exercise and cognitive function · 2016-08-12 · 1.1 varieties of exercise effects on...

Exercise and Cognitive Function

Editors

Terry McMorrisUniversity of Chichester, UK

Phillip D. TomporowskiUniversity of Georgia, Athens, USA

Michel AudiffrenUniversity of Poitiers, France


Editors

Terry McMorrisUniversity of Chichester, UK

Phillip D. TomporowskiUniversity of Georgia, Athens, USA

Michel AudiffrenUniversity of Poitiers, France

This edition first published 2009, # 2009 by John Wiley & Sons, Ltd

Wiley-Blackwell is an imprint of John Wiley & Sons, formed by the merger of Wiley’s global Scientific,Technical and Medical business with Blackwell Publishing.

Registered office: John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

Other Editorial Offices:9600 Garsington Road, Oxford, OX4 2DQ, UK

111 River Street, Hoboken, NJ 07030-5774, USA

For details of our global editorial offices, for customer services and for information about how to apply forpermission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell

The right of the author to be identified as the author of this work has been asserted in accordance with theCopyright, Designs and Patents Act 1988.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, inany form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted bythe UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher.

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not beavailable in electronic books.

Designations used by companies to distinguish their products are often claimed as trademarks. All brand namesand product names used in this book are trade names, service marks, trademarks or registered trademarks of theirrespective owners. The publisher is not associated with any product or vendor mentioned in this book. Thispublication is designed to provide accurate and authoritative information in regard to the subject matter covered.It is sold on the understanding that the publisher is not engaged in rendering professional services. If professionaladvice or other expert assistance is required, the services of a competent professional should be sought.

Library of Congress Cataloguing-in-Publication Data

Exercise and cognitive function / edited by Terry McMorris, Phillip D. Tomporowski, Michel Audiffren.p. cm.

Includes bibliographical references and index.ISBN 978-0-470-51660-7 (cloth)1. Cognition–Effect of exercise on. 2. Exercise–Psychological aspects. I. McMorris, Terry. II. Tomporowski,Phillip D., 1948- III. Audiffren, Michel.

BF311.E877 2009612.8’233–dc22 2008046669

ISBN: 978-0-470-51660-7

A catalogue record for this book is available from the British Library.

Typeset in 10.5/12.5pt Minion by Thomson Digital, Noida, India.Printed in Great Britain by Antony Rowe Ltd., Chippenham, WiltshireFirst impression 2009

www.wiley.com/wiley-blackwell

www.wiley.com/wiley-blackwell

Contents

Preface ix

Contributors xi

PART 1 THEORETICAL AND METHODOLOGICAL ISSUES 1

1 Acute exercise and psychological functions: a cognitive-energeticapproach 3Michel Audiffren

1.1 Varieties of exercise effects on psychological variables 41.2 The cognitive psychology approach 91.3 The energetic approach 111.4 Exercise effects and cognitive-energetic models 141.5 Sensorimotor and cognitive functions affected by exercise 241.6 Limits of the cognitive-energetic approach and future perspectives 331.7 Conclusion 39

2 Exercise and cognitive function: a neuroendocrinological explanation 41Terry McMorris

2.1 Catecholamines and 5-hydroxytryptamine as brain neurotransmitters 412.2 How exercise induces increases in brain concentrations of

noradrenaline, dopamine, cortisol and 5-hydroxytryptamine 432.3 Exercise, catecholamines, cortisol and cognition: research 502.4 Task type 592.5 Discussion 632.6 Developing a neuroendocrinological model for an interaction between

exercise and cognition 67

3 The transient hypofrontality theory and its implicationsfor emotion and cognition 69Arne Dietrich

3.1 Clearing the ground 713.2 Exercise-induced transient hypofrontality 73

3.3 Implications for emotion 793.4 Implications for cognition 813.5 Reconceptualizing the existing data in the field 87

4 Methodological issues: research approaches, research design,and task selection 91Phillip D. Tomporowski

4.1 Research approaches 924.2 Research design issues 994.3 Task selection issues 1064.4 Conclusions and recommendations 112

PART 2 ACUTE EXERCISE AND COGNITION 115

5 Exercise, dehydration and cognitive function 117Terry McMorris

5.1 Exercise-induced dehydration and cognitive function 1195.2 Discussion 1285.3 Conclusions 134

6 Exercise, nutrition and cognition 135Adam David Cunliffe and Gulshanara Begum

6.1 Fatigue and limits to human performance 1366.2 Assessing the effects of exercise and nutrition on cognitive

performance 1386.3 Nutrition, exercise and cognitive performance 1426.4 Micronutrients, exercise and cognitive performance 1456.5 Nutritional ergogenic aids and cognitive performance 1476.6 Integration of research observations 1486.7 Challenges in research 1506.8 Conclusion 151

7 A chronometric and electromyographic approach to the effectof exercise on reaction time 153Karen Davranche and Michel Audiffren

7.1 Research 1567.2 Conclusion 159

8 Acute aerobic exercise effects on event-related brain potentials 161Charles H. Hillman, Matthew Pontifex and Jason R. Themanson

8.1 Executive control 1638.2 Neuroelectric measurement 1648.3 Event-related brain potentials during exercise 165

vi CONTENTS

8.4 Event-related brain potentials following exercise 1708.5 Future directions and conclusions 177

9 Exercise and decision-making in team games 179Terry McMorris

9.1 Designing a decision-making test 1809.2 Research results 1839.3 Ecological validity and future research 1899.4 Implications for team games players and coaches 192

10 Blood glucose and brain metabolism in exercise 193Niels H. Secher, Thomas Seifert, Henning B. Nielsen and Bjørn Quistorff

10.1 Cerebral metabolism during exercise 19410.2 Cerebral oxygenation 20210.3 Cerebral metabolism 20310.4 Acute hypoglycemia 20910.5 Conclusions 20910.6 Future research 210Acknowledgements 210

PART 3 CHRONIC EXERCISE AND COGNITION 211

11 An integrated approach to the effect of acute and chronic exerciseon cognition: the linked role of individual and task constraints 213Caterina Pesce

11.1 The gap between acute and chronic exercise research 21311.2 Individual constraints on the acute exercise–cognition relationship:

the role of chronic exercise effects 21511.3 Effect of physical fitness: links to exercise intensity and to the time

relation between physical exercise and cognitive task 21811.4 Effect of cognitive expertise: links to cognitive task complexity,

exercise intensity and duration, and age 21911.5 Effect of motor coordination skills: links to physical exercise

complexity, intensity and duration 22311.6 Bridging the gap between acute and chronic exercise studies 225

12 Chronic exercise and cognition in older adults 227Jennifer Etnier

12.1 Theoretical underpinnings 22812.2 Empirical evidence 23012.3 Moderators of the relationship 24512.4 Practical conclusions 24512.5 Challenges 24612.6 Future research 247

viiCONTENTS

13 Exercise and cognition in children 249Catherine L. Davis and Kate Lambourne

13.1 Definition of terms 24913.2 Literature review 25013.3 The Medical College of Georgia study 25413.4 Potential mechanisms 26213.5 Summary and recommendations for future research 266

14 Chronic exercise and developmental disabilities 269James Zagrodnik and Michael Horvat

14.1 Defining terms 26914.2 Research investigating the effects of exercise on cognition

among the developmentally disabled 27214.3 Problems to address and future research considerations 27914.4 Practical applications and conclusions 282

15 Chronic exercise in brain diseases 285Laura Eggermont and Erik Scherder

15.1 Observational studies of physical activity 28615.2 Physical activity intervention studies 28815.3 Physical activity, cognition and different types of dementia 29815.4 Role of vascular disease 30215.5 Neurodegenerative disease, nitric oxide, vascular disease and

physical activity 30515.6 Final conclusion 305

PART 4 DISCUSSION AND CONCLUSION 307

16 Summary and direction for future research 309Terry McMorris, Phillip D. Tomporowski, and Michel Audiffren

16.1 Summary: emerging theoretical approaches 30916.2 Summary of research results 31216.3 Future theoretically driven research 31416.4 Future applied research 31616.5 General summary 317

References 319

Index 375

viii CONTENTS

Preface

Formore years thanwe care to remember, wehave been studying the effect of exerciseon cognitive function. To us it is a fascinating subject and we are delighted to haveseen this field of research blossom in recent years. Themain purpose of this book is tostimulate further research and scholarship in this area. Not only do we need moreresearch but we also require the development of sound theoretical frameworks. Someof you reading this book will already have been involved in research and/ortheorising. However, many will not, but we hope that you will join us. We shouldnote here that the study of exercise and cognitive function also helps us to understandhow other stressors affect cognition and, indeed, provides useful information foranyone interested in how the brain works.

As well as encouraging more research, we are keen to develop inter-disciplinarywork. We believe that this is the only way in which we will increase our knowledge ofthe exercise-cognition interaction. In this book we have drawn together researchersfromdiffering disciplines, e.g. cognitive psychology, cognitive neuroscience, exercisebiochemistry, psychophysiology. In Part 1 we examine theoretical and methodologi-cal aspects. Michel Audiffren begins with an interesting overview of how theory hasdeveloped before outlining the input that we can receive from cognitive-energeticmodels of the arousal--cognition interaction. In Chapter 2 Terry McMorris takes aneuroendocrinological approach, while in Chapter 3 Arne Dietrich puts forward hissomewhat controversial theory of transient hypofrontality. Whether you agree withArne or not, this is a very interesting read and it raises some important factors. In thelast chapter in this section, Phillip Tomporowski discusses themanymethodologicalissues involved in undertaking such research. Phillip’s great experience in this fieldmakes this chapter a ‘must’ for anyone whowishes to undertake research in this area.

Parts 2 and 3 examine research findings with Part Two examining research usingacute exercise and Part 3 looking at chronic exercise. In Part 3 Terry McMorrisexamines research into the effect of exercise and dehydration (Chapter 5) andexercise and performance in team games (Chapter 9). Both are areas of limitedresearch but ones which have practical importance. In Chapter 6 Adam Cunliffe andGulshanara Begum examine another under-researched area, the interaction betweenexercise, nutrition and cognition and they raise some very interesting and practicalissues. In Chapter 7, Karen Davranche and Michel Audiffren provide an overview ofsome of the research that they have undertaken to examine whether exercise affects

motor or premotor aspects of reaction time tasks. The results of their studies, so far,suggest that this is a fruitful area of research. In Chapter 8, Charles Hillman,MatthewPontifex and Jason Themanson from the University of Illinois at Urbana-Champaignexamine research that has been carried out using psychophysiological techniques.The Urbana-Champaign group have provided some very exciting research in thisarea. In the final chapter, Niels Secher, Thomas Seifert, Henning Nielsen and BjørnQuistorff from the University of Copenahgen provide the very latest on the processesaffecting blood glucose and brainmetabolism during exercise. This chapter providesuswith invaluable information concerning the underlying processes thatmight affectthe exercise--cognition interaction.

The first chapter of Part 3, by Caterina Pesce, actually straddles both acute andchronic exercise and, as Caterina says, we researchers should probably take each intoaccount rather than just focusing on the one. In Chapter 12, Jennifer Etnier provides afascinating review of the vast amount of work that has examined chronic exercise andcognition in the elderly. While in Chapter 13, Catherine Davis and Kate Lambourneexamine the other end of the age spectrum looking at children. In particular theyreport on a fascinating study undertaken in Georgia, USA. The final two chaptersexamine exercise and disabilities. In Chapter 14, James Zagrodnik and MichaelHorvat look at research with children, while in Chapter 15 Laura Eggermont and ErikScherder examine exercise and brain diseases. The potential for exercise to be used toprevent diseases such as Alzheimer’s and Parkinson’s is a very exciting one. In Part 4we present an overview of the book and raise some topics for future research.

Acknowledgements

We would like to thank, very much, all of the contributors to this book, who havetaken time out from very busy schedules to provide us with their expertise. Also wegive great thanks to the staff at Wiley-Blackwell in Chichester, especially CeliaCarden, for all their help. We would also like to thank the many individuals from allover theWorld who have given of their time to be subjects in research. It can be veryhard and even when they are paid it is not enough.

Terry McMorrisPhillip D. Tomporowski

Michel Audiffren

x PREFACE

Contributors

Michel Audiffren

Centre de Recherche sur la Cognition etl’ApprentissageUniversit�e de Poitiers, France

Gulshanara Begum

School of BiosciencesUniversity of Westminster, UK

Adam David Cunliffe

School of BiosciencesUniversity of Westminster, UK

Catherine L. Davis

Medical College of GeorgiaAugusta, Georgia, USA

Karen Davranche

Sport, Exercise and Health SciencesResearch CentreUniversity of Chichester, UK

Arne Dietrich

Department of Social and BehavioralSciencesAmerican University of Beirut, Lebanon

Laura Eggermont

Department of Clinical NeurophysiologyVrije Universiteit, Amsterdam,Netherlands

Jennifer Etnier

Department of Exercise and Sport ScienceUniversity of North Carolina,Greensboro, USA

Charles H. Hillman

Department of Kinesiology and CommunityHealthUniversity of Illinois at Urbana-Champaign, USA

Michael Horvat

Department of KinesiologyUniversity of Georgia, Athens, USA

Kate Lambourne

Medical College of GeorgiaGeorgia, USA

Terry McMorris

Sport, Exercise and Health SciencesResearch CentreUniversity of Chichester, UK

Henning B. Nielsen


Caterina Pesce

Institute for Motor ControlUniversity of Rome, Italy

Matthew Pontifex


Bjørn Quistorff

Rigshospitalet and Department ofBiomedical SciencesThe Panum Institute University ofCopenhagen, Denmark

Erik Scherder

Department of Clinical NeurophysiologyVrije Universiteit, Amsterdam,Netherlands

Niels H. Secher


Thomas Seifert


Jason R. Themanson


Phillip D. Tomporowski


James Zagrodnik


xii CONTRIBUTORS

PART 1

THEORETICAL ANDMETHODOLOGICAL ISSUES

1Acute exercise and psychologicalfunctions: a cognitive-energeticapproachMichel Audiffren

Feelings of efficiency or inefficiency are often reported by people practising a mentaltask during or following a physical activity. For instance, athletes know fromexperience that performing a warm-up exercise improves their reactivity to a startsignal; soldiers carrying heavy loads complain about the debilitating effect of centralfatigue on decision making while walking long distances; and the elderly are awarethat regular physical activity improves mental health and autonomy. In spite of thesepopular and professional beliefs, the interactions between the physiological changes,induced by exercise and the psychological functions potentially affected by thesechanges were, and still are, the object of an extended debate. Today, a large number ofscientific studies support a positive effect of exercise on psychological functions (forrecent narrative and meta-analysis reviews see Colcombe and Kramer, 2003; Etnieret al., 2006;McMorris andGraydon, 2000; Tomporowski, 2003b). The purpose of thischapter is to present different theories and methods drawn from cognitive andenergetic approaches that explain the effects of exercise on psychological processes.In addition, wewill attempt to determine the location of these effects in the cognitive-energetic architecture of the information processing system.

The chapter is organized as follows. The first sectionwill be devoted to the proposalof common taxonomies for the effects of exercise on psychological and physiologicalvariables. These taxonomies will guide the choice of adequate physiological inter-ventions, cognitive tasks and theories or models, which can explain the differenteffects of exercise on psychological processes. In the second section, the cognitivepsychology approach and the information processing paradigm will be brieflypresented. This approach provides very powerful methods that allow the study ofthe effect of exercise on psychological functions. The energetic approach will be

Exercise and Cognitive Function, Edited by Terry McMorris, Phillip D. Tomporowski and Michel Audiffren

� 2009 John Wiley & Sons, Ltd

assessed in the third section, as it supplies a very useful theoretical framework fromwhich to understand the acute effects of exercise on cognition. In the fourth section,four cognitive-energetic models explaining the facilitating and debilitating effects ofacute exercise on information processing will be proposed. The fifth section willpresent six methods from cognitive psychology that explain the localization of theeffects of exercise in the architecture of cognitive-energetic systems. Finally, in thesixth section, some limits of the cognitive-energetic approach will be discussed andcomplementary approaches that could give additional information on the mechan-isms underlying the effects of exercise on cognition will be suggested.

1.1 Varieties of exercise effects on psychological variables

The literature on the effects of exercise on psychological functions is abundant. From1960, several hundred experiments have been conducted throughout the world onthis topic. Four kinds of results are generally reported: (1) an improvement inpsychological functions; (2) an impairment of psychological functions; (3) a changein strategy to maintain psychological test performance; and (4) no effects onpsychological functions. The fourth category is problematic because researcherscannot distinguish between two possible reasons for the failure: (1) there is no effectof exercise on psychological functions in the real world or (2) the methodology usedin the experiment was not appropriate to show a significant effect of exercise onpsychological functions. The three other categories of results are themost interestingbecause they lead to new questions: (1) what are the adequate conditions to observethese phenomena (intensity and duration of exercise, time of observation duringor after exercise, age and cardiovascular fitness of participants, etc.); (2) whichphysiological mechanisms explain these positive or negative effects on psychologicalprocesses; and (3) which psychological processes are affected by these mechanisms.The aim of this first section is to propose taxonomies for the interaction betweenexercise and cognitive tasks, which can help scientists to conceptualize futureresearch on the topic.

Aerobic versus anaerobic exercise

Studies that assessed the effects of exercise on psychological functions have used alarge variety of exercise protocols (see Tomporowski and Ellis, 1986, for a narrativereview). Physical exercises performed by the participants can be classified accordingto their mode of progress (constant or incremental load), their intensity andduration, and the two general metabolic pathways that supply the energy to themuscle (aerobic and anaerobic). According to the review by Tomporowski and Ellis,four categories of exercises can be distinguished (see Table 1.1).

The intensity of exercise is generally defined according to the capabilities of theparticipants. Different indices have been used by researchers to determine the samerelative intensity of exercise for all participants: for instance, the percentage of

4 CH1 ACUTE EXERCISE AND PSYCHOLOGICAL FUNCTIONS

maximal oxygen uptake (e.g. Deligni�eres, Brisswalter and Legros, 1994; Travlos andMarisi, 1996), the percentage of maximal power output (e.g. Arcelin, Deligni�eres andBrisswalter, 1998; Hogervorst et al., 1996; McMorris and Graydon, 1996a; Paas andAdam, 1991), the percentage of ventilatory threshold (e.g. Davranche et al., 2005a,Davranche, Audiffren and Denjean, 2006a), the percentage of lactate threshold(Chmura et al., 1998; Kashihara and Nakahara, 2005), the percentage of maximalheart rate (e.g. McGlynn et al., 1979), the percentage of heart rate reserve (e.g. Pesce,Casella and Capranica, 2004), or the rating of perceived exertion (e.g. Kamijo et al.,2004b). In Table 1.1, the intensity of exercise is expressed in percentage of maximalvolume of oxygen uptake (%VO2max), a function of the cardiorespiratory systemreflecting the total amount of oxygen that the individual can utilize. This index iscommonly used in exercise physiology and exercise psychology.

Physiologicalmechanisms providing energy to themuscle and physiological statesinduced by exercise vary considerably from one category of exercise to the other.Very brief and intense exercise, described in the first line of Table 1.1, mainly involvesthe anaerobic metabolism. It may be localized, such as a handgrip at maximal force,or involve the whole body, such as in the 100m sprint. The anaerobic metabolismdoes not require oxygen and yields small quantities of adenosine 50-triphosphate(ATP) per mole of glucose or muscle glycogen. Three separate anaerobic mechan-isms, initiated in parallel but varying markedly in duration, provide ATP to themuscle following the initiation of this type of exercise (Ward-Smith, 1999): (1) a smallquantity of endogenous ATP molecules stored in muscle, which are depleted inapproximately 5 s; (2) the breakdown of intramuscular phosphocreatine (PCr) thatcontributes to the production of ATP and lasts approximately 10 s; and (3) theoxygen-independent glycolysis that supplies predominantly theATP from 10 to 100 s.Blood lactate andHþ ions are the endproducts that are released into the bloodby thisoxygen-independent glycolysis. The availability of phosphagens (ATP and PCr), and

Table 1.1 Taxonomy of physical exercises as a function of duration and intensity of thephysiological intervention.

Mode ofexercise Duration Intensity

Metabolicpathway Example

Constant-loadexercise

Very brief<3 min

Supra-maximal> 100% VO2max

Anaerobic Handgrip atmaximal force

Ramp incrementalexercise

Moderate10 to 25min

Up to maximalUp to 100%VO2max

Aerobic andAnaerobic

VO2max test


Short tomoderate5 to 60min

Sub-maximal30 to 80% VO2max

Aerobic Walking on atreadmill


Long > 60min Sub-maximal30 to 80%VO2max

Aerobic Marathon race

51.1 VARIETIES OF EXERCISE EFFECTS ON PSYCHOLOGICAL VARIABLES

blood acidosis are generally accepted to be two of the most likely limitations toanaerobic muscle performance. Very brief and intense exercise presents two pro-blems for studying the effects of exercise on cognition. On the one hand, due to theshort duration of the exercise, the number of trials recorded during the exercise forthe cognitive task is necessarily small. On the other hand, peripheral fatiguemechanisms take place during this category of exercise. Studies not concerned withthese peripheral fatigue phenomena should choose one of the other categories ofexercise.

Graded exercise, described in the second line of Table 1.1, in which exerciseintensity is progressively increased from light intensity to exhaustion, involvesanaerobic as well as aerobic mechanisms. When exercise is sustained beyond100 s, aerobic metabolism, which requires oxygen and yields large quantities of ATPpermole of glucose, progressively replaces anaerobicmetabolism as themain sourceof energy. In a graded exercise, the aerobicmechanism provides the larger part of theenergy until the anaerobic lactate threshold is reached. Anaerobic lactate thresholdrepresents the critical point atwhichmetabolicmodifications bring about the energy-demand transition from aerobic to anaerobic exercise. Incremental exercise untilexhaustion presents a serious limit for studying the effect of exercise on cognitionbecause the participants’ physiological state changes throughout the exercise. Duringthe first minutes of the exercise, the energy is mainly supplied by aerobic mechan-isms, and there is no fatigue nor any real mental effort required to sustain therelatively low intensity of exercise. By contrast, during the last minutes of this kind ofexercise, the energy ismainly supplied by anaerobic glycolysis, participants generallyfeel peripheral fatigue and a high level of mental effort is required to continue theexercise.

Aerobic exercise of moderate duration, described in the third line of Table 1.1,seems to be the most interesting category of exercise to study the positive effect ofexercise on cognition. There are three main reasons for this: (1) participants performat steady-state throughout the entire exercise session and aerobicmechanisms are themain source of energy soon after the beginning and until the end of exercise; (2) acognitive task involving a large number of trials can be performed simultaneously tothe exercise; and (3) central as well as peripheral fatigue phenomena are limited,which is in contrast to the three other categories of exercise. Finally, aerobic exerciseof long duration, described in the fourth line of Table 1.1, is more interesting forstudying the effects of central fatigue on cognition. The last two categories of exercisegenerally require whole-body activation.

Acute versus chronic exercise

There is another very important distinction concerning the protocols used to studythe effects of exercise on cognition. Both acute and chronic exercise have beenextensively used, but they must be distinguished because they induce differentchanges in the organism (see Table 1.2). While acute exercise concerns itself with asingle bout of exercise, chronic exercise concerns itself with the repetition of bouts of


exercise over time, lasting from weeks to years (Audiffren et al., 2007a). Thebehavioural and psychological changes induced by a single bout of exercise generallyappear quite rapidly after the beginning of exercise (seconds to minutes) anddisappear relatively quickly after its cessation (minutes to hours). Neurophysiologi-cal changes, which underlie the transitory behavioural and psychological changesinduced by exercise, can be viewed as a transient modulation of the activity of theneural networks involved in the cognitive task or the mental state of interest. Incontrast, chronic effects of exercise reflect structural and durable changes in theorganism, like angiogenesis (e.g. Swain et al., 2003), synaptogenesis (e.g. Chu andJones, 2000) or neurogenesis (e.g. van Praag et al., 1999).

The behavioural and psychological changes induced by the regular practice of aphysical activity typically appear a few weeks after the beginning of the exerciseprogramme and can be maintained several weeks after its termination. Neurophysi-ological changes underlying stable behavioural and psychological changes inducedby exercise can be viewed as durable anatomical changes in the brain structure atdifferent possible levels (e.g. neuroreceptor, synapse, neuron, neural network, brainstructure). In this chapter, wewill focus on theories andmodels which can explain theacute effects of exercise on cognitive processes.

Bottom-up versus top-down processes

Psychometric tasks used to examine the acute effects of exercise on psychologicalprocesses involve a large variety of sensorimotor and cognitive processes. Accordingto Paillard (1986, 2005), it is useful to distinguish these two levels of informationprocessing (see Figure 1.1). On the one hand, the sensorimotor level can be conceivedas an interface between the brain and its environment. Ascending informationgathered by sensory organs for further processing through attentional control anddescending commands for self-generated movements require this sensorimotorinterface. The sensorimotor level is mainly genetically pre-wired and supplies vitalfunctions such as stretch reflex and orienting reaction. On the other hand, thecognitive level is underpinned by an apparatus endowed with the whole resourcesof neocortical and limbic structures and able to process a large variety of mentalstates that characterize higher brain functions. The sensorimotor level mainlyfunctions in a reactive way, but possesses its own adaptive loops, whereas the

Table 1.2 Taxonomy of physical exercises as a function of physiological changes they induce.

Type of effectMode ofexercise

Type of physiologicalchange

Type of brain mechanismunderpinning the effect

Acute effect Single bout ofexercise

Transient Modulation of theactivity of a neuralnetwork

Chronic effect Regular exercise Durable Anatomical changes in thebrain structure

71.1 VARIETIES OF EXERCISE EFFECTS ON PSYCHOLOGICAL VARIABLES

cognitive level anticipates events and functions in a predictive way on the basisof abstract representations of internal and external worlds stored in long-termmemory.

Each of these two levels of information processing can be separated into differentmodules, stages or functions (Fodor, 1983; Sternberg, 2001). For instance, informationprocessing may be separated into sensory, perceptual, decisional and motor stages(Sanders, 1983), while executive processes can be separated into three basic functions,inhibitingpre-potent responses,updatingworkingmemoryandshiftingbetweentasksor rules (Miyake et al., 2000). Recent empirical data suggest that steady-state aerobicexercise does not affect all these stages and functions, but influences some of them indifferentways (Audiffren,TomporowskiandZagrodnik,2008;Davrancheetal., 2005a,2006b; Dietrich and Sparling, 2004). Bottom-up stimulus-driven processes (see leftcolumn of Table 1.3) would be improved by acute aerobic exercise, whereas top-downeffortful processes (see right column of Table 1.3) would be impaired. This idea of abi-directional effect of arousing stimulation on information processing is not new, but

Figure 1.1 Sensorimotor and cognitive levels of information processing. (Modified from Paillard,1986, 2005.)

Table 1.3 The hypothetical bidirectional effect of acute bout of steady-state aerobic exercise oncognitive processes.

Improvement of performance Impairment of performance

Bottom-up Top-downStimulus-driven Goal-drivenAutomatic EffortfulImplicit ExplicitUnconscious Conscious

Note: When participants perform a cognitive task while exercising, performance of the cognitive task may beeither improved or impaired. Tasks which tap processes showing several characteristics described in the right

column tend to be improved,whereas taskswhich involve processes presenting several characteristics described

in the left column tend to be impaired (see text for more explanations). Each line describes a bi-dimensional

continuum (e.g. Automatic--Effortful). There is no dimensional overlap between the five bi-dimensionalcontinua. For instance, a top-down process may be totally unconscious, or a stimulus-driven process may

require allocation of effort.


similar to the distinction between sustained information processes and short-termmemory processes made by Humphreys and Revelle (1984) and tested in exerciseprotocols by Paas andAdam, and their co-workers (Adam et al., 1997; Paas andAdam,1991). Currently, there is no general theory providing a synthesis for all these newdata,approachesandhypotheses concerning the effects of acuteboutsof aerobic exerciseoncognitive processes. Several chapters of this book lay the foundation for a newtheoretical framework explaining the multiple interactions between acute exerciseand cognition.

The assessment of the effects of exercise on cognition should include: (1) the choiceof an adequate physiological intervention; (2) the selection of tasks according to thesensorimotor and cognitive processes they involve; and (3) the separation ofcognitive stages and functions within the same task. Different methods worked outby cognitive psychologists and cognitive neuroscientists in order to separate senso-rimotor stages and cognitive functions will be presented later.

1.2 The cognitive psychology approach

Cognitive psychology emerged in the middle of the twentieth century. This branchof psychology is based on the theoretical framework of the cognitive sciences (i.e.artificial intelligence, mathematics, linguistics, philosophy, neurosciences andpsychology). Therefore, mathematics inspired three main ideas of cognitivepsychology (Andler, 1986). The first idea considers the language of the mind asa formal system. A formal system is a set of symbols that can be combined accordingto rules based on the shape of the symbols. Puzzles, alphabets and languages are allexamples of formal systems. Themind uses these symbols to generate thoughts. Thesecond idea considers the mind as a computing machine. The first universalcomputing machine was conceptualized by the English mathematician Alan Tur-ing. A Turing machine is composed of an input/output system that can write andread a finite set of symbols, and of a memory system that can assume a finite set ofstates. According to Turing (1950), a machine could operate in the same way as thehumanmind. Thoughts, then, would simply be a combination of symbols. The thirdidea considers the mind as a channel of communication conveying information.This idea was inspired by the seminal paper entitled ‘The Mathematical Theory ofCommunication’ published by Claude Shannon in 1948. In this article, Shannonproposed a linear system of information transmission and provided a quantitativedefinition of information, or uncertainty, measured in bits. Different laws pre-dicting human behaviour, like Hick’s law (Hick, 1952) or Fitts’ law (Fitts, 1954), wereformalized by cognitive psychologists thanks to Shannon’s theory of communica-tion. The quantitative definition of information is very useful to objectivelydetermine the complexity of the task or the uncertainty the participants have tocope with.

These threemain ideas ledcognitivepsychologists toconceptualize the information-processingparadigm.The term‘paradigm’ isusedhere in thesenseofKuhn(1962), that

91.2 THE COGNITIVE PSYCHOLOGY APPROACH

is to say, the ideas, theories, methods, techniques and applications shared by acommunity of researchers to solve important scientific problems. Typically, a para-digmis first establishedby thepublicationofarevolutionarybookorpaperthatsetsoutscientific problems and possible solutions. This chapter focuses on the information-processing paradigm because the larger part of studies interested in the effect ofexercise on psychological functions used this specific approach.

Five basic assumptions underlie the information-processing paradigm and areshared by the community of cognitive psychologists: (1) there are mental statesinvolved inmental processes; (2) mental states possess physical existence as physicalstates (i.e. they result from the electrochemical activity of neurons); (3) mental statescannot be reduced to physical states; (4) mental states take place in a computingsystem like the Turingmachine (i.e. themind is a symbol-processing system); and (5)a mental state corresponds to a representation and a mental process operatestransformations of representations. One of the central concerns of cognitive psy-chology is to study these representations andmental processes. Because these objectsare not directly observable, cognitive psychologists developedmethods to infer someof their characteristics from the measurement of different variables, such as reactiontime, and correct and incorrect response rates.

Behavioural psychologists assumed that scientific psychology must be based ondirectly observable facts such as stimuli and responses without any reference tohypothetical inner states. By contrast, cognitive psychologists are interested by statesand processes that take place between sensory inputs and motor outputs of theinformation-processing system. These hypothetical inner states and processes areconsidered as intermediary or latent variables; intermediary because they take placebetween stimuli and responses, and latent because they are not directly observable,but inferred from variations of dependent variables.

Typical topics of cognitive psychology include attention, pattern recognition,memory, motor control, reasoning, problem-solving, language, decision-making,learning and, more recently, executive functions. Typical goals of cognitive psychol-ogy are to formulate general principles and laws concerning cognitive processes thatare true for everyone; to separate the human information processing system intocomponents; to identify the nature and duration of the component processesinvolved in the performance of a cognitive task; and to describe the architectureof the cognitive system in order to explain human mental performances in a largevariety of situations. The information-processing paradigm was still dominant andconsidered by themajority of cognitive psychologists as the appropriate way to studyhuman cognition in the last decades of the twentieth century (Eysenck and Keane,1990). Today, thanks to the technological progresses in different branches ofneurosciences, such as brain imagery, it becomes necessary to shift from a ‘pure’cognitive psychology approach to a cognitive neuroscience approach to have a betterunderstanding of cognitive functions.

Since the beginning of cognitive psychology, two major conceptual frameworksof human information processing were in competition: the resource-driven modelsand the data-driven models (Rabbitt, 1979; Sanders, 1983). Generally speaking,


resource-driven models consider that performance of the human operator in a taskrequiring processing of information depends on three things: (1) the amount ofavailable resources; (2) the amount of resources required to perform the task; and(3) the amount of resources actually allocated to the different processes involved inthe task. Resources can be defined as energizing forces necessary to perform tasks(Gopher and Donchin, 1986). Human information processing and digital computersof the 1960s were conceived as limited capacity systems (Sanders, 1997), that is to say,systems which possess a limited amount of resources. In contrast, data-drivenmodels consider that performance depends on the quality of the processing realizedby a sequence of stages that transform input representations into output representa-tions. A processing stage refers to an aggregate of computational processes thatparticipate in the same mental operation, such as feature extraction or responseselection (Gopher and Donchin, 1986). Pure data-driven models inspired by thecomputer metaphor do not explain variability of performance under differentenvironmental or internal states of the organism (e.g. heat, stressful noise, emotion,effects of drugs, fatigue and arousal induced by exercise) (Hockey, Coles andGaillard,1986). In order to explain such variability in the information-processing system, itbecomes necessary to combine resource- and data-driven approaches. The resource-driven approach shares several ideas with the energetic approach presented in thenext section. For instance, the concept of resources is closely related to the concept ofarousal.

Cognitive psychologists developed ingenious methods to infer mental processes,which takeplacebetween stimuli and responses, and themake-upof the architectureof the information-processing system (e.g. Sternberg’s additive factors method,1969a). Several of thesemethods can be also used to determine the locus of influenceof aerobic exercise within the information-processing architecture and will bepresented later in this chapter. The architecture of the information-processingsystem has been described in different models (e.g. Sanders’ discrete/serialinformation processingmodel, Sanders, 1990). In the fourth section of this chapter,cognitive energetic models that synthesize the theoretical frameworks from cogni-tive and energetic approaches will be presented. They provide a very usefultheoretical framework, which explains the different effects of exercise on cognitiveprocesses.

1.3 The energetic approach

The energetic approach is concerned with the intensive or energizing aspects ofbehaviour as opposed to its directional or semantic aspects. Concepts such as arousaland activation were associated early with energy mobilization or energy releasewithin the organism (Duffy, 1962) and their relation to performance can be traced tothe earliest decades of experimental psychology and neurophysiology. For instance,Yerkes and Dodson (1908) observed an inverted-U shaped function of efficiencydepending upon the degree of arousal of the organism. These findings led to the

111.3 THE ENERGETIC APPROACH

general supposition that under- or over-aroused individuals perform poorly,whereas optimal performance occurs in a moderately aroused state (Hebb, 1955).The U-shaped conceptualization of arousal has maintained a place in mainstreampsychology and has been adopted by a number of applied researchers to explainperformance in human factors and sport settings (e.g. Easterbrook, 1959; Oxendine,1984; Raglin and Hanin, 2000). In this perspective, physical exercise has beenconsidered an arousing stimulation of the organism (Cooper, 1973; Davey, 1973)and a U-shaped function between exercise and cognitive performance has beenexpected (N€a€at€anen, 1973). However, empirical data did not support the hypothesisthat performance is an inverted-U function of exercise intensity (for a review, seeMcMorris and Graydon, 2000). The view that acute aerobic exercise can be consid-ered an arousing stressor is central in the present chapter.

Arousal and activation, terms often used interchangeably, were initially conceivedby researchers as unidimensional constructs that rangedona continuumfromsleep towakefulness (Duffy,1957, 1962; Malmo, 1959). Changes in arousal levels were oftenlinked to the activity of the ascending reticular formation (e.g. Lindsley, Bowdenand Magoun, 1949; Moruzzi and Magoun, 1949). According to the unidimensionalperspective, thereexists a generalnonspecific poolof energetic resources that supportsall cognitive functions and the amount of available resources allocated to a taskdepends, among other variables, on an individual’s arousal level (e.g. Kahneman,1973). The unitary perspective has been criticised, however. The low correlationsamong different measures of arousal obtained by researchers are inconsistent with aunidimensional view of arousal (Lacey, 1967; Eysenck, 1982; Thayer, 1989). Theundifferentiated resource view is not compatible with perfect time-sharing of tworesource-demanding tasks (Wickens, 1984). Further, neurophysiological evidencereveals that the reticular formation is not a homogenous systembut, rather, one that ishighly differentiated (Robbins and Everitt, 1995).

In response to these shortcomings, several researchers have proposed thatarousal is a multidimensional construct. For instance, on the basis of many animaland human neuropsychological and psychophysiological data, (Pribram andMcGuinness,1975; McGuinness and Pribram, 1980) suggested that there is aninvoluntary and a voluntary mode of attentional control. The involuntary modeinvolves two basal mechanisms: arousal, a phasic short-lived and reflex response toinput; and activation, a tonic long-lasting and involuntary readiness to respond. Athirdmechanism, effort, coordinates arousal and activation, and allows a voluntarycontrol of attention. Pribram and McGuinness restricted the use of the concept ofarousal to be synonymous with the orienting reaction discovered by Sharpless andJasper (1956), and Sokolov (1960). A response of the arousal mechanism occurswhen an input change produces a measurable phasic change in a physiological orbehavioural variable over a baseline. In accordance with studies by Berlyne (1960),arousal results when, in the history of the organism’s experience, an input issurprising, complex or novel. Such a reaction involves the assumption that theinput is matched against some residual of past experience in the organism, that is aresidual neuronal model of events. Activation differs from arousal, therefore, in


maintaining a tonic readiness to respond, reflected in an increase in corticalnegativity (contingent negative variation -- CNV) and tonic heart rate deceleration.Under many circumstances, arousal and activation appear to be forcibly linked. Instressful situations they share the function of reflex coupling input to output (e.g.startle reflex). In the absence of controlled arousal and activation, behavingorganisms would be constantly aroused by their movements and moved byarousing inputs. One function of the effort mechanism is to uncouple arousal andactivation in order to avoid undesirable reactions.

Recent advances in methods of assessing the structure and function of thebrain have provided researchers the means to identify more precisely the neuro-physiological components of arousal and activation. The reticular activating systemhas, for example, been shown to consist of several inter-related arousal systems thatare differentiated by specific neurotransmitters (Robbins and Everitt, 1995). Threemain systems of neuromodulators have been distinguished: the noradrenergic, thedopaminergic and the serotoninergic systems. Several studies conducted on animalsand humans showed that acute physical exercise results in a releasing of braincatecholamines (noradrenaline and dopamine) and indolamines (serotonin or5-hydroxytryptamine) (see Meeusen and De Meirleir, 1995, for a review). Therefore,a large part of acute and chronic effects of exercise on cognitive processes maybe closely related to the catecholaminergic and indolaminergic neuromodulationsof neural networks involved in information processing. The noradrenergic systemoriginates from the locus coeruleus in the pons. Neural cell bodies send projectionsthroughout a large part of the neocortex and the hippocampus. Fluctuations inlocus coeruleus activity can be observed in cortical electroencephalograms (EEG)and in P300 event-related potentials. Tone, lights or tactile stimuli, as well as noxiousor stressful events, result in an increase of the activity of the locus coeruleus, whichhas been linked to preservation of alertness that aids in detecting sensorysignals under high levels of arousal (see Foote, BloomandAston-Jones, 1983; Berridgeand Waterhouse, 2003; Posner, 1995; Pribram and McGuinness, 1975; Ramos andArnsten, 2007; Robbins and Everitt, 1995; for reviews). The coeruleo-cortical norad-renergic system appears to have a protective function of maintaining an individual’scapacity to maintain discrimination processes under stressful or arousingcircumstances. An increase in brain noradrenergic transmission improves thesignal-to-noise ratio of evoked responses to environmental stimuli, either byenhancing evoked responses, by suppressing ‘background activity’ or by a combi-nation of these two effects in several cortical terminal regions, whatever the sensorymodality (e.g. Kasamatsu and Heggelund, 1982; Foote, Freedman and Oliver, 1975;Hurley, Devilbiss and Waterhouse, 2004; Moxon et al., 2007; Waterhouse andWoodward, 1980).

The dopaminergic system originates from cell bodies located in the substantianigra pars compacta and from the ventral tegmentum. Projections from these areasmodulate neural activity in (a) the dorsal and ventral striatum, which, in turn, affectthe supplementary motor area, premotor area and primary motor cortex and (b) thefrontal lobe, and more particularly the medial prefrontal cortex, that underlies

131.3 THE ENERGETIC APPROACH

executive functions. These pathways affect the activation or energization of behav-iour and account for the vigour and frequency of behavioural outputs (Robbins andEveritt, 2007). Characteristics of noradrenergic and dopaminergic neuromodulatormechanisms are summarized in Table 1.4. The serotoninergic system originates fromcell bodies located in the raphe nucleus. Neurons from this system dampen theactions of each of the two preceding systems and promote behavioural inhibition andcortical deactivation (see also Chapter 2).

We saw that exercise has been considered an arousing stressor (e.g. Davey, 1973;N€a€at€anen, 1973; Thayer, 1987), but few investigators have provided a theoreticalrationale for a causal link between exercise, arousal, brain catecholamines andimprovement in cognitive performance. As stated by McMorris and Graydon(2000), Cooper (1973) was the first author to propose a set of clear arguments basedon different studies conducted with animals and humans: (1) synthesis of noradren-aline increases in the brain of the rat during severe and prolonged forced exercise;(2) concentration of plasma catecholamines increases during exercise; (3) brainnoradrenergic activity increases during cortical activation; (4) the level of corticalarousal is related to the level of activity of the reticular formation; and (5) exercise canincrease the activation of the reticular formation via somatosensory feedback due tothe movement of the limbs. Considerable research has provided support for thesearguments and today, the peripheral and central arousing effects of exercise are welldocumented. Acute exercise is widely known to activate both the sympatheticnervous system and the hypothalamo-pituitary-adrenal system, resulting in a releaseof catecholamines and indolamines, both centrally and peripherally (e.g. Wittert,2000; Meeusen andDeMeirleir, 1995, for reviews). Models andmethods presented inthe two following sections are compatible with this hypothetical explanation of acuteeffects of aerobic exercise on cognition.

1.4 Exercise effects and cognitive-energetic models

During the second half of the twentieth century, several cognitive-energetic modelssynthesising the two main approaches were proposed (see Figure 1.2). In the presentsection, four of these models will be presented. They provide a heuristic frameworkfor the study of the acute effects of exercise on cognition.

Table 1.4 Two energetical mechanisms activated by an acute bout of steady-state aerobic exercise.

Energeticalmechanism

Neurotransmittersystem Brain localization Main function ERP index

Arousal Noradrenaline Locus coeruleus Filtering inputs P300Activation Dopamine Substantia nigra

pars compactaEnergizing outputs CNV

Note: ERP: Event-related potential; P300: Positive cortical wave observed 300milliseconds after the occurrenceof response signal in a reaction time task; CNV: contingent negative variation.


Kahneman’s model (1974)

The first cognitive-energeticmodel can be found inKahneman’s book ‘Attention andEffort’ (1973). Kahneman viewed the amount of resources available at any time aslimited. The amount of available resources depends on the level of arousal, which isdetermined by two sets of factors: (1) the demands imposed by the activities in whichthe organism engages, or prepares to engage in; and (2) miscellaneous sources ofarousal such as intensity of stimulations, psychostimulant effect of drugs, anxiety oracute effect of aerobic exercise. Resources are accumulated in a single undifferenti-ated pool of resources (see Figure 1.3). The amount of available resources and theeffort invested performing a task can be measured through several measures ofarousal (e.g. pupil dilatation, heart-rate variability). An allocation policymechanismdirects and supervises the allocation of resources. The strategy of allocation isinfluenced by enduring dispositions (e.g. pre-wired and automatic behaviours suchas the automatic and involuntary orientation towards a novel stimulus), momentaryintentions (e.g. reaching a task-related goal) and feedback from ongoing activities(e.g. feedback of success).

According to Kahneman, the level of arousal corresponds to the amount ofavailable attentional resources, while effort is understood to be the voluntaryattention allocated to a task. In this perspective, decrements in performance aredue to demands of concurrent activities or processes which exceed the amountof available resources. The notions of ‘processing limitations’ and ‘effort invested toperform a task’ led Norman and Bobrow (1975) to introduce the very interestingconcepts of data-limited and resource-limited processes. Whenever an increase inthe amount of processing resources can result in improved performance, theperformance on that task is said to be resource-limited; and whenever performanceis independent of processing resources, the performance is said to be data-limited(see Figure 1.4). In the resource-limited region of the performance-resource function,

Figure 1.2 A schematic representation of the theoretical roots of cognitive-energetic models.

151.4 EXERCISE EFFECTS AND COGNITIVE-ENERGETIC MODELS

performance is assumed to be a monotonically increasing function of the amount ofallocated resources to perform the task. Norman and Bobrow distinguished twoforms of data limitations: (1) signal data limitations, when the limit to performancedepends primarily upon the signal-to-noise ratio and the quality of the input data

Figure 1.4 The Norman and Bobrow performance-resource function (see text for more explanation).

Figure 1.3 Kahneman’s cognitive-energetic model. (Modified from Kahneman, 1973.)


exercise and cognitive function · 2016-08-12 · 1.1 varieties of exercise effects on...

Documents