UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Exploration and identification of neural correlates in healthy young adults during a graded cognitive,… Porter, Shaun 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_2017_may_porter_shaun.pdf [ 4.97MB ]
Metadata
JSON: 24-1.0342970.json
JSON-LD: 24-1.0342970-ld.json
RDF/XML (Pretty): 24-1.0342970-rdf.xml
RDF/JSON: 24-1.0342970-rdf.json
Turtle: 24-1.0342970-turtle.txt
N-Triples: 24-1.0342970-rdf-ntriples.txt
Original Record: 24-1.0342970-source.json
Full Text
24-1.0342970-fulltext.txt
Citation
24-1.0342970.ris

Full Text

		Exploration and Identification of Neural Correlates in Healthy Young Adults During a Graded Cognitive, Physical, and Combined Task: An EEG Study  by Shaun Porter BSc. With Specialization in Human Kinetics, University of Ottawa, 2014  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF   MASTER OF SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES  (Rehabilitation Sciences)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  February 2017  © Shaun Porter, 2017     	ii	Abstract Returning to play following a sports related concussion remains a controversial process due to the emphasis placed on subjective symptom reporting. The development of an objective measure capable of assessing cortical recovery remains elusive, however EEG has shown promise with its ability to record during exercise. The objective of this pilot study was to examine the association between EEG metrics and behavioural changes in healthy young adults.  The study involved 13 participants who performed a novel graded working memory task, a graded exercise session and a task combining the two together while EEG was recorded over 3 separate sessions. The tasks consisted of 5 levels of increasing difficulty and each participant performed the tasks in a randomized order. Participant heart rate, perceived exertion and accuracy were recorded between levels and tasks. EEG analysis applied power spectrum analysis and graph theoretical analysis to identify cortical activity and cortical networks changes.  When graded exercise and cognition were combined, there was a significant change in behaviour and neural activity compared to when each task was completed individually. The combined task led to significant changes in brain and behavior as seen in EEG activation pattern, power output and frontal functional connectivity measures.   These results suggest that following sports-related concussion individuals would require increased neural resources to complete a combined cognitive and exercise task. Following injury, these additional resources may not be available and result in a decrease in task performance. This data has the potential to be used in addition to existing concussion recovery tests in assuring full recovery prior to the return to play.       	iii	Preface  The present thesis contains results from a study completed by the candidate Shaun Porter, under the supervision of Dr. Naznin Virji-Babul. Experimental design and conception was a joint effort between Dr. Virji-Babul and Shaun Porter. The Anti-Saccade and Serial Addition Task, developed by Dr. Noah Silverberg and colleagues, was used in this study with full permission. The candidate was responsible for all data acquisition and analysis, data interpretation, and documentation.   The study and all associated methods were approved by the University of British Columbia (UBC)’s Clinical Research Ethics Board (H15-0214). The author would like to thank Saurabh Garg MSBME and Arnold Young B.Eng. for their tremendous help throughout the data analysis process.         	iv	Table Of Contents Abstract ......................................................................................................................................................... ii Preface .......................................................................................................................................................... iii Table Of Contents ........................................................................................................................................ iv List Of Tables .............................................................................................................................................. vii List Of Figures ........................................................................................................................................... viii Acknowledgments ....................................................................................................................................... xii Chapter 1: Overview of Sport-Related Concussion and Recovery ............................................................... 1 1.1 Sports-Related Concussions ............................................................................................................... 1 1.2 Behavioural And Cognitive Impact Of Concussions .......................................................................... 2 1.3 Neurometabolic Impact Of SRC ......................................................................................................... 3 1.4 Brain Structure And Function Changes Following SRC .................................................................... 5 1.5 Summary Of Changes In Brain And Behaviour Following Concussion. ........................................... 5 1.6 SRC Management ............................................................................................................................... 6 1.7 Return To Play Protocol ...................................................................................................................... 6 1.8 Effect Of Cognition And Exercise On SRC Recovery ....................................................................... 8 1.9 Purpose Of Thesis ............................................................................................................................... 9 1.10 Objective And Aims .......................................................................................................................... 9 1.11 Hypotheses ...................................................................................................................................... 10 Chapter 2: Basic Principles Of EEG And Analysis .................................................................................... 12 2.1 What Is EEG? .................................................................................................................................... 12 2.2 Traditional EEG Analysis ................................................................................................................. 13 2.3 Source Analysis ................................................................................................................................. 14 2.4 Power Spectrum Analysis ................................................................................................................. 18 2.4.1 EEG Frequency Bands ............................................................................................................... 19 2.4.2 External Factors ......................................................................................................................... 21 2.5 Functional Connectivity .................................................................................................................... 22 Chapter 3 Literature Review Of Cognitive Function And Exercise ........................................................... 27 3.1 Cognitive Function ............................................................................................................................ 27 3.1.1 EEG Measures Of Cognitive Function ...................................................................................... 28 3.2 The Brain And Exercise .................................................................................................................... 32 3.3 Combining Cognition And Exercise ................................................................................................. 35 Chapter 4: Methods ..................................................................................................................................... 38     	v	4.1 Participant Information ..................................................................................................................... 38 4.2 Anti-Saccade Serial Addition Task (ASAT) ..................................................................................... 38 4.3 EEG Data Collection ......................................................................................................................... 40 4.4 EEG Data Analysis ........................................................................................................................... 40 4.4.1 Power Spectrum Analysis .......................................................................................................... 41 4.4.2 Brain Connectivity Network Modeling ...................................................................................... 42 4.4.3 Graph Theoretical Analysis ....................................................................................................... 42 4.5 Experimental Protocol ....................................................................................................................... 43 4.5.1 Cognitive Task: ASAT ............................................................................................................... 44 4.5.2 Exercise ...................................................................................................................................... 45 4.5.3 ASAT-Exercise Dual Task ......................................................................................................... 46 4.6 Outcome Measures ............................................................................................................................ 47 4.6.1 Rated Perceived Exertion ........................................................................................................... 47 4.6.2 ASAT Accuracy ......................................................................................................................... 47 4.6.3 Power Spectrum Analysis .......................................................................................................... 47 4.6.4 Functional Connectivity ............................................................................................................. 47 4.7 Statistical Analysis ............................................................................................................................ 48 4.7.1 Experimental Outcome Measures .............................................................................................. 48 Chapter 5: Results ....................................................................................................................................... 50 5.1 Participant Characteristics ................................................................................................................. 50 5.2 Behavioural Results .......................................................................................................................... 51 5.2.1 Heart Rate And RPE .................................................................................................................. 51 5.2.2 Accuracy Of Anti-Saccade Serial Addition Task. ..................................................................... 52 5.3 EEG Results ...................................................................................................................................... 53 5.3.1 ERP ............................................................................................................................................ 53 5.3.2 Power Spectrum Analysis .......................................................................................................... 56 5.3.3 Functional Connectivity ............................................................................................................. 61 Chapter 6: Discussion ................................................................................................................................. 65 6.1 Limitations ........................................................................................................................................ 68 6.2 Applications For Sports Related Concussion .................................................................................... 69 Chapter 7: Conclusion ................................................................................................................................. 70 References ................................................................................................................................................... 71 Appendices .................................................................................................................................................. 78     	vi	Appendix A: Godin Leisure-Time Exercise Questionnaire .................................................................... 78 Appendix B: Study Recruitment Flyer .................................................................................................... 79 Appendix C: Study Consent Form .......................................................................................................... 80 Appendix D: Recent Physical Activity Questionnaire (RPAQ) ............................................................. 86 Appendix E: Borg Scale .......................................................................................................................... 94       	vii	List Of Tables Table 1 – 1 An example of the graduated return to play protocol for concussion recovery. …... 7  Table 4 - 1 Block progressions for the Anti-Saccade Serial Addition Task and Exercise Intensities .................................................................................................................. 44  Table 5 – 1 Participant Characteristics including Age, Gender, Godin Total Leisure Activity score and RPAQ scores……..……………………………………………………... 50     	viii	List Of Figures Figure 2 – 1 Example of an ERP and the how they are created……...…………………….….   13 Figure 2 – 2 Example of the effect orientation has on a single dipole positioned at electrode position Cz……………...………………………………………………………… 15 Figure 2 – 3 Virtual electrode placement in 10-10 Average montage…………...…………….. 17 Figure 2 – 4 Location of Sources in the Virtual Source Montage, Includes Transverse, Coronal and Sagittal 2D views and 3D views for better visualization of region distribution……………………………………………………………………...... 18 Figure 2 – 5 Representation of process to create structural and functional connectivity networks using graph theoretical analysis…...……………..……………………………….. 24  Figure 2 – 6 Graphical representation of network measures…………….…………………….  25 Figure 4 – 1 Example of a single ASAT task trial……………………………………………..  46 Figure 5 – 1 HR and RPE for the Exercise and Combined Cog-Ex Tasks…………………….. 51 Figure 5 – 2 Working memory task accuracy ………………………………...……………..... 52 Figure 5 – 3 Visualization of ERP signal over the entire scalp……………………………...... 54 Figure 5 – 4 Group Average ERP during rest and all difficulty blocks……………………..... 56 Figure 5 – 5 Absolute Power of the EEG signal for each frequency band by task....……...…. 57 Figure 5 – 6 Absolute Power of the EEG signal for each frequency band by Region………... 58 Figure 5 – 7 Absolute Power of the EEG signal throughout the blocks……………………… 60 Figure 5 – 8 Changes in graph theoretical measures between the two cognitive tasks………. 61 Figure 5 – 9 Changes in Theta Clustering Coefficient through levels……………………….. 62 Figure 5 – 10 Theta Degree Region X Task Interactions……………………………………. 63 Figure 5 – 11 Beta Betweeness Region X Task Interaction…………………………………. 64    	ix	List Of Abbreviations ACC: Anterior Cingulate Cortex ANCOVA: Analysis of Covariance ANOVA: Analysis of Variance ASAT: Anti-Saccade Serial Addition Task ATP: Adenosine Triphosphate BDNF: Brain Derived Neurotrophic Factor BESA: Brain Electrical Signal Analysis BOLD: Brain Oxygen Level Dependent CBH: Centre for Brain Health CSF: Cerebral Spinal Fluid CT: Computerized Tomography Cz: Central Electrode DLPFC: Dorsal Lateral Prefrontal Cortex DTI: Diffuse Tensor Imaging EEG: Electroencephalography EF: Executive Function ERP: Evoked Response FEF: Frontal Eye Fields FFT: Fast Fourier Transform FM: Frontal Midline fMRI: Functional Magnetic Resonance Imaging HR: Heart Rate     	x	HRR: Heart Rate Reserve Hz: Hertz ICA: Independent Component Analysis IFG-1: Insulin Growth Factor – 1 LFP: Local Field Potential LMM: Linear Mixed Model MEG: Magnetoencephalography MFG: Medial Frontal Gyrus MRI: Magnetic Resonance Imaging MS: Multiple Sclerosis mTBI: mild Traumatic Brain Injury PASAT: Paced Auditory Serial Addition Task PCFDR: PC False Discovery Rate PCS: Post Concussion Syndrome PET: Positron Emission Tomography PVSAT: Paced Visual Serial Addition Task RPAQ: Recent Physical Activity Questionnaire RPM: Rotation per Minute RTP: Return to Play SEF: Supplementary Eye Fields SRC: Sports-Related Concussion TBI: Traumatic Brain Injury VLPFC: Ventro-lateral Prefrontal Cortex     	xi	W: Watts WCST: Wisconsin Card Sorting Task WM: Working Memory     	xii	Acknowledgments   At the completion of this degree, I would like to thank the following for their tremendous amount of support, without which this thesis would not be possible: I would like to thank my supervisor, Dr. Naznin Virji-Babul, for the patience, guidance, encouragement, advice and opportunity that she has provided during my time as her student. I have been extremely lucky to have a supervisor that pushed me to go beyond what I though I was capable of and accomplish so much in two years. Thank you for always having an open door and your prompt responses to my emails regardless of the time or day.  I would also like to extend my gratitude to my committee members Dr. Noah Silverberg, Dr. Will Panenka and Dr. Teresa Liu-Ambrose for taking their time to provide their expertise and guidance during my time here at UBC.  A big thank you to all my past and present members of the Perception-Action Lab at the Djavad Mowafaghian Centre for Brain Health – Courtney Hilderman, Jenna Forrester, Naama Rotem-Kohavi, Najah Alhajri, Saurabh Garg, Arnold Young, and Vrinda Munjal. Your help whether it was with data collection or analysis, helping me talk through my results or discussion or just blowing off stress has been immeasurable to the successful completion of this thesis. Your expertise and friendships are invaluable and I will cherish our time together forever.  To my parents, thank you for always believing in me and for instilling in me the importance of hard work and perseverance. Thank you for taking my many phone calls and for listening to my constant ramblings about what was going wrong today and for trying to understand my topic no matter how poorly I was able to explain it.  To my amazing fiancée Kira, who constantly amazes me with her patience and words of encouragement, thank you for your support, your words of wisdom during my moments of self-    	xiii	doubt and mental road blocks, and for sticking by my side throughout the crazy journey that has been grad school. None of this would have been possible without you. I love you.  Finally, I would like to thank the Natural Sciences and Engineering Research Council of Canada, for providing the funding which allowed me to undertake this research.         	1	Chapter 1: Overview of Sport-Related Concussion and Recovery 1.1 Sports-Related Concussions Sports-related concussions (SRC) are prevalent in all age groups and sports. Classified as a mild form of traumatic brain injury (TBI), sport related concussions (SRC) are caused by linear and/or rotational forces exerted on the brain and are defined as temporarily disturbed brain function resulting from a traumatic force [1]. TBI is a serious condition that ranges in severity, from mild to severe and is classified based on the length of time spent unconscious, post-injury amnesia, and the Glasgow Coma Scale in addition to the information derived from neuroimaging [2]. Concussion and mild TBI are both viewed as being on the mild side of the TBI spectrum, and are often presented as one and the same.  However, there are significant differences in the classification of these injuries. Mild TBI is classified based on a loss of consciousness for no more than 30 minutes or amnesia, and a Glasgow Coma Score of 13 to 15 (Ref: American Congress of Rehabilitation Medicine and the Centers for Disease Control). A concussion meanwhile, is classified as an alteration of mental status that does not necessarily involve loss of consciousness (American Academy of Neurology). Clinical diagnosis of SRC is based on a combination of subjective symptom reporting, balance testing, neuropsychological testing and clinical exam. Concussion often presents with a diverse number of signs and symptoms, the most common of which being headache and dizziness [3]. Other concussion symptoms include nausea, vomiting, balance problems, fatigue, sensitivity to light or noise, dazed, trouble concentrating, trouble paying attention, forgetfulness, confusion, irritability, increased emotions, drowsiness, increased sleep, difficulty falling asleep and difficulty staying asleep [4]. These symptoms are often grouped into physical, cognitive/emotional and sleep categories and the number and     	2	severity of the symptoms vary greatly by individual [5]. The signs and symptoms of concussion can be subtle and present immediately following the injury or in the hours and days to follow. In most cases, symptoms resolve within the first 7-10 days, however in some cases symptoms can persist for weeks to months [6, 7].  Currently a large component of SRC diagnosis and recovery management is based on subjective symptom reporting. In order to improve diagnosis and care following concussion there is a need to develop reliable objective measures for clinical application. Prichep and colleagues (2013) developed an EEG-based discriminant index, to act as an objective measure for concussion recovery and identification, which has shown to be sensitive to the presence of concussion as well as severity [7]. The index was capable of distinguishing concussed individuals even after clinical symptoms and neuropsychological testing scores had recovered. This supports the finding from other studies, which suggests that clinical symptoms and cognitive function are not directly related to brain function[8, 9]. Further research is needed to identify which neurophysiological measures are most sensitive in assessing cognitive function and the impact concussion has on these processes. 1.2 Behavioural And Cognitive Impact Of Concussions  Concussion identification and management relies on a combination of subjective symptom reporting and neurocognitive testing, as neuroimaging has shown difficulties in identifying changes following concussion. Multiple tests have been developed to assess a wide range of cognitive functions and have been suggested to aid in tracking the recovery of the athlete prior to return to play. Concussion has been found to influence a wide range of cognitive functions that include attention and concentration, processing speed, learning, working memory, executive function and verbal fluency [10]. As each concussion is unique, the cognitive domain     	3	impacted by the injury can vary greatly both between individuals and between concussions within the same individual. When used to track concussion recovery, neurocognitive testing has shown varying results. A study by McCrea and colleagues (2003) identified impaired cognitive processing and verbal memory in concussed athletes two days following concussion, which along with their reported symptoms had returned to normal by day 7 in the majority of the participants [6]. Fazio and colleagues (2007) found that individuals who were asymptomatic continued to have impaired cognitive function compared to healthy controls [11]. The conflicting results between symptom reporting and neurocognitive testing show the need for more research and the potential role for neuroimaging in the assessment of concussion recovery. A recent systematic review of neurocognitive testing following concussion found a large number of neurocognitive tests, however there were a limited number of studies. The authors stated concerns regarding the adoption of neurocognitive tests into clinical practice while validity and reliability of the tests remain questionable. The authors stressed the importance of being familiar with factors that can influence performance of the tests and considering them in their interpretation [12].  The authors also noted the worrisome trend for athletes to “sandbag” their baseline tests in order to decrease their chance of missing playing time if they suffer a SRC. The incorporation of neuroimaging alongside the neurocognitive testing provides an increased level of objectivity in assessing SRC recovery.  1.3 Neurometabolic Impact Of SRC  Initial concussion models used to comprehend the neurophysiological basis of the injury originated from experimental animal studies. These studies showed that following a concussive injury there is an indiscriminate release of excitatory neurotransmitter glutamate, in addition to a large depolarization of neurons through a sodium and potassium ionic shift [13]. In order to     	4	return the system to homeostasis, a large amount of ATP (adenosince triphosphate) is required and results in an increase in cerebral glucose metabolism [14]. The period of hyperglycolysis can persist for days following injury, with the duration found to be linked to the severity of the injury [15]. These changes require increased cerebral blood flow and if this need is unmet, can result long-term damage to brain cells [16]. Following the hyperacute phase of increased glucose metabolism, there is a prolonged period of metabolic depression that can last 7 – 10 days in adult rats, which prevents the brain from functioning normally and places the brain in a vulnerable state for a secondary injury [17]. In terms of cortical activity, animal studies have shown an immediate suppression of cortical activity following injury. This widespread suppression persisted between seconds and minutes, and was followed by a period of slowed activity that gradually returned to baseline levels within the hour [18].  The neurophysiological response to concussion has been similarly reported in humans, showing the increased release of glutamate and potassium, as well as the changes in glucose metabolism [19]. Using positron emission tomography (PET), changes in glucose metabolism have been recorded in vivo in patients following both TBI and concussion [20]. The neurophysiological changes in the brain post-SRC follow a similar timeframe as the symptom reporting and some neurocognitive studies, suggesting a potential link between the cellular alterations and observable symptom and cognitive results. However, how these neurophysiological changes influence the structural and functional state of the brain remains underexplore and could provide more information regarding any acute damage to the brain and the recovery process following SRC.      	5	1.4 Brain Structure And Function Changes Following SRC Structurally, studies incorporating DTI have shown anterior regions of the brain to be more vulnerable to injury [21], while functionally, fMRI studies consistently find decreased activation in frontal regions such as right medial frontal gyrus (MFG), anterior cingulate cortex (ACC) and right precentral gyrus [22]. A bilateral decrease in dorsal lateral prefrontal cortex (DLPFC) activation has also been recorded in multiple studies [23-25]. These structural and functional changes occur in regions associated with executive function and working memory performance, and correspond with significant decreases in executive function and working memory task performance.  Another neuroimaging technique, Electroencephalography (EEG) has shown early promise in identifying changes following TBI and concussion. An overall decrease in EEG power has been shown across all frequency bands following concussion [26]. This could be a potential physiological underpinning for the impaired cognitive function seen following concussion. Other changes associated with concussion include increase in hemispherical power asymmetry, decreases in hemisphere coherence [27], reduction of low frequency power and increase in higher beta power around medial frontal brain areas [28], and changes in local connectivity networks in the prefrontal cortex [29]. These changes suggest that following concussion the connections in the brain are altered, moving away from short, densely connected networks to more widespread networks. These new networks properties require increased effort and energy, potentially increasing the vulnerability of the brain to a secondary injury.  1.5 Summary Of Changes In Brain And Behaviour Following Concussion. Concussion is often viewed through the reported cognitive and physical symptoms that present following the injury and often recover within 7 – 10 days. The injury is much more     	6	complex and causes a series of cognitive, behavioural, neurometabolic changes, and alterations to the connectivity of the brain. Cognitively, SRCs most commonly cause impaired memory and verbal fluency, with a wide range of other cognitive processes often affected. The connectivity changes within the brain show the frontal brain regions to be more sensitive to injury following SRC.  Accurate understanding of the various impacts concussion and SRC have on the brain is critical for the development of effective treatment and management plans.  1.6 SRC Management Following sport-related concussions, the current guidelines for recovery include rest until acute symptoms resolve [30]. However, there is debate in the literature regarding the length of rest for optimal recovery. A recent study by Thomas and colleagues (2015) found that strict rest following concussion had no benefit and led to increased number of symptoms and delayed resolution [31]. Current recommendations are that individuals should rest for 3 days prior to a gradual return to pre-injury activities [32]. A review by Brolinson (2014) attempted to assess the evidence available for management of sports-related concussion and concluded that the evidence was so poor that they could not form a conclusion in regard to the benefit of rest, or exercise in improving recovery [33]. While optimal management of SRC is still in need of study there is reliable evidence supporting the implementation of a graduated return to play protocol.   1.7 Return To Play Protocol The return to activity and sport following a SRC has been regulated through the return-to-play (RTP) protocol, originally created by the Concussion in Sport Group [34] and widely adopted, both in the following consensus statements and in practice around the world. Safe RTP     	7	is of critical importance due to the potential risk of both short-term [35, 36] and long-term damage [37]. The graduated RTP protocol consists of six progressive stages of incremental tasks related to sport performance (Table 1 – 1). The protocol begins with no activity during the recovery stage, followed by light aerobic exercise and gradually progresses to sport specific activities. Following each stage, the athlete is assessed for concussion symptoms, if none reappear; they are permitted to move to the next stage. Upon completion of the six stages and receiving medical clearance, the athlete is deemed ready to return to play. The current RTP protocol relies heavily on subjective symptom reporting.  Table 1 – 1 Example of the graduate return to play protocol for concussion recovery Rehabilitation Stage Objective of Stage No Activity Recovery Light aerobic exercise Increase heart rate Sport-specific exercise Add movement Non-contact training drills Exercise, coordination and cognitive load Full-contact practice  Restore athlete’s confidence; coaching staff assesses functional skills Return to play   The RTP protocol relies on subjective symptom reporting, using the post-concussion scale (PCS). While the PCS has been shown to be reliable [38], the validity of using symptoms is highly questionable as concussed athletes presenting with no symptoms continued to show deficits in neurocognitive testing [11]. In addition, athletes are known to underreport symptoms in order to return to play [39], decreasing the validity the testing. Furthermore, concussion assessments often occur immediately following the incident or the stage of exercise, this is critical, as the literature shows that exercise can result in post concussion-like symptoms in healthy individuals [40-42]. There remains a need for an objective measure to reliably track     	8	concussion recovery and provide accurate assessment of brain recovery within the return to play protocol. The results from this thesis provide preliminary objective measures of healthy brain activity in healthy young adults during graded exercise and cognitive tasks. These results can be used in future studies to determine changes in individuals recovering from concussion.  1.8 Effect Of Cognition And Exercise On SRC Recovery Sports are a complex combination of physical activity and cognitive function. Therefore when testing for recovery, both exercise and cognition should be assessed. Lee and colleagues (2015) incorporated a cognitive task into a standardized exercise protocol in order to investigate if this would provoke a greater number or severity of symptoms in healthy individuals [40]. Although no increase in symptom reporting was reported with the addition of the cognitive task, the researchers noted that changes in exercise intensity and cognitive task difficulty could influence these results. When recently concussed athletes were tested following exercise, a significant difference was found in neurocognitive testing. McGrath and colleagues (2013) had athletes who were asymptomatic at rest and returned to baseline on the ImPACT (a commercially available neurocognitive test) complete an exercise session and repeat the neurocognitive testing. They found 27.7% of athletes showed a post-exertion decline in neurocognitive function that was not attribute to overall performance but specifically in memory ability [43]. The introduction and application of objective measures such as EEG is critical for the continuation to develop the understanding of the concussed brain and the recovery process. Advancements in EEG software and hardware has allowed for recording of brain activity during more strenuous exercise. This provides an exciting opportunity to identify neurophysiological underpinnings of the brain involved during exercise and cognition, with the goal of potentially     	9	discovering new biomarkers of concussion. In order to accurately characterize the concussed and recovering brain, it is imperative to first fully understand how the healthy brain is influenced by cognition and exercise. 1.9 Purpose Of Thesis When the brain is injured or in a vulnerable state such as following a SRC, the brain will react differently as it completes the tasks. Understanding the neurophysiological impact of load on the healthy brain is of critical importance as it can then be used as a baseline for pathological populations. This thesis set out to explore and identify the characteristics of the healthy brains response to multi modal loading.  1.10 Objective And Aims  The objective of the current study was to examine the association between EEG metrics and behavioural changes in healthy normal adults as a foundation for evaluating individuals with concussion.  The aims of this study are: 1. To evaluate the association between EEG power and functional connectivity metrics and performance during cognitive loading. 2. To evaluate the association between EEG power and performance during physical loading.  3. To evaluate the association between EEG power and functional connectivity metrics and performance during a combination of cognitive and physical loading.      	10	4. To compare EEG power and functional connectivity metrics between the types of loading.  1.11 Hypotheses 1. Increased cognitive load will result in a decrease in task accuracy.  2. Cognitive load will result in significantly increased power in all frequency bands.  3. A positive load dependent relationship will emerge between the working memory task and local connectivity measures (degree, clustering coefficient, betweenness) within the frontal brain regions.  4. The combined cognitive – exercise task will be associated with a significantly greater increase in frontal activity than either task individually.  In Chapter 2, I cover the basics of EEG and the traditional methods of analysis, followed by detailed explanation of the analyses incorporated in this work. This includes source analysis, power spectrum analysis, and functional connectivity through graph theoretical analysis. In Chapter 3, there is a literature review exploring the current understanding of cognitive, physical and combined loading and it’s affect on behaviour and neurophysiological measures. The methods of my study will be discussed in Chapter 4, detailing the study design, data analysis techniques used, and the statistical analyses. In order to understand how the healthy brain is affected by multi-modal loading, I present the findings of my research in Chapter 5, which utilized a repeated measure design to compare brain activation during a cognitive task, an exercise task, and a combined cognitive-exercise task. Thirteen healthy young participants (22.5 years old ± 0.65) completed the three tasks on three     	11	separate days in a randomized order. During all conditions, brain activity was collected through EEG, along with behavioural measures. This adds to the current literature on the effect cognition, exercise and a combination of the two influence the neurophysiological features of the brain and provides healthy control data for future studies to assess individuals recovering from SRC.        	12	Chapter 2: Basic Principles Of EEG And Analysis 2.1 What Is EEG? Electroencephalography (EEG) has been used to measure human brain signal for almost a century. Discovered by Hans Berger in 1924, the German psychiatrist was able to successfully measure brain activity in humans during various states including: sleep, wakefulness, and focused attention [44]. EEG is a graphic representation of voltage changes between two cortical locations plotted over time [45]. The voltage changes, or signal, are the postsynaptic potentials of the cortical neurons. Or in other terms, the EEG measures voltage changes on the skull at all electrodes [46]. The electrical potential of a single neuron is much too small to be recorded through EEG and therefore in order to be measured, a large group of neurons (i.e. 107 neurons [47]) must activate simultaneously to produce a strong enough signal. The electrical potentials created by neurons also form an electric field. When large groups of neurons activate simultaneously the sum of the activity creates a local field potential. Local field potentials (LFPs) can be open or closed depending on their orientation. Open LFPs are orientated perpendicular to the scalp while Closed LFPs run parallel, as such surface EEG is only capable of measuring Open LFPs [48].  All regions of the brain produce local field potentials, however due to the layers that surround the brain: the cerebrospinal fluid (CSF), the skull, and the scalp; the signal is greatly attenuated by the time it reaches the surface [44]. The main source of EEG signal originates from the cerebral cortex, as a limitation of EEG is its inability to measure deeper cortical structures. While the EEG has low spatial resolution, it is one of the only neuroimaging techniques with sufficient temporal resolution to record the fast dynamic changes of cortical activity. EEG can be recorded between 250 and 2000 Hz or samples per second.      	13	2.2 Traditional EEG Analysis Early studies using EEG aimed to examine the raw signal to identify changes due to a specific task or stimuli.  As the technique advanced, studies began taking advantage of averaging the signal. This led to the development of the event-related potential (ERP) technique and became the primary analysis of EEG in cognitive neuroscience. ERPs are time-locked to a specific stimulus and as such have allowed for the observation of many specific aspects of cognitive function including task preparation, stimuli identification and cognitive function [46]. The ERP waveform is composed of peaks and dips that allow for the visualization of neural processing throughout the trial (Figure 2-1). The peaks and dips of the ERP are known as components and are defined by their polarity (negative or positive), timing, scalp distribution and sensitivity to task changes [46].   Figure 2 – 1. Example of how an average ERP waveform is created.      	14	Over the years many components of the waveform have been identified and attributed to specific functions. For example, P1 and N1 are indicative of information processing in the visual cortex and perceptual analysis, respectively [49, 50]. Further along the waveform is another component known as P3, which is attributed to working memory encoding and maintenance. Advances in EEG methodology have made it possible to record brain activity from a large number of electrodes over the entire head. This had led to very dense ERP data sets that can be cumbersome and difficult to interpret. New analysis techniques have emerged to discern additional information not fully reflected within the ERP waveforms.   2.3 Source Analysis The LFPs mentioned above are in the form of dipoles, with both positive and negative charges. An important aspect of EEG is the location, strength and orientation of the dipoles as these measures can greatly influence the recorded signal. For example in Figure 2-2 a single dipole is shown near the central electrode (Cz). On the left, the dipole is orientated towards the scalp and this is where the max activity will be found. On the right, the dipole is orientated tangential to the scalp. This results in a change of the signal and almost no activity is present above the dipole.       	15	 Figure 2 – 2 Example of the effect orientation has on a single dipole positioned at electrode position Cz [48].  The influence of dipole orientation and the effect on EEG activity stresses the importance of understanding the cortical source of EEG activity, which is one of the major limitations with using EEG. Therefore extensive work has been done in attempts to develop methods capable of localizing the activity to a particular cortical source. There are two methods used to approach this issue: (1) investigate which cortical region is responsible for the recorded EEG signal, or (2) investigate how particular brain regions contribute to the recorded brain activity. The first method is addressed through source localization being applied to the signal to determine the location of the dipoles. Source localization is a signal processing technique that takes the voltage potentials at the various scalp locations and estimates the current sources inside the brain that best fit this data [51]. This requires solving the inverse problem. That is, each electrical potential     	16	measured at the scalp can be explained by activity of an infinite number of cortical configurations. This problem can only be solved by applying multiple a priori assumptions regarding the generation of the EEG signal [52]. Many source localization algorithms exist, each attempting to optimally explain the scalp activity by cortical sources.   An alternative method in understanding the cortical source of the scalp activity is through the application of source montages, a type of virtual montage. This digital EEG reconstruction calculates the topography of the signal using all the recorded electrodes. The signal is then reconstructed at each recorded electrode site as well as any ‘virtual’ electrode located on the scalp. This allows for the construction of standard EEG montages such as the reference free, 10-10 and 10-20 systems (Figure 2-3).          	17	 Figure 2 – 3 Virtual electrode placement in 10-10 Average montage.  Virtual montages are excellent at recording radial activity underneath each electrode, but have shown to be less sensitive to tangential activity and require the addition of whole-head spherical spline maps to reflect cortical source activity [53]. Once a source montage is applied the resulting traces can be viewed as large virtual electrodes that are roughly 3-4 cm in diameter placed on the cortical area it is supposed to be modeling. Source montages reconstruct approximate source waveforms, which are calculated using the generalized montage, previous knowledge on scalp topographies results from the EEG recording, and linear algebra. For a more detailed description of the creation of source montages refer to the review by Scherg and colleagues [53]. These source waveforms are an estimate of the magnitude of the activity of each region over time and allow for a simple representation of the cortical activity.  The source     	18	montage used in this thesis consisted of 15 sources created to best estimate cortical activity of the entire brain (Figure 2 – 4).   Figure 2 – 4 Location of Sources in the Virtual Source Montage. Transverse, Coronal and Sagittal 2D views and 3D views for better visualization of region distribution.   2.4 Power Spectrum Analysis In order to understand EEG signals, interest has developed in analyzing specific frequency bands, as they have been shown to represent specific neurological processes [54]. When recorded from the scalp, the EEG signal is the culmination of all the frequency bands     	19	together. In order to assess the individual frequency bands, a fast Fourier transform (FFT) is used. The algorithm works by taking the EEG data, collected in the time domain, and transforms it into the frequency domain allowing for the frequency bands to be divided for separate analysis. The FFT is an algorithm that rapidly converts a signal from the time domain into a representation in the frequency domain using the discrete Fourier transform (DFT). The DFT creates a voltage by frequency spectral graph known as a “power spectrum”, where power is equal to EEG magnitude squared. EEG power represents the distribution of signal power over frequency and has been reliably shown to relate to cortical activity. When combined with blood oxygen level-dependent (BOLD) signal in fMRI, results show a frequency dependent relationship with low frequency EEG being negatively correlated and higher frequencies being positively correlated [55, 56].   2.4.1 EEG Frequency Bands Following the FFT, the EEG signal is transformed from a single signal into multiple signals with various frequencies. The EEG signal is commonly composed of frequencies between 1 – 50 Hz. The frequencies are grouped into the following 5 bands ranging from slow (delta) to fast (gamma).  § Delta (0.5 – 3.5 Hz): Composed of the slowest frequencies, delta waves are the dominant frequency during early developmental stages [57], and sleep in adults [54]. In addition, delta is associated with learning, motivation and the brain reward system [58]. In terms of cognition, delta found to be linked to the P3 ERP component in various cognitive tasks. This connection is theorized to be related to motivation, with delta activity seeming to help motivate the brain to pay attention to the stimuli for task completion [58].     	20	§ Theta (3.5 – 7.5 Hz): Associated with working memory and inhibitory control. Theta and particularly in frontal midline is thought to have an active role in the maintenance phase of memory [58, 59], which coincides with its recorded connection to the hippocampus [60].  § Alpha (7.5 – 12.5 Hz): Appear spontaneously during wakefulness, relaxed states and mental inactivity or resting state. Most pronounced within the occipital lobe during eyes closed conditions [54], while also linked to working memory and short-term memory functions [61].  § Beta (12.5 – 30 Hz): Historically linked to motor functions, more recent research suggests beta frequency involved in maintaining the status quo within the sensorimotor system [62]. Further studies have indicated elevated beta activity during the delay phase in working memory tasks [63]. § Gamma (30 – 60 Hz): Highest frequency band shown to be involved in wide variety of actions including stimulus selection, attention, arousal, object recognition, memory formation [64, 65]. Gamma activity often found to be locked with slower frequencies indicating a potential interaction required for proper memory function [65]. Cognitive functions involve oscillations from multiple frequency bands as many brain regions must interact and communicate for successful functioning. Oscillations of different frequencies are indicative of global state changes in the brain [62]. Higher frequency oscillations are indicative of arousal and more distinct activation patterns, whereas low frequency waves are present in of low arousal and global state changes.     	21	2.4.2 External Factors The various frequency bands have well identified cognitive functions and are often used to assess cortical activity in response to a particular task. When there is an increase in a specific frequency it is interpreted to be the result of the task. It is important to note that there are several external factors that can influence cortical activity as measured through EEG. Drugs can have a large influence on EEG activity that can confound results and cause in accurate interpretation if not taken into account. Antidepressants such as selective serotonin reuptake inhibitors (SSRIs), have shown to significantly impact frontal theta power, alpha and beta frequencies [66]. Antipsychotic drugs, such as Clozapine, have been linked to increased delta and theta power in frontal brain regions as well as decreased alpha and beta. Mood stabilizers have shown to lead to increased delta and theta wave activity, decreased alpha activity and varying effects on beta depending on the medication. Stimulants, such as caffeine, are shown to acutely increase attention, alertness, and restore performance due to fatigue [67]. However, studies show caffeine to lead to a significant decrease in EEG power across the spectrum in fronto-parieto-occipital and central electrodes [68]. Other common external factors that can influence EEG activity include depression, anxiety, and attention-deficit hyperactivity disorder (ADHD). Depression has been shown to cause significant increase in absolute beta power, in addition to an overall shift to faster frequencies across the spectrum [69].  Meanwhile, anxiety is linked to an increase in global alpha power in both males and females [70]. ADHD, a neurodevelopmental disorder, has been shown to cause an increase in theta power and decrease in beta power when compared to controls [71]. These changes that result form medication, or neurological disorders stress the importance of proper screening and identification of potential factors that can influence EEG signal. This ensures that     	22	the interpretations of the resulting changes in EEG are in response to the task and not from any other cause.  2.5 Functional Connectivity  The ability of the human brain for higher cognitive function is theorized to be the result of structural and functional connections forming complex networks between widespread brain regions. These networks have been proposed to represent the physiological basis for information processing and mental representation [72-74]. A variety of methods and imaging modalities have been used to characterize the many structural and functional networks that allow for the integration and segregation required for information processing[75]. Graph theoretical analysis (or Graph theory) is an analysis technique that allows for noninvasive mapping of these structural and functional networks and their properties. It begins by modeling the brain as a series of complex networks and identifies the many topological properties of these brain networks [75]. Both structural and functional brain networks can be constructed using the following four steps provided by Bullmore and Sporns [76] (see Figure 2 - 5.): 1. Define the nodes of the network. These can be defined by electroencephalography electrodes or alternatively as anatomically defined regions (MRI, DTI).  2. Estimate a continuous measure of association between nodes. This is possible through a variety of ways, including spectral coherence, Granger causality or through conditional dependence and independence between any two regions based on all other brain regions.  3. Generate association matrix by calculating all pairwise associations between nodes and apply a threshold each element to produce a binary adjacency matrix. This matrix is composed of the number of edges between each pair of nodes. In most cases, this is binary, 1 (edge present) or 0 (no edge between nodes).      	23	4. Calculate network parameters of interest for this graphical model of brain network and proceed to compare against a series of random networks with the same number of nodes.   A network in graph theory is stated to be composed of a series of nodes and edges. The topological properties of these networks can be defined by a variety of measures. While not exhaustive, below is a list of measures commonly applied to explain structural and functional brain networks:  Node Degree  A critical measure for the network, degree represents the number of connections or edges each particular node has. Individual node degree denotes the importance of the node to the network. Degree is also a measure centrality, in that a node with high degree interacts with many other nodes in the network [74].  Clustering Coefficient Another important measure of connectivity, if a node’s nearest neighbors are also connected to each other, the graph forms a cluster. Clustering coefficient represents the local connectivity of a graph. Small-world networks, like the brain, have high clustering and small path lengths [77]. Path Length And Efficiency  Path length is the minimum number of edges needed to go from one node to another and represents the level of global integration of the network. The average shortest path of a network is the average of all shortest paths between all pairs of nodes. Global efficiency is the inverse of the average shortest path. Local efficiency of an individual node is the inverse of the average shortest path connecting to that node. Global and local efficiency measure the ability of a network to transmit information at the global and local levels [76-78].  	 24	                Figure 2 - 5. Representation of process to create structural and functional connectivity networks using graph theoretical analysis.    	25	Modularity Modularity is a measure of structural networks tendency to form modules. Modules are a group of nodes that are strongly connected to each other but not to other nodes. Modules play an important role in complex networks as they often have different functional roles within the network [76, 77].  Centrality And Hubs Centrality is a measure of how many shortest paths between other nodes pass through a particular node. High centrality reflects that nodes importance to the network. A hub is a node with high degree or centrality.  If considered in conjunction with modularity, there are two types of hubs: provincial (hubs connected to vertices in the same module) and connector hubs (hubs connected to nodes in other modules) [76, 77] (See Figure 2 - 6).             Figure 2 - 6. Graphical representation of network measures.     	26	By utilizing graph theoretical analysis of brain connectivity it is possible to get a complete and thorough description of the structural and functional networks within the brain, providing critical information for both research and clinical applications.       	27	Chapter 3 Literature Review Of Cognitive Function And Exercise 3.1 Cognitive Function  Executive function (EF) is a general term that includes task switching, planning, attention, working memory, and inhibitory control among others. The prefrontal cortex (PFC) is known to play a crucial role in all executive function processes, however the PFC is a heterogeneous neuro-anatomical region and different areas have been proposed to be responsible for separate cognitive functions [79]. The DLPFC, ventrolateral PFC (VLPFC) and ACC have all been identified as regions involved in two sub-processes of EF: response inhibition and working memory.  The ability to inhibit external distracting stimuli and focus on a specific task is a crucial component of everyday life and an important component of executive function. Common inhibitory control tasks include the Stroop task, the trail making test, and the anti-saccade task. The anti-saccade task [80] has been used extensively to examine visual attention, reflexive inhibition, and neurophysiological status [81]. This task involves looking away from a peripheral target by suppressing the automatic response to look at the target and then create a voluntary motor command to look away [82]. The task requires input from many subcortical structures, including the pontine and midbrain nuclei, as well as cortical structures such as the PFC and ACC [83, 84].  Working memory (WM) is a term that describes a cognitive function that can hold and manipulate information within the brain for a limited amount of time, for the purpose of a specific cognitive activity or task.  It also includes sustained attention and focus on particular information, while rejecting distractors [85].  WM is required for complex cognitive tasks such     	28	as comprehension, reasoning, planning, learning and mental arithmetic [86, 87]. Baddeley and Hitch [88] first defined working memory in 1974, wherein they proposed a multicomponent model that had separate verbal and visual systems, under the control of a central executive. Extensive animal and human studies have demonstrated that the prefrontal cortex (PFC) is a key area involved in working memory [86]. In particular, the DLPFC, VLPFC and ACC [89] have been associated with working memory. Their specific roles are still debated but the DLPFC and ACC are active during retention of information. ACC activity also seems to be particularly influenced by task difficulty [90, 91]. 3.1.1 EEG Measures Of Cognitive Function Neuroimaging is critical to understanding the brain regions involved in EF and its sub-processes. EEG is sensitive to changes in brain activation at rest and during cognitive tasks. As a non-intrusive imaging modality with high temporal resolution, EEG is an optimal neuroimaging modality to study EF. Findings with EEG suggest the presence of a load effect, wherein increasing cognitive load through working memory task difficulty, elicits a greater response in brain activity in frontal brain regions [55, 92-95]. However, this finding has shown to be frequency dependent as well as location specific.  Frequency of the EEG oscillations are important as they can control the timing of neuronal firing and coordinate information transfer between different brain regions [62]. This study focused on theta, alpha, beta, and gamma frequency bands as these have shown to be particularly sensitive to working memory and exercise. Theta activity has been reliably shown to have a major role in working memory [95]. In particular, theta activity has been linked to encoding and retention of information. Multiple     	29	studies incorporated a modified Sternberg task to test working memory and theta activity. These studies found with increasing WM load there was a corresponding increase in frontal midline (FM) theta activity [55, 93, 94]. Maurer and colleagues also found that the increase in FM theta was correlated with accuracy: the larger the increase in theta, the greater the decrease in accuracy [93]. This effect is not task specific as Ku and colleagues found similar relationship with FM Theta using both a visual and auditory mental addition task [92]. Through the application of intracranial EEG and source localization, the medial PFC and ACC have been identified as the cortical origin of the FM theta activity [91, 93, 96].  Activity in the alpha frequency band has traditionally been linked to steady state or idling activity with amplitude suppressed by eye opening and visual stimuli.  Recently, high resting alpha power has been linked to increased WM performance as well as successful saccadic control [84]. Alpha is suggested to play an important role in saccadic control network circuits and top-down control of suppressing external saccades [84]. Other studies show alpha power involved in working memory and mental arithmetic [92] suggesting that alpha is an active contributor in attention and consciousness. During WM tasks, alpha activity is reduced and negatively correlated with cognitive load [93, 97, 98]. The alpha frequency is often separated into low (8-10 Hz) and high (10-13 Hz) bands due to their varying sources of origin and activity. Low alpha is found over parietal, temporal and superior frontal regions and is shown to be more sensitive to load and decreases with increasing WM load [93, 98, 99]. High alpha on the other hand originates from occipital and occipital-parietal regions and is sensitive to visuospatial factors, and to a lesser extent cognitive load [85, 93]. Successful WM performance is also linked to high resting alpha power, which is thought to correlate with successful saccadic performance.      	30	Activity in beta has been linked to a variety of cognitive processes including movement related, sensory, cognitive and emotional stimuli. Studies have identified its role in sensorimotor functions [62] and been associated with sensory processing [100]. For WM performance, there have been mixed findings as some studies show increased beta activity while others show a decrease [101, 102]. Guntekin and colleagues (2013) identified the presence of a physical response to be the cause of the result variations. When no physical response is required, beta power increases during WM tasks and is sensitive to cognitive load [103].   Gamma activity has been linked to a wide range of cognitive processes including movement preparation, attention, sensorimotor integration and memory formation [62]. A critical component of WM is the maintenance phase, where the stimulus is no longer present and the information must be temporarily stored prior to the required manipulation to complete the task. Increased gamma synchronization has been reported during attention and maintenance phase of WM [104]. Gamma activation found to be partially location specific, depending on the type of cognitive task. Auditory processing regions such as the putative auditory dorsal and ventral processing streams found to have increased gamma activation following auditory WM tasks [104, 105]. Visual WM tasks such as the delayed-matching-to-sample task used by Tallon-Baudry and colleagues [106], caused increased gamma activation at occipital and temporal EEG electrodes during the maintenance phase. Gamma activation is also shown to be linearly correlated with WM load [107].  WM requires well-organized communication between various brain regions. This communication occurs along the WM networks where groups of neurons activate at the same frequency cycle. By measuring the connectivity of the network it is possible to determine how distant brain regions cooperate and transmit information [108]. Various neuroimaging modalities     	31	have been used to identify WM networks, regions that are activated during visual WM tasks, composed of frontal, temporal, parietal and occipital lobes. When assessing neuronal interaction and communication, oscillatory phase synchrony is important as each frequency band can have a specific role [109]. Payne and Kounios used a Sternberg Recognition task and found an increase in theta coherence between frontal midline and temporal-parietal regions and alpha coherence in midline parietal and left temporal-parietal [110]. Furthermore, a MEG/EEG study found inter-area phase synchrony was strengthened with increase WM load, particularly in fronto-parietal regions in alpha, beta and gamma bands [111]. Using graph theoretical analysis, visual WM networks have been described, indicating connection density to be load dependent in alpha (10-13 Hz), beta (18-24 Hz) and gamma (30-40 Hz) within the fronto-parietal and visual areas [112]. Other graph theory components can provide more information about the visual WM functional networks. Clustering coefficient and path lengths have been found to be WM load dependent. Li and colleagues (2011) showed variance in changes depending on the hemisphere of the task. Clustering coefficient significantly increased with load in alpha, beta and gamma bands in the left visual field and all the bands in the right visual field. Path length was also influenced by load but in fewer frequency bands (Left: theta, beta and gamma; Right: beta). These results suggest that WM load changes the local connectedness of the brain networks [113].  Recently, Zhang and colleagues (2016) found load dependent connectivity changes in theta along frontal midline areas. The curve of the connection strength increased from loads 1 – 4 before decreasing. The accuracy of the task also decreased with load [108]. This data supports the concept of WM capacity, which is the amount of information that can be temporarily stored and manipulated. Behavioural studies have suggested that WM has a capacity of holding up to four items before faltering [114, 115]. The connectivity results suggest that once WM exceeds     	32	capacity there is a decrease in efficiency and activity within the WM network that leads to a decrease in accuracy.   3.2 The Brain And Exercise Exercise is known have many health benefits including improved cognitive learning, executive function and even protection from age related decline [116]. Through animal models, the mechanisms responsible for these benefits have been studied extensively. They include: improved plasticity and neurogenesis within the hippocampus (particularly the dentate gyrus [117]) and increased levels of synaptic proteins [118] and glutamate receptors (NR2b and GluR5) [119]. Exercise is also responsible for increased availability of growth factors, such as brain derived neurotrophic factor (BDNF) and insulin growth factor 1 (IGF-1) [120, 121]. Exercise also contains neuroprotective properties aiding in both recovery and reducing the severity of many types of injuries and illnesses including depression [116]. These benefits have primarily been reported as a result of exercise sustained over an extended period of time (3-12 months) [122-125]. While research has shown benefits from acute exercise, these studies primarily focus on immediately following exercise completion and not during the exercise session. The specific influences exerted on the frontal brain regions during acute exercise remain to be fully explored.  Motor skills such as those involved in sport participation, require the integration of information from a wide range of areas including peripheral sensors, spinal locomotor networks as well as motor and premotor cortices [126]. Due to the temporal resolution that EEG provides, it is possible to study the neural control of movement. As mentioned above, by studying the oscillatory components of the EEG signal, it is possible to investigate neuronal interactions and     	33	communications.  With exercise, theta, alpha, beta and low gamma frequencies have shown to be of interest and reflect different aspects of motor planning, execution and control [127, 128].  During sport specific activities, increases in frontal midline theta power was linked to improved performance in expert golfers [129] , and rifle shooters [130]. The source of theta FM has been localized to the medial frontal cortex and the ACC, areas important for focused attention. Furthermore, along with playing a major role in cognition, learning, and memory; theta oscillations have been proposed to be involved in the integration of sensory and motor information during sensorimotor actions [131]. A recent study by Cruikshank (2012) found increased theta power during a sensorimotor task in motor areas (C3 and C4) [132]. Similarly, theta power was found to increase during movement onset of upper limb ballistic movements in contralateral motor cortices [133].  During and immediately following large body movement such as treadmill walking and cycling, studies have shown EEG activity to increase across frequency ranges [134, 135]. Bailey and colleagues (2008) found increased activity across EEG frequencies, including theta, during a graded cycling session to fatigue. This increase in activity was shown to occur at multiple electrode sites, leading the authors to question if peripheral physiology was the driver of the EEG activity [135]. Alpha activity in response to exercise has been difficult to quantify reliably, as studies show both increased power [135, 136] and decreased power [126, 137]. An important distinction to take under consideration is the time of cortical recording in relation to the exercise. Kubitz & Mott (1996) recorded EEG during a 15-minute (three 5-minute stages) session of progressively more intense exercise (50 to approximately 150 W) on a cycle ergometer. They found an exercise related decrease in alpha activity and corresponding increase in Beta activity over baseline values at each exercise load. The activity returned to baseline after completion of     	34	the exercise [137].  More recently, Enders and colleagues recorded EEG during high intensity cycling and found increase in alpha, beta and gamma activity over the left frontal cortex. This increase in activity corresponded with fatigue, specifically for the alpha and beta frequency bands. The gamma activity was unchanged with fatigue. They proposed that their results indicate involvement of the cerebral cortex during cycling in an executive control and motor planning capacity.  When cortical activity is measured following exercise completion there is a similar activation pattern [126], however, the frequency band and region of activation is dependent on exercise type and familiarity with the modality. Brummer and colleagues (2011) had experienced runners perform treadmill running, cycling on an ergometer, arm crank, and isokinetic movement at 50% and 80% intensity and recorded EEG prior and following the exercise. Following moderate intensity exercise, alpha activity showed an increase with all exercise modalities, while beta increased over the parietal cortex during the bicycle trial only. In contrast, during high intensity exercise alpha activity had modality dependent changes. Cycling and hand crank resulted in no change in activity; treadmill running had decreased alpha activation and isokinetic trials showed increased alpha activity.  Beta activity meanwhile showed a decrease in activation in frontal brain regions for the treadmill trial [138]. Another study meanwhile found increased absolute power in beta following a graded cycling test [139]. These results suggest that exercise type and intensity alters cortical activation and that the region of activation is related to participant familiarity with the modality.  A meta-analysis by Crabbe and Dishman (2004) assessed the literature on EEG changes during and immediately following exercise. They concluded that EEG activity in delta, theta, and beta frequencies increased both during and immediately following exercise. Activity in alpha     	35	showed changes in absolute power, but not in terms of relative power to the other frequencies. The authors also noted that the changes in activity were widespread and not grouped to specific cortical sites [134].  Few studies have attempted to assess the impact of acute exercise on functional connectivity measures. The sparse existing literature suggests that immediately following exercise there is an increase in functional connectivity in sensorimotor areas, while no impact has been reported in frontal regions [140].  3.3 Combining Cognition And Exercise  The incorporation of a cognitive task with the graduated exercise for the RTP protocol strives to increase the load on the brain in order to elicit symptoms from the post-concussed brain. The hypofrontality theory proposed by Dietrich (2003) suggests that during exercise there is a high level of activation within the motor and sensory cortexes which leads to a re-allocation of the limited resources normally needed for information processing [141]. This results in an inhibition of neural networks not involved in exercise. Therefore if an individual attempts to perform a task using the PFC, a brain region not heavily involved with exercise, it is much more difficult and requires increased effort. Davranche and McMorris (2009) assessed cognitive function during steady state cycling at individual lactate thresholds and found improvement in RT during trials of the Simon task but impaired function in response inhibition [142]. A similar study had participants cycling at 30%, 50% and 80% of their heart rate reserve (HRR) while performing the Wisconsin Card Sorting Test [143]. They found significant impairment in performance during high-intensity exercise but not moderate or low intensity [144]. They concluded that their results lent support for the hypofrontality theory. Recent meta-analyses on the effect of acute exercise on cognition found that exercise has a consistent positive effect on     	36	cognitive tasks completed following exercise completion [145, 146]. When cognition is assessed during exercise, the results are less conclusive. The variations in results were proposed to be methodological, as Chang and colleagues included a wider range of studies and participants, resulting in a small positive effect on cognition [145], whereas Lambourne and Tomporowski incorporated a narrow scope of studies, only including healthy young participants and a negative effect. The meta-analyses identified moderators that were found to influence the outcome of studies that included exercise intensity, time of cognitive testing and duration, cognitive task type, and fitness of participants. Following a meta-analysis on the effect of walking on cognition, Al-Yahya and colleagues found working memory and executive function tasks to be the most consistently and strongly influenced by exercise [147].  Neuroimaging provides an opportunity to understand the cortical origination of the changes in cognition with exercise. A study by Li and colleagues (2014) had participants complete an N-back test following rest and immediately following a 20-minute cycle at 60-70% HR Max. The authors found that behaviorally, the exercise did not influence the accuracy of the task but did have a significant effect on brain activation during the most difficult level of the N-back. Brain activation was increased in the right middle prefrontal gyrus, the right lingual gyrus, and left fusiform gyrus, areas involved in executive function. Decrease activation was noted in the anterior cingulate cortexes, left inferior frontal gyrus and the right paracentral lobule [148], which the authors implied was a transition to a default mode status and compensatory mechanism for executive processes.  EEG remains sparsely used for exploring cognition during exercise. One study had participants cycle at 60% of the HR max and found that as difficulty increased, in a modified flanker task, accuracy decreased and cortical activation increased [149]. Specifically, they found     	37	increased amplitude in the time-locked EEG signal in frontal brain regions. The authors suggested that the increased activation and decreased accuracy are indicative of increased inefficiency of the neuroelectric system. This leads to an increase in effort and resources to perform the task. Another, more recent paper by Olson and colleagues (2016) supported these results, as they reported a decrease in accuracy and corresponding increase in EEG activity at both 40% and 60% exercise intensity during a modified flanker task [150]. This increase in effort to complete the cognitive task was further supported by a study with adolescents that had participants exercise for 20 minutes at 60% of their HR max and complete a cognitive task composed of an Eriksen flanker task and Go No-Go task [151]. Using coherence analysis, which explores the synchrony of brain oscillations across different scalp locations, they found unfit adolescents showed increased coherence in Alpha and Beta frequencies. Increase coherence is suggested to represent increased effort. These results support two meta-analyses that identified fitness level to be a moderator that influences cognition during exercise [145, 146].  Overall, the results from the behavioural and neuroimaging studies support the theory of transient hypofrontality [141], in that with exercise there is a decrease in cognitive task accuracy and increase in cortical activity. This increase represents more resources being required to perform the task. The results from the neuroimaging studies suggest a decrease in neural efficiency during combined cognition and exercise. Graph theoretical analysis has the potential to provide more insight into how the various attributes of the functional networks are impacted by this paradigm design.        	38	Chapter 4: Methods  The present study has received approval from UBC’s Clinical Research Ethics Board (H15-02714). All participants independently provided written and verbal informed consent, in accordance with the principles outlined by the Declaration of Helsinki.  4.1 Participant Information All interested participants were required to meet the following inclusion criteria:  o 18-25 years old o Right handed o No history of prior concussion or head injury  o No history of drug or alcohol abuse  o No diagnosis of learning disability or other neurological disorders; In addition, interested individuals were screened over the phone using the Godin Leisure Questionnaire (see Appendix A). In order to participate, individuals had to be moderately active or a score of greater than 14 on the questionnaire. This was selected as familiarity with exercise and fitness level is known to influence the effect of exercise on cortical activity [145, 151].  All participants were recruited from the University of British Columbia and surrounding area through flyers (Appendix B) and word of mouth.  4.2 Anti-Saccade Serial Addition Task (ASAT) The Anti-Saccade Serial Addition Task is a novel task that is comprised of two (2) well established tests: the Anti-Saccade Task and the Paced Serial Addition Task. The Anti-Saccade     	39	task is a classic test used to assess cognitive control and requires participants to inhibit a reflexive saccade [80]. The task is composed of two steps. First the participant must suppress the reflex to look at the stimulus and then create a voluntary motor command to look away from the target. The task requires a range of cognitive processes including inhibitory control, attention, working memory, and decision-making [152]. Extensive neuroimaging has been done using the Anti-Saccade task, identifying the frontal eye fields (FEF), supplementary eye fields (SEF), ACC, and DLPFC to be involved in the completion of this task [83, 153].  The Paced Auditory Serial Addition Task (PASAT) was originally developed for individuals with TBI [154]. Since then it has been used in many clinical populations including MS, whiplash, chronic fatigue syndrome and depression [154]. The test is a validated measure used to evaluate attention, executive control, working memory, and information processing speed [154]. Although shown to be valid and sensitive to many clinical conditions, the PASAT is very difficult and often causes extreme frustration and anxiety in participants [155]. The Paced Visual Serial Addition Task (PVSAT) was developed as an alternative to the PASAT and is moderately correlated at all difficulty levels [155]. Clear differences exist between the two in terms of difficulty, with the PVSAT being significantly easier and correspondingly a possible ceiling effect has been reported.  One benefit of the visual version is that it solves the input-output interference problem with the PASAT [155]. The cortical regions involved in the PVSAT are mainly in the frontal and parietal lobes (superior and inferior parietal lobe bilaterally, superior frontal gyrus bilaterally, left medial frontal gyrus, left inferior frontal gyrus, and adjacent part of the insula, anterior part of the cingulate gyrus and some cerebellar areas [156].     	40	4.3 EEG Data Collection Electroencephalographic data was recorded using a 32 channels EEG ASAlab system with Waveguard Technology cap (Advanced Neuro Technology, Enschede, Netherlands). This system is supplied with shielded wires to make recordings less susceptible to external noise and movements. EEG data was continuously recorded using a 500 Hz sampling frequency. The ground electrode (AFz) and common average reference was positioned between Fpz and Fz to ensure low impedance values (generally < 5 KΩ). The 32 electrodes were distributed along the scalp according to the 10/5 system [157]. The cap was fixed with a chinstrap to prevent shifting during the exercise trials and was permeable to air in order to prevent an increase in heat during exercise. Each electrode was filled with OneStep EEG-Gel (H + H Medizinprodukte GbR, Münster, Germany) for improved signal transduction. To ensure consistent cap placement, the vertex (Cz) electrode was placed midway between ears, and midway between the nasion and inion. On the first day, the participants were asked to sit still with their eyes closed for 5 minutes to collect resting state data.  4.4 EEG Data Analysis Analysis of the EEG signal was completed offline after each participant completed all of his or her visits. The EEG data was exported from the collection device and brought into Brain Electrical Source Analysis® Research (BESA) for analysis. The EEG signals were first filtered using a band-pass filter (4 – 50 Hz) and notch filter (60 Hz) to remove signal drift, line noise and motion artifacts. Independent Component Analysis (ICA) was used to decompose the signal and identify eye blinks, which were then removed from analysis, as were channels with excessive noise. An automated artifact scan was performed to check signal for noise. Participant data was included in analysis as long as 70% of trials were clean of artifacts. The task sent triggers to the     	41	EEG system allowing for the identification of each stimuli presentation. Using the accuracy data, all trials in which the participant responded incorrectly, were removed from further analysis. The data from all accurate trials for each task and block were then averaged using 1.24 seconds epoch (-0.24 – 1000 ms).  A 10-10 average virtual montage was applied to the data, resulting in 27 channels. This data was then exported into MATLAB (Version R2013b, The Mathworks, Inc., Natick, MA, USA)). Within MATLAB, using scripts developed in the lab, the average signal from each participant was averaged together to form a single “Grand Average” signal that represented the group’s ERP in response to the task. The data was plotted (Figure 5 – 3) to allow for visualization of the ERP over the entire scalp. Using this plot, the signal was visually evaluated and regions that showed significant peaks of activation were highlighted. The brain regions of interest for this study were chosen a priori and this process was used to confirm that the chosen regions were in fact involved in the tasks. The visual inspection of the data confirmed our chosen areas of interest with the left, central, and right frontal regions and occipital midline undergoing further analysis. 4.4.1 Power Spectrum Analysis In BESA, a virtual source montage was applied to the signal. This montage, as discussed in Chapter 2, reconstructs approximate source waveforms that can be used to represent cortical activity. Fast Fourier Transforms (FFT) was then applied to EEG signal, which transforms the data from the time domain to the frequency domain. This allows for the calculation of power at each frequency band. For this study, the data was segmented into theta (4-8 Hz), low alpha (8-10 Hz), high alpha (10-13 Hz), beta (13 - 30 Hz) and gamma (30 – 45 Hz). The output of the FFT was absolute power (nAm2) and these values were used for statistical analyses.     	42	4.4.2 Brain Connectivity Network Modeling The data was exported to MATLAB once more under the Virtual Source montage. A local script was used to segment the signal into theta (4-8 Hz), alpha (8-13 Hz), beta (13-30 Hz) and gamma (30-45 Hz). EEG signal in each frequency band was then run through the connectivity analysis. To compute the brain connectivity networks, the PC controlled false discovery rate (PCFDR) algorithm was used[158]. PCFDR is a computational method based on the error rate criterion of the discovered network. Partial correlation is used to evaluate the conditional independence, which estimates the directed interactions between any two-brain regions after removing the effects of all other brain areas. The PC algorithm starts from a complete graph and tests for conditional independence in an efficient way. The PCFDR algorithm asymptotically controls the false discovery rate (FDR) below the predefined levels, which evaluates the proportion between the connections that are falsely detected to all those detected. Compared to the traditional Type-1 and Type-2 error rates, FDR has more conservative error rate criteria for modeling brain connectivity due to its direct relation to the uncertainty of the networks of interest. The PCFDR algorithm and pseudo-code are described in details in [159]. The FDR threshold was set at the 5% level. The learned connectivity networks are binary undirected graph with the inferred connections at the 5% significance level. The binary undirected networks were computed for each individual for all frequencies, task conditions and blocks independently.   4.4.3 Graph Theoretical Analysis Graph theoretical analysis was used to extract structural network features from the learned networks [76]. Traditional graph theoretical measures were used to characterize the network features in terms of density, global efficiency, clustering coefficient, and modularity.     	43	The definitions of these measures can be reviewed in Chapter 2. The Brain Connectivity Toolbox [74] was used to perfrom the graph theoretical analysis. For the network, each of the 15 virtual source montage sources represented a brain region within the network.  4.5 Experimental Protocol This study employed a single group randomized repeated measures design. Interested participants were appropriately screened by a phone interview to ensure they met all of the stated inclusion criteria. Participants deemed eligible were invited to visit the Djavad Mowafaghian Centre for Brain Health (CBH) at the University of British Columbia. Participants were asked to come in to CBH on three (3) different days within a 7-day period, at the same time of day, in order to account for the effect of circadian rhythm on cognition [160]. On the testing days the participants performed one of the three tasks: ASAT, Exercise or ASAT-Exercise dual task. The order of the tasks was randomized to account for any order effect. After initial telephone screening, each participant had his or her task order randomized prior to his or her first visit.   During the first visit to the Centre for Brain Health, each participant was met by the research coordinator and asked to review a written consent form (see Appendix C), as well as encouraged to ask questions as they arose. Physical activity level was quantified using the Recent Physical Activity Questionnaire (RPAQ), which has undergone extensive reliability and validity testing [161]. The RPAQ consists of questions across three (3) different activity domains to determine the individuals’ average activity level over the last four (4) weeks (Appendix D). The participant then completed the ASAT task training. This included two decomposed proponents of the task. The total learning period took 30 minutes. Following the task training, the EEG cap was placed on the participants’ head, after which they were instructed to sit still for 5 minutes, while keeping their mind clear to allow for the collection of resting state brain activity.      	44	The study procedure was composed of the three testing conditions. The order was randomly assigned immediately following acceptance into the study. After the EEG cap was placed on his or her head, the participants were instructed to sit on the stationary bicycle (Zhejiang Everbright Industry, Inc, Taichung City, Taiwan). Participants were asked to adjust the seat to ensure optimal pedal distance and the bike was then adjusted to keep the wall mounted screen a set distance based off of their height in order to maintain the correct angles for the anti-saccade component of the ASAT. The participants were then instructed to complete one of the three (3) task conditions. 4.5.1 Cognitive Task: ASAT The task began with a fixation point in the middle of the screen, followed by a distractor stimulus (red dot) being shown for 100 ms at a fixed distance to the left or right of the centre (in random sequence). In the opposite direction to the distractor stimulus, the target stimulus (single digit) was presented for 200 ms (Figure 4 – 1). The task required the participant to mentally track and verbally add the sequentially presented digits. The participants also had to ignore the distractor stimulus, or else they would miss the target stimulus as a result of the time cost attributed to the erroneous pro-saccade. The ASAT was presented using a commercially available stimulus presentation software (E-Prime 2.0, Psychology Software Tools). The task consisted of 5 blocks, with each block increasing in difficulty through changes in the anti-saccade angle, the inter-stimulus interval and number of trials (Table 4 - 1). Participants were instructed to respond verbally as quickly as possible following each stimulus.   Table 4 – 1. Block progressions for the Anti-Saccade Serial Addition Task and Exercise Intensities.     	45	 Block 1 Block 2 Block 3 Block 4 Block 5 ASAT inter-stimulus interval (sec) 3.0 2.5 2.0 1.5 1.0 ASAT distractor off-set (degrees) 5 7.5 10 12.5 15 Heart rate reserve (%) 30 40 50 60 70  4.5.2 Exercise The participants were asked to change into comfortable clothing for exercise. At this time they were asked to put on a Polar Heart rate monitor band (Polar Electro, Oy, Kempele, Finland). The participant was then asked to sit quietly for 5 minutes in order to measure their resting HR. Using their age, their max HR was calculated using equation 1. Their target HR was then calculated using Equation 2.       !"!"# = 208− 0.7 × !"#                                      (1)     !"!" = !"!"# −  !"!"#$  ×%+ !"!"#$    (2) The exercise task was divided into 5 separate blocks of increasing intensity (Table 4 - 1). Based off of the target HR calculated using equation (2) the participant was asked to begin biking at 30% of their heart rate reserve. The research coordinator monitored the participant’s heart rate and modulated the intensity of the exercise appropriately by increasing or decreasing the resistance, as the participant was asked to maintain a steady RPM of 60 throughout the task. Once a steady state HR was obtained, the participant continued for 5 minutes. After 5 minutes the participant increased their intensity to 40% and repeated for 50%, 60% and finally 70%. Steady state HR was defined as within +/- 5 beats per minute and was constantly monitored by the research administrator to ensure compliance and adjust the necessary settings.       	46	During the last 30 seconds of each exercise block the participants were shown a Borg intensity scale (Appendix E) and asked to identify their perceived exertion using the scale. Following the final exercise block the participants were instructed to cool down for 5 minutes.  4.5.3 ASAT-Exercise Dual Task  Participants changed into comfortable clothing for exercising and placed the Polar Heart Rate monitor band under their clothing. The participants were asked to sit on the exercise bike while the cap was placed on their head. Once the cap was firmly placed on their head the participants was asked to begin pedaling at 30% of their HRR. Once reaching steady state the ASAT began. At each increase of exercise intensity, a corresponding increase in ASAT difficulty occurred, reaching a max difficulty at the end (Table 1). During the final 30 seconds of each exercise block the participants was shown a Borg intensity scale and asked to identify their perceived exertion on the scale.   Figure 4 – 1. An example of a single ASAT task trial.      	47	4.6 Outcome Measures 4.6.1 Rated Perceived Exertion The participants’ rated perceived exertion was recorded during each block of the Exercise – only and Combined Cog-Ex tasks using the Borg scale.  4.6.2 ASAT Accuracy Participant responses were recorded using an electronic recording device. All responses were then input into a database for analysis. The main outcome variable of interest for the ASAT task was response accuracy (%). Accuracy was determined for each block by calculating the correct number of responses and dividing the sum by the total number of trials.  4.6.3 Power Spectrum Analysis The main outcome variable for the EEG Power spectrum analysis was absolute power, a measure that has been shown to be more responsive to change in exercise studies [134]. Absolute power was compared across tasks (3), loads (5) and brain regions for each frequency.  4.6.4 Functional Connectivity Based on prior research in our lab, two graph theoretical measures were chosen a priori as the main outcome measures for assessing the networks during the various tasks. These critical measures of network connectivity included degree, betweenness centrality, and clustering coefficient.       	48	4.7 Statistical Analysis 4.7.1 Experimental Outcome Measures The dependent variables for this study included accuracy, absolute power, and the connectivity measures. All statistical tests were performed using IBM SPSS Statistics (Version 23.0.0.0, IBM Corporation, Armonk, New York, USA). A statistician was consulted to ensure appropriate analyses were run to test each of the hypotheses. A linear mixed model (LMM) was used to model the effect of task condition, levels, regions, and interactions had on the task accuracy and the EEG measures. A LMM was used as it offered more flexibility and a better fit for the data than traditional analyses such as repeated measure ANOVA, ANCOVA, and regression models. Using a mixed model allows for the incorporation of each participant’s variability by computing a random slope and a random intercept for each participant, which takes into account the participant differences at each level.  For the behavioural statistical analysis, the model was used to identify differences in task accuracy during the Cognitive-only and Combined Cog-Ex tasks, as well as between difficulty levels. Since the study design had multiple conditions within a single subject, correlation within subject data had to be accounted for. For the EEG data, the model had Task, Level and Region as fixed effects. Participant and Task type were used as random effects. Task type was considered a random effect as participants performed each task on a separate day. Day to day differences within each of the participants was thereby accounted for within the model. The model used for most of the analysis was the following:     	49	! =  ! + !!!"#$ +  !!!"#"$ +  !!!"#$%& + !! + !!!"#$ + !!!"#$%&%'"($ + ∈  In this model, Y is the dependent variable, either power amplitude or graph theory measure of the EEG signal. α represents the average dependent variable baseline value. !!!"#$, !!!"#"$, and !!!"#$%& represent the average task, level, and region effects on the slope of the model. !! , !!!"#$,!"# !!!"#$%&%!!"# represent the random variation in the intercept and slope of the model.  represents the residuals of the model. Residuals from the model were tested for symmetry. A reference for substantial departure from normality was suggested to be an absolute skew of 2 [162]. If the residuals surpassed the skew threshold the data underwent a logarithmic transformation and re-run through the model. Only significant main effects and interactions present in both models are reported. Significance was set at 0.05 and all significant pairwise comparisons were corrected for multiple comparisons through a Bonferroni correction.        	50	Chapter 5: Results 5.1 Participant Characteristics Due to the extremely limited time available with the EEG cap and system, thirteen (13) participants were recruited to participate in the study. Of this group, eight (8) were males and the remaining five (5) were females. The mean age of the participants was 22.5 years (0.65).  The mean years of education was 16.4 years. The average total leisure activity score was 48.0 (5.6) from the Godin Leisure-Time Exercise Questionnaire, indicating the group was very active. [163]. The participants’ physical activity level was confirmed by the RPAQ, which revealed a mean of 1.38 hours (0.22) of moderate to vigorous active per day. Table 5-1 provides a summary view of the participant characteristics.  Table 5-1 Participant Characteristics including Age, Gender, Godin Total Leisure Activity score and RPAQ scores. ID Age Gender Godin Total Leisure Activity Score RPAQ  P001 22 M 71 3.7 P002 20 F 21 1.4 P003 18 F 64 1.1 P004 21 F 74 1.0 P005 24 M 27 0.9 P006 19 F 34 1.5 P007 25 M 66 1.4 P008 24 M 77 1.5 P009 24 M 44 1.4 P010 23 M 28 1.3 P011 24 F 39 0.9 P012 24 M 50 1.6 P013 25 M 29 0.3 AVERAGE 22.5 M = 8/F=5 48.0 1.38 SME 0.65  5.6 0.22     	51	5.2 Behavioural Results 5.2.1 Heart Rate And RPE Mean Heart Rate and RPE are presented in Figure 5-1 for the Exercise – only and Combined Cog-Ex tasks at each level. Both HR and RPE show no significant effect of task condition, with values being almost identical throughout the levels. The LMM identified a significant effect on Level for both HR and RPE, F (17.9, 107) = 1143.5, p = < .001, F (52.7, 95.05) = 165.88, p = < .001. Post hoc analyses using Bonferroni correction for multiple comparisons revealed significant increase in HR and RPE between each level (p < .001) as both values gradually increased from Block A to Block E.   Figure 5 – 1 HR and RPE for the Exercise and Combined Cog-Ex Tasks. Participant Heart Rate (HR) (Left) showed significant increases between each of the levels in both conditions (p < .001). Rated perceived exertion (RPE) showed a similar pattern, with increased ratings every block (p < .001). Both HR and RPE showed no significant differences between conditions.   0	20	40	60	80	100	120	140	160	180	A	 B	 C	 D	 E	HR	0	5	10	15	20	A	 B	 C	 D	 E	RPE	 Exercise	Combined	    	52	5.2.2 Accuracy Of Anti-Saccade Serial Addition Task.   Mean ASAT accuracy was compared between the Cognitive only task and Combined Cog-Ex task, showing that the Combined Cog-Ex task had a lower combined accuracy than the Cognitive-only task although this did not reach significance F (30.5, 55.35) = 3.71, p = .078 (Figure 5 – 2 A). Both task conditions were then combined to assess changes by level of difficulty and are presented in Figure 5 – 2 B. There was a significant main effect for Level, F (30.5, 95.3) = 63.7, p < .001.  Post hoc analysis revealed Accuracy was stable during Blocks A – C and showed a significant decrease in the final two blocks.     Figure 5 – 2 Working memory task accuracy. 80	85	90	95	100	Cognitive	-	Only	 Combined	Accuracy	%	A50	55	60	65	70	75	80	85	90	95	100	Block	A	 Block	B	 Block	C	 Block	D	 Block	E	Accuracy	%	B	*			*	    	53	(A) Accuracy differences between task, although not significant (p = .078), the Combined Cog-Ex task showed a lower accuracy overall. (B) Combined accuracy (%) of both tasks for each block. There was a significant decrease in accuracy in Blocks D and E (p < .001). * Indicates significance (p < .05)  5.3 EEG Results 5.3.1 ERP The grand averaged ERP was calculated using the 10-10 montage for all blocks for each condition. The frontal regions (F3, Fz, F4) and occipital midline (Oz) were selected a priori as regions of interest due to the task design. Figure 5 – 3 illustrates large peaks in activation in these regions during both the Cognitive-only and Combined Cog-Ex tasks, with both tasks showing very similar patterns of activation.       	54	     Figure 5 – 3 Visualization of ERP signal over the entire scalp.  Each box represents the ERP signal for that corresponding electrode within the 10-10 montage. Cognitive – only (Blue) and Combined Cog-Ex (Green) show no significant differences in activation pattern. Regions of interest: F3, Fz, F4, O2, Oz, O1 show the expected increase in amplitude during the tasks.  Cognitive	–	only	Combined     	55	The ERP data was used to confirm the regions of interest and their activation during the various tasks. As a result, the data was evaluated strictly qualitatively and no statistical analyses were performed, as the ERP were not the variables of interest in this study. Figure 5 – 4 shows the changes in the grand average ERP at each block for the regions of interest. Compared to rest, there is an increase in peak amplitude during the first found blocks of the tasks, with all frontal electrodes showing a similar pattern of activation. The occipital region showed a different pattern with less of a significant peak and a more continuous pattern of activation throughout the trial.  Within the frontal electrodes there is a change in pattern of the EEG response during the final block, moving from a large singular peak of activation to a pattern with decreased amplitude but more continuous activation throughout the task.                   	56	  Figure 5 – 4 Group Average ERP during rest and all difficulty blocks.  A clear ERP is discernable from all task conditions compared to the Resting state. For all of the frontal electrodes (F3, Fz, F4) there is little change in activation throughout the first 4 blocks (A-D) before a substantially different signal in Block E. The ERP in Oz is different from the frontal electrodes and shows a signal with increased latency throughout the blocks.      5.3.2 Power Spectrum Analysis The power spectrum analysis separated the EEG signal into five (5) different frequencies: Theta, low Alpha, high Alpha, Beta and Gamma. Each frequency was run separately through the LMM and will be presented by main effect followed by interactions. Main Effect: Condition Figure 5 – 5 shows a significant Main Effect for Condition was found for Theta, F (268.2, 359.7) = 10.68, p = .001. Post hoc analysis revealed Theta power was significantly higher during the 	F3 FZ																							 F4 OZ 	Rest A B C D E     	57	Exercise – only (p < .001) and Combined Cog-Ex (p = .042) tasks compared to the Cognitive – only task condition.  Figure 5 – 5 Absolute Power of the EEG signal for each frequency band. Minimal fluctuations were seen in most frequency bands. Theta power showed significant differences between conditions, with increased power in both Exercise-only and Combined Cog-Ex tasks over the Cognitive-only condition.  * Indicates significance (p < .05) Main Effect: Region A significant main effect of region for theta F (268.2, 665) = 399.55, p < .001 was observed (Figure 5 – 6). Post hoc analysis revealed significantly higher power in OpM compared to all frontal regions (p < .001) and lower power in FM (p < .001).  A significant main effect of region for low alpha power F (219.0, 665) = 148.27, p < .001 was reported (Figure 5 – 6). Post hoc analysis revealed significantly higher power in OpM compared to all frontal regions (p < .001).  0	200	400	600	800	1000	1200	1400	Theta	 L-Alpha	 H-Alpha	 Beta	 Gamma	Absolute	Power	(nAm2)	Cognitive	-	only	Exercise	-	only	Combined		*	 	*	    	58	A significant main effect of region for high alpha power F (331.0, 665) = 428.69, p < .001 was reported (Figure 5 – 6). Post hoc analysis revealed significantly higher power in OpM compared to all frontal regions (p < .001) and lower power in FM compared to all other regions (p ≤ .001). A significant main effect of region for beta power (F (226.5, 665) = 214.7, p < .001) was reported (Figure 5 – 6). Post hoc analyses revealed significantly higher beta power in OpM and significantly lower in FM compared to other regions (p < .001). A significant main effect of region for gamma power (F (242.4, 665) = 148.6, p < .001) was reported (Figure 5 – 6). Post hoc analyses revealed significantly higher gamma power in OpM (p < .001) and significantly lower in FM compared to other regions (p < .05).  Figure 5 – 6 Absolute Power of the EEG signal from each frequency band by Region. With all tasks combined, power in OpM was significantly higher than any other region for all frequency bands. FM was significantly lower than the other frontal regions and OpM in every frequency except low alpha.  * Indicates significance (p < .05)  0	500	1000	1500	2000	2500	3000	3500	Theta	 L-Alpha	 H-Alpha	 Beta	 Gamma	Absolute	Power	(nAm2)	FL	FM	FR	OpM								*												*						*						*						*					*	 				*	 					*	 				*	    	59	Main Effect: Level   Figure 5 – 7 shows a significant main effect for level was found for beta, F (226.5, 665) = 4.46, p = .001.  Post hoc analysis revealed beta power was significantly higher in Block E (p < .005) compared to Blocks B and C when all tasks were considered together. A significant main effect for level was found for gamma, F (242.4, 665) = 5.87, p < .001 (Figure 5 – 7). Post hoc analysis revealed gamma power was significantly higher in Block E (p < .005) compared to Blocks A, B and C (p ≤ .001) and approached significance in Block D (p = .066) when all tasks were considered together.            	60	  Figure 5 – 7 Absolute Power of EEG signal throughout the blocks. When tasks and regions are combined, there is a significant increase in Absolute power in the final block (E) in Beta (Blue, top) and Gamma (Red, bottom) frequencies (p < .005). * Indicates significance (p < .05)     0	200	400	600	800	1000	1200	1400	1600	1800	A	 B	 C	 D	 E	Absolute	Power	(nAm2 )	Beta		*			0	100	200	300	400	500	600	A	 B	 C	 D	 E	Absolute	Power	(nAm2 )	Gamma	*	    	61	5.3.3 Functional Connectivity Global Connectivity Changes in global functional connectivity were explored using graph theoretical measures and showed no significant changes in density, global efficiency, clustering coefficient, or modularity. The analysis then focused on more localized changes through connectivity between regions.  Local Connectivity The local connectivity analysis was focused entirely on the frontal brain regions, frontal left (FL), frontal midline (FM) and frontal right (FR), as this was the area primarily expected to be involved in the task.  Main Effect: Task  A significant main effect for task was reported for beta Degree, F (12, 348) = 6.09, p = .014 (Figure 5 – 8). Post hoc analysis revealed a significantly higher value of degree in the Cognitive – only task –keep terms consistent between task and condition compared to the Combined Cog-Ex condition (p = .014).   Figure 5 – 8 Changes in graph theoretical measures between the two cognitive tasks. Beta Degree is significantly lower during the Combined Cog-Ex task than the Cognitive-only task (p < .014). * Indicates significance (p < .05) 3.4	3.5	3.6	3.7	3.8	3.9	4	4.1	4.2	Beta	Degree	Cognitive	-	only	Combined	*	    	62	Main Effect: Level  Figure 5 – 9 shows a significant main effect for level with theta Clustering Coefficient, F (12, 336) = 2.59, p = .037. Post hoc analysis revealed Block D to be significantly lower than Blocks B (p = .011) and C (p = .005).   Figure 5 – 9 Changes in Theta Clustering Coefficient through levels. With both tasks combined, theta clustering coefficient is significantly lower in Block D compared to Blocks B and C.  * Indicates significance (p < .05)  Interaction: Region X Task A significant Interaction was found for theta Degree, F (12, 348) = 4.18, p = .016 (Figure 5 – 10). Post hoc analysis revealed a significantly higher value of theta Degree in the Cognitive – only task at FM compared to the Combined Cog-Ex condition (p = .029).   0	0.05	0.1	0.15	0.2	0.25	0.3	A	 B	 C	 D	 E	Clustering	CoefSicient	*	    	63	 Figure 5 – 10 Theta Degree Region X Task Interactions. Theta Degree presented with significantly lower values during the Combined Cog-Ex task compared to the Cognitive –only task in FM only (p = .03). The Frontal left and right regions did have the same pattern.  * Indicates significance (p < .05)   There was a significant Region by Task interaction for beta Betweeness, F (360,360) = 3.87, p = .021 (Figure 5 – 11). Post hoc analysis revealed a significantly higher value of beta Betweenness in the Cognitive – only task at FL compared to the Combined Cog-Ex task (p = .007).    2.5	2.6	2.7	2.8	2.9	3	3.1	3.2	3.3	3.4	3.5	FL	 FM	 FR	Degree	 Cognitive	-	only	Combined	*	    	64	 Figure 5 – 11 Beta Betweeness Region X Task Interaction Beta Betweeness presented with significantly lower values during the Combined Cog-Ex task compared to the Cognitive – only task in FL only (p = .007). * Indicates significance (p < .05)    0	0.01	0.02	0.03	0.04	0.05	0.06	0.07	0.08	0.09	FL	 FM	 FR	Cognitive	-	only	Combined	*	    	65	Chapter 6: Discussion The objective of the current study was to examine the association between EEG metrics and behavioural changes in healthy normal adults as a foundation for evaluating individuals with concussion.  In order to accomplish this objective, healthy and physically active participants were asked to complete three tasks: a challenging graded cognitive task, a graded cycling session, and a combined task that had the participants complete both tasks simultaneously. Regarding the exercise paradigm, this study employed a paradigm that forced participants’ to place priority on the exercise, by maintaining a speed of 60 rpm throughout the levels and keeping HR within the set level. This forced priority had an influence on the results, as participants were not able to decide themselves on which task to prioritize. This study design was used to reduce the variability in the results and to match the existing literature, which utilized similar paradigms [143, 144, 149, 150]. Our results suggest the graded levels of exercise induced significant changes in both the perceived exertion and heart rate, with both values increasing at each progressive level. This result was seen in both the Exercise-only and Combined Cog-Ex task and suggests that the stepwise progressions were large enough to induce a significant change in the physiological and perceived exertion response.  Between the Cognitive-only and Combined Cog-Ex task there were no statistically significant differences in the accuracy of the task, however there were significant differences in the neural activity. Specifically, theta absolute power was found to be modulated by the task condition. The Combined Cog-Ex task condition elicited significantly greater activity in the frontal brain regions within the theta frequency band compared to the Cognitive-only condition. These results suggest that when exercise is added to the cognitive task, greater attentional resources are required in order to maintain accuracy. Multiple studies have identified theta     	66	activity to be positively correlated with cognitive load and accuracy of the task [92, 93, 95]. This result contributes to the current literature regarding theta‘s active role in cognition.  Consistent with literature using cognitive tasks, as the ASAT increased in difficulty, there was a significant decrease in the accuracy. Specifically, a significant decrease was noted in the final two levels (Blocks D and E), while no significant changes were recorded in the first three blocks. These results suggest that only tasks with a high level of cognitive difficulty result in a decrease in accuracy. The addition of the EEG data allows for the exploration of the impact task type and cognitive load has on neural activity as well as the relationship between the behavioural and neural levels of measurement.  Between the tasks levels there were significant changes in activity within the beta and gamma frequency bands. In particular, a significant increase was found during the final level of the tasks in both frequencies, coinciding with the largest decrease in accuracy. Beta and gamma activity has previously been identified to be involved in many aspects of cognition, including working memory. Beta activity has shown to be sensitive to working memory load, especially during tasks requiring no physical response, as physical movement attenuates activity in this frequency band [103]. Gamma activity meanwhile, is associated with attention and memory formation. Within working memory, gamma has shown to be involved in the maintenance phase of the task and increase almost linearly with load [104, 107]. The reported increased beta and gamma power as the cognitive task increased in difficulty suggests that increased activation was required to complete the task. Furthermore, there was a change in the pattern of activation of the EEG signal within the frontal brain regions during the hardest level of the cognitive task. The signal pattern went from a single, easily identified peak to a more continuous activation throughout the trial. These two results suggest the brain utilizes two strategies to manage the     	67	increase in difficulty – increases activity in frequency bands involved with cognition and alters the pattern of activation from a singular peak to a more continuous level of activation. This results in an overall increase in neural activity, which extends over a longer period of time.  During the Combined Cog-Ex task there were changes in connectivity within the frontal brain regions. In particular, beta degree was significantly lower during the combined Cog-Ex task condition compared to the Cognitive-only condition. Degree is a measure of network connectedness and nodal importance to the network. The higher the value of degree of a particular node, the more central and important the node is to the network as a whole. The reported decrease in degree during the Combined Cog-Ex task suggests that with the addition of exercise, the frontal network becomes less central, or important, to the performance of the task. Future studies should explore how the network changes and which regions become more involved during the combined task condition by including a greater number of nodes in a variety of brain regions.  As with the behavioural and earlier neural measures, there was a significant load effect within the functional connectivity measures. Theta clustering coefficient was shown to decrease as the cognitive load increased. Clustering coefficient is an important measure of local connectivity, as it measures the network being composed of small local clusters of connections. Healthy brain networks are characterized as being a small-world network, meaning it has short path lengths and high levels of clustering coefficient. During WM tasks clustering coefficient has been shown to be load dependent, increasing with high cognitive load [113]. The results from this study contrast with this consensus as clustering coefficient was found to decrease with increasing difficulty. A possible explanation for this contrasting result is the fact that both the Cognitive – only and Combined Cog-Ex task conditions were assessed together. Due to the small     	68	sample size, it was not feasible to assess the conditions individually.  It is possible that the current results suggest that as the task increases in difficulty, the frontal brain regions shift from an efficient and densely connected network to a larger, less efficient and less exclusive network involving other nodes outside of the frontal brain regions. Another possible explanation is that when the added exercise component reaches a certain level of difficulty, the neural activity shifts to accommodate the exercise, thereby altering the network away from a frontally centered network. Future studies should explore this possibility by increasing the sample size and assessing the two conditions individually.  6.1 Limitations Several limitations are present in this pilot study exploring the association between EEG metrics and behavioural changes in healthy normal adults during various sources of load. As note previously, the statistical power was limited in this study due to a small sample size, further research should be done using a larger sample as well as a group with concussion history. Although significant differences were noted between tasks, a larger group would be beneficial in assessing the differences between levels as well as the Task X Level interactions.  This study utilized a novel task that has not been compared directly to similar tasks. A study incorporating this new task along with the PASAT or PVSAT could be beneficial and allow for an improved comparison between the scores as we noted large differences in accuracy scores between the ASAT and the PASAT/PVSAT in similar populations. Future studies should include both tasks to elucidate how comparable the tasks truly are. A final limitation of this study was the small number of nodes used in the analysis. This greatly limited the ability to assess the cortical changes caused by the various sources of load.     	69	Future studies should include areas outside the PFC to get a better understanding of the impact on the whole brain. The functional connectivity analysis was particularly limited in this study as we were unable to identify how the connectivity was influenced by the exercise.   6.2 Applications For Sports Related Concussion This study provides the framework for future studies to investigate the differences in neural activation between healthy individuals and those recently recovered from sports-related concussion. The results imply that the addition of exercise has a measurable impact on neural activity and individuals recovering from SRCs would require greater neural resources that may not be available following injury. This would result in a measurable decrease in accuracy during a cognitive and exercise combined task condition. This information could be used in addition to symptom reporting to aid in identifying when athletes are able to return to play safely. To the author’s knowledge this is the first study to explore brain activation using power spectrum analysis and graph theoretical analysis during a combination of exercise and cognitive function and the relationship with behavioural scores. Future research should consider larger sample sizes in addition to including both healthy and concussed groups.        	70	Chapter 7: Conclusion  This study explored the cortical activity associated with the completion of a novel working memory cognitive task, an exercise task and a combined task in healthy young adults. Our results indicate that combining graded exercise and cognition results in significant changes in both behaviour and cortical activity. When compared to either the Exercise - only or Cognitive – only task condition, the Combined Cog-Ex condition results in significant overall changes in brain and behaviour that is observable in EEG signal pattern, EEG power output, and in local functional connectivity metrics within the frontal regions of the brain. These results provide new information regarding the impact exercise has on neural activity and could have applications in future studies involved in concussion recovery.         	71	References 1. Harmon, K.G., et al., American Medical Society for Sports Medicine position statement: concussion in sport. Clin J Sport Med, 2013. 23(1): p. 1-18. 2. Malec, J.F., et al., The mayo classification system for traumatic brain injury severity. J Neurotrauma, 2007. 24(9): p. 1417-24. 3. Meehan, W.P., 3rd, P. d'Hemecourt, and R.D. Comstock, High school concussions in the 2008-2009 academic year: mechanism, symptoms, and management. American Journal of Sports Medicine, 2010. 38(12): p. 2405-9. 4. Nuwer, M.R., et al., Routine and quantitative EEG in mild traumatic brain injury. Clin Neurophysiol, 2005. 116(9): p. 2001-25. 5. Porter, S., et al., The Child Sport Concussion Assessment Tool (Child SCAT3): normative values and correspondence between child and parent symptom scores in male child athletes. BMJ Open Sport Exerc Med, 2015. 1(1): p. e000029. 6. McCrea, M., et al., Acute effects and recovery time following concussion in collegiate football players: the NCAA Concussion Study. JAMA, 2003. 290(19): p. 2556-63. 7. Prichep, L.S., et al., Time course of clinical and electrophysiological recovery after sport-related concussion. Journal of Head Trauma Rehabilitation, 2013. 28(4): p. 266-73. 8. Jantzen, K.J., et al., A prospective functional MR imaging study of mild traumatic brain injury in college football players. AJNR Am J Neuroradiol, 2004. 25(5): p. 738-45. 9. Chen, J.K., et al., Recovery from mild head injury in sports: evidence from serial functional magnetic resonance imaging studies in male athletes. Clin J Sport Med, 2008. 18(3): p. 241-7. 10. Covassin, T., et al., Immediate post-concussion assessment and cognitive testing (ImPACT) practices of sports medicine professionals. Journal of Athletic Training, 2009. 44(6): p. 639-44. 11. Fazio, V.C., et al., The relation between post concussion symptoms and neurocognitive performance in concussed athletes. NeuroRehabilitation, 2007. 22(3): p. 207-16. 12. Resch, J.E., M.A. McCrea, and C.M. Cullum, Computerized neurocognitive testing in the management of sport-related concussion: an update. Neuropsychol Rev, 2013. 23(4): p. 335-49. 13. Katayama, Y., et al., Massive increases in extracellular potassium and the indiscriminate release of glutamate following concussive brain injury. J Neurosurg, 1990. 73(6): p. 889-900. 14. Yoshino, A., et al., Dynamic changes in local cerebral glucose utilization following cerebral conclusion in rats: evidence of a hyper- and subsequent hypometabolic state. Brain Res, 1991. 561(1): p. 106-19. 15. Bartnik-Olson, B.L., et al., Insights into the metabolic response to traumatic brain injury as revealed by (13)C NMR spectroscopy. Front Neuroenergetics, 2013. 5: p. 8. 16. Giza, C.C. and D.A. Hovda, The Neurometabolic Cascade of Concussion. J Athl Train, 2001. 36(3): p. 228-235. 17. Hovda, D.A., et al., Diffuse prolonged depression of cerebral oxidative metabolism following concussive brain injury in the rat: a cytochrome oxidase histochemistry study. Brain Res, 1991. 567(1): p. 1-10. 18. Shaw, N.A., The neurophysiology of concussion. Prog Neurobiol, 2002. 67(4): p. 281-344. 19. Jalloh, I., et al., Glucose metabolism following human traumatic brain injury: methods of assessment and pathophysiological findings. Metab Brain Dis, 2015. 30(3): p. 615-32. 20. Hovda, D.A., The neurophysiology of concussion. Prog Neurol Surg, 2014. 28: p. 28-37. 21. Lipton, M.L., et al., Diffusion-tensor imaging implicates prefrontal axonal injury in executive function impairment following very mild traumatic brain injury. Radiology, 2009. 252(3): p. 816-24. 22. Eierud, C., et al., Neuroimaging after mild traumatic brain injury: review and meta-analysis. Neuroimage Clin, 2014. 4: p. 283-94.     	72	23. Chen, J.K., et al., A validation of the post concussion symptom scale in the assessment of complex concussion using cognitive testing and functional MRI. J Neurol Neurosurg Psychiatry, 2007. 78(11): p. 1231-8. 24. Mayer, A.R., et al., Functional connectivity in mild traumatic brain injury. Hum Brain Mapp, 2011. 32(11): p. 1825-35. 25. McAllister, T.W., Neurobiological consequences of traumatic brain injury. Dialogues Clin Neurosci, 2011. 13(3): p. 287-300. 26. Thompson, J., W. Sebastianelli, and S. Slobounov, EEG and postural correlates of mild traumatic brain injury in athletes. Neurosci Lett, 2005. 377(3): p. 158-63. 27. McCrea, M., et al., Acute effects and recovery after sport-related concussion: a neurocognitive and quantitative brain electrical activity study. J Head Trauma Rehabil, 2010. 25(4): p. 283-92. 28. Balkan, O., et al., Source-domain spectral EEG analysis of sports-related concussion via Measure Projection Analysis. Conf Proc IEEE Eng Med Biol Soc, 2015. 2015: p. 4053-6. 29. Virji-Babul, N., et al., Changes in Functional Brain Networks following Sports-Related Concussion in Adolescents. J Neurotrauma, 2014. 30. McCrory, P., et al., Consensus statement on concussion in sport: the 4th International Conference on Concussion in Sport held in Zurich, November 2012. Br J Sports Med, 2013. 47(5): p. 250-8. 31. Thomas, D.G., et al., Benefits of strict rest after acute concussion: a randomized controlled trial. Pediatrics, 2015. 135(2): p. 213-23. 32. Silverberg, N.D. and G.L. Iverson, Is rest after concussion "the best medicine?": recommendations for activity resumption following concussion in athletes, civilians, and military service members. J Head Trauma Rehabil, 2013. 28(4): p. 250-9. 33. Brolinson, P.G., Management of sport-related concussion: a review. Clin J Sport Med, 2014. 24(1): p. 89-90. 34. McCrory, P., et al., Summary and agreement statement of the 2nd International Conference on Concussion in Sport, Prague 2004. Br J Sports Med, 2005. 39(4): p. 196-204. 35. McCrory, P., G. Davis, and M. Makdissi, Second impact syndrome or cerebral swelling after sporting head injury. Curr Sports Med Rep, 2012. 11(1): p. 21-3. 36. Cantu, R.C. and A.D. Gean, Second-impact syndrome and a small subdural hematoma: an uncommon catastrophic result of repetitive head injury with a characteristic imaging appearance. J Neurotrauma, 2010. 27(9): p. 1557-64. 37. McKee, A.C., et al., Chronic traumatic encephalopathy in athletes: progressive tauopathy after repetitive head injury. J Neuropathol Exp Neurol, 2009. 68(7): p. 709-35. 38. Lovell, M.R., et al., Measurement of symptoms following sports-related concussion: reliability and normative data for the post-concussion scale. Appl Neuropsychol, 2006. 13(3): p. 166-74. 39. Meier, T.B., et al., The underreporting of self-reported symptoms following sports-related concussion. J Sci Med Sport, 2015. 18(5): p. 507-11. 40. Lee, H., S.J. Sullivan, and A.G. Schneiders, Does a standardised exercise protocol incorporating a cognitive task provoke postconcussion-like symptoms in healthy individuals? J Sci Med Sport, 2015. 18(3): p. 245-9. 41. Balasundaram, A.P., et al., Symptom response following acute bouts of exercise in concussed and non-concussed individuals - a systematic narrative review. Phys Ther Sport, 2013. 14(4): p. 253-8. 42. Alla, S., et al., Does exercise evoke neurological symptoms in healthy subjects? J Sci Med Sport, 2010. 13(1): p. 24-6. 43. McGrath, N., et al., Post-exertion neurocognitive test failure among student-athletes following concussion. Brain Injury, 2013. 27(1): p. 103-13. 44. Niedermeyer, E. and F.H. Lopes da Silva, Electroencephalography : basic principles, clinical applications, and related fields. 4th ed. 1999, Philadelphia: Lippincott Williams & Wilkins. xi, 1258 p., 8 p. of plates.     	73	45. Olejniczak, P., Neurophysiologic basis of EEG. J Clin Neurophysiol, 2006. 23(3): p. 186-9. 46. Woodman, G.F., A brief introduction to the use of event-related potentials in studies of perception and attention. Atten Percept Psychophys, 2010. 72(8): p. 2031-46. 47. Cooper, R., et al., Comparison of Subcortical, Cortical and Scalp Activity Using Chronically Indwelling Electrodes in Man. Electroencephalogr Clin Neurophysiol, 1965. 18: p. 217-28. 48. Koenig, T., Basic Principles of EEG and MEG Analysis. 2014. 49. Heinze, H.J., et al., Combined spatial and temporal imaging of brain activity during visual selective attention in humans. Nature, 1994. 372(6506): p. 543-6. 50. Vogel, E.K. and S.J. Luck, The visual N1 component as an index of a discrimination process. Psychophysiology, 2000. 37(2): p. 190-203. 51. Grech, R., et al., Review on solving the inverse problem in EEG source analysis. J Neuroeng Rehabil, 2008. 5: p. 25. 52. Michel, C.M., et al., EEG source imaging. Clin Neurophysiol, 2004. 115(10): p. 2195-222. 53. Scherg, M., et al., Advanced tools for digital EEG review: virtual source montages, whole-head mapping, correlation, and phase analysis. J Clin Neurophysiol, 2002. 19(2): p. 91-112. 54. Freeman, W.J., R. Quian Quiroga, and SpringerLink ebooks - Biomedical and Life Sciences (2013), Imaging brain function with EEG advanced temporal and spatial analysis of electroencephalographic signals. 2013, Springer,: New York. p. 1 online resource. 55. Meltzer, J.A., et al., Effects of working memory load on oscillatory power in human intracranial EEG. Cereb Cortex, 2008. 18(8): p. 1843-55. 56. Scheeringa, R., et al., Trial-by-trial coupling between EEG and BOLD identifies networks related to alpha and theta EEG power increases during working memory maintenance. Neuroimage, 2009. 44(3): p. 1224-38. 57. Knyazev, G.G., J.Y. Slobodskoj-Plusnin, and A.V. Bocharov, Event-related delta and theta synchronization during explicit and implicit emotion processing. Neuroscience, 2009. 164(4): p. 1588-600. 58. Knyazev, G.G., Motivation, emotion, and their inhibitory control mirrored in brain oscillations. Neurosci Biobehav Rev, 2007. 31(3): p. 377-95. 59. Jensen, O. and C.D. Tesche, Frontal theta activity in humans increases with memory load in a working memory task. Eur J Neurosci, 2002. 15(8): p. 1395-9. 60. Mitchell, D.J., et al., Frontal-midline theta from the perspective of hippocampal "theta". Prog Neurobiol, 2008. 86(3): p. 156-85. 61. Palva, S. and J.M. Palva, New vistas for alpha-frequency band oscillations. Trends Neurosci, 2007. 30(4): p. 150-8. 62. Engel, A.K. and P. Fries, Beta-band oscillations--signalling the status quo? Curr Opin Neurobiol, 2010. 20(2): p. 156-65. 63. Siegel, M., M.R. Warden, and E.K. Miller, Phase-dependent neuronal coding of objects in short-term memory. Proc Natl Acad Sci U S A, 2009. 106(50): p. 21341-6. 64. Herrmann, C.S., M.H. Munk, and A.K. Engel, Cognitive functions of gamma-band activity: memory match and utilization. Trends Cogn Sci, 2004. 8(8): p. 347-55. 65. Herrmann, C.S. and T. Demiralp, Human EEG gamma oscillations in neuropsychiatric disorders. Clin Neurophysiol, 2005. 116(12): p. 2719-33. 66. Aiyer, R., V. Novakovic, and R.L. Barkin, A systematic review on the impact of psychotropic drugs on electroencephalogram waveforms in psychiatry. Postgrad Med, 2016. 128(7): p. 656-64. 67. Weiss, B. and V.G. Laties, Enhancement of human performance by caffeine and the amphetamines. Pharmacol Rev, 1962. 14: p. 1-36. 68. Siepmann, M. and W. Kirch, Effects of caffeine on topographic quantitative EEG. Neuropsychobiology, 2002. 45(3): p. 161-6. 69. Knott, V., et al., EEG power, frequency, asymmetry and coherence in male depression. Psychiatry Res, 2001. 106(2): p. 123-40.     	74	70. Knyazev, G.G., A.N. Savostyanov, and E.A. Levin, Alpha oscillations as a correlate of trait anxiety. Int J Psychophysiol, 2004. 53(2): p. 147-60. 71. Snyder, S.M. and J.R. Hall, A meta-analysis of quantitative EEG power associated with attention-deficit hyperactivity disorder. J Clin Neurophysiol, 2006. 23(5): p. 440-55. 72. Bressler, S.L., Large-scale cortical networks and cognition. Brain Res Brain Res Rev, 1995. 20(3): p. 288-304. 73. Mesulam, M.M., From sensation to cognition. Brain, 1998. 121 ( Pt 6): p. 1013-52. 74. Rubinov, M. and O. Sporns, Complex network measures of brain connectivity: uses and interpretations. Neuroimage, 2010. 52(3): p. 1059-69. 75. He, Y. and A. Evans, Graph theoretical modeling of brain connectivity. Curr Opin Neurol, 2010. 23(4): p. 341-50. 76. Bullmore, E. and O. Sporns, Complex brain networks: graph theoretical analysis of structural and functional systems. Nat Rev Neurosci, 2009. 10(3): p. 186-98. 77. Stam, C.J., et al., Graph theoretical analysis of magnetoencephalographic functional connectivity in Alzheimer's disease. Brain, 2009. 132(Pt 1): p. 213-24. 78. Latora, V. and M. Marchiori, Efficient behavior of small-world networks. Phys Rev Lett, 2001. 87(19): p. 198701. 79. Elliott, R., Executive functions and their disorders. Br Med Bull, 2003. 65: p. 49-59. 80. Hallett, P.E., Primary and secondary saccades to goals defined by instructions. Vision Res, 1978. 18(10): p. 1279-96. 81. Everling, S. and B. Fischer, The antisaccade: a review of basic research and clinical studies. Neuropsychologia, 1998. 36(9): p. 885-99. 82. Munoz, D.P. and S. Everling, Look away: the anti-saccade task and the voluntary control of eye movement. Nat Rev Neurosci, 2004. 5(3): p. 218-28. 83. McDowell, J.E., et al., Neurophysiology and neuroanatomy of reflexive and volitional saccades: evidence from studies of humans. Brain Cogn, 2008. 68(3): p. 255-70. 84. Hamm, J.P., D. Sabatinelli, and B.A. Clementz, Alpha oscillations and the control of voluntary saccadic behavior. Exp Brain Res, 2012. 221(2): p. 123-8. 85. Gevins, A. and M.E. Smith, Neurophysiological measures of working memory and individual differences in cognitive ability and cognitive style. Cereb Cortex, 2000. 10(9): p. 829-39. 86. Baars, B.J. and N.M. Gage, Cognition, brain, and consciousness: Introduction to cognitive neuroscience. 2010: Academic Press. 87. Baddeley, A., Working memory. Science, 1992. 255(5044): p. 556-9. 88. Baddeley, A.D. and G. Hitch, Working memory. Psychology of learning and motivation, 1974. 8: p. 47-89. 89. Marrocco, R.T. and M.C. Davidson, Neurochemistry of attention. 1998. 90. Gould, R.L., et al., fMRI BOLD response to increasing task difficulty during successful paired associates learning. Neuroimage, 2003. 20(2): p. 1006-19. 91. Onton, J., A. Delorme, and S. Makeig, Frontal midline EEG dynamics during working memory. Neuroimage, 2005. 27(2): p. 341-56. 92. Ku, Y., et al., Sequential roles of primary somatosensory cortex and posterior parietal cortex in tactile-visual cross-modal working memory: a single-pulse transcranial magnetic stimulation (spTMS) study. Brain Stimulation, 2015. 8(1): p. 88-91. 93. Maurer, U., et al., Frontal midline theta reflects individual task performance in a working memory task. Brain Topography, 2015. 28(1): p. 127-34. 94. Gevins, A., et al., High-resolution EEG mapping of cortical activation related to working memory: effects of task difficulty, type of processing, and practice. Cereb Cortex, 1997. 7(4): p. 374-85. 95. Sauseng, P., et al., Control mechanisms in working memory: a possible function of EEG theta oscillations. Neurosci Biobehav Rev, 2010. 34(7): p. 1015-22.     	75	96. Ishii, R., et al., Medial prefrontal cortex generates frontal midline theta rhythm. Neuroreport, 1999. 10(4): p. 675-9. 97. Gevins, A., et al., Monitoring working memory load during computer-based tasks with EEG pattern recognition methods. Hum Factors, 1998. 40(1): p. 79-91. 98. Gevins, A., et al., Electroencephalographic imaging of higher brain function. Philos Trans R Soc Lond B Biol Sci, 1999. 354(1387): p. 1125-33. 99. Ku, Y., et al., Spectra-temporal patterns underlying mental addition: an ERP and ERD/ERS study. Neurosci Lett, 2010. 472(1): p. 5-10. 100. Haenschel, C., et al., Gamma and beta frequency oscillations in response to novel auditory stimuli: A comparison of human electroencephalogram (EEG) data with in vitro models. Proc Natl Acad Sci U S A, 2000. 97(13): p. 7645-50. 101. Kukleta, M., et al., Beta 2-band synchronization during a visual oddball task. Physiol Res, 2009. 58(5): p. 725-32. 102. Ishii, R., et al., Cortical oscillatory power changes during auditory oddball task revealed by spatially filtered magnetoencephalography. Clin Neurophysiol, 2009. 120(3): p. 497-504. 103. Guntekin, B., et al., Beta oscillatory responses in healthy subjects and subjects with mild cognitive impairment. Neuroimage Clin, 2013. 3: p. 39-46. 104. Jensen, O., J. Kaiser, and J.P. Lachaux, Human gamma-frequency oscillations associated with attention and memory. Trends Neurosci, 2007. 30(7): p. 317-24. 105. Pesaran, B., et al., Temporal structure in neuronal activity during working memory in macaque parietal cortex. Nat Neurosci, 2002. 5(8): p. 805-11. 106. Tallon-Baudry, C., et al., Induced gamma-band activity during the delay of a visual short-term memory task in humans. J Neurosci, 1998. 18(11): p. 4244-54. 107. Howard, M.W., et al., Gamma oscillations correlate with working memory load in humans. Cerebral cortex, 2003. 13(12): p. 1369-1374. 108. Zhang, D., et al., Functional connectivity among multi-channel EEGs when working memory load reaches the capacity. Brain Res, 2016. 1631: p. 101-12. 109. Womelsdorf, T., et al., Modulation of neuronal interactions through neuronal synchronization. Science, 2007. 316(5831): p. 1609-12. 110. Payne, L. and J. Kounios, Coherent oscillatory networks supporting short-term memory retention. Brain Res, 2009. 1247: p. 126-32. 111. Palva, S., S. Monto, and J.M. Palva, Graph properties of synchronized cortical networks during visual working memory maintenance. Neuroimage, 2010. 49(4): p. 3257-68. 112. Palva, J.M., et al., Neuronal synchrony reveals working memory networks and predicts individual memory capacity. Proc Natl Acad Sci U S A, 2010. 107(16): p. 7580-5. 113. Li, L., J.X. Zhang, and T. Jiang, Visual working memory load-related changes in neural activity and functional connectivity. PLoS One, 2011. 6(7): p. e22357. 114. Fukuda, K., E. Awh, and E.K. Vogel, Discrete capacity limits in visual working memory. Curr Opin Neurobiol, 2010. 20(2): p. 177-82. 115. Luck, S.J. and E.K. Vogel, Visual working memory capacity: from psychophysics and neurobiology to individual differences. Trends Cogn Sci, 2013. 17(8): p. 391-400. 116. Cotman, C.W., N.C. Berchtold, and L.A. Christie, Exercise builds brain health: key roles of growth factor cascades and inflammation. Trends Neurosci, 2007. 30(9): p. 464-72. 117. Eadie, B.D., V.A. Redila, and B.R. Christie, Voluntary exercise alters the cytoarchitecture of the adult dentate gyrus by increasing cellular proliferation, dendritic complexity, and spine density. J Comp Neurol, 2005. 486(1): p. 39-47. 118. Vaynman, S.S., et al., Exercise differentially regulates synaptic proteins associated to the function of BDNF. Brain Res, 2006. 1070(1): p. 124-30. 119. Farmer, J., et al., Effects of voluntary exercise on synaptic plasticity and gene expression in the dentate gyrus of adult male Sprague-Dawley rats in vivo. Neuroscience, 2004. 124(1): p. 71-9.     	76	120. Berchtold, N.C., et al., Exercise primes a molecular memory for brain-derived neurotrophic factor protein induction in the rat hippocampus. Neuroscience, 2005. 133(3): p. 853-61. 121. Trejo, J.L., E. Carro, and I. Torres-Aleman, Circulating insulin-like growth factor I mediates exercise-induced increases in the number of new neurons in the adult hippocampus. J Neurosci, 2001. 21(5): p. 1628-34. 122. van Praag, H., et al., Exercise enhances learning and hippocampal neurogenesis in aged mice. J Neurosci, 2005. 25(38): p. 8680-5. 123. Schweitzer, N.B., et al., Exercise-induced changes in cardiac gene expression and its relation to spatial maze performance. Neurochem Int, 2006. 48(1): p. 9-16. 124. Radak, Z., et al., The effects of training and detraining on memory, neurotrophins and oxidative stress markers in rat brain. Neurochem Int, 2006. 49(4): p. 387-92. 125. O'Callaghan, R.M., R. Ohle, and A.M. Kelly, The effects of forced exercise on hippocampal plasticity in the rat: A comparison of LTP, spatial- and non-spatial learning. Behav Brain Res, 2007. 176(2): p. 362-6. 126. Enders, H., et al., Changes in cortical activity measured with EEG during a high-intensity cycling exercise. Journal of Neurophysiology, 2016. 115(1): p. 379-88. 127. Gwin, J.T., et al., Electrocortical activity is coupled to gait cycle phase during treadmill walking. Neuroimage, 2011. 54(2): p. 1289-96. 128. Taniguchi, M., et al., Movement-related desynchronization of the cerebral cortex studied with spatially filtered magnetoencephalography. Neuroimage, 2000. 12(3): p. 298-306. 129. Baumeister, J., et al., Cortical activity of skilled performance in a complex sports related motor task. Eur J Appl Physiol, 2008. 104(4): p. 625-31. 130. Doppelmayr, M., T. Finkenzeller, and P. Sauseng, Frontal midline theta in the pre-shot phase of rifle shooting: differences between experts and novices. Neuropsychologia, 2008. 46(5): p. 1463-7. 131. Bland, B.H., The physiology and pharmacology of hippocampal formation theta rhythms. Prog Neurobiol, 1986. 26(1): p. 1-54. 132. Cruikshank, L.C., et al., Theta oscillations reflect a putative neural mechanism for human sensorimotor integration. J Neurophysiol, 2012. 107(1): p. 65-77. 133. Ofori, E., S.A. Coombes, and D.E. Vaillancourt, 3D Cortical electrophysiology of ballistic upper limb movement in humans. Neuroimage, 2015. 115: p. 30-41. 134. Crabbe, J.B. and R.K. Dishman, Brain electrocortical activity during and after exercise: a quantitative synthesis. Psychophysiology, 2004. 41(4): p. 563-74. 135. Bailey, S.P., et al., Changes in EEG during graded exercise on a recumbent cycle ergometer. Journal of Sports Science & Medicine, 2008. 7(4): p. 505-11. 136. Fumoto, M., et al., Ventral prefrontal cortex and serotonergic system activation during pedaling exercise induces negative mood improvement and increased alpha band in EEG. Behav Brain Res, 2010. 213(1): p. 1-9. 137. Kubitz, K.A. and A.A. Mott, EEG power spectral densities during and after cycle ergometer exercise. Research Quarterly for Exercise & Sport, 1996. 67(1): p. 91-6. 138. Brummer, V., et al., Brain cortical activity is influenced by exercise mode and intensity. Med Sci Sports Exerc, 2011. 43(10): p. 1863-72. 139. Moraes, H., et al., Beta and alpha electroencephalographic activity changes after acute exercise. Arquivos de Neuro-Psiquiatria, 2007. 65(3A): p. 637-41. 140. Rajab, A.S., et al., A single session of exercise increases connectivity in sensorimotor-related brain networks: a resting-state fMRI study in young healthy adults. Front Hum Neurosci, 2014. 8: p. 625. 141. Dietrich, A., Functional neuroanatomy of altered states of consciousness: the transient hypofrontality hypothesis. Conscious Cogn, 2003. 12(2): p. 231-56. 142. Davranche, K. and T. McMorris, Specific effects of acute moderate exercise on cognitive control. Brain Cogn, 2009. 69(3): p. 565-70.     	77	143. Etnier, J.L. and Y.K. Chang, The effect of physical activity on executive function: a brief commentary on definitions, measurement issues, and the current state of the literature. J Sport Exerc Psychol, 2009. 31(4): p. 469-83. 144. Wang, C.C., et al., Executive function during acute exercise: the role of exercise intensity. J Sport Exerc Psychol, 2013. 35(4): p. 358-67. 145. Chang, Y.K., et al., The effects of acute exercise on cognitive performance: a meta-analysis. Brain Res, 2012. 1453: p. 87-101. 146. Lambourne, K. and P. Tomporowski, The effect of exercise-induced arousal on cognitive task performance: a meta-regression analysis. Brain Res, 2010. 1341: p. 12-24. 147. Al-Yahya, E., et al., Cognitive motor interference while walking: a systematic review and meta-analysis. Neurosci Biobehav Rev, 2011. 35(3): p. 715-28. 148. Li, L., et al., Acute aerobic exercise increases cortical activity during working memory: a functional MRI study in female college students. PLoS ONE [Electronic Resource], 2014. 9(6): p. e99222. 149. Pontifex, M.B. and C.H. Hillman, Neuroelectric and behavioral indices of interference control during acute cycling. Clin Neurophysiol, 2007. 118(3): p. 570-80. 150. Olson, R.L., et al., Neurophysiological and behavioral correlates of cognitive control during low and moderate intensity exercise. Neuroimage, 2016. 131: p. 171-80. 151. Hogan, M., et al., The interactive effects of physical fitness and acute aerobic exercise on electrophysiological coherence and cognitive performance in adolescents. Experimental Brain Research, 2013. 229(1): p. 85-96. 152. Hutton, S.B., Cognitive control of saccadic eye movements. Brain Cogn, 2008. 68(3): p. 327-40. 153. Hamm, J.P., et al., Pre-cue fronto-occipital alpha phase and distributed cortical oscillations predict failures of cognitive control. Journal of Neuroscience, 2012. 32(20): p. 7034-41. 154. Tombaugh, T.N., A comprehensive review of the Paced Auditory Serial Addition Test (PASAT). Arch Clin Neuropsychol, 2006. 21(1): p. 53-76. 155. Fos, L.A., et al., Paced Visual Serial Addition Test: an alternative measure of information processing speed. Appl Neuropsychol, 2000. 7(3): p. 140-6. 156. Lazeron, R.H., et al., A paced visual serial addition test for fMRI. J Neurol Sci, 2003. 213(1-2): p. 29-34. 157. Oostenveld, R. and P. Praamstra, The five percent electrode system for high-resolution EEG and ERP measurements. Clin Neurophysiol, 2001. 112(4): p. 713-9. 158. Li, J., Z. Wang, and M.J. McKeown, Learning brain connectivity with the false-discovery-rate-controlled PC-algorithm. Conf Proc IEEE Eng Med Biol Soc, 2008. 2008: p. 4617-20. 159. Li, J. and Z. Wang, Controlling the false discovery rate of the association/causality structure learned with the PC algorithm. Journal of Machine Learning Research, 2009. 10(2): p. 475-514. 160. Blatter, K. and C. Cajochen, Circadian rhythms in cognitive performance: methodological constraints, protocols, theoretical underpinnings. Physiol Behav, 2007. 90(2-3): p. 196-208. 161. Golubic, R., et al., Validity of electronically administered Recent Physical Activity Questionnaire (RPAQ) in ten European countries. PLoS One, 2014. 9(3): p. e92829. 162. West, S.G., J.F. Finch, and P.J. Curran, Structural equation models with nonnormal variables: Problems and remedies., in Structural equation modeling: Concepts, issues, and applications, R.H. Hoyle, Editor. 1995, Sage Publications, Inc: Thousand Oaks, CA, US. p. 56 - 75. 163. Godin, G. and R.J. Shephard, A simple method to assess exercise behavior in the community. Can J Appl Sport Sci, 1985. 10(3): p. 141-6.     	78	Appendices Appendix A: Godin Leisure-Time Exercise Questionnaire      	79	Appendix B: Study Recruitment Flyer      	80	Appendix C: Study Consent Form      	81	      	82	      	83	      	84	      	85	      	86	Appendix D: Recent Physical Activity Questionnaire (RPAQ)      	87	     	88	     	89	     	90	     	91	     	92	     	93	      	94	Appendix E: Borg Scale  Rating Description   6 NO EXERTION AT ALL 7  8 EXTREMELY LIGHT 9  10 VERY LIGHT 11  12 LIGHT 13  14 SOMEWHAT HARD 15  16 HARD (HEAVY) 17  18 VERY HARD 19 EXTREMELY HARD 20 MAXIMAL EXERTION    

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.24.1-0342970/manifest

Comment

Related Items