02 Physical Activity and Psychological Well-Being

1111
2 Physical Activity and Psychological

3
4 Well-Being
5111
6
7
8
9
1011
1
2
3111 The ‘feel-good’ effect of physical activity is often reported by exercisers
4 but has not been sufficiently recognised in health services and health
5 promotion. In the modern world of stress and inactivity there is an
6 increasing need to assess the potential benefits of physical activity for our
7 mental health. Physical Activity and Psychological Well-Being provides a
8 research consensus on the relationship between physical exercise and
9 aspects of mental health. Whilst reviewing and integrating relevant infor-
20111 mation, the book also considers physical activity in relation to the different
1 aspects of mental health:
2
3 • anxiety
4 • depression
5 • mood and emotion
6 • self-esteem
7 • cognitive functioning
8 • psychological dysfunction
9
30111 Physical Activity and Psychological Well-Being is an important resource
1 and foundation for those in health services, health psychology, clinical
2 psychology, psychiatry and sport and exercise settings for promoting the
3 benefits of physical activity for improving mental health. The text is also
4 invaluable reading for undergraduate and postgraduate students in sport
5 and exercise science.
6
7 Stuart J.H. Biddle is Professor of Exercise and Sport Psychology at Lough-
8 borough University, Kenneth R. Fox is Professor and Head of Exercise
9 and Health Sciences at the University of Bristol, Stephen H. Boutcher is
40111 Reader in Psychophysiology at De Montfort University.
1
2
3
44
45111
ii Contents
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
4
45111
Contents iii
1111
2
3
Physical Activity and
4
5 Psychological Well-Being
6
7
8
9
1011
1
2
3111 Edited by Stuart J.H. Biddle,
4
5 Kenneth R. Fox, Stephen H. Boutcher
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
LE
9 UT D
RO
GE
40111
1
•
•
2
p
Ta
ou
or
y
3 r
l
& F r n cis G
a
4
45111 London and New York
iv Contents
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
First published 2000
5 by Routledge
6 11 New Fetter Lane, London EC4P 4EE
7 Simultaneously published in the USA and Canada
8 by Routledge
9 29 West 35th Street, New York, NY 10001
20111 Routledge is an imprint of the Taylor & Francis Group
1 This edition published in the Taylor & Francis e-Library, 2002.
2
© 2000 Stuart J.H. Biddle, Kenneth R. Fox and Stephen H. Boutcher,
3 selection and editorial matter; individual chapters © the contributors.
4
All rights reserved. No part of this book may be reprinted or
5 reproduced or utilised in any form or by any electronic, mechan-
6 ical, or other means, now known or hereafter invented, including
7 photocopying and recording, or in any information storage or
retrieval system, without permission in writing from the publishers.
8
9 British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
30111
Library of Congress Cataloging-in-Publication Data
1 Physical activity and psychological well-being / edited by Stuart J.H. Biddle,
2 Kenneth R. Fox, Stephen H. Boutcher.
3 p.cm.
Includes bibliographical references and index.
4 1. Exercise—Psychological aspects. 2. Mental illness—Physiological
5 aspects. I. Biddle, Stuart. II. Fox, Kenneth R., 1949– III. Boutcher,
6 Stephen H. (Stephen Hugh), 1949–
RA781.P562 2000
7 613.7′1′019—dc21 00–028617
8
9 ISBN 0–415–23481–6 (hbk) ISBN 0–415–23439–5 (pbk)
ISBN 0-203-46832-5 Master e-book ISBN
40111
1
2
ISBN 0-203-77656-9 (Adobe eReader Format)
3
4
45111
Contents v
1111
2 Contents
3
4
5
6
7
8
9
1011
1
2
3111 List of figures vii
4 List of tables viii
5 List of contributors x
6 Acknowledgements xii
7
8
9 1 The case for exercise in the promotion of mental
20111 health and psychological well-being 1
1 KENNETH R. FOX, STEPHEN H. BOUTCHER,
2 GUY E. FAULKNER AND STUART J.H. BIDDLE
3
4 2 Physical activity, anxiety, and stress 10
5 ADRIAN H. TAYLOR
6
7 3 The relationship between physical activity and
8 clinically defined depression 46
9 NANETTE MUTRIE
30111
1 4 Emotion, mood and physical activity 63
2
STUART J.H. BIDDLE
3
4
5 The effects of exercise on self-perceptions and
5
self-esteem 88
6
KENNETH R. FOX
7
8
9 6 Cognitive performance, fitness, and ageing 118
STEPHEN H. BOUTCHER
40111
1
2 7 Physical activity as a source of psychological
3 dysfunction 130
4 ATTILA SZABO
45111
vi Contents
1111 8 The way forward for physical activity and the
2 promotion of psychological well-being 154
3 STUART J.H. BIDDLE, KENNETH R. FOX,
4 STEPHEN H. BOUTCHER AND GUY E. FAULKNER
5
6
7 References 169
8 Index 197
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
4
45111
Contents vii
1111
2 Figures
3
4
5
6
7
8
9
1011
1
2
3111 3.1 Relative risk of developing depression at follow-up from
4 different levels of baseline physical activity 50
5 3.2 Relative risk of developing depression at follow-up from
6 different levels of physical activity at baseline 51
7 5.1 A hierarchical model of self-concept 91
8 5.2 Self-perception constructs measurable at different levels
9 of specificity 92
20111 5.3 The Exercise and Self-Esteem Model for intervention
1 studies 94
2 5.4 Adaptation of the Exercise and Self-Esteem Model for use
3 with the Physical Self-Perception Profile 95
4 6.1 Standard scores averaged across eight cognitive tests
5 (Culture Fair IQ, Digit Span, Digit Symbol, Dots, RT,
6 Stroop Interference and Stroop Total) for aerobically
7 trained, strength trained, and control subjects 126
8 7.1 A psychosocial model for the development of eating disorder
9 symptoms in athletes 151
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
4
45111
viii Contents
1111
2 Tables
3
4
5
6
7
8
9
1011
1
2
3111 2.1 Summary of key meta-analytic review papers 14
4 2.2 Longitudinal relationship between physical activity and
5 anxiety 19
6 2.3 Effects of acute exercise on anxiety 28
7 2.4 Effects of chronic exercise on psychosocial stress
8 reactivity 34
9 2.5 Effects of acute exercise on psychosocial stress reactivity 38
20111 3.1 Changes in physical activity status and subsequent
1 depression 50
2 3.2 Randomised controlled studies of exercise treatment
3 for clinically defined depression 54
4 4.1 Common measures of health-related quality of life 65
5 4.2 Summary of mood and affect measures commonly used
6 in physical activity research 68
7 4.3 Factors of psychological well-being for physical activity
8 identified by Moses et al. (1989) 71
9 4.4 Summary of findings from meta-analytic and narrative
30111 reviews investigating the relationship between physical
1 activity, mood and affect 72
2 4.5 Summary of findings from British population surveys
3 investigating the relationship between physical activity
4 and psychological well-being 76
5 4.6 Controlled experimental trials from Britain investigating
6 physical activity, exercise and psychological well-being 78
7 5.1 Randomised control trials addressing the effect of
8 exercise on self-esteem and physical self-perceptions 100
9 6.1 Examples of cognitive tests used in fitness, cognitive
40111 performance and ageing research 120
1 6.2 Intervention studies examining the effect of fitness on
2 cognitive performance of older adults 124
3 7.1 Components or symptoms of exercise dependence 135
4 7.2 Summary table of research into exercise dependence 138
45111
Tables ix
1111 7.3 (a) Exercise dependence self-evaluation check list 143
2 (b) Self-help strategies for the modification of exercise
3 habits 143
4 7.4 Summary table of published papers on physical activity
5 and eating disorders relationships 145
6 8.1 Psychological and physiological mechanisms that could
7 underlie the relationship between physical activity and
8 mental health 164
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
4
45111
1111
2 Contributors
3
4
5
6
7
8
9
1011
1
2
3111 Professor Stuart J.H. Biddle is Professor of Exercise and Sport Psychology
4 at Loughborough University where he is associated with the British
5 Heart Foundation National Centre for Physical Activity and Health
6 and the Institute of Youth Sport, both within the Department of Physical
7 Education, Sports Science and Recreation Management. He was
8 President of the European Federation of Sport Psychology 1991–1999,
9 and in 1998 was awarded Distinguished International Scholar of the
20111 Association for the Advancement of Applied Sport Psychology. He is
1 the inaugural Editor-in-Chief of the journal Psychology of Sport &
2 Exercise. Professor Biddle’s research interests are in motivational
3 influences on health-related physical activity and exercise, as well as
4 emotional and other psychological outcomes of physical activity; and
5 he enjoys cycling to work.
6
Professor Kenneth R. Fox is Professor and Head of Exercise and Health
7
Sciences at the University of Bristol. His research interests include exer-
8
cise and physical self-perceptions, the role of physical activity in obesity
9
prevention and treatment, and health-related behaviour change inter-
30111
ventions. Publications include the edited text The Physical Self: From
1
Motivation to Well-being (Champaign, IL: Human Kinetics, 1997) and
2
the Physical Self-Perception Profile Manual (DeKalb, IL: Northern
3
Illinois University, Office for Health Promotion, 1990) as well as papers
4
on exercise focusing on motivation, obesity, and well-being.
5
6 Dr Stephen H. Boutcher is currently Director of the Physical Activity and
7 Health Research Unit at De Montfort University. His research focuses
8 on the effect of physical activity on the mental and cardiovascular health
9 of hypertensives, older men and women, and individuals with arthritis.
40111 He has extensive experience of publishing and is a member of the
1 Editorial Board of the Journal of Sport & Exercise Psychology. Prior
2 to his return to the UK, Dr Boutcher was a staff member of universities
3 in the United States and Australia.
4
45111
Contributors xi
1111 Guy Faulkner is currently a researcher in exercise psychology at Lough-
2 borough University. After graduating in physical education at the
3 University of Sydney, Australia, Guy completed a Masters degree in
4 Sport and Exercise Psychology at the University of Exeter, England,
5 before lecturing in Higher Education for three years. Research inter-
6 ests are mainly in physical activity and mental health and its promotion.
7 The role of exercise as an adjunctive treatment for schizophrenia is
8 also of particular interest and he has published on this topic in the
9 Journal of Sport & Exercise Psychology and the Journal of Mental
1011 Health.
1
Professor Nanette Mutrie is a Visiting Researcher at the MRC Social and
2
Public Health Sciences Unit at Glasgow University and is affiliated to
3111
the Centre for Exercise Science and Medicine at the university. She
4
has researched ways of increasing active living in both clinical popula-
5
tions and the community. Recent publications have addressed the
6
psychological benefits of physical exercise for women, physical activity
7
and its link with mental, social and moral health in young people and
8
exercise adherence issues for clinical populations. Professor Mutrie is
9
an Accredited Sport and Exercise Psychologist with the British
20111
Association of Sport and Exercise Sciences (BASES) and was Chair
1
of the BASES psychology section for three years until 2000. She is also
2
an elected member of the European Federation of Sport Psychology’s
3
Managing Council. In an effort to enjoy the mental health benefits of
4
activity, she plays squash and walks with her dog.
5
6 Dr Attila Szabo is Senior Lecturer in Sport and Exercise Psychology at
7 The Nottingham Trent University having been a researcher in Canada
8 and Hungary. His research interests focus on exercise and effect, exer-
9 cise dependence and deprivation, and Internet-based research.
30111
Professor Adrian Taylor is the Alexander Chair of Health and Physical
1
Activity at De Montfort University. He spent ten years at the University
2
of Brighton before moving to De Montfort in 2000. On behalf of the
3
British Association of Sport and Exercise Sciences he has led the devel-
4
opment of a Department of Health-funded National Quality Assurance
5
Framework for GP exercise referral schemes, and a national audit of
6
work with older people, sport and exercise in higher education. Adrian
7
has published and presented on physical activity interventions in
8
primary care, psychological aspects of sports injury rehabilitation, and
9
exercise and mental health. He is currently leading a three-year project
40111
looking at stress reactivity and exercise among trainee musicians.
1
2
3
4
45111
xii Contents
1111
2 Acknowledgements
3
4
5
6
7
8
9
1011
1
2
3111 The editors would like to thank the following for their valued input to
4 this project:
5
6 • Somerset Physical Activity Group and Somerset Health Authority, in
7 particular Trudy Grant
8 • British Association of Sport & Exercise Sciences, in particular Dr
9 Andy Smith
20111 • The British Psychological Society
1 • All members of the Academic Symposium for their helpful feedback
2 on earlier drafts of chapters
3 • Guy Faulkner (Loughborough University) for his editorial work.
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
4
45111
1111
2 1 The case for exercise in the
3
4 promotion of mental health
5111
6
and psychological well-being
7
8 Kenneth R. Fox, Stephen H. Boutcher,
9 Guy E. Faulkner and Stuart J.H. Biddle
1011
1
2
3111
4 Exercise and community health
5 There is now a worldwide acceptance among medical authorities that phys-
6 ical activity is an important element of healthy living (WHO, 1995).
7 Syntheses of studies (Berlin & Colditz, 1990; Powell, Thompson, Caspersen,
8 & Kendrick, 1987) have indicated that sedentary lifestyles carry at least
9 twice the risk of serious disease and premature death. This is on a par with
20111 the relative risk of hypertension and hyperlipidemia and not far behind
1 smoking and has led to suggestions that inactivity should be considered the
2 fourth primary risk factor for coronary heart disease and stroke. Sedentary
3 living is also the most prevalent risk factor with around 40% of the middle-
4 aged and elderly population taking part in infrequent or no moderate to
5 vigorous physical activity (Sports Council/HEA, 1992).
6 The public burden of inactivity is therefore high and activity promo-
7 tion could provide a cost-effective strategy for public health improvement
8 (Morris, 1994). In the US it has been estimated that inactivity results
9 in one third of all deaths from CHD, colon cancer and diabetes (Powell
30111 & Blair, 1994). The strength of the evidence has led to a US Surgeon
1 General’s Report entitled Physical Activity and Health (1996) calling for
2 nationally driven initiatives to promote physical activity. In the UK,
3 the Health of the Nation Task Force on Physical Activity produced the
4 consultation paper More people, more active, more often (Department of
5 Health, 1994b). Also, the Health Education Authority expert consensus
6 conference was held to determine the recommended amount of activity
7 for health and targets for physical activity promotion (Killoran, Fentem,
8 & Caspersen, 1994). Policy documents and agendas for physical activity
9 promotion were also produced by organisations such as the National
40111 Forum for Coronary Heart Disease Prevention (1995).
1 Since that time, substantial amounts of public funds have been provided
2 through the Health Education Authority to deliver Active for Life, a public
3 media and community support campaign to promote physical activity. This
44 has finished now and in its latter phases had a more specific focus on
45111 groups such as young people, women, and ethnic minorities. For instance,
2 Kenneth Fox et al.
1111 a consensus conference, recommendations document and book summar-
2 ising existing literature (Biddle, Sallis, & Cavill, 1998) has been produced
3 concerning young people and physical activity. A similar campaign has
4 been launched by the Health Education Board for Scotland.
5 This rapid boom in interest in the role of physical activity in health has
6 not only taken place at the central policy-making level. There have also
7 been significant initiatives at grass roots level since the early 1980s.
8 For example, many schools and local education authorities have been
9 attempting to promote children’s physical activity through a greater
1011 emphasis on health in the curriculum and through schemes to promote
1 walking and cycling to school. Leisure services have teamed up with
2 primary health care units to develop exercise prescription schemes. These
3111 mushroomed in the 1990s (see Fox, Biddle, Edmunds, Bowler, & Killoran,
4 1997) largely in the absence of rigorous evaluation but have thrived
5 because of a general belief in their efficacy and value by patients and
6 personnel. In contrast, the use of exercise in secondary care has been slow
7 and mainly restricted to cardiac rehabilitation. Similarly, commerce and
8 industry in Britain have not mirrored the tremendous growth in corporate
9 wellness programmes seen in the United States.
20111 In summary, developments in the use of exercise as a medium for health
1 promotion have been built on increasingly sound evidence from epidemio-
2 logical and well-controlled training studies (see HEA, 1995; and Pate
3 et al., 1995, for summaries) as well as a grass roots interest among various
4 groups of professionals. It is clear that the case has been constructed
5 around the impact of exercise on reducing the risk of physical health prob-
6 lems such as CHD, some cancers, obesity, diabetes, and to a lesser extent
7 musculo-skeletal problems such as low back pain and osteoporosis. To
8 date, much less attention has been paid to the contribution of exercise
9 to the prevention and treatment of the increasingly burgeoning problem
30111 of mental disorders, illnesses, and general mental malaise.
1
2
The mental health problem
3
4 The 1995 Health Survey for England showed that 20% of women and
5 14% of men may have at some time suffered mental illness. It has been
6 estimated that one in seven adults in the UK will suffer some form of
7 psychiatric morbidity at some point in their lifespan. Even among chil-
8 dren it has been estimated that up to 20% will suffer mild and 7–10%
9 moderate to severe mental health problems that hinder normal develop-
40111 ment (Kurtz, 1992) and there is evidence of a worsening trend, particularly
1 in socially disadvantaged populations (Rutter & Smith, 1995).
2 The most prevalent psychiatric disorder is depression, affecting 5–10%
3 of the population of most developed countries (Weismann & Klerman,
4 1992), with some estimates suggesting that 20% of the population will be
45111
The case for exercise 3
1111 affected by ‘depressive disorders’ at one time in their lives (Richards,
2 Musser, & Gerson, 1999). It is more common in the older middle-aged
3 and elderly populations with the result that 20% of consultees in primary
4 care in Britain have recognisable degrees of symptomology (Paykel &
5 Priest, 1992). In addition, more than half with mental health problems
6 seek help from their GP yet GPs have no specialised training in this
7 area (Richards et al., 1999). Treatment is generally through serotonin-
8 enhancing pharmaceuticals. Less common is the use of psychotherapy,
9 sometimes incorporating stress management techniques and occasionally
1011 exercise.
1 Such prevalance is not without great cost. The Department of Health
2 (1996) estimated that in 1992–93, 17% of expenditure in the health services
3111 that amounted to more than £5 billion was spent on mental illness and
4 disorders. In 1992, the Office of Health Economics estimated costs of
5 treating depression at £333 million, which included £55 million for drugs,
6 £250 million for hospitalisation, and £28 million for primary care consul-
7 tations. Wider cost implications were estimated at £6 billion when social
8 services provision, sickness and invalidity benefits, and loss of productivity
9 were included. Cooper and Cartwright (1996) estimated that half of all
20111 absenteeism due to sickness is stress related. The Department of Health
1 (1996) estimated that 15% and 26% of days of certified incapacity in men
2 and women respectively were due to mental disorders.
3 Problems with mental health are also associated with suicide ideation,
4 suicide attempts, and successful suicides, contributing to human distress
5 and further service costs. There are also increasing signs that less than
6 optimal mental well-being is common in the population. The impact of
7 emotional distress, low self-esteem, poor body image, chronic anxiety and
8 stress that is not diagnosed as a clinical disorder has not been possible to
9 estimate. However, it adds to the demands of primary care and social
30111 services, is linked to drug abuse problems, alcoholism and increased absen-
1 teeism from work. Furthermore, mental well-being is a critical element of
2 quality of life.
3 Recently, the Department of Health acknowledged the problems of
4 increasing mental illness and poor mental well-being in their White Paper
5 Saving Lives: Our Healthier Nation (1999) and wrote ‘The national strategy
6 must reflect more than just the absence of physical disease and be a basis
7 for efforts which acknowledge a more rounded idea of good health’. The
8 promotion of mental health has been included as one of four health targets
9 in the proposed national health contract so that regional and local strate-
40111 gies will be developed to address the problem. As part of this contract,
1 new primary care groups consisting of multi-professional teams are being
2 established. In promoting mental health, these groups will have to consider
3 a broader range of approaches that can be incorporated within local health
4 improvement programmes.
45111
1111 Physical activity and the promotion of psychological
2 well-being
3
4 Although it is clear that the case for exercise in reducing physical illness is
5 well established, there has also been a growing interest in the contribution
6 of exercise to the alleviation of the problems of mental illness. Specifically,
7 there has been increasing consideration of the role of exercise as
8
9 • a therapy for the treatment of mental illness and disorders;
1011 • a means of coping and managing mental illness;
1 • a means of improving quality of life for the mentally ill; and
2 • a means of preventing the onset of mental health problems.
3111
4 In addition, there is a growing recognition of a widespread mental malaise
5 in the general public that is expressed as mild depression, low self-esteem,
6 high stress and anxiety and poor coping. This has been accompanied by
7 institutional and cultural reductions in physical activity levels and it has
8 been suggested that increases in exercise participation may have a substan-
9 tial impact on the incidence of sub-clinical levels of mental ill health among
20111 the general public. This has accompanied a greater focus, at the demand
1 of the research councils and National Health Service funding bodies, on
2 the assessment of quality of life and related constructs such as life satis-
3 faction and mental well-being.
4 These concerns have been reflected in an increasing interest in research
5 and policy concerning exercise and mental health. In 1987 the US National
6 Institute of Mental Health consensus workshop statements regarding the
7 contribution of exercise to mental health were published in a book by
8 Morgan and Goldston (1987). This comprehensive summary of the liter-
9 ature has since been followed and updated by others such as Biddle and
30111 Mutrie (1991), Leith (1994) and more recently Morgan (1997), and several
1 published narrative and meta-analytical reviews. There have also been
2 further consensus conferences which have, at least in part, addressed the
3 mental benefits of exercise, the most notable of these being the ‘Physical
4 activity, health and well-being’ conference held in Quebec in 1995 (see
5 Biddle, 1995; Blair & Hardman, 1995) and the San Diego conference on
6 adolescence and physical activity (see Sallis & Patrick, 1994). Additionally,
7 in a review of the treatment of depression in primary care services, the
8 Centre for Health Economics (Freemantle et al., 1993) recommended
9 the funding of research into the effectiveness of non-drug therapies, partic-
40111 ularly for those who do not respond well to medication.
1 Accordingly, either in terms of clinical or non-clinical conditions, exer-
2 cise may offer substantial potential alone or as an adjunct in improving
3 the mental well-being of many individuals. There are five important
4 benefits that are associated with the potential use of exercise in such a
45111 role. First, exercise is cheap. Second, exercise carries negligible delete-
1111 rious side-effects. Third, exercise can be self-sustaining in that it can be
2 maintained by the individual once the basic skills have been learnt
3 (Martinsen, 1993). Fourth, given that many common non-drug treatments,
4 such as cognitive behavioural therapy, can be expensive and often in short
5 supply (Mutrie, this book), there is much to commend other strategies.
6 The need for treatment in psychiatry can never be fully met by health
7 professionals. Promoting exercise could reach a broader audience of indi-
8 viduals who cannot access therapy or would prefer not to use medication.
9 Greater time and effort could also be targeted at more acute and complex
1011 cases by mental health services. Finally, given the inherent physical bene-
1 fits, exercise should be promoted regardless of any impact on mental
2 health. In particular, the physical health needs of psychiatric clients are
3111 poorly served (McCarrick, Manderscheid, Bertolucci, Goldman, & Tessler
4 (1986). These benefits have important implications for the quality of life
5 of many individuals as well as the financial burden imposed on the NHS
6 by mental ill health.
7 In some countries, the evidence for exercise and mental health has
8 already been accepted and formalised into delivery systems. In Belgium,
9 for instance, psychomotor therapy to treat depression and anxiety is now
20111 established in the health system. Unfortunately, in the UK it remains
1 unusual for mental health services to use exercise as a therapy or preven-
2 tive medium. Furthermore, the case for exercise and its potential to
3 improve the general mental well-being of the population and prevent
4 mental illness has either not been widely publicised or seen as a priority
5 by health services, a situation paralleled in the US, where exercise is also
6 not a more popular treatment option despite the supportive evidence
7 (Tkachuk & Martin, 1999). One reason for this lack of recognition may
8 be the ineffective diffusion of such research to other health professionals,
9 which this book seeks to rectify.
30111
1
Purposes of this book
2
3 Somerset Health Authority (SHA) has commissioned this series of papers.
4 SHA through its appointed officers and the Somerset Physical Activity
5 Group are already committed to physical activity promotion. Notably they
6 have developed the first recognition/accreditation and scientific advisory
7 system for exercise prescription in the UK. SHA has also seen the poten-
8 tial for exercise in the promotion of mental as well as physical health and
9 already has promotion and research projects underway. They saw a need
40111 to update and summarise existing evidence for the case of exercise and
1 mental health enhancement. This is particularly timely in the light of the
2 recent Department of Health White Paper (DoH, 1999) that has targeted
3 mental health as one of four key health outcomes in the national health
4 contract. In relation to mental health, exercise is specifically suggested as
45111 a strategy that ‘people’ can participate in for improving mental health.
1111 The purpose of these chapters is therefore to provide an updated
2 overview of the case for exercise and the promotion of psychological
3 well-being. Leading researchers have been recruited to produce chapters
4 on the effects of exercise on anxiety and stress, depression, mood and
5 affect, self-perceptions including self-esteem, cognitive performance
6 and also the negative effects of exercise on psychological well-being.
7 Attention has been directed at (a) treating mental illness and disorders,
8 and (b) enhancing psychological well-being in the general public. Priority
9 has been given to evidence from randomised controlled trials, large scale
1011 epidemiological studies, and meta-analytic reviews. Authors have been
1 asked to summarise their findings in tables and closing sections featuring
2 ‘what we know’ and ‘what we need to know’ statements. A closing chapter
3111 draws together key findings and implications for further research and
4 practice.
5 Throughout, attention has been paid to the relevance of findings to
6 health service commissioners and providers and every effort has been
7 made to present the information in familiar terminology. The Consensus
8 statements that were adapted from this work have been supported by
9 professional and governing bodies such as the BPS (British Psychology
20111 Society), BASES (British Association of Sport and Exercise Sciences) and
1 Exercise England. The strong evidence base exists! It is hoped that this
2 resource will be valuable for practitioners in ‘making the case’ for phys-
3 ical activity and mental health at a local and national level.
4 In closing, it is important to acknowledge a range of other conditions
5 for which exercise has been suggested as a therapeutic possibility and
6 that are not covered fully in this book. For example, exercise has been
7 suggested as an adjunctive therapy for schizophrenia (see Faulkner &
8 Biddle, 1999), developmental disorders (see Gabler-Halle, Halle & Chung,
9 1993), somatoform disorders (see Tkachuk & Martin, 1999), substance
30111 abuse disorders such as alcohol dependence and drug addiction (see
1 Mutrie, 1997) and smoking cessation (see Ussher, Taylor, West &
2 McEwen, in press). Additionally, promising research on the effects
3 of exercise on sleep (see Youngstedt, O’Connor, & Dishman, 1997) is
4 emerging. These areas have been given limited treatment here largely
5 because they are emergent areas of research or an insufficient amount of
6 research has been conducted to warrant reliable conclusions. However,
7 they point to the exciting and as yet untapped potential exercise may offer
8 within the growing field of mental health promotion.
9
40111
Definitions
1
2 Throughout the papers, the following definitions apply.
3
Affect (emotion): specific feeling states generated in reaction to certain
4
events or appraisals.
45111
1111 Anorexia nervosa: a psychological disorder characterised by excessive
2 weight loss through dieting and over-exercising motivated by a false
3 perception of fatness.
4
Anxiety: an emotional state, typified by a cognitive component (e.g. worry,
5
self-doubt and apprehension) and a somatic component (e.g. heightened
6
awareness of physiological responses such a heart rate, sweaty palms
7
and tension). Anxiety can also be described as a relatively stable dispo-
8
9 sition to be anxious across multiple situations and the term used is trait
1011 anxiety.
1 Bulimia: an eating disorder in which binge eating is followed by self-
2 induced vomiting or purging.
3111
4 Clinically defined depression: a level of depression that has been diag-
5 nosed with standard instruments (e.g. above 16 on the Beck Depression
6 Inventory) and/or clinical interviews (e.g. satisfying criteria from the
7 Diagnostic and Statistical Manual of Mental Disorders IV).
8 Commitment to exercise: the degree of dedication or devotion to an
9 adopted exercise behaviour.
20111
1 Deprivation sensations: (see withdrawal symptoms).
2 Exercise: a subset of physical activity that is volitional, planned, struc-
3 tured, repetitive and aimed at improvement or maintenance of an aspect
4 of fitness or health.
5
6 Exercise addiction: a condition in which exercise becomes the most impor-
7 tant component of one’s life; exercise is performed even when other
8 commitments are more important or against medical advice; a process that
9 has negative consequences on both the individual and her/his environment
30111 including career and personal relationships (see also exercise dependence
1 and obligatory exercising).
2 Exercise dependence: a behavioural process in which the need for exer-
3 cising is so strong that it controls the individual’s life with several
4 identifiable symptoms such as salience, euphoria, exercise tolerance, with-
5 drawal symptoms, conflict, relapse, loss of control over life activities and
6 exercise behaviour, identifiable negative consequences, risk of self-injury,
7 social withdrawal, lack of compromise, and denial but awareness of the
8 problem (see also exercise addiction and obligatory exercising).
9
40111 Exercise deprivation: the forced need to abstain from exercising due to
1 other commitments or injury that often results in negative psychological
2 and/or physical symptoms (see also withdrawal symptoms).
3
Mood: the global set of affective states we experience on a day-to-day
4
basis.
45111
1111 Negative addiction: a term used to contrast positive addiction and it refers
2 to the harmful effects of socially endorsed behaviours, such as exercise
3 or work, when they are performed in a compulsive manner.
4
Obligatory exercising: used as a synonym to exercise addiction and exercise
5
dependence; the individual loses control over exercise whilst the latter
6
takes control over the individual (see also exercise addiction and exercise
7
dependence).
8
9 Physical activity: an umbrella term describing any bodily movement
1011 produced by the skeletal muscles resulting in energy expenditure.
1
Physical fitness: a multidimensional indicator of several functional capac-
2
ities such as cardiovascular endurance, muscular strength, or mobility,
3111
which in varying degrees are a result of genetics and stage in the lifespan,
4
as well as physical activity levels.
5
6 Positive addiction: often confounded with commitment to exercise, the term
7 refers to compulsion with a socially accepted behaviour such as exercise
8 or work in contrast to socially condemned behaviours such as alcohol or
9 drug abuse (see also commitment to exercise).
20111
Primary exercise dependence: a behavioural process in which exercise
1
performed in a compulsive manner represents an end in itself to achieve
2
something positive such as euphoria and/or to avoid something negative
3
such as withdrawal symptoms (see also exercise dependence and secondary
4
exercise dependence and withdrawal symptoms).
5
6 Psychological dysfunction: undesirable state of mental well-being that has
7 adverse effect(s) on the individual’s normal daily functioning.
8
Secondary exercise dependence: a behavioural process in which exercise
9
performed in a compulsive manner represents a means to achieve another
30111
objective such as weight loss (see also exercise dependence and primary
1
exercise dependence).
2
3 Self-esteem or self-worth: the awareness of good possessed by the self and
4 represents how positive individuals feel about themselves in general.
5
Self-perception: an umbrella term that denotes all types of self-referring
6
statements about the self ranging from those that have specific content to
7
those that express general feelings.
8
9 Social physique anxiety: anxiety or over-concern about one’s physical
40111 appearance in situations in which comparison (in physical appearance)
1 with others is inevitable.
2
Sport: physical activity that involves structured competitive situations
3
governed by rules. In mainland Europe, the term ‘sport’ is often used in
4
the wider context to include all exercise and leisure-time physical activity.
45111
1111 Stress: has been defined in many ways but an interactional perspective is
2 assumed within this book. As a result of appraisal of perceived capabili-
3 ties and demands associated with sources of stress (stressors) stress
4 manifests itself in emotional states, and physiological, psychological and
5 behavioural responses.
6
Withdrawal symptoms: psychological and physical symptoms one experi-
7
ences when a planned exercise session is missed; they are more severe in
8
people addicted to exercise and may surface even after missing a single
9
bout of planned exercise (see also exercise deprivation).
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
4
45111
1111
2 2 Physical activity, anxiety, and
3
4 stress
5
6 Adrian H. Taylor
7
8
9
1011
1
2
3111 Stress-related disorders have been widely recognised. The Department of
4 Health (DoH) estimated that in 1994 80 million working days were lost
5 due to anxiety and depression, at a cost of £5.3 billion. In addition, NHS
6 expenditure on the treatment of anxiety and depression has been esti-
7 mated at over £1 billion. Stress has been linked to all current Department
8 of Health (Our Healthier Nation) priority areas including cancer, coronary
9 heart disease/stroke, accidents, and mental health/suicide (DoH, 1999).
20111 A growing body of literature has focused on the relationship between
1 physical activity (PA), stress and anxiety (Landers & Petruzzello, 1994),
2 and yet in a recent review of the effectiveness of mental health promo-
3 tion interventions, only three studies were identified that had focused on
4 the use of exercise (Tilford, Delaney, & Vegells, 1997). The notion that
5 physical fitness somehow protects people from the stress of daily life is
6 not a new one. For example, Kobasa, Maddi and Pucetti (1982) reported
7 that executive businessmen, operating in a high stress environment, who
8 both participated in an aerobic exercise programme and had a ‘hardy’
9 personality had a reduced frequency of illness. The purpose of this review,
30111 therefore, is to examine the evidence for using exercise for the preven-
1 tion and treatment of anxiety disorders.
2 The review will begin with definitions of stress, anxiety and PA and a
3 brief introduction to their interrelationships. The chapter will then focus
4 on the evidence for these relationships from, firstly, a synthesis of quali-
5 tative and quantitative reviews, and then from a systematic review of
6 recent research studies. This will be followed by statements concerning
7 ‘what we know’ and ‘what we need to know’, and how our existing knowl-
8 edge may influence future interventions and professional practice among
9 those seeking to enhance mental health.
40111
1
Definitions of stress, anxiety and physical activity
2
3 Sources of stress (stressors) in our daily lives are rarely just objective, such as
4 extreme heat or noise. The interactional perspective on stress suggests that
45111 stress arises from an imbalance between our perceived capabilities and
Physical activity, anxiety, and stress 11
1111 perceived situational demands (Cox, 1985). Stress manifests itself in emo-
2 tional states, as well as physiological, psychological and behavioural
3 responses. The emotional state, anxiety, reflects negative cognitive appraisal,
4 typified by worry, self-doubt and apprehension (i.e. cognitive anxiety). State
5 (situational) anxiety may also be elevated due to increased awareness of
6 physiological responses to stress (i.e. somatic anxiety). As an example,
7 phobic anxiety may only occur in response to specific circumstances in which
8 a person perceives a lack of control or uncertainty. An individual may also
9 have a tendency to become anxious across many situations and this pre-
1011 disposition has been termed trait anxiety. Both state and trait anxiety,
1 because they are largely cognitive phenomena, have usually been assessed
2 using questionnaires. However, anxiety has also been inferred from mea-
3111 sures of physiological arousal (e.g. blood pressure).
4 For the purposes of this review, physical activity is defined as gross
5 movement of sufficient intensity and duration to potentially increase
6 aerobic or anaerobic capacity. Forms of movement such as those involved
7 in yoga, t’ai chi and stretching exercises are not included. However, the
8 review will consider the effects of such movement forms on anxiety when
9 included in a study as a comparison treatment intervention.
20111
1
The role of physical activity in anxiety prevention and
2
treatment
3
4 The main concern in the prevention and treatment of anxiety is that perfor-
5 mance in a wide range of daily tasks may suffer (e.g. examination failure;
6 serious accidents), and that enduring anxiety may lead to chronic health
7 problems, such as fear of social interaction or even suicide. This review
8 will focus mostly on the effects of physical activity on self-reported anxiety,
9 with reference to some physiological measures where understanding of
30111 underlying processes is enhanced. Typically, the chronic effects of exercise
1 training (over weeks to months) on trait anxiety have been examined.
2 One proposed mechanism for such effects is from the accumulated doses
3 of single exercise sessions. Acute exercise studies have therefore focused
4 on changes in anxiety from pre to post single exercise sessions. Both
5 chronic and acute exercise studies have employed research designs to
6 determine the specific effects of exercise by comparing changes in anxiety
7 with either control groups or alternative treatment groups.
8 In a related area of research, the effects of exercise on psychophysio-
9 logical reactivity to stress have also been examined. The main questions
40111 have been: (1) Does fitness level influence reactivity to stress? (2) Does
1 fitness training (chronic exercise) reduce reactivity to stress? (3) Does a
2 single (acute) session of exercise reduce reactivity to stress? Most studies
3 have used laboratory simulation of stress (e.g. cognitive tasks performed
4 under pressure) to answer these questions. The main implication of this
45111 work is that regular exercise may attenuate the response to daily stressors
12 Adrian Taylor
1111 (Selye, 1956) which for the most part require no such physiological reac-
2 tivity since the coping response rarely requires a fight or flight. Prolonged
3 inappropriate response to stressors may result in cardiovascular, meta-
4 bolic and immunological changes linked to chronic health problems such
5 as hypertension, diabetes mellitus and other diseases of maladaptation
6 (Selye, 1956).
7
8
Review of reviews
9
1011 The PA, stress and anxiety literature has advanced to the point where there
1 are reviews of reviews (Landers & Petruzzello, 1994; Gauvin & Spence,
2 1996). The two papers cited have reviewed 29 and 57 reviews, respectively,
3111 revealing great variability in completeness. For the purposes of the present
4 chapter, computerised search techniques (see later) and cross-checking in
5 recent research papers were conducted. Reviews (articles and book chap-
6 ters) were classified as qualitative (without meta-analysis) or quantitative
7 (with meta-analysis). The findings from the latter are shown in Table 2.1.
8
9
Strength of exercise effects from meta-analyses
20111
1 Meta-analyses have become an increasingly acceptable way to provide a
2 summary of both the strength of relationships between variables of interest
3 and also to identify the key moderator variables that might influence the
4 strength of such relationships (e.g. gender differences). As a guideline,
5 effect sizes (ES), usually determined through the calculation of differences
6 between an exercise and control group or from pre- to post-exercise inter-
7 vention, of up to 0.39, 0.40–0.69, and above 0.69, are generally described
8 as reflecting a small, moderate and large effect, respectively (Thomas &
9 Nelson, 1996).
30111 The reviews shown in Table 2.1 indicate that exercise has a low to
1 moderate anxiety-reducing effect across different populations and that
2 there are some important moderator variables. While some original
3 research studies have been reviewed in more than one of the reviews in
4 Table 2.1, each one offers a distinct focus.
5 Calfas and Taylor (1994) identified only 11 studies concerned with young
6 people that prevented further identification of moderator variables in the
7 meta-analysis. The anxiety-reducing effect of exercise was minimal, but
8 only four studies involved a control group and a further four had a poor
9 conceptual basis for inclusion in the review.
40111 In contrast, Long and Van Stavel (1995) included 40 studies involving
1 healthy adults (those over 18 years). There was a moderate anxiety-
2 reducing effect (ES 0.45) among exercise groups engaged in activity
3 with a dosage sufficient to increase aerobic fitness. This effect was slightly
4 reduced (ES 0.36) when contrasting change with other non-aerobic
45111 exercise groups. The effects seemed to be stronger for those reporting
1111 greater stress and among adults rather than college students. No other
2 moderators were identified.
3 McDonald and Hodgdon (1991) limited their review to 22 studies that
4 demonstrated aerobic fitness gain. Interestingly, the ES for reduction in
5 trait anxiety was lower (ES 0.25) than the previous review. Anxiety-
6 reducing effects were only apparent for males and young/middle-aged
7 participants.
8 Kugler, Seelbach and Kruskemper (1994) also restricted their review to
9 studies in which the exercise was of sufficient dosage to increase fitness,
1011 but only among cardiac rehabilitation patients (n 13 studies). The overall
1 ES was again small (0.31) with no identified moderator variables.
2 Petruzzello, Landers, Hatfield, Kubitz and Salazar (1991) provided the
3111 most comprehensive meta-analysis, with over 100 studies examining
4 the effects of both acute and chronic exercise on psychological self-report
5 measures and psychophysiological indices of stress. The overall ES ranged
6 from low (0.24) to moderate/high (0.65), with greater effects on self-report
7 measures from chronic exercise and greater effects on psychophysiolog-
8 ical measures from acute exercise. A number of moderator variables were
9 identified which inform us of the most likely subject, exercise, and method-
20111 ological characteristics that influence the strength of relationships.
1 For self-report measures, chronic exercise reduced trait anxiety more
2 when: exercise sessions lasted 21–30, or more than 40, minutes (in com-
3 parison to < 21); the training period was more than 15 weeks (in
4 comparison to < 10). Studies involving random or matched assignment
5 showed the greatest anxiety-reducing effects. In terms of acute effects,
6 state anxiety was reduced more when the exercise was aerobic and it
7 lasted 21–30 minutes (in comparison to < 21); and the ES was determined
8 from within-subject comparison from pre- to post-exercise as opposed to
9 changes from pre- to post-exercise being compared with other anxiety-
30111 reducing treatments.
1 For psychophysiological measures combining chronic and acute exercise
2 effects, there were more likely to be reductions in skin conductivity,
3 electromyography (EMG), and central nervous system (CNS) measures
4 when exercise intensity was 40–59% of heart rate max/VO2max (compared
5 to 70–79%); exercise duration was less than 31 minutes; participants were
6 18–30 years of age; there was no random assignment to exercise; and a
7 within-subject, pre- to post-exercise design was used.
8 In the most notable review on the psychophysiological reactivity to
9 stress, Crews and Landers (1987) conducted a meta-analysis involving 34
40111 studies. The findings generally indicated reduced reactivity to stressors (in
1 terms of heart rate, systolic and diastolic blood pressure, skin conductivity,
2 muscle tension and self-reported psychological symptoms), or faster
3 recovery following a stressor, for those who were fitter or improved their
4 fitness with training, or had just undertaken a single exercise session,
45111 compared with a baseline measure or a control group.
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Table 2.1 Summary of key meta-analytic review papers
Author No. of Number of Delimitations of review Overall ES Key moderators

studies effect sizes
Calfas & Taylor, 10 non-clinical Not given Various searches & 0.15 Insufficient studies so did
1994 1 clinical cross-checking, qualitative review
11–21 year olds
Crews & Landers, 34 92 from 1,449 Reactivity to stressors 0.48 No follow-up analysis due
1987 participants clearly stated to heterogeneity in ES.
But larger ES for
published, chronic,
random assigned studies
Kugler et al., 13 Not given Only studies designed 0.31 None (only considered
1994 to improve fitness in duration of training
post MI/angina/cardiac and time for follow-up).
patients Smaller n related to
No combined interventions greater ES
Long & Van 40 76 1975–93 w. gp = 0.45 Low-stressed = 0.28 v
Stavel, 1995 Only studies with b. gp = 0.36 High-stressed = 0.51
Type = gross, freq. = 2/3 p.w. Student = 0.16 v
Duration = 20 mins, > 6wks Adult = 0.53
Only >18 yrs, healthy No effect of gender,
No x-sectional designs anxiety measure, training
length or freq. of exercise
(p.w.), study drop-out,
leader background, aerobic
v non-aerobic, random
assignment, publication
type, follow-up
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Table 2.1 (continued)
Author No. of Number of Delimitations of review Overall E.S. Key moderators

studies effect sizes
McDonald & 22 Not given 1973–89 A state = 0.28 No effect for females
Hodgdon, 12 A state Only studies with fitness gain (0.31m/0.16f) No effect for older
1991 17 A trait – ACSM aerobic guidelines A trait = 0.25 (cf. young & middle age)
3 TMAS A state/trait & TMAS (0.28m/0.07f) Design not important
(pre-exp. had smaller ES)
Petruzzello et al., 104 408 from 3,048 Fully reported, up to 1989 A state = 0.24 Aerobic > non-aerobic
1991 participants Within 5s = 0.47 and
cardiac rehab = 0.48
A trait = 0.54 < 21mins = –0.12,
> 20min = 0.41
No effect for 18–30 yr olds
< 10 wks = 0.16,
> 10 wks= 0.36 to 0.63
random > intact groups
phys. = 0.56 GSR/EMG/CNS>
SBP/DBP/HR
(acute = 0.65) 40–59% > 70–79%
(chronic = 0.40) < 31 mins = 0.77, >
30 mins = 0.29
random < intact or single
gp repeated
18–30 > 45 yrs+
Notes:
Per week (p.w.); state anxiety (A state); trait anxiety (A trait); Taylor Manifest Anxiety Scale (TMAS: Taylor, 1953); galvanic skin response (GSR);
electromyography (EMG); central nervous system (CNS); systolic blood pressure (SBP); diastolic blood pressure (DBP); heart rate (HR); within-group
comparison (w. gp); between-group comparison (b. gp); male (m); female (f); myocardial infarction (MI); American College of Sports Medicine (ACSM);
physiological (phys.); effect size (ES); pre-experimental design (pre-exp.).
16 Adrian Taylor
1111 Qualitative reviews
2
There have been many qualitative (narrative) reviews, varying in scope
3
and rigour. Perhaps one of the most comprehensive involved 56 studies
4
concerned with chronic exercise and self-reported anxiety (Leith, 1994; also
5
see Leith & Taylor, 1990). Overall, 73% of the studies reported anxiety-
6
reducing effects. Leith suggested that the only exercise characteristic to
7
influence the outcome was duration. Chronic exercise lasting nine or more
8
weeks fairly consistently led to greater reductions in trait anxiety.
9
1011
1 Criteria for selecting original sources for the present review
2
3111 The present review involved a systematic on-line computer search of
4 Medline, CINAHL, Psychinfo, and Sport Discus. The following key words
5 were used in the searches: exercise, physical activity, physical fitness,
6 anxiety, and psychological stress. Review and original research literature
7 was also checked for additional references. Sources were limited to full
8 refereed journal articles, this excluding abstracts, conference proceedings,
9 book chapters, and dissertations. Because the most comprehensive review
20111 of the field (Petruzzello et al., 1991) included sources up to 1989, a deci-
1 sion was made to limit the search to articles appearing from 1989 onwards.
2 No sources were excluded on the basis of research design. The review
3 also includes papers which have considered the effects of exercise on
4 specific sub-domains of anxiety [e.g. test anxiety, treadmill and dyspnea
5 anxiety among Coronary Obstructive Pulmonary Disease (COPD)
6 patients, self-presentational anxiety, such as physical appearance anxiety
7 and social physique anxiety (Leary, 1992)]. Previous reviews have not
8 included these but reducing such types of anxiety may have important
9 consequences on future behaviour and well-being.
30111
1 Limitations of the search
2
3 Previous reviews have included studies using the anxiety/tension scale from
4 the Profile of Mood States (POMS) questionnaire. An editorial decision,
5 on behalf of the overall project, was made to include these studies in the
6 chapter on exercise and mood (Biddle, this volume). For the most part,
7 only studies that included a standardised, readily available anxiety measure
8 were reviewed.
9 Studies involving multifaceted interventions were excluded. For
40111 example, a number of cardiac rehabilitation programmes included exercise,
1 stress management and cognitive-behavioural counselling. While such
2 programmes are common they also make interpretation of mental health
3 changes impossible to attribute to exercise alone.
4 Studies that focused on exercise in a form not normally prescribed for
45111 the general population were also excluded. For example, some studies
1111 focused on changes in competitive state anxiety from before to after a
2 sports competition, or responses to maximal exercise testing.
3 This review will therefore consider whether literature from 1989 to the
4 present has confirmed and clarified our understanding of:
5
6 1 The anxiety-reducing effects of chronic physical activity.
7 2 The anxiety-reducing effects of acute physical activity.
8 3 The effects of chronic and acute physical activity on reactivity to stress.
9
1011
The effects of chronic exercise
1
2
Cross-sectional studies
3111
4 The 11 studies identified (Aldana, Sutton, Jacobson, & Quirk, 1996;
5 Brandon & Loftin, 1991; Frederick & Morrison, 1996; King & Cotes, 1989;
6 Lobstein, Ismail, & Rasmussen, 1989; Martinsen, Strand, Paulsson, &
7 Kaggestad, 1989; Muraki, Maehara, Ishii, Ajimoto, & Kikuchi, 1993; Nouri
8 & Beer, 1989; Stephens, 1988; Stewart et al., 1994; Szabo, Brown, Gauvin,
9 & Seraganian, 1993) vary considerably in terms of quality. A typical
20111 research design has involved the comparison of anxiety levels between
1 less and more active, or less and more fit, using a wide range of activity
2 measures or fitness tests. Clearly, such a pre-experimental design may
3 offer little in the way of causal evidence for the anxiety-reducing effects
4 of exercise, in comparison with quasi and true experimental designs
5 (Campbell & Stanley, 1963). Nevertheless, some large cross-sectional
6 studies (e.g. Stephens, 1988) have provided good generalisability (if weak
7 internal validity), or may offer an initial insight into stronger associations
8 among subsets of the population not previously identified.
9
30111
Summary of cross-sectional studies
1
2 The largest study (Aldana et al., 1996), of over 32,000 individuals enrolling
3 in a health insurance screening programme, used the Strain Question-
4 naire (Lefebvre & Sanford, 1984). A self-report activity survey was used
5 to determine typical energy expenditure levels which was then used to
6 dichotomise participants into high and low active. The less active were
7 twice as likely to report high stress levels. However, a causal relationship
8 cannot be inferred. It may well be that less anxious individuals were
9 attracted to physical activity and exercise, more than were anxious indi-
40111 viduals. Also, studies that have contrasted high and low fit people are
1 severely limited by the fact that fitness measures are partly genetic and
2 partly behaviour-related. Only a few studies examined both activity levels
3 and fitness measures.
4
45111
18 Adrian Taylor
1111 Design of longitudinal studies
2
The 27 studies involving some form of exercise intervention over a period
3
of time are summarised in Table 2.2. A typical design has involved one
4
or more measures of anxiety at baseline, followed by a period of exercise
5
over weeks or months, and one or more follow-up assessment(s) of anxiety.
6
In order to control for a variety of alternative explanations for anxiety-
7
reducing effects, a comparison has been made with a control group
8
receiving either no intervention or a placebo group receiving an inactive
9
intervention with equal contact time. Some studies have also contrasted
1011
the anxiety-reducing effects of exercise and other stress management inter-
1
ventions. All but five studies assessed fitness, using a variety of stan-
2
dardised testing protocol, at baseline and after the training period.
3111
4
5 Findings from longitudinal studies
6
Nine (33%) of the studies in the present review reported no anxiety-
7
reducing effects on one or more of the self-report measures administered.
8
Two of these studies appeared to have inadequate power to detect anxiety
9
reductions (see Kugler et al., 1994 for a discussion on sampling sizes in
20111
this field of research).
1
2
3 PARTICIPANT CHARACTERISTICS
4
Previous reviews have been critical of exercise scientists for using conve-
5
nient samples of college students. Indeed, 32% of the 56 studies involving
6
chronic exercise reviewed by Leith (1994) included college students. In
7
the present review only 15% involved college students, demonstrating a
8
new focus on older populations. Over 25% of the studies included partic-
9
ipants over 60 years of age. Unfortunately, there has been no increase in
30111
the concern for under 18-year-olds, with only one study identified since
1
1988 (Norris, Carroll, & Cochrane, 1992). A variety of unhealthy groups
2
were studied, including four with cardiovascular disease, one with cancer,
3
one with COPD, and six with mild to severe mental disorders. Almost all
4
the studies involved initially inactive individuals. There was no difference
5
in the number of males and females in the studies.
6
7
8 RESEARCH DESIGN
9
There appears to have been an increase in the proportion of studies that
40111
have employed random assignment. The present review revealed 67%
1
compared with about 40% using random assignment in the reviews by
2
Petruzzello et al. (1991) and Leith (1994). Of the nine studies that showed
3
no anxiety-reducing effect, one involved an intact group comparison and
4
the other eight involved random assignment. This suggests that there is
45111
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Table 2.2 Longitudinal relationship between physical activity and anxiety
Author(s) Participants Design Comparison group Duration Type of exercise Fitness Anxiety Outcome/
change () measure comments
Altchiller & 38f/5m Pre/post – 1. Aerobic ex. (AE) 8 wks 70–85% (AE only) N/A TAI Only AE ↓ TAI
Motta, 1994 32yrs Post/post 2. Non-aerobic (NAE) 3 p.w. (ES = –0.60)
(20–67) r.a. to 2 grps (calisthenics) Limitations: AE grp
initially
less anxious
Only 48% adherence to
AE
Bartlewski, 43f Pre-post 1. Aerobic ex. (AE) 10 wks High impact aerobics class N/A SPAS AE ↓ SPAS
Van Raalte, college 2 intact grps 2. Psychology class (C) (ES = 0.54 cf.
& Brewer, control = –0.10)
1996
Blumenthal 50m/51f Pre-post- 1. Aerobic ex. (AE) 16 wks 1. 70% HR, 30 min aerobic, Only AE TAI No ↓ TAI (for f or m)
et al., 1991 67yrs post r.a. to 2. Yoga (Y) (3 gps) + 30 mins non-aerobic, grp at
(60–83) 3 grps 3. Wait-list control (C) + 16 wks 3 p.w. 16 wks
inactive, (gp 1) 2. Yoga = 60 min, 2 p.w.
healthy + 6 mth
Brown et al., 69f/66m Pre-post 1. Mod. int. walk (MW) 16 wks 1. 65–75% HR, 30–40 mins, Only in TAI No ↓ TAI (for any grp)
1995 53 yrs r.a. to 5 grps 2. Low int. walk (LW) 3 p.w. MW &
(40–69) 3. (2) + relaxation 2. 45–55% HR, 40-50 mins, LW
inactive, (LWR) 3 p.w.
healthy 4. t’ai chi (TC) 3. As 2 with relaxation tape
5. No ex. control grp 4. 45 mins, 3 p.w.
Carrieri- 26f/25m Pre-post- 1. Aerobic ex. (AE) 5 wks 1. 12 sessions over 4–6 wks AE & SE SAI AE & SE ↓ all measures
Kohlman 67 +/– 7yrs post r.a. to 2. Self efficacy + 8 wks 70–85%, 20–30 mins TRANX (at 5 and 13 wks)
et al., 1996 mod-severe 2 grps then enhancement 2. (1) + coaching (nurse-led DA
COPD both did (SE) + 1 + video)
patients home based 3. Home walk, 70%,
ex. prog. 20 mins, 4 p.w.
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Cramer, 50f Pre-during- 1. Walking 15 wks Walking = 60% HR, Only SAI Walking ↓ SAI
Nieman, & 34 yrs post r.a. to 45 mins, 5 p.w. Ex. grp (at 6 wks only)
Lee, 1991 (20–45), 2 grps 2. No ex. control Supervised on outdoor route SAI not related to
inactive, fitness
healthy
Dixhoorn 147m/9f Pre-post r.a. 1. Aerobic ex. (AE) 5 wks AE = 70% peak HR, Both grps TAI No ↓ TAI (for either
et al., 1990 56 yrs to 2 grps 2. Aerobic ex. + (AET) 30 mins, 7 p.w. ↑ 50% grp)
(36–76) therapy cycle erg.
MI patients AET = additional 60 mins
p.d. of relaxation and
breathing
Fisher & 54f Pre-post r.a. 1. Control 6 wks 2. Aerobic and weight – N/A PASTAS Both therapy grps ↓
Thompson, 23 yrs to 3 grps 2. Ex. therapy 1 supervised and 2 home PASTAS
1994 (17–45) 3. Cog.-behav. therapy sessions p.w.
low 3. 6 1 hr sessions on
appearance stress, relaxation, image
evaluation etc.
Jambor 30 Pre-post Ex. v quiet rest 8 wks 60–90%, 15–45 mins, Ex. Grp ↑ Mod. Ex. & rest ↓ cognitive
et al., 1994 29 yrs 2 non- 3 p.w. aqua CSAI-2 and somatic anx.
(20–45) equivalent running program
inactive grps
King et al., 357 m, f Pre-post r.a. 1. No ex. control 12 mths Vigorous ex. = 73–88%, Gains 2, 3, TMAS 2, 3, & 4 ↓ TMAS and
1993 57 yrs to 4 grps 2. Vig. group ex. aerobic, 3 p.w. &4 PSS PSS
(50–65) 3. Vig. home ex. Moderate ex. = 60–73%, Higher adherence – more
inactive, 4. Mod. home ex. 30 mins, 5 p.w. ↓ TMAS and PSS
healthy Group ex. = at an ex. centre anxiety not related to
fitness or body wt.
Home based gp. did ↑ ex.
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Kugler 35m Pre-post 1. Ex. advice control 8 wks 2. 75%, 20 mins, 3 p.w., Gains 2, 3, TMAS 3 & 4 ↓ TMAS
et al., 1990 32–66 yrs r.a. to 4 grps 2. Hospital based cycle erg. &4 but small n
cardiac (from wks 3. Hospital + home 3.1 (hospital) + 2 (home)
rehab. 7–14 post cycle p.w., cycle
MI) 4. Hospital + home 4. 1 (hospital) + 2 (home)
walk p.w., cycle + walk
Martinsen, 70 m, f Pre-post 1. Aerobic ex. (AE) 8 wks 1. 70%, 60 mins, 3 p.w., Only AE CPRS AE and NAE ↓ anxiety
Hoffart, & 39 yrs r.a. to 2 grps 2. Non-aerobic ex. walk/jog grp PARS fitness not related
Solberg, anx. (NAE) 2. Strength & flexibilty ACS to anx.
1989 patients exercises plus
rehabilitation
Martinsen, 89 m, f Pre-post-post None 8 wks 50–70%, 60 mins daily, yes SRT ↓ anxiety (particularly
Sandvik & 40 +/–7 yrs + 12 ski, cycle, walk, jog, alcohol abusers and
Kolbjornsrud psych. mths aerobics + normal depressed) after 8 wks
1989 in-patients treatment and 12 mths
sedentary 65% adherence > greater
↓ anxiety
McAuley 58f,56m Pre-post None 20 wks Brisk walking, 15–40 mins, N/A SPAS ↓ SPAS – related to: ↓ hip
et al., 1995 55 yrs 3 p.w., supervised circ. for females, ↑ initial
(45–64) outcome expectancy, ↑
inactive, self efficacy (walking),
healthy being male, but not
adherence to ex. or age
Mock et al., 46f Pre-post 1. Walking (W) 3 wks 1. Walking, 20–30 mins, No VAS Walking ↓ anxiety
1997 49 yrs alternately 2. No ex. control (C) 4/5 p.w. 86% ex. adherence
(35–65) assigned to self-directed with limited
inactive, 2 grps therapist contact
newly
diagnosed,
breast
cancer
Norris, 77m Pre-post 1. Aerobic ex. (AE) 10 wks AE = 45 min, 3 p.w., Only AE ↑ JSQ AE ↓ job stress
Carroll, & 20–50 Intact grps 2. Non-aerobic ex. run, aerobic
Cochrane, healthy (NAE) NAE = 30 min, 3 p.w.,
1990 policemen 3. Control weights
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Norris, 30f, 30m Pre-post 1. Vig. ex. 10 wks 1, 2, 3 = 25–30 mins, Only vig. MAACL-A Vig ex. ↓ MAACL-A
Carroll, & 13–17 yrs, 4 intact 2. Mod. ex. 2 p.w., ex. ↑ PSS Vig ex. ↓ strength of
Cochrane, healthy grps r.a. 3. Flexibility (NAE) 70–75% (vig.), 50–60% stress-anxiety relationship
1992 to 4 conditions 4. Control (mod.)
Norvell & 43 m Pre-post Ex. v wait list controls 16 wks Ex. = 12 station circuit, 20 mins, ex. grp ↑ PSS Ex. ↓ anx. and PSS
Belles, 33 +/– 8yrs, r.a. to 2 grps 3 p.w. SCL-90-A
1993 healthy (3rd group =
policemen ex. drop-outs)
Norvell, 43 f Pre-post 1. Aerobic ex. (AE) 12 wks AE = 70–85%, 30 mins, only AE PSS No ↓ Anx./PSS
Martin, & 59 +/– 6yrs r.a. to 3 grps 2. Passive ex. (PE) 2 p.w. grp ↑ SCL-90-A But ↑ in fitness
Salamon, inactive, 3. No ex. control PE – on exercise tables related to ↓ PSS
1991 healthy
O’Connor 19f, 13m Pre-post Ex. v no ex. control 8 wks Ex. = 60–85%, 30 mins, only Ex. SAI No ↓ anx. (pre-post on
et al., 1995 college r.a. to 2 grps 3 p.w. grp ↑ maximal ex. test)
inactive, aerobics
healthy
Pierce, 90 m,f Pre-post 1. Aerobic ex. (AE) 16 wks AE = 70%, 35 mins, only AE SAI No ↓ SAI/TAI
Madden, et al., 45 yrs r.a. to 3 grps 2. Non-aerobic ex. (NAE) 3 p.w. walk/jog grp ↑ TAI
1993 (29-59) 3. Wait-list control NAE = circuit, weights,
mild flexibility, 30 mins,
hypertensives 2/3 p.w.
Pistacchio, 301 m, f Pre-post Post hoc comparisons to 10 wks 60%, 20–30 mins, 3 p.w. N/A TAI No ↓ in anxiety overall
Weinberg college identify determinants aerobics class but ↓ TAI predicted by
& Jackson, healthy initial ↑ ex. self efficacy
1989 & ↑ TAI
Sexton, 28f/25m Pre-post-post 1. Walking (W) 8 wks Both = 70%, 30 mins J ↑ at BPRS, + W and J ↓ Anx.
Maere, & 38 yrs r.a. to 2 grps 2. Jogging (J) + 6 mths 3/4 p.w., unsupervised 8 wks SCL-90-A., fitness (pre-6 mths)
Dahl, (19–60) after leaving hospital J&W↑ +& related to 6 mths
1989 neurotic (4–5 weeks) at 6 mths STAI anxiety
inpatients More drop-outs from
jogging
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Steptoe 33 m,f Pre-post- 1. Aerobic ex. (AE) 10 wks AE = 60–65%, 20 mins, AE ↑ TAI At 10 wks – AE ↓ POMS
et al., 1989 37 +/– 9 yrs post r.a. 2. Non-aerobic ex. + 3 mths jog/walk POMS (no ↓ in TAI)
(20–60) to 2 grps (NAE) NAE = weight training + At 3 mths – AE ↓ POMS
anxious, flexibility and TAI 70% adherence
overweight Both did 1 supervised and to supervised session
3 unsupervised p.w.
Topp, 1989 49 m,f Pre-post 1. Aerobic ex. (AE) 7 wks AE = running, 3 p.w. AE ↑ TAQ AE and R ↓ test anxiety
college 3 intact 2. Relaxation (R) R = 30 mins, 3 p.w. fitness not related to
healthy groups 3. Controls anx.
(classes)
Veale et al., 42f, 23m Pre-post 1. No ex. controls 12 wks AE = running, 3 p.w. AE ↑ TAI AE ↓ TAI
1992 35 yrs r.a. to 2 grps 2. Aerobic ex. (AE) fitness related to anx.
(19–58) 71% adherence to ex.
depressed
outpatients
Worcester 173 m Pre-post-post 1. Low ex. (NAE) 8 wks LE = low intensity, Only HE ↑ SAI No ↓ SAI for LE
et al., 1993 54 +/– 2yrs r.a. to 2 grps 2. High ex. (HE) + 12 calisthenics and inter- at 8 wks or HE at 8 wks or
< 70 yrs (over 3 yrs) mths mittent 60 mins, 2 p.w. 12 mths
post MI HE = high intensity, 60 mins,
patients 3 p.w. aerobic ex.
Notes:
Per week (p.w.); Spielberger State & Trait Anxiety Inventory (SAI & TAI: Spielberger et al., 1983); Taylor Manifest Anxiety Scale (TMAS: Taylor, 1953); Physical
Appearance State and Trait Anxiety Scale (PASTAS: Reed et al., 1991); Treadmill anxiety (TRANX); Dyspnea anxiety (DA); Profile of Mood States (POMS: McNair
et al., 1971) – Tension/Anxiety scale only; Modified Competitive State Anxiety Inventory (Mod. CSAI-2: Martens et al, 1990); Visual Analogue Scale (VAS); Symptom
Rating Test (SRT); Comprehensive Psychopathological Rating Scale (CPRS); Phobic Aviodance Rating Scale (PARS); Agoraphobic Cognitions Scale (ACS); Perceived
Stress Scale (PSS); Social Physique Anxiety Scale (SPAS: Hart et al., 1989); Multiple Affect Adjective Check List – Anxiety (MAACL-A: Zuckerman & Lubin, 1965);
Test Anxiety Questionnaire (TAQ: Sarason, 1975); Hopkins Symptom Check List (SCL-90: Derogatis, 1980); Job Stress Questionnaire (JSQ: Seamonds, 1982); Brief
Psychiatric Rating scale (BPRS); within-group comparison (w.gp); between-group comparison (b.gp); male (m); female (f); myocardial infarction (MI); coronary obstruc-
tive pulmonary disease (COPD); effect size (ES); N/A Not assessed; random assignment (r.a.); ergometer (erg.); circumference (circ.). NB. Scales without citations are
described in the respective sources shown in the table.
24 Adrian Taylor
1111 either bias in the publishing process (i.e. studies involving randomised
2 designs are more likely to be published when there are no anxiety-reducing
3 effects) or that anxiety-reducing effects are less likely to be observed in
4 carefully controlled studies; a point which cannot be resolved easily. There
5 appears to have been an increase in the proportion of studies with longer
6 periods of training and follow-up assessments. Petruzzello et al. (1991)
7 reported only 7% of studies involving training periods over 15 weeks
8 compared to 26% in the present review. Encouragingly, four studies also
9 included a follow-up assessment.
1011 The shortest study (3 weeks) by Mock and colleagues (1997) used a visual
1 analogue scale to assess anxiety. They found significantly reduced anxiety
2 among a group of newly diagnosed breast cancer patients assigned to a self-
3111 directed walking programme with only limited nurse contact (compared
4 with a no exercise control group of similar health status). Interestingly,
5 while there were only small increases in fitness for the walking group, there
6 were small losses in fitness for the control group. Another relatively brief
7 training study (just 5 weeks) by Dixhoorn, Duivenvoorden, Pool, &
8 Verhage (1990) involved 156 post-MI patients who either did just daily
9 exercise or daily exercise and relaxation therapy. Neither group reduced
20111 anxiety despite exceptionally large gains in fitness of 50%.
1
2
EXERCISE CHARACTERISTICS
3
4 The majority of the studies appeared to be concerned with increasing
5 aerobic fitness, with at least one group in 22 studies following American
6 College of Sports Medicine (ACSM, 1991) guidelines for frequency (3 times
7 per week), intensity (c. 60–90%) and duration (at least 20 minutes). Six
8 studies did not report complete information about the exercise dose. This
9 should not necessarily be seen as a criticism as studies involving free-living
30111 exercise in a natural environment are obviously very difficult to monitor,
1 and yet may well be most favoured by participants. Only two studies
2 reported some exercise sessions lasting less than 20 minutes (within a range
3 from 15 to 45 minutes) and both involved progression from shorter to
4 longer sessions over time. This is supported by the fact that of the 22 studies
5 reporting assessment of aerobic fitness change, only one (Mock et al., 1997)
6 failed to find an increase. Perhaps this is not surprising given the inactive
7 status of participants in the majority of studies. Interestingly, of the nine
8 studies reporting mixed or no anxiety-reducing effects, eight (the other did
9 not report fitness change) reported fitness gains for the exercise groups. Of
40111 the seven studies which specifically examined the effects of fitness change
1 on anxiety change, only three found a positive relationship. This confirms
2 previous reviews (e.g. Leith, 1994).
3 If aerobic fitness gain is not essential for anxiety reduction, is there a
4 difference between the effects of non-aerobic and aerobic exercise? From
45111 only two non-aerobic studies identified in the meta-analysis of Petruzzello
1111 et al. (1991) there was no evidence for anxiety-reducing effects. In the
2 present review, only one study (Martinsen, Hoffart, & Solberg, 1989) out
3 of seven (see Table 2.2) involving a non-aerobic exercise group (e.g. calis-
4 thenics, flexibility, circuit training, weight training), reported an anxiety-
5 reducing effect. This study involved initially anxious hospital inpatients,
6 and the non-aerobic group was involved in not only strength and flexi-
7 bility exercises, but also relaxation therapy. Both these facts may have
8 influenced the findings.
9 It is also important to consider adherence to exercise, since any anxiety-
1011 reducing effect is unlikely to remain if the treatment is discontinued. Only a
1 few studies examined whether greater adherence was related to greater
2 effects, or whether specific intervention groups differed in their adherence
3111 levels. There are, of course, important methodological implications too if
4 only those likely to reduce anxiety remain in a study. King, Taylor, &
5 Haskell (1993) reported that those who did more exercise reduced their
6 scores more on the Taylor Manifest Anxiety Scale and the Perceived Stress
7 Scale over a 12-month period. Interestingly, those in vigorous and moderate
8 home-based exercise groups had higher levels of adherence (compared with
9 group-based supervised sessions). Martinsen, Sandrik and Kolbjornsrud
20111 (1989) also identified a positive relationship between adherence and anxiety
1 reduction. In a study comparing the effects of walking and jogging on neu-
2 rotic hospital inpatients, Sexton, Maere and Dahl (1989) reported reduced
3 anxiety for both groups but better adherence to the walking programme.
4
5
OTHER MODERATORS
6
7 One criticism of the literature is that it may be impossible to separate the
8 real effects of exercise on anxiety from expected effects (Ojanen, 1994). In
9 other words, will exercise reduce anxiety even if participants don’t expect
30111 it to? Only a few studies have attempted to control for expectancy (e.g.
1 Steptoe, Edwards, Moses, & Mathews, 1989). In an interesting study by
2 McAuley, Bane, Ruddolph and Lox (1995), those with higher initial
3 expectancy reduced their social physique anxiety (SPA) more than those
4 with lower expectancy. Similarly, those with greater belief in their ability to
5 walk (self-efficacy) also reduced SPA by more. In the past, researchers have
6 focused on objective exercise and group characteristics in search of moder-
7 ators. The work by McAuley et al. reflects an increasing concern for the psy-
8 chosocial factors which may determine the strength of the anxiety-reducing
9 effects. For example, Wankel (1993) discussed the importance of enjoyment
40111 during exercise as a determinant of changes in mental health outcomes.
1
2
TREATMENT COMPARISONS
3
4 If aerobic exercise reduces anxiety, are these effects greater than other
45111 anxiety-reducing interventions? Only four studies addressed this question.
26 Adrian Taylor
1111 The alternative treatment included, yoga, t’ai chi, cognitive-behavioural
2 therapy, and relaxation. Fisher and Thompson (1994) reported enhanced
3 scores on the Physical Appearance State and Trait Anxiety Scale, among
4 females with low appearance evaluation, following either exercise therapy
5 or cognitive-behavioural therapy after six weeks. Topp (1989) also reported
6 reduced test anxiety after seven weeks of either aerobic running or relax-
7 ation. Blumenthal et al. (1991) and Brown et al. (1995) reported no
8 reduction in anxiety (using Spielberger’s Trait Anxiety Inventory)
9 following either aerobic exercise or alternative therapies (see Table 2.2).
1011 These findings confirm Petruzzello et al.’s (1991) review that suggested
1 aerobic exercise has no greater benefit in reducing anxiety than alterna-
2 tive anxiety-reducing treatments. However, only aerobic exercise groups
3111 increased cardiovascular fitness. There are, of course, many other health
4 benefits known to result from physical activity, such as reduced risk of
5 cardiovascular disease, osteoporosis, and colon cancer (Bouchard,
6 Shephard, & Stephens, 1994).
7 There is clearly a need for more studies to compare exercise and other
8 treatments but if the anxiety-reducing effects are similar, unsupervised
9 exercise could be a cheaper alternative. Adherence to different treatments
20111 should also be considered.
1
2
Mechanisms for anxiety-reducing effects of chronic exercise
3
4 A number of mechanisms have been proposed in the literature, including
5 the development of more positive affect and cognitions associated with
6 lower anxiety. These might involve positive mood states, self esteem and
7 perceptions of the physical self, and perceptions of control over behaviours
8 and outcomes (see Biddle, this volume; Fox, this volume). Exercise train-
9 ing may also develop enhanced coping resources, and social networks which
30111 may, in turn, reduce anxiety. While fitness changes appear not to be essen-
1 tial for anxiety reduction, it is likely that improved physical health status
2 will enhance health-related quality of life and reduce fear and anxiety often
3 associated with ageing (Rejeski, Brawley, & Schumaker, 1996). Improved
4 health status may be due to enhanced immunological, cardiovascular, neu-
5 rological and physical functioning associated with physical activity.
6 A further suggested mechanism is through the accumulated effects of
7 single exercise sessions. The next section will examine the effects of a
8 single exercise session on state anxiety.
9
40111
The effects of acute exercise
1
2 The 24 studies involving a single exercise session are summarised in Table
3 2.3. Changes in self-reported anxiety are shown together with psycho-
4 physiological measures (e.g. blood pressure) which provide an insight into
45111 some of the mechanisms for anxiety-reduction effects.
1111 Only three studies failed to show an anxiety-reducing effect on self-
2 report measures (Head, Kendall, Ferner, & Eagles, 1996; O’Connor,
3 Petruzzello, Kubitz, & Robinson, 1995; Youngstedt, Dishman, Cureton,
4 & Peacock, 1993) from pre- to post-exercise. Eight of the nine studies
5 that reported effect sizes showed greater anxiety-reducing effects of exer-
6 cise than the overall mean effect size (0.24) reported in the meta-analysis
7 by Petruzzello et al. (1991).
8
9
PARTICIPANT CHARACTERISTICS
1011
1 It is clear that most studies involved convenience samples with only four not
2 involving college students. Twenty studies involved males and 12 included
3111 females. Only single studies were located involving each of the inactive,
4 physically challenged or high test-anxious people. The research, therefore,
5 offers little generalisability to the general population of adults and special
6 groups. The only study (Petruzzello, Jones, & Tate, 1997) to compare active
7 with inactive individuals revealed anxiety-reducing effects for both (from 6
8 to 30 minutes post-exercise) with greater effects for the active (ES 1.04)
9 than the inactive (ES 0.56). Anxiety increased during exercise among the
20111 inactive group, suggesting a need for practitioner support and different cog-
1 nitive strategies used by the exercise neophyte.
2
3
RESEARCH DESIGN
4
5 Most studies (14) involved within-subject designs (i.e. participants were
6 tested in all conditions, serving as their own control), while three studies
7 involving random assignment to different groups, three had intact groups,
8 and three had no control group. Nine studies involved comparison with
9 a non-exercise control group/condition and two also contrasted the effects
30111 of exercise with another anxiety-reducing treatment.
1 Across the 24 studies, anxiety was measured between 0 and 120 minutes
2 post exercise, with 14 studies using more than one measurement time.
3 Petruzzello et al. (1991) reported the lowest ES (though not significantly
4 less) for assessments up to 5 minutes post-exercise and yet five studies in
5 the present review only measured anxiety immediately after exercise.
6
7
EXERCISE CHARACTERISTICS
8
9 The studies finding no anxiety-reducing effects were all unique in terms
40111 of exercise characteristics, and the findings largely confirm previous liter-
1 ature (Petruzzello et al., 1991). Head and colleagues (1996) reported that
2 participants had not reduced anxiety 15 minutes after a 60 minute low
3 intensity treadmill walk. In the study by O’Connor and colleagues (1995),
4 trained participants performed maximal fitness tests under two different
45111 protocol. Although Petruzzello et al. identified no detrimental effect of
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Table 2.3 Effects of acute exercise on anxiety
Author(s) Participants Design Comparison group Exercise Anxiety Outcome/comments

characteristics measure
Berger & 74, m, f Pre-post (0 mins) 1. Pool (swim) 1 = 81%, 20 mins, in pool SAI Swim and recording ↓ Anx.
Owen, college Intact class or pool grps 2. Pool (record (i.e. ‘time out’ support)
1992 healthy r.a. to swim/record keeping) Only women ↓ Anx.
3. Class control
Brown, 10 m, f Pre-post (2/3 mins) Ex. v quiet rest Aerobic cycle or treadmill SAI Quiet rest ↓ SAI, SBP, DBP
Morgan, & college Within subj. r.a. to Phys: BP Ex. ↑ SBP, ↓ SAI
Raglin, 1993 Physically both conditions
challenged
Crocker & 85 m, f Pre-post (0 mins) 1. Aerobic ex. (AE) AE = 70–80% HR, 20 min, SAI AE and AR ↓ SAI
Grozelle, college r.a. to 3 grps 2. Autogenic aerobic program No gender differences
1991 healthy relaxation (AR) AR = 30 mins
3. Control (C) C = free to wander for 30 mins
Dishman, 23 m Pre-post (1 min) 1. low fit Preferred intensity SAI Ex. ↓ SAI (only for high fit)
Farquhar, college between 2 grps (VO2 = 43ml. (no difference in RPE between
& Cureton, healthy kg.min.) fitness level)
1994 2. High fit 20 mins, cycle erg.
(VO2 = 57 ml.
kg.min.)
Doan et al., 52 m, f Pre-post (0 mins) 1. Aerobic ex. (AE) AE = moderate, 15 mins, cycle erg. MAACL AE ↓ MAACL (ES) = –0.80 cf.
1995 college r.a. to 3 grps 2. Relaxation (R) R = blindfold, relaxation tape control)
healthy, 3. Control (C) C = magazine reading R ↓ MAACL (ES) = –0.70 cf.
high test- control)
anxious
Head et al., 20 m Pre-post (15 mins) 1. Placebo 50%, 60 mins, treadmill walk with SAI No ↓ SAI
1996 college Within subj. r.a. to 2. & 3. Low/high placebo drug or 2 types of (for ex. & drug conditions)
healthy all 5 conditions propranolol betablockers (low/high dose)
Double blind 4. & 5. Low/high
metoprolol
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
McAuley, 34, m, f Pre-during-post (0, 15 1. Lab ex. (LE) 1. & 2. RPE = 14–16, 20 mins, SAI LE ↓ SAI (ES = –0.82)
Mihalko, & college mins) 2. Natural aerobic NEE ↓ SAI (ES = –0.60)
Bane, 1996 healthy Within subj. r.a. to environment ex. 3. Quiet rest Somatic items ↑ during, ↓
all 3 conditions (NEE) after ex.
3. Control (C) Cog. items ↓ during & after ex.
O’Connor & 14 m Pre-post (10, 20 min) 1. 0800 hrs Ex. at each time = 70%, 20 mins SAI Ex ↓ SAI and SBP
Davis, 1992 college Within subj. r.a. to 2. 1200 hrs treadmill run Phys: BP (at both 10 and 20 mins post ex.)
healthy all 4 conditions 3. 1600 hrs No effect of time of day
4. 2000 hrs
O’Connor 14 f Pre-post (1–120 mins) 1. Control 2, 3, 4 = 6 exercises 10 reps SAI No ↓ HR and BP
et al., 1993 college Within subj. r.a. to 2., 3., 4. Weight 3 sets, 30 mins at different loads Phys: BP Pre-post (120 min):
healthy all 4 conditions training (2 = 40%, 3 = 60%, 4 = 80% HR 40% ↓ SAI (ES = –0.25)
of max.) 60% ↓ SAI (ES = –0.58)
80% ↓ SAI (ES = –0.43)
O’Connor 16 m Pre-post (2, 10 mins) None 2 min stages to maximum SAI Max. ex. ↓ SAI (ES = –0.68 at
et al., 1995 college intensity post-2 min, –0.98 at post-10 min)
healthy
O’Connor 12 m Pre-post (10 mins) 2 protocols for max. 1. 3 min stages to maximum SAI No ↓ in SAI for either
et al., 1995 28 yrs Within subj. r.a. to test 2. 5 min stages to maximum protocol. ES = –0.14 to 0.52
trained both conditions
Petruzzello 19 m Pre-post (10–30 mins) None 75%, 30 mins, treadmill run SAI Ex. ↓ SAI (at 10–30 min post)
& Landers, college Phys: Ex. ↑ left frontal activation (cf.
1994a active EEG right frontal)
S’s with initial ↑ left sided
activation
↓ SAI more (at 30 min post)
Petruzzello 16 m Pre-post (0–30 mins) 1. 15 min ex. Both at 75%, treadmill run SAI Both conditions ↓ SAI (at 10,
& Landers, college Within subj. r.a. to 2. 30 min ex. 20, 30 mins). ES = –0.53 to
1994b active both conditions –0.96 (less at 10 mins)
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Author(s) Participants Design Comparison group Exercise Anxiety Outcome/comments

characteristics measure
Petruzzello 20 m Pre-post (0–30 mins) 1. Cooler situ All at 75%, 30 mins treadmill run SAI All 3 conditions ↓ SAI (at 20
et al., 1993 college Within subj. r.a. to 2. Normal situ & 30 mins post). ES = –0.52
active all 3 conditions 3. Warmer situ to –0.98) (less at 10 mins with
↑ SAI for warmer situ)
core body temp. not
necessary for SAI
Petruzzello 18 active Pre-during-post (6–30 Active v inactive RPE = 13, 24 mins, cycle erg. SAI Both grps ↓ SAI (at 6–30 mins).
et al., 12 inactive mins) ES = –1.04 (active)
(1997) college ES = –0.56 (inactive)
No. diff. bet. grps but inactive
↑ SAI during ex.
Petruzzello 20 m,f Pre-post (0–30 mins) 1. No ex. control 30 mins, cycle erg. SAI 55% and control didn’t ↓ SAI
·
& Tate, college Within subj. r.a. to all 2. 55% VO2max Phys: 70% ex. ↓ SAI (at 5 and 10
·
1997 trained 3 conditions 3. 70% VO2max EEG min post) predicted by greater
relative left frontal activation
pre-ex.
·
Raglin & 25 m,f Pre-post (5, 60, 120 mins) 1. 40% VO2max 20 mins, cycle erg. SAI All conditions ↓ SAI (at 60
·
Wilson, 24 yrs Within subj. r.a. to all 3 2. 60% VO2max and 120 mins post)
·
1996 trained conditions 3. 70% VO2max 70% ex ↑ SAI (at 5 min post)
but only S’s with a low pre-ex.
SAI
Raglin, 25 m,f Pre-post (0, 20, 60 mins) Cycling v wt. training 70–80%, 30 mins SAI Only cycling ↓ SAI (at 60 min
Turner, & college Within subj. r.a. to both Phys: BP post)
Eksten, 1993 trained conditions No ↓ DBP
Rejeski, 30 m Pre-during-post None 75%, 15 mins, treadmill run SAI ex ↓ SAI (at 10 min post)
Hardy, & college (10 mins) AD-ACL ex ↓ AD-ACL (energetic
Shaw 1991 healthy arousal)
ex ↑ AD-ACL (tiredness &
calmness)
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Roth, 40 active, Pre-post (15 mins) Ex. v waiting grp Ex. = 60–80% (6–8 mins) POMS Ex ↓ Anxiety/tension
1989 40 inactive r.a. to 2 grps + 50–65% (12–14 mins) (for both active and inactive)
m, f
college
healthy
Szabo et al., 40 runners 11.9 active days Runners studied Active days = 8 km in 40.6 mins SAI SAI lower on days when
1998 compared with in situ on average participants ran, cf. didn’t run.
9.1 inactive days
Tate & 20 m, f Pre-during-post (5–30 1. No ex. 30 mins, cycle erg. SAI ex. ↑ SAI (during ex.)
Petruzzello, college mins) 2. 55% AD-ACL Only 70% ex. ↓ SAI (at 30
1995 active Within subj. r.a. to all 3 3. 70% min post). ES = –0.39
conditions 55% & 70% ex. ↓ AD-ACL
(energetic arousal) during &
post
Trine & 30 m, f Pre-post (5–15 mins) 1. 0600 hrs Preferred intensity, run on indoor SAI ex. at all times ↓ SAI and DBP
Morgan, 41 yrs (m) Within subj. r.a. to all 3 2. 1100 hrs track Phys: HR, No effects of time of day,
1997 34 yrs (f) conditions 3. 1600 hrs No diff. between conditions BP preferred time to run, or
runners gender
Youngstedt 11 m Pre-post (15, 25 mins) 1. Cycling in 70%, 20 mins SAI No ↓ SAI or EEG (alpha,
et al., 1993 college Within subj. r.a. to all warm water Phys: BP, beta, theta wave)
fit 4 conditions 2. Cycling in EEG 1 & 3 ↓ mean BP (at 15 & 25
cold water mins post ex.)
3. Sitting in
warm water
4. Quiet rest
Notes:
Per week (p.w.); State & Trait Anxiety Inventory (SAI & TAI: Spielberger et al., 1983); Profile of Mood States (POMS: McNair et al., 1971) – Tension/Anxiety
scale only; Multiple Affect Adjective Check List – Anxiety (MAACL-A: Zuckerman & Lubin, 1965); Activation-Deactivation Adjective Check List (AD-
ACL): Thayer, 1967; male (m); female (f); effect size (ES); random assignment (r.a.); ergometer (erg.); physiological measure (phys.); systolic & diastolic blood
pressure (SBP & DBP); electroencephalography (EEG); Ratings of Perceived Exertion (RPE: Borg, 1973). Under design column, time(s) in brackets is time(s)
of post-assessment(s) after exercise ends.
32 Adrian Taylor
1111 high intensity exercise (> 80%) on anxiety reduction, O’Connor’s study
2 suggests that there is a ceiling at which benefits can be gained. They
3 reported some non-significant increases in anxiety 10 minutes after exercise
4 (ES 0.52). Youngstedt et al. (1993) also reported no reduction in state
5 anxiety but participants cycled in a pool, a rather unnatural environment,
6 to test the effects of temperature, as an anxiety-reducing mechanism.
7 Recent attention has focused on the effects of the exercise environment
8 on anxiety reduction. McAuley, Mihalko, and Bane (1996) revealed that
9 both exercising in a laboratory and in a natural environment reduced
1011 anxiety. O’Connor and Davis (1992), and Trine and Morgan (1997), inves-
1 tigated the effects of time of day on anxiety reduction. They both
2 concluded that time of day, whether it be actual or different from preferred
3111 time, had no influence.
4 Only two studies involved weight training. Raglin, Turner and Eksten
5 (1993) compared weight training with cycling and found only the latter
6 led to a reduction in anxiety. This confirmed Petruzzello et al’s (1991)
7 meta-analysis, albeit with very few studies. However, O’Connor, Bryant,
8 Veltri and Gebhardt (1993) compared a non-exercising control group with
9 weight training, holding repetitions constant but setting intensity at 40%,
20111 60% and 80% of one maximum repetition. ESs were 0.25, 0.58, and
1 0.43, but only 120 minutes after exercise completion.
2
3
TREATMENT COMPARISONS
4
5 Only two studies (Crocker and Grozelle 1991; Doan, Plante, DiGregorio,
6 & Manuel, 1995) compared the anxiety-reducing effects of an aerobic
7 exercise group with a relaxation group. Both groups reduced anxiety
8 compared with the control group, thereby supporting Petruzzello et al.’s
9 meta-analytic findings.
30111
1
Mechanisms for anxiety-reducing effects of acute exercise
2
3 A number of physiological changes occur from pre- to post-acute exercise
4 and these have been suggested as possible mechanisms for anxiety-reducing
5 effects. These include temperature increase, increased beta endorphins,
6 reduced muscle tension, increase in parasympathetic activity, and reduced
7 excitability of the central nervous system. The findings largely confirm those
8 revealed by Petruzzello et al. (1991) in that exercise-related anxiety-reducing
9 effects are associated with changes in psychophysiological measures,
40111 although these are less apparent for blood pressure and heart rate measures.
1 Another proposed mechanism for anxiety-reducing effects has been the
2 ‘time-out’ or ‘distraction’ hypothesis. In support of this, Brown, Morgan
3 and Raglin (1993) reported reduced anxiety following both quiet rest and
4 exercise (following earlier studies by Raglin & Morgan, 1987, and Ruck
45111 & Taylor, 1991). This suggests that any distracting activity may result in
1111 reduced anxiety. However, it may be that some forms of exercise (perhaps
2 with greater external stimulation such as music) may provide a greater
3 guarantee that distraction will occur, and of course there are other benefits
4 from physical activity. The anxiety-reducing effects of exercise may also
5 last longer.
6
7
Effects of exercise on reactivity to stress
8
9
Summary of findings from cross-sectional exercise studies
1011
1 Nine of the fourteen cross-sectional studies reviewed since 1989 showed
2 that fit and/or active individuals were less reactive to psychosocial stressors.
3111 Due to a variety of possible confounding factors, this literature will not be
4 reported in detail here (Blaney, Sothmann, Raff, Hart, & Horn, 1990;
5 Boutcher, Nugent, & Weltman, 1995; Buckworth, Dishman, & Cureton,
6 1994; Choi, & Salmon, 1995; Czajkowski et al., 1990; de Geus, Lorenz, van
7 Doornen, de Visser, & Orlebeke, 1990; de Geus, Lorenz, van Doornen, &
8 Orlebeke, 1993; Graham, Zelchner, Peacock, & Dishman, 1996; Houtmann,
9 & Bakker, 1991; Long, 1991; McCubbin, Cheung, Montgomery, Bulbulian,
20111 & Wilson, 1992; Sothmann, Hart, & Horn, 1991; Steptoe, Moses, Mathews,
1 & Edwards, 1990; Van Doornen, & de Geus, 1989).
2
3
Summary of findings from chronic exercise studies
4
5 Twelve studies, published since 1988, were identified which compared
6 reactivity to and/or recovery from a psychosocial stressor pre- and post-
7 training. A summary of these studies is shown in Table 2.4.
8 All studies showed improvements in aerobic fitness over periods from
9 5 to 16 weeks with the exception of one that examined the effects of exer-
30111 cise withdrawal. This study lasted only one week and showed no change
1 in reactivity (Szabo & Gauvin, 1992). Six studies showed no effects and
2 six showed some positive effect of training on psychological and physio-
3 logical measures during and/or after a stressor. Given the wide variety of
4 exercise training in terms of frequency, intensity and duration, the nature
5 of the stressors (i.e. passive v. active response, novel v. familiar, natural
6 v. simulated), and the indices of psychophysiological reactivity, it is diffi-
7 cult to draw conclusions. However, Claytor (1991) suggested that reactivity
8 differences were only likely to be observed when the task was familiar.
9 In another recent review, Sothmann et al. (1996) suggested that more
40111 studies were needed involving clinical populations with compromised stress
1 systems (e.g. depressed, anxious, etc.), and studies to examine gender
2 differences given that the neuroendocrine stress response may be different
3 for males and females. They also suggested that consideration of how
4 exercise training may impact on both the sympathetic and parasympathetic
45111 response should be considered more in the future.
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Table 2.4 Effects of chronic exercise on psychosocial stress reactivity
Author(s) Participants Design Stressor Exercise characteristics Anxiety Outcome/comments

measure
·
Blaney 14 m Pre (TI) – post (T2) Stroop test (18 mins) Ex. training group ↑ VO2max from SAI No in reactivity or recovery
–1 –1
et al., 1990 42 yrs 16 wks 45 → 53 ml.kg. .min TAI on all variables, from T1 to T2
(35–50) self-select to 2 grps Control grp = 42 & 43 ml.kg–1.min–1 Phys:
healthy ACTH,
active cortisol,
HR
Blumenthal 37 m Pre (T1) – post (T2) Maths task Aerobic ex. (AE) grp ↑ Phys: AE less reactive (HR, DBP, E)
·
et al., 1990 42 yrs over 12 wks 3 5 min blocks VO2max by 14% HR, BP, from T1 to T2
(30–52) r.a. to 2 grps Strengths training (ST) grp by 3% E, NE, AE recoverd faster from T1
·
healthy, ↑ VO2max RPP to T2 cf. ST grp
Type As
Calvo, 21m, 58f (pre (T1) – post (T2) Speech, Control = waiting list (with some 1. Overt Ex. grp reduced 1 & 2 from
Szabo, & college, over 12 wks maths, contact) behaviour pre-post and cf. control grp.
Capafons, healthy r.a. to 2 grps fine motor task Ex. grp = Combined strength, 2. CSAQ Ex. grp had faster HR
1996 flexibility & endurance training, 3. HR recovery, post-stress
·
VO2 ↑
·
de Geus 22 m Pre (T1) – post (T2) 1. Memory search Aerobic ex. grp ↑ VO2max Phys: HR, No in reactivity or recovery
et al., 1990 college over 7 wks task from 46.6 → 51.9 RR, PEP, on all variables from T1 to T2
inactive, r.a. to 2 grps 2. Tone avoidance No. ex. grp didn’t change RSA
healthy task SAI
3. Cold pressor test
2
1
1
9
8
7
6
5
4
3
2
1
1
9
8
7
6
5
4
3
2
1
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
4011
3011
2011
3111
1011
1111
de Geus 62 m Pre (T1) – mid (T2) 1. Memory search Training + training grp:
·
Phys: HR, Training ↓ HR and DBP
et al., 1993 33 yrs – post (T3) over task VO2max = 47 → 52 → 54 BP, TPR, overall but not reactivity to
(24–40) 12 + 12 wks 2. Tone avoidance Training + de-training grp: PEP stressors
·
inactive, r.a. to 2 grps task VO2max = 47 → 51 → 50 No sig. moderators
healthy self select 3. Cold pressor test No training + training grp:
·
to 2 grps VO2max = 45 → 45 → 52
No training + no train. grp:
·
VO2max = 44 → 43→ 44
Kubitz & 24 m,f Pre (T1) – post (T2) Pre-stress – recovery AE = 60–80%, 40 mins 3 p.w., SAI No training effects on SAI
Landers, college r.a. to 2 grps at T1 and T2 8 wks, cycle erg. Phys: or BP
1993 healthy (AE, C) Stressor = Stroop test C = 8 wks HR, RSA, Ex. training ↓ HR and changed
(3 min) and maths BP, EEG RSA and alpha laterality
(3 min) during stressor
LaPerriere 50 m Pre-post ex. Notification of HIV Ex. = 80% (3 mins) + 60–70% POMS Only controls with +ve HIV
et al., 1990 18–40 yrs r.a. to 2 grps then test results (2 mins) 45 mins, 3 p.w., Phys: results ↑ anx. (ie. ex.
inactive, HIV test aerobic fitness gains over 5 wks immune attenuated anxiety response
healthy, control = no. ex. profile to natural stressor). Also
HIV risk supported by immune profile
results
Sherwood, 27 m Pre (T1)-post (T2) Letter response task Aerobic training grp Phys: AE training for hypertensives
·
Light, & 41 yrs over 12 wks in competition with (VO2max = 34 → 38) HR, BP, ↓ DBP, HR, TPR and DBP
Blumenthal, (33–56) r.a. to 2 grps others Strength training grp TPR, CO reactivity to stressors
·
1989 Type A’s (5 mins) (VO2max = 34 → 35)
Stein & 33 m Pre (T1)-post (T2) 1. Passive responding Aerobic training (peak Phys: Training ↓ HR overall and
·
Boutcher, 46 +/– over 8 wks 2. Push button Stroop VO2 = 2.9–3.3 l.min–1) HR, BP, reactivity to stressors
1992 6 yrs r.a. to 2 grps 3. Verbal Stroop No ex. controls (no change) R-R int, No change in other variables
inactive, (11 mins total) PTT,
healthy TWA,
RWA,
Temp
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111

measure
·
Steptoe 20m, 55f Pre (T1) – post (T2) Easy and hard Vig. aerobic (est. VO2max ↑ 22%) Phys: Training → no change in TAI
·
et al., 39 yrs over 10 wks problem solving Mod. aerobic (est. VO2max ↑ 12%) HR, BP, No diff in reactivity to
·
1990 (18–60) r.a. to 4 grps (6 mins each) Weight training (est. VO2max ↑ 6%) RR stresors in any variable
inactive. No ex. control (no change) TAI
healthy
Steptoe, 3 m, 29 f Pre (T1) – post (T2) Easy and hard Mod. aerobic ex. Phys: AE training – no differences in
·
Moses et al., 37 +/– over 10 wks problem solving (VO2max = 41 → 48) HR, BP, reactivity to stressors in any
1993 9 yrs matched 2 grps (6 mins each) Non-aerobic ex. (no change) SCL variable from T1 to T2
(20–60) Raven’s progressive
anxious, matrices
overwght
Szabo & 16 m, 8 f Pre (T1) – post (T2) Math task (5 min) 1. Regular ex. grp Phys: HR Ex. withdrawal had no effect
Gauvin, college over 1 week 2. Ex. withdrawal grp on HR reactivity to stressor
1992 v. active, r.a. to 2 grps
healthy
Notes:
Pulse transit time (PTT); skin temperature (Temp); Pre-injection period (PEP); Cardiac output (CO); Norepinephrine (NE); T-wave amplitude (TWA); R-wave
amplitude (RWA); Skin conductance level (SCL); Total peripheral resistance (TPR); diastolic blood pressure (DBP); Respiration rate (RR), Epinephrine (E);
Rate pressure product (RRP); inter heart beat interval (R-R int); respiratory sinus arrhythmia (RSA); electroencephalography (EEG); adrenocorticotrophic hor-
mone (ACTH); State & Trait Anxiety Inventory (SAI & TAI: Spielberger et al., 1983); Cognitive–Somatic Anxiety Questionnaire (CSAQ: Calvo et al., 1990);
Profile of Mood States (POMS: McNair et al., 1971); random assignment (r.a.); group (grp); weeks (wks); years (yrs); male, female (m, f); Stroop test requires fast
response only when rapidly presented word matches colour of word.
1111 Summary of findings from acute exercise studies
2
Fourteen studies, published since 1988, were identified that compared reac-
3
tivity to and/or recovery from a psychosocial stressor following a single
4
exercise session. A summary of these studies is shown in Table 2.5.
5
Only four studies failed to show a reduction in reactivity to a stressor
6
following exercise. Exercise sessions lasted between 10 and 120 minutes,
7
and were performed at between 28–80% of maximum heart rate. The
8
stressors were largely brief (3–5 minutes duration) and included both
9
passive and active responses. A wide range of biochemical, cardiovascular,
1011
psychological and cerebral measures were employed. All studies involved
1
healthy participants, mostly of college age.
2
As examples of the type of study showing a positive outcome, Rejeski,
3111
Thompson, Brubaker and Miller (1992) reported that exercising on a
4
cycle ergometer for 40 minutes at 70% of maximum heart rate, followed
5
by 30 minutes of rest before a public speech, led to less increase in
6
systolic and diastolic blood pressure during the stressor, in comparison
7
with a non-exercising control condition. Also, Steptoe, Kearsley and
8
Walters (1993) compared a light exercise group with groups exercising on
9
a cycle ergometer at 50% and 70% of their maximum oxygen uptake.
20111
They reported lower systolic blood pressure following the public speech
1
among the 70% group compared with the light exercise group.
2
3
4 Implications for the researcher
5
The review conducted provides ideas for further research. In the broadest
6
terms, research should examine the anxiety-reducing effects of exercise
7
among specific sub-groups of the population and more carefully consider
8
the social context in which exercise takes place. Hopefully we can deter-
9
mine, through experimental manipulation of cognitive processes such as
30111
self-evaluation, how the exercise practitioner can increase the anxiety-
1
reducing effects of PA.
2
Many questions remain about the importance of reducing psychophys-
3
iological reactivity to psychosocial stressors through exercise training or
4
a single session of exercise. For example, Carroll, Smith, Sheffield, Shipley
5
and Marmot (1995) suggested that reactivity to a psychological stressor
6
only accounted for 1% of the variance in blood pressure 4.9 years later
7
when controlling for initial blood pressure among a large sample. However,
8
the authors did accept that the follow-up period was rather short, and
9
that reactivity among a younger sample may better predict hypertension
40111
in later life. Clearly, an implication is that more prospective studies are
1
necessary to examine how exercise may serve as a mediator in the devel-
2
opment of hypertension, particularly among those with a predisposition
3
to physiologically react more strongly to psychosocial stressors.
4
45111
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Table 2.5 Effects of acute exercise on psychosocial stress reactivity

measure
Boone et al., 8 41 +/– Within subj. randomly Stroop test (5 min) Ex. = 60%, 60 mins, treadmill Phys: Ex. ↓ MAP, SBP & DBP
1993 7 yrs ordered to (ex. or no No ex. = quiet rest, 60 mins BP, HR reactivity to stressor
borderline ex.) then pre-during- Both followed by 10 mins rest
hyper- post stressor
tensives
Doan et al., 52 m,f r.a. to 3 grps (AE, R, C) IQ test (3 min) AE = mod., 15 mins, cycle erg. MAACL C ↑ MAACL (cf. AE & R)
1995 college, then pre-post stressor R = blindfold, relaxation tape Effects of AE greater for m
healthy C = magazine reading (cf. f)
high test-
anxious
Flory & 18 f Within subj. (AD + SP) Study period (40 mins) AD = 28%, 20 mins dance MAACL No diff. in MAACL at post task
Holmes, college then (15 mins) pre-post SP = 40 mins ‘studying’ Phys: AD ↑ HR (cf. SP)
1991 healthy stressor HR, BP
Hobson & 80 f Within subj. randomly Modified Stroop test C = no ex. control Phys: BP 40 min ex. ↓ DBP & MAP
Rejeski, college ordered to C, 10, 25 (3 min) following 10 & 25 & 40 mins. ex. on reactivity to stressor
1993 healthy & 40 min then pre- 20 mins rest cycle erg at 70% cf. no ex. condition
during stressor No effects for 10 & 25 min
conditions
Jin, 1992 96 m, f r.a. to 4 grps MS = math & tests TC = t’ai chi, 60 mins SAI All 4 grps ↓ cortisol
36 yrs (TC, TC-V, W,C). (60 mins) TC-V = t’ai chi video, 60 mins Phys: Only TC ↓ SAI (cf. control)
healthy Within subj. (MS + ES = horror film W = walking, 60 mins, 6 km/h HR, BP, but not after expectancy
from ES) then treatment (60 mins) C = neutral reading CA effects controlled
t’ai chi On different days Longer post treatment period
clubs needed
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Kubitz & 28 m, f r.a. to 2 ghrps (AE, C) Vigilance task AE = 75–80%, 15 mins, cycle erg. AD-ACL AE ↓ beta & ↑ alpha & theta
Pothakos, college then (5 min) pre-post (15 mins) C = relaxation tape Phys: waves
1997 healthy stressor EEG AE ↑ AD=ACL
All effects only after ex. and
at 5 mins in stressor
Perronet 7m Pre-post Stroop test (3 mins) 1. 53%, 120 mins, cycle erg. Phys: Ex ↓ E (by 50%)
et al., 1989 23 yrs Within subj. r.a. to both 2. rest (120 mins) HR, BP,
healthy conditions E, NE
Rejeski, 12 m Within subj. randomly Modified Stroop test C = no ex. control Phys: HE had ↓ SBP during stress
Gregg, et al., 31 yrs ordered to C, LE, HE, (2 3 mins) LE = 50%, 30 mins, cycle erg. HR, BP (cf. LE & C)
1991 (23–38) then pre, during, and following 30 mins rest HE = 80%, 60 mins, cycle erg. LE & HE had ↓ DBP during
highly post stressor stress (cf. C)
trained HE < LE < C on MAP
cyclists during stressor
No differences post stressor
Rejeski 48 f Within subj. random Modified Stroop test C = no ex. control Phys: Ex. had ↓ SBP & DBP during
et al., 1992 25–40 yrs ordered to C, Ex., then 3 mins) + Public Ex. + 70% HR, 40 mins cycle erg. HR, BP stressors (cf. C)
low & pre-during-post stressor speech (3 mins) MAACL-R- Ex. had fewer & less intense
mod. fit, following 30 mins rest Anxiety anx.-related thoughts prior to
healthy Anxiety – speech (not Stroop)
self report Ex. had no effect on recovery
rates or MAACL
Roth, 1989 40 active Stressor 1 then r.a. to ex. Numerical tasks Ex. = 60–80% (6–8 mins) Phys: Ex. had no effect on HR and
40 inactive, v control then stressor 2 + 50–65% (12–14 mins) HR, BP BP reactivity to stressors
m, f, Control = waiting POMS stressors
college No diff. between active v
healthy inactive or fit v unfit
Roy & 30 m r.a. to 3 grps (C, LE, HE) Numerical tasks (4 5 C = no ex. control SAI No effects on SAI & CSAQ
Steptoe, college then pre-during-post mins) following 20 mins LE = 25 Watts, 20 mins, cycle erg. Phys: C > LE > HE on SBP, DBP,
1991 healthy stressor rest He = 100 Watts, 20 mins, cycle erg. HR, BP HR reactivity to stress
CSAQ
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111

measure
Sedlock & 58 f 4 grps then (1–10 mins) Numerical task (5 mins) 1. High TAI / low fit Phys: HR Fitness level had no mediating
duda, college pre-during-post stressor follwing 20 mins rest 2. Low TAI / low fit ASDS effect on reactivity or recovery
1994 healthy 3. High TAI / high fit to a stressor following ex.
4. Low TAI / high fit High TAI had higher HR and
All grps = 50%, 15 mins, cycle erg. ASDS post stressor
Steptoe, 36 inactive r.a. to 3 grps (C, LE, HE) Numerical tasks C = light ex. control Phys: Only HE had lower SBP during
Kearsley, 36 active then pre-during-post (5 mins) + Public LE = 50%, 20 mins, cycle erg. HR, BP, and post maths (cf. C), lower
& Walters, 27 yrs stressors speech (3 mins) HE = 70%, 20 mins cycle erg. BRS SBP post speech (cf. C), lower
1993 (20–35), m mixed order, following POMS BRS & DBP post maths (cf. C)
30 mins rest No differences for HR
No diff. between active v
inactive
Szabo 9m Within subj. to both ex. 1. Numerical task Ex. = 60%, 30 mins cycle erg. SAI Ex. had no effect on reactivity
et al., healthy or film then during (3 mins) Film = control = neutral Phys: to stressor
1993 stressors 2. Stroop test (3 mins) HR, BP,
E, NE
Notes:
Pulse transit time (PTT); skin temperature (Temp); Pre-injection period (PEP; Cardiac output (CO); Norepinephrine (NE); T-wave amplitude (TWA); R-wave ampli-
tude (RWA); Skin conductance level (SCL); Total peripheral resistance (TPR); Respiration rate (RR); Epinephrine (E); Rate pressure product (RRP); inter heart beat
interval (R-R int); systolic/diastolic blood pressure (SBP/DBP0; heart rate (HR); Mean arterial pressure (MAP); electroencephalography (EEG); Baroceptor reflex sensi-
tivity (BRS); State & Trait Anxiety Inventory (SAI & TAI: Spielberger et al., 1983); Cognitive–Somatic Anxiety Questionnaire (CSAQ: Calvo et al., 1990); Multiple
Affect Adjective Check List (MAACL: Zuckerman & Lubin, 1965); Profile of Mood States (POMS: McNair et al., 1971); Adjective Semantic Differential Scale (ASDS:
Hull et al., 1984); Activation-Deactivation Adjective Check List (AD-ACL: Thayer, 1967; random assignment (r.a.); male, female (m, f).
1111 Implications for the exercise practitioner
2
3 The physical benefits of exercise have received the greatest attention in
4 the past. This review provides ample evidence to support the use of exer-
5 cise for the reduction of trait anxiety over a period of time, and for the
6 anxiety-reducing effects of a single exercise session. While it appears clear
7 that exercise of an aerobic rhythmic type is most beneficial, other social
8 factors need to be carefully considered to maximise the benefits.
9 Exercise sessions and programmes should provide a distraction from
1011 worry and anxiety-inducing thoughts and provide the exerciser with a sense
1 of mastery and achievement. If an anxious exerciser feels threatened or
2 under evaluation in an exercise setting then these perceptions should be
3111 discussed with an exercise leader with appropriate exercise counselling
4 skills. To avoid social anxiety, exercise testing and programming should
5 involve individual contact and goal setting in an environment that supports
6 positive change in self-perceptions. The promotion of physical activity for
7 anxiety reduction should be discussed and exercise sessions incorporated
8 into daily living. It may be that individuals find less motivation to exercise
9 during more intense periods of stress (e.g. prior to exams or deadlines). An
20111 exercise counsellor/leader should identify such risks of inactivity and
1 emphasise the importance of exercise for stress management.
2 There is some evidence that improvements in fitness can have a cross-
3 training effect, enabling the more fit to physiologically react less and
4 recover more quickly from stressors. The accumulated effects of reduced
5 reactivity may result in lower levels of stress-related fatigue from normal
6 occupation-related activities and other psychosocial demands. The impli-
7 cations of this are that individuals may feel less tired through enhanced
8 recovery from daily stressors. They may also be less likely to cope with
9 stress by other means such as smoking and alcohol consumption. There
30111 is some evidence that physical activity may be linked to reduced cravings
1 for a cigarette and improved success in smoking cessation (Ussher, Taylor,
2 West, & McEwen, in press).
3
4
Implications for the health analyst and policy maker
5
6 Despite the minimal reference to physical activity in a recent review of
7 the effectiveness of different mental health promotion interventions
8 (Tilford et al., 1997), there is considerable evidence, from rigorously
9 designed and conducted research, supporting the anxiety-reducing effects
40111 of physical activity. Evidence for a variety of physiological, psychological
1 and social mechanisms has partially supported the logical supposition that
2 these effects will occur. While the effects may be similar to other non-
3 medication anxiety treatments, the simplicity of promoting physical
4 activity, such as walking, and at low cost, makes the promotion of phys-
45111 ical activity for the treatment of various types of anxiety appealing.
42 Adrian Taylor
1111 A supervised exercise programme, such as in an exercise prescription
2 scheme, may provide an important setting for the initiation of an anxiety-
3 reducing intervention if the exercise practitioner has an adequate
4 understanding of the anxiety-reducing mechanisms involved.
5 In conclusion, summary statements of ‘what we know’ and ‘need to
6 know’ will be presented.
7
8
9
1011 What we know
1
• The literature suggests a low-to-moderate anxiety-reducing effect
2
of physical activity, with some studies suggesting a potentially
3111
greater effect.
4
• Studies have fairly consistently shown that a period of exercise
5
training can reduce trait anxiety.
6
• Single exercise sessions will result in reductions in state anxiety.
7
• The research evidence comes from studies employing various
8
designs but the stronger effects are shown by randomised
9
controlled trials.
20111
• The limited number of comparisons between physical activity
1
and other medication-free anxiety treatments suggest compa-
2
rable anxiety-reducing effects.
3
• Single sessions of moderate exercise can reduce short-term
4
physiological reactivity to and enhance recovery from brief
5
psychosocial stressors.
6
• Exercise training has successfully reduced trait anxiety across
7
a wide range of sub-groups in the population, including active
8
and inactive, anxious and non-anxious, healthy and unhealthy
9
individuals (e.g. undergoing cardiac rehabilitation, or with cancer,
30111
COPD, and a variety of mental disorders), and in both males
1
and females.
2
• Exercise training has been used successfully to reduce trait
3
anxiety in a wide range of clinical and non-clinical settings.
4
However, some anxiety disorders (e.g. panic disorder with agora-
5
phobia) may not respond to exercise training.
6
• Exercise training appears to have the greatest trait anxiety-
7
reducing effects when the duration is at least 10 weeks (with
8
greatest benefits over 15 weeks).
9
• Trait anxiety-reducing effects are not dependent on changes in
40111
physical fitness.
1
• Single (acute) exercise sessions appear to have the greatest
2
state anxiety-reducing effects when the exercise type is aerobic
3
and rhythmic.
4
45111
1111
2 What we need to know
3
4 • There has been relatively little examination of the anxiety-
5 reducing effects of PA for young people.
6 • There is scope for better understanding of how special groups
7 (e.g. the obese with social physique anxiety; asthmatics and
8 COPD patients who experience fears about breathing; older
9 people with a fear of falling) may benefit from a programme of
1011 exercise.
1 • Only a few studies have examined the trait anxiety-reducing
2 effects of exercise with sessions less than 20 minutes.
3111 • We need to know more about the long-term effects of accumu-
4 lated doses of activity (in line with current recommendations for
5 physical activity for cardiovascular disease prevention).
6 • Only a few studies have examined the long-term anxiety-reducing
7 effects (i.e. over 4 months) of PA. We need to know, for example,
8 whether a 10 week exercise programme will have lasting anxiety-
9 reducing effects, and if not what dose of exercise is necessary
20111 to maintain the effects.
1 • We know little about the anxiety-reducing effects of short bouts
2 (<15 minutes) of free-living, unsupervised aerobic physical
3 activity, which can be most easily integrated into an active
4 lifestyle, as a low-cost intervention.
5 • We know little about the anxiety-reducing effects of non-aerobic
6 exercise such as weight and circuit training.
7 • Few studies have examined the influence of social interactions
8 on anxiety in the exercise setting. We need to know whether
9 exercise practitioner manipulations of self-efficacy, outcome
30111 expectancy, perceived competence, goal setting, feedback, atten-
1 tional focus and perceived exertion and enjoyment can increase
2 anxiety-reducing effects, particularly among inactive and inex-
3 perienced exercisers.
4 • The cost-effectiveness of physical activity as an anxiety treat-
5 ment has not been considered. Studies need to compare PA
6 with other anxiety-reducing interventions, not only in terms of
7 anxiety reduction (which appears to be similar) but also in terms
8 of cost. Related to this would be careful consideration of adher-
9 ence to the respective interventions.
40111 • Fitness change has been extensively investigated as a mediator
1 in the PA–anxiety relationship, with equivocal findings. Other
2 possible mediators should be investigated including correlates
3 of trait anxiety such as hardiness/mental toughness, and coping
4 resources.
45111
44 Adrian Taylor
1111
2 • Adherence to exercise training appears to be greater when it is
3 of moderate intensity (e.g. walking), and integrated into an active
4 lifestyle. Nevertheless, we need to know much more about
5 the determinants of adherence to free-living and facility-based
6 exercise programmes.
7 • Further evidence is needed to show how improved fitness may
8 reduce cardiovascular, neuroendocrine, and cerebral reactivity
9 to and recovery from psychosocial stressors.
1011
1
Effects of a single session of exercise on state anxiety
2
3111 • Most studies have examined anxiety-reducing effects among
4 college-age participants in laboratory settings. We need to know
5 more about the effects of a single session of exercise in natu-
6 ralistic settings, such as in workplace gyms following anxiety-
7 evoking work, or in schools prior to or after examinations.
8 • Studies have generally shown limited anxiety-reducing effects
9 from a session of resistance exercise, but this may be due to
20111 limited follow-up periods. We need to know if the anxiety-reducing
1 effects are observed later than 2 hours post-exercise before
2 dismissing this form of activity.
3 • Few studies have examined the effects of low–moderate inten-
4 sity physical activity (e.g. slow walking). We need to know more
5 about the minimum intensity necessary for anxiety-reducing
6 effects.
7 • We need to know more about the anxiety-reducing effects of
8 different lengths of exercise sessions, particularly comparing
9 shorter (5–10 minute sessions) with longer sessions (20–30
30111 minutes).
1
2
Effects of exercise training on reactivity to stress
3
4 • Few studies have examined how people with low stress buffering
5 capabilities (e.g. the depressed, anxious) can reduce their reac-
6 tivity to and enhance recovery from stressors with exercise
7 training.
8 • Few studies have examined the effects of low–moderate inten-
9 sity physical activity on stress reactivity; most have focused on
40111 improving aerobic fitness.
1 • Few studies have examined the effects of exercise on naturally
2 occurring stressors which may elicit more intense stress reactions
3 and emotions such as anger, hostility, helplessness, and fear.
4
45111
1111
2 Effects of acute physical activity on reactivity to stress
3
• We need to know more about the dose (intensity and duration)
4
of exercise necessary for reducing reactivity, and how affective
5
outcomes from exercise interact with reactivity.
6
• We need to know how exercise can impact on naturally occur-
7
ring stressors, in terms of reactivity
8
9
1011
1 Acknowledgements
2
3111 Sincere thanks are expressed to Dr Helen Carter for her tireless assist-
4 ance in collecting the material and preparing the tables in the manuscript
5 while completing her Ph.D at the University of Brighton.
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
4
45111
1111
2 3 The relationship between
3
4 physical activity and
5
6
clinically defined depression
7
8 Nanette Mutrie
911
1011
1
2
3111 This review aims to provide a comprehensive picture of what we know about
4 the relationship between physical activity and clinically defined depression.
5 It will cover epidemiological evidence, evidence from meta-analytic reviews
6 and will highlight key studies in the area. Discussion is included on whether
7 or not a case can be made for a causal link between physical activity and
8 depression. Finally, what we know, what we need to know, and what future
9 research and practice should focus on will be itemised.
20111
1
Introduction
2
3 According to the NHS Health and Advisory Service (1995), one in seven
4 adults in the UK will suffer some form of psychiatric morbidity and the
5 prevalence of mental health problems among children is estimated at up
6 to 20% with 7–10 % having moderate to severe problems which prevent
7 normal functioning (Kurtz, 1992). Depression is one of the most common
8 psychiatric problems. An estimated 20% of consultees in primary care
9 report symptoms of depression (Paykel & Priest, 1992). In terms of work-
30111 place incidence of depression it has been suggested, by analysing American
1 employee health insurance data, that depression is the most common
2 complaint with a higher prevalence in women than men (Anspaugh,
3 Hunter & Dignan, 1996). It has been estimated that clinically defined
4 depression affects 5–10% of the population of most developed countries
5 (Weismann & Klerman, 1992). Taken together this evidence suggests a
6 large and expensive burden in healthcare resources in the treatment of
7 depression.
8 Definitions of depression range from episodes of unhappiness, which
9 affect most people from time to time, to persistent low mood and inability
40111 to find enjoyment that would probably be classified as clinical depression.
1 In addition, depression may be secondary to other medical conditions such
2 as alcohol addiction. Most cases of depression are treated in general prac-
3 tice but more severe cases are referred to psychiatric services. In the UK,
4 drugs continue to be the most frequently used treatment for depression
45111 although psychotherapy and ECT are also used (Hale, 1997). Hale (1997)
Physical activity and depression 47
1111 states that all anti-depressant drugs are equally effective given correct
2 dosage and six to eight weeks of treatment. The common usage of drugs
3 clearly has cost implications and while non-drug treatment, such as weekly
4 sessions of cognitive behavioural therapy, may be suitable for a number
5 of patients, it is often in short supply and is also costly.
6 A recent overview of depression and its treatment in the UK did not
7 mention the value of exercise at all (Hale, 1997). Over the past 20 years
8 the literature in the area of physical activity/exercise and mental health
9 has been growing, but as Dishman (1995) points out the evidence has not
1011 apparently persuaded mental health agencies, such as the American
1 Psychiatric Association, to endorse the role of exercise in treating mental
2 illness such as depression. This is in contrast to coronary artery disease
3111 in which inactivity is now recognised as a primary risk factor (Pate et al.,
4 1995). Perhaps the evidence for the role of exercise in treating and
5 preventing mental illness is not convincing or perhaps the mental health
6 literature is suffering from a dualist tendency to treat the mind (mental
7 health) and body (physical health) as separate issues and therefore fail to
8 see as a priority the mental outcomes of a physical treatment such as
9 exercise (Beesley & Mutrie, 1997).
20111 Patients often report that they do not want drugs (Scott, 1996) and yet
1 drugs are the most common treatment for depression. Patient choice is,
2 therefore, another aspect of the treatment of depression that suggests it
3 is worthwhile to pursue the possibility of the use of exercise. Exercise
4 could be a reasonable option which has few negative side effects and could
5 be cost-effective in comparison to both drug and non-drug options such
6 as psychotherapy.
7 McEntee and Halgin (1996) reported that while many psychotherapists
8 believe in the therapeutic value of exercise, very few (around 10%) recom-
9 mend exercise to their clients. From their survey of 110 practising psycho-
30111 therapists they concluded that one of the major reasons for the reluctance
1 to discuss exercise was that it was perceived as inappropriate. Exercise was
2 perceived as being very directive and perhaps dealt with better by physicians
3 or physical recreation specialists: ‘Many therapists simply do not see their
4 work as pertaining to the body, and they believe that most clients come to
5 therapy to discuss psychological ailments, not physical or exercise-related
6 ones’ (McEntee & Halgin, 1996, p. 55). It would therefore seem that there
7 is much work to be done to convince those who deliver mental health
8 services to focus on the links between mind and body and to look more
9 positively on the role of exercise in mental health issues.
40111
1
Definition of clinically defined depression
2
3 One issue that has plagued our understanding of the relationship between
4 physical activity and depression is the lack of agreement amongst
45111 researchers of criteria that define depression. Many previous reviews have
48 Nanette Mutrie
1111 included cases of ‘depression’ that would not reach clinically defined
2 criteria and may be better defined as transitory negative affect. In this
3 chapter, only clinically defined depression will be included; that is, patients
4 will have sought help for their symptoms and a diagnosis made via
5 standard instruments or interviews. The most common questionnaire
6 used, especially in exercise studies, is the Beck Depression Inventory
7 (BDI) (Beck, Ward, Mendelsohn, Mock, & Erbaugh, 1961). Moderate
8 depression on the BDI is defined as a score of 16 or above. However,
9 many exercise studies have included individuals with scores lower than 16
1011 at baseline which would be considered as a transitory or normal score;
1 such studies will not be included in this review. In terms of clinical inter-
2 view, diagnosis of depression is made via criteria listed in the Diagnostic
3111 and Statistical Manual of Mental Disorders (DSM-IV) (American
4 Psychiatric Association, 1994) or the International Classification of
5 Diseases (ICD-10) (World Health Organisation, 1993). In research studies,
6 the Research Diagnostic Criteria are often used (Spitzer, Endicott, &
7 Robins, 1978).
8 Depression often occurs with other chronic diseases and mental disorders
9 and such cases may well be included but all will have met the criteria for
20111 clinical depression. In this review, and throughout the text (see Chapter 1),
1 physical activity is used as the term to describe any form of activity result-
2 ing in energy expenditure while exercise is used to describe systematic
3 programmes of activity, which are often supervised. Thus in discussing
4 epidemiological evidence physical activity will be the term which is used,
5 but when discussing treatment the term exercise is more appropriate.
6
7
Epidemiological evidence
8
9 We owe a great debt to the work of Morgan (see Morgan, 1994) who
30111 pioneered much of the initial research into the role of exercise and mental
1 health. It was perhaps his early findings that fitness levels for both male
2 (Morgan, 1968, 1969) and female (Morgan, 1970b) psychiatric patients
3 were lower than non-hospitalised controls, which led to experimental work
4 in using exercise as part of a treatment regime for such patients. Martinsen,
5 Strand, Paulson and Kaggestad (1989) replicated these findings, with
6 Norwegian psychiatric patients. Morgan (1970a) also showed that patients
7 admitted to a psychiatric hospital, but discharged in a short period of time
8 (on average 61 days), had higher levels of muscular endurance on admis-
9 sion than patients with similar initial levels of depression who remained
40111 in hospital for longer (at least one year). Such cross-sectional data raised
1 intriguing questions about whether lack of exercise can cause depression
2 or whether depression causes lack of exercise, and whether increasing
3 fitness levels could influence recovery. There were also methodological
4 questions about the relative contribution of genetics and motivation in
45111 the fitness estimates that were obtained. However, some of these early
1111 questions have now been answered. In the next section, the review of
2 epidemiological evidence suggests that depression is indeed associated
3 with low activity/fitness and that those who maintain activity are less likely
4 to develop depression.
5 The strongest epidemiological evidence comes from four prospective
6 studies that have followed cohorts over time. In all of the studies depres-
7 sion was clinically defined and in one study depression was diagnosed by
8 psychiatric interview (Weyerer, 1992). Statistical adjustments for potential
9 confounding variables, such as age and socio-economic background, were
1011 also made in each of the studies.
1 Farmer et al. (1988) reported a follow-up of 1,497 respondents to a
2 large survey with particular regard to activity and depression. This study
3111 showed that, over a period of eight years, women who had engaged in
4 ‘little or no’ activity were twice as likely to develop depression as those
5 who had engaged in ‘much’ or ‘moderate’ activity. The effects of age,
6 employment, income, education and chronic medical conditions were all
7 statistically accounted for. There was no significant association over the
8 same time period for men, but for those men who were depressed at base-
9 line, inactivity was a strong predictor of continued depression at the
20111 eight-year follow-up.
1 Camacho, Roberts, Lazarus, Kaplan and Cohen (1991) also found an
2 association between inactivity and incidence of depression in a large popu-
3 lation from Alameda County in California who provided baseline data in
4 1965 and were followed up in 1974 and 1983. Physical activity was cate-
5 gorised as low, medium or high. In the first wave of follow-up (1974) the
6 relative risk (RR) of developing depression was significantly greater for
7 both men and women who were low active in 1965 (RR 1.8 for men, 1.7
8 for women) compared to those who were high active. There is some
9 evidence for a dose-response relationship with those who were moder-
30111 ately active in 1965 showing lower risk of developing depression than
1 those who were low active (see Figure 3.1).
2 In the second follow-up in 1983, four categories of activity status were
3 created. These categories are shown in Table 3.1 and are defined as follows:
4
5 1 those who were low active in 1965 and remained low in 1974 (low/low);
6 2 those who had been low active in 1965 but had increased activity level
7 in 1974 (low/high);
8 3 those who had been high active in 1965 and decreased activity by 1974
9 (high/low);
40111 4 those who had been high active at both times points (high/high).
1
2 Those who were inactive in 1965 but had increased activity in 1974 were
3 at no greater risk of developing depression in 1983 than those who had
4 been active at both times points (the reference group for computing the
45111 odds ratio). This perhaps suggests a protective effect of physical activity.
50 Nanette Mutrie
1111
2 1.8
relative risk of developing depression
3 1.6
4
5 1.4
6 1.2
at follow-up
7 1
8
0.8
9
1011 0.6
1 0.4
2
0.2
3111
4 0
low moderate high
5
6 physical activity level at baseline
7
Figure 3.1 Relative risk of developing depression at follow-up from different levels
8
of baseline physical activity (adapted from Camacho et al., 1991).
9
20111
1 None of the odds ratios computed for risk of depression in 1983 showed
2 significant differences between the four activity categories. The largest
3 odds ratio, however, was for those who had relapsed from activity in 1965
4 to inactivity in 1974. They were 1.6 times more likely to develop depres-
5 sion in 1983 than those who had maintained activity, but it must be
6 remembered that this odds ratio did not reach significance. The authors
7 note, however, that this odds ratio was relatively unaffected by adjust-
8 ments for age, sex, physical health, socio-economic status, social support,
9 life events, anomie, smoking status, relative weight, 1965 level of depres-
30111 sion and alcohol consumption. This led them to believe it is a robust
1 finding. Given that only 137 people were in this category, it is perhaps
2 not surprising that the odds ratio did not reach significance. However, the
3 evidence from the 1974 follow-up did provide statistically significant
4 evidence that low activity preceded the reported depression.
5
6
7 Table 3.1 Changes in physical activity status and subsequent depression (from
8 Camacho et al., 1991)
9
40111 Activity status Odds ratio for developing Confidence interval
1 1965–74 depression in 1983 for odds ratio
2 1 low/low 1.22 0.62–2.38
3 2 low/high 1.11 0.52–2.21
4 3 high/low 1.61 0.80–3.22
45111 4 high/high 1.00 Reference group
1111 Paffenbarger, Lee and Leung (1994) have reported similar findings from
2 the Harvard Alumni studies that followed men for 23–27 years. In that
3 study men who engaged in three or more hours of sport activity per week
4 at baseline had a 27% reduction in the risk of developing depression at
5 follow-up compared to those who played for less than one hour per week.
6 When the authors combined the various indices of physical activity (sports
7 play, walking, stair climbing) evidence for a dose-response relationship
8 emerged; those who had expended 2,500 kcal or more per week were 28%
9 less at risk of developing clinically recognised depression than men who
1011 expended less than 1,000 kcal/week. Those who expended between 1,000
1 and 2,499 kcal/week had a 17% reduction in risk compared to the least
2 active group. This dose-response trend was significant and is illustrated in
3111 Figure 3.2 in terms of relative risk. These findings also suggest that inac-
4 tivity precedes depression.
5 All of these studies have been conducted on North American popula-
6 tions but Weyerer (1992) showed that in a community sample from Bavaria
7 (n 1,536), the physically inactive were 3.15 times more likely to have
8 depression than those who were regularly active. All were interviewed
9 by a research psychiatrist and 8.3% were identified as depressive using a
20111 clinical scale. There was some evidence for a dose-response relationship
1 since those reporting only occasional physical activity were 1.55 times
2 more likely to have depression than those who were regularly physically
3 active, although this was not statistically significant. This cross-sectional
4 data is open to the criticism that the relationship is created because the
5 depressed are inactive. The strongest counter to this argument, which we
6 have seen in the three studies already reviewed, is follow-up data which
7
8
1
9
30111 0.9
relative risk of depression
1 0.8
2 0.7
3 0.6
4 0.5
5
0.4
6
7 0.3
8 0.2
9 0.1
40111 0
1 < 1,000 1,000–2,499 > 2,500
2 physical activity index (kcals/week)
3
4 Figure 3.2 Relative risk of developing depression at follow-up from different levels
45111 of physical activity at baseline (adapted from Paffenbarger et al., 1994).
52 Nanette Mutrie
1111 shows the least active are most at risk of developing depression at a later
2 point in time. However, low physical activity was not a predictor of depres-
3 sion at a five-year follow-up to this study. The time scale of the Weyerer
4 follow-up (1992) was shorter than any of the other studies and this may
5 be the reason for this apparent difference in results.
6 All four of these studies show an association between activity and
7 depression with the least active having the greatest incidence of depression.
8 In three of the four studies follow-up data suggest that inactivity at base-
9 line is predictive of developing depression at follow-up. This suggests that
1011 inactivity precedes depression. It is important to reiterate that other
1 possible variables, such as physical health status, were accounted for since
2 people may well be inactive because they are disabled or prevented from
3111 taking part in activity because of a medical condition. However, there are
4 other reasons such as lack of social skills or socio-economic status that
5 could also predict both inactivity and depression that may not have been
6 fully accounted for. Hopefully, there will be more epidemiological data
7 of this nature that will help us form a picture of the time course of the
8 onset of depression in relation to inactivity and allow for further explo-
9 ration of variables that predict inactivity and depression. In particular,
20111 longitudinal studies are required to elucidate the possible benefits and
1 risks of involvement in PA for youth or adult psychological functioning.
2 Steptoe and Butler (1996) suggested that this could be done with British
3 data from a cohort study initiated in 1970 which they have already used
4 to show a positive association between sport participation and emotional
5 well-being for the cohort during adolescence. Thus the epidemiological
6 data are strongly suggestive of a protective effect from activity but yet
7 more data are required. A final point to note is that there is no evidence
8 to suggest that increasing physical activity or exercise increases the risk
9 of depression.
30111
1
Meta-analyses
2
3 Two meta-analytic reviews of exercise as a treatment for depression have
4 provided further substantial evidence for positive effects (McDonald &
5 Hodgdon, 1991; North, McCullagh, & Tran, 1990). Both report effect sizes
6 of around one half of a standard deviation of change in depression scores
7 which suggests that exercise does have an anti-depressant effect. Calfas
8 and Taylor (1994) report a small meta-analysis of five randomised control
9 trials (RCT) on healthy and psychologically ‘at risk’ adolescents. They
40111 reported an effect size of 0.38 for exercise on depression, although the
1 small number of studies involved means that this must be a cautious
2 conclusion.
3 The outcomes of meta-analytic reviews are subject to the quality of the
4 input. Dishman (1995) suggests that averaging results from studies with
45111 different designs and methods of measurement is not helpful and concludes
1111 that there are too few studies with similar features to warrant confidence
2 in the results of meta-analysis in this area. In addition, the issue of whether
3 depression has been clinically defined is particularly important for this
4 chapter. In fact very few studies included in the meta-analyses by North
5 et al. (1990) or McDonald and Hodgdon (1991) had clinically diagnosed
6 individuals.
7 Craft and Landers (1998) have addressed this issue and conducted a
8 meta-analysis confined to those with clinically defined depression. This
9 meta-analysis included 30 studies, many of which were unpublished disser-
1011 tations. The average effect size (ES) was 0.72. Further analysis of the
1 moderating variables showed that the effect sizes for mode of exercise
2 (aerobic versus non-aerobic) did not differ, and there was no difference
3111 between exercise treatment and psychotherapeutic or behavioural inter-
4 ventions. However, there was a greater ES for those initially classified as
5 moderate to severe in depression compared to those classified as mild to
6 moderate. The results of Craft and Landers’ meta-analysis are therefore
7 very encouraging. However, even with a well conducted meta-analysis
8 such as this one, some of Dishman’s (1995) criticisms remain since in some
9 of the comparisons (e.g. with different entry level of depression) there
20111 are very few studies. It seems best, therefore, to also look at individual
1 studies in detail rather than relying solely on meta-analytic conclusions.
2
3
Key studies
4
5 Most narrative reviews in this area (e.g. Biddle & Mutrie, 1991; Byrne &
6 Byrne, 1993; Gleser & Mendelberg, 1990; Martinsen, 1989; Morgan, 1994)
7 make cautious positive conclusions but note the methodological limitations
8 of many studies and this criticism has been echoed many times (Dishman,
9 1995). However, with the exception of Martinsen (1989, 1993, 1994),
30111 reviews have included non-clinically defined depression and Morgan (1994)
1 noted that one of the most reliable findings in this area is that exercise
2 will not decrease depression in those who are not depressed in the first
3 place. It would seem appropriate, therefore, to examine all studies in which
4 exercise has been used to treat clinically defined depression and to limit
5 the discussion to studies which have the best design features.
6 Literature was searched using BIDS, accessing Social Science Citation
7 Index and Embase which searches medical literature, PsychLit, Firstsearch
8 and Sport Discus. All studies from 1970 onwards which could be located
9 incorporating random assignment of subjects to groups and including a
40111 clinically defined measure of depression were reviewed. This process
1 excludes some well-designed studies, such as McCann and Holmes (1984),
2 because depression levels were below 16 on the BDI. The key studies are
3 summarised in Table 3.2.
4
45111
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Table 3.2 Randomised controlled studies of exercise treatment for clinically defined depression
Authors/location Participants Design Treatment groups Measures Results (statistically

significant at 0.05)
Greist et al. n = 28 (15 women), 10 weeks of 1. 10 sessions of time-limited SCL The running treatment was as
(1979) USA RDC criteria for treatment, 1 and psychotherapy effective as the two
depression 3 month follow-up 2. time-unlimited psychotherapy; psychotherapy treatments
3. running with a leader
3 30–45 mins/week
Klein et al. n = 74 (53 women), 12 weeks of 1. running with a leader, SCL and The running treatment was
(1985) USA mean age 30 years, treatment and 2 45 mins/week psychiatric as effective as the other two
recruited via media, 1, 3 & 9 month 2. group meditation, interview treatments
RDC criteria for follow-up 2 hours/week
depression 3. group therapy, 2 hours/week
Martinsen n = 43, mean age 9 weeks of 1. exercise group, aerobic BDI; predicted The exercise group decreased
·
et al. (1985) 40 years, treatment training,
· 50–70% max. max. VO2 depression scores and
Norway hospitalised VO2, 1 hour, 3/week increased fitness more than
depressives, 2. control group, occupational the control group
clinical assessment therapy, 1 hour, 3/week
by DSM-III
Doyne et al. n = 40 (all women) 8 weeks of 1. aerobic group (running); BDI; HRSD; Both exercise conditions
(1987) USA recruited through treatment; 1, 7, 4/week cardiovascular reduced depression more than
mass media; mean 12 month follow-up 2. non-aerobic group (weight- fitness (METS) waiting list control. Levels of
age 29 years; lifting); 4/week from sub-maximal depression remained lower
clinical assessment 3. waiting list control group test than baseline to 1 year
by RDC follow-up
Fremont & n = 49, recruitment 10 weeks of 1. cognitive therapy 1 hour/week BDI All three groups improved.
Craighead via advertisement treatment and 2. running with a leader, Improvements maintained at
(1987) USA BDI scores of 16 2 month follow-up 3 20 mins/week 2 month follow-up
and above 3. both cognitive therapy and
running.
Mutrie (1988) n = 24 (20 women), 8 weeks of 1. aerobic exercise conducted at BDI, POMS, After 4 weeks only the
UK mean age 42 years, treatment; home (walk/jog) 3 20–30 standard step-test aerobic group made
clinical assessment assessment at mins/week for aerobic fitness, significant reductions on BDI.
by GP diagnoses 4 weeks, 8 weeks 2. non-aerobic strengthening and standard sit-up After 8 weeks all groups
and BDI scores of and 20 week stretching exercise conducted test for strength decreased BDI scores and
16 and above follow-up at home, 3 20–30 mins/week these scores were maintained
3. no treatment for 4 weeks, then at 20 weeks with no group
combination of aerobic and differences. There were no
non-aerobic exercise, 3 20 group differences in fitness
mins/week test results with no change
noted at 4 weeks and all
groups improving by 8 weeks
Martinsen, n = 99 (63 women), 8 weeks of 1. aerobic training, 3 1 Montgomery- Both groups decreased
Strand et al. mean age 41 years, treatment hour/week Asberg rating scale, depression scores. Only the
(1989) Norway hospitalised 2. strength & flexibility training, BDI,
· predicted aerobic· group made gains on
depressives, RDC 3 1 hour/week VO2 max. max. VO2
classification
Veale et al. (1992) n = 83 (53 women) 12 weeks of 1. standard treatment CIS, BDI, Exercise group reduced
·
Trial 1 UK mean age 36 years, treatment 2. aerobic exercise (3/week predicted VO2 depressive symptoms (CIS)
clinical assessment running) adjunctive to max. and trait anxiety more than
by CIS standard treatment standard group despite
incomplete adherence by
some S’s
Veale et al. (1992) n = 41; clinical 12 weeks of each group received standard CIS, BDI, Both exercise groups showed
·
Trial 2 UK assessment by CIS treatment treatment and either: predicted VO2 similar changes to that seen in
1. aerobic exercise (3/week) or max. study 1 above. No differences
2. non-aerobic exercise between groups on any
(stretching, yoga) (3/week) measures
Bosscher (1993) n = 24 (12 women), 8 weeks of 1. standard movement therapy SDS Only the running group
Netherlands mean age 34 years, treatment of mixed games and exercises, showed significant decreases
hospitalised 50 mins 3 week in depression although scores
depressives, RDC 2. running 45 mins 3 week still above entry level criteria.
No fitness measures taken
Notes:
RDC (Research Diagnostic Criteria: Spitzer, Endicott & Robins, 1978) DSM-III (Diagnostic and Statistical manual of Mental Disorders: American Psychiatric
Association, 1980) SCL (Symptom Checklist: Derogatis, Lipman & Covi, 1973) BDI (Beck Depression Inventory: Beck, Ward, Mendelsohn, Mock & Erbaugh,
1961) HRSD (Hamilton Rating Scale: Hamilton, 1960) POMS (Profile of Mood States: McNair, Lorr & Droppleman, 1971) CIS (Clinical Interview Schedule,
Goldberg, Cooper, Eastwood, Kedward, Shepherd, 1970) SDS (Zung depression scale, Zung, Richards & Short, 1965).
56 Nanette Mutrie
1111 Conclusions from key studies
2
The first and obvious conclusion from Table 3.2 is that more studies are
3
required. Only three could be found which had been conducted in the
4
1990s. It is also concluded that both internal and external validity are
5
high, given that only those studies with good design features were included,
6
and given that they have been conducted in North America and Europe,
7
all with similar results. Table 3.2 also shows that exercise programmes
8
(both aerobic and non-aerobic) can reduce clinically defined depression
9
and that the reduction of depression is of the same order as that found
1011
for a variety of standard psychotherapeutic treatments. Furthermore, these
1
anti-depressant effects are feasible in a short time frame (4–8 weeks) and
2
persist from 2 months to 1 year. These findings seem to be very similar
3111
to those from meta-analyses and thus add confidence to the meta-analytic
4
conclusion that exercise can have a substantial anti-depressant effect.
5
What we do not know is the comparative effects of exercise treatment
6
with drug treatment. This seems surprising given that drugs are the most
7
common treatment for depression in the UK, and it is also surprising that
8
so few studies have been conducted in the UK. The studies in Table 3.2
9
have mean ages of participants between 29 and 42 years suggesting that
20111
exercise effects for depression levels for youth or older adults have not been
1
studied. Only one study on the use of exercise as an adjunctive treatment
2
in clinically diagnosed mental illness in children was found (Brown, Welsh,
3
Labbe, Vitulli, & Kulkarni, 1992), but it was excluded from the table of key
4
studies since it did not reach the design criteria. Similarly, a well-designed
5
study (McNeil, LeBlanc, & Joyner, 1991) of exercise and social contact with
6
older adults was excluded since the adults were entered into that study with
7
scores below 16 on the BDI. In addition, we do not know much about
8
adherence levels to exercise especially in the follow-up phases. Some
9
studies do report this but in most cases the details are missing.
30111
It is clearly difficult to conduct studies with good design features in this
1
area. Even by selecting those studies with random assignment to treat-
2
ment conditions, methodological difficulties remain which limit the
3
strength of the conclusions. These difficulties include:
4
5
• achieving a big enough sample to ensure statistical power (could the
6
findings of ‘no difference’ between some conditions be a Type 2 statis-
7
tical error?);
8
• equalising time in contact with professionals in the different treatment
9
conditions;
40111
• avoiding resentful demoralisation in a no-treatment group or a group
1
given the ‘routine’ as opposed to the ‘new’ treatment;
2
• controlling for the effects of the positive characteristics of an exercise
3
leader;
4
• conducting long term follow-up; and
45111
1111 • finding adequate measures of the variables of interest including fitness
2 changes.
3
4 The next stage for research must attempt to overcome these methodo-
5 logical difficulties.
6
7
8
How good is the evidence for an anti-depressant effect
9
from exercise?
1011 It is still difficult to conclude that there is a causal link between exercise
1 and reduction in depression because there are many peripheral issues (such
2 as the effect of an exercise leader, or a class effect) associated with most
3111 of the successful programmes. There are also relatively few experimental
4 studies. However, given the more recent addition of epidemiological data
5 to the discussion it may be appropriate to use Hill’s (1965) classic criteria
6 for deciding whether there is an association or a causal link between
7 observed illness (in this case depression) and some environmental condi-
8 tion (in this case exercise status). Hill (1965) suggested eight criteria which
9 can be used to help scientists and practitioners decide if a causal interpre-
20111 tation of evidence can be made. I will use these eight criteria to look at the
1 evidence for an anti-depressant effect of exercise.
2
3
4 Strength of association
5 The first of Hill’s criteria is strength of the association. Meta-analytic
6 studies show an effect size of 0.53 – 0.72 for exercise on depression.
7 Epidemiological studies suggest a relative risk of around 1.7 for the inac-
8 tive reporting depression at a later date. This evidence is not quite as
9 strong as that for exercise and coronary heart disease, where range of
30111 relative risk of between 1.5 to 2.5 for the inactive have been reported
1 (Pate et al., 1995). Nevertheless, the exercise effect for depression does
2 show strength of association.
3
4
5 Consistency
6 The second of Hill’s criteria is consistency where the question is whether
7 or not the association between exercise and depression has been shown
8 in different places, with different people, at different times and in different
9 circumstances. If we look at Table 3.2 we can see that experimental
40111 evidence has been found in the US, UK, and other parts of Europe. The
1 same is true for the epidemiological evidence. Men and women have been
2
included, the data spans three decades of work, and the circumstances
3
include community, hospitals and primary care settings. It does indeed
4
seem that the findings are consistent.
45111
58 Nanette Mutrie
1111 Specificity
2
Specificity, the third of Hill’s criteria, refers to whether or not other asso-
3
ciations exist between the conditions and the disease. Hill argues that
4
specificity, which is limiting the conditions to the disease, for example
5
smoking and lung cancer, strengthens the argument for causation. In exer-
6
cise studies specificity does not exist. Depression is not the only condition
7
that has been linked to inactivity. Indeed, others have shown a link between
8
exercise and all-cause mortality (Blair et al., 1989), and depression has
9
been shown to have more than one contributing factor (Kaplan, Roberts,
1011
Camacho, & Coyne, 1987). However, while Hill argued that if specificity
1
exists it strengthens the argument for causation, he also argued that if
2
specificity is not present other criteria might supply additional evidence
3111
and it is not a fatal flaw in the case.
4
5
6 Temporal sequence
7
The fourth of Hill’s criteria is temporal sequence. In order to conclude
8
that there is a causal link between inactivity and depression we must be
9
able to judge whether or not inactivity precedes the onset of depression.
20111
Early cross-sectional studies could not answer this because it was equally
1
likely that depression preceded inactivity. However, at least three prospec-
2
tive population studies have shown that the inactive are more likely to
3
develop depression. Thus there is some evidence for the temporal sequence
4
which strengthens the case for causation.
5
6
7 Dose-response
8
Hill’s fifth criterion is evidence for a dose-response curve or biological
9
gradient. Two of the prospective epidemiological studies showed a dose-
30111
response gradient with the least active at baseline being most at risk of
1
developing depression at follow-up, while the most active had the lowest
2
risk. In terms of experimental studies there is insufficient evidence at
3
present to suggest that different doses of exercise produce different psycho-
4
logical outcomes. Although both aerobic and non-aerobic exercise have
5
produced an anti-depressant effect, almost all the aerobic exercise has
6
been based on moderate intensity (60–75%) levels with a typical 3 times
7
per week, 20–60 minute prescription. However, it has also been noted that
8
negative effects in terms of mood occur in athletes who far exceed the
9
typical prescription (Morgan, 1994). Thus the evidence for a dose-response
40111
curve is modest. There is obviously a need for additional data to complete
1
this imperfect picture.
2
3
4
45111
1111 Plausibility
2
The sixth criterion is biological plausibility. Here we are looking for the
3
explanation of the observed association. There is considerable agreement
4
that the underlying mechanisms that relate to the positive effects from
5
exercise on mental illness are not yet known (Biddle & Mutrie, 1991;
6
Morgan & Goldston, 1987; Plante, 1993). Several possible mechanisms,
7
including biochemical changes such as increased levels of endorphins, and
8
psychological changes such as an increased sense of mastery, have been
9
proposed (La Forge, 1995; Petruzzello, Landers, Hatfield, Kubitz, &
1011
Salazar, 1991). The studies showing an anti-depressant effect for non-
1
aerobic exercise suggest that an improvement in aerobic fitness is not a
2
key issue. However, objective measures of all possible fitness parameters
3111
(aerobic, strength, flexibility and body composition) should be included
4
in studies to provide evidence that the exercise programme has had the
5
desired fitness effect and to shed light on potential mechanisms.
6
The fact that we do not know which mechanism operates should not
7
prevent us saying that they remain ‘plausible’. Dishman (1995), in his
8
excellent review of this topic, concludes that our lack of knowledge about
9
the biological plausibility of the association between exercise and mental
20111
health is a major shortcoming in the literature. It is likely that this short-
1
coming contributes to the lack of acceptance of the role of exercise by
2
psychiatrists (Hale, 1997). Hill (1965, p. 298) reminds us that we should
3
not demand too much of this criteria because ‘What is biologically plau-
4
sible depends upon the biological knowledge of the day’. Determining the
5
mechanisms for the psychological effects of exercise in general, and for
6
depression in particular, is perhaps the greatest challenge to exercise scien-
7
tists trying to illuminate the relationship between exercise and mental
8
health. It would appear that much of the knowledge has to be developed
9
using animal models until such times when we have technology to study
30111
brain function in humans during exercise. Brain imaging is one possible
1
technology that may advance our understanding of the mechanisms. It is
2
clear that the answer to this complex question will not be found in exercise
3
laboratories alone. We must collaborate with colleagues in neuroscience
4
and psychological medicine to expand our knowledge. La Forge (1995,
5
p. 28) provided this sensible guide to future practice:
6
7
The mechanism is likely an extraordinary synergy of biological trans-
8
actions, including genetic, environmental, and acute and adaptive
9
neurobiological processes. Inevitably, the final answers will emerge
40111
from a similar synergy of researchers and theoreticians from exercise
1
science, cognitive science and neurobiology.
2
3
4
45111
60 Nanette Mutrie
1111 Coherence
2
The possible mechanisms should not conflict with what is understood to
3
be the natural history and biology of mental illness. This is Hill’s seventh
4
criterion of coherence. While, as with many other aspects of these criteria,
5
the evidence is far from complete, one example might show coherence.
6
More women than men report depression and women report less activity
7
than men. Development of animal models to study inactivity and depres-
8
9 sion and the use of exercise to combat depression will provide further
1011 evidence for coherence.
1
2 Experimental evidence
3111
4 Perhaps the best evidence comes under Hill’s eighth criterion of experi-
5 mental evidence already discussed in the conclusions from the key studies
6 in Table 3.2. The experimental evidence supports a causal link with exer-
7 cise programmes and depression reduction.
8 In reviewing the evidence in terms of these criteria it can be seen that
9 the only criterion which the link between inactivity and depression does
20111 not fulfil is specificity. Other criteria, such as temporal sequence and dose-
1 response have only modest support, but it does seem reasonable to
2 conclude that there is supportive evidence for a causal link between inac-
3 tivity and depression although much work remains to be done. There are
4 those who might say that the evidence is still insufficient and therefore
5 we should not recommend the use of exercise in the treatment of depres-
6 sion or consider inactivity to be a factor in the onset of depression.
7 However, as Hill (1965, p. 12) reminded us,
8
9 All scientific work is incomplete – whether it be observational or
30111 experimental. All scientific work is liable to be upset or modified
1 by advancing knowledge. That does not confer upon us a freedom to
2 ignore the knowledge we already have, or postpone the action that
3 it appears to demand at a given time.
4
5 The potential benefit of advocating the use of exercise as part of a treat-
6 ment package for depression far outweighs the potential risk that no effect
7 will occur. There are very few possible negative side effects (e.g. injury, exer-
8 cise dependence) and there have been no negative outcomes reported in the
9 literature. In addition, there are potential physical health benefits such as an
40111 increase in fitness, weight reduction, and decreased coronary artery disease
1 risks. Therefore, physical activity/exercise should be advocated as part of
2 the treatment for clinically defined depression. The evidence presented
3 here adds further strength to the arguments used by health promoters for
4 the need to prevent diseases (including mental illness) by increasing the
45111 percentage of the population engaged in regular physical activity.
1111 Future practice and research
2
Having considered the evidence, this section will consider guidelines and
3
future directions for those promoting and researching physical activity
4
and depression.
5
6
7 Guidelines for practice
8
• Physical activity/exercise should be advocated as part of the treatment
9
of clinically defined depression.
1011
• Health promotion campaigns aimed at increasing the level of physical
1
activity in the population should include the prevention of depression
2
as part of the rationale.
3111
• Exercise leaders, general practitioners and other para-medical staff
4
working with depressed patients need in-service training on how
5
exercise may have an anti-depressant effect.
6
• Pre-service training for doctors, psychiatrists and clinical psychologists
7
is required on the topic of the anti-depressant effects of exercise.
8
9
20111 Guidelines for research
1
• The exploration of physical activity and depression levels in large
2
population sets should be encouraged.
3
• Randomised controlled trials of exercise therapy versus standard drug
4
therapy and psychotherapy are required which attempt to overcome
5
the methodological difficulties of earlier studies.
6
• Qualitative studies of how different patients and medical staff perceive
7
the role of exercise in the treatment of depression should be encour-
8
aged.
9
• Practical and experimental application of exercise is required for both
30111
young and old people suffering from depression.
1
• Exercise scientists must collaborate with medical and neurobiological
2
scientists to explore the mechanisms of the anti-depressant effects of
3
exercise.
4
5
6
7 What we know
8 • Epidemiological evidence has demonstrated that physical activity
9 is associated with a decreased risk of developing clinically
40111 defined depression.
1 • Evidence from experimental studies shows that both aerobic
2 exercise and resistance training exercise may be used to treat
3 moderate and more severe depression, usually as an adjunct to
4 standard treatment.
45111
62 Nanette Mutrie
1111
2 • The anti-depressant effect of exercise is of the same magnitude
3 as that found from psychotherapeutic techniques. However, the
4 range of psychotherapies used in these studies does not perhaps
5 mirror currently available ‘best practice’ such as cognitive behav-
6 ioural therapy.
7 • There is support for a causal link between exercise and
8 decreased depression.
9 • No negative effects of exercise have been noted in depressed
1011 populations.
1
2
3111
4
6 • Is there an anti-depressant effect from exercise for younger and
7 older adults?
8 • Are the psychological effects of physical activity the same for dif-
9 ferent modes of activity (e.g. aerobic, strength-based, flexibility-
20111 based)?
1 • Do different intensities and durations of physical activity make
2 a difference and do fitness levels modulate that effect?
3 • What is the time course of the effects?
4 • How do the potential mechanisms of the effects interact?
5 • How do effects of exercise compare to those of drug treatments
6 and what adjunctive value does exercise have along with drug
7 treatment?
8 • If drugs are also administered is the interaction of drug and exer-
9 cise safe?
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
4
45111
1111
2 4 Emotion, mood and physical
3
4 activity
5
6 Stuart J.H. Biddle
7
8
9
1011
1
2 This chapter summarises the evidence on exercise and physical activity
3111 (PA) and emotional feelings and mood. Appropriate evidence-based
4 conclusions are drawn and implications for guiding future research
5 and practice in health settings provided. Because the areas of mood and
6 emotion are not easy to delimit, studies investigating clinical depression,
7 self-esteem, and cognitive functioning have not been addressed as these
8 are reviewed elsewhere in this volume. Tension-related mood states will
9 be considered, but not the wider literature on anxiety and stress as this
20111 also is covered elsewhere. Only studies investigating exercise and ‘lifestyle’
1 physical activity, rather than competitive sport, are considered.
2 Evidence has been drawn from papers located through electronic
3 searches using Sport Discus, PsychLit, and Medline, as well as searching
4 extensive personal files and references cited in other reviews. Preference
5 was given to papers published since 1987, and particularly meta-analyses,
6 epidemiological surveys, and controlled trials.
7 Given the potentially disparate nature of the topic being addressed, the
8 review is organised around the themes of emotion (affect) and mood but
9 within the overall context of health-related quality of life (HRQL).
30111 Research consensus statements are provided in conclusion.
1
2
Health-related quality of life
3
4 Rejeski, Brawley and Shumaker (1996) suggest that it is typical for HRQL
5 to be defined in terms of participants’ perceptions of function. They outline
6 six types of HRQL measures:
7
8 • global indices of HRQL. These might include general life satisfaction,
9 or self-esteem.
40111 • physical function. Perceptions of function; physical self-perceptions;
1 health-related perceptions.
2 • physical symptoms. Fatigue; energy; sleep.
3 • emotional function. Depression, anxiety, mood, affect (emotion).
4 • social function. Social dependency; family/work roles.
45111 • cognitive function. Memory; attention; problem-solving.
64 Stuart Biddle
1111 Rejeski et al. (1996) state that the National Institutes of Health in the
2 US now mandate researchers to include measures of HRQL in most
3 clinical trials. HRQL outcomes are justified in physical activity research
4 for several reasons. First, health-related perceptions of patients are impor-
5 tant in their own right. Second, they may act as a motivator of future
6 participation in physical activity. Third, correlations between HRQL and
7 performance measures of dysfunction, such as perceived ability to climb
8 stairs, are usually stronger than between HRQL and changes in physical
9 fitness, such as aerobic capacity. This suggests that performance measures
1011 are much more easily detected and are more salient to patients than
1 objective fitness measures.
2 HRQL measures, however, are usually viewed simply in terms of
3111 physical function. This is a narrow view. This chapter, therefore, will
4 review studies that fall primarily into the emotional function category
5 of HRQL. Chapters elsewhere in this volume cover other aspects of
6 HRQL.
7
8
Measurement of HRQL
9
20111 There are many HRQL instruments and these include affective measures,
1 although agreement over measurement is difficult to come by (see Bennett
2 & Murphy, 1997). Key measures are summarised in Table 4.1 and include
3 the SF-36, The Nottingham Health Profile and the EuroQol. The SF-36
4 is a 36-item questionnaire designed to assess 8 health dimensions covering
5 functional status, well-being, and overall evaluation of health (Dixon,
6 Heaton, Long, & Warburton, 1994). Dixon et al. (1994) conclude that the
7 SF-36 is not designed for specific patient groups, and that it is not directly
8 based on lay views; also there is little evidence that it detects change.
9 Indeed, health changes detected through interview by Hill, Harries and
30111 Popay (1996) went undetected by the SF-36 (see also Jenkinson, Layte,
1 Coulter & Wright, 1996). Overall, therefore, an over-reliance on the use
2 of the SF-36 in HRQL studies is not recommended. Where possible, if
3 mood, affect and HRQL are to be assessed in physical activity interven-
4 tions, more specific measures should be sought.
5 The Nottingham Health Profile (NHP) (Hunt, McEwan & McKenna,
6 1986) has been used since the 1970s in intervention and outcome trials
7 and, like the EuroQol (Buxton, O’Hanlon & Rushby, 1990, 1992), assesses
8 six dimensions of HRQL. (See Bowling, 1995, for a review of quality of
9 life measures.)
40111
1
HRQL and physical activity
2
3 Rejeski et al. (1996) provide a comprehensive review of HRQL and phys-
4 ical activity. The present chapter, therefore, will focus on the emotional
45111 dimension of HRQL, and specifically mood and affect. Chapters elsewhere
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Table 4.1 Common measures of health-related quality of life
Label Reference(s) Measures Comments
SF-36 Dixon et al. (1994) • general health Mental health measures include
• physical functioning 5 items assessing ‘anxiety’,
• mental health depressed mood, happiness. Social
• role limitations (physical and functioning subscale assesses
emotional) whether physical health or emotional
• bodily pain problems have interfered with social
• energy/tiredness activities
• social functioning
Nottingham Health Hunt et al. (1986) • energy

Profile • pain
• physical mobility
• emotional reactions
• social isolation
• sleep
EuroQol Buxton et al. (1990, 1992) • mobility

• self-care
• ‘usual activities’
• pain/discomfort
• anxiety/depression
• general health change
66 Stuart Biddle
1111 in this volume deal with other aspects of HRQL, such as clinical depres-
2 sion (Mutrie), anxiety (Taylor), self-esteem (Fox), and cognitive function-
3 ing (Boutcher). Rejeski et al. (1996) offer the following conclusions from
4 their review of physical activity and HRQL studies:
5
6 • HRQL test batteries should include general and condition- or popu-
7 lation-specific measures.
8 • The degree of change observed in HRQL through PA will depend
9 on baseline levels.
1011 • The degree of impact of PA on HRQL will depend on both the physio-
1 logical stimulus as well as social and behavioural characteristics of the
2 treatment or intervention.
3111 • People vary in the extent to which they value certain health-related
4 outcomes from physical activity, hence this will affect HRQL percep-
5 tions of those in intervention studies.
6
7
Emotion and mood
8
9 The mood states and emotions associated with physical activity have a
20111 potentially important role in health promotion. If we believe that phys-
1 ical activity is a positive health behaviour to be encouraged and promoted,
2 how people feel during and after activity may be critical in determining
3 whether they maintain their involvement. Hence, emotion and mood may
4 have motivational properties for an important health-related behaviour.
5 In addition, positive mood and affect are important health outcomes
6 in their own right. For example, Morgan, in the preface to his book on
7 physical activity and mental health stated that ‘it is our belief that preven-
8 tion – not treatment – offers the best solution to the pandemic mental
9 health problems that characterize modern society. This book attempts to
30111 demonstrate the extent to which physical activity, a nonpharmacological
1 strategy, can be effective in this regard’ (1997, p. xv).
2 Mood is the global set of affective states we experience on a day-to-
3 day basis. Although mood can be conceptualised in terms of distinct mood
4 states, such as vigour and depression, it differs from emotion, which is
5 normally defined in terms of specific feeling states generated in reaction
6 to certain events or appraisals. However, the distinction between mood
7 and emotion in PA research studies is often not clear.
8
9
Measuring emotion and mood
40111
1
Emotion
2
3 Emotion has sometimes been distinguished from ‘affect’. Lazarus (1991,
4 p. 6) suggests that emotion is more generic than affect and defines the
45111 latter as the subjective experience of emotion, and emotion itself as
Emotion, mood and physical activity 67
1111 ‘complex, patterned . . . reactions to how we think we are doing in our
2 lifelong efforts to survive and flourish and to achieve what we wish for
3 ourselves’. However, the two terms are sometimes not distinguished from
4 each other, and this is often the case in physical activity research. In the
5 present paper the two terms will be used synonymously.
6 In psychology, there is a debate concerning the nature of emotion. Some
7 prefer to define emotion in terms of discrete emotional reactions, such as
8 pleasure, fear, happiness, and excitement (Clore, Ortony, & Foss, 1987;
9 Lazarus, 1991; Weiner, 1995). Others suggest that emotions are best
1011 defined in terms of their common properties, or dimensions, such as posi-
1 tive and negative affect (Watson & Tellegen, 1985).
2 Lazarus (1991) argues that the distinct qualities of emotional reactions
3111 are lost, or blurred, when reduced to a few dimensions. According to his
4 ‘cognitive-motivational-relational’ theory of emotion, he argues that each
5 emotion is unique because it is created by a different appraisal of the per-
6 ceived significance of an event. However, it is also logical to see emotions
7 clustered according to common categories. Watson and colleagues
8 (Watson, Clark, & Carey, 1988; Watson, Clark, & Tellegen, 1988; Watson
9 & Tellegen, 1985) have shown that two major factors emerge from an
20111 analysis of emotions – positive affect and negative affect. The former
1 refers to feelings such as alertness and activeness, whereas negative affect
2 refers to unpleasant affective states such as anger and fear.
3 Russell and colleagues have also advocated a dimensional approach to
4 the study of emotions (see Russell, 1980). In their ‘circumplex’ model,
5 they suggest that emotion can best be defined in terms of the two dimen-
6 sions of valence (i.e. pleasant–unpleasant) and arousal (i.e. high–low). This
7 gives rise to emotions being classified along these two dimensions, such
8 as tense (high arousal/low pleasure), excited (high arousal/high pleasure),
9 relaxed (low arousal/high pleasure), and depressed (low arousal/low plea-
30111 sure) (see Warr, 1990).
1
2
Mood
3
4 Measures of mood have typically involved the Profile of Mood States
5 (POMS) (McNair, Lorr, & Droppleman, 1971), although McDonald
6 and Hodgdon (1991) also located exercise studies using the Multiple
7 Affect Adjective Check List (MAACL) (Zuckerman & Lubin, 1965). The
8 MAACL, however, assesses only anxiety, depression and hostility and is
9 prone to social desirability distortion. Similarly, the POMS is comprised
40111 of five negative mood scales and only one positive scale (vigour). Studies
1 using the MAACL, therefore, cannot be considered adequate for the study
2 of psychological well-being (PWB) or mood, and those using the POMS
3 are limited due to the single positive mood subscale. The POMS can also
4 be varied according to the instructions, such as participants describing how
45111 they feel/have felt ‘right now’ or ‘over the past few weeks’.
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Table 4.2 Summary of mood and affect measures commonly used in physical activity research
Instrument Reference Measures Comments
POMS (Profile of McNair et al. (1971) 65-item scale assessing: • only one positive subscale
Mood States) • tension • used extensively in PA research
• depression • short and bipolar forms available
68 Stuart Biddle
• anger • time instructions can be varied

• vigour • can be a state or trait scale
• fatigue • general scale not specific to
• confusion physical activity
PANAS (Positive Watson, Clark, & Two 10-item affect scales assessing: • good psychometric properties
and Negative Tellegen (1988) • positive affect: e.g. excited, • assesses only two general dimensions
Affect Schedule) enthusiastic, inspired • time instructions can be varied
• negative affect: e.g. • can be a state or trait scale
distressed, hostile, irritable • general scale not specific to physical
activity
BFS Abele & Brehm (1993) 40-item scale devised in German to • extensive German research
(Befindlichkeitsskalen) assess two-dimensional model of mood: supporting validity of scale
activation (high/low) and evaluation in sport and exercise settings
(positive/negative). 8 subscales: • state scale
• activation (high/positive)
• elation (high/positive)
• calmness (low/positive)
• contemplativeness (low/positive)
• excitation (high/negative)
• anger (high/negative)
• fatigue (low/negative)
• depression (low/negative)
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
MAACL (Multiple Zuckerman & Lubin • scale comprises 132 adjectives • time instructions can be varied
Affect Adjective (1965) • assesses anxiety, depression • can be a state or trait scale
Check List) and hostility • general scale not specific to
physical activity
• some doubts expressed about
psychometric properties (see
McDonald & Hodgdon (1991)
FS (Feeling Scale) Hardy & Rejeski (1989) • single-item scale assessing • developed for exercise research
hedonic tone (pleasure/displeasure) • state scale
• 11-point scale ranging from –5 to +5
EFI (Exercise-Induced Gauvin & Rejeski (1993) 12-item adjective scale assessing • developed for exercise research
Feeling Inventory) four dimensions: • sound psychometric properties
• positive engagement • state scale
• tranquillity
• revitalisation
• physical exhaustion
SEES (Subjective McAuley & Courneya 12-item adjective scale assessing • developed for exercise research
Exercise Experiences (1994) three dimensions: • sound psychometric properties
Scale) • positive well-being • state scale
• psychological distress
• fatigue
70 Stuart Biddle
1111 Abele and Brehm (1993) report a number of studies they have conducted
2 in Germany using the ‘Befindlichkeitsskalen’ (BFS). This scale places mood
3 states along the continua of high/low activation and positive/negative
4 mood. This is similar to Russell’s (1980) circumplex model. For example,
5 one can contrast high activation moods that differ in their positive or
6 negative evaluation. A high activation/negative mood might be ‘anger’
7 whereas a high activation/positive mood state might be ‘elation’.
8
9
Measures of mood and affect in physical activity research
1011
1 The majority of studies investigating exercise, PA and PWB have assessed
2 mood and affect using scales such as the POMS, MAACL and the PANAS
3111 (see Table 4.2). The POMS, in particular, has been criticised mainly on
4 the basis that it is restricted to just one positive factor (vigour). Steptoe
5 (1992, pp. 208–9), for example, argues that ‘measures like the POMS . . .
6 fail to capture the positive feelings of well-being that are more than the
7 mere absence of anxiety, depression or irritation’. The PANAS, while
8 being used increasingly in exercise research, is ‘restricted to the assess-
9 ment of global affect’ (Gauvin & Rejeski, 1993, p. 404) and fails to
20111 differentiate more specific types of emotion.
1 Steptoe (1992) reports on his work on mood and exercise by suggesting
2 that ‘context-specific’ measures might be particularly useful. For example,
3 his research group developed a list of 36 items thought to be associated
4 with exercise and mood. Factor analysis revealed the three factors of
5 ‘coping assets’, ‘coping deficits’, and ‘physical well-being’. Such a scale
6 (shown in Table 4.3) allows for a more balanced view of PWB in compar-
7 ison to the POMS. However, Steptoe’s scale has not been used by other
8 researchers.
9 Gauvin and Rejeski (1993) developed the Exercise-induced Feel-
30111 ing Inventory (EFI) in an effort to capture four distinct feeling states
1 in exercise: revitalisation, tranquillity, positive engagement, and physical
2 exhaustion. Psychometric support has been reported for adults (Gauvin
3 & Rejeski, 1993) and children (Vlachopoulos, Biddle, & Fox, 1996).
4 In a similar vein, McAuley and Courneya (1994) developed the Subjective
5 Exercise Experiences Scale (SEES) comprising three factors of positive
6 well-being, psychological distress, and fatigue and this, too, now has support
7 with children (Markland, Emberton, & Tallon, 1997). Both the SEES and
8 EFI are scales can that easily be used in field assessments of exercise affect.
9 A summary of key measures of mood and affect in physical activity research
40111 is shown in Table 4.2 (see also Gauvin & Spence, 1998).
1
2
Emotion, mood and physical activity
3
4 There are a very large number of studies investigating the relationship
45111 between physical activity and affective states. As such, I will draw my
1111 Table 4.3 Factors of psychological well-being for physical activity identified by
2 Moses et al. (1989)
3
Factor 1 Factor 2 Factor 3
4 Coping assets Coping deficits Physical well-being
5
6 self-confident easily irritated refreshed
enthusiastic disappointed with self healthy
7
uplifted calm (–) strong
8 proud of self drained supple
9 elated easily upset fit
1011 invigorated distressed well
1 coping bothered
achieving something overwhelmed
2
overcoming difficulties under too much pressure
3111 getting close to goals run down
4 competent
5 under control
6 attractive
well-organised
7
8 Nicholas G. Norgan, Physical Activity and Health, Steptoe, 1992, Cambridge University Press.
9
20111
1 conclusions from three types of studies, as shown in Tables 4.4 and 4.5.
2 Table 4.4 summarises narrative and meta-analytic reviews on PA and
3 mood/affect, as defined for this review. Only two published meta-analyses
4 were located (McDonald & Hodgdon, 1991; Schlicht, 1994), but neither was
5 in a refereed journal. Table 4.5 summarises large epidemiological surveys
6 conducted in Britain on physical activity and psychological well-being.
7 Studies of mood, affect and PA involve both acute and chronic effects
8 of PA. Acute effects are assessed in terms of state responses to single
9 exercise or PA sessions. Chronic effects are assessed more in terms of
30111 generalised traits from involvement in PA over time. Typically, epidemi-
1 ological surveys assess chronic PA effects whereas experimental trials
2 measure the effects of acute exercise.
3
4
Narrative and meta-analytic reviews
5
6 Of the 20 reviews listed in Table 4.4, there is cautious support for the
7 proposition that PA is associated with enhanced affect and mood. The
8 caution comes not from the lack of apparent evidence but from the rela-
9 tively weak research designs utilised. For example, the comprehensive
40111 review by Leith (1994) showed that experimental evidence was less
1 convincing than for pre- or quasi-experimental studies. Similarly, reviews
2 by Berger (1996; Berger & McInman, 1993; Wankel & Berger, 1990)
3 conclude that mood effects can be positive after exercise but causal links
4 cannot be supported, and certain conditions, such as non-competitive
45111 aerobic exercise, might need to be met for such effects to occur. However,
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Table 4.4 Summary of findings from meta-analytic and narrative reviews investigating the relationship between physical activity, mood
and affect
Study Review design and scope Results and conclusions
Leith & Taylor (1990) Narrative review of pre-, quasi- and actual Pre-experimental: 1 mood study reported. Effect of
experimental studies of exercise and PWB, exercise was positive
broadly defined Quasi-experimental: 10 mood studies; all showed
improvement with exercise. 1 affect study; no
change detected
Experimental: 4 mood studies; 1 showed positive
change, 3 no change
Wankel & Berger (1990) Narrative review of the social psychological Enjoyment reported as a main reason for sport
benefits of recreational sport. Categories of involvement. ‘. . . consistency of the accumulated
benefit classified as: personal enjoyment, results . . . is impressive’ (p. 170)
personal growth, social harmony, and social Support for link between PA and psychological
change well-being but the nature of this relationship
remains unclear
Sport has potential for developing positive values
and social integration, but only if certain conditions
are met
Biddle & Mutrie (1991) Narrative review of psychological effects of Support for positive mood being associated with
exercise in non-clinical populations exercise, but causal and experimental links still to be
established
Jex (1991) Narrative review of psychological benefits of The small number of studies reported show a
exercise in work settings positive effect for exercise, but other factors
accounting for this relationship cannot be ruled out
McDonald & Hodgdon (1991) Meta-analysis of the effects of aerobic fitness POMS effect sizes: tension (–0.322), depression
training on mood (–0.284), anger (–0.182), vigour (–0.399), fatigue
(–0.271), confusion (–0.402)
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Brown (1992) Narrative review of the relationship between Mood: 3 of 7 studies report positive effect.
PA and PWB in the elderly Life satisfaction: none of 3 studies report effect
Steptoe (1992) Narrative review of psychological well-being Exercise has a positive long-term effect on mood and
and PA PWB
This is not accounted for by selection, group
socialisation, expectations or attentional factors.
Research indicates that high intensity exercise may
confer less benefit than exercise of a more moderate
nature
Abele & Brehm (1993) Narrative review of psychological effects of Mood effects of aerobic exercise: all 15 studies
exercise and sport report positive effect
Mood effects of sport competition: studies show
increases before a game and decreases after a game
in activation and excitation
Evidence supports a ‘disequilibrium’ model of sport
mood (i.e. competitors compete to change their
mood and seek excitement through sport) and an
‘equilibrium’ model of exercise mood (i.e. to seek
tension reduction and a ‘feel better’ effect)
Berger & McInman (1993) Narrative review of the association between Mood: support for a relationship between exercise
exercise and aspects of quality of life and positive mood, but only if certain conditions are
met
Fillingim & Blumenthal (1993) Narrative review of psychological effects of Mood: aerobic exercise studies have yielded less
exercise in the elderly consistent findings than for younger adults
Hutzler & Bar-Eli (1993) Narrative review of psychological effects of Mood: 4 of 5 studies showed similar mood profiles
sport for those with disabilities for athletes with disabilities compared to those
without disabilities
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Study Review design and scope Results and conclusions
Tuson & Sinyor (1993) Narrative review of the effects of acute Anger: 6 of 12 studies showed improved scores after
exercise on affect/mood exercise; 1 worsened
Anxiety: 21 of 39 showed improvement; 4 worsened
Depression: 9 of 28 showed improvement; 0 worsened
Vigour: 6 of 15 showed improvement; 0 worsened
Fatigue: 3 of 13 showed improvement; 1 worsened
Confusion: 3 of 12 showed improvement; 0 worsened
Wykoff (1993) Narrative review of the psychological effects Non-clinical populations: support for mood
of exercise for adult women enhancing effect of exercise
Clinical populations: no studies reported on mood
Leith (1994) Narrative review of exercise and mood 26 of 34 studies reported showed improvements in
mood after exercise: this includes 2 of 2
pre-experimental studies, 19 of 24 quasi-
experimental studies, and 5 of 8 experimental
studies
Martinsen & Stephens (1994) Narrative review of exercise and mental Good population surveys are rare.
health in both clinical and ‘free-living’ Studies indicate mental health benefits; only limited
populations support for the mental health benefits of exercise
for those initially ‘well’
McAuley (1994) Narrative review of physical activity and Results of 23 published studies investigating PA and
psychosocial outcomes affect (excluding anxiety, depression, stress
reactivity, and mood state scales primarily negative
in nature) reviewed. 69% showed a positive
relationship PA and PWB
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Schlicht (1994) Meta-analysis of ‘sport’ and PWB Overall ES (0.15) was not significantly different
from zero, but ranged from –0.31 to 0.81
Studies not listed so quality and appropriateness of
selection criteria not possible to check
McAuley & Rudolph (1995) Narrative review of PA and PWB in older Affect & mood: 25 studies reported of which all but
adults 2 showed positive effects. Positive mood changes
generally more evident in men than women
Life satisfaction/HRQL: limited evidence for a
positive effect
Mutrie & Biddle (1995) Narrative review of exercise and mental Mood: generally positive effects for exercise, but
health in non-clinical populations, with an there is a limited amount of experimental research
emphasis on European research
Berger (1996) Narrative review of the psychological benefits Mood: support for a relationship between PA and
of an active lifestyle positive mood, but causal links not established
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Table 4.5 Summary of findings from British population surveys investigating the relationship between physical activity and
psychological well-being
Study Survey design and scope Results and conclusions
Sports Council & Health National Fitness Survey for England of 16–74 Small but consistent trend showing relationship
Education Authority (1992) year olds (n = 4316). One section of interview between PA and well-being. Same trend evident
assessed perceived well-being for those in poorest health, reducing the chance that
those ‘well’ choose to exercise, thus reversing the
direction of possible influence
Association between PA and well-being stronger for
those 55 years and over
Trends evident for all age groups and both sexes
Thirlaway & Benton (1996) National Health & Lifestyle Survey data. Higher PA associated with better mental health in
Representative British sample (n = 6,200) women over 30 years and men over 50 years
Assessed on PA and General Health No relationship for those under 30 years of age
Questionnaire. (Unpublished survey data
reported in book chapter)
Steptoe & Butler (1996) Investigation of the association between Greater sport/vigorous PA was positively associated
emotional well-being and regular sport/vigorous with emotional well-being independent of gender,
PA in 16-year-olds (n = 5,061). Data from 1986 SES or health status
follow-up to 1970 British Cohort Study Participation in non-vigorous activity was associated
with high psychological and somatic symptoms on
Malaise Inventory
1111 the cautiously positive conclusions from Table 4.4 are further enhanced
2 by the fact that the reviews span several countries and populations (e.g.
3 workplace, women, people with disabilities) with diverse methods and
4 measuring instruments. Hardly any studies report negative mood effects.
5 McDonald and Hodgdon (1991) report a meta-analysis on PA and mood.
6 They delimited their review to aerobic fitness training studies and found
7 that researchers used mainly the POMS (tension, depression, anger,
8 fatigue, vigour, and confusion) or MAACL (anxiety, depression, and
9 hostility). Results are shown in Table 4.4 and suggest a clear relationship
1011 between exercise and vigour and a lack of negative mood, although the
1 effect sizes are generally small-to-moderate.
2 Interestingly, Schlicht (1994) reported a brief meta-analytic review of
3111 exercise and mental health in which the proposed relationship between
4 physical activity and PWB was not supported. The analyses involved 44
5 samples from 39 studies and 8,909 research participants. However, the
6 studies selected for the meta-analysis were not listed in his paper and addi-
7 tional information is required on study selection criteria before more can
8 be concluded. Given that the paper was published in German it would be
9 interesting to see how many of the studies included are written in German,
20111 and whether many are in English and thus more likely to be included in
1 the North American meta-analyses that have supported a link between
2 physical activity and other indices of mental health (e.g. Petruzzello,
3 Landers, Hatfield, Kubitz, & Salazar, 1991). Nevertheless, Schlicht (1994)
4 reported an overall ES of only 0.15, but with a large range, and the over-
5 all ES was not significantly different from zero (see Table 4.4). This may
6 reflect that similar forms of PA are viewed favourably by some individuals
7 but not by others, thus reducing any psychological effect. In other words,
8 the effects of PA on mental health may be quite individual.
9
30111
Epidemiological surveys
1
2 Epidemiological surveys, while often suffering from methodological short-
3 comings such as lacking internal validity and control (see Chapter 8), have
4 the advantage over some other studies in so far as they usually have large
5 samples, are representative of the population, and hence allow good gener-
6 alisability of findings. Three such studies from Britain are summarised in
7 Table 4.5 and show clear positive relationships between PA and psycholog-
8 ical well-being. Confidence in these results is enhanced by noting that the
9 surveys cover both adolescents and adults, use clinical and non-clinical
40111 assessment tools, and cover a total sample of 15,577. However, as noted by
1 Thirlaway and Benton (1996), not all groups seemed to benefit from PA.
2 The three British studies are comparable to Stephens’ (1988) secondary
3 analysis of four North American surveys with over 55,000 people. Across
4 several measures, there was a clear association between PA and psycho-
45111 logical well-being. For example, positive affect was associated with PA
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Table 4.6 Controlled experimental trials from Britain investigating physical activity, exercise and psychological well-being
Study Participants Design and treatment Results and conclusions
Steptoe & Cox (1988) Female Single session experiment testing the Moderate intensity exercise produced more
students (n = 32) effects of exercise intensity and music negative mood states (increased tension-
on mood anxiety, reduced vigour and exhilaration)
78 Stuart Biddle
All participants exercised for four Low intensity exercise produced

periods: at both low and moderate favourable mood state changes
intensity with music and metronome Ratings of Perceived Exertion were
slightly lower when exercising with
music rather than a metronome. No mood
effects for music
Steptoe & Bolton Female students Replication and extension of Steptoe Immediately after higher intensity exercise,
(1988) (n = 40) & Cox (1988) participants reported higher tension-anxiety
Exercised for 15 mins at either and mental fatigue than those in the low
moderate or low intensity intensity condition
Both groups showed a decline in these
states during the exercise recovery period
Moses et al. (1989) Sedentary adults An experimental study of the effects Only the moderate intensity exercise group
(n = 109) of exercise training on mental well-being. showed reductions on the tension-anxiety
Participants assigned to either high and confusion mood scales and a measure
intensity aerobic exercise, moderate coping deficits
intensity aerobic exercise, attention-
placebo, or wait-list control
10-week training period undertaken
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Parfitt et al. (1994) Students (n = 80) Experimental test on affective reactions High-active participants reported greater
to exercise as a function of exercise positive affect in the high intensity condition
intensity and exercise history. in comparison to the low-active group.
High and low active participants No differences at the lower intensity
reported psychological affect in the last More positive affect reported 5 mins
30 secs and 5 mins after exercising at post-exercise compared with last minute
60% and 90% of V·O2max of the exercise bout
Parfitt et al. (1996) High (n = 15) Test of affective reactions during and High-active reported more positive affect
and low active after exercise on cycle ergometer. than low-active
(n = 15) women Exercise took place at 3 different Affect was more positive 5 mins
in mid-20s levels of RPE: 9, 13 & 17 post-exercise than in the last 20 secs of the
Affect assessed with FS (Feeling Scale) exercise bout
Affect was progressively less positive as
RPE increased
80 Stuart Biddle
1111 for both men and women in the two age groups under and over 40 years.
2 Stephens (1988, pp. 41–42) provided the following clear conclusion:
3
4 the inescapable conclusion of this study is that the level of physical
5 activity is positively associated with good mental health in the house-
6 hold populations of the United States and Canada, when mental health
7 is defined as positive mood, general well-being, and relatively infre-
8 quent symptoms of anxiety and depression. This relationship is
9 independent of the effects of education and physical health status,
1011 and is stronger for women and those age 40 years and over than for
1 men and those age under 40. The robustness of this conclusion derives
2 from the varied sources of evidence: four population samples in two
3111 countries over a 10-year period, four different methods of opera-
4 tionalizing physical activity and six different mental health scales.
5
6
Evidence from experimental trials
7
8 Few studies have investigated the effects of PA on affect and mood through
9 controlled experimental trials. Five British studies are summarised in Table
20111 4.6. These show clearly that the intensity of exercise is important in deter-
1 mining the effects of exercise on mood, something not suggested in other
2 types of studies such as epidemiological surveys. The three studies by
3 Steptoe and his colleagues (Steptoe & Bolton, 1988; Steptoe & Cox, 1988;
4 Moses, Steptoe, Mathews, & Edwards, 1989) show that moderate, but not
5 high intensity exercise, has mood enhancing effects. Similarly, Parfitt,
6 Markland, & Holmes (1994) show that feeling states in exercise are signifi-
7 cantly worse at a higher intensity for less active individuals. Moreover,
8 Parfitt, Eston and Connolly (1996) showed that low-active women reported
9 more negative affect at a higher exercise intensity than high-active women.
30111 Raglin has suggested that ‘sensations associated with exertion and post-
1 exercise fatigue following high-intensity activity delay, but do not eliminate,
2 post-exercise anxiety reductions’ (1997, p. 117, emphasis added). The
3 increases in negative mood after high intensity exercise reported in
4 Steptoe’s research may be due to the higher exertion required, but studies
5 have shown that positive mood is still enhanced some time later. Indeed,
6 Parfitt et al. (1996) showed that for both high-active and low-active women,
7 positive affect was higher 5 minutes after exercise in comparison to the
8 last 20 seconds of the exercise bout, and this was most pronounced for
9 the highest level of exercise intensity, thus supporting Parfitt et al. (1994).
40111 The temporal nature of changes in mood after different intensities of
1 exercise requires further investigation. However, even if the post-exercise
2 negative mood effect is transitory, it may be enough to affect adherence
3 and reduce physical activity participation.
4 In a controlled trial of healthy American adults, King, Taylor, Haskell
45111 and DeBusk (1989) found that participants assigned to a 6-month PA
1111 intervention, in comparison to controls, showed significant improvements
2 in body appearance satisfaction, perceived physical fitness, and satisfaction
3 with weight. No differences, however, were found for depressed mood,
4 tension/anxiety, or confidence/well-being. These results suggest that
5 psychological changes are more likely if they are closely linked to the
6 physical changes associated with an exercise programme.
7
8
Exercise and sub-clinical depression
9
1011 Exercise and clinical depression is covered by Nanette Mutrie in this
1 volume. However, many studies have investigated participants who have
2 not reached clinically defined levels of depression, but may be suffering
3111 transitory negative depressive mood. McDonald and Hodgdon (1991) iden-
4 tified five measures of depression in their meta-analysis of aerobic training
5 studies. These were the BDI, the Centre for Epidemiological Studies
6 Depression Scale (CES-D) (Radloff, 1977), Lubin’s (1965) Depression
7 Adjective Check List (DACL), the Symptom Check List 90 (SCL-90)
8 (Derogatis, Lipman, & Covi, 1973), and Zung’s (1965) Self-Rating
9 Depression Scale (SDS). In addition, the POMS depression subscale has
20111 been used (see Leith, 1994) although McDonald and Hodgdon used this
1 as part of their analysis of mood rather than depression per se.
2
3
Meta-analytic reviews
4
5 Two meta-analyses have been conducted on exercise and depression.
6 McDonald and Hodgdon (1991) have also meta-analysed depression as
7 an outcome variable for their study of aerobic fitness training. In addi-
8 tion, North et al. (1990) reported a meta-analysis of 80 studies yielding
9 290 effect sizes on exercise and depression. The mean ES was 0.53.
30111 Similarly, McDonald and Hodgdon found an overall ES of 0.55 for their
1 ‘depression cluster’, which included various depression and related mood
2 scales.
3 Although these meta-analyses probably constitute the best evidence
4 available to date, there are a number of issues that should caution over-
5 confidence. Many of these are argued well by Dunn and Dishman (1991)
6 and Dishman (1995). For example, some studies in the meta-analyses may
7 have included individuals suffering from depression with a primary anxiety
8 component. Dunn and Dishman argue this point on the basis of evidence
9 that shows a large number of people meeting Diagnostic and Statistical
40111 Manual (DSM-II-R) criteria for agoraphobia and panic attacks also suffer
1 from depression or have a history of depression, hence exercise may reduce
2 state anxiety and elevate mood which could then produce changes in
3 depression. North et al.’s (1990) meta-analysis is also questioned on the
4 basis of non-uniformity in defining depression, as well as the discrepancy
45111 between the results of the meta-analysis and other studies.
82 Stuart Biddle
1111 Epidemiological surveys
2
The large-scale survey analysis reported by Stephens (1988) provides
3
evidence at the level of epidemiological data. Concerning depression
4
assessed with the CES-D, results for over 3,000 North American adults
5
from the first National Health and Nutrition Examination Survey
6
(NHANES-I) showed that depression was highest for those reporting
7
‘little/no exercise’ in comparison to those classified in the ‘moderate’ and
8
9 ‘much’ exercise categories. Interestingly, this difference is suggestive that
1011 only moderate exercise may be sufficient for anti-depressant effects and
1 that additional activity yields no additional benefit. Further support
2 was provided in follow-up data in NHANES-II (Farmer, et al., 1988).
3111
4 Evidence from experimental trials
5
6 Leith (1994) reports 42 studies investigating exercise and depression and
7 81% show anti-depressant effects. Of these, 25 appear to deal with sub-
8 clinical depression and 84% show positive effects for exercise. Of the 6
9 sub-clinical studies Leith classifies as experimental, all but one reported
20111 positive changes in depression.
1
2 Do mood and emotion effects vary across people and settings?
3
4 There is some evidence in epidemiological surveys (e.g. Stephens, 1988)
5 that more positive affect results from physical activity for women and
6 those over 40 years of age. However, when specific aspects of affect are
7 studied, such as depression, the picture is not clear. Similarly, it is not
8 known whether specific forms of exercise, such as aerobic exercise, are
9 more beneficial than others when affect is considered. Some might argue
30111 that we will never resolve this issue since people will ‘feel good’ after
1 exercise they prefer, and feel ‘less good’ after exercise that is not to their
2 liking. Such individual differences will mask any population trends linking
3 PA and affect. However, this has yet to be shown.
4 Nevertheless, one area of research in sport psychology may shed some
5 light on contextual factors influencing exercise affect. The approach people
6 have in some physical activities (their goals), and the environment
7 perceived by the participants (climate) may be important. We have studied
8 two main achievement goals in PA. A task goal orientation is held when
9 success is defined primarily in terms of self-improvement and task mastery.
40111 It is highly correlated with the belief that effort will bring success. An
1 ego goal orientation is held when success is defined in terms of winning
2 and demonstrating superiority over others. This correlates highly with the
3 belief that ability is necessary for success (see Duda, 1993).
4 We conducted a meta-analysis of 37 studies, with a total of 41 inde-
45111 pendent samples (n 7,950), investigating the relationship between task
1111 and ego goals and positive (PA) and negative affect (NA) (Ntoumanis &
2 Biddle, 1999a). After correcting for measurement and sampling error, the
3 correlation between task orientation and positive affect was moderate-
4 to-high (0.55). The other correlations were generally small: task – NA
5 0.18; ego – PA 0.10; ego – NA 0.04. These results suggest that
6 adopting a task goal orientation in exercise will lead to more positive
7 affective reactions. This may be due to greater perceptions of control and
8 higher intrinsic motivation.
9 In another review, we calculated effect sizes for the relationship between
1011 task and ego climates in PA and positive and negative affect (Ntoumanis
1 & Biddle, 1999b). Climates refer to the perception of contextual cues in
2 a situation (e.g. exercise class) that may emphasise more of a task climate
3111 or an ego climate. In the former, group members perceive they have
4 greater decision-making involvement, success is defined and evaluated in
5 terms of individual effort and improvement, and new learning strategies
6 are encouraged. An ego climate, on the other hand, emphasises inter-
7 personal comparison, and evaluation is based on normative standards.
8 Calculations from 14 studies (n 4,484) revealed that a task climate
9 was associated quite strongly with positive affective and motivational
20111 outcomes, such as satisfaction and intrinsic motivation (Effect Size [ES]
1 0.71). Conversely, an ego climate was associated with positive outcomes
2 in a negative direction (ES 0.30). Negative outcomes, such as worry,
3 were negatively associated with a task climate (ES 0.26) and positively
4 with an ego climate (ES 0.46). These results suggest that a task climate
5 is associated with greater positive and less negative affect. Of course we
6 cannot conclude whether those in a task climate, or those with a task goal,
7 are better off than non-participants, but given the evidence presented else-
8 where in this review, it seems highly likely that they will be.
9
30111
A brief overview of possible mechanisms
1
2 The review so far suggests that physical activity is associated with posi-
3 tive mood and affect. However, this is not enough. We also need to know
4 more about why and how such effects occur. This necessitates a brief
5 discussion on the mechanisms of the links between PA and PWB, although
6 a more generic discussion is provided in the final chapter.
7 Mechanisms for the effects of PA on mood and affect have not been
8 clearly identified. Several proposed mechanisms are plausible, including
9 biochemical, physiological, and psychological (see Biddle & Mutrie, 1991;
40111 Boutcher, 1993; Morgan, 1997). Possible biochemical and physiological
1 mechanisms include changes associated with an increase in core body
2 temperature with exercise (thermogenic hypothesis; see Koltyn, 1997),
3 increase in endorphin production following exercise (endorphin hypoth-
4 esis; see Hoffmann, 1997), changes in central serotonergic systems from
45111 exercise (serotonin hypothesis; see Chaouloff, 1997), and the effects of
84 Stuart Biddle
1111 exercise on neurotransmitters (e.g. norepinephrine hypothesis; see Dish-
2 man, 1997). In addition, the ‘feel better’ effect from PA may result from
3 changes in self-esteem from mastering new tasks, or from time away
4 from negative or more stressful aspects of our lives.
5 In an elegant analysis of possible mechanisms and their interaction with
6 exercise experience, Boutcher (1993) proposes that for those just starting
7 exercise (i.e. in the ‘adoption phase’), greater emphasis should be placed
8 on psychological mechanisms since the exerciser had not adapted, physi-
9 ologically, to the PA stimulus. In the maintenance phase, Boutcher suggests
1011 that both psychological and physiological mechanisms are likely to be
1 important, and in the final habituation phase, he suggests that emphasis
2 should be placed on physiological mechanisms and the influence of behav-
3111 ioural conditioning.
4
5
Guidelines for research and practice in health settings
6
7 It is clear from this review that physical activity is associated, in a positive
8 way, with emotion, mood and psychological well-being. Despite method-
9 ological difficulties in this area, health professionals should have some
20111 confidence in the promotion of physical activity for the purposes of
1 enhancing psychological well-being, however defined.
2
3
Guidelines for research
4
5 This review has drawn evidence from numerous narrative reviews, two
6 meta-analytic reviews, several large-scale surveys, and a few well-designed
7 experimental trials. Despite this, research into emotion, mood and PWB
8 and physical activity is fraught with difficulties. First, the measurement of
9 PWB has been inconsistent or restrictive in nature. Affect and mood are
30111 difficult to define, at least in operational terms, and when they have been
1 assessed, many different instruments have been used. In the assessment
2 of mood, the POMS has been the dominant instrument, yet, as already
3 alluded to, this is narrow and allows for the assessment of only one posi-
4 tive factor. In short, generalised measures of affect and mood are in need
5 of development to be suitable for physical activity research.
6 To this end, two recent measures of exercise-related affect have been
7 developed (EFI and SEES). These allow for the acute effects of exercise
8 to be assessed and should be tested further in this country during exer-
9 cise trials. However, they present difficulties when assessing the effects of
40111 alternative treatments and do not solve the problem of using a consistent
1 measurement technology for assessing the chronic effects of physical
2 activity and exercise.
3 The work of Steptoe and his colleagues (see Steptoe, 1992) has demon-
4 strated that moderate rather than high intensity exercise is associated with
45111 positive well-being. This requires further testing with more varied intensities
1111 and activities, as well as investigating whether high intensity exercise is
2 associated with PWB but only after some delay.
3 It is not possible, at this stage, to draw clear conclusions concerning the
4 characteristics of participants or exercise modalities. Some surveys have
5 suggested that women and older individuals gain most from physical
6 activity, and that aerobic over anaerobic exercise produces superior
7 psychological effects. However, these trends have not always been
8 confirmed and further work is required.
9
1011
Guidelines for practice
1
2 It has been concluded that physical activity is consistently associated with
3111 positive affect and mood. Although we are unable to state with confidence
4 that physical activity causes positive affect, it is clear that an association
5 exists for both acute and chronic involvement. Health professionals can,
6 with confidence, promote moderate physical activity in the knowledge that
7 ‘good mental health’ will be correlated with such behaviours. The current
8 message associated with the Active for Life campaign in England is wholly
9 consistent with this approach. Advocating more vigorous exercise appears
20111 to be less easy to justify, except in fitter individuals.
1 If positive affect is a desired outcome of physical activity promotion by
2 health professionals, evidence supports the use of moderate aerobic PA and
3 the adoption of a task goal orientation. Moderate PA, such as walking and
4 cycling, are currently promoted as key activities in Active for Life. These
5 types of activities have the added advantage of being possible to fit into a
6 typical daily schedule with minimal disruption and thus aid adherence.
7 As far a task orientation is concerned, this means that individuals playing
8 sport or taking part in a physical activity where some form of ‘achieve-
9 ment’ is salient, should adopt reference standards that are internally
30111 focused. In other words, success should be judged in terms of personal
1 progress and effort. Similarly, a mastery (task-oriented) motivational
2 climate is associated with greater PWB. Promotion of this group envi-
3 ronment entails promoting individual challenges, involving group members
4 in decision making, recognising individual progress, having evaluation
5 based on individual progress, and providing opportunities for practice and
6 improvement (Ames, 1992).
7 Although there will be a wide variety of types of physical activities that
8 are associated with PWB, the promotion of moderate aerobic PA, empha-
9 sising personal effort, progress and participation, seems an appropriate
40111 strategy at this point in our knowledge.
1
2
3
4
45111
86 Stuart Biddle
1111
2 What we know
3
4 Based on the evidence reviewed, the following statements are offered
5 for what we currently know about the relationship between affect,
6 mood and physical activity:
7
8 • Physical activity is consistently associated with positive affect
9 and mood.
1011 • Where quantified trends have been identified, aerobic exercise
1 has a small-to-moderate effect on tension (–), depression (–),
2 vigour (+), fatigue (–), and confusion (–), and a small effect on
3111 anger (–).
4 • A positive relationship between physical activity and psycholog-
5 ical well-being has been confirmed in several large epidemi-
6 ological surveys, including in the UK, using different measures
7 of activity and well-being.
8 • Experimental trials support a positive effect for moderate inten-
9 sity exercise on psychological well-being.
20111 • Meta-analytic evidence shows that adopting a goal in exercise
1 that is focused on personal improvement, effort, and mastery
2 has a moderate-to-high association with positive affect.
3 • Meta-analytic evidence shows that a group climate in exercise
4 settings focused on personal improvement and effort has a
5 moderate-to-high association with positive affect.
6
7
8
30111
1 Our knowledge is far from complete. The following questions
2 address key concerns and represent what we need to know:
3
4 • Are the associations between physical activity and psycholog-
5 ical well-being causal?
6 • Do different types of physical activity produce different affec-
7 tive responses?
8 • Is physical activity likely to produce superior psychological
9 effects for some groups, such as women?
40111 • When might high intensity exercise produce positive affective
1 responses?
2 • Are current psychometric measures of HRQL and exercise-
3 related affect adequate for capturing the range of affective
4 responses in physical activity?
45111
1111
2 • Are current psychometric measures of HRQL and exercise-
3 related affect adequate for assessing change over time?
4 • What mechanisms explain the link between PA, affect and
5 mood?
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
4
45111
1111
2 5 The effects of exercise
3
4 on self-perceptions and
5
6
self-esteem
7
8 Kenneth R. Fox
9
1011
1
2
Background
3111
4 Volumes of research have been generated on the topic of self-esteem and
5 self-concept to the point where it is difficult to find a psychological
6 construct that has attracted more academic attention. Self-esteem is also
7 one of the few psychological terms that has acquired a meaning among
8 the general public. It regularly crops up in informal conversations, usually
9 in the context of explaining particular mental states and behaviours.
20111 Reference to self-esteem also features in formal policy documents of a
1 range of organisations and institutions. The National Curriculum for
2 schools in England and Wales, for example, places enhancement of self-
3 esteem as a major curricular goal. Corporations include improvement in
4 mental well-being and self-esteem as an important target for the welfare
5 of their workforce. Health interventions, particularly programmes to facili-
6 tate rehabilitation from substance abuse, acute and chronic injury and
7 disease, often focus on improved self-esteem as a primary objective. More
8 recently, self-esteem has been considered as an important aspect of quality
9 of life and mental well-being and as such has been considered as a possible
30111 target for public health campaigns.
1 Why is so much significance attached to a phenomenon that merely
2 exists in the mind of the individual as a mental abstraction? Several
3 features make self-esteem and other self-perception constructs very rele-
4 vant to health.
5
6 • Self-esteem is widely accepted as a key indicator of emotional stability
7 and adjustment to life demands. High self-esteem has been related to
8 a range of positive qualities such as life satisfaction, positive social
9 adjustment, independence, adaptability, leadership, resilience to stress,
40111 and high level of achievement in education and work. Self-esteem has
1 emerged therefore as one of the strongest predictors of subjective well-
2 being (Diener, 1984) and is consequently an important element of
3 mental well-being and quality of life.
4 • Self-esteem and related self-perceptions are closely implicated with
45111 choice and persistence in a range of achievement and health behaviours.
Exercise and self-esteem 89
1111 Many contemporary theories of human motivation feature elements of
2 the self. Needless to say, we enjoy feeling good about ourselves and it is
3 clear that we tend to gravitate to those settings in life which provide
4 opportunities for high self-ratings. High self-esteem is associated with
5 healthy behaviours (particularly in adolescents) such as not smoking,
6 lower suicide risk, greater involvement in sport and exercise, and
7 healthier eating patterns (Torres & Fernandez, 1995). Self-esteem
8 and self-perceptions of ability are therefore critical to understanding
9 determinants of health behaviours and this is evident in patients in the
1011 primary care setting (Hurst, Boswell, Boogard, & Watson, 1997).
1 • Low self-esteem is closely related to mental illness and absence of
2 mental well-being. It frequently accompanies depression, trait anxiety,
3111 neuroses, suicidal ideation, sense of hopelessness, lack of assertive-
4 ness and low perceived personal control. Improved self-esteem has
5 therefore been used frequently as a target for change and also as a
6 success marker for psychotherapy (Wylie, 1979).
7
8 In essence, beyond the satisfaction of basic physiological needs such as
9 food and warmth, there is little of more importance to an individual than
20111 maintaining a high degree of self-esteem. The search for self-esteem is
1 considered so strong that it has been termed by Campbell (1984) as the
2 First Law of Human Nature and many theorists believe that sense of self
3 is central to the possession of mental and even physical health.
4
5
Self-esteem and the self-system
6
7 In order to meaningfully overview the literature on exercise and its impact
8 on self-esteem, it is necessary to first unravel the tangle of constructs
9 involved. The self is best described as a complex system of constructs.
30111 Theorists believe that these may be organised by a self director who acts as
1 an information processor and decision maker. Information relevant to the
2 self is gathered and organised to form a self-description, termed self-
3 concept, identity, or set of identities based on its abilities, qualities, traits and
4 the roles it performs. Murphy (1947, p. 996) describes the self-concept as
5 ‘the individual as known to the individual’. Roles in several life domains
6 may contribute to the self-concept and might include perceptions of self at
7 work, in social relationships, in the family, and also the physical self which
8 is dictated by qualities related to our appearance and physical prowess.
9 The self director will invest time in directing choice and persistence in
40111 activities and use a range of self-promotion and self-presentation strategies
1 (including self-serving biases and defensiveness) to achieve the best results
2 for self. The bank balance at the end of the day constitutes self-esteem or
3 self-worth. Whereas self-concept is a self-description, self-esteem is a self-
4 rating of how well the self is doing. Campbell (1984) defines it as ‘an aware-
45111 ness of good possessed by self’. The criteria and content used to determine
90 Kenneth Fox
1111 worth are dictated both by the individual and the primary culture in which
2 he/she operates. Additionally, individuals might ascribe to subcultures that
3 value other aspects of life such as athletic ability, higher spiritual or moral
4 ground, or even criminal behaviour. Within these constraints, each person
5 will draw upon a personal menu of attributes and achievements, dependent
6 on exposure and experience, placing greater value on some elements than
7 others. Some personal menus may closely conform to cultural norms and
8 expectations, while others might be more individualised. However, the
9 criteria on which self-esteem is based are ultimately set by the individual.
1011 Self-esteem is therefore essentially phenomenological and based on being
1 an ‘OK person’ dependent on what the individual considers as ‘OK’. This
2 is an important principle as it suggests that the effect of exercise on self-
3111 esteem cannot be explained in the absence of consideration of the past
4 experiences and values of the individual.
5
6
Measuring self-esteem and self-perceptions
7
8 A great deal of research was generated on self-esteem in the 1970s and
9 80s. Instruments consisted of banks of items, each calling for a response
20111 on perceived possession of some personal quality or competence such as
1 having attractive facial features, academic ability, or lots of friends. These
2 responses were simply totalled to produce a self-esteem score, a technique
3 that has since been widely criticised as it does not take into account the
4 multidimensionality of the self (Marsh, 1997; Wylie, 1979, 1989). More
5 recently, a profile approach has been adopted where instruments are made
6 up of several subscales each assessing self-ratings in different aspects of
7 life or domains of competence such as work, family and friendships. Overall
8 or global self-esteem or self-worth is best measured by a separate subscale
9 using items which avoid specific domain content and refer to pride in self,
30111 general competence, and equal worth to others. Rosenberg’s 10-item
1 Global Self-Esteem Scale (Rosenberg, 1965) has been widely used and
2 validated for adolescents onward. Harter’s self-perception profiles for
3 adolescents and adults include a General Self-Worth subscale of this nature
4 (Harter, 1988; Messer & Harter, 1986), as does Marsh’s Self-Description
5 Questionnaire series (Marsh, 1992a, 1992b)
6 Following the use of self-perception profiles and separate global or self-
7 esteem scales, models depicting how dimensions are related to self-esteem
8 have been offered and in some cases tested. Some support has been
9 provided for a hierarchical structure like the roots of a tree, with self-
40111 esteem forming the stable apex or tree trunk. Domains of life form the
1 main roots, with increasingly finer roots that search out closer contact
2 with life experiences and represent more specific content (see Figure 5.1).
3 Of particular significance to exercise and mental health is the physical
4 self. This is consistently featured as a strong root in the self system with
45111 an overall physical self-worth underpinned by a range of physical attributes
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
General
General Self-Concept
Nonacademic Self-Concept
Academic and
Nonacademic Academic Social Emotional Physical
Self-Concept Self-Concept Self-Concept Self-Concept Self-Concept
Significant Particular Physical Physical

Subareas of English History Math Science Peers Others Emotional Ability Appearance
Self-Concept States
Evaluation of
Behaviour in
Specific
Situations
Figure 5.1 A hierarchical model of self-concept.

From R.J. Shavelson, J.J. Hubner, and G.C. Stanton. (1976). ‘Self-concept: Validation of Construct Interpretations.’ Review of Educational
Research, 46, p.413. Copyright by the American Educational Research Association, reprinted by permission of the publisher.
92 Kenneth Fox
1111 and competencies. With the development and validation of hierarchical
2 models, it has become possible to locate measures of self-perception
3 according to the specificity of their content (see Figure 5.2).
4 The Tennessee Self-Concept Scale (TSCS) was one of the first instru-
5 ments to utilise a multidimensional structure and the review featured
6 later, in this chapter will identify it as the instrument of choice in many
7 of the earlier exercise/self-esteem studies. It consists of a physical self
8 subscale as well as subscales to assess moral/ethical, personal, family
9 and social dimensions in addition to a lie scale. Unfortunately, the phys-
1011 ical self subscale totals diverse items, many of which have little relevance
1 to exercise. Marsh and Richards (1988) through confirmatory factor
2 analysis, criticised the instrument for poor psychometric qualities.
3111 Results using the TSCS are likely to be less convincing than they
4 might be, although the instrument has recently been upgraded and may
5 be particularly useful for clinical settings with patients with psychiatric
6 disorders (Byrne, 1996).
7 Two well-validated comprehensive instruments have been developed in
8 recent years to assess self-ratings at two levels of the physical domain.
9 The Physical Self-Perception Profile (Fox & Corbin, 1989) measures
20111 perceptions of sport competence, physical strength, physical condition,
1 body attractiveness and overall self-worth. The Physical Self-Description
2 Questionnaire measures nine elements of the physical self (Marsh,
3 Richards, Johnson, Roche, & Tremayne, 1994), general physical self
4 and general self-esteem. In addition to profiles, there are also instruments
5 to measure singular aspects of the physical self. Aspects of body appear-
6 ance have been assessed for many years such as body image, body
7 satisfaction (with whole and parts), body acceptance, and more recently
8 social physique anxiety (anxiety associated with displaying the body in
9 public settings).
30111
1 LEVEL 1 Global SELF-ESTEEM
2
3 LEVEL 2 Domain Physical self-worth
4
5 LEVEL 3 Subdomain Sport competence Attractiveness
6 LEVEL 4 Facet Soccer competence Fatness
7
8 LEVEL 5 Subfacet Shooting ability Fatness of hips
9
40111 LEVEL 6 Situation-specific Scoring efficacy Feeling fat hips
in these clothes
1
2
Figure 5.2 Self-perception constructs measurable at different levels of specificity.
3
From K.R. Fox (1998). ‘Advances in the Measurement of the Physical Self’. In J.L. Duda,
4 Advances in Sport and Exercise Psychology Measurement. Morgantown, WV: Fitness
45111 Information Technology.
1111 It is also possible to assess quite specific aspects of the physical self that
2 have particular relevance to an intervention. Self-efficacy measures fall
3 into this category and represent an individual’s perceived confidence in
4 their ability to successfully complete a task such as climbing stairs, visiting
5 a swimming pool, maintaining a three-times-a-week exercise programme.
6 Although these self-ratings do not represent self-esteem change, they may
7 help identify possible mechanisms through which self-esteem might be
8 enhanced. These measures are of particular significance to interventions
9 with older people and those in rehabilitation where confidence in move-
1011 ment ability may initially be low.
1 What is clear is that an array of measures to assess self-perceptions and
2 self-esteem is available to the health professional or researcher who wishes
3111 to document possible change due to participation in an exercise programme.
4 It is not within the scope of this chapter to review all available instruments.
5 The interested reader should see Byrne (1996) for the whole range of self-
6 concept instruments. For the assessment of physical self-perceptions, see
7 Fox (1998), for body image, see Bane and McAuley (1998) and for self-
8 efficacy and exercise-related confidence, see McAuley and Mihalko (1998).
9 Unfortunately, it is only recently that the more theoretically grounded
20111 and comprehensive instrumentation has appeared in the exercise/self-
1 esteem intervention literature so that much of the evidence is based on
2 poorly validated assessments. Furthermore, there has not been a systematic
3 approach adopted. The general view is that specific elements are more
4 accessible to change, and Sonstroem and Morgan (1989) have presented a
5 testable model indicating how experiences with exercise might improve self-
6 efficacy and eventually affect physical self-worth and self-esteem (Figure
7 5.3). Models such as this could offer a consistent framework for furthering
8 exercise and self-esteem research but to date published studies are not in
9 evidence.
30111 Recently, Sonstroem, Harlow and Josephs (1994) modified the original
1 model of Sonstroem and Morgan (1989) in conjunction with the Physical
2 Self-Perception Profile (Figure 5.4). The relationships in the model were
3 supported through structural equation modelling. This provides an
4 example of how improved instrumentation in combination can offer a
5 more comprehensive and systematic framework for the study of self-
6 perception change through exercise.
7
8
The potential for exercise in the promotion of self-esteem
9
40111 However, the focus of interventions has been the simple description of
1 change and little research has been conducted on the identification of mech-
2 anisms of change. There are many potential candidates. High value is
3 attached in the dominant western culture to physical attractiveness
4 (particularly for women), and a range of competencies and status indica-
45111 tors such as educational attainment, success at work, physical and artistic
94 Kenneth Fox
1111
2
3
4 General Self- Self-
esteem esteem
5
6 I
7
Self-perceptions
n
8 t
9 e Physical Physical
Physical Physical
r competence acceptance
1011 competence acceptance
v ∆ ∆
1 e
2 n
3111 t
4 i
5 Specific Physical o Physical
self-efficiacy n self-efficiacy
6
7
8
9 Physical measures
20111
1
Test 1 Test 2 . . . nth test
2 Time
3
4 Figure 5.3 The Exercise and Self-Esteem Model for intervention studies.
5 W.P. Morgan, and R.J. Sonstroem (1989). ‘Exercise and self-esteem: Rationale and model.’
6 Medicine and Science in Sports and Exercise, 21, 329–337.
7
8 skills, sporting ability, and to some extent wealth and material possessions.
9 With improved competence comes a sense of effectiveness, feelings of self-
30111 determination and personal control. These are tied to self-esteem. Further-
1 more, self-acceptance – the degree to which we accept our strengths and
2 weaknesses, may also influence self-esteem, based on the assumption that
3 we cannot all excel at everything. These are important considerations in the
4 design of interventions to promote self-esteem. It is possible, for example,
5 that self-esteem can be lowered through experiences in the physical domain
6 if the conditions raise awareness and self-criticism without increasing per-
7 ceived competence. In addition, several theorists argue that humans have
8 a need to sense social significance as reflected by feelings of power, impor-
9 tance, relatedness, belonging, love worthiness, and unconditional worth.
40111 This social need may offer a further route to self-esteem enhancement.
1 The physical domain features strongly in the value system of the
2 western culture and as a result is consistently included in models of
3 self-esteem. Elements of physical self are particularly significant as the
4 body functions as the public interface of the self with the social world and
45111 is used to project characteristics such as status, sexuality, youthfulness,
1111
2
3
SELF
4
ESTEEM
5
6
7
8
9
1011
1
2 PHYSICAL PHYSICAL
SELF-WORTH ACCEPTANCE
3111
4
5
6
7
8 Sport Physical Physical Attractive
Competence Strength Condition Body
9
20111
1
2
3
4 Exercise
5 Efficacy
6
7
Figure 5.4 Adaptation of the Exercise and Self-Esteem Model for use with the
8 Physical Self-Perception Profile (after Sonstroem, Harlow, & Josephs,
9 1994).
30111 From K. R.Fox (1998). ‘Advances in the Measurement of the Physical Self.’ In J.L. Duda,
1 Advances in Sport and Exercise Psychology Measurement. Morgantown, WV: Fitness
Information Technology.
2
3
4 and prowess. For this reason, the physical self may be particularly im-
5 portant in the development of self-esteem. Cross-sectional research has
6 indicated that body image provides the strongest correlation with self-
7 esteem (r 0.6–0.8) throughout the lifespan. Physical skills, fitness, and
8 sport competencies are also important to many, especially youngsters
9 as they grow and learn to make comparisons, but the strengths of
40111 these correlations vary among populations. The potential mechanisms by
1 which involvement in exercise or sport might promote self-esteem there-
2 fore are broadly:
3
4 • An undetermined psychophysiological mechanism that enhances mood
45111 and positive self-regard.
96 Kenneth Fox
1111 • Enhanced body image, body satisfaction or body acceptance through
2 weight loss or improved muscle tone.
3 • Enhanced perceived physical competence through improved abilities,
4 prowess, and aspects of fitness such as strength and cardiorespiratory
5 function.
6 • Enhanced sense of autonomy and personal control over the body, its
7 appearance, and functioning.
8 • Improved sense of belonging and significance through relationships
9 with exercise leaders or others in the exercise group.
1011
1 For further discussion of potential mechanisms some of which have yet
2 to be adequately researched, see Fox (1997) and Sonstroem (1997a, 1997b).
3111 It is in the context of these recent developments in theory and instru-
4 mentation that the literature investigating the influence of exercise on
5 self-esteem should be examined. This is particularly the case for the phys-
6 ical self which has emerged as a strong correlate of global self-esteem and
7 a likely location for the operation of the main mechanisms of its change.
8
9
Evidence of the effect of exercise on self-esteem
20111
1 Against this background, the second section of this paper will provide an
2 update and overview of existing research that is relevant to the effect
3 of exercise on self-esteem and self-perceptions. The following sources
4 were used:
5
6 • electronic data bases including Medline, Psychinfo, PsychLit, Sport
7 Discus, with follow up using BIDS;
8 • previous reviews with at a least a section on self-esteem/self-concept
9 including Berger and McInman (1993), Calfas and Taylor (1994), Doan
30111 and Scherman (1987), Gleser and Mendelberg (1990), Leith (1994),
1 Leith and Taylor (1990), Sonstroem, (1984), Sonstroem (1997a, 1997b),
2 and Spence (unpublished);
3 • personal records of papers and abstracts.
4
5
Cross-sectional research
6
7 The early descriptive literature related to self-esteem, physical activity,
8 exercise and sport is vast, with much of it lacking theoretical grounding
9 and generated with poor instrumentation. Studies generally fall into three
40111 categories:
1
2 • those comparing groups who take part in specific sports or exercise
3 activities with similar groups who are not involved;
4 • those comparing groups who are fit or low in body fat with those who
45111 are unfit or overweight; and
1111 • those involving larger population samples where level of leisure
2 activity has been related through correlation analyses to aspects of
3 well-being including self-esteem.
4
5 In many studies, measures were composite self-esteem scales or assess-
6 ments of body cathexis or body image. More recent research has used
7 perception profiles that have provided a richer documentation of relation-
8 ships. The following general conclusions can be drawn from this literature:
9
• Taking part in regular sport or exercise is moderately associated with
1011
more positive physical self-perceptions, including body image, from
1
late adolescence onwards (e.g. Fox & Corbin, 1989; Sonstroem,
2
Speliotis, & Fava, 1992).
3111
• Being fit and slim are weakly associated with positive physical self-
4
perceptions, body image and in some populations body satisfaction
5
(e.g. Balogun, 1987; Fox, Page, Armstrong, & Kirby, 1994; Tucker,
6
1987).
7
• Sport and exercise participation are weakly associated with global self-
8
esteem in many studies but this relationship is inconsistent and is
9
probably dependent on population, environmental, and individual
20111
characteristics.
1
2 There are several anomalies to these patterns, particularly among
3 females who exercise heavily (Davis, 1997; Sonstroem, 1997a, 1997b), and
4 athletes involved in activities where maintenance of low weight or a slim
5 body is required for elite performance. Here, the benefits of activity appear
6 not quite so apparent. Participation of this kind may heighten awareness,
7 body centrality, and self-criticism. Also, some females seem susceptible
8 to a ‘shifting goal posts’ phenomenon whereby their body satisfaction,
9 body acceptance and self-esteem does not improve when they exercise,
30111 even though they acknowledge some success with weight loss or improved
1 fitness. This demonstrates that confounding and mediating factors are often
2 present to weaken self-esteem-exercise associations.
3 For the vast majority of the public who are the likely targets of exer-
4 cise-based interventions in health care, the positive relationships seem to
5 hold firm, particularly with increasing age. Those who are involved in
6 sport or exercise generally have a higher level of physical self-perceptions,
7 including physical self-worth and body image and there is a tendency for
8 them also to have higher self-esteem than their age-group peers.
9 However, cross-sectional research tells us little about causality. Although
40111 participation in sport and exercise is associated with a higher degree of
1 well-being, it is impossible to determine the degree to which positive self-
2 perceptions are the determinants or outcomes of sport and physical activity
3 participation. There is likely to be a high degree of simultaneous processing
4 as self-perception benefits are experienced and this increases motivation
45111 to participate. It is also likely that associations are strengthened by previous
98 Kenneth Fox
1111 drop-out from sport and exercise of those who have experienced failure,
2 embarrassment and whose physical self-perceptions and self-esteem have
3 been under threat. Also those suffering mental disorders such as depres-
4 sion are more likely to avoid structured physical activity.
5
6
Intervention research
7
8 In comparison to other aspects of exercise and mental well-being, surpris-
9 ingly few reviews of exercise and self-esteem research have been published.
1011 Sonstroem conducted a narrative review in 1984 and reported the find-
1 ings of 16 intervention studies. He concluded ‘Exercise programs are
2 associated with significant increases in self-esteem scores of participants’
3111 (p. 138). However, only ten studies had control groups, only four were
4 randomised, and half the studies had 20 or fewer subjects in the experi-
5 mental treatment. Nine studies employed physical self as well as
6 self-esteem measures. Sonstroem went on to state, ‘At this time it is not
7 known why or in what manner exercise programs affect self-esteem, or
8 which people are responsive’ (p. 150).
9 Although several reviews have included sections on self-concept, and
20111 Gruber (1985) conducted a meta-analysis of studies with children, the only
1 other comprehensive review has been Leith (1994). He reported 16 exper-
2 imental, 21 quasi-experimental, and 10 pre-experimental studies and
3 concluded ‘Approximately one-half of the studies reviewed reported signif-
4 icant changes in self-concept/self-esteem following participation in an
5 exercise programme. These results appear quite inconsistent’. However,
6 he also went on to make several interesting summary statements and
7 observations, several of which are taken on board in the remaining section
8 of this paper.
9 In the context of evidence-based health initiatives, this paper prioritises
30111 randomised controlled studies (RCSs) and these are summarised in Table
1 5.1. RCSs are particularly important given the special difficulties encoun-
2 tered in self-esteem research which include effects from socially desirable
3 responding, expectancy, self-presentation strategies, pleasing the leader/
4 researcher, and temporary versus lasting effects. In addition, a further 44
5 non-randomised controlled studies were considered.
6
7
General conclusions
8
9 Table 5.1 shows that only 36 RCSs were identified in the literature since
40111 1970, and these included nine unpublished masters and doctoral disserta-
1 tions. This represents little more than a study per year and contrasts vividly
2 with the many hundreds of studies in some areas of health services
3 research. These investigations involved a wide range of populations, exer-
4 cise modes and instrumentation, making comparisons and generalisations
45111 difficult and in some instances meaningless.
1111 Of the 36 studies, 28 (78%) indicated positive changes in some aspects
2 of physical self-esteem or self-concept. This is a robust and significant
3 finding that gives clear evidence that exercise helps people see themselves
4 more positively. The results appear stronger for aspects of the physical
5 self (particularly aspects of body image). This is important as they are
6 consistently related to global self-esteem throughout the lifespan. Further-
7 more, an important recent study has indicated that physical self-worth
8 (from the PSPP) which is the global summary of all perceptions in the
9 physical domain carries important emotional adjustment qualities. This
1011 has been established independently of self-esteem and socially desirable
1 responding on questionnaires (Sonstroem & Potts, 1996). This suggests
2 that physical self-worth and related constructs should be regarded as key
3111 mental health indicators in their own right and should be assessed system-
4 atically in interventions. As physical self-perception profiles featuring
5 global physical self-worth subscales have only appeared in the last 10
6 years, a critical construct has not been assessed in most of the studies.
7 Where global self-esteem was assessed, there were mixed findings with
8 about half the studies showing generalised improvement. This is similar
9 to the conclusions of Leith (1994) and also Berger and McInman (1993)
20111 who found that 44% of reviewed studies indicated positive change. It is
1 also supported by conclusions from a recent meta-analysis by Spence and
2 Poon (1997) that has yet to be published as a full paper. They conclude
3 that a small (0.22) but significant effect size emerges for the effect of exer-
4 cise on self-concept or self-esteem. The inconsistency or weakness in
5 findings across studies may be partly due to differences in instrumentation.
6 Where the Rosenberg scale was used, which is one of the better validated
7 scales, significant change beyond controls was rarely reported. Significant
8 improvements were more likely to be recorded with TSCS (which does
9 not contain a true global self-esteem scale) or the summed-item composite
30111 self-concept measures which have been highly criticised.
1 Certainly the evidence suggests that increases in self-esteem (a) do not
2 automatically arise through exercise involvement, and (b) may not always
3 accompany positive changes in physical self-perceptions. However, this is
4 entirely in line with theoretical projections which suggest that the self-
5 esteem construct is the stable outcome from a wide array of life events.
6 Exercise would have to be a particularly powerful experience to instigate
7 a group change in a matter of a few weeks (especially if sense of mastery
8 is the key mechanism), although it may occur with particularly receptive
9 individuals.
40111 Another explanation for the discrepant findings with self-esteem is the
1 likely interaction between the nature of the population studied and
2 the type and setting of exercise. What works for some may not be effec-
3 tive for others and mediating or confounding factors have rarely been
4 assessed in studies.
45111
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Table 5.1 Randomised control trials addressing the effect of exercise on self-esteem and physical self-perceptions
Author(s) Date Subjects Groups Treatment Instruments Results
(i) Children
McGowan et al. 1974 37 grade 7 boys with 1. endurance training 18 weeks 3–4 per week TSCS Greater increase
low self-esteem and team sport (with in exercise group
winning enhanced)
2. Control
Neala 1977 60 grade 9 boys 1. CV fitness 10 weeks CSEI No change
(15 per group) 2. Counselling (goals) Exercise program
3. CV fitness + counselling not outlined
4. Control
Martinek et al. 1978 344 boys and girls 1. Motor activities 45 minutes 1 per week MZ Increase in the
from grades 1–5 and gymnastics treatment group
2. Control
Percy et al. 1981 30 grades 5 and 6 1. Running 7 weeks 3 per week CSEI Marked increase
girls and boys 2. Control in treatment
group
Schempp et al. 1983 208 boys and girls 1. Shared decision Movement ed./gym MZ Greater increases
from grades 1–5 2. Teacher dominated 8 weeks in both treatment
3. Control 1 45 mins per week groups.
Effect size 0.59
Smitha 1984 49 boys and girls 1. Running 10 weeks 3 per week PHSCS No significant
grades 4 and 5 2. Yoga change
3. Control (PE class)
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Marsh & Peart 1988 137 grade 8 girls 1. Competitive exercise 6 weeks 14 sessions of SDQ II Increases in
2. Cooperative exercise 35 mins aerobic exercise physical self in
3. Volleyball control cooperative
group and
decline in
competitive
group.
No change in
global self.
Calfas & Cooper 1996 44 adolescent girls 1. CV exercise + 5 weeks at summer school SPPA Athletic
(Abs.) computer classes 10 hours CV exercise/week competence
2. Computer classes increased in
Group 1 and
self-esteem in
both groups
(ii) Adults
Johnstona 1970 73 male college 1. Physical condition Not stated Q-Sort No change in
students of low– 2. Sport skills technique self-concept
average fitness 3. Control or movement
concept
Davisa 1971 39 male college 1. CV exercise I Not stated TSCS No change
students 2. CV exercise II
3. Control
Whitea 1974 152 college students 1. Individualised circuit 10 weeks (frequency TSCS Increased self-
enrolled in PE training and time not stated) esteem and
classes 2. Normal PE classes subscales beyond
controls
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Hilyer & 1979 120 college students 1. Running & stretching 10 weeks 3 per week TSCS Increased self-
Mitchell 2. Same + counselling concept in
3. Controls Group 2
Trujillo 1983 35 female college 1. Weight training 16 weeks typical college TSCS Groups 1 & 2
students 2. Running activity classes? increased self-
3. Active controls concept. Group
1 improved
significantly
greater than
active control
Brown & 1986 85 mature and 1. Weight training 12 weeks 3 per week TSCS Increases in
Harrison young women 2. Inactive control for 60 minutes physical self-
concept and self-
satisfaction
for both young
and mature
groups
Ben Shlomo 1986 sedentary females 1. Arm training 6 weeks 3 per week TSCS No change in
& Short 2. Leg training of leg or arm ergometry self-concept,
3. Control at 60–80% max HR BCS physical self-
concept or
body cathexis
O’Neilla 1989 53 non-athlete 1. Aerobic exercise 4 weeks (frequency & TSCS Physical self-
female college sessions time not stated) concept
students 2. Aerobics lectures increased in
Group 1.
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Cocklina 1989 69 women aged 1. Exercise (mod. int.) 8 weeks 3 per week IAV Self-acceptance
20–50 2. Exercise (light int.) 1. To improve CV fitness increase in both
3. Inactive controls 2. As placebo BCS exercise groups
but body
cathexis
unrelated
Cusumano & 1992 95 female college 1. Hatha yoga 3 weeks 380 mins RSES PSES decreased
Robinson students 2. Progressive relaxation session PSES while self-esteem
increased in
both groups. No
group differences
Desharnais 1993 male and female Running RSES Positive changes

et al. adults
Brown et al. 1995 135 middle-aged 1. Mod. int. walking 16 weeks 3 per week RSES Positive changes
sedentary men and 2. Low int. walking (must have completed at SPES in exercise
women 3. low int walking least 42 sessions). All BCS groups for body
+ relaxation response sessions at indoor facility Success cathexis and
4. Group t’ai chi expectancies physical
5. Control competence,
time effect for
self-esteem but
not group
differences
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
King et al. 1989 120 middle-aged 1. Home-based exercise 6 months of walk/jog at Perceptions Increase in
mainly sedentary 2. Inactive controls 65–77% max HR. of satisfaction satisfaction
men and women 5 sessions per week with levels and health
prescribed appearance, ratings in
fitness, exercise group
weight and
ratings of
health
behaviours
King et al. 1993 357 50–65 year-old 1. High int. ex. group 12 months programme Self-perception Higher rating of
sedentary men and 2. High int. home ex. High int. – 3 40 mins of change in change in all 3
women 3. Low int. home ex. per week at 73–83% health, exercise groups
4. Inactive controls max. appearance, than controls
Low int. – 5 30 mins fitness, and
weight
Tucker & 1993 60 early middle-aged 1. Walking 12 weeks 3 per week BCS Improvements in
Mortell women 2. Weight training home-based weights ratings of both groups but
or walking fitness greater in weight
and fitness weight training
improvement
Alferman & 1995 Sedentary 1. Exercise 6 months 1–2 sessions Not stated Increase in
Stoll middle-aged 2. Waiting list control per week for 60 mins physical concept
adults but not self-
esteem
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Alferman & 1995 Sedentary 1. Fitness 6 months 1–2 sessions Not stated All groups
Stoll middle-aged 2. Jogging per week for 60 mins increase in
adults 3. Relaxation physical
4. Back exercises self-concept and
self-esteem but
exercise groups
significantly
greater
improvement in
physical self-
concept
Palmer 1995 27, nonclinical, 1. Supervised walking 8 weeks building from RSES Increase in self-
premenopausal 2. Non-walking control 20 mins per session. esteem in
women Frequency not stated walking group
only
Talbot & Taylor 1998 142 middle-aged 1. Exercise prescription CV exercise + weights PSPP Increase in
(Abs.) patients with risk scheme (shortened) physical self-
of CHD 2. Non-referred controls 10 weeks 2 per week PIP worth, condition,
appearance, and
health
(iii) Special populations
Collingwood 1972 50 male adult 1. Physical training 4 weeks, 5 per week Body attitude Improvements in
rehabilitation 2. Control for 60 mins of general scales body attitude,
patients fitness work IAV self-concept and
self-acceptance
in exercise group
Whitinga 1981 80 alcoholics 1. Therapy Exercise 5 days per week TSCS Increases in both
2. Therapy + exercise incl. 2 miles walk, gym groups but no
activity and swim additional effect
due to exercise
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Hilyer et al. 1982 60 males, 15–18 1. Fitness group + Strength, CV exercise Self-esteem Increase
years, 55% black counsellor support and flexibility Inv (but impact of
youth offenders 2. Team sports 20 weeks 3 90 mins Form A counselling not
known)
Hannaforda 1984 25 depressed males 1. Exercise Jogging 3 days per week RSES No increase in
2. Corrective therapy 30 mins (length not self-esteem even
3. Waiting list controls stated) with increase in
CV fitness and
reduced
depression
Short et al. 1984 45 overweight/obese 1. Instruction + 8 weeks of 90 mins TSCS Greater
policemen 29–52 conditioning lifestyle instruction and increases
years old 2. Instruction alone 3 45 mins walk/jog. (2–3 times) in
exercise group
in physical self,
personal self
and self-
satisfiction
MacMahon & 1988 54 learning disabled 1. High int. exercise and 20 weeks PH with Increase in both
Gross boys sports 1. HR >160. assistance groups with high
2. Low int. exercise and 2. Intermittent with intensity greatest.
sports HR < 160
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Ossip-Klein 1989 32 depressed female 1. Running 8 weeks 3–4 per week Beck Increase in self-
et al. adults 2. Weights 20 mins running or Self-Concept concept,
3. Delayed control weights + warmup & Test perceptions of
cool down Osgood’s energy and
Semantic fitness in both
Differential exercise groups
Mactavish & 1992 26 middle-aged 1. Physical activity 5 weeks of subject Perceived Greater increases
Searle males 2. No activity selected activity and Leisure in perceived
and females with sports Competence competence and
mental retardation Scale self-esteem in
RSES treatment group
Donaghy & 1998 117 men and 1. Group and home 3 weeks formal exercise PSPP Change in
Mutrie women in alcohol exercise followed by 12 weeks physical self-
rehabilitation 2. Control home-based activity worth, perceived
condition and
strength at 4
weeks month and
condition and
strength at 8
weeks in exercise
group. NS at 5
months
Notes:
a
= PhD Abstract only.
BCS (Body Cathexis Scale: Secord & Jourard, 1953) CSEI (Coopersmith Self-Esteem Inventory: Coopersmith, 1967) IAV (Index of Adjustment and
Values: Bills, Vance, & McLean, 1951) MZ (Martinek-Zaichkowsky Self-Concept Scale for Children: Martinek & Zaichowsky, 1977) PHSCS (Piers-Harris
Self-Concept Scale for Children: Piers, 1984) PSES (Physical Self-Efficacy Scale: Ryckman, Robbins, Thornton & Cantrell, 1982) PSPP (Physical Self-
Perception Profile: Fox, 1990; Fox & Corbin, 1989) PIP (Perceived Importance Profile: Fox, 1990) PSDQ (Physical Self-Description Questionnaire: Marsh
et al, 1994) RSES (Rosenberg Self-Esteem Scale: Rosenberg, 1965) SDQ II (Self-Description Questionnaire for Adolescents: Marsh, 1992a) SPES
(Sonstroem’s Physical Estimation Scale: Sonstroem, 1978) SPPA (Self-Perception Profile for Adolescents: Harter, 1988) TSCS (Tennessee Self-Concept
Scale: Fitts, 1965)
108 Kenneth Fox
1111 Which populations can benefit?
2
3 Children and adolescents
4
5 Gruber (1986) conducted a meta analysis of studies with children and in
6 1994, Calfas and Taylor reviewed the impact of physical activity on the
7 psychological well-being of adolescents, and included a section on self-
8 concept. Gruber concluded that the effect of activity programmes was
9 positive, particularly for those already low in self-esteem. Physical fitness
1011 and aerobics programmes produced superior results to motor skill and
1 sport programmes. There is also some evidence (Marsh & Peart, 1988;
2 Schempp, Cheffers, & Zaichowsky, 1983) that cooperative and more demo-
3111 cratic exercise settings produce stronger effects and this is supported by
4 recent literature on motivational climate in sport, exercise and physical
5 education. Calfas and Taylor (1994) when comparing the effects on the
6 range of mental benefits found that the strongest changes were for self-
7 esteem, self-concept or self-efficacy with nine out of ten studies revealing
8 positive results. Of the eight RCSs conducted with children in Table 5.1
9 (which include two reported by Calfas & Taylor) five report self-concept
20111 or self-esteem changes and a sixth found changes in physical self-concept
1 accompanying exercise. One trial (Hilyer et al., 1982) reported self-esteem
2 improvements with youth offenders when exercise was combined with
3 counsellor support. Although exercise/self-esteem studies have been
4 conducted with obese children, it is not possible to single out the effects
5 of exercise from weight loss (French, Story, & Perry, 1995)
6 The evidence is sufficient to conclude that exercise is an effective
7 medium for developing a positive self in children, is particularly effective
8 for those with low self-esteem, and has greatest potential when presented
9 in a style that will encourage mastery and self-development. It must also
30111 be kept in mind that school-based programmes have potential to lower
1 self-esteem, as youngsters are not in the same position as adults to drop
2 out if experiences are negative.
3
4 Young adults
5
6 Seven RCTs were located with young adults who were mainly US college
7 males and females. Six of the studies used the Tennessee Self-Concept
8 Scale. Six showed positive change with one restricted to change in the
9 physical self and another in the group where exercise was combined with
40111 exercise counselling (Hilyer & Mitchell, 1979). A range of exercise modes
1 was used with the majority being aerobics or running, circuit training,
2 and weight training. A further 20 controlled studies with intact groups
3 were identified, 13 being unpublished theses or dissertations. Fifteen of
4 these studies involved students across a range of college activity classes,
45111 with the majority focusing on aerobic dance, cardiovascular fitness and
1111 weight training. Most studies used the TSCS and the Body Cathexis Scale.
2 Half the studies reported non-significant results and some of the remaining
3 studies showed only weak gains.
4 This literature is biased towards US college students, the majority
5 already having average to high self-esteem and being involved in regular
6 physical activity. It is unlikely therefore that young adults as a group will
7 experience the greatest mental benefits from exercise. They are also a low
8 health risk population and therefore unlikely to receive priority attention
9 for mental health service provision.
1011
1
Middle-aged adults
2
3111 Middle-age appears a particularly crucial time for exercise interventions
4 as the population becomes less active, increases in weight and symptoms
5 of ageing become more apparent and yet there seems to remain some
6 potential for lifestyle change. This is a population therefore that is of
7 particular interest to health service providers. Seven of the best designed
8 and most recent RCSs were with large groups of previously sedentary
9 populations in this age range (Alferman & Stoll, 1995 [two studies]; Brown
20111 et al., 1995; King, Taylor, & Haskell, 1993; King, Taylor, Haskell, &
1 DeBusk, 1989; Talbot & Taylor, 1998; Tucker & Mortell, 1993). Studies
2 ranged in length from 10 weeks to a year with four studies lasting at least
3 6 months. Programmes included moderate versus low intensity walking,
4 home-based versus group exercise, walking versus weight training, and
5 fitness and jogging versus relaxation. All studies indicated positive
6 improvement in physical self-perceptions. These included ratings of fitness,
7 appearance, physical health, body cathexis, and physical self-worth.
8 However, only three studies assessed global self-esteem and this did not
9 indicate significant improvement. Even in the absence of global self-esteem
30111 change, this produces a robust picture of psychological improvement in
1 these groups and warrants further consideration for health service invest-
2 ment. These studies are also backed up by a growing literature on
3 self-efficacy and physique anxiety in middle-aged individuals which reveals
4 potential exercise benefits (McAuley, Courneya, & Lettunich, 1991;
5 McAuley, Mihalko, & Bane, 1995). It is well established that this popu-
6 lation can benefit a great deal in reduced incidence of morbidity and
7 mortality through heart disease, obesity, diabetes and some cancers.
8 Improvements in mental well-being may increase motivation so that the
9 full array of benefits is experienced.
40111
1
Older adults
2
3 No RCSs were located for elderly adults. This is surprising given the poten-
4 tial of exercise to increase functionality, independence, and life quality.
45111 Studies have tended to address other constructs such as life satisfaction,
110 Kenneth Fox
1111 subjective well-being and quality of life. Four unpublished controlled stud-
2 ies involving exercising groups were identified. Flexibility programmes
3 were ineffective (Bozoian & McAuley, 1994; Yeagle, 1982) but strength and
4 fitness programmes indicated positive self-concept changes (Bozoian &
5 McAuley, 1994; Olfman, 1987). Recently, Mutrie and Davison (1994)
6 recruited 83 adults with a mean age of 61 years who self-selected into
7 either a home-based or class-based exercise programme lasting 3 months.
8 Measures using the PSPP were applied pre- and post-programme and
9 6 months after the programme. Positive changes in physical self worth were
1011 found for both groups and these were greatest for the class-based exer-
1 cisers. Further trials with this population are required that combine self-
2 efficacy, physical self-perception and self-esteem measures.
3111
4
Special populations
5
6 Several groups fall into this category including those with mental disor-
7 ders, those who are in rehabilitation from substance abuse, ill-health or
8 injury, those who have physically disability, and those who are obese. In
9 reviewing the effect of exercise on self-esteem of special groups, Leith
20111 (1994) pointed out that only 3 of 13 studies did not produce significant
1 change. This was attributed in part to these groups initially being low in
2 self-esteem. However, there are few well-designed studies with special
3 groups, possibly because of recruitment and randomisation difficulties.
4 There are RCSs to support change due to exercise in adults with mental
5 retardation (Mactavish & Searle, 1992), depressed females (Ossip-Klein
6 et al., 1989), youth offenders (Hilyer et al., 1982), obese males (Short,
7 DiCarlo, Steffee, & Pavlou, 1984), male rehabilitation patients (Colling-
8 wood, 1972) and problem drinkers (Donaghy & Mutrie, 1998). There is
9 a body of literature on disability and sport participation indicating that
30111 there is great potential for involvement to improve mental well-being in
1 people with disabilities (Sherrill, 1997). However, as yet there are no well-
2 controlled studies with this population. Clearly, the potential benefits for
3 a range of special populations is high and much further research is required.
4
5
Gender differences
6
7 There is evidence that exercise has a beneficial effect for males and
8 females. There may be greatest potential for females as they consistently
9 score lower initially on self-confidence in physical activity and also body
40111 image, physical self-worth and self-esteem (Lirgg, 1991).
1
2
3
4
45111
1111 What are the characteristics of effective exercise?
2
3 Type of activity
4
The effect of a wide range of physical activities and sports has been inves-
5
tigated. Various forms of cardiovascular exercise including running,
6
walking, aerobic dance, and circuit training are most common in studies.
7
All of these activities have indicated that they can be effective in improving
8
self-perceptions, although there is a reflection of the general finding that
9
only approximately 50% produce significant change. Studies with other
1011
activities such as swimming, flexibility training, martial arts, and expressive
1
dance have generally failed to indicate significant change, however they
2
are too few in number to make firm conclusions. Many of these studies
3111
have been conducted on college-age males and females, a population prone
4
to non-significant results. Where endurance exercise and walking has been
5
used with middle-aged adults, there is more conclusive evidence of success
6
than with younger adults.
7
Weight training has attracted increasing attention and ten studies
8
including two RCSs have recorded improvements in body image and other
9
physical self-perceptions in men and both young and middle-aged women.
20111
Some RCSs have made comparisons between resistance exercise and other
1
activities including running (Trujillo, 1983), walking (Tucker & Mortell,
2
1993) and also swimming, aerobics, and PE classes. There is some indi-
3
cation that resistance training is superior to endurance exercise in
4
improving body image and physical self-esteem.
5
6
7 Exercise frequency, intensity and duration
8
There is insufficient variance in the studies to assess the impact of frequency
9
of exercise with the vast majority of programmes opting for three sessions
30111
per week. This is in contrast to current recommendations for health-related
1
activity which is five or six occasions of moderate intensity activity per
2
week. Intensity is rarely reported and only two RCSs compared low with
3
high intensity exercise. King et al. (1993) found that both treatments were
4
effective in stimulating psychological improvement in adults with no dif-
5
ferences in degree of change. MacMahon and Gross (1988) found that
6
higher intensity sports were more successful with learning disabled boys.
7
Most studies report the length of the exercise session and programmes that
8
last longer than 60 minutes are more likely to produce positive change. This
9
may reflect an increasing level of commitment in the participants.
40111
1
2 Programme duration
3
Programmes have varied in length from a single session to 12 months. Leith
4
(1994) recently divided studies into those lasting 8 weeks or less, 9 to 12
45111
112 Kenneth Fox
1111 weeks, and more than 12 weeks. He concluded that although there was
2 evidence of change in some studies in all of these groups, there was a higher
3 likelihood of self-esteem change in longer programmes. The RCSs reported
4 here support this observation. However, the longer well-controlled studies
5 have not assessed global self-esteem and the time required for lasting
6 change; also, how long change lasts is still not known.
7
8
What are the mechanisms?
9
1011 Although there is sound evidence that exercise can produce positive
1 changes in well-being through improved physical self-perceptions and
2 sometimes self-esteem, the question still remains as to the main mecha-
3111 nisms underpinning such change. For the fine tuning of intervention design,
4 it is important not only for mechanisms to be determined but also for the
5 conditions under which they optimally function to be identified.
6 Returning to possible mechanisms outlined earlier in this paper:
7
8 1 An undetermined psychophysiological mechanism. The cross sectional
9 and longitudinal evidence fail to show a consistent relationship between
20111 global self-esteem and exercise participation. This suggests the absence
1 of a generic or generalised psychophysiological or psychobiochemical
2 effect. The variance among studies both for populations and charac-
3 teristics of the exercise setting suggests that mechanisms are more
4 likely to be psychosocial in origin.
5 2 Improvements in fitness or weight loss. There is evidence from several
6 studies that fitness change (as measured by standard laboratory or
7 field tests of fitness) is not necessary for enhanced self-esteem
8 or improved physical self-perceptions (Ben-Shlomo & Short, 1986;
9 King et al. 1989; Ossip-Klein et al., 1989; Palmer, 1995). This paral-
30111 lels the obesity treatment literature where amount of weight lost is
1 not consistently reflected in the psychological benefits (French, Story
2 & Perry, 1995). Perceptions of health, physical competence, fitness
3 and body image may arise simply because there is a feeling that the
4 body is improving through exercise. There is some indication that
5 muscular fitness reflected in improved tone or strength can have a
6 more rapid and powerful sensory effect than cardiovascular or flexi-
7 bility change.
8 3 Autonomy and personal control. There is no direct evidence to estab-
9 lish sense of control over the body, its appearance and functioning as
40111 the main route to self-esteem change. Instrumentation has not been
1 systematically used to test this hypothesis. Furthermore, because
2 autonomy is tied to identity change, it is unlikely that many studies
3 have been conducted for long enough for any effect to be adequately
4 documented. However, the cross-sectional evidence that changes in
45111 self-efficacy for exercise is associated with adherence in middle-aged
1111 and older people (see McAuley et al., 1995) suggests a promising line
2 for intervention research.
3 4 Sense of belonging and significance. Group or individual social support
4 can produce mental benefits and this may be possible in the exercise
5 setting. However, the effect of exercising regularly in a group or regular
6 contact with an exercise counsellor on social well-being has not been
7 adequately tested. Two studies have included some form of counselling
8 with the exercise programme and reported positive effects. To date, there
9 is insufficient evidence to show that group exercise produces greater
1011 improvements in self-esteem or self-perceptions than home-based or
1 individual exercise. It is likely to vary with the individual, and the popu-
2 lation, and there may be gender differences in preference. Leisure-centre
3111 based exercise prescription schemes, for example, attract mainly middle-
4 aged females who report social benefits (Fox, Biddle, Edmunds, Bowler,
5 & Killoran, 1997). Exercise groups for the older middle-aged and elderly
6 held in community and leisure centre settings seem to rely on social inter-
7 action as a key component to successful attendance patterns.
8
9 In summary, we still do not know what it is about exercise that helps
20111 people feel better about themselves. It is likely that there are several mech-
1 anisms operating, some tied to improvements in the body, others linked to
2 social significance and the exercise setting. Firmer conclusions cannot be
3 drawn at this time. The greater use of multidimensional self-perception
4 instruments that are capable of more comprehensive documentation of the
5 nature of change may be more revealing in future studies.
6 Cross-sectional research has already indicated that there may be several
7 mediating variables involved in exercise/self-esteem links. These probably
8 include factors such as the degree of autonomy experienced by the exer-
9 ciser, the centrality or importance of exercise to the individual, and the
30111 nature of the exercise leadership. It is conceivable that some factors are
1 necessary conditions under which self-esteem enhancement can take place.
2 There are also factors attached to the exercise setting that may work
3 against self-esteem development and this is particularly crucial for captive
4 audiences such as schoolchildren.
5 Possibly more than any other element of well-being, self-esteem, because
6 of its essentially subjective nature, is likely to be more consistently explained
7 with a ‘horses for courses’ explanation. It is unlikely that a group mean
8 approach will be sufficiently sensitive to individual differences in response
9 to the many different characteristics of the exercise environment and moti-
40111 vational climate in which it is conducted.
1 It must also be realised that almost all studies report results of those
2 who remain in the programme. Those who choose to drop-out may form
3 an interesting group to study as they may reveal elements of the pro-
4 gramme that are potentially negative in their effect such as increasing
45111 social physique anxiety or initiating feelings of incompetence and failure.
114 Kenneth Fox
1111 Implications for research
2
3 • Given the importance of the self to human functioning and health
4 outlined at the beginning of this document, there has been a pitiful
5 amount of well-designed research conducted, particularly in the form
6 of true experiments (one trial per year). The reasons for this are not
7 clear but this has been an underfunded area that has been seen to
8 have little more than academic appeal and consequently has been
9 ascribed low priority for health services funding.
1011 • Research has largely been conducted by physical educators and sport
1 and exercise scientists whose interests and needs are often quite
2 different from those of health services professionals. This is reflected
3111 in an absence of evidence-based health principles underpinning
4 research. For example, intention to treat statistics are not included in
5 any of the studies reported in this paper and there is little evidence
6 that cost-effectiveness has been considered.
7 • Randomised controlled trials are important but will not tell us all we
8 need to know. Individual responses to the conditions of exercise and
9 exercise settings will vary and it will be necessary to use time series
20111 case studies and a range of qualitative techniques to adequately
1 unravel the mechanisms at work.
2 • Generally, studies have been too short to fully test out the influence
3 of exercise on self-esteem. It is likely that a construct so critical to
4 mental functioning as self-esteem will take some time for lasting
5 change to occur. Unfortunately, where studies have lasted for 6 or 12
6 months, they have not assessed self-esteem (see Sonstroem, 1997a).
7 Longer studies are required which utilise comprehensive self-percep-
8 tion measures and evidence-based health principles.
9 • Studies are required that investigate degree of well-being change
30111 (such as emotional adjustment, reductions in depression, and life satis-
1 faction) alongside self-perception change. As yet, clinical criteria
2 attached to self-esteem or physical self-perception levels have not been
3 developed, so it remains difficult to attach practical significance to
4 self-esteem change scores.
5
6
Implications for practice
7
8 • Greatest self-perception/self-esteem improvements are likely to
9 occur in those groups who have the most to gain physically from
40111 exercise participation. This includes those who are in poor physical
1 condition such as the middle-aged, the elderly and the overweight and
2 obese.
3 • Greatest improvements are also likely to occur in those who are
4 initially low in self-confidence, self-esteem, physical self-worth, and
45111 body image, including women in general, those with mild depression,
1111 physically disabled children and adults, overweight and obese adults
2 and children, and perhaps offenders.
3 • Currently there is greatest support for the effectiveness of cardiovas-
4 cular exercise and weight training programmes.
5 • Not enough is known about the effectiveness of specific exercise char-
6 acteristics but it seems wise to focus on exercise that is moderately
7 demanding for the population, with sessions optimally lasting in the
8 region of 60 minutes.
9 • Programmes should last at least 12 weeks with some form of contact
1011 continuing for 6 months or more. Limited evidence presented here
1 suggests that global changes in self-esteem and identity are more likely
2 given longer intervention.
3111 • Adherence factors cannot be separated from those which promote
4 self-esteem. On the one hand, the programme cannot be effective
5 without participation and on the other, mental benefits are associated
6 with sustained adherence. In this sense, conditions which affect the
7 attractiveness of the exercise programme, such as the qualities of the
8 leader or the exercise setting, may be critical to changes in self-esteem.
9
20111
Final comments
1
2 Exercise and sports participation are associated with more positive self-
3 perceptions but this does not allow us to determine whether participation
4 causes enhanced well-being or helps prevent mental disorders and ill-
5 health. The evidence from intervention studies shows clearly that exercise
6 helps people feel better about themselves and this contributes to their
7 mental well-being and presumably their quality of life. This in itself
8 suggests that health professionals should consider physical activity as an
9 important element of health promotion. In addition, improved self-esteem
30111 is an important marker of recovery from clinical symptoms of depression
1 and anxiety and should be systematically assessed. Similarly, physical self-
2 worth has been shown to be independently associated with elements of
3 well-being and should provide an important benchmark for success.
4 Finally, self-esteem and physical self-perceptions are inextricably linked
5 to motivation through choice and persistence in health behaviours,
6 including exercise. For this reason alone, it has to be given serious consid-
7 eration in any intervention. A problem facing recognition of the
8 importance of self-esteem is that it is often seen by health professionals
9 as an outcome rather than a cause of either well-being or ill-health.
40111 Its centrality to human functioning demands that it be given serious
1 consideration as a determinant.
2
3
4
45111
116 Kenneth Fox
1111
2 What we know
3
4 • Exercise can be used as a medium to promote physical self-
5 worth and other important physical self-perceptions such as body
6 image. In some situations, this improvement is accompanied by
7 improved self-esteem.
8 • Physical self-worth carries mental well-being properties in its own
9 right and should be considered as a valuable end-point of exer-
1011 cise programmes.
1 • Positive effects can be experienced by all age groups but there
2 is greater evidence of change in children and middle-aged adults.
3111 • These effects can be experienced by men and women.
4 • Effects are likely to be greater for those with low self-esteem
5 but these individuals may be difficult to attract into programmes.
6 • Several types of exercise are effective in changing self-
7 perceptions but there is most evidence to support aerobic exer-
8 cise and weight training, with weight training indicating greatest
9 effectiveness in the short term.
20111
1
2
3
5
• The degree to which self-perception and self-esteem change is
6
accompanied by reductions in clinical symptoms, indicators of
7
emotional adjustment and general well-being.
8
• More about the mechanisms of change.
9
• More about the optimal conditions under which mechanisms
30111
might operate.
1
• More about which populations are responsive to which mecha-
2
nisms.
3
• More about which individual characteristics increase respon-
4
siveness to mechanisms of change.
5
• More about some populations that might particularly benefit from
6
exercise including the elderly, the obese, those with mental disor-
7
ders, and those with physical disability.
8
• More about the dynamics of change. Little is known about how
9
long it takes to produce changes, and how long they last.
40111
• More about the conditions under which improvements in self-
1
esteem and self-perceptions are inhibited.
2
• More about those who do not volunteer for studies or who drop
3
out and do not feature in the results.
4
45111
1111
2 • How much change in self-esteem scores is necessary for a
3 meaningful impact on functioning, behaviours and well-being. To
4 date, insufficient evidence has become available to develop clin-
5 ical criteria and targets of change.
6
7
8
9 Acknowledgements
1011 Appreciation is extended to Emma Stratton M.Sc. and Dave Carless M.Sc.
1 for their assistance in preparing this review.
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
4
45111
1111
2 6 Cognitive performance,
3
4 fitness, and ageing
5
6 Stephen H. Boutcher
7
8
9
1011
1
2
3111 It is well established that a decrease in cognitive performance is an
4 inevitable consequence of growing old (Botwinick, 1973; Dustman,
5 Emmerson, & Shearer, 1990). Behavioural slowing is the most significant
6 aspect of the reduced cognitive performance that accompanies ageing.
7 Almost all studies that have examined speed task performance have found
8 older adults to be slower than their younger counterparts (Salthouse, 1985).
9 This increased slowness of response is of major consequence for older
20111 individuals. For example, behavioural slowing affects the ability to drive
1 a car, thus older adults are prone to higher accident rates and, as a result,
2 pay higher insurance premiums. Slower movement speeds also contribute
3 to higher accident rates at home and at work. Also, Hertzog (1989) has
4 demonstrated that behavioural slowing in older adults is the major cause
5 of their decline in intelligence. He has shown that the deterioration in
6 mental abilities, typically found in older adults, may not reflect a loss of
7 thinking but more a slowing of the rate of intelligent thought.
8 If age-related degradation in cognitive performance could be prevented
9 then health care costs and suffering could be significantly reduced and
30111 functional ability and quality of life improved. Physical fitness is one of
1 the few non-pharmacological interventions that could offset the cognitive
2 decline accompanying ageing. However, the relationship between fitness,
3 ageing, and cognitive performance is extremely complex and the method-
4 ological barriers are daunting. Also the mechanisms underlying the effects
5 of fitness on the cognitive performance of older adults are unclear and
6 those that have been proposed have little empirical support.
7 This review draws evidence from a number of narrative reviews, cross-
8 sectional and intervention studies, and two meta-analyses. Papers were
9 located through electronic searches using Sport Discus, Medline, and
40111 PsychLit. The major emphasis of the review is on the effect of aerobic fit-
1 ness on cognitive performance of older adults from non-clinical popula-
2 tions. The ability of physical exercise to improve the cognitive performance
3 of older adults has important health implications and consequently has
4 attracted the greatest research interest. Firstly, the concepts of ageing,
45111 cognitive performance, and physical fitness will be discussed. Then the
Cognitive performance and physical fitness 119
1111 cross-sectional and intervention research results in this area will be sum-
2 marised, after which possible mechanisms underlying the fitness and
3 enhanced cognitive performance of older adults will be outlined.
4
5
Concepts
6
7
Ageing
8
9 Ageing concerns processes in humans that occur with the passing of time.
1011 Chronological age refers to the number of years a person has existed,
1 whereas functional age is the age at which a person functions. Biological age
2 is to do with biological rather than chronological processes. Researchers in
3111 the fitness, ageing, cognitive performance area focus on both chronological
4 and biological ageing as they assume that fit older adults are biologically
5 younger than unfit older adults (Chodzko-Zajko & Moore, 1994). Age
6 categories used in this review are: ‘young-old’ between 65–74 years, ‘old’
7 between 75–84 years, and ‘old-old’ between 85–99 years. Gerontologists
8 also disassociate the ageing process (Busse, 1969) that occurs in the absence
9 of disease (primary ageing) from physiological and mental decline caused by
20111 poor health (secondary ageing).
1
2
Cognitive performance
3
4 Spirduso (1994) suggests that cognition can be viewed as functions of the
5 brain that include memory, association, abstract reasoning, and spatial
6 ability. She further suggests that attention, information-processing speed,
7 and perception are some of the processes of cognition that support the
8 cognitive functions. These processes interact and allow humans to process
9 information and make decisions. Tests used to assess these different cogni-
30111 tive functions and processes are extensive. One of the most used tests in
1 the fitness and cognitive research area is the reaction time (RT) task that
2 mainly assesses attention and speed of response. In this task participants
3 attend to a stimulus, usually a light, and depress a button as quickly as
4 possible when the stimulus light goes off or the colour of the light changes.
5 Examples of other tests that have been used to assess cognitive perfor-
6 mance in ageing are listed in Table 6.1.
7
8
Physical fitness
9
40111 Fitness is a multifactorial concept that includes flexibility, strength,
1 stamina, and anaerobic and aerobic fitness. Individuals who are aerobically
· ·
2 fit possess high maximal oxygen uptake (VO2max). Thus, VO2max represents
3 the maximal amount of aerobic work an individual can accomplish. It is
4 represented as millilitres of oxygen consumed per kilogram of body mass
45111 per minute of exercise. The great majority of studies in this area have
1111 Table 6.1 Examples of cognitive tests used in fitness, cognitive performance and
2 ageing research
3
Test Aspect of cognition
4 that test measures
5
6 Reaction time Speed and attention
Choice reaction time Speed and attention
7
Wechsler Adult Intelligence Scale (WAIS) Intelligence
8 Recall and recognition Memory
9 Dual-time attention task Attention
1011 Culture Fair Intelligence Scale Fluid intelligence
1 Digit Symbol WAIS subtest Fluid intelligence
Dots Estimation Memory
2
Stroop Colour Test Attention
3111 Line matching tests Memory
4 Verbal comparison tests Verbal fluency
5 Stanford-Binet Intelligence Quotient Intelligence
6 Wechsler Memory Scale Memory
Sternberg Number task Memory
7
Raven’s Progressive Matrices Test Memory
8
9
20111
1 examined the relationship between aerobic fitness and cognitive perfor-
2 mance. Aerobic fitness has been assessed directly and indirectly by
·
3 VO2max, sub maximal tests, field tests, resting cardiovascular parameters,
4 and self-report of physical activity levels. This review will primarily focus
5 on the relationship between chronic aerobic exercise (e.g. regular walking,
6 jogging, swimming), ageing, and cognitive performance.
7
8
Ageing and cognition
9
30111 As people age not all aspects of cognitive performance decline. Chodzko-
1 Zajko and Moore (1994) have suggested that cognitive decline with ageing
2 is more likely to occur in processes that require attention. Using the
3 concepts of Hasher and Zacks (1988) and others (see Chodzko-Zajko &
4 Moore, 1994) they view cognitive processes as being distributed along an
5 automatic-to-effortful processing continuum. Thus, tasks that require
6 effort, such as attention and speed, are likely to be performed less well
7 by older adults. The literature on ageing supports this view as behavioural
8 slowing is a typical characteristic of advancing age (Dustman et al., 1990;
9 Salthouse, 1985; Stelmach, 1994). Although the mechanisms underlying
40111 these slowing of behaviours have not been determined, the prevalent view
1 is that reduced speed of performance occurs because of central processing
2 limitations rather than peripheral factors (Chodzko-Zajko & Moore, 1994).
3 Another factor that influences ageing and cognitive performance is
4 health status or secondary ageing. As people age the incidence of disease
45111 increases and disease has been linked to cognitive decline (Milligan,
1111 Powell, Harley, & Furchtgott, 1984). Because people who exercise gener-
2 ally possess less disease (e.g., cardiovascular disease, adult onset diabetes)
3 fitness may positively influence cognitive performance by keeping people
4 healthy. Thus, the effect of fitness on both primary and secondary ageing
5 are important issues for this research area.
6
7
The effects of fitness on cognitive performance of older
8
adults
9
1011
Cross-sectional studies
1
2 Researchers using cross-sectional designs have typically been interested
3111 in comparing the cognitive response of physically fit to unfit older adults.
4 They have usually compared the cognitive performance of groups of older
5 athletes or adults engaged in physical activity with older adults who are
6 inactive (Hart, 1981; Sherwood & Selder, 1979; Spirduso, 1975; Spirduso
7 & Clifford, 1978). Other cross-sectional studies conducted into the effects
8 of fitness on cognitive performance of older individuals include Abourez
9 and Toole (1995), Arito and Oguri (1990), Baylor and Spirduso (1988),
20111 Clarkson-Smith and Hartley (1989), Del Rey (1982), Dustman, et al.
1 (1990), Era (1988), Hart and Shay (1964), Hoyer, Labouvie and Baltes
2 (1973), Ismail and El-Naggar (1981), Milligan et al. (1984), Molloy,
3 Beerschoten, Borrie, Crilly and Cape (1988), Offenbach, Chodko-Zajko
4 and Ringel (1990), Perlmutter and Nyquist (1990), Rikli and Busch (1986),
5 Roberts (1990), Stacey, Kozma and Stones (1985), Stones and Kozma
6 (1989), Suominen, Heikkinen, Parkatti, Forsberg and Kiiskinen (1980) and
7 Szafran (1966).
8 The first study to use this approach was conducted by Spirduso (1975).
9 Both young (20 year olds) and older males (above 60 years) who regularly
30111 exercised (racquetball) for three years were compared to young and old
1 non-exercising controls. Findings were positive and reflect the results of the
2 majority of cross-sectional studies conducted in this area. Older males who
3 regularly exercised possessed significantly quicker RTs than older inactive
4 males. It is significant that the RTs of the exercising old were similar to
5 those of the exercising young who were more than 40 years younger.
6 Other studies using differing exercise modalities such as walking (Arito &
7 Oguri, 1990), jogging (Baylor & Spirduso, 1988), and swimming (Hawkins,
8 Kramer, & Capaldi, 1992) have also found similar results. These effects have
9 also been found using varying methods of fitness assessment. Studies exam-
40111 ining other aspects of cognitive functioning, such as memory and fluid and
1 crystallised intelligence, have generally not found consistent differences
2 between fit and unfit older adults (Chodzko-Zajko & Moore, 1994).
3 The meta-analysis by Thomas, Landers, Salazar and Etnier (1994)
4 indicates that the effect size for cross-sectional studies examining the rela-
45111 tionship between fitness and cognitive performance in older adults was
1111 significant but moderate (effect size [ES] 0.31; standard deviation [SD]
2 0.54). The most reliable effects were found for RT, maths, and acuity
3 tasks. Also the analysis revealed that exercising women (ES 0.47; SD
4 0.59) showed greater levels of cognitive performance than exercising
5 men (ES 0.15; SD 0.68). These large standard deviations reflect the
6 inherent variability of the results of studies in this area.
7 Because of the limitations of the cross-sectional design the relationships
8 described may not be causal and thus results should be viewed with caution.
9 For example, there are a number of problems concerning physical activity
1011 that pose problems for the cross-sectional design (see Chapter 8 for discus-
1 sion on methodological issues regarding physical activity). Also the use
2 of valid measures of cognition is another issue for both cross-sectional
3111 and intervention research (Spirduso, 1994). Measures of cognitive perfor-
4 mance in the exercise gerontology research area have included a range
5 of tests and tasks that are described in Table 6.1. Most cognitive measures
6 have involved speed such as simple and choice RT tasks. The attainment
7 of maximum speed on these tasks is influenced by a range of factors such
8 as the motivation of the individual. Also it may take considerable prac-
9 tice on the task before people can perform at their fastest. Few researchers
20111 have allowed participants to practice long enough to produce well-learned
1 performances. Thus, these studies are limited by the amount of practice
2 received (Spirduso, 1994).
3
4
Summary
5
6 Results of the majority of cross-sectional studies suggest that physically
7 fit or active older adults typically process information more efficiently than
8 their less fit counterparts. These effects are most pronounced in tasks that
9 are attention demanding and rapid (e.g., RT tasks).
30111
1
Intervention studies
2
3 As described in Chapter 8 there are numerous internal validity threats to
4 the cross-sectional design. Furthermore, this design does not allow causal
5 relationships to be inferred. The pre- to post-study design circumvents
6 many of the problems that besiege cross-sectional research. Intervention
7 studies typically administer an aerobic fitness programme to older subjects
8 and assess cognitive performance and fitness before and after the
9 programme. Fourteen published intervention studies in this area were
40111 located and these are shown in Table 6.2.
1 One of the first intervention studies was carried out by Dustman et al.
2 (1984). Authors used adults between the ages of 55 and 70 years (non-
3 exercise control, strength-and-flexibility exercise control, and an aerobic
4 exercise group) with similar socio-economic and intelligence levels.
45111 Cognitive performance was assessed though a battery of tests (see Figure
1111 6.1). After 4 months of exercise training the cognitive performance of the
2 aerobic exercise group was significantly improved. In contrast, the cognitive
3 performance of the sedentary group did not change. This study provides
4 strong support for the relationship between fitness and cognitive perfor-
5 mance because authors used appropriate control groups and trained sub-
6 jects significantly improved their aerobic power (a 27% increase).
7 In another intervention study both RT and choice RT improved after
8 a one-year exercise programme and then remained at that level with two
9 more years of exercising (Rikli & Edwards, 1991). Three years after the
1011 start of the exercise programme RTs of the non-exercising subjects were
1 significantly slower suggesting that the exercise programme may have
2 delayed age-related slowing in RT.
3111 A number of investigations have attempted to replicate Dustman
4 et al.’s findings. For example, in a series of experiments Blumenthal and
5 colleagues (1988, 1989) administered aerobic exercise, strength training,
6 yoga or sedentary living and examined pre-to-post changes in a number
7 of cognitive parameters. Results indicated that increases in aerobic power
8 did not result in enhanced cognitive performance. Chodzko-Zajko and
9 Moore (1994) have raised a number of criticisms about this study that
20111 have included the young age of participants, lack of a control group, and
1 the high fitness levels of participants before the exercise intervention.
2 Blumenthal et al. (1991) attempted to control for some of these factors
3 in a further study. Although exercisers increased aerobic power by 11.6%
4 their cognitive performance did not improve. These non-significant findings
5 are similar to the results of another training study by Panton, Graves,
6 Pollock, Hayhery and Chen (1990). In contrast, a study by Hawkins et al.
7 (1992), of similar design to that of Blumenthal et al. (1991), did find signifi-
8 cant improvements in dual-task performance after a ten-week aerobic
9 exercise programme. However, fitness change in this study was measured
30111 by changes in resting heart rate and not by direct fitness assessment.
1 As can be seen from Table 6.2, five of the intervention studies did find
2 an increase in cognitive performance after aerobic training, whereas the
3 others did not. Interestingly, the biggest increase in cognitive performance
4 was found in participants who showed the biggest increase in aerobic
5 fitness (Dustman et al., 1984). Spirduso (1994) has pointed out, however,
6 that these participants’ initial level of aerobic power was unusually low.
7 Nevertheless, it is feasible that older individuals may have to record large
8 increases in aerobic power before cognitive performance is enhanced. The
9 meta-analysis by Etnier et al. (1997) indicates that the overall effect size
40111 for intervention studies examining the relationship between fitness and
1 cognitive performance was significant and large for adults aged between
2 45 and 60 years (ES 1.02; SD 1.15) and significant but small for
3 adults aged between 60 and 90 years (ES 0.19; SD 0.37). Method-
4 ological issues concerning physical activity and the intervention design are
45111 discussed in Chapter 8.
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Table 6.2 Intervention studies examining the effect of fitness on cognitive performance of older adults
Authors Participants Design Treatment Cognitive Cognitive Fitness results

measures results
Barry et al. n = 13 males 3 months training aerobic exercises RT, Raven’s no change increase in agility
(1966) and females Matrices muscular endurance
(71 yrs)
Elsayed et al. n = 36 old and 4 months training jogging, fluid and increases in fluid increase in a
(1980) young males calisthenics crystallised intelligence composite fitness
(24–64 yrs) intelligence tests score
Dustman n = 43 indivi- 4 months training aerobic, strength a battery of tests increased 27% increase in
et al. (1984) duals (55–70 yrs) training performance V·O2max
Blumenthal n = 28 males 3 months training jogging, strength RT in a no change 15% increase in

et al. (1988) (43 yrs) training memory-search V·O2max
task
Blumenthal n = 101 males 4 months training aerobic, a battery of tests no change 11.6% increase in
·
et al. (1989) and females flexibility, VO2max
(67 yrs) control
Madden et al. n = 85 males 4 months training aerobic, RT, memory no change 11% increase in
(1989) and females anaerobic, retrieval control V·O2max
(60–80 yrs)
Stones & n = 200 males 12 months training aerobic exercises RT, digit symbol increased increase in an
Kozma and females subtest performance aggregated fitness
(1989) score
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Emery & n = 48 males 3 months training aerobic exercise, digit span subtest, no change unclear
Gatz (1990) and females social activity, digit symbol
(61–86 yrs) control waiting subtest
Panton et al. n = 49 males 6 months training walk/jog, strength RT, PMT, MT no change 20% increase in
(1990) USA and females training, control V·O2max
Blumenthal n = 101 males 14 months training aerobic exercise, a battery of tests no change 10–15% increase in
et al. (1991) and females yoga, waiting list V·O2max
(60–83 yrs)
Hawkins n = 37 males 10 weeks training swimming, time-sharing, increase in not measured

et al. (1992) and females control dual-task dual-task
(63–82 yrs) performance performance
Hill et al. n = 87 adults 12 months training endurance, Wechsler no change increase in fitness
(1993) exercise memory
control scale
Paas et al. n = 58 adults 10 months training running, control RT, CRT, letter no change 20% increase in
(1994) recognition fitness
Moul et al. n = 30 males 4 months training walking, weights, Ross Information increased 15.8 % increase in
(1995) and females control Processing performance V·O2max
Assessment
Notes:
RT = reaction time, MT = movement time, PMT = pre-motor time, CRT = choice reaction time.
1111
2 108
3
4 * = P < 0.05
5
6
106 *
7
8
9 104
STANDARD SCORES
1011
1
2 102
3111
4
5
100
*
6
7
8
9 98
20111
1
2
96
3 PRE POST PRE POST PRE POST
4 AEROBIC EXERCISE NON-EXERCISE
5 CONTROL CONTROL
6
7 Figure 6.1 Standard scores averaged across eight cognitive tests (Culture Fair IQ,
8 Digit Span, Digit Symbol, Dots, RT, Stroop Interference and Stroop
Total) for aerobically trained, strength trained, and control subjects.
9
Adapted from R.E. Dustman, R.O. Ruhling, E.M. Russell, D.E. Shearer, H.W. Bonekat,
30111 J.W. Shigeoka, J.S. Wood, & D.C. Bradford (1984). ‘Aerobic exercise training and improved
1 neuropsychological function of older individuals.’ Neurobiology of Aging, 5, 35–42, with
2 permission from Elsevier Science.
3
4
5
Summary
6
7 Results of intervention studies examining the effect of exercise on cogni-
8 tive functioning in older individuals are equivocal. Five intervention studies
9 demonstrate improvement in cognitive performance following aerobic
40111 training whereas a number of well-controlled experiments have not (Table
1 6.2). The meta-analysis by Etnier et al. (1997) indicates that the overall
2 effect size for intervention studies examining the relationship between
3 fitness and cognitive performance in older adults is significant but small.
4
45111
1111 Mechanisms underlying the fitness/cognitive performance
2 relationship
3
4 Although the mechanisms underlying the fitness/cognitive relationship
5 have not been identified, several lines of evidence provide preliminary
6 support for a number of hypotheses (see Chodzo-Zajko & Moore, 1994;
7 Spirduso, 1980; Spirduso, 1994). These include cerebral circulation, neuro-
8 trophic stimulation, neural efficiency, secondary ageing, and psychosocial
9 mechanisms.
1011
1 Cerebral circulation
2
3111 Evidence suggests that impaired cerebral circulation is associated with
4 reduced cognitive performance (see Chodzo-Zaiko & Moore, 1994; Speith,
5 1965). For instance, induced hypoxia has resulted in cognitive decline on a
6 number of cognitive tasks (Kennedy, Dunlap, Bandert, Smith, & Houston,
7 1989; McFarland, 1963). Significantly, cognitive decline was similar to that
8 seen in non-hypoxic elderly subjects. In contrast, oxygen supplementation
9 has been found to result in improved cognitive performance (Ben-Yashai
20111 & Diller, 1973). Thus, chronic exercise could maintain cerebrovascular
1 integrity by enhancing oxygen transportation to the brain and the enhanced
2 blood flow may positively influence cognitive performance (Chodzo-Zaiko
3 & Moore, 1994).
4
5 Neurotrophic stimulation
6
7 Reduced neurotransmitter synthesis, structural alterations to neurones,
8 and a general degradation of the central nervous system are associated
9 with ageing (Chodzo-Zajko & Moore, 1994; Cottman & Holets, 1985). It
30111 is possible that participation in physical activity may offset these changes
1 associated with advancing age (Chodzo-Zaiko & Moore, 1994). For
2 example, chronic exercise is associated with increased brain weight in both
3 primates (Floeter & Greenough, 1979) and rats (Pysh & Weiss, 1979).
4 Thus, regular exercise may improve cognitive functioning by offsetting the
5 general degradation of the central nervous system.
6
7
Neural efficiency
8
9 Whereas cerebral circulation and neurotrophic adaptations are direct
40111 mechanisms, other explanations are indirect, and focus on the informa-
1 tion processing efficiency of the central nervous system (Chodzo-Zaiko &
2 Moore, 1994). For example, electroencephalographic (EEG) responses
3 have been examined to determine if fitness improves cognitive processing
4 in old age (Spirduso, 1994). It has been found that old age is associated
45111 with changes in EEG responses (Dustman, Emmerson et al., 1990).
1111 Dustman, Emmerson et al. (1990) recorded the EEG response of fit and
2 unfit young (20–31 years) and older (50–62 years) adults and found that
3 both the young and older fit adults possessed faster components to their
4 visual evoked potentials than those who did not exercise. Thus, there is
5 preliminary evidence to indicate that participation in regular aerobic exer-
6 cise by older adults is associated with enhanced central nervous system
7 processing.
8
9
Secondary ageing mechanisms
1011
1 Regular exercise may also have a secondary effect on the cognitive perfor-
2 mance of older individuals by preventing disease. Cardiovascular disease,
3111 adult onset diabetes, and hypertension are thought to impair cognitive
4 function (Birren, Woods, & Williams, 1980; Hertzog, Schaie, & Gribbin,
5 1978). As people age the incidence of these diseases increases significantly
6 (Birren et al., 1980). In contrast, regular exercise has been associated with
7 reduced incidence of cardiovascular disease and hypertension and is impor-
8 tant in helping to control adult onset diabetes by enhancing glucose
9 tolerance and insulin sensitivity.
20111
1
Psychosocial mechanisms
2
3 Chronic exercisers differ in a number of ways from inactive individuals.
4 For example, people who exercise regularly have generally higher incomes,
5 educational levels, and cognitive abilities (Stones & Kozma, 1989), and
6 typically have more positive attitudes (Clarkson-Smith & Hartley, 1989).
7 Thus, the development of more positive attitudes accompanying partici-
8 pation in a fitness programme could result in greater motivation to try
9 harder on the cognitive tasks during the post-programme test.
30111
1
Summary
2
3 A number of direct and indirect mechanisms could underpin the relation-
4 ship between fitness and the cognitive performance of older adults. These
5 include enhanced cerebral blood-flow, increased neurotrophic stimulation,
6 increased neural efficiency, improved health, and greater motivation.
7 Research into these mechanisms is preliminary at this stage and has not
8 produced conclusive findings.
9
40111
1
Future research
2 There is a need for exercise scientists to collaborate with medical and
3 gerontological researchers to identify the mechanisms underlying the cog-
4 nitive performance enhancement effects of exercise. In addition, random-
45111 ised controlled trials need to be long enough to bring about physiological
1111 adaptation and need to use valid measures of cognitive performance and
2 valid measurement of aerobic fitness. Finally, cross-sectional studies com-
3 paring highly trained and untrained older adults’ cognitive performance
4 whilst controlling for subject differences are required.
5
6
7 What we know
8
9 • The majority of cross-sectional studies show that fit older adults
1011 display better cognitive performance than less fit older adults.
1 • The association between fitness and cognitive performance is
2 task-dependent with most pronounced effects in tasks that are
3111 attention-demanding and rapid (e.g. RT tasks).
4 • Results of intervention studies are equivocal but meta-analytic
5 findings indicate a small but significant improvement in cognitive
6 performance of older adults who experience an increase in
7 aerobic fitness.
8 • Research into possible mechanisms underlying the effects of
9 fitness on cognitive performance in older adults is preliminary at
20111 this stage and has not produced conclusive findings.
1
2
3
5
6 • What are the mechanisms underlying the effects of fitness on
7 the enhanced cognitive performance of older adults?
8 • What is the dose-response relationship? In other words how
9 much of a fitness improvement needs to occur before an increase
30111 in cognitive performance is demonstrated?
1 • Is it necessary to develop physiological adaptations to an
2 exercise regimen before cognitive performance increases occur?
3 • What aspect of cognitive functioning of older adults is most
4 greatly affected by increases in fitness? For instance, is speed
5 in contrast to memory recall most likely to be improved with
6 increases in fitness?
7 • If regular exercise results in enhanced cognitive performance
8 does it do so at all ages or are the effects more significant in
9 older adults?
40111 • What happens to the cognitive performance of those individuals
1 who stop exercising?
2 • Do older females record greater increases in cognitive perfor-
3 mance after exercise training than older males?
4
45111
1111
2 7 Physical activity as a source
3
4 of psychological dysfunction
5
6 Attila Szabo
7
8
9
1011
1
2
3111 The mental and physiological benefits of physical activity are almost undis-
4 puted. There is a strong consensus in scientific circles with regard to the
5 value of integrating physical activity in one’s regular lifestyle (Bouchard,
6 Shephard, & Stephens, 1994). However, in isolated cases (a key term
7 throughout this article is isolated cases) physical activity can lead to unde-
8 sirable or harmful psychological states. Although these states are mostly
9 transient, their impact may have irreversible damage on the life of the
20111 individual. The aim of this chapter is by no means to take a stand against
1 physical activity, but simply to discuss the relatively rare circumstances in
2 which exercise or physical activity may inflict substantial or even perma-
3 nent damage on physically active people. The aim of the chapter is to
4 help the reader become aware of exercise related factors that are associated
5 with psychological dysfunction. It is hoped that knowledge and timely
6 recognition of these factors may prevent the development of ‘unhealthy’
7 exercise behaviours and their negative consequences.
8
9
The dark side of physical activity
30111
1
Negative consequences of physical activity
2
3 Negative psychological experiences associated with exercise participation
4 may occur early in the adoption phase or later in the exercise mainte-
5 nance phase. They may be classified as having acute (transient) and chronic
6 (persistent) effects. In the adoption phase, psychological hardship is mostly
7 transient and it may lead to drop-out from the exercise programme
8 (Dishman, 1988, 1993; Gauvin, 1990). In the exercise maintenance phase,
9 more severe psychological dysfunction is associated with one’s attitude
40111 towards physical activity and, therefore, it may persist for a relatively long
1 period.
2
3
During the adoption phase
4
45111 At the early phases of adoption of physical activity there are two general
outcomes. The person will either like the activity and continue exercising,
Physical activity and psychological dysfunction 131
1111 or dislike it – for a number of reasons – and discontinue the activity. In
2 the case of exercise maintenance the early experiences may often be
3 stressful. Levels of physical effort exerted by the novice exerciser may
4 be high which in turn may lead to negative affect (Raglin, 1993). Further,
5 if socially undesirable physical appearance was among the motives for the
6 adoption of physical activity, a possible anxiety associated with one’s
7 appearance – ‘social physique anxiety’ – may have aversive psychological
8 effects (Hart, Leary, & Rejeski, 1989). Social physique anxiety is relatively
9 common and it results from the comparison of the self with the appearance
1011 of co-exercisers. Therefore, this exercise-related anxiety is most prevalent
1 in physical activities practised in social settings.
2 Physical activity can also induce anxiety in ‘exercise sensitive’ individuals.
3111 Cameron and Hudson (1986) reported that 20% of 66 patients diagnosed
4 with anxiety disorders, on the basis of the Diagnostic and Statistical Manual
5 (DSM-III), as well as 22% of 37 controls reported experiencing anxiety
6 during exercise. Whilst the patients responded with greater anxiety in gen-
7 eral, both groups experienced increasingly greater anxiety when exercise
8 intensity was increased sequentially from ‘very mild’ to ‘very vigorous’
9 levels through five stages. The findings of Cameron and Hudson (1986)
20111 were expanded to healthy individuals in a recent inquiry that demon-
1 strated that a 20-minute bout of high intensity resistance exercise increased
2 state anxiety whereas low intensity exercise had the opposite effect
3 (Bartholomew & Linder, 1998). High intensity exercise training, especially
4 in competitive athletes, was also associated with negative changes in mood
5 states (Raglin, 1993; Raglin, Eksten, & Garl, 1995; Raglin, Koceja, Stager,
6 & Harms, 1996). Consequently, high intensity exercise may lead to anxiety
7 and negative mood which than can lead to the abandonment of the exercise
8 regimen (see also Biddle, this volume).
9 Often the early phases of adoption of physical activity will end with
30111 non-adherence to exercise. There may be many reasons for this including
1 inappropriate physical challenge, unpleasant social and/or environmental
2 settings, or an inability to handle other life commitments effectively. Non-
3 adherence is extremely common. In fact about 60% of those who intend to
4 adopt an exercise activity will discontinue the selected activity within the first
5 six months (Dishman, 1988, 1993). Non-adherence may have a significant neg-
6 ative impact on the mental health of the individual (Brewer, 1993). For exam-
7 ple, feelings of guilt, inadequacy, lack of determination, and social critique
8 can trigger decrements in one’s self-concept and/or self-esteem, predisposing
9 the person to a number of psychological dysfunctions such as depression and
40111 anxiety. Regrettably, no scientific inquiries have been undertaken to identify
1 the psychological consequences of abandoning exercise regimens.
2
3
After the adoption phase
4
45111 Once physical activity has been adopted as a regular part of the indi-
vidual’s lifestyle the most critical issue is the pattern or habit of practising
132 Attila Szabo
1111 the activity. As with all behaviours, moderation is important. Overdoing
2 the adopted physical activity can lead to both injuries and to the neglect
3 of other important responsibilities in life. In fact, for various reasons,
4 exercisers may lose control over their exercise and travel a ‘path of self-
5 destruction’ (Morgan, 1979). The most common psychological dysfunction
6 associated with over-exercising, is exercise dependence. This disorder is
7 classified as ‘primary exercise dependence’ when it manifests itself in the
8 form of behavioural addiction, and as ‘secondary exercise dependence’
9 when it appears in conjunction with eating disorders (De Coverley Veale,
1011 1987). The remainder of the text will examine primary and secondary
1 exercise dependence.
2
3111
Primary exercise dependence
4
5 An abnormal reliance on physical activity is frequently termed exercise
6 addiction (e.g. Thaxton, 1982), exercise dependence (e.g. Cockerill &
7 Riddington, 1996), or obligatory exercising (Pasman & Thompson, 1988).
8 The term exercise is often replaced by the term physical activity or the
9 name of the activity, such as running (e.g. Chapman & De Castro, 1990).
20111 In some instances the concept is referred to as negative addiction (e.g.
1 Rudy & Estok, 1989) to contrast the idea of positive addiction (Glasser,
2 1976) that was introduced as a more favourable perspective. Unfortunately,
3 the use of the latter terminology yielded confusion in the scientific liter-
4 ature and will be discussed later. The currently preferred terminology is
5 exercise dependence (Cockerill & Riddington, 1996; De Coverley Veale,
6 1987). The prevalence of exercise dependence is not well known, but it
7 is speculated that only a very small percentage of regular exercisers are
8 affected (De Coverley Veale, 1987; Morris, 1989).
9
30111
The conceptual meaning of exercise dependence
1
2 In 1976, Glasser introduced the term positive addiction to denote the per-
3 sonally and socially beneficial aspects of a regular physical activity regimen
4 in contrast to some self-destructive behaviours. The ‘positive’ perception
5 and subsequent adoption of the terminology led to its widespread but loose
6 usage within both athletic and scientific circles. Morgan (1979) was first to
7 realise the dilemma. To discuss some rare but existing compulsive exercise
8 behaviours, he introduced the term negative addiction as a counter-term to
9 positive addiction. The fact is, however, that all addicted behaviours are
40111 pathological and they are always negative (Rozin & Stoess, 1993).
1 Glasser’s (1976) ‘positive’ prefix referred to the benefits of commitment
2 to physical activity (a healthy behaviour) as opposed to the negative effects
3 of ‘unhealthy’ addictions, such as smoking, drinking, or drug abuse.
4 In fact, positive addiction can be viewed as a synonym for commitment
45111 to physical activity (Carmack & Martens, 1979; Pierce, 1994). However,
1111 when commitment to physical activity is equated to addiction or to depen-
2 dence on exercise (Conboy, 1994; Sachs, 1981; Thornton & Scott, 1995) a
3 conceptual confusion is created. For example, Thornton and Scott (1995)
4 found that they could classify 77% of a sample of 40 runners as moder-
5 ately or highly addicted to running. Since any addiction is a serious
6 psychopathological condition, this figure is certainly exaggerated. Some
7 researchers have realised this problem and attempted to draw a fine line
8 between commitment and addiction to physical activity (Chapman & De
9 Castro, 1990; Summers & Hinton, 1986; Szabo, Frenkl, & Caputo, 1997).
1011 Most recent evidence indicates that addiction to exercise and commitment
1 to exercise are independent concepts (Szabo et al., 1997).
2
3111
Drawing a fine line between commitment and addiction to
4
physical activity
5
6
Commitment to physical activity
7
8 This is a reflection of how dedicated or devoted a person is to her/his
9 physical activity. It is a measure of the strength of adherence to an adopted
20111 physical activity regimen. For the committed exerciser, satisfaction, enjoy-
1 ment, and achievement derived from exercise are incentives that motivate
2 the continuity of the behaviour (Chapman & De Castro, 1990). Sachs
3 (1981) viewed commitment to exercise as a result of intellectual analysis
4 of the rewards gained from exercise, including social relationships, health
5 benefits, status, prestige, or monetary advantages. Committed exercisers,
6 in light of Sachs’ (1981) description (a) often exercise for extrinsic rewards,
7 (b) view their exercise as an important, but not central, part of their lives,
8 and (c) may not suffer severe withdrawal symptoms when they cannot
9 exercise for some reason (Summers & Hinton, 1986). The committed exer-
30111 ciser controls her/his physical activity (Johnson, 1995).
1
2
Addiction to physical activity
3
4 In contrast to committed exercisers, addicted exercisers, according to Sachs
5 (1981) are more likely to (a) exercise for internal satisfaction, (b) view
6 exercise as the chief part of their lives, and (c) experience strong depri-
7 vation sensations when they are unable to exercise (Summers & Hinton,
8 1986). Motivation for exercise is another distinguishing feature between
9 commitment and addiction to exercise. Exercise-dependent individuals
40111 may be motivated by negative (avoiding withdrawal symptoms) and/or
1 positive (‘runners’ high’) reinforcements (Pierce, 1994; Szabo, 1995).
2 However, negative reinforcement, or avoidance behaviour, is not a usual
3 motivating factor for the committed exerciser (Szabo, 1995). Finally, in
4 contrast to the committed exerciser, the exercise dependent person loses
45111 control over her/his physical activity (Johnson, 1995).
134 Attila Szabo
1111 Definition and symptoms of exercise dependence
2
An abnormal or unhealthy exercise pattern is a pre-requisite in the clas-
3
sification of exercise dependence (Cockerill & Riddington, 1996). A
4
commonly used definition for exercise dependence stems from Sachs (1981,
5
p. 118) who described addiction to running as ‘addiction of a psycholog-
6
ical and/or physiological nature, upon a regular regimen of running,
7
characterized by withdrawal symptoms after 24 to 36 hours without partic-
8
ipation’. This definition is well accepted in the literature (Furst &
9
1011 Germone, 1993; Morris, 1989; Sachs & Pargman, 1984). However, there
1 is a problem with this definition because withdrawal symptoms are only
2 one of the many characteristics of addictive behaviours (Brown, 1993;
3111 Griffiths, 1997). Further, many empirical studies have simply assessed the
4 mere presence, rather than the type, frequency, and intensity of with-
5 drawal symptoms (Szabo, 1995; Szabo et al., 1997). Yet negative psycho-
6 logical symptoms are reported by all committed, not necessarily addicted,
7 exercisers for the times when exercise is prevented for an involuntary
8 reason (Szabo, 1997; Szabo, Frenkl, & Caputo, 1996). Even participants
9 in physically less strenuous leisure activities, such as bowlers, report depri-
20111 vation sensations when bowling is not possible. However, the intensity of
1 the symptoms reported by this group is less than that reported by aerobic
2 dancers, weight-trainers, cross-trainers, or fencers (Szabo et al., 1996).
3 Consequently, it is clear that the presence of withdrawal symptoms in
4 itself is insufficient for the diagnosis of exercise dependence. Cockerill and
5 Riddington (1996) do not even mention withdrawal symptoms in their
6 listing of characteristics associated with exercise dependence. In fact the
7 presence of withdrawal symptoms, in many forms of physical activity,
8 suggests that exercise has a positive effect on people’s psychological well-
9 being. This effect is missed when an interruption of the habitual activity
30111 is commanded by unforeseen circumstances.
1 A close analysis of the literature reveals that there is a lack of a sound
2 definition of exercise dependence. Johnson’s (1995) point, that the exercise-
3 dependent individual is controlled by her/his physical activity, may be a
4 starting point in the quest for a more appealing conceptual definition. The
5 joint presence of a number of commonly reported symptoms should then be
6 closely observed in the classification of genuine exercise dependence (see
7 Table 7.1). These symptoms, unfortunately, were either ignored or only
8 loosely attended to (i.e. the degree of their severity was overlooked) in
9 empirical research.
40111
1
Assessment of exercise dependence
2
3 There are major limitations in the assessment of exercise dependence. The
4 available ‘addiction’ scales (see Table 7.2) are embedded, at least in part,
45111 with questions pertaining to commitment rather than to dependence on
1111 Table 7.1 Components or symptoms of exercise dependence
2
Components of addiction Description
3
4 1 Salience The physical activity assumes a primary role in
5 (Brown, 1993; De Coverley the person’s life and there is an obsessive
6 Veale, 1987; Griffiths, 1997) preoccupation with the practiced activity many
times a day, and even during other activities
7
8 2 Euphoria/satisfaction A psychological and/or physical experience that
9 (Brown, 1993; Cockerill & may be identified as the main driving force
Riddington, 1996; Griffiths, (i.e. the key source of motivation) for the chosen
1011 1997) physical activity and its pattern
1
3 Tolerance The ‘dose-dependent’ aspect of exercise addiction.
2
(Brown, 1993; De Coverley The person needs to progressively increase the
3111 Veale 1987; Griffiths, 1997) frequency, duration, and/or the intensity of the
4 adopted physical activity to derive the level of
5 satisfaction previously attained with lesser exercise
6 4 Withdrawal symptoms Severe negative physical or psychological feelings
7 (Brown, 1993; De Coverley experienced when the adopted physical activity
8 Veale 1987; Griffiths, 1997) cannot be performed for an involuntary reason
9 (i.e. injury or other commitments)
20111 5 Conflict Interpersonal problems arising from over-exercising
1 (Brown, 1993; Griffiths, and intrapersonal (intrapsychic) conflict arising
1997) from feelings of guilt and dissatisfaction
2
associated with the neglect of other (than
3 exercise) life obligations
4
6 Relapse A predisposition to re-establishment of the
5 (Brown, 1993; De Coverley previously ‘unhealthy’ pattern of exercise behaviour
6 Veale 1987; Griffiths, 1997) after a period of abstinence or ‘normal’
7 exercising. (Due to the effort involved in most
8 physical activities, this component is argued to be
9 weak in exercise addiction (Cockerill &
Riddington, 1996))
30111
1 7 Loss of control over The urge for exercising becomes so intense that it
life-activities interferes with other activities and the need for
2
(Griffiths, 1997) exercise often becomes immediate. Until that need
3 is satisfied, the other life-activities are deficiently
4 performed
5 8 Loss of control over The exerciser cannot resist the urge to exercise.
6 one’s exercise behaviour While the exercise dependent person may try to
7 (Cockerill & Riddington, set limits in her/his exercise patterns, she/he is
8 1996; Johnson, 1995) unable to respect those self-set limits. In short, a
9 lack of ability to exercise with moderation
40111 9 Negative consequences The negative outcomes directly associated with
1 (Griffiths, 1997) over-exercising. It may involve the loss of
employment, poor academic performance,
2
break-up in marriage or other relationship(s) and
3 other consequences generally considered to have
4 an undesirable effect on a person’s life
45111
136 Attila Szabo
1111 Table 7.1 (continued)
2
Components of addiction Description
3
4 10 Risk of self-injury At times of mild injuries the addicted exerciser
5 (De Coverley Veale, 1987; cannot abstain from exercise and, thus, assumes
6 Wichmann & Martin, 1992) the risk of self-injury by maintaining her/his
physical activity
7
8 11 Social withdrawal A search for approval from other exercisers, and
9 (Cockerill & Riddington, avoidance of the company of those who criticise
1996) the physical activity pattern of the exercise-
1011 dependent person
1
12 Lack of compromise Signs of neglecting family or work responsibilities
2
(Wichmann & Martin, 1992) to spend more time exercising. Other life-
3111 commitments are ignored even though this carries
4 significant negative consequences
5 13 Denial of a problem or The rationalisation of the problem via conscious
6 self-justification search for reasons why exercise should take
7 (Wichmann & Martin, 1992) priority over all other life commitments
8 14 Full awareness of the The exercise dependent person knows well that
9 problem there is a problem with her/his exercise
20111 (De Coverley Veale,1987) behaviour through feedback from other people
1 or from the negative life-events directly resulting
from over-exercising
2
3
4
5 exercise. Furthermore, like most Likert scales, they were developed to
6 measure the degree of addiction, rather than to positively diagnose exer-
7 cise dependence. Then there is confusion between the distinct concepts
8 of addiction and commitment. For example, Conboy (1994) used the
9 Commitment to Running Scale (Carmack & Martens, 1979) to measure
30111 addiction. Such confusion hinders the valid assessment of exercise depen-
1 dence. In-depth interviews, structured around the classic symptoms of
2 addiction (Table 7.1), conducted with both the exercisers and her/his rela-
3 tives and/or colleagues may be the most appropriate technique for the
4 positive identification of exercise dependence.
5
6 Possible causes of addiction to physical activity
7
8 Physiological explanations
9
40111 While the symptoms of exercise addiction may be recognised, the causes of
1 exercise dependence are far from clear. A number of hypotheses have been
2 put forward. One of them is the popular endorphin hypothesis. According
3 to this model, exercise stimulates the release of beta-endorphins, a natural
4 opiate in the brain, that induces feelings of euphoria such as the well known
45111 ‘runners’ high’ (Pierce, 1994). This hypothesis views exercisers as ‘opiate
1111 junkies’ and is highly challenged. Indeed, there is psychopharmacological
2 counter-evidence for this explanation (De Coverley Veale, 1987; Pierce,
3 1994). Further, there is a lack of correlation between exercise dependence
4 scores and beta-endorphin immunoreactivity (Pierce, 1994). Finally, the
5 biological obstacle presented by the blood brain barrier in the transport of
6 the peripheral beta-endorphins to the brain also argues against the endor-
7 phin hypothesis (De Coverley Veale, 1987; Pierce, 1994).
8 A more plausible physiological explanation, known as the ‘sympathetic
9 arousal hypothesis’, was proposed by Thompson and Blanton (1987).
1011 According to this model, regular exercise leads to decreased sympathetic
1 arousal at rest. This adaptation leads to a lethargic state with lack of exer-
2 cise. To maintain an optimal level of arousal, and to overcome the lethargic
3111 state at rest, the habituated exerciser needs to exercise to increase her/his
4 level of arousal. The associated symptoms match the symptoms of exercise
5 dependence. Therefore, on physiological grounds this explanation may be
6 plausible.
7
8
Psychological explanations
9
20111 Some runners try to run away from their psychological problems (Morris,
1 1989). They are very few in number and use exercise as a means of coping
2 or escape from serious distress. Like others, who turn to drugs and alcohol
3 in cases of severe distress, exercisers may abuse their exercise so that
4 behavioural addiction becomes evident (Griffiths, 1997). Since exercise,
5 in contrast to alcohol or drugs, requires substantial physical effort
6 (Cockerill & Riddington, 1996), it is a less ‘convenient’ coping mechanism
7 that requires strong self-determination and self-discipline. Therefore, the
8 cases of exercise addiction or dependence are very rare in contrast to
9 other forms of escape behaviours.
30111
1
Correlates of exercise dependence
2
3 Researchers have looked at the correlates of exercise dependence (see
4 Table 7.2), but have been unable to identify when or why a transition
5 takes place from ‘healthy’ to ‘unhealthy’ exercise behaviour (Johnson,
6 1995). Exercise dependence appears to be positively related to anxiety
7 (Morgan, 1979; Rudy & Estok, 1989) and negatively related to self-esteem
8 (Estok & Rudy, 1986; Rudy & Estok, 1989). Further, the length of expe-
9 rience with a particular physical activity appears to be positively associated
40111 with exercise dependence (Furst & Germone, 1993; Hailey & Bailey, 1982;
1 Thaxton, 1982). If experience is associated with exercise dependence, it
2 is reasonable to speculate that a major life event change (or stress) may
3 trigger addiction that is exhibited through ‘revolutionary’ rather than evo-
4 lutionary changes in the habitual physical activity pattern of the individual.
45111 The affected individual may see this form of coping as healthy on the
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Table 7.2 Summary table of research into exercise dependence
Author(s) Participants and type Measure(s) of exercise Correlates of addiction to Main findings and/or
of physical activity dependence/addiction physical activity conclusions
Anshel (1991) 60 exercisers in health amount (frequency and restlessness, withdrawal differences observed
and fitness centre hours) or participation in symptoms, positive mood between the addicted and
(30 males & exercise sessions after exercise, ignorance non-addicted exercisers
30 females) of physical discomfort as well as between males
(i.e., injuries) and females
Carmack & 315 runners subjectively perceived levels examined as a secondary perceived addiction was
Martens (1979) (250 males & of addiction (‘very much’, variable; subjective found to be a predictor
65 females) ‘somewhat’, & ‘not at all’) conceptualisation by the of commitment to
participants runnning
Chapman & 47 runner Running Addiction Scale addiction scores were

De Castro (32 males & (RAS – self-developed) related to high frequency (1) large gender differences
(1990) 17 females) of running and positive observed; (2) duration of
psychological characteristics run, but not addiction,
was related to mood
enhancement
Conboy (1994) 61 runners Commitment to Running withdrawal symptoms on withdrawal effects can
(51 males & Scale (CR – Carmack & non-running days only be seen when
10 females) Martens, 1979) commitment and
dependence are jointly
tested
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Crossman, 31 runners self-perceived addiction exercise withdrawal was postulates the existence of
Jamieson, & (15 males & rated on a 9-point Likert unrelated to the scores of two opposing processes;
Henderson (1987) 16 females – scale ranging from ‘not at all’ subjectively reported one negative due to
study 1) and to ‘extremely‘ addiction exercise withdrawal and
20 swimmers one positive due to rest
(12 females & during periods of exercise
8 males – study 2) withdrawal
Estok & Rudy 57 marathon and 38 Running Addiction Scale higher addiction scores were higher addiction scores
(1986) non-marathon runner (RAS – self-developed) related to lower self-esteem were not related to more
females physical injuries and no
differences were seen
between the groups
Furst & Germone 98 runners (72 males Negative Addiction Scale duration of involvement in no gender or physical
(1993) & 26 females) and (Hailey & Bailey, 1982) physical activity was related activity (type) differences
90 other exercisers to higher addiction scores were seen in relation to
(60 males & 30 the addiction scores
females)
Griffiths (1997) 1 female jiu-jitsu interview modelled on salience, tolerance, exercise has taken over a
practitioner Brown’s (1993) addictive withdrawal, euphoria, person’s life which argues
components conflict, loss of control, against the concept of
relapse ‘positive addiction’
Hailey & Bailey 60 male runners Negative Addiction Scale duration of involvement in with a few exceptions
(1982) placed in groups on running was related to negative addiction is
the basis of their higher negative addiction linearly related to length
running history scores of running experience
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Author(s) Participants and type Measure(s) of exercise Correlates of addiction to Main findings and/or
of physical activity dependence/addiction physical activity conclusions
Morgan (1979) 8 case studies self-reported compulsive withdrawal symptoms, runners (exercisers) may
(6 runners, patterns of exercising and exercise in spite of injuries, lose perspective and the
1 swimmer and withdrawal symptoms in lack ignorance of major life- control over their exercise
1 wrestler) of exercise commitments, multiple and finally undergo the
exercise sessions on the path of self-destruction
same days
Pasman & 90 participants Obligatory Exercise withdrawal symptoms and related obligatory
Thompson, (1988) (45 males & Questionnaire (modified from the maintenance of exercise exercising to eating
45 females who the Obligatory Running in spite of injuries disturbances
were represented Questionnaire (Blumenthal
equally in three et al., 1985))
groups: runners,
weightlifters
and controls)
Pierce, Daleng, & 102 females Negative Addiction Scale type of activity may be dancers showed greater
McGowan (47 dancers, (Hailey & Bailey, 1982) related to exercise addiction addiction to their
(1993) 39 runners, and physical activity than
16 field-hockey runners, but only
athletes) speculative explanations
are provided
Rudy & Estok 202 marathon Running Addiction Scale addiction was positively proposes that the
(1989) runners (104 (Estok & Rudy, 1986) that related to anxiety and it Running Addiction
females & 98 males) was further improved here was negatively related to Scale is a useful tool for
self-esteem gauging negative addiction
in runners
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Rudy & Estok 35 runners Running Addiction Scale addiction to running was speculates that intense
(1990) (22 females & (Rudy & Estok, 1989) unrelated to dyadic running jeopardises
13 males and adjustment, but spouses’ family relationships and
their spouses) rating of addiction was dyadic adjustments
negatively related to
their own adjustment
Sachs & Pargman 12 runners in-depth interview related running addiction to health reasons play an
(1979) (all males) the high life-priority of important role in running
running and feelings of addiction and the
accomplishment and withdrawal symptoms
relaxation after running (when not running) are
key signs of addiction
Thaxton (1982) 33 runners self-reported on a 10-point related running addiction to concludes that even slight
(24 males & rating scale (ranging from running experience (years variation from the running
9 females) ‘non-addicted’ to ‘extremely of running) and weekly schedule can have
addicted’) frequency of running negative effects on
habitual runners
Thornton & Scott 40 male runners Running Addiction Scale related running addiction to stresses the individual
(1995) (Rudy & Estok, 1989) and the frequency and the differences in the
Negative Addiction Scale distance of running, examination of running
(Hailey & Bailey, 1982) rather than to running addiction
experience
142 Attila Szabo
1111 basis of popular knowledge and the media-spread information about
2 the positive aspects of exercise.
3 Indeed, the media plays an important role in what people believe about
4 and expect from their exercise. The media-propagated positive image of
5 the exercising individual provides a mask behind which some exercisers
6 with severe emotional distress can hide. Thus the media-projected positive
7 information about physical activity can be used to deny the existence of
8 the problem (a characteristic of addictive behaviours) and to delay its
9 detection to the advanced stages when all symptoms of addiction are
1011 vividly present. Because of such a possible delay, it is likely that only case
1 studies, presented in the literature, reflect genuine cases of exercise depen-
2 dence. Indeed, a random sample of habitual exercisers may contain very
3111 few cases, if any (!), of exercise addicts (Morris, 1989).
4
5
Recommendations for the individual
6
7 Physically active people should keep their physical activity in perspective.
8 The person who feels that she/he is at risk may wish to evaluate the state-
9 ments in Table 7.3(a). The statements are based on some critical addictive
20111 components presented in Table 7.1 and Zaitz’s (1989) opinion. If most of
1 the statements are ‘true’ there is a need to acknowledge the possibility
2 of a problem and to take immediate action. For the rebuilding of a healthy
3 exercise pattern, Zaitz (1989) proposes some self-help strategies (see Table
4 7.3(b)). However, if no self-set changes in exercise habits seem to be
5 viable, the person should consider seeking professional help. At this stage
6 it is important to remember that over-exercising may have serious detri-
7 mental and irreversible effects and the root of the problem (i.e. the
8 reason(s) for over-exercise) can and must be identified. The mere self-
9 acknowledgement of the problem is already a significant step towards the
30111 rectification of the problem.
1
2
Recommendations for health practitioners
3
4 Exercise-dependent people (Wichmann & Martin, 1992) frequently visit
5 orthopaedic clinics because they continue exercising even at times of minor
6 injuries. These injuries then become more severe and force the individual
7 to seek medical help. Therefore, orthopaedic surgeons, physiotherapists,
8 and occupational therapists should be familiar with the symptoms of exer-
9 cise dependence. When recognising the symptoms in a patient, they should
40111 refer the affected individual to their colleagues specialised in the area of
1 behavioural addictions. Exercise dependence, like other behavioural addic-
2 tions, should be considered to be a serious condition. Once a positive
3 diagnosis has emerged, the principal concern should be to find the main
4 causes of the dependence. The treatment should be geared toward the
45111 cause, not the symptom, of over-exercising.
1111 Table 7.3(a) Exercise dependence self-evaluation check list
2
1 Exercise has the highest priority in my life.
3
2 I have experienced major losses due to exercise.
4 3 I cannot miss a scheduled exercise no matter what.
5 4 I will exercise against medical advice or when injured.
6 5 I am irritated and intolerable when I miss my exercise.
7 6 I will make no compromises when it comes to exercise.
8
9
1011 Table 7.3(b) Self-help strategies for the modification of exercise habits (based
1 on Zaitz (1989)
2
3111 1 Fulfil the urge of exercising with different modalities of exercise (i.e. cross-
training).
4
2 Consciously schedule a ‘reasonable’ rest period between two bouts of
5 exercise to prevent mental and physical fatigue.
6 3 Exercise your mind by getting involved in mental and social activities that
7 can lower anxiety and give a burst to the self-esteem.
8 4 Learn stress-management techniques, relaxation. and/or meditation.
9
20111
Recommendations for research
1
2 Future research should treat exercise dependence as a potentially serious
3 disorder that may be part of a class of disorders often referred to as behav-
4 ioural addictions. Since most likely only a very small portion of the
5 exercising population is exercise dependent (Morris, 1989; Pierce, 1994),
6 clinical case studies should form the infrastructure of research on exer-
7 cise dependence. Excessive exercise, whether conceptualised in terms of
8 high frequency, intensity, duration, or history, should not be used as the
9 only criterion in the diagnosis of exercise dependence. If the physically
30111 active individual has not experienced major negative life-events directly
1 traceable to her or his exercising behaviour and she/he is not jeopardising
2 personal health and social relationships, the presence of exercise depen-
3 dence is unlikely. Further, because existing tools for gauging exercise
4 dependence are mostly Likert scales, they always yield a score of depen-
5 dence that ranges from low to high. Graded scales are perhaps effective
6 in the evaluation of tendencies for exercise dependence, but not in the
7 diagnosis of dependence. People scoring in the upper end of these scales
8 should be followed-up over time. This method will allow researchers to
9 identify the life-event(s) that trigger dependence.
40111
1
Secondary exercise dependence
2
3 De Coverley Veale (1987) differentiated between primary and secondary
4 exercise dependence. In the previous section primary dependence was
45111 examined. Secondary exercise dependence is a common characteristic of
144 Attila Szabo
1111 eating disorders such as anorexia nervosa and bulimia nervosa (De
2 Coverley Veale, 1987). In these disorders, excessive exercise is considered
3 to be an auxiliary feature used in caloric control and weight loss. Secondary
4 exercise dependence occurs in different ‘doses’ in people affected by eating
5 disorders. It has been estimated that one third of anorectics might be
6 affected (Crisp, Hsu, Harding, & Hartshorn, 1980).
7
8
The relationship between exercise and eating disorders
9
1011 A team of long-distance runners and scholars specialised in eating disor-
1 ders (Yates, Leehey, & Shisslak, 1983) observed a striking resemblance
2 between the psychology of anorectic patients and the very committed
3111 runners whom they named obligatory runners. They interviewed sixty
4 marathoners and closely examined the traits of a subgroup of male athletes
5 who corresponded to the ‘obligatory’ category. They reported that male
6 obligatory runners resembled anorexic women in some personality traits,
7 such as expression of anger, high self-expectation, tolerance of pain, and
8 depression as well as in some demographic details. Yates et al. (1983)
9 related these observations to a unique and hazardous way of establish-
20111 ment of self-identity. This work has marked the foundation of research
1 into the relationship between exercise and eating disorders.
2
3
The analogy between anorexia and excessive exercising
4
5 Since Yates et al. (1983) published their article, a large number of studies
6 have examined the relationship between exercise and eating disorders. A
7 close examination of these studies (Table 7.4) reveals some opposing find-
8 ings to the original report. For example, three studies that compared
9 anorectic patients with high level, or obligatory, exercisers (Blumenthal,
30111 O’Toole & Chang, 1984; Davis et al., 1995; Knight, Schocken, Powers,
1 Feld, & Smith, 1987) failed to demonstrate an analogy between anorexia
2 and excessive exercising. The differences in methodology between these
3 inquiries are, however, significant. They all looked for an analogy between
4 excessive exercise and anorexia, but from a different perspective.
5 Blumenthal et al. (1984) and Knight et al. (1987) examined a mixed gender
6 sample’s scores on a popular personality test (the Minnesota Multiphasic
7 Personality Inventory – MMPI). Davis et al. (1995) tested a female sample
8 using specific questionnaires aimed at assessing compulsiveness, commit-
9 ment to exercise, and eating disorders. Finally, Yates et al. (1983) looked
40111 to some demographic and personality parallels between obligatory runners
1 and anorectic patients. Further, the classification of the exercise behav-
2 iour may have differed in these studies. Therefore, these studies are not
3 easily comparable.
4 The controversy between the above studies may be partly solved by
45111 considering the results of a more recent study. Wolf and Akamatsu (1994)
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
Table 7.4 Summary table of published papers on physical activity and eating disorders relationships
Author(s) Participants Objectives Measurements Conclusion about the

relationship between
exercise and eating
disorders
Blumenthal, compared 24 anorectics to assess the similarity Minnesota Multiphasic Runners and anorectics
O’Toole, & to 43 obligatory runners between anorexia nervosa Personality Inventory are different. The
Chang (1984) and obligatory running (MMPI); Clinical diagnosis relationship is superficial
based on the DSM II and on the basis of the ten
the DSM III (Diagnostic and subscales of the MMPI
Statistical Manual of Mental
Disorders)
Brewerton, 110 anorexic, bulimic or to compare compulsively Diagnostic Survey of the Compulsive exercising
Stellefson, both females grouped exercising and non- Eating Disorders; was related to elevated
Hibbs, Hodges, into compulsive (n = 31) exercising patients Clinical diagnosis based on body dissatisfaction in
& Cochrane and non-compulsive suffering from eating the DSM III for anorexia patients with eating
(1995) exercise (n = 79) groups disorders nervosa and bulimia nervosa disorders and it was more
prevalent (39%) in
anorectics than in
bulimics (23%)
Davis (1990a) 86 exercising and 72 to compare body image Eysenk Personality Inventory; Body-dissatisfaction was
non-exercising women and weight preoccupation Body Image Questionnaire; related to poorer
between exercising and and Subjective Body Shape; emotional well-being in
non-exercising women Eating Disorder Inventory the exercise group only.
(EDI) EDI scores did not
differ between the groups
Davis (1990b) 53 exercising and 43 to study addictiveness, Addictiveness with the Eysenk Addictiveness was related
non-exercising women weight preoccupation, and Personality Questionnaire to weight and dieting
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111

exercise and eating
disorders
exercise patterns in a non- (EPQ); Body Focus; Eating variables in both groups
clinical population Disorder Inventory (EDI) and to perfectionism in
the exercise group. EDI
scores did not differ
between the groups
Davis, Brewer, 88 men and 97 women to present a new Addictiveness; Commitment to Presents validity and
& Ratusny ‘Commitment to Exercise’ Obsessive compulsiveness; reliability data for the two
(1993) questionnaire and to study Eating Disorder Inventory factor (obligatory
the relationship between (EDI), ‘Drive for Thinness’ exercising and
exercising and obsessive subscale pathological exercising)
compulsiveness, weight ‘Commitment to Exercise
preoccupation and Questionnaire’. Excessive
addictiveness exercising was found to
be distinct from eating
disorders
Davis et al. 46 anorexic patients, to test the relationship Commitment to exercise, Weight preoccupation
(1995) 76 high-level exercisers, between obsessive Obsessive compulsiveness; and excessive exercising
55 moderate exercisers, compulsiveness and Eating Disorder Inventory were related in both
all females exercise in anorectics in (EDI), ‘Drive for Thinness’ high-level exercisers and
contrast to moderate and subscale anorectics
high-level exercising
controls
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
French, Perry, 852 female students to observe changes, over Negative emotionality; Dieting habits were not
Leon, & a three-year period, in Self-concept; Eating related to physical
Fulkerson (1995) psychological and health Disorders Symptom Scores activity levels over three
variables in dieting and (based on DSM III), years, but dieters
non-dieting women Restrained Eating Scale, reported greater decreases
Eating Disorder Inventory in physical activity than
non-dieters
French, Perry, 1494 adolescents to examine correlates of Food preference and eating High-performance sport
Leon, & symptoms of eating patterns questionnaires and participation was found to
Fulkerson (1994) disorders, including food Eating Disorders Symptoms be a predictor of eating
preferences, eating disorders symptoms
patterns, and physical
activity
Levine, Marcus, 77 females (44 assigned to examine the effects of Beck Depression Inventory; Binge-eating disorder was
& Moulton to regular walking & an exercise intervention in Eating Disorder examination successfully managed
(1996) 33 control) suffering the treatment of obese (a semi-structured clinical through a 24-week (aimed
from binge-eating women with binge-eating interview) to burn 1000 calories per
disorder disorder week) walking programme
Pasman & 90 participants (45 males to examine body image Obligatory Exercise Runners and weightlifters
Thompson & 45 females) equal in and eating disturbance in Questionnaire; Eating reported greater eating
(1988) three groups: obligatory obligatory runners and Disorders Inventory (EDI); disturbance than controls.
runners, obligatory weight-lifters and in Body Self-relations Females also reported
weight-lifters and sedentary controls Questionnaire (BSRQ) greater eating disturbance
sedentary controls than males
Richert & 345 students to examine the relationship Eating Attitude Test (EAT) Exercise was positively
Hummers between exercise pattern correlated with EAT
(1986) risk for eating disorders scores and participants
with relatively high EAT
scores showed a preference
for jogging
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111

exercise and eating
disorders
Szymanski & 66 female athletes and to examine the Bem Sex-Role Inventory; Athletes scored higher on
Chrisler 20 non-athletes relationships between Eating Disorder Inventory most subscales of the EDI
(1990) eating disorders, gender (EDI) than non-athletes
roles, and athletic activity
Thiel, Gottfried, 84 low-weight male to study the prevalence of Eating Disorder Inventory 52% of the athletes
& Hesse athletes (25 wrestlers eating disorders in male (EDI) reported binging and 11%
(1993) & 59 rowers) athletes who, by the nature of the respondents
of their sport, are pressured evinced subclinical eating
to maintain low weight disorders. Concludes that
low-weight wrestlers
and rowers should be
considered at risk for
eating disorders
Williamson 98 female college to study the risk factors Social Influence; Sports Validated a psychosocial
et al. (1995) athletes involved in the Competition Anxiety model of risk factors
development of eating Test (SCAT); Athletic for the development of
disorders in female college self-appraisal; Interview for eating disorders in female
athletes Diagnosis of Eating Disorders college athletes. The
model suggests that social
influence, performance
anxiety and self-appraisal
together influence
body-size concern which
in turn is a strong
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
3
2
1
9
8
7
6
5
4
2
1
9
8
7
6
5
4
3
2
3111
1011
1111
45111
40111
30111
20111
determinant of eating
disorder symptoms
Wolf & 120 male and 168 female to study the relationship Eating Disorder Inventory Women involved in
Akamatsu students classified as between exercise and (EDI); Eating Attitude Test athletics demonstrated
(1994) 159 exercisers and eating disorders in college (EAT) more anorectic/bulimic
129 non-exercisers students attitudes and greater
weight preoccupation than
non-exercising women but
they did not manifest the
same personality
characteristics as female
non-exercisers with the
same level of eating
disorder
Yates, Leehey, 60 male long-distance or to study the similarity Interview Found a strong
& Shissiak trail runners between obligatory running resemblance between the
(1983) and anorexia nervosa characteristics of
obligatory running and
anorexia nervosa and,
thus, marked the interest
in further exploration of
the relationship
150 Attila Szabo
1111 studied female athletes who exhibited tendencies for eating disorders.
2 These females, however, did not manifest the personality characteristics
3 associated with eating disorders. Thus, in agreement with Blumenthal et
4 al.’s (1984) and Knight et al.’s (1987) explanation, differences between
5 obligatory exercisers and anorectic patients may outweigh substantially
6 the similarities reported by Yates et al. (1983). In a more recent theo-
7 retical article, Yates, Shisslak, Crago and Allender (1994) also admit that
8 the comparison of excessive exercisers with eating-disordered patients is
9 erroneous because the two populations are significantly different.
1011
1
Prevalence of eating disorder symptoms in exercisers and
2
non-exercisers
3111
4 Davis (1990a, 1990b) and Davis, Brewer and Ratusny (1993) conducted
5 a series of studies (see Table 7.4) in which they examined exercising and
6 non-exercising individuals and their tendency for eating disorders. In none
7 of these studies was exercise behaviour clearly related to eating disorders.
8 Opposing these conclusions are the results reported by French, Perry,
9 Leon and Fulkerson (1994), Pasman and Thompson (1988), Richert and
20111 Hummers (1986), Szymanski and Chrisler (1990), and Wolf and Akamatsu
1 (1994). Because similar measurements were used in general, the discrep-
2 ancy between the two sets of studies may be most closely related to the
3 definition of exercise. In the latter set of studies either excessive exer-
4 cisers or athletes were tested in contrast to those tested in the first set.
5 However, the definition of ‘excessive exercise’ needs to be standardised
6 in research. Four factors, including mode, frequency, intensity, and dura-
7 tion, must be reported otherwise it is unclear what ‘excessive exercise’ or
8 ‘athlete’ means. Reporting only one or two exercise parameter(s) is often
9 insufficient, especially in studies dealing with eating disorders because the
30111 latter is suspected to occur only in a very limited segment of the physi-
1 cally active population.
2 The majority of the reviewed studies (Table 7.4) suggest that high level
3 of exercise or athleticism is associated with symptoms of eating disorders.
4 The determinants of this relationship are not well known. Recently,
5 Williamson et al. (1995) proposed a psychosocial model for the develop-
6 ment of eating disorder symptoms in female athletes (see Figure 7.1). The
7 authors revealed that over-concern with body size, that was mediated
8 by social influence for thinness, anxiety about athletic performance, and
9 negative appraisal of athletic achievement, was a primary and strong deter-
40111 minant of the aetiology of eating disorder symptoms. This model should
1 be given serious consideration in the future and tested in several segments
2 of the exercising population.
3 Although women are at higher risk for developing eating disorders
4 (Yates et al., 1994), male athletes may be at risk too. For example, Thiel,
45111 Gottfried and Hesse (1993) reported a high frequency of eating-disorder
1111
2
3
4 Social Athletic
5 Pressure for Performance
Thinness Anxiety
6
7 Athletic
8 Achievement
9 Self-appraisal
1011
1
2
3111
4
5
6
Body-concern
7
8
9
20111
1
2
3
4 Eating
5 Disorder
6 Symptoms
7
8
9
30111
Figure 7.1 A psychosocial model for the development of eating disorder symp-
1 toms in athletes (based on Williamson et al., 1995).
2
3
4 symptoms and even subclinical incidences of eating disorders in low weight
5 male wrestlers and rowers (Table 7.4). This report attracts attention to
6 the fact that in some sports (i.e. gymnastics, boxing, and wrestling), in
7 which weight maintenance is critical, athletes may be at high risk for devel-
8 oping eating disorders. Athletes in these sports may turn to often
9 ‘unhealthy’ weight control methods (Enns, Drewnowski, & Grinker, 1987).
40111 This high-risk population, however, has received little attention in the
1 literature. In the future more research should be aimed at this segment
2 of the athletic population.
3 The relationship between physical activity and eating disorders is not
4 always negative. It is wrong to assume that exercise is directly related to eat-
45111 ing disorders. Only a very small segment of the physically active population
152 Attila Szabo
1111 is affected negatively. One study, purposefully included in Table 7.4, has
2 used physical activity successfully as a means of treatment for eating disor-
3 ders. Levine, Marcus and Moulton (1996) have shown that a simple walking
4 regimen, performed three to five times a week and aimed to expend 1,000
5 kcal, was efficient in managing binge-eating disorders in a clinically diag-
6 nosed sample of obese women. In fact, about 71% of the experimental group
7 was abstinent from binge eating by the end of the 6-month study.
8
9
Is excessive physical activity the cause or the consequence of
1011
eating disorders?
1
2 In view of De Coverley Veale’s (1987) classification of ‘secondary exer-
3111 cise dependence’ excessive exercise is a consequence of eating disorders.
4 In these conditions exercise is used as a means for decreasing body weight
5 (Blumenthal, Rose, & Chang, 1985; De Coverley Veale, 1987). However,
6 Davis (1990a) argues that exercise may foster a higher degree of body
7 narcissism and a distorted perception of one’s body size which, in turn,
8 may trigger eating disturbances. She suggests that it may be inappropriate
9 to perceive exercise simply as the consequence of eating disorders. Indeed
20111 some exercisers may resort to dieting for the sake of better performance
1 (De Coverley Veale, 1987). However, to date there is insufficient evidence
2 to claim that exercise may be a contributing factor to eating disorders.
3 Therefore, the hypothesis that was proposed by Davis (1990a) needs
4 further scrutiny. The model proposed by Williamson et al. (1995) might
5 be a valuable starting point in future studies.
6
7
Recommendations for future research
8
9 The majority of studies on the relationship between exercise and eating
30111 disorders have no conceptual foundation. Therefore, future studies need
1 to use psychosocial models, such as that proposed by Williamson et al.
2 (1995), to test the relationship and causality between physical activity and
3 eating disorders. A clear definition of what is meant by excessive exercise
4 or high-level exercise or athlete must be presented to allow for compara-
5 bility with other studies. Longitudinal studies that monitor both exercise
6 behaviour and eating habits, along with psychological factors such as
7 anxiety, self-concept, body-image or body-concern, may be the most
8 promising in the quest for a clearer understanding of the relationship
9 between exercise and eating disorders.
40111
1
General conclusions
2
3 Physical exercise in moderation seldom carries negative consequences. In
4 most cases of psychological dysfunction associated with exercise behav-
45111 iour, physical activity is a means of coping with emotional problems. The
1111 coping mechanism, whether exercise, alcohol, or medication can be abused.
2 The abuse of the former is rare because there is physical effort involved,
3 in contrast to the latter two (Cockerill & Riddington, 1996). Thus genuine
4 exercise dependence is extremely rare in the exercising population (Morris,
5 1989). People affected by eating disorders often use exercise as a means
6 of weight control. Some correlates of excessive exercise are anxiety, low
7 self-esteem, and long-term fidelity to the activity, as well as distorted body
8 image in some cases of eating disorders. Generally, excessive physical
9 activity is not a cause of psychological dysfunction, but rather a symptom
1011 of the latter.
1
2
3111
4 What we know
5 • Exercise dependence should not be equated with commitment
6 to exercise.
7 • Exercise dependence is best understood in runners since other
8 exercisers have seldom been examined.
9 • Exercise dependence is very rare.
20111 • A significant proportion of people suffering from eating disorders
1 resort to high levels of physical activity to lose weight.
2 • The personality characteristics of anorectics and highly committed
3 exercisers are significantly different.
4 • A relationship between exercise and eating disorders is evident
5 in athletic populations, particularly high-level exercisers and
6 professional athletes.
7 • Female athletes, and those in sports participated in within weight
8 categories, are at greater risk than other athletes of developing
9 eating disorders.
30111
1
2
3
5
6 • What factors cause exercise dependence?
7 • What is the role of exercise history, anxiety, and self-esteem in
8 exercise dependence?
9 • To what extent is excessive exercising a consequence of eating
40111 disorders?
1 • Could exercise have a positive effect on some eating disorders?
2 • What are the relationships between aspects of athleticism, body
3 image/concern, exercise and eating disorders?
4 • What is the risk of male athletes developing eating disorders?
45111
1111
2 8 The way forward for physical
3
4
activity and the promotion of
5 psychological well-being
6
7
8
Stuart J.H. Biddle, Kenneth R. Fox,
9 Stephen H. Boutcher and Guy E. Faulkner
1011
1
2
3111
4 This volume has covered key aspects of the literature on physical activity
5 and psychological well-being – anxiety, depression, affect and mood, self-
6 esteem, cognitive functioning, and psychological dysfunction. Space and
7 time limitations have precluded coverage of sleep, wider aspects of quality
8 of life, or certain mental illnesses.
9 The purpose of the review was to draw evidence-based conclusions for
20111 health professionals and to inform the development of health policy. This
1 final chapter will summarise the evidence in each of the areas addressed,
2 consider some limitations of the existing literature, summarise what we
3 need to know, and finally, address key issues for the design and delivery
4 of physical activity for the promotion of psychological well-being.
5
6
Summary of research findings
7
8 In this section, we summarise the key findings from each research area
9 examined in the book. Much of this will be based on the statements in
30111 each chapter that represent ‘what we know’ about the field. These have
1 been arrived at either through clear existing evidence or are based on our
2 best judgement at this time.
3
4
Anxiety and stress
5
6 The importance of tackling anxiety and stress-related disorders is best
7 highlighted by the huge cost to industry and the NHS, as well as to the
8 quality of life of individuals. In addition, stress can be linked to many of
9 the current political priorities in health, such as coronary heart disease
40111 (CHD), mental illness and accidents.
1 The area of anxiety and stress is complex. Physical activity interven-
2 tions might affect immediate anxiety feelings (state anxiety), relatively
3 stable anxiety characteristics of the individual (trait anxiety), or psycho-
4 physiological markers of anxiety, such as blood pressure or heart rate.
45111 When taking into account such complexity, some might argue that the
The way forward 155
1111 consistency of findings is impressive, providing convincing evidence of an
2 important effect for physical activity. This said, the evidence shows rather
3 low effects in terms of strength when viewed overall. Meta-analytic effect
4 sizes typically are in the low-to-moderate range (about 0.25–0.50). How-
5 ever, this is achieved across diverse methods using different measures of
6 anxiety and physical activity. In addition, stronger effects appear in what
7 are considered superior research designs, such as experimental trials. These
8 observations help strengthen the case for the anxiety-reduction effects
9 of activity. The key findings on physical activity, anxiety and stress are
1011 summarised below.
1
2
3111
4 Anxiety and stress: what we know
5 • Exercise has a low-to-moderate anxiety-reducing effect.
6 • Exercise training can reduce trait anxiety and single exercise
7 sessions can result in reductions in state anxiety.
8 • The strongest anxiety-reduction effects are shown in randomised
9 controlled trials.
20111 • Single sessions of moderate exercise can reduce short-term
1 physiological reactivity to and enhance recovery from brief
2 psychosocial stressors.
3
4
5
Depression
6
7 As with stress-related disorders, depression is a major health problem in the
8 UK and other Western countries. It represents a major economic cost to
9 the NHS and industry and leads to considerable human misery and suffer-
30111 ing (see Chapter 1). Common day-to-day depressive moods are addressed
1 in the section on mood and affect (see Chapter 4), while depression in this
2 text has been delimited to clinically defined depression (see Chapter 3).
3 Overall, the evidence is strong enough for us to conclude that there is
4 support for a causal link between physical activity and reduced clinically
5 defined depression. This is the first time such a statement has been made.
6 It is worth recalling that the first published statement on the causal link
7 between physical activity and CHD was only in 1987 (Powell, Thompson,
8 Caspersen, & Kendrick, 1987). We arrive at our conclusion by judging the
9 evidence against eight standard criteria in epidemiological research.
40111 Summarising, it is clear that the criteria of strength of association, consis-
1 tency of evidence from different populations and settings, the temporal
2 sequencing (of inactivity preceding depression), biological plausibility, and
3 experimental evidence support the view that (a) inactivity produces higher
4 risk of subsequent depression, and (b) physical activity can be an effective
45111 medium for its treatment. Summary findings are shown below.
156 Stuart Biddle et al.
1111
2 Clinically defined depression: what we know
3 • There is support for a causal link between exercise and
4 decreased depression.
5 • Epidemiological evidence has demonstrated that physical activity
6 is associated with a decreased risk of developing clinically
7 defined depression.
8 • Evidence from experimental studies shows that both aerobic
9 and resistance exercise may be used to treat moderate and more
1011 severe depression, usually as an adjunct to standard treatment.
1 • The anti-depressant effect of exercise can be of the same magni-
2 tude as that found for other psychotherapeutic interventions.
3111 • No negative effects of exercise have been noted in depressed
4 populations.
5
6
7
8 Emotion and mood
9 Although day-to-day moods and emotions may not be seen as a high
20111 priority in some clinical settings, we should not underestimate their impor-
1 tance in public health promotion. First, if people ‘feel good’ through
2 exercise, they are more likely to adhere to a physical activity programme
3 or active lifestyle; that is, enhanced mood acts as a positive reinforcer of
4 activity. This enables them to receive important protective physical health
5 benefits from exercise. Second, enhanced mood is an important part of
6 overall quality of life – something seen as increasingly important in health
7 promotion. Third, an emotionally satisfied individual is less likely to
8 become one of the ‘worried well’ who frequently present in primary care.
9 Evidence from diverse mood-related studies in physical activity is fairly
30111 consistent. Whether we draw on meta-analytic or narrative reviews, exper-
1 imental studies, or large-scale epidemiological surveys, evidence points to
2 a convincing relationship between physical activity and improved positive
3 mood. Key findings are summarised below.
4
5
6
Emotion and mood: what we know
7
• Physical activity and exercise have consistently been associated
8
with positive mood and affect.
9
• Meta-analytic evidence shows that aerobic exercise has a small-
40111
to-moderate effect on vigour (), tension (), depression (),
1
fatigue () and confusion (), and a small effect on anger ().
2
• A positive relationship between physical activity and psycho-
3
logical well-being has been confirmed in several large-scale
4
45111
The way forward 157
1111
2 epidemiological surveys, including in the UK, using different
3 measures of activity and well-being.
4 • Experimental trials support a positive effect for moderate inten-
5 sity exercise on psychological well-being.
6 • Meta-analytic evidence shows that adopting a goal in exercise
7 that is focused on personal improvement, effort, and mastery
8 has a moderate-to-high association with positive affect.
9 • Meta-analytic evidence shows that a group climate in exercise
1011 and sport settings that is focused on personal improvement and
1 effort has a moderate-to-high association with positive affect.
2
3111
4
5 Self-esteem
6 Self-esteem – itself widely recognised as a critical indicator of mental
7 health – is important in many contexts. An individual with high self-esteem
8 is more likely to be emotionally stable, and to cope better with life
9 demands, and is likely to be less dependent on support services. In addition,
20111 self-esteem and more specific self-perceptions can determine motivation
1 and drive individuals towards positive health behaviours.
2 Although the number of randomised controlled trials (RCTs) in this
3 area is small, the evidence does point to an effect for physical activity on
4 self-esteem, particularly in terms of physical self-perceptions. This is an
5 important finding in the light that physical self-perceptions, such as body
6 image and perceived physical self-worth, may carry important mental
7 health properties in their own right. Effects on generalised self-esteem
8 tend to be mixed, which is not surprising given the large array of life
9 opportunities for affecting the relatively stable quality of self-esteem.
30111 Summary evidence is shown below.
1
2
3
Self-esteem: what we know
4
• Exercise can be used as a medium to promote physical self-
5
worth and other important physical self-perceptions such as body
6
image. In some situations, this improvement is accompanied by
7
improved self-esteem.
8
• Physical self-worth carries mental well-being properties in its own
9
right and should be considered as a valuable end-point of exer-
40111
cise programmes.
1
• Positive effects of exercise on self-perceptions can be experi-
2
enced by all age groups but there is strongest evidence for
3
change for children and middle-aged adults.
4
45111
1111
2 • Positive effects of exercise on self-perceptions can be experi-
3 enced by men and women.
4 • Positive effects of exercise on self-perceptions are likely to be
5 greater for those with initially low self-esteem.
6 • Several types of exercise are effective in changing self-percep-
7 tions but there is most evidence to support aerobic exercise and
8 resistance training, with the latter indicating greatest effective-
9 ness in the short term.
1011
1
2
3111 Cognitive functioning
4
The population distribution progressively includes a higher percentage of
5
older people and the consequences of this trend are profound. If evidence
6
can demonstrate that physical activity affects the functional capacity and
7
quality of life of older adults, this will have important implications for the
8
way we promote health-enhancing activities and environments for older
9
people. Although there are several ways in which physical activity might
20111
benefit the older adult, this issue has largely been approached through
1
the effect of activity on the slowing or prevention of age-related decline
2
in cognitive functioning. Research in this area has remained rather
3
constrained in its approach and the evidence for a broad-based causal
4
effect from physical activity is weak. Summary findings are shown below.
5
6
7
Cognitive functioning: what we know
8
• The majority of cross-sectional studies show that fit older adults
9
display better cognitive performance than less fit older adults.
30111
• The association between fitness and cognitive performance is
1
task-dependent, with most pronounced effects in tasks that are
2
attention-demanding and rapid (e.g. reaction time tasks).
3
• Results of intervention studies are equivocal but meta-analytic
4
findings indicate a small but significant improvement in cognitive
5
functioning of older adults who experience an increase in aerobic
6
fitness.
7
8
9
40111
1 Psychological dysfunction
2 Although the benefits of physical activity are becoming increasingly well
3 documented, all sides of the argument must be addressed to arrive at a cred-
4 ible evidence base. For this reason, it is important to investigate whether
45111 exercise can have negative psychological effects in some individuals.
The way forward 159
1111 Indeed, the media have shown great interest over the past decade in the
2 issue of exercise ‘addiction’ (‘dependence’), and researchers have debated
3 whether exercise addiction and eating disorders are related. Summary
4 evidence is shown below and highlights that the negative effects of exercise
5 that have been shown in some individuals are rare and need not be seen as
6 an issue of concern for public health.
7
8
9 Psychological dysfunction: what we know
1011 • Exercise dependence is extremely rare.
1 • Many people suffering from eating disorders undertake high
2 levels of physical activity.
3111 • The personality characteristics of anorectics are significantly
4 different from highly committed exercisers.
5
6
7
8
Limitations of research on physical activity and
9
20111
1 In arriving at a summary of the evidence linking physical activity and
2 psychological well-being, all authors have drawn on a variety of sources
3 and methods, including narrative reviews, meta-analyses, experimental
4 designs, and large-scale surveys. Despite this, a common theme running
5 through all of the literature in this field is that strong research designs
6 are in the minority. Much of the research is in the form of small-scale
7 cross-sectional studies, or experiments without controls and/or randomi-
8 sation, the weaknesses of which are well known. The reasons for this are
9 not fully obvious, but it is probably due to the emergent status of the field
30111 of exercise science where much of the work has been conducted, and the
1 paucity of research funding. It is not our intention to dwell on the method-
2 ological weaknesses of this literature, however we consider it important
3 to provide some summary remarks regarding the general nature of research
4 in this area.
5
6
Measurement
7
8 It is clear from the studies reviewed that there is little consensus concern-
9 ing the measurement of physical activity and domains of psychological well-
40111 being. For example, studies often adopt differing methods for assessing
1 physical activity, such as different self-report scales, or different ‘objective’
2 measures such as heart rate monitors or movement sensors. In addition,
3 each area of psychological well-being calls on many different scales to
4 assess the construct of interest. For example, in self-esteem research, some
45111 researchers assess generalised self-esteem only, others assess physical
1111 domains of self-perceptions. Even within the latter approach, several
2 instruments have been used. In addition, the best method of assessing phys-
3 ical fitness in the area of physical activity and psychological well-being is
4 undetermined.
5
6
Research designs
7
8 A number of authors have highlighted the methodological inadequacy
9 of the cross-sectional approach (Boutcher, 1990; Folkins & Sime, 1981;
1011 Spirduso, 1994). Often non-equivalent groups have been studied and the
1 temporal effects of physical activity cannot be established. These types of
2 data are not likely to be accepted as admissable evidence of the effect
3111 of exercise on aspects of psychological well-being by health authorities
4 and fortunately is now rarely found in refereed literature.
5 However, even when an intervention approach is adopted, many studies
6 do not randomise to treatment and control groups. This is undoubtedly due
7 to the convenience of studying intact exercising groups. Clearly, however,
8 the results of such studies are likely to be influenced by self-selecting vari-
9 ables such as personality disposition, attitudes, and socio-economic back-
20111 ground (Boutcher, 1990). No less important is a problem facing many field
1 trials, even if there is randomisation. Those who finally accept participation
2 in a trial will have signed a consent form and may well be more inclined
3 towards exercise than those who refuse. Almost all findings are based on the
4 effect of exercise on those who volunteered for an exercise study.
5
6
Attrition
7
8 Unfortunately, attrition often occurs for non-exercisers starting a training
9 programme for the first time. Attrition itself may not be uniform among
30111 types of individuals within each group or between treatment and control
1 groups, often leading to non-equivalence by the end of the study. Such
2 effects are rarely reported or corrected through statistical treatment.
3 Conclusions are usually drawn upon those who have volunteered for the
4 study and who have remained in the study and results are usually not
5 presented on an intention-to-treat basis. The characteristics of drop-outs
6 are also rarely considered.
7
8
Exercise dose-response
9
40111 The issue of intensity and duration of exercise has important public health
1 implications yet remains poorly controlled in many studies. With the adop-
2 tion of the ‘moderate’ message for physical activity (see Killoran, Fentem,
3 & Caspersen, 1994), we need to identify the psychological effects of more
4 moderate forms of physical activity and also the effect of intermittent
45111 versus continuous exercise bouts. For example, Murphy and Hardman
The way forward 161
1111 (1998) have found similar improvements in fitness for previously seden-
2 tary women between those undertaking a brisk walking programme
3 involving one 30-minute walk per day and those doing a 10-minute walk
4 three times per day. Currently, because many reports have not clearly
5 stated the exercise mode and dosage, the effect of different regimes on
6 psychological well-being is not possible to determine.
7
8
Cost-effectiveness
9
1011 Research to date has focused on the efficacy of physical activity to improve
1 aspects of psychological well-being. The use of field trials to test the cost-
2 effectiveness of delivery of physical activity programmes within health
3111 service settings is rare. The language of economic analysis is notably absent
4 in the literature. For example, numbers-needed-to-treat statistics have not
5 been calculated. The use of physical activity either in the treatment of
6 mental disorders or the promotion of general well-being has rarely been
7 compared with alternative interventions. This is undoubtedly due to the
8 emergent state of the field of study, but is perhaps also due to its roots.
9 With the exception of the literature on clinically diagnosed populations,
20111 the research has been conducted by exercise and sport scientists whose
1 main objective until recently has been academic progress rather than health
2 service design and delivery.
3 When judged against the rigours of the evidence-based health move-
4 ment, where the randomised controlled trial is treated as the gold standard
5 and less credence is offered to other approaches, some areas of the liter-
6 ature appear weak or at best in their infancy. The body of knowledge on
7 clinical depression offers a clear exception to this. In addition, for an area
8 of research that has received relatively little government or commercial
9 sponsorship, the body of knowledge could be considered surprisingly
30111 convincing. Finally, the study of exercise and psychological well-being may
1 not lend itself quite so well to the randomised controlled trial design. It
2 is not possible to offer a placebo, ethical clearance requires the control
3 group to be offered some form of intervention, and the public nature of
4 exercise (versus medication, for example) means there is a strong chance
5 of contamination across groups. Furthermore, with clinically diagnosed
6 populations where there is limited access to limited subject numbers, RCTs
7 may not be feasible at all. This suggests a more flexible and forgiving
8 approach to the interpretation of the existing literature and the planning
9 for future research.
40111 Following our consideration of a number of different issues and method-
1 ological difficulties, we have summarised points requiring research
2 attention in the future. These are shown below and are grouped according
3 to measurement, different populations, exercise and physical activity pro-
4 gramming, economic issues and mechanisms of the relationship between
45111 physical activity and psychological well-being.
1111
3
4 Measurement
5 • Are current psychometric measures adequate for capturing the
6 range of affective responses in physical activity and for assessing
7 change over time?
8 • How much change in scores is necessary for a meaningful impact
9 on functioning, behaviours and well-being? To date, insufficient
1011 evidence is available in many areas to develop clinical criteria
1 and targets of change.
2
3111 Populations
4 • How do special groups (e.g. the obese with social physique
5 anxiety; asthmatics and chronic obstructive pulmonary disease
6 (COPD) patients who experience fears about breathing; older
7 people with a fear of falling) differ in the benefits of a programme
8 of exercise?
9 • Is the mental health effect from exercise the same across all
20111 ages and both genders?
1 • More information is needed on those who do not volunteer for
2 studies or who drop out and do not feature in the results.
3
4 Exercise and physical activity programming
5 • What are the long-term effects of accumulated doses of activity
6 (in line with current recommendations for physical activity for
7 cardiovascular disease prevention)?
8 • What are the longer term effects (i.e. over 4 months) of phys-
9 ical activity? We need to know, for example, whether a 10-week
30111 exercise programme will have lasting effects, and if not what is
1 necessary to maintain the effects?
2 • What are the effects of short bouts (< 15 minutes) of free-living,
3 unsupervised aerobic physical activity, which can be most easily
4 integrated into an active lifestyle, as a low-cost intervention?
5 • What are the social effects of exercise on mental well-being?
6 • We need to know whether exercise practitioner manipulations of
7 self-efficacy, outcome expectancy, perceived competence, goal
8 setting, feedback, attentional focus and perceived exertion and
9 enjoyment can have effects, particularly among inactive and inex-
40111 perienced exercisers.
1 • Adherence to exercise training appears to be greater when it is
2 of moderate intensity (e.g. walking), and integrated into an active
3 lifestyle. What are the determinants of adherence to free-living
4 and facility-based exercise programmes?
45111
The way forward 163
1111
2 • What are the competencies and skills required by exercise
3 professionals to most effectively promote physical activity for
4 psychological well-being? What is the role of other mental health
5 professionals?
6 • Are the psychological effects of physical activity the same for dif-
7 ferent modes of activity (e.g. aerobic, strength-based, flexibility-
8 based)?
9 • Do different intensities and durations of physical activity make
1011 a difference and do fitness levels modulate that effect?
1 • When might high intensity exercise produce positive affective
2 responses?
3111
4 Economic issues
5 • The cost-effectiveness of physical activity as a treatment for
6 mental health has not been considered. More studies need to
7 compare activity with other interventions, not only in terms of
8 mental health but also cost. Related to this would be careful
9 consideration of adherence to the respective interventions.
20111
1 Mechanisms
2 • How do the potential mechanisms underlying the effects interact?
3 • How do effects of exercise compare to those of drug treatments
4 and what adjunctive value does exercise have along with drug
5 treatment?
6 • If drugs are also administered is the interaction of drug and exer-
7 cise safe?
8 • Under what conditions are the associations between physical
9 activity and psychological well-being causal?
30111 • What mechanisms explain the link between activity and psycho-
1 logical well-being?
2 • We need information on the dynamics of change. Little is known
3 about how long it takes to produce changes and how long they
4 last.
5 • Is it necessary to develop physiological adaptations to an
6 exercise regimen before psychological well-being increases
7 occur?
8
9
40111
1
2
3
4
45111
1111 A key theme emerging from this review of ‘what we need to know’
2 concerns the mechanisms explaining the relationship between physical
3 activity and psychological well-being. Given that the identification of
4 underlying mechanisms for the explanation for why physical activity may
5 have beneficial effects is so important, some brief expansion of these points
6 is made here.
7
8
9 Explanatory mechanisms
1011 Table 8.1 illustrates some psychological and physiological mechanisms that
1 have the potential to influence psychological well-being (see Boutcher,
2 1993). This is an important issue and crucial for physical activity promo-
3111 tion, exercise prescription and intervention. For example, the mechanisms
4 underlying emotional change caused by a bout of physical activity may
5 be very different for the neophyte and the experienced jogger. The begin-
6 ning jogger may be more likely to be influenced by cognitive factors during
7 exercise whereas the experienced jogger is more likely to be influenced
8 by physiological factors, particularly as they may operate at different exer-
9 cise intensities by choice. The rationale here is that greater exposure to
20111 an exercise stimulus will cause greater physiological adaptation that in
1 turn will bring about a greater physiological influence on well-being.
2 Initially cognitive factors may influence affect but with greater physio-
3 logical adaptation to the exercise stimulus and increased activity at higher
4 percentages of maximum capacity, physiological mechanisms may also
5 become influential.
6 In the initial phase of starting to exercise, cognitive processes in the form
7 of attributions and self-efficacy may be especially important influences on
8 psychological well-being. Consequently, key determinants of mental health
9 status for the unfit, neophyte exerciser may be their expectations, level of
30111 self-efficacy, and the type of attributions and appraisals they make. This
1
2
3 Table 8.1 Psychological and physiological mechanisms that could underlie the
4 relationship between physical activity and mental health
5
6 Psychological Physiological
7
8 Time-out Endorphins
Mastery Biochemical changes
9
Confidence (e.g. catecholamines)
40111 Fellow exercisers’ characteristics Hyperthermic change
1 Exercise leader characteristics Autonomic changes
2 Exercise facilities Visceral feedback
3 Exercise environment
4 Self esteem
45111 Social support
The way forward 165
1111 suggestion is supported by research that has indicated that self-efficacy is
2 most influential during the adoption phase of exercise.
3 When the individual starts to physiologically adapt to exercise or is
4 exposed to repeated exercise, physiological mechanisms developed
5 through adaptation to regular exercise may play a more prominent role.
6 This suggestion is supported indirectly by past research that has indicated
7 that trained and untrained participants can possess different patterns
8 of emotional response to exercise stimuli. It is also feasible, however,
9 that individuals may exercise repeatedly but exercise intensity may not
1011 be great enough to bring about significant physiological adaptations. For
1 these individuals repeated exposure to light exercise may bring about
2 little physiological change but may result in behavioural conditioning to
3111 the characteristics of the exercise stimulus. For instance, a recreational
4 runner’s positive post-exercise psychological state may be generated
5 by the conditioning of pleasant cognitions by jogging in an attractive
6 park.
7 Throughout the 1990s, research into physical activity and psychological
8 well-being has expanded, quite often producing better quality research
9 than before. We also have a number of large-scale surveys to consult, as
20111 well as meta-analytic reviews of specific topics in this field. However,
1 despite this impressive research effort, we seem no closer to identifying
2 a clear set of mechanisms explaining why physical activity might have
3 effects on psychological well-being. Researchers must continue to attempt
4 to identify the mechanisms underlying this relationship. As authors in this
5 review have stated, the mechanisms explaining why physical activity might
6 be associated with different parameters of mental health require a great
7 deal of consideration. Readers are referred to Morgan (1997) for compre-
8 hensive coverage of this area.
9
30111
Key issues in the promotion of physical activity for
1
2
3 One purpose of this text is to provide a consensus on the evidence linking
4 physical activity and psychological well-being with a view to reaching
5 health professionals. With the growing evidence of the psychological bene-
6 fits of physical activity, as presented here, the increasing recognition of
7 the importance of addressing the mental health of the nation, as devel-
8 oped at length in the latest British government policy documents (DoH,
9 1999), it is opportune to consider the issue of physical activity interven-
40111 tions through health services. While there might be some acceptance of
1 the role of physical activity in psychological well-being by health profes-
2 sionals, such as through self-reports of enhanced mood from patients on
3 primary-care exercise schemes (Fox, Biddle, Edmunds, Bowler, & Killoran,
4 1997; Riddoch, Puig-Ribera, & Cooper, 1998), some real difficulties have
45111 yet to be overcome.
1111 Difficulties facing health professionals
2
Physical activity promotion for mental health is a new area for health
3
4 service professionals to consider. As with any new venture, barriers may
5 impede its progress. Firstly, there are attitudinal barriers that are related
6 to the status of physical activity and beliefs as to its efficacy in promoting
7 psychological well-being. On consideration of their review of the evidence
8 in the US, Tkachuk and Martin (1999) were surprised that ‘exercise’ had
9 not become a more popular treatment for certain mental health condi-
1011 tions. Similarly, in the UK there are likely to be factors that constrain the
1 development of physical activity as an accepted element of interventions
2 to promote psychological well-being.
3111 One key factor concerns dualistic tendencies to treat the mind and body
4 as separate entities. For example, McEntee and Halgin (1996, pp. 55, 58),
5 in their survey of psychotherapists’ use of exercise, concluded that ‘many
6 therapists simply do not see their work as pertaining to the body’ and
7 that ‘topics such as exercise are viewed as unimportant by some mental
8 health workers who fail to appreciate the relationship between physical
9 and psychological health’. This may be exacerbated by the increasingly
20111 narrow scope of specialisation developing within health services that
1 encourages differential diagnosis and treatment along physical or psycho-
2 logical lines.
3 Another subtle barrier concerns the nature of ‘exercise’ itself. Martinsen
4 and Stephens (1994) first suggested that the status of exercise interven-
5 tions was low in the field of psychiatry. Specifically, if one spent years
6 learning sophisticated techniques for treating clinical conditions then some
7 reluctance could be predicted if something as ‘simple’ as exercise was
8 suggested as having comparable treatment effects. This ‘simplicity’ also
9 invokes notions of a common sense approach to health. Intuitively, being
30111 physically active is good for psychological well-being. However, becoming
1 more active may be something that individuals can attempt themselves
2 without the help of mental health professionals. Exercise then becomes a
3 non-professional type of intervention and consequently may not be consid-
4 ered.
5 Such barriers may be more relevant when developing the role of phys-
6 ical activity in alleviating clinical conditions and less relevant in the general
7 promotion of psychological well-being. In relation to clinical conditions,
8 physical activity should not be seen as competing with other strategies for
9 promoting psychological well-being and rather as being an important addi-
40111 tion to a range of therapeutic options. Given the highlighted weaknesses
1 in the research such positioning is deserved. Further experimental studies
2 and the cost-effectiveness of physical activity interventions are undoubt-
3 edly needed but more importantly, future research findings must be
4 effectively disseminated outside the traditional boundaries of sport and
45111 exercise research fields.
The way forward 167
1111 In general, some of the systemic barriers faced by primary care profes-
2 sionals in promoting physical activity will certainly be applicable to the
3 current context. For example, McKenna, Naylor and McDowell (1998)
4 suggested time constraints, lack of incentive or reimbursement, lack of
5 standard protocols, lack of success in the counselling role, lack of appro-
6 priate training, and the absence of a coordinated and systematic daily
7 approach in practice operations were all influential.
8 These barriers may only exacerbate the relative absence of exercise
9 specialists in health care settings which limits the likelihood that physical
1011 activity promotion for mental health will be comprehensively adopted as
1 part of health improvement plans or activities of primary care groups.
2 Given the short protocols that are now being developed for exercise coun-
3111 selling, the shift to lifestyle physical activity as the promotional message
4 in addition to the growth of subsidised exercise referral schemes and other
5 leisure opportunities, such barriers can be reduced further through the
6 creative efforts of health care professionals. Specifically, greater partner-
7 ship will be essential between health and leisure services. Exercise
8 specialists, in particular, can play a more proactive role in creating links
9 with health care professionals in developing physical activity opportunities.
20111 It must be reiterated that physical activity should be promoted regard-
1 less of its impact on psychological well-being due to the associated physical
2 health benefits. However, this book highlights that there is sufficient
3 evidence to consider physical activity in promoting psychological well-
4 being. Further justification is now available for health professionals in
5 developing physical activity promotion schemes. Health professionals can
6 evaluate such efforts to not only strengthen the evidence but also inform
7 as to how physical activity is best promoted for psychological well-being,
8 of which little is presently known. We may have an emerging evidence
9 base, but not necessarily the evidence-based practice. To further this aim,
30111 research must also continue to develop accepted and standardised
1 measures of physical activity, and reach consensus on the most effective
2 types of instruments and indicators to measure mental health outcomes,
3 in order to assist the practitioner in implementing and evaluating practice.
4
5
Pointers for interventions
6
7 Interventions can be focused on mental health and psychological well-
8 being per se, such as dealing with patients with clinical depression, or they
9 can be focused on well-being as a determinant of adherence to an exer-
40111 cise programme. If exercise produces feelings of well-being, adherence is
1 likely to be enhanced. Mutrie (1999) provides a summary of issues asso-
2 ciated with exercise adherence for various clinical groups such as COPD,
3 cardiac rehabilitation, diabetes, cancers, HIV/AIDS, arthritis, osteoporosis,
4 and low back pain. Physical activity can have an important role to play
45111 for all of these conditions yet adopting and maintaining involvement is
1111 critical. If programmes are devised with psychological as well as physical
2 outcomes in mind, adherence should be enhanced. Programmes, there-
3 fore, will need an element of choice, a climate of self-improvement and
4 personal mastery, enjoyment, and will normally be of moderate intensity.
5 Direct interventions could involve a number of psychological condi-
6 tions, such as alcohol abuse and drug rehabilitation (Mutrie, 1999),
7 schizophrenia (Faulkner & Biddle, 1999; Faulkner & Sparkes, 1999), as
8 well as the areas addressed in this book. The most likely avenue for inter-
9 ventions will occur through the expanding number of GP referral schemes
1011 in the UK. GPs will be increasingly likely to refer patients to such
1 programmes to alleviate more common conditions such as depression and
2 anxiety. A recent development involves referrals from Community Mental
3111 Health Team (CMHT) members, particularly Community Psychiatric
4 Nurses (CPNs), onto existing GP referral programmes or programmes and
5 services established specifically for mental health service clients. Initiatives,
6 similar to a brokerage service, are also developing that provide physical
7 activity and leisure opportunities for mental health clients in the commu-
8 nity. Finally, many in-patient and out-patient units provide exercise and
9 sporting activities for patients and clients. For further information on such
20111 initiatives, readers are referred to the Physical Activity and Mental Health:
1 National Consensus Statements and Guidelines for Practice document
2 produced in parallel with this research review (Grant, in press).
3 However, as discussed, barriers to implementation are still significant
4 and promoting physical activity to improve psychological well-being is far
5 from a nationally applied standard, nor is physical activity commonly
6 considered in developing care plans for mental health service clients. This
7 book has provided a realistic overview of the state of evidence for the
8 effect of physical activity on psychological well-being. Positive results have
9 been reported and some advice regarding the difficulties that need to be
30111 addressed for physical activity promotion to be put into practice has been
1 offered. Physical activity is a critical domain within public health and its
2 role in the promotion of psychological well-being is important. It demands
3 serious consideration in contemporary health promotion efforts.
4
5
6
7
8
9
40111
1
2
3
4
45111
1111
2 References
3
4
5
6
7
8
9
1011
1
2
3111 Abele, A., & Brehm, W. (1993). Mood effects of exercise versus sports games:
4 Findings and implications for well-being and health. International Review of
Health Psychology, 2, 53–80.
5
Abourez, T., & Toole, T. (1995). Effect of task complexity on the relationship
6
between physical fitness and reaction time in older women. Journal of Aging
7 and Physical Activity, 3, 251–260.
8 Aldana, S.G., Sutton, L.D., Jacobson, B.H., Quirk, M.G. (1996). Relationships
9 between leisure time, physical activity and perceived stress. Perceptual and Motor
20111 Skills, 82, 315–321.
1 Alfermann, D., & Stoll, O. (1995). Effects of physical exercise on self-concept and
2 subjective well-being: Results of field experiments. In R. Vanfraechem-Raway
3 & Y. Vanden Auweele (Eds), Proceedings of the 9th European Congress on
4 Sport Psychology, Part 1 (pp. 238–246). Brussels: Belgian Society of Sport
5 Psychology.
Altchiler, L., & Motta, R. (1994). Effects of aerobic and nonaerobic exercise on
6
anxiety, absenteeism, and job satisfaction. Journal of Clinical Psychology, 50,
7
829–840.
8 American College of Sports Medicine (ACSM)(4th Edition)(1991). Guidelines for
9 exercise testing and prescription (4th edn). Philadelphia: Lea & Febiger.
30111 American Psychiatric Association. (1980). Diagnostic and statistical manual (3rd
1 edn). Washington, DC: APA.
2 American Psychiatric Association (1994). Diagnostic and statistical manual of
3 mental disorders, IV. Washington DC: APA.
4 Ames, C. (1992). Achievement goals, motivational climate, and motivational
5 processes. In G.C. Roberts (Ed.), Motivation in sport and exercise (pp. 161–176).
6 Champaign, IL: Human Kinetics.
Anshel, M.H. (1991). A psycho-behavioral analysis of addicted versus non-addicted
7
male and female exercisers. Journal of Sport Behavior, 14, 145–154.
8
Anspaugh, D.J., Hunter, S., & Dignan, M. (1996). Risk factors for cardiovascular
9 disease among exercising versus nonexercising women. American Journal of
40111 Health Promotion, 10, 171–174.
1 Arito, H., & Oguri, M. (1990). Contingent negative variation and reaction time
2 of physically-trained subjects in simple and discriminative tasks. Industrial
3 Health, 28, 97–106.
4 Balogun, J.A. (1987). The interrelationships between measures of physical fitness
45111 and self-concept. Journal of Human Movement Studies, 13, 255–265.
170 References
1111 Bane, S., & McAuley, E. (1998). Body image and exercise. In J.L. Duda (Ed.),
2 Advances in sport and exercise psychology measurement (pp. 311–324).
3 Morgantown, WV: Fitness Information Technology.
4 Barry, A.J., Steinmetz, J.R., Page, H.F., Birkhead, N.C., & Rodahl, K. (1966). The
effects of physical conditioning on older individuals, II: Motor performance and
5
cognitive function. Journal of Gerontology, 21, 192–199.
6
Bartholomew, J.B., & Linder, D.E. (1998). State anxiety following resistance exer-
7 cise: The roles of gender and exercise intensity. Journal of Behavioral Medicine,
8 21, 205–219.
9 Bartlewski, P.P., Van Raalte, J.L., & Brewer, B.W. (1996). Effects of aerobic exer-
1011 cise on the social physique anxiety and body esteem of female college students.
1 Women in Sport and Physical Activity Journal, 5, 49–62.
2 Baylor, A.M., & Spirduso, W.W. (1988). Systematic aerobic exercise and compo-
3111 nents of reaction time in older women. Journal of Gerontology, 43, 121–126.
4 Beck, A.T., Ward, C.H., Mendelsohn, M., Mock, J., & Erbaugh, H. (1961). An
5 inventory for measuring depression. Archives of General Psychiatry, 4, 561–571.
Beesley, S., & Mutrie, N. (1997). Exercise is beneficial adjunctive treatment in
6
depression. Letter to Editor, British Medical Journal, 315, 1542.
7
Bennett, P., & Murphy, S. (1997). Psychology and health promotion. Buckingham:
8 Open University Press.
9 Ben-Shlomo, L.S., & Short, M.A. (1986). The effects of physical conditioning on
20111 selected dimensions of self-concept in sedentary females. Occupational Therapy
1 in Mental Health, 5, 27–46.
2 Ben-Yishai, Y., & Diller, L. (1973). Changing of atmospheric environment to
3 improve mental and behavioural function. New York State Journal of Medicine,
4 12, 2877–2880.
5 Berger, B. (1996). Psychological benefits of an active lifestyle: what we know and
6 what we need to know. Quest, 48, 330–353.
Berger, B.G., & McInman, A. (1993). Exercise and the quality of life. In R.N. Singer,
7
M. Murphey, & L.K. Tennant (Eds). Handbook of research on sport psychology
8
(pp. 729–760). New York: Macmillan.
9 Berger, B.G., & Owen, D.R. (1992). Preliminary analysis of a causal relationship
30111 between swimming and stress reduction: Intense exercise may negate the effects.
1 International Journal of Sport Psychology, 23, 70–85.
2 Berlin J.A., & Colditz, G.A. (1990). A meta-analysis of physical activity in the
3 prevention of coronary heart disease. American Journal of Epidemiology, 132,
4 612–628.
5 Biddle, S. (1995). Exercise and psychosocial health. Research Quarterly for Exercise
6 and Sport, 66, 292–297.
7 Biddle, S.J.H., & Mutrie, N. (1991). Psychology of physical activity and exercise:
A health-related perspective. London: Springer-Verlag.
8
Biddle, S., Sallis, J., & Cavill, N, (1998). Young and active? Young people and
9
health-enhancing physical activity: Evidence and implications. London: HEA.
40111 Bills, R.E., Vance E.L., & McLean, O.S. (1951). An index of adjustment and
1 values. Journal of Consulting Psychology, 15, 257–261.
2 Birren, J.E., Woods, A.M., & Williams, M.V. (1980). Behavioural slowing with
3 age: Causes, organisation, and consequences. In L.W. Poona (Ed.), Aging in the
4 1980s: Psychological issues (pp. 293–308). Washington, DC: American Psycho-
45111 logical Association.
References 171
1111 Blair, S.N., & Hardman, A. (1995). Introduction: Physical activity, health and well-
2 being – an international consensus conference. Research Quarterly for Exercise
3 and Sport, 66, ii.
4 Blair, S.N., Kohl, H.W., Paffenberger, R.S., Clark, D.G., Cooper, K.H., & Gibbons,
L.W. (1989). Physical fitness and all-cause mortality: A prospective study of
5
healthy men and women. Journal of the American Medical Association, 262,
6
2395–2401.
7 Blaney, J., Sothmann, M., Raff, H., Hart, B., & Horn, T. (1990). Impact of exercise
8 training on plasma adrenocorticotropin response to a well-learned vigilance task.
9 Psychoneuroendocrinology, 15, 453–62.
1011 Blumenthal, J.A., Emery, C.F. Madden, D.J., George, L.K., Coleman, E.,
1 Riddle, M.W., McKee, D.C., Reasoner, J., & Williams, R.S. (1989). Cardio-
2 vascular and behavioral effects of aerobic exercise training in healthy older men
3111 and women. Journal of Gerontology: Medical Sciences, 44, M147-M157.
4 Blumenthal, J.A., Emery, C.F., Madden, D.J., Schneibolk, S.S., Walsh-Riddle, M.W.,
5 George, L.K., McKee, D.C., Higgenbotham, M., Cobb, F.R., & Colman, R.E.
(1991). Long term effects of exercise on psychological functioning in older men and
6
women. Journal of Gerontology, 46, 352–361.
7
Blumenthal, J.A., Fredrikson, M., Kuhn, C., Ulmer, R.L., Walsh-Riddle, M., &
8 Applebaum, M. (1990). Aerobic exercise reduces levels of cardiovascular and
9 perceived stress in subjects without prior evidence of myocardial ischemia.
20111 American Journal of Cardiology, 65, 93–98.
1 Blumenthal, J.A., O’Toole, L.C., & Chang, J.L. (1984). Is running an analogue of
2 anorexia nervosa? Journal of the American Medical Association (JAMA), 252,
3 520–523.
4 Blumenthal, J.A., & Madden, D.J. (1988). Effects of aerobic exercise training, age, and
5 physical fitness on memory-search performance. Psychology and Aging, 3, 280–285.
6 Blumenthal, J.A., Rose, S., & Chang, J.L. (1985). Anorexia nervosa and exercise:
Implications from recent findings. Sports Medicine, 2, 237–247.
7
Boone, J.B, Jr., Probst, M.M., Rogers, M.W., & Berger, R. (1993). Post-exercise
8
hypotension reduces cardiovascular responses to stress. Journal of Hypertension,
9 11, 449–453.
30111 Borg, G.A.V. (1973) Perceived exertion: A note on history and methods. Medicine
1 & Science in Sport, 5, 190–193.
2 Bosscher, R.J. (1993). Running and mixed physical exercises with depressed psychi-
3 atric patients. International Journal of Sport Psychology, 24, 170–184.
4 Botwinick, J. (1973). Aging and Behaviour. New York: Springer.
5 Bouchard, C., Shephard, R.J., & Stephens, T. (Eds) (1994). Physical activity, fitness,
6 and health. Champaign, IL: Human Kinetics.
7 Boutcher, S.H. (1990). Aerobic fitness: Measurement and issues. Journal of Sport
and Exercise Psychology, 12, 235–247.
8
Boutcher, S. (1993). Emotion and aerobic exercise. In R.N. Singer, M. Murphey,
9
& L.K. Tennant (Eds), Handbook of research on sport psychology (pp. 799–814).
40111 New York: Macmillan.
1 Boutcher, S.H., Nugent, F.W. & Weltman, A.L. (1995). Heart rate response to
2 psychological stressors of individuals possessing resting bradycardia. Behavioural
3 Medicine, 21, 40–46.
4 Bowling, A. (1995). Measuring health: A review of quality of life measurement
45111 scales. Buckingham: Open University Press.
172 References
1111 Bozoian, S., & McAuley, E. (1994). Strength training effects on subjective well-
2 being and physical function in the elderly (Abs.). Medicine and Science in Sports
3 and Exercise, 26, S156.
4 Brandon, J.E., & Loftin, J.M. (1991). Relationship of fitness to depression, state
and trait anxiety, internal health locus of control, and self-control. Perceptual
5
and Motor Skills, 73, 563–568.
6
Brewer, B.W. (1993). The dark side of exercise and mental health. In S. Serpa
7 (Ed.), Proceedings: VIII World Congress of Sport Psychology (pp. 531–534).
8 Lisbon: International Society of Sport Psychology.
9 Brewerton, T.D., Stellefson, E.J., Hibbs, N., Hodges, E.L., & Cochrane, C.E.
1011 (1995). Comparison of eating disorder patients with and without compulsive
1 exercising. International Journal of Eating Disorders, 17, 413–416.
2 Brown, D.R. (1992). Physical activity, ageing, and psychological well-being: An
3111 overview of the research. Canadian Journal of Sports Science, 17, 185–193.
4 Brown, D.R., & Harrison, J.M. (1986). The effects of a strength training program
5 on the strength and self-concept of two female age groups. Research Quarterly
for Exercise and Sport, 57, 315–320.
6
Brown, D.R., Morgan, W.P., & Raglin, J.S. (1993). Effects of exercise and rest on
7
the state anxiety and blood pressure of physically challenged college students.
8 Journal of Sports Medicine and Physical Fitness, 33, 300–305.
9 Brown, D.R., Wang, Y., Ward, A., Ebbeling, C.B., Fortlage, L., Puleo, E., Benson,
20111 H., & Rippe, J.M. (1995). Chronic psychological effects of exercise and exer-
1 cise plus cognitive strategies. Medicine and Science in Sports and Exercise, 27,
2 765–775.
3 Brown, R.I.F. (1993). Some contributions of the study of gambling to the study
4 of other addictions. In W.R. Eadington & J.A. Cornelius (Eds), Gambling
5 behavior and problem gambling (pp. 241–272). Reno: University of Nevada
6 Press.
Brown, S.W., Welsh, M.C., Labbe, E.E., Vitulli, W.F., & Kulkarni, P. (1992).
7
Aerobic exercise in the psychological treatment of adolescents. Perceptual and
8
Motor Skills, 74, 555–560.
9 Buckworth, J., Dishman, R.K., & Cureton, K.J. (1994). Autonomic responses of
30111 women with parental hypertension: Effects of physical activity and fitness.
1 Hypertension, 24, 576–584.
2 Busse, E.W. (1969). Theories of aging. In E.W. Busse and E. Pfeiffer (Eds),
3 Behaviour and adaptation in later life (pp. 11–32). Boston: Little Brown.
4 Buxton, M.J., O’Hanlon, M., & Rushby, J. (1990). A new facility for the measure-
5 ment of health-related quality of life. Health Policy, 16, 199–208.
6 Buxton, M.J., O’Hanlon, M., & Rushby, J. (1992). EuroQol: A reply and reminder.
7 Health Policy, 20, 329–332.
Byrne, A., & Byrne, D.G. (1993). The effect of exercise on depression, anxiety
8
and other mood states. Journal of Psychosomatic Research, 37, 565–574.
9
Byrne, B.M. (1996). Measuring self-concept across the lifespan. Washington DC:
40111 American Psychological Association.
1 Calfas, K., & Cooper, D. (1996). Effect of a 5 week exercise training program on
2 self-worth among adolescent girls: A randomised controlled study. Medicine and
3 Science in Sports and Exercise, 28, S135.
4 Calfas, K. J., & Taylor, C. (1994). Effects of physical activity on psychological
45111 variables in adolescents. Pediatric Exercise Science, 6, 406–423.
References 173
1111 Calvo, M.G., Alamo, L., & Ramos, P. (1990). Test anxiety, motor performance
2 and learning: Attentional and somatic interference. Personality and Individual
3 Differences, 11, 29–38.
4 Calvo, M.G., Szabo, A., & Capafons, J. (1996). Anxiety and heart rate under
psychological stress: The effects of exercise training. Anxiety, Stress and Coping,
5
9, 321–337.
6
Camacho, T.C., Roberts, R.E., Lazarus, N.B., Kaplan, G.A., & Cohen, R.D. (1991).
7 Physical activity and depression: Evidence from the Alameda county study.
8 American Journal of Epidemiology, 134, 220–231.
9 Cameron, O.G., & Hudson, C.J. (1986). Influence of exercise on anxiety level in
1011 patients with anxiety disorders. Psychosomatics, 27, 720–723.
1 Campbell, D.T., & Stanley, J.C. (1963). Experimental and quasi-experimental
2 designs for research. Chicago: Rand McNally.
3111 Campbell, R.N. (1984). The new science: Self-esteem psychology. Lanham, MD:
4 University Press of America.
5 Carmack, M.A., & Martens, R. (1979). Measuring commitment to running: A survey
of runners’ attitudes and mental states. Journal of Sport Psychology, 1, 2542.
6
Carrieri-Kohlman, V., Gormley, J.M., Douglas, M.K., Paul, S.M., & Stulbarg, M.S.
7
(1996). Exercise training decreases dyspnea and the distress and anxiety asso-
8 ciated with it. Monitoring alone may be as effective as coaching. Chest, 110,
9 1526–1535.
20111 Carroll, D., Smith, G.D., Sheffield, D., Shipley, M.J., & Marmot, M.G. (1995).
1 Pressor reactions to psychological stress and prediction of future blood pres-
2 sure: Data from the Whitehall II study. British Medical Journal, 310, 771–776.
3 Chaouloff, F. (1997). The serotonin hypothesis. In W.P. Morgan (Ed.), Physical
4 activity and mental health (pp. 179–198). Washington, DC: Taylor & Francis.
5 Chapman, C.L., & De Castro, J.M. (1990). Running addiction: Measurement and
6 associated psychological characteristics. Journal of Sports Medicine and Physical
Fitness, 30, 283–290.
7
Chodzko-Zajko. W.J., & Moore, K.A. (l994). Physical fitness and cognitive func-
8
tioning in aging. Exercise and Sport Sciences Reviews, 22, 195–220.
9 Choi, P.Y.L., & Salmon, P. (1995). Stress responsivity in exercisers and non-exer-
30111 cisers during different phases of the menstrual cycle. Social Science and Medicine,
1 6, 769–777.
2 Clarkson-Smith, L., & Hartley, A.A. (1989). Relationships between physical exer-
3 cise and cognitive abilities in older adults. Psychology and Aging, 4, 183–189.
4 Claytor, R.P. (1991). Stress reactivity: Hemodynamic adjustments in trained and
5 untrained humans. Medicine and Science in Sports and Exercise, 23, 873–881.
6 Clore, G.L., Ortony, A., & Foss, M.A. (1987). The psychological foundations of
7 the affective lexicon. Journal of Personality and Social Psychology, 53, 751–766.
Cockerill, I.M., & Riddington, M.E. (1996). Exercise dependence and associated
8
disorders: A review. Counselling Psychology Quarterly, 9, 119–129.
9
Cocklin, J.C. (1989). The effects of physical fitness and body cathexis on self-
40111 concept change in women after aerobic conditioning. Doctoral dissertation,
1 Oklahoma State University, Dissertation Abstracts International, 49 (10), 4594–B.
2 Collingwood, T.R. (1972). The effects of physical training upon behaviour and
3 self-attitudes. Journal of Clinical Psychology, 28, 583–585.
4 Conboy, J.K. (1994). The effects of exercise withdrawal on mood states of runners.
45111 Journal of Sport Behavior, 17, 188–203.
174 References
1111 Cooper, C. & Cartwright, S. (1996). Mental health and stress in the workplace.
2 London: HMSO.
3 Coopersmith, S. (1967). The antecedents of self-esteem. San Francisco, CA: Freeman.
4 Cottman, C.W., & Holets, V.R. (1985). Structural changes at synapses with age:
Plasticity regeneration. In C.E. Finch & E.L. Schneider (Eds), Handbook of the
5
biology of aging (2nd edn) (pp. 617–644). New York: Nostrand & Rinehold.
6
Cox, R.H. (1985). Sport psychology: Concepts and applications. Dubuque, IA:
7 W.C. Brown.
8 Craft, L.L., & Landers, D.M. (1998). The effect of exercise on clinical depression
9 and depression resulting from mental illness: A meta-analysis. Journal of Sport
1011 and Exercise Psychology, 20, 339–357.
1 Cramer, S.R., Nieman, D.C., & Lee, J.W. (1991). The effects of moderate exercise
2 training on psychological well-being and mood state in women. Journal of
3111 Psychosomatic Research, 35, 437–49.
4 Crews D.J. & Landers, D.M. (1987). A meta-analytic review of aerobic fitness and
5 reactivity to psychosocial stressors. Medicine and Science in Sports and Exercise,
19 (Suppl.), S114–S120.
6
Crisp, A.H., Hsu, L.K.G., Harding, B., & Hartshorn, J. (1980). Clinical features
7
of anorexia nervosa: A study of a consecutive series of 102 female patients.
8 Journal of Psychosomatic Research, 24, 179–191.
9 Crocker, P.R., & Grozelle, C. (1991). Reducing induced state anxiety: Effects of
20111 acute aerobic exercise and autogenic relaxation. Journal of Sports Medicine and
1 Physical Fitness, 31, 277–282.
2 Crossman, J., Jamieson, J., & Henderson, L. (1987). Responses of competitive
3 athletes to lay-offs in training: Exercise addiction or psychological relief? Journal
4 of Sport Behavior, 10, 28–38.
5 Cusumano, J.A., & Robinson, S.E. (1992). The short term psychophysiological
6 effects of hatha yoga and progressive relaxation on female Japanese students.
Applied Psychology: An International Review, 42, 77–90.
7
Czajkowski, S.M., Hidelang, R.D., Dembroski, T.M., Mayerson, S.E., Parks, E.B.,
8
& Holland, J.C. (1990). Aerobic fitness, psychological characteristics, and cardio-
9 vascular reactivity to stress. Health Psychology, 9, 676–692.
30111 Davis, C. (1990a). Body image and weight preoccupation: A comparison between
1 exercising and non-exercising women. Appetite, 15, 13–21.
2 Davis, C. (1990b). Weight and diet preoccupation and addictiveness: The role of
3 exercise. Personality and Individual Differences, 11, 823–827.
4 Davis, C. (1997). Body image, exercise, and eating behaviors. In K.R. Fox (Ed.),
5 The physical self: From motivation to well-being (pp. 143–174). Champaign, IL:
6 Human Kinetics.
7 Davis, C., Brewer, H., & Ratusny, D. (1993). Behavioral frequency and psycho-
logical commitment: Necessary concepts in the study of excessive exercising.
8
Journal of Behavioral Medicine, 16, 611–628.
9
Davis, C., Kennedy, S.H., Ralevski, E., Dionne, M., Brewer, H., Neitzert, C., &
40111 Ratusny, D. (1995). Obsessive compulsiveness and physical activity in anorexia
1 nervosa and high-level exercising. Journal of Psychosomatic Research, 39,
2 967–976.
3 Davis, W.M. (1971). The effects of a cardiovascular conditioning program on
4 selected psychological responses of college males (Doctoral dissertation, Univer-
45111 sity of Oklahoma, 1970). Dissertation Abstracts International, 32, 221A.
References 175
1111 De Coverley Veale, D.M.W. (1987). Exercise dependence. British Journal of
2 Addiction, 82, 735–740.
3 de Geus, E.C.J., Lorenz, J.P., Van Doornen, L.J.P., de Visser, D.C., & Orlebeke,
4 J.F. (1990). Existing and training induced differences in aerobic fitness: Their
relationship to physiological response patterns during different types of stress.
5
Psychophysiology, 27, 457–478.
6
de Geus, E.J., Lorenz, J.P., Van Doornen, L.J., & Orlebeke, J.F. (1993). Regular
7 exercise and aerobic fitness in relation to psychological make-up and physio-
8 logical stress reactivity. Psychosomatic Medicine, 55, 347–363.
9 Del Rey, P. (1982). Effects of contextual interference on the memory of older
1011 females differing in levels of physical activity. Perceptual and Motor Skills, 55,
1 171–180.
2 Department of Health (1994a). Health of the nation: Key area handbook mental
3111 illness. London: HMSO.
4 Department of Health (1994b). More people, more active, more often. London:
5 Department of Health.
Department of Health. NHS Executive. (1996). Burdens of disease: a discussion
6
document. London: Department of Health.
7
Department of Health (1999). White Paper. Saving lives: Our healthier nation.
8 London: HMSO.
9 Derogatis, L.R. (1980). SCL-90: Administration, scoring and interpretation manual
20111 (Rev. edn). Baltimore, MD: Clinical Psychometrics Research Unit, John Hopkins
1 University School of Medicine.
2 Derogatis, L., Lipman, R., & Covi, L. (1973). The SCL-90: An outpatient psychi-
3 atric rating scale. Psychopharmacology Bulletin, 9, 13–28.
4 Desharnais, R., Jobin, J., Cote, C., Levesque, L., & Godin, G. (1993). Aerobic
5 exercise and the placebo effect. Psychosomatic Medicine, 55, 149–154.
6 Diener, E. (1984). Subjective well-being. Psychological Bulletin, 95, 542–575.
Dishman, R.K. (Ed.) (1988). Exercise adherence: Its impact on public health.
7
Champaign, IL: Human Kinetics.
8
Dishman, R.K. (1993). Exercise adherence. In R.N. Singer, M. Murphey, & L.K.
9 Tennant (Eds), Handbook of research on sport psychology (pp. 779–798). New
30111 York: Macmillan.
1 Dishman, R.K. (1995). Physical activity and public health: Mental health. Quest,
2 47, 362–385.
3 Dishman, R.K. (1997). The norepinephrine hypothesis. In W.P. Morgan (Ed.),
4 Physical activity and mental health (pp. 199–212). Washington, DC: Taylor &
5 Francis.
6 Dishman, R.K., Farquhar, R.P., & Cureton, K.J. (1994). Responses to preferred
7 intensities of exertion in men differing in activity levels. Medicine and Science
in Sports and Exercise, 26, 783–790.
8
Dixhoorn, J., Duivenvoorden, H.J., Pool, J., & Verhage, F. (1990). Psychic effects
9
of physical therapy after myocardial infarction. Journal of Psychosomatic
40111 Research, 34, 327–337.
1 Dixon, P., Heaton, J., Long, A., & Warburton, A. (1994). Reviewing and applying
2 the SF-36. Outcomes Briefing, 4, 3–25.
3 Doan, B.T., Plante, T.G., DiGregorio, M., & Manuel, G. (1995). The influence of
4 aerobic exercise and relaxation training on coping with test-taking anxiety.
45111 Anxiety, Stress, and Coping, 8, 101–111.
176 References
1111 Doan, R.E., & Scherman, A. (1987). The therapeutic effect of physical fitness on
2 measures of personality: A literature review. Journal of Counseling and
3 Development, 66, 28–36.
4 Donaghy, M.E., & Mutrie, N. (1998). A randomised controlled study to investi-
gate the effect of exercise on the physical self-perceptions of problem drinkers.
5
Physiotherapy, 84, 169.
6
Doyne, E.J., Ossip-Klein, D.J., Bowman, E., Osborn, K.M., McDougall-Wilson,
7 I.B., & Neimeyer, R.A. (1987). Running versus weight lifting in the treatment
8 of depression. Journal of Consulting and Clinical Psychology, 55, 748–754.
9 Duda, J.L. (1993). Goals: A social-cognitive approach to the study of achievement
1011 motivation in sport. In R.N. Singer, M. Murphey, & L.K. Tennant (Eds).
1 Handbook of research on sport psychology (pp. 421–436). New York: Macmillan.
2 Dunn, A.L. & Dishman, R.K. (1991). Exercise and the neurobiology of depres-
3111 sion. Exercise and Sport Sciences Reviews, 19, 41–98.
4 Dustman, R.E., Emmerson, R.Y., Ruhling, R.O., Shearer, D.E., Steinhaus, L.A.,
5 Johnson, S.C., Bonekat, H.W., & Shigeoka, J.W. (1990). Age and fitness effects
on EEG, ERPs, visual sensitivity, and cognition. Neurobiology of Aging, 11,
6
193–200.
7
Dustman, R.E., Emmerson, R.Y., & Shearer, D.E. (1990). Electrophysiology and
8 aging: slowing, inhibition, and aerobic fitness. In M.L Howe, M.J. Stones, &
9 C.J. Brainerd (Eds), Cognitive and behavioural performance factors in atypical
20111 aging (pp. 103–149). New York: Springer-Verlag.
1 Dustman, R.E., Ruhling, R.O., Russell, E.M., Shearer, D.E., Bonekat, H.W.,
2 Shigeoka, J.W., Wood, J.S., & Bradford, D.C. (1984). Aerobic exercise training
3 and improved neuropsychological function of older individuals. Neurobiology
4 of Aging, 5, 35–42.
5 Elsayed, M., Ismail, A.H., & Young, R.J. (1980). Intellectual differences of adult
6 men related to age and physical fitness before and after an exercise program.
Journal of Gerontology, 35, 383–387.
7
Emery, C.F., & Gatz, M. (1990). Psychological and cognitive effects of an exercise
8
program for community-residing older adults. The Gerontologist, 30, 184–188.
9 Enns, M.P., Drewnowski, A., & Grinker, J.A. (1987). Body composition, body-
30111 size estimation and attitudes toward eating in male college athletes. Psycho-
1 somatic Medicine, 49, 56–64.
2 Era, P. (1988). Sensory, psychomotor, and motor functions in men of different
3 ages. Scandinavian Journal of Social Medicine, 39 (Suppl.), 9–77.
4 Estok, P.J., & Rudy, E.B. (1986). Physical, psychosocial, menstrual changes/risks
5 and addiction in female marathon and nonmarathon runners. Health Care for
6 Women International, 7, 187–202.
7 Etnier, J.L., Salazar, W., Landers, D.M., Petruzzello, S.J., Han, M., & Nowell, P.
(1997). The influence of physical fitness and exercise upon cognitive function-
8
ing: A meta-analysis. Journal of Sport and Exercise Psychology, 19, 249–274.
9
Farmer, M., Locke, B., Moscicki, E., Dannenberg, A., Larson, D., & Radloff, L.
40111 (1988). Physical activity and depressive symptoms: The NHANES I epidemio-
1 logic follow-up study. American Journal of Epidemiology, 128, 1340–1351.
2 Faulkner, G., & Biddle, S.J.H. (1999). Exercise as therapy for schizophrenia:
3 A review. Journal of Mental Health, 8, 441–457.
4 Faulkner, G., & Sparkes, A.C. (1999). Exercise as therapy for schizophrenia: An
45111 ethnographic study. Journal of Sport and Exercise Psychology, 21, 52–69.
References 177
1111 Fillingim, R.B., & Blumenthal, J.A. (1993). Psychological effects of exercise among
2 the elderly. In P. Seraganian (Ed.), Exercise psychology (pp. 237–253). New
3 York: John Wiley.
4 Fisher, E., & Thompson, J.K. (1994). A comparative evaluation of cognitive-behav-
ioral therapy (CBT) versus exercise therapy (ET) for the treatment of body
5
image disturbance. Preliminary findings. Behavior Modification, 18, 171–85.
6
Fitts, W.H. (1965). Tennessee Self-Concept Scale: Manual. Los Angeles: Western
7 Psychological Services.
8 Floeter, M.K., & Greenough, W.T. (1979). Cerebrellar plasticity: Modification of
9 Purkinge cell structure by differential rearing in monkeys. Science, 206, 227–229.
1011 Flory, J.D., & Holmes, D.S. (1991). Effects of an acute bout of aerobic exercise
1 on cardiovascular and subjective cognitive work. Journal of Psychosomatic
2 Research, 35, 225–230.
3111 Folkins, C.H., & Sime, W.E. (1981). Physical fitness training and mental health.
4 American Psychologist, 36, 373–389.
5 Fox, K.R. (1990). The Physical Self-Perception Profile Manual. DeKalb, IL: Office
for Health Promotion, Northern Illinois University.
6
Fox. K.R. (1997). The physical self and processes in self-esteem development. In
7
K.R. Fox (Ed.), The physical self: From motivation to well-being (pp. 111–140).
8 Champaign, IL: Human Kinetics.
9 Fox, K.R. (1998). Advances in the measurement of the physical self. In J.L. Duda
20111 (Ed.), Advances in sport and exercise psychology measurement (pp. 295–310).
1 Morgantown, WV: Fitness Information Technology.
2 Fox, K., Biddle, S.J.H., Edmunds, L., Bowler, I., & Killoran, A. (1997). Physical
3 activity promotion through primary health care in England. British Journal of
4 General Practice, 47, 367–369.
5 Fox, K.R., & Corbin, C.B. (1989). The Physical Self-Perception Profile: Develop-
6 ment and preliminary validation. Journal of Sport and Exercise Psychology, 11,
408–430.
7
Fox, K.R., Page, A., Armstrong, N., & Kirby, B. (1994). Dietary restraint and self-
8
perceptions in early adolescence. Personality and Individual Differences, 17, 87–96.
9 Frederick, C.M., & Morrison, C.S. (1996). Social physique anxiety: personality
30111 constructs, motivations, exercise attitudes, and behaviors. Perceptual and Motor
1 Skills, 82, 963–72.
2 Freemantle, N., Long, A., Mason, J., Sheldon, T., Song, F., Watson, P., & Wilson,
3 C. (1993). The treatment of depression in primary care. Effective Health Care,
4 Issue 5.
5 Fremont, J., & Craighead, L. W. (1987). Aerobic exercise and cognitive therapy in
6 the treatment of dysphoric moods. Cognitive Therapy and Research, 11, 241–251.
7 French, S.A., Perry, C.L., Leon, G.R., & Fulkerson, J.A. (1994). Food preferences,
eating patterns, and physical activity among adolescents: Correlates of eating
8
disorders symptoms. Journal of Adolescent Health, 15, 286–294.
9
French, S.A., Perry, C.L., Leon, G.R., & Fulkerson, J.A. (1995). Changes in psycho-
40111 logical variables and health behaviors by dieting status over a three-year period
1 in a cohort of adolescent females. Journal of Adolescent Health, 16, 438–447.
2 French, S.A., Story, M., & Perry, C.L. (1995). Self-esteem and obesity in children
3 and adolescents: A literature review. Obesity Research, 3, 479–490.
4 Furst, D.M., & Germone, K. (1993). Negative addiction in male and female runners
45111 and exercisers. Perceptual and Motor Skills, 77, 192–194.
178 References
1111 Gabler-Halle, D., Halle, J.W., & Chung, Y.B. (1993). The effects of aerobic exer-
2 cise on psychological and behavioral variables of individuals with developmental
3 disabilities: A critical review. Research in Development Disabilities, 14, 359–386.
4 Gauvin, L. (1990). An experiential perspective on the motivational features of
exercise and lifestyle. Canadian Journal of Sport Sciences, 15, 51–58.
5
Gauvin, L., & Rejeski, W.J. (1993). The Exercise-Induced Feeling Inventory:
6
Development and initial validation. Journal of Sport and Exercise Psychology,
7 15, 403–423.
8 Gauvin, L., & Spence, J.C. (1996). Physical activity and psychological well-being:
9 Knowledge base, current issues, and caveats. Nutrition Reviews, 54, S53–65.
1011 Gauvin, L., & Spence, J.C. (1998). Measurement of exercise-induced changes in
1 feeling states, affect, mood, and emotions. In J.L. Duda (Ed.), Advances in sport
2 and exercise psychology measurement (pp. 325–336). Morgantown, WV: Fitness
3111 Information Technology.
4 Glasser, W. (1976). Positive Addiction. New York, NY: Harper & Row.
5 Gleser, J., & Mendelberg, H. (1990). Exercise and sport in mental health: A review
of the literature. Israel Journal of Psychiatry and Related Sciences, 27, 99–112.
6
Goldberg, D.P., Cooper, B., Eastwood, M.R., Kedward, H.B., & Shepherd, M.
7
(1970). A standardised psychiatric interview for use in community surveys.
8 British Journal of Preventative and Social Medicine, 24, 18–23.
9 Graham, R.E., Zelchner, A., Peacock, L.J., & Dishman, R.K. (1996). Bradycardia
20111 during baroreflex stimulation and active or passive stressor tasks. Cardio-
1 respiratory Fitness and Psychophysiology, 33, 566–575.
2 Grant, T. (Ed.)(in press). Physical activity and mental health: National consensus
3 statements and guidelines for practice. London: Health Education Authority.
4 Greist, J.H., Klein, M.H., Eischens, R.R., Faris, J.W., Gurman, A.S., & Morgan,
5 W.P. (1979). Running as a treatment for depression. Comprehensive Psychiatry,
6 20, 41–54.
Griffiths, M. (1997). Exercise addiction: A case study. Addiction Research, 5,
7
161–168.
8
Gruber, J.J. (1986). Physical activity and self-esteem development in children: A
9 meta-analysis. American Academy of Physical Education Papers, 19, 330–48.
30111 Hailey, B.J., & Bailey, L.A. (1982). Negative addiction in runners: A quantitative
1 approach. Journal of Sport Behavior, 5, 150–153.
2 Hale, A.S. (1997). ABC of mental health: Depression. British Medical Journal,
3 315, 43–46.
4 Hamilton, M. (1960). A rating scale for depression. Journal of Neurosurgical
5 Psychiatry, 23, 56–61.
6 Hannaford, C. (1984). The psychophysiological effects of a running program on
7 depression, self-esteem and anxiety. Dissertation Abstracts International, 44, 3527B.
Hardy, C.J., & Rejeski, W.J. (1989). Not what, but how one feels: The measurement
8
of affect during exercise. Journal of Sport and Exercise Psychology, 11, 304–317.
9
Hart, B. (1981). The effect of age and habitual activity on the fractionated compo-
40111 nents of resisted and unresisted response time. Medicine and Science in Sports
1 and Exercise, 13, 78.
2 Hart, E.A., Leary, M.R., & Rejeski, W.J. (1989). The measurement of social
3 physique anxiety. Journal of Sport and Exercise Psychology, 11, 94–104.
4 Hart, M.E., & Shay, C.T. (1964). Relationship between physical fitness and acad-
45111 emic success. Research Quarterly, 35, 443–445.
References 179
1111 Harter, S. (1988). Manual for the Self-Perception Profile for Adolescents. Denver,
2 CO: University of Denver.
3 Hasher, L., & Zacks, R.T. (1988). Automatic and effortful processes in memory.
4 Journal of Experimental Psychology, 108, 356–388.
Hawkins, H.L., Kramer, A.F., & Capaldi, D. (1992). Aging, exercise, and attention.
5
Psychology and Aging, 7, 643–653.
6
Head, A., Kendall, M.J., Ferner, R., & Eagles, C. (1996). Acute effects of beta
7 blockade and exercise on mood and anxiety. British Journal of Sports Medicine,
8 30, 238–242.
9 Health Education Authority (1995). Health update 5: Physical activity. London:
1011 Health Education Authority.
1 Hertzog, C.K. (1989). Influences of cognitive slowing on age differences in intel-
2 ligence. Developmental Psychology, 25, 636–651.
3111 Hertzog, C., Schaie, K.W., & Gribbin, K. (1978). Cardiovascular disease and
4 changes in intellectual functioning from middle to old age. Journal of Geron-
5 tology, 33, 872–883.
Hill, A.B. (1965). The environment and disease: Association or causation?
6
Proceedings of the Royal Society of Medicine, 58, 295–300.
7
Hill, R.D., Storandt, M., & Malley, M. (1993). The impact of long-term exercise
8 training on psychological function in older adults. Journal of Gerontology:
9 Psychological Sciences, 48, 12–17.
20111 Hill, S., Harries, U., & Popay, J. (1996). Is the short form 36 (SF-36) suitable for
1 routine health outcomes assessment in health care for older people? Evidence
2 from preliminary work in community based health services in England. Journal
3 of Epidemiology and Community Health, 50, 94–98.
4 Hilyer, J.C., & Mitchell, W. (1979). Effect of systematic physical fitness training
5 combined with counseling on the self-concept of college students. Journal of
6 Counseling Psychology, 26, 427–436.
Hilyer, J.C., Wilson, D.G., Dillon, C., Caro, L., Jenkins, C., Spencer, W.A., &
7
Booker, A. (1982). Physical fitness training and counseling treatment for youth
8
offenders. Journal of Counseling Psychology, 29, 292–303.
9 Hobson, M.L., & Rejeski, W.J. (1993). Does the dose of acute exercise mediate
30111 psychophysiological responses to mental stress? Journal of Sport and Exercise
1 Psychology, 15, 77–87.
2 Hoffmann, P. (1997). The endorphin hypothesis. In W.P. Morgan (Ed.), Physical
4 Houtmann, I.L.D., & Bakker, F.C. (1991). Individual differences in reactivity to
5 and coping with the stress of lecturing. Journal of Psychosomatic Research, 35,
6 11–24.
7 Hoyer, W.J., Labouvie, G.V., & Baltes, P.B. (1973). Modification of response
speed deficits and intellectual performance in the elderly. Human Development,
8
16, 233–242.
9
Hull, E., Young, S., & Ziegler, M. (1984). Aerobic fitness affects cardiovascular
40111 responses to stressors. Psychophysiology, 21, 353–360.
1 Hunt, S.M., McEwan, J., & McKenna, S.P. (1986). Measuring health status. London:
2 Croom Helm.
3 Hurst, D.F., Boswell, D.L., Boogard, S.E., & Watson, M.W. (1997). The rela-
4 tionship of self-esteem to the health-related behaviors of the patients of a primary
45111 care clinic. Archives of Family Medicine, 6, 67–70.
180 References
1111 Hutzler, Y., & Bar-Eli, M. (1993). Psychological benefits for sports for disabled
2 people: A review. Scandinavian Journal of Medicine and Science in Sports, 3,
3 217–228.
4 Ismail, A.H., & El-Naggar, A.M. (1981). Effect of exercise on cognitive processing
in adult men. Journal of Human Ergology, 10, 83–91.
5
Jambor, E.A., Rudisill, M.E., Weekes, E.M., & Michaud, T.J. (1994). Association
6
among fitness components, anxiety, and confidence following aerobic training
7 in aquarunning. Perceptual and Motor Skills, 78, 595–602.
8 Jenkinson, C., Layte, R., Coulter, A., & Wright, L. (1996). Evidence for the sensi-
9 tivity of the SF-36 health status measure to inequalities in health: Results from
1011 the Oxford Healthy Lifestyles Survey. Journal of Epidemiology and Community
1 Health, 50, 377–380.
2 Jex, S.M. (1991). The psychological benefits of exercise in work settings: A review,
3111 critique, and dispositional model. Work and Stress, 5, 133–147.
4 Jin, P. (1992). Efficacy of Tai Chi, brisk walking, meditation, and reading in
5 reducing mental and emotional stress. Journal of Psychosomatic Research, 36,
361–370.
6
Johnson, R. (1995). Exercise dependence: When runners don’t know when to quit.
7
Sports Medicine and Arthroscopy Review, 3, 267–273.
8 Johnston, B. (1970). A study of the relationships among self-concept, movement
9 concept and physical fitness and the effects of a physical conditioning program
20111 upon self-concept and movement concept (Doctoral dissertation, Florida State
1 University, 1969). Dissertation Abstracts International, 30, 5270A–5271A.
2 Kaplan, G.A., Roberts, R.E., Camacho, T.C., & Coyne, J.C. (1987). Psychosocial
3 predictors of depression. American Journal of Epidemiology, 125, 206–220.
4 Kennedy, R.S., Dunlap, W.P., Bandert, L.E., Smith, M.G., & Houston, C.S. (1989).
5 Cognitive performance in a simulated climb of Mount Everest: Operation
6 Everest II. Aviation and Space Environmental Medicine, 60, 99–104.
Killoran, A., Fentem, P., & Caspersen, C. (Eds) (1994). Moving on: International
7
perspectives on promoting physical activity. London: Health Education Authority.
8
King, B., & Cotes, J.E. (1989). Relation of lung function and exercise capacity to
9 mood and attitudes to health. Thorax, 44, 402–409.
30111 King, A.C., Taylor, C. & Haskell, W.L. (1993). Effects of differing intensities and
1 formats of 12 months of exercise training on psychological outcomes in older
2 adults. Health Psychology, 12, 292–300.
3 King, A.C., Taylor, C.B., Haskell, W.L., & DeBusk, R.F. (1989). Influence of
4 regular aerobic exercise on psychological health: A randomized, controlled trial
5 of healthy middle-aged adults. Health Psychology, 8, 305–324.
6 Klein, M.J., Griest, J.H., Gurman, A.S., Neimeyer, R.A., Lesser, D.P., Bushnell,
7 N.J., & Smith, R.E. (1985). A comparative outcome study of group psycho-
therapy vs. exercise treatments for depression. International Journal of Mental
8
Health, 13, 148–177.
9
Knight, P.O., Schocken, D.D., Powers, P.S., Feld, J., & Smith, J.T. (1987). Gender
40111 comparison in anorexia nervosa and obligate running (abstract). Medicine and
1 Science in Sports and Exercise, 19(2: Suppl.), S66.
2 Kobasa, S.C., Maddi, J.R., & Puccetti, M.C. (1982). Personality and exercise as
3 buffers in the stress-illness relationship. Journal of Behavioral Medicine, 5, 391–404.
4 Koltyn, K.F. (1997). The thermogenic hypothesis. In W.P. Morgan (Ed.), Physical
References 181
1111 Kubitz, K.A., & Landers, D.M. (1993). The effects of aerobic training on cardio-
2 vascular responses to mental stress: An examination of underlying mechanisms.
3 Journal of Sport and Exercise Psychology, 15, 326–337.
4 Kubitz, K.A., & Pothakos, K. (1997). Does aerobic exercise decrease brain acti-
vation? Journal of Sport and Exercise Psychology, 19, 291–301.
5
Kugler, J., Dimsdale, J.E., Hartley, L.H., & Sherwood, J. (1990). Hospital super-
6
vised vs. home exercise in cardiac rehabilitation: Effects on aerobic fitness,
7 anxiety, and depression. Archives of Physical Medicine and Rehabilitation, 71,
8 322–325.
9 Kugler, J., Seelbach, H., & Kruskemper, G.M. (1994). Effects of rehabilitation
1011 exercise programmes on anxiety and depression in coronary patients: A meta-
1 analysis. British Journal of Clinical Psychology, 33, 401–410.
2 Kurtz, Z. (Ed.) (1992). With health in mind. London: Action for Sick Children.
3111 La Forge, R. (1995). Exercise-associated mood alterations: A review of interac-
4 tive neurobiological mechanisms. Medicine, Exercise, Nutrition and Health, 4,
5 17–32.
Landers, D.M., & Petruzzello, S.J. (1994). Physical activity, fitness and anxiety. In
6
C. Bouchard, R.J. Shephard, & T. Stephens (Eds), Physical activity, fitness and
7
health: International proceeedings and consensus statement. Champaign, IL:
8 Human Kinetics.
9 LaPerriere, A.R., Antoni, M.H., Schneiderman, N., Ironson, G., Klimas, N., Caralis,
20111 P., & Fletcher, M.A. (1990). Exercise intervention attenuates emotional distress
1 and natural killer cell decrements following notification of positive serologic
2 status for HIV-1. Biofeedback and Self Regulation, 15, 229–242.
3 Lazarus, R.S. (1991). Emotion and adaptation. New York: Oxford University Press.
4 Leary, M.R. (1992). Self-presentational processes in exercise and sport. Journal
5 of Sport and Exercise Psychology, 14, 339–351.
6 Lefebvre, R., & Sanford, S. (1984). A multi-modal questionnaire for stress. Journal
of Human Stress, 11, 69–75.
7
Leith, L.M. (1994). Foundations of exercise and mental health. Morgantown, WV:
8
Fitness Information Technology.
9 Leith, L.M., & Taylor, A.H. (1990). Psychological aspects of exercise: A decade
30111 literature review. Journal of Sport Behavior, 13, 219–239.
1 Levine, M.D., Marcus, M.D., & Moulton, P. (1996). Exercise in the treatment of
2 binge eating disorder. International Journal of Eating Disorders, 19, 171–177.
3 Lirgg, C. (1991). Gender differences in self-confidence in physical activity: A
4 meta-analysis of recent studies. Journal of Sport and Exercise Psychology, 13,
5 294–310.
6 Lobstein, D.D., Ismail, A.H., & Rasmussen, C.L. (1989). Beta-endorphin and
7 components of emotionality discriminate between physically active and seden-
tary men. Biological Psychiatry, 26, 3–14.
8
Long, B.C. (1991). Physiological and psychological stress recovery of physically fit
9
and unfit women. Canadian Journal of Behavioural Science, 23, 53–65.
40111 Long, B.C., & Van Stavel, R. (1995). Effects of exercise training on anxiety: A
1 meta-analysis. Journal of Applied Sport Psychology, 7, 167–189.
2 Lubin, A. (1965). Adjective checklists for measurement of depression. Archives
3 of General Psychiatry, 12, 57–62.
4 MacMahon, J.R., & Gross, R.T. (1988). Physical fitness and psychosocial effects of
45111 aerobic exercise in delinquent adolescent males. Sports Medicine, 142, 1361–1366.
182 References
1111 Mactavish, J.B., & Searle, M.S. (1992). Older individuals with mental retardation
2 and the effect of a physical activity intervention on selected social psychological
3 variables. Therapeutic Recreation Journal, 26, 38–47.
4 Madden, D.J., Blumenthal, J.A., Allen, P.A., & Emery, C.F. (1989). Improving
aerobic capacity in healthy older adults does not necessarily lead to improved
5
cognitive performance. Psychology and Aging, 4, 307–320.
6
Markland, D., Emberton, M., & Tallon, R. (1997). Confirmatory factor analysis
7 of the Subjective Exercise Experiences Scale among children. Journal of Sport
8 and Exercise Psychology, 19, 418–433.
9 Marsh, H.W. (1992a). Self-Description Questionnaire II: Manual. Publication Unit,
1011 Faculty of Education, University of Western Sydney.
1 Marsh, H.W. (1992b). Self-Description Questionnaire (SDQ) III: A theoretical
2 and empirical basis for the measurement of multiple dimensions of late adoles-
3111 cent self-concept. An interim test manual and research monograph. Macarthur,
4 New South Wales, Australia: University of Western Sydney, Faculty of
5 Education.
Marsh, H.W. (1997). The measurement of physical self-concept: A construct vali-
6
dation approach. In K.R.Fox (Ed.), The physical self: From motivation to
7
well-being (pp. 27–58). Champaign, IL: Human Kinetics.
8 Marsh, H.W., & Peart, N.D. (1988). Competitive and co-operative physical fitness
9 training programs for girls: Effects on physical fitness and multidimensional self-
20111 concepts. Journal of Sport and Exercise Psychology, 10, 390–407.
1 Marsh, H.W., & Richards, G.E. (1988). The Tennesse Self-Concept Scales:
2 Reliability, internal structure, and construct validity. Journal of Personality and
3 Social Psychology, 55, 612–624.
4 Marsh, H.W., Richards, G.E., Johnson, S., Roche, L., & Tremayne, P. (1994).
5 Physical Self-Description Questionnaire: Psychometric properties and multitrait-
6 multimethod analysis of relations to existing instruments. Journal of Sport and
Exercise Psychology, 16, 270–305.
7
Martens, R., Vealey, R.S., & Burton, D. (1990). Competitive Anxiety in Sport.
8
9 Martinek, T.J., Cheffers, J.T., & Zaichkowsky, L.D. (1978). Physical activity, motor
30111 development, self-concept: Race and age differences. Perceptual and Motor
1 Skills, 26, 147–154.
2 Martinek, T.J., & Zaichkowsky, L.D. (1977). Manual and scale for the Martinek-
3 Zaichkowsky self-concept scale for children. Jacksonville, IL: Psychologists and
4 Educators.
5 Martinsen, E.W. (1989). The role of aerobic exercise in the treatment of depres-
6 sion. Stress Medicine, 3, 93–100.
7 Martinsen, E.W. (1993). Therapeutic implications of exercise for clinically anxious
and depressed patients. International Journal of Sport Psychology, 24, 185–199.
8
Martinsen, E.W. (1994). Physical activity and depression: Clinical experience. Acta
9
Psychiatrica Scandinavica, 377, 23–27.
40111 Martinsen, E.W., Hoffart, A., & Solberg, O. (1989). Comparing aerobic and non-
1 aerobic forms of exercise in the treatment of clinical depression: A randomized
2 trial. Comprehensive Psychiatry, 30, 324–331.
3 Martinsen, E.W., Medhus, A., & Sandvik, L. (1985). Effects of aerobic exercise
4 on depression: A controlled trial. British Medical Journal, 291, 100.
45111 Martinsen, E.W., Sandvik, L., & Kolbjornsrud, O.B. (1989). Aerobic exercise in
References 183
1111 the treatment of non-psychotic mental disorders: An exploratory study. Nord.
2 Psykiatr. Tidsskr., 44, 35–43.
3 Martinsen, E.W. & Stephens, T. (1994). Exercise and mental health in clinical and
4 free-living populations. In R.K. Dishman (Ed.), Advances in exercise adherence
(pp. 55–72). Champaign, IL: Human Kinetics.
5
Martinsen, E.W., Strand, J., Paulson, G., & Kaggestad, J. (1989). Physical fitness
6
level in patients with anxiety and depressive disorders. International Journal of
7 Sports Medicine, 10, 58–61.
8 McAuley, E. (1994). Physical activity and psychosocial outcomes. In C. Bouchard,
9 R.J. Shephard, & T. Stephens (Eds), Physical activity, fitness and health
1011 (pp. 551–568). Champaign, IL: Human Kinetics.
1 McAuley, E., Bane, S.M., Rudolph, D.L., & Lox, C.L. (1995). Physique anxiety
2 and exercise in middle-aged adults. Journal of Gerontology, 50B, 229–235.
3111 McAuley, E., & Courneya, K. (1994). The Subjective Exercise Experiences Scale
4 (SEES): Development and preliminary validation. Journal of Sport and Exercise
5 Psychology, 16, 163–177.
McAuley, E., Courneya, K.S. & Lettunich, J. (1991). Effects of acute and long-
6
term exercise on self-efficacy responses in sedentary, middle-aged males and
7
females. The Gerontologist, 31: 4, 534–542.
8 McAuley, E., & Mihalko, S.L. (1998). Measuring exercise-related self-efficacy. In
9 J.L. Duda (Ed.), Advances in sport and exercise psychology measurement
20111 (pp. 371–392). Morgantown, WV: Fitness Information Technology.
1 McAuley, E., Mihalko, S.L., & Bane, S.M. (1995). Exercise and self-esteem in
2 middle-aged adults: Multidimensional relationships and physical fitness and self-
3 efficacy influences. Journal of Behavioral Medicine, 20, 67–83.
4 McAuley, E., Mihalko, S.L., & Bane, S.M. (1996). Acute exercise and anxiety
5 reduction: Does the environment matter? Journal of Sport and Exercise
6 Psychology, 18, 408–419.
McAuley, E., & Rudolph, D. (1995). Physical activity, aging, and psychological
7
well-being. Journal of Aging and Physical Activity, 3, 67–96.
8
McCann, I.L., & Holmes, D. S. (1984). Influence of aerobic exercise on depression.
9 Journal of Personality and Social Psychology, 46, 1142–1147.
30111 McCarrick, A.K., Manderscheid, R.W., Bertolucci, D.E., Goldman, H., & Tessler,
1 R.C. (1986). Chronic medical problems in the chronic mentally ill. Hospital and
2 Community Psychiatry, 37, 289–291.
3 McCubbin, J.A., Cheung, R., Montgomery, T.B., Bulbulian, R., & Wilson, J.F.
4 (1992). Aerobic fitness and opioidergic inhibition of cardiovascular stress reac-
5 tivity. Psychophysiology, 29, 687–697.
6 McDonald, D.G., & Hodgdon, J.A. (1991). Psychological effects of aerobic fitness
7 training: Research and theory. New York: Springer-Verlag.
McEntee, D.J., & Halgin, R.P. (1996). Therapists’ attitudes about addressing the
8
role of exercise in psychotherapy. Journal of Clinical Psychology, 52, 48–60.
9
McFarland, R.A. (1963). Experimental evidence of the relationship between aging
40111 and oxygen want: In search of a theory of aging. Ergonomics, 6, 339–366.
1 McGowan, R.W., Jarman, B,O. & Pedersen, D.M. (1974). Effects of a competi-
2 tive endurance training program on self-concept and peer approval. Journal of
3 Psychology, 86, 57–60.
4 McKenna, J., Naylor, P-J., & McDowell, N. (1998). Barriers to physical activity pro-
45111 motion by GPs and practice nurses. British Journal of Sports Medicine, 32, 242–247.
184 References
1111 McNair, D.M., Lorr, M., & Droppleman, L.F. (1971). Manual for the Profile of
2 Mood States. San Diego, CA: Educational and Industrial Testing Service.
3 McNeil, J.K., LeBlanc, E.M., & Joyner, M. (1991). The effects of exercise on
4 depressive symptoms in the moderately depressed elderly. Psychology and
Aging, 6, 487–488.
5
Milligan, W.L., Powell, D.A., Harley, C., & Furchtgott, E. (1984). A comparison
6
of physical health and psychosocial variables as predictors of reaction time and
7 serial learning performance in elderly men. Journal of Gerontology, 39, 704–710.
8 Mock, V., Dow, K.H., Meares, C.J., Grimm, P.M., Dienemann, J.A., Haisfield-
9 Wolfe, M.E., Quitasol, W., Mitchell, S., Chakravarthy, A., & Gage, I. (1997).
1011 Effects of exercise on fatigue, physical functioning, and emotional distress during
1 radiation therapy for breast cancer. Oncology Nursing Forum, 24, 991–1000.
2 Molloy, D.W., Beerschoten, D.A., Borrie, M.J., Crilly, R.G., & Cape, R.D.T.
3111 (1988). Acute effects of exercise on neuropsychological function in elderly
4 subjects. Journal of the American Geriatrics Society, 36, 29–33.
5 Morgan, W.P. (1968). Selected physiological and psychomotor correlates of depres-
sion in psychiatric patients. Research Quarterly, 39, 1037–1043.
6
Morgan, W.P. (1969). A pilot investigation of physical working capacity in
7
depressed and non-depressed psychiatric males. Research Quarterly, 40, 859–861.
8 Morgan, W.P. (1970a). Physical fitness correlates of psychiatric hospitalization. In
9 G.S. Kenyon (Ed.), Contemporary psychology of sport (pp. 297–300). Chicago:
20111 Athletic Institute.
1 Morgan, W.P. (1970b). Physical working capacity in depressed and non-depressed
2 psychiatric females: A preliminary study. American Corrective Therapy Journal,
3 24, 14–16.
4 Morgan, W.P. (1979). Negative addiction in runners. The Physician and Sport-
5 medicine, 7, 57–63; 67–70.
6 Morgan, W.P. (1994). Physical activity, fitness and depression. In C. Bouchard,
R. J. Shephard, & T. Stephens (Eds), Physical activity, fitness, and health
7
(pp. 851–867). Champaign, IL: Human Kinetics.
8
Morgan, W.P. (Ed.) (1997). Physical activity and mental health. Washington, DC:
9 Taylor & Francis.
30111 Morgan, W.P., & Goldston, S.E. (Eds) (1987). Exercise and mental health.
1 Washington: Hemisphere.
2 Morgan, W.P., & Sonstroem, R.J. (1989). Exercise and self-esteem: Rationale and
3 model. Medicine and Science in Sports and Exercise, 21, 329–337.
4 Morris, M. (1989). Running round the clock. Running, 104, 44–45.
5 Morris, J. (1994). Exercise in the prevention of coronary heart disease: Today’s
6 best buy in public health. Medicine and Science in Sports and Exercise, 26,
7 807–814.
Moses, J., Steptoe, A., Mathews, A., & Edwards, S. (1989). The effects of exercise
8
training on mental well-being in the normal population: A controlled trial.
9
Journal of Psychosomatic Research, 33, 47–61.
40111 Moul, J.L., Goldman, B., & Warren, B. (1995). Physical activity and cognitive
1 performance in the older population. Journal of Aging and Physical Activity, 3,
2 135–145.
3 Muraki, S., Maehara, T., Ishii, K., Ajimoto, M., & Kikuchi, K. (1993). Gender
4 difference in the relationship between physical fitness and mental health. Annals
45111 of Physiological Anthropology, 12, 379–384.
References 185
1111 Murphy, G. (1947). Personality: A biosocial approach to origins and structure. New
2 York: Harper & Row.
3 Murphy, M.H., & Hardman, A.E. (1998). Training effects of short and long bouts
4 of brisk walking in sedentary women. Medicine and Science in Sports and
Exercise, 30, 152–157.
5
Mutrie, N. (1988). Exercise as a treatment for moderate depression in the UK
6
National Health Service. In Sport, Health, Psychology and Exercise Symposium
7 Proceedings (pp. 96–105). London: The Sports Council and Health Education
8 Authority.
9 Mutrie, N. (1997). The therapeutic effects of exercise on the self. In K.R. Fox
1011 (Ed.), The physical self: From motivation to well-being (pp. 287–314). Champaign,
1 IL: Human Kinetics.
2 Mutrie, N. (1999). Exercise adherence and clinical populations. In S.J. Bull
3111 (Ed.), Adherence issues in sport and exercise (pp. 75–109). Chichester: John
4 Wiley.
5 Mutrie, N., & Biddle, S.J.H. (1995). The effects of exercise on mental health in
non-clinical populations. In S.J.H. Biddle (Ed.), European perspectives on exer-
6
cise and sport psychology (pp. 50–70). Champaign, IL: Human Kinetics.
7
Mutrie, N., & Davison, R. (1994). Physical self-perceptions in exercising older
8 adults. Journal of Sports Sciences, 12, 203.
9 National Forum for Coronary Heart Disease Prevention (1995). Physical activity:
20111 An agenda for action. London: Health Education Authority.
1 Neal, R. (1977). Effect of group counseling and physical fitness programs on self-
2 esteem and cardiovascular fitness (Doctoral dissertation, Boston University,
3 1977). Dissertation Abstracts International, 38, 1911A–1912A.
4 NHS Health and Advisory Service (1995). Child and adolescent mental health
5 services. London: HMSO.
6 Norris, R., Carroll, D., & Cochrane, R. (1990). The effects of aerobic and anaerobic
training on fitness, blood pressure and psychological stress and well-being.
7
Journal of Psychosomatic Research, 34, 367–375.
8
Norris, R., Carroll, D., & Cochrane, R. (1992). The effects of physical activity and
9 exercise training on psychological stress and well-being in an adolescent popu-
30111 lation. Journal of Psychosomatic Research, 36, 55–65.
1 North, T.C., McCullagh, P., & Tran, Z.V. (1990). The effect of exercise on depres-
2 sion. Exercise and Sport Sciences Reviews, 18, 379–415.
3 Norvell, N., & Belles, D. (1993). Psychological and physical benefits of circuit
4 weight training in law enforcement personnel. Journal of Consulting and Clinical
5 Psychology, 61, 520–527.
6 Norvell, N., Martin, D., & Salamon, A. (1991). An examination of the psycho-
7 logical and physiological benefits of passive and aerobic exercise in sedentary
middle-aged women. Journal of Nervous and Mental disease, 147, 95–96.
8
Nouri, S., & Beer, J. (1989). Relations of moderate physical exercise to scores
9
on hostility, aggression, and trait-anxiety. Perceptual and Motor Skills, 68,
40111 1191–1194.
1 Ntoumanis, N., & Biddle, S.J.H. (1999a). Affect and achievement goals in phys-
2 ical activity: A meta-analysis. Scandinavian Journal of Medicine and Science in
3 Sports, 9, 315–332.
4 Ntoumanis, N., & Biddle, S.J.H. (1999b). A review of motivational climate in phys-
45111 ical activity. Journal of Sports Sciences, 17, 643–665.
186 References
1111 O’Connor, P.J., Bryant, C.X., Veltri, J.P., & Gebhardt, S.M. (1993). State anxiety
2 and ambulatory blood pressure following resistance exercise in females. Medicine
3 and Science in Sports and Exercise, 25, 516–521.
4 O’Connor, P.J., & Davis, J.C. (1992). Psychobiologic responses to exercise at
different times of day. Medicine and Science in Sports and Exercise, 24, 714–719.
5
O’Connor, P.J., Petruzzello, S.J., Kubitz, K.A., & Robinson, T.L. (1995). Anxiety
6
responses to maximal exercise testing. British Journal of Sports Medicine, 29,
7 97–102.
8 Offenbach, S.I., Chodko-Zajko, W.J., & Ringel, R.L. (1990). Relationship between
9 physiological status, cognition, and age in adult men. Bulletin of the Psychonomic
1011 Society, 28, 112–114.
1 Ojanen, M. (1994). Can the true effects of exercise on psychological variables be
2 separated from placebo effects? International Journal of Sport Psychology, 25,
3111 63–80.
4 Olfman, S. (1987). Relationships among physical fitness, cognitive performance
5 and self-concept in older adults. (Doctoral dissertation, Concordia University,
1986). Dissertation Abstracts International, 48, B271.
6
O’Neill, M.P. (1989). Physical self-concept and psychological mood states: Their
7
relationship to aerobic exercise. (Doctoral dissertation, Washington State
8 University). Dissertation Abstracts International, 50, B4780.
9 Ossip-Klein, D.J., Doyne, E.J., Bowman, E.D., Osborn, K.M., McDougall-Wilson,
20111 J.B., & Neimeyer, R.A. (1989). Effects of running or weight lifting on self-
1 concept in clinically depressed women. Journal of Consulting and Clinical
2 Psychology, 57, 158–161.
3 Paas, F.G., Adams, J.J., Janssen, G.M.E., Vrencken, J.G.P., & Bovens, A.M.P.M.
4 (1994). Effects of a 10-month endurance training program on performance of
5 a speeded perceptual-motor task. Perceptual and Motor Skills, 78, 1267–1273.
6 Paffenbarger, R.S., Lee, I.-M., & Leung, R. (1994). Physical activity and personal
characteristics associated with depression and suicide in American college men.
7
Acta Psychiatrica Scandinavia, 89, 16–22.
8
Palmer, L.K. (1995). Effects of a walking program on attributional style, depression
9 and self-esteem in women. Perceptual and Motor Skills, 81, 891–898.
30111 Panton, L.B., Graves. J.E., Pollock, M.L., Hayhery, J.M., & Chen, W. (1990).
1 Effect of aerobic and resistance training on fractionated reaction time and speed
2 of movement. Journals of Gerontology, 45, M26–M31.
3 Parfitt, G., Eston, R., & Connolly, D. (1996). Psychological affect at different
4 ratings of perceived exertion in high- and low-active women: A study using a
5 production protocol. Perceptual and Motor Skills, 82, 1035–1042.
6 Parfitt, G., Markland, D., & Holmes, C. (1994). Responses to physical exertion in
7 active and inactive males and females. Journal of Sport and Exercise Psychology,
16, 178–186.
8
Pasman, L., & Thompson, J.K. (1988). Body image and eating disturbance in oblig-
9
atory runners, obligatory weight lifters, and sedentary individuals. International
40111 Journal of Eating Disorders, 7, 759–777.
1 Pate, R.R., Pratt, M., Blair, S.N., Haskell, W.L., Macera, C.A., Bouchard, C.,
2 Buchner, D., Ettinger, W., Heath, G.W., King, A.C., Kriska, A., Leon, A.S.,
3 Marcus, B.H., Morris, J., Paffenbarger, R.S., Patrick, K., Pollock, M.L., Rippe,
4 J.M., Sallis, J., & Wilmore, J.H. (1995). Physical activity and public health.
45111 Journal of the American Medical Association, 273, 402–407.
References 187
1111 Paykel, E.S., & Priest, R.G. (1992). Recognition and management of depression
2 in general practice: A consensus statement. British Medical Journal, 305,
3 1198–1202.
4 Percy, L.E., Dzuiban, C.D., & Martin, J.B. (1981). Analysis of the effects of distance
running on self-concepts of elementary students. Perceptual and Motor Skills,
5
52, 42.
6
Perlmutter, M., & Nyquist, L. (1990). Relationships between self-reported physical
7 and mental health and intelligence performance across adulthood. Journal of
8 Gerontology: Sciences, 45, 145–155.
9 Perronet, F., Massicotte, D., Paquet, J.E., Brisson, G., & de Champlain, J. (1989).
1011 Blood pressure and plasma catecholamine responses to various challenges
1 during exercise recovery in man. European Journal of Applied Physiology, 58,
2 551–555.
3111 Petruzzello, S.J., Jones, A.C., & Tate, A.K. (1997). Affective responses to acute
4 exercise: A test of opponent process theory. Journal of Sports Medicine and
5 Physical Fitness, 37, 205–212.
Petruzzello, S.J., & Landers, D.M. (1994a). State anxiety reduction and exercise:
6
Does hemispheric activation reflect such changes? Medicine and Science in Sports
7
and Exercise, 26, 1028–1035.
8 Petruzzello, S.J., & Landers, D.M. (1994b). Varying the duration of acute exercise:
9 Implications for changes in affect. Anxiety, Stress and Coping, 6, 301–310.
20111 Petruzzello, S.J., Landers, D.M., Hatfield, B.D., Kubitz, K.A., & Salazar, W. (1991).
1 A meta-analysis on the anxiety-reducing effects of acute and chronic exercise:
2 Outcomes and mechanisms. Sports Medicine, 11, 143–182.
3 Petruzzello, S.J., Landers, D.M., & Salazar, W. (1993). Exercise and anxiety reduc-
4 tion: Examination of temperature as an explanation for affective change. Journal
6 Petruzzello, S.J., & Tate, A.K. (1997). Brain activation, affect and aerobic exercise:
An examination of both state-independent and state-dependent relationships.
7
8
Pierce, E.F. (1994). Exercise dependence syndrome in runners. Sports Medicine,
9 18, 149–155.
30111 Pierce, E.F., Daleng, M.L., & McGowan, R.W. (1993). Scores of exercise depen-
1 dence among dancers. Perceptual and Motor Skills, 76, 531–535.
2 Pierce, T.W., Madden, D.J., Siegel, W.C., & Blumenthal, J.A. (1993). Effects of
3 aerobic exercise on cognitive and psychosocial functions in patients with hyper-
4 tension. Health Psychology, 12, 286–291.
5 Piers, E.V. (1984). Piers-Harris Children’s Self-Concept Scale: Revised manual. Los
6 Angeles: Western Psychological Services.
7 Pistacchio, T., Weinberg, R., & Jackson, A. (1989). The development of a psycho-
biologic profile of individuals who experience and those who do not experi-
8
ence exercise related mood enhancement. Journal of Sport Behavior, 12,
9
151–166.
40111 Plante, T.G. (1993). Aerobic exercise in prevention and treatment of psycho-
1 pathology. In P. Seraganian (Ed.), Exercise psychology: The influence of physical
2 exercise on psychological processes (pp. 358–379). New York: Wiley.
3 Powell K.E., & Blair S. (1994). The public health burdens of sedentary living
4 habits: Theoretical but realistic estimates. Medicine and Science in Sports and
45111 Exercise, 26, 851–856.
188 References
1111 Powell K.E., Thompson, P.D., Casperson, C.J., & Kendrick, J.S. (1987). Physical
2 activity and the incidence of coronary heart disease. Annual Review of Public
3 Health, 8, 253–287.
4 Pysh, J.J., & Weiss, G.M. (1979). Exercise during development induces an increase
in Purkinge cell dendritic tree size. Science, 206, 230–231.
5
Radloff, L.S. (1977). The CES-D scale: A self-report depression scale for research
6
in the general population. Applied Psychological Measurement, 1, 385–401.
7 Raglin, J.S. (1993). Overtraining and staleness: Psychometric monitoring of
8 endurance athletes. In R.N. Singer, M. Murphey, & L.K. Tennant (Eds),
9 Handbook of research on sport psychology (pp. 840–850). New York: Macmillan.
1011 Raglin, J.S. (1997). Anxiolytic effects of physical activity. In W.P. Morgan (Ed.),
1 Physical activity and mental health (pp. 107–126). Washington, DC: Taylor &
2 Francis.
3111 Raglin, J.S., Eksten, F., & Garl, T. (1995). Mood state responses to a pre-season
4 conditioning program in male collegiate basketball players. International Journal
5 of Sport Psychology, 26, 214–225.
Raglin, J.S., Koceja, D.M., Stager, J.M., & Harms, C.A. (1996). Mood, neuro-
6
muscular function, and performance during training in female swimmers.
7
Medicine and Science in Sports and Exercise, 28, 372–375.
8 Raglin, J.S., & Morgan, W.P. (1987). Influence of exercise and quiet rest on state
9 anxiety and blood pressure. Medicine and Science in Sports and Exercise, 19,
20111 456–463.
1 Raglin, J.S., Turner, P.E., & Eksten, F. (1993). State anxiety and blood pressure
2 following 30 minutes of leg ergometry or weight training. Medicine and Science
3 in Sports and Exercise, 25, 1044–1048.
4 Raglin, J.S., & Wilson, M. (1996). State anxiety following 20 minutes of bicycle
5 ergometer exercise at selected intensities. International Journal of Sports
6 Medicine, 17, 467–471.
Reed, D.L., Thompson, J.K., & Coovert, D.L. (1991) Development and valida-
7
tion of the Physical Appearance State and Trait Anxiety Scale (PASTAS).
8
Journal of Anxiety Disorders, 5, 323–332.
9 Rejeski, W.J., Brawley, L.R., & Schumaker, S.A. (1996). Physical activity and
30111 health-related quality of life. Exercise and Sport Sciences Reviews, 24, 71–108.
1 Rejeski W.J., Gregg, E., Thompson, A., & Berry, M. (1991). The effects of varying
2 doses of acute aerobic exercise on psychophysiological stress responses in highly
3 trained cyclists. Journal of Sport and Exercise Psychology, 13, 188–199.
4 Rejeski, W.J., Hardy, C.J., & Shaw, J. (1991). Psychometric confounds of assessing
5 state anxiety in conjunction with acute bouts of vigorous exercise. Journal of
6 Sport and Exercise Psychology, 13, 65–74.
7 Rejeski, W.J., Thompson, A., Brubaker, P.H., & Miller, H.S. (1992). Acute exer-
cise: Buffering psychosocial stress responses in women. Health Psychology, 11,
8
355–362.
9
Richards, S., Musser, W.S., & Gershon, S. (1999). Maintenance pharmacotherapies
40111 for neuropsychiatric disorders. Philadelphia, PA: Brunner/Mazel.
1 Richert, A.J., & Hummers, J.A. (1986). Patterns of physical activity in college
2 students at possible risk for eating disorder. International Journal of Eating
3 Disorders, 5, 775–763.
4 Riddoch, C., Puig-Ribera, A., & Cooper, A. (1998). Effectiveness of physical activity
45111 promotion schemes in primary care: A review. London: Health Education Authority.
References 189
1111 Rikli, R.E., & Busch, S. (1986). Motor performance of women as a function of
2 age and physical activity level. Journal of Gerontology, 41, 645–649.
3 Rikli, R.E., & Edwards, D.J. (1991). Effects of a three year exercise program on
4 motor function and cognitive processing speed in older women. Research
Quarterly of Exercise and Sport, 62, 61–67.
5
Roberts, B.L. (1990). Effects of walking on reaction and movement times among
6
elders. Perceptual and Motor Skills, 71, 131–140.
7 Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ:
8 Princeton University.
9 Roth, D.L. (1989). Acute emotional and psychophysiological effects of aerobic
1011 exercise. Psychophysiology, 26, 593–602.
1 Roy, M., & Steptoe, A. (1991). The inhibition of cardiovascular responses to
2 mental stress following aerobic exercise. Psychophysiology, 28, 680–699.
3111 Rozin, P., & Stoess, C. (1993). Is there a general tendency to become addicted?
4 Addictive Behaviors, 18, 81–87.
5 Ruck, J., & Taylor, A.H. (1991). The effects of aerobic exercise and quiet rest on
state anxiety and blood pressure following a psychological stressor (abstract).
6
Journal of Sports Sciences, 9, 447–448.
7
Rudy, E.B., & Estok, P.J. (1989). Measurement and significance of negative addic-
8 tion in runners. Western Journal of Nursing Research, 11, 548–558.
9 Rudy, E.B., & Estok, P.J. (1990). Running addiction and dyadic adjustment.
20111 Research in Nursing and Health, 13, 219–225.
1 Russell, J.A. (1980). A circumplex model of affect. Journal of Personality and
3 Rutter, M., & Smith D.J. (Eds). (1995). Psychosocial disorders in young people:
4 Time trends and their causes. New York: John Wiley.
5 Ryckman, R.M., Robbins, M.A., Thornton, B. & Cantrell, P. (1982). Development
6 and validation of a physical self-efficacy scale. Journal of Personality and Social
Psychology, 42, 891–900.
7
Sachs, M.L. (1981). Running addiction. In M. Sacks & M. Sachs (Eds), Psychology
8
of Running (pp. 116–126). Champaign, IL: Human Kinetics.
9 Sachs, M.L., & Pargman, D. (1979). Running addiction: A depth interview exam-
30111 ination. Journal of Sport Behavior, 2, 143–155.
1 Sachs, M.L., & Pargman, D. (1984). Running addiction. In M.L. Sachs &
2 G.W. Buffone (Eds), Running as therapy: An integrated approach (pp. 231–252).
3 Lincoln, NE: University of Nebraska Press.
4 Sallis, J.F., & Patrick, K. (1994). Physical activity guidelines for adolescents: Con-
5 sensus statements. Pediatric Exercise Science, 6, 302–314.
6 Salthouse, T.T. (1985). A theory of aging. Amsterdam: North-Holland.
7 Sarason, I.G. (1975). Test anxiety, attention and general problems of anxiety. In
C.D. Spielberger & I.G. Sarason (Eds), Stress and anxiety: Volume 1 (pp. 381–
8
403). Washington DC: Hemisphere/Wiley.
9
Schempp, P.G., Cheffers, J.T.F., & Zaichowsky, L.D. (1983). Influence of deci-
40111 sion-making on attitudes, creativity, motor skill and self-concept in elementary
1 children. Research Quarterly for Exercise and Sport, 54, 183–189.
2 Schlicht, W. (1994). Sport und seelische gesundheit: Eine meta-analyse. In J. Nitsch
3 & R. Seiler (Eds), Health sport – movement therapy: Proceedings of the 8th
4 European Congress of Sport Psychology 1991: Volume 4 (pp. 57–63). Sankt
45111 Augustin, Germany: Academia Verlag.
190 References
1111 Scott, J. (1996). Cognitive therapy of affective disorders: A review. Journal of
2 Affective Disorders, 37, 1–11.
3 Seamonds, B.C. (1982). Stress factors and their effects on absenteeism in a corpo-
4 rate employee group. Journal of Occupational Medicine, 24, 393–397.
Secord, P.F., & Jourard, S.M. (1953). The appraisal of body cathexis: Body cathexis
5
and the self. Journal of Consulting Psychology, 17, 343–347.
6
Sedlock, D.A., & Duda, J.L. (1994). The effect of trait anxiety and fitness level
7 on heart rate and state anxiety responses to a mental arithmetic stressor among
8 college-age women. International Journal of Sport Psychology, 25, 218–229.
9 Selye, H. (1956). The stress of life. New York: McGraw-Hill.
1011 Sexton, H., Maere, A., & Dahl, N.H. (1989). Exercise intensity and reduction
1 in neurotic symptoms: A controlled follow-up study. Acta Psychiatrica
2 Scandinavica, 80, 231–235.
3111 Shavelson, R.J., Hubner, J.J., & Stanton, G.C. (1976). Self-concept: Validation of
4 construct interpretations. Review of Educational Research, 46, 407–41.
5 Sherrill, C. (1997). Disability, identity, and involvement in sport and exercise. In
K.R. Fox (Ed.), The physical self: From motivation to well-being (pp. 257–286).
6
7
Sherwood, A., Light, K.C., & Blumenthal, J.A. (1989). Effects of aerobic exercise
8 training on hemodynamic responses during psychosocial stress in normotensive
9 and borderline hypertensive type A men: A preliminary report. Psychosomatic
20111 Medicine, 51, 123–136.
1 Sherwood, D.E., & Selder, D.J. (1979). Cardiorespiratory health, reaction time,
2 and aging. Medicine and Science in Sports, 11, 186–189.
3 Short, M.A., DiCarlo, S., Steffee, W.P., & Pavlou, K. (1984). Effects of physical
4 conditioning on self-concept of adult obese males. Physical Therapy, 64, 194–198.
5 Smith, T.P. (1984). An evaluation of the psychological effects of physical educa-
6 tion on children (Doctoral dissertation, De Paul University, 1983). Dissertation
Abstracts International, 44, 3260B.
7
Sonstroem, R.J. (1978). Physical estimation and attraction scales: Rationale and
8
research. Medicine and Science in Sports, 10, 97–102.
9 Sonstroem, R.J. (1984). Exercise and self-esteem. Exercise and Sport Sciences
30111 Reviews, 12, 123–155.
1 Sonstroem, R.J. (1997a). Physical activity and self-esteem. In W.P. Morgan (Ed.),
2 Physical activity and mental health (pp. 128–143). Washington DC: Taylor &
3 Francis.
4 Sonstroem, R.J. (1997b). The physical self-system: A mediator of exercise and
5 self-esteem. In K.R. Fox (Ed.), The physical self: From motivation to well-being
6 (pp. 3–26). Champaign, IL: Human Kinetics.
7 Sonstroem, R.J., Harlow, L.L., & Josephs, L. (1994). Exercise and self-esteem:
Validity of model expansion and exercise associations. Journal of Sport and
8
Exercise Psychology, 16, 29–42.
9
Sonstroem, R.J., & Morgan, W.P. (1989). Exercise and self-esteem: Rationale and
40111 model. Medicine and Science in Sports and Exercise, 21, 329–337.
1 Sonstroem, R.J., & Potts, S.A. (1996). Life adjustment correlates of physical self-
2 concepts. Medicine and Science in Sports and Exercise, 28, 619–625.
3 Sonstroem, R.J., Speliotis, E.D., & Fava, J.L. (1992). Perceived physical compe-
4 tence in adults: An examination of the Physical Self-Perception Profile. Journal
References 191
1111 Sothmann, M.S., Buckworth, J., Claytor, R.P., Cox, R.H., White-Welkley, J.E., &
2 Dishman, R.K. (1996). Exercise training and the cross-stressor adaptation
3 hypothesis. Exercise and Sport Sciences Reviews, 24, 267–287.
4 Sothmann, M.S., Hart, B.A., & Horn, T.S. (1991). Plasma catecholamine response
to acute psychological stress in humans: Relation to aerobic fitness and exer-
5
cise training. Medicine and Science in Sports and Exercise, 23, 7.
6
Speith, W. (1965). Slowness of task performance and cardiovascular diseases.
7 In A.T. Welford & J.E. Birren (Eds), Behavior, aging, and the nervous system
8 (pp. 366–400). Springfield, IL: Thomas.
9 Spence, J.C. (unpublished). The effects of exercise on self-esteem. Edmonton:
1011 University of Alberta.
1 Spence, J.C., & Poon, P. (1997). The effect of physical activity on self-concept:
2 A meta-analysis. Alberta Centre for Well-Being: Research Update, 4, 4.
3111 Spielberger, C.D., Gorsuch, R.L., Lushene, R., Vagg, P.R., & Jacobs, G.A. (1983).
4 Manual for the State-trait Anxiety Inventory (Form Y1). Palo Alto, CA:
5 Consulting Psychologists Press.
Spirduso, W.W. (1975). Reaction and movement time as a function of age and
6
physical activity level. Journal of Gerontology, 30, 435–440.
7
Spirduso, W.W. (1980). Physical fitness, aging, and psychomotor speed: A review.
8 Journal of Gerontology, 35, 850–865.
9 Spirduso, W.W. (1994). Physical dimensions of aging. Champaign, IL: Human
20111 Kinetics.
1 Spirduso, W.W., & Clifford, P. (1978). Replication of age and physical activity
2 effects on reaction and movement time. Journal of Gerontology, 33, 26–30.
3 Spitzer, R.L., Endicott, J., & Robins, E. (1978). Research diagnostic criteria.
4 Archives of General Psychiatry, 35, 773–782.
5 Sports Council & Health Education Authority (1992). Allied Dunbar National
6 Fitness Survey: Main findings. London: Author.
Stacey, C., Kozma, A., & Stones, M.J. (1985). Simple cognitive and behavioral
7
changes resulting from improved physical fitness in persons over 50 years of
8
age. Canadian Journal on Aging, 4, 67–73.
9 Stein, P.K., & Boutcher, S.H. (1992). The effect of participation in an exercise
30111 training program on cardiovascular reactivity in sedentary middle-aged males.
1 International Journal of Psychophysiology, 13, 215–223.
2 Stelmach, G. (1994). Physical activity and aging: Sensory and perceptual process-
3 ing. In C. Bouchard (Ed.), Physical activity, fitness, and health: International
4 proceedings and consensus statement (pp.521–529). Champaign, IL: Human
5 Kinetics.
6 Stephens, T. (1988). Physical activity and mental health in the United States and
7 Canada: Evidence from four population surveys. Preventive Medicine, 17, 35–47.
Steptoe, A. (1992). Physical activity and psychological well-being. In N. Norgan
8
(Ed.), Physical activity and health (pp. 207–229). Cambridge: Cambridge Univer-
9
sity Press.
40111 Steptoe, A., & Bolton, J. (1988). The short-term influence of high and low inten-
1 sity physical exercise on mood. Psychology and Health, 2, 91–106.
2 Steptoe, A., & Butler, N. (1996). Sports participation and emotional well-being in
3 adolescents. The Lancet, 347, 1789–1792.
4 Steptoe, A., & Cox, S. (1988). Acute effects of aerobic exercise on mood. Health
45111 Psychology, 7, 329–340.
192 References
1111 Steptoe, A., Edwards, S., Moses, J., & Mathews, A. (1989). The effects of exercise
2 training on mood and perceived coping ability in anxious adults from the general
3 population. Journal of Psychosomatic Research, 33, 537–547.
4 Steptoe, A., Kearsley, N., & Walters, N. (1993). Cardiovascular activity during
mental stress following vigorous exercise in sportsmen and inactive men.
5
6
Steptoe, A., Moses, J., Edwards, S., & Mathews, A. (1993). Exercise and respon-
7 sivity to mental stress: Discrepancies between the subjective and physiological
8 effects of aerobic training. International Journal of Sport Psychology, 24, 110–129.
9 Steptoe, A.M., Moses, J., Mathews, A., & Edwards, S. (1990). Aerobic fitness,
1011 physical activity, and psychophysiological reactions to mental tasks. Psycho-
1 physiology, 27, 264–274.
2 Stewart, A.L., Hays, R.D., Wells, K.B., Rogers, W.H., Spritzer, K.L., & Greenfield,
3111 S. (1994). Long-term functioning and well-being outcomes associated with phys-
4 ical activity and exercise in patients with chronic conditions in the Medical
5 Outcomes Study. Journal of Clinical Epidemiology, 47, 719–730.
Stones, M.J., & Kozma, A. (1989). Age, exercise, and coding performance.
6
Psychology and Aging, 4, 190–194.
7
Summers, J.J., & Hinton, E.R. (1986). Development of scales to measure partic-
8 ipation in running. In L.E. Unestahl (Ed.), Contemporary Sport Psychology
9 (pp. 73–84). Orebro, Sweden: Veje.
20111 Suominen, H., Heikkinen, E., Parkatti, T., Forsberg, S., & Kiiskinen, A. (1980).
1 Effects of ‘lifelong’ physical training on functional aging in men. Scandinavian
2 Journal of Social Medicine, 55, 225–240.
3 Szabo, A. (1995). The impact of exercise deprivation on well-being of habitual
4 exercisers. The Australian Journal of Science and Medicine in Sport, 27,
5 68–75.
6 Szabo, A. (1997). Cross sectional research on the Internet. Journal of Physical
Education and Sport Sciences, 10, 14–22.
7
Szabo, A., Brown, T.G., Gauvin, L., & Seraganian, P. (1993). Aerobic fitness does
8
not influence directly heart rate reactivity to mental stress. Acta Psychologica
9 Hungarica, 81, 229–237.
30111 Szabo, A., Frenkl, R., & Caputo, A. (1996). Deprivation feelings, anxiety, and
1 commitment to various forms of physical activity: A cross-sectional study on
2 the Internet. Psychologia, 39, 223–230.
3 Szabo, A., Frenkl, R., & Caputo, A. (1997). Relationships between addiction to
4 running, commitment to running, and deprivation from running. European
5 Yearbook of Sport Psychology, 1, 130–147.
6 Szabo, A., Frenkl, R., Janek, G., Kalman, L., & Laszay, D. (1998). Runners’
7 anxiety and mood on running and non-running days: An in situ daily moni-
toring study. Psychology, Health and Medicine, 3, 193–199.
8
Szabo, A., & Gauvin, L. (1992). Reactivity to written mental arithmetic: Effects
9
of exercise lay-off and habituation. Physiology and Behavior, 51, 501–506.
40111 Szabo, A., Peronnet, F., Boudreau, G., Cote, L., Gauvin, L., & Seraganian, P.
1 (1993). Psychophysiological profiles in response to various challenges during
2 recovery from acute aerobic exercise. International Journal of Psychophysiology,
3 14, 285–292.
4 Szafran, J. (1966). Age, cardiac output, and choice reaction time. Nature, 209,
45111 836–837.
References 193
1111 Szymanski, L.A., & Chrisler, J.C. (1990). Eating disorders, gender role, and athletic
2 activity. Psychology: A Journal of Human Behavior, 27, 20–29.
3 Talbot, H.M., & Taylor, A.H. (1998). Changes in physical self-perceptions: Findings
4 from a randomised controlled study of a GP exercise referral scheme. Journal
of Sports Sciences, 16, 105–106.
5
Tate, A.K., & Petruzzello, S.J. (1995). Varying the intensity of acute exercise:
6
Implications for changes in affect. Journal of Sports Medicine and Physical
7 Fitness, 35, 295–302.
8 Taylor, J.A. (1953). A personality scale of manifest anxiety. Journal of Abnormal
1011 Thaxton, L. (1982). Physiological and psychological effects of shortterm exercise
1 addiction on habitual runners. Journal of Sport Psychology, 4, 7380.
2 Thayer, R.E. (1967). The measurement of activation through self-report.
3111 Psychological Reports, 20, 663–678.
4 Thiel, A., Gottfried, H., & Hesse, F.W. (1993). Subclinical eating disorders in male
5 athletes. Acta Psychiatrica Scandinavica, 88, 259–265.
Thirlaway, K., & Benton, D. (1996). Exercise and mental health: The role of
6
activity and fitness. In J. Kerr, A. Griffiths, & T. Cox (Eds), Workplace health,
7
employee fitness and exercise (pp. 69–82). London: Taylor & Francis.
8 Thomas, J.R., Landers, D.M., Salazar, W., & Etnier, J. (1994). Exercise and
9 cognitive function. In C. Bouchard, R.J. Shepherd, & T. Stephens (Eds),
20111 Physical activity, fitness, and health (pp. 521–529). Champaign, IL: Human
1 Kinetics.
2 Thomas, J.R., & Nelson, J.K. (1996). Research Methods in Physical Activity (3rd
3 edn). Champaign, IL: Human Kinetics.
4 Thompson, J.K., & Blanton, P. (1987). Energy conservation and exercise depen-
5 dence: A sympathetic arousal hypothesis. Medicine and Science in Sports and
6 Exercise, 19, 91–97.
Thornton, E.W., & Scott, S.E. (1995). Motivation in the committed runner:
7
Correlations between self-report scales and behaviour. Health Promotion
8
International, 10, 177–184.
9 Tilford, S., Delaney, F., & Vegells, M. (1997). Effectiveness of mental health promo-
30111 tion interventions: A review. London: Health Education Authority.
1 Tkachuk, G.A., & Martin, G.L. (1999). Exercise therapy for patients with psychi-
2 atric disorders: Research and clinical implications. Professional Psychology:
3 Research and Practice, 30, 275–282.
4 Topp, R. (1989). Effect of relaxation or exercise on undergraduates’ test anxiety.
5 Perceptual and Motor Skills, 69, 35–41.
6 Torres, R., & Fernandez, F. (1995). Self-esteem and the value of health as deter-
7 minants of adolescent health behavior. Journal of Adolescent Health Care, 16,
60–63.
8
Trine, MR., & Morgan, WP. (1997). Influence of time of day on the anxiolytic
9
effects of exercise. International Journal of Sports Medicine, 18, 161–168.
40111 Trujillo, C.M. (1983). The effect of weight training and running exercise inter-
1 vention programs on the self-esteem of college women. International Journal of
2 Sport Psychology, 14, 162–173.
3 Tucker, L.A. (1987). Effect of weight training on body attitudes: Who benefits
4 most? Journal of Sports Medicine, 27, 70–78.
45111 Tucker, L.A., & Mortell, R. (1993). Comparison of the effects of walking and weight
194 References
1111 training programs on body image in middle-aged women: An experimental study.
2 American Journal of Health Promotion, 8, 34–42.
3 Tuson, K.M., & Sinyor, D. (1993). On the affective benefits of acute aerobic exer-
4 cise: Taking stock after twenty years of research. In P. Seraganian (Ed.), Exercise
psychology (pp. 80–121). New York: John Wiley.
5
US Department of Health and Human Services (PHS). (1996). Physical activity
6
and health: A report of the Surgeon General (executive summary). Pittsburgh,
7 PA: Superintendent of Documents.
8 Ussher, M., Taylor, A.H., West, R., & McEwen A. (in press). Does exercise aid
9 smoking cessation? A sytematic review. Journal of Addiction.
1011 Van Doornen, L.J., & de Geus, E.J. (1989). Aerobic fitness and the cardiovas-
1 cular response to stress. Psychophysiology, 26, 17–28.
2 Veale, D., Le Fevre, K., Pantelis, C., de Souza, V., Mann, A., & Sargeant, A.
3111 (1992). Aerobic exercise in the adjunctive treatment of depression: A randomised
4 controlled trial. Journal of the Royal Society of Medicine, 85, 541–544.
5 Vlachopoulos, S., Biddle, S., & Fox, K. (1996). A social-cognitive investigation
into the mechanisms of affect generation in children’s physical activity. Journal
6
of Sport and Exercise Psychology, 18, 174–193.
7
Wankel, L.M. (1993). The importance of enjoyment to adherence and psycho-
8 logical benefits from physical activity. International Journal of Sport Psychology,
9 24, 151–169.
20111 Wankel, L.M., & Berger, B.G. (1990). The psychological and social benefits of
1 sport and physical activity. Journal of Leisure Research, 22, 167–182.
2 Warr, P. (1990). The measurement of well-being and other aspects of mental
3 health. Journal of Occupational Psychology, 63, 193–210.
4 Watson, D., Clark, L.A., & Carey, G. (1988). Positive and negative affectivity
5 and their relation to anxiety and depressive disorders. Journal of Abnormal
6 Psychology, 97, 346–353.
Watson, D., Clark, L.A., & Tellegen, A. (1988). Development and validation of
7
brief measures of positive and negative affect: The PANAS scales. Journal
8
of Personality and Social Psychology, 54, 1063–1070.
9 Watson, D., & Tellegen, A. (1985). Toward a consensual structure of mood.
30111 Psychological Bulletin, 98, 219–235.
1 Weiner, B. (1995). Judgments of responsibility: A foundation for a theory of social
2 conduct. New York: The Guilford Press.
3 Weismann, M.M., & Klerman, G.L. (1992). Depression: Current understanding
4 and changing trends. Annual Review of Public Health, 13, 319–339.
5 Weyerer, S. (1992). Physical inactivity and depression in the community: Evidence
6 from the Upper Bavarian Field Study. International Journal of Sports Medicine,
7 13, 492–496.
White, A. (1974). The interrelationships between measures of physical fitness and
8
measures of self-concept of selected Mississippi State University male students
9
(Doctoral dissertation, Mississippi State University, 1973). Dissertation Abstracts
40111 International, 34, 4849A.
1 Whiting, J. (1981). The effect of a mild exercise program on the psychological
2 treatment of in-patient alcoholics. Dissertation Abstracts International, 42,
3 394B–395B.
4 Wichmann, S., & Martin, D.R. (1992). Exercise excess. The Physician and Sports-
45111 medicine, 20, 193–200.
References 195
1111 Williamson, D.A., Netemeyer, R.G., Jackman, L.P., Anderson, D.A., Funsch, C.L.,
2 & Rabalais, J.Y. (1995). Structural equation modeling of risk factors for the
3 development of eating disorder symptoms in female athletes. International
4 Journal of Eating Disorders, 17, 387–393.
Wolf, E.M., & Akamatsu, T.J. (1994). Exercise involvement and eating disordered
5
characteristics in college students. Eating Disorders, 2, 308–318.
6
Worcester, M.C., Hare, D.L., Oliver, R.G., Reid, M.A., & Goble, A.J. (1993).
7 Early programmes of high and low intensity exercise and quality of life after
8 acute myocardial infarction. British Medical Journal, 307, 1244–1247.
9 World Health Organisation (1993). The ICD-10 classification of mental and behav-
1011 ioural disorders: Diagnostic criteria for research. Geneva: WHO.
1 World Health Organisation (1995). Exercise for health. WHO/FIMS Committee
2 on Physical Activity for Health. Bulletin of the World Health Organisation, 73,
3111 135–136.
4 Wykoff, W. (1993). The psychological effects of exercise on non-clinical and clin-
5 ical populations of adult women: A critical review of the literature. Occupational
Therapy in Mental Health, 12, 69–106.
6
Wylie, R.C. (1979). The self-concept, Volume 2. Theory and research on selected
7
topics. Lincoln, NE: University of Nebraska.
8 Wylie, R.C. (1989). Measures of self-concept. Lincoln, NE: University of Nebraska.
9 Yates, A., Leehey, K., & Shisslak, C.M. (1983). Running: An analogue of anorexia?
20111 New England Journal of Medicine, 308, 251–255.
1 Yates, A., Shisslak, C.M., Crago, M., & Allender, J. (1994). Overcommitment to
2 sport: Is there a relationship to the eating disorders? Clinical Journal of Sport
3 Medicine, 4, 39–46.
4 Yeagle, G.W. (1982). The effect of a physical recreation program on the self-
5 concept and flexibility of senior citizens (Doctoral dissertation, University of
6 Utah). Dissertation Abstracts International, 42, B3184.
Youngstedt, S.D., Dishman, R.K., Cureton, K.J., & Peacock, L.J. (1993). Does
7
body temperature mediate anxiolytic effects of acute exercise? Journal of
8
Applied Physiology, 74, 825–831.
9 Youngstedt, S.D., O’Connor, P.J., & Dishman, R.K. (1997). The effects of acute
30111 exercise on sleep: A quantitative synthesis. Sleep, 20, 203–214.
1 Zaitz, D. (1989). Are you an exercise addict? Idea Today, 7, 44.
2 Zuckerman, M., & Lubin, B. (1965). Manual for the Multiple Affect Adjective
3 Checklist. San Diego, CA: Educational and Industrial Testing Service.
4 Zung, W. (1965). A self-rating depression scale. Archives of General Psychiatry,
5 12, 63–70.
6 Zung, W.W.K., Richards, C.B., & Short, M.J. (1965). Self-rating depression scale
7 in an out-patient clinic. Archives of General Psychiatry, 13, 508–515.
8
9
40111
1
2
3
4
45111
1111
2
3
4
5
6
7
8
9
1011
1
2
3111
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
4
45111
1111
2 Index
3
4
5
6
7
8
9
1011
1
2
3111 ACSM see American College of Sports cognitive performance/fitness
4 Medicine relationship 118–29
5 Active for Life Campaign 1, 85 concept 119
activity status categories 49–50, 50 primary ageing 119, 121
6
acute exercise studies secondary ageing 128
7
anxiety-reduction 11, 26–7, 28–31, slowness of response 118
8 32–3 alcohol 41, 46, 168
9 psychosocial stress reactivity 37, Aldana, S.G. 17
20111 38–40 American College of Sports Medicine
1 adaptation to exercise 165 (ACSM, 1991) guidelines 24
2 addiction to exercise American Psychiatric Association 47
3 correlates 138 anorexia nervosa
4 definition 7 definition 7
5 health practitioners 142 excessive exercising 144, 145–9
6 physiological causes 136–7 anxiety
psychological causes 137 acute exercise 11, 26–7, 28–31, 32
7
research recommendations 143 chronic exercise 17–26
8 cross-sectional chronic exercise
reviews 138–41
9 self-help 143 studies 17
30111 symptoms 135–6 definition 7, 10–11
1 adherence to regimes 167–8 longitudinal relationship to physical
2 adolescents, self-esteem 108 activity 18, 19–23
3 adoption phase, exercise programmes physical activity studies 10–45
4 130–2 research findings 154–5
5 aerobic fitness self-reported 11
6 definition 119–20 treatment comparisons 25–6
7 intervention studies 122–3, 124–5 visual analogue assessment scale 24
non-aerobic comparison 24–5 anxiety; see also social physique
8
affect (emotion) anxiety
9 arousal levels, addiction to exercise 137
definition 6
40111 measurement 66–7, 68–9, 70 assessment
1 mood distinction 66–7 addiction 138–41
2 physical activity effects 63–87 depression 81
3 research findings 156–7 exercise dependence 134, 135–6
4 ageing 118–29 HRQL 64, 65, 66
45111 cognition 120–1 limitations 159–60
198 Index
1111 mood/emotion 66–7, 68–9, 70 Carroll, D. 37
2 self-esteem/self-perceptions 90, causal links 57–60, 155
3 92–3 Centre for Health Economics 4
4 attitudinal barriers, health cerebral circulation, fitness/cognitive
professionals 166–7 performance relationship 127
5
attrition 160 CHD see coronary heart disease
6 children
autonomy, self-esteem 96, 112–13
7 mental health problems 46
8 BASES see British Association of physical activity 2
9 Sport and Exercise Sciences self-esteem 108
1011 Bavarian populations 51 chronic exercise studies
1 Beck Depression Inventory (BDI) 48, anxiety-reduction 11, 17–26
2 53, 56 psychosocial stress reactivity 33,
3111 Befindlichkeitsskalen (BFS) 68, 70 34–6
4 behaviours chronological ageing 119
5 behavioural addictions 143 ‘circumplex’ model (Russell) 67,
escape behaviours 137 70
6
slowing with age 118, 120 clinical diagnosis, depression 53
7 clinically-defined depression
unhealthy 130–53
8 Belgium, psychomotor therapy 5 definition 7, 47–8
9 benefits of exercise 4–5 epidemiological evidence 48–52
20111 beta-endorphins, exercise dependence physical activity relationship
1 136, 137 46–62
2 BFS see Befindlichkeitsskalen cognitive performance
3 Biddle, Stuart J.H. 1–9, 63–87 concept 119
4 biochemical mechanisms, physical fitness/ageing 118–29
5 activity effects on mood/emotion intervention studies 122–3,
6 83–4 124–5
biological ageing 119 physiological mechanisms 127–8
7
biological plausibility (Hill’s criterion) research findings 158
8 slowing with age 118, 120
59
9 blood flow, cerebral 127 tests 119, 120
30111 body acceptance, heavy exercise in coherence (Hill’s criterion) 60
1 females 97 college students
2 body image convenience for study 18, 27
3 research 97 Harvard Alumni studies 51
4 self-esteem 95–6 commitment
5 Boutcher, Stephen H. 1–9, 118–29 addiction distinction 133, 136
6 BPS see British Psychological Society definition 7
7 British Association of Sport and Commitment to Running Scale
Exercise Sciences (BASES) 6 (Conboy) 136
8
British controlled experimental trials community health, exercise 1–2
9 Community Mental Health Team
78–9
40111 British population surveys 76, 77, 80 168
1 British Psychological Society (BPS) 6 Community Psychiatric Nurses 168
2 bulimia nervosa 7, 144 compulsive exercising see exercise
3 dependence
4 Calfas, K.J. 12 consensus workshops 4
45111 Camacho, T.C. 49–50, 50 consistency (Hill’s criterion) 57
Index 199
1111 constructs, self-perception 89, 92 risk versus physical activity 51
2 controlled experimental trials sub-clinical 81
3 cost-effectiveness 161 women 49
4 exercise treatment for clinically- deprivation see withdrawal symptoms
defined depression 54–5, 56–7 Diagnostic and Statistical Manual of
5
mood/emotion 78–9, 80–1 Mental Disorders (DSM-IV) 48
6 Dishman, R.K. 47, 52, 53, 81
self-esteem 98, 100–7
7 coronary heart disease (CHD) 2 ‘distraction’ hypothesis 32–3
8 correlates, exercise dependence 137, dose-response 49, 58, 160–1
9 142 drug rehabilitation 168
1011 costs drug treatment, depression 47, 56
1 anxiety/depression 10 DSM-IV see Diagnostic and Statistical
2 cost-effectiveness 161, 164 Manual of Mental Disorders
3111 depression treatments 46–7 Dustman, R.E. 122, 123
4 mental illness 3
5 counselling for fitness 167 Eating Attitude Test (EAT) 147,
Craft, L.L. 53 149
6
Crews, D.J. 13 Eating Disorder Inventory (EDI) 145,
7 146, 147, 148, 149
cross-sectional studies
8 chronic exercise 17 eating disorders
9 fitness/cognitive performance/ageing exercise relationship 144
20111 121–2 females 149, 150
1 psychosocial stress reactivity 33 psychosocial model 151
2 self-esteem 96–8, 113 weight control 151
3 cultural values, self-esteem 90 Effect Size (ES), mood/emotion in
4 different settings 83
5 definitions effect sizes (ES)
6 addiction 133 anxiety/stress reduction 12, 13
aerobic fitness 119 exercise as depression treatment
7
ageing 119 52–3
8 effectiveness, exercise for self-esteem
clinically-defined depression 47–8
9 cognitive performance 119 111–12
30111 commitment 133 EFI see Exercise-induced Feeling
1 exercise 7 Inventory
2 exercise dependence 132–3, 134 ego goal orientation 82–3
3 general 6–9 electroencephalograms (EEGs) 127–8
4 physical fitness 119–20 electromyography (EMG),
5 stress/anxiety 10–11 anxiety/stress measurement 13
6 Department of Health, working days electronic searches 16, 63
7 lost 10 EMG see electromyography
depression emotion
8
clinical diagnosis 53 measurement 66–7, 68–9, 70
9 mood distinction 66–7
clinically-defined 7, 46–62
40111 drug treatment 56 physical activity effects 63–87
1 future research 62 research findings 156–7
2 incidence 2–3 endorphins, exercise dependence 136
3 measurement 81 epidemiological research
4 meta-analytic reviews 81 depression 48–52, 155, 156
45111 research findings 61–2, 155–6 mood/emotion 76, 77, 80, 82
200 Index
1111 ES see Effect Size Farmer, M. 49
2 escape behaviours 137 Faulkner, Guy E. 1–9
3 euphoria 135 Feeling Scale (FS) 69
4 EuroQol 64, 65 females
evidential criteria, exercise anti- cognitive performance 122
5
depressant effects 57–60 eating disorders 149, 150
6 heavy exercise 97
excessive exercise
7 eating disorders 144, 145–9 inactivity/depression 49
8 standards requirement 150, 152 First Law of Human Nature
9 exercise (Campbell) 89
1011 characteristics fitness
1 acute studies 27, 32 aerobic 24–5, 119–20, 122–3, 124–5
2 chronic studies 24–5 ageing 118–29
3111 definition 7 cognitive performance 127–8
4 exercise addiction see exercise concept 119–20
5 dependence counselling for 167
exercise dependence definition 8
6
correlates 138 psychiatric patients 48
7 Fox, Kenneth R. 1–9, 88–117
definition 7
8 health practitioners 142 FS see Feeling Scale
9 physiological causes 136–7
20111 psychological causes 137 gender differences, self-esteem 110
1 research recommendations 143 General Practitioners (GPs), mental
2 reviews 138–41 health patients 3
3 self-help 143 Glasser, W. 132
4 symptoms 135–6 global self-esteem 90, 99
5 exercise deprivation, definition 7 Global Self-Esteem Scale (Rosenberg)
6 exercise deprivation; see also 90
withdrawal symptoms group activities, self-esteem 113
7
exercise effectiveness 111–12
8 Harvard Alumni studies,
exercise practitioners 41
9 exercise regimes depression/physical activity 51
30111 adoption phase 130–2 health analysts 41–2
1 non-adherence 131 Health Education Authority 1
2 psychosocial stress reactivity 44–5 Health of the Nation Task Force on
3 Exercise and Self-Esteem Model 94, Physical Activity 1
4 95 health professionals
5 exercise treatment, depression, RCSs attitudinal barriers 166–7
6 54–5 difficulties facing 166–7
7 Exercise-induced Feeling Inventory exercise dependence treatments
(EFI) 69, 70, 84 142
8
experimental evidence (Hill’s criterion) health promotion 66, 85
9 Health Survey for England (1995) 2–3
60
40111 experimental trials, mood/emotion health-related quality of life (HRQL)
1 78–9, 80–1, 82 63
2 experimental trials; see also measurement 64, 65, 66
3 randomised controlled studies; heavy exercise
4 research; reviews anxiety states 131
45111 Eysenck Personality Inventory 145 females 97
Index 201
1111 hierarchical models, self-esteem 90, 91 Marsh’s Self-Description
2 high intensity exercise see heavy Questionnaire 90
3 exercise measurement
4 Hill’s criteria 57–60 addiction 138–41
coherence 60 depression 81
5
consistency 57 HRQL 64, 65, 66
6 limitations 159–60
dose-response 58
7 experimental evidence 60 mood/emotion 66–7, 68–9, 70
8 plausibility 59 self-esteem/self-perceptions 90,
9 specificity 58 92–3
1011 strength of association 57 mechanisms
1 temporal sequence 58 fitness/cognitive performance
2 HRQL see health-related quality of relationship 127–8
3111 life PA effects on mood/affect 83–4
4 hypertension 1, 37 physiological 164
5 psychological 164
inactivity research directions 163
6
depression 49–52 self-esteem 95–6, 112–13
7 mental illness 2–3, 4; see also addiction
epidemiological studies 48–52
8 odds ratios 50 to exercise; anxiety; depression
9 physical activity index 51 meta-analyses
20111 individual preferences, exercise 82 exercise as depression treatment
1 individual worth perception, self- 52–3
2 esteem 89–90 fitness/cognition/ageing 121–2
3 intervention studies mood/emotion 71, 72–5, 77, 81
4 exercise effects on self-esteem 94, strength of exercise effects 12–13,
5 98–9 14–15
6 fitness effects on cognitive methodological difficulties, study
performance 124–5 design 56–7
7
fitness/cognitive performance/ageing middle-aged adults, self-esteem
8 109
122–3
9 interventions, recommendations Minnesota Multiphasic Personality
30111 167–8 Inventory (MMPI) 144, 145
1 MMPI see Minnesota Multiphasic
2 Kobasa, S.C. 10 Personality Inventory
3 Kugler, J. 13 Mock, V. 24
4 models, self-esteem 90, 91, 92, 93, 94,
5 Lazarus, R.S. 66–7 94
6 Leith, L.M. 16, 18 moderator variables, anxiety/stress
7 Likert scales 135–6, 136, 143 reduction 12, 13
literature see reviews mood
8
Long, B.C. 12 definition 7
9 emotion distinction 66–7
longitudinal studies, anxiety/stress
40111 reduction 18, 19–23 heavy exercise 131
1 measurement 66–7, 68–9, 70
2 MAACL see Multiple Affect Adjective physical activity effects 63–87
3 Check List research findings 156–7
4 McAuley, E. 25 Morgan, W.P. 4, 48, 53, 66, 132
45111 McDonald, D.G. 13, 52, 53 Moses, J. 71
202 Index
1111 motivation older adults
2 addiction/commitment distinction cognitive performance 118–29
3 133 fitness effects on cognitive
4 emotion/mood 66 performance 121–3, 124–5,
self-esteem 89 126–9
5
Multiple Affect Adjective Check List self-esteem 109–10
6 on-line searches 16, 63
(MAACL) 67, 69, 70, 77
7 Murphy, G. 89
8 Mutrie, Nanette 46–62 PA see physical activity
9 Paffenbarger, R.S. 51, 51
1011 narrative reviews see qualitative PANAS see Positive and Negative
1 reviews Affect Schedule
2 National Curriculum, self-esteem 88 Perceived Stress Scale 25
3111 National Health and Nutrition personal control, self-esteem 94, 96,
4 Examination Survey (NHANES) 112–13
5 82 Petruzzello, S.J. 13, 16, 18, 26, 27,
National Institutes of Health (USA) 32
6
64 phobic anxiety, characteristics 11
7 physical activity (PA)
negative addiction
8 definition 8 anxiety/stress reduction 10–45
9 primary exercise dependence 132 definition 8, 10–11
20111 Negative Addiction Scale 139, 140 HRQL 64
1 negative effects role in anxiety prevention 11–12
2 physical activity 130–53 Physical Appearance State and Trait
3 transient 130 Anxiety scale 26
4 neural efficiency, fitness 127–8 physical fitness
5 neurotransmitters ageing 118–29
6 fitness/cognitive performance 127 cognitive performance relationship
mood/emotion 84 mechanisms 127–8
7
neurotrophic stimulation, concept 119–20
8 definition 8
fitness/cognitive performance
9 relationship 127 psychiatric patients 48
30111 NHANES see National Health and physical health problems 1, 2
1 Nutrition Examination Survey Physical Self-Description
2 NHS Health and Advisory Service Questionnaire 92
3 (1995) 46 Physical Self-Perception Profile 92, 93,
4 non-adherence, exercise regimes 131 95
5 non-aerobic exercise, aerobic physical self-worth 99
6 comparison 24–5 physiological factors
7 Nottingham Health Profile (NHP) 64, adaptations 165
65 exercise addiction 136–7
8
mechanisms 83–4, 164
9 plausibility (Hill’s criterion) 59
obligatory exercising see exercise
40111 dependence POMS see Profile of Mood States
1 odds ratios, depression/inactivity 50 questionnaire
2 Office of Health Economics, mental population studies
3 illness costs 3 depression 48–52, 155, 156
4 Office of Population Censuses and mood/emotion 76, 77, 80, 82
45111 Surveys 46 self-esteem 108–10
Index 203
1111 positive addiction psychosocial stress reactivity
2 definition 8 exercise effects 33, 34–6, 37–40
3 primary exercise dependence 132 exercise training effects 44–5
4 Positive and Negative Affect Schedule public health, exercise 1
(PANAS) 68, 70 PWB see psychological well-being
5
practice recommendations
6 qualitative reviews
exercise dependence treatments
7 142 anxiety/stress 16
8 health promotion 85 depression 53
9 self-esteem 114–15 mood/emotion 71, 72–5
1011 primary ageing, definition 119, 121
1 primary exercise dependence 132–4, random assignment research method
2 135, 136–7, 138–41, 142–3 18
3111 definition 8 randomised controlled studies
4 Profile of Mood States (POMS) (RCSs)
5 questionnaire 16, 67, 68, 70, 77, cost-effectiveness 161
84 exercise treatment for clinically-
6
programmes of exercise defined depression 54–5, 56–7
7 self-esteem 98, 100–7
adoption phase 130–2
8 non-adherence 131 reaction times (RTs)
9 self-esteem 111–12, 130–2 cognitive performance 119
20111 promotion of physical activity 5, 41–2, older adults 121–2, 123
1 166–7 reactivity to stress see psychosocial
2 psychiatric patients, fitness 48 stress reactivity
3 psychological dysfunction reasons for exercise 1–9
4 definition 8 Rejeski, W.J. 63, 64, 66
5 eating disorders 144–53 research
6 exercise addiction 130–43 guidelines 84–5
research findings 158–9 limitations 159–61, 164
7
psychological mechanisms study designs 160
8 research; see also research
eating disorders 144–53
9 exercise addiction 130–43 directions; research findings;
30111 fitness/mental health relationship reviews
1 164 research directions
2 physical activity-related anxiety/stress 43–4
3 130–53 cognitive performance 128–9
4 psychological well-being (PWB) eating disorders 152, 153
5 factors of 71 exercise addiction 143
6 measurement 70 general 162–3
7 mood 67 mood/emotion 84–5
PA studies 76, 77, 78–9 self-esteem 114, 116–17
8
research guidelines 84–5 research findings
9 anxiety/stress 154–5
psychomotor therapy, Belgium 5
40111 psychophysiological reactivity, stress anxiety/stress studies 42
1 11, 13 cognitive functioning 158
2 psychosocial mechanisms, cognitive cognitive performance/fitness
3 performance/fitness 128 relationship 129
4 psychosocial model, eating disorders depression 155–6
45111 151 eating disorders 153
204 Index
1111 mood 156–7 global 99
2 self-esteem 116, 157–8 intervention research 98–9
3 summarised 154–8 low 110
4 reviews measurement 90, 92–3
anxiety/stress studies 12–41, 14–15, middle-aged adults 109
5
19–23, 28–31, 34–6, 38–40 models 90, 91, 92, 93, 94, 94
6 older adults 109–10
cognitive performance 121–9
7 criteria for selection 16 randomised controlled studies 98,
8 depression 48–57, 53, 54–5, 56–7 100–7
9 eating disorders 145–9 research findings 116, 157–8
1011 exercise dependence 138–41 reviews 96–9, 100–7, 108–15
1 limitations 16–17 significance 88–9
2 methodological difficulties 56–7 weight loss 112
3111 mood/emotion 71, 72–5, 77, 81 young adults 108–9
4 self-esteem 96–9, 100–7, 108–15 self-help strategies, addiction 143
5 risk self-perception 88–117
depression development 51 constructs 89, 92
6
sedentary lifestyle 1 definition 8
7 measurement 90, 92–3
Rosenberg scale 99
8 RT see reaction time self-reported anxiety
9 Running Addiction Scale 138, 140, cross-sectional studies 17
20111 141 meta-analyses 11, 13
1 Russell, J.A. 67, 70 SF-36 questionnaire 64, 65
2 single exercise sessions, state anxiety
3 SAI see State Anxiety Inventory effects 44
4 Saving Lives: Our Healthier Nation situational anxiety see state anxiety
5 (1999) 3 slowness of response, ageing 118, 120
6 SCAT see Sports Competition Anxiety smoking 41
Test social physique anxiety (SPA) 25,
7
schools, physical activity 2 131
8 definition 8
search methods 16, 53, 63, 96, 118
9 secondary ageing Somerset Health Authority (SHA) 5
30111 definition 119, 120–1 Sonstroem, R.J. 93, 95, 98
1 mechanisms 128 Sothmann, M.S. 33
2 secondary exercise dependence 143–4, SPA see social physique anxiety
3 145–9, 150–2 specificity (Hill’s criterion) 58
4 definition 8 Spirduso, W.W. 121, 122, 123, 127
5 sedentary lifestyle, risk 1 sport, definition 8
6 SEES see Subjective Exercise Sports Competition Anxiety Test
7 Experiences Scale (SCAT) 148
self director, construct 89 standard scores, cognitive tests 126
8
self-acceptance 94 state anxiety
9 characteristics 11
self-concept 89–90, 91, 98, 99
40111 self-efficacy measures 93 definition 154
1 self-esteem 88–117 research findings 155
2 adolescents 108 single exercise sessions 44
3 children 108 State Anxiety Inventory (SAI) 28–31
4 definition 8 Stephens, T. 80
45111 exercise-based promotion 93–6 Steptoe, A. 70, 78, 80, 84
Index 205
1111 Strain Questionnaire 17 Trait Anxiety Inventory (TAI) 19, 20,
2 strength of association (Hill’s criterion) 22, 23, 26, 36
3 57 transient negative effects 130
4 strength of exercise effects, meta- TSCS see Tennessee Self-Concept
analyses 12–13, 14–15 Scale
5
stress
6
definition 9, 10–11 unhealthy exercise behaviours 130–53
7 physical activity studies 10–45 United Kingdom
8 psychophysiological reactivity to 11 British Association of Sport and
9 research findings 154–5 Exercise Sciences (BASES) 6
1011 stress reactivity see psychosocial stress British controlled experimental trials
1 reactivity 78–9
2 studies see reviews British population surveys 76, 77,
3111 sub-clinical depression 81 80
4 Subjective Exercise Experiences Scale British Psychological Society (BPS)
5 (SEES) 69, 70, 84 6
suicide risk 3, 89 drug treatment for depression 56
6
sympathetic arousal hypothesis 137 US National Institute of Mental Health
7 4
Szabo, Attila 130–53, 133, 134
8
9 TAI see Trait Anxiety Inventory visual analogue scale, anxiety
20111 task goal orientation 83 assessment 24
1 Taylor, Adrian 10–45
2 Taylor Manifest Anxiety Scale 20, 21, weight loss
3 25 eating disorders 151
4 temporal sequence (Hill’s criterion) 58 self-esteem 112
5 Tennessee Self-Concept Scale (TSCS) weight training, acute studies 32
6 92, 99 Weyerer, S. 51–2
‘time-out’ hypothesis 32–3 withdrawal symptoms 133, 134, 135
7
trait anxiety definition 9
8 women, depression 49
characteristics 11
9 chronic exercise 16 working days lost, Department of
30111 definition 154 Health figures 10
1 meta-analyses 13
2 reduction by exercise practitioners Yates, A. 144, 149, 150
3 41 young adults, self-esteem 108–9
4 research findings 42, 155
5 treatment comparisons 26 Zaitz, D. 142, 143
6
7
8
9
40111
1
2
3
4
45111

02 Physical Activity and Psychological Well-Being

Uploaded by

Copyright:

Available Formats

02 Physical Activity and Psychological Well-Being

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

02 Physical Activity and Psychological Well-Being

Uploaded by

Copyright:

Available Formats

1111

2 Physical Activity and Psychological

Author No. of Number of Delimitations of review Overall ES Key moderators

Author No. of Number of Delimitations of review Overall E.S. Key moderators

Author(s) Participants Design Comparison group Exercise Anxiety Outcome/comments

Author(s) Participants Design Comparison group Exercise Anxiety Outcome/comments

Author(s) Participants Design Stressor Exercise characteristics Anxiety Outcome/comments

Author(s) Participants Design Stressor Exercise characteristics Anxiety Outcome/comments

Author(s) Participants Design Stressor Exercise characteristics Anxiety Outcome/comments

Author(s) Participants Design Stressor Exercise characteristics Anxiety Outcome/comments

Authors/location Participants Design Treatment groups Measures Results (statistically

Label Reference(s) Measures Comments

Nottingham Health Hunt et al. (1986) • energy

EuroQol Buxton et al. (1990, 1992) • mobility

Instrument Reference Measures Comments

• anger • time instructions can be varied

Study Review design and scope Results and conclusions

Study Review design and scope Results and conclusions

Study Survey design and scope Results and conclusions

Study Participants Design and treatment Results and conclusions

All participants exercised for four Low intensity exercise produced

Significant Particular Physical Physical

Figure 5.1 A hierarchical model of self-concept.

Author(s) Date Subjects Groups Treatment Instruments Results

Author(s) Date Subjects Groups Treatment Instruments Results

Desharnais 1993 male and female Running RSES Positive changes

Author(s) Date Subjects Groups Treatment Instruments Results

Author(s) Date Subjects Groups Treatment Instruments Results

Authors Participants Design Treatment Cognitive Cognitive Fitness results

Blumenthal n = 28 males 3 months training jogging, strength RT in a no change 15% increase in

Hawkins n = 37 males 10 weeks training swimming, time-sharing, increase in not measured

Chapman & 47 runner Running Addiction Scale addiction scores were

Author(s) Participants Objectives Measurements Conclusion about the

Author(s) Participants Objectives Measurements Conclusion about the

Author(s) Participants Objectives Measurements Conclusion about the

You might also like