UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Ready to move? The effect of stroke on attention and motor planning of voluntary leg movements. Peters, Susan (Sue) 2017

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2018_february_peters_susan.pdf [ 4.03MB ]
Metadata
JSON: 24-1.0360746.json
JSON-LD: 24-1.0360746-ld.json
RDF/XML (Pretty): 24-1.0360746-rdf.xml
RDF/JSON: 24-1.0360746-rdf.json
Turtle: 24-1.0360746-turtle.txt
N-Triples: 24-1.0360746-rdf-ntriples.txt
Original Record: 24-1.0360746-source.json
Full Text
24-1.0360746-fulltext.txt
Citation
24-1.0360746.ris

Full Text

READY TO MOVE? THE EFFECT OF STROKE ON ATTENTION AND MOTOR PLANNING OF VOLUNTARY LEG MOVEMENTS. by  Susan (Sue) Peters  B.A., University of Western Ontario, 2005 M.P.T., University of Western Ontario, 2007  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Rehabilitation Sciences)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   November 2017  © Susan (Sue) Peters, 2017 ii  Abstract  Background: Many movements individuals perform in a given day are voluntary and goal directed, requiring the ability to focus attention and plan movements according to those goals. Altered cognition and motor impairments after stroke limit functional balance and walking performance. Little research has examined the neurophysiology of attention and planning for standing balance and walking after stroke. Examining the mechanisms underlying attention and planning of balance and walking is paramount to understanding the factors that may be important for stroke recovery and functional community living.  Methods: The overarching objective of this thesis was to examine planning and attention for leg movements in individuals with stroke. There were two primary objectives: (1) to examine whether planning differs between the paretic and non-paretic legs measured at the muscle and brain, and (2) to determine how attention may alter, or gate, the processing of irrelevant somatosensory information used for planning.   Summary of findings: In Chapters 2 and 3, contrary to expectations, no differences in planning were found between stepping with the non-paretic and paretic legs. However, individuals with greater motor impairment showed larger levels of muscle co-contraction during planning (Chapter 2), and greater cognitive effort and longer planning durations (Chapter 3) than individuals with less motor impairment. For attention, irrelevant somatosensory information was gated by attention during planning plantarflexion movements in young adults compared with rest iii  (Chapter 4). In Chapter 5, the main finding was that some irrelevant somatosensory information was not gated by attention after stroke, while other information was gated. This indicates possible dysfunction in pathways connected to the somatosensory cortex after stroke that can be mediated by attention. More importantly, gating levels during early planning explained a significant amount of variability in a measure of community balance and mobility.   Conclusions: This dissertation contributes new knowledge toward understanding the effects of stroke on planning and attention of leg movements. The findings suggest planning and attention are important factors in community levels of balance and mobility that require consideration in future development of targeted neurophysiological assessment and treatment of attention and planning after stroke, with potential impact on balance and walking performance.     iv  Lay Summary  A stroke can change the way a person thinks and moves, which can limit balance and walking. Little research has tested how brain and muscle circuits for attention and planning for balance and walking are affected after a stroke. This research examined if there were differences between the legs in the planning of movements after stroke, and if stoke affected how the brain handles distracting sensory information. No differences were found in planning between the legs; yet, people with greater movement difficulties after stroke showed larger muscle activation and greater planning effort. Some distracting information was not eliminated after stroke, like it was in older adults. This may show that parts of the brain that manage distracting information do not work well after stroke. This thesis shows that attention and planning are important factors to consider for the performance of balance and walking after stroke.   v  Preface The work in this dissertation was conceived, designed, conducted, analyzed, and written by Susan (Sue) Peters. The research described was approved by the University of British Columbia’s Clinical Research Ethics Board certificate #H12-00837 and #H15-03511. Chapters 1 and 6 were written by Susan (Sue) Peters. Drs. Jayne Garland, Lara Boyd, Bimal Lakhani, and Kathryn Hayward assisted in editing these chapters. Chapter 2 is based on work conducted by Susan (Sue) Peters, Drs. Jayne Garland, Kimberly Miller, Tanya Ivanova, Michael Hunt, and Mr. Christopher Cochrane. Susan (Sue) Peters was responsible for study design, data collection, analysis, interpretation, writing, and revising the manuscript. Drs. Miller and Ivanova assisted with data collection, interpretation, and revising the manuscript. Mr. Cochrane assisted with data analysis and with revising the manuscript. Drs. Garland and Hunt assisted in study design, analysis, interpretation, and editing the manuscript. Chapter 3 is based on work conducted by Susan (Sue) Peters, Drs. Tanya Ivanova, Bimal Lakhani, Jayne Garland, Todd Handy, Richard Staines, and Lara Boyd. Susan (Sue) Peters was responsible for study design, data collection, analysis, interpretation, writing, and revising the manuscript. Drs. Ivanova and Lakhani assisted with data collection, analysis, interpretation, and revising the manuscript. Drs. Garland and Boyd assisted with study design, interpretation, and revising the manuscript. Drs. Handy and Staines assisted with interpretation, editing, and revising the manuscript. vi  Chapter 4 is based on work conducted by Susan (Sue) Peters, Katlyn Brown, Bea Francisco, and Drs. Richard Staines, Jayne Garland, Todd Handy, and Lara Boyd. Susan (Sue) Peters was responsible for study design, data collection, analysis, interpretation, writing, and revising the manuscript. Ms. Brown assisted with study design, data collection, interpretation, and revising the manuscript. Ms. Franscisco assisted with data collection, analysis, and revising the manuscript. Drs. Staines, Garland, Handy, and Boyd assisted with study design, interpretation, and revising the manuscript. Chapter 5 is based on work conducted by Susan (Sue) Peters, Katlyn Brown, and Drs. Richard Staines, Jayne Garland, Todd Handy, and Lara Boyd. Susan (Sue) Peters was responsible for study design, data collection, analysis, interpretation, writing, and revising the manuscript. Ms. Brown assisted with study design, data collection, interpretation, and revising the manuscript. Drs. Staines, Garland, Handy, and Boyd assisted with study design, interpretation, and revising the manuscript. In Appendix A, an additional manuscript is presented. This work was developed by Susan (Sue) Peters and Drs. Todd Handy, Bimal Lakhani, Lara Boyd, and Jayne Garland. Susan (Sue) Peters was responsible for generating the ideas, literature review, and manuscript writing. Drs. Handy, Lakhani, Boyd and Garland delivered critical contribution to the ideas and manuscript editing. The manuscript was developed from a literature search over the course of Susan (Sue)’s PhD comprehensive exam. A version of Chapter 2 has been published: Peters S, Garland SJ, Miller KJ, Cochrane CK, Ivanova TD, Hunt MA. Motor Planning for Loading During Gait in Subacute Stroke. Arch Phys Med Rehabil. 2016 Apr;97(4):528-35. doi: 10.1016/j.apmr.2015.11.003. Epub 2015 Nov 26. vii  A version of Appendix A has been published: Peters S, Handy TC, Lakhani B, Boyd LA, Garland SJ. Motor and Visuospatial Attention and Motor Planning After Stroke: Considerations for the Rehabilitation of Standing Balance and Gait. Phys Ther. 2015 Oct;95(10):1423-32. doi: 10.2522/ptj.20140492. Epub 2015 Apr 30. Chapters of this dissertation that have been published may include additional details to increase clarity and continuity across the chapters of this dissertation work. viii  Table of Contents  Abstract .......................................................................................................................................... ii	Lay Summary ............................................................................................................................... iv	Preface ............................................................................................................................................ v	Table of Contents ....................................................................................................................... viii	List of Tables .............................................................................................................................. xvi	List of Figures ............................................................................................................................ xvii	List of Abbreviations ................................................................................................................. xix	Acknowledgements .................................................................................................................... xxi	Dedication ................................................................................................................................. xxiii	Chapter 1: General introduction and background .................................................................... 1	1.1	 Problem of stroke on community levels of balance and walking ...................................... 1	1.1.1	 Epidemiology and economic burden of stroke ........................................................... 1	1.1.2	 Normal motor performance of standing balance and walking .................................... 2	1.1.3	 Motor performance of standing balance and walking with aging and stroke ............. 3	1.1.4	 Changes to standing balance and walking under cognitive demand ........................... 4	1.2	 Planning and attention as unexplored factors that may affect balance and walking ......... 5	1.3	 Planning ............................................................................................................................. 7	1.3.1	 Early and late stages of planning ................................................................................ 8	1.3.2	 Neural substrates of planning ...................................................................................... 9	1.3.2.1	 Brain regions involved in planning ...................................................................... 9	ix  1.3.2.2	 Connections between regions and generators of planning (Figure 1.1) ............. 10	1.3.3	 Aging and stroke effects on planning ....................................................................... 11	1.4	 Attention .......................................................................................................................... 13	1.4.1	 Neural substrates of attention .................................................................................... 14	1.4.1.1	 Gating function of the thalamus ......................................................................... 15	1.4.1.2	 Role of prefrontal cortex .................................................................................... 15	1.4.2	 Impact of relevance and spatial aspects of directed attention ................................... 16	1.4.3	 Aging and stroke effects on attention ....................................................................... 17	1.5	 Modulation of planning by attention ................................................................................ 18	1.6	 Methodological approaches ............................................................................................. 19	1.6.1	 Electromyogram (EMG) ........................................................................................... 19	1.6.2	 Electroencephalogram (EEG) ................................................................................... 21	1.6.2.1	 Somatosensory evoked potentials (SEPs) .......................................................... 22	1.7	 Thesis overview ............................................................................................................... 23	1.7.1	 Specific research aims and hypotheses for each chapter .......................................... 24	1.7.2	 Significance ............................................................................................................... 25	Chapter 2: Planning for loading during walking in sub-acute stroke ................................... 27	2.1	 Introduction and background ........................................................................................... 27	2.2	 Methods ............................................................................................................................ 29	2.2.1	 Participants ................................................................................................................ 29	2.2.2	 Clinical assessment ................................................................................................... 30	2.2.3	 Laboratory assessment .............................................................................................. 30	2.2.4	 Data analysis ............................................................................................................. 31	x  2.2.5	 Statistical analysis ..................................................................................................... 32	2.3	 Results .............................................................................................................................. 33	2.3.1	 Participant demographics .......................................................................................... 33	2.3.2	 Within- and between-leg EMG timing and amplitude .............................................. 34	2.3.3	 Between-leg co-contraction ...................................................................................... 36	2.4	 Discussion ........................................................................................................................ 41	2.4.1	 Limitations ................................................................................................................ 43	2.5	 Conclusions ...................................................................................................................... 43	Chapter 3: Symmetry of cortical planning for initiating stepping in sub-acute stroke ....... 44	3.1	 Introduction and background ........................................................................................... 44	3.2	 Methods ............................................................................................................................ 46	3.2.1	 Participants ................................................................................................................ 46	3.2.2	 Functional assessment ............................................................................................... 47	3.2.3	 Experimental protocol ............................................................................................... 47	3.2.3.1	 Behavioural task ................................................................................................. 47	3.2.3.2	 Movement .......................................................................................................... 48	3.2.3.3	 Electromyography (EMG) ................................................................................. 48	3.2.3.4	 Electroencephalography (EEG) ......................................................................... 49	3.2.4	 Data analysis ............................................................................................................. 50	3.2.4.1	 Goniometer and EMG analysis .......................................................................... 50	3.2.4.2	 EMG analysis ..................................................................................................... 51	3.2.4.3	 Co-contraction index .......................................................................................... 51	3.2.4.4	 EEG analysis ...................................................................................................... 52	xi  3.2.5	 Statistical analysis ..................................................................................................... 53	3.3	 Results .............................................................................................................................. 54	3.3.1	 Participants ................................................................................................................ 54	3.3.2	 No differences between legs ..................................................................................... 56	3.3.3	 Correlations with EEG measures .............................................................................. 58	3.3.4	 EMG .......................................................................................................................... 60	3.3.5	 Sub-analysis of slow versus fast steppers ................................................................. 61	3.4	 Discussion ........................................................................................................................ 62	3.4.1	 Planning for stepping is similar between limbs ........................................................ 62	3.4.2	 Planning a step linked with anticipatory knee flexor muscle activity bilaterally ..... 63	3.4.3	 Motor performance of paretic stepping differentiates planning measures ................ 64	3.4.4	 Limitations ................................................................................................................ 64	3.5	 Conclusions ...................................................................................................................... 66	Chapter 4: Cortical processing of irrelevant somatosensory information from the leg is altered by attention during early planning ............................................................................... 67	4.1	 Introduction and background ........................................................................................... 67	4.2	 Methods ............................................................................................................................ 69	4.2.1	 Participants ................................................................................................................ 69	4.2.2	 Experiment design .................................................................................................... 70	4.2.2.1	 Behavioural task ................................................................................................. 70	4.2.2.2	 Irrelevant somatosensory stimulation ................................................................ 71	4.2.2.3	 Vibration to direct attention toward or away ..................................................... 71	4.2.2.4	 Electromyography (EMG) ................................................................................. 73	xii  4.2.2.5	 Electroencephalography (EEG) ......................................................................... 73	4.2.2.6	 Data collection protocol (Figure 4.2) ................................................................. 74	4.2.2.7	 Functional assessment ........................................................................................ 74	4.2.3	 Data analysis ............................................................................................................. 75	4.2.3.1	 EEG data ............................................................................................................ 75	4.2.3.2	 Behavioural analysis .......................................................................................... 76	4.2.4	 Statistical analysis ..................................................................................................... 76	4.3	 Results .............................................................................................................................. 77	4.3.1	 Electrophysiology ..................................................................................................... 77	4.3.2	 Behavioural and functional measures ....................................................................... 81	4.4	 Discussion ........................................................................................................................ 82	4.4.1	 Behavioural response time is different between conditions ..................................... 84	4.4.2	 SFAQ not related to neurophysiological measures of attention in the leg ............... 84	4.4.3	 Methodological considerations ................................................................................. 85	4.5	 Conclusions ...................................................................................................................... 86	Chapter 5: Attention-mediated suppression of irrelevant somatosensory stimuli during early planning explains significant variability in community balance and mobility after stroke ............................................................................................................................................ 88	5.1	 Introduction and background ........................................................................................... 88	5.2	 Methods ............................................................................................................................ 90	5.2.1	 Participants ................................................................................................................ 90	5.2.2	 Irrelevant somatosensory stimulation ....................................................................... 91	5.2.3	 Behavioural task ........................................................................................................ 92	xiii  5.2.3.1	 Vibration to direct attention toward or away ..................................................... 92	5.2.3.2	 Data collection protocol (Figure 4.2) ................................................................. 93	5.2.4	 Electromyography (EMG) ........................................................................................ 94	5.2.5	 Electroencephalography (EEG) ................................................................................ 94	5.2.6	 Functional assessments ............................................................................................. 95	5.2.7	 Data analysis ............................................................................................................. 96	5.2.8	 Statistical analysis ..................................................................................................... 97	5.2.8.1	 Exploratory stepwise linear regression analysis ................................................ 98	5.3	 Results .............................................................................................................................. 98	5.3.1	 Participants ................................................................................................................ 98	5.3.2	 Electrophysiology ................................................................................................... 100	5.3.3	 Behavioural and functional measures ..................................................................... 105	5.3.4	 Regression results ................................................................................................... 107	5.4	 Discussion ...................................................................................................................... 110	5.4.1	 No attentional modulation of the N40 after stroke ................................................. 110	5.4.2	 Community balance and mobility explained by change scores on the N70 and N40 ……………………………………………………………………………………..111	5.4.3	 No leg differences for individuals with stroke ........................................................ 112	5.4.4	 Limitations .............................................................................................................. 113	5.5	 Conclusions .................................................................................................................... 113	Chapter 6: General discussion ................................................................................................. 114	6.1	 Overview ........................................................................................................................ 114	xiv  6.2	 Planning differs between individuals with high and low levels of motor performance, but not between the paretic and non-paretic legs (Chapters 2 and 3) ........................................... 114	6.3	 Stroke alters attention-mediated gating of irrelevant somatosensory stimuli during early planning in the leg for the N40 (Chapters 4 and 5) ................................................................. 116	6.4	 Between-leg differences in performance but symmetry in planning after stroke (Chapters 2,3,5) ....................................................................................................................................... 118	6.5	 Individuals with stroke and older adults differ from young adults for attention-mediated gating of irrelevant somatosensory stimuli for the P50 (Chapters 4,5) .................................. 119	6.6	 The early planning phase is essential to motor performance (Chapters 4,5) ................. 121	6.7	 Limitations ..................................................................................................................... 122	6.8	 Proposed theoretical framework for brain regions supporting attention and planning of voluntary goal-directed leg movements .................................................................................. 124	6.9	 Implications and future directions ................................................................................. 127	6.9.1	 Extend methodology to develop knowledge of planning and attention of leg movements .......................................................................................................................... 127	6.9.2	 Recovery of leg function through targeted planning and attention training ........... 129	6.9.3	 Develop clinical attention test specific for the leg .................................................. 131	6.10	 Conclusions .................................................................................................................. 131	References .................................................................................................................................. 133	Appendix: Motor and visuospatial attention and motor planning after stroke: Considerations for the rehabilitation of standing balance and gait ..................................... 159	Appendix A ............................................................................................................................. 159	A.1	 Background ............................................................................................................... 159	xv  A.2	 Operational definitions .............................................................................................. 160	A.3	 Clinical Relevance of Attention after Stroke ............................................................ 161	A.4	 Clinical Importance of Motor Planning after Stroke ................................................ 161	A.5	 Objectives .................................................................................................................. 163	A.6	 Visuospatial Attention and its Influence on Motor Planning .................................... 163	A.7	 Visuospatial attention after a stroke .......................................................................... 166	A.8	 Motor Attention and its influence on Motor Planning .............................................. 169	A.9	 Dual tasking as a means to assess motor attention .................................................... 171	A.10	 Motor Planning After Stroke ................................................................................... 173	A.11	 Type of movement cue influences activity of cortical regions ............................... 174	A.12	 Proposed theoretical model for brain regions supporting motor and visuospatial attention, and motor planning of voluntary goal-directed movements ............................... 175	A.13	 Conclusions ............................................................................................................. 177	 xvi  List of Tables Table 2.1: EMG measures. ......................................................................................................... 35	Table 2.2: Characteristics of paretic leg low and high co-contraction groups compared with healthy controls. .......................................................................................................................... 39	Table 3.1: Demographics and lesion location information. ..................................................... 55	Table 3.2: Results of behaviour, EEG, and EMG measures with significant correlations indicated. ...................................................................................................................................... 57	Table 4.1: Electrophysiology and behavioural results. ............................................................ 80	Table 5.1: Demographics and clinical measures for all participants. .................................... 99	Table 5.2: Electrophysiology and behavioural results. .......................................................... 102	Table 5.3: Regression results. .................................................................................................. 108	Table A.1: Summary of anatomical and functional regions associated with motor and visuospatial attention and motor planning ………………………………………………….165 xvii  List of Figures Figure 1.1: Brain regions for planning and attention. ............................................................... 8	Figure 2.1: Co-contraction index pre versus co-contraction post. ......................................... 37	Figure 2.2: EMG activity during one walking cycle for representative participants. .......... 38	Figure 2.3: ANOVA results. ....................................................................................................... 40	Figure 3.1: Schematic of experimental set up (A) and measurement parameters (B). ........ 50	Figure 3.2: Representative participant's EEG, EMG, and goniometer signals. ................... 56	Figure 3.3: MRCP amplitude and duration correlations. ....................................................... 59	Figure 3.4: Correlations between MRCP and BF onset latency. ............................................ 60	Figure 4.1: Stepper and plinth set up. ....................................................................................... 70	Figure 4.2: Data collection and analysis protocol with attention conditions. ........................ 72	Figure 4.3: Representative participant's EEG over Cz during testing of three attention conditions on the right. ............................................................................................................... 78	Figure 4.4: ANOVA and post-hoc results of SEP component amplitudes. ............................ 79	Figure 4.5: Subjective reports of attention and behavioural response time. ......................... 81	Figure 5.1: SEP traces at Cz electrode of representative older healthy control (A) and participant with stroke (B). ...................................................................................................... 101	Figure 5.2: ANOVA results (main effects panel A) and post-hoc results (panel B) of SEP component amplitudes. ............................................................................................................. 104	Figure 5.3: Movement response times (A) and subjective reports of attention (B). ........... 106	Figure 5.4: Partial regression plots of predictors on the x-axis with CB&M on the y-axis. ..................................................................................................................................................... 109	xviii  Figure 6.1: Theoretical framework for how attention impacts planning for voluntary leg movements after a stroke that affects attention-mediated gating. ....................................... 125	Figure A.1: Brain regions for visuospatial and motor attention………….…….………….164 Figure A.2: Proposed theoretical model………………………………………….………….175 xix  List of Abbreviations ANOVA  analysis of variance APA   anticipatory postural adjustment AU   arbitrary units BF   biceps femoris CB&M  community balance and mobility scale CCI   co-contraction index CNS   central nervous system CMSA   chedoke mcmaster stroke assessment EEG   electroencephalography  EMG   electromyography EMGi   integrated electromyography F   female FM   fugl-meyer fMRI   functional magnetic resonance imaging GON   goniometer xx  HSD   honest significant difference M   male M1   primary motor cortex MRCP   movement related cortical potential PFC   prefrontal cortex PMC   premotor cortex PPC   posterior parietal cortex RF   rectus femoris S   seconds S1   primary sensory cortex SD   standard deviation SEP   somatosensory evoked potential SMA   supplementary motor area TMS   transcranial magnetic stimulation µV   microvolts  xxi  Acknowledgements  I would like to thank my doctoral co-supervisors, Drs. Lara Boyd and Jayne Garland. Jayne, I appreciate your honest feedback and encouragement of an exploratory approach to research. Lara, I am grateful for the freedom to explore and pursue questions driven by my inquisitive nature. Both of you have given me such a gift with a supportive and nurturing work environment which has allowed me to follow a line of inquiry that has become important and satisfying to me. You have both inspired me while demonstrating that a fulfilling scientific career can include passionate inquiry and meaningful family and leisure time. I would also like to thank my supervisory committee members Drs. Todd Handy and Richard Staines. To Todd, for your amazing ability to ask deep and thoughtful questions while maintaining humility. To Rich, for facilitating and honing my interest in neurophysiology and electroencephalography.   To the members of the Brain Behaviour and Neural Control of Force Production and Movement Laboratories. You are my lab mates and colleagues without whom this work and experience could not have happened. A special thanks to Kate Brown for being a sounding board and support from the comprehensive exam right through to the dissertation defense. For Drs. Bimal Lakhani and Kate Hayward, for your mentorship and friendship. I have had the privilege of working with many intelligent, supportive, and creative people.   I would also like to thank all of the study participants and funding support from the Canadian Institutes for Health Research, the Four Year Fellowship from UBC, and to the IODE War Memorial Scholarship. This research would not have been possible without their contribution. xxii   Outside of academia, I have been blessed with amazing friends who are a source of encouragement and comic relief. Though there are too many to name here, I would not have come this far without their constant support and friendship. Last but certainly not least, I would like to thank my family. First, for my husband, Carl Peters. Carl, you are my support and rock. I would never have had the courage to start, let alone finish, my dissertation work without you. To my parents, sisters, and family-in-law. You have supported me throughout my graduate work and I will forever be thankful for your unconditional love and encouragement. To my friends and family, I hope that I can be as supportive to all of you as you have been for me.   xxiii  Dedication  For Carl. Words cannot express … 1 Chapter 1: General introduction and background  1.1 Problem of stroke on community levels of balance and walking 1.1.1 Epidemiology and economic burden of stroke  Stroke affects the lives of many Canadians. The Heart and Stroke Foundation of Canada’s 2017 report states that there are 62,000 strokes in Canada every year, with an 80% survival rate [1]. About two-thirds of individuals who sustain a stroke survive and need to cope with differing types and degrees of impairments (defined as deviation or loss in body function or structure [2]) that include motor deficits [3].  The effects of stroke continue for many individuals long after discharge from hospital. In a study of individuals in the chronic phase of recovery after a severe stroke, 3% had no motor impairments, 34% were mildly to moderately impaired, and 63% were severely impaired [4]. More than 400,000 Canadians live with long-term disability post-stroke, with the number of survivors expected to nearly double in the next 20 years [1]. Furthermore, the economic costs associated with stroke are high. Two thirds of stroke survivors return home with family caregivers playing an essential role after a stroke; approximately 8 million family caregivers in Canada provide 25 billion per year in unpaid care [1]. The cost to the economy, the community, and the individual is high. Thus, improvements in motor functions (defined as all body functions, activities and participation [5]), such as standing balance and walking, through rehabilitation of individuals post-stroke have the potential to improve the lives of Canadians living with stroke and their families.    2 1.1.2 Normal motor performance of standing balance and walking  The human body is not rigid. Standing still and upright against gravity is a dynamic process that requires an ongoing series of small adjustments to keep the body’s center of mass over the base of support [6]. To help control the body sway, muscles produce stiffness [7]. Muscles that cross two joints, like the biceps femoris and rectus femoris that traverses the hip and knee, generate indispensable stability in the presence of these fluctuations [8,9]. In particular, low levels of concurrent muscle activity in agonist and antagonist muscles increase stability around the knee [10]. Body segments can be moved independently by muscle activity thus redistributing the center of mass [6]. These adjustments maintain the center of mass within the base of support in a wide variety of environmental conditions, while feedback mechanisms direct the level and direction of corrective muscle activity to maintain standing balance [11].   To start walking, the center of mass needs to move ahead of one foot by leaning forward as the opposite foot swings onward to form a new base of support; to keep walking, the center of mass must move beyond the new base of supports’ edge in a reciprocal manner [6]. For successful walking performance, the change in the body segments must be anticipated and controlled; with inadequate anticipation for controlling the center of mass over the base of support, a fall can occur [6]. Anticipation of the next movement allows for improved speed of performance, and better control of the upcoming movement [6].   Both feedback and feedforward control strategies are involved in standing balance and walking [12]. Anticipatory postural adjustments (APAs) are feedforward muscle activity that occur prior to movement onset, for example, the burst of hamstring muscle activity seen prior to a unilateral arm raise in standing that acts to provide stability in advance of a forward movement  3 that shifts the center of mass backward [13]. Typically in healthy adults, the ipsilateral hamstring muscle shows a burst of muscle activity approximately 200ms prior to the onset of a unilateral arm raise, followed by a contralateral burst in hamstrings [14], to prepare the body to keep the center of mass within the base of support.  1.1.3 Motor performance of standing balance and walking with aging and stroke  Motor performance changes with aging and after stroke. Aging is accompanied with loss in muscle area and fiber number [15], reduced number of motor units, and altered capacity of the neuromuscular junction to adapt to changes in physical activity [16]. These alterations may influence standing balance and walking with aging. After a stroke, different postural control strategies are used, typically resulting in decreased weight-bearing on the paretic leg to maintain standing balance, described as stance asymmetry [17]. Though the paretic leg is often examined, the non-paretic leg is part of the compensation pattern seen in weight bearing asymmetry [18]. Compensatory patterns are behavioural strategies to circumvent impairments that allow standing and walking performance to occur [19]. The effect of compensation can be both positive and negative as it is a mechanism to allow a movement goal to be achieved, however, it can shape post-stroke neuroplasticity in maladaptive or suboptimal ways [19]. With increased uncertainty for walking in the community compared with inside the home, compensation through increased muscle co-contraction may increase stability as an adaptation to allow for safe community walking [20]. Alternatively, spatiotemporal asymmetry in walking is closely related to dynamic standing balance and may be linked to number of falls after a stroke [21]. The non-paretic leg exhibits altered muscle activity with increased EMG activity and compensatory EMG patterns compared with healthy controls [22]. Another postural control mechanism where alterations are seen after stroke is APA. After a stroke, both non-existent and smaller bursts of hamstring  4 muscle are seen in anticipation of an arm raise performed in standing considered to be an altered APA [23]. Likewise, smaller magnitudes of ankle muscle activity are seen for stepping initiation post-stroke compared with healthy controls [24].  1.1.4 Changes to standing balance and walking under cognitive demand    Altered cognition and motor impairments after stroke limits functional balance and walking performance with the consequence of increased fall risk. Increased falls risk after stroke is proposed to be due to decreased attentional capacity and impaired postural responses, and so is linked with performance of balance and mobility [25]. Falls frequently occur during walking due to perturbations to standing balance [26]. Within 6 months of being discharged from hospital, up to 73% of individuals have at least 1 fall [27], and are more than twice as likely to fall compared to healthy adults [28]. The risk of a fall is increased after stroke due to paretic leg weakness, reduced standing balance, cognitive impairments, sensory loss [27], declines in the ability to divide attention to perform a dual task, as well as inappropriate utilization of limited cognitive resources [29].  Little research has examined the neurophysiology of attention for standing and walking with aging or stroke, outside of a dual tasking paradigm. Combined, changes in the brain and muscle can result in the altered execution of performing two things at once. For example, the inability to talk while walking, thought to measure functional dual tasking, is significantly associated with increased fall risk among older adults [30]. When young adults stand on a force platform during a dual cognitive/motor task, they have increased body sway relative to single task performance; that is suggested to signify decreased postural control and prioritization of the cognitive task over the motor task [31]. When older adults perform a similar dual task, body  5 sway decreases with increased ankle muscle activity compared with young adults [32]. It is possible that older adults use a stiffening or co-contraction strategy during the dual task to reduce the cognitive demand for postural control.   Walking requires attention, even in healthy adults, and dual task costs generally increase as walking becomes more difficult, such as in patient populations [29]. Beyond spatiotemporal asymmetries during walking or standing alone, walking around obstacles requires disproportionate attention in individuals with stroke [33]. Secondary tasks performed during walking may have reduced performance [34-36], which indicates the prioritization of postural control for walking is high after stroke. For example, when performing a quick voluntary step, individuals take longer to initiate the step in the paretic leg when asked to perform a cognitive task during stepping [37]. This may mean that cognitive processes are involved in step initiation. Beyond walking and step performance, weight-bearing asymmetry increases when individuals are asked to stand as symmetrically as possible during a cognitive task, suggesting that symmetric weight bearing is attention demanding after stroke [38]. Thus, after a stroke, more attention is required to maintain stance symmetry and step initiation time is delayed with the paretic leg. 1.2 Planning and attention as unexplored factors that may affect balance and walking  Many of the movements that individuals perform in a given day are voluntary and goal directed, requiring the ability to plan movements according to those goals. Motor planning, hereafter termed planning, is operationally defined as the integration of sensory afferent information [39] with known internal representations of limb constants (i.e. limb length, muscle force) [40] together with a movement goal [41], for the purpose of generating an upcoming  6 movement [42]. Likewise, safe living in the community requires the ability to ignore irrelevant stimuli to allow attention to be directed toward a movement goal. Motor attention, hereafter termed attention, is operationally defined as the ability to selectively process somatosensory input pertinent to an individual’s movement goals [43]. Since many individuals experience sensorimotor impairments that disrupt normal performance of balance and walking after a stroke [4], paying attention to the appropriate sensorimotor information to plan movements may be vital to the safety of many community dwelling adults post-stroke. Thus, the importance of examining the mechanisms underlying planning and attention of balance and walking is paramount to understanding factors that may be important for stroke recovery and integration into functional community living [4].   The ability to pay attention and plan movements well is vital to community dwelling adults post-stroke [39,44]. Decreased motor control of balance and walking in older adults who are prone to falling is linked with decreased performance on cognitive tests involving attention [45]. Poor attention affects motor performance and functional recovery [46], and increases the risk of falls [28], lending support to the importance of examining attention post-stroke. The reciprocal action of walking is driven through central pattern generators in the spinal cord [47], yet walking would not be possible without the brain. Evidence suggests cortical involvement in maintaining upright stability while walking [48], step initiation [49], planning [50], and attention [51]. The majority of attention and planning research examines the arm [52-55]. While some parallels can be made between the arm and leg, due to the known differences in anatomy and physiology and the different functional use of the leg, how planning and attention interact to produce leg movements has yet to be determined. The extent to which planning and attentional deficits limit the performance of balance and walking after a stroke is uncertain. Filling this  7 knowledge gap is particularly important for individuals who have sustained a stroke, as there are known performance impairments in the leg that may be due to deficiencies in attention and planning.  1.3 Planning  Deficits in motor performance of the paretic arm are, in part, related to poor planning as seen with increased cognitive effort required to plan a movement [56]. Beyond cognitive effort, more time is needed to plan a movement post-stroke, which is associated with reduced precision and coordination of the arm [52,57]. Importantly, planning time for an arm movement can decrease with rehabilitation [52]. Certain task-specific interventions improve planning of the hand [52] and increase cortical blood flow in planning areas [58], which may indicate active neuroplasticity of planning with rehabilitation.    The function of planning is to: (1) prepare for an upcoming voluntary action, and (2) maintain a state of readiness or preparedness for possible unplanned movements [42]. At rest, motor preparedness is not for integrating a movement goal with the intent of future movement; rather, it functions to sustain readiness for possible movements. Motor preparedness is essentially a perpetual “readiness to move”, while planning is a construct that encompasses both the maintained state of readiness and the willed preparation of motor actions; both states occur without movement for the purposes of this dissertation. The state of motor preparedness and the process of planning likely occur in parallel with performance; however, for this dissertation the assumption that it can be studied in series will be made.  8 1.3.1 Early and late stages of planning  Linking action plans to movement requires parallel processing with distinct brain regions active in early and late planning stages. Brain activity in some regions is linked to the motor goal (e.g., stepping), whereas activity in other areas corresponds with lower-level features of the movement (e.g., muscles, forces, and direction) [59]. Movement planning has been studied since the 1960’s [60], with corresponding brain regions that adjust activity in defined stages, measured with electroencephalogram (EEG) [61]. Early planning involves activity in the supplementary motor area (SMA), premotor cortex (PMC) and subcortical structures such as the basal ganglia and cerebellum thought to involve integrating the motor goal with sensory processing (Figure 1.1) [62,63].   Figure 1.1: Brain regions for planning and attention.   M1 = primary motor cortex, PFC = prefrontal cortex, PMC = premotor cortex, PPC = posterior parietal cortex, SMA = supplementary motor cortex, S1 = primary somatosensory cortex, TH = thalamus.   9 Just prior to movement onset, during late planning, cortical activity in the primary motor cortex (M1) increases [64]. Brain activity in the SMA and PMC consistently precedes M1 activity [65], so that SMA and PMC are active during early stage of planning and M1 is active during the late stage.  1.3.2 Neural substrates of planning  Stepping requires a plan for movement together with the appropriate APA of trunk and leg muscles [66]. Brain regions associated with planning and performance of step initiation are associated with generation and execution of APAs in healthy adults [66]. This brain activity can be linked to APA in standing preparation also for rising onto tiptoes [67]. It has been proposed that a distributed brain network involving the thalamus, basal ganglia, cerebellum, SMA, PMC, M1, and the prefrontal cortex (PFC) is required to plan and execute the postural control required to maintain upright standing balance to initiate stepping [66]. The reticular formation [68], locus coeruleus [69], and the mesencephalic locomotor region [70] in the brainstem are also essential for successful execution of postural control of balance and stepping. Indeed, impairments with initiating stepping, such as that found in Parkinson’s disease and progressive supranuclear palsy, are associated with alterations to brainstem structures [70]. Though the basal ganglia, cerebellum, and brainstem are important for planning and maintaining postural control, the focus of this dissertation will be on the role of the SMA, PMC, and M1 due to the planning functions outlined, and PFC and thalamus because of the twofold planning and attention functions that will be described below. 1.3.2.1 Brain regions involved in planning  Brain activity for planning depends on the type of movement cue given. The SMA plays  10 a greater role in movements produced without external stimuli (or self-initiated) [71-76] and the PMC is more engaged with movements based on external stimuli (externally-cued), such as a “go” cue (Figure 1.1) [74,75,77,78]. Self-initiated hand movement planning produces higher activity together with an earlier response in SMA compared with externally-cued planning [74]. This is opposite to the earlier activity occurring in PMC with externally-cued movements compared with self-initiated [78]. Brain activity in the SMA, sensorimotor cortices, and other brain structures such as the thalamus reflect the decision-making for “when to move” for self-initiated movement preparation that is not present in externally-cued conditions [75-77,79]. Further supporting this finding is that self-initiated movement planning has increased activity in the PFC compared with an externally-cued condition, thought to reflect the need for movement selection [80].   The function of the PFC for planning is for movement goal selection. The PFC controls goal-directed grasping actions within multiple contexts, and exerts a wide influence on parietal and premotor areas to activate the regions appropriate for the intended action during planning and performance [81]. With aging, older adults over-recruit the PFC compared with younger adults with an earlier onset of planning and larger PFC activation, with tonic involvement of PFC regardless of task complexity; this suggests that older adults plan with greater anticipation at a higher cost [82]. Taken together, planning requires a coordinated network of brain activity in multiple regions, thus it is conceivable that leg planning may be negatively influenced by stroke. 1.3.2.2 Connections between regions and generators of planning (Figure 1.1)  Using tracers and intracortical stimulation in non-human primates, the homuncular and parallel structure of the SMA, PMC, and M1 regions [83] suggest each area makes a distinct  11 contribution to movement production at select effectors [84]. Each premotor area (SMA, PMC) projects to M1 as well as the spinal cord through corticospinal tract neurons [85], which means that: (1) each premotor area can potentially influence motor output, and (2) motor output can be independent of M1. Using diffusion imaging, the parietal cortex is connected with M1 and PMC for planning of grasp [86]. The posterior parietal cortex (PPC) encodes potential as well as selected motor plans for reaching suggesting that goal-directed decision-making and planning are integrated [87]. Previously, the SMA, PMC, and subcortical regions were thought to influence generation and control of movement through connections with M1, but it is now recognized that the premotor areas each receive a unique pattern of inputs. The PMC activity is primarily inhibitory during movement preparation and cancels out the M1 excitatory command [88]. To begin movement, PMC activity decreases resulting in reduced inhibition, revealing a network already prepared to activate muscles at or near threshold. In this way movement is preplanned but not executed prematurely [88]. Cortically, it is established that the SMA is the generator of the early planning phase [89] for self-initiated movements, and PMC for externally cued [78]. From intracranial recordings, a subcortical generator for planning is found in the thalamus [90].  Via intracortical electrodes, the generators of the early and late phases of planning in the leg are the bilateral SMA and M1, respectively, and that in the leg with scalp electrodes, differentiation between these two regions are difficult for the foot considering their close proximity [91]. 1.3.3 Aging and stroke effects on planning  After a stroke, deficits in motor performance of the arm are in part related to poor planning [56], and planning can involve altered levels of brain activity or longer planning duration [57,92,93]. These deficits identified for arm movements after stroke may include the loss of ipsilesional activity during planning of paretic arm movements [57]. Non-paretic finger  12 flexion produces activity in the contralateral M1 similar to healthy adults, whereas paretic finger flexion activates bilateral sensorimotor cortices [94], together with increased SMA activity [95]. This suggests that more cortical resources are demanded for planning an equivalent motor task indicating that after a stroke, compensation through altered network activity and altered timing occurs to allow planning to happen.  Increased time and effort level to plan an arm movement [53] is a reflection of higher cognitive demand for planning [53,79]. Taking more time to plan a movement post-stroke is also associated with altered motor performance, such as reduced precision and coordination of the arm after stroke [52,57]. The time to plan an arm movement after stroke can decrease with rehabilitation [52]. Yet, it is undetermined if longer planning duration after stroke is a beneficial or harmful compensatory mechanism. Long planning durations may be advantageous for individuals after stroke that may need more time to generate force and coordinate limbs, but it may also be a negative compensation if there is a need to move quickly to avoid injury. Irrespective of the consequence, compensation through altered network activity likely occurs to allow for planning after a stroke; however, planning stepping has not been examined in individuals after stroke comparing the paretic and non-paretic legs. Most of the evidence for planning after stroke is with the arm. Considering that improving standing balance and walking after stroke is a common rehabilitation focus, and that planning deficits are evident for the arm, it follows that examining the planning of leg movements is a necessary next step. Of equal importance to how planning may influence stepping, is attention.  13 1.4 Attention  Steady state walking is considered an essentially autonomous process that requires very little attention in young adults [96]. However, specific components of walking, such as initiating stepping, appear to require cortical activity [49,97], with increased cortical involvement amid aging or pathology [98]. This implies that walking places a greater demand on the cortex, which may leave less cognitive resources for planning and attention and thus, increases the risk of altered or reduced motor performance potentially leading to a fall. Considering the potential functional consequences of limitations to the performance of walking, little is known regarding the neurophysiological substrates of attention for planning stepping. In the arm, motor performance is modulated by attention [99]. Altering the spatial focus of attention toward the hand, facilitated or increased the size of motor evoked potentials than attention away from the hand, showing that attentional variation matters in young healthy adults [99]. Thus, examination of the attentional processes required to plan certain components of walking, such as step initiation, is necessary to increase understanding of the underlying neurophysiology of walking.   There are many types of attention required for safe community living. Broadly speaking, attention can have an internal (endogenous) or external (exogenous) focus. Endogenous attention refers to focusing on sensory stimuli, such as somatosensation, while exogenous attention is focused outside the body such as to visual stimuli. Selective attention enables stimulus filtering and suppression of distractors, while sustained attention refers to maintained attention over time. Divided attention is the capability of carrying out more than one task simultaneously. Lastly, alternating attention refers to rapid attention shifting from one task to another [29]. Though all of these types of attention may be impacted by a stroke, this work will limit discussion to the attention required to plan a movement, and will be considered to be selective and sustained  14 spatial somatosensory attention. This type of attention is like an internal spotlight that can be focused or intentionally directed by disengaging, moving, and engaging attention to a certain part of the body (i.e. skin over the lateral ankle) [100].   1.4.1 Neural substrates of attention  Attention requires the integration of several brain areas; it can be considered a network with specific anatomy, connectivity and function [100,101]. This network functions to maintain alertness and sensory orientation with damage to any part of the network possibly resulting in deficits in attention [102]. The frontoparietal network generates attention to the spatial features of a movement being planned [103,104]. Spatial attention can be voluntarily or involuntarily shifted to a body part with voluntary shifts (under conscious control) producing greater activity in the dorsal frontoparietal network than the ventral network [105]. The dorsal frontoparietal network is involved with the goal-directed selection of relevant sensory information and for preparing responses to this information, in contrast to the ventral frontoparietal network which identifies relevant stimuli, and functions to interrupt the dorsal system when an important or salient event occurs, such as an obstacle during walking [106]. The dorsal frontoparietal network includes the PFC and the PPC (Figure 1.1). Thus, the function of the PFC is for both movement goal selection to assist in planning (Section 1.3.2.1) as well as for attention processes such as gating, described below (Section 1.4.1.2).   The PPC has a large role in attention [42,107]. Sensory signals converge on the PPC from many different modalities including the primary sensory cortex (S1), where the combined signals allow for altering the sensory gain (up or down) depending on the attentional priorities given to the sensory signals [108], allowing the PPC to encode these variables in the output to planning  15 regions [109]. The parietal cortex functions in attention as evidenced by: (1) neuroimaging studies establishing parietal activity when young adults attend to finger movements they are planning to make [110], and (2) lesions in the left parietal lobe affecting the attention of changing from one movement to another [111,112]. The dominant role for the parietal cortex for disengaging attention can explain why some stroke patients find movement sequencing difficult with even the ipsilesional (non-paretic) hand [111]. In summary, the role of the parietal cortex for attention during planning is for orienting a limb in space, for altering planned movements, and for movement sequencing.  1.4.1.1 Gating function of the thalamus  Relevant somatosensory information can be incorporated into a motor plan while irrelevant information can be disregarded through gating, which is the mechanism for regulating the levels of somatosensory stimuli that reach the cortex [43,51,113]. This can be measured with somatosensory evoked potentials (SEPs) (Section 1.6.2.1). Gating can occur during movement, the late phase of planning, or with attentional tasks in the arm [114-116]. When healthy adults attend to stimuli relevant to an arm plan, the amplitude of certain SEP components are facilitated while others are gated [116]. Hence, attention toward a motor plan appears to focus relevant sensory stimuli for the motor goal while reducing irrelevant information. Considering that the circuitry of the PFC has extensive reciprocal thalamocortical connections [117], the thalamus gates somatosensory information reaching the cortex under PFC control [113] (Figure 1.1).  1.4.1.2 Role of prefrontal cortex  A key to optimal motor performance is the ability to retain information and prevent interference from irrelevant information [118]. Based on task-relevance, the PFC has a role in  16 suppressing irrelevant information by influencing the thalamus; this spares limited cognitive resources from being overwhelmed by too much sensory input with suppression of irrelevant or unattended stimuli [119]. The PFC is not significantly active during automatic performance, but is active when young adults are asked to pay attention to performing an overlearned task [120]. This illustrates that the PFC has a role in the volitional aspect of attention. In patients with unilateral PFC damage, median nerve stimulation at rest produces increased SEP amplitudes over S1 compared with healthy controls, thought to reflect the reduced inhibition of the PFC on sensory processing [121]. Though that particular study did not manipulate attention, there may be a reduction in gating for somatosensory information from the leg as a result of a stroke. If this is the case, it is unclear whether attention toward the leg would also be impacted as a result.  1.4.2 Impact of relevance and spatial aspects of directed attention  The way that attention affects somatosensory information processing can differ based on how relevant the information is to the planned task, as well as the spatial orientation of the locus of attention. Task-irrelevant information is filtered out by the PFC at early cortical processing stages in the hand [54]. In healthy adults, attention toward a task-relevant vibrotactile stimulus applied to the finger leads to enhanced S1 contralateral activity together with suppression of ipsilateral S1, and is associated with activity in the PFC [122]. Normal aging processes are associated with decreased suppression of irrelevant somatosensory information during an attention-demanding task consistent with age-related changes to the PFC’s decline in executive control [123]. For the leg, somatosensory information is facilitated selectively in S1 during leg movement when the information is task-relevant [124]. In addition, PFC may regulate task-related sensory information used during control of standing balance [125]. While the PFC has an important role in sustaining inhibition in early sensory cortical processing and modifying  17 transmission based on task-relevant processes, these processes can be altered with aging and after a stroke in the PFC [119].  1.4.3 Aging and stroke effects on attention  Decreases in attentional resources occur with aging. As individuals age, structural changes arise in cortical PFC areas, a region related to attention and goal-directed planning [29]. Furthermore, white matter lesions, loss of dendritic branching in the PFC, tissue loss in gray matter, and decreased dopaminergic activity in the frontal areas are also associated with decline in cognitive functions, including attention [126]. Over and above structural changes, older adults experience functional changes such as increased activity in the PFC during dual task [127]. There is competition for attentional resources for standing balance during a cognitive task, and it has been suggested that with aging an impairment of the distribution of attentional resources exists [128,129].   Functional motor recovery after stroke is influenced by the attentional system. For example, the ability to attend, measured by clinical attention tests administered two months post-stroke, significantly correlated with arm and leg motor function two years later [46]. Individuals with parietal cortex lesions have difficulty disengaging attention from a planned hand movement to another movement signifying the parietal cortex has a role in attention [112]. After thalamic stroke, difficulty with detecting vibrotactile stimulus in the hand in the presence of a competing contralateral irrelevant stimulus suggests that the ability to control sensory input to the cortex through the thalamus may be altered after stroke [130]. When individuals with neglect post-stroke plan to move their left hand, irrelevant information from the right hand can interfere and cause delays that result in failure to inhibit ipsilesional hand motor plans [131]. Current  18 rehabilitation for neglect primarily focuses on cognitive interventions, however, a recent Cochrane review stated the need for an attentional control condition for future clinical trials [132]. Although attention has been studied post-stroke [133], the link between leg performance and spatially directing attention endogenously within the body has not. There are several gaps within the area of attention and leg movements. Though gating of irrelevant somatosensory stimuli has been examined in healthy adults and individuals after stroke for the arm, whether gating or altered inhibition may occur during planning of leg movements is unexamined. The potential impact on this process by stroke is unmapped; alterations to gating by attention may influence community levels of balance and mobility. 1.5 Modulation of planning by attention  Though attention is vital for functional tasks such as walking, little is known about how stroke affects attention in the early planning phase. Neuroanatomically, planning and attention share common brain regions. The Premotor Theory of Attention provides a starting point for examining attention and planning after stroke, in that (1) there is some overlap between the brain regions involved with planning and attention, (2) planning shifts attention, and that (3) attention is shifted with goal directed planning [134]. In other words, an individual can voluntarily shift attention to a body part for the purpose of planning a movement, and the brain regions involved with both attention and planning may be similar. Indeed, regions important for planning also have attention functions, such as the PFC and thalamus [119,135]. In the arm, meaningful objects that can be grasped, such as a tool or a cup, can draw attention to a location in space during planning, with this type of spatial attention suggested to include the bilateral PMC [55], a region also involved with planning [136]. Performance of reaching and grasping as well as walking in the community requires the integration of sensory information from many systems, including  19 vision. To focus this dissertation, the emphasis within the following chapters will be on how somatosensory information is used in planning and attention. Considering the altered ability for some individuals after stroke to suppress irrelevant somatosensory information from the arm, impaired attention may produce difficulty with selecting the appropriate feedback to compare the current limb position to the goal-directed position being planned. Most attention research has been performed in the hand, and the contribution of attention to leg planning and performance has not been established.  1.6 Methodological approaches 1.6.1 Electromyogram (EMG)  The central nervous system controls the contraction and relaxation of muscle activity [137]. EMG is a recording of the electrical activity of muscles at rest, or during a voluntary or electrically-elicited contraction [138]. It can be reliably recorded over the skin of a muscle or intramuscularly with a wire [139]. An EMG signal’s time-course and morphological properties can be processed to extract information to infer the anatomical structure and physiological function of the muscle being studied [138]. With a voluntary muscle contraction, a burst (defined as EMG signal crossing a predetermined threshold) of EMG activity is seen. The crossing of the threshold represents the muscle activation onset time, with the highest point delineated as the peak amplitude, and the time to peak as the duration between onset and peak; these variables may allow for the evaluation of postural control strategies within pathologies [140]. Another EMG parameter, the M-wave, can be elicited with electrical stimulation and is a signal that depends on stimulation intensity, fatigue, and contraction level of the muscle [137]. EMG is used  20 both clinically for diagnosis and for research purposes to examine activation patterns with the timing and amplitude of muscle activity.   Co-contraction during walking in individuals with neurological impairment has been examined since 1979, and as of 2014, only 8 articles have been published in individuals with stroke examining co-contraction indices in the leg during walking [141]. Co-contraction is defined as the activation of pairs of muscles that contribute to joint stability, such as the biceps femoris (BF) and rectus femoris (RF) in the thigh that support the knee joint during the load acceptance phase of walking. Muscle co-contraction is important for joint stability, movement accuracy, adapting to environmental demands [141], and to improve postural stability during walking after stroke [142]. Although higher ankle co-contraction can increase stability while walking, it is associated with decreased balance and paretic ankle muscle weakness [142]. Reduced balance after stroke has been attributed to sensorimotor impairment, altered postural control and APA, as well as abnormal co-contraction linked to impairments in attention [143]. Increased levels of co-contraction of the paretic leg muscles contribute to altered timing of muscle activation in the stance phase of walking compared with healthy adults [144], and are associated with greater impairment [145,146]. Beyond impairment, higher levels of paretic ankle co-contraction during walking are linked with lower functional balance and ankle muscle weakness [142]. Individuals with stroke show longer duration of thigh muscle co-contraction during walking [20]. While levels of paretic ankle co-contraction can decrease with weight-shifting practice, these changes are not associated with improvements in walking speed [147].   21 1.6.2 Electroencephalogram (EEG)  EEG is a recording of the electrical activity of neurons in the brain through voltage fluctuations. It is used clinically for diagnosis and for research purposes. In research, EEG can be recorded at rest or during evoked potentials, which are time-locked to a stimulus that typically requires a response. Currently, neuroimaging technology does not allow non-invasive recording in healthy humans at the neuronal level. At best, the recording at one surface electrode reflects the summed activity of many neurons. Planning for the arm [148] or leg [149] can be measured via a movement related cortical potential (MRCP) which occurs bilaterally over the SMA or PMC beginning up to 2 seconds prior to movement in healthy adults (defined as early phase of planning for this dissertation). At approximately 500ms before movement onset, the MRCP is strongest laterally over M1 contralateral to the moving body part (defined as late phase of planning for this dissertation) [71,72]. This potential is similar to the contingent negative variation, readiness potential, or bereitschaftspotential; they differ based on experimental protocols among other factors [57,60,150]. However, the overall spatiotemporal profiles are similar with early phase SMA and PMC activity transitioning to late phase M1 [74]. Planning is an important aspect of the MRCP since spontaneous movements, like eye blinks, lack anticipatory SMA activity [151]. The amplitude and duration of the MRCP are thought to provide measures of cognitive effort and duration, respectively, for planning [52].  For examining walking after stroke, EEG measures in an animal model correlated with walking recovery [152]. In humans, EEG is used to identify MRCP in individuals with stroke to explore walking performance. In a chronic stroke proof of principle study, MRCPs detected the intention to walk in individuals with stroke [153], and was used to decrease the time taken to  22 calibrate brain-computer interfaces [154]. Though EEG is increasingly being used to examine MRCPs in individuals with stroke, an explicit comparison of the planning between the paretic and non-paretic legs has not been done.  1.6.2.1 Somatosensory evoked potentials (SEPs)  An SEP is the measurement of the electrical activity of the brain in response to a tactile stimulation in the periphery. It is used to assess the transmission of the afferent volley by examining the timing and amplitude of the alterations to the electrical activity recorded. For this thesis, the level of gating can be measured with SEPs, which produce a measurable alteration of cortical activity at predictable time intervals after a stimulus. Peripheral nerve stimulation alters cortical activity with peaks and valleys in the EEG signal, termed ‘components’, that can be measured from base to peak, or peak to peak (N40, P50, and N70). The N40, P50, and N70 are the nomenclature used to describe the SEP polarity (N = negative, P = positive) and the approximate latency after the stimulus (N40 = 40ms after stimulus) after tibial nerve stimulation. The positive and negative components depend on the referencing method and direction of the dipole in the brain region analyzed [91].   Neurophysiologically, the amplitude of SEPs from the leg are thought to measure the arrival of afferent information to sensory areas; amplitudes can be altered by both thalamic and/or cortical gating mechanisms [155]. With tibial nerve stimulation, area 3b in the foot region of S1 generates the N40 component through thalamo-cortical afferent projections in a modality specific manner [156], and has been confirmed by magnetoencephalogram [157], and with intracerebral and scalp recordings [158]. For the P50 as the primary positivity after tibial nerve stimulation, the generator has been proposed to be S1 [159]. Yamada (1996) proposed that the  23 N40 and P50 generators of the tibial nerve are area 3b and area 1, respectively, based on the known thalamic projections to the sensory cortex [159]. Hari (1996) extended this finding to report that activity in area 5 closely followed area 3b, suggesting that PPC may be involved in components generated after the N40, such as for the N70 [156]; however, the source of the N70 is still a question that requires further anatomical and neurophysiological data to support area 5 as its generator [157]. These components and generators will further be discussed in Chapters 4 and 5 through the examination of gating of irrelevant somatosensory information in young adults, then comparing individuals with stroke to older adults.   1.7 Thesis overview  The overarching objective of this thesis is to examine planning and attention for leg movements in individuals with stroke. There are two primary objectives: (1) to examine whether planning differs between the paretic and non-paretic legs at the levels of the muscle and brain, and (2) to determine how attention can alter the somatosensory information used for planning. Chapters 2 through 5 comprise original research contributions of the dissertation. Chapter 2 investigates EMG of the BF and RF muscle co-contraction, timing, and amplitude in the planning phase of initial contact during walking in individuals with stroke. In Chapter 3, the MRCP amplitude and duration and the EMG in planning step initiation are examined comparing the paretic to the non-paretic legs. Chapters 4 and 5 examine attention-mediated gating of irrelevant somatosensory information during early planning in young adults (Chapter 4), and then compares individuals with stroke to older adults (Chapter 5). Chapter 6 presents a general discussion of the main findings of the thesis and limitations. Finally, Chapter 6 concludes with the development of a framework from which future research can ask meaningful questions.  24 1.7.1 Specific research aims and hypotheses for each chapter  Chapter 2 Aims: The primary purpose of this chapter was to determine the characteristics of planning surrounding initial contact during walking through examination of the timing, amplitude, and co-contraction of thigh muscle activity surrounding the paretic and non-paretic knee in individuals in the sub-acute phase after stroke compared with healthy controls. The secondary aim was to investigate what characteristics of muscle timing, amplitude, and functional balance and mobility differentiated between individuals after stroke with high or low levels of co-contraction in the paretic leg surrounding initial contact during walking.   Chapter 2 Hypotheses: First, EMG during terminal swing would be similar between legs indicating symmetry of planning. Second, individuals after stroke with high levels of co-contraction would be associated with longer durations and higher amplitude EMG as well as low functional balance and mobility compared with lower co-contraction levels.  Chapter 3 Aims: The first aim of this chapter was to determine whether planning a voluntary step differs between the paretic and non-paretic legs at the cortical and muscle levels. The second aim was to examine whether measures of planning related to muscle and clinical measures of balance and mobility.    Chapter 3 Hypotheses: First, planning a step with the paretic versus non-paretic leg would result in larger MRCP amplitude, and longer MRCP duration compared to planning a step with the non-paretic leg. Second, there would be a positive relationship between longer planning time (MRCP duration) and greater cortical effort (MRCP amplitude) with muscle measures (onset, slope, and co-contraction index) as well as with lower scores on clinical measures.  25  Chapter 4 Aim: The aim of this chapter was to examine whether attention gates irrelevant somatosensory stimuli reaching the cortex during the early planning phase of plantarflexion movements in young healthy individuals.   Chapter 4 Hypothesis: SEP amplitudes of irrelevant stimuli would be lower when attention is directed toward planning plantarflexion compared with both (1) attention directed away and (2) during rest.  Chapter 5 Aims: The first aim of this chapter was to determine whether gating of irrelevant somatosensory information was present in individuals with stroke during the early planning phase of a plantarflexion movement. The second aim was to examine whether the level of gating explained variability in a measure of community balance and mobility.   Chapter 5 Hypotheses: Gating would be absent in individuals with stroke, and that levels of gating would explain a significant amount of variability in a measure of community balance and mobility. 1.7.2 Significance  This dissertation work was designed to contribute basic scientific knowledge toward the idea that planning and attention processes may be essential for community levels of balance and walking after stroke. These findings will provide insight into the neurophysiological underpinnings of specific aspects of planning and attention in individuals after stroke and healthy adults. Further these studies will offer an important starting point for future research with neurophysiological measures of attention and planning that are tested in the leg, which might have meaningful applications in clinical settings. The potential clinical importance of work related to this dissertation is further elaborated in Appendix 1, which includes a perspective  26 paper on the theoretical basis for considering planning and attention in the context of rehabilitation of balance and walking after stroke. Additionally, this dissertation highlights the need for more research that examines planning and attention, in addition to and separate from other forms of attention (e.g., visuospatial attention), considering these factors are strong influencers of motor performance in individuals after stroke. It is conceivable that a better understanding of the neurophysiological underpinnings of planning and attention and the application of this knowledge to clinical interventions may assist with motor function outcomes after stroke, with repercussions to the Canadian economy and quality of life of individuals post-stroke.     27 Chapter 2: Planning for loading during walking in sub-acute stroke  2.1  Introduction and background  After a stroke, many individuals experience sensorimotor impairments that disrupt their balance and walking abilities [34-36], and this is a major focus in rehabilitation [4]. Often overlooked in the retraining of walking is the consideration of how planning during terminal swing influences motor performance. Of particular importance for walking retraining after stroke is the ability to maintain or regain motor control of the knee upon initial contact and into the load acceptance period of walking [160]. Without sufficient motor control during load acceptance, a leg may “give way”, placing the individual at risk for a fall [161]. The motor control of the knee during load acceptance is largely due to proper planning for, and response to, loading. This requires precise planning of the magnitude and timing of muscle activity [162].   Planning typically involves co-contraction of the quadriceps and hamstrings during terminal swing, with increased force generation from the quadriceps after initial contact to control the knee from flexing under the added load of body weight [163]. Given that stroke produces muscle weakness and altered muscle activation patterns [164], it is possible that a disruption of planning may influence the motor control, and subsequent stability, of the knee during stance.  Differences in the temporal patterns of muscle activity in individuals following stroke have been reported previously for those in chronic stroke [20,165]. Significantly greater duration of co-contraction of quadriceps and hamstrings muscle groups in the paretic and non-paretic legs  28 in single leg support during walking has been observed compared to healthy controls [20,165]. In a sample of individuals with chronic stroke approximately nine months post-stroke, Den Otter et al. found that the period of co-contraction of the thigh muscles during the stance phase was longer compared with a control group (61% in the paretic, 62% in the non-paretic leg, 25% control) [165], consistent with the findings of a recent review paper [20]. Additionally, this review paper found only three studies examining thigh muscle co-contraction after stroke (with only one high-quality study), all of which reported differences in co-contraction timing during the acute or chronic stages of recovery [20]. The authors suggest that altered control of the knee with a longer duration of co-contraction may be an adaptation strategy during the acute phase of stroke. What is unknown is whether the altered timing found between healthy adults and individuals after stroke is also present between the paretic and non-paretic legs in the subacute phase of recovery as this stage has not been specifically examined, or whether both legs post-stroke have symmetrical alterations in the temporal pattern of activity. Additionally, these studies did not specifically examine planning as reflected by the amplitude of muscle co-contraction of the thigh musculature specific to loading during walking in individuals after stroke.   Aims: The primary purpose of this chapter was to determine the characteristics of planning surrounding initial contact during walking through examination of the timing, amplitude, and co-contraction of thigh muscle activity surrounding the paretic and non-paretic knee in individuals in the sub-acute phase after stroke compared with healthy controls. The secondary aim was to investigate what characteristics of muscle timing, amplitude, and functional balance and mobility differentiated between individuals after stroke with high or low levels of co-contraction in the paretic leg surrounding initial contact during walking.   29  Hypotheses: First, EMG during terminal swing would be similar between legs indicating symmetry of planning. Second, individuals after stroke with high levels of co-contraction would be associated with longer durations and higher amplitude EMG as well as low functional balance and mobility compared with lower co-contraction levels. 2.2 Methods 2.2.1 Participants  Participants with stroke were recruited from two inpatient rehabilitation facilities. These participants were enrolled in a randomized clinical trial (RCT; PI Garland) investigating the effect of balance retraining on functional balance and mobility post-discharge from inpatient rehabilitation; all data presented in this paper were collected at the baseline testing session prior to commencing the intervention [166]. Inclusion criteria for this study comprised: (1) first-ever clinically diagnosed stroke, (2) ability to ambulate independently for ten meters without walking aids or orthotics, (3) hemiparesis in the leg (Chedoke McMaster Stroke Assessment (CMSA) scores ≥ 3 and < 7 in the leg or foot subscales) [3], (4) less than 6 months post-stroke similar to other work in subacute stroke [167,168], and (5) living in the community. Exclusion criteria included: (1) co-existing neurological (other than stroke) or musculoskeletal disorders (e.g. acute ankle pain), and (2) cognitive (score on the Mini-Mental State Examination < 23/30) [169,170] or language impairment that prevented the individual from understanding instructions and giving consent. Healthy participants were recruited from the community. Inclusion criteria were: (1) over 60 years old, (2) free from significant ankle, knee, hip or back pain in the past 3 years. Exclusion criteria were: (1) diagnosis of arthritis in the leg, and (2) any significant neuromuscular impairment as a result of illness or disease (e.g. stroke, Parkinson’s disease).  30 Participants provided informed written consent and the institution’s Clinical Research Ethics Board approved this study [166]. No previous studies have published on magnitudes of co-contraction surrounding initial contact during sub-acute stroke; therefore, a study examining EMG activity for the rectus femoris and biceps femoris muscles during stepping after stroke was used to estimate sample size, based on an average effect size of 0.77 [171]. Based on an alpha of 0.05, power of 0.80, and d = 0.77, a sample size of 26 participants was needed to detect differences in EMG metrics for this study. Sample size was calculated from reference table C.2 out of the Portney & Watkins textbook [172]. 2.2.2 Clinical assessment  Balance and mobility was evaluated with the Community Balance and Mobility scale (CB&M). The CB&M is scored out of 96 (higher scores indicate better function), has strong content and construct validity for community-level walking balance [173], and has been validated for use in community-dwelling persons after stroke [174]. The CMSA is descriptive measure representing the level of movement impairment, with a score of 0 out of 7 being no palpable movement and a score of 7 out of 7 being normal movement; the scale has high intra- and inter- rater, and test-retest reliability as well as high validity [3]. Both the CB&M and the CMSA were assessed for all participants post-stroke, and the CB&M for healthy individuals, by the same trained physical therapist. 2.2.3 Laboratory assessment  Walking was evaluated in a motion analysis laboratory using ten high-speed digital cameras, sampling at 100Hz to record the movements of twenty-two reflective markers affixed to the participant according to a modified Helen Hayes marker configuration [175]. Participants,  31 wearing their own walking shoes, performed five trials of ten meters of over-ground walking at a self-selected preferred speed, without harnesses, walking aids, or orthotics. For those participants who required assistance, a physical therapist walked alongside at arms-length to prevent falls. All participants were able to complete the five trials with rest given as needed, and no falls occurred during testing. The skin over recording sites was prepared with an alcohol pad prior to electrode application. Surface electromyography (EMG) data were recorded bilaterally with a TrignoTM Wireless System, sampling at 2000Hz from the biceps femoris (BF) and rectus femoris (RF) muscles. 2.2.4 Data analysis  The leg used for analysis for the healthy participants was randomly selected, and all analyses were completed on the randomly selected leg only. Initial contact (0% of the walking cycle) was defined as the onset of the lowest point of vertical excursion of the heel marker and one full walking cycle was taken until the next initial contact of the same heel marker (100% of the walking cycle). Walking speed was calculated for the healthy, paretic, and non-paretic legs separately based on the forward excursion of the sacral marker over the course of each leg’s walking cycle [176]. Trials that were performed at walking speeds more than 10% different from the mean speed for the given leg were discarded.  All EMG data were band-pass filtered (20-400 Hz), rectified and subsequently low-pass filtered (25Hz). Walking EMG data were normalized to the maximum EMG recorded during a maximal voluntary isometric contraction (MVIC) [166]. The mean amplitude of EMG activity over a moving 100ms window was calculated to determine maximal muscle activation amplitude (EMGmax) [166]. Muscle onset and offset times during walking trials were determined as the  32 point where the rectified EMG activity was three standard deviations above baseline, before (onset time) and after (offset time) initial contact [24]. Muscle onset and offset times were converted to a percentage of the total walking cycle time (% walking cycle) with negative values indicating that onset occurred before initial contact. Total muscle activity duration was calculated as the time (% walking cycle) from onset to offset. Mean EMG amplitude was calculated 10% of the walking cycle prior to initial contact (EMGpre) and 10% of the walking cycle after initial contact (EMGpost).  Planning was quantified by calculating the co-contraction index (CCI) between BF and RF muscles point-by-point using:  CCI = Σ (lower EMGi/higher EMGi) x (lower EMGi + higher EMGi) based on work by Rudolph et al 2000 [177] and Kean et al 2009 [178]. The mean CCI was calculated for the 10% walking cycle window prior to initial contact (CCIpre) and 10% after initial contact (CCIpost) that were chosen to reflect co-contraction and muscle activity in planning for, and during loading, respectively. A large CCI score indicates a high level of muscle activity for both the BF and RF in each 10% walking cycle window. A low CCI score indicates one muscle produced low activity relative to the other OR that both muscles had low activity. All variables (onset, offset, EMGpre, EMGpost, CCIpre, CCIpost) were measured for the 5 trials and then averaged [177,178]. All analysis was done using Spike 2 software. 2.2.5 Statistical analysis  The Shapiro-Wilk test was used to assess for normality (p < 0.001) [179]. If data were not normally distributed, the variable was log transformed prior to subsequent analysis [179].  33 Differences for each muscle between the healthy, paretic, and non-paretic legs within individuals in the averaged EMG onset and offset time, EMG total duration time, EMGpre, EMGpost, CCIpre and CCIpost were evaluated with one-way analysis of variance (ANOVA) and Tukey’s HSD post-hoc tests. Tukey’s HSD was chosen as it is in the midrange of the power spectrum balancing Type 1 and Type 2 error risk [179]. The associations among EMG parameters (EMG onset and offset time, EMG total duration time, EMGpre, EMGpost,) and CCI (CCIpre and CCIpost) for participants post-stroke were explored using Pearson correlations. Low and high magnitudes of co-contraction for the paretic leg only were defined by being equal to or below the mean healthy participant value (low CCI group) or above the mean healthy value for both the CCIpre and CCIpost (high CCI group). Participants post-stroke with one high CCI (for CCIpre or CCIpost) and one low CCI value (for CCIpre or CCIpost) were excluded from further analysis. After high or low CCI classification, one-way ANOVAs with Tukey’s post-hoc comparisons of CB&M, walking speed, and all EMG measures were performed between legs for the healthy and paretic legs only. Significance levels for all statistical tests were set at p ≤ 0.05. All statistical tests were conducted using the Statistical Package for the Social Sciences (SPSS v. 21). Results are presented as mean ± standard deviations unless otherwise stated. Given the exploratory nature of this study, no adjustments for multiple comparisons were made [180,181]. 2.3  Results 2.3.1 Participant demographics  Twenty-seven individuals who were 68.0 ± 10.9 (mean ± SD) years old and 95.2 ± 38.1 days post-stroke took part in the study. The mean CB&M score of the stroke group was 46.4 ± 18.9, indicating a moderate level of functional balance and mobility. Their average walking  34 speed was 1.09 ± 0.27 m/s during the stance phase of the paretic leg, and 1.08 ± 0.27 m/s for the stance phase of the non-paretic leg. Their median (range) CMSA Leg and Foot scores were 6 (3-7) and 5 (3-7), respectively. Eight healthy controls were tested. These participants were 68.6 ± 7.2 years old, with CB&M scores of 78.8 ± 7.5. Average walking speed for controls was 1.42 ± 0.11 m/s. 2.3.2 Within- and between-leg EMG timing and amplitude  All variables except for EMGpost for the RF and BF on the non-paretic leg were normally distributed; after log transforming these variables, normality was achieved. A main effect was present for RF offset time only (p = 0.028), being significantly earlier in the walking cycle in healthy controls (at 17.8% post initial contact) than the paretic (25.6%) and non-paretic (28.1%) legs (p = 0.03) (Table 2.1).   35 Table 2.1: EMG measures.   Paretic Non-paretic Healthy Control EMG Measure BF RF BF RF BF RF Onset (% walking cycle)  -14.4±5.4 -5.6±4.3 -12.5±4.9 -5.2±3.1 -16.3±2.3 -6.2±2.8 Offset (% walking cycle) 17.1±8.0 25.6±9.1 14.9±7.1 28.1±9.9 11.1±7.2 17.8±7.1* Total Duration (% walking cycle) 31.4±9.4 31.1±8.8 27.4±9.7 33.2±9.9 27.4±7.9 24.0±8.3 EMGpre (% EMGmax) 25.4±15.4 7.3±4.7 24.4±17.6 6.6±6.1 29.6±21.1 5.0±2.3 EMGpost (% EMGmax) 22.8±18.7 14.8±6.9 19.8±18.6 13.5±12.1 18.0±21.1 10.2±6.5 CCIpre (a.u.) 0.09±0.07 0.08±0.06 0.06±0.03 CCIpost (a.u.) 0.17±0.10 0.14±0.11 0.09±0.04 Values are mean ± SD. % walking cycle indicates percentage of the total walking cycle time. AU, arbitrary units; CCIpost, co-contraction index in the 10% window post-initial contact; CCIpre, co-contraction index in the 10% window pre-initial contact; EMG, electromyography; EMGmax, electromyography maximum amplitude; EMGpost, electromyography post-initial contact; EMGpre, electromyography pre-initial contact.* p = 0.03 for difference between healthy control and paretic/non-paretic limbs.   36 No other EMG parameter of timing or amplitude was different between the legs. For BF in participants post-stroke, the onset time was significantly associated with the mean EMGpre muscle activity in both legs (paretic r = -0.37, p = 0.05; non-paretic r = -0.39, p = 0.05), with earlier onsets being related to greater mean muscle activity. For RF, there was a similar relationship between onset time and EMGpre activity in the non-paretic leg (r = -0.39, p = 0.04) and in the paretic leg (r = -0.60, p = 0.001). There was no significant association between EMG offset times and mean EMGpost activity in BF or RF on either leg (r < 0.36). 2.3.3 Between-leg co-contraction  The mean CCI values for the healthy control group were 0.06 ± 0.03 for CCIpre and 0.09 ± 0.05 for CCIpost. Co-contraction magnitudes surrounding initial contact were similar between the healthy, paretic, and non-paretic legs (Table 2.1). Of the twenty-seven participants post-stroke, fifteen participants post-stroke had CCIpre and CCIpost values both above the healthy mean and were classified as having a high CCI (Figure 2.1, Figure 2.2B) and five participants had CCI values both below the mean of the healthy group and were classified as having a low CCI (Figure 2.1, Figure 2.2C,D).  37  Figure 2.1: Co-contraction index pre versus co-contraction post. Co-contraction index (CCI) of biceps femoris (BF) and rectus femoris (RF) muscles for 10% walking cycle prior to initial contact (CCIpre) versus co-contraction index 10% after the initial contact (CCIpost) in the non-paretic (white circles) and paretic legs.  For the paretic leg, the CCI was stratified based on the healthy mean CCIpre (vertical dashed line) and the healthy mean CCIpost (horizontal dashed line). The low CCI group (black circles) had co-contraction levels equal to or smaller than the two medians. The high CCI group (black triangles) had co-contraction levels larger than the two means. CCIs that are outside the selected criteria (seven cases marked with x) were excluded from the group analysis.   38  Figure 2.2: EMG activity during one walking cycle for representative participants.  EMG activity before and during one walking cycle for representative participants: one healthy control participant (A), a participant with levels of co-contraction above the mean of the sample, or high CCI (B), two participants with low levels of co-contraction, or lowCCI, (C and D). Biceps femoris (BF) (dark gray, pointing down) and rectus femoris (RF) (light gray, pointing up) activity is shown. Time of the initial contact is marked with solid vertical line (time = 0) and two vertical dashed lines denote 10% walking cycle prior to and after initial contact in each panel. EMG activity was normalized to the maximal voluntary isometric contraction (MVIC) of each muscle prior to the walking test. Inserts in each panel show the co-contraction index before (CCIpre) and after (CCIpost) the initial contact.   39 This excluded seven participants post-stroke with one high CCI value and one low CCI value from further analysis (Figure 2.1).   ANOVAs were run to compare CB&M, walking speed, and EMG measures between the three groups: healthy (n = 8), low CCI (n = 5), and high CCI (n = 15). There was a significant main effect of group examining differences between low CCI, high CCI, and healthy at the p ≤ 0.05 level (Figure 2.3), such that the low CCI group had a pattern of RF activity that more closely resembled the healthy EMG activity (Table 2.2).   Table 2.2: Characteristics of paretic leg low and high co-contraction groups compared with healthy controls.  Values are mean ± SD; Abbreviations: EMG, electromyography; CB&M, Community Balance and Mobility scale; CCI, co-contraction index. * p ≤ 0.05. Post-hoc results: a = high CCI different from healthy control; b = low CCI different from healthy control; c = high CCI different from low CCI.  Participants with stroke Control   Low CCI High CCI    (n = 5) (n = 15) (n = 8) p value Paretic Walking Speed (m/s) 1.19±0.21 1.02±0.25 a 1.42±0.11 0.001*      CB&M (out of 96) 53.6±14.6 b 37.5±13.5 a,c 78.8±7.5 <0.001*      EMG Measure     Biceps Femoris     Onset (% walking cycle)  -11.3±1.8 b -13.8±3.5 -16.3±2.3 0.025* Offset (% walking cycle) 13.3±7.1 20.0±5.7 a 11.1±7.1 0.009* Total Duration (% walking cycle) 24.6±8.5 33.8±6.3 c  27.4±7.9 0.031* EMGpre (% EMGmax) 15.7±14.0 30.9±15.7 29.6±21.1 0.238 EMGpost (% EMGmax) 12.0±15.2 31.9±18.8 18.0±21.1 0.087      Rectus Femoris     Onset (% walking cycle)  -5.4±3.0 -7.0±4.4 -6.2±2.8 0.700 Offset (% walking cycle) 21.0±3.4 27.0±8.6 a 17.8±7.1 0.029* Total Duration (% walking cycle) 26.4±5.2 34.0±7.7 a 24.0±8.3 0.013* EMGpre (% EMGmax) 2.8±1.3 10.0±4.6 a,c 5.0±2.3 0.001* EMGpost (% EMGmax) 5.9±1.8 18.9±5.9 a,c 10.2±6.5 <0.001*      CCIpre (a.u.) 0.04±0.02 0.14±0.06 a,c 0.06±0.03 0.001* CCIpost (a.u.) 0.05±0.01 0.24±0.06 a,c 0.09±0.05 <0.001* CCItotal (a.u.) 0.04±0.01 0.19±0.05 a,c 0.08±0.04 <0.001*  40    Figure 2.3: ANOVA results. EMG parameters of healthy control participants (black bars), participants after stroke in the low co-contraction group (white bars, lowCCI) and the high co-contraction group (gray bars, highCCI) for biceps femoris (BF) and rectus femoris (RF) muscles from the paretic and healthy control legs. In the top row, the onset of the EMG burst (A) prior to initial contact and total duration of EMG burst (B) are presented. Mean EMG amplitude for 10% walking cycle prior to initial contact (EMGpre, panel C) and after initial contact (EMGpost, panel D) is shown in bottom row. Post-hoc statistically significant differences between the lowCCI and healthy control groups are denoted with an ‘a’, while differences between healthy control and the lowCCI group are denoted with a ‘b’, and finally, the differences between the lowCCI and highCCI groups are denoted with a ‘c’. Significance was set at p ≤ 0.05. Values are mean ± SD.    Scores of the low CCI group for the paretic RF and BF offsets and total durations, as well as RF EMGpre/EMGpost, were similar to the healthy control group and were different than the high CCI group. Additionally, the CCIpre, CCIpost and CCItotal for the paretic leg in the low CCI  41 group were different than the high CCI group. Most interestingly, functional measures such as walking speed and the CB&M for low CCI group were also closer to the pattern of activity of the healthy participants.   Representative examples of the EMG patterns in three participants post-stroke and one healthy participant are depicted in Figure 2.2. The participant in Figure 2.2B has a high CCI. A low CCI could reflect low EMG activity in both the RF and BF as in the participant in Figure 2.2C or reciprocally-activated muscles suggesting potentially more coordinated movement as in the participant in Figure 2.2D.  2.4 Discussion  This study is unique in exploring the characteristics of planning in control of the knee surrounding the load acceptance period of walking in subacute stroke. With the exception of the duration of paretic BF activation, the amplitude, timing and co-contraction indices of thigh muscle activity were characterized by symmetry and were found to be comparable between the paretic and non-paretic legs when compared with healthy muscle activity. Additionally, for those participants post-stroke with high levels of co-contraction surrounding the knee, both altered timing and amplitude characterized muscle function, whereas participants post-stroke with low levels of co-contraction more closely approximated the pattern of muscle activity found in healthy controls, and presented with higher levels of functional balance and mobility.   Significant relationships between muscle onset time and amount of activity were present in all muscles bilaterally. For those participants post-stroke with late onset timing, a low amount of muscle activity was also observed suggesting that a trade off for late timing with increased EMG activity was not occurring. Despite known weaknesses present in the paretic leg [182], the  42 non-paretic leg appears to produce a similar pattern of muscle activity to that in the paretic leg during loading, suggesting that the motor control of walking after subacute stroke prioritizes maintenance of thigh muscle symmetry between legs [22]. Though asymmetries in this population have been reported for specific variables such as step length and swing time [183], we found that indices of co-contraction were similar within each leg (CCIpre and CCIpost) and between legs (paretic, non-paretic, and healthy), lending support to symmetry as an important motor control strategy.   Participants with stroke in the subacute stage who showed with low levels of co-contraction had better functional balance and mobility, and more closely approximated the levels of balance and mobility in healthy participants. The CB&M evaluates performance of braided sideways walking, single leg scooting and hopping with both legs; tasks that demand a high level of leg coordination and reciprocal muscle activation. High levels of functional balance and mobility, together with lower EMG amplitude surrounding initial contact during walking, are consistent with the finding that higher levels of motor control is accompanied by low co-contraction and overall muscle activity [184]. Low co-contraction surrounding initial contact may indicate more reliance on passive stabilization of the knee through intact ligaments/cartilage than an active control system. It is also possible compensation may be occurring in ways not measured by this study, or that there are other anatomical constraints to aid with motor control of the knee. Alternatively, individuals with low co-contraction levels may maintain knee stability through both altered timing and amplitude similar to those with higher levels of co-contraction, but with differences in the relative contribution of each muscle so that both the quadriceps and hamstrings are not contracting at the same level and at the same time, allowing for a more coordinated, adaptable and reciprocally patterned movement.  43  In those individuals with high CCI, characteristics included high levels of muscle activity in the paretic leg, and altered timing of the paretic quadriceps compared with healthy controls and the stroke group with low CCI values. It is possible that individuals with high levels of co-contraction about the knee do so to compensate for a lack of motor control in an effort to establish a sufficient amount of knee stability in the presence of neuromuscular weakness. Alternatively, high CCI could be maladaptive as high levels of co-contraction at the ankle were associated with increased falls risk in older adults [185].  2.4.1 Limitations  This study included participants with a range of stroke impairments, and did not stratify based on degree of impairment; however, it is representative of the range seen within typical outpatient rehabilitation departments and thus, may increase generalizability of the results. Additional longitudinal studies are needed to address potential changes to co-contraction over time such as after outpatient physical therapy. 2.5 Conclusions  During subacute stroke, planning for the loading response period of walking is characterized by symmetry between legs in both muscle timing and amplitude. High levels of co-contraction surrounding the knee joint was associated with lower levels of function whereas low co-contraction levels were more strongly related to higher functional balance and mobility. These findings suggest a compensatory strategy of increased co-contraction in those with more impairment while maintaining symmetry of muscle activity between legs.  44 Chapter 3: Symmetry of cortical planning for initiating stepping in sub-acute stroke  3.1 Introduction and background  After stroke, many people exhibit altered movement patterns making normal performance of balance and walking difficult [4]. However, the influence of planning on performance of balance and walking post-stroke is not well understood. Planning is defined as the integration of sensory afferent information [39] such as limb position and muscle force [40], with a functional goal [41], to generate a movement [42]. Planning links with motor performance in healthy individuals [186,187]. Deficits in arm performance after stroke are influenced by poor planning [56]; however, there is limited research connecting planning with motor performance in the leg. It is possible that stroke rehabilitation could provide more effective treatments if more was known about how the brain plans functional movements such as standing balance and walking after stroke.   During sub-acute stroke, planning during the loading phase of the walking cycle is characterized by symmetry between limbs in both knee muscle timing and amplitude,  (Results - Chapter 2) [188]. For those individuals with stroke who show high levels of co-contraction of the thigh muscles, both altered timing and amplitude of electromyography (EMG) characterized patterns of muscle activity during walking; these individuals also presented with poorer functional balance and mobility [188]. Planning hand movements occurs in the cortex [78] as well as at the level of the muscle, yet it is not known whether cortical activity is important for planning step initiation after stroke.  45  To examine planning processes in the cortex, the movement related cortical potential (MRCP) amplitude and duration are commonly measured using electroencephalography (EEG). In young adults, the MRCP begins as a deflection in the EEG signal approximately 2 seconds before voluntary movement onset [72], can be observed for planning voluntary movements in the foot [61], and for rising up on toes in standing [67]. The spatiotemporal profile of the MRCP for functional leg movements such as stepping after a stroke is unknown. Current knowledge is limited to performance and recovery of the arm or hand. In the acute phase of recovery after a stroke, the MRCP cannot be reliably recorded due to difficulty with self-initiated movements of the paretic limb [189]. In individuals with sub-acute stroke, differences between planning of paretic and non-paretic hand movements has been observed [190]. These differences are sustained into the chronic phase post-stroke, and shown via longer duration and larger amplitude MRCP’s for planning movements of the paretic hand compared with the non-paretic hand [93]. Overall, non-paretic hand flexion produces activity in the contralateral primary sensorimotor cortex similar to healthy adults, whereas paretic hand flexion activates the bilateral sensorimotor cortex [94], as well as increased supplementary motor area activity [95]. This suggests that more cortical resources are demanded for planning paretic arm tasks. Altered planning after stroke is associated with increased time and level of effort level to plan an arm movement [53]; this is a reflection of higher cognitive demand for planning [53,79].   Aims: The first aim of this chapter was to determine whether planning a voluntary step differs between the paretic and non-paretic legs at the cortical and muscle levels. The second aim was to examine whether measures of planning related to muscle and clinical measures of balance and mobility.    46  Hypotheses: First, planning a step with the paretic versus non-paretic leg would result in larger MRCP amplitude, and longer MRCP duration compared to planning a step with the non-paretic leg. Second, there would be a positive relationship between longer planning time (MRCP duration) and greater cortical effort (MRCP amplitude) with muscle measures (onset, slope, and co-contraction index) as well as with lower scores on clinical measures. 3.2 Methods  3.2.1 Participants  Participants provided informed written consent, and ethics for this study was approved by the Clinical Research Ethics Board of the University of British Columbia as part of a randomized clinical trial [166]. Participants were recruited from two local hospitals, Holy Family Hospital and Lions Gate Hospital, after discharge from in-patient care. Individuals were included if informed consent was provided, stroke occurred < 6 months ago (however, one participant was 7 months post stroke at the time of testing). Stroke was confirmed by computed tomography or magnetic resonance imaging scan at hospital admission. Participants also had leg and foot motor deficits defined as scoring between 3 to 6 of a possible 7 on the Chedoke-McMaster Stroke Assessment (CMSA), minimum Berg Balance Score (BBS) of 30 out of 56, and living in the community. Exclusion criteria were severe co-morbidity preventing safe participation, co-existing disease affecting postural responses such as peripheral neuropathies or vestibular pathology, bilateral stroke, any musculoskeletal condition that would interfere with testing, and cognitive (determined by the Mini-Mental Status Examination score < 23/30 [169,170]) or language impairment to provide informed consent. Descriptive data such as location of stroke, sex, age, and time since stroke were also collected. No previous studies have published on  47 amplitudes of MRCP for planning stepping in individuals after stroke; therefore, data from a previous study measuring MRCP amplitude for an arm task in post-stroke individuals was used to estimate sample size [53]. Based on an alpha of 0.05, power of 0.80, and d = 1.005, a sample size of 13 participants was needed. Sample size was calculated from reference table C.2 out of the Portney & Watkins textbook [172].  3.2.2 Functional assessment  A registered physiotherapist assessed the following two clinical measures. The leg and foot components of the Chedoke-McMaster Stroke Assessment (CMSA) were assessed. The CMSA provides a functional measure of stage of recovery and impairment level with a score of 0 being no palpable movement to a score of 7 being normal movement. It also has high intra- and inter-rater reliability [3]. The Community Balance and Mobility (CB&M) scale is a valid and sensitive measure for assessing change in functional standing and walking balance and mobility in ambulatory individuals post-stroke with mild to moderate impairments [174]. The CB&M reflects a range of walking balance necessary for the broad challenges that may be anticipated at the community level of walking [173]. It is a functional performance measure with 13 tasks (with a maximum score of 96): 0 denotes inability to perform the task, and 5 effective completion of the task.  3.2.3 Experimental protocol 3.2.3.1 Behavioural task  Participants performed self-initiated stepping whereby the foot was placed onto a 10cm high step. A researcher stood within arms length to ensure safety of the participant as needed.  48 Four sets of 20 steps (6 to 15 seconds between each step) were completed totaling 60 to 80 steps per leg, similar to the number of repetitions in other planning studies in stroke using the arm [53]. The stepping leg (paretic or non-paretic) alternated between sets resulting in 2 conditions: 1) paretic stepping, or 2) non-paretic stepping. All participants were given rest as needed, in addition to rest after every 20 repetitions. Participants were given the instruction to step when they felt ready. No other instructions were given regarding timing. If steps were too close in time to allow for EEG analysis and measurement of the MRCP, participants were asked to wait longer between steps. A fixation point was affixed to the wall in front of the participants at eye level.  3.2.3.2 Movement  Movement onset was identified with two electro-goniometers (2D goniometer, model 508 with Noraxon DTS equipped with Analog Input module, Noraxon USA Inc., Scottsdale, AZ, USA) affixed to the lateral knees to measure angular displacement of the stepping leg in the sagittal plane (X axis of the goniometer). The analog signal from the goniometer on the stepping leg was used to generate a trigger (5V, 100µs, Power 1401, CED, Cambridge, UK) each time an angle threshold was reached during knee flexion. The trigger was recorded simultaneously on electroencephalography (EEG) and electromyography (EMG) systems [53]. The goniometer signals, pre-amplified (500x) and band-pass filtered 10-500 Hz, were subsequently sampled with CED Power 1401 data acquisition interface with Spike 2 software (version 6.17, CED, Cambridge, UK) at 1000 Hz. 3.2.3.3 Electromyography (EMG)  Skin over recording sites was prepared with an alcohol pad prior to electrode application. EMG was recorded with two bipolar EMG electrodes (Pre-Gelled, Disposable, 10mm diameter,  49 20mm spacing; MVAP Medical Supplies Inc.) placed over the muscle belly of biceps femoris (BF) and rectus femoris (RF) muscles. EMG signals, pre-amplified (500x) and band-pass filtered 10-500 Hz, were collected using Noraxon Telemyo DTS equipped with Analog Input module. Subsequently, the EMG analog signals were sampled with CED Power 1401 data acquisition interface with Spike 2 software at 2000 Hz and stored for offline analysis together with the goniometer and trigger signals. EMG was collected to examine the pattern of muscle activity during planning (feedforward muscle activity) and performance [164].  3.2.3.4 Electroencephalography (EEG)  For the study of planning and movement preparation EEG provides high temporal resolution within milliseconds, which allows movement preparation to be examined separately from movement execution [191]. Electroencephalography (EEG) was recorded with a direct current full-band EEG system (NEURO PRAX EEG, NeuroConn, Ilmenau, Germany). A 64-channel cap with Ag-AgCl EEG electrodes was placed using the International 10-20 System. The ground electrode was located on the skull, 1 cm posterior and 1 cm lateral to the Cz electrode. All 64 electrodes were prepared and recorded. Electrodes were referenced to bilateral mastoids and placed in the midline of nasion-to-inion and preauricular-to-preauricular lines. Electrode impedance was kept below 5 kΩ at each scalp location. EEG data were continuously collected and digitized at 2000 Hz before being stored on a computer for off-line analysis.   50 3.2.4 Data analysis 3.2.4.1 Goniometer and EMG analysis  For each trial, a window for 4s prior and 4s post trigger time was used for analysis, with the time from movement onset to peak knee flexion defined as the stepping duration (seconds) (Figure 3.1). EMG signals were inspected offline and trials with large movement artifacts were excluded (not more that 3-5 trials in 4 out of the 10 participants).  Figure 3.1: Schematic of experimental set up (A) and measurement parameters (B).  Panel A schematic of the experimental set up with the goniometers on the lateral knee (black is stance leg, dotted is stepping leg), with EMG placed on the BF and RF muscles and the EEG cap on the head. Panel B indicates measurement parameters for the MRCP (amplitude = difference between zero and offset peak, duration = onset – offset time), as well as EMG parameters of onset, peak and slope. The goniometer trace (GON) indicates movement onset (0) to the end of the step (100%). BF = biceps femoris, EEG = electroencephalogram, EMG = electromyogram, GON = goniometer, MRCP = movement related cortical potential, RF = rectus femoris.   51 3.2.4.2 EMG analysis  The EMG signals (band-pass filtered 10-500 Hz) were rectified and trials were aligned to the movement onset identified on the goniometer trace of the stepping leg (Figure 3.1). Goniometer traces from the stepping leg and all EMGs were averaged and the following measures were taken. First, the end of the step was identified as the lowest point on the goniometer trace after the maximal flexion of the knee when the foot was resting on the step, and average step duration was calculated (Figure 3.1). Second, the root mean square of the baseline EMG was measured for 0.5 s at the beginning of the average trace. Burst onset, peak amplitude, and average slope were measured (Figure 3.1). EMG burst onset was determined as the point in time where the rectified EMG activity crossed 2 SDs above the baseline. EMG burst onset was converted to a percentage of the total stepping duration time, with negative values indicating EMG onset occurred before stepping was initiated. The EMG slope was calculated as the change of EMG amplitude from the burst onset and normalized to the peak amplitude. If no burst was found for a muscle, this was treated as missing data in analysis. Thus, EMG parameters entered into analysis were the burst onset and average slope for the BF and RF muscles. 3.2.4.3 Co-contraction index  For each trial, RF and BF EMG of the stepping and stance legs were rectified, normalized to the maximum EMG in each trial [192] and low-pass filtered at 25 Hz to produce a linear envelope. The co-contraction index (CCI) between BF and RF muscles was quantified point-by-point using:  CCI = Σ (lower EMGi/higher EMGi) x (lower EMGi + higher EMGi)  52 based on work by Rudolph et al 2000 [177], Kean et al 2009 [178], and Peters et al 2016 [188]. The mean CCI was calculated for the time window from onset of movement to peak knee flexion (CCIpost) (Figure 3.1). The same absolute time before the onset of movement was chosen to reflect co-contraction and muscle activity in planning for stepping (CCIpre). A large CCI score indicates a high level of muscle activity for both the BF and RF in each time window. A low CCI score indicates one muscle produced low activity relative to the other or that both muscles had low activity. 3.2.4.4 EEG analysis  EEG analysis was performed using the EEGLAB toolbox (v. 13) for Matlab (v. 2014b, MathWorks, Natick, MA) [193]. First, EEG data were visually inspected to detect and interpolate bad EEG channels, after which all 3-4 runs of data collection for each limb were concatenated for further analysis. The EEG data were down sampled to 500 Hz, filtered at 0.25-100Hz, and cleaned from line noise (60Hz plus harmonics removed with the CleanLine plugin for EEGLAB). Independent components analysis (ICA) using the Infomax ICA algorithm was run with the results used to eliminate eye blinks, movement, and environmental artifact from the data [66]. ICA is commonly used for decomposing signals into their underlying sources or parts, and maximizes signal independence producing components. Each component from an ICA has a spatial map with an associated time course [194]. A benefit of ICA is that it can be used to distinguish between movement artifact, noise, and non-task related signal components, as well as identify transient task-related activations [195]. ICA conserves data that may otherwise be removed by other data cleaning methods [196], and has been used examining walking initiation in healthy individuals [149], lateral step initiation [66] and stroke population for this purpose [197]. ICA-pruned data were visually inspected with any noisy time periods rejected manually  53 [66]. Continuous EEG activity was then epoched into: (a) 4000 ms prior to, and (b) 4000 ms after the onset of movement, with a 500ms baseline period to ensure the epochs were long enough to support collection of longer duration MRCP and slower task performance within a stroke population [52]. The Cz electrode was chosen as the electrode for data analysis as the MRCP can reflect M1 and SMA processes [61,91]. Epochs were averaged separately for the paretic and non-paretic stepping conditions. On the average cleaned trace, the mean ± 2SD was calculated for the baseline period (500ms) and the point the EEG signal deflected from the baseline mean ± 2SD was identified as the ‘onset’ of the MRCP. The peak negativity prior to the step was also identified as the ‘offset’ and was used to calculate amplitude (Figure 3.1). The difference between onset and offset was defined as the MRCP duration (seconds) and reflects the time taken for planning; the difference in amplitude between baseline and offset was defined as the MRCP amplitude (µV), and reflects the cognitive effort level associated with the task planned and performed (Figure 3.1) [52,53].  3.2.5 Statistical analysis  Statistical analysis was performed with SPSS for Windows v. 23 (SPSS Inc, Chicago, Illinois). The Shapiro-Wilk test was used to assess for normality with p < 0.001 [179]. One-way within-subjects analysis of variance (ANOVA) was used to compare each of EEG, EMG, and step duration between limbs. Relationships among planning variables (MRCP duration, MRCP amplitude), stepping performance (stepping duration; EMG parameters of burst onset, average slope, CCI), and clinical measures (CMSA, CB&M) were examined using Pearson’s correlation coefficient. A sub-analysis of ‘group’ was performed using the median paretic step duration to divide the sample into two groups to examine if brain and muscle activity differ between those who step fast or slow. Levene’s test was used to examine the assumption of homogeneity of  54 variances prior to (ANOVA). For any variables that failed to meet this assumption, a Welch U test was run [198]. Means and standard deviations are reported unless otherwise indicated. Significance level was set to p ≤ 0.05.  Given the exploratory nature of this study, no adjustments for multiple comparisons were made [180,181]. 3.3 Results 3.3.1 Participants  Thirteen individuals participated in this study; three were excluded post-hoc due to insufficient duration between steps. Ten individuals’ data were analyzed (70.9 ± 7.7 years of age; 6m, 4f, 3.9 ± 1.5 months after stroke) (Table 3.1).   55  Table 3.1: Demographics and lesion location information. Subject ID Age Months since stroke Paretic limb Sex CMSA total (leg + foot) Slow or fast stepper CB&M  (/96) Lesion location          1 84 4 L M 10 Slow 28 R CR 2 73 7 L M 8 Slow 34 R BG 3 77 5 R M 9 Fast 37 R PICA, R lateral medulla 4 68 3 L M 11 Fast 44 R Frontal cortical 5 68 3 R F 10 Fast 35 L CR, BG, lacunar 6 63 3 L F 9 Fast 38 R pons r 77 3 L M 10 Slow 45 R TH, BG, PLIC 8 57 4 R F 9 Slow 40 L BG, CR 9 73 5 L F 12 Slow 21 L posterior communicating artery 10 69 2 R M 12 Fast 39 L CB, medulla, vertebral artery Mean 70.9 3.9   10.0  36.1  SD 7.7 1.5   1.3  7.2           CR = corona radiata, BG = basal ganglia, PICA = posterior inferior cerebellar artery, TH = thalamus, PLIC = posterior limb internal capsule, CB = cerebellum, CMSA = Chedoke McMaster Stroke Assessment, CB&M = Community Balance and Mobility scale.   56 Individuals were mild to moderately impaired with average CMSA scores for the leg and foot 4.9 ± 0.9 and 5.1 ± 0.6, respectively. The CB&M scores were on average 36.1 ± 7.2 / 96. See Table 3.1 for demographic and lesion location information.  3.3.2 No differences between legs  All variables were found to be normally distributed (p ≥ 0.001 [179]). There were no statistically significant differences in performance between the paretic and non-paretic legs in time to peak knee flexion (p = 0.087), or EEG measures (p ≥ 0.069) (Figure 3.2, Table 3.2).   Figure 3.2: Representative participant's EEG, EMG, and goniometer signals.		Vertical line on Panel A and B represents movement onset. A: Paretic stepping condition, B: Non-paretic stepping condition. BF = biceps femoris, GON = goniometer, RF = rectus femoris, MRCP = movement related cortical potential. 57 Table 3.2: Results of behaviour, EEG, and EMG measures with significant correlations indicated.   Movement Condition    Dependent Variables Paretic leg stepping Non-paretic leg stepping p-value Effect Size Movement time to peak knee flexion (s) 0.84 ± 0.23a 1.00 ± 0.41 0.087 0.291 EEG        MRCP Amplitude (µV) -33.27 ± 25.80 -21.32 ± 15.17 0.069 0.322  MRCP Duration (s) 3.41 ± 0.28 3.41 ± 0.35 0.933 0.001 EMG       Stance leg Muscle EMG Measure      RF Slope (au) 0.67 ± 0.21 0.72 ± 0.30 0.444 0.075   Onset (%step) -10.58 ± 24.2 -7.15 ± 24.55 0.534 0.050  BF Slope (au) 1.07 ± 0.71a,b 1.07 ± 0.42 0.998 0.000   Onset (%step) -5.29 ± 26.96 a,b 5.08 ± 14.90 c 0.102 0.269   CCIpre (au) 0.12 ± 0.06 0.13 ± 0.06 0.681 0.020   CCIpost (au) 0.18 ± 0.07 0.20 ± 0.08 0.558 0.039  Stepping leg        RF Slope (au) 1.41 ± 0.61 3.08 ± 1.91 0.080 0.425   Onset (%step) 22.34 ± 15.5 33.96 ± 16.35 0.227 0.232  BF Slope (au) 1.33 ± 0.92 1.44 ± 0.76 0.851 0.005   Onset (%step) -12.94 ± 23.13 a 2.10 ± 20. 38 0.210 0.214   CCIpre (au) 0.11 ± 0.07 0.10 ± 0.04 0.497 0.053   CCIpost (au) 0.11 ± 0.03 0.13 ± 0.06 0.511 0.049  EMG = electromyography, EEG = electroencephalography, CCI = co-contraction index, AU = arbitrary units, RF = rectus femoris, BF = biceps femoris, µV = microvolts. a = correlated with MRCP amplitude during paretic stepping, b = correlated with MRCP duration during paretic stepping, c = correlated with MRCP duration during non-paretic stepping. There were no correlations with MRCP amplitude during non-paretic stepping and movement time or EMG measures during the non-paretic stepping condition. Correlations indicated with a,b,c are significant with p ≤ 0.05.  58 For EMG measures, there were no differences between limbs for onset, slope, CCIpre, or CCIpost for the stepping leg or stance leg for either condition (p ≥ 0.080) (Figure 3.2, Table 3.2). Table 3.2 also includes effect sizes. 3.3.3 Correlations with EEG measures  Behaviour: For the paretic leg stepping condition, movement time to peak knee flexion was related to MRCP amplitude (r = -0.741, p = 0.014) where slower movement times correlated with higher levels of cognitive effort.  EEG: The MRCP amplitude between paretic and non-paretic stepping conditions was correlated (r = 0.717, p = 0.020), with higher amplitudes to plan a paretic step related to higher amplitudes to plan a non-paretic step (Figure 3.3B).    59  Figure 3.3: MRCP amplitude and duration correlations.  A: Correlation between paretic (y-axis) and non-paretic (x-axis) stepping conditions and MRCP durations. B: Correlation between paretic (x) and non-paretic (y) stepping conditions and MRCP amplitudes. C: Correlations between MRCP amplitude (y) and MRCP duration (x) in the non-paretic stepping condition. D: Correlations between MRCP amplitude (y) and MRCP duration (x) in the paretic stepping condition. Blue diamonds indicate participants with stroke who are fast steppers, and orange diamonds indicate participants with stroke who are slow steppers.  For MRCP duration, a relationship was also found between limbs (r = 0.650, p = 0.042) so that longer durations for planning a paretic step were related to longer planning durations for the non-paretic leg (Figure 3.3A). MRCP amplitude during paretic stepping was correlated with MRCP duration during paretic stepping (r = -0.725, p = 0.018) so that individuals with longer planning  60 durations on one leg also took longer to plan a movement with the other leg (Figure 3.3D). This was not significant for the non-paretic limb (r = - 0.475, p = 0.165; Figure 3.3B). 3.3.4 EMG  Paretic stepping condition: The BF muscle onset and slope of the non-paretic stance leg were correlated with the MRCP amplitude and duration during the paretic stepping condition (r ≥ 0.678, p ≤ 0.031; Table 3.2, Figure 3.4).   Figure 3.4: Correlations between MRCP and BF onset latency.  EEG measures on the x axis (µV) and BF latency onset (% step duration) on the y axis. Blue diamonds and line indicate the stepping leg, orange squares and line indicate the stance leg.   61 The onset of BF on the paretic stepping leg was also related to MRCP amplitude during the paretic leg stepping condition (r = -0.702, p = 0.024; Table 3.2, Figure 3.4). No correlations were found for CCIpre or CCI post, or the RF muscle during the paretic stepping condition (r ≤ -0.753, p ≥ 0.051).   Non-paretic stepping condition: The BF muscle onset of the paretic stance leg was correlated with MRCP duration (r = 0.707, p = 0.022; Figure 3.4). No other correlations were found for CCIpre or CCI post, the RF muscle, or for slope for the BF muscle (r ≤ 0.621, p ≥ 0.056). 3.3.5 Sub-analysis of slow versus fast steppers  All variables met the homogeneity of variance assumption except for the onset for RF, CCIpre and CCIpost for the paretic stance leg, and onset for RF for the non-paretic stance leg. For these variables, the result of the Welch’s U test is stated below.  Behaviour: The median time to peak knee flexion with the paretic leg stepping was 0.864 seconds, with five participants above the median classified as ‘slow’ steppers, and the participants below the median as ‘fast’ steppers (Table 3.1, Figure 3.3)   EEG: One-way ANOVAs of the EEG and EMG measures were statistically significant for between group differences in paretic stepping MRCP amplitude (F1,8 = 6.052, p = 0.039) and non-paretic stepping MRCP duration (F1,8 = 6.295, p = 0.036; Figure 3.3), with the slow steppers having a larger MRCP amplitude during the paretic stepping condition (-49.3 ± 27.6 µV versus -17.2 ± 9.5 for fast steppers), and longer MRCP duration for the non-paretic stepping condition (3.63 ± 0.33s versus 3.19 ± 0.20 for fast steppers).   62  EMG: There were no differences between slow and fast steppers for CCIpre, CCIpost or other EMG measures, or for clinical measures. 3.4 Discussion  Contrary to the hypothesis, no differences were found between limbs with respect to cortical or muscle measures of planning for initiating stepping with the paretic or non-paretic legs. In fact, MRCP measures were correlated demonstrating that planning for this particular stepping task may require more symmetrical cortical effort and duration to maintain postural control of standing balance (Table 3.2, Figure 3.3). Differences in the MRCP were found when duration of stepping performance was factored in; individuals who step more slowly with the paretic leg demonstrated greater MRCP amplitudes and durations for planning a paretic and non-paretic leg step, respectively (Figure 3.2). Further, planning of a paretic step was related to knee muscle flexor activity, with the onset timing of the burst of knee flexor activity in the stance (non-paretic) leg and timing in the stepping (paretic) leg (Table 3.2, Figure 3.4). Planning of a non-paretic step also was related to planning the timing of knee flexor muscle activity in the paretic stance leg only (Table 3.2, Figure 3.4). 3.4.1 Planning for stepping is similar between limbs  It is interesting that correlations, not differences, were observed for planning a step between the paretic and non-paretic legs (Figure 3.2, Table 3.2). It is possible that the motor plan required to move the paretic leg to initiate stepping, is similar to that required for maintaining postural control on the paretic leg in stance with a non-paretic leg step. Alternatively, there may be one plan that encompasses both the control of posture and movement. There is evidence that the planning for postural control associated with step initiation  63 may be comparable between limbs and for similar types of motor output. For example, in young adults, MRCP amplitude and duration are similar amongst varying types of lateral stepping [66], as well as between the right and left arm and leg [89]. The authors speculated that this lack of difference suggests comparable motor programs are needed to plan the shifting of the center of mass, irrespective of whether it is to plan functional stepping or a focal leg task [66]. It is conceivable that in the sub-acute phase after stroke, that the motor plan required for stepping reflects this pattern of planning. Fang et al. 2007 hypothesized that longer MRCP durations after stroke could be due to the longer time needed to plan as well as the physiologically increased time required for signal transmission from the central nervous system to the muscle [53]. It is also possible that motor equivalence remains intact for planning stepping in individuals after sub-acute stroke. Motor equivalence is where a task goal can be identified independent of the various ways the goal can be accomplished with various effectors [199]. If voluntary movements have an abstract representation in the brain to allow for flexibility and adaptability with higher order motor plans affording this flexibility [40], the motor plan needed for initiating stepping is likely general in nature. The results in this experiment suggest that some individuals follow this pattern after a stroke.  3.4.2 Planning a step linked with anticipatory knee flexor muscle activity bilaterally  Preparatory knee flexor activity was associated with planning of a paretic step (Table 3.2, Figure 3.4), while the timing of knee flexor muscle activity related to planning of a non-paretic step in the paretic stance leg only (Table 3.2). It is possible that planning after stroke is focused on anticipatory postural activity (APA) of the hamstrings muscle to maintain control of the center of mass. Indeed, individuals with lesions in the supplementary motor area have impaired APA in the arm linking altered planning to reductions in muscle preparatory processes  64 [200]. In stroke, APA of the hamstring muscles in preparation for a unilateral arm raise has been found to be altered in standing [23]; however, amelioration of the APA of the hamstrings can be linked to recovery of standing balance [23]. Thus, the planning through the MRCP may be an important process for the timing of muscle preparation for initiating stepping. 3.4.3 Motor performance of paretic stepping differentiates planning measures  When sub-divided by duration of paretic leg stepping, slow steppers had higher MRCP amplitudes and longer durations (Figure 3.3, Orange diamonds). This may indicate greater cortical effort and that a longer duration is needed to plan stepping movements after stroke. Individuals who step slowly may have more difficulty with planning the task, reflected in more effort and time needed by the cortex. These findings are consistent with other research in the arm where individuals with stroke show correlations between performance with both the paretic or non-paretic arm for circle drawing and EEG planning measures, showing worse performance related to higher effort or duration [53]. This differs from healthy older adults who can scale up postural responses and increase anticipated postural responses with no change to EEG preparatory activity [201]. In young adults, there is also some evidence to suggest that EEG measures of planning can differ depending on knowledge of the upcoming task [202], increases in knee extensor load [203], and rate of force development and intensity [204]. Future studies comparing individuals with unilateral stroke in the SMA to M1 (generators of the early and late phases of the MRCP, respectively) may show differences between legs. 3.4.4 Limitations There are a few limitations to this work. This study recruited individuals in the sub-acute phase of stroke that meet a minimum level of physical ability. Participants were moderately to  65 minimally impaired to ensure that the stepping task could be completed. This restricts generalizability, yet, highlights the need for future work in individuals with severe impairment after stroke in the acute and chronic phases. Fatigue is also prevalent in the stroke population affecting 35 to 92% [205]. This was a concern as this study required many repetitions of the task [89]. To limit fatigue, rest breaks were provided at the participants’ request in addition to after every 20 repetitions. The onset of movement was defined as the point where the goniometer signal deflected from baseline. The participants’ body may have moved prior to knee flexion onset, which could have been measured with force plates [66,73]. As the research question did not require the linking of the timing of onset of MRCP to EMG muscles, it was not necessary for us to add measurements of center of pressure to the protocol. Additionally, previous MRCP studies have already examined the timing of movement of the center of pressure to the onset of MRCP [66,97]. Age-matched controls were not recruited. There are a few reasons why. First, the research question aimed to examine whether there were differences between the paretic and non-paretic legs. Second, initiation of stepping has already been examined in healthy individuals [66]. Attention can also influence MRCP amplitude and latency in individuals with stroke and healthy individuals [206]. This influence was mitigated by having a fixation point, and providing rest breaks as needed. Last, EEG experiments involving the leg provide a special challenge regarding separation of signal sources and ensuing interpretation. In the arm, cortical potentials are measured in the contralateral hemisphere such as in the more laterally oriented electrodes C3 or C4, whereas measurement of MRCPs for the leg occurs in the midline at Cz over the motor and sensory homunculus within the interhemispheric fissure for either leg. This is a limitation that requires addressing by the field to answer further questions probing differences between the legs.  66 3.5 Conclusions  This study demonstrated that planning of a step is similar for a paretic or non-paretic leg and suggests that the postural control elements of step initiation require comparable levels of cortical effort and duration. Additionally, planning measures were associated with measures of muscle activation onset on the both the stance and the stepping leg. The ability to control one’s standing balance to start stepping has implications for the everyday lives of people living with the effects of stroke.  67 Chapter 4: Cortical processing of irrelevant somatosensory information from the leg is altered by attention during early planning  4.1 Introduction and background  The ability to actively suppress irrelevant sensory information is needed for safe and efficient movements in sensory-rich environments. The process as it specifically relates to the facilitation of relevant and the inhibition of irrelevant somatosensory stimuli for a motor plan is broadly termed “attention” [51,207]. Attention uses gating to reduce irrelevant somatosensory information to allow planning to occur effectively [51,124].  The level of gating influences the amount of irrelevant somatosensory information reaching primary motor (M1) and somatosensory (S1) cortices for the arm [208], as well as for the leg [124], which can be measured with somatosensory evoked potentials (SEPs). The amplitude of SEPs may be inhibited (gated) or facilitated, depending on whether the somatosensory stimuli are attended to as a relevant cue to initiate a movement ipsi- or contra-laterally in the arm [114-116,208]. The important finding is that cortical responses are modulated by relevancy of a sensory cue in the leg where sensory stimuli that are more closely attended to will evoke a larger response in S1 than irrelevant sensory information [124]. Hence, attention appears to focus relevant somatosensory stimuli on the motor goal. It remains unknown to what extent irrelevant stimuli are gated in the presence of another potentially competing relevant sensory cue in the leg.  68  Planning, measured as movement related cortical potentials (MRCP), occurs in early and late stages [57], with late planning (~ <500ms to movement onset) being influenced by movement-specific features and not cognitive load [209]. In young healthy individuals, the early stage of planning (~1500ms to 500ms prior to movement) is thought to reflect attentional processes in frontal cortices; the planning of walking initiation demonstrates larger MRCP amplitudes in the early stage of planning than when a simple foot movement is planned [50]. These authors suggest that larger MRCP amplitudes during early planning of walking are attributed to greater cognitive levels of attention required to plan the task. Moreover, for the hand, attending to the timing of movement initiation increases early planning neural activity [210], with differences present as early as 1200ms before movement onset [209]. Taken together, attention to the timing of movement initiation generates larger cortical responses in the early planning stage compared to the late planning stage.  During the late stage of planning, gating of irrelevant SEPs occurs in the arm and leg compared with rest [116,211,212]. For the leg, this gating during late planning occurs when participants plan movements of either the ipsi- or contralateral ankle muscles [211,212]. Additionally, when early planning is examined by dividing it into sub-phases, differences in the levels of gating are found [211,212], but no difference is present between early planning and rest. This suggests that neurophysiological processes may differ even within the phases of planning. Although some work has been done in the early phase of movement planning, whether attentional manipulation during early planning engages the gating mechanism found in late planning is undetermined. It is conceivable that attentional demands during early planning in the leg also produce gating of irrelevant stimuli, subject to whether attention is directed toward the leg. As both aging and stroke can alter gating of irrelevant stimuli in the arm [119,123], it was  69 important to first establish a neurophysiological baseline in the leg prior to further experimentation in older adults and stroke (Chapter 5).   Aim: The aim of this chapter was to examine whether attention gates irrelevant somatosensory stimuli reaching the cortex during the early planning phase of plantarflexion movements in young healthy individuals.   Hypothesis: SEP amplitudes of irrelevant stimuli would be lower when attention is directed toward planning plantarflexion compared with both (1) attention directed away and (2) during rest. 4.2 Methods 4.2.1 Participants  A convenience sample of thirteen healthy individuals participated on 2 separate days of data collection (7 males, 6 females, age 27.8 ± 4.7, range 20-36 years). All participants provided written informed consent, and the University of British Columbia Clinical Research Ethics board approved all experimental procedures and protocols. Individuals were right footed as indicated by a self-reported preference for kicking a ball [213,214]. Individuals were excluded if they had a history of acute or chronic musculoskeletal disorders in the lower extremity and neurological, psychiatric, or psychological diagnoses such as Attention Deficit Disorder or Attention Deficit Hyperactivity Disorder [215]. As no previous studies have published on attention-mediated gating during early planning in the leg, data from a previous study of tibial nerve SEPs in healthy adults recorded during early planning compared with rest was used to calculate an effect size of 1.067 [212]. Assuming a power of 0.80 and alpha of 0.05, and d = 1.067, a sample size of 13  70 participants was needed to detect a difference. Sample size was calculated from reference table C.2 out of the Portney & Watkins textbook [172]. 4.2.2 Experiment design 4.2.2.1 Behavioural task  Participants, lying supine, performed voluntary ankle plantarflexion movements against a custom-made foot pedal mounted on a plinth (Figure 4.1).   Figure 4.1: Stepper and plinth set up.   71 The angle and position of the pedal was adjusted so that the participant pushed on the pedal with the ball of the foot. The pedal compressed an air bladder; a custom computer program read and recorded the pressure changes associated with the task. All participants received a brief practice session to become familiar with the device. A maximum voluntary contraction (MVC) was measured during a 3-second plantarflexion movement against the pedal prior to starting the behavioural task. 4.2.2.2 Irrelevant somatosensory stimulation  SEPs were evoked with electrical stimulation to the tibial nerve in the popliteal fossa (square-wave pulse duration of 0.2 ms and interstimulus interval of 2 Hz) via surface electrodes with the anode in the popliteal fossa and the cathode over the distal quadriceps muscle [212]. Electrical pulses were triggered with Spike2 (version 6.03, CED, Cambridge, UK) software and were delivered using a Digitimer DS7AH stimulator unit (Digitimer, Welwyn Garden City, UK). Stimuli were delivered at 20% above motor threshold [216,217], defined as the point at which tibial nerve stimulation produced a small but consistent visible muscle twitch. Participants were asked to concentrate on detecting the vibration to cue plantarflexion movement and not to pay attention to the continuous electrical stimulation [211,212]. 4.2.2.3 Vibration to direct attention toward or away  To direct attention, vibration (80Hz) was applied in separate blocks ipsilateral to the stimulated leg: (attend toward condition, parameters outlined below) or contralateral to the stimulated leg (attend away condition, parameters outlined below) at random 4-8 s intervals (Figure 4.2A).  72  Figure 4.2: Data collection and analysis protocol with attention conditions. In panel A (Collect - Time), the late planning phase is defined as the reaction time between the onset of the vibration cue and the onset of movement (green star). The green star indicates when the participant starts moving, while the grey rectangle (Collect – Vibration Cue) indicates the on/off of the vibration cue to move. Panel A (SEP analysis) indicates which SEPs were retained and discarded before statistical analysis. In panel B, the green star indicates the moving limb, and the grey rectangle indicates vibration location. R = right, L = left. The lightning bolt indicates location of tibial nerve stimulation.   Vibration was delivered over the medial soleus muscle approximately 3-5cm proximal to the electromyography (EMG) recording electrodes described below. As soon as the participant detected the vibration, they were instructed to plantarflex that ankle as quickly as possible until the vibration stopped, which was set to 20% of the MVC (Figure 4.2A). The same custom computer program controlled the on/off of the vibration device; the ‘on’ signal occurred at random 4-8 s intervals (Figure 4.2A). Each participant completed 35 plantarflexion repetitions  73 of each condition, for a total of 4 blocks of movement (2 per leg), and 2 blocks of rest (1 per leg). Participants were required to sustain attention toward the limb receiving the vibration for the duration of the block (approx. 5 minutes per block). Previous literature suggests that subjective reports of attention toward irrelevant somatosensory stimuli may differ depending on perceived task difficulty [218]. To measure and control for this potential confound, after each block, participants were asked to answer the question, “How well could you focus on the task (i.e. vibration onset) and ignore the stimulus (i.e. tibial nerve stimulation)?” by giving a number out of 10, with 0 indicating not able to focus and 10 having perfect focus throughout the entire block. 4.2.2.4 Electromyography (EMG)  An M-wave was recorded in the medial soleus muscle with 3cm diameter circular surface recording electrodes (Covidien, Mansfield, MA) [219], after the skin over recording sites was prepared with an alcohol pad prior to electrode application. The soleus muscle was selected as it is innervated by the tibial nerve [220].  EMG data were collected using Noraxon Telemyo DTS equipped with Analog Input system. The analog signals were sampled with Power 1401 data acquisition interface with Spike 2 software (version 6.03, CED, Cambridge, UK) and at 5000 Hz, pre-amplified (500x) and band-pass filtered at 10-1500 Hz. EMG was monitored to ensure the M-wave had consistent amplitude throughout the experiment to ensure that a constant stimulus was delivered [211,212].  4.2.2.5 Electroencephalography (EEG)  SEPs were recorded with a direct current full-band EEG system (NEURO PRAX EEG, NeuroConn, Ilmenau, Germany). A 64-channel cap with Ag-AgCl EEG electrodes was placed using the International 10-20 System. Electrodes were referenced to bilateral mastoids and  74 placed in the midline of nasion-to-inion and preauricular-to-preauricular lines. The ground electrode was located on the cap, 1 cm posterior and 1 cm lateral to the Cz electrode. Twenty-nine electrodes were prepared and recorded (Fz, F1, F2, F3, F4, FCz, FC1, FC2, FC3, FC4, Cz, C1, C2, C3, C4, CPz, CP1, CP2, CP3, CP4, Pz, P1, P2, P3, P4, AFz, FPz, FP1, FP2). Electrode impedance was kept below 5 kΩ at each scalp location. EEG data were digitized at 2000 Hz before being stored on a computer for off-line analysis. Participants were asked to keep their eyes closed and their face and body relaxed throughout each experimental condition. Continuous EEG was collected with separate triggers indicating both (1) the timing of tibial nerve stimulation, and (2) the timing of vibration onset/offset.  4.2.2.6 Data collection protocol (Figure 4.2)  One trial of 300 SEP stimulations (approximately 3 minutes) was collected without movement or vibration and defined as the ‘rest’ condition [216,217]. Three conditions were tested in both right and left legs (total of 6 blocks): (1) rest, where no attentional demands were placed on participants, 2) attention toward the stimulated limb (ipsilateral vibration), and 3) attention away from the stimulated limb (contralateral vibration). Participants were randomized to start the protocol on the right or left leg. Next, to optimize time the rest condition was collected followed by randomized attention toward and attention away conditions. 4.2.2.7 Functional assessment  To determine the global level of attention for each participant, a standardized neuropsychological test, the Integrated Visual Auditory Continuous Performance Test (IVA-CPT), was administered [221]. The IVA-CPT is a computerized test where participants click a button on a mouse when they see or hear the number 1 and not click when they see or hear the  75 number 2. The test takes approximately 15-20 minutes and examines sustained attention. From the IVA-CPT battery output, a priori a single variable, the SFAQ (Full-Scale Attention Quotient), was selected as a global measure of the ability to respond to stimuli under low demand conditions such as cued plantarflexion. It combines both accuracy of response (i.e. number of errors) and response time (RT) into a standardized measure that indicates level of sustained attention and this score was used for analysis [222,223].  4.2.3 Data analysis 4.2.3.1 EEG data  SEPs were analyzed off-line with open-source EEGLAB software (version 13) [193] running in the MATLAB (version R2013b) environment. Nerve stimulations that occurred during the vibration and plantarflexion movement were discarded prior to further analysis so that only stimulations that occurred during the early planning period were averaged and measured (Figure 4.2A). Cz was chosen as: (1) it is over the sensorimotor cortex, (2) is commonly used to examine SEPs delivered to the tibial nerve, and (3) has detected gating during planning of ankle movements in previous studies [124,211,212]. The EEG data were filtered at 0.5-120Hz, and cleaned from line noise (60Hz plus harmonics removed with the CleanLine plugin for EEGLAB). EEG data were segmented into epochs: 100 ms prior and 300 ms after tibial nerve stimulation, with the 100ms prior to stimulation used as a baseline. Epochs with significant noise and/or artifact were identified and rejected with the default automated thresholding method in EEGLAB. The N40, P50, N70 components were measured using the peak-to-peak method and used in subsequent statistical analysis. The N40 was measured from the P30 to N40, the P50 was measured from the N40 to P50, and the N70 was measured from the P50 to N70. These  76 components were selected since the amplitude of SEPs from the leg represents the arrival of somatosensory stimuli to sensory areas reflecting both thalamic and cortical gating mechanisms [116,155]. Following tibial nerve stimulation at the ankle, the generators of the N40, P50, and N70 are thought to be contralateral areas 3b and 1 in S1 and area 5 in the posterior parietal cortex, respectively [156]. Subsequent studies suggest similar generators for SEPs after tibial nerve stimulation at the knee [212]. 4.2.3.2 Behavioural analysis  The plantarflexion behavioural task was analyzed for response time (RT, seconds) from the vibratory stimulus onset until time to reach 20% MVC (or vibratory stimulus offset; Figure 4.2A). Mean RT was calculated for each condition per individual. Subjective reports of attention were recorded, and means with standard deviations calculated for each movement condition. 4.2.4 Statistical analysis  All analyses were performed with SPSS software (version 23). The Shapiro-Wilk test was used to assess normality with p < 0.001 [179]. To analyze the assumption of sphericity prior to the repeated measures ANOVA, Mauchly’s test of sphericity was used; if the result of the test was significant indicating sphericity assumption was violated, the Greenhouse-Geisser adjustment was used to correct for this violation. For electrophysiology measures, three separate two-way repeated-measures ANOVAs were performed for each dependent measure (N40, P50, N70 amplitudes) including within subject factors of LIMB (right, left), and CONDITION (rest, attend toward, attend away) as factors. Post-hoc tests were performed for significant differences following the ANOVAs with Bonferroni correction for multiple comparisons. Considering the number of comparisons being performed, the Bonferroni correction was selected to reduce the  77 potential for Type 1 error [179]. For behavioural measures, two separate two-way repeated-measures ANOVA were performed for the dependent measures of: (1) mean response time (seconds) and (2) subjective reports of attention (/10) with LIMB (right, left), and CONDITION (attend toward, attend away) as factors. Associations between electrophysiology (SEP amplitudes), behaviour (RT, subjective attention), and functional measures (SFAQ) were examined using Pearson’s correlation coefficient. Statistical significance was set at p ≤ 0.05. Given the exploratory nature of this study, no adjustments for multiple comparisons were made [180,181]. Results are reported in means ± standard deviation unless otherwise stated. 4.3 Results 4.3.1 Electrophysiology  All variables were normally distributed (p ≥ 0.001). A representative participant’s SEP traces are shown in Figure 4.3.    78  Figure 4.3: Representative participant's EEG over Cz during testing of three attention conditions on the right.  For the N40 amplitude over Cz, no interaction (F(2, 24) = 1.068, p = 0.360) or main effect of LIMB was found (F(1, 12) = 0.022, p = 0.885) but there was a main effect of CONDITION (F(2, 24) = 6.613, p = 0.005). Post hoc tests showed that the N40 amplitude in the attend toward condition was smaller than the rest (p = 0.013) (Figure 4.4A, Table 4.1).    79  Figure 4.4: ANOVA and post-hoc results of SEP component amplitudes. Bars indicate standard error. Stars indicate significant post-hoc results. The stars in Panel A and C indicate the results of the ANOVA with the significant main effect of condition (collapsed between legs).   80  Table 4.1: Electrophysiology and behavioural results.   Electrophysiology Behaviour Leg Condition N40 (µV) P50 (µV) N70 (µV) Movement RT (s) Subjective Attention (/10)        Right Rest 2.45 ± 1.28 2.17 ± 1.71 5.26 ± 3.23 --- ---  Attend Toward 1.62 ± 1.02 1.78 ± 1.25 4.47 ± 2.75 0.618 ± 0.124 7.46 ± 1.66  Attend Away 2.07 ± 1.60 1.98 ± 1.35 4.70 ± 2.77 0.626 ± 0.101 7.54 ± 1.56        Left Rest 2.07 ± 1.31 2.41 ± 1.56 5.25 ± 4.10 --- ---  Attend Toward 1.78 ± 1.35 1.90 ± 0.98 4.22 ± 2.44 0.685 ± 0.115 7.58 ± 1.24   Attend Away 2.15 ± 1.68 2.10 ± 1.36 4.84 ± 3.21 0.615 ± 0.108 7.85 ± 1.07 Values are shown as mean ± standard deviation. µV = microvolts; RT = response time; s = seconds.    81 For the P50 amplitude, no interaction (F(2, 24) = 0.191, p = 0.827) or main effects were significant for either LIMB (F(1, 12) = 0.339, p = 0.57) or CONDITION (F(1, 12) = 2.785, p = 0.08; Figure 4.4B, Table 4.1). For the N70 amplitude, using the Greenhouse-Geisser correction, no interaction (F(1.229, 14.748) = 0.167, p = 0.740), or main effect of LIMB was present (F(1.00, 12.00) = 0.005, p = 0.95); however, a main effect of CONDITION was found (F(1.85, 22.19) = 4.985, p = 0.02; Table 4.1, Figure 4.4C). Post-hoc tests indicated that the N70 amplitude in the attend toward condition was smaller than the rest condition (p = 0.05; Figure 4.4C). 4.3.2 Behavioural and functional measures  The response times for all conditions ranged from 0.615 ± 0.108 s to 0.686 ± 0.115 s (Figure 4.5B, Table 4.1).   Figure 4.5: Subjective reports of attention and behavioural response time.  The star indicates significant post-hoc results.  82  The ANOVA was significant with a LIMB X CONDITION interaction effect (F(1, 12) = 5.596, p = 0.036), revealing a slower RT for left plantarflexion during the attend toward condition on the left (Figure 4.5). Subjective reports of attention ranged from 7.46 ± 1.66 / 10 (attend toward the right) to 7.85 ± 1.07 /10 (attend away on the left) (Figure 4.5A, Table 4.1). The ANOVA did not show an interaction effect (F(1, 11) = 0.647, p = 0.438) and was not significant for LIMB (F(1,11) = 0.379, p = 0.551) or CONDITION  (F(1,11) = 0.478, p = 0.504) for subjective scores. Within limb, the attend toward condition was rated lower (i.e. harder to ignore irrelevant stimulus) than attend away, however, there was no difference between conditions (p = 0.551). The mean score on the SFAQ was 91.21 ± 27.87, indicating an average level of sustained attention function. There were no significant correlations between behaviour (RT and subjective reports of attention), function, and neurophysiological measures. 4.4 Discussion  In this study, attention impacted somatosensory processing of irrelevant information during the early stage of planning plantarflexion movements in young healthy individuals, through gated N40 and N70 amplitudes when attending toward the leg. The N40 and N70 SEP amplitudes from irrelevant somatosensory information are gated during the early stage of planning compared to rest when attending toward the planned leg movement (Figure 4.4). Additionally, this attend toward gating effect was maintained whether or not a right or left leg movement was being planned. In healthy adults, gating of the N40 is known to be present in the late planning stage (~500ms before movement) in the ankle compared with rest [211,212]. These data extend past work by showing that gating of irrelevant somatosensory stimuli occurs during  83 the early stage of planning a leg movement (>500ms before movement) when attending toward the leg.   Two potential neurophysiological factors may explain why gating occurs in the early planning stage with attention toward a leg: (1) the neuroanatomical connections between the prefrontal cortex, the thalamus, and planning regions such as the supplementary motor area (SMA) and premotor cortex (PMC), and (2) attentionally-driven spatial conflict resolution. First, the amplitude of SEPs from the leg, including the N40 and N70, is suggested to measure the arrival of a sensory stimulus to sensory areas reflecting both thalamic and cortical gating [155]. Wasaka et al. 2005 suggest that the SMA is one of the main sources of cortical gating in late planning of self-initiated plantarflexion due to the known activation of SMA during planning and the cortico-cortical anatomical connections of SMA with both the primary motor and sensory cortices [212]. From functional neuroimaging studies, it is known that the PMC is preferentially involved in externally cued planning [74]; it is conceivable that the PMC may function as a source of cortical gating in this externally cued protocol. The prefrontal cortex is also involved in attentional suppression of irrelevant sensory signals through the thalamus [122]. It is possible, therefore, that the prefrontal cortex inhibits the irrelevant somatosensory stimulus through the thalamus mediated by the PMC. Further examination with neuroimaging is needed to map these pathways in the leg.  Second, in the attend away condition there is no spatial conflict to resolve as the sensory stimuli and vibration are applied to different limbs, with SEP amplitudes being similar to rest. Greater inhibition of the competing irrelevant somatosensory stimulus is required to resolve the spatial conflict during the attend toward condition resulting in a smaller SEP compared with rest. This is a novel finding that requires additional study to further investigate the potential  84 generators of these gating effects in the leg and how they may relate to levels of community balance and walking.  4.4.1 Behavioural response time is different between conditions  Participants responded slower to the vibration cue when attending toward the left during left irrelevant stimulation (Figure 4.5B, Table 4.1). It is known that attention toward a task can decrease response times in the detection of signals at that location [224]; considering that subjective reports of attention between conditions were the same, there are a few potential explanations for this finding. First, it is possible that participants found the spatial conflict resolution more difficult in the attend toward condition on the left reflected in a greater duration needed to detect the vibration onset. Second, leg dominance, or footedness, has been shown to predict asymmetries in strength, balance, and walking in young healthy individuals [225]. In this study, response times were slower on the non-dominant left side when attending toward the irrelevant stimulus. Thus, it is conceivable that the factor of ‘leg’ may affect behavioural output, but not neurophysiological processes of attention during the early stage of planning. Third, visuospatial attention shows biases toward the right visual field, which manifests as quicker reaction times when targets are presented on the right, relative to targets on the left [226]. The findings in this chapter may show a similar attentional bias in somatosensory processing with faster reaction times when attending toward the right leg.  4.4.2 SFAQ not related to neurophysiological measures of attention in the leg  No significant correlations were found between the SFAQ, a global attention function measure, and neurophysiological measures of attention during early planning of plantarflexion. This is in contrast to clinical studies that link neuropsychological measures of attention with  85 clinical measures of walking [227], or use computer-based tests completed with the arm to characterize attention after brain injury such as stroke [228]. The neuroanatomy and neurophysiology of the arm and leg are well known to be separate and distinct with functional and anatomical specialization of the brain’s attention systems [229]. These findings underscore the need to develop a tool specifically for evaluating the attention of leg movements that can be administered outside of the laboratory to: (1) characterize normal and pathological capabilities of the attention system to attend to the leg and (2) to evaluate the efficacy of interventions thought to improve lower limb attention. 4.4.3 Methodological considerations  There are strengths to the study’s methodology. First, the level of stimulus used to evoke a SEP remained the same throughout the protocol, and the M-wave amplitude was monitored and remained consistent, so measured alterations to SEP amplitudes were unlikely to be the result of changing stimuli among conditions. The attentional gating effects found in this study are consistent with studies of irrelevant somatosensory information in the upper extremity during late planning [116].   Second, considerable research that examines attention for the arm has used stimuli, such as vision, to cue the experimental motor task [51,124,207]. Neurophysiological responses differ for attended stimuli of touch and vision [230]. Thus, it may be more difficult to quantify the contribution of attention in an experimental paradigm that uses vision to draw attention to the leg as there may be crossmodal interference when multiple modalities are used [231]. For many functional activities of the foot such as walking, visualization of the foot is not required; this is in contrast to the need for vision for many functional activities in the hand such as chopping  86 vegetables. Accordingly, to probe attention for the leg in a functional manner, a vision-free methodology is preferred.   One limitation of this study is that planning and attentional processes share similar cortical output so that it is difficult to disentangle attention from planning processes. The methodology in this study attempted to address this limitation by maintaining consistency of planning demands between conditions so that changes to the levels of SEP amplitudes could be attributed to attentional manipulations of the irrelevant somatosensory information. This limitation could be addressed in future studies with a clinical population with focal lesions to known planning regions, such as the SMA or PMC, or in an aged population with known changes to the prefrontal cortex.  Unlike in the arm where cortical potentials are measured in the contralateral hemisphere to the stimulated limb such as in C3 or C4, measurement of SEPs evoked in the leg occurs in the midline, over Cz [124,211]. These SEP amplitudes reflect the summation of facilitation and inhibition so that what is measured is the result of cortical attentional processes occurring for both the right and left legs simultaneously. It is possible that the location of the leg in the motor and sensory homunculus within the interhemispheric fissure does not allow for differentiation of hemispheric differences within the EEG modality. This is a limitation that the field may need to address to answer questions related to the generators of SEP potentials and frontally-mediated responses to irrelevant stimuli in the leg. 4.5 Conclusions  Attention alters cortical processing of irrelevant somatosensory information during early planning in the leg as indicated by reduced amplitudes of the N40 and N70. If attention alters the  87 level of somatosensory stimulus that reaches the cortex for early planning of a leg movement, then where attention is directed may need to be examined to understand how these processes relate to rehabilitation research of the leg. Further study is required to determine whether these indices are relevant measures for older adults and for individuals with brain injury, such as stroke.  88 Chapter 5: Attention-mediated suppression of irrelevant somatosensory stimuli during early planning explains significant variability in community balance and mobility after stroke  5.1 Introduction and background  After stroke, attention deficits are common and significantly impact motor function [222]. Paying attention and planning movements is vital for the safety of many community dwelling adults post-stroke; this group of individuals is known to have difficulty performing a second task while walking [44], and are at an increased risk of falling [161]. Many movements are voluntary and goal directed, and require planning. Thus, understanding the mechanisms underlying planning and attention is vital to post-stroke outcomes. It is currently unknown how attention and planning interact to influence motor performance of voluntary, goal-directed movements, such as balance and walking, after a stroke.   Though many types of attention are required for safe and functional community living, few have studied motor attention during leg movements. By influencing gating in the thalamus [135], attention selects somatosensory feedback related to the movement goal [51] and disregards irrelevant information [43]. For arm movements, gating of thalamocortical signals has been shown during movement, late planning, and when attention is directed to other tasks [114-116]. Additionally, one animal study showed that thalamocortical gating can occur during walking [232]; this finding has not been replicated in humans. Consequently, attention toward a motor plan appears to gate sensory stimuli that are unrelated to arm movement. Gating  89 mechanisms that control sensory input to the cortex may be important for motor outcomes after stroke [130]; however, it is not known how attention may mediate gating of sensory information from the leg.  After a stroke in the prefrontal cortex (PFC), very little attention-mediated gating of somatosensory stimuli from the hand is observed, which indicates a role for the PFC in inhibition [119]. This is consistent with work showing that individuals with a PFC lesion demonstrate a decline in inhibition in early cortical somatosensory evoked potentials (SEPs) with median nerve stimulation [121], though these SEPs were delivered at rest without manipulating attention. This suggests a reduction in gating of somatosensory stimuli from the arm as a result of a stroke, that attention does little to alter. Based on these findings, it is unclear whether stroke affects the gating of somatosensory information from the leg.   Three peaks of interest are produced by SEPs as a result of tibial nerve stimulation generated by area 3b in the primary somatosensory cortex (S1; N40 component) [156], area 1 in S1 (P50 component) [159], and area 5 in the posterior parietal cortex (PPC; N70 component), though debate over this generator persists [157]. While a stroke may affect the transmission of an afferent volley to the cortex with absent SEPs at rest depending on the lesion location [233], the question is whether attention can alter the level of stimulus reaching the cortex in components that do remain after a stroke. Compared with the arm, the leg shows greater levels of recovery after a stroke [234]; thus, it is unknown whether attention-mediated gating functions differently for leg movements after a stroke.  Planning occurs in early and late phases, with attention-mediated gating occurring in both of these phases in young adults in the leg (Chapter 4 for early phase, [235] for late phase). Most  90 of what is currently known about attention after stroke centers on planning arm movements, yet, attention is paramount to planning safe community balance and walking. Considering the goals of improved standing balance and walking, and reducing falls after stroke, it is essential that the impact of stroke on attention during the planning of leg movements be examined.   Aims: The first aim of this chapter was to determine whether gating of irrelevant somatosensory information was present in individuals with stroke during the early planning phase of a plantarflexion movement. The second aim was to examine whether the level of gating explained variability in a measure of community balance and mobility.   Hypotheses: Gating would be absent in individuals with stroke, and that levels of gating would explain a significant amount of variability in a measure of community balance and mobility. 5.2 Methods 5.2.1 Participants  A convenience sample of healthy older adults and individuals with stroke were recruited from the community and participated on 2 separate days of data collection. All participants provided written informed consent, and the University of British Columbia Clinical Research Ethics board approved all experimental procedures and protocols. To be included, participants demonstrated right footedness determined as a preference for kicking a ball [213,214] as footedness may affect electroencephalogram (EEG) amplitudes [236], and the cognitive capacity to provide consent (Montreal Cognitive Assessment: MoCA <25/30 [237]). For individuals with stroke, additional inclusion criteria included stroke > 6 months ago, and a minimum active range  91 of ankle motion from 10 to 30 degrees of plantarflexion to be able to perform the experimental task. Exclusion criteria for all participants included acute trauma or other reasons where use of the lower limbs are contraindicated, and history of psychiatric or psychological diagnoses such as attention deficit disorder [215]. Additionally, individuals were excluded if the stroke involved the anterior cerebral artery (blood supply to S1 of the leg), and the PFC for the following reasons: (1) with an S1 lesion, depressed or abolished N30 with median nerve stimulation is found for some individuals at rest [238], (2) the PFC regulates suppression of task irrelevant somatosensory stimuli [122]. The Bells Cancellation Test for neglect was used as a screen to exclude individuals with stroke with unilateral neglect [239]. Participants were excluded from the study if the number of bells cancelled on the contralesional side minus the number of bells cancelled on the ipsilesional side was ≥ 3 [239]. Descriptive data such as location of stroke, sex, age, and time since stroke were collected to characterize participants. All participants were screened prior to testing with monofilaments (Baseline® Tactile™ Monofilaments, Fabrication Enterprises Inc., 2008) over the lateral ankle (location of delivery of vibration cue) to ensure sensory thresholds were within normal ranges (e.g. no evidence of peripheral neuropathy). The N40 amplitudes for the rest and attend toward conditions from Chapter 4 was used to estimate sample size, based on an effect size of 0.911. Based on an alpha of 0.05, power of 0.80, and d = 0.911, a sample size of 13 participants for each group was needed to detect differences in gating between conditions. Sample size was calculated from reference table C.2 (sample size calculation table for paired t-tests) out of the Portney & Watkins textbook [172]. 5.2.2 Irrelevant somatosensory stimulation  SEPs were evoked with electrical stimulation of the tibial nerve in the popliteal fossa (square-wave pulse duration of 0.2 ms and interstimulus interval of 2 Hz) via surface electrodes  92 with the anode in the popliteal fossa and the cathode over the distal quadriceps muscle [212]. Electrical pulses were triggered with Spike2 (version 6.03, CED, Cambridge, UK) software and were delivered using a Digitimer DS7AH stimulator unit (Digitimer, Welwyn Garden City, UK). Stimuli were delivered at 20% above motor threshold [216,217], defined as the point at which the tibial nerve stimulation produced a small but consistent visible muscle twitch in the ankle plantarflexors.  5.2.3 Behavioural task  All participants, lying supine, performed voluntary ankle plantarflexion movements against a custom-made foot pedal mounted on a plinth (Figure 4.1). The angle and position of the pedal were adjusted so that the participant pushed on the pedal with the ball of the foot. The pedal compressed an air bladder; a custom computer program read and recorded the pressure changes associated with the task. The same custom computer program controlled the on/off of the vibration described below; the ‘on’ signal occurred at random 4-8 s intervals (Figure 4.2A). All participants received a brief practice session to become familiar with the device. A maximum voluntary contraction (MVC) was measured during a 3-second plantarflexion movement against the pedal prior to starting the behavioural task.  5.2.3.1 Vibration to direct attention toward or away  To direct attention, vibration (80Hz) was applied to the lateral ankle in separate conditions ipsilateral to the stimulated leg (attend toward condition, parameters outlined below), or contralateral to the stimulated leg (attend away condition, parameters outlined below) (Figure 4.2B). As soon as the participant detected the vibration, they were instructed to plantarflex that ankle as quickly as possible until the vibration stopped, which was set to 20% of the MVC  93 (Figure 4.2A). The target level of force was 20% MVC as previous work has demonstrated decreased tibial nerve N40 SEP amplitudes at higher force levels [240]. A researcher observed and recorded performance for each step so that trials with anticipated steps or incorrect responses (i.e. plantarflexion without vibration cue) could be excluded prior to analysis.  5.2.3.2 Data collection protocol (Figure 4.2)  One trial of 300 SEP stimulations (3 minutes) was collected without movement or vibration and defined as the ‘rest’ condition where participants were asked to remember a random 5 digit number given by the experimenter [216,217]. Two planning conditions were tested in both right and left legs for a total of 4 movement conditions: (1) attention toward the stimulated limb (ipsilateral vibration and movement), and (2) attention away from the stimulated limb (contralateral vibration and movement) (Figure 4.2B). Each participant completed 35 plantarflexion repetitions in each of the attend toward and attend away conditions, for a total of 70 plantarflexion movements per leg, and 2 rest conditions (1 per leg). Participants were required to sustain attention toward the limb receiving the vibration for the duration of the condition (5 minutes per movement condition). Participants were randomized to start the protocol on the right or left leg. Next, to optimize time the rest condition was collected followed by randomized attention toward and attention away conditions.  Subjective reports of attention toward irrelevant somatosensory stimuli may differ depending on the task [218]. To measure and control for this potential confound, after each block, participants were asked to give a number out of 10 to describe how well they felt they were able to focus on the task (i.e. vibration onset cue) and ignore the stimulus (i.e. tibial nerve  94 stimulation), with 0 indicating not able to focus and 10 having perfect focus throughout the entire block. 5.2.4 Electromyography (EMG)  An M-wave was recorded in the medial soleus muscle with 3cm diameter circular surface recording electrodes placed 2-cm apart (Covidien, Mansfield, MA) [219], after the skin over recording sites was prepared with an alcohol pad prior to electrode application. The soleus muscle was selected as it is innervated by the tibial nerve [220].  EMG data were collected using Noraxon Telemyo DTS equipped with an Analog Input system. The analog signals were sampled with Power 1401 data acquisition interface with Spike 2 software (version 6.03, CED, Cambridge, UK) at 5000 Hz, pre-amplified (500x) and band-pass filtered at 10-1500 Hz. Soleus EMG was monitored to ensure the M-wave had consistent amplitude throughout the experiment to ensure that a stable stimulus was delivered [211,212].  5.2.5 Electroencephalography (EEG)  SEPs were recorded with a direct current full-band EEG system (NEURO PRAX EEG, NeuroConn, Ilmenau, Germany). A 64-channel cap with Ag-AgCl EEG electrodes was placed using the International 10-20 System in the midline of nasion-to-inion and preauricular-to-preauricular lines. The ground electrode was located on the cap, 1 cm posterior and 1 cm lateral to the Cz electrode. Twenty-nine electrodes (Fz, F1, F2, F3, F4, FCz, FC1, FC2, FC3, FC4, Cz, C1, C2, C3, C4, CPz, CP1, CP2, CP3, CP4, Pz, P1, P2, P3, P4, AFz, FPz, FP1, FP2), ground, and bilateral mastoid electrodes were prepared, recorded, and then referenced to bilateral mastoids offline. Electrode impedance was kept below 5 kΩ at each scalp location. EEG data were digitized at 2000 Hz before being stored on a computer for off-line analysis. Participants  95 were asked to keep their eyes closed and their face and body relaxed throughout each experimental condition. Continuous EEG was collected with separate triggers indicating both: (1) the timing of tibial nerve stimulation, and (2) the timing of vibration onset/offset.  5.2.6 Functional assessments  Integrated Visual Auditory Continuous Performance Test: To determine the global level of attention for each participant on day 2 the Integrated Visual Auditory Continuous Performance Test (IVA-CPT) was administered [221]. The IVA-CPT is a standardized neuropsychological test that examines sustained attention. A priori a single variable, the SFAQ (Full-Scale Attention Quotient), was selected as a global measure of the ability to respond to stimuli under low demand conditions such as cued plantarflexion. It combines both accuracy of response (i.e. number of errors) and response time (RT) into a standardized measure that indicates level of sustained attention [222,223].   Balance and mobility tests: A registered physical therapist assessed the Community Balance and Mobility (CB&M) scale for all individuals. The CB&M is scored out of 96 (higher scores indicate better function), has strong content and construct validity for community-level walking balance [173] for older adults [241], and for community-dwelling persons after stroke [174]. For individuals with stroke, the leg portion of the Fugl-Meyer (FM-LE) was also completed. The FM-LE is valid and reliable for the measurement of leg motor impairment of community-dwelling individuals with stroke [242].   96 5.2.7 Data analysis  EEG data: SEPs were analyzed off-line with open-source EEGLAB software (version 13) [193] running in the MATLAB (version R2013b) environment [193]. Cz was chosen as: (1) it is over the sensorimotor cortex, (2) is commonly used to examine SEPs delivered to the tibial nerve, and (3) has detected gating during planning of ankle movements in previous studies [124,211,212]. Nerve stimulations that occurred during the vibration and plantarflexion movement were discarded prior to further analysis so that only stimulations that occurred during the early planning period were averaged and measured (Figure 4.2A). The EEG data were filtered at 0.5-120Hz, and cleaned from line noise (60Hz plus harmonics removed with the CleanLine plugin for EEGLAB). EEG data were segmented into epochs: 100 ms prior and 300 ms after tibial nerve stimulation, with the 100ms prior to stimulation used as a baseline. Epochs with significant noise and/or artifact were identified and rejected with the default automated thresholding method in EEGLAB. The N40, P50, N70 components were measured using the peak-to-peak method and used in subsequent statistical analysis. The N40 was measured from the P30 to N40, the P50 was measured from the N40 to P50, and the N70 was measured from the P50 to N70.   Behavioural analysis: The plantarflexion behavioural task was analyzed for response time (RT, seconds) from the vibratory stimulus onset until time to reach 20% MVC (or vibratory stimulus offset; Figure 4.2A). Mean RT was calculated for each condition per individual. Subjective reports of attention were recorded, and means with standard deviations were calculated for each movement condition.  97 5.2.8 Statistical analysis  All analyses were performed with SPSS software (version 23). The Shapiro-Wilk test was used to assess normality with p < 0.001 [179]. If data were not normally distributed, the variable was log transformed and then re-tested for normality; if data still did not meet the assumption of normality, data were square root transformed prior to subsequent analysis [179]. Independent t-tests assessed whether SEP amplitudes at rest differed between groups prior to analysis of variance (ANOVA) as both stroke and aging can affect resting SEPs [121,123]. To analyze the assumption of sphericity prior to the repeated measures ANOVA, Mauchly’s test was used; if the result of the test was significant indicating sphericity assumption was violated, the Greenhouse-Geisser adjustment was used to correct for this violation. For electrophysiology measures, three separate 3-way mixed-model ANOVAs were performed for each dependent measure (N40, P50, N70 amplitudes) with GROUP (stroke, older adult) as the between-subject factor; and LIMB (paretic/left, non-paretic/right), and CONDITION (rest, attend toward, attend away) as within-subject factors. Post-hoc tests using the Bonferroni correction were performed following the ANOVAs. Considering the number of comparisons being performed, the Bonferroni correction was selected to reduce the potential for Type 1 error [179]. For behavioural measures two, 2-way repeated-measures ANOVA were performed for the dependent measures of (1) mean response time (seconds) and (2) subjective reports of attention (/10) with LIMB (non-paretic/right, paretic/left), and CONDITION (attend toward, attend away) as factors. Pearson’s correlations were calculated between response times and electrophysiological measures (SEP amplitudes). Given the exploratory nature of this study, no adjustments for multiple comparisons were made [180,181].  98 5.2.8.1 Exploratory stepwise linear regression analysis  The dependent variable, CB&M, was explored with the following independent variables using all participants’ data: (1) RT, and (2) change scores of SEP amplitudes between rest and the two attention conditions for all EEG components and both legs. Change scores were used instead of raw amplitudes to examine whether level of gating (i.e. difference in SEP amplitude between rest and attention condition) predicted scores within the CB&M. For regression modeling, the assumptions of normality and independence of the residuals were tested using the Durbin–Watson statistic (d = 1.886) [243]. Collinearity between multiple predictors was evaluated using variance inflation factor (VIF < 2) and tolerance levels (above 0.1) [244]. Statistical significance was set at p ≤ 0.05. Results are reported in means ± standard deviation unless otherwise stated. 5.3 Results 5.3.1 Participants  Demographic information including lesion location and months post-stroke are included in Table 5.1.   99  Table 5.1: Demographics and clinical measures for all participants. Participant Sex Age CB&M SFAQ FM-LE Paretic side Months post stroke Lesion location S01 M 74 61 123.2 25 L 66 R TH S02 M 78 48 119.7 24 L 20 R PICA, lateral medulla S03 M 74 54 109.5 25 R 65 L BG S04 M 78 14 88 19 R 109 L Pons S05 F 73 70 107.6 28 L 13 R IC S06 F 53 37 111.8 23 L 71 R MCA S07 M 78 58 88.6 20 R 44 L Pons S08 F 78 34 108 24 R 45 L IC, CR S09 F 57 53 101 20 R 9 L BG, CR S10 F 56 50 111.3 26 R 27 L TH S11 F 36 32 62.3 18 R 68 L BG Mean ± SD  66.8 ± 14.1 46.5 ± 15.9 102.8 ± 17.4 22.9 ± 3.2  48.8 ±  30.4  HC01 M 76 64 108     HC02 F 72 68 111.1     HC03 M 73 71 117     HC04 F 74 87 120.2     HC05 F 60 82 83.6     HC06 F 74 64 118.2     HC07 M 75 71 115.2     HC08 F 56 92 108.3     HC09 M 81 77 112.6     HC10 F 66 84 121.9     HC11 F 71 69 103.2     HC12 F 62 75 112.7     HC13 F 61 86 120.1     Mean ± SD  69.3 ± 7.5 76.2 ± 9.3 111.7 ± 10.1     S = stroke, HC = healthy control, SD = standard deviation, M = male, F = female, CB&M = community balance and mobility scale, FM-LE = fugl-meyer lower extremity, R = right, L = left, TH = thalamus, PICA = posterior inferior cerebellar artery, IC = internal capsule, MCA = middle cerebral artery, CR = corona radiata, BG = basal ganglia  100  After screening, one healthy older adult and two individuals with stroke were excluded due to inability to detect the onset of vibration, the cue for the experimental task. Thus, a total of 24 individuals participated in this study: thirteen healthy older adults (69.3 ± 7.5 years old; 4m, 9f), and eleven individuals with stroke (66.8 ± 14.1 years; 5m, 6f) (Table 5.1). Individuals with stroke had FM-LE scores between 18 to 28 (of 28) with an average level of impairment of 22.9 ± 3.2. Seven individuals’ paresis involved the right leg and for four individuals, the left. Individuals after stroke were 48.8 ± 30.4 months post-stroke at assessment. 5.3.2 Electrophysiology  SEP traces are shown in Figure 5.1 from a representative older healthy adult and stroke participant.   101  Figure 5.1: SEP traces at Cz electrode of representative older healthy control (A) and participant with stroke (B).  All variables were normally distributed according to the Shapiro-Wilk test (p > 0.001) [179], except the N40 component paretic/left attend toward and attend away, and the P50 rest non-paretic/right SEPs (Table 5.2).  102  Table 5.2: Electrophysiology and behavioural results.    Electrophysiology Behaviour Group Leg Condition N40 (µV) P50 (µV) N70 (µV)      Raw Change Score Raw Change Score Raw Change Score Movement RT (s) Subjective Attention (/10)            Older healthy controls Left Rest 2.2 ± 1.6 -- 2.0 ± 0.9 -- 3.1 ± 1.6 -- -- -- Attend Toward 1.3 ± 1.0 -0.9 ± 1.1 1.7 ± 1.0 -0.3 ± 0.6 2.2 ± 1.1 -0.9 ± 1.2 1.0 ± 0.4 8.3 ± 1.4 Attend Away 1.6 ± 1.4 -0.6 ± 0.9 1.8 ± 1.2 -0.2 ± 0.8 2.7 ± 1.4 -0.4 ± 0.9 1.1 ± 0.5 8.5 ± 1.0           Right Rest 2.2 ± 1.1 -- 1.7 ± 0.6 -- 4.2 ± 2.3 -- -- -- Attend Toward 1.2 ± 1.2 -1.0 ± 0.9 1.7 ± 0.8 0.0 ± 0.6 2.5 ± 1.3 -1.7 ± 1.7 1.2 ± 0.5 8.2 ± 1.6 Attend Away 1.4 ± 0.9 -0.9 ± 0.7 1.9 ± 1.1 0.3 ± 1.0 3.6 ± 2.3 -0.5 ± 1.4 1.2 ± 0.6 8.3 ± 1.5            Stroke Paretic Rest 1.5 ± 1.2 -- 2.7 ± 1.8 -- 4.0 ± 2.1 -- -- -- Attend Toward 1.3 ± 1.0 -0.2 ± 0.8 1.7 ± 1.7 -1.0 ± 1.3 2.2 ± 1.4 -1.9 ± 1.1 1.2 ± 0.5 8.5 ± 1.1 Attend Away 1.5 ± 1.0 0.0 ± 0.5 2.3 ± 1.3 -0.4 ±1.3 2.5 ± 1.8 -1.5 ± 1.3 1.2 ± 0.7 8.8 ± 0.7           Non-Paretic Rest 1.7 ± 0.8 -- 2.8 ± 1.7 -- 3.8 ± 2.0 -- -- -- Attend Toward 1.5 ± 0.9 -0.2 ± 1.0 2.3 ± 1.7 -0.5 ± 0.7 2.8 ± 1.5 -1.0 ± 1.8 1.1 ± 0.4 8.7 ± 1.1 Attend Away 1.5 ± 0.8 -0.2 ± 0.7 2.5 ±1.5 -0.2 ± 0.7 2.9 ± 1.7 -1.0 ± 1.6 1.1 ± 0.4 8.3 ± 1.3 Values are shown as mean ± standard deviation; µV = microvolts; RT = response time; s = seconds.   103 After log transformation, normality was not achieved; however, after square root transformation, normality was achieved for all N40 and P50 variables [179]. The independent t-tests were all non-significant (p > 0.05) indicating similar SEP amplitudes between groups, except for the P50 rest non-paretic/right (p = 0.046) with larger P50 amplitudes at rest in individuals with stroke. Using the Greenhouse-Geisser adjustment for the N40 amplitude over Cz, there was a CONDITION X GROUP interaction (F(1.579, 34.729) = 4.239, p = 0.030), with post-hoc tests indicating, for older healthy adults only, smaller amplitudes for attend toward and attend away when compared with rest (p = 0.001, p < 0.001, respectively, Figure 5.2B).    104  Figure 5.2: ANOVA results (main effects panel A) and post-hoc results (panel B) of SEP component amplitudes.  Blue column indicates rest condition; red the attend toward; green the attend away. Bars indicate standard error. Stars indicate significant post-hoc results where p ≤ 0.05.  No other interaction effects were significant (p ≥ 0.449). Individuals with stroke did not demonstrate differences between conditions for the N40 amplitude (p ≥ 0.901). For the P50 amplitude, no interaction effects were significant (p ≥ 0.143). A main effect of CONDITION was found (F(2, 44) = 4.920, p = 0.012). Post-hoc tests showed smaller amplitudes for attend toward than rest for both groups (p = 0.013, Figure 5.4A). For the N70 amplitude, no interaction effects were found (p ≥ 0.060); however, a main effect of CONDITION was present (F(2, 44) =  105 19.754, p < 0.001; Figure 5.2A). Post-hoc tests demonstrated differences between conditions (Figure 5.2A). Specifically, compared with rest, N70 amplitudes were smaller for both attend toward and attend away (p < 0.001, p < 0.001, respectively; Figure 5.2A). 5.3.3 Behavioural and functional measures  Across all movement conditions, there were 3.4 ± 3.7 and 3.5 ± 3.4 trials with anticipated steps or incorrect responses (i.e. plantarflexion without vibration cue) for individuals with stroke and older healthy adults, respectively. These trials were excluded prior to further analysis. After exclusion, the response times for all conditions ranged from 1.01 ± 0.39 s (attend toward left) to 1.22 ± 0.61 s (attend away right) for older healthy adults, and for individuals with stroke, from 1.06 ± 0.38 s (attend toward non-paretic) to 1.24 ± 0.51 s (attend toward paretic)  (Figure 5.3A, Table 5.2).    106  Figure 5.3: Movement response times (A) and subjective reports of attention (B).   The star indicates significant post-hoc results. Purple bars indicate the attend toward condition; Orange bars indicate the attend away condition.  The ANOVA was significant with a LIMB X GROUP interaction (F(1, 19) = 6.031, p = 0.024), with post-hoc tests indicating slower RT for the right leg compared with the left for older healthy adults (p = 0.047; Figure 5.3A). Response times were not correlated with any electrophysiological measure (p ≥ 0.112).  Subjective reports of attention for the older healthy adults rated the attend toward condition on the right as the most difficult (8.15 ± 1.625 of 10), whereas individuals with stroke rated the ability to pay attention the highest for attend away from the paretic limb (8.77 ± 0.684 of 10; Figure 5.3B, Table 5.1). The ANOVA did not show an interaction effect and was not significant for LIMB (F(1, 22) = 1.078, p = 0.310) or CONDITION  (F(1, 22) = 0.167, p = 0.686) for subjective scores.   107  For the CB&M, older healthy adults scored on average 76.2 ± 9.3 of 96 and individuals with stroke scored 46.5 ± 15.9 of 96. The mean score on the SFAQ was 111.7  ± 10.1 for older healthy adults, and individuals with stroke scored 102.8  ± 17.4, indicating an average level of sustained attention function for both groups.  5.3.4 Regression results  For the CB&M, three variables accounted for 69% of the variance in CB&M scores (Table 5.3, Figure 5.4).    108  Table 5.3: Regression results.  Model Predictors R R2 R2 change Sig. change Overall F Overall Sig.  1  (a)RT attend away paretic/left  0.569 0.323 --- --- 9.078 0.007 2 (a)RT attend away paretic/left,  (b)N70 change score attend toward paretic/left  0.763 0.583 0.260 0.004 12.574 <0.001 3 (a)RT attend away paretic/left,  (b)N70 change score attend toward paretic/left,  (c)N40 change score attend toward non-paretic/right  0.831 0.691 0.108 0.026 12.669 <0.001 RT = response time; a,b,c predictors.   109  Figure 5.4: Partial regression plots of predictors on the x-axis with CB&M on the y-axis.  A: Response time during the attend away condition from the paretic/left vs. CB&M; B: Change scores for the N70 during attend toward on the paretic/left; C: Change scores for the N40 during attend toward on the non-paretic/right. Orange indicates participants with stroke, blue indicates healthy control. CB&M = community balance and mobility scale, S = seconds, µV = microvolts.  RT for attend away on the paretic/left accounted for 32% variance, so that slower RT were associated with lower CB&M scores (R2 = 0.323, p = 0.007; Table 5.3, Figure 5.4A). The SEP amplitude change from rest in the attend toward condition in: 1) N70 amplitude on the paretic/left (R2 change = 0.260, p = 0.004; Table 5.3, Figure 5.4B), and 2) the N40 for the non-paretic/right limbs (R2 change = 0.108, p = 0.026; Table 5.3, Figure 5.4C) added significantly to the model. For the N70 change scores, greater levels of gating linked with lower CB&M scores, whereas for the N40 change scores, greater levels of gating linked with higher CB&M scores (Figure 5.4B,C). The final model, including all three variables, accounted for 69% variance in CB&M scores (F(3,17) = 12.669, p < 0.001).   110 5.4 Discussion  This is the first study establishing that individuals with stroke do not show attention-mediated gating of the N40 component from irrelevant somatosensory information during the early planning phase in the leg compared with older adults; however, gating was present for the P50 and N70 for both groups (Table 5.2, Figure 5.2). Second, neurophysiological measures of this gating during early planning explained significant and unique variance in a measure of community balance and mobility when combined with response time for the movement (Table 5.3, Figure 5.4). Thus, the ability to gate irrelevant somatosensory information appears to be important during early planning for both older adults and after stroke. 5.4.1 No attentional modulation of the N40 after stroke  In this study, the amplitude of the N40 did not vary based on attention condition for individuals with stroke. With tibial nerve stimulation, the N40 component is generated by S1, specifically area 3b, and represents the arrival of somatosensory information through thalamocortical afferent projections to the somatosensory cortex [156,159,245-247]. In the arm, task relevant modulation of sensory information and gating involves the PFC [122]. Considering that the participants in this study did not have lesions in S1 or PFC, it may be that dysfunction in interneurons or disruption in pathways connecting to area 3b, may explain the lack of attention-mediated inhibition of the N40. Otherwise, the loss of N40 modulation may be due to the altered influence of the PFC on the thalamic gate, though these changes would need to be beyond changes related to aging as the older adults in this study did show gating of the N40.  111 5.4.2 Community balance and mobility explained by change scores on the N70 and N40   The N70 and N40 change scores were included as two of the significant predictors for the CB&M (Table 5.3, Figure 5.4B,C). The direction of these predictors are in positive and negative directions so that less gating on the N70 is related to higher function, with the converse relationship for the N40 (Figure 5.4). A potential neurophysiological explanation for these findings involves the generators for these components. The N40 (area 3b) amplitude when attending toward the leg may represent suppression of irrelevant information with higher gating related to higher function; individuals with the ability to suppress extraneous somatosensory information at the earliest processing stage can perform challenging balance and mobility tasks.   Area 5 in the posterior parietal cortex generates the N70 with tibial nerve stimulation [212,240]. The function of the posterior parietal cortex is to integrate sensory information with a planned goal [248], with lesions in area 5 affecting the accuracy of step placement [249]. The positive relationship of the N70 with CB&M suggests that lower levels of gating relate to higher levels of function so that in essence, more irrelevant information is being received by area 5. For the posterior parietal cortex to accurately integrate sensory information compatible with the movement goal, receiving more irrelevant information may serve as a redundancy mechanism to ensure the irrelevant information is not required for successful planning and performance. To do this well, more irrelevant information can facilitate better performance, and may require more time to process the received sensory information. The cost for ignoring potentially relevant information in the leg is high; inaccurate performance may lead to a fall.   For some participants, instead of gating, a loss of inhibition of irrelevant somatosensory information was shown: the N70 for a select few older adults, and the N40 for some individuals  112 with stroke (Figure 5.4B,C). Age-related changes also occur in the posterior parietal cortex where greater parietal recruitment improved behavioural performance on a cognitive task compared with young adults [250]. The greater recruitment may mean less gating is associated with a normal aging process for the N70. Despite this variability, the regression results show the importance of levels of gating as explaining variability in functional balance and mobility (Table 5.3).  5.4.3 No leg differences for individuals with stroke  As the aim of this study was to examine attentional processes during planning, the task was designed so that it could be performed with the paretic limb with relative ease. It was surprising that no neurophysiological or behavioural differences were found between the paretic and non-paretic legs (Table 5.2, Figure 5.2, Figure 5.3). Differences in motor performance between the paretic and non-paretic legs are well established [251-253]. It is possible that the neurophysiological measures may have shown differences between the legs with a more difficult task, similar to previous work that showed facilitation of SEP amplitudes when balance challenge increased [254]. Another possibility is that a more severely affected group, or a larger sample size would show between-leg neurophysiological differences. Regardless of the task, gating may still be preserved. Indeed, the results of the regression indicate that performance of more challenging community based tasks is explained by the levels of gating (Table 5.3). Even with a relatively easy task, neurophysiological measures of attention predicted motor performance (CB&M regression Table 5.3).  113 5.4.4 Limitations  There are a few limitations to this study. The sample was not homogenous for lesion location thus, it is difficult to draw conclusions regarding the effects of lesion location; however, it is consistent with other studies that use heterogeneous lesion locations to increase clinical utility and applicability [255-257]. Also, the sample size was small but consistent with other attention based studies examining neurophysiology with EEG [119]. Lastly, both right and left side lesions were included in this study. There is some evidence to support that the left supramarginal gyrus in the parietal cortex plays a role in attention [42]. Individuals with stroke in the left parietal cortex have difficulty disengaging attention from one planned arm movement to another, which suggests that it has a role in attention after a stroke [112]. It is possible that the individuals who lack attention-mediated gating may have had micro or silent lesions that encompass this brain region that standard MRIs or CT scans do not capture. None of the participants with stroke had a stroke within the PFC or M1/S1 for the leg (Table 5.1), so differences between groups are likely due to stroke and not from aging alone.  5.5 Conclusions  Though both groups show some attention-mediated gating of irrelevant somatosensory information from the leg during early planning, individuals with stroke had abolished N40 component gating. Additionally, the levels of gating, along with response time, explained variance in a measure of community balance and mobility. If attention alters the somatosensory stimuli from a leg movement, then directing attention may affect safe community walking.  114 Chapter 6: General discussion  6.1 Overview  The experiments in this thesis were designed to evaluate attention and planning of voluntary leg movements in individuals after stroke. Summarized below are the major findings from each research chapter, followed by a more thorough discussion of overarching dissertation themes. General limitations based on experimental techniques and approaches are considered, with a proposed theoretical framework that depicts the possible interactions between planning and attention. The dissertation concludes with a discussion of the implications and recommendations for future work that builds on the findings of this thesis and the proposed framework.  6.2 Planning differs between individuals with high and low levels of motor performance, but not between the paretic and non-paretic legs (Chapters 2 and 3)  Given previous work had shown differences in planning between the paretic and non-paretic arms, the first two experiments were designed to examine planning processes as evidenced at the level of the muscle (Chapter 2) and brain (Chapter 3). In both Chapters 2 and 3, contrary to expectations, no differences were found in planning between stepping with non-paretic and paretic legs. However, individuals with greater impairment showed larger levels of co-contraction (Chapter 2), and greater cognitive effort and longer planning durations (Chapter 3).   115  Chapter 2 investigated the characteristics of planning for the initial foot contact phase of walking and examined thigh muscle activity. Measures of muscle onset and offset timing, amplitude and co-contraction were found to be similar between the paretic and non-paretic legs, regardless of the level of motor impairment. When normal co-contraction levels measured from healthy controls were used to stratify the participants with stroke, the high co-contraction group had lower balance and mobility scores together with longer durations and higher levels of thigh muscle activity. The low co-contraction group had muscle activity patterns similar to healthy controls together with moderate community balance and mobility scores. These findings suggest that individuals with higher motor impairment compensate through increased levels of co-contraction. Compensation is a behavioural strategy to bypass impairments to allow for standing and walking performance to occur with both positive and negative effects on function [19]. Though high co-contraction can be used as a compensatory strategy, it may also limit these individuals from achieving better motor outcomes if transition to lower levels of co-contraction is not achieved. Relying on compensation for motor performance can come with costs. With aging, increased muscle co-contraction may be necessary to counteract decreases in postural control [258], but it increases the risk of falls during a perturbation [259].  Planning regions contribute to the timing of muscle onset for step initiation in healthy adults [260]. Disruption to planning regions as a result of stroke alters preparatory ankle muscle activity [192]. Therefore, examining planning involved with step initiation in individuals after stroke was a logical next step (Chapter 3). The paretic leg was examined as both the stepping leg and the stance leg. Interestingly, there were no differences between legs with respect to step duration, planning or muscle parameters. Planning duration and amplitude were correlated between paretic and non-paretic legs suggesting symmetry in the planning of stepping, as  116 stepping duration in individuals who stepped more slowly with the paretic leg required greater cognitive effort and time to plan both a paretic and non-paretic step, respectively. Further, planning of a paretic or non-paretic step was associated with the onset timing of knee muscle flexor activity. It is possible that the planning required for postural control to move the paretic leg or to stand on the paretic leg is so similar, that the measurement of the MRCP is comparable. Another conceivable explanation is that the MRCP reflects the planning of both the stance and stepping leg simultaneously. However, potential planning differences between legs with other types of stepping, such as for a lateral or backwards step, cannot be ruled out. 6.3 Stroke alters attention-mediated gating of irrelevant somatosensory stimuli during early planning in the leg for the N40 (Chapters 4 and 5)  Considering the impact of attention on performance [261] and recovery after stroke [222], the next two experiments (Chapter 4 and Chapter 5) were designed to explore the influence of attention on irrelevant somatosensory information reaching the cortex. Gating during the early phase of planning had not yet been examined. Therefore, establishing if gating of irrelevant somatosensory stimuli could be mediated by attention during early planning for the leg in young adults was the first step (Chapter 4). Then, a similar experimental protocol was employed in older adults and individuals with stroke to determine whether stroke affected this pathway (Chapter 5). In both chapters, only irrelevant somatosensory stimuli were investigated.   Attention was found to impact somatosensory processing of irrelevant information during the early stage of planning plantarflexion movements in young adults. This was demonstrated through gated N40 and N70 amplitudes when attending ipsilaterally to the planned leg movement. Based on the results of Chapter 4, it was speculated that stroke would affect the way  117 attention alters the gating of irrelevant somatosensory information reaching the cortex. This line of reasoning was further supported by the results of Chapter 4 in that the somatosensory information from the leg in the early planning phase was altered by attention in young adults.  The first aim of this study was to determine whether gating of irrelevant somatosensory information was present in individuals with stroke compared with older healthy controls during the early planning phase of plantarflexion. The second aim was to examine whether the level of gating was important for community balance and mobility scores. The main finding was that the N40 was not gated by attention after stroke. Both young adults (Chapter 4) and older adults (Chapter 5) show a similar attention-mediated gating pattern for the N40 when attending toward a leg receiving irrelevant stimulation. Considering that the generator of the N40 is area 3b [156], and that individuals with lesions in S1 and PFC were excluded, a plausible yet speculative neurophysiological interpretation is that disruption or dysfunction in pathways connected to area 3b may explain the lack of inhibition with attention after stroke.   This is the first study establishing that individuals with stroke and older adults gate irrelevant somatosensory information from the P50 and N70 components during early planning for stepping. Thus, gating appears to be intact for some components during early planning for both older adults and after stroke. A more thorough theoretical exploration of the P50 findings will take place in Section 6.5. The N70 and N40 gating explained significant and unique variance in the community balance and mobility scale, in addition to response time for cued plantarflexion. Higher levels of attention-mediated gating of the N70 toward the paretic or non-dominant leg (i.e. the more negative the change score) were linked to lower balance and mobility. Area 5 (N70 generator) in the posterior parietal cortex for higher-level integration of gated sensory information with a movement goal [248] may need more time to process the  118 incoming sensory information. If more time is taken, it may result in better performance of balance and mobility tasks.    6.4 Between-leg differences in performance but symmetry in planning after stroke (Chapters 2,3,5)  There are many differences between the paretic and non-paretic legs for the motor performance of walking and standing balance outlined in previous work. These differences include asymmetry in spatiotemporal parameters including step length and swing time [183], and decreased weight bearing on the paretic leg in standing [253]. Higher asymmetry is linked to more severe impairments in the leg and foot [252]. The paretic leg has more pronounced timing abnormalities than the non-paretic leg during walking [165]. These between-leg differences in performance are also present in young adults where leg dominance has been shown to predict asymmetries in strength, balance, and walking [225]. Thus, it was plausible that the factor of ‘leg’ may affect early planning and attentional processes. However, this was not the case in any of the experiments within this thesis.   Despite evidence supporting asymmetries in motor performance after stroke, in this dissertation, planning was found to be symmetrical in the leg. This finding was repeated across electrophysiological modalities of EMG (Chapter 2) and EEG (Chapters 3,5), with three distinctive experimental methodologies (Chapters 2,3,5), and in multiple phases of stroke recovery (subacute in Chapters 2,3; chronic in Chapter 5). This is contrary to studies of planning and attention in the arm [52,53,56,262,263], where paretic arm planning takes longer with greater cognitive effort, and irrelevant sensory information impairs performance on attention-demanding tasks. Many arm tasks can be unilateral and asymmetrical, whereas leg tasks such as  119 stepping and walking inherently require high amounts of symmetry. Perhaps the cognitive effort and time to plan the postural control needed for stepping and stance simultaneously, overshadows that for planning to move the paretic leg to step, or to maintain balance for stance. Indeed, there is evidence in young adults that multiple types of leg movements show equivalent planning implying comparable motor programs are needed to plan the control of upright posture [66].   It is important to recognize that methodological limitations potentially counter the previous speculation concerning symmetry of planned leg movements. After stroke, the paretic leg is weaker and produces force more slowly [164] so that normalizing to a smaller maximum amplitude and step duration could produce similar relative amplitudes and timing to the non-paretic leg. For EEG, SEP evaluation of neurophysiological processes in the leg utilizes the Cz electrode located at the vertex, regardless of which leg is being examined [124,211], unlike in the arm where cortical potentials can be measured in the contralateral hemisphere to the stimulated limb. The location of the leg in the motor and sensory homunculus within the interhemispheric fissure may not allow for differentiation of ipsilesional and contralateral potentials within the EEG modality. What is measured at Cz reflects the summation of facilitation and inhibition as a result of attention and planning processes occurring for both the right and left legs concurrently.  6.5 Individuals with stroke and older adults differ from young adults for attention-mediated gating of irrelevant somatosensory stimuli for the P50 (Chapters 4,5)  Individuals with stroke and older adults had significant gating for the P50 during the attend toward condition (Chapter 5), which was absent in young adults (Chapter 4). The significant gating may be explained by changes to the PFC that occur with aging. At rest, SEP  120 amplitudes from median nerve stimulation are larger for older adults than young adults [264], due to the reduced inhibition of the PFC [123]. After a stroke in the PFC, enlarged SEP amplitudes at rest are also found compared with healthy controls [121]. Older adults have reduced attention-based modulation of irrelevant somatosensory information from the hand compared with young adults, consistent with age related declines in the executive function of the PFC [123]. It is proposed that the magnitude of gating depends on SEP amplitudes at rest, and that augmented gating in older adults is a result of these enlarged resting SEPs [265]. If older adults have higher amplitudes of SEPs at rest for the P50 (Chapter 5), they may need to have added gating to attain the normal levels shown by young adults (Chapter 4); however, this has not been tested statistically between these groups.  Older age is associated with an increase in susceptibility to distraction by task-irrelevant stimuli in the visual [266] and auditory [267] modalities, and for the hand in the somatosensory [123] modality; these are related to decreased inhibitory function of the PFC [123,267]. Further, this distraction is modality-dependent and linked to the unique gating mechanisms used by different modalities (i.e. auditory gating at central and peripheral levels vs. visual gating at central levels of processing [266]). However, Gmehlin et al. 2011 found that gating of irrelevant stimuli was preserved with aging within the auditory modality [268]. Considering the role of the PFC for attention-mediated gating [119,122], and the potential for gating to be preserved, it is plausible that gating of irrelevant somatosensory information from the leg is maintained for some individuals in the somatosensory modality, despite aging being linked to increased levels of distraction by task-irrelevant stimuli [123].   121 6.6 The early planning phase is essential to motor performance (Chapters 4,5)  During an irrelevant somatosensory stimulus, a vibrotactile stimulus was used to cue plantarflexion to draw attention toward or away from the stimulated leg. Similar experimental protocols were employed in previous work demonstrating attenuated SEP amplitudes from the leg during late planning [235]. However, key differences in this paradigm and analysis approach enabled a novel perspective. Importantly, this protocol allowed us to compare attention directed toward and away from the leg, and the analysis technique enabled measurement in the early phase of planning. Early processes are distinct from late planning. Early planning is vital as it reflects attention toward task complexity [50], timing of movement initiation [268], and functions to plan multiple movements ahead of a single action [269]. Decisions regarding movement details, such as specific effector parameters, are made at the last possible moment in late planning [270]. The ability to suppress irrelevant somatosensory information during early planning may serve to enhance motor performance of stepping by shortening response time, with better prediction of imminent movements.  The N70 and N40 change scores measured during early planning were included as two of the predictors for CB&M scores in older adults and individuals with stroke (Chapter 5). Bearing in mind that individuals with stroke did not show gating of the N40 during the attention conditions (Figure 5.2), it may be that individuals with the ability to suppress extraneous somatosensory information measured by the N40 during early planning can perform balance and mobility at a more challenging level (Chapter 5, Figure 5.4). Otherwise, it may be a difference related to aging or stroke. Considering that there was no significant gating of the N40 in this study in individuals with chronic stroke (Chapter 5), it would be interesting to examine planning  122 neurophysiology longitudinally to see whether the lack of N40 gating is occurs immediately after stroke, or if it is a negative compensation from the stroke. 6.7 Limitations  There are several important limitations to the interpretation of the research findings in this dissertation associated with the design of the experiments. These are detailed below.   (1) The participants in these experiments were all independent, community living, and able to walk without assistive devices. Due to the need for a minimal level of standing balance for experiments in Chapters 2 and 3, and the requirement for controlled plantarflexion in Chapter 5, a severity bias may be introduced as individuals with the most severe strokes were not tested. To minimize this bias, recruitment aimed to include individuals along the range of severity allowed by the inclusion criteria. Though ranges of stroke severity were included, no individuals with severe leg impairment participated, as measured by the CMSA and FM-LE. As such, the findings cannot be fully generalizable to the range of impairments seen post-stroke. Additionally, individuals with cognitive impairment were excluded, so the understanding of planning and attention gained from this dissertation applies only to those individuals with scores within normal limits on common cognitive screens (MMSE, MOCA). It is possible that planning and attention differences between the legs may be found in individuals with cognitive impairment. Despite these limitations to generalizability, the findings from this dissertation are applicable to mild to moderately impaired individuals after stroke with normal levels of cognition.     (2) For Chapters 3 through 5, one electrode (Cz) was chosen to measure planning and attention, similar to previous work [212,235]. It is a logical choice considering it is the only electrode directly over the leg region in the primary sensorimotor cortices. With a single  123 electrode, however, it is not possible to examine the differences between the hemispheres in a stepping task (Chapter 3). In the arm, with the more lateral orientation of the hand region in the sensorimotor cortex, differentiation can be made. Despite this limitation, many studies examine the leg using Cz and have made important discoveries using this approach [212,235,254].  (3) There are also a few potential limitations to generalization based on stroke lesion location. To mitigate this potential confound, individuals were excluded with lesions involving the PFC (functions for attention-mediated gating) and anterior cerebral artery (blood supply to primary sensory cortex) in Chapter 5. There are known differences in clinical presentation for some individuals with right hemisphere lesions such as pusher syndrome or visuospatial neglect [271]. To mitigate the effect of such impairments, individuals with stroke were screened for neglect using the Bell’s test [272]. Interpretation of these findings must be made in light of this limitation.   (4) The experiments in individuals with stroke were at a single time point and not recovery focused. Thus, the findings in this thesis cannot be used to inform stratification in clinical trials. A future longitudinal study examining attention and planning throughout recovery may be important to understand how these factors emerge over time.  (5) The strength of the research supporting areas 1 and 5 as generators of the P50 and N70, respectively, with tibial nerve stimulation is limited (Chapters 4 and 5). While MEG, intracerebral, and scalp recordings have confirmed that area 3b in the foot area of S1 generates the N40 component [156-158], scalp recordings with source localization methods show that the P50 generator is area 1 [159], and area 5 generates the N70 [156]. Intracerebral and MEG studies  124 of the P50 and N70 may strengthen findings from research that examines attention-modulated gating. 6.8 Proposed theoretical framework for brain regions supporting attention and planning of voluntary goal-directed leg movements  A recent review of the parietal cortex and its role with reaching and grasping [273] recommended that future work develop network-based models for the recovery of reach and grasp versus the past emphasis on lesion-based studies. These network studies can potentially identify impairment and recovery based on functional systems that characterize deficits, predict who can benefit the most from therapy, and identify remaining regions that can be accessed for better functional recovery. The same reasoning is extended to future studies of planning, attention, and movements of the leg after stroke.   A proposed model including anatomical and functional regions known to be involved with attention and planning is demonstrated in Figure 6.1.    125  Figure 6.1: Theoretical framework for how attention impacts planning for voluntary leg movements after a stroke that affects attention-mediated gating.  M1 = primary motor cortex, PFC = prefrontal cortex, PMC = premotor cortex, PPC = posterior parietal cortex, S1 = primary somatosensory cortex, SMA = supplementary motor cortex, TH = thalamus.  In the early stage of planning, the prefrontal cortex (PFC) selects a movement-related goal and based on this goal, regulates the flow of information leaving the thalamus (Figure 6.1A). The thalamus directs the remaining afferent somatosensory information to the primary somatosensory cortex (S1). Through the frontoparietal network, the PFC concurrently sends goal-directed information to the posterior parietal cortex (PPC). Somatosensory signals from S1 converge on the PPC where integration of the sensory information with the movement goal from PFC occurs. The SMA and PMC areas receive this information and generate a plan based on the attention-mediated somatosensory signals and movement goals. In the late stage of planning, the SMA and  126 PMC link to the primary motor cortex where the specific features of the plan are determined (Figure 6.1B). In this way, walking and balance movements are voluntary, goal directed, and pre-planned, but not executed prematurely. If dysfunction with inhibition of irrelevant somatosensory information occurs after stroke, it is possible that the PFC allows too much information to leave the thalamus that floods S1 with unnecessary sensory information (Figure 6.1C). The PPC attempts to integrate this sensory information with the movement goal, with the SMA and PMC generating a plan based on some irrelevant sensory information. Levels of cognitive effort and duration may plausibly increase in this situation, with detrimental effects on performance of balance and walking (Figure 6.1D).  In general, stroke may impact attention and planning beyond the lesion location within the network by interfering with input and/or output of the regions involved. Functional performance of balance and walking can be impacted as a result. It is unknown whether lesions affecting attention produce compensatory activity, or which regions may compensate if stroke occurs. Perhaps the planning functions of the SMA and PMC could be interchangeable if a lesion was present in one planning region. Though there appears to be overlap or redundancy in the regions involved with planning and attention that may be potentially exploited for recovery after a stroke, multiple strokes or bilateral stroke may overload these redundancies with exponential impact on balance and walking. If multiple regions are affected by stroke, less capacity for compensation remains. While stroke may produce local disruption in these regions, it is also possible that stroke alters the structure or function of any region within this network with a cascade of impact. As this is the first rehabilitation focused theoretical model integrating attention within the early phase of planning, future studies may be warranted to compare how the arm may be similar or different to the leg. Furthermore, longitudinal studies focused on lesion  127 location may help to further elucidate the nuances in the network, and how the network may evolve with time post-stroke.  6.9 Implications and future directions  This thesis contributes unique knowledge to the neurophysiology of attention and planning of the leg in healthy young and older adults, as well as stroke. There is a great need in both the scientific and clinical community to understand appropriate measurement, diagnoses, and interventions for these deficits after stroke. The first study to examine the neurophysiology of planning in 1964 [60] generated great interest with the number of papers examining this topic now numbering in the thousands. Attention research followed with specific aspects of attention examined in the late 1980’s [51]. Despite years of study, planning and attention deficits and how they may interact with stroke recovery are not wholly understood. This dissertation work was an important first step toward this line of research. Potential applications of this dissertation to neurorehabilitation after stroke are expanded further in a narrative review article (Appendix A). Notwithstanding the decades long study of planning and attention, much is left to be explored. The following paragraphs outline prospective topics of study. 6.9.1 Extend methodology to develop knowledge of planning and attention of leg movements  Though between-leg symmetry is a common theme through this dissertation, future work is required to probe this finding further. One way to test this discovery is to apply the same experimental protocol in Chapter 2 where individuals walked at a comfortable speed, and instead ask individuals to walk as fast and slow as possible. Within individuals after stroke, fast walking compared with comfortable walking improves step length without increasing common  128 compensations of hip hiking or circumduction [274]. Walking slowly increases the mechanical work of the paretic leg 1.7 times more than normal [275]. Taken together, testing different walking speeds may show differences between legs and/or the vary thigh muscle activity to plan initial contact during walking.  For Chapter 3, alterations to stepping duration may show between-leg differences. As the magnitude of the MRCP scales with the speed of elbow movement [204], so it is possible that future studies may show differences between legs with: (1) stepping or walking at a fast pace, (2) higher levels of severity, or (3) a larger sample size. The planning required for complex movements, such as obstacle avoidance can be compared with step initiation to determine whether task complexity alters planning after a stroke at the level of the muscle and brain.   Based on Chapters 4 and 5, future study may be warranted with electrocorticography (electrodes placed intracranially on the exposed brain) to clarify whether between-leg differences are present, and whether differences relate to attention and planning. Alternatively, a new method or technology could be created considering the specific needs of examining cortical regions involved with leg movements. Using the current experimental protocol, higher levels of stroke severity could be tested and compared with lower levels to probe whether impairment influences attention during early planning.    Beyond singular modalities, combining imaging such as EEG with transcranial magnetic stimulation (TMS) may provide novel insights into potential treatments for planning deficits. In young adults, manipulating sensory stimuli to the hand increased corticospinal excitability measured by TMS and enlarged the MRCP in a single session [276]. Additionally, an inhibitory TMS technique reduced the MRCP amplitude for planning a thumb movement [277]. As many  129 individuals with stroke have altered cortical excitability [278], and in Chapter 3, slow stepping is related to higher MRCP amplitudes, it is possible that TMS techniques can be examined to assess whether alterations to excitability can influence planning and performance. Structural magnetic resonance imaging (MRI) could be used to ratify or expand on the theoretical model (Section 6.8) by linking structural images to the function measured by the neurophysiological parameters tested in this dissertation (MRCP, SEP). A stroke lesion could be mapped onto the network and with the extracted structure and function measurements, related to planning (MRCP parameters), attention (SEP components) and performance. 6.9.2 Recovery of leg function through targeted planning and attention training   After 5 weeks of motor skill practice of a novel arm task, smaller cortical planning amplitudes were found suggesting more efficient planning with training [186]. For the leg, extensor muscle resistance training decreased planning effort levels, measured with MRCP, with earlier planning onsets thought to be through increased efficiency [187]. Furthermore, if given information that an upcoming external perturbation will be large, planning amplitudes increase in magnitude to match the perturbation [202]. Considering that planning can change through training in young adults, future work could test whether planning improves with neurorehabilitation in individuals with higher impairment who tend to have deficits with planning (findings from Chapters 2 and 3). Currently, neurorehabilitation after stroke utilizes three strategies, among others, to promote leg recovery after stroke: motor skill practice [279], muscle strengthening [280], and perturbation training [281]. However, it is unknown whether current neurorehabilitation practices improve planning after stroke. Based on the research described, it is a speculative but plausible physiological explanation for why some individuals show recovery of leg motor function through more efficient planning processes. Planning could  130 likewise become more efficient with targeted therapies yet to be developed and tested, such as rehabilitation to increase the speed of stepping movement (Chapter 3). Planning related changes could be identified with MRCP measurements before and after a therapeutic intervention and correlated to changes in muscle strength, coordination patterns, or standing balance and walking speed. It would be interesting to test whether individuals who respond to such therapeutic focus show gains in planning.  Auditory or visual sustained and alternating attention deficits, while common after stroke, do show improvement over time [222]. Training of auditory and visual sustained and alternating attention with the Integrated Visual Auditory Continuous Performance Test (IVA-CPT) (used in Chapter 4 and 5) with the arm after stroke improves certain aspects of cognitive function; however, motor attention again has not been studied, nor the impact of motor attention on performance or recovery [223]. There are a few interesting approaches to rehabilitation of attention that can be tested based on this dissertation. Attention to touch in the arm plays a role in cortical plasticity by modulating ongoing processing with long lasting effects on the cortex [282]. Applying this concept to the leg together with the findings outlined in Chapters 4 and 5, suggest that employing the same experimental protocol to train attention in the leg may have therapeutic effects of adaptive plasticity and thus, deserves further study. Importantly, it is unknown whether the effectiveness of walking and balance rehabilitation after a stroke is affected by neurophysiological deficits in attention and planning, and it is also unknown whether these deficits can be successfully remediated. These questions can now be examined longitudinally as this dissertation has tested dependent measures for both planning and attention. Current therapy may improve neurophysiological processes of attention and planning; however, if it does not, novel therapeutic interventions should be tested.  131 6.9.3 Develop clinical attention test specific for the leg  Given current knowledge of the importance of attention during early planning for motor function (explains significant variance in CB&M – Chapter 5), there is a need for a leg-based measure of attention that can be administered clinically. The IVA, an auditory and visual sustained and alternating attention test, did not relate to any of neurophysiological attention measures in healthy controls or individuals after stroke. Currently, clinical tests of attention to touch in the leg may probe attention indirectly through proprioceptive perception assessments like the great toe ‘up and down’ test [283], or with light touch tests for extinction [284]. A starting point for a more direct approach is using the methodology from Chapters 4 and 5. Future studies could consider collecting a rest and attend toward condition only as the attend away condition did not seem to differentiate levels of attention beyond the two mentioned. During this testing, a longer collection time for the attend toward condition would likely tax the sustained attention network more, and generate more errors in performance. The attention-mediated gating of trials without errors could be compared to trials with performance errors to elucidate whether the level of gating differs. This could be a starting point for a laboratory-based test that could become a proxy for a clinical measure to be developed. Both of these tests could help to determine if attention deficits that fall outside of neglect are going undiagnosed, untreated, or at least unmeasured. The full impact of these deficits in impairment or in recovery has yet to be determined. 6.10 Conclusions  This dissertation is the first to investigate the neurophysiology of planning and attention of leg movements after stroke. The question posited in the title of this dissertation asked whether  132 individuals after stroke are ready to move; the short answer is both ‘yes’ and ‘no’. In some ways, the planning processes appear to be intact bilaterally; however, the timing and magnitude of planning is altered and reflects greater levels of both cortical and muscle effort in those with higher leg impairment. For attention during planning, some gating processes are intact while others are lacking. A promising avenue for discovery is whether the residual capacity of the planning and attention network could compensate for local deficits, and alternatively, by identifying the gaps in these networks, researchers have targeted areas to examine for recovery and neuroplasticity.   133 References   [1] Heart&Stroke. Different strokes: Recovery triumphs and challenges at any age report  2017.  [2] Towards a common language for functioning, disability and health ICF. Geneva: World Health Organization; 2002.  [3] Gowland C, Stratford P, Ward M, Moreland J, Torresin W, Van Hullenaar S, et al. Measuring physical impairment and disability with the Chedoke-McMaster Stroke Assessment. Stroke. 1993 Jan;24(1):58-63.  [4] Duncan PW, Goldstein LB, Matchar D, Divine GW, Feussner J. Measurement of motor recovery after stroke. Outcome assessment and sample size requirements. Stroke. 1992 Aug;23(8):1084-9.  [5] Kostanjsek N. Use of The International Classification of Functioning, Disability and Health (ICF) as a conceptual framework and common language for disability statistics and health information systems. BMC public health. 2011 May 31;11 Suppl 4:S3.  [6] Kreighbaum Ellen BKM. Biomechanics: A quantitative approach for studying human movement. 4th Edition ed. Toronto: Allyn and Bacon 1996.  [7] Winter DA, Patla AE, Prince F, Ishac M, Gielo-Perczak K. Stiffness control of balance in quiet standing. Journal of neurophysiology. 1998 Sep;80(3):1211-21.  [8] Gunther M, Grimmer S, Siebert T, Blickhan R. All leg joints contribute to quiet human stance: a mechanical analysis. Journal of biomechanics. 2009 Dec 11;42(16):2739-46.  [9] Gunther M, Putsche P, Leistritz L, Grimmer S. Phase synchronisation of the three leg joints in quiet human stance. Gait & posture. 2011 Mar;33(3):412-7.  [10] Lee HJ, Chang WH, Choi BO, Ryu GH, Kim YH. Age-related differences in muscle co-activation during locomotion and their relationship with gait speed: a pilot study. BMC geriatrics. 2017 Jan 31;17(1):44.  [11] Peterka RJ. Sensorimotor integration in human postural control. Journal of neurophysiology. 2002 Sep;88(3):1097-118.  [12] Lam T, Anderschitz M, Dietz V. Contribution of feedback and feedforward strategies to locomotor adaptations. Journal of neurophysiology. 2006 Feb;95(2):766-73.   134 [13] Mochizuki G, Ivanova TD, Garland SJ. Postural muscle activity during bilateral and unilateral arm movements at different speeds. Experimental brain research. 2004 Apr;155(3):352-61.  [14] Garland SJ, Willems DA, Ivanova TD, Miller KJ. Recovery of standing balance and functional mobility after stroke. Archives of physical medicine and rehabilitation. 2003 Dec;84(12):1753-9.  [15] Booth FW, Weeden SH, Tseng BS. Effect of aging on human skeletal muscle and motor function. Medicine and science in sports and exercise. 1994 May;26(5):556-60.  [16] Deschenes MR. Motor unit and neuromuscular junction remodeling with aging. Current aging science. 2011 Dec;4(3):209-20.  [17] Mansfield A, Danells CJ, Zettel JL, Black SE, McIlroy WE. Determinants and consequences for standing balance of spontaneous weight-bearing on the paretic side among individuals with chronic stroke. Gait & posture. 2013 Jul;38(3):428-32.  [18] Genthon N, Rougier P, Gissot AS, Froger J, Pelissier J, Perennou D. Contribution of each lower limb to upright standing in stroke patients. Stroke. 2008 Jun;39(6):1793-9.  [19] Jones TA. Motor compensation and its effects on neural reorganization after stroke. Nature reviews Neuroscience. 2017 May;18(5):267-80.  [20] Rosa MC, Marques A, Demain S, Metcalf CD. Lower limb co-contraction during walking in subjects with stroke: A systematic review. Journal of electromyography and kinesiology. 2014 Feb;24(1):1-10.  [21] Lewek MD, Bradley CE, Wutzke CJ, Zinder SM. The relationship between spatiotemporal gait asymmetry and balance in individuals with chronic stroke. Journal of applied biomechanics. 2014 Feb;30(1):31-6.  [22] Raja B, Neptune RR, Kautz SA. Coordination of the non-paretic leg during hemiparetic gait: expected and novel compensatory patterns. Clinical biomechanics. 2012 Dec;27(10):1023-30.  [23] Peters S, Ivanova TD, Teasell R, Garland SJ. Is the Recovery of Functional Balance and Mobility Accompanied by Physiological Recovery in People With Severe Impairments After Stroke? Neurorehabilitation and neural repair. 2014 Mar 13.  [24] Sousa AS, Silva A, Santos R. Ankle anticipatory postural adjustments during gait initiation in healthy and post-stroke subjects. Clinical biomechanics. 2015 Nov;30(9):960-5.   135 [25] Liu-Ambrose T, Pang MY, Eng JJ. Executive function is independently associated with performances of balance and mobility in community-dwelling older adults after mild stroke: implications for falls prevention. Cerebrovasc Dis. 2007;23(2-3):203-10.  [26] Bugdayci D, Paker N, Dere D, Ozdemir E, Ince N. Frequency, features, and factors for falls in a group of subacute stroke patients hospitalized for rehabilitation in Istanbul. Archives of gerontology and geriatrics. 2011 May-Jun;52(3):e215-9.  [27] Hebert D, Lindsay MP, McIntyre A, Kirton A, Rumney PG, Bagg S, et al. Canadian stroke best practice recommendations: Stroke rehabilitation practice guidelines, update 2015. International journal of stroke. 2016 Jun;11(4):459-84.  [28] Jorgensen L, Engstad T, Jacobsen BK. Higher incidence of falls in long-term stroke survivors than in population controls: depressive symptoms predict falls after stroke. Stroke. 2002 Feb;33(2):542-7.  [29] Yogev-Seligmann G, Hausdorff JM, Giladi N. The role of executive function and attention in gait. Movement Disorders. 2008 Feb 15;23(3):329-42; quiz 472.  [30] Beauchet O, Annweiler C, Dubost V, Allali G, Kressig RW, Bridenbaugh S, et al. Stops walking when talking: a predictor of falls in older adults? European journal of neurology. 2009 Jul;16(7):786-95.  [31] Mitra S, Knight A, Munn A. Divergent effects of cognitive load on quiet stance and task-linked postural coordination. Joural of Experimental Psychological Human Perception Performance. 2013 Apr;39(2):323-8.  [32] Melzer I, Benjuya N, Kaplanski J. Age-related changes of postural control: effect of cognitive tasks. Gerontology. 2001 Jul-Aug;47(4):189-94.  [33] Smulders K, van Swigchem R, de Swart BJ, Geurts AC, Weerdesteyn V. Community-dwelling people with chronic stroke need disproportionate attention while walking and negotiating obstacles. Gait & posture. 2012 May;36(1):127-32.  [34] Bowen A, Wenman R, Mickelborough J, Foster J, Hill E, Tallis R. Dual-task effects of talking while walking on velocity and balance following a stroke. Age Ageing. 2001 Jul;30(4):319-23.  [35] Canning CG, Ada L, Paul SS. Is automaticity of walking regained after stroke? Disabil Rehabil. 2006 Jan 30;28(2):97-102.  [36] Yang YR, Chen YC, Lee CS, Cheng SJ, Wang RY. Dual-task-related gait changes in individuals with stroke. Gait & posture. 2007 Feb;25(2):185-90.  [37] Melzer I, Goldring M, Melzer Y, Green E, Tzedek I. Voluntary stepping behavior under single- and dual-task conditions in chronic stroke survivors: A comparison between the  136 involved and uninvolved legs. Journal of electromyography and kinesiology. 2010 Dec;20(6):1082-7.  [38] Roerdink M, Geurts AC, de Haart M, Beek PJ. On the relative contribution of the paretic leg to the control of posture after stroke. Neurorehabilitation and neural repair. 2009 Mar-Apr;23(3):267-74.  [39] Ghez C, Favilla M, Ghilardi MF, Gordon J, Bermejo R, Pullman S. Discrete and continuous planning of hand movements and isometric force trajectories. Experimental brain research. 1997 Jun;115(2):217-33.  [40] Kandel E, Schwartz J, Jessell T. Part VI: Movement. In: Butler J, Lebowitz H, eds. Principles of Neural Science. 4th ed. USA: McGraw-Hill Companies, Inc. 2000:653-872.  [41] Zimmermann M, Meulenbroek RG, de Lange FP. Motor planning is facilitated by adopting an action's goal posture: an fMRI study. Cerebral cortex. 2012 Jan;22(1):122-31.  [42] Peters S, Handy TC, Lakhani B, Boyd LA, Garland SJ. Motor and Visuospatial Attention and Motor Planning After Stroke: Considerations for the Rehabilitation of Standing Balance and Gait. Physical therapy. 2015 Oct;95(10):1423-32.  [43] Greenberg AS, Gmeindl L. Strategic control of attention to objects and locations. The Journal of neuroscience. 2008 Jan 16;28(3):564-5.  [44] Takatori K, Okada Y, Shomoto K, Ikuno K, Nagino K, Tokuhisa K. Effect of a cognitive task during obstacle crossing in hemiparetic stroke patients. Physiother Theory Pract. 2012 May;28(4):292-8.  [45] Whitney J, Close JC, Jackson SH, Lord SR. Understanding risk of falls in people with cognitive impairment living in residential care. J Am Med Dir Assoc. 2012 Jul;13(6):535-40.  [46] Robertson IH, Ridgeway V, Greenfield E, Parr A. Motor recovery after stroke depends on intact sustained attention: a 2-year follow-up study. Neuropsychology. 1997 Apr;11(2):290-5.  [47] Zehr EP, Duysens J. Regulation of arm and leg movement during human locomotion. Neuroscientist. 2004 Aug;10(4):347-61.  [48] Knaepen K, Mierau A, Tellez HF, Lefeber D, Meeusen R. Temporal and spatial organization of gait-related electrocortical potentials. Neuroscience letters. 2015 Jul 10;599:75-80.  [49] Koenraadt KL, Roelofsen EG, Duysens J, Keijsers NL. Cortical control of normal gait and precision stepping: an fNIRS study. NeuroImage. 2014 Jan 15;85 Pt 1:415-22.  137  [50] Yazawa S, Shibasaki H, Ikeda A, Terada K, Nagamine T, Honda M. Cortical mechanism underlying externally cued gait initiation studied by contingent negative variation. Electroencephalography and clinical neurophysiology. 1997 Oct;105(5):390-9.  [51] Goldberg ME, Segraves MA. Visuospatial and motor attention in the monkey. Neuropsychologia. 1987;25(1A):107-18.  [52] Daly JJ, Fang Y, Perepezko EM, Siemionow V, Yue GH. Prolonged cognitive planning time, elevated cognitive effort, and relationship to coordination and motor control following stroke. IEEE transactions on neural systems and rehabilitation engineering. 2006 Jun;14(2):168-71.  [53] Fang Y, Yue GH, Hrovat K, Sahgal V, Daly JJ. Abnormal cognitive planning and movement smoothness control for a complex shoulder/elbow motor task in stroke survivors. Journal of the neurological sciences. 2007 May 15;256(1-2):21-9.  [54] Bolton DA, Staines WR. Transient inhibition of the dorsolateral prefrontal cortex disrupts attention-based modulation of tactile stimuli at early stages of somatosensory processing. Neuropsychologia. 2011 Jun;49(7):1928-37.  [55] Handy TC, Grafton ST, Shroff NM, Ketay S, Gazzaniga MS. Graspable objects grab attention when the potential for action is recognized. Nature neuroscience. 2003 Apr;6(4):421-7.  [56] Dean PJ, Seiss E, Sterr A. Motor planning in chronic upper-limb hemiparesis: evidence from movement-related potentials. PloS one. 2012;7(10):e44558.  [57] Colebatch JG. Bereitschaftspotential and movement-related potentials: origin, significance, and application in disorders of human movement. Mov Disord. 2007 Apr 15;22(5):601-10.  [58] Kononen M, Kuikka JT, Husso-Saastamoinen M, Vanninen E, Vanninen R, Soimakallio S, et al. Increased perfusion in motor areas after constraint-induced movement therapy in chronic stroke: a single-photon emission computerized tomography study. J Cereb Blood Flow Metab. 2005 Dec;25(12):1668-74.  [59] Gallivan JP, McLean DA, Valyear KF, Pettypiece CE, Culham JC. Decoding action intentions from preparatory brain activity in human parieto-frontal networks. The Journal of neuroscience. 2011 Jun 29;31(26):9599-610.  [60] Walter WG, Cooper R, Aldridge VJ, McCallum WC, Winter AL. Contingent negative variation: An electric sign of sensorimotor association and expectancy in the human brain. Nature. 1964 Jul 25;203:380-4.   138 [61] Terada K, Ikeda A, Yazawa S, Nagamine T, Shibasaki H. Movement-related cortical potentials associated with voluntary relaxation of foot muscles. Clinical neurophysiology. 1999 Mar;110(3):397-403.  [62] Thut G, Hauert C, Viviani P, Morand S, Spinelli L, Blanke O, et al. Internally driven vs. externally cued movement selection: a study on the timing of brain activity. Brain research Cognitive brain research. 2000 Jun;9(3):261-9.  [63] Elsinger CL, Harrington DL, Rao SM. From preparation to online control: reappraisal of neural circuitry mediating internally generated and externally guided actions. NeuroImage. 2006 Jul 1;31(3):1177-87.  [64] Ball T, Schreiber A, Feige B, Wagner M, Lucking CH, Kristeva-Feige R. The role of higher-order motor areas in voluntary movement as revealed by high-resolution EEG and fMRI. NeuroImage. 1999 Dec;10(6):682-94.  [65] Cunnington R, Windischberger C, Moser E. Premovement activity of the pre-supplementary motor area and the readiness for action: studies of time-resolved event-related functional MRI. Hum Mov Sci. 2005 Oct-Dec;24(5-6):644-56.  [66] Varghese JP, Merino DM, Beyer KB, McIlroy WE. Cortical control of anticipatory postural adjustments prior to stepping. Neuroscience. 2016 Jan 28;313:99-109.  [67] Saitou K, Washimi Y, Koike Y, Takahashi A, Kaneoke Y. Slow negative cortical potential preceding the onset of postural adjustment. Electroencephalography and clinical neurophysiology. 1996 Jun;98(6):449-55.  [68] Drew T, Prentice S, Schepens B. Cortical and brainstem control of locomotion. Progress in brain research. 2004;143:251-61.  [69] McGregor R, Siegel JM. Illuminating the locus coeruleus: control of posture and arousal. Nature neuroscience. 2010 Dec;13(12):1448-9.  [70] Jahn K, Zwergal A. Imaging supraspinal locomotor control in balance disorders. Restorative neurology and neuroscience. 2010;28(1):105-14.  [71] Barrett G, Shibasaki H, Neshige R. Cortical potentials preceding voluntary movement: Evidence for three periods of preparation in man. Electroencephalography and clinical neurophysiology. 1986 4//;63(4):327-39.  [72] Shibasaki H, Hallett M. What is the Bereitschaftspotential? Clinical neurophysiology. 2006 Nov;117(11):2341-56.  [73] Rajachandrakumar R, Fraser JE, Schinkel-Ivy A, Inness EL, Biasin L, Brunton K, et al. Atypical anticipatory postural adjustments during gait initiation among individuals with sub-acute stroke. Gait & posture. 2017 Feb;52:325-31.  139  [74] Cunnington R, Windischberger C, Deecke L, Moser E. The preparation and execution of self-initiated and externally-triggered movement: a study of event-related fMRI. NeuroImage. 2002 Feb;15(2):373-85.  [75] Taniwaki T, Okayama A, Yoshiura T, Togao O, Nakamura Y, Yamasaki T, et al. Functional network of the basal ganglia and cerebellar motor loops in vivo: different activation patterns between self-initiated and externally triggered movements. NeuroImage. 2006 Jun;31(2):745-53.  [76] Cunnington R, Windischberger C, Deecke L, Moser E. The preparation and readiness for voluntary movement: a high-field event-related fMRI study of the Bereitschafts-BOLD response. NeuroImage. 2003 9//;20(1):404-12.  [77] Richter W, Andersen PM, Georgopoulos AP, Kim SG. Sequential activity in human motor areas during a delayed cued finger movement task studied by time-resolved fMRI. Neuroreport. 1997 Mar 24;8(5):1257-61.  [78] Lu MK, Arai N, Tsai CH, Ziemann U. Movement related cortical potentials of cued versus self-initiated movements: double dissociated modulation by dorsal premotor cortex versus supplementary motor area rTMS. Human brain mapping. 2012 Apr;33(4):824-39.  [79] Wiese H, Stude P, Nebel K, Osenberg D, Ischebeck W, Stolke D, et al. Recovery of movement-related potentials in the temporal course after prefrontal traumatic brain injury: a follow-up study. Clinical neurophysiology. 2004 Dec;115(12):2677-92.  [80] Jahanshahi M, Jenkins IH, Brown RG, Marsden CD, Passingham RE, Brooks DJ. Self-initiated versus externally triggered movements. I. An investigation using measurement of regional cerebral blood flow with PET and movement-related potentials in normal and Parkinson's disease subjects. Brain. 1995 Aug;118 ( Pt 4):913-33.  [81] Gallivan JP, McLean DA, Flanagan JR, Culham JC. Where one hand meets the other: limb-specific and action-dependent movement plans decoded from preparatory signals in single human frontoparietal brain areas. The Journal of neuroscience. 2013 Jan 30;33(5):1991-2008.  [82] Berchicci M, Lucci G, Pesce C, Spinelli D, Di Russo F. Prefrontal hyperactivity in older people during motor planning. NeuroImage. 2012 Sep;62(3):1750-60.  [83] Strick PL, Dum RP, Picard N. Motor areas on the medial wall of the hemisphere. Novartis Found Symp. 1998;218:64-75; discussion -80, 104-8.  [84] Dum RP, Strick PL. Motor areas in the frontal lobe of the primate. Physiol Behav. 2002 Dec;77(4-5):677-82.   140 [85] Dum RP, Strick PL. Frontal lobe inputs to the digit representations of the motor areas on the lateral surface of the hemisphere. The Journal of neuroscience. 2005 Feb 9;25(6):1375-86.  [86] Koch G, Cercignani M, Pecchioli C, Versace V, Oliveri M, Caltagirone C, et al. In vivo definition of parieto-motor connections involved in planning of grasping movements. NeuroImage. 2010 May 15;51(1):300-12.  [87] Cui H, Andersen RA. Different representations of potential and selected motor plans by distinct parietal areas. The Journal of neuroscience. 2011 Dec 7;31(49):18130-6.  [88] Cohen O, Sherman E, Zinger N, Perlmutter S, Prut Y. Getting ready to move: transmitted information in the corticospinal pathway during preparation for movement. Curr Opin Neurobiol. 2010 Dec;20(6):696-703.  [89] Boschert J, Hink RF, Deecke L. Finger movement versus toe movement-related potentials: further evidence for supplementary motor area (SMA) participation prior to voluntary action. Experimental brain research. 1983;52(1):73-80.  [90] Rektor I. Scalp-recorded Bereitschaftspotential is the result of the activity of cortical and subcortical generators--a hypothesis. Clinical neurophysiology. 2002 Dec;113(12):1998-2005.  [91] Ikeda A, Luders HO, Burgess RC, Shibasaki H. Movement-related potentials recorded from supplementary motor area and primary motor area. Role of supplementary motor area in voluntary movements. Brain. 1992 Aug;115 ( Pt 4):1017-43.  [92] Fisher BE, Boyd L, Winstein CJ. Contralateral cerebellar damage impairs imperative planning but not updating of aimed arm movements in humans. Experimental brain research. 2006 Oct;174(3):453-66.  [93] Yilmaz O, Birbaumer N, Ramos-Murguialday A. Movement related slow cortical potentials in severely paralyzed chronic stroke patients. Frontiers in human neuroscience. 2014;8:1033.  [94] Takeda K, Gomi Y, Imai I, Shimoda N, Hiwatari M, Kato H. Shift of motor activation areas during recovery from hemiparesis after cerebral infarction: a longitudinal study with near-infrared spectroscopy. Neurosci Res. 2007 Oct;59(2):136-44.  [95] Kato H, Izumiyama M, Koizumi H, Takahashi A, Itoyama Y. Near-infrared spectroscopic topography as a tool to monitor motor reorganization after hemiparetic stroke: a comparison with functional MRI. Stroke. 2002 Aug;33(8):2032-6.  [96] Clark DJ. Automaticity of walking: functional significance, mechanisms, measurement and rehabilitation strategies. Frontiers in human neuroscience. 2015;9:246.   141 [97] Saradjian AH, Tremblay L, Perrier J, Blouin J, Mouchnino L. Cortical facilitation of proprioceptive inputs related to gravitational balance constraints during step preparation. Journal of neurophysiology. 2013 Jul;110(2):397-407.  [98] Mihara M, Miyai I, Hatakenaka M, Kubota K, Sakoda S. Sustained prefrontal activation during ataxic gait: a compensatory mechanism for ataxic stroke? NeuroImage. 2007 Oct 01;37(4):1338-45.  [99] Stefan K, Wycislo M, Classen J. Modulation of associative human motor cortical plasticity by attention. Journal of neurophysiology. 2004 Jul;92(1):66-72.  [100] Posner MI, Petersen SE. The attention system of the human brain. Annu Rev Neurosci. 1990;13:25-42.  [101] Fan J, Posner M. Human attentional networks. Psychiatr Prax. 2004 Nov;31 Suppl 2:S210-4.  [102] Mesulam MM. A cortical network for directed attention and unilateral neglect. Ann Neurol. 1981 Oct;10(4):309-25.  [103] Peelen MV, Mruczek RE. Sources of spatial and feature-based attention in the human brain. The Journal of neuroscience. 2008 Sep 17;28(38):9328-9.  [104] Chica AB, Paz-Alonso PM, Valero-Cabre A, Bartolomeo P. Neural bases of the interactions between spatial attention and conscious perception. Cerebral cortex. 2013 Jun;23(6):1269-79.  [105] Kincade JM, Abrams RA, Astafiev SV, Shulman GL, Corbetta M. An event-related functional magnetic resonance imaging study of voluntary and stimulus-driven orienting of attention. The Journal of neuroscience. 2005 May 4;25(18):4593-604.  [106] Corbetta M, Shulman GL. Control of goal-directed and stimulus-driven attention in the brain. Nature reviews Neuroscience. 2002 Mar;3(3):201-15.  [107] Szczepanski SM, Pinsk MA, Douglas MM, Kastner S, Saalmann YB. Functional and structural architecture of the human dorsal frontoparietal attention network. Proceedings of the National Academy of Sciences of the United States of America. 2013 Sep 24;110(39):15806-11.  [108] Kam JW, Dao E, Farley J, Fitzpatrick K, Smallwood J, Schooler JW, et al. Slow fluctuations in attentional control of sensory cortex. Journal of cognitive neuroscience. 2011 Feb;23(2):460-70.  [109] Cohen YE, Andersen RA. A common reference frame for movement plans in the posterior parietal cortex. Nature reviews Neuroscience. 2002 Jul;3(7):553-62.   142 [110] Rushworth MF, Krams M, Passingham RE. The attentional role of the left parietal cortex: the distinct lateralization and localization of motor attention in the human brain. Journal of cognitive neuroscience. 2001 Jul 1;13(5):698-710.  [111] Rushworth MF, Johansen-Berg H, Gobel SM, Devlin JT. The left parietal and premotor cortices: motor attention and selection. NeuroImage. 2003 Nov;20 Suppl 1:S89-100.  [112] Rushworth MF, Nixon PD, Renowden S, Wade DT, Passingham RE. The left parietal cortex and motor attention. Neuropsychologia. 1997 Sep;35(9):1261-73.  [113] Brunia CH. Waiting in readiness: gating in attention and motor preparation. Psychophysiology. 1993 Jul;30(4):327-39.  [114] Nakata H, Inui K, Wasaka T, Nishihira Y, Kakigi R. Mechanisms of differences in gating effects on short-and long-latency somatosensory evoked potentials relating to movement. Brain topography. 2003 Summer;15(4):211-22.  [115] Cheron G, Dan B, Borenstein S. Sensory and motor interfering influences on somatosensory evoked potentials. Journal of clinical neurophysiology. 2000 May;17(3):280-94.  [116] Brown MJ, Staines WR. Modulatory effects of movement sequence preparation and covert spatial attention on early somatosensory input to non-primary motor areas. Experimental brain research. 2015 Feb;233(2):503-17.  [117] Popovich C, Staines WR. Acute aerobic exercise enhances attentional modulation of somatosensory event-related potentials during a tactile discrimination task. Behavioural brain research. 2015 Mar 15;281:267-75.  [118] Bancroft TD, Hockley WE, Servos P. Diffusion modeling of interference in vibrotactile working memory. Neuroreport. 2012 Mar 07;23(4):255-8.  [119] Bolton DA, Staines WR. Attention-based modulation of tactile stimuli: a comparison between prefrontal lesion patients and healthy age-matched controls. Neuropsychologia. 2014 May;57:101-11.  [120] Jueptner M, Stephan KM, Frith CD, Brooks DJ, Frackowiak RS, Passingham RE. Anatomy of motor learning. I. Frontal cortex and attention to action. Journal of neurophysiology. 1997 Mar;77(3):1313-24.  [121] Yamaguchi S, Knight RT. Gating of somatosensory input by human prefrontal cortex. Brain research. 1990 Jun 25;521(1-2):281-8.  [122] Staines WR, Graham SJ, Black SE, McIlroy WE. Task-relevant modulation of contralateral and ipsilateral primary somatosensory cortex and the role of a prefrontal-cortical sensory gating system. NeuroImage. 2002 Jan;15(1):190-9.  143  [123] Bolton DA, Staines WR. Age-related loss in attention-based modulation of tactile stimuli at early stages of somatosensory processing. Neuropsychologia. 2012 Jun;50(7):1502-13.  [124] Staines WR, Brooke JD, McIlroy WE. Task-relevant selective modulation of somatosensory afferent paths from the lower limb. Neuroreport. 2000 Jun 5;11(8):1713-9.  [125] Bolton DA, Brown KE, McIlroy WE, Staines WR. Transient inhibition of the dorsolateral prefrontal cortex disrupts somatosensory modulation during standing balance as measured by electroencephalography. Neuroreport. 2012 Apr 18;23(6):369-72.  [126] Burke SN, Barnes CA. Neural plasticity in the ageing brain. Nature reviews Neuroscience. 2006 Jan;7(1):30-40.  [127] Ohsugi H, Ohgi S, Shigemori K, Schneider EB. Differences in dual-task performance and prefrontal cortex activation between younger and older adults. BMC Neurosci. 2013;14:10.  [128] Bernard-Demanze L, Dumitrescu M, Jimeno P, Borel L, Lacour M. Age-related changes in posture control are differentially affected by postural and cognitive task complexity. Current aging science. 2009 Jul;2(2):139-49.  [129] Lacour M, Bernard-Demanze L, Dumitrescu M. Posture control, aging, and attention resources: models and posture-analysis methods. Clinical neurophysiology. 2008 Dec;38(6):411-21.  [130] Staines WR, Black SE, Graham SJ, McIlroy WE. Somatosensory gating and recovery from stroke involving the thalamus. Stroke. 2002 Nov;33(11):2642-51.  [131] Coulthard E, Rudd A, Husain M. Motor neglect associated with loss of action inhibition. Journal of neurology, neurosurgery, and psychiatry. 2008 Dec;79(12):1401-4.  [132] Bowen A, Hazelton C, Pollock A, Lincoln NB. Cognitive rehabilitation for spatial neglect following stroke. The Cochrane database of systematic reviews. 2013 Jul 01(7):CD003586.  [133] Brown LA, Sleik RJ, Winder TR. Attentional demands for static postural control after stroke. Archives of physical medicine and rehabilitation. 2002 Dec;83(12):1732-5.  [134] Smith DT, Schenk T. The Premotor theory of attention: time to move on? Neuropsychologia. 2012 May;50(6):1104-14.  [135] Schiff ND, Shah SA, Hudson AE, Nauvel T, Kalik SF, Purpura KP. Gating of attentional effort through the central thalamus. Journal of neurophysiology. 2013 Feb;109(4):1152-63.  144  [136] Mattia M, Pani P, Mirabella G, Costa S, Del Giudice P, Ferraina S. Heterogeneous attractor cell assemblies for motor planning in premotor cortex. The Journal of neuroscience. 2013 Jul 3;33(27):11155-68.  [137] Raez MB, Hussain MS, Mohd-Yasin F. Techniques of EMG signal analysis: detection, processing, classification and applications. Biological procedures online. 2006;8:11-35.  [138] Merletti R, Lo Conte LR. Surface EMG signal processing during isometric contractions. Journal of electromyography and kinesiology. 1997 Dec;7(4):241-50.  [139] Bogey R, Cerny K, Mohammed O. Repeatability of wire and surface electrodes in gait. American journal of physical medicine & rehabilitation. 2003 May;82(5):338-44.  [140] Chimera NJ, Benoit DL, Manal K. Influence of electrode type on neuromuscular activation patterns during walking in healthy subjects. Journal of electromyography and kinesiology. 2009 Dec;19(6):e494-9.  [141] Rosa MC, Marques A, Demain S, Metcalf CD, Rodrigues J. Methodologies to assess muscle co-contraction during gait in people with neurological impairment - a systematic literature review. Journal of electromyography and kinesiology. 2014 Apr;24(2):179-91.  [142] Kitatani R, Ohata K, Hashiguchi Y, Sakuma K, Yamakami N, Yamada S. Clinical factors associated with ankle muscle coactivation during gait in adults after stroke. NeuroRehabilitation. 2016 Apr 06;38(4):351-7.  [143] Tasseel-Ponche S, Yelnik AP, Bonan IV. Motor strategies of postural control after hemispheric stroke. Clinical neurophysiology. 2015 Nov;45(4-5):327-33.  [144] Dyer JO, Maupas E, de Andrade Melo S, Bourbonnais D, Nadeau S, Forget R. Changes in activation timing of knee and ankle extensors during gait are related to changes in heteronymous spinal pathways after stroke. Journal of neuroengineering and rehabilitation. 2014 Oct 24;11:148.  [145] Dyer JO, Maupas E, Melo Sde A, Bourbonnais D, Forget R. Abnormal coactivation of knee and ankle extensors is related to changes in heteronymous spinal pathways after stroke. Journal of neuroengineering and rehabilitation. 2011 Aug 02;8:41.  [146] Lamontagne A, Richards CL, Malouin F. Coactivation during gait as an adaptive behavior after stroke. Journal of electromyography and kinesiology. 2000 Dec;10(6):407-15.  [147] Kitatani R, Ohata K, Sakuma K, Aga Y, Yamakami N, Hashiguchi Y, et al. Ankle muscle coactivation during gait is decreased immediately after anterior weight shift practice in adults after stroke. Gait & posture. 2016 Mar;45:35-40.   145 [148] Gu Y, Farina D, Murguialday AR, Dremstrup K, Birbaumer N. Comparison of movement related cortical potential in healthy people and amyotrophic lateral sclerosis patients. Front Neurosci. 2013;7:65.  [149] Jiang N, Gizzi L, Mrachacz-Kersting N, Dremstrup K, Farina D. A brain-computer interface for single-trial detection of gait initiation from movement related cortical potentials. Clinical neurophysiology. 2015 Jan;126(1):154-9.  [150] Brunia CHM, Bosch DA, Speelman JD, Berg-Lenssen MMCVd, Boxtel GJMV. Chapter 28 The thalamic contribution to the emergence of the readiness potential. In: Z. Ambler SNZK, Rossini PM, eds. Supplements to Clinical Neurophysiology: Elsevier 2000:207-9.  [151] Mota IA, Lins OG. Bereitschaftspotential preceding spontaneous and voluntary eyelid blinks in normal individuals. Clinical neurophysiology. 2017 Jan;128(1):100-5.  [152] Zhang SJ, Ke Z, Li L, Yip SP, Tong KY. EEG patterns from acute to chronic stroke phases in focal cerebral ischemic rats: correlations with functional recovery. Physiological measurement. 2013 Apr;34(4):423-35.  [153] Sburlea AI, Montesano L, Cano de la Cuerda R, Alguacil Diego IM, Miangolarra-Page JC, Minguez J. Detecting intention to walk in stroke patients from pre-movement EEG correlates. Journal of neuroengineering and rehabilitation. 2015 Dec 12;12:113.  [154] Sburlea AI, Montesano L, Minguez J. Advantages of EEG phase patterns for the detection of gait intention in healthy and stroke subjects. Journal of neural engineering. 2017 Jun;14(3):036004.  [155] Tinazzi M, Fiaschi A, Mauguiere F, Manganotti P, Polo A, Bonato C, et al. Effects of voluntary contraction on tibial nerve somatosensory evoked potentials: gating of specific cortical responses. Neurology. 1998 Jun;50(6):1655-61.  [156] Hari R, Nagamine T, Nishitani N, Mikuni N, Sato T, Tarkiainen A, et al. Time-varying activation of different cytoarchitectonic areas of the human SI cortex after tibial nerve stimulation. NeuroImage. 1996 Oct;4(2):111-8.  [157] Hashimoto I, Sakuma K, Kimura T, Iguchi Y, Sekihara K. Serial activation of distinct cytoarchitectonic areas of the human S1 cortex after posterior tibial nerve stimulation. Neuroreport. 2001 Jul 03;12(9):1857-62.  [158] Valeriani M, Insola A, Restuccia D, Le Pera D, Mazzone P, Altibrandi MG, et al. Source generators of the early somatosensory evoked potentials to tibial nerve stimulation: an intracerebral and scalp recording study. Clinical neurophysiology. 2001 Nov;112(11):1999-2006.  [159] Yamada T, Matsubara M, Shiraishi G, Yeh M, Kawasaki M. Topographic analyses of somatosensory evoked potentials following stimulation of tibial, sural and lateral femoral  146 cutaneous nerves. Electroencephalography and clinical neurophysiology. 1996 Jan;100(1):33-43.  [160] Dobkin BH. Training and exercise to drive poststroke recovery. Nat Clin Pract Neuro. 2008 02//print;4(2):76-85.  [161] Weerdesteyn V, de Niet M, van Duijnhoven HJ, Geurts AC. Falls in individuals with stroke. Journal of rehabilitation research and development. 2008;45(8):1195-213.  [162] Perry J. Gait analysis: Normal and pathological function.: Thorofare: SLACK Inc. 2010.  [163] Schmitz A, Silder A, Heiderscheit B, Mahoney J, Thelen DG. Differences in lower-extremity muscular activation during walking between healthy older and young adults. Journal of Electromyography and Kinesiology. 2009 12//;19(6):1085-91.  [164] Garland SJ, Gray VL, Knorr S. Muscle activation patterns and postural control following stroke. Motor control. 2009 Oct;13(4):387-411.  [165] Den Otter AR, Geurts AC, Mulder T, Duysens J. Abnormalities in the temporal patterning of lower extremity muscle activity in hemiparetic gait. Gait & posture. 2007 Mar;25(3):342-52.  [166] Miller KJ, Hunt MA, Pollock CL, Bryant D, Garland SJ. Protocol for a randomized controlled clinical trial investigating the effectiveness of Fast muscle Activation and Stepping Training (FAST) for improving balance and mobility in sub-acute stroke. BMC neurology. 2014;14:187.  [167] Yavuzer G, Selles R, Sezer N, Sutbeyaz S, Bussmann JB, Koseoglu F, et al. Mirror therapy improves hand function in subacute stroke: a randomized controlled trial. Archives of physical medicine and rehabilitation. 2008 Mar;89(3):393-8.  [168] Bang DH, Shin WS, Choi SJ. The effects of modified constraint-induced movement therapy combined with trunk restraint in subacute stroke: a double-blinded randomized controlled trial. Clinical rehabilitation. 2015 Jun;29(6):561-9.  [169] Agrell B, Dehlin O. Mini mental state examination in geriatric stroke patients. Validity, differences between subgroups of patients, and relationships to somatic and mental variables. Aging. 2000 Dec;12(6):439-44.  [170] Tombaugh TN, McIntyre NJ. The mini-mental state examination: a comprehensive review. Journal of the American Geriatrics Society. 1992 Sep;40(9):922-35.  [171] Gray VL, Ivanova TD, Garland SJ. Effects of fast functional exercise on muscle activity after stroke. Neurorehabilitation and neural repair. 2012 Oct;26(8):968-75.   147 [172] Portney L, Watkins M. Foundations of Clinical Research: Applications to Practice. 3rd ed. Upper Saddle River, New Jersey: Julie Levin Alexander 2009.  [173] Pollock C, Eng J, Garland S. Clinical measurement of walking balance in people post stroke: a systematic review. Clinical rehabilitation. 2011 Aug;25(8):693-708.  [174] Knorr S, Brouwer B, Garland SJ. Validity of the Community Balance and Mobility Scale in community-dwelling persons after stroke. Archives of physical medicine and rehabilitation. 2010 Jun;91(6):890-6.  [175] Kadaba MP, Ramakrishnan HK, Wootten ME, Gainey J, Gorton G, Cochran GV. Repeatability of kinematic, kinetic, and electromyographic data in normal adult gait. J Orthop Res. 1989;7(6):849-60.  [176] Orthotrak 6.6 gait analysis software user manual. Santa Rosa: Motion analysis corp 2009.  [177] Rudolph KS, Axe MJ, Snyder-Mackler L. Dynamic stability after ACL injury: who can hop? Knee Surg Sports Traumatol Arthrosc. 2000;8(5):262-9.  [178] Kean CO, Birmingham TB, Garland JS, Jenkyn TR, Ivanova TD, Jones IC, et al. Moments and muscle activity after high tibial osteotomy and anterior cruciate ligament reconstruction. Medicine and science in sports and exercise. 2009 Mar;41(3):612-9.  [179] Gamst G, Meyers LS, Guarino AJ. Analysis of variance designs: A conceptual and computational approach with SPSS and SAS. New York, NY: Cambridge University Press 2008.  [180] Nakagawa S. A farewell to Bonferroni: the problems of low statistical power and publication bias. Behavioral Ecology. 2004;15(6):1044-5.  [181] Pocock SJ. Clinical trials with multiple outcomes: a statistical perspective on their design, analysis, and interpretation. Controlled clinical trials. 1997 Dec;18(6):530-45; discussion 46-9.  [182] Ramsay JW, Barrance PJ, Buchanan TS, Higginson JS. Paretic muscle atrophy and non-contractile tissue content in individual muscles of the post-stroke lower extremity. Journal of biomechanics. 2011 Nov 10;44(16):2741-6.  [183] Patterson KK, Mansfield A, Biasin L, Brunton K, Inness EL, McIlroy WE. Longitudinal changes in poststroke spatiotemporal gait asymmetry over inpatient rehabilitation. Neurorehabilitation and neural repair. 2015 Feb;29(2):153-62.  [184] Semmler JG. Motor unit synchronization and neuromuscular performance. Exercise and sport sciences reviews. 2002 Jan;30(1):8-14.   148 [185] Nelson-Wong E, Appell R, McKay M, Nawaz H, Roth J, Sigler R, et al. Increased fall risk is associated with elevated co-contraction about the ankle during static balance challenges in older adults. European journal of applied physiology. 2012 Apr;112(4):1379-89.  [186] Wright DJ, Holmes P, Di Russo F, Loporto M, Smith D. Reduced motor cortex activity during movement preparation following a period of motor skill practice. PloS one. 2012;7(12):e51886.  [187] Falvo MJ, Sirevaag EJ, Rohrbaugh JW, Earhart GM. Resistance training induces supraspinal adaptations: evidence from movement-related cortical potentials. European journal of applied physiology. 2010 Jul;109(5):923-33.  [188] Peters S, Garland SJ, Miller KJ, Cochrane CK, Ivanova TD, Hunt MA. Motor Planning for Loading During Gait in Subacute Stroke. Archives of physical medicine and rehabilitation. 2016 Apr;97(4):528-35.  [189] Gerloff C. Movement-related cortical potentials in patients with focal brain lesions. In: Jahanshahi M, Hallett M., ed. The Bereitschaftspotential. 1 ed: Springer US 2003:175-89.  [190] Honda M, Nagamine T, Fukuyama H, Yonekura Y, Kimura J, Shibasaki H. Movement-related cortical potentials and regional cerebral blood flow change in patients with stroke after motor recovery. Journal of the neurological sciences. 1997 Mar 10;146(2):117-26.  [191] Platz T, Kim IH, Pintschovius H, Winter T, Kieselbach A, Villringer K, et al. Multimodal EEG analysis in man suggests impairment-specific changes in movement-related electric brain activity after stroke. Brain : a journal of neurology. 2000 Dec;123 Pt 12:2475-90.  [192] Richard A, Van Hamme A, Drevelle X, Golmard JL, Meunier S, Welter ML. Contribution of the supplementary motor area and the cerebellum to the anticipatory postural adjustments and execution phases of human gait initiation. Neuroscience. 2017 Jul 01.  [193] Delorme A, Makeig S. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. Journal of neuroscience methods. 2004 Mar 15;134(1):9-21.  [194] Smith SM, Miller KL, Salimi-Khorshidi G, Webster M, Beckmann CF, Nichols TE, et al. Network modelling methods for FMRI. NeuroImage. 2011 Jan 15;54(2):875-91.  [195] McKeown MJ, Makeig S, Brown GG, Jung TP, Kindermann SS, Bell AJ, et al. Analysis of fMRI data by blind separation into independent spatial components. Human brain mapping. 1998;6(3):160-88.   149 [196] Drisdelle BL, Aubin S, Jolicoeur P. Dealing with ocular artifacts on lateralized ERPs in studies of visual-spatial attention and memory: ICA correction versus epoch rejection. Psychophysiology. 2017 Jan;54(1):83-99.  [197] Jung TP, Makeig S, Westerfield M, Townsend J, Courchesne E, Sejnowski TJ. Removal of eye activity artifacts from visual event-related potentials in normal and clinical subjects. Clinical neurophysiology. 2000 Oct;111(10):1745-58.  [198] Fagerland MW, Sandvik L. Performance of five two-sample location tests for skewed distributions with unequal variances. Contemporary clinical trials. 2009 Sep;30(5):490-6.  [199] Fu Q, Zhang W, Santello M. Anticipatory planning and control of grasp positions and forces for dexterous two-digit manipulation. The Journal of neuroscience. 2010 Jul 07;30(27):9117-26.  [200] Viallet F, Massion J, Massarino R, Khalil R. Coordination between posture and movement in a bimanual load lifting task: putative role of a medial frontal region including the supplementary motor area. Experimental brain research. 1992;88(3):674-84.  [201] Smith BA, Jacobs JV, Horak FB. Effects of magnitude and magnitude predictability of postural perturbations on preparatory cortical activity in older adults with and without Parkinson's disease. Experimental brain research. 2012 Oct;222(4):455-70.  [202] Mochizuki G, Boe S, Marlin A, McIlRoy WE. Perturbation-evoked cortical activity reflects both the context and consequence of postural instability. Neuroscience. 2010 Oct 13;170(2):599-609.  [203] Wheaton LA, Mizelle JC, Forrester LW, Bai O, Shibasaki H, Macko RF. How does the brain respond to unimodal and bimodal sensory demand in movement of the lower extremity? Experimental brain research. 2007 Jun;180(2):345-54.  [204] Siemionow V, Yue GH, Ranganathan VK, Liu JZ, Sahgal V. Relationship between motor activity-related cortical potential and voluntary muscle activation. Experimental brain research. 2000 Aug;133(3):303-11.  [205] Duncan F, Wu S, Mead GE. Frequency and natural history of fatigue after stroke: a systematic review of longitudinal studies. Journal of psychosomatic research. 2012 Jul;73(1):18-27.  [206] Aliakbaryhosseinabadi S, Kostic V, Pavlovic A, Radovanovic S, Nlandu Kamavuako E, Jiang N, et al. Influence of attention alternation on movement-related cortical potentials in healthy individuals and stroke patients. Clinical neurophysiology. 2017 Jan;128(1):165-75.   150 [207] Rushworth MF, Ellison A, Walsh V. Complementary localization and lateralization of orienting and motor attention. Nature neuroscience. 2001 Jun;4(6):656-61.  [208] Brown KE, Ferris JK, Amanian MA, Staines WR, Boyd LA. Task-relevancy effects on movement-related gating are modulated by continuous theta-burst stimulation of the dorsolateral prefrontal cortex and primary somatosensory cortex. Experimental brain research. 2015 Mar;233(3):927-36.  [209] Bortoletto M, Cunnington R. Motor timing and motor sequencing contribute differently to the preparation for voluntary movement. NeuroImage. 2010 Feb 15;49(4):3338-48.  [210] Baker KS, Piriyapunyaporn T, Cunnington R. Neural activity in readiness for incidental and explicitly timed actions. Neuropsychologia. 2012 Apr;50(5):715-22.  [211] Wasaka T, Kida T, Nakata H, Kakigi R. Pre-movement modulation of tibial nerve SEPs caused by a self-initiated dorsiflexion. Clinical neurophysiology. 2006 Sep;117(9):2023-9.  [212] Wasaka T, Nakata H, Kida T, Kakigi R. Gating of SEPs by contraction of the contralateral homologous muscle during the preparatory period of self-initiated plantar flexion. Brain research Cognitive brain research. 2005 May;23(2-3):354-60.  [213] Rustagi SM, Thakyal S, Vv Gopichand P. Foot index in right footed adults. Journal of clinical and diagnostic research. 2014 Jun;8(6):AM01.  [214] Chapman JP, Chapman LJ, Allen JJ. The measurement of foot preference. Neuropsychologia. 1987;25(3):579-84.  [215] Itagaki S, Yabe H, Mori Y, Ishikawa H, Takanashi Y, Niwa S. Event-related potentials in patients with adult attention-deficit/hyperactivity disorder versus schizophrenia. Psychiatry research. 2011 Sep 30;189(2):288-91.  [216] Kanovsky P, Bares M, Rektor I. The selective gating of the N30 cortical component of the somatosensory evoked potentials of median nerve is different in the mesial and dorsolateral frontal cortex: evidence from intracerebral recordings. Clinical neurophysiology. 2003 Jun;114(6):981-91.  [217] Feys H, Van Hees J, Bruyninckx F, Mercelis R, De Weerdt W. Value of somatosensory and motor evoked potentials in predicting arm recovery after a stroke. Journal of neurology, neurosurgery, and psychiatry. 2000 Mar;68(3):323-31.  [218] Sugimoto F, Katayama J. Somatosensory P2 reflects resource allocation in a game task: assessment with an irrelevant probe technique using electrical probe stimuli to shoulders. International journal of psychophysiology. 2013 Feb;87(2):200-4.   151 [219] Jenkins ND, Palmer TB, Cramer JT. Comparing the reliability of voluntary and evoked muscle actions. Clinical physiology and functional imaging. 2014 Nov;34(6):434-41.  [220] Dalmau-Pastor M, Fargues-Polo B, Jr., Casanova-Martinez D, Jr., Vega J, Golano P. Anatomy of the triceps surae: a pictorial essay. Foot and ankle clinics. 2014 Dec;19(4):603-35.  [221] Tinius TP. The Integrated Visual and Auditory Continuous Performance Test as a neuropsychological measure. Archives of clinical neuropsychology. 2003 Jul;18(5):439-54.  [222] Barker-Collo S, Feigin V, Lawes C, Senior H, Parag V. Natural history of attention deficits and their influence on functional recovery from acute stages to 6 months after stroke. Neuroepidemiology. 2010;35(4):255-62.  [223] Barker-Collo SL, Feigin VL, Lawes CM, Parag V, Senior H, Rodgers A. Reducing attention deficits after stroke using attention process training: a randomized controlled trial. Stroke. 2009 Oct;40(10):3293-8.  [224] Jehu DA, Desponts A, Paquet N, Lajoie Y. Prioritizing attention on a reaction time task improves postural control and reaction time. The International journal of neuroscience. 2015 Feb;125(2):100-6.  [225] Schneiders AG, Sullivan SJ, O'Malley KJ, Clarke SV, Knappstein SA, Taylor LJ. A valid and reliable clinical determination of footedness. Physical medicine & rehabilitation. 2010 Sep;2(9):835-41.  [226] Mangun GR, Luck SJ, Plager R, Loftus W, Hillyard SA, Handy T, et al. Monitoring the visual world: hemispheric asymmetries and subcortical processes in attention. Journal of cognitive neuroscience. 1994 Summer;6(3):267-75.  [227] Arsic S, Konstantinovic L, Eminovic F, Pavlovic D, Popovic MB, Arsic V. Correlation between the Quality of Attention and Cognitive Competence with Motor Action in Stroke Patients. BioMed research international. 2015;2015:823136.  [228] Wallmark S, Lundstrom E, Wikstrom J, Ronne-Engstrom E. Attention deficits after aneurysmal subarachnoid hemorrhage measured using the test of variables of attention. Stroke. 2015 May;46(5):1374-6.  [229] Hopf JM, Mangun GR. Shifting visual attention in space: an electrophysiological analysis using high spatial resolution mapping. Clinical neurophysiology. 2000 Jul;111(7):1241-57.  [230] Staines WR, Popovich C, Legon JK, Adams MS. Early modality-specific somatosensory cortical regions are modulated by attended visual stimuli: interaction of vision, touch and behavioral intent. Frontiers in psychology. 2014;5:351.  152  [231] Lukas S, Philipp AM, Koch I. Crossmodal attention switching: auditory dominance in temporal discrimination tasks. Acta psychologica. 2014 Nov;153:139-46.  [232] Marlinski V, Sirota MG, Beloozerova IN. Differential gating of thalamocortical signals by reticular nucleus of thalamus during locomotion. The Journal of neuroscience. 2012 Nov 07;32(45):15823-36.  [233] Chu NS. Short- and long-latency tibial somatosensory evoked potentials in cerebral lesions affecting Rolandic leg areas. Journal of neurology. 1987 Feb;234(2):74-82.  [234] Lee KB, Lim SH, Kim KH, Kim KJ, Kim YR, Chang WN, et al. Six-month functional recovery of stroke patients: a multi-time-point study. International journal of rehabilitation research. 2015 Jun;38(2):173-80.  [235] Staines WR, Brooke JD, Cheng J, Misiaszek JE, MacKay WA. Movement-induced gain modulation of somatosensory potentials and soleus H-reflexes evoked from the leg. I. Kinaesthetic task demands. Experimental brain research. 1997 Jun;115(1):147-55.  [236] Boschert J, Deecke L. Handedness, footedness and finger and toe movement-related cerebral potentials. Human neurobiology. 1986;5(4):235-43.  [237] Pendlebury ST, Mariz J, Bull L, Mehta Z, Rothwell PM. MoCA, ACE-R, and MMSE versus the National Institute of Neurological Disorders and Stroke-Canadian Stroke Network Vascular Cognitive Impairment Harmonization Standards Neuropsychological Battery after TIA and stroke. Stroke. 2012 Feb;43(2):464-9.  [238] Sonoo M, Shimpo T, Takeda K, Genba K, Nakano I, Mannen T. SEPs in two patients with localized lesions of the postcentral gyrus. Electroencephalography and clinical neurophysiology. 1991 Nov-Dec;80(6):536-46.  [239] Molenberghs P, Sale MV. Testing for spatial neglect with line bisection and target cancellation: are both tasks really unrelated? PloS one. 2011;6(7):e23017.  [240] Wasaka T, Nakata H, Kida T, Kakigi R. Changes in the centrifugal gating effect on somatosensory evoked potentials depending on the level of contractile force. Experimental brain research. 2005 Sep;166(1):118-25.  [241] Balasubramanian CK. The community balance and mobility scale alleviates the ceiling effects observed in the currently used gait and balance assessments for the community-dwelling older adults. Journal of geriatric physical therapy (2001). 2015 Apr-Jun;38(2):78-89.  [242] Park EY, Choi YI. Psychometric Properties of the Lower Extremity Subscale of the Fugl-Myer Assessment for Community-dwelling Hemiplegic Stroke Patients. Journal of physical therapy science. 2014 Nov;26(11):1775-7.  153  [243] Fields C. Metaphorical motion in mathematical reasoning: further evidence for pre-motor implementation of structure mapping in abstract domains. Cogn Process. 2013 Mar 5.  [244] Field A. Discovering statistics using SPSS, 3rd Ed. Los Angeles: Sage 2009.  [245] Allison T, McCarthy G, Wood CC, Jones SJ. Potentials evoked in human and monkey cerebral cortex by stimulation of the median nerve. A review of scalp and intracranial recordings. Brain. 1991 Dec;114 ( Pt 6):2465-503.  [246] Allison T, McCarthy G, Wood CC. The relationship between human long-latency somatosensory evoked potentials recorded from the cortical surface and from the scalp. Electroencephalography and clinical neurophysiology. 1992 Jul-Aug;84(4):301-14.  [247] Mouchnino L, Fontan A, Tandonnet C, Perrier J, Saradjian AH, Blouin J, et al. Facilitation of cutaneous inputs during the planning phase of gait initiation. Journal of neurophysiology. 2015 Jul;114(1):301-8.  [248] Calton JL, Taube JS. Where am I and how will I get there from here? A role for posterior parietal cortex in the integration of spatial information and route planning. Neurobiology of learning and memory. 2009 Feb;91(2):186-96.  [249] Lajoie K, Drew T. Lesions of area 5 of the posterior parietal cortex in the cat produce errors in the accuracy of paw placement during visually guided locomotion. Journal of neurophysiology. 2007 Mar;97(3):2339-54.  [250] Huang CM, Polk TA, Goh JO, Park DC. Both left and right posterior parietal activations contribute to compensatory processes in normal aging. Neuropsychologia. 2012 Jan;50(1):55-66.  [251] Balasubramanian CK, Bowden MG, Neptune RR, Kautz SA. Relationship between step length asymmetry and walking performance in subjects with chronic hemiparesis. Archives of physical medicine and rehabilitation. 2007 Jan;88(1):43-9.  [252] Alexander LD, Black SE, Patterson KK, Gao F, Danells CJ, McIlroy WE. Association between gait asymmetry and brain lesion location in stroke patients. Stroke. 2009 Feb;40(2):537-44.  [253] Hendrickson J, Patterson KK, Inness EL, McIlroy WE, Mansfield A. Relationship between asymmetry of quiet standing balance control and walking post-stroke. Gait & posture. 2014 Jan;39(1):177-81.  [254] McIlroy WE, Bishop DC, Staines WR, Nelson AJ, Maki BE, Brooke JD. Modulation of afferent inflow during the control of balancing tasks using the lower limbs. Brain research. 2003 Jan 24;961(1):73-80.   154 [255] Edwards JD, Meehan SK, Linsdell MA, Borich MR, Anbarani K, Jones PW, et al. Changes in thresholds for intracortical excitability in chronic stroke: more than just altered intracortical inhibition. Restorative neurology and neuroscience. 2013;31(6):693-705.  [256] Borich MR, Brown KE, Boyd LA. Motor skill learning is associated with diffusion characteristics of white matter in individuals with chronic stroke. Journal of neurologic physical therapy. 2014 Jul;38(3):151-60.  [257] Butefisch CM, Wessling M, Netz J, Seitz RJ, Homberg V. Relationship between interhemispheric inhibition and motor cortex excitability in subacute stroke patients. Neurorehabilitation and neural repair. 2008 Jan-Feb;22(1):4-21.  [258] Nagai K, Yamada M, Uemura K, Yamada Y, Ichihashi N, Tsuboyama T. Differences in muscle coactivation during postural control between healthy older and young adults. Archives of gerontology and geriatrics. 2011 Nov-Dec;53(3):338-43.  [259] Nagai K, Okita Y, Ogaya S, Tsuboyama T. Effect of higher muscle coactivation on standing postural response to perturbation in older adults. Aging clinical and experimental research. 2017 Apr;29(2):231-7.  [260] Jacobs JV, Lou JS, Kraakevik JA, Horak FB. The supplementary motor area contributes to the timing of the anticipatory postural adjustment during step initiation in participants with and without Parkinson's disease. Neuroscience. 2009 Dec 1;164(2):877-85.  [261] Wen X, Yao L, Liu Y, Ding M. Causal interactions in attention networks predict behavioral performance. The Journal of neuroscience. 2012 Jan 25;32(4):1284-92.  [262] Ptak R, Schnider A. The dorsal attention network mediates orienting toward behaviorally relevant stimuli in spatial neglect. The Journal of neuroscience. 2010 Sep 22;30(38):12557-65.  [263] Snow JC, Mattingley JB. Goal-driven selective attention in patients with right hemisphere lesions: how intact is the ipsilesional field? Brain. 2006 Jan;129(Pt 1):168-81.  [264] Ogata K, Okamoto T, Yamasaki T, Shigeto H, Tobimatsu S. Pre-movement gating of somatosensory-evoked potentials by self-initiated movements: the effects of ageing and its implication. Clinical neurophysiology. 2009 Jun;120(6):1143-8.  [265] Touge T, Takeuchi H, Sasaki I, Deguchi K, Ichihara N. Enhanced amplitude reduction of somatosensory evoked potentials by voluntary movement in the elderly. Electroencephalography and clinical neurophysiology. 1997 Mar;104(2):108-14.   155 [266] Guerreiro MJ, Murphy DR, Van Gerven PW. The role of sensory modality in age-related distraction: a critical review and a renewed view. Psychological bulletin. 2010 Nov;136(6):975-1022.  [267] Kisley MA, Davalos DB, Engleman LL, Guinther PM, Davis HP. Age-related change in neural processing of time-dependent stimulus features. Brain research Cognitive brain research. 2005 Dec;25(3):913-25.  [268] Gmehlin D, Kreisel SH, Bachmann S, Weisbrod M, Thomas C. Age effects on preattentive and early attentive auditory processing of redundant stimuli: is sensory gating affected by physiological aging? The journals of gerontology Series A, Biological sciences and medical sciences. 2011 Oct;66(10):1043-53.  [269] Tanji J, Shima K. Role for supplementary motor area cells in planning several movements ahead. Nature. 1994 Sep 29;371(6496):413-6.  [270] Todorov E, Jordan MI. Optimal feedback control as a theory of motor coordination. Nature neuroscience. 2002 Nov;5(11):1226-35.  [271] Vossel S, Weiss PH, Eschenbeck P, Fink GR. Anosognosia, neglect, extinction and lesion site predict impairment of daily living after right-hemispheric stroke. Cortex. 2013 Jul-Aug;49(7):1782-9.  [272] Gautheir L DF, Joanette Y. The Bells Test: A quantitative and qualitative test for visual neglect. Int J Clin Neuropsych. 1989;11:49-54.  [273] van Vliet P, Pelton TA, Hollands KL, Carey L, Wing AM. Neuroscience findings on coordination of reaching to grasp an object: implications for research. Neurorehabilitation and neural repair. 2013 Sep;27(7):622-35.  [274] Tyrell CM, Roos MA, Rudolph KS, Reisman DS. Influence of systematic increases in treadmill walking speed on gait kinematics after stroke. Physical therapy. 2011 Mar;91(3):392-403.  [275] Stoquart G, Detrembleur C, Lejeune TM. The reasons why stroke patients expend so much energy to walk slowly. Gait & posture. 2012 Jul;36(3):409-13.  [276] Sato D, Yamashiro K, Onishi H, Yasuhiro B, Shimoyama Y, Maruyama A. Whole-hand water flow stimulation increases motor cortical excitability: a study of transcranial magnetic stimulation and movement-related cortical potentials. Journal of neurophysiology. 2015 Feb 01;113(3):822-33.  [277] Ortu E, Ruge D, Deriu F, Rothwell JC. Theta Burst Stimulation over the human primary motor cortex modulates neural processes involved in movement preparation. Clinical neurophysiology. 2009 Jun;120(6):1195-203.   156 [278] Borich MR, Neva JL, Boyd LA. Evaluation of differences in brain neurophysiology and morphometry associated with hand function in individuals with chronic stroke. Restorative neurology and neuroscience. 2015;33(1):31-42.  [279] Winstein C, Lewthwaite R, Blanton SR, Wolf LB, Wishart L. Infusing motor learning research into neurorehabilitation practice: a historical perspective with case exemplar from the accelerated skill acquisition program. Journal of neurologic physical therapy : JNPT. 2014 Jul;38(3):190-200.  [280] Ada L, Dorsch S, Canning CG. Strengthening interventions increase strength and improve activity after stroke: a systematic review. The Australian journal of physiotherapy. 2006;52(4):241-8.  [281] Mansfield A, Wong JS, Bryce J, Knorr S, Patterson KK. Does perturbation-based balance training prevent falls? Systematic review and meta-analysis of preliminary randomized controlled trials. Physical therapy. 2015 May;95(5):700-9.  [282] Johansen-Berg H, Christensen V, Woolrich M, Matthews PM. Attention to touch modulates activity in both primary and secondary somatosensory areas. Neuroreport. 2000 Apr 27;11(6):1237-41.  [283] Hillier S, Immink M, Thewlis D. Assessing Proprioception: A Systematic Review of Possibilities. Neurorehabilitation and neural repair. 2015 Nov-Dec;29(10):933-49.  [284] Berti A, Oxbury S, Oxbury J, Affanni P, Umilta C, Orlandi L. Somatosensory extinction for meaningful objects in a patient with right hemispheric stroke. Neuropsychologia. 1999 Mar;37(3):333-43.  [285] Proto D, Pella RD, Hill BD, Gouvier WD. Assessment and rehabilitation of acquired visuospatial and proprioceptive deficits associated with visuospatial neglect. NeuroRehabilitation. 2009;24(2):145-57.  [286] Pierce SR, Buxbaum LJ. Treatments of unilateral neglect: a review. Archives of physical medicine and rehabilitation. 2002 Feb;83(2):256-68.  [287] Tan C, Tretriluxana J, Pitsch E, Runnarong N, Winstein CJ. Anticipatory planning of functional reach-to-grasp: a pilot study. Neurorehabilitation and neural repair. 2012 Oct;26(8):957-67.  [288] Kelso JA, Fuchs A, Lancaster R, Holroyd T, Cheyne D, Weinberg H. Dynamic cortical activity in the human brain reveals motor equivalence. Nature. 1998 Apr 23;392(6678):814-8.  [289] Hitier M, Besnard S, Smith PF. Vestibular pathways involved in cognition. Frontiers in integrative neuroscience. 2014;8:59.   157 [290] Stiles L, Smith PF. The vestibular - basal ganglia connection: Balancing motor control. Brain research. 2014 Dec 8.  [291] Cullen KE. The vestibular system: multimodal integration and encoding of self-motion for motor control. Trends in neurosciences. 2012 Mar;35(3):185-96.  [292] Schieber MH. Constraints on somatotopic organization in the primary motor cortex. Journal of neurophysiology. 2001 Nov;86(5):2125-43.  [293] Almeida J, Mahon BZ, Zapater-Raberov V, Dziuba A, Cabaco T, Marques JF, et al. Grasping with the eyes: the role of elongation in visual recognition of manipulable objects. Cogn Affect Behav Neurosci. 2014 Mar;14(1):319-35.  [294] Goodale MA, Milner AD. Separate visual pathways for perception and action. Trends Neurosci. 1992 Jan;15(1):20-5.  [295] Goodale MA, Milner AD, Jakobson LS, Carey DP. A neurological dissociation between perceiving objects and grasping them. Nature. 1991 Jan 10;349(6305):154-6.  [296] De Renzi E. Disorders of visual recognition. Semin Neurol. 2000;20(4):479-85.  [297] Martinaud O, Pouliquen D, Gerardin E, Loubeyre M, Hirsbein D, Hannequin D, et al. Visual agnosia and posterior cerebral artery infarcts: an anatomical-clinical study. PloS one. 2012;7(1):e30433.  [298] Goodale MA. How (and why) the visual control of action differs from visual perception. Proc Biol Sci. 2014 Jun 22;281(1785):20140337.  [299] Dawson AM, Buxbaum LJ, Duff SV. The impact of left hemisphere stroke on force control with familiar and novel objects: neuroanatomic substrates and relationship to apraxia. Brain research. 2010 Mar 4;1317:124-36.  [300] Cloutman LL. Interaction between dorsal and ventral processing streams: where, when and how? Brain Lang. 2013 Nov;127(2):251-63.  [301] Goodale MA, Kroliczak G, Westwood DA. Dual routes to action: contributions of the dorsal and ventral streams to adaptive behavior. Progress in brain research. 2005;149:269-83.  [302] Handy TC, Schaich Borg J, Turk DJ, Tipper CM, Grafton ST, Gazzaniga MS. Placing a tool in the spotlight: spatial attention modulates visuomotor responses in cortex. NeuroImage. 2005 May 15;26(1):266-76.  [303] di Pellegrino G, Rafal R, Tipper SP. Implicitly evoked actions modulate visual selection: evidence from parietal extinction. Curr Biol. 2005 Aug 23;15(16):1469-72.   158 [304] Ptak R. The frontoparietal attention network of the human brain: action, saliency, and a priority map of the environment. Neuroscientist. 2012 Oct;18(5):502-15.  [305] Greene CM, Soto D. Functional connectivity between ventral and dorsal frontoparietal networks underlies stimulus-driven and working memory-driven sources of visual distraction. NeuroImage. 2014 Jan 1;84:290-8.  [306] Ghafouri M, McIlroy WE, Maki BE. Initiation of rapid reach-and-grasp balance reactions: is a pre-formed visuospatial map used in controlling the initial arm trajectory? Experimental brain research. 2004 Apr;155(4):532-6.  [307] Chica AB, Bartolomeo P, Lupianez J. Two cognitive and neural systems for endogenous and exogenous spatial attention. Behavioural brain research. 2013 Jan 15;237:107-23.  [308] DePaul VG, Wishart LR, Richardson J, Thabane L, Ma J, Lee TD. Varied Overground Walking Training Versus Body-Weight-Supported Treadmill Training in Adults Within 1 Year of Stroke: A Randomized Controlled Trial. Neurorehabilitation and neural repair. 2014 Aug 12.  [309] Colby CL, Goldberg ME. Space and attention in parietal cortex. Annu Rev Neurosci. 1999;22:319-49.  [310] Cavada C, Goldman-Rakic PS. Posterior parietal cortex in rhesus monkey: I. Parcellation of areas based on distinctive limbic and sensory corticocortical connections. J Comp Neurol. 1989 Sep 22;287(4):393-421.  [311] Cavada C, Goldman-Rakic PS. Posterior parietal cortex in rhesus monkey: II. Evidence for segregated corticocortical networks linking sensory and limbic areas with the frontal lobe. J Comp Neurol. 1989 Sep 22;287(4):422-45.  [312] Woollacott M, Shumway-Cook A. Attention and the control of posture and gait: a review of an emerging area of research. Gait & posture. 2002 Aug;16(1):1-14.  [313] Bourlon C, Lehenaff L, Batifoulier C, Bordier A, Chatenet A, Desailly E, et al. Dual-tasking postural control in patients with right brain damage. Gait & posture. 2014 Jan;39(1):188-93.  [314] Little CE, Woollacott M. Effect of attentional interference on balance recovery in older adults. Experimental brain research. 2014 Jul;232(7):2049-60.  [315] Passingham RE. Premotor cortex: sensory cues and movement. Behavioural brain research. 1985 Nov-Dec;18(2):175-85.      159 Appendix: Motor and visuospatial attention and motor planning after stroke: Considerations for the rehabilitation of standing balance and gait   The manuscript presented below was developed over the course of my PhD comprehensive examination prior to conducting my thesis research. The information included in this appendix discusses related work to my thesis and the potential implications for neurorehabilitation; however, no thesis data are presented in this manuscript.  Appendix A   A.1 Background   After stroke, many people experience sensorimotor impairments that disrupt motor performance of balance and gait [4]. Most of the movements executed in a given day are voluntary and goal directed, requiring the capacity to plan movements according to those goals. Also, the ability to attend to certain relevant stimuli while ignoring others is needed for living in the community. Paying attention and planning movements is vital to many community dwelling adults post-stroke, as they are known to have difficulty performing another task while walking [44]. Thus, the importance of understanding the mechanisms underlying attention and planning is paramount to post-stroke recovery and integration into functional community living.  160 A.2 Operational definitions   Motor planning is defined as the integration of sensory afferent information [39] with known internal representations of body anthropometrics (e.g. leg length, joint range of motion, muscle force) [40] based on previous experience, integrated with a movement goal [41] for the purpose of generating an upcoming movement. For example, if the goal is to stand up and start walking, motor planning combines the visuospatial information about the environment, with somatosensory feedback of current leg and trunk position, and past experiences of walking – all prior to the onset of movement. The function of motor planning is to (1) prepare for an upcoming voluntary action, and (2) to maintain a state of readiness or preparedness for possible unplanned perturbations to the current movement goal.   Motor attention is defined as the ability to selectively process somatosensory input relevant to the movement goal [51]. It can prime the motor plan with relevant somatosensory feedback while disregarding irrelevant information. These types of attentional functions require integration among several brain areas and are sub-served by various networks [100,101]. In the previous example, where an individual is preparing to stand up and walk, motor attention can modulate sensory afferent information that is being processed during motor planning. The feedback related to where the leg is in space is directed towards motor planning regions, whereas auditory information may not reach the planning regions, if it is not deemed important to the motor plan.  Visuospatial attention selects the relevant visual and spatial input in the environment to be accounted for in the motor plan. This type of attention ensures the motor plan contains pertinent visuospatial information. Clinical examinations of attentional processes often consider  161 the influence of visuospatial neglect on motor output, and have been reviewed extensively in the literature [285,286]. However, with or without neglect, difficulties with motor planning and attention may be present after stroke. Consequently, this review will not directly examine neglect but rather focuses on the integration of visuospatial attention into motor planning. A.3 Clinical Relevance of Attention after Stroke  Functional motor recovery after stroke is influenced by the attentional system.  For example, one study found that the ability to attend, as measured by clinical attention tests administered two months post-stroke, significantly correlated with motor and functional outcomes two years later [46]. A prospective observational study in older adults identified those who fell as having poor attention and increased postural sway when standing with eyes closed [45]. After a stroke, individuals are more than twice as likely to fall compared to healthy controls [28], and therefore, it is possible that altered motor attention may be a factor in falls incidence post-stroke. Although the role of motor attention has been studied post-stroke using dual tasking paradigms [133], the link between lower extremity motor performance and motor attention or visuospatial attention has yet to be examined in a stroke population. A.4 Clinical Importance of Motor Planning after Stroke  Of equal importance to motor attention, is considering how motor planning influences motor performance. Most of the evidence for motor planning after stroke has been obtained from upper extremity (UE) movements. Deficits in motor performance of the UE are, in part, related to poor motor planning [56]. Taking more time to plan a movement post-stroke is also associated with altered motor performance, such as reduced precision and coordination of the UE after stroke [52,57], and the time required to plan an UE movement after stroke has been shown to  162 decrease with rehabilitation [52]. Certain physical therapy treatments, such as constraint induced movement therapy, improve motor planning of the hand [287] and increase cortical blood flow in motor planning areas [58], which may indicate active cortical reorganization with rehabilitation.  Considering the oft-cited goals of improved standing balance and gait, and reducing falls after stroke, it is essential that we begin to consider the impact of altered motor planning on lower extremity (LE) movements. To date, there is limited research to inform scientists or clinicians about the role of motor planning in the performance of functional movements made by the LE after a stroke.   In the LE, past work assumed the reciprocal action of walking was primarily driven through central pattern generators in the spinal cord [47], which may explain the lack of neurophysiological studies of motor planning and visuospatial cortical networks relating to gait.  However, there is evidence to suggest that there is cortical involvement in step initiation [49], LE motor planning [50], and motor attention. Studies of attention and motor planning in the UE can inform how motor performance in the LE may be influenced by motor attention and planning [112],[288]. Other potential subcortical regions influencing motor performance in the LE include the vestibular system and the basal ganglia, and their contributions to motor control have been reviewed elsewhere [289-291]. Though the central nervous system (CNS) affords significant flexibility and some overlapping functions, anatomical evidence suggests that certain brain regions may be linked to movement of specific body parts [292]. Since neuroanatomical specificity likely plays a role in functional motor performance, it is important to determine the differences and similarities in the brain activity for motor attention in UE and LE voluntary movement planning.  163 A.5 Objectives   This perspectives paper has three main objectives: (1) to propose a theoretical model for an anatomical and functional network of cortical brain regions that supports motor and visuospatial attention, and motor planning of voluntary goal-directed movements; (2) to discuss how stroke may impact this network and lead to altered LE function, and poor balance and gait, and (3) to suggest considerations for rehabilitation of standing balance and gait after stroke. A.6 Visuospatial Attention and its Influence on Motor Planning  Visuospatial attention selects relevant sensory information and supports the preparation of responses to this information.  The part of the environment that an individual is attending to can modulate the planning of a goal-directed movement. The Premotor Theory of Attention has been influential in understanding the relationship between motor planning and visuospatial attention [134]. It postulates that (1) motor planning and visuospatial attention use the same neural circuitry, (2) planning a movement directs visuospatial attention toward the upcoming movement goal, and (3) the ocular system is specialized to orient visuospatial attention to the movement being prepared. As a result, attention is drawn to the sensory and motor feedback related to the upcoming motor goal.  There are two different visual pathways in the cortex that are affected by visuospatial attention – the dorsal and ventral streams, which are anatomically consistent with the superior and inferior longitudinal fasciculi (Figure A.1).     164  Figure A.1: Brain regions for visuospatial and motor attention.   The dorsal and ventral streams include the occipital cortex and the posterior parietal cortex (PPC) and inferior temporal cortex (ITC), respectively.  The dorsal frontoparietal network includes the PPC and the prefrontal cortex (PFC). The ventral frontoparietal network includes the inferior frontal gyrus (IFG) and temporoparietal junction (TPJ).  Motor attention includes primary sensory cortex (S1) and PPC.  The PPC is important for the dorsal visual stream (part of visuospatial attention) and the dorsal frontoparietal network, and motor attention.  PPC = posterior parietal cortex, ITC = inferior temporal cortex, PFC = prefrontal cortex, IFG = inferior frontal gyrus, TPJ = temporoparietal junction, S1 = primary sensory cortex.  The ventral stream functions to enable visual object recognition while the dorsal stream is important for visually-guided action directed towards an object [293]. The ventral stream, which includes the occipital cortex and the inferior temporal cortex, processes visual input regarding colour, size, and shape of an object or the environment [294,295]. If a lesion is present in the ventral stream, difficulty with object recognition, or visual agnosia, can occur [296] (Table A.1).  165    Table A.1: Summary of anatomical and functional regions associated with motor and visuospatial attention and motor planning. Anatomical/ Functional Region Known Function Potential Functional Effects After a Lesion Dorsal stream Visually-guided action directed towards an object  Apraxia Ventral stream Object recognition  Visual agnosia Dorsal frontoparietal network Selects relevant sensory information and prepares responses to this information  Difficulty sifting visuospatial information to identify cues for the motor plan Ventral frontoparietal network Identifies relevant stimuli and interrupts dorsal system if an important event occurs  May have difficulty changing from one motor goal to another Motor attention (PPC) Orienting a limb in space based on attentional priorities  Limited prediction of where limb will be in space, reduced movement sequencing Supplementary motor area (SMA) Self-initiated movement planning Increased cognitive effort and slower motor planning of self-initiated movements  Premotor cortex (PMc) Externally-cued movement planning Increased cognitive effort and slower motor planning of externally-cued movements PPC = posterior parietal cortex  Visual agnosia after a stroke can include difficulty with recognizing faces, words, or common objects [297]. The dorsal stream, comprising the occipital cortex together with the posterior parietal cortex (PPC), carries information about the position and nature of goal-oriented objects or the environment as it relates to actions that can be performed [298]. Lesions of the dorsal stream may produce apraxia [299], where difficulty with imitation of an action (i.e. difficulty following gestured commands from a physical therapist) may occur. Importantly, the functional separation of the two visual streams allows for partial preservation of visuospatial processing after a lesion in one of them [300]. However, interaction between the dorsal and ventral streams  166 may be required for purposeful actions [301], such as reaching and grasping, or standing up to walk in a complex environment. Other brain regions are also involved in visuospatial attention and have overlapping functions with motor planning. For example, when a graspable “tool-like” object such as a coffee cup enters the visual field, increased brain activity is observed in the supplementary motor area (SMA) [302], inferior parietal lobule, and the premotor cortex (PMC), indicating that motor planning can be tied to visuospatial attention for an object [55,302]. To date nearly all research in this field has focused on reaching and grasping, yet it is likely that the functional demands of posture and gait when interacting with objects or the environment require similar patterns of brain network activity. However, it remains unclear whether stroke impacting visuospatial attention impacts LE movement planning. A.7 Visuospatial attention after a stroke   A stroke may alter visuospatial attention. However, one study demonstrated that despite a lesion that may directly alter visuospatial attention, cues in the environment can influence subsequent motor planning and performance [303]. That is, although a stroke in the right parietal cortex resulted in a reduction in visuospatial attention toward the left visual field, this was attenuated when cup handles were presented allowing for a left-hand grasp, suggesting that the visual system may be correctly and unconsciously extracting action-related information for grasping and then modulating attention by activating the specific motor plan the object represents [303]. This is encouraging for clinicians working with individuals who have difficulty with visuospatial attention, as object-related cues to direct attention may be useful during therapy. During rehabilitation, this may mean that LE-specific objects, like a patient’s own shoes or socks, may direct visuospatial attention to the object location. Also, gait retraining may be more effective in the patient’s own home or community, if visuospatial attention is impaired.  167 Currently, it is unknown whether visuospatial attention can also be attenuated by LE-specific environmental cues, yet it is very likely that safe and independent mobility requires intact visuospatial attentional processing.   Frontoparietal networks in the brain are also involved with visuospatial attention by generating attention to the spatial features of a planned movement [103]. Visuospatial attention can be voluntarily alerted to a location in space with attentional shifts producing activity in the dorsal frontoparietal attention network [105] (Figure A.1). This dorsal network, comprising the prefrontal cortex (PFC) together with the posterior parietal cortex (PPC), is engaged when prioritized shifts of attention in space are related to movement goals [304] (Table A.1). Similar to the dorsal stream, the dorsal frontoparietal network functions for goal-directed selection of relevant sensory information and for preparation of responses to this information. A physical therapist may engage the dorsal frontoparietal network by verbally directing the patient’s attention to the movement goal. For example, if the goal of therapy is to increase gait distance, the physical therapist may draw the patient’s attention to a destination, i.e. “walk to the kitchen.”  In contrast, the ventral frontoparietal network, inferior frontal gyrus and temporoparietal junction, identifies relevant stimuli and interrupts the dorsal system when an important or salient event occurs, such as an obstacle during gait [305] (Figure A.1). If the movement goal were to stand up and walk across the room, the dorsal frontoparietal network would identify the relevant visual and spatial sensory cues prior to movement, such as current position/angle of hips and knees relative to feet, in addition to the position of environmental obstacles needed to walk around to get to the destination. This dorsal frontoparietal network helps to prepare a response by outputting the selected relevant visuospatial information downstream to motor planning regions. The ventral frontoparietal system serves to interrupt activity in the dorsal system if an  168 unexpected sensory event occurs. This likely allows for a quick balance correction in the event of an unexpected perturbation to standing balance [306]. The cortical networks required to process incoming sensory stimuli likely utilize the dorsal and ventral frontoparietal attentional network together with the dorsal and ventral streams [307] (Figure A.1). If a stroke affects the dorsal frontoparietal attention system, it may be difficult to sift through incoming visuospatial sensory information to identify relevant cues for the motor plan [262] (Table A.1). If the ventral frontoparietal system is lesioned, unexpected perturbations to gait may not interrupt the current motor plan, potentially resulting in slower balance corrections to the perturbation.  Based on the aforementioned evidence, rehabilitation of standing balance and gait for people with visuospatial attention deficits may be more effective in an environment containing only the items needed for the task. For example, if the treatment goal is to improve skilled walking by walking on multiple surfaces (like concrete to carpet), clearing the environment of all items except for the ones needed for the task may facilitate task relevant visuospatial attention. Progression to real-world situations may include gradually adding more visual stimuli as visuospatial attention improves. If the ventral frontoparietal network is damaged, difficulty with changing from one motor goal to another may occur, possibly influencing the ability to step quickly in response to a perturbation while walking increasing the chance for a fall. In rehabilitation, the therapist may engage the ventral frontoparietal network by providing obstacles during gait in a safe and controlled gait training environment, such as body weight supported treadmill training [308]. During walking, the physical therapist might place obstacles on the treadmill at unexpected intervals and ask the patient to step over it as quickly as possible practicing interrupting gait.    169 A.8 Motor Attention and its influence on Motor Planning  Motor attention is the selection of relevant somatosensory input for a movement goal and it primes the motor plan with relevant somatosensory feedback while disregarding irrelevant information [43]. Visuospatial and motor attention are thought to be similar with attentional processes selecting relevant visuospatial or somatosensory information for the motor plan [51]. According to Cohen and Andersen (2002), a goal-directed behavior (e.g. voluntary movement to interact with objects or the environment) can be considered as “a dynamic link between a sensory stimulus and a motor act” [109]. This dynamic link requires the intermediary steps of altering visuospatial and motor attention toward salient stimuli, including the transformation of external space to internal coordinates [109], to form an appropriate motor plan. Motor attention and visuospatial attention converge with this “coordinate transformation.” The environment is encoded in the brain in several egocentric reference frames [309]. A coffee cup sitting on a table can have a variety of reference frames – relative to one’s eyes, relative to one’s arm, relative to the table – and performing these frame-of-reference calculations is important for completing a motor task successfully, calculating the difference between the current limb position and the desired limb position to complete the goal [109]. These calculations are possible as the CNS has internal representations for constants, such as UE and LE length, and knowledge that production of force at certain joint angles will put the body part in a known space (based on experience), and this can be identified prior to movement and incorporated into a motor plan [40]. This allows for prediction of where a hand or foot will be in space prior to movement onset. Impaired visuospatial attention may produce difficulty with encoding the space around us, whereas motor attentional impairments may produce difficulty with selecting the appropriate feedback for this current-to-desired limb calculation. Many of these calculations are performed in the posterior  170 parietal cortex (PPC) - a part of both the dorsal frontoparietal and the motor attention networks [107] (Figure A.1). Sensory signals converge on the PPC from many different sensory modalities including the primary sensory cortex (S1), where the combined signals allow for altering the sensory gain (up or down) depending on the attentional priorities given to the sensory signals [108], allowing the PPC to encode these variables in the output to motor planning regions [109] (Figure A.1). For example, if the movement goal is to stand on a moving bus, sensory signals from the visual, vestibular, and somatosensory systems converge on the PPC where sensory information relevant to the goal of standing on a moving surface may take priority over somatosensation from the arm/hand. The PPC sends the prioritized sensory information to motor planning areas.    More broadly, the parietal cortex is known to be involved with motor attention as evidenced by: (1) neuroimaging studies of healthy adults demonstrating parietal activity during movement preparation [110], and (2) lesions in the left parietal lobe affect the ability to disengage attention from one planned movement to another [111,112]. The supramarginal gyrus (SMG), part of the parietal cortex, functions for orienting a limb in space and is connected with S1 [310] and PMc/SMA cortices [311]. Additionally, motor attention activity is present in the left SMG and the anterior intraparietal sulcus – both left parietal regions, even with left-hand responses [110]. The dominant role for the left parietal cortex, specifically the SMG, for disengaging motor attention can explain why some individuals after stroke with left but not right parietal lesions find movement sequencing difficult even with the ipsilesional hand. Adults with lesions in the left hemisphere have difficulty disengaging motor attention from a planned movement to another suggesting the left parietal cortex has a role in motor attention [112]. In summary, the role of the parietal cortex, in particular the left hemisphere, for motor attention  171 during motor planning is for orienting a limb in space, for altering planned movements, and for movement sequencing. If the PPC is damaged after a stroke, failing to perform the appropriate coordinate transformations, and movement sequences will likely influence the motor plan and subsequent motor performance (Table A.1). As a result, reduced visuospatial and motor attention could increase the incidence of falls. A.9 Dual tasking as a means to assess motor attention   Studies of motor attention in the LE conventionally examine dual tasking, the most common being a cognitive task during standing or walking, though dual motor tasks have also been examined. The types of combined tasks, whether directing attention to a motor or cognitive task, likely produce differing cortical demands.  The use of dual-tasking paradigms to study attention is based on the following assumptions: (1) that the capacity for information processing is restricted, (2) that each task performed requires a finite capacity for information processing, and (3) if the two tasks are performed together requiring more than the total capacity, performance on one or both tasks decreases [312]. However, if the performance of both tasks decreases, the exact attentional cost is difficult to determine [312]. When healthy young adults are standing still on a force platform during a dual cognitive/motor task, they demonstrate increased body sway relative to single task performance, which suggests prioritization of the attentionally-demanding cognitive task over the motor attention for postural control as evidenced by the decreased performance on the motor task [31]. In contrast, when older adults perform a similar dual motor/cognitive task, they tend to prioritize postural control of the motor task with compensatory motor strategies [313]. Maintaining postural control through compensation during a dual task may indicate that older adults who use this strategy have reduced adaptability to  172 perturbations of standing balance [314]. A possible outcome of this type of compensatory pattern may be the need to resort to a stepping strategy to avoid a fall if perturbed [312].    Beyond standing, motor attention has been studied during dual tasks including stepping and gait. Altered performance during dual tasking, such as the inability to talk while walking, is significantly associated with increased fall risk among older adults [30]. Safe ambulation requires attention, even in healthy adults, and dual task costs generally increase with pathology [29]. The risk of falls may be exacerbated by basic motor impairments, a decline in the ability to divide attention to perform a dual task, as well as altered executive function [29]. Following a stroke, altered gait parameters (such as decreased stride length and gait speed) during dual tasking suggest that motor attention for gait remains a high priority after stroke [34-36]. Additionally, individuals after stroke demonstrate diminished cognitive function while performing a dual cognitive/motor task, which may indicate that common daily tasks such as obstacle crossing while walking require disproportionate attention and prioritization of the motor task over the cognitive task [33]. In addition to gait and stepping performance, altered motor attention during dual task performance after stroke was seen when participants were instructed to stand as symmetrically as possible while force platforms assessed the contribution of each LE to weight-bearing symmetry [38]. During the cognitive task, weight-bearing asymmetry increased suggesting that symmetric weight bearing is attention demanding.   During rehabilitation, a physical therapist may improve motor attention by using a dual task training paradigm increasing postural task demands as motor performance improves, perhaps starting in a symmetrical weight bearing position. Therapy after a lesion in the left parietal cortex may require the therapist to specifically train lower extremity movement sequences.  For example, if the patient requires re-training of transferring from the bed to a chair,  173 rehabilitation may be more effective if the pattern of movements remains the same even if the bed or chair surfaces or heights change.  Motor attention selects relevant somatosensory information required for the motor goal, and the motor plan accounts for this factor. A.10 Motor Planning After Stroke   After stroke, brain activity in regions associated with motor planning is often altered.  Motor planning is known to involve activity in the SMA, PMc and subcortical structures such as the basal ganglia and cerebellum [62]. Motor planning deficits identified for UE movements after stroke include altered regional brain activity, such as the loss of ipsilesional activity during motor planning of paretic UE movements [57]. In other studies, non-paretic hand flexion produced activity in the contralateral primary motor cortex (M1) similar to healthy adults, whereas paretic hand flexion activated bilateral S1 [94], in addition to increased SMA activity [95]. This suggests that more cortical resources are demanded for motor planning of tasks in the paretic hand. Using EEG analysis, longer planning duration was associated with increased time taken to plan a movement [53], another reflection of a higher cognitive demand during planning [53,79]. However, it is unknown whether slower motor planning is a beneficial or harmful compensatory mechanism. In typical day-to-day gait and balancing tasks, taking more time to plan a movement may be a positive protective feature of the post-stroke motor planning network to prevent errors. After a stroke, increased time in planning a route that navigates around obstacles is likely to be safer than being inadequately prepared. Alternatively, an individual may experience harmful effects from a prolonged planning duration such as slow adjustments for obstacles during gait, which may result in a fall. Irrespective of the consequence, compensation through altered network activity likely occurs to allow for motor planning after a stroke.   174 A.11 Type of movement cue influences activity of cortical regions  Motor planning regions produce different activity depending on the type of movement cue given. The SMA is considered essential for movements produced without external stimuli (or self-initiated) and the PMc is engaged in selecting movements based on external stimuli (externally-cued), such as a “go” cue [315]. During an externally-cued movement, a person responds to a signal without prior knowledge of the precise timing of signal presentation. Using functional magnetic resonance imaging (fMRI) during hand movements in healthy adults, self-initiated movement produced higher basal ganglia activity together with an earlier response in the SMA, while both self-initiated and externally-cued movements had similar activity in the contralateral M1 just before movement onset [74]. This is supported by other fMRI studies reporting that brain activity in the SMA, sensorimotor cortices, and deeper brain structures reflect the demands of self-initiated movement preparation not present in externally-cued conditions [75-77,79]. Because these studies suggest that motor planning requires a coordinated network of brain activity in multiple regions, it is conceivable that motor planning may be negatively influenced by stroke.    Rehabilitation of motor planning for the LE after a stroke should consider the type of cueing provided, as well as the amount of cortical effort required. Although the majority of goal-directed movements are self-initiated, if a lesion is present in the SMA, therapy may be more successful if external cues are provided (Table A.1). Rehabilitation can use external cues with the therapist indicating when and how to move, essentially providing a “go” cue. On the other hand, if the SMA is spared, self-initiated movement may take advantage of the increased amount of brain activity and increased subcortical activity observed experimentally. Tailoring  175 rehabilitation to the lesion location found on admission computed tomography (CT) or magnetic resonance imaging (MRI) may be useful in the provision of therapy (Table A.1).   A.12 Proposed theoretical model for brain regions supporting motor and visuospatial attention, and motor planning of voluntary goal-directed movements  A proposed network including anatomical and functional regions known to be involved with visuospatial and motor attention, and how these attentional processes influence motor planning is demonstrated in Figure A.2.     Figure A.2: Proposed theoretical model.   176 The prefrontal cortex (PFC) selects a movement-related goal while sensory signals from the dorsal stream and primary sensory cortex (S1) converge on the posterior parietal cortex (PPC).  The PPC narrows (or converges) the attentionally-selected signals pertaining to the goal, based on visuospatial and motor attentional priorities.  The supplementary motor area (SMA) and premotor cortex (PMc) areas receive this information and generate a motor plan based on the attention-filtered sensory signals and movement goals. The PPC is important for the dorsal visual stream (part of visuospatial attention), the dorsal frontoparietal network, and motor attention. PFC = prefrontal cortex, S1 = primary sensory cortex, PPC = posterior parietal cortex, SMA = supplementary motor area, PMc = premotor cortex.  The PFC selects a movement-related goal while sensory signals from the dorsal visual stream and S1 converge on the PPC. While the ventral stream identifies what an object is, the dorsal stream provides location and spatial orientation of those objects; therefore, the dorsal stream is considered to be integral to the model as an object can be determined to be an obstacle to the movement-related goal without identifying what the object is, though in extreme circumstances, such as navigating through a dangerous environment, both streams are required for safety. The PPC narrows (or converges) the attentionally-selected signals pertaining to the goal, based on visuospatial and motor attentional priorities. The SMA and PMc areas receive this information and generate a motor plan based on the attention-filtered sensory signals and movement goals.   Stroke may impact this functional network both locally and distally by interfering with input and/or output of the anatomical or functional regions involved. Functional performance can be impacted as a result. This model requires testing comparing healthy controls to individuals who have sustained a stroke, and with paretic and nonparetic UE and LE motor planning, as it is the first rehabilitation focused theoretical model integrating motor and visuospatial attention with motor planning. It is unknown whether lesions affecting visuospatial or motor attention produce compensatory activity, or which regions might participate in compensation if it occurs. It is also unknown whether the motor planning functions of the SMA/PMc could be interchangeable if a lesion was present in one motor planning region.  177 A.13 Conclusions  Though there is limited research directly examining motor and visuospatial attention and motor planning in the LE, strokes in different brain regions could alter visuospatial and motor attention, and motor planning, which would impact motor performance during gait and standing. The likely brain networks important for LE motor and visuospatial attention and motor planning involve the dorsal and ventral visual streams for visuospatial processing, the dorsal and ventral frontoparietal network for visuospatial and motor attention, as well as the premotor regions (SMA/PMc) for motor planning. Future studies are needed to link brain activity involved in attention and motor planning with clinical measures of motor performance in the LE. Additionally, lesion location may influence the functions of visuospatial and motor attention and planning such that health care professionals involved in rehabilitation after stroke may benefit from detailed lesion descriptions based on computed tomography (CT) or magnetic resonance imaging (MRI). It is still unknown the extent to which motor planning and attentional deficits limit the remediation of motor function of balance and gait after a stroke.    This perspectives paper highlights the need of future research aimed at determining how motor and visuospatial attention and motor planning interact to produce voluntary, goal-directed LE movements such as gait, after a stroke. This type of research would allow health care professionals involved in post-stroke care to better understand how these different types of impairments affect motor function and to better tailor therapeutic interventions. Future research should attempt to identify which current and novel rehabilitation treatments improve planning and attention for standing balance and gait using the model developed in this perspectives paper. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0360746/manifest

Comment

Related Items