|UFDC Home||myUFDC Home | Help|
1 PRESENCE, ASSESSMENT AND PLASTICITY OF DESCENDING PATHWAYS AFTER SPINAL CORD INJURY By MARTINA REBEKKA SPIESS 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 2012
2 2012 Martina Spiess
3 To my parents Robert and Beatrice Spiess
4 ACKNOWLEDGMENTS So many people have touched my life and helped me to achieve my goals so far! First of all, I would like to thank my parents Robert and Beatrice Spiess They provide continuous and stable support and encouragement no matter where life and my adventures carry me. They are the ones that taught me that education is not only important, but that acquiring knowledge is fun and very rewarding. Without them, I would not be where I am today. I would like to thank my mentors : Dr. Andrea Behrman, Dr. Carolynn Patten and Dr. Dena Howland who, four years ago, believed in me and accepted me as their student. In these four years, they shared their scientific knowledge and offered their guidance and dedication in educational and research matters. However, they also cared about what kind of person I am and am becoming. Their personal support in all sorts of life situations beyond just school was and is incredible! They all live by example and I have learned so much more from each one of them than just science! This is also true for my fourth co mmittee member, Dr. Orit Shechtman. Further I would like to mention Dr. Mary Thigpen, Dr. Emily Fox and Dr. Nicole Tester. They provided me with so much support ( scientific, editorial and moral ) during this thesis and belonged to my extended circle of ment ors during the last four and a half years I have also enjoyed learning from my teaching mentors, Dr. Steven George and Dr. Donovan Lott. Helping in their classes was always a pleasure and the teaching skills I learned from the m are invaluable! There are a lot of people at the VA BRRC that have helped me with my data collections. This research would not have been possible without a wonderful team. Above all, this team included our outstanding lead engineer Theresa McGuirk two
5 incredibly talented undergraduate assistants, Yasmin Islam and Tali Yohann and an a mazing research PT, Helen Emery. These ladies went way beyond what I had ever imagined t o assist with my dissertation research. Other people without whom this research would not have been possible include Lindsay Perry, Mike Rademaker and all the other undergraduate assistants who assisted with training and testing sessions Finally of course, I would like to thank the wonderful individuals that participated in our studies! It is their generous contributi on of their time and effort that allows us to conduct studies that will hopefully make a difference for people with spinal cord injury! In the first two years of my PhD, I learned a lot in the classes offered by many departments on campus. I would like to thank all those professors for sharing their knowledge, for their enthusiasm and their passion. Specifically, I would like to mention Dr. Chris Hass, Dr. Michael Marsiske and Dr. David Fuller. They made learning so much fun! In addition, there are countless teachers, instructors and mentors that have had a positive influence on my life in my entire school career leading up to the PhD. In particular, I would like to thank Dr. Martin Schubert and Dr. Huub van Hedel for their su pport and guidance and for sharing their passion for spinal cord injury research with me. Last but certainly not least, there are my friends and family both new ones that I made here in the US and the old ones from back home, which include, but are not limited to: Ginny Little, Leandro Neves, Manuela Corti, Petra Dokladal and family Simone and Stephan Gasser, Christian and Daniela Bhler and family Sandra Sprri Meyer and family, Carmela Flury and Maarten Graaskam p Serge Duflon and family, Laurie and Roger Phill ips and family Becky and Jeff Grant, Joachim von Zitzewitz,
6 Richard and Linda Finch and family, Susan and Donald McClure, the RSD and DPT students of the University of Florida the members of the Gainesville cycling club and the Swiss American society in Atlanta, G A and all the wonderful instructors of UF RecSports Finally, the list also includes my sister Annette Liebi Spiess with her husband Marcel and my two adorable nieces Yael and Samira. All these friends and family are the people that pr ovided support when I was struggling and who shared my successes when I had them they are the people that helped me take my mind of f work on bike rides during fitness classes on skype calls and who showed me the natural beauty of Florida and the Southeast of the US or discovered it together with me. They are the people that welcomed me back home during my vacations and adjusted their schedules to be able to spend time with me in my beloved mountains! In short, they are the people that helped me keep my s anity throughout the process! All together, its really the people that made my Graduate School experience and my life in the Gatornation such a fantastic experience! I have been blessed and will cherish the memories and friendships forever!
7 TABLE OF CONT ENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES .......................................................................................................... 11 LIST OF FIGURES ........................................................................................................ 13 LIST OF ABBREVIATIONS ........................................................................................... 15 ABSTRACT ................................................................................................................... 18 1 INTRODUCTION, AIMS AND HYPOTHESIS ......................................................... 20 1.1 Aims and Hypotheses ................................................................................... 24 1.2 Project I Defining BrownSquardSyndrome: Asymmetrical Sparing of Descending Pathways ......................................................................................... 24 1.3 Project II The Acoustic Startle Response: Assessing Plasticity in a Descending Pathway after Spinal Cord Injury ..................................................... 25 1.4 Project III Inducing Plasticity in Descending Pathways through Locomotor Training ................................................................................................................ 26 2 LITERATURE REVIEW .......................................................................................... 28 2.1 Human Spinal Cord Injury ................................................................................. 28 2.1.1 Incidence, Prevalence and Demographics of Human Spinal Cord Injury ............................................................................................................. 28 2.1.2 Functional Consequences after SCI ........................................................ 30 2.1.3 Other Consequences after SCI ............................................................... 31 2.2 Descending Systems in Humans ...................................................................... 32 2.2.1 Corticospinal Tract ................................................................................... 32 2. 2.2 Vestibulospinal Tract .............................................................................. 33 2.2.3 Reticulospinal Tract ................................................................................. 33 2.3 The BrownSquard Syndrome ........................................................................ 34 2.3.1 Definition of the BrownSquard and BrownSquard Plus Syndrome .... 34 2.3.2 Functional Recovery in Individuals with BSS ........................................... 35 2.3.3 Outcome Measures Used to Des cribe the BSS Population ..................... 37 2.3.4 Treatment Effects in Individuals with BSS ............................................... 39 2.3.5 Underlying Mechanisms of Recovery in BSS and BSPS ......................... 40 2.4 Neural Control of Walking ................................................................................. 42 2.4.1 Spinal Level ............................................................................................. 43 2.4.2 Supraspinal Level .................................................................................... 46 220.127.116.11 Gait initiation .................................................................................. 46 18.104.22.168 Online control of gait ...................................................................... 47 2.5 Walking Adaptability .......................................................................................... 48 2.5.1 What is Needed for Community Ambulation? .......................................... 48
8 2.5.2 How Well Can Individuals with iSCI Adapt their Walking Pattern to Environmental Demands? ............................................................................. 52 22.214.171.124 Adaptation to speed ....................................................................... 52 126.96.36.199 Adaptation to inclines ..................................................................... 53 188.8.131.52 Adaptation to obstacles .................................................................. 53 184.108.40.206 Adaptation to demands of community walking ............................... 54 220.127.116.11 Summary ........................................................................................ 55 2.6 The Nervous System Hardwired or Capable of Plasticity? ............................. 55 2.6.1 Paradigm Shift ......................................................................................... 55 2.6.2 Definition of Neuroplasticity ..................................................................... 56 2.6.3 Mechanisms of Neuroplasticity ................................................................ 56 2.6.4 Negative Neuroplasticity .......................................................................... 58 2.6.5 Principles of Neuroplasticity .................................................................... 59 2.7 Compensation vs. Recovery ............................................................................. 60 2.8 Locomotor Training ........................................................................................... 61 2.8.1 Basic Neuroscience Research Leading to the Development of Locomotor Training ....................................................................................... 61 2.8.2 The Principles of Locomotor Training ...................................................... 62 18.104.22.168 Maximizing weight bear ing on the legs (principle 1) ...................... 63 22.214.171.124 Optimize sensory cues (principle 2) ............................................... 63 126.96.36.199 Optimizing kinematics (principle 3) ................................................. 64 188.8.131.52 Maximizing recovery, minimizing compensatory strategies (principle 4) ............................................................................................. 64 2.8.3 Clinical Studies Investigating Locomotor Training ................................... 64 2.9 Outcome Measurement .................................................................................... 66 2.9.1 Selection of Outcome Measures ............................................................. 66 2.9.2 Measures of Body Structure and Body Function ..................................... 68 184.108.40.206 Neurophysiological measurements ................................................ 68 220.127.116.11 International Standards for Neurological Classification of Spinal Cord Injury .............................................................................................. 79 2.9.3 Measures of Activity ................................................................................ 80 18.104.22.168 Activity capacity ............................................................................. 80 2. 9.3.2 Activity performance ....................................................................... 83 2.9.4 Measures of Performance ....................................................................... 87 2.10 Summary and Conclusion of the Literature Review ........................................ 87 2.11 Tables and Figures ......................................................................................... 89 3 PROJECT I DEFINING BROWN SEQUARD SYNDROME: ASSYMETRICAL SPARING OF DESCENDING PATHWAYS ............................................................ 92 3.1 Background ....................................................................................................... 92 3.2 Methods ............................................................................................................ 95 3.2.1 Data ......................................................................................................... 95 3.2.2 Outcome Meas ures ................................................................................. 96 3.2.3 Definitions of BSS ................................................................................... 96 3.2.4 Statistical Analysis ................................................................................... 97 3.3 Results .............................................................................................................. 97
9 3.3.1 Available Data ......................................................................................... 97 3.3.2. A ssessments Classified as BSS/BSPS .................................................. 98 3.3.3 Stability of Being Defined as BSS/BSPS ................................................. 98 3.3.4 Demographics ......................................................................................... 99 3.3.5 Functional Status of Individuals Classified as BSS/BSPS ....................... 99 3.3.6 Improvement in Functional Status over Time ........................................ 100 3.4 Discussion ...................................................................................................... 101 3.5 Tables and Figures ......................................................................................... 106 4 PROJECT II THE ACOUSTIC STARTLE RESPONSE: ASSESSING PLASTICITY OF A DESCENDING PATHWAY AFTER SPINAL CORD INJURY 113 4.1 Background ..................................................................................................... 113 4.2 Methods .......................................................................................................... 118 4.2.1 Participants ............................................................................................ 118 4.2.2 Auditory Stimulation ............................................................................... 118 4.2.3 Recording .............................................................................................. 119 4.2.4 Positions for Testing .............................................................................. 119 4.2.5 Testing Protocol .................................................................................... 120 4.2.6 Outcome Measures ............................................................................... 120 22.214.171.124 Percentage of response in the group of individuals with SCI and in the group of ablebodied control participants .................................... 120 126.96.36.199 Rates of response in each individual ............................................ 121 188.8.131.52 Subjectively perceived startle ....................................................... 122 4.2.7 Data Analysis ........................................................................................ 122 184.108.40.206 Data processing ........................................................................... 122 220.127.116.11 Short term habituation .................................................................. 122 18.104.22.168 Long term habituation .................................................................. 122 22.214.171.124 Effect of position .......................................................................... 123 126.96.36.199 Statistical analysis ........................................................................ 123 4.3 Results ............................................................................................................ 126 4.3.1 Demographic Information ...................................................................... 126 4.3.2 Percentages of Response ..................................................................... 126 188.8.131.52 Short term habituation .................................................................. 126 184.108.40.206 Long term habituation .................................................................. 127 220.127.116.11 Effect of position on specific muscle groups ................................ 128 18.104.22.168 Additional exploration of the data ................................................. 129 4.3.3 Subjectively Perceived Startle ............................................................... 130 4.4 Discussion ...................................................................................................... 131 4.4.1 Orbicularis Oculi Responses ................................................................. 131 4.4.2 Decrease in Short Term Habituation ..................................................... 132 4.4.3 Long term Habituation in AbleBodied Individuals ................................. 133 4.4.4 Influence of Position on Percentage of Response of Specific Muscle Groups ........................................................................................................ 134 4.4.5 Withingroup Differences ....................................................................... 137 4.4.6 Shape of Responses ............................................................................. 138 4.4.7 Meaning of Increased ASR Responses ................................................. 139
10 4.4.8 Differences in P erceived Startle ............................................................ 140 4.6 Tables and Figures ......................................................................................... 143 5 PROJECT III INDUCING PLASTICITY IN DESCENDING PATHWAYS THROUGH ADAPTABILITY LOCOMOTOR TRAINING ....................................... 166 5.1 Background ..................................................................................................... 166 5.2 Methods .......................................................................................................... 169 5.2.1 Participants ............................................................................................ 169 5.2.2 Training ................................................................................................. 170 5.2.3 Testing ................................................................................................... 172 22.214.171.124 Measures of body structure and function ..................................... 172 126.96.36.199 Measur es of activity ..................................................................... 175 5.3 Results ............................................................................................................ 177 5.3.1 Demographics ....................................................................................... 177 5.3.2 Completeness of Data ........................................................................... 177 5.3.3 Measures of Body Function and Structure ............................................ 178 188.8.131.52 TMS and CCT and iSP Excitability and inhibition of the motor cortex and conduction along the CST ................................................... 178 184.108.40.206 H Reflex Excitability of the spinal cord ...................................... 178 220.127.116.11 VSR Excitability of the vestibular nuclei and conduction along the VST ................................................................................................. 179 18.104.22.168 ASR Excitability of the brainstem areas and conduction along the ReST ............................................................................................... 179 22.214.171.124 ISNCSCI Neurological function at baseline ............................... 180 5.3.4 Measures of Activity .............................................................................. 180 126.96.36.199 Spatiotemporal data and foot trajectories ..................................... 180 188.8.131.52 Clinical assessments .................................................................... 182 5.4 Discussion ...................................................................................................... 184 5.5 Tables and Figures ......................................................................................... 191 6 CONCLUSIONS ................................................................................................... 228 LIST OF REFERENCES ............................................................................................. 232 BIOGRAPH ICAL SKETCH .......................................................................................... 262
11 LIST OF TABLES Table page 2 2 Minimally necessary walking distances and speeds as well as curb heights reported by three different studies ...................................................................... 90 3 1 Classification as BSS/BSPS over time. ............................................................ 106 3 2 Demographic information. ................................................................................ 108 4 1 Inclusion and exclusion criteria for both ablebodied control participants and individuals with SCI for the ASR experiment. ................................................... 143 4 2 Demographic information ................................................................................. 144 4 3 Pattern of elicited EMG responses in reaction to an acoust ic startle in one participant with iSCI with low response rates. ................................................... 145 4 4 Pattern of elicited EMG responses in reaction to an acoustic startle in one participant with iSCI with high response rates. ................................................. 146 4 5 Pattern of elicited EMG responses in reaction to an acoustic startle in one ablebodied control person with low response rates. ........................................ 147 4 6 Pattern of elicited EMG responses in reactio n to an acoustic startle in one able bodied control person with high response rates. ....................................... 148 5 1 Summary of assessments completed by each of the control participants. ....... 191 5 2 Demographic information for individuals with iSCI. ........................................... 192 5 3 Demographic information for ablebodied control participants. ......................... 193 5 4 TMS Recruitment curve para meters in individuals with iSCI ............................ 194 5 5 TMS Recruitment curve parameters in ablebodied control participants. ......... 196 5 6 F wave measurements. .................................................................................... 197 5 7 Central conduction times. ................................................................................. 198 5 8 Ipsilateral si lent pe riods in participants with iSCI ............................................. 199 5 9 Ipsilateral silent period in ablebodied control participants. .............................. 201 5 10 H Reflex measures as an indication of spinal cord excitability in individuals with SCI. ........................................................................................................... 202
12 5 11 H Reflex measures as a measure of spinal cord excitability healthy control participants. ...................................................................................................... 203 5 12 Latencies and amplitudes of Vestibulospinal reflex responses. ........................ 204 5 13 Latencies and amplitudes of Vestibulospinal reflex responses: EMG respons e. .......................................................................................................... 204 5 14 Acoustic startle reflex responses in individuals with iSCI. ................................ 205 5 15 Acoustic startle reflex responses in ablebodied control participants. ............... 206 5 16 Spatiotemporal gait characteristics. .................................................................. 207 5 17 Step clearance when stepping over a 6inch box ........................................... 208 5 18 Step clearance when stepping over a 10inch box. .......................................... 209 5 19 Results from clinical outcome measures. ......................................................... 210
13 LIST OF FIGURES Figure page 2 1 Algorithm used to compute the AIS .................................................................... 91 3 1 Algorithm for the definition of iSCI syndromes .................................................. 111 3 2 Number of available data sets from the Neuromuscular Recovery Network and number of assessments classified as BrownSquard Syndrome ............. 112 4 1 Short term habituation. ..................................................................................... 149 4 2 Long term habituation ....................................................................................... 150 4 3 Differential effect of position on the average percent of responses in flexor and extensor muscles. ...................................................................................... 151 4 4 Differential effect of position on the average percent of responses in trunk and extremity muscles. ..................................................................................... 152 4 5 Differential effect of position on the average percent of responses in proximal and distal muscles. ........................................................................................... 153 4 6 Example of ASR responses from an individual with iSCI that showed generally low response rates to acoustic startle ............................................... 154 4 7 Example of ASR responses from an individual with iSCI that showed generally high response rates to acoustic startle .............................................. 155 4 8 Example of ASR responses from an ablebodied control participant that showed generally low response rates to acoustic startle .................................. 156 4 9 Example of ASR responses from an ablebodied control participant that showed generally high response rates to acoustic startle ................................ 157 4 10 Influence of sex on percentage of response in individuals with iSCI ................ 158 4 11 Influence of age on percentage of response in individuals with iSCI ................ 159 4 12 R elationship between ASR responses rates and motor scores in individuals with SCI ............................................................................................................ 160 4 13 Example of reduced EMG activity after ASR stimulation .................................. 161 4 14 Relationship betweensubjectively perceived startle and EMG responses ....... 162 4 15 Subjectively perceived degree of startle to an acoustic stimulation .................. 163
14 4 16 Subjectively perceived degree of startle to an acoustic stimulation. ................. 164 4 17 Subjectively perceived degree of startle to an acoustic stimulation with regards to position ............................................................................................ 165 5 1 Overview of training and testing days ............................................................... 211 5 2 TMS Recrui tment curve from the right TA muscle in SCIAdapt04 .................... 212 5 3 iSPs recorded in the left TA muscle of SCIAdapt 04. ....................................... 213 5 4 F waves from the right TA muscle in SCIAdapt04. ........................................... 214 5 5 H Reflex recruitment curves and slopes in SCIAdapt01 ................................... 215 5 6 Averaged EMG responses and sway in response to galvanic vestibular stimulation in SCIAdapt02 at MID testing ......................................................... 216 5 7 Average self selected and fastest comfortable walking speed ......................... 217 5 8 Average stride length and cadence .................................................................. 218 5 9 Average distance between heel or toe and the upper edge of the 6 inch box .. 219 5 10 Average distance between heel or toe and the upper edge of the 10 inch box 220 5 11 Average h orizontal distances when stepping over the 6 box ........................... 221 5 12 Average horizontal distances when stepping over the 10 box ......................... 222 5 13 Spinal Cord Injury Functional Ambulation Profile. ............................................. 223 5 14 Neuromuscular recovery scale. ........................................................................ 224 5 15 Enhanced kinem atics while stepping over a box ............................................. 225 5 16 Activity Balance Confidence Scale ................................................................... 226 5 17 Steps taken with the right foot per day in the usual environment ...................... 227
15 LIST OF ABBREVIATION S 10MWT 10 Meter Walking Test ABC Activities Balance Confidence Scale AdaptLT Adaptability Locomotor Training AIS ASIA Impairment Scale AP Anal Pressure ASIA American Spinal Injury Association ASR Acoustic Startle Response AUC Area Under the Curve BasicLT Basic Locomotor Training BB Biceps Brachii Muscle BSPS Brown Squard Plus Syndrome BSS Brown Squard Syndrome BWS Body Weight Support CCT Central Conduction Time CLR Cerebellar Locomotor Region CPG Central Pattern Generator cSCI Complete Spinal Cord Injury CST Corticospinal Tract DGI Dynamic Gait Index EC Extensor Carpi Radialis Muscle ES Errector Spinae Muscle FC Flexor Carpi Radialis Muscle FES Functional Electrical Stimulation FIM Functional Independence Measure
16 ICF International Classification of Functioning, Disability and Health iSCI Incomplete Spinal Cord Injury ISCOS International Spinal Cord Society ISNCSCI International Standards for Neurological Classification of Spinal Cord Injury LEMS Lower Extremity Motor Score LT Locomotor Training LT Light Touch Score MEP Motor Evoked Potential MH Medial Hamstring Muscles MLR Mesencephalic Locomotor Region MS Motor Score MTh Motor Threshold NLI Neurological Level of Injur y NMRS Neuromuscular Recovery Scale (Phases) OO Orbicularis Oculi Muscle PLR Pontine Locomotor Region PP Pin Prick Score PPN Pedunculopontine Nucleus RA Rectus Abdominis Muscle ReST Reticulospinal Tract SAM Step Activity Monitor SCI Spinal Cord Injury SCI FAP Spinal Cord Injury Functional Ambulation Profile SCIM Spinal Cord Independence Measure SCOM/SC Sternocleido occipito mastoideus Muscle
17 SLR Subthalamic Locomotor Region Sol Soleus Muscle SSEP Somatosensory Evoked Potential TA Tibialis Anterior Musc le TB Triceps Brachii Muscle TMB Timed Movement Battery TMS Transcranial Magnetic Stimulation UEMS Upper Extremity Motor Score VAC Voluntary Anal Contraction VM Vastus Medialis Muscle VSR Vestibulospinal Reflex VST Vestibulospinal Tract WHO World Health Or ganization
18 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 PRESENCE, ASSESSMENT AND PLASTICITY OF DESCENDING PATHWAYS AFTER SPINAL CORD INJURY By Martina Rebekka Spiess December 2012 Chair: Andrea Behrman, PT, PhD Cochair: Carolynn Patten, PhD, PT Major: Rehabilitation Science Spinal cord injury (SCI) results in diminished voluntary muscle control and consequent walking deficits. This research focuses on the plasticity of the descending motor pathways in response to iSCI and training. The goals were to: 1) describe and assess the presence of spared or recovered neural connec tions and 2) modify existing locomotor training approaches in order to enhance the supraspinal center involvement. We first describe the functional abilities of individuals with BrownSquard Syndrome (BSS). Human neurologic deficits after BSS are comparab le to the hemisection model prevalently used in SCI animal research. Our research showed that individuals with BSS, when classified according to three recently used definitions, are not a homogeneous group, but differ substantially in their functional abil ities. A consensus in definitions would improve the use of this model and the comparability to the animal literature. Secondly, we examined the function of a motor tract. The Reticulospinal Tract (ReST) relays information to the spinal cord that is important especially for initiation of walking. The neurophysiological test used to examine the functional integrity of this
19 tract, the acoustic startle response (ASR), has not been well described in persons with SCI. Specifically, habituation of the response and means to avoid or delay this habituation are poorly understood. Our research shows that more stable responses could be elicited in individuals standing on compliant ground rather than lying supine. Also, in healthy control participants, when tested a sec ond time after 48hrs, responses are reduced when compared to the first data collection. Finally, t he ASR and other neurophysiological and clinical test s were used to show that neural plasticity between centers above and below the spinal lesion can be induc ed by 6 weeks of locomotor training. No difference between 3 weeks of basic locomotor training and 3 weeks of adaptive training could be detected in our sample of 3 participants. This case series provides important data to adequately plan a larger study in vestigating the differential effects of basic locomotor training and specific adaptability training designed to induce plasticity in the spinal connections below and above the lesion.
20 CHAPTER 1 INTRODUCTION, AIMS A ND HYPOTHESIS Motor and sensory pathways necessary for walking function are often damaged following SCI, resulting in walking dysfunction or complete loss of walking ability. Until recently the central nervous system (CNS) was considered hard wired and incapable of regeneration. ( Ramon y Cajal 1928) Current research provides evidence that this is not true, and that positive changes in the brain and spinal cord can occur after injury; i.e., the CNS is plastic (for example, refer to Dobkin, Edgerton et al. 1992; Dobkin 1993; Dobkin, Harkema et al. 1995; Byl and Melnick 1997; Edgerton, de Leon et al. 1997; Edgerton, Roy et al. 1997; Basso 2000; Edgerton, Leon et al. 2001; Behrman, Bowden et al. 2006; Cha, Heng et al. 2007; Edgerton, Courtine et al. 2008; Lynskey, Belanger et al. 2008) It is also clear that this plasti city can be induced by behavioral interventions/training ( Dobkin 1993 ; Byl and Melnick 1997; Barbeau, Norman et al. 1998; de Leon, Hodgson et al. 1998; Dietz, Wirz et al. 1998; Barbeau, Fung et al. 2 002 ; Barbeau, Ladouceur et al. 2002 ; Behrman, Bowden et al. 2006; Behrman and Harkema 2007; Edgerton, Courtine et al. 2008; Lynskey, Belanger et al. 2008 ; Rojas Vega, Abel et al. 2008 ) Classic studies of spinalized cats have demonstrated that basic stepping movements can be generated by the lumbar spinal cord in the absence of supraspinal input. ( Edgerton, Roy et al. 1997; de Leon, Hodgson et al. 1998; Edgerton, Leon et al. 2001; Edgerton, Tillakaratne et al. 2004) Alternatively, sensory input from the periphery can stimulate stepgenerating neural netw orks in the lumbar spinal cord (this has been referred to as bottom up approach) Based on this knowledge, locomotor training
21 strategies were developed to promote walking recovery in humans by providing the spinal cord with appropriate sensory inputs dur ing task specific practice. These approaches have been successful in aiding individuals with incomplete SCI (iSCI) to relearn basic overground stepping ( Wernig and Muller 1992 ; Dietz, Wirz et al. 1997; Dietz, Wirz et al. 1998; Barbeau, McCrea et al. 1999; Wernig, Nanassy et al. 1999; Behrman and Harkema 2000; Harkema 2001 ; Barbeau, Fung et al. 2002; Behrman, Lawless Dixon et al. 2005; Field Fote, Lindley et al. 2005; Wirz, Zemon et al. 2005; Dietz 2009; Field Fote and Roach 2011) However, functional walking requires skills beyond basic stepping such as stepping over an obstacle or managing uneven terrains. Animal research suggests that spinal cord mechanisms are not sufficient for such tasks. Supraspinal input is necessary for these adaptive behaviors (top down input) ( Tester and Howland 2008; Doperalski, Tester et al. 2011) Human bipedal locomotion involves significantly more supraspinal input than quadraped animal locomotion due to the additional requirement of maintaining upright posture ( Barthelemy, Grey et al. 2011) This demand increases even more when individuals are challenged with more demanding walking tasks such as stepping over a puddle or keeping ones balance while walking in a crowded environment ( Kandel, Schwartz et al. 2000; Capaday 2002 ; Jahn, Deutschlander et al. 2008) The focus of this research was to investigate the plasticity of the descending motor pathways in response to injury, and then to training. Specifically we examined the corti co spinal tract, the vestibulospinal tract and the reticulospinal tract. These tracts play a role in voluntary movement, posture, balance and initiation of walking ( Baloh and
22 Honrubia 1979; Davis, Gendelman et al. 1982 ; Koch 1999; Kandel, Schwartz et al. ; Day and Fitzpatrick 2005; McKay, Lee et al. 2005; Bronstein 2006; Netter 2006; Jahn, Deutschlander et al. 2008; Baloh and Kerber 2010; Felten and Shettz 2010; Day, Ramsay et al. 2011 ; Le Ray, Juvin et al. 2011) The overall goal was to examine if locomotor training approaches could be designed to recruit supraspinal pathways necessary for the performance of successful adaptive walking behaviors. The functional integrity of each of the aforementioned tracts can be measured noninvasively using available neurophysiological tests. It has to be mentioned that none of the stimuli used in these tests will elicit responses isolated to the particul ar tract of interest. Rather, parallel transmission through several different pathways is more likely ( Britton, Day et al. 1993; Liechti, Muller et al. 2008) Regardless, they can still provide important information about the descending motor pathways. The corticospinal tract can be readily assessed using Transcranial Magnetic Stimulation (TMS), a well established clinical and research tool (for example, refer to McKay, Stokic et al. 1997; Thomas and Gorassini 2005; Diehl, Kliesch et al. 2006; Ellaway, Catley et al. 2007; Gorassini, Norton et al. 2009) Measurement of the vestibulospinal tract (VST) using the vestibulospinal reflex (VSR) has received less attention, but has potential as a useful tool ( Goldberg, Fernandez et al. 1982; Fitzpatrick and Day 2004; Day and Fitzpatrick 2005; Liechti, Muller et al. 2008; Day, Ramsay et al. 2011; GuzmanLopez, Buisson et al. 2011) T he acoustic startle response (ASR) however, believed to measure functional integrity of the reticulospinal tract (ReST) is only described minimally in a human population with SCI ( Brown, Day et al. 1991; Yeomans and Frankland 1995 ; Jankelowitz and Colebatch 2004 ; Kumru, Vidal et al.
23 2008) This test may serve as a useful prognostic and/or outcome measurement tool similar to TMS and VSR me asurements. However, there is a lack of information about appropriate stimulation parameters and normative values, especially in a population with SCI. The hemisection SCI model has often been used in animal (cat) models to study the spinal cords response to, and recovery from, injury (for example ( Bernstein and Bernstein 1971; Murray and Goldberger 1974 ; Kato, Murakami et al. 1984; Kato, Muraka mi et al. 1985; Eidelberg, Nguyen et al. 1986; Basso, Murray et al. 1994; Kuhtz, Boczek et al. 1996; Tester, Plaas et al. 2007; Tester and Howland 2008; Arvanian, Schnell et al. 2009; Hall and Traystman 2009) While cats regain considerable stepping abilities after a spinal cord hemisection, ( Eidelberg, Nguyen et al. 1986; Kato 1989; Basso, Murray et al. 1994; Kuhtz Buschbeck, Boczek Funcke et al. 1996; Kuhtz, Boczek et al. 1996) and even remaster higher skilled walking tasks to a certain extent ( Tester and Howland 2008 ; Jefferson, Tester et al. 2011) recovery in humans with similar lesions has not been as successful ( Dietz, Wirz et al. 1998; Mehrholz, Kugler et al. 2008) This apparent g ap in translation of therapies between cats and humans could be in part due to the fact that hemisectioned cats possess intact descending systems on the nonaffected side of the body. These intact tracts might bear potential for plasticity. In human iSCI, there is a very similar model, the BrownSquard Syndrome (BSS). Individuals with a BSS suffered from a hemi disruption of the spinal cord with loss of motor function and light touch sensation ipsilaterally below the lesion as well as loss of pain and tem perature sensation contralesionally ( ASIA 2011) Similar to the hemisectioned cats, they have intact descending systems on the nonaffected side of
24 the body which possibly are conductive to recovery of function. There is debate in the literature as to whether these individuals recover to a lar ger extent than other individuals with iSCI. However, vastly different definitions of BSS have been used when describing the extent of recovery in these patients and it is therefore challenging to compare the results from the individual papers ( Taylor and Gleave 1957; Little and Halar 1985 ; Koehler and Endtz 1986; Roth, Park et al. 1991; Hayes, Hsieh et al. 2000; Wirz, Zorner et al. 2009; Ye, Jia et al. 2009; Pouw, van de Meent et al. 2010; Wu, Ho et al. 2010) 1.1 Aims and Hypotheses 1.2 Project I Defining BrownSquardSyndrome: Asymmetrical Sparing of Descending Pathway s A controversy in the literature is whether patients with BrownSquard Syndrome (BSS) profit disproportionately from rehabilitative strategies or have more spontaneous recovery when compared to other individuals with iSCI. We will apply three definitions recently used by three different authors to define BSS to a large set of patients and examine the effect of using the different definitions on the estimation of outcomes for the BSS population. Aim: Examine the relationship between BBS definition and func tional abilities We hypothesize that the definition of BSS used will influence: 1) the percentage of people defined as having BSS; and 2) how the group of patients classified as having BSS are performing on functional outcome measures (Berg scale, 10meter wal king test, Neuromuscular recovery scale).
25 1. 3 Project II The Acoustic Startle Response: Assessing Plasticity in a Descending Pathway after Spinal Cord Injury Of the three main descending systems involved with human locomotion, the cortico spinal tract (C ST), the vestibulospinal tract (VST) and the reticulospinal tract (ReST) ( Kandel, Schwartz et al. 2000 ; Basso, Beattie et al. 2002; Dietz 2002 ; Yang and Gorassini 2006) neurophysiological correlates have been well described for the CST and the VST, while much less is known regar ding the ReST. The ASR may serve as an effective prognostic tool and outcome measure of neuroplasticity in response to locomotor training and potentially other therapeutic interventions targeting neuromuscular recovery ( Harkema, Gerasimenko et al. 2011) A particular difficulty when using the ASR as a measurement tool to assess ReST functi on is the rapid habituation, preventing repeated testing to reach stable results ( Brown, Rothwell et al. 1991) Aim: Develop a protocol to minimize short term and long term habituation of the ASR and ther efore render this test more useful as an outcome tool. We hypothesize that: 1) long term habituation will not be present in healthy controls over the timeframe of 48hrs; 2) habituation will be delayed when compared to descriptions from the literature (responses extinguished after 26 stimulations) if stimulations are presented while body position is randomly varied; and 3) when the ASR is conducted in healthy controls and individuals with iSCI in 3 different body positions (supine, standing and standing on compliant
26 ground), the probability of eliciting a response will be increased in parallel with the increasing postural demands of the tested positions 1. 4 Project III Inducing Plasticity in Descending Pathways through Locomotor Training The environment of the body weight support and treadmill provides a safe and permissive environment to help individuals with iSCI recover basic overground stepping patterns. ( Wernig, Muller et al. 1995 Barbeau, 2003 #266 ; Behrman and Harkema 2000 ; Yang and Gorassini 2006; Behrman, Nair et al. 2008; Harkema, Schmidt Read et al. 2011) Human upright bipedal walking also requires supraspinal input ( Capaday, Lavoie et al. 1999; Capaday 2002; Yang and Gorassini 2006) and therefore engaging supraspinal centers during training could lead to improved outcomes. Also, it has been suggested that the ability to engage such supraspinal centers is dependent on the amount of spared functional integrity in the spinal tracts connecting those centers to the alphamotor neurons and interneurons in the spinal cord ( Eidelberg, Walden et al. 1981; Cao, Zhang et al. 2005 ; Hill, Zhang et al. 2009) and that some descending tract func tion is in fact necessary to regain independent walking function overground ( Basso, Beattie et al. 1996; Dobkin, Barbeau et al. 2007 ; Sturt, Holland et al. 2009) Aim : Examine the relationship between the functional integrity of the main motor tracts and functional outcome measures in the individuals with iSCI before and after 6 weeks of standard locomotion traini ng and 6 weeks of adaptive locomotor training. We hypothesize that: 1) individuals capable of basic overground stepping with difficulties in more challenging walking tasks will profit from training adaptive walking tasks
27 (AdaptLT) in the permissive, treadmil l with body weight support environment and overground; and 2) that neuroplastic changes will be apparent at both the body structure and function and activities levels.
28 CHAPTER 2 LITERATURE REVIEW 2.1 Human Spinal Cord Injury 2.1.1 Incidence, Prevalence and Demographics of Human Spinal Cord Injury An estimated 12,000 Americans (40 cases per m illion population) suffer from a spinal cord injury (SCI) ( NSCISC 2011 ) and its functional, psychological and economical consequences ( Anderson 2004 ; Jannings and Pryor 2012; Marti, Reinhardt et al. 2012) annually. Incidence rates have been reported between 27.1 and 83 cases per Million population in different US states ( Warren, Moore et al. 1995; Surkin, Gilbert et al. 2000; Burke, Linden et al. 2001) and between 10.4 and 46.2 cases per m illion population worldwide ( Martins, Freitas et al. 1998; van Asbeck, Post et al. 2000 ; Wyndaele and Wynda ele 2006) Prevalence of SCI is less frequently reported ( Blumer and Quine 1995; Dahlberg, Kotila et al. 2005; Wyndaele and Wyndaele 2006) An international comparison of prevalence estimates of SCI found that values between 11 and 112 cases per 100,000 persons have been reported ( Blumer and Quine 1995) With life expectancy increasing f or individuals with SCI ( NSCISC 2011 ) higher prevalence is to be expected in the future, assuming the incidence rate remains stable. Spinal cord injury is most common in young adults although the average age at injury increased in parallel to the average age of the general population in the US from 28.7 years in the seventies to 40.7 years since 2005 ( NSCISC 2011 ) Worldwide, most demographic studies of patients with SCI report average ages between 30 and 40 years at the time of injury ( Wyndaele and Wyndaele 2006) In the US, over 80% ( NSCISC 2011) o f the patients are male and this has changed only minimally over the last 30 years. Different rates of SCI can also be seen among different ethnic groups in the US
29 ( NSCISC 2011 ) Since 2005, 66.5% of injured persons are Caucasian (compared to 72% in the general population in a similar time period), 26.8% are African Americans (compared to 12.6% of the general population in a similar time period) and 2% are Asian (compared to 4.8% of the general population in a similar time period) ( Humes, Jones et al. 2011; NSCISC 2011 ) Motor vehicle crashes are the most common traumatic cause of spinal cord injury, followed by falls and acts of violence ( NSCISC 2011 ; Devivo 2012) Thereby, falls are the most common traumatic etiology for persons age of 60 or older. It is anticipated that the incidence of SCI will increase as Baby Boomers age, and falls will become a progressively more important cause for SCI ( Devivo 2012) Depending on the report, between 25 and 80% of total admissions for SCI ( New and Sundararajan 2008; Cosar, Yemisci et al. 2010; Scivoletto, Farchi et al. 2011) are due to nontraumatic events, such as vertebral spondylosis, tumors or vascular and congenital diseases ( McKinley, Seel et al. 1999; Sturt, Holland et al. 2009) These individuals are generally older ( McKinley, Seel et al. 1999; New and Sundararajan 2008 ; Sturt, Holland et al. 2009; Cosar, Yemisci et al. 2010) and have more comorbiditi es ( Sturt, Holland et al. 2009) than individuals that su ffer a traumatic SCI. Non traumatic SCI occurs more often at a thoracic level, leading to paraplegia or paraparesis rather than quadriplegia and it more often causes incomplete injuries ( Sturt, Holland et al. 2009) With the increase in age in the general population, it is to be expected that nontraumatic etiologies will become increasingly important in the future ( Sturt, Holland et al. 2009) Individuals with incomplete spinal cord injur ies (iSCI) are becoming a large (between 47.9% and 61.2%, ( Kurtzke 1977; van Asbeck, Post et al. 2000; Dahlberg,
30 Kotila et al. 2005; Wyndaele and Wyndaele 2006; NSCISC 2011 ) ) and increasing ( NSCISC 2011 ) portion of all SCIs. These individuals possess some spared or restored function below the level of injury. Hereby, please note that the same definition of the term incomplete was not used in the different studies cited. According to the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI), an injury is classified as incomplete if the individual exhibits any sensory or motor function in the last sacral segm ents (S4/5, voluntary anal contraction or deep anal pressure) ( ASIA 2011 ) The ISNCSCI divide incomplete injuries into 5 clinical syndromes depending on the pattern of functional loss ( ASIA 2011 ) One of them, the Brown Sequard syndrome (BSS) is characterized by damage primarily to one half of the spinal cord, and occurs in about 24% of all patients with SCI ( Roth, Park et al. 1991; Hayes, Hsieh et al. 2000; McKinley, Santos et al. 2007) or 17.1% diagnosed with one of the five clinical SCI syndromes ( McKinley, Santos et al. 2007 ; ASIA 2011 ) 2.1.2 Functional Consequences after SCI One third to one fourt h of all individuals with SCI regain some walking function after rehabilitation ( Barbeau, Norman et al. 1998; Barbeau, Ladouceur et al. 1999; Sisto, Forrest et al. 2008) Regaining walking was considered the highest priority in 15.9% of patients with paraplegia and 7.8% of those with tetraplegia (no information on injury severity o f respondents was provided in this study) ( Anderson 2004) Peo ple with initial sparing of the sacral segments ( Waters, Adkins et al. 1991 ; Waters, Adk ins et al. 1994; Kirshblum and O'Connor 1998; Geisler, Coleman et al. 2001; Fawcett, Curt et al. 2007; Kirshblum, Botticello et al. 2011) or spared motor function below the level of the lesion ( Dobkin, Barbeau et al. 2007; Sturt, Holland et al. 2009) have a high probability of regaining independent walking function at leisurely velocities. However, these
31 individuals experience a wide range o f deviations from severely disabled to near normal with regards to gait mechanics ( Barbeau, Norman et al. 1998; Barbeau, Ladouceur et al. 1999; Nadeau, Arsenault et al. 1999; Pepin, Ladouceur et al. 2003 ; Pepin, Norman et al. 2003) muscle stiffness ( Dietz, Quintern et al. 1981; Fung and Barbeau 1994) balance ability ( Shumway Cook and Woollacott 2001; van Hedel, Wirth et al. 2005; Day, Kautz et al. 2011) walking speed ( Barbeau, Norman et al. 1998 ; van Hedel, Dietz et al. 2007; Sturt, Holland et al. 2009) and dependency on assistive devices ( Barb eau, Fung et al. 2002; Sisto, Forrest et al. 2008; Harkema, Behrman et al. 2011) Aberrant gait patterns and accompanying joint pathomechanics can lead to degenerative joint disease and pain. These factors in turn have been associated with an increased risk of falls in individuals with iSCI ( Brotherton, Krause et al. 2007; Brotherton, Krause et al. 2007) 2.1.3 Other Consequences after SCI Due to the young age of m any SCI patients, individuals are either in school or are developing their careers when they are injured ( NSCISC 2011 ) Because of the deficits these patients experience after the injury, barriers to return to gainful employment are high with returnto work rates low ( Anderson, Dumont et al. 2007) These rates differ between countries; for example, 35.2% are working at 20 years post injury in the US ( NSCISC 2011 ) compared to 63.8% in a sample of 495 indiv iduals at least 1 year post injury in Switzerland ( Marti, Reinhardt et al. 2012) Low returnto work rates in these young individuals can lead to substantial economic consequences for the patients and their families in addition to the functional and psychological burden ( NSCISC 2011 ) I n contrast, sufficient i ncome and employment were associated with better QOL ( Dajpratham and Kongkasuwan 2011) Though individuals with SCI report the need for
32 psychological support ( Jannings and Pryor 2012) rates of depression ar e only marginally higher than in ablebodied individuals, unlike in patients with stroke or multiple sclerosis. ( Hassanpour, Hotz Boendermaker et al. 2011) 2.2 Descending Systems in H umans Several descending tracts relay information from the brain to the spinal cord and finally to the muscles. Not all tracts are developed to the same extent in all species. In humans, mainly three tracts are of importance: the corticospinal tract (CST), the vestibulospinal tract (VST) and the reticulospinal tract (ReST) ( Nathan, Smith et al. 1996; Kandel, Schwartz et al. 2000; Lemon 2008) The anatomy and physiology of each of these tracts will be described in the following sections. 2.2.1 Corticospinal Tract The corticospinal tract originates in the primary motor cortex, and supplemental and premotor cortical areas. Pyramidal cell axons descend from these areas to the ventral spinal cord horns ( Lemon 2008) Approximately 80% ( Felten and Shettz 2010) to 90% ( Kandel, Schwartz et al. 2000) of the fibers cross to the contralateral side in the decussation of the medullary pyramids. The contralateral fibers travel in the lateral CST, while those that continue to descend ipsilaterally form the anterior CST. Most of these anterior CST fibers cross at the spinal segmental level through the anterior white commissure. Fibers from the lateral CST terminate on alpha and gamma motorneurons ( Capaday 2002) of distal muscles, especially the hand and fingers in the upper extremity ( Felt en and Shettz 2010) and ankle dorsiflexors in the lower extremity ( Yang and Gorassini 2006) either directly or through interneurons at the segmental level ( Felten and Shettz 2010) The monosynaptic connections with the alphamotorneurons are particularly important for fine voluntary control of the muscles ( Lemon 2008 ) These
33 are connections that are present only in primates and are more developed in humans than in nonhuman primates ( Lemon 2008 ) Lastly, the terminations on interneurons play an essential role in intermuscular and interlimb coordination ( Kandel, Schwartz et al. 2000; Lemon 2008) 2.2.2 Vestibulo spinal Tract The vestibular labyrinth consists of three semicircular canals, a saccule and utricule, filled with endolymph on each side of the body. Stereocilia of hair cells in the epithelium of these organs deflect according to linear and angular acceleration of the body in space and transduce those mechanical stimuli into action potentials ( Kandel, Schwartz et al. 2000) Action potentials are transmitted through the vestibular nerve to the vestibular nuclei. Of the four nuclei, the lateral nucleus (Deuters nucleus) is the one mostly involved in postural control; the other nuclei are dedicated to the occulomotor centers for stabilization of gaze and transmitting sensory input to the thalamus and the primary sensory cort ex. Further connections are formed with the reticular formation as well as with the contralateral vestibular nuclei ( Kandel, Schwartz et al. 2000) 2.2.3 Reticulospinal Tract The ReST consists of two pathways: 1) the lateral, medullary pathway that originates in the nucleus gigantocellularis of the medulla and 2) the medial, pontine pathway originating in the nucleus reticularis pons caudalis There is debate as to where the axons o f the ReST are traveling in the spinal cord. Nathan et al ( Nathan, Smith et al. 1996) pointed out the vast differences in eight major anatomical text books with regard to the location of the reticulospinal fibers in the spinal cord. This controversy is likely due to the fact that most of the knowledge has been derived from different animal models ( Davis, Gendelman et al. 1982; Nathan, Smith et al. 1996 ; Koch 1999)
34 However, Nathan et al (for example Nathan, Smith et al. 1996) confirmed findings derived from animal work that axons of the pontine ReST travel ipsilaterally in the anterior spinal cord while the medullary ReST travels bilaterally in th e lateral spinal cord. ReST fibers are scattered throughout the columns rather than forming distinct tracts like the fibers of the VST or the CST ( Nathan, Smith et al. 1996) Fibers from both the pontine and the medullary ReST terminate on interneurons and lower motor neurons that finally innervate axial and proximal muscles ( Davis, Gendelman et al. 1982; Koch 1999; Felten and Shettz 2010) 2.3 The BrownSquard Syndrome 2. 3 .1 Definition of the BrownSquard and BrownSquard Plus Syndrome A large and increasing number of individuals who suffer from a spinal cord injury have some spared or restored function below the level of injury ( NSCISC 2011 ) Depending on the symptomatic impairments present in each of these iSCI patients, the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) differentiate between five clinical syndromes ( ASIA 2011 ) One of those five, occurring in about 2 4% ( Roth, Park et al. 1991; Hayes, Hsieh et al. 2000; McKinley, Santos et al. 2007) of all SCI, is the Brown Squard Syndrome (BSS), a condition arising from a hemi severance of the spinal cord. Due to the specific individual decussation heights of the spinal tracts, this lesion leads to a characteristic asymmetric pattern of neurologic deficits below the level of lesion ( Brown Squard 1855) (Table 21). Individuals experience reduced muscle function, as well as vibration, positionand light touch sensation on the ipsilesional side with simultaneous contralesional loss of sensitivity to pain and temperature ( Mumenthaler and Mattle 2002 ) A variety of classification systems for BSS have been developed. The term Brown Squard Pl us
35 Syndrome (BSPS) is used for those cases with only partial losses of the respective functions ( Koehler and Endtz 1986) More recently, authors attempted to quantify the term relatively wh en classifying patients into the BSS or BSPS group. Wirz et al ( Wirz, Zorner et al. 2009) regarded differences of at least 19 points between the ISNCSCI motor scores of the right and left key muscles as BSS syndrome. Their definition did not include sensory scores. Pouw et al included patients with BSPS into their analysis if they had any asymmetry (difference>0 in muscle or pin prick scores) with greater losses of the two modalities being on opposite sides. Sensation to light touch was not used to classify participants ( ( Pouw, van de Meent et al. 2010) and personal communication). Both studies included only participants with cervical injuries. Using absolute thresholds while declaring individuals as having BSS or BSPS seems to introduce possible bias towards patients with higher levels of injury reaching that threshold more readily. Hayes et al. ( Hay es, Hsieh et al. 2000) developed a series of different decision algorithms based on the ISNCSCI to classify the 5 syndromes described in the standards. Due to the absence of an external criterion, no validation of the algorithms in the classical sense w as possible. However, the authors favor one algorithm (presented in Figure 31), due to its clinical reasonableness in a set of 56 subjects 2. 3 .2 Functional Recovery in Individuals with BSS Individuals with BSS syndrome traditionally have been considered to have a better prognosis when compared to other individuals with iSCI ( Bosch, Stauffer et al. 1971; Little and Halar 1985; Koehler and Endtz 1986; Na than 1994; McKinley, Santos et al. 2007; Moreh, Meiner et al. 2009) However, more recent articles do not confirm this impression ( Wirz, Zorner et al. 2009; Pouw, van de Meent et al. 2010) Reasons for this
36 disagreement, besides the already mentioned dif ferences in defining BSS or BSPS, could be differences in the study designs and sizes, endpoints chosen and described, etiologies and injury severity of patients included, assessment time points relative to the time of injury, and the control group to whic h the individuals with BSS have been compared. Only a few studies clearly define inclusion and exclusion criteria for both BSS/BSPS and control groups ( McKinley, Santos et al. 2007; Wirz, Zorner et al. 2009; Pouw, van de Meent et al. 2010) The specific etiologies and mechanisms that typically lead to BSS also mi ght introduce some bias. The major causes of BSS are stab wounds, gunshot wounds and motor vehicle accidents T he incidence of the particular type of injury can vary according to geographic location. Additionally, individuals with BSS are often young males more so than in a general SCI group ( Peacock, Shrosbree et al. 1977 ; Hayes, Hsieh et al. 2000 ; McKinley, Santos et al. 2007) Younger adults a ppear to have an advantage in neurologic recovery, so the relatively younger age of the BSS group might explain some of their greater functional improvement ( Burns, Golding et al. 1997; Kirshblum and O'Connor 2000; Jakob, Wirz et al. 2009) Lipschitz et al. ( Lipschitz 1967; Lipschitz 1973) suggested that the underlying mechanisms of injury, even in those with a stab wound being the common cause of injury, could influence the amount of recovery one can reach. They postulated three different mechanisms of injury as a r esult of a stab wound: direct trauma by the weapon or bone fragments, damage secondary to vessels responsible for spinal cord blood supply or countercoup spinal contusions on the opposite side of weapon penetration. They describe that individuals with damage to
37 the spinal cord secondary to vascular damage had the most improvement while the functional deficits due to direct trauma by the weapon were usually irreversible ( Lipschitz 1967 ; Lipschitz 1973; Pascual, Alcacer et al. 2011) Ultimately, these factors require consideration and control when designing an intervention study comparing individuals with BSS to other iSCI groups. 2. 3 .3 Outcome Measures Used to Describ e the BSS Population Outcome measures in the older studies were often not clearly defined. Interestingly, one paper that is often cited to show the favorable outcome of BSS over other iSCI individuals included outcome measures as crude as "the outcome of s urgical treatment was good, "showing good functional recovery", or "recovered/disabled/dead" ( Koehler and Endtz 1986) The term recovery also has been used with different meanings over time. Taylor et al. ( Taylor and Gleave 1957) for example defined recovery as any movement that was observed in a previously paralyzed limb. Today, we tend to think of recovery as full restoration of function to preinjury levels ( Levin, Kleim et al. 2009) When information on walking function is presented, it is usually restricted to walking function attained/not attained, in certain cases in combination with information on the assistive device used ( Bosch, Stauffer et al. 1971 ; Peacock, Shrosbree et al. 197 7 ; Ye, Jia et al. 2009 ; Wu, Ho et al. 2010) Barthel index, FIM or SCIM scores ( Roth, Park et al. 1991; McKinley, Santos et al. 2007; Moreh, Meiner et al. 2009; Wirz, Zorner et al. 2009; Pouw, van de Meent et al. 2010) or functional walking tests ( Moreh, Meiner et al. 2009; Wirz, Zorner et al. 2009) Only one study ( Little and Halar 1985) included minimal qualitativ e descriptions of walking patterns in their two patients.
38 A few larger studies, all of newer date, could be found that stated clearly defined outcome measures, assessment time points and study populations. McKinley et al. ( McKinley, Santos et al. 2007) compared the syndromes as defined by the ISNCSCI (with an addition of a sixth, very rare posterior cord syndrome) at admission and discharge from initial rehabilitation. The authors concluded that BSS patients have a slightly better outcome, as well as slightly more improvements over their inpatient stay in most subscales of the Functional Independence measure (FIM) ( Keith, Granger et al. 1987; Hall, Cohen et al. 1999) including FIM motor and FIM mobility. They also noticed that BSS patients were the youngest of all groups and the one with the highest number of AIS C and second highest number (after Central Cord SyndromeCCS) of AIS D patients. Those patients are known to have favorable outcomes and the authors did not control for these potential confounders. Also, pre, and post assessments were conducted at admission and discharge from the r ehabilitation facility. No information is available as to the delay between injury onset and admission of the patient to inpatient rehabilitation, but it can be imagined that individuals with BSS, who generally tend to have less severe injuries (AIS C or D ), were transferred earlier during a phase where there is still considerable spontaneous regeneration present ( Fawcett, Curt et al. 2007) This could explain the higher gains in individuals with BSS compared to all other groups during inpatient rehabilitation. Pouw et al. ( Pouw, van de Meent et al. 2010) compared the different parameters from the ISNCSCI and the Spinal Cord Injury Independence Measure (SCIM II) of BSPS patients to those of nonBSPS pat ients at 15 days and 612 months after injury. Pure BSS patients were excluded from the study with the rational that the pure form of t his
39 syndrome is too rare to have clinical relevance. When controlling for severity of injury (AIS), the only differences between BSPS and nonBSPS individuals were found in bladder management. Based on their results, the authors question the importance of differentiating between the different clinical syndromes in iSCI. Wirz et al. ( Wirz, Zorner et al. 2009) compared neurophysiological (Amplitudes of Somatosensory evoked Potentials SSEPs), neurological (ISNCSC I Motor scores) and functional (10m walking test, Walking Index for Spinal Cord Injury WISCI II and SCIM II) outcomes of individuals with BSS and those with central cord syndrome ( CCS ) at 1 month and 6 months post injury. No differences in outcome between the two groups were found. Authors therefore suggest these two anatomically dissimilar types of SCI possess equally efficient mechanisms of functional recovery. The studies by Pouw and Wirz used standardized assessment time points with regards to the time of injury. The amount and nature of therapeutic interventions the individuals received is not documented, but patients with paraplegia received approximately 6 months, those with tetraplegia 9 months of inpatient rehabilitation in the European centers that contributed data to those studies. These numbers tend to be lower for individuals with iSCI. 2. 3 .4 Treatment Effects in Individuals with BSS In individuals with BSS, reports specifically investigating treatment effects are often case studies ( Moreh, Meiner et al. 2009 ; Ye, Jia et al. 2009; Wu, Ho et al. 2010) Most of them focus on acute care orthopedic or surgical treatment ( Ye, Jia et al. 2009 ; Wu, Ho et al. 2010) rather than on locomotor training or any other form of physical therapy. As one of the few studies focusing on a rehabilitation strategy, Moreh et al. ( Moreh, Meiner et al. 2009) assessed the usefulness of robotic locomotor training in a
40 patient with a severe decompression injury due to a diving accident and resulting in a BSS. They concluded that the favorable outcome (improvement in WISCI II from 1 to 15 along with improvements in motor and sensory scores and a conversion from AIS C to AIS D) in this patient despite the severity of his injury hints at the usefulness of locomotor therapy in this population. 2. 3 .5 Underlying Mechanisms of Recovery in BSS and BSPS Very few studies in humans have investigated the underlying mechanisms of functional recovery in individuals with BSS. Most of the evidence available today arose from animal models. Unlike the human BSS, which only occurs in a small part of the SCI population, the hemisection model is regularly used in animal experiments of SCI ( Helgren and Goldberger 1993; Ni, Li et al. 2005 ; Hill, Zhang et al. 2009; Rossignol, Barriere et al. 2009) Most studies using this model investigated the effect of certain therapy modalities in g roups of hemisectioned animals. However, from the results of the placebo control groups, some information can be gained of the underlying mechanisms of recovery from hemisection in animals. In one study investigating the effect of Chondroitinase ABC in SC I, hemisectioned cats in the control group quickly regained a fair amount of basic walking skills as well as a limited amount of skilled walking abilities. Skilled walking tasks were executed using compensation strategies, for example by means of only three limbs or altered kinematics (more knee flexion) when compared to preinjury. These limited functional improvements went along with the count of a limited amount of descending axons below the lesion ( Tester and Howland 2008; Jefferson, Tester et al. 2011) suggesting that axonal regeneration might play a role in functional recovery.
41 Incomplete recovery of skilled hand function of between 20 and 90% has been shown in macaque monkeys after cervical hemisectio n. This functional recovery was shown to be inversely correlated to the extent of the lesion and accompanied by an increase in total axonal arbor length, an increase in terminal buttons in the axons and by an increase of labeled axons below the lesion site. Most of these axons arose from the contralesional side ( Freund, Schmidlin et al. 2006) Kato et al were able to show a certain degree of functional recovery of locomotor function in a cat after staggered hemisections, leading to the conclusion, that collateral sprouting and connections through i nterneurons were established between the spared parts of the descending tracts ( Kato, Murakami et al. 1985) One very similar study is available from human data. Nathan et al. conducted cordotomies in 44 terminally ill cancer patients for the purpose of reducing neuropathic pain. They assessed motor recovery and were able to obtain histological data from the spinal cords after the patient s were deceased ( Nathan 1994) Fiber tracts below the level of the lesion were interpreted as fibers being spared from the induced injury. No mentioning of the possibility that some of the fibers were regenerated or had formed new connections was made. However, they also describe a case where staggered hemisections were performed at different time points. The patient recovered walking fun ction after the first hemisection, but recovery was lost after the second surgery on the opposite side. Nathan et al. concluded that recovery must have been due to contralateral fibers that compensated for the loss on the ipsilateral side after the first surgery ( Nathan and Smith 1973) He also proposed that, as recovery occurred at such a fast rate, the contralateral corticospinal tract already was exerting bilateral effects
42 before the first injury. Therefore, the connections were already present and when used increasingly, those already present connections might have become strengthened enough to lead to the observed functional recovery ( Cooke and Bliss 2006) Another human study attempting to investigate underlying mechanisms of recovery in individuals with BSS in vivo used Somatosensory Evoked Potentials (SSEPs), a neurophysiological correlate for dorsal column function ( Wirz, Zorner et al. 2009) They were unable to show a relationship between the amplitudes and latencies of SSEPs and the functional recovery in their individuals with BSS. However, whereas sensory inputs are important for motor recovery ( Rossignol 2002) the inputs are not propagated through the same spinal tracts and the authors state that they did therefore not expect a strong correlation. In fact, it has been shown that measurements of the spinothalamic tract, probably due to its closer anatomical vicinity with the corticospinal tract would provide a better correlate for motor function than the dorsal columns ( Crozi er, Graziani et al. 1991; Burns, Golding et al. 1997) 2. 4 Neural Control of Walking Walking consists of three main components: the ability to take steps, the ability to keep ones balance and the ability to adapt ones stepping pattern to demands of the environment ( Behrman, Bowden et al. 2006) Neural control of these different components in healthy human adults occurs at both the spinal and supraspinal levels, albeit to different degrees for the different components ( Dimitrijevic, Gerasimenko et al. 1998; Capaday, Lavoie et al. 1999; Schubert, Curt et al. 1 999 ; Capaday 2002; Dietz 2003)
43 2. 4 .1 Spinal Level In his experiments with cats, dogs and other animal models in the early 20th century, Charles Sherrington discovered a reciprocal pattern of flexor and extensor activity in the decerebrated animal that he contributed to patterned reflex activity ( Sherrington 1910) During the same time period (1911), Thomas Graham Brown, a research Fellow in Sherringtons lab, was able to show a decerebrated cat still produced rhythmic contractions of the tibialis anterior and gastrocnemius muscles when it was also deafferented, concluding that this rhythmic pattern could not be due to patterned reflexes but was spinal in origin. He described lumbar centers organized in antagonistic pairs that evoked these patterns ( Brown 1911) His research was not well appreciated or cited until Lundberg discovered the cellular substrate of Browns half center theory ( Stuart and Hultborn 2008) Lundberg et al. were investigating the influence of descending monaminergic noradrenergic and serotoninergic pathways on the spinal networks using intravenous injection of DOPA (L3,4 dihydroxyphenyl alanine). They were able to show that there is a network with mutual reciprocal inhibitory action between flexor and extensor neurons that lead to alternating activation of the respective muscles. Lundberg realized that these networks possibly provided the cellular substrate to the half center oscillators that were proposed by Thomas Graham Brown ( Stuart and Hultborn 2008) Since then, the idea of the central pattern generator (CPG) has been widely studied in both animal and human models, not only in locomotion, but also for respiration, swallowing, scratching or whisker movements ( Andersson, Grillner et al. 197 8 ; Grillner and Wallen 1985 ; McClellan and Sigvardt 1988 ; Dietz 1995; Bellingham 1998; Zehr 2005; Ijspeert 2008) Recently, it also has been shown that each half center
44 or hemisegmental locomotor network is capable of producing rhythmic output ( Yang and Gorassini 2006 ; Cangiano, Hill et al. 2012) As for the human CPG, it has been proposed to consist of a combination of major pacemakers in the form of self oscillating cells in the upper, and so called half center oscillators as separate generator cells in the lower spinal cord ( Ivanenko, Poppele et al. 2009) Endogenous oscillator cells are capable of producing rhythmical neurological activity, e.g. by self limitation of the excitatory activity through an inhibitory gene expression product ( Kandel, Schwartz et al. 2000) Half oscillators are two groups of cells that reciprocally inhibit each other, turning a tonic input into a rhythmic output ( Zehr 2005) Evidence suggests that peripheral somatosensory input as well as descending input from supraspinal centers such as the cortex, the cerebellum and the brainstem are integrated at the spinal level and used by the CPG to select an appropriate motor program ( Horak and Nashner 1986; Dietz 2002 ) The CPG has been studied in humans with motor complete SCIs where supraspinal input is minimal to absent ( Dobkin, Edgerton et al. 1992 ; Dobkin, Edgerton et al. 1992; Dobkin, Harkema et al. 1995; Dietz, Muller et al. 2002; Beres Jones and Harkema 2004; Lunenburger, Bolliger et al. 2006; Spiess, Jaramillo et al. 2012) These patients were unable to produce voluntary movement. However, a reciprocal EMG pattern that was influenced by induced changes in the somatosensory feedback (such as alterations in body weight support, treadmill speed and range of motion of the joint s) emerged during manual or robot assisted treadmill training. Calancie et al. describe a very interesting single case of a participant with incomplete injury that experienced involuntary stepping movements when lying supine. These movements could be reliably
45 triggered and modulated by sensory input, the participant was able to voluntarily enhance, but not suppress them ( Calancie, Needham Shropshire et al. 1994) A ver y similar case is described by Nadeau et al. ( Nadeau, Jacquemin et al. 2010) in a participant with complete severance of the spinal cord. This participant showed continuous alternating movements that were sustained over 23 days even when in a sitting position. In this participant, it was unclear how the movement periods were triggered, but they could be modulated through sensory input similar to the case described by Calancie et al. ( Calancie, Needham Shr opshire et al. 1994) These examples suggest the existence of a CPG capable of producing patterned output without supraspinal input also in humans. Confirming evidence is provided by work in newborn infants. Newborns are capable of producing reciprocal flexor and extensor movem ents in their legs (spontaneously or triggered by peripheral input), even those infants born anencephalic. Therefore, these reciprocal movements may have a spinal origin ( Forssberg 1986; Dietz 2003) It further appears that the CPG is capable of adapting the locomotor pattern to speed changes in a limited fashion ( Beres Jones and Harkema 2004) However, neither patients with complete SCI nor anencephalic infants were able to step with full body weight bearing ( Dietz 1987; Dobkin, Edgerton et al. 1992; Dobkin, Harkema et al. 1995 ; Lunenburger, Bolliger et al. 2006) as is possible in spinal animal models ( Brown 1911; Dobkin, Edgerton et al. 1992; Dobkin 1993 ; Edgerton, Roy et al. 1997; de Leon, Hodgson et al. 1998) Ther efore, while many of the basic mechanisms underlying locomotion seem to have been preserved during the transmission from quadru, to bipedalism ( Nilsson, Thorstensson et al. 1985; Nicol, Granat et al. 1995;
46 Duysens, Van de Crommert et al. 2002; MacKay Lyons 2002; Dietz 2003) there likely is a higher dependence of bipedal upright gait on supraspinal input ( Dietz 1 987 ; Jahn, Deutschlander et al. 2008) Dobkin et al. provided evidence supporting the need of supraspinal input in human uprig ht gait by showing that those individuals that recovered some overground walking function all had at least partially intact descending tract function ( Dobkin 1993) 2. 4 .2 Supraspinal Level 2. 4 .2.1 Gait initiation Th e midbrain locomotor region (MLR), corresponding to the nucleus cuneiformis and the dorsal and anterior part of the pedunculopontine nucleus (PPN) is involved in increasing the muscle tone and initiating gait ( Steeves and Jordan 1980; Douglas, Noga et al. 1993; Whelan 1996; Mori, Matsui et al. 1998; Kandel, Schwartz et al. 2000; Jordan, Liu et al. 2008 ; Le Ray, Juvin et al. 2011) This area was active in humans during gait initiation and an electrical stimulation thereof in decerebrated cats initiated walking ( Shik, Severin et al. 1969; Whelan 1996; Mori, Matsui et al. 1998; Jahn, Deutschlander et al. 2008; Takakusaki, Tomita et al. 2008) Also when stimulated electrically, the pontine locomotor region (PLR) and the subthalamic motor region (SLR) can initiate walking in the cat ( Mori, Matsuyama et al. 1992; Jordan, Liu et al. 2008) Of these, the PLR, but not the SLR was activated in human initiation of (imaginary) gait ( Jahn, Deutschlander et al. 2008) The MLR has connections via the medial formation reticularis through the reticulospinal pathway to both the ventro, and dorsolateral funiculus to activate the CPG in the spinal cord, allowing initiation of gait ( Douglas, Noga et al. 1993; Whelan 1996 ; Jahn, Deutschlander et al. 2008 )
47 2. 4 .2.2 Online control of gait The speed of locomotion seems to be guided much through the same pathways as initiation of gait. In the experiments with decerebrated cats, speed of walking depended on the intensity of the stimulation ( Shik, Severin et al. 1969; Whelan 1996 ; Le Ray, Juvin et al. 2011) while in humans, the cerebellar loco motor region (CLR) seems to have an influence on gait spe ed ( Jahn, Deutschlander et al. 2008) Other regions of the brain are responsible for ref ining the motor pattern in response to feedback from the limbs and again others in guiding limb movement in response to visual input ( Greenlee 2000; Vaillancourt, Thulborn et al. 2003; Lajoie, Andujar et al. 2010) Visual information is essential in skilled walking, e.g. to avoid an upcoming obstacle. It is processed in the visual and motor cortex and while some descending tracts from the motor cortex connect to the cerebellum for integration, there are also direct connections to the spinal cord ( Cooper, Miya et al. 1998) allowing regulation of CPG activity. The cerebellum is mainly involved in refining the motor pattern, receiving proprioceptive inputs from the peripher y as well as from the spinal locomotor system. The cerebellum also integrates visual and vestibular afferent information and controls balance ( Morton and Bastian 2004; Jahn, Deutschlander et al. 2008) Lesions in this area lead to an ataxic or drunken gait, when balance, interlimb coordination and motor fine tuning are impaired ( Morton and Bastian 2004) Finally, the motor cortex is heavily involved in online modification of gait patterns such as increasin g flexor activity when stepping over an obstacle ( Orlovsky 1972; Armstrong 1988; Drew 1988)
48 2. 5 Walking Adaptability 2. 5 .1 What is Needed for Community Ambulation? Successful ambulation in the community requires more than just the ability to take step s on the level floor of the hallway of a physical therapy clinic ( Lerner Frankiel Varcas et al. 1986; Robinett and Vondran 1988; Patla and Shumway Cook 1999; Ladouceur, Bar beau et al. 2003 ; Pepin, Norman et al. 2003; Shumway Cook, Patla et al. 2003; Musselman and Yang 2007; Musselman, Fouad et al. 2009; Andrews, Chinworth et al. 2010; Harkema, Behrman et al. 2011; Jannings and Pryor 2012) Researchers have targeted defining the locomotor abilities necessary to meet walking challenges in everyday life ( Lerner Frankiel, Varcas et al. 1986; Robinett and Vondran 1988; Patla and Shumway Cook 1999; Shumway Cook, Patla et al. 2003; Musselman and Yang 2007; Andrews, Chinworth et al. 2010; Jannings and Pryor 2012) Initially, these reports focused mainly on the minimal distance one has to be able to cover (walkin g endurance) as well as the minimal speed one has to achieve and the height of curbs one encounters in the community ( Lerner Frankiel, Varcas et al. 1986; Robinett and Vondran 1988; Andrews, Chinworth et al. 2010) Lerner Frankiel et al. ( Lerner Frankiel, Varcas et al. 1986) measured curb hei ghts, speeds necessary to cross intersections and distances needed to cross intersections or run common errands such as going to the grocery store, post office or the bank in Los Angeles county. These results are presented in Table 22 They further tested 7 individuals post stroke and 3 individuals post lower extremity ambulation who had been rated as independent community ambulators at their recent discharge from rehabilitation. Only one out of these ten persons achieved the necessary speed to cross a comm ercial crosswalk. All ten participants were able to negotiate curbs if permitted to use their assistive devices.
49 No information is provided as to how many people could complete the task without an assistive device. Only 5/10 persons could cover the distanc e determined necessary to visit a drugstore, 8/10 could reach distances necessary to visit a department store and supermarket. Participants generally considered themselves more independent than they were. The authors suggest that individuals with deficits in community walking had already adapted their lifestyles to avoid certain challenging situations. Robinett et al. ( Robinett and Vondran 1988) used the same measuring protocol as Lerner Frankiel in seven communities in New Mexico and Texas. They divided the communities into rural towns, small towns and cities depending on their populations. While the requirements for successful community ambulation varied according to population size of the community, all distances far exceeded what patients are expected to walk in a PT setting to be considered community ambulators. Over 10 years later, Andrews et al. ( Andrews, Chinworth et al. 2010 ) repeated the study by Lerner Frankiel. They adapted the protocol to meet current conditions by adding distances for superstores, club warehouses and hardware stores. Andrews stated that since the earlier studies in the eighties, a few changes have taken place that influence the walking speeds and distances necessary. First of all, the Americans with Disabilities Act forced retail establishments to provide handicapped accessible parking. Second, so called big box retail stores (large, rectangular singlestory buildings, isolated with immense parking lots) have started appearing and became very common. Andrews therefore added club warehouses, superstores and hardware stores to the list of sites measured. He took measurements at 141 establishments in15 towns and cities in North Carolina. The largest m inimal necessary distance measured was in club
50 warehouses, followed by superstores and hardware stores. The longest overall necessary distance measured was at a club warehouse with 922 meters. However, almost all businesses provided some sort of motorized mobility device or benches. Interestingly in this study, smaller towns required faster crosswalk crossing speeds than larger cities. In summary, the work by Lerner Frankiel, Robinett and Andrews shows that minimal distances and speeds necessary to successf ully ambulate in the community exceed by far the distances that are used to rate somebody as independent community walker upon discharge from a rehabilitation facility ( Lerner Frankiel, Varcas et al. 1986; Robinett and Vondran 1988; Andrews, Chinworth et al. 2010) A person that has enough endurance and speed to fulfill the requirements put forth by the three discussed studies will still face considerable challenges during everyday walking. Patla and Shumway Cook therefore proposed a new framework of communit y mobility. Besides the already mentioned factors minimal walking distance and time constraints, they recommend to take into consideration 6 further dimensions of community ambulation: ambient conditions, terrain characteristics, external physical load, attentional demands, postural transitions and traffic level ( Patla and Shumway Cook 1999) They propose that these dimens ions interact with each other and that the total degree of challenge to the patient is defined through this interaction. What a patient is able to do results from a combination of the abilities of the individual and the environmental demands. For example, walking 20m at a given speed during the day on a flat surface is less challenging than walking the same distance at the same speed at dusk on packed snow. A person can improve by learning a new task or by
51 accomplishing the same tasks under more challenging conditions. The authors investigated how many times older adults with disability encounter or avoid challenges in all 8 dimensions of community ambulation compared to healthy controls ( Shumway Cook, Patla et al. 2003 ) Older adults with disability encounter less challenges in all 8 dimensions of community ambulation, they avoid walking l onger distances, crossing streets with traffic lights, walking in low light and snow/ice conditions, they encounter less stairs, avoid carrying two or more heavy items, avoid making postural transitions, avoid going to unfamiliar places and encounter less situations with high density of traffic ( Shumway Cook, Patla et al. 2003) The fram ework suggested by Patla and Shumway Cook and its emphasis of patient environment interaction illustrated the need for an individual to be able to constantly adapt ones walking to each situation. Musselman et al. ( Musselman and Yang 2007) also were interested in defining the challenges one has to adapt to while walking in the community. They documented walking tasks most commonly en countered in daily living by 50 able bodied adults and 16 ambulatory, community dwelling persons with iSCI. Frequently encountered tasks (more than 10 times per day by at least 50% of the control participants) were walking on smooth and rough surfaces, ope ning and closing doors and carrying objects. Moderately frequent tasks (at least 1 time per day by 75% of the control participants) were: negotiating obstacles, walking on uneven and sloped surfaces, in crowded environments, narrow spaces and steps and stairs. Similar to the findings by Patla and Shumway Cook in the disabled elderly, participants with iSCI in Musslemans study
52 clearly encountered fewer of these challenging tasks than ablebodied participants throughout the day. In ambulatory individuals wit h SCI, encountering so many obstacles in the community can lead to enough fatigue and frustration that can cause them to get discouraged from walking in the community As a result, they prefer to use a wheelchair ( Jannings and Pryor 2012) 2. 5 .2 How Well Can Individuals with iSCI Adapt their Walking Pattern to Environmental Demands? Several studies have investigated the ability of persons with iSCI to adapt to specific challenges ( Barbeau, Fung et al. 2002) such as increased speed ( Pepin, Ladouceur et al. 2003 ; Pepin, Norman et al. 2003) inclines ( Leroux, Fung et al. 1999) stepping over obstacles ( Ladouceur, Barbeau et al. 2003; van Hedel, Wirth et al. 2005) or managing walking challenges in the community ( Lapointe, Lajoie et al. 2001) 2. 5 .2.1 Adaptation to speed Pepin et al. ( Pepin, Norman et al. 2003) measured angular displacements of lower limbs and EMG activity for seven able bodied individuals and seven persons with iSCI walking at several different speeds (0.11m/s)on a treadmill to measure their ability to adapt their walking pattern to the different speeds. SCI subjects could only adapt to a narrow range of speeds and were limited in increasing stride length and stride frequency. Only three out of the seven individuals could reach the maximum speed. Participants with iSCI adapted to increases in speed from 0.10.5m/s by increas ing stride lengths versus cadence compared to healthy controls. However, at higher speeds (0.7 and 1.0m/s), participants with iSCI seem unable to further increase their stride length and adapt to these speeds. The authors suggest that the limited ability to
53 increase stride frequency is likely due to the altered neural drive. Limitations in the ability to decrease double limb stance and therefore to increase cadence is the greatest constraint for these individuals with iSCI when attempting to reach maximal speeds. They prefer increasing stride lengt h over cadence to adapt to increased speed of the treadmill and have a limited range of cadence frequencies available ( Pepin, Ladouceur et al. 2003) 2. 5 .2.2 Adaptation to inclines Leroux et al. ( Leroux, Fung et al. 1999; Leroux, Fung et al. 2002; Leroux, Fung et al. 2006) investigated the ability of individuals with SCI to adapt their walking pattern to inclines of 5, 10 and 15 Individuals with SCI were able to adapt to this challenge with limitations. Only 3/7 participants were able to walk on the 15 incline. Although able bodied control participants adapt their walking pattern at the hip, knee and ankle joints, individuals with SCI only adapt at the hip. Able bodied individuals all used very similar adaptation strategies and their hip to knee angle relationship remained very stable across different incline grades. In contrast, the group of participants with SCI showed notable v ariability both within and between subject s. EMG in controls increased in duration and peak amplitude in all muscles with increasing grade. In individuals with SCI, the only increase in EMG was in the VL. Level of lesion was not related to the ability to adapt ones walking pattern to the inclined surface, but subjects with faster maximal walking speeds on level grounds were the ones who could adapt to steeper grades. 2. 5 .2.3 Adaptation to obstacles In a study by Ladouceur et al. ( Ladouceur, Barbeau et al. 2003) both individuals with SCI (AIS C and D, more than one year post injury) and healthy controls walked
54 over a unilateral 3cm and 5mm obstacle. While controls adapted their hip and knee angles to step over the obstacle, the individuals with SCI subjects increased knee flexion only. However, participants with SCI also compensated by lifting their greater trochanter more than ablebodied participants for obstacles of low height. Similar to the adaptations to speed changes or incline walking, individuals with SCI were mostly able to adapt to the challenge of stepping over an obstacle (only one person could not clear the 3cm obstacle) but used a variety of different strategies. Authors mention that the lower speeds chosen by the persons with SCI compared to controls could be a reason for the differences between the groups ( Ladouceur, Barbeau et al. 2003) Similar results were found when individuals with SCI were asked to learn a novel high precision task. Participants were stepping on a treadmill with restricted vision preventing them from seeing an approaching obstacle. Without vision, they had to learn to lift their foot high enough to just clear the obstacle. Again, individuals with SCI were able to learn the task and improve their performance, but showed different adaptation from ablebodied individuals. They had higher numbers of hitting the obstacle an d their steps were more variable. Of note, the two groups did not differ from each other regarding distance walked in 6min, but individuals with SCI had longer double limb support phases when walking on a treadmill, indicating that their basic locomotor fu nction was not fully recovered ( van Hedel, Wirth et al. 2005) 2. 5 .2.4 Adaptation to demands of community walking The former studies mainly focused on one task and investigated peoples ability to adapt to that task. Lapointe et al. ( Lapointe, Lajoie et al. 2001) chose a battery of tasks that people encounter regularly during community walking. He did not assess how individuals adapted to the different tasks, but only if they were able to adapt/complete
55 the following tasks: achieve the speed they considered necessary to cross an average intersection (>1.06m/s at self selected and/or fastest possible walking speed), step up and down a 18.5cm curb, walk up and down a 4.7 ramp, stopping within 750ms from an auditory cue and walking at least 340m. Nine healthy controls and 9 individuals with iSCI participated in the study. Most individuals with iSCI were able to adapt to these tasks in the sense of being able to co mplete them. However, only 3 people reached the necessary walking speed to cross an average intersection at their fastest possible and only 1 and their self selected walking speeds. Also, some people were walking so slowly when the auditory cue started tha t they did not have to even move to stop. As outlined above, successful community ambulation also requires individuals to be able to pay attention to several physical and environmental aspects simultaneously ( Shumway Cook, Patla et al. 2002) Individuals with SCI have to allocate greater attentional resourc es to the tasks of walking, especially during the single leg stance of the gait cycle ( Lajoie, Barbeau et al. 1999) 2. 5 .2.5 Summary In general, most s tudies found that individuals with iSCI are able to adapt their walking pattern to meet the demands posed to them by changes in the environment. However, these participants were limited in their capacity to adapt, their strategies were variable, and the ad aptation strategies used clearly differed from those seen in healthy controls. 2. 6 The Nervous System Hardwired or Capable of Plasticity? 2. 6 .1 Paradigm Shift The traditional view of the nervous system is that of a hardwired network, incapable of regeneration after an injury ( Ramon y Cajal 1928) However, in the last 30
56 years, this paradigm has been increasingly challenged and a substantial and growing body of evidence indicates that even the adult neural system is highly plastic ( Dobkin, Edgerton et al. 1992; Dobkin 1993 ; Dobkin, Harkema et al. 1995 ; Calancie, Lutton et al. 1996; Byl and Melnick 1997; Edgerton, de Leon et al. 1997; Edgerton, Roy et al. 1997; Basso 2000; Edgerton, Leon et al. 2001; Behrman, Bowden et al. 2006; Cha, Heng et al. 2007; Edgerton, Courtine et al. 2008; Lynskey, Belanger et al. 2008 ) 2. 6 .2 Definition of Neuroplasticity The term neuroplasticity refers to any functional or cellular changes in the nervous system and can occur in response to specific stimuli during both development and throughout life. For example, practicing a motor task will lead to adaptations throughout the motor pathway (motor cortex and associated brain areas, spinal cord, peripheral nervous syst em) responsible for the trained movement ( Nudo, Milliken et al. 1996) while inte grating information for an exam will lead to adaptations in the frontal lobe and associated areas. Besides this activity dependent plasticity, injury of the nervous system itself can also stimulate plastic adaptations ( Dunlop 2008; Blesch and Tuszynski 2009) Neuroplasticity permits the system to accommodate for new demands while maintaining previously acquired skills ( Wolpaw 2007) 2. 6 .3 Mechanisms of Neuroplasticity Neuroplasticity happens through different mechanisms ( Dobkin 1993; Curt, Schwab et al. 2004) such as sy naptic potentiation/alteration of excitability and synaptogenesis ( Cooke and Bliss 2006; Kerchner and N icoll 2008 ) increased efficacy of spared pathways or increased use of alternate descending pathways ( Dobkin 1993) outgrowth of axons or neurogenesis in certain parts of the nervous system ( Chen, Yu et al. 2007; Hagg 2009) or reorganization of representations within the brain ( Dobkin 1993;
57 Adkins, Boychuk et al. 2006) While modulation (for example serotonin influences) only lasts as long as the modulator is present, effects of long term potentiation outlast the stimulus duration ( Cooke and Bliss 2006) This is true even in the early stages, when there are no structural (e.g., protein) changes yet, but receptor s have improved their functions due to phosphorylation ( Cooke and Bliss 200 6 ) Synaptic potentiation describes the process of strengthening synapses that receive sufficient potentials arriving close enough to allow for temporal summation. Temporal summation unblocks a certain class of receptors (through elimination of the chan nel blocking Mg+), allowing calcium to enter the cell, which then in turn triggers a cascade of enzymes that commence short and long term potentiation. This process finally leads to expression of new proteins and enhanced synthesis of postsynaptic receptor s ( Kandel, Schwartz et al. 2000; Cooke and Bliss 2006) Donald Hebb created the famous saying, What fires together wires together, which accurately describes this process in a very simple way ( Hebb 1949 ; Boak ye 2009) Recently with technical advances, it was also discovered that receptors are not always stationary, but are capable of moving rapidly within the cell from one synapse to another ( Ottersen 2010) and even between the cell and extracellular matrix ( Petrini, Lu et al. 2009) This mobility of receptors is thought to play a major role in strengthening synaptic connections ( Newpher and Ehlers 2008) Certain parts of the adult human nervous system have the ability for axonal outgrowth and/or neurogenesis. Neurogenesis (in the adult) pre dominantly happens in the dentate gyrus of the hippocampus ( Hagg 2009 ) and axonal outgrowth in the more
58 permissive (no olygodendrocytes with its inhibitory proteins present) environment of the peripheral nervous system ( Chen, Yu et al. 2007) In case of injury to the lateral corticospinal tract (CST), axons of the ventral CST, the vestibulospinal tract (VST) and the reticulospinal tract (ReST) are most likely to undergo plastic changes and form novel connections ( Eidelberg, Nguyen et al. 1986; Dobkin 1993) Additiona lly, there is plasticity of the propriospinal networks ( Calancie, Lutton et al. 1996) Calancie et al. were able to show that after cervical SCI, connections were formed between the upper and lower ex tremities that were not present in healthy controls. These connections, although seemingly of no functional advantage to the patient, were seen as a sign of neuroplasticity of the spinal cord below the lesion. Last but not least, and in part a consequence of the former mechanisms, cortical representation maps can be altered as response to certain repetitive stimuli ( JohansenBerg, Dawes et al. 2002) Generally, larger representations are associated with enhanced corresponding function. All of these neuroplastic processes seem to be influenced or governed by growth factors, specifically by brain derived neurotrophic factor BDNF ( Lu and Figurov 1997; Kafitz, Rose et al. 1999 ; Adkins, Boychuk et al. 2006; Rojas Vega, Abel et al. 2008) 2. 6 .4 Negative Neuroplasticity In case of overuse and especially if different sensory inputs follow so closely that they cannot be distinguished from one another anymore, the borders between the dif ferent areas in the cortical representation maps can become blurred. This dedifferentiation of the cortical representation is associated with focal dystonia, a disorder characterized by mass movements and the inability for isolated motion ( Byl, Merzenich
59 et al. 1997) This co uld be considered an example of detrimental neuroplasticity. In patients with SCI, unwanted neuroplasticity is also of importance, since evidence suggests that it can lead to negative consequences such as pain or autonomic dysreflexia ( Brown and Weaver 2011) Also, long term potentiation has a counterpart, long term depression (and both can be either ben eficial or detremental to the individual). The function of the nervous system can deteriorate in a certain area if there are not enough stimuli present. Use it or lose it! This phenomenon, has been called learned nonuse and has been discussed for example in SCI ( Dietz 2011) and stroke ( Wolf, Lecraw et al. 1989; Sterr, Freivogel et al. 2002) 2. 6 .5 Principles of Neuroplasticity Certain fundamental principles seem to govern neuroplasticity and the ability of the nervous system to learn new motor tasks. These are important to consider when designing therapies that ai m to induce neuroplastic changes. According to the power law of practice ( Fitts 1964 ) (or exponential law of practice ( Heathcote, Brown et al. 2000) ), increments of improvement s decrease continuously as the performer reaches higher skill levels. If there is no progression in training, a patient will continuously decrease his improvement rate until he reaches a ceiling. It is therefore important that the patient continuously remains challenged during the training sessions ( Fitts 1964 ; Nudo 2003; Schmidt and Lee 2005; Musselman, Fouad et al. 2009) Further, practice needs to be task specific ( Barnett, Ross et al. 1973) meaning that the practice needs to be as similar as possible to the task that is to be acquired. For example, cats that were trained to stand after spinal cord injury learned to stand, but not to walk and vice versa ( De Leon, Hodgson et al. 1998) Similarly, rats that were trained to swim after moderately severe spinal contusion, although improving body position and hindlimb
60 activity, did not improve overground walking funct ion ( Magnuson, Smith et al. 2009) It has been suggested that task specificity can go as far as improving performance in the trained task (reaching) at the expense of a non trained task (walking) in spinal cord injured rats ( Girgis, Merrett et al. 2007 ) On the other hand, in rats that were trained to walk sideways, forward walking improved more than in those animals trained to only walk forward. Authors suggested that adding variability to the training by having the rats walk sideways while keeping the main requirements of the task, reciprocal flexor and extensor movements in the leg, the same as in forward walking made the training more challenging and therefore lead to more improvement ( Shah, Gerasimenko et al. unpublished submitted for publication Feb 2012) Introducing variability to practice, while keeping the main aspect of the trained task specific to the task that one wants to acquire increases retention of skills and transfer of skills from a learned task to a novel task ( Wulf an d Schmidt 1988; Memmert 2006 ) This can also be seen in the fact that inducing errors during training is crucial for motor learning ( Emken, Benitez et al. 2007; Hidler, Nichols et al. 2009) 2. 7 Compensation vs. Recovery The evidence of neuropl asticity, as it was discussed in the last section, opens new doors in rehabilitation ( Dobkin 1993 ; Behrman, Bowden et al. 2006; Backus 2010; Bowden, Behrman et al. 2012) In a paradigm of a rigid, unchangeable nervous system, the only way to regain certain functions and independence is by learning alternative motor strategies and using alternative body structures or assistive devices to compensate for the function that has been lost ( Bowden, Behrman et al. 2012) For example, using a wheelchair as a means of mobility rather than walking, bearing weight through arms and assistive devices instead of through the legs or compensating with
61 hip circumduction for a lack of hip, knee and ankle flexion are all examples of compensation. Under the assumption of existing activity dependent plasticity on the other hand, recovery, meaning the ability to perform movements in the same manners as preinjury ( Levin, Kleim et al. 2009) is possible. The terms recovery and compensation can be used on different levels. A patient that is capable of using and getting around in a wheelchair might be able to return to work, care for a family or attend school. This person has therefore recovered the ability to participate in life situations. By no means has this person recovered any walking functi on, but is instead compensating for his lacking walking function by using a wheelchair. A person that can step on a walker or crutches has recovered the task of walking. However, walking patterns are different from preinjury patterns and the patient compe nsates for reduced function of the lower limbs by loading through his arms. A person walking without assistive devices has recovered the ability to use the appropriate limbs for the task of walking, but might still not have recovered the appropriate mechanics. We tend to think of a person as having recovered walking function if preinjury walking kinematics and kinetics reappear. However, such a person might still be using alternative neural pathways and have compensated for a loss of neural structure by us ing residual pathways or alternate brain areas ( Levin, Kleim et al. 2009) 2. 8 Locomotor Training 2. 8 .1 Basic Neuroscience Research Leading to the Development of Locomotor Training Knowledge gained about central pattern generator networks in the spinal cord as described above on one hand and the increasing evidenc e of plastic properties along
62 the whole neural axis on the other hand have lead to the development of therapies aimed at regaining stepping movements in animals with different degrees of spinal lesions such as the rat ( Grillner and Zangger 1979; Thota, Carlson et al. 2001; Cha, Heng et al. 2007; Beaumont, Kaloustian et al. 2008) and the cat ( Lovely, Gregor et al. 1986; Barbeau and Rossignol 1987 ; Lovely, Gregor et al. 1990; Edgerton, Guzman et al. 1991; Edgerton, Roy et al. 1991; Edgerton, Roy et al. 1992; de Leon, Hodgson et al. 1998; Edgerton, Leon et al. 2001) Adult, spinalized cats were given partial body weight support while being positioned on a moving treadmill. Whereas no connection was present between supraspinal centers and the networks in the spinal cord in these cats, the spinal cord was still capable of receiving sensory input from the periphery. With training, these cats were able to perform well coordinated stepping movements on the treadmill under full weight bearing ( Barbeau and Rossignol 1994) 2. 8 .2 The Principles of Locomotor Training Since the first experiments in spinalized cats, locomotor therapy has come a long way. Evidence from animal research has been translated and made available for humans with neurological deficits ( Harkema, Behrman et al. 2011) On the technical side, body weight support systems have been developed that help unload the participant while he or she is stepping on a treadmill. On the therapeutic side, evidence based training protocol s have been developed that follow the principles of motor learning, for example task specificity, power law of practice, variability of training and challenge ( Shumway Cook and Woollacott 2001; Schmidt and Lee 2005; Harkema, Behrman et al. 2011)
63 Four principles of locomotor training have been defined that should be the basis of any kind of locomotor training, be it on the treadmill, overground or in the community ( Behrman and Harkema 2007; Harkema, Behrman et al. 2011) 2. 8 .2.1 Maximizing weight bearing on the legs (principle 1) During walking while relying on assistive devices, a part of the body weight is loaded through the arms rather than through the legs. However, studies have shown t hat locomotor EMG activity increases with increased loading through the legs ( Visintin and Barbeau 1989; Dietz, Muller et al. 2002) This effect was shown even in individuals with complete SCI, leading to the conclusion that the sensory input of load receptors leads to activation of the spinal networks and therefore also to activation of muscles that are not under voluntary control ( Mller 2006) Similar to assistive devices, a body weight support system also leads to unloading of the limbs. However, it has been shown that unloading through a harness supporting the body weight at the trunk does not lead to a reduction in leg EMG activity as unloading through the arms while using an assistive device does ( Visintin and Barbeau 1989) This in combination with efforts to keep the unloading at the minimal level that still allows for appropriate kinematics, will increase locomotor EMG activity in the legs. 2. 8 .2.2 Optimize sensory cues (principle 2) In the sense of task specificity, pre injury kinematics should be targeted to provide optimal sensory feedback to the nervous system ( Dietz 2009; Harkema, Behrman et al. 2011) Manual assistance as required to achieve those joint trajectories is applied, but kept to the necessary minimum and active patient participation is highly encouraged. Optimal sensory cues also include walking speed. Increased speeds are associated
64 with increases in locomotor EMG activity, especially at spe eds above 2.5km/h ( Lunenburger, Bolliger et al. 2006; Spiess, Jaramillo et al. 2012) 2. 8 .2.3 Optimizing kinematics (principle 3) Appropriate hip extension is a key factor to activate the flexors and transition from stance to swing phase ( Dietz, Muller et al. 20 02) The body weight support system encourages an upright posture, making an adequate hip extension more feasible than when an individual is bent over to unload body weight through an assistive device. 2. 8 .2.4 Maximizing recovery, minimizing compensator y strategies (principle 4) As mentioned above, compensation through assistive devices will lead to a reduction in locomotor EMG patterns, will produce inappropriate sensory cues and will lead to undesirable neuroplasticity. A compensation strategy will be learned rather than pr e injury movement patterns. 2. 8 .3 Clinical Studies Investigating Locomotor Training Locomotor training protocols have been tested in clinical studies and proved to be effective in participant with incomplete injuries with regards to improving gait speed and spatiotemporal gait parameters, EMG patterns, decreasing use of assistive devices, ability to take steps and walking independence ( Wernig and Muller 1992; Wernig, Muller et al. 1995; Barbeau, Fung et al. 1999; Behrman and Harkema 2000 ; Barbeau, Ladouceur et al. 2002 ; Behrman, Flynn et al. 2002; Field Fote and Tepavac 2002; Behrman, Lawless Dixon et al. 2005; Field Fote, Lindley et al. 2005 ; Wernig 2006; Dietz 2008; Nooijen, Ter Hoeve et al. 2009; Manella, Torres et al. 2010; Spiess, Jaramillo e t al. 2012) However, several studies that compared LT to other training methods aiming to improve gait found no differences between the training modalities with the selected outcome measures (gait speed, Functional Independence Measure, WISCI, harmful
65 side effects; for a review refer to Mehrholz, Kugler et al. 2008). LT is desig ned to promote recovery of preinjury walking patterns. It is therefore possible that differences between approaches that are more based on compensatory strategies and LT cant be detected with the outcome measures used. Measures that assess the manner in which the movements are performed, such as spatiotemporal assessments of kinematic analyses might be more effective in differentiating effects between LT and other therapy modalities. While most studies investigating locomotor training focused on regaini ng basic overground stepping patterns, Mussleman et al. ( Musselman, Fouad et al. 2009) aimed to specifically target locomotor adaptability with their training protocol. They trained four individuals with chronic iSCI (> 10 months post injury) that were able to walk 5m overground with or without walking devices or physical assistance (patients scored 612 on the WISCI). The researcher uses an ABA design with the different blocks consisting of either 3 months (35 times 1hr per week) of basic LT (BasicLT) or 3 months of what they refer to as skill training. BasicLT was completed on a treadmill with body weight support and followed the principles of locomotor training, with the exception that hand rails at chest height were used for balance. Skill training included practicing tasks overground that are important for daily walking and for example included walking on different surfaces, in windy conditions, walking and reaching, navigating steps, curbs, slopes, crowded environm ents, walking while completing a secondary task and increasing walking speed (crossing streets at traffic lights) and endurance. Ankle foot orthoses were worn by all participants in either training modality and walking aids (individuals needed different levels of assistance from a 4 wheeled walker and assist
66 from one person to two forearm crutches) were used during the skill training. In both training modalities, the researchers aimed to challenge the participant so at least on near fall occurred per sess ion. Outcome measures included the Borg Balance Scale, the Activity Balance and Confidence Scale, a 6min walk, a 10MTS and the modified Emory Functional Ambulation Profile. Participants improved during both training modalities, but with the exception of the BBS, they improved more with the skilled walking training; this was true regardless of the order of training. Of note, according to the authors, participants differed in their willingness to take risks. The risk was increased during the skill training, because unlike the basic LT, it happened outside of the safe and permissive BWS environment and in real traffic situations. 2. 9 Outcome Measurement 2. 9 .1 Selection of Outcome Measures With the recent paradigm shift in SCI research and in an effort to prepar e for future clinical trials, efforts have been intensified to develop, select and optimize outcome measurement tools and make recommendations for clinical trials including participants with spinal cord injury ( Blight and Tuszynski 2006; Marino 2007; Steeves, Lam mertse et al. 2007; Anderson, Aito et al. 2008; Alexander, Anderson et al. 2009 ) The Internati onal classification of functioning, disability and health (ICF), endorsed by the World Health Organization (WHO) Assembly in 2001 ( WHO 2001) provides an independent reference frame of what is relevant to a given population from a patient, clinician and researcher point of view ( Cieza, Kirchberger et al. 2010) Based on a biopsychosocial model, the ICF paradigm consists of a hierarchically organized list of body structures, body functions, activities and participation items as well as environmental and personal factors that can be used to describe ones level of functioning. Core sets, listing all
67 relevant items for a speci fic health condition, are a valuable tool when choosing or designing outcome measures for research and clinic. The ICF does not define how the individual domains should be measured, but it indicates what should be measured ( Cieza, Kirchberger et al. 2010) The ICF core set for chronic SCI includes 168 categories that were deemed important for people with SCI by patients, caregivers and health professionals ( Cieza, Kirchberger et al. 2010) Of these, the following body structures are directly related to locomotion: cervical spinal cord, thorac ic spinal cord, lumbosacral spinal cord, cauda equina, spinal nerves, structures of thigh, structure of lower leg. At the level of activity and participation, the following items related to locomotion were deemed important: standing, bending, shifting the bodys center of gravity, maintaining a body position, lifting and carrying objects, pulling, pushing, walking short and long distances, walking on different surfaces, walking around obstacles, moving around within the home, within buildings other than hom e, moving around outside and doing homework ( Cieza, Kirchberger et al. 2010 ) The choice of outcome measures to use also depends on the current knowledge in the respective clinical field. In the field of SCI or neurology in the broader sense, a paradigm shift has taken place within the last f ew decades ( Behrman, Bowden et al. 2006) With evermore increasing evidence of neuroplasticity after neurological injury ( Darian Smith 2009) and the knowledge that this plasticity is highly activity dependent ( Edgerton, Tillakaratne et al. 2004; Behrman, Bowden et al. 2006; Wolpaw 2007 ; Dunlop 2008; Lynskey, Belanger et al. 2008; Johnston 2009) rehabilitation efforts are now focusing on providing the individual with the appropriate stimuli to induce this plasticity. Recovery of preinjury movement patterns, rather than compensation through
68 spared muscles or assistive devices, is the main goal of contemporary rehabilitation. Whereas teaching compensatory strategies had effects mainly in the activity and participation domains, provoking neuronal changes results in changes on the body structure and function level. Therefore, outcome measures that ca n differentiate between compensation and recovery are to be preferred when measur ing outcome in clinical trials of SCI ( Bowden, Behrman et al. 2012) The test batteries for the three proposed experiments described in this dissertation were developed based on the considerations presented above. Below is a list of outcome measures used along with a rational why they are specifically suited for use in these experiments. 2. 9 .2 Measures of Body Structure and Body Function 2. 9 .2.1 Neurop hysiological measurements Motor evoked potentials as a test of corticospinal tract integrity. Motor evoked potentials (MEPs), elicited by transcranial magnetic stimulation (TMS) are a noninvasive tool to assess the excitability of the motor cortex, the functional integrity of intracortical neuronal structures and the corticospinal tract (CST), nerve roots, alpha motorneurons and the muscle ( Rothwell 1997 ; Kobayashi and Pascual Leone 2003 ; Butler and Wolf 2007) This technique is well established in the clinic and research. A strong current passing through a coil that is placed over the brain of the participant will lead to a rapidly changing magnetic field that penetrates the skull without attenuation. This magnetic field in turn induces a secondary current of opposite direction to the first one within the brain ( Kobayashi and Pascual Leone 2003 ) A current flow parallel to the long axis of the pyramidal axon depolarizes a cell most readily ( Rossini, Barker et al. 1994) Therefore, for the lower extremities, due to the orientation of the
69 cells, the current likely enters the pyramidal cells dir ectly at the cell body and exits at the first or second node of the axon, where the membrane is then depolarized. The resulting descending volleys are termed D waves ( Rothwell 1997; Kobayashi and Pascual Leone 2003 ) Hereby, fast conducting fibers have the lowest threshold ( Kobayashi and Pascual Leone 2003) For the upper extremities and due to the different orientation of the pyramidal cells, the current likely depolarizes interneurons, which then in turn excite the pyramidal cells through excitatory post synaptic potentials. These waves are termed I waves with slow conducting fibers being depolarized at lowest thresholds ( Rothwell 1997; Kobayashi and Pascual Leone 2003) Both D and I waves lead to a brief muscle contraction, which can be documented with surface EMG electrodes on the muscle belly. Parameters of motor evoked potentials used in this study. For this study, we assessed the motor threshold, central conducti on time, recruitment curve and the ipsilateral silent period. The Resting Motor Threshold is defined as the lowest stimulator intensity that elicits MEP responses in the resting target muscle of at least 50V in at least 50% of a series of consecutive stim ulations ( Kobayashi and Pascual Leone 2003) Other authors ( Rossini, Barker et al. 1994) recommend a 100V threshold. In participants where no responses can be elicited in the resting muscle, we assess active motor threshold. Active contraction of the target muscle of 515% of maximal voluntary contraction facilitates MEP responses in response to TMS such as that amplitudes are increased and latencies decreased ( Rossini, Barker et al. 1994) Active motor threshold is defined as the lowest stimulator intensity that elicits an MEP response of (>100uV) that is clearly
70 distinguishable from the background EMG ( Shirota, Hamada et al. 2012) in five out of ten consecutive trials. Motor threshold is a very reliably measurable parameter that is believed to reflect membrane excitability along the neural axis f rom the pyramidal cells and interneurons projecting to those all the way down to the muscle. It provides insight into the functional integrity of the long tracts and the efficacy of the synapses ( Rossini, Barker et al. 1994; Kobayashi and Pascual Leone 2003) MEP can be more consistently elicited in the lower extremities in subjects that are standing ( Rossini, Barker et al. 1994) However, due to the wealth of parameters that are being assessed in this study, study time lasts up to 3 hrs and for most people, standing (quietly) for that amount of time would not be feasible. Particip ants were therefore seated half reclined with their head against a head rest and their legs supported by leg rests. Recruitment curves display the increase in MEP amplitude as a function of increase in stimulator output. After defining the motor threshold, stimulation intensity is increased in steps of <5% and >5 stimulations per level until a plateau is reached regarding MEP amplitudes ( Cacchio, Cimini et al. 2009 ) or until the participant wished not to increase the intensity anymore. MEP signals are filtered, areas under the curve are then extracted for each stimulation level, normalized to the maximal M wave of the respective muscle ( McKay, Stokic et al. 1997 ) averaged for each level of stimulation and then plotted and fitted with a Boltzman equation ( Capaday 1997) The slope of the Bo ltzman equation serves as a measure of excitability of the neural structures involved. The smaller the slope parameter, the larger the increments in MEP area per unit of TMS
71 intensity ( Carroll, Riek et al. 2001) This value has been show valid and reliable ( Carroll, Riek et al. 2001; Cacchio, Cimini et al. 2009) Ipsilateral silent periods (ISP) are assessed by instructing the participant to sustain a 10% contraction in th e muscle ipsilateral to the stimulated hemisphere (and thus contralateral to the target muscle). This contraction will be inhibited for a period of time after the MEP can be recorded in the target muscle ( Rossini, Barker et al. 1994; Kobayashi and Pascual Leone 2003; Lo and Fook Chong 2004) The length of the SP depends on the intensity of the TMS. In healthy controls, intensities needed to elicit this inhibition are about the same as threshold intensities to elicit an MEP; however, in individuals with a neurological disorder, SPs can be elicited without eliciting a clear M EP ( Rossini, Barker et al. 1994) The SP is most likely mediated by GABA receptors in the brain, but spinal inhibitory mechanisms (Renshaw) are believed to contribute to the first 5060ms ( Kobayashi and Pascual Leone 2003 ; Lo and Fook Chong 2004) ISPs have not been elicited very often in the lower extremities and especially not in the TA muscle. Some authors present results from ISP measurements in the abductor hallucis ( Lo and Fook Chong 2004) The central motor conduction time in this study is calculated as follows ( Rossini, Barker et al. 1994) : CCT = Latency of MEP [ ( M+F + 1 ) ] (2 1) Whereas CCT: Central Conduction Time MEP: Latency of Motor Evoked Potential M: Latency of M wave F: Latency of F wave
72 M waves and F waves are obtained by stimulating the peripheral nerve close to the neuromuscular junction. At supramaximal stimulation intensities, an exciting volley is sent orthograde in the nerve and leads to a brief muscle c ontraction. At the same time, a volley travels retrogradely towards the alphamotorneuron cell body, where it leads to a depolarization which in turn travels orthogradely to the muscle and leads to another, much weaker and variable contraction. Therefore, the time that it takes for the volley to travel from the stimulation site (M wave latency), plus the time it takes from the stimulation site to the alphamotorneuron and back to the muscle (F wave latency) corresponds to twice the length of the peripheral nerve plus an estimated 1ms turnover time at the cell body ( Rossini, Barker et al. 1994; Kobayashi and Pascual Leone 2003) The formula proposed by Rossini et al. provides a very elegant method of estimating the true CCT without an additional synaptic delay at the spinal level and without having to stimulate the efferent roots over the spinal cord. However, there are two drawbacks. First, the turn around time of 1ms is an estimate and might not be accurate in all cases. And second, the speed of propagation in the peripheral nerve might be influenced by the fact that the volley has to travel along a partially refractory axon on its way back to the muscle ( Rossini, Barker et al. 1994) The CCT can provide information about the myelination of the fastest conducting axons, which can be of particular interest in individuals with SCI ( Rossini, Barker et al. 1994; Kobayashi and Pascual Leone 2003) Motor evoked potentials used in studies on SCI. MEPs elicited by TMS have been used frequently in individuals with SCI to assess the functional integrity of the cortico spinal tract ( McKay, Stokic et al. 1997; McKay, Lee et al. 2005; Diehl, Kliesch et al. 2006; Ellaway, Catley et al. 2007; van Hedel, Wirth et al. 2010; Freund, Rothwell et
73 al. 2011) to predict functional outcome after SCI ( Clarke, Modarres Sadeghi et al. 1994; Macdonell and Donnan 1995; Curt and Dietz 1997; Curt, Keck et al. 1998; Kirshblum and O'Connor 19 98; Curt and Dietz 1999) as well as to document change over time ( van Hedel, Murer et al. 2007; Ellaway, Kuppuswamy et al. 2011) and outcome after locomotor training ( Thomas and Gorassini 2005; Benito Penalva, Opisso et al. 2009) MEPs will be used in the third experiment of this dissertation for two purposes: first, to describe the functional integrity of the corticospinal tracts at baseline and second to assess change in functional integrity over time with training. Having intact or partially intact function of the corticospinal tracts at baseline might be associated with a better response to training. Furthermore, the block of AdaptLT in particular is designed to induce corticospinal plasticity. If effective, this might be visibl e using the MEP parameters described above. Vestibulospinal reflex as a test of vestibulospinal tract integrity. The VestibuloSpinal Reflex assesses functional integrity of the vestibulospinal tract by measuring EMG responses in reaction to a galvanic st imulus applied to the labyrinth. This test will be used in experiment t hree of this dissertation. The galvanic stimulation leads to a reduction in firing of the vestibular nerve on the side of the anode producing a triphasic sway response with the initial sway being in the direction of the anode. Based on this, responses that are inverted when the anode and cathode are switched are regarded as being caused by the vestibular stimulation ( Britton, Day et al. 1993; Fitzpatrick, Burke et al. 1994; Liechti, Muller et al. 2008) Participants are standing with their heads turned to either the right or the left. Consequently, the sway occurs in a saggital plane with respect to the trunk and
74 extremities and the corresponding muscle contractions are observ ed in the flexors and extensors of the trunk and lower extremities. Traditionally, the most pronounced responses have been recorded from the soleus muscle opposite of the head turn. The VSR provokes larger EMG responses in muscles that are active in postural control ( Wydenkeller, Liechti et al. 2006; Liechti, Muller et al. 2008) .This can be seen in both participants with iSCI as well as in healthy controls. To activate postural musculature and assess the VSR, participants will stand on a foam pad (Airex Balance Pad, 47cm x 38.5cm x 7cm), with eyes closed ( Horak 1987 ) For project three of this dissertation, the galvanic stimulation is delivered using rubber surface electrodes, attached over the mastoid processes. Stimulations consist of 30 rect angular unipolar currents at 3mA, lasting for 400ms that are presented either positively (anode left) or negatively (anode right) in a pseudorandomized fashion. We record Center of Pressure (CoP) displacement and surface EMG from 7 muscle pairs: Sternocl eido mastoid, Biceps brachii, Erector spinae, Vastus medialis of the M.quadriceps femoris, Tibialis anterior and Soleus. EMG signals are lowpass filtered (40Hz 5th order butterworth filter), rectified and averaged over 15 positive or the 15 negative stimul ations. Preactivity (mean activity over the last 100ms before stimulation) is subtracted and the signal is expressed as a percentage of the average EMG signal 100ms before the stimulus). Several parameters are derived from the averaged traces. For CoP dis placement, latencies of the first three anteroposterior deviations after stimulus onset are defined (tA1, tA2, tA3). The maximal corresponding amplitudes are defined as half the maximal difference between the two averaged traces ( Liechti, Muller et al. 2008) (A1, A2, A3). For EMG traces, short (SL), medium (ML) and off response
75 (end of ML) are def ined. Mean amplitudes of these responses are calculated by averaging the differences between the two traces (SLAmp, MLAmp) ( Liechti, Muller et al. 2008) The Acoustic startle response as a test of reticulospinal tract integrity. Humans as well as other living creatures react to a sudden, intense stimulus with startle, a fast response including musc ular contractions in a craniocaudal sequence ( Koch 1999 ; Ilic, Potter Nerger et al. 2011) The sound stimulus depolarizes the vestibulocochlear nerve and the impulse is relayed through the nuclei cochlearis to the nuclei caudal pontine reticular nucleus ( Davis, Gendelman et al. 1982) From there, connections to the eye muscles are made through the oculomotorius (III), trochlearis (IV) and abducencs (VI) nerves, to the mimical musculature through the facialis (VII), to the sternocleido occipito mastoideus (SCOM) and the superior trapeze muscles through the accessorius (XI). Connections to muscles mostly of the trunk and proximal limbs are made through the Reticulospinal Tract (ReST) ( Davis, Gendelman et al. 1982; Koch 1999; Felten an d Shettz 2010) The acoustic startle response measures the functional integrity of the Reticulospinal Tract (ReST) ( Szabo 1965; Rossignol 1975; Davis, Gendelman et al. 1982; Davis 1984; Wilkins, Hallett et al. 1986 ; Yeomans and Frankland 1995; Koch 1999; Kofler, Muller et al. 2006; Kumru, Vidal et al. 2008; Valls Sole 2012) This tool will be used in experiments two and three of this dissertation with slightly different configurations. The ASR has been described by several investigators in the animal model as well as in healthy human controls and different populations with neurologic injury ( Szabo 1965; Rossignol 1975; Davis, Gendelman et al. 1982; Davis 1984; Wilkins, Hallett et al. 1986 ; Yeomans and Frankland 1995; Kofler, Muller et al.
76 2006; Kumru, Vidal et al. 2008; Valls Sole 2012) Validity of the measurement has been established on the animal model by comparing data to results from neuroanatomical investigations, lesion studies and electrical stimulation of the reticular formation( Davis, Gendelman et al. 1982 ; Davis 1984) In humans, validity of t he ASR as a measure of the functional integrity of the ReST is assumed due to the similar response patterns in humans and animals ( Brown, Rothwell et al. 1991) ASRs are more readily elicited in axial and proximal muscles ( Rossignol 1975; Davis 1984; Brown, Day et al. 1991; Brown, Rothwell et al. 1991; Grosse and Brown 2003; Kofler, Muller et al. 2006; Kumru, Vidal et al. 2008) which is in line with the finding that the ReST fibers terminate on alpha motorneurons and interneurons that provide inner vations to those muscle groups. Also, increasing latencies in muscles that are further away from the reticular formation suggest that the responses actually are relayed through the ReST. One of the challenges with the ASR described in the literature is the rapid habituation after 26 stimuli ( Davis and Wagner 1969; Brown, Rothwell et al. 1991; Valsamis and Schmid 2011) This phenomenon prevents researchers from being able to do repeated measurements to increase reliability. Acoustic startle in healthy persons. In healthy controls, responses have mainly been recorded from neck, trunk and arm muscles. There is debate as to whether more flexor or extensor responses are elicited and responses may be dependent upon the position of the participant with increased responses in muscles that are actively involved in maintaining posture ( Brown, Day et al. 1991; Brown, Rothwell et al. 1991; Kumru, Vidal et al. 2008; Baker 2011)
77 Acoustic startle in individuals with SCI. Only very limited data are available on ASR mea surement in individuals with SCI. Kumru et al. ( Kumru, Vidal et al. 2008) elicited acoustic startle responses in participants with complete and incomplete SCI and concluded that the rate of elicited responses was increased in all participants with SCI when compared to healthy controls. They suggested this indicates that neuroplastic changes have taken place and the ReST in this population might conduct more signals than in healthy controls. Jankelowitz et al. ( Jankelowitz and Colebatch 2004) confirmed this finding in a small sample of individuals with cervical SCI (they also found similar results in a sample of participants with stroke). Exc itability of the spinal cord. H reflexes are an accepted measure of spinal cord excitability ( Zehr 2002; Misiaszek 2003; Muller and Dietz 2006) Measuring H reflex parameters along with measures of cortical excitability can allow differentiation between plastic effects that take place at the level of the spinal cord and those taking place at the level of the brain. H reflexes of the Soleus muscle are elicited by stimulating the posterior tibial nerve in the popliteal fossa. Different parameters are used to quantify H reflex responses. Input/output curves are used to quantify the spinal cord excitability. There are several accepted methods of describing input/output (or recruitment) curves ( Klimstra and Zehr 2008; Phadke, Robertson et al. 2012) One includes expressing both the M wave and the H reflex in terms of stimulator output and divides the slope of the linear portion of the steepest part of the H reflex by the slope of the steepest part of the M wave to create a H/M ratio ( Phadke, Robertson et al. 2012 ) This method allows appreciating changes that take place at the muscular level or the central nervous system level
78 separately. In connection with TMS, changes can be assigned to either muscular, spinal cord or cortical/ corticospinal level. Klimstra et al. suggested normalizing H reflex amplitudes to maximal M waves and expressing those with respect to the stimulation intensity that has been normalized to the intensity that produces and M wave of 50% of the maximal M wave. According to Klimstra, this is the preferred method if the stimulation intensity for each stimulation is known ( Klimstra and Zehr 2008) In the last method for analyzing H reflex recruitment curves, H Responses of each st imulation are expressed as percentage of the corresponding M wave and those ratios are plotted against the M waves normalized to the maximal M wave, creating an H/M wave ratio. This method can be applied even if the exact stimulation intensity for each H r eflex/M wave pair is not known. Slopes of the linear portion of the ascending limb of the sigmoid fit to the recruitment curves are calculated for both methods by Klimstra ( Klimstra and Zehr 2008) H reflexes in healthy persons. H reflex latencies have a very high test retest reliability (r>0.92). Amplitudes are less reliable(r=0.29 while lying prone) ( Ali and Sabbahi 2001) H reflexes responses are dependent on body position and, if assessed during walking, on the position in the gait cycle( Crenna and Frigo 1987; Al Jawayed, Sabbahi et al. 1999; Andersen and Sinkjaer 1999; Ali and Sabbahi 2000; Ali and Sabbahi 2001) H reflexes in persons with SCI. In persons with SCI, H reflex to M wave ratios at thr ee points during the gait cycle were larger than in ablebodied control participants ( Trimble, Behrman et al. 2001) This increase in H/M ratios has also been shown in H reflexes evoked in the upper extremity ( Little and Halar 1985) However, a reduction of
79 the H reflex modulation has been shown during gait ( Knikou, Angeli et al. 2009; Phadke, Thompson et al. 2010) ,specifically a decreased depression of the response during the swing phase ( Knikou, Angeli et al. 2009) It has been hypothesized that this increased activity is due to reactive plasticity in response to reduced supraspinal input at the spinal level. An increase of H reflex modulation may therefore indicate functional recovery ( Phadke, Thompson et al. 2010) With a single locomotor training session, H/M ratio s were depressed when compared to pre training ( Trimble, Behrman et al. 2001) 2. 9 .2.2 International S tandards for N eurological C lassification of S pinal C ord I njury The International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) (previously known as AISA Standards) are the most widely used tool to measure severity of the neurological impairment as a consequence of SCI. This test is used in all three experiments of this dissertation. The ISNCSCI have been developed as a bedside test and can be conducted noninvasively, with minimal equipment in almost all stages of spinal cord injury (allowing for continuous monitoring over time). It is based on various preexisting measures and was adopted by the International Spinal Cord Society (ISCoS) in 1992 ( Ditunno 1994 ; Furlan, Fehlings et al. 2008; Ditunno 2010) The assessment consists of a sensory and a motor exam. The sensory exam tests light touch as well as pin prick sensation in 28 dermatomes on both sides of the body. Sensation is rated as either absent, impaired or normal. The motor exam consists of rating 10 key muscles on each side on a 5 point scale from no visible or palpable contraction to normal active movement over the full range of motion against full resistance. Additi onally, anal pressure (AP) and voluntary contraction of the anal sphincter muscle (VAC) are rated as present or absent by means of a digital exam.
80 From the collected data, sensory and motor summary scores are defined for each modality and body side. Sensor y and motor levels are defined as the most caudal level with normal function in the respective modality. The neurological level of injury (NLI) is defined by the single most caudal level with normal function and is equal to the more rostral of the two mod ality levels. Completeness of the injury is defined as absence of any function at the last sacral level (S4/5, AS, VAC). This definition, termed sacral sparing definition was chosen for its stability and ability to predict motor recovery ( Waters, Adkins et al. 1991) Finally, the above information is integrated into a sing le measure of severity of injury, the American Spinal Injury Impairment Scale (AIS). The corresponding algorithm is presented in Figure 21 ( ASIA 2011 ) 2. 9 .3 Measures of Activity Several measures of activity have been chosen, both capacity measures (measures of what an individual is capable of doing in a standardized environment) as well as performance measures (measures of what an individual is capable of doing in his or her usual environment, aka real life situations) ( WHO 2001; WHO 2002) All these tests will be used in experiment two of this dissertation. 2. 9 .3.1 Activity capacity Capacity measures include the 10 meter walking test (10MWT) ( Wolf, Catlin et al. 1999; van Hedel, Dietz et al. 2007; Jackson, Carnel et al. 2008) the Walking Index for Spinal Cord Injury II (WISCI II) ( Ditunno, Ditunno et al. 2000; Dittuno and Dittuno 2001) the short version of the Balance Evaluation Systems Test (Mini BESTest) ( Franchignoni, Horak et al. 2010) the Spinal Cord Injury Functional Ambulation Profile (S CI FAP) ( Musselman, Brunton et al. 2011) as well as the stair items from the Dynamic Gait Index (DGI) ( Shumway Cook and Woollacott 2001) the Neuromuscular Recovery
81 Scale ( Behrman, Ardolino et al. 2012) as well as the stair item from the Timed Movement Battery (TMB) ( Creel, Light et al. 2001) The 10MWT is one of the most commonly used tests for gait speed ( Jackson, Carnel et al. 2008) While some authors recommend a distance of 10meters being timed and 2m added each at the beginning and the end to account for acceleration and deceleration ( van Hedel, Dietz et al. 2007) others recommend to time the middle 6meters of a 10meter walk. ( Wolf, Catlin et al. 1999) We chose the latter based on t he similarity with item 13 (step over obstacles) from the MINI BESTest ( Franchignoni, Horak et al. 2010) The WISCI is a SCI specific, ordinal walking scale created by a Delphi procedure. It lists different levels of independence while walking over 10 meters ( Ditunno, Ditunno et al. 2000; Dittuno and Dittuno 2001 ; Morganti, Scivoletto et al. 2005 ) The WISCI is considered a valid and reliable measure for walking capacity, especially in lower functioning individuals. However, it has been criticized for its floor and ceiling effects and the corresponding missing responsiveness ( van Hedel, Wirz et al. 2006; Jackson, Carnel et al. 2008) Suggestions to overcome the ceiling problems are to use the WISCI as a timed test in the future ( Jackson, Carnel et al. 2008) The MiniBESTest encompasses 14 items analyzing different postural control systems that may contribute to dynamic balance ( Franchignoni, Horak et al. 2010) Each item is scored on a three point scale as normal, mildly or severely impaired to a total maximal score of 28. This test has been validated in a convenience sample of 115 patients with balance disorders of varied origin ( Franchignoni, Horak et al. 2010)
82 The SCI FAP includes 7 timed walking tasks encountered regularly in daily life such as negotiating obstacles or carrying objects while walking. ( Musselman, Brunton et al. 2011) Unlike the MiniBESTest, this test has been specifically develope d for the SCI population. Spatiotemporal data. Spatiotemporal parameters include walking speed (distance covered per time), cadence (steps taken per time), step length (distance from heel strike to heel strike of the next footfall), stride length (distance from heel strike to consecutive heelstrike of the same foot), step width (lateral distance between two consecutive foot falls), step time (time to complete the step), single limb stance (percentage of the time of a gait cycle during which one limb is in swing phase), double limb stance (percentage of the time of a gait cycle during which both feet are touching the ground) and swing time (percentage of the time the leg was in swing. Spatiotemporal and kinematic parameters in individuals with SCI differ from those in ablebodied control persons to different extents. Individuals generally walk at slower speeds, both overground and on a treadmill ( Pepin, Ladouceur et al. 2003; Pepin, Norman et al. 2003; van Hedel, Dietz et al. 2007) although the group o f individuals with SCI presents with a wide variety from incapacity to walk to near normal walking speeds ( Barbeau, Fung et al. 2002) Their cadence and step and stride length are reduced and there is asymmetry between the left and right leg regarding these parameters ( Barbeau, Fung et al. 2002; Pepin, Norman et al. 2003; Nooijen, Ter Hoeve et al. 2009) Step cycles are longer than in controls and more hip excursion and knee flexion during swing and at initial foot contact ( Barbeau, Fung et al. 2002) While foot motion can recover near normal trajectories, at least in a laboratory setting, it is
83 achieved by relying more on proximal and axial muscles when compared to healthy controls ( Ivanenko, Poppele et al. 2009) 2. 9 .3.2 Activity performance Performance measures include the Activity Balance Confidence Scale (ABC) ( Powell and Myers 1995 ; Myers, Fletcher et al. 1998; Lajoie and Gallagher 2004) an inventory of walking tasks most commonly encountered by able bodied adults ( Musselman and Yang 2007) a retrospective falls history assessment ( Arnold and Faulkner 2007) and a standardized diary to assess falls during and after the training period ( Wirz, Muller et al. 2010) assessing the number of steps taken measured by a Step Activity Monitor (SAM) and a Computer Adaptive ICF Activity Measure. ( Velozo, Kielhofner et al. 1999) All activity performance assessment tools will be used in experiment two of this dissertation. The ABC is a measure of perceived self efficacy in everyday tasks requiring balance. ( Powell and Myers 1995) The ABC consists of 16 items of varied demands on ones balance. For each of the questions, the individuals is asked how confi dent they feel that they can accomplish the task without falling on a scale from 0% (will certainly fall) to 100% (will definitely not fall). The ABC has been proven valid, reliable and responsive in community dwelling elderly and in different neurological populations ( Powell and Myers 1995,Botner, 2005 #2456; Lajoie and Gallagher 2004; Steffen and Seney 2008; Talley, Wyman et al. 2008; Huang and Wang 2009) No validity testing has been conducted in a population with iSCI. However, it has been used as an outcome tool in studies investigating iSCI ( Musselman, Fouad et al. 2009; Lam, Pauhl et al. 2011) Mussleman, who used the ABC as an outcome measurement tool in a study comparing BWSTT to overground skill training discovered more increase in the ABC
84 with skill training than with BWSTT ( Musselman, Fouad et al. 2009) Lam et al. found an increase of 22% in the ABC score in response to 36 sessions of resistance augmented robotic BWSTT in their single case study ( Lam, Pauhl et al. 2011) While the Falls Efficacy scale has also been used in iSCI ( Behrman and Harkema 2000; Wirz, Muller et al. 2010) it has been shown to be less reliable and less responsive in a direct comparison ( Powell and Myers 1995) We therefore chose to use the ABC as a subjective measure of self efficacy in everyday balance challenges. The inventory of walking tasks most commonly encountered by able bodied adults has been validated in healthy controls and people with SCI ( Musselman and Yang 2007) Participants were convenience samples from community groups in Edmonton and Ottawa, Canada. Participants with iSCI were recruited through the Edmonton branch of the Canadian Paraplegic Association. Proportions of able bodied adults encountering each of the tasks 13, 4 9 or more than 10 times per day are presented and data from persons with SCI are compared to these normative values. Mussleman defined a frequent walking task as a task that at least 50% of able bodied participants perform more than 10 times per day. These tasks included opening and closing of doors, walking on smooth surfaces, walking on rough surfaces, and carrying objects while walking. They further defined tasks that are encountered less frequently (at least once per day), but by at least 75% of able bodies participants as moderately frequent task ( Musselman and Yang 2007) Falls are a significant problem in individuals with iSCI. They occur frequently and can have significant consequences ( Brotherton, Krause et al. 2007) Fall related injuries limit people with iSCI in their ability to get out in the community, the ability to care for
85 themselves or even the ability to spend time out of bed ( Brotherton, Krause et al. 2007) Brotherton ( Brotherton, Krause et al. 2007) developed a falls questionnaire specifically for individuals with iSCI through consolidation of existing balance measures and the addition of fall related items. However, they do not provide more information on specific questions asked ( Arnold and Faulkner 2007) In addition to assessing the falls history retrospectively, we ask our participants to keep a falls diary for up to 3 months after study inclusion. This specifically designed calendar has been developed and used by Wirz et al. ( Wirz, Muller et al. 2010 ) and has been provided to us with the permission to use it by Markus Wirz. Partic ipants are provided calendar sheets with instructions to note falls as every event in which they involuntarily lose balance and end up on a lower level (chair, bed or floor). Additionally, injuries are classified as injury to the skin, injury to joins or fractures and treatment by a doctor of hospitalization are noted if they were necessary. The SAM (Step watchTM; Orthocare Innovations LLC, Oklahoma City, OK) measures the number of steps taken over a certain period ( Coleman, Smith et al. 1999 ; Bowden and Behrman 2007) The participant wears the small device (5x7x1cm) on his stronger leg. The device is calibrated to suit the step frequency and acceleration of each individual. It has been shown t o be valid and reliable in different populations ( Coleman, Smith et al. 1999 ) including iSCI ( Bowden and Behrman 2007) In a comparison to observed counts during a 6min walk, Bowden et al. found 97% accuracy and excellent test retest reliability (ICC=0.99 for two 6min walk tests at least 1 week apart). The minimally detectable differenc e as defined by the 95% confidence interval was 177203.7 steps. Proper individualized calibration is key due to the fact that individuals with
86 iSCI may have impaired function on both legs. We calibrated the SAM according to the recommendations in Bowden e t al. ( Bowden and Behrman 2007) Participants were asked to wear the device over two 24hr periods during the week and two 24hr periods on a weekend. In addition, participants were asked to keep an activity diary over those four days, indicating approximately how many minutes per hour they spent wal king, standing, sitting or lying down. This procedure is done to confirm the results obtained by the SAM. T he ICF Activity Measure ( Velozo, Wang et al. 2008) is a self report computer adaptive test that is freely available on the internet (www.icfmeasure.phhp.ufl.edu). It consists of a list of activities related to the constructs maintaining and changing positions, carrying, lif ting, pulling or pushing, using hands, walking or climbing, or self care activities. Rasch item response theory models were used to investigate item characteristics and arrange the individual items in a rank order based on their difficulty. The diff erent constructs were calibrated on various populations including participants with back pain, UE injury, LE injury and spinal cord injury. During the computer adaptive testing, participants self report on the level of difficulty they experience while perf orming a presented task. Hereby, the difficulty of the next item presented always depends on how difficult the previous one was rated by the participant. If an individual expresses considerable difficulties, the next task presented will be less challenging if he or she expresses little to no difficulty, the next task presented will be more challenging. This process allows presentation of the most relevant questions to the individual and therefore reduction of testing time. Following answering of each quest ion, the standard error is calculated and the program stops
87 once the criterion is reached, meaning that any further questions will not add to the classification of this individuals abilities. A participants ability is presented on a 100 point scale and t he scale values are connected to the item bank of tasks (value and range are provided) relative to task difficulty ( Velozo, Kielhofner et al. 1999; Velozo, Wang et al. 2008 ) The participants abilities have improved significantly if results from subsequent evaluations are above two standard errors of the previous evaluation results (personal communication Dr. Velozo). 2. 9 .4 Measures of Performance According to recommendations for clinical trials in SCI, measures of participation and quali ty of life measures should only be included as secondary measures and only in larger phase III trials ( Steeves, Lammertse et al. 2007) We therefore refrained from including such a measure in our test battery. 2. 10 Summary and Conclusion of the Literature Review A spinal cord injury leads to highly variable degrees of functional deficits. Amongst other complaints, individuals experience difficulties using appropriate walking kinematics and at normal walking speeds. These wal king difficulties are especially pronounced during higher skilled walking tasks and interfere with successful community ambulation. Functional deficits, amongst other factors, are dependent on the amount of spared and functional descending tracts. Individuals with BrownSquard Syndrome have sparing of one half of the spinal cord and therefore might show more potential for recovery of preinjury function when compared to others. However, evidence regarding this subject is inconclusive, which might in part be due to the different definitions that were used to classify individuals as having BSS.
88 Given the importance of descending tracts in human gait, it is important to be able to reliably assess their functional integrity. The ASR, assessing functional integr ity of the ReST has received little attention, especially in a population with SCI, but might be an important measure to document neural plasticity. Therefore, testing parameters need to be better defined to be able to record ASR responses in a consistent and reliable manner. Finally, the question remains if behaviorally based therapies are able to induce plasticity in these descending tracts to improve adaptability and performance in higher skilled walking tasks. A further development of the established l ocomotor training, using the same permissive environment and based on the same principles, but including more challenging tasks might lead to increased plasticity in the three major descending tracts. This, in turn might lead to improved adaptability and t herefore to more successful community ambulation. The following chapters describe three experiments that will address the above mentioned issues. Experiment one will assess the effect of using different definitions of BSS on the functional characteristics in this group with descending sparing. Experiment two will address the assessment of descending tract function in one of the three major descending tracts involved in locomotion. Finally, experiment three will address the potential induction of neuroplasti city in the descending tracts and improved performance in adaptability walking tasks in response to a newly developed variation of locomotor training.
89 2.1 1 Tables and Figures Table 21. Symptoms of BSS Ipsilesional deficits Contralesional deficits Pyramidal tracts Paresis Lateral spinothalamic tract Pain and temperature sense missing or highly reduced (dissociated sensory dysfunction) Anterior spinothalamic tract Touch slightly reduced (fibers in the ipsilateral dorsal tracts assure remaining function) Vasomotoric fibers of the lateral columns Initially overly warm and red skin, possibly less sweat secretion Overuse of the contralateral spinothalamic tract with hypersensitization Initial hypersensitivity to touch Dorsal Columns Position and vibration sense compromised Anterior column and anterior roots Segmental atrophy and flaccid paresis Entering dorsal root Segmental anesthesia and analgesia Translated into English and adapted from Mumenthaler, M. and H. Mattle (2002). Grundkurs Neurologie, Illustriertes Basiswissen fr das Studium. Stuttgart, Thieme.
90 Table 22 M inimally necessary walking distances and speeds as well as curb heights reported by three different studies Lerner Frankiel 1986 Mean(Range) Robinett 1988 (data from cities) Median (Range) Andrews 2010 Mean (Range) Distances in m Post office 64 (33 102) 60.5 (36.0 84.0) 52.0 (25.1 98.4) Bank 98 (50 179) 63.0 (30.0 217.0) 57.1 (25.0 102.0) Medical office 98 (43 159) 47.5 (26.5 122.0) 65.8 (30.5 149.4) Pharmacy/ Drugstore 332 (148 597) 139.5 (103.5 213.0) 206.3 (153.9 255.1) Department store 286 (174 381) 327.0 (169.0 505.0) 345.9 (241.3 512.0) Grocery/ supermarket 267 (233 338) 342.0 (201.0 480.0) 380.6 (162.1 526.0) Hardware 565.5 (499.2 626.7) Superstore 606.6 (472.0 792.0) Club warehouse 676.8 (506.3 922.0) Crosswalk commercial 27 (22 34) 16.5 (8.5 27.5) Crosswalk residential 13 (8 16) 10.0 (9.0 12.0) Crosswalk 13.29 (3.84) Speed in m/s Commercial Crosswalk Residential Crosswalk Crosswalk 1.32 1.06 0.49 (0.21 0.89) Height in cm Curb commercial 20 (15 22) Curb residential 18 (14 22) Curb 18.5 (14.0 21.5) Lerner Frankiel et all collected data in Los Angeles county; Robinett et al. collected data in 7 communities of three different sizes in New Mexico and Texas (rural towns, small towns and cities; data from cities with populations ab ove 95000 persons are presented here); Andrews et al. collected data in 15 towns and cities in multiple regions of central and western North Carolina. Both Andrews and Robinett indicate that they have based their measuring methods on the work by Lerner Frankiel (Andrews, A. W., S. A. Chinworth, et al. (2010). "Update on distance and velocity requirements for community ambulation." J Geriatr Phys Ther 33(3): 128134: Lerner Frankiel, M. B., S. Varcas, et al. (1986). "Functional community ambulation: what ar e your criteria?" Clinical Management in Physical Therapy 6: 1215; Robinett, C. S. and M. A. Vondran (1988). "Functional ambulation velocity and distance requirements in rural and urban communities. A clinical report." Phys Ther 68(9): 1371 1373.)
91 Figure 21. Algorithm used to compute the AIS ( adapted from International Standards for Neur ological Classification of Spinal Cord Injury. Atlanta, GA, American Spinal Injury Association. revised 2011) ; S4/5: 4th and 5th sacral segment; VAC: Voluntary Anal Contraction; AS: Anal Sensation, AIS: ASIA Impairment Scale.
92 CHAPTER 3 PROJECT I DEF INING BROWN SEQUARD SYNDROME : ASSYMETRICAL SPARING OF DESCENDING PATHWAYS 3.1 Background The International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) classify incomplete spinal cord injuries (SCI) into 5 clinical syndromes, depending on the pattern of functional loss ( ASIA 2011 ) One of those five, Brown Squard Syndrome (BSS) occurs in about 1 4% of all traumatic SCI or in 17% of all clinical syndromes ( McKinley, Santos et al. 2007) and is due to a severance of one half of the spinal cord. Due to the specific individual decussation levels of the spinal tracts, this lesion leads to a characteristic asymmetric al pattern of neurologic deficits below the level of lesion ( Brown Squard 1855) This lesion is def ined by the ISNCSCI as an ipsilateral loss of proprioception, vibration and motor control at and below the level of the lesion, sensory loss of all modalities at the level of the lesion, and contralateral loss of pain and temperature sensation ( ASIA 2011 ) The term BrownSquard Plus Syndrome (BSPS) is used for those cases with only partial losses of the respective functions ( Roth, Park et al. 1991,Koehler, 1986 #5905) The BS/BSP syndrome is of particular interest due to its resemblance in symptoms to the hemisection model often used in animal research ( Filli, Zorner et al. 2011) Individuals with BSS have been said to have a favorable prognosis when compared to other individuals with iSCI ( Bosch, Stauffer et al. 1971; Little and Halar 1985; Koehler and Endtz 1986; Nathan 1994; McKinley, Santos et al. 2007; Moreh, Meiner et al. 2009) However, studies reporting favorable outcomes for individuals with BSS or BSPS are often case descriptions or case series ( Little and Halar 1985; Koehler and Endtz 1986 ; Moreh, Meiner et al. 2009; Wu, Ho et al. 2010) Outcome measures in
93 the ol der studies were often not clearly defined and could be as crude as "the outcome of surgical treatment was good, "showing good functional recovery", or "recovered/disabled/dead" ( Koehler and Endtz 1986) Three more recent studies with larger samples and clearly defined outcome measures, assessment time points and study populations examined this issue more closely. McKinley et al. ( McKinley, Santos et al. 2007) compared the syndromes as defined by the ISNCSCI at admission and discharge from initial rehabilitation. The authors concluded that BSS patients have a slightly better outcome, as well as slightly more improvements over their inpatient stay in most of the Functional Independence measure (FIM) ( Keith, Granger et al. 1987; Hall, Cohen et al. 1999) subscales including FIM motor and FIM mobility. O f note, BSS patients were the youngest of all t he five groups, had the highest percentage of AIS C scores, and the second highest percentage (after Central Cord Syndrome) of AIS D scores. They suspected that these parameters could be confounding factors when interpreting the data. Pouw et al. ( ( Pouw, van de Meent et al. 2010) compared the different parameters from the ISNCSCI and the Spinal Cord Injury Independence Measure (SCIM II) of BSPS patients to those of non BSPS pa tients at 15 days and 612 months after injury. When controlling for severity of injury (AIS A impairment scale AIS ), the only differences between BSPS and nonBSPS individuals were found in bladder management. Based on their results, the authors questioned the importance of differentiating between the different clinical syndromes in iSCI. Wirz et al. ( Wirz, Zorner et al. 2009) compared neurophysiological ( a mplitudes of somatosensory evoked p otentials SSEPs), neurological (ISNCSCI Motor scores) and functional (10m walking test, Walking Index for Spinal Cord Injury WISCI II
94 and SCIM II) outcomes of individuals with BSS to those with Central Cord Syndrome (CCS) at 1 month and 6 months post injury. No differences in outcome between the two groups were found. Authors therefore suggest these two anatomically dissimilar types of SCI equally e fficient mechanisms of functional recovery. One reason for the disagreement in study outcomes over time could lie within the different definitions used to classify individuals as BSS or BSPS. For example, McKinley et al. ( McKinley, Santos et al. 2007) defined BSS according to the 2000 version of the ISNCSCI ( ASIA 2002 ) as a lesion that produces relatively greater ipsilateral proprioceptive and motor loss with contralateral loss of sensitivity to pain and temperature ( McKinley, Santos et al. 2007) According to personal communication (E ma il request) with the main author, an average of 2 grades of difference in motor scores with contralateral deficits in pain/temperature sensation was used as a threshold. Wirz et al. ( Wirz, Zorner et al. 2009) regarded differences of at least 19 points between the ISNCSCI motor scores of the right and left key muscles as BSS syndrome. Their definition did not include sensory scores. Pouw et al. included patients with BSPS into their analysis if they had any asymmetry (difference>0 in muscle and pin prick scores) with greater losses of the two modalities being on opposite sides. Sensation to light touch was not used to classify participants ( ( Pouw, van de Meent et al. 2010) and personal communication). Both Wirz et al. ( Wirz, Zorner et al. 2009) and Pouw et al. ( Pouw, van de Meent et al. 2010) included only participants with cervical injuries. Using absolute thresholds (e. g. a 1 point, 2 point or 19 point difference between right and left sides as explained above) to decide if individuals have BSS or BSPS seems to int roduce possible bias towards patients with higher levels of injury reaching
95 that threshold more readily. Hayes et al. ( Hayes, Hsieh et al. 2000) developed a series of different decision algorithms with relat ive threshold values (% difference between right and left or upper and lower extremity scores) based on the ISNCSCI to classify the 5 syndromes described in the ISNCSCI. Due to the absence of an external criterion, no validation of the algorithms in the cl assical sense was possible. However, the authors favor one algorithm and cutoff value ( 10%, presented in Figure 31), due to its clinical reasonableness. It is likely that the definition used for BSS or BSPS plays an important role when describing clini cal or research outcomes. The effect of using different definitions can be best determined if those definitions are applied to a single dataset. Therefore, the goal of this project is to first apply the three definitions by Wirz, Pouw and Hayes to a large dataset of patients with iSCI and classify individuals as either having BSS/BSPS or not according to each definition. Then, the functional abilit ies will be compared between the groups Further, in a subgroup of individuals for whom data was available for both baseline and assessment after 3560 training sessions improvement in their functional abilit ies in response to locomotor training will be quantified for and contrasted between, each of the groups defined as having BSS/BSPS. We hypothesize that the d efinition used will have an effect on the percentage of people classified as having BSS/BSPS, as well as on their functional abilities. 3.2 Methods 3.2.1 Data A large dataset was obtained from the Christopher and Dana Reeve Foundation NeuroRecovery Network with IRB approval The NeuroRecovery network (NRN) provides locomotor training to individuals with incomplete SCI in out patient settings and
96 assesses participants every four weeks via standardized assessments Assessments include balance and gait outcomes as indicated below For a detailed description of the NRN, please see Harkema, Schmidt Read et al. 2011. 3.2.2 Outcome M easures Outcome measures used in the present analysis include: the IS NCSCI ( ASIA 2011) the Berg Balance Scale (BBS) ( Wirz, Muller et al. 2010) the 6min (6minWT) and 10meter (10mWT) walking tests ( van Hedel, Wirz et al. 2005; van Hedel, Dietz et al. 2007; Jackson, Carnel et al. 2008) and the Neuromuscular Recovery Scale (NMRS) ( Harkema, Behrman et al. 2011 ; Behrman, Ardolino et al. 2012) If an assistive walking device was used during the first data collection for a specific test, then this device was used during later data collections. The data have been controlled for correct computation of ISNCSCI parameters by recalculating these values with a computer algorithm ( Schuld, Wiese et al. 2 011 ) 3.2.3 Definitions of BSS Algorithms in SPSS (IBM SPSS Statistics, Version 20) were developed to apply the definitions described by Wirz, Pouw and Hayes. For the Wirz definition, individuals were classified as having BSS/BSPS if the absolute diff erence between right and left motor scores (MS, according to the ISNCSCI) reached at least 19 ( Wirz, Zorner et al. 2009) For the Pouw definition, individuals were classified as having BSS/BSPS if there was any difference between the left and right MS as well as a difference in the pinprick scores (PP) with the larger PP deficit occurring co ntralaterally to the larger motor deficit ( Pouw, van de Meent et al. 2010 ) Finally, the Hayes definition was used and had several criteria : 1) the level of injury was above T12, 2) a motor loss of at least 10% (average motor score below the level of lesion <4.5), 3) the motor loss was less
97 than 10% greater in the upper limbs than the lower limbs, 4) at least 10% difference between left and right MS and l ight touch scor es, and 5) a 10% difference in the opposite direction in the pinprick scores ( Hayes, Hsieh et al. 2000) (figure 31). This convention potentially creates seven groups: BSS/BSPS according to Wirz, Pouw and Hayes and individuals that are defined as having BSS/BSPS by any combination of these definitions The groupings are compared regarding their functional outcome. 3.2.4 Statistical Analysis Descriptive statistics i.e. m eans and standard deviations of continuous variables and frequencies and percentages of noncontinuous variables will be derived Due to the considerable differences in group size and nonnormal distributions of most parameters, nonparametric statistics w ere chosen to compare parameters between groups that were defined as having BSS/BSPS according to the different definitions. Statistical significance was set at an alpha level of p= 0.05. Statistical testing was conducted with IBM SPSS 20. 3.3 Results 3.3. 1 Available D ata The NRN dataset contains ISNCSCI data from 999 different assessments from 408 different participants In 735 ISNCSCI datasets from 384 different participants, sufficient information was available to define individuals as having BSS/BSPS or not (figure 3 2) Further, 2091 functional assessment sets f rom 421 different participants were available. These clinical assessment sets do not always contain measurements for all
98 tests under investigation. For example, it is possible that the 10mWT was conducted, but data would be missing for the NMRS during the same assessment period. Data sets from ISNCSCI and functional tests were merged if the data w ere collected within 7 days of each other. This step produced 321 datasets from 213 participants that c ould be classified as having BSS/BSPS or not (sufficient ISNCSCI data obtainable) and for which at least part of the clinical test results were also available. Th is product will be referred to as classifiable datasets 3.3.2. Assessments C lassified as BS S/BSPS A total of 52 of 321 assessments (over different assessment periods) were classified as BSPS according to Pouw ; 14 of 321 were classified as BSS according to Wirz ; 6 of 321 were classified as BSPS/BSS by both Wirz and Pouw and 1 out of 321 was defined as BSPS/BSS according to both Pouw and Hayes. When looking at the definition by Pouw, 18.3% of all assessments were defined as BSPS. In contrast, 6% of all assessments were defined as BSS by Wirz and 0.3% by Hayes when using their 10% threshol d criterion (figure 3 2) 3.3.3 Stability of B eing D efined as BSS/BSPS An individual's classification as having BSS/BSPS actually could change over time, as illustrated in table 31. Of the 17 individuals classified at baseline as having BSPS according to Pouw, only 6 were still classified as having BSPS during at least one later data collection. Of the 8 individuals classified as having BSS by Wirz at baseline, only two were still classified as having BSS by that same definition at a later time point. How ever, three individuals who were initially classified as BSS by Wirz and did not meet the criteria of this definition anymore in a later testing session, were instead classified a s having BSPS according to Pouw All three individuals defined by both Wirz and
99 Pouw as having BSS/BSPS at baseline were defined as having BSS/BSPS at all available later time points by at least one of the definitions ( o ne individual was classified as having BSS by both definitions after 40 and 54 training sessions, one individual was classified as having BSPS according to Pouw after 110 and 131 training sessions and finally, one individual was classified as having BSS according to Pouw after 40 training sessions) 3.3.4 Demographics Table 32 provides demographic information for the 132 classifiable datasets from baseline testing as well as on the 17 of them that were classified as having BSPS according to Pouw, the 8 that were classified as having BSS according to Wirz and the three that met criteria for both definitions. D iffe rences in the distribution of age occurred between the three groups (Pouw, Wirz, Wirz and Pouw; Kruskal Wallis p=0.013). Individuals that fulfilled the criteria by both definitions were the youngest and less than half the age of individuals classified only by Pouw. A ll means and standard deviations are presented in table 32 With regards to the number of years post injury individuals classified as BSS/BSPS were in a more acute stage, but this difference did not reach significan ce (Kruskal Wallis p=0.798). Only participants with cervical injuries were classified as having BSS/BSPS. When looking at all classifiable baseline datasets, cervical injuries were also by far the most common (73.5%). 3.3.5 Functional S tatus of I ndividuals C lassified as BSS/BSPS Only one of the three people classified as BSS/BSPS by both definitions was able to complete the 10MWT and none completed the 6mWT D ifferences were therefore
100 only tested between the individuals defined as having BSPS by Pouw and those defined as having BSS by Wirz. There were no differences in the 1 0mWT (Mann Whitney U, p=0.393) or 6mWT (Mann Whitney U, p=0.571) between the two groups. The effect size comparing the group defined as BSPS by Pouw and the group defined as BSS by Wirz is 0.61 for the 10MWT with the Pouw group ha ving faster walking speeds than the Wirz group. The effect size comparing the two groups regarding the distance covered in the 6mWT is 0.67 with individuals in the Pouw group covering larger distances ( all means and standard deviations are presented in table 33). The large effect sizes indicate that there are potentially important, yet not statistically significant differences between the two groups. No difference in the distribution of the Berg Balance Score (Kruskal Wallis, p=0.555) an y of the SCIFAI subscales (only Pouw and Wirz included, Mann Whitney U, p>0.624) or the NMRS (only Pouw and Wirz included, Mann Whitney U, p=0.075) was detected across the different groups It is worth mentioning that regardless of the lack of significant differences, the Pouw group scored higher than the Wirz group in all examined assessment tools except for the assistive device subscale of the SCI FAI where their scores were nearly identical ( mean 3.3/14 points for Wirz group and 3.1/14 points for Pouw group) 3.3.6 Improvement in F unctional S tatus over T ime L imited data w ere available from later time points from the 28 individuals that were initially identified as having BSS/BSPS Table 34 shows mean change scores per group. No differences in change scores between any of the groups are present in any of the assessments. However, as a whole group, improvement was seen in the 6min walking test ( initial assessment mean: 65.9m, std: 113.7m N=28 ; follow up assessment
101 mean: 180.0m, std: 182.5m N=12 ; Wilcoxon signed rank test, p=0.011), with a very strong t endency to wards improvement in the 10MWT ( initial assessment mean: 28.2sec, std: 35.6sec N=28 ; follow up assessment mean, N=12 : 16.7sec, std: 16.2sec; Wilcoxon signed rank test, p=0.051) 3.4 Discussion Identifying individuals with BBS may be helpful to bridge the translational gap between research in animal models and humans due to the anatomical similarity of human BSS with the often used hemisection model in animals ( Filli, Zorner et al. 2011) Also, the selection of human study participants towards more homogeneous study samples can lead to more powerful detection of treatment effects ( Hayes, Hsieh et al. 2000; Fawcett, Curt et al. 2007) We investigated the effect of applying three recently used definitions for BSS/BSPS to a single dataset. Our results show that there are no significant differences in functional status or in improvement in functional status over 3560 locomotor training sessions based on the definition used. Also, individuals defined as having BSS/BSPS according to at least one of the three definitions at baseline, might no longer meet the classification criteria for BSS/BSPS during follow up assessments. The definition by Pouw lead to the largest number of individuals classified as BSPS. This is not surprising, since they had the least stringent definition regarding motor scores of the three definitions. I ndividuals were classified as having BSPS if there was a t least a 1 point side difference in their motor scores with the corresponding difference in pin prick sensation in the opposite direction ( Pouw, van de Meent et al. 2010) Authors stated that a pure BSS occurs too rarely to be of any clinical significance and they consequently chose a very low threshold of right to left difference to define individuals as having BSPS. The definition by Wirz ignores sensory function, but
102 requires higher motor score differences between right and left sides of 19 points ( Wirz, Zorner et al. 2009) in an attempt to select cases that were as close to a pure BSS as possible. As could be expected, thi s definition was more specific and led to fewer individuals meeting the criteria of their definition. Both definitions led to a low rate of individuals being consistently defined as having BSS/BSPS during follow up assessments The question remains to if t his is due to an inconsistency of the measurement/definitions or rather to the fact that neuroplastic changes that have occurred in response to training, and have lead to more symmetry in motor and/or sensory function. In the latter case, not being classif ied as having BSS (and therefore asymmetric motor function) could mean a positive response to training. However, there were individuals of whom measurements were available over several assessment periods that were repeatedly, but inconsistently classified as having BSS/BSPS. This which seems to indicate that there is no consistent improvement from having BSS to more symmetrical motor abilities. Interestingly, when individuals were defined as having BSS/BSPS by both definitions, they were then consistently identified in that way during later testing sessions. This could indicate that a combining the aspects of both of these definitions into one (u sing sensory and motor scores and setting higher thresholds for minimal right to left differences) could increase specificity and lead to more stable reclassifications duri ng consecutive testing sessions. The definitions used by Wirz and Pouw only identi fied individuals with cervical injuries as having BSS/BSPS. T he high threshold set by Wirz of 19 points difference between right and left may bias the distribution of individuals classified as having BSS
103 towards those with cervical injuries. However, Pouw s definition using a very low threshold of 1 point difference also leads to only cervical injuries being defined as having BSPS. Therefore, this concentration on cervical injuries of both definitions could just be an expression of the fact that the majorit y of all participants in this sample had suffered from cervical injuries. This, on the other hand is reflective of the general population with spinal cord injury ( NSCISC 2011 ) The definition algorithm used by Hayes ( Hayes, Hsieh et al. 2000) l ed to only one assessment being classified as BSS/BSPS of all classifiable datasets with available functional test results W hen ignoring the functional outcome measures and only looking at the 735 classifiable ISNCSCI datasets, four participants were defined as having BSS by Hayes thus 0.5% of all available datasets. This number is much lower than previously reported ( McKinley, Santos et al. 2007) Interestingly, when Hayes tested their algorithms in a group of 56 patients with incomplete SCI nobody was defined as having BBS according to their algorithm All their patients failed to satisfy the criteria for ipsilateral dorsal column dysfunction and contralateral hemianal gesia ( Hayes, Hsieh et al. 2000) N either of the other two definitions included both dorsal column and spinothalamic tract dysfunction. The BSS is thought of as a syndrome that includes all three modalities, ispilateral proprioceptive and motor l o ss and contralateral loss of sensitivity to pain and temperature. This is based on anatomical considerations, namely that a hemisection leads to damage of the motor and dorsal sensory tracts that innervate the ipsilateral body structures and the spinothalamic tract that innervates the contralater al side. By omission of the sensory aspects from the definition, an asymmetrical damage of the anterior cords leading to a difference in motor scores,
104 could be defined as a BSS. The definition algorithm by Hayes is based on the fundamental anatomical ideas and replicates it most precisely. Individuals defined as BSS by Hayes therefore potentially form the most homogeneous group. However, we were not able to test this because not enough individuals met all criteria to be defined as BSS by this stringent def inition. We did not find any significant differences in functional abilities among the three definition groups. This might be surprising since while the definitions by Hayes and Wirz lead to inclusion of only individuals with syndromes close to a pure BSS, the definition by Pouw leads to including individuals with BSPS. Earlier studies, looking at differences in functional outcome between BSS and BSPS participants found differences between the two groups ( Koehler and Endtz 1986; Roth, Park et al. 1991) However, they were not in agreement as to which group had better outcomes. Roth et al. report that individuals with BSPS have a better prognosis, higher Barthel scores and shorter length of stay than pure BSS patients ( Roth, Park et al. 1991) Koehler et al. on the other hand indicate that individuals with pure BS S recover better than those with BSPS, however, the outcome measure they reported was a three point scale of recovery/ disablement/ death ( Koehler and Endtz 1986) T here were no significant functional differences between groups in this study However individuals de fined as BSPS by Pouw scored better than those defined by Pouw on all tests other than the assistive device subscale of the SCI FAI where they had equal scores This, in combination with the low number of individuals included could indicate that with more statistical power, differences between the groups could be man ifest and individuals defined as BSPS by Pouw might have a better functional status than those classified as BSS by Wirz.
105 Another relevant question is whether individuals with BSS are particularly suited to undergo locomotor training. O ne case study has pr eviously investigated the reaction of an individual with BSS to (robotic) locomotor training ( Moreh, Meiner et al. 2009) This study concluded that since this patient had a very severe injury with a nonfavorable prognos is but still showed marked improvement s, it may be concluded that locomotor training could be beneficial in individuals with BSS/BSPS In the present study, individuals who were classified as BSS by any of the definitions were able to walk longer distances in 6 minutes and had a strong t endenc y to walking faster over 10 meters after 3560 sessions of locomotor training, confirming that locomotor training can be beneficial to these individuals. In conclusion, the definition that is used to define BSS/BSPS influences the est i mate of the percentage of people classified as having BSS/BSPS, and possibly their functional outcome. A consensus is therefore needed in order to increase comparability over different studies and even across different animal and human models. Based on our data, we recommend that at least one sensory modality be included in the definition of BSS/BSPS and that thresholds are set to identify those with a clear asymmetry. This could lead to more stable definitions and to definitions that are more reflective of anatomical conditions and therefore to a better ability to compare individuals with human BSS to animal hemisection models.
106 3. 5 Tables and Figures Table 31 Classification as BSS/BSPS over time. Individual Sessions Wirz 2009 Pouw 2010 Hayes 2000 10% Threshold Hayes 2000 5% Threshold 1 0 X 16 23 2 0 X 20 X 3 0 X 4 0 X 40 5 0 X 57 6 0 X 53 96 7 0 X 20 40 55 X 8 0 X 20 9 0 X 94 10 0 X 61 X 11 0 X 108 X 12 0 X 13 X 13 0 X 14 0 X 13 15 0 X 16 0 X 36 17 0 X 44 X For individuals that were classified as BSS by at least one of the definitions at initial testing, we show their classification during later available testing times. Sessions indicate how many training sessions the participant had completed at the time of the corresponding testing session.
107 Table 31. Continued Individual Sessions Wirz 2009 Pouw 2010 Hayes 2000 10% Threshold Hayes 2000 5% Threshold 18 0 X 14 35 X 19 0 X 20 39 X 20 0 X 21 0 X 79 X 22 0 X 23 0 X 24 0 X 47 X 25 0 X 133 X 26 0 X X 40 X X 54 X X 27 0 X X 110 X 131 X 28 0 X X 40 X
108 Table 32 Demographic information. All partici pants (132) Pouw (17) Wirz (8) Both (3) Sex Male Female 102 (77.3%) 30 (22.7%) 15 (88.2%) 2 (11.8%) 6 (75.0%) 2 (25.0%) 1 (33.3%) 2 (66.6%) Age 37.7(17.0) 49.0(15.62) 38.38(15.96) 21(2.65) Years post injury 2.30(4.03) 2.05(4.64) 3.36(7.12) 1.17(1.50) Race African American Hispanic White Undetermined 27(20.5%) 6(4.5%) 98(74.2%) 1(5.9%) 1(5.9%) 14(82.4%) 1(5.9%) 1(12.5%) 7(87.5%) 3(100%) NLI C1 C2 C3 C4 C5 C6 C7 C8 T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11 T12 L1 L2 L3 L4 NT 2(1.5%) 8(6.1%) 10(7.6%) 19(14.4%) 22(16.7%) 17(12.9%) 16(12.1%) 2(1.5%) 3(2.3%) 2(1.5%) 5(3.8%) 3(2.3%) 2(1.5%) 3(2.3%) 3(2.3%) 3(2.3%) 3(2.3%) 3(2.3%) 3(2.3%) 1(0.8%) 1(0.8%) 1(0.8%) 1(0.8%) 1(0.8%) 1(0.8%) 4(23.5%) 3(17.6%) 3(17.6%) 2(11.8%) 3(17.6%) 2(11.8%) 4(50.0%) 1(12.5%) 2(25.0%) 1(12.5%) 1(33.3%) 2(66.6%) AIS NT B C D 6(4.5%) 1(0.8%) 31(23.5%) 94(71.2%) 1(5.9%) 2(11.8%) 14(82.4%) 2(25.5%) 6(75.0%) 1(33.3%) 2(66.6%) Means and standard deviations are reported for age and years post injury. Frequencies and percentages are reported for all other values. NLI: Neurological level of injury; AIS: American Spinal Injury Association Impairment Scale: C: cervical; T: thoracic; L: lumbar; NT: not tested/testable
109 Table 33. Clinical outcomes at baseline. Wirz Pouw Both Ability to perform 10MWT 4 of 8 11 of 17 1 of 3 T ime 10MWT in s ec 67.4 (40.3)s 45.1(33.1)s 22.4s Ability to perform 6m WT 4 of 8 11 of 17 0 of 3 D istance 6mWT in m 72.5(56.4) 141.3(148.5) Ability to perform Berg Balance Test 8 of 8 15 of 15* 3 of 3 Score Berg Balance Test 14.6 (11.4) 19.7(17.1) 16.7(25.4) Ability to perform SCI FAI 4 of 4** 13 of 1 3** 1 of 1* Score SCI FAI (AD) 3.3(2.5) 3. 1 (2. 4 ) 8.0 Score SCI FAI (Braces) 5.8(0.5) 5.9(0.3) 6.0 Score SCI FAI (Gait) 11.5(3.3) 13.0(3.2) 18.0 NMRS score available NMRS (median and range) 4 of 8 2A (1C 2A) 13 of 17 2B(1 3C) 2 of 3 1C(1A 2B) Means and standard deviations are presented unless otherwise noted. 10MWT: 10 meter walking test; sec: seconds; 6mWT: 6 min walking test; SCI FAI: Spinal cord injury functional ambulation inventory; NMRS: Neuromuscular recovery scale; Wirz: individuals defined as having BSS according to the definition by Wirz; Pouw: individuals defined as having BSPS according to the definition by Pouw; Both: individuals meeting criteria of both definitions; *data from 2 participants missing; **data from 4 participants missing
110 Table 34. Improvement in c linical outcomes between baseline and 3560 training sessions. Wirz Pouw Both Classified as having BSS/BSPS after 35 60 training sessions 3 ( Pouw) 1 (Pouw) 1 (Pouw and Wirz) 1 (Pouw) Ability to perform 10MWT 2 of 3 6 of 7 1 of 2 Time reduction 10MWT in sec 49.4(28.9) 37.5(29.4) Ability to perform 6mWT 2 of 3 5 of 7 1 of 2 Distance increase 6mWT in m 174.3(106.5) 171.1(136.9) Ability to perform Berg Balance Test 3 of 3 7 of 7 2 of 2 Increase in Berg Balance Test score 13.3(15.0) 15.7(13.8) 3.5(6.4) Ability to perform SCI FAI 2 of 3 7 of 7 1 of 2 Score SCI FAI (AD) 0 0.7(1.6) 0 Score SCI FAI (Braces) 0 0 0 Score SCI FAI (Gait) 7.0(1.4) 5.5(4.2) 2 NMRS score available NMRS (median and range) ** ** ** Means and standard deviations of the change scores are presented from those participants of whom data for the respective test was available for both the baseline and the follow up testing session. 10MWT: 10 meter walking test; sec: seconds; 6mWT: 6 min walking test; SCI FAI: Spinal cord injury functional ambulation inventory; NMRS: Neuromuscular recovery scale; Wirz: individuals defined as having BSS according to the definition by Wirz; Pouw: individuals defined as having BSPS according to the definition by Pouw; Both: indivi duals meeting criteria of both definitions; *this person was unable to complete the test at baseline, therefore no change score could be calculated; **no data available from this test during any of the assessments after 3560 training sessions.
111 Figure 3 1 Algorithm for the definition of iSCI syndromes adapted from Hayes et al ( Hayes, Hsieh et al. 2000) ; aUEMS/aLEMS average Upper and Lower Extremity Motor Score, T : threshold value (10% will be used as suggested by the authors), aMS : average Motor Score, NLG : Neurological level of injury, LMS/RMS/TMS : Left/Right/Total Motor Scores, 1LT/2LT : Light touch on one and the opposite side, 1PP/2PP : Pin Prick on one and the opposite side, CCS : Central Cord Syndrome, BSS : Brown Squard Syndrome, ACS : Anterior Cord Syndrome, PCS : Posterior Cord Syndrome, CES : Cauda Equina Syndrome.
112 Figure 32 Number of available data sets from the Neuromuscular Recovery Network and number of assessments classified as Brown Squard Syndrome (BSS) or Brown Squard Plus Syndrome (BSPS) Numbers in the Venn diagrams correspond to the numbers of assessments that were classified as having BSS/BSPS by each of the three authors definitions.
113 CHAPTER 4 PROJECT II THE ACOUSTIC STARTLE RESPONSE: ASSESSING PLASTICITY OF A DESCENDING PATHWAY AFTER SPIN AL CORD INJURY 4.1 Background Therapeutic strategies aimed to behaviorally induce plasticity in locomotor neural circuits after incomplete spinal cord injury (iSCI) are currently being tested by the rehabilitation community. Such therapies activate the neuromuscular system below the lesion, thereby promoting walking recovery with pre injury movement patterns as opposed to compensatory strategies to achieve mobility ( Behrman, Lawless Dixon et al. 2005; Behrman and Harkema 2007; Harkema, Behrman et al. 2011 ) To assess possible neural mechanisms mediating these behavioral changes, reliable outcome measures to quantify induced neural changes are needed. One of the nervous centers involved in locomotion is the mesencephalic locomotor region (MLR). In cats ( Shik, Severin et al. 1969; Mori, Matsuyama et al. 1992) and lampreys ( Le Ray, Juvin et al. 2011 ) stimulation of this area leads to increased tone ( Le Ray, Juvin et al. 2011) and initiation of stepping and swimming movements, respectively Increased stimulation intensity is associated with increased stepping/swimming movement speeds ( Shik, Severin et al. 1969; Mori, Ma tsuyama et al. 1992; Le Ray, Juvin et al. 2011) The MLR activates the medial medullary reticular formation, specifically the nucleus gigantocellularis ( Garcia Rill and Skinner 1987) from which the medullary (lateral) reticulospinal tract (ReST) projects to the alpha motor neurons and interneurons that innervate proximal and axial muscl es ( Davis, Gendelman et al. 1982; Davis 1984 ; Garcia Rill and Skinner 1987; Whelan 1996; Koch 1999) The ReST is an important descending pathway, involved in balance and posture as well as in gait initiation ( Steeves and Jordan 1980; Davis, Gendelman et al. 1982; Koch 1999;
114 Kandel, Schwartz et al. 2000; Jahn, Deutschlander et al. 2008) It consists of the mentioned medullary (lateral) part and a pontine (anterior) part that originat es in the caudal pontine reticular nucleus ( Davis, Gendelman et al. 1982; Davis 1984; Koch 1999) The ReST provides tonic input to the spinal networks and is involved in finetuning levels of force output ( Armstrong 1988) fMRI studies have shown these structures and circuits are preserved ac ross species and throughout the transition from quadrupedal to human bipedal upright gait ( Jahn, Deutschlander et al. 2008; Jahn, Deutschlander et al. 2008) In humans, the reticular formation and the ReST are involved in preparation and execution of voluntary actions, especially rapid movements, by providing an increased excitability of the cells involved with the movement ( Valls Sole, Kumru et al. 2008; Valls Sole 2012) In addition to the direct muscle contractio ns, processes in the reticular formation can cause changes in the excitability of diverse motor structures, such as the cortex or the spinal cord ( Rossignol 1975; Valls Sole 2012) Interestingly, these data support clinical observations that reciprocal stepping movements after SCI are ofte n first observed when the patient is in a state of general excitement (personal communication Dr. A.L. Behrman). In ablebodied persons, the main tract that relays information about fine motor control to the limbs is the corticospinal tract (CST). Human bi pedal upright walking depends on the primary motor cortex and the CST for normal stepping as well as perturbed walking and for gait adaptations ( Barthelemy, Grey et al. 2011) After SCI, if the CST is damaged, the ReST could be a possible alternative pathway and a substrate for plasticity and recovery of function ( Jankelowitz and Colebatch 2004 ; Kumru, Vidal et al. 2008) This could be due to th e reason that in contrast to the well bundled fibers of the CST and the
115 vestibulospinal tract, the axons of the ReST are more distributed across the spinal cord ( Nathan, Smith et al. 1996) which can lead to more spared fibers, especially in very localized injuries. The ReST also mediates startle responses, which are defense reactions to unexpected stimuli. This response therefore can provide an indirect measure related to integrity and/or plasticity of this pathway. Visual, auditory or ki nest h etic stimuli can be used, but responses are most consistently elicitable in response to acoustic stimuli. The sensory side of the acoustic startle response (ASR) consists of the cochlear nerve and the cochlear nuclei. Action potentials are generated i n the brainstem (nucleus pontis caudalis) and in motorneurons of the brain stem and the spinal cord. Volleys travel up through the reticulobulbar and down through the medial ReST to alpha motorneurons (directly or via interneurons) and finally to the skeletal muscles of the trunk and limbs, where electromyographic activity in reaction to the startling stimulus can be measured ( Davis, Gendelman et al. 1982; Davis 1984; Brown, Rothwell et al. 1991; Koch 1999; Kumru, Vidal et al. 2008) An exception exists with regards to the Orbicularis Oculi (OO) muscle. Due to its disproportionally short response latency, this reaction is likely not mediated through the ReST, but through other pathways. Therefore, the OO response is considered a separate part of the acoustic startle response ( Hori, Yasuhara et al. 1986; Valls Sole, Kumru et al. 2008) The ASR has been described in ablebodied individuals ( Rossignol 1975; Wilkins, Hallett et al. 1986; Br own, Rothwell et al. 1991; Grosse and Brown 2003) and in populations with neurological ( Jankel owitz and Colebatch 2004; Nieuwenhuijzen, Horstink et al. 2006 ; Kumru, Vidal et al. 2008 ) and psychological ( Cadenhead, Carasso
116 et al. 1999; Csomor, Yee et al. 2009) diagnoses. In ablebodied controls, the ASR can be elicited more readily in proximal and axial muscles. Positive r esponses are comparatively lower in the distal leg muscles than more proximal muscle groups Response latencies increase as the distance of the corresponding muscle from the brainstem increases ( Brown, Rothwell et al. 1991) Resp onses can be more readily elicited and are of shorter latencies when the muscle is involved in maintaining posture ( Brown, Day et al. 1991) Increased numbers of positive responses, as well as increased EMG amplitudes often are observed in neurologic populations when compared to ablebodied controls ( Jan kelowitz and Colebatch 2004; Kumru, Vidal et al. 2008) Researchers suggest this could be due to plastic changes in the R eST in response to damage to the spinal cord ( Jankelowitz and Colebatch 2004; Kumru, Vidal et al. 2008) Normally the ASR habituates rapidly in both animals and humans ( Davis, Parisi et al. 1982; Valsamis and Schmid 2011; Valls Sole 2012) leading to lower response rates, smaller amplitudes, and shorter durations over repeated exposures to the startling stimulus ( Wilkins, Hallett et al. 1986 ) Habituation typically occurs within 26 stimulations, even with interstimulus intervals as large as 20 min ( Brown, Rothwell et al. 1991) Th e only exception to this is the Orbicularis Oculi muscle, which does not habituate. This is another indication that this muscles reaction is likely mediated through an alternative pathway ( Valls Sol e, Kumru et al. 2008) Both short term habituation (within a testing session), as well as long term habituation (across testing sessions over five days), have been described in rodents ( Valsamis and Schmid 2011) T hese types of habituation have not been well characterized in the human population. If
117 present th e effect decreases the ability to establish a reliable assessment of the ASR through repetitive testing and thus compromises the ability to use the ASR as an outcome measure. As a result, determining ways to effectively prevent, decrease, or postpone habituation during ASR testing, thereby increasing the consistency of ASR responses would significantly enhance the utility of this response as an outcome measure. Most studies investigating the ASR in individuals with neurological deficits tested participants while sitting rel axed, reclining or lying supine ( Brown, Day et al. 1991; Brown, Rothwell et al. 1991; Voordecker, Mavroudakis et al. 1997; Jankelowitz and Colebatch 2004 ; Kumru, Vidal et al. 2008) However, it has been shown that the ASR is modulated during human gait ( Nieuwenhuijzen, Schillings et al. 2000) ; and that ASR responses are more consistently elicited during standing in ablebodied individuals ( Brown, Day et al. 1991) Additionally, effects of posture on other neurophysiological measurements such as transcranial magnetic stimulation ( Rossini, Barker et al. 1994) and galvanic vestibular stimulation ( Lund and Broberg 1983; Wydenkeller, Liechti et al. 2006; Liechti, Muller et al. 2008) already have been established. In order to obtain more consistent responses, it may therefore be more appropriate to examine the ASR in individuals while their muscles are actively maintaining posture The goal of this study therefore was twofold. First, we aimed to de scribe short term (within session) habituation of EMG responses to acoustic startling stimuli over 15 stimulations in individuals with iSCI and ablebodied control participants. We investigated in the effect of position on this short term habituation and hypothesized that more consistently recordable responses would be elicited in positions requiring more
1 18 activity to maintain posture and balance. Second, in able bodied individuals, we aimed to assess the amount of long term habituation with this protocol. We hypothesized that if habituation occurred within a series of 15 stimulations, ASR responses would reappear if tested 48hrs later; however responses w ould be less frequently elicited than during the first testing session Establishing the characteristic s of both short term and long term habituation to the ASR in human subjects is important in order to optimize use of this response as an outcome tool to assess neuroplastic changes in the ReST in clinical trials. 4.2 Methods 4.2.1 Participants A total of 9 individuals with iSCI and 9 sex and age (+/ 5 years) matched ablebodied control participants were included in this study. All individuals with iSCI were classified as ASIA impairment scale (AIS) D and able to stand independently and repeatedly for the d uration of the testing. Please refer to Table 41 for a complete list of inclusion and exclusion criteria and table 42 for complete demographic information on all participants 4.2.2 Auditory Stimulation During a testing session, 15 computer generated audi tory stimuli (locochrome, 120ms, 105db) were delivered binaurally using earbud headphones at random interstimulus intervals of 28 min distributed over a 45 min session The locochrome sound is a blowing train horn, which is meant to warn of a potential danger in real life situations. We therefore considered it particularly suitable to induce a startle response.
119 4.2. 3 Recording Surface EMG was recorded from 13 muscles bilaterally: Orbicularis Oculi (Pars Orbitalis, OO), Sterno Cleido Occipito Mastoideus (SC ), Biceps Brachii (BB), Triceps Brachii (TB), Extensor Carpi Radialis (ECR), Flexor Carpi Radialis (FCR), Rectus Abdominis (RA), Errector Spinae (ES), Gluteus Medius (GM), Vastus medialis (VM), Medial hamstrings (MH), Tibialis Anterior (TA) and Soleus (Sol ). The skin was shaved, if necessary, and cleaned with alcohol to optimize electrode contact and reduce skin impedance. Single sweeps of 2000ms including 100ms pre stimulus, were recorded through the Nexus 1.7.1 data acquisition system (Version 1.7.1; Vicon Motion Systems Ltd.). EMG were sampled at 2 kHz, amplified, filtered and exported for analysis with custom written programs (Matlab, The MathWorks Inc, Version R2008a). 4.2. 4 Positions for Testing The positions tested include lying supine (referred to a s supine) standing on solid ground (referred to as standing) and standing on a foam pillow (Airex Balance Pad, 47cm x 38.5cm x 7cm ; referred to as foam ). Participants were wearing a harness for safety while standing and standing on foam, but littl e to no weight support was provided ( 2 lbs). Positional orders were block randomized, and every matched pair (person with iSCI and their matched, ablebodied control participant ) w as tested using the same position orders and inter stimulus durations. Five blocks of three stimulations were presented, with each of the blocks consisting of a random order of the 3 positions. Random orders were generated by a list randomizer (random.org), which uses white noise for random list generation.
120 4.2. 5 Testing P rotocol After instrumentation, participants were standing relaxed with eyes open. Talking was kept to a minimum during the entire testing session. A first stimulation was delivered once the participant was standing quietly. After this first reference stimulation, 15 stimulations were delivered in a random order of positions and at random interstimulus interval s as described above. During each stimulation, the participant was observed and visi bly noticeable muscle contractions/limb movements in response to the stimuli were noted on the data collection sheet. After each stimulation the participant was asked to rate the perceived intensity of startle as described earlier The participant then was instructed to adopt the next position on the randomized list, with assistance provided as necessary. This procedure was repeated for all 15 stimulations (approximately 45min). Finally, another reference stimulation was conducted in standing. In ablebodie d controls, th is entire protocol was repeated 4250hrs after the first experiment. 4.2.6 Outcome M easures 184.108.40.206 Percentage of response in the group of individuals with SCI and in the group of a ble bodied control participants For each stimulation (total 1 5 stimulations consisting of 5 blocks of one stimulation in each of the three positions) and each muscle (total 22 muscles excluding the OO), the percentage of individuals responding to the stimulation with an evoked potential out of all valid trials was d efined. Responses in OO muscles served as a positive control to determine whether the startle was elicited in each participant and trial. Trials were only considered valid if there was a positive response in at least one of the OO muscles in this trial and participant. If all participants showed a response in at least one of the OO
121 muscles in one specific trial, then the total possible number of responses for this trial was 9. Further, if three out of nine individuals with SCI showed a response in a specifi c muscle in this trial, then the percentage of response for this trial and muscle would be 33.3% (3 out of 9 responses). If however, one person did not show a response in either OO in this trial, the trial of this person was eliminated. Percentages of resp onses for all muscles in this trial were then calculated out of 8 instead of 9. In the above example, the percentage of response for this particular muscle would be 3 out of 8, hence 37.5%. The percentages of responses for each muscle were then averaged ov er all muscles with the exception of the OO in order to get an overall percentage of response for a specific stimulation (i.e. one position in one block). This variable was computed to make comparisons between groups, stimulations and muscles, but could not be used to make comparisons or establish relationships between individuals. 220.127.116.11 Rates of response in each individual Rates of response for each individual and stimulation were computed by calculating the number of muscles with a positive response in this trial and individual relative to the 22 possible responses (one in each muscle other than the OO). These rates were then averaged over all valid trials in this individual. Valid trials were defined again as only trials with at least one positive response in either of the OO muscles. Rates of response were used when investigating differences between individuals within the same group. This variable was computed to make comparisons or establish relationships between individuals, but could not be used to m ake comparisons between individual muscles.
122 18.104.22.168 Subjectively perceived startle In addition to the EMG measurements, participants rated their subjectively perceived startle on a scale from 010 with 0 indicating no startle at all and 10 indicating the m ost imaginable startle after each stimulation. 4.2. 7 Data Analysis For each stimulation and muscle, presence of an EMG response was defined visually from both rectified and nonrectified EMG traces if they were: A) clearly distinguishable from background EMG and B) repeated in at least 2 out of the 15 trials. Percentages of response in each group/testing session as well as rates of response in each individual participant were then created as described above. 4.2. 7 .1 Data processing EMG data were demeaned and filtered. A second order bi directional butterworth filter was applied (10200Hz). Data were rectified to define the presence or absence of a response for each stimulation and muscle in every individual Perce ntages of responses for each individual trial and muscle and rates of response for each individual trial and participant were then calculated as described above. 4.2. 7 2 Short t erm h abituation To assess short time habituation of the ASR within one testing session, percentages of response per stimulation were averaged for the three stimulations (lying, standing, foam) in each of the 5 blocks and these block averages were compared against each other 4.2. 7 3 Longt erm h abituation In able bodied control subjects, percentages of response were averaged over the three stimulations in each block per testing session. The average over the three
123 stimulations in the first block, first session were then com pared to the average over the three stimulations in the first block, second session and so forth. 4.2. 7 4 Effect of position Total percentages of response per position (lying supine, standing, standing on foam) were created by averaging percentages of resp onse for all 5 trials in each position for both group and both session. 4.2. 7 5 Statistical analysis Descriptive statistics. Means and standard deviations are presented for demographic factors where not otherwise noted. ANOVA models. We used analyses of variance (ANOVA) models to test for differences between testing conditions and groups. Alpha levels of 0.05 were used for all analyses and GreenhouseGe is ser corrections were used to correct for violation of the sphericity assumption w here appropriate. Post hoc tests were conducted if the omnibus model indicated a significant difference between at l east two of the comparisons Where not otherwise indicated, Bonferroni corrections were used to account for multiple testing. Perceived star tle. A mixed model analysis of variance was conducted to test for changes in perceived startle over the 5 blocks of 3 stimulations each ( short term habituation) and for differences in perceived startle between individuals with iSCI and ablebodied control participants. This first model included perceived startle measurements as the dependent variable. Block (15) and position (lying, standing, foam) were entered as withinsubject factors and diagnosis (SCI/control) was entered as the betweensubject factor.
124 A repeatedmeasures ANOVA to test for differences in perceived startle between the first and second testing session in ablebodied participants ( longterm habituation) included perceived startle measurements as dependent variable. Session (session 1/session 2), position (lying, standing, foam) and block number (115) were used as within subjects factors. Percentages of response to acoustic startle. Preliminary analyses indicated that there is an effect of body side (which was different in different muscles) for the comparison between individuals with iSCI and ablebodied controls as well as for the comparison between the first and second session in ablebodied controls. We explored this effect by qualitatively comparing the muscle response with handedness as well as motor scores in the corresponding segments for individuals with iSCI. We also searched for patterns regarding which muscles were concerned by the side differences i.e. w hether specific muscle groups were involved such as only flexors or only muscles of the lower limbs However, the effect seemed to be nonsystematic and was therefore considered a noninformative epiphenomenon. The factor side was not included in the final statistical models. A mixed model ANOVA was conducted, using percentage of response in individuals with iSCI and ablebodied control participants during their first testing session as dependent variable. Block (1 to 5) and position (supine, standing, foam) were entered as within and diagnosis as betweensubject s factors. To investigate whether long term habituation occurs for the ASR, a repeatedmeasures ANOVA was conducted on outcomes from two testing sessions, approximately 48 hrs apart. Percentage of response to acoustic startle in ablebodied
125 participants during their fir st and second testing session was entered as the dependent variable, session (1st and 2nd), block (1 5) and position were entered as withinsubject factors. Effect of Position on specific muscle groups. To assess the effect of position during an acoustic s tartle on different muscle groups, we selected and grouped certain muscles as follows. The trunk muscles RA and ES were compared against the lower extremity muscles TA and Sol and the upper extremity muscles FF and FE to describe the effect of position on different limbs. The hip and thigh muscles GM, VM and the upper arm muscles TB and BB were compared against the shank muscles TA and Sol and the forearm muscle groups FF and FE in order to test for the effect of position on proximal and distal muscles. Finally, the extensors TB, FE, ES, GM, VM and Sol were compared against the flexors BB, FF, SCOM, RA, MH and TA to describe the effect of position on extensor and flexor muscles. Three different repeatedmeasures ANOVA models were conducted to investigate the differential effect of position on muscle groups as defined above. The first model included percentage of response as dependent variable, muscle type and position as within group factors and group as between group factor. The second model included percent age of response as dependent variable, body part and position as within factors and group as between factor. The third model included percentage of response as dependent variable, muscle location and position as within factors and group as between factor.
126 4.3 Results 4.3.1 Demographic I nformation Table 42 presents demographic information for all participants. It is worth not ing that the average time since injury was skewed by one participant that was injured 51 years ago. The mode of time since injury was 5years. 4.3. 2 Percentages of R esponse 4.3. 2 .1 Short term habituation The mixedmeasures ANOVA model indicated that t here was a significant interaction between block, position and group (F(4.881,107.390)=2.958, p=0.016). Post hoc testing revealed that in individuals with SCI, there was no habituation in any position and no differences between the conditions. In ablebodied control participants while lying supine, there was a reduction in the percentage of resp onse between block 1 (mean: 28.125%, std: 19.31%) and all other blocks (block 2 mean: 14.8% std 14.46 % ; block 3: mean 15.73%, std 16.04%; block 4: mean 13.85%, std 12.63%; block 5: mean 3.7%, std: 9.85%; p for all pairwise comparisons <0.007). In addition, the percentage of responses in block 5 was significantly smaller than the percentage of responses in all other blocks (all p values<0.008), suggesting short term habituation did exist in this group and position. Reductions in the percentage of response, and therefore habituation, could also be seen in the standing condition between blocks 1 (mean: 27.76%, std: 15.35%) and 4 (mean: 17.58%, std: 12.03 % p=0.001) 1 and 5 (mean: 3.7%, std: 9.85%, p<0.001) 2 (mean: 24.06%, std: 17.62%) and 5 (p=0.002) and 3 (mean: 21.86%, std: 14.23%) and 5 (p= 0.003) while no significant pairwise comparisons were seen when participants (ablebodied) were standing on foam, indicating that there
127 was no short term habituation in these control participants when standing on foam ( Figure 4 1 ). 4.3. 2 .2 Longterm habituation The repeated measures analysis including data from the first and second session of able bodied control participants revealed a significant interaction of block, session and position (F(8,88)=3.663, p=0.001). Differences between sessions were found in block 1 while individuals were lying supine. Able bodied i ndividuals show significantly lower percentages o f response during the second testing session (mean: 9.375%, std: 12.07%) when compared to the first session (mean: 28.125%, std: 19.31%; p=0.002). Further differences between session 1 and 2 can be found in block 3 while on foam, with individuals showing significantly fewer responses during the second testing session (session 1: mean 26.04%, std 12.45%; session 2: mean 9.53%, std 11.13%; p=0.01) and in standing block 5, where percentages of responses were significantly in creased during the second testing session (mean: 23.125%, std: 14.56%) when compared to the first (mean: 7.4%, std: 8.64%; p=0.004). While there is short term habituation during the first testing session as described above, there were few significant differences between blocks in any of the positions dur ing the second testing session No differences between block were present while lying supine. In standing there was a reduction in percentage of response from block 1 (mean: 27.76 % std: 15.35% ) to block 3 (mean: 20.35% std: 15.58% ; p=0.046). While stand ing on foam there were reductions from block 1 (mean: 25.00% std: 7.54 % ) to block 3 (mean: 9.53% std: 11.13% ; p=0.037) and block 2 (26.04 % std: 9.9 1%) to block 3 ( p=0.037) To summarize when testing happens in supine, individuals initially have lower percentages of response during the second testing session However, due to the fact that there is short t erm habituation during the
128 first, but not the second testing session, percentages are not different between sessions one and two in later blocks. While standing on solid ground, individuals have similar responses during the first stimulation blocks of each session, but due to short term habituation during the first, but not the second testing session in this position, responses are larger during the second testing session in block 5. Finally, when testing on foam, no longterm habituation could be detected. A temporary reduction in percentage of response in block 3 (mean:9.53%, std: 11.13%) was followed by an increas e in block four leading to no short ter m habituation as expressed by a lack of significant differences between blocks 1 (mean: 25.00%, std: 7.54%) or 2 (mean: 26.04%, std: 9.9%) and blocks 4 (mean: 23.13%, std: 17.36%) or 5 (mean: 18.5%, std: 13.66% ; Figure 42 ) 4.3. 2 .3 Effect of p osition on specific muscle groups Flexor and extensor muscles. This statistical model did not reveal any significant interaction effects and only a main effect of muscle type (F(1,58)=15.37, p<0.001). G enerally a larger percentage of responses were seen in flexor (m ean: 24.85% std: 15.88% ) than extensor muscles (mean: 18.95 % std: 13.79% ; Figure 43 ). Muscles in the trunk and extremities. A main effect of limb (F(1.154,20.768)=7.134, p=0.012) and post hoc comparisons indicated that higher percentages of response we re found in the trunk (mean: 24.29%, std: 13.71%) when compared to the lower extremity (mean:13.93%, std: 13.13% p=0.036; Figure 44 ) Proximal and distal muscles. This model revealed a significant interaction of position and muscle location (F(1. 589,60.385 )=4. 567,p=0.0 21). Proximal muscles showed higher percentages of response when standing on foam (mean: 25.89 std : 14.50) than when standing (mean: 19.88 % std : 13.44 % ;p= 0.006 ) or lying supine
129 (mean: 19.185 % std : 11.88% ; p=0.0 09). Distal muscles showed lower percentages of response when lying supine (mean: 12.733, std: 11.17) in comparison to when standing (mean: 19.635, std: 10.96; p=0.001) and a tendency towards lower responses compared to when standing on foam (mean: 18.880, std: 14.25; p=0.05 1 ; Figure 45 ) 4.3. 2 .4 Additional exploration of the data The following are additional results from qualitative analyses that were not predetermined, but worth noting. Considerable withingroup variability. We noted substantial differences in individuals reactions to acoustic startle that were not dependent on their diagnosis. Tables 43 to 4 6 and Figures 46 to 4 9 provide examples of a person with low response rate and a person with high response rate for the individuals with SCI and the ablebodied control participants. We explored if these differences could be dependent on sex (figure 41 0 ) or age (figure 41 1 ) Analyses did not indicate a relationship between ASR response rates and age (R2=0.165, p=0.513) or differences dependent on sex (MannWhitney U: 38.00, p=0.892). Relationship between ASR responses and motor scores. No correlation was revealed between ASR response rate and upper extremity (R2=0.434, p=0.243) or lower extremity motor scores (R2= 0.039, p=0.921; figure 41 2 ). Periods of EMG silen ce. In the course of data analysis, w e noticed a phenomenon of periods of EMG silence either with or without a previous period of increased EMG activity (ASR response) in several participants and across different positions and muscles For an example, please refer to figure 4 1 3 Relationship between rates of ASR response and subjectively perceived startle. A relationship betweensubject ively perceived startle and rates of ASR
130 responses could not be shown statistically (R2=0.334, p=0.176). However, this could be driven mainly by the small number of individuals and there could potentially be a relationship between those factors (figure 4 1 4 ) 4.3.3 Subjectively Perceived Startle The mixedmodel ANOVA testing for differences over the course of 15 stimulati ons in individuals with iSCI and ablebodied control participants revealed no interactions between any of the factors; no differences between individuals with iSCI and ablebodied control participants but a main effect of block (F(2.021,32.336)=5.551,p=0.008). Post Hoc tests revealed no differences in pair wise comparisons when correcting for multiple testing with a Bonferroni adjustment. Significant reductions in subjectively perceived startle between all blocks (block 1: mean 6.48, std 2.17; block 2: mean 6.54, std 2.01, block 3: mean 6.11, std 2.23, block 4: mean 5.78, std: 2.60) and block 5 (mean: 5.39, std: 2.73; all p values<0.021) as well as between block 2 and 4 (p=0.029) were revealed when using the less restrictive LSD correction (Figure 41 5 A.). W hen normalizing subjectively perceived startle values to the value indicated during the reference stimulation, the same pattern of differences was present (main effect of block; F(1.862,29.795)=5.388, p=0.011) with LSD corrected post hoc tests revealing re ductions in perceived startle between block 1 to 4 (block 1: mean 0.89, std 0.26; block 2: mean 0.89, std 0.26, block 3: mean 0.83, std 0.29, block 4: mean 0.79, std: 0.36) and block 5 (mean: 0.73, std: 0.37; all pvalues<0.022; Figure 41 5 B). The statisti cal model comparing perceived startle during the first and second testing sessions in ablebodied control participants showed no significant effect (Figure 4 16A). However, when subjectively perceived startle responses were normalized to the values indicat ed during the reference stimulation and entered into the model, a main
131 effect of block was revealed (F(1.758,14.064)=3.395, p=0.048). Post hoc testing using a Bonferroni correction for multiple testing showed no differences in pair wise comparisons, post h oc testing using LSD correction for multiple testing showed differences between blocks 1 (mean: 0.94, std: 0.21) and 3 (mean: 0.76, std: 0.37; p=0.031) and blocks 1 and 5 (mean: 0.71, std: 0.39; p=0.043) (Figure 416 B). Figure 417 shows subjectively perceived startle by position. Participants subjectively perceived startle was not dependent on their position. 4.4 Discussion The ASR is a potentially important test for ReST function and plasticity after SCI and training. However, previous studies have descr ibed habituation of the response within 2 6 t rials, precluding researchers from test ing repeatedly or us ing signal averaging to achieve stable and reliable measurements with this neurophyisological test. We investigated if changing positions and increasing postural demands resulted in a reduction of said habituation. 4.4. 1 Orbicularis Oculi R esponses Several authors have suggested that the first response in the Orbicularis Oculi is not mediated by the same reticulospinal pathway as all the other muscles ( Brown, Rothwell et al. 1991; Kumru, Vidal et al. 2008 ) based on their relatively faster response time as well as their lack of habituation. Our data supports this notion. OO responses were consistently elicitable. In fact, there was only one ablebodied c ontrol participant in wh om responses from the OO could not be recorded consistently. In this participant, eyeblink was visible in about onehalf of the trials where no evoked potential could be distinguished from background EMG. Therefore, it is possible electrode placement may have been ineffective/poor. This possibility, however, seems less likely when
132 considering that similar results were obtained during the second data collection 48hrs later. 4.4. 2 Decrease in S hort T erm H abituation With our protocol, i ndividuals with iSCI did not appear to habituate, short term, to the acoustic startle. In ablebodied control participants, however, habituation was present in supine and standing, but not while standing on foam. This supports our hypothesis and suggests t hat changing positions and increasing postural demands may provide one method to postpone habituation in ablebodied control participants Several authors have previously described that individuals with SCI ( Kumru, Vidal et al. 2008) or other neurological deficits ( Voordecker, Mavroudakis et al. 1997; Jankelowitz and Colebatch 2004; Kofler, Mull er et al. 2006; Csomor, Yee et al. 2009) have increased response parameters (probability of response, amplitudes, latencies etc) to acoustic startle compared to ablebodied control s. O ur data also provide supporting evidence of this phenomenon. Furthermore, we were able to show that individuals with iSCI had similar percentages of response over all stimulation blocks (no habituation) while percentages of response in ablebodied control participants decreased with increasing stimulation numbers (habituation present) U sing stability of responses to repeated stimulation has been suggested as a stable marker for disease severity in schizophrenia (however, they only tested the OO muscle) ( Cadenhead, Carasso et al. 1999) Considering the differences in habituation detected between controls a nd individuals with SCI th e response to the ASR over time might also be useful as an additional parameter to classify ReST function in populations with movement related neurological disorders An increase in the number of stimulations
133 required until responses have habituated could be an expression of ongoing neuroplastic processes. 4.4. 3 Longterm H abituation in A ble B odied I ndividuals If the ASR is to be used as a measurement tool to assess changes over time, for example neuroplasticity in response to an intervention, then test retest reliability and absence of long term habituation is critical. However, long term habituation has been established in an animal model ( Valsamis and Schmid 2011) In these rodent models, whole body startle movements rather than EMG responses are assessed. A few studies assessed test retest reliability in humans ( Braff, Stone et al. 1978; Schwarzkopf, McCoy et al. 1993) and authors concluded that responses to an acoustic startle are a rather stable phenomenon when measured over three weeks. However, they focused on a different technique (prepulse inhibition), an d assessed only the OO muscle in a population with schizophrenia. In fact, considering that there does not appear to be any short term habituation to acoustic startle in the OO muscle ( Hori, Yasuhara et al. 1986 ; Valls Sole, Kumru et al. 2008) it is not surprising that no longterm habituation was found ov er three weeks. Our data looking at 12 muscle pairs throughout the body, illustrate that in ablebodied control participants, there is an effect of long term habituation over 48hrs. This effect is most pronounced during the first block in the supine posit ion. Yet, over 5 blocks of three stimulations within one testing session, percentages of response decrease to a different extent leading to similar responses between sessions i n the last testing blocks. This pattern is specifically true when individuals are tested in supine, suggest ing that in order to achieve stable, reliable test retest results when testing in this position, responses from initial stimulations should be discarded and only later responses reliably be compared. However, if individuals are
134 standing on solid ground or on foam, no longterm habituation seems to be present. Therefore, in individuals who are capable of standing or standing on foam, it m ay be advisable to test the ASR in either of these posi tions to achieve stable test retest results. We showed long term habituation after 48hrs in ablebodied individuals. A t this point, it remains unclear whether long term habituation of the ASR in muscles other than the OO may still b e present at later time points n or if it is present at all in persons with SCI. Considering that individuals with SCI do n o t show the short term habituation that ablebodied control participants exhibit, it is not clear that the long term habituation we found in ablebodied cont rols is also present in individuals with SCI. Therefore, it would be of interest in future studies to investigate long term habituation in participants with SCI and focus on time frames typically used in intervention studies This would determine the impac t habituation may have on outcomes associated with intervention studies. 4.4. 4 Influence of P osition on P ercentage of R esponse of S pecific M uscle G roups We chose to collect EMG from a wide range of flexor and extensor muscles, from upper and lower extremity muscles and from proximal and distal muscles to determine patterns and changes in muscle r esponses. This was based on some neurophysiologic thoughts. T he two parts of the ReST (medullary and pontine parts) are believed to activate different muscle groups. The medullary ReST is thought to reinforce the corticospinal tract and the rubrospinal tract by enhancing flexor movements while the pontine ReST rein forces the vestibulospinal tract in enhancing extensor movements ( Felten and Shettz 2010) With the ASR being mediated mainly through the pontine ReST ( Davis, Gendelman et al. 1982; Davis 1984; Delwaide and Schepens 1995; K och
135 1999; Jahn, Deutschlander et al. 2008; Jahn, Deutschlander et al. 2008; Felten and Shettz 2010) one could hypothesize that the response to a startle in able bodied individuals is more pronounced in the extensors than in the flexors Extensor responses seem to be present in animals such as the cat ( Wright and Barnes 1972) or in rats where a startle leads to rapid extension of the lim bs so that the rat jumps into the air ( Davis 1984) If this would hold true in ablebodied humans, then changes in this pattern after SCI could be an indication of neuroplastic changes in the ReST. A flexor pattern could indicate that ReST fibers are now terminating on synapses that preinjury have been occupied by fibers from the corticospinal tract. However, our results showed higher responses in flexors in both the ablebodied contr ol participants as well as in individuals with iSCI and no difference in patterns between the two groups. Therefore, o ur results are consistent with earlier work that described a general flexor response in humans ( Rossignol 1975; Davis 1984) Rossignol for example studied responses to acoustic startle in the TA and gastrocnemius muscles and out of 97 responses that he was able to record, 74 were found in the flexor TA. It seems that a more complex relationship exists between the function of the ReST and responses to acoustic startle and that a change in flexor extensor distribution in responses might not be a good indicator for neuroplastic changes in the ReST. We expected differences in percentage of response based on position to be dependent upon the body part involved. We had based this on earlier work by Brown et al. ( Brown, Day et al. 1991) who indicated that a weight bearing activity is necessary to increase responses in corresponding muscles Specifically, they did not find a change in responses in the BB when participants went from sitting to standing, but they did find
136 increased response probability and an additional early component to the response when individuals were crouching on all fours, hence loading weight through their arms. We therefore expected that differences between lying supine and standing would be more pronounced in the lower extremity and trunk muscles when compared to the upper extremity muscles. Howev er, this was not the case. In the six muscles that we included in this analysis, we did not find differences in percentages of response that were dependent on the position. In general, trunk muscles always had the largest percentages, followed by upper and then lower extremity muscles The percentage of responses increased in parallel in all three limbs when positional demands were increased. In individuals with iSCI, it is possible that standing was a very challenging postural task and therefore, individuals were in a more general state of increased descending drive. This state may contribute to the fact that percentages of response increased not only in the weight baring lower extremities and the trunk, but also in the UE when position is changed from lying supine to standing and standing on foam. However, this would not explain why the same effect was shown in ablebodied control participants when changing from lying supine to standing. Further research is necessary to study the differential effects of pos itional changes on the percentage of ASR responses in the different limbs. All our participants were able to stand independently. While standing on foam, a therapist was guarding the participants in addition to the safety harness, but no contact was provided during or close to the stimulations. It is unclear how using body weight support or contact assistance from a therapist would alter the ASR responses Presently, we can only draw conclusions for higher functioning individuals with iSCI.
137 Testing the ASR might however be useful early after injury, both for prognostic purposes as well as to track changes in response to therapeutic interventions. With our protocol, we presented the participants with a random order of positions. We can therefore not conclude with certainty if the reduced habituation was due to the fact that we constantly switched positions or if it was due to the position with different postural demands themselves. We have reason to believe that it was the positions themselves that were the in fluencing factor, due to the fact that regardless of block or group or session, responses almost always seemed largest when standing on foam and smallest in supine, even if some of those differences did not reach statistical significance. However, in order to draw this conclusion with certainty, a study would have to be designed that compares series of stimulations in the individual positions to a series with randomly changing positions. 4.4. 5 Within group D ifferences Large withingroup differences were seen in participants with iSCI as well as in a ble bodied control participants Both groups contained individuals that habituated rapidly over only a few trials, while others responded in multiple muscles over all 15 trials. This was the case despite the fact that our group of individuals with iSCI consisted uniquely of people that were rated AIS D, and had high ASIA motor scores in both the upper and lower extremities. In preliminary testing during protocol development we suspected an effect of sex on reactions to acoustic startle and decided to match our participants with SCI to control participants regarding that factor in order not to induce sex as a potentially confounding factor. However, qualitative analysis of our results did not indicate a relationship between rates of ASR responses in individuals and sex Similarly, age did not explain
138 within group differences Informal interviews with participants indicated that some individual differences may be explained with prior startling experiences. In this cas e we would expect individuals with higher percentages of response to indicate higher subjectively perceived startle. An explorative analysis indicated that this could be the case However, this study was not designed to test any of these effects and these results can only be looked at as possible directions for future research. 4.4. 6 Shape of R esponses Due to voluntary activation of muscles while standing or standing on foam the onset and mostly offset of responses were sometimes difficult or impossible t o evaluate. In many of these cases, we did, however observe a phenomenon of decreased activity that was time locked to the stimulus and clearly reproducible over multiple stimulations. This phenomenon has previously been described in ablebodied humans ( Rossignol 1975) and in anim al models ( Eccles, Ito et al. 1967; Wright and Barnes 1972; Ito 2006) It has been suggested that the startling sound leads to an inhibition of intra cerebellar nuclei and therefore to a reduction of their facilitation of neurons of the caudal pontine and medullary reticular formation. This in turn leads to a reduced excitation of spinal motorneurons by the ReST. In fact, these periods of reduced activi ty a re diminished or abolished after cere bellectomy ( Eccles, Ito et al. 1967; Wright and Barnes 1972; Ito 2006 ) Rossig n ol described that this period of EMG silence mostly occurred after a period of excitation that is clearly distinguishable from background EMG However, he describes occasions where the period of silence was the only response. Also, even in the cases where a n ASR response (increased EMG activity) was present, latency and duration of the period of silence were more consistent than latency and amplitudes of
139 ASR responses ( Rossignol 1975) Alth ough neuroplastic changes of supraspinal centers have been described in reaction to SCI ( Curt, Alkadhi et al. 2002) it is likely that the majority of recordable EMG changes are due to processes, both damage and neuroplasticity, at subcortical levels. Therefore, a change in occurrence, latency or duration of the period of electromyographic silence might be reflective of changes in function of the ReST as both excitatory and inhibitory volleys might travel slower. Our study was not designed to investigate these periods of EMG silence, in fact, they were a serendipitous finding. However, further investigations might be valuable in an effort to increase the value of the ASR as a test of ReST function. 4.4. 7 Meaning of I ncreased ASR R esponses We have investigated ASR responses in individuals with SCI that had a considerably high level of function. W e qualitatively compared their ASR response rate (muscles with a response in each individual out of the 22 muscles measured) to their motor scores and failed to show a relationship. This is not enough to draw conclusions about the meaning of increased responses for the individual. Several authors suggested that increased ASR responses can be a sign of neuroplasticity in the ReST to compensate for loss of neurons or axons in other tracts connecting supraspinal centers with the spinal cord ( Jankelowitz and Colebatch 2004; Kumru, Vidal et al. 2008) In this case, increased responses should be related to increased functional capabilities. Alternatively, it can be hypothesized that the ReST terminates on synapses that were previously occupied by corticospinal tract axons. It can be imagined that they are competing with axons from the CST, preventing the regeneration of preinjury synapses between the CST and interneurons/alpha motorneurons. In this case, the increased percentages of response in individuals with iSCI might not be advantageous. It can
140 even be imagined that this question depends on the stage of recovery after SCI. Potentially, neuroplasticity in the ReST is advantageous initially after injury, while in later recovery stages, it is advantageous if these synapses are occupied by CST fibers again. However, at this point, only speculatio n is possible. In order to investigate the relationship of percentages of ASR response with functional abilities a study including both measurement types would need to be conducted. 4.4.8 Differences in P erceived S tartle We included a measurement of subjectively perceived startle into the testing protocol in reaction to feedback from individuals with SCI about an increase in their proneness to startle since their injury. Increased startle reactions could be an expression of a general increase in excitabili ty in reaction to the spinal cord damage, which could be mediated by the ReST ( Kumru, Vidal et al. 2008; Kumru Kofler et al. 2009) We did not find a significant difference in perceived startle between the individuals with iSCI and the ablebodied control participants. However, when looking at raw startle values, individuals with SCI consistently reported hi gher scores. The lack of betweengroup differences could be driven by the large amount of within group variability. Within group variability could, apart from personal factors, in part be due to how participants reported the startle. All participants were instructed to rate how startling they experienced each sound. However, several individuals provided feedback in the end that indicated that they rated the perceived intensity of the sound (volume) rather than their perceived startle. When looking at the v alues that were normalized to the perceived startle during the reference stimulation, individuals with SCI had lower values. This could be due to
141 the fact that their perceived startle to the reference stimulation was larger. Indeed, the average perceived s tartle of all individuals with SCI to the reference stimulation was 8.22 (std 1.30), while the mean value of the ablebodied control participants was 6.44 (std 1.24). Finally, it is important to note that perceived startle did not always coincide with the presence or absence of EMG responses. This is illustrated for example in the first control participant, who rated the perceived startle at 0 for the last 4 stimulations in the second session, but still showed repeated EMG responses in the lower extremities while standing or standing on foam. On the other hand, the seventh control participant had very few recordable EMG responses, but consistently rated the perceived startle intensity at 7 over all trials. We have proposed a protocol that uses change in pos itions to decrease habituation of evoked responses to acoustic startle. The goal of this protocol and associated research was to enhance stability of ASR testing in individuals with SCI This in turn could render the ASR more useful to assess plastic chang es in the ReST Our results indicate that habituation was less pronounced when compared to values previously presented in the literature. We recommend that if possible, ASR be assessed while participants are standing or standing on foam if the goal is to increase the number of stimulations that lead to a response. Our data also indicate that when using the ASR as a longitudinal test, long term habituation might have to be considered and initial stimulations might have to be discarded to r each stable results. This study further raised additional questions for future research. We recommend that in individuals unable to stand on their own, the effect of standing with the
142 assistance of a body weight support harness and trainers should be investigated. Being able to obtain stable results in those individuals would allow using the ASR in an early stage after SCI. Future research should also investigate the relationship between parameters of the ASR and functional abilities of individuals with SC I to be able to determine if increased responses are related to functional improvement. Further, more knowledge about the periods of EMG silence in response to an acoustic startle could provide more useful insight.
143 4.6 Tables and Figures Table 41. Inclusion and exclusion criteria for both ablebodied control participants and individuals with SCI for the ASR experiment. Able bodied controls Individuals with SCI Inclusion Criteria Adults age and sex matched to individuals with SCI Adults of at least 18 years of age Incomplete SCI with the ability to stand independently and repeatedly over the testing session lasting up to one hour (using a harness for safety reasons only) Able to provide written informed consent Able to provide written informed consent Exclusion Criteria History of neurological disease Neurological diseases other than the SCI History of hearing loss that would compromise testing (able to hear 1000Hz tones) History of hearing loss that would compromise testing (able to hear 1000Hz tones) History of psychological disorders such as schizophrenia, diseases from the autism spectrum, anxiety disorders or depression History of psychological disorders such as schizophrenia, diseases from the autism spectrum anxiety disorders or depression Pain or any orthopedic disease that would prevent the participant from following study procedures Pain or any orthopedic disease that would prevent the participant from following study procedures
144 Table 42. Demographic information Individuals with iSCI Mean(STD) Able bodied control participants Mean(STD) Sex 3 females 3 females Age 47.3y (19.1y) 48.9y (18.4y) Time since injury 10.4y (15.5y); range 1.5 to 51y Etiology 6 traumatic, 1 tumor, 1 epidural hematoma, 1 AVM rupture Levels of injury 8 cervical from C1 to C8 1 thoracic: T3 AIS D for all participants UEMS 44.7 (5.1) LEMS 45.3 (3.2) Handedness All right handed before as well as after the injury 8 right handed, 1 left handed AIS: American Spinal Injury Association Impairment Scale, UEMS: Upper Extremity Motor Score, LEMS: Lower Extremity Motor Score, y: years, AVM: Arteriovenous malformation, C: cervical injuries, T: thoracic injuries
145 Table 43. Pattern of elicited EMG responses in reaction to an acoustic startle in one participant with iSCI with low response rates Supine Standing Foam Total/muscle LOO XXXXX 5/5 OXXXX 4/5 XXXXX 5/5 14/15 93.3% ROO XXXXX 5/5 OXXXX 4/5 XXXXX 5/5 14/15 93.3% LSC OOOOO 0/5 OOOOO 0/4 XXXXO 4/5 4/14 28.6% RSC OOOOO 0/5 OXXOX 3/4 XXXXO 4/5 7/14 50.0% LBB OOOOO 0/5 OOOOO 0/4 XXXXO 4/5 4/14 28.6% RBB OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% LTB OOOOO 0/5 OOOOO 0/4 XXXXX 5/5 5/14 35.7 % RTB OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% LFF XXOOX 3/5 OOOOO 0/4 OXXXX 4/5 7/14 50.0% RFF OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% LFE OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% RFE OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% LRA OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% RRA OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% LES OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% RES OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% LGM OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% RGM OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% LVM OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% RVM OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% LMH OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% RMH OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% LTA OOOOO 0/5 OXOOO 1/4 XXXOO 3/5 4/14 28.6% RTA OOOOO 0/5 OOOOO 0/4 XXXXO 4/5 4/14 28.6% LSol OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% RSol OOOOO 0/5 OOOOO 0/4 OOOOO 0/5 0/14 0.0% Total/ condition 3/120 2.5% 4/96 4.2% 28/120 23.3% 35/336 8.9% L: Left; R: Right; OO: Orbicularis Oculi; SC: SternoCleido Occipito Mastoideus; BB: Biceps Brachii; TB: Triceps Brachii; FC: Flexor Carpi Radialis; EC: Extensor Carpi Radialis; RA: Rectus Abdominis; ES: Errector Spinae; GM: Gluteus Maximus; VM: Vastus Medialis; MH: Medial Hamstrings; TA: Tibialis Anterior; Sol: Soleus; X: a response was evoked in this trial; O: no response was evoked in this trial. OO mu scles were used as a control, but not included in averages over all muscles. The first stimulation in standing was not included in the averaging process because no response could be elicited in either OO muscle.
146 Table 44. Pattern of elicited EMG responses in reaction to an acoustic startle in one participant with iSCI with high response rates. Supine Standing Foam Total/muscle LOO XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% ROO XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% LSC XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% RSC XOXXX 4/5 XOXXO 3/5 XXXXX 5/5 12/15 80.0% LBB XXXXX 5/5 XXXXX 5/5 XXXOX 4/5 14/15 93.3% RBB XXXXX 5/5 XXXXX 5/5 XXOXX 4/5 14/15 93.3% LTB XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% RTB XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% LFF OOOOO 0/5 XXXXX 5/5 XXXOX 4/5 9/15 60.0% RFF XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% LFE XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% RFE XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% LRA OXXXX 4/5 XXXXX 5/5 XXXOX 4/5 13/15 86.6% RRA XXXXX 5/5 XXXXX 5/5 XXXOO 3/5 13/15 86.6% LES OXXXX 4/5 XXXXX 5/5 XXXOX 4/5 13/15 86.6% RES XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% LGM OOOOO 0/5 XXXXX 5/5 XXXXX 5/5 10/15 66.6% RGM OXXXX 4/5 XXXXX 5/5 XXXOX 4/5 13/15 86.6% LVM XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% RVM XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% LMH XXXXX 5/5 OOOOO 0/5 OOOOO 0/5 5/15 33.3% RMH OXXXX 4/5 XXXXX 5/5 XXXOX 4/5 13/15 86.6% LTA XXXXX 5/5 XXXXX 5/5 XXOOX 3/5 13/15 86.6% RTA XXXXX 5/5 XXXXX 5/5 XXXOX 4/5 14/15 93.3% LSol XXXXX 5/5 OOOOO 0/5 OOOOO 0/5 5/15 33.3% RSol OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% Total/ condition 100/120 83.3% 103/120 85.8% 93/120 77.5% 296/360 82.2% L: Left; R: Right; OO: Orbicularis Oculi; SC: SternoCleido Occipito Mastoideus; BB: Biceps Brachii; TB: Triceps Brachii; FC: Flexor Carpi Radialis; EC: Extensor Carpi Radialis; RA: Rectus Abdominis; ES: Errector Spinae; GM: Gluteus Maximus; VM: Vastus Medialis; MH: Medial Hamstrings; TA: Tibialis Anterior; Sol: Soleus; X: a response was e voked; O: no response was evoked. OO muscles were only used as a control and not included in the totals per condition.
147 Table 45. Pattern of elicited EMG responses in reaction to an acoustic startle in one ablebodied control person with low response rates. Supine Standing Foam Total/muscle LOO XXXXX 5/5 XXOXX 4/5 XXXXX 5/5 14/15 93.3% ROO XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% LSC OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% RSC OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% LBB OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% RBB OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% LTB OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% RTB OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% LFF OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% RFF OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% LFE OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% RFE OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% LRA OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% RRA OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% LES OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% RES OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% LGM OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% RGM OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% LVM OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% RVM OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% LMH OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% RMH OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% LTA OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% RTA OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% LSol OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% RSol OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% Total/ condition 0/120 0/120 0/120 0/360 0.0% L: Left; R: Right; OO: Orbicularis Oculi; SC: SternoCleido Occipito Mastoideus; BB: Biceps Brachii; TB: Triceps Brachii; FC: Flexor Carpi Radialis; EC: Extensor Carpi Radialis; RA: Rectus Abdominis; ES: Errector Spinae; GM: Gluteus Maximus; VM: Vastus Medialis; MH: Medial Hamstrings; TA: Tibialis Anterior; Sol: Soleus; X: a response was evoked; O: no response was evoked. OO musc les were only used as a control and not included in the totals per condition.
148 Table 46. Pattern of elicited EMG responses in reaction to an acoustic startle in one able bodied control person with high response rates. Supine Standing Foam Total/muscle LOO XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% ROO XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% LSC XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% RSC XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% LBB OXXXO 3/5 XXXXX 5/5 OXXOX 3/5 11/15 73.3% RBB OOOOO 0/5 OOOOO 0/5 OXXOX 3/5 3/15 20.0% LTB OXXXO 3/5 OXXXX 4/5 OXXOX 3/5 10/15 66.7% RTB OXXXO 3/5 XXXXX 5/5 XXXOX 4/5 12/15 80.0% LFF OOOOO 0/5 XXXXX 5/5 XXXOX 4/5 9/15 60.0% RFF OOXXO 2/5 OXOOO 1/5 OXXOX 3/5 6/15 40.0% LFE OXXXO 3/5 XXXXX 5/5 XXXXX 5/5 13/15 86.7% RFE XOXXO 3/5 XOXXX 4/5 OXXOX 3/5 10/15 66.7% LRA XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% RRA XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% LES XXXXX 5/5 XXXXX 5/5 OXXXX 4/5 14/15 93.3% RES XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% LGM XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% RGM XXXXX 5/5 XXXXX 5/5 XXXXX 5/5 15/15 100.0% LVM XXXXO 4/5 OXXXX 4/5 XXXXX 5/5 13/15 86.7% RVM XXXXX 5/5 XXXXX 5/5 XXXOX 4/5 14/15 93.3% LMH XXXXX 5/5 OXXXX 4/5 XXXOO 3/5 12/15 80.0% RMH OXXXO 3/5 OXXXX 4/5 XXXOX 4/5 11/15 73.3% LTA OOXOO 1/5 XOXXX 4/5 OXXOO 2/5 7/15 46.7% RTA OOXOO 1/5 OXXXX 4/5 OOOXX 2/5 7/15 46.7% LSol OOXXX 3/5 OOOXO 1/5 OOOXX 2/5 6/15 40.0% RSol OOOOO 0/5 OOOOO 0/5 OOOOO 0/5 0/15 0.0% Total/ condition 79/120 65.8% 95/120 79.2% 89/120 74.4% 263/360 73.1% L: Left; R: Right; OO: Orbicularis Oculi; SC: SternoCleido Occipito Mastoideus; BB: Biceps Brachii; TB: Triceps Brachii; FC: Flexor Carpi Radialis; EC: Extensor Carpi Radialis; RA: Rectus Abdominis; ES: Errector Spinae; GM: Gluteus Maximus; VM: Vastus Medialis; MH: Medial Hamstrings; TA: Tibialis Anterior; Sol: Soleus; X: a response was evoked; O: no response was evoked. OO muscles were only used as a control and not included in the totals per condition.
149 Figure 41 Short term habituation in individuals with SCI (panel A, orange) and able/bodied control participants during their first assessment session (panel B, blue). Percentages of r esponse were averaged over all 12 muscle pairs for each stimulation.
150 Figure 42 Long term habituation in able bodied control participants Panel A) First testing session, panel B) second testing session. Percentages of r esponse were averaged over all 12 muscle pairs for each stimulation.
151 Figure 43 Differential effect of position on the average percent of responses in flexor and extensor muscles. Panel A) Individuals with iSCI. Panel B) Ablebodied control participants. Flexor muscles include: SCOM, RA, BB, FF, MH and TA; extensor muscles include: ES TB, FE, VM and Sol Means and STDs are presented.
152 Figure 44 Differential effect of position on the average percent of responses in trunk and extremity muscles Panel A) Individuals with iSCI Panel B) Ablebodied control participants. Trunk muscles include: ES and RA; upper extremity (UE) muscles include: FF and FE; lower extremity (LE) muscles include TA and Sol. Means and STDs are presented.
153 Figure 45 Differential effect of position on the average percent of responses in proximal and distal muscles. Panel A) Individuals with iSCI. Panel B) Ablebodied control participants. Proximal muscles include: BB, TB, GM and VM; distal muscles include: FF, FE, TA and Sol. Means and STDs are presented.
154 Figure 46 Example of ASR responses from an individual with iSCI that showed generally low response rates to acoustic startle. Data from the left OO (panel A), left SCM (panel B) and left TA (panel C) muscle are presented. 5 stimulations in each position are shown. T he first stimulation of each block is presented in black and stimulations 24 are presented in consecutively lighter shades of grey
155 Figure 47 Example of ASR responses from an individual with iSCI that showed generally high response rates to acous tic startle. Data from the left OO (panel A), left SCM (panel B) and left TA (panel C) muscle are presented. 5 stimulations in each position are shown. The first stimulation of each block is presented in black and stimulations 24 are presented in consecut ively lighter shades of grey
156 Figure 48 Example of ASR responses from an ablebodied control participant that showed generally low response rates to acoustic startle. Data from the left OO (panel A), left SCM (panel B) and left TA (panel C) muscle are presented. 5 stimulations in each position are shown. The first stimulation of each block is presented in black and stimulations 24 are presented in consecutively lighter colors.
157 Figure 49 Example of ASR responses from an ablebodied control participant that showed generally high response rates to acoustic startle. Data from the left OO (panel A), left SCM (panel B) and left TA (panel C) muscle are presented. 5 stimulations in each position are shown. The first stimulation of each block is pres ented in black and stimulations 24 are presented in consecutively lighter colors.
158 Figure 41 0 Influence of sex on percentage of response in individuals with iSCI. Mean percentages of response over all muscles other than the OO were plotted against sex for the three females and 6 males with iSCI and the control persons. Please note that due to the small groups size in females, the median and the endings of the whiskers directly indicate the values of the three participants each.
159 Figure 41 1 Influence of age on percentage of response in individuals with iSCI. Mean percentages of response over all muscles other than the OO were plotted against age for the participants with iSCI (orange, R2=0.119) and the ablebodied control part icipants (blue, R2=0.001).
160 Figure 41 2 Relationship between ASR responses rates and motor scores in individuals with SCI Mean percentages of response over all muscles other than the OO were plotted against their upper extremity (UEMS, panel A) or lower extremity (LEMS, panel B) motor score.
161 Figure 41 3 Example of reduced EMG activity after ASR stimulation. A) Stimul us artifact, B) rectified EMG response of the left GM muscle in a participant with SCI. Responses to 5 stimulations collected while the participant was standing were rectified and superimposed. A period of reduced EMG activity is clearly visible starting 7 5ms after stimulation onset and lasting for about 50ms.
162 Figure 41 4 Relationship betweensubject ively perceived startle and EMG responses. Mean percentages of response over all muscles other than the OO were plotted against average values of perceived startle for the participants with iSCI (orange, R2=0.037) and the ablebodied control participants (blue, R2=0.207).
163 Figure 41 5 Subjectively perceived degree of startle to an acoustic stimulation in individuals with SCI (orange) and able bodied control participants (blue) during the first testing session. Perceived startle was rated on a scale from 010 (0 being no startle at all and 10 being most imaginable startle) and averaged over the three positions per block and all 9 participants per group, error bars represent standard deviations. Panel A) shows raw indicated values, panel B shows values that were normalized to the value indicated during the reference stimulation.
164 Figure 416. Subjectively perceived degree of startle to an ac oustic stimulation in ablebodied control participants during the first (blue) and second testing session (green). Perceived startle was rated on a scale from 010 (0 being no startle at all and 10 being most imaginable startle) and averaged over the three positions per block and all 9 participants, error bars represent standard deviations. Panel A) shows raw indicated values, panel B shows values that were normalized to the value indicated during the reference stimulation.
165 Figure 417. Subjectively perceived degree of startle to an acoustic stimulation with regards to position in individuals with iSCI (A), ablebodied control participant during the first testing session (B) and during the second testing session 48hrs later (C). Perceived startle was rated on a scale from 010 and values were normalized to the value indicated by the participant during the reference stimulation prior to the 15 test stimulations. Data points represent mean values over 9 participants each.
166 CHAPTER 5 PROJECT III INDUCING PLASTICITY IN DESCENDING PATHWA YS THROUGH ADAPTABILITY LOCOMOT OR TRAINING 5.1 Background Since the late 1990s, rehabilitation approaches for individuals with iSCI have shifted from primarily emphasizing the use of compensation (e.g. braces, devices, alternate movement strategies) to promoting neuromuscular activation below the level of the lesion to enhance recovery of function ( Edgerton, Guzman et al. 1991; Dobkin 1993; Edgerton, Leon et al. 2001; Behrman, Bowden et al. 2006; Edgerton, Kim et al. 2006; Rojas Vega, Abel et al. 2008) Basic locomotor training ( B asicLT) is a rehabilitation approach to promote the recovery of walking after iSCI lesion ( Barbeau, Danakas et al. 1993 ; Dobkin, Apple et al. 2003; Behrman, Lawless Dixon et al. 2005 ; Harkema, Behrman et al. 2011) A body weight supported treadmill system (BWST) is often used as a controlled and permissive trai ning environment During intense, repetitive stepping, t rainers can continually assess the effect of the intervention and assure that locomotor training principles are being followed ( Behrman and Harkema 2007; Harkema, Behrman et al. 2011) These principles include maximal weight bearing through the lower extremities, optimized sensory cues through manual contact by the trainers to promote upright posture and normal stepping mechanics and the minimization of compensatory in favor of more recovery based strategies. Adhering to these principles is argued to enhance sensory input from the lower extremities and trunk to continuously be sent to the spinal cord during repetitive walking practice. It is believed that over time, repetitive prac tice promotes neuroplastic changes and optimizes the function of the spinal cor d below the level of the lesion ( Barbeau, Danakas et al. 1993 ; Dobkin, Apple et al. 2003; Behrman, Lawless Dixon et al. 2005 ;
167 Harkema, Behrman et al. 2011) This strategy has proven successful to help some individuals with iSCI recover basic overgr ound stepping patterns Specifically, LT improves aspects of gait that are strongly influenced by sensory input and m ainly dependent on networks in the spinal cord (central pattern generator, CPG) such as plantarflexor and knee and hip extensor muscle act ivity necessary for weight bearing ( Wernig, Muller et al. 1995; Behrman and Harkema 2000; Barbeau 2003; Yang and Gorassini 2006 ; Behrman, Nair et al. 2008; Harkema, Schmidt Read et al. 2011) However recovery of upright human bipedal walking is not possible by sol ely increasing the function of the spinal networks below the lesion. Rather it also requires descending input especially in muscles primarily controlled by supraspinal centers such as the ankle dorsiflexors ( Capaday, Lavoie et al. 1999; Capaday 2002; Yang and Gorassini 2006) T herefore, engaging supraspinal centers during training and promoting neuroplastic changes in their connections to the spinal networks below the lesion could lead to improved outcomes and a potentially greater response to locomotor training Even more important successful community ambulation requires not only basic stepping, but also being able to keep ones balance and to adapt limb trajectories to environmental conditions such as negotiating obstacles or stairs ( Patla and Shumway Cook 1999; Shumway Cook, Patla et al. 2002 ; Shumway Cook, Patla et al. 2003; Musselman and Yang 2007) For t hese more challenging, socalled adaptive locomot or tasks, descending control from supraspinal centers to modulate the patterned output by the CPG is critical. Training paradigms stimulating neuroplasticity in these descending systems, therefore, could lead to meaningful improvements in these challenging
168 adaptive walking tasks. This in turn could lead to a greater ability of individuals with iSCI to master challenges they encounter during everyday community ambulation. F ew studies have focused on retraining walking beyond basic step training and the effe ct of training of adaptive walking tasks after spinal cord injury is not well known. Mussleman et al. ( Musselman, Fouad et al. 2009) compared the effect of basic locomotor training (B asicLT ) on a treadmill to skilled locomotor training overground in four individuals with iSCI They concluded that the skill training overground led to greater improvements in walking speed, walking endurance and in a battery of functional walking adaptability tasks. However, due to fear of falling in the absence of a safety harness, practicing some of the more difficult overground walking tasks was avoided by several participants possibly decreasing their response to the int ervention. In addition, orthoses and assistive devic es as well as physical assistance were used, as needed, during the skilled training sessions. These factors however promote compensation, such as weight loading through the upper extremities, and therefore may have decreased the sensory stimuli that promo te recovery of function below the level of the lesion ( Lunenburger, Bolliger et al. 2006; Behrman and Harkema 2007) We suggest that both of these challenges (fear of falling and reliance on compensatory approaches) can be addressed when adaptive tasks are trained in the same safe and permissive BWSTT environment while observing the same locomotor principles that have proven successful to retrain basic stepping In this environment, challenging tasks can be practi ced repeatedly without fear of falling and with appropriate weight loading through the trunk and lower instead of upper extremities This may lead to improved outcomes and a more thorough exploitation of the recovery
169 potential. The primary aim of this st udy, therefore, is to establish and test a training protocol that makes use of the safe and permissive BWST environment during practice of adaptive locomotor tasks. We compare the effect of 15 sessions of adaptive locomotor training (AdaptLT) to 15 sessions of BasicLT on basic stepping ability and stepping adaptability in individuals with iSCI that are able to take steps, but are challenged by higher skilled (adaptive) walking tasks. We hypothesize that the beneficial effects of training will be apparent across two primary domains of the International classification of functioning, disability and health, body structure and function and activity ( WHO 2001; WHO 2002) Specifically, t o examine changes in body structure and function, we assess ed neurophysiological function below the level of the lesion as well as within descending spinal pathways associated with walking function To assess changes at an activity level we included a battery of specifically designed adaptive walking tasks and standardized clinical assessments The findings of this study will contribute to the development of optimal training modalities and parameters to optimize neuroplasticity and improv e walking function in individuals with iSCI. 5.2 Methods 5.2.1 Participants Individuals over 18 years of age with a first occurrence of iSCI that are full time ambulators using one or two canes crutches or no assistive device were. Participants were elig ible if the event that le d to the SCI occurred at least 12 months previous if they suffered from an upper motor neuron lesion at a cervical or thoracic level and if they were medically stable. Individuals were asked to maintain their dosages of anti spast icity medication stable throughout the course of the study. Individuals were not
170 eligible to participant in this study if they were currently participating in a rehabilitation or research protocol that could interfere or influence the outcome of this study if they had a history of congenital SCI or degenerative spinal disorders or if they could not safely participate in the training protocol. Additionally, for each test, t hree able bodied control participants were recruited to obtain normative values. Since not all control participants completed all the testing, a total of 6 persons were included. 5.2. 2 Training Three p articipants with iSCI underwent two blocks of 15 training sessions each, which included one block of AdaptLT and one block of BasicLT i n a pseudorandomized order. Each training session, regardless of the block, included 30min stepping time on the treadmill and 10min stepping overground. A frequency of 5 trainings per week was targeted with each participant. Both training blocks integrate d the principles of LT described by Harkema et al. ( Harkema, Behrman et al. 2011) These principles are: maximal weight bearing through the lower extremities (minimizing body weight support as possible) providing optimal sensory cues, optimize kinematics and maximize recovery of preinjury kinematics while minimizing compensatory mechanisms. Both blocks included periods of retraining during which manual assistance by the therapists was provided to optimize kinem atics while walking at high walking speeds and, if necessary to attain proper kinematic walking patterns, increased BWS The goal of these bouts was to maximize sensory input to the spinal cord from the periphery. These periods of retraining were complemented in both training blocks by periods of independence training, during which manual assistance as well as BWS was continuously decreased ( Harkema, Behrman et al. 2011) Speed during those training bouts was decr eased as needed to allow the participant to take steps (both basic steps
171 as well as steps adapted to objects or other environmental demands) independently with the goal to increase engagement of supraspinal centers Pulse and blood pressure were monitored in regular intervals. During BasicLT, participants practiced basic stepping while concentrating on a ppropriate kinematics. Gait deviations were identified for each individual participant and goals were set to recover preinjury trajectories and eliminate the use of compensatory strategies. Further goals during this block were increasing walking speed and endurance. During AdaptLT, participants practiced various adaptive tasks according to their deficits. Tasks most often practiced included stepping over objects, carrying objects while walking, talking or reading while walking, walking sideways or backwards, changing speed, changing step length, walking in shoes with soft foam attached to the sole to mimic walking on soft grounds, walking with high heeled shoes or stepping while being bumped into. An example of how we individualized training for each participant is that participant SCIAdapt01 mentioned being afraid to walk her dogs because she did not feel stable enough to resist their pull. This was addressed by having the participant bring a dog leash to training and the trainers pulling on the leash into multiple random directions during walking. Training focused on feedforward tasks, meaning that the element of surprise was held as minimal as possible. For example, pulls on the leash were not unexpected, nor was the presentation of objects on the treadmill. Rather, the participant was warned by counting backwards to the event to allow them to prepare for the task to be mastered.
172 5.2. 3 Testing A comprehensive battery of tests at the body structure, body function and activity domains of health was chosen for this pilot study in order to thoroughly explore changes in the ne u r o muscul o skeletal system in response to the different types of training. The assessments included neurophysiological assessments of the descending spinal tracts and the excitability of the spinal cord, clinical assessment tools as well as instrumented gait testing during basic and adaptive walking tasks. Testing for individuals with iSCI was performed at baseline, between the two training blocks and after the final training session. Control parti cipants were measured once to obtain normative values. Table 5 1 provides an overview over the tests undergone by the ablebodied control participants. 5.2. 3.1 Measures of body structure and function Motor evoked potentials (MEP) central conduction time (CCT) and ipsilateral silent periods (iSP) To test the functional integrity of the corticospinal tract (CST), MEPs were evoked by transcranial magnetic stimulation (TMS) applied over the leg representation of the primary motor cortex. EMG s were recorded f rom the right and left first dorsal interosseus (FDI) and the right and left tibialis anterior (TA) muscles. Motor thresholds and recruitment curves were obtained. Experimental motor thresholds are defined as the stimulation intensity level at which at least 5 out of 10 stimulations lead to EMG responses larger than 50 V in peak to peak amplitude ( Borckardt, Nahas et al. 2006) Recruitment curves were fit with a Boltzmann equ ation and the slope of the linear fit curve of the steepest part of the Boltzmann fit was used to determine the actual motor threshold by determining the stimulation level at which the slope of the linear fit line intersects with the abscissa ( Perez, LundbyeJensen et al. 2006 )
173 F wave s were obtained and central motor conduction times were calculated according to Rossini et al. ( Rossini and Pauri 1999) to detect changes in conduction time of the descending spinal tracts, specifically the corticospinal tract. In addition, as a measure of function of the periph eral nervous system, latencies and persistenc y of F waves are presented. We assessed the duration of the ipsilateral silent period ( iSP ) as the time during which EMG activity falls below mean baseline EMG by more than 1 standard deviation (iSP onset) after stimulation of the ipsilateral cortex until the EMG was visually evaluated to recover to a stable level clearly distinguishable from the iSP period (iSP offset) Percentage of inhibition is calculated as difference between the area calculated by multiplying the duration of the ISP by the pre stimulus mean (total block area) and the area calculated by integrating the area under the curve of the ISP (iSP area) ( Chen 2004) H reflex. H reflex es a measure of spinal excitability, were obtained from the soleus muscles. Participants were seated with their legs supported on leg rests while their tibial nerve was stimulated in the fossa poplitea at 0.3Hz using a square wave of 0.5 1ms. Responses were recorded with surface electrodes placed on the belly of the soleus muscle Stimulation intensities were increased in 0.5mV increments per stimulation starting at subthreshold levels to levels at which a maximal M wave was elicited. Latencies were calculated as the time between stimulation and onset of the H reflex M wave and H reflex amplitudes were pl otted against normalized stimulation intensit ies A ratio between the slope of the linear portion of the H reflex recruitment
174 curve and the slope of the linear portion of the M wave recruitment curve was computed ( Funase, Higashi et al. 1996; Phadke, Robertson et al. 2012) Vestibulos pinal Reflex (VSR). To test the functional integrity of the vestibulospinal tract, EMG w as recorded from the right soleus muscle and body sway in response to galvanic vestibular stimulation over the mastoid processes w as measured. Participants were standi ng on a foam pillow ( Airex Balance Pad, 47cm x 38.5cm x 7cm ) during testing with their head turned either to the left or right side (one trial each) arms crossed and eyes closed. This protocol has been show to result in the largest and most reliable response ( Wydenkeller, Liechti et al. 2006; Liechti, Muller et al. 2008) Onset latencies of the displacement of the center of pressure and onset of medium latency EMG responses of the soleus muscle are reported as described by Liechti et al. ( Liechti, Muller et al. 2008) Acoustic Startle Response ( ASR ) To test the functional integrity of the reticulospinal tract, EM G activity in seven muscle pairs (SternoCleido Mastoideus, Biceps Brachii, Errector Spinae, Vastus Medialis, Medial Hamstring, Tibialis Anterior, Soleus) in reaction to a loud, unexpected acoustic startle was measured. The startling sound was delivered th rough headphones and consisted of a 120ms long train horn. Participants were either supine or standing on solid ground The room was quiet and participants were encouraged to relax but try not to fall asleep ( Kumru, Vidal et al. 2008) Presence or absence of a response was decided from visual inspection for two stimulations in each of the two positions. The occurrence of responses for each of the muscles is reported.
175 International Standards for Ne urological Classification of Spinal Cord Injury ( ISNCSCI ) Assessments of neurological function were conducted according to the guidelines by the American Spinal Injury Association (ASIA) ( ASIA 2011 ) This assessment tests motor and sensory function (pin prick as well as light touch) in each my otome and dermatome respectively. It was used to characterize each individuals injury and motor and sensory impairment but was not used as an outcome tool to assess change over time. The free online calculator www.ais.emsci.org ( Schuld, Wiese et al. 2011) was used to define levels, motor and sensory scores and impairment levels. The ISNCSCI recommends testing of proprioception as an additional test. We therefore added proprioceptive testing of the lower extremities to this exam. 5.2. 3 .2 Measures of a ctivity Spatiotemporal data and foot trajectories during basic stepping and while performing adaptability tasks overground. Assessments w ere conducted while participants were both walking overground and on a treadmill. Data in this dissertation focus on results collected from overground testing. All trials were performed over a distance of approximately 15meters. Participants were instrumented with reflective markers and trajectories of the heel and tip of the second toe were recorded using a 12camera infrared system (Vicon Motion Systems Ltd). The camera setup allowed tracking of trajectories for approximately the middle 5m of the 12m walk way. During the first six trials, individuals walked at a self selected speed (three trials) and at their fastest comfortable speed (three trials) ( Wolf, Catlin et al. 1999; van Hedel, Dietz et al. 2007 ; Jackson, Carnel et al. 2008 ) without any obstacles present. Subsequently, for three trials each, individuals stepped over a 6 inch (6 inches wide and 2 inches deep) or a 10 inch high box (6 inches wide and 2 inches deep) placed halfway
176 al ong the walkway. The object was positioned on the left side of the walkway; thus, individuals only stepped over the object with the left (more involved) foot. Individuals were asked to attempt to clear the box wi th a minimal clearing distance. Spatiotemporal parameters such as gait speed, cadence, step and stride length, step width and single and double limb stance in percent of the whole gait cycle were calculated. Further, the minimal clearing distances for both heel and toe w ere calculated a s minimal distance between the top edge of the box and the heel marker and the top edge of the box and the toe marker respectively. Lastly, the distance between the toe position at toe off and the near edge of the box before stepping over the box and the d istance between the heel and the far edge of the box at heel strike were defined as measure of horizontal precision. Clinical Assessments. The following clinical assessments were conducted: the Walking Index for Spinal Cord Injury II (WISCI II) ( Ditunno, Ditunno et al. 2000; Dittuno and Dittuno 2001) the Spinal Cord Injury Functional Ambulation Profile (SCI FAP) ( Musselman, Brunton et al. 2011) as well as the items focusing on trunk and lower extremity function from the Neuromuscular Recovery Scale (stand retraining, stand adaptability, step retraining, step adaptability, sit, reverse sit up, sit up, trunk extension in sitting, sit to stand, stand and walking) ( Harkema, Behrman et al. 2011; Behrman, Ardolino et al. 2012) Participants also completed the Activity specific Balance Confidence Scale (ABC), a questionnaire assessing their subjective con fidence in their walking skills ( Powell and Myers 1995 ; Myers, Fletcher et al. 1998; Lajoie and Gallagher 2004) Subsequent to each testing session, individuals wore a step activity monitor on
177 their less involved ankle for two weekdays and two weekend days to quantify number of steps taken daily in their familiar environment ( Bowden and Behrman 2007) 5.3 Results 5.3.1 Demographics Three females with iSCI and six ablebodied females Demographic information for all participants is provided in Tables 5 2 and 53 The study protocol was completed within 9 to 15 weeks by the participants. Adjustments to the training and testing schedule were due to F ederal holidays, vacations and work obligations by the participants Participant SCIAdapt04 had an unforeseen 7week break between the two training blocks, therefore an additional testing session was conducted before beginning the second training block. Figure 51 presents an overview of the dispersion of training and testing sessions for the three participants 5.3.2 Completeness of D ata Not all assessments could be performed by all participants. Reasons for missing data are diverse. One test (SCI FAP) was only included into the testing battery after the first participant had already been tested. Participant SCIAdapt0 4 was unable to attend two full days of testing due to work obligations, therefore the test battery was reduced. Further, technical issues, for example a complete computer outage during a testing session, lead to missing data.
178 5.3. 3 Measures of B ody F unction and S tructure 5.3. 3 .1 TMS and CCT and iSP Excitability and inhibition of the motor cortex and conduction along the CST Parameters obtained from transcranial magnetic stimulation were highly variable and no clear patterns with regard to the type of training were evident regardless of the parameter observed. For example in participant SCIAdapt04, a decrease i n the motor threshold took place when stimulating the left hemisphere and recording the MEP from the right TA muscle between pre and mid testing, indicating an increase in neural excitability. During the training free period, this threshold further diminis hed, but then increased again during the second training block ( please refer to table 54 for participants with SCI and table 55 for ablebodied control participants, as well as figure 5 2 for exemplary recruitment curve data) Similarly variable data w er e available for the CCT. This variability is mostly driven by the differences in MEP latencies, while F wave latencies were more stable (tables 56 and 57 and figure 54 ). Parameters from iSP measurements are presented in table 58 for individuals with SC I and in table 59 for ablebodied control participants. A n example is presented in figure 53. 5.3. 3 .2 H Reflex Excitability of the spinal cord H Reflexes were not consistently obtainable from these individuals with iSCI. When obtainable, they often were not typical with regards to the intensity at which they occurred and with regards to their form. At times, H Reflexes occurred at subthreshold stimulation intensities, but were very heterogeneous in shape. At other times, homogeneous responses occurred around 30ms however, only at suprathreshold
179 stimulation intensities. This led to a difficulty to distinguish H Reflexes from F waves, given that certain characteristics of the response were consistent with typical F waves while others were consistent with typical H reflex responses. In addition, H to M wave ratios were very small, especially when compared to healthy control participants as a result of the small, difficult to detect H Reflexes ( please refer to table 510 for individuals with SCI and table 51 1 for ablebodied controls and figure 55 for an example of H Reflex responses ) 5.3. 3 3 VSR Excitability of the vestibular nuclei and conduct ion along the VST Responses from the distal leg muscles have been elicited most consistently and therefore, results are presented from the TA and soleus muscle contralateral to the side the individual was facing. Data are presented from participant SCIAdapt0 2 (Figure 5 6 ). The participant had undergone 15 sessions of AdaptLT at the time of testing. No A1 latency could be defined in this participant, since the traces of left and right stimulations already deviated before stimulus onset. Latencies and Amplitudes of the responses can be seen in Table 51 2 and 51 3 5.3. 3 4 ASR Excitability of the brainstem areas and conduction along the ReST EMG responses in reaction to an acoustic startle could be recorded from muscles below the level of the injury in all three participants. There was considerable between participant difference when it came to the number of muscles in which responses could be elicited. Within each participant on the other hand, the tendency to respond remained constant over time, regardless of the type of training that took place between two assessment periods (tables 51 4 and 51 5 )
180 5.3. 3 5 ISNCSCI Neurological function a t baseline Results from the ISNCSCI are presented in table 51. All three participants had injuries classified as AIS D, indicating considerable motor function below the level of the lesion was present. For all participants, the left side of the body exhibited greater motor impairments than the right side. One participant had a thoracic injury and therefore no involvement of the upper extremities. Moderately reduced sensitivity to l ight touch was p re sent in all individuals, and a marked reduction in the abi lity to differentiate sharp from dull was noted in SCIAdapt02 on the side opposite of the most pronounced motor deficit. All participants had normal proprioception except for SCIAdapt04 who had impaired proprioception for the digitus I of the left foot. 5. 3 4 Measures of Activity 5.3. 4 1 Spatiotemporal data and foot trajectories Self selected and fastest comfortable walking speed. Individuals increased their self selected overground walking speed between consecutive testing sessions regardless of the type of training provided. Two individuals initially walked more than 1std slower than the mean of the control participants (1. 18 m/s std: 0.14m/s ). By the second testing session, all individuals with iSCI walked faster at their self selected walking speed than the mean of control participants and at post testing, their speed exceeded 1std above the mean of the control participants. In contrast, while two out of three participants with iSCI also improved their fastest comfortable walking speed over the two block of training (with no obvious difference between the different types of training), all remained below the average speed chosen by the three control participants (1.80m/s, std: 0.33m/s) ( figure 57 and table 5 16)
181 Stride length and cadence. Regardless of t he type of training provided between two training sessions, individuals with iSCI increased their stride length at self selected walking speeds continuously. Two out of three participants stared out at stride lengths of 1 standard deviation or more below t he average stride length of healthy control participants (1.25m, std: 0.19m) and reached values above the mean at the second testing session. Further i ncreases were seen during the second training block. The cadence on the other hand initially increased fr om values more than one standard deviation below the mean of healthy controls (114 steps per minute) to values more than one standard deviation above this reference, thus having a higher cadence than able bodied control participants. These values decreased during the second training block to reach values within 1standard deviation of the mean of able bodied controls in 2 out of 3 participants (figure 58 and table 516) Clearing distance of he e l and toe with regards to the box. Data from the left (in all participants more involved) leg stepping over the box es are presented. When stepping over the 6 inch box, i ndividuals with iSCi generally tended to lift their heel higher above the upper edge of the box than able bodied control participants, although with one exception this increase amounted to less than 1std from the mean of ablebodied controls. The clearing distance increased after the block of AdaptLT in all participants and decreased after BasicLT. The toes on the other hand, were not lifted higher in individuals with iSCi than in control participants. There was a tendency to decrease the clearing distance during blocks of AdaptLT, while during BasicLT, they remained the same or increased.
182 Differences between heel and toe clearing distances were larger in individuals with iSCI than in control participants. This was even more pronounced after blocks of Adapt LT than after blocks of BasicLT (figure 5 9 and table 517). These results were not identical to the ones obtained when individuals stepped over a box with 10 inch height No clear pattern depending on the type of training could be detected and the heel to box clearing distance of individuals with iSCI was within 1 standard deviation from the mean of ablebodied individuals with only one exception (fi gure 510 and table 518) Toe clearing distance tended to decrease over both training blocks and was either within one standard deviation above the mean of control participants or below the mean. Horizontal distances between foot and box at toe off and heel strike. No specific pattern could be detected in the change of position of the foot relative to the box at toe off and heel strike of the swing phase stepping over the object. Regardless of the size of the box, SCIAdapt01 seemed to choose a different strategy from the other two participants. When stepping over the 6inch box, this participant lifted the foot further away from the box and landed closer to the box on the opposite side, while over the 10 inch box, the opposite was the case (figures 5 11 and 512) 5.3. 4 .2 Clinical a ssessments All participants reached the maximally possible 20 points at all assessment time points on the WISCI II. Results for the different standardized clinical assessment tools are presented in table 519 and figures 513 to 5 17 Spinal cord injury functional ambulation profile. All individuals showed lower (and therefore improved) SCI FAP scores at the last testing session compared to the first session. For SCIAdapt02 and 04, greater improvements were evident following the
183 block of AdaptLT. No comparison between the improvement s during the two training blocks can be made for SCIAdapt01 because this test was not conducted at baseline with this participant (Figure 5 13) Neuromuscular recovery scale. Each participant improved one level on the NMRS during the first training block regardless of the type of training conducted. During the second training block, results were different. O ne participant improved by another level another one remained constant. The last partic ipant, SCIAdapt04 was special insofar as that this person improved by one NMRS level during the training free period between the two training blocks, but reverted to a level lower again after the second training block which for this patient was the Basic LT block (Figure 5 14) Figure 515 shows visual assessment of kinematics while stepping over a box in a participant at baseline (panel A) and after two weeks of AdaptLT). The participant improved her kinematics, using more hip flexion and less circumducti on. In addition, no manual assistance by the therapist was necessary after two weeks of AdaptLT to safely complete the task and previously pronounced associated movements (consensual involuntary movement s which acco mpany voluntary efforts ( Dun glison and Lathrop Stedman 1903) ) in the left upper extremity were not apparent after two weeks of training Activities specific balance confidence scale. Walking confidence, as assess ed by the ABC improved in two out of three participants over the two training blocks. In the remaining participant, scores improved during both of the training blocks However, there was a pronounced decrease in the scores during the training free period, which contributed to an overall reduction in this participants score on this scale. The
184 confidence according to the ABC of all participants increased during the block of BasicLT, while only one out of three individuals improved their confidence during the block of AdaptLT (Figu re 5 16) Step activity monitor steps taken during daily life. All individuals took fewer st rides ( measured are steps with the right foot) per day as measured on 4 days of their normal daily routine after a training block when compared to before the trai ning regardless of the type of training provided For one participant, only data before and after AdaptLT, but not BasicLT was available (Figure 5 17) 5.4 Discussion In this small pilot study, three adults with iSCI completed two block s of training, 15 sessions of A daptLT and 15 sessions of B asicLT. While it is not clear which block was most beneficial for improving walking for the participants in this small cohort, overall across both blocks, the three individuals improved in several of the selected out come measures. This shows that these high functioning individuals can still progress, even in rather chronic states of between one and a half and seven years after injury, when a relative plateau in recovery has been reached ( Burns, Marino et al. 2012) S elf selected walking speed increased for all partici pants mostly driven by an increase in stride length. This is in accordance with a study by Pepin et al. who found that individuals with iSCI prefer increasing stride length over cadence to adapt to increased speed of the treadmill and have a limited range of cadence frequencies available ( Pepin, Ladouceur et al. 2003) The authors in this study assumed that this is due to the decrease in neural drive to the respective muscles.
185 Participants also improved on certain gait related standardized outcome measures that are used in the clinic and showed improvement of movement strategies towards pre injury approaches as assessed upon visual inspection. The tests of body structure and function were more variable within and betweensubject s. Our neurophysiological tests sought to describe the functional integrity of the nervous system on different levels (supraspinal str uctures, spinal cord white matter, spinal cord grey matter and the peripheral nervous system) and for different descending tracts (cortico spinal, vestibul o spinal and reticulospinal tracts). The variability in results is perhaps a reflection of the diver sity of participants injuries in terms of location and extent of the lesion. Interesting, however, was that the ISNCSCI, a commonly used clinical measure of severity of injury ( ASIA 2011 ) provided very consistent results, assigning an AIS D level to all three participants. This clinical test perhaps did not capture the extent of between participant diversity as much as the neurophysiological tests did. Further, our aim was to compare the change over time of these neurophysiological measures in response to the two different types of training Our hypothesis was that BasicLT, mainly targeting the neural networks in the spinal cord, would lead to relatively more changes within those networks and therefore changes in H Reflex measures while AdaptLT would lead to relatively more changes in descending systems and the associated measurements However, due to the amount of between and within subject variability, no such trends could be detected. D ue to the location of the injury being in the spinal cord, we suspected that a measure of central conduction time could provide interesting information about the
186 integrity and neuroplastic changes within the spinal cord ( Rossini, Barker et al. 1994) Changes in this measure with measures of brain excitability, s uch as the slope of the recruitment curve remaining stable, could help isolate changes to the spinal cord white matter. A decrease in CCT could be evidence for remyelinization processes or potentially strengthening of spared or newly formed synapses. Howev er, our data showed no consistent trend, which was mainly due to variability in MEP latencies. In general, none of the neurophysiological tests we used detected a difference in effect between the two types of training This could either mean that the meas ures were not stable or sensitive enough to detect such a difference, or, on the other hand that there was no differential effect present A lack of difference between the two types of locomotor training could indicate that the 15 sessions of each type of training provided in this pilot study were not sufficient to induce different responses Alternatively, it could be due to the fact that the two types of training were not diverse or specific enough to lead to differential effects in differ ent parts of the nervous system During both training blocks, time spent practicing independence w as high when compared to time spent retraining. This means that even during BasicLT, there might have been a considerable amount of descending tract involvement. Participants were very cognitively engaged during those independence bouts to attempt walking with the best possible kinematics. Finally, this is the first study to describe AdaptLT using the BWST environment and although the training protocol was based on the principles of LT ( Behrman and Harkema 2007 ; Harkema, Behrman et al. 2011) it is possible that adaptations in the AdaptLT protocol are necessary and would lead to more improvements with this type of training.
187 Possibly, the reason for the lack of a differential e ffect of the two training types in this study is due to a combination of above mentioned outcome measure and training protocol related factors. In addition, t he individuals in this study all were full time community ambulators without assistive devices and a WISCI score of 20 ( considerably higher than the participants in the earlier study investigating adaptive locomotor training by Musselman et al. ( Musselman, Fouad et al. 2009) ) These individuals therefore had a high intensity of adaptive walking training already in their home and daily life outside their home. The additional 30min per day of training provided in the framework of this study focused on tasks specifically tailored to their needs. Wh ile they successfully improved their performance on those tasks, such as walking in high heels or walking with a dog on a leash, this improvement might have been very specific and not well reflected in more general assessment tools, even if those tools wer e specifically designed to measure neuromuscular recovery. Besides the neurophysiological outcome measures, there are also other measures that did not discover a differential effect between the two training types as we had expected. The NMRS for example me asures improvement in functional tasks with an emphasis on recovery of preinjury moving patterns rather than just assessing the ability to perform a task regardless of kinematics. Therefore, this scale seems very well suited to measure outcome in a therapy aimed at restoring such preinjury movement patterns. While all participants improved by one level on this score, it might have still underestimated the amount of recovery that took place in the lower extremities, specifically in walking function. While we choose to only assess those items of the
188 NMRS that were related to trunk and lower extremity function, some of the ratings still required upper extremity function. For example, to receive the highest level in the sit item, appropriate kinematics of the scapulae, shoulder, elbows, wrists and fingers are required ( Harkema, Behrman et al. 2011; Behrman, Ardolino et al. 2012) when shifting weight from a sitting position. While appropriate arm swing was included in both BasicLT and AdaptLT, we did not otherwise specifically concentrate our training on upper extremity function. It is therefore not surprising that individuals did not progress in these items and due to the importance of such trailing items when computing the overall NMRS, their recovery towards preinjury movement patterns as could be vis ually assessed ( for an example refer to figure 515 ) was underestimated by this scale. Another clear discrepancy worth mentioning was the difference between participants confidence in their walking skills as assessed by the ABC scale and the informal feedback w e received from them during training and testing sessions This is true despite the fact that the ABC has been shown to be test retest reliable in community dwelling elderly over a two week interval (r=0.92) ( Powell and Myers 1995) Participants in the current study all indicated that their largest personal benefit from both blocks of training, and specifically from AdaptLT was that they gained a substantial amount of walking confidence and now dare to attempt higher skilled walking tasks that they had, in some cases, been avoiding since their injury. For example, participants SCIAdapt01 and 04 both started wearing high heels again, a declared goal at the beginning of the training. Also, Participant 04 went dancing with her husband, indicating that her newly gained confidence allowed her to do so. Participant 02 mentioned leaving
189 her hiking pole at home for a hike on a forest trail and daring to walk across a busy fairground without taking a wheelchair for safety like she used to do. These reported improvements were, however, not reflected in the ABC scores. The idea that there were indeed changes in response to training despite the fact that they are not reflected in neurophysiological or some standardized clinical measurement tools is supported by the fact that certain clinical and spatiotemporal parameters, as well as parameters of qualitative gait analysis did change. For example, people did increase both in their s elf selected walking speed and their stride length and improved their score on the SCI FAP. Numerous factors negatively influenced the completeness and consistency of the data that was collected. One of these factors was t he duration of the data collections. Data collections lasted approximately 8hrs a day, over two days. Individuals were given the opportunity to spread out the assessments over three days if they preferred. However, two out of the three participants with SC I were employed full time and therefore had limited time resources to allocate to testing sessions. Participant availability and fatigue and a very dense testing schedule lead to considerable time pressure. On occasion, certain assessments were not conduct ed due to that reason. In addition, no margin for repetition of any parts of an experiment to improve data quality if deemed necessary was available. However, the large amount of assessments was planned with the intention to acquire a complete picture in three participants and to then be able to choose outcome measures that were most useful for future research. Further refinement of the outcome measur es to be included in a study assessing walking adaptability might lead to more meaningful outcomes in the future. Based on the results
190 of this study, analysis of spatiotemporal parameters during specifically designed adaptability tasks could be a promising way to assess responses to AdaptLT. Potentially, for the specific adaptability tasks (stepping over the box), using more different sizes of obstacles to demand more adaptability within a testing session could better reflect the effect of such a training paradigm. In summary, our training strategy was beneficial in improving certain parameters of walking, which was detected with some of the used outcome measures, but not with others We suggest that test batteries should be further developed and used to determine not only the effect of different training modalities but also to be able to predict which individuals might profit most from the different types of training approaches
191 5. 5 Tables and Figures Table 51. Summary of assessments completed by each of the control participants. TMS, F&H ASR VSR Treadmill Overground SCIAdaptC01 X X SCIAdaptC02 X X SCIAdaptC03 X X SCIAdaptC04 X X X SCIAdaptC05 X SCIAdaptC06 X X X X X TMS, F&H: Transcranial magnetic stimulation, F wave and H Reflex testing, ASR: Acoustic startle response, VSR: Vestibulospinal Reflex testing; Treadmill: Instrumented testing of basic and adaptive locomotor tasks on the treadmill; Overground: Instrumented testing of basic and adaptive locomotor tasks overground.
192 Table 52 Demographic information for i ndividuals with iSCI. SCIAdapt01 SCIAdapt02 SCIAdapt04 Sex female female female Age 53 45 59 Time since injury (years) 1.5 7 2 Type of injury Transverse myelitis Stab wound MVA NLI SLR SLL MLR MLL T1 T1 T2 T1 T2 C4 C4 C5 T1 C5 C2 T10 C2 T1 C2 AIS D D D UEMS R UEMS L 25 25 25 21 24 21 LEMS R LEMS L 25 22 24 17 25 22 LT R LT L PPR PP L 54 48 41 40 42 53 11 36 47 39 46 35 Proprioception: Knee R Knee L Ankle R Ankle L Digitus I R Digitus I L Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Impaired MVA: Motor vehicle accident, NLI: Neurological level of injury, SLR: Sensory level right, SLL: Sensory level left, MLR: Motor level right, MLL: Motor level left, AIS: American Spinal Injury Association Impairment Scale, UEMS R and UEMS L: Upper Extremity M otor Score right and left, LEMS R and LEMSL: Lower Extremity Motor Score right and left (out of 25). LT R and LT L: Right and left Light touch scores (out of 56); PP R and PP L: Right and left Pin Prick scores (out of 56); Sensory and motor exams were con ducted according to the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) ( ASIA 2011 ) Neurological levels, the AIS and motor scores were computed with a free online calculator that implements the rules of the SNCSCI (www.emsci.org) ( Schuld, Wiese et al. 2011) Joint appreciation and position sense were assessed and defined according to the suggestions for optional sensory testing elements by the ISNCSCI ( ASIA 2011 ) .
193 Table 53 Demographic information for able bodied control participants. SCIAdaptC01 SCIAdaptC02 SCIAdaptC03 SCIAdaptC04 SCIAdaptC05 SCIAdaptC06 Sex female female female female female female Age 22 31 26 23 44 53 Adapt OG X X X Adapt TM X X X TMS, F&H X X X VSR X X X ASR X X X This table includes information about which tests were performed by which participant. Adapt OG: Adaptability tests overground, Adapt TM: Adaptability tests on the treadmill, TMS: Transcranial Magnetic Stimulation, F: F wave testing, H: H Reflex testing, VSR: Vestibulospinal Reflex, ASR: Acoustic Startle Response.
194 Table 54 TMS Recruitment curve parameters in individuals with iSCI : this test was not conducted in this participant; *: this test was conducted, but the data did not fit a Boltzman curve; MTexp: experimental motor threshold; MTtheoretical: th eoretical motor threshold; V50: midpoint of Boltzman fit curve; Slope: slope of regression line through steepest part of Boltzman fit curve; MEPMax: Maximal MEP; Int: Stimulator output intensity at MEPMax in % of maximal stimulator output; AN: MEP Area un der the curve normalized to M wave amplitude unless otherwise noted; AR: Area under the curve in V*msec (raw values); LAT: latency in ms; Amp: Amplitude; Area; Area under the curve; R TA: Right Tibialis Anterior; L TA: Left Tibialis Anterior; R FDI: Right first dorsal interosseus; N: Curve fit based on areas under the curves normalized to M wave; R: Curve fit based on raw areas under the curve. For each subjects training block sequence, refer to Figure 51.
195 Muscle SCIAdapt01 SCIAdapt02 SCIAdapt04 PRE MID POST PRE MID POST PRE MID MID RE POST MT exp R TA 73 65 58 58 45 60 60 50 53 L TA 63 64 70 55 72 70 58 R FDI 57 35 37 MT theoretical R TA 69.64 N 78.93 N 55.49 N 56.94 N 55.02 N 64.32 N 52.08 N 37.91 N 49.38 N L TA 53.84 N 47.13 N 43.36 N 44.42 N 35.19 N 81.50 N 84.07 N 86.19 N R FDI 59.8 R 29.76 R 40.52 R 38.35 R V50 R TA 71.48 N 85.17 N 57.96 N 68.51 N 68.40 N 77.64 N 75.03 N 68.88 N 69.79 N L TA 74.93 N 69.50 N 57.33 N 62.49 N 67.64 N 91.36 N 93.16 N 94.46 N R FDI 77.27 R 41.51 R 45.09 R 48.62 R Slope R TA 0.31 0.02 0.01 0.08 0.01 <0.01 <0.01 <0.01 <0.01 L TA 0.01 <0.01 <0.01 0.01 <0.01 0.01 <0.01 0.01 R FDI 439.63 R 557.8 R 3561.3 R 911.36 R MEP Max R TA Int 75 90 75 80 80 90 95 85 65 A N 0.26 0.07 2.15 0.11 0.09 0.15 0.15 0.17 A R 7491.41 1530.09 28691.58 12398.27 4585.08 4248.11 4896.28 7118.89 LAT 25.00 30.00 31.50 29.50 29.00 31.00 31.00 31.00 33.5 L TA Int 70 90 85 80 85 80 100 100 100 A N 0.22 0.09 0.11 0.07 0.07 0.03 0.07 0.11 A R 2896.46 1619.41 5500.91 5292.62 4028.20 1327.82 2123.44 6888.97 LAT 25.00 33.00 32.00 29.50 30.00 32.50 33.00 33.00 31.00 R FDI Int 95 55 55 70 A N A R 15357.32 13309.71 38540.07 18791.26 LAT 24.00 22.50 24.00 23.00 M wave R TA LAT 3.00 3.50 2.00 3.50 3.00 3.00 3.50 2.50 2.50 Amp 3265.00 3095.20 1494.76 12162.20 4460.84 3536.79 4352.38 4912.04 Area 28571.97 21762.30 13374.51 108176.54 50742.43 27750.64 32225.72 41002.91 L TA LAT 3.50 2.00 2.50 3.00 4.00 2.00 3.50 2.50 2.50 Amp 2624.09 2543.02 4510.98 4944.75 5303.15 3669.95 3392.96 6720.85 Area 25550.68 17320.42 50327.83 76569.26 59363.97 47544.71 31364.10 60731.13 R FDI LAT Amp Area
196 Table 55 TMS Recruitment curve parameters in ablebodied control participants. Muscle SCIAdaptC04 SCIAdaptC05 SCIAdaptC06 MT exp R TA 48 38 44 L TA 54 42 R FDI 36 R ? 60 MT theoretical R TA 42.15 N 23.30 N 47.59 N L TA 54.20 N 29.83 N R FDI 48.44 R 40.28 N 59.72 N V50 R TA 49.09 N 53.20 N 58.08 N L TA 68.30 N 44.09 N R FDI 70.635 R 67.43 N 71.16 N Slope R TA <0.01 0.01 0.03 L TA <0.01 0.01 R FDI 486.12 R 0.24 0.01 MEP Max R TA Intensity 65 75 70 A N 0.02 0.09 0.76 A R 1296.58 3259.7 28491.32 Lat 27.50 32.00 30.00 L TA Intensity 70 75 A N 0.02 0.08 A R 1629.13 3259.70 Lat 29.00 32.00 R FDI Intensity 80 63 85 A N 5.16 0.20 A R 20051.31 25671.17 1095.25 Lat 23.00 26.00 25.00 M wave R TA LAT 3.00 2.50 3.50 Amp 5378.87 4206.54 3023.43 Area 61407.55 40322.66 37519.65 L TA LAT 3.50 2.50 Amp 10464.02 4335.36 Area 87916.64 32025.79 R FDI LAT 2.50 3.00 Amp 829.17 1160.38 Area 4972.23 5358.29 : this test was not conducted in this participant; MTexp: experimental motor threshold; MTtheoretical: theoretical motor threshold; V50: midpoint of Boltzman fit curve; Slope: slope of regression line through steepest part of Boltzman fit curve; MEPMax: Maximal MEP; Int: Stimulator output intensity at MEPMax in % of maximal stimulator output ; AN: MEP Area under the curve normalized to Mwave amplitude unless otherwise noted; AR: Area under the curve in V*msec (raw values); LAT: latency in ms; Amp: Amplitude; Area; Area under the curve; R TA: Right Tibialis Anterior; L TA: Left Tibialis Anter ior; R FDI: Right first dorsal interosseus; N: curve fit based on areas under the curves normalized to Mwave; R: curve fit based on raw areas under the curve.
197 Table 56 F wave measurements. Participant Session Right Left SCIAdapt01 PRE Latency 30 29.2 Persistence 70 80 MID Latency 27.8 28.1 Persistence 80 80 POST Latency 28.9 28.7 Persistence 45 90 SCIAdapt02 PRE Latency 27.8 33.5 Persistence 40 30 MID Latency 28.6 35.1 Persistence 90 100 POST Latency 29.6 33.8 Persistence 90 100 SCIAdapt04 PRE Latency 28.7 27.8 Persistence 100 80 MID Latency 28.4 28.9 Persistence 100 100 MID RETEST Latency 28.9 27.4 Persistence 95 95 POST Latency 28.2 27.7 Persistence 90 100 SCIAdaptC04 CONTROL Latency 27.4 26.8 Persistence 65 65 SCIAdaptC05 CONTROL Latency 32.8 34.5 Persistence 50 35 SCIAdaptC06 CONTROL Latency 31.7 30.7 Persistence 45 60 Latencies (ms) and persistence of F waves (% of positive responses out of 20 trials) of the TA muscle are presented. For each subjects training block sequence, refer to Figure 51.
198 Table 57 Central conduction times. Participant Assessment Muscle MEP latency in ms M wave latency in ms F wave latency in ms CCT in ms SCIAdapt01 PRE TA Right 25.00 4.37 30.0 7.32 TA Left 25.00 4.03 29.2 7.89 MID TA Right 30.00 5.05 27.8 13.08 TA Left 33.00 5.04 28.1 15.93 POST TA Right 31.50 3.97 28.9 14.57 TA Left 32.00 4.03 28.7 15.14 SCIAdapt02 PRE TA Right 29.50 4.47 27.8 12.87 TA Left 29.50 4.03 33.5 10.24 MID TA Right 29.00 4.46 28.6 11.97 TA Left 30.00 4.62 35.1 9.64 POST TA Right 4.47 29.6 TA Left 32.50 4.77 33.8 12.72 SCIAdapt04 PRE TA Right 31.00 5.15 28.7 13.58 TA Left 33.00 5.02 27.8 16.09 MID TA Right 31.00 5.22 28.4 13.69 TA Left 4.77 28.9 MID RETEST TA Right 31.00 4.31 28.9 13.90 TA Left 33.00 27.4 POST TA Right 33.50 5.28 28.2 16.26 TA Left 31.00 5.16 27.7 14.07 SCIAdaptC04 TA Right 27.50 5.38 27.4 10.61 TA Left 29.00 5.25 26.8 12.48 SCIAdaptC05 TA Right 32.00 4.46 32.8 12.87 TA Left 32.00 4.64 34.5 11.93 SCIAdaptC06 TA Right 30.00 5.48 31.7 10.91 TA Left 5.21 30.7 Central conduction times (CCT) are presented as calculated from motor evoked potentials recorded from the Tibialis anterior m uscle and F waves of the Peroneal nerve according to Rossini et al. ( Rossini and Pauri 1999) For each subjects training block sequence, refer to Figure 5 1.
199 Table 58 Ipsilateral silent periods in participants with iSCI. The duration of the ipsilateral silent periods (iSP) are presented and represent the time during which EMG activity falls below mean baseline by more than 1 standard deviation after stimulation of the i psilateral cortex. : this test was not conducted in this participant; *: this test was conducted, but a response could not be identified or the data were uninterpretable. For each subjects training block sequence, refer to Figure 51.
200 Muscle SCIAda pt01 SCIAdapt02 SCIAdapt04 PRE MID POST PRE MID POST PRE MID MID RE POST Onset Time (ms) R TA 60.5 56.5 50.7 43.5 63.0 62.5 60.0 L TA 59.5 58.0 60.0 63.0 59.0 61.0 60.0 L FDI 36.5 44.0 39.6 * Offset Time (ms) R TA 279.6 284.5 115.0 115.1 275.7 281.9 277.3 L TA 251.2 208.5 131.5 185.0 227.5 260.2 282.6 L FDI 66.0 90.0 94.2 * Duration (ms) R TA 219.1 228.0 64.2 71.6 212.7 219.4 217.3 L TA 191.7 150.5 71.5 122.0 168.5 199.2 222.6 L FDI 29.5 46.0 54.6 * Total Block Area (V*ms) R TA 35248.2 74463.6 3168.5 2598.7 7704.9 12999.12 18979.7 L TA 32828.7 22693.6 4372.0 3940.2 23010.3 15975.0 33768.4 L FDI 36733.3 1864.0 2671.3 * iSP area (V*ms) R TA 25078.8 55293.2 1216.8 1005.7 4413.5 9985.8 15960.9 L TA 21677.1 16602.7 1674.7 1951.4 16582.1 10285.7 26458.6 L FDI 22040.7 365.0 564.0 * % of Inhibition R TA 71.1 74.3 38.4 38.7 57.3 76.8 84.1 L TA 66.0 73.2 38.3 49.5 72.1 65.0 78.4 L FDI 60.0 19.6 21.1 * MEP area (V*ms) R TA 12831.8 5935.0 1387.2 349.8 4254.9 7746.8 6270.8 L TA 4676.3 3313.1 582.9 2131.6 4200.3 3892.0 5839.6 L FDI 4325.1 216.5 * MEP latency (ms) R TA 28.0 28.7 29.0 30.0 26.2 27.3 26.0 L TA 28.5 29.7 33.0 32.6 32.6 32.0 31.5 L FDI 18.5 17.1 *
201 Table 59 Ipsilateral silent period in ablebodied control participants. Muscle SCIAdaptC04 SCIAdaptC05 SCIAdaptC06 Onset Time (ms) R TA 62.0 75.5 56.5 L TA 59.5 63.5 L FDI 27.5 30.3 Offset Time (ms) R TA 242.5 307.0 169.0 L TA 274.4 311.5 L FDI 74.9 75.0 Duration (ms) R TA 180.5 231.5 112.5 L TA 214.9 248.0 L FDI 47.5 44.7 Total Block Area (V*ms) R TA 24072.6 28461.5 27255.6.3 L TA 47799.4 25171.4 L FDI 35008.2 48454.4 CSP area (V*ms) R TA 12426.1 16402.0 12607.4.3 L TA 35178.9 10067.3 L FDI 20468.8 27903.62 % of Inhibition R TA 51.6 57.6 46.3.5 L TA 73.6 40.0 L FDI 58.5 57.6 MEP area (V*ms) R TA 5704.9 11435.3 13918.6.5 L TA 5717.8 11599.0 L FDI MEP latency (ms) R TA 25.2 27.5 26.9 L TA 26.0 27.5 L FDI The duration of the ipsilateral silent periods (iSP) are presented and represent the time during which EMG activity falls below mean baseline by more than 1 standard deviation after stimulation of the ipsilateral cortex. : this test was not conducted in this participant; *: this test was conducted, but a response could not be identified or the data were uninterpretable.
202 Table 510. H Reflex measures as an indication of spinal cord excitability in individuals with SCI. Participant PRE MID MID RETEST POST SCIAdapt01 H/M Ratio right 0.02 M Lat right in ms 2.8 4.5 5.06 H Lat right in ms 30.0 H/M Ratio left 0.06 0.02 0.15 M Lat left in ms 4.1 3.4 4.3 H Lat left in ms 29.2 29.1 29.7 SCIAdapt02 H/M Ratio right 0.06 0.03 0.08 M Lat right in ms 5.5 4.7 3.89 H Lat right in ms 32.6 31.9 31.8 H/M Ratio left 0.60 0.26 1.99 M Lat left in ms 7.1 5.1 5.1 H Lat left in ms 33.1 30.9 31.17 SCIAdapt04 H/M Ratio right 0.16 0.01 0.04 0.03 M Lat right in ms 5.8 6.0 4.2 4.3 H Lat right in ms 32.8 32.3 31.7 31.8 H/M Ratio left 0.07 0.03 0.04 0.02 M Lat left in ms 5.2 5.0 4.9 3.5 H Lat left in ms 31.4 32.4 30.8 31.5 M Lat: latency of the Mwave, H Lat: latency of the H reflex, H M ratio: ratio between slopes of the ascending part of the H reflex to the slope of the ascending part of the Mwave as described in the text; *: no clear H reflex could be identified for this participant. For each subjects training block sequence, re fer to Figure 51.
203 Table 51 1 H Reflex measures as a measure of spinal cord excitability healthy control participants. Participant SCIAdaptC04 H/M Ratio right M Lat right in ms 3.9 H Lat right in ms H/M Ratio left 0.16 M Lat left in ms 4.4 H Lat left in ms 30.0 SCIAdaptC05 H/M Ratio right 0.26 M Lat right in ms 5.5 H Lat right in ms 32.4 H/M Ratio left M Lat left in ms 6.5 H Lat left in ms SCIAdaptC06 H/M Ratio right 0.27 M Lat right in ms 4.8 H Lat right in ms 32.6 H/M Ratio left 0.21 M Lat left in ms 4.7 H Lat left in ms 31.8 M Lat: latency of the Mwave, H Lat: latency of the H reflex, H M ratio: ratio between slopes of the ascending part of the H reflex to the slope of the ascending part of the Mwave as described in the text; *: no clear H reflex could be identified for this participant.
204 Table 51 2 Latencies and amplitudes of Vestibulospinal reflex responses A1 Response in ms A2 Response in ms A3 response in ms Latency (in ms) Amplitude (in cm) Latency (in ms) Amplitude (in cm) Latency (in ms) Amplitude (in cm) Not defined Not defined 139.2 1.56 580.9 2.40 Center of pressure. A1, A2 and A3: Crossings of Center of pressure traces with anodal stimulation on the left and on the right side. Please see text for further explanations. Table 51 3 Latencies and amplitudes of Vestibulospinal reflex responses: EMG response. Muscle Short latency response Medium latency response Off response response Latency (in ms) Amplitude (in %) Latency (in ms) Amplitude (in %) Latency (in ms) Amplitude (in %) TA 119.5 184.3 242.2 162.8 421.5 183.0 Soleus 112.4 242.1 304.3 127.4 437.0 315.7
205 Table 51 4 Acoustic startle reflex responses in individuals with iSCI. SCIAdapt01 SCIAdapt02 SCIAdapt04 PRE MID POST PRE MID POST PRE MID MID RE POST Supine Stand Supine Stand Supine Stand Supine Stand Supine Stand Supine Stand Supine Stand Supine Stand Supine Stand Supine Stand LSC 2 1 2 2 2 2 2 1 2 2 1 1 0 2 1 0 0 0 RSC 2 2 2 2 2 2 2 1 2 2 1 1 0 1 1 0 0 0 LBB 2 0 1 0 2 2 2 1 2 2 1 1 0 0 1 0 2 0 RBB 2 0 1 2 2 2 2 1 1 0 0 1 0 0 0 0 0 0 LES 2 0 1 2 2 2 2 2 2 2 0 1 0 1 0 0 0 2 RES 2 0 1 2 2 2 2 1 2 2 0 0 0 1 0 0 0 0 LVM 1 0 0 0 2 2 2 1 1 2 0 0 0 1 0 0 0 0 RVM 1 0 0 2 2 2 1 1 1 2 0 0 1 2 0 0 0 0 LMH 2 0 0 0 2 2 1 1 2 1 0 0 0 2 0 0 0 0 RMH 0 0 0 2 2 2 2 1 0 1 0 2 0 1 0 0 0 1 LTA 1 1 1 1 2 2 1 1 2 1 0 2 0 2 0 0 1 2 RTA 1 0 0 1 2 2 0 2 0 1 0 0 0 1 0 1 0 0 LSol 0 0 0 0 2 2 2 1 2 2 0 1 0 0 0 0 0 1 RSol 0 0 0 0 2 2 0 1 0 2 1 2 0 1 0 0 0 1 Number of responses visually defined as present/absent from rectified EMG traces recorded from 14 muscles (left). For each subjects training block sequence, refer to Figure 51. *: no clear response could be identified.
206 Table 51 5 Acoustic startle reflex responses in ablebodied control participants. SCIAdaptC01 SCIAdaptC02 SCIAdaptC06 Supine Stand Supine Stand Supine Stand LSC 1 0 2 2 2 2 RSC 2 1 2 2 2 2 LBB 0 1 0 1 2 2 RBB 0 1 0 2 2 2 LES 1 2 2 2 2 2 RES 1 2 2 2 2 2 LVM 0 0 2 2 2 2 RVM 0 0 1 1 2 1 LMH 0 0 1 1 2 2 RMH 0 0 1 1 2 2 LTA 0 1 0 2 2 2 RTA 0 1 0 2 2 2 LSol 0 0 1 2 2 0 RSol 0 0 0 0 2 0 Number of responses visually defined as present/absent from rectified EMG traces recorded from 14 muscles (left).
207 Table 51 6 Spatiotemporal gait characteristics. SCI Adapt01 SCI Adapt02 SCI Adapt04 Controls Pre Mid Post Pre Mid Post Pre Mid MidRe Post SSWS (m/s) Mean 1.16 1.26 1.28 1.00 1.27 0.93 1.28 1.33 1.41 1.18 Std 0.04 0.03 0.04 0.17 0.14 0.09 0.11 0.03 0.10 0.14 FCWS (m/s) Mean 1.76 1.58 1.67 1.54 1.58 1.70 1.5 1.54 1.74 1.65 1.80 Std 0.20 0.04 0 0.01 0.02 0.00 0.04 0.04 0.00 0.09 0.33 Cadence (st/min) Mean 0.62 0.63 0.65 0.57 0.64 0.66 0.54 0.66 0.68 0.71 114.35 Std 0.02 0.02 0.02 0.02 0.03 0.04 0.04 0.03 0.02 0.06 3.74 Step Length (m) Mean 0.62 0.63 0.65 0.57 0.64 0.66 0.54 0.66 0.68 0.71 0.33 Std 0.02 0.02 0.02 0.02 0.03 0.04 0.04 0.03 0.02 0.06 0.33 Step Width (m) Mean 0.11 0.12 0.11 0.06 0.04 0.09 0.09 0.08 0.08 0.08 0.07 Std 0.04 0.05 0.03 0.04 0.03 0.05 0.03 0.04 0.03 0.05 0.04 Stride Length (m) Mean 1.25 1.27 1.3 1.15 1.28 1.32 1.08 1.31 1.36 1.41 0.65 Std 0.03 0.02 0.02 0.04 0.02 0.07 0.08 0.05 0.03 0.09 0.67 Stance (% GC) Mean 65.72 64.58 64.7 62.04 60.59 63.39 69.23 64.15 63.76 64.88 33.95 Std 1.74 1.97 1.86 4.51 4.98 2.33 3.39 1.03 0.77 2.11 35.90 Single Limb Support (% GC) Mean 35.29 36.43 35.85 38.47 39.66 37.21 32.15 36.9 37.07 36.21 17.87 Std 2.22 2.81 1.82 4.90 5.17 2.99 3.77 1.53 0.83 3.13 18.21 Swing (% GC) Mean 34.74 35.91 35.8 38.34 39.88 37.08 31.19 36.34 36.73 35.63 17.59 Std 1.74 1.96 1.86 4.51 4.97 2.34 3.36 1.03 0.77 2.10 18.10 Spatiotemporal gait characteristics for individuals with iSCI and ablebodied controls are presented. The mean data for each subject is the average value calculated from 69 steps during a single walking trial. The mean data values for the control subjects is the average value calculated from data recorded in the 3 able bodied controls. SSWS: Self selected walking speed in meters per second (m/s); FCWS: Fastest comfortable walking speed in m/s. The values calculated for cadence (steps per minute, st/min), step length, step width, stride length, stance onset indicated at the percentage of the gait cycle at which stance begins (stance, % GC), single limb support onset (single limb support, % GC) and swing onset (swing, % GC) were calculated from values record ed during walking at a self selected walking speed. *: parameters could not be computed for this parameter due to software problems; for each subjects training block sequence, refer to Figure 51.
208 Table 51 7 Step clearance when stepping over a 6inch box. SCIAdapt01 SCIAdapt02 SCIAdapt04 Controls Distance (mm) PRE MID POST PRE MID POST PRE MID MID RE POST Vertical H Mean 147.86 142.34 162.14 160.98 240.83 167.16 181.51 190.44 194.91 189.91 140.33 Std 26.83 13.75 6.19 19.24 101.89 3.05 25.64 32.49 18.91 31.21 86.79 Vertical T Mean 123.05 125.47 109.67 154.09 124.74 123.73 148.22 142.81 109.42 146.89 132.41 Std 38.21 13.34 16.82 21.01 21.65 5.73 11.50 8.31 10.43 18.58 14.68 Horizontal T Mean 479.87 639.85 684.32 531.94 477.85 438.13 571.92 498.37 511.31 538.20 540.25 Std 46.63 42.5 7.48 62.44 153.33 19.35 34.35 52.93 39.21 36.29 60.33 Horizontal H Mean 228.52 182.99 188.55 433.96 572.62 531.17 441.49 514.98 576.42 498.54 368.01 Std 71.61 26.00 12.21 25.11 183.42 14.85 41.88 75.27 18.97 60.35 211.51 Step clearance distances for the heel (H) and great toe (T) when stepping over a 6inch box in individuals with iSCI and ablebodied controls are presented. The goal of the task was to step over the box, but to remain as close to it as possible. Therefore smaller values indicate improvement. The vertical clearance distances are presented as well as the horizontal distance from the box once the foot contacts the gro und following the object clearance. The mean data for each subject is the average value calculated from 3 trials. The mean data values for the control subjects is the average value calculated from data recorded in the 3 able bodied controls. For each subjects training block sequence, refer to Figure 51.
209 Table 51 8 Step clearance when stepping over a 10inch box. SCI Adapt01 SCI Adapt02 SCI Adapt04 Controls PRE MID POST PRE MID POST PRE MID POST PRE Vertical H Mean 131.12 153.18 123.14 148.85 188.26 178.02 199.79 165.85 189.64 158.88 169.04 Std 7.22 1.46 28.2 32.07 15.24 30.38 20.12 41.49 52.44 41.67 58.35 Vertical T Mean 109.74 103.85 103.85 135.65 141.68 122.66 157.96 118.51 106.4 104.07 127.99 Std 14.83 17.47 17.47 7.47 22.26 11.44 9.04 11.87 13.72 33.73 19.89 Horizontal T Mean 578.5 679.73 555.24 488.28 474.77 448.61 523.73 505.33 415.25 511.12 492.22 Std 34.14 32.27 4.21 28.58 51.55 28.86 21.74 109.22 55.46 56.23 82.92 Horizontal H Mean 149.59 187.34 230.02 490.61 589.74 598.29 415.45 509.62 640.32 556.98 458.49 Vertical H Std 46.87 66.19 26.04 44.93 15.95 55.26 37.05 122.03 76.04 85.22 177.29 Step clearance distances for the heel (H) and great toe (T) when stepping over a 10inch box in individuals with iSCI and ablebodied controls are presented. The goal of the task was to step over the box, but to remain as close to it as possible. Therefore smaller values indicate improvement. The vertical clearance distances are presented as well as the horizontal distance from the box once the foot contacts the gro und following the object clearance. The mean data for each subject is the average value calculated from 3 trials. The mean data values for the control subjects is the average value calculated from data recorded in the 3 able bodied controls. For each subjects training block sequence, refer to Fig ure 51.
210 Table 51 9 Results from clinical outcome measures. SCIAdapt01 SCIAdapt02 SCIAdapt04 PRE MID POST PRE MID POST PRE MID MID RE POST WISCI 20 20 20 20 20 20 20 20 20 20 SCI FAP 8.01 7.09 10.58 7.42 7.61 8.09 6.19 7.39 6.92 NMRS 3A 3B 3B 2B 2C 3A 3B 3C 4A 3C ABC 80.0 93.8 91.3 76.1 69.4 82.5 83.9 91.3 65.0 76.3 SAM mean(std) 2964 (1060) 2822 (1344) 5345 (1834) 4795 (1276) 3282 (2395) 5050 (601) 4326 (1125) 5999 (1454) 4189 (681) Raw values of the individual test scores are presented for the WISCI, SCIFAP, NMRS and the ABC. Means and standard deviations over four days are shown for the SAM data. : data not available for these assessments; For each subjects training block sequence, refer to Figure 51.
211 Figure 51. Overview o f training and testing days. T: Training took place on this day; *: testing took place on this day; blue shading: Basic LT took place, orange shading: AdaptLT took place.
212 Figure 52 TMS Recruitment curve from the right TA muscle in SCIAdapt04. Panel A: Pre testing session; Panel B: testing post AdaptLT; Panel C: testing pre BasicLT; Panel D: testing post BasicLT.
213 Figure 53 iSPs recorded in the left TA muscle of SCIAdapt 04. Representative data are shown. Panel A : Pre testing session; P anel B : testing post AdaptLT; Panel C : testing pre BasicLT; Panel D : testing post BasicLT.
214 Figure 5 4 F waves from the right TA muscle in SCIAdapt04. Panel A: Pre testing session; Panel B: testing post AdaptLT; Panel C: testing pre BasicLT; Panel D: testing post BasicLT. Vertical lines: F wave onset.
215 Figure 55 H Reflex recruitment curves and slopes in SCIAdapt01. A: Pre testing session; B: post AdaptLT; C: post BasicLT. Stimulation intensity was normalized to the intensity at motor threshold.
216 Figure 56. Averaged EMG responses and sway in response to galvanic vestibular stimulation in SCIAdapt02 at MID testing. The participant stood on a foam pillow with their head turned to the left side. The average of all stimulations with the anode behind the left mastoid process is shown in the red traces; the average of all stimulations with the anode behind the right mastoid process is shown in the black traces SL: short latency EMG response, ML: medium latency EMG response; OFF: EMG response after termination of the stimulus. A2, A3: alternating body sway in the anterior posterior direction.
217 Fi gure 57 Average self selected (Panel A) and fastest comfortable (Panel B) walking speed before, between and after the two training blocks.SCIAdapt04 was tested four times, before and after the first block of training and before and after the second bloc k of training. Colored lines indicate the kind of training that took place in each participant between two consecutive assessments. Black line and grey zones indicate the mean and 1standard deviation of ablebodied control participants.
218 Figure 58 Average stride length (Panel A) and cadence (Panel B) before, between and after the two training blocks.SCIAdapt04 was tested four times, before and after the first block of training and before and after the second block of training. Colored lines indicate the kind of training that took place in each participant between two consecutive assessments. Black line and grey zones indicate the mean and 1standard deviation of ablebodied control participants.
219 Figure 59 Average distance between heel (Panel A) or toe (Panel B) and the upper edge of the 6 inch box before, between and after the two training blocks.SCIAdapt04 was tested four times, before and after the first block of training and before and after the second block of training. Colored lines indicate the kind of training that took place in each participant between two consecutive assessments. Black line and grey zones indicate the mean and 1standard deviation of ablebodied control participants.
220 Figure 510. Average distance between heel (Panel A) or toe (Panel B) and the upper edge of the 10 inch box before, between and after the two training blocks.SCIAdapt04 was tested four times, before and after the first block of training and before and after the second block of training. Colored l ines indicate the kind of training that took place in each participant between two consecutive assessments. Black line and grey zones indicate the mean and 1standard deviation of ablebodied control participants.
221 Figure 511. Average distance between the toe and the front end of the box at toe off (Panel A) and the heel and the back end of the box at heel strike (Panel B) when stepping over the 6 box before, between and after the two training blocks.SCIAdapt04 was tested four times, before and after t he first block of training and before and after the second block of training. Colored lines indicate the kind of training that took place in each participant between two consecutive assessments. Black line and grey zones indicate the mean and 1standard dev iation of ablebodied control participants.
222 Figure 512. Average distance between the toe and the front end of the box at toe off (Panel A) and the heel and the back end of the box at heel strike (Panel B) when stepping over the 10 box before, between and after the two training blocks.SCIAdapt04 was tested four times, before and after the first block of training and before and after the second block of training. Colored lines indicate the kind of training that took place in each participant between two consecutive assessments. Black line and grey zones indicate the mean and 1standard deviation of ablebodied control participants.
223 Figure 513. Spinal Cord Injury Functional Ambulation Profile. SCI FAP scores are composed of the time needed to complete 7 tasks and the assistance needed to do so. Therefore, decreasing scores indicate improvement. Raw scores from PRE, MID and POST testing are presented for SCIAdapt02. Only PRE and MID test results are available for SCIAdapt01. SCIAdapt04 was t ested four times, before and after the first block of training and before and after the second block of training. Colored lines indicate the kind of training that took place in each participant between two consecutive assessments.
224 Figure 514. Neuromuscular recovery scale. Individuals are assigned to a phase according to their degree to which they can perform certain activities with recovered preinjury kinematic movement patterns. Higher scores indicate more recovery. Raw scores from PRE, MID a nd POST testing are presented. SCIAdapt04 was tested four times, before and after the first block of training and before and after the second block of training. Colored lines indicate the kind of training that took place in each participant between two consecutive assessments.
225 Figure 515. Enhanced kinematics while stepping over a box. Panel A shows participant SCIAdapt02 stepping over a box during the baseline testing session. Panel B shows the same participant stepping over the box after two weeks o f AdaptLT. Please note improved hip flexion, reduced circumduction and that no manual assistance is provided to complete the task after two weeks of Adapt LT ( Photos courtesy of Martina Spiess )
226 Figure 516. Activity Balance Confidence Scale. Participants subjective confidence in their ability to conduct 16 daily walking tasks without falling. Higher scores indicate more confidence. Raw scores from PRE, MID and POST testing are presented. SCIAdapt04 was tested four times, before and after the first bl ock of training and before and after the second block of training. Colored lines indicate the kind of training that took place in each participant between two consecutive assessments.
227 Figure 517. Steps taken with the right foot per day in the usual environment. Averaged steps per day over two weekdays and two weekend days are shown. Average daily steps from PRE, MID and POST testing are presented for SCIAdapt02. Data from SCIAdapt01 is only available for MID and POST testing. SCIAdapt04 was tested fo ur times, before and after the first block of training and before and after the second block of training. Colored lines indicate the kind of training that took place in each participant between two consecutive assessments.
228 CHAPTER 6 CONCLUSIONS I ncomplete spinal cord injury (iSCI) results in different degrees of damage to the descending spinal tracts. This dissertation investigated spared and regained function of these descending tracts from three different perspectives. First, we described the fu nctional ability in individuals with Brown Squard Syndrome (BSS)/ Brown Squard Plus syndrome (BSPS), a type of iSCI with a specific pattern of descending tract damage. Secondly, we suggest modifications to an existing descending tract test, the acoustic s tartle response (ASR) in order to improve its usefulness in assessing the functional integrity of the reticulospinal tract (ReST) in individuals with iSCI. Finally, in a pilot study, we explored the feasibility and usefulness of a training protocol aimed at inducing plasticity in the descending tracts. Individuals with BSS/BSPS suffer from a hemi severance of their spinal cord, leading to diminished motor function and sensitivity to vibration and proprioception ipsilateral to the lesion and contralesional r eduction of sensitivity to pain and temperature. Historically, these patients are believed to have a better prognosis than individuals that suffered from other types of iSCI. However, results of newer studies do not confirm this notion. One reason for this disagreement could be that different definitions have been used by different authors. We hypothesized that the definition used would affect the percentage of individuals classified as having BSS as well as the estimate of their functional outcome. We clas sified 132 participants from a single dataset as having BSS or not using to three recently used, and well described, definitions in the literature ( Hayes, Hsieh et al. 2000; Wirz, Zorner et al. 2009; Pouw, van de Meent et al. 2010) Seventee n, eight and zero participants were classified as
229 BSS/BSPS, using definitions proposed by Pouw, Wirz and Hayes respectively. While there were no statistically significant differences between the groups defined as BSS/BSPS by the different definitions, the group defined as having BSPS by Pouw et al. ( Pouw, van de Meent et al. 2010) reached the highest scores on all of the functional tests used. No difference in the amount of fu nctional improvement over 35 to 60 sessions of locomotor training could be shown between the groups. This lack of difference could partly be due to the fact that a limited number of people were classified as having BSS/BSPS. The evidence supported our hypothesis as follows: 1) the definition used to define BSS/BSPS influenced the estimate of the number of people classified as BSS/BSPS and 2) the definition used possibly influenced their described level of functional ability. A consensus in definitions used is needed to increase comparability between studies, especially between studies in humans with BSS/BSPS and studies in animals after spinal hemisections. In the second experiment, we investigated the presence of habituation and a method to decrease habituation of evoked muscle responses to an acoustic startle. The ASR as a test of functional ReST integrity has only been minimally described in persons with SCI. Short term (during one testing session) as well as long term (over several days) habituation have been described in humans with other disorders as well as in animal models, but the effects of habituation are not well understood in humans with iSCI. However, habituation can potentially decrease the utility of this test, especially as an outcome measure in a longitudinal study. Our study found that in ablebodied control participants, habituation of muscle responses to an acoustic startle is reduced if individuals are standing on a foam pillow during testing, rather than lying supine. In
230 individuals with iSCI, no short term habitation could be found in lying supine, standing or standing on foam when individuals were changing positions randomly. In studies where individuals with iSCI are compared to ablebodied individuals, we recommend that, if possible, t esting should be conducted while individuals are in a position that challenges the postural system in order to reduce short term habituation. Our data further provides evidence of long term habituation in ablebodied control participants. Responses are mos tly reduced in reaction to initial stimulations when tested again 48hrs later. Due to reduced short term habituation during the second testing session when compared to the first, later stimulations yield similar percentages of response between the first and second testing session. This indicates that initial stimulations might have to be discarded when using this test as an outcome measure in a longitudinal study. The proposed protocol potentially enhances stability of ASR testing and therefore its utility as a test of ReST function. These methodological steps could help to detect important plastic changes in this tract in response to iSCI and treatment. One potential treatment inducing plasticity in the ReST and other descending systems is locomotor training (LT). Traditional LT (BasicLT), focusing on basic stepping to target spinal cord plasticity, has lead to recovery of basic stepping function in many individuals with iSCI. However these persons are still limited in every day community walking due to thei r deficits in managing higher skilled walking tasks. Locomotor training including practice of these more demanding tasks (AdaptLT), and thus targeting descending tract plasticity, could potentially increase the functional walking ability of persons with iS CI beyond basic stepping and allow them to successfully manage demands of every day community ambulation. In this pilot investigation, we compared
231 three weeks of BasicLT to 3 weeks of AdaptLT in three individuals with iSCI that were able to walk independently but reported difficulties with more demanding walking tasks. Individuals showed improvement in several outcome measures over the total training time of 6 weeks, but no differential effect between the two different training types could be shown. Several reasons could be responsible for this, including the low number of participants and the relatively short duration of the study with 15 training sessions each. In summary, when describing a group of individuals with spared or recovered descending tract function, definitions matter and homogeneous groups can only be described if the same definitions are used. When assessing ReST function using the ASR, we recommend that individuals stand on foam during the testing if results from ablebodied persons would b e compared to those of individuals with iSCI. Further, long term habituation must be taken into consideration when using this test as an outcome measure in a treatment study. Finally, while no difference between the two modalities could be shown in our thr ee participants, locomotor training improved the functional walking ability of individuals with iSCI regardless of whether training targeted mainly spinal cord or descending tract plasticity. These findings contribute to the understanding of descending tra ct function in individuals with iSCI. Understanding the presence, function and plasticity of these tracts will ultimately help to improve treatment after iSCI, specifically with regards to walking function. The overarching goal is for individuals with iSCI to recover walking function at a higher level to allow them to successfully navigate challenges of every day walking in the community.
232 LIST OF REFERENCES Adkins, D. L., J. Boychuk, et al. (2006). "Motor training induces experiencespecific patterns of plasticity across motor cortex and spinal cord." J Appl Physiol 101(6): 17761782. Al Jawayed, I. A., M. Sabbahi, et al. (1999). "The H reflex modulation in lying and a semi reclining (sitting) position." Clin Neurophysiol 110(12): 20442048. Alexander, M. S., K. D. Anderson, et al. (2009). "Outcome measures in spinal cord injury: recent assessments and recommendations for future directions." Spinal Cord 47(8): 582591. Ali, A. A. and M. A. Sabbahi (2000). "H reflex changes under spinal loading and unloading conditions in normal subjects." Clin Neurophysiol 111(4): 664670. Ali, A. and M. A. Sabbahi (2001). "Test retest reliability of the soleus H reflex in three different positions." Electromyogr Clin Neurophysiol 41(4): 209214. Andersen, J. B. and T. Sink jaer (1999). "The stretch reflex and H reflex of the human soleus muscle during walking." Motor Control 3(2): 151157. Anderson, D., S. Dumont, et al. (2007). "Determinants of return to work among spinal cord injury parients: A literature review." Journal of Vocational Rehabilitation 27: 5768. Anderson, K., S. Aito, et al. (2008). "Functional recovery measures for spinal cord injury: an evidencebased review for clinical practice and research." J Spinal Cord Med 31(2): 133144. Anderson, K. D. (2004). "Tar geting recovery: priorities of the spinal cordinjured population." J Neurotrauma 21(10): 13711383. Andersson, O., S. Grillner, et al. (1978). "Peripheral control of the spinal pattern generators for locomotion in cat." Brain Res 150(3): 625630. Andrews, A. W., S. A. Chinworth, et al. (2010). "Update on distance and velocity requirements for community ambulation." J Geriatr Phys Ther 33(3): 128134. Armstrong, D. M. (1988). "The supraspinal control of mammalian locomotion." J Physiol 405: 137. Arnold, C. M. and R. A. Faulkner (2007). "The history of falls and the association of the timed up and go test to falls and near falls in older adults with hip osteoarthritis." BMC Geriatr 7: 17.
233 Arvanian, V. L., L. Schnell, et al. (2009). "Chronic spinal hemisectio n in rats induces a progressive decline in transmission in uninjured fibers to motoneurons." Exp Neurol 216(2): 471480. ASIA (2002). International Standards for Neurological Classification of Spinal Cord Injury. Chicago IL, American Spinal Injury Associat ion. revised 2002. ASIA (2011). International Standards for Neurological Classification of Spinal Cord Injury. Atlanta, GA, American Spinal Injury Association. revised 2011. Backus, D. (2010). "Exploring the potential for neural recovery after incomplete t etraplegia through nonsurgical interventions." PM R 2(12 Suppl 2): S279285. Baker, S. N. (2011). "The primate reticulospinal tract, hand function and functional recovery." J Physiol 589(Pt 23): 56035612. Baloh, R. W. and V. Honrubia (1979). "The central vestibular system." Contemp Neurol Ser 18: 47 99. Baloh, R. W. and K. Kerber (2010). Clinical neurophysiology of the vestibular system. New York, NY, Oxford University Press. Barbeau, H. (2003). "Locomotor training in neurorehabilitation: emerging rehabili tation concepts." Neurorehabil Neural Repair 17(1): 311. Barbeau, H., M. Danakas, et al. (1993). "The effects of locomotor training in spinal cord injured subjects: a preliminary study." Restor Neurol Neurosci 5: 8184. Barbeau, H., J. Fung, et al. (2002) "A review of the adaptability and recovery of locomotion after spinal cord injury." Prog Brain Res 137: 925. Barbeau, H., J. Fung, et al. (1999). "New approach to retrain gait in stroke and spinal cord injured patients." Neurorehabilitation and Neural R epair 13(3): 177178. Barbeau, H., M. Ladouceur, et al. (2002). "The effect of locomotor training combined with functional electrical stimulation in chronic spinal cord injured subjects: walking and reflex studies." Brain Res Rev 40(13): 274291. Barbeau, H., M. Ladouceur, et al. (1999). "Walking after spinal cord injury: evaluation, treatment, and functional recovery." Arch Phys Med Rehabil 80(2): 225235. Barbeau, H., D. A. McCrea, et al. (1999). "Tapping into spinal circuits to restore motor function." Brain Res Brain Res Rev 30(1): 2751. Barbeau, H., K. Norman, et al. (1998). "Does neurorehabilitation play a role in the recovery of walking in neurological populations?" Ann N Y Acad Sci 860: 377392.
234 Barbeau, H. and S. Rossignol (1987). "Recovery of loc omotion after chronic spinalization in the adult cat." Brain Res 412(1): 8495. Barbeau, H. and S. Rossignol (1994). "Enhancement of locomotor recovery following spinal cord injury." Curr Opin Neurol 7(6): 517 524. Barnett, M. L., D. Ross, et al. (1973). Motor skills learning and the specificity of training principle." Res Q 44(4): 440447. Barthelemy, D., M. J. Grey, et al. (2011). "Involvement of the corticospinal tract in the control of human gait." Prog Brain Res 192: 181197. Basso, D. M. (2000). "Neuroanatomical substrates of functional recovery after experimental spinal cord injury: implications of basic science research for human spinal cord injury." Phys Ther 80(8): 808817. Basso, D. M., M. S. Beattie, et al. (1996). "Graded histological and locom otor outcomes after spinal cord contusion using the NYU weight drop device versus transection." Experimental Neurology 139(2): 244256. Basso, D. M., M. S. Beattie, et al. (2002). "Descending systems contributing to locomotor recovery after mild or moderat e spinal cord injury in rats: experimental evidence and a review of literature." Restor Neurol Neurosci 20(5): 189218. Basso, D. M., M. Murray, et al. (1994). "Differential recovery of bipedal and overground locomotion following complete spinal cord hemis ection in cats." Restorative Neurology and Neuroscience 7: 95110. Beaumont, E., S. Kaloustian, et al. (2008). "Training improves the electrophysiological properties of lumbar neurons and locomotion after thoracic spinal cord injury in rats." Neurosci Res 62(3): 147154. Behrman, A. L., E. Ardolino, et al. (2012). "Assessment of functional improvement without compensation reduces variability of outcome measures after human spinal cord injury." Arch Phys Med Rehabil accepted. Behrman, A. L., M. G. Bowden, et al. (2006). "Neuroplasticity after spinal cord injury and training: an emerging paradigm shift in rehabilitation and walking recovery." Phys Ther 86(10): 14061425. Behrman, A. L., S. M. Flynn, et al. (2002). "Locomotor training outcomes for persons with incomplete spinal cord injury." American Paraplegia Society. Behrman, A. L. and S. J. Harkema (2000). "Locomotor training after human spinal cord injury: a series of case studies." Phys Ther 80(7): 688700. Behrman, A. L. and S. J. Harkema (2007). "Physical rehabilitation as an agent for recovery after spinal cord injury." Phys Med Rehabil Clin N Am 18(2): 183202, v.
235 Behrman, A. L., A. R. Lawless Dixon, et al. (2005). "Locomotor training progression and outcomes after incomp lete spinal cord injury." Phys Ther 85(12): 13561371. Behrman, A. L., P. M. Nair, et al. (2008). "Locomotor training restores walking in a nonambulatory child with chronic, severe, incomplete cervical spinal cord injury." Phys Ther 88(5): 580590. Belling ham, M. C. (1998). "Driving respiration: the respiratory central pattern generator." Clin Exp Pharmacol Physiol 25(10): 847856. Benito Penalva, J., E. Opisso, et al. (2009). "H reflex modulation by transcranial magnetic stimulation in spinal cord injury s ubjects after gait training with electromechanical systems." Spinal Cord. Beres Jones, J. A. and S. J. Harkema (2004). "The human spinal cord interprets velocity dependent afferent input during stepping." Brain 127(Pt 10): 22322246. Bernstein, J. J. and M E. Bernstein (1971). "Axonal regeneration and formation of synapses proximal to the site of lesion following hemisection of the rat spinal cord." Experimental Neurology 30: 336351. Blesch, A. and M. H. Tuszynski (2009). "Spinal cord injury: plasticity, regeneration and the challenge of translational drug development." Trends Neurosci 32(1): 4147. Blight, A. R. and M. H. Tuszynski (2006). "Clinical trials in spinal cord injury." J Neurotrauma 23(34): 586593. Blumer, C. E. and S. Quine (1995). "Prevalen ce of spinal cord injury: an international comparison." Neuroepidemiology 14(5): 258268. Boakye, M. (2009). "Implications of neuroplasticity for neurosurgeons." Surg Neurol 71(1): 510. Borckardt, J. J., Z. Nahas, et al. (2006). "Estimating resting motor thresholds in transcranial magnetic stimulation research and practice: a computer simulation evaluation of best methods." J ECT 22(3): 169175. Bosch, A., E. S. Stauffer, et al. (1971). "Incomplete traumatic quadriplegia. A tenyear review." JAMA 216(3): 4 73478. Bowden, M. G. and A. L. Behrman (2007). "Step Activity Monitor: Accuracy and test retest reliability in persons with incomplete spinal cord injury." J Rehabil Res Dev 44(3): 355362. Bowden, M. G., A. L. Behrman, et al. (2012). "Advancing measureme nt of locomotor rehabilitation outcomes to optimize interventions and differentiate between recovery versus compensation." J Neurol Phys Ther 36(1): 3844.
236 Braff, D., C. Stone, et al. (1978). "Prestimulus effects on human startle reflex in normals and schi zophrenics." Psychophysiology 15(4): 339343. Britton, T. C., B. L. Day, et al. (1993). "Postural electromyographic responses in the arm and leg following galvanic vestibular stimulation in man." Exp Brain Res 94(1): 143151. Bronstein, A. M. (2006). "What the vestibulospinal system does and what it doesn't." Suppl Clin Neurophysiol 58: 120133. Brotherton, S. S., J. S. Krause, et al. (2007). "Falls in individuals with incomplete spinal cord injury." Spinal Cord 45(1): 3740. Brotherton, S. S., J. S. Kraus e, et al. (2007). "A pilot study of factors associated with falls in individuals with incomplete spinal cord injury." J Spinal Cord Med 30(3): 243250. Brown Squard, C. E. (1855). "Experimental and Clinical Researches on the Physiology and Pathology of th e Spinal Cord, and some other parts of the Nervous Centres." American Journal of the medical Sciences 30(59): 161165. Brown, A. and L. C. Weaver (2011). "The dark side of neuroplasticity." Exp Neurol. Brown, P., B. L. Day, et al. (1991). "The effect of posture on the normal and pathological auditory startle reflex." J Neurol Neurosurg Psychiatry 54(10): 892897. Brown, P., J. C. Rothwell, et al. (1991). "New observations on the normal auditory startle reflex in man." Brain 114 ( Pt 4): 1891 1902. Brown, T. G. (1911). "The intrinsic factors in the act of progression in the mammal." Proceedings of the Royal Society of London Series B Containing Papers of a Biological Character 84(572): 308319. Burke, D. A., R. D. Linden, et al. (2001). "Incidence rates and p opulations at risk for spinal cord injury: A regional study." Spinal Cord 39(5): 274278. Burns, A. S., R. J. Marino, et al. (2012). "Clinical diagnosis and prognosis following spinal cord injury." Handb Clin Neurol 109: 4762. Burns, S. P., D. G. Golding, et al. (1997). "Recovery of ambulation in motor incomplete tetraplegia." Arch Phys Med Rehabil 78(11): 11691172. Butler, A. J. and S. L. Wolf (2007). "Putting the brain on the map: use of transcranial magnetic stimulation to assess and induce cortical pl asticity of upper extremity movement." Phys Ther 87(6): 719736. Byl, N. N. and M. Melnick (1997). "The neural consequences of repetition: clinical implications of a learning hypothesis." J Hand Ther 10(2): 160174.
237 Byl, N. N., M. M. Merzenich, et al. (199 7). "A primate model for studying focal dystonia and repetitive strain injury: effects on the primary somatosensory cortex." Phys Ther 77(3): 269284. Cacchio, A., N. Cimini, et al. (2009). "Reliability of transcranial magnetic stimulationrelated measurem ents of tibialis anterior muscle in healthy subjects." Clin Neurophysiol 120(2): 414419. Cadenhead, K. S., B. S. Carasso, et al. (1999). "Prepulse inhibition and habituation of the startle response are stable neurobiological measures in a normal male population." Biol Psychiatry 45(3): 360364. Calancie, B., S. Lutton, et al. (1996). "Central nervous system plasticity after spinal cord injury in man: interlimb reflexes and the influence of cutaneous stimulation." Electroencephalogr Clin Neurophysiol 101(4) : 304315. Calancie, B., B. Needham Shropshire, et al. (1994). "Involuntary stepping after chronic spinal cord injury. Evidence for a central rhythm generator for locomotion in man." Brain 117 ( Pt 5): 1143 1159. Cangiano, L., R. H. Hill, et al. (2012). "T he hemisegmental locomotor network revisited." Neuroscience. Cao, Q., Y. P. Zhang, et al. (2005). "Functional and electrophysiological changes after graded traumatic spinal cord injury in adult rat." Exp Neurol 191 Suppl 1: S3S16. Capaday, C. (1997). "Neurophysiological methods for studies of the motor system in freely moving human subjects." J Neurosci Methods 74(2): 201218. Capaday, C. (2002). "The special nature of human walking and its neural control." Trends Neurosci 25(7): 370376. Capaday, C., B. A Lavoie, et al. (1999). "Studies on the corticospinal control of human walking. I. Responses to focal transcranial magnetic stimulation of the motor cortex." J Neurophysiol 81(1): 129139. Carroll, T. J., S. Riek, et al. (2001). "Reliability of the input output properties of the cortico spinal pathway obtained from transcranial magnetic and electrical stimulation." J Neurosci Methods 112(2): 193202. Cha, J., C. Heng, et al. (2007). "Locomotor ability in spinal rats is dependent on the amount of activity i mposed on the hindlimbs during treadmill training." J Neurotrauma 24(6): 10001012. Chen, R. (2004). "Interactions between inhibitory and excitatory circuits in the human motor cortex." Exp Brain Res 154(1): 110.
238 Chen, Z. L., W. M. Yu, et al. (2007). "Per ipheral regeneration." Annu Rev Neurosci 30: 209233. Cieza, A., I. Kirchberger, et al. (2010). "ICF Core Sets for individuals with spinal cord injury in the long term context." Spinal Cord 48(4): 305312. Clarke, C. E., H. Modarres Sadeghi, et al. (1994). "Prognostic value of cortical magnetic stimulation in spinal cord injury." Paraplegia 32(8): 554560. Coleman, K. L., D. G. Smith, et al. (1999). "Step activity monitor: long term, continuous recording of ambulatory function." J Rehabil Res Dev 36(1): 81 8. Cooke, S. F. and T. V. Bliss (2006). "Plasticity in the human central nervous system." Brain 129(Pt 7): 16591673. Cooper, B. G., D. Y. Miya, et al. (1998). "Superior colliculus and active navigation: role of visual and nonvisual cues in controlling cellular representations of space." Hippocampus 8(4): 340372. Cosar, S. N., O. U. Yemisci, et al. (2010). "Demographic characteristics after traumatic and nontraumatic spinal cord injury: a retrospective comparison study." Spinal Cord 48(12): 862866. Cree l, G. L., K. E. Light, et al. (2001). "Concurrent and construct validity of scores on the Timed Movement Battery." Phys Ther 81(2): 789798. Crenna, P. and C. Frigo (1987). "Excitability of the soleus H reflex arc during walking and stepping in man." Exp B rain Res 66(1): 4960. Crozier, K. S., V. Graziani, et al. (1991). "Spinal cord injury: prognosis for ambulation based on sensory examination in patients who are initially motor complete." Arch Phys Med Rehabil 72(2): 119121. Csomor, P. A., B. K. Yee, et al. (2009). "Impaired prepulse inhibition and prepulseelicited reactivity but intact reflex circuit excitability in unmedicated schizophrenia patients: a comparison with healthy subjects and medicated schizophrenia patients." Schizophr Bull 35(1): 244255 Curt, A., H. Alkadhi, et al. (2002). "Changes of nonaffected upper limb cortical representation in paraplegic patients as assessed by fMRI." Brain 125(Pt 11): 25672578. Curt, A. and V. Dietz (1997). "[Prognosis of traumatic spinal cord lesions. Significance of clinical and electrophysiological findings]." Nervenarzt 68(6): 485495. Curt, A. and V. Dietz (1999). "Electrophysiological recordings in patients with spinal cord injury: significance for predicting outcome." Spinal Cord 37(3): 157165.
239 C urt, A., M. E. Keck, et al. (1998). "Functional outcome following spinal cord injury: significance of motor evoked potentials and ASIA scores." Arch Phys Med Rehabil 79(1): 8186. Curt, A., M. E. Schwab, et al. (2004). "Providing the clinical basis for new interventional therapies: refined diagnosis and assessment of recovery after spinal cord injury." Spinal Cord 42(1): 16. Dahlberg, A., M. Kotila, et al. (2005). "Prevalence of spinal cord injury in Helsinki." Spinal Cord 43(1): 4750. Dajpratham, P. and R. Kongkasuwan (2011). "Quality of life among the traumatic spinal cord injured patients." J Med Assoc Thai 94(10): 12521259. Darian Smith, C. (2009). "Synaptic plasticity, neurogenesis, and functional recovery after spinal cord injury." Neuroscientist 15(2): 149165. Davis, M. (1984). The Mammalian startle response. Neural mecahnisms of startle behavior. New York, Plenum Press. Davis, M., D. S. Gendelman, et al. (1982). "A primary acoustic startle circuit: lesion and stimulation studies." J Neurosci 2(6): 791805. Davis, M., T. Parisi, et al. (1982). "Habituation and sensitization of startle reflexes elicited electrically from the brainstem." Science 218(4573): 688690. Davis, M. and A. R. Wagner (1969). "Habituation of startle response under incremental sequence of stimulus intensities." J Comp Physiol Psychol 67(4): 486492. Day, B. L. and R. C. Fitzpatrick (2005). "The vestibular system." Curr Biol 15(15): R583586. Day, B. L., E. Ramsay, et al. (2011). "The human semicircular canal model of galvanic ves tibular stimulation." Exp Brain Res 210(34): 561568. Day, K. V., S. A. Kautz, et al. (2011). "Foot placement variability as a walking balance mechanism post spinal cord injury." Clin Biomech (Bristol, Avon) 27(2): 145150. De Leon, R. D., J. A. Hodgson, et al. (1998). "Full weight bearing hindlimb standing following stand training in the adult spinal cat." J Neurophysiol 80(1): 8391. de Leon, R. D., J. A. Hodgson, et al. (1998). "Locomotor capacity attributable to step training versus spontaneous recover y after spinalization in adult cats." J Neurophysiol 79(3): 13291340. Delwaide, P. J. and B. Schepens (1995). "Auditory startle (audiospinal) reaction in normal man: EMG responses and H reflex changes in antagonistic lower limb muscles." Electroencephalogr Clin Neurophysiol 97(6): 416423.
240 Devivo, M. J. (2012). "Epidemiology of traumatic spinal cord injury: trends and future implications." Spinal Cord. Diehl, P., U. Kliesch, et al. (2006). "Impaired facilitation of motor evoked potentials in incomplete spinal cord injury." J Neurol 253(1): 5157. Dietz, V. (1987). "Role of peripheral afferents and spinal reflexes in normal and impaired human locomotion." Rev Neurol (Paris) 143(4): 241254. Dietz, V. (1995). "Central pattern generator." Paraplegia 33(12): 739. Dietz, V. (2002). "Proprioception and locomotor disorders." Nat Rev Neurosci 3(10): 781790. Dietz, V. (2003). "Spinal cord pattern generators for locomotion." Clin Neurophysiol 114(8): 13791389. Dietz, V. (2008). "Body weight supported gait training: from laboratory to clinical setting." Brain Res Bull 76(5): 459463. Dietz, V. (2009). "Body weight supported gait training: from laboratory to clinical setting." Brain Res Bull 78(1): I VI. Dietz, V. (2011). "Neuronal plasticity after a human spinal cord injury: Positive and negative effects." Exp Neurol. Dietz, V., R. Muller, et al. (2002). "Locomotor activity in spinal man: significance of afferent input from joint and load receptors." Brain 125(Pt 12): 26262634. Dietz, V., J. Quintern, et al. (1981). "Electrophysiological studies of gait in spasticity and rigidity. Evidence that altered mechanical properties of muscle contribute to hypertonia." Brain 104(3): 431449. Dietz, V., M. Wirz, et al. (1998). "Locomotor capacity and recovery of spinal cord fun ction in paraplegic patients: a clinical and electrophysiological evaluation." Electroencephalogr Clin Neurophysiol 109(2): 140153. Dietz, V., M. Wirz, et al. (1998). "Locomotor pattern in paraplegic patients: training effects and recovery of spinal cord function." Spinal Cord 36(6): 380 390. Dietz, V., M. Wirz, et al. (1997). "Locomotion in patients with spinal cord injuries." Phys Ther 77(5): 508516. Dimitrijevic, M. R., Y. Gerasimenko, et al. (1998). "Evidence for a spinal central pattern generator in humans." Ann N Y Acad Sci 860: 360376. Dittuno, P. L. and J. F. Dittuno, Jr. (2001). "Walking index for spinal cord injury (WISCI II): scale revision." Spinal Cord 39(12): 654656.
241 Ditunno, J. F. (2010). "Outcome measures: evolution in clinical trials of neurological/functional recovery in spinal cord injury." Spinal Cord. Ditunno, J. F., Jr. (1994). "American spinal injury standards for neurological and functional classification of spinal cord injury: past, present and future. 1992 Heiner Sell Lecture of the American Spinal Injury Association." J Am Paraplegia Soc 17(1): 711. Ditunno, J. F., Jr., P. L. Ditunno, et al. (2000). "Walking index for spinal cord injury (WISCI): an international multicenter validity and reliability study." Spinal Cord 38(4): 234243. Dobkin, B., H. Barbeau, et al. (2007). "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 Neur al Repair 21(1): 2535. Dobkin, B., V. Edgerton, et al. (1992). "Training induces rhythmic locomotor EMG patterns in subjects with complete SCI." Neurology 42 (Suppl 3): 207208. Dobkin, B. H. (1993). "Neuroplasticity. Key to recovery after central nervous system injury." West J Med 159(1): 5660. Dobkin, B. H., D. Apple, et al. (2003). "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(3): 153167. Dobkin, B. H., R. V. Edgerton, et al. (1992). Sensory Input During Treadmill Training Alters Rhythmic Locomotor EMG Output in Subjects with Complete Spinal Cord Injury. Annual meeting of the Soc iety for NEuroscience, Anaheim, California. Dobkin, B. H., S. Harkema, et al. (1995). "Modulation of locomotor like EMG activity in subjects with complete and incomplete spinal cord injury." J Neurol Rehabil 9(4): 183190. Doperalski, A. E., N. J. Tester, et al. (2011). "Altered obstacle negotiation after low thoracic hemisection in the cat." J Neurotrauma 28(9): 19831993. Douglas, J. R., B. R. Noga, et al. (1993). "The effects of intrathecal administration of excitatory amino acid agonists and antagonists on the initiation of locomotion in the adult cat." J Neurosci 13(3): 9901000. Drew, T. (1988). "Motor cortical cell discharge during voluntary gait modification." Brain Research 457: 181187. Dunglison, R. and T. Lathrop Stedman (1903). A dictionary of m edical science. Philadelphia and New York, Lea Brothers & Co.
242 Dunlop, S. A. (2008). "Activity dependent plasticity: implications for recovery after spinal cord injury." Trends Neurosci 31(8): 410418. Duysens, J., H. W. Van de Crommert, et al. (2002). "A w alking robot called human: lessons to be learned from neural control of locomotion." J Biomech 35(4): 447453. Eccles, J. C., M. Ito, et al. (1967). The cerebellum as a neuronal machine. Berlin, Springer Verlag. Edgerton, V. R., G. Courtine, et al. (2008). "Training locomotor networks." Brain Res Rev 57(1): 241254. Edgerton, V. R., R. D. de Leon, et al. (1997). "Usedependent plasticity in spinal stepping and standing." Adv Neurol 72: 233247. Edgerton, V. R., C. P. Guzman, et al. (1991). Trainability of t he Spinal Cord to Generate Hindlimb Stepping Patterns in Adult Spinalized Cats. Neurobiological Basis of Human Locomotion. M. Shimamura, S. Grillner and V. R. Edgerton. Tokyo, Japan, Japan Scientific Societies Press: 411423. Edgerton, V. R., S. J. Kim, et al. (2006). "Rehabilitative therapies after spinal cord injury." J Neurotrauma 23(34): 560570. Edgerton, V. R., R. D. Leon, et al. (2001). "Retraining the injured spinal cord." J Physiol 533(Pt 1): 1522. Edgerton, V. R., R. R. Roy, et al. (1997). "Does Motor Learning Occur in the Spinal Cord?" The Neuroscientist 3(5): 18. Edgerton, V. R., R. R. Roy, et al. (1991). Recovery of full weight supporting locomotion of the hindlimbs after complete thoracic spinalization of adult and neonatal cats. Plasticity of Motoneuronal Connections. A. Wernig, Elsevier Science Publishers BV: 405 418. Edgerton, V. R., R. R. Roy, et al. (1992). "Potential of adult mammalian lumbosacral spinal cord to execute and acquire improved locomotion in the absence of supraspinal input ." J Neurotrauma 9 Suppl 1: S119128. Edgerton, V. R., N. J. Tillakaratne, et al. (2004). "Plasticity of the spinal neural circuitry after injury." Annu Rev Neurosci 27: 145167. Eidelberg, E., L. H. Nguyen, et al. (1986). "Recovery of locomotor function after hemisection of the spinal cord in cats." Brain Res Bull 16(4): 507515. Eidelberg, E., J. G. Walden, et al. (1981). "Locomotor control in macaque monkeys." Brain 104(Pt 4): 647663.
243 Ellaway, P. H., M. Catley, e t al. (2007). "Review of physiological motor outcome measures in spinal cord injury using transcranial magnetic stimulation and spinal reflexes." J Rehabil Res Dev 44(1): 6976. Ellaway, P. H., A. Kuppuswamy, et al. (2011). "Development of quantitative and sensitive assessments of physiological and functional outcome during recovery from spinal cord injury: a clinical initiative." Brain Res Bull 84(45): 343357. Emken, J. L., R. Benitez, et al. (2007). "Motor adaptation as a greedy optimization of error an d effort." J Neurophysiol 97(6): 39974006. Fawcett, J. W., A. Curt, et al. (2007). "Guidelines for the conduct of clinical trials for spinal cord injury as developed by the ICCP panel: spontaneous recovery after spinal cord injury and statistical power needed for therapeutic clinical trials." Spinal Cord 45(3): 190205. Felten, D. L. and A. N. Shettz (2010). Netter's Atlas of Neuroscience. Philadelphia, USA, Saunders, Elsevier. Field Fote, E. C., S. D. Lindley, et al. (2005). "Locomotor training approaches for individuals with spinal cord injury: a preliminary report of walking related outcomes." J Neurol Phys Ther 29(3): 127137. Field Fote, E. C. and K. E. Roach (2011). "Influence of a locomotor training approach on walking speed and distance in people wi th chronic spinal cord injury: a randomized clinical trial." Phys Ther 91(1): 4860. Field Fote, E. C. and D. Tepavac (2002). "Improved intralimb coordination in people with incomplete spinal cord injury following training with body weight support and elec trical stimulation." Phys Ther 82(7): 707715. Filli, L., B. Zorner, et al. (2011). "Motor deficits and recovery in rats with unilateral spinal cord hemisection mimic the BrownSequard syndrome." Brain 134(Pt 8): 22612273. Fitts, P. M. (1964). Perceptual Motor Skill Learning. Categories of Human Learning. A. W. Melton. New York, Academic Press243285. Fitzpatrick, R., D. Burke, et al. (1994). "Task dependent reflex responses and movement illusions evoked by galvanic vestibular stimulation in standing humans." J Physiol 478 ( Pt 2): 363372. Fitzpatrick, R. C. and B. L. Day (2004). "Probing the human vestibular system with galvanic stimulation." J Appl Physiol 96(6): 23012316. Forssberg, H. (1986). Development and integration of human locomotor functions. D evelopment and Plasticity of the Mammalian Spinal Cord. M. Goldberger, A. Gorio and M. Murray. Padova, Livinia Press. 3: 5363.
244 Franchignoni, F., F. Horak, et al. (2010). "Using psychometric techniques to improve the Balance Evaluation Systems Test: the mi ni BESTest." J Rehabil Med 42(4): 323331. Freund, P., J. Rothwell, et al. (2011). "Corticomotor representation to a human forearm muscle changes following cervical spinal cord injury." Eur J Neurosci 34(11): 18391846. Freund, P., E. Schmidlin, et al. (2006). "NogoA specific antibody treatment enhances sprouting and functional recovery after cervical lesion in adult primates." Nat Med 12(7): 790792. Funase, K., T. Higashi, et al. (1996). "Evident difference in the excitability of the motoneuron pool betw een normal subjects and patients with spasticity assessed by a new method using H reflex and M response." Neurosci Lett 203(2): 127130. Fung, J. and H. Barbeau (1994). "Effects of conditioning cutaneomuscular stimulation on the soleus H reflex in normal and spastic paretic subjects during walking and standing." J Neurophysiol 72(5): 20902104. Furlan, J. C., M. G. Fehlings, et al. (2008). "Motor and sensory assessment of patients in clinical trials for pharmacological therapy of acute spinal cord injury: p sychometric properties of the ASIA Standards." J Neurotrauma 25(11): 12731301. Garcia Rill, E. and R. D. Skinner (1987). "The mesencephalic locomotor region. I. Activation of a medullary projection site." Brain Research 411: 112. Garcia Rill, E. and R. D. Skinner (1987). "The mesencephalic locomotor region. II. Projections to reticulospinal neurons." Brain Research 411: 1320. Geisler, F. H., W. P. Coleman, et al. (2001). "The Sygen multicenter acute spinal cord injury study." Spine 26(24): 8798. Girgi s, J., D. Merrett, et al. (2007). "Reaching training in rats with spinal cord injury promotes plasticity and task specific recovery." Brain 130(Pt 11): 29933003. Goldberg, J. M., C. Fernandez, et al. (1982). "Responses of vestibular nerve afferents in the squirrel monkey to externally applied galvanic currents." Brain Research(252): 156160. Gorassini, M. A., J. A. Norton, et al. (2009). "Changes in locomotor muscle activity after treadmill training in subjects with incomplete spinal cord injury." J Neurop hysiol 101(2): 969979. Greenlee, M. W. (2000). "Human cortical areas underlying the perception of optic flow: brain imaging studies." Int Rev Neurobiol 44: 269292.
245 Grillner, S. and P. Wallen (1985). "Central pattern generators for locomotion, with specia l reference to vertebrates." Annu Rev Neurosci 8: 233261. Grillner, S. and P. Zangger (1979). "On the central generation of locomotion in the low spinal cat." Exp Brain Res 34(2): 241261. Grosse, P. and P. Brown (2003). "Acoustic startle evokes bilaterally synchronous oscillatory EMG activity in the healthy human." J Neurophysiol 90(3): 16541661. GuzmanLopez, J., Y. Buisson, et al. (2011). "Interaction between vestibulospinal and corticospinal systems: a combined caloric and transcranial magnetic stimulation study." Exp Brain Res 214(1): 3745. Hagg, T. (2009). "From neurotransmitters to neurotrophic factors to neurogenesis." Neuroscientist 15(1): 2027. Hall, E. D. and R. J. Traystman (2009). "Role of animal studies in the design of clinical t rials." Front Neurol Neurosci 25: 1033. Hall, K. M., M. E. Cohen, et al. (1999). "Characteristics of the Functional Independence Measure in traumatic spinal cord injury." Arch Phys Med Rehabil 80(11): 14711476. Harkema, S., Y. Gerasimenko, et al. (2011). "Effect of epidural stimulation of the lumbosacral spinal cord on voluntary movement, standing, and assisted stepping after motor complete paraplegia: a case study." Lancet 377(9781): 19381947. Harkema, S. J. (2001). "Neural plasticity after human spinal cord injury: application of locomotor training to the rehabilitation of walking." Neuroscientist 7(5): 455468. Harkema, S. J., A. L. Behrman, et al. (2011). Locomotor Training: Principles and Practice. New York, NY, Oxford University Press. Harkema, S. J ., M. SchmidtRead, et al. (2011). "Balance and Ambulation Improvements in Individuals With Chronic Incomplete Spinal Cord Injury Using Locomotor Training Based Rehabilitation." Arch Phys Med Rehabil. Hassanpour, K., S. Hotz Boendermaker, et al. (2011). "L ow depressive symptoms in acute spinal cord injury compared to other neurological disorders." J Neurol. Hayes, K. C., J. T. Hsieh, et al. (2000). "Classifying incomplete spinal cord injury syndromes: algorithms based on the International Standards for Neurological and Functional Classification of Spinal Cord Injury Patients." Arch Phys Med Rehabil 81(5): 644652. Heathcote, A., S. Brown, et al. (2000). "The power law repealed: the case for an exponential law of practice." Psychon Bull Rev 7(2): 185207.
246 Hebb, D. (1949). The organization of behavior: a neuropsychological theory. New York, Wiley. Helgren, M. E. and M. E. Goldberger (1993). "The recovery of postural reflexes and locomotion following low thoracic hemisection in adult cats involves compensati on by undamaged primary afferent pathways." Experimental Neurology 123(1): 1734. Hidler, J., D. Nichols, et al. (2009). "Multicenter randomized clinical trial evaluating the effectiveness of the Lokomat in subacute stroke." Neurorehabil Neural Repair 23(1): 5 13. Hill, R. L., Y. P. Zhang, et al. (2009). "Anatomical and functional outcomes following a precise, graded, dorsal laceration spinal cord injury in C57BL/6 mice." J Neurotrauma 26(1): 115. Horak, F. B. (1987). "Clinical measurement of postural cont rol in adults." Phys Ther 67(12): 18811885. Horak, F. B. and L. M. Nashner (1986). "Central programming of postural movements: adaptation to altered support surface configurations." J Neurophysiol 55(6): 13691381. Hori, A., A. Yasuhara, et al. (1986). "B link reflex elicited by auditory stimulation in the rabbit." J Neurol Sci 76(1): 4959. Huang, T. T. and W. S. Wang (2009). "Comparison of three established measures of fear of falling in community dwelling older adults: psychometric testing." Int J Nurs S tud 46(10): 13131319. Humes, K. R., N. A. Jones, et al. (2011). Overview of Race and Hispanic Origin: 2010. 2010 Census Briefs. U. S. C. Bureau, U.S. Department of Commerce, Economics and Statistics Administration. Ijspeert, A. J. (2008). "Central pattern generators for locomotion control in animals and robots: a review." Neural Netw 21(4): 642653. Ilic, T. V., M. Potter Nerger, et al. (2011). "Startle stimuli exert opposite effects on human cortical and spinal motor system excitability in leg muscles." P hysiol Res 60 Suppl 1: S101106. Ito, M. (2006). "Cerebellar circuitry as a neuronal machine." Prog Neurobiol 78(35): 272303. Ivanenko, Y. P., R. E. Poppele, et al. (2009). "Distributed neural networks for controlling human locomotion: lessons from normal and SCI subjects." Brain Res Bull 78(1): 1321.
247 Jackson, A. B., C. T. Carnel, et al. (2008). "Outcome measures for gait and ambulation in the spinal cord injury population." J Spinal Cord Med 31(5): 487499. Jahn, K., A. Deutschlander, et al. (2008). "Su praspinal locomotor control in quadrupeds and humans." Prog Brain Res 171: 353362. Jahn, K., A. Deutschlander, et al. (2008). "Imaging human supraspinal locomotor centers in brainstem and cerebellum." Neuroimage 39(2): 786792. Jakob, W., M. Wirz, et al. (2009). "Difficulty of elderly SCI subjects to translate motor recovery --"body function" --into daily living activities." J Neurotrauma 26(11): 20372044. Jankelowitz, S. K. and J. G. Colebatch (2004). "The acoustic startle reflex in ischemic stroke." Neurology 62(1): 114116. Jannings, W. and J. Pryor (2012). "The experiences and needs of persons with spinal cord injury who can walk." Disabil Rehabil. Jefferson, S. C., N. J. Tester, et al. (2011). "Chondroitinase ABC promotes recovery of adaptive limb movem ents and enhances axonal growth caudal to a spinal hemisection." J Neurosci 31(15): 57105720. JohansenBerg, H., H. Dawes, et al. (2002). "Correlation between motor improvements and altered fMRI activity after rehabilitative therapy." Brain 125(Pt 12): 27 312742. Johnston, M. V. (2009). "Plasticity in the developing brain: implications for rehabilitation." Dev Disabil Res Rev 15(2): 94 101. Jordan, L. M., J. Liu, et al. (2008). "Descending command systems for the initiation of locomotion in mammals." Brain Res Rev 57(1): 183191. Kafitz, K. W., C. R. Rose, et al. (1999). "Neurotrophinevoked rapid excitation through TrkB receptors." Nature 401(6756): 918921. Kandel, E., J. Schwartz, et al. (2000). Principles of Neural Science, Mc Graw Hill Companies, Inc. Kato, M. (1989). "Chronically isolated lumbar half spinal cord, produced by hemisection and longitudinal myelotomy, generates locomotor activities of the ipsilateral hindlimb of the cat." Neuroscience Letters 98: 149153. Kato, M., S. Murakami, et al. (198 5). "Recovery of postural control following chronic bilateral hemisections at different spinal cord levels in adult cats." Exp Neurol 90(2): 350364.
248 Kato, M., S. Murakami, et al. (1984). "Disruption of foreand hindlimb coordination during overground loc omotion in cats with bilateral serial hemisection of the spinal cord." Neuroscience Research 2: 2747. Keith, R., C. Granger, et al. (1987). The Functional Independence Measure; a new tool for rehabilitation. Advances in Clinical Rehabilitation. N. Eisenberg and R. Grezesiak. New York, Springer. Kerchner, G. A. and R. A. Nicoll (2008). "Silent synapses and the emergence of a postsynaptic mechanism for LTP." Nat Rev Neurosci 9(11): 813825. Kirshblum, S., A. Botticello, et al. (2011). "The impact of sacral s ensory sparing in motor complete spinal cord injury." Arch Phys Med Rehabil 92(3): 376383. Kirshblum, S. C. and K. C. O'Connor (1998). "Predicting neurologic recovery in traumatic cervical spinal cord injury." Arch Phys Med Rehabil 79(11): 14561466. Kirs hblum, S. C. and K. C. O'Connor (2000). "Levels of spinal cord injury and predictors of neurologic recovery." Phys Med Rehabil Clin N Am 11(1): 127. Klimstra, M. and E. P. Zehr (2008). "A sigmoid function is the best fit for the ascending limb of the Hoff mann reflex recruitment curve." Exp Brain Res 186(1): 93105. Knikou, M., C. A. Angeli, et al. (2009). "Soleus H reflex modulation during body weight support treadmill walking in spinal cord intact and injured subjects." Exp Brain Res 193(3): 397407. Kobayashi, M. and A. Pascual Leone (2003). "Transcranial magnetic stimulation in neurology." Lancet Neurol 2(3): 145156. Koch, M. (1999). "The neurobiology of startle." Prog Neurobiol 59(2): 107128. Koehler, P. J. and L. J. Endtz (1986). "The BrownSequard s yndrome. True or false?" Arch Neurol 43(9): 921 924. Kofler, M., J. Muller, et al. (2006). "Auditory startle responses as a probe of brainstem function in healthy subjects and patients with movement disorders." Suppl Clin Neurophysiol 58: 232248. Kuhtz Bu schbeck, J. P., A. BoczekFuncke, et al. (1996). "Recovery of Locomotion after Spinal Cord Hemisection: An X Ray Study of the Cat Hindlimb." Experimental Neurology 137: 212224. Kuhtz, B. J., F. A. Boczek, et al. (1996). "Recovery of locomotion after spinal cord hemisection: an X ray study of the cat hindlimb." 137(2): 212224.
249 Kumru, H., M. Kofler, et al. (2009). "Brainstem reflexes are enhanced following severe spinal cord injury and reduced by continuous intrathecal baclofen." Neurorehabil Neural Repair 23(9): 921927. Kumru, H., J. Vidal, et al. (2008). "Exaggerated auditory startle responses in patients with spinal cord injury." J Neurol 255(5): 703709. Kurtzke, J. F. (1977). "Epidemiology of spinal cord injury." Neurol Neurocir Psiquiatr 18(23 Suppl ): 157 191. Ladouceur, M., H. Barbeau, et al. (2003). "Kinematic adaptations of spinal cordinjured subjects during obstructed walking." Neurorehabil Neural Repair 17(1): 2531. Lajoie, K., J. E. Andujar, et al. (2010). "Neurons in area 5 of the posterior parietal cortex in the cat contribute to interlimb coordination during visually guided locomotion: a role in working memory." J Neurophysiol 103(4): 22342254. Lajoie, Y., H. Barbeau, et al. (1999). "Attentional requirements of walking in spinal cord injur ed patients compared to normal subjects." Spinal Cord 37(4): 245250. Lajoie, Y. and S. P. Gallagher (2004). "Predicting falls within the elderly community: comparison of postural sway, reaction time, the Berg balance scale and the Activities specific Bala nce Confidence (ABC) scale for comparing fallers and nonfallers." Arch Gerontol Geriatr 38(1): 1126. Lam, T., K. Pauhl, et al. (2011). "Using robot applied resistance to augment body weight supported treadmill training in an individual with incomplete spinal cord injury." Phys Ther 91(1): 143151. Lapointe, R., Y. Lajoie, et al. (2001). "Functional community ambulation requirements in incomplete spinal cord injured subjects." Spinal Cord 39(6): 327335. Le Ray, D., L. Juvin, et al. (2011). "Chapter 4-sup raspinal control of locomotion: the mesencephalic locomotor region." Prog Brain Res 188: 5170. Lemon, R. N. (2008). "Descending pathways in motor control." Annu Rev Neurosci 31: 195218. Lerner Frankiel, M. B., S. Varcas, et al. (1986). "Functional community ambulation: what are your criteria?" Clinical Management in Physical Therapy 6: 1215. Leroux, A., J. Fung, et al. (1999). "Adaptation of the walking pattern to uphill walking in normal and spinal cord injured subjects." Exp Brain Res 126(3): 359368. Leroux, A., J. Fung, et al. (2002). "Postural adaptation to walking on inclined surfaces: I. Normal strategies." Gait Posture 15(1): 6474.
250 Leroux, A., J. Fung, et al. (2006). "Postural adaptation to walking on inclined surfaces: II. Strategies following s pinal cord injury." Clin Neurophysiol 117(6): 12731282. Levin, M. F., J. A. Kleim, et al. (2009). "What do motor "recovery" and "compensation" mean in patients following stroke?" Neurorehabil Neural Repair 23(4): 313319. Liechti, M., R. Muller, et al. (2 008). "Vestibulospinal responses in motor incomplete spinal cord injury." Clin Neurophysiol. Lipschitz, R. (1967). "Associated injuries and complications of stab wounds of the spinal cord." Paraplegia 5(2): 7582. Lipschitz, R. (1973). "Stab wounds of the spine." Proc Veterans Adm Spinal Cord Inj Conf(19): 165169. Little, J. W. and E. Halar (1985). "Temporal course of motor recovery after BrownSequard spinal cord injuries." Paraplegia 23(1): 3946. Little, J. W. and E. M. Halar (1985). "H reflex changes f ollowing spinal cord injury." Arch Phys Med Rehabil 66(1): 1922. Lo, Y. L. and S. Fook Chong (2004). "A transcranial magnetic stimulation study of the ipsilateral silent period in lower limb muscles." Neurosci Lett 368(3): 337340. Lovely, R. G., R. J. Gr egor, et al. (1986). "Effects of training on the recovery of full weight bearing stepping in the adult spinal cat." Exp Neurol 92(2): 421435. Lovely, R. G., R. J. Gregor, et al. (1990). "Weight bearing hindlimb stepping in treadmill exercised adult spinal cats." Brain Res 514(2): 206218. Lu, B. and A. Figurov (1997). "Role of neurotrophins in synapse development and plasticity." Rev Neurosci 8(1): 112. Lund, S. and C. Broberg (1983). "Effects of different head positions on postural sway in man induced by a reproducible vestibular error signal." Acta Physiol Scand 117(2): 307309. Lunenburger, L., M. Bolliger, et al. (2006). "Modulation of locomotor activity in complete spinal cord injury." Exp Brain Res 174(4): 638 646. Lynskey, J. V., A. Belanger, et al. (2008). "Activity dependent plasticity in spinal cord injury." J Rehabil Res Dev 45(2): 229240. Macdonell, R. A. and G. A. Donnan (1995). "Magnetic cortical stimulation in acute spinal cord injury." Neurology 45(2): 303306. MacKay Lyons, M. (2002). "Central pattern generation of locomotion: a review of the evidence." Phys Ther 82(1): 6983.
251 Magnuson, D. S., R. R. Smith, et al. (2009). "Swimming as a model of task specific locomotor retraining after spinal cord injury in the rat." Neurorehabil Neural Repair 23(6): 535545. Manella, K. J., J. Torres, et al. (2010). "Restoration of walking function in an individual with chronic complete (AIS A) spinal cord injury." J Rehabil Med 42(8): 795798. Marino, R. J. (2007). "Domains of outcomes in spinal cord injury for clinical trials to improve neurological function." J Rehabil Res Dev 44(1): 113122. Marti, A., J. D. Reinhardt, et al. (2012). "To work or not to work: labour market participation of people with spinal cord injury living in Switzerland." Spinal Cord. Martins, F., F. Freitas, et al. (1998). "Spinal cord injuries --epidemiology in Portugal's central region." Spinal Cord 36(8): 574578. McClellan, A. D. and K. A. Sigvardt (1988). "Features of entrainment of spinal pattern generators for locomotor activity in the lamprey spinal cord." J Neurosci 8(1): 133145. McKay, W. B., D. C. Lee, et al. (2005). "Neurophysiological examination of the corticospinal system and voluntary motor control in motor incomplete human spinal cord injury." Exp Brain Res 163(3): 379 387. McKay, W. B., D. S. Stokic, et al. (1997). "Assessment of corticospinal function in spinal cord injury using transcranial motor cortex stimulation: a review." J Neurotrauma 14(8): 539548. McKinley, W., K. Santos, et al. (2007). "Incidence and outcom es of spinal cord injury clinical syndromes." J Spinal Cord Med 30(3): 215224. McKinley, W. O., R. T. Seel, et al. (1999). "Nontraumatic spinal cord injury: incidence, epidemiology, and functional outcome." Arch Phys Med Rehabil 80(6): 619623. Mehrholz, J., J. Kugler, et al. (2008). "Locomotor training for walking after spinal cord injury." Cochrane Database Syst Rev(2): CD006676. Mehrholz, J., J. Kugler, et al. (2008). "Locomotor training for walking after spinal cord injury." Spine (Phila Pa 1976) 33(21): E768777. Memmert, D. (2006). "Long term effects of type of practice on the learning and transfer of a complex motor skill." Percept Mot Skills 103(3): 912916. Misiaszek, J. E. (2003). "The H reflex as a tool in neurophysiology: its limitations and use s in understanding nervous system function." Muscle Nerve 28(2): 144160.
252 Moreh, E., Z. Meiner, et al. (2009). "Spinal decompression sickness presenting as partial BrownSequard syndrome and treated with robotic assisted body weight support treadmill training." J Rehabil Med 41(1): 8889. Morganti, B., G. Scivoletto, et al. (2005). "Walking index for spinal cord injury (WISCI): criterion validation." Spinal Cord 43(1): 2733. Mori, S., T. Matsui, et al. (1998). "Cerebellar induced locomotion: reticulospinal control of spinal rhythm generating mechanism in cats." Ann N Y Acad Sci 860: 94105. Mori, S., K. Matsuyama, et al. (1992). "Neuronal constituents of postural and locomotor control systems and their interactions in cats." Brain Dev 14 Suppl: S109120. Mo rton, S. M. and A. J. Bastian (2004). "Cerebellar control of balance and locomotion." Neuroscientist 10(3): 247259. Muller, R. and V. Dietz (2006). "Neuronal function in chronic spinal cord injury: divergence between locomotor and flexionand H reflex ac tivity." Clin Neurophysiol 117(7): 14991507. Mller, R. M. (2006). Lokomotions und Reflexaktivitt bei kompletter Querschnittlhmung: Einflussfaktoren und zeitliche Entwickling. PhD, Eidgenssische Technische Hochschule ETH. Mumenthaler, M. and H. Mattl e (2002). Grundkurs Neurologie, Illustriertes Basiswissen fr das Studium. Stuttgart, Thieme. Murray, M. and M. E. Goldberger (1974). "Restitution of function and collateral sprouting in the cat spinal cord: the partially hemisected animal." J Comp Neurol 158(1): 1936. Musselman, K., K. Brunton, et al. (2011). "Spinal cord injury functional ambulation profile: a new measure of walking ability." Neurorehabil Neural Repair 25(3): 285293. Musselman, K. E., K. Fouad, et al. (2009). "Training of walking skills overground and on the treadmill: case series on individuals with incomplete spinal cord injury." Phys Ther 89(6): 601611. Musselman, K. E. and J. F. Yang (2007). "Walking tasks encountered by urbandwelling adults and persons with incomplete spinal cord injuries." J Rehabil Med 39(7): 567574. Myers, A. M., P. C. Fletcher, et al. (1998). "Discriminative and evaluative properties of the activities specific balance confidence (ABC) scale." J Gerontol A Biol Sci Med Sci 53(4): M287294.
253 Nadeau, S., A. B. Ars enault, et al. (1999). "Analysis of the clinical factors determining natural and maximal gait speeds in adults with a stroke." Am J Phys Med Rehabil 78(2): 123130. Nadeau, S., G. Jacquemin, et al. (2010). "Spontaneous motor rhythms of the back and legs in a patient with a complete spinal cord transection." Neurorehabil Neural Repair 24(4): 377383. Nathan, P. W. (1994). "Effects on movement of surgical incisions into the human spinal cord." Brain 117: 337346. Nathan, P. W., M. Smith, et al. (1996). "Vesti bulospinal, reticulospinal and descending propriospinal nerve fibres in man." Brain 119 ( Pt 6): 18091833. Nathan, P. W. and M. C. Smith (1973). "Effects of two unilateral cordotomies on the motility of the lower limbs." Brain 96(3): 471494. Netter, F. (2006). Atlas der Anatomie des Menschen. Stuttgart, Thieme. New, P. W. and V. Sundararajan (2008). "Incidence of nontraumatic spinal cord injury in Victoria, Australia: a populationbased study and literature review." Spinal Cord 46(6): 406411. Newpher, T. M. and M. D. Ehlers (2008). "Glutamate receptor dynamics in dendritic microdomains." Neuron 58(4): 472497. Ni, W., Y. M. Li, et al. (2005). "[Evaluation of neurological function following establishment of spinal cord hemisection model in rhesus]." Sich uan Da Xue Xue Bao Yi Xue Ban 36(3): 328330. Nicol, D. J., M. H. Granat, et al. (1995). "Evidence for a human spinal stepping generator." Brain Res 684(2): 230232. Nieuwenhuijzen, P. H., M. W. Horstink, et al. (2006). "Startle responses in Parkinson pati ents during human gait." Exp Brain Res 171(2): 215224. Nieuwenhuijzen, P. H., A. M. Schillings, et al. (2000). "Modulation of the startle response during human gait." J Neurophysiol 84(1): 6574. Nilsson, J., A. Thorstensson, et al. (1985). "Changes in leg movements and muscle activity with speed of locomotion and mode of progression in humans." Acta Physiol Scand 123(4): 457475. Nooijen, C. F., N. Ter Hoeve, et al. (2009). "Gait quality is improved by locomotor training in individuals with SCI regardless of training approach." J Neuroeng Rehabil 6: 36.
254 NSCISC (2011). Spinal Cord Injury Facts and Figures at a Glance, National Spinal Cord Injury Statistical Center, Birmingham, Alabama. Nudo, R. J. (2003). "Adaptive plasticity in motor cortex: implications f or rehabilitation after brain injury." J Rehabil Med(41 Suppl): 7 10. Nudo, R. J., G. W. Milliken, et al. (1996). "Use dependent alterations of movement representations in primary motor cortex of adult squirrel monkeys." J Neurosci 16(2): 785807. Orlovsky, G. N. (1972). "The effect of different descending systems on flexor and extensor activity during locomotion." Brain Research 40: 359371. Ottersen, O. P. (2010). "How hardwired is the brain? Technological advances provide new insight into brain malleabil ity and neurotransmission." Nutr Rev 68 Suppl 2: S60 64. Pascual, M. G., V. S. Alcacer, et al. (2011). "BrownSequardplus syndrome after a stab injury." NeuroRehabilitation 29(4): 353357. Patla, A. E. and A. Shumway Cook (1999). "Dimensions of mobility: defining the complexity and difficulty associated with community mobility." Journal of Aging and Physical Activity 7: 7 19. Peacock, W. J., R. D. Shrosbree, et al. (1977). "A review of 450 stabwounds of the spinal cord." S Afr Med J 51(26): 961964. Pepi n, A., M. Ladouceur, et al. (2003). "Treadmill walking in incomplete spinal cord injured subjects: 2. Factors limiting the maximal speed." Spinal Cord 41(5): 271279. Pepin, A., K. E. Norman, et al. (2003). "Treadmill walking in incomplete spinal cord inju red subjects: 1. Adaptation to changes in speed." Spinal Cord 41(5): 257270. Perez, M. A., J. Lundbye Jensen, et al. (2006). "Changes in corticospinal drive to spinal motoneurones following visuomotor skill learning in humans." J Physiol 573(Pt 3): 8438 55. Petrini, E. M., J. Lu, et al. (2009). "Endocytic trafficking and recycling maintain a pool of mobile surface AMPA receptors required for synaptic potentiation." Neuron 63(1): 92105. Phadke, C. P., C. T. Robertson, et al. (2012). "Upper extremity H ref lex measurement post stroke: Reliability and inter limb differences." Clin Neurophysiol. Phadke, C. P., F. J. Thompson, et al. (2010). "Soleus H reflex modulation after motor incomplete spinal cord injury: effects of body position and walking speed." J Spi nal Cord Med 33(4): 371378.
255 Pouw, M. H., H. van de Meent, et al. (2010). "Relevance of the diagnosis traumatic cervical Brown Sequardplus syndrome: an analysis based on the neurological and functional recovery in a prospective cohort of 148 patients." Spinal Cord. Powell, L. E. and A. M. Myers (1995). "The Activities specific Balance Confidence (ABC) Scale." J Gerontol A Biol Sci Med Sci 50A(1): M2834. Ramon y Cajal, S. (1928). Degeneration and Regeneration of the Nervous System. London, Oxford Universit y Press. Robinett, C. S. and M. A. Vondran (1988). "Functional ambulation velocity and distance requirements in rural and urban communities. A clinical report." Phys Ther 68(9): 13711373. Rojas Vega, S., T. Abel, et al. (2008). "Impact of exercise on neur oplasticity related proteins in spinal cord injured humans." Neuroscience 153(4): 10641070. Rossignol, S. (1975). "Startle responses recorded in the leg of man." Electroencephalogr Clin Neurophysiol 39(4): 389397. Rossignol, S. (2002). "Locomotion and its recovery after spinal injury in animal models." Neurorehabil Neural Repair 16(2): 201206. Rossignol, S., G. Barriere, et al. (2009). "Reexpression of locomotor function after partial spinal cord injury." Physiology (Bethesda) 24: 127139. Rossini, P. M., A. T. Barker, et al. (1994). "Noninvasive electrical and magnetic stimulation of the brain, spinal cord and roots: basic principles and procedures for routine clinical application. Report of an IFCN committee." Electroencephalogr Clin Neurophysiol 91(2): 79 92. Rossini, P. M. and F. Pauri (1999). "Central motor conduction time studies." Electroencephalogr Clin Neurophysiol Suppl 51: 199211. Roth, E. J., T. Park, et al. (1991). "Traumatic cervical BrownSequard and BrownSequardplus syndromes: the sp ectrum of presentations and outcomes." Paraplegia 29(9): 582589. Rothwell, J. C. (1997). "Techniques and mechanisms of action of transcranial stimulation of the human motor cortex." J Neurosci Methods 74(2): 113122. Schmidt, R. and T. Lee (2005). Motor c ontrol and learning: a behavioral emphasis. Champaign, IL, Human Kinetics. Schubert, M., A. Curt, et al. (1999). "Voluntary control of human gait: conditioning of magnetically evoked motor responses in a precision stepping task." Exp Brain Res 126(4): 583588.
256 Schuld, C., J. Wiese, et al. (2011). "Computer implementation of the international standards for neurological classification of spinal cord injury for consistent and efficient derivation of its subscores including handling of data from not testable se gments." J Neurotrauma 29(3): 453461. Schwarzkopf, S. B., L. McCoy, et al. (1993). "Test retest reliability of prepulse inhibition of the acoustic startle response." Biol Psychiatry 34(12): 896900. Scivoletto, G., S. Farchi, et al. (2011). "Traumatic and non traumatic spinal cord lesions: an Italian comparison of neurological and functional outcomes." Spinal Cord 49(3): 391396. Shah, P., Y. Gerasimenko, et al. (unpublished submitted for publication Feb 2012). "Variability in Step Training Enhances Locomotor Recovery After a Spinal Cord Injury." European Journal of Neuroscience. Sherrington, C. S. (1910). "Flexion reflex of the limb, crossed extension reflex, and reflex stepping and standing." Journal of Physiology 40: 28121. Shik, M. L., F. V. Severin, et al. (1969). "Control of walking and running by means of electrical stimulation of the mesencephalon." Electroencephalogr Clin Neurophysiol 26(5): 549. Shirota, Y., M. Hamada, et al. (2012). "Increased primary motor cortical excitability by a single pul se transcranial magnetic stimulation over the supplementary motor area." Exp Brain Res. Shumway Cook, A., A. Patla, et al. (2003). "Environmental components of mobility disability in community living older persons." J Am Geriatr Soc 51(3): 393398. Shumway Cook, A., A. E. Patla, et al. (2002). "Environmental demands associated with community mobility in older adults with and without mobility disabilities." Phys Ther 82(7): 670681. Shumway Cook, A. and M. Woollacott (2001). Motor Control: Theory and Practic al Applications. Philadelphia, Lippincott Williams and Wilkins. Sisto, S. A., G. F. Forrest, et al. (2008). "Technology for mobility and quality of life in spinal cord injury." IEEE Eng Med Biol Mag 27(2): 5668. Spiess, M. R., J. P. Jaramillo, et al. (201 2). "Unexpected Recovery after Robotic Locomotor Training at Physiologic Stepping Speed a single case design." Arch Phys Med Rehabil. Steeves, J. D. and L. M. Jordan (1980). "Localization of a descending pathway in the spinal cord which is necessary for controlled treadmill locomotion." Neuroscience Letters 20: 283288.
257 Steeves, J. D., D. Lammertse, et al. (2007). "Guidelines for the conduct of clinical trials for spinal cord injury (SCI) as developed by the ICCP panel: clinical trial outcome measures." S pinal Cord 45(3): 206221. Steffen, T. and M. Seney (2008). "Test retest reliability and minimal detectable change on balance and ambulation tests, the 36item short form health survey, and the unified Parkinson disease rating scale in people with parkinsonism." Phys Ther 88(6): 733746. Sterr, A., S. Freivogel, et al. (2002). "Neurobehavioral aspects of recovery: assessment of the learned nonuse phenomenon in hemiparetic adolescents." Arch Phys Med Rehabil 83(12): 17261731. Stuart, D. G. and H. Hultborn ( 2008). "Thomas Graham Brown (1882--1965), Anders Lundberg (1920), and the neural control of stepping." Brain Res Rev 59(1): 7495. Sturt, R. N., A. E. Holland, et al. (2009). "Walking ability at discharge from inpatient rehabilitation in a cohort of nont raumatic spinal cord injury patients." Spinal Cord 47(10): 763768. Surkin, J., B. J. Gilbert, et al. (2000). "Spinal cord injury in Mississippi. Findings and evaluation, 19921994." Spine (Phila Pa 1976) 25(6): 716721. Szabo, I. (1965). "Elicitability of the acoustic startle reaction after brain stem lesions." Acta Physiol. Acad. Sci. Hung 27: 155165. Takakusaki, K., N. Tomita, et al. (2008). "Substrates for normal gait and pathophysiology of gait disturbances with respect to the basal ganglia dysfunction." J Neurol 255 Suppl 4: 1929. Talley, K. M., J. F. Wyman, et al. (2008). "Psychometric properties of the activities specific balance confidence scale and the survey of activities and fear of falling in older women." J Am Geriatr Soc 56(2): 328333. Taylor, R. G. and J. R. Gleave (1957). "Incomplete spinal cord injuries; with BrownSequard phenomena." J Bone Joint Surg Br 39B(3): 438 450. Tester, N. J. and D. R. Howland (2008). "Chondroitinase ABC improves basic and skilled locomotion in spinal cord injured cats." Exp Neurol 209(2): 483496. Tester, N. J., A. H. Plaas, et al. (2007). "Effect of body temperature on chondroitinase ABC's ability to cleave chondroitin sulfate glycosaminoglycans." J Neurosci Res 85(5): 11101118. Thomas, S. L. and M. A. Go rassini (2005). "Increases in corticospinal tract function by treadmill training after incomplete spinal cord injury." J Neurophysiol 94(4): 28442855.
258 Thota, A., S. Carlson, et al. (2001). "Recovery of locomotor function after treadmill training of incomplete spinal cord injured rats." Biomed Sci Instrum 37: 6367. Trimble, M. H., A. L. Behrman, et al. (2001). "Acute effects of locomotor training on overground walking speed and H reflex modulation in individuals with incomplete spinal cord injury." J Spinal Cord Med 24(2): 7480. Vaillancourt, D. E., K. R. Thulborn, et al. (2003). "Neural basis for the processes that underlie visually guided and internally guided force control in humans." J Neurophysiol 90(5): 33303340. Valls Sole, J. (2012). "Assessment of excitability in brainstem circuits mediating the blink reflex and the startle reaction." Clin Neurophysiol 123(1): 1320. Valls Sole, J., H. Kumru, et al. (2008). "Interaction between startle and voluntary reactions in humans." Exp Brain Res 187(4): 497507. Valsamis, B. and S. Schmid (2011). "Habituation and prepulse inhibition of acoustic startle in rodents." J Vis Exp(55): e3446. van Asbeck, F. W., M. W. Post, et al. (2000). "An epidemiological description of spinal cord injuries in The Netherlands in 1994." Spinal Cord 38(7): 420424. van Hedel, H. J., V. Dietz, et al. (2007). "Assessment of walking speed and distance in subjects with an incomplete spinal cord injury." Neurorehabil Neural Repair 21(4): 295301. van Hedel, H. J., C. Murer, et al. (2007). "The amplitude of lower leg motor evoked potentials is a reliable measure when controlled for torque and motor task." J Neurol 254(8): 10891098. van Hedel, H. J., B. Wirth, et al. (2010). "Ankle motor skill is intact in spinal cord injury, unlike strok e: implications for rehabilitation." Neurology 74(16): 12711278. van Hedel, H. J., B. Wirth, et al. (2005). "Limits of locomotor ability in subjects with a spinal cord injury." Spinal Cord 43(10): 593603. van Hedel, H. J., M. Wirz, et al. (2006). "Improv ing walking assessment in subjects with an incomplete spinal cord injury: responsiveness." Spinal Cord 44(6): 352356. van Hedel, H. J., M. Wirz, et al. (2005). "Assessing walking ability in subjects with spinal cord injury: validity and reliability of 3 w alking tests." Arch Phys Med Rehabil 86(2): 190196. Velozo, C. A., G. Kielhofner, et al. (1999). "The use of Rasch analysis to produce scalefree measurement of functional ability." Am J Occup Ther 53(1): 8390.
259 Velozo, C. A., Y. Wang, et al. (2008). "Utilizing Rasch measurement models to develop a computer adaptive self report of walking, climbing, and running." Disabil Rehabil 30(6): 458467. Visintin, M. and H. Barbeau (1989). "The effects of body weight support on the locomotor pattern of spastic paretic patients." Can J Neurol Sci 16(3): 315325. Voordecker, P., N. Mavroudakis, et al. (1997). "Audiogenic startle reflex in acute hemiplegia." Neurology 49(2): 470473. Warren, S., M. Moore, et al. (1995). "Traumatic head and spinal cord injuries in Al aska (19911993)." Alaska Med 37(1): 1119. Waters, R. L., R. Adkins, et al. (1994). "Prediction of ambulatory performance based on motor scores derived from standards of the American Spinal Injury Association." Arch Phys Med Rehabil 75(7): 756760. Waters R. L., R. H. Adkins, et al. (1991). "Definition of complete spinal cord injury." Paraplegia 29(9): 573581. Wernig, A. (2006). "Longterm body weight supported treadmill training and subsequent follow up in persons with chronic SCI: effects on functional walking ability and measures of subjective well being." Spinal Cord 44(4): 265266; author reply 267268. Wernig, A. and S. Muller (1992). "Laufband locomotion with body weight support improved walking in persons with severe spinal cord injuries." Paraplegia 30(4): 229238. Wernig, A., S. Muller, et al. (1995). "Laufband therapy based on 'rules of spinal locomotion' is effective in spinal cord injured persons." Eur J Neurosci 7(4): 823829. Wernig, A., A. Nanassy, et al. (1999). "Laufband (treadmill) therapy in incomplete paraplegia and tetraplegia." J Neurotrauma 16(8): 719726. Whelan, P. J. (1996). "Control of locomotion in the decerebrate cat." Prog Neurobiol 49(5): 481515. WHO (2001). International Classification of Functioning, Disability and Health: ICF. Geneva, Switzerland, World Health Organization. WHO (2002). Towards a Common Language for Functioning, Disability and Health ICF. ICF Training Beginner's guide. Geneva, WHO: 22. Wilkins, D. E., M. Hallett, et al. (1986). "Audiogenic startle reflex of man and its relationship to startle syndromes. A review." Brain 109 ( Pt 3): 561573.
260 Wirz, M., R. Muller, et al. (2010). "Falls in persons with spinal cord injury: validity and reliability of the Berg Balance Scale." Neurorehabil Neural Repair 24(1): 7077. Wirz, M., D. H. Zemon, et al. (2005). "Effectiveness of automated locomotor training in patients with chronic incomplete spinal cord injury: a multicenter trial." Arch Phys Med Rehabil 86(4): 672680. Wirz, M., B. Zorner, et al. (2009). "Outcome after incomplete spinal cord injury: central cord versus BrownSequard syndrome." Spinal Cord 48(5): 407414. Wolf, S. L., P. A. Catlin, et al. (1999). "Establishing the reliability and validity of measurements of walking time using the Emory Functional Ambulati on Profile." Phys Ther 79(12): 11221133. Wolf, S. L., D. E. Lecraw, et al. (1989). "Forced use of hemiplegic upper extremities to reverse the effect of learned nonuse among chronic stroke and head injured patients." Exp Neurol 104(2): 125132. Wolpaw, J. R. (2007). "Spinal cord plasticity in acquisition and maintenance of motor skills." Acta Physiol (Oxf) 189(2): 155169. Wright, C. G. and C. D. Barnes (1972). "Audiospinal reflex responses in decerebrate and chloralose anesthetized cats." Brain Res 36(2): 307331. Wu, Y. T., C. W. Ho, et al. (2010). "BrownSequard syndrome caused by type III odontoid fracture: a case report and review of the literature." Spine (Phila Pa 1976) 35(1): E2730. Wulf, G. and R. A. Schmidt (1988). "Variability in practice: facil itation in retention and transfer through schema formation or context effects?" J Mot Behav 20(2): 133149. Wydenkeller, S., M. Liechti, et al. (2006). "Impaired scaling of responses to vestibular stimulation in incomplete SCI." Exp Brain Res 175(1): 1911 95. Wyndaele, M. and J. J. Wyndaele (2006). "Incidence, prevalence and epidemiology of spinal cord injury: what learns a worldwide literature survey?" Spinal Cord 44(9): 523529. Yang, J. F. and M. Gorassini (2006). "Spinal and brain control of human walki ng: implications for retraining of walking." Neuroscientist 12(5): 379389. Ye, T. W., L. S. Jia, et al. (2009). "BrownSequard syndrome due to penetrating injury by an iron fence point." Spinal Cord. Yeomans, J. S. and P. W. Frankland (1995). "The acousti c startle reflex: neurons and connections." Brain Res Brain Res Rev 21(3): 301314.
261 Zehr, E. P. (2002). "Considerations for use of the Hoffmann reflex in exercise studies." Eur J Appl Physiol 86(6): 455468. Zehr, E. P. (2005). "Neural control of rhythmic human movement: the common core hypothesis." Exerc Sport Sci Rev 33(1): 5460.
262 BIOGRAPHICAL SKETCH Martina Rebekka Spiess was born in Schlieren, Switzerland in 1976. She grew up with her parents Robert and Beatrice and her younger sister Annette in Dieti kon, a suburb of Zurich, Switzerland. She completed her K 12 education in 1996 when she graduated from the Kantonsschule Enge in Zrich, Switzerland. Her high school education included one year as an exchange student at the Indiana Area Senior High School in Indiana, PA, USA in 1992/93. After completing high school in Switzerland, Ms. Spiess was accepted into physical therapy school in Leukerbad, Switzerland and graduated in February 2001. Classes at the physical therapy school Leukerbad are held in either French or German, depending on the native tongue of the respective instructor. Martina also completed several internships in the French speaking part of Switzerland, which allowed her to become fluent in this language. Upon achievement of her PT degree, Ms Spiess became very interested in research and at the same time, the University of Zurich, in cooperation with the University of Maastricht, The Netherlands, offered a Master of Physical Therapy Science degree for the first time. Ms. Spiess was accepted i nto the pilot class and graduated in 2005, receiving her MPTSc. The program was a part time academic program during which Ms. Spiess also worked part time as a staff physical therapist in different clinical settings from acute care in both regional and Uni versity hospitals to private clinics. After graduation, she was accepted for a position as a research fellow at Balgrist University Hospital, Zurich. She worked in the Spinal Cord Injury research team and was mainly involved with the European Multicenter S pinal Cord Injury Study. In 2008, she returned to school to pursue a PhD in Rehabilitation Sciences at the University of Florida on an Alumni Fellowship from the University of
263 Florida. Under the mentoring of Drs: Behrman, Patten and Howland, she studied re covery of locomotion after spinal cord injury, specifically in those individuals that possess the ability for basic stepping overground, but experience limitations when having to complete more challenging walking tasks, such as stepping over a puddle or up a curb. Ms. Spiess is interested in the underlying mechanisms of recovery of these functions and in the neurophysiological measurement tools to quantify them. She received her Ph.D. from the University of Florida in the fall of 2012. Ms. Spiess was named outstanding International Student by the College of Public Health and Health Professions in 2010 and received the Alex Courtelis Award for Outstanding International Students from the University of Florida in 2011. After graduation, Ms. Spiess plans to return to her home country Switzerland and pursue a post doctoral fellowship at the University of Zurich. Martina enjoys spending time in nature, especially hiking, climbing, biking or skiing with good friends or family members in her beloved mountains. She loves to share her passion for sports and nature with others, has fun volunteering for different causes and enjoys cooking and baking for her friends and family.