. .
.

-
am.
a.
A if.
«wfiilnmwmflp ‘

#1

n4 . a,
. . 33.3w.“
“imam.“

V may; ,
ErJJIUJ; in”?!
Liv. . L .

0t

.0‘

3
r .. A”. t r r .
_...fi.u.1m.
- a. “flaw
.2, . tr 1. i
,3 z

r.
a. :

7....
a. .

5. v...
95:.
4... v

 

 

. fifl

1m

flaw” ..
w 4:11...”

5,... .i%

. Lt”? 33¢? n

.gmfiwfifa p
3.. wtw
‘1‘... v

. ,. i

i. .

.3:

. .
JR“

r
en;

‘ its,
Z, W1:

.3.

:
2-,.

in.
a J... .
.

.
1. £3.14,

. .
.5v .1

 

. 2. ..
. L .1. s.
.2 1?. at

 

 

 

 

. py
3.1..

1.:

 

 

;. xauu .
1320 “i .aifiww‘wwp.
. x. it”. :1. 1. vm :3
V .1: .5453. £ :
Ufl,‘ an“: 41%
l.

V 4. .
“W35...“

u. ‘ x ‘
Fumfifimfifiumw :9.

.

.. rtufiflfiq

13;
@311:
35%

«any:
. .-

wfifimm

9.4. ".03.
r :9
33w: u.

. .32;

«W. {I
. - ; 2
3.1.3....

 

.a
.9 Jw.

1,...

1‘4.

wt»... ..
‘44.».
iguana

V x.

.3?

 

THESiS

200i

This is to certify that the

dissertation entitled

QUANTIFICATION OF THE LOCOMOTION OF GAITED HORSES

presented by
Molly Christine Nicodemus

has been accepted towards fulfillment
of the requirements for

PHD degree in Animal Science

 

 

Date // *§\8* oo . (a

 

 

MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

 

 

Li
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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11/00 c'lCIRC/DatoDmpBS-p."

QUANTIFICATION OF THE LOCOMOTION OF GAITED HORSES
By

Molly Christine Nicodemus

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY

Department of Animal Science

2000

t .‘

ABSTRACT
QUANTIFICATION OF THE LOCOMOTION 0F GAITED HORSES
By

Molly Christine Nicodemus

The majority of equine kinematic studies have utilized two-dimensional analysis
since the horse’s limbs have evolved to move primarily in the sagittal plane. However, in
horse breeds noted for their limbs moving outside of the sagittal plane, such as the

Penivian Paso, three-dimensional analysis of the joints is necessary.

Experiment I

The Application of Virtual Markers to a Joint Coordinate System for Equine
Three-Dimensional Joint Motion. Earlier equine three-dimensional studies applying a
joint coordinate system (JCS) encountered errors in the data due to marker displacement
and visibility. This study presented a targeting scheme for the carpus and fetlock that met
the requirements for developing a JCS while addressing marker problems. With this
targeting scheme, an additional set of three-dimensional markers called virtual markers
was applied to the forelimb. The earlier problems were minimized, and results compared

favorably with previous equine studies of the walk.

Experiment 11
Comparison of a Joint Coordinate System versus Multi-Planar Analysis. Multi-

planar analysis (MPA) is more commonly used in equine three-dimensional kinematic

M...

 

'

ow!

IV‘

1.-

 

“'3
“.5

.4.-

Ex

.—»

t

:5

YA
Aurb

.—

\

3x

\A
.nys

analysis, and therefore, this study compared JCS and MPA measurements of the Peruvian
Paso walk. The Peruvian Paso "termino", a characteristic outward motion of the
forelimbs during the swing phase, gave an ideal opportunity to study out-of-sagittal plane
joint motion. MPA was found to be appropriate for flexion/extension when corrections
were made for the angle of travel of the horse. The MPA abduction error at the carpus
was a function of axial rotation and flexion when the antebrachial segment was vertical to
the ground, and adduction/abduction error at the fetlock occurred as a result of
metacarpal rotation when the pastem segment was horizontally aligned along the MPA x-
axis. Adduction/abduction differences were attributed to the JCS axis moving with the

horse’s anatomy during locomotion while the MPA axis was static.

Experiment III

Three-Dimensional Motion of the Corpus and F etlock in the F orelimb of the
Missouri Fox Trotter. Three-dimensional kinematic analysis using a JCS was applied to
the Missouri Fox Trotter to describe the carpal and fetlock joint motion of the flat walk
and fox trot. Although velocity was faster at the fox trot, stride length and ranges of joint
motions were similar between gaits. At both gaits, the carpus extended, abducted, and
externally rotated during stance and flexed, abducted, and internally rotated during swing.
The fetlock demonstrated a single peak of extension during stance and a double peak of
flexion during swing with peak external rotation occurring around the first peak of
flexion. The velocity, diagonal step duration, stride and stance durations, peak carpal

external rotation, and the adduction/abduction joint patterns were different between gaits.

*TI‘HG 'd

DEDICATION
I dedicate this dissertation to my parents. This dissertation is a product of our shared love
for horses. From giving me my first horse to driving across several states to watch me
compete at an intercollegiate horse show, your continued support has helped me reach my

dreams. All that I have achieved is because of you.

iv

ACKNOWLEDGEMENTS

First, I would like to extend my appreciation for all the patience, support, and
guidance from my graduate committee: Dr. Robert Bowker, Dr. Susan Holcombe, and
Dr. Brian Nielsen. A special thanks to my advisors, Dr. Hilary Clayton and Dr. Christine
Corn. Your willingness to explore new areas and to share in my enthusiasm for my
project is greatly appreciated.

I am eternally grateful to Joel Lanovaz for being such a patient and thought
provoking teacher, and the graduate students of the McPhail Lab for going out of their
way to assist in my project. This research would not have been possible without the
Mary Anne McPhail Endowment. I would like to recognize and thank the following
students for their help with data collection and analysis: Sara Blechinger, Anna
Brookhouse, Niki Cilli, Audrey Fletcher, Jenni Himebaugh, Kristen Slater, Jacki
Svetkovich, Katie Swartz, and Diane Ursu. Your excitement for the gaited horse brought
new life to my project!

A personal thanks to my office—mates and all of the other graduate students,
faculty, and staff of the Animal Science Department for all of their help, support, and
friendship through my graduate program.

Finally, I would like to give my deepest thanks to the gaited horse community
including the Michigan Fox Trotter Association and Los Grandes Peruvian Paso Horse
Club for their support and enthusiasm for my project. A special recognition to the
Bulmer, Byerly, Ostrom, and Potter families for the donation of their horses to my

project. Your love for your horses was an inspiration for the development of this gaited

horse research program. I hope that you will learn as much from this dissertation as I
have learned from you. I dedicate my continued research of the gaited horse to your

"four-legged" family members. May this research help in their long and healthy lives.

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................... x
LIST OF FIGURES ............................................................................................................ xi
CHAPTER I
GENERAL INTRODUCTION AND OBJECTIVES ......................................................... 1
CHAPTER 11
GENERAL REVIEW OF LITERATURE ........................................................................... 7
History of the Gaited Horse ..................................................................................... 7
Defining the Gaited Horse ....................................................................................... 9
Changing the Speed... ........................................................................................... 12
Understanding the Gaited Horse ............................................................................ 14
Biomechanical Analysis of Gaited Horses ............................................................ 15
Ground Reaction Force Analysis ............................................................... 15
Two-Dimensional Kinematic Analysis ...................................................... l6
Three-Dimensional Kinematic Analysis .................................................... 20
Equine Net Joint Moments and Joint Powers ............................................ 23
Conclusions ............................................................................................................ 26
CHAPTER IH

EXPERIMENT I: THE APPLICATION OF VIRTUAL MARKERS TO A JOINT
COORDINATE SYSTEM FOR EQUINE THREE-DINIENSIONAL JOINT

MOTION ............................................................................................................................ 27
Summary ................................................................................................................ 27
Introduction ................................................................... -. ........................................ 27
The Theory of Three-Dimensional Analysis ......................................................... 28

Joint Coordinate System ............................................................................ 28
Three-Dimensional Marker Methodology ................................................. 3O
Transformations ......................................................................................... 32
Euler Angles ............................................................................................... 34
Application of a Joint Coordinate System and Virtual Markers ............................ 37
Subject ....................................................................................................... 37
Markers ...................................................................................................... 37
Video Recording and Analysis .................................................................. 39
Statistics ..................................................................................................... 40
Results .................................................................................................................... 4O
Carpal Joint Motion ................................................................................... 40
Fetlock Joint Motion .................................................................................. 43
Discussion .............................................................................................................. 43

vii

CHAPTER IV
EXPERIMENT II: COMPARISON OF A JOINT COORDINATE SYSTEM VERSUS

MULTI-PLAN AR ANALYSIS ......................................................................................... 49
Summary ................................................................................................................ 49
Introduction ............................................................................................................ 50
Materials and Methods ........................................................................................... 51

Subjects ...................................................................................................... 51

JCS Segmental Axes .................................................................................. 51

MPA Segmental Axes ................................................................................ 52

Data Collection .......................................................................................... 52
Marker Placement ...................................................................................... 53

Data Reduction ........................................................................................... 56
Results .................................................................................................................... 58
Carpal Joint ................................................................................................ 58
Fetlock Joint .............................................................................................. 59
Discussion ............................................................................................................. 66
Conclusions. . . ....................................................................................................... 74

CHAPTER V

EXPERIMENT III: THREE-DIMENSIONAL MOTION OF THE CARPUS AND

FETLOCK OF THE FORELIMB OF THE MISSOURI FOX TROTTER ...................... 77
Summary ................................................................................................................ 77
Introduction ............................................................................................................ 78
Materials and Methods ........................................................................................... 78

Subjects ...................................................................................................... 78
Segmental Coordinate Systems .................................................................. 80
Marker Placement ...................................................................................... 81
Data Collection .......................................................................................... 82
Data Reduction .......................................................................................... 82
Data Analysis ............................................................................................. 83
Results .................................................................................................................... 87
Flat Walk ................................................................................................... 87
Fox Trot ..................................................................................................... 87
Discussion .............................................................................................................. 91
Carpal Joint ................................................................................................ 93
Fetlock Joint ............................................................................................... 93
Gait Comparisons ....................................................................................... 95

viii

CH4
C0:

P

Au.

\IT.

'1
’H

CHAPTERVI

CONCLUSIONS ............................................................................................................... 98
Chapter Summaries .................................................................................... 99

Experiment I ................................................................................... 99

Experiment I] ............................................................................... 101

Experiment III .............................................................................. 106

Gait Adaptations ...................................................................................... 109
Recommendation for Future Studies ....................................................... 116

Final Conclusions... ................................................................................ 120
APPENDICES ................................................................................................................ 121
Appendix A: Experiment 11 (Graphs).. ................................................... 122

Appendix B: Experiment II (Derivations) ............................................... 133

Appendix C: Experiment III ................................................................ 135
REFERENCES ............................................................................................................... 147
VITA ............................................................................................................................... 155

ix

s as
u . n .u
Wu AU
0 n VII

5 .rl
.l.v ¢
n‘. I.
.<M A.\
.4 F

..‘
3&1

.\ '
....C‘..

D"

LIST OF TABLES

Table 5.1. Gait characteristics (mean and + SD) measured for the flat walk
and fox trot strides used in this study ................................................................................. 84

Table 5.2. Standing angles for carpal and fetlock flexion/extension,
adduction/abduction, and intemal/extemal rotation of the six
Fox Trotters and the mean and + SD ............................................................. 85

Table 5.3. Mean (SD) for the peak values of three-dimensional joint
motions of the carpus and fetlock for the flat walk and fox trot ........................................ 86

~p'a
:u.5

F:

3—»

pr;

Tn.»

a»

v.-
:1

'1’:
1“-~
b

F

‘ vwn
II ,
n kAn‘o

.o

'1'
f

TABLE OF FIGURES

Figure 2.1. Hildebrand’s graph (1965) demonstrating the distinction
between various gaits according to hind limb stance duration and

% of stride interval that footfall of forefoot follows hind on the same side ...................... 17
Figure 3.1. Three noncollinear markers, A, B, and C, fixed to a rigid body .................... 33
Figure 3.2. The location of vectors used in transformation calculations .......................... 33

Figure 3.3. Caudolateral view of an exploded limb showing the radius, carpus,
and metacarpal segments to illustrate the segmental axes used to develop the joint
coordinate system ............................................................................................................... 36

Figure 3.4. Caudolateral view of an exploded limb showing the metacarpal and
proximal phalangeal segments to illustrate the segmental axes used to develop the
joint coordinate system ...................................................................................................... 36

Figure 3.5. Location of the virtual markers, tracking markers and markers
used as both virtual and tracking markers during the standing file
to establish virtual marker locations relative to tracking markers ..................................... 38

Figure 3.6. Mean + 1 SD. walking carpal flexion/extension angle
relative to the standing carpal flexion joint angle .............................................................. 41

Figure 3.7. Mean + 1 SD. walking carpal adduction/abduction angle
relative to the standing carpal adduction/abduction joint angle ......................................... 41

Figure 3.8. Mean + 1 SD. walking carpal internal/extemal rotation
angle relative to the standing axial rotation joint angle ..................................................... 42

Figure 3.9. Mean + 1 SD. walking fetlock flexion/extension angle
relative to the standing carpal flexion joint angle .............................................................. 42

Figure 3.10. Mean + 1 SD. walking fetlock adduction/abduction
angle relative to the standing fetlock adduction/ abduction joint angle ............................. 44

Figure 3.11. Mean + 1 SD. walking fetlock intemal/extemal rotation
angle relative to the standing fetlock axial rotation joint angle ......................................... 44

Figure 4.1. Data collection with tracking markers attached to the right forelimb
during a walking trial ......................................................................................................... 53

Figure 4.2. Marker placement during the standing file as seen in the lateral
and frontal vrews54

xi

- rr'fi
.L‘.b
s

Ro-o. -.

. .
-» b-s-\

.v';

.;—.b
§

y.
un‘

, . '3
.“.~

 

,
“ 4 ’u
. 0““
:-
‘~.‘
- - e
~3“ "
\ -

Figure 4.3. Calculation used to correct sagittal plane measurements using MPA
for the horse's angle of travel determined from the back markers ............................. 55

Figure 4.4. Measurement of the segmental angles in the sagittal plane.......................57

Figure 4.5. Mean carpal flexion/extension, adduction/abduction, intemal/external

rotation using JCS. . ........................................................................................................... 60
Figure 4. 6. Mean carpal flexion/extension for JCS and MPA ......................................... 60
Figure 4.7. Mean carpal adduction/abduction for JCS and MPA ..................................... 61

Figure 4.8. Absolute difference between carpal adduction/abduction measured
using the JCS and MPA ..................................................................................................... 61

Figure 4.9. Top graph: Carpal flexion/extension and intemal/external rotation measured
using the JCS. Middle graph: Segmental angles in GCS sagittal plane for antebrachium
and metacarpus. Bottom graph: 2 displacement between the proximal and distal

tracking markers for the antebrachium and metacarpus62

Figure 4.10. Mean fetlock flexion/extension, adduction/abduction, internal/extemal

rotation using JCS... ... .................................................................................................... 63
Figure 4.11. Mean fetlock flexion/extension for JCS and MPA ..................................... 63
Figure 4.12. Mean fetlock adduction/abduction for JCS and MPA .............................. 64

Figure 4.13. Absolute difference between fetlock adduction/abduction
measured using the JCS and MPA ..................................................................................... 64

Figure 4.14. Top graph: Fetlock flexion/extension and intemal/external rotation
measured using the JCS. Middle graph: Segmental angles in GCS sagittal plane for
metacarpus and pastem. Bottom graph: Z displacement between the proximal and distal
tracking markers for the metacarpus and pastem .......................................................... 65

Figure 4.15. The true flexion/extension and adduction/abduction angles and the angular
errors created in the apparent position of the distal segment by a rotation of the proximal
segment when measured using MPA ................................................................................ 68

Figure 4.16. Contour surface showing adduction/abduction error measured

by MPA in the frontal plane as a result of different combinations of created

by axial rotation and flexion when the proximal segment is vertical in the

sagittal plane68

xii

44. r... 4 an. vi 1*. Jam .3 1w .w. .s. .x. .L .s. .u
N w .o . . » . . . L N «l.
9» .Jr 5 » ”K as o . a» O\ s
v . . P u“ d . 9. PA r a ‘ n c o. n
. . S . . .... 5 . .x .5 . S _ . he .. . A
W-I‘b F vu-V I— F‘. 7-hb firm \- %.-. F k W... VJ.‘ o.“ ‘-.b O.
:1 A 3. t. a C12 .5 n. :1 G r. t. .7. ti 7.

mg“; 0"
-..

1.1. “I"
.-_.

For.

Figure 4.17. Angle between the JCS adduction/abduction axis and the x-axis of the

GCS for the carpal joint and the fetlock joint ................................................................ 71
Figure 4.18. Y displacement of the carpal joint center throughout a single

stride and the corresponding orientation of the JCS carpal adduction/abduction

axis within the sagittal plane shown at intervals of 3% of the stride. 71
Figure 4.19. Y displacement of the fetlock joint center throughout a single

stride and the corresponding orientation of the JCS carpal adduction/abduction

axis within the sagittal plane shown at intervals of 3% of the stride................... .....72
Figure 4.20. Angle between the JCS flexion/extension axes and the z—axis of the

GCS for the carpal joint and the fetlock joint ................................................................ 72
Figure 5 .1. Tracking marker placement for horses 1-3 and horses 4-6 ............................ 81
Figure 5.2. Mean carpal flexion/extension, adduction/abduction, intemal/extemal
rotation of the flat walk ...................................................................................................... 88
Figure 5.3. Mean fetlock flexion/extension, adduction/abduction, intemal/extemal
rotation of the flat walk ...................................................................................................... 88
Figure 5.4. Mean carpal flexion/extension, adduction/abduction, intemal/extemal
rotation of the fox trot ........................................................................................................ 90
Figure 5.5. Mean fetlock flexion/extension, adduction/abduction, intemal/extemal
rotation of the fox trot ........................................................................................................ 90
Figure 6.1. Skewed frontal view measurements as a result of external rotation

that creates an unreal abduction angle ............................................................................. 103

xiii

. n . . o . ll. . . 'IA .1. 11 .l. . fin .. . :4 ‘ s

. v . . -u .4 .un 44 1 ... .- .. .. ... . .L ... .....C w. ..u
H... 3m L. 3*. v... ..m m... IL m... .4. *3 1‘ T: 2* 7: Em f: .Um T: v .a. 3: 4M
.— _ C1 m _ F on [I E :1 9: :1 .3 Cl ML 3. u» C. “A 3.. 4.x

film. .3 . .
F T C.

LIST OF APPENDD( A FIGURES

Figure A.1. JCS carpal flexion/extension angles measured for Peruvian walking
trial 1, 2, and 3 ............................................................................................................. 122

Figure A. 2. MPA carpal flexion/extension angles measured for Peruvian walking
trial 1, 2, and 3 ............................................................................................................. 122

Figure A. 3. JCS carpal adduction/abduction angles measured for Peruvian walking
trial 1, 2, and 3 ............................................................................................................. 123

Figure A. 4. MPA carpal adduction/abduction angles measured for Peruvian walking
trial 1, 2, and 3 ............................................................................................................. 123

Figure A. 5. JCS carpal intemal/extemal rotation angles measured for Peruvian
walking trial 1, 2, and 3 ............................................................................................... 124

Figure A. 6. Carpal absolute difference between JCS and MPA adduction/abduction
angles measured for Peruvian walking trial 1, 2, and 3 ........................................ 124

Figure A. 7. JCS fetlock flexion/extension angles measured for Peruvian walking
trial 1, 2, and 3 ............................................................................................................. 125

Figure A. 8. MPA fetlock flexion/extension angles measured for Peruvian walking
trial 1, 2, and 3 ............................................................................................................. 125

Figure A. 9. JCS fetlock adduction/abduction angles measured for Peruvian walking
trial 1,2, and 3... ............................................................................................................. 126

Figure A. 10. MPA fetlock adduction/abduction angles measured for Peruvian
walking trial 1, 2, and 3.. ............................................................................................... 126

Figure A. 11. JCS fetlock intemal/extemal rotation angles measured for Peruvian
walking trial 1, 2, and 3 ................................................................................................. 127

Figure A. 12. Fetlock absolute difference between JCS and MPA adduction/
abduction angles measured for Peruvian walking trial 1, 2, and 3127

Figure A. 13. Radius segmental flexion/extension angles measured for Peruvian
walking trial 1, 2, and 3 .................................................................................................. 128

Figure A. 14. Metacarpal segmental flexion/extension angles measured for Peruvian
walking trial 1, 2, and 3 .. ............................................................................................... 128

Figure A. 15. Pastem segmental flexion/extension angles measured for Peruvian
walking trial 1, 2, and 3 .................................................................................................. 129

xiv

p; evog

A I..’Q‘
»

0'; ~55;

s “‘5

1

l
I

Figure A. 16. Differences between the z-coordinates of the proximal and distal
markers of the radius segment during Peruvian walking trial 1, 2, and 3129

Figure A. 17. Differences between the z-coordinates of the proximal and distal markers
of the metacarpal segment during Peruvian walking trial 1, 2, and 3 ....................... 130

Figure A. 18. Differences between the z-coordinates of the proximal and distal markers
of the pastem segment during Peruvian walking trial 1, 2, and 3130

Figure A. 19. Angles measured between the JCS carpal flexion/extension axes '
and the GCS z-axis for Peruvian walking trial 1, 2, and 3 ..................................... 131

Figure A. 20. Angles measured between the JCS carpal adduction/abduction
axes and the GCS x-axis for Peruvian walking trial 1, 2, and 3 .............................. 131

Figure A. 21. Angles measured between the JCS fetlock flexion/extension axes
and the GCS z-axis for Peruvian walking trial 1, 2, and 3... .......................................... 132

Figure A. 22. Angles measured between the JCS fetlock adduction/abduction axis
and the GCS x-axis for Peruvian walking trial 1, 2, and 3 ..................................... 132

XV

.....,.:
3....-
-

’4‘ ..

“ab-ob

..
1......
.

A ¢--.
v

A4,.
a“ a.

.' m

‘5’

”11 9 ‘

lb '1

my“

LIST OF APPENDIX C FIGURES

Figure C.1. Fox Trotter #1 carpal mean and SD +/- flexion/extension,

adduction/abduction, intemal/extemal rotation of the flat walk ............................

Figure C. 2. Fox Trotter #1 fetlock mean and SD +/- flexion/extension,

adduction/abduction, intemal/extemal rotation of the flat walk .................................

Figure C. 3. Fox Trotter #1 carpal mean (thick line) and SD +/- flexion/extension,

adduction labduction, intemal/extemal rotation of the fox trot .

Figure C. 4. Fox Trotter #1 fetlock mean and SD +/— flexion/extension,

adduction/abduction, intemal/extemal rotation of the fox trot................ .

Figure C.5. Fox Trotter #2 carpal mean and SD +/- flexion/extension,

adduction/abduction, intemal/extemal rotation of the flat walk ..................................

Figure C. 6. Fox Trotter #2 fetlock mean and SD +/- flexion/extension,

adduction/abduction, internal/extemal rotation of the flat walk

Figure C. 7. Fox Trotter #2 carpal mean and SD +/- flexion/extension,

adduction/abduction, intemal/extemal rotation of the fox trot

Figure C. 8. Fox Trotter #2 fetlock mean and SD +/- flexion/extension,

adduction/abduction, intemal/extemal rotation of the fox trot

Figure C.9. Fox Trotter #3 carpal mean and SD +l- flexion/extension,

adduction/abduction, intemal/extemal rotation of the flat walk........... . . . . . . .

Figure C. 10. Fox Trotter #3 fetlock mean and SD +/- flexion/extension,

adduction/abduction, intemal/extemal rotation of the flat walk .

Figure C. 11. Fox Trotter #3 carpal mean and SD +/- flexion/extension,

adduction/abduction, intemal/extemal rotation of the fox trot........ . . .

Figure C. 12. Fox Trotter #3 fetlock mean and SD +/- flexion/extension,

adduction/abduction, intemal/external rotation of the fox trot

Figure C. 13. Fox Trotter #4 carpal mean and SD +/- flexion/extension,

adduction/abduction, intemal/extemal rotation of the flat walk....... . . . . . .

Figure C. 14. Fox Trotter #4 fetlock mean and SD +/- flexion/extension,

adduction/abduction, intemal/extemal rotation of the flat walk ............................

Figure C. 15. Fox Trotter #4 carpal mean and SD +/- flexion/extension,

adduction/abduction, intemal/extemal rotation of the fox trot ..............................

xvi

..l35

..l35

136

...136

137

137

138

138

139

139

140

140

141

..141

..142

.tm-i

“if. -‘.~

 

 

.Ir-w-g

...ueoS
v

.
,4-...o.

49.4. ..

3'.”
-;-.§
5

14.4.. ..
“Lanes

r‘ 1“":

1 --_..
»

<Q..‘
‘9‘»

Figure C. 16. Fox Trotter #4 fetlock mean and SD +/- flexion/extension,
adduction/abduction, intemal/external rotation of the fox trot ................................ 142

Figure C.l7. Fox Trotter #5 carpal mean and SD +/- flexion/extension,
adduction/abduction, intemal/extemal rotation of the flat walk .............................. 143

Figure C. 18. Fox Trotter #5 fetlock mean and SD +/- flexion/extension,
adduction/abduction, intemal/external rotation of the flat walk .............................. 143

Figure C. 19. Fox Trotter #5 carpal mean and SD +/- flexion/extension,
adduction/abduction, intemal/external rotation of the fox trot ................................ 144

Figure C. 20. Fox Trotter #5 fetlock mean and SD +/- flexion/extension,
adduction/abduction, intemal/extemal rotation of the fox trot ................................ 144

Figure C.21. Fox Trotter #6 carpal mean and SD +/- flexion/extension,
adduction/abduction, intemal/external rotation of the flat walk .............................. 145

Figure C. 22. Fox Trotter #6 fetlock mean and SD +/- flexion/extension,
adduction/abduction, internal/extemal rotation of the flat walk .............................. 145

Figure C. 23. Fox Trotter #6 carpal mean and SD +/- flexion/extension,
adduction/abduction, intemal/extemal rotation of the fox trot ................................ 146

Figure C. 24. Fox Trotter #6 fetlock mean and SD +l- flexion/extension,
adduction/abduction, internal/extemal rotation of the fox trot ................................ 146

xvii

 

TH; .va-l‘a

uLV Alba.»

.34.”)- q
xflv¥..\

-“Q. '3
0

;u... Ab?

. 0

'fl_\' a-v-

“X- It“.

‘v-uq-..
~ I!\
'm..~.,

“ma- .

a may, .‘

‘-

! .‘_
may. .b

‘4‘
-‘ ,
\_ N.
.“Z ;
K‘,‘

v :
'\‘w
-

s_v.
.. ,a‘

.
.L-IQ

CHAPTER I

GENERAL INTRODUCTION AND OBJECTIVES

Locomotion, the act of moving from one place to another, is achieved by cycling
the limbs back and forth in a variety of patterns called gaits. Each gait is defined by the
sequence and timing of the movements. Horses are particularly versatile in terms of their
gait repertoire. Gait analysis is applied to measure and describe the kinematics, the
movements of the body, and the kinetics, the forces creating the movements. Human
curiosity concerning the movement of the horse has been evident for many years. The
American photographer Eadweard Muybridge made a breakthrough in 1877 when he
provided photographic evidence of the existence of a suspension (airborne) phase in the
gait of the trotting horse. He went on to take thousands of sequential photographs of
people and animals engaged in various activities (Muybridge 1899). Muybridge's
recordings were later applied to determine the footfall sequences for normal equine gait
(Clayton 1989).

Over the years the technology used for gait analysis has evolved beyond simply
documenting footfall sequences to applying gait analysis for clinical and performance
evaluation. The mechanics of the walk, trot, canter, and gallOp have been extensively
studied, but the application of gait analysis to exotic equine gaits, such as the fox trot and
paso gaits, is limited. Knowledge of the kinematics of these gaits as measured in the
following chapters will assist in the evaluation of a horse's level of training and
performance, in the detection of pathological gait, and in assessing the response to

surgical or medical therapy.

Gaited horses perform a range of gaits that are not performed by the non-gaited
breeds. The gaited horse breeds have the ability to replace the trot with various four-beat
stepping gaits described as non-walking stepping gaits. A four-beat gait is one in which
each limb contacts the ground separately. A stepping gait lacks a period of suspension.
The walk is a lateral sequence, stepping gait with a regular four-beat rhythm and at least
two feet supporting the body at one time (Speirs 1997). The walk is described as “the
mother of gaits” since the non-walking stepping gaits are considered variations of the
walk (Imus 1995).

In contrast to the stepping gaits, the trot, pace, canter, and gallop have a
suspension phase, a period where the horse is in flight. Gaits with a suspension phase are
classified as leaping gaits. In contrast to the stepping gaits, the leaping gaits use the
suspension phase to assist in increasing the speed of the gait (Clayton 1994; Leach et a1.
1987).

Although the non-walking stepping gaits and the walk share similar
characteristics, some of the kinematic variables must change in order to produce the
quicker four-beat stepping gait. Humans demonstrated an increased range of joint motion
as the speed increased at the walk (Mann 1975). This increased joint flexion reduces the
risk of musculoskeletal damage by alternating the increased impact forces associated with
ground contact at a higher speed (I-Iershman and Nicholas 1995). Furthermore, increased
joint flexion may also act as a mechanism for increasing the inverted pendulum effect.
The decrease in limb height due to flexion increases the upward centrifugal force, and in

turn, exceeds the downward gravitational force. This ratio is described as the Froude

39-5 fig
itdt\h . \

$3685. 01 I

.Q
. . ..
ACKQ,“

I
.Q ‘

'3 ‘7“ a;
b‘nte u.~

.

7"“':,".'

h. e u...
. i

1

'“'a'~ .-

c.5‘--'l

‘ “'1’ v- -
he...:~’§

number. As the Froude number exceeds 1.0, the horse limbs are more apt to fly off the
ground and increase stride turnover (Gray 1968).

In horses, van Weeren et a1. (1993) reported an increase in maximal carpal and
fetlock flexion and hind limb translational motion as Standardbred trotters increase the
speed of the trot from 6 to 9 m/s. In humans, increased flexion is coupled with rotation of
the distal limb joints. Maximal internal rotation of the knee occurs at 15° to 45° of
flexion. Maximal external rotation occurs at 60° to 90° of flexion. Internal rotation is
attributed to the application of an anterior force and external rotation results from the
application of a posterior force. The amount of rotational force and the stability of the
anterior and posterior cruciate ligaments determine the amount of coupled rotation in
humans (Mann 1980).

Without the benefit of a suspension phase, the gaited horse may utilize increased
flexion and coupled rotation to produce a faster four-beat gait though this has not been
studied scientifically. Although axial rotation and abduction/adduction are limited in the
distal joints of the equine limb, a noted exception is the "termino" motion that occurs
when the Peruvian Paso performs the paso gaits. Terrnino is characterized by the
forelimbs of the horse swinging outward from the shoulder during the swing phase of the
stride (Harris 1993). This movement is a desirable characteristic of the Peruvian Paso's
gaits, but is regarded as a defect in other breeds. Analysis of these unique gait
characteristics, as described in the following chapters, will assist in the further
understanding of the mechanisms of the non-walking stepping gaits.

These unique characteristics of the gaited horse, such as the additional

adduction/abduction and axial rotation, create difficulties for measuring joint motion

c

. . 'v-v' 4'

51:: “AW

5

it; v 91

tiulfi \“O-
J ' r~~°

1.1.1: T014.

.‘. ‘.
mum-3..;
C

rSot'm-‘L

using traditional kinematic analysis techniques. Axial rotation can not be measured using
either two-dimensional analysis or multi-planar analysis. Furthermore, the effects of
axial rotation and flexion on frontal plane measurements of human joint motion using
multi-planar analysis have found distortion in the adduction/abduction measurements
(Soutas—Little et a1. 1987). Therefore, in the studies described in this dissertation,
standard gait analysis techniques were adapted to measure three-dimensional analysis
using a joint coordinate system that produced acceptable accuracy. The availability of
these techniques facilitated measurement of such gait characteristics as coupling of
flexion and axial rotation. The termino motion of the Peruvian Paso was studied as an
extreme example of motion occurring outside of the sagittal plane. These techniques

were further applied to the gaits of the Missouri Fox Trotter.

 

:TC 1T.“

«J
\.\k

p

it far 1:;

The objectives of this study were:

1) To develop a virtual marker methodology that creates an anatomically meaningful and

repeatable joint coordinate system.

2) To compare the three-dimensional joint motion measured using a joint coordinate

system versus multi-planar analysis using the Peruvian Paso.

3) To measure the three—dimensional carpal and fetlock joint motion during the flat walk

and fox trot of the Missouri Fox Trotter.

Hypothesis:

The first chapter of this dissertation describes the development of a reliable and
repeatable methodology for the establishment of a joint coordinate system for equine
three-dimensional motion analysis. In the later chapters, the methodology was applied to

test the following hypotheses:

1) The measured values of abduction/adduction and flexion/extension of the carpal and
fetlock joints are different when measured with an anatomically-based joint coordinate

system versus multi-planar analysis.

2) The three-dimensional joint motion measured for the flat walk and fox trot
demonstrates characteristics unique to each gait and different from the kinematics

measured in earlier studies for the walk and trot.

History

21F.I'
pl“.-

”‘1

did-law.
'v i-y’éasa.
l

.2.
fi ‘\
I
9h,‘
. A
it.“ x ,
I ‘-\
rs...

 

CHAPTER II

GENERAL REVIEW OF THE LITERATURE

History of the Gaited Horse

The 55 million-year history of the horse has been documented through drawings,
paintings, sculptures, and stories. Unlike modern day art that demonstrates the horse
performing the trot, earlier artistic accounts show the horse performing a non-walking
stepping gait. Evidence for the existence of gaited horses that performed a four-beat
stepping gait other than the walk can be traced as far back as the Ice Age. Ancient ice
carvings discovered in France depict a lateral stepping gait similar to the amble (Imus
1995). Hipparion equid fossil footprints, dating from 3.5 million years ago, were
discovered in 1979 in Tanzania under volcanic acid. Comparisons of the fossil footfall
pattern with contemporary gait patterns determined the Hipparion equid was performing
a running walk (Budiansky 1997). The flying horse of Kansu, a bronze horse statue of
the Eastern Han Dynasty (25 AD—220 AD), displays the racking gait (Ensminger 1977).

During the middle ages in Western Europe, the Palfrey, an ambling saddle horse,
flourished among the royalty and society people. Wealthy individuals bred for horses
that performed a smooth, easy riding gait. The Palfrey was a very refined, elegant horse
that performed a fast, ground covering walking gait that lacked the jarring motion of the
suspension phase while still maintaining speeds comparable to the trot and pace.
Chaucer’s Canterbury Tales makes references to the Palfrey when he writes about the

Wife of Bath, “Upon an Amblere easily she sat” (Imus 1995). However, the longer

..i a-Q‘Aa". l
Lani“ no»; 4 . :

hoses. T:

(mast;
Q

«:71 us, \

distances traveled by coaches called for horses with greater speed and strength and this
led to the rise of the trotting horse (Imus 1995).

The English began to settle the Americas in the 17th century. Most of the settlers
were poor so the horses they brought over the oceans were grade trotting horses. Around
the same time as the English settlers arrived, the Spanish military were actively
conquering America as they rode north through Mexico on their majestic Spanish gaited
horses. The Conquistadors traveled through America battling the Native Indians. If the
Conquistadors lost a battle, their horses were left to run wild or were captured by the
Indians. Gradually, these gaited horses were bred with the grade horses of the Indians
and the poor English settlers. Some of the genetic traits of the gaited horses were passed
along, but over time these gaits were seen less frequently (Imus 1995).

The popularity of the gaited horse returned to America with the growth of
plantation farming. Plantation owners bred horses that could perform a ground-covering
stride while moving at a smooth, comfortable gait. The horses were ridden between the
rows of crops as the owners observed the conditions of the crops and work of the slaves
(Imus 1995). The further rise in the popularity of the gaited horse is attributed to the
current growth in "baby boomers," the population that is aged fifty and older (Pascoe
1999). As this population ages while maintaining financial security, older adults are
taking up horseback riding. Unlike the younger riders, this older adult population is

looking for a smooth, easy riding horse (Imus 1995).

A Defining the Gaited Horse

What defines a horse as being gaited? A gaited horse is one that has the ability to
perform non-walking stepping gaits that can reach speeds equivalent to the trot or the
pace (Imus 1995). The trot is a two-beat gait in which a diagonal pair of legs move
synchronously. Diagonal stance is followed by a suspension (airborne) phase, a period
when all four hooves are off the ground. The-pace is similar to the trot, but instead of the
diagonal limb pairs moving synchronously, movements of the lateral limb pairs are
synchronized. Due to the presence of suspension phases, the trot and pace are defined as
leaping gaits. This is in contrast to the stepping gaits that lack a period of suspension.

The most common stepping gait is the walk, which is performed by both gaited
and non-gaited horses. The walk is described as “the mother of gaits” since the other
four-beat gaits are considered variations of the walk (Imus 1995). The most obvious
difference between the walk and the non-walking stepping gaits is the speed of
progression, which is faster in the non-walking stepping gaits. However, stylistic
differences between breeds (e. g. animation and head nod) play a role in defining the
gaits, in addition to the temporal characteristics.

The footfall sequence of the walk is left hind limb (LH)- left forelimb (LF)— right
hind limb (RH)- right forelimb (RF) and the footfalls occur in a regular four-beat rhythm.
Regularity is defined kinematically as the footfalls being spaced evenly in time so that
limb motions are 90° out of phase and the duty factor of each limb is 25% (Hildebrand
1978). The regularity of the footfalls of the gait classifies the walk as a square gait, as

opposed to a gait that shows lateral or diagonal couplets.

A symmetrical gait is one in which the left and right sides move 180° out of phase
(Hildebrand 1965) so that the interval between the left footfall and right footfall is equal
to the interval between the right footfall and left footfall in both the fore and hind limb
pairs (Hildebrand 1978). The walk, trot, pace, fox trot, rack, slow gait, paso gaits, and
running walk are all symmetrical gaits. The timing and sequence of the limb movements
of symmetrical gaits can be described using two kinematic variables: the hind limb stance
duration and the lateral advanced placement, both expressed as a percentage of stride
duration (Hildebrand 1965).

The non-walking stepping gaits can be classified according to the rhythm of the
footfalls. The rhythm may be regular, irregular with lateral couplets, or irregular with
diagonal couplets. The rhythm affects the duration of the limb support phases of the
stride. Horses performing a regular (or square) four-beat gait spend equal amounts of
time in diagonal and lateral support phases (Imus 1995). Besides the walk, other square
gaits include the flat walk, rack, tolt, flying pace, paso llano, and paso largo.

In a true lateral gait, movements of the lateral pairs of limbs are synchronized.
The pace is usually described as a lateral gait in which the lateral support phase is
followed by a period of suspension. At racing speed, however, pacers have been shown
to dissociate the movements of the lateral limb pairs so that contact of the hind limb
precedes that of the lateral forelimb by 3-10% of the stride duration. Thus, ground
contact of the lateral limb pair occurs as a lateral couplet (Wilson et al. 1988). This
dissociation of the pacing lateral limb pairs so that the hind limb makes contact before the
ipsilateral forelimb creates a lateral footfall sequence similar to the walk and non-walking

stepping gaits. A lateral stepping gait also demonstrates an irregular rhythm with lateral

lO

 

curs rats:

of 1020‘}

mind: the

 

‘. " <
I1:235 .99-

C“-‘Iho -5
‘ ‘\
' 'ifibhvua
o

L”;-

".V'U-
Anabk t‘.‘-

‘ h“\‘

Feb”:
L ‘1‘.~

”431B"

"

couplets but differs from the pace by not having a period of suspension. A lateral couplet
occurs when the hind footfall is followed by the ipsilateral fore footfall after an interval
of 10-20% of the stride (Hildebrand 1965). Four-beat stepping gaits with lateral couplets
include the broken pace, slow gait, amble, running walk, sobreandado, and single foot
(Imus 1995). During these gaits the hind footfall precedes the fore footfall by 10-15% of
the stride duration (Hildebrand 1965).

In a true diagonal gait, movements of the diagonal limb pairs occur
synchronously. Similar to the modified lateral gait, the modified diagonal gait is created
by the dissociation of diagonal limb pairs, resulting in an irregular rhythm with diagonal
couplets. Diagonal couplets occur when the diagonal limb movements dissociate by 10-
15% of the stride duration with the forelimb making contact just prior to the diagonal
hind limb (Hildebrand 1978).

The trot is described as a diagonal gait, though even at slow speeds the diagonal
footfalls may occur slightly asynchronously (Clayton 1994; Holstrom et al. 1994). The
diagonal advanced placement is defined as being positive when the hind hoof contacts the
ground before the diagonal fore hoof and negative when the fore hoof contacts the ground
before the diagonal hind hoof. A positive diagonal advanced placement has been
associated with good quality of the trot in warmblood horses (Holstrom et al. 1994). At
racing speeds the trot demonstrates a marked dissociation of the diagonal limb pairs
creating diagonal couplets (Drevemo et al. 1980) with the fore hoof contacting the
ground before the diagonal hind hoof, which is a negative diagonal advanced
displacement. This dissociation of the diagonal limb pair at the racing trot creates a

lateral footfall sequence (LH, LF, RH, RF) that resembles the walk and the non-walking

11

rrw

“3.635;“

H -‘

'1H...

-.—

\Q

~:-. ‘(
Emlw‘

“115le
“2:15:71

35.11213,

r‘x.- _
J,“ .‘x
..._ ‘
Vi
.-u

I .?
~‘e1v a
‘N
“
Jr?

stepping gaits. The fox trot is an example of a four-beat stepping gait with diagonal

couplets (Imus 1995).

Changing the Speed

To increase speed, horses use transitions between gaits combined with changes of
stride length and stride duration within each gait. Leach et al. (1987) studied the changes
in stride length and stride duration in relation to speed. In general, an increase in speed
was associated with reducing stride duration by increasing stride length. Shortening the
stride duration is achieved as a result of reductions in stance duration and overlap and an
increase in the duration of the suspension phase. The suspension phase is a large
contributor to maintaining and increasing the speed of the gait at the trot (Clayton 1998),
canter (Clayton 1994), and gallop (Leach et al. 1987). The non-walking stepping gaits,
which do not show a suspension phase, are performed at speeds equivalent to the trot
while maintaining a stepping sequence (Imus 1995). The mechanisms by which this is
achieved have not been investigated in detail.

The non-gaited horse initially increases speed by walking faster, which is
achieved by a combination of lengthening the stride and increasing the stride rate. At a
certain walking speed there is a transition to the trot. In ponies, the walk-trot transition
occurs at a speed range of 1.5 to 2.2 m/s (Hoyt and Taylor 1981). Gait transitions may
occur as a means of improving energetic and metabolic efficiency (Hoyt and Taylor
1981) or as a means of changing the forces on the limbs (Taylor 1985; Niki et al. 1984).
Vertical ground reaction forces increase at the walk-trot transition and decrease at the

trot-canter transition. This decrease in forces is attributed to changes in limb movement

12

II
C" $th ~v
-‘A~‘ 5‘
b r

H.312;

L...‘ -
“1“?) IT

’?v~
5...,1-3 r

..
a
\-

occurs 1

patterns (Rubin and Lanyon 1982). The transition from canter to gallop occurs more
gradually as the diagonal limb pair (leading hind and trailing fore) become more
dissociated with the hind limb making contact prior to the diagonal forelimb. The hind
limb propulsive forces then act to lift the forelimbs, thus reducing forelimb forces (Pratt
1983).

The gallop usually has one suspension phase in each stride. The speed of the
gallop increases in correspondence to stride length, but anatomically the horse reaches a
maximum stride length (Hiraga er al. 1994). Velocity is increased further by cycling the
limbs more rapidly to increase the stride rate. However, at maximum stride length and
stride rate, some horses add another suspension phase to the gallop sequence. This
occurs between lift off of the leading hind limb and contact of the trailing forelimb. A
few horses even add a very short third suspension phase between the stance phases of the
two forelimbs (Deuel et al. 1986; Rooney 1998).

The gaited horse does not make a distinct transition from the walk to a non-
walking stepping gait. Instead, there is a gradual adjustment of the timing of the limb
movements. Hildebrand (1965) describes the gaits of the gaited horse as a continuum that
does not show distinct boundaries between the non-walking stepping gaits. Therefore,
the arbitrary boundaries of non-walking stepping gaits form a gait continuum in which
breed and style contribute to the differentiation between gaits.

Gaited horses continue to perform stepping gaits at speeds equivalent to those
performed at the trot of the non-gaited horses (Hildebrand 1965). The Tennessee Walking
horse, for example, maintains the footfall pattern of the walk at speeds up to 20 miles per

hour (Harris 1993). The Missouri Fox Trotter can travel long distances at a relatively fast

13

fer t'c

- .0:

in.» h.

’flfi.“ .

o»fi-'\AI

o. -. M
A

A‘.'-.

Ind:

5“

a
‘ f
I

fox trot without tiring. This apparent energetic efficiency is why the US. Forest Service
selected these horses for use by their park rangers (Ridings and Bradbury 1976). The
lack of a transition to the trot in gaited horses may indicate that the non-walking stepping
gaits are more energetically efficient or that they create lower forces on the limbs.

The mechanics utilized for increasing the speed of the non-walking stepping gaits
involves adjusting the stride length and stride frequency. Slight adjustments in the
footfall timing of the various non-walking stepping gaits have also been recognized as a

factor for influencing the speed of the gait (Imus 1995).

Understanding the Gaited Horse

So why is it important to understand the mechanics of the non-walking stepping
gaits and the similarities and differences between these gaits and other equine gaits? The
answer to this question relates both to understanding and judging performance and to the
detection of pathological gaits. For lameness evaluations, the trot is the gait of choice
due to its rhythmic, symmetric nature, combined with the fact that the forces on the limbs
are higher than during the walk. Clinicians observe the symmetry of the gait from left to
right, the motion of the head, withers and croup, and the rhythrrric sound of the two-beat
gait (Pasquini et al. 1995; Adams 1975). Most veterinarians have a very limited
knowledge of the mechanics of the non—walking stepping gaits. Judging the soundness of
the non-walking stepping gaits according to the criteria used at the walk or trot is not
appropriate and may limit the detection of lameness problems specific to the breed.

Even though the walk and non-walking stepping gaits share similar gait

characteristics, the horse performing a non-walking stepping gait protracts and retracts

14

‘A 1.9-.
ltfifutn

96302.“;

1995 l,

Blomec

3°10 \ C"

0'°hr~

a ..‘i‘

;n-{ 'h

Hal

‘f".

the limbs more rapidly so the footfalls occur at a quicker rate than the walk. Therefore,
the horse must change the temporal variables in order to produce the quicker four-beat
gait. The head, neck and croup movements also differ between the walk and the non- -
walking stepping gaits and these differences cause confusion between normal and
pathological gaits. Clinicians rely heavily on the head nodding to indicate forelimb
lameness during lameness evaluations (Speirs 1997). However, a sound gaited horse
performing the fox trot or the running walk will naturally display a head nod (Imus

1995).

Biomechanical Analysis of Gaited Horses

Ground Reaction F orce Analysis. Kinetic analysis describes the forces that cause
movement. In the past several years equine kinetic analysis has focused on the collection
of force plate data to describe the ground reaction force (GRF) pattern of sound horses
and the effects of lameness on gait symmetry.

Each gait has a typical force profile. In the walk, the vertical GRF is higher in the
forelimbs than the hind limbs, which reflects the distribution of the body weight. At a
constant walking velocity, the fore and hind limbs have approximately equal braking and
propulsive impulses, changing from braking to propulsion at 50% of stance in the
forelimb and 45% in the hind limb. During walking vertical force curves for both the fore
and hind limbs demonstrate two peaks: the first is higher in the forelimbs and the second
is higher in the hind limbs (Merkens et al. 1985). The two peaks become more distinct at

higher walking velocities (Khumsap et al. in press).

15

 

firm:
we».

Luau“

' -
n»-
H.

are!

\—

'1." .~

u-..~
-

.’

:1
E’

"1)

'N

During the trot, the vertical GRF curve has a single peak, the amplitude of which
is higher in the forelimb than the hind limb. Both the fore and hind limbs have higher
vertical forces at the trot than the walk. As for the longitudinal forces at the trot, the
braking impulse is larger in the forelimbs, while the hind limbs provide more propulsion
(Merkens er al. 1993).

In the canter, the trailing forelimb has the highest vertical GRF with a magnitude
of 1.5 times the horse's body weight. The trailing fore and trailing hind limbs contribute
more to propulsion than the leading limbs. The diagonal limb pair (leading hind limb and
trailing fore limb) provide more braking, while the leading forelimb has the highest
braking forces (Merkens er al. 1993).

Two-Dimensional Kinematic Analysis. Kinematic analysis provides a graphical
record of the movements of the horse's limbs during locomotion. Hildebrand ( 1965)
originally took on the role of objectively classifying the spectrum of gaits performed by
various species including horses. The temporal characteristics of the symmetrical gaits
were displayed on a graph in which the x-axis represented the percent of stride duration
that each hind foot was on the ground and the y-axis represented percent of stride
duration by which the fore foot footfall followed the ipsilateral hind footfall. Fourty-four
possible limb support sequences of various symmetrical gaits made up the superimposed

grid system from which he proposed a scheme of names for equine symmetrical gaits

(Figure 2.1).

16

 

 

.hvnv..aHU‘-AUI- he .IlIa“ . .-a‘ . I
. {1‘ "1 h=~
9

 

 

 

 

 

 

 

 

 

 

% of stride interval that footfall of forefoot
follows hind on same side

 

 

 

 

 

% of stride interval that each hind foot' rs on the gro'nnd

Figure 2.1. Hildebrand’s graph (1965) demonstrating the distinction between
various gaits according to hind limb stance duration and % of stride interval that
footfall of forefoot follows hind on same side.

Hildebrand's graph was based on data describing the symmetrical gaits of sixty-
eight horses performing the pace, trot, walk, paso gait, slow gait, rack, running walk, and
variations of these basic symmetrical gaits (Figure 2.1). On the vertical axis, the paso
gait, rack, running walk, and slow gait fell within the same region as the walk and mid-
way between the regions occupied by the pace and the trot. On the horizontal axis, the
walk and the paso gait have overlapping areas on the graph, but with the walk having a

relatively longer hind stance duration than the paso gaits.

l7

 

TL“:
Ad‘.

"2‘;

A...

F

f.“
‘/

.-.

(’1)

(fl

Both the trot and pace have a relatively large range of percentage of stride that
each hind limb is on the ground. This range is attributed to the large speed variation
within these gaits; at slower speeds the hind limbs are on the ground longer than at faster
speeds. The stepping gaits represent a wide range of percentages of stride that each hind
limb is on the ground with the running walk having the greatest range. Along the
horizontal axis, areas of overlap between the different stepping gaits represent the fact
that these gaits form a continuum rather than having distinct boundaries between them.
The interval by which the forelimb follows the hind limb on the same side falls within the
range of 20—40% of the stride for the stepping gaits with the walk having the greatest
range. The stepping gaits lie between the trot at 45-60% and the pace at 0-10%
dissociation.

Through graphing the various symmetrical gaits based on temporal
measurements, Hildebrand developed a classification scheme based on these temporal
gait characteristics. He concluded that the boundaries between the gaits are not distinct.
Instead the gaits are more like a continuum, with different gaits being arbitrarily
separated by styles within the breed (Hildebrand 1965). Thesebreed styles play a major
role in the distinction between the symmetrical gaits. The following are examples of
stylistic differences between breeds: the head nod and an extreme overstriding (the hind
limb showing a large amount of over-tracking relative to the forelimb on the same side)
separates the running walk from the rack; the ternrino (lateral motion of the forelimb
during the swing phase) separates the gait of the Peruvian Paso from the Paso Fino

(Hildebrand 1965).

18

 

scram 2;.

5.1"“? m

unkn‘
»

r
v“.“~\ .\
mum

r:
I“

.1

I"

it!

 

The running walk is described as a stepping gait with lateral couplets, which
implies that ground contact of the hind limb precedes that of the forelimb on the same
side by 10-20% of the stride (Hildebrand 1965) and the primary support phase is lateral
bipedal support. If the percentage of lateral support becomes too great, however, the
horse switches to a pace. On the other hand, some horses spend less time in lateral
support and slightly more time in a diagonal support while still technically performing a
running walk by maintaining other gait characteristics such as hind limb overreach and a
rhythmic head nod. If the percentage of diagonal support becomes too great, some horses
may switch to a fox trot, which is not as common in a purebred Tennessee Walker. Slade
(1993) measured sixty-four conformational traits in the Tennessee Walker, Icelandic, and
Walkony horse, then made correlations between the measured traits and the percentage of
time spent in lateral and diagonal support sequences during the running walk. The
Tennessee Walker and Icelandic horses were found to have a more lateral running walk
while the Walkony horses had a more diagonal running walk (Slade 1993). The
following eight conformational characteristics were related to the degree of lateral or
diagonal support of the running walk: rear cannon length, pelvis length, hip to hock
length, neck length, stifle joint angle, hip joint angle, shoulder joint angle, and hock joint
angle.

Just as kinematic analysis assisted in the further understanding of the running
walk and the gait characteristics unique to different breeds performing these gaits, the
mechanics of the fox trot have been explored using high speed cinematography. Clayton
and Bradbury (1995) measured the temporal characteristics, including footfall and limb

support sequences and classified the fox trot as a fast, lateral sequence, four-beat gait

l9

 

‘I‘

..-
,E...
e

b“

$.4\

r.‘

h?
L.

a“

0L-

.‘f

with diagonal couplets. The support sequences consisted of 60.6% of the stride duration
spent in diagonal support, 21.7% in lateral support, and 17.8% in tripedal support.

Along with describing the normal timing of the stepping gaits, kinematic analysis
can detect defects in the gait. Clayton (1995) measured the stride kinematics of the
collected, medium, and extended walks. Only one horse out of the six hi ghly-trained
dressage horses had a regular four-beat rhythm during all of these types of walk. Four of
the horses displayed lateral couplets with irregularities evident more often in the medium
and extended walks. Temporal measurements of the half pirouette at the walk indicated
that horses performing this movement displayed irregularities in the rhythm. These
irregularities were associated with a shorter interval between footfalls of the outside
forelimb and inside hind limb and a longer interval between footfalls of the inside hind
limb and inside forelimb. The absence of forward motion in the half pirouette causes
instability in the horse's balance, and the horse compensates for this by using a longer
tripedal support and substituting a period of quadripedal support for bipedal support
(I-Iodson et al. 1998).

Three-Dimensional Kinematic Analysis. Previous studies of stepping gaits, along
with the majority of other equine kinematic research, has been limited to two-
dimensional analysis in the sagittal plane, which ignores movements in the frontal and
transverse planes. This is not an unreasonable simplification with regards to the normal
motion patterns of the equine limbs. Over millions of years, the horse’s limbs have
evolved in such a way that the joints distal to the shoulder and the hip move primarily in
the sagittal plane, which is energetically efficient. However, small out—of-plane

movements also occur. Herring er al. (1992) described fetlock joint motion in sagittal and

20

frontal planes. By making comparisons between the walk, trot, and canter, it was
determined that the amount of motion in the frontal plane was velocity-dependent.
Although motion outside of the sagittal plane is restricted anatomically, even
simple visual observations can detect out of plane motion in the gaited horse. The
termino of the Peruvian Paso is a good example of this motion. The forelimbs show an

outward rotation during the swing phase that starts at the shoulders and moves down

through the distal joints. Judges score the quality of the termino and during a horse show,

the Peruvian Paso is evaluated from different angles to determine the degree of termino
as the horse performs the paso gaits (Imus 1995).

Three-dimensional kinematic analysis has been applied to various aspects of
human movement through the establishment of specialized laboratories in which patients
are evaluated for clinical diagnosis and assessment of treatment using three-dimensional
kinematic analysis. In this analysis technique, certain prerequisites need to be followed
including the definition of a coordinate system in three-dimensional space. The purpose
of a coordinate system is to define the relative position between two bodies. During
kinematic analysis, a calibration frame is used to orient the movements in relation to the
space in which they are occurring, which is described as a global coordinate system
(GCS).

Grood and Suntay (1983) were the first to introduce an anatomically-based
coordinate system for the description of threeedimensional rotational and translational
motion of the knee joint. Based on it's anatomical definition, this is described as a joint
coordinate system (JCS). Subsequently, Soutas-Little et al. (1987) developed a JCS to

describe the motion of the ankle joint.

21

 

 

 

I 'V‘ “'7“.
Kink-hull

a
l

f

With regard to equine three-dimensional motion analysis, Fredricson and
Drevemo (1971) described the three-dimensional movements of the hoof using markers
attached to a glass-fiber-molded shoe. Their methodology applied aerospace technology
to describe the hoof movements within a GCS, which was related to the local
environment rather than the horse's anatomy. Three-dimensional analysis using a GCS
has also been applied to describe quantitative variables associated with equine carpal
lameness (Peloso et al. 1993).

The results of these studies were described in terms of three, mutually
perpendicular, anatomical planes: the sagittal, transverse and frontal planes. A drawback
to this type of multi-planar analysis is that it cannot measure intemal/extemal rotational
motion, and therefore, we can not determine the relationship between intemal/external
rotation and abduction/adduction or flexion/extension of the joint. Furthermore, since
each plane is measured separately, the relationship between flexion/extension and
adduction/abduction is speculative. The true relationship between these joint motions can
be evaluated only if all of these motions are measured concurrently within the same
anatomically-based coordinate system. A JCS is an anatomical coordinate system based
upon the limb segments comprising the joint. The joint motions are tracked and
calculated within a JCS, which allows for the measurement of intemal/extemal rotation
and the determination of the relationship between joint motions.

Heleski (1991) described a methodology for measuring three-dimensional carpal
and fetlock joint motions using a JCS. The shortcoming to the methodology was
associated with marker placement restrictions. Three-dimensional analysis requires the

use of three, non-collinear markers on each limb segment. Two of these markers are

n
u

a \
.fiu

..9 .o

- .-

L44

. ~

aligned with an anatomical axis of the segment and the third is perpendicular to that axis.
This requirement limits the locations that can be used for placement of the markers and
does not allow the selection of marker sites that fulfill other criteria, such as minimization
of skin displacement relative to underlying bony landmarks. To date the primary
linritation to equine three-dimensional motion analysis using a JCS has been the lack of a
targeting scheme for three-dimensional analysis of equine joint motion that overcomes
the limitations described above without disturbing the natural movement of the horse.

Equine Net Joint Moments and Joint Powers. Kinematic analysis describes the
movements without considering the forces that cause the movement. Kinetic analysis
measures these forces and the resulting mechanical energetics (Winter 1990). During
locomotion, continuously changing internal and external forces during the gait are
characterized by joint moments and energy patterns throughout the stride (Winter and
Robertson 1978). Although the kinematics may be identical, the patterns of energy
production across joints can be very different (Winter 1990). Understanding the
synergistic patterns of these forces gives a more complete picture of the mechanics of the
gait.

The net joint moment indicates the torque produced across a joint by the muscles,
tendons, and ligaments, and is calculated by combining videographic and force data with
morphometric measurements using a link-segment model and an inverse dynamic
solution (Clayton et al. 1998). Certain assumptions are made concerning the link
segment model: 1) the mass of each segment is located at its center of mass, 2) the center
of mass remains fixed, 3) joints are considered to be hinges, 4) each segment's mass

moment of inertia about the center of mass is constant, and 5) each segment's length

23

Asu'.

w. 5‘

.r.
“av-A

0’1

 

J
D

f.

remains constant. The forces acting on a link segment model include gravitational forces,
external forces, and muscle and ligament forces. This model requires the following
information for the calculation of net joint moments: segment positions and joint angular
movements, ground reaction forces and the center of pressure, and segment
morphometric and inertial data (Winter 1990)

Joint power is a measure of the mechanical work done across the joint and is
calculated as the product of the net joint moment and the joint angular velocity. Positive
joint power occurs when the net joint moment and joint angular velocity have the same
polarity, and negative power occurs when they have opposite polarities (Colbome et al.
1997). Positive joint power indicates that power is being generated and negative joint
power indicates power absorption (Colbome et al. 1997). During power generation, the
muscles are performing a concentric contraction, and in power absorption, the muscles
are performing an eccentric contraction (Colbome et al. 1997 ).

Most of the studies on joint moments and powers have been in human subjects.
The findings have helped to differentiate sub-samples of the population, to detect gait
pathologies, to make recommendations for treatment of gait pathologies, and to assess the
outcome of such treatments. For example, Chao et al. (1983) found older adults had a
higher knee flexion moment and a higher planter flexion moment in the ankle than
younger adults. The same age differences may be true for the horse, and if so, age would
need to be a consideration when selecting test subjects. Davids et al. (1996) found
differences in the net joint moments of the hip, knee and ankle between children with and
without cerebral palsy. Contrary to the normal walking pattern, children with cerebral

palsy use the hip rather than the ankle as the primary power generator and the ankle

24

rather than the knee for power absorption. This demonstrates how adaptable the
biomechanics of the gait are to changes in the body. Analysis of gait adaptations due to
disease should be evaluated for the horse.

Analysis of net joint moments and joint powers may give insight into how the
gaited horse maintains a stepping gait at faster speeds when non-gaited horses are
performing a leaping gait. In human subjects, the patterns of energy transfer have been
compared for the slow and fast walk. Peak power at push-off in the plantar flexor
muscles increased from 180 W (Watts) at a slow walk to 490 W at a fast walk. Energy
absorption during the initial deceleration ranged from 30 W at the slow walk to 125 W at
the fast walk (Winter and Robertson 1978).

In horses, net joint moments have been described during the stance phase at the
walk (Colbome et al. 1997; Colbome et al. 1998; Clayton et al. 2000a), the stance phase '
at the trot (Clayton et al. 1998) and the swing phase at the trot (Lanovaz et al. 1999). Net
joint moments and joint powers have been measured to assist in understanding the effects
of musculoskeletal pathologies and therapeutic shoeing. The effects of superficial digital
flexor tendinitis on forelimb net joint moments and joint powers during the stance phase
at the trot have been described (Clayton et al. 2000b) and the response to treatment with a
heel wedge has been assessed (Clayton et al. 2000c). In another study, it was shown that
horses compensated for the weight of shoes by increased energy generation on the flexor
aspect of the elbow during early stance to overcome the additional inertia due to the
weight of the shoes, and increased energy generation on the extensor aspect of the elbow
in late stance to overcome the extra momentum as the limb was swinging forward

(Singleton et al. 2000).

25

 

Show

need it

:1".-
lxcfll
CK£?\

Conch

Differences in energetic patterns across the joints at different walking speeds have
shown that the hip joint is the primary site of increased energy generation as walking
speed increases, assisted by the tarsus (Khumsap et al. 2000). Investigation of how the
sites of power generation differ between the walk and the faster stepping gaits would give
further insight into the mechanics of the non-walking stepping gait. Comparisons of this
nature will elucidate the mechanism for increasing the speed of the gait without
disturbing the footfall patterns. The information presented in this dissertation can be
further applied for analysis of net joint moments and joint powers of the non-walking

stepping gaits.

Conclusions

Over the years, great advances have been made in the study of equine
biomechanics. Kinematic and kinetic studies of the walk and trot are providing
significant insights into the mechanics these gaits. However, most of the evaluation to
date have been two-dimensional analysis, which ignores adduction/abduction and
internal/extemal rotation. Furthermore, the application of kinematic and kinetic analysis
to the gaits of the gaited horse have been limited to analysis of the temporal variables.
This limited analysis may be due to the unique gait characteristics that they exhibited
such as out-of-sagittal plane motion and the lack of kinematic analysis techniques that
can accurately measure these characteristics. The following studies provide a technique
for measuring these unique gait characteristics and preliminary analysis of these gaits that

will serve as a foundation for further investigation.

26

EXPE

C0

 

Sum

[mt

CHAPTER III
EXPERIMENT I: THE APPLICATION OF VIRTUAL MARKERS TO A JOINT
COORDINATE SYSTEM FOR EQUINE THREE-DIMENSIONAL JOINT

MOTION

Summary

Joint motions that occur in planes other than the sagittal plane affect the esthetics
of a horse's movement and may change in response to certain types of locomotor
pathology. Measurement of these motions has been limited due to the complexity of
establishing a targeting scheme that creates a meaningful and repeatable joint coordinate
system. This study presents a targeting scheme that meets the requirements for
developing a joint coordinate system. The targeting scheme was applied to the equine

carpal and fetlock joints, and the results compare favorably with previous equine studies.

Introduction

Although human kinematic research has utilized three-dimensional analysis in
many applications (Grood and Suntay 1983; Soutas-Little et al. 1987; Cappozzo 1991),
equine kinematic research has primarily focused upon two-dimensional motion analysis
(Back et al. 1996; Galisteo et al. 1996; Hodson et al. 2000). Since the anatomy of the
horse and human are so different, adaptation of the methodology for measuring three-
dimensional motion in humans for use in horses is necessary. The objective of this study
is to develop a methodology for measuring equine three-dimensional joint motion

accurately and consistently. This study includes the theoretical and practical

27

A QMN‘;
s0;.).u~

rm: 3:

' .l .3
’11.:th

The '

~11.
h.“

’4‘.’
.d‘“
-
l‘l‘ ‘
h‘ .
r
’h-
..
‘1‘

ill
of”,

(A -'

[Jr

"J

considerations involved in using virtual marker methodology and the establishment of a
joint coordinate system for measuring three—dimensional motion at the equine carpal and

fetlock joints.

The Theory of Three-Dimensional Motion Analysis

Joint Coordinate System. Three-dimensional joint motions can be described by
six independent coordinates comprised of three translations and three rotations. The
majority of joint motion studies have measured the relative rotational motion between the
two segments using Euler angles (Grood and Suntay 1983). The Cardarric/Eulerian
rotation sequences describe three-dimensional joint motion through an ordered sequence
of rotations about the axes of a selected, Cartesian coordinate system. The limitation to
the Cardanic/Eulerian rotational method is that the joint orientation is attained by a
specific temporal sequence of rotations about the selected axes. The primary rotation
about the selected Cartesian coordinate system is followed by a second rotation about
another, displaced coordinate axis, and a third rotation about a doubly displaced
coordinate axis. Any variation in the ordered sequence will result in different values of
the three-dimensional joint angles. For example, if the temporal sequence designates the
first rotation as flexion/extension and this sequence is changed so that flexion/extension
becomes the third rotation, the flexion/extensionangle would be different for the two
sequences even though the actual angle was identical. This limitation has caused
problems when repeating research trials, and in interpreting and comparing clinical data

(W oltring 1994).

28

 

To overcome the restriction of the Cardanic/Eulerian rotation sequence, Grood
and Suntay (1983) proposed a coordinate system that is independent of the order of the
occurrences of the translations and rotations. The three rotational axes consist of two
axes embedded into the limb segments that make up the joint of interest and a third that is
the common perpendicular to the fixed axes of the two limb segments. The orientation of
the third axis is determined by the cross product of the unit base vectors of the fixed axes.
The third axis, which is not fixed to either segment, is described as the floating axis.
Since the two fixed axes are within the segments, the rotations are dependent upon the
choice of the embedded axes. Thus, the rotational sequence is classified as a geometrical
sequence.

In the case of the human knee, the joint coordinate system designates that flexion/
extension occurs about the femoral fixed axis, internal/extemal rotation about the tibial
fixed axis, and adduction/abduction about the floating axis (Grood and Suntay 1983).
Therefore, the angles between the floating axis and the anterior axis of each bone
determine joint flexion and tibial rotation. Joint adduction/abduction is measured from
the angles between the tibial and femoral segmental fixed axes. These three calculated
angles are termed Euler angles.

The first component in measuring equine three-dimensional motion is to adapt the
joint coordinate system developed for human joints and to apply it to the equine joints.
Heleski (1991) utilized the joint coordinate system described by Grood and Suntay
(1983) to measure equine three-dimensional joint motion. The study encountered

difficulties in identifying suitable locations for the markers used to track segmental

29

movements. In this study, we overcame these difficulties by deve10ping a virtual marker
targeting system.

Three-Dimensional Marker Methodology. The marker placement guidelines for
three-dimensional kinematic analysis are that the markers should be located in areas that
are subject to minimal skin displacement, visible to at least two cameras throughout the
stride, and allow the construction of a meaningful and repeatable segmental coordinate
system. In previous studies, the requirements for three-dimensional marker placement
have been restrictive (Heleski 1991). Furthermore, Heleski (1991) attached the 3-D
tracking markers to a strap that was placed around the horse's limbs. This strap was
prone to motion error so that as the limb moved the strap would shift from it's original
location. In order to overcome these limitations while achieving the requirements stated
above, two sets of markers are used to fulfill the marker placement guidelines: virtual
markers establish the segmental coordinate system while tracking markers are used to
follow the segmental movements. The tracking markers are placed on sites that are
readily visible throughout the stride and for which skin displacement is minimal. The
spatial relationship between the two sets of markers is established with the horse in a
standing position. During locomotion, the locations of the virtual markers are calculated
from the tracking marker coordinates using transformations developed from the standing
position.

The development of this type of targeting system assumes that the biological
system being studied is composed of a series of rigid bodies. In other words, each limb
segment is treated as rigid, and movements of the soft tissues are ignored. This is a

reasonable simplification in the distal limbs of the horse. For each of these rigid rods (or

30

limb segments), a segmental coordinate system is created as a reference for movement
measurement. The use of virtual markers is based on a knowledge of the distance and
orientation between the origin of the segmental coordinate system and a defined point on
the segment, together with a known transformation between the global coordinate system
and the segmental coordinate system. When these conditions are fulfilled, then at any
instant in time the same point on the segment can be located in the global coordinate
system (Soutas-Iittle 1996).

The virtual marker locations are subject to the requirements listed earlier for
three-dimensional targeting schemes in that two markers must lie along the anatomical
axis of the segment and the third must be placed perpendicular to that axis. The virtual
markers define the anatomical segmental coordinate system. Their location is recorded
videographically on a standing file, but it is not necessary for the virtual markers to
remain on the limb while the horse is in motion. In addition to the virtual markers, three
tracking markers are attached to each segment at readily visible locations where skin
displacement is minimal. The tracking markers are recorded on the standing file and
while the horse is in motion. They define an arbitrary tracking segmental coordinate
system. A transformation between the tracking segmental coordinate system and the
global coordinate system is calculated from the tracking files. Transformations between
the anatomical and tracking segmental coordinate systems are developed from the
standing file. These transformations are used to calculate the coordinates of the virtual
markers and to locate the virtual markers as if they had been present during locomotion.
The virtual marker system allows compensation for skin displacement and facilitates

tracking in areas of the limb that may be hidden from the camera during locomotion.

31

Transformations. Transformations are used to determine the location of the virtual
markers during locomotion from a knowledge of the position of the tracking markers.
The tracking segmental coordinate system (TSCS) is determined by three tracking
markers fixed to each segment that move with the segment throughout the motion. The
origin of the TSCS is determined (Figure 3.1).

The first axis is calculated from the cross product of the vectors from the origin to
the two other markers (1):

(l) Ii=r1xr2/Ir1xrzl
The second axis is calculated by the cross product of the vector from the origin to the
third marker and the first axis (2):

(2) k=r2xil|r2xi|
Lastly, the third axis is calculated by the cross product of the first and second axis (3):

a) jzkxi
The coefficient of these unit vectors are the directional cosines, it, between the TSCS and
the global coordinate system (GCS):

(4) i = 1,021 + A” 3+ szk

(5) j: Mi + M 3+ 1,212

(6) 1? = Ami + Azyi + 1,112
These directional cosines, 2t, between the TSCS and GCS, expressed in terms of a 3 x 3
rotational transformation matrix, [T], allows for the reference from the TSCS to the GCS.

The translational vector, R0, is drawn from the origin of the GCS to the TSCS.

Vector Rv is drawn between the origin of the GCS to the virtual marker (Figure 3.2).

32

 

Figure 3.1. Three noncollinear markers, A, B, and C, fixed to a rigid body.
The origin, marker A, of the tracking segmental coordinate system and the x, y,
and z axes are indicated as well as the unit vectors i, j, k of the axes,
respectively.

 

Figure 3.2. The virtual vector location is calculated from R, the vector from the
global coordinate system (GCS) origin to the virtual marker D; R0, the vector from the
GCS origin to the tracking segmental coordinate system (TSCS) origin; and P, the
vector from the TSCS origin to the virtual marker D.

33

..\.

‘ .a
i

Q‘.

Vectors from the TSCS origin to the virtual targets are calculated and expressed in terms
of the GCS (7):

(7) P = R, - R0
The virtual vector P is further transformed from the GCS to the TSCS using the rotational
transformation matrix [T] to form the virtual vector p (8):

(8) p = [T] P
Since the segment is assumed to be rigid, the virtual vector expressed in terms of TSCS
will not change regardless of the orientation of the segment. For each frame, the
orientation of the TSCS can be calculated from the tracking markers. The transformation
matrix and the virtual vectors can then be used to transform each virtual vector for that
segment from the TSCS back to the GCS (9):

(9) R.= R0 + m" p
Therefore, for each frame of motion data, the location of the virtual marker is expressed
in terms of the GCS as if it had actually been tracked.

Euler Angles. After the transformations have been calculated and the segmental
coordinate systems have been established, the joint coordinate system is determined for
the measurement of three-dimensional joint motion. The joint angles can be measured
from the embedded segmental coordinate system determined by the virtual vectors. The
axes of the embedded segmental coordinate systems are denoted as the following: 32,, 3",,
and 2, for the radius; it... 9..., and gm for the metacarpus; and 2p, 9p, and 2,, for the proximal
phalanx. For each segment, the z-axis runs through the long axis of the segment, the x-
axis runs medial/lateral, and the y-axis runs dorsal/palmar. At the carpal joint, for

example, the xr axis of the radius rs desrgnated as the flexron/extensron axrs, ’61, and the

34

:1

1‘;

l“)

Z. 'l

2m axis of the third metacarpus is the intemal/extemal rotation axis, ’63. The floating axis,

the adduction/abduction axis, is calculated by the cross product of the e] and $3 axes, 222

(Figure 3.3):
(1) $1 = x, Flexion and Extension
(2) €33 = 2m Internal and External Rotation
(3) $2 = (8, x 83) / | 8, x e, | Adduction and Abduction

The joint Euler angles are (Figure 3.3):

(4) Flexion (-) / Extension (+) = -asin (cl; ’ z?)

(5) Adduction () / Abduction (+) = asin ($1 '63)

(6) Internal (-) / External Rotation (+) = asin ($2 ' 52m)

The joint angles are defined so that joint motion is described relative to their respective
axes in that flexion/extension is relative to $1, intemal/extemal rotation to $3, and
adduction/abduction to 82.

At the fetlock joint, the in, axis of the third metacarpus is designated as the
flexion/extension axis, 31, and the 2,, axis of the proximal phalanx is the intemal/extemal
rotational axis, ’63. The floating axis, the adduction/abduction axis, is calculated by the
cross product of the 81 and $3 axes, 32 (Figure 3.4). Joint motion is measured using the
same calculations as the carpal joint with 2, replaced with 2m and it... replaced with it,
The above calculations follows that of Soutas-Little et al. (1987) in which the coordinate
system is rotated 90° from the coordinate system used in Grood and Suntay (1983) so that
the relationship between the two approaches can be expressed as follows:

sin or = -cos(0t+90). This approach is more applicable for clinical evaluation.

35

Adduction(-)/Abduction(+)

 

Flexion(-)/Extension(+)

 

1 i Internal(-)/External(+) Rotation

 

Figure 3.3. Caudolateral view of an exploded limb showing the radius, carpus,
and metacarpal segements to illustrate the segmental axes used to develop the
joint coordinate system. The orientations and locations of the flexion/extension
(e1), adduction/abduction (e2), and intemal/external rotation (e3) axes are shown.

   

Adduction(-)/Abduction(+)

Flexion(-)/Extension(+)

Internal(-)/External(+) Rotation

 

Figure 3.4. Caudolateral view of an exploded limb showing the metacarpal and
proximal phalangeal segements to illustrate the segmental axes used to develop
the joint coordinate system. The orientations and locations of the flexion/
extension (e1), adduction/abduction (e2), and intemal/extemal rotation (e3) axes
are shown.

36

Application of the Joint Coordinate System and Virtual Markers

Subject. The methodology for measuring three-dimensional joint motion was
developed at the walk using a non-gaited horse that was determined to be clinically
sound. The results were compared with previously published data of joint angles
calculated using well—established techniques for two-dimensional analysis and with joint
angles calculated from three-dimensional analysis.

Markers. To track the three-dimensional motion of the equine carpal joint using a
joint coordinate system and virtual marker methodology, two sets of markers were
applied to the antebrachial, metacarpal, and proximal phalangeal segments of the right
forelimb. The virtual and tracking markers consisted of polystyrene hemispheres covered
with retro-reflective tape.

Three virtual markers were attached to a rigid “L” structure (Figure 3.5). A virtual
marker "L" structure was taped securely to the skin overlying the proximal radius, the
distal third metacarpus, and along the length of the proximal phalanx of the right forelimb
with the long axis of the "L" parallel to the anatomical axis of each segment (Figure 3.5).
Three tracking markers were attached to the antebrachial, metacarpal, and proximal
phalangeal segments of the right forelimb using super glue (Future Glue, Pacer
Technology, Ranch Cucamonga, California).

The tracking markers were placed over palpable anatomical landmarks on the
bones where skin displacement is minimal or for which correction algorithms are
available. The antebrachial tracking markers were placed on the distal end of the
segment in a triangular formation (Figure 3.5). The metacarpal tracking markers were

placed in a similar arrangement on the proximal end of the segment (Figure 3.5). On the

37

 

 

Figure 3.5. Location of the virtual markers (black spheres), tracking markers
(gray spheres) and markers (clear spheres) used as both virtual and tracking
markers during the standing file to establish virtual marker locations relative
to tracking markers. Picture shows the virtual "L" structure on the proximal
phalanx.

38

proximal phalanx, the locations of the two virtual markers aligned along the long axis of
the segment were used for two of the tracking markers and the third tracking marker was
placed on the dorsal aspect of the segment to form a triangular formation (Figure 3.5).

Video Recording and Analysis. Two 60 Hz CCD camcorders attached to two
Super VHS tapedecks (Panasonic AG-450, Matsushita Electric Corp., Secausus, NJ) were
oriented at 60° to the horse’s plane of motion, separated from each other by a distance of
609 cm, and located at a distance of 500 cm from, and perpendicular to, the line of
motion. A SOD-watt lamp was placed behind each camcorder to illuminate the retro-
reflective markers during recording. The rectangular calibration volume measuring 342
cm x 190 cm x 210 cm was defined by thirty points.

A three-second standing file was recorded with the horse in a normal standing
position with both the tracking markers and the virtual marker structures in place. The
virtual marker structures were then removed, except for the two markers on the proximal
phalanx that were also used as tracking markers. The horse was recorded on videotape as
it moved at a walk with only the tracking markers attached.

The videotapes were digitized with the aid of the Ariel Performance Analysis
System (Ariel Dynamics, Trabuco Canyon, California) and analyzed using custom
software. For the standing file, the positions of the tracking and virtual markers were
digitized manually. For the walking sequence, the tracking markers were digitized
automatically in both camcorder views. Using the calibration frame previously described
and the direct linear transformation method (Abdel-Aziz and Karara 1971), the three-
dimensional coordinate values of the digitized markers were calculated. The raw data

were filtered using a fourth order Butterworth digital filter, with a cut off frequency of 6

39

limit

~44

due.
’1 _
.~_

RES:

a...

l‘\‘ k.

Hz. The GCS was determined by the calibration frame. Using the digitized locations of
the virtual and tracking markers, the segmental coordinate system was determined and
tracked through the stride using the previously described transformations. The joint
coordinate system and the three-dimensional joint motion were calculated and measured
using the previously described Euler angle methodology. The joint angles measured
during the standing file were expressed as the zero value and the angles recorded during
walking were measured relative to the standing angle.

Statistics. Four strides of the walk were used to measure the three-dimensional
joint motion of the carpus and the fetlock. Graphs were plotted for flexion/extension,
adduction/abduction, and intemal/external rotation of each joint. Peak values for each

type of motion were determined and means and standard deviations were calculated.

Results

Carpal Joint Motion. During the stance phase, the carpus maintained an angle
equivalent to standing angles, then showed a single flexion cycle during the swing phase.
Peak flexion angle was ~65° i 2.6° at 80% of the stride. The small standard deviation
(Figure 3.6) indicates that there was little variability between strides.

The carpus was adducted during early and late stance and reached peak adduction
(-36° :1: 64°) at 85% of the stride (Figure 3.7). The carpus showed very little rotational
motion, with peak value for internal rotation (-6° i 3.2°) at nridswing (Figure 3.8).

The three joint motions appeared to be coupled. During rrridswing, the carpus
flexed, adducted, and internally rotated. Towards the end of the swing phase, the carpus

approached the standing joint angles in preparation for the start of stance.

40

4— Stance Phase ——>l1—Swing Phase—p

 

 

 

 

 

 

 

0 20 4O 60 80 100
Stride (%)
Figure 3.6. Mean (dark line) i 1 SD. (light line) walking carpal flexion

(-)/extension (+) angle relative to the standing carpal flexion joint
angle.

‘__ Stance Phase —->|1-Swing Phase—p

 

 

 

 

 

 

0 20 40 60 80 100
Stride (%)

Figure 3.7. Mean (dark line) :1; 1 SD. (light line) walking carpal
adduction (-)/abduction (+) angle relative to the standing carpal
adduction/abduction joint angle.

41

<— Stance Phase ——> H—Swing Phase—p

 

 

 

 

 

 

0 20 40 60 80 100
Stride (%)
Figure 3.8. Mean (dark line) i 1 SD. (light line) walking carpal internal

(-)/extemal (+) rotation angle relative to the standing axial rotation joint
angle.

4— Stance Phase —>H—Swing Phase—p

20'r
0- A-

-20 4

 

 

 

 

 

 

0 20 40 60 80 100
Stride (%)
Figure 3.9. Mean (dark line) i 1 SD. (light line) walking fetlock

flexion (-)/extension (+) angle relative to the standing carpal flexion
joint angle.

42

 

 

 

um-

0?ng

 

Um

HQ?

103

FOL

F etlock Joint Motion. During the stance phase, the fetlock extended, abducted,
and internally rotated. Flexion/extension showed 50° range of motion at the fetlock joint
(Figure 3.9). Peak extension (19° i 4.l°) occurred in midstance. During breakover the
fetlock flexed and showed two flexion peaks during the swing phase.

The fetlock showed a maximal abduction of 5° 1; 84° around the time of
midstance, and was then adducted during late stance and throughout the swing phase
(Figure 3.10). The fetlock had 10° range of adduction/abduction joint motion.

The fetlock was internally rotated by about 15° during most of stance (Figure
3.11). During breakover, the fetlock showed a small peak of external rotation, then was
internally rotated through the remainder of swing. The fetlock had 20° range of

internal/extemal rotational motion.

Discussion

The joint coordinate system established in this study allowed a meaningful
description of three-dimensional joint motion for the equine carpus and fetlock.
Heleski (1991) first proposed the application of a joint coordinate system to the equine
carpal and fetlock joints and performed some preliminary measurements at the walk and
trot. Marker visibility and difficulty in tracking the markers throughout the stride were '
identified as problems in the methodology. This study addresses these problems by
developing a virtual tracking system.

To date, Heleski's study of two young Arabian horses is the only comparable
three-dimensional study using a JCS to measure carpal and fetlock joint motions of the

walk and trot. In comparing the overall joint motion trends reported by Heleski with

43

<—— Stance Phase ——flk—Swing Phase—b

 

 

 

 

 

 

0 20 40 60 80 100
Stride (%)
Figure 3.10. Mean (dark line) i 1 SD. (light line) walking fetlock

adduction (-)/abduction (+) angle relative to the standing fetlock
adduction/abduction joint angle.

{-— Stance Phase —D<—Swing Phase—p

 

 

 

 

 

 

 

0 20 40 60 80 100
Stride (%)
Figure 3.11. Mean (dark line) i 1 SD. (light line) walking fetlock

internal (-)/ external (+) rotation angle relative to the standing fetlock
axial rotation joint angle.

lose ‘.

 

amlw;

’ in“?

$1.72.:

w. I‘ be

gm» L
u

that

 

“lash,

those found in the present study, the motion patterns were quite similar and the peaks
occurred at approximately the same times in the stride. Fetlock flexion/extension had a
similar range of motion in both studies. Carpal flexion/extension range of motion was 17°
greater in the two horses measured in Heleski's study, but there was 20° less carpal
adduction than was measured in this study. The subject used in this study was also an
Arabian, with the only obvious difference being that it was much older than Heleski's
subjects. To date, no studies have measured the effect of age on equine three-
dimensional joint motion. Fetlock adduction/abduction was very inconsistent from stride
to stride in Heleski's study which was ascribed to problems with added motion of the
straps to which the tracking markers were attached and difficulties in marker visibility.
The smaller variability in the present study was a consequence of the improved accuracy
resulting from the use of virtual markers. Both carpal and fetlock intemal/extemal
rotational ranges of motion and timing of the peaks were similar in the two studies.

In the following chapters, the virtual marker “L” structure was replaced with
virtual markers attached directly onto the skin with some of the virtual markers also used
as tracking markers. The changes in virtual marker methodology were described in
greater detail within the respective chapters. The adjustments in methodology reduced
the number of markers, which facilitated the clear distinction between markers, especially
on the distal aspect of the limb, and reduced the time taken for marker placement. The
new placement of the markers was used in chapter IV (Experiment 11) to allow for
simultaneous collection of two-dimensional data for comparison of the flexion/extension

values. Although the virtual marker placement was slightly different, the established

45

1135 C

CO:

 

 

 

 

,7

1

(Ti

r”

axes of the joint coordinate system were the same so that joint motion was measured
consistently throughout all three experiments.

A three-dimensional study of limb motions in the sagittal and frontal planes was
performed using a multi-planar analysis (MPA) instead of a JCS (Hening et al. 1992).
Since that study was performed at the trot and adduction/abduction angles were measured
for the entire limb rather than the individual joints, it is not appropriate to make direct
comparisons with this study. In the following chapter, the relationship between a JCS
and MPA was further evaluated to facilitate the most appropriate methodology for
analysis of equine locomotion, specifically the gaits of the gaited horse.

Two-dimensional studies have reported flexion/extension patterns and peak
angles that are similar to this study. In Dutch warmblood horses, Back et al. (1996) found
that the swing phase started at 60% of the stride with the peak carpal flexion angle of 60°
occurring at mid-swing at 80% of the stride. Hodson et al. (2000) found that peak carpal
flexion of 55° occurred at 75% of the stride in horses of mixed breeds.

Previous two-dimensional studies found similar fetlock flexion peaks during the
swing phase (Back et al. 1995; Hodson et al. 2000). However, during the stance phase,
Back et al. (1996) and Galisteo et al. (1996) found two extension peaks while this study
along with Hodson et al. (2000) showed only one distinct peak. This difference is
attributed to velocity. The studies by Back et al. (1996) and Galisteo et al. (1996) were
performed at faster velocities (1.6 to 1.7 m/s) than Hodson et al. (2000) and this study
(1.5 m/s). The slower velocity is associated with a reduced rate of limb loading which, in
turn, decreases the rebound effect that is responsible for the slight flexion during mid-

stance (Khumsap et al. in press). Thus, the two peaks of fetlock extension become more

46

distinct as velocity of the walk increases. The flexion/extension range of motion of the
fetlock in this study was similar to Hodson et al. (2000). Both studies were conducted in
the same laboratory and performed under similar experimental conditions, including
velocity.

Since there have been so few studies of equine three-dimensional analysis using a
JCS, comparisons with the present study are quite limited. The effects of variables such
as age, velocity, and footing on walking kinematics have not even been studied using
two-dimensional or multi-planar analysis. Now that the technique has been developed,
these variables should be studied using a JCS to evaluate the effects in three dimensions.
Just as human three-dimensional analysis has been applied to clinical settings, it is the
recommendation of this researcher that further studies should also be conducted to test
the usefulness of this methodology for clinical applications in horses.

In conclusion, a technique for measuring three-dimensional joint motion in a joint
coordinate system using virtual markers was developed. A joint coordinate system
measures joint motion within an anatomically-based coordinate system compared to
multi-planar analysis, which measures joint angles using planar projections. Measuring
three-dimensional joint motion using multi-planar analysis does not allow for the
measurement of axial rotation and coupled joint motion, which was found in this study
with the application of a joint coordinate system to a horse walking. A more detailed
comparison of multi-planar analysis with a joint coordinate system is made in chapter
four. The technique described in this study was applied to a horse walking to measure
carpal and fetlock flexion/extension, adduction/abduction, and internal/extemal rotation.

Comparison with an earlier three-dimensional study using a joint coordinate system

47

(Heleski 1991) indicated that the virtual marker methodology showed less variability
from one stride to the next. Flexion/extension angles were similar to previous two-
dimensional kinematic studies (Back et al. 1996; Hodson et al. 2000). The joint
coordinate system using a virtual marker methodology described in this chapter was
applied to the following chapters with slight alterations made, which are described in the

respective chapters, to continue to improve upon the methodology.

48

Sum]

e..-
30.x

v~-Q,31
5‘35»;

' .50.
‘ 1.1.

I
F.)

. .a' -,
Eye's:-

CHAPTER IV
EXPERIMENT II: COMPARISON OF A JOINT COORDINATE SYSTEM

VERSUS MULTI-PLANAR ANALYSIS

Summary

A kinematic multi-planar analysis (MPA) reduces a subject's three-dimensional
motion to two-dimensional projections onto planes defined by a global fixed coordinate
system (GCS). An alternative to this kinematic method is a joint coordinate system
(JCS), which describes the three-dimensional orientation of the segments comprising the
joint with respect to each other. This study compares carpal and fetlock joint kinematics
measured using a MPA with those measured using a JCS. A Peruvian Paso was recorded
during three walking trials using 60 Hz video camcorders. Skin markers tracked the
movements and defined the anatomical axes of antebrachial, metacarpal, and proximal
phalangeal segments. A JCS was established between the two segments comprising the
carpal and fetlock joints to measure flexion/extension, intemal/extemal rotation, and
adduction/abduction at each joint. The MPA model used two markers aligned on the long
axis of each segment and measured flexion/extension angles projected onto the sagittal
plane of the coordinate system and adduction/abduction angles projected onto the frontal
plane of the coordinate system. Carpal and fetlock flexion/extension angles for the walk
were sinrilar for the JCS and MPA indicating that sagittal plane analysis using a MPA is
adequate when flexion and extension are the only measurements being made provided the
horse’s plane of motion is aligned with the plane of calibration. There were large

differences in carpal and fetlock adduction/abduction angles measured using a MPA

49

nq‘F'

5LAnnf

firm}
.-

.V,b\

2625:

Intro

b). :15
A v

.1: (1:1
‘I

compared with a JCS. Analysis of the reasons for these differences indicated that the
accuracy of frontal plane analysis to measure adduction/abduction is limited by it’s
inability to correct for out-of-plane rotations along the long axis of the segments
comprising the joint. This finding clearly illustrates the inherent errors in using planar
projections to measure multi-planar motions. It is recommended that a JCS be used to

measure joint motions other than flexion/extension in the sagittal plane.

Introduction

Most of the studies of equine locomotion have measured movements of the limb
segments and joints of the horse relative to a global coordinate system (GCS) that is fixed
in space and oriented along the direction of travel of the subject. Often, segments are
defined by two points placed along the long axis of the segment with the relative angles
between two adjacent segments defining the joint angles. The three-dimensional
coordinates of the points are projected onto the two-dimensional sagittal and frontal
planes to give a multi-planar analysis (MPA) of segmental and joint motions. MPA has
been used to describe vertical displacement, flexion/extension, and adduction/abduction
of the fetlock joint (Hening et al. 1992) and the stifle and tarsal joints (Back et al. 2000).
A limitation to MPA occurs when the segmental motion is not parallel to the projection
plane causing the angular measurements in that plane to be distorted.

An alternative to MPA is to use an anatomically-based coordinate system
established on each rigid segment. The joint angle can then be expressed as relative
rotations between two adjacent segments. One method used to accomplish this is known

as a joint coordinate system (JCS) (Grood and Suntay 1983). Heleski (1991) used a JCS

50

 

V"

v

 

we.

.7?

to describe three-dimensional carpal and fetlock joint angles of walking and trotting
horses. The results were expressed in terms of flexion/extension, adduction/abduction,
and intemal/external rotation. It was concluded that the three-dimensional marker
placement requirements were restrictive and prone to marker motion error. In order to
overcome these limitations, a virtual marker methodology has been developed and used
to establish a JCS for the carpus and fetlock (Chapter 1]], Experiment I).

The objectives are to compare carpal and fetlock joint kinematics measured

simultaneously using a JCS with those from MPA in the sagittal and frontal planes.

Materials and Methods

Subject. The subject was a sound Peruvian Paso ridden at the walk. This breed
was chosen because it is noted for a characteristic outward motion of the forelimbs during
the swing phase of all it's gaits, which is known as a termino. Limb motion during
termino offers an ideal opportunity to study exaggerated three-dimensional limb motion.

JCS Segmental Axes. Coordinate systems for the antebrachial, metacarpal, and
proximal phalangeal segments were established as described in chapter three. The origin
of the antebrachial coordinate system was located on the lateral styloid process on the
distal radius. The antebrachial x-axis was directed laterally from the origin perpendicular
to the antebrachial sagittal plane and on the same vector as a line from the medial to
lateral styloid processes. The antebrachial z-axis, contained in the antebrachial frontal
plane, ran from the origin proximally along the long axis of the segment at 90° to the x-
axis. The antebrachial y-axis ran dorsally from the origin, perpendicular to the

antebrachial frontal plane.

51

”It 6 . \ w
.lk n h . WI“
Id 1

fr

. .. ... C
. h ~$ a k»
T. ex ..
..a LL A.»

The origin of the metacarpal coordinate system was located on the lateral aspect
of the distal end of the third metacarpus. The metacarpal x-axis was directed laterally
from the origin perpendicular to the metacarpal sagittal plane and on the same vector as a
line from the media] to lateral distal aspect of the third metacarpus. The metacarpal z-
axis, contained in the metacarpal frontal plane, ran from the origin proximally along the
long axis of the segment at 90° to the x-axis. The metacarpal y-axis ran dorsally from the
origin, perpendicular to the metacarpal frontal plane.

The origin of the proximal phalangeal segment was located on the lateral aspect
of the distal end of the proximal phalanx. The z—axis of the proximal phalangeal segment
ran from the origin proximally along the long axis of the segment. Similar to the other
segmental axes, the proximal phalangeal x and y-axes ran laterally and dorsally from the
origin, respectively.

MPA Segmental Axes. The axes of the antebrachial, metacarpal, and proximal
phalangeal segments used for the MPA corresponded to the z-axes of the segments as
defined by the JCS. Therefore, the MPA segmental axes passed along the long axes of
the antebrachial, metacarpal, and proximal phalangeal segments.

Data Collection. Kinematic data were collected at 60 Hz using four Super VHS
camcorders (Panasonic AG-450, Matsushita Electric Corp., Secausus, NJ), arranged
around the data collection area in increments of 60°. One camcorder was located towards
the front of the horse, one towards the back of the horse and two towards the lateral
aspect of the right side of the horse. A 500-watt lamp was placed behind each camcorder
to illuminate the retroreflective markers during recording. The calibration volume (342 x

190 x 210 cm3) was defined using 30 equally spaced points.

52

The horse was ridden along the runway at a walk. A trial was considered
acceptable if the horse traveled at a consistent velocity (2.50-2.56 m/s) along a straight
line that was parallel to the plane of calibration with all markers being clearly visible in
the two lateral camcorder views (Figure 4.1).

Marker Placement. Reflective markers were attached to the skin overlying well-
defined bony landmarks on the proximal and distal ends of the antebrachial, metacarpal,
and proximal phalangeal segments on the lateral aspect of the limb (Figure 4.2). The
tracking markers were located at sites that have been shown to have relatively little skin
displacement during the walk (van Weeren et al. 1988). Corrections for skin
displacement were not performed since 3-D correction algorithms and even 2-D
corrections are not available for the Peruvian Paso breed or for Peruvian gaits. Current 2-
D correction algorithms are based on Dutch Warmbloods and are very dependent on the ‘
gaits being measured (van Weeren et al. 1988). As for this study, the same markers were

used for the two methodologies so any errors due to skin displacement were the same.

 

Figure 4.1. Data collection with tracking markers attached to the right forelimb
during a walking trial.

53

 

 

   

 

 

 

 

 

 

Caudal Cranial Lateral Medial
.Y I
l i
Tl
Lateral View Frontal View

Figure 4.2. Marker placement during the standing file as seen in the lateral (left)
and frontal (right) views. Virtual markers on the lateral aspect of the limb (black
circles) for establishing the joint coordinate systems (JCS) were placed over bony
landmarks on the proximal and distal ends of the antebrachial, metacarpal, and
proximal phalangeal segments. A third virtual marker (open circles) was placed on
the medial aspect of the distal extrenrity of each segment. Two of the JCS tracking
markers (black circles) were the same as the lateral virtual markers. The third
tracking marker (gray circles) was placed on the dorsal surface, midway between
the other two tracking markers. The tracking markers (black circles) on the
proximal and distal ends of each segment were used for multi-planar analysis
(MPA). The virtual markers on the medial aspect were removed during the walking
trials.

54

Segmental coordinate systems for the JCS were defined by the two lateral markers
used for MPA together with a third marker attached on the medial styloid process of the
radius, the metacarpal attachment of the medial collateral ligament of the fetlock, and the
medial aspect of the distal end of the proximal phalanx. The medial markers were tracked
using a virtual targeting system as described in Experiment I (Chapter III; Soutas-Little
1996) utilizing a fourth marker located on the cranial aspect of each segment, midway
between the two lateral markers.

Two reflective markers were placed along the spine of the horse to deternrine the
horSe's angle of travel relative to the GCS during the trials. This information was used to
align the horse's plane of motion with the GCS using a simple 2-D rotational

transformation (Figure 4.3).

[cos(9) -sin(®):l x0 =[x]
sin(®) cos(@) 2., z

 

Figure 4.3. Calculation used to correct sagittal plane measurements using MPA
for the horse's angle of travel determined from the back markers (black spheres),

where G is the angle that the horse is out-of-plane, x0 and 20 are the uncorrected
marker coordinates, and x and z are the corrected marker coordinates.

55

Data Reduction. The skin markers were automatically tracked and digitized using .
a video analysis system (Ariel Performance Analysis System, Ariel Dynamics Inc,
Trabuco Canyon, California). Three-dimensional locations of markers were obtained
using direct linear transformation (Abdel-Aziz and Karara 1971). The three—dimensional
coordinates were filtered using a fourth-order Butterworth digital filter at a cutoff
frequency of 6 Hz. Joint motionhin the MPA was described for the distal segment relative
to the proximal segment (Winter 1990). The MPA was used to measure flexion/
extension in the sagittal plane and adduction/abduction in the frontal plane. The JCS was
established based on the system described by Grood and Suntay (1983). For the carpal
joint, flexion/extension was measured around the antebrachial fixed x-axis,
internal/external rotation was measured around the metacarpal fixed z-axis, and
adduction/abduction was measured around a floating axis perpendicular to both the
flexion/extension axis and the intemal/external rotation axis. For the fetlock joint,
flexion/extension was measured around the metacarpal fixed x-axis, internal/extemal
rotation was measured around the proximal phalangeal fixed z-axis, and
adduction/abduction was measured around a floating axis perpendicular to both the
flexion/extension axis and the internal/external axis. For example, flexion occurred when
the angle on the caudal/palmar aspect of the joint decreased, internal rotation occurred '
when the distal segment rotated medially relative to the proximal segment, and adduction
occurred when the angle on the medial aspect of the floating axis decreased. Zero joint
angle was defined as alignment of the proximal and distal segments. Flexion, adduction,
and internal rotation were assigned negative values. Extension, abduction, and external

rotation were assigned positive values. For carpal flexion/extension angles greater than

56

90°, an additional calculation was applied to overcome errors arising from taking the sine
of an angle greater than 90° in the JCS calculations (Chapter III, Experiment I).

The mean curves for flexion/extension, abduction/adduction, and intemal/extemal
rotation of the carpal and fetlock joints at the walk measured by the JCS were graphed.
The mean absolute difference between the adduction/abduction measured with the JCS
and MPA were calculated. Segmental angles were determined in the sagittal view using
the raw x coordinates for the proximal and distal markers that were aligned along the
long axis. Segmental angles in the sagittal plane were measured from an axis
perpendicular to the ground with the segment rotating around it's proximal end. An
angle of 0° indicated the segment was vertical and an angle of 90° indicated the segment
was horizontal. Protraction (cranial rotation of the distal end of the segment) was
assigned a positive value and retraction (caudal rotation of the distal end of the segment)

was assigned a negative value (Figure 4.4).

Caudal

  

Retraction (-)

 

 

Angle = 0°

Figure 4.4. Measurement of the segmental angles in the sagittal plane.
Negative values indicate retraction, positive values indicate protraction, and
0° indicates the segment is vertical.

57

The anatomical orientation and the location of the segmental axes relative to the
GCS was explored in an attempt to understand the differences between the two methods.
The mean difference between the raw 2 coordinates of the proximal and distal markers of
the segmental long axis was calculated. A difference of zero indicates that the segment is
aligned along the x-axis of the GCS (longitudinal axis of the horse). When 2 coordinates
of the proximal and distal tracking markers were aligned with the corresponding axis of
the GCS, only the sagittal view showed an angulation of the segment relative to the GCS.
The mean angle between the JCS adduction/abduction axes and the x-axis of the GCS
(longitudinal axis of the horse) and between the JCS flexion/extension axes and the z-axis

of the GCS (mediolateral axis of the horse) were calculated.

Results

The following section describes the joint motion of the Peruvian Paso walk using
the JCS and differences between the JCS and MPA measurements. The orientations of
the joints and the proximal and distal segments comprising the joints are described at the
times when there are marked differences between the two methods. Values for all
individual trial measurements were graphed and included in Appendix A.

Carpal Joint. JCS standing angles showed that conformation of the right forelimb
was such that the carpus was extended 2°, abducted 8°, and internally rotated -7°. JCS
measurements showed that during the stance phase the carpus was extended, abducted,
and internally rotated (Figure 4.5). Breakover began at 40% of the stride and lift off

occurred at 55% of the stride. Throughout breakover and early swing, the carpus showed

58

flexion, abduction, and external rotation. At mid swing, the carpus reached peak flexion

(-94°), abduction (23°), and external rotation (7°).

Flexion/extension patterns were sinrilar when measured using the MPA and JCS,
and the peak flexion angles, which occurred during mid-swing showed rrrinimal
differences between the two techniques (Figure 4.6). The adduction/abduction curves
showed differences in the general pattern and timing of peaks (Figure 4.7) with the peak
values demonstrating large differences between the MPA and JCS during the transition
from stance to swing (Figure 4.8). Peak absolute difference (16°) between JCS and MPA
adduction/abduction angles occurred around the time of lift off at 53% of the stride
(Figure 4.8). At this point in the stride, the carpus was flexing and externally rotating
(Figure 4.9). The radial segment was vertical in the GCS (sagittal plane angle 0°) at 53%
of the stride (Figure 4.9). The difference in the z coordinates between the proximal and
distal markers along the radial long axis was zero at 53% of the stride indicating the long

axis was aligned along the x-axis of the GCS (Figure 4.9).

F etlock Joint. JCS standing angles showed that the fetlock was extended 18°,
adducted -11°, and internally rotated ~13°. JCS measurements showed that during the
stance phase the fetlock was extended and adducted (Figure 4.10). During breakover,
which began at 40% of the stride, the fetlock flexed, abducted, and internally rotated.
The direction of rotation changed just before lift off. Peak flexion (41°), abduction (2°),
and external rotation (3°) occurred in early swing. During late swing, the fetlock

extended, adducted, and internally rotated.

59

<-— Stance Phase —-> {— Swing Phase -—->
40 *

20 r :

 

 

— f
a. aaaaa
o - ~
'..,‘ .. .I - _ _. u.-
-20 -

.40 -
-60 .
-so 1

-100

 

Angle (°)

 

 

 

 

O 20 4O 6O 80 100
Stride (%)

Figure 4.5. Mean carpal flexion (-)/extension (+) (black line), adduction (-)/abduction
(+) (dark gay line), internal (-)/external (+) rotation (light gay line) using JCS. '

4— Stance Phase —> d—Swing Phase —>

 

 

 

 

Angle (°)

 

 

 

 

0 20 40 60 80 100
Stride (%)

Figm'e 4.6. Mean carpal flexion (-)/extension (+) for JCS (black line) and MPA
(any line).

60

 

 

 

<— Stance Phase —> <— Swing Phase —>
40
C 20 '
2 {
M H ‘-
E
< O I I I I
-20

 

 

 

 

0 20 40 6O 80 100
Stride (%)

Figure 4.7. Mean carpal adduction (-)/ abduction (+) for JCS (black line) and
MPA (gray line).

4— Stance Phase —->|<—Swing Phase —>

 

 

 

4.

O I I I

 

 

 

 

0 20 40 60 80 100
Stride (%)

Figure 4.8. Absolute difference between carpal adduction (-)/abduction (+)
measured using the JCS and MPA.

61

 

 

 

 

 

 

Angle (°)

 

 

 

 

 

 

 

 

 

 

 

 

' "é‘ 10 ‘ 0cm=No wntalZDisplacment

3 A Lateral

a-I /'

=

o

E 0 —

8 ' \ ’
'5. Medial
.2 .16 a n a n n n n a a
G o 20 4o 60 so 100

Stride (%)

Figure 4.9. Top graph: Carpal flexion (-)/extension (+) (solid line) and internal (-)
lextemal (+) rotation (dotted line) measured using the JCS. Middle graph: Segnental
angles in GCS sagittal plane for antebrachium (solid line) and metacarpus (dotted line).
Bottom graph: Z displacement between the proximal and distal tracking markers for
the antebrachium (solid line) and metacarpus (dotted line). The thick vertical line at

53% of the stride is the time when carpal adduction/abduction angles measured with the
JCS and MPA have the maximal difference.

62

(— Stance Phase —> d—Swing Phase —>

 

80
60
40

 

{Pia

_. A
O r I” 17 ‘Ca -.w-“- ' . _%‘ ‘

0 ‘15 -. N“: ‘

g" -20 - I. '6'"

0 20 4o . 6O 80 100
Stride (%)

Figure 4.10. Mean fetlock flexion (-)/extension (+) (black line), adduction (-)/
abduction (+) (dark gay line), internal (-)/external (+) rotation (light gay line)
using JCS.

 

 

<— Stance Phase —> ng Phase—9

 

 

 

 

 

 

 

0 20 40 6O 80 100
% Stride

Figure 4.11. Mean fetlock flexion (-)/extension (+) for JCS (black line) and MPA
(gray line)-

63

4— Stance Phase —> 4-—Swing Phase—>

 

 

 

 

Angle (°)

 

 

 

 

 

O 20 40 60 80 100
Stride (%)

Figure 4.12. Mean fetlock adduction (-)/abduction (+) for JCS (black line) and
MPA (gay line).

4— Stance Phase —i> 4—Swing Phase—b
140

120 J fl
100 4

 

 

 

 

 

Angle (°)
.h.
C

 

 

 

0 20 4O 60 80 100
Stride (%)

Figure 4.13. Absolute difference between fetlock adduction (-)/ abduction (+)
measured using the JCS and MPA.

 

1% I *I I I _l I I' 1

 

 

6.. so- .
E ....
é” 0, ' _ ‘V‘v'wm .

 

.50. _l_ _L 1
0° = Vertical Segmental Angle -90° = Horizontal Segmental Angle

rsol . .
30 M i

   
   

 

 

 

 

 

Angle (°)
in
e:
fl

 

 

 

 

 

 

Displacement (cm)

Stride (%)

Figure 4.14. Top graph: Fetlock flexion (-)/extension (+) (solid line) and internal (-)
lextemal (+) rotation (dotted line) measured using the JCS. Middle graph: Segmental
angles in GCS sagittal plane for metacarpus (solid line) and pastem (dotted line).
Bottom graph: Z displacement between the proximal and distal tracking markers for
the metacarpus (solid line) and pastem (dotted line). The thick vertical line at 61% of
the stride is the time when carpal adduction/abduction angles measured with the JCS
and MPA have the maximal difference.

65

Fetlock flexion/extension patterns were similar when measured using the MPA
and JCS (Figure 4.11) with extension during stance and flexion during swing. Both MPA

and JCS showed adduction during stance with peak abduction during early swing, but the

peak abduction angles were quite different between JCS and MPA (Figure 4.12). The
peak absolute difference (136°) between the JCS and MPA adduction/abduction angles
occurred at 61% of the stride (Figure 4.13), which coincided with peaks of flexion and a
small amount of external rotation (Figure 4.14). As the absolute difference in
adduction/abduction between the JCS and MPA increased, the pastem segment angle
decreased reaching -90° around the time the adduction/abduction absolute difference
peaked (Figure 4.14). The difference in z coordinates of the proximal and distal markers
along the pastem long axis was zero at this point indicating that the pastem long axis was

aligned with the x-axis of the MPA global coordinate system (Figure 4.14).

Discussion

Since the JCS is an anatomically-based coordinate system, it moves dynamically
with the horse's anatomy so that the axes of the coordinate system make adjustments for
the subject's angle of travel, conformation, and unique limb motion patterns, such as the
ternrino. Since MPA is based upon a global coordinate system, the axes are not dynamic
and do not adjust to the changes in anatomical orientation as the horse moves. This is the
basis for the differences in adduction/abduction angles measured using JCS and MPA.
Understanding the movement of the horse during locomotion will assist in locating where

this limitation for MPA creates error in the measurements of adduction/abduction.

66

During the stance phase, the horse’s body moved laterally over the supporting
forelimb as shown by the z displacements in Figure 4.9, which shows the proximal end of
the radius is lateral to it’s distal end throughout stance. The antebrachium and
metacarpus were retracted with approximately equal angular velocities until breakover
when the antebrachium reversed its direction of rotation reaching a vertical position at lift
off, while the metacarpus continued to be retracted. This combination of segmental
motions resulted in carpal joint flexion. During breakover the horse’s body moved
medially until at lift off the proximal and distal ends of the antebrachium were aligned
vertically when viewed in the frontal plane (Figure 4.9). Therefore, at lift off the
antebrachium was vertical when viewed in either the sagittal or frontal plane. Under
these conditions any axial rotation of the antebrachial segnent would have a large effect
on the adduction/abduction angle of the carpal joint as seen in the frontal plane using
MPA (Figure 4.15). This orientation of the proximal segnent and the error that is created
under the conditions described above is illustrated in the 3-D surface gaph (Figure 4.16).
This shows the effects of different combinations of axial rotation of the proximal segnent
and flexion of the joint due to retraction of the distal segment on the adduction/abduction
error in the MPA measurements. The derivations of this relationship is given in Appendix
B. When axial rotation of the proximal segment occurs, it creates an adduction/abduction

error in the frontal plane measurements.

At lift off (55% of the stride), the pastem segment had rotated beyond the
horizontal (sagittal plane angle -92°). At this time, the difference between the z
coordinates of the proximal and distal markers on the pastem long axis was zero so that

the distal aspect of the segment was not visible in the frontal view during the time that the

67

 

Sagittal View Frontal View

Figure 4.15. The true flexion/extension (sagittal view) and adduction/abduction
(frontal view) angles (solid line) and the angular errors (dashed line) created in the
apparent position of the distal segment by a rotation of the proximal segment when
measured using MPA. The illustration shows the apparent rotation of the distal
segnent after rotation of the proximal segment although there is no real change in
the flexion/extension (left) and adduction/abduction (right).

    

70
10
20 30 40 50 60 Flexion of the
70 80 90 Joint (deg)
Axial Rotation of the
Proximal Segment (deg)

Figure 4.16. Contour surface showing adduction (—)/abduction (+) error measured by
MPA in the frontal plane as a result of different combinations of created by axial
rotation and flexion when the proximal segment is vertical in the sagittal plane. The
shading along the contour indicates the error (deg), the vertical line along the contour
indicates axial rotation of only the proximal segment (deg), and the horizontal line

indicates the joint flexion angle (deg). The error is based on calculations shown in
Appendix B.

68

pastem segnent was rotated beyond -90°. Under these conditions, the angle of the

pastem segment was difficult to measure in the frontal plane because the pastem segnent
was behind the metacarpal segnent when the sagittal plane angle was —90° or the pastem
segnent was hidden behind the metacarpal segnent when it rotated beyond -90°. Under
these circumstances, any axial rotation of the metacarpal segment, even without flexion/
extension or adduction/abduction of the fetlock joint, allowed the pastem segment to
become visible in the frontal view. Thus, axial rotation of the metacarpus was
misinterpreted as adduction/abduction of the fetlock joint in MPA. The sudden large
peak in MPA fetlock adduction/abduction just after lift off (Figure 4.12) occurred during
a period of axial rotation of the metacarpus while the pastem segnent angle was past the
horizontal, but the adduction/abduction angle of the fetlock was not changing. Internal
rotation of the metacarpus created an artificial abduction of the fetlock in the frontal
view. External rotation would have created the impression of fetlock adduction. If there
were no metacarpal axial rotation, the pastem segment could not be tracked using MPA

until the amount of retraction angle decreased to less than -90°.

Using MPA, Back et al. (2000) studied the effect of heel wedges on frontal plane
motion of the hind pastem relative to the metatarsus. The problems described above for
frontal plane measurements of the fetlock joint were avoided by measuring the changes in
adduction/abduction with and without heel wedges during the stance phase only. The
findings presented here show that smaller differences were found between JCS and MPA
adduction/abduction measurements during stance. The hind fetlock joint was found to

show increased abduction with the heel wedges during stance. Limiting frontal plane

69

measurements to the stance phase reduced some of the problems in using MPA, but

would not completely resolve errors due to differences in the JCS and MPA axes.

Based on in vitro studies of pastem joint loading, Degueurce et al. (2000)
described the three movements of the joint as always being associated. The relationship
between the different types of movement affects the axes of the JCS. For example, the
JCS adduction/abduction axis is perpendicular to both the flexion/extension and
intemal/extemal rotational axes. Therefore, flexion/extension and/or intemal/extemal
rotation will cause a shift in the orientation of the adduction/abduction axis. This
dynamic quality of the JCS axes allows for the depiction of the relationship or coupling
between the three joint movements. On the other hand, the MPA axes do not move with
the anatomy nor do they account for the relationship between the three joint movements.
Therefore, comparing measurements made using the dynamic axes of the JCS with those
made using the static axes of NIPA will undoubtedly show differences, especially when
the orientation of the segments and joints fulfil the conditions described above. The
angle between the JCS adduction/abduction axis and the GCS x—axis is illustrated in
Figure 4.17. The nrinima and maxima of the traces represent the minimal and maximal
absolute differences between the adduction/abduction angles measured using the JCS and
MPA indicating that the adduction/abduction axis of the JCS has a significant effect on
the measured angle.

The dynamic quality of the JCS adduction/abduction axis is illustrated in Figure
4.18 for the carpus and Figure 4.19 for the fetlock. Each gaph represents a single stride.
The JCS adduction/abduction axis originating from the y displacement curve of the joint

center was plotted for every 3% of the stride. It can be seen that the orientation of the

70

 

 

 

 

 

 

<— Stance Phase—b 4— Swing Phase—>
100
80 -
‘€3 60
0
"‘° 4
.5 40 “’
20
O I I I f I I
0 20 40 60 80 100

Stride (%)

Figure 4.17. Angle between the JCS adduction/abduction axis and the x-axis of
the GCS for the carpal joint (black line) and the fetlock joint (gay line).

 

 

  

 

<— Stance Phase—p <— Swing Phase—p
55'
’8‘50’
3
ii .
.5 45
I
40 \

 

 

0 i0 40 60 so 100

Stride (%)
Figure 4.18. Y displacement of the carpal joint center throughout a single

stride and the corresponding orientation of the JCS carpal adduction/abduction
axis within the sagittal plane shown at intervals of 3% of the stride.

71

<— Stance Phase —fi ‘— Swing Phase—p

1

 

f T

35'

*7

30

Height (cm)

 

 

 

A

0 20 do 80
Stride (%)

 

 

Figure 4.19. Y displacement of the fetlock joint center throughout a single
stride and the corresponding orientation of the JCS carpal adduction/abduction
axis within the sagittal plane shown at intervals of 3% of the stride.

 

 

 

 

 

 

4— Stance Phase -—> {—8ng Phase—p
40
E
,. 20 \ K.‘
<1
0 r I I I I
0 20 40 6O 80 100

Stride (%)

Figure 4.20. Angle between the JCS flexion/extension axes and the z—axis of the
GCS for the carpal joint (black line) and the fetlock joint (gay line).

72

adduction/abduction axis as seen from the sagittal view changes as the joint center moves
vertically throughout the stride. The y displacement of the joint center was chosen as the
variable for the vertical axis to facilitate visualization of the motion of the joint in the
sagittal plane throughout the stride. For the MPA measurements, the adduction]
abduction axis does not move with the limb motion. Therefore, when the orientation of
the JCS adduction/abduction axis is not parallel with the corresponding GCS axis as seen
during breakover (Figure 4.18) there are more differences measured between the JCS and
MPA adduction/abduction angles. Around 20% of the stride (Figure 4.18), the JCS and

MPA adduction/abduction axes are almost parallel resulting in less error (Figure 4.8).

The angle at which the horse traveled through the data collection area was
transformed during post-processing of the data so that it was aligned with z-axis of the
JCS. This resulted in a relatively small difference between the flexion/extension axis of
the JCS and z—axis of the GCS (4° to 26°; Figure 4.20) compared with the relatively large
difference between the adduction/abduction axis of the JCS and x-axis of the GCS (14° to
92°; Figure 4.17). Thus, differences between the flexion/extension measurements for the
two techniques were minimized. However, even after correcting for the horse's angle of
travel, there were small differences in the flexion/extension measurements due to the fact
that axial rotation was not accounted for when using MPA. The angular difference
between the JCS and MPA flexion/extension axes at the fetlock were fairly consistent
throughout the stride with peak differences corresponding with peak adduction/abduction
axes differences. The fact that the peak angular difference in the JCS and MPA carpal
flexion/extension axes occurred during stance whereas the peak angular difference in

adduction/abduction axes occurred during swing may be due to the lateral movement of

73

the body over the limb during stance and the accompanying change in radial orientation
due to this rotation. Differences due to radial orientation may be minimized by choosing

appropriate axes for establishing the radial segnental coordinate system.

A limitation to the JCS used in this study was the selection of the primary axis
used for establishing the segnental coordinate system. In order to facilitate direct
comparisons with earlier three-dimensional analyses using a JCS, the axis created by the
two virtual markers along the segmental long axis was used as the base axis for
establishing the segnental coordinate system. However, the radius is wider proximally
than distally and the proximal marker is somewhat lateral to the distal marker. This
affects the orientation of the coordinate system so that it may not truly represent the load
bearing axis of the antebrachium. The geater the discrepancy, the geater will be the off-
set of the base axis from the load bearing axis of the segment. This also explains why the '
peak absolute difference between the JCS and MPA carpal adduction/abduction is
slightly offset from the point at which the radius is vertical as seen on the bottom gaph in
Figure 4.9. In future studies, it may be preferable to use the axis through the medial and
lateral markers on the distal aspect of the segnent as the base axis of the antebrachial
coordinate system so that difference in size of the proximal and distal radial epiphyses

has less effect on measuring the load bearing axis of the segment.

Conclusions

Human kinematic studies have applied three-dimensional analysis using a JCS to
describe the relative rotations between segnents (Grood and Suntay 1983; Soutas-Little

et al. 1987). Since the JCS is oriented to the subject's anatomy, rather than to the

74

calibrated space through which the movement occurs, the subject does not have to move

completely straight through the data collection area. The JCS method has the additional

benefits of measuring axial rotations and having the ability to detect coupling between
the three types of joint motion. However, the JCS setup for tracking an anatomical
coordinate system is lengthy and in equine subjects it may not be practical for most
clinical applications. The additional effort involved in using a JCS rather than MPA is
justified if it offers improvement in the accuracy in the measurements of joint motion.
The results of this study show that carpal and fetlock joint kinematics measured in the
sagittal plane using MPA compared well with flexion/extension angles measured using a
JCS. Therefore, MPA appears to be adequate for measurement of flexion and extension
of the carpus and fetlock at the walk provided corrections are made to align the angle of

travel with the GCS established by the calibration frame.

Previous studies have measured carpal and fetlock adduction/abduction angles
between the forelimb and the gound (Hening et al. 1992; Peloso et al. 1993) rather than
between the proximal and distal segments comprising the joint, which limits the accuracy
of this measurement. Comparisons between the adduction/abduction angles of the carpal
and fetlock joints using MPA and JCS showed quite different pattenrs and magritudes
with the two techniques. The fact that the JCS gave a more accurate interpretation of
movements in the frontal plane indicates that it is more appropriate than MPA for
describing adduction/abduction. The reasons for problems with frontal plane
measurements cited in this chapter should be considered in measuring any type of equine

locomotion using MPA.

75

It is concluded that the need to use a JCS versus MPA depends on the objectives
of the study. If the only measurements required are flexion/extension in a sagittal plane,
then planar analysis using a MPA appears to be adequate, provided care is taken to ensure
that the subject moves straight and parallel with the plane of calibration. For
measurements of other types of joint motion, it is recommended that more than two
tracking markers are used per segnent and a JCS or a similar anatomically-based system

be used.

76

CHAPTER V
EXPERIMENT III: THREE-DIMENSIONAL MOTION OF THE CARPUS AND

FETLOCK IN THE FORELIMB OF THE MISSOURI FOX TRO'I'TER

Summary

The objective of this study was to measure carpal and fetlock joint kinematics of
the flat walk and fox trot to provide a more comprehensive description of these gaits and
to assist in distinguishing between them. During the flat walking stance phase, the carpus
and the fetlock were extended and abducted. At breakover, the carpus was internally
rotated as the fetlock was externally rotated. During early swing, the carpus and fetlock
decreased the abduction angle as the joints flexed. At mid swing, the carpus reached peak
internal rotation and the fetlock reached peak external rotation. At the fox trot, the carpus
was abducted and extended during stance. There was a period of carpal external rotation
that peaked before breakover followed by a cycle of flexion that peaked in mid-swing.
The fetlock was extended throughout stance and showed a flexion cycle during swing.
This joint showed a small amount of adduction and internal rotation during stance and
switched to abduction and external rotation at breakover that continued throughout the
swing phase. Although magnitudes between gaits were insignificant, except for carpal
external rotation, the carpal and fetlock adduction/abduction patterns are different.
Velocity, stride and stance durations, diagonal step duration, peak carpal external
rotation, and carpal and fetlock adduction/abduction patterns distinguish the flat walk

from the fox trot.

77

Introduction

The four-beat stepping gaits of the gaited horse are recognized by the sequence
and timing of the limb placements, together with the characteristic limb movement styles.
These are measured as the temporal, linear, and angular kinematics. The Missouri Fox
Trotter gaits are the flat walk and the fox trot, both of which are described as four-beat,
lateral sequence, stepping gaits that lack a period of suspension. The flat walk has a
regular rhythm while the fox trot has an irregular rhythm with diagonal couplets. The flat
walk is described as a faster version (5 mph) of the walk (Ziegler 2000). Hildebrand
(1965) described the fox trot as a fast, lateral sequence, diagonal couplets walk. Clayton
and Bradbury (1994) measured the temporal variables of the fox trot and found that
lateral advanced placement was significantly longer than diagonal advancement
placement, thus confirming that the rhythm of the gait is characterized by diagonal
couplets. As a consequence of the irregular rhythm, the diagonal bipedal support periods
were significantly longer than the lateral bipedal support periods. The angular kinematics
of the fox trot have not been described. In chapters three (Experiment I) and four
(Experiment II) the three-dimensional kinematics of the walk of the non-gaited horse and
the Peruvian Paso were shown to have carpal and fetlock joint motion in the frontal and
transverse planes. The objectives of this study are to describe the three-dimensional

carpal and fetlock joint motions of the flat walk and fox trot of the Missouri Fox Trotter.

Materials and Methods
Subjects. The subjects were six Missouri Fox Trotters. The criteria for inclusion

in the study were that the horses were sound, were registered as Missouri Fox Trotters,

78

and were naturally able to perform the flat walk and the fox trot gaits. The horses were
ridden by an experienced gaited horse rider. Trials were excluded from the study if the
horses were determined not to be performing the gait properly according to the following

gait standards (Ziegler 2000):

Flat Walk: The flat walk is a regular, four—beat gait with a lateral sequence of footfalls:
LH-LF—RH-RF. The support sequence alternates between periods of tripedal
and bipedal support with approximately equal amounts of time being spent in
diagonal bipedal and lateral bipedal support phases. The flat walk is
performed at a faster velocity than the walk while maintaining a regular

rhythm. If the rhythm becomes irregular, the horse is perfornring a fox walk.

Fox Trot: The fox trot is a four-beat stepping gait with a lateral sequence of limb
placements similar to the walk. Unlike the flat walk, the fox trot is
characterized by an irregular rhythm with diagonal couplets and longer
periods of diagonal bipedal support than lateral bipedal support. A horse
performing this gait gives the appearance of walking with the front limbs and
trotting with the hind limbs. The fox trot is performed at a faster speed than

the flat walk, otherwise the horse is performing a fox walk.

The fox walk is a transitional gait between the flat walk and the fox trot. It is

characterized by a four-beat rhythm with lateral sequence of footfalls and diagonal

79

couplets, but does not demonstrate the visual characteristics of the fox trot and is
described by the breed association as an undesirable gait (Ziegler 2000).

In this study, temporal variables were measured for the flat walk and fox trot.
Forelirnb protraction/retraction was determined from markers overlying the proximal
aspect of the scapular spine and the lateral aspect of the distal hoof. Angles were
measured relative to a line drawn perpendicular to the gound in the sagittal plane.
Protraction angles were assigned positive values and retraction angles were assigned
negative values.

Segmental Coordinate Systems. The axes of the coordinate systems for the
antebrachial, metacarpal, and proximal phalangeal segnents were established as
described in chapter four. Briefly, the origin of the antebrachial coordinate system was
located on the lateral styloid process of the distal radius. The x-axis was directed
laterally from the origin perpendicular to the antebrachial sagittal plane and on the same
vector as a line from the medial to lateral styloid process. The z-axis, contained in the
antebrachial frontal plane, ran from the origin proximally along the antebrachial long axis
at 90° to the x-axis. The y-axis ran dorsally from the origin, perpendicular to the
antebrachial frontal plane.

The origin of the metacarpal coordinate system was located on the lateral aspect
of the distal third metacarpus. The x-axis was directed laterally from the origin
perpendicular to the metacarpal sagittal plane and on the same vector as a line from the
medial to lateral aspect of the third metacarpus. The z-axis, contained in the metacarpal

frontal plane, ran from the origin proximally along the metacarpal long axis at 90° to the

80

x-axis. The y-axis ran dorsally from the origin, perpendicular to the metacarpal frontal
plane.

The origin of the proximal phalangeal segment was located on the lateral aspect
of the distal end of the proximal phalanx. The z-axis ran from the origin proximally
along the long axis of the segnent. Similar to the other segnental axes, the x and y-axes
ran laterally and dorsally from the origin, respectively.

Marker Placement. Both tracking and virtual markers were attached to the right
forelimb to establish segnental coordinate systems for the antebrachial, metacarpal, and
proximal phalangeal sements. In horses 1-3, the segnental coordinate systems were
established and tracked using virtual marker "L" structures together with the tracking
markers described in chapter three (Figure 5.1). In horses 4-6, the segnental coordinate
systems were established with two virtual markers along the lateral side of the long axis .

and a third virtual marker on the medial side of each segnent. The tracking markers

 

Figure 5.1. Tracking marker placement for horses 1-3 (lefi) and horses 4—6 (right).

81

were placed on the lateral and dorsolateral aspects of the segments as described in chapter
four (Figure 5.1). Although the marker placements were different between horses 1-3 and
horses 46, the segnental axes established by the markers were the same for all horses.

Two croup markers were placed along the sacral spine of horses 4—6 to deterrrrine
vertical excursions calculated from the raw y coordinates. Markers were attached to the
point of the elbow and the lateral aspect of the distal hoof for horses 1-3 to determine
mediolateral shift of the body over the hoof calculated from the raw 2 coordinates.

Data Collection. The calibration volume (342 x 190 x 210 cm3) was defined
using 30 points. Kinematic data of the flat walk and fox trot were recorded using two 60
Hz camcorders (Panasonic AG-450, Matsushita Electric Corp., Secausus, NJ) placed on
the lateral aspect of the right side of the horse and oriented at 60° to each other. Two
additional camcorders placed in front of and behind the horse in increments of 60° to the
lateral camcorders were used to locate the virtual markers on the medial aspect of the
limb in the standing file. A 500-watt lamp was placed behind each camcorder to
illuminate the retroreflective markers during recording.

The horses were ridden along the runway at a flat walk and fox trot. A trial was
considered acceptable if the horse traveled at a consistent speed along a straight line that
was aligned with the calibration volume with all markers being clearly visible in the two
lateral camcorder views. Four trials were recorded for each horse at each gait.

Data Reduction. The frames of hoof contact, heel off and toe off were recorded
for each limb and used to calculate stride duration, stance and swing durations of the right
forelimb and right hind limb, and the time of onset and duration of breakover. The

markers were automatically tracked and digitized using a video analysis system (Ariel

82

Performance Analysis System, Ariel Dynamics Inc, Trabuco Canyon, California). Three-
dimensional coordinates of markers were obtained using direct linear transformation
(Abdel-Aziz and Karara 1971). The raw data were filtered using a fourth-order
Butterworth digital filter at a cutoff frequency of 6 Hz. The joint coordinate system
(JCS), which was based on Grood and Suntay ( 1983), measured flexion/extension
relative to the x-axis of the proximal segment, intemal/extemal rotation around the z-axis
of the distal segnent, and adduction/abduction around a floating axis perpendicular to the
proximal segnental x-axis and the distal segnental z-axis. Joint angles were expressed
relative to the position of alignment of the proximal and distal segnents. Flexion,
adduction, and internal rotation were designated negative. Extension, abduction, and
external rotation were designated positive.

Data Analysis. The angular measurements were time normalized to stride
duration. Ensemble averages were calculated and plotted for the flexion/extension,
adduction/abduction, and intemal/external rotation of the carpal and fetlock joints at the
flat walk and fox trot. Mean values 1: SD were determined for the temporal variables,
protraction/retraction angles, and maximal carpal and fetlock joint angles for both gaits.
Mean values i SD for each horse were gaphed and included in Appendix C. Paired t-
tests were used to determine differences for temporal variables and peak joint angular

measurements between the flat walk and fox trot (P < 0.05).

83

Table 5.1. Gait characteristics (mean and i SD) measured for the flat
walk and fox trot strides used in this study. Significant differences (*)
between gaits are given (P < 0.05). Measurements taken from horses 1-6

(n=6) unless indicated on table.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Flat Walk Fox Trot
Velocity (m/s) 1.75 i 0.06* 3.17 i 003*
Stride Duration (5) 1.17 _-i_- 006* 0.63 i: 0.03*
Forelimb 1.95 i 0.05 2.0 i 0.03
Stride Length (in)
Forelimb 0.77 i’ 0.06* 0.32 _-l_~_ 0.02*
Stance Duration (5)
Forelimb 0.40 10.05 0.31 i 0.03
Swing Duration (3)
Forelimb 0.04 i 0.01 0.03 i 0.01
Breakover Duration (5)
Hindlimb 0.78 :1; 0.06* 0.33 i 0.03*
Stance Duration (5)
Hindlimb 0.35 i 0.04 0.31 i 0.01
Swing Duration (5)
Hindlimb 0.04 i 0.01 0.03 i 0.01
Breakover Duration (5)
Lateral Step Duration (5) 0.30 i 0.01 0.23 :l: 0.02
Diagonal Step Duration (5) 0.32 i 003* 0.10 _-+_; 0.01*
Protraction/Retraction 57 i 3 64 i 2
Range of Motion (°)
Protraction (°) 22 i 4 28 i 3
Retraction (°) -35 i 4 —36 i 2
Croup Vertical Excursions 7 i l 9 i 2
(cm) (Horses 4-6)
Mediolateral Shift over 6 j; 3 4 i 1

Hoof (cm) (Horses 1-3)

 

 

 

84

 

Table 5.2. Standing angles for carpal and fetlock flexion/extension (FIE),
adduction/abduction (Ad/Ab), and intemal/extemal (I/E) rotation of the six
Fox Trotters and the mean and :t SD for all horses. Angles are measured
relative to the position of alignment of the proximal and distal segnent.
Flexion, adduction, and internal rotation are assigned negative values.

 

 

 

 

 

 

 

 

 

Horse #
#1 #2 #3 #4 #5 #6 Mean
(1- SD)

Carpal 3 5 5 6 3 -3 3 (3)
F/E (°)
Carpal 3 7 10 4 3 2 4 (3)
Ad/Ab (°)
Carpal 2 -2 -4 5 -2 -3 -1 (4)
VB
Rotation (°)
Fetlock 3O 19 31 24 20 33 26 (7)
FIE (°)
Fetlock 6 8 4 6 8 2 6 (2)
Ad/Ab (°)
Fetlock 3 7 5 2 2 4 4 (2)
VB
Rotation (°)

 

 

 

 

 

 

 

 

85

 

Table 5.3. Mean (SD) for peak values of three—dimensional joint motions of
the carpus and fetlock for the flat walk and fox trot (n=6). Significant

differences (*) between gaits are given (P < 0.05). Joint motions occurring at
different phases of stride were not statistically compared (1'). The percent of
stride at which the peaks occurred is in square brackets under the peak value.

 

 

 

 

 

 

 

Flexion/extension Intemal/extemal Adduction/abduction
(deg) rotation (deg) (deg)
Flexion Extension Internal External Adduction Abduction
(minimal (minimal
external abduction)
rotation)
Flat -64 (3) 8 (4) -17 (6) 12 (8)* 4 (6) ‘l' 10 (6)*
Walk [72%] [11%] [69%] [39%] [83%] [94%]
Carpus
Fox -67 (6) 10(4) -11 (6) 3 (4)* 4(2) 1' 20 (3)*
Trot [74%] [17%] [86%] [28%] [8%] [78%]
Carpus
Flat -30 (1) 56(3) 2 (9) 27 (6)’[ 3 (3) T 27(6) T
Walk [67%] [23%] [48%] [73%] [82%] [40%]
Fetlock
Fox -26 (7) 58 (4) -2 (8) 30 (9)1“ —6 (7) T 18(7) T
Trot [65%] [37%] [23%] [68%] [16%] [54%]
Fetlock

 

 

 

 

 

 

 

86

 

Results

Flat Walk. The velocity of 1.75 ~_t~_ 0.06 m/s was achieved using a stride length of
1.95 i 0.05 m and a stride duration of 1.170 _-l_- 0.06 3 (Table 5.1). The forelimb stance
phase occupied the period from 0 to 48% of the stride with breakover starting at 40% of
the stride and lasting 0.04 i 0.01 s. The forelimbs demonstrated a maximum protraction

angle of 22° _-t_- 4 and maximum retraction angle of -35° _-l_- 4.

The three-dimensional standing angles of the carpus and fetlock for each horse are
shown in Table 5.2. During stance, the carpus was slightly extended and abducted
(Figure 5.2). Maximal extension of 8° was recorded at 11% of the stride, but maximal
abduction of 10° did not occur until the swing phase around 94% of the stride (Table 5.3).
The carpus rotated externally throughout most of stance, with peak value occuning
around the start of breakover. A peak in internal rotation occurred around lift off and a

larger peak (-17°) in mid swing around the time of peak flexion (Figure 5.2, Table 5 .3).

The fetlock was extended, abducted, and externally rotated throughout the stance
phase with a distinct extension peak (56°) in mid stance. During breakover, there were
decreases in abduction and external rotation. In the swing phase the fetlock joint had two
flexion peaks. The larger flexion peak (-30°) at 67% of stride was followed by a smaller
peak (-16°) at 85% of the stride (Figure 5 .3). Peak abduction of the fetlock (27°) occurred
at the start of breakover. The fetlock was externally rotated throughout the stride with
peak external rotation (27°) occurring at 73% of the stride between the two flexion peaks
(Figure 5.3).

Fox Trot. The velocity of 3.17 i 0.03 m/s was significantly faster and was

achieved using a significantly shorter stride duration of 0.63 i 0.03 s and fore

87

4—Stance Phase —> 4— Swing Phase —*
4O
20

g I I "1‘.

_20 _, V

 

Angle (°)

40%

 

 

 

 

0 20 40 60 80 100
Stride (%)

Figure 5.2. Mean carpal flexion (-)/extension (+) (black line), adduction (-)
labduction (+) (dark gay line), internal (-)/external (+) rotation (light gay line)
of the flat walk (n=6).

4——Stance Phase —-H1— Swing Phase —>

 

80
6O -

  

 

 

Angle (°)

 

 

 

 

O 20 40 60 80 100
Stride (%)
Figure 5.3. Mean fetlock flexion (-)/extension (+) (black line), adduction (-)

/abduction (+) (dark gay line), internal (-)/external (+) rotation (light gay line)
of the flat walk (n=6).

88

(0.32 i 0.02 s) and hind limb (0.33 _-1_-_ 0.03 s) stance durations (Table 5.1). Stride length
did not differ between the flat walk and the fox trot. The forelimb stance phase occupied
the period from 0 to 44% of the stride with breakover starting at 37% of the stride. There
was a trend toward a reduction in breakover time in the fox trot, Compared with the flat
walk, especially in the forelimbs. An irregularity of rhythm with diagonal couplets was
achieved using a shorter diagonal step duration of 0.10 i 0.01 s and a longer lateral step
duration of 0.23 i 0.02 5 compared to the flat walk. The forelimbs demonstrated a
maximum protraction angle of 28° i 3 and maximum retraction angle of -36° _-l_-_ 2.
Although these angles did not differ significantly between gaits, there was a trend toward
geater forelimb protraction at the fox trot. In switching from the flat walk to fox trot, the
range of croup vertical excursions increased by 2 cm (horses 4-6) and mediolateral shift

over the hoof decreased by 2 cm (horses 1-3; Table 5.1).

The carpus was extended through stance with peak extension (10°) at 17% of the
stride, after which the carpus flexed slowly at first, then more rapidly during breakover.
The carpus reached peak flexion (-67°) around 74% of stride (Figure 5.4, Table 5.3).
Throughout the entire stride, the carpus was abducted with maximum abduction (20°) at
mid swing. Maximum carpal abduction angles were significantly different between gaits
(Table 5.2). The carpus showed minimal axial rotation during stance, but was internally
rotated from the start of breakover through the swing phase with a peak value of —1 1°
(Figure 5.4). Peak external rotation was significantly less in the fox trot than the flat

walk (Table 5.3).

The fetlock joint was extended throughout stance with peak extension (58°)

during late stance around the start of breakover. The fetlock extension angle gadually

89

‘— Stance Phase —-> 4— Swing Phase —>

 

40

 

Angle (°)

 

 

 

 

 

0 20 40 6O 80 100

Stride (%)
Figure 5.4. Mean carpal flexion (-)/extension (+) (black line), adduction (-)/

abduction (+) (dark gay line), internal (-) /extemal (+) rotation (light gay line)
of the fox trot (n=6).

4—Stance Phase —> 4— Swing Phase —"

 

 

 

 

 

 

 

O 20 40 60 80 100
Stride (%)
Figure 5.5. Mean fetlock flexion (-)/extension (+) (black line), adduction (-)/

abduction (+) (dark gay line), internal (-)/ external (+) rotation (light gay line)
of the fox trot (n=6).

90

decreased throughout breakover and into early swing. There was a flexion peak (-26°) in

early swing, followed by a slightly smaller flexion peak (~25°) in mid swing (Figure 5.5).

The fetlock showed a small amount of adduction in the middle of the stance phase, but
was abducted during breakover and throughout swing, reaching peak abduction (18°) in
early swing. There was little axial rotation during stance, but the fetlock showed external
rotation during breakover and throughout swing with a distinct peak in early swing (30°),

coinciding with the first flexion peak.

Discussion

In the equine limbs, motions of the joints distal to the hip and shoulder are largely
restricted to the sagittal plane as a result of anatomical constraints that have evolved to
minimize energy expenditure during limb protraction and retraction (Budiansky 1997).
Therefore, motions in the frontal and transverse planes are relatively small in most
horses. However, in chapter three and four, the carpus and fetlock were found to
demonstrate measurable amounts of adduction/abduction and intemal/extemal rotation
during walking. This chapter shows that adduction/abduction and intemal/extemal

rotation of the carpal and fetlock joints occur at both gaits.

Measurement of joint motions outside of the sagittal plane may assist in the
distinction between the gaits of the gaited horse. For example, when comparing the flat
walk and fox trot, the differences in flexion/extension are quite small, but there are larger
differences in adduction/abduction and intemal/extemal rotation. At the carpus, although
there are slight differences in the peak flexion and extension, the motion outside of the

sagittal plane shows greater differences between gaits, particularly between the peak

91

abduction and external rotation. At the fetlock, flexion/extension profiles are similar for
the two gaits, except peak extension occurs later in the fox trot, but the motions outside of
the sagittal plane show larger differences, particularly in the timing of the peak values.
The similarities between the forelimb sagittal plane joint motions in the two gaits may

explain why the fox trot is described as walking in the forelimbs (Imus 1995).

The measurement of temporal variables further assists in differentiating between
gaits in that velocity, stride duration, diagonal step duration, and fore and hind limb
stance durations Show significant differences between gaits. The stride length was similar
for both gaits, which indicates that the Fox Trotter increases stride rate rather than stride
length to increase speed. This is in contrast to dressage horses moving from a collected
to extended walk in which stride length makes a much larger contribution to changes in
speed than stride rate (Clayton 1994), but is similar to recreational horses that increase
walking speed primarily by increasing stride rate (Nicodemus et al. 2000c).

Hind limb kinematics were not measured in detail in this study, but visual
observations along with croup vertical excursions measurements indicate an increase in
hindquarters vertical excursions in the fox trot, which is a common characteristic noted in
the fox trot (Imus 1995). Croup markers on horses 4-6 showed that vertical excursions

increased by 2 cm at the fox trot compared with the flat walk (Table 5.1).

Earlier studies have described the temporal characteristics of the fox trot (Clayton
and Bradbury 1994; Hildebrand 1965), but the angular kinematics of the limbs have not
been reported. The Missouri Fox Trotter was chosen for this study to further describe the
gaits of this breed, the flat walk and fox trot, using three-dimensional kinematic analysis.

The results have shown that the ranges of abduction/adduction and intemal/external

92

rotation of the carpal and fetlock joints of the Missouri Fox Trotter were larger than the
ranges of motion of the non-gaited horses analyzed in chapter three and in earlier equine

three-dimensional joint motion studies (Heleski 1991).

Carpal Joint. Although significant differences were found between the peak
magritudes of external rotation and abduction, the patterns of motion in all three planes
were sinrilar for the flat walk and fox trot. The carpus was abducted through the majority

of the stride for both gaits, but the abduction was geater at the fox trot during swing.

F etlock Joint. The flexion/extension patterns were similar for the two gaits
except that peak extension occurred later in the stance phase for the fox trot than the flat
walk. The extension pattern of the flat walk was more sirrrilar to a trotting profile than a
walking profile (Heleski 1991). During breakover, the decrease in extension and
beginning of flexion were more gradual in the flat walk due to a longer period of
breakover. The flat walk showed more abduction during stance than the fox trot in which
there was a slight adduction in mid stance. The intemal/extemal rotation patterns were

quite similar for the two gaits.

The most distinct kinematic difference between gaits was the reversal of the
adduction/abduction curve during the stance phase of the fetlock (abducted in the flat
walk, adducted in the fox trot). The following observations assist in explaining this
reversal. At the flat walk, the horse’s body spends most of the stride in tripedal support,
and it is during a tripedal support phase with two fore and one hind limbs that peak
fetlock abduction occurs. In measuring the raw 2 coordinates of elbow and hoof markers
placed on horses 1-3, it was found that during this tripedal support phase the body shifted

laterally over the stationary hoof in preparation for the following lateral bipedal support.

93

After the tripedal support phase with two hind limbs and one forelimb, the body moved
medially in preparation for the following diagonal bipedal support phase. During the flat
walk, the three horses demonstrated a mediolateral shift over the hoof of 6 cm compared
to 4 cm in the fox trot in which diagonal bipedal support phases predominate with the
body more centered to maintain balance (Table 5.1). Further analysis of the body motion
relative to the limb segments using more appropriate marker placement on the horse’s

trunk would assist in substantiating these findings.

Differences between gaits for the carpal and fetlock adduction/abduction patterns
are also apparent during the swing phase with peak carpal and fetlock adduction
occurring during the flat walk swing phase and peak carpal and fetlock abduction during
the fox trot swing phase. As described earlier, mediolateral movements of the body
relative to the forelimb during stance cause adduction/abduction joint angular patterns
that are specific to the gait. As a consequence of these body movements, the flat walk
shows a decrease in abduction during breakover whereas the body motion of the fox trot
shows increased abduction during breakover. Due to the twisting of the joint at lift off
and the inactivity of the tendons and ligaments around the joints during early swing, the
flat walk continues the pattern of decreasing abduction that started during breakover and
the fox trot continues the pattern of increasing abduction. During late swing, when the
tendon and ligaments around the joint become active to prepare the distal limb for gound
contact, the direction of movement changes the abduction trends and in both gaits the
abduction angles revert back toward the standing angles. Measurements of muscular

activity and calculation of net joint moments and joint powers of the Missouri Fox Trotter

94

would assist in supporting these explanations of the differences in swing phase
adduction/abduction patterns between the two gaits.

Gait Comparisons. Previous two-dimensional kinematic studies of the forelimb at
the walk (Hodson et al. 2000) measured a slower velocity (1.4 m/s) with a similar stride
duration (1.27 s) to the flat walk (1.170 i 0.06 s) and a longer stride duration than the fox
trot (0.631 _-_l-_ 0.027 s). Hodson et al. (2000) reported a somewhat shorter walking
forelimb stance duration at the walk (0.662 s) than the flat walk (0.77 i 0.06 s) and a
protraction/retraction range of motion (43.9°), which was less than either the flat walk
(57° i 3) or the fox trot (64° i 2). At the trot, Back et al. (1994) reported a higher
velocity (4.0 m/s) than the fox trot (3.17 :1; 0.03 s), which was achieved using a slightly
longer stride duration (0.67 s) with a shorter forelimb stance duration (0.27 s), and a
smaller range of protraction/retraction (443°). The likely explanation for this is that the '
trotting horse utilizes the suspension phase, rather than the protraction/retraction range of
motion to increase velocity. As for hind limb temporal variables, the flat walk hind stance
duration (0.78 i 0.06 s) fell within the range measured for a slow (1.033 s) to fast
(0.383 3) walk (Nicodemus et al. 2000c), but the fox trot (0.33 _-i_-_ 0.03 5) had a shorter
hind stance duration. However, Holmstrom et al. (1994) measured a shorter hind stance

duration at the trot (0.22 s) than the 0.352 5 measured for the fox trot.

The flat walk and fox trot were found to have different flexion/extension
characteristics than either the walk or trot. The fetlock flexion/extension range of motion
measured by Hodson et al. (2000) in the walk (54°) was considerably smaller than the flat
walk (86°) and the fox trot (84°), but Holmstrom et al. (1994) measured a larger range in

the trot (92°). Back et al. (1994) found a larger range of flexion/extension at the carpus

95

(91°) of trotting Dutch Warmbloods compared with the flat walk (72°) and fox trot (77°).
All of these gaits showed two distinct flexion peaks in the swing phase of the trotting

fetlock that were also detected in the flat walk and fox trot.

Compared with the walk the flat walk has faster velocity with shorter stride and
swing durations, but compared with the trot the flat walk has a slower Speed with longer
stride and stance durations. The fox trot has a faster velocity than the walk with shorter
fore and hind stance and swing durations, but has a slower velocity than the trot with
longer fore and hind limb stance durations. The stride duration of the fox trot is shorter
than either the walk or trot giving a quicker stride rate. Compared with the walk and trot,
both the flat walk and the fox trot have geater protraction/retraction, fetlock
flexion/extension, carpal and fetlock adduction/abduction, and carpal and fetlock
intemal/extemal rotation ranges of motion. In the kinematic study of Back et al. (1994),
trotting Dutch Warmbloods have a larger carpal flexion/extension range of motion than
either the flat walk or fox trot. However, Heleski (1991) found a smaller range of carpal
flexion/extension (76°) in trotting Arabians, which was similar to the ranges recorded in
this study for the flat walk and fox trot. Differences in carpal flexion/extension between
studies may be related to breed or gait. Galisteo et al. (1996) studied Andalusians, which
are known to perform their gaits with much animation, and reported that the range of
carpal flexion/extension was 92°, which was geater than either the flat walk or fox trot,
while the fetlock range of motion in the Andalusians (84°) was sirrrilar to the flat walk
and fox trot. Overall, the gait differences found in comparing two-dimensional studies of
the walk and trot with this study demonstrate that the flat walk and fox trot are not only

distinct from each other, but also distinct from the walk and trot.

96

Furthermore, comparisons between Heleski's three-dimensional data of the walk
and trot with the flat walk and fox trot also show differences between the gaits outside of
the sagittal plane. At the carpus, the adduction/abduction ranges of motion of the walk
(14°) and trot (18°) were Similar to the flat walk ( 14°) and fox trot (16°), respectively. As
for the intemal/extemal rotation, the trot (13°) was comparable to the fox trot (13°), but
the walk (9°) was quite different from the flat walk (29°). At the fetlock, the
adduction/abduction range of motion was smaller at the walk (7°) and trot (20°) than the
flat walk (24°) and fox trot (24°). Intemal/external rotation at the walk (9°) and trot ( 13°)
were also smaller than for the fox walk (25°) and fox trot (32°). Comparing all the joint
motions measured by Heleski (1991) with those measured in this study, the geatest
differences were found not in sagittal plane motion, but in the measurements outside of
sagittal plane. Therefore, it is concluded that further studies on the gaits of gaited horses

need to apply three-dimensional analysis using a JCS to distinguish between their gaits.

In conclusion, in comparison to walk and trot, the flat walk and fox trot have
distinct gait characteristics, particularly in the range of joint motion. The flat walk and
fox trot show distinct similarities in the temporal variables and joint angular patterns and
magnitudes. The main differences found between the two gaits were seen in the motion
outside of the sagittal plane with significant differences in peak values of carpal external
rotation and differences in the timing of the carpal abduction peak and the fetlock
adduction/abduction peaks. Therefore, the Missouri Fox Trotter demonstrates similar
ranges of joint motion between it’s two gaits, especially in the terms of flexion/extension
range of motion. Differences between the gaits affected the timing rather than the

magnitude of the peaks.

97

CHAPTER VI

CONCLUSIONS

"And God took a handful of the West Wind and created the horse..."
-The Koran

From a "handful of wind" to the superb athlete seen today, the anatomy of the
horse has evolved in order to survive in a changing environment and through these
changes the horse, at the same time, has become more suitable to man's needs. The fox-
sized eohippus dating back 55 million years has adapted over time into the modem-day's
athletically efficient horse (Harris 1993). Some of the evolutionary changes in the equine
anatomy were directed toward locomotor efficiency, a characteristic that was necessary to
outrun predators. One such adaptation involved restricting the range of motion of the
joints distal to the shoulder and hip so the movements of the distal limb are primarily
those of flexion and extension in the sagittal plane. Planar motion of the limb is
energetically efficient (Budiansky 1997).

The scientific study of equine locomotion has exploded during the past 20 years
as a result of the availability of more powerful computers, which are needed to process
the vast amounts of numerical data. The initial studies were two—dimensional analyses in
the sagittal plane, which described flexion and extension of the joints, but neglected the
relatively small amounts of adduction/abduction and intemal/external rotation. More
recently, interest in three-dimensional analysis has increased and the availability of more
powerful computers has facilitated these complex analyses. Human kinematic analyses

paved the way with the establishment of joint coordinate systems for the human knee and

98

ankle (Grood and Suntay 1983; Soutas-Little et al. 1987). In 1991, Heleski applied this
technique to the horse, and her results showed that even in the athletically efficient limbs
of the horse there were detectable amounts of adduction/abduction and intemal/extemal
rotation. The presence of joint motion outside of the sagittal plane raised the question of
its importance in relation to the performance of different gaits and athletic activities.

The limitation of Heleski's (1991) equine three-dimensional study was that it
encountered noise due to displacement between the strap used to anchor the tracking
markers to the skin and visual loss of tracking markers in one of the camera views. In
vitro studies have corrected for these problems by using bone pins (Degueurce et al.
2000; Chateau et al. 2000), but none of these Studies has offered an opportunity to

measure three-dimensional equine motion without disturbing the horse's natural gait.

Chapter Summaries

Experiment I. In this research progam, the development of a virtual marker
technique overcame the problems that Heleski (1991) encountered. The benefits of this
new technique were particularly obvious in the repeatability of measurements from stride
to stride, which resulted in consistent patterns and amplitudes from stride to stride in the
flexion/extension and adduction/abduction angles for both the carpal and fetlock joints.
The fact intemal/extemal rotation showed relatively large variability even when using the
virtual marker technique may be related to the fact that this joint motion is the last in the
series of calculations so it is more sensitive to any noise due to factors such as skin
displacement. Human studies have found that intemal/extemal rotation is highly effected

by skin displacement showing oscillations of :1; 1° (Cappozzo 1991). Skin displacement

99

algorithms were not applied in this dissertation since only 2-D algorithms are available
and application of these algorithms to a 3-D model may create unforeseen error.
Furthermore, human three-dimensional analysis has described variations in the magnitude
of the intemal/extemal rotation due to imbalances in or instability of the step (Mann
1980). The same may be true in horses. Indeed hoof imbalances induced by adding
lateral or medial wedges were associated with increases in fetlock adduction/abduction
and axial rotation in an in vitro model under vertical loading (Chateau et al. 2000).
Further studies of hoof imbalances in live horses, which induce abnormal shear forces,
may show even geater three-dimensional joint motion, and perhaps more variability
from one stride to the next. Greater three-dimensional equine joint motion is also likely
to be evident with imbalances due to surface irregularities. However, the earlier equine
three-dimensional studies (Heleski 1991) and the results presented in the third chapter of
this dissertation, have shown that intemal/extemal rotation is small for the walk in
comparison to the range of flexion/extension and adduction/abduction.

The walk of the non-gaited horse measured in this study had smaller ranges of
joint motion than the other gaits measured in this dissertation and similar ranges to those
reported in earlier three-dimensional studies (Heleski 1991) except for carpal
adduction/abduction. The horse measured in this study was an older pleasure horse.
Analysis of other non-gaited horses trained to move with more animation, such as Park
horses, may show larger ranges of motion. Another consideration is that horses may
show an age-related increase in adduction as noted in human subjects (Chao et al. 1983),
or adduction/abduction may be affected by hoof/shoe angle. Back et al. (2000) found that

increasing the lateral angle of the hoof was associated with increased adduction of the

100

stifle joint and increased abduction of the tarsal and hind fetlock joints during the stance
phase. Further research using a larger sample population would give insight into the
effect of these variables on adduction/abduction.

Experiment II. Multi-planar analysis (MPA) measures flexion/extension and
adduction/abduction. This technique is less time consuming and the joint calculations are
less complex than using a joint coordinate system (JCS) to measure joint motions out-of-
the sagittal plane making MPA more applicable to a clinical environment. Earlier studies
(Herring et al. 1992; Back et al. 2000) have applied MPA to measure sagittal and frontal
plane kinematics during the stance phase. An obvious limitation to MPA is that axial
rotation can not be measured. For years, equine gait analysis has neglected measurement
of axial rotation because the anatomy of the horse has evolved to primarily move in the
sagittal plane. However, even in a visual evaluation of normal equine locomotion some
axial rotation can be detected, but this motion can not be measured with MPA. One of
the questions to consider is whether it is important to measure small amounts of axial
rotation, especially in the Peruvian Paso, which is noted for "termino," an outward
paddling motion of the forelimb originating from the shoulder. Furthermore, how does
neglecting axial rotation effect the flexion/extension and adduction/abduction
measurements using MPA? In studies of running humans, Soutas-Little et al. (1987)
compared ankle joint kinematics measured using a JCS and MPA. The results showed
large differences between the adduction/abduction angles measured by the two
techniques. In this study, not only were the JCS and MPA measurements of a walking
Peruvian Paso compared, but the reasons for the differences were explored in terms of

joint, segnental, and axes orientations. It was concluded that when the horse moved

101

along the plane of calibration or corrections were made for the angle of travel,
flexion/extension angles were sirrrilar between MPA and JCS. In comparing
adduction/abduction angles for JCS and MPA, differences were found between the two
techniques for both the carpus and fetlock with the largest differences occurring during
the swing phase when the joints are actively externally rotating. Visually, the
relationship between axial rotation and the MPA adduction/abduction error can be seen
on Figure 6.1 when axial rotation is not accounted for it is interpreted as adduction!
abduction in the frontal plane. Therefore, we may simply conclude that in horses known
to demonstrate large amounts of rotation, such as the Peruvian Paso, MPA measurements
are skewed. However, the extent of MPA adduction/abduction error is not limited to the
Peruvian Paso or other horses with large amounts of axial rotation of the limbs. In fact,
MPA adduction/abduction error is created by a combination of factors that includes axial
rotation, though peak error does not coincide with peak axial rotation. For the carpal
joint, peak difference between JCS and MPA adduction/abduction measurements
occurred when the radius was vertical with the gound and the proximal and distal aspects
of the bone were aligned. At this point, the carpus was flexed and the joint was slightly
externally rotated. When the proximal segment of a joint is oriented as described for the
radius, then adduction/abduction error is a function of flexion and axial rotation. When
either flexion or axial rotation increases, the MPA adduction/abduction error increases
and this relationship is calculated using the steps described in Appendix B. As for the
fetlock, during the period of peak difference between the adduction/abduction measured
by the JCS and MPA, the pastem segnent was either flexed to the horizontal so that the

bone was viewed in the frontal plane along it’s long axis or the pastem segment was

102

rotated beyond the horizontal so that it was hidden behind the metacarpus. At the same
time, the proximal and distal ends of the pastem segnent were aligned horizontally.
With this combination of segnental orientations, any axial rotation causes adduction/

abduction error in that MPA misinterprets the rotation as adduction/ abduction.

No Axial
Rotation . "External fl
‘ ‘ ' , Rotation,“

 

Figure 6.1. Frontal view measurements of the horse’s forelimb are skewed due to
external rotation of the antebrachium creating an abduction angle that is not real.

103

Another flaw in MPA is that it’s axes do not move dynamically with the joint
motion so that if the horse does not move parallel with the plane of calibration or if it has
conformational or tracking abnormalities, such as knock-knees or paddling, the axes will
not adjust for these changes creating errors in all three joint motion measurements. In
this study, the horse’s angle of travel was transformed so that it was aligned with the
calibration frame. Consequently, only slight differences between the JCS and MPA
flexion/extension measurements were found. When using the JCS, the adduction/
abduction axis moves dynamically with the other axes demonstrating the coupling of
joint motions. MPA adduction/abduction is measured around a stationary axis so that
coupling of joint motions and particularly intemal/external rotation is neglected.
Therefore, MPA does not address the effect of axial rotation on adduction/abduction, nor
can it assess the coupling of joint motions. It is concluded that when using MPA the
horse’s angle of travel must be aligned with the plane of calibration to measure
flexion/extension and that measurements outside of the sagittal plane should be done
using a JCS.

The description of the joint motion of the Peruvian Paso was based on a single
horse in order to limit variability due to individual styles of movement and
conformational differences. Further studies utilizing larger sample populations should be
conducted to better describe the joint motion of the ternrino. However, the Peruvian Paso
joint motion measured in this study is briefly described in chapter four (Experiment II)
because it is the first scientific description of the kinematics of the Peruvian Paso. The
stance kinematics reflected the horse’s standing conformation. The carpus was extended,

abducted, and internally rotated during stance until breakover where it started to flex,

104

increase abduction, and externally rotate. These movements continued until late swing
when the carpus extended, decreased abduction, and rotated internally. The fetlock was
extended and adducted during stance with a small peak of external rotation in mid stance.
The fetlock reaches peak extension just before breakover, then began to flex, abduct, and
internally rotate during breakover. In early swing, the fetlock reached peak flexion,
abduction, and external rotation, then began to extend, adduct, and internally rotate
throughout the rest of the stride. For both joints, the majority of joint motion outside of
the sagittal plane occurred during the swing phase, which is when the horse exhibits
termino.

Limited generalization of the motion of the termino can be made from this single
subject. Preliminary analysis of other Peruvian Pasos not used in this dissertation showed
that intra-horse variability is small whereas inter-horse variability is larger, particularly
with regard to axial rotation, which represents different degees of animation of the
termino. In the human comparison study of JCS and MPA measurements of the ankle
(Soutas—Little et al. 1987), two runners were measured, but the subjects’ measurements
were not combined since their style of running created differences in the patterns of joint
motion, especially during breakover and early swing. The horse measured in chapter four
was an older pleasure Peruvian Paso. Other Peruvian Pasos measured, but not reported in
this dissertation, included a nationally competitive Show horse and a young, pleasure
horse. The younger horse demonstrated less axial rotation, smaller protraction/retraction
angles, and shorter stride, stance and swing durations than the show horse. All of these
horses were barefoot and similarly trimmed so that in this preliminary analysis shoeing

and trimming were not confounding variables. In addition to the collection of more data

105

describing normal Peruvian Paso joint motion, further research into the effects of training
and shoeing are needed. Back et al. (1995) found significant differences in the temporal
characteristics of the trot stride in young, Dutch Warmbloods after 70 days in training.
Similar results may be seen in the Peruvian Paso. Motion outside of the sagittal plane
may increase with training, especially in regards to the termino. The effect of medial and
lateral wedges on stance phase frontal plane motion of the stifle, tarsus, and hind fetlock
of Shetland ponies found that increased adduction/abduction coincided with the
application of the wedges (Back et al. 2000). Research on the relationship of shoeing and
the three-dimensional joint motion of the termino may find similar results.

Experiment III. In chapter five, the technique of three-dimensional analysis using
a JCS was applied to the forelimb of Missouri Fox Trotters to measure the carpal and
fetlock joint motion of the flat walk and the fox trot. The fox trot has been described as a
gait in which the horse appears to walk with the forelimbs and trot with the hind limbs
(Imus 1995). By changing the coordination patterns of the fore and hind limbs, the
Missouri Fox Trotter can produce characteristics of the walk in the forelimbs and the trot
in the hind limbs. However, the hind limbs during the fox trot do not have a suspension
phase like the trot, but instead, have a period of overlap between the hind limb stance
phases (Clayton and Bradbury 1995). Clayton and Bradbury (1995) defined the fox trot
according to measured temporal characteristics as a fast, lateral sequence, diagonal
couplets walk.

In chapter five (Experiment III) temporal comparisons between the flat walk and
fox trot showed that the fox trot has a faster velocity, which is the result of a shorter

stride duration due to reduced stance and swing durations, while stride length does not

106

change. By changing the footfall timing of the flat walk, the Missouri Fox Trotter was
able to create another four-beat stepping gait with its own unique characteristics, the fox
trot. In the sport of dressage, horses are required to maintain the tempo and rhythm of the
walk during manipulations between collected and extended walk. However, it has been
shown that dressage horses often demonstrate an irregular rhythm with lateral couplets at
the extended walk (Clayton 1995). Lateral couplets are also a characteristic of the
running walk, a gait performed by some gaited horse breeds. Perhaps this shared gait
adaptation between the dressage extended walk and the four-beat non-walking stepping
gaits allow horses to perform the walk at a faster velocity without changing to a leaping
gait. Gait adaptations will be discussed later in more detail.

The walk of the Missouri Fox Trotter demonstrates a relatively long stance
duration and breakover during which the carpus shows a geater range of intemal/extemal
rotation and the same fetlock adduction/abduction range. In comparison, the fox trot
demonstrated geater or similar ranges of joint motion except for carpal intemal/extemal
rotation, while the horse was rapidly picking up and placing down its hooves at a more
rapid rate, but with a similar stride length. These gait characteristics of the fox trot were
associated with higher vertical forces, which result in more extension of the fetlock as the
joints absorb the forces during stance. Preliminary analysis of gound reaction forces
collected during the Missouri Fox Trotter trials show a slightly higher vertical gound
reaction force at the fox trot than the flat walk, which may be the reason for the slightly
larger fetlock extension found in the fox trot. With a shorter period of breakover during
the fox trot, fetlock flexion is more rapid than in the flat walk. Interestingly, the fetlock

extension profile during stance of the flat walk resembles a normal fetlock trotting

107

profile, which may explain why the differences between the vertical gound reaction
forces are not as large as seen between the walk and trot. Larger differences were found
between the mediolateral gound reaction forces of the flat walk and fox trot, which may
be associated with the differences in patterns and amplitudes of adduction/abduction and
axial rotation. Further kinetic analysis of the Missouri Fox Trotter will assist in
explaining the similarities and differences of the flat walk and fox trot.

For both the carpal and fetlock joints, the flat walk and fox trot Show similar
ranges of motion, but there are differences in the adduction/abduction patterns. As
described in chapter five, these differences may be explained by the shifting of the horse's
body over the supporting limb and adjustments in the weight shifts between gaits. In the
fox trot, the horse spends most of stance in a diagonal bipedal support phase, similar to
the trot, and spends less time in a tripedal support phase unlike the flat walk.

Therefore, the body has less time to rotate over the limb so the horse lifts up and
over the body instead of a mediolateral shift increasing vertical excursions of the croup
that reaches peak y displacement during swing and decreasing mediolateral range of
motion of the elbow relative to the hoof during stance. This may be an adaptation for
shortening the time taken to rotate over the limb. During breakover, when the body is
shifting from a tripedal to diagonal bipedal support and preparing for lift off of the
forelimb, is when the largest shift in carpal internal rotation and fetlock abduction are
measured. During a diagonal bipedal support phase the weight of the body is more
centralized where as the weight of the body in a tripedal support phase is offset toward
the side that has two limbs on the gound. The longer period of tripedal support during

the flat walk means the weight of the body is more offset throughout the stride.

108

Therefore, the differences in limb support and how the body moves over the limb, such as
the longer period of rotation of the body over the limb during the flat walk compared to
the increased vertical excursions and decreased mediolateral limb motion during the fox
trot, may explain differences in adduction/abduction patterns. This adaptation to rapidly
move the body over the limb during the fox trot is also apparent in the rapid flexing of the
fetlock during the breakover as seen in a trotting flexion/extension profile.

In comparing the flat walk to the fox trot, while exhibiting some similarities, both
have unique gait characteristics, particularly in the three-dimensional joint angular
motion. Therefore, using the ranges measured in earlier two-dimensional kinematics of
the walk of non-gaited horses for determining abnormalities in the Missouri Fox Trotter
would be limiting even with the flat walk. In fact, the flat walk demonstrated a larger
range of motion, except for carpal adduction/abduction, than what was measured for the °
walk of the non-gaited horse in chapter three. On the other hand, differences between the
walk of the Missouri Fox Trotter compared to the fox trot were small apart from the few
exceptions discussed earlier, which may be useful in the detection and treatment of
lameness. These conclusions emphasize the importance of knowing the normal ranges of
motion for the gaits of the gaited horses in order to distinguish normal and pathological

motion patterns.

Gait Adaptations
Measuring only sagittal plane kinematics is not the whole story of the gait. In
some gaits such as the walk, the internal/extemal rotation may be small, but neglecting

the influence of internal/extemal rotation can distort the other measurements. In the gaits

109

performed by gaited horse breeds, out-of—sagittal plane joint motion is a significant
characteristic of the gait that may help to explain how these horses are able to use
stepping gaits at faster velocities while non-gaited horses switch to the trot. Although
this dissertation only studied a few gaits for a few breeds, each gait demonstrated unique
characteristics that may be described as adaptations. The following section describes
some theories regarding the origins of these adaptations.

Limb interference has been identified as a cue to the horse to make a transition to
another gait (Hildebrand 1978). The larger ranges of flexion/extension, adduction]
abduction, and intemal/extemal rotational motion in the Missouri Fox Trotter may allow
these horses to avoid interference between limbs and to continue using stepping gaits at
faster velocities. Since only forelimb kinematics were measured, the relationship
between hoof placements and interference could not be determined. Future studies
measuring the over-tracking distances and arcs of the distal (hoof) limb movement in the
different planes relative to the position of the hoof that is in stance will assist in
determining the presence of interference. This would be particularly important for the
Tennessee Walking Horse, which is known for having a large amount of over-tracking
with the hind limbs (Imus 1995).

Gaited horses may be able to perform their gaits with this geater range of motion
due to anatomy, training, and/or shoeing. Shoeing studies have shown that the
application of medial/lateral hoof wedges increase the out-of-sagittal plane motion in the
stifle, tarsus, and fore and hind fetlocks (Back et al. 2000; Chateau et al. 2000). The
subjects used in this study had no special training or shoeing, but were described to have

selective breeding. This breeding may have resulted in the inheritance of certain

110

conformational traits needed to perform these gaits. Slade (1993) found particular
conformation correlated well with the quality of the performance of the running walk in
several gaited breeds. Similar studies on conformation of the Peruvian Paso and Missouri
Fox Trotter may lead to similar results.

At faster velocities the horse's limbs are subjected to increased vertical gound
reaction forces (Khumsap et al. in press), and increased forces have also been shown to
act as a cue for the horse to switch gaits (Rubin and Lanyon 1982). Horses increase the
flexion/extension of the joints to absorb increased gound reaction forces (Leach 1990).
Preliminary analysis of the Missouri Fox Trotter gound reaction forces found higher
mediolateral gound reaction forces in the fox trot than the flat walk, which suggests that
the increased out-of-sagittal plane motion helps to absorb the out-of-sagittal plane gound
reaction forces. Therefore, the larger range of joint motion outside of the sagittal plane,
may allow the gaited horse to absorb the additional out-of-sagittal plane gound reaction
forces producing stepping gaits at faster speeds without being cued to switch to a leaping
gait. Understanding of the relationship of the mediolateral gound reaction forces and
joint motion outside of the sagittal plane will need to entail further analysis of the kinetics
and net joint moments and joint powers.

Impulse is the amount of force applied over a period of time, such as a stance
phase. If stance duration decreases, the force must increase to maintain the impulse.
Studies of non-gaited horses (Khumsap et al. in press) have shown that as walking
velocity increases, the hind limbs are subjected to significantly higher vertical gound
reaction forces distributed over a shorter stance duration. Conversely, lame horses often

use an increased stance duration as a means of reducing the peak forces by distributing

lll

the force on the lame limb over a longer period of time (Keegan et al. 2000). Therefore,
the longer stance duration of the fox trot compared to that reported in the trot (Back et al.
1994) may decrease the peak vertical gound reaction forces, thus delaying the cue to
switch gaits.

The gaits of the gaited horse are described as being efficient (Imus 1995), and this
is important because energy expenditure has been described as another cue to the horse to
switch gaits (Hoyt and Taylor 1981). In competitive human race walkers, as speed
approaches the upper limit for the walk, efficiency decreases until the walker breaks into
a run by adding a suspension phase. This results in an increase in gait efficiency
(Cavagna and Franzetti 1981). Therefore, it must be questioned whether the faster
stepping gaits of the gaited horse are truly energetically efficient. Although gait
efficiency was not measured in this study, some speculations can be made as to how the
locomotor adaptations of the gaited horse adjust for gait efficiency. As the human
competitive race walker increases speed, the flexion/extension of the joints increases to
take advantage of the elastic properties of the muscles, ligaments, and tendons. This
increased flexion/extension allows for geater storage and release of elastic energy
(Cavagna et al. 1976). By increasing the fetlock extension of the fox trot during the
stance phase, the Missouri Fox Trotter may also be utilizing this elastic property to help
conserve energy. By preserving gait efficiency, the gaited horse can continue stepping
through the gait at faster speeds without adding a suspension phase.

The amount of out-of-sagittal plane motion during stance may also be important
to energy conservation through utilization of the elastic properties of the muscles,

ligaments, and tendons responsible for movement. However, the amount of energy

112

conservation is limited due to the small amounts of muscles, ligaments, and tendons
stretching as a consequence of adduction/abduction and intemal/external rotation.
Generally, elastic energy storage and release is thought to be more effective in leaping
gaits. Wickler (2000) has studied the gait efficiency of the walk and trot and has made
some observations on the gaits of gaited horses. From preliminary data, the non-walking
stepping gaits were concluded to be inefficient, which suggests other factors such as
training and environment have a geater influence on performance of these gaits than
energy expenditure.

The large amount of out-of—sagittal plane motion in the gaited horse may be
disadvantageous to structural soundness. This additional movement of the joints may be
considered to contribute to instability of the joint, which may make the production of
leaping gaits more difficult and uncomfortable for the horse. Due to the historically
small numbers of gaited breeds, clinical histories are limited. Gaited horse owners have
described suspensory ligament problems with the Peruvian Paso and hip dysplasia with
the Tennessee Walking Horse. It is not known how much heredity and environment
contribute to these problems. Evaluation of the anatomical structures of the limb joints
and histories of joint problems specific to the breeds may give further insight into
anatomical problems associated with these breeds and disadvantages of these gaits.

If gait efficiency is not a driving factor for production of the non-walking
stepping gaits, external conditions may also be influential in gait production. A study of
the effects of incline (Wickler et al. 1999) and added back weight (W ickler et al. 2000)
on gait speed selection showed that a 10% incline was associated with a decrease in

preferred speed of 0.31 m/s, whereas carrying extra weight equivalent to increasing body

113

weight by 18.5% was associated with a decrease in preferred speed of 0.16 m/s. A 10%
increase in body weight and a 3.5% increase in treadmill incline increased stride
frequency (Barrey et al. 1993). Along with being described as a mechanism for
decreasing muscle stress (W ickler et al. 1999), these gait adaptations have been attributed
to a fear of loss of balance (Zatisiorky 1994). Pat Parelli, a nationally recogrized horse
trainer, finds training a horse to lie down on cue is the most difficult maneuver to teach
because of their instinctive fear of losing balance (Parelli 2000). This fear motivates the
horse to make adaptations in their body stance so that they will always be balanced
enough to flee from predators. By adding weight to their backs or inclining the terrain,
the horse's balance is shifted and they must make adaptations to their stance and their gait
to maintain balance. Dressage horses performing the half pirouette at the walk adapted
for the loss of balance due to dynamic instability by adjusting the footfalls and support
sequences creating an irregular walking rhythm with a longer hind stance and tripedal
support (Hodson et al. 1999). The prolongation of hind stance led to increased overlap
with the increased number of hooves on the gound giving a larger base of support. At
slower walking velocities (0.96 m/s) horses increase their overlap and base of support by
having a longer fore (1050 ms) and hind (1033 ms) stance durations and tripedal support
(83% of stride duration). With increased velocity (2.89 m/s), the fore (367 ms) and hind
(383 ms) stance durations and tripedal support (17% of stride duration) decreased
(Nicodemus et al. 2000c). The North American Single-Footing Horse performing the
road gait, a non-walking stepping gait with a regular rhythm, replaced the tripedal
support phase of the walk with a single support phase that made up 41% of the stride

(Nicodemus et al. 2000b). The horse relied more on the dynamic stability of the gait than

114

the overlap and the number of hooves on the gound to maintain balance. When looking
back at the history of the breeds studied in this dissertation, the Missouri Fox Trotter and
Peruvian Paso, these breeds were established to move over rugged terrain usually
carrying heavy loads (Imus 1995). These horses were selected and bred to be able to
adapt to these conditions without losing their balance, and so through time, the gaits that
were suited for these conditions were established into the breeds.

This balance is maintained through the longer stance duration, comparative to the
trot as discussed earlier, and the presence of tripedal support phases during the gait. The
longer period of diagonal bipedal support of the fox trot (Clayton and Bradbury 1995)
also maintains balance during locomotion because the line of support runs close to the
horse's center of gavity (Hodson et al. 1998). These gait adaptations are necessary to
maintain balance over uneven terrain.

Most of the studies on animal adaptations for balance have utilized deccerebrate
cats in which the stepping patterns are adapted in accordance with changes in the nature
of gound support (Yanagihara et al. 1993). Adjustment of limb patterns in an attempt to
achieve dynamic stability is a learning process, especially in more complex adaptations.
It involves several structures of the body including the head, vertebrae, and limbs.
Studies of postural adjustment of cats with forelimb movement determined two patterns.
A diagonal bipedal limb support initially was detected as a forelimb of the cat was
moved. This limb support phase represented the least displacement of the center of
gavity. After several trials, cats made a postural adjustment by shifting the head to the
opposite forelimb and leaving three instead of two limbs on the gound (Bush and Clarac

1985). Although a learned process, it was more stable because the vertebral column

115

flexed laterally, which shifted the center of gavity to the center of the triangle formed by
the three supporting limbs. Similar adaptations may be used by the gaited horses in the
non—walking Stepping gaits with complex postural adjustments being learned through
adjusting the head, vertebrae, and limbs to find the most stable position.

This adaptation for increased stability may have been learned for survival in
rugged terrain in breeds including the Missouri Fox Trotter and Peruvian Paso that
originated from such an environment. This postural adaptation is seen in the Missouri
Fox Trotter as each hind limb hits the gound the head shifts toward the single forelimb
already on the gound (Nicodemus et al. 2000a). This action shifts to the center of the
horse into the triangular base of support comprised of one fore and two hind hooves.

This adjustment of the head and vertebrae is also an adaptation seen in equine forelimb
lameness (Peloso et al. 1993), which further demonstrates the highly adaptive ability of '
the horse. Although terrain may not be a factor for gait adaptation for the modern
Missouri Fox Trotter or Peruvian Paso, human interaction through training may remain as
the driving force in encouraging these horses to make these adaptations, as seen in the
temporal variables and three-dimensional joint motion, to produce the non-walking

stepping gaits.

Recommendation for Future Studies

The application of three-dimensional analysis is necessary for quantifying such
gait adaptations as out-of-sagittal plane joint motion. Furthermore, since multi-planar
analysis was found inaccurate for measuring adduction/abduction and did not measure

internal/extemal rotation, a joint coordinate system using the virtual marker methodology

116

is more suitable for the comprehensive and accurate measurement of joint motion. In all
three experiments described in this dissertation, out-of—sagittal plane joint motion was
detected and for each gait the pattern of joint motion was quite unique. For example,
even in the walk of each of the three breeds in this dissertation showed different joint
patterns and amplitudes of motion. The ability to measure movements outside of the
sagittal plane is important in the complete understanding of the locomotion of the gaited
horse. Furthermore, as seen in chapter three (Experiment I), even in the non-gaited horse
joint motion outside of the sagittal plane is evident so that the non-gaited horse would
benefit from three-dimensional analysis. Interestingly, the non-gaited horse
demonstrated the largest range of carpal adduction/abduction measured in this
dissertation and showed more adduction/abduction than earlier measurements of the non-
gaited horse's walk (Heleski 1991). The differences in carpal adduction/abduction may be
related to such factors as age and should be further investigated. Although adduction!
abduction differences between horses could be measured using MPA, in chapter four
(Experiment II) this technique was found to create error that is a function of flexion and
axial rotation in frontal plane measurements. The axial rotation measured for the walk of
the non-gaited horse is small, but as determined in chapter four any amount of this
rotation will create adduction/abduction error using MPA. Therefore, further three-
dimensional analysis studies of both gaited and non-gaited horses should include the use
of a JCS.

In the Missouri Fox Trotter, the flat walk and fox trot demonstrated similar
patterns of joint motion, especially in the sagittal plane, which may explain why the fox

trot is described as the horse walking on the forelimbs. Both gaits demonstrated similar

117

stride length and similar motion characteristics in that during stance the carpus extended,
abducted, and externally rotated and the fetlock extended, and during swing the carpus
flexed and internally rotated and the fetlock flexed and externally rotated. At the fox trot,
an increase in velocity was associated with decreases in stride duration, fore and hind
limb stance durations and tripedal limb support, while diagonal bipedal support
increased. These changes in temporal characteristics from the flat walk to the fox trot
coincide with kinematic differences in that the fetlock adduction/abduction pattern is
flipped so that peak values of adduction/abduction occur at different phases of the stride
and the carpus demonstrates a peak of abduction during swing. The fetlock extension
peak occurred later in the stance phase for the fox trot than the flat walk, after which the
fetlock began to flex rapidly during breakover and early swing. At the same time there
was a change from tripedal support to diagonal bipedal support. These kinematic
changes in joint motion patterns gave the appearance that the horse was walking on the
forelimbs. Further analysis of the fox trotting hind limbs may find joint motion similar to
a trot as indicated by the increase in vertical excursions of the croup during the fox trot.
Differences such as the adduction/abduction patterns between gaits give insight into the
joint mechanics for performing the two gaits, but further research on the muscle
contribution and the effects of external conditions would give geater insight into the
gaits of the Missouri Fox Trotter. Kinematics of other non-walking stepping gaits should
be measured including further research on the Peruvian Paso. Research concerning the
gaits of the gaited horse are. important as the gaited horse population in the United States
is gowing rapidly. The number of veterinarians that are knowledgeable of these gaits is

limited so that this research can assist in objectively describing the gaits and be a tool for

118

lameness and treatment evaluations. As seen in the flat walk and fox trot, the differences
between gaits may be small, but these differences may prove to be important in assessing
the quality of the gait or in the detection of lameness.

In all gaits, particularly those of the gaited horse in which gait adaptations are
made so that the horse can continue the stepping gait at faster speeds, the application of
three-dimensional analysis using a JCS is necessary for a complete picture of the gaits
and these adaptations. With these special gait characteristics may come unique problems
concerning both the health and the performance of the gaited horse that can only be fully
addressed through analysis of motion outside of the sagittal plane. However, after
finding measurable amounts of out-of-sagittal plane motion in the non—gaited walk and
determining that differences in adduction/abduction measurements may be related to such
factors as age, training and shoeing, it is the recommendation of this researcher that three-
dimensional analysis should not exclude the non-gaited horse. Three-dimensional
analysis using a JCS is recommended since MPA is restrictive due to corrections for the
horse's angle of travel for flexion/extension measurements is needed, frontal plane
measurements are skewed, and axial rotation can not be measured. Nevertheless, through
these experiments it has come to the attention of this researcher that the following aspects
of the JCS technique need to be further improved including the development of 3-D
correction algorithms for skin displacement that encompass many different gaits and
breeds and development of computer progams to facilitate tracking and processing three-
dimensional data. As for the clinical application of a JCS, along with determining the
ranges of normal joint motion of various gaits, it is necessary to deve10p gaphing

progams to assist in the visualization of three-dimensional motion. The JCS

119

methodology presented in this dissertation offers an opportunity to gain a more complete
understanding of joint motion, which is important in differentiating between the non-

walking stepping gaits of gaited horses for both clinical and performance applications.

Final Conclusions

The aim of this dissertatibn was to contribute to the discipline of equine
biomechanics by developing a methodology for quantifying the locomotion of the gaited
horse. A virtual marker methodology to facilitate three-dimensional analysis using a JCS
was developed for the carpal and fetlock joints. The virtual marker methodology was
found to be feasible and practical for three-dimensional analysis and the carpal and
fetlock joint angles were repeatable over several strides. Comparisons were made
between flexion/extension and adduction/abduction measurements using a JCS and MPA.
The results showed that MPA was acceptable for flexion/extension measurements if the
horse was moving parallel to the plane of calibration. Adduction/abduction error was
found with MPA that contributed to axial rotations and segmental orientations along with
differences in the JCS and MPA adduction/abduction axes. Three-dimensional analysis
using a JCS was applied to the carpus and fetlock of the Missouri Fox Trotter during the
flat walk and fox trot. Although stride length and carpal and fetlock ranges of motion 6
were similar, velocity, stride and stance durations, diagonal step duration, carpal external
rotation peaks, and carpal and fetlock adduction/abduction patterns were different
between the gaits. Further application of a JCS using the virtual marker methodology as

described in this dissertation is recommended for the continued development of the

understanding of the gaits of the gaited horse.

APPENDICES

121

APPENDIX A: EXPERIMENT II

4— Stance Phase —> 4—Swing Phase —>

 

 

 

Angle (°)

 

 

 

 

 

0 20 40 60 80 100
Stride (%)

Figure A. l . JCS carpal flexion (-)/ extension (+) angles (°) measured for Peruvian
walking trial 1 (black line), 2 (dark gay line), and 3 (light gay line).

4-— Stance Phase —-> d—Swing Phase —§

 

20

0
-20 .
-40 r
-60 u
-30 .
-100 r
~120

 

  

 

Angle (°)

 

 

 

 

0 20 40 60 80 100
Stride (%)

Figure A.2. MPA carpal flexion (-)/ extension (+) angles (°) measured for Peruvian
walking trial 1 (black line), 2 (dark gay line), and 3 (light gay line).

122

 

 

 

 

 

 

+— Stance Phase —> d—-Swing Phase —>
40
20 - 75%
C
0
Tab
E
< 0 fl I I I
-20

0 20 40 60 80 100
Stride (%)

Figure A.3. JCS carpal adduction (-)/ abduction (+) angles (°) measured for Peruvian
walking trial 1 (black line), 2 (dark gay line), and 3 (light gay line).

4— Stance Phase —> 4—Swing Phase—D

 

 

 

 

 

 

 

O 20 40 6O 80 100
Stride (%)

Figure A.4. MPA carpal adduction (-)/ abduction (+) angles (°) measured for Peruvian
walking trial 1 (black line), 2 (dark gay line), and 3 (light gay line).

123

<— Stance Phase —> fi—Swing Phase—->

 

 

 

 

 

 

 

 

o 20 4o 60 80 100
Stride (%)

Figure A.5. JCS carpal internal (-)/ external (+) rotation angles (°) measured for
Peruvian walking trial 1 (black line), 2 (dark gay line), and 3 (light gay line).

4— Stance Phase—b d—Swing Phase—>

 

40

 

 

 

 

 

 

0 20 40 60 80 100
Stride (%)
Figure A.6. Carpal absolute difference between JCS and MPA adduction (-)/

abduction (+) angles (°) measured for Peruvian walking trial 1 (black line), 2 (dark gay
line), and 3 (light gay line).

124

<— Stance Phase —> l4—Swing Phase—D

 

 

 

 

 

 

 

0 20 40 60 80 100
Stride (%)

Figure A.7. JCS fetlock flexion (-)/ extension (+) angles (°) measured for Peruvian
walking trial 1 (black line), 2 (dark gay line), and 3 (light gay line).

. ‘— Stance Phase ——> N—Swing Phase—->

 

 

 

 

 

 

 

Stride (%)

Figure A.8. MPA fetlock flexion (-)/ extension (+) angles (°) measured for Peruvian
walking trial 1 (black line), 2 (dark gay line), and 3 (light gay line).

125

<—- Stance Phase —> fi—Swing Phase —->

 

 

 

 

 

 

 

 

O 20 40 60 80 100
Stride (%)

Figure A.9. JCS fetlock adduction (-)/ abduction (+) angles (°) measured for
Peruvian walking trial 1 (black line), 2 (dark gay line), and 3 (light gay line).

4— Stance Phase —> 4—Swing Phase -——>

 

200
160 '
120 -
8O '
4O .
0 ea
-40 -
-80 r
-120 r
-160 ..
-200

 
 
 
 
 
  

Angle (°)

 

 

 

 

 

0 20 4O 60 80 100
Stride (%)

Figure A. 10. MPA fetlock adduction (-)/ abduction (+) angles (°) measured for
Peruvian walking trial 1 (black line), 2 (dark gay line), and 3 (light gay line).

126

4— Stance Phase —> §-—Swing Phase ——>

 

 

 

 

 

 

 

 

0 20 4O 60 80 100
Stride (%)

Figure A.11. JCS fetlock internal (-)/ external (+) rotation angles (°) measured for
Peruvian walking trial 1 (black line), 2 (dark gay line), and 3 (light gay line).

{-— Stance Phase —> <—Swing Phase ——>
200

160 r
120 -
80 r
404

0 .
~40

 

 

 

Angle (°)

 

 

 

0 20 4O 60 80 100
Stride (%)
Figure A. 12. Fetlock absolute difference between JCS and MPA adduction (-)/

abduction (+) angles (°) measured for Peruvian walking trial 1 (black line), 2
(dark gay line), and 3 (light gay line).

127

<-— Stance Phase —> fl—Swing Phase —>

 

 

Angle (°)

 

 

 

 

 

0 20 40 60 80 100
Stride (%)

Figure A. 13. Radius segnental flexion (-)/ extension (+) angles (°) measured for
Peruvian walking trial 1 (black line), 2 (dark gay line), and 3 (light gay line).

4— Stance Phase -—-> l—Swing Phase —>

 

 

Angle (°)

 

 

 

 

 

0 20 4O 60 80 100
Stride (%)

Figure A. 14. Metacarpal segnental flexion (-)/ extension (+) angles (°) measured
for Peruvian walking trial 1 (black line), 2 (dark gay line), and 3 (light gay line).

128

<— Stance Phase —> +—Swing Phase—b
100 ’ *

 

 

 

Angle (°)

 

 

 

 

0 20 40 60 80 100
Stride (%)

Figure A. 15. Pastem segmental flexion (-)/ extension (+) angles (°) measured for
Peruvian walking trial 1 (black line), 2 (dark gay line), and 3 (light gay line).

 

 

 

{—— Stance Phase —-> 4——Swing Phase—b
12 *
E 8
81
~ 4 f 1
fl /
“é o " 7"" \ /
I I \ I r /
0 .
3 4- /
a. _
.2
n 8-
-12

 

 

 

20 4o 60 80 100
Stride (%)

Figure A. 16. Differences (cm) between the z-coordinates of the proximal and distal
markers of the radius segnent during Peruvian walking trial 1 (black line), 2 (dark
gay line), and 3 (light gay line).

C

129

+— Stance Phase —>|4—Swing Phase —>

.h

 

 

 

L o

 

Displacement (cm)

 

 

 

 

O 20 40 60 80 100

Stride (%)
Figure A.17. Differences (cm) between the z-coordinates of the proximal and distal

markers of the metacarpal segnent during Peruvian walking trial 1 (black line), 2
(dark gay line), and 3 (light gay line).

4— Stance Phase —->J<—— Swirrg Phase —->

 

 

 

O

  

Displacement (cm)
do is

 

 

 

.L.
N

 

0 20 40 60 80 100
Stride (%)
Figure A. 1 8. Differences (cm) between the z-coordinates of the proximal and

distal markers of the pastem segnent timing Peruvian walking trial 1 (black line),
2 (dark gay line), and 3 (light gay line).

130

 

 

 

 

 

 

 

O 20 40 60 80 100
Stride (%)
Figure A. 19. Angles (°) measured between the JCS carpal flexion/extension axes

and the GCS z-axis for Peruvian walking trial 1 (black line), 2 (dark gay line),
and 3 (light gay line).

4— Stance Phase —>l<-—Swing Phase-->

 

 

 

 

 

 

Stride (%)

Figure A.20. Angles (°) measured between the JCS carpal adduction/abduction
axes and the GCS x-axis for Peruvian walking trial 1 (black line), 2 (dark gay
line), and 3 (light gay line).

131

<— Stance Phase -—> fi—Swing Phase —->

 

 

 

 

 

 

 

O 20 40 6O 80 100
Stride (%)

Figure A.21. Angles (°) measured between the JCS fetlock flexion/extension axes
and the GCS z-axis for Peruvian walking trial 1 (black line), 2 (dark gay line), and
3 (light mv line).

4— Stance Phase ——->|<-—Swing Phase—->
100

 

Angle (°)

 

 

 

 

 

 

0 20 40 60 80 100
Stride (%)

Figure A.22. Angles (°) measured between the JCS fetlock adduction/abduction axis
and the GCS x-axis for Peruvian walking trial 1 (black line), 2 (dark gay line), and 3

(light gay line).

132

APPENDIX B: EXPERIMENT II (DERIVATION)

Adduction/Abduction Error Derivation Steps. The following illustrates and describes the
derivation of the 3-D surface gaph (Figure 4.16). The illustration shows the angle and
orientation of two lines viewed from the side, the front, and the top before (solid line) and
after (dotted line) a rotation of the top line. The angle of flexion (9;) is determined from
the side view by subtracting the known flexion angle between the two lines from 180°.
Since the rotation is given, the rotation angle (9,) is known for the top view. The angle
of adduction/abduction error (8,) due to the rotation is calculated in the following steps:

 

 

 

O ‘0
y I I Y 2
D
x z
x
Side View Front View Top View

The length (L) of the lower line in the sagittal view is known, which is used with 9f to
determine the distance (S) from the location of the lower line before the rotation to a line
perpendicular to the gound before the rotation.

l
S=sinOfL ()

This distance (S) represents the length of the lower line before and after rotation in the
top view. Using the determined S value and the known rotation in the top view, the
distance the line rotates (D) is calculated.

D = sin 9,5 (2)

133

’-

When S from equation 1 is substituted into equation 2, D becomes a function of Of and
er.

D=sin 8, sin OIL (3)

The height (H) of the line in the front view is determined from 9f and the length of lower
line as measured from the side view.

H =cos€-)fL (4)

From the determined distance (D) and the height (H), the angle of error is determined.
Therefore, the angle of error is a function of the flexion and rotation angles as shown on

Figure 4.15.

 

D (5)
tan 6, = —
H
(6)
tan 9. = sin 9, sin GIL
cos OIL
(7)

tanOe=sin(-3,tan(~)f

134

APPENDIX C: EXPERIMENT III

d—Stance Phase —> 4— SwinLPhase —>

 

60

 

 

 

 

 

 

o 20 4o 60 80 100
Stride (%)

Figure C.l. Fox Trotter #1 carpal mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the flat walk

<——Stance Phase —> <— Swing Phase —->

 

80
60 r
40 '1

 

 

 

   
  

 
   
 

 

 

20 ' V’
t A I42

8 0 . .‘fiw’ :2:
'E'a ‘ “(I/
= -20 ~
<

.40 .1

-6O

 

 

 

 

0 20 40 60 80 100
Stride (%)

Figure C.2. Fox Trotter #1 fetlock mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the flat walk.

135

d—Stance Phase —> 4— Swing Phase —*

 

6O
40 -

 

 

 

 

 

 

 

0 20 40 6O 80 100
Stride (%)

Figure C.3. Fox Trotter #1 carpal mean (thick line) and SD +/- (thin line) flexion
(—)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the fox trot.

4-—Stance Phase—D +— Swing Phase ‘-—'>

 

 

 

 

 

 

 

 

 

0 20 40 60 80 100
Stride (%)

Figure C.4. Fox Trotter #1 fetlock mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (—)/abduction (+) (dark gay line), internal
(-)/extemal (+) rotation (light gay line) of the fox trot.

136

4—Stance Phase-—> <— Swing Phase—P

 

 

\f f\
-~"/—‘\

- ..- ,'~’:\
1933335”; ~“‘\\ <WA\ MI.’/ 4

 

 

 

 

 

0 20 40 60 80 100
Stride (%)
Figure C.5. Fox Trotter #2 carpal mean (thick line) and SD +/- (thin line) flexion

(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
9’ external (+) rotation (light gay line) of the flat walk

80 4-—Stance Phase—b 4— Swing Phase—D

 

 

 

 

Angle (°)

 

 

 

 

 

o 20 4o 60 80 100
Stride (%)

Figure C.6. Fox Trotter #2 fetlock mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the flat walk.

137

4—Stance Phase—i 4—- Swing Phase—5

 

 

 

 

 

 

 

0 20 40 60 80 100
Stride (%)

Figure C.7. Fox Trotter #2 carpal mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the fox trot.

4— Swing Phase ——§

 

 

 

 

 

 

 

 

 

 

o 20 4o 60 80 100
Stride (%)

Figure C.8. Fox Trotter #2 fetlock mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the fox trot.

138

6O 4—Stance Phase —-> <— Swing Phase -—>

40 r
20 r
0 6 . .
-20 ..
-40 ‘-
-60 a
-30 ..
-100

 

 

 

  
 
 

Angle (°)

 

 

 

 

o 20 4o 60 80 100
Stride (%)

Figure C.9. Fox Trotter #3 carpal mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the flat walk.

4—Stance Phase —> 4— Swing Phase —->

 

 

 

 

 

 

 

 

0 20 4o 60 80 100
- Stride (%)

Figure C. 10. Fox Trotter #3 fetlock mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (-)/ abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the flat walk.

139

60 <-—StancePhase—> <— S ’ - Phase—a

 

 

 

 

 

 

o 20 4o 60 80 100
Stride (%)

Figure Cl 1. Fox Trotter #3 carpal mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the fox trot.

d—Stance Phase —-> 4— Swing Phase —>

 

 

 

 

 

 

 

 

 

60 80 100
Stride (%)

Figure C. 12. Fox Trotter #3 fetlock mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the fox trot.

140

60 d—Stanee Phase —>

 

 

 

 

 

 

0 20 40 60 80 100
Stride (%)

Figure C. 13. Fox Trotter #4 carpal mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/extemal (+) rotation (light gay line) of the flat walk.

4—Stance Phase —> ‘— Swing Phase —>

 

 

 

 

 

 

 

0 20 40 6O 80 100

Stride (%)

Figure C. 14. Fox Trotter #4 fetlock mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the flat walk.

141

 

4o-
20-1 ,5

,__, \V‘.)
O I. .‘w M R \ .

 

-20 d x .I ‘ " MSW->31 '

.40 4
-60 .
-80 .
-100

Angle (°)

 

 

 

 

0 20 40 60 80 100
Stride (%)

Figure C. 1 5. Fox Trotter #4 carpal mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the fox trot.

4—Stance Phase—> 4— Swing Phase—v

 

 

 

 

 

 

 

 

0 20 40 60 80 100
Stride (%)

Figure C.l6. Fox Trotter #4 fetlock mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the fox trot

142

4——Stanee Phase —>

 

 

 

 

 

 

 

0 20 40 6O 80 100
Stride (%)
Figure C.l7. Fox Trotter #5 carpal mean (thick line) and SD +/- (thin line) flexion

(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the flat walk. ,

4—Stance Phase —> <— Swing Phase —’

 

60-1
40]

 

 

 

 

 

 

 

0 20 4O 60 80 100
Stride (%)

Figure C. 18. Fox Trotter #5 fetlock mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the flat walk.

143

 

60 <—-Stance Phase—9 4— SwingPhase—>

 

 

 

 

 

 

o 20 4o 60 80 100
Stride (%)

Figure C. 19. Fox Trotter #5 carpal mean (thick line) and SD +/- (thin line) flexion
(-)/ extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the fox not.

4—Stance Phase —> 4— Swing Phase —'

 

 

 

 

 

 

 

 

 

0 20 40 60 80 100

Stride (%)

Figure C.20. Fox Trotter #5 fetlock mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/externa1 (+) rotation (light gay line) of the fox trot.

144

1::

6O 4—Stance Phase-> <— SwingPhase—D

 

 

 

 

 

 

Stride (%)
Figure C.21. Fox Trotter #6 carpal mean (thick line) and SD +/- (thin line) flexion

(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the flat walk.

<—Stance Phase—-> 4— Swing Phase —>

 

 

 

 

 

 

 

 

 

0 20 4O 60 80 100
Stride (%)
Figure C.22. Fox Trotter #6 fetlock mean (thick line) and SD +/- (thin line) flexion

(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the flat walk.

145

 

 

 

 

 

 

 

o 20 40 6O 80 100
Stride (%)

Figure C.23. Fox Trotter #6 carpal mean (thick line) and SD +/- (thin line) flexion
(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the fox trot.

§—Stance Phase —> ‘— Swing Phase _’

 

 

 

 

 

 

 

 

 

O 20 40 60 80 100
Stride (%)
Figure C.24. Fox Trotter #6 fetlock mean (thick line) and SD +/— (thin line) flexion

(-)/extension (+) (black line), adduction (-)/abduction (+) (dark gay line), internal
(-)/external (+) rotation (light gay line) of the fox trot.

146

REFERENCES

147

REFERENCES

Abdel-Aziz, Y. I. and Karara, HM. (1971) Direct linear transformation from comparator
coordinates into object space coordinates into object space coordinates in close-range
photogammetry. ASP Symposium on Close Range Photogrammetry. American Society
of Photogammetry, Falls Church.

Adams, on. (1975) lameness in Horses, 3rd edn. Lea & Febiger, Philadelphia pp 331-
359,482.

Back, W., Barneveld, A., Bruin, G., Schamhardt, H.C. and Hartman, W. (1994)
Kinematic detection of superior gait quality in young trotting Warmbloods. Vet. Quart.
16, S91-96.

Back, W., Hartman, W., Schamhardt, H.C., Bruin, G. and Barneveld, A. (1995)
Kinematic response to a 70-day training period in trotting Dutch Warmbloods. Equine vet
J. Suppl. 18, 127-131.

Back, W., Gerritsen, J., Ahne, I., Gouwerok, A., Ebell, P., Klarenbeck, A. and Koning,
J.J.de. (2000) The effect of laterally wedged shoes on the sagittal and transversal plane
kinematics of the Shetland ponies. Fourth International Workshop on Animal
Locomotion. 50.

Back, W., Schamhardt, H.C. and Bameveld. A. ( 1996) Are kinematics of the walk related
to the locomotion of a warmblood horse at the trot? Vet. Quart. 18 Suppl 2, $71-$76.

Barrey, E., Galloux, P., Valette, J .P., Auvinet, B. and Wolter, R. (1993) Stride
characteristics of overgound versus treadmill locomotion in the saddle horse. Acta
Anatomica. 146(2-3), 90-94.

Budiansky, S. (1997) The Nature of Horses. Simon & Schuster Macmillan Company,
New York. pp 24-25.

Bush, B.M.H. and Clarac, F. (1985) Coordination of motor behavior. Cambridge,
Cambridge University Press. pp 201-220, 221-227.

Cappozzo, A. (1991) Three-dimensional analysis of human walking: Experimental
methods and associated artifacts. Human Movement Sc. 10, 589-602.

Cavagna, GA. and Franzetti, P. (1981) Mechanics of competition walking. J Physiol.
315, 243-251.

Chao, E.Y., Laughman, R.K., Schneider, E. and Stauffer, RN. (1983) Normative data of

knee joint motion and gound reaction forces in adult level walking. J. Biomech. 16(3),
219-233.

148

Chateau, H., Degueurce, C., Pasqui-Boutard, V. and Denoix, J .M. (2000) 3D behaviour
of the equine metacarpo-phalangeal joint. Fourth International Workshop on Animal
Locomotion. 46.

Clayton, HM. (1989) Gait analysis as a predictive tool in performance horses. J. equine
vet. Sci. 9, 335-336.

Clayton, HM. (1994) Comparison of the stride kinematics of the collected, working,
medium and extended trot in horses. Equine vet. J. 26, 230-234.

Clayton, HM. (1995) Comparison of the stride kinematics of the collected, medium and
extended walks in horses. Am J. vet. Res. 56, 849-852.

Clayton, HM. (1998) The suspension is killing me! Dressage & CT. July, 14-16.

Clayton, HM. and Bradbury, J .W. (1995) Temporal characteristics of the fox trot, a
symmetrical equine gait. Appl. Anim. Behavior Sc. 42, 153-159.

Clayton, H.M., Hodson, E., Lanovaz, J .L. and Colbome, G.R. (2000a) The hind limb in
walking horses: 2. Net joint moments and joint powers. Equine vet. J. (in press).

Clayton, H.M., Lanovaz, J.L., Schamhardt, H.C., Willemen, MA. and Colbome, GR.
(1998) Net joint moments and powers in the equine forelimb during during the stance
phase of the trot. Equine vet. J. 30, 384-389.

Clayton, H.M., Schamhardt, H.C., Lanovaz, J .L., Colbome, GR. and Willemen, M.A.
(2000b) Net joint moments and joint powers in horses with superficial digital flexor
tendinitis. Am. J. vet. Res. 61, 191-196.

Clayton, H.M., Willemen, M.A., Lanovaz, J.L. and Schamhardt, H.C. (2000c) Effects of
a heel wedge in horses with superficial digital flexor tendinitis. Vet. Comp. Ortho.
Traumatology. 13, 1-8.

Colbome, G.R., Lanovaz, J L, Sprigings, E.J., Schamhardt, H.C. and Clayton, HM.
(1997) Joint moments and power in equine gait: a preliminary study. Equine vet. J.
Suppl. 23, 33-36.

Colbome, G.R., Lanovaz, J.L., Sprigings, E.J., Schamhardt, H.C. and Clayton, HM.
(1998) Forelimb joint movements and power during the walking stance phase of horses.
Am. J. vet. Res. 59, 609-614.

Davids, J.R., Bagley, A. and Bryan, M. (1996) Kinematic and kinetic analysis of running
in children with cerebral palsy. Gait and Posture. 4, 177.

149

 

 

Degueurce, C., Chateau, H., Pasqui-Boutard, V., Geiger, D. and Denoix, J.M. (2000)
Kinematics of the proximal interphalangeal joint in the horse. Fourth International
Workshop on Animal Locomotion. 43.

Deuel, N .R. and Lawerence, L.M. (1986) Effects of velocity on gallop limb contact
variables. J. equine vet. Sci. 6, 143-146.

Drevemo, S., Dalin, D. and Fredricson, I. (1980) Equine locomotion: l. The analysis of
linear and temporal stride characteristics of trotting standardbreds. Equine vet. J. 12(2),
60-65. ~-

Ensrrringer, ME. (1977) Horses and Horsemanship, 5th edn. The Interstate Printers &
Publishers, Inc., Danville, Illinois. pp 31-44, 363-364.

Fredricson, I. and Drevemo, S.(197l) A new method investigating equine locomotion.
Equine vet. J. 3, 137-140.

Galisteo, A.M., Cano, M.R., Miro, F., Vivo, J ., Morales, J.L. and Aguera, E. (1996)
Angular joint parameters in the Andalusian horse at walk, obtained by normal
videogaphy. J. equine vet. Sci. 16, 73-77.

Gray, J. (1968) Animal Locomotion. Wiendenfeld & Nicholson, London.

Grood, E. S. and Suntay, WJ. (1983) A joint coordinate system for the clinical
description of three-dimensional motions: Application to the knee. J. Biomech. 105, 136-
144.

Harris, SE. (1993) Horse Gaits, The natural mechanics of balance and movement
common to all breeds movement. Simon & Schuster Macmillan Company, New York.

pp 102-114.

Heleski, C. (1991) The application of three-dimensional kinematic methodology to the
equine knee and ankle joints. Thesis. Michigan State University, East Lansing, MI.

Herring, L., Thompson, KN. and Jarret, S. (1992) Defining normal three—dimensional
kinematics of the lower forelimb in horses. J. equine vet. Sc. 12(3), 172-176.

Hershman, EB. and Nicholas, J.A. (1995) The Lower Extremity & Spine, Vol. 1(2).
Mosby,.St. Louis, Missouri. pp 335-337.

Hildebrand, M. (1960) How animals run. Sci. Am. 202 (MaY). 148-157.
Hildebrand, M. (1965) Symmetrical gaits of horses. Science. 150, 701-708.

Hildebrand, M. (1978) The adaptive significance of tetrapod gait selection. Annual
Meeting of the American Society of Zoologists. December, Richmond, VA. 255-267.

150

Hiraga, A., Yarnanobe, A. and Kubo, K. (1994) Relationships between stride length,
stride frequency, step length and velocity at the start dash in a racehorse. J. equine vet.
Sci. 5(4), 127-130.

Hodson, E.F., Clayton, HM. and Lanovaz, J.L. (1998) Temporal analysis of walk
movements in the Grand Prix dressage test at the 1996 Olympic Games. Appl. Anim. Beh.
Sci. 62, 89-97.

Hodson, E.F., Clayton, HM. and Lanovaz, J .L. (2000) The forelimb in walking horses: 1.
kinematics and gound reaction forces. Equine vet. J. 32(4), 287-294.

Holmstrom, M., Fredricson, I. and Drevemo, S. (1994) Biokinematic differences between
riding horses judged as good and poor at the trot. Equine vet J. Suppl. 17, 51-56.

Hoyt, DR and Taylor, CR. (1981) Gait and the energetics of locomotion in horses.
Nature. 292, 239-240.

Imus, B. (1995) Heavenly Gaits: The Complete Guide to Gaited Riding Horses.
CrossOver Publications, New York. pp 11-24.

Keegan, K.G., Wilson, D.A., Smith, B.K. and Wilson, DJ. (2000) Changes in kinematic
variables seen with lameness induced by applying pressure to the frog and to the toe in
adult horses trotting on a treadmill. Fourth International Workshop on Animal
Locomotion. 66. '

Khumsap, 8., Clayton, HM. and Lanovaz, J .L. (in press) Effect of walking velocity on
gound reaction force variables in the hind limb of normal horses. Amer. J. vet. Res.

Khumsap, 8., Clayton, HM. and Lanovaz, J.L. (2000) Effect of walking velocity on hind
limb kinetics during stance in normal horses. Fourth International Workshop on Animal
Locomotion. l9.

Lanovaz, J .L., Clayton, HM. Colbome, G.R., Schamhardt, H.C. (1999)
Forelimb kinematics and net joint moments during the swing phase of the trot. Equine
vet. J. Suppl. 30, 235-239.

Leach, DH. (1990) Normal equine locomotion. Proceedings for the Post Graduate
Committee in Veterinary Science: Equine Lameness and Foot Conditions. 25-35.

Leach, DH. and Cymbaluk, NP. (1986) Relationships between stride length, stride
frequency, velocity, and morphometrics of foals. Am. J. vet. Res. 45, 888-892.

Mann, RA. (1975) Biomechanics of the foot. In: American Academy of Orthopedic
Surgens: Atlas of Orthotics. Mosby, St. Louis.

151

 

 

SE

 

Mann, R.A. (1980) Running, jogging and walking: a comparative biomechanical study.
In: Bateman, IE and Trot, A. (eds.): The foot and ankle. Theime-Stratton, New York.

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 vet. J. 18,
207-214.

Merkins, H.W., Schamhardt, H.C., van Osch, G.J.V.M. and van den Bogert, AJ. (1993)
Ground reaction force patterns of Dutch Warmblood horses at normal trot. Equine vet. J.
25, 134-137.

Muybridge, E. (1899) Animal in motion. Republished (1957) Ed: L.S. Brown. Dover
Publications, Inc. New York.

Nicodemus, M.C., Lanovaz, J.L. and Clayton, H.M. (2000a) Forelimb kinematics and
kinetics of the fox trot. J. Anim. Sci. Suppl. 1. 78, 150.

Nicodemus, M.C., Lanovaz, J.L. and Clayton, H.M. (2000b) Temporal stride variables of
4-beat square gaits. Proceedings of the Association of Equine Sports Medicine Meetings:
Equine F itness- The Olympic Way, 329, v.

Nicodemus, M.C., Lanovaz, J.L. and Clayton, H.M. (2000c) The effect of velocity on
temporal variables of the equine walk. Conference on Equine Sports Medicine and
Science: The Elite Show Jumper, 155.

Niki, Y, Ueda, Y, Yoshida, K and Masumitsu, H. (1982) A force plate study in equine
biomechanics 2. The vertical and fore-aft components of floor reaction forces and
motion of equine limbs at walk and trot. Bull. Equine Res. Inst. 19, l-l7.

Niki, Y, Ueda, Y, Yoshida, K and Masumitsu, H. (1984) A force plate study in equine
biomechanics 3. The vertical and fore-aft components of floor reaction forces and
motion of equine limbs at the canter. Bull. Equine Res. Inst. 21, 8-18.

Parelli, P. (2000) In the Company of Horses. A&E Television Documentary.

Pascoe, E. (1999) Great Gliders. Horse & Rider. September, 83-85.

Pasquini, C., Pasquini, S., Bahr, R. and Jann, H. (1995) Guide to Equine Clinics:
Lameness Diagnosis, Vol. 3, Sudz Publishing, Pilot Point, Texas. pp 66-67, 142-143.

Peloso, J. G., J. A. Stick, R. W. Soutas-Little, J. P.Caron and C. E. DeCamp. (1991)
Computer-assisted three-dimensional gait analysis of amphotericin-induced equine carpal
lameness. Vet. Surg. 20(5), 344.

Pratt, G.W., Jr. (1983) Remarks on gait analysis. In: Equine Exercise Physiology. Snow,
D.H., Persson, S.G.B. and Rose, R.J. (eds.) Granta Editions, Cambridge. 245-262.

152

Ridings, RN. and Bradbury, J.W. (1976) The fox trotter, a public servant. West.
Horseman. 41, 6.

Rooney, J .R. (1998) The Lame Horse. The Russell Meerdink Company, Ltd, Neenah,
Wisconsin. pp 201-214.

Rubin, CT. and Lanyon, LE. (1982) Limb mechanics as a function of speed and gait: a
study of functional strains on the radius and tibia of horse and dog. J. Exp. Biol. 101, 187-
211.

Singleton, W.H., Lanovaz, J.L., Prades, ML. and Clayton, HM. (2000) Swing phase net
joint moments and joint powers under different shoeing conditions. Fourth International
Workshop on Animal Locomotion. 45.

Slade, L.M. (1993) Conformation and gait characteristics of Icelandic, Tennessee Walker
and Walkony horses. SIWAI/AESM Abstracts. 2, 8.

Soutas-Little, R.W. (1996) The use of virtual markers in human movement analysis. Gait
and Posture. 4, 176-177.

Soutas-Little, R. W., C. G. Beavis, M. C. Verstraete, and T. L. Markus. (1987) Analysis
of foot motion during running using a joint coordinate system. Med. Sci. Sports. Ex. 19,
185-293.

Speirs, V.C. (1997) Clinical Examination of Horses. W.B. Saunders Company,
Philadelphia. pp 102-103.

Taylor, CR. (1985) What determines the cost of locomotion? A closer look at what
muscles do. Proc. 5‘h Annual Scientific Meeting of Assoc. for Equine Sports Medicine.
15-30.

Weeren, P.R. van, Bogert, A.1. and Barneveld, A. (1998) Quantification of skin
displacement near the carpal, tarsal and fetlock joints of the walking horse. Equine vet. J.
20(3), 203-208.

Weeren, P.R. van, Bogert, A.1. van den, Back, W., Bruin, G. and Barneveld, A. (1993)
Kinematics of the Standardbred trotter measured at 6, 7, 8 and 9 m/s on a treadmill before
and after 5 months of pre-race training. Acta Anatomica. 146, 154-161.

Wickler, SF. (2000) Personal communications. Sydney, Australia.
Wickler, SF. and Hoyt, DP. (2000) Preferred speed and the effect of weight.

Proceedings of the Association of Equine Sports Medicine Meetings: Equine F imess- The
Olympic Way, 329, iii.

153

Wickler, S.F., Hoyt, D.F., Lewis, CC, Cogger, E.A., Zarate, 8., Bradley, Q. and
McGuire, R. (1999) Proc. 16‘” Equine Nutrition Physiol. Symp. 2425.

Wilson, B.D., Neal, R.J., Howard, A. and Groenendyk, S. (1988) The gait of pacers: I.
Kinematics of the racing stride. Equine vet. J. 20, 347-351.

Winter, DA. (1990) Kinetics: Forces and Moments of Force. In: Biomechanics and
Motor Control ofHuman Movement. 2“(1 edn. John Wiley & Sons, New York. pp 75-
102.

Winter, DA. and Robertsons, D.G.E. (1978) Joint torque and energy patterns in normal
gait. Biol. Cybernetics. 29, 137-147.

Woltring, H. J. (1994) 3-D attitude representation'of human joints: A standardization
proposal. J. Biomech. 27(12), 1399-1414.

Yanaghara, D., Udo, M., Kondo, I. and Yoshida, TA. (1993) A new learning paradigm:
adaptive changes in interlimb coordination during perturbed locomotion in decerebrate
cats. Neurosci. Res. 18, 241-244.

Zatsiorky, V.M., Werner, MS and Kaimin, M.A. (1994) Basic kinematics of walking. J
Sports Med. Phys. F imess. 34(2), 109-134.

Ziegler, L. Gaits of the gaited horse. [Online] Available email: nicodem3 @pilot.msu.edu
from smithrider7 @email.msn.com, July 26, 2000.

154

VITA

155

Department of Animal Science
MOLLY CHRISTINE NICODEMUS Michigan State University

Birthplace: Webster, Texas
DOB: 5-29- 74

HIGHLIGHTS OF PROFESSIONAL EXPERIENCE
Teaching Experience

0 Three years college level teaching: Horsemanship, Dressage, Independent Study,
Horse Selection and Judging, Advanced Horse Judging, Equine Exercise Physiology,
and Horse Management

Intercollegiate Coaching

o Coached the MSU Intercollegiate Huntseat, Stockseat, and Dressage Teams
0 Coached the MSU Horse Judging Team

Extension Experience

0 Three years of adult extension presentations: MSU Adult Riding Clinic, Upper
Peninsula Workshop, ANR Week, and Grand Rapids Expo

0 Three years of youth extension presentations: 4-H Exploration Days, Horse Jamboree,
Summer Riding Clinic, and Hoosier Horse Fair

0 Extension publications: MSU Equine Extension Newsletter (Winter 1998, Spring
1999), Journal of Mo. Fox Trotter Assoc. (Mar 1998), and Equine Times (Aug 1999)

Employment History

0 Mississippi State University-Department of Animal & Dairy Sciences 2000-
Assistant Professor

0 Michigan State University-Department of Animal Science 1997-2000
Graduate Assistant

0 Texas Twin Gates Farm 1995-1997
Assistant Barn Manager/Horse Trainer

0 Manes Tennessee Walking Horse Stables 1994-1995
Assistant Horse Trainer

0 Texas Lions Club Camp for Handicap Children 1993-1994

Director of Therapeutic Riding Program

156

In— MOLLY CHRISTINE NICODEMUS I

Academic Conferences: Presentations and Publications

0 Agriculture and Natural Resources Research Week March 1998
East Lansing, Michigan
0 Kentucky Equine Research Nutrition and Physiology Meetings April 1998

Lexington, Kentucky
6 Phi Zeta Research Day: Conference on MSU Veterinary Medicine October 1998
East Lansing, Michigan

0 Equine Nutrition and Physiology Symposium June 1999
Raleigh, North Carolina

9 Association of Equine Sports Medicine Conference February 2000
Sydney, Australia

6 Conference on Equine Sports Medicine and Science May 2000
Sicily, Italy

0 International Workshop on Animal Locomotion May 2000
Vienna, Austria

9 Annual Animal Science Meetings July 2000
Baltimore, Maryland

Education

Ph.D., Animal Science, Michigan State University, Fa112000

East Lansing, MI

Dissertation: Quantification of the Locomotion of Gaited Horses

M.S., Agricultural Sciences, Sam Houston State University, Summer 1997
Huntsville, TX
Thesis: Stride Duration Response to Weight Training in Horses

B.S., Agricultural Communications, Southwest Missouri State University, Summer 1995

Springfield, MO
Minor: Animal Science

157