<%BANNER%>

Neuromechanical and Neurophysiological Examination of Walking with and without an Ankle Foot Orthosis in Non-Injured Ind...

Permanent Link: http://ufdc.ufl.edu/UFE0021651/00001

Material Information

Title: Neuromechanical and Neurophysiological Examination of Walking with and without an Ankle Foot Orthosis in Non-Injured Individuals and Persons with Incomplete Spinal Cord Injury
Physical Description: 1 online resource (158 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: Rehabilitation Science -- Dissertations, Academic -- UF
Genre: Rehabilitation Science thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Clinicians often use orthotic devices to compensate for walking related impairments after incomplete spinal cord injury (ISCI). Orthotic devices such as an ankle foot orthosis (AFO) are commonly used to stabilize the ankle joint and aid toe clearance during walking. Compensatory stepping achieved with an AFO has led therapists to assume that such devices could be integrated in newer, neurobiologically driven, recovery-based interventions such as locomotor training (LT) for individuals with ISCI. In spite of the appeal of such compensatory strategies, their use during LT is still controversial. This is due to the lack of information about the possible effect of the device in optimizing or hindering afferent input from lower limb motion; joint, muscle and cutaneous receptors fundamental to the training. After ISCI, pattern generating neural network within the spinal cord increases its reliance on motion-related afferent input from these receptors for maintaining locomotor control. Limiting ankle excursion with an AFO may alter the interconnected limb joint assembly specific to walking and in turn influence the afferent information critical for stepping. Our study explored the therapeutic use of such devices from a walking recovery based paradigm. The aim of this project was to investigate the mechanical and neurophysiological implications of the use of an AFO during stepping in non-injured individuals and persons with ISCI. Specifically, we examined the effect of wearing a posterior leaf spring ankle foot orthosis (PAFO) on transition phase joint kinematics and kinetics and soleus H-reflex modulation during walking. In the first experiment, we examined the transition phase mechanics with and without a PAFO in healthy, non-injured individuals. Our study identified and measured the changes that occurred in normal joint kinematics and kinetics as a result of wearing a PAFO. The results suggested that proximal hip extension; crucial for the transition from stance-to-swing and the rate of loading during the swing- to-stance phase were significantly decreased. In the second experiment, we compared transition phase mechanics observed while walking with and without the PAFO in individuals with ISCI to normal mechanics. The comparison assessed the effect of the PAFO on pre-existing stepping related deficits in individuals with ISCI and also measured deviance or likeness of the change observed in these individuals from normal. The results suggested that the use of a PAFO decreased hip extension thereby impacting the provision of at least one critical afferent input key to the restoration of walking. In the third experiment, soleus H-reflexes were compared in non-injured individuals while walking with and without the PAFO in ten different phases of the gait cycle. The result showed that walking with the PAFO did not affect soleus H-reflex excitability in these individuals. In the fourth and final experiment, soleus H-reflexes were compared in the mid-stance and mid-swing phase in individuals with ISCI, while walking with and without the PAFO. A significant increase in the soleus H-reflex amplitude was observed in the mid-swing phase of walking. Our findings suggest that the PAFO increased afferent inflow and modulated reflex activity. However, increase in afferent input in the mid-swing phase of the gait cycle may not be favorable to retraining the task of walking. In summary, our results suggest that walking with a minimally restrictive PAFO alters transition phase mechanics and soleus H-reflex modulation during mid-swing phase of walking. Therefore, during LT, use of a compensatory PAFO to achieve stepping may not coincide with the principles of training.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Behrman, Andrea L.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021651:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021651/00001

Material Information

Title: Neuromechanical and Neurophysiological Examination of Walking with and without an Ankle Foot Orthosis in Non-Injured Individuals and Persons with Incomplete Spinal Cord Injury
Physical Description: 1 online resource (158 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: Rehabilitation Science -- Dissertations, Academic -- UF
Genre: Rehabilitation Science thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Clinicians often use orthotic devices to compensate for walking related impairments after incomplete spinal cord injury (ISCI). Orthotic devices such as an ankle foot orthosis (AFO) are commonly used to stabilize the ankle joint and aid toe clearance during walking. Compensatory stepping achieved with an AFO has led therapists to assume that such devices could be integrated in newer, neurobiologically driven, recovery-based interventions such as locomotor training (LT) for individuals with ISCI. In spite of the appeal of such compensatory strategies, their use during LT is still controversial. This is due to the lack of information about the possible effect of the device in optimizing or hindering afferent input from lower limb motion; joint, muscle and cutaneous receptors fundamental to the training. After ISCI, pattern generating neural network within the spinal cord increases its reliance on motion-related afferent input from these receptors for maintaining locomotor control. Limiting ankle excursion with an AFO may alter the interconnected limb joint assembly specific to walking and in turn influence the afferent information critical for stepping. Our study explored the therapeutic use of such devices from a walking recovery based paradigm. The aim of this project was to investigate the mechanical and neurophysiological implications of the use of an AFO during stepping in non-injured individuals and persons with ISCI. Specifically, we examined the effect of wearing a posterior leaf spring ankle foot orthosis (PAFO) on transition phase joint kinematics and kinetics and soleus H-reflex modulation during walking. In the first experiment, we examined the transition phase mechanics with and without a PAFO in healthy, non-injured individuals. Our study identified and measured the changes that occurred in normal joint kinematics and kinetics as a result of wearing a PAFO. The results suggested that proximal hip extension; crucial for the transition from stance-to-swing and the rate of loading during the swing- to-stance phase were significantly decreased. In the second experiment, we compared transition phase mechanics observed while walking with and without the PAFO in individuals with ISCI to normal mechanics. The comparison assessed the effect of the PAFO on pre-existing stepping related deficits in individuals with ISCI and also measured deviance or likeness of the change observed in these individuals from normal. The results suggested that the use of a PAFO decreased hip extension thereby impacting the provision of at least one critical afferent input key to the restoration of walking. In the third experiment, soleus H-reflexes were compared in non-injured individuals while walking with and without the PAFO in ten different phases of the gait cycle. The result showed that walking with the PAFO did not affect soleus H-reflex excitability in these individuals. In the fourth and final experiment, soleus H-reflexes were compared in the mid-stance and mid-swing phase in individuals with ISCI, while walking with and without the PAFO. A significant increase in the soleus H-reflex amplitude was observed in the mid-swing phase of walking. Our findings suggest that the PAFO increased afferent inflow and modulated reflex activity. However, increase in afferent input in the mid-swing phase of the gait cycle may not be favorable to retraining the task of walking. In summary, our results suggest that walking with a minimally restrictive PAFO alters transition phase mechanics and soleus H-reflex modulation during mid-swing phase of walking. Therefore, during LT, use of a compensatory PAFO to achieve stepping may not coincide with the principles of training.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Behrman, Andrea L.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021651:00001


This item has the following downloads:


Full Text





NEUROMECHANICAL AND NEUROPHYSIOLOGICAL EXAMINATION OF WALKING
WITH AN ANKLE FOOT ORTHOSIS IN NON-INJURED INDIVIDUALS AND PERSONS
WITH INCOMPLETE SPINAL CORD INJURY


















By

PREETI MOHANDAS NAIR


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2008

































2008 Preeti Mohandas Nair
































To Lord Ganesha and Shri Gajanan Maharaj for dwelling upon me at all times!!









ACKNOWLEDGMENTS

I am grateful to several individuals who have been my guiding angels throughout this

process. The greater part of this work was made possible by the instruction of my teachers and

by the love and support of my family and friends. It is with my heartfelt gratitude that I

acknowledge each of them.

First and foremost, I am thankful to my mentor, Dr. Andrea L. Behrman, for helping me

clear this milestone. I will always be indebted to her for her excellent guidance throughout and

hope that in the future I am blessed with more opportunities to work with a remarkable teacher,

researcher and human being such as herself. Also, I am grateful to my committee members, Dr.

Steven Kautz, Dr. John Rosenbek, Dr. Floyd Thompson and Dr. Paul Zehr, for their valuable and

significant contributions to better my work.

Sincere thanks to all the volunteers who have contributed significantly to these studies.

This work would not have been possible or complete without their willingness to participate.

My heartfelt thanks also to my friends from both the Locomotor training and HMPL lab with

whom I have had the privilege to work. They have been my family while I have been away from

home. Special thanks to Dr. Mark Bishop, Dr. Mary Thigpen and Dr. Claudia Senesac for their

encouragement and emotional support during the home stretch.

I extend my thanks to the University of Florida for bestowing me with the Alumni

Fellowship that financially supported my education here. Many thanks also, to the personnel in

the Department of Physical Therapy, who have helped me in one way or another during my

graduate studies.

Finally, I would like to express my deepest gratitude to my family members. My parents

have provided me unconditional love and emotional support and have been my pillars of strength

through trying times. They have sacrificed significantly to make my dreams come true and I can









only hope that I can make them proud and happy always. I am also indebted to my sister and her

family for their support and encouragement during the past few years. Last but not least, I am

grateful to GOD ALMIGHTY for dwelling upon me at all times and keeping me focused, strong

and always appreciative of the bigger good in life. This project would never have happened if it

was not for the blessings of my family and HIM.









TABLE OF CONTENTS

page

A CK N O W LED G M EN T S ................................................................. ........... ............. .....

L IST O F T A B L E S ......... .... .............. ....................................... .............................10

LIST OF FIGURES .................................. .. .. .... .... .......... ....... 12

A B S T R A C T ............ ................... ............................................................ 14

CHAPTER

1 NEUROMECHANICAL AND NEUROPHYSIOLOGICAL EXAMINATION OF
WALKING WITH AN ANKLE FOOT ORTHOSIS IN HEALTHY, NON-INJURED
INDIVIDUALS AND PERSONS WITH INCOMPLETE SPINAL CORD INJURY..........17

2 L ITE R A TU R E R E V IE W ........................................................................ ... ......................22

Overview of Human Spinal Cord Injury: Consequence and Rehabilitation.........................22
Introduction to the Problem .......................................................................... 22
Consequence of Injury ................................................................ .... ..... 22
W walking Potential after SCI............................................. ............. ............... 23
Rehabilitation of Individuals w ith ISCI ................................. ............... ....................24
Orthotic Devices: Rationale for Use and Prescription ......................................... 25
N eurobiological Control of W walking ......................................................... .............. 26
Stepping M mechanics ..................... ....... .......... ...... ...... .. .................. 30
Basic Instrum entation and Term inology ........................................ ...... ............... 30
Characterization of Stepping Pattern................ ............... ....................32
N eural A ssem bly for Stepping............................................. ................... ............... 33
Central Pattern G enerator for Stepping ........................................ ....................... 33
Sensory D rive R required for Stepping......................................... ......................... 35
H-reflex Elicitation and Modulation During Stepping............................... ...............38
E licitation ...................... ............ ..... ... ............ ... .. ............ .... ........ 3 8
Task-specific/ Phase-dependent modulation of the H-reflex .............................. 39
R recovery of W walking after SC I ........................................................... ....................... 40
Plasticity at the Level of the Spinal Cord........................................................ .......... 40
Plasticity of the Spinal Stretch Reflex (SSR) ................................................................ 42
Locomotor Training and Skill-dependent Plasticity of the Nervous System..................44
Is Orthotic use Appropriate during Locomotor Training?....................................................45
Altered Stepping Mechanics Resulting from the Use of Orthotic/Assistive Devices.....46
Altered H-reflexes Resulting from the Use of Orthotic Devices ..................................47
R ationale of the Studies........... ................................................................ ...... ..... .. 48









Methodological Considerations for the Measurement of Mechanics and Soleus H-reflex
D during Stepping .............................................................................49
M easurem ent of Stepping M echanics ........................................ ........................ 49
Study considerations .............................................. ................... .... ........ .... 49
Interpretation of kinematic and kinetic data..........................................................49
Reliability of Vicon and Force Plate Measures.............. ......... ..................50
M easurem ent of Soleus H -reflex......................... ......... ........................ ............... 52
Study considerations ............................... ......... ... .......... ... ........ 52
Interpretation of H-reflex amplitude ......................... ...............53
R liability of Soleus H -reflex.................................................. ............................ 54
Clinical and Scientific Relevance of the Study ...................................... ................. 54
C clinical R elevance of the Study ........................................................... .....................54
Scientific R elevance of the Study...................... ................................. ............... 55

3 EXAMINATION OF WALKING WITH AND WITHOUT AN ANKLE FOOT
ORTHOSIS IN NON-INJURED INDIVIDUALS AND PERSONS WITH
INCOMPLETE SPINAL CORD INJURY: A NEUROMECHANICAL PERSPECTIVE ...61

4 EXAMINATION OF WALKING WITH AND WITHOUT AN ANKLE FOOT
ORTHOSIS IN HEALTHY, NON-INJURED INDIVIDUALS-1A........................ ...64

H ypotheses.......... ...........................................................64
Methods .........................................64
Subject Selection .......................................................................... .. ......... ........................64
Experim mental Set-up ........................ .................... .. .. ........... .... ....... 65
Subject Preparation ......... ... ............ ............ .......... ......... ... .. ............ 66
P procedure ............. ........................................................................... ..... 66
D ata P ro c e ssin g ............................................................................................................... 6 7
D ata A n a ly sis ............................................................................................................. 6 8
R e su lts ......... ... ........................................................ ..................................... 6 8
D discussion ......... .... ................................................................................... ......69
L im itatio n ................................................................7 1
C o n c lu sio n ................................................................................................................. 7 1

5 COMPARISON OF WALKING WITH AND WITHOUT ANKLE FOOT ORTHOSIS
IN PERSONS WITH INCOMPLETE SPINAL CORD INJURY-1B ................................79

Specific Aim s and Hypothesis............................................. 79
Methods .........................................80
Subject Selection .......................................................................... .. ......... ........................80
Experimental Set-up ...................................... ...................81
Subject Preparation .............................................. ............. ........... 82
P procedure ............. ......... .................................................................. ..... 82
D ata P ro c e ssin g ............................................................................................................... 8 3
D ata A n a ly sis ............................................................................................................. 8 4
R e su lts ...... ........ ... ... .. ..................................... ..................................... 8 4
Within-Subject Comparisons for Individuals with ISCI ...........................................84









Between-Subject Comparisons for Individuals with ISCI Walking Without a PAFO
and T heir M watched C control ................................................................................ ...85
Between-Subject Comparisons for Individuals with ISCI Walking with a PAFO
and Their M watched C control ................................................. .... ....................... 85
Temporal and Spatial Comparisons Within Subjects...................................................85
D iscu ssio n ................... ...................8...................5..........
Clinical Im plication .................................................................... .. ..... .. 87
L im ita tio n s ................................................................................................................. 8 7
C o n c lu sio n ....................................................................................................................... 8 8

6 PHASE DEPENDENT MODULATION OF SOLEUS H-REFLEX IN HEALTHY,
NON-INJURED INDIVIDUALS WHILE WALKING WITH AN ANKLE FOOT
O R T H O S IS ......... .... .. ....... .............................................................................................. 10 3

In tro d u ctio n ..........................................................................................................1 0 3
Sp ecific A im s......................................................104
M methods ................ ... ..............................................................105
Subject Selection .......................................................................... ......... ........ ......... 105
Experimental Set-up ........................................................................ ......... .................. 105
Subject Preparation ...................................................... 106
Procedure ......... .......... ................................................... 106
D ata P ro c e ssin g ............................................................................................... 10 8
D ata Analysis ...................... ........ .......................................................................108
R e su lts ..........................................................................................................1 0 8
D iscu ssion .......... ..... .... ............. ............................................109
L im station s ........................................................................................111
C o n c lu sio n ............................................................................................................... 1 1 2

7 IMMEDIATE, PHASE DEPENDENT, SOLEUS H-REFLEX MODULATION IN
PERSONS WITH INCOMPLETE SPINAL CORD INJURY WHILE WALKING
W ITH AN AN KLE FOOT ORTH O SIS .......................................................................... 119

In tro d u ctio n ................... ...................1.............................9
Sp ecific A im s......................................................12 0
M methods ................ ... ..............................................................12 1
Subject Selection ................................. ....................................... 121
Experim mental Set-up ......................................... .................. ................. 121
Subject P rep aration ............................................................................... 122
P ro c e d u re ................................................................................12 3
D ata P ro c e ssin g ................................................................... .................................. 12 4
D ata A n a ly sis ........................................................................................................... 12 4
R e su lts ................... ...................1.............................5
D iscu ssio n ................... ................................................2 5
C clinical Im plication s ............................................................... 127
L im station s ..................................................... 12 8
C o n c lu sio n s ............................................................................................................. 12 8



8









L IST O F R E F E R E N C E S ............................................................................. ..........................138

B IO G R A PH IC A L SK E T C H ......................................................................... .. ...................... 158





















































9









LIST OF TABLES


Table page

2-1 Coefficient multiple correlations (CMC) reflecting stride to stride variability between
and w within days. .............................................................................60

4-1 Demographics of the study participants.................... ..... ... .. ................ ..77

4-2 Hip, knee and ankle joint kinematic and kinetic data while walking with and without
an ankle foot orthosis (AFO) during the stance-to-swing phase of the gait cycle.............77

4-3 Hip, knee and ankle joint kinematic kinematic and kinetic data while walking with
and without an ankle foot orthosis (AFO) during swing-to-stance phase of the gait
cycle ...........................................................................78

4-4 Average interlimb temporal and spatial data while walking with and without an ankle
foot orthosis (A F O ) .............................................................. ...................... ... 78

5-1 Participant demographics of individuals with ISCI and control subjects.........................95

5-2 Hip, knee and ankle joint kinematic and kinetic data during the stance-to-swing
phase of the gait cycle while walking with and without an ankle foot orthosis (AFO)
in individuals with incomplete spinal cord injury.................................. ............... 96

5-3 Hip, knee and ankle joint kinematic and kinetic data during the swing-to-stance
phase of the gait cycle while walking with and without an ankle foot orthosis (AFO)
in individuals with incomplete spinal cord injury.................................. ............... 97

5-4 Kinematic and kinetic data during the stance-to-swing phase of the gait cycle at the
hip, knee and ankle joints while walking without an ankle foot orthosis (AFO) in
individuals with incomplete spinal cord injury compared to their matched, non-
injured controls. ......................................................... ................. 98

5-5 Kinematic and kinetic data during the swing-to-stance phase of the gait cycle at the
hip, knee and ankle joints while walking without an ankle foot orthosis (AFO) in
individuals with incomplete spinal cord injury compared to their matched, non-
injured controls. ......................................................... ................. 99

5-6 Changes in the hip, knee and ankle joint kinematics and kinetics during the stance-
to-swing phase of the gait cycle while walking with an ankle foot orthosis (AFO) in
individuals with incomplete spinal cord injury compared to their matched, non-
injured controls. ..........................................................................100

5-7 Kinematic and kinetic changes at the hip, knee and ankle joints during the swing-to-
stance phase of the gait cycle while walking without an ankle foot orthosis (AFO) in
individuals with incomplete spinal cord injury compared to their matched, non-
injured controls. ..........................................................................101









5-8 Average interlimb temporal and spatial data while walking with and without an ankle
foot orthosis (AFO) in individuals with incomplete spinal cord injury...........................102

6-1 Demographics of non-injured participants recruited for the study. ...............................118

7-1 P participant dem graphics .............................................................................. ...............135

7-2 Ipsilateral H/M ratio with and without the AFO.................................. 135

7-3 Contralateral H/M ratio with and without the AFO..................... .............................. 136

7-4 Normalized soleus EMG amplitude with and without AFO ipsilaterally......................136

7-5 Normalized TA EMG amplitude with and without AFO ipsilaterally. .........................136

7-6 Normalized soleus EMG amplitude with and without AFO contralaterally....................137

7-7 Normalized TA EMG amplitude with and without AFO contralaterally......................137









LIST OF FIGURES


Figure page

2-1 Reduced animal preparation showing the impact of transaction at different levels of
the nervous system .................................. ... .... .. .. .......... ........ 56

2-2 Stance-to-swing and swing-to-stance phases of the gait cycle. ....................................57

2-4 Neural control of locomotion in an intact nervous system (A) and compromised
nervous system (B ).. ........................ ...... .................... .. .......................58

2-5 Electrical stimulus (shown here by the grey ellipse) applied to the mixed nerve
conducts the stimuli orthodromically in the motor and sensory axons to evoke the M-
wave and the H-reflex respectively...................... ...... ............................. 59

4-1 Participant with safety harness walking on an instrumented treadmill............................72

4-2 Ipsilateral and contralateral average joint angles with and without the ankle foot
orthosis (A FO ) ipsilaterally.. .............................. ... ......................................... 73

4-3 Stick figure representing the changes in individual joint motion in the stance-to-
swing and swing-to-stance phase of the gait cycle with and without the ankle foot
orthosis (A FO ). ............................................................................74

4-4 Ipsilateral and contralateral average joint powers during the stance-to-swing phase of
the gait cycle with and without the ankle foot orthosis (AFO) ipsilaterally...................75

4-5 Ipsilateral and contralateral vertical and horizontal (AP) ground reaction forces
(GRF) during the swing-to-stance phase of the gait cycle with and without the ankle
foot orthosis (A FO ) ipsilaterally............................................... ............................. 76

5-1 Ipsilateral and contralateral average joint angles during the swing-to-stance and
stance-to-swing phase of the gait cycle with and without the ankle foot orthosis
(A F O ) ipsilaterally ...............................................................................89

5-2 Ipsilateral and contralateral average joint powers during the swing-to-stance and
stance-to-swing phase of the gait cycle with and without the ankle foot orthosis
(A F O ) ipsilaterally ...............................................................................90

5-3 Ipsilateral and contralateral vertical and horizontal (AP) ground reaction forces
(GRF) during the swing-to-stance phase of the gait cycle with and without the ankle
foot orthosis (A FO ) ipsilaterally............................................... ............................. 91

5-4 Ipsilateral hip extension values with and without the AFO during the stance-to-swing
phase of the gait cycle for spinal cord injured individuals and their matched controls.....92









5-5 Step length while walking with and without the AFO in individuals with incomplete
spinal cord injury. .......................................... ............................ 93

5-6 Double limb support time while walking with and without the PAFO in individuals
with income plete spinal cord injury. ............................................................................ 94

6-1 Experimental design for the examination of changes in soleus H-reflex amplitude in
healthy, non-injured individuals while walking with and without an ankle foot
orthosis (AFO). .................................... ................................. ..........113

6-2 Ipsilateral raw soleus H-reflex data while walking with and without AFO at 300ms
from heel strike (H S) and toe-off (TO)................................................. ....... ........ 114

6-3 Contralateral raw soleus H-reflex data while walking with and without AFO at
300ms from heel strike (HS) and toe-off (TO). ............ .............................................. 114

6-4 Ipsilateral mean H-reflex amplitudes with and without AFO normalized to M-max in
each phase of the gait cycle.. ............................................................................ .... 115

6-5 Contralateral mean H-reflex amplitudes with and without AFO normalized to M-max
in each phase of the gait cycle.. ........................................................................... .. .. 115

6-6 Ipsilateral [A] and contralateral [B] M-max amplitude with and without the AFO
across the gait cycle. .................................................................. .... .. ........... 116

6-7 Ipsilateral [A] and contralateral [B] actual M wave amplitude used to evoke the
soleus H-reflex with and without the AFO across the gait cycle...................................117

7-1 Experimental design for testing the effect of walking with and without an ankle foot
orthosis (AFO) in individuals with incomplete spinal cord injury (ISCI) at their self-
selected (SS) w walking speed.................................................. ............................... 130

7-2 Average H/M ratio values with and without an AFO in mid-stance and mid-swing
phase of walking relative to static standing in the ipsilateral limb..................................131

7-3 Average H/M ratio values with and without AFO in mid-stance and mid-swing phase
of walking relative to static standing in the contralateral limb.............. ..................131

7-4 Ipsilateral and contralateral raw soleus H-reflex data while walking with and without
AFO during mid-stance (MSt) and mid-swing (MSw) phase of the gait cycle.............132

7-5 Ipsilateral [A] and contralateral [B] M-max amplitude with and without the AFO in
the mid-stance and mid-swing phase of walking .......................................................... 133

7-6 Ipsilateral [A] and contralateral [B] actual M wave amplitude used to evoke the
soleus H-reflex with and without the AFO across the gait cycle...................................134









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

NEUROMECHANICAL AND NEUROPHYSIOLOGICAL EXAMINATION OF WALKING
WITH AN ANKLE FOOT ORTHOSIS IN NON-INJURED INDIVIDUALS AND PERSONS
WITH INCOMPLETE SPINAL CORD INJURY

By

Preeti Mohandas Nair

May 2008

Chair: Andrea L. Behrman
Major: Rehabilitation Science

Clinicians often use orthotic devices to compensate for walking related impairments after

incomplete spinal cord injury (ISCI). Orthotic devices such as an ankle foot orthosis (AFO) are

commonly used to stabilize the ankle joint and aid toe clearance during walking. Compensatory

stepping achieved with an AFO has led therapists to assume that such devices could be

integrated in newer, neurobiologically driven, recovery-based interventions such as locomotor

training (LT) for individuals with ISCI. In spite of the appeal of such compensatory strategies,

their use during LT is still controversial. This is due to the lack of information about the possible

effect of the device in optimizing or hindering afferent input from lower limb motion; joint,

muscle and cutaneous receptors fundamental to the training. After ISCI, pattern generating

neural network within the spinal cord increases its reliance on motion-related afferent input from

these receptors for maintaining locomotor control. Limiting ankle excursion with an AFO may

alter the interconnected limb joint assembly specific to walking and in turn influence the afferent

information critical for stepping. Our study explored the therapeutic use of such devices from a

walking recovery based paradigm.









The aim of this project was to investigate the mechanical and neurophysiological

implications of the use of an AFO during stepping in non-injured individuals and persons with

ISCI. Specifically, we examined the effect of wearing a posterior leaf spring ankle foot orthosis

(PAFO) on transition phase joint kinematics and kinetics and soleus H-reflex modulation during

walking. In the first experiment, we examined the transition phase mechanics with and without a

PAFO in healthy, non-injured individuals. Our study identified and measured the changes that

occurred in normal joint kinematics and kinetics as a result of wearing a PAFO. The results

suggested that proximal hip extension; crucial for the transition from stance-to-swing and the

rate of loading during the swing- to-stance phase were significantly decreased. In the second

experiment, we compared transition phase mechanics observed while walking with and without

the PAFO in individuals with ISCI to normal mechanics. The comparison assessed the effect of

the PAFO on pre-existing stepping related deficits in individuals with ISCI and also measured

deviance or likeness of the change observed in these individuals from normal. The results

suggested that the use of a PAFO decreased hip extension thereby impacting the provision of at

least one critical afferent input key to the restoration of walking.

In the third experiment, soleus H-reflexes were compared in non-injured individuals while

walking with and without the PAFO in ten different phases of the gait cycle. The result showed

that walking with the PAFO did not affect soleus H-reflex excitability in these individuals. In the

fourth and final experiment, soleus H-reflexes were compared in the mid-stance and mid-swing

phase in individuals with ISCI, while walking with and without the PAFO. A significant increase

in the soleus H-reflex amplitude was observed in the mid-swing phase of walking. Our findings

suggest that the PAFO increased afferent inflow and modulated reflex activity. However,

increase in afferent input in the mid-swing phase of the gait cycle may not be favorable to









retraining the task of walking. In summary, our results suggest that walking with a minimally

restrictive PAFO alters transition phase mechanics and soleus H-reflex modulation during mid-

swing phase of walking. Therefore, during LT, use of a compensatory PAFO to achieve stepping

may not coincide with the principles of training.









CHAPTER 1
NEUROMECHANICAL AND NEUROPHYSIOLOGICAL EXAMINATION OF WALKING
WITH AN ANKLE FOOT ORTHOSIS IN HEALTHY, NON-INJURED INDIVIDUALS AND
PERSONS WITH INCOMPLETE SPINAL CORD INJURY

Walking is a fundamental motor function of human beings. It consists of three neurally

controlled and coordinated tasks that allow 1) generation of a basic reciprocal, stepping pattern

required for propulsion of the body 2) maintenance of equilibrium during propulsion and 3)

adaptability of the walking pattern to the environment and to the behavioral goals of the

individual.1-3 Humans can move around from one place to another, maintain an upright posture,

interact with the environment and perform a flurry of activities characteristic to human nature

due to their ability to walk. Although walking is an essential element in daily living, its

importance is usually only recognized when it is impaired or lost.

Spinal cord injury (SCI) is a debilitating condition resulting in walking impairment or

inability secondary to deficits in voluntary strength (of the limbs and trunk) and sensation.4'5

Although all persons with SCI express a desire to walk only twenty five to thirty three percent of

these individuals regain the ability to do so.6-9 Persons with incomplete spinal cord injury (ISCI)

have a greater potential for walking recovery compared to individuals with complete injury due

to some sparing of motor and sensory function below the level of the lesion.10 However, walking

in persons with ISCI may be slow with asymmetrical steps, flexed posture and impaired balance

and adaptability.

Current rehabilitation strategies after SCI are based on the assumption that deficits due to

SCI are irremediable from surgical, medical, or therapeutic means.4'11 In particular, the spinal

cord, being a hard-wired conduit of information from supraspinal structures to the muscles is

viewed post-injury as irreparable and the deficits permanent and irreversible.1113 To compensate

for irremediable deficits post-SCI, therapists employ braces, assistive devices, and wheelchairs to









teach new behavioral strategies to achieve mobility.4'14-16 Assistive devices, such as canes,

crutches, and walkers provide upright support through upper extremity weight bearing to

compensate for lower limb and trunk muscle weakness. Likewise, single joint (ankle foot

orthosis-AFO) or multi joint (knee ankle foot orthosis-KAFO) are used to stabilize joints and aid

stepping.16 For example, posterior leaf spring ankle foot orthosis (PAFO) is often used to

compensate inadequate toe clearance and loss of heel strike during stepping.16,17

In contrast to the assumption of ceased neurological function below the level of injury,

neuroscientists examining the neurobiological control of walking in both animals and humans

have provided convincing evidence for recovery. 18-26 Evidence suggests that the adult

mammalian spinal cord is plastic and is known to reorganize after injury when provided with the

appropriate stimulus that is intense, task-specific and repetitive in nature.3'22'27-29 Reciprocal

stepping, for example, defined as the repetitious, mechanical sequence of limb motion and

weight shift such that, one limb maintains contact with the ground and supports the body while

the other limb swings forward is facilitated by a host of afferent inputs that modulate the

transition from support-to-swing and vice-versa.30-32'33

Afferent inputs for the neurobiological control of stepping involve motion related changes

in joint position, muscular force and limb load sensed by joint, muscle and cutaneous receptors

in the lower extremity.34-37 Evidence from animal and human studies indicate that terminal hip

extension and unloading are critical sensory inputs required for the afferent initiation of the

transition from support-to-swing.31'34'38 Limiting hip extension and/or transfer of weight on either

limb delays the transition from support-to-swing and swing-to-support and limits forward

progression of the body.









Evolving rehabilitation approaches such as locomotor training have effectively integrated

the provision of afferent input required for the neural control of stepping with repeated task

specific practice to retrain the ability to step in individuals with ISCI.25,37,39,40 Using a treadmill,

body weight support and manual assistance the training facilitates repeated cueing of critical

motion related inputs including hip extension and limb unloading required for stepping.24-26'41-43

The repeated volley of motion related input via the training can drive the pattern generating body

within the spinal cord, induce reorganization of the nervous system and facilitate the recovery of

walking in persons with SCI having limited access to supraspinal input.25'26,44 Therefore, this

training strategy personifies the intrinsic repair potential of the spinal cord which can be tapped

into using appropriate physical rehabilitation strategies to facilitate recovery.

Ironically, addition of orthotic devices as training variables to the physiological-based

training may hinder rather than facilitate recovery.24 Single joint, rigid orthotic devices such as

AFOs might restrict the range of excursion of the distal and proximal linked joints and alter the

afferent information related to joint position and load pivotal for stepping.24 Furthermore, the use

of an assistive/orthotic device might change the basic stepping pattern thereby modifying the

afferent information utilized for stepping. In addition, the acquisition of a new motor skill such

as walking with an orthotic device could induce a new pattern of skill-dependent plasticity that

could impact their ability to reacquire normal walking potential.45 Thus the role of such

compensatory strategies in promoting use of the intrinsic mechanisms for stepping lack

mechanical and neurophysiological evidence justifying their use and need to be investigated.24'46

The overall purpose of this study was therefore to examine the mechanical and

neurophysiological effect of walking with one such single joint, range limiting orthotic device

the AFO. In this four part project, I examined the effect of wearing an AFO on 1) immediate









phase specific H-reflex modulation and 2) walking mechanics in healthy, non-injured individuals

and 3) immediate phase specific H-reflex modulation and 4) walking mechanics in persons with

ISCI.

In these experiments, healthy, non-injured individuals and persons with ISCI walked on an

instrumented treadmill at speeds close to normal walking speeds. Kinematic, kinetic and H-reflex

data was collected while these individuals walked with and without wearing an AFO.

Specifically, the mechanical changes in joint position, joint powers, vertical and horizontal

ground reaction forces in the transition phases were examined in these experiments accompanied

by the neurophysiological changes in the soleus H-reflex amplitude. Interpretation of the

mechanical and neurophysiological data served as a critical first step in interpreting how

restricting lower extremity single joint excursion while walking with a AFO modified the

consequent sensory information i.e. hip joint position and load required for reciprocal stepping.

In summary, these studies enhanced our understanding of the impact of altered walking

mechanics associated with these compensatory devices on the neurally controlled task of

reciprocal stepping in healthy, non-injured individuals and persons with SCI. The findings of this

study will be useful in 1) clinical decision-making for the use of such devices in physiological-

based training interventions and 2) designing neuromechanically compatible assistive and

orthotic devices.

The following literature review is composed of sections that will serve to orient the reader

to the foundation principles underlying the purpose of this project. An overview of SCI related

walking impairment and the use of orthotic devices to compensate for impairment is described

first. Juxtaposed are the neurobiological control of walking and the subtask of stepping

highlighting the mechanical and neural characteristics of stepping. The newer evolving recovery









based intervention called locomotor training that has stemmed from our knowledge of the neural

control of walking and plasticity of the nervous system is described next. This section is

followed by the gap in knowledge pertaining to the use of an AFO from a neural control

standpoint. Methodological considerations for investigation of an AFO and interpretation of

results in relation to the neural control of stepping is described that will provide the basis for the

experimental paradigm used in this project. The clinical and scientific relevance of the studies

are discussed in the final section.









CHAPTER 2
LITERATURE REVIEW

Overview of Human Spinal Cord Injury: Consequence and Rehabilitation

Introduction to the Problem

Spinal cord injury (SCI), an injury to the neural elements within the spinal cord, results in

a multitude of dysfunction, loss of sensation and motor function being the most profound.47'48

Injury could occur as a result of motor vehicular accidents, diving accidents that fractures,

dislocates or compresses the vertebrae protecting it.7'49 or may also result from a gunshot50 or

knife wound51 that penetrates and cuts the cord. Additionally, secondary damage usually occurs

with a traumatic injury as a result of bleeding, swelling and inflammatory processes that

compress the cord.52 Non-traumatic injuries to the cord occur as a result of tumors, vascular

problems, spina bifida and several other conditions.53-55

Although there is a significant rate of mortality associated with injury, survival after SCI

has improved considerably because of efficient critical care and improved urinary rehabilitation

and respiratory management.5256 About 253,000 people currently in the US live with SCI and

there are an additional 11,000 new cases every year.57 The current ten-year survival rate of spinal

cord injured patients is approximately 86% of normal.57

Consequence of Injury

The effects of SCI vary according to the level and type (complete or incomplete) of

injury.58'59 In a complete injury, there is bilateral, total sensory and motor loss below the level of

injury. A person with an incomplete SCI (ISCI) retains some sensation below the level of injury.

Incomplete injuries are variable, and a person with such an injury may have patchy motor

involvement such as he or she might be able to move one limb more than another, may be able to

feel parts of the body that cannot be moved, or may have functioning on one side of the body









more than the other. Depending on the level of injury, paralysis can involve all four extremities a

condition called quadriplegia or tetraplegia occurring as a result of cervical injuries or only the

trunk and lower body a condition called paraplegia occurring as a result of injury at the thoracic

level or below.

The American Spinal Injury Association (ASIA) impairment scale (AIS) is used to classify

the level and severity of injury in relation to the loss of sensation and motor function.58'60 The

scale consists of five categories that classify sensory and motor function. ASIA A is defined as a

complete injury with no motor or sensory function preserved in the sacral segments S4-S5. ASIA

B, C and D are all incomplete injuries but classification varies based on the level of motor

involvement below the level of the lesion. ASIA B is defined as an incomplete injury with some

sensory but no motor function preserved below the level of injury including the sacral segments

S4-S5. To be classified as an ASIA C more than half of the muscles are graded less than 3/5

voluntary strength. ASIA D is defined as an incomplete injury with at least half of the muscles

graded more than 3/5. A person with SCI is classified as an ASIA E if he/she has no

neurological deficits that are detectable on a neurological examination of this type.

Walking Potential after SCI

Walking ability after SCI has been defined in the literature in several different ways. For

example, studies have characterized it based on the ability to ambulate upright fifty feet without

assistance61 or the ability to walk in the community or in the household8'9 or the ability to walk

reciprocally for at least two hundred feet with/without orthotics or assistive devices62 or walking

functional independence measure of > 3/7.63 The ability to walk ranks as one of the top five

priorities of individuals with SCI based on the level and severity of injury but interestingly only

25-33% of these individuals are able to do so.52'6'64 Key predictors of ambulation potential after

SCI include: ASIA score D or E at admission,8'9 age,6164 ASIA lower extremity motor scores









greater than or equal to 10 by one month,65 manual muscle testing score in the quadriceps greater

than 2/566 and sparing of pin prick sensation below the level of injury.62 Based on these factors, a

significant number of persons who regain the ability to walk are persons with ISCI compared to

persons with complete SCI.10,61 Maynard et al. (1979), for example, reported that of 123 patients

with incomplete sensory deficits 72 hours after injury, 47% were ambulatory and 87% of the

patients with incomplete motor lesions were walking at one year.67

Although upright mobility may be spared or achieved with assistance, walking is typically

impaired in persons with ISCI as a result of varying levels of muscular paralysis, sensory

deficits, spasticity and poor trunk control.68-70 Gait in an individual with ISCI is often

characterized by one or combination of the following deviations (i) inadequate active hip

extension during stance; (ii) limited hip flexion; (iii) limited knee flexion; (iv) excess ankle

plantar flexion during swing; and (v) impaired initial foot contact.5 Consequently, these

individuals are often seen taking slow, asymmetrical and uncoordinated steps over a wide base of

support and having limited adaptability to the environment.5'71

Rehabilitation of Individuals with ISCI

The International Classification of Functioning, Disability and Health (ICF) is the

framework developed by the World Health Organization (WHO) to describe functioning and

disability at both the individual and population levels.72 Conventional physical therapy

interventions for improving walking function in persons with ISCI targets two main domains of

the model namely, body function/ structures and activities. At the level of body function and

structures, the level and severity of injury using the ASIA scores is assessed and interventions

maximizing residual muscle strength and endurance in muscles that can be voluntarily activated

above and below the lesion are implemented.14,15 Similarly, interventions at the activity level

emphasize the use of assistive and orthotic devices to improve ambulation potential14'16 and teach









new strategies for upright mobility.14,15,73 Patient performance at this level however is a critical

indicator of type and extent of orthotic supports and the patient's tolerance to ambulation.

Orthotic Devices: Rationale for Use and Prescription

Orthotic devices range from simple single joint braces such as the ankle foot orthosis

(AFO) to multi joint orthotics such as reciprocating gait orthosis (RGOs).15,74,75 The prescription

of these devices varies based on several factors including the ASIA impairment score of motor

complete or incomplete made by the clinician to the needs and desires of the client.74 The goals

of prescribing orthotic devices for walking are to support the paralyzed or weakened

musculoskeletal structure, add stability to joints, improve mobility, correct alignment and

improve overall functional independence.16 Posterior leaf spring ankle foot orthosis (PAFOs) are

usually prescribed for higher functioning individuals with ISCI to provide support for weakened

musculature around the ankle joint.16,76 Since it supports one single joint it is considered least

cumbersome and called a device of minimal assistance. The guiding principles for

recommendation are to control the ankle joint by limiting excursion range, provide safe joint

mechanics, prevent toe drag during the stance-to-swing transition, minimize the risk of falls and

enhance the ability to walk faster and efficiently.16,17

The current rationale for use of orthotic devices for persons with ISCI has stemmed from

the hierarchical model of motor control that has been well accepted by rehabilitation

practitioners and continues to serve as the basis for rehabilitation to date.11'12 The model portrays

a hard-wired, immutable central nervous system (CNS) that controls all voluntary movements by

sending commands from the cerebral cortex to the periphery. The spinal cord serves simply as a

cable between the brain and the peripheral musculature receiving stimuli from the periphery and

relaying cortical commands to the periphery. Therefore if the spinal cord is injured, the damage

is considered irreparable, non-malleable and permanent.12'77 Conventional rehabilitation









therefore typically consists of a compensatory approach to deal with walking impairments in

persons after ISCI.78 The approach utilizes the use of compensatory orthotic devices that utilize

other spared abilities to accomplish the task or modify the task and/or the environment to make it

easier for a person to accomplish the goal.79

Neurobiological Control of Walking

The prevailing assumption that neural recovery is not possible following SCI has led to the

aggrandizement of orthotic devices that compensate for walking impairments.80-82 Compelling

evidence from neuroscience examining the neural control of walking however contradicts this

assumption.31'37'8386 Neuroscientists have investigated the role of the nervous system,

particularly the spinal cord to adapt and reorganize after complete transactions at the level of the

cord.20'87'88 The ability of the spinal cord to respond to peripheral sensory input, generate and

modulate rhythmic activity in the lower limbs and reorganize after injury make it a viable

substrate for intervention.31'84'89'90 The subsequent sections expand on the role of the spinal cord

in the control of walking and its plasticity after injury that have led to a proposed paradigm shift

in SCI rehabilitation from compensation-based to a recovery-based model.24

Conceptual models portray walking as three neurally controlled and coordinated tasks: 1)

Generation of a basic reciprocal, stepping pattern required for propulsion of the body, 2)

Maintenance of equilibrium during propulsion and 3) Adaptability of the walking pattern to the

environment and to the behavioral goals of the individual.1'3 However, convergence and

processing of multiple afferent inputs occurs at every level of the nervous system to bring about

the smooth, patterned orchestration of several joints and muscular synergies that characterize

walking.91-93 While walking, a single alpha motor neuron might receive as many as ten thousand

inputs.94 Determining which input has a greater relative influence on sculpting the locomotor

pattern has been a challenging task in the complex human nervous system. Neuroscientists have









therefore utilized reduced, non-human, vertebrate preparations such as cat models to study the

influence of exclusive afferent inputs on locomotor behavior and extrapolated their research

findings to human control of locomotion.31'86'95 The premise for extrapolation being the

locomotor framework which is remarkably similar throughout the vertebrate phylum, in spite of

the form of locomotion which is species specific.84'96

For example, the basic principle of organizing rhythmic stepping behavior is similar in

humans compared to other non-human vertebrates.9597 In a spinalized cat model the reflex or

automatic behavior associated with walking i.e. reciprocal stepping does not require control by

cerebral cortex rather it is controlled by subcortical and spinal centers which are subject to

cortical intervention.98 Reciprocal stepping is generated at the level of the spinal cord as long as

the weight is supported and the ground is moving under their feet. Assemblies of premotor

interneurons in the spinal cord are synaptically interconnected with each other and with

motorneuronal pools that are capable of sustaining alternating movements required for walking.99

These networks of neurons and interneurons are called central pattern generators (CPGs).100 The

central pattern generators simplify the control of locomotion by harnessing the large degrees of

freedom and provide the basic framework for walking.99'101'102 Sensory input driven from the

periphery is known to control the rhythm of walking by making the required phase transitions,

shaping the pattern of activity and reinforcing ongoing activity. Sensory information from

muscle spindles which are sensitive to changes in the muscle length,103 golgi tendon organs that

are responsive to muscle tension104,105 and flexor reflex afferents involving the

mechanoreceptors, cutaneous afferents and nociceptors is processed at the level of the CPG.104,105

When the task is consistent and unaltered, like during stable state walking, the spinal cord









demonstrates remarkable autonomy in producing the reflexive stepping pattern without much

cortical control.

Neural descending pathways from the brain to the cord are predominantly involved in fine-

tuning this pattern. For example, the corticospinal tracts influence the locomotor performance by

acting monosynaptically or oligosynaptically on the alpha-motorneurons or indirectly via

connections on the CPG.99 Thus, selective muscle control required for fine tuning and/ or

modulation of locomotor synergy i.e. speed of locomotion could be achieved. Also, the

excitability of the CPGs is governed strongly by locomotor centers in the midbrain and brainstem

i.e. mesencephalic, pontine or subthalamic locomotor regions106-108 that are under limbic and

cortical control and that dictate purposeful locomotion such as starting or stopping.

Similarly, higher brain centers such as the cortex, cerebellum98 and basal ganglia109 are

responsible for integrating afferent information from different sensory sources with cortical

motor commands. Integration of the sensory input with the motor output facilitates control of the

other two subtasks i.e. maintaining balance and adapting to the environment by engaging

corrective and reflexive postural mechanisms required for walking. When the cerebellum, the

brainstem and the spinal cord are spared in cats (decerebrate preparation), it is seen that the

animal is able to generate rhythmic activity, support body weight and propel itself. However, the

other two subtasks for successful locomotion i.e. dynamic balance and adaptability to the

environment are noticeably deficient2'102 (Figure 2-1).

Dynamic balance in the decerebrate cat requires integration of sensory systems: visual,

kinesthetic and vestibular system.7 Walking in a cluttered environment for example requires

integration of information from visual, somatosensory and vestibular inputs to maintain balance

and perform the task of walking successfully.110 The reticulospinal, rubrospinal, vestibulospinal









and corticospinal tracts111,112 are descending spinal tracts that project motor commands for fine

postural adjustment to the spinal cord after processing afferent information. Corrective postural

responses are then integrated and adjusted to the current state or phase of stepping by the

CPG.113

Lack of cortical control above the level of transaction also leads to a loss of adaptability to

the environment and to the person's behavioral goals (Figure 2-1). The motor cortex and the

basal ganglia are important for "skilled" locomotion in which the feet must be guided to establish

firm contact with narrowly specified points in the environment.97 Studies on decorticate cats

showed that although the loss of the cortex has minimal impact on the locomotor process, the

context in which the locomotor movements are performed is affected.84'114 For example,

decorticate cats exhibit limited range of options in locomotor movements and are hyperactive to

stimuli that would tend to elicit a minimal response. Likewise, ablation of the caudate nucleus of

the basal ganglia in cats results in the animal following anything that moves termed "compulsory

approach syndrome"; while diencephalic cats (whose thalamus and hypothalamus have been

removed) demonstrate "obstinate progression" i.e. walking into obstacles and not attending to

environmental stimuli.114 Therefore cortical centers and the basal ganglia are important for

adaptive control of movement to the environment.

While the different levels of control exerted by the nervous system are important for the

coordination of the overall task of walking, the proposed studies focus on one of the subtasks of

walking i.e. stepping. The next two sections elaborate on characterizing the stepping pattern

mechanically and describing the neural assembly required for stepping.









Stepping Mechanics

Basic Instrumentation and Terminology

Stepping is a motor task that obeys laws governing static and dynamic bodies and can be

quantified using our knowledge of basic mechanics. Kinematics is the branch of mechanics

dealing with the motion of body segments without being concerned with the forces that cause the

motion.115 Kinematic analysis, using automated motion analysis systems measures positions,

angles, velocities, and accelerations of body segments and joints during motion. Joint angle (also

called inter-segmental angle) is defined as the angle between the two segments on either side of

the joint, usually measured in degrees. 116117 For the interconnected chain of segments involving

the ankle, knee and hip joints of the lower limbs, joint angles are particularly useful in

determining the relative motion of one joint with respect to the other. However, these

measurements only describe the motion performed and are limited in what they can tell us about

the cause of the motion.

Kinetic analysis on the other hand measures forces acting at a particular joint, segment or

body as a whole that cause the specific walking pattern.116,117 Forces in walking can be internal

such as muscle activity, ligamentous constraint, or external such as ground-reaction forces

created from external loads. During walking concentric and eccentric contraction of the limb

musculature around a joint results in the generation and absorption of mechanical energy

necessary to accomplish the movement that we observe and is referred to as "joint power". Joint

power therefore is the rate at which energy is either generated or absorbed and is the product of a

joint moment and the joint angular velocity.115 The joint moment (also known as torque or

moment of force) being the rotational potential of the forces acting on a joint. The joint moment

usually is calculated around a joint center. The units used to express moments or torques are

Newton-meters (N-m) and for research purposes usually are normalized to the subject's body









mass. Joint power is generated when the moment and the angular velocity are in the same

direction and said to be absorbed when they are in opposite directions. The units used to measure

joint power are Watts (W).

During stepping, ground reaction forces (GRFs) produced as a result of body weight

transferring onto and moving across the supporting foot constitutes the external forces acting on

the body. GRF as the name denotes, is basically the reaction to the force the body exerts on the

ground. GRF is comprised of three components: 1) vertical force, 2) fore-aft shear and 3) medial-

lateral shear.115-117 Information on these forces is obtained from a force platform or force plate,

which is a transducer set into the floor to measure the forces and torques applied by the foot to

the ground. These devices provide a quantified measure of the three components of the resultant

GRF vector.

Measurement of vertical ground reaction forces produced during walking provides

information on the load imposed on the joints during weight bearing.33 Normally this force is

represented as two peaks with a valley in between. The first peak occurs in response to weight

acceptance while loading the limb. The second peak is caused by acceleration of the body

forward. Fore-aft or horizontal ground reaction forces measure propulsion as the body weight

shifts from one lower limb to the other. Horizontal forces have a negative and a positive

component. The negative component is referred to as the braking force and is indicative of a

backward horizontal friction force between the floor and the foot to prevent the foot from sliding

forward. The positive component is referred to as the propulsive force and is indicative of the

foot pushing back on the floor to propel the body forward. The medio-lateral forces measure

stability of the body during walking.117 The exchange of body weight from one limb to the other









generate the medio-lateral shear forces. Peak medial shear occurs while loading the limb and

peak lateral shear occurs while unloading the limb.

Characterization of Stepping Pattern

Reciprocal stepping can therefore be functionally characterized using biomechanical

measures. Stepping is defined as the repetitious sequence of limb motion such that, one limb

maintains contact with the ground and supports the body while the other limb advances

forward.33 The supporting or weight bearing phase is termed as the stance phase and the limb is

referred to as the "stance limb" while the forward stepping phase is termed as the swing phase

and the limb is referred to as the "swing limb". A single sequence of support (stance) and

advancement (swing) executed by one limb is called a "gait cycle".33

The transition from stance-to-swing is characterized by two phases; terminal stance phase

and pre-swing phase. The terminal stance phase begins with the heel rise of the supporting limb

and continues till the other foot strikes the ground. The pre-swing phase begins with foot strike

of the other limb and continues till toe-off of the supporting limb. The objective of these phases

together is the progression of the body forward. Power is generated in this phase to propel the

limb and body forward.118'119 The transition from stance-to-swing is kinematically characterized

by extension of the hip joint, flexion at the knee joint and plantarflexion at the ankle joint33

(Figure 2-2). Kinetically this phase is characterized by a horizontal propulsive force to aid

forward progression of the body.

Similarly, the transition from swing-to-stance is characterized by two phases: initial

contact and loading response.33 Initial contact begins when the swinging limb strikes the floor.

Loading response begins with initial floor contact and continues until the other foot is lifted for

swing. The demand for immediate transfer of body weight onto the limb as soon as it contacts

the ground requires initial limb stability and shock absorption while simultaneously preserving









the momentum of progression. Therefore the objective of these phases together is to provide

weight bearing stability and preserve progression. Swing-to-stance is kinematically characterized

by flexion of the hip joint, extension of the knee joint and plantarflexion of the ankle joint

(Figure 2-2).33 Kinetically this phase is characterized by a horizontal braking force to aid weight

acceptance.

The transition from stance-to-swing occurs between 30 to 60% of the gait cycle and the

transition from swing-to-stance occurs between 0 to 10 % of the gait cycle.33 The temporal

sequence of these transitions is the result of interactions between a tripartite neural system

consisting of supraspinal, spinal and sensory input.95

Neural Assembly for Stepping

Central Pattern Generator for Stepping

Rhythmic and reciprocal stepping can be triggered by descending supraspinal command,

which delegate the motor coordination to specialized spinal circuitry for pattern generation called

the CPG. Primarily, the concept of specialized spinal circuitry existed as far back as 1911, when

Brown et al, showed that cats with a transected spinal cord and with cut dorsal roots still

produced rhythmic alternating contractions in ankle flexors and extensors.120 This provided the

basis of the concept of a spinal locomotor center that Brown referred to as the 'half-center'

model. One half of this center induced activity in flexors, the other in extensors.

Much later, Grillner and Wallen100 coined and demonstrated the existence of CPGs as

assemblies of premotor intemeurons that were synaptically connected with each other and with

the motorneuron pools and capable of creating an elaborate flexor and extensor synergy between

different muscles of a limb required for locomotion. Although the anatomical details of CPGs are

known for a few cases only, the motor commands originate from the spinal cords of a variety of

vertebrates.100 For example in cats, the nature of pattern generation is still uncertain because the









exact connections of only a few interneurons such as Ia interneurons, Renshaw cells, ventral

spinocerebellar cells and spinobulbar cells connecting to lateral reticular nucleus are known with

certainty.108 These neurons could be driven by the CPG or could well be a part of the CPG itself.

However, what is clear is the fact that no rhythmic input is required to activate these circuits.121

These circuits can function in vitro when isolated from the brain, as evidenced by locomotion in

decerebrate cats122 and when isolated from the motor and sensory apparatus of the limbs.120 The

rhythms can often be initiated by simple tonic (i.e. non-oscillating) electric or pharmacological

stimulation.

In humans, rhythmic, alternating electromyographic activity of the lower limbs in the

absence of supraspinal and movement related afferent input has been interpreted as evidence for

central pattern generation.123 Such evidence was also provided in a study of six subjects with

complete SCI, where researchers were able to induce rhythmic, alternating, locomotor-like EMG

pattern on continuous epidural spinal cord stimulation.124

Therefore the CPG is considered the elementary building block on which rhythmic

movement is based. As soon as rhythmic movement is initiated, feedback from the moving

limbs, termed as "motion-related feedback" in this review, arrives at the spinal cord to inform the

nervous system of the local conditions. This feedback assists in shaping the pattern of walking,

reinforcing ongoing activity and controlling the phase transitions.38'125 An ensemble of motion-

related input arising from skin, joint receptors, muscle spindle, golgi tendon organs,

mechanoreceptors, nociceptors is believed to influence the pattern of stepping.91'108'126

Specifically, hip joint position sensed by muscle spindle and load sensed by the golgi tendon

organs are two of the several motion- related inputs contributing to the control of stepping.31'127









Several animal and human studies examining the neurobiological control of stepping have

validated these findings.

Sensory Drive Required for Stepping

Sherrington was the first to propose that proprioceptors responding to hip extension are

important for initiating swing.128 Grillner and Rossignoll129 found that preventing the hip from

attaining an extended position in chronic spinal cats inhibited the generation of the flexor burst

and hence the onset of the swing phase. The most direct evidence for this conclusion however

came from vibrating the hip flexor muscle (iliopsoas) during stance which led to an earlier onset

of swing in walking decerebrate cats.130 The receptors signaling hip extension were probably the

primary and secondary endings of muscle spindles in hip flexor muscles (Group Ia afferents).

Similarly, in humans, involuntary and alternating stepping-like movements were observed

in an individual after incomplete SCI upon extending the hip in the supine position.131

Furthermore, hip walking movements (i.e. facilitating hip joint excursion with the knees fixed in

an extended position) induced by a driven gait orthosis (DGO) in individuals with complete SCI

produced pattern of leg muscle EMG activity that corresponded to that normal stepping in

healthy, non-injured individualssubjects.127 Researchers examining infant stepping also support

the role of hip extension position for the initiation of swing. From the recorded hip motion and

electromyographic data these scientists concluded that the preferred hip position was always one

directly opposite the direction of walking during infant stepping.38 It was thus suggested that the

hip position is important in initiating the stance-to-swing transition.

Another important sensory input regulating the stance-to-swing transition is the extensor

load relayed by the Golgi tendon organs (Group Ib afferents) in the ankle extensor muscles.31'132

During locomotor activity, electrical stimulation of the group Ib afferents from the ankle

extensor inhibits the generation of flexor bursts and hence prolongs the duration of extensor









activity. Duysens and Pearson (1980) observed that gradually increasing load applied to the

achilles tendon resulted in an increase in both the amplitude and duration of the rhythmic EMG

bursts of ankle extensors.89 Similarly, cutaneous afferents innervating the skin of the foot (group

II afferents) are also load monitors. Electrical pulses applied to the foot pad innervated by the

sural nerve were able to prolong the extensor burst in the stance phase in pre-mammilary cats

preparations thereby providing evidence that load-related cutaneous input from the foot can

inhibit the CPG for the generation of flexion during swing.36'133

In humans, researchers found that unloading the ankle extensors by a portable device in the

stance phase of walking reduced the soleus EMG activity and the reduction was maintained even

when transmission in Ia afferents was blocked by local anesthesia. This finding thus pointed to

group Ib and/or group II afferents contributing to the extensor EMG activity in the stance

phase.134 Harkema et al. (1997) observed that the amplitude of extensor muscle activation in the

legs was directly related to the level of body weight loading on the legs during stepping of

healthy, non-injured individualsand SCI subjects during manually-assisted stepping on a

treadmill.135 Dietz et al. (2002) also found that physiological locomotor-like leg movements

alone (100% body unloading) generated by the application of the DGO on a treadmill are not

sufficient to generate leg muscle activation in either healthy, non-injured individualssubjects or

in subjects with complete para-/tetraplegia.127 In this study, leg movements in combination with

loading of the legs led to appropriate leg muscle activation.

In summary, during stance phase, load of the lower limb is detected by group I extensor

muscle afferents and group II cutaneous afferents which activate the extensor half center (EHC)

of the CPG. Extensor activity is reinforced during the loading period of the stance phase. At the









end of the stance phase, group Ia afferents of flexor muscles excite the flexor half center (FHC)

which inhibits the EHC and thereby initiates the onset of the swing phase (Figure 2-3).

The importance of motion-related sensory input for locomotor control is evident when

descending supraspinal input is compromised or altered such as after SCI. In contrast to an intact

nervous system processing multiple sources of afferent input, after complete or incomplete SCI,

the spinal circuitry does not become silent (Figure 2-4 A & B). The circuitry instead adapts to its

altered combination of inputs and predominantly utilizes motion-related afferent input to

facilitate locomotor response.37 The weighted response of the spinal circuitry to ascending

afferent input illustrates the high level of spinal automaticity for locomotor control. Therapeutic

strategies that optimize motion-related afferent input to the spinal cord can therefore be utilized

to regain locomotor control after SCI. For example, stepping can be initiated by shifting the body

weight to one leg and moving the head and trunk so that the hip position of the contralateral leg

is extended.41'42

Apart from influencing the stepping pattern, sensory input also modulates spinal reflex

behavior during stepping.136 Spinal reflexes are those in which the sensory stimuli arise from

receptors in muscles, joints and skin, and in which the neural circuitry responsible for the motor

response is entirely contained within the spinal cord. Spinal stretch reflexes are the simplest

"stimulus-response" behaviors exhibited by the mammalian nervous system. The stretch reflex as

defined by Wolpaw is "the initial, purely spinal, largely monosynaptic response to sudden

muscle stretch that is accessible anatomically and physiologically."137 H-reflex is an electrical

analogue of the spinal stretch reflex and is a commonly studied spinal reflex in human beings.138

It was originally described by Paul Hoffman in 1910 and is electrically elicited leading the

stimulus to bypass the effect of the gamma motorneuron.139 Therefore, H-reflex is a valuable tool









in assessing modulation of monosynaptic reflex activity in the spinal cord and has been used

since to assess the response of the nervous system to various neurologic conditions,140-142

application of therapeutic modalities,143'144 exercise training and motor task performance.145-147

H-reflex Elicitation and Modulation During Stepping

Elicitation

The technique used to evoke the H-reflex involves electrical stimulation of a mixed

peripheral nerve.148 When a percutaneous stimulation of increasing intensity is applied to a

mixed nerve, action potentials travel along afferent portion of the reflex arc along the Ia sensory

afferents, until they synapse on the alpha motorneuron. The efferent portions of the reflex

pathway results from action potentials generated by the alpha motor neuron until they reach the

myoneural junction and produce a twitch response which is recorded by surface electrodes on the

muscle of interest (Figure 2-5A).

In addition to the H-reflex, electrical stimulation of the peripheral nerve also causes direct

activation of the efferent fibres that conduct orthodromically to produce a response in the EMG

known as a muscle response or M-wave.148,149 When the stimulus intensity is really low only the

la afferent fibres of the mixed nerve undergo depolarization leading to the appearance of the H-

reflex tracing on the EMG. As the stimulus intensity is increased, more Ia afferent fibres are

recruited, resulting in activation of more alpha motor neurons and increasing the amplitude of the

H-reflex. Continuing to increase the intensity beyond the point of elicitation of the H-reflex also

results in direct stimulation of the motor axons and thereby production of the M-wave (Figure 2-

5A). The M-wave on a recording is usually seen before the H-reflex due to the relatively short

path that the action potentials need to travel. Compared to this the H-reflex is characterized as a

latency response since it does not occur right after the application of the stimulus (Figure 2-5B).

The latency of the reflex response is a result of the length of the H-reflex pathway, which









involves the afferent and efferent length of the path and also the overall length of the limb. For

example, in the soleus muscle, the H-reflex tracing usually appears approximately 30

milliseconds after the delivery of the stimulus whereas the M-wave is usually seen after 6 to 9

milliseconds of stimulus application.138

The amplitudes of the H-reflex and M-wave both increase fairly linearly with increase in

stimulation intensity until the maximum H- reflex (H-max) representing the fullest extent of

reflex activation is reached.138 Increasing stimulus intensity further beyond H-max, maximum M

response (M-max) representing the maximum muscle activation is reached.138'149 Therefore, by

increasing the stimulation intensity a recruitment curve depicting, stimulation intensity sufficient

to evoke a sequence of H-reflex, H-max, disappearing H-reflex tracing, M-max and M-max

plateau can be obtained (Figure 2-5C).138,149

Task-specific/ Phase-dependent modulation of the H-reflex

Soleus H-reflexes are the most widely assessed reflexes in locomotor studies.138 In healthy,

non-injured individuals, soleus H-reflexes are strongly modulated during the gait cycle with the

highest amplitude registered during the stance phase and the lowest amplitude recorded during

the swing phase.146 H-reflexes are minimal at the time of heel contact, rise to a maximum shortly

after midstance, decrease rapidly at the time of toe-off and are minimal during swing in both

young and older age groups.150 Therefore, soleus H-reflexes in healthy, non-injured individuals

demonstrate phase-dependent modulation during stepping. Along with the phase-dependent

modulation of the reflex during the step cycle, H-reflexes are also known to display considerable

differences in modulation between different motor tasks.147'151 Both phase-dependent and task-

dependent modulation of the reflex is critical to the optimal performance of motor behaviors.

Modulation of the H-reflex during stepping has been attributed to both, sensory input152'153 and to

the higher central mechanisms controlling motion.154 Although the bulk of literature supports the









primary role of central reflex modulation during walking, the secondary role of sensory inputs is

critical as well.

The role of sensory inputs is especially critical when the central control of reflex

modulation is compromised for e.g. post SCI. As a result of injury to the spinal cord, phase-

dependent modulation is impaired and reflex amplitudes are higher than normal throughout the

gait cycle.155,156 In such a scenario, the reflex muscle response is likely to be dependent on

peripheral sensory inputs deficient of central modulation. Several studies have reported that the

soleus H-reflexes can be modulated by sensory inputs of peripheral origin such as hip joint

position, leg load, and cutaneous receptors in sole of the foot.134,157 Sensory input related to hip

joint position or from hip joint proprioceptors for example is shown to markedly influence soleus

H-reflexes during passive movement of the hip from flexion to extension phase in both healthy,

non-injured individualsand persons with spinal cord injury.158,159

Similarly, mechanical loading of the foot sole, ranging from 15 to 70 N is known to

significantly inhibit soleus H-reflex amplitude in both seated healthy, non-injured individualsand

persons with complete SCI.136 Although performed in non-locomotor tasks, the above studies are

suggestive of a possible alteration in H-reflex amplitude via peripheral sensory inputs even after

SCI. Thus post-SCI, provision of critical motion related sensory inputs may play a role in phase-

dependent H-reflex modulation and could be useful in optimizing task performance.

Recovery of Walking after SCI

Plasticity at the Level of the Spinal Cord

The lifelong ability of the nervous system to reorganize neural pathways structurally or

functionally in response to experience or activity termed as "neuroplasticity" was once solely

considered to be a supraspinal phenomenon.7 Acquisition and maintenance of normal motor

performance however involved skill-dependent plasticity at multiple sites including the spinal









cord.85'160 Early evidence of this phenomenon was seen as early as in 1951 when Shurrager and

Dykman reported that treadmill walking in spinalized cats improved with training.161

Locomotion was much better in cats exposed to treadmill training than in cats that received

standard care after injury.90 The improvements in training seen after training persisted even after

training was stopped. Similarly, De Leon et al performed a series of experiments in spinalized

cats either trained to stand or step on the treadmill. Their results revealed that cats trained to

stand improved standing function and those trained to step improved in stepping.28'90 Locomotor

ability was exactly reversed in these groups when they were retrained to perform the other task.

These researchers also tested the effect of glycinergic inhibitor, strychnine in spinalized

cats that were trained to either walk or stand on the treadmill.162 They observed that locomotion

improved in cats trained to stand when strychnine reduced glycinergic inhibition and had no

effect on the cats trained to step. These findings suggested that glycinergic inhibition in the

spinal cord interfered with stepping ability in spinal animals and locomotor training improved

stepping ability by reducing the levels of inhibition on the spinal networks. Skill-dependent

training can therefore markedly change or modify physiological and biochemical state of

multiple neurotransmitter-modulator systems in the spinal cord and enhance locomotor

recovery. 162-164 Studies of motor unit properties after spinal cord transaction, with and without

training, have also provided supplemental evidence by indicating that training induced

improvements in walking and standing are not attributable to peripheral changes in muscle

strength. 165

Several studies in humans have also provided compelling evidence for activity dependent

plasticity of the spinal cord caudal to injury. Similar to cats, treadmill training using a body

weight support showed significant improvements in walking behavior in people after severe









spinal cord injuries.27,41,42,135 Dietz et al. (2003) also acknowledges that alteration in glycinergic

and GABAergic systems seen in animal models as a result of the training could be true for

humans as well.166 The spinal cord as a result of the training learns to respond to specific sensory

inputs associated with locomotion which would potentially reorganize the circuitry involved in

locomotion. Histologic data supporting the possibility of sprouting or synaptogenesis in the

human spinal cord although not direct are also evident.167

Plasticity of the Spinal Stretch Reflex (SSR)

Contradictory to their conceptualization as "hard-wired", the spinal stretch-reflexes are

modulated during movement and adapt to training.85 Reflex modulation refers to the change in

strength (or amplitude) of the reflex over the course of a behavior and is essential for the optimal

performance of the motor behavior. Researchers have shown that in monkeys, rats and humans

these reflexes can be operantly conditioned i.e. the amplitude of the response can be either up-

trained or down-trained. 168-172 In these protocols the amplitude is measured as electromyographic

activity and reward occurs when the amplitude is either above or below a criterion level.85 While

change in the tonic descending activity motivated by the probability to seek reward initiates

reflex change, for this reflex to respond consistently in this fashion over several sessions certain

alterations occur somewhere in the spinal arc of the reflex.

Similarly, short term and long term changes have been noted in the reflex amplitude in

animals and humans as a result of task-specific training.173-175 A single training session of short

bouts of balancing on an unstable platform in normal subjects demonstrated a progressive

decrease in the soleus H-reflex amplitude of about 26%.176 Hess et al also observed a modulation

of the reflex size during adaptation to a new motor task.177 H-reflex modulation over time was

evaluated for normal subjects over five runs of treadmill walking (three with normal treadmill

walking and two with randomly stepping over the obstacle 100 times). The largest adaptations









with a significant increase of reflex amplitude occurred during the first obstacle run. This

increase lasted only briefly and the reflex amplitudes decreased to their previous values. During

the later obstacle run, no H-reflex modulation occurred. Similarly, normal subjects training to

walk backward showed progressive adaptation of their soleus H-reflex in mid-swing phase as

early as twenty minutes after training.175 Long-term maintenance of this modulation was also

noted as late as five months after cessation of training indicating acquisition and maintenance of

novel motor skills.

Operant or training induced conditioning of the reflex can have potential implication in

rehabilitation. Studies based on application of this phenomenon in subjects with neurological

injury have shown significant changes in the reflex behaviors of these individuals. For example,

studies have shown that operant conditioning of the spinal reflex reduced hyperactive biceps

stretch reflex in people with spinal cord injuries with the reduction persisting for four months

following cessation of training.178 Similar results in locomotor studies have only been

documented in spinalized animal models thus far but might have potential ramifications in

improving human locomotion as well.45'179'180 In four subjects with incomplete spinal cord injury,

for example, a single bout of step training over the treadmill increased overground walking speed

by 25% and reduced soleus H-reflex amplitude during overground walking providing evidence

for activity- dependent plasticity of the reflex.181

To this end, since reflex behaviors function as parts of complex behaviors, conditioning

them in accordance to the requirements of the task i.e. task specific training may help improve

functional outcomes in the targeted population.182 Also from a rehabilitation standpoint,

provision of appropriate task-specific stimuli and/or adopting different feedback strategies that

could facilitate or depress these reflex responses in desired ways could help in retraining or









reeducation of function. For example, providing optimal hip extension during stepping would up-

train the stretch response of the hip flexors to facilitate hip flexion for swing.

In summary, short-term and long-term adaptive changes occur within the reflex

components as a result of conditioning. These changes are also known to persist despite removal

of the descending input thereby exhibiting learned behavior at the level of the spinal cord. This

adaptability and amenability of the reflex to persistent change or conditioning constitutes spinal

cord plasticity.

Locomotor Training and Skill-dependent Plasticity of the Nervous System

Animal and human research on the neurobiological control of stepping and the skill-

dependent plasticity of the nervous system have challenged the compensation based approach of

rehabilitation after SCI. Research has revealed that task-specific, repetitive training following

SCI in animals and humans promotes skill-dependent plasticity in the spinal cord and plays a

critical role in the recovery of locomotor abilities including stepping.88 21,22,27,83 The emerging

training strategy is to provide the CNS with peripheral sensory input related to locomotion in

order to stimulate a stepping response.19,24-26,41,43,44,135,183 Processing of task-specific kinematic

and kinetic information facilitates performance of the task and learning.184 This strategy is based

upon evidence that the lumbosacral spinal cord is capable of recognizing and processing

functional sensory cues to produce a functional motor response.125 Simply put, generation of the

stepping pattern would involve the provision of motion-related afferent input associated with

stepping. Two of these inputs related to stepping are loading and hip position, which have been

discussed earlier.38'127'132'135'185 A therapeutic intervention termed as locomotorr training" has

been developed based on such research.

In the locomotor training environment this specific sensory input is made available by

having a person with SCI walk over a treadmill in a harness connected to a BWS system.43 This









assembly provides an environment where normal walking speeds, bilateral limb loading and

proper limb and trunk kinematics can be safely and effectively trained to enhance the neural

output generating walking.43 Evidence supporting the benefits of locomotor training in

facilitating the recovery of walking in humans has been recorded in the literature as early as the

1990s.25,39,41,42,186-188 Persons trained in this training environment demonstrated significant

improvements not only in pattern of stepping, walking velocity, endurance but also in

neurophysiological predictors of improved coordination like the electromyographic activity in

the muscles of the lower extremity.189-191 Soleus H-reflexes have also been shown to respond to

training post-SCI. A single bout of training has shown to increase walking speed with a

corresponding decrease in reflex amplitude, which is usually high after injury.181

Is Orthotic use Appropriate during Locomotor Training?

Despite the compelling evidence supporting the concept of locomotor training controversy

and variability still exists in training parameters in the training protocol.24-26,42,46,192 Not all the

parameters for optimizing recovery are known thus far and those that are, for example, BWS,

manual assistance and speed are still being reviewed for mode of delivery and dosage.193

Therefore although the principles of training have been established, there is no universal

agreement on ideal training parameters for locomotor training.4'46'82 For example, controversy

still exists about the use of parallel bars during training. In a study done by Conrad et al

improvements in step symmetry were observed using conventionally used parallel bars during

training.68'194 On the contrary, Visintin et al reported improvements in symmetry using vertical

body weight support instead of parallel bars providing upright support.195

Similarly, although widely used in conventional gait training, guiding evidence for the use

of orthotic devices such as the AFO during LT is lacking.24 Proponents of LT hesitate to use an

AFO during training based on the assumption that it would interfere with optimizing the









kinematics and therefore afferent input required to drive the neural circuitry.24,26,196 Logical

concerns expressed are that the AFO might restrict the range of excursion of the ankle and linked

knee and hip joints altering the afferent information related to hip position and load.

Additionally, although an AFO might facilitate an alternate stepping pattern, it would serve as a

fixed variable unlike BWS and manual assistance that can be adjusted to facilitate independent

stepping over the course of the training. Therefore the use of the device during LT qualifies

further investigation.

Altered Stepping Mechanics Resulting from the Use of Orthotic/Assistive Devices

Although the goal is to improve ambulation potential, prescription of assistive and/or

orthotic devices often does not take into account the influence of the device on the user's

resultant stepping pattern.197'198 For example, rigid support provided at the ankle by the PAFO

might limit the excursion of the contingent joints affecting the control of the transition from

stance-to-swing. Similarly, loading through the fixed upper extremities over the parallel bars

alters the prerequisite loading pattern through the lower extremities required for stepping.195

However, the hierarchical framework assuming irreversible walking deficits post-injury have

promoted compensatory function with such devices rather than examining their implications on

the neural sub-tasks of walking.

Under the supposition that neurological function ceases to exist below the level of injury,

clinical assessments for the prescription of ambulatory devices utilize a frame work for

substitution of impaired function rather than the restitution of function.78 Therefore, the

ambulatory ability of a person in moving from one place to another rather than the walking

pattern utilized in achieving this mobility is emphasized. Individuals requiring assistive devices

have a decreased ability to provide the supporting, stabilizing, propulsive or restraining forces at

the lower extremities necessary for forward progression.199 With assistive devices the upper









extremities are the structures bearing the load and are being employed in providing the

complementary forces for walking.200'201 Melis et al. (1999) reported a decrease in hip excursion

and a decrease in the ability to unload the limb while walking with a walker.202 Visintin and

Barbeau (1994) investigated the consequences of weight bearing on the upper extremities

compared to weight bearing through the legs, both with 40% BWS provided. Her results

indicated a decrease in electromyographic activity in the lower limbs and more asymmetry in the

limb kinematics.195

Similarly while wearing orthotic devices like the PAFO, Ounpuu et al. (1996) reported an

inability to generate sufficient power in the transition from stance-to-swing phase of stepping.203

However, conclusions drawn from these studies contend that ambulatory devices are still capable

of fulfilling various assistive functions during walking, although they affect posture and walking

pattern. The inability to generate normal walking mechanics was concluded to be the result of the

irreversible nature of the injury rather than the inability of the device to provide or assist normal

stepping. Therefore the assessment of an assistive and/or orthotic device based on its ability to

restore normal walking mechanics is a novel perspective.

Altered H-reflexes Resulting from the Use of Orthotic Devices

The H-reflex has been commonly employed as a neural probe in describing and

interpreting neural interplay for the control of movement. Elicitation of the reflex and

measurement of its amplitude has provided insights into the changes in transmission in spinal

pathways during the performance of a motor task.85 Since afferent information is processed at the

multiple levels of the nervous system including the spinal cord, examination of the soleus H-

reflex modulation at the ankle joint would be a favorable tool in linking device-dependent

sensory information to a motor response. Schneider et al, for example, demonstrated that in

normal subjects bracing the ankle joint angle to 900, such as in preventing foot drop, the burst of









TA activity was eliminated and so too the inhibition of the H-reflex.154 A similar inhibition in H-

reflex was observed by Brooke et al on bracing the ankle joint suggestive of premotorneuronal

mechanism for reflex inhibition.204 Nishikawa et al. (1999) examined the effect of applying an

ankle brace in sitting to an uninjured ankle.205 They reported an increase in H-reflex amplitude in

the braced compared to the unbraced condition concluding that ankle bracing increased

stimulation of the cutaneous mechanoreceptors around the joint. However, in these experiments

the subjects were tested during passive movement of the limb or in a static position instead of

active walking.

Garrett et al. studied the effect an knee orthosis on soleus H-reflex modulation during

walking.206 They reported that even though the walking pattern changed in normal individuals,

phase-dependent H-reflex modulation was not influenced by the use of a knee orthosis. In this

study however, the ankle was not restricted and hence the specific effect of an AFO is unknown.

Therefore differential modulation of the H-reflex while walking with and without wearing an

AFO will help determine if walking with an AFO is an inherently different motor task compared

to normal walking.

Rationale of the Studies

Demonstrating the presence of numerous parallel systems within the CNS that reorganize

after injury using intense, task-dependent interventions such as LT has catapulted the goal for

rehabilitation of walking after SCI from compensation to recovery.21'207'25'208 Recovery of the

sub-component of stepping, in this intervention depends on careful selection of training variables

that deliver critical motion-related sensory input to facilitate stepping. Conversely, aptness of a

training variable such as an AFO to facilitate stepping can be determined by examining the

mechanical and neurophysiological deviations observed during stepping.









Therefore, I investigated the effect of an AFO on mechanics and soleus H-reflex

modulation in healthy, non-injured individuals and persons with ISCI during stepping.

Measuring these two factors in healthy, non-injured individuals provided a frame of reference for

subsequently comparing normal and impaired stepping. Where as, examining changes in the two

factors in persons with ISCI and comparing it to normal helped identify the degree of correction

or deterioration in stepping provided by the AFO. Therefore, the cross-sectional studies have

gleaned specific information about the use of AFOs in physiological paradigms that optimize

recovery of walking after SCI.

Methodological Considerations for the Measurement of Mechanics and Soleus H-reflex
During Stepping

Measurement of Stepping Mechanics

Study considerations

In the study protocol, measurement of joint excursion at the hip, knee and ankle with and

without wearing an AFO were used to assess critical kinematics required for the transition from

stance-to-swing. The AFO was fitted on the dominant side for healthy, non-injured control

subject and the more involved side for persons with ISCI. Kinematic and kinetic data are

susceptible to change with walking speed.115 Therefore, the speed of walking was matched in

both the walking with and without the AFO conditions. Also kinematic data is affected by limb

length and height of the person.115 Therefore, for comparisons between people with ISCI and

their non-injured counterparts, height, age and weight matched individuals were selected for the

study.

Interpretation of kinematic and kinetic data

Walking is a motor task in which each of the interconnected segments of the lower limbs

undergoes a characteristic excursion to move the body forward.33 If the terminal excursion of any









one of the joints is restricted it has the potential to alter the excursion of the other joints linked to

it thereby disrupting stepping pattern. The stance-to-swing transition phase is usually

characterized by ankle plantarflexion, knee flexion and hip extension to aid push-off.33 If ankle

range is restricted using an AFO, plantarflexion will be limited and the knee and hip excursion

will also be altered. Similarly, the swing-to-stance transition phase is characterized by ankle

plantarflexion, knee extension and hip flexion.33 If ankle plantarflexion is limited the knee and

hip range will also be altered. Assessment of kinematic data while walking with an AFO will be

helpful in elucidating this phenomenon.

Furthermore assessment of kinetic data will indicate the discrepancies in the functional

task requirements of stepping. For example, during stance-to-swing phase, plantarflexor power is

generated to propel the limb and body forward.33'203 If plantarflexor power is limited while

walking with an AFO,203 the knee and hip joint powers might demonstrate a compensatory

increase or decrease to propel the limb forward. Also, with limited propulsion, peak braking and

propulsive forces generated in the fore-aft direction will decrease compared to normal push-off

Similarly, the rate of loading will be used to quantify the ability to shift weight from one lower

limb to the other and measured by calculating the slope of the vertical ground reaction

force.209'210 A prolonged loading rate onto the limb donning the AFO will be indicative of an

inability to plantarflex the ankle and extend the knee to support the transferred weight.210

Collective assessment and interpretation of the above parameters will therefore provide valuable

information with regards to the mechanics of stepping with an AFO.

Reliability of Vicon and Force Plate Measures

Reliability refers to the ability of an instrument to provide consistent, stable and repeatable

measurements. In this project, Vicon motion analysis system (Vicon, Oxford Metrics

Ltd,Oxford, England) and kistler forceplates (Kistler Instruments, Inc.,Amherst, NY) were used









to identify the mechanical changes during stepping with an without a PAFO. The Vicon motion

analysis system is an automated, high-speed three-dimensional (3-D) motion system where

cameras track the motion of retro-reflective markers that are placed on subject's body landmarks.

Kadaba et al (1989) investigated the repeatability of kinematic, kinetic and electromyographic

data using the Vicon motion analysis system on forty normal subjects three times a day on three

separate days.211 Excellent intra-rater reliability was seen for kinematic data in the sagittal plane

both within and between test days (Table 2-1). Similarly, frontal and transverse planes joint

angle motion yielded good repeatability within test days leading them to conclude that gait

variables measured by the Vicon system are quiet repeatable for subjects walking at their normal

speed. Similarly, Richards et al (1999) compared accuracy of several automated motion analysis

systems by placing two markers 50cms apart on an aluminum bar rotating in the horizontal plane

in the camera capture volume.212 The results of his study indicated that the Vicon was the most

accurate with a maximum error of 0.183 cm that is the lowest among all other systems. Several

other studies have also reported excellent intra-rater, inter-rater and test-retest reliability of the

Vicon motion analysis system making it the current "gold standard" for motion analysis.213-215

Ground reaction forces were measured with two kistler force plates that use triaxial

piezoelectic force transducers mounted at the covers of each plate to measure the three

components of the ground reaction force vectors. These force plates have been shown to be the

reliable standard for measuring dynamic transition from bipedal to single limb stance in healthy,

non-injured adults.216 Intraclass correlation coefficient (ICC) for the magnitude of the propulsive

and braking force has been reported to be greater or equal to 0.73 for fast movements and greater

or equal to 0.88 at the natural speed.









Force plates have also been used for comparing ground reaction force patterns in non-

injured individuals and individuals with ISCI.217 Repeatability of initial vertical force peak and

time to peak variables measured by force platforms while donning foot orthosis has also been

reported.218 Excellent ICC results were demonstrated for the vertical force variables, with power

greater than 0.80. Measurement of peak vertical ground reaction force during a vertical jump at

two time points 48 hours apart was demonstrated to be very reliable (ICC [2,1] = .94).219

Measurement of Soleus H-reflex

Study considerations

Elicitation of soleus H-reflex during walking with and without a PAFO requires control of

several extraneous factors that could potentially confound the reflex response. These factors have

been identified below and methods to control them have been discussed. For the purpose of

consistency, soleus H-reflex will be evoked, on the dominant side of healthy, non-injured control

subject and on the more involved side for persons with ISCI. The reflex will be evoked by

localizing the tibial nerve in the popliteal fossa.138,220 Subject positioning is critical during H-

reflex testing since several factors affect the soleus H-reflex. H-reflex is sensitive to various

inputs including posture,221 joint position,149 reciprocal and recurrent inhibition,222 behavioral

state,223'224 caffeine intake225 and muscle activity.226 However, if the above factors are

sufficiently controlled, then H-reflex can provide information of the state of the reflex arc.

Stimulation intensity is another factor that affects reflex response. Intensity was maintained

between 8-12% of M-max so as to evoke a direct muscle response. This procedure helped to

safeguard against movement of the stimulating electrode that might alter the relative activation

of the Ia afferent axons and alter the H-reflex amplitude without changes in synaptic efficacy.

Soleus H-reflex amplitude is affected by background EMG activity of the muscle.149'226

Background EMG activity was measured 100 ms prior to stimulation in each condition to ensure









similar level of motor neuronal excitability. Stimulus frequency was maintained between 3-5

seconds to avoid post-activation depression of the response.222 Since profound reductions in M-

max amplitude have been reported to occur across the time course of an experiment, M-max was

elicited in each condition and in each tested phase of the gait cycle throughout the experiment for

subsequent normalization of the data.227

Interpretation of H-reflex amplitude

The H-reflex demonstrates a phase-dependent regulation of its amplitude during walking

which is an important component of motor control, allowing afferent feedback to have differing

effects in different phases of the step cycle.228 This regulation is required to accommodate the

functional requirements of the task. For example, the soleus H- reflex is minimal at heel contact,

increases to maximum during stance and decreases rapidly just prior to toe-off and is minimal

during swing. The observed changes in size facilitate weight support and ankle extension during

mid-to-late stance while allowing ankle dorsiflexion during swing and while the body moves

over the foot during early to midstance.146,228,229 Therefore if the task of walking with an AFO

was similar to walking without one then the neural control would be preserved between the two

conditions and the soleus H-reflex modulation in the step cycle would be similar. An increase in

H-reflex amplitude while walking with an AFO (assuming all other conditions including

stimulus strength are maintained constant) compared to walking without one in selective phases

of the step cycle would be indicative of an altered task.

Similarly, with regards to adaptation of the reflex to a new motor task studies have shown

this to occur in a biphasic fashion.137 For example, for the novel task of stepping over an

obstacle, Hess et al demonstrated a progressive adaptation in soleus H-reflex amplitude during

repetitive stepping. He showed that in normal subjects the soleus H/M ratio increased strongly at

onset of the motor learning task and reduced over the course of exercise reflecting the nervous









systems capacity to adapt the locomotor pattern to the actual requirements.177 The initial increase

in soleus H-reflex amplitude is attributable to descending influence of the corticospinal tracts on

the spinal reflex arc.230,231 The eventual reduction in reflex amplitude is speculated to be the

effect of the acquired task being automatically performed and controlled at a spinal or brainstem

level.169 Therefore prolonged successive stepping with an AFO in normal subjects if different

from walking without one would also exhibit adaptive changes in reflex amplitude as seen with

the acquisition of a new motor task.

Reliability of Soleus H-reflex

The reliability of the soleus H-reflex testing in supine and standing position has been

confirmed for inter-session and intra-session procedures.138 ICC (2, 1) for H-max, M-max and H-

max/M-max has been reported to be 0.99 +/- 0.007, 0.95 +/-0.08,0.97 +/-0.009 respectively.

Similarly, the intra-session and inter-session reliability of soleus H-reflex over five consecutive

days in a standing position have also been established.220 The standing intra-session and inter-

session reliability was established to be 0.85 and 0.80 respectively.

Clinical and Scientific Relevance of the Study

Clinical Relevance of the Study

Walking with orthotic and assistive devices has been the quintessential approach for

improving walking potential in persons with incomplete spinal cord injuries. Assessment of such

devices in physiological-based training paradigms like locomotor training will provide

information about the degree of conformity of such clinical strategies with the principles of

neurobiological control of walking. Accordingly, in this study, mechanics of stepping and H-

reflex modulation with the AFO will be evaluated which characterize the motion-related afferent

input being processed at the level of the spinal cord. Altered stepping mechanics and reflex

modulation with compensatory devices would reflect the failure to provide task-specific sensory









input driving the intrinsic spinal circuitry necessary for walking recovery. Knowledge of results

from this study would influence clinical decision-making for the use of such devices and

strategies in physiological-based training interventions.

Scientific Relevance of the Study

Apart from the clinical implications, this study stimulates a strong rationale for assessing

archetypical strategies that have historically guided clinical practice so far. Generation of

appropriate task mechanics associated with stepping is hypothesized as essential for recovery of

stepping pattern in physiological-based training interventions. The transition phase mechanics

are modulated by the motion-related information processed in every gait cycle. Assessment of

the task mechanics of stepping in these transition phases while walking with orthotic devices is

critical to examine how motion related information generated by such strategies influence the

ability to step successfully.

Assessment of conventional strategies from a neural framework is relevant to the

neurological population to which these strategies commonly apply. The tools used for

assessment of walking are relevant to address the task and function of specific events

characterizing the walking behavior. Therefore the integration of two different perspectives,

biomechanical and neurophysiological, will provide an effective framework for understanding

the control of movement and addressing the questions posed.



















A




Spinal preparation Decerebrate Decorticate Intact system
preparation preparation

* Near normal Improved Dynamic Adaptable
inter/intra limb coordination of equilibrium locomotor
activation activation Initiate control system
patterns patterns reasonably to meet goals
* Functionally Weight support normal goal of the animal in
modulate Active propulsion directed any
reflex action behavior in environment
* Execute other neonatally
rhythmic decorticate
movements animal
concurrently Repertoire of
options limited
Altered context
of locomotor
movements

Figure 2-1. Reduced animal preparation showing the impact of transaction at different levels of
the nervous system. Adapted from Patla, A. E. in Evaluation and management of gait
disorders (ed. Spivack, B. S.) (New York, 1995).














'r ar


9


Swing to- stance


,*rs,^ : ,-.. ^- \v^^

... ...o .. .,
T -- I -- I-- -- -- I ---- I-I ',. ~
***p-- "p'(^,^e~gait:^ite


I
I 4WY,,
? yms.


I 0L5c~


aIHRUNIUMM


Loft
Right |_~__ _


Figure 2-2. Stance-to-swing and swing-to-stance phases of the gait cycle.


Figure 2-3. Model of the flexor and extensor half centers (FHC & EHC) and afferent input
regulating stance and swing phase during stepping. Adapted from Van de Crommert, H. W.,
Mulder, T. & Duysens, J. Neural control of locomotion: sensory control of the central pattern
generator and its relation to treadmill training. Gait Posture 7, 251-63 (1998).


I


_ _


Stance-to-swing














SENSORY INFORMATION
Visual O1ffEtion
Taste Auditory ProI ricepti


Certliuponil N4|sealml m larmitf i

% iWid~hpm moler ntrt

asotT information
m\ prooccpHiev



l|rrinsc ailhit\

A ULTOMA TICI TY


MUTOR1OUrPIT
_ieaim. ipfet efl com'rifo. PasuPra l am


SENSORY INFORMATION
Visual Olfaction
Taste Aditory PropriKepli'n


Cor C ikspinal Nu spnali mduluyatr
tfol mcloqhl C mlin n
flVeslimdKHinl Cm UoplsPln4u1i
SVsory dinformation


Figure 2-4. Neural control of locomotion in an intact nervous system (A) and compromised
nervous system (B). Adapted from Edgerton, V. R., Tillakaratne, N. J., Bigbee, A. J.,
de Leon, R. D. & Roy, R. R. Plasticity of the spinal neural circuitry after injury. Annu
Rev Neurosci 27, 145-67 (2004).












MuscCl Spindle

Stimulus Cell body of la
afferent in DRG
. n,


-- M-wave


Muscle
B H-reflex
D


M-wave


stilmulus artifact
20 rns


-+J-Hre41ex


I 3 3


4 5 6 7


Stimulus Intensity

Figure 2-5. Electrical stimulus (shown here by the grey ellipse) applied to the mixed nerve
conducts the stimuli orthodromically in the motor and sensory axons to evoke the M-
wave and the H-reflex respectively (A). Stimulus triggered recording of the H-reflex
also known as latency response occurring after the M-wave (B). Adapted from Zehr,
E. P. Considerations for use of the Hoffmann reflex in exercise studies. Eur J Appl
Physiol. 86, 455-68 (2002). Elicitation of the recruitment curve showing maximum
H-reflex and M-max amplitude in response to stimulus intensity (C). Adapted from
Palmieri, R. M., Ingersoll, C. D. & Hoffman, M. A. The Hoffmann reflex:
methodologic considerations and applications for use in sports medicine and athletic
training research. J Athl Train. 39, 268-77 (2004).


C 0-4
i-35
0-3
S025
S0.2
a 0.1
< 0.05
0.05
0


a 9









Table 2-1. Coefficient multiple correlations (CMC) reflecting stride to stride variability between
and within days.


CMC within day


Vertical GRF
Horizontal GRF
Hip Flexion/Extension
Knee Flexion/Extension
Ankle dorsiflexion/ plantarflexion


Right
0.99 0.00
0.99 0.00
0.99 0.00
0.99+ 0.01
0.970.02


Left
0.99 0.00
0.99 0.00
0.99+ 0.01
0.99 0.00
0.970.01


CMC between days
Right Left
0.99 0.00 0.990.00
0.980.01 0.980.01
0.970.01 0.970.01
0.980.01 0.980.01
0.930.03 0.930.03









CHAPTER 3
EXAMINATION OF WALKING WITH AND WITHOUT AN ANKLE FOOT ORTHOSIS IN
NON-INJURED INDIVIDUALS AND PERSONS WITH INCOMPLETE SPINAL CORD
INJURY: A NEUROMECHANICAL PERSPECTIVE

Spasticity, muscular weakness and co-activation are key motor impairments limiting

walking potential in individuals with incomplete spinal cord injury (ISCI).68-70 As one

component of rehabilitation, clinicians often use orthotic devices to compensate for these

impairments and aid walking.15,75,78 Individuals lacking muscular control at the ankle are

prescribed single joint ankle foot orthosis (AFO) to stabilize the joint, supplement deficient push-

off and aid toe clearance during stepping. 1617 An AFO limits ankle excursion and simultaneously

influences excursion of the knee and hip, thereby allowing the person to gain more proximal

control for walking.232235 AFOs also improve overground walking speed and interlimb

kinematics in persons with ISCI.236 These broader benefits may lead to the assumption that a role

exists for AFOs as permissive devices in recovery based interventions such as locomotor training

(LT).

The AFO induced alterations in stepping mechanics, however, might not coalesce with the

training principles of LT.24 During LT, stepping is retrained by practice of task-specific

repetitive, rhythmic, stepping kinematics over a treadmill using task facilitatory training

variables such as body weight support and manual assistance.3'25 44'125 The training is based on

the facilitation of intrinsic mechanisms within the spinal cord that respond to specific afferent

input associated with the task of stepping. The spinal cord processes afferent information arising

from muscle, joint and cutaneous receptors during stepping to adapt the motor output to the

phase of stepping.34-37 For example, hip extension and limb unloading are critical afferent inputs

required to initiate the transition from stance-to-swing during stepping.38'93'129 Limiting hip

extension and/or unloading delay this transition and limit the forward progression of the body.









Proponents of LT therefore hesitate to train with an AFO due to the assumption that it

impacts optimal kinematics and motion-related sensory input required for stepping.24'26'237

Logical concerns expressed are that the AFO might restrict the range of excursion of the ankle

and linked knee and hip joints altering the afferent information related to hip position and load.

Additionally, although stance-to-swing transition with the AFO might be alternatively possible,

the orthosis is a passive element in an otherwise active training protocol where other facilitatory

variables can be adjusted to facilitate independence. Therefore the purpose of the study was to

investigate the use of the device during stepping in healthy, non-injured individuals and persons

with ISCI.

To assess the production of critical ankle, knee and hip joint kinematics required for

stepping the range of excursion of these joints were measured during stepping.70,238-241

Furthermore, measurement of the vertical and horizontal forces reflected the rate of

loading/unloading the lower limbs and peak braking/propulsive force required for forward

progression of the body respectively.24242 As secondary variables of interest, interlimb temporal

and spatial measures of symmetry correlated to the functional task requirements of walking in

these transition phases such as double limb support time and step length were also assessed.5'243

In summary, two experiments were conducted interpreting mechanical information in transition

phases (stance-to-swing and swing-to-stance) while walking with and without the AFO. The first

experiment (Refer to Chapter 4) examined transition phase mechanics with and without an AFO

in healthy, non-injured individuals that provided normative data for subsequently comparing

with persons with ISCI. The second experiment (Refer to Chapter 5) compared transition phase

mechanics observed while walking with and without the AFO in individuals with ISCI to normal









mechanics. The comparison assessed changes in pre-existing stepping related deficits while

walking with the AFO and measured deviance or likeness of the observed change from normal.









CHAPTER 4
EXAMINATION OF WALKING WITH AND WITHOUT AN ANKLE FOOT ORTHOSIS IN
HEALTHY, NON-INJURED INDIVIDUALS-1A

The purpose of the experiment was to examine the effect of an AFO on transition phase

mechanics during walking in healthy, non-injured individuals. Specifically, I assessed the

immediate effect of wearing a PAFO on ipsilateral lower extremity 1) kinematics and 2) kinetics

during the stance-to-swing and swing-to-stance phase of walking.

Hypotheses

1A: Compared to walking without an AFO on a treadmill at speed approximating 1.2 m/s,
wearing an AFO in healthy, non-injured individuals will affect the stance-to-swing transition
mechanics observed on the AFO side: specifically, decrease peak ankle plantar flexion and hip
extension and increase peak knee flexion range and increase peak knee and hip flexor powers
and decrease peak ankle plantarflexor power.

1B: Compared to walking without an AFO on a treadmill at speed approximating 1.2 m/s,
wearing an AFO in healthy, non-injured individuals will affect the swing-to-stance transition
mechanics observed on the AFO side: specifically, decrease ankle plantarflexion and increase
hip and knee flexion range of motion and decrease rate of loading and peak braking force.

Methods

Subject Selection

A sample of convenience consisted of fourteen healthy, non-injured individuals living

independently in the Gainesville community (Table 4-1). Each subject provided informed

consent before participating in the study. The University of Florida Institutional Review Board

and the Veteran Affairs Subcommittee approved the study for clinical investigation. Mean age

and standard deviation was between 26.9+ 3.7 years. Subjects were screened for a medical

history of any neurological, musculoskeletal or orthopedic problem that would affect their

walking performance over the treadmill. Power analysis for determination of sample size was

based on pilot data from three healthy, non-injured participants. The change in hip joint angle

and peak braking force were selected for calculating sample size since these are the variables of









primary interest. It was determined that a sample size of fourteen subjects was required to reach

an alpha level of 0.05 and a power of 0.80

Experimental Set-up

Once the subject had read and signed the informed consent form, motion data was

collected and analyzed using a 3-D motion analysis system in conjunction with an ADAL3D

instrumented split-belt treadmill custom manufactured and calibrated by TECMACHINE

(Cedex, France), mounted flush with the floor and anchored to the foundation. Four Kistler

piezoelectric sensors on each half treadmill allow calculation of the two-dimensional location of

the center of pressure (COP) and the moment about the vertical axis, in addition to the three-

dimensional ground reaction force, under each foot. Belt speeds can be controlled as slow as 0.1

m/s.

Ground reaction forces were recorded at 1000Hz for each limb when in contact with the

treadmill belt. The force plates were allowed to warm-up for at least 15 minutes as per

manufacturer guidelines and calibrated prior to data collection. The walking pattern of the

subjects were captured and analyzed by a Vicon three-dimensional motion analysis system. The

system consists of the VICON 612 Datastation with twelve active video channels and a 64

Channel A/D Board for analog signals. There were twelve 1000Hz M2-cameras (Digital CMOS

M2 series cameras have a resolution of 1280 x 1024). Included software was: Workstation,

Polygon, BodyBuilder, Plug-In Gait, Plug-In Modeller, and Real Time II. The twelve cameras

had a frame rate of 60-120 fps and used infrared (IR) light-emitting diode strobes, which were

gen-locked. Static calibration of the system used the clinical L-frame, which contains 4 retro

reflective markers, being placed in a predetermined position on the motion analysis force

platforms. Following this a dynamic calibration was done using a 500mm wand that was moved

around the capture area for approximately 20 seconds. Analog video data were also collected









using a standard camcorder recording at 100 fps with its optical axis perpendicular to the plane

of interest (i.e. the sagittal plane of motion).

Subject Preparation

Subjects were asked to wear tennis shoes and change into appropriate clothing (dark

colored cycling shorts and shirt) for testing. For trials using the AFO, each subject was fitted

with an off the shelf posterior leaf spring ankle foot orthosis (PAFO). Fitting was assessed by

measuring fit inside shoe, length of the calf shell and that of the footplate. Standardized fitting

included using a PAFO whose length fits an inch to two below the fibular head when donned and

whose footplate length extends till the tip of the toes.244 Lightweight retro-reflective markers

were attached to the following bony landmarks: posterior superior iliac spines (PSIS), anterior

superior iliac spines (ASIS), knee-joint axes, lateral malleoli, medial malleoli, clavicular notch,

sternum, C7, T10, and acromium processes. The second foot ray, base of the 5th metatarsal and

the heel markers were approximated on the subjects' shoes. Clusters of markers were attached to

the pelvis, thigh, shank, and foot segments. This modified Helen Hayes marker set is commonly

used to capture bilateral 3D kinematics using a twelve-camera VICON motion analysis

system.208

Each subject was fitted with a body weight supporting harness equipped with an additional

overhead safety catch. The harness and safety catch when used either with or without BWS

provided safety to the person walking on the treadmill and holds or catches the person if he or

she should lose their balance, stumble or begin to fall (Figure 4-1).

Procedure

After equipment set-up and subject preparation, the walking trials over the instrumented

treadmill were recorded. First, subjects were asked to stand with one leg on each belt of the

instrumented treadmill to record a static trial. The static trial was used to create the subject









specific model by defining joint center locations and segment lengths. The leg chosen for

donning the AFO and the order of testing with and without it was randomized for each subject.

For the AFO trial, each subject was requested to wear a unilateral, size-fitted PAFO. The insole

of the shoe was removed in order to fit the AFO and to even out the limb length on both sides.

The subject walked on the instrumented treadmill for the collection of kinematic and kinetic data

with the overhead safety and harness. Subjects were permitted to practice walking on the

treadmill until they achieved steady state walking at the speed of 1.2m/s and comfort while

walking in this environment. Once the subject felt comfortable at the set speed and the

investigator viewed a steady-state pattern of walking, kinematic and kinetic data was collected

for 30 seconds in each of the two conditions. After data collection, the trial was processed to

verify if all the desired data was collected properly. Rest was provided during testing, as

requested. This experiment took approximately two hours from the start for set-up and data

collection.

Data Processing

Kinetic data (Ground reaction forces and moments) and segment kinematic data was low

pass filtered with zero lag digital Butterworth filter (20 and 9 Hz cut-off frequencies

respectively). Software for Interactive Musculoskeletal Modeling (SIMM) was used to create

subject specific models. Segment inertial properties were calculated for each subject based on the

subject's mass and segment lengths. SIMM and SDFast performed an inverse dynamics analysis

for each trial.115'117 All data were averaged across trials for each subject. The kinematic and

kinetic data from each trial was normalized to percent stride using Matlab code and then

compared between the two conditions.









Data Analysis

For each phase of interest, the two conditions (with and without AFO) were compared

using a Hotelling's T2-test, which is a multivariate analogue of the paired t-test. The test is a

multivariate extension of the Student's t-test for paired data in comparison of mean difference

vectors, i.e. the differences of two or more dependent variables considered twice in the same

subjects.245 The primary dependent variables compared included the rate of limb loading, peak

braking force, excursion and power at the ankle, knee and hip joints. For the stance to swing

phase the excursion and power at the ankle, knee and hip joints were analyzed collectively. For

the swing to stance phase the joint excursion, rate of limb loading and peak braking force were

analyzed collectively. Interlimb temporal and spatial measures of symmetry i.e. double limb

support time and step length, our secondary dependent variables were compared between

conditions using a paired t-test. Significance level was set at p< 0.05. To correct for multiple

comparisons, a Holm's step down method was used that adjusted p-values for each research

question.246

Results

Tables 4-2, 4-3 and 4-4 and Figures 4-2, 4-3, 4-4 & 4-5 show the mean and standard

deviation in kinematic and kinetic measures related to each phase with and without the PAFO

ipsilaterally. In the stance-to-swing phase a significant decrease in peak ankle plantar flexion, hip

extension and peak plantarflexor power were noted while walking with a PAFO. In the stance-to-

swing phase, while walking with a PAFO, a significant increase in hip flexion, decrease in the

rate of loading and peak braking force were observed. With regards to interlimb coordination

with the PAFO, double limb support time increased significantly on the ipsilateral limb.









Discussion

The main finding of the study was that in the stance-to-swing phase, donning a PAFO

kinematically decreased peak ankle plantarflexion and hip extension and kinetically decreased

plantarflexor power. Additionally, in the swing-to-stance phase, wearing a PAFO kinematically

increased hip flexion, decreased ankle dorsiflexion and kinetically decreased peak braking force

and increased rate of limb loading.

In stance-to-swing phase, with a PAFO, non-injured subjects demonstrated a significant

decrease in hip extension and ankle plantarflexion which coincided with a decrease in propulsive

force. Generation of plantarflexor power is vital for forward progression of the body. The ankle

plantar flexors provide -70% of the joint work during walking.119,247 However, bracing the ankle

decreased plantarflexor power generation by 17%. Reduction in power may have resulted from a

decrease in angular velocity or moment and subsequently contributed to slowing limb

progression.115 Therefore the PAFO reduced the ability of the ankle to contribute to push-off in

the stance-to-swing transition phase. Interestingly, the effect of the brace was not only isolated to

the ankle but also observed at the hip joint. A decrease in hip extension observed as a result of

the brace could lead to the poor stretch of the hip flexor muscles thereby increasing the difficulty

in initiating swing.38

In the swing-to-stance phase, the primary function of the PAFO is to prevent footdrop.

However, with a PAFO non-injured individuals demonstrated a significant increase in hip

flexion and a decrease in ankle dorsiflexion thereby affecting the heel rocker. The heel rocker is

the first phase in the gait cycle after initial contact that determines the limb's loading response.33

The momentum generated by the fall of body weight onto the stance limb is preserved by this

heel rocker. Normal initial contact is made by the calcaneal tuberosity, which becomes the

fulcrum about which the foot and tibia move. With a PAFO, an increase in hip flexion and a









simultaneous decrease in ankle dorsiflexion in the swing-to-stance phase limit the smooth

transfer of body weight onto the stance limb. Kinetically, this coincided with the decrease in

peak braking force on the PAFO side and a delay in the rate of loading. Temporally, a decreased

loading rate corresponded with an increase in the initial phase of double support time on the

ipsilateral side indicative of a delay in shifting the weight from one limb to another.

In non-injured individuals, our study demonstrated that an orthosis altered the transition

phase kinematics and kinetics crucial to stepping. Our findings could have potential implications

in neurologically impaired individuals in whom brace walking is common. Past studies have

evaluated the benefits of an ankle foot orthosis in different neurological populations.248'249

However these studies have evaluated the compensatory benefits of using an orthosis on

temporal and spatial patterns of walking with the device without accounting for changes in joint

kinematics and kinetics in the transition phases of walking. Appropriate joint kinematics and

kinetics in the transition phases are crucial for providing optimal motion- related sensory input to

a compromised nervous system. Utilization of these sensory inputs has been shown to aid in

retraining the neuromuscular system for walking recovery.24'26'44 Our study explored this

paradigm shift by re-examining the use of such devices for walking by examining joint

kinematics and kinetics during the transition phase of walking. In our study, use of an orthosis

failed to produce desired proximal joint kinematics such as hip extension in the stance-to-swing

phase of walking and meet the functional task requirements such as rate of loading in able-

bodied individuals. Since the orthosis affected walking in non-injured individuals its effect on

gait in individuals nervous system disorders could be more pronounced. Therefore, the results of

our study suggest that the purpose and functional implication of an ankle foot orthosis needs to

be evaluated rigourously in neurological populations. Neurobiologically driven recovery based









interventions such as locomotor training targeted at providing normal walking kinematics and

thereby appropriate motion related sensory input to a compromised nervous system need to

weigh the use of such devices cautiously.

Limitation

We did not collect EMG in these individuals which would help correlate our findings with

lower limb muscle activity. Also, examination of kinematic and kinetic changes with a PAFO,

limits the generalizability of our results to other more rigid devices such as solid or hinged ankle

foot orthoses.

Conclusion

A minimally restrictive device such as a PAFO in non-injured individuals impacted the

provision of critical afferent input during the transition phases of walking. Proximal hip

extension crucial for the transition from stance-to-swing and the rate of loading during the

swing-to-stance phase were decreased. Intuitively, use of more rigid devices could exaggerate

these findings. Non-injured individuals were able to adapt to walking with the PAFO by

increasing the double support time ipsilaterally. Given the clinical relevance of our study, the use

of a PAFO for neurological populations needs to be systematically assessed.








































Figure 4-1. Participant with safety harness walking on an instrumented treadmill. Reflective
markers were applied to bony landmarks on the pelvis and bilateral lower extremities
for this study.













Non AFO side Hip Angle


40 60
percent gait cycle


AFO side Ankle Angle


N
N /

N\
',


20 40 60
percent gait cycle


Non AFO side Ankle Angle


20 40 60
percent gait cycle


stance-to-swing


80 100


20 40 60
percent gait cycle


Figure 4-2. Ipsilateral and contralateral average joint angles with and without the ankle foot

orthosis (AFO) ipsilaterally. The ipsilateral decrease in peak hip extension and
increase in hip flexion [A], no change in peak knee flexion [B] and decrease in peak

ankle plantarflexion and dorsiflexion [C] during the swing-to-stance and stance-to-
swing phase of the gait cycle are highlighted by dotted circles. Vertical lines represent

point of toe-off in the gait cycle. Significant changes represented as p<0.05.


I~f
/ N


/
/ /
1 /I


80 100


W -LU
a)
aD
03
a>
*o -40


80 100


AFO side Hip Angle












stance-to-swing phase


Hip extension


Knee
flexion


Ankle
plantarflexion [ Ankle
plantarflexion
With AFO


Without AFO


With AFO


THip
flexion


SKnee
flexion


Ankle
dorsiflexion


Without AFO


Ankle
dorsiflexion
with AFO

With AFO


Figure 4-3. Stick figure representing the changes in individual joint motion in the stance-to-
swing and swing-to-stance phase of the gait cycle with and without the ankle foot
orthosis (AFO).


swing-to-stance phase










AFO Side Hip Power


D
0.25 r


20 40 60 80 100
percent gait cycle
AFO Side Knee Power


E
0.15


Non AFO Side Hip Power


20 40 60 80
percent gait cycle
Non AFO Side Knee Power


100


20 40 60 80 100 20 40 60
percent gait cycle percent gait cycle


AFO Side Ankle Power


20 40 60 80 100
percent gait cycle


Non AFO Side Ankle Power

- with AFO
no AFO


20 40 60
percent gait cycle


80 100


Figure 4-4. Ipsilateral and contralateral average joint powers during the stance-to-swing phase of
the gait cycle with and without the ankle foot orthosis (AFO) ipsilaterally. The
ipsilateral hip flexor power [A], knee flexor power [B] and ankle plantarflexor power
[C] are highlighted by dotted circles. Vertical lines represent point of toe-off in the
gait cycle. Significant changes represented as p<0.05.


B
0.15r


C
0.25 r













AFO Side GRF Vertical
P<0.05 I


20 40 60 80 100
percent gait cycle


AFO Side GRF AP


20 40 60
percent gait cycle


80 100


Non AFO Side GRF Vertical


40 60
percent gait cycle


Non AFO Side GRF AP

with AFO
- no AFO /


20 40 60
percent gait cycle


80 100


Figure 4-5. Ipsilateral and contralateral vertical and horizontal (AP) ground reaction forces
(GRF) during the swing-to-stance phase of the gait cycle with and without the ankle
foot orthosis (AFO) ipsilaterally. The prolonged rate of loading during vertical
loading [A] and decrease in horizontal braking force [C] are highlighted by dotted
circles (p<0.05) ipsilaterally.


C
0.2r









Table 4-1.
ID
N1
N2
N3
N4
N5
N6
N7
N8
N9
N10
N11
N12
N13
N14


Demographics of the study participants.
Age Sex
24 F
27 M
30 M
28 F
26 F
31 F
27 M
24 F
22 M
27 F
23 M
29 M
23 M
36 F


Orthotic
R
R
R
R
R
R
R
R
R
L
L
L
L
L


Size
Small
Large
X-large
Medium
Medium
Medium
Large
Medium
Large
Small
Large
Medium
Large
Small


Table 4-2. Hip, knee and ankle joint kinematic and kinetic data while walking with and without
an ankle foot orthosis (AFO) during the stance-to-swing phase of the gait cycle.
Stance-to-swing phase Without AFO Standard With AFO Standard p-
deviation deviation value
Peak hip ioint extension -8.67 5.58 -6.77+ 5.51 .001*


(degrees)
Peak knee joint flexion
(degrees)
Peak ankle joint
plantarflexion
(degrees)
Hip joint power
(Watts/ body weight)
Knee joint power
(Watts/ body weight)
Ankle joint power
(Watts/ body weight)
*Significant changes.


-64.87+ 3.80

-19.45+ 5.71


0.11+ 0.04

-0.15+ 0.02

0.18+ 0.03


-64.54+ 4.15

-12.01+ 5.28


0.12+ 0.04

-0.15+ 0.03

0.15+ 0.03


.000*


.899

.893


.000*









Table 4-3. Hip, knee and ankle joint kinematic kinematic and kinetic data while walking with
and without an ankle foot orthosis (AFO) during swing-to-stance phase of the gait
cycle.


Swing-to-stance phase
Peak hip joint flexion
(degrees)
Peak knee joint flexion
(degrees)
Peak ankle joint
plantarflexion
(in degrees)
Rate of loading
(N/kg)
Peak braking force
(N/kg)
*Significant changes.


Without AFO Standard
deviation
32.75+ 4.73

-17.86+ 3.47


-8.40+ 3.55


0.06 0.01

-0.16 0.02


With AFO Standard
deviation


35.60+ 5.13

-17.81+ 3.55


-10.90+ 4.77


0.05+ 0.01

-0.15+ 0.02


Table 4-4. Average interlimb temporal and spatial
foot orthosis (AFO).


data while walking with and without an ankle


No AFO Unilateral AFO
Primary Contralateral Primary Contralateral
side side pae side side p-v
Step
length 0.57+ 0.14 0.58+ 0.14 .329 0.56+ 0.15 0.58+ 0.13 .326
(meters)
Double
support
support 0.19 0.02 0.19 0.02 .111 0.22 0.03 0.17 0.02 .000*
time
(seconds)
*Significant changes.


p-
value
.028*


.924


.000*


.018*

.013*









CHAPTER 5
COMPARISON OF WALKING WITH AND WITHOUT ANKLE FOOT ORTHOSIS IN
PERSONS WITH INCOMPLETE SPINAL CORD INJURY-1B

Given the gait deficits after ISCI, two investigative steps that would serve to inform

clinical-decision making for use of an AFO during walking retraining are 1) to examine the

transition phase mechanics in persons with ISCI while walking with and without the AFO and 2)

to compare the observed mechanics in each of the conditions to normal walking mechanics.

Therefore we proposed to assess the changes in walking mechanics in individuals with ISCI

while walking with an AFO and to examine the proximity or deviation of the observed change to

normal matched control values.

Specific Aims and Hypothesis

* Aim 1: To compare the immediate effect of walking with and without wearing an AFO on
lower extremity kinematics and kinetics during the stance-to-swing and swing-to-stance
phase of walking in persons after ISCI.

* Hypothesis la: Compared to walking without wearing a AFO, wearing an AFO in persons
with ISCI will significantly change the stance-to-swing transition observed on the AFO side:
specifically, peak ankle plantar flexion, knee flexion and hip extension and peak knee, hip
and plantarflexor powers.

* Hypothesis Ib: Compared to walking without wearing an AFO, wearing an AFO in persons
with ISCI will significantly change the swing-to-stance transition observed on the AFO side:
specifically, ankle plantarflexion, knee and hip flexion and rate of loading and braking force.

* Aim 2: To compare lower extremity kinematics and kinetics during the stance-to-swing and
swing-to-stance phase while walking with and without wearing an AFO in individuals with
ISCI to that of healthy, non-injured age, weight, height and speed matched controls.

* Hypothesis 2a: In persons with ISCI, the stance-to-swing transition will be significantly
deviated from normal while walking with an AFO compared to walking without one.
Specifically, with an AFO there will be a decrease in peak ankle plantar flexion and hip
extension and increase in peak knee flexion and an increase in peak knee and hip flexor
powers and decrease peak ankle plantarflexor power.

* Hypothesis 2b: In persons with ISCI, the swing-to-stance transition will be significantly
deviated from normal while walking with an AFO compared to walking without one.
Specifically, with an AFO there will be a decrease in ankle plantarflexion and increase in
knee and hip flexion and a decrease in rate of loading and peak braking force.









Methods


Subject Selection

Eight persons with ISCI ranging between 18-80 years and their height, weight and age

matched controls were recruited for this experiment and signed an informed consent form

approved by University of Florida Institutional Review Board and the Veteran Affairs

Subcommittee for Clinical Investigation. Participant demographics are tabulated in Table 5-1.

American Spinal Injury Association (ASIA) motor score and impairment scale data were

collected from all participants with ISCI to assess the degree of impairment in each leg. The

criteria for inclusion in the study were as follows:

1) Persons with ISCI classified as ASIA D

2) Medically stable

3) Have quadriceps strength of at least 3/5

4) Have decreased ankle strength (dorsiflexor strength of less than or equal to 4/5)

5) and/or absent or impaired proprioception at the ankle

6) Can stand unaided for one minute

7) Can walk with minimal assistive device such as the cane but does not use an ankle foot
orthosis.

Exclusion criteria included persons who were unable to follow 3 step commands,

amputation, medical instability, significant musculoskeletal problems other than SCI that limit

hip and knee extension or ankle plantarflexion to neutral. Sample size was determined from

previous studies examining similar gait characteristics in these individuals.250,251 Pepin et al.

found significant differences (p<0.01) in hip extension in a sample size of seven persons with

ISCI compared to their non-injured counterparts at matched speeds while comparing the

adaptability of gait pattern in individuals with ISCI to different walking speeds.250 The change in

hip joint angle (hip extension change from pilot data=5.25 degrees, SD from Pepin's study= + 5)









were used to calculate sample size for our study. It was determined that a sample of 8 subjects

will be required to reach an alpha level of 0.05 and a power of 0.80.

Experimental Set-up

Once the subject had read and signed the informed consent form, motion data was

collected and analyzed using a 3-D motion analysis system in conjunction with an ADAL3D

instrumented split-belt treadmill custom manufactured and calibrated by TECMACHINE

(Cedex, France), mounted flush with the floor and anchored to the foundation. Four Kistler

piezoelectric sensors on each half treadmill allow calculation of the two-dimensional location of

the center of pressure (COP) and the moment about the vertical axis, in addition to the three-

dimensional ground reaction force, under each foot. Belt speeds can be controlled as slow as 0.1

m/s.

Ground reaction forces were recorded at 1000Hz for each limb when in contact with the

treadmill belt. The force plates were allowed to warm-up for at least 15 minutes as per

manufacturer guidelines and calibrated prior to data collection. The walking pattern of the

subjects were captured and analyzed by a Vicon three-dimensional motion analysis system. The

system consists of the VICON 612 Datastation with twelve active video channels and a 64

Channel A/D Board for analog signals. There were twelve 1000Hz M2-cameras (Digital CMOS

M2 series cameras have a resolution of 1280 x 1024). Included software was: Workstation,

Polygon, BodyBuilder, Plug-In Gait, Plug-In Modeller, and Real Time II. The twelve cameras

had a frame rate of 60-120 fps and used infrared (IR) light-emitting diode strobes, which were

gen-locked. Static calibration of the system used the clinical L-frame, which contains 4 retro

reflective markers, being placed in a predetermined position on the motion analysis force

platforms. Following this a dynamic calibration was done using a 500mm wand that was moved

around the capture area for approximately 20 seconds. Analog video data were also collected









using a standard camcorder recording at 100 fps with its optical axis perpendicular to the plane

of interest (i.e. the sagittal plane of motion).

Subject Preparation

Subjects were asked to wear tennis shoes and change into appropriate clothing (dark

colored cycling shorts and shirt) for testing. For trials using the AFO, each subject was fitted

with an off the shelf posterior leaf spring ankle foot orthosis (PAFO). Fitting was assessed by

measuring fit inside shoe, length of the calf shell and that of the footplate. Standardized fitting

included using a PAFO whose length fits an inch to two below the fibular head when donned and

whose footplate length extends till the tip of the toes.244 Lightweight retro-reflective markers

were attached to the following bony landmarks: posterior superior iliac spines (PSIS), anterior

superior iliac spines (ASIS), knee-joint axes, lateral malleoli, medial malleoli, clavicular notch,

sternum, C7, T10, and acromium processes. The second foot ray, base of the 5th metatarsal and

the heel markers were approximated on the subjects' shoes. Clusters of markers were attached to

the pelvis, thigh, shank, and foot segments. This modified Helen Hayes marker set is commonly

used to capture bilateral 3D kinematics using a twelve-camera VICON motion analysis

system.208

Each subject was fitted with a body weight supporting harness equipped with an additional

overhead safety catch. The harness and safety catch when used either with or without BWS

provided safety to the person walking on the treadmill and holds or catches the person if he or

she should lose their balance, stumble or begin to fall (Figure 4-1).

Procedure

After equipment set-up and subject preparation, the walking trials over the instrumented

treadmill were recorded. First, subjects were asked to stand with one leg on each belt of the

instrumented treadmill to record a static trial. The static trial was used to create the subject









specific model by defining joint center locations and segment lengths. The leg chosen for

donning the AFO and the order of testing with and without it was randomized for each subject.

For the AFO trial, each subject was requested to wear a unilateral, size-fitted PAFO. The insole

of the shoe was removed in order to fit the AFO and to even out the limb length on both sides.

The subject walked on the instrumented treadmill for the collection of kinematic and kinetic data

with the overhead safety and harness. Subjects were permitted to practice walking on the

treadmill until they achieved steady state walking at the speed of 1.2m/s and comfort while

walking in this environment. Once the subject felt comfortable at the set speed and the

investigator viewed a steady-state pattern of walking, kinematic and kinetic data was collected

for 30 seconds in each of the two conditions. After data collection, the trial was processed to

verify if all the desired data was collected properly. Rest was provided during testing, as

requested. This experiment took approximately two hours from the start for set-up and data

collection.

Data Processing

Kinetic data (Ground reaction forces and moments) and segment kinematic data was low

pass filtered with zero lag digital Butterworth filter (20 and 9 Hz cut-off frequencies

respectively). Software for Interactive Musculoskeletal Modeling (SIMM) was used to create

subject specific models. Segment inertial properties were calculated for each subject based on the

subject's mass and segment lengths. SIMM and SDFast performed an inverse dynamics analysis

for each trial.115'117 All data were averaged across trials for each subject. The kinematic and

kinetic data from each trial was normalized to percent stride using Matlab code and then

compared between the two conditions.









Data Analysis

For each phase of interest, the two conditions (with and without AFO) were compared

using a Hotelling's T2-test, which is a multivariate analogue of the paired t-test. The test is a

multivariate extension of the Student's t-test for paired data in comparison of mean difference

vectors, i.e. the differences of two or more dependent variables considered twice in the same

subjects.245 The dependent variables compared included the rate of limb loading, peak braking

force, excursion and power at the ankle, knee and hip joints. For the stance-to-swing phase, the

excursion and power at the ankle, knee and hip joints were analyzed collectively. For the swing-

to-stance phase, the joint excursion, limb loading and peak braking force were analyzed

collectively. Interlimb temporal and spatial measures of symmetry i.e. double limb support time

and step length, our secondary dependent variables were compared between conditions using a

paired t-test. Significance level was set at p< 0.05. To correct for multiple comparisons, a Holm's

step down method was used that adjusted p-values for each research question. The same analyses

were repeated comparing the control data to individuals with ISCI walking with and without the

AFO.246

Results

Figures 5-1 through Figure 5-6 and Tables 5-2 through Table 5-8 show the change in the

kinematic and kinetic measures related to each gait phases with and without the PAFO

ipsilaterally in individuals with ISCI and their matched non-injured controls.

Within-Subject Comparisons for Individuals with ISCI

In the stance-to-swing phase, with a PAFO, a decrease in hip extension was observed

within subjects (Figure 5-1 & Table 5-2). Likewise, in the swing-to-stance phase, with a PAFO,

an increase in peak knee flexion was observed within subjects (Figure 5-1 & Table 5-3). After









correcting for multiple comparisons in these phases, weak statistical support existed for these

variables.

Between-Subject Comparisons for Individuals with ISCI Walking Without a PAFO and
Their Matched Control

In the stance-to-swing phase, without a PAFO, a decrease in peak knee joint flexion and

knee joint power were observed in individuals with ISCI. After correcting for multiple

comparisons in this phase, weak statistical support existed for these variables (Figure 5-1 &

Table 5-4). In the swing-to-stance phase, without a PAFO, a significant increase in peak hip

flexion was observed in individuals with ISCI, which after correcting for multiple comparisons

was statistically significant (Figure 5-1 & Table 5-5).

Between-Subject Comparisons for Individuals with ISCI Walking with a PAFO and Their
Matched Control

In the stance-to-swing phase, with a PAFO, a decrease in peak knee joint flexion and knee

joint power were observed in individuals with ISCI. After correcting for multiple comparisons in

this phase, weak statistical support existed for these variables (Figure 5-1 & Table 5-6). In the

swing-to-stance phase, with a PAFO, an increase in peak hip flexion was observed in individuals

with ISCI, which after correcting for multiple comparisons was statistically significant (Figure 5-

1 & Table 5-7).

Temporal and Spatial Comparisons Within Subjects

Both the spatial and temporal measures of symmetry namely step length and double limb

support time did not demonstrate significant differences between the two walking conditions

(Figure 5-5, 5-6 & Table 5-8 ).

Discussion

The main finding of the study was that in individuals with ISCI, donning a PAFO

decreased hip extension in the stance-to-swing phase of walking compared to walking without it









(Figure 5-4 & 5-8). While walking with a PAFO, step length value and interlimb symmetry did

not change between conditions. Interestingly, double limb support time on the ipsilateral limb

increased in 5/8 subjects while walking with a PAFO (Figure 5-6). Additionally, symmetry

indices for the interlimb double limb support time increased concurrent with the increase in

ipsilateral double limb support time.


Furthermore, compared to normal walking, gait in individuals with ISCI walking without a

PAFO was characterized by a significant decrease in knee flexion in the stance-to-swing phase

that also correlated to a decrease in knee flexor power. Likewise, a significant increase in hip

flexion in the swing-to-stance phase of walking was observed. Interestingly, we did not observe a

trend for improvement or deterioration in these pre-existing gait deviations with the PAFO when

compared to matched controls. The increase in hip flexion in the swing-to-stance transition phase

and decrease in knee flexion in the stance-to-swing phase are common gait deviations

characteristic to individuals with ISCI. Pepin et al (2003) has demonstrated a significant increase

in hip flexion at the time of heel contact in individuals with ISCI.250 Likewise the reduction in

knee angular velocity during walking is a common gait deviation observed in individuals with

ISCI which could account for the decrease in knee flexion.70


With regards to power generation, during the stance-to-swing phase, in both the conditions,

individuals with ISCI demonstrated an ipsilateral decrease in ankle plantarflexor, knee flexor and

hip flexor power compared to the contralateral limb. Generation of plantarflexor power is vital

for forward progression of the body. The ankle plantar flexors provide -70% of the joint work

during walking.119,247 However, wearing a PAFO did not augment power generation at the ankle.

Additionally, a decrease in horizontal propulsive and braking force were also noted on the

ipsilateral side with and without PAFO compared to the contralateral limb. The decrease in









propulsive force could result from inability to generate sufficient power ipsilaterally. Likewise,

the reduced braking force compared to the contralateral limb could result from the significant

increase in hip flexion observed in these individuals.

Clinical Implication

Unlike past studies reporting the compensatory benefits of using an ankle foot orthosis,

this study uniquely examined the ability of the PAFO to meet the normal kinematic and kinetic

task requirements of stepping. Failure to improve limb kinematics and kinetics with a PAFO

during treadmill walking is suggestive of the inability of the device to provide a normal walking

pattern in individuals after ISCI. Importantly, a distally worn PAFO impacted proximal joint

excursion by limiting hip extension. Hip extension is one of the essential kinematic features of

the stance-to-swing transition during walking. Studies have shown that preventing the hip from

attaining an extended position inhibited the generation of the flexor burst and hence the onset of

the swing phase.129130, 38 The use of PAFO for step retraining on a treadmill after SCI may thus

hinder achievement of the task-specific, locomotion-related afferent input used to retrain

stepping via sensorimotor activation of the neuromuscular system.37

Limitations

The results of our study demonstrated weak statistical support for within subject

comparisons. We had performed our initial power and sample estimates based on preliminary

data for our primary outcome of interest: hip extension. Consequently, we found this variable

was different between conditions. However, the secondary variables remained under powered to

find true differences if they existed. With a larger sample size very small differences would be

detected as significant. However, statistical significance needs to be assessed with caution since

it does not imply if the difference between the variables is large or important. Additionally, apart









from the kinematic and kinetic data, collection of electromyographic data in these individuals

would have helped correlate our findings with lower limb muscle activity.

Conclusion

For the rehabilitation specialist, the characterization of gait deficits observed in ISCI subjects

is important for treatment purposes. Likewise, characterization of improvement or deterioration

in these deficits with the use of orthotic devices is important particularly for developing and

bettering new rehabilitation approaches. As newer interventions are being developed, the

therapeutic rationales for the use of orthotic devices might change based on the guiding

principles of these interventions. Traditionally, wearing an AFO in individuals with ISCI has

been considered compensatory solution or a "quick fix" for remediating gait deficits resulting

from muscular weakness, incoordination and spasticity. However, from a neurobiological control

of walking based perspective, an AFO may alter the sensory experience necessary for retraining

the nervous system and might not produce the desired therapeutic effect.24'252

Interestingly, in our study, the use of a minimally restrictive PAFO decreased hip extension in

participants with ISCI. The observed decrease could impact the provision of at least one critical

afferent input key to the restoration of walking. Furthermore, use of more rigid devices is likely

to exaggerate our findings. Consequently, if the goal of recovery based interventions such as

locomotor training is to provide optimal limb kinematics, the use of a PAFO for stepping would

not coincide with the principles of training.












Non AFO side -i, Angle


L,:, 4,1 r':'
percent gait cycle


AFO side Knee Angle


40 60
percent gait cycle


Non AFO side Knee Angle


20 40 60 80 100 20 40 60 80 100
percent gait cycle percent gait cycle


C
AFO side Ankle Ang
20

10 0 *.






-;'n" '....


20 40 60
percent gait cycle


le


\tllIl '-1,-t I- I"ll'


80 100


Non AFO side Ankle Angle











l^ Control
Switch AFO
no AFO


20 40 60
percent gait cycle


80 100


Figure 5-1. Ipsilateral and contralateral average joint angles during the swing-to-stance and
stance-to-swing phase of the gait cycle with and without the ankle foot orthosis
(AFO) ipsilaterally. The ipsilateral peak hip flexion and extension [A], peak knee
flexion [B] and peak ankle plantarflexion and dorsiflexion [C] are highlighted by
dotted circles. The gray shaded area represents matched control data. Vertical lines
represent point of toe-off in the gait cycle. Significant changes in the joint angles
represented as p<0.05.


AFO side Hip Angle


2.
~
5`
r
--ii -




.











Non AFO side Hip Power


0.1



0.06

0 04- \ ,: .




-0.02 / .
sta
A AJA


20 40 60
percent ,:v:i cycle

AFO side Knee Power


~- N,


ce-to-swing


80 100


20 40 60 80 100
percent pal cycle


percent gait cycle


E Non AFO side Knee Power
0 04

0,021-







11 ,
- 004

-0.06
-n r. -


20 40 60
percent 'pal cycle


80 100


AFO side Ankle Power


20 40 60
percent ,'ill cycle


F Non AFO side Ankle Power
0 15
01 Control
0.1 with AFO
no AFO

0 05
,---) ,. / / '
[/


80 100


20 40 60
percent 'i cycle


80 100


Figure 5-2. Ipsilateral and contralateral average joint powers during the swing-to-stance and
stance-to-swing phase of the gait cycle with and without the ankle foot orthosis
(AFO) ipsilaterally. The ipsilateral hip flexor power [A], knee flexor power [B] and
peak ankle plantarflexor power [C] are highlighted by dotted circles. The gray shaded
area represents matched control data. Vertical lines represent point of toe-off in the
gait in the gait cycle. Significant changes in the joint powers represented as p<0.05.


0,02 [


-I[r lI[1 1


AFO side -,ii, Power










A AFO side Vertical GRF


20 40 60 80 100
percent gait cycle

AFO side Anterior/Posterior GRF


20 40 60
percent .:,I cycle


-0.05


80 100


20 40 60
percent gait cycle


Non AFO side Anterior/Posterior GRF


20 40 60
percent gait cycle


Figure 5-3. Ipsilateral and contralateral vertical and horizontal (AP) ground reaction forces
(GRF) during the swing-to-stance phase of the gait cycle with and without the ankle
foot orthosis (AFO) ipsilaterally. The rate of loading during vertical loading [A] and
the horizontal braking force [C] are highlighted by dotted lines ipsilaterally. The gray
shaded area represents matched control data.


C
0.15r


80 100


80 100


Non AFO side Vertical GRF










20.00

15.00


S10.00
a,
0)
. 5.00

o 0.00

| -5.00

I -10.00


t AFO
1 No AFO
110 *MC


-15.00

-20.00


Subject ID


Figure 5-4. Ipsilateral hip extension values with and without the AFO during the stance-to-
swing phase of the gait cycle for spinal cord injured individuals and their matched
controls.


L [A


-

-


16mo--V






















0.45
E

- 0.40

.O
- 0.35


0.30


0.25


* Unilateral AFO Ipsilateral
* Unilateral AFO Contralateral
* No AFO Ipsilateral
* No AFO Contralateral


0.20 ---
Ipsilateral Contralateral Ipsilateral Contralateral

Unilateral AFO No AFO



Figure 5-5. Step length while walking with and without the AFO in individuals with incomplete
spinal cord injury.
















S0.50
0

E 0.45

E Unilateral AFO Ipsilateral
04 Unilateral AFO Contralateral
o 0.40
m No AFO Ipsilateral
m No AFO Contralateral
.0
E 0.35


0
o 0.30



0.25
Ipsilateral Contralateral Ipsilateral Contralateral

Unilateral AFO No AFO



Figure 5-6. Double limb support time while walking with and without the PAFO in individuals
with incomplete spinal cord injury.









Table 5-1. Participant demographics of individuals with ISCI and control subjects.
Injury duration Speed
ID Age Sex Height Orthotic Injury level ASIA score months (m/sec) Assitive/Orthotic device
11 46 M 5'6" L C5-C6 D 10 0.8 NA
12 33 M 5'11" R C6-7 D 14 0.6 Cane on left
13 66 M 6' 3" L C7 D 79 0.5 Cane on right
14 49 F 5'5" L C4-C5 D 46 0.7 NA
16 49 F 5'10" L C7 D 23 0.7 NA
17 40 F 5'8" L C2-T1 D 253 0.5 NA
19 25 M 5'11" R T4-5 D 90 0.4 Solid right AFO
110 57 M 6'2" L C5 D 122 0.3 Cane on right

Cl M 5'7" L 0.8


C2 32
C3 62
C4 49
c C6 52
C7 40
C9 27
C10 52


M 5'10"
M 6'3"
F 5'3"
F 5'7"
F 5'6"
M 5'11"
M 5"10"











Table 5-2. Hip, knee and ankle joint kinematic and kinetic data during the stance-to-swing phase
of the gait cycle while walking with and without an ankle foot orthosis (AFO) in
individuals with incomplete spinal cord injury.


Without AFO With AFO


Peak hip joint extension
-2.57+ 10.57
(degrees)
Peak knee joint flexion -48.14+ 8.94
(degrees)
Peak ankle joint plantarflexion 6.32 8.23
(degrees)
Hip joint power
(Watts/ Body weight)
Knee joint power
(Watts/ Body weight)
Ankle joint power
(Watts/ Body weight) 0
* Represents significant changes


-1.18+ 9.72

-47.61+ 8.85

-2.97 5.22

0.04 0.02

-0.03+ 0.01

0.08+ 0.09


Confidence Interval
Lower Upper


-2.862

-3.114


-.313

2.057


-6.863 .173


-.006

-.006

-.111


.006

.001

.051


Stance-to-swing


p-value


.022*

.643

.059

1.000

.170

.411









Table 5-3. Hip, knee and ankle joint kinematic and kinetic data during the swing-to-stance phase
of the gait cycle while walking with and without an ankle foot orthosis (AFO) in
individuals with incomplete spinal cord injury.


Swing-to-stance

Peak hip joint
flexion
(degrees)
Peak knee joint
flexion
(degrees)
Peak ankle joint
plantarflexion
(degrees)
Rate of loading
Peak braking
force


Without AFO


31.15 8.10


-18.31 7.50


-6.73 6.98


0.04 0.01

-0.06 0.02


With AFO


32.13+ 6.81


-19.82+ 6.16


-6.87 + 5.19


0.04+ 0.01

-0.06 + 0.03


Confidence interval
Lower Upper


-3.096 1.126


.016


3.006


-2.637 2.917


-.001

-.011


.006


* Represents significant changes


p-value


.306


.048*


.908


.170


1.000










Table 5-4. Kinematic and kinetic data during the stance-to-swing phase of the gait cycle at the
hip, knee and ankle joints while walking without an ankle foot orthosis (AFO) in
individuals with incomplete spinal cord injury compared to their matched, non-
iniured controls.


Stance-to-swing

Peak hip joint
extension
(degrees)
Peak knee joint
flexion
(degrees)
Peak ankle joint
plantarflexion
(degrees)
Hip joint power
(Watts/ Body
weight)
Knee joint power
(Watts/ Body
weight)
Ankle joint power
(Watts/ Body
weight)


Without AFO


-2.57 10.57


-48.14+ 8.94


-6.32 8.23


0.04 0.02


-0.03 0.02


0.05+ 0.03


Control


-5.61 6.42


-57 5.20


-9.07 7.51


0.03 0.02


-0.05 0.02


0.07 0.03


Confidence Interval
Lower Upper

-10.437 4.747


-15.327 -2.396


-12.602 7.087


-.018


-.037


-.009


.003


-.001


.042


* Represents significant changes


p-value


.405


.014*


.529


.142


.044*


.178










Table 5-5. Kinematic and kinetic data during the swing-to-stance phase of the gait cycle at the
hip, knee and ankle joints while walking without an ankle foot orthosis (AFO) in
individuals with incomplete spinal cord injury compared to their matched, non-
iniured controls.


Swing-to-stance

Peak hip joint
flexion
(degrees)
Peak knee joint
flexion
(degrees)
Peak ankle joint
plantarflexion
(degrees)
Rate of loading
Peak braking
force


Without AFO


31.15+ 8.10


-18.31+ 7.50


-6.73 6.98

0.04+ 0.01

-0.06 + 0.02


Control


24.93+ 6.18


-14.98+ 4.57


-6.35+ 5.07

0.04+ 0.01

-0.07 0.02


Confidence interval


Lower


Upper


p-value


-10.087 -2.353 .007*


-3.894


-5.455

-.007

-.015


10.561 .312


6.230 .880

.004 .598


.005


.275


* Represents significant changes










Table 5-6. Changes in the hip, knee and ankle joint kinematics and kinetics during the stance-to-
swing phase of the gait cycle while walking with an ankle foot orthosis (AFO) in
individuals with incomplete spinal cord injury compared to their matched, non-
iniured controls.


Stance-to-swing

Peak hip joint
extension
(degrees)
Peak knee joint
flexion
(degrees)
Peak ankle joint
plantarflexion
(degrees)
Hip joint power
(Watts/ Body
weight)
Knee joint power
(Watts/ Body
weight)
Ankle joint power
(Watts/ Body
weight)


With AFO


-1.18 9.72


-47.61 8.85


-2.97 5.22


0.04 0.02


-0.03 0.01


0.08 0.09


Control


-5.61 6.42


-57 5.20


-9.07 7.51


0.03 0.02


-0.05 0.02


0.07 0.03


Confidence Interval
Lower Upper

-11.556 2.694


-16.098 -2.675


-14.275 2.070


-.018


-.035


-.112


.003


-.005


.085


* Represents significant changes


p-value


.185


.013*


.121


.142


.018*


.751









Table 5-7. Kinematic and kinetic changes at the hip, knee and ankle joints during the swing-to-
stance phase of the gait cycle while walking without an ankle foot orthosis (AFO) in
individuals with incomplete spinal cord injury compared to their matched, non-
injured controls.
Confidence interval


Swing-to-stance

Peak hip joint
flexion
(degrees)
Peak knee joint
flexion
(degrees)
Peak ankle joint
plantarflexion
(degrees)
Rate of loading
Peak braking
force


With AFO


32.13+ 6.81


-19.82+ 6.16


-6.87 + 5.19


0.04+ 0.01

-0.06 + 0.03


Control


24.93+ 6.18


-14.98+ 4.57


-6.35+ 5.07


0.04+ 0.01

-0.07 0.02


Lower


-11.159


-1.533


-3.939


-.001

-.020


Upper


p-value


-3.251 .004*


11.218 .116


4.996 .788


.006

.005


.170

.197


* Represents significant changes









Table 5-8. Average interlimb temporal and spatial data while walking with and without an ankle
foot orthosis (AFO) in individuals with incomplete spinal cord injury.
No AFO Unilateral AFO
Ipsilateral Contralateral p- Ipsilateral Contralateral p-
side side value side side value


0.31+ 0.02



0.36+ 0.09


0.31+ 0.03



0.36+ 0.14


.921 0.31+ 0.02



.935 0.38+ 0.13


0.31+ 0.02



0.35+ 0.10


.301



.273


Step
length
(meters)
Double
support
time
(seconds)









CHAPTER 6
PHASE DEPENDENT MODULATION OF SOLEUS H-REFLEX IN HEALTHY, NON-
INJURED INDIVIDUALS WHILE WALKING WITH AN ANKLE FOOT ORTHOSIS

Introduction

Individuals with incomplete spinal cord injury (ISCI) have weakness and/or spasticity of

the musculature below the level of injury making it difficult to meet the functional demands of

gait.68-70 In conventional rehabilitation practice, spasticity or loss of muscle strength are

substituted by compensatory orthotic devices that stabilize, realign and control the range of

excursion of the weakened joint or limb segment to assist with walking.15'75'78 For example, the

posterior leaf spring ankle foot orthosis (PAFO) is used to compensate for deficient push off in

terminal stance and foot drag in swing.16,17,203 With a PAFO, improvement in overground

walking outcomes such as walking speed, stride length, stance knee position and walking

energetic have been documented.236

In spite of the appeal of such compensatory strategies, their use in neurobiologically

driven, recovery-based interventions such as locomotor training for individuals with ISCI is still

controversial.24 This is due to the lack of information about the use of the device in optimizing or

hindering afferent input from joint, muscle and cutaneous receptors fundamental to the

training.34'36'37'133 After SCI, pattern generating neural networks within the spinal cord increases

their reliance on motion-related afferent input from these receptors for maintaining locomotor

control.25'26'44 Limiting ankle excursion with a PAFO may alter the interconnected limb joint

assembly specific to walking and in turn negatively influence the afferent information critical

for stepping.

The soleus H-reflex has been commonly employed as a neural probe in interpreting the

interplay of afferent input and movement control.149,228,253,254 Elicitation of the reflex and

measurement of its amplitude have provided insights in spinal transmission during the









performance of a motor task. During walking, the modulation of reflex amplitude is a measure of

the regulation of afferent feedback during different phases of the step cycle.146,228,229,253 For

example, reflex amplitude increases to maximum during stance and decreases rapidly to

minimum during swing. This regulation is required to accommodate the functional requirements

of the task, i.e. facilitate weight support and plantarflexion during stance while allowing ankle

dorsiflexion during swing.146,228,229,253

In healthy, non-injured individuals, an increase in peroneal H-reflex amplitude has been

observed while wearing a brace in static sitting or standing position suggestive of a heightened

sensorimotor response due to stimulation of the cutaneous mechanoreceptors.154'205'255 Increasing

cutaneous input from the sole of the foot leads to a reduction in Ia presynaptic inhibition of the

soleus muscle. The predicted outcome of this is facilitation of soleus H-reflex amplitude.

Therefore, use of an orthotic device touching the plantar surface of the foot and limiting the

range of motion at the ankle could alter the rich sensory information processed from the ankle-

foot complex and potentially modulate reflex activity in non-injured individuals. However, due

to task-specific nature of H-reflex amplitude, it is difficult to extrapolate the results of a static

task to the dynamic task of walking.256 Examination of reflex amplitude while walking with a

PAFO will be useful in determining functional implication of the device in the task of walking.

The purpose of this study was to examine the phase dependent modulation of the H-reflex in the

gait cycle with and without a PAFO in non-injured individuals.

Specific Aims

Aim: To compare phase specific modulation of the soleus H-reflex amplitude in non-

injured individuals while walking with and without a PAFO,

Hypothesis: In non-injured individuals, soleus H-reflex amplitude while walking with an

AFO will be significantly larger compared to the H-reflex amplitude without an AFO









Methods


Subject Selection

A sample of convenience consisted of fourteen healthy, non-injured individuals living

independently in the Gainesville community. Each participant provided informed consent before

participating in the study. The University of Florida Institutional Review Board and the Veteran

Affairs Subcommittee had approved the study for clinical investigation. Age range for our

participants was between 18-60 years. Participant demographics are tabulated in Table 6-1. This

study did not include any subjects with any detectable gait and postural disorders. Subjects were

screened for a medical history of any neurological, musculoskeletal or orthopedic problem that

may affect their walking performance over the treadmill.

Using the effect size and standard deviation from pilot data for non-injured subjects (H/M

ratio post-pre =0.08, SD from previous study=0.11), to achieve statistical power of 80% at an

alpha level of 0.05 we needed 14 normal subjects.

Experimental Set-up

Soleus H-reflexes were evoked, for the purpose of consistency, on the dominant side of

healthy, non-injured individuals subjects. Skin was shaved and cleaned for application of

electrodes. A bipolar (2 cm inter-electrode distance) Ag-AgCl surface electrode (Therapeutics

Unlimited, Iowa City, Iowa) was placed longitudinally over the soleus muscle. These electrodes

are embedded in an epoxy mount with preamplifier circuitry and a 2-cm interelectrode distance.

The preamplifier and second-stage amplifier provide a total amplification of 1000x with a low-

frequency cut off of 20 Hz.

To evoke H-reflexes, one millisecond current pulses were delivered via a constant-current

stimulator (Grass Instruments, model S8800 with a modified CCU1) using a 2 cm 1/2 sphere

silver cathode placed in the popliteal fossa and a 10 cm silver anode positioned just superior to









the patella. The tibial nerve was localized, in the popliteal fossa by the electrode placement, to

evoke a soleus H-reflex at the least current intensity required. Data were acquired at a sample

rate of 10 kHz per channel and stored digitally with a commercially available data acquisition

system (Data-Pac III by Run Technologies) in a personal computer (Dell Systems, Intel

Celeron).

Subject Preparation

Subjects were asked to wear tennis shoes and change into appropriate clothing (dark

colored cycling shorts and shirt). Skin was shaved and cleaned for application of surface

electrodes. For trials using the PAFO, each subject was fitted with an off-the-shelf PAFO. The

leg chosen for donning the PAFO and the order of testing with and without it was randomized

for each subject. For the PAFO trial, each subject was requested to wear a unilateral, size-fitted

PAFO. The insole of the shoe was removed in order to fit the AFO and to even out the limb

length on both sides. Fitting of the AFO was assessed by measuring fit inside shoe, length of the

calf shell and that of the footplate. Standardized fitting included using an AFO whose length fits

an inch to two below the fibular head when donned and whose footplate length extends till the

tip of the toes.244 During treadmill walking, footswitches were placed inside the shoes that were

helpful in determining the phases of walking. Each subject was fitted with a body weight

supporting harness equipped with an additional overhead safety catch. The harness and safety

catch when used either with or without BWS provided safety to the person walking on the

treadmill and held or caught the person if he or she lost their balance, stumbled or began to fall.

Procedure

The order of testing was randomized for walking with and without PAFO (Figure 6-1).

Prior to eliciting reflexes during walking, H-reflexes were first elicited in static standing position

for use as a control reference across trials.146,257 For this purpose, participants were asked to









stand quietly and H-reflexes were collected in this position. Stimulus intensity was maximized

and three maximum M-waves were recorded in the static condition. At least fifteen H-reflexes

were then elicited at stimulus intensity within a range of 8-12 % of the M-max in the static

position. A recruitment curve was constructed in the static stance position to ensure that the H-

reflex was on the ascending limb of the curve.138,149

During walking trials in either of the conditions, H-reflexes were elicited across ten time

divided phases of the gait cycle determined by footswitches namely heel strike (HS), HS+100ms,

HS+200ms, HS+300ms, HS+400ms and toe-off (TO), TO+100ms, TO+200ms, TO+300ms,

TO+400ms. The event and time point of stimulation was achieved by connecting the

footswitches to a Schmidt trigger that sensed the event and delivered the pulse. For example, for

the phase of heel strike the pulse was delivered at Oms. For mid-stance, the pulse was delivered

after a time delay from heel strike.

Once the subject began stepping on the treadmill at 1.2 m/s, three maximum M-waves

were recorded in each of the ten phases of the gait cycle (Figure 6-1).227,257 These recordings

were used in determining the stimulus intensity for each tested phase in the gait cycle and were

also used for subsequent normalization of the data. Subsequently, stimulation was delivered at

stimulus intensity within a range of 8-12 % of the M-max calculated in each of the phases of the

gait cycle.138,149 At least fifteen stimuli were delivered in a consecutive or an alternating fashion

in each of the ten phases of the gait cycle.138'149

During testing, in both the static and walking condition, the activity in the soleus and TA

muscle was recorded over a 100 ms duration prior to electrical stimulation.149 This activity was

normalized to three maximum voluntary contractions of the TA and soleus collected at the

beginning of the experiment. Also, the M-wave was constantly monitored to make readjustments









to the stimulus intensity if required. A fifteen-minute sitting break was provided after completion

of static and walking trials in one condition before proceeding with the other.

Data Processing

After filtering and rectification of the data, mean peak-to-peak amplitude of 10 H-reflexes

for each phase of the examined gait cycle was calculated and compared between the two

conditions (with and without PAFO). Prior to this comparison, the H-reflex values were

normalized to M-max values procured in the respective phases (H/M ratio).149

Data Analysis

For the ten phases of the gait cycle, a Hotelling's T2-test, which is a multivariate analogue

of the paired t-test, was performed. The test is a multivariate extension of the Student's t-test for

paired data in comparison of mean difference vectors, i.e. the differences of two or more

dependent variables considered twice in the same subjects.245 The dependent variables compared

include the H/Mmax amplitude in thel0 phases of the gait cycle (HS, HS+100ms, HS+200ms,

HS+300ms, HS+400ms, TO, TO+100ms, TO+200ms, TO+300ms, TO+400ms). Significant

changes between the two conditions (with and without PAFO) were identified using the Holm's

correction which corrects for multiple comparisons by adjusting alpha value.246 Additionally, a

repeated measures ANOVA was also performed to compare M-max amplitude, actual M wave

amplitude used for stimulation of the H-reflex and electromyographic activity recorded 100ms

prior to stimulation in the TA and soleus muscles between the two conditions across the gait

cycle. Significance level was set at p< 0.05. The same analyses were repeated for the

contralateral limb.

Results

Both ipsilaterally and contralaterally the mean H/M ratios were not significantly different

between the two walking conditions (p>0.05) for any of the phases (Figures 6-2 through Figure









6-5). The mean EMG of soleus and tibialis anterior muscles 100 ms prior to the electrical

stimulation was not significantly different in both conditions (p<0.05). Additionally, M-max

amplitude (Figure 6-6) and the actual M wave amplitude used to evoke the soleus H-reflexes

(Figure 6-7) were not significantly different across the gait cycle in both the conditions.

Discussion

Peripheral input from the muscle and cutaneous receptors of the ankle-foot complex have

been known to modulate soleus H- reflex amplitude during different tasks. The soleus H-reflex is

facilitated by excitation of the plantar cutaneous afferents located around the heel.258 Like wise,

the change in ankle joint angle has been shown to modulate H-reflex excitability.259'260,154

Therefore, use of an orthotic device touching the plantar surface of the foot and limiting the

range of motion at the ankle could alter the rich sensory information processed from the ankle-

foot complex and potentially modulate reflex activity. Although we hypothesized that there

would be an increase in soleus H-reflex amplitude with an ankle brace the results of our current

study show that soleus H-reflex amplitude remains unchanged while walking with a PAFO in

healthy, non-injured individuals.

Previous studies have shown that ankle bracing impacts reflex amplitude in static tasks.

For example, Nishikawa et al (1999) reported a 10% increase in peroneus longus (PL) H-reflex

amplitude after application of a semi-rigid ankle support in the seated, non-weight bearing

position.205 The non-weight bearing position and the testing of the peroneus muscle may account

for the differing results between the Nishikawa study and this work. Likewise, Schneider et al

reported an increase in soleus H-reflex amplitude on passively imposing rapid knee flexion from

static stance position.154 Interestingly, not all studies done in a static task have demonstrated an

increase in H-reflex amplitude as a result of bracing the ankle joint. For example, Sefton et al

found no effect of a semi-rigid ankle brace on the PL H-reflex during an inversion









perturbation261 and no effect on soleus H-reflex during a single limb stance task.262 They

rationalized their findings to the fact that the external ankle support provided an increase in

mechanical stability, negating the need for neuronal adaptation to maintain upright stance.

The essential difference between our studies and previous studies was that our experiment

involved the dynamic task of walking with or without the brace rather than a static task. Since

the brace-related afferent input did not alter the lower limb reflexes during walking, it appears

that this reflex is centrally modulated in healthy, non-injured individuals. Therefore, even if a

peripheral influence is shown to have an effect in one task, it does not follow that the same input

will be effective in another task. Several studies in the literature have demonstrated more central

modulation of reflex activity during locomotion and locomotor like tasks. For example, Garrett

et al (1999) and Schneider et al (2000) reported that the soleus H-reflex amplitude did not change

when the knee was braced thereby blocking the normal excursion during locomotion.154,206

Similarly, Yang and Wheelan (1993) have shown that inactivity of the tibialis anterior or activity

of the soleus muscle during the swing phase of gait did not affect phase-specific modulation of

the soleus H-reflex during walking.254

In our study, a kinematically significant decrease in ankle plantarflexion and hip extension

were observed as a result of walking with the brace compared to walking without one (Refer

Figure 4-1). However, in non-injured individuals, kinematic changes during brace walking did

not change the modulation pattern through out the gait cycle suggestive of central modulation of

reflex activity. In non-injured subjects, such an occurence where changes in afferent input from

the periphery do not alter H-reflex excitability is probably for maintenance of the locomotor task.

In the event of impaired locomotor control as exists after SCI, the reflex modulation might

change during brace walking. This is because the ability of the spinal cord to modulate sensory









input and presynaptic inhibition are both altered post-SCI.263 Previous studies, have

systematically reported that greater H/M ratios were recorded in post-SCI subjects than recorded

in non-injured controls.263-265 For example, electrical excitation of the plantar cutaneous afferents

has been shown to facilitate soleus H-reflexes in persons with SCI and depress reflex amplitude

in non-injured subjects in sitting.266 Therefore, although, reflex modulation during walking did

not change between conditions in non-injured individuals, our results may not extend to

individuals with neurological impairment. However, this remains to be tested experimentally.

Limitations

As has been advocated, soleus H-reflex in our study was evaluated during walking rather

than at rest because the reflex undergoes task specific modulation. The factors that could be

potential confounds in the study such as speed of walking, testing order, background EMG

activity and stimulus intensity were controlled. However, extraneous peripheral afferent input

from other that sources such as stimulus generated perturbations or pain associated with repeated

stimulation could affect the measured H-reflex amplitude. Also, although a maximal M wave

was evoked in each of the tested phases for normalization purposes, a recruitment curve was not

constructed during the walking trials. A recruitment curve would assess reflex modulation over

the range of intensities during walking and also ensured evaluation of the same proportion of the

motor neuron pool. Although the testing phases were randomized, as a result of repeated

stimulation, post activation depression of the reflex could have occurred affecting the results of

the study. However, this is unlikely because studies have shown that synaptic transmission from

la fibers to motor neurons depends in a complex fashion on the rate of nerve impulses.267 During

movement tasks, stimuli that produce one or two extra impulses in a neuron that is already

conducting tens of impulses per second will not produce significant depression.256









Conclusion

In summary, non-injured individuals demonstrate central modulation of reflex activity which

attenuates extraneous sensory input from the periphery for the maintenance of the locomotor

task. A spinal injury disrupting supraspinal pathways can affect this phase specific reflex

modulation. In the presence of impaired central modulation, persistent cutaneous input that is

usually presynaptically inhibited during the gait cycle could be facilitated reinforcing the

walking related impairment. In the light of these findings, an ankle foot orthosis needs to be

evaluated systematically in individuals after SCI for the generation of reflex modulation

characteristic of normal walking.










N subjects


With AFO


Without AFO


Standing
H-reflexes


Oms 100ms!
0
I I


200m' 300m4 400mk Oms
50
I I I


100Oms I 200ms I 300ms I

I I I I


ns


400ms
Time
100 ~1% of gait cycle


Walking on treadmill at 1.2 m/s

Figure 6-1. Experimental design for the examination of changes in soleus H-reflex amplitude in
healthy, non-injured individuals while walking with and without an ankle foot
orthosis (AFO).












Without AFO at
HS+300ms


SA1


With AFO at
HS+300ms


Without AFO at
TO+300ms


With AFO at
TO+300ms


Figure 6-2. Ipsilateral raw soleus H-reflex data while walking with and without AFO at 300ms
from heel strike (HS) and toe-off (TO). Vertical red and blue guide bars capture
soleus H-reflex event.


V-- -- ^ ----..-.- I


Without AFO at
HS+300ms


With AFO at
HS+300ms


Without AFO at
TO+300ms


With AFO at
TO+300ms


Figure 6-3. Contralateral raw soleus H-reflex data while walking with and without AFO at
300ms from heel strike (HS) and toe-off (TO). Vertical red and blue guide bars
capture soleus H-reflex event.


L
I


,I











100
90
80
70
60
50 -
40
30
20
10 0
0


-AFO
-no AFO
-Standing


HS 100 200 300 400 TO 100 200 300 400

Gait cycle (milliseconds)



Figure 6-4. Ipsilateral mean H-reflex amplitudes with and without AFO normalized to M-max in
each phase of the gait cycle. The gait cycle is represented in 100ms increments from
heel strike (HS) and toe off (TO).


100
90
80
70


-AFO
-no AFO
-Standing


HS 100 200 300 400 TO 100 200 300 400
Gait cycle (milliseconds)


Figure 6-5. Contralateral mean H-reflex amplitudes with and without AFO normalized to M-
max in each phase of the gait cycle. The gait cycle is represented in 100ms
increments from heel strike (HS) and toe off (TO).














7.00

6.00

5.00

E 4.00

E 3.00

2.00

1.00

0.00

x x x x ,0 x xx xx

Stimulation phases of the gait cycle


B
9.00


8.00

7.00

w 6.00

S5.00
E
S4.00
E
| 3.00

2.00

1.00

0.00


m Without AFO
m With AFO


m Without AFO
m With AFO


Stimulation phases of the gait cycle


Figure 6-6. Ipsilateral [A] and contralateral [B] M-max amplitude with and without the AFO
across the gait cycle. The gait cycle is represented in 100ms increments from heel
strike (HS) and toe off (TO).




















m Without AFO
m With AFO


ix pa ex o x t x g t x cx

Stimulation phases of the gait cycle


m Without AFO
m With AFO


X X X Xg

Stimulation phases of the gait cycle


Figure 6-7. Ipsilateral [A] and contralateral [B] actual M wave amplitude used to evoke the
soleus H-reflex with and without the AFO across the gait cycle. The gait cycle is
represented in 100ms increments from heel strike (HS) and toe off (TO).


1.00

0.90

0.80

0.70

0.60

0.50

0.40

0.30

0.20

0.10

0.00









Table 6-1. Demographics of non-injured participants recruited for the study.
Subject Orthotic
Age Sex .
ID side
N1 24 F R
N2 27 M R
N3 30 M R
N4 28 F R
N5 26 F R
N6 31 F R
N7 27 M R
N8 24 F R
N9 22 M R
N10 27 F L
N11 23 M L
N12 29 M L
N13 23 M L
N14 36 F L









CHAPTER 7
IMMEDIATE, PHASE DEPENDENT, SOLEUS H-REFLEX MODULATION IN PERSONS
WITH INCOMPLETE SPINAL CORD INJURY WHILE WALKING WITH AN ANKLE
FOOT ORTHOSIS

Introduction

The soleus H-reflex has often been employed to examine the neural regulation of afferent

information during walking.228,253 Elicitation of the reflex and measurement of its amplitude have

provided a method to evaluate the modulation of the spinal pathways during walking.138 In

healthy, non-injured individuals, the soleus H-reflex undergoes phase-specific modulation to

accommodate to the functional requirements of the task.146,228,229 Specifically, the size of the

soleus H-reflex is higher during the stance phase and lower during the swing phase of the step

cycle. However, in individuals after ISCI, the soleus H-reflex demonstrates decreased depth of

modulation. As a consequence, there is a reduced amplitude modulation across the step cycle and

it simply remains increased throughout the step cycle.155,156,268 Consequently, lack of reflex

modulation contributes to their walking impairments.155,156,268

During conventional gait training, individuals with ISCI are often prescribed orthotic

devices to assist with walking. Orthotic devices such as an ankle foot orthosis (AFO) are used for

assisting foot clearance, increasing gait speed, and improving walking endurance.16,269 Apart

from mechanically aiding foot clearance,270 studies in healthy, non-injured individuals suggest

that an AFO increases afferent feedback from cutaneous receptors in the foot and shank to

improve ankle positioning.255,271 For example, ankle bracing in a variety of static motor tasks

such as sitting or standing have reported an increase in peroneal205'255 and soleus H-reflex

amplitude in healthy, non-injured individuals.154 However, the increase in reflex amplitude with

the brace has only been documented in healthy, non-injured individuals under static conditions

limiting the extrapolation of results to persons with ISCI in dynamic tasks. In individuals with









ISCI with impaired reflex modulation, an orthotic device touching the plantar surface of the foot

and limiting the range of motion at the ankle could alter the rich sensory information processed

from the ankle-foot complex and potentially modulate reflex activity.

Additionally, in persons with ISCI, cutaneous stimulation of the foot and sole while

walking has been suggested as a potential method to restore reflex modulation comparable to that

seen in healthy, non-injured individuals.155 Furthermore, simulation of walking kinematics using

manual assistance, bodyweight support and a treadmill in persons with ISCI improved reflex

modulation and overground stepping speed without bracing the ankle. 181 Consequently,

conclusive evidence supporting or not supporting the use of the brace in improving phase-

specific modulation during gait training in individuals with ISCI is not apparent. Also, since

afferent stimulation of one limb is known exert a considerable influence on the reflex activity of

the contralateral limb, the influence of an AFO on contralateral soleus H-reflex modulation also

warrants investigation.153

Therefore, the immediate phase-dependent modulation of the soleus H-reflex in persons

with ISCI during walking with and without an AFO was examined ipsilaterally and

contralaterally. Specifically, we hypothesized an increase in reflex amplitude only ipsilaterally

while walking with the AFO compared to walking without one in individuals with ISCI.

Specific Aims

* Aim 1: In persons with ISCI and ambulatory, to compare immediate phase-dependent
modulation of the soleus H-reflex with and without an AFO in the mid-stance phase of
walking.

* Hypothesis 1: In persons with ISCI, soleus H-reflex amplitude will be significantly larger in
mid-stance while walking with an AFO compared to walking without an AFO.

* Aim 2: In persons with ISCI and ambulatory, to compare immediate phase-dependent
modulation of the soleus H-reflex in with and without an AFO in the mid-swing phase of
walking.









*Hypothesis 2: In persons with ISCI, soleus H-reflex amplitude PAFO will be significantly
larger in mid-swing while walking with an AFO compared to walking without an AFO.

Methods

Subject Selection

Nine persons with ISCI ranging between 18-80 years were recruited for this experiment

and sign an informed consent form approved by University of Florida Institutional Review Board

and the Veteran Affairs Subcommittee for Clinical Investigation (Table7-1). American Spinal

Injury Association (ASIA) motor score and impairment scale data were collected from all

participants with ISCI to assess the degree of impairment in each leg.

The criteria for inclusion in the study were: Medically stable persons with ISCI classified

as ASIA D, having quadriceps strength of at least 3/5, having decreased ankle strength and/or

impaired or absent proprioception at the ankle, having ankle dorsiflexor strength of < 4/5, able to

stand unaided for one minute and walking with minimal assistive devices but not using an AFO.

Exclusion criteria include persons who are unable to follow 3 step commands, amputation,

medical instability, significant musculoskeletal problems other than SCI that limit hip and knee

extension or ankle plantarflexion to neutral.

Experimental Set-up

Soleus H-reflexes were evoked, for the purpose of consistency, on the dominant side of

healthy, non-injured individuals subjects. Skin was shaved and cleaned for application of

electrodes. A bipolar (2 cm inter-electrode distance) Ag-AgCl surface electrode (Therapeutics

Unlimited, Iowa City, Iowa) was placed longitudinally over the soleus muscle. These electrodes

are embedded in an epoxy mount with preamplifier circuitry and a 2-cm interelectrode distance.

The preamplifier and second-stage amplifier provide a total amplification of 1000x with a low-

frequency cut off of 20 Hz.









To evoke H-reflexes, one millisecond current pulses were delivered via a constant-current

stimulator (Grass Instruments, model S8800 with a modified CCU1) using a 2 cm 1/2 sphere

silver cathode placed in the popliteal fossa and a 10 cm silver anode positioned just superior to

the patella. The tibial nerve was localized, in the popliteal fossa by the electrode placement, to

evoke a soleus H-reflex at the least current intensity required. Data were acquired at a sample

rate of 10 kHz per channel and stored digitally with a commercially available data acquisition

system (Data-Pac III by Run Technologies) in a personal computer (Dell Systems, Intel

Celeron).

Subject Preparation

Subjects were asked to wear tennis shoes and change into appropriate clothing (dark

colored cycling shorts and shirt). Skin was shaved and cleaned for application of surface

electrodes. For trials using the PAFO, each subject was fitted with an off-the-shelf PAFO. The

leg chosen for donning the PAFO and the order of testing with and without it was randomized

for each subject. For the PAFO trial, each subject was requested to wear a unilateral, size-fitted

PAFO. The insole of the shoe was removed in order to fit the AFO and to even out the limb

length on both sides. Fitting of the AFO was assessed by measuring fit inside shoe, length of the

calf shell and that of the footplate. Standardized fitting included using an AFO whose length fits

an inch to two below the fibular head when donned and whose footplate length extends till the

tip of the toes.244 During treadmill walking, footswitches were placed inside the shoes that were

helpful in determining the phases of walking. Each subject was fitted with a body weight

supporting harness equipped with an additional overhead safety catch. The harness and safety

catch when used either with or without BWS provided safety to the person walking on the

treadmill and held or caught the person if he or she lost their balance, stumbled or began to fall.









Procedure

H-reflexes were elicited in two testing conditions (with and without PAFO) and two parts

of gait cycle (mid-stance and mid-swing) during walking. The order of testing was randomized

for with and without PAFO conditions (Figure 7-1). Treadmill speed was maintained constant for

the both the testing conditions. H-reflexes were first be elicited in static standing position.

Collecting, H-reflexes in static position served as a control reference across trials since the reflex

is not modulated in a static position.146,257 For this purpose, participants were asked to stand

quietly and H-reflexes were collected in this position. Stimulus intensity was maximized and

three maximum M-waves were recorded in the static standing condition. Fifteen H-reflexes were

then elicited at stimulus intensity within a range of 8-12 % of the M-max in the static position. A

recruitment curve was constructed in the static stance position to ensure that the H-reflex was on

the ascending limb of the curve.138,149

Once the subject began stepping on the treadmill at self- selected speed, three maximum

M-waves were recorded in mid-stance phase and mid-swing phase respectively.227'257 These

recordings were used to determine the stimulus intensity for each tested phase in the gait cycle

and were also useful for subsequent normalization of the data.138,149 Subsequently, stimulation

was delivered at stimulus intensity within a range of 8-12 % of the M-max in the mid-stance and

mid-swing phase of the gait cycle.138'149 At least 15 H-reflexes were recorded in each of the two

selected phases of the gait cycle at each of the time points. 138,149

During testing, in both the static and walking condition, the activity in the soleus and TA

muscle was recorded over duration of 100 ms prior to electrical stimulation. This activity was

normalized to the average of three maximum voluntary contractions of the TA and soleus

collected at the start of the experiment. Also, the M-wave was constantly monitored to make

readjustments to the stimulus intensity if required. Duration between two consecutive electrical









stimulations were randomly maintained at a minimum of three seconds and maximum 5 seconds

to avoid post activation depression and habituation. 138,149 After recording the H-reflexes in

walking, the procedure for the static position was repeated again. After data collection, the data

was stored for subsequent analysis. A fifteen-minute sitting break was provided after completion

of static and walking testing in one condition before proceeding with the other.

As part of a secondary question examining the effect of the ipsilateral brace on

contralateral limb the same procedure was repeated on the contralateral side. Only eight of the

above participants participated in this part of the study.

Data Processing

After filtering and rectification of the data, mean peak-to-peak amplitude of 10 H-reflexes

for each phase of the examined gait cycle was calculated and compared between the two

conditions (with and without PAFO). Prior to this comparison, the H-reflex values were

normalized to M-max values procured in the respective phases (H/M ratio).149

Data Analysis

For the mid-stance and mid-swing phase on the ipsilateral limb, a paired t-test with bracing

condition (with or without PAFO) as the independent variable and the H-reflex amplitude as the

dependent variable was performed. The same analysis was repeated for the contralateral limb.

Significant changes between the two conditions (with and without PAFO) were identified using

the Holm's step-down correction which corrects for multiple comparisons by adjusting alpha

value.246 Additionally, a repeated measures ANOVA was also performed to compare M-max

amplitude, actual M wave amplitude used for stimulation of the H-reflex and electromyographic

activity recorded 100ms prior to stimulation in the TA and soleus muscles between the two

conditions across the gait cycle. Significance level was set at p<0.05.









Results

Figure 7-2 through Figure 7-4 shows the ipsilateral and contralateral H-reflex modulation

in mid-stance and mid-swing phase with and without the PAFO ipsilaterally. After correcting for

multiple comparisons, the mean H/M ratio was significant importantly for the ipsilateral mid-

swing phase of walking (without PAFO: 0.130.10 & with PAFO 0.290.14).The values for the

H/M ratio for all the tested conditions are reported in Table 7-2 and Table 7-3. The mean EMG

of soleus and tibialis anterior muscles 100 ms prior to the electrical stimulation did not change

systematically in both conditions (Tables 7-4 through Table 7-7). Furthermore the M-max

amplitude and the actual Mwave amplitude used for evoking the soleus H-reflex did not change

significantly (Figure 7-5 and Figure 7-6).

Discussion

This is the first study to systematically examine the effect of bracing on soleus H-reflex

modulation during the task of walking in individuals with ISCI. The main finding of the study

was that, there was a significant increase in soleus H-reflex amplitude ipsilaterally in the mid-

swing phase while walking with a PAFO. In the absence of a change in the ankle-foot orientation

or stretch at the ankle joint with a PAFO, these results are suggestive of an increase in afferent

inflow in the mid-swing phase of walking.

In our study, the background EMG activity in both the walking conditions (with and

without the PAFO) was similar despite the changes in H-reflex amplitude reinforcing the fact

that the modulation of the reflex is not directly dependent on the excitation level of the alpha-

motorneurons.146 Likewise the presence or absence of ankle clonus did not affect reflex

amplitude in the two walking conditions. Subject 19 who had clonus had a change in mid-swing

reflex amplitude similar to subject 14 who did not demonstrate any clonic activity in the soleus

muscle.









In individuals with ISCI, an increase in reflex amplitude with the PAFO during the mid-

swing phase could be the result of altered presynaptic inhibition. Presynaptic inhibition is

important because during walking sensory input from cutaneous and proprioceptive receptors

continuously converges on the spinal circuits.263'272 These surplus inputs must either be

synergistically combined with the motor commands or be appropriately suppressed to minimize

interference. Since only certain input requires selective modulation, presynaptic inhibition of

sensory input allows suppression of specific inputs to a neuron without influencing other

synaptic inputs. The predominant sources of presynaptic inhibition are peripheral inputs from

cutaneous afferents and central descending pathways.272,273 During walking, cutaneous input

from the foot sole is known to modulate reflex activity and change muscle synergies thereby

contributing to adaptive locomotor strategies. For example, Bastiaanse et al (2000) observed that

load receptors are involved in the regulation of cutaneous reflex responses in order to adapt the

locomotor pattern to the environmental conditions.274

The repertoire of adaptive movement strategies is usually limited in individuals with ISCI

because the ability of the spinal cord to modulate sensory input and presynaptic inhibition are

usually impaired post-injury.263,275 Previous studies, have systematically reported that greater

H/M ratios were recorded in post-SCI subjects than recorded in non-injured controls. 156,263-265

The increase in reflex amplitude during brace walking in our study may have occurred because

of the stimulation of plantar cutaneous afferents caused by the PAFO. Excitation of the plantar

cutaneous afferents facilitates soleus H-reflex in persons with SCI in sitting.266 In the absence of

supraspinal modulation of reflex activity, peripheral cutaneous inputs may be beneficial to

modulate reflex activity.155 Nakajima et al (2006) has shown that reflex connections from

cutaneous nerves in the foot on to the lower limb muscles are arranged in a highly topographical









manner and may play an important role in limb loading and ground contact in response to tactile

sensation.276 However, from the results of our study two interesting observations can be made

about the PAFO during walking.

An increase in reflex amplitude observed in the mid-swing phase during walking with a

PAFO suggests that a simple, off-the-shelf orthosis could potentially increase afferent inflow and

modulate reflex activity in high functioning individuals with ISCI. However, in individuals with

shallow modulation of reflex activity throughout the gait cycle, the increase in afferent inflow

especially during the mid-swing phase might be unfavorable for reflex modulation and ultimately

walking.156 Therefore, prior to its use, the type and purpose of orthotic device, the movement

task of interest such as walking or cycling and the targeted population needs to be assessed

carefully.

Interestingly, no increase in reflex amplitude was noted with the brace in the contralateral

limb in mid-stance or mid-swing. Also motion data collected previously did not show a change

in limb kinematics with or without the PAFO contralaterally. Our inference further strengthens

the idea of a localized cutaneous response of the PAFO on the reflex modulation ipsilaterally.

Clinical Implications

Soleus H-reflexes are exaggerated post-SCI. If reflex dysregulation is secondary to

disruption of supraspinal inhibitory control mechanisms, then training strategies inhibiting the

hyperactive reflex segmentally in a task and phase specific manner may be beneficial for the

proper restoration of locomotion. Locomotor training is one such strategy that works on the

above principle of providing optimal sensory input to the nervous system to recover walking

ability after ISCI. The training provides sensory cues and phasic information related to

locomotion. One of the cues pivotal to training is to minimize sensory stimulation that would

conflict with sensory information associated with locomotion.155,277,278 The potency of the









training stems from the fact that a single bout of locomotor training is capable of producing

significant depression of the exaggerated soleus H-reflexes and improved walking speed in

persons with ISCI.181 Therefore, research efforts are being directed at systematically determining

the critical components of locomotor training, that optimize the provision of critical sensory

input that aid walking recovery. Within the realms of this goal, our study demonstrates that the

integration of clinically acknowledged stepping aids such as the PAFO's during locomotor

training could be counter productive to the recovery of walking post SCI. Therefore such devices

should be chosen only after careful consideration of outcome for training purposes.

Limitations

First, we only examined the effect of the PAFO in the mid-stance and mid-swing phase of

walking limiting our inferences to only two specific phases of walking. A thorough examination

in different phases of the gait cycle might reveal the unique effects of the PAFO within the entire

gait cycle. Second, for standardization purposes, we examined the effect of only one type of

AFO which limits generalizability to other types of AFO's. Based on their impairments

individuals with ISCI might use customized AFOs which could yield different results.

Conclusions

In persons with ISCI, soleus H-reflex amplitude increased significantly in the mid-swing

phase of walking with an AFO compared to walking without it. Our findings suggest that, in the

presence of impaired central and peripheral modulation of reflex activity, an ankle foot orthosis

that provides persistent cutaneous inputs from the foot sole might contribute to modulating reflex

activity. However, increase in afferent input in certain phases of the gait cycle might not always

be favorable to the task of walking. Therefore therapeutic interventions targeted at promoting

walking recovery in individuals with ISCI should carefully consider the use of such non-

adaptable, compensatory orthotic devices that could potentially hinder retraining or reeducation









of function. Instead, adaptable strategies that have been documented to provide appropriate phase

specific sensory input such as functional cutaneous stimulation of the foot or appropriate cueing

using manual assistance should be incorporated during training to assist with walking and

promote recovery.













N subjects


With AFO


Mid-stance


H-reflex in
Testing order walking at
rTaninored Standing H-reflex walking at
randomized SS speed

Without AFO Mid-swing


Figure 7-1. Experimental design for testing the effect of walking with and without an ankle foot
orthosis (AFO) in individuals with incomplete spinal cord injury (ISCI) at their self-
selected (SS) walking speed.

















m snoAFO
m AFO
---standing


Phase of stimulation in the gait cycle


Figure 7-2. Average H/M ratio values with and without an AFO in mid-stance and mid-swing
phase of walking relative to static standing in the ipsilateral limb.




100
90
80
70
60 noAFO
E 50 AFO
40 -rn-standing


30
20
10
0


Midstance


Midswing


Phase of stimulation in the gait cycle


Figure 7-3. Average H/M ratio values with and without AFO in mid-stance and mid-swing
phase of walking relative to static standing in the contralateral limb.


100
90
80
70
60-
60
40-
50
20-
40
30
20
10
0


Midstance


Midswing










Primary side


MSt without AFO


MSt with AFO


MSw without AFO MSw with AFO


Contralateral side


MSt without AFO


MSt with AFO


MSw without AFO
MSw without AFO


MSw with AFO


Figure 7-4. Ipsilateral and contralateral raw soleus H-reflex data while walking with and without
AFO during mid-stance (MSt) and mid-swing (MSw) phase of the gait cycle. Vertical
red and blue guide bars capture soleus H-reflex event.























* No AFO
EAFO


Mid-stance Mid-swing

Phase of stimulation in the gait cycle


* No AFO
SAFO


Mid-stance Mid-swing

Phase of stimulation in the gait cycle


Figure 7-5. Ipsilateral [A] and contralateral [B] M-max amplitude with and without the AFO in
the mid-stance and mid-swing phase of walking.


8


7 -


6 -


5


4 -


3


2 -


1-


0
































Mid-stance


Mid-swing


Phase of stimulation in the gait cycle


Mid-stance


Mid-swing


Phase of stimulation in the gait cycle


Figure 7-6. Ipsilateral [A] and contralateral [B] actual M wave amplitude used to evoke the
soleus H-reflex with and without the AFO across the gait cycle.


0
z 0.7

x 0.6
E
a 0.5
4-
o 0.4

. 0.3
Co
0.2

0.1

0.0


* No AFO
NAFO


Kl


01
= 0.7

x 0.6
E
a 0.5
-0.4
Ci
, 0.3
CO
0.2

0.1

0


* No AFO
mAFO









Table 7-1. Participant demographics.


Sex Height


Orthotic Injury
level


5'11"
6' 3"
5'5"
6'0"
5'10"
5'8"
5"

5'11"
6'2"


C6-7
C7
C4-5
C5-6
C7
C2-T1
C6

T4-5
C5


ASIA Injury
duration
score
sc in months
D 14
D 79
D 46
D 58
D 23
D 253
D 180


Speed
(m/sec)

0.6
0.5
0.7
0.4
0.7
0.5
0.3


Assitive/Orthotic
device

Cane on left
Cane on right
NA
Cane on left
NA
NA
NA
Solid AFO on
right
Cane on right


ASIA: American Spinal Injury Association


Table 7-2. Ipsilateral H/M ratio with and without the AFO.
No AFO AFO
Subject
Mid- Mid-
ID Standing stance Mid-swing Standing stance Mid-swing
12 0.71 0.59 0.28 0.59 0.58 0.31
13 0.35 0.20 0.09 0.40 0.40 0.16
14 0.69 0.53 0.02 0.68 0.68 0.39
15 0.68 0.51 0.24 0.78 0.59 0.40
16 0.05 0.04 0.04 0.03 0.07 0.05
17 0.08 0.10 0.04 0.04 0.21 0.19
18 0.79 0.66 0.20 0.77 0.70 0.41
19 0.70 0.58 0.21 0.68 0.60 0.49
110 0.36 0.48 0.10 0.41 0.50 0.23
Avg 0.49 0.41 0.13 0.49 0.48 0.29
Std.Dev 0.29 0.23 0.10 0.29 0.22 0.14


ID Age


19 25
110 57









Table 7-3. Contralateral H/M ratio with and without the AFO.
No AFO


Mid-stance
0.45
0.45
0.66
0.19
0.19
0.10
0.13
0.71
0.36
0.24


Mid-swing
0.04
0.16
0.05
0.02
0.03
0.01
0.04
0.06
0.05
0.05


Standing
0.26
0.43
0.55
0.16
0.48
0.06
0.23
0.60
0.35
0.20


AFO


Mid-stance
0.41
0.50
0.65
0.23
0.51
0.03
0.14
0.68
0.39
0.24


Mid-swing
0.02
0.17
0.02
0.02
0.04
0.01
0.03
0.08
0.05
0.05


Normalized soleus EMG amplitude with and without AFO ipsilaterally.


with
AFO
0.16
0.05
0.13
0.12
1.65
4.73
2.17
2.66
NT


Midstance Midswing
without with without with
AFO AFO AFO AFO
0.22 0.13 0.23 0.16
0.05 0.05 0.05 0.05
0.15 0.20 0.16 0.17
0.16 0.14 0.14 0.14
0.28 0.31 1.25 1.22
4.68 4.73 4.69 4.73
2.26 2.19 2.18 2.17
2.65 2.66 2.66 2.67
NT NT NT NT


Table 7-5. Normalized TA EMG amplitude with and without AFO ipsilaterally.


Midswing
without with
AFO AFO
0.21 0.17
0.12 0.10
0.97 0.96
0.42 0.44
0.02 0.02
0.60 0.55
0.46 0.44
0.82 0.90
NT NT


Standing
0.31
0.40
0.51
0.22
0.44
0.06
0.36
0.70
0.38
0.19


Subject
ID
12
13
14
15
16
17
18
19
Avg
Std.dev


Table 7-4.
Soleus


Subject
ID
12
13
14
15
16
17
18
19
110


Standing
without
AFO
0.12
0.04
0.14
0.13
1.60
4.68
2.17
2.64
NT


Standing


Midstance


Subject
ID
12
13
14
15
16
17
18
19
110


without
AFO
0.14
0.10
0.96
0.30
0.02
0.73
0.46
0.56
NT


with
AFO
0.12
0.12
0.96
0.40
0.01
0.62
0.43
0.74
NT


without
AFO
0.14
0.11
0.96
0.38
0.02
0.48
0.42
0.79
NT


with
AFO
0.13
0.10
0.95
0.41
0.03
0.57
0.44
0.89
NT









Table 7-6. Normalized soleus EMG amplitude with and without AFO contralaterally.


Soleus
Standing Midstan
Subject without with without


AFO
0.14
0.09
0.12
0.12
0.10
3.33
1.60
1.37
NT


AFO
0.16
0.09
0.08
0.13
0.12
3.33
1.59
1.37
NT


AFO
0.16
0.09
0.19
0.15
0.11
3.33
1.61
1.37
NT


ce Midswing
with without with
AFO AFO AFO
0.18 0.15 0.16
0.09 0.08 0.08
0.09 0.26 0.37
0.15 0.10 0.12
0.12 0.33 0.56
3.34 3.56 3.39
1.61 1.59 1.62
1.37 1.37 1.37
NT NT NT


Table 7-7. Normalized TA EMG amplitude with and without AFO contralaterally.
TA


g Midstance
with without w


AFO
0.18
0.43
0.47
0.56
0.43
1.22
0.44
0.67
NT


Midswing
ith without with
FO AFO AFO
18 0.22 0.17
39 0.46 0.44
47 0.47 0.47
56 0.54 0.58
44 0.42 0.44
25 1.34 1.26
42 0.46 0.43
67 0.59 0.61
[T NT NT


Standin


Subject
ID
12
13
14
15
16
17
18
19
110


without
AFO
0.15
0.48
0.47
0.57
0.44
1.25
0.54
0.68
NT


AFO
0.17
0.45
0.46
0.57
0.44
1.24
0.48
0.61
NT









LIST OF REFERENCES


1. Forssberg, H. in Brainstem control of spinal mechanisms (eds. Sjolund, B. & Bjorklund,
R. A.) 253-71 (Elsevier Biomedical, Amsterdam, 1982).

2. Patla, A. E., Prentice, S., Robinson, C. & Neufeld, J. Visual control of locomotion:
strategies for changing direction and for going over obstacles. JExp Psychol Hum
Percept Perform 17, 603-34 (1991b).

3. Barbeau, H. Locomotor training in neurorehabilitation: emerging rehabilitation concepts.
Neurorehabil Neural Repair 17, 3-11. (2003).

4. Bell, D. T. Principles and techniques of physiotherapy in the treatment of spinal
paraplegia. Phys Ther Rev 35, 235-7 (1955).

5. van der Salm, A. et al. Gait impairments in a group of patients with incomplete spinal
cord injury and their relevance regarding therapeutic approaches using functional
electrical stimulation. ArtifOrgans 29, 8-14 (2005).

6. Barbeau, H., Ladouceur, M., Norman, K. E., Pepin, A. & Leroux, A. Walking after spinal
cord injury: evaluation, treatment, and functional recovery. Arch Phys MedRehabil 80,
225-35 (1999).

7. Sekhon, L. H. & Fehlings, M. G. Epidemiology, demographics, and pathophysiology of
acute spinal cord injury. Spine 26, S2-12 (2001).

8. Waters, R. L., Adkins, R. H., Yakura, J. S. & Sie, I. Motor and sensory recovery
following incomplete tetraplegia. Arch PhysMedRehabil 75, 306-11 (1994).

9. Waters, R. L., Adkins, R. H., Yakura, J. S. & Sie, I. Motor and sensory recovery
following incomplete paraplegia. Arch Phys Med Rehabil 75, 67-72 (1994).

10. Raineteau, O. & Schwab, M. E. Plasticity of motor systems after incomplete spinal cord
injury. Nat Rev Neurosci 2, 263-73 (2001).

11. Craik, R. L. in Contemporary management of motor problems (ed. Liser, M. J.) 165-74
(Bookcrafters, Inc., Fredericksburg, VA, 1991).

12. Gordon, J. in Movement science foundations for physical therapy in rehabilitation (eds.
Carr, J. H. & Shepherd, R. B.) 1-30 (Aspen Publishers Inc, Rockville, 1987).

13. Basso, D., Behrman, AL, Harkema, SJ. Recovery of walking after central nervous system
insult: basic research in the controlled locomotion as a foundation for developing
rehabilitation strategies. Neurology Report 24, 47-54 (2000).

14. Somers, M. Spinal cord injury: functional rehabilitation (Prentice-Hall Inc., London,
2001).









15. 0' Sullivan, S. B. & Schmitz, T. J. Physical rehabilitation assessment and treatment
(eds. O' Sullivan, S. B. & Schmitz, T. J.) (F.A. Davis Company, Philadelphia, 2000).

16. Atrice, M. B. Lower extremity orthotic management for the spinal-cord-injured client.
Spinal Cord Inj Rehabil 5, 1-10 (2000).

17. Lehmann, J. F. Biomechanics of ankle-foot orthoses: prescription and design. Arch Phys
MedRehabil 60, 200-7 (1979).

18. Barbeau, H. & Rossignol, S. Recovery of locomotion after chronic spinalization in the
adult cat. Brain Res 412, 84-95 (1987).

19. Barbeau, H. & Blunt, R. A novel interactive locomotor approach using body weight
support to retrain gait in spastic paretic subjects (ed. Wernig, A.) (Elsevier Science
Publishers, 1991).

20. Barbeau, H., Danakas, M. & Arsenault, B. The effects of locomotor training in spinal
cord injured subjects: a preliminary study. Restor Neurol Neurosci 5, 81-84 (1993).

21. Dobkin, B. H. Neuroplasticity. Key to recovery after central nervous system injury. West
JMed 159, 56-60 (1993).

22. Edgerton, V. R. et al. Use-dependent plasticity in spinal stepping and standing. Adv
Neurol 72, 233-47 (1997).

23. Basso, M. Neuroplasticity of descending and segmental systems after spinal cord
contusion. neurology Report 22, 48-53 (1998).

24. Behrman, A. L., Bowden, M. G. & Nair, P. M. Neuroplasticity after spinal cord injury
and training: an emerging paradigm shift in rehabilitation and walking recovery. Phys
Ther 86, 1406-25 (2006).

25. Behrman, A. L. & Harkema, S. J. Locomotor training after human spinal cord injury: a
series of case studies. Phys Ther 80, 688-700. (2000).

26. Behrman, A. L. et al. Locomotor training progression and outcomes after incomplete
spinal cord injury. Phys Ther 85, 1356-71 (2005).

27. Barbeau, H. et al. Tapping into spinal circuits to restore motor function. Brain Res Brain
Res Rev 30, 27-51 (1999).

28. De Leon, R. D., Hodgson, J. A., Roy, R. R. & Edgerton, V. R. Full weight-bearing
hindlimb standing following stand training in the adult spinal cat. JNeurophysiol 80, 83-
91(1998).

29. Edgerton, V. R. et al. Potential of adult mammalian lumbosacral spinal cord to execute
and acquire improved locomotion in the absence of supraspinal input. JNeurotrauma 9
Supply 1, S119-28 (1992).









30. Zehr, E. P. Neural control of rhythmic human movement: the common core hypothesis.
Exerc Sport Sci Rev 33, 54-60 (2005).

31. Van de Crommert, H. W., Mulder, T. & Duysens, J. Neural control of locomotion:
sensory control of the central pattern generator and its relation to treadmill training. Gait
Posture 7, 251-63 (1998).

32. Dietz, V. Human neuronal control of automatic functional movements: interaction
between central programs and afferent input. Physiol Rev 72, 33-69 (1992).

33. Perry, J. Gait analysis: normal and pathological function (Slack, Thorofare,NJ, 1992).

34. Duysens, J., Van de Crommert, H. W., Smits-Engelsman, B. C. & Van der Helm, F. C. A
walking robot called human: lessons to be learned from neural control of locomotion. J
Biomech 35, 447-53 (2002).

35. Duysens, J., Clarac, F. & Cruse, H. Load-regulating mechanisms in gait and posture:
comparative aspects. Physiol Rev 80, 83-133 (2000).

36. Duysens, J. Reflex control of locomotion as revealed by stimulation of cutaneous
afferents in spontaneously walking premammillary cats. JNeurophysiol 40, 737-51
(1977).

37. Edgerton, V. R., Tillakaratne, N. J., Bigbee, A. J., de Leon, R. D. & Roy, R. R. Plasticity
of the spinal neural circuitry after injury. Annu Rev Neurosci 27, 145-67 (2004).

38. Pang, M. Y. & Yang, J. F. The initiation of the swing phase in human infant stepping:
importance of hip position and leg loading. JPhysiol 528 Pt 2, 389-404 (2000).

39. Dietz, V. Locomotor training in paraplegic patients. Ann Neurol 38, 965 (1995).

40. Barbeau, H., Fung, J., Leroux, A. & Ladouceur, M. A review of the adaptability and
recovery of locomotion after spinal cord injury. Prog Brain Res 137, 9-25 (2002).

41. Wernig, A., Muller, S., Nanassy, A. & Cagol, E. Laufband therapy based on 'rules of
spinal locomotion' is effective in spinal cord injured persons. Eur JNeurosci 7, 823-9
(1995).

42. Wernig, A. & Muller, S. Laufband locomotion with body weight support improved
walking in persons with severe spinal cord injuries. Paraplegia 30, 229-38 (1992).

43. Barbeau, H., Wainberg, M. & Finch, L. Description and application of a system for
locomotor rehabilitation. MedBiolEng Comput 25, 341-4 (1987).

44. Harkema, S. J. Neural plasticity after human spinal cord injury: application of locomotor
training to the rehabilitation of walking. Neuroscientist 7, 455-68 (2001).









45. Chen, Y. et al. The interaction of a new motor skill and an old one: H-reflex conditioning
and locomotion in rats. JNeurosci 25, 6898-906 (2005).

46. Hidler, J. M. What is next for locomotor-based studies? Journal ofRehabilitaion
Research and Development 42, 10-6. (2005).

47. Stover, S. L., DeLisa, J. A. & Whiteneck, G. G. Spinal cord injury: clinical outcomes
from the model systems (Aspen Publishers, Gaithesburg (MD), 1995).

48. Staas, W. E., Freedman, M. K. & Fried, G. F. Spinal cord injury and spinal cord injury
medicine. (eds. DeLisa, J. A. & Gans, B. M.) (Lippincott-Raven Publishers.,
Philadelphia, PA, 1998).

49. Burney, R. E., Maio, R. F., Maynard, F. & Karunas, R. Incidence, characteristics, and
outcome of spinal cord injury at trauma centers in North America. Arch Surg 128, 596-9
(1993).

50. Waters, R. L., Adkins, R. H., Yakura, J. & Sie, I. Profiles of spinal cord injury and
recovery after gunshot injury. Clin Orthop RelatRes, 14-21 (1991).

51. Waters, R. L., Sie, I., Adkins, R. H. & Yakura, J. S. Motor recovery following spinal cord
injury caused by stab wounds: a multicenter study. Paraplegia 33, 98-101 (1995).

52. McDonald, J. W. & Sadowsky, C. Spinal-cord injury. Lancet 359, 417-25. (2002).

53. Moore, A. P. & Blumhardt, L. D. A prospective survey of the causes of non-traumatic
spastic paraparesis and tetraparesis in 585 patients. Spinal Cord 35, 361-7 (1997).

54. McKinley, W. O., Seel, R. T. & Hardman, J. T. Nontraumatic spinal cord injury:
incidence, epidemiology, and functional outcome. Arch Phys MedRehabil 80, 619-23.
(1999).

55. New, P. W., Rawicki, H. B. & Bailey, M. J. Nontraumatic spinal cord injury:
demographic characteristics and complications. Arch Phys MedRehabil 83, 996-1001
(2002).

56. DeVivo, M. J. & Stover, S. L. in Spinal cord injury: clinical outcomes from the model
systems. (eds. Stover, S. L., DeLisa, J. A. & Whiteneck, G. G.) 289- 316. (Aspen
Publishers, Inc., Gaithersburg, Maryland, 1995).

57. Annual report (National Spinal Cord Injury Statistical Center, Birmingham, Alabama,
2004).

58. Maynard, F. M., Jr. et al. International standards for neurological and functional
classification of spinal cord injury. American Spinal Injury Association. Spinal Cord 35,
266-74 (1997).









59. Ditunno, J. F., Jr., Young, W., Donovan, W. H. & Creasey, G. The international
standards booklet for neurological and functional classification of spinal cord injury.
American Spinal Injury Association. Paraplegia 32, 70-80 (1994).

60. Standards for neurological classification of spinal injury patients. (Chicago, IL, 1984).

61. Burns, S. P., Golding, D. G., Rolle, W. A., Jr., Graziani, V. & Ditunno, J. F., Jr. Recovery
of ambulation in motor-incomplete tetraplegia. Arch Phys Med Rehabil 78, 1169-72
(1997).

62. Crozier, K. S., Graziani, V., Ditunno, J. F., Jr. & Herbison, G. J. Spinal cord injury:
prognosis for ambulation based on sensory examination in patients who are initially
motor complete. Arch Phys MedRehabil 72, 119-21 (1991).

63. Dobkin, B. et al. The evolution of walking-related outcomes over the first 12 weeks of
rehabilitation for incomplete traumatic spinal cord injury: the multicenter randomized
Spinal Cord Injury Locomotor Trial. Neurorehabil Neural Repair 21, 25-35 (2007).

64. Daverat, P. et al. Early prognostic factors for walking in spinal cord injuries. Paraplegia
26, 255-61 (1988).

65. Waters, R. L., Adkins, R., Yakura, J. & Vigil, D. Prediction of ambulatory performance
based on motor scores derived from standards of the American Spinal Injury Association.
Arch Phys Med Rehabil 75, 756-60 (1994).

66. Crozier, K. S. et al. Spinal cord injury: prognosis for ambulation based on quadriceps
recovery. Paraplegia 30, 762-7 (1992).

67. Maynard, F. M., Reynolds, G. G., Fountain, S., Wilmot, C. & Hamilton, R. Neurological
prognosis after traumatic quadriplegia. Three-year experience of California Regional
Spinal Cord Injury Care System. JNeurosurg. 50, 611-6 (1979).

68. Conrad, B., Benecke, R. & Meinck, H. M. Gait disturbances in paraspastic patients (eds.
Delwaide, P. J. & Young, R. R.) (Elsevier Science Publishers, 1985).

69. Fung, J. & Barbeau, H. A dynamic EMG profile index to quantify muscular activation
disorder in spastic paretic gait. Electroencephalogr Clin Neurophysiol 73, 233-44 (1989).

70. Krawetz, P. & Nance, P. Gait analysis of spinal cord injured subjects: effects of injury
level and spasticity. Arch Phys MedRehabil 77, 635-8 (1996).

71. Nene, A. V., Hermens, H. J. & Zilvold, G. Paraplegic locomotion: a review. Spinal Cord
34, 507-24 (1996).

72. Organization, W. H. (2001).

73. Umphred, D. A. Neurological rehabilitation (Mosby Inc., St.Louis, 2001).









74. Mytrice, B. A. Lower extremity orthotic management for the spinal-cord-injured client.
Spinal Cord nj Rehabil 5(4), 1 -10 (2000).

75. Nawoczenski, D. A. & Epler, M. E. Oi thitii in functional rehabilitation of the lower
limb (W.B. Saunders, Philadelphia, 1997).

76. Maxwell, D. J., Granat, M. H., Baardman, G. & Hermens, H. J. Demand for and use of
functional electrical stimulation systems and conventional orthoses in the spinal lesioned
community of the UK. Artif Organs 23, 410-2 (1999).

77. Ramon y Cajal, S. Degeneration and regeneration of the nervous system. (Oxford
University Press, London, 1928).

78. Guide tophysical therapy practice (American Physical Therapy Association, Fairfax,
VA, 2001).

79. van Tuijl, J. H., Janssen-Potten, Y. J. & Seele, H. A. Evaluation of upper extremity motor
function tests in tetraplegics. Spinal Cord 40, 51-64. (2002).

80. Kasaoka, K. & Sankai, Y. in IEEE/RSJInternational Conference on Intelligent Robots
and Systems (2001).

81. Kawamoto, H. & Sankai, Y. in 13th IEEE International Workshop on Robot and Human
Interactive Communication (2004).

82. Ferris, D. P., Sawicki, G. S. & Domingo, A. Powered lower limb orthoses for gait
rehabilitation. Top Spinal Cord Inj Rehabil 11, 34-49 (2005).

83. Grillner, S. Interaction between central and peripheral mechanisms in the control of
locomotion. Prog Brain Res 50, 227-35 (1979).

84. Grillner, S. Neurobiological bases of rhythmic motor acts in vertebrates. Science 228,
143-9 (1985).

85. Wolpaw, J. R. Acquisition and maintenance of the simplest motor skill: investigation of
CNS mechanisms. MedSci Sports Exerc 26, 1475-9 (1994).

86. Rossignol, S. et al. Locomotor capacities after complete and partial lesions of the spinal
cord. Acta NeurobiolExp 56, 449-63 (1996).

87. Lovely, R. G., Gregor, R. J., Roy, R. R. & Edgerton, V. R. Effects of training on the
recovery of full-weight-bearing stepping in the adult spinal cat. Exp Neurol 92, 421-35
(1986).

88. Barbeau, H., Julien, C. & Rossignol, S. The effects of clonidine and yohimbine on
locomotion and cutaneous reflexes in the adult chronic spinal cat. Brain Res 437, 83-96
(1987).









89. Duysens, J. & Pearson, K. G. Inhibition of flexor burst generation by loading ankle
extensor muscles in walking cats. Brain Res 187, 321-32 (1980).

90. de Leon, R. D., Hodgson, J. A., Roy, R. R. & Edgerton, V. R. Locomotor capacity
attributable to step training versus spontaneous recovery after spinalization in adult cats.
JNeurophysiol 79, 1329-40 (1998).

91. Pearson, K. G. Common principles of motor control in vertebrates and invertebrates.
Annu Rev Neurosci 16, 265-97 (1993).

92. Grillner, S., McClellan, A., Sigvardt, K., Wallen, P. & Wilen, M. Activation of NMDA-
receptors elicits "fictive locomotion" in lamprey spinal cord in vitro. Acta Physiol Scand
113, 549-51 (1981).

93. Grillner, S. & Dubuc, R. Control of locomotion in vertebrates: spinal and supraspinal
mechanisms. Adv Neurol 47, 425-53 (1988).

94. Nolte, J. The human brain, 5th edition an introduction to its functional anatomy
(Mosby-Year Book, St. Louis, MO, 2002).

95. Duysens, J. & Van de Crommert, H. W. Neural control of locomotion; the central pattern
generator from cats to humans. Gait Posture 7, 131-41 (1998).

96. Delcomyn, F. Neural basis of rhythmic behavior in animals. Science 210, 492-8 (1980).

97. Patla, A. E. in Evaluation and management of gait disorders (ed. Spivack, B. S.) (New
York, 1995).

98. Brooks, V. B. Cerebellar functions in motor control. Hum Neurobiol 2, 251-60 (1984).

99. Armstrong, D. M. The supraspinal control of mammalian locomotion. JPhysiol. 405, 1-
37 (1988).

100. Grillner, S. & Wallen, P. Central pattern generators for locomotion, with special
reference to vertebrates. Annu Rev Neurosci., 233-61 (1985).

101. Patla, A. E., Prentice, S. D., Robinson, C. & Neufeld, J. Visual control of locomotion:
strategies for changing direction and for going over obstacles. JExp Psychol Hum
Percept Perform 17, 603-34 (1991).

102. Patla, A. E., Beuter, A. & Prentice, S. A two stage correction of limb trajectory to avoid
obstacles during stepping. Neurosci Res Commun 8, 153-9 (1991a).

103. Weytjens, J. L., Viberg, D. A., Caputi, A. A., Kallesoe, K. & Hoffer, J. A. New
transducers for measuring muscle length in unrestrained animals. JNeurosci Methods 45,
217-25 (1992).









104. Schomburg, E. D. Spinal sensorimotor systems and their supraspinal control. Neurosci
Res. 7, 265-340 (1990).

105. Schomburg, E. D. Spinal functions in sensorimotor control of movements. Neurosurg
Rev. 13, 179-85 (1990).

106. Grillner, S. & Shik, M. L. On the descending control of the lumbosacral spinal cord from
the "mesencephalic locomotor region". Acta Physiol Scand 87, 320-33 (1973).

107. Shik, M. L., Severin, F. V. & Orlovsky, G. N. Control of walking and running by means
of electrical stimulation of the mesencephalon. Electroencephalogr Clin Neurophysiol
26, 549 (1969).

108. Orlovsky, G., Deliagina, T. & Grillner, S. Neural control of locomotion: from mollusc to
man. (Oxford Univ. Press, Oxford, UK, 1999).

109. Connor, N. P. & Abbs, J. H. Sensorimotor contributions of the basal ganglia: recent
advances. Phys Ther 70, 864-72 (1990).

110. Berthoz, A. The role of gaze in compensation of vestibular dysfunction: the gaze
substitution hypothesis. Prog Brain Res 76, 411-20 (1988).

111. Eidelberg, E. & Yu, J. Effects of corticospinal lesions upon treadmill locomotion by cats.
Exp Brain Res 43, 101-3 (1981).

112. Yu, J. & Eidelberg, E. Effects of vestibulospinal lesions upon locomotor function in cats.
Brain Res 220, 179-83 (1981).

113. Nashner, L. M. & Forssberg, H. Phase-dependent organization of postural adjustments
associated with arm movements while walking. JNeurophysiol 55, 1382-94 (1986).

114. Garcia-Rill, E. The basal ganglia and the locomotor regions. Brain Res 396, 47-63
(1986).

115. Winter, D. A. Biomechanics and motor control of human movement (John Wiley & Sons,
New Jersey, 2005).

116. Knudson, D. Fundamentals ofbiomechanics (Springer, 2003).

117. Robertson, D. G. E., Caldwell, D. E., Hamill, J., Kamen, G. & Whittlesey, S. N. Research
methods in biomechanics (Human Kinetics Publishers, 2004).

118. Gottschall, J. S. & Kram, R. Energy cost and muscular activity required for propulsion
during walking. JAppl Physiol 94, 1766-72 (2003).

119. Meinders, M., Gitter, A. & Czerniecki, J. M. The role of ankle plantar flexor muscle
work during walking. ScandJRehabilMed30, 39-46 (1998).









120. Brown, T. G. The intrinsic factors in the act of progression in the mammal. Proc R Soc
Lond[Biol] 84, 308-19 (1911).

121. Grillner, S. & Zangger, P. The effect of dorsal root transaction on the efferent motor
pattern in the cat's hindlimb during locomotion. Acta Physiol Scand 120, 393-405 (1984).

122. Shik, M. L., Severin, F. V. & Orlovsky, G. N. Control of walking and running by means
of electric stimulation of the midbrain. Biofizika, 659-66. (1966).

123. Maegele, M., Muller, S., Wemig, A., Edgerton, V. R. & Harkema, S. J. Recruitment of
spinal motor pools during voluntary movements versus stepping after human spinal cord
injury. JNeurotrauma 19, 1217-29. (2002).

124. Dimitrijevic, M. R., Gerasimenko, Y. & Pinter, M. M. Evidence for a spinal central
pattern generator in humans. Ann N YAcadSci 860, 360-76 (1998).

125. Edgerton, V. R. et al. A physiological basis for the development of rehabilitative
strategies for spinally injured patients. JAm Paraplegia Soc 14, 150-7 (1991).

126. Bassler, U. Sensory control of leg movement in the stick insect carausius morosus. Biol
Cybern 25, 61-72 (1977).

127. Dietz, V., Muller, R. & Colombo, G. Locomotor activity in spinal man: significance of
afferent input from joint and load receptors. Brain 125, 2626-34 (2002).

128. Sherrington, C. S. Flexion reflex of the limb, crossed extension reflex, and reflex
stepping and standing. Journal ofPhysiology 40, 28-121 (1910).

129. Grillner, S. & Rossignol, S. On the initiation of the swing phase of locomotion in chronic
spinal cats. Brain Res 146, 269-77 (1978).

130. Hiebert, G. W., Whelan, P. J., Prochazka, A. & Pearson, K. G. Contributions of hindlimb
flexor muscle afferents to the timing of phase transitions in the cat step cycle.
J..Neurophysiol 75, 1126-37 (1996).

131. Calancie, B. et al. Involuntary stepping after chronic spinal cord injury. Evidence for a
central rhythm generator for locomotion in man. Brain 117 ( Pt 5), 1143-59 (1994).

132. Dietz, V. & Duysens, J. Significance of load receptor input during locomotion: a review.
Gait Posture 11, 102-10. (2000).

133. Duysens, J. & Pearson, K. G. The role of cutaneous afferents from the distal hindlimb in
the regulation of the step cycle of thalamic cats. Exp Brain Res 24, 245-55 (1976).

134. Sinkjaer, T., Andersen, J. B., Ladouceur, M., Christensen, L. O. D. & Nielsen, J. B.
Major role for sensory feedback in soleus EMG activity in the stance phase of walking in
man. JPhysiol. 523, 817-827 (2000).









135. Harkema, S. J. et al. Human lumbosacral spinal cord interprets loading during stepping. J
Neurophysiol 77, 797-811 (1997).

136. Knikou, M. & Conway, B. A. Reflex effects of induced muscle contraction in normal and
spinal cord injured subjects. Muscle Nerve 26, 374-82 (2002).

137. Wolpaw, J. R. & O'Keefe, J. A. Adaptive plasticity in the primate spinal stretch reflex:
evidence for a two-phase process. JNeurosci 4, 2718-24 (1984).

138. Palmieri, R. M., Ingersoll, C. D. & Hoffman, M. A. The Hoffmann reflex: methodologic
considerations and applications for use in sports medicine and athletic training research. J
Athl Train. 39, 268-77 (2004).

139. Hoffmann, P. Beitrag zur Kenntnis der menschlichen Reflexe mit besonderer
Berucksichtigung der elektrischen Erscheinungen. Arch Anat Physiol. 1, 223-46 (1910).

140. Fisher, M. A. AAEM Minimonograph #13: H reflexes and F waves: physiology and
clinical indications. Muscle Nerve. 15, 1223-33 (1992).

141. Braddom, R. I. & Johnson, E. W. Standardization of H reflex and diagnostic use in Sl
radiculopathy. Arch Phys Med Rehabil 55, 161-6 (1974).

142. Braddom, R. L. & Johnson, E. W. H reflex: review and classification with suggested
clinical uses. Arch Phys Med Rehabil 55, 412-7 (1974).

143. Bell, K. R. & Lehmann, J. F. Effect of cooling on H- and T-reflexes in normal subjects.
Arch Phys Med Rehabil 68, 490-3 (1987).

144. Krause, B. A., Hopkins, J. T., Ingersoll, C. D., Cordova, M. L. & Edwards, J., E. The
relationship of ankle temperature during cooling and rewarming to the human soleus H
reflex. Journal of Sport Rehabilitation 9, 253-62 (2000).

145. Hoffman, M. A. & Koceja, D. M. The effects of vision and task complexity on Hoffmann
reflex gain. Brain Res. 700, 303-7 (1995).

146. Capaday, C. & Stein, R. B. Amplitude modulation of the soleus H-reflex in the human
during walking and standing. JNeurosci 6, 1308-13 (1986).

147. Capaday, C. & Stein, R. B. Difference in the amplitude of the human soleus H reflex
during walking and running. JPhysiol392, 513-22 (1987).

148. Schieppati, M. The Hoffman reflex: a means of assessing spinal reflex excitability and its
descending control in man. Progress in Neurobiology 28, 345-76 (1987).

149. Zehr, E. P. Considerations for use of the Hoffmann reflex in exercise studies. Eur JAppl
Physiol. 86, 455-68 (2002).









150. Chalmers, G. R. & Knutzen, K. M. Soleus Hoffmann-reflex modulation during walking
in healthy elderly and young adults. J GerontolA Biol Sci Med Sci 55, B570-9 (2000).

151. Simonsen, E. B. & Dyhre-Poulsen, P. Amplitude of the human soleus H reflex during
walking and running. JPhysiol 515 ( Pt 3), 929-39 (1999).

152. Brooke, J. D., Cheng, J., Misiaszek, J. E. & Lafferty, K. Amplitude modulation of the
soleus H reflex in the human during active and passive stepping movements. J
Neurophysiol. 73, 102-11 (1995).

153. Brooke, J. D. et al. Sensori-sensory afferent conditioning with leg movement: gain
control in spinal reflex and ascending paths. Prog Neurobiol 51, 393-421 (1997).

154. Schneider, C., Lavoie, B. A. & Capaday, C. On the origin of the soleus H-reflex
modulation pattern during human walking and its task-dependent differences. J
Neurophysiol. 83, 2881-90 (2000).

155. Fung, J. & Barbeau, H. Effects of conditioning cutaneomuscular stimulation on the soleus
H-reflex in normal and spastic paretic subjects during walking and standing. J
Neurophysiol 72, 2090-104 (1994).

156. Yang, J. F. et al. H-reflex modulation during walking in spastic paretic subjects. Can J
Neurol Sci 18, 443-52 (1991).

157. Knikou, M. & Conway, B. A. Modulation of soleus H-reflex following ipsilateral
mechanical loading of the sole of the foot in normal and complete spinal cord injured
humans. Neurosci Lett 303, 107-10 (2001).

158. Kamiya, A., Tanabe, S., Muraoka, Y. & Masakado, Y. Modulation of the soleus H-reflex
during static and dynamic imposed hip angle changes. Int JNeurosci 116, 1045-53
(2006).

159. Knikou, M. & Rymer, W. Z. Hip angle induced modulation of H reflex amplitude,
latency and duration in spinal cord injured humans. Clin Neurophysiol 113, 1698-708
(2002).

160. Wolpaw, J. R. & Carp, J. S. Plasticity from muscle to brain. ProgNeurobiol 78, 233-63
(2006).

161. Shurrager, P. S. & Dykman, R. A. Walking spinal carnivores. J Comp Physiol Psychol.
44, 252-62(1951).

162. de Leon, R. D., Tamaki, H., Hodgson, J. A., Roy, R. R. & Edgerton, V. R. Hindlimb
locomotor and postural training modulates glycinergic inhibition in the spinal cord of the
adult spinal cat. JNeurophysiol 82, 359-69 (1999).

163. Chau, C., Barbeau, H. & Rossignol, S. Early locomotor training with clonidine in spinal
cats. JNeurophysiol 79, 392-409 (1998).









164. Chau, C., Barbeau, H. & Rossignol, S. Effects of intrathecal alphal- and alpha2-
noradrenergic agonists and norepinephrine on locomotion in chronic spinal cats. J
Neurophysiol 79, 2941-63 (1998).

165. Roy, R. R. & Acosta, L., Jr. Fiber type and fiber size changes in selected thigh muscles
six months after low thoracic spinal cord transaction in adult cats: exercise effects. Exp
Neurol 92, 675-85 (1986).

166. Dietz, V. Spinal cord pattern generators for locomotion. Clin Neurophysiol 114, 1379-89
(2003).

167. Krassioukov, A. V., Bunge, R. P., Pucket, W. R. & Bygrave, M. A. The changes in
human spinal sympathetic preganglionic neurons after spinal cord injury. Spinal Cord 37,
6-13 (1999).

168. Wolpaw, J. R. & Herchenroder, P. A. Operant conditioning of H-reflex in freely moving
monkeys. JNeurosci Methods 31, 145-52 (1990).

169. Wolpaw, J. R. The complex structure of a simple memory. Trends Neurosci 20, 588-94
(1997).

170. Wolpaw, J. R. & Tennissen, A. M. Activity-dependent spinal cord plasticity in health and
disease. Annu Rev Neurosci 24, 807-43 (2001).

171. Chen, X. Y., Chen, Y., Chen, L., Tennissen, A. M. & Wolpaw, J. R. Corticospinal tract
transaction permanently abolishes H-reflex down-conditioning in rats. JNeurotrauma 23,
1705-12 (2006).

172. Chen, Y. et al. Operant conditioning of H-reflex can correct a locomotor abnormality
after spinal cord injury in rats. JNeurosci 26, 12537-43 (2006).

173. Chen, X. Y., Chen, L. & Wolpaw, J. R. Conditioned H-reflex increase persists after
transaction of the main corticospinal tract in rats. JNeurophysiol 90, 3572-8 (2003).

174. Perez, M. A., Lungholt, B. K. & Nielsen, J. B. Presynaptic control of group Ia afferents in
relation to acquisition of a visuo-motor skill in healthy humans. JPhysiol 568, 343-54
(2005).

175. Schneider, C. & Capaday, C. Progressive adaptation of the soleus H-reflex with daily
training at walking backward. JNeurophysiol 89, 648-56 (2003).

176. Trimble, M. H. & Koceja, D. M. Downtraining of triceps surae H-reflexes during a
balance task in humans. Intern JNeurosci 76, 293-303 (1994).

177. Hess, F., Van Hedel, H. J. & Dietz, V. Obstacle avoidance during human walking: H-
reflex modulation during motor learning. Exp Brain Res 151, 82-9 (2003).









178. Segal, R. L. & Wolf, S. L. Operant conditioning of spinal stretch reflexes in patients with
spinal cord injuries. Exp Neurol 130, 202-13 (1994).

179. Bouyer, L. J., Whelan, P. J., Pearson, K. G. & Rossignol, S. Adaptive locomotor
plasticity in chronic spinal cats after ankle extensors neurectomy. JNeurosci 21, 3531-41
(2001).

180. Carrier, L., Brustein, E. & Rossignol, S. Locomotion of the hindlimbs after neurectomy
of ankle flexors in intact and spinal cats: model for the study of locomotor plasticity. J
Neurophysiol 77, 1979-93 (1997).

181. Trimble, M. H., Behrman, A. L., Flynn, S. M., Thigpen, M. T. & Thompson, F. J. Acute
effects of locomotor training on overground walking speed and H- reflex modulation in
individuals with incomplete spinal cord injury. JSpinal CordMed 24, 74-80. (2001).

182. Wolpaw, J. R. The education and re-education of the spinal cord. Prog Brain Res 157,
261-280 (2006).

183. Dobkin, B. H., Harkema, S., Requejo, P. & Edgerton, V. R. Modulation of locomotor-
like EMG activity in subjects with complete and incomplete spinal cord injury. JNeurol
Rehabil9, 183-90 (1995).

184. Schmidt, R. A. & Young, D. E. Methodology for motor learning: A paradigm for
kinematic feedback. JMot Behav 23, 13-24 (1991).

185. Dietz, V. Evidence for a load receptor contribution to the control of posture and
locomotion. Neurosci Biobehav Rev 22, 495-9 (1998).

186. Wernig, A., Nanassy, A. & Muller, S. Laufband (treadmill) therapy in incomplete
paraplegia and tetraplegia. JNeurotrauma 16, 719-26 (1999).

187. Gardner, M. B., Holden, M. K., Leikauskas, J. M. & Richard, R. L. Partial body weight
support with treadmill locomotion to improve gait after incomplete spinal cord injury: a
single-subject experimental design. Phys Ther 78, 361-74 (1998).

188. Wirz, M., Colombo, G. & Dietz, V. Long term effects of locomotor training in spinal
humans. JNeurol Neurosurg Psychiatry 71, 93-6. (2001).

189. Barbeau, H., Pepin, A., Norman, K. E., Ladouceur, M. & Leroux, A. Walking after spinal
cord injury: control and recovery. The neuroscientist 4, 14-24 (1998).

190. Behrman, A. L. & Harkema, S. J. Recovery of walking: locomotor training after human
spinal cord injury. (submitted).

191. Dietz, V., Colombo, G., Jensen, L. & Baumgartner, L. Locomotor capacity of spinal cord
in paraplegic patients. Ann Neurol 37, 574-82. (1995).









192. Field-Fote, E. C., Lindley, S. D. & Sherman, A. L. Locomotor training approaches for
individuals with spinal cord injury: a preliminary report of walking-related outcomes. J
NeurolPhys Ther 29, 127-37 (2005).

193. Chen, G. & Patten, C. Treadmill training with harness support: selection of parameters
for individuals with poststroke hemiparesis. JRehabil Res Dev. 43, 485-98 (2006).

194. Conrad, B., Benecke, R. & Carnehl, J. in Motor control mechanisms in health and
disease (ed. Desmedt, J. E.) 717-726 (Raven Press, New York, 1983).

195. Visintin, M. & Barbeau, H. The effects of parallel bars, body weight support and speed
on the modulation of the locomotor pattern of spastic paretic gait. A preliminary
communication. Paraplegia 32, 540-53 (1994).

196. Dobkin, B. in Joint meeting ofACRMaIndASNR (Tuscon, AZ, 2003).

197. Mandzak-McCarron, K. & Drayton-Hargrove, S. Ambulation aids. RehabilNurs 12, 139-
41(1987).

198. Mulley, G. Walking frames. Bmj 300, 925-7 (1990).

199. Parziale, J. R. & Daniels, J. D. The mechanical performance of ambulation using spring-
loaded axillary crutches. A preliminary report. Am JPhys MedRehabil 68, 192-5 (1989).

200. Waters, R. L., Yakura, J. S. & Adkins, R. H. Gait performance after spinal cord injury.
Clin Orthop Relat Res, 87-96 (1993).

201. Opila, K. A., Nicol, A. C. & Paul, J. P. Forces and impulses during aided gait. Arch Phys
MedRehabil 68, 715-22 (1987).

202. Melis, E. H., Torres-Moreno, R., Barbeau, H. & Lemaire, E. D. Analysis of assisted-gait
characteristics in persons with incomplete spinal cord injury. Spinal Cord 37, 430-9.
(1999).

203. Ounpuu, S., Bell, K. J., Davis, R. B., 3rd & DeLuca, P. A. An evaluation of the posterior
leaf spring orthosis using joint kinematics and kinetics. JPediatr Orthop 16, 378-84
(1996).

204. Brooke, J. D., McIlroy, W. E. & Collins, D. F. Movement features and H-reflex
modulation. I. Pedalling versus matched controls. Brain Res 582, 78-84 (1992).

205. Nishikawa, T. & Grabiner, M. D. Peroneal motoneuron excitability increases
immediately following application of a semirigid ankle brace. J Orthop Sports Phys Ther
29, 168-73; discussion 174-6 (1999).

206. Garrett, M., Kerr, T. & Caulfield, B. Phase-dependent inhibition of H-reflexes during
walking in humans is independent of reduction in knee angular velocity. JNeurophysiol
82, 747-53 (1999).









207. Bach-y-Rita, P. Rehabilitation versus passive recovery of motor control following central
nervous system lesions. Adv Neurol 39, 1085-92 (1983).

208. Kadaba, M. P., Ramakrishnan, H. K. & Wootten, M. E. Measurement of lower extremity
kinematics during level walking. J Orthop Res 8, 383-92 (1990).

209. Cappozzo, A. in Adaptability of human gait. (ed. Patla, A. E.) 167-85 (Elsevier Science
Pub, North-Holland, 1991).

210. Lafortune, M. A., Lake, M., J. & Hennig, E. M. Differential shock transmission response
of the human body toimpact severity and lower limb posture. JBiomech 29, 1531-7.
(1996).

211. Kadaba, M. P. et al. Repeatability of kinematic, kinetic, and electromyographic data in
normal adult gait. J Orthop Res 7, 849-60 (1989).

212. Richards, J. G. The measurement of human motion: a comparison of commercially
available systems. Human Movement Science 18, 589-602 (1999).

213. Baker, R. Gait analysis methods in rehabilitation. JNeuroengineeringRehabil 3, 4
(2006).

214. Schwartz, M. H., Trost, J. P. & Wervey, R. A. Measurement and management of errors in
quantitative gait data. Gait Posture 20, 196-203 (2004).

215. Barker, S. P. in School of Biomedical Engineering, Science and Health Systems (Drexel
University, 2004).

216. Hanke, T. A. & Rogers, M. W. Reliability of ground reaction force measurements during
dynamic transitions from bipedal to single-limb stance in healthy adults. Phys Ther 72,
810-6 (1992).

217. Lukacs, R. U. et al. in Society for Neuroscience (2002).

218. Morarty, R. & Agosta, J. in Australian Conference of Science and Medicine in Sport
(Adelaide, 1998).

219. Cordova, M. L. & Armstrong, C. W. Reliability of ground reaction forces during a
vertical jump: implications for functional strength assessment. JAthl Train 31, 342-5
(1996).

220. Hopkins, J. T., Ingersoll, C. D., Cordova, M., L, & Edwards, J., E. Intrasession and
intersession reliability of the soleus H-reflex in supine and standing positions.
Electromyogr Clin Neurophysiol. 40, 89-94 (2000).

221. Paquet, N. & Hui-Chan, C. W. Human soleus H-reflex excitability is decreased by
dynamic head-and-body tilts. J Vestib Res 9, 379-83 (1999).









222. Pierrot-Deseilligny, E. & Mazevet, D. The monosynaptic reflex: a tool to investigate
motor control in humans. Interest and limits. Neurophysiol Clin 30, 67-80 (2000).

223. Honore, J., Demaire, C. & Coquery, J. M. Effects of spatially oriented attention on the
facilitation of the H reflex by a cutaneous stimulus. Electroencephalogr Clin
Neurophysiol 55, 156-61 (1983).

224. Bonnet, M., Decety, J., Jeannerod, M. & Requin, J. Mental simulation of an action
modulates the excitability of spinal reflex pathways in man. Brain Res Cogn Brain Res 5,
221-8 (1997).

225. Walton, C., Kalmar, J. & Cafarelli, E. Caffeine increases spinal excitability in humans.
Muscle Nerve 28, 359-64 (2003).

226. Funase, K. & Miles, T. S. Observations on the variability of the H reflex in human soleus.
Muscle Nerve 22, 341-6 (1999).

227. Crone, C., Johnsen, L. L., Hultbom, H. & Orsnes, G. B. Amplitude of the maximum
motor response (Mmax) in human muscles typically decreases during the course of an
experiment. Exp Brain Res 124, 265-70 (1999).

228. Stein, R. B. & Capaday, C. The modulation of human reflexes during functional motor
tasks. Trends Neurosci 11, 328-32 (1988).

229. Crenna, P. & Frigo, C. Excitability of the soleus H-reflex arc during walking and
stepping in man. Exp Brain Res 66, 49-60 (1987).

230. Chen, X. Y. & Wolpaw, J. R. Dorsal column but not lateral column transaction prevents
down-conditioning of H reflex in rats. JNeurophysiol 78, 1730-4 (1997).

231. Segal, R. Plasticity in the central nervous system: operant conditioning of the spinal
stretch reflex. Top Stroke Rehabilitation 3, 76-87 (1997).

232. Sawicki, G. S., Domingo, A. & Ferris, D. P. The effects of powered ankle-foot orthoses
on joint kinematics and muscle activation during walking in individuals with incomplete
spinal cord injury. JNeuroengineering Rehabil 3, 3 (2006).

233. Ferris, D. P., Gordon, K. E., Sawicki, G. S. & Peethambaran, A. An improved powered
ankle-foot orthosis using proportional myoelectric control. Gait Posture 23, 425-8
(2006).

234. Gordon, K. E., Sawicki, G. S. & Ferris, D. P. Mechanical performance of artificial
pneumatic muscles to power an ankle-foot orthosis. JBiomech 39, 1832-41 (2006).

235. Hylton, N. Dynamic orthotic concepts: background and experiences. (Orthopaedie
technology, Dortmund, Germany, 2000).









236. Beekman, C. E., Miller-Porter, L. & Schoneberger, M. Energy cost of propulsion in
standard and ultralight wheelchairs in people with spinal cord injuries. Phys Ther 79,
146-58 (1999).

237. Dobkin, B. H. et al. Methods for a randomized trial of weight-supported treadmill
training versus conventional training for walking during inpatient rehabilitation after
incomplete traumatic spinal cord injury. Neurorehabil Neural Repair 17, 153-67 (2003).

238. Lacquaniti, F., Ivanenko, Y. P. & Zago, M. Kinematic control of walking. Arch ItalBiol
140, 263-72 (2002).

239. Patrick, J. H. Case for gait analysis as part of the management of incomplete spinal cord
injury. Spinal Cord 41, 479-82 (2003).

240. Smith, P. A., Hassani, S., Reiners, K., Vogel, L. C. & Harris, G. F. Gait analysis in
children and adolescents with spinal cord injuries. JSpinal CordMed27 Suppl 1, S44-9
(2004).

241. Ferris, D. P., Gordon, K. E., Beres-Jones, J. A. & Harkema, S. J. Muscle activation
during unilateral stepping occurs in the nonstepping limb of humans with clinically
complete spinal cord injury. Spinal Cord 42, 14-23 (2004).

242. Hidler, J. Robotic-assessment of walking in individuals with gait disorders. ConfProc
IEEE Eng MedBiol Soc 7, 4829-31 (2004).

243. Balasubramanian, C. K., Bowden, M. G., Neptune, R. R. & Kautz, S. A. Relationship
between step length asymmetry and walking performance in subjects with chronic
hemiparesis. Arch Phys MedRehabil. 88, 43-9 (2007).

244. Aisen, M. L. O/ tlltiik in neurologic rehabilitation (Demos, New York, NY, 1992).

245. Portney, L. G. & Watkins, M. P. Foundations of clinical research: applications to
practice. (Prentice-Hall, Upper Saddle River, NJ, 2000).

246. Westfall, P. H. & Young, S. S. Resampling-based multiple testing: examples and methods
forp-value adjustment (John Wiley & Sons, Inc., New York, 1993).

247. Kuo, A. D., Donelan, J. M. & Ruina, A. Energetic consequences of walking like an
inverted pendulum: step-to-step transitions. Exerc Sport Sci Rev. 33, 88-97 (2005).

248. Beekman, C. E., Perry, J., Boyd, L. A., Newsam, C. J. & Mulroy, S. J. The effects of a
dorsiflexion-stopped ankle foot orthosis on walking in individuals with incomplete spinal
cord injury. Topics in Spinal Cord Injury Rehabilitation 5, 54-62 (2000).

249. Hesse, S., Werner, C., Matthias, K., Stephen, K. & Berteanu, M. Non-velocity-related
effects of a rigid double-stopped ankle-foot orthosis on gait and lower limb muscle
activity of hemiparetic subjects with an equinovarus deformity. Stroke. 30, 1855-61.
(1999).









250. Pepin, A., Norman, K. E. & Barbeau, H. Treadmill walking in incomplete spinal-cord-
injured subjects: 1. Adaptation to changes in speed. Spinal Cord 41, 257-70. (2003).

251. Pepin, A., Ladouceur, M. & Barbeau, H. Treadmill walking in incomplete spinal-cord-
injured subjects: 2. Factors limiting the maximal speed. Spinal Cord 41, 271-9. (2003).

252. Dromerick, A. W., Lum, P. S. & Hidler, J. Activity-based therapies. NeuroRx. 3, 428-38
(2006).

253. Yang, J. F., Stein, R. B. & James, K. B. Contribution of peripheral afferents to the
activation of the soleus muscle during walking in humans. Exp Brain Res 87, 679-87
(1991).

254. Yang, J. F. & Whelan, P. J. Neural mechanisms that contribute to cyclical modulation of
the soleus H-reflex in walking in humans. Exp Brain Res 95, 547-56 (1993).

255. Cordova, M. L. & Ingersoll, C. D. Peroneus longus stretch reflex amplitude increases
after ankle brace application. Br JSports Med 37, 258-62 (2003).

256. Stein, R. B. & Thompson, A. K. Muscle reflexes in motion: how, what, and why? Exerc
Sport Sci Rev 34, 145-53 (2006).

257. Frigon, A., Collins, D. F. & Zehr, E. P. Effect of rhythmic arm movement on reflexes in
the legs: modulation of soleus H-reflexes and somatosensory conditioning. J
Neurophysiol 91, 1516-23 (2004).

258. Sayenko, D. G. et al. Facilitation of the soleus stretch reflex induced by electrical
excitation of plantar cutaneous afferents located around the heel. Neurosci Lett 415, 294-
8 (2007).

259. Hwang, I. S. Assessment of soleus motoneuronal excitability using the joint angle
dependent H reflex in humans. JElectromyogr Kinesiol 12, 361-6 (2002).

260. Patikas, D. A., Kotzamanidis, C., Robertson, C. T. & Koceja, D. M. The effect of the
ankle joint angle in the level of soleus Ia afferent presynaptic inhibition. Electromyogr
Clin Neurophysiol 44, 503-11 (2004).

261. Sefton, J. M., Hicks-Little, C. A., Koceja, D. M. & Cordova, M. L. Effect of inversion
and ankle bracing on peroneus longus Hoffmann reflex. ScandJMedSci Sports 17, 539-
46 (2007).

262. Sefton, J. M., Hicks-Little, C. A., Koceja, D. M. & Cordova, M. L. Modulation of soleus
H-reflex by presynaptic spinal mechanisms during varying surface and ankle brace
conditions. Neurophysiol Clin 37, 15-21 (2007).

263. Stein, R. B. Presynaptic inhibition in humans. Prog Neurobiol 47, 533-44 (1995).









264. Little, J. W. & Halar, E. M. H-reflex changes following spinal cord injury. Arch Phys
MedRehabil 66, 19-22 (1985).

265. Hiersemenzel, L. P., Curt, A. & Dietz, V. From spinal shock to spasticity: neuronal
adaptations to a spinal cord injury. Neurology 54, 1574-82 (2000).

266. Knikou, M. Plantar cutaneous input modulates differently spinal reflexes in subjects with
intact and injured spinal cord. Spinal Cord. 45, 69-77 (2007).

267. Luscher, H., Ruenzel, P. W. & Henneman, E. Effects of impulse frequency, PTP, and
temperature on responses elicited in large populations of motoneurons by impulses in
single Ia-fibers. JNeurophysiol 50, 1045-58 (1983).

268. Trimble, M. H., Kukulka, C. G. & Behrman, A. L. The effect of treadmill gait training on
low-frequency depression of the soleus H-reflex: comparison of a spinal cord injured man
to normal subjects. Neurosci Lett 246, 186-8. (1998).

269. Kim, C. M., Eng, J. J. & Whittaker, M. W. Effects of a simple functional electric system
and/or a hinged ankle-foot orthosis on walking in persons with incomplete spinal cord
injury. Arch Phys Med Rehabil 85, 1718-23 (2004).

270. Cordova, M. L., Ingersoll, C. D. & Palmieri, R. M. Efficacy of prophylactic ankle
support: an experimental perspective. JAthl Train 37, 446-57 (2002).

271. Feuerbach, J. W., Grabiner, M. D., Koh, T. J. & Weiker, G. G. Effect of an ankle orthosis
and ankle ligament anesthesia on ankle joint proprioception. Am J Sports Med 22, 223-9
(1994).

272. Iles, J. F. Evidence for cutaneous and corticospinal modulation of presynaptic inhibition
of la afferents from the human lower limb. JPhysiol 491 ( Pt 1), 197-207 (1996).

273. Rudomin, P. & Schmidt, R. F. Presynaptic inhibition in the vertebrate spinal cord
revisited. Exp Brain Res 129, 1-37 (1999).

274. Bastiaanse, C. M., Duysens, J. & Dietz, V. Modulation of cutaneous reflexes by load
receptor input during human walking. Exp Brain Res 135, 189-98 (2000).

275. Calancie, B. et al. Evidence that alterations in presynaptic inhibition contribute to
segmental hypo- and hyperexcitability after spinal cord injury in man.
Electroencephalogr Clin Neurophysiol 89, 177-86 (1993).

276. Nakajima, T., Sakamoto, M., Tazoe, T., Endoh, T. & Komiyama, T. Location specificity
of plantar cutaneous reflexes involving lower limb muscles in humans. Exp Brain Res.
175, 514-25. (2006).

277. Aniss, A. M., Gandevia, S. C. & Burke, D. Reflex responses in active muscles elicited by
stimulation of low-threshold afferents from the human foot. JNeurophysiol 67, 1375-84
(1992).









278. Yang, J. F. & Stein, R. B. Phase-dependent reflex reversal in human leg muscles during
walking. JNeurophysiol 63, 1109-17 (1990).









BIOGRAPHICAL SKETCH

Preeti Nair was born in Mumbai, India in 1979. She received her bachelor's degree in

physical therapy from Pune University, India, in 2001. She worked for a year at a government

hospital in Mumbai and for the non-profit Multiple Sclerosis Society of India. Her specific

interest in neurological rehabilitation developed from two areas of interest; first, the elusive

workings of the nervous system for the control of movement and learning. Secondly, her work

experience with individuals with neurological impairment and the challenges faced in restoring

them to their activities of daily living. The process of disablement unleashed by the disease state

and perpetuated by underdeveloped infrastructure and apathy for neurological rehabilitation

motivated her to pursue higher education in the United States; a country that has set the mark for

its multidimensional approach towards enablement and empowerment of an individual with

impairment. Duly, she chose the interdisciplinary, Rehabilitation Sciences Doctoral program

offered at the University of Florida in 2002. Under the expert tutelage of Dr. Andrea Behrman

and Dr. Steven Kautz, her research deals with examining the neuromechanical control of walking

with orthotic devices in individuals after spinal cord injury that integrates the principles of

neurological control of walking, motor control and movement mechanics. She is a recipient of

the Alumni Fellowship which provided financial support for her doctoral education.





PAGE 1

1 NEUROMECHANICAL AND NE UROPHYSIOLOGICAL EXAM INATION OF WALKING WITH AN ANKLE FOOT ORTHOSIS IN NO N-INJURED INDIVIDUALS AND PERSONS WITH INCOMPLETE SPINAL CORD INJURY By PREETI MOHANDAS NAIR A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Preeti Mohandas Nair

PAGE 3

3 To Lord Ganesha and Shri Gajanan Maha raj for dwelling upon me at all times!!

PAGE 4

4 ACKNOWLEDGMENTS I am grateful to several i ndividuals who have been my guiding angels throughout this process. The greater part of this work was ma de possible by the instruction of my teachers and by the love and support of my family and friends. It is with my hear tfelt gratitude that I acknowledge each of them. First and foremost, I am thankful to my mentor, Dr. Andrea L. Behrman, for helping me clear this milestone. I will always be indebted to her for her excellen t guidance throughout and hope that in the future I am blessed with more opportunities to work with a remarkable teacher, researcher and human being such as herself. Also, I am grateful to my committee members, Dr. Steven Kautz, Dr. John Rosenbek, Dr. Floyd Thomps on and Dr. Paul Zehr, for their valuable and significant contributions to better my work. Sincere thanks to all the vol unteers who have contributed si gnificantly to these studies. This work would not have been possible or complete without their willingness to participate. My heartfelt thanks also to my friends from both the Locomotor training and HMPL lab with whom I have had the privilege to work. They have been my family while I have been away from home. Special thanks to Dr. Mark Bishop, Dr. Ma ry Thigpen and Dr. Claudia Senesac for their encouragement and emotional support during the home stretch. I extend my thanks to the University of Florida for bestowing me with the Alumni Fellowship that financially supported my education here. Many thanks also, to the personnel in the Department of Physical Therapy, who have helped me in one way or another during my graduate studies. Finally, I would like to express my deepest gr atitude to my family members. My parents have provided me unconditional love and emotiona l support and have been my pillars of strength through trying times. They have sacrificed significantly to make my dreams come true and I can

PAGE 5

5 only hope that I can make them proud and happy always I am also indebted to my sister and her family for their support and encouragement during the past few years. Last but not least, I am grateful to GOD ALMIGHTY for dwelling upon me at all times and keeping me focused, strong and always appreciative of the bigger good in life. This project would never have happened if it was not for the blessings of my family and HIM.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ........10 LIST OF FIGURES.......................................................................................................................12 ABSTRACT...................................................................................................................................14 CHAP TER 1 NEUROMECHANICAL AND NEUROPHY SIOLOGIC AL EXAMINATION OF WALKING WITH AN ANKLE FOOT ORTH OSIS IN HEALTHY, NON-INJURED INDIVIDUALS AND PERSONS WITH INCOMPLETE SPINAL CORD INJURY..........17 2 LITERATURE REVIEW.......................................................................................................22 Overview of Human Spinal Cord Injury: Consequence and R ehabilitation........................... 22 Introduction to the Problem.............................................................................................22 Consequence of Injury..................................................................................................... 22 Walking Potential after SCI............................................................................................. 23 Rehabilitation of Individuals with ISCI.......................................................................... 24 Orthotic Devices: Rationale for Use and Prescription.................................................... 25 Neurobiological Control of Walking...................................................................................... 26 Stepping Mechanics................................................................................................................30 Basic Instrumentation and Terminology......................................................................... 30 Characterization of Stepping Pattern...............................................................................32 Neural Assembly for Stepping................................................................................................ 33 Central Pattern Generator for Stepping........................................................................... 33 Sensory Drive Required for Stepping.............................................................................. 35 H-reflex Elicitation and M odulation During Stepping .................................................... 38 Elicitation.................................................................................................................38 Task-specific/ Phase-dependent modulation of the H-reflex................................... 39 Recovery of Walking after SCI.............................................................................................. 40 Plasticity at the Level of the Spinal Cord........................................................................ 40 Plasticity of the Spinal Stretch Reflex (SSR)..................................................................42 Locomotor Training and Skill-dependent Plas ticity of the Nervous System.................. 44 Is Orthotic use Appropriate during Locomotor Training?...................................................... 45 Altered Stepping Mechanics Resulting from the Use of Orthotic/Assistive Devices ..... 46 Altered H-reflexes Resulting from the Use of Orthotic Devices .................................... 47 Rationale of the Studies...................................................................................................48

PAGE 7

7 Methodological Considerations for the Measurem ent of Mechanics and Soleus H-reflex During Stepping..................................................................................................................49 Measurement of Stepping Mechanics............................................................................. 49 Study considerations................................................................................................49 Interpretation of kinematic and kinetic data............................................................. 49 Reliability of Vicon a nd Force Plate Measures ............................................................... 50 Measurement of Soleus H-reflex..................................................................................... 52 Study considerations................................................................................................52 Interpretation of H-reflex amplitude........................................................................ 53 Reliability of Soleus H-reflex..........................................................................................54 Clinical and Scientific Relevance of the Study...................................................................... 54 Clinical Relevance of the Study......................................................................................54 Scientific Relevance of the Study.................................................................................... 55 3 EXAMINATION OF WALKING WITH AND W ITHOUT AN ANKLE FOOT ORTHOSIS IN NON-INJURED INDI VIDUALS AND PERSONS WITH INCOMPLETE SPINAL CORD INJURY: A NEUROMECHANICAL PERSPECTIVE... 61 4 EXAMINATION OF WALKING WITH AND W ITHOUT AN ANKLE FOOT ORTHOSIS IN HEALTHY, N ON-INJURED INDIVIDUALS-1A...................................... 64 Hypotheses..............................................................................................................................64 Methods..................................................................................................................................64 Subject Selection.............................................................................................................64 Experimental Set-up........................................................................................................65 Subject Preparation..........................................................................................................66 Procedure.........................................................................................................................66 Data Processing...............................................................................................................67 Data Analysis...................................................................................................................68 Results.....................................................................................................................................68 Discussion...............................................................................................................................69 Limitation..................................................................................................................... ...71 Conclusion.......................................................................................................................71 5 COMPARISON OF WALKING WITH AND WITHOUT ANKLE F OOT ORTHOSIS IN PERSONS WITH INCOMPLE TE SPINAL CORD INJURY-1B................................... 79 Specific Aims and Hypothesis................................................................................................ 79 Methods..................................................................................................................................80 Subject Selection.............................................................................................................80 Experimental Set-up........................................................................................................81 Subject Preparation..........................................................................................................82 Procedure.........................................................................................................................82 Data Processing...............................................................................................................83 Data Analysis...................................................................................................................84 Results.....................................................................................................................................84 Within-Subject Comparisons for Individuals with ISCI................................................. 84

PAGE 8

8 Between-Subject Comparisons for Individuals with ISCI W alking Without a PAFO and Their Matched Control.......................................................................................... 85 Between-Subject Comparisons for Individua ls with ISCI W alking with a PAFO and Their Matched Control.......................................................................................... 85 Temporal and Spatial Comparisons Within Subjects...................................................... 85 Discussion...............................................................................................................................85 Clinical Implication......................................................................................................... 87 Limitations.................................................................................................................... ...87 Conclusion.......................................................................................................................88 6 PHASE DEPENDENT MODULATION OF S OLEUS H-REFLEX IN HEALTHY, NON-INJURED INDIVIDUALS WHILE WA LKING WITH AN ANKLE FOOT ORTHOSIS...........................................................................................................................103 Introduction................................................................................................................... ........103 Specific Aims........................................................................................................................104 Methods................................................................................................................................105 Subject Selection...........................................................................................................105 Experimental Set-up......................................................................................................105 Subject Preparation........................................................................................................106 Procedure.......................................................................................................................106 Data Processing.............................................................................................................108 Data Analysis.................................................................................................................108 Results...................................................................................................................................108 Discussion.............................................................................................................................109 Limitations.................................................................................................................... .111 Conclusion.....................................................................................................................112 7 IMMEDIATE, PHASE DEPENDENT, SOLEUS H-REFLEX MODULATION IN PERSONS W ITH INCOMPLETE SPINAL CORD INJURY WHILE WALKING WITH AN ANKLE FOOT ORTHOSIS............................................................................... 119 Introduction................................................................................................................... ........119 Specific Aims........................................................................................................................120 Methods................................................................................................................................121 Subject Selection...........................................................................................................121 Experimental Set-up......................................................................................................121 Subject Preparation........................................................................................................122 Procedure.......................................................................................................................123 Data Processing.............................................................................................................124 Data Analysis.................................................................................................................124 Results...................................................................................................................................125 Discussion.............................................................................................................................125 Clinical Implications..................................................................................................... 127 Limitations.................................................................................................................... .128 Conclusions...................................................................................................................128

PAGE 9

9 LIST OF REFERENCES.............................................................................................................138 BIOGRAPHICAL SKETCH.......................................................................................................158

PAGE 10

10 LIST OF TABLES Table page 2-1 Coefficient multiple correlations (CMC) refl ec ting stride to stride variability between and within days..................................................................................................................60 4-1 Demographics of the study participants............................................................................. 77 4-2 Hip, knee and ankle joint kinematic and ki netic data while walking with and without an ankle foot orthosis (A FO) during the stance-to-swing phase of the gait cycle............. 77 4-3 Hip, knee and ankle joint kinematic kinematic and kinetic data while walking with and without an ankle foot orthosis (A FO) during swing-to-stance phase of the gait cycle.......................................................................................................................... .........78 4-4 Average interlimb temporal and spatial da ta while walking with and without an ankle foot orthosis (AFO). ...........................................................................................................78 5-1 Participant demographics of individua ls with ISCI and control subjects. ......................... 95 5-2 Hip, knee and ankle joint kinematic and kinetic data during the stance-to-swing phase of the gait cycle w hile walking with and without an ankle foot orthosis (AFO) in individuals with incomp lete spinal cord injury.............................................................. 96 5-3 Hip, knee and ankle joint kinematic and kinetic data during th e swing-to-stance phase of the gait cycle while walking with and without an ankle foot orthosis (AFO) in individuals with incomp lete spinal cord injury.............................................................. 97 5-4 Kinematic and kinetic data during the stan ce-to-swing phase of the gait cycle at the hip, knee and ankle joints while walking w ithout an ankle foot orthosis (A FO) in individuals with incomplete spinal cord injury compared to their matched, noninjured controls..................................................................................................................98 5-5 Kinematic and kinetic data during the swi ng-to-stance phase of the gait cycle at the hip, knee and ankle joints while walking w ithout an ankle foot orthosis (A FO) in individuals with incomplete spinal cord injury compared to their matched, noninjured controls..................................................................................................................99 5-6 Changes in the hip, knee and ankle joint kinem atics and kinetics during the stanceto-swing phase of the gait cycle while walki ng with an ankle foot orthosis (AFO) in individuals with incomplete spinal cord injury compared to their matched, noninjured controls................................................................................................................100 5-7 Kinematic and kinetic changes at the hi p, knee and ankle joints during the swing-tostance ph ase of the gait cycle while walking without an ankle foot orthosis (AFO) in individuals with incomplete spinal cord injury compared to their matched, noninjured controls................................................................................................................101

PAGE 11

11 5-8 Average interlimb temporal and spatial da ta while walking with and without an ankle foot orthosis (AFO) in in dividuals with incom plete spinal cord injury........................... 102 6-1 Demographics of non-injured par ticipants recruite d for the study. ................................. 118 7-1 Participant demographics................................................................................................. 135 7-2 Ipsilateral H/M ratio with and without the AFO.............................................................. 135 7-3 Contralateral H/M ratio with and without the AFO. ........................................................136 7-4 Normalized soleus EMG amplitude with and without AFO ipsilaterally. ....................... 136 7-5 Normalized TA EMG amplitude wi th and without AFO ipsilaterally. ...........................136 7-6 Normalized soleus EMG amplitude w ith and without AFO contralaterally. ...................137 7-7 Normalized TA EMG amplitude with and without AFO contralaterally. ....................... 137

PAGE 12

12 LIST OF FIGURES Figure page 2-1 Reduced animal preparation showing the imp act of transection at different levels of the nervous system ............................................................................................................. 56 2-2 Stance-to-swing and swing-to-s tance phases of the gait cycle. ......................................... 57 2-4 Neural control of locomotion in an in tact nervous system (A) and compromised nervous system (B)............................................................................................................58 2-5 Electrical stimulus (shown here by the grey ellipse) applied to the mixed nerve conducts the stim uli orthodromically in th e motor and sensory axons to evoke the Mwave and the H-reflex respectively.................................................................................... 59 4-1 Participant with safety harness walking on an instrum ented treadmill.............................. 72 4-2 Ipsilateral and contralate ral average joint angles with and without the ankle foot orthosis (AFO) ipsilaterally.. ............................................................................................. 73 4-3 Stick figure representing the changes in indiv idual joint motion in the stance-toswing and swing-to-stance phase of the ga it cycle with and without the ankle foot orthosis (AFO)...................................................................................................................74 4-4 Ipsilateral and contralateral average join t powers during the stan ce-to-swing phase of the gait cycle with and w ithout the ankle foot orthos is (AFO) ipsilaterally.. ....................75 4-5 Ipsilateral and contralateral vertical and horizontal (AP) ground reaction forces (GRF) during the swing-to-sta nce phase of the gait cycle with and without the ankle foot orthosis (AFO) ipsilaterally.. ...................................................................................... 76 5-1 Ipsilateral and contralate ral averag e joint angles during the swing-to-stance and stance-to-swing phase of the gait cycle w ith and without the a nkle foot orthosis (AFO) ipsilaterally............................................................................................................ .89 5-2 Ipsilateral and contrala teral averag e joint powers during the swing-to-stance and stance-to-swing phase of the gait cycle w ith and without the a nkle foot orthosis (AFO) ipsilaterally............................................................................................................ .90 5-3 Ipsilateral and contralateral vertical and horizontal (AP) ground reaction forces (GRF) during the swing-to-sta nce phase of the gait cycle with and without the ankle foot orthosis (AFO) ipsilaterally.. ...................................................................................... 91 5-4 Ipsilateral hip extension values with and without the AFO during the stance-to-swing phase of the gait cycle for spin al cord injured individuals and their m atched controls..... 92

PAGE 13

13 5-5 Step length while walking with and without the AFO in individuals with incom plete spinal cord injury............................................................................................................. ..93 5-6 Double limb support time while walking w ith and without the PAFO in individuals with incom plete spinal cord injury.................................................................................... 94 6-1 Experimental design for the examination of changes in soleus H-ref lex amplitude in healthy, non-injured individuals while wa lking with and without an ankle foot orthosis (AFO).................................................................................................................113 6-2 Ipsilateral raw soleus H-reflex data w hile walking with and without AFO at 300ms from heel strike (HS) and toe-off (TO)............................................................................ 114 6-3 Contralateral raw soleus H-reflex data while walking with and without AFO at 300m s from heel strike (HS) and toe-off (TO)................................................................ 114 6-4 Ipsilateral mean H-reflex amplitudes with and without AFO norm alized to M-max in each phase of the gait cycle............................................................................................. 115 6-5 Contralateral mean H-reflex amplitudes with and without AFO nor malized to M-max in each phase of the gait cycle......................................................................................... 115 6-6 Ipsilateral [A] and contralateral [B] M-m ax amplitude with and without the AFO across the gait cycle......................................................................................................... 116 6-7 Ipsilateral [A] and contralateral [B] act ual M wave am plitude used to evoke the soleus H-reflex with and without the AFO across the gait cycle..................................... 117 7-1 Experimental design for testing the effect of walking with and w ithout an ankle foot orthosis (AFO) in individuals with incom plete spinal cord injury (ISCI) at their selfselected (SS) walking speed............................................................................................. 130 7-2 Average H/M ratio values with and without an AFO in m id-stance and mid-swing phase of walking relative to static standing in the ipsilateral limb.................................. 131 7-3 Average H/M ratio values with and wit hout AFO in m id-stance and mid-swing phase of walking relative to static st anding in the contralateral limb........................................ 131 7-4 Ipsilateral and contralatera l raw soleus H-reflex data wh ile walking with and without AFO during m id-stance (MSt) and mid-swing (MSw) phase of the gait cycle............... 132 7-5 Ipsilateral [A] and contralateral [B] Mm ax amplitude with and without the AFO in the mid-stance and mid-swing phase of walking............................................................. 133 7-6 Ipsilateral [A] and contralateral [B] act ual M wave am plitude used to evoke the soleus H-reflex with and without the AFO across the gait cycle..................................... 134

PAGE 14

14 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy NEUROMECHANICAL AND NE UROPHYSIOLOGICAL EXAM INATION OF WALKING WITH AN ANKLE FOOT ORTHOSIS IN NO N-INJURED INDIVIDUALS AND PERSONS WITH INCOMPLETE SPINAL CORD INJURY By Preeti Mohandas Nair May 2008 Chair: Andrea L. Behrman Major: Rehabilitation Science Clinicians often use orthotic devices to compensate for walking related impairments after incomplete spinal cord injury (ISCI). Orthotic de vices such as an ankle foot orthosis (AFO) are commonly used to stabilize the ankle joint and aid toe clearance during walking. Compensatory stepping achieved with an AFO has led therapis ts to assume that such devices could be integrated in newer, neurobiologi cally driven, recovery-based in terventions such as locomotor training (LT) for individuals with ISCI. In spite of the appeal of such compensatory strategies, their use during LT is still controversial. This is due to the lack of information about the possible effect of the device in optimiz ing or hindering afferent input from lower limb motion; joint, muscle and cutaneous receptors fundamental to the training. After ISCI, pattern generating neural network within the spinal cord increases its reliance on motion-rela ted afferent input from these receptors for maintaining locomotor contro l. Limiting ankle excursion with an AFO may alter the interconnected limb joint assembly specifi c to walking and in turn influence the afferent information critical for stepping. Our study explored the therapeutic use of such devices from a walking recovery based paradigm.

PAGE 15

15 The aim of this project was to investig ate the mechanical and neurophysiological implications of the use of an AFO during steppi ng in non-injured individuals and persons with ISCI. Specifically, we examined the effect of wear ing a posterior leaf spri ng ankle foot orthosis (PAFO) on transition phase joint kinematics and ki netics and soleus H-re flex modulation during walking. In the first experiment, we examined the transition phase mechanics with and without a PAFO in healthy, non-injured individuals. Our st udy identified and measured the changes that occurred in normal joint kinematics and kinetics as a result of wearing a PAFO. The results suggested that proximal hip ex tension; crucial for the transi tion from stance-to-swing and the rate of loading during the swi ngto-stance phase were signifi cantly decreased. In the second experiment, we compared transition phase mech anics observed while walking with and without the PAFO in individuals with IS CI to normal mechanics. The co mparison assessed the effect of the PAFO on pre-existing steppi ng related deficits in individual s with ISCI and also measured deviance or likeness of the change observed in these individuals from normal. The results suggested that the use of a PAFO decreased hip extension thereby impacting the provision of at least one critical affe rent input key to the restoration of walking. In the third experiment, soleus H-reflexes were compared in non-inju red individuals while walking with and without the PAFO in ten different phases of the gait cycle. The result showed that walking with the PAFO did not affect soleus H -reflex excitability in these individuals. In the fourth and final experiment, sole us H-reflexes were compared in the mid-stance and mid-swing phase in individuals with ISCI, while walking with and without th e PAFO. A significant increase in the soleus H-reflex amplitude was observed in the mid-swing phase of walking. Our findings suggest that the PAFO increased afferent inflow a nd modulated reflex activity. However, increase in afferent input in the mid-swing phase of the gait cycle may not be favorable to

PAGE 16

16 retraining the task of walking. In summary, our results suggest that walking with a minimally restrictive PAFO alters transi tion phase mechanics and soleus H-reflex modulation during midswing phase of walking. Therefore, during LT, us e of a compensatory PAFO to achieve stepping may not coincide with the principles of training.

PAGE 17

17 CHAPTER 1 NEUROMECHANICAL AND NE UROP HYSIOLOGICAL EXAM INATION OF WALKING WITH AN ANKLE FOOT ORTHOSIS IN HE ALTHY, NON-INJURE D INDIVIDUALS AND PERSONS WITH INCOMPLETE SPINAL CORD INJURY Walking is a fundamental motor function of hum an beings. It consists of three neurally controlled and coordinated tasks th at allow 1) generation of a basi c reciprocal, stepping pattern required for propulsion of the body 2) maintenance of equilibrium during propulsion and 3) adaptability of the walking pattern to the e nvironment and to the behavioral goals of the individual.1-3 Humans can move around from one place to another, maintain an upright posture, interact with the environment a nd perform a flurry of activities characteristic to human nature due to their ability to walk. Although walking is an essentia l element in daily living, its importance is usually only recognize d when it is impaired or lost. Spinal cord injury (SCI) is a debilitating condition resulti ng in walking impairment or inability secondary to deficits in voluntary strength (of the limbs and trunk) and sensation.4,5 Although all persons with SCI express a desire to wa lk only twenty five to thirty three percent of these individuals regain the ability to do so.6-9 Persons with incomplete spinal cord injury (ISCI) have a greater potential for walking recovery comp ared to individuals with complete injury due to some sparing of motor and sensory function below the level of the lesion.10 However, walking in persons with ISCI may be slow with asymmetr ical steps, flexed posture and impaired balance and adaptability. Current rehabilitation strategies after SCI are based on the assumption that deficits due to SCI are irremediable from surgi cal, medical, or therapeutic means.4,11 In particular, the spinal cord, being a hard-wired conduit of information from supraspinal structures to the muscles is viewed post-injury as irreparable and th e deficits permanent and irreversible.11-13 To compensate for irremediable deficits post-SCI, therapists em ploy braces, assistive devices, and wheelchairs to

PAGE 18

18 teach new behavioral strategies to achieve mobility.4,14-16 Assistive devices, such as canes, crutches, and walkers provide upright suppor t through upper extremity weight bearing to compensate for lower limb and trunk muscle weakness. Likewise, si ngle joint (ankle foot orthosis-AFO) or multi joint (knee ankle foot orthosis-KAFO) are used to stabilize joints and aid stepping.16 For example, posterior leaf spring ankle foot orthosis (PAFO) is often used to compensate inadequate toe clearance a nd loss of heel strike during stepping.16,17 In contrast to the assumption of ceased neurological function below the level of injury, neuroscientists examining the neurobiological co ntrol of walking in both animals and humans have provided convincing evidence for recovery.18-26 Evidence suggests that the adult mammalian spinal cord is plastic and is known to reorganize after injury when provided with the appropriate stimulus that is intense, task-specific and repetitive in nature.3,22,27-29 Reciprocal stepping, for example, defined as the repeti tious, mechanical sequence of limb motion and weight shift such that, one limb maintains cont act with the ground and supports the body while the other limb swings forward is facilitated by a host of afferent inputs that modulate the transition from support-to-swing and vice-versa.30-32,33 Afferent inputs for the neurobiological contro l of stepping involve motion related changes in joint position, muscular force and limb load se nsed by joint, muscle and cutaneous receptors in the lower extremity.34-37 Evidence from animal and human st udies indicate that terminal hip extension and unloading are critic al sensory inputs required for the afferent initiation of the transition from support-to-swing.31,34,38 Limiting hip extension and/or transfer of weight on either limb delays the transition from support-to-s wing and swing-to-suppor t and limits forward progression of the body.

PAGE 19

19 Evolving rehabilitation approaches such as locomo tor training have effectively integrated the provision of afferent input required for the neural control of steppi ng with repeated task specific practice to retrain the ability to step in individuals with ISCI.25,37,39,40 Using a treadmill, body weight support and manual assistance the traini ng facilitates repeated cueing of critical motion related inputs including hip extension and limb unloa ding required for stepping.24-26,41-43 The repeated volley of motion re lated input via the training can drive the patter n generating body within the spinal cord, induce reor ganization of the nervous system and facilitate th e recovery of walking in persons with SCI having limited access to supraspinal input.25,26,44 Therefore, this training strategy personifies the in trinsic repair potential of the spinal cord which can be tapped into using appropriate physical rehabilitation strategies to facilitate recovery. Ironically, addition of orthotic devices as training variab les to the physio logical-based training may hinder rather than facilitate recovery.24 Single joint, rigid orthotic devices such as AFOs might restrict the range of excursion of th e distal and proximal linke d joints and alter the afferent information related to joint position and load pivotal for stepping.24 Furthermore, the use of an assistive/orthotic device might change the basic stepping pattern thereby modifying the afferent information utilized for stepping. In addition, the acquisition of a new motor skill such as walking with an orthotic device could induce a new pattern of skill-dependent plasticity that could impact their ability to reacquire normal walking potential.45 Thus the role of such compensatory strategies in promoting use of the intrinsic mechanisms for stepping lack mechanical and neurophysiological evidence justif ying their use and need to be investigated.24,46 The overall purpose of this study was ther efore to examine the mechanical and neurophysiological effect of walking with one su ch single joint, range limiting orthotic device the AFO. In this four part project, I examined the effect of wearing an AFO on 1) immediate

PAGE 20

20 phase specific H-reflex modula tion and 2) walking mechanics in healthy, non-injured individuals and 3) immediate phase specific H-reflex modula tion and 4) walking mechanics in persons with ISCI. In these experiments, healthy, non-injured individuals and pers ons with ISCI walked on an instrumented treadmill at speeds close to normal walking speeds. Kinematic, kinetic and H-reflex data was collected while these individuals walked with and without wearing an AFO. Specifically, the mechanical changes in joint position, joint powers, vertical and horizontal ground reaction forces in the transition phases were examined in these experiments accompanied by the neurophysiological changes in the soleus H-reflex amplitude. Interpretation of the mechanical and neurophysiological data served as a critical first st ep in interpreting how restricting lower extremity single joint excurs ion while walking with a AFO modified the consequent sensory information i.e. hip joint pos ition and load required for reciprocal stepping. In summary, these studies enhanced our unde rstanding of the impact of altered walking mechanics associated with these compensatory devices on the neurally controlled task of reciprocal stepping in healthy, non -injured individuals and persons with SCI. The findings of this study will be useful in 1) clin ical decision-making for the use of such devices in physiologicalbased training interventions and 2) designing neuromechanically compatible assistive and orthotic devices. The following literature review is composed of s ections that will serve to orient the reader to the foundation principles underlying the purpose of this project. An ov erview of SCI related walking impairment and the use of orthotic devi ces to compensate for impairment is described first. Juxtaposed are the neurobiological control of walking and the subtask of stepping highlighting the mechanical and neural character istics of stepping. The newer evolving recovery

PAGE 21

21 based intervention called locomo tor training that has stemmed fr om our knowledge of the neural control of walking and plasticity of the nervous system is described next. This section is followed by the gap in knowledge pertaining to the use of an AFO from a neural control standpoint. Methodological consider ations for investigation of an AFO and interpretation of results in relation to the neural control of stepping is described that will provide the basis for the experimental paradigm used in this project. The clinical and scientific relevance of the studies are discussed in the final section.

PAGE 22

22 CHAPTER 2 LITERATURE REVIEW Overview of Human Spinal Cord Injury: Consequence and Rehabilitation Introduction to the Problem Spinal co rd injury (SCI), an in jury to the neural elements with in the spinal cord, results in a multitude of dysfunction, loss of sensation and motor function being the most profound.47,48 Injury could occur as a result of motor vehicu lar accidents, diving accidents that fractures, dislocates or compresses the vertebrae protecting it.7,49 or may also result from a gunshot50 or knife wound51 that penetrates and cuts the cord. Add itionally, secondary damage usually occurs with a traumatic injury as a result of bleeding, swelling and inflammatory processes that compress the cord.52 Non-traumatic injuries to the cord occur as a result of tumors, vascular problems, spina bifida and several other conditions.53-55 Although there is a signif icant rate of mortality associated with injury, survival after SCI has improved considerably because of efficient critical care and improve d urinary rehabilitation and respiratory management.52,56 About 253,000 people currently in the US live with SCI and there are an additional 11,000 new cases every year.57 The current ten-year su rvival rate of spinal cord injured patients is approximately 86% of normal.57 Consequence of Injury The effects of SCI vary accord ing to the level and type (complete or incomplete) of injury.58,59 In a complete injury, there is bilateral, total sensory and motor loss below the level of injury. A person with an incomplete SCI (ISCI) re tains some sensation below the level of injury. Incomplete injuries are variable, and a person with such an injury may have patchy motor involvement such as he or she might be able to move one limb more than another, may be able to feel parts of the body that cannot be moved, or may have functioning on one side of the body

PAGE 23

23 more than the other. Depending on the level of injury, paralysis can involve all four extremities a condition called quadriplegia or tetraplegia occurri ng as a result of cervica l injuries or only the trunk and lower body a condition called paraplegia occurring as a resu lt of injury at the thoracic level or below. The American Spinal Injury Association (ASIA) impairment scale (AIS) is used to classify the level and severity of injury in relati on to the loss of sensat ion and motor function.58,60 The scale consists of five categories that classify sensory and motor function. ASIA A is defined as a complete injury with no motor or sensory functio n preserved in the sacral segments S4-S5. ASIA B, C and D are all incomplete injuries but cl assification varies base d on the level of motor involvement below the level of the lesion. ASIA B is defined as an incomplete injury with some sensory but no motor function preserved below the le vel of injury includi ng the sacral segments S4-S5. To be classified as an ASIA C more than half of the muscles are graded less than 3/5 voluntary strength. ASIA D is defined as an incomplete injury with at least half of the muscles graded more than 3/5. A person with SCI is classified as an ASIA E if he/she has no neurological deficits that are detectable on a neurological examination of this type. Walking Potential after SCI W alking ability after SCI has been defined in the literature in several different ways. For example, studies have characteri zed it based on the ability to am bulate upright fifty feet without assistance61 or the ability to walk in th e community or in the household8,9 or the ability to walk reciprocally for at least two hundred feet with/without orthotics or assistive devices62 or walking functional independence measure of > 3/7.63 The ability to walk ranks as one of the top five priorities of individuals with SC I based on the level and severity of injury but interestingly only 25-33% of these individuals are able to do so.52,6,64 Key predictors of ambulation potential after SCI include: ASIA score D or E at admission,8,9 age,61,64 ASIA lower extremity motor scores

PAGE 24

24 greater than or equal to 10 by one month,65 manual muscle testing score in the quadriceps greater than 2/566 and sparing of pin prick sensa tion below the level of injury.62 Based on these factors, a significant number of persons who regain the ability to walk are persons with ISCI compared to persons with complete SCI.10,61 Maynard et al. (1979), for exampl e, reported that of 123 patients with incomplete sensory deficits 72 hours afte r injury, 47% were ambulatory and 87% of the patients with incomplete motor le sions were walking at one year.67 Although upright mobility may be spared or achie ved with assistance, walking is typically impaired in persons with ISCI as a result of varying levels of muscular paralysis, sensory deficits, spasticity a nd poor trunk control.68-70 Gait in an individual with ISCI is often characterized by one or combination of the fo llowing deviations (i) inadequate active hip extension during stance; (ii) li mited hip flexion; (iii) limited knee flexion; (iv) excess ankle plantar flexion during swing; and (v ) impaired initial foot contact.5 Consequently, these individuals are often seen taking slow, asymmetrical and uncoordina ted steps over a wide base of support and having limited adaptability to the environment.5,71 Rehabilitation of Individuals with ISCI The International Classif icat ion of Functioning, Disability and Health (ICF) is the framework developed by the World Health Orga nization (WHO) to describe functioning and disability at both the individual and population levels.72 Conventional physical therapy interventions for improving walking function in persons with ISCI targets two main domains of the model namely, body function/ structures and activities. At the level of body function and structures, the level and severity of injury using the ASIA scores is assessed and interventions maximizing residual muscle strength and endurance in muscles that can be voluntarily activated above and below the lesion are implemented.14,15 Similarly, interventions at the activity level emphasize the use of assistive and orthotic devices to improve ambulation potential14,16 and teach

PAGE 25

25 new strategies for upright mobility.14,15,73 Patient performance at this level however is a critical indicator of type and extent of orthotic supports and the patient's tolerance to ambulation. Orthotic Devices: Rationale for Use and Prescription Orthotic dev ices range from simple single join t braces such as the ankle foot orthosis (AFO) to multi joint ortho tics such as reciprocating gait orthosis (RGOs).15,74,75 The prescription of these devices varies based on several factor s including the ASIA impairment score of motor complete or incomplete made by the clinicia n to the needs and de sires of the client.74 The goals of prescribing orthotic devices for walking are to support the paralyzed or weakened musculoskeletal structure, add stability to joints, improve mobility, correct alignment and improve overall functional independence.16 Posterior leaf spring ankl e foot orthosis (PAFOs) are usually prescribed for higher functioning individuals with ISCI to provide support for weakened musculature around the ankle joint.16,76 Since it supports one single jo int it is considered least cumbersome and called a device of minimal assistance. The guiding principles for recommendation are to control the ankle joint by limiting excursion range, provide safe joint mechanics, prevent toe drag during the stance-to-swing transition, minimize the risk of falls and enhance the ability to walk faster and efficiently.16,17 The current rationale for use of orthotic devi ces for persons with ISCI has stemmed from the hierarchical model of motor control th at has been well accepted by rehabilitation practitioners and continues to serve as the basis for rehabilitation to date.11,12 The model portrays a hard-wired, immutable central nervous system (CNS) that cont rols all voluntary movements by sending commands from the cerebral cortex to the periphery. The spinal cord serves simply as a cable between the brain and the peripheral musculature receiving s timuli from the periphery and relaying cortical commands to th e periphery. Therefore if the spin al cord is injured, the damage is considered irreparable, non-malleable and permanent.12,77 Conventional rehabilitation

PAGE 26

26 therefore typically consists of a compensatory approach to deal with walking impairments in persons after ISCI.78 The approach utilizes the use of comp ensatory orthotic devices that utilize other spared abilities to accomplish the task or m odify the task and/or the environment to make it easier for a person to accomplish the goal.79 Neurobiological Control of Walking The prevailing assum ption that neural recovery is not possible following SCI has led to the aggrandizement of orthotic devices th at compensate for walking impairments.80-82 Compelling evidence from neuroscience examining the neural control of walking however contradicts this assumption.31,37,83-86 Neuroscientists have investigated the role of the nervous system, particularly the spinal cord to ad apt and reorganize after complete transections at the level of the cord.20,87,88 The ability of the spinal cord to respond to peripheral sensory input, generate and modulate rhythmic activity in the lower limbs and reorganize af ter injury make it a viable substrate for intervention.31,84,89,90 The subsequent sections expand on the role of the spinal cord in the control of walking and its plasticity after injury that have led to a proposed paradigm shift in SCI rehabilitation from compensati on-based to a recovery-based model.24 Conceptual models portray walking as three ne urally controlled and coordinated tasks: 1) Generation of a basic reciproc al, stepping pattern required for propulsion of the body, 2) Maintenance of equilibrium during propulsion and 3) Adaptability of the walking pattern to the environment and to the behavi oral goals of the individual.1,3 However, convergence and processing of multiple afferent inputs occurs at every level of the nervous system to bring about the smooth, patterned orchestrati on of several joints and muscular synergies that characterize walking.91-93 While walking, a single alpha motor neuron might receive as many as ten thousand inputs.94 Determining which input has a greater re lative influence on sculpting the locomotor pattern has been a challenging ta sk in the complex human nervous system. Neuroscientists have

PAGE 27

27 therefore utilized reduced, non-hum an, vertebrate preparations su ch as cat models to study the influence of exclusive afferent inputs on locomo tor behavior and extra polated their research findings to human control of locomotion.31,86,95 The premise for extrapolation being the locomotor framework which is remarkably similar throughout the vertebrate phylum, in spite of the form of locomotion which is species specific.84,96 For example, the basic principle of organizi ng rhythmic stepping behavior is similar in humans compared to other non-human vertebrates.95,97 In a spinalized cat model the reflex or automatic behavior associated with walking i.e. reciprocal stepping doe s not require control by cerebral cortex rather it is c ontrolled by subcortical and spinal centers which are subject to cortical intervention.98 Reciprocal stepping is ge nerated at the level of th e spinal cord as long as the weight is supported and the ground is m oving under their feet. Assemblies of premotor interneurons in the spinal cord are synaptic ally interconnected with each other and with motorneuronal pools that are capab le of sustaining alternating movements required for walking.99 These networks of neurons and interneurons are called central pattern generators (CPGs).100 The central pattern generators simplif y the control of locomotion by harnessing the large degrees of freedom and provide the basic framework for walking.99,101,102 Sensory input driven from the periphery is known to control th e rhythm of walking by making the required phase transitions, shaping the pattern of activit y and reinforcing ongoing activity. Sensory information from muscle spindles which are sensitiv e to changes in the muscle length,103 golgi tendon organs that are responsive to muscle tension104,105 and flexor reflex a fferents involving the mechanoreceptors, cutaneous afferents and nocicep tors is processed at the level of the CPG.104,105 When the task is consistent and unaltered, lik e during stable state walking, the spinal cord

PAGE 28

28 demonstrates remarkable autonomy in producing the reflexive stepping pattern without much cortical control. Neural descending pathways from the brain to the cord are predominantly involved in finetuning this pattern. For example, the corticospinal tracts influence the locomotor performance by acting monosynaptically or oligosynaptically on the alpha-motorneuron s or indirectly via connections on the CPG.99 Thus, selective muscle control re quired for fine tuning and/ or modulation of locomotor synergy i.e. speed of locomotion could be achieved. Also, the excitability of the CPGs is governed strongly by locomotor centers in th e midbrain and brainstem i.e. mesencephalic, pontine or subthalamic locomotor regions106-108 that are under limbic and cortical control and that di ctate purposeful locomotion su ch as starting or stopping. Similarly, higher brain centers su ch as the cortex, cerebellum98 and basal ganglia109 are responsible for integrating afferent information from different sensory sources with cortical motor commands. Integration of the sensory input with the motor output facilitates control of the other two subtasks i.e. maintaining balan ce and adapting to the environment by engaging corrective and reflexive postura l mechanisms required for walking. When the cerebellum, the brainstem and the spinal cord are spared in cats (decerebrate preparation), it is seen that the animal is able to generate rhythmic activity, support body weight and prope l itself. However, the other two subtasks for successful locomotion i.e. dynamic balance and adaptability to the environment are noticeably deficient2,102 (Figure 2-1). Dynamic balance in the decerebrate cat require s integration of sensory systems: visual, kinesthetic and vestibular system.97 Walking in a cluttered envi ronment for example requires integration of information from visual, somatosensory and vestibular inputs to maintain balance and perform the task of walking successfully.110 The reticulospinal, rubr ospinal, vestibulospinal

PAGE 29

29 and corticospinal tracts111,112 are descending spinal tracts that project motor commands for fine postural adjustment to the spinal cord after pr ocessing afferent information. Corrective postural responses are then integrated and adjusted to the current st ate or phase of stepping by the CPG.113 Lack of cortical control above the level of transection also le ads to a loss of adaptability to the environment and to the persons behavioral goals (Figure 2-1). The motor cortex and the basal ganglia are important for "s killed" locomotion in which the feet must be guided to establish firm contact with narrowly spec ified points in the environment.97 Studies on decorticate cats showed that although the loss of the cortex has minimal impact on the locomotor process, the context in which the locomotor move ments are performed is affected.84,114 For example, decorticate cats exhibit limited range of options in locomotor movements and are hyperactive to stimuli that would tend to elicit a minimal respons e. Likewise, ablation of the caudate nucleus of the basal ganglia in cats results in the animal following anything that moves termed "compulsory approach syndrome"; while diencephalic cats (whose thalamus and hypothalamus have been removed) demonstrate "obstinate progression i.e. walking into obstacles and not attending to environmental stimuli.114 Therefore cortical centers and the basal ganglia are important for adaptive control of movement to the environment. While the different levels of control exerte d by the nervous system are important for the coordination of the overall task of walking, the pr oposed studies focus on one of the subtasks of walking i.e. stepping. The next two sections elaborate on characterizi ng the stepping pattern mechanically and describing the neur al assembly required for stepping.

PAGE 30

30 Stepping Mechanics Basic Instrumentation and Terminology Stepping is a m otor task that obeys laws governing static and dynamic bodies and can be quantified using our knowledge of basic mechan ics. Kinematics is the branch of mechanics dealing with the motion of body segments without being concerned with the forces that cause the motion.115 Kinematic analysis, using automated motion analysis systems measures positions, angles, velocities, and accelerations of body segm ents and joints during motion. Joint angle (also called inter-segmental angle) is defined as the a ngle between the two segmen ts on either side of the joint, usually measured in degrees.116,117 For the interconnected chain of segments involving the ankle, knee and hip joints of the lower limbs, joint angles are particularly useful in determining the relative motion of one joint with respect to the other. However, these measurements only describe the motion performed and are limited in what they can tell us about the cause of the motion. Kinetic analysis on the other hand measures forces acting at a particular joint, segment or body as a whole that cause the specific walking pattern.116,117 Forces in walking can be internal such as muscle activity, ligamentous constraint or external such as ground-reaction forces created from external loads. During walking concentric and ec centric contraction of the limb musculature around a joint results in the generation and absorption of mechanical energy necessary to accomplish the movement that we observe and is referred to as joint power. Joint power therefore is the rate at which energy is eith er generated or absorbed and is the product of a joint moment and the joint angular velocity.115 The joint moment (also known as torque or moment of force) being the rota tional potential of the forces ac ting on a joint. The joint moment usually is calculated around a joint center. The units used to express moments or torques are Newton-meters (N-m) and for research purposes usually are normalized to the subject's body

PAGE 31

31 mass. Joint power is generated when the moment and the angular velocity are in the same direction and said to be absorbed when they ar e in opposite directions. The units used to measure joint power are Watts (W). During stepping, ground reaction forces (GRF s) produced as a result of body weight transferring onto and moving across the supporting foot constitutes the external forces acting on the body. GRF as the name denotes, is basically the reaction to the force the body exerts on the ground. GRF is comprised of three co mponents: 1) vertical force, 2) fore-aft shear and 3) mediallateral shear.115-117 Information on these forces is obtained from a force platform or force plate, which is a transducer set into the floor to meas ure the forces and torques applied by the foot to the ground. These devices provide a quantified measure of the three compone nts of the resultant GRF vector. Measurement of vertical ground reaction forces produced during walking provides information on the load imposed on the joints during weight bearing.33 Normally this force is represented as two peaks with a valley in between The first peak occurs in response to weight acceptance while loading the limb. The second peak is caused by acceleration of the body forward. Fore-aft or horizontal ground reacti on forces measure propulsion as the body weight shifts from one lower limb to the other. Hori zontal forces have a negative and a positive component. The negative component is referred to as the braking force and is indicative of a backward horizontal friction force between the floor and the foot to prevent the foot from sliding forward. The positive component is referred to as the propulsive force and is indicative of the foot pushing back on the floor to propel the body forward. The medio-lateral forces measure stability of the body during walking.117 The exchange of body weight from one limb to the other

PAGE 32

32 generate the medio-lateral shear forces. Peak me dial shear occurs while loading the limb and peak lateral shear occurs while unloading the limb. Characterization of Stepping Pattern Reciprocal stepping can ther efore be functionally charac terized using biom echanical measures. Stepping is defined as the repetitious sequence of limb motion such that, one limb maintains contact with the ground and supports the body while the other limb advances forward.33 The supporting or weight b earing phase is termed as th e stance phase and the limb is referred to as the "stance limb" while the forwar d stepping phase is termed as the swing phase and the limb is referred to as the "swing limb". A single sequence of support (stance) and advancement (swing) executed by one limb is called a gait cycle.33 The transition from stance-to-swing is charac terized by two phases; terminal stance phase and pre-swing phase. The terminal stance phase be gins with the heel ri se of the supporting limb and continues till the other foot strikes the grou nd. The pre-swing phase begins with foot strike of the other limb and continues till toe-off of the supporting limb. The objective of these phases together is the progression of the body forward. Po wer is generated in this phase to propel the limb and body forward.118,119 The transition from stance-to-swi ng is kinematically characterized by extension of the hip joint, flexion at the knee joint and plantarflexion at the ankle joint33 (Figure 2-2). Kinetically this phase is characterized by a horizontal pr opulsive force to aid forward progression of the body. Similarly, the transition from swing-to-stance is characterized by two phases: initial contact and loading response.33 Initial contact begins when th e swinging limb strikes the floor. Loading response begins with initial floor contact and continues until the other foot is lifted for swing. The demand for immediate transfer of body we ight onto the limb as soon as it contacts the ground requires initial limb stability and shoc k absorption while simultaneously preserving

PAGE 33

33 the momentum of progression. Therefore the objec tive of these phases toge ther is to provide weight bearing stability and pres erve progression. Swing-to-stance is kinematically characterized by flexion of the hip joint, extension of the knee joint and plantarflexion of the ankle joint (Figure 2-2).33 Kinetically this phase is characterized by a horizontal br aking force to aid weight acceptance. The transition from stance-to-swing occurs be tween 30 to 60% of the gait cycle and the transition from swing-to-stance occurs between 0 to 10 % of the gait cycle.33 The temporal sequence of these transitions is the result of interactions between a tripartite neural system consisting of supraspinal, spinal and sensory input.95 Neural Assembly for Stepping Central Pattern Generator for Stepping Rhythm ic and reciprocal stepping can be triggered by descending supraspinal command, which delegate the motor coordination to specialized spinal circuitry for pa ttern generation called the CPG. Primarily, the concept of specialized spinal circuitry existed as far back as 1911, when Brown et al, showed that cats with a transected spinal cord and with cut dorsal roots still produced rhythmic alternating contrac tions in ankle flexors and extensors.120 This provided the basis of the concept of a spinal locomotor center that Brown re ferred to as the half-center model. One half of this center induced activ ity in flexors, the other in extensors. Much later, Grillner and Wallen100 coined and demonstrated the existence of CPGs as assemblies of premotor interneurons that were synaptically connected wi th each other and with the motorneuron pools and capable of creating an elaborate flexor and ex tensor synergy between different muscles of a limb required for locomoti on. Although the anatomical details of CPGs are known for a few cases only, the motor commands orig inate from the spinal cords of a variety of vertebrates.100 For example in cats, the nature of pattern generation is still uncertain because the

PAGE 34

34 exact connections of only a few interneurons su ch as Ia interneurons, Renshaw cells, ventral spinocerebellar cells and spinobulba r cells connecting to lateral re ticular nucleus are known with certainty.108 These neurons could be driv en by the CPG or could well be a part of the CPG itself. However, what is clear is the f act that no rhythmic input is re quired to activate these circuits.121 These circuits can function in vitro when isolat ed from the brain, as evidenced by locomotion in decerebrate cats122 and when isolated from the motor and sensory apparatus of the limbs.120 The rhythms can often be initiated by simple tonic (i.e. non-oscillating) elec tric or pharmacological stimulation. In humans, rhythmic, alternating electromyogr aphic activity of the lower limbs in the absence of supraspinal and moveme nt related afferent input has b een interpreted as evidence for central pattern generation.123 Such evidence was also provided in a study of six subjects with complete SCI, where researchers were able to induce rhythmic, alternating, locomotor-like EMG pattern on continuous epidural spinal cord stimulation.124 Therefore the CPG is considered the elem entary building block on which rhythmic movement is based. As soon as rhythmic move ment is initiated, feedback from the moving limbs, termed as "motion-related feedback" in this review, arrives at the spinal cord to inform the nervous system of the local conditions. This feedback assists in shaping the pattern of walking, reinforcing ongoing activity and cont rolling the phase transitions.38,125 An ensemble of motionrelated input arising from skin, joint recep tors, muscle spindle, golgi tendon organs, mechanoreceptors, nociceptors is believed to influence the pattern of stepping.91,108,126 Specifically, hip joint position sensed by muscle spindle and load sensed by the golgi tendon organs are two of the several motionrelated inputs contributing to the control of stepping.31,127

PAGE 35

35 Several animal and human studies examining th e neurobiological contro l of stepping have validated these findings. Sensory Drive Required for Stepping Sherrington was the first to propose that pr oprioceptors responding to hip extension are im portant for initiating swing.128 Grillner and Rossignol129 found that preventing the hip from attaining an extended position in chronic spinal cats inhibited th e generation of the flexor burst and hence the onset of the swi ng phase. The most direct evidence for this conclusion however came from vibrating the hip flexor muscle (iliopsoas) during stance which led to an earlier onset of swing in walking decerebrate cats.130 The receptors signaling hip extension were probably the primary and secondary endings of muscle spindles in hip flexor muscles (Group Ia afferents). Similarly, in humans, involuntary and alternat ing stepping-like movements were observed in an individual after incomplete SCI upon extending the hip in the supine position.131 Furthermore, hip walking movements (i.e. facilitating hip joint excursion with the knees fixed in an extended position) induced by a driven gait ort hosis (DGO) in individuals with complete SCI produced pattern of leg muscle EMG activity that corresponded to that normal stepping in healthy, non-injured individualssubjects.127 Researchers examining in fant stepping also support the role of hip extension positi on for the initiation of swing. Fr om the recorded hip motion and electromyographic data these scient ists concluded that the prefe rred hip position was always one directly opposite the direction of walking during infant stepping.38 It was thus suggested that the hip position is important in initia ting the stance-to-swing transition. Another important sensory input regulating the stance-to-swing transition is the extensor load relayed by the Golgi tendon organs (Group Ib afferents) in the ankle extensor muscles.31,132 During locomotor activity, electrical stimulati on of the group Ib afferents from the ankle extensor inhibits the generati on of flexor bursts and hence prolongs the duration of extensor

PAGE 36

36 activity. Duysens and Pearson (198 0) observed that gradually increasing load applied to the achilles tendon resulted in an increase in both the amplitude and duration of the rhythmic EMG bursts of ankle extensors.89 Similarly, cutaneous afferents innerv ating the skin of the foot (group II afferents) are also load monitors. Electrical pu lses applied to the foot pad innervated by the sural nerve were able to prolong the extensor burst in the stance phase in pre-mammilary cats preparations thereby providing evidence that lo ad-related cutaneous input from the foot can inhibit the CPG for the generation of flexion during swing.36,133 In humans, researchers found that unloading th e ankle extensors by a portable device in the stance phase of walking reduced the soleus EMG activity and the reduction was maintained even when transmission in Ia afferent s was blocked by local anesthesia This finding thus pointed to group Ib and/or group II afferents contributing to the extensor EMG activity in the stance phase.134 Harkema et al. (1997) observed that the amp litude of extensor muscle activation in the legs was directly related to the level of body weight loading on the legs during stepping of healthy, non-injured individualsand SCI subj ects during manually-assisted stepping on a treadmill.135 Dietz et al. (2002) also found that physiological locomotor-like leg movements alone (100% body unloading) gene rated by the application of the DGO on a treadmill are not sufficient to generate leg muscle activation in either healthy, non-injured individualssubjects or in subjects with complete para-/tetraplegia.127 In this study, leg movements in combination with loading of the legs led to appr opriate leg muscle activation. In summary, during stance phase load of the lower limb is detected by group I extensor muscle afferents and group II cutaneous afferents which activate the extensor half center (EHC) of the CPG. Extensor activity is reinforced dur ing the loading period of the stance phase. At the

PAGE 37

37 end of the stance phase, group Ia afferents of flexor muscles excite the flexor half center (FHC) which inhibits the EHC and thereby initiates the onset of the swing phase (Figure 2-3). The importance of motion-related sensory input for locomotor control is evident when descending supraspinal input is compromised or altere d such as after SCI. In contrast to an intact nervous system processing multiple sources of afferent input, after complete or incomplete SCI, the spinal circuitry does not become silent (Figur e 2-4 A & B). The circuitry instead adapts to its altered combination of inputs and predominantly utilizes motion-related afferent input to facilitate locomotor response.37 The weighted response of the spinal circuitry to ascending afferent input illustrates the high level of spinal automaticity for locomotor control. Therapeutic strategies that optimize motion-related afferent input to the spin al cord can therefore be utilized to regain locomotor control afte r SCI. For example, stepping can be initiated by shifting the body weight to one leg and moving the head and trunk so that the hip position of the contralateral leg is extended.41,42 Apart from influencing the stepping pattern, sensory input also modul ates spinal reflex behavior during stepping.136 Spinal reflexes are those in which the sensory stimuli arise from receptors in muscles, joints and skin, and in which the neural circuitry responsible for the motor response is entirely contained within the spinal cord. Spinal stretch reflexes are the simplest stimulus-response behaviors exhibited by the mamma lian nervous system. The stretch reflex as defined by Wolpaw is the initial, purely spinal, largely monosynaptic response to sudden muscle stretch that is accessibl e anatomically and physiologically.137 H-reflex is an electrical analogue of the spinal stretch reflex and is a commonly studied spinal reflex in human beings.138 It was originally described by Paul Hoffman in 1910 and is electrically elicited leading the stimulus to bypass the effect of the gamma motorneuron.139 Therefore, H-reflex is a valuable tool

PAGE 38

38 in assessing modulation of monosynaptic reflex activity in the sp inal cord and has been used since to assess the response of the nervous system to various neurologic conditions,140-142 application of therapeutic modalities,143,144 exercise training and motor task performance.145-147 H-reflex Elicitation and Modulation During Stepping Elicitation The technique used to evoke the H-reflex involves electrical s tim ulation of a mixed peripheral nerve.148 When a percutaneous stimulation of increasing intensity is applied to a mixed nerve, action potentials travel along affere nt portion of the reflex arc along the Ia sensory afferents, until they synapse on the alpha moto rneuron. The efferent portions of the reflex pathway results from action potentials generate d by the alpha motor neur on until they reach the myoneural junction and produce a twitch response wh ich is recorded by su rface electrodes on the muscle of interest (Figure 2-5A). In addition to the H-reflex, elec trical stimulation of the periph eral nerve also causes direct activation of the efferent fibres that conduct orthodromically to produce a response in the EMG known as a muscle response or M-wave.148,149 When the stimulus intens ity is really low only the Ia afferent fibres of the mixed nerve undergo de polarization leading to th e appearance of the Hreflex tracing on the EMG. As the stimulus intens ity is increased, more Ia afferent fibres are recruited, resulting in activation of more alpha motor neurons and increasing the amplitude of the H-reflex. Continuing to increase the intensity bey ond the point of elicitation of the H-reflex also results in direct stimulation of the motor axons and thereby production of the M-wave (Figure 25A). The M-wave on a recording is usually seen before the H-reflex due to the relatively short path that the action potentials need to travel. Compared to this the H-reflex is characterized as a latency response since it does not occur right after the a pplication of the stimulus (Figure 2-5B). The latency of the reflex response is a result of the length of the H-reflex pathway, which

PAGE 39

39 involves the afferent and efferent length of the path and also th e overall length of the limb. For example, in the soleus muscle, the H-refl ex tracing usually appears approximately 30 milliseconds after the delivery of the stimulus wh ereas the M-wave is usua lly seen after 6 to 9 milliseconds of stimulus application.138 The amplitudes of the H-reflex and M-wave bot h increase fairly linearly with increase in stimulation intensity until the maximum Hrefl ex (H-max) representing the fullest extent of reflex activation is reached.138 Increasing stimulus intensity further beyond H-max, maximum M response (M-max) representing the maxi mum muscle activation is reached.138,149 Therefore, by increasing the stimulation intensity a recruitment curve depicting, stimulation intensity sufficient to evoke a sequence of H-reflex, H-max, di sappearing H-reflex tracing, M-max and M-max plateau can be obtai ned (Figure 2-5C).138,149 Task-specific/ Phase-dependent modulation of the H-reflex Soleus H-ref lexes are the most widely assessed reflexes in locomotor studies.138 In healthy, non-injured individuals, soleus H-reflexes are strongly modulated during the gait cycle with the highest amplitude registered during the stance phase and the lowest amplitude recorded during the swing phase.146 H-reflexes are minimal at the time of h eel contact, rise to a maximum shortly after midstance, decrease rapidly at the time of toe-off and are minima l during swing in both young and older age groups.150 Therefore, soleus H-reflexes in healthy, non-injured individuals demonstrate phase-dependent modulation duri ng stepping. Along with the phase-dependent modulation of the reflex during th e step cycle, H-reflexes are also known to display considerable differences in modulation between different motor tasks.147,151 Both phase-dependent and taskdependent modulation of the reflex is critical to the optimal pe rformance of motor behaviors. Modulation of the H-reflex during stepping has been attributed to both, sensory input152,153 and to the higher central mechanisms controlling motion.154 Although the bulk of literature supports the

PAGE 40

40 primary role of central reflex modulation during walking, the secondary role of sensory inputs is critical as well. The role of sensory inputs is especially cr itical when the central control of reflex modulation is compromised for e.g. post SCI. As a result of injury to the spinal cord, phasedependent modulation is impaired and reflex amplitudes are higher than normal throughout the gait cycle.155,156 In such a scenario, the reflex muscle response is likely to be dependent on peripheral sensory inputs deficien t of central modulation. Several st udies have reported that the soleus H-reflexes can be modul ated by sensory inputs of periphe ral origin such as hip joint position, leg load, and cutaneous receptors in sole of the foot.134,157 Sensory input related to hip joint position or from hip joint proprioceptors for example is shown to markedly influence soleus H-reflexes during passive movement of the hip from flexion to extension phase in both healthy, non-injured individualsand persons with spinal cord injury.158,159 Similarly, mechanical loadi ng of the foot sole, ranging from 15 to 70 N is known to significantly inhibit soleus H-reflex amplitude in both seated he althy, non-injured individualsand persons with complete SCI.136 Although performed in non-locomoto r tasks, the above studies are suggestive of a possible alteration in H-reflex am plitude via peripheral se nsory inputs even after SCI. Thus post-SCI, provision of critical motion related sensory inputs may play a role in phasedependent H-reflex modulation and could be useful in optimizing task performance. Recovery of Walking after SCI Plasticity at the Level of the Spinal Cord The lifelong ability of the nervous system to reorganize ne ural pathways structurally or functionally in response to experi ence or activity termed as neuroplasticity was once solely considered to be a supraspinal phenomenon.77 Acquisition and maintenance of normal motor performance however involved skill-dependent plas ticity at multiple sites including the spinal

PAGE 41

41 cord.85,160 Early evidence of this phenomenon was seen as early as in 195 1 when Shurrager and Dykman reported that treadmill walking in spinalized cats improved with training.161 Locomotion was much better in cats exposed to treadmill training than in cats that received standard care after injury.90 The improvements in training seen after training persisted even after training was stopped. Similarly, De Leon et al perf ormed a series of experiments in spinalized cats either trained to stand or st ep on the treadmill. Their results revealed that cats trained to stand improved standing function and thos e trained to step improved in stepping.28,90 Locomotor ability was exactly reversed in th ese groups when they were retrai ned to perform the other task. These researchers also tested the effect of gl ycinergic inhibitor, stry chnine in spinalized cats that were trained to either walk or stand on the treadmill.162 They observed that locomotion improved in cats trained to stand when strych nine reduced glycinergi c inhibition and had no effect on the cats trained to step. These findings suggested that glycinergic inhibition in the spinal cord interfered with stepping ability in spinal an imals and locomotor training improved stepping ability by reducing the levels of inhibi tion on the spinal networks. Skill-dependent training can therefore markedly change or m odify physiological and biochemical state of multiple neurotransmitter-modulator systems in the spinal cord and enhance locomotor recovery.162-164 Studies of motor unit prope rties after spinal cord transection, with and without training, have also provided supplemental evidence by indicating that training induced improvements in walking and standing are not at tributable to peripheral changes in muscle strength.165 Several studies in humans have also provide d compelling evidence for activity dependent plasticity of the spinal cord caudal to injury. Similar to cats, treadmill training using a body weight support showed significant improvements in walking behavior in people after severe

PAGE 42

42 spinal cord injuries.27,41,42,135 Dietz et al. (2003) also acknowledge s that alteration in glycinergic and GABAergic systems seen in animal models as a result of the training could be true for humans as well.166 The spinal cord as a result of the training learns to respond to specific sensory inputs associated with locomoti on which would potentially reorganize the circuitry involved in locomotion. Histologic data supporting the possibility of sprouting or synaptogenesis in the human spinal cord although not direct are also evident.167 Plasticity of the Spinal Stretch Reflex (SSR) Contradictory to their conceptualization as h ard -wired, the spinal stretch-reflexes are modulated during movement and adapt to training.85 Reflex modulation refers to the change in strength (or amplitude) of the reflex over the course of a behavior and is essential for the optimal performance of the motor behavior Researchers have shown that in monkeys, rats and humans these reflexes can be operantly conditioned i.e. the amplitude of the response can be either uptrained or down-trained.168-172 In these protocols the amplitude is measured as electromyographic activity and reward occurs when the amplitude is either abov e or below a criterion level.85 While change in the tonic descending activity motivated by the probabil ity to seek reward initiates reflex change, for this reflex to respond consiste ntly in this fashion over several sessions certain alterations occur somewhere in the spinal arc of the reflex. Similarly, short term and long term changes ha ve been noted in the reflex amplitude in animals and humans as a result of task-specific training.173-175 A single training session of short bouts of balancing on an unstable platform in normal subjects demonstrated a progressive decrease in the soleus H-reflex amplitude of about 26%.176 Hess et al also observed a modulation of the reflex size during adap tation to a new motor task.177 H-reflex modulation over time was evaluated for normal subjects over five runs of treadmill walking (three with normal treadmill walking and two with randomly stepping over the obsta cle 100 times). The largest adaptations

PAGE 43

43 with a significant incr ease of reflex amplitude occurred du ring the first obstacle run. This increase lasted only briefly and the reflex amp litudes decreased to their previous values. During the later obstacle run, no H-refl ex modulation occurred. Similarl y, normal subjects training to walk backward showed progressive adaptation of their soleus H -reflex in mid-swing phase as early as twenty minutes after training.175 Long-term maintenance of this modulation was also noted as late as five months after cessation of trai ning indicating acquisition and maintenance of novel motor skills. Operant or training induced c onditioning of the reflex can ha ve potential implication in rehabilitation. Studies based on application of this phenomenon in subjects with neurological injury have shown significant changes in the refl ex behaviors of these individuals. For example, studies have shown that operant conditioning of the spinal reflex reduced hyperactive biceps stretch reflex in people with spinal cord injuri es with the reduction pe rsisting for four months following cessation of training.178 Similar results in locomotor studies have only been documented in spinalized animal models thus far but might have potential ramifications in improving human locomotion as well.45,179,180 In four subjects with incomplete spinal cord injury, for example, a single bout of step training ove r the treadmill increased overground walking speed by 25% and reduced soleus H-reflex amplitude during overground walking providing evidence for activitydependent plasticity of the reflex.181 To this end, since reflex beha viors function as parts of co mplex behaviors, conditioning them in accordance to the requireme nts of the task i.e. task specific training may help improve functional outcomes in the targeted population.182 Also from a rehabilitation standpoint, provision of appropriate ta sk-specific stimuli and/or adopting di fferent feedback strategies that could facilitate or depress thes e reflex responses in desired wa ys could help in retraining or

PAGE 44

44 reeducation of function. For example, providing optimal hip extension during stepping would uptrain the stretch response of the hip fle xors to facilitate hip flexion for swing. In summary, short-term and long-term adaptive changes occur within the reflex components as a result of conditioning. These cha nges are also known to persist despite removal of the descending input thereby ex hibiting learned behavior at the level of the spinal cord. This adaptability and amenability of th e reflex to persistent change or conditioning constitutes spinal cord plasticity. Locomotor Training and Skill-dependent Plasticity of the Nervous System Ani mal and human research on the neurobiol ogical control of stepping and the skilldependent plasticity of the nervous system have challenged the compensation based approach of rehabilitation after SCI. Research has revealed that task-specific, repe titive training following SCI in animals and humans promotes skill-dependen t plasticity in the spinal cord and plays a critical role in the recovery of locomotor abilities including stepping.88 21,22,27,83 The emerging training strategy is to provide the CNS with peri pheral sensory input related to locomotion in order to stimulate a stepping response.19,24-26,41,43,44,135,183 Processing of task-specific kinematic and kinetic information facilitates performance of the task and learning.184 This strategy is based upon evidence that the lumbosacral spinal cord is capable of recognizing and processing functional sensory cues to produ ce a functional motor response.125 Simply put, generation of the stepping pattern would involve th e provision of motion-related affe rent input associated with stepping. Two of these inputs related to stepping are loading and hip position, which have been discussed earlier.38,127,132,135,185 A therapeutic intervention termed as "locomotor training" has been developed based on such research. In the locomotor training environment this specific sensory input is made available by having a person with SCI walk over a treadmill in a harness connected to a BWS system.43 This

PAGE 45

45 assembly provides an environment where normal walking speeds, bilateral limb loading and proper limb and trunk kinematics can be safely a nd effectively trained to enhance the neural output generating walking.43 Evidence supporting the benefi ts of locomotor training in facilitating the recovery of walki ng in humans has been recorded in the literature as early as the 1990s.25,39,41,42,186-188 Persons trained in this training environment demonstrated significant improvements not only in pattern of stepping, walking velocity, endurance but also in neurophysiological predictors of improved coor dination like the electr omyographic activity in the muscles of the lower extremity.189-191 Soleus H-reflexes have also been shown to respond to training post-SCI. A single bout of training ha s shown to increase walking speed with a corresponding decrease in reflex amplitu de, which is usually high after injury.181 Is Orthotic use Appropriate during Locomotor Training? Despite the com pelling evidence supporting the concept of locomotor training controversy and variability still exists in training parameters in the training protocol.24-26,42,46,192 Not all the parameters for optimizing recovery are known thus far and those that are, for example, BWS, manual assistance and speed are still being reviewed for mode of delivery and dosage.193 Therefore although the principles of training ha ve been established, there is no universal agreement on ideal training parameters for locomotor training.24,46,82 For example, controversy still exists about the use of parallel bars dur ing training. In a study done by Conrad et al improvements in step symmetry were observed us ing conventionally used parallel bars during training.68,194 On the contrary, Visintin et al reported improvements in symmetry using vertical body weight support instead of parallel bars providing upright support.195 Similarly, although widely used in conventiona l gait training, guiding ev idence for the use of orthotic devices such as the AFO during LT is lacking.24 Proponents of LT hesitate to use an AFO during training based on the assumption that it would interfere with optimizing the

PAGE 46

46 kinematics and therefore afferent input re quired to drive th e neural circuitry.24,26,196 Logical concerns expressed are that the AFO might restrict the range of excursion of the ankle and linked knee and hip joints altering the afferent in formation related to hip position and load. Additionally, although an AFO might facilitate an alternate steppi ng pattern, it would serve as a fixed variable unlike BWS and manua l assistance that can be adjust ed to facilitate independent stepping over the course of the training. Ther efore the use of the device during LT qualifies further investigation. Altered Stepping Mechanics Resulting from the Use of Orthotic/Assistive Devices Although the goal is to improve am bulation pot ential, prescription of assistive and/or orthotic devices often does not take into acc ount the influence of th e device on the user's resultant stepping pattern.197,198 For example, rigid support provi ded at the ankle by the PAFO might limit the excursion of the contingent joints affecting the control of the transition from stance-to-swing. Similarly, loading through the fi xed upper extremities over the parallel bars alters the prerequisite loading pattern thr ough the lower extremities required for stepping.195 However, the hierarchical framework assuming i rreversible walking defi cits post-injury have promoted compensatory function with such device s rather than examining their implications on the neural sub-tasks of walking. Under the supposition that neurological function ceases to exist below the level of injury, clinical assessments for the prescription of ambulatory devices utilize a frame work for substitution of impaired function rath er than the restitution of function.78 Therefore, the ambulatory ability of a person in moving from one place to another rather than the walking pattern utilized in achieving this mobility is em phasized. Individuals requiring assistive devices have a decreased ability to provide the supporting, stabiliz ing, propulsive or re straining forces at the lower extremities necessary for forward progression.199 With assistive devices the upper

PAGE 47

47 extremities are the structures bearing the lo ad and are being employed in providing the complementary forces for walking.200,201 Melis et al. (1999) reported a decrease in hip excursion and a decrease in the ability to unload the limb while walking with a walker.202 Visintin and Barbeau (1994) investigated the consequences of weight bearing on the upper extremities compared to weight bearing through the le gs, both with 40% BWS provided. Her results indicated a decrease in electrom yographic activity in the lower limbs and more asymmetry in the limb kinematics.195 Similarly while wearing orthotic devices like the PAFO, Ounpuu et al. (1996) reported an inability to generate sufficient power in the transition from stance-to-swing phase of stepping.203 However, conclusions drawn from these studies c ontend that ambulatory de vices are still capable of fulfilling various assistive functions during walking, although they affect posture and walking pattern. The inability to generate normal walking m echanics was concluded to be the result of the irreversible nature of the injury rather than the inability of the device to provide or assist normal stepping. Therefore the assessment of an assistive and/or orthotic device base d on its ability to restore normal walking mechanics is a novel perspective. Altered H-reflexes Resulting from the Use of Orthotic Devices The H-reflex has been comm only employed as a neural probe in describing and interpreting neural interplay for the control of movement. Elicitation of the reflex and measurement of its amplitude has provided insigh ts into the changes in transmission in spinal pathways during the perfor mance of a motor task.85 Since afferent informa tion is processed at the multiple levels of the nervous system including the spinal cord, examination of the soleus Hreflex modulation at the ankle joint would be a favorable tool in linking device-dependent sensory information to a motor response. Schneider et al, for example, demonstrated that in normal subjects bracing the ankle joint angle to 90, such as in preventing foot drop, the burst of

PAGE 48

48 TA activity was eliminated and so too the inhibition of the H-reflex.154 A similar inhibition in Hreflex was observed by Brooke et al on bracing the ankle joint suggestiv e of premotorneuronal mechanism for reflex inhibition.204 Nishikawa et al. (1999) examin ed the effect of applying an ankle brace in sitting to an uninjured ankle.205 They reported an increase in H-reflex amplitude in the braced compared to the unbraced conditi on concluding that ankle bracing increased stimulation of the cutaneous mechanoreceptors ar ound the joint. However, in these experiments the subjects were tested during passive movement of the limb or in a static position instead of active walking. Garrett et al. studied the effect an knee orthosis on soleus H-reflex modulation during walking.206 They reported that even though the walki ng pattern changed in normal individuals, phase-dependent H-reflex modula tion was not influenced by the use of a knee orthosis. In this study however, the ankle was not restricted and hence the specific effect of an AFO is unknown. Therefore differential modulation of the H-reflex while walking with and without wearing an AFO will help determine if walking with an AFO is an inherently different motor task compared to normal walking. Rationale of the Studies De monstrating the presence of numerous parall el systems within the CNS that reorganize after injury using intense, task-dependent inte rventions such as LT has catapulted the goal for rehabilitation of walking after SC I from compensation to recovery.21,207,25,208 Recovery of the sub-component of stepping, in this intervention depends on careful selection of training variables that deliver critical motion-rela ted sensory input to facilitate stepping. Conversely, aptness of a training variable such as an AFO to facilitate stepping can be determined by examining the mechanical and neurophysiological de viations observed during stepping.

PAGE 49

49 Therefore, I investigated th e effect of an AFO on mechanics and soleus H-reflex modulation in healthy, non-injured individua ls and persons with ISCI during stepping. Measuring these two factors in healthy, non-injured individuals provided a frame of reference for subsequently comparing normal and impaired st epping. Where as, examining changes in the two factors in persons with ISCI and comparing it to normal helped identify the degree of correction or deterioration in stepping provided by the AFO. Therefore, the cross-sectional studies have gleaned specific information about the use of AFOs in physiological paradigms that optimize recovery of walking after SCI. Methodological Considerations for the Measure ment of Mechanics an d Soleus H-reflex During Stepping Measurement of Stepping Mechanics Study considerations In the study protocol, m easurement of joint ex cursion at the hip, kn ee and ankle with and without wearing an AFO were used to assess crit ical kinematics required for the transition from stance-to-swing. The AFO was fitted on the dominant side for healthy, non-injured control subject and the more involved side for persons with ISCI. Kinematic and kinetic data are susceptible to change with walking speed.115 Therefore, the speed of walking was matched in both the walking with and without the AFO conditions. Also kinematic data is affected by limb length and height of the person.115 Therefore, for comparisons between people with ISCI and their non-injured counterparts, height, age and weight matched individuals were selected for the study. Interpretation of kinematic and kinetic data Walking is a m otor task in which each of th e interconnected segments of the lower limbs undergoes a characteristic excursion to move the body forward.33 If the terminal excursion of any

PAGE 50

50 one of the joints is restricted it has the potential to alter the excu rsion of the other joints linked to it thereby disrupting stepping pattern. The st ance-to-swing transition phase is usually characterized by ankle planta rflexion, knee flexion and hi p extension to aid push-off.33 If ankle range is restricted using an AFO, plantarfle xion will be limited and the knee and hip excursion will also be altered. Similarly, the swing-to-sta nce transition phase is characterized by ankle plantarflexion, knee exte nsion and hip flexion.33 If ankle plantarflexion is limited the knee and hip range will also be altered. Assessment of kinematic data while walking with an AFO will be helpful in elucidating this phenomenon. Furthermore assessment of kinetic data will indicate the discrepanc ies in the functional task requirements of stepping. For example, duri ng stance-to-swing phase, plantarflexor power is generated to propel the limb and body forward.33,203 If plantarflexor power is limited while walking with an AFO,203 the knee and hip join t powers might demonstrate a compensatory increase or decrease to prope l the limb forward. Also, with limited propulsion, peak braking and propulsive forces generated in the fore-aft dire ction will decrease compared to normal push-off. Similarly, the rate of loading will be used to quantify the ability to shift weight from one lower limb to the other and measured by calculating the slope of the vertical ground reaction force.209,210 A prolonged loading rate onto the limb donning the AFO will be indicative of an inability to plantarflex the ankle and extend the knee to support the transferred weight.210 Collective assessment and interpre tation of the above parameters will therefore pr ovide valuable information with regards to the m echanics of stepping with an AFO. Reliability of Vicon and Force Plate Measures Reliability r efers to the ability of an instrument to provide co nsistent, stable and repeatable measurements. In this project, Vicon motion analysis system (Vicon, Oxford Metrics Ltd,Oxford, England) and kistler forceplates (Kistler Instruments, Inc.,Amherst, NY) were used

PAGE 51

51 to identify the mechanical changes during ste pping with an without a PAFO. The Vicon motion analysis system is an automated, high-speed three-dimensional (3-D) motion system where cameras track the motion of retro-reflective ma rkers that are placed on subject's body landmarks. Kadaba et al (1989) investigated the repeatability of kinematic kinetic and electromyographic data using the Vicon motion analysis system on forty normal subjects three times a day on three separate days.211 Excellent intra-rater reliability was seen for kinematic data in the sagittal plane both within and between test days (Table 2-1). Similarly, frontal and tr ansverse planes joint angle motion yielded good repeatability within te st days leading them to conclude that gait variables measured by the Vicon system are quiet repeatable for subjects walking at their normal speed. Similarly, Richards et al (1999) compared accuracy of several automated motion analysis systems by placing two markers 50cms apart on an al uminum bar rotating in the horizontal plane in the camera capture volume.212 The results of his study indicated that the Vicon was the most accurate with a maximum error of 0.183 cm that is the lowest among all other systems. Several other studies have also reported ex cellent intra-rater, inter-rater a nd test-retest reliability of the Vicon motion analysis system making it the current gold standard for motion analysis.213-215 Ground reaction forces were measured with tw o kistler force plates that use triaxial piezoelectic force transducers mounted at the corners of each plate to measure the three components of the ground reaction force vectors. These force plates have been shown to be the reliable standard for measuring dynamic transition from bipedal to single limb stance in healthy, non-injured adults.216 Intraclass correlation coefficient (ICC) for the magnitude of the propulsive and braking force has been reported to be greater or equal to 0.73 for fast movements and greater or equal to 0.88 at the natural speed.

PAGE 52

52 Force plates have also been used for co mparing ground reaction force patterns in noninjured individuals and individuals with ISCI.217 Repeatability of initial vertical force peak and time to peak variables measured by force platfo rms while donning foot orthosis has also been reported.218 Excellent ICC results were demonstrated fo r the vertical force va riables, with power greater than 0.80. Measurement of peak vertical ground reaction force during a vertical jump at two time points 48 hours apart was demonstrated to be very reliable (ICC [2,1] = .94).219 Measurement of Soleus H-reflex Study considerations Elicitation of soleus H-reflex during walking with and withou t a PAFO requires control of several extraneous factors that could potentially confound the reflex response. These factors have been identified below and m ethods to control them have been discussed. For the purpose of consistency, soleus H-reflex will be evoked, on the dominant side of healthy, non-injured control subject and on the more involved side for pers ons with ISCI. The reflex will be evoked by localizing the tibial nerve in the popliteal fossa.138,220 Subject positioning is critical during Hreflex testing since several factors affect the so leus H-reflex. H-reflex is sensitive to various inputs including posture,221 joint position,149 reciprocal and recurrent inhibition,222 behavioral state,223,224 caffeine intake225 and muscle activity.226 However, if the above factors are sufficiently controlled, then H-refl ex can provide information of the state of the reflex arc. Stimulation intensity is another factor that affects reflex response. Intensity was maintained between 8-12% of M-max so as to evoke a direct muscle response. This procedure helped to safeguard against movement of the stimulating el ectrode that might alter the relative activation of the Ia afferent axons and alter the H-reflex amplitude without changes in synaptic efficacy. Soleus H-reflex amplitude is affected by background EMG activity of the muscle.149,226 Background EMG activity was measured 100 ms prio r to stimulation in each condition to ensure

PAGE 53

53 similar level of motor neuronal excitability. Stimulus frequenc y was maintained between 3-5 seconds to avoid post-activation depression of the response.222 Since profound reductions in Mmax amplitude have been reported to occur across the time course of an experiment, M-max was elicited in each condition and in each tested phase of the gait cy cle throughout the experiment for subsequent normalization of the data.227 Interpretation of H-reflex amplitude The H-ref lex demonstrates a phase-dependent regulation of its amplitude during walking which is an important component of motor contro l, allowing afferent feedback to have differing effects in different phases of the step cycle.228 This regulation is requ ired to accommodate the functional requirements of the task. For example, the soleus Hreflex is minimal at heel contact, increases to maximum during stance and decreases rapidly just prior to toe-off and is minimal during swing. The observed changes in size facil itate weight support and ankle extension during mid-to-late stance while allowing ankle dorsifl exion during swing and while the body moves over the foot during early to midstance.146,228,229 Therefore if the task of walking with an AFO was similar to walking without one then the neur al control would be preserved between the two conditions and the soleus H-reflex modulation in th e step cycle would be similar. An increase in H-reflex amplitude while walking with an AFO (assuming all other conditions including stimulus strength are maintained constant) compared to walking without one in selective phases of the step cycle would be i ndicative of an altered task. Similarly, with regards to adaptation of the re flex to a new motor task studies have shown this to occur in a biphasic fashion.137 For example, for the nove l task of stepping over an obstacle, Hess et al demonstrated a progressive adaptation in soleus H-reflex amplitude during repetitive stepping. He showed that in normal s ubjects the soleus H/M ratio increased strongly at onset of the motor learning task and reduced over the course of exercise reflecting the nervous

PAGE 54

54 systems capacity to adapt the locomoto r pattern to the actual requirements.177 The initial increase in soleus H-reflex amplitude is attributable to descending influence of the corticospinal tracts on the spinal reflex arc.230,231 The eventual reduction in reflex amplitude is speculated to be the effect of the acquired task being automatically pe rformed and controlled at a spinal or brainstem level.169 Therefore prolonged successive stepping with an AFO in normal subjects if different from walking without one would also exhibit adap tive changes in reflex am plitude as seen with the acquisition of a new motor task. Reliability of Soleus H-reflex The reliability of the soleus H-reflex test ing in supine and standing position has been confirm ed for inter-session a nd intra-session procedures.138 ICC (2, 1) for H-max, M-max and Hmax/M-max has been reported to be 0.99 +/0.007, 0.95 +/-0.08,0.97 +/-0.009 respectively. Similarly, the intra-session and inter-session reliabili ty of soleus H-reflex over five consecutive days in a standing position have also been established.220 The standing intra-session and intersession reliability was established to be 0.85 and 0.80 respectively. Clinical and Scientific Relevance of the Study Clinical Relevance of the Study Walking with orthotic and assistive devices h as been the quintessential approach for improving walking potential in persons with incomplete spinal cord injuries. Assessment of such devices in physiological-based training para digms like locomotor training will provide information about the degree of conformity of such clinical strategies with the principles of neurobiological control of walking. Accordingl y, in this study, mechanics of stepping and Hreflex modulation with the AFO will be evaluate d which characterize the motion-related afferent input being processed at the le vel of the spinal cord. Altere d stepping mechanics and reflex modulation with compensatory devices would reflect the failure to provide task-specific sensory

PAGE 55

55 input driving the intrinsic spinal circuitry neces sary for walking recovery. Knowledge of results from this study would influen ce clinical decision-making for the use of such devices and strategies in physiologicalbased training interventions. Scientific Relevance of the Study Apart from the clinical implications, this st udy stimulates a strong rationale for assessing archetypical strategies that have historically guided clinical practice so far. Generation of appropriate task mechanics associated with steppi ng is hypothesized as essential for recovery of stepping pattern in physiologica l-based training interventions. The transition phase mechanics are modulated by the motion-related information pr ocessed in every gait cycle. Assessment of the task mechanics of stepping in these transiti on phases while walking with orthotic devices is critical to examine how motion related informa tion generated by such strategies influence the ability to step successfully. Assessment of conventional strategies from a neural framework is relevant to the neurological population to which these stra tegies commonly apply. The tools used for assessment of walking are relevant to addre ss the task and functi on of specific events characterizing the walking behavior. Therefore th e integration of two different perspectives, biomechanical and neurophysiological, will provide an effective framework for understanding the control of movement and a ddressing the questions posed.

PAGE 56

56 Figure 2-1. Reduced animal preparation showing the impact of transection at different levels of the nervous system. Adapted from Patla, A. E. in Evaluation and management of gait disorders (ed. Spivack, B. S.) (New York, 1995). Spinal preparation Decerebrate preparation Decorticate preparation Intact system Near normal inter/intra limb activation patterns Functionally modulate reflex action Execute other rhythmic movements concurrently Improved coordination of activation patterns Weight support Active propulsion Dynamic equilibrium Initiate reasonably normal goal directed behavior in neonatally decorticate animal Repertoire of options limited Altered context of locomotor movements Adaptable locomotor control system to meet goals of the animal in any environment

PAGE 57

57 Figure 2-2. Stance-to-swing and swing-to-stance phases of the gait cycle. Figure 2-3. Model of the flexor and extensor half centers (F HC & EHC) and afferent input regulating stance and swing phase during stepping. Adapted from Van de Crommert, H. W., Mulder, T. & Duysens, J. Neural control of locomotion: sensory control of the central pattern generator and its relation to treadm ill training. Gait Posture 7, 251-63 (1998). Stance-to-swing Swing tostance

PAGE 58

58 Figure 2-4. Neural control of locomotion in an intact ner vous system (A) and compromised nervous system (B). Adapted from Edgerton, V. R., Tillakaratne, N. J., Bigbee, A. J., de Leon, R. D. & Roy, R. R. Plasticity of the spinal neural circ uitry after injury. Annu Rev Neurosci 27, 145-67 (2004).

PAGE 59

59 Figure 2-5. Electrical stimulus (shown here by the grey ellipse) applied to the mixed nerve conducts the stimuli orthodromically in th e motor and sensory axons to evoke the Mwave and the H-reflex respect ively (A). Stimulus triggere d recording of the H-reflex also known as latency respons e occuring after the M-wave (B). Adapted from Zehr, E. P. Considerations for use of the Hoffma nn reflex in exercise studies. Eur J Appl Physiol. 86, 455-68 (2002). Elicitation of the recruitment curve showing maximum H-reflex and M-max amplitude in response to stimulus intensity (C). Adapted from Palmieri, R. M., Ingersoll, C. D. & Hoffman, M. A. The Hoffmann reflex: methodologic considerations and applications for use in sports me dicine and athletic training research. J Athl Train. 39, 268-77 (2004). C Stimulus

PAGE 60

60 Table 2-1. Coefficient multiple correlations (CMC ) reflecting stride to stride variability between and within days. CMC within day CMC between days Right Left Right Left Vertical GRF 0.99 0.00 0.99 0.00 0.99 0.00 0.99.00 Horizontal GRF 0.99 0.00 0.99 0.00 0.98.01 0.98.01 Hip Flexion/Extension 0.99 0.00 0.99 0.01 0.97.01 0.97.01 Knee Flexion/Extension 0.99 0.01 0.99 0.00 0.98.01 0.98.01 Ankle dorsiflexion/ plantarflexion 0.97.02 0.97.01 0.93.03 0.93.03

PAGE 61

61 CHAPTER 3 EXAMINATION OF WALKING WITH AND WI THOUT AN ANKLE FOOT ORTHOSIS IN NON-INJURED INDIVIDUALS AND PERSONS WITH INCOMPLETE SPINAL CORD INJURY: A NEUROMECHANICAL PERSPECTIVE Spasticity, muscular weakness and co-activ ation are key motor impairments limiting walking potential in individuals with in complete spinal cord injury (ISCI).68-70 As one component of rehabilitation, c linicians often use orthotic de vices to compensate for these impairments and aid walking.15,75,78 Individuals lacking muscular control at the ankle are prescribed single joint ankle foot orthosis (AFO) to stabilize the joint, supplement deficient pushoff and aid toe clearance during stepping.16,17 An AFO limits ankle ex cursion and simultaneously influences excursion of the knee and hip, ther eby allowing the person to gain more proximal control for walking.232-235 AFOs also improve overground walking speed and interlimb kinematics in persons with ISCI.236 These broader benefits may lead to the assumption that a role exists for AFOs as permissive devices in recove ry based interventions such as locomotor training (LT). The AFO induced alterations in stepping mechanics, however, might not coalesce with the training principles of LT.24 During LT, stepping is retraine d by practice of task-specific repetitive, rhythmic, stepping kinematics over a treadmill using task facilitatory training variables such as body weight support and manual assistance.3,25,44,125 The training is based on the facilitation of intrinsic mechanisms within the spinal cord that respond to specific afferent input associated with the task of stepping. The spinal cord pro cesses afferent information arising from muscle, joint and cutaneous receptors duri ng stepping to adapt the motor output to the phase of stepping.34-37 For example, hip extension and limb unloading are critical afferent inputs required to initiate the transition from stance-to-swing during stepping.38,93,129 Limiting hip extension and/or unloading dela y this transition and limit the forward progression of the body.

PAGE 62

62 Proponents of LT therefore hesita te to train with an AFO due to the assumption that it impacts optimal kinematics and motion-rela ted sensory input re quired for stepping.24,26,237 Logical concerns expressed are th at the AFO might restrict the ra nge of excursion of the ankle and linked knee and hip joints alte ring the afferent information re lated to hip position and load. Additionally, although stanceto-swing transition with the AFO might be alternatively possible, the orthosis is a passive element in an otherwis e active training protocol where other facilitatory variables can be adjusted to facilitate independe nce. Therefore the purpose of the study was to investigate the use of the device during stepping in healthy, noninjured individuals and persons with ISCI. To assess the production of critical ankle, knee and hip joint kinematics required for stepping the range of excursion of thes e joints were measured during stepping.70,238-241 Furthermore, measurement of the vertical a nd horizontal forces reflected the rate of loading/unloading the lower limbs and peak br aking/propulsive force required for forward progression of the body respectively.240-242 As secondary variables of interest, interlimb temporal and spatial measures of symmetry correlated to the functional task requi rements of walking in these transition phases such as double limb suppor t time and step length were also assessed.5,243 In summary, two experiments were conducted inte rpreting mechanical information in transition phases (stance-to-swing a nd swing-to-stance) while walking w ith and without the AFO. The first experiment (Refer to Chapter 4) examined tran sition phase mechanics with and without an AFO in healthy, non-injured indivi duals that provided normative da ta for subsequently comparing with persons with ISCI. The sec ond experiment (Refer to Chapte r 5) compared transition phase mechanics observed while walking with and without the AFO in individuals with ISCI to normal

PAGE 63

63 mechanics. The comparison assessed changes in pre-existing stepping re lated deficits while walking with the AFO and measured deviance or likeness of the observed change from normal.

PAGE 64

64 CHAPTER 4 EXAMINATION OF WALKING WITH AND WI THOUT AN ANKLE FOOT ORTHOSIS IN HEALTHY, NON-INJURED INDIVIDUALS-1A The purpose of the experiment was to examine the effect of an AFO on transition phase mechanics during walking in h ealthy, non-injured individuals. Specifically, I assessed the immediate effect of wearing a PA FO on ipsilateral lower extremity 1) kinematics and 2) kinetics during the stance-to-swing and swing-to-stance phase of walking. Hypotheses 1A: Compared to walking without an AFO on a treadmill at speed approximating 1.2 m/s, wearing an AFO in healthy, noninjured individuals will affect the stance-to-swing transition mechanics observed on the AFO side: specifically, decrease peak ankle plantar flexion and hip extension and increase peak knee flexion range and increase peak knee and hip flexor powers and decrease peak ankle plantarflexor power. 1B: Compared to walking without an AFO on a treadmill at speed approximating 1.2 m/s, wearing an AFO in healthy, noninjured individuals will affect the swing-to-stance transition mechanics observed on the AFO side: specifically decrease ankle planta rflexion and increase hip and knee flexion range of motion and decrease rate of load ing and peak braking force. Methods Subject Selection A sample of convenience consisted of fourt een healthy, non-injur ed individuals living independently in the Gainesville community (T able 4-1). Each subject provided informed consent before participating in the study. The University of Florida Institutional Review Board and the Veteran Affairs Subcommittee approved th e study for clinical inve stigation. Mean age and standard deviation was between 26.9 3.7 ye ars. Subjects were screened for a medical history of any neurological, musculoskeletal or orthopedic problem that would affect their walking performance over the treadmill. Power analysis for determination of sample size was based on pilot data from three healthy, non-injure d participants. The cha nge in hip joint angle and peak braking force were selected for calcula ting sample size since thes e are the variables of

PAGE 65

65 primary interest. It was determined that a sample size of fourteen subjects was required to reach an alpha level of 0.05 and a power of 0.80 Experimental Set-up Once the subject had read and signed the informed consent form, motion data was collected and analyzed using a 3-D motion analysis system in conjunction with an ADAL3D instrumented split-belt treadmill custom manufactured and calibrated by TECMACHINE (Cdex, France), mounted flush with the floor a nd anchored to the found ation. Four Kistler piezoelectric sensors on each half treadmill allow calculation of the two-dimensional location of the center of pressure (COP) and the moment about the vertical axis, in addition to the threedimensional ground reaction force, under each foot. Belt speeds can be controlled as slow as 0.1 m/s. Ground reaction forces were recorded at 1000H z for each limb when in contact with the treadmill belt. The force plates were allowed to warm-up for at least 15 minutes as per manufacturer guidelines and calib rated prior to data collection. The walking pattern of the subjects were captured and analyzed by a Vicon three-dimensional motion analysis system. The system consists of the VICON 612 Datastation with twelve active vi deo channels and a 64 Channel A/D Board for analog signals. There were twelve 1000Hz M2-cameras (Digital CMOS M2 series cameras have a reso lution of 1280 x 1024). Included software was: Workstation, Polygon, BodyBuilder, Plug-In Gait, Plug-In Mode ller, and Real Time II. The twelve cameras had a frame rate of 60-120 fps and used infrared (IR) light-emitting diode strobes, which were gen-locked. Static calibration of the system used the clinical L-frame, which contains 4 retro reflective markers, being placed in a predet ermined position on the motion analysis force platforms. Following this a dynamic calibrati on was done using a 500mm wand that was moved around the capture area for approximately 20 sec onds. Analog video data were also collected

PAGE 66

66 using a standard camcorder record ing at 100 fps with its optical ax is perpendicular to the plane of interest (i.e. the sagittal plane of motion). Subject Preparation Subjects were asked to wear tennis shoes and change into appropriate clothing (dark colored cycling shorts and shirt) for testing. For trials using the AFO, each subject was fitted with an off the shelf posterior leaf spring ankl e foot orthosis (PAFO). Fitting was assessed by measuring fit inside shoe, length of the calf shell and that of the footplate. Standardized fitting included using a PAFO whose lengt h fits an inch to two below the fibular head when donned and whose footplate length extends till the tip of the toes.244 Lightweight retro-reflective markers were attached to the following bony landmarks: poste rior superior iliac sp ines (PSIS), anterior superior iliac spines (ASIS), knee-joint axes, late ral malleoli, medial malleoli, clavicular notch, sternum, C7, T10, and acromium processes. The second foot ray, base of the 5th metatarsal and the heel markers were approximated on the subjects shoes. Clusters of mark ers were attached to the pelvis, thigh, shank, and foot segments. Th is modified Helen Hayes marker set is commonly used to capture bilateral 3D kinematics using a twelvecamera VICON motion analysis system.208 Each subject was fitted with a body weight s upporting harness equipped with an additional overhead safety catch. The harness and safety catch when used either with or without BWS provided safety to the person walking on the treadmill and holds or catches the person if he or she should lose their balance, stum ble or begin to fall (Figure 4-1). Procedure After equipment set-up and subject preparation, the walking trials over the instrumented treadmill were recorded. First, subjects were as ked to stand with one leg on each belt of the instrumented treadmill to record a static trial. The static trial was used to create the subject

PAGE 67

67 specific model by defining joint center locations and segment lengths. The leg chosen for donning the AFO and the order of testing with and without it wa s randomized for each subject. For the AFO trial, each subject was requested to wear a unilatera l, size-fitted PAFO. The insole of the shoe was removed in order to fit the AF O and to even out the limb length on both sides. The subject walked on the instrumented treadmill fo r the collection of kinematic and kinetic data with the overhead safety and harness. Subj ects were permitted to practice walking on the treadmill until they achieved steady state walking at the speed of 1.2m/s and comfort while walking in this environment. Once the subject felt comfortable at the set speed and the investigator viewed a steady-stat e pattern of walking, kinematic and kinetic data was collected for 30 seconds in each of the two conditions. Afte r data collection, the trial was processed to verify if all the desired data was collected properly. Rest was provi ded during testing, as requested. This experiment took approximately two hours from the start for set-up and data collection. Data Processing Kinetic data (Ground reaction forces and moments) and segm ent kinematic data was low pass filtered with zero lag di gital Butterworth filter (20 and 9 Hz cut-off frequencies respectively). Software for Interactive Musculoskeletal Modeling (SIMM) was used to create subject specific models. Segment inertial proper ties were calculated for each subject based on the subjects mass and segment length s. SIMM and SDFast performed an inverse dynamics analysis for each trial.115,117 All data were averaged across trials for each subject. The kinematic and kinetic data from each trial was normalized to percent stride using Matlab code and then compared between the two conditions.

PAGE 68

68 Data Analysis For each phase of interest, the two conditi ons (with and without AFO) were compared using a Hotellings T2-test, which is a multivariate analogue of the paired t-test. The test is a multivariate extension of the Student's t-test for paired data in comparison of mean difference vectors, i.e. the differences of two or more dependent variable s considered twice in the same subjects.245 The primary dependent variables compared included the rate of limb loading, peak braking force, excursion and power at the ankle, knee and hip joints. For the stance to swing phase the excursion and power at the ankle, knee and hip joints were analyzed collectively. For the swing to stance phase the joint excursion, ra te of limb loading and peak braking force were analyzed collectively. Interlimb temporal and spatial meas ures of symm etry i.e. double limb support time and step length, our secondary dependent variables we re compared between conditions using a paired t-test. Significance level was set at p< 0.05. To correct for multiple comparisons, a Holm's step down method was used that adjusted p-values for each research question.246 Results Tables 4-2, 4-3 and 4-4 and Figures 4-2, 4-3, 4-4 & 4-5 show the mean and standard deviation in kinematic and kine tic measures related to each ph ase with and without the PAFO ipsilaterally. In the stance-to-swing phase a signif icant decrease in peak a nkle plantar flexion, hip extension and peak plantarflexor power were noted while walking with a PAFO. In the stance-toswing phase, while walking with a PAFO, a significant increase in hip flexion, decrease in the rate of loading and peak braking force were obs erved. With regards to interlimb coordination with the PAFO, double limb support time increas ed significantly on the ipsilateral limb.

PAGE 69

69 Discussion The main finding of the study was that in the stance-to-swing phase, donning a PAFO kinematically decreased peak ankle plantarflexion and hip extension and kinetically decreased plantarflexor power. Additionally, in the swing-to-stance phase, wearing a PAFO kinematically increased hip flexion, decreased ankle dorsiflexion and kinetically decreased peak braking force and increased rate of limb loading. In stance-to-swing phase, with a PAFO, non-injured subjects demonstrated a significant decrease in hip extension and a nkle plantarflexion which coincide d with a decrease in propulsive force. Generation of plantarflexor power is vital for forward progression of the body. The ankle plantar flexors provide ~70% of the joint work during walking.119,247 However, bracing the ankle decreased plantarflexor power generation by 17%. Reduction in power may have resulted from a decrease in angular velocity or moment and subsequently contri buted to slowing limb progression.115 Therefore the PAFO reduced the ability of the ankle to contribute to push-off in the stance-to-swing transition phase. Interestingly, the effect of the brace was not only isolated to the ankle but also observed at the hip joint. A de crease in hip extension ob served as a result of the brace could lead to the poor stretch of the hip flexor muscle s thereby increasi ng the difficulty in initiating swing.38 In the swing-to-stance phase, the primary function of the PAFO is to prevent footdrop. However, with a PAFO non-inju red individuals demonstrated a significant increase in hip flexion and a decrease in ankle dorsiflexion thereby affecting the heel rocker. The heel rocker is the first phase in the gait cycle after initial contact that determines the limbs loading response.33 The momentum generated by the fall of body wei ght onto the stance limb is preserved by this heel rocker. Normal initial contact is made by the calcaneal tuberosity, which becomes the fulcrum about which the foot and tibia move. With a PAFO, an increase in hip flexion and a

PAGE 70

70 simultaneous decrease in ankle dorsiflexion in the swing-to-stance phase limit the smooth transfer of body weight onto the stance limb. Kinetically this coincided with the decrease in peak braking force on the PAFO si de and a delay in the rate of loading. Temporally, a decreased loading rate corresponded with an increase in the initial ph ase of double support time on the ipsilateral side indicative of a delay in shif ting the weight from one limb to another. In non-injured individuals, our study demonstrat ed that an orthosis altered the transition phase kinematics and kinetics crucial to stepping. Our findings could have potential implications in neurologically impaired individuals in whom brace walking is common. Past studies have evaluated the benefits of an ankle foot or thosis in different ne urological populations.248,249 However these studies have evaluated the compen satory benefits of using an orthosis on temporal and spatial patterns of walking with the device without accounting for changes in joint kinematics and kinetics in the transition phases of walking. Appropriate joint kinematics and kinetics in the transition phases are crucial for providing optimal motionrelated sensory input to a compromised nervous system. Utilization of th ese sensory inputs has been shown to aid in retraining the neuromuscular system for walking recovery.24,26,44 Our study explored this paradigm shift by re-examining the use of su ch devices for walking by examining joint kinematics and kinetics during the transition phase of walking. In our study, use of an orthosis failed to produce desired proximal joint kinematics such as hip extension in the stance-to-swing phase of walking and meet the functional task re quirements such as rate of loading in ablebodied individuals. Since the orthosis affected walking in non-injured in dividuals its effect on gait in individuals nervous system disorders coul d be more pronounced. Therefore, the results of our study suggest that the purpose an d functional implication of an ankle foot orthosis needs to be evaluated rigourously in neurological populations. Neurobiologically driven recovery based

PAGE 71

71 interventions such as locomotor training target ed at providing normal walking kinematics and thereby appropriate motion relate d sensory input to a compromised nervous system need to weigh the use of such devices cautiously. Limitation We did not collect EMG in these individuals which would help correlate our findings with lower limb muscle activity. Also, examination of kinematic and kinetic changes with a PAFO, limits the generalizability of our results to other more rigid devices such as solid or hinged ankle foot orthoses. Conclusion A minimally restrictive device such as a PA FO in non-injured indi viduals impacted the provision of critical a fferent input during the transition phases of walking. Proximal hip extension crucial for the transition from stan ce-to-swing and the rate of loading during the swing-to-stance phase were decreased. Intuitively, use of more rigid devices could exaggerate these findings. Non-injured individuals were able to adapt to walking with the PAFO by increasing the double support time ip silaterally. Given the clinical relevance of our study, the use of a PAFO for neurological populations needs to be systematically assessed.

PAGE 72

72 Figure 4-1. Participant with safety harness walking on an instrumented treadmill. Reflective markers were applied to bony landmarks on th e pelvis and bilateral lower extremities for this study.

PAGE 73

73 Figure 4-2. Ipsilateral and cont ralateral average joint angles w ith and without the ankle foot orthosis (A FO) ipsilaterally. The ipsilateral decrease in peak hip extension and increase in hip flexion [A], no change in peak knee flexion [B] and decrease in peak ankle plantarflexion and dorsi flexion [C] during the swing-to-stance and stance-toswing phase of the gait cycle are highlighted by dotted circles. Vertical lines represent point of toe-off in the gait cycle. Sign ificant changes repr esented as p<0.05. P<0.05 P<0.05 sw in g -t o s tan ce stance-to-swing sw in g -t o s tan ce stance-to-swing swing-to-stance stance-to-swing P<0.05 P<0.05 A B C D E F

PAGE 74

74 Figure 4-3. Stick figure repres enting the changes in individual joint motion in the stance-toswing and swing-to-stance phase of the ga it cycle with and without the ankle foot orthosis (AFO). Ankle dorsiflexion with AFO Ankle plantarflexion With AFO Knee flexion Hip extension Hip extension Ankle plantarflexion Knee flexion Ankle dorsiflexion Hip flexion Knee flexion Hip flexion Knee flexion stance-to-swing phase swing-to-stance phase Without AFO Without AFO With AFO With AFO

PAGE 75

75 Figure 4-4. Ipsilateral and cont ralateral average joint powers duri ng the stance-to-swing phase of the gait cycle with and wit hout the ankle foot orthosis (AFO) ipsila terally. The ipsilateral hip flexor power [A], knee fle xor power [B] and ankl e plantarflexor power [C] are highlighted by dotted ci rcles. Vertical lines represent point of toe-off in the gait cycle. Significant changes represented as p<0.05. stance-to-swing stance-to-swing P<0.05 stance-to-swing A B C F E D

PAGE 76

76 Figure 4-5. Ipsilateral and contralateral vertical and horizon tal (AP) ground reaction forces (GRF) during the swing-to-stance phase of the gait cycle with and without the ankle foot orthosis (AFO) ipsilaterally. The pr olonged rate of loading during vertical loading [A] and decrease in horizontal br aking force [C] are highlighted by dotted circles (p<0.05) ipsilaterally. P<0.05 P<0.05 swing-to-stance swing-to-stance A B C D

PAGE 77

77 Table 4-1. Demographics of the study participants. ID Age Sex Orthotic Size N1 24 F R Small N2 27 M R Large N3 30 M R X-large N4 28 F R Medium N5 26 F R Medium N6 31 F R Medium N7 27 M R Large N8 24 F R Medium N9 22 M R Large N10 27 F L Small N11 23 M L Large N12 29 M L Medium N13 23 M L Large N14 36 F L Small Table 4-2. Hip, knee and ankle joint kinematic and kinetic data while walking with and without an ankle foot orthosis (AFO) during the stance-to-swing phase of the gait cycle. Stance-to-swing phase Without AFO Standard deviation With AFO Standard deviation pvalue Peak hip joint extension (degrees) -8.67 5.58 -6.77 5.51 .001* Peak knee joint flexion (degrees) -64.87 3.80 -64.54 4.15 .541 Peak ankle joint plantarflexion (degrees) -19.45 5.71 -12.01 5.28 .000* Hip joint power (Watts/ body weight) 0.11 0.04 0.12 0.04 .899 Knee joint power (Watts/ body weight) -0.15 0.02 -0.15 0.03 .893 Ankle joint power (Watts/ body weight) 0.18 0.03 0.15 0.03 .000* *Significant changes.

PAGE 78

78 Table 4-3. Hip, knee and ankle joint kinematic kinematic and ki netic data while walking with and without an ankle foot orthosis (AFO) during swing-to-stance phase of the gait cycle. Swing-to-stance phase Without AFO Standard deviation With AFO Standard deviation pvalue Peak hip joint flexion (degrees) 32.75 4.73 35.60 5.13 .028* Peak knee joint flexion (degrees) -17.86 3.47 -17.81 3.55 .924 Peak ankle joint plantarflexion (in degrees) -8.40 3.55 -10.90 4.77 .000* Rate of loading (N/kg) 0.06 0.01 0.05 0.01 .018* Peak braking force (N/kg) -0.16 0.02 -0.15 0.02 .013* *Significant changes. Table 4-4. Average interlimb te mporal and spatial data while walking with and without an ankle foot orthosis (AFO). No AFO Unilateral AFO Primary side Contralateral side p-value Primary side Contralateral side p-value Step length (meters) 0.57 0.14 0.58 0.14 .329 0.56 0.15 0.58 0.13 .326 Double support time (seconds) 0.19 0.02 0.19 0.02 .111 0.22 .03 0.17 0.02 .000* *Significant changes.

PAGE 79

79 CHAPTER 5 COMPARISON OF WALKING WITH AND WI THOUT ANKLE F OOT ORTHOSIS IN PERSONS WITH INCOMPLETE SPINAL CORD INJURY-1B Given the gait deficits after ISCI, two invest igative steps that w ould serve to inform clinical-decision making for use of an AFO dur ing walking retraining are 1) to examine the transition phase mechanics in persons with ISCI while walking with and without the AFO and 2) to compare the observed mechanics in each of the conditions to normal walking mechanics. Therefore we proposed to assess the changes in walking mechanics in in dividuals with ISCI while walking with an AFO and to examine the pr oximity or deviation of the observed change to normal matched control values. Specific Aims and Hypothesis Aim 1: To compare the immediate effect of wa lking with and without wearing an AFO on lower extremity kinematics and kinetics duri ng the stance-to-swing and swing-to-stance phase of walking in persons after ISCI. Hypothesis 1a: Compared to walking without weari ng a AFO, wearing an AFO in persons with ISCI will significantly change the stance-to-swing transition observed on the AFO side: specifically, peak ankle plantar flexion, knee flexion and hip extension and peak knee, hip and plantarflexor powers. Hypothesis 1b: Compared to walking without wearing an AFO, wearing an AFO in persons with ISCI will significantly change the swing-to-stance transition observed on the AFO side: specifically, ankle plantarflexion, knee and hip flexion and rate of loading and braking force. Aim 2: To compare lower extremity kinematics a nd kinetics during the stance-to-swing and swing-to-stance phase while walking with and without wearing an AFO in individuals with ISCI to that of healthy, n on-injured age, weight, height and speed matched controls. Hypothesis 2a: In persons with ISCI, the stance-to-swing transition will be significantly deviated from normal while walking with an AFO compared to walking without one. Specifically, with an AFO there will be a d ecrease in peak ankle plantar flexion and hip extension and increase in peak knee flexion a nd an increase in peak knee and hip flexor powers and decrease peak a nkle plantarfle xor power. Hypothesis 2b: In persons with ISCI, the swing-to-stance transition will be significantly deviated from normal while walking with an AFO compared to walking without one. Specifically, with an AFO there will be a d ecrease in ankle plantarf lexion and increase in knee and hip flexion and a decrease in ra te of loading and peak braking force.

PAGE 80

80 Methods Subject Selection Eight persons with ISCI ranging between 18-80 years and th eir height, weight and age matched controls were recruited for this experiment and signed an informed consent form approved by University of Florida Institutional Review Board and the Veteran Affairs Subcom mittee for Clinical Investigation. Particip ant demographics are tabulated in Table 5-1. American Spinal Injury Association (ASIA) motor score and impairment scale data were collected from all participants with ISCI to a ssess the degree of impair ment in each leg. The criteria for inclusion in the study were as follows: 1) Persons with ISCI classified as ASIA D 2) Medically stable 3) Have quadriceps strength of at least 3/5 4) Have decreased ankle strength (dorsiflexor strength of le ss than or equal to 4/5) 5) and/or absent or impaired proprioception at the ankle 6) Can stand unaided for one minute 7) Can walk with minimal assistive device such as the cane but does not use an ankle foot orthosis. Exclusion criteria included persons who were unable to follow 3 step commands, amputation, medical instability, significant muscul oskeletal problems other than SCI that limit hip and knee extension or ankle plantarflexion to neutral. Samp le size was determined from previous studies examining similar gait characteristics in these individuals.250,251 Pepin et al. found significant differences (p<0.0 1) in hip extension in a samp le size of seven persons with ISCI compared to their non-injured counterpa rts at matched speeds while comparing the adaptability of gait pattern in individuals with ISCI to different walking speeds.250 The change in hip joint angle (hip extension ch ange from pilot data=5.25 degrees, SD from Pepin's study= 5)

PAGE 81

81 were used to calculate sample size for our study. It was determined that a sample of 8 subjects will be required to reach an alpha level of 0.05 and a power of 0.80. Experimental Set-up Once the subject had read and signed the informed consent form, motion data was collected and analyzed using a 3-D motion analysis system in conjunction with an ADAL3D instrumented split-belt treadmill custom manufactured and calibrated by TECMACHINE (Cdex, France), mounted flush with the floor a nd anchored to the found ation. Four Kistler piezoelectric sensors on each half treadmill allow calculation of the two-dimensional location of the center of pressure (COP) and the moment about the vertical axis, in addition to the threedimensional ground reaction force, under each foot. Belt speeds can be controlled as slow as 0.1 m/s. Ground reaction forces were recorded at 1000H z for each limb when in contact with the treadmill belt. The force plates were allowed to warm-up for at least 15 minutes as per manufacturer guidelines and calib rated prior to data collection. The walking pattern of the subjects were captured and analyzed by a Vicon three-dimensional motion analysis system. The system consists of the VICON 612 Datastation with twelve active vi deo channels and a 64 Channel A/D Board for analog signals. There were twelve 1000Hz M2-cameras (Digital CMOS M2 series cameras have a reso lution of 1280 x 1024). Included software was: Workstation, Polygon, BodyBuilder, Plug-In Gait, Plug-In Mode ller, and Real Time II. The twelve cameras had a frame rate of 60-120 fps and used infrared (IR) light-emitting diode strobes, which were gen-locked. Static calibration of the system used the clinical L-frame, which contains 4 retro reflective markers, being placed in a predet ermined position on the motion analysis force platforms. Following this a dynamic calibrati on was done using a 500mm wand that was moved around the capture area for approximately 20 sec onds. Analog video data were also collected

PAGE 82

82 using a standard camcorder record ing at 100 fps with its optical ax is perpendicular to the plane of interest (i.e. the sagittal plane of motion). Subject Preparation Subjects were asked to wear tennis shoes and change into appropriate clothing (dark colored cycling shorts and shirt) for testing. For trials using the AFO, each subject was fitted with an off the shelf posterior leaf spring ankl e foot orthosis (PAFO). Fitting was assessed by measuring fit inside shoe, length of the calf shell and that of the footplate. Standardized fitting included using a PAFO whose lengt h fits an inch to two below the fibular head when donned and whose footplate length extends till the tip of the toes.244 Lightweight retro-reflective markers were attached to the following bony landmarks: poste rior superior iliac sp ines (PSIS), anterior superior iliac spines (ASIS), knee-joint axes, late ral malleoli, medial malleoli, clavicular notch, sternum, C7, T10, and acromium processes. The second foot ray, base of the 5th metatarsal and the heel markers were approximated on the subjects shoes. Clusters of mark ers were attached to the pelvis, thigh, shank, and foot segments. Th is modified Helen Hayes marker set is commonly used to capture bilateral 3D kinematics using a twelvecamera VICON motion analysis system.208 Each subject was fitted with a body weight s upporting harness equipped with an additional overhead safety catch. The harness and safety catch when used either with or without BWS provided safety to the person walking on the treadmill and holds or catches the person if he or she should lose their balance, stum ble or begin to fall (Figure 4-1). Procedure After equipment set-up and subject preparation, the walking trials over the instrumented treadmill were recorded. First, subjects were as ked to stand with one leg on each belt of the instrumented treadmill to record a static trial. The static trial was used to create the subject

PAGE 83

83 specific model by defining joint center locations and segment lengths. The leg chosen for donning the AFO and the order of testing with and without it wa s randomized for each subject. For the AFO trial, each subject was requested to wear a unilatera l, size-fitted PAFO. The insole of the shoe was removed in order to fit the AF O and to even out the limb length on both sides. The subject walked on the instrumented treadmill fo r the collection of kinematic and kinetic data with the overhead safety and harness. Subj ects were permitted to practice walking on the treadmill until they achieved steady state walking at the speed of 1.2m/s and comfort while walking in this environment. Once the subject felt comfortable at the set speed and the investigator viewed a steady-stat e pattern of walking, kinematic and kinetic data was collected for 30 seconds in each of the two conditions. Afte r data collection, the trial was processed to verify if all the desired data was collected properly. Rest was provi ded during testing, as requested. This experiment took approximately two hours from the start for set-up and data collection. Data Processing Kinetic data (Ground reaction forces and moments) and segm ent kinematic data was low pass filtered with zero lag di gital Butterworth filter (20 and 9 Hz cut-off frequencies respectively). Software for Interactive Musculoskeletal Modeling (SIMM) was used to create subject specific models. Segment inertial proper ties were calculated for each subject based on the subjects mass and segment length s. SIMM and SDFast performed an inverse dynamics analysis for each trial.115,117 All data were averaged across trials for each subject. The kinematic and kinetic data from each trial was normalized to percent stride using Matlab code and then compared between the two conditions.

PAGE 84

84 Data Analysis For each phase of interest, the two conditions (with and without AFO) were compared using a Hotellings T2-test, which is a multivariate analogue of the paired t-test. The test is a multivariate extension of the Student's t-test for paired data in comparison of mean difference vectors, i.e. the differences of two or more dependent variable s considered twice in the same subjects.245 The dependent variables compared included the rate of limb loading, peak braking force, excursion and power at the ankle, knee a nd hip joints. For the stance-to-swing phase, the excursion and power at the ankle, knee and hip joints were analyzed collectively. For the swingto-stance phase, the joint excursion, limb lo ading and peak braking force were analyzed collectively. Interlimb temporal and spatial measures of symmetry i.e. double limb support time and step length, our secondary dependent variab les were compared betw een conditions using a paired t-test. Significance level was set at p< 0. 05. To correct for multiple comparisons, a Holm's step down method was used that adjusted p-values for each research question. The same analyses were repeated comparing the control data to indi viduals with ISCI walking with and without the AFO.246 Results Figures 5-1 through Figure 5-6 and Tables 5-2 through Table 58 show the change in the kinematic and kinetic measures related to each gait phases with and without the PAFO ipsilaterally in individuals with ISCI and their matc hed non-injured controls. Within-Subject Comparisons for Individuals with ISCI In the stance-to-swing phase, with a PAFO, a decrease in hip extension was observed within subjects (Figure 5-1 & Ta ble 5-2). Likewise, in the swi ng-to-stance phase, with a PAFO, an increase in peak knee flexion was observed wi thin subjects (Figure 51 & Table 5-3). After

PAGE 85

85 correcting for multiple comparisons in these phases, weak statistical support existed for these variables. Between-Subject Comparisons for Individuals with ISCI Walking Without a PAFO and Their Matched Control In the stance-to-swing phase, without a PAFO, a decrease in peak knee joint flexion and knee joint power were observed in individua ls with ISCI. After correcting for multiple comparisons in this phase, weak statistical support existed fo r these variable s (Figure 5-1 & Table 5-4). In the swing-to-stance phase, wit hout a PAFO, a significant increase in peak hip flexion was observed in individuals with ISCI, which after correcting for multiple comparisons was statistically significant (Figure 5-1 & Table 5-5). Between-Subject Comparisons for Individuals with ISCI Walking with a PAFO and Their Matched Control In the stance-to-swing phase, with a PAFO, a decrease in peak knee joint flexion and knee joint power were observed in individuals with IS CI. After correcting for multiple comparisons in this phase, weak statistical suppor t existed for these variables (F igure 5-1 & Table 5-6). In the swing-to-stance phase, with a PAFO an increase in peak hip flex ion was observed in individuals with ISCI, which after correcting for multiple comparisons was statistically significant (Figure 51 & Table 5-7). Temporal and Spatial Comparisons Within Subjects Both the spatial and temporal measures of symmetry namely step length and double limb support time did not demonstrate significant di fferences between the two walking conditions (Figure 5-5, 5-6 & Table 5-8 ). Discussion The main finding of the study was that in individuals with ISCI, donning a PAFO decreased hip extension in the st ance-to-swing phase of walking compared to walking without it

PAGE 86

86 (Figure 5-4 & 5-8). While walki ng with a PAFO, step length va lue and interlimb symmetry did not change between conditions. Inte restingly, double limb support time on the ipsilateral limb increased in 5/8 subjects while walking with a PAFO (Figure 5-6). Additionally, symmetry indices for the interlimb double li mb support time increased concurrent with the increase in ipsilateral double limb support time. Furthermore, compared to normal walking, gait in individuals with ISCI walking without a PAFO was characterized by a signi ficant decrease in knee flexion in the stance-to-swing phase that also correlated to a decrea se in knee flexor power Likewise, a significant increase in hip flexion in the swing-to-stance pha se of walking was observed. Inte restingly, we did not observe a trend for improvement or deterioration in these pre-existing gait deviations with the PAFO when compared to matched controls. The increase in hi p flexion in the swing-to -stance transition phase and decrease in knee flexion in the stance -to-swing phase are common gait deviations characteristic to individuals w ith ISCI. Pepin et al (2003) has demonstrated a significant increase in hip flexion at the time of heel contact in individuals with ISCI.250 Likewise the reduction in knee angular velocity during walking is a common gait deviation observed in individuals with ISCI which could account for the decrease in knee flexion.70 With regards to power generation, during the stance-to-swing phase, in both the conditions, individuals with ISCI de monstrated an ipsilatera l decrease in ankle planta rflexor, knee flexor and hip flexor power compared to the contralateral limb. Generation of planta rflexor power is vital for forward progression of the body. The ankle plan tar flexors provide ~70% of the joint work during walking.119,247 However, wearing a PAFO did not au gment power generation at the ankle. Additionally, a decrease in horiz ontal propulsive and braking force were also noted on the ipsilateral side with a nd without PAFO compared to the co ntralateral limb. The decrease in

PAGE 87

87 propulsive force could result from inability to generate sufficient power ipsilaterally. Likewise, the reduced braking force compared to the cont ralateral limb could result from the significant increase in hip flexion observed in these individuals. Clinical Implication Unlike past studies reporting th e compensatory benefits of us ing an ankle foot orthosis, this study uniquely examined the ability of the PAFO to meet th e normal kinematic and kinetic task requirements of stepping. Failure to im prove limb kinematics and kinetics with a PAFO during treadmill walking is suggestive of the inab ility of the device to provide a normal walking pattern in individuals after ISCI. Importantly, a distally worn PAFO impacted proximal joint excursion by limiting hip extension. Hip extension is one of the essential kinematic features of the stance-to-swing transition duri ng walking. Studies have shown that preventing the hip from attaining an extended position inhi bited the generation of the flexor burst and hence the onset of the swing phase.129,130, 38 The use of PAFO for step retraini ng on a treadmill after SCI may thus hinder achievement of the task-specific, locomo tion-related afferent i nput used to retrain stepping via sensorimotor activati on of the neuromuscular system.37 Limitations The results of our study demonstrated weak statistical support for within subject comparisons. We had performed our initial power and sample estimates based on preliminary data for our primary outcome of interest: hip ex tension. Consequently, we found this variable was different between conditions. However, the secondary variables remained under powered to find true differences if they existed. With a larg er sample size very sm all differences would be detected as significant. However, statistical significance needs to be assessed with caution since it does not imply if the difference between the variables is large or important. Additionally, apart

PAGE 88

88 from the kinematic and kinetic data, collection of electromyographic data in these individuals would have helped correlate our findings with lower limb muscle activity. Conclusion For the rehabilitation specialist, the characterizati on of gait deficits observed in ISCI subjects is important for treatment purpos es. Likewise, characterization of improvement or deterioration in these deficits with the use of orthotic devi ces is important particul arly for developing and bettering new rehabilitation approaches. As ne wer interventions are being developed, the therapeutic rationales for the use of orthotic devices might change based on the guiding principles of these interventions. Traditionally, wearing an AFO in individuals with ISCI has been considered compensatory solution or a qui ck fix for remediating gait deficits resulting from muscular weakness, incoordination and spas ticity. However, from a neurobiological control of walking based perspective, an AFO may alte r the sensory experience necessary for retraining the nervous system and might not pr oduce the desired th erapeutic effect.24,252 Interestingly, in our study, the use of a minima lly restrictive PAFO decr eased hip extension in participants with ISCI. The observe d decrease could impact the provi sion of at least one critical afferent input key to the restorat ion of walking. Furthermore, use of more rigid devices is likely to exaggerate our findings. Consequently, if th e goal of recovery based interventions such as locomotor training is to provide optimal limb ki nematics, the use of a PAFO for stepping would not coincide with the principles of training.

PAGE 89

89 Figure 5-1. Ipsilateral and contralateral average joint angles during the swing-to-stance and stance-to-swing phase of the gait cycle w ith and without the a nkle foot orthosis (AFO) ipsilaterally. The ipsilateral peak hip flexion and extension [A], peak knee flexion [B] and peak ankle plantarflexi on and dorsiflexion [C] are highlighted by dotted circles. The gray shaded area repres ents matched control data. Vertical lines represent point of toe-off in the gait cycle. Significant changes in the joint angles represented as p<0.05. P<0.05 P<0.05 swing-to-stance stance-to-swing stance-to-swing swing-to-stance swing-to-stance stance-to-swing P<0.05 P<0.05 A B C D E F

PAGE 90

90 Figure 5-2. Ipsilateral and contralateral average joint powers during the swing-to-stance and stance-to-swing phase of the gait cycle w ith and without the a nkle foot orthosis (AFO) ipsilaterally. The ipsilateral hip fle xor power [A], knee fl exor power [B] and peak ankle plantarflexor power [C] are highl ighted by dotted circles. The gray shaded area represents matched control data. Vertical lines represent point of toe-off in the gait in the gait cycle. Significant changes in the joint powers represented as p<0.05. stance-to-swing stance-to-swing stance-to-swing P<0.05 C B A D E F

PAGE 91

91 Figure 5-3. Ipsilateral and contralateral vertical and horizon tal (AP) ground reaction forces (GRF) during the swing-to-stance phase of the gait cycle with and without the ankle foot orthosis (AFO) ipsilaterally. The rate of loading during vert ical loading [A] and the horizontal braking force [C] are highlighted by dotted lines ipsilaterally. The gray shaded area represents matched control data. A C B D

PAGE 92

92 -20.00 -15.00 -10.00 -5.00 0.00 5.00 10.00 15.00 20.00 I1I2I3I4I6I7I9I10 Subject ID Hip extension in degrees AFO No AFO MC Figure 5-4. Ipsilateral hip ex tension values with and without the AFO during the stance-toswing phase of the gait cycle for spinal co rd injured individuals and their matched controls.

PAGE 93

93 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 IpsilateralContralateralIpsilateralContralateral Unilateral AFO No AFO Step length in meters Unilateral AFO Ipsilateral Unilateral AFO Contralateral No AFO Ipsilateral No AFO Contralateral Figure 5-5. Step length while walking with and without the AFO in individua ls with incomplete spinal cord injury.

PAGE 94

94 0.25 0.30 0.35 0.40 0.45 0.50 0.55 IpsilateralContralateralIpsilateralContralateral Unilateral AFO No AFO Double limb support time in seconds Unilateral AFO Ipsilateral Unilateral AFO Contralateral No AFO Ipsilateral No AFO Contralateral Figure 5-6. Double limb support time while walki ng with and without the PAFO in individuals with incomplete spinal cord injury.

PAGE 95

95Table 5-1. Participant demographics of i ndividuals with ISCI and control subjects. ID Age Sex Height Orthot icInjury levelASIA score Injury duration months Speed (m/sec)Assitive/Orthotic device I1 46 M 5'6" L C5-C6 D 10 0.8 NA I2 33 M 5'11" R C6-7 D 14 0.6 Cane on left I3 66 M 6' 3" L C7 D 79 0.5 Cane on right I4 49 F 5'5" L C4-C5 D 46 0.7 NA I6 49 F 5'10" L C7 D 23 0.7 NA I7 40 F 5'8" L C2-T1 D 253 0.5 NA I9 25 M 5'11" R T4-5 D 90 0.4 Solid right AFO I10 57 M 6'2" L C5 D 122 0.3 Cane on right C1 M 5 L 0.8 C2 32 M 5'10" R 0.6 C3 62 M 6'3" L 0.5 C4 49 F 5'3" L 0.7 C6 52 F 5'7" L 0.7 C7 40 F 5'6" L 0.5 C9 27 M 5'11" R 0.4 C10 52 M 5"10" L 0.3

PAGE 96

96 Table 5-2. Hip, knee and ankle joint kinematic and kinetic data during the stance-to-swing phase of the gait cycle while walking with and w ithout an ankle foot orthosis (AFO) in individuals with incomplete spinal cord injury. Confidence Interval Stance-to-swing Without AFOWith AFO Lower Upper p-value Peak hip joint extension (degrees) -2.57 10.57 -1.18 9.72 -2.862 -.313 .022* Peak knee joint flexion (degrees) -48.14 8.94 -47.61 8.85-3.114 2.057 .643 Peak ankle joint plantarflexion (degrees) -6.32 8.23 -2.97 5.22 -6.863 .173 .059 Hip joint power (Watts/ Body weight) 0.04 0.02 0.04 0.02 -.006 .006 1.000 Knee joint power (Watts/ Body weight) -0.03 0.02 -0.03 0.01 -.006 .001 .170 Ankle joint power (Watts/ Body weight) 0.05 0.03 0.08 0.09 -.111 .051 .411 Represents significant changes

PAGE 97

97 Table 5-3. Hip, knee and ankle joint kinematic and kinetic data during the swingto-stance phase of the gait cycle while walking with and w ithout an ankle foot orthosis (AFO) in individuals with incomplete spinal cord injury. Confidence interval Swing-to-stance Without AFO With AFO Lower Upper p-value Peak hip joint flexion (degrees) 31.15 8.10 32.13 6.81 -3.096 1.126 .306 Peak knee joint flexion (degrees) -18.31 7.50 -19.82 6.16 .016 3.006 .048* Peak ankle joint plantarflexion (degrees) -6.73 6.98 -6.87 5.19 -2.637 2.917 .908 Rate of loading 0.04 0.01 0.04 0.01 -.001 .006 .170 Peak braking force -0.06 0.02 -0.06 0.03 -.011 .011 1.000 Represents significant changes

PAGE 98

98 Table 5-4. Kinematic and kineti c data during the stance-to-swing phase of the gait cycle at the hip, knee and ankle joints while walking w ithout an ankle foot orthosis (AFO) in individuals with incomplete spinal cord injury compared to their matched, noninjured controls. Confidence Interval Stance-to-swing Without AFO Control Lower Upper p-value Peak hip joint extension (degrees) -2.57 10.57 -5.61 6.42 -10.437 4.747 .405 Peak knee joint flexion (degrees) -48.14 8.94 -57 5.20 -15.327 -2.396 .014* Peak ankle joint plantarflexion (degrees) -6.32 8.23 -9.07 7.51 -12.602 7.087 .529 Hip joint power (Watts/ Body weight) 0.04 0.02 0.03 0.02 -.018 .003 .142 Knee joint power (Watts/ Body weight) -0.03 0.02 -0.05 0.02 -.037 -.001 .044* Ankle joint power (Watts/ Body weight) 0.05 0.03 0.07 0.03 -.009 .042 .178 Represents significant changes

PAGE 99

99 Table 5-5. Kinematic and kineti c data during the swing-to-stance phase of the gait cycle at the hip, knee and ankle joints while walking w ithout an ankle foot orthosis (AFO) in individuals with incomplete spinal cord injury compared to their matched, noninjured controls. Confidence interval Swing-to-stance Without AFO Control Lower Upper p-value Peak hip joint flexion (degrees) 31.15 8.10 24.93 6.18 -10.087 -2.353 .007* Peak knee joint flexion (degrees) -18.31 7.50 -14.98 4.57 -3.894 10.561 .312 Peak ankle joint plantarflexion (degrees) -6.73 6.98 -6.35 5.07 -5.455 6.230 .880 Rate of loading 0.04 0.01 0.04 0.01 -.007 .004 .598 Peak braking force -0.06 0.02 -0.07 0.02 -.015 .005 .275 Represents significant changes

PAGE 100

100 Table 5-6. Changes in the hip, knee and ankle joint kinematics a nd kinetics during the stance-toswing phase of the gait cycle while walking with an ankle foot orthosis (AFO) in individuals with incomplete spinal cord injury compared to their matched, noninjured controls. Confidence Interval Stance-to-swing With AFO Control Lower Upper p-value Peak hip joint extension (degrees) -1.18 9.72 -5.61 6.42 -11.556 2.694 .185 Peak knee joint flexion (degrees) -47.61 8.85 -57 5.20 -16.098 -2.675 .013* Peak ankle joint plantarflexion (degrees) -2.97 5.22 -9.07 7.51 -14.275 2.070 .121 Hip joint power (Watts/ Body weight) 0.04 0.02 0.03 0.02 -.018 .003 .142 Knee joint power (Watts/ Body weight) -0.03 0.01 -0.05 0.02 -.035 -.005 .018* Ankle joint power (Watts/ Body weight) 0.08 0.09 0.07 0.03 -.112 .085 .751 Represents significant changes

PAGE 101

101 Table 5-7. Kinematic and kineti c changes at the hip, knee and a nkle joints duri ng the swing-tostance phase of the gait cycle while walking without an ankle foot orthosis (AFO) in individuals with incomplete spinal cord injury compared to their matched, noninjured controls. Confidence interval Swing-to-stance With AFO Control Lower Upper p-value Peak hip joint flexion (degrees) 32.13 6.81 24.93 6.18 -11.159 -3.251 .004* Peak knee joint flexion (degrees) -19.82 6.16 -14.98 4.57 -1.533 11.218 .116 Peak ankle joint plantarflexion (degrees) -6.87 5.19 -6.35 5.07 -3.939 4.996 .788 Rate of loading 0.04 0.01 0.04 0.01 -.001 .006 .170 Peak braking force -0.06 0.03 -0.07 0.02 -.020 .005 .197 Represents significant changes

PAGE 102

102 Table 5-8. Average interlimb te mporal and spatial data while walking with and without an ankle foot orthosis (AFO) in individuals with incomplete spinal cord injury. No AFO Unilateral AFO Ipsilateral side Contralateral side pvalue Ipsilateral side Contralateral side pvalue Step length (meters) 0.31 0.02 0.31 0.03 .921 0.31 0.02 0.31 0.02 .301 Double support time (seconds) 0.36 0.09 0.36 0.14 .935 0.38 0.13 0.35 0.10 .273

PAGE 103

103 CHAPTER 6 PHASE DEPENDENT MODULATION OF SO LEUS HREFLEX IN HEALTHY, NONINJURED INDIVIDUALS WHILE WALKING WITH AN ANKLE FOOT ORTHOSIS Introduction Individuals with incomplete spinal cord inju ry (ISCI) have weakness and/or spasticity of the musculature below the level of injury maki ng it difficult to meet the functional demands of gait.68-70 In conventional rehabilitati on practice, spasticity or loss of muscle strength are substituted by compensatory orthotic devices th at stabilize, realign and control the range of excursion of the weakened joint or limb segment to assist with walking.15,75,78 For example, the posterior leaf spring ankle foot or thosis (PAFO) is used to compensate for deficient push off in terminal stance and foot drag in swing.16,17,203 With a PAFO, improvement in overground walking outcomes such as walking speed, stri de length, stance kn ee position and walking energetics have been documented.236 In spite of the appeal of such compensatory strategies, their use in neurobiologically driven, recovery-based interventions such as locomotor training for individuals with ISCI is still controversial.24 This is due to the lack of information about the use of the de vice in optimizing or hindering afferent input from joint, muscle and cutaneous receptors fundamental to the training.34,36,37,133 After SCI, pattern generati ng neural networks within the spinal cord increases their reliance on motion-related afferent input from these receptors for maintaining locomotor control.25,26,44 Limiting ankle excursion w ith a PAFO may alter the in terconnected limb joint assembly specific to walking and in turn negatively influence the afferent information critical for stepping. The soleus H-reflex has been commonly empl oyed as a neural probe in interpreting the interplay of afferent input and movement control.149,228,253,254 Elicitation of the reflex and measurement of its amplitude have provided in sights in spinal transmission during the

PAGE 104

104 performance of a motor task. During walking, the modulation of reflex amplitude is a measure of the regulation of afferent feedback dur ing different phases of the step cycle.146,228,229,253 For example, reflex amplitude increases to maxi mum during stance and decreases rapidly to minimum during swing. This regulation is requi red to accommodate the functional requirements of the task, i.e. facilitate weight support a nd plantarflexion during st ance while allowing ankle dorsiflexion during swing.146,228,229,253 In healthy, non-injured individuals, an increase in peroneal H-reflex amplitude has been observed while wearing a brace in static sitting or standing position suggestive of a heightened sensorimotor response due to stimulati on of the cutaneous mechanoreceptors.154,205,255 Increasing cutaneous input from the sole of the foot leads to a reduction in Ia presynaptic inhibition of the soleus muscle. The predicited outcome of this is facilitation of soleus H-reflex amplitude. Therefore, use of an orthotic device touching the plantar surfa ce of the foot and limiting the range of motion at the ankle coul d alter the rich sensory inform ation processed from the anklefoot complex and potentially modulate reflex ac tivity in non-injured i ndividuals. However, due to task-specific nature of H-reflex amplitude, it is difficult to extrapolate the results of a static task to the dynamic task of walking.256 Examination of reflex amplitude while walking with a PAFO will be useful in determining functional imp lication of the device in the task of walking. The purpose of this study was to examine the phase dependent modulation of the H-reflex in the gait cycle with and without a PAFO in non-injured individuals. Specific Aims Aim: To compare phase specific modulation of the soleus H-reflex amplitude in noninjured individuals while walki ng with and without a PAFO, Hypothesis: In non-injured individuals, soleus H-reflex amplitude while walking with an AFO will be significantly larger compared to the H-reflex amplitude without an AFO

PAGE 105

105 Methods Subject Selection A sample of convenience consisted of fourt een healthy, non-injur ed individuals living independently in the Gainesville community. Each participant provided informed consent before participating in the study. The University of Fl orida Institutional Review Board and the Veteran Affairs Subcommittee had approved the study for clinical investigati on. Age range for our participants was between 18-60 years. Participant demographics are tabulated in Table 6-1. This study did not include any subjects with any detectable gait and pos tural disorders. Subjects were screened for a medical history of any neurologic al, musculoskeletal or orthopedic problem that may affect their walking performance over the treadmill. Using the effect size and standa rd deviation from pilot data for non-injured s ubjects (H/M ratio post-pre =0.08, SD from previous study=0.11) to achieve statistical power of 80% at an alpha level of 0.05 we needed 14 normal subjects. Experimental Set-up Soleus H-reflexes were evoked, for the purpos e of consistency, on the dominant side of healthy, non-injured individuals subjects. Skin was shaved and cleaned for application of electrodes. A bipolar (2 cm inter-electrode di stance) AgAgCl surface el ectrode (Therapeutics Unlimited, Iowa City, Iowa) was placed longitudi nally over the soleus muscle. These electrodes are embedded in an epoxy mount w ith preamplifier circuitry and a 2-cm interelectrode distance. The preamplifier and second-stage amplifier provide a total amp lification of 1000 with a lowfrequency cut off of 20 Hz. To evoke H-reflexes, one millisecond current pul ses were delivered via a constant-current stimulator (Grass Instruments, model S8800 with a modified CCU 1) using a 2 cm 1/2 sphere silver cathode placed in the popliteal fossa and a 10 cm silver anode positioned just superior to

PAGE 106

106 the patella. The tibial nerve was localized, in the popliteal fossa by the electrode placement, to evoke a soleus H-reflex at the least current intensity required. Data were acquired at a sample rate of 10 kHz per channel and stored digitally with a commercially available data acquisition system (Data-Pac III by Run T echnologies) in a personal com puter (Dell Systems, Intel Celeron). Subject Preparation Subjects were asked to wear tennis shoes and change into appropriate clothing (dark colored cycling shorts and shirt). Skin was shaved and cleaned for application of surface electrodes. For trials using the PAFO, each subj ect was fitted with an off-the-shelf PAFO. The leg chosen for donning the PAFO and the order of testing with and without it was randomized for each subject. For the PAFO tria l, each subject was requested to wear a unilatera l, size-fitted PAFO. The insole of the shoe was removed in order to fit the AFO and to even out the limb length on both sides. Fitting of the AFO was assesse d by measuring fit inside shoe, length of the calf shell and that of the footplate. Standardized fitting included using an AFO whose length fits an inch to two below the fibular head when donned and whose footplate length extends till the tip of the toes.244 During treadmill walking, footswitches we re placed inside the shoes that were helpful in determining the phases of walking. Each subject was fitted with a body weight supporting harness equipped with an additional overhead safety catch. The harness and safety catch when used either with or without BWS provided safety to the person walking on the treadmill and held or caught the person if he or she lost their balance, stumbled or began to fall. Procedure The order of testing was randomized for walk ing with and without PAFO (Figure 6-1). Prior to eliciting reflexes during walking, H-reflex es were first elicited in static standing position for use as a control reference across trials.146,257 For this purpose, participants were asked to

PAGE 107

107 stand quietly and H-reflexes were collected in this position. Stimulus intensity was maximized and three maximum M-waves were recorded in the static condition. At least fifteen H-reflexes were then elicited at stimulus intensity within a range of 8-12 % of the M-max in the static position. A recruitment curve was constructed in th e static stance position to ensure that the Hreflex was on the ascending limb of the curve.138,149 During walking trials in either of the conditions, H-reflexes were elicited across ten time divided phases of the gait cycle determined by fo otswitches namely heel strike (HS), HS+100ms, HS+200ms, HS+300ms, HS+400 ms and toe-off (TO), TO+100ms, TO+200ms, TO+300ms, TO+400ms. The event and time point of s timulation was achieved by connecting the footswitches to a Schmidt trigger that sensed the event and delivered the pulse. For example, for the phase of heel strike the pul se was delivered at 0ms. For mi d-stance, the pulse was delivered after a time delay from heel strike. Once the subject began steppi ng on the treadmill at 1.2 m/s, three maximum M-waves were recorded in each of the ten phases of the gait cycle (Figure 6-1).227,257 These recordings were used in determining the stimulus intensity for each tested phase in the gait cycle and were also used for subsequent normalization of the da ta. Subsequently, stimulation was delivered at stimulus intensity within a range of 8-12 % of th e M-max calculated in each of the phases of the gait cycle.138,149 At least fifteen stimuli were delivered in a consecutive or an alternating fashion in each of the ten phases of the gait cycle.138,149 During testing, in both the static and walking condition, the activity in the soleus and TA muscle was recorded over a 100 ms dur ation prior to electrical stimulation.149 This activity was normalized to three maximum voluntary contractions of the TA and soleus collected at the beginning of the experiment. Also, the M-wave wa s constantly monitored to make readjustments

PAGE 108

108 to the stimulus intensity if required. A fifteen -minute sitting break was provided after completion of static and walking trials in one co ndition before proceeding with the other. Data Processing After filtering and rectification of the data, mean peak-to-peak amplitude of 10 H-reflexes for each phase of the examined gait cycle was calculated and comp ared between the two conditions (with and without PAFO). Prior to this comparison, the H-reflex values were normalized to M-max values procured in the respective phases (H/M ratio).149 Data Analysis For the ten phases of the gait cycle, a Hotelling's T2-test, which is a multivariate analogue of the paired t-test, was performed. The test is a multivariate extension of the Student's t-test for paired data in comparison of mean difference ve ctors, i.e. the differe nces of two or more dependent variables considered twice in the same subjects.245 The dependent variables compared include the H/Mmax amplitude in the10 phases of the gait cycle (HS, HS+100ms, HS+200ms, HS+300ms, HS+400ms, TO, TO+100ms, TO+2 00ms, TO+300ms, TO+400ms). Significant changes between the two conditions (with and w ithout PAFO) were identified using the Holms correction which corrects for multiple comparisons by adjusting alpha value.246 Additionally, a repeated measures ANOVA was also performed to compare M-max amplitude, actual M wave amplitude used for stimulation of the H-refl ex and electromyographic activity recorded 100ms prior to stimulation in the TA and soleus muscles between the two conditions across the gait cycle. Significance level was set at p< 0.05. The same analyses were repeated for the contralateral limb. Results Both ipsilaterally and contralaterally the mean H/M ratios were not significantly different between the two walking conditions (p>0.05) for any of the phases (Figures 6-2 through Figure

PAGE 109

109 6-5). The mean EMG of soleus and tibialis an terior muscles 100 ms pr ior to the electrical stimulation was not significan tly different in both conditions (p<0.05). Additionally, M-max amplitude (Figure 6-6) and the ac tual M wave amplitude used to evoke the soleus H-reflexes (Figure 6-7) were not signifi cantly different across the gait cy cle in both the conditions. Discussion Peripheral input from the muscle and cutaneous receptors of the ankle-foot complex have been known to modulate soleus Hreflex amplitude during different tasks. The soleus H-reflex is facilitated by excitation of the plantar cutaneous afferents lo cated around the heel.258 Like wise, the change in ankle joint angle has been shown to modulate H-re flex excitability.259,260,154 Therefore, use of an orthotic device touching the plantar surfa ce of the foot and limiting the range of motion at the ankle coul d alter the rich sensory inform ation processed from the anklefoot complex and potentially modulate reflex activity. Although we hypothesized that there would be an increase in soleus H-reflex amplitude with an ankle brace the results of our current study show that soleus H-reflex amplitude re mains unchanged while walking with a PAFO in healthy, non-injured individuals. Previous studies have shown that ankle braci ng impacts reflex amplitude in static tasks. For example, Nishikawa et al (1999) reported a 10% increase in peroneus longus (PL) H-reflex amplitude after application of a semi-rigid ankle support in th e seated, non-weight bearing position.205 The non-weight bearing position and the tes ting of the peroneus muscle may account for the differing results between the Nishikawa study and this work. Likewi se, Schneider et al reported an increase in soleus H-reflex amplit ude on passively imposing rapid knee flexion from static stance position.154 Interestingly, not all studies done in a static task have demonstrated an increase in H-reflex amplitude as a result of bracing the ankle joint. For example, Sefton et al found no effect of a semi-rigid ankle brace on the PL H-reflex during an inversion

PAGE 110

110 perturbation261 and no effect on soleus H-reflex during a single limb stance task.262 They rationalized their findings to the fact that the external ankle su pport provided an increase in mechanical stability, negating the need for ne uronal adaptation to maintain upright stance. The essential difference between our studies and prev ious studies was that our experiment involved the dynamic task of walking with or wit hout the brace rather than a static task. Since the brace-related afferent input did not alter the lowe r limb reflexes during walking, it appears that this reflex is centrally modulated in healthy, non-injured indi viduals. Therefore, even if a peripheral influence is show n to have an effect in one task, it does not follow that the same input will be effective in another task. Several studies in the literature have demonstrated more central modulation of reflex activity during locomotion a nd locomotor like tasks. For example, Garrett et al (1999) and Schneider et al (2000) reported that the soleus H -reflex amplitude did not change when the knee was braced thereby bloc king the normal excursion during locomotion.154,206 Similarly, Yang and Wheelan (1993) have shown that inactivity of the tibialis anterior or activity of the soleus muscle during the swing phase of gait did not aff ect phase-specific modulation of the soleus H-reflex during walking.254 In our study, a kinematically si gnificant decrease in ankle pl antarflexion and hip extension were observed as a result of walking with the brace compared to walk ing without one (Refer Figure 4-1). However, in non-injured individual s, kinematic changes during brace walking did not change the modulation pattern through out the gait cycle sugges tive of central modulation of reflex activity. In non-injured subjects, such an occurence where changes in afferent input from the periphery do not alter H-reflex excitability is probably for maintenance of the locomotor task. In the event of impaired locomo tor control as exists after SC I, the reflex modulation might change during brace walking. This is because the ability of the spinal cord to modulate sensory

PAGE 111

111 input and presynaptic inhibi tion are both altered post-SCI.263 Previous studies, have systematically reported that greate r H/M ratios were recorded in post-SCI subjects than recorded in non-injured controls.263-265 For example, electrica l excitation of the plantar cutaneous afferents has been shown to facilitate soleus H-reflexes in persons with SCI and depress reflex amplitude in non-injured subjects in sitting.266 Therefore, although, reflex modulation during walking did not change between conditions in non-injured individuals, ou r results may not extend to individuals with neurolog ical impairment. However, this remains to be tested experimentally. Limitations As has been advocated, soleus H-reflex in our study was evaluated during walking rather than at rest because the reflex undergoes task specific modulation. The factors that could be potential confounds in the study such as speed of walking, testi ng order, background EMG activity and stimulus intensity were controlled. However, extraneous peripheral afferent input from other that sources such as stimulus generate d perturbations or pain a ssociated with repeated stimulation could affect the measured H-refl ex amplitude. Also, although a maximal M wave was evoked in each of the tested phases for normalization purposes, a recruitment curve was not constructed during the walking tr ials. A recruitment curve would assess reflex modulation over the range of intensities during walking and also ensured evaluati on of the same proportion of the motor neuron pool. Although the testing phases were randomized, as a result of repeated stimulation, post activation depression of the refl ex could have occurred affecting the results of the study. However, this is unlikely because stud ies have shown that synaptic transmission from Ia fibers to motor neurons depends in a complex fashion on the rate of nerve impulses.267 During movement tasks, stimuli that produce one or tw o extra impulses in a neuron that is already conducting tens of impulses per second will not produce significant depression.256

PAGE 112

112 Conclusion In summary, non-injured individu als demonstrate central modulation of reflex activity which attenuates extraneous sensory input from the periphery for the maintenance of the locomotor task. A spinal injury disrupting supraspinal pa thways can affect this phase specific reflex modulation. In the presence of im paired central modulation, persis tent cutaneous input that is usually presynaptically inhibited during the ga it cycle could be facilitated reinforcing the walking related impairment. In the light of thes e findings, an ankle foot orthosis needs to be evaluated systematically in i ndividuals after SCI for the generation of reflex modulation characteristic of normal walking.

PAGE 113

113 Figure 6-1. Experimental design for the examination of changes in soleus H-reflex amplitude in healthy, non-injured individuals while wa lking with and without an ankle foot orthosis (AFO). Testing order randomized With AFO Without AFO N subjects Walking on treadmill at 1.2 m/s Standing H-reflexes 0ms 100ms 200ms 300ms 400ms 0ms 100ms 200ms 300ms 400ms

PAGE 114

114 Figure 6-2. Ipsilateral raw soleus H-reflex data while walking with and without AFO at 300ms from heel strike (HS) and toe-off (TO). Vertical red and blue guide bars capture soleus H-reflex event. Figure 6-3. Contralateral raw soleus H-reflex data while walking with and without AFO at 300ms from heel strike (HS) and toe-off (TO). Vertical red and blue guide bars capture soleus H-reflex event. Without AFO at HS+300ms With AFO at HS+300ms Without AFO at TO+300ms With AFO at TO+300ms Without AFO at HS+300ms With AFO at HS+300ms Without AFO at TO+300ms With AFO at TO+300ms

PAGE 115

115 0 10 20 30 40 50 60 70 80 90 100 HS100200300400TO100200300400Gait cycle (milliseconds)% Mmax AFO no AFO Standing Figure 6-4. Ipsilateral mean H -reflex amplitudes with and without AFO normalized to M-max in each phase of the gait cycle. The gait cycle is represented in 100ms increments from heel strike (HS) and toe off (TO). 0 10 20 30 40 50 60 70 80 90 100 HS100200300400TO100200300400Gait cycle (milliseconds)% M-max AFO no AFO Standing Figure 6-5. Contralateral mean H-reflex am plitudes with and without AFO normalized to Mmax in each phase of the gait cycle. The gait cycle is represented in 100ms increments from heel strike (HS) and toe off (TO).

PAGE 116

116 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00HS HS + 100 HS + 200 HS+ 300 HS+ 400 TO TO + 100 T O+ 200 TO+ 300 TO+400Stimulation phases of the gait cycleMmax amplitude Without AFO With AFO 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00HS HS+1 0 0 H S+ 20 0 HS+3 0 0 H S+ 40 0 TO TO + 100 T O+ 20 0 T O+ 30 0 TO+400Stimulation phases of the gait cycleMmax amplitude Without AFO With AFO Figure 6-6. Ipsilateral [A] and contralateral [B] M-max amplitude with and without the AFO across the gait cycle. The ga it cycle is represented in 100ms increments from heel strike (HS) and toe off (TO). A B

PAGE 117

117 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00HS HS +10 0 HS +20 0 HS+3 0 0 HS+4 0 0 TO TO+1 0 0 TO+200 TO+300 TO+400Stimulation phases of the gait cycle 8 to 12 % M-max value Without AFO With AFO 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00H S HS+100 HS+ 2 00 HS+300 HS+400 TO T O+ 1 0 0 TO+200 TO+ 3 00 T O+ 4 0 0Stimulation phases of the gait cycle8 to12 % of M max value Without AFO With AFO Figure 6-7. Ipsilateral [A] and contralateral [B] actual M wave amplit ude used to evoke the soleus H-reflex with and without the AFO across the gait cycle. The gait cycle is represented in 100ms increments from h eel strike (HS) and toe off (TO). A B

PAGE 118

118 Table 6-1. Demographics of non-injured participants recru ited for the study. Subject ID Age Sex Orthotic side N1 24 F R N2 27 M R N3 30 M R N4 28 F R N5 26 F R N6 31 F R N7 27 M R N8 24 F R N9 22 M R N10 27 F L N11 23 M L N12 29 M L N13 23 M L N14 36 F L

PAGE 119

119 CHAPTER 7 IMMEDIATE, PHASE DEPENDENT, SOLEUS H-REFLEX MODULATION IN PERSONS W ITH INCOMPLETE SPINAL CORD INJURY WHILE WALKING WITH AN ANKLE FOOT ORTHOSIS Introduction The soleus H-reflex has often been employed to examine the neural regulation of afferent information during walking.228,253 Elicitation of the reflex and m easurement of its amplitude have provided a method to evaluate the modulati on of the spinal path ways during walking.138 In healthy, non-injured individuals, the soleus H-reflex undergoes phase-s pecific modulation to accommodate to the functional requirements of the task.146,228,229 Specifically, the size of the soleus H-reflex is higher during the stance phase and lower during the swing phase of the step cycle. However, in individuals after ISCI, the so leus H-reflex demonstrates decreased depth of modulation. As a consequence, there is a reduced amplitude modulation across the step cycle and it simply remains increased throughout the step cycle.155,156,268 Consequently, lack of reflex modulation contributes to their walking impairments.155,156,268 During conventional gait training, individuals with ISCI are often prescribed orthotic devices to assist with walking. Or thotic devices such as an ankle foot orthosis (AFO) are used for assisting foot clearance, increasing gait speed, and improving walking endurance.16,269 Apart from mechanically aiding foot clearance,270 studies in healthy, non-in jured individuals suggest that an AFO increases afferent feedback from cutaneous receptors in the foot and shank to improve ankle positioning.255,271 For example, ankle bracing in a variety of static motor tasks such as sitting or standing have reported an increase in peroneal205,255 and soleus H-reflex amplitude in healthy, non-injured individuals.154 However, the increase in reflex amplitude with the brace has only been documented in healthy, non-injured individuals under static conditions limiting the extrapolation of results to persons with ISCI in dynamic tasks. In individuals with

PAGE 120

120 ISCI with impaired reflex modulation, an orthotic device touching the plan tar surface of the foot and limiting the range of motion at the ankle could alter the rich sensory information processed from the ankle-foot complex and potentially modulate reflex activity. Additionally, in persons with ISCI, cutaneous stimulation of the foot and sole while walking has been suggested as a potential method to rest ore reflex modulation comparable to that seen in healthy, non-injured individuals.155 Furthermore, simulation of walking kinematics using manual assistance, bodyweight support and a trea dmill in persons with ISCI improved reflex modulation and overground stepping speed without bracing the ankle.181 Consequently, conclusive evidence supporting or not supporting the use of the brace in improving phasespecific modulation during gait tr aining in individuals with ISCI is not apparent. Also, since afferent stimulation of one limb is known exert a c onsiderable influence on the reflex activity of the contralateral limb, the influe nce of an AFO on contralateral soleus H-reflex modulation also warrants investigation.153 Therefore, the immediate phase-dependent modu lation of the soleus H-reflex in persons with ISCI during walking w ith and without an AFO was examined ipsilaterally and contralaterally. Specifically, we hypothesized an increase in reflex amplitude only ipsilaterally while walking with the AFO compared to walk ing without one in individuals with ISCI. Specific Aims Aim 1: In persons with ISCI and ambulatory, to compare immediate phase-dependent modulation of the soleus H-reflex with and without an AFO in the mid-stance phase of walking. Hypothesis 1: In persons with ISCI, soleus H-reflex amplitude will be significantly larger in mid-stance while walking with an AFO co mpared to walking without an AFO. Aim 2: In persons with ISCI and ambulatory, to compare immediate phase-dependent modulation of the soleus H-refl ex in with and without an AFO in the mid-swing phase of walking.

PAGE 121

121 Hypothesis 2: In persons with ISCI, soleus H-reflex amplitude PAFO will be significantly larger in mid-swing while walking with an AFO compared to walk ing without an AFO. Methods Subject Selection Nine persons with ISCI rangi ng between 18-80 years were recr uited for this experiment and sign an informed consent form approved by Univ ersity of Florida Institutional Review Board and the Veteran Affairs Subcommittee for Clinical Investigation (Table7-1). American Spinal Injury Association (ASIA) moto r score and impairment scale data were collected from all participants with ISCI to assess the degree of impairment in each leg. The criteria for inclusion in the study were: Medically stable persons with ISCI classified as ASIA D, having quadriceps stre ngth of at least 3/5, having decreased ankle strength and/or impaired or absent proprioception at the ankle, having ankle dorsifle xor strength of < 4/5, able to stand unaided for one minute and walking with mi nimal assistive devices but not using an AFO. Exclusion criteria include persons who are una ble to follow 3 step commands, amputation, medical instability, significant musculoskeletal pr oblems other than SCI that limit hip and knee extension or ankle plan tarflexion to neutral. Experimental Set-up Soleus H-reflexes were evoked, for the purpos e of consistency, on the dominant side of healthy, non-injured individuals subjects. Skin was shaved and cleaned for application of electrodes. A bipolar (2 cm inter-electrode di stance) AgAgCl surface el ectrode (Therapeutics Unlimited, Iowa City, Iowa) was placed longitudi nally over the soleus muscle. These electrodes are embedded in an epoxy mount w ith preamplifier circuitry and a 2-cm interelectrode distance. The preamplifier and second-stage amplifier provide a total amp lification of 1000 with a lowfrequency cut off of 20 Hz.

PAGE 122

122 To evoke H-reflexes, one millisecond current pul ses were delivered via a constant-current stimulator (Grass Instruments, model S8800 with a modified CCU 1) using a 2 cm 1/2 sphere silver cathode placed in the popliteal fossa and a 10 cm silver anode positioned just superior to the patella. The tibial nerve was localized, in the popliteal fossa by the electrode placement, to evoke a soleus H-reflex at the least current intensity required. Data were acquired at a sample rate of 10 kHz per channel and stored digitally with a commercially available data acquisition system (Data-Pac III by Run T echnologies) in a personal com puter (Dell Systems, Intel Celeron). Subject Preparation Subjects were asked to wear tennis shoes and change into appropriate clothing (dark colored cycling shorts and shirt). Skin was shaved and cleaned for application of surface electrodes. For trials using the PAFO, each subj ect was fitted with an off-the-shelf PAFO. The leg chosen for donning the PAFO and the order of testing with and without it was randomized for each subject. For the PAFO tria l, each subject was requested to wear a unilatera l, size-fitted PAFO. The insole of the shoe was removed in order to fit the AFO and to even out the limb length on both sides. Fitting of the AFO was assesse d by measuring fit inside shoe, length of the calf shell and that of the footplate. Standardized fitting included using an AFO whose length fits an inch to two below the fibular head when donned and whose footplate length extends till the tip of the toes.244 During treadmill walking, footswitches we re placed inside the shoes that were helpful in determining the phases of walking. Each subject was fitted with a body weight supporting harness equipped with an additional overhead safety catch. The harness and safety catch when used either with or without BWS provided safety to the person walking on the treadmill and held or caught the person if he or she lost their balance, stumbled or began to fall.

PAGE 123

123 Procedure H-reflexes were elicited in two testing conditions (with and without PAFO) and two parts of gait cycle (mid-stance and mid-swing) during walking. The order of testing was randomized for with and without PAFO conditions (Figure 7-1) Treadmill speed was maintained constant for the both the testing conditions. Hreflexes were first be elicit ed in static standing position. Collecting, H-reflexes in static position served as a control refere nce across trials since the reflex is not modulated in a static position.146,257 For this purpose, particip ants were asked to stand quietly and H-reflexes were coll ected in this position. Stimul us intensity was maximized and three maximum M-waves were recorded in the sta tic standing condition. Fi fteen H-reflexes were then elicited at stimulus intensity within a rang e of 8-12 % of the M-max in the static position. A recruitment curve was constructed in the static stance position to ensure that the H-reflex was on the ascending limb of the curve.138,149 Once the subject began stepping on the treadmill at selfselected speed, three maximum M-waves were recorded in mid-stance phase and mid-swing phase respectively.227,257 These recordings were used to determine the stimulus intensity for each tested phase in the gait cycle and were also useful for subse quent normalization of the data.138,149 Subsequently, stimulation was delivered at stimulus intensity within a ra nge of 8-12 % of the M-max in the mid-stance and mid-swing phase of the gait cycle.138,149 At least 15 H-reflexes were recorded in each of the two selected phases of the gait cycle at each of the time points. 138,149 During testing, in both the static and walking condition, the activity in the soleus and TA muscle was recorded over durati on of 100 ms prior to electrical stimulation. This activity was normalized to the average of three maximum vol untary contractions of the TA and soleus collected at the start of the experiment. Also, the M-wave was constantly monitored to make readjustments to the stimulus intensity if required. Duration between two consecutive electrical

PAGE 124

124 stimulations were randomly maintained at a mi nimum of three seconds and maximum 5 seconds to avoid post activation de pression and habituation.138,149 After recording the H-reflexes in walking, the procedure for the static position was re peated again. After data collection, the data was stored for subsequent anal ysis. A fifteen-minute sitting br eak was provided after completion of static and walking testing in one cond ition before proceeding with the other. As part of a secondary question examini ng the effect of the ipsilateral brace on contralateral limb the same procedure was repeated on the contralateral side. Only eight of the above participants participated in this part of the study. Data Processing After filtering and rectification of the data, mean peak-to-peak amplitude of 10 H-reflexes for each phase of the examined gait cycle was calculated and comp ared between the two conditions (with and without PAFO). Prior to this comparison, the H-reflex values were normalized to M-max values procured in the respective phases (H/M ratio).149 Data Analysis For the mid-stance and mid-swing phase on the ip silateral limb, a paired t-test with bracing condition (with or without PAFO) as the independen t variable and the H-reflex amplitude as the dependent variable was performed. The same analysis was repeated for the contralateral limb. Significant changes between the two conditions (with and without PAFO) were identified using the Holms step-down correction which corrects for multiple comparisons by adjusting alpha value.246 Additionally, a repeated measures ANOVA was also performed to compare M-max amplitude, actual M wave amplitude used for stim ulation of the H-reflex and electromyographic activity recorded 100ms prior to stimulation in the TA and so leus muscles between the two conditions across the gait cycle. Si gnificance level was set at p<0.05.

PAGE 125

125 Results Figure 7-2 through Figure 7-4 shows the ipsila teral and contralateral H-reflex modulation in mid-stance and mid-swing phase with and without the PAFO ip silaterally. After correcting for multiple comparisons, the mean H/M ratio was si gnificant importantly for the ipsilateral midswing phase of walking (without PAFO: 0.13.10 & with PAFO 0.29.14).The values for the H/M ratio for all the tested conditions are repor ted in Table 7-2 and Table 7-3. The mean EMG of soleus and tibialis anterior muscles 100 ms prio r to the electrical stim ulation did not change systematically in both conditi ons (Tables 7-4 through Table 7-7). Furthermore the M-max amplitude and the actual Mwave amplitude used for evoking the soleus H-reflex did not change significantly (Figure 7-5 and Figure 7-6). Discussion This is the first study to systematically exam ine the effect of bracing on soleus H-reflex modulation during the task of walking in individuals with ISCI The main finding of the study was that, there was a significant increase in sole us H-reflex amplitude ipsilaterally in the midswing phase while walking with a PAFO. In the abse nce of a change in the ankle-foot orientation or stretch at the ankle joint with a PAFO, these re sults are suggestive of an increase in afferent inflow in the mid-swing phase of walking. In our study, the background EMG activity in both the walking conditions (with and without the PAFO) was similar de spite the changes in H-reflex amplitude reinforcing the fact that the modulation of the reflex is not direct ly dependent on the excita tion level of the alphamotorneurons.146 Likewise the presence or absence of ankle clonus did not affect reflex amplitude in the two walking conditions. Subject I9 who had clonus had a change in mid-swing reflex amplitude similar to subject I4 who did not demonstrate any clonic activity in the soleus muscle.

PAGE 126

126 In individuals with ISCI, an increase in re flex amplitude with the PAFO during the midswing phase could be the result of altered pr esynaptic inhibition. Pr esynaptic inhibition is important because during walking sensory input from cutaneous and proprioceptive receptors continuously converges on the spinal circuits.263,272 These surplus inputs must either be synergistically combined with the motor comman ds or be appropriately suppressed to minimize interference. Since only certain input requires se lective modulation, pres ynaptic inhibition of sensory input allows suppression of specific i nputs to a neuron without influencing other synaptic inputs. The predominant sources of pr esynaptic inhibition are peripheral inputs from cutaneous afferents and ce ntral descending pathways.272,273 During walking, cutaneous input from the foot sole is known to modulate reflex activity and change mu scle synergies thereby contributing to adaptive locomotor strategies. For example, Bastiaanse et al (2000) observed that load receptors are involved in th e regulation of cutaneous reflex responses in order to adapt the locomotor pattern to the environmental conditions.274 The repertoire of adaptive movement strategies is usually limited in individuals with ISCI because the ability of the spinal cord to modul ate sensory input and presynaptic inhibition are usually impaired post-injury.263,275 Previous studies, have system atically reported that greater H/M ratios were recorded in post-SCI subjects than reco rded in non-injured controls.156,263-265 The increase in reflex amplitude during brace wa lking in our study may have occurred because of the stimulation of plantar cu taneous afferents caused by the PAFO. Excitation of the plantar cutaneous afferents facilitates soleus H -reflex in persons with SCI in sitting.266 In the absence of supraspinal modulation of reflex activity, peripheral cutaneous inputs may be beneficial to modulate reflex activity.155 Nakajima et al (2006) has shown that reflex connections from cutaneous nerves in the foot on to the lower limb muscles are arranged in a highly topographical

PAGE 127

127 manner and may play an important role in limb lo ading and ground contact in response to tactile sensation.276 However, from the results of our study two interesting observations can be made about the PAFO during walking. An increase in reflex amplitude observed in the mid-swing phase during walking with a PAFO suggests that a simple, offthe-shelf orthosis could potentia lly increase afferent inflow and modulate reflex activity in high functioning individua ls with ISCI. However, in individuals with shallow modulation of reflex activity throughout the gait cycle, the increas e in afferent inflow especially during the mid-swing phase might be unfavorable for reflex modulation and ultimately walking.156 Therefore, prior to its use, the type a nd purpose of orthotic device, the movement task of interest such as walking or cycling and the targeted populati on needs to be assessed carefully. Interestingly, no increase in reflex amplitude was noted with the brace in the contralateral limb in mid-stance or mid-swing. Also motion data collected prev iously did not show a change in limb kinematics with or without the PAFO contralate rally. Our inference further strengthens the idea of a localized cutaneous response of the PAFO on the reflex modulation ipsilaterally. Clinical Implications Soleus H-reflexes are exaggerated post-SCI If reflex dysregulation is secondary to disruption of supraspinal inhibito ry control mechanisms, then tr aining strategies inhibiting the hyperactive reflex segmentally in a task and pha se specific manner may be beneficial for the proper restoration of locomotion. Locomotor trai ning is one such strategy that works on the above principle of providing optimal sensory inpu t to the nervous system to recover walking ability after ISCI. The tr aining provides sensory cues and phasic information related to locomotion. One of the cues pivotal to training is to minimize sensory stimulation that would conflict with sensory information asso ciated with locomotion.155,277,278 The potency of the

PAGE 128

128 training stems from the fact th at a single bout of locomotor tr aining is capable of producing significant depression of the exaggerated soleus H-reflexes and improved walking speed in persons with ISCI.181 Therefore, research efforts are being directed at systematically determining the critical components of locomotor training, that optimize the provisi on of critical sensory input that aid walking recovery. W ithin the realms of this goal, our study demonstrates that the integration of clinically ac knowledged stepping aids such as the PAFO's during locomotor training could be counter productive to the recovery of walking pos t SCI. Therefore such devices should be chosen only after careful consid eration of outcome for training purposes. Limitations First, we only examined the effect of the PA FO in the mid-stance and mid-swing phase of walking limiting our inferences to only two specific phases of walking. A thorough examination in different phases of the gait cycl e might reveal the unique effects of the PAFO within the entire gait cycle. Second, for standard ization purposes, we examined th e effect of only one type of AFO which limits generalizability to other t ypes of AFO's. Based on their impairments individuals with ISCI might us e customized AFOs which coul d yield different results. Conclusions In persons with ISCI, soleus H-reflex amplitude increased significantly in the mid-swing phase of walking with an AFO compared to walk ing without it. Ou r findings suggest that, in the presence of impaired central and peripheral modulat ion of reflex activity, an ankle foot orthosis that provides persistent cutaneous inputs from the foot sole might contribute to modulating reflex activity. However, increase in a fferent input in certai n phases of the gait cycle might not always be favorable to the task of walking. Therefore therapeutic interventions targeted at promoting walking recovery in individuals with ISCI should carefully consider the use of such nonadaptable, compensatory orthotic devices that could potentially hinder re training or reeducation

PAGE 129

129 of function. Instead, adaptable st rategies that have been documen ted to provide appropriate phase specific sensory input such as functional cutaneou s stimulation of the foot or appropriate cueing using manual assistance should be incorporated during training to assi st with walking and promote recovery.

PAGE 130

130 Figure 7-1. Experimental design for testing the effect of walk ing with and without an ankle foot orthosis (AFO) in individuals with incomplete spinal cord injury (ISCI) at their selfselected (SS) walking speed. Testing order randomized With AFO Without AFO Mid-stance N subjects Standing H-reflex H-reflex in walking at SS speed Mid-swing

PAGE 131

131 0 10 20 30 40 50 60 70 80 90 100 MidstanceMidswing Phase of stimulation in the gait cycle% Mmax noAFO AFO standing Figure 7-2. Average H/M ratio values with a nd without an AFO in mi d-stance and mid-swing phase of walking relative to static standing in the ipsilateral limb. 0 10 20 30 40 50 60 70 80 90 100 MidstanceMidswing Phase of stimulation in the gait cycle% Mmax noAFO AFO standing Figure 7-3. Average H/M ratio values with and without AFO in mi d-stance and mid-swing phase of walking relative to static standing in the c ontralateral limb.

PAGE 132

132 Primary side Contralateral side Figure 7-4. Ipsilateral and contra lateral raw soleus H-reflex data while walking with and without AFO during mid-stance (MSt) and mid-swing (M Sw) phase of the gait cycle. Vertical red and blue guide bars captur e soleus H-reflex event. MSw with AFO MSw without AFO MSt without AFO MSt with AFO MSw with AFO MSw without AFO MSt without AFO MSt with AFO

PAGE 133

133 0 1 2 3 4 5 6 7 8 9 Mid-stanceMid-swing Phase of stimulation in the gait cycleMmax amplitude No AFO AFO 0 1 2 3 4 5 6 7 8 Mid-stanceMid-swing Phase of stimulation in the gait cycleMmax amplitude No AFO AFO Figure 7-5. Ipsilateral [A] and contralateral [B ] M-max amplitude with and without the AFO in the mid-stance and mid-swing phase of walking. A B

PAGE 134

134 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Mid-stanceMid-swing Phase of stimulation in the gait cycle8 to 12% of Mmax value No AFO AFO 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Mid-stance Mid-swing Phase of stimulation in the gait cycle8 to 12% of Mmax value No AFO AFO Figure 7-6. Ipsilateral [A] and contralateral [B] actual M wave amplit ude used to evoke the soleus H-reflex with and without the AFO across the gait cycle. A B

PAGE 135

135 Table 7-1. Participant demographics. ID Age Sex Height Orthotic Injury level ASIA score Injury duration in months Speed (m/sec) Assitive/Orthotic device I2 33 M 5'11" R C6-7 D 14 0.6 Cane on left I3 66 M 6' 3" L C7 D 79 0.5 Cane on right I4 49 F 5'5" L C4-5 D 46 0.7 NA I5 59 M 6'0" R C5-6 D 58 0.4 Cane on left I6 49 F 5'10" L C7 D 23 0.7 NA I7 40 F 5'8" L C2-T1 D 253 0.5 NA I8 59 F 5" R C6 D 180 0.3 NA I9 25 M 5'11" R T4-5 D 90 0.4 Solid AFO on right I10 57 M 6'2" L C5 D 122 0.3 Cane on right ASIA: American Spinal Injury Association Table 7-2. Ipsilateral H/M rati o with and without the AFO. Subject No AFO AFO ID Standing Midstance Mid-swingStanding Midstance Mid-swing I2 0.71 0.59 0.28 0.59 0.58 0.31 I3 0.35 0.20 0.09 0.40 0.40 0.16 I4 0.69 0.53 0.02 0.68 0.68 0.39 I5 0.68 0.51 0.24 0.78 0.59 0.40 I6 0.05 0.04 0.04 0.03 0.07 0.05 I7 0.08 0.10 0.04 0.04 0.21 0.19 I8 0.79 0.66 0.20 0.77 0.70 0.41 I9 0.70 0.58 0.21 0.68 0.60 0.49 I10 0.36 0.48 0.10 0.41 0.50 0.23 Avg 0.49 0.41 0.13 0.49 0.48 0.29 Std.Dev 0.29 0.23 0.10 0.29 0.22 0.14

PAGE 136

136 Table 7-3. Contralateral H/M ra tio with and without the AFO. Subject No AFO AFO ID Standing Mid-stance Mid-swing Standing Mid-stance Mid-swing I2 0.31 0.45 0.04 0.26 0.41 0.02 I3 0.40 0.45 0.16 0.43 0.50 0.17 I4 0.51 0.66 0.05 0.55 0.65 0.02 I5 0.22 0.19 0.02 0.16 0.23 0.02 I6 0.44 0.19 0.03 0.48 0.51 0.04 I7 0.06 0.10 0.01 0.06 0.03 0.01 I8 0.36 0.13 0.04 0.23 0.14 0.03 I9 0.70 0.71 0.06 0.60 0.68 0.08 Avg 0.38 0.36 0.05 0.35 0.39 0.05 Std.dev 0.19 0.24 0.05 0.20 0.24 0.05 Table 7-4. Normalized soleus EMG amp litude with and without AFO ipsilaterally. Soleus Standing Midstance Midswing Subject ID without AFO with AFO without AFO with AFO without AFO with AFO I2 0.12 0.16 0.22 0.13 0.23 0.16 I3 0.04 0.05 0.05 0.05 0.05 0.05 I4 0.14 0.13 0.15 0.20 0.16 0.17 I5 0.13 0.12 0.16 0.14 0.14 0.14 I6 1.60 1.65 0.28 0.31 1.25 1.22 I7 4.68 4.73 4.68 4.73 4.69 4.73 I8 2.17 2.17 2.26 2.19 2.18 2.17 I9 2.64 2.66 2.65 2.66 2.66 2.67 I10 NT NT NT NT NT NT Table 7-5. Normalized TA EMG amplit ude with and without AFO ipsilaterally. TA Standing Midstance Midswing Subject ID without AFO with AFO without AFO with AFO without AFO with AFO I2 0.14 0.12 0.14 0.13 0.21 0.17 I3 0.10 0.12 0.11 0.10 0.12 0.10 I4 0.96 0.96 0.96 0.95 0.97 0.96 I5 0.30 0.40 0.38 0.41 0.42 0.44 I6 0.02 0.01 0.02 0.03 0.02 0.02 I7 0.73 0.62 0.48 0.57 0.60 0.55 I8 0.46 0.43 0.42 0.44 0.46 0.44 I9 0.56 0.74 0.79 0.89 0.82 0.90 I10 NT NT NT NT NT NT

PAGE 137

137 Table 7-6. Normalized soleus EMG amplit ude with and without AFO contralaterally. Soleus Standing Midstance Midswing Subject ID without AFO with AFO without AFO with AFO without AFO with AFO I2 0.14 0.16 0.16 0.18 0.15 0.16 I3 0.09 0.09 0.09 0.09 0.08 0.08 I4 0.12 0.08 0.19 0.09 0.26 0.37 I5 0.12 0.13 0.15 0.15 0.10 0.12 I6 0.10 0.12 0.11 0.12 0.33 0.56 I7 3.33 3.33 3.33 3.34 3.56 3.39 I8 1.60 1.59 1.61 1.61 1.59 1.62 I9 1.37 1.37 1.37 1.37 1.37 1.37 I10 NT NT NT NT NT NT Table 7-7. Normalized TA EMG amplitude with and without AFO contralaterally. TA Standing Midstance Midswing Subject ID without AFO with AFO without AFO with AFO without AFO with AFO I2 0.15 0.17 0.18 0.18 0.22 0.17 I3 0.48 0.45 0.43 0.39 0.46 0.44 I4 0.47 0.46 0.47 0.47 0.47 0.47 I5 0.57 0.57 0.56 0.56 0.54 0.58 I6 0.44 0.44 0.43 0.44 0.42 0.44 I7 1.25 1.24 1.22 1.25 1.34 1.26 I8 0.54 0.48 0.44 0.42 0.46 0.43 I9 0.68 0.61 0.67 0.67 0.59 0.61 I10 NT NT NT NT NT NT

PAGE 138

138 LIST OF REFERENCES 1. Forssberg, H. in Brainstem control of spinal mechanisms (eds. Sjolund, B. & Bjorklund, R. A.) 253-71 (Elsevier Biomedical, Amsterdam, 1982). 2. Patla, A. E., Prentice, S., Robinson, C. & Neufeld, J. Visual control of locomotion: strategies for changing direc tion and for going over obstacles. J Exp Psychol Hum Percept Perform 17, 603-34 (1991b). 3. Barbeau, H. Locomotor training in neuroreh abilitation: emerging re habilitation concepts. Neurorehabil Neural Repair 17, 3-11. (2003). 4. Bell, D. T. Principles an d techniques of physiotherapy in the treatment of spinal paraplegia. Phys Ther Rev 35, 235-7 (1955). 5. van der Salm, A. et al. Gait impairments in a group of patients with incomplete spinal cord injury and their relevance regardi ng therapeutic approaches using functional electrical stimulation. Artif Organs 29, 8-14 (2005). 6. Barbeau, H., Ladouceur, M., Norman, K. E., Pe pin, A. & Leroux, A. Walking after spinal cord injury: evaluation, treat ment, and functional recovery. Arch Phys Med Rehabil 80, 225-35 (1999). 7. Sekhon, L. H. & Fehlings, M. G. Epidem iology, demographics, and pathophysiology of acute spinal cord injury. Spine 26, S2-12 (2001). 8. Waters, R. L., Adkins, R. H., Yakura, J. S. & Sie, I. Motor and sensory recovery following incomplete tetraplegia. Arch Phys Med Rehabil 75, 306-11 (1994). 9. Waters, R. L., Adkins, R. H., Yakura, J. S. & Sie, I. Motor and sensory recovery following incomplete paraplegia. Arch Phys Med Rehabil 75, 67-72 (1994). 10. Raineteau, O. & Schwab, M. E. Plasticity of motor systems after incomplete spinal cord injury. Nat Rev Neurosci 2, 263-73 (2001). 11. Craik, R. L. in Contemporary management of motor problems (ed. Liser, M. J.) 165-74 (Bookcrafters, Inc., Fr edericksburg, VA, 1991). 12. Gordon, J. in Movement science foundations for phy sical therapy in rehabilitation (eds. Carr, J. H. & Shepherd, R. B.) 1-30 (Asp en Publishers Inc, Rockville, 1987). 13. Basso, D., Behrman, AL, Harkema, SJ. Rec overy of walking after central nervous sytem insult: basic research in the controlle d locomotion as a foundation for developing rehabilitation strategies. Neurology Report 24, 47-54 (2000). 14. Somers, M. Spinal cord injury: functional rehabilitation (Prentice-Hall Inc., London, 2001).

PAGE 139

139 15. O' Sullivan, S. B. & Schmitz, T. J. Physical rehabilitation assessment and treatment (eds. O' Sullivan, S. B. & Schmitz, T. J.) (F.A. Davis Company, Philadelphia, 2000). 16. Atrice, M. B. Lower extremity orthotic ma nagement for the spinal-cord-injured client. Spinal Cord Inj Rehabil 5, 1-10 (2000). 17. Lehmann, J. F. Biomechanics of ankl e-foot orthoses: pres cription and design. Arch Phys Med Rehabil 60, 200-7 (1979). 18. Barbeau, H. & Rossignol, S. Recovery of lo comotion after chronic spinalization in the adult cat. Brain Res 412, 84-95 (1987). 19. Barbeau, H. & Blunt, R. A novel interactive locomoto r aproach using body weight support to retrain gait in spastic paretic subjects (ed. Wernig, A.) (Elsevier Science Publishers, 1991). 20. Barbeau, H., Danakas, M. & Arsenault, B. The effects of locomotor training in spinal cord injured subjects : a preliminary study. Restor Neurol Neurosci 5, 81-84 (1993). 21. Dobkin, B. H. Neuroplasticity. Key to recovery after central nervous system injury. West J Med 159, 56-60 (1993). 22. Edgerton, V. R. et al. Use-dependent pl asticity in spinal stepping and standing. Adv Neurol 72, 233-47 (1997). 23. Basso, M. Neuroplasticity of descending and segmental systems after spinal cord contusion. neurology Report 22, 48-53 (1998). 24. Behrman, A. L., Bowden, M. G. & Nair, P. M. Neuroplasticity after spinal cord injury and training: an emerging paradigm shift in rehabilitation and walking recovery. Phys Ther 86, 1406-25 (2006). 25. Behrman, A. L. & Harkema, S. J. Locomo tor training after human spinal cord injury: a series of case studies. Phys Ther 80, 688-700. (2000). 26. Behrman, A. L. et al. Locomotor trai ning progression and outcomes after incomplete spinal cord injury. Phys Ther 85, 1356-71 (2005). 27. Barbeau, H. et al. Tapping into spinal circuits to restore motor function. Brain Res Brain Res Rev 30, 27-51 (1999). 28. De Leon, R. D., Hodgson, J. A., Roy, R. R. & Edgerton, V. R. Full weight-bearing hindlimb standing following stand tr aining in the adult spinal cat. J Neurophysiol 80, 8391 (1998). 29. Edgerton, V. R. et al. Potential of adult mammalian lumbosacral spinal cord to execute and acquire improved locomotion in the absence of supraspinal input. J Neurotrauma 9 Suppl 1, S119-28 (1992).

PAGE 140

140 30. Zehr, E. P. Neural control of rhythmic human movement: the common core hypothesis. Exerc Sport Sci Rev 33, 54-60 (2005). 31. Van de Crommert, H. W., Mulder, T. & D uysens, J. Neural control of locomotion: sensory control of the central pattern generator and its relation to treadmill training. Gait Posture 7, 251-63 (1998). 32. Dietz, V. Human neuronal control of automatic functional movements: interaction between central programs and afferent input. Physiol Rev 72, 33-69 (1992). 33. Perry, J. Gait analysis: normal and pathological function (Slack, Thorofare,NJ, 1992). 34. Duysens, J., Van de Crommert, H. W., Smits -Engelsman, B. C. & Van der Helm, F. C. A walking robot called human: lessons to be l earned from neural control of locomotion. J Biomech 35, 447-53 (2002). 35. Duysens, J., Clarac, F. & Cruse, H. Load-regulating mechanisms in gait and posture: comparative aspects. Physiol Rev 80, 83-133 (2000). 36. Duysens, J. Reflex control of locomoti on as revealed by stimulation of cutaneous afferents in spontaneously walking premammillary cats. J Neurophysiol 40, 737-51 (1977). 37. Edgerton, V. R., Tillakaratne, N. J., Bigbee, A. J., de Leon, R. D. & Roy, R. R. Plasticity of the spinal neural circuitry after injury. Annu Rev Neurosci 27, 145-67 (2004). 38. Pang, M. Y. & Yang, J. F. The initiation of the swing phase in human infant stepping: importance of hip position and leg loading. J Physiol 528 Pt 2, 389-404 (2000). 39. Dietz, V. Locomotor training in paraplegic patients. Ann Neurol 38, 965 (1995). 40. Barbeau, H., Fung, J., Leroux, A. & Ladouceur, M. A review of the adaptability and recovery of locomotion after spinal cord injury. Prog Brain Res 137, 9-25 (2002). 41. Wernig, A., Muller, S., Nanassy, A. & Ca gol, E. Laufband thera py based on 'rules of spinal locomotion' is effective in spinal cord injured persons. Eur J Neurosci 7, 823-9 (1995). 42. Wernig, A. & Muller, S. Laufband lo comotion with body weight support improved walking in persons with severe spinal cord injuries. Paraplegia 30, 229-38 (1992). 43. Barbeau, H., Wainberg, M. & Finch, L. De scription and application of a system for locomotor rehabilitation. Med Biol Eng Comput 25, 341-4 (1987). 44. Harkema, S. J. Neural plasticity after huma n spinal cord injury: a pplication of locomotor training to the rehab ilitation of walking. Neuroscientist 7, 455-68 (2001).

PAGE 141

141 45. Chen, Y. et al. The interaction of a new motor skill and an old one: H-reflex conditioning and locomotion in rats. J Neurosci 25, 6898-906 (2005). 46. Hidler, J. M. What is ne xt for locomotor-based studies? Journal of Rehabilitaion Research and Development 42, 10-6. (2005). 47. Stover, S. L., DeLisa, J. A. & Whiteneck, G. G. Spinal cord injury: clinical outcomes from the model systems (Aspen Publishers, Gaithesburg (MD), 1995). 48. Staas, W. E., Freedman, M. K. & Fried, G. F. Spinal cord injury and spinal cord injury medicine. (eds. DeLisa, J. A. & Gans, B. M.) (Lippincott-Raven Publishers., Philadelphia, PA, 1998). 49. Burney, R. E., Maio, R. F., Maynard, F. & Karunas, R. Incidence, characteristics, and outcome of spinal cord injury at trauma centers in North America. Arch Surg 128, 596-9 (1993). 50. Waters, R. L., Adkins, R. H., Yakura, J. & Sie, I. Profiles of spinal cord injury and recovery after gunshot injury. Clin Orthop Relat Res, 14-21 (1991). 51. Waters, R. L., Sie, I., Adkins, R. H. & Yakur a, J. S. Motor recovery following spinal cord injury caused by stab wounds: a multicenter study. Paraplegia 33, 98-101 (1995). 52. McDonald, J. W. & Sadows ky, C. Spinal-cord injury. Lancet 359, 417-25. (2002). 53. Moore, A. P. & Blumhardt, L. D. A prospective survey of the causes of non-traumatic spastic paraparesis and tetraparesis in 585 patients. Spinal Cord 35, 361-7 (1997). 54. McKinley, W. O., Seel, R. T. & Hardman, J. T. Nontraumatic spinal cord injury: incidence, epidemiology, and functional outcome. Arch Phys Med Rehabil 80, 619-23. (1999). 55. New, P. W., Rawicki, H. B. & Bailey, M. J. Nontraumatic spinal cord injury: demographic characteristics and complications. Arch Phys Med Rehabil 83, 996-1001 (2002). 56. DeVivo, M. J. & Stover, S. L. in Spinal cord injury: clinical outcomes from the model systems. (eds. Stover, S. L., DeLisa, J. A. & Whiteneck, G. G.) 289316. (Aspen Publishers, Inc., Gaithersburg, Maryland, 1995). 57. Annual report (Nati onal Spinal Cord Injury Statistical Center, Birmingham, Alabama, 2004). 58. Maynard, F. M., Jr. et al. Internationa l standards for neurol ogical and functional classification of spinal cord injury. American Spinal Injury Association. Spinal Cord 35, 266-74 (1997).

PAGE 142

142 59. Ditunno, J. F., Jr., Young, W., Donovan, W. H. & Creasey, G. The international standards booklet for neurologi cal and functional classification of spinal cord injury. American Spinal Injury Association. Paraplegia 32, 70-80 (1994). 60. Standards for neurological classifica tion of spinal injury patients. (Chicago, IL, 1984). 61. Burns, S. P., Golding, D. G., Rolle, W. A., Jr., Graziani, V. & Ditunno, J. F., Jr. Recovery of ambulation in motor-i ncomplete tetraplegia. Arch Phys Med Rehabil 78, 1169-72 (1997). 62. Crozier, K. S., Graziani, V., Ditunno, J. F ., Jr. & Herbison, G. J. Spinal cord injury: prognosis for ambulation based on sensory exam ination in patients who are initially motor complete. Arch Phys Med Rehabil 72, 119-21 (1991). 63. Dobkin, B. et al. The evolution of walki ng-related outcomes over th e first 12 weeks of rehabilitation for incomplete traumatic sp inal cord injury: th e multicenter randomized Spinal Cord Injury Locomotor Trial. Neurorehabil Neural Repair 21, 25-35 (2007). 64. Daverat, P. et al. Early prognostic fact ors for walking in sp inal cord injuries. Paraplegia 26, 255-61 (1988). 65. Waters, R. L., Adkins, R., Yakura, J. & Vi gil, D. Prediction of ambulatory performance based on motor scores derived from standards of the American Spinal Injury Association. Arch Phys Med Rehabil 75, 756-60 (1994). 66. Crozier, K. S. et al. Spinal cord inju ry: prognosis for ambulation based on quadriceps recovery. Paraplegia 30, 762-7 (1992). 67. Maynard, F. M., Reynolds, G. G., Fountain, S., Wilmot, C. & Hamilton, R. Neurological prognosis after traumatic quadriplegia. Thr ee-year experience of California Regional Spinal Cord Injury Care System. J Neurosurg. 50, 611-6 (1979). 68. Conrad, B., Benecke, R. & Meinck, H. M. Gait disturbances in paraspastic patients (eds. Delwaide, P. J. & Young, R. R.) (Elsevier Science Publishers, 1985). 69. Fung, J. & Barbeau, H. A dynamic EMG profile index to quantify muscular activation disorder in spastic paretic gait. Electroencephalogr Clin Neurophysiol 73, 233-44 (1989). 70. Krawetz, P. & Nance, P. Gait analysis of sp inal cord injured subjects: effects of injury level and spasticity. Arch Phys Med Rehabil 77, 635-8 (1996). 71. Nene, A. V., Hermens, H. J. & Zilvol d, G. Paraplegic locomotion: a review. Spinal Cord 34, 507-24 (1996). 72. Organization, W. H. (2001). 73. Umphred, D. A. Neurological rehabilitation (Mosby Inc., St.Louis, 2001).

PAGE 143

143 74. Mytrice, B. A. Lower extremity orthotic management for the spinal-cord-injured client. Spinal Cord Inj Rehabil 5(4), 1 -10 (2000). 75. Nawoczenski, D. A. & Epler, M. E. Orthotics in functional re habilitation of the lower limb (W.B. Saunders, Philadelphia, 1997). 76. Maxwell, D. J., Granat, M. H., Baardman, G. & Hermens, H. J. Demand for and use of functional electrical stimulation systems and conventional orthoses in the spinal lesioned community of the UK. Artif Organs 23, 410-2 (1999). 77. Ramon y Cajal, S. Degeneration and regeneration of the nervous system. (Oxford University Press, London, 1928). 78. Guide to physical therapy practice (American Physical Therapy Association, Fairfax, VA, 2001). 79. van Tuijl, J. H., Janssen-Potten, Y. J. & S eele, H. A. Evaluation of upper extremity motor function tests in tetraplegics. Spinal Cord 40, 51-64. (2002). 80. Kasaoka, K. & Sankai, Y. in IEEE/RSJ International Conference on Intelligent Robots and Systems (2001). 81. Kawamoto, H. & Sankai, Y. in 13th IEEE International Workshop on Robot and Human Interactive Communication (2004). 82. Ferris, D. P., Sawicki, G. S. & Domingo, A. Powered lower limb orthoses for gait rehabilitation. Top Spinal Cord Inj Rehabil 11, 34-49 (2005). 83. Grillner, S. Interaction between central and peripheral mechanisms in the control of locomotion. Prog Brain Res 50, 227-35 (1979). 84. Grillner, S. Neurobiological bases of rhythmic motor acts in vertebrates. Science 228, 143-9 (1985). 85. Wolpaw, J. R. Acquisition and maintenance of the simplest motor sk ill: investigation of CNS mechanisms. Med Sci Sports Exerc 26, 1475-9 (1994). 86. Rossignol, S. et al. Locomotor capacities afte r complete and partial lesions of the spinal cord. Acta Neurobiol Exp 56, 449-63 (1996). 87. Lovely, R. G., Gregor, R. J., Roy, R. R. & Edgerton, V. R. Effects of training on the recovery of full-weight-bearing st epping in the adult spinal cat. Exp Neurol 92, 421-35 (1986). 88. Barbeau, H., Julien, C. & Rossignol, S. The effects of clonidine and yohimbine on locomotion and cutaneous reflexes in the adult chronic spinal cat. Brain Res 437, 83-96 (1987).

PAGE 144

144 89. Duysens, J. & Pearson, K. G. Inhibiti on of flexor burst gene ration by loading ankle extensor muscles in walking cats. Brain Res 187, 321-32 (1980). 90. de Leon, R. D., Hodgson, J. A., Roy, R. R. & Edgerton, V. R. Locomotor capacity attributable to step training ve rsus spontaneous recovery afte r spinalization in adult cats. J Neurophysiol 79, 1329-40 (1998). 91. Pearson, K. G. Common prin ciples of motor control in vertebrates and invertebrates. Annu Rev Neurosci 16, 265-97 (1993). 92. Grillner, S., McClellan, A., Sigvardt, K., Wallen, P. & Wilen, M. Activation of NMDAreceptors elicits "fictive locomotion" in lamprey spinal cord in vitro. Acta Physiol Scand 113, 549-51 (1981). 93. Grillner, S. & Dubuc, R. Control of locomo tion in vertebrates: spinal and supraspinal mechanisms. Adv Neurol 47, 425-53 (1988). 94. Nolte, J. The human brain, 5th edition an intr oduction to its f unctional anatomy (Mosby-Year Book, St. Louis, MO, 2002). 95. Duysens, J. & Van de Crommert, H. W. Neur al control of locomotion; the central pattern generator from cats to humans. Gait Posture 7, 131-41 (1998). 96. Delcomyn, F. Neural basis of rhythmic behavior in animals. Science 210, 492-8 (1980). 97. Patla, A. E. in Evaluation and management of gait disorders (ed. Spivack, B. S.) (New York, 1995). 98. Brooks, V. B. Cerebellar functions in motor control. Hum Neurobiol 2, 251-60 (1984). 99. Armstrong, D. M. The supraspina l control of mammalian locomotion. J Physiol. 405, 137 (1988). 100. Grillner, S. & Wallen, P. Central pattern generators for locomotion, with special reference to vertebrates. Annu Rev Neurosci., 233-61 (1985). 101. Patla, A. E., Prentice, S. D., Robinson, C. & Neufeld, J. Visual control of locomotion: strategies for changing direc tion and for going over obstacles. J Exp Psychol Hum Percept Perform 17, 603-34 (1991). 102. Patla, A. E., Beuter, A. & Prentice, S. A two stage correct ion of limb trajectory to avoid obstacles during stepping. Neurosci Res Commun 8, 153-9 (1991a). 103. Weytjens, J. L., Viberg, D. A., Caputi, A. A., Kallesoe, K. & Hoffer, J. A. New transducers for measuring muscle length in unrestrained animals. J Neurosci Methods 45, 217-25 (1992).

PAGE 145

145 104. Schomburg, E. D. Spinal sensorimotor systems and their supraspinal control. Neurosci Res. 7, 265-340 (1990). 105. Schomburg, E. D. Spinal functions in sensorimotor control of movements. Neurosurg Rev. 13, 179-85 (1990). 106. Grillner, S. & Shik, M. L. On the descending control of the lumbosacral spinal cord from the "mesencephalic locomotor region". Acta Physiol Scand 87, 320-33 (1973). 107. Shik, M. L., Severin, F. V. & Orlovsky, G. N. Control of walking and running by means of electrical stimulation of the mesencephalon. Electroencephalogr Clin Neurophysiol 26, 549 (1969). 108. Orlovsky, G., Deliagina, T. & Grillner, S. Neural control of locomotion: from mollusc to man. (Oxford Univ. Press, Oxford, UK, 1999). 109. Connor, N. P. & Abbs, J. H. Sensorimotor contributions of the basal ganglia: recent advances. Phys Ther 70, 864-72 (1990). 110. Berthoz, A. The role of gaze in compen sation of vestibular dysfunction: the gaze substitution hypothesis. Prog Brain Res 76, 411-20 (1988). 111. Eidelberg, E. & Yu, J. Effects of corticospinal lesions upon treadmill locomotion by cats. Exp Brain Res 43, 101-3 (1981). 112. Yu, J. & Eidelberg, E. Effects of vesti bulospinal lesions upon locomotor function in cats. Brain Res 220, 179-83 (1981). 113. Nashner, L. M. & Forssberg, H. Phasedependent organization of postural adjustments associated with arm movements while walking. J Neurophysiol 55, 1382-94 (1986). 114. Garcia-Rill, E. The basal ganglia and the locomotor regions. Brain Res 396, 47-63 (1986). 115. Winter, D. A. Biomechanics and motor control of human movement (John Wiley & Sons, New Jersey, 2005). 116. Knudson, D. Fundamentals of biomechanics (Springer, 2003). 117. Robertson, D. G. E., Caldwell, D. E., Hamill, J., Kamen, G. & Whittlesey, S. N. Research methods in biomechanics (Human Kinetics Publishers, 2004). 118. Gottschall, J. S. & Kram, R. Energy cost and muscular activity required for propulsion during walking. J Appl Physiol 94, 1766-72 (2003). 119. Meinders, M., Gitter, A. & Czerniecki, J. M. The role of ankle plantar flexor muscle work during walking. Scand J Rehabil Med 30, 39-46 (1998).

PAGE 146

146 120. Brown, T. G. The intrinsic factors in the act of progression in the mammal. Proc R Soc Lond [Biol] 84, 308-19 (1911). 121. Grillner, S. & Zangger, P. The effect of dorsal root transection on the efferent motor pattern in the cat's hindlimb during locomotion. Acta Physiol Scand 120, 393-405 (1984). 122. Shik, M. L., Severin, F. V. & Orlovsky, G. N. Control of walking and running by means of electric stimulation of the midbrain. Biofizika, 659-66. (1966). 123. Maegele, M., Muller, S., We rnig, A., Edgerton, V. R. & Harkema, S. J. Recruitment of spinal motor pools during voluntary movements versus stepping after human spinal cord injury. J Neurotrauma 19, 1217-29. (2002). 124. Dimitrijevic, M. R., Gerasimenko, Y. & Pi nter, M. M. Evidence for a spinal central pattern generator in humans. Ann N Y Acad Sci 860, 360-76 (1998). 125. Edgerton, V. R. et al. A physiological basis for the developm ent of rehabilitative strategies for spinally injured patients. J Am Paraplegia Soc 14, 150-7 (1991). 126. Bassler, U. Sensory control of leg moveme nt in the stick insect carausius morosus. Biol Cybern 25, 61-72 (1977). 127. Dietz, V., Muller, R. & Colombo, G. Loco motor activity in spinal man: significance of afferent input from joint and load receptors. Brain 125, 2626-34 (2002). 128. Sherrington, C. S. Flexion reflex of th e limb, crossed extension reflex, and reflex stepping and standing. Journal of Physiology 40, 28-121 (1910). 129. Grillner, S. & Rossignol, S. On the initiation of the swing phase of locomotion in chronic spinal cats. Brain Res 146, 269-77 (1978). 130. Hiebert, G. W., Whelan, P. J., Prochazka, A. & Pearson, K. G. C ontributions of hindlimb flexor muscle afferents to the timing of phase transitions in the cat step cycle. J..Neurophysiol 75, 1126-37 (1996). 131. Calancie, B. et al. Involun tary stepping after chronic spinal cord injury Evidence for a central rhythm generator for locomotion in man. Brain 117 ( Pt 5), 1143-59 (1994). 132. Dietz, V. & Duysens, J. Significance of lo ad receptor input duri ng locomotion: a review. Gait Posture 11, 102-10. (2000). 133. Duysens, J. & Pearson, K. G. The role of cutaneous afferents from the distal hindlimb in the regulation of the step cycle of thalamic cats. Exp Brain Res 24, 245-55 (1976). 134. Sinkjaer, T., Andersen, J. B., Ladouceur, M ., Christensen, L. O. D. & Nielsen, J. B. Major role for sensory feedback in soleus EM G activity in the stance phase of walking in man. J Physiol. 523, 817-827 (2000).

PAGE 147

147 135. Harkema, S. J. et al. Human lumbosacral spinal cord interprets loading during stepping. J Neurophysiol 77, 797-811 (1997). 136. Knikou, M. & Conway, B. A. Reflex effect s of induced muscle contraction in normal and spinal cord injured subjects. Muscle Nerve 26, 374-82 (2002). 137. Wolpaw, J. R. & O'Keefe, J. A. Adaptive pl asticity in the primate spinal stretch reflex: evidence for a two-phase process. J Neurosci 4, 2718-24 (1984). 138. Palmieri, R. M., Ingersoll, C. D. & Hoff man, M. A. The Hoffmann reflex: methodologic considerations and applications for use in s ports medicine and athl etic training research. J Athl Train. 39, 268-77 (2004). 139. Hoffmann, P. Beitrag zur Kenntnis der menschlichen Reflexe mit besonderer Berucksichtigung der elektrischen Erscheinungen. Arch Anat Physiol. 1, 223-46 (1910). 140. Fisher, M. A. AAEM Minimonograph #13: H reflexes and F waves: physiology and clinical indications. Muscle Nerve. 15, 1223-33 (1992). 141. Braddom, R. I. & Johnson, E. W. Standardi zation of H reflex and diagnostic use in Sl radiculopathy. Arch Phys Med Rehabil 55, 161-6 (1974). 142. Braddom, R. L. & Johnson, E. W. H reflex: review and classification with suggested clinical uses. Arch Phys Med Rehabil 55, 412-7 (1974). 143. Bell, K. R. & Lehmann, J. F. Effect of c ooling on Hand T-reflexes in normal subjects. Arch Phys Med Rehabil 68, 490-3 (1987). 144. Krause, B. A., Hopkins, J. T., Ingersoll, C. D., Cordova, M. L. & Edwards, J., E. The relationship of ankle temperat ure during cooling and rewarming to the human soleus H reflex. Journal of Sport Rehabilitation 9, 253-62 (2000). 145. Hoffman, M. A. & Koceja, D. M. The eff ects of vision and task complexity on Hoffmann reflex gain. Brain Res. 700, 303-7 (1995). 146. Capaday, C. & Stein, R. B. Amplitude m odulation of the soleus H-reflex in the human during walking and standing. J Neurosci 6, 1308-13 (1986). 147. Capaday, C. & Stein, R. B. Difference in the amplitude of the human soleus H reflex during walking and running. J Physiol 392, 513-22 (1987). 148. Schieppati, M. The Hoffman reflex: a means of assessing spinal reflex excitability and its descending control in man. Progress in Neurobiology 28, 345-76 (1987). 149. Zehr, E. P. Considerations for use of the Hoffmann reflex in exercise studies. Eur J Appl Physiol. 86, 455-68 (2002).

PAGE 148

148 150. Chalmers, G. R. & Knutzen, K. M. Sole us Hoffmann-reflex m odulation during walking in healthy elderly and young adults. J Gerontol A Biol Sci Med Sci 55, B570-9 (2000). 151. Simonsen, E. B. & Dyhre-Poulsen, P. Am plitude of the human soleus H reflex during walking and running. J Physiol 515 ( Pt 3), 929-39 (1999). 152. Brooke, J. D., Cheng, J., Misiaszek, J. E. & Lafferty, K. Amplitude modulation of the soleus H reflex in the human during active and passive stepping movements. J Neurophysiol. 73, 102-11 (1995). 153. Brooke, J. D. et al. Sensori-sensory af ferent conditioning with leg movement: gain control in spinal reflex and ascending paths. Prog Neurobiol 51, 393-421 (1997). 154. Schneider, C., Lavoie, B. A. & Capaday, C. On the origin of the soleus H-reflex modulation pattern during human walking and its task-depen dent differences. J Neurophysiol. 83, 2881-90 (2000). 155. Fung, J. & Barbeau, H. Effects of conditioning cutaneomuscular stimulation on the soleus H-reflex in normal and spastic pareti c subjects during walking and standing. J Neurophysiol 72, 2090-104 (1994). 156. Yang, J. F. et al. H-re flex modulation during walking in spastic paretic subjects. Can J Neurol Sci 18, 443-52 (1991). 157. Knikou, M. & Conway, B. A. Modulation of soleus H-reflex following ipsilateral mechanical loading of the sole of the foot in normal and complete spinal cord injured humans. Neurosci Lett 303, 107-10 (2001). 158. Kamiya, A., Tanabe, S., Muraoka, Y. & Masa kado, Y. Modulation of the soleus H-reflex during static and dynamic imposed hip angle changes. Int J Neurosci 116, 1045-53 (2006). 159. Knikou, M. & Rymer, W. Z. Hip angle induced modulation of H reflex amplitude, latency and duration in spin al cord injured humans. Clin Neurophysiol 113, 1698-708 (2002). 160. Wolpaw, J. R. & Carp, J. S. Plasticity from muscle to brain. Prog Neurobiol 78, 233-63 (2006). 161. Shurrager, P. S. & Dykman, R. A. Walking sp inal carnivores. J Comp Physiol Psychol. 44, 252-62 (1951). 162. de Leon, R. D., Tamaki, H., Hodgson, J. A., Roy, R. R. & Edgerton, V. R. Hindlimb locomotor and postural training modulates glycin ergic inhibition in the spinal cord of the adult spinal cat. J Neurophysiol 82, 359-69 (1999). 163. Chau, C., Barbeau, H. & Rossignol, S. Early locomotor training with clonidine in spinal cats. J Neurophysiol 79, 392-409 (1998).

PAGE 149

149 164. Chau, C., Barbeau, H. & Rossignol, S. Ef fects of intratheca l alpha1and alpha2noradrenergic agonists and norepinephrine on locomotion in chr onic spinal cats. J Neurophysiol 79, 2941-63 (1998). 165. Roy, R. R. & Acosta, L., Jr. Fiber type a nd fiber size changes in selected thigh muscles six months after low thoracic spinal cord tr ansection in adult cats: exercise effects. Exp Neurol 92, 675-85 (1986). 166. Dietz, V. Spinal cord patt ern generators for locomotion. Clin Neurophysiol 114, 1379-89 (2003). 167. Krassioukov, A. V., Bunge, R. P., Pucket, W. R. & Bygrave, M. A. The changes in human spinal sympathetic preganglioni c neurons after spinal cord injury. Spinal Cord 37, 6-13 (1999). 168. Wolpaw, J. R. & Herchenroder, P. A. Operant conditioning of H-reflex in freely moving monkeys. J Neurosci Methods 31, 145-52 (1990). 169. Wolpaw, J. R. The complex structure of a simple memory. Trends Neurosci 20, 588-94 (1997). 170. Wolpaw, J. R. & Tennissen, A. M. Activity-de pendent spinal cord plasticity in health and disease. Annu Rev Neurosci 24, 807-43 (2001). 171. Chen, X. Y., Chen, Y., Chen, L., Tennissen, A. M. & Wolpaw, J. R. Corticospinal tract transection permanently abolishes H-reflex down-conditioning in rats. J Neurotrauma 23, 1705-12 (2006). 172. Chen, Y. et al. Operant conditioning of H-reflex can correct a locomotor abnormality after spinal cord injury in rats. J Neurosci 26, 12537-43 (2006). 173. Chen, X. Y., Chen, L. & Wolpaw, J. R. Conditioned H-reflex increase persists after transection of the main co rticospinal tract in rats. J Neurophysiol 90, 3572-8 (2003). 174. Perez, M. A., Lungholt, B. K. & Nielsen, J. B. Presynaptic control of group Ia afferents in relation to acquisition of a visuomotor skill in healthy humans. J Physiol 568, 343-54 (2005). 175. Schneider, C. & Capaday, C. Progressive ad aptation of the soleus H-reflex with daily training at walking backward. J Neurophysiol 89, 648-56 (2003). 176. Trimble, M. H. & Koceja, D. M. Downtr aining of triceps sura e H-reflexes during a balance task in humans. Intern J Neurosci 76, 293-303 (1994). 177. Hess, F., Van Hedel, H. J. & Dietz, V. Obstacle avoidance du ring human walking: Hreflex modulation during motor learning. Exp Brain Res 151, 82-9 (2003).

PAGE 150

150 178. Segal, R. L. & Wolf, S. L. Operant conditioni ng of spinal stretch refl exes in patients with spinal cord injuries. Exp Neurol 130, 202-13 (1994). 179. Bouyer, L. J., Whelan, P. J., Pearson, K. G. & Rossignol, S. Adaptive locomotor plasticity in chronic spinal cats after ankle exte nsors neurectomy. J Neurosci 21, 3531-41 (2001). 180. Carrier, L., Brustein, E. & Rossignol, S. Locomotion of the hindlimbs after neurectomy of ankle flexors in intact and spinal cats : model for the study of locomotor plasticity. J Neurophysiol 77, 1979-93 (1997). 181. Trimble, M. H., Behrman, A. L., Flynn, S. M., Thigpen, M. T. & Thompson, F. J. Acute effects of locomotor training on overground wa lking speed and Hreflex modulation in individuals with incomplete spinal cord injury. J Spinal Cord Med 24, 74-80. (2001). 182. Wolpaw, J. R. The education a nd re-education of the spinal cord. Prog Brain Res 157, 261-280 (2006). 183. Dobkin, B. H., Harkema, S., Requejo, P. & Edgerton, V. R. Modulation of locomotorlike EMG activity in subjects with complete and incomplete spinal cord injury. J Neurol Rehabil 9, 183-90 (1995). 184. Schmidt, R. A. & Young, D. E. Met hodology for motor learning: A paradigm for kinematic feedback. J Mot Behav 23, 13-24 (1991). 185. Dietz, V. Evidence for a load receptor contribution to the control of posture and locomotion. Neurosci Biobehav Rev 22, 495-9 (1998). 186. Wernig, A., Nanassy, A. & Muller, S. La ufband (treadmill) therapy in incomplete paraplegia and tetraplegia. J Neurotrauma 16, 719-26 (1999). 187. Gardner, M. B., Holden, M. K., Leikauskas J. M. & Richard, R. L. Partial body weight support with treadmill locomotion to improve gait after incomplete sp inal cord injury: a single-subject experimental design. Phys Ther 78, 361-74 (1998). 188. Wirz, M., Colombo, G. & Dietz, V. Long te rm effects of locomotor training in spinal humans. J Neurol Neurosurg Psychiatry 71, 93-6. (2001). 189. Barbeau, H., Pepin, A., Norman, K. E., Ladou ceur, M. & Leroux, A. Walking after spinal cord injury: cont rol and recovery. The neuroscientist 4, 14-24 (1998). 190. Behrman, A. L. & Harkema, S. J. Recove ry of walking: locomo tor training after human spinal cord injury. (submitted). 191. Dietz, V., Colombo, G., Jensen, L. & Baumga rtner, L. Locomotor cap acity of spinal cord in paraplegic patients. Ann Neurol 37, 574-82. (1995).

PAGE 151

151 192. Field-Fote, E. C., Lindley, S. D. & Sherman, A. L. Loco motor training approaches for individuals with spinal cord injury: a prel iminary report of walking-related outcomes. J Neurol Phys Ther 29, 127-37 (2005). 193. Chen, G. & Patten, C. Treadmill training with harness support: selection of parameters for individuals with poststroke hemiparesis. J Rehabil Res Dev. 43, 485-98 (2006). 194. Conrad, B., Benecke, R. & Carnehl, J. in Motor control mechani sms in health and disease (ed. Desmedt, J. E.) 717-726 (Raven Press, New York, 1983). 195. Visintin, M. & Barbeau, H. The effects of parallel bars, body weight support and speed on the modulation of the locomotor pattern of spastic paretic gait. A preliminary communication. Paraplegia 32, 540-53 (1994). 196. Dobkin, B. in Joint meeting of ACRM and ASNR (Tuscon, AZ, 2003). 197. Mandzak-McCarron, K. & Drayton-Hargrove, S. Ambulation aids. Rehabil Nurs 12, 13941 (1987). 198. Mulley, G. Walking frames. Bmj 300, 925-7 (1990). 199. Parziale, J. R. & Daniels, J. D. The m echanical performance of ambulation using springloaded axillary crutches. A preliminary report. Am J Phys Med Rehabil 68, 192-5 (1989). 200. Waters, R. L., Yakura, J. S. & Adkins, R. H. Gait performance afte r spinal cord injury. Clin Orthop Relat Res, 87-96 (1993). 201. Opila, K. A., Nicol, A. C. & Paul, J. P. Forces and impul ses during aided gait. Arch Phys Med Rehabil 68, 715-22 (1987). 202. Melis, E. H., Torres-Moreno, R., Barbeau, H. & Lemaire, E. D. Analysis of assisted-gait characteristics in persons with incomplete spinal cord injury. Spinal Cord 37, 430-9. (1999). 203. Ounpuu, S., Bell, K. J., Davis, R. B., 3rd & DeLuca, P. A. An evaluation of the posterior leaf spring orthosis using joint kinematics and kinetics. J Pediatr Orthop 16, 378-84 (1996). 204. Brooke, J. D., McIlroy, W. E. & Collins, D. F. Movement features and H-reflex modulation. I. Pedalling versus matched controls. Brain Res 582, 78-84 (1992). 205. Nishikawa, T. & Grabiner, M. D. Pe roneal motoneuron excitability increases immediately following applicati on of a semirigid ankle brace. J Orthop Sports Phys Ther 29, 168-73; discussi on 174-6 (1999). 206. Garrett, M., Kerr, T. & Caul field, B. Phase-dependent in hibition of H-reflexes during walking in humans is independent of reduction in knee angular velocity. J Neurophysiol 82, 747-53 (1999).

PAGE 152

152 207. Bach-y-Rita, P. Rehabilitation versus passi ve recovery of motor control following central nervous system lesions. Adv Neurol 39, 1085-92 (1983). 208. Kadaba, M. P., Ramakrishnan, H. K. & Wootten, M. E. Measurement of lower extremity kinematics during level walking. J Orthop Res 8, 383-92 (1990). 209. Cappozzo, A. in Adaptability of human gait. (ed. Patla, A. E.) 167-85 (Elsevier Science Pub, North-Holland, 1991). 210. Lafortune, M. A., Lake, M., J. & Hennig, E. M. Differential shock transmission response of the human body toimpact seve rity and lower limb posture. J Biomech 29, 1531-7. (1996). 211. Kadaba, M. P. et al. Repeatability of ki nematic, kinetic, and electromyographic data in normal adult gait. J Orthop Res 7, 849-60 (1989). 212. Richards, J. G. The measurement of hum an motion: a comparison of commercially available systems. Human Movement Science 18, 589-602 (1999). 213. Baker, R. Gait analysis methods in rehabilitation. J Neuroengineering Rehabil 3, 4 (2006). 214. Schwartz, M. H., Trost, J. P. & Wervey, R. A. Measurement and management of errors in quantitative gait data. Gait Posture 20, 196-203 (2004). 215. Barker, S. P. in School of Biomedical Engineering, Science and Health Systems (Drexel University, 2004). 216. Hanke, T. A. & Rogers, M. W. Reliability of ground reaction force measurements during dynamic transitions from bipedal to single-limb stance in healthy adults. Phys Ther 72, 810-6 (1992). 217. Lukacs, R. U. et al. in Society for Neuroscience (2002). 218. Morarty, R. & Agosta, J. in Australian Conference of Science and Medicine in Sport (Adelaide, 1998). 219. Cordova, M. L. & Armstrong, C. W. Re liability of ground reaction forces during a vertical jump: implications fo r functional strength assessment. J Athl Train 31, 342-5 (1996). 220. Hopkins, J. T., Ingersoll, C. D., Cordova M., L, & Edwards, J., E. Intrasession and intersession reliability of the soleus H-reflex in supine and standing positions. Electromyogr Clin Neurophysiol. 40, 89-94 (2000). 221. Paquet, N. & Hui-Chan, C. W. Human sole us H-reflex excitab ility is decreased by dynamic head-and-body tilts. J Vestib Res 9, 379-83 (1999).

PAGE 153

153 222. Pierrot-Deseilligny, E. & Mazevet, D. The monosynaptic reflex: a tool to investigate motor control in humans. Interest and limits. Neurophysiol Clin 30, 67-80 (2000). 223. Honore, J., Demaire, C. & Coquery, J. M. Effects of spatially or iented attention on the facilitation of the H reflex by a cutaneous stimulus. Electroencephalogr Clin Neurophysiol 55, 156-61 (1983). 224. Bonnet, M., Decety, J., Jeannerod, M. & Requin, J. Mental si mulation of an action modulates the excitability of sp inal reflex pathways in man. Brain Res Cogn Brain Res 5, 221-8 (1997). 225. Walton, C., Kalmar, J. & Cafarelli, E. Caff eine increases spinal excitability in humans. Muscle Nerve 28, 359-64 (2003). 226. Funase, K. & Miles, T. S. Observations on th e variability of the H reflex in human soleus. Muscle Nerve 22, 341-6 (1999). 227. Crone, C., Johnsen, L. L., Hultborn, H. & Orsnes, G. B. Amplitude of the maximum motor response (Mmax) in human muscles t ypically decreases duri ng the course of an experiment. Exp Brain Res 124, 265-70 (1999). 228. Stein, R. B. & Capaday, C. The modula tion of human reflexes during functional motor tasks. Trends Neurosci 11, 328-32 (1988). 229. Crenna, P. & Frigo, C. Ex citability of the soleus H -reflex arc during walking and stepping in man. Exp Brain Res 66, 49-60 (1987). 230. Chen, X. Y. & Wolpaw, J. R. Dorsal column but not lateral column transection prevents down-conditioning of H reflex in rats. J Neurophysiol 78, 1730-4 (1997). 231. Segal, R. Plasticity in the central ner vous system: operant conditioning of the spinal stretch reflex. Top Stroke Rehabilitation 3, 76-87 (1997). 232. Sawicki, G. S., Domingo, A. & Ferris, D. P. The effects of powere d ankle-foot orthoses on joint kinematics and muscle activation duri ng walking in individuals with incomplete spinal cord injury. J Neuroengineering Rehabil 3, 3 (2006). 233. Ferris, D. P., Gordon, K. E., Sawicki, G. S. & Peethambaran, A. An improved powered ankle-foot orthosis using pr oportional myoelectric control. Gait Posture 23, 425-8 (2006). 234. Gordon, K. E., Sawicki, G. S. & Ferris, D. P. Mechanical performance of artificial pneumatic muscles to power an ankle-foot orthosis. J Biomech 39, 1832-41 (2006). 235. Hylton, N. Dynamic orthotic concepts: background and experiences. (Orthopaedie technology, Dortmund, Germany, 2000).

PAGE 154

154 236. Beekman, C. E., Miller-Porter, L. & Sc honeberger, M. Energy cost of propulsion in standard and ultralight wheelchairs in people with spinal cord injuries. Phys Ther 79, 146-58 (1999). 237. Dobkin, B. H. et al. Methods for a randomized trial of weight-supported treadmill training versus conventional training for walking during in patient rehabilitation after incomplete traumatic spinal cord injury. Neurorehabil Neural Repair 17, 153-67 (2003). 238. Lacquaniti, F., Ivanenko, Y. P. & Za go, M. Kinematic control of walking. Arch Ital Biol 140, 263-72 (2002). 239. Patrick, J. H. Case for gait analysis as part of the management of incomplete spinal cord injury. Spinal Cord 41, 479-82 (2003). 240. Smith, P. A., Hassani, S., Reiners, K., Voge l, L. C. & Harris, G. F. Gait analysis in children and adolescents with spinal cord injuries. J Spinal Cord Med 27 Suppl 1, S44-9 (2004). 241. Ferris, D. P., Gordon, K. E., Beres-Jones, J. A. & Harkema, S. J. Muscle activation during unilateral stepping occurs in the nonstepping limb of humans with clinically complete spinal cord injury. Spinal Cord 42, 14-23 (2004). 242. Hidler, J. Robotic-assessment of walking in individuals with gait disorders. Conf Proc IEEE Eng Med Biol Soc 7, 4829-31 (2004). 243. Balasubramanian, C. K., Bowden, M. G., Neptune, R. R. & Kautz, S. A. Relationship between step length asymmetry and walki ng performance in subjects with chronic hemiparesis. Arch Phys Med Rehabil. 88, 43-9 (2007). 244. Aisen, M. L. Orthotics in neurologic rehabilitation (Demos, New York, NY, 1992). 245. Portney, L. G. & Watkins, M. P. Foundations of clinical research: applications to practice. (Prentice-Hall, Upper Saddle River, NJ, 2000). 246. Westfall, P. H. & Young, S. S. Resampling-based multiple testing: examples and methods for p-value adjustment (John Wiley & Sons, Inc., New York, 1993). 247. Kuo, A. D., Donelan, J. M. & Ruina, A. Energetic consequences of walking like an inverted pendulum: step-to-step transitions. Exerc Sport Sci Rev. 33, 88-97 (2005). 248. Beekman, C. E., Perry, J., B oyd, L. A., Newsam, C. J. & Mulroy, S. J. The effects of a dorsiflexion-stopped ankle foot orthosis on walking in individua ls with incomplete spinal cord injury. Topics in Spinal Cord Injury Rehabilitation 5, 54-62 (2000). 249. Hesse, S., Werner, C., Matthias, K., Stephe n, K. & Berteanu, M. Non-velocity-related effects of a rigid double-stopped ankle-foot orthosis on gait and lower limb muscle activity of hemiparetic subjects with an equinovarus deformity. Stroke. 30, 1855-61. (1999).

PAGE 155

155 250. Pepin, A., Norman, K. E. & Barbeau, H. Treadmill walking in incomplete spinal-cordinjured subjects: 1. Adaptation to changes in speed. Spinal Cord 41, 257-70. (2003). 251. Pepin, A., Ladouceur, M. & Barbeau, H. Treadmill walking in incomplete spinal-cordinjured subjects: 2. Factors limiting the maximal speed. Spinal Cord 41, 271-9. (2003). 252. Dromerick, A. W., Lum, P. S. & Hidler, J. Activity-based therapies. NeuroRx. 3, 428-38 (2006). 253. Yang, J. F., Stein, R. B. & James, K. B. Contribution of periph eral afferents to the activation of the soleus muscle during walking in humans. Exp Brain Res 87, 679-87 (1991). 254. Yang, J. F. & Whelan, P. J. Neural mechanis ms that contribute to cyclical modulation of the soleus H-reflex in walking in humans. Exp Brain Res 95, 547-56 (1993). 255. Cordova, M. L. & Ingersoll, C. D. Peroneus longus stretch reflex amplitude increases after ankle brace application. Br J Sports Med 37, 258-62 (2003). 256. Stein, R. B. & Thompson, A. K. Muscle reflexes in motion: how, what, and why? Exerc Sport Sci Rev 34, 145-53 (2006). 257. Frigon, A., Collins, D. F. & Zehr, E. P. Ef fect of rhythmic arm movement on reflexes in the legs: modulation of soleus H-reflexes and somatosensory conditioning. J Neurophysiol 91, 1516-23 (2004). 258. Sayenko, D. G. et al. Facilitation of th e soleus stretch reflex induced by electrical excitation of plantar cutaneous afferents located around the heel. Neurosci Lett 415, 2948 (2007). 259. Hwang, I. S. Assessment of soleus mot oneuronal excitability using the joint angle dependent H reflex in humans. J Electromyogr Kinesiol 12, 361-6 (2002). 260. Patikas, D. A., Kotzamanidis, C., Robertson, C. T. & Koceja, D. M. The effect of the ankle joint angle in the level of soleus Ia afferent pres ynaptic inhibition. Electromyogr Clin Neurophysiol 44, 503-11 (2004). 261. Sefton, J. M., Hicks-Little, C. A., Koceja, D. M. & Cordova, M. L. Effect of inversion and ankle bracing on peroneus longus Hoffmann reflex. Scand J Med Sci Sports 17, 53946 (2007). 262. Sefton, J. M., Hicks-Little, C. A., Koceja, D. M. & Cordova, M. L. Modulation of soleus H-reflex by presynaptic spinal mechanis ms during varying surface and ankle brace conditions. Neurophysiol Clin 37, 15-21 (2007). 263. Stein, R. B. Presynapt ic inhibition in humans. Prog Neurobiol 47, 533-44 (1995).

PAGE 156

156 264. Little, J. W. & Halar, E. M. H-refl ex changes following spinal cord injury. Arch Phys Med Rehabil 66, 19-22 (1985). 265. Hiersemenzel, L. P., Curt, A. & Dietz, V. From spinal shock to spasticity: neuronal adaptations to a spinal cord injury. Neurology 54, 1574-82 (2000). 266. Knikou, M. Plantar cutaneous input modulates differently spinal reflexes in subjects with intact and injured spinal cord. Spinal Cord. 45, 69-77 (2007). 267. Luscher, H., Ruenzel, P. W. & Henneman, E. Effects of impulse frequency, PTP, and temperature on responses elicited in large populations of motoneurons by impulses in single Ia-fibers. J Neurophysiol 50, 1045-58 (1983). 268. Trimble, M. H., Kukulka, C. G. & Behrman, A. L. The effect of treadmill gait training on low-frequency depression of the soleus H-refl ex: comparison of a spinal cord injured man to normal subjects. Neurosci Lett 246, 186-8. (1998). 269. Kim, C. M., Eng, J. J. & Whittaker, M. W. Effects of a simple functional electric system and/or a hinged ankle-foot or thosis on walking in persons with incomplete spinal cord injury. Arch Phys Med Rehabil 85, 1718-23 (2004). 270. Cordova, M. L., Ingersoll, C. D. & Palmieri, R. M. Efficacy of prophylactic ankle support: an experimental perspective. J Athl Train 37, 446-57 (2002). 271. Feuerbach, J. W., Grabiner, M. D., Koh, T. J. & Weiker, G. G. Effect of an ankle orthosis and ankle ligament anesthesia on ankle joint proprioception. Am J Sports Med 22, 223-9 (1994). 272. Iles, J. F. Evidence for cutaneous and corticospinal modulation of presynaptic inhibition of Ia afferents from the human lower limb. J Physiol 491 ( Pt 1), 197-207 (1996). 273. Rudomin, P. & Schmidt, R. F. Presynaptic inhibition in the vert ebrate spinal cord revisited. Exp Brain Res 129, 1-37 (1999). 274. Bastiaanse, C. M., Duysens, J. & Dietz, V. Modulation of cutaneous reflexes by load receptor input during human walking. Exp Brain Res 135, 189-98 (2000). 275. Calancie, B. et al. Evidence that altera tions in presynaptic inhibition contribute to segmental hypoand hyperexcitability after spinal cord injury in man. Electroencephalogr Clin Neurophysiol 89, 177-86 (1993). 276. Nakajima, T., Sakamoto, M., Tazoe, T., Endoh, T. & Komiyama, T. Location specificity of plantar cutaneous reflexes involv ing lower limb muscles in humans. Exp Brain Res. 175, 514-25. (2006). 277. Aniss, A. M., Gandevia, S. C. & Burke, D. Reflex responses in active muscles elicited by stimulation of low-threshold a fferents from the human foot. J Neurophysiol 67, 1375-84 (1992).

PAGE 157

157 278. Yang, J. F. & Stein, R. B. Phase-dependent reflex reversal in human leg muscles during walking. J Neurophysiol 63, 1109-17 (1990).

PAGE 158

158 BIOGRAPHICAL SKETCH Preeti Nair was born in Mum bai India in 1979. She received her bachelors degree in physical therapy from Pune University, India, in 2001. She worked for a year at a government hospital in Mumbai and for the non-profit Multiple Sclerosis Society of India. Her specific interest in neurological rehabi litation developed from two areas of interest; first, the elusive workings of the nervous system for the control of movement and learning. Secondly, her work experience with individuals with neurological impairment and th e challenges faced in restoring them to their activities of daily living. The process of disablement unleashed by the disease state and perpetuated by underdeveloped infrastructure and apathy for neurol ogical rehabilitation motivated her to pursue higher educ ation in the United States; a country that has set the mark for its multidimensional approach towards enableme nt and empowerment of an individual with impairment. Duly, she chose the interdisciplinar y, Rehabilitation Scien ces Doctoral program offered at the University of Florida in 2002. U nder the expert tutelage of Dr. Andrea Behrman and Dr. Steven Kautz, her research deals with examining the neuromechani cal control of walking with orthotic devices in individu als after spinal cord injury that integrates the principles of neurological control of walking, motor control and movement mech anics. She is a recipient of the Alumni Fellowship which provided financial support for her doctoral education.