability to move

The horses used in this study were selected from a groups of 18 horses that had
recently completed on exercise study. Each horse was led at the trot on a straight line on a
hard surface to evaluate the degree of lameness. After grading the lameness, flexion of

the tarsal/stifle joints was performed on both hind limbs of each horse by holding the

40

fourth metatarsal bone parallel to the ground for 90 seconds, then trotting the horse in a
straight line on a hard surface. Lameness after the flexion test was also graded. Eight
horses with lameness equal or less than grade 1 before and after the flexion tests

proceeded to radiography.

Radiography. Routine radiography of the tarsal joints was performed on both
hind limbs. The radiographic views included lateromedial, dorsoplantar, laterodorsal-
plantaromedial oblique and mediodorsal-plantarolateral oblique views. Radiographic
films from each horse were evaluated by a Board Certified Veterinary Radiologist.
Horses with abnormal tarsal bones or arthritic changes in both distal intertarsal and
tarsometatarsal joints on the same limb were excluded from the study. Juvenile spavin
has been reported in young horses before training (Laverty, et al., 1991; Bameveld and
van Weeren, 1999), and this condition has been described as an incidental finding that
may not be clinically significant. Therefore, horses with mild osteophytic changes in only
one joint (distal intertarsal or tarsometatarsal joint) were not excluded at this stage. The

remaining six horses proceeded to nuclear scintigraphy.

Nuclear scintigraphy. Since radiographic changes might not be apparent in the
early stage of degenerative joint disease, nuclear scintigraphy, which is more sensitive
than radiography for detecting degenerative changes at an early stage, was included in the
selection process. A Board Certified Veterinary Radiologist performed the nuclear
scintigraphic procedures. Each horse was catheterized intravenously in the jugular vein

using a 14 gauge, 13.3 cm catheter (Angiocath, Infusion Therapy Systems Inc., Sandy,

41

UT) to deliver Technetium-99M bound with methylene diphosphonate at a dosage of 10
mCi per 45.4 kg body weight. The horse was isolated in an equine nuclear medicine stall
for at least 2 hours before the bone phase scintillation images were taken. The horse was
sedated using detomidine hydrochloride (Dormosedan, Pfizer Animal Health, Exton, PA)
approximately 3 mg intravenously then was placed in front of the scintillation detector
for nuclear imaging. Two bone phase scinti graphic images, lateral and plantar views,
were taken from the tarsal joints of both hind limbs. The images were evaluated for
increased radiopharmaceutical uptake. After nuclear scinti graphic images were taken, the
jugular catheter was removed and the horse was returned to an equine nuclear medicine
stall for 24 hours isolation. Horses with normal or only mildly increased

radiopharmaceutical uptake in the tarsal region proceeded to gait analysis.

Ground reaction force analysis. The peak vertical force and vertical impulse are
very consistent with mean symmetry of 97:2 % symmetrical values between the left and
right sides of the forelimb and hind limb pairs (Merkens, et al., 1993). Therefore, left-
right symmetry of vertical ground reaction force (GRF) was used as a selection criterion
in this study. Velocities, peak vertical forces and vertical impulses of 5 trials fi'om left
and right forelimbs and hind limbs during the sound condition were analyzed. Symmetry
values of peak vertical force and vertical impulse were obtained by dividing the values
from one limb by the values from the other limb. The numerator was always the lesser
value and denominator was always the larger value. The result was as a percentage.
Horses with force variables that differed by more than 3 standard deviations from values

reported in the literature was considered asymmetrical and was excluded form further

42

analysis. There were two horses excluded at this stage, one with marked asymmetrical

forces in the forelimbs and the other with marked asymmetrical forces in the hind limbs.

TARSAL SYNOVITIS INDUCTION

Endotoxin preparation. The endotoxin injection was prepared in a biological
safety cabinet. A 5 mg vial of lipopolysaccharide endotoxin from Escherichia coli
OSS:B5 (L2637, Sigma-Aldrich, Saint Louis, MO) was diluted with 5 ml sterile water
and kept as endotoxin stock solution at concentration of 1 mg/ml. The stock solution was
diluted ten-fold by adding 9 ml of 6.6 pH diluent (Normosol-R, Abbott Laboratories,
North Chicago, IL) to 1 ml of stock solution to obtain a second stock at 100 jig/ml. The
same procedure was repeated by mixing 1 ml of the second stock with 9 ml of 6.6 pH
diluent to obtain the final endotoxin solution of 10 jig/ml. Five sterile vials (American
Pharmaceutical Partners Inc., Los Angeles, CA) were used to keep the final endotoxin

solution, with 2 ml in each vial. The solution was stored at 46°F until used.

Intraarticular injection of endotoxin. Non-infectious synovitis was induced in
the right tarsal joints of each horse by a clinician from the Large Animal Hospital,
Michigan State University. Before induction of synovitis, a physical examination was
performed to determine baseline values of rectal temperature, pulse rate, respiratory rate,
mucous membrane color, capillary refill time and joint appearance. The horse was
sedated by intravenous injection of a mixture of 3 mg detomidine hydrochloride and 3 mg
butorphanol tartrate (Torbugesic, Fort Dodge Animal Health, Overland Park, KS).

Synovitis was induced in the distal intertarsal and tarsometatarsal joints by injection of 5

43

ug endotoxin into each joint. The skin over the sites of entry into the tarsometatarsal joint
on the lateral side, and the distal intertarsal joint on the medial side of the right hind limb
were aseptically prepared by hair clipping and scrubbing with betadine scrub, alcohol and
betadine solution. A 20 gauge, 3.8 cm needle used to inject 0.5 ml of 10 ug/ml endotoxin
solution into the tarsometatarsal joint, which was approached by inserting the needle over
the proximal end of the fourth metatarsal bone. A 22 gauge, 3.8 cm needle was used to
inject 0.5 ml of 10 jig/ml endotoxin solution into the distal intertarsal joint, which was
entered at the gap between the central tarsal, the third tarsal and the fused first and second

tarsal bones.

Care after endotoxin injection. Physical examination of each horse was
performed every two hours after injection of endotoxin for the first 12 hours, then at
intervals of 6 hours, until 24 hours after induction. The parameters assessed were rectal
temperature, pulse rate, respiratory rate, mucous membrane color, capillary refill time,
joint appearance and assessment of appetite. Lameness evaluation was performed
approximately every 12 hours. If the horse developed a lameness of grade 3 or greater,
0.5-l g of phenylbutazone (Phenylzone paste, Schering-Plough Animal Health Corp.,
Union, NJ) was administered orally according to the severity of lameness. When the
horse had a mild but consistent lameness (grade 2) at the trot on a hard surface, gait
analysis was performed. After gait analysis, each horse was kept in a stall and examined
daily until it had no sign of lameness at the walk and minimal lameness at the trot (grade

1 or less). At this time, the horse was returned to pasture.

44

GAIT ANALYSIS

Gait analysis was performed twice in each horse. After the horse had been
selected as suitable for the study, gait analysis was performed to obtain data in the sound
condition, from both left and right sides of the horse. After synovitis induction, gait
analysis was repeated to obtain data for the lame condition fiom both left and right sides.

A flow chart showing the data collection procedures is shown in Figure 2-2.

Figure 2-2 Flow chart for gait analysis procedures.

 

 

 

 

 

 

 

 

 

Video System
Horse Camera positions identified and
data collection volume calibrated
Skin marker attachment l
on the first side

 

 

 

 

 

 

Captured force plate comer data

/

Captured first side reference pose

1

Data collection at trot
5 trials from each limb Zero adjustment

J.

Remove skin markers
Attached skin markers on the second side

i

Captured second side reference pose

1

Data collection at trot
5 trials from each limb

 

 

 

 

 

 

 

Force plate

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

45

Kinematic data.

The video system (Motion Analysis Corporation, Santa Rosa, CA) consisted of 6
video cameras, a video processor computer (MIDAS), tracking computer, SVGA CRT
monitor and multisync digitizing monitor (Figure 2-3). Each camera had a ring light
surrounding the camera lens that produced a bright red strobed light to illuminate
spherical, retro-reflective markers attached to the horse. The video cameras captured the
location of the markers in space. All six cameras were synchronized and the video analog
inputs were sent to a video processor computer to be converted to digital image data, then
forwarded to a tracking computer via an Ethernet connection. The tracking computer
combined the digital input from the six cameras to calculate the location of markers in
space relative to the camera locations. At any point in time, each marker had to be seen
by at least two cameras in order to identify the marker location. After all the marker
locations had been identified, the computer generated a picture of the markers linked
according to a predetermined marker template. These pictures were generated in real time

as the horse moved across the field of view.

46

 

Figure 2-3 Motion analysis video system.

 

 

Video system with 6 cameras (1-6) arranged on the far side of the runway. The computer
system is located behind camera 5. The runway is in the foreground.

The calibration system Calibration consisted of two steps: identification of the
camera locations in space and calibration of the data collection volume. To identify the
camera locations, a rigid L-shaped frame with 4 spherical, retro-reflective markers
attached at predefined locations was placed in the middle of data collection field. This
frame was also used to identify the origin and orientation of the global coordinate system
by aligning its two arms with the longitudinal and transverse axes of the force plate
(Figure 2-4). All six cameras captured the markers on the L-shaped frame and sent the
input to the computer units. Based on the known locations of the markers, the computer
inversely calculated the location of each camera, and after this procedure, the cameras

were stationary until data collection had been completed.

47

figure 2-4 Position of a rigid L-shaped frame in the data collection field.

 

The global coordinate system was identified for the horizontal longitudinal (X) and
horizontal transverse (Y) axes. The vertical axis (Z) was perpendicular to the X and Y
axes (not shown in this picture). The force plate can be seen above the L-shaped frame.
The marker on the left comer of the flame (0,0) was used to identify the origin of the
global coordinate system in the X and Y axes. The origin in the Z axis was at ground
level directly beneath the (0,0) point.

Calibration of the data collection volume was performed to refine the calculated
locations of the cameras and to reduce the effects of lens distortion. It involved the use of
a 1000 mm wand calibration stick with three 3.8 cm spherical retro-reflective control
markers. All cameras captured the wand moving throughout the data collection volume in
directions parallel to each of the three axes (X, Y and Z). The calibrated volume
measured approximately 5 x 2 X 3 meters (Figure 2-5). The computer units combined the

information from the six cameras. Based on the fact that the wand length was constant at

48

1000 mm, computer algorithms adjusted each camera view along the three axes using
least square residual error to optimize the six camera views.

Location of the force plate was obtained by placing a 2.54 cm diameter, spherical
retro-reflective marker at each corner of the force plate, then capturing these markers on
all cameras as the force plate comer data file. The location of the force plate within the

data collection volume could then be determined.

Figure 2-5 Data collection volume after calibration.

4,1—

 

The position of the six cameras (1-6), the orientation and origin of the global coordinate
system (X,Y,Z axes) and the location of the force plate on the runway (rectangle) were
identified.

49

Skin marker setup The whole body skin marker setup was used to obtain two-
dirnensional kinematic data for the forelimb and hind limb, and three-dimensional

kinematic data for the tarsal joint.

Figure 2-6 Position of retro-reflective markers on the cranial part of the body.

 

On the cranial part of the body (Figure 2-6), 2.54 cm spherical, retro-reflective
markers were attached to the skin over the following anatomical landmarks: 1. wing of
first cervical vertebra; 2. base of the neck; 3. tuberosity of scapular spine; 4. caudal part
of greater tubercle of humerus (representing center of rotation of scapulohumeral joint);
5. attachment of lateral collateral ligament of cubital joint on distal humerus (representing

center of rotation of cubital joint); 6. distal edge of ulnar carpal bone midway between the

50

lateral styloid process of radius and proximal third metacarpus (representing center of
rotation of carpal joint); 7. distal end of third metacarpus at attachment of the lateral
collateral ligament of metacarpophalangeal joint (representing the center of rotation of
metacarpophalangeal joint); 8. lateral hoof wall distal to the coronary band; 9. hoof wall

at toe and 10. hoof wall at heel.

Figure 2-7 Position of retro-reflective markers on the caudal part of the body.

 

On the caudal part of the body (Figure 2-7), 2.54 cm spherical, retro-reflective
markers were attached on the skin over 1 1. dorsal midline of the trunk; 12. ventral part of
the tuber coxae; l3. caudodorsal projection of the greater trochanter; l4. cranial part of
distal femur; 15. femoral condyle at the attachment of the lateral collateral ligament of
femorotibial joint (representing center of rotation of femorotibial joint); 16. proximal

tibia at the attachment of lateral collateral ligament of femorotibial joint; 17. lateral

51

malleolus of distal tibia; 18. dorsal edge of the head of fourth metatarsus; l9. distal end of
third metatarsus at the attachment of the lateral collateral ligament of
metatarsophalangeal joint (representing center of rotation of metatarsophalangeal joint);
20. lateral hoof wall distal to the coronary band; 21. hoof wall at toe; and 22. hoof wall at
heel. In order to track the three-dimensional motion of tibia and metatarsal segments,
additional markers were attached on those two segments: 23. distal edge of tibial
tuberosity; 24. midpoint of a line connecting l6 and 17; 25. caudal aspect of the tibia
approximately two-third of the distance along a line from 16 to 24; 26. cranial aspect of
the tibia approximately one-third of the distance along a line from 24 to 17; 27.
approximately one-third of the distance along a line from 18 to 19; 28. approximately
two-third of the distance along a line from 18 to 19; 29. dorsal edge of third metatarsus
between 18 and 27; and 30. dorsal edge of third metatarsus between 28 and 19. All skin
markers could be captured simultaneously and identified correctly by the computer units
(Figure 2-8).

Since the center of rotation of the distal interphalangeal joint cannot be directly
palpated, its position relative to the hoof markers was located radiographically. The three
hoof markers on the coronary band, toe and heel were marked with metal beads. A
lateromedial radiograph of each foot was taken to identify the position of the center of
rotation of the distal interphalangeal joint relative to the position of the three hoof
markers. During post-processing of the data, the positions of the hoof markers were used

to determine the distal interphalangeal joint location, as described later.

52

Figure 2-8 Stick figure of whole body marker setup displayed by the computer monitor.

12 ‘2

10‘9

Force plate data. A 60 cm X 120 cm force plate (AMTI LG6, AMTI, Watertown,
MA) embedded in the middle of the runway, was used to collect GRF data with a
sampling frequency of 1,200 Hz. The force plate reading was adjusted to zero in all three
directions before each data collection session. When a horse stepped on the plate, forces
in the vertical, horizontal longitudinal and horizontal transverse directions, and the
location of the center of pressure were recorded to the same computer units as the video
system. The kinematic axes were aligned with the force plate axes using the force plate
corner data file. The conventional axes for force plate in this study were defined as

vertical (Z), horizontal longitudinal (X) and horizontal transverse (Y) axes. Force data

53

were synchronized temporally and spatially with the kinematic data within the computer

units.

Data collection procedures.

Normalization for horse size Horse height affects stride length, with shorter
horses using higher stride frequencies than taller horses at the same velocity, which
requires more energy expenditure over the same distance. Comparisons between horses
of different sizes are facilitated by scaling the data so that each animal moves at the same
velocity relative to its body size. Scaling in human locomotion uses F roude numbers to
adjust the velocity of progression using a characteristic length, such as leg length. When
people walk with equal F roude numbers, they have dynamic similarity (Zatsiorky, et al.,
1994). Scaling can also be used to obtain dynamic similarity in horses. Height at the
withers, which is proportional to the segment lengths of the forelimbs and hind limbs
(Holmstrom, 1990), is applied in gait analysis to scale the gait variables.

The important GRFs are those used to accelerate/decelerate the body mass and
those used to oppose gravitational forces. In order to achieve dynamic similarity, the
proportion of acceleration and gravitational forces should be similar for animals of
different sizes (Hof, 2001).

m131 = mzaz

mlg ng

 

where m1 is mass of subject 1, a1 is the acceleration of subject 1, m2 is mass of subject 2,
and a2 is the acceleration of subject 2.
From this equation, the scaling for absolute acceleration of any horse should be

adjusted for gravitational acceleration (g = 9.81 m/sz). When analyzing other variables

54

 

that relate to change in height at the withers (distance in the direction of gravitational
acceleration), they should be scaled to achieve dynamic similarity as well. The equations

used to scale time and velocity are as follow :

. distance . distance 10
acceleration = . 2 then trme = ——.— = —
time acceleration g

 

 

 

 

. distance distance . .
velocrty = , = . = Jdrstancexacceleratron = ,/10 X g
time J distance
acceleration

When 10 is height at the withers in meters, the term used to describe the adjusted
variables is “Dimensionless units ” because the numerator and denominator have the
same units. In this study, normalization for velocity was used to select trials for analysis
from each horse. The dimensionless units for subject velocity (VDU) were calculated
using the following equation :

measured velocity (m/s)
\/ 1o x g

 

velocity =

A previous study (Clayton, et al., 2000b) reported the mean preferred trotting
velocity for normal horses was 0.83 VDU and moderate lameness reduced the preferred
trotting velocity by approximately 10%. The velocity in dimensionless units in this study
was selected in the range of 0.73-0.77 VDU, which was considered to be a reasonable
range for achieving comparable velocities before and after synovitis induction.

Data collection The first session for data collection was used to obtain data
describing the horse in the sound condition. Video cameras were set to record at a rate of
120 frames per second, with a shutter speed of 1/1000 and f-stop setting around 4.
Recordings of the force plate comer data, the rigid L-shaped frame and the wand

calibration were used to identify camera locations, force plate location, and global

55

coordinate system, respectively, and to calibrate the data collection volume. The whole
body marker setup was attached to the first side of the horse’s body. Two additional 2.54
cm spherical, retro-reflective markers (reference markers) were attached on the medial
malleolus of the tibia and the dorsal edge of the head of the second metatarsus. The horse
was positioned obliquely to the longitudinal axis of the data collection volume in such a
way that all the markers could be seen by at least two cameras. A video recording, named
the reference pose, was captured, then the two reference markers were removed.

The handler led the horse at the trot through the data collection volume. A good
hit was one in which the horse placed its first side forelimb and/or hind limb on the force
plate without any other limb being on the force plate at the same time. Custom software
was used to calculate velocity and velocity in dimensionless units (VDU) of the marker
overlying the tuberosity of the scapular spine and used as the approximate forward
velocity for each trial. The information was instantly available after completion of the
trial. Successful trials were selected on the basis of a good hit on the force plate by the
first side forelimb and/or hind limb, combined with velocity in the range of 0.73-0.77
VDU. The data collection was continued until 5 successfirl trials had been obtained from
both the forelimb and hind limb. After the data collection on the first side of the horse
was completed, the skin around each marker was stained by waterproof ink then the
markers were removed. A whole body marker setup was then attached on the second side
of the horse’s body. Data collection procedures were repeated including attachment of
medial reference markers, recording of the reference pose, capture of 5 successful trials
from both the forelimb and hind limb, and skin marking with waterproof ink around each

marker before removal.

56

After induction of synovitis, when the horse was assessed as having a grade 2
lameness or less, data were recorded for the lame condition on both left and right sides as
described previously. Marker placement could be reproduced from the sound condition
by placing markers at the positions indicated by waterproof ink. Data collection
procedures were performed including whole body marker set attachment on the first side,
attachment of the medial reference markers and recording of the reference pose. After 5
successful trials from the first side forelimb and hind limb had been captured, the markers
were removed and a whole body marker setup attached in the positions marked by
waterproof ink on the second side of the horse. The data collection procedures were
repeated until 5 successful trials from the second side forelimb and hind limb had been

obtained.

DATA PROCESSING AND ANALYSIS

Two-dimensional variables. Data from the tracking computer units were
transferred to a processing computer to be manipulated before analysis. Kinematic
information from successful trials was analyzed on a stride-by-stride basis. For the right
side data, a stride started and ended at the instant of ground contact, so the complete
stride consisted of a stance phase followed by a swing phase. For the left side data, the
stride started and ended at lift off, so each stride consisted of a swing phase followed by a
stance phase. This difference between the two sides was necessary because the calibrated
volume extended to one side of the force plate. Custom software (MAC Project
Organizer, Mary Anne McPhail Equine Performance Center, East Lansing, MI) was used

to organize kinematic and force information for each trial in preparation for further

57

processing. Matlab software (Version 6.1, The Math Works, Inc., Natick, MA) was used
for the mathematical calculations. A fourth order Butterworth digital filter was used to
filter the raw kinematic data with a cut off frequency of 10 Hz, and to filter the raw force
plate data with a cut off frequency of 40 Hz. Skin displacement correction for the trot
(van Weeren, 1989) was applied to the kinematic data obtained from markers at the
greater tubercle of humerus, proximal radius, distal edge of ulnar carpal bone,
caudodorsal projection of the greater trochanter, femoral condyle, proximal tibia, the
head of the fourth metatarsus and distal end of the third metatarsus before any further
kinematic data analysis.

Kinematic variables Temporal variables were obtained from the time recording of
the limb events. Stride duration was determined from the number of frames elapsing
between successive ground contacts of each right hoof for the right side data or
successive lift offs of each left hoof for the left side data. The force plate is very accurate
for identifying the time the hoof was on the ground (Schamhardt and Merkens, 1994) so
it was used to detect the ground contact and lift off frames for each limb. The limb
patterns from the ground contact and lift off frame were calculated from all markers on
the limb and be used as the patterns to identify the successive ground contact frame for
the right limbs and former lift off frame for the left limbs. Stance and swing durations
were determined from the number of frames elapsing while the limb was on the ground
and off the ground, respectively.

Angular displacement, velocity and acceleration were obtained for each joint
during the complete stride. The retro-reflective markers over the anatomical landmarks

on the fore and hind limbs were used to construct stick figures with the lines representing

58

segments of the limbs, and the joint angles being formed where the segment lines
intersected. The lateral radiograph of the hoof was used to identify the location of the
center of rotation of the distal interphalangeal joint relative to the position of the three
hoof markers. The center of rotation of the distal interphalangeal joint was determined
on the radio graph using a standard method to identify the center of circular object (Figure
2-9). Two straight lines were drawn across the distal articular surface of the second
phalanx then a perpendicular was drawn from the middle of each line toward the center

of the bone. The intersection of these perpendicular lines was the center of joint rotation.

Figure 2-9 Diagram showing the method used to identify the center of a circular object.

\/

Custom software (Radcof, Mary Anne McPhail Equine Performance Center, East
Lansing, MI) was used to identify the center of rotation of the distal interphalangeal joint
from the lateromedial radiograph of the hoof (Figure 2-10). The spatial relationship
between the hoof markers and the joint center of rotation was then determined. The
standard procedure was to use the markers at the coronary band and toe, but other
markers combinations could be used if these ones had not been tracked well by the video
system. The hoof coordinate system was established using the chosen markers and the

reference length was measured between them. The hoof segment was aligned parallel

59

with the dorsal edge of the third phalanx from the center of the distal interphalangeal
joint to the distal hoof wall. Standard two-dimensional transformation (Winter, 1990) was
used to locate the distal interphalangeal joint, distal hoof wall and hoof angle with the
respect to the hoof coordinate system and reference length. During kinematic analysis,
the hoof coordinate system was re-established in each frame, the locations of the distal
interphalangeal joint and distal hoof wall were determined in the hoof coordinate system,
then the standard two-dimensional transformation and reference length were used to
define those locations in the kinematic global coordinate system. This procedure was

performed for every frame during the stride.

Figure 2-10 Lateral radiograph of the hoof and hoof coordinate system.

 

Custom software was used to calculate the location of distal interphalangeal joint center
of rotation using two straight lines drawn across the distal arctic surface of the second
phalanx (short, dark lines). The joint center of rotation was located at the intersection of
lines (dashed lines) drawn perpendicular to their midpoints. Alignment of hoof segment
(long black line) relative to the hoof coordinate system (white lines) was determined from
retro-reflective markers at the coronary band and toe.

60

A similar transformation was used to locate the center of rotation of the
tarsocrural joint on a lateromedial radiograph. The standard method for identifying the
center of a circular object, in this case the tibial tarsal bone, was applied and the tarsal
coordinate system was established (Figure 2-1 1). The reference length was 10 cm long
and was drawn distally from the markers on the fourth metatarsal bone, parallel to the
long axis of the third metatarsal bone. Standard two-dimensional transformation was used
to locate the tarsocrural joint center of rotation in the tarsal coordinate system using the
reference length, and was transposed to the kinematic global coordinate system. This

procedure was repeated for every frame during the stride.

Figure 2-11 Lateral radiograph of the tarsus and tarsal coordinate system.

 

Custom software was used to calculate the location of the tarsocrural joint center of
rotation relative to the tarsal coordinate system (white lines) constructed from the retro-
reflective marker at the dorsal edge of the head of the fourth metatarsal bone and a
reference line parallel to the long axis of the third metatarsal bone.

61

The coxofemoral joint center of rotation lies medial to the cranial part of the
greater trochanter. Its position was calculated from the locations of the skin markers on
the femur. The skin marker correction algorithm was applied, then the position of this
marker was transformed in the X-Y plane. The reference length for the femoral segment
was the distance between the markers on the caudal part of greater trochanter and the
femoral condyle. It was used to establish the hip coordinate system (Figure 2-12).
Anatomical measurements on the equine femurs showed that the center of the femoral
head was cranial and distal to the caudal part of greater trochanter by 7.8% and 17.8%,
respectively, of the reference length. Standard two-dimensional transformation was
performed to locate the coxofemoral center of rotation in the hip coordinate system. The
hip coordinate system was re-established fi'om kinematic data using skin markers on the
caudal greater trochanter and femoral condyle. Two-dirnensional transformation was

performed to define the location of the coxofemoral joint in the global coordinate system.

Figure 2-12 Lateral side of equine femur and hip coordinate system.

 

I.

The hip coordinate system established from markers at the caudal part of greater
trochanter and femoral condyle. The position of the coxofemoral joint center of rotation
relative to these markers was calculated by two-dimensional transformation.

62

The angle of each joint was measured on its anatomical flexor side. Thus, the
angles of the scapulohumeral, carpal, femorotibial, metacarpophalangeal,
metatarsophalangeal and distal interphalangeal joints were measured on the caudal or
palmar/plantar side of the limb and the angles of the cubital, coxofemoral and tarsal joints
were measured on the cranial or dorsal side of the limb. During locomotion, the angle
between adjacent segments was measured; flexion was defined as a decrease in the angle
and extension occurred when the angle increased. The peak magnitudes of joint flexion
and extension and their times of occurrence were obtained. Each trial was standardized to
stance or swing duration, and the time of occurrence of each peak was expressed as
percent of stance or swing duration. The range of j oint motion was the difference
between maximal and minimal values of joint angle during stance or swing phases of
each trial. The kinematic variables from the sound and lame conditions were obtained for
comparison.

Linear variables were calculated from changes in location of digitized markers on
anatomical landmarks of interest within the calibrated volume. Stride length was the
forward longitudinal displacement of each limb between successive ground contacts or
lift offs. Vertical displacements of the tuber coxae, coxofemoral joint, femorotibial joint,
tarsal joint, metatarsophalangeal joint, distal interphalangeal joint and distal hoof wall
were measured during the stance or swing phases as the differences between the lowest
and the highest position of each landmark. After the temporal and linear variables were
obtained, the exact velocity in m/s of the limb in each trial was calculated by stride length
divided by stride duration. The velocity in dimensionless units then was calculated as

described previously.

63

F orce plate variables GRF data from each trial were analyzed to determine the
peak magnitudes of the vertical force peak, negative longitudinal or braking force and
positive longitudinal or propulsive force and their times of occurrence expressed as
percent of stance duration. Irnpulses were calculated by time integration of the force
curves and expressed as vertical impulse, braking impulse and propulsive impulse. Force
peaks and impulses were normalized to horse mass and expressed as N/kg and Ns/kg,
respectively.

Net joint moment, net ioint power and net Loint energy The kinematic and GRF

 

data were synchronized in the time domain and combined with inertial parameters of the
limb segments (Buchner, et al., 1997). An inverse dynamic analysis (Winter, 1990) was
used to calculate the sagittal plane net joint moments for the distal interphalangeal,
metacarpophalangeal, metatarsophalangeal, carpal, tarsal, cubital, femorotibial,
scapulohumeral and coxofemoral joints. Moments on the anatomical flexor side were
assigned a negative value and those on the extensor side were assigned a positive value.
Net joint power was calculated as the product of net joint moment and joint angular
velocity. The value was positive when the net joint moment and joint angular velocity
had the same polarity (concentric muscular action) and negative when they had opposite
polarities (eccentric muscular action). The net joint moment and net joint power were
normalized to stance or swing duration and to horse mass, and was expressed as moment
and power per kg body weight (Nm/kg and W/kg). Maximal moment and power values
were determined from the obvious and consistent peaks in the net joint moment and net
joint power curve. Bursts of energy absorption and generation were calculated by time

integration of the negative and positive parts, respectively, of the net joint power curves.

64

The area under a positive power curve was net joint energy generation and the area under
a negative power curve was net joint energy absorption. The net joint energy variables
were measured as Joule/kg (J/kg) body weight. The net joint moment peaks, net joint
power peaks, net joint energy absorption and generation, and time of occurrence of each
peak expressed in percent of stance or swing duration fi'om sound and lame conditions

were obtained for comparison.

Three-dimensional kinematic variables for the tarsal joint. One method of
expressing three-dimensional (3 D) joint kinematics is to describe the relative movement
between two segmental coordinate systems. In this study, segmental coordinate systems
were constructed for the tibial and metatarsal segments. Because skin displacement of the
marker locations relative to the underlying bones is an important source of errors in
kinematic analysis, the true bone motion was identified using reference bone data
(Lanovaz, et al., 2002) from a same group of horses. Skin marker correction algorithms
(Appendix A) were obtained fiom the reference bone data using the same marker
locations for the tibia and metatarsal segments. The tibial segmental coordinate system
(Figure 2-13) was constructed from the markers over the proximal tibia at the attachment
of the lateral collateral ligament of the femorotibial joint, the lateral malleolus and the
medial malleolus. The first axis to be defined was the flexion/extension axis (Y), which
was a vector running from lateral malleolus to medial malleolus. The second axis was the
abduction/adduction axis (X), which was calculated as a vector perpendicular to both the
Y axis and a line from the lateral malleolus to the proximal tibia. This axis was positive

in a caudal to cranial direction. The third axis was the intemal/external rotation axis (Z),

65

which was calculated as a vector perpendicular to both X and Y axes, and positive in a
distal to proximal direction. The origin of the tibial segmental coordinate system was
embedded in the bone midway between lateral and medial malleoli. The metatarsal
segmental coordinate system (Figure 3-13) was constructed using markers over the
dorsal edge of the heads of the fourth and second metatarsals, and the distal condyle of
the third metatarsus on its lateral side. The procedures used to obtain X, Y and Z axes of
the metatarsal segmental coordinate system were similar to those for the tibial segmental
coordinate system. The flexion/extension (Y) axis was a vector from the fourth to the
second metatarsus, which was positive in the medial direction. The abduction/adduction
(X) axis was perpendicular to the Y axis and a line from the head of the fourth metatarsus
to the distal third metatarsal condyle. The Y axis was positive in the dorsal direction. The
internal/extemal rotation (Z) axis was mutually perpendicular to both X and Y axes, and
was positive in a proximal direction. The origin of the metatarsal coordinate system was
embedded in the bone midway between the dorsal edge of the heads of the fourth and

second metatarsals (Figure 2-13).

66

Figure 2-13 Segmental coordinate system of tibial and metatarsal segments.

e-"a Z
Lateral ' - .6 Q; x Medral

. :,
1-
f.

 

Position of retro-reflective markers (black circles) relative to tibia (left) and third
metatarsus (right). Segmental coordinate systems were located at the middle of the distal
tibia and at the middle of the proximal third metatarsus. The flexion/extension axis (Y)
had a lateral to medial direction. The abduction/adduction axis (X) had a caudal/plantar to
cranial/dorsal direction. The intemal/extemal rotation axis (Z) had a distal to proximal
direction.

The reference pose was used to relate the location of the skin markers overlying
each segment to its segmental coordinate system. A 3D transformation matrix (Winter,
1990) was used to transform the locations of these skin markers from the global
coordinate system to their corresponding segmental coordinate system. During the
trotting trials, skin markers were tracked and expressed in terms of the global coordinate
system. A singular-value decomposition method (Soderkvist and Wedin, 1993) was used
to calculate the position and orientation of each segmental coordinate system for every

frame during the stride based on the skin markers of each segment.

67

Complete 3D kinematic analysis of the tarsal joint complex involved the
measurement of relative angular and translational motions between the tibial and
metatarsal segmental coordinate systems. The relative angular motion between these two
coordinate systems was expressed in terms of a spatial attitude vector (Woltring, 1994).
This method is based on the finite helical axis representation of relative motion, in which
the position and orientation of one body relative to the other can be described as a scalar
rotation and translation between two segments (Woltring, et al., 1985). The attitude
vector is calculated by multiply the scalar finite helical angle with the component of the
unit vector describing the helical axis. The results are three orthogonal angular values that
are expressed in term of the coordinate axes of one of the bodies. The magnitude of
relative motion of the first segment relative to the second segment was the same as that of
the second segment to the first segment, but in the opposite direction.

There were three relative angular motions between the two segments;
flexion/extension, abduction/adduction and intemal/external rotation. In this study,
angular motion was measured as movement of the metatarsus relative to the fixed tibia
(Figure 2-14). Flexion/extension was motion around the Y axis; negative values were
assigned for flexion, and positive values were assigned for extension.
Abduction/adduction was motion around the X axis; negative values were assigned for
abduction, and positive values were assigned for adduction. Intemal/ external rotation was
motion around the Z axis; negative values were assigned for external rotation, and

positive values were assigned for internal rotation.

68

Figure 2-14 Three-dimensional angular motion of the metatarsal segment relative to the
tibial segment; flexion(-)/extension(+) (left), abduction(-)/adduction(+) (middle) and
internal(+)/extemal(-) rotation (right). Figure illustrates the right hind limb.

 

Medial view Dorsomedial view Dorsomedial view

Translation was measured in three directions; cranial/caudal, medial/lateral and
proximal/distal translations. The relative translation was expressed as the metatarsal
translation relative to the fixed tibia (Figure 2-15). Cranial/caudal translation was motion
along the X axis; negative values were assigned for caudal translation, and positive
values were assigned for cranial translation. Medial/lateral translation was motion along
the Y axis; negative values were assigned for lateral translation, and positive values were
assigned for medial translation. Proximal/distal translation was motion along the Z axis;
negative values were assigned for distal translation, and positive values were assigned for

proximal translation.

69

Figure 2-15 Three-dimensional translational motion of the metatarsal segment relative to
the tibial segment; cranial(+)/caudal(-) translation (left), medial(+)/lateral(-) translation
(middle) and proximal(+)/distal(-) translation (right). Figure illustrates the right hind
limb.

     

Medial view Dorsomedial view Dorsomedial view

The ranges of 3D angular and translational motion were obtained from every trial
during the stance or swing phases for comparison between the sound and lame

conditions.

STATISTICAL ANALYSIS

Power analysis is most useful when planning a study to estimate the sample size
for the experiment (Thomas, 1997). It was not possible to perform a power analysis
during the design phase of this study because there are no comparable data that can be

used to estimate power for most of the variables.

70

Statistical software (SPSS for Windows version 10.1, SPSS Inc., Chicago, IL)
was used to analyze the data. Independent t-tests were used to test the difference in
velocity between left and right limbs. For two-dimensional variables, kinematic, GRF,
net joint moment, net joint power, and net joint energy variables were obtained from 5
trials of each limb. The values of each variable were combined to obtain the mean values
that were regarded as being representative values for each limb. The descriptive analysis
for the sound condition was performed using the representative values of each variable
from left and right hind limbs of all horses to find the mean and standard deviation.

Independent t-tests were used to test whether there was any difference in velocity
between sound and lame conditions of each limb. The representative variables of each
limb from sound and lame conditions were compared using dependent t-test. Significant
differences were obtained when the P value was less than 0.05. A trend toward a
significant difference was reported when the P value was less than 0.1 but more than
0.05.

For 3D variables, flexion/extension, adduction/abduction and internal/external
rotation, as well as cranial/caudal, medial/lateral and proximal/distal translation were
obtained from 5 trials of the right limb of each horse to calculate the representative mean
values of each limb. The 3D motion of the tarsal joint from the motion of the tibial and
metatarsal bones was obtained (Lanovaz, et al., 2002) to illustrate the true reference
kinematics of the bones. The root mean square (RMS) errors of tarsal joint 3D rotation
and translation from each individual horse during the whole stance or swing duration
were calculated between reference bone data and the kinematic data from both

uncorrected and corrected skin markers. The variables that showed reasonable RMS

71

errors and shape agreement of the motion curves between the reference bone data and
corrected skin markers would be analyzed for the ranges of motion during the stance and
swing phase. The dependent t-test was used to compare the rotational and translational

ranges of motion before and after the synovitis induction. Significant differences were

obtained when the P value was less than 0.05.

72

CHAPTER 3

SAGITTAL PLANE KINEMATICS AND KINETICS OF THE HIND LIMBS IN
SOUND TROTTING HORSES: RESULTS AND DISCUSSION

HORSE SELECTION

The lameness examination, radiographic and nuclear scintigraphic results of the

four horses used in the study are reported in Table 3-1.

Table 3-1 Examination results from lameness evaluation, radiography and nuclear

 

 

 

 

 

scintigraphy.
Horse . . . .
no Lameness evaluation Radiography Nuclear scmtrgraphy
Grade 1 on the right Mild arthritic change in $2315 lfrzrme:::frltlic a1
1 hind limb before and the right P . .
. . . uptake 1n the right
after flexron test tarsometatarsal jornt . .
tarsometatarsal jornt
Grade 1 on the right Mild arthritic change in
2 hind limb before and the right Normal
after flexion test tarsometatarsal joint
Grade 0 before flexron Mild 1 tie change
test and grade 1 after .
3 . . on the left distal Normal
flexron test on the right . l . .
hind limb Intertarsa jornt
Mild increase in
Grade 1 on the right radiopharrnaceutical
4 hind limb before and Normal uptake in the left and
after flexion test right tarsometatarsal
joints

 

 

 

 

 

73

 

TEST FOR DIFFERENCE IN VELOCITY AND SYMMETRY

Data from the left and right sides of the body were collected fiom different trials
with the horses moving in opposite directions. Since most of gait variables vary with
velocity, it is important to check for consistency of velocity before proceeding with
further analyses. There were no significant differences (P>0.05) in limb velocity or
velocity in dimensionless units between the left and right forelimbs or left and right hind
limbs in the sound condition (Table 3-2). Having established equality of velocity for data
collections on the left and right sides, the next step was to assess symmetry between the
left and right limbs. Symmetry values calculated according to Merkens (Merkens, et al.,
1993) for peak vertical force and vertical impulse were close to 100%, which is indicative
of a symmetrical weight-bearing pattern (Table 3—3). The independent t-test of linear
variables, indicate no significant differences (P>0.05) between left and right hind limbs.
Hence the variables could be combined between left and right hind limbs for descriptive

analysis of the gait variables.

Table 3-2 Statistical analysis results for testing differences of velocity (m/s) and velocity

in dimensionless units between left and right limbs. Values are mean and (SD).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Horse no. Veloci (m/s) Velocity in dimensionless units
Right fore Left fore Right fore Left fore P value
1 2.91 (0.09) 2.77 (0.13) 0.77 (0.02) 0.74 (0.04) 0.09
2 2.91 (0.06) 2.85 (0.05) 0.76 (0.02) 0.74 (0.01) 0.82
3 2.83 (0.07) 2.88 (0.07) 0.74 (0.02) 0.75 (0.02) 0.28
4 2.82 (0.06) 2.83 (0.06) 0.75 (0.02) 0.75 (0.02) 0.87
Horse no. Velocity (m/s) Velocity in dimensionless units
Right hind Left hind Right hind Left hind P value
1 2.85 (0.09) 2.74 (0.06) 0.76 (0.02) 0.73 (0.02) 0.05
2 2.85 (0.07) 2.87 (0.07) 0.74 (0.02) 0.75 (0.02) 0.74
3 2.82 (0.11) 2.85 (0.08) 0.74 (0.03) 0.75 (0.02) 0.67
4 2.83 (0.04) 2.79 (0.04) 0.76 (0.01) 0.75 (0.01) 0.13

 

74

 

Table 3-3 Symmetry values in sound condition of vertical force peaks (N/kg) and vertical
impulses (N s/kg) between left and right limbs. Standard values represent left-right
symmetry of vertical force peaks and vertical impulses expressed as mean and (SD)
(Merkens, et al., 1993).

 

 

 

 

 

 

 

Horse no. Vertical force peak Vertical impulse Standard
Forelimbs Hind limbs F orelimbs Hind limbs values
1 94 97 93 98 97 (2)
2 100 99 95 97 97 (2)
3 99 97 96 99 97 (2)
4 98 100 95 98 97 (2)

 

 

 

 

 

 

 

GROUND REACTION FORCE VARIABLES

The three ground reaction force (GRF) components (vertical, longitudinal and
transverse) from both hind limbs in the sound condition are illustrated in Figure 3-1.
Vertical GRF was always positive. After a small impact spike, it gradually increased to
reach its maximal value around 50% of the stance phase then gradually decreased.
Longitudinal GRF had negative and positive phases. The initial negative (braking) impact
spike (Minl) was followed by a period of oscillation with peaks Maxl, Min2, Max2.
After that, there was a more prolonged negative phase with the peak braking force (Min3)
that acted to decelerate the horse’s forward movement during the first half of stance. It
was followed by a positive (propulsive) phase during the second half of stance, with peak
Max3 as a peak propulsive force. The longitudinal positive impulse had higher values
than longitudinal negative impulse, indicating the important role of the hind limb to
propel the body forward. Transverse GRF was least consistent among the three force
components. The transverse force was directed laterally during the impact period until
around 30% stance, after which the value remained small and quite variable between

horses through the end of the stance phase. The means, standard deviations, coefficients

75

of variation (% CV) of the force peaks and their time of occurrence (% stance duration),
together with the corresponding impulses are reported in Table 3-4. Due to the high %
CV of the transverse GRF, this variable was not used in the analysis comparing sound
and lame conditions.

The center of vertical load distribution between the four limbs could be calculated
to observe the pattern of load distribution in each horse. The craniocaudal distribution
was obtained from the percent contribution to the load from the front side by dividing
vertical impulse of the two forelimbs by total vertical impulse of the four limbs. The
distribution between left to right sides was calculated by dividing total left limb vertical
impulse by total vertical impulse of the four limbs. A value of 50% on the vertical axis
indicates equal contributions to vertical load from the forelimbs and hind limbs, whereas
a value of 50% on the horizontal axis indicates equal contributions from the left limbs
and right limbs. Table 3-5 reports the vertical impulses in the four horses used in this
study. Figure 3-2 illustrates the vertical impulse distribution plotted from the information

in Table 3-5.

76

 

 

Figure 3-1 Vertical (above), longitudinal (middle) and transverse (below) components of
the ground reaction forces from hind limbs of sound trotting horses. Thick solid line
indicates mean values from both hind limb. Thin solid lines indicate one standard
deviation above and below the mean.

10.0 - Max

Vertical force
(legl

 

 

 

 

Longitudinal force
(Niko)

 

 

 

 

 

8
g 0.5-
% g 00 "Li M
a v ' \/ \/' Min3 '
E new
mm
-1.0-
o 20 40 so so 100

Percent stance duration

77

 

 

Table 3-4 Ground reaction force (GRF) variables, expressed as mean values for both
hind limbs, in sound trotting horses. Values are mean and (SD).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variables Mean peak Coefficient Time of
force or impulse of variation occurrence
(%) (% Stance)
Vertical force Max (N/kg) 9.052 (0.402) 4 50.5 (1.6)
Vertical impulse (NS/kg) 1.453 (0.077) 5 -
Longitudinal force Maxl (N/kg) -0.134 (0.200) 149 10.0 (0.6)
Longitudinal force Max2 (N/kg) -0.638 (0.126) 20 17.9 (1.3)
Longitudinal force Max3 (N/kg) 1.037 (0.114) 11 75.1 (1.7)
Longitudinal force Minl (N/kg) -1.621 (0.237) 15 4.6 (0.7)
Longitudinal force Min2 (N/kg) -0.773 (0.108) 14 14.3 (0.9)
Longitudinal force Min3 (N/kg) -0.717 (0.068) 9 26.4 (1.7)
Longitudinal negative impulse (N s/kg) -0.073 (0.008) 11 -
Longitudinal positive impulse (NS/kg) 0.087 (0.016) 18 -
Transverse force Maxl (N/kg) -0.040 (0.097) 243 8.9 (1.2)
Transverse force Max2 (N/kg) 0.039 (0.247) 633 32.8 (3.2)
Transverse force Max3 (N/kg) 0.156 (0.238) 153 63.7 (6.0)
Transverse force Minl (N/kg) -0.435 (0.212) 49 4.3 (0.5)
Transverse 1 force Min2 (N/kg) -0.322 (0.135) 42 15.1 (1.4)
Transverse force Min3 (N/kg) 0.018 (0.271) 1506 46.5 (3.8)
Transverse negative impulse (N s/kg) -0.032 (0.032) 100 -
Transverse positive impulse (N s/kg) 0.022 (0.021) 95 -

 

78

 

 

Table 3-5 Vertical impulse (N s/kg) on all four limbs of each horse.

 

 

 

 

 

0 . . o . .
Left Right Left Right /o contribution /0 contribution
Horse . . from the from the left
fore fore hrnd hrnd . .
forelimbs limbs
l 1.90 1.76 1.32 1.34 57.91 50.89
2 1.97 1.88 1.50 1.47 56.44 50.97
3 1.97 2.05 1.50 1.51 57.22 49.39
4 1.99 1.90 1.48 1.50 56.66 50.47

 

 

 

 

 

 

 

 

Figure 3-2 Load distribution of vertical impulse of 4 sound horses. Each dot represents
the value from one horse.

59%
0
Front
0
I
0
55 % L ~ ' '

Figure 3-2 shows that all horses had more vertical load on forelimbs than the hind
limbs. Three horses had a higher percentage contribution to vertical loading from the left
limbs, and only one horse had a higher percentage contribution to vertical loading from

the right limbs. This might be an indication of sidedness.

79

 

TWO-DIMENSIONAL BIOMECHANICAL PROFILES

Coxofemoral joint. Figures 3-3 and 3-4 illustrate the joint angle, net joint
moment and net joint power profiles of coxofemoral joint during stance and swing
phases, respectively. The joint gradually extended throughout stance, with the net joint
moment on the extensor side of the joint through most of stance, changing to the flexor
side during breakover. Two positive power peaks corresponded with bursts of energy
generation on the extensor aspect of the coxofemoral joint from 10-40% stance and 45-
80% stance. During the swing phase, the joint had a single cycle of flexion and extension,
reaching its minimal angle around 60% swing. The net joint moment was on the flexor
aspect until 30% swing, then the value remained low until moving to the extensor aspect
around 75% swing. There were two peaks of positive net joint power corresponding with
a burst of energy generation on the flexor aspect in early swing and a burst of energy

generation on the extensor aspect in late swing.

Femorotibial joint. Figures 3-5 and 3-6 illustrate the joint angle, net joint
moment and net joint power profiles of the femorotibial joint during stance and swing
phases, respectively. During stance, the joint angle reveals fluctuating patterns of flexion
and extension with 2 flexion peaks around 20% and 60% stance. After oscillations during
impact, the net joint moments were on the flexor aspect of the joint, moving to the
extensor aspect during breakover. During the impact period, the femorotibial joint had a
large positive net joint power, indicative of energy generation and positive work
performed on the flexor aspect of the joint. Between 20% to 60% of stance, the joint

extended then flexed against a flexor moment, resulting in negative net joint power and

80

energy absorption followed by a similar magnitude of positive net joint power and energy
generation. This profile is the typical pattern for elastic energy storage and release by
tendons and ligaments. After 60% stance, little work was performed around the
femorotibial joint.

In the first half of swing, the joint flexed against an extensor moment to reach the
minimal flexion around 50% swing. In the second half of swing, the joint extended
against a flexor moment. Consequently, net joint power was negative through most of the
swing phase, with two distinct negative peaks around 10% and 80% swing. Therefore,
flexion and extension of the femorotibial joint during the swing phase were controlled by

eccentric activity in the extensor and flexor muscles, respectively.

Tarsal joint. Figures 3-7 and 3-8 illustrate the joint angle, net joint moment and
net joint power profiles of the tarsal joint during stance and swing phases, respectively.
The tarsal joint extended during the initial 5% stance, after which the joint flexed as the
limb accepted the body weight and reached the first flexion peak (Minl) around 20%
stance. The joint slightly extended then flexed to reach the second flexion peak (Min2)
around 60% stance. The net joint moment had small magnitude oscillations during
impact, then remained on the extensor aspect throughout stance. The net joint power
profile appeared to show two cycles of elastic energy storage and release. The first cycle,
which occupied the initial 30% stance, had a relative large burst of energy absorption
followed by a smaller burst of energy generation. The second cycle immediately followed
the first cycle. It had a longer duration with a larger burst of energy generation than the

first cycle.

81

During the swing phase, the tarsal joint flexed then extended with a similar
pattern to the femorotibial joint, reaching its minimal flexion angle around 50% swing.
The net joint moment was on the flexor aspect during the first 50% of swing, then moved
to the extensor aspect. The net joint moment and joint angle had the same polarity
throughout swing: the joint flexed with a flexor moment in early swing and extended with
an extensor moment in late swing. The corresponding positive net joint power peaks
occurred around 25% (Maxl) and 85% (Max2) swing. Positive work was performed
across the tarsal joint to move the joint during the swing phase by two bursts of energy

generation, first on the flexor aspect then on the extensor aspect of the joint.

Metatarsophalangeal joint. Figures 3-9 and 3-10 illustrate the joint angle, net
joint moment and net joint power profiles of the metatarsophalangeal joint during stance
and swing phases, respectively. The metatarsophalangeal joint extended then flexed
during the stance phase, with maximal extension around 50% stance. The net joint
moment was on the flexor aspect of the joint throughout the stance phase. Net joint power
was negative as the joint extended against a flexor moment and positive as the joint
flexed with the flexor moment. This resulted in a burst of energy absorption followed by
a burst of energy generation, which were of similar magnitude. It was a classic profile of
elastic energy storage and release by ligaments and tendons.

During the swing phase, the joint flexed to reach a peak (Minl) around 25%
swing. The corresponding net joint moment was on the extensor aspect, resulting in
negative net joint power and a burst of energy absorption. From 35% swing, the joint

gradually extended against a small flexor moment. Soft tissues around the

82

metatarsophalangeal joint performed mainly negative work, with the largest burst of

energy absorption on the extensor aspect, acting to control flexion in early swing.

Distal interphalangeal joint. Figures 3-11 and 3-12 illustrate the joint angle, net
joint moment and net joint power profiles of the distal interphalangeal joint during stance
and swing phases, respectively. During the stance phase, the joint flexed to accept body
weight with the net moment on the flexor aspect, which resulted in positive net joint
power and a small burst of energy generation during 0-40% stance. The joint reached its
minimal flexion angle around 40% stance then extended against a flexor moment, which
was associated with a large burst of energy absorption. During breakover, the joint
reached maximal extension then flexed with the flexor moment, yielding a very small
burst of positive work in terminal stance.

During the swing phase, the joint flexed then experienced fluctuating motion like
the metatarsophalangeal joint. The net joint moment and net joint power profiles were
also similar to those from the metatarsophalangeal joint, but the values were an order of

magnitude smaller.

83

Figure 3-3 Joint angle, net joint moment and net joint power of coxofemoral joint during
the stance phase. Thick solid line indicates mean values from both hind limbs of sound
trotting horses. Thin solid lines indicate one standard deviation above and below the

 

 

 

 

 

 

 

mean.
140' Max2
0 A
.5, to
c g 120- Max1
“' a
a
C o
E 3 in2
1001
mm
80 f T r T 1
0 20 4O 60 80 100
’E 1'5 Max1
or:
E A 1'0 Max2 Max3
2 °’ 0.5 A
g E 004\ /\ .
.2 v ' \V/ N
*6 -o.5-
z in1
-1.0-
-1.5‘
0 20 40 60 80 100
4.0a
3
g A 2.0. Max1 Max2
“ 2’
fl
:
:§ 5
'2‘
-2.04
-4.0-
O 20 40 60 80 100

Percent stance duration

84

Figure 3-4 Joint angle, net joint moment and net joint power of coxofemoral joint during
the swing phase. Thick solid line indicates mean values fi'om both hind limbs of sound
trotting horses. Thin solid lines indicate one standard deviation above and below the
mean.

 

 

 

 

 

 

150 7
130 4
3 a
at
110 -
s f
o- c»
I: .
.5 g 90
3
7° ' Min
50 I r I V 1
0 20 40 60 80 100
1 4
H
r:
g 0.5 -
0 ’5 Max
E E o /
a I fl
:
E E
H
0 -0.5 - -
2 MIN
-1 a
0 20 40 60 80 100
2.5 - Max2
1.5 - Max1

Net joint power
(W/k9)

 

0 20 4O 60 80 1 00

Percent swing duration

85

Figure 3-5 Joint angle, net joint moment and net joint power of femorotibial joint during
the stance phase. Thick solid line indicates mean values from both hind limbs of sound
trotting horses. Thin solid lines indicate one standard deviation above and below the
mean.

 

 

130-

o

— A Max1

2’ g Max2

,2 a 110-

'5 i mm mm

a
90 T I I I I

0 20 40 60 80 100

1.0-

 

 

Net joint moment
(lekg)

 

 

Net joint power
(W/kgl

 

0 20 4O 60 80 1 00

Percent stance duration

86

Figure 3-6 Joint angle, net joint moment and net joint power of femorotibial joint during
the swing phase. Thick solid line indicates mean values from both hind limbs of sound
trotting horses. Thin solid lines indicate one standard deviation above and below the
mean.

 

 

 

 

140 -

120 . Max
9 a
a:

100 -
s f
o- a:
c .1
.5 i 80
.j

60 - Min

40 I I I fi I

O 20 40 60 80 100

0.4 -
2% Max
g A 0.2 K

as

E N
g E 0 1 w I j
_c_>_ v
‘6 -o.2 -
2

-0.4 -

0 20 4O 60 80 100

SEW W

Net joint power
(Wlkgl

 

-1.0 -
Min2
-1.5 d
0 20 40 60 80 100

Percent swing duration

87

Figure 3-7 Joint angle, net joint moment and net joint power of tarsal joint during the
stance phase. Thick solid line indicates mean values from both hind limbs of sound
trotting horses. Thin solid lines indicate one standard deviation above and below the
mean.

 

 

 

 

130-
2 A Max1
a. U)
c
Q g 110i Max2
-- O
.E o
0 3
” mm mm
90 I r I I I
0 20 40 60 80 100
2.01
g 1.5 d Max2
g A
I» q
E Max1
5 5 0.51 -
"‘ / m2
é’ 0.0 E: . . . . ~
mm
-0.5
0 20 40 60 80 100
4.0-
Max2

 

Net joint power
(W/kg)

 

Min1

 

Min2
-4.0 4
0 20 40 60 80 100

Percent stance duration

88

Figure 3-8 Joint angle, net joint moment and net joint power of tarsal joint during the
swing phase. Thick solid line indicates mean values from both hind limbs of sound
trotting horses. Thin solid lines indicate one standard deviation above and below the

mean.
140 -

120 - Max
100 ~

80a

Joint angle
(degrees)

60 . Min

 

 

40 r V U I fi
0 20 40 60 80 100

0.2 -

 

 

 

Net Joint moment
(lekg)

 

0 20 40 60 80 100

Max1 Max2

 

 

Net joint power
(WIkg)

 

0 20 40 60 80 1 00

Percent swing duration

89

Figure 3-9 Joint angle, net joint moment and net joint power of metatarsophalangeal joint
during the stance phase. Thick solid line indicates mean values from both hind limbs of
sound trotting horses. Thin solid lines indicate one standard deviation above and below

the mean.

Net joint moment Joint angle

Net Joint power

(degrees)

(Nm/kg)

MM)

260 -
Max
240 -
220 -
200 -

180-

 

 

160 I I I r
0 20 40 60 80 100

 

 

Min

 

0 20 40 60 80 1 00

8.0 - Max

4.0 '-

 

0.0 " I I r t

-4.0 -
Min

 

-8.0 ‘

0 20 40 60 80 1 00

Percent stance duration

90

Figure 3-10 Joint angle, net joint moment and net joint power of metatarsophalangeal
joint during the swing phase. Thick solid line indicates mean values from both hind limbs
of sound trotting horses. Thin solid lines indicate one standard deviation above and below

the mean.

220.0 -

200.0 -

180.0 -

160.0 -

Joint angle
(degrees)

140.0 -

Max2

Max1

Min2

 

 

1 20.0

D

0.03 -
0.02 -

Net joint moment
(lekg)

 

20 100

 

-0.10 -

(W/kg)

Net joint power

-0.20 -

 

-0.30 -

Min1

 

20 40 60 80 1 00

Percent swing duration

91

Figure 3—11 Joint angle, net joint moment and net joint power of distal interphalangeal
joint during the stance phase. Thick solid line indicates mean values from both hind limbs
of sound trotting horses. Thin solid lines indicate one standard deviation above and below
the mean.

 

 

 

 

 

 

 

220w Max
zool
1’ 1;
O
180-
s t
g 8, 1601
O 3
.,
1404 Min
120 I T I I I
0 20 40 60 80 100
u..-
5
g A o I I I I
Oi
E E
E
-- z
0 v
E
Min
-0.5d
0 20 40 60 80 100
is
3
O A r u
°' i”
H
C
:6; E
g,- -2.0-
2
-4.0J '"
0 20 40 60 80 100

Percent stance duration

92

Figure 3-12 Joint angle, net joint moment and net joint power of distal interphalangeal
joint during the swing phase. Thick solid line indicates mean values from both hind limbs
of sound trotting horses. Thin solid lines indicate one standard deviation above and below
the mean.

 

 

 

 

 

 

230.0 -
210.0 -
0 A
T» ‘0 Max2
g 8 190.0 '
E a, Max1
E .5: 170.0 -
150.01 M'" m2
130-0 I I I I I
0 20 40 60 80 100
0.004 -
.- Max
5 0 002
8 A ' /\
m
c E I V\
'6 5
3: Min
0 -0.002 -
2
-0.004 ‘
0 20 40 60 80 100
0.020 -
a; Max1 Max2
/
a g, 0.000 V . V .
a -
: Min2
:§. 5
7, -0.020 -
Z
Min1
-0.040 J
0 20 40 60 80 100

Percent swing duration

The values of joint angles, net joint moments, net joint powers and net joint
energies of all joints are reported in Tables 3-6 to 3-9 for the stance phase and in Tables
3-10 to 3-13 for the swing phase. During the stance phase, the distal interphalangeal joint
had the largest range of motion, followed by the metatarsophalangeal joint (Table 3-6).
The more proximal joints had a considerably smaller range of motion than the distal
joints. The highest net joint moment peaks were the extensor moment at the tarsal joint
(Max2), followed by the flexor moment at the metatarsophalangeal joint (Min) and the
extensor moment at the coxofemoral joint (Max1) (Table 3-7). The highest positive
power peaks were at the metatarsophalangeal joint (Max), followed by the tarsal joint
(Max2) and femorotibial joint (Maxl). Therefore, these three joints had high rates of
performing positive work. The highest negative power peak was on the flexor aspect of
the metatarsophalangeal joint (Min), followed by the distal interphalangeal joint (Min)
and tarsal joint (Minl) (Table 3-8). Therefore, these three joints had high rates of
performing negative work. The femorotibial, tarsal and metatarsophalangeal joints had
similar magnitudes of total positive and total negative net joint energies (Table3-9) that
reflected the patterns of elastic energy storage and release. The coxofemoral joint
performed mostly positive work to protract and retract the limb, while the distal
interphalangeal joint performed mostly the negative work.

During the swing phase, all joints had similar ranges of motion (Table 3-10). The
magnitude of the net joint moment decreased in a proximal to distal sequence (Table 3-

1 1). The coxofemoral joint had the largest magnitude of net joint moment peak (Min1) on
its flexor aspect, and the distal interphalangeal joint had the smallest magnitude net joint

moment peak (Min) on its flexor aspect. The absolute values of the net joint power peaks

94

also decreased from the proximal to distal joints (Table 3-12). The largest net joint power
peak was on the extensor aspect of the coxofemoral joint (Max2) during late swing. From
the net joint energy profiles (Table 3-13), the coxofemoral joint performed the largest
amount of positive work to swing the limb forward, assisted by positive work at the tarsal
joint. The femorotibial joint performed the largest amount of negative work to control
joint movement. The metatarsophalangeal and distal intertarsal joints showed only small
amounts of both energy generation and absorption, which indicated that their motion was
driven by inertia and segmental energy transfers from the more proximal joint during
swing.

Magnitudes of the mean net energy generation and absorption of the hind limb
joints during stance phase, swing phase and total stride are summarized in Table 3-14.
The net energy generation and absorption during stance had similar magnitudes, with a
small net mechanical energy absorption of -0.0l 8 J/kg. During swing, there was a net
mechanical energy generation of 0.1 194 J/kg. This resulted in a net mechanical energy

generation ofO. 1014 J/kg during the whole stride.

95

Table 3-6 Joint angle peaks (degrees) averaged from both hind limbs in sound trotting

horses during the stance phase. Values are mean and (SD).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Joint angle Time of
occurrence
(% Stance)
Coxofemoral Max1 109.78 (7.46) 41.0 (3.4)
Max2 125.20 (6.64) 88.0 (6.6)
Min1 101.12 (7.18) 17.8 (4.9)
Min2 114.84 (7.78) 60.2 (4.0)
Range 26.27 (1.23) -
Femorotibial Max1 113.24 (1.90) 40.3 (2.8)
Max2 109.97 (2.19) 85.6 (5.4)
Min1 108.46 (2.00) 22.3 (1.2)
Min2 107.36 (2.08) 63.3 (3.6)
Range 10.97 (2.67) -
Tarsal Max1 114.79 (2.87) 5.2 (0.7)
Max2 106.45 (3.74) 37.36 (4.30)
Min1 105.44 (4.20) 25.9 (3.6)
Min2 103.33 (3.41) 58.2 (2.4)
Range 12.12 (0.96) -
Metatarsophalangeal Max 237.34 (11.49) 55.2 (2.9)
Range 47.79 (9.16) -
Distal interphalangeal Max 206.92 (6.45) 95.8 (1.2)
Min 151.55 (10.20) 44.9 (2.3)
Range 55.66 (6.42) -

 

96

 

Table 3-7 Net joint moment peaks (Nm/kg) averaged from both hind limbs in sound

trotting horses during the stance phase. Values are mean and (SD).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Joint moment Time of
occurrence
(% Stance)
Coxofemoral Max1 1.101 (0.129) 10.0 (0.5)
Max2 0.575 (0.126) 23.8 (3.6)
Max3 0.557 (0.185) 60.0 (3.4)
Mini -0.372 (0.248) 4.5 (0.7)
Min2 0.367 (0.096) 15.2 (2.2)
F emorotibial Max1 0.365 (0.100) 4.5 (0.7)
Max2 0.135 (0.055) 94.7 (2.0)
Mini -0.59 (0.131) 9.8 (0.6)
Min2 -0.537 (0.144) 4.07 (2.1)
Max1 0.422 (0.086) 10.3 (0.6)
Tarsal Max2 1.406 (0.154) 50.4 (2.1)
Mini -0.132 (0.052) 4.4 (0.7)
Min2 0.325 (0.076) 13.7 (0.5)
Metatarsophalangeal Min -1.l49 (0.140) 53.8 (1.7)
Distal interphalangeal Min -0.304 (0.058) 62.4 (2.3)

 

97

 

Table 3-8 Net joint power peaks (W/kg) averaged from both hind limbs in sound trotting

horses during the stance phase. Values are mean and (SD).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Joint power Time of
occurrence
(% Stance)

Coxofemoral Max1 1.501 (0.419) 25.3 (5.3)
Max2 1.473 (0.554) 63.3 (4.0)

Femorotibial Maxl 2.488 (0.875) 10.2 (0.6)
Max2 1.150 (0.493) 49.9 (3.2)

Mini -0.503 (0.188) 4.7 (0.8)

Min2 -1.085 (0.384) 32.1 (2.0)

Tarsal Max1 0.862 (0.762) 31.7 (4.1)
Max2 2.627 (0.565) 68.2 (2.7)

Mini -2.219 (0.496) 17.9 (1.9)

Min2 -2.159 (1.073) 47.5 (4.0)

Metatarsophalangeal Max 6.237 (1.284) 77.5 (1.4)
Min -4.653 (1.316) 31.4 (5.3)

Distal interphalangeal Max 0.588 (0.258) 28.0 (4.5)
Min -3.037 (0.625) 77.1 (1.6)

 

98

 

Table 3-9 Net joint energies (J/kg) calculated by time integration of the corresponding
net joint power peaks during the stance phase. Values were averaged from both hind
limbs in sound trotting horses. Values are mean and (SD).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Joint energy
Coxofemoral Max1 0.063 (0.023)
Max2 0.073 (0.034)

Total positive energy 0.140 (0.140)
Total negative energy -0.014 (0.009)

Femorotibial Max1 0.037 (0.016)
Max2 0.041 (0.018)
Min1 -0.004 (0.001)
Min2 -0.031 (0.011)

Total positive energy 0.079 (0.027)
Total negative energy -0.043 (0.015)

Tarsal Max1 0.024 (0.014)
Max2 0.125 (0.023)
Min1 -0.075 (0.016)
Min2 -0.076 (0.043)

Total positive energy 0.147 (0.023)

Total negative energy -0.151 (0.038)

Metatarsophalangeal Max 0.394 (0.094)
Min -0.386 (0.108)

Total positive energy 0.394 (0.094)

Total negative energy -0.387 (0.108)

Distal interphalangeal Max 0.038 (0.018)
Min -0.222 (0.049)

Total positive energy 0.040 (0.017)

Total negative energy -0.223 (0.049)

 

 

 

99

 

Table 3-10 Joint angle peaks (degrees) averaged from both hind limbs in sound trotting

horses during the swing phase. Values are mean and (SD).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Joint angle Time of
occurrence
(% Swing)
Coxofemoral Min 76.85 (5.43) 64.7 (1.1)
Range 48.63 (1.41) -
Femorotibial Max 114.97 (4.03) 90.2 (2.6)
Min 68.35 (3.09) 50.2 (2.9)
Range 50.44 (2.63) -
Tarsal Max 113.38 (4.99) 91.1 (2.6)
Min 70.77 (3.09) 54.5 (2.7)
Range 44.67 (1.90) -
Metatarsophalangeal Maxl 160.54 (6.94) 42.9 (7.8)
Max2 185.90 (7.63) 88.7 (2.0)
Min1 147.70 (6.67) 25.3 (2.5)
Min2 156.61 (6.76) 56.3 (7.2)
Range 45.31 (3.38) -
Distal interphalangeal Max1 170.08 (6.34) 31.9 (4.2)
Max2 180.41 (8.29) 90.9 (6.7)
Min1 162.60 (5.69) 19.1 (3.4)
Min2 155.33 (6.13) 46.1 (6.4)
Range 45.29 (5.05) -

 

100

 

Table 3-11 Net joint moment peaks (Nm/kg) averaged fi'om both hind limbs in sound
trotting horses during the swing phase. Values are mean and (SD).

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Joint moment Time of
occurrence
(% Swing)

Coxofemoral Max 0.1190 (0.0299) 39.1 (1.1)
Min1 -0.4313 (0.0575) 16.8 (12.1)

Min2 -0.1347 (0.0650) 60.1 (4.5)

Femorotibial Max 0.1993 (0.0285) 8.9 (1.9)
Tarsal Max 0.0185 (0.0115) 48.8 (7.3)
Min -0.0928 (0.0160) 10.3 (2.2)

Metatarsophalangeal Max 0.0170 (0.0021) 13.9 (3.0)
Min -0.0070 (0.0025) 44.0 (5.8)

Distal interphalangeal Max 0.0023 (0.0007) 14.7 (4.4)
Min -0.0010 (0.0002) 41.7 (6.7)

 

 

 

 

 

 

101

Table 3-12 Net joint power peaks (W/kg) averaged from both hind limbs in sound
trotting horses during the swing phase. Values are mean and (SD).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Joint power Time of
occurrence
(% Swing)

Coxofemoral Max1 1.1630 (0.1830) 18.3 (3.4)
Max2 1.8785 (0.6698) 92.0 (4.7)

Min -O.6299 (0.1565) 38.8 (1.0)

Femorotibial Max 0.1595 (0.1381) 41.0 (3.7)
Min1 -0.5193 (0.1201) 12.8 (7.4)

Min2 -0.9414 (0.3216) 86.1 (3.6)

Tarsal Max1 0.3340 (0.0590) 25.6 (5.0)
Max2 0.3179 (0.1124) 86.2 (5.8)

Metatarsophalangeal Maxl 0.0187 (0.0121) 30.6 (3.0)
Max2 0.0150 (0.0057) 48.5 (6.9)

Min1 -0.1663 (0.0245) 12.3 (1.7)

Min2 -0.0167 (0.0101) 40.5 (6.4)

Min3 -0.0315 (0.0086) 80.6 (5.0)

Distal interphalangeal Maxl 0.0056 (0.0028) 25.4 (3.3)
Max2 0.0059 (0.0017) 39.9 (6.1)

Minl -0.0230 (0.0117) 8.6 (1.9)

Min2 -0.0055 (0.0024) 87.1 (3.8)

 

102

 

Table 3-13 Net joint energies (J/kg) calculated by time integration of the corresponding
net joint power peaks during the swing phase. Values were averaged from both hind
limbs in sound trotting horses. Values are mean and (SD).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Joint energy
Coxofemoral Max1 0.1041 (0.0095)
Max2 0.1223 (0.0419)
Min -0.0269 (0.0101)
Total positive energy 0.2291 (0.0365)
Total negative energy -0.0277 (0.0100)
Femorotibial Min1 -0.0485 (0.0103)
Min2 -0.0786 (0.0131)
Total positive energy 0.01 19 (0.0088)
Total negative energy -0.1290 (0.0207)
Tarsal Max1 0.0291 (0.0075)
Max2 0.0221 (0.0088)
Min -0.0016 (0.0014)
Total positive energy 0.0520 (0.0129)
Total negative energy -0.0040 (0.0029)
Metatarsophalangeal Max1 0.0006 (0.0004)
Max2 0.0004 (0.0002)
Minl -0.0094 (0.0019)
Min2 -0.0004 (0.0002)
Min3 -0.0024 (0.0007)
Total positive energy 0.0011 (0.0005)
Total negative energy -0.0131 (0.0019)
Distal interphalangeal Maxl 0.0002 (0.0001)
Max2 0.0002 (0.0001)
Min1 -0.0009 (0.0004)
Min2 -0.0004 (0.0001)
Total positive energy 0.0004 (0.0002)
Total negative energy -0.0013 (0.0002

 

103

 

Table 3-14 Summary of means energy generation and absorption (J/kg) at the joints of
the hind limb during stance phase, swing phase and the total stride.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Energy Energy Energy Energy Net
generated absorbed generated absorbed energy
stance stance swing swing stride
Coxofemoral 0.140 -0.014 0.2291 -0.0277 0.3274
F emorotibial 0.079 -0.043 0.01 19 -0. 1290 -0.081 1
Tarsal 0.147 -0.151 0.0520 -0.0040 0.044
Metatarsophalangeal 0.394 -0.387 0.001 1 -0.013 1 -0.005
Distal 0.040 -0.223 0.0004 -0.0013 -0. 1839
interphalangal
Total 0.800 -0.818 0.2945 -0.1751 0.1014
DISCUSSION

Kinematic data. The application of skin correction algorithms to the raw
kinematic data is very important, especially proximal to the tarsal joint. The coupled
motion of the femorotibial and tarsal joints via the function of the reciprocal apparatus
could not be identified in data without correction for skin motion (Kobluk, et al., 1989),
but could be observed afier the movement of skin markers had been corrected (van den
Bogert, 1990; van Weeren, et al., 1990). The models for correction of skin displacement
may not be able to account for all individual variation. It has been suggested that only
70% of skin displacement at the femorotibial joint center of rotation (femoral condyle)
can be predicted correctly (van Weeren, et al., 1992a). Nevertheless, much improvement
was found for the reciprocal movement between femorotibial and tarsal joints flexion and
extension after the application of skin correction algorithms to the femorotibial joint,

including the joint extension during the end of stance and joint flexion/extension during

104

swing (van den Bogert, et al., 1990; van Weeren, et al., 1992b; Back, et al., 1995b). Even
the skin correction algorithm could not eliminate all the errors from skin displacement,
but it helps improve the kinematic information. Therefore, the skin correction algorithms
should be applied in studies based on data from skin markers, especially in the limb
segment proximal to the tarsal joint.

Distal interphalangeal joint angles have the highest variation among hind limb
joints (Back, et al., 1995b; Degueurce, et al., 1997; Pourcelot, et al., 1997b). The
coefficient of variation for the range of motion of the distal interphalangeal joint in this
study was 12% during stance and 11% during swing, which is similar to the value (14%)
reported during the whole stride (Back, et al., 1995b). Radiographic identification of the
distal interphalangeal joint center of rotation reduced inter-individual variations due to
errors in marker placement on the hoof wall.

Trotting velocity affects range of motion at the hind limb joints. An increase in
trotting velocity resulted in increases in tarsal joint flexion, tarsal and coxofemoral joint
ranges of motion, and range of retraction-protraction of the entire limb (Galisteo, et al.,
1998). Velocity also seems to influence the tarsal joint angle profile. In Dutch
warrnbloods trotting at 4 m/s (Back, et al., 1995b) the angle profile showed two flexion
peaks similar to the pattern in this study, whereas Standardbreds trotting at 10.7 to 11.5
m/s (Johnston, et al., 1996) had different profiles with a single flexion peak around 40%
stance. At faster trotting velocities, peak tarsal flexion tends to occur earlier than at
slower velocities (Clayton, personal communication). Therefore, comparisons of angle-
time graphs for the tarsal joint should be performed cautiously and only within a similar

range of trotting velocities.

105

The joint angle diagrams of the hind limb joints in this study were similar to those
reported previously (Back, et al., 1994; Back, et al., 1995b). During early stance, the hind
limb was compressed by flexion of the femorotibial, tarsal, and distal interphalangeal
joints and hyperextension of the metatarsophalangeal joint (Hjerten, et al., 1994) while
the coxofemoral joint gradually extend thorough the stance phase. The decrease in total
limb length during stance was described as a mechanism to absorb impact shock (Back,
et al., 1995b). Movements of the femorotibial and tarsal joints were completely
synchronized by the reciprocal apparatus during the swing phase. During the stance
phase, small discrepancies of flexion and extension between the femorotibial and tarsal
joint could be observed during impact and prior to lift off. In Back, et al. (1995b), the
femorotibial and tarsal joint profiles during the early stance before and after skin
correction showed the same profile in each joint, therefore, the profiles of these two joints
obtained from the study that did not use skin correction could be compared during early
stance. During impact, tarsal joint extended while the femorotibial joint slightly flexed,
similar to the finding from previous studies (Holmstrom, etal., 1994; Back, et al., 1995b;
Holmstrom, et al., 1995b). This discrepancy during early stance was not observed at the
walk (Galisteo, et al., 1996; Hodson, et al., 2001). The higher GRFs in the trot compared
with the walk might be sufficient to stretch the peroneus tertius, which shows a brief
period of increased strain during this time (van Weeren, et al., 1992b). At the end of
stance, the femorotibial joint maintained its angle while the tarsal joint extended.
Consequently, the peroneus tertius was stretched and reached its maximal strain just prior
to lift off. Strain decreased rapidly after lift off, which indicated a release of stored

energy (van Weeren, et al., 1992b).

106

Force plate data. Vertical and longitudinal ground reaction forces (GRFs) and
impulses in this study had similar magnitudes but lower % CV than Merkens, et al.
(1993). In that study, the velocity ranged from 3.9 to 4.3 m/s. Subject velocity influences
peak vertical force and impulse at walk (Khumsap, et al., 2001b) and vertical impulse at
trot (McLaughlin, et al., 1996). Due to the velocity dependent effects on GRF, subject
velocity was controlled by standardizing the dynamic similarity and consequently the
variation between horses was reduced.

The center of load distribution indicated that the forelimbs contributed more than
the hind limbs in accepting the total body load. In this study, the horses’ forelimb carried
56.44% to 57.91%, which is in accordance with the fact that the center of gravity is
located closer to the forelimbs (Sprigings and Leach, 1986). Each horse has a slight
asymmetry loading at the trot between the left and right sides in the forelimb or hind limb
pairs (Merkens, et al., 1993). Quarter horses showed a lead limb preference (Deuel and
Lawrence, 1987), and sidedness may evident in Standardbred trotters as young as 8
months old (Drevemo, et al., 1987). A slight asymmetry in vertical impulse might be an
indication of sidedness in each horse that is equivalent to handedness in people. In this
study, three of the four horses loaded the left limbs a little more than the right limbs.

Impact oscillations of the GRF ended earlier in the vertical GRF than the braking
GRF, which is similar to the results in previous studies (Merkens and Schamhardt, 1994;
Schamhardt and Merkens, 1994; Clayton, 2001). The hind limb made heel first contact
with the ground just prior to the start of the main loading phase (Merkens and
Schamhardt, 1994). The hind hoof had more horizontal velocity than vertical velocity

during trotting impact (Back, et al., 1995c). This resulted in a little forward sliding of

107

hind hoof at impact, which may be a mechanism to reduce the oscillation in the
metatarsophalangeal joint. Therefore, the metatarsophalangeal joint angle curve was
smoother than that of the metacarpophalangeal joint during early stance (Merkens and
Schamhardt, 1994; Back, et al., 1995c).

There have been limited studies reporting GRFs in the hind limb of trotting
horses. In contrast, there have been several studies of GRFs in normal trotting dogs and
dogs with orthopedic problems. Studies in humans have shown a similar GRF pattern to
the equine hind limb, but most of the human clinical studies focus on GRF analysis at the
walk. The studies in humans and dogs offer some insight into the role of normal joints on
GRF generation and the effect of specific pathologies on those forces. In dogs with
degenerative joint disease of the coxofemoral joint, total hip replacement resulted in
significant increases in vertical and propulsive impulses in the treated limb (Budsberg, et
al., 1996). This suggests that the mechanical role of the coxofemoral joint is to generate a
propulsive force and impulse to push the body upward and forward.

In humans, quadriceps action modulates the magnitude of the heel strike transient,
and quadriceps paralysis resulted in a large increase in the heel strike transient (Jefferson,
et al., 1990). In dogs with induced-synovitis of the femorotibial joint (Cross, et al., 1997),
and cats with degeneration of the femorotibial joint (Suter, etal., 1998), GRF analysis
indicated decreases in peak vertical and braking GRFs without a significant change in
propulsive impulse. These findings suggest that the femorotibial joint plays a role in
deceleration during the first half of stance.

Comparison of canine GRFs, before and after temporary impairment of the tibial

nerve, revealed a significant decrease in vertical GRF without a change in braking or

108

propulsive forces (Rumph and Steiss, 1998). The function of the tarsal joint, therefore,

seems to be important in generating vertical force, rather than longitudinal forces.

Inverse dynamic analysis. Inverse dynamic analysis is sensitive to the protocol
for data collection and the method for data analysis. The fact that the horse is not a set of
perfectly rigid body segments, and kinematic data are contaminated by measurement
errors and soft tissue deformation, affects the results of inverse dynamic analysis. The
results are acceptable, however, as long as the same methodology is used to evaluate and
compare locomotion in each horse (van den Bogert, 1998).

The two important forces that the limb joints have to overcome are the GRF and
gravitational force. During stance, the main effect of the net joint moments is to resist the
collapse of the limb joints under the influence of a large GRF. The forces needed to
overcome gravity and to accelerate the limb can be considered negligible. During swing,
the only external force acting on the limb is the gravitational force, and the muscular
work performed across the proximal joints to accelerate and decelerate the limb causes
large inertial forces in the distal limb. Therefore, it is important to apply inertial
parameters for inverse dynamic analysis of the limb segments during swing (Schamhardt,
1998; Lanovaz and Clayton, 2001).

The equine hind limb has many muscles that cross two or more joints and have
complicated actions. Several coxofemoral extensors also act as femorotibial flexors,
while femorotibial extensors also act as coxofemoral flexors. The reciprocal apparatus
also has an important role in coordinating movement of the femorotibial and tarsal joints.

Some of tarsal extensors, such as the gastrocnemius and deep digital flexor muscle, also

109

influence femorotibial joint flexion. The same muscle or tendon may perform positive
work at one joint and negative work at another joint. Calculation of power flows between
segments (Colborne, et al., 1997b) may give more information about these energy
transfers, but the technique is beyond the scope of this study. The objective of using
inverse dynamic analysis here is to develop a deeper understanding of limb function and

to relate these findings to the previous literature.

Coxofemoral joint. Extensor muscles performed positive work to extend the joint
throughout stance. Electromyographic (EMG) studies (Wentink, 1978; Tokuriki and
Aoki, 1995; Robert, et al., 1998; Robert, et al., 1999) have indicated that gluteus medius,
biceps femoris and semitendinosus are active during these bursts of energy generation.
EMG activity of tensor fascia latae commenced during the second half of stance, but its
activity was initially less than that of the extensor muscles, resulting in a net moment on
the extensor aspect. The net moment moved to the flexor aspect during late stance due to
either a decrease in extensor activity or an increase in flexor activity or both. Since joint
angle did not change during late stance, net muscular work was diminished, probably due
to the simultaneous development of tension in the coxofemoral extensors and flexors at
this time.

Tensor fascia latae activity continued until the middle of swing, during which
time the muscle performed positive work to flex the joint and protract the limb.
Semitendinosus also showed activity during early swing that ceased just before the
middle of swing, but its activity was probably more important for flexing the femorotibial

joint. During the second half of swing, EMG activity was present in semitendinosus,

110

gluteus medius and biceps femoris (Wentink, 1978; Tokuriki and Aoki, 1995; Robert, et
al., 1998), which performed positive work to extend the coxofemoral joint in preparation

for impact.

F emorotibial joint. The femorotibial joint plays a role in controlling the braking
force (Jefferson, et al., 1990; Cross, etal., 1997; Suter, etal., 1998). In this study, an
extensor moment was present with negative power during 0-5% stance, which
corresponded with the first negative longitudinal force peak (Min1) that acted to
decelerate the forward movement of the hoof. The following oscillation in the
longitudinal forces matched the oscillation pattern in the femorotibial net joint moment in
early stance. EMG studies in horses reported activity of rectus femoris and vastus
lateralis from terminal swing into the stance phase (Wentink, 1978; Robert, etal., 1999).
Our findings indicated that the role of these muscles was to perform negative work on the
extensor aspect of the femorotibial joint, which had the effect of smoothing the
longitudinal impact shock. The biarticular coxofemoral extensors, which also acted as
femorotibial flexors, were active during stance. These muscles performed positive
muscular work to flex the joint until 20% stance. The joint extended from 20-40% stance,
corresponding to the extension of the coxofemoral joint. During this period, the GRP
vector passed cranial to both the coxofemoral and femorotibial joints, which tended to
flex the former joint and extend the latter joint. The muscles needed to generate forces
that counteracted the GRF to control the motion, which resulted in a net extensor moment
at the coxofemoral joint and a net flexor moment at the femorotibial joint. The

coxofemoral joint was actively extended by positive work of semitendinosus and biceps

lll

femoris. The femorotibial joint passively extended, controlled by a flexor moment
produced by the coxofemoral extensors. This resulted in energy absorption at the
femorotibial joint during 20-40% stance. During this period, the tarsal joint also extended
but to a lesser degree than the femorotibial joint, which caused stretching of superficial
digital flexor (SDF) tendon, which shows an increase in load at this time (Riemersma, et
al., 1988). Hence, energy absorption on the flexor aspect of the femorotibial joint
appeared to correspond with elastic energy storage in the SDF tendon. Elastic energy was
released as the tendon recoiled between 40-60% stance as the femorotibial joint flexed. In
late stance, EMG activity in the vastus lateralis and tensor fascia latae created a net
extensor moment at the femorotibial joint but the joint angle changed only slightly during
this period, so very little mechanical work was performed.

Energy was absorbed through most of swing. During the first half of swing, the
femorotibial joint flexed against a net extensor moment. That was likely provided by
tensor fascia latae, which was actively flexing the coxofemoral joint. This resulted in net
extensor moment and net negative work at the femorotibial joint. During the second half
of swing, the femorotibial joint extended against a net flexor moment, which was likely
due to contraction of the biarticular semitendinosus and biceps femoris that were actively
extending the coxofemoral joint during this time. The bursts of energy absorption at the
femorotibial joint (Min1, Min2) occurred at approximately the same time as energy
generation at the coxofemoral joint (Max 1 , Max2), which supports the importance of bi-

articular muscle activities between coxofemoral and femorotibial joints.

112

Tarsal joint. The tarsal joint appears to play an important role in vertical force
generation. The oscillating pattern of the net joint moment profile during early stance
reflected the oscillating pattern found in the vertical GRF. During 0-5% stance, there was
a brief period of joint extension which continued from the swing phase until the hoof was
flat on the ground, in agreement with a previous kinematic study (Hjerten, et al., 1994).
From approximately 5-25% stance, the net joint moment was on the extensor aspect
corresponding with EMG activity in the gastrocnemius (Wentink, 1978), deep digital
flexor (DDF) muscle (Jansen, et al., 1992) and increasing load in the SDF and DDF
tendons (Riemersma, et al., 198 8). The negative work came from a combination of
energy absorption during eccentric muscular contraction and elastic energy storage in the
tendons. Stored elastic energy was released between 20-40% stance, but the amount of
energy released was less than the preceding burst of energy absorption, indicating the
important role of gastrocnemius in absorbing impact shock. The tarsal flexor tendons
continued to store elastic energy, with peak energy storage (Min2) coinciding with the
vertical force peak. The tarsus was the only joint in the hind limb that had a negative net
joint power peak coinciding with the vertical force peak, which may reflect its role in
absorbing peak vertical force from body weight. Peak Min2 in the net joint power profile
was highly variable between horses, which may reflect different mechanisms for
managing the load. The elastic energy released from 60—90% (Max2) exceeded the
amount of energy absorbed (Min2). Therefore, the tarsal extensors performed positive
work in addition to the elastic energy release. Gastrocnemius had incidental periodical
activity during this time (Wentink, 1978). The DDF muscle, which was active (Jansen, et

al., 1992) and its tendon reached maximal load in the second half of stance (Riemersma,

113

et al., 1988), might have more responsible than gastrocnemius to move the tarsal joint.
These two muscles performed positive work to extend the tarsus in terminal stance. Since
the femorotibial joint angle did not change during this period, the peroneus tertius tendon
was stretched. The strain peaked just prior to lift off, but strain magnitude did not exceed
3% at the walk or the trot (van Weeren, et al., 1992b). The cranial tibial muscle
(Wentink, 1978) became active in late stance through early swing.

In early swing, the tarsal joint flexed due to positive work performed by
contraction of cranial tibial muscle and elastic energy recoil of the peroneus tertius.
Activity of the cranial tibial muscle ceased before the middle of swing, therefore, the
continuing flexion of the tarsus came from the action of the reciprocal apparatus. The
gastrocnemius muscle became active from the middle of swing through early stance
(Wentink, 197 8), performing positive work on the extensor aspect to extend the joint
from around 60% to the end of swing. The positive work performed by gastrocnemius
during this time might contribute to the passive extension of the femorotibial joint
thorough the function of reciprocal apparatus, resulting in the femorotibial joint extension

against flexor moment and energy absorption during the second half of swing.

Metatarsophalangeal joint. The structures distal to the tarsal joint are mostly
tendons and ligaments that can support the distal joints mechanically. During stance the
metatarsophalangeal joint had a burst of negative work followed by a burst of positive
work, that were typical of elastic energy storage and release. The joint angle, net joint
moment and net joint power profiles of the metatarsophalangeal joint were similar to

those of the metacarpophalangeal joint (Clayton, et al., 1998; Clayton, eta1., 20000), but

114

peak power was lower in the hind limbs due to the lower vertical GRF. During the first
half of stance, the joint extended to accept body weight, with energy absorption in the
SDF and DDF tendons and suspensory ligament (Riemersma, et al., 1988). Later in the
stance phase, elastic energy was released as these structures recoiled during the joint
flexion, and the net joint power became positive. The transition from negative to positive
power occurred a little later in the hind limb than the forelimb, in accordance with
differences in the joint angle profiles. The metacarpophalangeal joint extended faster,
reached maximal extension earlier (Back, eta1., 1995c) and had a shorter period of
energy absorption than the metatarsophalangeal joint.

The coupling mechanism of the reciprocal apparatus on the plantar aspect of the
limb includes the metatarsophalangeal joint via the SDF tendon (Molenaar, 1983; Back,
et al., 1995b). This causes the metatarsophalangeal joint to flex when the femorotibial
and tarsal joints flex in early swing. This flexion is controlled by an extensor moment
during 0-35% swing, probably from negative work of the long digital extensor tendon
(Robert, et al., 1999). In the second half of the swing, the metatarsophalangeal joint was
passively extended due to the inertial motion of pastem and hoof segments. The digital
flexor muscles, which were active during this time (van den Bogert, 1989), controlled the
motion of those two segments to prevent the overextension of the joint and to position the

distal segments for impact.

Distal interphalangeal joint. The joint angle, net joint moment and net joint

power profiles of the distal interphalangeal joint in the hind limb were similar to those of

the forelimb (Clayton, et al., 1998; Clayton, eta1., 2000c). During 0-45% stance, the joint

115

actively flexed due to the positive work of the deep digital flexor muscle (Jansen, et al.,
1992). After that, the joint extended under control of the same muscle, which was
performing negative work. The change in net joint power from positive to negative
occurred a little earlier in the forelimb, around 35% stance, than in the hind limb. Other
researchers have shown that the hoof sole angle relative to the ground at contact was
greater in the hind limb than the forelimb (Back, et al., 19950), indicating a more
exaggerated heel-first placement in the hind hoof. Consequently, the hind hoof moved
into a flat position and the distal interphalangeal joint reached its maximal joint flexion
later than the fore hoof. The longer duration of distal interphalangeal joint flexion in the
hind limb resulted in a longer period of energy generation.

During swing, the movement and power profiles of the distal interphalangeal joint
were similar to those of the metatarsophalangeal joint. The mechanism of j oint motion
was thought to be similar, involving control of inertial forces. Since the hoof segment has
such a small mass, the magnitude of negative work needed to restrain flexion and
extension of the distal interphalangeal joint motion was only about one tenth of that
required at the metatarsophalangeal joint. The most influential single inertial parameter in
net joint moment calculations during swing is the mass of the hoof segment (Lanovaz and
Clayton, 2001). Adding heavy shoes or pads to the hoof increases the segmental mass,
gravitational force and inertial forces, requiring more negative work to restrain the joint

motion.

Hind limb coordination. Muscular work is used to transform joint rotations into

forward translation of the entire hind limb. Muscles also transfer energy from one

116

segment to the other (Winter, 1990). Mechanical energy transfer among proximal and
distal limb segments causes an efficient conversion between rotational motions of body
segments and translation of the body center of gravity. In human sprinters, the rectus
femoris muscle transferred power from coxofemoral to femorotibial joint, while the
harnstrings transferred power from femorotibial to coxofemoral joint. Gastrocnemius
transferred power from femorotibial to tarsal joint (Jacobs, et al., 1996). In the equine
limb, several muscles and tendons cross more than one joint. The tendinous structures
connecting the proximal and the distal parts of the limbs also have the capacity to transfer
mechanical, muscular energy to other parts of the limbs (Schamhardt, 1998). Muscles and
their tendons can transfer energy between segments if the two segments are rotating in
the same direction (Winter, 1990). During most of stance, the femoral, tibial and
metatarsal segments rotated in the same direction. During terminal stance, only the
femoral and metatarsal segments rotated in the same direction (Hodson, et al., 2001).
Energy transfers among these three segments could enhance the efficiency of trotting
gait.

Trotting horses use elastic energy conservation to minimize energy expenditure
(Cavagna, et al., 1977; Minetti, et al., 1999). Net energy during stance in the trot was
almost balanced (-0.018 J/kg), while at the walk energy was generated 0.098 J/kg during
stance (Clayton, et al., 2001). Elastic energy storage and release in the hind limb
contributed two-thirds of overall energy storage (Biewener, 1998). In this study the
metatarsophalangeal, tarsal and femorotibial joints contributed to elastic energy storage
and release. The elastic energy storage may come fi'om a mechanical energy transfer in a

proximal to distal direction when the body center of mass is decelerated downward and

117

backward until the middle of stance (Buchner, et al., 2000). Energy transfers might occur
via the musculotendinous and ligamentous structures that connected the femoral, tibial
and metatarsal segments. During the second half of stance, elastic energy release may
result in energy transfer in a distal to proximal direction to accelerate the body center of
mass upward and forward during the second half of stance. The gait during stance was
balanced in net energy generation and absorption due to elastic spring around the
metatarsophalangeal, tarsal and femorotibial joints, while energy generation at the
coxofemoral joint was balanced by absorption at the distal interphalangeal joint, which
acted as an energy damper.

Energy transfers might assist the proximal joints in moving the distal segments.
During most of the swing, the femoral, tibial, metatarsal, pastem and hoof segments
rotated accordingly (Hodson, et al., 2001). Most of the muscular activities during swing
are around the proximal limb joints. During the first half of swing, the tensor fascia latae
flexed the coxofemoral joint, while semitendinosus flexed the femorotibial joint. The
tarsal joint flexed under the activity of cranial tibial muscle. During the second half of
swing, the coxofemoral joint was extended by hamstring muscles, while the tarsal joint
was extended by gastrocnemius. There was little activity in the digital flexors and
extensors during swing. Therefore, the distal joints moved under the influence of the
activity around proximal joints, probably via energy transfer through digital extensor and
flexor tendons. In the bouncing gait, such as the trot, the elastic energy release from the
end of stance may also contribute to the flexion of the limb during swing. Joint flexions
bring the moment of inertia of the hind limb closer to the pivot point in the acetabulum,

thus reducing the needed to accelerate and decelerate the entire limb during swing.

ll8

Therefore, the internal mechanical work required to move the body segments relative to
the body center of mass at the trot is less than that at the walk (Minetti, et al., 1999). The
combination of elastic energy storage and release and reduction in moment of inertia
might explain the lower net mechanical energy of the limb during the whole stride at the
trot (0.1014 J/kg) than that at the walk (0.161 J/kg) (Clayton, et al., 2001).

A proximal to distal temporal sequence of peak positive powers was consistently
found in human sprinters, and was thought to be the appropriate sequence to generate
high sprinting speed (Jacobs and van Ingen Schenau, 1992; Johnson and Buckley, 2001).
In this study, the peak positive powers during the second half of stance were found
sequentially in the coxofemoral joint (Max2) followed by the tarsal joint (Max2) and the
metatarsophalangeal joint (Max), indicating a proximal to distal sequence of energy

generation.

CONCLUSION

The coxofemoral and tarsal joints were the source of energy generation to move
the limb throughout the stride. The coxofemoral joint protracts and retracts the entire
limb, whereas the tarsal joint is more concerned with proximal and distal movement. The
femorotibial joint was an important site of energy absorption during swing. The sources
of elastic energy storage and release in the hind limb were the metatarsophalangeal, tarsal
and femorotibial joints. The distal joints were driven by inertial forces during swing,

following the active motion of the proximal joints.

119

CHAPTER 4

EFFECT OF UNILATERAL SYNOVITIS OF DISTAL INTERTARSAL AND
TARSOMETATARSAL JOINTS ON SAGITTAL PLANE KINEMATICS AND
KINETICS OF TROTTING HORSES: RESULTS AND DISCUSSION

LAMENESS AND PHYSICAL EXAMINATION

Before lameness induction, every horse had normal rectal temperature, pulse rate,
respiratory rate, mucous membrane color, and capillary refill time, and the tarsal joint had
a normal appearance. Rectal temperature increased by 0.6-2.0°F within 6-12 hours after
induction, after which the rectal temperature gradually decreased to baseline values
within 18 hours. None of the horses became inappetant during the observation period. All
horses became lame within 4 hours after injection of endotoxin, but there were minimal
changes in joint appearance and no swelling of the periarticular tissues.

Lameness evaluation performed 12 hours after endotoxin injection indicated
lameness more severe than grade 3 in 3 horses (Horses no. 2, 3, 4), so 0.5-1 g of
phenylbutazone was given orally to those horses. At the lameness evaluation 24 hours
after endotoxin injection, three horses (Horses no.1, 2, 3) were lame at grade 2 or less and
were ready for gait analysis. Horse no.4 was assessed as grade 2 lame at 30 hours after
endotoxin injection. Within one week after synovitis induction, none of the horses was
lame at the walk and lameness was minimal at the trot (less than grade 1). At this time the

horses were returned to pasture.

120

EFFECT OF SYNOVITIS ON INTRA-LIMB COORDINATION

Test for differences in velocities. There was no significant difference in velocity

between sound and lame conditions (P>0.05) (Table 4-1), therefore the differences in

magnitude of the variables were not affected by subject velocities and could be

interpreted as effects of synovitis of the distal intertarsal and tarsometatarsal joints.

Table 4-1 Statistical analysis comparing velocity (m/s) and velocity in dimensionless
units of each limb between sound and lame conditions. Values are mean and (SD).

 

 

 

 

 

 

   

   
 

 

  
  

 

 

 

    
  

Horse no. Veloci (m/s) Velocity in dimensionless units
RF sound RF lame RF sound RF lame P value
1 2.91 (0.09) 2.87 (0.09) 0.77 (0.02) 0.76 (0.02) 0.45
2 2.91 (0.06) 2.88 (0.04) 0.76 (0.02) 0.75 (0.01) 0.38
3 2.83 (0.07) 2.87 (0.08) 0.74 (0.02) 0.75 (0.02) 0.31
2.82 (0.06) 0.75 0.02) 0.74 (0.01 0.36

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Horse no. Velocity (m/s) Velocity in dimensionless units
LF sound LF lame LF sound LF lame P value
1 2.77 (0.13) 2.89 (0.18) 0.74 (0.04) 0.77 (0.05) 0.27
2 2.85 (0.05) 2.87 (0.03) 0.74 (0.01) 0.75 (0.01) 0.47
3 2.88 (0.07) 2.91 (0.07) 0.75 (0.02) 0.76 (0.02) 0.47
4 2.83 (0.06) 2.81 (0.07) 0.75 (0.02) 0.75 (0.02) 0.72
Horse no. Velocity (m/s) Velocity in dimensionless units
RH sound RH lame RH sound RH lame P value
1 2.85 (0.09) 2.85 (0.10) 0.74 (0.02) 0.76 (0.03) 0.92
2 2.88 (0.06) 2.90 (0.07) 07440.02) 0.75 (0.02) 0.31
3 2.82 (0.11) 2.83 (0.04) 0.74 (0.03) 0.74 (0.01) 0.89
2.78 (0.07 0.16

 

 

 

 

 

 

 

 

 

 

 

 

Horse no. Velocity (m/s) Velocity in dimensionless units
LH sound LH lame LH sound LH lame P value
1 2.74 (0.06) 2.81 (0.05) 0.73 (0.02) 0.75 (0.01) 0.10
2 2.87 (0.07) 2.88 (0.05) 0.75 (0.02) 0.75 (0.01) 0.66
3 2.85 (0.08) 2.86 (0.07) 0.75 (0.02) 0.75 (0.02) 0.83
4 2.79 (0.04) 2.79 (0.06) 0.75 (0.01) 0.74 (0.02) 0.92

 

121

 

Kinematic variables. There were no significant differences (P>0.05) in stride
duration, stance duration, swing duration or stride length between sound and lame
conditions on the right hind limb. The range of vertical displacement of the talus during
the stance phase was significantly smaller in the lame condition, with the mean difference
(SD) between sound and lame conditions of 0.32 (0.18) cm. There were also significant
decreases in tarsal and distal interphalangeal joint ranges of motion and a trend toward a
decrease in coxofemoral joint peak extension (Max1) in the middle of the stance phase
(Table 4-2). At the tarsal joint during stance, angular displacement between impact and
peak Min1 (tarsal joint range of flexion) and angular displacement between peak Min2
and the angle obtained at the end of stance (tarsal joint range of extension), had trends
toward a decrease in range of motion (P=0.06 and P=0.09, respectively), with mean
differences (SD) of 2.68 (1.81) and 1.42 (1.12) degrees, respectively. During the swing
phase, there was a significant increase in vertical displacement of the tuber coxae and a
trend toward a decrease in distal interphalangeal joint vertical displacement (P=0.09),

with mean differences (SD) of 0.61 (0.37) and 1.31 (1.10) cm, respectively.

Ground reaction forces. Figure 4-1 illustrates the vertical and longitudinal
ground reaction force (GRF) profiles from the right hind limb in sound and lame
conditions. Peak vertical force of the right hind limb was significantly lower in the lame
condition compared with the sound condition. There were trends toward decreases in
vertical impulse (P=0.06) and peak pr0pu1sive force (P=0.09) in the right hind limb

(Table 4-2).

122

Net joint moments, net joint powers and net joint energies. Several variables
in this category showed significant differences and trends toward differences between
sound and lame conditions. During stance (Table 4-2), there was a decrease in positive
work (energy Max2) at the femorotibial joint. At the tarsal joint, the main findings were a
lower negative power peak (power Min1) with a trend toward decreased energy
absorption (energy Min1) during initial stance, and a significant reduction in peak power
and energy generation during late stance (power Max2 and energy Max2). The distal
interphalangeal joint showed trends toward decreases in energy generation (energy Max)
during initial stance and energy absorption (power Min and energy Min) during the
second half of stance.

During the swing phase (Table 4-3), the femorotibial joint absorbed less energy as
the joint was passively extended during the second half of swing (energy Min2). The
tarsal joint showed a decrease in peak flexor moment (Min1) and a trend toward decrease
in energy generation (energy Max1) in early swing. In late swing, the tarsal extensors
produced a smaller power peak (power Max2) without a decrease in energy generation.
Both the metatarsophalangeal and distal interphalangeal joints showed significant
reductions in energy absorption (energy Min1, both joints) as the joints flexed during the
initial swing phase.

The kinematic and kinetic variables reported in Tables 4.2 and 4.3 are those that
showed significant differences (P<0.05) and trends toward differences (0.05<P<0.1)
between sound and lame conditions as shown in Figures 4-1 to 4-11. Tables with

complete data for all variables are shown in Appendix B.

123

Table 4-2 Differences between sound and lame conditions in variables that differed

significantly or showed a trend toward a significant difference during the stance phase of

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

the right hind limb.
Joint Variables Mean SD
difference
Ground reaction force Vertical force Max (N/kg) 0312 a 0182
Vertical impulse (N s/kg) 41052 b 0,034
Longitudinal force Max3 (N/kg) 00896 0,073
Coxofemoral Joint angle Max1 (degrees) -123 b 0.97
Femorotibial Net joint energy Max2 (J/kg) -0.008 a 0,004
Tarsal Joint range of motion (degrees) 43.20 3 128
Net joint power Max2 (W/kg) -0374 a 0.060
Net joint power Min1 (W/kg) -0.597 a 0.306
Net joint energy Max2 (J/kg) 0023 a 0,005
Net joint energy Min1 (J/kg) -0022 b 0.015
Total positive energy (J/kg) -0023 a 0.015
Total negative energy (J/kg) -003 3 a 0.020
Metatarsophalangeal Net joint moment Min (Nm/kg) ,0034 a 0,017
Distal interphalangeal Joint range of motion (degrees) _5.21 3 200
Net joint moment Min (Nm/kg) -0.058 b 0,044
Net joint power Min (W/kg) .0770 b 0.620
Net joint energy Max (J/kg) 0012 b 0.009
Net joint energy Min (J/kg) -0.056 b 0.043

 

 

Negative mean differences indicate values were higher in the sound condition than in the

lame condition. a P<0.05. b 0.05<P<0.1.

124

 

Table 4-3 Differences between sound and lame conditions in variables that differed

significantly or showed a trend toward a significant difference during the swing phase of

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

the right hind limb.
Joint Variables Mean SD
difference

Femorotibial Net joint moment Max (Nm/kg) 00142 b 0,0094
Net joint power Min2 (W/kg) -0.1138B 0,0897
Net joint energy Min2 (J/kg) -0.0069 3 00035
Total negative energy (J/kg) -001216 0.0091

Tarsal Net joint moment Min (Nm/kg) -0.0065 3 0,0036
Net joint power Max2 (W/kg) -0.0859 3 00431
Net joint energy Maxl (J/kg) _00042 b 00029
Total positive energy (J/kg) -0.0089 b 0,0063

Metatarsophalangeal Net joint power Max1 (W/kg) -0.0058 3 0,0033
Net joint power Max2 (W/kg) -00047 b 0.0040
Net joint power Min1 (W/kg) -00247 b 0.0161
Net joint energy Max1 (J/kg) -00002 b 0,0002
Net joint energy Max2 (J/kg) -00001 a 0.00003
Net joint energy Min1 (J/kg) -00013 a 0.0005
Total negative energy (J/kg) 0002 5 a 00014

Distal interphalangeal Net joint moment Max (Nm/kg) _00009 3 0,0006
Net joint power Max1 (W/kg) -00021 a 0.0010
Net joint power Max2 (W/kg) _00015 3 0,0008
Net joint power Min1 (W/kg) -00109 a 0.0057
Net joint energy Maxl (J/kg) 410001 3 000004
Net joint energy Min1 (J/kg) _00004 a 0.0003
Net joint energy Min2 (J/kg) _00001F 00001
Total positive energy (J/kg) 00002 b 00001
Total negative energy (J/kg) -00005 a 0.0003

 

 

 

 

 

 

 

 

 

 

 

 

 

Negative mean differences indicate values were higher in the sound condition than in the

lame condition. a P<0.05. b 0.05<P<0. 1.

125

 

Figure 4-1 Vertical and longitudinal ground reaction force data from the right hind limb.
The thick black line and dashed lines indicate the mean value and one standard deviation
above and below the mean of the sound condition. The thick gray line indicates mean of
the lame condition. Dot represents variable that difi‘ers significantly between sound and

lame conditions. 3 P<0.05. b 0.05<P<0. 1.

 

a
10.0 1 Max
8.0 -
o
2
e a 5.0 ‘ .
To
.2 g 4.0.
t V
g
2.0 .f;
0.0

 

 

Percent stance duration

 

 

Longitudinal force
(Niko)

 

0 20 40 60 80 1 00

Percent stance duration

126

Figure 4-2 Kinematic and kinetic variables of the right coxofemoral joint during stance.
The thick black line and dashed lines indicate the mean value and one standard deviation
above and below the mean of the sound condition. The thick gray line indicates mean of
the lame condition. Dot represents variable that differs significantly between sound and

lame conditions. 3 P<0.05. b 0.05<P<0.1.

1401

(degrees)

 

 

Joint angle

 

 

0 20 40 60 80 1 00

 

 

Net joint moment
(lekg)

 

 

 

-1.0 -
-1.5 r
0 20 40 60 80 100
4.0 -
3 l
g 2.0
O A
o. g ,x
C
E E 0.0 w
‘6
2 -2.0 -
-4.0 -

0 20 40 60 80 1 00

Percent stance duration

127

Figure 4-3 Kinematic and kinetic variables of the right coxofemoral joint during swing.
The thick black line and dashed lines indicate the mean value and one standard deviation
above and below the mean of the sound condition. The thick gray line indicates mean of
the lame condition.

150 -

Joint angle
(degrees)

 

 

 

 

 

 

Net joint moment
(lekg)

 

-1.0 J
o 20 40 so so 100

2.5 -

 

 

Net joint power
(Wlkg)

 

 

-1.5 -
0 20 40 60 80 100

Percent swing duration

128

Figure 4-4 Kinematic and kinetic variables of the right femorotibial joint during stance.
The thick black line and dashed lines indicate the mean value and one standard deviation
above and below the mean of the sound condition. The thick gray line indicates mean of
the lame condition. Dot represents variable that differs significantly between sound and

lame conditions. a P<0.05. b 0.05<P<O.l.

130 -

 

 

913
9::
23, 110-
.so
03

_,

 

 

 

 

 

Net joint moment
(Nm/kg)

 

 

Net joint power
(Wlkg)

 

-4.0 .
0 20 40 60 80 100

Percent stance duration

129

Figure 4-5 Kinematic and kinetic variables of the right femorotibial joint during swing.
The thick black line and dashed lines indicate the mean value and one standard deviation
above and below the mean of the sound condition. The thick gray line indicates mean of
the lame condition. Dot represents variable that differs significantly between sound and

lame conditions. a P<0.05. b 0.0S<P<0. 1.
140

120

Joint angle
(degrees)

 

 

 

 

 

Net joint moment
(Mm/kg)

 

 

 

 

Net joint power
(Wlkg)

 

-1 .5 -
0 2O 4O 60 80 100

Percent swing duration

130

Figure 4-6 Kinematic and kinetic variables of the right tarsal joint during stance. The
thick black line and dashed lines indicate the mean value and one standard deviation
above and below the mean of the sound condition. The thick gray line indicates mean of
the lame condition. Dot represents variable that differs significantly between sound and

lame conditions. a P<0.05. b 0.05<P<O. 1.

130 !

 

 

 

2 A
m m no.
a n.
E 9
g 3
90 I U t I 1
0 20 40 60 80 100
2.0-

 

 

Net joint moment
(lekg)

 

0 20 40 60 80 1 00
4.0 -

 

Net joint power
(Wlkgl

 

 

60 80 100

Percent stance duration

131

Figure 4-7 Kinematic and kinetic variables of the right tarsal joint during swing. The
thick black line and dashed lines indicate the mean value and one standard deviation
above and below the mean of the sound condition. The thick gray line indicates mean of
the lame condition. Dot represents variable that differs significantly between sound and

lame conditions. 3 P<0.05. b 0.05<P<O.1.

 

 

 

 

 

 

 

 

 

 

140 j
120 -
i; 1;; 100-
a i
E 9 8° ‘
'5 3
'3 60 q
40 f I I I I
O 20 40 60 80 100
0.2 -
‘E 0.1 -
0
g A
O
E E 0.0
‘E
"5 .2. t .
:7 -0.1 d '
0 .
2
-0.2 -
O 20 40 60 80 100
0.5 - a
_. - Max2
a ......... .‘ll...’ 3’ “u
g ,
O A ’1' -------- .
Q m 0 O "
:2: a . ,t' I
.2 2 Max]
‘6
2
-0.5 .
0 20 40 60 80 100

Percent swing duration

132

Figure 4-8 Kinematic and kinetic variables of the right metatarsophalangeal joint during
stance. The thick black line and dashed lines indicate the mean value and one standard
deviation above and below the mean of the sound condition. The thick gray line indicates
mean of the lame condition. Dot represents variable that differ significantly between

sound and lame conditions. 3 P<0.05. b 0.0S<P<O. 1.

 

 

 

260~
240- ---------------------------

+2, ’5; 220-

C

: s 200w

0 E

" 180-
160 I I r I I

o 20 40 so so 100

 

Net joint moment
(lekg)

 

 

O 20 4O 60 80 1 00

 

Net joint power
(Wlkgl

 

 

O 20 40 60 80 1 00

Percent stance duration

133

Figure 4-9 Kinematic and kinetic variables of the right metatarsophalangeal joint during
swing. The thick black line and dashed lines indicate the mean value and one standard
deviation above and below the mean of the sound condition. The thick gray line indicates
mean of the lame condition. Dot represents variable that differs significantly between

sound and lame conditions. a P<0.05. 0.05<P<0.1.

Joint angle
(degrees)

 

 

 

 

 

Net joint moment
(lekg)

-0.02 -

 

-0.03 -

0.10 - b

 

 

Net joint power
(W/kg)

i'.:-.',"'a
-0.20 - M1nl

 

-0.30 -
0 20 40 60 80 1 00

Percent swing duration

134

Figure 4-10 Kinematic and kinetic variables of the right distal interphalangeal joint
during stance. The thick black line and dashed lines indicate the mean value and one
standard deviation above and below the mean of the sound condition. The thick gray line
indicates mean of the lame condition. Dot represents variable that differs significantly

between sound and lame conditions. a P<0.05. b 0.05<P<O.1.

220 n

.......

 

 

 

2 A
a) m
s t
SE 8’
O E

3 a

range
120 I I I T I
O 20 40 60 80 100

 

Net joint moment
(Nm/kg)

 

 

Net joint power
(W/kg)

 

 

O 20 40 60 80 1 00

Percent stance duration

135

Figure 4-11 Kinematic and kinetic variables of the right distal interphalangeal joint
during swing. The thick black line and dashed lines indicate the mean value and one
standard deviation above and below the mean of the sound condition. The thick gray line
indicates mean of the lame condition. Dot represents variable that differs significantly

between sound and lame conditions. a P<0.05. 0.05<P<0.1.

Net joint moment Joint angle

Net joint power

(degrees)

(Nm/kg)

(W/kg)

230.0 -
210.0
190.0 -

170.0 ‘

 

1 50.0 <

 

1 30.0 I I I I I
0 20 40 60 80 100

0.004 -

 

0.002 '-

0.000 "I

 

 

-0.002 -

 

-0.004 J
0 20 40 60 80 1 00

0.020 -

 

 

 

o 20 4o 60 80 100
Percent swing duration

136

FORELIMB VARIABLES IN THE SOUND CONDITION

Before comparing forelimb variables between the sound and lame conditions, a
brief explanation of the biomechanical profiles of the forelimbs in the sound condition
will be given. Kinematic, ground reaction force, net joint moment and net joint power
patterns from both forelimbs in this study were similar to previous studies (Clayton, et
al., 1998a; Lanovaz, etal., 1999; Clayton, et al., 2000c). The peak values of some
variables were slightly different, which may be due to breed differences. During the
stance phase, the net joint moment was mainly on the cranial side of the scapulohumeral
joint, which was the extensor side. The other joints had the net joint moment on the
caudal/palmar side, which was the extensor side of the cubital joint and the flexor side of
the carpal, metacarpophalangeal and distal interphalangeal joints (Figure 4-12). During
limb protraction in the first half of swing, the net joint moment was on the cranial/dorsal
side of all joints, which was an extensor moment for all except the cubital joint. When the
limb was retracted in the second half of swing, the net joint moment moved to the
caudal/palmar side of all joints (Figure 4-13).

Net joint power profiles during the stance phase (Figure 4-14) showed evidence of
elastic energy storage and release patterns at the scapulohumeral, cubital, carpal and
metacarpophalangeal joints. The distal interphalangeal joint performed mostly negative
work. During the swing phase (Figure 4-15), the cubital joint was the only joint that
performed positive work, with bursts of energy generation in early swing and late swing.
At the other joints, the amount of energy absorption decreased in a proximal to distal

sequence.

137

Figure 4-12 Net joint moment profiles at the scapulohumeral, cubital, carpal,
metacarpophalangeal and distal interphalangeal joints averaged from both forelimbs of
sound trotting horses during the stance phase. The heavy line is the mean curve and the
two lighter lines indicate one standard deviation.

 

 

 

2i)-
Max
2 in 1.0 a /_/—//\
e E o o /\/\ Scapulohumeral
2 v o /V\V/ I T I I I
Min
4.0"
O 20 40 60 80 100
10 ‘ Max

Moment

(N m/kg)
O O
'o in

' r ' \\/\’j' Cubital

J Min
0 20 4O 60 80 100

 

 

Carpal

Moment
(N m/kg)

 

 

0 20 40 60 80 100

 

Metacarpophalangeal

Moment
(N m/kg)

 

 

0 20 40 60 80 100

 

Distal
interphalangeal

 

Moment
(N m/kg)

 

o 2 o 4 o 6 o 8 o 1 o 0
Percent stance duration

138

Figure 4-13 Net joint moment profiles at the scapulohumeral, cubital, carpal,
metacarpophalangeal and distal interphalangeal joints averaged fiom both forelimbs of
sound trotting horses during the swing phase. The heavy line is the mean curve and the
two lighter lines indicate one standard deviation.

.600

.300

.000

Moment
(N m/kg)

.300

.600

.400
.200

.000

Moment
(N Mg)

.200

.400

.150

.050

.050

Moment
(Nm/kg)

.150

.040
.020

.000

Moment
(N m/kg)

.020

.040

.004
.002

.000

Moment
(N m/kg)

.002

.004

l

-l

\

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

. N: scapulohumeral
J
0 20 40 60 80 100

. f . Cubital
o 2 o 4 o 6 o 8 o 1 o o
1

r . . Carpal
o 2 o 4 o 6 o a o 1 o o

. f . Metacarpophalangeal
o 2 o 4 o 6 o 8 o 1 o o
1
1

. . Distal
v interphalangeal

0 20 40 60 80 100

Percent swing duration

139

Figure 4-14 Net joint power profiles at the scapulohumeral, cubital, carpal,
metacarpophalangeal and distal interphalangeal joints averaged from both forelimbs of
sound trotting horses during the stance phase. The heavy line is the mean curve and the
two lighter lines indicate one standard deviation.

 

 

 

 

 

 

 

 

 

 

 

”0‘0
‘1
E Scapulohumeral
3
B
G
a.
o 2 o 4 o 6 o 8 o 1 o 0
’6‘0
i‘
E Cubital
3
B
e
an
1 o o
’53
E
‘5 Carpal
3
O
a.
1 0 0
”6‘0
1%
E . Metacarpophalangeal
Ln
‘3’
e
a.
1 o 0
1&0 Distal
E interphalangeal
a":
3
e
G-t
o 2 o 4 o 6 o 8 0 1 o 0

Percent stance duration

140

Figure 4-15 Net joint power profiles at the scapulohumeral, cubital, carpal,
metacarpophalangeal and distal interphalangeal joints averaged from both forelimbs of
sound trotting horses during the swing phase. The heavy line is the mean curve and the
two lighter lines indicate one standard deviation.

1.50

.50

-050

Power (W/kg)
O

-1.50
1.50
A
on
.5 100
Ea 0.50
h
o
3 0.00
c
A:

-050

0&0

Power (W/kg)

0J0

J0

Power (W/kg)
6

Power (W/kg)

20

. // Scapulohumeral

40 60 80 100

 

 

 

 

 

 

 

 

 

20

20

20

 

40 60 80 100
' ' Carpal
4 o 6 o 8 o 1 o o
\/r/—\' Metacarpophalangeal
40 60 80 100

 

*1 . Distal
interphalangeal

4 o 6 0 a o 1 o 0
Percent swing duration

141

INTER-LIMB COMPENSATION FOR LAMENESS OF THE RIGHT HIND
LIMB

Kinematic variables. Data were collected at the similar velocity in the sound and
lame conditions. There were no significant differences (P>0.05) in stride, stance and
swing duration, or stride length in the compensating 1efi hind limb, left forelimb or right
forelimb. To assess the asymmetrical motion of the tuber coxae in the lame condition, the
quotient of tuber coxae range of vertical displacement between the right and lefi hind
limbs during stance of each horse was calculated (Buchner, et al., 1996c). In the sound
condition, all horses had a slightly bigger range of tuber coxae displacement on the right
side than on the left side as shown by the quotient values greater than one (Table 4.4). In
the lame condition, three of the four horses had smaller quotient than in the sound
condition, indicating a relative increase in displacement of the left tuber coxae. Further
evaluation showed a significant increase in vertical displacement of the left tuber coxae
during the stance phase with a mean difference (SD) of 0.23 (0.13) cm. There was also a
trend toward an increase in distal hoof wall vertical displacement (P=0.09) during the

swing phase of the left hind limb, with a mean difference of 0.5 (0.4) cm.

Ground reaction forces. There were changes in GRF in the left (contralateral)
fore and hind limbs. In the left hind limb, there was a trend toward an increase in peak
vertical force (P=0.08) and a significant decrease in braking impulse (P<0.05), with mean
differences (SD) ofO. 170 (0.133) N/kg and 0.007 (0.004) Ns/kg, respectively. The lefi
forelimb showed a significant decrease in vertical impulse (P<0.05) in the lame

condition, with a mean difference of 0.050 (0.023) Ns/kg.

142

To assess the compensating pattern of load distribution, the center of vertical load
distribution between the four limbs in the lame condition was calculated in each horse
and plotted with the values from the sound condition (Figure 4-16). In three horses, the
load distribution was shifted to the left in the lame condition, while Horse no.2 had
different pattern by shifting the load to the right side of the body.

The symmetry index expressed as a percent quotient between the hind limb pair
(Weishaupt, et al., 2001) was applied to the vertical impulse in this study. The symmetry
index (Table 4-5) of Horses no.1 and 4 indicated a smaller vertical impulse in the right
hind limb in the lame condition, whereas the symmetry index in Horses no.2 and 3 did
not change. A similar calculation was applied within the same limb to determine a
quotient for the vertical impulse (Table 4-6). All horses, except Horse no.2, showed a
decrease in vertical impulse in the lame condition of the right hind limb, but only Horses

no.1 and 4 showed a compensatory increase in vertical impulse in the left hind limb.

Net joint moments, net joint powers and net joint energies. Significant
differences or trends toward differences were found in limbs that showed significant
differences in kinematic and GRF variables. During stance, the only significant change in
the left hind limb (Table 4-7) was an increase in the metatarsophalangeal joint peak
flexor moment (Min). During swing, there was an increase in magnitude of peak flexor
moment at the coxofemoral joint (moment Min1) and an increase in peak positive power
(power Maxl) in early swing. In terminal swing, the coxofemoral joint generated less

energy (energy Max2) and the femorotibial joint had smaller peak negative power (power

Minl).

143

In the left forelimb stance phase (Table 4-8), there were significant decreases in
peak flexor moments at the carpal and metacarpophalangeal joints. The
metacarpophalangeal joint also showed a trend toward a decrease in peak negative power
(power Min1) and elastic energy storage (energy Min1) in early stance. The only joint

that showed a significant decrease in energy generation was the carpal joint (energy

Max2).

Figure 4-16 Load distribution of vertical impulse of 4 horses. The black dot represents
the value in the sound condition. The gray dot represents the value in the lame condition.
Arrow represents the direction of load distribution changes in each horse.

 

59 %

Horse 1
Front

 

Horse 2

1w:

t
m:

 

 

55 %

 

 

 

Left 4 50% > Right

144

Table 4-4 Symmetry index as a quotient value of difference of tuber coxae displacement
between the hind limb pair from the sound and lame conditions. Values greater than one
mean the displacement range is greater in the right hind limb than the left hind limb.

 

 

 

 

 

 

 

 

Horse 1 Horse 2 Horse 3 Horse 4
Sound condition 1.00 1.1 1 1.03 1.08
Lame condition 0.98 1.13 1.00 0.97

 

 

Table 4-5 Symmetry index as a percent quotient of difference of vertical impulses
between the hind limb pair from the sound and lame conditions. Negative values mean
the impulse is greater in the left hind limb than the right hind limb.

 

 

 

 

 

 

 

 

Horse 1 Horse 2 Horse 3 Horse 4
Sound condition 1% -l% 0% -1%
Lame condition -2% -l% 0% -4%

 

 

Table 4-6 Symmetry index as a percent quotient of difference of vertical impulses
between the sound and lame conditions in the same hind limb. Negative values mean the
impulse is greater in the sound condition than the lame condition.

 

 

 

 

 

 

 

 

Horse 1 Horse 2 Horse 3 Horse 4
Right hind limb -3% 0% -2% -3%
Left hind limb 1% -l% -l % 2%

 

145

 

Table 4-7 Differences between sound and lame conditions in variables that differed

significantly or showed a trend toward a significant difference during stance and swing
phases of the left hind limb.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Mean SD
(Stance or Swing) difference
Coxofemoral (Swing) Net joint moment Min1 (Nm/kg) 0047 a 0,021
Net joint moment Min2 (Nm/kg) 0023 b 0,018
Net joint power Maxl (Nm/kg) 0139 a 0.073
Net joint energy Max2 (J/kg) -0020 b 0.017
F emorotibial (Swing) Net joint power Minl (W/kg) _0115 a 0,029
gfgfitczgfophalangeal Net jOlnt power Mln (W/kg) 0.291 a 0.128
Metatarsophalangeal Net joint power Maxl (W/kg) _0003 5 b 0,0029
(Swing) Total positive energy (J/kg) -00003 a 0.0001
313:: gi;1terphalangeal Net joint energy Max2 (J/kg) 410001 b 0.00004

 

Negative mean differences indicate values were higher in the sound condition than in the
lame condition. a P<0.05. b 0.05<P<0. 1.

Table 4-8 Differences between sound and lame conditions in variables that differed
significantly or showed a trend toward a significant difference during stance phases of the

 

 

 

 

 

 

left fore limb.
Joint Variables Mean SD
difference
Carpus Net joint moment Min Him/kg) 0104 a 0049
Net joint energy Max2 (J/kg) -0013 a 0.005
Metacarpophalangeal Net joint moment Min (Nm/kg) -0051 a 0.028
Net joint power Min (W/kg) -0.468 b 0.327
Net joint energy Min (J/kg) -0019 F 0.015

 

 

 

 

 

Negative mean differences indicate values were higher in the sound condition than in the
lame condition. a P<0.05. b 0.05<P<O. 1.

146

 

 

DISCUSSION

Lameness model. The ability to induce a transient lameness in sound horses
offers an important and reliable method to study the locomotor pattern in horses with a
specific, well-defined lameness in a controlled manner that minimizes individual
variations (Buchner, 2001b). The lameness model used in this study has been used
previously to study the effect of pain in the distal tarsal joints on two-dimensional hind
limb kinematics (Kramer, et al., 2000). In that study, the induction procedures were
slightly different from those reported here; phosphate buffered saline was used to dilute
the endotoxin and there was an injection of contrast solution into the distal tarsal joints
before endotoxin injection. Soft tissue swelling extended distally to the proximal or mid
metatarsus in all horses in that study, which may be for two reasons. Firstly, the use of
phosphate buffered saline could cause irritation to the joint. Secondly, the distal
intertarsal joint is narrow, and if the volume of fluid injected into the joint exceeded the
total volume of the joint space, there might be some leakage into periarticular tissue
during withdrawal of the needle. The non-steroidal anti-inflammatory medication
administered in this study helped to reduce inflammation and decreased the time between

induction and gait analysis.

Endotoxin induced synovial membrane inflammation. The lameness model
used in this study induced pain in the distal tarsal joints, which are the most common site
of tarsal osteoarthritis. The painful sensation may be slightly different from that of
naturally occurring osteoarthritis. Degeneration of high-load low motion joints is likely

due to trauma to the periarticular soft tissues (Pool, 1996), leading to degeneration of

147

cartilage and subchondral bone sclerosis. Furthermore, in the natural disease pain may be
associated with different events in the stride cycle, and horses may learn to accommodate
over a period of time. The effects of the tarsal synovitis model represent a general
response to pain in the distal tarsal joints, but the effects may not be identical to the

natural disease.

Effects of synovitis on symmetry of motion and vertical impulse of hind
limbs. Clinical evaluation in the equine hind limb is based on symmetry of left and right
gluteal rises. But it is recognized that, in mild lameness, the gluteal rise may be
symmetrical, which makes it difficult to identify the lame limb (Stashak, 1987). Several
kinematic studies have assessed asymmetrical motion in hind limb lameness (Buchner, et
al., 1993; Buchner, et al., 1996c; Pourcelot, et al., l997a). Comparison between vertical
displacement of the left and right tuber coxae or tuber sacrale is simple and corresponds
to the usual method of subjective assessment by clinicians. In this study, comparison of
the height of the tuber coxae and vertical impulse of the left and right hind limbs in the
lame condition did not yield significant differences (P>0.05) in the whole group of
horses. The time that elapsed between subjectively grading the lameness and data
collection allowed further improvement of clinical signs. Therefore, lameness was less
than grade 2 at the time of data collection. When the horses were analyzed individually,
it was found on the basis of tuber coxae displacement (Table 4-4) and analysis of vertical
impulses (Table 4-6) that Horse no.2 was only slightly lame when the data were collected
for the lame condition. The tuber coxae displacement quotient in Horses no. 1, 3 and 4

decreased in the lame condition, while Horse no.2 had a slight increase. In a previous

148

study (Buchner, et al., 1996c) found that the mean differences in the tuber coxae
displacement quotient between sound and grade 1 lame, and between sound and grade 2
lame were 0.25 and 0.34, respectively. This was much larger than the difference seen in
three horses (0.02-0.1 1) in this study. The previous study used the lameness grading
system of Stashak (1987) in which the definition of grade 1 lame implies a more marked
lameness than grade 1 lame in the AAEP grading system (American Association of
Equine Practitioners, 1991) used in this study. Certainly, the lameness observed in this
study was very mild in all horses.

Criteria for grading the degree of lameness by vertical impulse have been
proposed based on the mean difference between the two hind limbs: 0-2% mean
difference for grade 1, 2-4% mean difference for grade 2, and more than 4% mean
difference for grade 3 (Weishaupt, et al., 2001). The clinical grading system in that study
was similar to the one used in this study, so the criteria might be applied as an objective
evaluation in the present study. The mean difference of vertical impulse between the left
and right hind limbs in the lame condition (Table 4-5) indicated that Horses no.1 and 4
would be assessed as grade 2 lame, while Horses no.2 and 3 would be assessed as grade 1
lame.

The analysis of the center of vertical loading distribution indicated that, even in
the sound condition, the vertical loading of the left and right limbs was not symmetrical,
which is probably a reflection of sidedness. In sound horses, asymmetry in tuber coxae
displacement (Buchner, et al., 1996c; Pourcelot, et al., 1997a; Pourcelot, et al., l997b), in
peak vertical force (Williams, et al., 1999) and in vertical impulse (Merkens, et al., 1993)

have been reported at the trot. Sidedness is known to exist in horses (Deuel and

149

Lawrence, 1987; Drevemo, et al., 1987), and trainers are well aware of the fact that
horses naturally have a slightly asymmetrical movement pattern, preferring to carry more
weight on one hind limb, usually the left hind limb (Ljungquist, 1976). The effect of
slight sidedness on gait variables and the evaluation of asymmetry in mild lameness have
not been investigated. In this study, Horse no.3 had its center of vertical loading on the
right side in the sound condition. Although the center of vertical loading moved toward
the left when the horse was lame in the right hind limb, the effect was to produce a more
symmetrical loading pattem. If the lameness had been more severe, it is anticipated that
the center of loading would have shifted further to the left. The shift in load distribution
in Horse no.2 also brought the center of vertical loading closer to the midline of the body,
but it is not known why the loading shifted toward the lame limb. In mild lameness, it is
possible that certain gait variables actually became more symmetrical, though a more
severe lameness would be expected to produce the more typical patterns of asymmetry.
When evaluating very mild lameness, it is important to take account of the fact
that even sound horses may show asymmetrical motion and weight-bearing patterns. It
can be difficult to determine whether these asymmetries are simply individual variation
due to sidedness or the effect of a real pathology. There is a gray zone between sidedness
and slight lameness (Buchner, 2001a). When comparing mean difference of vertical
impulse in the right hind limb (Table 4-6), three horses were assessed as grade 2 lame,
and one horse was assessed as grade 1 lame. By making direct comparisons within the
same limb between sound and lame conditions, the effects of sidedness in the individual

horses were reduced. Therefore, comparisons between conditions in the same limb might

150

be more suitable in evaluating mild lameness than the asymmetry indices that have been
used in more severe lameness.

Due to the small sample size in this study, there was a risk of Type II statistical
errors (Vincent, 1999). Therefore, variables that showed a trend (P<0.10) toward
differences between sound and lame conditions are reported in addition to those showing
significant differences (P<0.05). If a larger number of subjects had been available, the
increase in statistical power might reveal additional differences between sound and lame

conditions.

Intra-limb coordination. During the stance phase, the main functions of the
tarsal joint are to absorb energy during the impact phase and to generate energy during
push off. The effects of distal tarsal joint pain were apparent in the tarsal joint mechanical
profiles during stance. The mean decrease in tarsal joint range of motion during stance
was 2.2 degrees. Since there was no significant difference in peak flexion or extension, it
seems likely that the decrease in range of motion was due to small decreases in both peak
flexion and peak extension, as shown by a trend toward a decrease in tarsal joint range of
flexion during the first half of stance, and a trend toward a decrease in tarsal joint range
of extension during the second half of stance.

Most of the significant changes in tarsal joint function occurred in early and late
stance. During early stance, the decreased rate of performing negative work (tarsal power
Minl) on the extensor aspect of the tarsus could be due to reduced activity of the
gastrocnemius. The SDF tendon, assisted by the DDF tendon, might be responsible for

the trend toward a decrease in energy absorption during this time by absorbing a little less

151

elastic energy as the joint accepted load. During terminal stance, there was a trend toward
a decrease in tarsal joint range of extension. A kinematic study of distal tarsal joint
lameness (Kramer, et al., 2000) also reported a decrease in maximum tarsal joint
extension at the end of stance. The decrease in rate of performing positive work during
late stance was probably also an attempt to reduce peak power on the extensor aspect.
The DDF muscle is active to provide propulsion at this time (Jansen, et al., 1992) and the
decrease in energy generation in late stance was most likely due to a decrease in muscular
work from this muscle. The DDF muscle also affected the energy profile at the distal
interphalangeal joint, causing a trend toward decreases in peak negative power and
energy absorption. Changes in the tarsal joint due to synovitis during early and late stance
corresponded with the occurrence of peak braking and propulsive forces. The propulsive
force showed a trend toward a decrease in peak magnitude. This might be an attempt to
adjust the limb function to reduce shear force acting on the tarsal joint complex.

The main firnction of the tarsal joint during swing is to raise and lower the distal
limb, so that the hoof swings clear of the ground. The mechanical events in late stance
affected the subsequent swing phase. The decrease in positive work of the tarsal extensor
muscles during late stance led to a more gradual increase in joint extension with less
stretching of peroneus tertius. The elastic energy storage in this tendon decreased, with a
corresponding decrease in the amount of elastic energy released during early swing. This
explains the significant decrease in peak net flexor moment (Min1) at the tarsal joint
around 10% of swing. During early swing, the rate of performing positive work (power
Max1) on the flexor aspect did not change, which might indicate a constant rate of

performing positive work by the cranial tibial muscle during this period. Therefore, the

152

trend toward decreased energy generation in early swing was thought to be due to a
decrease in elastic energy release from peroneus tertius.

In the second half of swing, tarsal extensors, most likely gastrocnemius muscle,
reduced the rate of performing positive work without a decrease in energy generation.
This again suggests an attempt to reduce the peak power. The reduction in rate of work
performed continued into early stance as described previously.

In the present study, there was a significant decrease in tarsal joint range of
motion and a trend toward a decrease in tarsal joint range of flexion during stance. On the
contrary, lameness located in the hind hoof was associated with an increase in peak tarsal
joint flexion during mid stance (Buchner, et al., 1996b). This suggests a different
mechanism of compensation for lameness at different sites within the limb. In the hind
hoof lameness, increased flexion of the tarsal joint was thought to reflect a greater role in
shock absorption by using the tarsal extensor muscles to adjust limb loading (Buchner, et
al., 1996b). This implies an increase in negative work performed by the tarsal extensor
muscles, though this was not confirmed by inverse dynamic analysis. As a result, the
loading increased more gradually during mid stance and the peak force on the hoof was
reduced. In the distal tarsal joint lameness in this study, there were no compensatory
increases in shock absorbing duty from other joints within the lame limb. Instead, there
was a reduction of loading of the entire limb.

Kinematic characteristics of distal tarsal osteoarthritis (bone spavin) are a
reduction in tarsal joint flexion during swing and a lower height of the foot flight arc

(Gough and Munroe, 1998). Synovitis induction in this study was associated with a trend

153

toward a decrease in vertical displacement of distal interphalangeal joint during swing,
but there was no change in tarsal range of motion during swing.

One of the surgical procedures for treating bone spavin is cunean tenectomy
(Cottage, et al., 1997 ; Bohanon, 1999), the objective of which is to relieve tension from
this tendon during contraction of the cranial tibial muscle. Reports on the efficacy of this
procedure are limited. Since the rate of performing work by the cranial tibial muscle and
the range of joint flexion during early swing in the present study did not change in the
lame condition, cunean tenectomy might not give a good response in mild lameness due
to synovitis. In horses affected with natural bone spavin, assessment of the owner’s
satisfaction with the outcome of the cunean tenectomy appeared to be good (Eastman, et
al., 1997), but that study did not report the level of lameness in the affected hind limb
before surgery. If lameness were more severe, it might have reduced the activity of the
cranial tibial muscle and relieved tension on the cunean tendon, so that cunean tenectomy
allowed an improvement in function. This could be investigated using inverse dynamic
analysis before and after surgery. Changes in net joint power and energy generation
during early swing might be a suitable quantitative indicator to assess the outcome and
might even be used as a prognostic tool.

The ability of a horse to advance to a high level in equestrian sports may be
limited by subtle lameness in the distal tarsal joints. In its early stages, osteoarthritis of
low-motion joints tends to cause performance-related problems rather than obvious
lameness. Most of these problems are subtle and slowly progressive (Moyer, etal., 1993).
Since the medial branch of cranial tibial tendon comes under tension during tarsal flexion

(Molenaar, 1983), the distal tarsal joints experience compression on their medial side

154

during joint flexion. When the horse performs collected work, especially piaffe and
passage, the femorotibial and tarsal joints are more flexed during early and mid stance
(Holmstrom, et al., 1995a). In the study reported here, peak joint flexion occurred closely
in time with peak negative power at the tarsal joint. Repetitive loading on the medial side
of the joint may lead to synovitis. In an experimental study of repetitive impulsive
loading in rabbit femorotibial joints, synovial inflammation was apparent after eight
weeks of loading. Cartilage breakdown was focal and limited to the weight-bearing area
(Lukoschek, et al., 1986). Low impact produced acute tissue stresses below the injury
threshold, while hi gh-intensity impacts produced stresses that exceeded the threshold of
joint degeneration (N ewberry, et al., 1998). In cartilage regions subjected to high contact
stress, synthesis of large aggregating proteoglycans was decreased and synthesis of
decorin was increased (Little, et al., 1997). Moreover, articular cartilage from aged horses
has markedly less overall metabolic activity, compared with cartilage from young horses
(Morris and Treadwell, 1994). In contrast to the catabolic effects of exercise, synovial
fluid from ponies that had been exercised for a week had anabolic effects on explanted
cartilage by enhancing glycosaminoglycan synthesis (Hoogen, et al., 1998). Ideally, the
training regimen of athletic horses should be adjusted to balance the catabolic and
anabolic effects. If the training regimen exceeds the physiological limit, joint
inflammation commences. The resulting pain may lead to a reluctance to perform

collected movements, which require a lot of tarsal joint flexion.

Inter-limb compensation. Quadrupeds show complex interactions and patterns

of compensation between the four limbs. This study showed a trend toward an increase in

155

vertical peak force and a significant increase in peak net joint moment on the flexor
aspect of metatarsophalangeal joint of the contralateral hind limb. Kinematic adaptations
suggestive of a compensatory increase in loading of the contralateral hind limb have been
reported in horses in which lameness was induced by pressure on the sole of the hoof
(Buchner, et al., 1996b), with the observed changes including an increase in maximal
flexion of the tarsal joint during stance. An increase in peak vertical force in the
compensating hind limb was also reported in dogs with anterior cruciate ligament
transection (Rumph, et a1, 1995). If horses that are mildly lame due to inflammation of
the distal intertarsal and tarsometatarsal joints continue to work, the increase in vertical
force on the contralateral hind limb may be sufficient to cause inflammation in the distal
tarsal joints, which may explain the fact that osteoarthritis in the distal tarsal joints (bone
spavin) often occurs bilaterally.

Lameness due to unilateral hind limb lameness may give the appearance of
lameness in the ipsilateral forelimb, when assessed by head acceleration asymmetry
(Uhlir, et al., 1997). In sound horses at trot, the head accelerates downward during the
first half of stance, then accelerates upward during the second half of stance. In unilateral
hind limb lameness, the vertical head acceleration during the contralateral forelimb stance
was more than that during the ipsilateral forelimb (Uhlir, et al., 1997). An increase in
head vertical acceleration amplitude has been reported on the diagonal of the lame hind
limb in horses with hoof lameness (Buchner, et al., 19960), but the finding was not
statistically significant among lameness grade 0, 1 and 2. Therefore, compensating head

movements are not a consistent finding in mild hind limb lamenesses.

156

In the present study, the lame condition showed a decrease in vertical impulse in
the contralateral forelimb, which is indicative of unloading of the diagonal lame limb
pair. There was a similar finding of decreased vertical force peaks in both forelimbs in a
dog with unilateral acute synovitis of the femorotibial joint (Rumph, et al., 1993), though
the exact mechanism of redistribution could not be identified. In induced carpal joint
lameness in horses, there also appeared to be a tendency toward overall reduction in
vertical forces. It might be that the horse maintained forward velocity with less vertical
displacement of the body center of mass, resulting in a less bouncy gait with lameness
(Morris and Seeherman, 1987). In induced fore hoof lameness, the vertical acceleration
of the body center of mass, which represents the total vertical force, reduced significantly
during the diagonal lame limb stance (Buchner, et al., 2001). In the synovitis condition,
the metacarpophalangeal and carpal joints of the contralateral forelimb, which are crossed
by several tendinous structures on their flexor aspect, showed decreases in peak flexor
moment in both joints, a trend toward a decrease in elastic energy absorption at the
metacarpophalangeal joint and a decrease in energy generation at the carpal joint (energy
Max2). This might reflect decreased bouncing of the body mass during lameness.

Horses affected with osteoarthritis of the distal tarsal joints often present with
back pain (Dyson, 1995; Cottage, et al., 1997 ; Gough and Munroe, 1998), which may
develop over a period of time in association with chronic pain in the distal tarsal joints.
This might indicate adaptation of back motion with this kind of lameness. There is an
interaction between movement of the limbs and movement of the vertebral column at the
trot (Faber, et al., 2001a). At the trot, the vertebral column can rotate in three dimensions:

flexion/extension, lateral bending and axial rotation. During the first half of hind limb

157

stance at the trot, which corresponds with the first half of the contralateral hind limb
swing, the inter-vertebral joints show extension, lateral bending toward the supporting
hind limb and axial rotation toward the swinging hind limb. During the second half of
stance, which corresponds with the second half of swing of the contralateral hind limb,
inter-vertebral motion involves flexion, lateral bending toward the swinging hind limb
and axial rotation toward the supporting hind limb (Faber, et al., 2001a). The skin marker
set in our study was not designed for direct analysis of back motion, but movements of
the sacrum might be inferred from the relative displacements of the left and right tubera
coxarum. During a complete stride, each tuber coxae had two cycles of down and up
motion. The highest position occurred during the suspension period and the lowest
position around mid stance of the ipsilateral and contralateral hind limbs. Vertical
displacement of the tuber coxae could be a result of either rotational motion of the
sacrum or adjustment of hind limb joints to accept load. In the lame condition in this
study, there was an increase in vertical displacement of the left tuber coxae during left
hind stance but, at the same time there was an increase in vertical displacement of the
right tuber coxae during swing of the right hind limb. Because the joint range of motion
in the left hind limb did not change, this is probably an indication of more lumbosacral
extension during the first half of stance of the left hind limb. After mid stance, a
propulsive force from the hind limb flexed the vertebral column by providing a forward
and upward force. The back flexion is thought to restrain by longissimus dorsi (Faber, et
al., 2001a), which is active twice around the period of back flexion during the complete
stride (Robert, et al., 1998). This muscle provides physiological rigidity of the vertebral

column, thus enhancing the forward propulsive thrust (Rooney, 1977). A trend toward an

158

increase in peak vertical force during stance might tend to increase flexion of the
vertebral column, which would require more activity from the longissimus dorsi to resist
the flexion. An increase in peak positive power at the left coxofemoral joint during early
swing, probably fi'om tensor fascia latae, might increase the momentum of the limb as it
swings forward, thus helping to propel the body forward, and maintain the body’s
forward velocity.

During the second half of stance in sound trotting horses, the coxofemoral joint
extends and the pelvis rotates axially toward the supporting hind limb, which facilitates
ground clearance of the contralateral hind limb as it is protracted (Faber, et al., 2001a). In
the lame right hind limb in this study, there was a decrease in the coxofemoral joint peak
extension (Maxl) during stance, which would make it more difficult for the left hind limb
to be protracted clear of the ground during the second half of swing. Gluteus medius
works primarily as a coxofemoral extensor and secondarily as a coxofemoral abductor
(Dyce, et al., 1987). A trend toward a decrease in coxofemoral energy generation (energy
Max2) during swing of the left hind limb might indicate a transition of gluteus medius
from performing positive work in the sagittal plane to performing more abduction. This
adjustment might help the protracting left hind limb to clear the ground but, at the same
time, might change the pattern of back motion. Possible increase in back extension during
stance, in combination with the possible changes in back motion during swing of the left

coxofemoral joint, may lead to fatigue of the back muscles over a period of time.

159

CONCLUSION

Differences in kinetic variables between sound and lame conditions were
identified using inverse dynamic analysis. In mild lameness, comparison of variables in
the same limb between conditions was more suitable than comparison of asymmetrical
variables between hind limbs. The energetic variables calculated from inverse dynamic
analysis gave more informative changes in the lame tarsal joint than kinematics or GRF
alone. Important characteristics of lameness due to synovitis in the distal intertarsal and
tarsometatarsal joints were a decrease in impact energy absorption during early stance
and a decrease in energy generation during push off at the tarsal joint in the lame hind
limb. These changes, in combination with a decrease in vertical impulse in the
contralateral forelimb, indicated unloading of the diagonal lame limb pair. Compensatory
changes were found in the contralateral hind limb, which showed a trend toward an
increase in peak vertical force and an increase in positive power at the coxofemoral joint
during early swing. A possible mechanism for causing back pain secondarily to lameness

in the distal tarsal joints was described.

160

CHAPTER 5

THREE—DIMENSIONAL KINEMATICS OF THE EQUINE TARSAL JOINT:
RESULTS AND DISCUSSION

REFERENCE BONE DATA

The true rotational and translational motion of the tarsal joint complex was
obtained from retro-reflective markers attached to bone pins inserted on the lateral side of
the tibia and metatarsus (Lanovaz, et al., 2002). Motion of the metatarsal segment was
described relative to the fixed tibial segment in terms of three rotational motions:
flexion/extension, abduction/adduction and intemal/external rotation, and three
translational motions: cranial/caudal, medial/lateral and proximal/distal.

The three-dimensional (3D) motions are illustrated as means and standard
deviations of rotational (Figure 5-1) and translational (Figure 5-2) motions of the tarsal
joint from reference bone data during stance from 4 horses. The values were expressed
relative to the impact value. During stance, the tarsal joint flexed, then extended, similar
to the motion found in the two-dimensional (2D) analysis. The distal part of the
metatarsal segment moved away from the mid line of the body relative to the tibia,
resulting in abduction during the first half of stance, after which the angle returned to the
impact value. Flexion peaked around 40% stance, which corresponded with the peaks in
abduction, cranial translation and proximal translation. The joint gradually extended,
adducted and caudally translated passing through the impact angle around 80% stance,

then continued these movements until the end of stance. The metatarsal segment was

161

internally rotated relative to the tibial segment throughout stance (Figure 5-1). The
medial and lateral translation patterns differed between horses, resulting in wide standard
deviations. The mean curve indicated that the metatarsal segment translated laterally
relative to the tibial segment throughout stance.

During swing, the rotational (Figure 5-3) and translational (Figure 5-4) motions of
the tarsal joint from reference bone data are illustrated and expressed relative to the
impact values in the previous stance phase. During swing, the joint flexed, abducted and
externally rotated with the peaks of these motions occurring at the same time around the
middle of swing, after which the joint moved back to the impact values. The
cranial/caudal and proximal/distal translational motion had similar patterns. The
metatarsal segment translated cranially and proximally, reaching peaks around the same
time as the rotational motion, then returned to impact values. The medial/lateral
translations, however, were quite varied among horses. During early and late swing, the
metatarsal segment translated laterally showing a large variation in the amount of lateral

translation in the middle of swing.

162

Figure 5-1 Three-dimensional rotation of the tarsal joint from reference bone data during
stance: flexion(-)/extension(+) angle (top), abduction(-)/adduction angle (middle) and
intemal(+)/external(-) rotation angle (bottom). The thick black line indicates mean value
from 4 horses. Thin lines indicate one standard deviation above and below the mean.
Zero indicates the impact value.

 

 

 

 

  

 

 

 

 

 

1° ' Extension
7; //
f
m ‘l
o
'3
2
ca
:
<
Flexion
-15 ..
0 20 40 60 80 100
5 ..
Adduction
7:?
g /\
.19; 0 ff
3.7
a:
C
<
Abduction
-5 a
0 20 4O 60 80 100
5 -
A Internal
(I:
g L
3. M
E 0 Mr I I I I
2
c»
C
<
External
-5 -
O 20 4O 60 80 100

°/o stance duration

163

Figure 5-2 Three-dimensional translation of the tarsal joint from reference bone data
during stance: cranial(+)/caudal(-) translation (top), medial(+)/lateral(-) translation
(middle) and proximal(+)/distal(-) translation (bottom). The thick black line indicates
mean value from 4 horses. Thin lines indicate one standard deviation above and below
the mean. Zero indicates the impact value.

 

 

 

 

 

 

 

 

20 -
A Cranial
E
E.
u..-
1:
o
E V
8 \
fl
'3. -10 -
.2
n
-20 - Caudal
0 20 40 60 80 100
5 '1
E Medial
E
‘E
g 0 I v I “$1
8
2
a.
.2
a
-5 . Lateral
0 20 40 60 80 100
5 q
’E‘
é
a
2
g o
8
2
a.
.93
o
'5 ‘ Distal
0 20 40 60 80 100

°/o stance duration

164

Figure 5-3 Three-dimensional rotation of the tarsal joint from reference bone data during
swing: flexion(-)/extension(+) angle (top), abduction(-)/adduction angle (middle) and
intemal(+)/ external(-) rotation angle (bottom). The thick black line indicates mean value
from 4 horses. Thin lines indicate one standard deviation above and below the mean.
Zero indicates the impact value.

15-

 

 
   

 

    

 

 

 

 

 

Extension
A 5 i\
h
g -15 .
v -25 -
.2
g -35 -
'45 ‘ Flexion
-55 -
O 20 4O 60 80 100
5 a
Adduction
<3 _
t
or
o
3?.
2
2’ -10 d
< I
Abduction
-15 -
0 20 40 60 80 100
5 -
,- Internal
CD
13
m
o
3
2
or
C
<
External
-10 -
O 20 4O 60 80 100

°/c swing duration

165

Figure 5-4 Three-dimensional translation of the tarsal joint fi'om reference bone data
during swing: cranial(+)/caudal(-) translation (top), medial(+)/lateral(-) translation
(middle) and proximal(+)/distal(-) translation (bottom). The thick black line indicates
mean value from 4 horses. Thin lines indicate one standard deviation above and below
the mean. Zero indicates the impact value.

 

 

 

 

 

 

 

60 -
E 50 - Cranial
5, 4o -
E
a, 30 ~
.5, 20 -
8
a 10 "
2 0 /
o I
-10 4% Caudal
0 20 40 60 80 100
5 1
’g‘ Medial
é
a
c /\/\
g o w . .
w /v\/‘/
0
.9.
a. M
.l’
D
-5 - Lateral
0 20 40 60 80 100
40 -
E 30 “ Proximal
E 20 -
o
E
g 10 -
2
a
'2 0 '5 ' I r I
o
'10 ‘ Distal
0 20 40 60 80 100

°/o swing duration

166

COMPARISON BETWEEN REFERENCE BONE DATA AND SKIN MARKER
DATA

Root mean square (RMS) errors were calculated as an indication of the amount of
deviation between the shapes of the two curves from reference bone data and skin marker
data. The values were calculated during stance for the three joint angles (Table 5-1) and
the three joint displacements (Table 5-2), and during swing for the three joint angles
(Table 5-3) and the three joint displacements (Table 5-4). In addition, RMS errors
between the reference bone data and skin data corrected for flexion/extension angle using
2D skin correction algorithms from the literature (van Weeren, 1989) were also
calculated during stance and swing phases. RMS errors between the reference bone data
and corrected skin data were smaller than those between reference bone data and
uncorrected skin data, indicating an improvement of the data after the application of skin
correction algorithms. Uncorrected skin data for flexion/extension obtained using 6 skin
markers per segment had smaller RMS errors than data from 2 markers per segment
corrected using algorithms published for 2D data.

The RMS errors only identified the magnitude of the errors. To explore
similarities and differences in the shape of the curves, the reference bone data and
corrected skin data from individual horses were plotted and evaluated qualitatively.
Agreement between curves was assessed as ‘good’ when the shape and direction of the
curves were similar for the two sets of data; ‘fair’ when the shape and direction of the
curves were less similar; and ‘poor’ when neither the shape nor direction of the curves
were similar. During stance, the shape agreement in individual horses for three rotational

motions (Figures 5-5 to 5-7) and three translational motions (Figures 5-8 to 5-10), and the

167

mean shape from all horses (Figures 5-11 and 5-12) were evaluated and reported in
Tables 5-1 and 5-2. During swing, the shape agreement from individual horses for three
rotational motions (Figures 5-13 to 5-15) and three translational motions (Figures 5-16 to
5-18), and the mean shape from all horses (Figures 5-19 and 5-20) were evaluated and
reported in Tables 5-3 and 5-4.

The curves showed better agreement in shape during the swing phase than during
the stance phase. Three motions were assessed as having fair shape agreement. In
abduction/adduction (Figure 5-1) and proximal/distal translation (Figures 5-2 and 5-4),
the reference bone data showed relatively wide standard deviations, indicating a
considerable amount of variation among the horses. The directions of motion were the
same but there was some disagreement between reference bone data and skin corrected
data (Figures 5-11, 5-12 and 5-20) in the magnitude of the movement. The fair agreement
for cranial/caudal displacement (Figure 5-12) during stance was mainly from the shape
difference during early stance, while the information around the middle of stance was
quite similar. Therefore, comparisons between conditions for these ranges of motion in
the same horse should be acceptable and, based on these findings, further analyses were
performed on the range of joint rotational and translational motions in the sound and lame
conditions. Due to poor agreement in shape of the curves for medial/lateral translation

and internal/ external rotation, they were not used for further analysis.

168

Table 5-1 Root mean square (RMS) errors of three rotational motions during stance

between reference bone data and corrected and uncorrected skin data. Shape agreement is

the assessment between reference bone data and 3D corrected skin data.

Flexion/extension (degrees)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Horse Bone data and 3D Shape Bone data and Bone data and 2D
corrected skin agreement uncorrected skin corrected skin
1 2.61 Good 2.65 4.23
2 3.31 Good 4.61 6.17
3 2.24 Good 3.28 5.21
4 1.34 Good 1.65 3.22
Mean 2.38 Good 3.05 4.71
Abduction/adduction (degrees)
Horse Bone data and 3D Shape Bone data and
corrected skin agreement uncorrected skin
1 0.37 Good 0.96
2 0.97 Fair 1.06
3 0.7 5 Fair 1.37
4 2.67 Fair 2.42
Mean 1.19 Fair 1.45

 

 

 

 

 

 

Internal/external rotatiorfidcgrees)

 

 

 

 

 

 

 

 

 

 

Horse Bone data and 3D Shape Bone data and
corrected skin agreement uncorrected skin
1 0.98 Fair 8.67
2 2.77 Poor 8.86
3 3.29 Poor 9.44
4 2.10 Poor 7.51
Mean 2.29 Poor 8.62

 

 

169

 

Table 5-2 Root mean square (RMS) errors of three translational motions during stance
between reference bone data and corrected and uncorrected skin data. Shape agreement is
the assessment between reference bone data and 3D corrected skin data.

Cranial/caudal translation (mm)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Horse Bone data and 3D Shape Bone data and
corrected skin agreement uncorrected skin
1 3.29 Fair 4.41
2 3.87 Fair 5.15
3 5.08 Fair 5.80
4 8.01 Fair 8.22
Mean 5.06 Fair 5.89
Medial/lateral translation (mm)
Horse Bone data and 3D Shape Bone data and
corrected skin agreement uncorrected skin
1 1.86 Fair 3.08
2 1.60 Poor 4.47
3 1.52 Fair 4.46
4 2.34 Poor 5.42
Mean 1.83 Poor 4.36
Proximal/distal translation (mm)
Horse Bone data and 3D Shape Bone data and
corrected skin agreement uncorrected skin
1 2.65 Fair 5.74
2 2.72 Fair 7.00
3 2.03 Fair 6.09
4 2.65 Fair 5.89
Mean 2.51 Fair 6.18

 

 

 

 

 

 

170

Table 5-3 Root mean square (RMS) errors of three rotational motions during swing

between reference bone data and corrected and uncorrected skin data. Shape agreement is

the assessment between reference bone data and 3D corrected skin data.

F lexion/extension (degrees)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Horse Bone data and 3D Shape Bone data and Bone data and 2D
corrected skin agreement uncorrected skin corrected skin
1 1.74 Good 2.00 6.10
2 2.73 Good 2.21 4.82
3 3.68 Good 4.98 8.29
4 3.15 Good 5.02 6.41
Mean 2.83 Good 3.55 6.41
Abduction/adduction (degrees)
Horse Bone data and 3D Shape Bone data and
corrected skin agreement uncorrected skin
1 2.35 Fair 6.73
2 0.69 Good 5.28
3 1.57 Good 6.01
4 1.89 Good 5.71
Mean 1.63 Good 5.93
Internal/external rotation (degrees)
Horse Bone data and 3D Shape Bone data and
corrected skin agreement uncorrected skin
1 3.41 Poor 8.40
2 1.68 Poor 8.68
3 1.34 Fair 8.96
4 4.50 Poor 11.34
Mean 2.73 Poor 9.35

 

 

 

 

 

 

 

Table 5-4 Root mean square (RMS) errors of three translational motions during swing
between reference bone data and corrected and uncorrected skin data. Shape agreement is
the assessment between reference bone data and 3D corrected skin data.

Cranial/caudal translation (mm)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Horse Bone data and 3D Shape Bone data and
corrected skin agreement uncorrected skin
1 9.78 Fair 10.46
2 2.60 Good 6.58
3 5.27 Good 11.71
4 8.59 Good 13.58
Mean 6.56 Good 10.58
Medial/lateral translation (mm)
Horse Bone data and 3D Shape Bone data and
corrected skin ageement uncorrected skin
1 2.87 Poor 3.56
2 1.98 Poor 10.23
3 1.59 Fair 9.31
4 5.14 Poor 7.04
Mean 2.89 Poor 7.54
Proximal/distal translation @m)
Horse Bone data and 3D Shape Bone data and
corrected skin agreement uncorrected skin
1 6.33 Fair 10.80
2 1.43 Good 12.06
3 8.12 Fair 18.50
4 4.89 Good 7.78
Mean 5.19 Fair 12.29

 

 

 

 

 

172

 

 

 

Figure 5-5 F1exion(-)/extension(+) angle of horses 1-4 (top to bottom) during stance.
The gray line indicates angle obtained from skin markers. The black line indicates angle
obtained after application of 3D skin correction algorithm. The dotted line indicates angle
obtained from bone pins. Zero indicates the impact value.

  

 

 

 

 

 

 

 

 

 

 

 

 

10 a
A s d 0
£2 / Extensron
i-
5!) 0
8
j: -5 1 Flexion
E“
- J
4 10
-15 ..
0 20 4O 60 80 100
’3
3 Extension
In
on
8
: Flexion
3
.15 ..
o 20 4o 60 80 100
10 -
A
(I)
Q C
2 Extensron
on
d)
'6
V
‘9 Flexion
‘ob
:1
«r
-15 ..
o 20 4o 60 so 100
10 -
A
U)
o O
2 Extensron
on
0
'1:
v
2 Flexion
on
i:
<1:
0 20 4o 60 80 100

% stance duration

173

Figure 5-6 Abduction (-)/adduction (+) angle of horses 1-4 (top to bottom) during stance.
The gray line indicates angle obtained from skin markers. The black line indicates angle
obtained after application of 3D skin correction algorithm. The dotted line indicates angle
obtained from bone pins. Zero indicates the impact value.

 

 

 

 

 

 

 

 

 

 

 

 

5 1
’8
O
Q
'55 Adduction
8
° Abduction
'30
i:
<
.5 .1
o 20 4o 60 so 100
5 '1
7r?
o O
‘5}, Adduction
8
2 Abduction
on
:1
<
.5 .1
0 2O 40 60 80 100
5 -
7r?
0
8‘0 Adduction
8
:7 Abduction
a
:1
<
-5 J
0 20 40 60 80 100
5 .
7r?
0
to
g f Adduction
Q)
a) Abduction
<

 

 

o 20 4o 60 so too
% stance duration

174

Figure 5-7 Internal (+)/external (-) rotation angle of horses 1-4 (top to bottom) during
stance. The gray line indicates angle obtained from skin markers. The black line indicates
angle obtained after application of 3D skin correction algorithm. The dotted line indicates
angle obtained from bone pins. Zero indicates the impact value.

 

 

 

 

 

 

10
’3 5
3 Internal
I-
§ 0 \ I I a
a: \
5; -5 - \_ ”/r//_\\ External
'51) \\
2 401 ‘\~-\

-15 1 \

o 20 4o 60 80 100

10 a
”a 5 a
8 Internal
EDD o 4 ‘2 FAQ
3 \ A External
2 '5 .1 s-..’ X
on \
g ‘10 " ‘.\\‘\\

“rm—«x
-15 .
0 20 40 60 80 100
10 -

5 ‘ Internal

 

 

Angle (degrees)
in o
' l
.‘ ?

 

External
. \\
‘10 " K \\ \
-15 - ‘xww
o 20 4o 60 80 100

10 -
A
U)
0
2 Internal
on
0
-o
v
‘9 Ex 1
.30 tema
:1
<1

 

 

o 20 4o 60 so 100
% stance duration

175

Figure 5-8 Cranial (+)/caudal (-) translation of horses 1-4 (top to bottom) during stance.
The gray line indicates angle obtained from skin markers. The black line indicates angle
obtained after application of 3D skin correction algorithm. The dotted line indicates angle
obtained fiom bone pins. Zero indicates the impact value.

 

 

 

 

 

 

 

 

 

20 .
E
E.
E
E Cranial
O
O
.2!
8‘
a Caudal
E
E
E
a Cranial
Q)
U
.2
8
'5 Caudal
o 20 4o 60 so 100
A 20 -
g 15 .
E 10 - / M - .. w
E 5 4 / Cranial
0
5 O f T I I
8'
'5 '5 ‘ Caudal
-10 a
o 20 4o 60 so 100
”I5
5
E’
0
as: Cranial
0
.2
o.
.21
‘3 Caudal

 

o 20 40 so so 100
% stance duration

176

Figure 5-9 Medial (+)/lateral (-) translation of horses 1-4 (top to bottom) during stance.
The gray line indicates angle obtained from skin markers. The black line indicates angle
obtained after application of 3D skin correction algorithm. The dotted line indicates angle
obtained from bone pins. Zero indicates the impact value.

 

 

 

 

 

 

 

 

 

 

10 -
A
E
E
«a 5 "
5
a Medial
Q)
U
.25.
8‘
'5 Lateral
A
E
E
v
H
5
E Medial
O
C)
.2!
8‘
5 Lateral
-5 a
0 20 40 60 80 100
10 ~
A
E
e ,flw/«m
a / Medial
o /
o A
.E , , . . .
b Lateral
-5 -
0 20 40 60 80 100
A 10 -
E,
1.: are-“NM“ fl-r" M"
5 5 T /\.//
E ' Medial
O
o.
2
a 5 Lateral
o 20 4o 60 80 100

% stance duration

177

Figure 5-10 Proximal (+)/distal (-) translation of horses 1-4 (top to bottom) during
stance. The gray line indicates angle obtained fiom skin markers. The black line indicates
angle obtained after application of 3D skin correction algorithm. The dotted line indicates
angle obtained from bone pins. Zero indicates the impact value.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

15 -
a r~~~x~~
1o .1 /,/" \
5 ,v
*5 Proximal
Q
E
0
8
'5. Distal
.2
G -10 .
o 20 4o 60 80 100
A
E
E
v
*5 Proximal
4)
E
Q)
8
3.. _5 _ Distal
.E
G
-10 -
o 20 4o 60 so 100
15 1
A
E
E
v
‘5 Proximal
0
E
0
8
"' Distal
% -5 1
n
-10 .
o 20 4o 60 80 100
A 15 -
E
E 10 -
v
‘a‘
o 5 ‘ Proximal
E
o 0
.E
,2 -5 Distal
G
-10
o 20 4o 60 so 100

% stance duration

178

Figure 5-11 Three-dimensional rotation of the tarsal joint obtained fiom all horses during
stance: flexion(-)/extension(+) angle (top), abduction(-)/adduction angle (middle) and
intemal(+)/extemal(-) rotation angle (bottom). The black line indicates mean value from
corrected skin data. The dotted line indicates mean value from reference bone data. Zero

indicates the impact value.

 

 

 

 

 

 

 

 

 

 

 

1° ' Extension
1?
E
m
a
3
2
2’
< '10 ‘ Flexion
-15 d
0 20 40 60 80 100
5 -
Adduction
1?
i
a:
o
3
2
m
C
< I
Abduction
-5 -
0 2O 40 60 80 100
5 -
A Internal
a:
E
a:
o
3
2
m
c
<
External
-5 -
0 20 40 60 80 100

% stance duration

179

Figure 5-12 Three-dimensional translation of the tarsal joint obtained fi'om all horses
during stance: cranial(+)/caudal(-) translation (top), medial(+)llateral(-) translation
(middle) and proximal(+)/distal(-) translation (bottom). The black line indicates mean
value from corrected skin data. The dotted line indicates mean value from reference bone
data. Zero indicates the impact value.

   

 

 

 

   
 

 

   

 

 

20 -
E Cranial
g 10 -
“
C
2
O
8
.9!
a
,9 -1o -
a Caudal
-20 J
o 20 40 so so 100
5 -
E Medial
E
0
g o < . 5‘3.
0 "'"n..e"h ............
2 ......................................
n. ...................
.2
0
Lateral
-5 J
o 20 4o 60 so 100
5 -
E
E Proximal
: ......
C .........
o .
25; o 4 ......
2 aaaaaaaaa
a
.2
a .
Distal
-5 -
o 20 4o 60 so 100

% stance duration

180

Figure 5-13 Flexion (-)/extension (+) angle of horses 1-4 (top to bottom) during swing.
The gray line indicates angle obtained fi'om skin markers. The black line indicates angle
obtained after application of 3D skin correction algorithm. The dotted line indicates angle
obtained from bone pins. Zero indicates the impact value.

 

 

 

 

 

 

 

 

 

 

Extension

A
m
8
‘60 Flexron
d)
1:
v
d)
3

15 Extension
A 5 LA
0 -5 . -
'5‘, Flexron
Q) -15 ‘
3
a -25 1
a -35 4

.55 -

o 20 4o 60 80 100
9 Extension
0
in
ca Flexion
1:
v
.9.
an
E
<
0 20 40 60 80 100

’5? Extension
0
0
'6‘»
o Flexion
'6
v
O
'51)
E
<1

 

 

0 20 40 60 80 100
% swing duration

181

Figure 5-14 Abduction (-)/adduction (+) angle of horses 1-4 (top to bottom) during
swing. The gray line indicates angle obtained from skin markers. The black line indicates
angle obtained afier application of 3D skin correction algorithm. The dotted line indicates
angle obtained from bone pins. Zero indicates the impact value.

 

 

 

 

 

 

 

 

 

 

1o -
§ Adduction
E‘s
%
z: Abduction
"6's
E
<1
§ Adduction
5'»
£3
5; Abduction
3b
i -10 ~

-15 4

o 20 40 60 so too

10 -
§ 5 ‘ Adduction
lab .
’5’
: Abduction
'S'o
E
<

-15 -

o 20 4o 60 so 100

10 «
§ Adduction
‘6'»
8
2.: Abduction
E”

 

 

o 20 4o 60 so 100
% swing duration

182

Figure 5-15 Internal (+)/external (-) rotation angle of horses 1-4 (top to bottom) during
swing. The gray line indicates angle obtained from skin markers. The black line indicates
angle obtained afier application of 3D skin correction algorithm. The dotted line indicates
angle obtained from bone pins. Zero indicates the impact value.

 

 

 

 

 

 

 

10 5
Internal
A
m
0
d)
to
Q
1, External
v
2
on
E
<1
0 20 4o 60 80 100
10 -
A 5 + Internal
8
o O
5::
.g '5 ‘ External
V d -
a '10 //
'15 "N I/
a \_/
<1 -20 .
-25 d
o 20 4o 60 80 100
Internal
A
(0
Q)
Q
'50
0
.5 External
v
2
5°
0 20 4o 60 80 100
*3 Internal
8
h
an
O
'3 External
O
E”

 

 

o 20 4o 60 80 100
% swing duration

183

Figure 5-16 Cranial (+)/caudal (-) translation of horses 1-4 (top to bottom) during swing.
The gray line indicates angle obtained from skin markers. The black line indicates angle
obtained afier application of 3D skin correction algorithm. The dotted line indicates angle
obtained from bone pins. Zero indicates the impact value.

 

 

 

 

 

 

 

 

 

 

 

70 7
E 5° ‘
g 50 -
IT 4M
5 304
E 20 a .
g 10 4 Cranial
'3. -
.2
G Caudal
A
E
E
v
H
:1
Q
E O
3 Cranial
.2
a.
.9
G Caudal
o 20 4o 60 80 100
A
E
E
v
H
s:
0
E O
8 Cranial
.2
a.
.Z‘.
G Caudal
o 20 4o 60 so 100
A
E
E
V
H
:a
Q
E
O
o O
.2 Cranial
a.
.9
G
Caudal
o 20 4o 60 80 100

% swing duration

184

Figure 5-17 Medial (+)/lateral (-) translation of horses 1-4 (top to bottom) during swing.
The gray line indicates angle obtained from skin markers. The black line indicates angle
obtained after application of 3D skin correction algorithm. The dotted line indicates angle
obtained from bone pins. Zero indicates the impact value.

 

 

 

 

 

 

 

 

20 -
’a‘
15 «
é
a 10 ~
5
E Media]
0
U
.E
a.
m
'5 Lateral
E
E
v
H
5
E Media]
6)
U
2
3'
'5 Lateral
-10 ..
0 20 4o 60 so 100
20w
E
15 J
5 10 ”My“
H -I ,
s / \ .
E 5 3“” W/ ‘~ Medial
0
E o . .
E
a '5 Lateral
A
E
E
V
H
5
E Medial
0
O
.2
8'
'5 Lateral

 

 

0 20 40 60 80 100
% swing duration

185

Figure 5-18 Proximal (+)/distal (-) translation of horses 1-4 (top to bottom) during
swing. The gray line indicates angle obtained from skin markers. The black line indicates
angle obtained after application of 3D skin correction algorithm. The dotted line indicates
angle obtained from bone pins. Zero indicates the impact value.

 

 

 

 

 

 

 

 

 

 

E
E
E
Q)
E
8
5 Proximal
a.
.2
G
Distal
E
E
E
Q
E
8
5 Proximal
a.
.2
a I
Distal
E
E
H
:1
Q
E
8
2 O
9: Proximal
.5
o 20 40 so so 100 mm“
E
E
H
t:
O
E
8
2 .
a. Proxnmal
a
-10 d
Distal

o 20 4o 60 80 100
% swing duration

186

Figure 5-19 Three-dimensional rotation of the tarsal joint obtained from all horses during
swing: flexion(-)/extension(+) angle (top), abduction(-)/adduction angle (middle) and
intemal(+)/extemal(-) rotation angle (bottom). The black line indicates mean value from
corrected skin data. The dotted line indicates mean value from reference bone data. Zero
indicates the impact value.

 

 

 

 

 

 

 

 

 

15 Extension
’0?
:3.
c»
o
3
2
a:
c Fl x n
< e io
5
Adduction
it?
.8.
c»
o
3
2
a:
c '10 ‘ Abduction
<
-15 d
0 20 40 60 80 100
5
A Internal
(0
t
a:
o
3
2
m
c n
< Exter al
-10 ..
O 20 4O 60 80 100

% swing duration

187

Figure 5-20 Three-dimensional translation of the tarsal joint obtained from all horses
during swing: cranial(+)/caudal(-) translation (top), medial(+)/lateral(-) translation
(middle) and proximal(+)/distal(-) translation (bottom). The black line indicates mean
value from corrected skin data. The dotted line indicates mean value from reference bone
data. Zero indicates the impact value.

  
  

 

 

 

 

 

   

 

 

60 -
E 50 - ---------
E 40 . Cranial
E 30 1
20 q
§
3“ 1O - .
a 0 ' ........ "
-1o ,. Caudal
o 20 40 6° 3° 10°
5 .
E Medial
5
‘E
o
E
a.
.3
a Lateral
-10 -
4o -.
E 30 . '° " Proximal
E 20 -
o
E
8 10 -
2 ’°
.3 o . """"
n
.10 Distal
o 20 40 5° 8° 10°

% swing duration

188

EFFECTS OF TARSAL LAMENESS ON 3D MOTION

Several values were measured during stance and swing to represent each type of
rotational or translational motion. The curve obtained from cranial/caudal translation
during stance indicated that the second peak from skin corrected data had the closest
value to the peak cranial translation from the reference bone data, and this peak was used
to represent peak cranial translation during stance. Angular and linear displacements
during stance were obtained from the impact value to peak flexion, to peak abduction, to
the second peak of cranial translation and to peak proximal translation. The ranges of
motion were calculated as the differences between the minimal and maximal values
during stance. During swing, the angular and linear displacements were obtained from the
impact value to peak flexion, to peak abduction, to peak cranial translation and to peak
proximal translation. The total ranges of motion during swing were calculated as
described for the stance phase. Dependent t-tests were used to compare the measured
variables in the sound and lame conditions (Table 5-5). The corresponding ranges of
motion from impact to peak flexion from 2D analysis of the right hind limb in the sound
condition (Chapter 3) were 9.77 (1.49) and 41.95 (2.65) degrees during stance and during
swing, respectively. The total ranges of joint flexion/extension (Chapter 3) were 12.44
(1.14) and 44.05 (1.78) degrees during stance and swing, respectively.

From Table 5-5, the significant decrease in range of flexion during the first half of
stance corresponded with the significant decrease in range of cranial translation and a
trend toward a decrease in proximal translation of the metatarsal segment relative to the
fixed tibia. In the normal tarsal joint, the rotational and translational motions occur

synchronously. Due to the shape of the talus, rotation is coupled with translation. To

189

quantify the magnitude of coupling between the other motions with the joint
flexion/extension, reference bone data were used to determine the ratios of the joint
rotation and translation relative to changes in flexion (Table 5-6). These ratios were
selected corresponding to the variables that differed significantly between sound and

lame conditions in Table 5-5.

190

Table 5-5 Differences of three-dimensional variables between sound and lame conditions

of the right hind limb. Values are mean and (SD).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Stance phase
Variables Sound Lame Mean
condition condition difference
Impact to peak flexion (degrees) 10.6 (1.7) 8.1 (0.6) _2. 5 (1.4) a
Total range of flexion/extension (degrees) 15.9 (2.8) 14.5 (1.9) -l.4 (1.3)
Impact to peak abduction (degrees) 1.6 (1.0) 1.6 (1.5) 0.0 (0.7)
Total range of abduction/adduction (degrees) 3.1 (0.8) 3.2 (1.8) 0.1 (1.1)
Impact to peak cranial translation (mm) 17.1 (3.0) 12.9 (2.1) -42 (2.5) a
Total range of cranial/caudal translation a
(mm) 20.4 (1.9) 16.9 (2.6) -35 (1,8)
Impact to peak proximal translation (mm) 5.1 (1.0) 3.8 (0.9) -13 (1_1)b
Total range of prox1mal/distal translation 8.2 (2.0) 7.1 (1.4) -l.l (1. 6)
(M)
Swing phase
Variables Sound Lame Mean
condition condition difference
Impact to peak flexion (degrees) 46.8 (4.0) 44.0 (3.4) -2.8 (3.4)
Total range of flexion/extension (degrees) 52.2 (3.0) 50.4 (1.7) -1.9 (3.4)
Impact to peak abduction (degrees) 9.4 (2.1) 9.1 (1.5) -0.3 (2.8)
Total range of abduction/adduction (degrees) 10.5 (2.4) 10.1 (2.1) -0.3 (3.4)
Impact to peak cranial translation (mm) 49.7 (11.9) 43.2 (8.5) -6.5 (5.9)
Total range of cranial/caudal translation 52 1 (7 4) 46 7 (5 5) _5 4 (6 2)
(mm) . . . . . .
Impact to peak proximal translation (mm) 39.5 (9.3) 37.1 (8.7) -24 (1.4) a
T t l f ' l/d' t l l '
o a range 0 proxima IS a trans ation 44.2 (8.4) 40.3 (7.7) -4.0 (3.1)b

(mfg)

 

 

 

 

 

Negative mean differences indicated values fromb sound condition were higher than
values from lame condition. 3 indicated P<0.05. indicated 0.05<P<0.1.

l9l

 

 

Table 5-6 Range of motion obtained from reference bone data during stance and swing

 

 

 

 

 

 

phases.
Variables Mean (SD) Magnitude
change *
Impact to peak flexion, stance (degrees) 6.23 (1.5) 1.00
Impact to peak cranial translation, stance (mm) 7.71 (0.62) 1.24
Impact to peak proximal translation, stance (mm) 1.87 (0.95) 0.30
Impact to peak flexion, swing (m) 33.43 (3.22) 1.00
Impact to peak proximal translation, swing (m) 23.47 (5.19) 0.70

 

 

 

 

* Magnitude change means the change in a variable with l-degree change in

flexion/extension angle.

DISCUSSION

The tarsal joint complex consists of four joints. The major motion is the

flexion/extension at the tarsocrural joint. The motions at the other 3 joints, proximal

intertarsal, distal intertarsal and tarsometatarsal joints, are considered to be small. Due to

the oblique orientation of the talar ridges, flexion/extension of the tarsocrural joint is

accompanied by translational motion, which is characteristic of helical motion (Badoux,

1987). The largest force loading the distal tibia is a torsional force (Schneider, et al.,

1982; Hartman, et al., 1984). The combination of torsional loading from the tibia and the

oblique orientation of the talus may lead to motion at the other three low-motion joints in

the tarsal joint complex. The objective of this study was to develop a non-invasive

method of measuring 3D motion of the tarsal joint complex. Since the segments between

the three distal tarsal joints are extremely short, it is not possible to attach skin markers to

represent each segment of the tarsal joint complex, though some assumptions could be

made regarding the motion of these joints. The tarsocrural joint has a screw motion

192

 

(Badoux, 1987; Lanovaz, et al., 2002), which implies that rotational and translational
motions of the tarsocrural joint in any direction are highly coupled with flexion/extension
of this joint. Loss of coupling between flexion/extension and the other motions suggests
movement at tarsal joints outside the tarsocrural joint (Lanovaz, et al., 2002).

The flexion/extension angle obtained from skin markers using 2D skin correction
algorithms (van Weeren, 1989) had higher RMS errors compared to the flexion/extension
angle obtained from uncorrected 3D skin data (Tables 5-1 and 5-3). This was because the
number of markers per segment in 3D skin marker set was higher and the markers were
more widely distributed over the segment than in the 2D skin marker set. The tarsal joint
motion from the 2D marker set was constructed from 3 markers connected by two
straight lines representing the tibial and metatarsal segments. Tarsal joint motion from the
3D marker set was constructed from 12 markers distributed over the two segments. The
motions of all 12 markers were taken into account for calculating the segmental motion.
Therefore, even uncorrected, the 3D skin marker set could represent the motion of the
segment better than the straight line in the 2D marker set. The tibial markers underwent
much larger displacements than those on the third metatarsus (Appendix A). The 3D skin
correction algorithm reduced errors due to skin displacement, resulting in lower RMS
errors in joint flexion/extension compared with the reference bone data. It was shown in
this study that the skin corrected 2D marker set might underestimate the flexion/extension
ranges of motion during both stance and swing phases.

The correction algorithms for skin displacement are important for obtaining
accurate kinematic data from skin-based markers. In order to study 3D kinematics of the

tarsal joints using skin-based markers, it was necessary to develop 3D skin correction

193

algorithms based on knowledge of the motion of the underlying bones. The advantage of
this study was that the same horses were used for two data collection sessions.
Approximately one month afier recovery from synovitis, the same horses were used in
the bone pin study of tibial and third metatarsal kinematics (Lanovaz, et al., 2002), the
data from which were used to construct skin correction algorithms (Appendix A). By
using the same horses for both studies, it enhanced the ability to test how closely the skin
correction algorithms could correct the skin marker data to the real bone motion in the
individual horse.

Data from the reference bone motion indicated that some variation among horses
existed in rotational and translational motions. Therefore, the skin correction algorithms,
which were calculated as the best fit for a group of horses, might not be able to account
for individual variation in all motions. Even though the RMS errors for internal/external
rotation and medial/lateral translation were small, their ranges of motion were also small.
In addition, the shape agreement between the reference bone data and skin corrected data
were poor. Therefore, it is not suggested that this 3D marker set be used for testing any
changes in these two motions. During stance, abduction/adduction, proximal/distal
translation and cranial/caudal translation were evaluated as having fair shape agreement.
As described previously, the shape disagreement for those first two motions were due to
different in the magnitude of the movement, but the directions of motion were correct.
The fair agreement for cranial/caudal displacement was mainly from different shapes in
early stance, while the information around the middle of stance was quite similar.

Therefore, the comparisons between conditions for these ranges of motion in the same

194

horse should be acceptable, but direct comparison between different horses should be
avoided due to underestimation or overestimation of joint motion from skin-based data.

During swing, the shape agreements for internal/external rotation and
medial/ lateral translation were also poor as described during stance. Therefore, it is not
suggested to use this 3D skin marker set to obtain information describing these two
motions during swing. Proximal/distal translation showed good agreement in shape,
although proximal displacement in mid swing was overestimated (Figure 5-20), the value
still fell within one standard deviation of the reference bone data. Therefore, swing phase
flexion/extension, abduction/adduction, cranial/caudal translation and proximal/distal
translation ranges of motion can be estimated with acceptable accuracy from corrected
skin data.

Since the placements of 3D skin markers in the same horse were similar between
sound and lame conditions, the skin correction algorithms should correct for skin
displacement in a similar way. Therefore, changes in the 3D range of motion should
represent the effect of synovitis in the distal intertarsal and tarsometatarsal joints on the
3D tarsal joint motion. The 2.5 degrees decrease in joint flexion when pain existed in the
distal intertarsal and tarsometatarsal joints was similar to the 2.2 degrees decrease in the
tarsal joint range of motion during stance obtained from 2D analysis. From reference
bone data, the cranial/caudal translation was not completely proportional to the
flexion/extension angles during stance. The additional cranial translation of the
metatarsus relative to tibia might come from superposition of motion outside the
tarsocrural joint (Lanovaz, et al., 2002). According to Table 5-6, the 2.5 degrees decrease

in flexion angle during stance should be coupled with 3.1 mm decrease in cranial

195

translation. The significant decrease of 4.2 mm in cranial translation exceeded the
calculated values accompanying the decrease in flexion angle, which suggests a decrease
in cranial translation outside the tarsocrural joint of around 1.1 mm.

During stance when the tarsal joint flexes, the medial branch (cunean tendon) of
the cranial tibial tendon comes under tension (Molenaar, 1983). The insertion of the
tendon might pull the fused first and second tarsal bone forward, consequently,
compressing and pushing the central and third tarsal bones forward. On the right hind
limb in the lame condition, the gastrocnemius adjusted by reducing the high peak
negative power during the first half of stance (Chapter 4). This corresponded to the time
when tarsal joint range of motion decreased from impact to peak flexion. This might
reduce compression on the central tarsal bone in an attempt to reduce the motion due to
pain from synovitis, leading to reduction of cranial translation in the lame limb.

In the normal joint, motion is essential for maintaining healthy cartilage. The
movement of nutrients, water and waste within the cartilage matrix is facilitated when the
articular cartilage is cyclically loaded during locomotion (Pool, 1996). Due to the flat
shape of the central tarsal and third tarsal bones, motion of the distal intertarsal and
tarsometatarsal joints is likely restricted to translation and axial rotation. Information
obtained from the reference bone data (Lanovaz et al, 2002) indicated the presence of
axial rotation and cranial/caudal translation of the metatarsal segment relative to the fixed
tibial segment during stance that was not coupled with flexion/extension of the
tarsocrural joint. Although synovitis might be associated with changes in internal/external
rotation, the correction algorithm for estimating internal/external rotation from the skin

markers was shown not to be sufficiently accurate for general use. The decrease in cranial

I96

translation that occurred in horses with synovitis could not be separated among the joints
in the tarsal joint complex, but the change between the sound and lame conditions was
assumed to be the effect of synovitis on the motion of the distal intertarsal and
tarsometatarsal joints. This small, but significant, decrease in cranial translation might
represent a stage in pathogenesis of distal tarsal osteoarthritis. Examination of specimens
in different stages in development of bone spavin strongly suggests that the cause is
probably trauma to the periarticular soft tissues (Pool, 1996). The resulting pain and
inflammation reduces the joint movement and gliding motion ceases. Formation of
enthesophytes and periosteal new bone formation at the joint capsule insertion line
strengthens the traumatized tissues, and also restricts the gliding motion of the distal
tarsal joints. Repetitive forces of weight bearing are transmitted through the stationary
distal tarsal joints, resulting in full-thickness necrosis of articular cartilage and focal
subchondral bone remodeling at the area of force transmission through the joints.

Strain gauges mounted on the third metatarsus of walking ponies showed that the
point of force application at the tarsometatarsal joint was located in the dorsomedial
portion of the joint surface during the first peak of vertical GRF, and was located in the
dorsolateral portion during the second peak of vertical GRF (Schamhardt, et al., 1989).
At trot and canter, the third metatarsus was loaded with a higher compressive force than
at walk and the third metatarsus was closely aligned with the ground reaction force
throughout most of the stance (Biewener, et al., 1988). Unfortunately, the latter study did
not indicate the point of force application at the tarsometatarsal joint. The radiographic
changes in horses affected with bone spavin are mostly found on the dorsomedial side of

the distal tarsal joints (Cottage, et al., 1997). This area of the joint surfaces was examined

197

histologically in studies designed to assess early pathological changes in the distal tarsal
joints (Laverty, et al., 1991; Bameveld and van Weeren, 1999), which might be the area
through which the load is transmitted across the distal tarsal joints. The reduction in
cranial translation in horses with synovitis of the distal tarsal joints would lead to
stationary joint surfaces in this area. If the horses are subjected to high intensity work, the
load transmitted through this area might exceed the physiological limit and lead to
necrosis of articular cartilage. If this hypothesis is correct, it could support the
pathogenesis of bone spavin proposed by Pool (1996).

During swing, there was a significant decrease of 2.4 mm in the range of proximal
translation at the tarsal joint complex without a change in the range of joint flexion. This
decrease might be related to the snapping motion of the tarsal joint that is due to the
eccentric attachment of the collateral ligaments relative to the center of rotation of the
tarsal joint. The collateral ligaments are stretched when the joint is fully extended at the
end of stance, thus storing elastic energy. As the joint flexes, the tibia moves around the
talar groove until it passes the critical point at which there is no tension on the ligament.
At that point, the collateral ligament recoils and elastic energy is released (Rooney,
1990), resulting in a snapping motion of the tarsal joint. Because the collateral ligaments
attach with a proximal/distal orientation, some of the elastic energy release might
contribute to the proximal/distal translation of the tarsal joint complex. None of the other
tested variables changed in association with the decrease in proximal translation found in
this study. In the reference bone data (Figures 5-1 and 5-2), joint extension at the end of
stance was accompanied by internal rotation and lateral translation. Since the skin

corrections were inadequate to assess these two motions, information between sound and

198

lame conditions could not be compared. It is theorized that when lameness occurs in the
distal intertarsal and tarsometatarsal joints, it limits joint motion by decreasing internal
rotation and lateral translation, which might decrease the distance between proximal and
distal attachments of the lateral collateral ligament when the joint was fiilly extended.
Therefore, elastic energy storage decreases, with a consequent decrease in elastic energy
release to generate the proximal displacement.

The range of motion from impact to peak flexion during stance decreased by 2.5
degrees when synovitis was present, corresponding to a decrease in tarsal joint range of
flexion/extension of 2.2 degrees as determined by 2D skin markers. The 2D kinematic
analysis offers a relatively quick and easy marker setup and appears to provide
reasonable information to assess the effect of different conditions on flexion and
extension of the joint. In horses affected with natural bone spavin, the point of force
application to the hoof, assessed by a force plate analysis, was more caudal and lateral
than in normal horses, which was believed to be an adaptation to reduce loading in the
painful medial aspect of the tarsal joint (Boswell, et al., 2000). Even though the 3D skin
marker set in this study provided fair assessment of magnitude of 3D motion during
stance, it might be interesting to apply the 3D marker set to observe the pattern of tarsal
joint motion which might correspond to the more caudal and lateral loading in Boswell,
et al. (2000). Conservative treatment of bone spavin includes balancing and trimming the
foot and the application of special shoes, such as the rolled toe shoe with a lateral
extension or lateral trailers (Dyson, 1995), raised heel/rolled toe shoe (Cottage, et al.,
1997) or lateral extension shoe (Gough and Munroe, 1998). The 3D kinematics, in

combination with the analysis of 2D tarsal net joint energy profiles might enhance the

199

quantification of biomechanical changes after the application of special shoes, as a
method of testing the usefulness of each type of orthopedic shoes. The relatively good 3D
information obtained from the 3D skin marker set during swing allows analysis and
comparison of information between horses. It may be useful for assessing motion other
than flexion/extension, such as the extreme motion in the athlete during performance, or
the stabbing motion of the tarsal joint during swing (Dyson, 1995; Gough and Munroe,

1998).

CONCLUSION

The tarsal joint complex has motion outside the sagittal plane. The motions during
stance include flexion, abduction, internal rotation, cranial translation, lateral translation
and proximal translation of the metatarsus relative to the tibia. The motions during swing
were similar to those during stance, except for external rotation of the metatarsus relative
to the tibia. The skin marker set in this study could be used to assess the motion during
swing fairly accurately. During stance, it is acceptable to compare the motion within
horses. The ranges of motion during swing were larger than those during stance. Distal
tarsal synovitis resulted in decreases in ranges of tarsal joint flexion, cranial translation
during stance and a decrease in proximal translation during swing. The findings might
support the hypothesis that synovitis is the primary event in the development of

osteoarthritis in distal tarsal joints.

200

CONCLUSIONS

Gait analysis techniques, including kinematic and force plate analysis, have been
successfully applied to study normal and pathological gaits in horses. In the present
study, inverse dynamic analysis was used to study the two-dimensional net joint
mechanical work of the equine hind limb at the trot. Net joint moment, net joint power
and net joint energy profiles of the coxofemoral, femorotibial, tarsal, metatarsophalangeal
and distal interphalangeal joints in sound trotting horses have been identified. The
biomechanical profiles of the hind limb joints were similar between horses after
adjustment for the effect of subject velocity and body mass, which supports Hypothesis 1.
During stance, the coxofemoral and tarsal joints were the main sources of energy
generation to propel the body forward. During trotting, energy expenditure reduced by
elastic energy conservation, especially at the metatarsophalangeal joint, assisted by the
tarsal and femorotibial joints. The distal interphalangeal joint absorbed energy and acted
as an energy damper. During swing, the main sources of energy generation were the
coxofemoral and tarsal joints. The femorotibial joint performed mostly negative work to
control the limb motion, while the distal interphalangeal and metatarsophalangeal joints
were driven by the inertia, following the motion of the proximal joints.

Induction of synovitis by injecting lipopolysaccharide endotoxin in the distal
intertarsal and tarsometatarsal joints was performed to study the effect of tarsal lameness
on equine locomotion at the trot. The lameness in this study appeared to be mild, which
made it difficult to identify asymmetrical motion between the left and right hind limbs.

Comparisons between lame and sound conditions in the same limb proved more suitable

201

for evaluating mild lameness than the asymmetry indices that have been used in more
severe lameness. There were several kinematic and kinetic changes within the lame limb
and between different limbs when compared between the sound and lame conditions,
which supports Hypothesis 2. In the lame tarsal joint during stance, there were decreases
in peak vertical force, tarsal joint range of motion, peak negative net joint power during
early stance, peak positive power and positive energy during push off. Mechanical energy
deficits were found in the lame tarsal joint by decrease in total positive and negative net
joint energy. During swing, there were decreases in peak flexor moment during early
swing and peak positive net joint power during late swing. There was a trend toward a
decrease in total positive net joint energy at the tarsal joint during swing. There was no
compensatory increase in energy generation from other joints within the lame limb.
Instead, the load was redistributed between the limbs, so that the lame diagonal pair was
unloaded, with a reduction of the diagonal forelimb vertical impulse in the lame
condition. The contralateral hind limb compensated by showing a trend toward an
increase in peak vertical force. During early swing, there was an increase in positive
power at the left (compensating) coxofemoral joint, which increases the momentum of
the limb as it swings forward, thus helping to propel the body forward, and maintain the
body’s forward velocity.

Three-dimensional (3D) kinematics of the right tarsal joint was studied using an
array of skin-based markers to study rotational and translational motions of the third
metatarsus relative to the tibia. Six markers were attached on each bone segment.
Comparison between reference bone data and skin-based marker data indicated that tarsal

joint complex has the motion outside the sagittal plane and the motion can be identified

202

using skin markers, which supports Hypothesis 3. F lexion/extension and
abduction/adduction rotational motions and proximal/distal and cranial/caudal
translational motions could be measured fairly accurately during swing, and the motions
during stance were acceptable for comparison within horses. Internal/external rotational
and medial/lateral translational motions during stance and swing were not successfully
obtained from this skin marker set. Tarsal synovitis in the distal intertarsal and
tarsometatarsal joints resulted in a reduction of the joint flexion and cranial translation
during stance, and a decrease in proximal translation of the third metatarsus relative to
the tibia during swing.

From this study, it can be stated that tarsal joint is an important source of positive
work to move the limb throughout the stride. Energy generation at the tarsal joint came
from both muscular work and elastic energy conservation. Elastic energy storage and
release occurs mainly during stance. When mild lameness occurs in the distal tarsal
joints, adjustments in the locomotion pattern were identified using the inverse dynamic
analysis. The horse maintained its forward velocity with minimal changes in kinematics
making visual detection of lameness difficult in the early stage of synovitis. The
availability of data describing the horse’s locomotor characteristics when sound might be
worthwhile for monitoring that horse through time. The technique could be applied
objectively to evaluate the outcome of treatment regimes in horses affected with pain in

the distal intertarsal and tarsometatarsal joints.

203

APPENDIX A

THE DEVELOPMENT OF CORRECTIONS FOR SKIN DISPLACEMENT
ARTIF ACTS AT THE EQUINE TARSAL JOINT AND APPLICATION TO 3D
JOINT KINEMATICS

Joel L. Lanovaz, Siripom Khumsap, Hilary M. Clayton

ABSTRACT

In horses, the tarsus is arguably the most critical hind limb joint with respect to
both performance and injury. Previous kinematic investigations have focused on 2D
planar motion. Routine study of 3D tarsal kinematics is hampered by errors due to the
displacement of skin surface tracking markers relative to the underlying bones. Reliable
kinematics can be obtained with bone-fixed markers, but an accurate, non-invasive
method would have more applications.

Simultaneous kinematic data from skin-based and bone-fixed markers attached to
the tibia and third metatarsus were collected from three trotting subjects. The 3D skin
displacement patterns for the skin-based markers were parameterized using a truncated
Fourier series model. These displacements were expressed in terms of the local
coordinate systems for each bone. The 3D kinematics of the segments for any stride
could then be calculated by relating the motion of the skin-based markers in global
coordinates and the local skin displacements using an SVD based algorithm. The results
showed that the displacement model could reduce the RMS error in the joint kinematics
35 to 65% over uncorrected skin-based data. When the models were applied to two

additional independent subjects, the shapes of the resulting traces matched well with the

204

bone-fixed kinematics. The errors in tarsal internal/external rotation estimation and
translation with respect to the mediolateral axis of the tibia from the corrected data were
still too high to accurately reproduce those degrees of freedom. However, the remaining
rotations and translations were estimated with sufficient accuracy to be used in general

analysis.

INTRODUCTION

The tarsal joint is one of the primary sites of lameness in the equine hind limb
(Gabel, 1983; Winter et al., 1996; Gough and Munroe, 1998) and its basic kinematic
parameters have been correlated with sport horse performance (Holmstrom et al., 1995).
Several studies have reported planar 2D equine tarsal joint motion (Kobluk et al., 1989;
Holmstrém et al., 1994; Back et al., 1995; Hodson et al., 2001), but only a few studies
have examined 3D equine tarsal kinematics either in vitro (Schamhardt et al., 1984) or in
vivo (Lanovaz et al., 2002).

When skin surface markers are used to track segmental motion, 3D kinematics
calculated from these markers are especially sensitive to skin motion artifacts. Relative
displacements between skin and bone at some sites over the equine tibia have a range of
motion of over 80 mm during a stride at the trot (van Weeren et al., 1992). Marker
cluster design and placement can reduce some errors (Cappozzo et al., 1997), but actual
underlying bone motion is still difficult to reproduce (Cappozzo et al., 1996). This is true
even when utilizing a redundant marker set and a least-squares estimator of the

transformation data (Reinschmidt et al., 1997 a).

205

Kinematic data collected using bone-fixed markers have been used to describe the
true underlying in vivo kinematics of joints in human subjects (Lafortune et al., 1992;
Cappozzo et al., 1996; Reinschmidt et al., 1997b) and in horses (van Weeren et al., 1992;
Lanovaz et al., 2002). The use of bone-fixed markers is invasive and it is desirable to
develop procedures that avoid the need for surgical intervention but still produce an
acceptable degree of accuracy.

Skin displacement artifact correction algorithms have been successfully
developed for 2D kinematics in horses (van Weeren et al., 1992) and this procedure has
been shown to improve joint angular data over non-corrected kinematics. The feasibility
of using 3D correction algorithms has not been investigated in equine subjects.

Alexander and Andriacchi (2001) suggested a 3D skin correction procedure based
on the point cluster technique (Andriacchi et al., 1998) that utilized information about
skin displacement in the local bone coordinate system. The current study employs a
similar but more direct approach using a singular value decomposition (SVD) algorithm
(Séderkvist and Wedin, 1993). In this study, simultaneous kinematic data collected from
skin surface and bone-fixed markers from the tibia and third metatarsus of equine
subjects are used to parameterize the general motion of skin marker sites relative to the
underlying bone. Models of the skin motion are developed and these displacement
models are then used in the SVD procedure to calculate 3D kinematics from skin surface

markers.

206

METHODS

Subjects. The subjects were 5 sound Quarter Horses of similar mass and size.
The horses were judged to be free of obvious lameness during a clinical examination
prior to the study. The protocol used for this investigation was approved by the

university’s ethical use committee.

Bone based markers. Subjects 1-3 were utilized for the following bone-fixed
kinematic protocol. A detailed explanation of the bone-fixed procedure has been
presented elsewhere (Lanovaz et al., 2002). In summary, 4.75 mm diameter Steinmann
pins were inserted percutaneously under general anesthesia into the tibia and third
metatarsus of the right hind limb of each of the three subjects. A marker triad with 25
mm diameter reflective spherical markers was rigidly attached to each pin during data
collection session. Analgesics were administered systemically and locally, and all

subjects appeared pain free and moved normally during data collection.

Skin-based markers. For all 5 subjects, a set of six 25 mm diameter, spherical
reflective markers were attached at predetermined locations on the surface of the skin
over the right tibia and the right metatarsus of each subject (Fig 1). This was in addition
to the bone-fixed marker triads for subjects 1-3. The distance between markers TIB-A
and TIB-B served as the reference length for the tibia and the distance between markers
MET-A and MET-B served as the reference length for the third metatarsus. The skin-

based markers were distributed over the lateral aspect of each segment, since most studies

207

track kinematic data from the lateral side. The tibial markers were concentrated

proximally in order to minimize visual merging with the bone pin marker triads.

Bone coordinate systems. In order to define local bone-based coordinate systems
(BCS), all subjects were placed in a normal standing position and an additional reflective
marker was placed over the medial malleolus of the tibia (TIB-C) and the dorsal edge of
the head of second metatarsal bone (MET-C). These markers were removed for data
collection.

A right-handed coordinate system was developed for the tibia by first defining the
flexion/extension axis as the vector running from TIB-B to TIB-C. The
adduction/abduction axis of the tibia was defined as a vector pointing cranially and
perpendicular to the plane formed by the flexion/extension axis and the vector running
from TIB-B to TlB-A. Finally, the internal/external rotation axis of the tibia was defined
as a vector pointing proximally along the long axis of the bone perpendicular to the plane
formed by the flexion/extension and adduction/abduction axes. The origin of the tibial
BCS was embedded in the bone midway between TIB-B and TIB-C.

The third metatarsal BCS was defined in a similar manner to the tibial system
with the flexion/extension axis defined as a vector running from MET-B to MET-C. The
adduction/abduction axis was defined as a vector pointing cranially and perpendicular to
the plane formed by the flexion/extension axis and the vector running from MET-B to
MET-A. The internal/external rotation axis was defined as pointing proximally and
perpendicular to the other two axes, forming a right-handed coordinate system. The

origin of the third metatarsal BCS was embedded in the bone midway between MET-B

208

and MET-C. For both the tibia and third metatarsus, the adduction/abduction,
flexion/extension and internal/external rotation axes are also referred to as the x, y and z
axes respectively.

Data Collection. All subjects were led in hand at the trot along a 40 m rubber
covered, concrete runway. Three dimensional kinematic data were collected using a 6
camera analysis system (ExpertVision RealTime, Motion Analysis Corporation, Santa
Rosa, CA) recording at 120 frames per second. A volume measuring 5 m by 2 m by 3 m
was calibrated and the mean error in measuring a known length within the volume was
0.88 mm. Each successful trial consisted of a single stride of the right hind limb, starting
with stance. Data collected from a force platform (LG6-4-8000, AMTI, Watertown, MA)
embedded in the runway were used to detect the onset and termination of the right hind
stance. Termination of a right hind stride was determined by an automatic algorithm that
matched terminal swing phase kinematics to the kinematic pattern at the beginning of the
previous stance phase. The kinematic data from the trot trials were optimally filtered
using a Generalized Cross Validation (GCV) spline routine (Woltring, 1986). For
subjects 1-3, the bone-fixed triads and the skin-based markers were tracked
simultaneously for each trial. Four strides from each horse were selected in order to

closely match forward velocity between subjects.

Reference Kinematics. The 3D kinematics of the tibia and third metatarsus, as
calculated by the motion of the bone-fixed triads, were defined as the “true” reference
kinematics of the bones. The orientation matrices and displacement vectors for the

reference kinematics were calculated by relating the locations of the triads in their

209

corresponding BCS with their global locations during each frame of motion data using an
SVD method (Sliderkvist and Wedin, 1993). The 3D attitudes of the segments and tarsal

joint were expressed in terms of a spatial attitude vector (Woltring, 1994). The attitude

vector components are referred to as 0x, By and 02 and correspond generally to the

abduction/adduction, flexion/extension, and internal/external rotation angles respectively.
The segment attitude vectors were expressed with respect to the global coordinate
system. The joint angular motion of the tarsal joint complex was calculated as the motion
of the third metatarsus with respect to the tibia (Lanovaz et al., 2002). Translations of the
tarsal joint were calculated as the motion of the origin of the third metatarsal BCS with
respect to the origin of the tibial BCS, expressed in terms of the tibial BCS. The three
components of translation along the x, y and z-axes of the tibia are referred to as DX, DY

and DZ respectively.

Quantification of Skin Displacement. The 3D movement of the skin-based
markers in the global coordinate system were collected for the motion data. For subjects
l-3, the data from each frame were transformed using the orientation matrices and
displacement vectors calculated from the bone-fixed triads and expressed in terms of their
corresponding BCS. The skin-based data from each subject were normalized to a
percentage of the segment reference length calculated from the standing pose.

The data from the skin-based markers were parameterized by representing the
data from each coordinate of each marker as a truncated Fourier series (Capozzo et al.,

1975; van Weeren et al., 1992) given by

210

n n
27ch . 27d“

C(t)=p0+ 2 pk cos[—)+ E qk s1n(—] (l)
k=l 100 k=l 100

where C (t) is an x, y or z coordinate of a marker, t is expressed as a percentage of cycle

time, p0 is the constant offset, pk and qk are Fourier series parameters and n is the

number of harmonics used to describe the data.

In order to determine n for each coordinate, the pooled data from all 12 trials from
subjects 1-3 were fit to Eq. (1) using a standard least-squares method, with n ranging
from I to 30. A GCV criterion method (W oltring, 1990) was used to find the statistically
optimal n for each coordinate of each marker and the Fourier parameters p0, pk, and qk
were then calculated. Model parameters were also calculated for each subject, using the
optimum number of harmonics derived from the fit to the complete pool of trials (van
Weeren et al., 1992).

In order to minimize the effects of any marker placement variation between
subjects, the standing pose value of the coordinate (expressed as a percentage of segment

length) was subtracted from the skin displacement prior to fitting the data to the Fourier

series. The pa value in the model then became a mean offset from the standing pose.

The standing pose value for the subject was added to the results from the Fourier model
to obtain a final value.

As an estimate of the goodness of fit for the pooled data Fourier parameters, the
RMS fit error between the model and the actual data was calculated for each coordinate
of each marker. The RMS and peak to peak amplitudes of the skin displacement for each

coordinate of each marker were calculated from the fitted Fourier parameters.

211

For comparison, skin displacements for the x-axis and z-axis of TlB-A, TIB-B, MET-B
and MET-A were calculated based on published 2D models (van Weeren et al., 1992)
using mean kinematics from the 12 trials from subjects l-3 of the current study. The axes
from van Weeren et al. (1992) were transformed to correspond with the conventions used

in this investigation.

Three dimensional skin correction. The Fourier models of the skin
displacement were used to estimate the actual three dimensional orientation and position
of the tibia and third metatarsus from the skin-based marker data alone. It is assumed
that the reference standing pose does not contain any skin displacement artifacts. The
marker positions in the reference pose, expressed in the local BCS, are used as the base

position for the SVD method (Séderkvist and Wedin, 1993). It can be shown that, using

the SVD method, the orientation matrix R0 and the displacement vector do describing the

three-dimensional position of a segment in global coordinates can be found by solving

”1
min ZIIROxi +d0 — y,-||2 (2)
REQ,d i=1

where m is the number of markers on the segment, x,- are the three-dimensional reference

pose locations of the markers, expressed in the local BCS, and y,- are the three-

dimensional locations of the markers at some point in time during the movement,

expressed in the global coordinate system.

212

When skin displacement artifacts are present, the solution of R0 and do from Eq.

(2) contains errors since the skin artifacts are incorrectly interpreted as changes in
position of the segment. Although these errors are minimized in the least-squares sense,
they can still result in large discrepancies when calculating kinematic variables

(Reinschmidt et al., l997a). However, if the nature of the skin artifact is known in

relation to the local BCS, the values of x,- can be adjusted for the skin displacement and

then used in Eq. (2) to recover the true orientation and position of the segment.

The Fourier parameters obtained from the skin displacement quantification
procedures described above were used to generate the positions of the markers in their
local BCS for each frame of data. The SVD algorithm was used to calculate orientation
matrices and position vectors for each segment using the generated local BCS marker

positions containing the skin displacement as x,- and the measured global coordinate skin

marker locations as yi in Eq. (2).

The orientation matrices derived from the corrected skin data for the tibia and
third metatarsal segments were expressed as spatial attitude vectors with respect to the
global coordinate system. The three-dimensional kinematics of the tarsal joint were also
calculated from the corrected skin data and expressed in the same manner as those
calculated from the bone-fixed kinematics. The kinematics of the tibia and third
metatarsal segment were calculated fi'om the uncorrected skin-based markers. Using all
of the trials from subjects l-3, the RMS error was calculated between the bone-fixed
kinematic data and both the uncorrected and corrected skin-based data for all six degrees

of freedom.

213

Application. The data from subjects 4 and 5 were used to test the general
applicability of the skin correction model. Corrected tibial and third metatarsal
kinematics calculated from the skin-based marker data of subjects 4 and 5 were compared

to the mean bone-fixed kinematics from subjects 1-3.

RESULTS

Descriptive data from the subjects and the kinematic trials are given in Table A-1.

The reconstructions of the skin displacements from the Fourier model are shown
in Figures A-22 and A-3. The skin displacement based on the pooled data and on the
data for each of the three subjects are given on the same graph in order to visualize any
variation between horses. The data from the model are plotted without adding the
standing pose value so that they can be compared at similar scales. The mean values for
the standing pose locations of the markers are given in Table A-2. The skin displacement
patterns for the x-axis and z-axis of the tibial markers showed similar patterns with higher
amplitudes at the proximal locations. The y-axis of the tibial markers showed little
movement except for TIB-l and TIB-2, which were located furthest from the reference
line of the segment. On the third metatarsus, the displacement patterns for the x-axis and
z-axis of the markers had similar amplitudes and patterns during the swing phase, but
during the stance phase the displacement at the proximal end of the segment was in the
opposite direction to that of the distal end. The y-axis displacements of the third
metatarsal segment markers had small amplitudes and higher variability between sites
and subjects. Overall, the tibial markers underwent much larger displacements than those

on the third metatarsus.

214

The optimal number of harmonics, the po value, the RMS fit error of the pooled

model to the actual skin displacement, and the RMS and peak to peak amplitudes of the
skin displacement models are given in Table 3. The RMS amplitude of the skin
displacement for the tibial markers ranged from 0.1% to 7.1% of the segment length, with
most ranging from 2 to 5 % of segment length. This corresponds to approximate
displacements of 7 to 17 mm. The largest peak to peak amplitude of skin displacement in
the tibial markers was for the z-axis of marker TIB—A which showed a range of motion of
15.9% of segment length or approximately 55 mm. The tibial p0 values averaged 2 to 3%
of segment length (approximately 8 mm). For the third metatarsal segment, the RMS
skin displacement amplitudes ranged from 0.4 % to 2.6% with most values in the 1.5 to 2
% range corresponding to approximate values of 4 to 5 mm. The largest peak to peak

amplitude of skin displacement in the third metatarsal markers was 5.7% of the segment

length (approximately 14 mm), seen in the z-axis of marker MET-4. The pa values for

the third metatarsus were relatively low.

The average number of harmonics required to model the skin displacements was
between 7 and 8 for the tibia and between 4 and 5 for the third metatarsus. The pooled
trial Fourier model showed a good fit to the actual skin data, with most values for the
RMS fit error being well below the peak to peak amplitude of the displacement itself
(Table A-3).

The skin displacements for the x-axis and z-axis of TIB-A, TIB-B, MET-B and

MET-A calculated from 2D models given by van Weeren (van Weeren, et al (1992) are

215

plotted against the current data (Figure A-4). In general, the 2D study tended to estimate
a larger range of skin displacement than the current study.
The segmental kinematics were calculated with respect to the global coordinate

system. The average RMS difference between the bone-fixed segment angles and the

corrected skin segment angles was 1.7 ° (Table A-4), with the best fit in the 6y angle of

the third metatarsus (0.8 0) and the largest RMS difference in the third metatarsus (92

angle (3.4 °). The average RMS difference in segment displacements was 2.9 mm.
Angular kinematics calculated from the bone-fixed data are compared to those calculated
from the skin marker data corrected using the Fourier displacement models (Figure A-5).
The mean trace of all 12 trials is presented for each data set.

The mean kinematics of the tarsal joint from bone-fixed data, corrected skin
marker data and uncorrected skin marker data are shown in Figure A-6. There is good
agreement between the bone-fixed data and the corrected skin data, and there is a

significant improvement between corrected and uncorrected skin data. The uncorrected

skin data underestimates the 6x range of motion and overestimates the DY and DZ ranges

of motion. Some of the largest errors in the uncorrected skin data occur in the 62 angle,

where the average range of motion is 4 times larger than the bone-fixed or corrected skin
values.

Kinematics derived from uncorrected and corrected skin marker data in subjects 4
and 5 are compared with the bone-fixed data from subjects 1-3 in Figure A-7. The

corrected skin kinematics show some offsets relative to the bone-fixed data, but there is

216

good shape agreement for 6x, 6y, DX and DZ. The 62 angle and the DY translation have

an overall poor shape agreement. To account for the inter-individual offsets, the data
were also expressed relative to their values at the start of stride (Figure A-7). The RMS
differences between the corrected kinematic data from subjects 4 and 5 and the mean
bone-fixed data are given in Table A-6 and are expressed both in terms of absolute data

and data referenced to the start of the stride.

DISCUSSION

The periodic nature of the skin displacement lends itself to being modeled with a
truncated Fourier series. The general displacement patterns were reproducible within and
between subjects and this was reflected in the low fit errors.

The only previous equine skin displacement investigation (van Weeren et al.
1992) developed regression equations for 4 of the sites looked at in this study. Despite
differences between that study and the current one in terms of the nature of the data (i.e.
2D vs 3D) and differences in subject breeds and sizes, the data are quite comparable.
Although van Weeren (van Weeren, et al., 1992) gave models for the x-axis
displacements of both ends of the third metatarsus, their original data were within the
systematic error range of their study and they deemed the models not useful for
correction. This may explain the differences in those axes compared with the results
presented here. The other difference between the two studies is an offset seen in the x-

axis displacement data of the proximal tibia. Van Weeren (van Weeren, et al., 1992)

assumed that the skin oscillates around the original marker position and therefore the po

217

value of their Fourier model was set to zero before fitting the data. The current study
found that this assumption is probably valid for the third metatarsus and distal tibia, but is

not true for the proximal tibia.

In all cases, the 62 value showed the highest RMS error when skin marker data

were compared to bone-fixed data. The skin-based markers were distributed as widely as
possible on each segment, given the requirement of visibility from the lateral aspect and
the physical limitations of segment size. However, on the third metatarsus, the skin
markers tended to be clustered around the segmental z-axes. This type of marker

configuration results in an increased sensitivity to marker location error (S6derkvist and

Wedin, 1993; Cappozzo et al., 1997), especially in the calculation of (92 which is

generally the most sensitive to errors (Cappozzo et al., 1996). In the case of the 611 value

for the tarsal joint kinematics, the RMS error of the corrected skin data is as high as the
average range of motion. This indicates that a generalized skin correction model based

on a pool of horses for the current marker configuration still contains too much error to

accurately reproduce 62 values for the tarsal joint. This was apparent when the skin

correction model was applied to subjects 4 and 5.

The ex value calculated from the corrected skin markers, however, shows both a

drastic improvement over the uncorrected skin data and a low RMS error when compared

with the bone-fixed data. The application of the model on the last two subjects also

showed a good shape agreement. The 6y values showed a low RMS error for the

218

corrected skin markers and the correction model gave good results when applied to the

independent subjects. The RMS error for the 6y values from the uncorrected skin

markers was also low. This is an indication of the general robustness of the SVD
algorithm and shows that, with a suitable redundant marker set, the flexion/extension
angles of the tarsal joint can be calculated fairly accurately without correction for skin
artifacts.

The structure of the equine tarsal joint, with the tibia moving along the grooves of
a circular talus roughly 80 mm in diameter (Badoux, 1987), dictates that most joint

rotations have corresponding coupled translations. The translation errors in the

uncorrected skin-based data are largely due to the errors in the estimation of the 6,, and 62

values. The skin correction model significantly reduced the RMS error in all translations,
with the greatest RMS error improvement in the DX value and the best shape
improvement in the DY translation. When the correction model was applied to the
independent subjects, the DZ translation matched well with the bone-fixed data. There
was an offset in the DX and DY curves but a good shape agreement in the DX curve.
Even though the DY translation estimation is improved with the correction model, it still

had a relatively poor shape agreement, especially during swing which can be attributed to
errors in estimating the (92 rotation and possibly to an overestimation of the 6x rotation at
mid-swing (Figure A-7).

In general, inter-subject differences were reduced when the data were expressed

relative to a reference pose. For example, when the data were expressed with respect to

the start of the stride, the RMS error for the DX translation was reduced by an average of

219

70% in the independent subjects. The unguligrade stance of horses raises the phalanges
and metatarsal bones off the ground and allows a range of limb segment lengths and
angulations, even within uniform populations. The use of the start of the stride as a
normalization point to account for between subject conformational effects is one
approach and has been shown to be fairly consistent for the equine tarsal joint (Lanovaz
et al., 2002).

The use of the SVD algorithm to adjust for skin displacement artifacts is straight
forward and easy to implement. The 3D motion of the tibia and third metatarsus was
reproduced well and the relative joint motion of the tarsus calculated using the corrected

skin-based markers showed a marked improvement over the uncorrected skin markers.

The 6x, 6y, DX and DZ estimations are suitable for use in general analysis. It is

suggested that for between-subject comparisons, the skin corrected kinematics should be

expressed relative to a reference orientation such as the start of the stride.

ACKNOWLEDGEMENTS

This work was supported by the McPhail Endowment, College of Veterinary
Medicine, Michigan State University. The authors would like to thank Drs. John Stick
and Jennifer Brown for performing the surgical procedures and Carissa Wickens and
Sarah Fox for assistance during data collection. The authors would also like to thank Dr.
Ton van den Bogert for providing additional information regarding the 2D correction

procedures.

220

REFERENCES

Alexander, E.J., Andriacchi, T.P., 2001. Correcting for deformation in skin-based marker
systems. Journal of Biomechanics 34, 355-361.

Andriacchi, T.P., Alexander, E.J., Toney, M.K., Dyrby, C.O., Sum, J., 1998. A point
cluster method for in vivo motion analysis: applied to a study of knee kinematics.
Journal of Biomechanical Engineering 120, 743-749.

Badoux, D.M., 1987. Some biomechanical aspects of the structure of the equine tarsus.
Anatomischer Anzeiger 164, 53-61.

Cappozzo, A., Leo, T., Pedotti, A., 1975. A general computing method for the analysis
of human locomotion. Journal of Biomechanics 8, 307-320.

Cappozzo, A., Catani, F., Leardini, A., Benedetti, M.G., Della Croce, U., 1996. Position
and orientation in space of bones during movement: experimental artifacts. Clinical
Biomechanics 11, 90-100.

Cappozzo, A., Cappello, A., Della Croce, U., Pensalfini, F., 1997. Surface-marker
cluster design criteria for 3-D bone movement reconstruction. IEEE Transactions on
Biomedical Engineering 44, 1 165-1174.

Back, W., Schamhardt, H.C., Savelberg, H.H.C.M., van den Bogert, A.J., Bruin, G.,
Hartman, W., Bameveld, A., 1995. How the horse moves: 2. Significance of graphical
representations of equine hind limb kinematics. Equine Veterinary Journal 27, 39-45.

Gabel, AA, 1983. Prevention, diagnosis and treatment of inflammation of the distal
hock. In Proceedings of the 28th Annual Convention of the American Association of
Equine Practitioners. Manhattan, NY, pp 287-298.

Gough, M., Munroe, (3., 1998. Decision making in the diagnosis and management of
bone spavin in horses. In Practice 20, 252-259.

221

Hodson E.F., Clayton, H.M., Ianovaz, J .L., 2001. The hind limb in walking horses: 1.
Kinematics and ground reaction forces. Equine Veterinary Journal 33, 38-43.

Holmstréim, M., Fredricson, 1., Drevemo, S., 1994. Biokinematic analysis of the Swedish
Warmblood riding horse at trot. Equine Veterinary Journal 26, 23 5-240.

Holmstréim, M., Fredricson, I., Drevemo, S., 1995. Biokinematic effects of collection on
the trotting gaits in the elite dressage horse. Equine Veterinary Journal 27, 281-287.

Kobluk, C.N., Schnurr, D., Horney, F.D., Hearn, T.C., Summer-Smith, G., Willoughby,
RA, DeKleer, VS, 1989. Use of high speed cinematography and computer generated

gait diagrams for the study of equine hind limb kinematics. Equine Veterinary Journal 21,
48-58.

Lafortune, M.A., Cavanagh, P.R., Sommer III, H.J., Kalenak, A. 1992. Three-
dimensional kinematics of the human knee during walking. Journal of Biomechanics 25,
347-357.

Lanovaz, J .L., Khumsap, S., Clayton, H.M., Stick, J .A., Brown, J ., 2002. Three
dimensional kinematics of the tarsal joint at the trot. Equine Veterinary Journal
Supplement 34, 308-313.

Reinschmidt, C., van den Bogert, A.J., Nigg, B.M., Lundberg, A., Murphy, N., l997a.
Effect of skin movement on the analysis of skeletal knee joint motion during running.
Journal of Biomechanics 30, 729-732.

Reinschmidt, C., van den Bogert, A.J., Murphy, N., Lundberg, A., Nigg, B.M., l997b.
Tibiocalcaneal motion during running — measured with external and bone markers.
Clinical Biomechanics 12, 8-16.

Schamhardt, H.C., Hartman, W. , De Lange, A., 1984. Kinematics of the equine tarsus.
Abstracts of the XV Congress of the European Association of Veterinary Anatomists,
Utrecht, The Netherlands, 178-179.

S6derkvist, I., Wedin, P., 1993. Determining the movements of the skeleton using well-
configured markers. Journal of Biomechanics 26, 1473-1477.

222

van Weeren, P.R., van den Bogert, A.J., Bameveld, A., 1992. Correction models for skin
displacement in equine kinematic gait analysis. Journal of Equine Veterinary Science 12,
178-192.

Winter, D., Bruns, E., Glodek, P., Hertsch, B., 1996. Genetic disposition of bone diseases
in sport horses. Zuchtungskunde 68, 92-108.

Woltring, H.J., Huiskes, R., De Lange, A., Veldpaus, F., 1985. Finite centroid and
helical axis estimation from noisy landmark measurements in the study of human joint
kinematics. Journal of Biomechanics 18, 379-389.

Woltring, H.J., 1986. A FORTRAN package for generalized, cross-validatory spline
smoothing and differentiation. Advances in Engineering Software 8 (2), 104-113.

Woltring, H.J., 1990. Model and measurement error influences in data processing. In:
Biomechanics of human movement: applications in rehabilitation, sports and ergonomics.
Bertec Corporation, Worthington, Ohio, pp. 203-237.

Woltring, H.J., 1994. 3D attitude representation of human joints: a standardization
proposal. Journal of Biomechanics 27, 1399-1414.

223

 

 

3.3 3.2:- 3.3 3.3- 3.3 35 $52
3.3 3.3- 3.3 3.3 3.3 3.: 3.52
3.3 3.:- 3.3 3.3- 3.3 3.3 352
3.3 3.2- 3.3 3.3- 3.3 3....- 3.32
3.3 3.3- 3.3 38. 3.3 3.2 7.52
3.3 33 3.3 3.3- 3.3 35 952
3.3 35 3.: 3.2- 3.3 3.o m-mfi
3.3 3.8 3.3 3.3- 3.3 3.3 rm:
3.3 3.3 3.: 3.8- 3.3 3.3 m-mfi
3.3 3.3 3.3 3.3- 3.3 3.3- 2:.
3.3 3.3 3.3 3.8- 3.3 3.: 7m:-
3.3 3.03 3.3 3.:- 3.3 3.o arm:

3 > x 3...:

 

Ema-2 Eon-mom mo own-5953 E 53m 8s Ea £58.? BEE-coo
28: :82 2: 8 8838- 55, 38835 cs «:5 .m; $8.33 30 883 $5388 05 E83 macs-82 EVE-E mo amv E82 ~-< «Ewa-

 

 

 

3. a. 3 3. £- 23 8- 33.3 EN :3 33 a: a? 33 m2 383 3: 63:82
3.3% 3: :3 v2 3- 3m 3: m
3.38 8- EN NR 3..- wt“ 31 a.

3. 3m 33 33 SN 8- mm :2 m
3. N3. NE 83 98 am am :1 N
3. 38m 33 M3...- an 3.- 3.. 31 _
own-c853 WEV EH5 AEEV Ewan-A AEEV AEV mac-EB Sufism
3.3 3.2
853m cEC- cur-Hm 3620> mama-802 Em Ewan-A «EC- 3 Emma:

 

28b 833E053 05 E83 8% 038583 58E 05 98 38.33 05 E83 9% 95%-omen —.< 03:.

4
2
2

 

 

 

 

 

3.3m 3cm . 4 3R6 3.6. 4- 6 3. 4 .4 3.3.6 3.3.6 3.66 m 3&6 3.34 3. 4 .4 3&6- 3 474-4442
3.56 3.66 3.6. 4 3&.&. 3 3.5. 4 3&6 3& . 4 3R6 4 3.64 3.3.6 3.66 3.56 3 3--4.m44>4
3.5. 4 3&6 3.6. 4 3.66- 6 3.5.6 3&6 3.6.6 3.6.6 3 3. 4 .3 3&4 3.6.6 3.64 mm m-.4-m4>4
3.6.3 3.6. 4 3&. 4 3&- 4 - m 3. 4 .4 3.3.6 3.64 3&. 4 3 3.3.N 3&6 3.34 3.6.6 m N-.4-m44>4
3.6.m 3.6. 4 3.6.6 3. 4 . 4 - 5 3R. 4 3R6 3. 4 .& 3&6- 3 3.66 3& .6 3.m . 4 3.66. N 4--4-m44>4
3.3m 3&4 3.56 3.66- n 3&4 3.5.6 3.6. 4 3. 4 6- m 3.36 3&4 3. 4 .4 3&6- 6 m4--4-m44>4
3.6.44 3&4 3.m .6 3&6- N4 3&6 3. 4 .6 3R6 3&6 N 3.6.3 3&4 3.6.6 3&6- 64 m-m44-4-

3. 4 .6 3.66 3.56 3. 4 .6 w 3. 4 .N 3&6 3R6 3.m. 4- 6 3% .3 3.6.& 3.66 3.66- N 3..m44-4-
3. 4 .4 4 3.53 3.64 3.3. 4- 3 3&.& 3&6 3.6.6 3&6 m 3&6 3.56 3&6 3.6.3- 4 4 m-m44.4-
3.664 3.4 .n 3.m. 4 3.66- 3 3.66 3&.& 3.34 3.64 & 3&6 3&3 3&4 3&6- 6 &-m44-4-
3&64 3. 4 .3 3.64 3.3.6- 6 3.3.5 3&6 3&4 3&3- b 3&3 3.6. 4 3.6. 4 3. 4 .m- m Tm:-
3.6.m4 3. 4 .3. 3.64 3.66- 6 3&6 3.34 3.6.6 3&.&- 6 3.664 3&6 3. 4 .m 3.3.5.. 64 <-m44.4-

9E. 6E. Etc cm. AS 448.. 6E3 Etc cm Q6 9E. 6E... .85 em 45 484342
V4.86 m4>4~4 m4>4~4 84.5 .486 m5 m4>4y4 84.5 .486 m4>4y4 m5 84:0

N > X

 

 

225

arm-64 EcEwom mam-880E 4.44444

8 «E44 244 60 E6824 8 8.48 8:4? 4:886 443. .4».on 24. wEm: 4.834348 Eon-02.4634. Exm 244 .40 84.34495. V486 9 c486 4.5.
m4>4~4 86 ch. 6E3 v4.8.4 4.5... 6E3 m4>4~4 24% .404.oE 24. 4.5. EoEcom4-4m44. 4332. 846 52586 858.4644. 244 m4 coho m4>4y4 24.4- 384884
824 6.244.488 844 E06 486-40 58E 844 m4 634...» cm. 244 4.5. 404.08 42.503 244 E moEoE-a: .40 86:5: REES 26 84844.5 $480 24.4-

3843.4: Dom-443m 5.4m 4408 8.4 83:44.88 N 4.8. > .X 244 404 358834-424. Exm 48464.0:- 244 E06 mozmufim 63:44.5ch m-< «4.4-«.4.

mouanEx E42. 43.4.8 98 m 4.48 3 $8.33 8.4 Sam

6

E049? 08:44.88 43046 244 8 486mm: 543 23 moumEoEx REQEwom RES—30E 4:45 4.5 «54.4...

 

 

 

 

64 4 0m m.m 3.3 6.6 64 65.8 4.88950 m> meégm 953434 ”m 48.3%
m .3 Om 6.5 m6 6.m m . 4 nExm 6808.80 m> 4.86-334 033434 ”3 48.35
34 4 M: 6.m& w64 m.m 3m nExm 4.88980 m> 4.88-2484 328443. ”m 48.34%
Nam M64 3.2 &.m4 NM 6.4 nExm 4.80880 m> 4.846-384 3:48.43. ”3 48.35
66 4.6 n64 N64 m.m m.m E4.4m 4.083.845 m> 45.34-2434 E404. 438,4.
6.m 6.3 &.3 6.3 m4 4.& E446 4582.80 m> coxmbgm E404. 483.4.
n.& 4.3 m.m 3.m &6 m4 «Exm coBoEoU m> 4.36-2844 mSESSoE 6.44.4.4.
EN n4 6.& Wm 64 6.6 «5.4m 4.284.400 m> 4.344-845 334.4.
NE E .8 N3 Am. .3
AEEV mEoEoomama $04.6 mo4wc<

 

 

 

84495454 cofifiéEm 4.5 4.346-286 59524 $808.43. mEM 3.3. 04.45.

226

Figure A-l Lateral view of the tibia (left) and third metatarsus (right) of a right limb
showing the locations of the skin surface markers relative to the underlying bones.

 

 

 

 

 

TIB-B

TIB-A: distal site of attachment of the lateral collateral femoro-tibial ligament, TIB—l:
1/6 of the segment length distal to TIB-A and over the tibial crest, TIB-Z: 1/6 of the
segment length proximally from TIB-3 and approximately 1/6 of the segment length
caudal to the TIB-A/TIB-B line, TIE-3: mid-way along the line between TIB-A and TIB-
B, TIE-4: 1/6 of the segment length distally from "FIB-3 and about 1/6 of the segment
length cranial to the line between TIB-A and TIB-B, TIB-B: over the lateral malleolus of
the tibia, MET-B: dorsal edge of the head of the fourth metatarsal bone, MET-1: dorsal
aspect of the bone mid-way between MET-B and MET-2, MET-2: 1/4 of the segment
length distally from MET-B, MET-3: 1/4 of the segment length proximally from MET-A,
MET-4: dorsal aspect of the bone mid-way between MET-A and MET-3, MET-A:
metatarsal attachment of the lateral collateral ligament of the metatarsophalangeal joint.
Not shown are the markers TIB—C and MET-C which are used during the standing
reference pose and are placed on the medial side of the limb, opposite to TIB-B and
MET-B respectively.

227

Figure A-2 Skin displacements for each coordinate of each tibial marker. The data are
generated from the truncated Fourier series models. The dashed and dotted lines are from
models fit to the three individual horses, while the thick solid line is from the model
based on the pool of 12 trials from subjects 1-3. The columns represent the x, y and z
coordinates respectively. Each row represents a marker with the top row being TIB-A
and bottom row being TIB-B.

Skin Displacement (percent of segment length)

L. LL-
60:60: 0.6

-15

31:60-

 

31:60:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TIB-B - l

 

0 50 100 0

 

 

 

 

 

100 0 50 100

Percent of Stride

228

 

Figure A-3 Skin displacements for each coordinate of each third metatarsal marker. The
data are generated from the truncated Fourier series models. The dashed and dotted lines
are from models fit to the three individual horses, while the thick solid line is from the
model based on the pool of 12 trials from subjects 1-3. The columns represent the x, y
and z coordinates respectively. Each row represents a marker with the top row being
MET-B and bottom row being MET-A.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Skin Displacement (percent of segment length)

 

 

 

 

 

 

 

-5 L MET-A 1"" J

 

 

 

0 50 100 0 50 100 0 5‘0 100
Percent of Stride

229

h -.~ 6.3 ..

Figure A-4 Comparison of skin displacement models for the proximal and distal ends of
the tibia and third metatarsus. The solid lines are from the models for the TIB-A, TIB-B,
MET-A and MET-B markers from the current study. The open dotted lines are from the
models given in van Weeren (van Weeren, et al., 1992) for the corresponding locations.
The first column is the displacement data for the x-axis and the second column is the
displacement data for the z-axis. The data are given as a percentage of segment length.

X Displacement Z Displacement

 

 

 

 

GUI

 

A
v

i—n
O

   

Prpximal Tibia
z'n

 

 

 

 

 

 

 

 

Distal Tibia

 

 

 

 

 

 

 

 

Proxnmal
Metatarsus

 

 

 

 

 

 

 

 

    
 

 

 

I

Distal
Metata rsus

 

 

 

 

 

 

0 5‘0 100 0 5‘0 100
Percent of Stride

230

Figure A-5 Angular kinematics of the tibia and third metatarsus with respect to the
global coordinate system. The solid lines are from the bone-fixed kinematics and the
dotted lines are from corrected skin kinematics. Data are a mean of the 12 trials from
subjects 1-3, normalized to percent of stride.

Tibia 3rd Metatarsus

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 5‘0 100 0 50 100
Percent of Stride

231

Figure A-6 Kinematics of the tarsal joint (motion of the third metatarsus relative to the
tibia). The right hand colurrm is the angular data (in degrees) and the left hand column is
the displacements (in mm). The solid lines represents bone-fixed kinematics, the filled
dotted lines represents corrected skin kinematics, and the open dotted lines are kinematics
from uncorrected skin markers. Data are a mean of the 12 trials from subjects 1-3.

Angles (deg) Displacements (mm)

 

fl,

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Percent of Stride

232

Figure A-7 Application of correction models to kinematics of the tarsal joint (motion of
the third metatarsus relative to the tibia) in absolute values and referenced to the start of
the stride. The right hand columns are the angular data (in degrees) and the left hand
columns are the displacements (in mm). The solid lines represents bone-fixed kinematics
from subjects 1-3, the filled dotted lines represents corrected skin kinematics from

subject 4, and the open dotted lines are kinematics from corrected skin markers of subject
5.

Angles (deg) Displacements (mm)
Absolute Relative Absolute Rejitive

DX 0x A

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Percent of Stride

233

APPENDIX B

BIOMECHANICAL VARIABLES IN SOUND AND LAME CONDITIONS

Table B-1 Ground reaction force (GRF) variables in the hind limbs. Values from sound
condition are averaged from both hind limbs. Lame RH is the limb in which synovitis
was induced. Compensating LH is the contralateral hind limb. Values are mean and (SD).

3 P<0.05. b 0.05<P<O. 1.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variables Sound Lame RH CompLeIiiisating
Vertical force Max (N/kg) 9.052 8.780 9.183
(0.402) (0.558) a (0.455) b
Vertical impulse (N s/kg) 1.453 1.404 1.459
(0.077) (0.089) b (0.087)
Longitudinal force Maxl (N/kg) -0.l34 -0.128 -0.l90
(0.200) (0.144) (0.262)
Longitudinal force Max2 (N/kg) -0.638 -0.511 -0.645
(0.126) (0.1 15) (0.126)
Longitudinal force Max3 (N/kg) 1.037 1.001 1.086
(0.114) (0.094) b (0.032)
Longitudinal force Minl (N/kg) -l.621 -l.481 -l.580
(0.237) (0.320) (0.542)
Longitudinal force Min2 (N/kg) -0.773 -0.640 -0.783
(0.108) (0.149) (0.1 17)
Longitudinal force Min3 (N/kg) -0.717 -0.566 -0.722
(0.068) (0.1 12) (0.040)
Longitudinal negative impulse (N s/kg) -0.073 -0.069 -0.071
(0.008) (0.007) (0.008) a
Longitudinal positive impulse (N s/kg) 0.087 0.086 0.091
(0.016) (0.016) (0.002)

 

 

 

234

Table B-2 Joint angle (degrees) during stance in the hind limbs. Values from sound
condition are averaged from both hind limbs. Lame RH is the limb in which synovitis
was induced. Compensating LH is the contralateral hind limb. Values are mean and (SD).

3 P<0.05. b 0.05<P<O.l.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Sound Lame RH Complefisating
COXOfemora‘ Max1 109.78 (7.46) 107.22 (5.19) b 111.35 (9.57)
Max2 125.20 (6.64) 123.95 (4.39) 127.89 (8.36)
Min1 101.12 (7.18) 99.06 (4.34) 102.73 (8.13)
Min2 114.84 (7.78) 112.77 (5.33) 116.84 (9.01)
Range 26.27 (1.23) 25.64 (1.23) 26.38 (0.91)
Femorotibial Max1 113.24 (1.90) 110.96 (3.03) 114.04 (1.78)
Max2 109.97 (2.19) 107.46 (1.96) 111.13 (2.29)
Minl 108.46 (2.00) 106.72 (3.17) 108.94 (1.19)
Min2 107.36 (2.08) 104.87 (2.19) 108.59 (1.64)
Range 10.97 (2.67) 10.69 (2.74) 10.34 (2.85)
Tarsus Maxl 114.79 (2.87) 111.07 (3.89) 116.08 (3.20)
Max2 106.45 (3.74) 104. 60 (3.74) 107.89 (2.87)
Minl 105.44 (4.20) 103.53 (3.34) 107.86 (4.02)
Min2 103.33 (3.41) 101.80 (2.94) 104.98 (4.83)
Range 12.12 (0.96) 10.24 (0.99) a 11.72 (1.86)
Metatarsophalangeal Max 237.34 (11.49) 233.25 (9.87) 240.65 (11.04)
Range 47.79 (9.16) 45.86 (8.56) 48.98 (10.41)
Distal Max 206.92 (6.45) 206.64 (5.30) 202.53 (5.39)
mterphalangeal Min 151.55 (10.20) 157.25 (8.86) 146.77 (9.21)
Range 55.66 (6.42) 54.60 (5.41) a 56.27 (7.97)

 

235

 

Table B-3 Net joint moment peaks (Nm/kg) during stance in the hind limbs. Values from
sound condition are averaged fi'om both hind limbs. Lame RH is the limb in which
synovitis was induced. Compensating LH is the contralateral hind limb. Values are mean

and (so). 3 P<0.05. b 0.05<1><0.1.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Sound Lame RH compfgsa‘ing

Coxofemoral Maxl 1.101 (0.129) 1.055 (0.078) 1.044 (0.108)

Max2 0.575 (0.126) 0.542 (0.134) 0.644 (0.059)

Max3 0.557 (0.185) 0.353 (0.062) 0.689 (0.160)

Min1 -0372 (0.248) -0219 (0.396) -0313 (0.563)

Min2 0.367 (0.096) 0.391 (0.167) 0.378 (0.087)

Femorotibial Maxl 0.365 (0.100) 0.293 (0.186) 0.329 (0.330)

Max2 0.135 (0.055) 0.176 (0.032) 0.105 (0.058)

Minl -0.59 (0.131) -0529 (0.125) -0553 (0.138)

Min2 -0537 (0.144) -O.416 (0.078) -O.676 (0.082)

Max1 0.422 (0.086) 0.410 (0.059) 0.413 (0.089)

Tarsus Max2 1.406 (0.154) 1.293 (0.128) 1.471 (0.204)

Min1 -0132 (0.052) -0.065 (0.089) -0114 (0.163)

Min2 0.325 (0.076) 0.351 (0.107) 0.347 (0.139)

Metatarsophalangeal Min -1.149 (0.140) -l.064 (0.167)a -1.199 (1.143)

Distal Min -0304 (0.058) -0214 (0.046?3 -0321 (0.072)
1nterphalangeal

 

236

 

Table B-4 Net joint power peaks (W/kg) during stance in the hind limbs. Values from
sound condition are averaged from both hind limbs. Lame RH is the limb in which

synovitis was induced. Compensating LH is the contralateral hind limb. Values are mean
and (so). 3 P<0.05. b 0.05<P<O. 1.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Sound Lame RH CompIeInIsating
Coxofemoral Max1 1.501 (0.419) 1.150 (0.301) 1.698 (0.425)
Max2 1.473 (0.554) 0.917 (0.101) 1.827 (0.497)
Femorotibial Max1 2.488 (0.875) 2.058 (0.768) 2.325 (0.996)
Max2 1.150 (0.493) 0.690 (0.258) 1.461 (0.462)
Minl -0.503 (0.188) -0.415 (0.313) -0.396 (0.434)
Min2 -l.085 (0.384) -0.661 (0.165) -l.402 (0.210)
Tarsus Maxl 0.862 (0.762) 0.812 (0.374) 0.727 (0.815)
Max2 2.627 (0.565) 2.377 (0.348) a 2.648 (0.670)
Min1 -2.219 (0.496) -1.758 (0.316)a -2.119 (0.608)
Min2 -2.159 (1.073) -1.820 (0.681) -2.281 (1.476)
Metatarsophalangeal Max 6.237 (1.284) 5.113 (0.614) 6.566 (1.076)
Min -4.653 (1.316) -4.230 (1.046) -5.095 (l.615)a
Distal Max 0.588 (0.258) 0.379 (0.030) 0.796 (0.441)
interphalangeal b
Min -3.037 (0.625) -2.049 (0.323) -3.l37 (0.746)

 

 

237

marsmm

Table B-5 Net joint energies (J/kg) during stance in the hind limbs. Values fi'om sound
condition are averaged from both hind limbs. Lame RH is the limb in which synovitis
was induced. Compensating LH is the contralateral hind limb. Values are mean and (SD).

a P<0.05. b 0.05<P<0. 1.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Sound Lame RH Complefisatmg

Coxofemoral Maxl 0.063 (0.023) 0.048 (0.030) 0.077 (0.024)

MaxZ 0.073 (0.034) 0.038 (0.014) 0.099 (0.018)

Total +13 0.140 (0.140) 0.090 (0.044) 0.182 (0.032)

Total -13 -0.014 (0.009) -0.026 (0.016) -0.005 (0.003)

Femorotibial Max1 0.037 (0.016) 0.034 (0.017) 0.040 (0.024)

Max2 0.041(0.018) 0.022(0.017)a 0.051(0.015)

Min1 -0004 (0.001) -0004 (0.001) —0.003 (0.001)

Min2 -0031 (0.011) -0020 (0.006) -0042 (0.010)

Total +13 0.079 (0.027) 0.061 (0.022) 0.092 (0.027)

Total -E 0.043 (0.015) 0027 (0.009) -0055 (0.03

Tarsus Maxl 0.024 (0.014) 0.018 (0.009) 0.022 (0.005)

Max2 0.125(0.023) 0.107 (0.018)a 0.125(0.026)

Min1 -0.075(0.016) -0.059(0.011)b -0.074 (0.013)

Min2 -0.076 (0.043) -0.060 (0.026) -0.080 (0.058)

Total +13 0.147 (0.023) 0.126 (0.008)a 0.142 (0.033)

Tota1-E -O.151(0.038) -O.ll9(0.035)a -O.155(0.068)

Metatarsophalangeal Max 0.394 (0.094) 0.311 (0.053) 0.408 (0.069)

Min -0.386 (0.108) 0.359 (0.100) -0422 (0.129)

Total +13 0.394 (0.094) 0.311 (0.053) 0.408 (0.069)

Total -E -O.387 (0.108) -0359 (0.100) -0422 (0.129)

913““ Max 0.038(0.018) 0.020(0007)b 0.052 (0.033)
interphalangeal b

Min -0.222 (0.049) -0.146 (0.021) -0.229 (0.059)

Total +13 0.040 (0.017) 0.022 (0.006) 0.053 (0.033)

Total -E -0223 (0.042 -O.146 (0.021) -0.229 (0.059)

 

 

238

 

Table B-6 Joint angle (degrees) during swing in the hind limbs. Values from sound
condition are averaged from both hind limbs. Lame RH is the limb in which synovitis
was induced. Compensating LH is the contralateral hind limb. Values are mean and (SD).

a P<0.05. b 0.05<P<O. 1.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Sound Lame RH Complegsating
Coxofemoral Min 76.85 (5.43) 74.83 (4.68) 78.73 (6.00)
Range 48.63 (1.41) 48.49 (1.20) 48.49 (2.77)
F emorotibial Max 114.97 (4.03) 114.29 (4.05) 115.07 (3.98)
Min 68.35 (3.09) 68.07 (2.82) 68.38 (3.48)
Range 50.44 (2.63) 49.08 (1.39) 50.87 (2.06)
Tarsus Max 113.38 (4.99) 110.85 (4.96) 114.35 (5.17)
Min 70.77 (3.09) 70.48 (3.16) 72.32 (3.30)
Range 44.67 (1.90) 41.66 (1.47) 44.21 (1.35)
Metatarsophalangeal Max1 160.54 (6.94) 157.89 (1.02) 161.79 (4.74)
MaxZ 185.90 (7.63) 183.48 (4.60) 186.74 (7.20)
Min1 147.70 (6.67) 144.88 (6.59) 149.94 (5.99)
Min2 156.61 (6.76) 154.67 (1.49) 158.51 (7.17)
Range 45.31 (3.38) 44.20 (2.43) 43.90 (5.02)
Distal Max1 170.08 (6.34) 169.07 (4.67) 167.77 (4.99)
interphalangeal Max2 180.41 (8.29) 183.20 (7.36) 181.96 (7.39)
Min1 162.60 (5.69) 160.89 (3.10) 161.57 (6.07)
Min2 155.33 (6.13) 155.70 (3.41) 154.38 (6.42)
Range 45.29 (5.05) 43.64 (6.77) 43.24 (3.17)

 

239

 

Table B-7 Net joint moment peaks (N m/kg) during swing in the hind limbs. Values from
sound condition are averaged from both hind limbs. Lame RH is the limb in which
synovitis was induced. Compensating LH is the contralateral hind limb. Values are mean

and (so). 8 P<0.05. b 0.05<P<O.l.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Sound Lame RH compffisa‘ing
Comfemo‘al Max 0.119 (0.030) 0.138 (0.035) 0.118 (0.016)
Min1 -0431 (0.058) -0445 (0.067) -O.463 (0.064)a
Min2 -0.135 (0.065) -0117 (0.072) -0120 (0.059)Y
Femorotibial Max 0.199 (0.029) 0.180 (0.033) b 0.200 (0.026)
Tarsus Max 0.019 (0.012) 0.018 (0.010) 0.017 (0.013)
Min -0093 (0.016) -0.085 (0.014)a -0093 (0.016)
Metatarsophalangeal Max 0.017 (0.002) 0.016 (0.003) 0.016 (0.003)
Min -0007 (0.003) -0.006 (0.001) -0007 (0.003)
Distal Max 0.002 (0.001) 0.002 (0.001) a 0.002 (0.001)
interphalangeal
Min -0001 (0.000) -0001 (0.000) -0.001 (0.001)

 

240

 

 

Table B-8 Net joint power peaks (W/kg) during swing in the hind limbs. Values from
sound condition are averaged from both hind limbs. Lame RH is the limb in which
synovitis was induced. Compensating LH is the contralateral hind limb. Values are mean

and (SD). a P<0.05. b 0.05<P<0. l.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Sound Lame RH CompIeInIsating

COXOfemora‘ Maxl 1.163 (0.183) 1.275 (0.177) 1.243 (0.253) a
Max2 1.879 (0.670) 1.645 (0.532) 1.539 (0.567)

Min -0.630 (0.157) 0.735 (0.174) 0.616 (0.116)

Femorotibial Max 0.160 (0.138) 0.144 (0.104) 0.141 (0.128)

Min1 0.519 (0.120) 0.428 (0.137) 0.457 (0.129)21

Min2 0.941 (0.322) 0.843 (0.316)b -O.801 (0.132)

Tarsus Max1 0.334 (0.059) 0.296 (0.064) 0.316 (0.068)
Max2 0.318 (0.112) 03-25 “W99 a 0.244 (0.047)

Metatarsophalangeal Maxl 0.019(0012) 0.014(0011)a 0.014(0011)b
Max2 0.015 (0.006) 0.011 (0.006) b 0.012 (0.009)

Min1 0.166 (0.025) 0.142 (0.034)b 0.159 (0.038)

Min2 0.017 (0.010) -0.016 (0.008) 0.014 (0.014)

Min3 0.032 (0.009) 0.027 (0.004) 0.022 (0.003)

1318‘“ Maxl 0.006 (0.003) 0.004 (0.003) a 0.004 (0.002)

interphalangeal

Max2 0.006 (0.002) 0.005 (0.001) a 0.005 (0.005)

Min1 0.023 (0.012) 0.013 (0.006)a 0.015 (0.007)

Min2 0.006 (0.002) 0.004 (0.003) 0.004 (0.002)

 

241

 

Table B-9 Net joint energies (J/kg) during swing in the hind limbs. Values from sound
condition are averaged from both hind limbs. Lame RH is the limb in which synovitis
was induced. Compensating LH is the contralateral hind limb. Values are mean and (SD).

3 P<0.05. b 0.05<1><0.1.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Joint Variables Sound Lame RH CompensatingLH
Coxo- Maxl 0.104 (0.010) 0.106 (0.010) 0.106 (0.017)
femoral Max2 0.122 (0.042) 0.125 (0.052) 0.100 (0.040) b

Min 0.025 (0.010) 0.029 (0.019) 0.022 (0.005)
Total +13 0.229 (0.037) 0.236 (0.047) 0.210 (0.037)
Total -13 0.028 (0.010) 0.035 (0.012) 0.026 (0.008)
Femoro- Minl 0.049 (0.010) 0.044 (0.013) 0.049 (0.013
“13131 Min2 0.079 (0.013) 0.074 (0.014) a -O.75 (0.004)
Total +5 0.012 (0.009) 0.016 (0.012) 0.011 (0.008)
Total -E 0.129 (0.021) 0.118 (0.027) b 0.124 (0.013)
Tarsus Maxl 0.029 (0.008) 0.026 (0.008) b 0.027 (0.009)
Max2 0.022 (0.009) 0.017 (0.009 0.019 (0.004)
Min 0.002 (0.001) 0.001 (0.001) 0.002 (0.001)
Total +13 0.052 (0.013) 0.043 (0.013? 0.046 (0.011)
Total -13 0.004 (0.003) 0.004 (0.003) 0.003 (0.003)
Metatarso' Max1 0.001 (0.000) 0.0004 (000039” 0.0004 (0.0003)
phalangeal a
Max2 0.0004 (0.0002) 0.0003 (0.0002) 0.0004 (0.0003)
Minl 0.009 (0.002) 0.008 (0.002) a 0.009 (0.003)
Min2 0.0004 (0.0002) 0.0004 (0.0003) 0.0004 (0.0003)
Min3 0.002 (0.001) 0.002 (0.001) 0.002 (0.000)
Total +13 0.0011 (0.0005) 0.0008 (0.0005) 0.0008 (0.0004)a
Total -13 0.013 (0.002) 0.012 (0.002) a 0.012 (0.003)
Distal Maxl 0.0002 (0.0001) 0.0001 (0.0001)a 0.0001 (0.0001)
1n CI"
phalangeal Max2 0.0002 (0.0001) 0.0001 (0.0001) 0.0001 (0.0001)b
Min1 0.001(0000) 0.001 (0.000) a 0.001 (0.000)
Min2 0.0004 (0.0001) 0.0003 (0.0001)b 0.0003 (0.0001)
Total +13 0.0004 (0.0002) 0.0003 (0.0002)b 0.0003 (0.0002)
T061144; 0.001 (0.001) 0.001 (0.000) a 0.001 (0.000)

 

242

 

Table B-10 Ground reaction force (GRF) variables in the forelimbs. Values from sound

condition are averaged from both forelimbs. Lame RF and Lame LF are the
compensating limb. Values are mean and (SD). a P<0.05. b 0.05<P<0. l.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variables Sound Lame RF Lame LF
Vertical force Max (N/kg) 10.385 10.195 10.249
(0.492) (0.547) (0.754)
Vertical impulse (NS/kg) 1.929 1.879 1.910
(0.088) (0.105) (0.039) 3
Longitudinal force Maxl (N/kg) -0.206 -0.256 -0.249
(0.091) (0.116) (0.163)
Longitudinal force Max2 (N/kg) -1.010 -1.040 -1 .012
(0.134) (0.201) (0.207)
Longitudinal force Max3 (N/kg) 0.872 0.810 0.845
(0.146) G). 140) (0.207)
Longitudinal force Min1 (N/kg) -1.206 -1.200 -l.328
(0.146) (0.043) (0.123)
Longitudinal force Min2 (N/kg) -l.197 -l.188 -1.150
(0.157) (0.210) (0.185)
Longitudinal force Min3 (N/kg) -1.l62 -1.l75 -l.168
(0.162) (0.212) (0.119)
Longitudinal negative impulse (N s/kg) -0.134 0.134 -0.134
(0.022) (0.025) (0.014)
Longitudinal positive impulse (N s/kg) 0.074 0.067 0.073
(0.019) (0.017) (0.019)

 

243

 

REFERENCES

Anonymous. 1991. Definition and classification of lameness, p. 19. In American
Association of Equine Practitioners Guide for Judging of Equestrian Events, 4 ed.

Audigie, F ., Pourcelot, P., Degueurce, C., Geiger, D. and Denoix, J. M. 2001. Kinematic
analysis of the symmetry of limb movements in lame trotting horses. Equine Veterinary
Journal. 33(supplement): 128-134.

Axelsson, M., Eksell, P., Roneus, B., Brostrom, H., Haggstrom, J. and Carlsten, J. 1998.
Relationship between hind limb lameness and radiographic signs of bone spavin in
Icelandic horses in Sweden. Acta Veterinary Scandinavia. 39:349-357.

Back, W., Bameveld, A., van Weeren, P. R. and van den Bogert, A. J. 1993. Kinematic
gait analysis in equine carpal lameness. Acta Anatomica. 146(2-3):86-89.

Back, W., Bameveld, A., Bruin, G., Schamhardt, H. and Hartman, W. 1994. Kinematic
detection of superior gait quality in young trotting warmbloods. The Veterinary
Quarterly. 16(suppl 2):S91-S96.

Back, W., Schamhardt, H., Savelberg, H., van den Bogert, A., Bruin, G., Hartman, W.
and Bameveld, A. 1995a. How the horse moves: 1. significance of graphical
representations of equine forelimb kinematics. Equine Veterinary Journal. 27(1):31-38.

Back, W., Schamhardt, H., Savelberg, H., van den Bogert, A., Bruin, G., Hartman, W.
and Bameveld, A. 1995b. How the horse moves: 2. significance of graphical
representations of equine hind limb kinematics. Equine Veterinary Journal. 27(1):39-45.

Back, W., Schamhardt, H., Hartman, W. and Bameveld, A. 1995c. Kinematic differences
between the distal portions of the forelimbs and hind limbs of horses at the trot. American
Journal of Veterinary Research. 56(1 1):] 522-1528.

Badoux, D. 1987. Some biomechanical aspects of the structure of the equine tarsus.
Anatomischer Anzeiger. 164253-61 .

244

Bameveld, B. and van Weeren, P. 1999. Early changes in the distal intertarsal joint of
Dutch Warmblood foals and the influence of exercise on bone density in the third tarsal
bone. Equine Veterinary Journal. 31(supplement):67-73.

Barr, A., Dow, S. and Goodship, A. 1995. Parameter of forelimb ground reaction force in
48 normal ponies. Veterinary Record. 136:283-286.

Becker, C. K., Savelberg, H. H., Buchner, H. H. and Bameveld, A. 1998. Long-term
consequences of experimental desmotomy of the accessory ligament of the deep digital
flexor tendon in adult horses. American Journal of Veterinary Research. 59(3):347-351.

Biewener, A. 1998. Muscle-tendon stresses and elastic energy storage during locomotion
in the horse. Comparative Biochemistry and Physiology Part B. 120:73-87.

Biewener, A., Thomason, J. and Lanyon, L. 1988. Mechanics of locomotion and jumping
in the horse (Equus): in vivo stress in the tibia and metatarsus. Journal of Zoology
London. 214:547-565.

Bjomsdottir, S., Axelsson, M., Eksell, P., Sigurdsson, H. and Carlsten, J. 2000.
Radiographic and clinical survey of degenerative joint disease in the distal tarsal joints in
Icelandic horses. Equine Veterinary Journal. 32(3):268-272.

van den Bogert, A. 1989. Computer stimulation of locomotion in the horse. Thesis.
University of Utrecht. Utretch.

van den Bogert, A. 1998. Computer-assisted gait analysis in equine orthopedic practice:
the case for inverse dynamic analysis. Equine Veterinary Journal. 30(5):362-363.

van den Bogert, A., van Weeran, P. and Schamhardt, H. 1990. Correction for skin
displacement errors in movement analysis of the horse. Journal of Biomechanics.
23(1):97-101.

Bohanon, T. C. 1999. The tarsus, p. 848-862. In J. A. Auer and J. A. Stick (ed.), Equine
surgery, 2nd ed. W. B. Saunders Company, London.

245

Boswell, J. C., McGuigan, M. P., Schramme, M., Newman, S. and Wilson, A. M. 2000.
The position of the point of zero moment relative to the foot in normal horses and horses
suffering from osteoarthritis of the small tarsal joints (bone spavin), 67. In Proceeding of
the Fourth International Workshop on Animal Locomotion.

Buchner, H. H. F. 2001a. The advance of clinical biomechanics. Equine Veterinary
Journal. 33(5):430—432.

Buchner, H. H. F. 2001b. Gait adaptation in lameness, p. 251-279. In W. Back and H.
Clayton (ed.), Equine locomotion. W.B. Saunders, London.

Buchner, F ., Kastner, J ., Girtler, D. and Knezevic, P. 1993. Quantification of hind limb
lameness in the horse. Acta Anatomica. 146: 196-199.

Buchner, H. H. F., Savelberg, H. H. C. M. and Becker, C. K. 1996a. Load redistribution

after desmotomy of the accessory ligament of the deep digital flexor tendon in adult
horses. Veterinary Quarterly. 18(suppl 2):S70—S4.

Buchner, H., Savelberg, H., Schamhardt, H. and Bameveld, A. 1996b. Limb movement
adaptations in horses with experimentally induced fore- or hindlimb lameness. Equine
Veterinary Journal. 28(1):63-70.

Buchner, H., Savelberg, H., Schamhardt, H. and Bameveld, A. 19960. Head and trunk
movement adaptations in horses with experimentally induced fore- or hindlimb
lameness. Equine Veterinary Journal. 28(1):71-76.

Buchner, H., Savelberg, H., Schamhardt, H. and Bameveld, A. 1997. Inertial properties
of Dutch Warmblood horses. Journal of Biomechanics. 30(6):653-658.

Buchner, H. H. F., Oberrnuller, S. and Scheidl, M. 2000. Body centre of mass movement
in the sound horse. Veterinary Journal. 160:225-234.

Buchner, H. H. F ., Obermuller, S. and Scheidl, M. 2001. Body centre of mass movement
in the lame horse. Equine Veterinary Journal. 33(supplement):122-127.

246

 

Budsberg, S. C., Chambers, J. N., van Inc Stephen, L., Foutz, T. L. and Reece, L. 1996.
Prosepctive evaluation of ground reaction forces in dogs undergoing unilateral total hip
replacement. American Journal of Veterinary Research. 57(12): 1781-1785.

Cavagna, G. A., Heglund, N. C. and Taylor, C. R. 1977. Mechanical work in terrestrial
locomotion: two basic mechanisms for minimizing energy expenditure. American Journal
of Physiology. 233(5):R243-R26l.

Chateau, H., Degueurce, C., Jerbi, H., Crevier-Denoix, N., Pourcelot, P., Audigie, F .,
Pasqui-Boutard, V. and Denoix, J. M. 2001. Normal three-dimensional behavior of the
metacarpophalangeal joint and the effect of uneven foot bearing. Equine Veterinary
Journal. 33(supplement):84-88.

Clayton, H. 1990. The effect of an acute angulation of the hind hooves on diagonal
synchrony of trotting horses. Equine Veterinary Journal. 9(supplement):91-94.

Clayton, H. 1996. Instrumentation and techniques in locomotion and lameness.
Veterinary Clinics of North America: Equine practice. 12(2):337-350.

Clayton, H. M. 2001. Performance in Equestrian Sports, p. 193-226. In W. Back and H.
Clayton (ed.), Equine locomotion. W.B. Saunders, London.

Clayton, H., Lanovaz, J ., Schamhardt, H., Willemen, M. and Colborne, G. 1998a. Net
joint moments and powers in the equine forelimb during the stance phase of the trot.
Equine Veterinary Journal. 30(5):384-3 89.

Clayton, H., Theoret, C., Barber, S. and Curle, R. 1998b. Effects of carpal synovectomy
on stride kinematics of trotting horses. Veterinary and Comparative Orthopedic &
Traumatology. 1 1:80-84.

Clayton, H., Hodson, E. and Lanovaz, J. 2000a. The forelimb in walking horses: 2. net
joint moments and joint powers. Equine Veterinary Journal. 32(4):295-299.

Clayton, H., Schamhardt, H., Willemen, M., Lanovaz, J. and Colborne, G. 2000b.
Kinematics and ground reaction forces in horses with superficial digital flexor tendinitis.
American Journal of Veterinary Research. 61(2): 191-196.

247

Clayton, H., Schamhardt, H., Willemen, M., Lanovaz, J. and Colborne, G. 2000c. Net
joint moments and joint powers in horses with superficial digital flexor tendinitis.
American Journal of Veterinary Research. 61(2):197-201.

Clayton, H., Willemen, M., Lanovaz, J. and Schamhardt, H. 2000d. Effects of heel wedge

in horses with superficial digital flexor tendinitis. Veterinary and Comparative
Orthopedic & Traumatology. 13:1-8.

Clayton, H. M., Hodson, E., Lanovaz, J. L. and Colborne, G. R. 2001. The hindlimb of
walking horses: 2. net joint moments and joint powers. Equine Veterinary Journal.
33(1):44-48.

Colborne, G., Lanovaz, J ., Sprigings, E., Schamhardt, H. and Clayton, H. l997a. Joint
moments and power in equine gait: a preliminary study. Equine Veterinary Journal.
23(supplement):33-36.

Colborne, G., Lanovaz, J ., Sprigings, E., Schamhardt, H. and Clayton, H. 1997b. Power
flow in the equine forelimb. Equine Veterinary Journal. 23(supplement):37-40.

Colborne, G., Lanovaz, J ., Sprigings, E., Schamhardt, H. and Clayton, H. 1998. Forelimb
joint moments and power during the walking stance phase of horses. American Journal of
Veterinary Research. 59(5):609—614.

Cottage, 1., Lane, M. and F errars, W. 1997. Review of current methods available for the
treatment of bone spavin. Equine Veterinary Education. 9(5):258-264.

Cross, A. R., Budsberg, S. C. and Keefe, T. J. 1997. Kinetic gait analysis assessment of
meloxicam efficiency in a sodium urate-induced synovitis model in dogs. American
Journal of Veterinary Research. 58(6):626-631.

Dalin, G. and Jeffcott, L. 1985. Locomotion and gait analysis. Veterinary Clinics of
North America: Equine practice. 1(3):549—572.

Degueurce, C., Pourcelot, P., Audigie, F., Denoix, J. and Geiger, D. 1997. Variability of
the limb joint patterns of sound horses at trot. Equine Veterinary Journal.
23(supplement):89-92.

248

Degueurce, C., Charteau, H., Pasqui-Boutard, V., Pourcelot, P., Audigie, F ., Crevier-
Denoix, N., Jerbi, H., Geiger, D. and Denoix, J. 2000. Concrete use of the joint
coordinate system for the quantification of articular rotations in the digital joints of the
horse. Veterinary Research. 31 :297-3 1 l.

Degueurce, C., Charteau, H., Jerbi, H., Crevier-Denoix, N., Pourcelot, P., Pasqui-
Boutard, V., Geiger, D. and Denoix, J. 2001. Three-dimensional kinematics of the
proximal interpahalangeal joint: effects of raising the heels or the toe. Equine Veterinary
Journal. 33(supplement):79-83.

Deuel, N. R. and Lawrence, L. M. 1987. Laterality in the gallop gait of horses. Journal of
Biomechanics. 20:645-649.

Dow, S., Leendertz, J ., Silver, 1. and Goodship, A. 1991. Identification of subclinical

tendon injury from ground reaction force analysis. Equine Veterinary Journal. 23(4):266-
272.

Drevemo, S., Fredricson, I., Hjerten, G. and McMiken, D. 1987. Early development of
gait asymmetries in trotting Standardbred colts. Equine Veterinary Journal. 19:189-191.

Dyce, K. M., Sack, W. O. and Wensing, C. J. G. 1987. The hind limb of the horse, p.
576-595. In D. Pedersen (ed.), Textbook of veterinary anatomy. W. B. Saunders
Company, Philadelphia.

Dyson, S. 1995. Distal hock joint pain: diagnosis and management. lppologia. 6(3):7-23.

Eastman, T. G., Bohanon, T. C., Beeman, G. M. and Swanson, T. D. 1997. Owner survey
on cunean tenectomy as a treatment for bone spavin in performance horses, p. 121-122.
In 43th Proceedings of the American Association of Equine Practitioners.

Eksell, P., Axelsson, M., Brostrom, H., Roneus, B., Haggstrom, J. and Carlten, J. 1998.
Prevalence and risk factors fo bone spavin in Icelandic horses in Sweden: a radiographic
field study. Acta Veterinary Scandinavia. 39:339-348.

Faber, M., Schamhardt, H., van Weeren, R., Johnston, C., Roepstorff, L. and Bameveld,
A. 2000. Basic three-dimensional kinematics of the vertebral column of horses walking
on a treadmill. American Journal of Veterinary Research. 61(4):399-406.

249

Faber, M., Johnston, C., Schamhardt, H., van Weeren, R., Roepstorff, L. and Bameveld,

A. 2001 a. Basic three-dimensional kinematics of the vertebral column of horses trotting
on a treadmill. American Journal of Veterinary Research. 62(5):757-764.

Faber, M., Johnston, C., Schamhardt, H., van Weeren, R., Roepstorff, L. and Bameveld,
A. 2001b. Three-dimensional kinematics of the equine spine during canter. Equine
Veterinary Journal. 33(supplement): 145-149.

Gabel, A. 1983. Prevention, diagnosis and treatment of inflammation of the distal hock,
p. 287-298. In 28th Proceedings of the American Association of Equine Practitioners.

Galisteo, A., Cano, M., Miro, F ., Vivo, J ., Morales, J. and Aguera, E. 1996. Angular joint
parameters in the Andalusian horse at walk, obtained by normal videography. Journal of
Equine Veterinary Science. 16(2):73-77.

Galisteo, A., Cano, M., Morales, J ., Miro, F., Vivo, J. and Aguera, E. 1997. Kinematics in
horses at the trot before and after an induced forelimb supporting lameness. Equine
Veterinary Journal. 23(supplement):97-101 .

Galisteo, A., Cano, M., Morales, J ., Vivo, J. and Miro, F. 1998. The influence of speed
and height at the withers on the kinematics of sound horses at the hand-led trot.
Veterinary Research Communications. 22(6):415-423.

Goodship, A., Brown, P., MacFIE, H., Lanyon, L. and Silver, 1. 1987 . A quantitative
force plate assessment of equine locomotor performance, p. 236-270. In J. R. Gillespie
and N. E. Robinson (ed.) Proceedings of the Second International Conference on Equine
Exercise Physiology. ICEEP Publications, Davis.

Gough, M. and Munroe, G. 1998. Decision making in the diagnosis and management of
bone spavin in horses. In Practice. 20:252-259.

van Harreveld, P. D., Lillich, J. D., Kawcak, C. E., Gaughan, E. M., McLaughlin, R. M.

and DeBowes, R. M. 2002. Clinical evaluation of the effects of immobilization followed
by reimmobilization and exercise on the metacarpophalangeal joint in horses. American

Journal of Veterinary Research. 63(2):282-288.

250

Hartman, W., Schamhardt, H., Lammertink, J. and Badoux, D. 1984. Bone strain in the
equine tibia: an in vivo strain gauge analysis. American Journal of Veterinary Research.
45(5):880-884.

Hjerten, G. and Drevemo, S. 1993. Shortening of the forelimb in the horse during the
stance phase. Acta Anatomica. 146(2-3):193-195.

Hjerten, G., Drevemo, S. and Eriksson, L. 1994. Shortening of the hind limb in the horse
during the stance phase. Equine Veterinary Journal. l7(supplement):48-50.

Hodson, E., Clayton, H. M. and Lanovaz, J. L. 2001. The hindlimb in walking horses: 1.
kinematics and ground reaction forces. Equine Veterinary Journal. 33(1):38-43.

Hof, A. 2001. Scaled energetics of locomotion, p. 351-363. In W. Back and H. Clayton
(ed.), Equine locomotion. W.B. Saunders, London.

Holmstrom, M. 1990. Variation in conformation of Swedish Warmblood horses and
conformational characteristics of elite sport horses. Equine Veterinary Journal.
22(3):186-193.

Holmstrom, M., Fredricson, I. and Drevemo, S. 1994. Biokinematic analysis of the
swedish warmblood riding horse at trot. Equine Veterinary Journal. 26(3):235-240.

Holmstrom, M., Fredricson, I. and Drevemo, S. 1995a. Biokinematic effects of collection
on the trotting gaits in the elite dressage horse. Equine Veterinary Journal. 27(4):281-287.

Holmstrom, M., Fredricson, I. and Drevemo, S. 1995b. Variation in angular pattern
adaptation from trot in hand to passage and piaffe in the Grand Prix dressage horse.
Equine Veterinary Journal. 18(supplement):l32-l37.

van den Hoogen, B. M., van de Lest, C. H., van Weeren, P. R., Lafeber, F. P., Lopes-
Cardozo, M., van Golde, L. M. and Bameveld, A. 1998. Loading-induced changes in

synovial fluid affect cartilage metabolism. British Journal of Rheumatology. 37(6):671-
676.

251

Jacobs, R. and van Ingen Schenau, G. J. 1992. Intermuscular coordination in a sprint
push-off. Journal of Biomechanics. 25(9):953-965.

Jacobs, R., Bobbert, M. F. and van Ingen Schenau, G. J. 1996. Mechanical output from
individual muscles during explosive leg extensions: the role of biarticular muscles.
Journal of Biomechanics. 29(4):513-523.

Jansen, M., van Raaij, J ., van den Bogert, A., Schamhardt, H. and Hartman, W. 1992.
Quantitative analysis of computer-averaged electromyographic profiles of intrinsic limb

muscles in ponies at the walk. American Journal of Veterinary Research. 53(2):2343-
2349.

Jefferson, R. J ., Collins, J. J ., Whittle, M. W., Radin, E. L. and O’Connor, J. J. 1990. The
role of the quadriceps in controlling impulsive forces around heel strike. Journal of
Engineering in Medicine. 204(1):21-28.

Johnson, M. D. and Buckley, J G. 2001. Muscle power patterns in the mid-acceleration
phase of sprinting. Journal of Sports Sciences. 19:263-272.

Johnston, C., Roepstorff, L., Drevemo, S. and Kallings, P. 1996. Kinematic of the distal
hindlimb during stance phase in the fast trotting Standardbred. Equine Veterinary
Journal. 28(4):263-268.

Keegan, K. G., Wilson, D. J ., Wilson, D. A., Barnett, C. D. and Smith, B. 1998a. Effects
of balancing and shoeing of the forelimb feet on kinematic gait analysis in five horses
with navicular disease. Journal of Equine Veterinary Science. 18(8):522-527.

Keegan, K. G., Wilson, D. A., Wilson, D. J ., Smith, B., Gaughan, E .M., Pleasant, R. S.,
Kramer, J ., Howard, R. D., Bacon-Miller, C., Davis, E. G., May, K. A., Valentino, W. L.
and van Harreveld, P. D. 1998b. Evaluation of mild lameness in horses trotting on a
treadmill by clinicians and interns or residents and correlation of their assessments with
kinematic gait analysis. American Journal of Veterinary Research. 59(11):1370-1377.

Keegan, K., Wilson, D., Smith, B. and Wilson, D. 2000. Changes in kinematic variables
observed during pressure-induced forelimb lameness in adult horses trotting on a
treadmill. American Journal of Veterinary Research. 61(6):612-619.

252

Keg, P. R., Bameveld, A., Schamhardt, H. C. and van den Belt, A. J. 1994. Clinical and
force plate evaluation of the effect of a high plantar nerve block in lameness caused by
induced mid-metatarsal tendinitis. Veterinary Quarterly. 16(suppl 2):S70-S75.

Keg, P. R., Schamhardt, H. C. van Weeren, P. R. and Bameveld, A. 1996. The effect of
diagnostic regional nerve blocks in the fore limb on the locomotion of clinically sound
horses. Veterinary Quarterly. 1 8(suppl 2):S 106-S 109.

Khumsap, S., Clayton, H. M. and Lanovaz, J. L. 2001a. Effect of walking velocity on
hindlimb kinetics during stance in normal horses. Equine Veterinary Journal.
33(supplement):21-26.

Khumsap, 8., Clayton, H. M., and Lanovaz, J. L. 2001b. Effect of walking velocity on

ground reaction force variables in the hind limb of clinically normal horses. American
Journal of Veterinary Research. 62(6):901-906.

Khumsap, S., Clayton, H. M., Lanovaz, J. L. and Bouchey, M. 2002. Effect of walking
velocity on forelimb kinematics and kinetics. Equine Veterinary Journal.
34(supplement):in press.

Kirker-Head, C. A., F ackelman, G. E., Hoogasian, J. J. and Doyle, E. M. 1993. Studies
on propentofylline for the treatment of navicular diasease. Journal of Equine Veterinary
Science. 13(2):106-l l3.

Kobluk, C., Schnurr, D., Horney, F., Summer-Smith, G., Willoughby, R., Dekleer, V. and
Hearn, T. 1989. Use of high-speed cinematography and computer generated gait
diagrams for the study of equine hindlimb kinematics. Equine Veterinary Journal.
21(1):48-58.

Kramer, J ., Keegan, K. G., Wilson, D. A., Smith, B. K. and Wilson, D. J. 2000.
Kinematic of the hind limb in trotting horses afier induced lameness of the distal

intertarsal and tarsometatarsal joints and intra-articular administration of anesthetic.
American Journal of Veterinary Research. 61(9):1031-1036.

Lanovaz, J. L. and Clayton, H. M. 2001. Sensitivity of forelimb swing phase inverse
dynamics to inertial parameter errors. Equine Veterinary Journal. 33(supplement):27-3 1.

253

Lanovaz, J ., Clayton, H., Colborne, G. and Schamhardt, H. 1999. Forelimb kinematics
and net joint moments during the swing phase of the trot. Equine Veterinary Journal.
30(supplement):235-239.

Lanovaz, J. L., Khumsap, S., Clayton, H. M., Stick, J. A. and Brown, J. 2002. Three
dimensional kinematics of the tarsal joint at the trot. Equine Vetemary Journal.
34(supplement):308-3 l 3.

Laverty, S., Stover, S., Belanger, D., O'Brien, T., Pool, R., Pascoe, J ., Taylor, K. and
Harrington, T. 1991. Radiographic, high detail radiographic, microangiographic and
histological findings of the distal portion of the tarsus in weanling, young and adult
horses. Equine Veterinary Journal. 23(6):413-421.

Leach, D. 1987. Locomotion analysis technology for evaluation of lameness in horses.
Equine Veterinary Journal. 19(2):97-99.

Leach, D. 1993. Recommended terminology for researchers in locomotion and
biomechanics of quadrupedal animals. Acta Anatomica. 146: 130-136.

Leach, D. and Crawford, W. 1983. Guideline for the future of equine locomotion
research. Equine Veterinary Journal. 15(2):103-110.

Leach, D. and Dagg, A. 1983. A review of research on equine locomotion and
biomechanics. Equine Veterinary Journal. 15(2):93-102.

Leach, D., Ormrod, K. and Clayton, H. 1984. Standardized terminology for the
description and analysis of equine locomotion. Equine Veterinary Journal. 16(6):522-528.

Little, C. B., Ghosh, P. and Rose, R. 1997. The effect of strenuous versus moderate
exercise on the metabolism of proteoglycans in articular cartilage from different weight-
bearing regions of the equine third carpal bone. Osteoarthritis Cartilage. 5(3): 161-172.

Ljungquist, B. 1976. Training of the young or untrained horse, p. 50-78. In Practical
dressage manual. Whittet & Shepperson, Richmond, Virginia.

254

Lukoschek, M., Boyd, R. D., Schaffler, M. B., Burr, D. B. and Radin, E. L. 1986.
Comparison of joint degeneration models. Surgical instability and repetitive impulsive
loading. Acta Orthopedic Scandinavia. 57(4):349-353.

May, S. A. and Wyn-J ones, G. 1987. Identification of hindlimb lameness. Equine
Veterinary Journal. 19(3):185-188.

McLaughlin, R., Gaughan, E., Roush, J. and Skaggs, C. 1996. Effects of subject velocity
on ground reaction force measurements and stance times in clinically normal horses at
the walk and trot. American Journal of Veterinary Research. 57(1):7-11.

Meershoek, L. S., Schamhardt, H. C., Roepstorff, L. and Johnston, C. 2001. F orelimb
tendon loading during jump landings and the influence of fence height. Equine
Veterinary Journal. 33(supplement):6-10.

Merkens, H. and Schamhardt, H. 1988. Distribution of ground reaction forces of the
concurrently loaded limbs of the Dutch Warmblood horse at the normal walk. Equine
Veterinary Journal. 20(3):209-213.

Merkens, H. W. and Schamhardt, H. C. 1994. Relationships between ground reaction
force patterns and kinematics in the walking and trotting horse. Equine Veterinary
Journal. l7(supplement):67-70.

Merkens, H. W., Schamhardt, H. C., Hartman, W. and Kersjes, A. W. 1985. Ground
reaction force patterns of Dutch warmblood horses at normal walk. Equine Veterinary
Journal. 18(3):207-214.

Merkens, H., Schamhardt, H., Hartman, W. and Kersjes, A. 1988. The use of H(orse)
index: a method of analysing the ground reaction force patterns of lame and normal
gaited horses at the walk. Equine Veterinary Journal. 20(1):29-36.

Merkens, H., Schamhardt, H., van Osch, G. and van den Bogert, A. 1993. Ground
reaction force patterns of Dutch Warmblood horses at normal trot. Equine Veterinary
Journal. 25(2):l34-l37.

255

Minetti, A., Ardigo, L., Reinach, E. and Saibene, F. 1999. The relationship between
mechanical work and energy expenditure of locomotion in horses. Journal of
Experimental Biology. 202:2329-2338.

Molenaar, G. J. 1983. Kinematics of the reciprocal apparatus in the horse. Anatomia
Histologia Embryologia. 12(3):278-287.

Morales, J ., Manchado, M., Vivo, J ., Galisteo, A., Aguera, E. and Miro, F. 1998. Angular
kinematic patterns of limbs in elite and riding horses at trot. Equine Veterinary Journal.
30(6):528-533.

Morris, E. A. and Seeherman, H. J. 1987. Redistribution of ground reaction forces in
experimentally induced equine carpal lameness, p. 553-563. In J. R. Gillespie and N. E.
Robinson (ed.) Proceedings of the Second International Conference on Equine Exercise
Physiology. ICEEP Publications, Davis.

Morris, E. A. and Treadwell, B. V. 1994. Effect of interleukin 1 on articular cartilage
from young and aged horses and comparison with metabolism of osteoarthritic cartilage.
American Journal of Veterinary Research. 55(1): 138-146.

Moyer, W., Brokken, T. and Raker, C. 1983. Bone spavin in thoroughbred race horses, p.
81-92. In 29th Proceedings of the American Association of Equine Practitioners.

Moyer, W., Genovese, R. and Corley, J. 1993. Diagnosing subtle lameness problems by
employing specific therapy, p. 203-205. In 39th Annual Convention Proceedings of
American Association of Equine Practitioners.

Newberry, W. N., Garcia, J. J., Mackenzie, C. D., DeCamp, C. E. and Haut, R. C. 1998.
Analysis of acute mechanical insult in an animal model of post-traumatic osteoarthrosis.
Journal of Biomechanical Engineer. 120(6):704-709.

Nicodemus, M. C. 2000. Quantification of the locomotion of gaited horses. Thesis.
Michigan State University, East Lansing.

Peham, C., Licka, T., Scheidl, M. and Girtler, D. 1998. Speed dependency of motion
pattern consistency. Journal of Biomechanics. 31(9):769-772.

256

Peham, C., Licka, T., Girtler, D. and Scheidl, M. 1999a. Supporting forelimb lameness:
clinical judgement vs computerised symmetry measurement. Equine Veterinary Journal.
31(5):417-421.

Peham, C., Scheidl, M. and Licka, T. 1999b. Limb locomotion - speed distribution
analysis as a new method for stance phase detection. Journal of Biomechanics. 32:11 19-
1 124.

Peham, C., Licka, T., Mayr, A. and Scheidl, M. 2000. Individual speed dependency of
forelimb lameness in trotting horses. Veterinary Journal. 160: 135-138.

Peloso, J. G., Stick, J. A., Caron, J. C., Peloso, P. M. and Soutas-Little, R. W. 1993a.
Effects of hylan on amphotericin-induced carpal lameness in equids. American Journal of
Veterinary Research. 54(9): 1527-1534.

Peloso, J. G., Stick, J. A., Soutas-Little, R. W., Caron, J. C., DeCamp, C. E. and Leach,
D. H. 1993b. Computer-assisted three-dimensional gait analysis of amphotericin-induced
carpal lameness in horses. American Journal of Veterinary Research. 54(9): 1535-1543.

Pool, R. R. 1996. Pathologic manifestations of joint disease in the athletic horse, p. 87-
104. In C. W. Mcllwraith and G. W. Trotter (ed.) Joint disease in the horse. W. B.
Saunders Company, London.

Pourcelot, P., Audigie, F., Degueurce, C., Denoix, J. and Geiger, D. 1997a. Kinematic
symmetry index: a method for quantifying the horse locomotion symmetry using
kinematic data. Veterinary Research. 28:525-538.

Pourcelot, P., Degueurce, C., Audigie, F ., Denoix, J. and Geiger, D. l997b. Kinematic
analysis of the locomotion symmetry of sound horses at a slow trot. Equine Veterinary
Journal. 23(supplement):93-96.

Riemersma, D., Schamhardt, H., Hartman, W. and Lammertink, L. 1988. Kinetics and

kinematics of the equine hind limb: in vivo tendon loads and force plate measurements in
ponies. American Journal of Veterinary Research. 49(8):1344-1352.

257

 

Robert, C., Valette, J. and Denoix, J. 1998. Surface electromyographic analysis of the
normal horse locomotion: a preliminary report, p. 80—85. In Conference on Equine Sports
Medicine and Science. Wageningen Pers, Wageningen.

Robert, C., Valette, J. P., Degueurce, C. and Denoix, J. M. 1999. Correlation between
surface electromyography and kinematics of the hindlimb of horses at trot on a treadmill.
Cells Tissues Organs. 165(2):] 13-122.

Rooney, J. 1977. The vertebral column, p. 90-95. In Biomechanics of lameness in horses.
Williams & Wilkins, Baltimore.

Rooney, J. 1990. The jumping behavior of the humeroradial and tarsocrural joints of the
horse. Journal of Equine Veterinary Science. 10(4):3l 1-314.

Rooney, J. R., Thompson, K. N. and Shapiro, R. 1991. A contribution to the study of
velocity, stride length, and frequency in the horse. Journal of Equine Veterinary Science.
1 1(4):208-209.

Rumph, P. F. and Steiss, J. E. 1998. Ground reaction forces in Greyhounds with tibial
nerve injury. American Journal of Veterinary Research. 59(4):375-378.

Rumph, P. F ., Kincaid, S. A., Baird, D. K., Karnmermann, J. R., Visco, D. M. and
Goetze, L F. 1993. Vertical ground reaction force distribution during experimentally

induced acute synovitis in dogs. American Journal of Veterinary Research. 54(3):365-
369.

Rumph, P. F ., Kincaid, S. A., Visco, D. M., Baird, D. K., Karnmermann, J. R. and West,
M. S. 1995. Redistribution of vertical ground reaction force in dogs with experimentally
induced chronic hindlimb lameness. Veterinary Surgery. 24(5):384-3 89.

Savelberg, H., Buchner, H. and Becker, C. 1997. Recovery of equine forelimb fimction
after desmotomy of the accessory ligament of the deep digital flexor tendon. Equine
Veterinary Journal. 23(supplement):27-29.

Schamhardt, H. C. 1998. The mechanics of quadrupedal locomotion. How is the body
propelled by muscles? European Journal of Morphology. 36(4-5):272-279.

258

Schamhardt, H. and Merkens, H. 1994. Objective determination of ground contact of
equine limbs at the walk and trot: comparison between ground reaction forces,
accelerometer data and kinematics. Equine Veterinary Journal. l7(supplement):75-79.

Schamhardt, H., Hartman, W. and Lammertink, J. 1989. Forces loading the tarsal joint in
the hind limb of the horse, determined from in vivo strain measurements of the third
metatarsal bone. American Journal of Veterinary Research. 50(5):728-733.

Schneider, R., Milne, D., Gabel, A., Groom, J. and Bramlage, L. 1982. Multidirectional
in vivo strain analysis of the equine radius and tibia during dynamic loading with and
without a cast. American Journal of Veterinary Research. 43(9): 1541-1550.

Seeherman, H. J ., Morris, E. A. and Fackelman, G. E. 1987. Computerized force plate
determination of equine weight-bearing profiles, p. 536-552. In J. R. Gillespie and N. E.
Robinson (ed.) Proceedings of the Second International Conference on Equine Exercise
Physiology. ICEEP Publications, Davis.

Soderkvist, I. and Wedin, P. 1993. Determining the movements of the skeleton using
well-configured markers. Journal of Biomechanics. 26: 1473-1477.

Sprigings, E. and Leach, D. 1986. Standardized technique for determining the centre of
gravity of body and limb segments of horses. Equine Veterinary Journal. 18(1):43-49.

Stashak, T. S. 1987. Diagnosis of lameness, p. 840-877. In T. S. Stashak (ed.), Adam’s
lameness in horses, 4th ed. Lea & F ebiger, Philadelphia.

Steiss, J. E., White, N. A. and Bowen, J. M. 1989. Electroacupuncture in the treatment of
chronic lameness in horses and ponies: a controlled clinical trial. Canadian Journal of
Veterinary Research. 53(2):239-243.

Suter, E., Herzog, W., Leonard, T. R. and Nguyen, H. 1998. One-year changes in hind
limb kinematics, ground reaction forces and knee stability in an experimental model of
osteoarthritis. Journal of Biomechanics. 31(6):51 1-517.

Thomas, L. 1997. A review of statistical power analysis software. Bulletin of the
Ecological Society of America. 78(2):126-139.

259

Tokuriki, M. and Aoki, O. 1995. Electromyographic activity of the hindlimb muscles
during the walk, trot and canter. Equine Veterinary Journal. 18(supplement):152-155.

Uhlir, C., Licka, T., Kubber, P., Peham, C., Scheidl, M. and Girtler, D. 1997.
Compensatory movements of horses with a stance phase lameness. Equine Veterinary
Journal. 23(supplement): 102-105.

Vincent, W. J. 1999. The t test: comparing means from two sets of data, p. 117-148. In
Statistics in kinesiology, 2nd ed. Human Kinetics, Champaign.

Vorstenbosch, M. A., Buchner, H. H., Savelberg, H. H., Schamhardt, H. C. and
Bameveld, A. 1997. Modeling study of compensatory head movements in lame horses.
American Journal of Veterinary Research. 58(7):713-718.

van Weeren, P. 1989. Skin displacement in equine kinematic gait analysis. Thesis.
University of Utretch. N.V. SDU, Utretch.

van Weeren, P. 2001. History of locomotion research, p. 1-35. In W. Back and H.
Clayton (ed.), Equine locomotion. W.B. Saunders, London.

van Weeren, P., van den Bogert, A., Bameveld, A., Hartman, W. and Kersjes, A. 1990.
The role of the reciprocal apparatus in the hind limb of the horse investigated by a

modified CODA-3 opto-electronic kinematic analysis system. Equine Veterinary Journal.
9(supplement):95-100.

van Weeren, P. R., van den Bogert, A. J. and Bameveld, A. 1992a. Correction models for
skin displacement in equine kinematic gait analysis. Journal of Equine Veterinary
Sciences. 12(3):178-192.

van Weeren, P., Jansen, M., van den Bogert, A. and Bameveld, A. 1992b. A kinematic
and strain gauge study of the reciprocal apparatus in the equine hind limb. Journal of
Biomechanics. 25(11):]291-1301.

Weishaupt, M. A., Wiestner, T., Hogg, H. P., Jordan, P., Auer, J. A. and Barrey, E. 2001.
Assessment of gait irregularities in the horse: eye vs. gait analysis. Equine Veterinary
Journal. 33(supplement): 135-140.

260

 

Wentink, G. 1978. Biokinetical analysis of the movements of the pelvic limb of the horse
and the role of the muscles in the walk and the trot. Anatomy and Embryology. 152:261-
272.

Willemen, M .A., Savelberg, H. H. C. M., Jacobs, M. W. H. and Bameveld, A. 1996.
Biomechanical effects of rocker-toe shoes in sound horses. Veterinary Quarterly.
18(suppl 2):S75-S78.

Willemen, M .A., Savelberg, H. H. C. M. and Bameveld, A. 1997. The improvement of
the gait quality of sound trotting warmblood horses by normal shoeing and its effect on
the load on the lower forelimb. Livestock Production Science. 52(2):]45-153.

Willemen, M .A., Savelberg, H. H. C. M. and Bameveld, A. 1999. The effect of
orthopedic shoeing on the force exerted by the deep digital flexor tendon on the navicular
bone in horses. Equine Veterinary Journal. 31(1):25-30.

Williams, G. E. 2001. Locomotor characteristics of horses with navicular disease.
American Journal of Veterinary Research. 62(2):206-210.

Williams, G., Silverman, B., Wilson, A. and Goodship, A. 1999. Disease-specific
changes in equine ground reaction force data documented by use of principal component
analysis. American Journal of Veterinary Research. 60(5):549-555.

Willms, F., Rohe, R. and Kalm, E. 1996. Significance of genetic aspects of bone diseases
in horses. Pferdeheilkunde. 12:345-346.

Wilson, A. M., Seelig, T. J ., Shield, R. A. and Silverman, B. W. 1998. The effect of foot
imbalance on point of force application in the horse. Equine Veterinary Journal.
30(6):540-545.

Winter, D. A. 1990. Biomechanics and motor control of human movement. John Wiley &
Sons, Inc., New York.

Woltring, H. J. 1994. 3D attitude representation of human joints: a standardization
proposal. Journal of Biomechanics. 27:1399-1414.

261

 

Woltring, H. J ., Huiskes, R., De Lange, A. and Veldpaus, F. 1985. Finite centroid and
helical axis estimation from noisy landmark measurements in the study of human joint
kinematics. Journal of Biomechanics. 18:379-389.

Wu, G. and Ladin, Z. 1996. Limitations of quasi-static estimation of human joint loading
during locomotion. Medical and Biological Engineering and Computing. 34(6):472-476.

Zatsiorky, V., Werner, S. and Kaimin, M. 1994. Basic kinematics of walking: step length
and step frequency, a review. Journal of Sports Medicine and Physical Fitness.
34(2): 109-134.

262

MICHIGAN STATE UNIVERSITY LIB AR

11111111 1111111 11111111111111

3 1293 02372 7468