Exercise for women receiving adjuvant therapy
of breast-cancer: a systematic review
Vorgelegt von
Diplom-Biologin
Martina Markes
aus Köln
an der Technischen Universität Berlin
Fakultät VII – Wirtschaft und Management
zur Erlangung des akademischen Grades
Doktorin der Gesundheitswissenschaften / Public Health
Dr. P.H.
Genehmigte Dissertation
Promotionsausschuss:
Vorsitzender: Prof. Dr. med. Reinhard Busse
Gutachterin: Prof. Dr. phil. Ulrike Maschewsky-Schneider
Gutachter: Prof. Dr. med. Karl-Ludwig Resch
Tag der wissenschaftlichen Aussprache: 9.12.2010
Berlin 2011
D 83
Abstract
Abstract
Exercise has become an integral part of breast cancer rehabilitation. A growing
body of evidence shows health benefits such as increased physical fitness and
reduced fatigue through exercise, not only after breast cancer treatment has
finished, but during treatment as well.
The aim of this systematic review was to determine the effectiveness of aerobic
and resistance exercise interventions during adjuvant treatment of breast can-
cer with respect to physical functioning and health-related physical fitness,
among other important health outcomes. The findings of controlled trial stud-
ies of aerobic, strength, and combined exercise interventions among women
undergoing adjuvant treatment of breast cancer were critically evaluated and
summarised. The potential harm associated with exercise in this vulnerable
period was also examined and discussed. This review was conducted in co-
operation with the Cochrane Breast Cancer Group, and followed the rigorous
review methodology of the Cochrane Collaboration.
Results indeed indicate that important physical and mental health outcomes,
such as health-related physical fitness, fatigue, anxiety, and depression, can be
improved through regular exercise training. There is no evidence currently in
the literature showing an association of moderate physical exercise training
during adjuvant treatment of breast cancer with increased risk of injuries, lym-
phedema, or other harm. Thus, the results of this systematic review support
recommending and encouraging women undergoing adjuvant treatment of
breast cancer to exercise.
If the benefits of participation in exercise programmes are to be preserved over
the course of cancer survivorship, sustained physical exercise is essential. Exer-
cise adherence thus plays a vital role in maintaining the benefits associated with
exercise. The review concludes that understanding the determinants of exercise
adherence and the barriers to participation is important. Applying strategies for
behaviour change to individual situations can help in developing and maintain-
ing the habits associated with a healthy lifestyle. Besides developing the evi-
dence base through further research, it is recommended that further activities
should focus as well on how research findings can be translated in to public
health practice.
Table of contents
I
Table of Contents
List of figures _______________________________________________IV
List of tables ________________________________________________ V
List of abbreviations, acronyms and units of measurement___________VI
Executive Summary ___________________________________________ 1
1 Introduction _____________________________________________ 5
2 Setting the context ________________________________________ 8
2.1 Logic model of the review__________________________________ 8
2.2 Breast cancer treatment __________________________________ 12
2.3 Rationale for exercise interventions ________________________ 13
2.3.1 Health-related physical fitness__________________________________ 13
2.3.2 Treatment-related symptoms___________________________________ 19
2.3.3 Biological and physiological problems ___________________________ 21
2.3.4 Mental health_______________________________________________ 23
2.3.5 Health-related quality of life ___________________________________ 23
2.3.6 Recurrence and survival ______________________________________ 26
2.3.7 Prevalence of treatment-related adverse effects ____________________ 26
2.4 The intervention: exercise ________________________________ 27
2.4.1 Exercise patterns in breast cancer patients ________________________ 27
2.4.2 Behaviour change theories for program planning ___________________ 28
2.4.3 Exercise adherence __________________________________________ 30
2.4.4 Exercise prescription _________________________________________ 32
2.4.5 Exercise Testing ____________________________________________ 33
2.5 Potential harm associated with exercise _____________________ 34
2.6 Previous research _______________________________________ 36
2.7 Preliminary work: rehabilitation sport in Germany ___________ 37
3 Methods _______________________________________________ 39
3.1 The role of systematic reviews in evidence based medicine______ 40
3.2 Execution of the review___________________________________ 41
3.2.1 Review protocol ____________________________________________ 41
3.2.2 Identification of research______________________________________ 41
3.2.3 Search updates within duration of review implementation ____________ 44
3.2.4 Selection of studies __________________________________________ 45
3.2.5 Data extraction _____________________________________________ 47
3.2.6 Study quality assessment______________________________________ 49
3.2.7 Meta-analysis_______________________________________________ 55
3.3 Methods for the assessment of harm ________________________ 60
4 Results: Identification and description of studies_______________ 63
4.1 Studies of benefit ________________________________________ 63
4.1.1 Process of retrieval of included trials in basic search ________________ 66
4.1.2 Process of retrieval of ongoing trials_____________________________ 67
4.1.3 Characteristics of excluded studies ______________________________ 68
4.1.4 Characteristics of included studies ______________________________ 70
4.2 Studies of harm _________________________________________ 86
Table of contents
II
5 Results: Evidence synthesis ________________________________ 89
5.1 Physical health effects____________________________________ 89
5.1.1 Physical functioning _________________________________________ 89
5.1.2 Health-related physical fitness__________________________________ 90
5.1.3 Treatment-related symptoms___________________________________ 92
5.1.4 Biological and physiological effects _____________________________ 93
5.2 Mental health effects_____________________________________ 95
5.2.1 Emotional distress ___________________________________________ 95
5.2.2 Positive psychological function_________________________________ 95
5.3 Quality of life effects _____________________________________ 96
5.3.1 Cancer- and cancer-site-specific quality of life _____________________ 96
5.3.2 Generic health-related quality of life_____________________________ 96
5.4 Adherence and contamination _____________________________ 96
5.5 Forest plots for immediate post-intervention effects __________ 100
5.6 Sustainability of effects__________________________________ 106
5.6.1 Exercise behaviour _________________________________________ 106
5.6.2 Long-term intervention effects ________________________________ 107
5.6.3 Recurrence and survival _____________________________________ 108
5.6.4 Forest plots for long-term intervention outcomes __________________ 109
5.7 Harm from exercise interventions _________________________ 111
5.8 Perception of exercise during breast cancer treatment ________ 112
6 Discussion_____________________________________________ 113
6.1 Summary of main results ________________________________ 113
6.2 Applicability and overall completeness of evidence___________ 115
6.3 Strengths and limitations of included studies________________ 116
6.3.1 Inconsistencies of effects across studies _________________________ 117
6.3.2 Replicability of the intervention and mediational pathways __________ 118
6.4 Potential biases in the review process ______________________ 120
6.5 Agreements with other reviews ___________________________ 121
7 Conclusions and forecast_________________________________ 122
7.1 Implications for research ________________________________ 122
7.2 Implications for practice_________________________________ 123
7.2.1 Setting up exercise classes____________________________________ 123
8 Bibliography___________________________________________ 126
9 Appendices ____________________________________________ 138
9.1 Appendix 1 – Protocol___________________________________ 138
9.2 Appendix 2 – Search Activities ___________________________ 149
9.2.1 Journals handsearched_______________________________________ 149
9.2.2 Conference proceedings searched ______________________________ 149
9.2.3 Study registers searched _____________________________________ 150
9.2.4 Experts contacted __________________________________________ 150
9.2.5 Search strategy for electronic databases _________________________ 151
9.3 Appendix 3 – Study eligibility form________________________ 154
Table of contents
III
9.4 Appendix 4 – Data extraction forms _______________________ 155
9.4.1 Data extraction of study descriptors ____________________________ 155
9.4.2 Instructions for extraction of study descriptors ____________________ 156
9.4.3 Effect size data coding_______________________________________ 158
9.5 Appendix 5 – Study quality forms_________________________ 159
9.5.1 Methodological quality form__________________________________ 159
9.5.2 Code of practice for implementing methodological quality criteria ____ 160
9.5.3 Intervention quality form_____________________________________ 162
9.6 Appendix 6 – Identification/critical appraisal of studies of harm 163
9.6.1 Medline search strategy for studies of harm (WinSPIRS 5.0)_________ 163
9.6.2 Data extraction form for studies of harm_________________________ 164
9.6.3 Quality assessment form for studies of harm _____________________ 165
9.7 Appendix 7 – Reference list of excluded studies of benefit _____ 166
9.8 Appendix 8 – Included studies of benefit ___________________ 168
9.8.1 Study descriptors and effect size statistics________________________ 168
9.8.2 Intervention quality _________________________________________ 214
9.9 Appendix 9 – Characteristics of studies adressing harm_______ 215
9.9.1 Reference list of included studies of harm _______________________ 215
9.9.2 Reference list of excluded studies of harm _______________________ 215
9.9.3 Quality of studies of harm ____________________________________ 216
9.9.4 Study descriptors and effect size data of studies of harm ____________ 217
9.10 Appendix 10 – Outcome assessment instruments_____________ 219
9.11 Appendix 11 – Effect size calculation: inputs into meta-analyses 221
9.12 Appendix 12 – Forest plots_______________________________ 222
9.12.1 Immediate post-intervention outcomes________________________ 222
9.12.2 Long-term outcomes______________________________________ 225
Table of contents
I
V
List of figures
Figure 1: Logic model for evaluation of exercise interventions..........................11
Figure 2: Breast cancer treatment overview...........................................................14
Figure 3: A functional performance framework....................................................15
Figure 4: Work breakdown structure ......................................................................42
Figure 5: Process flow of review steps....................................................................48
Figure 6: Flow diagram of study selection process ...............................................64
Figure 7: Meta-analysis for cardiorespiratory fitness ..........................................100
Figure 8: Meta-analysis for strength......................................................................101
Figure 9: Meta-analysis for strength with resistance exercise training studies 101
Figure 10: Meta-analysis for body composition ..................................................102
Figure 11: Meta-analysis for cancer-related fatigue.............................................103
Figure 12: Meta-analysis for cancer-related depression......................................104
Figure 13: Meta-analysis for anxiety......................................................................104
Figure 14: Meta-analysis for cancer-site-specific quality of life.........................105
Figure 15: Meta-analysis for physical activity.......................................................109
Figure 16: Meta-analysis for long-term effect on fatigue ...................................109
Figure 17: Meta-analysis for long-term effect on depression............................110
Figure 18: Meta-analysis for long-term effect on cancer-specific quality of life
....................................................................................................110
Table of contents
V
List of tables
Table 1: Inclusion criteria for reviewing benefits of exercise..............................46
Table 2: Quality criteria assessed with the vanTulder scale .................................51
Table 3: Criteria for assessing quality of aerobic endurance training .................53
Table 4: Criteria for assessing quality of muscular endurance training..............54
Table 5: Inclusion criteria for reviewing harm associated with exercise............61
Table 6: Reference list of included studies.............................................................65
Table 7: Identification of included studies per database in the basic search.....67
Table 8: Origin of reports of included trials in the basic search.........................67
Table 9: Origin of ongoing trials..............................................................................68
Table 10: Characteristics of excluded studies.........................................................68
Table 11: Description of interventions and study characteristics.......................73
Table 12: Outcomes reported ..................................................................................78
Table 13: Quality criteria met by studies.................................................................81
Table 14: Number of studies meeting individual quality criteria .......................82
Table 15: Characteristics of included studies.........................................................83
Table 16: Characteristics of studies of harm..........................................................88
Table 17: Summary of findings: immediate post-intervention effects..............99
Table 18: Summary of findings: long-term effects..............................................108
Table 19: Summary of findings: harm...................................................................112
Table of contents
V
I
List of abbreviations, acronyms and units of measurement
ACSM American College of Sports Medicine
AET Aerobic exercise training
ANOVA Analysis of variance
BDI Beck Depression Inventory
%BF Body fat percentage
BMD Bone mineral density
CACE Complier-Average Causal Effect
CBCG Cochrane Breast Cancer Group
CES-D Center for Epidemiological Studies – Depression Scale
CBCSR Cochrane Breast Cancer Group Specialised Register
CDSR Cochrane Database of Systematic Reviews
CI Confidence interval
CINAHL Cumulative Index to Nursing and Allied Health Literature
CONFSCI Conference Papers Index
CONSORT CONsolidated Standards of Reporting Trials
DBS German Federation for Disabled Sports,
Deutscher Behindertensportverband
df Degrees of freedom
DOSB German Sports Federation, Deutscher Olympischer Sportbund
FACT Functional Assessment of Cancer Therapy instrument
FACT–An Functional assessment of cancer therapy – Anemia scale
FACT–B Functional assessment of cancer therapy – Breast cancer scale
FACT–ES Functional assessment of cancer therapy – Endocrine symptom
scale
FACT–F Functional assessment of cancer therapy – Fatigue
FACT–G Functional assessment of cancer therapy – General scale
g Hedge’s g
HRmax Maximum heart rate
HRreserve Heart rate reserve
ICC Intra-class correlation coefficient
I2 Degree of inconsistency across studies in meta-analysis
IQR Interquartile range
LBM Lean body mass
M Mean
Min minutes
12 MWT 12-minute walk test
Mo months
NA Negative affects
NCCN National Comprehensive Cancer Network
NKCA Natural Killer Cytotoxic Activity
PFS Piper Fatigue Scale
Table of contents
V
II
POMS Profile of mood states
PA Positive affects
PANAS Positive and Negative Affect Schedule
PSQI Pittsburgh Sleep Quality Index
Q Cochran’s Q
RCT Randomised controlled trial
RET Resistance exercise training
RevMan RevMan Analyses (Computer Programme)
1-RM One repetition maximum
RPE Rating of Perceived Exertion
r-PFS Revised Piper Fatigue Scale
RR Relative Risk
SAS Symptom assessment scale
SD Standard deviation
SD Self-directed exercise training
21 m-m
SE Standard error of the difference in means
SF-36 Medical Outcomes Study 36-Item Short Form
SGB IX Book 9 of the German social code, Sozialgesetzbuch
SIGLE System for Information on Grey Literature
SMD Standardised mean difference
SPSS Statistical Product and Service Solutions
STAI Spielberger State Anxiety Inventory
SU Supervised exercise training
TMD Total mood disturbance
TNM Tumour node metastasis system
VAS Visual analogue scale
VO2max Maximal oxygen consumption
WHO World Health Organization
WHOQOL World Health Organization Quality of Life project
Wk weeks
WISE Women International Space Simulation for Exploration
WMD Weighted mean difference
Yr years
Executive Summary
1
Executive Summary
Breast cancer remains an important public health problem in Europe and the ageing of the
European population will cause cancer incidence data to continue to increase. The principal
treatments for breast cancer are surgery, chemotherapy, radiotherapy, and hormonal therapy, and
evidence suggests that these are very effective at improving disease-free and overall survival.
However, these therapies can compromise women’s physical, mental, or social health. Evidence
continues to accumulate on the positive impact of exercise on women’s physical or mental health
and on quality of life improvements during treatment. The literature suggests, however, that
many women who exercised prior to being diagnosed with breast cancer do not continue to
exercise during their treatment. Over time, prolonged inactivity leads to decreased muscle
strength, progressive loss of physical functioning and a decreased ability to perform the activities
of daily living. Women, clinicians, and health policy makers need reliable, up-to-date information
from controlled trials on the benefits and potential harm from exercise during adjuvant cancer
treatment, to make evidence-based decisions about interventions. At present, a diverse range of
primary studies exists in the scientific literature regarding the role of exercise during treatment for
breast cancer, and this systematic review seeks to summarise this evidence to date.
Methods of the systematic review
This review was conducted in co-operation with the Cochrane Breast Cancer Group, and fol-
lowed the rigorous review methodology of the Cochrane Collaboration. The research question is:
should women undergoing adjuvant treatment of breast cancer be encouraged to exercise? Study
identification was based on a comprehensive search strategy with a variety of commonly used
electronic databases and manual search methods. Inclusion and exclusion criteria were developed
in accordance with the review question, and were defined in terms of the population, interven-
tions, outcomes, and the study designs of interest. Trials were included that reported on women
receiving adjuvant treatment (chemotherapy, hormonal therapy, or radiotherapy) for breast can-
cer. Breast cancer was restricted to stages 0-III, and trials which included women with stage IV
breast cancer (i.e., with distant metastasis) were excluded from the review. Trials that included
women who had completed adjuvant cancer treatment, or who were being treated for other can-
cers were excluded. Trials with an intervention consisting of aerobic or resistance exercise were
included, but those that examined complex exercise interventions (e.g., a program of exercise and
diet, or a program of exercise and behavioural therapy) were excluded. Trials in which exercise
interventions were restricted to selected body functions only (e.g., arm mobility) were also ex-
cluded. Trials were included that employed at least one of the following outcome measures:
physical functioning, health-related physical fitness, symptom experience, biological or physio-
logical outcomes, mental health, health-related quality of life and harm. Both randomised con-
Executive Summary
2
trolled trials and non-randomised controlled trials were eligible for inclusion in the review. No
language restrictions were applied. All studies were critically and systematically evaluated with
respect to their methodological quality (i.e., design, implementation and analysis), to determine
the extent to which the results were reliable. The exercise intervention was evaluated separately
regarding its potential to provide an adequate training stimulus. Data from included trials were
extracted following a standardised format, and then combined using meta-analysis; this is a statis-
tical procedure that integrates the results of several independent studies to give one overall esti-
mate of intervention effects. A random-effects model was chosen for the meta-analysis since
heterogeneity between trials was expected.
In order to answer the research question, whether women receiving adjuvant treatment of breast
cancer should be encouraged to exercise, the benefit of exercise has to outweigh the potential
harm. This review aimed to evaluate both benefits and harms, and has subsequently required a
more complex design, using different search strategies and eligibility criteria for studies of harm
to handle different sets of studies for various outcomes. Study designs eligible for assessing po-
tential harm were (randomised) controlled trials to identify well-recognised and easily detectable
harmful effects. The study of harm had to be a key trial objective, which means that the harm
had to be a major primary or secondary outcome of trials to be included. All studies were eligible
that investigated harmful effects in the rehabilitation of breast cancer, either during or after adju-
vant cancer treatment.
Main results
Fifteen studies that had assessed the benefits of aerobic or resistance exercise training (or both)
on physical and mental health outcomes in women during adjuvant treatment for early breast
cancer were included in the review. The results were based on 1,042 participants from these 15
included studies. These studies were predominantly performed in North America (United States
and Canada); only two of these trials, a pilot study followed by a subsequent RCT, were imple-
mented in Europe. Sample sizes across trials ranged from 10 to 242 participants; only recently
have large-scale trials of 100 participants or more emerged in the literature, with two of these
including more than 200 participants.
With respect to health-related physical fitness, meta-analyses were performed for cardiorespira-
tory fitness, muscular fitness (strength) and body composition. Exercise was shown to be an
effective intervention for improving cardiorespiratory fitness in comparison to usual care (SMD
0.54; 95% CI 0.32 to 0.77), even during breast cancer treatment. This medium effect was based
on the results from eight studies with a total of 709 participants. Pooling the data from four
studies, with a total of 328 participants, also yielded a medium effect size for a statistically signifi-
Executive Summary
3
cant increase in strength observed in exercising participants, compared to controls (SMD 0.42;
95% CI 0.06 to 0.78). Body composition outcomes, which describe the relative amounts of fat
and lean tissue, were pooled from four trials (n=414). Exercise was more effective than usual care
in preventing unfavourable changes in body composition (SMD -0.29; 95% CI -0.55 to -0.03)).
Self-reported physical functioning was measured in three studies but a meta-analysis for this
outcome was not performed due to limited available data. Consequently, this review is unable to
make any conclusions about the impact of exercise in self-reported physical functioning. Regard-
ing treatment-related symptoms, a meta-analysis for fatigue was undertaken. Pooling the data
from seven studies (n=714) that provided adequate data for fatigue showed that exercise was
more effective than usual care in reducing feelings of fatigue (SMD -0.17; 95% CI -0.32 to -0.02),
with the SMD indicating a small effect size. Bone health (bone mineral density) was the primary
outcome measure in one trial which reported that aerobic exercise preserved lumbar spine bone
mineral density better than usual care alone. For mental health outcomes, the effect sizes were
small. Pooling the effects of three trials yielded a reduction in cancer-related depression in the
exercise groups compared to controls (n=443) (SMD -0.24; 95% CI -0.43 to -0.04). Results for
anxiety suggested a small, but non-significant effect of exercise (n=269) (SMD -0.25; 95% CI -
0.54 to 0.04) from two studies.
Four studies were included in this review that assessed harm (lymphedema and injuries). Only
one of these studies assessed potential harm (i.e., lymphedema) during adjuvant treatment of
breast cancer. The other three studies examined harm in the post-adjuvant setting. No increases
in injury rates, lymphedema or other harm were revealed in the exercise groups compared to
non-exercising control groups. Thus, preliminary evidence from exercise intervention studies
indicates that there is no increased risk of harm associated with exercise.
Conclusions
Exercise during adjuvant treatment for breast cancer can be regarded as a supportive self-care
intervention; exercise results in improved physical fitness and thus the capacity for performing
activities of daily life, which may otherwise be impaired due to inactivity during treatment. Fur-
thermore, small improvements in fatigue, depression and anxiety can be observed.
The present available evidence regarding safety of exercise during adjuvant breast cancer treat-
ment is limited. However, safety concerns are highly relevant to exercise promotion in the period
of adjuvant cancer treatment because uncertainty about the safety of engaging in exercise may act
as a barrier to prescribing exercise despite the growing body of evidence which supports the
benefits of moderate intensity exercise during breast cancer treatment. It is also possible that
regular exercisers could be discouraged from being physically active during and after breast can-
Executive Summary
4
cer treatment by doctors, other health professionals and their families. Therefore, safety needs to
be systematically addressed in future exercise studies.
Since exercise interventions (for sedentary participants) require behaviour change, strategies for
behaviour change should underpin these interventions. Theory-based interventions are important
in the development and promotion of healthy physical activity habits since theories are a general-
ised and careful interpreted systematic summary of empirical evidence related to behaviour
change and thus, application of theory should improve the likelihood of effectiveness of exercise
interventions.
What is known about exercise in cancer patients has been derived largely from research on mid-
dle-aged women. Only one study has examined exercise during adjuvant therapy in older breast
cancer patients. It is unclear whether older women derive similar benefits from exercise. Possibly,
they have lower exercise participation rates and more difficulty adhering to an exercise program.
Exercise during adjuvant treatment should be approached from an aging perspective as well.
The included studies provide important guidance to cancer care clinicians as well as those who
operate community-based health and fitness programs. Exercise is becoming increasingly recog-
nised as beneficial to cancer patients during treatment. Of necessity, these studies were predomi-
nantly conducted as part of academic oncology care programs. Many breast cancer patients, how-
ever, approach community-based fitness centres for health promotion training. Finally, existing
research findings have to be used to develop evidence-based guidelines or prescriptions for can-
cer patients. It is recommended that exercise programs for breast cancer patients need to be
individualized. However, there is insufficient evidence to allow for the identification of factors
that have to be considered in the development of individualized exercise programs. There is a
need for an evidence-based set of exercise guidelines to be developed. The exercise rehabilitation
programs established in Germany for women who have been treated for breast cancer is an ex-
ample of how exercise could be offered to breast cancer patients during treatment within a health
care context. Germany has embraced the importance of regular exercise in the rehabilitation of
cancer patients and may provide a model that could be adopted in other countries.
Introduction
5
1 Introduction
Breast cancer is a major public health burden, both in Germany and worldwide. Breast cancer is
by far the most common cancer of women, with an estimated 1.15 million new cases worldwide
in 2002 (Parkin et al. 2005). Breast cancer incidence and mortality vary considerably by world
region. More than half of the cases are in industrialized countries—about 361,000 in Europe and
230,000 in North America 2002 (Parkin et al. 2005). In general, the incidence is high (greater than
80 per 100,000) in developed regions of the world and low (less than 30 per 100,000) in develop-
ing regions, although this latter rate is increasing. Variability in mortality rates is much lower
(approximately 6-23 per 100,000) because of the more favourable chances of surviving breast
cancer in the high-incidence developed regions (Parkin et al. 2005). There was a trend of increas-
ing breast cancer incidence almost everywhere, partly due to increases in risk factors such as
decreased childbearing and breast-feeding, increased exogenous hormone exposure, and detri-
mental dietary and lifestyle changes, such as obesity and reduced physical activity. However,
between 2001 and 2004, incidence rates of invasive breast cancer declined more than 8% in the
United States, with the greatest drops observed for estrogen receptor-positive tumours among
women aged 50 years and over (Ravdin et al. 2007). Patterns of falling incidence reflect a major
influence of reductions in hormone therapy use after the early termination of the Women’s
Health Initiative trial on postmenopausal hormone therapy (Hausauer et al. 2009). Also reduc-
tions in the pool of previously unscreened women due to the saturation of mammographic
screening programs could have lowered incidence (Hausauer et al. 2009). Breast cancer-related
mortality is now decreasing in many high-risk countries due to a combination of intensified early
detection efforts and advances in treatment (Parkin and Fernandez 2006).
In Germany, over 57,000 women are currently diagnosed with breast cancer every year. Breast
cancer accounts for well over a quarter (27.8%) of all cancers among women. The average age at
onset is 63, six years below the average for all cancer sites (Batzler et al. 2008). Overall, breast
cancer incidence in Germany has been rising continuously since 1980, while the mortality rate has
been falling slightly since the mid-1990s (Batzler et al. 2008) but it remains unclear whether this
decrease is due to early diagnosis or improved treatment (Giersiepen et al. 2005).
Being diagnosed with breast cancer usually means undergoing significant and prolonged medical
treatments. Although the benefit of current treatments for breast cancer is clear, namely im-
proved survival, treatment-related adverse effects are of considerable importance. In the course
of diagnosis and treatment of breast cancer, women are likely to experience situations that cause
considerable distress. Not only clinical factors (e.g., type of treatment, presence of pain) and
social factors (e.g., availability of support from friends and family) contribute to the distress, but
Introduction
6
also role changes may be needed due to the potential inability to continue work and dependency
on others; life goals may be disrupted or life plans may need to be modified facing a life-
threatening illness. For women, issues related to body image are crucial: body image for women
includes feeling feminine and attractive or enjoying the body as a symbol of social expression.
Breasts are a symbol of femininity and sexuality, and thus a woman with breast cancer may feel
that her body has betrayed her, or the loss of a breast can feel like the end of being female.
Women with breast cancer do not only have to cope with changes in physical appearance; often
they no longer perceive their body as an intact, properly functioning entity. Breast cancer does
not just lead to surgery scars: a loss of feeling in the affected breast may be a further consequence
of surgery, radiotherapy can lead to redness and soreness on the affected area and chemotherapy
often causes hair loss, weight gain, and premature menopause. Experiencing these types of body
changes can be especially challenging for younger women, who may be bothered more by these
changes than older women. Breast cancer can result in a subsequent dissatisfaction with appear-
ance, perceived loss of femininity, and body integrity, reluctance to look at oneself naked, and
feeling less sexually attractive. A woman dealing with the effects of breast cancer may begin to
avoid intimacy, dress alone or in the dark, or even minimise the time that she spends bathing.
Such concerns about body image can undermine women as they try to adjust. Altogether, diagno-
sis and treatment of breast cancer affects women physically as well as psychologically. There is a
need for adequate attention to promoting functioning and psychological well-being among
women receiving adjuvant treatment of breast cancer.
Medical treatment focuses on beating cancer, but women whose lives are affected by breast can-
cer have to find how to live with it and what works best for them. Health promotion strategies
that place a focus on the restoration of physical, functional, emotional, and social aspects of
health can support women to realign. Health promotion strategies become imperative in the face
of the widespread use of adjuvant treatment with its increase in time, complexity and dose-
intensity, and the steadily decreasing mortality rates from breast cancer, with a five-year survival
rate from breast cancer of 79% in Germany (Bertz et al. 2006). In particular, exercise interven-
tions are increasingly prominent in the nursing, medical, and sports science literature that repre-
sents a growing body of knowledge, which suggests numerous benefits of regular physical exer-
cise for breast cancer patients. Furthermore, the many health and fitness benefits (e.g., physio-
logical, metabolic, psychological) associated with regular exercise, both in the general population
and in clinical populations, support the rationale advanced for exercise interventions in the popu-
lation of women receiving adjuvant treatment of breast cancer. According to the United States
Surgeon General’s report on physical activity and health (United States Department of Health
and Human Services 1996), people of all ages, both male and female, benefit from regular exer-
Introduction
7
cise. The extent and strength of the evidence linking physical activity to many health improve-
ments is impressive: amongst others, physical activity reduces the risk of developing diabetes,
hypertension, and colon cancer; it enhances mental health, fosters health-related physical fitness,
and helps maintain function and preserve independence in older adults (United States Depart-
ment of Health and Human Services 1996). It seems plausible that breast cancer patients as well
could maintain function and improve the quality of their lives through a practice of moderate
physical activity.
The overall question that guided the implementation of this research project was:
Should women, who are undergoing adjuvant treatment of breast cancer, be recommended physical exercise?
A systematic review appeared to provide an adequate methodology for this research project. The
Cochrane Breast Cancer Group (CBCG) was approached and provided inputs and peer review
for this research project. Accordingly, this systematic review of exercise for women receiving
adjuvant therapy of breast cancer was implemented as a Cochrane review, pursuing the respective
formal, methodological, and procedural features of a Cochrane review. The first result of the
cooperation with the CBCG was that an excerpt of this research project has already been pub-
lished in the Cochrane Library (Markes et al. 2006).
This systematic review followed three objectives. The primary objective was to assess the effects
of a structured exercise training program on the physical functioning and health-related physical
fitness in the target group of women, who are undergoing adjuvant treatment of breast cancer. A
secondary objective was to determine the effect of exercise training on secondary outcome meas-
ures such as treatment-related symptoms, biological and physiological problems, mental health,
and health-related quality of life. Given the potential for harmful effects of exercise interventions
in this vulnerable target group, a final objective was to identify and assess harm associated with
exercise during adjuvant treatment of breast cancer.
Setting the context
8
2 Setting the context
2.1 Logic model of the review
This systematic review of exercise for women receiving adjuvant treatment of breast cancer
started with a logic model in order to guide the review process and to ensure that the review
would yield relevant, useful information (Figure 1). Basically, a logic model provides a visual
representation of the conceptualisation of the overall topic area; it describes and illustrates ex-
pected relationships between pre-intervention conditions, activities and short- and long-term
outcomes. Logic models show the program logic and can help review authors understanding how
the program works. Logic models are also useful for identifying outcomes that need to be con-
sidered in the systematic review and thus can also aid in the design of a strategy for synthesizing
data across studies.
The logic model of this review model described the linkage between planned activities and their
expected outcomes. The following elements are acknowledged for and further outlined in the
following paragraphs:
• Problems that require intervention
• Inputs and resources for a successful exercise program
• Activities that have to be accomplished if access to those resources is given
• Outcomes that can be expected if exercise programs are delivered as planned.
The elements of the logic model also work as a framework for the information regarding the
context of the review.
Problems
Being diagnosed with breast cancer usually means undergoing a series of diverse treatments (sur-
gery, radiotherapy, systemic chemotherapy, and hormonal treatment). These treatments often
have a persistent deleterious effect on a woman’s health status, physical and emotional well-
being, and health lifestyle behaviours. Breast cancer rehabilitation programs are often based on
psychotherapy or social support. Such interventions do not usually deal with the physical prob-
lems faced by women receiving adjuvant treatment of breast cancer, such as fatigue, loss of
physical functioning, or weight gain. Exercise is a potential intervention that may both improve
health status and boost breast cancer patients’ well-being.
Inputs and resources
Implementing an exercise program requires sustainable financing, based on diverse sources of
funding and affordable fees. Qualified instructors, together with facilities for exercise that ac-
Setting the context
9
commodate and encourage participation, are equally important. Social marketing, networking
with a varied group of partners and cooperation with health care providers provide a basis for
recruiting participants and for delivering exercise programs (implementation of exercise classes).
Further, it is essential that certain management skills and resources are committed to the pro-
gram.
Outputs and activities
Based on these resources, an exercise program providing ongoing training can be established for
breast cancer patients. Exercise instructors are faced with the challenge to improve exercise hab-
its through appropriate exercise prescriptions and monitoring. Women receiving adjuvant treat-
ment of breast cancer may require further physical evaluation, i.e., pre-participation stress testing,
before starting the exercise program. Furthermore, specific instruction should be given to these
women concerning the type, frequency, intensity, and duration of exercise. A variety of safe,
effective, and attractive programs – with a range of offerings, including strength and aerobic
endurance training, flexibility, and balance elements – can facilitate the development of regular
exercise habits. Unfortunately, efforts to help participants adopt an exercise program are often
unsuccessful. This is why a variety of methods and tools need to be employed to reach partici-
pants – especially those who have been sedentary. Motivation to exercise can be improved and
attrition minimised, if program activities are guided through behaviour change strategies to sup-
port participation and personal commitment (e.g., pre-activity exercise counselling, individual
goal setting sessions, and telephone or mail follow-up of women with repeated absences).
Outcomes
There is a chain of outcomes, since not all outcomes will occur at the same time. Some outcomes
must occur before the achievement of other outcomes and program goals. It is expected that the
following outcomes will occur over the short-, intermediate, and long-term as a result of the
exercise intervention. First, exercise programming, together with exercise counselling and pre-
scription, allows breast cancer patients to change the way they perceive physical activity and
exercise during breast cancer treatment. So, as a short-term health promotion outcome, breast cancer
patients are expected to recognise that exercise is a safe, practical intervention that is effective in
combating treatment-related adverse effects and which can increase quality of life and health-
related fitness. Exercise self-efficacy, i.e., the belief and conviction that one can successfully exer-
cise, is expected to increase and the intent to regularly exercise should arise. “Health promotion
outcomes” such as knowledge about exercise and its benefits or exercise self-efficacy (Nutbeam
2000) are the most immediate outcomes: they represent those personal, social and structural
factors that can be modified in order to change intermediate health outcomes. Intermediate health
outcomes (Nutbeam 2000) are changes in the determinants of health, notably changes in lifestyle
Setting the context
10
such as exercise patterns. In the long-term regular exercise improves health outcomes such as physi-
cal functioning, health-related fitness and quality of life. Exercise-induced harms, such as muscu-
loskeletal injuries or cardiovascular complications during exercise, are generally expected to be
under control due to appropriate conditioning and the gradual increase in duration and intensity
of exercise, together with exercise testing, prior to engaging in a regular exercise program.
Conclusions for evaluating exercise programs
Inputs and outputs on the program level can be evaluated when focusing on structure and proc-
esses. Process evaluation, especially, could provide crucial information on participation rates and
program implementation (e.g., Were all sessions delivered? Did the program offer exercise op-
tions with instruction in proper technique, and provide qualified supervision?). Such information
would be crucial because a program can fail either because the intervention was not well suited to
the problems or because the intervention was not implemented as planned. However, these ques-
tions are not treated in this review because primary studies only dealt with outcome evaluation.
The focus of this review is the evaluation of health outcomes.
Setting the context
11
SITUATION
Women receiving treatment of breast cancer often experience reductions in health-related physical fitness and quality of life.
Exercise is an intervention that may improve these problems after diagnosis of breast cancer.
INPUTS
OUTPUTS
Activities Participation
OUTCOMES – IMPACT
Short Medium Long-term
¾ Organisational
commitment of
leadership, re-
sources and staff
¾ Networking
¾ Social marketing
¾ Sustainable fi-
nancing
¾ Qualified/ certi-
fied instructors
¾ Facilities for
physical activity
Exercise program
¾ Exercise testing
¾ Exercise prescrip-
tion
¾ Exercise facilitation
or exercise counsel-
ling
¾ Application of
psychological and
behavioural theories
to effect behaviour
change
¾ Addressing exercise
barriers and facilita-
tors
¾ Continuous exercise
training
¾ Sedentary
women re-
ceiving adju-
vant treat-
ment of
breast cancer
Health promotion
outcomes
Increase in:
¾ Knowledge about
exercise during
cancer treatment
¾ Attitudes, behav-
ioural intentions:
planning partici-
pation
¾ Exercise self-
efficacy
Intermediate health
outcomes
Increase in:
¾ Initial adoption of
exercise
¾ Continued partici-
pation or relapse/
resumption
¾ Exercise intensity,
frequency and du-
ration of sessions
as prescribed
Health outcomes
Increase in:
¾ Physical functioning
¾ Health-related fitness
¾ Health-related quality
of life
¾ Monitoring safety:
No Harm
Figure 1: Logic model for evaluation of exercise interventions
Setting the context
12
2.2 Breast cancer treatment
The mainstay of breast cancer treatment is surgery with possible adjuvant local and systemic
therapy. One important approach for the selection of adjuvant treatment is based on cancer
staging. The global standard in cancer staging is the tumour-node-metastasis (TNM) system
which represents an attempt to classify malignant tumours based on the major morphological
attributes which are thought to influence disease prognosis, namely: size of the primary tumour
(T), presence and extent of regional lymph node involvement (N), and presence of distant metas-
tases (M) (Singletary and Connolly 2006). Breast cancer is diagnosed as having progressed to any
of four stages: carcinoma in situ (stage 0), early stage breast cancer (stages I, II), locally advanced
breast cancer (stage III), and metastatic breast cancer (stage IV).
Depending on the staging and type of the tumour, possibly just a lumpectomy (removal of the
lump) or mastectomy (surgical removal of the entire breast) is needed. Although women with
lumpectomy do not have overt metastases at the time of staging, they remain at risk of local
recurrence and of metastatic spread: undetected micrometastatic deposits of the disease may
remain, either locally or at distant sites, that eventually develop into clinically detectable recur-
rence. These deposits can be treated with local or systemic adjuvant therapies.
Radiotherapy is delivered as an adjuvant local therapy; it is an essential component of breast
conserving therapy and is usually given after surgery has been performed. Radiotherapy consists
of the use of high powered X-rays that precisely target the area needing treatment. X-rays work
by eliminating the microscopic cancer cells that may remain near the area where the tumour was
removed, either in the chest wall or in regional lymph nodes. Evidence indicates a significant
decrease in local recurrence rates for patients receiving radiotherapy (Poortmans 2007). Radio-
therapy administered to the remaining breast tissue is typically delivered over a total duration of 6
weeks.
Systemic treatments include chemotherapy and hormonal therapy. Chemotherapy refers to the
use of cytotoxic drugs, which affect either cell division or DNA-synthesis and function. Chemo-
therapy is designed to eradicate microscopic deposits of cancer cells that may have spread from
the primary breast cancer. It has been shown to substantially improve the long-term relapse-free
and overall survival in premenopausal and postmenopausal women with both node-positive and
node-negative disease (Early Breast Cancer Trialists’ Collaborative Group 2005). Combination
chemotherapy involves treating a patient with a number of different drugs simultaneously which
differ in their mechanism and side effects. There are several different chemotherapy regimens
that may be used. Polychemotherapy regimens commonly comprise four to six cycles which are
delivered over four to six months. Most chemotherapy is delivered intravenously, although some
Setting the context
13
agents are administered orally. It may be given on either an inpatient or an outpatient basis de-
pending on the patient, the cancer, the stage of cancer, the type of chemotherapy, and the dosage.
Breast cancer may be endocrine-dependant needing oestrogen for proliferation and is thus ex-
pected to be responsive to endocrine treatment. Patients with oestrogen receptor-positive tu-
mours typically receive a hormonal treatment after chemotherapy is completed. The goal of hor-
monal therapy is to prevent the stimulation of breast cancer cells by oestrogen. Oestrogen depri-
vation can be achieved either by anti-oestrogens (e.g., tamoxifen) that bind to the oestrogen
receptor and thus block the receptors or by suppression of oestrogen synthesis (e.g., aromatase
inhibitors). Hormonal therapy for hormone receptor-positive breast cancer is generally delivered
over five years.
In sum, cancer treatments such as surgery, radiotherapy, chemotherapy, and hormonal therapy
reduce both the risk of relapse and death and are well-established treatments. They are offered to
a large proportion of patients (Kuerer et al. 2004, Chia et al. 2005, Early Breast Cancer Trialists’
Collaborative Group 2005).
Treatment options depending on diagnosis are summarised in the breast cancer treatment over-
view (Figure 2).
2.3 Rationale for exercise interventions
During primary treatment of breast cancer, a broad range of treatment-related adverse effects
occur which compromise physical health, mental health, social health, and subsequently impact
quality of life. In this chapter, it is described how primary treatment of breast cancer interferes
with patient-reported and objective health outcomes. Additionally, for each group of health out-
comes, the rationale for exercise intervention is depicted.
2.3.1 Health-related physical fitness
Physical fitness has been defined in different ways; however it is generally accepted to define
physical fitness as the “ability to carry out daily tasks with vigour and alertness, without undue
fatigue, and with ample energy to enjoy leisure-time pursuits and to meet unforeseen emergen-
cies” (United States Department of Health and Human Services 1996). Physical fitness thus in-
cludes cardiorespiratory endurance, skeletal muscular endurance, skeletal muscular strength,
skeletal muscular power, speed, flexibility, agility, balance, reaction time, and body composition
(United States Department of Health and Human Services 1996). Because these attributes differ
in their importance to athletic performance as compared to health, a distinction has been made
between performance-related fitness and health-related fitness (Caspersen et al. 1985).
Setting the context
14
Figure 2: Breast cancer treatment overview
(Alberta Breast Cancer Program 2003)
Health-related fitness includes cardiorespiratory fitness, muscular strength and endurance, body
composition, and flexibility. Skill-related components of physical fitness such as balance or coor-
dination are important pre-conditions for safe exercise. Balance for example relates to the “main-
tenance of equilibrium while stationary or moving” (United States Department of Health and
Human Services 1996) and thus is an important skill for preventing falls. Coordination relates to
the “ability to use the senses, such as sight and hearing, together with body parts in performing
motor tasks smoothly and accurately” (United States Department of Health and Human Services
1996); a lack of coordination could be perceived as a substantial barrier for exercise.
Setting the context
15
In women receiving adjuvant treatment of breast cancer, physical functioning is particularly im-
portant, as the prevalence of physical disability increases with age and can further be enhanced
during debilitating treatments. Adequate physical function plays a major role in maintaining par-
ticipation. Declining physical functioning contributes to the need of assistance in performing
basic tasks and may have various social consequences, at its worst permanent incapacity for work
or institutionalisation.
There is a progressive relationship between physical fitness, functional performance, and partici-
pation in life activities (Rikli and Jones 1997). Participation in life activities (e.g., personal care,
shopping, and travelling) requires the ability to perform functions such as walking and stair
climbing, that require a certain level of physical fitness (Figure 3). Exercise could improve physi-
cal functioning by increasing physical fitness.
Muscle strength/
endurance
Aerobic
endurance
Balance
Coordination
Speed/Agility
Power
Walking
Stair climbing
Standing up from
chair
Lifting/Reaching
Bending/Kneeling
Jogging/Running
Personal care
Shopping/
Errands
Housework
Gardening
Sports
Travelling
Physical fitness Mobility functions Participation in
life activities
Figure 3: A functional performance framework
(adopted from Rikli and Jones 1997)
Physical functioning can be measured through (1) objective mobility using performance tests
such as sit-to-stand-to-sit test, lift-and-reach test; (2) perceived mobility using self-report ques-
tionnaires, and (3) participation in life activities using self-report questionnaires as well (Bennett
et al. 2006). Performance-based methods measure what the patient can do, and self-reports meas-
ure what the woman thinks she can do (Terwee et al. 2006). While self-report questionnaires have
the advantages of being easier to use, less time-consuming, less of a burden to patients, and are
not influenced by observer bias, performance-based methods are described as being less influ-
enced by psychological factors such as expectations and beliefs, cognitive impairments, culture,
language, and education level (Terwee et al. 2006). Furthermore, it has been suggested that per-
formance-based measures may identify deficits in physical functioning before they can be identi-
fied by self-report (Bennett et al. 2006, Terwee et al. 2006).
Setting the context
16
2.3.1.1 Cardiorespiratory fitness
Cardiorespiratory fitness is a health-related component of physical fitness that relates to the abil-
ity of the circulatory and respiratory systems to supply oxygen during sustained physical activity
(United States Department of Health and Human Services 1996). Cardiorespiratory endurance
can be severely impaired as a consequence of breast cancer and its treatment through several
pathological mechanisms. To begin with, chemotherapy can damage bone marrow and subse-
quently impair the production of erythrocytes, so that the resulting anaemia leads to a compro-
mised oxygen saturation of blood (Dimeo 2001). Then cardiotoxic chemotherapeutic agents such
as anthracyclines and cyclophosphamide can cause a reduction of cardiac output (Dimeo 2001).
There are other functional and anatomic changes due to cancer treatment that may affect oxygen
transport and utilisation, such as alterations of the bronchial tree, lung and plasma volume, alveo-
lar surface, pulmonary perfusion, and concentration of oxidative enzymes (Dimeo 2001).
As a consequence of these treatment-related changes, more cardiac and respiratory work is
needed to sustain an adequate oxygen supply to the cells during rest and physical activity. How-
ever, the oxygen transport systems can be overwhelmed even by normal daily activities such as
climbing stairs or housekeeping. Sedentary habits and bed rest aggravate this further with reduced
muscle blood flow, red cell volume, capillarization and oxidative enzymes (Convertino et al.
1997) causing a shift away from aerobic metabolism to anaerobic glycolysis. Anaerobic glycolysis
is a less effective means of energy production; moreover, anaerobic glycolysis contributes to
metabolic acidosis since lactic acid is one of its end products (Dimeo 2001). The combined bur-
den of increased heart rate and respiratory work, shift of aerobic to anaerobic metabolism, and
metabolic acidosis can lead to the inability to carry out any intense physical effort.
Evidence suggests that healthy older adults elicit the same 10% to 30% increases in maximum
oxygen consumption with prolonged endurance exercise training as young adults (Balady et al.
2000), which may be true for women undergoing adjuvant treatment of breast cancer as well.
This increase in maximal oxygen uptake in older adults is a function of training intensity, with
light-intensity training eliciting minimal or no changes (Balady et al. 2000).
2.3.1.2 Muscular fitness
Also muscular endurance and strength – the ability of the muscle to continue to perform without
fatigue and to exert force (United States Department of Health and Human Services 1996) – can
be diminished associated with adjuvant treatment of breast cancer. This decrease is caused by an
abnormal accumulation of muscle metabolites, changes in neuromuscular function (National
Comprehensive Cancer Network 2007), and an increased release of cytokines as a consequence
of tissue necrosis after radiotherapy (Dimeo 2001).
Setting the context
17
The benefits of resistance training have been well documented: resistance training can enhance
strength, muscle mass, and muscle quality (strength relative to muscle mass) (Deschenes and
Kraemer 2002, Hunter et al. 2004).
Common recommendations to cancer patients undergoing adjuvant treatment have focused on
reduced activity and rest (Dimeo 2001, Pinto and Trunzo 2005) in order to achieve “less discom-
fort” (Dimeo 2001) through avoiding breathlessness and tachycardia from physical effort. How-
ever, sedentary habits and bed rest may further aggravate physiological changes of cancer treat-
ment, resulting in severe atrophy of the peripheral muscle mass and deleterious effects on the
cardiovascular system. One bed rest study of 10 days with older adults found a large loss of skele-
tal muscle, particularly from the lower extremities, that was greater than that observed in young
individuals after 28 days (Kortebein et al. 2007). A 2-months bed rest study within the women
international space simulation for exploration (WISE) 2005 program showed that prolonged bed
rest caused microcirculatory endothelial dysfunction which might participate in cardiovascular
deconditioning (Demiot et al. 2007). Thus, a prolonged sedentary lifestyle can precipitate a “dis-
use syndrome” (Pinto and Trunzo 2005) and further deconditioning (Dimeo 2001) with an in-
crease in both sense of disability and risk of injury (Dimeo 2001, Pinto and Trunzo 2005). Pro-
longed sedentarism creates a self-perpetuating condition of diminished activity leading to de-
creased health-related fitness, and vice versa.
2.3.1.3 Body composition
Body composition is a component of physical fitness that relates to the relative amounts of mus-
cle, fat, bone, and other vital parts of the body (United States Department of Health and Human
Services 1996). Weight gain and obesity are common occurrences in women diagnosed with
breast cancer (Demark-Wahnefried et al. 1997, Rock 1999, Demark-Wahnefried et al. 2001,
McInnes and Knobf 2001, Lankester et al. 2002). Significant weight gain occurs in 50% to 96%
of women receiving adjuvant chemotherapy for early-stage breast cancer, as indicated by a com-
pilation of study results, and is greater and more prevalent in younger patients (Demark-
Wahnefried et al. 2001). Gains in weight usually range from 2 kg to 6 kg during the first year of
diagnosis but greater gains can be observed as well (McInnes and Knobf 2001). There is evidence
that the majority of the weight gained during breast cancer treatment is composed of body fat;
studies that measured changes in body composition, rather than postdiagnosis weight gain only,
noted increased body fat in women undergoing adjuvant chemotherapy (Demark-Wahnefried et
al. 2001). These alterations can not only lead to declines in physical function but can also predis-
pose women to weight-related chronic illnesses such as diabetes, heart disease, or orthopaedic
problems. Moreover, weight gain can have negative impact on psychosocial aspects of quality of
Setting the context
18
life (McInnes and Knobf 2001). Clinical and epidemiological studies have identified obesity as an
important negative prognostic factor for survival after the diagnosis of breast cancer: specifically,
increased body weight and waist-to-hip ratio have been associated with greater recurrence and
mortality in women with breast cancer (Kumar et al. 2000, Daling et al. 2001, Chlebowski et al.
2002, Barnett et al. 2008).
Causes underlying postdiagnosis weight gain in breast cancer patients are not fully understood,
but multiple reasons have been suggested, including being or becoming postmenopausal after
diagnosis, decreased physical activity, and increased total caloric intake (McInnes and Knobf
2001). Furthermore, the issue of possible direct effects of adjuvant chemotherapy itself on me-
tabolism has been addressed in an exploratory study (Demark-Wahnefried et al. 1997), in which a
transient decrease in resting metabolic rate was observed. Both a potentially decreased resting
metabolic rate and reduced physical activity levels postdiagnosis result in a decline in total energy
expenditure associated with adjuvant cancer treatment (Irwin et al. 2003). This decline in total
energy expenditure justifies an approach to weight control emphasising increased physical activ-
ity.
Increased exercise to promote weight loss and maintenance and to restore lean body mass is a
strategy whose benefits are fairly well supported by evidence. Exercise can be recognised as an
important weight management intervention since exercise can induce a negative energy balance –
i.e., where energy expenditure exceeds energy intake – by directly enhancing energy expenditure
(the energy cost of the exercise episode) and possibly by affecting resting metabolic rate as well.
Because resting metabolic rate and energy expenditure of physical activity account for around
90% of total energy expenditure (Lemmer et al. 2001), any intervention that increases these two
components of energy expenditure could be useful in restoring energy balance and preventing
weight gain in women receiving adjuvant treatment of breast cancer. Breast cancer patients may
benefit from resistance training to restore muscle mass since resistance training in particular has
been reported to increase lean body mass, which is associated with an increase in the resting
metabolic rate (Lemmer et al. 2001). Putting this together, this suggests that increases in resting
metabolic rate and energy expenditure of physical activity could explain the decreases in fat mass
that have been reported in response to resistance training (Lemmer et al. 2001). However, current
evidence suggests that changes in the resting metabolic rate in response to resistance training may
be influenced by sex with resting metabolic rate not being affected in women (Lemmer et al.
2001). Aerobic endurance training is associated with significant metabolic adaptations in skeletal
muscles (United States Department of Health and Human Services 1996): besides adaptations
that greatly enhance the oxidative capacity of the endurance-trained muscle, the ability of trained
Setting the context
19
muscles to use fat as an energy source is also improved. This improved capacity to oxidise fat
may lead to reductions in fat mass.
2.3.2 Treatment-related symptoms
Women with breast cancer experience a variety of symptoms as a result of their disease or as a
result of treatments for their disease. These symptoms are a major problem for the women, as
well for their family caregivers, because the management of these symptoms is often the respon-
sibility of the patients themselves (Dodd and Miaskowski 2000). In addition, unrelieved symp-
toms can have deleterious effects on patient outcomes such as functional status, mood states, and
quality of life (Dodd et al. 2001). Symptoms are “subjective perceptions of alterations in normal
bodily function and sensation or cognition” (Parker et al. 2005). The cancer symptom experience
is complex: individuals undergoing adjuvant treatment of cancer are likely to experience multiple
concurrent and related symptoms, i.e., symptoms occur not in isolation but in pairs, groups, or
clusters (Dodd et al. 2004). Also interactions between symptoms can be observed: they happen
when two or more symptoms coexist, precipitate, or synergize each other, or where they trigger
the development of other symptoms (Parker et al. 2005). For example, fatigue and depression
have been often documented as examples of concurrent, related symptoms in patients with can-
cer (Barsevick et al. 2006); likewise with sleep disturbances and pain (Parker et al. 2005). Up to 40
different symptoms may be experienced by breast cancer patients receiving adjuvant chemother-
apy (Groenvold et al. 2006). These symptoms include acute consequences of chemotherapy (e.g.,
nausea, vomiting, loss of appetite, diarrhoea, hair loss, sore mouth), symptoms related to ovarian
ablation resulting from chemotherapy (e.g., hot flushes, amenorrhea, vaginal dryness), and symp-
toms like fatigue, loss of energy or sleep disturbances (Groenvold et al. 2006). Also more specific
concerns such as an altered sense of femininity, feelings of decreased attractiveness, and lymphe-
dema (Brady et al. 1997) are important. The experience of a range of symptoms may result in a
perceived decrease in quality of life.
2.3.2.1 Cancer-related fatigue
Patients with cancer commonly report a lack of energy during the course of their disease and
treatment: fatigue. Fatigue is the symptom that has, by far, the largest impact on limiting function
and overall quality of life in breast cancer patients (Arndt et al. 2006). Cancer-related fatigue is
defined as a “distressing, persistent, subjective sense of tiredness or exhaustion related to cancer
or cancer treatment that is not proportional to recent activity and interferes with usual function-
ing” (National Comprehensive Cancer Network 2007). Fatigue is a subjective and multidimen-
sional concept with distinct symptoms concerning physical functioning (e.g., diminished energy
Setting the context
20
or need to rest), cognitive functioning (diminished concentration or attention), and affective
functioning (decreased motivation or interest) (Servaes et al. 2002).
As indicated by the US National Comprehensive Cancer Network (NCCN), fatigue is a nearly
universal symptom in patients receiving cytotoxic chemotherapy, radiotherapy, or treatment with
biological response modifiers. It is a problem that affects 70% to 100% of cancer patients and
has been exacerbated by the increased use of fatigue-inducing multimodal treatments and of
dose-dense, dose-intense protocols (National Comprehensive Cancer Network 2007). Months or
even years after treatment has ended, people living with cancer report that fatigue is a disruptive
symptom (National Comprehensive Cancer Network 2007).
Despite high prevalence, the exact mechanisms involved in the pathophysiology of fatigue are
unknown (Winningham 2001, de Jong et al. 2002, Servaes et al. 2002, Tavio et al. 2002, Mock and
Olsen 2003, Stasi et al. 2003). Although the relative importance of biological, psychological, and
situational factors is usually unclear, these and other factors appear to be important in the patho-
genesis. Proposed biological mechanisms include abnormal accumulation of muscle metabolites,
production of cytokines, changes in neuromuscular function, and abnormalities in adenosine
triphosphate synthesis and anaemia (National Comprehensive Cancer Network 2007) as well as
loss of muscle mass (Baracos 2001, Stasi et al. 2003). Psychological factors discussed as contribut-
ing to symptoms of fatigue comprise anxiety and depression (Stasi et al. 2003). Finally, situational
factors contributing to fatigue include sleep disorders (Stasi et al. 2003) and physical inactivity
(Winningham 2001).
Differing definitions and theoretical frameworks within which fatigue is being studied result in
differing assessment tools. The Piper Fatigue Scale was among the first multidimensional instru-
ments developed specifically for cancer-related fatigue. It measures the dimensions of subjective
fatigue on four scales: behavioural/severity scale (relating to the severity, distress, and degree of
disruption in activity of daily living); affective meaning scale (relating to the emotional meaning
attributed to fatigue); sensory scale (relating to the physical symptoms of fatigue); and cogni-
tive/mood scale (relating to mental and mood states) (Piper et al. 1989, Piper et al. 1998).
Typically, non-pharmacological interventions and pharmacological therapy are advocated in order
to manage cancer-related fatigue: non-pharmacological interventions include education, exercise,
rest and sleep, energy conservation and stress reduction (Ahlberg et al. 2003, National Compre-
hensive Cancer Network 2007), whereas pharmacological therapy comprises psychostimulants,
sleep medication, or anaemia treatment as indicated (National Comprehensive Cancer Network
2007).
Setting the context
21
One promising intervention for combating feelings of low energy and fatigue is regular exercise.
However, since mechanisms involved in the pathophysiology of fatigue are unknown, mecha-
nisms potentially contributing to a decrease in feelings of fatigue are also unclear. There may be
biological, psychological, and social aspects of a regular exercise experience contributing to im-
proved fatigue status. Neurobiological adaptations that accompany regular exercise (Dishman et
al. 2006) may be of importance as well as social interactions that are involved in exercise interven-
tions frequently.
2.3.2.2 Nausea and vomiting
Nausea is a significant problem for most women receiving chemotherapy for breast cancer (Dib-
ble et al. 2003). Nausea is a complex process and can be regarded as a protective reflex against
the ingestion of toxins. Cytotoxic antineoplastic agents vary greatly in their potential to induce
emesis and in the severity of this symptom (Miller and Kearney 2004).
Besides pharmacological antiemetic therapy, several non-pharmacological interventions have
been explored: progressive muscle relaxation, guided imagery, self-hypnosis, biofeedback and
cognitive distraction (Markman 2002). Given the complexity of the emetic process, and with only
preliminary available evidence showing that benefits can be achieved through exercise, no
mechanisms have been proposed yet in the literature for how exercise may affect nausea. How-
ever, exercise may induce distraction and thus contribute to relief of nausea symptoms.
2.3.3 Biological and physiological problems
There are substantial short- and long-term biological and physiological problems deriving from
adjuvant treatment of breast cancer such as a depression of the immune system, premature
menopause with subsequently decreasing bone mineral status, or cardiovascular effects. Short-
term problems typically include those effects encountered during the course of treatment that
resolve within month of the completion of therapy. By contrast, long-term problems include
adverse effects of treatment arising after the conclusion of adjuvant treatment and may have a
sustained impact (Partridge et al. 2001).
2.3.3.1 Depression of the immune system
Virtually all chemotherapeutic regimens and radiotreatment can cause myelosupression (Partridge
et al. 2001, Shapiro and Recht 2001), a depression of the immune system, often by paralysing the
bone marrow and leading to a decrease of white blood cells (neutropenia), red blood cells (anae-
mia) and platelets (thrombocytopenia).
It has been suggested that enhanced immune system function and reduced susceptibility to can-
cer may occur with regular moderate exercise, whereas exhaustive exercise and overtraining may
Setting the context
22
lead to suppressed immune system function (Woods et al. 1999). Based on this working theory,
exercise-induced alterations in cancer-related immune system components have received in-
creased research attention: there is evidence that “physical exercise training may improve a num-
ber of immune parameters that may be important in cancer defence” (Fairey et al. 2002).
2.3.3.2 Premature menopause and postmenopausal bone loss
Premenopausal women treated with adjuvant chemotherapy frequently develop permanent ovar-
ian failure, or early menopause (Ramaswamy and Shapiro 2003), especially after administration of
alkylating agents (e.g., cyclophosphamide, which is a common component in several chemother-
apy regimens for breast cancer). The major determinants of ovarian failure are the age at treat-
ment and the total cumulative dose of cyclophosphamide. The overall incidence of ovarian failure
related to cyclophosphamide regimens is about 70% (Ramaswamy and Shapiro 2003), with a
timeframe between 2 and 16 months to onset of ovarian failure. In chemotherapy-induced ovar-
ian failure, oestrogen levels decline rapidly resulting in oestrogen deficiency-associated bone loss.
Aromatase inhibitors (administered for early-stage, oestrogen receptor-positive breast cancer) are
likely to increase the risk of osteoporosis in some breast cancer survivors as well (Ramaswamy
and Shapiro 2003). Altogether, women with breast cancer are at higher risk for osteoporosis and
osteoporotic fractures than other women (Adler 2007, Brown and Guise 2009).
Regular exercise has shown effects on bone density, size, shape, and geometry resulting in sub-
stantial improvements in mechanical strength (Wolff et al. 1999, Kelley et al. 2001, Turner and
Robling 2005, Hamilton et al. 2009). It has been suggested that gravitational forces and muscle
pull related to exercise may affect bone density through the production of strains within the
skeleton. These strains are perceived by bone cells as osteogenic resulting in bone formation
(Wolff et al. 1999).
2.3.3.3 Cardiovascular injuries
Each chemotherapeutic agent used in breast cancer management is associated with unique acute
and long-term cardiac complications (Jones et al. 2007). The majority of complications are tran-
sient effects that do not persist after completing chemotherapy which shows the need for action
especially in the period during cancer treatment. Regarding radiotherapy, it can be said that while
modern radiation techniques provide lower cardiac mortality risks than older techniques, cardio-
pulmonary damage does nonetheless occur (Jones et al. 2007). The direct and indirect effects of
adjuvant therapy coupled with an unhealthy lifestyle and the presence of modifiable risk factors
all contribute to an elevated risk of future cardiovascular disease especially in middle-aged and
elderly women with early breast cancer who are already at risk for cardiovascular disease (Jones et
al. 2007). There is the chance that lifestyle modifications, especially exercise interventions, may
Setting the context
23
mitigate these adverse cardiac effects. At the same time, there is the risk that cardiovascular inju-
ries from adjuvant treatment predispose women to harmful effects of exercise training. These
risks have to be evaluated adequately (surveillance for harm).
2.3.4 Mental health
According to the WHO, mental health can be conceptualised as a “state of wellbeing in which
the individual realizes his or her own abilities, can cope with the normal stresses of life, can work
productively and fruitfully, and is able to make a contribution to his or her community” (World
Health Organization 2007). In this positive sense, mental health is the foundation for wellbeing
and effective functioning for an individual and for a community.
Identification of emotional distress is a key step in the assessment of mental health problems.
Being diagnosed with cancer marks the beginning of a series of difficult but essential life adjust-
ments in response to this challenging life stressor. Frequently, cancer patients develop subsequent
mood disturbances such as depression and anxiety. Depressive symptoms can be associated with
certain psychiatric disorders (e.g., adjustment disorder, major depressive disorder), or they can be
present in the absence of a psychiatric diagnosis. Diagnosis of major depression may be compli-
cated by overlapping effects and interrelationships of psychological distress (anxiety, anger,
shock, bereavement), major depressive episode (depressed mood, loss of interest/pleasure, low
self-esteem or guilt, suicidal ideation) and symptoms induced by cancer and/or cancer treatment
(Somerset et al. 2004).
Besides depression, anxiety is a further concern in the category of emotional distress. Frequently,
depression and anxiety coexist in cancer patients. Reactive anxiety, which is related to fear, is by
far the most common type of anxiety seen in patients and can be regarded as a normal, however
debilitating, reaction to facing the unknown. It includes fear of death, treatment, disability, pain,
suffering, and loss or disruption of relationships (Sivesind and Baile 2001).
Several mechanisms have been proposed for how exercise may affect depression: diversion of
negative thoughts, mastery of new skills, social contact, positive feedback, and increased sense of
self-worth together with physiological effects such as changes in endorphin levels (Lawlor and
Hopker 2001). A meta-analysis examining the effectiveness of exercise on general mood (includ-
ing phenomena such as depression, anxiety, anger, vigour) in the elderly found exercise to be
associated with significant improvements in mood in this age group (Arent et al. 2000).
2.3.5 Health-related quality of life
Health-related quality of life plays an important role in helping establish the optimal treatment
and care approach for cancer patients: improving the quality of life, not merely the length of life,
Setting the context
24
is an important parameter of benefit. However, there is only limited information available on the
nature and time course of the impact that breast cancer treatment has on health-related quality of
life. Quality of life is an important outcome describing the subjective experience of the breast
cancer disease and treatment, but there is no consensus on what it really is. Contrary to the vast
body of literature relating to happiness, wellbeing and satisfaction, quality of life has been associ-
ated with an “astonishing lack of theory” (Bullinger 2002). The World Health Organisation
(WHO) has defined health as “a state of complete physical, mental and social wellbeing and not
merely the absence of disease or infirmity” (WHO 1948). The WHO’s definition of health is
widely used and strongly supports the multidimensional aspects of health. This definition has
influenced the notion of quality of life as a multidimensional concept. There are several sugges-
tions about how this multidimensional character of the quality of life could be incorporated into
its definition. The WHO Quality of Life Assessment group (WHOQOL) defines quality of life as
the “individuals’ perceptions of their position in life in the context of the culture and value sys-
tems in which they live and in relation to their goals, expectations, standards and concerns”
(WHOQOL group 1996). Quality of life refers to a subjective evaluation with a focus upon re-
spondents’ perceived quality of life, and subsequently quality of life measures assess the value of an
intervention based on wellbeing and function rather than based on symptoms, diseases, or condi-
tions. According to Bullinger, health-related quality of life can be viewed as a psychological con-
struct which describes the psychological, social, physical and functional aspects of wellbeing and
function from the individual’s perspective (Bullinger 2000). Some authors extend the dimensions
by further aspects such as spirituality (Bullinger 2002) or financial concerns, job satisfaction, and
living conditions (WHOQOL group 1996). The various miscellaneous definitions stress the mul-
tidimensionality of the quality of life concept as well as the relevance of respondents’ self-report.
There are several valid, reliable questionnaires available to measure a person’s health-related
quality of life with two basic approaches: first, generic instruments that provide a summary health
profile and secondly, disease-specific instruments which focus on specific problems associated
with a disease or area of functioning. Currently, one of the most commonly used generic health
status instruments is the Medical Outcomes Study 36-Item Short Form (SF-36), a 36-item meas-
ure encompassing eight domains: physical functioning, social functioning, mental health, role
limitations due to physical problems, role limitations due to emotional problems, vitality, bodily
pain, and general health perceptions. The domains and items selected for disease-specific instru-
ments are directly associated with the impairments caused by a specific disease process or its
therapeutic interventions. Disease-specific instruments are designed to assess health-related qual-
ity of life in patient populations, and so disease-specific instruments are likely to be more reactive
than generic instruments to intervention-related changes. A number of such instruments have
Setting the context
25
been designed for use with cancer populations. Because cancer at different sites may have widely
differing effects on health-related quality of life, cancer-specific instruments still represent a
rather broad approach. Site-specific instruments are those which attempt to measure the impact
on the health-related quality of life for cancer at a particular anatomical site. One example of
these instruments is the Functional Assessment of Cancer Therapy (FACT) instrument, which is
comprised of a core instrument and subscales. Items are contained within four validated sub-
scales (physical wellbeing, social/family wellbeing, emotional wellbeing, and functional wellbe-
ing).
Despite the extensive use of chemotherapy, the literature investigating the impact of adjuvant
chemotherapy on health-related quality of life of breast cancer patients in prospective, longitudi-
nal studies is relatively limited. Many studies either employ a cross-sectional or retrospective
design, or lack control groups. Furthermore, a wide range of chemotherapeutic regimens, which
may have different toxicity profiles, has been studied. Moreover, most studies have assessed only
relatively few variables of health-related quality of life.
Randomised clinical studies were primarily conducted by the large oncological clinical trial groups
in North America and Europe. In testing of chemotherapy regimens in breast cancer patients,
quality of life has been considered as an outcome parameter. Health-related quality of life during
adjuvant chemotherapy for breast cancer shows transient adverse effects during therapy, but it
improves after cessation of therapy – either standard therapy (Hürny et al. 1996, Fairclough et al.
1999, de Haes et al. 2003, Land et al. 2004, Martin et al. 2005, Bernhard et al. 2007, Bernhard et
al. 2008), dose-intensive therapy (Del Mastro et al. 2002), or high-dose therapy (Macquart-Moulin
et al. 2000, Brandberg et al. 2003, Peppercorn et al. 2005). Overall quality of life improves sub-
stantially over the first two years (Hürny et al. 1996). For the majority of breast cancer patients,
most aspects of health-related quality of life recover without long-term effects after adjuvant
chemotherapy has ended – except for vasomotor symptoms and sexual dysfunction (Grimison
and Stockler 2007).
Reported changes in quality of life over time have to be treated cautiously since they need not
necessarily derive from actual changes in health or symptoms: people with severe chronic ill-
nesses report quality of life equal or superior compared to reports from less severely ill or healthy
people, and consistent disparities arise between clinical measures of health and patients’ own
evaluations (Rapkin and Schwartz 2004). In fact, this paradox is now understood to reflect a
psychological adaptation (a “response shift”) that occurs in cancer patients as well as in patients
with other chronic diseases (Rapkin and Schwartz 2004). The internal standard by which patients
appraise their current health state shifts and the same questionnaire items on wellbeing can elicit
Setting the context
26
fundamentally different answers over time. To the extent that subjective wellbeing reflects psy-
chological adaptation, the connection between subjective quality of life and disease course (or
treatment response) weakens (Schwartz and Rapkin 2004).
In light of the observation that quality of life is predominantly disrupted during treatment, inter-
ventions to sustain quality of life are most needed in this period. Physical exercise is regarded as
an intervention with significant potential to influence health-related quality of life in various
populations (United States Department of Health and Human Services 1996). The most direct
effects are likely to be in the areas of psychological wellbeing (e.g., self-concept, self-esteem,
mood, and affect), perceived physical function (e.g., perceived ability to perform activities of daily
living), and physical wellbeing (e.g., perceived symptoms and fatigue).
2.3.6 Recurrence and survival
Epidemiological evidence from a recent prospective cohort study of almost 3,000 women sug-
gests that higher levels of physical activity are associated with reduced risks for breast cancer
recurrence, breast-cancer-specific mortality, and all-cause of mortality (Holmes et al. 2005).
2.3.7 Prevalence of treatment-related adverse effects
Irrespective of the respective prevalence, there is evidence that a number of health-care profes-
sionals do not indicate that patients have certain treatment-related adverse effects – even though
those effects are reported by patients. Subsequently, treatment-related adverse effects remain
underdiagnosed and managed inadequately (Groenvold et al. 2006).
Prevalence estimates for treatment-related adverse effects are not presented here due to several
methodological issues regarding prevalence estimates which have been observed in studies. First,
most studies assessing prevalence estimates for treatment-related adverse effects are cross-
sectional, and so these studies neglect the time course of the impact of adjuvant cancer treatment.
Secondly, no standard has been adopted and a variety of different (self-report) instruments were
used across studies. This makes comparison across studies difficult, if not impossible.
Finally, besides the type of adjuvant cancer treatment there are various factors that contribute to
the rate of occurrence of treatment-related adverse effects such as age, culture, or co-morbidities,
and so studies in different settings will reveal different prevalence estimates for treatment-related
adverse effects. Estimates of prevalence are difficult to compare across studies if different time
frames are employed, different assessment tools are used, or different inclusion or exclusion
criteria are being relied upon.
Setting the context
27
2.4 The intervention: exercise
Exercise, physical activity, and fitness describe different concepts, where physical activity is the
broadest of the three concepts and can be characterised as “any bodily movement, produced by
skeletal muscles that results in energy expenditure” (Caspersen et al. 1985). Physical activity can
occur in any setting and can be categorised into occupational, household, or recreational physical
activity and sports. Exercise is a subset of physical activity and is “planned, structured, repetitive
and purposive in the sense that improvement or maintenance of one or more components of
physical fitness is an objective” (Caspersen et al. 1985). Physical fitness refers to a “set of attrib-
utes that are either health- or skill-related” (Caspersen et al. 1985) and that can be achieved
through exercise. Health-related fitness encompasses cardiorespiratory endurance, muscular
endurance, muscular strength, body composition and flexibility (Caspersen et al. 1985).
2.4.1 Exercise patterns in breast cancer patients
Overall, the stress associated with the diagnosis of breast cancer and cancer treatment is known
to disrupt health behaviours such as exercise behaviour (Pinto and Trunzo 2005). Many women
with breast cancer are reluctant to exercise and may decrease their exercise levels during the pe-
riod of adjuvant therapy. Several studies have addressed the important questions of whether
declines in physical activity are evident during adjuvant therapy and whether observed declines
during adjuvant therapy persist beyond treatment completion. These studies suggest that physical
activity after treatment completion may remain depressed relative to pre-diagnosis levels.
The evidence regarding exercise patterns in breast cancer patients is predominantly based on
cross-sectional and retrospective research. Data from the U. S. National Health Interview Survey
with a total of nearly 130,000 adults (Bellizzi et al. 2005) showed that the proportion of cancer
survivors who met the recommendations for physical activity level was 29.6%, compared with
36.6% for those without a history of cancer. Courneya and Friedenreich (1997a, 1997b) required
retrospective reports from cancer survivors (breast and colorectal cancer) regarding how their
physical activity changed after cancer diagnosis and treatment and found decreases in exercise
during treatment followed by an increase in exercise after treatment completion. Exercise activity,
however, did not return to pre-diagnosis levels. Blanchard et al. (2003) found that 30% of cancer
survivors, of whom nearly half were breast cancer survivors reported decreased exercise activity,
whereas 15% reported increased activity since diagnosis. Finally, Irwin et al. (2003) compared
physical activity levels before and after diagnosis of breast cancer using retrospective reports of
pre-diagnosis activity. This showed that physical activity declined by about two hours per week
with the greatest decreases in sports-related physical activity in women treated with adjuvant
radiotherapy plus chemotherapy. This evidence from cross-sectional or retrospective research
Setting the context
28
suggests that cancer treatment can be associated with reductions in physical activity. In contrast,
one study with a prospective, longitudinal design – while it did indeed find decreased activity
levels, relative to the pre-diagnosis baseline– showed that the decreased activity levels seemed to
recover quickly and (on average) did not differ from the pre-diagnosis baseline a few months
after completion of adjuvant therapy (Andrykowski et al. 2007). But over a third of the sample
evidenced a clinically important change in leisure-time exercise over the course of the study:
whereas some women reported decreases in leisure-time exercise after diagnosis, other women
reported increases in exercise six months after completion of treatment, relative to pre-diagnosis
levels. Although the decline in exercise participation during adjuvant therapy was followed by
recovery to pre-diagnosis levels in this study, participation in leisure-time exercise activity six
months after conclusion of adjuvant therapy was still suboptimal, with only a moderate percent-
age of participants meeting current guidelines for exercise activity.
In sum, activity levels during adjuvant therapy for breast cancer appear to decline. Even if they
recover to pre-diagnosis levels, activity levels among breast cancer survivors seem to be sub-
optimal and health promotion programs with exercise interventions are required as the following
data from health surveillance in Germany indicate: the rate of women who participate in regular
physical exercise (weekly exercise activities of two hours or more) declines with age: from 40% of
women aged between 20 and 29 years to 22% of women aged between 70 and 79 years (Ruetten
et al. 2005). Consequently, even if a decline in exercise participation during adjuvant treatment
for breast cancer is followed by recovery to pre-diagnosis levels after the conclusion of treatment,
exercise participation may still be suboptimal.
Possibly, the period of adjuvant therapy is the ideal time point for the delivery of health promo-
tion interventions in the cancer diagnosis-rehabilitation continuum. Intervening at the ideal time
point would ensure that the women diagnosed with breast cancer are most amenable to both
make changes in their behaviour and maintain those changes. It could be assumed that women
may be “catapulted in their readiness to alter life-style risk factors by a diagnosis of cancer and its
treatment” (Pinto et al. 2000). Receptivity towards health promotion interventions may be high
during the period of initial treatment, a period that turned out to be a “teachable moment” (De-
mark-Wahnefried et al. 2000, Demark Wahnefried et al. 2005). On the other hand, the distress of
breast cancer diagnosis and cancer treatment could overwhelm women regarding coping re-
quirements.
2.4.2 Behaviour change theories for program planning
Health promotion interventions based on exercise have the potential to reduce dysfunction and
improve long-term health. On the level of personal skills, “health promotion supports personal
Setting the context
29
and social development through providing information, education for health, and enhancing life
skills. By so doing, it increases the options available to people to exercise more control over their
own health and over their environments, and to make choices conducive to health” (WHO
1986). Effective health promotion programs can help women to manage their breast cancer ex-
perience and reduce further disease risks. Usually, behaviour changes are required in order to
achieve success, e.g., previously sedentary women have to adopt a regular exercise program.
Health behaviour and health promotion theories can play a crucial role in program planning.
They provide tools for developing appropriate interventions and evaluating their success based
on findings from various disciplines, such as psychology, sociology, or marketing. These tools are
based on an understanding of behaviour and thus allow a program planning to be grounded in
more than intuition – but in a grasp of the dynamics of health behaviour and the social/ envi-
ronmental influences that affect health behaviour. Health behaviour theories suggest how pro-
gram strategies can be devised that reach the target group and have an impact. Moreover, they
help to identify appropriate indicators for monitoring and evaluation. Health behaviour theories
can consequently make a major contribution to improving the design of programs and maximis-
ing potential effects. There are numerous theories to draw on and no single theory dominates
health education and promotion. Some theories focus on individuals as the focus of change (e.g.,
Health Belief Model). Others examine change within families, institutions, or communities. Some
constructs, such as self-efficacy, i.e., the confidence in one’s ability to take action and overcome
barriers, are central to several theories.
For understanding the determinants of exercise participation, various distinct behaviour change
theories have been examined in exercise studies with women undergoing active treatment of
breast cancer: the Theory of Planned Behaviour, that assumes that behavioural intentions are the
most important determinant of behaviour (Courneya and Friedenreich 1999a, Courneya et al.
2001, Hunt-Shanks et al. 2006); the Transtheoretical model (Stages of Change), whose basic
premise is that behaviour change is a process, not an event (Rogers et al. 2007); and finally the
Social Cognitive Theory, which identifies self-efficacy, goals, and outcome expectancies as the
main factors governing health behaviour change (Rogers et al. 2004, Rogers et al. 2005a, Rogers
et al. 2005b).
Using the Transtheoretical Model and the Social Cognitive Theory as examples, the requirements
regarding exercise interventions within the framework of a behaviour change theory are high-
lighted in the following. According to the Transtheoretical Model (Prochaska and DiClemente
1983), changes are processes involving progress through a series of five stages: regarding exercise
activity, precontemplation is the first stage at which people are not yet considering exercise adoption
in the foreseeable future; contemplation is the next stage at which people are intending to adopt
Setting the context
30
exercise in the next six months, being aware of the pros of exercise adoption but also of the cons;
preparation is the third stage at which people are already intending to start exercise in the immedi-
ate future: at this stage, they have a plan of action, such as joining an exercise class or they have
already started exercising but not regularly. Following preparation, action is the stage in which
people have started a regular exercise program within the past six months. In the Transtheoretical
Model, action is only one of five stages and can not be regarded as behaviour change. This stage
is the stage where vigilance against relapse is critical. Maintenance, finally, is the stage in which
people are regularly exercising for six months or more. Relapse – the regression from action or
maintenance to an earlier stage – tends to be the rule for most health behaviours (Cancer Preven-
tion Research Center). Programs based on the Transtheoretical Model are designed to develop
interventions that are matched to the specific needs of the participants and recognise that differ-
ent individuals will be in different stages of change. Interventions are tailored to the participants’
motivational readiness to start exercising, and so participants who are ready for action may re-
ceive guidance on exercise adoption such as recommendations regarding specific types of exer-
cise or where to exercise. Such action-oriented recommendations would be regarded as ineffec-
tive for those who have no intention of adopting exercise. These individuals might need more
detailed information regarding the benefits of exercise benefits or its safety.
According to the Social Cognitive Theory (Bandura 1997), three main factors affect the likeli-
hood that a person will change health behaviour: self-efficacy, goals, and outcome expectancies.
Exercise interventions based on the Social Cognitive Theory often focus on self-efficacy – the
confidence in one’s ability to take action and overcome barriers. If individuals have a sense of
self-efficacy, they can change behaviours even when faced with obstacles. Regarding exercise
adoption, the Social Cognitive Theory implies that individuals are not motivated to act, or to
persist through challenges if they do not feel that they can have control over their exercise behav-
iour. Interventions attempt to enhance self-efficacy through techniques such as setting realistic,
easily attainable goals, focusing on participants’ progress and reinforcing successes.
2.4.3 Exercise adherence
The biggest challenge in encouraging health-related lifestyle change is not the initial change but
rather the ability of participants to adhere to a change in the long run. The term adherence refers
to the level of participation achieved in a behavioural regimen once the individual has agreed to
undertake it (King 1994). Adherence to exercise is a critical concern because the benefits of exer-
cise may not persist when exercise is discontinued (Lennon et al. 2004, Herrero et al. 2007). Re-
garding exercise adherence, adoption, maintenance, and resumption of exercise are distinguished:
for exercise prescriptions to be successful, women must persist and adhere to the recommended
Setting the context
31
program, i.e., they must initiate and maintain their participation in exercise. Then, adherence to
exercise regimens after interventions have ended is a crucial issue. Success in maintaining exer-
cise depends on repeated adoption or resumption of exercise after periods of inactivity (Dishman
1994). There are two components of adherence: attendance at exercise sessions and the exercise
intensity achieved when compared with the target intensity as prescribed. Despite the importance
of adhering to exercise, exercise adherence appears to be a neglected issue in cancer research.
Given that adherence to exercise is a problem in both asymptomatic and diseased populations,
with an average dropout rate of 50% (Dishman 1994), it can reasonably be assumed that compa-
rable rates of non-adherence will occur among breast cancer patients. However, better adherence
may lead to better outcomes.
Understanding the ways in which individuals sustain healthier long-term lifestyle choices is
needed to help cancer patients maintain recommended levels of physical activity. Knowing the
key barriers women face when receiving adjuvant treatment for breast cancer may serve as a basis
for the development of interventions to assist women in exercising regularly during this difficult
time. Several studies have investigated predictors of adherence to exercise in women but these
have mainly recruited young healthy women. Predictors of exercise adherence in younger healthy
adults may be quite different to those for elderly populations of women with breast cancer. The
identification of reliable predictors of exercise adherence will allow healthcare providers to effec-
tively intervene and support sedentary women during the process of changing their patterns of
physical activity. Many barriers (real or perceived) exist, which represent obstacles to maintaining
regular exercise in this period of cancer treatment. In this respect, the assessment of barriers to
supervised exercise in breast cancer patients participating in a randomised controlled trial (Cour-
neya et al. 2008a) showed the importance of disease- and treatment-related barriers: feeling sick,
fatigue, chemotherapy day, nausea/vomiting and depression, amongst others, accounted for
more than half of the missed sessions among breast cancer patients.
Issues related to gender also appear to be important. A focus-group study (Emslie et al. 2007) of
the experiences of women with breast cancer who took part in an exercise trial, suggested that
the woman’s traditional role as a caregiver and concerns about body image act as potential barri-
ers to physical activity and that a gender-sensitive approach may help overcoming barriers to
physical activity, e.g., exercise classes solely for women with breast cancer provided within a
supportive group environment. It is clear that addressing barriers, particularly disease- and gen-
der-related barriers, is a critical issue that will impact upon exercise adherence during treatment
for breast cancer.
Setting the context
32
Inactivity, i.e., non-adherence to exercise prescriptions, may actually intensify such treatment-
related adverse effects as weight gain, deterioration of lean muscle mass, overall reduction in
physical functioning and fatigue, which in turn may subsequently act as further barriers to exer-
cise.
2.4.4 Exercise prescription
According to the American College of Sports Medicine (ACSM), the art of exercise prescription
is the “successful integration of exercise science with behavioural techniques that result in long-
term program compliance and attainment of the individual’s goals” (Balady et al. 2000). Exercise
prescription for breast cancer patients must be highly individualised owing to the extreme vari-
ability of the effects of cancer and treatment regimens on functional capacity. Furthermore, since
cancer is strongly associated with advancing age, other concurrent or prior health problems
should be anticipated and taken into account when developing any exercise prescription. For the
breast cancer patient, the exercise prescription provides a guideline for safe and effective levels of
exercise training, basing the intensity level on exercise testing, the present physical status, and the
current phase of treatment. Essential components of a systematic individualised exercise pre-
scription include the appropriate mode(s), intensity, duration, frequency, and progression of
exercise.
Regarding exercise intensity, low-fit, sedentary and clinical populations can improve physical
fitness with lower intensity, longer duration exercise sessions (Balady et al. 2000). The ability of
individuals to undertake exercise successfully at a given absolute intensity is directly related to
their relative effort as reflected by heart rate response and perceived exertion. Consequently, the
most common method of setting exercise intensity use heart rate and perceived exertion. Public
health recommendations for the general population as well as for clinical populations have
evolved from emphasizing vigorous activity for cardiorespiratory fitness to including the option
of moderate levels of activity for numerous health benefits. Cardiorespiratory fitness gains seem
to be similar when exercise occurs in several short sessions (e.g., 10 minutes) as when the same
total amount and intensity of exercise occurs in one longer session (e.g., 30 minutes) (Jakicic et al.
1995, United States Department of Health and Human Services 1996). Moreover, for women
who might be unable to exercise 30 minutes continuously due to treatment-related side effects,
shorter episodes are clearly better than none and, quite significantly, short-bouts of exercise may
enhance exercise adherence (Jakicic et al. 1995).
According to the updated recommendation from the American College of Sports Medicine (Has-
kell et al. 2007), to promote and maintain health adults need moderate-intensity aerobic (endur-
ance) physical activity for a minimum of 30 minutes on five days each week or vigorous-intensity
Setting the context
33
aerobic physical activity for a minimum of 20 minutes on three days each week. Moderate-
intensity aerobic activity is generally equivalent to a brisk walk and noticeably accelerates the
heart rate while vigorous-intensity activity causes rapid breathing and a substantial increase in
heart rate. Combinations of moderate- and vigorous-intensity activity, for example by walking
briskly for 30 minutes twice during the week and then jogging for 20 minutes on two other days,
are feasible to meet this recommendation. In addition to endurance physical activity, every adult
should perform activities that maintain or increase muscular strength and endurance (i.e., resis-
tance exercise training) a minimum of two days each week. The intensity of resistance exercise
training can be adjusted by varying the weight, the number of repetitions, the length of the rest
interval between exercises, or the number of sets of exercises completed. Courneya (Courneya et
al. 2000) proposed exercise guidelines for early stage cancer patients, to be modified as needed
for specific patients. These guidelines recommend walking and cycling as safe and generally well
tolerated exercise modes involving large muscle groups, with a recommended frequency of three
to five times per week. More deconditioned patients should begin with daily sessions of shorter
duration and lower intensity. In general, moderate intensity exercise (50-75% heart rate reserve,
rate of perceived exertion 11-14) in sessions of between 20 and 30 minutes duration are recom-
mended, with modifications as needed, including very short exercise bouts (three to five minutes)
followed by rest periods. Perceived exertion is based on the physical sensations a person experi-
ences during physical exercise training, including increased heart rate, increased breathing rate,
sweating, and muscle fatigue. This is a subjective measure, but exertion rating provides a fairly
good means for self-monitoring and adjusting exercise intensity. Perceived exertion ratings be-
tween 11 (“light”) and 14 (“somewhat hard”) on the Borg Scale suggest that exercise is being
performed at a light to moderate level of intensity. The Borg Scale ranges from 6 to 20, where 6
mean “no exertion at all” and 20 mean “maximal exertion”. With regard to progression, patients
should meet frequency and duration goals before they increase exercise intensity, with slower and
more gradual progression for deconditioned patients or those who are experiencing severe side
effects of treatment. For an exercise program to be successful, participants must be exposed to a
therapeutic dose of exercise.
2.4.5 Exercise Testing
Women undergoing adjuvant treatment for breast cancer are generally capable of first completing
exercise testing, and then following an exercise prescription based on their functional ability and
limitations, regardless of whether disease- or treatment-related (Schwartz 1997). The ACSM
recommends pre-participation health screening relative to risk factors (Balady et al. 2000) and
thus, the exercise evaluation for breast cancer patients currently receiving adjuvant therapy
should consist of a medical examination and exercise testing prior to participation. The objective
Setting the context
34
of exercise testing is to prescribe a safe and effective exercise level through evaluation of exercise
tolerance and aerobic power. Exercise tests for persons with cancer are typically submaximal;
submaximal exercise testing is preferred for women receiving adjuvant treatment of breast cancer
whose performance may be limited because of fatigue or physical symptoms. Indeed, maximal
exercise testing is considered the gold standard for assessing maximal aerobic capacity but only a
few individuals reach maximum oxygen consumption, and, furthermore, these tests require par-
ticipants to exercise to the point of volitional fatigue, so additional monitoring equipment is
needed. Predictive and performance tests are the two major categories of submaximal tests. In
predictive tests, exercise response to a given workload is used to predict maximal aerobic capac-
ity. Motorised treadmills, stationary cycle ergometers, or weight machines are the most commonly
used devices for the assessment of physical fitness. Exercise tests follow standard graded proto-
cols with increments in work rate (e.g., grade, speed of treadmill). Typically, heart rate or oxygen
consumption at two or more workloads is measured; results are used to predict maximum oxygen
consumption.
Besides predictive exercise testing, performance tests (field tests) are employed that involve
measuring the responses to standardised physical activities that are typically encountered in eve-
ryday life (Noonan and Dean 2000). Although less precise, submaximal testing provides a rea-
sonably accurate reflection of an individual’s fitness: if the heart rate response decreases over a
period of exercise training, it is likely that cardiorespiratory fitness has improved, independent of
the accuracy with which maximum oxygen consumption is predicted.
2.5 Potential harm associated with exercise
Understanding the potential for harm is important for recommending exercise to women receiv-
ing adjuvant treatment of breast cancer. Harms are defined to be all possible adverse conse-
quences of an intervention (Ioannidis et al. 2004); they are the direct opposite of benefits, against
which they must be weighed. Since the benefit of an intervention may be outweighed by its harm,
a balanced decision requires reliable evidence not only on the benefit, but also on the harm asso-
ciated with an intervention. Clinical concerns about prescribing exercise to cancer patients are
manifold: first, cancer patients may be at an increased risk of musculoskeletal injury from exer-
cise, due to prolonged, reduced levels of overall physical activity, or changes in weight; there is an
increased likelihood of pathological bone fractures arising from compromised bone integrity. The
risk of musculoskeletal injuries, especially in older exercisers, is substantial. Acute stress from
sudden forceful movement, or from repetitive movement, can cause strains, tears, and fractures,
with there being a greater risk of injury associated with increased exercise frequency (McTiernan
2004). However, most musculoskeletal injuries related to physical exercise are believed to be
Setting the context
35
preventable by gradually working up to a desired level of activity and by avoiding excessive
amounts of activity (United States Department of Health and Human Services 1996). Second,
vigorous exercise has the potential to act as an immunosuppressive, depending upon exercise-
type, intensity, duration, and the immune parameters examined. Third, cancer patients may have
increased cardiac risks, especially if they receive high or extended doses of cardiotoxic medica-
tions (e.g., Adriamycin), or have had their chest exposed to radiation. Moreover, a worsening of
side effects such as severe pain and nausea may be triggered by exercising, and fatigue may be
exacerbated by physical exercise.
Not least, physical activity is particularly feared as a possible risk factor for arm edema, which is a
common complication of breast cancer therapy and can result in substantial functional impair-
ment and psychological morbidity. The risk of arm edema increases when axillary dissection and
axillary radiotherapy are used. Lymphedema develops when the lymphatic load exceeds lymphatic
transport capacity (Földi et al. 1989). Breast cancer-related arm lymphedema is caused when the
axillary lymphatic system is interrupted by axillary lymph node dissection or radiotherapy-induced
fibrosis (Johansson et al. 2005), which results in the accumulation of fluid in subcutaneous tissue
in the arm. The most commonly applied treatments for arm lymphedema are compression ther-
apy, manual lymph drainage and intermittent pneumatic compressions (Johansson et al. 2005).
Although risk factors for lymphedema are poorly characterised, there are fears that physical activ-
ity is a possible risk factor (Johansson et al. 2005, Ahmed et al. 2006). In prospective research,
however, no form of physical activity has been associated with lymphedema (Ahmed et al. 2006),
and guidelines that warn breast cancer patients against vigorous, repetitive or excessive upper
body exercise are considered problematic (Ahmed et al. 2006). In contrast to these guidelines,
some researchers argue that exercise would act as a treatment modality rather than a causative
factor for lymphedema (McKenzie and Kalda 2003, Cheema and Gaul 2006), and subsequently
suggest the use of a gradual, progressive upper-body exercise program in the rehabilitation and
prevention of lymphedema. McKenzie and Kalda argue that skeletal muscle pump action, which
increases lymph flow actions, is stimulated by exercise (McKenzie and Kalda 2003). This argu-
ment is further supported by investigations of lymph flow dynamics in healthy people that dem-
onstrate both increased lymph flow rates during exercise and higher lymph clearance rates at rest
in trained subjects (Havas et al. 1997). Cheema especially advocates upper-body resistance train-
ing, because strong muscles would be more advantageous for “mobilizing stagnant interstitial
fluid” (Cheema and Gaul 2006).
In addition, it is important to consider that women receiving adjuvant chemotherapy or radio-
therapy may not respond normally to exercise training (Winningham et al. 1986). Due to the
increased risks cancer patients face, the following contraindications to exercise for cancer patients
Setting the context
36
have been given (Winningham et al. 1986): unusual fatigability or muscle weakness, development
of irregular pulse, leg pain or cramps, chest pain, acute onset of nausea during exercise, vomiting
or severe diarrhoea within previous 24-36 hours, disorientation/confusion, dizziness/blurred
vision/faintness, pallor/cyanosis, sudden-onset dyspnoea, decreased heart rate and/ or blood
pressure with increased workload and intravenous chemotherapy within previous 24 hours. Fi-
nally cancer patients may be unable or unwilling to tolerate exercise in their weakened physical
and emotional condition (Courneya et al. 2000). Considering the potentially increased risks, the
importance of assessing the harm of exercise interventions in cancer patients is made stronger.
In general, exercise facilitators have the responsibility for not exposing a participant to an unrea-
sonable risk of harm through monitoring, instruction, supervision, or advice. Negligent training,
e.g., leaving a participant without instructions about running on the treadmill may constitute an
unreasonable risk of harm. Also, allowing or encouraging a participant to bench press more
weight than is appropriate without adequate supervision may be seen to be subjecting the partici-
pant to an unreasonable risk to her health.
2.6 Previous research
Assessment of previous research indicated that the research question of this review had not been
examined yet through a comprehensive systematic review including quantitative synthesis of
results.
The protocol of this review was prepared and published in 2004 and at that time there were only
narrative, non-quantitative reviews of exercise interventions among cancer patients during and
after cancer treatment, which had considered a number of outcomes. Some highlighted specific
outcomes and populations, such as weight loss in breast cancer patients (Chlebowski et al. 2002),
immune function (Fairey et al. 2002), fatigue (Stricker et al. 2004, Watson and Mock 2004) and
quality of life (Courneya and Friedenreich 1999b); others reviewed effects on a variety of physio-
logical and psychosocial outcomes (Friedenreich and Courneya 1996, Courneya 2001, Courneya
2003, Irwin and Ainsworth 2004, Oldervoll et al. 2004, Galvao and Newton 2005, Knols et al.
2005). Narrative reviews of exercise intervention studies have been qualitative; they have noted
methodological weaknesses but have been unable to quantify outcomes. For the most part, these
reviews included all types of cancer and a range of study designs (i.e., controlled and uncontrolled
trials).
Since 2004, a few systematic reviews have emerged that included quantitative analyses (Stevinson
et al. 2004, Schmitz et al. 2005, Conn et al. 2006, McNeely et al. 2006). However, all these reviews
examined the effectiveness of exercise interventions applying a different scope to the research
question, as the following shows:
Setting the context
37
Schmitz et al. (2005) examined the effectiveness of exercise interventions, either alone or com-
bined with diet modification, during and after cancer treatment across all types of cancers. Effect
size estimates were calculated separately for interventions during and after cancer treatment;
however, no peculiar data for women undergoing adjuvant treatment of breast cancer were pre-
sented. Significant effect sizes were seen in studies, conducted with cancer patients during treat-
ment, showing improvements in cardiorespiratory fitness, physiological outcomes, symptoms and
side effects, and for immune parameters. For fatigue as well as for vigour/ vitality quantitative
null findings were reported with regard to the period during treatment.
Likewise, Stevinson at al. (2004) examined the effectiveness of exercise interventions during and
after cancer treatment across all types of cancers. Meta-analysis on physical function in controlled
trials of women with breast cancer (during and after cancer treatment) yielded evidence that
exercise improved physical function; but no overall effect of exercise on fatigue could be ob-
served.
Conn et al. (2006) investigated exercise interventions during cancer treatment across all types of
cancers. There were significant improvements in physical function, symptoms other than fatigue
and body composition. Furthermore, modest improvements were documented for mood, quality
of life, and fatigue.
McNeely et al. (2006) examined the sub-group of women with breast cancer and presented ef-
fects of exercise interventions during and after cancer treatment jointly. Significant effects were
observed of improvements in cardiorespiratory fitness, fatigue, and quality of life.
In sum, previous reviews have presented uniformly affirmative conclusions regarding exercise.
However, many of the findings from the early studies were based on small sample sizes and were
weak in terms of methodological quality. In recent years, several larger scale experimental studies
have complemented previous work.
2.7 Preliminary work: rehabilitation sport in Germany
Exercise opportunities (tailored to the needs of breast cancer survivors who have completed their
adjuvant treatment) are already implemented in the communities in Germany: the opportunity for
regular exercise participation in rehabilitation sports groups is provided within the organised
sports. These rehabilitation sports groups are available to women who have been treated for
breast cancer previously (post-treatment); currently these exercise programs are not determined by
the needs of women during treatment.
Exercise for cancer patients in Germany is counted among rehabilitation sport. Currently, there
are more than 600 rehabilitation sports groups for women with breast cancer; 300 alone in
Setting the context
38
Nordrhein-Westfalen (Landessportbund Nordrhein-Westfalen 2009). In 1991, a training curricu-
lum for exercise facilitators was established and since 1992 exercise for cancer patients has been
included in the German Sports Federation guidelines for exercise facilitator training (Große-
Waechter et al. 2004). The legal foundations for rehabilitation within organised sports are con-
tained in Book 9 of the German social code, SGB IX (covering rehabilitation and participation of
people with disabilities). According to SGB IX, exercise rehabilitation programs are to be pro-
vided and funded as supplementary benefits to rehabilitation. Rehabilitation sports groups for
breast cancer patients are formally prescribed by the treating physician and led by qualified in-
structors (Übungsleiter/innen) to ensure competent and skilful guidance and surveillance of
classes. However, a physician is always on-hand to counsel participants and the exercise facilita-
tors when required. The cost of participation in these programs is reimbursed to sports clubs by
health or retirement insurance funds. Provision of rehabilitation sports groups for women with
breast cancer is growing but has not yet achieved full coverage, particularly not in rural areas.
Altogether, Germany has embraced the importance of regular exercise in the rehabilitation of
cancer patients. The implementation of exercise opportunities within the national sports federa-
tion for this target group seems to be unique. This provides a model of how exercise programs
for breast cancer survivors can be offered within a health care context. However, there is still the
failure of exercise opportunities for women during adjuvant treatment of breast cancer.
Methods
39
3 Methods
The objectives of this piece of work were to assess first, the effects of a structured exercise training
program on physical functioning and health-related physical fitness in women receiving adjuvant
treatment of breast cancer and secondly, to determine the effect of exercise training on secondary
outcome measures such as treatment-related symptoms, biological and physiological problems, men-
tal health, and health-related quality of life. Finally, this review aimed at assessing harm associated
with physical exercise during adjuvant treatment of breast cancer.
The scope of application of these objectives was restricted with respect both to the type of cancer
(breast cancer) and to the timing of the intervention with regard to cancer treatment (during adjuvant
treatment). This restriction is for two reasons: First, cancer is a common denominator for a large
number of neoplastic diseases, each with a different aetiology, course, and prognosis and a disease
pathophysiology that may vary considerably. Similarly, cancer patient groups are clinically heteroge-
neous in terms of their demographic profile (e.g., age, sex distribution), behavioural profile (e.g.,
smoking status, alcohol consumption, and obesity), treatment protocols, and symptoms and side
effects. Accordingly, summarising the effects of exercise interventions across such disparate groups
appeared debateable and problematic. Secondly, cancer patients, who are undergoing adjuvant cancer
treatment, are confronted with a number of specific problems. To recap, current cancer treatments
are toxic in numerous ways and produce harmful short- and long-term health effects. Clinicians who
wish to prescribe exercise for women currently undergoing cancer treatment need to know whether
exercise will improve physical and mental health in this period of treatment as well. Clinical advice
needs to be based on studies conducted on patients who are at a similar point of the breast cancer
experience (pre-treatment, during treatment, post treatment). For example, a study that reports that
exercise is useful to combat fatigue among women who have completed treatment will not assist a
clinician in deciding whether to prescribe exercise for cancer-related fatigue during adjuvant cancer
treatment.
A systematic review was chosen as an appropriate research design in order to answer the research
question since several primary studies already exist. The methodology of a systematic review an-
swered the purposes of reducing the quantity of data, of improving precision and reducing bias, and
of checking consistency across results and explaining inconsistencies if present.
Methods
40
3.1 The role of systematic reviews in evidence based medicine
There is a body of published medical literature that is often overwhelming; moreover, individual
research studies may vary in quality, have conflicting results, or fail to set their findings in the context
of previous studies. Systematic reviews, however, identify, appraise, and synthesise research evidence
from individual studies. A systematic review can be defined as “a review of the evidence on a clearly
formulated question that uses systematic and explicit methods to identify, select and critically ap-
praise relevant primary research, and to extract and analyse data from the studies that are included in
the review” (NHS Centre for Reviews and Dissemination 2001). Compared to narrative reviews that
use informal, unsystematic and subjective methods to collect and interpret information, and that
often summarise information subjectively and in a narrative way, systematic reviews have the advan-
tages of reducing bias, being replicable and of providing a reliable basis for decision making. A sys-
tematic review establishes “whether scientific findings are consistent and can be generalised across
populations, settings, and treatment variations, or whether findings vary significantly by particular
subsets” (Mulrow 1994) and thus conducting a systematic review is central to the practice of evi-
dence based medicine, the “conscientious, explicit, and judicious use of current best evidence in
making decisions about the care of individual patients” (Sackett et al. 1996).
Nevertheless, systematic reviews are retrospective research, and are thus potentially subject to many
of the same biases that affect other retrospective studies (Cook et al. 1995). Therefore, a rigorous
review methodology is required. This review was conducted in co-operation with the CBCG and was
first published in the Cochrane Database of Systematic Reviews (CDSR) in October 2006 (Markes et
al. 2006). Each systematic review conducted under the umbrella of the Cochrane Collaboration fol-
lows a quality standard which provides a prescribed format and feedback during the review process
from submission of a title, through development and publication of a protocol, to publication of the
completed review. Each Cochrane review is prepared within one of the Collaborative Review
Groups with a focus on a particular area of health. All these groups have an editorial base where a
small team of people supports the production of Cochrane reviews. This review conformed to the
formal, methodological, and procedural features of a Cochrane review. Cochrane reviews have the
following general features (Higgins and Green 2006):
• Structured format
• Detailed methods section
• Thorough and systematic search strategy
• Quality assessment of included studies using predefined criteria
Methods
41
• Quantitative data synthesis in a meta-analysis if appropriate
• Co-operation with ‘consumers’ in order to account for issues important to people receiving certain
interventions
• Peer review in multi-national editorial teams in order to ensure that a review is applicable in differ-
ent parts of the world, and a
• Regular update of the review through incorporating results from newly completed or identified
clinical trials after publication.
3.2 Execution of the review
The work breakdown structure of the review (Figure 4) describes how the review was executed. All
methodological steps are shown here and described in detail within this methods section.
3.2.1 Review protocol
Since systematic reviews are by their nature retrospective, the methods to be used were established
and documented in advance. A protocol for the review (see Appendix 1) was published prior to
knowledge of the available studies. The aim of this procedure was to promote transparency of meth-
ods and processes, to reduce the potential for duplication, and finally to allow for peer review of the
planned methods.
3.2.2 Identification of research
The aim of this step was to generate as comprehensive as possible a list of primary studies, both
published and unpublished, which are suitable for answering the questions whether exercise contrib-
utes to the improvement of physical and mental health and subsequently, whether women, who are
undergoing adjuvant treatment of breast cancer, should be offered exercise. A variety of search
methods (both computerised and manual) was used to ensure that the search would be as compre-
hensive and unbiased as possible.
Methods
42
Figure 4: Work breakdown structure
Systematic review: Exercise for women
receiving adjuvant treatment of breast cancer
Planning Identification
of research
Selection of
studies Data extraction
Study quality
assessment
Data
synthesis Publication
Review title
Electronic
databases
Guidance of
selection
Guidance of
extraction
Methodological
quality
Descriptive
synthesis Cochrane Library
Compile
proposal
Develop
search
strategy
Develop eligibil-
ity criteria/form
Develop
extraction form
Choose quality
assessment tool
Describe
studies
Compile report
Submit to
CBCG
Conduct
search
Formulate in-
structions for
implementing
Formulate instruc-
tions for imple-
menting
Formulate instruc-
tions for applying
assessment tool
Discuss
hetero-
geneity
Submit to CBCG
Peer review Conduct pretests
Cooperate in peer
review process
Publication Assess quality Publication
Calculate Inter-
Rater reliability
Review
protocol
Other
sources
Selection
process
Extraction
process
Intervention
quality
Quantitative
synthesis
Compile
protocol
Conduct
Hand-
searching
Screen title,
abstracts
Conduct pretests
Develop quality
assessment tool
Conduct
meta-
analyses
Submit to
CBCG
Contact
experts
Screen full
reports
Extract data
Formulate instruc-
tions for applying
assessment tool
Conduct
sensitivity
analyses
Cooperate in
peer review
process
Check con-
ference
proceedings
Assess quality
Publication
Check refer-
ence lists
Methods
43
Electronic databases were used as the main source for identifying relevant research. Since ex ante,
no single electronic database could be identified as comprehensive in the review topic and be-
cause databases also tend to have geographical and language biases, the search strategy included a
range of databases:
• MEDLINE with records of biomedical literature
• EMBASE with records of biomedical literature
• CINAHL (Cumulative Index to Nursing and Allied Health Literature) with records of litera-
ture on all aspects of nursing and allied health disciplines
• PsycInfo with records of research in psychology and related behavioural and social sciences
• SPORTDiscus with records of research in sport, health, fitness and sports medicine
• Cochrane Breast Cancer Specialised Register (CBCSR) with records of trials in healthcare
regarding breast cancer identified through the work of the CBCG.
A separate search strategy was developed for each database. In order to maximise the yield of
relevant articles, the search strategies were built to make the literature search sensitive, rather than
specific. So, it was expected to retrieve a large proportion of articles, which would not be relevant
to the review question. After an initial search for relevant studies, a series of update searches were
conducted to identify new studies which appeared after the initial search.
The following structured approach was used in order to develop search terms that could be effec-
tively combined for searching electronic databases: first, the review question was broken down
into its facets of population, intervention, and study design. Then, the search terms in each facet
which best captured the subject were identified. The group of search terms covering each facet of
the review question included textwords (free text) in the title and abstract of studies as well as any
available subject indexing terms that were assigned by the database producer. Textwords and
their variants were chosen based on relevant studies on the review topic that had already been
identified. Information on the subject indexing used by databases was thesaurus-derived. Terms
within one facet were joined with the Boolean ‘OR’ operator; these sets of terms, describing each
facet, were then joined together with the ‘AND’ operator. The search strategy for MEDLINE
implemented the standard Cochrane search strategy for randomised controlled trials (RCTs), i.e.,
the search strategy used for identification of RCTs in Cochrane reviews. Because of the complex-
ity of the question it seemed to be more feasible for most databases to omit the facet describing
adjuvant cancer treatment from the search strategy. The periods “during treatment” and “post-
treatment” were distinguished later, at the stage of title and abstract checking. The development
Methods
44
of the search strategy was an iterative process and the final strategy was built up from a series of
pilot searches.
Reference lists from publications (primary studies and reviews) that were found through database
searches were scanned to identify further studies for consideration. These search activities were
complemented by handsearching, and by searching grey literature and conference proceedings.
Key journals, relevant in the fields of exercise and health promotion in the cancer diagnosis-
rehabilitation continuum, were hand-searched to identify articles that might had been missed in
database and reference list searches. Hand-searching means that a journal was searched page by
page (i.e., by hand), including editorials, letters, etc., to identify reports of randomised controlled
trials and controlled clinical trials. Grey literature was searched in the System for Information on
Grey Literature (SIGLE), where reports, technical reports, discussion papers or other formats,
which are not indexed in the major databases, are indexed. Dissertations and theses were
searched for in ProQuest Digital Dissertations, which provides access to American dissertations,
and Masters’ theses and in DissOnline.de, the information system for German dissertations.
Conference proceedings that provide information on research in progress as well as completed
research were searched for via relevant specific conferences and Conference Papers Index
(CONFSCI), which is a bibliographic database from Cambridge Scientific Abstracts that provides
access to international research papers and findings presented at scientific and technical confer-
ences and meetings throughout the world. To further supplement identification of research, eight
research registers were searched for ongoing research. Finally, after a thorough and systematic
search had been conducted, a list of studies that met the inclusion criteria was sent to experts in
the field of exercise interventions in cancer patients. They were requested to check the list for
completeness, and to provide information on any ongoing research that could be considered for
inclusion in the review. A detailed summary of search activities can be found in Appendix 2.
3.2.3 Search updates within duration of review implementation
The principal search for study identification was conducted in 2004 after the review protocol had
been published. This search was updated quarterly during the implementation phase of the re-
view in order to capture newly published reports; the last search for primary studies was con-
ducted in September 2008. MEDLINE and CINAHL turned out to be comprehensive databases
in the review topic. Updating searches were conducted in MEDLINE and the Cochrane Breast
Cancer Specialised Register; CINAHL was not accessible over the whole review period and sub-
sequently was not searched again after the first basic search.
Methods
45
3.2.4 Selection of studies
Study selection was a multi-stage process that started with the development of study selection
criteria. Inclusion and exclusion criteria were developed directly from the review question: Should
women, who are undergoing adjuvant treatment of breast cancer, be offered exercise? These
criteria were defined in terms of the population, the intervention, the outcomes, and the study
designs of interest. Only studies that met the inclusion criteria were included in the review. With
regard to the population, trials were included that reported on women receiving adjuvant treat-
ment (chemotherapy, hormonal therapy, or radiotherapy) for breast cancer. Breast cancer was
restricted to stages 0-III. If trials had included women with stage IV breast cancer, i.e., with dis-
tant metastasis, these trials would have been excluded from this review. Also trials that included
women who had already completed adjuvant cancer treatment or were treated for cancers other
than breast cancer were excluded. With regard to the intervention, trials with an intervention
consisting of aerobic exercise training or resistance exercise training were included. Aerobic exer-
cise training was defined as a training to improve cardiorespiratory endurance and the efficiency
of the aerobic-energy producing systems. This type of exercise involves the use of large muscle
groups over prolonged periods in activities that are rhythmic and aerobic in nature (e.g., swim-
ming, walking, hiking, cycling, rowing, or endurance games). Resistance training was defined as a
training to increase muscle endurance – i.e., the ability of a muscle group to execute repeated
contractions over a period of time sufficient to cause muscular fatigue or to maintain a specific
percentage of the maximum voluntary contraction for a prolonged period of time. Trials that
examined complex exercise interventions – such as a program of exercise training, and diet, or a
program of exercise training and behavioural therapy – were excluded. Also, trials with exercise
interventions restricted to selected body functions only (e.g., arm mobility) were excluded. With
regard to outcome measures, trials were included that employed at least one outcome measure
out of the following outcome measurement categories: physical functioning, health-related physi-
cal fitness, symptom experience, biological or physiological outcomes, mental health, health-
related quality of life, and harm. Study designs that qualified for inclusion into the review were
randomised as well as non-randomised controlled trials. No language restrictions were applied.
Selection criteria are tabulated in Table 1.
Methods
46
Table 1: Inclusion criteria for reviewing benefits of exercise
Inclusion criteria Exclusion criteria
Participants - Women receiving adjuvant treatment of
breast cancer
- Breast cancer stages I-III
- Adjuvant treatment com-
pleted
- Cancer other than breast
cancer
Intervention - Aerobic exercise
- Resistance exercise
- Mixed exercise (aerobic and resistance)
- Exercise as part of a com-
plex intervention
- Exercise restricted to local
muscular endurance
- Stretching or flexibility
training
Outcome - Physical functioning (objective mobility,
perceived mobility, participation in life
activities)
- Physical fitness (cardiorespiratory fit-
ness, muscular strength, muscular en-
durance, body composition)
- Symptom experience
- Biological and physiological variables
- Mental health (e.g., depression, anxiety)
- Health-related quality of life
- Harm
- Adherence data only
Study design - Randomised controlled trials
- Non-randomised controlled trials
- Uncontrolled trials
- Group assignment based
on self-selection
Initially, the selection criteria were applied liberally to the citations generated from computer
database searching: reports potentially fulfilling the inclusion criteria of this review were selected
based on title and abstract. The titles and abstracts identified as being potentially relevant were
provisionally included. Then, full text articles of any possibly relevant reports were retrieved for
more detailed evaluation. For the final selection of trials an eligibility form was used. The eligibil-
ity form was designed for this review in order to systematically assess eligibility based on the pre-
defined inclusion and exclusion criteria (see Appendix 3). The decision algorithm required the
fulfilment of the predefined eligibility criteria (see Table 1). A report was excluded according to
the first criterion that it did not fulfil. Study selection was performed independently by two re-
view authors to improve reliability. Disagreements were resolved by consensus; if necessary a
third person was consulted to reach a final decision. Figure 5 describes the process flow of the
review – from study selection to meta-analysis. These steps are also described in the following
sections.
Methods
47
3.2.5 Data extraction
The aim of data extraction is to produce a data repository out of which the analysis could emerge
by means of condensing information from that reported by primary investigators. Because this
can be a subjective process and prone to error, several measures were implemented in order to
minimise bias at all stages of this process: first, the review protocol already contained a data ex-
traction form which listed the data items to be extracted from each of the primary studies and
additionally, a statement of practice for implementing data extraction (see Appendix 4) was de-
veloped that included instructions and decision rules about extracting data because this improves
accuracy and consistency in data extraction which is extremely important. Finally, data extraction
was performed independently by two data extractors to improve reliability. Since published re-
ports did not always provide all the information that needed to be extracted, authors of the re-
spective studies were contacted in such circumstances to provide further information. The data
extracted by the review authors were compared and any disagreements were discussed and re-
solved by consensus among the data extractors. A separate form was used to denote and correct
errors or disagreements; thus, a historical record of the decisions and refinements that occurred
during the conduct of the review was obtained.
The final data extraction form comprised two rather different parts: one part that contained
information about study characteristics, i.e., the study descriptors (see Appendix 4), and the part
that contained information about the empirical findings of the study, i.e., effect measure data (see
Appendix 4). There is a practical reason for distinguishing between study descriptors and effect
size data: study descriptors apply to the entire study and need to be extracted only once for a
given study because it is the same for different variables, follow-ups, sample breakdowns, and the
like. In contrast, a study can include numerous measured variables that represent different con-
structs and so data for several distinct effect sizes have to be extracted for each study. Therefore,
a complete data extraction per study included one set of data on the study level and as many sets
of effect measure data as needed to extract all the relevant quantitative findings that the study
reported. A study was regarded as consisting of a set of data collected under a single research
plan from a designated sample of respondents. Subsequently, data from multiple written reports
of one study were assembled together in the process of data extraction and analysis.
The data extraction form for study descriptors comprised study characteristics such as study
methodology, samples, intervention, and setting, all of which representing factors that may influ-
ence the nature and magnitude of empirical findings. It was based on open-ended items (as op-
posed to pre-coded items) to which information relevant to the issue at hand was recorded from
the study. These responses were later examined for commonalities and then coded into a man
ageable set of categories with the aim of building data files for data synthesis.
Methods
48
Studies retrieved for
more detailed
evaluation
Study eligible?
Eligibility form
Data collection
form for study
descriptors
Code of practice
for data
collection
Data extraction
Description of
reasons for
exclusion
no
Listing of included
studies
yes
Assessment of
methodological
quality
Study quality
form
Code of practice
for implementing
qualiy criteria
Evaluation of
Inter-Rater-
Reliability
Criteria for
assessing
training stimulus
Assessment of
intervention
quality
Description of
included studies
Meta-Analysis
appropriate?
Meta-Analysis
yes
Narrative
synthesis
no
Screening of full
reports
Data collection
form for effect
measure data
Figure 5: Process flow of review steps
Methods
49
The coding frame comprised methodological, participant, and intervention characteristics need-
ing categorisation. Methodological characteristics included study design (e.g., randomisation) and
all information needed for quality assessment. Participant characteristics included age, breast
cancer staging, previous physical activity, and type of cancer treatment. Exercise attributes coded
included the setting of the intervention, whether a behaviour change theory was applied or not,
type of exercise, intensity, frequency per week, and duration per session. With regards to out-
come measures, outcome constructs and assessment instruments for each construct were ex-
tracted.
Intervention effect data was recorded as a profile of associated information describing the par-
ticulars of the statistical findings. As a first step, data on outcomes were extracted in the format in
which they were reported; in a second step, data were transformed regarding the requirements of
RevMan Analyses, the computer program of the Cochrane Library used for statistical pooling.
The following data was extracted from trial reports for analyses: For dichotomous outcomes, the
number of events and the number of participants in each group was extracted. For continuous
outcomes the mean values of the outcomes, the standard deviations of the outcomes, and the
number of participants for whom the outcome had been assessed were extracted in each of the
two groups. Further, some supporting information was extracted such as nature of the variable
(e.g., construct measured, measurement operationalisation, or presentation of variable as mean
change from baseline to final values or as final value) or point in time when a variable was meas-
ured and time covered (end of intervention or follow-up).
3.2.6 Study quality assessment
The assessment of trial quality was a crucial part of the process of this systematic review, since
the validity of the conclusions of a systematic review depends on the methodological quality of
the primary studies included. Assessment of methodological quality of the included trials was
guided by the definition of methodological quality as “the likelihood of the trial design to gener-
ate unbiased results that are sufficiently precise and allow application in clinical practice” (Verha-
gen et al. 2001). Components of methodological quality can influence effect estimates in two
directions: both underestimation and overestimation of the true effects seem to be possible when
study quality is low (Verhagen et al. 2001).
As a quality assessment instrument, a variant of the vanTulder scale was applied (vanTulder et al.
1997): this instrument is based on 17 quality items, which are scored numerically to provide a
quantitative estimate of overall study quality. Scores were generated by weighting all items
equally. The vanTulder scale is a quality assessment instrument that has already been used in a
number of systematic reviews in the field of physiotherapy and therefore appeared to be adequate
Methods
50
for the purposes of this review. This scale includes a set of generic core items with all criteria of
the Jadad score (Jadad et al. 1996) and the Delphi List (Verhagen et al. 1998); both instruments
were developed according established principles used for creating measurement scales. The items
of the scale consist of internal validity criteria, descriptive criteria, and criteria of statistical analy-
sis (Table 2). The internal validity criteria refer to characteristics of primary studies that might be
related to selection bias (i.e., systematic differences between comparison groups in prognosis or
responsiveness to the intervention), performance bias (i.e., systematic differences in care pro-
vided to the participants in the comparison groups apart from the intervention under investiga-
tion), attrition bias (i.e., systematic differences between comparison groups arising from with-
drawals or exclusions of participants from the study sample), and detection bias (i.e., systematic
differences between comparison groups with respect to how outcomes are determined). Accord-
ingly, randomisation and allocation concealment, co-interventions, observer blinding, and inten-
tion-to-treat analysis were among those checked in the quality scale. Trials designed to compare
exercise interventions to their absence cannot blind neither participants nor exercise facilitators.
Therefore, methodological quality regarding blinding was judged solely on blinding assessors who
collect outcome data: the aim of blinding assessors is to reduce differential assessment of out-
comes. Descriptive criteria of the vanTulder scale refer to external validity: information derived
from descriptive criteria (eligibility criteria, similarity of baseline characteristics, and description of
intervention, adverse effects, and timing of follow-up measurements) allow assessment of
whether effects observed in primary studies may be generalised. Quality forms were used to per-
form quality assessment (see Appendix 5) and a code of practice for implementing quality criteria
was developed in order to allow reliable quality assessment (see Appendix 5).
Each criterion was answered with ‘yes’, ‘no’ or ‘n/a’ (not available); answers were coded with:
‘yes’ as ‘+’, ‘no’ as ‘’ and ‘n/a’ as ‘?’. A quality score was calculated for each study by summing
positive scores for individual items, resulting in a possible score from 0 to 17. Following van-
Tulder (vanTulder et al. 1997), moderate quality was defined as fulfilling a score between nine
and 12 (50 to 75%) of all methodological quality criteria; high quality was defined as fulfilling at
least 13 (> 75%) of all methodological quality criteria.
The process of assessing methodological quality was guided by the intent to distinguish between
the quality of the trial and the quality of its report. Following Moher, the quality of a report was
regarded as “providing information about the design, conduct, and analysis of the trial” (Moher
et al. 1995).
Methods
51
Table 2: Quality criteria assessed with the vanTulder scale
Quality aspect Quality criteria assessed with the vanTulder scale
Internal validity
- Selection bias
- Performance bias
- Attrition bias
- Detection bias
- Was a method of randomisation performed?
- Was the treatment allocation concealed?
- Were co-interventions either avoided or comparable?
- Was adherence acceptable in all groups?
- Was the withdrawal/ drop-out rate acceptable and were reasons
described?
- Did the analysis include intention-to-treat analysis?
- Was the outcome assessor blinded to the intervention?
- Was timing of outcome assessment comparable in both groups?
- Were outcome measures relevant?
Description - Were the eligibility criteria specified?
- Were groups similar at baseline?
- Were interventions explicitly described?
- Were adverse effects described?
- Was short-term follow-up measurement performed?
- Was long-term follow-up measurement performed?
Statistics - Was the sample size described for each group at randomisation?
- Were point estimates and measures of variability presented for
the primary outcome measures?
This distinction between the trial and its report is important to avoid the case where a well re-
ported trial with several biases would receive a high-quality score or where conversely, a well
designed and conducted but poorly reported trial would receive a low quality score. Therefore,
where it was unclear whether it was the reporting of the trial that was weak or its design and
conduct, authors of primary studies were contacted to ask for more detailed information. With
over-reporting, on the other hand, the attempt to identify and then subject to quality assessment
was made by means of items asking for the details of trial design and conduct: for example, trials
were assessed with respect to whether they had performed a method of randomisation, some-
thing some “randomised” studies may actually not have done (Verhagen et al. 2001). Methodo-
logical quality based on the summary quality score was intended to be used as one criterion in
sensitivity analyses by estimating a pooled effect with studies above different methodological cut-
off points.
However, scoring systems tend to be subjective and the reliability of this list is unknown, as is the
case for most criteria lists. Therefore methodological quality was assessed independently by two
Methods
52
raters in a subset of studies. Disagreements in quality assessments between the two raters were
resolved by consensus and, if disagreement persisted, a third person was consulted for a final
decision. Inter-rater agreement for the overall methodological quality assessment was measured
using the intra-class correlation coefficient (ICC). The ICC is based on the analysis of variance
(ANOVA) model. Variance in quality scoring between cases (here: primary studies) was distin-
guished from the variance in quality scoring within primary studies that indicates lacking agree-
ment between raters. When differences in quality scores are high between primary studies with
simultaneously low variance in quality scores within primary studies, observations are regarded to
be reliable. ICC calculation in SPSS (Statistical Product and Service Solutions) was based on the
two-way mixed model with measures of absolute agreement (Shrout and Fleiss 1979): differences
among primary studies and variability among the raters were treated as the two systematic sources
of variance; differences among primary studies were treated as a random factor in the ANOVA
model, variability among the raters becomes the second factor in the two-way ANOVA model
which is treated as a fixed factor, resulting in the two-way mixed model. In the mixed model,
inferences are confined to the particular set of raters used in the quality assessment process. The
ICC is interpreted as the proportion of total variance accounted for by variation within primary
studies. The level of reliability for the ICC was described in this review using the same categories
that were suggested by Landis and Koch (1977) for Kappa, as another measure of inter-rater
agreement. ICC values of 0.21 to 0.40 were considered as ‘fair agreement’, values from 0.41 to
0.60 as ’moderate agreement’, from 0.61 to 0.80 as ‘substantial agreement’ and from 0.81 to 1.00
as ‘almost perfect agreement’. The ICC was calculated for a sample of included studies solely to
assure sufficient reliable quality assessment.
The exercise intervention was evaluated separately regarding its potential to provide any adequate
training stimulus. Aspects that were assessed are exercise intensity, exercise frequency, duration
of single exercise sessions, and duration of the exercise intervention program. Analysis of the
training stimulus roughly followed the ACSM guidelines for exercise prescription for the elderly
(Balady et al. 2000), the ACSM references for exercise programming for cancer patients
(Schwartz 1997) and the exercise prescription guidelines for early-stage cancer patients and can-
cer survivors as recommended by Courneya (Courneya et al. 2000). Trials were classified as pro-
viding an adequate training stimulus if they met three of four predefined requirements (see Table
3 and Table 4).
The prescribed exercise intensity should be above the minimal level required to induce a “training
effect”, yet below the metabolic load that evokes abnormal clinical signs or symptoms. For
women receiving adjuvant treatment of breast cancer aerobic endurance training of moderate
intensity, which is dependent on current fitness level and medical treatments, seems to be feasi-
Methods
53
ble. Guidelines recommend 50% to 75% of maximal oxygen consumption (VO2max), i.e., the
highest rate of oxygen consumption attainable during maximal or exhaustive exercise; or 50% to
75% of heart rate reserve (HRreserve), i.e., the difference between the maximum heart rate and
resting heart rate; or 60% to 80% of the maximum heart rate (HRmax), i.e., the heart rate that a
person could achieve during maximal physical exertion; or an rate of perceived exertion (RPE) of
11 to 14 (Courneya et al. 2000). HRreserve is the best guideline if HRmax is estimated rather than
measured. Furthermore, program duration had to be at least six weeks with a frequency of at least
three times a week with sessions of at least 20 minutes duration (see Table 3).
Table 3: Criteria for assessing quality of aerobic endurance training
Training stimulus aspect Quality criteria
Intensity - 60-80% of maximum heart rate(HRmax) or
- 50-75% of heart rate reserve(HRreserve) or
- 50-75% of maximum oxygen consumption (VO2max)
- 11-14 rate of perceived exertion (RPE)
Duration per session 20-60 min (continuous bouts of minimum 10 minutes
each throughout the day) or exercise to tolerance
Frequency per week At least three days per week
Program duration At least six weeks
Regarding resistance training, a minimum of eight to ten separate exercises that train the major
muscle groups (arms, shoulders, chest, abdomen, back, hips, and legs) should be performed.
Muscular strength as well as muscular endurance is developed by the overload principle, i.e., by
increasing the resistance to movement or the frequency or duration of activity to levels above
those normally experienced (Balady et al. 2000). The intensity of resistance training can be ma-
nipulated by varying the weight, the number of repetitions, the length of the rest interval between
exercises, or the number of sets of exercises completed (Balady et al. 2000). The training stimulus
of resistance training was regarded as adequate when a minimum of one set, with ten to 15 repeti-
tions of each of these exercises, was performed to the point of volitional fatigue. As with aerobic
endurance training, resistance training in primary studies had to be performed on two to three
days per week, with program duration of at least six weeks, to be assessed as adequate (see Table
4). Intervention quality forms were used to perform quality assessment (see Appendix 5).
Each criterion was answered with ‘yes’, ‘no’ or ‘n/a’ (not available); answers scored as ‘yes’ were
coded as ‘+’, ‘no’ as ‘’ and ‘n/a’ as ‘?’. A quality score was then calculated for each study by
summing scores for individual items, resulting in a possible score from zero to four. In trials
using exercise interventions with both aerobic and resistance modules, the training stimulus was
assessed for each module and then the corresponding mean was calculated.
Methods
54
Table 4: Criteria for assessing quality of muscular endurance training
Training stimulus aspect Quality criteria
Intensity - 10-15 repetitions to near fatigue or
- At least 60% of one Repetition Maximum
Duration per session At least 1 set
Frequency per week At least two or three days per week
Program duration At least six weeks
The next phase following quality assessment and data extraction was to collate and summarise
the data that had been extracted from the primary studies. This was accomplished through a
descriptive, non-quantitative synthesis that was complemented by the use of meta-analysis. In
descriptive synthesis, study characteristics were tabulated to qualitatively assess variation in par-
ticipants, interventions, and measurement of outcomes across studies. Furthermore, the tabular
summaries were used to plan the quantitative synthesis.
3.2.7 Summarising effects across studies
Quantitative synthesis through meta-analysis focuses on the aggregation and comparison of the
findings of different research studies. The reason for combining data from individual studies is
that these may not be able to estimate effects precisely, because of small sample sizes. By com-
bining the data from these studies, a meta-analysis acquires the statistical power to increase the
precision of the estimate of effect. However, the focus lays on magnitude and direction of a
relationship, not merely its statistical significance.
The steps involved in meta-analysis are to create an independent set of relevant effect sizes for
each construct to be analysed, compute the weighted mean and the confidence interval for the
mean, and to test for homogeneity.
3.2.7.1 Creating an independent set of effect sizes
Statistical independence was defined at the study level: if a study presented more than one effect
measure for a construct by using different measurement operationalisations (e.g., body composi-
tion assessment through weight, body mass index, relative lean body mass, skinfolds), they were
not included in the same analysis. In such cases where multiple effect sizes were presented for
one outcome, only the effect size that was most comparable with those in other studies was se-
lected for inclusion in the meta-analysis.
The effect measure (treatment effect, estimate of effect) is the statistical representation of the
observed relationship between an intervention and an outcome. Outcome assessments across
included studies predominantly employed either measurements (e.g., body weight or aerobic
Methods
55
capacity) or assessment scales (e.g., Piper Fatigue Scale or Beck Depression Inventory). These
assessment scales actually yield ordinal data, but were analysed as continuous outcomes in the
studies included and were subsequently treated as continuous outcomes in meta-analysis as well.
This procedure appeared to be adequate because with increasing numbers of categories in as-
sessment scales, ordinary outcomes acquire similar properties to continuous outcomes. Thus,
meta-analyses had to deal predominantly with continuous data. For continuous data, the treat-
ment effect is expressed as a difference in means or standardised difference in means. The unstan-
dardised mean difference effect size statistic is constructed directly from the differences between the
group means; it is applicable when the same operationalisation of a variable of interest, with the
same measurement procedures and the same numerical scale, is used in all of the research find-
ings to be meta-analysed. In the Cochrane Database of Systematic reviews, the unstandardised
mean difference is termed the weighted mean difference (WMD) which is a confusing term since no
weighting is involved in calculating the statistical summary of a single trial.
If, however, different measurements or scales are used for assessing the same outcome the stan-
dardised mean difference (SMD) is used. This effect size statistic applies to research findings that
contrast two groups on their respective mean scores on some dependent variable that is not
operationalised in the same way across study samples. Under these circumstances, study results
are standardised on a uniform scale before they are combined. The standardised mean difference
effect size statistic expresses the size of the intervention effect in each study relative to the vari-
ability observed in that trial and is calculated according to:
SMD
pooledsd
mm 21 −
=
where m1 is the mean for group 1, m2 is the mean for group 2, and sdpooled is the pooled standard
deviation. By standardising the difference between exercise and control group means on the
corresponding pooled standard deviation, intervention effects are represented in terms of stan-
dard deviation units irrespective of the original operationalisation and so can be meaningfully
combined and compared across studies.
For outcomes with dichotomous data (e.g., onset of lymphedema) the effect measure was gener-
ated as a measure of relative effect using the risk ratio (relative risk), i.e., the ratio of the probabil-
ity of the event occurring in the exercise group versus a non-exercising control group.
3.2.7.2 Meta-analysis
Results from individual studies were statistically combined to provide a weighted average estimate
of the overall intervention effect. A weighted average is defined as
Methods
56
Weighted average
d
i
ii
W
WES
∑
∑
=
Where ESi is the intervention effect (effect size) estimated in the ith study, Wi is the weight given
to the ith study, and the summation is across all studies.
Meta-analysis employs weighted analysis for all data-analyses involving effect sizes, because from a
statistical perspective effect size values based on larger samples do more precisely estimate the
corresponding population value than those based on smaller samples, due to a smaller sampling
error in large samples compared to that in small samples. Hence, there is the problem of variable
reliability of the information that the effect sizes carry. The way this problem is handled in meta-
analysis is to weigh each effect size by a term that represents its precision, so that its contribution
to any statistical analysis is proportional to its reliability. Effect sizes based on larger samples are
weighted more in statistical analysis than those based on smaller samples because they embody
less sampling error. The optimal weights are based on the standard error of the effect size
(Hedges and Olkin 1985). Each effect size is weighted by the inverse of its variance (calculated as
the square of the standard error).
For computing the mean effect size with the corresponding confidence interval, a random effects
model was chosen, since the differences between studies arising from variations in the exercise
program, settings etc. whose sources cannot be identified, are expected to be random. The ran-
dom effects model includes another random component in the statistical model in addition to
subject-level sampling error to represent the variation among effect sizes. This means that the
weight applied to each effect size represents both subject-level sampling error and the additional
random variance component assumed by the model.
Data were analysed using the RevMan Analyses statistical package in Review Manager (RevMan
2003). RevMan does accept summary data in only two formats – events and sample size (for
dichotomous outcomes), or means and standard deviations (for continuous outcomes). If any
studies provide data in another format (such as mean and confidence intervals) it was necessary
to manually manipulate the statistical information available in the reports to estimate required
statistics for those studies. These estimation procedures can be distinguished into several catego-
ries with different degrees of approximation to the required statistics:
• Descriptive data from which means (m) and standard deviations (SD) can be computed (e.g.,
standard errors or confidence intervals)
• Complete significance tests along with sample sizes (e.g., t-values from a t-test or F-values
from a one-way ANOVA)
Methods
57
• Exact p-values for a t-test or a one-way ANOVA, and the sample size for each group, or the
total for both, or
• Categorical p-values for a t-test or a one-way ANOVA, and the sample size for each group, or
the total for both.
In the meta-analyses conducted within this review, the required statistics were estimated from
descriptive data and from significance tests, the approaches with higher levels of approximation.
Standard deviations were estimated from the results of significance tests as follows. First, the
standard error of the difference in means ( 21 m-m
SE ) was calculated by dividing the difference in
means by the t-value (the t-value in independent t-tests is the ratio of the difference in means to
the standard error of the difference in means).
Standard error of the difference in means ( 21 m-m
SE )t
mm 21
−
=
The standard deviation was then obtained from the 21 m-m
SE using the following formula:
21
m-m
11
SE 21
nn
sd
+
=
This standard deviation was then allocated to both, the exercise and to the control group. This
method was also used for test statistics from a one-way F-ratio based on only two groups,
where Ft =². Theoretically it is possible to obtain an estimate of the standard deviation for a
specific measure from another study that used that measure with a very similar sample. Neverthe-
less, within these meta-analyses, no effect size was calculated based on the standard deviation for
a particular variable of another study, since samples did not appear to be comparable.
The formulation of the standardised mean difference implemented in RevMan Analyses is
Hedges adjusted g, which is very similar to Cohen’s d, but includes an adjustment for small sam-
ple bias. Further, the version of random effects meta-analysis implemented in RevMan Analyses
is that described by DerSimonian and Laird (DerSimonian and Laird 1986). This version incorpo-
rates the heterogeneity of effects for obtaining an estimate of the random effects variance com-
ponent. Weighted individual effect sizes are combined to give a pooled estimate (DerSimonian
and Laird Random Effect): each effect size value is multiplied by its respective weight (subject-
level sampling error and the additional random variance component), then summed and divided
by the sum of weights. The standard error of the pooled intervention effect is used to derive a
confidence interval which communicates the precision (or uncertainty) of the summary estimate,
Methods
58
and to derive a p-value which communicates the strength of the evidence against the null hy-
pothesis of no intervention effect. The graphical approach for displaying effect estimates in
RevMan Analyses is the forest plot: forest plots for continuous outcomes illustrate (Deeks et al.
2006):
• The raw data (means, standard deviations and sample sizes) for each arm in each study
• Point estimates and confidence intervals for the chosen effect measure, both as blocks and
lines and as text
• A meta-analysis using the chosen effect measure (unstandardised mean difference or standard-
ised mean difference) and chosen method (fixed or random effects), both as a diamond and as
text
• The total numbers of participants in the experimental and control groups
• Heterogeneity statistics (the chi-squared test and the I2 statistic)
• A test for overall effect (overall average effect for random effects meta-analyses)
• Percent weights given to each study.
As suggested by Deeks (Deeks et al. 2006), presentation of data as a forest plot (displaying a
summary effect of meta-analysis , the grand mean) was skipped when only a single study was found
for a particular outcome. Instead, results of single studies were presented in an additional table.
Trials that contributed to meta-analyses with two exercise groups were incorporated into the
meta-analysis with both exercise groups; a control group was allocated to each of them, with a
number of participants according to the proportion of participants in the exercise groups.
3.2.7.3 Analysing heterogeneity
The homogeneity of the effect size distribution was analysed to assess the adequacy of the mean
effect size for representing the entire distribution, because a mean effect size would not be repre-
sentative for a distribution with a large variance. In a homogenous distribution the difference of
an individual effect size from the population mean arises only by sampling error, whereas in a
heterogeneous distribution the variability of the effect sizes is larger than would be expected from
sampling error and so each effect size does not estimate a common population mean. Homoge-
neity testing is based on a comparison of the observed variability in effect size values with an
estimate of the variance that would be expected from sampling error alone.
The classical measure of heterogeneity is Cochran’s Q, which examines the null hypothesis that
all studies are evaluating the same effect. Cochran’s Q is calculated as the weighted sum of
squared differences between individual study effects and the pooled effect across studies, with
Methods
59
the weights being those used in the meta-analysis. Q is distributed as a chi-square statistic with k
(number of studies) minus 1 degrees of freedom (df). Q has low power as a comprehensive test
of heterogeneity with small numbers of effect sizes, especially if they are based on small subject
samples. While a statistically significant result may indicate a problem with heterogeneity, a non-
significant result must not be taken as evidence of no heterogeneity.
Higgins (Higgins et al. 2003) developed an alternative approach that quantifies the effect of het-
erogeneity, providing a measure of the degree of inconsistency in studies’ results. The quantity I2
describes the percentage of total variation across studies, which is due to heterogeneity rather
than chance (sampling error). The statistic I2 is calculated as
I² Q
dfQ −
=x 100%
Where Q is the chi-squared statistic and df is its degree of freedom (Higgins and Thompson
2002, Higgins et al. 2003). I² is an intuitive and simple expression of the inconsistency of studies’
results. Following Higgins (Higgins et al. 2003), I2 values of 25% were considered to indicate low
heterogeneity, I2 values of 50% to indicate moderate heterogeneity and I2 values of 75% to indi-
cate large heterogeneity.
Potential effect modifications due to different populations or exercise intervention characteristics
(such as setting or training stimulus) were investigated and discussed descriptively but were not
analysed quantitatively using subgroup-analysis or meta-regression since the number of studies
was too low. Deeks et al. (2006) advise that for undertaking simple regression analyses at least ten
observations (i.e., ten studies in a meta-analysis) should be available for each characteristic mod-
elled, since it is very unlikely that an investigation of heterogeneity would produce useful findings
unless the number of studies is substantial. Therefore, possible sources of heterogeneity were
investigated and discussed only descriptively. As potential explanations for heterogeneity, the exer-
cise dose, adherence, the use of a health promotion theory and the exercise history of participants
were examined. Furthermore, sensitivity analyses were employed to investigate heterogeneity,
where heterogeneity due to the presence of one outlying trial with results that conflict with the
rest of the trials was examined by performing analyses both with and without outlying trials as
part of a sensitivity analysis. However, as Deeks et al. (2006) point out, explorations of heteroge-
neity that are devised after heterogeneity is identified can at best lead to the generation of hy-
potheses and furthermore, investigations of heterogeneity are of questionable value when there
are very few studies.
Reporting biases make positive studies easier to find than those with non-significant results;
subsequent over-representation of positive studies in a systematic review may mean that the
Methods
60
respective review is biased toward a positive result. Efforts were made to minimise this by exten-
sive searching. Estimating how big this problem is in this review was not feasible: the method
available in RevMan for estimating the problem of publication bias is the funnel plot, which were
not performed because the power of this method is limited when it comprises only a few small
studies, that is, in those situations when bias is most likely to distort the results of the meta-
analyses (Deeks et al. 2006). Furthermore, true heterogeneity in treatment effects may also lead to
funnel plot asymmetry and so the option is not available in RevMan Analyses when the random
effects model is chosen, which incorporates heterogeneity among trials.
3.3 Methods for the assessment of harm
In order to answer the research question, whether exercise should be offered to women receiving
adjuvant treatment of breast cancer, benefits of exercise have to be outweighed with potential
harm and thus, the third objective of the review was to answer the research question “what harm
is associated with exercise during adjuvant treatment of breast cancer”. This means that this
review aimed to evaluate both the benefit and harm and subsequently required a more complex
design to handle different sets of studies for various outcomes. In fact, using different search
strategies and eligibility criteria for studies of harm, compared to studies of benefit, generated a
second group of eligible studies. However, compared to the methods for conducting systematic
reviews of efficacy, which are well-established, guidance on how systematic reviews of harm
should be performed is limited (Golder et al. 2006).
The supplementary search became necessary because there was insufficient evidence regarding
the potential harm of exercise interventions during cancer treatment based on information from
the studies of benefit. The need for a supplementary search was further underpinned by theoreti-
cal reasoning: first, in a (randomised) controlled trial designed to assess aspects of efficacy or
effectiveness, this study design may not be able to identify all possible harms caused by exercise
because in RCTs, only what was looked for will be seen (Busse et al. 2002). Also, the reporting of
RCTs in regard to quality and quantity of harmful effects is currently largely inadequate (Ioanni-
dis et al. 2004).
The following inclusion criteria were applied for reviewing harmful effects of exercise: study
designs eligible for assessing potential harm were controlled trials (randomised and non-
randomised). This procedure facilitated identification of data on a well-recognised and easily
detectable harmful effect. In contrast, it would not allow the identification of information on
harmful effects that were new, rare, or long-term. However, this procedure seemed adequate
since potential harmful effects of exercise in the context of breast cancer are well-known (e.g.,
injuries). The study of harm had to be a key trial objective, which means that the harm had to be
Methods
61
one of the major primary or secondary outcomes of trials to be included. This approach ensured
that surveillance of harm was active, with structured questionnaires or interviews or predefined
laboratory or other diagnostic tests. A preliminary literature search suggested that there would be
quite few publications reporting harmful effects of moderate exercise training during adjuvant
cancer treatment and thus, all studies were eligible that investigated harmful effects in the reha-
bilitation of breast cancer, either during or after adjuvant cancer treatment. The rationale behind
this is the following: if harm appears to be substantial in breast cancer patients who have com-
pleted cancer treatment, then harm is probably to be expected during adjuvant treatment as well.
Evidence of no harm however, does not indicate that there is no harm for breast cancer patients
under adjuvant treatment as well. For an overview on inclusion criteria see Table 5.
Table 5: Inclusion criteria for reviewing harm associated with exercise
Inclusion criteria Exclusion criteria
Participants - Women with breast cancer
- Breast cancer stages I-III
- Cancer other than breast
cancer
Intervention - Aerobic exercise
- Resistance exercise
- Mixed exercise (aerobic and resistance)
- Stretching or flexibility
training
Outcome - Harm - Other outcomes only
Study design - Randomised controlled trials
- Non-randomised controlled trials
- Cohort
- Case-control
- Cross-sectional
- Case series
MEDLINE search was conducted with WinSPIRS 5.0 in October 2006 (WinSPIRS is the local
access, Windows-based version of – SPIRS – the SilverPlatter Information Retrieval System);
October 2006 is the cut-off point for inclusion of trials. The search strategy employed two ap-
proaches: first, text words for harms that could already be specified were used, such as lymphe-
dema or injury. This approach is useful for harms under consideration that are already known.
The second approach was to use synonyms of harms and related terms (text words). Additionally,
references of identified trials were checked for identifying further reports. The search strategy is
presented in detail in the Appendix 6. Selection of trials was performed, through title and abstract
checking based on the inclusion criteria. For data extraction in selected and finally included stud-
ies, a coding form was designed (see Appendix 6), which considered information on the follow-
ing facets of studies: type of harm assessed, time period of observation, type of exercise, type of
study, frequencies or proportions of harms, and statistical tests.
Methods
62
Quality assessment was based on an 8-criteria quality-rating instrument for assessing harm (Chou
and Helfand 2005). The following factors were considered: selection of participants, description
of population, loss to follow-up, pre-specification and definition of harms, description of ascer-
tainment technique, ascertainment of adverse events, statistical analysis of potential confounders,
and duration of follow-up. The quality assessment form for studies of harm can be found in
Appendix 6. Each quality criterion was coded with “1” if the study did meet this criterion; other-
wise it was coded with “0”. The total quality score was calculated by summing up the criteria. The
maximum total quality score was eight on a scale from zero to eight: following Chou and Helfand
(2005) a quality score of more than six was rated as good, a quality score from four to six as fair
and a quality score of less than four as poor quality. Information from primary studies examining
harm was synthesised descriptively.
Results: Identification and description of studies
63
4 Results: Identification and description of studies
Beneficial effects and harmful effects of exercise interventions were reviewed with similar rigour
in this systematic review. Therefore, two independent sets of studies were included in this review
and are described separately in the following paragraphs. Each included study was assigned one
identification label (study-ID) – composed of the name of the first author and the year of publi-
cation. This study-ID is used for reporting purposes.
4.1 Studies of benefit
Fifteen studies (as of September 2008) could be identified within the international literature
which explored the effects of aerobic or resistance training, or both on physical and mental
health outcomes in women during adjuvant treatment of early breast cancer.
The basic search in 2004 resulted in eight studies for inclusion in the review. The process of
study selection, starting from initial hits, through potentially relevant studies, and finally to in-
cluded studies, is illustrated in Figure 6. Through comprehensive searches of bibliographic data-
bases a total of 1688 citations were identified; contact with experts and scanning bibliographies of
reviews revealed another 202 studies that deemed potentially relevant. Subsequently, 1890 titles
and abstracts were screened. The majority of these citations (n=1858) were excluded because
they were not about breast cancer-related health promotion through exercise. These citations
included duplicates arising from different databases. Thirty-two study reports were identified as
being potentially relevant for inclusion in the review. Full reports were obtained and processed.
After screening of the full reports, 23 were excluded from the review. At this stage, the single
most important reason for excluding full reports was because the exercise intervention was not
concurrent with adjuvant cancer treatment (n=6). One dissertation thesis (Battaglini 2004) that
was identified in the basic search had to be excluded temporarily because the author asked for
exclusion until publication of his thesis. A total of eight studies were available for inclusion in the
review based on the basic search in 2004.
Regular quarterly search updates (overall 17) as well as publication of already identified studies
yielded seven more studies to be included in this review. One additional study was identified for
potential inclusion based on its abstract only (So et al. 2006); however this study is still awaiting
assessment due to language barriers (it is written in Korean) and so is not yet listed in the in-
cluded studies.
Results: Identification and description of studies
64
Searches of
electronic
databases
N=1688
Titles and abstracts
screened
N=1890
Potentially relevant
N=32
Full document screened
N=32
Citations
excluded
N=1858
Reports excluded
N=23
9 reports of 8 relevant
studies included in systematic
review
Basic search: Studies
included in meta-analyses
N=5
Criterion 2: N=4
Criterion 3: N=3
Criterion 4: N=6
Criterion 6: N=1
Criterion 7: N=5
Key for exclusion criteria
(1) Did NOT compare two
groups
(2) NOT about breast cancer
patients
(3) NOT about exercise
interventions
(4) Exercise intervention and
adjuvant cancer treatment not
concurrent
(5) Exercise programme less
than six weeks
(6) Did NOT report on health-
related outcomes
(7) Did NOT report a clinical
trial
(8) Author’s request (publication
status)
Review update: Studies
included in meta-analyses
N=12
Search updates in
MEDLINE
N=4
Publication of
identified studies
N=3
Expert contacts;
bibliographies in
reviews; N=202
Criterion 8: N=1
Criterion 1: N=3
Figure 6: Flow diagram of study selection process
Results: Identification and description of studies
65
The following 15 studies were included into the review:
Table 6: Reference list of included studies
Study-ID Reference
1. Battaglini
2007
Battaglini C, Bottaro M, Dennehy C, Barfoot D, Shields E, Kirk D, Hackney
A. The effects of resistance training on muscular strength and fatigue levels in
breast cancer patients. Rev Bras Med Esporte 12: 139e-144e; 2006.
Battaglini C, Bottaro M, Dennehy C, Rae L, Shields E, Kirk D, Hackney A.
The effects of an individualized exercise intervention on body composition in
breast cancer patients undergoing treatment. Sao Paulo Med J 125: 22-28;
2007.
2. Campbell
2005
Campbell A, Mutrie N, White F, McGuire F, Kearney N. A pilot study of a
supervised group exercise programme as a rehabilitation treatment for women
with breast cancer receiving adjuvant treatment. Eur J Oncol Nurs 9: 56-63;
2005.
3. Crowley
2003
Crowley, SA. The effect of a structured exercise program on fatigue, strength,
endurance, physical self-efficacy, and functional wellness in women with early
stage breast cancer. PhD [dissertation]. Ann Arbor: University of Michigan;
2003.
4. Courneya
2007
Courneya KS, Segal RJ, Gelmon K, Reid RD, Mackey JR, Friedenreich CM,
Proulx C, Lane K, Ladha AB, Vallance JK, Liu Q, Yasui Y, McKenzie DC.
Six-month follow-up of patient-rated outcomes in a randomized controlled
trial of exercise training during breast cancer chemotherapy. Cancer Epide-
miol Biomarkers Prev 16: 2572-8; 2007.
Courneya KS, Segal RJ, Mackey JR, Gelmon K, Reid RD, Friedenreich CM,
Ladha AB, Proulx C, Vallance JK, Lane K, Yasui Y, McKenzie DC. Effects of
aerobic and resistance exercise in breast cancer patients receiving adjuvant
chemotherapy: a multicenter randomized controlled trial. J Clin Oncol 25:
4396-404; 2007.
5. Drouin 2002 Drouin J. Aerobic exercise training effects on physical function, fatigue and
mood, immune status, and oxidative stress in subjects undergoing radiation
treatment for breast cancer. PhD [dissertation]. Detroit: Wayne State Univer-
sity; 2002.
6. Kim 2006 Kim CJ, Kang DH, Smith BA, Landers KA. Cardiopulmonary responses and
adherence to exercise in women newly diagnosed with breast cancer undergo-
ing adjuvant therapy. Cancer Nurs 29: 156-65; 2006.
7. MacVicar
1986
MacVicar M, Winningham M. Response of cancer patients on chemotherapy
to a supervised exercise program. Cancer Bull 13: 265-74; 1986.
8. MacVicar
1989
MacVicar MG, Winningham ML, Nickel JL. Effects of aerobic interval train-
ing on cancer patients’ functional capacity. Nurs Res 38: 348-51; 1989.
Winningham ML, MacVicar MG, Bondoc M, Anderson JI, Minton JP. Effect
of aerobic exercise on body weight and composition in patients with breast
cancer on adjuvant chemotherapy. Oncol Nurs Forum 16: 683-9; 1989.
9. Mock 1997 Mock V, Dow KH, Meares CJ, Grimm PM, Dienemann JA, Haisfield Wolfe
ME, Quitasol W, Mitchell S, Chakravarthy A, Gage I. Effects of exercise on
fatigue, physical functioning, and emotional distress during radiation therapy
for breast cancer. Oncol Nurs Forum 24: 991-1000; 1997.
10. Mock 2005 Mock V, Frangakis C, Davidson NE, Ropka ME, Pickett M, Poniatowski B,
Stewart KJ, Cameron L, Zawacki K, Podewils LJ, Cohen G, McCorkle R.
Exercise manages fatigue during breast cancer treatment: a randomized con-
trolled trial. Psychooncology 14: 464-77; 2005.
Results: Identification and description of studies
66
11. Mutrie 2007 Mutrie N, Campbell AM, Whyte F, McConnachie A, Emslie C, Lee L, Kear-
ney N, Walker A, Ritchie D. Benefits of supervised group exercise programme
for women being treated for early stage breast cancer: pragmatic randomised
controlled trial [online]. BMJ. 334: 517; 2007. Available from:
http://www.bmj.com/cgi/reprint/334/7592/517
12. Payne 2008 Payne JK, Held J, Thorpe J, Shaw H. Effect of exercise on biomarkers, fa-
tigue, sleep disturbances, and depressive symptoms in older women with
breast cancer receiving hormonal therapy. Oncol Nurs Forum 35: 635-42;
2008.
13. Schwartz
2007
Schwartz AL, Winters-Stone K, Gallucci B. Exercise effects on bone mineral
density in women with breast cancer receiving adjuvant chemotherapy. Oncol
Nurs Forum 34(3): 627-33, 2007.
14. Segal 2001
Segal R, Evans W, Johnson D, Smith J, Colletta S, Gayton J, Woodard S,
Wells G, Reid R. Structured exercise improves physical functioning in women
with stages I and II breast cancer: results of a randomized controlled trial. J
Clin Oncol 19: 657-65; 2001.
15. Winningham
1988
Winningham ML, MacVicar MG. The effect of aerobic exercise on patient
reports of nausea. Oncol Nurs Forum 15: 447-50; 1988.
4.1.1 Process of retrieval of included trials in basic search
MEDLINE played a major role in study identification for the review. This is demonstrated
through tabulation of initial hits, potentially relevant and finally included studies per database
(Table 7). Likewise, from the perspective of the included study, the origin of retrieval shows the
importance of MEDLINE for study identification (Table 8). Besides MEDLINE (either
searched via WinSPIRS or via PubMed in the internet), CINAHL and the CBCSR were the data-
bases which yielded most studies for final inclusion in the review. Exercise interventions may be
regarded as nursing interventions, which is why the CINAHL database with its nursing focus
yielded best results. Three studies were identified via MEDLINE (MacVicar 1989, Mock 1997,
Segal 2001) and four via CINAHL (Winningham 1988, MacVicar 1989, Mock 1997, Drouin
2002). Two studies were identified exclusively via CINAHL (Winningham 1988, Drouin 2002).
ProQuest Digital Dissertations was another important source for identifying relevant studies
performed as dissertation research. Two studies were identified this way (Crowley 2003, Battag-
lini 2004). One problem that emerged with dissertations was that data from dissertations pending
publication could not be included in the review. One trial could not be identified via searches in
electronic databases but only via reference checking (MacVicar 1986). Via EMBASE, only one
trial (Segal 2001) was identified that was also identified by MEDLINE. PsycInfo, SIGLE,
CONFSCI and DissOnline.de yielded no studies to include in the review and thus appeared to be
of minor importance for the question of the review.
These results substantiate the pragmatic approach to limit search updates to MEDLINE and the
Cochrane Breast Cancer Specialised Register: apart from CINAHL which could not be accessed
when updating searches, MEDLINE yielded best results. Three dissertation theses and the first
Results: Identification and description of studies
67
study at all in this topic were not identified via MEDLINE: however, these dissertations did not
contribute to meta-analyses in a noteworthy degree due to a small number of participants and
fragmentary data availability.
Table 7: Identification of included studies per database in the basic search
Electronic database
Initial hits
No.
included at
first stage
Unobtainable
or not
received
No. of trials
included at
second stage
MEDLINE 596 10 0 3
PubMed (web based) 89 9 0 3
EMBASE 559 3 0 1
CINAHL 79 10 0 4
PsycInfo 89 4 0 0
SPORTDiscus 250 4 0 1
Cochrane Breast Cancer
Specialised Register (CBCSR) 14 8 0 3
SIGLE 1 0 0 0
CONFSCI 5 2 0 0
ProQuest Digital Dissertations 6 2 1 2
DissOnline.de 183 0 0 0
Table 8: Origin of reports of included trials in the basic search
MEDLINE
Pubmed
EMBASE
CINAHL
PsyvInfo
SPORTDiscus
CBCSR
SIGLE
CONFSCI
ProQuest
DissOnline.de
Notes
Battaglini 2007 9 Thesis
Crowley 2003 9 Thesis
Drouin 2002 9 Thesis
MacVicar 1986
MacVicar 1989 9 9 9 9
Mock 1997 9 9 9 9
Segal 2001 9 9 9 9 9
Winningham 1988 9
4.1.2 Process of retrieval of ongoing trials
Ongoing trials were successfully identified through expert contacts and through clinical trials
registers. Table 9 presents how ongoing studies were retrieved for inclusion in this review. In
sum, this tabulation indicates that ongoing trials were primarily identified through contact with
experts in the field. Clinical trial registers of unpublished and ongoing clinical trial research did
not yield as many results as had been expected facing the fact that registration facilitates the dis-
semination of information among clinicians, researchers, and patients. However, when this re-
Results: Identification and description of studies
68
view started in 2004, study registration was the exception; only now is it the rule (Laine et al.
2007) and updating searches for ongoing studies in trial registers may yield more results. Tracking
of ongoing studies resulted in three studies which had been published during the course of the
review and thus could be included in the actual review (Campbell 2005, Courneya 2007, Mutrie
2007). One more study has been wound up in the meantime (Kleine-Tebbe 2006), however data
has not been provided despite many queries. Hence the only German study that fulfilled the
inclusion criteria for this review could not be included.
Table 9: Origin of ongoing trials
Experts Conferences Trial registers Notes
Campbell 2005 9 Published
Chetiyawardana 2004 9
Courneya 2005 9 Published
Kleine-Tebbe 2006 9
Mutrie 2005 9 Published
4.1.3 Characteristics of excluded studies
27 studies were excluded. Reasons for exclusion fell into several categories. The exercise inter-
vention was not concurrent with adjuvant cancer treatment in seven trials; exercise was either
part of a complex intervention, or exercise intervention period was too short (less than six
weeks), or no exercise intervention at all was implemented in five trials; in five trials the partici-
pants were not predominantly breast cancer patients; four trials could not be characterised as
controlled trials (they were study protocols or reviews). Four more trials did not compare two
groups assigned by the investigator (they compared high exercisers with low exercisers) and one
trial did not assess health-related outcome measures. One study (Battaglini 2004) with publication
pending was excluded only temporarily. For a detailed description of the reasons for exclusion
see Table 10. All unfulfilled exclusion criteria are listed here; however only the first reason ac-
cording to the inclusion algorithm was counted as reason for exclusion. This table does not only
contain clinical studies but also review articles, etc, which were part of the full text retrieval in
order to assure decision of exclusion when abstracts were ambiguous. A reference list of excluded
studies is presented in Appendix 7.
Table 10: Characteristics of excluded studies
Study-ID Reason for exclusion
1. Aghili 2007 Participants not predominantly breast cancer patients
2. Battaglini 2004 Transient exclusion due to publication pending (inclusion after publi-
cation of dissertation thesis in 2007)
Results: Identification and description of studies
69
Study-ID Reason for exclusion
3. Burnham 2002 Exercise not concurrent with adjuvant cancer treatment
4. Courneya 2003a Participants not predominantly breast cancer patients,
exercise as part of a complex intervention
5. Courneya 2003b Exercise not concurrent with adjuvant cancer treatment
6. Daley 2007 Exercise not concurrent with adjuvant cancer treatment
7. Demark-
Wahnefried 2002
No clinical trial, protocol status,
exercise as part of a complex intervention
8. Demark-
Wahnefried 2003 No clinical trial, design paper
9. Dimeo 1999 Participants not predominantly breast cancer patients,
no chemotherapy, radiotherapy or hormonal treatment
10. Fairey 2003 Exercise not concurrent with adjuvant cancer treatment
11. Given 2002 Participants not predominantly breast cancer patients,
no exercise intervention
12. Ho 1986 No exercise intervention
13. Hwang 2008 Duration of intervention program less than 6 weeks
14. Latikka 1997 No clinical trial, review
15. McKenzie 2003 Exercise not concurrent with adjuvant cancer treatment
16. Mock 1994 Exercise as part of a complex intervention
17. Mock 2001 Trial does not compare two groups as assigned by investigator
18. Mock 2002 No exercise intervention
19. Mustian 2002 No clinical trial, review
20. Pickett 2002 No health-related outcome measure (adherence study)
21. Pinto 2003 Exercise not concurrent with adjuvant cancer treatment
22. Schneider 2007 Trial does not compare two groups as assigned by investigator
23. Schwartz 1999 Trial does not compare two groups as assigned by investigator
24. Schwartz 2001 Trial does not compare two groups as assigned by investigator
25. Segar 1998 Exercise not concurrent with adjuvant cancer treatment
26. Shaw 2003 No clinical trial, protocol status,
exercise as part of a complex intervention
27. Wilkie 2003 Participants not predominantly breast cancer patients,
duration of intervention program less than 6 weeks
Results: Identification and description of studies
70
4.1.4 Characteristics of included studies
The final selection resulted in 15 studies being included in this review. Data extraction tables can
be found in Appendix 8. The included studies were predominantly performed in North America
(United Sates and Canada); only two studies (a pilot study and a subsequent RCT) were imple-
mented in Europe from one United Kingdom study group. One German study could not be
included into the review because results were not provided by the study group. Sample sizes
across studies ranged from 10 to 242; just recently, four large-scale trials emerged exceeding 100
participants, two of them even included more than 200 participants. For trial characteristics and
outcomes see Table 15.
4.1.4.1 Participants
As specified in the review’s inclusion criteria, all studies included women with breast cancer who
were undergoing adjuvant cancer treatment. Modes of adjuvant treatment were heterogeneous
and varied across the 15 exercise intervention studies. Some studies included women following
various treatment regimens and other studies focused either on chemotherapy, radiotherapy, or
hormonal therapy, and thus included exclusively women following the respective treatment regi-
men. Women under various treatments were included in six studies: one study (Segal 2001) in-
cluded patients on chemotherapy, radiotherapy or hormonal therapy; patients received chemo-
therapy or radiotherapy in five studies (Battaglini 2007, Campbell 2005, Kim 2006, Mock 2005,
Mutrie 2007). Nine studies included exclusively women who received the same mode of adjuvant
treatment: just chemotherapy in six studies (Courneya 2007, Crowley 2003, MacVicar 1986, Mac-
Vicar 1989, Schwartz 2007, Winningham 1988); radiotherapy in two studies (Drouin 2002, Mock
1997) and hormonal therapy in one study (Payne 2008).
Regarding the pre-intervention lifestyle, three trials explicitly included women who had a seden-
tary lifestyle before the intervention (Campbell 2005, Mock 2005, Mutrie 2007) and thus might
benefit more from a moderate training stimulus than exercisers do and who, on the other hand,
might exhibit greater difficulties with adhering to exercise; four trials (Courneya 2007, Crowley
2003, Schwartz 2007, Segal 2001) included sedentary women and exercisers. The remaining eight
trials did not specify whether participants were sedentary before intervention; several of these
studies used sedentary lifestyle as an inclusion criterion without presenting data for exercise his-
tory. Subsequently no information regarding sedentarism was available within these studies.
Women included in primary studies were relatively young: mean age was 51 years, with SD=5.1.
Only one study (Payne 2008) included exclusively women older than 65 years. Participants were
overweight with a mean body mass index of more than 25 in eight of nine trials that reported
relevant data (Battaglini 2007, Courneya 2007, Drouin 2002, Kim 2006, MacVicar 1989, Mock
Results: Identification and description of studies
71
2005, Mutrie 2007, Winningham 1988). Participants were normal weight with a mean body mass
index of 24 in only one study (Mock 1997). Study participants were predominantly diagnosed
with early breast cancer: five studies included women with breast cancer stages I and/or II
(Crowley 2003, MacVicar 1986, MacVicar 1989, Mock 1997, Segal 2001), and seven studies in-
cluded women with stages 0 to III (Campbell 2005, Courneya 2007, Drouin 2002, Kim 2006,
Mock 2005, Mutrie 2007, Schwartz 2007). Only one trial was open for women with metatstatic
breast cancer (Winningham 1988). This trial was included although metastatic breast cancer was
an exclusion criterion because the proportion of women with metastatic breast cancer was low in
the intervention group (13%) and even lower in the control group. No information on disease
stage was available for two studies (Battaglini 2007, Payne 2008). Further information on patient
characteristics was extracted for each study, e.g., menopausal status or exact type of chemother-
apy regimen. However reporting across studies was not consistent at all concerning these patient
characteristics: menopausal status, for example, was reported in five trials only.
4.1.4.2 Intervention
Exercise interventions across studies were heterogeneous in several aspects. Three studies com-
pared two exercise intervention groups with one control group; the two interventions varied
either in supervision of exercise training (Segal 2001) or in mode of exercise (Courneya 2007,
Schwartz 2007).
The exercise interventions were implemented in different settings ranging from community based
exercise classes and home-based exercise programs to cancer centre-based exercise programs:
interventions in two studies were community based (Campbell 2005, Mutrie 2007) with exercise
classes, that were run by specifically trained exercise specialists and took place in several loca-
tions. Four studies were centre based, e.g., in fitness centres of Universities or Cancer Centres
(Battaglini 2007, Courneya 2007, Kim 2006, Segal 2001), three studies were based in exercise
physiology laboratories (MacVicar 1986, MacVicar 1989, Winningham 1988) and a further seven
studies implemented a home based intervention (Crowley 2003, Drouin 2002, Mock 1997, Mock
2005, Payne 2008, Schwartz 2007, Segal 2001), where women exercised on their own following
their exercise prescriptions. Thus, in seven studies participants’ exercise was self-directed and not
supervised.
Also mode of exercise differed across studies: studies either tested aerobic exercise interventions,
resistance training programs, or applied a combined aerobic-resistance program. Two studies
tested two different exercise interventions: an aerobic exercise program and a resistance exercise
program (Courneya 2007, Schwartz 2007). Aerobic exercise interventions were implemented in
11 studies using walking, cycle ergometer training, treadmill, or elliptical. Three of these studies
Results: Identification and description of studies
72
employed interval training with a cycle ergometer (MacVicar 1986, MacVicar 1989, Winningham
1988); the mode of aerobic exercise in home-based interventions was predominantly walking.
Four studies (Battaglini 2007, Campbell 2005, Crowley 2003, Mutrie 2007) applied a combined
aerobic-resistance program which was implemented as a group exercise program in two of the
studies (Campbell 2005, Mutrie 2007). Two studies tested the effects of an isolated resistance
training program in a second intervention group (Courneya 2007, Schwartz 2007). Resistance
training across studies was either implemented with weight machines or elastic tubing or bands.
Besides one study (Drouin 2002) that used a stretching intervention as the comparison arm,
studies compared an exercise intervention with no intervention.
Exercise intervention periods in the included studies lasted from six weeks to six months. Exer-
cise interventions were between six and 12 weeks in eight studies (Campbell 2005, Drouin 2001,
Kim 2006, MacVicar 1986, MacVicar 1989, Mock 1997, Mutrie 2007, Winningham 1988) and
more than 12 weeks in the other seven studies (Battaglini 2007, Courneya 2007, Crowley 2003,
Mock 2005, Payne 2008, Schwartz 2007, Segal 2001). The longest intervention period of any trial
was 26 weeks (Segal 2001). Studies with shorter intervention periods (six to seven weeks) in-
volved breast cancer patients receiving exclusively radiotherapy, which is of shorter duration than
chemotherapy (Drouin 2002, Mock 1997). In one trial (Mock 2005), the exercise intervention was
implemented to span the period of time from initiation to cessation of the participant’s adjuvant
therapy and subsequently participants in the intervention arm of the study had intervention peri-
ods of variable length (either six weeks with radiotherapy or three to six months with chemother-
apy). In three trials participants exercised less than three times a week (Battaglini 2007, Campbell
2005, Mutrie 2007); in all other trials exercise frequency was three times or more per week.
Two studies only, both studies of the UK study group (Campbell 2005, Mutrie 2007), reported
that the intervention was led by a model of behaviour change in order to give women support to
change exercise behaviour. Here, the exercise program itself was amended by group discussions
after exercise classes. Themes addressed in these discussions were the health benefits from exer-
cise, barriers, goal setting, supportive environments and activity options in the community. Dis-
cussions were based on the “Stages of change” approach and were designed to promote inde-
pendent exercise after the intervention.
Information regarding exercise interventions is tabulated in Table 11. Note that some interven-
tion characteristics number up to more than 15 because of those studies with two intervention
groups.
Results: Identification and description of studies
73
Table 11: Description of interventions and study characteristics
Intervention or study characteristic No. of studies
Country United States
Canada
United Kingdom
Other
10
2
2
1
Year of publication 1986-1990
1991-2000
2001-2005
after 2005
3
1
5
6
Cancer diagnosis Breast cancer 15
Timing of intervention During treatment 15
Type of adjuvant
cancer treatment
Chemotherapy
Radiotherapy
Chemo- or radiotherapy
Chemo- or radio- or hormonal therapy
Hormonal therapy
6
2
5
1
1
Exercise history of
participants
Sedentary
Sedentary and exercisers
Not specified (or inclusion criterion without assessment)
3
4
8
Setting of intervention Laboratory based
Centre based
Home based
Community based
3
4
7
2
Supervision of exercise Supervised
Self-directed exercise
9
7
Theory of behaviour
change
Theory based
Not theory based
2
13
Exercise mode Aerobic exercise
Aerobic interval exercise
Resistance exercise
Mixed exercise
8
3
2
4
Exercise frequency Three times or more per week
Less than three times per week
12
3
Intervention length Six to 12 weeks – chemotherapy
Six to 12 weeks – radiotherapy
More than 12 weeks
6
2
7
Study design Randomised controlled trial
Non-randomised controlled trial
13
2
Two trial arms
Three trial arms
10
5
Two intervention groups
Two control groups
3
2
Nature of control
group
No intervention
Stretching exercise
14
1
Results: Identification and description of studies
74
4.1.4.3 Outcomes
Exercise interventions are expected to affect several health outcomes, both clinical outcomes and
patient-reported health outcomes. Accordingly, multiple outcomes were assessed across exercise
studies during adjuvant treatment of breast cancer. Every included study provided information
on immediate post-intervention effects. Three studies additionally provided information on sus-
tainability of effects with data on health and/or physical activity outcomes assessed four respec-
tively six months post-intervention (Kim 2006, Courneya 2007, Mutrie 2007).
Exercise interventions tailored to increase aerobic fitness, muscle strength, and other fitness
domains can be appropriately measured using physical fitness and physical functioning outcomes.
Apart from two studies (Payne 2008, Winningham 1988), all studies assessed physical fitness
which is an important determinant of physical functioning. Physical fitness was assessed through
either cardiorespiratory or muscular fitness or through body composition. Cardiorespiratory
fitness was assessed in all of the 13 studies that assessed physical fitness; six trials additionally
measured body composition (Battaglini 2007, Courneya 2007, Drouin 2002, MacVicar 1989,
Mutrie 2007, Segal 2001), and five trials measured muscular fitness/ strength (Battaglini 2007,
Courneya 2007, Crowley 2003, Drouin 2002, Schwartz 2007).
Assessment of cardiorespiratory fitness across the included studies was typically performed using
two major categories of submaximal exercise tests, i.e., predictive and performance tests. Sub-
maximal exercise testing is adequate in people whose performance may be limited because of
fatigue or other side-effects of cancer treatment and in cases where maximal exercise testing is
contraindicated due to cardiopulmonary, musculoskeletal, and neuromuscular impairments. The
predictive submaximal exercise tests applied were the modified Bruce treadmill test (Battaglini
2007, Drouin 2002, Kim 2006), the Cornell treadmill test (Crowley 2003), the modified Canadian
aerobic fitness test, a step test (Segal 2001), and one investigator developed a protocol for a bicy-
cle test on a stationary cycle ergometer (MacVicar 1986, MacVicar 1989, Winningham 1988).
These (modified) protocols typically involve modest stage-to-stage increments in energy require-
ments in order to assess cardiorespiratory fitness based on exercise time or peak work rate using
predictive equations for maximal oxygen uptake. Courneya 2007 employed a maximal incre-
mental exercise protocol on a treadmill which was not further specified. In five trials a perform-
ance test, the 12-minute walk test (12 MWT), was applied (Campbell 2005, Mock 1997, Mock
2005, Mutrie 2006, Schwartz 2007). Performance tests involve measuring the responses to stan-
dardised physical activities that are typically encountered in everyday life; in the 12-minute walk
test, the total distance covered in 12 minutes is reported.
Results: Identification and description of studies
75
Assessment of muscular strength was based on submaximal tests as well since maximal strength
testing may produce test-induced muscle soreness and muscular injury from muscle strain in
previously untrained individuals. In submaximal strength tests, a woman’s one repetition maxi-
mum (1-RM) lifting capacity is predicted by means of regression equations from the number of
repetitions of submaximal weight she could lift. However, only one trial (Battaglini 2007) pro-
vided information on how muscular strength assessments were performed, i.e., the submaximal
muscle endurance protocol that was used (Kuramoto and Payne 1995); one study (Courneya
2007) employed the eight repetition maximum, i.e., the maximum weight one can lift with eight
repetitions. However, again no information was provided concerning the muscle endurance pro-
tocol that was used.
Physical fitness measures are important indicators of physiological improvements gained through
regular exercise; but effects of improved physical fitness on physical functioning are of even
more interest. Physical functioning means the ability to carry out various activities that require
physical capability, ranging from basic mobility functions (e.g., climbing stairs) to participation in
life activities such as travelling. Physical functioning, the primary outcome of this systematic
review, was assessed in three trials only (Crowley 2003, Mock 2005, Segal 2001). In these three
trials, physical functioning was measured by perceived mobility only, using the physical function-
ing scale of the SF-36. Performance tests of objective mobility were not included. Twelve trials
thus failed to measure outcomes beyond physical fitness.
Cancer-related fatigue as a symptom of disturbed physical health was measured in ten trials
(Battaglini 2007, Campbell 2005, Courneya 2007, Crowley 2003, Drouin 2002, MacVicar 1986,
Mock 1997, Mock 2005, Mutrie 2007, Payne 2008) and three trials measured vitality as a related
construct (Crowley 2003, MacVicar 1986, Segal 2001). The multifactorial nature of fatigue
seemed to be generally acknowledged across studies since multidimensional assessment instru-
ments were employed. Studies predominantly assessed fatigue through either the revised Piper
fatigue scale (R-PFS) or the Piper Fatigue Scale (PFS) (Piper et al. 1986, Piper et al. 1989). The
original version contains 40 questions and uses a visual analogue scale. The revised PFS incorpo-
rates 22 items, which evaluate perception of current fatigue with an 11-point Likert scale. Two
studies employed the Functional Assessment of Cancer Therapy (FACT) instrument – either the
FACT-Fatigue (FACT-F) consisting of the FACT-General (FACT-G), a core questionnaire, plus
13 fatigue items (the fatigue subscale) or the fatigue subscale only. The fatigue subscale employs a
5-point Likert scale and contains 13 items that attempt to identify the intensity of fatigue experi-
enced during the seven days before questionnaire administration (Cella et al. 1993, Cella 1997,
Cella 1998). PFS and FACT fatigue subscale are similar in length and content but differ with
respect to evaluation period (i.e., one week versus current).
Results: Identification and description of studies
76
Two studies assessed sleep disturbances; sleep disturbances are important since alterations in the
amount or quality of sleep have been associated with impaired alertness and with impaired cogni-
tive and emotional function and learning. Furthermore, physical health was assessed through
bone health, immune conditions and hormonal regulation. A bone mineral density (BMD) test
was used to determine bone health (measured density in g/cm3) in one study (Schwartz 2007).
As biological markers for describing immune system conditions, T-cells, natural killer cells, and
oxidative stress were examined in one study (Drouin 2002). Cortisol, serotonin, interleukin-6,
and bilirubin were used as indicators for hormonal regulation in a further study (Payne 2008).
Besides physical health outcomes, mental health outcomes were assessed across included studies.
As mental health outcomes, emotional distress was measured in six studies (Courneya 2007,
Drouin 2002, MacVicar 1986, Mock 1997, Mutrie 2007, Payne 2008). Emotional distress refers to
unpleasant feelings or emotions and typically comprises aspects of anxiety, depression, and anger.
Positive psychological function was assessed in four trials: self-efficacy and self-esteem were
measured in one trial, respectively indicating mastery and control (Patient Reported Outcomes
Measurement Information System Network 2008). Additionally, positive affects and satisfaction
with life were assessed.
One study (Payne 2008) studied a symptom cluster and combined three commonly seen, concur-
rent symptoms in breast cancer patients (i.e., sleep disturbances, fatigue and depression) into one
symptom cluster. They employed a conceptual model which suggested that fatigue, sleep distur-
bances, and depressive symptoms may result from a dysregulation of hormones. The idea was
that aerobic exercise might effect positive changes in involved hormones with subsequent im-
provements in symptoms from the symptom cluster - supporting the approach of managing
these three symptoms simultaneously.
Health-related quality of life was assessed in five studies (Campbell 2005, Courneya 2007, Crow-
ley 2003, Mutrie 2007, Segal 2001). Across these studies, this outcome was assessed with generic
quality of life instruments, cancer-specific quality of life instruments, and breast-cancer-specific
instruments. These instruments differ in their ability to capture fine changes related to the diag-
nosis and treatment of breast cancer: the more specific instruments can be expected to better
capture breast-cancer-specific experiences compared to such generic instruments as SF-36.
Information on physical activity during the intervention period was presented in six studies
(Campbell 2005, Courneya 2007, Crowley 2003, Kim 2006, Mock 2005, Mutrie 2007). Physical
activity measures provide an idea of additional physical activity and exercise outside of the study
in supervised studies, and of adherence to exercise prescriptions in studies where participants are
required to organise their own exercise. Thus, the outcome of physical activity in exercise trials is
Results: Identification and description of studies
77
closely related to adherence. Results across included studies were formatted either in kcal per
week (Mock 2005), minutes of leisure time activity or exercise (Campbell 2005, Crowley 2003,
Mutrie 2007) and average weekly frequency of exercise, average duration of exercise per session,
and average duration of exercise within prescribed target heart rate range (Kim 2006) or ratio of
participants meeting exercise prescription in intensity and duration (Courneya 2007). Studies
employed training journals or devices such as heart rate monitors, pedometers, or accelerometers
for measuring physical exercise, but predominantly presented participation rates in scheduled
exercise sessions (e.g., Drouin 2002, Segal 2001). Only two studies analysed physical activity
regarding intensity of prescribed exercise (Courneya 2007, Kim 2006).
According to training principles, effects of exercise training are the result of adaptive processes
affecting working muscles and the cardiovascular system that occur during regular training.
Physiological adaptations and the subsequent overall increase in fitness levels take place when the
magnitude of the training stimulus is greater than normal (overload). Adaptations associated with
any training effect are lost within a short period of stopping training (de-training). Thus, long-
term exercise adherence is worthwhile and regular physical activity would be an important long-
term effect of exercise interventions. Three trials assessed physical activity after four months and
six months respectively following intervention (Kim 2006, Courneya 2007, Mutrie 2007).
Two trials (Crowley 2003, Campbell 2005) assessed constructs which are included in health be-
haviour models: The construct of perceived benefits (assessed in Campbell 2005) comprises
beliefs about the positive outcomes associated with the exercise behaviour during adjuvant breast
cancer treatment and is included in the Transtheoretical Model (Stages of change). Self-efficacy
(assessed in Crowley 2003) is a similar construct and embraces the belief that a person has the
ability to complete an action.
Only in one of the included studies was harm a major secondary outcome of the study (Courneya
2007). In this trial, arm volume measurements were performed in order to assess lymphedema.
Possible adverse consequences of exercise were reported in nine trials (Battaglini 2007, Campbell
2005, Courneya 2007, Crowley 2003, Drouin 2002, Mock 1997, Mock 2005, Schwartz 2007, Segal
2001). However, it was not clarified how harm-related information was collected – neither con-
cerning the mode of data collection, nor whether surveillance for harm was active or passive.
Poor reporting practice for harm-related data was common with vague statements such as “no
adverse reactions to taking part in the exercise intervention” (Campbell 2005). Prior evidence of
harm was not systematically integrated in the trials. For detailed information on outcome meas-
ures see Table 12.
Results: Identification and description of studies
78
Table 12: Outcomes reported
Outcome category
No. of trials
assessing
the outcome
Construct assessed
No. of trials
assessing the
construct
Physical health
Physical function 3 Physical function (perceived) 3
Physical fitness 13
Cardiorespiratory fitness
Strength
Body composition
13
5
6
Symptoms 11
Fatigue
Nausea
Endocrine symptoms
Other symptoms (Pain, skin change, diar-
rhoea, mouth sores, constipation)
10
1
1
1
Sleep/ wake functions 2 Sleep disturbances 2
Biological markers 3
T-cells, natural killer cells, oxidative stress
Cortisol, serotonin, interleukin-6, bilirubin
Bone mineral density
1
1
1
Mental health
Emotional distress 6
Anxiety
Depression
Mood disturbance
Negative affects
2
4
2
1
Positive psychological
function 4
Self-esteem
Self-efficacy
Positive affects
Satisfaction with life
1
1
1
1
Health-related qual-
ity of life 5
Generic quality of life:
Cancer-specific quality of life
Cancer-site-specific quality of life
2
4
3
Harm 1 Lymphedema 1
Physical activity 6 Physical activity during intervention
Physical activity post-intervention
6
3
4.1.4.4 Other study characteristics
Small sample size was common among included trials. Only four trials (Courneya 2007, Mock
2005, Mutrie 2007, Segal 2001) had more than 30 participants per group; sample size was based
on power calculations in only five trials (Courneya 2007, Mock 1997, Mock 2005, Mutrie 2007,
Segal 2001). The median sample size was 42 patients, interquartile range (IQR) 22 to 119. Two
trials were controlled clinical trials (CCT) without random allocation of women to intervention
and control group (MacVicar 1986, Mock 1997); the other trials were randomised controlled
trials.
Results: Identification and description of studies
79
4.1.4.5 Quality of studies
Methodological quality
On average, the quality of the studies – following the vanTulder quality thresholds described in
the methods section – was moderate with a mean quality score of 11.4 (range 7 to 15 points out
of 17 possible points). The methodological quality score was associated with year of publication.
The mean overall methodological score in the four studies published before 2000 was 8.8 (range
7 to 10 points), compared to a mean of 12.4 (range 10 to 15 points) in the studies published since
2000. Methodological scores attributed to each study are presented in Table 13. Two of these
studies (MacVicar 1986 and Mock 1997) were non-randomised controlled trials; but, results from
these studies did not appear to deviate from those of randomised controlled trials.
Studies identified during the basic search were critically appraised by two persons in order to
explore the reliability of the quality assessment process. The inter-rater reliability was assessed
through the intraclass correlation coefficient as described in the methods section. The intraclass
correlation coefficient of 0.82 (ICC 0.82, 95% CI – 0.03 to 0.97) for those (nine) studies that
were assessed by two quality assessors regarding their methodological quality indicated an almost
perfect agreement among the two quality assessors; all disagreements could be resolved by dis-
cussion. Based on this result, this process of quality assessment was deemed to be sufficiently
reliable and so subsequent quality assessments were performed by only one quality assessor.
Table 14 summarises how many studies there were that met the individual quality criteria. The
following methodological issues emerged as problematic: Firstly, lack of observer blinding was
common in included studies; only two studies (Crowley 2003, Mutrie 2007) attempted observer
blinding, however, the success of blinding activities was not evaluated in those two studies. Sec-
ondly, seven studies either did not use adequate methods or failed to describe how they concealed
the allocation, i.e., separated the process of randomisation from the recruitment of participants.
This is a problem because studies using inadequate allocation concealment are more likely to
report significant findings than those using adequate concealment. Moreover, adherence prob-
lems were of concern. In deciding if exercise is likely to work for an individual woman, it is nec-
essary to know the effect of the exercise intervention in women who are physically active as pre-
scribed, i.e., who adhered to the exercise intervention. However, participants in the included
studies did not adhere to the exercise protocol to the full extent, or withdrew from the exercise
intervention for various reasons, e.g., adverse effects of the breast cancer treatment. When exer-
cise during adjuvant treatment of breast cancer is effective but non-adherence is substantial, the
analysis following the intention-to-treat principle (i.e., including all eligible participants) underes-
timates the magnitude of the intervention effect that will occur in adherent participants. On the
Results: Identification and description of studies
80
other hand, applying the intention-to- treat principle provides an unbiased assessment of the
efficacy of the exercise intervention at the level of adherence observed in the study. This level of
adherence could be similar to that observed in the community, and the results could inform
community-based decisions about the effectiveness of the exercise intervention. Only five studies
applied the intention-to- treat principle (or another adequate statistical procedure) to assess the
effect of exercise interventions during adjuvant cancer treatment. For example, Mock 2005 ap-
plied an innovative, valid statistical analysis and estimated the complier average causal effect
(CACE) which maintains randomisation-based properties and addresses non-adherence to the
exercise intervention. The other studies predominantly restricted the analysis to women for
whom there was complete data on the outcomes involved in the analysis (available case analysis).
Estimates from such analysis can be biased, especially if the women who are included in the
analysis are systematically different from those who were excluded in terms of one or more key
outcomes. Finally, long-term follow ups were considered in the quality score. Long-term follow-
ups for physical activity and health outcomes are of relevance because potential adverse effects of
adjuvant cancer treatment such as fatigue and weight gain are long-term adverse effects. Long-
term follow-ups were performed in three recent, included studies that assessed health and/or
physical activity outcomes beyond the period of adjuvant therapy (Courneya 2007, Kim 2006,
Mutrie 2007).
Results: Identification and description of studies
81
Table 13: Quality criteria met by studies
Study-ID Quality Score Patient
selection Intervention Outcome measurement Statistics
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Battaglini 2007 15 9 9 9 9 9 9 9
9 9 9 9 9
9 9 9
Campbell 2005 13 9
9
9 9
9
9
9
9
9
9
9
9 9
Courneya 2007 14 9 9 9 9 9 9 9 9 9 9 9 9
9 9
Crowley 2003 13 9
9
9
9
9
9 9
9
9
9
9
9 9
Drouin 2002 12 9
9
9
9 9
9
9
9
9
9
9 9
Kim 2006 11 9
9
9 9
9
9
9
9
9
9
9
MacVicar 1986 7 9
9
9
9
9
9
9
MacVicar 1989 8 9
9 9
9
9
9
9
9
Mock 1997 10 9
9 9
9
9
9
9
9
9
9
Mock 2005 14 9
9
9
9 9
9
9
9
9
9
9
9 9
9
Mutrie 2007 13 9
9
9 9
9
9 9
9
9
9
9
9 9
Payne 2008 7 9 9 9 9 9 9
9
Schwartz 2007 11 9
9
9 9 9
9
9 9
9
9 9
Segal 2001 13 9
9
9
9 9
9
9
9
9
9
9 9
9
Winningham 1988 10 9
9 9
9
9
9
9
9
9 9
Results: Identification and description of studies
82
Table 14: Number of studies meeting individual quality criteria
Quality criterion No. of studies
meeting criterion
Patient selection
1. Were the eligibility criteria specified?
2. Was a method of randomisation performed?
3. Was the treatment allocation concealed?
4. Were groups similar at baseline?
- 14
- 9
- 8
- 12
Intervention
5. Were interventions explicitly described?
6. Were co-interventions either avoided or comparable?
7. Was adherence acceptable in all groups?
- 10
- 12
- 8
Outcome measurement
8. Was the outcome assessor blinded to the intervention?
9. Were outcome measures relevant?
10. Were adverse effects described?
11. Was the drop-out rate acceptable and were reasons described?
12. Was short-term follow-up measurement performed?
13. Was long-term follow-up measurement performed?
14. Was timing of outcome assessment comparable in both groups?
- 2
- 15
- 9
- 12
- 15
- 3
- 15
Statistics
15. Was the sample size described for each group at randomisation?
16. Did the analysis include intention-to-treat analysis?
17. Were point estimates and measures of variability presented for the
primary outcome measures?
- 13
- 5
- 9
Training stimulus
The prescribed training stimulus was adequate in all included studies: all studies provided an
exercise prescription that can be supposed to produce the threshold intensity to maintain and
achieve improvements in cardiorespiratory and/or muscular fitness in that target group of cancer
patients. However, it was difficult to assess whether the actual training stimulus was as prescribed
since data on intensity and duration of exercise sessions performed by individuals were not pro-
vided. In Appendix 8 intervention quality is tabulated for all included studies.
High quality training studies
Twelve studies that had methodological quality scores of nine points or greater and provided
adequate exercise training stimuli were classified as high quality training studies (Battaglini 2007,
Campbell 2005, Courneya 2007, Crowley 2003, Drouin 2002, Kim 2006, Mock 1997, Mock 2005,
Mutrie 2007, Schwartz 2007, Segal 2001, Winningham 1988). Three studies were not classified as
high quality training studies; two of these studies belonged to the pioneer work regarding exercise
during adjuvant cancer treatment and were conducted in the eighties (MacVicar 1986, MacVicar
1989), one more study was published in 2008 (Payne 2008) and was the first study to examine
breast cancer patients older than 65 years.
Results: Identification and description of studies
83
Table 15: Characteristics of included studies
Study-ID Features Participants Intervention Key endpoints
Battaglini 2007
(US)
20 participants;
2 groups;
Control: no intervention;
Dissertation
Breast cancer patients;
Adjuvant treatment: chemotherapy
or radiotherapy;
Mean age 57 yr (SD 20)
• Mixed aerobic and resistance exercise
• 2 x/wk for 15 wk at 40%-60% of predicted exercise capac-
ity; 60 min per session
• Supervised
• Strength
• Fatigue
• Body composition
Campbell 2005
(UK)
22 participants;
2 groups;
Control: no intervention;
Pilot-study
Breast cancer patients;
Adjuvant treatment: chemotherapy
or radiotherapy;
Mean age 47.5 yr (SD 8)
• Mixed aerobic and resistance exercise
• 2 x/wk for 12 wk at 60%-75% of maximum heart rate; 10-
20 min per session
• Supervised
• Community-based exercise classes
• Aerobic fitness
• Fatigue
• Quality of life
Courneya 2007
(Canada)
242 participants;
3 groups with 2 interven-
tion groups;
Control: no intervention;
Follow-up 6 mo post-
intervention
Breast cancer patients;
Adjuvant treatment: chemotherapy;
Mean age 49 yr (range 25-78)
• Intervention 1: aerobic exercise (cycle ergometer, treadmill,
elliptical), 3 x/wk for 17 wk at 60-80% of aerobic capacity;
progressive increase from 15 to 45 min per session; super-
vised in fitness centre
• Intervention 2: resistance exercise (weight machines); 3
x/wk for 17 wk, set with 9 exercises; per session 2 sets of 8-
12 repetitions per exercise
• Aerobic fitness
• Strength
• Body composition
• Fatigue
• Emotional distress
• Quality of life
• Lymphedema
Crowley 2003
(US)
22 participants;
2 groups;
Control: no intervention;
Dissertation
Breast cancer patients;
Adjuvant treatment: chemotherapy;
Age range 36–58 yr
• Mixed aerobic (walking) and resistance (tubing) exercise.
• 3-5 x/wk for 13 wk at 60% of maximum heart rate; for 20-
60 min per session
• 2-3 sessions resistance exercise
• Self-directed
• Home-based
• Aerobic fitness
• Strength
• Fatigue
• Quality of life
Drouin 2002
(US)
23 participants;
2 groups;
Control: stretching;
Dissertation
Breast cancer patients;
Adjuvant treatment: radiotherapy;
Mean age 50.5 yr (SD 8.2)
• Aerobic exercise (walking)
• 3-5 x/wk for 7 wk at 50%-70% of maximum heart rate; for
20-45 min per session
• Self-directed
• Home-based
• Aerobic fitness
• Strength
• Body composition
• Fatigue
• Mood
• Biomarker
Results: Identification and description of studies
84
Study-ID Features Participants Intervention Key endpoints
Kim 2006
(South Korea)
41 participants;
2 groups;
Control: no intervention;
Follow-up 16 wk post-
intervention
Breast cancer patients;
Adjuvant treatment: chemotherapy
or radiotherapy;
Mean age 50 yr (SD7.6)
• Aerobic exercise (cycling or walking)
• 3 x/wk for 8 wk at 60-70% of peak aerobic capacity; for 30
min per session
• Supervised
• Self-directed aerobic exercise post-intervention
• Aerobic fitness
• Physical activity
MacVicar and
Winningham
1986
(US)
10 participants;
2 groups;
Control: no intervention
Breast cancer patients;
Adjuvant treatment: chemotherapy
• Aerobic (interval training on a stationary cycle ergometer)
• 3 x/wk for 10 wk at 60%-85% of maximum heart rate; for
20-30 min per session
• Supervised
• Aerobic fitness
• Mood
MacVicar 1989;
Winningham
1989
(US)
45 participants;
3 groups with 2 control
groups (no intervention
and placebo exercise)
Breast cancer patients;
Adjuvant treatment: chemotherapy;
Mean age 45 yr
• Aerobic (interval training on a stationary cycle ergometer)
• 3 x/wk for 10 wk at 60%-85% of maximum heart rate; for
20-30 min per session
• Supervised
• Aerobic fitness
• Body composition
Mock 1997
(US)
46 participants;
2 groups;
Control: no intervention
Breast cancer patients;
Adjuvant treatment: radiotherapy;
Mean age 49 yr
• Aerobic exercise (walking)
• 4-5 x/wk for 6 wk at 60%-80% of maximum heart rate; for
20-30 min per session
• Self-directed
• Home-based
• Aerobic fitness
• Fatigue
• Symptom experience
• Emotional distress
Mock 2005
(US)
119 participants;
2 groups;
Control: no intervention
Breast cancer patients;
Adjuvant treatment: chemotherapy
or radiotherapy;
Mean age 51.5 yr (SD 9.3)
• Aerobic exercise (walking)
• 5-6 x/wk for 6 wk (women receiving radiotherapy) or for 3-
6 mo (women receiving chemotherapy) at 50%-70% of
maximum heart rate; progressive increase from 15 to 30 min
per session
• Self-directed
• Home-based
• Physical functioning
• Aerobic fitness
• Fatigue
• Physical activity
Mutrie 2007
(UK)
201 participants,
2 groups,
Control: no intervention;
Follow-up 6 mo post-
intervention
Breast cancer patients;
Adjuvant treatment: chemotherapy
or radiotherapy;
Mean age 51.6 yr (SD 9.5)
• Mixed aerobic and resistance exercise
• 2 x/wk for 12 wk at 60%-75% of maximum heart rate; for
20 min per session (total session 35-50 min)
• Supervised
• Community-based exercise classes
• Aerobic fitness
• Body composition
• Fatigue
• Endocrine symptoms
• Breast-cancer-specific
symptoms
• Emotional distress
• Quality of life
Results: Identification and description of studies
85
Study-ID Features Participants Intervention Key endpoints
Payne 2008
(US)
20 participants,
2 groups,
Control: no intervention;
Pilot study
Breast cancer patients;
Adjuvant treatment: hormonal
therapy,
Mean age 65 yr (range 56-78)
• Aerobic exercise (walking)
• 4 x/wk at moderate intensity for 14 weeks, for 20 min per
session
• Self-directed
• Home-based
• Fatigue
• Depressive symptoms
• Sleep disturbances
• Biomarkers
Schwartz 2007
(US)
72 participants;
3 groups with 2 interven-
tion groups;
Control: no intervention
Breast cancer patients;
Adjuvant treatment: chemotherapy;
Mean age 48 yr
• Intervention 1: aerobic exercise (participants’ preferences), 4
x/wk for 6 mo at moderate intensity for 15-30 min per ses-
sion; self-directed, home based
• Intervention 2: resistance exercise (Thera band); 6 mo, 2 sets
of 8 exercises (4 upper, 4 lower body); 2 sets of 8-10 repeti-
tions per session
• Aerobic fitness
• Strength
• Bone mass density
Segal 2001
(Canada)
123 participants;
3 groups with 2 interven-
tion groups;
Control: no intervention
Breast cancer patients;
Adjuvant treatment: chemotherapy,
radiotherapy or hormonal therapy;
Mean age 51 yr (SD 8.7)
• Aerobic exercise (walking)
• 5 x/wk for 26 wk at 50-60% of peak aerobic capacity; for
20-30 min per session (total session 35-45 min)
• Intervention 1: 3 x supervised, 2 x self-directed exercise
• Intervention 2: self-directed home-based exercise
• Physical functioning
• Aerobic fitness
• Body composition
• Quality of life
Winningham
1988
(US)
42 participants;
3 groups with 2 control
groups (no intervention
and placebo exercise)
Breast cancer patients;
Adjuvant treatment: chemotherapy;
Mean age 46.6 yr
• Aerobic (interval training on a stationary cycle ergometer)
• 3 x/wk for 10 wk at 60%-85% of maximum heart rate; for
20-30 min per session
• Supervised
• Nausea
Results: Identification and description of studies
86
4.2 Studies of harm
Through MEDLINE search, 84 citations were identified. All 84 titles and abstracts were screened
for inclusion in reviewing harms. Based on title and abstract screening, 76 of these citations were
excluded because they were not reports of studies that examined harmful effects of exercise
interventions in the context of breast cancer. Eight reports representing seven studies were iden-
tified as being potentially relevant for inclusion. Full reports of these seven studies were obtained
and screened. Following screening of the full reports, a further four studies were excluded from
the review and three studies were included in the review of harmful effects of exercise. A refer-
ence list of included and excluded studies is presented in Appendix 9.
The four excluded studies were all excluded for reasons of study design. One study was a series
of case reports (Harries 2000) and three studies (Cheema 2006, Johansson 2005, Turner 2004)
represented before-and-after studies which, in contrast to studies with an intervention group and
a comparison group, were conducted within a single group. Outcomes were measured before and
after the exercise intervention. All of these excluded studies dealt with lymphedema comparable
to the included studies. It can be expected, therefore, that no qualitative information got lost
through the application of inclusion criteria.
The three included studies all examined potential harms of exercise after adjuvant cancer treat-
ment had been completed. No single study could be identified that examined harmful effects of
exercise in the time when women were still undergoing adjuvant treatment of breast cancer ─
besides that of Courneya 2007, one of the included studies of benefit that systematically ad-
dressed lymphedema. Methodological quality was assessed as good in two of the included trials
(Schmitz 2005, Basen-Engquist 2006) and as fair in one study (McKenzie 2003). All three studies
pre-specified and defined those harmful effects that would be assessed: lymphedema and injuries
were the only harmful effects that were examined. All three studies used adequate instruments
and techniques to ascertain harmful events. Schmitz 2005 was the only study that actively moni-
tored injuries. A standardised survey instrument was used to ask whether participants had experi-
enced injuries during the intervention period. If they responded yes, they were further asked
about the type of injury (e.g., heel spur, sprained ankle, shin splints, knee injury, hip injury,
pulled/strained muscle, sprained wrist, back problems). Furthermore, participants were asked
whether they believed the injury was the result of participation in the study. Other questions on
the injury survey were concerned with assessment of the effect of injury on exercise sessions, and
an assessment of how the injury was on the day of the injury survey. Ascertainment techniques
for lymphedema were the following: measuring and calculating arm volume, measuring arm cir-
Results: Identification and description of studies
87
cumference, and a validated survey instrument employed by Schmitz 2005 that measured self-
report of lymphedema diagnosis, symptoms, and treatment over the last three months.
Studies differed regarding the participants and the interventions examined. McKenzie 2003 in-
cluded women who already had developed arm lymphedema and examined the effects of an
upper-body exercise program (resistance and aerobic training) on secondary lymphedema.
Schmitz 2005 evaluated the effects of a resistance training program, which included exercises for
the upper and the lower body. Basen-Engquist 2006, finally, examined a lifestyle approach to
increasing physical activity, i.e., no detailed exercise prescription was provided but cognitive and
behavioural skills were addressed, together with ways to incorporate moderate physical activity
into the daily routine. Study characteristics are summarised in Table 16.
Results: Identification and description of studies
88
Table 16: Characteristics of studies of harm
Study-ID Methods Participants Interventions Outcomes
Basen-
Engquist
2006
RCT, 2 groups, randomisa-
tion by minimisation (simi-
lar to stratification);
Methodological quality
score: 7 (score range: 0-8).
60 breast cancer patients, after completion of
all treatments for breast cancer, within 7 years
of breast cancer diagnosis
Intervention: Lifestyle physical
activity program (taught women to
incorporate short periods of moder-
ate activity into their daily routines);
6 months, 21 sessions
Control: Standard care
Lymphedema: measured with
arm-circumference measurements
McKenzie
2003
RCT, 2 groups;
Methodological quality
score: 5 (score range: 0-8).
14 breast cancer patients, after completion of
all treatments for breast cancer more than 6
months before baseline; with unilateral lym-
phedema
Intervention: Mixed aerobic and
resistance exercise; 3 days/week;
8 weeks
Resistance exercise: 6 exercises, 10
repetitions, 2-3 sets per exercise.
Aerobic exercise (arm cycle ergome-
ter): 20 minutes
Control: no intervention
Arm volume: calculated from
arm-circumference measurements
and measured by water displace-
ment
Schmitz
2005
RCT, 2 groups, partial
crossover design, blocked
randomisation; stratifica-
tion for age and body fat;
Methodological quality
score: 7 (score range: 0-8).
(1) 85 breast cancer patients, after completion
of all treatments for breast cancer (except
hormonal therapy) 4 to 36 months before
baseline; sedentary to moderately physically
active; no weight training history; stable body
weight over the past year.
(2) for lymphedema assessment: analysis of a
sub-group of 45 trial participants who had
axillary node dissection beyond sentinel node
biopsy – a possible risk factor for lymphedema
Intervention: Resistance exercise;
60 minutes; 2 days/week; 26 weeks
(13 weeks supervised in small
groups; 13 weeks self-directed);
9 types of exercise using variable
resistance machines and free
weights; 8-12 repetitions, 3 sets per
exercise.
Control: no intervention in the first
six month; delayed intervention
from month 7 to 12.
(1) Injuries: standardised injury
reporting based on an injury
survey instrument
(2) Lymphedema: measured with
arm-circumference measurement,
self-report of diagnosis, and self-
report of symptoms (using a
validated survey instrument)
Outcomes were assessed 6-month
post-intervention in the interven-
tion group.
Results: evidence synthesis
89
5 Results: Evidence synthesis
From tabulation of study characteristics, one comparison emerged for quantitative analysis, that
is: exercise versus no exercise. The pre-planned comparison of exercise versus other interven-
tions (e.g., psychosocial interventions) was not feasible due to the absence of respective primary
studies. Meta-analysis was regarded as appropriate for the comparison of exercise versus no exer-
cise because studies were homogenous enough to be sensibly combined which was a conse-
quence of the rather focused review question and inclusion criteria concerning participants, inter-
vention, and study design. Data are reported in terms of effect size (SMD) and 95% CI. The
SMD can serve as a measure of strength of evidence but has limited value as a clinically meaning-
ful measure of intervention effect. In the context of the meta-analysis, the interpretation of an
effect as small, medium, or large was based on an operational definition with conventional crite-
ria: an effect size of 0.2 to 0.3 was regarded as a “small” effect, 0.5 to 0.7 a “medium” effect and
0.8 to 1.0 as a “large” effect. Meta-analyses were conducted using final values; change scores were
not used in meta-analysis. Effect sizes for outcomes that could not be pooled through meta-
analysis are presented in Table 17 and Table 18. Information on health assessment instruments
that were used in included studies (e.g., score range, interpretation of score) can be found in
Appendix 10. Appendix 11 shows the health assessment instruments and questionnaires that
contributed to pooling in meta-analyses for different outcomes.
5.1 Physical health effects
5.1.1 Physical functioning
Self-reported physical functioning was measured in three studies. Meta-analysis for this outcome
could not be performed. In one study (Segal 2001), physical functioning improved after exercise
training in both the self-directed and supervised exercise groups, whereas in the control group,
the physical functioning scale decreased: group comparisons of changes in physical functioning
revealed a significant mean difference between the self-directed exercise group and the control
group in favour of the exercise intervention. The mean difference of changes in physical func-
tioning from baseline to post-intervention between the supervised exercise group and the control
group did not achieve significance, although it showed a trend in favour of the exercise interven-
tion. However, the authors of this study discussed whether baseline differences in physical func-
tioning (higher physical functioning in the control group) may have accounted for the magnitude
and direction of changes in physical functioning observed over the intervention period.
No differences in physical functioning between exercise and control could be observed by Crow-
ley 2003. Mock 2005 reported physical functioning for “high walkers” – participants who exer-
Results: evidence synthesis
90
cised on average at least 60 minutes per week in three or mores sessions – compared to “low
walkers” who exercised not at all or less than 60 minutes per week (increased in high walkers and
decreased in low walkers). No results from analyses following the intention-to-treat principle (i.e.,
analysing participants in the groups to which they were randomised, without regard to whether
they adhered to the allocated exercise intervention) were presented in this study. In summary,
there is limited evidence that exercise during adjuvant therapy for breast cancer improves physical
functioning. The effect size is presented in Table 17.
5.1.2 Health-related physical fitness
5.1.2.1 Cardiorespiratory fitness
Cardiorespiratory fitness is the health-related component of physical fitness that relates to the
supply of oxygen to muscles during sustained physical activity (United States Department of
Health and Human Services 1996). Data from eight studies could be pooled in this meta-analysis
yielding eleven comparisons (due to three studies with two intervention arms) with a total of 709
participants. The overall pooled effect estimate (SMD 0.54, 95% CI 0.32 to 0.77) indicated that
participants in the exercise intervention groups experienced significantly increased cardiorespira-
tory fitness relative to participants in the control groups (Figure 7). The effect size of 0.54 stands
for a medium effect of physical exercise training on cardiorespiratory fitness. Cardiorespiratory
fitness was measured in four different units across the studies included in the meta-analysis: first,
oxygen uptake data were presented both with dimension mlVO2/kg/min (Courneya 2007,
Drouin 2002, Schwartz 2007, Segal 2001) and with dimension ml VO2/min (Kim 2006); second,
field test data (12-minute walk test) were provided using both walking distances in feet (Mock
1997) and meters (Campbell 2005, Mutrie 2007).
Other studies with cardiorespiratory fitness as an outcome did not provide sufficient data for
meta-analysis in cardiorespiratory fitness. One controlled clinical trial (MacVicar 1986), a pilot
study, reported only means without standard deviations. In one other trial, data for cardiorespira-
tory fitness were only presented by level exercise, i.e., high exercisers were compared with low
exercisers instead of following group assignment in statistical analysis (Mock 2005). Battaglini
2007 used physical fitness data to explain changes in primary outcomes, i.e., fatigue and lean body
mass and provided only means without standard deviations for cardiorespiratory fitness data.
Data from the remaining two trials with assessments of cardiorespiratory fitness (Crowley 2003,
MacVicar 1989) could not be transformed for meta-analysis requirements.
In a nutshell, exercise was an effective intervention for improving cardiorespiratory fitness rela-
tive to usual care even during breast cancer treatment. Preserving cardiorespiratory fitness during
Results: evidence synthesis
91
breast cancer treatment is important, because vital aspects of physical functioning require cardio-
respiratory fitness.
5.1.2.2 Muscular fitness
Muscular endurance and strength, i.e., the ability of the muscle to continue to perform without
fatigue and to exert force (United States Department of Health and Human Services 1996) are
important determinants of physical functioning as well. Pooling the effects of four studies –
yielding six comparisons due to multiple intervention arms in two studies – with a total of 328
participants yielded a total effect size (SMD 0.42; 95% CI 0.06 to 0.78 ) that indicated a statisti-
cally significant medium increase in strength for participants in the exercise intervention groups
compared to a control group (Figure 8). Muscle strength is an important outcome for breast
cancer patients because of the assumable impact on health-related quality of life: strength is pos-
sibly associated with physical functioning; further, a sense of return to feeling in control of the
own bodies (i.e., strength) may translate into feeling greater efficacy in other areas of life and
health-related quality of life (Ohira et al. 2006).
There are some inconsistencies in results across studies with moderate statistical heterogeneity (I2
=50%). Heterogeneity was reduced when only comparisons with interventions based on resistance
exercise training were included in this meta-analysis of strength (I2 =0%). Standardised mean
difference was 0.67, with 95% confidence interval 0.34 to 1.01 (Figure 9). This effect size indi-
cates that resistance exercise training in particular resulted in improved muscular strength when
compared to non-exercising control groups.
Studies assessed changes in muscular strength through the 1-repetition maximum (kg) of one
exercise or of several exercises combined. Because various exercises were used across studies to
assess muscular strength (e.g., overhead press, bench press, leg extension, or grip strength), the
standardised mean difference was used as the effect-size statistic.
5.1.2.3 Body composition
Body composition outcomes – relating to the relative amounts of muscle, fat, bone and other
vital parts of the body (United States Department of Health and Human Services 1996) – could
be pooled from four trials, yielding five comparisons with a total of 414 participants (Figure 10).
Exercise was more effective than usual care on maintaining or decreasing relative body fat and
weight (SMD -0.29; 95% CI -0.55 to -0.03). This effect was small but statistically significant. Body
composition is a relevant outcome since an increase in body mass index was observed to be asso-
ciated with a poorer prognosis (Barnett et al. 2008). Across studies body composition was meas-
Results: evidence synthesis
92
ured as weight, BMI, lean body mass, fat mass and skinfold thickness. For this meta-analysis BMI
and body fat percentage were combined.
5.1.3 Treatment-related symptoms
5.1.3.1 Cancer-related fatigue
Cancer-related fatigue was evaluated in ten studies. Pooling the effects of seven studies yielding
nine comparisons due to multiple intervention arms in two studies (see Figure 11) showed that
exercise was more effective than normal care on reduction of fatigue (SMD -0.17; 95% CI -0.32
to -0.02). The effect size indicates a small but statistically significant effect. There was no hetero-
geneity across these studies. Heterogeneity was moderate, with I2 of 47%, when the study (Battag-
lini 2007) was included in the meta-analysis (in the course of sensitivity analyses); the resulting
effect size was slightly greater with a standardised mean difference, then based on 734 partici-
pants, of -0.23, with 95% confidence interval -0.45 to -0.01.
The following assessments instruments for fatigue were employed within these eight studies: first,
the revised Piper-Fatigue scale was used in four studies (Battaglini 2007, Campbell 2005, Drouin
2002, Mock 2005); second, the Functional Assessment of Cancer Therapy (FACT) instrument
was used in two studies – either the FACT-Fatigue (FACT-F) consisting of the FACT-General
(FACT-G), a core questionnaire, plus 13 fatigue items (the Fatigue Subscale) or the Fatigue Sub-
scale only (Mutrie 2007, Courneya 2007); furthermore a visual analogue scale for fatigue was used
(Mock 1997); and finally assessments of vitality were made using the vitality scale of the SF-36 in
one study (Segal 2001). For statistical pooling in meta-analysis, effect sizes from SF-36 and
FACT-F were reversed following the direction of scales (in the fatigue subscale of the FACT
instrument, higher scores represent less fatigue). This procedure of combining reversed vitality
scores with fatigue scores is supported through substantial evidence supporting the adequacy of
the vitality scale of the SF-36 as a valid measure of energy and fatigue (O’Connor 2004).
5.1.3.2 Nausea
Nausea was measured in one trial only (Winningham 1988); the data from this trial show that
moderate aerobic exercise may provide relief from the symptom of chemotherapy-associated
nausea in some patients. Women in the control and placebo group significantly more frequently
experienced worsening or lack of amelioration of nausea: the relative risk (RR) for worsening or
lack of amelioration of nausea was 0.57, 95% CI 0.34 to 0.94. Thus, there is limited evidence for
an effect of exercise programs on reduction of nausea based on data from one study.
Results: evidence synthesis
93
5.1.3.3 Sleep disturbances
Sleep disturbances were examined in two studies (Mock 1997, Payne 2008). Due to methodologi-
cal limitations in one study (Payne 2008), data were not pooled in meta-analysis. In this study,
groups were not similar at baseline; a post-intervention group comparison yielded differences
between exercise and control group in favour of the control group, although the Pittsburgh Sleep
Quality Index (PSQI) scores for the exercise group decreased significantly over time, indicating
improved sleep quality. Sleep disturbances were reduced in Mock 1997, which provided limited
evidence for an effect of exercise programs on reduction of sleep disturbances based on data
from one study. The effect size is presented in Table 17.
Proper management of the multiple symptoms resulting from cancer and its treatment is impor-
tant, as symptoms can significantly distress patients and interfere with day-to-day functioning.
Furthermore, such symptoms might delay treatment or lead to premature treatment termination
(Cleeland 2007, Gapstur 2007). If treatment-related symptoms become so severe that patients
abandon important (and sometimes potentially curative) therapies or if they cause treatment
delays, they may diminish the chance of long-term remission or cure, and thus can directly affect
survival (Cleeland 2007). Residual treatment-related symptoms can also limit vocational activities
and inhibit social interaction (Cleeland 2007).
5.1.4 Biological and physiological effects
5.1.4.1 Bone health
Bone health (bone mineral density) was the primary outcome in one trial (Schwartz 2007). At the
end of the intervention period, it could be observed that aerobic exercise preserved lumbar spine
bone mineral density significantly better compared to usual care. Bone health is a particularly
relevant outcome measure for women who have had breast cancer because they are at higher risk
for osteoporosis and subsequent osteoporotic fractures than other women (Adler 2007, Brown
and Guise 2009). Furthermore, as many women with breast cancer will be long-term survivors,
the importance of skeletal health should not be underestimated and thus actively promoted with
this population of women.
5.1.4.2 Immune system
The effect of exercise on cell-mediated immunity is of concern since cytotoxic therapies, includ-
ing radiotherapy and chemotherapy, are discussed to be immunosuppressive. Exercise is associ-
ated with physiological changes in the immune system as well: moderate exercise may boost the
immune system function, but intense exercise can have the opposite effect.
Results: evidence synthesis
94
There is no evidence for an effect of exercise programs on immune functioning during breast
cancer treatment based on data from one study (Drouin 2002). No statistically significant differ-
ences were observed during radiotherapy for biological markers describing immune system con-
ditions (T-cells, natural killer cells, and oxidative stress). Both the exercise and control group
demonstrated significant declines in T-cell counts; since the final T-cell counts were not signifi-
cantly different between the two groups, this study helped support the safety of performing
moderate intensity aerobic exercise alongside radiotherapy of breast cancer. Similarly, exercise
was not able to preserve or improve natural killer cytotoxic activity during radiotherapy of breast
cancer. However, since final natural killer cytotoxic activity values were not significantly different
between exercise and control group following the intervention, aerobic exercise did not appear to
impair the immune function. The same was true for oxidative stress. The effect sizes are pre-
sented in Table 17.
5.1.4.3 Hormonal regulation
Hormonal regulation was addressed in one study (Payne 2008). The ideas behind addressing
hormonal regulation was a conceptual model suggesting that fatigue, sleep disturbances, and
depressive symptoms may result from a dysregulation of hormones. The authors hypothesized
that exercise interventions may effect positive changes in regulatory hormones, which may sug-
gest a mechanism that contributes to fatigue, sleep disturbances, and depressive symptoms in
patients with breast cancer. As biomarkers for hormonal regulation cortisol, serotonin, inter-
leukin -6, and bilirubin were measured.
There is limited evidence for an effect of exercise programs on hormonal regulation during breast
cancer treatment based on data from one study (Payne 2008). A significant intervention effect
could be observed on serotonin levels: serotonin levels decreased slightly over the intervention
period in the exercise group, suggesting that exercise may exert a negative influence on the pro-
duction of serotonin. Cortisol levels did not demonstrate a significant association with the exer-
cise intervention; however a downward trend in the intervention group did exist. Interleukin-6
levels were not significantly different between groups. A weak (but not significant) intervention
effect was seen for bilirubin levels, which may be an important factor related to fatigue. Effect
sizes could not be calculated due to insufficient reporting of results in the study.
Results: evidence synthesis
95
5.2 Mental health effects
5.2.1 Emotional distress
5.2.1.1 Depression
There were three trials that examined group differences for depression. For assessment of de-
pression, a symptom assessment scale (Mock 1997), the Beck Depression Inventory (Mutrie
2007) and the Center for Epidemiological Studies–Depression Scale (Courneya 2007) were used
and combined in meta-analysis. Pooling the effects of these three trials, which had four compari-
sons due to there being two intervention arms in Courneya 2007, yielded a significant reduction
of cancer-related depression in exercise groups when compared to a control group based on a
total of 443 participants: SMD -0.24, 95% CI -0.43 to – 0.04 (see Figure 12). There is evidence
that exercise is effective in reducing depression during adjuvant cancer treatment.
5.2.1.2 Anxiety
There is limited evidence for positive effects of exercise on anxiety reduction during adjuvant
therapy for breast cancer (see Figure 13). The effect size from a meta-analysis with two studies
(yielding three comparisons) suggested a small, but not statistically significant, effect of exercise
in reducing anxiety: SMD -0.25, 95% CI -0.54 to 0.04.
5.2.1.3 Other outcomes
There is no evidence for the effect of exercise on mood disturbances and negative affects, i.e., the
extent to which women felt aversive mood states and general distress. Results from single studies
indicated no statistically significant difference between groups. The effect sizes are presented in
Table 17.
5.2.2 Positive psychological function
Positive affects, i.e., the extent to which women felt enthusiastic, active, and alert, were increased
in the exercise group of one study (Mutrie 2007) compared to participants in the control group.
Likewise, exercising participants experienced higher self-esteem than participants in the control
group in a further study (Courneya 2007). Self-esteem is an important outcome for breast cancer
patients during difficult treatments.
No statistically significant differences between groups were found for satisfaction with life
(Campbell 2005), physical self-efficacy and attention performance (Crowley 2003) or for satisfac-
tion with body (Mock 1997).
Results: evidence synthesis
96
Thus, there is limited evidence that exercise lifts positive affects and self-esteem during adjuvant
therapy for breast cancer, but currently no evidence that exercise enhances other aspects of posi-
tive psychological function such as the feeling of mastery and control (self-efficacy).
5.3 Quality of life effects
5.3.1 Cancer- and cancer-site-specific quality of life
Four studies (Campbell 2005, Courneya 2007, Mutrie 2007, Segal 2001) examined effects of exer-
cise on cancer- and cancer-site-specific quality of life. There is evidence suggesting that exercise
increases breast-cancer-specific quality of life. Meta-analysis of three studies comprising 416
participants (see Figure 14) found a statistically significant improved cancer-specific quality of life
of participants in the exercise intervention groups when compared to non-exercising control
groups: the standardised mean difference was 0.23, with 95% confidence interval 0.03 to 0.43.
Cancer-specific quality of life was assessed using the Functional Assessment of Cancer Therapy
(FACT) instrument – the FACT-Breast (FACT-B) consisting of the FACT-General (FACT-G)
and a breast cancer subscale and the FACT-Anemia consisting of the FACT-General (FACT-G)
and the anemia subscale.
There is no evidence that exercise is effective in increasing cancer-specific quality of life during
adjuvant therapy for breast cancer: three studies examined cancer-specific quality of life (Camp-
bell 2005, Mutrie 2007, Segal 2001); meta-analysis could not be performed due to data availability
and compatibility (change scores or final values of results). Mutrie 2007 and Segal 2001 found no
significant intervention effect for FACT-G.
5.3.2 Generic health-related quality of life
Generic health-related quality of life (assessed via SF-36) was examined by Segal 2001 and Crow-
ley 2003 and no statistically significant differences between groups were found. Thus there is no
evidence for the effectiveness of exercise programs for generic health-related quality of life dur-
ing breast cancer treatment. Quantitative synthesis was not performed because the data for sub-
scales were presented without summary measures.
5.4 Adherence and contamination
Adherence was employed as a descriptive variable in the context of this review: Measuring adher-
ence to exercise is important because a failure of participants to adhere to exercise prescriptions
may bias results obtained by research studies, limiting the strength of the empirical evidence
generated. In general, exercise adherence can be calculated by comparing actual exercise behav-
Results: evidence synthesis
97
iours with the standards determined in the exercise recommendation. Different approaches were
used among the included studies to measuring adherence, that is, the level of exercise participa-
tion achieved once the woman had agreed to undertake it. First, the ratio of attendance to sched-
uled exercise sessions was widely used as an index of adherence in trials with supervised and self-
directed exercise interventions. Attendance in self-directed, home-based interventions was meas-
ured via participant diaries and self-report questionnaires. The ratio of attendance was calculated
by dividing the number of sessions attended by the number of scheduled sessions. Second, ad-
herence was operationalised as a dichotomous variable and assessed as the proportion of adher-
ent participants with various cut-points of exercise per week (Mock 1997, Mock 2005).
Adherence to exercise was said to be 70% or more of possible exercise sessions (supervised or
self-directed) in seven studies (Battaglini 2007, Campbell 2005, Courneya 2007, Crowley 2003,
Drouin 2002, Kim 2006, Segal 2001). Adherence may be of a similar magnitude in one more
study (MacVicar 1989), since they let participants repeat missed sessions. In one study (Mutrie
2007), less than 40% (38.8%) of participants attended at least 70% of the possible exercise ses-
sions. The proportion of adherent participants was found to be 86% by Mock 1997 and 72% by
Mock 2005 based on cut-points of 90, respectively 60, minutes of exercise per week.
Attendance rates alone do not inform on participants’ adherence behaviours during exercise ses-
sions; two of these trials additionally employed an evaluation of exercise behaviours during the
sessions and reported exercise adherence regarding exercise intensity and duration (Courneya
2007, Kim 2006). The aerobic exercise training group in Courneya 2007 met their prescribed
duration 96% of the time, and likewise for intensity 87% of the time. The resistance exercise
training group of the same study completed all prescribed nine exercises - two sets each, with
eight to 12 repetitions each set at least 95% of the time. Kim 2006 reported an average duration
of exercise of 43 minutes, with an average duration of exercise within prescribed target heart rates
of 28 minutes.
All but two of the studies (MacVicar 1986, Winningham 1988) monitored physical activity but
did not report the data. There was a range of methods used for monitoring physical activity:
some trials employed (electronic) devices such as heart rate monitors (e.g., Drouin 2002, Kim
2006); accelerometers (Schwartz 2007) or pedometers (Payne 2008); others used pulse rates or
perceived exertion (Mock 1997, Mock 2005).
Exercise levels in non-exercising control groups (contamination) were reported in three trials
(Crowley 2003, Courneya 2007, Mock 2005). Whereas Crowley 2003 stated the significant differ-
ence in activity levels between the two groups in terms of frequency and duration of exercise,
Mock 2005 observed 39% of the control group beginning regular walking during the study, at a
Results: evidence synthesis
98
level greater than 45 minutes per week, which was regarded as demonstrating considerable con-
tamination. However, when adopting the underlying cut-point of 45 minutes per week (Mock
2005), contamination can also be observed in the trial of Crowley 2003: participants in the con-
trol group walked a mean time of 53 minutes per week. In Courneya 2007, less than 15% of
participants in the control group reported regular exercise.
No information on adherence to prescribed exercise was given in four trials (MacVicar 1986,
Schwartz 2007, Payne 2008, Winningham 1988).
Attendance rates are difficult to compare due to different exercise stimuli prescribed across stud-
ies and due to various cut-points for determining rates of adherent participants. In sum, adher-
ence seemed to be a problem in some, but not all, studies.
One study (Courneya 2007) examined predictors of supervised exercise adherence in breast can-
cer patients receiving adjuvant chemotherapy. Exercise adherence in that study was good, with an
adherence rate of 70%, but was still not optimal. Independent predictors of adherence were loca-
tion/centre, aerobic fitness, disease stage, and depression. Results suggested that motivational
variables were not important predictors of adherence in those breast cancer patients already mo-
tivated enough to volunteer for an exercise study. Higher exercise adherence to supervised exer-
cise training during chemotherapy for breast cancer was achieved by women who were “fitter,
stronger, thinner, more advanced disease stage, better educated, less depressed, and not smoking”
(Courneya et al. 2008b). But most of the variation in exercise adherence remained unexplained
(21% was explained in multivariate analyses) indicating that there are other important predictors.
Furthermore, Courneya 2007 prospectively assessed exercise barriers by tracing reasons for each
session that any women in the intervention groups missed. By this means they obtained 2,090
reasons for missed exercise sessions. Based on content analysis, exercise barriers were assigned to
disease-/treatment-related barriers, life-related barriers, or motivation-related barriers. Disease-/
treatment-related barriers, such as fatigue, dizziness, depression, or coincidence with chemother-
apy day, accounted for more than half of all missed sessions. Life-related barriers, such as vaca-
tion or work issues, accounted for about one-third of missed sessions. Only a small portion of
barriers (13%) appeared to be motivation-related (Courneya et al. 2008a).
Results: evidence synthesis
99
Table 17: Summary of findings: immediate post-intervention effects
Outcome No. of
studies No. of participants Effect size
[95% CI]
Physical functioning 1 123 SMD 0.04 [-0.33; 0.42]
Physical fitness
Cardiorespiratory fitness
Strength
Body composition
8
4
4
709
328
414
SMD 0.54 [0.32; 0.77]
SMD 0.42 [0.06; 0.78]
SMD -0.29 [-0.55; -0.03]
Symptoms
Fatigue
Nausea
Endocrine symptoms
7
1
1
714
42
174
SMD -0.17 [-0.32; -0.02]
RR 0.57 [0.34, 0.94]
SMD -0.14 [-0.44; 0.16]
Biological markers
Natural killer cells
T-cells
Oxidative stress
Bone density
1
1
1
1
21
21
15
66
SMD 0.24 [-0.65, 1.12]
SMD 0.63 [-0.27, 1.54]
SMD -0.53 [-1.63, 0.57]
SMD 0.14 [-0.37; 0.65]
Sleep disturbances 1 46 SMD -0.67 [-1.27; -0.08]
Emotional distress
Anxiety
Depression
Mood disturbance
Negative affects
2
3
1
1
269
443
21
174
SMD -0.25 [-0.54; 0.04]
SMD -0.24 [-0.43; -0.04]
SMD -0.69 [-1.60, 0.22]
SMD -0.30 [-0.60, 0.00]
Positive psychological function
Self-esteem
Positive affects
1
1
223
174
SMD 0.28 [0.00; 0.56]
SMD 0.44 [0.14, 0.74]
Quality of life
Cancer-specific
3
416
SMD 0.23 [0.03; 0.43]
Harm
Lymphedema
1
242
RR 0.85 [0.31; 2.3]
Results: evidence synthesis
100
5.5 Forest plots for immediate post-intervention effects
For those outcomes, where combining of results was feasible, the pooled effect sizes are presented in a forest plot. Studies with two intervention groups
are included in meta-analyses with both intervention arms using the study-ID plus a code for the exercise intervention in order to distinguish between the
two interventions and their contribution the overall pooled effect (AET and RET for aerobic resp. resistance exercise training; and SD and SU for self-
directed resp. supervised exercise). More forest plots for immediate post-intervention outcomes (presenting sub-totals only) can be found in Appendix 12
Review : Exercise for w omen receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 01 Cardiorespiratory fitness
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Campbell 2005 10 1423.00(261.00) 9 1083.00(176.00) 3.92 1.44 [0.41, 2.48]
Drouin 2002 13 22.60(6.20) 8 16.60(2.20) 4.46 1.13 [0.17, 2.09]
Schw artz 2007 RET 21 1055.00(177.00) 11 944.00(241.00) 6.55 0.54 [-0.20, 1.28]
Schw artz 2007 AET 22 1228.00(322.00) 12 944.00(241.00) 6.56 0.93 [0.19, 1.68]
Mock 1997 22 3371.00(300.46) 22 3089.00(300.46) 8.26 0.92 [0.30, 1.55]
Kim 2006 22 1810.10(369.40) 19 1630.40(351.50) 8.27 0.49 [-0.14, 1.11]
Segal 2001 SD 40 26.30(5.30) 20 25.10(6.10) 9.85 0.21 [-0.33, 0.75]
Segal 2001SU 42 26.20(5.10) 21 25.10(6.10) 10.12 0.20 [-0.33, 0.72]
Courneya 2007 AET 71 25.70(7.40) 36 23.50(5.40) 13.02 0.32 [-0.08, 0.72]
Courneya 2007 RET 77 24.20(6.10) 37 23.50(5.40) 13.31 0.12 [-0.27, 0.51]
Mutrie 2007 82 1135.00(143.00) 92 984.00(221.00) 15.68 0.80 [0.49, 1.11]
Total (95% CI) 422 287 100.00 0.54 [0.32, 0.77]
Test for heterogeneity: Chi² = 18.14, df = 10 (P = 0.05), I² = 44.9%
Test for overall effect: Z = 4.71 (P < 0.00001)
-4 -2 0 2 4
Favours Control Favours exercise
Figure 7: Meta-analysis for cardiorespiratory fitness
Results: evidence synthesis
101
Review : Exercise for w omen receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 02 Strength
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Battaglini 2007 10 116.35(8.90) 10 102.70(15.20) 10.29 1.05 [0.10, 2.00]
Drouin 2002 13 31.30(6.50) 8 32.00(6.40) 11.40 -0.10 [-0.99, 0.78]
Schw artz 2007 RET 21 10.80(5.10) 11 9.50(4.10) 14.42 0.26 [-0.47, 1.00]
Schw artz 2007 AET 22 13.72(6.40) 12 9.50(4.10) 14.57 0.72 [-0.01, 1.45]
Courneya 2007 RET 77 31.90(10.80) 37 24.60(7.80) 24.61 0.73 [0.33, 1.13]
Courneya 2007 AET 71 24.70(7.50) 36 24.60(7.80) 24.71 0.01 [-0.39, 0.41]
Total (95% CI) 214 114 100.00 0.42 [0.06, 0.78]
Test for heterogeneity: Chi² = 10.08, df = 5 (P = 0.07), I² = 50.4%
Test for overall effect: Z = 2.29 (P = 0.02)
-4 -2 0 2 4
Favours control Favours exercise
Figure 8: Meta-analysis for strength
Review : Exercise for w omen receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 02 Strength
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Battaglini 2007 10 116.35(8.90) 10 102.70(15.20) 12.20 1.05 [0.10, 2.00]
Schw artz 2007 RET 21 10.80(5.10) 11 9.50(4.10) 20.45 0.26 [-0.47, 1.00]
Courneya 2007 RET 77 31.90(10.80) 37 24.60(7.80) 67.35 0.73 [0.33, 1.13]
Total (95% CI) 108 58 100.00 0.67 [0.34, 1.01]
Test for heterogeneity: Chi² = 1.88, df = 2 (P = 0.39), I² = 0%
Test for overall effect: Z = 3.99 (P < 0.0001)
-4 -2 0 2 4
Favours control Favours exercise
Figure 9: Meta-analysis for strength with resistance exercise training studies
Results: evidence synthesis
102
Review : Exercise for w omen receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 03 Body composition
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Battaglini 2007 10 25.90(2.90) 10 31.20(4.10) 5.98 -1.43 [-2.44, -0.42]
Drouin 2002 13 30.10(7.30) 8 31.70(6.10) 7.57 -0.22 [-1.11, 0.66]
Courneya 2007 AET 64 37.90(8.90) 34 39.80(8.80) 24.81 -0.21 [-0.63, 0.20]
Courneya 2007 RET 66 37.20(9.00) 35 39.80(8.80) 25.22 -0.29 [-0.70, 0.12]
Mutrie 2007 82 26.90(4.30) 92 27.90(6.90) 36.42 -0.17 [-0.47, 0.13]
Total (95% CI) 235 179 100.00 -0.29 [-0.55, -0.03]
Test for heterogeneity: Chi² = 5.61, df = 4 (P = 0.23), I² = 28.6%
Test for overall effect: Z = 2.21 (P = 0.03)
-4 -2 0 2 4
Favours exercise Favours control
Figure 10: Meta-analysis for body composition
Results: evidence synthesis
103
Review : Exercise for w omen receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 04 Fatigue
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Campbell 2005 10 2.43(1.94) 9 4.35(3.48) 2.66 -0.66 [-1.59, 0.27]
Drouin 2002 13 3.40(1.90) 8 3.90(2.40) 2.95 -0.23 [-1.11, 0.66]
Mock 1997 22 26.12(20.27) 24 43.05(36.37) 6.62 -0.56 [-1.15, 0.03]
Segal 2001 SD 40 -60.80(23.50) 20 -61.60(17.70) 8.01 0.04 [-0.50, 0.57]
Segal 2001SU 42 -55.80(24.00) 21 -61.60(17.70) 8.35 0.26 [-0.27, 0.78]
Courneya 2007 AET 74 -36.80(10.40) 36 -34.90(12.50) 14.50 -0.17 [-0.57, 0.23]
Courneya 2007 RET 76 -36.30(9.40) 37 -34.90(12.50) 14.92 -0.13 [-0.53, 0.26]
Mock 2005 54 3.50(2.40) 54 3.70(2.60) 16.21 -0.08 [-0.46, 0.30]
Mutrie 2007 82 -120.80(26.70) 92 -113.30(25.00) 25.78 -0.29 [-0.59, 0.01]
Total (95% CI) 413 301 100.00 -0.17 [-0.32, -0.02]
Test for heterogeneity: Chi² = 6.74, df = 8 (P = 0.57), I² = 0%
Test for overall effect: Z = 2.17 (P = 0.03)
-4 -2 0 2 4
Favours exercise Favours control
Figure 11: Meta-analysis for cancer-related fatigue
Results: evidence synthesis
104
Review : Exercise for w omen receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 05 Depression
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Mock 1997 22 9.51(23.50) 24 21.05(23.50) 10.82 -0.48 [-1.07, 0.10]
Courneya 2007 AET 74 9.70(9.30) 36 10.80(9.40) 23.52 -0.12 [-0.52, 0.28]
Courneya 2007 RET 76 10.60(9.50) 37 10.80(9.40) 24.21 -0.02 [-0.41, 0.37]
Mutrie 2007 82 8.60(6.80) 92 11.50(8.60) 41.45 -0.37 [-0.67, -0.07]
Total (95% CI) 254 189 100.00 -0.24 [-0.43, -0.04]
Test for heterogeneity: Chi² = 2.93, df = 3 (P = 0.40), I² = 0%
Test for overall effect: Z = 2.42 (P = 0.02)
-4 -2 0 2 4
Favours exercise Favours control
Figure 12: Meta-analysis for cancer-related depression
Review : Exercise for w omen receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 07 Anxiety
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Mock 1997 22 10.44(24.70) 24 26.93(24.70) 20.59 -0.66 [-1.25, -0.06]
Courneya 2007 AET 74 35.00(11.70) 36 37.40(12.00) 39.25 -0.20 [-0.60, 0.20]
Courneya 2007 RET 76 36.40(12.70) 37 37.40(12.00) 40.16 -0.08 [-0.47, 0.31]
Total (95% CI) 172 97 100.00 -0.25 [-0.54, 0.04]
Test for heterogeneity: Chi² = 2.55, df = 2 (P = 0.28), I² = 21.6%
Test for overall effect: Z = 1.66 (P = 0.10)
-4 -2 0 2 4
Favours exercise Favours control
Figure 13: Meta-analysis for anxiety
Results: evidence synthesis
105
Review : Exercise for w omen receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 06 Cancer-specific quality of life
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Campbell 2005 10 111.20(14.10) 9 94.30(28.40) 4.54 0.73 [-0.20, 1.67]
Courneya 2007 AET 74 144.70(25.20) 36 139.90(28.20) 25.09 0.18 [-0.22, 0.58]
Courneya 2007 RET 76 140.90(24.80) 37 139.90(28.20) 25.87 0.04 [-0.35, 0.43]
Mutrie 2007 82 106.50(21.90) 92 99.70(20.30) 44.50 0.32 [0.02, 0.62]
Total (95% CI) 242 174 100.00 0.23 [0.03, 0.43]
Test for heterogeneity: Chi² = 2.43, df = 3 (P = 0.49), I² = 0%
Test for overall effect: Z = 2.27 (P = 0.02)
-4 -2 0 2 4
Favours control Favours exercise
Figure 14: Meta-analysis for cancer-specific quality of life
Results: evidence synthesis
106
5.6 Sustainability of effects
Three exercise studies about breast cancer patients receiving adjuvant therapy have reported
longer-term follow-up of outcomes, going beyond the immediate post-intervention time point;
however, only two of these include health-related outcomes and moreover one study could not
be confirmed as assessing post-intervention outcomes after closer examination (Kim 2006).
These studies provide an insight into the question of whether any of the benefits of exercise
training during breast cancer therapy could be maintained into survivorship (i.e., longer-term
effects) or if any new benefits emerged (i.e., late effects). Furthermore, these studies provide
information on whether participants continued with exercise after the exercise study and whether
such continuation resulted in further improvements in outcomes.
5.6.1 Exercise behaviour
Maintaining exercise beyond the exercise intervention, i.e., long-term adherence to exercise is a
critical concern because the benefits of exercise may not persist when exercise is discontinued.
There is conflicting evidence regarding long-term exercise maintenance post-intervention. Three
studies provided post-intervention data on exercise behaviour or physical activity behaviour. Kim
2006 assessed levels of physical activity four months post-intervention: there were no overall
group differences after four months. However, within-group changes indicated significant in-
creases in voluntary exercise, as well as a significant decrease in sedentary activity in the interven-
tion group, indicating positive long-term changes during the post-intervention period. Compara-
ble changes could not be observed in the control group. However, the so-called post-intervention
period in this study is the same as the intervention period in other studies included in this review:
during the post-intervention period participants were continuously encouraged by a trained exer-
cise physiologist every two weeks to maintain their physical activity at home or in a community
setting. This procedure was employed in studies with self-directed home based interventions. The
main intervention of Kim 2006 was an on-site aerobic exercise program under direct supervision
by trained exercise physiologists in an exercise facility within the School of Nursing. Strictly
speaking, Kim 2006 employed a graded exercise intervention of six months, with two months
supervised exercise followed by four months of self-directed exercise. Merely, reporting in this
study was guided by a deviant terminology regarding exercise intervention and post-intervention
period.
This leaves two studies that present data on long-term adherence to exercise and/or physical
activity. Mutrie 2007 observed a decrease in the activity levels of the exercise group between the
end of the intervention period and the final assessment six months post-intervention. At the six-
Results: evidence synthesis
107
month follow-up assessment, there were no differences in physical activity between women who
had originally been assigned to the intervention group and those of the control group.
Courneya 2007 categorised participants into those meeting and those not meeting current guide-
lines for aerobic exercise (150 minutes of moderate-to-vigorous exercise per week) and for resis-
tance exercise (more than two resistance-training sessions per week). They observed that in the
aerobic exercise training group, as well as in the resistance-exercise training group, more than
30% of participants reported meeting at least one guideline at the six-month follow-up, and 18%
and 29% in the aerobic exercise training group and in the resistance exercise training group, re-
spectively, reported meeting both guidelines.
While data from Courneya 2007 suggest that a supervised exercise training program during adju-
vant chemotherapy may be an effective strategy for helping sedentary breast cancer patients’
transition towards an active lifestyle, data from Mutrie 2007 show difficulties of maintaining
physical activity without classes and supervision.
Pooling of physical activity data (assessed as minutes per week) from Mutrie 2007 and Kim 2006
indicates that participants in the exercise groups demonstrated higher physical activity levels
(standardised mean difference 0.25, with 95% confidence interval -0.02 to 0.51). However, it has
to be noted, that the participants described by Kim 2006 still received motivational support in the
post-intervention period. Results from meta-analysis based on the unstandardised mean differ-
ence shows that participants from the intervention group are 73 minutes more physically active
compared to participants from the control group (Figure 15).
5.6.2 Long-term intervention effects
Both studies with long-term follow-up of six months (Courneya 2007, Mutrie 2007) reported that
immediate post-intervention effects were largely maintained at six-month follow-up. Mutrie 2007
reported long-term intervention effects for cardiorespiratory fitness, endocrine symptoms, de-
pression, and positive affects. Furthermore, they observed a borderline-significant late effect for
general quality of life. Results of Courneya 2007 showed a longer-term effect on self-esteem and
a late effect on anxiety that emerged in the post-intervention period. The effect sizes are pre-
sented in Table 18.
Results could be pooled for three outcomes: fatigue, depression, and cancer-specific quality of
life. Meta-analyses for these outcomes indicated a small, and not statistically significant, effect of
exercise on reducing fatigue (see Figure 16), a small, statistically significant, effect of exercise on
reducing depression (see Figure 17), and a small, not statistically significant, effect on improving
cancer-specific quality of life (see Figure 18).
Results: evidence synthesis
108
Thus, there is increasing, though limited, evidence for sustained benefits of exercise at six months
post-intervention.
5.6.3 Recurrence and survival
There is no evidence that exercise during adjuvant therapy for breast cancer improves recurrence
and survival. These long-term outcomes were not examined in the body of studies included in
this review.
Table 18: Summary of findings: long-term effects
Outcome No. of
studies No. of participants Effect size
[95% CI]
Physical fitness
Cardiorespiratory fitness
1
177
SMD 0.63 [0.33; 0.94]
Body mass index 1 177 WMD 0.00 [-1.47; 1.47]
Symptoms
Fatigue
2
378
SMD -0.15 [-0.40; 0.10]
Endocrine symptoms 1 177 SMD -0.13 [-0.42; 0.17]
Emotional distress
Depression
2
378
SMD -0.27 [-0.48; -0.06]
Anxiety 1 201 SMD -0.30 [-0.60; 0.01]
Negative affects 1 177 SMD -0.26 [-0.56; 0.04]
Positive psychological func-
tion
Self-esteem
1
201
SMD 0.24 [-0.06; 0.54]
Positive affects 1 177 SMD 0.4 [0.10; 0.70]
Cancer-specific quality of life 2 378 SMD 0.24 [0.00; 0.49]
Physical activity 2 218 SMD 0.25 [-0.02; 0.51]
Results: evidence synthesis
109
5.6.4 Forest plots for long-term intervention outcomes
Results for long-term outcomes are presented in meta-analyses, where combining of results was feasible. To distinguish intervention arms of Courneya
2007, a code for aerobic and resistance exercise was created (AET resp. RET). More forest plots for long-term outcomes can be found in Appendix 12.
Review : Exercise for w omen receiving adjuvant therapy for breast cancer
Comparison: 02 Long-term effects
Outcome: 11 Physical activity
Study Exercise Control WMD (random) Weight WMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Mutrie 2007 82 492.00(327.00) 95 427.00(370.00) 45.86 65.00 [-37.69, 167.69]
Kim 2006 22 340.80(136.80) 19 261.00(167.40) 54.14 79.80 [-14.72, 174.32]
Total (95% CI) 104 114 100.00 73.01 [3.47, 142.56]
Test for heterogeneity: Chi² = 0.04, df = 1 (P = 0.84), I² = 0%
Test for overall effect: Z = 2.06 (P = 0.04)
-1000 -500 0 500 1000
Favours control Favours exercise
Figure 15: Meta-analysis for physical activity
Review : Exercise for w omen receiving adjuvant therapy for breast cancer
Comparison: 02 Long-term effects
Outcome: 02 Fatigue
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Courneya 2007 AET 68 -42.10(10.50) 30 -41.50(9.80) 26.96 -0.06 [-0.49, 0.37]
Courneya 2007 RET 73 -40.80(10.50) 30 -41.50(9.80) 27.41 0.07 [-0.36, 0.49]
Mutrie 2007 82 -41.30(9.70) 95 -37.60(11.80) 45.62 -0.34 [-0.64, -0.04]
Total (95% CI) 223 155 100.00 -0.15 [-0.40, 0.10]
Test for heterogeneity: Chi² = 2.69, df = 2 (P = 0.26), I² = 25.8%
Test for overall effect: Z = 1.18 (P = 0.24)
-1 -0.5 0 0.5 1
Favours exercise Favours control
Figure 16: Meta-analysis for long-term effect on fatigue
Results: evidence synthesis
110
Review : Exercise for w omen receiving adjuvant therapy for breast cancer
Comparison: 02 Long-term effects
Outcome: 01 Depression
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Courneya 2007 AET 68 7.20(7.50) 30 10.20(9.50) 24.08 -0.36 [-0.80, 0.07]
Courneya 2007 RET 73 9.60(10.40) 30 10.20(9.50) 24.95 -0.06 [-0.48, 0.37]
Mutrie 2007 82 8.40(7.20) 95 10.80(7.50) 50.97 -0.32 [-0.62, -0.03]
Total (95% CI) 223 155 100.00 -0.27 [-0.48, -0.06]
Test for heterogeneity: Chi² = 1.26, df = 2 (P = 0.53), I² = 0%
Test for overall effect: Z = 2.47 (P = 0.01)
-1 -0.5 0 0.5 1
Favours exercise Favours control
Figure 17: Meta-analysis for long-term effect on depression
Review : Exercise for w omen receiving adjuvant therapy for breast cancer
Comparison: 02 Long-term effects
Outcome: 03 Cancer-specific quality of life
Study Control Exercise SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Courneya 2007 AET 68 156.30(24.00) 30 152.40(26.40) 26.71 0.16 [-0.27, 0.59]
Courneya 2007 RET 73 152.90(26.00) 30 152.40(26.40) 27.23 0.02 [-0.41, 0.44]
Mutrie 2007 82 109.40(16.50) 92 101.20(21.70) 46.06 0.42 [0.12, 0.72]
Total (95% CI) 223 152 100.00 0.24 [0.00, 0.49]
Test for heterogeneity: Chi² = 2.54, df = 2 (P = 0.28), I² = 21.4%
Test for overall effect: Z = 1.92 (P = 0.05)
-1 -0.5 0 0.5 1
Favours control Favours exercise
Figure 18: Meta-analysis for long-term effect on cancer-specific quality of life
Results: evidence synthesis
111
5.7 Harm from exercise interventions
Adverse effects due to exercise were reported for nine of the studies included in the effectiveness
review (Battaglini 2007, Campbell 2005, Courneya 2007, Crowley 2003, Drouin 2002, Mock 1997,
Mock 2005, Schwartz 2007, Segal 2001). Harmful effects were observed in three of these studies:
Crowley 2003 reported onset of lymphedema in one participant, Drouin 2002 reported shoulder
tendonitis and decreases in strength due to overtraining in one participant, and Courneya 2007
reported some mild to moderate harmful effects related to maximal exercise testing in two par-
ticipants (hypotension, nausea, dizziness, weakness, and mild diarrhoea). Both participants recov-
ered quickly. No harmful effects were observed in the other six studies that reported on harmful
effects. However, none of these studies described how relevant information was collected. To
summarise, passive surveillance (spontaneous reports) of harm across the included studies of benefit
has not revealed any increase in injury rates or other adverse events in the exercise groups com-
pared to non-exercising control groups. No single study has performed active surveillance of
injuries in exercise interventions studies during adjuvant treatment for breast cancer. One study
(Courneya 2007) actively surveyed lymphedema and no increased risk of lymphedema associated
with exercise during adjuvant treatment for breast cancer was reported. Quite the contrary,
women in the exercise groups less frequently experienced an onset of lymphedema compared to
women in the control group (RR 0.85, 95% CI 0.31 to 2.3). Onset of lymphedema (defined as a
200 ml increase in the difference between affected and unaffected arm volumes, using volumetric
arm measurement based on water displacement) was a secondary outcome in this study and
therefore regularly monitored. The effect size is presented in Table 17.
The included studies of harm all performed active surveillance. Lymphedema and injuries were
evaluated in the included studies. The three included studies (see Table 19) yielded evidence of no
increased risk of lymphedema due to exercise. Active injury surveillance in exercise studies after
adjuvant treatment for breast cancer indicated that weight training is well tolerated, with rates of
minor injuries comparable to those observed in the general population. However, results relating
to harm were evaluated after adjuvant cancer treatment had been completed. Vulnerability may
be increased during adjuvant treatment.
Results: evidence synthesis
112
Table 19: Summary of findings: harm
Study-ID, n Lymphedema Injury (%)
McKenzie 2003
n=14
with lymphedema
No group differences in changes of arm
volume
n/a
Schmitz 2005
n=85
for injury assessment
n=45
for lymphedema as-
sessment
Arm circumference increase of ≥ 2cm:
Control: 1/22 (4.4%);
Intervention: 0/23 (0%).
Self-reported incidence:
Control: 2/16 (12.5%);
Intervention: 2/16 (12.5%).
Increase in symptoms:
Control: 3/22 (13.6%);
Intervention: 0/23 (0%).
Intervention:
6 months: 10.5%,
12 months: 22.5%
Control after delayed
intervention:
6 months: 20%
Basen-Engquist 2006
n=60
No group differences in numbers of arm
circumference increase of ≥ 2cm
n/a
5.8 Perception of exercise during breast cancer treatment
Besides the effects of the exercise intervention, experiences of participants are of interest as well.
One attendant focus-group study (Emslie et al. 2007) explored the perceptions of women exer-
cising in community-based groups established solely for women during their breast cancer treat-
ment. They found that classes helped them to adopt exercise or return to exercise. With regards
to the expert instruction in these groups, the women especially valued not having to explain their
problems to the instructor, and at the same time they could rebuild confidence in their ability to
participate in exercise in spite of feeling vulnerable. Women appreciated that through expert
instruction constant adjustment of pace and exercise to the composition of the group was made
possible. Then, with regards to the group context, the women emphasised how important it was
for them to exercise among women in similar circumstances; they valued the empathy, solidarity,
and acceptance they received from others, the friendships that evolved, the information-gathering
function of the group, and the freedom in choosing to share experiences. Finally, the exercise
focus in classes was seen as “upbeat, enjoyable and fun” (Emslie et al. 2007), giving them a
chance to forget about cancer for a while. Participants commonly complained that exercise
classes stopped abruptly at the end of the intervention of the supervised group exercise trial they
had taken part in (Emslie et al. 2007).
Discussion
113
6 Discussion
6.1 Summary of main results
Physical exercise appears to be an intervention that effectively improves health-related physical
fitness even in the period of breast cancer treatment. There is evidence for a medium effect of
exercise on cardiorespiratory fitness, strength, and body composition. Moreover, exercise offered
a small benefit against fatigue, anxiety and depression compared to no exercise. Evidence for
quality-of-life outcomes is limited; cancer-site-specific quality of life was better in exercisers than
in non-exercisers. Similar effects could not be observed for generic or cancer-specific quality of
life; generic or cancer-specific measures may be too broad for capturing changes in breast-cancer-
specific health outcomes.
For the purposes of forming an impression of the magnitude of effects, the effect size statistic
can be translated into a more intuitively understandable unit: e.g., for cardiorespiratory fitness, a
medium effect from physical exercise training could be established. The effect size of 0.54 means
that the average exercise group scored 0.54 standard deviation units higher on cardiorespiratory
fitness after exercise intervention compared to the average control group. In order to provide a
frame of reference and to put this effect size in some interpretable context, it can be translated
into walked distance (12 minutes) as a more intuitively comprehensible metric following a
method suggested by Lipsey and Wilson (2001). To do this, the overall mean and standard devia-
tion on the measure of interest (i.e., walked distance) for the control groups has to be generated
through pooling the results from multiple studies. With this information, an overall mean for the
intervention groups can be calculated on that same metric: it is the control group mean plus the
product of the effect size and the standard deviation value for the metric of interest (Lipsey and
Wilson 2001). So, for the control groups a mean of 977 meters walking distance was calculated,
together with a standard deviation of 208 meters, which – multiplying 0.54, the effect size for
cardiorespiratory fitness, by 201 meters – gives the equivalent of 112 meters Thus, with the con-
trol group’s mean of 977 meters, the average exercise group at post-intervention achieved 1089
meters, within these 12 minutes of test duration.
The potential harm of exercise during breast cancer treatment has generally been disregarded in
the literature. Where harm has been recorded, it has been spontaneously reported by participants,
rather than systematically assessed by active surveillance (where participants are asked about the
occurrence of specific harms in structured questionnaires or interviews, or predefined diagnostic
tests are performed at pre-specified time intervals). Passive surveillance of harm across the stud-
ies included in the review did not reveal any increase either in injury rates or for other harms in
the exercise groups, when compared to non-exercising control groups. No single study per-
Discussion
114
formed active surveillance of injuries occurring in exercise interventions during adjuvant treatment
for breast cancer. Of some relevance here however, is that active injury surveillance undertaken
in exercise studies after adjuvant treatment for breast cancer has indicated that weight training is
well tolerated, with rates of minor injuries comparable to those observed in the general popula-
tion. Active surveillance of lymphedema both during (one study) and after adjuvant treatment
(two studies) did not produce evidence of any increased risk associated with exercise during adju-
vant treatment for breast cancer. In summary, the present available evidence regarding safety of
exercise during adjuvant breast cancer treatment is limited.
Primary studies have examined exercise interventions over very broad ranges of exercise volume:
frequencies ranged from two to six times per week; duration from 15 to 60 minutes, intensity
from low to moderate with heart-rate ranges between 50 to 85 percent of maximum heart rate.
Within these broad ranges, there is little empirical data to give recommendations regarding exer-
cise volume. The available evidence does not allow the determining of the relationship between
the exercise volume and either the changes in health-related physical fitness and other physical or
mental health outcomes. Furthermore, there is insufficient evidence to determine whether aero-
bic exercise, resistance exercise, or mixed exercise is the most appropriate type of exercise for
women receiving adjuvant treatment of breast cancer. Finally, evidence is insufficient regarding
the individualisation of exercise prescriptions: treatment-related symptoms may require that pre-
scriptions need to be individualised; however, factors that must be considered in the develop-
ment of individualised exercise programs still have to be identified systematically as well as sub-
sequent adjustments of the exercise program.
Regarding long-term post-intervention effects of exercise, evidence is preliminary only. There
was a trend of benefits being maintained over time. However, differences between exercise and
control groups decreased compared to the immediate post-intervention time. Fitness outcomes
require attendance of participants and were examined in only one study at six-month follow up.
Pooling was feasible for patient-reported outcomes such as fatigue and depression. Pooling re-
sults from two studies still indicated a small benefit from exercise. However, the pivotal question
is whether women do continue their regular exercise program after the end of exercise interven-
tion. Maintenance of physical activity, however, appears to be a problem and particularly presents
a challenge for formerly sedentary women.
In sum, to answer the overall guiding review questions: yes, women should be offered exercise in
the period of adjuvant therapy for breast cancer. Exercise can be expected to provide benefits for
health-related physical fitness. There is modest evidence that benefits can also be expected for
other physical health outcomes such as alleviation of fatigue symptoms, and for mental health
Discussion
115
outcomes like anxiety and depression. The possibility of harm due to exercise in this sensitive
period could not be ruled out, since this problem has not been systematically examined in pri-
mary studies. There is limited relevant evidence from studies of benefit, but it does not indicate
any gateway for harm through exercise. Nevertheless, providing such evidence is necessary be-
cause, in the absence of such evidence, patients and physicians may avoid physical exercise train-
ing.
6.2 Applicability and overall completeness of evidence
One concern about the applicability of the results is the age of patients. Compared to the mean
age of 63 years at the time of breast cancer diagnosis in Germany (Batzler et al. 2008), the women
included in the primary studies of this review were relatively young, with a mean age of 50 years -
due to the restrictive inclusion criteria. Thus, evidence about exercise in women receiving adju-
vant treatment of breast cancer has been derived largely from research on middle-aged women. If
we try to apply these results in the context of older breast cancer patients, several problems
emerge: Do older breast cancer patients derive similar benefits from exercise? Are older breast
cancer patients more vulnerable to injury? Are exercise participation rates and adherence to an
exercise program comparable? Are determinants of exercise motivation and behaviour compara-
ble? One recent pilot study (Payne 2008) did focus on older women. Participants had a mean age
of 65 years, compared with the mean age of 50 years across the other studies. However, results
from this small study are only preliminary and thus older women remain an understudied patient
population.
One further problem regarding applicability are recruitment rates which vary considerably across
studies: for example, Mutrie 2007 approached 1,114 women and included 203 (18%) of them,
while Mock 1997 approached 65 women and 50 (77%) of these agreed to participate. Given the
range of reported recruitment rates, it can be assumed that studies were subject to selection bias,
and so participants might differ from the target population with respect to relevant characteris-
tics. For example, it is possible that consenting women in the control groups may have had more
disposition to increase exercise or activity than those declining to participate, simply as a function
of their general activity levels and motivation. Thus, it is difficult to assess the generalisability of
the results to the wider population.
The evidence of this review is not complete: harm appears to be disregarded in the literature on
exercise intervention studies during treatment. This gap in the evidence appears all the more
important in light of the finding that none of the reviewed studies have explored the appropri-
ateness, effectiveness, or safety of exercise in an older breast cancer population. Given the known
functional decline in older cancer survivors, going along with loss of skeletal muscle mass (sarco-
Discussion
116
penia), loss of flexibility, muscle weakness, and decreased bone density, which can be further
exacerbated by breast cancer disease and treatment, an increased risk of exercise-related traumatic
or overuse injuries has to be taken into account. Previously sedentary middle-aged and older
women who are treated with cardiotoxic drugs may be at increased risk of an adverse cardiovas-
cular event occurring during or following an acute bout of physical exercise and thus, cardiovas-
cular risks of exercise during cancer treatment should not be ignored. Sudden cardiac death and
acute myocardial infarction are serious complications of exercise in adults, and pre-participation
stress testing often does not reveal any abnormality in patients who subsequently suffer an acute
cardiac event (Thompson 2001). A commensurate statement regarding cardiovascular risks is
overdue because virtually all adjuvant therapies for breast cancer are associated with unique and
varying degrees of cardiovascular injuries (Jones et al. 2007), and as women progress through the
selected treatment regimens, they will be subjected to a series of cardiovascular insults.
Problems of applicability and the lack of evidence of harm may deter oncologists from recom-
mending exercise to women with breast cancer, although the evidence – as presented in this
review – suggests that exercise training may enhance health-related physical fitness, may alleviate
debilitating symptoms such as fatigue, and may improve such mental health outcomes as anxiety
and depression. Oncologists’ attitudes from a Canadian national survey indicate that there is
already consensus that exercise is beneficial (62%), important (56%) and safe (63%) for patients
with cancer during treatment (Jones et al. 2005). However, despite these positive attitudes, only
28% of oncologists had recommended exercise to their patients in the last month. Similarly, a
national survey in the United Kingdom showed that most clinicians (56%) did not routinely dis-
cuss physical activity with their patients (Daley et al. 2008). Evidence of the relationship, if any,
between level of exercise and increased harm from exercise, may contribute to further develop-
ment of the field of exercise interventions applied to breast cancer patients.
6.3 Strengths and limitations of included studies
Since the earliest exercise intervention studies during adjuvant therapy for breast cancer, the field
of exercise interventions has grown impressively and the methodological rigour has improved
within this time as well. The amount of evidence is based on 15 controlled studies that could be
included in this review. Recent studies more often, though not always, had larger sample sizes
(with nearly 100 participants per group), and were based on power calculations. Adequate sample
size is important, since benefits of exercise interventions may be relatively small. Therefore, the
number of participants included should be great enough to allow the detection of small differ-
ences between groups. Also reporting has improved, with nearly complete reporting according to
CONSORT (CONsolidated Standards of Reporting Trials) guidelines in some recent studies.
Discussion
117
Key methodological limitations of the included studies were observer blinding, concealment of
intervention allocation, and missing intention-to-treat analysis. Moreover, there were some incon-
sistencies between results, and for several results wide confidence intervals were observed.
A wide range of outcome constructs was assessed across the studies, which made it difficult to
combine outcomes in meta-analysis. For example, emotional distress was assessed by means of,
among other indicators, mood, anxiety, depression, and negative affects; then body composition
was evaluated via weight, body mass index, subcutaneous fat, and lean body mass. Within this
broad range of outcome measures, each outcome was assessed through several different assess-
ment instruments. Moreover, data reporting in early trials was very poor and did not provide
estimates of effect size that could be pooled.
6.3.1 Inconsistencies of effects across studies
Variation in the effects of exercise interventions could be observed for several outcomes. In the
sections below, some factors are discussed that may have affected heterogeneity. First, the per-
centage of total variation across studies that was due to heterogeneity rather than chance was
45% for cardiorespiratory fitness. Heterogeneity was 40% when the study arms with resistance
exercise training were excluded from meta-analysis in a sensitivity analysis, and it was further
reduced to barely 20% when Segal 2001 was excluded. The resulting standardised mean differ-
ence based on 440 participants was 0.73, with the 95% confidence interval of 0.48 to 0.97. Het-
erogeneity in the meta-analysis for cardiorespiratory fitness might be explained based on specific-
ity as a core exercise training principle. The principle of specificity states that training effects
derived from an exercise intervention are specific to the exercise performed and muscles involved
(Balady et al. 2000). To improve cardiorespiratory fitness, predominantly aerobic exercise train-
ing is required, and the effects from resistance exercise training programs can be expected to be
relatively small. Furthermore, in one study (Segal 2001), a non-specific exercise test was employed
as an assessment instrument for cardiorespiratory fitness: cardiorespiratory fitness was assessed
through a step test, but women had trained in walking. It might be expected that effects assessed
on a stepping ergometer would be lower compared to those assessed on a treadmill. Further-
more, compared to other studies, the training stimulus was relative low in the Segal study. This
low training stimulus becomes even more important in light of the observation that baseline
cardiorespiratory fitness of women in this study was relatively high compared to the fitness level
of women in other studies: oxygen uptake in two studies (Drouin 2002, Crowley 2003) was below
the 10th percentile rank for a 50-59 year-old female and, as such, these participants demonstrated
markedly low cardiorespiratory fitness (Balady et al. 2000). In contrast, oxygen uptake of partici-
pants in the Segal trial was around the 30th percentile rank for a 50-59 year-old female. Higher
Discussion
118
fitness levels require higher training stimuli to yield effects. The remaining heterogeneity can be
explained with clinical heterogeneity (various adjuvant treatment regimens) and differences in the
exercise interventions regarding frequency, program duration besides intensity.
Second, variability in results between studies was 50% in the meta-analysis for strength. As for
the meta-analysis for cardiorespiratory fitness, heterogeneity can be attributed to neglect of the
principle of specificity: across included studies, strength was measured not only after resistance
exercise training but also after aerobic exercise training. Typically, only resistance training is de-
signed to increase strength, power, and muscle endurance. Since effects of exercise training tend
to be specific, walking as an aerobic exercise for the lower body could not be expected to change
strength in the upper body. Heterogeneity resolved when only resistance exercise training studies
were pooled.
Furthermore, there was moderate heterogeneity of 47% in the meta-analysis for fatigue, which
could be resolved after exclusion of Battaglini 2007 from sensitivity analysis. One explanatory
approach for this heterogeneity is given in Battaglini’s operating schedule for fatigue assessments:
fatigue assessments were scheduled with some temporal distance to the cancer treatment cycles:
this assessment procedure avoided the highest levels of fatigue in the first days post-treatment. It
can be suspected that differences between groups do not become obvious before recovery from
acute treatment. No information regarding similar procedures was available from the other stud-
ies.
However, sensitivity analyses across outcomes indicated that Battaglini’s dissertation thesis, in-
volving around 20 participants, systematically contributed to heterogeneity in various outcomes;
results from this study suggest greater effects from exercise than other studies. However, since
predominant inconsistencies could be explained and resolved through sensitivity analyses, incon-
sistencies of studies’ results in meta-analyses do not reduce the confidence of results.
6.3.2 Replicability of the intervention and mediational pathways
Basically, exercise interventions in included studies were atheoretical. Theory-based interventions
are preferable because these interventions allow for results that are replicable and generalisable.
The intervention (package) has to be successful at helping sedentary individuals initiate and main-
tain a moderate-intensity exercise regimen. The planning of reasonable and effective exercise-
promotion activities calls for knowledge of the natural fluctuations of exercise in women receiv-
ing adjuvant treatment of breast cancer: determinants of planning for participation, initial adop-
tion of exercise, continued participation, or maintenance, and overall periodicity of participation
(e.g., relapse, resumption, and seasonal variations) have to be characterised and targeted in order
Discussion
119
to influence exercise patterns. Primary studies included in this review lacked a compelling con-
ceptual basis for exercise promotion and behaviour change: apart from two trials (Campbell
2005, Mutrie 2007), interventions were restricted to the delivery of an adequate training stimulus.
These interventions, that focus on exercise dosage only, neglect the complexity of exercise behav-
iour as a psychological, behavioural and social phenomenon.
The exercise studies of this review failed to include an assessment of program integrity, i.e., the
degree to which the exercise interventions were implemented as intended. Integrity information is
particular important in the context of exercise interventions during adjuvant cancer treatment,
since this context stands for dynamic and complex conditions that present numerous obstacles to
complete intervention delivery. Apart from two studies, only the prescribed training stimulus was
available with details for intensity, time, frequency and program period. Because integrity data, an
important source for information about the feasibility of exercise interventions in real-life set-
tings, are lacking, it remains difficult to determine whether the intervention has the potential to
be implemented as planned.
The training stimulus can be severely compromised by a lack of participants’ adherence to the
exercise intervention. For sedentary individuals, a change in personal health behaviour is required
in order to take up regular exercise. Thus, any exercise intervention can additionally be evaluated
according to the degree of behavioural change achieved in the intervention group. Adherence
problems do not only affect participation in exercise sessions and frequency of sessions, but also
affect the training intensity and duration achieved during each exercise session; insufficient exer-
cise intensity or duration may compromise the training stimulus as a whole. However, these two
facets of the training stimulus were poorly evaluated and reported in the included studies.
Besides adherence, the extent of exercise in the control group (contamination) is a further critical
component in exercise studies, and was described as a problem in a few trials. However, exercise
contamination was rarely reported and often only when the exercise program was home based.
This reporting practice indicates that contamination was not systematically investigated across
trials. Accordingly, it can not be determined based on available data whether small or non-
significant results are simply due to failure of implementation or are due to failure of the exercise
program.
The individual’s level of fitness is an important factor to consider before determining the level of
exercise intensity. According to ACSM (Balady et al. 2000), deconditioned individuals may dem-
onstrate increases in their cardiorespiratory fitness from exercise intensities at the lower end of
the intensity continuum, whereas fitter individuals need to work at the higher end of the intensity
continuum to improve fitness. The initial fitness level was considered in only a few studies, which
Discussion
120
were studies that limited participation to sedentary individuals; however, definitions of sedentary
varied.
Replicability of interventions across studies of atheoretical exercise interventions is compromised.
Theory-based interventions allow for results that are replicable and generalisable, and further-
more provide guidance in the development of the intervention and the mechanisms of change.
There was one study that additionally performed a predictor analysis of exercise adherence
(Courneya 2007); the predictor analysis was guided by the theory of planned behaviour, a social
cognitive model of human behaviour that proposes that intention is the most important determi-
nant of behaviour. However, results were not translated into intervention characteristics for
future studies. Moreover, important details of intervention components are not available: no
discussion guides, class outlines and materials, or motivational strategies for telephone counsel-
ling and other details of the intervention are given.
6.4 Potential biases in the review process
The basic search in 2004 was the most comprehensive one. Updating searches were less compre-
hensive and based only on MEDLINE and the Cochrane Central Register of Controlled Trials,
as already explained in detail. The practice of basing updates only on searches of these two regis-
ters means in particular that no further unpublished trials, or those that are difficult to locate
could be identified. However, trials that are difficult to locate are often of low quality (Egger et al.
2003). This tendency, of difficult to locate trials to be of low quality, “raises the worrying possi-
bility that rather than preventing bias through extensive literature searches, bias could be intro-
duced by including trials of low methodological quality” (Egger et al. 2003). Additionally, in spite
of comprehensive attempts to identify all relevant studies, predominantly English language stud-
ies were retrieved for inclusion in this review, with one German trial awaiting assessment after its
publication. This may reflect a language bias, where authors, whose native language is not Eng-
lish, are more likely to publish RCTs in an English-language journal if the results were statistically
significant. The policy taken in the registration of trials may provide an interesting basis for ad-
dressing publication bias in the future (Antes 2004).
The meta-analyses were performed using final values instead of change scores. However, it is
possible for baseline imbalances to occur between intervention groups for one or more variables
in a randomised controlled trial. Hence, statistically significant intervention effects may be due to
the baseline imbalances between intervention and control groups, rather than as a result of the
exercise intervention. There is the possibility, that effects from participating in an exercise pro-
gram may be in fact smaller and not statistically significant.
Discussion
121
The assessment of training stimulus was based on two scores, one for aerobic and one for resis-
tance exercise. For exercise interventions with both aerobic and resistance modules, applicability
of these two single scores (developed for assessing training stimulus of aerobic or resistance
interventions only) may lead to the underestimation of the size of the total training stimulus.
However, for all included trials an adequate training stimulus was determined.
Finally, critical appraisal of more than half of the included studies was performed by one reviewer
only. However, inter-rater reliability of the first charge of included studies indicated almost per-
fect agreement between the two reviewers. Thus, results from critical appraisal can be expected
to be still reliable.
6.5 Agreements with other reviews
Findings of this review are supported from results of other reviews: exercise was found to be
beneficial for cancer patients with any cancers during cancer treatment (Conn et al. 2006), as well
as for cancer survivors with various cancers in the post-adjuvant setting (Stevinson et al. 2004,
Schmitz et al. 2005, Conn et al. 2006, McNeely et al. 2006). Effects in the post-adjuvant setting
seem to be larger when compared to effects of exercise during treatment (Stevinson et al. 2004,
Schmitz et al. 2005). Finally, exercise was found to be beneficial for individuals with cancer-
related fatigue, both during and post cancer, in one Cochrane review (Cramp and Daniel 2008).
Benefits resulting from exercise included physical and mental health outcomes, and quality of life.
There is no evidence for harm in the other reviews that examined exercise in the continuum of
cancer experiences, regardless of focus.
.
Conclusions
122
7 Conclusions and forecast
7.1 Implications for research
There is a need for ongoing research, involving studies with adequate statistical power to further
develop the evidence base.
Achieving a consensus of researchers on outcome measures for exercise studies involving breast
cancer patients receiving adjuvant treatment is needed in order to assist interpretation and com-
parison of results across various interventions. Given the known functional decline, especially in
older women receiving adjuvant treatment of cancer, physical functioning becomes an important
outcome. Following the model of Bennett et al. (2006) for examining physical functioning in
cancer survivors, physical functioning assessments may include objective measures of mobility
such as balance, standing, reaching, and climbing stairs as well as self-reports of mobility and
lifestyle activities such as self-care, daily living, and occupational activities. Besides health-related
outcomes, the potential for harm should be assessed and reported as well. Reporting standards
for harm (Ioannidis et al. 2004) should help to inform both practitioners and the public on the
potential harm of exercise interventions during adjuvant cancer treatment. The long-term follow
up of exercise interventions requires attention, because some health problems linked to adjuvant
cancer treatment are either long-term, such as fatigue or deconditioning, or may be related to
recurrence and mortality, such as weight gain. Future research will need to improve understand-
ing of the ways in which individuals can sustain long-term lifestyle choices.
Regarding the exercise intervention, attention should be concentrated to tailoring type (e.g., aero-
bic or resistance training), dosage (i.e., frequency, intensity, and duration), progression, and con-
text (e.g., centre-based, community-based, or home-based, with either individual or group for-
mat) of exercise prescription to the specific needs of patients.
As described above, the actual training stimulus may substantially deviate from the assigned exer-
cise regimen. If participants allocated to the exercise group do not exercise (non-adherence), and
at the same time participants allocated to the control group do exercise (contamination), the
originally intended trial groups are degenerated into groups with participants who exercise and
those who do not, and moreover in unknown proportions. If those groups are analysed as ‘ran-
domised’, effects may be underestimated. In efficacy trials, investigators need to ensure adher-
ence to the intervention to determine whether exercise interventions in this population work.
Inclusion of sedentary participants only may be a way to deal with contamination issues: seden-
tary women can be expected not to start their own exercise program in the period of adjuvant
cancer treatment. In effectiveness trials, both adherence and contamination should be reported as
Conclusions
123
an outcome measure because poor adherence can render an efficacious intervention as ineffec-
tive.
Finally, further development of the field of exercise intervention will require more attention to
identifying more cost-effective ways to expose patients to the exercise prescription: outside Uni-
versity or cancer centres – inside the communities.
7.2 Implications for practice
The benefit of any exercise intervention is determined not only by its efficacy, but also by the
extent to which it is appropriately adopted and implemented in the community. The exercise
rehabilitation programs established in Germany for women who have been treated for breast
cancer provide a model of how exercise could be offered to breast cancer patients during treat-
ment within a health care context. This model of regular exercise in the rehabilitation of women
with breast cancer could be adopted in other countries. Currently, more than 600 rehabilitation
sports groups, providing exercise classes for women who have been treated for breast cancer,
exist in Germany within the framework of organised sports. A particular strength of these groups
is that they are local, thus allowing participation in structured exercise programs in the vicinity of
participants’ homes, a model which could easily be adapted for women undergoing treatment for
breast cancer.
Currently, being treated for breast cancer is widely regarded as a contraindication for participating
in existing rehabilitation sports groups. However, evidence indicates that there is no increased
risk of injury or lymphedema even during adjuvant cancer treatment, neither in home-based, self-
organised interventions nor in clinical settings, or community-based interventions. Where reha-
bilitation sports groups for breast cancer survivors are accessible, women who are still being
treated for breast cancer should be offered the opportunity to practise sport within one of these
groups. Exercise classes during adjuvant treatment should focus on maintaining exercise levels
among previously active women, and increasing exercise levels among inactive women.
7.2.1 Setting up exercise classes
There are some organisational pre-conditions that have to be ensured: first, there needs to be a
qualified group exercise instructor, to conduct the group exercise sessions for the target group of
women with breast cancer and a physician has to be enlisted to supervise the exercise classes.
Pre-activity participation screening by treating physicians and medical clearance are mandatory
before women can participate in the exercise program. Exercise facilities and sports equipment
such as balls, thera-bands or step boards are usually provided by the sports club. Venues should
be accessible by public transport and preferably barrier-free. An additional room for socializing
Conclusions
124
and informal exchange after exercise sessions would be an asset. Not least, recruitment and reten-
tion of participants is crucial to the success of rehabilitation sports programs. In contrast to exer-
cise programming for post-cardiac rehabilitation, recommending exercise to women diagnosed
with breast cancer is still made reluctantly by physicians. The idea of regular exercise as a benefi-
cial behaviour during breast cancer treatment has to be “sold” to treating physicians, to sports
clubs, and not least to women. Methods of social marketing may contribute to overcome re-
cruitment problems. Further, partnerships could be cultivated with physicians in breast centres/
rehabilitation facilities and with self-help initiatives to facilitate access to breast cancer patients.
One focus in the conduct of exercise classes is on helping sedentary individuals to initiate and
maintain a moderate-intensity exercise regimen. Conducive to this is a well-balanced exercise
program meeting the needs of participants. Women with low levels of fitness might need in-
struction and programs where they can experience enjoyment in movement and success in a
positive, non-competitive environment. This creates a challenge for exercise instructors to come
up with an interesting variety of movement combination, music, and teaching formations. Varied
teaching formations, other than front-facing formation, such as using circles, lines, and partner
groupings, also facilitate social interaction. Emotional and social components contribute to the
success of the exercise program and when programs meet exercise needs but fail to facilitate
social and emotional aspects, they may stagnate. For enjoyment, aerobic or resistance exercise,
flexibility, or balance exercises may be embedded in traditional and non-traditional movement
games that encourage physical activity. Performing movement to music may encourage feeling at
ease and experimenting with new movement activities. Safety is another important aspect of the
exercise program: monitoring exercise intensity does not only assure that the targeted training
dosage is achieved but also gives important feedback on how women are responding to the exer-
cise class. Using a variety of techniques to monitor exercise intensity increases the safety of the
program, e.g., taking exercise pulse and recovery pulse rates in conjunction with rates of per-
ceived exertion.
The art of exercise prescription is the “successful integration of exercise science with behavioural
techniques that result in long-term program compliance” (Balady et al. 2000): Especially for
groups of sedentary women, in addition to the pure exercise stimulus, a behaviour change ap-
proach is essential in order to motivate them to keep exercising. They not only need to feel that
they can carry out physical exercise (i.e., physical exercise self-efficacy), but they also need to feel
that they are fully responsible for initiating and maintaining exercise. Besides disease- and treat-
ment-related barriers, one cause for non-adherence to exercise prescriptions is motivational in
nature. Motivation is a critical variable in producing maintained change, especially with respect to
the post-treatment period, when disease- and treatment-related barriers can be expected to be less
Conclusions
125
relevant. The exercise instructor is challenged here to gain a clearer understanding of how to
facilitate motivation in the context of exercise behaviour.
Bibliography
126
8 Bibliography
Adler RA. Cancer treatment-induced bone loss. Curr Opin Endocrinol Diabetes Obes 14(6): 442-
6; 2007.
Ahlberg K, Ekman T, Gaston-Johansson F, Mock V. Assessment and management of cancer-
related fatigue in adults. Lancet 362: 640-50; 2003.
Ahmed RL, Thomas W, Yee D, Schmitz KH. Randomized controlled trial of weight training and
lymphedema in breast cancer survivors. J Clin Oncol 24: 2765-72; 2006.
Alberta Breast Cancer Program. Patient information: breast cancer treatment overview: [online].
2003. Available from: http://www.albertabreast.com/patientinfo/txoverview.php. Accessed
25.06.2007.
Andrykowski MA, Beacham AO, Jacobsen PB. Prospective, longitudinal study of leisure-time
exercise in women with early-stage breast cancer. Cancer Epidemiol Biomarkers Prev 16: 430-8;
2007.
Antes G. Registering clinical trials is necessary for ethical, scientific and economic reasons. Z
Arztl Fortbild Qualitatssich 98: 422; 2004.
Arent SM, Landers DM, Etnier JL. The effects of exercise on mood in older adults: a meta-
analytic review. J Aging Phys Act 8: 407-430; 2000.
Arndt V, Stegmaier C, Ziegler H, Brenner H. A population-based study of the impact of specific
symptoms on quality of life in women with breast cancer 1 year after diagnosis. Cancer 107:
2496-503; 2006.
Balady G, Berra, KA, Golding LA, Gordon NF, Mahler DA, Myers JN, Sheldahl LM. ACSM’s
guidelines for exercise testing and prescription. 6th ed. Philadelphia: Lippincott Williams & Wil-
kins; 2000.
Bandura A. Self-efficacy: The exercise of control. New York: Worth Publishers; 1997.
Baracos VE. Management of muscle wasting in cancer-associated cachexia: understanding gained
from experimental studies. Cancer 92: 1669-77; 2001.
Barnett GC, Shah M, Redman K, Easton DF, Ponder BA, Pharoah PD. Risk factors for the
incidence of breast cancer: do they affect survival from the disease? J Clin Oncol 26: 3310-6;
2008.
Barsevick AM, Dudley WN, Beck SL. Cancer-related fatigue, depressive symptoms, and func-
tional status: a mediation model. Nurs Res 55: 366-72; 2006.
Battaglini CL. A randomized study on the effects of a prescribed exercise intervention on lean
mass and fatigue changes in breast cancer patients during treatment. PhD [dissertation]. Colo-
rado: University of Northern Colorado; 2004.
Batzler WU, Giersiepen K, Hentschel S, Husman G, Kaatsch P, Kieschke J, Kraywinkel K,
Meyer M, Stabenow R, Stegmeier C. Cancer in Germany 2003-2004: Incidence and Trends. 6th
ed. Berlin: Robert Koch Institute and Association of Population-Based Cancer Registries in
Germany; 2008.
Bellizzi KM, Rowland JH, Jeffery DD, McNeel T. Health behaviors of cancer survivors: examin-
ing opportunities for cancer control intervention. J Clin Oncol 23: 8884-93; 2005.
Bennett JA, Winters-Stone K, Nail L. Conceptualizing and measuring physical functioning in
cancer survivorship studies. Oncol Nurs Forum 33: 41-9; 2006.
Bibliography
127
Bernhard J, Zahrieh D, Castiglione-Gertsch M, Hurny C, Gelber RD, Forbes JF, Murray E,
Collins J, Aebi S, Thurlimann B, Price KN, Goldhirsch A, Coates AS. Adjuvant chemotherapy
followed by goserelin compared with either modality alone: the impact on amenorrhea, hot
flashes, and quality of life in premenopausal patients -- the International Breast Cancer Study
Group Trial VIII. J Clin Oncol 25: 263-70; 2007.
Bernhard J, Zahrieh D, Zhang JJ, Martinelli G, Basser R, Hurny C, Forbes JF, Aebi S, Yeo W,
Thurlimann B, Green MD, Colleoni M, Gelber RD, Castiglione-Gertsch M, Price KN,
Goldhirsch A, Coates AS. Quality of life and quality-adjusted survival (Q-TWiST) in patients
receiving dose-intensive or standard dose chemotherapy for high-risk primary breast cancer. Br J
Cancer 98: 25-33; 2008.
Bertz J, Giersiepen K, Haberland J, Hentschel S, Kaatsch P, Katalinic A, Stabenow R, Stegmaier
C, Ziegler H. Krebs in Deutschland: Häufigkeiten und Trends. 5th ed. Saarbrücken: Association
of Population-Based Cancer Registries in Germany and Robert Koch Institute; 2006.
Blanchard CM, Denniston MM, Baker F, Ainsworth SR, Courneya KS, Hann DM, Gesme DH,
Reding D, Flynn T, Kennedy JS. Do adults change their lifestyle behaviors after a cancer diagno-
sis? Am J Health Behav 27: 246-56; 2003.
Brady MJ, Cella DF, Mo F, Bonomi AE, Tulsky DS, Lloyd SR, Deasy S, Cobleigh M, Shiomoto
G. Reliability and validity of the Functional Assessment of Cancer Therapy-Breast quality-of-life
instrument. J Clin Oncol 15: 974-86; 1997.
Brandberg Y, Michelson H, Nilsson B, Bolund C, Erikstein B, Hietanen P, Kaasa S, Nilsson J,
Wiklund T, Wilking N, Bergh J. Quality of life in women with breast cancer during the first year
after random assignment to adjuvant treatment with marrow-supported high-dose chemotherapy
with cyclophosphamide, thiotepa, and carboplatin or tailored therapy with Fluorouracil, epirubi-
cin, and cyclophosphamide: Scandinavian Breast Group Study 9401. J Clin Oncol 21: 3659-64;
2003.
Brown SA, Guise TA. Cancer treatment-related bone disease. Crit Rev Eukaryot Gene Expr
19(1): 47-60; 2009.
Bullinger M. Erfassung der gesundheitsbezogenen Lebensqualität mit dem SF-36-Health Survey.
Bundesgesundheitsbl 43: 190-7; 2000.
Bullinger M. Assessing health-related quality of life in medicine: an overview over concepts,
methods and applications in international research. Restor Neurol Neurosci 20: 93-101; 2002.
Busse R, Orvain J, Velasco M, Perleth M, Drummond M, Gurtner F, Jørgensen T, Jovell A,
Malone J, Ruther A, Wild C. Best practice in undertaking and reporting Health Technology As-
sessments: Working Group 4 Report. Int J Technol Assess Health Care 18: 361–422; 2002.
Cancer Prevention Research Center. Transtheoretical Model [online]. Available from:
http://www.uri.edu/research/cprc/TTM/detailedoverview.htm. Accessed 16.11.2008.
Caspersen CJ, Powell KE, Christenson GM. Physical activity, exercise, and physical fitness: defi-
nitions and distinctions for health-related research. Public Health Rep 100: 126-31; 1985.
Cella D. The Functional Assessment of Cancer Therapy-Anemia (FACT-An) Scale: a new tool
for the assessment of outcomes in cancer anemia and fatigue. Semin Hematol 34: 13-9; 1997.
Cella D. Factors influencing quality of life in cancer patients: anemia and fatigue. Semin Oncol
25: 43-6; 1998.
Cella DF, Tulsky DS, Gray G, Sarafian B, Linn E, Bonomi A, Silberman M, Yellen SB, Winicour
P, Brannon J. The Functional Assessment of Cancer Therapy scale: development and validation
of the general measure. J Clin Oncol 11: 570-9; 1993.
Bibliography
128
Cheema BS, Gaul CA. Full-body exercise training improves fitness and quality of life in survivors
of breast cancer. J Strength Cond Res 20: 14-21; 2006.
Chia S, Bryce C, Gelmon K. The 2000 EBCTCG overview: a widening gap. Lancet 365: 1665-6;
2005.
Chlebowski RT, Aiello E, McTiernan A. Weight loss in breast cancer patient management. J Clin
Oncol 20: 1128-43; 2002.
Chou R, Helfand M. Challenges in systematic reviews that assess treatment harms. Ann Intern
Med 142: 1090-9; 2005.
Cleeland CS. Symptom burden: multiple symptoms and their impact as patient-reported out-
comes. J Natl Cancer Inst Monogr 37: 16-21; 2007.
Conn VS, Hafdahl AR, Porock DC, McDaniel R, Nielsen PJ. A meta-analysis of exercise inter-
ventions among people treated for cancer. Support Care Cancer 31: 31; 2006.
Convertino VA, Bloomfield SA, Greenleaf JE. An overview of the issues: physiological effects of
bed rest and restricted physical activity. Med Sci Sports Exerc 29: 187-90; 1997.
Cook DJ, Sackett DL, Spitzer WO. Methodologic guidelines for systematic reviews of random-
ized control trials in health care from the Potsdam Consultation on Meta-Analysis. J Clin Epide-
miol 48: 167-71; 1995.
Courneya K, Mackey J, Jones LW. Coping with cancer: can exercise help? Phys Sportsmed 28:
49-73; 2000.
Courneya KS. Exercise interventions during cancer treatment: biopsychosocial outcomes. Exerc
Sport Sci Rev 29: 60-4; 2001.
Courneya KS. Exercise in cancer survivors: an overview of research. Med Sci Sports Exerc 35:
1846-52; 2003.
Courneya KS, Blanchard CM, Laing DM. Exercise adherence in breast cancer survivors training
for a dragon boat race competition: A preliminary investigation. Psychooncology 10: 444-452;
2001.
Courneya KS, Friedenreich CM. Relationship between exercise during treatment and current
quality of life among survivors of breast cancer. J Psychosoc Oncol 15: 35-57; 1997a.
Courneya KS, Friedenreich CM. Relationship between exercise pattern across the cancer experi-
ence and current quality of life in colorectal cancer survivors. J Altern Complement Med 3: 215-
26; 1997b.
Courneya KS, Friedenreich CM. Utility of the theory of planned behavior for understanding
exercise during breast cancer treatment. Psychooncology 8: 112-22; 1999a.
Courneya KS, Friedenreich CM. Physical exercise and quality of life following cancer diagnosis: a
literature review. Ann Behav Med 21: 171-9; 1999b.
Courneya KS, McKenzie DC, Reid RD, Mackey JR, Gelmon K, Friedenreich CM, Ladha AB,
Proulx C, Lane K, Vallance JK, Segal RJ. Barriers to supervised exercise training in a randomized
controlled trial of breast cancer patients receiving chemotherapy. Ann Behav Med 35: 116-22;
2008a.
Courneya KS, Segal RJ, Gelmon K, Reid RD, Mackey JR, Friedenreich CM, Proulx C, Lane K,
Ladha AB, Vallance JK, McKenzie DC. Predictors of supervised exercise adherence during
breast cancer chemotherapy. Med Sci Sports Exerc 40: 1180-7; 2008b.
Cramp F, Daniel J. Exercise for the management of cancer-related fatigue in adults. Cochrane
Database Syst Rev: CD006145; 2008.
Bibliography
129
Daley AJ, Bowden SJ, Rea DW, Billingham L, Carmicheal AR. What advice are oncologists and
surgeons in the United Kingdom giving to breast cancer patients about physical activity? Int J
Behav Nutr Phys Act 5: 46; 2008.
Daling JR, Malone KE, Doody DR, Johnson LG, Gralow JR, Porter PL. Relation of body mass
index to tumor markers and survival among young women with invasive ductal breast carcinoma.
Cancer 92: 720-9; 2001.
de Haes H, Olschewski M, Kaufmann M, Schumacher M, Jonat W, Sauerbrei W. Quality of life
in goserelin-treated versus cyclophosphamide plus methotrexate and fluorouracil-treated
premenopausal and perimenopausal patients with node-positive, early breast cancer: the Zoladex
Early Breast Cancer Research Association Trialists Group. J Clin Oncol 21: 4510-6; 2003.
de Jong N, Courtens AM, Abu-Saad HH, Schouten HC. Fatigue in patients with breast cancer
receiving adjuvant chemotherapy: a review of the literature. Cancer Nurs 25: 283-97; 2002.
Deeks JJ, Higgins JPT, Altman DJ. Analysing and presenting results. Cochrane Handbook for
Systematic Reviews of Interventions 4.2.6 [updated September 2006]. In: The Cochrane Library,
Issue 4. Chichester, UK: John Wiley & Sons; 2006.
Del Mastro L, Costantini M, Morasso G, Bonci F, Bergaglio M, Banducci S, Viterbori P, Conte P,
Rosso R, Venturini M. Impact of two different dose-intensity chemotherapy regimens on psycho-
logical distress in early breast cancer patients. Eur J Cancer 38: 359-66; 2002.
Demark-Wahnefried W, Peterson B, McBride C, Lipkus I, Clipp E. Current health behaviors and
readiness to pursue life-style changes among men and women diagnosed with early stage prostate
and breast carcinomas. Cancer 88: 674-84; 2000.
Demark-Wahnefried W, Aziz NM, Rowland JH, Pinto BM. Riding the crest of the teachable
moment: promoting long-term health after the diagnosis of cancer. J Clin Oncol 23: 5814-30;
2005.
Demark-Wahnefried W, Peterson BL, Winer EP, Marks L, Aziz N, Marcom PK, Blackwell K,
Rimer BK. Changes in weight, body composition, and factors influencing energy balance among
premenopausal breast cancer patients receiving adjuvant chemotherapy. J Clin Oncol 19: 2381-9;
2001.
Demark-Wahnefried W, Rimer BK, Winer EP. Weight gain in women diagnosed with breast
cancer. J Am Diet Assoc 97: 519-26; 1997.
Demiot C, Dignat-George F, Fortrat JO, Sabatier F, Gharib C, Larina I, Gauquelin-Koch G,
Hughson R, Custaud MA. WISE 2005: chronic bed rest impairs microcirculatory endothelium in
women. Am J Physiol Heart Circ Physiol 293: 3159-64; 2007.
DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 7: 177-88; 1986.
Deschenes MR, Kraemer WJ. Performance and physiologic adaptations to resistance training.
Am J Phys Med Rehabil 81: S3-16; 2002.
Dibble SL, Isreal J, Nussey B, Casey K, Luce J. Delayed chemotherapy-induced nausea in women
treated for breast cancer. Oncol Nurs Forum 30: E40-7; 2003.
Dimeo FC. Effects of exercise on cancer-related fatigue. Cancer 92: 1689-93; 2001.
Dishman RK. The measurement conundrum in exercise adherence research. Med Sci Sports
Exerc. 26: 1382-90; 1994.
Dishman RK, Berthoud HR, Booth FW, Cotman CW, Edgerton VR, Fleshner MR, Gandevia
SC, Gomez-Pinilla F, Greenwood BN, Hillman CH, Kramer AF, Levin BE, Moran TH, Russo-
Neustadt AA, Salamone JD, Van Hoomissen JD, Wade CE, York DA, Zigmond MJ. Neurobiol-
ogy of exercise. Obesity 14: 345-56; 2006.
Bibliography
130
Dodd MJ, Miaskowski C. The PRO-SELF-Program: a self-care intervention program for patients
receiving cancer treatment. Semin Oncol Nurs 16: 300-8; 2000.
Dodd MJ, Miaskowski C, Lee KA. Occurrence of symptom clusters. J Natl Cancer Inst Monogr:
76-8; 2004.
Dodd MJ, Miaskowski C, Paul SM. Symptom clusters and their effect on the functional status of
patients with cancer. Oncol Nurs Forum 28: 465-70; 2001.
Early Breast Cancer Trialists’ Collaborative Group. Effects of chemotherapy and hormonal ther-
apy for early breast cancer on recurrence and 15-year survival: an overview of the randomised
trials. Lancet 365: 1687-717; 2005.
Egger M, Jüni P, Bartlett C, Holenstein F, Sterne J. How important are comprehensive literature
searches and the assessment of trial quality in systematic reviews? Empirical study [online].
Health Technology Assessment 7(1). 2003. Available from:
http://www.ncchta.org/fullmono/mon701.pdf. Accessed 03.08.2009.
Emslie C, Whyte F, Campbell A, Mutrie N, Lee L, Ritchie D, Kearney N. “I wouldn’t have been
interested in just sitting round a table talking about cancer”: exploring the experiences of women
with breast cancer in a group exercise trial. Health Educ Res 22(6):827-38; 2007.
European Commission. The European model of sport: consultation document of DG X [online].
Brussels: European Commission; 1989. Available from:
http://ec.europa.eu/sport/library/doc/a/doc248_en.pdf. Accessed 18.03.2009.
Fairclough DL, Fetting JH, Cella D, Wonson W, Moinpour CM. Quality of life and quality ad-
justed survival for breast cancer patients receiving adjuvant therapy. Eastern Cooperative Oncol-
ogy Group (ECOG). Qual Life Res 8: 723-31; 1999.
Fairey AS, Courneya KS, Field CJ, Mackey JR. Physical exercise and immune system function in
cancer survivors: a comprehensive review and future directions. Cancer 94: 539-51; 2002.
Földi E, Földi M, Clodius L. The lymphedema chaos: A lancet. Ann Plast Surg 22: 505-15; 1989.
Friedenreich CM, Courneya KS. Exercise as rehabilitation for cancer patients. Clin J Sport Med 6:
237-44; 1996.
Gapstur RL. Symptom burden: a concept analysis and implications for oncology nurses. Oncol
Nurs Forum 34(3): 673-680; 2007.
Galvao DA, Newton RU. Review of exercise intervention studies in cancer patients. J Clin Oncol
23: 899-909; 2005.
Giersiepen K, Heitmann C, Janhsen K, Lang C. Brustkrebs. Gesundheitsberichterstattung des
Bundes: Heft 25. Berlin: Robert Koch-Institut; 2005.
Golder S, Loke Y, McIntosh HM. Room for improvement? A survey of the methods used in
systematic reviews of adverse effects [online]. BMC Med Res Methodol 6: 3; 2006. Available
from: http://www.biomedcentral.com/1471-2288/6/3. Accessed 18.03.2009.
Grimison PS, Stockler MR. Quality of life and adjuvant systemic therapy for early-stage breast
cancer. Expert Rev Anticancer Ther 7: 1123-1134; 2007.
Groenvold M, Fayers PM, Petersen MA, Sprangers MA, Aaronson NK, Mouridsen HT. Breast
cancer patients on adjuvant chemotherapy report a wide range of problems not identified by
health-care staff. Breast Cancer Res Treat 13: 13; 2006.
Große-Wächter E, Ledvina I, Schüle K, Uhlenbruck G, Markes M, Ebel H, Koenig G. Bewe-
gung uns Sport in der Krebsnachsorge [online]. Duisburg: Landessportbund Nordrhein-
Westfalen; 2004. Available from: http://www.wir-im-sport.de/ templates/
Bibliography
131
dokukategorien/dokumanagement/psdoc/file/77/2004_04_02434d15ad599bf.pdf. Accessed
18.03.2009.
Hamilton CJ, Swan VJ, Jamal SA. The effects of exercise and physical activity participation on
bone mass and geometry in postmenopausal women: a systematic review of pQCT studies. Os-
teoporos Int: 2009 Jun 6. [Epub ahead of print]
Haskell WL, Lee IM, Pate RR, Powell KE, Blair SN, Franklin BA, Macera CA, Heath GW,
Thompson PD, Bauman A. Physical activity and public health: updated recommendation for
adults from the American College of Sports Medicine and the American Heart Association. Med
Sci Sports Exerc 39: 1423-34; 2007.
Hausauer AK, Keegan TH, Chang ET, Glaser SL, Howe H, Clarke CA. Recent trends in breast
cancer incidence in US white women by county-level urban/rural and poverty status [online].
BMC Med 7:31; 2009. Available from: http://www.biomedcentral.com/1741-7015/7/31. Ac-
cessed 20.08.2009.
Havas E, Parviainen T, Vuorela J, Toivanen J, Nikula T, Vihko V. Lymph flow dynamics in exer-
cising human skeletal muscle as detected by scintography. J. Physiol 504: 233-9; 1997.
Hedges LV, Olkin I. Statistical Methods for Meta-Analysis. Orlando: Academic Press; 1985.
Herrero F, San Juan AF, Fleck SJ, Foster C, Lucia A. Effects of detraining on the functional
capacity of previously trained breast cancer survivors. Int J Sports Med 28: 257-64; 2007.
Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 21: 1539-58;
2002.
Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses.
Bmj 327: 557-60; 2003.
Higgins J, Green S, editors. Cochrane handbook for systematic reviews of interventions 4.2.6
[updated September 2006]. In: The Cochrane Library, Issue 4. Chichester, UK: John Wiley &
Sons; 2006.
Holmes MD, Chen WY, Feskanich D, Kroenke CH, Colditz GA.. Physical activity and survival
after breast cancer diagnosis. Jama 293: 2479-86; 2005.
Hunt-Shanks TT, Blanchard CM, Baker F, Hann D, Roberts CS, McDonald J, Livingston M,
Witt C, Ruiterman J, Ampela R, Kaw OC. Exercise use as complementary therapy among breast
and prostate cancer survivors receiving active treatment: examination of exercise intention. Integr
Cancer Ther 5: 109-16; 2006.
Hunter GR, McCarthy JP, Bamman MM. Effects of resistance training on older adults. Sports
Med 34: 329-48; 2004.
Hürny C, Bernhard J, Coates AS, Castiglione-Gertsch M, Peterson HF, Gelber RD, Forbes JF,
Rudenstam CM, Simoncini E, Crivellari D, Goldhirsch A, Senn HJ. Impact of adjuvant therapy
on quality of life in women with node-positive operable breast cancer. International Breast Can-
cer Study Group. Lancet 347: 1279-84; 1996.
Ioannidis JP, Evans SJ, Gotzsche PC, O’Neill RT, Altman DG, Schulz K, Moher D. Better re-
porting of harms in randomized trials: an extension of the CONSORT statement. Ann Intern
Med 141: 781-8; 2004.
Irwin ML, Ainsworth BE. Physical activity interventions following cancer diagnosis: methodo-
logical challenges to delivery and assessment. Cancer Invest 22: 30-50; 2004.
Irwin ML, Crumley D, McTiernan A, Bernstein L, Baumgartner R, Gilliland FD, Kriska A, Bal-
lard Barbash R. Physical activity levels before and after a diagnosis of breast carcinoma: the
Health, Eating, Activity, and Lifestyle (HEAL) study. Cancer 97: 1746-57; 2003.
Bibliography
132
Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ, McQuay HJ. Assess-
ing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials
17: 1-12; 1996.
Jakicic JM, Wing RR, Butler BA, Robertson RJ. Prescribing exercise in multiple short bouts ver-
sus one continuous bout: effects on adherence, cardiorespiratory fitness, and weight loss in
overweight women. Int J Obes Relat Metab Disord 19: 893-901; 1995.
Johansson K, Tibe K, Weibull A, Newton RC. Low intensity resistance exercise for breast cancer
patients with arm lymphedema with or without compression sleeve. Lymphology 38: 167-80;
2005.
Jones LW, Courneya KS, Peddle C, Mackey JR. Oncologists’ opinions towards recommending
exercise to patients with cancer: a Canadian national survey. Support Care Cancer 13: 929-37;
2005.
Jones LW, Haykowsky MJ, Swartz JJ, Douglas PS, Mackey JR. Early breast cancer therapy and
cardiovascular injury. J Am Coll Cardiol 50: 1435-41; 2007.
Kelley GA, Kelley KS, Tran ZV. Resistance training and bone mineral density in women: a meta-
analysis of controlled trials. Am J Phys Med Rehabil 80: 65-77; 2001.
King AC. Clinical and community interventions to promote and support physical activity partici-
pation. In: Dishman R. Advances in Exercise Adherence. Champaign: Human Kinetics; 183-212;
1994.
Knols R, Aaronson NK, Uebelhart D, Fransen J, Aufdemkampe G. Physical exercise in cancer
patients during and after medical treatment: a systematic review of randomized and controlled
clinical trials. J Clin Oncol 23: 3830-42; 2005.
Kortebein P, Ferrando A, Lombeida J, Wolfe R, Evans WJ. Effect of 10 days of bed rest on
skeletal muscle in healthy older adults. Jama 297: 1772-4; 2007.
Kuerer HM, Julian TB, Strom EA, Lyerly HK, Giuliano AE, Mamounas EP, Vicini FA.. Acceler-
ated partial breast irradiation after conservative surgery for breast cancer. Ann Surg 239: 338-51;
2004.
Kumar NB, Cantor A, Allen K, Cox CE. Android obesity at diagnosis and breast carcinoma
survival: Evaluation of the effects of anthropometric variables at diagnosis, including body com-
position and body fat distribution and weight gain during life span, and survival from breast
carcinoma. Cancer 88: 2751-7; 2000.
Kuramoto AK, Payne VG. Predicting muscular strength in women: a preliminary study. Res Q
Exerc Sport 66: 168-72; 1995.
Laine C, Horton R, De Angelis C, Drazen JM, Frizelle FA, Godlee F, Haug C, Hebert PC,
Kotzin S, Marusic A, Sahni P, Schroeder TV, Sox HC, Van Der Weyden MB, Verheugt FW.
Clinical trial registration: looking back and moving ahead. Med J Aust 186: 612-3; 2007.
Land SR, Kopec JA, Yothers G, Anderson S, Day R, Tang G, Ganz PA, Fisher B, Wolmark N.
Health-related quality of life in axillary node-negative, estrogen receptor-negative breast cancer
patients undergoing AC versus CMF chemotherapy: findings from the National Surgical Adju-
vant Breast and Bowel Project B-23. Breast Cancer Res Treat 86: 153-64; 2004.
Landessportbund Nordrhein-Westfalen: Sport in der Krebsnachsorge [homepage on the Inter-
net]. Duisburg: Landessportbund Nordrhein Westfalen; 2009. Available from: http://www.wir-
im-sport.de.
Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics
33: 159–74; 1977.
Bibliography
133
Lankester KJ, Phillips JE, Lawton PA. Weight gain during adjuvant and neoadjuvant chemother-
apy for breast cancer: an audit of 100 women receiving FEC or CMF chemotherapy. Clin Oncol
14: 64-7; 2002.
Lawlor DA, Hopker SW. The effectiveness of exercise as an intervention in the management of
depression: systematic review and meta-regression analysis of randomised controlled trials. Bmj.
322: 763-7; 2001.
Lemmer JT, Ivey FM, Ryan AS, Martel GF, Hurlbut DE, Metter JE, Fozard JL, Fleg JL, Hurley
BF. Effect of strength training on resting metabolic rate and physical activity: age and gender
comparisons. Med Sci Sports Exerc 33: 532-41; 2001.
Lennon SL, Quindry J, Hamilton KL, French J, Staib J, Mehta JL, Powers SK. Loss of exercise-
induced cardioprotection after cessation of exercise. J Appl Physiol 96: 1299-305; 2004.
Lipsey MW, Wilson DB. Practical Meta-Analysis. Thousands Oaks: Sage Publications; 2001.
Applied Social Research Methods Series, Volume 49.
Macquart-Moulin G, Viens P, Palangie T, Bouscary ML, Delozier T, Roche H, Janvier M, Fabbro
M, Moatti JP. High-dose sequential chemotherapy with recombinant granulocyte colony-
stimulating factor and repeated stem-cell support for inflammatory breast cancer patients: does
impact on quality of life jeopardize feasibility and acceptability of treatment? J Clin Oncol 18:
754-64; 2000.
MacVicar MG, Winningham ML. Promoting the functional capacity of cancer patients. Cancer
Bull 38: 235-9; 1986.
Markes M, Brockow T, Resch KL. Exercise for women receiving adjuvant therapy for breast
cancer. Cochrane Database Syst Rev: CD005001; 2006.
Markman M. Progress in preventing chemotherapy-induced nausea and vomiting. Cleve Clin J
Med 69: 609-17; 2002.
Martin M, Pienkowski T, Mackey J, Pawlicki M, Guastalla JP, Weaver C, Tomiak E, Al-Tweigeri
T, Chap L, Juhos E, Guevin R, Howell A, Fornander T, Hainsworth J, Coleman R, Vinholes J,
Modiano M, Pinter T, Tang SC, Colwell B, Prady C, Provencher L, Walde D, Rodriguez-Lescure
A, Hugh J, Loret C, Rupin M, Blitz S, Jacobs P, Murawsky M, Riva A, Vogel C. Adjuvant do-
cetaxel for node-positive breast cancer. N Engl J Med 352: 2302-13; 2005.
McInnes JA, Knobf MT. Weight gain and quality of life in women treated with adjuvant chemo-
therapy for early-stage breast cancer. Oncol Nurs Forum 28; 2001.
McKenzie DC, Kalda AL. Effect of upper extremity exercise on secondary lymphedema in breast
cancer patients: a pilot study. J Clin Oncol 21: 463-6; 2003.
McNeely ML, Campbell KL, Rowe BH, Klassen TP, Mackey JR, Courneya KS. Effects of exer-
cise on breast cancer patients and survivors: a systematic review and meta-analysis. Cmaj 175: 34-
41; 2006.
McTiernan A. Physical activity after cancer: physiologic outcomes. Cancer Invest 22: 68-81; 2004.
Miller M, Kearney N. Chemotherapy-related nausea and vomiting - past reflections, present prac-
tice and future management. Eur J Cancer Care 13: 71-81; 2004.
Mock V, Olsen M. Current management of fatigue and anemia in patients with cancer. Semin
Oncol Nurs 19: 36-41; 2003.
Moher D, Jadad A, Nichol G. Assessing the quality of randomised controlled trials: an annotated
bibliography of checklists. Control Clin Trials 16: 62-73; 1995.
Mulrow CD. Rationale for systematic reviews. Bmj 309: 597-9; 1994.
Bibliography
134
National Comprehensive Cancer Network. NCCN Clinical practice guidelines in oncology: can-
cer-related fatigue [online]. Washington: National Comprehensive Cancer Network; 2007. Avail-
able from: http://www.nccn.org. Accessed 21.02.2007.
NHS Centre for Reviews and Dissemination. Undertaking systematic reviews of research on
effectiveness. 2nd ed. CRD Report Number 4. York: University of York; 2001.
Noonan V, Dean E. Submaximal exercise testing: clinical application and interpretation. Phys
Ther 80(8):782-807. 2000.
Nutbeam D. Health literacy as a public health goal: a challenge for contemporary health educa-
tion and communication strategies into the 21st century. Health Promot Int 15: 259-67; 2000.
O’Connor PJ. Evaluation of four highly cited energy and fatigue mood measures. J Psychosom
Res 57: 435-41; 2004.
Ohira T, Schmitz KH, Ahmed RL, Yee D. Effects of weight training on quality of life in recent
breast cancer survivors: the Weight Training for Breast Cancer Survivors (WTBS) study. Cancer
106: 2076-83; 2006.
Oldervoll LM, Kaasa S, Hjermstad MJ, Lund JA, Loge JH. Physical exercise results in the im-
proved subjective well-being of a few or is effective rehabilitation for all cancer patients? Eur J
Cancer 40: 951-62; 2004.
Parker KP, Kimble LP, Dunbar SB, Clark PC. Symptom interactions as mechanisms underlying
symptom pairs and clusters. J Nurs Scholarsh 37: 209-15; 2005.
Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin 55: 74-
108; 2005.
Parkin DM, Fernandez LM. Use of statistics to assess the global burden of breast cancer. Breast J
12: S70-80; 2006.
Partridge AH, Burstein HJ, Winer EP. Side effects of chemotherapy and combined chemohor-
monal therapy in women with early-stage breast cancer. J Natl Cancer Inst Monogr: 135-42;
2001.
Patient Reported Outcomes Measurement Information System Network. Patient Reported Out-
comes Measurement Information System (PROMIS) Domain Framework [online]. 2008. Avail-
able from: http://www.nihpromis.org/. Accessed 9.9.2008.
Peppercorn J, Herndon J, Kornblith AB, Peters W, Ahles T, Vredenburgh J, Schwartz G, Shpall
E, Hurd DD, Holland J, Winer E. Quality of life among patients with Stage II and III breast
carcinoma randomized to receive high-dose chemotherapy with autologous bone marrow sup-
port or intermediate-dose chemotherapy: results from Cancer and Leukemia Group B 9066.
Cancer 104: 1580-9; 2005.
Pinto BM, Trunzo JJ. Health behaviors during and after cancer diagnosis. Cancer 104 (11 Suppl):
2614-23. 2005.
Pinto BM, Eakin E, Maruyama NC. Health behavior changes after a cancer diagnosis: what do
we know and where do we go from here? Ann Behav Med 22 (1): 38-52; 2000.
Piper BF, Dibble SL, Dodd MJ, Weiss MC, Slaughter RE, Paul SM. The revised Piper Fatigue
Scale: psychometric evaluation in women with breast cancer. Oncol Nurs Forum 25: 677-84;
1998.
Piper BF, Lindsey AM, Dodd MJ, Ferketich SL, Paul SM, Weller S. The development of an in-
strument to measure the subjective dimension of fatigue. In: Funk SG, Tornquist EM, Cham-
pagne MT. Key aspects of comfort: Management of pain, fatigue, and nausea. Springer Publish-
ing Company; 199-208; 1989.
Bibliography
135
Poortmans P. Evidence based radiation oncology: breast cancer. Radiother Oncol 84: 84-101;
2007.
Prochaska JO, DiClemente CC. Stages and processes of self-change of smoking: toward an inte-
grative model of change. J Consult Clin Psychol 51: 390-5; 1983.
Ramaswamy B, Shapiro CL. Osteopenia and osteoporosis in women with breast cancer. Semin
Oncol 30: 763-75; 2003.
Rapkin BD, Schwartz CE. Toward a theoretical model of quality-of-life appraisal: Implications of
findings from studies of response shift [online]. Health Qual Life Outcomes (14); 2004. Available
from: http://www.hqlo.com/content/2/1/14. Accessed 28.07.2009.
Ravdin PM, Cronin KA, Howlader N, Berg CD, Chlebowski RT, Feuer EJ, Edwards BK, Berry
DA: The decrease in breast-cancer incidence in 2003 in the United States. N Engl J Med
356:1670-1674; 2007.
Rhodes RE, Courneya KS, Bobick TM. Personality and exercise participation across the breast
cancer experience. Psychooncology 10: 380-8; 2001.
Rikli R, Jones J. Assessing physical performance in independent older adults: Issues and guide-
lines. J Aging Phys Act 5: 244-67; 1997.
Rogers LQ, Courneya KS, Shah P, Dunnington G, Hopkins-Price P. Exercise stage of change,
barriers, expectations, values and preferences among breast cancer patients during treatment: a
pilot study. Eur J Cancer Care 16: 55-66; 2007.
Rogers LQ, Courneya KS, Verhulst S, Markwell S, Lanzotti V, Shah P. Exercise barrier and task
self-efficacy in breast cancer patients during treatment. Support Care Cancer 9: 9; 2005a.
Rogers LQ, Matevey C, Hopkins-Price P, Shah P, Dunnington G, Courneya KS. Exploring social
cognitive theory constructs for promoting exercise among breast cancer patients. Cancer Nurs
27: 462-73; 2004.
Rogers LQ, Shah P, Dunnington G, Greive A, Shanmugham A, Dawson B, Courneya KS. Social
cognitive theory and physical activity during breast cancer treatment. Oncol Nurs Forum 32: 807-
15; 2005b.
Ruetten A, Abu-Omar K, Lampert T, Ziese T. Körperliche Aktivität [online]. Berlin: Robert
Koch Institute. Gesundheitsberichterstattung des Bundes. Heft 26; 2005. Available from:
http://www.rki.de. Accessed 21.08.2009.
Sackett DL, Rosenberg WM, Gray JA, Haynes RB, Richardson WS. Evidence based medicine:
what it is and what it isn’t. Bmj 312: 71-2; 1996.
Schmitz KH, Ahmed RL, Hannan PJ, Yee D. Safety and efficacy of weight training in recent
breast cancer survivors to alter body composition, insulin, and insulin-like growth factor axis
proteins. Cancer Epidemiol Biomarkers Prev 14: 1672-80; 2005.
Schmitz KH, Holtzman J, Courneya KS, Masse LC, Duval S, Kane R. Controlled physical activity
trials in cancer survivors: a systematic review and meta-analysis. Cancer Epidemiol Biomarkers
Prev 14: 1588-95; 2005.
Schwartz AL. Cancer. In: Durstine JL, Moore JE. ACSM’s exercise management for persons with
chronic diseases and disabilities. Champaign: Human Kinetics Inc: 166-172; 1997.
Schwartz CE, Rapkin BD. Reconsidering the psychometrics of quality of life assessment in light
of response shift and appraisal [online]. Health Qual Life Outcomes 2 (16); 2004. Available from:
http://www.hqlo.com/content/2/1/16. Accessed 28.07.2009.
Servaes P, Verhagen C, Bleijenberg G. Fatigue in cancer patients during and after treatment:
prevalence, correlates and interventions. Eur J Cancer 38: 27-43; 2002.
Bibliography
136
Shapiro CL, Recht A. Side effects of adjuvant treatment of breast cancer. N Engl J Med 344:
1997-2008; 2001.
Shrout PE, Fleiss JL. Intraclass correlations: Uses in assessing rater reliability. Psychol Bulletin
86:420-7; 1979.
Singletary SE, Connolly JL. Breast cancer staging: working with the sixth edition of the AJCC
Cancer Staging Manual. CA Cancer J Clin 56: 37-47; 2006.
Sivesind D, Baile WF. The psychologic distress in patients with cancer. Nurs Clin North Am 36:
809-25; 2001.
So HS, Kim IS, Yoon JH, Park OJ. Effects of aerobic exercise using a flex-band on physical
functions and body image in women undergoing radiation therapy after a mastectomy. Taehan
Kanho Hakhoe Chi 36: 1111-22; 2006.
Somerset W, Stout SC, Miller AH, Musselman D. Breast cancer and depression. Oncology 18:
1021-34; 2004.
Stasi R, Abriani L, Beccaglia P, Terzoli E, Amadori S. Cancer-related fatigue: evolving concepts
in evaluation and treatment. Cancer 98: 1786-801; 2003.
Stevinson C, Lawlor DA, Fox K. Exercise interventions for cancer patients: systematic review of
controlled trials. Cancer Causes & Control 15: 1035-56; 2004.
Stricker CT, Drake D, Hoyer KA, Mock V. Evidence-based practice for fatigue management in
adults with cancer: exercise as an intervention. Oncol Nurs Forum 31: 963-76; 2004.
Tavio M, Milan I, Tirelli U. Cancer-related fatigue (review). Int J Oncol 21: 1093-9; 2002.
Terwee CB, van der Slikke RM, van Lummel RC, Benink RJ, Meijers WG, de Vet HC. Self-
reported physical functioning was more influenced by pain than performance-based physical
functioning in knee-osteoarthritis patients. J Clin Epidemiol 59: 724-31; 2006.
Thompson PD. Cardiovascular Risks of Exercise: Avoiding Sudden Death and Myocardial In-
farction. Phys Sportsmed 29; 2001.
Turner CH, Robling AG. Exercises for improving bone strength. Br J Sports Med 39: 188-9;
2005.
United States Department of Health and Human Services. Centers for Disease Control and Pre-
vention. National Center for Chronic Disease Prevention and Health Promotion. Division of
Nutrition and Physical Activity: Physical Activity and Health: A Report of the Surgeon General.
Atlanta: US Supertendent of Documents; 1996.
vanTulder MW, Assendelft WJ, Koes BW, Bouter LM. Method guidelines for systematic reviews
in the Cochrane Collaboration Back Review Group for Spinal Disorders. Spine 22: 2323-30;
1997.
Verhagen AP, de Vet HC, de Bie RA, Boers M, van den Brandt PA. The art of quality assessment
of RCTs included in systematic reviews. J Clin Epidemiol 54: 651-4; 2001.
Verhagen AP, de Vet HC, de Bie RA, Kessels AG, Boers M, Bouter LM, Knipschild PG. The
Delphi list: a criteria list for quality assessment of randomized clinical trials for conducting sys-
tematic reviews developed by Delphi consensus. J Clin Epidemiol 51: 1235-41; 1998.
Watson T, Mock V. Exercise as an intervention for cancer-related fatigue. Phys Ther 84: 736-743;
2004.
WHOQOL group. WHOQOL-BREF: Introduction, administration, scoring and generic version
of the assessment [online]. Geneva: WHO; 1996. Available from:
http://www.who.int/mental_health/media/en/76.pdf. Accessed 28.07.2009.
Bibliography
137
Winningham M, MacVicar M, Burke C. Exercise for cancer patients: guidelines and precautions.
Phys Sportsmed 14: 125-34; 1986.
Winningham ML. Strategies for managing cancer-related fatigue syndrome: a rehabilitation ap-
proach. Cancer 92: 988-97; 2001.
Wolff I, van Croonenborg JJ, Kemper HC, Kostense PJ, Twisk JW. The effect of exercise train-
ing programs on bone mass: a meta-analysis of published controlled trials in pre- and postmeno-
pausal women. Osteoporos Int 9: 1-12; 1999.
Woods JA, Davis JM, Smith JA, Nieman DC. Exercise and cellular innate immune function. Med
Sci Sports Exerc 31: 57-66; 1999.
World Health Organization. WHO definition of Health [online]; Geneva: WHO; 1948. Available
from: http://www.who.int/about/definition/en/print.html. Accessed 28.07.2009.
World Health Organization. Ottawa Charter for Health Promotion [online]; Geneva: WHO;
1986. Available from: http://www.who.int/hpr/NPH/docs/ottawa_charter_hp.pdf. Accessed
28.07.2009.
World Health Organization. Mental health: strengthening mental health promotion [online].
Geneva: WHO; 2007. Available from: http://www.who.int/mediacentre/factsheets/fs220/en/.
Accessed 03.11.2008.
Appendix 1
138
9 Appendices
9.1 Appendix 1 – Protocol
BACKGROUND
Breast cancer detection and management have undergone dramatic changes over the past three
decades. Women are increasingly diagnosed with early stage disease leaving them with treatment
choices ranging from breast conserving options to mastectomy (Newman 2003). With the major-
ity of breast cancers diagnosed at an early stage, treatment is focused on cure and the prevention
of relapse due to micrometastatic disease. Because systemic adjuvant therapy effectively prevents
or delays some relapses and deaths in early-stage disease, this treatment approach has become
standard practice throughout most of the Western world (Hudis 2003). The mainstay of care is
local therapy, consisting of surgery (with or without radiotherapy), along with systemic adjuvant
therapy, which includes chemotherapy (cytotoxic agents), hormonal therapy or a combination of
these treatments.
Besides these major advances in managing early and locally advanced breast cancer, patients still
have to deal with severe side effects and psychological distress during adjuvant therapy and this
has a substantial impact on their quality of life.
• Side effects of radiotherapy can be distinguished in common short-term side effects such as
fatigue and skin erythema and in long-term side effects, including lymphedema, cardiac and
pulmonary toxicities and brachial plexopathy (Truong 2004).
• Chemotherapy is associated with short-term side effects such as emesis, nausea, stomatitis,
alopecia, myelosuppression, thromboembolism, myalgias, neuropathy and fatigue. Long-term
side effects are premature menopause, weight gain, fatigue, cardiac dysfunction, and cognitive
dysfunction (Partridge 2001).
• Patients receiving radiotherapy or chemotherapy furthermore report anxiety and depression
prior to, during and after therapy due to treatment side effects (Spiegel 1997).
• Adjuvant hormonal therapy produces symptoms secondary to oestrogen withdrawal, such as
hot flushes, bone demineralization and psychosexual effects (Rutqvist 2004).
Although research is producing increasingly hopeful insights into the causes and cures for cancer,
efforts to manage the side effects of adjuvant therapy have not kept pace (Patrick 2003). Exercise
interventions may be effective in managing some of these side effects such as fatigue, weight
gain, emotional distress, nausea/vomiting and premature menopause. The rationale for exercise
interventions is described in the following paragraphs:
Cancer-related fatigue is defined by the National Comprehensive Cancer Network as a “persis-
tent, subjective sense of tiredness related to cancer or cancer treatment that interferes with usual
functioning” (NCCN 2003). Fatigue results in substantial physical, psychosocial, cognitive and
socioeconomic consequences (Holley 2000). Acute and chronic radiotherapy-related fatigue oc-
Appendix 1
139
curs in up to 80% and 30% of patients respectively (Jereczek-Fossa 2001). During and after adju-
vant chemotherapy prevalence of fatigue is high and fluctuating (de Jong 2002) with a frequency
of 60 to 90% (Feyer 2001). Fatigue is also associated with factors such as depression, impaired
quality of sleep or pain (de Jong 2002). The rationale supporting exercise interventions for can-
cer-related fatigue is based on the proposition that the combined effects of toxic treatment and a
decreased level of activity during treatment cause a reduction in the capacity for physical per-
formance. Patients must therefore use greater effort and expend more energy to perform daily
activities leading to fatigue (NCCN 2003). Physical exercise training programmes may increase
functional capacity leading to reduced effort and decreased fatigue.
Weight gain is also common among breast cancer patients receiving adjuvant chemotherapy with
gains ranging from 0 to 22 kg influenced by menopausal status, nodal status and the type, dura-
tion and intensity of treatment (Demark 1997). Weight gain not only has a similar profound ef-
fect on quality of life as fatigue, but in addition represents a potentially poor prognostic factor
with higher relapse rates and poorer survival (Camoriano 1990). Evidence suggests that overeat-
ing is not a cause of weight gain among breast cancer patients who receive chemotherapy but the
result of reduced physical activity. Chemotherapy induced weight gain shows the distinctive pat-
tern of sarcopenic obesity, i.e., weight gain in the presence of lean tissue loss or absence of lean
tissue gain (Demark 2001). The development of sarcopenic obesity with evidence of reduced
physical activity supports the need for interventions focused on exercise, especially resistance
training.
Women treated for breast cancer also frequently experience higher levels of emotional distress
than the general population (Spiegel 1997). The rationale for considering exercise as an interven-
tion to reduce distress in women receiving adjuvant therapy for breast cancer is based upon the
literature that has demonstrated ameliorating effects of exercise on these problems: results of
studies with non-cancer populations indicate that aerobic exercise training has antidepressant and
anxiolytic effects and protects against harmful consequences of stress (Salmon 2001). There is
evidence that cognitive dysfunction may also occur in women receiving adjuvant chemotherapy
for breast cancer (O'Shaughnessy 2003; Rugo 2003; Tchen 2003). A meta-analytic study con-
ducted to examine the hypothesis that aerobic fitness training enhances the cognitive vitality of
healthy but sedentary older adults indicated that fitness training has robust benefits for cognition
(Colcombe 2003).
The incidence and severity of nausea and vomiting in patients receiving chemotherapy are af-
fected by numerous factors including the specific chemotherapeutic agents used, their dosage, the
schedule and route of administration and individual patient variability. Although standard chemo-
Appendix 1
140
therapy regimens for breast cancer are considered as mildly to moderately emetogenic (Dibble
2003), nausea and vomiting occur with these regimens: 73%- 82% of women receiving chemo-
therapy for breast cancer experience nausea (Dibble 2003) and if emesis is severe it can lead in its
turn to anticipatory nausea and vomiting (Morrow 1998). Winningham et al. suggest that aerobic
exercise may serve as a potential intervention for controlling or mitigating chemotherapy induced
nausea (Winningham 1988).
The role of exercise in breast cancer has been examined in retrospective, cross-sectional and
prospective studies (Courneya 2001; Courneya 2002; Courneya 2003; Pinto 1999) with the major-
ity of research focused on rehabilitation and health promotion in women who have completed
cancer treatment. This review aims to evaluate the role of exercise in managing common side
effects of adjuvant therapy for breast cancer.
OBJECTIVES
To systematically assess the effectiveness and safety of aerobic and/or muscular endurance exer-
cises – administered during adjuvant radiotherapy, chemotherapy or hormonal therapy – on
managing the following side effects in women with non-metastatic (stages I to III) breast cancer:
• fatigue
• weight gain
• emotional distress
• nausea and/or vomiting
• muscle weakness
• supression of immune functioning.
CRITERIA FOR CONSIDERING STUDIES FOR THIS REVIEW
Types of studies
Randomized and/or non-randomized controlled trials of exercise training during adjuvant treat-
ment (chemotherapy, hormonal and/or radiotherapy) in women with non-metastatic breast can-
cer will be considered for inclusion if they make any of the following comparisons:
• Aerobic and/or muscular endurance exercise versus no exercise
• Aerobic and/or muscular endurance exercise versus other interventions (e.g., psychosocial
interventions)
• Aerobic and/or muscular endurance exercise alone versus exercise as part of a complex in-
tervention
• Moderate intensity aerobic and/or muscular endurance exercise versus low intensity exercise
• Aerobic exercise versus muscular endurance exercise.
Appendix 1
141
Types of participants
Participants will include women with breast cancer stages I, II and III who are currently undergo-
ing adjuvant (including neo-adjuvant) chemotherapy, hormonal therapy or radiotherapy. Eligible
participants will include patients of any age, any reproductive status, any level of physical activity
before cancer diagnosis, and any weight/BMI. Patients will be ineligible if they have severe car-
diac disease, uncontrolled hypertension, orthopaedic contraindications etc.
Types of intervention
Studies that assess the effects of all forms of repeatedly performed aerobic and/or muscular
endurance exercise with program duration of at least six weeks will be considered for inclusion.
To be included in this review, the exercise intervention has to coincide with the adjuvant treat-
ment regimen rather than to follow it.
Interventions restricted to local muscular endurance only (e.g., training of shoulders, back or legs
only instead of including all major muscle groups) or to stretching exercises as well as to exercises
as part of complex interventions (e.g., complete decongestive lymphatic therapy) will be excluded.
Types of outcome measures
Outcomes assessed in this review will be any of the following:
•
Primary outcome measures
o Physical and functional well-being (e.g., fatigue, physical performance, sleep distur-
bances)
o Emotional and psychological well-being (e.g., depression, anxiety, self-esteem, cop-
ing, physical acceptance, uncertainty in disease)
o Psychosocial well-being (e.g., sex life, family life, social support, participation in pro-
fessional activities)
o Mental functioning (e.g., cognitive functions).
All types of outcome instruments such as interviews, patient self-report and clinical tests will be
considered.
•
Secondary outcome measures
o Cardiopulmonary function (e.g., maximum oxygen uptake, forced expiratory volume
in one second)
o Muscle strength
o Weight
o Immune function (e.g., natural killer cell activity, white blood count)
o Adverse effects (e.g., exercise-induced cardiovascular events or orthopaedic injuries).
•
Effect modifiers
o Compliance.
Appendix 1
142
SEARCH STRATEGY FOR IDENTIFICATION OF STUDIES
Electronic searching
We will apply the following MEDLINE search strategy (Silver Platter; Edition 2003) – based on
the Dickersin strategy for retrieving randomized controlled trials (Dickersin 1994), the CBCG's
strategy for the identification of populations with ‘breast neoplasms’ (Cochrane 2004) and key
words to identify exercise interventions. For sensitivity reasons, no keywords for non-metastatic
breast cancer and for adjuvant therapy will be applied: all reports retrieved will be checked for
stage of breast cancer and time point of exercise intervention.
1.RANDOMIZED-CONTROLLED-TRIAL in PT
2.CONTROLLED-CLINICAL-TRIAL in PT
3.RANDOMIZED-CONTROLLED-TRIALS
4.RANDOM-ALLOCATION
5.DOUBLE-BLIND-METHOD
6.SINGLE-BLIND-METHOD
7.#1 or #2 or #3 or #4 or #5 or #6
8.(TG=ANIMALS) not ((TG=HUMAN) and (TG=ANIMALS))
9.#7 not #8
10.CLINICAL-TRIAL in PT
11.explode CLINICAL-TRIALS/ all subheadings
12.(clin* near trial*) in TI
13.(clin* near trial*) in AB
14.(singl* or doubl* or trebl* or tripl*) near (blind* or mask*)
15.(#14 in TI) or (#14 in AB)
16.PLACEBOS
17.placebo* in TI
18.placebo* in AB
19.random* in TI
20.random* in AB
21.RESEARCH-DESIGN
22.#10 or #11 or #12 or 13 #14 or #15 or #16 or #17 or #18 or #19 or #20 or #21
23.((TG=ANIMALS) not (TG=HUMAN)) and (TG=ANIMALS)
24.#22 not #23
25.#24 not #9
26.TG=COMPARATIVE-STUDY
27.explode EVALUATION-STUDIES/ all subheadings
28.FOLLOW-UP-STUDIES
29.PROSPECTIVE-STUDIES
30.control* or prospectiv* or volunteer*
31.(#30 in TI) or (#30 in AB)
32.#26 or #27 or #28 or #29 or #31
33.(TG=ANIMALS) not ((TG=HUMAN) and (TG=ANIMALS))
34.#32 not #33
35.#34 not (#9 or #25)
36.#9 or #25 or #35
37.explode "Breast-Neoplasms"/ all subheadings
38.Cancer*
39.Carcinom*
40.Neoplas*
Appendix 1
143
41.Malignan*
42.#38 or #39 or #40 or #41
43.Breast*
44.#42 and #43
45.#37 or #44
46.explode "Exercise"/ all subheadings
47.explode "Exercise-Movement-Techniques"/ all subheadings
48.explode "Exercise-Therapy"/ all subheadings
49.explode "Physical-Education-and-Training"/ all subheadings
50.explode "Physical-Fitness"/ all subheadings
51.explode "Exertion"/ all subheadings
52.explode "Sports"/ all subheadings
53.(sport or sports) in ti, ab
54.physical activit* in ti,ab
55.kinesi?therap* in ti,ab
56.exercise* in ti, ab
57.walking in ti, ab
58.jogging in ti, ab
59.swimming in ti, ab
60.cycling in ti, ab
61.bicycling in ti, ab
62.weight in ti, ab
63.training in ti, ab
64.muscle in ti, ab
65.strengthening in ti, ab
66. endurance in ti, ab
67.#46 or #47 or #48 or #49 or #50 or #51 or #52 or #53 or #54 or #55 or #56 or 57# or
#58 or #59 or #60 or #61 or #62 or #63 or #64 or #65 or #66
68.#36 and #45 and #67
Other electronic databases that we will search are: EMBASE, CancerLit, PsycLit, SPORTDiscus,
CENTRAL and The System for Information on Grey Literature in Europe (SIGLE). Further-
more, we will search the specialized register maintained by the Cochrane Breast Cancer Group
and contact the Cochrane Collaboration Rehabilitation Field for help searching the literature in
the rehabilitation and physical therapy field. We will adapt the MEDLINE search strategy accord-
ing to the query requirements of the other databases.
Handsearching
Additionally, we will handsearch a series of relevant journals: Nursing Research, Oncology Nurs-
ing Forum, International Journal of Sports Medicine, Clinical Journal of Sports Medicine, Medi-
cine and Science in Sports and Exercise. Furthermore, we will check reference lists of retrieved
studies, reviews, as well as congress proceedings on sports medicine and oncology.
Experts
Relevant national and international experts will be contacted to identify trials either published or
unpublished where possible.
Appendix 1
144
METHODS OF THE REVIEW
Two reviewers will perform each of the following steps independently. Disagreements will be
resolved by discussion.
Selection of trials
Two reviewers (MM, TB) will independently assess potentially eligible trials for inclusion in the
review: first, titles (and abstracts if available) of articles identified will be checked for possible
inclusion; second, those believed to meet inclusion criteria will be obtained for a hard-copy re-
view. Discrepancies regarding eligibility will be resolved by discussion and/or a third reviewer.
Where it is necessary, additional information will be sought from the principal investigator of the
trial concerned.
Assessment of the methodological quality
Both reviewers will independently assess each eligible trial for quality; disagreements on scoring
will be resolved by discussion and a third reviewer will be used to resolve any discrepancies re-
garding quality that cannot be solved by discussion. Methodological assessment will be based on
the 19-point methodological quality scale of van Tulder (van Tulder 1997). Because blinding of
patients and care providers is not possible in exercise trials, we eliminated the corresponding
items from the original Tulder scale and thus reduced the maximum total van Tulder score possi-
ble to 17:
Methodological Quality Scale
1. Were the eligibility criteria specified?
2. Was a method of randomization performed?
3. Was the treatment allocation concealed?
4. Were groups similar at baseline regarding most important prognostic indicators?
5. Were interventions explicitly described?
6. Were co-interventions either avoided or comparable?
7. Was compliance acceptable in all groups?
8. Was the outcome assessor blinded to the intervention?
9. Were outcome measures relevant?
10. Were adverse effects described?
11. Was the withdrawal/ drop-out rate acceptable?
12. Was short-term follow-up measurement performed?
13. Was long-term follow-up measurement performed?
14. Was timing of outcome assessment comparable?
15. Was the sample size for each group described?
16. Did the analysis include intention-to-treat analysis?
17. Were point estimates and measures of variability presented for primary outcome measures?
Each criterion will be scored as ‘positive’, ‘negative’ or ‘unclear’. A total score will be computed
by counting the number of positive scores. Van Tulder et al. describe the use of these criteria to
Appendix 1
145
distinguish low and high quality by using an arbitrary preset cut-off point; following this ap-
proach of a preset cut-off point, moderate/high quality will be defined as fulfilling 9 (> 50%) or
more of the internal validity criteria.
Evaluation of adequacy of exercise dosage
Evaluation of the adequacy of the training stimulus will be based on the American College of
Sports Medicine (ACSM) exercise guidelines (ACSM 1998). These exercise guidelines represent
widely accepted criteria for achieving improvements in physical fitness; they are developed for
healthy individuals, and thus, we have adjusted them for non-metastatic breast cancer patients.
We will classify studies as ‘trials with an adequate exercise dosage’ by either meeting or not meet-
ing the dosage requirements of the ACSM guidelines.
Aerobic exercise: The dosage requirements for aerobic training interventions are:
• Intensity: 55-85% of maximum heart rate or 40-75% of maximum heart rate reserve or 40-
75% of maximum oxygen uptake reserve
• Duration of sessions: 20-60 min (minimum of 10 minutes continuous bouts throughout the
day) or exercise to tolerance
• Frequency: at least 2 days per week
• Total exercise period: at least 6 weeks.
Resistance exercise: The dosage requirements for muscular endurance training are:
• Intensity: 10-15 repetitions to near fatigue or at least 60% of one Repetition Maximum
• Number of sets completed (intensity): at least 1set
• Frequency: 2 or 3 days per week
• Total exercise period: at least 6 weeks.
Classification of high quality training studies
Studies which we classified as ‘methodological sound trials’ and as ‘trials with an adequate exer-
cise dosage’ will be considered as high quality training studies.
Data extraction
We will extract key information from all selected reports on a template data extraction form.
Extracted data will include:
• General information: Authors, title, source, contact address, country of performance of the
trial, language of publication, year of publication.
• Trial characteristics: Design (randomized controlled, non-randomized controlled), randomi-
zation method, concealment of allocation, configuration (parallel groups, waiting list, cross-
over), duration of intervention period, length of follow-up.
Appendix 1
146
• Patients: Sampling (random/consecutive/convenience), exclusion criteria, total and group
sample sizes, age, clinically relevant information (staging, menopausal status, hormone recep-
tor status, age, weight/BMI, activity level before diagnosis, compliance, attrition rates (rea-
sons/description).
• Experimental intervention: Mode of exercise training, intensity, duration, frequency, supervi-
sion (yes/no).
• Control intervention: Detailed description of the control intervention, particularly mode,
intensity duration, frequency, if available.
• Co-intervention: Detailed description of the co-intervention, particularly type of adjuvant
therapy (radiotherapy, chemotherapy, hormonal therapy, combination).
• Outcomes: Underlying concepts (e.g., depression), specific assessment instruments (e.g.,
Center for Epidemiologic Studies Depression Scale), score range (e.g., 0-60), direction (e.g.,
higher values indicate deterioration); we will extract outcome statistics according to the data-
entry requirements of RevMan Version 4.2. If the required data can not be extracted directly,
we will transform data as proposed in the Cochrane Reviewers’ Handbook.
Data combination
Regarding the multiplicity of outcome measures, meta-analysis will be considered where a group
of trials is sufficiently homogenous in terms of outcome measures. If the same outcome, e.g.,
depression or fatigue is assessed in several trials using different psychometric scales (continuous
data), the standardized mean difference (SMD) will be calculated; for dichotomous data (e.g.,
occurrence of adverse events) the relative risk (RR) will be calculated. Heterogeneity in treatment
effect will be assessed with a chi square test – provided that trials are enough in number - along
with the graphical presentation of confidence intervals of SMD/RR. If heterogeneity exists, we
will conduct a subgroup analysis (if at least ten studies will be available for each subgroup) on
intervention characteristics (type of exercise, adequacy of exercise, dosage) to explore effect
modification. I² of more than 50 % will be interpreted as evidence of heterogeneity (Cochrane
2004). The relationship between baseline risk and treatment effect will be elaborated through
discussion. Methodological quality will be used as a criterion (besides adequacy of intervention)
for inclusion in the quantitative or qualitative data synthesis; a sensitivity analysis may be per-
formed to determine whether overall results are the same when studies above different methodo-
logical cut-off points are analyzed. If possible, sensitivity analyses will be conducted on adequacy
of exercise dosage and ‘high quality training studies’. If meta-analysis is not reasonable data com-
bination will be performed only qualitatively.
Appendix 1
147
REFERENCES
ACSM. American College of Sports Medicine Position Stand. The recommended quantity and
quality of exercise for developing and maintaining cardiorespiratory and muscular fitness, and
flexibility in healthy adults. Med Sci Sports Exerc 1998; 30(6):975-91.
Ahlberg K, Ekman T, Gaston-Johansson F, Mock V. Assessment and management of cancer-
related fatigue in adults. Lancet 2003; 362(9384):640-50.
Camoriano JK, Loprinzi CL, Ingle JN, Therneau TM, Krook JE, Veeder MH.Weight change in
women treated with adjuvant therapy or observed following mastectomy for node-positive breast
cancer.J Clin Oncol 1990; 8(8):1327-34.
Cochrane Library. The Cochrane Library, Issue 1, 2004. Chichester, UK: John Wiley & Sons, Ltd.
2004.
Colcombe S, Kramer AF. Fitness effects on the cognitive function of older adults: a meta-
analytic study. Psychol Sci2003; 14(2):125-30.
Courneya KS. Exercise interventions during cancer treatment: biopsychosocial outcomes. Exerc
Sport Sci Rev 2001; 29(2):60-4.
Courneya KS, Mackey JR, McKenzie DC. Exercise for breast cancer survivors. Phys Sportsmed
2002; 30(8).
Courneya KS. Exercise in cancer survivors: an overview of research. Med Sci Sports Exerc 2003;
35(11):1846-52.
Curt GA, Breitbart W, Cella D, Groopman JE, Horning SJ, Itri LM, Johnson DH, Miaskowski C,
Scherr SL, Portenoy RK, Vogelzang NJ. Impact of cancer-related fatigue on the lives of patients:
new findings from the Fatigue Coalition. Oncologist 2000; 5(5):353-60.
de Jong N, Courtens AM, Abu-Saad HH, Schouten HC. Fatigue in patients with breast cancer
receiving adjuvant chemotherapy: a review of the literature. Cancer Nurs 2002; 25(4):283-97; quiz
298-9.
Demark-Wahnefried W, Rimer BK, Winer EP. Weight gain in women diagnosed with breast
cancer. J Am Diet Assoc 1997; 97(5):519-26, 529; quiz 527-8.
Demark Wahnefried W, Peterson BL, Winer EP, Marks L, Aziz N, Marcom PK, Blackwell K,
Rimer BK. Changes in weight, body composition, and factors influencing energy balance among
premenopausal breast cancer patients receiving adjuvant chemotherapy. J Clin Oncol 2001;
19(9):2381-9.
Dibble SL, Isreal J, Nussey B, Casey K, Luce J. Delayed chemotherapy-induced nausea in women
treated for breast cancer. Oncol Nurs Forum 2003; 30(2):E40-7.
Dickersin K, Scherer R, Lefebvre C. Identifying relevant studies for systematic reviews. Bmj
1994; 309(6964):1286-91.
Feyer P, Steingraeber M. Fatigue- a new therapeutic problem in oncology? [Fatigue-ein neues
therapeutisches Problem in der Onkologie?]. Im Focus Onkologie 2001; (7):59-64.
Holley S. Cancer-related fatigue. Suffering a different fatigue. Cancer Pract 2000; 8(2):87-95.
Hudis CA. Current status and future directions in breast cancer therapy. Clin Breast Cancer 2003;
4 Suppl 2:S70-5.
Jereczek-Fossa BA, Marsiglia HR, Orecchia R. Radiotherapy-related fatigue: how to assess and
how to treat the symptom. A commentary. Tumori 2001; 87(3):147-51.
Appendix 1
148
Morrow GR, Roscoe JA, Kirshner JJ, Hynes HE, Rosenbluth RJ. Anticipatory nausea and vomit-
ing in the era of 5-HT3 antiemetics. Support Care Cancer 1998; 6(3):244-7.
NCCN. Cancer-related fatigue. Clinical Practice Guidelines in Oncology 2003
NCCN. Antiemesis. 2004.
Newman LA, Sabel M. Advances in breast cancer detection and management. Med Clin North
Am 2003; 87(5):997-1028.
O'Shaughnessy J. Chemotherapy-related cognitive dysfunction in breast cancer. Semin Oncol
Nurs 2003; 19(4 Suppl 2):17-24.
Partridge AH, Burstein HJ, Winer EP. Side effects of chemotherapy and combined chemohor-
monal therapy in women with early-stage breast cancer. J Natl Cancer Inst Monogr 2001;
(30):135-42.
Patrick DL, Ferketich SL, Frame PS, Harris JJ, Hendricks CB, Levin B, Link MP, Lustig C,
McLaughlin J, Ried LD, Turrisi AT, Unutzer J, Vernon SW. National Institutes of Health State-
of-the-Science Conference Statement: Symptom Management in Cancer: Pain, Depression, and
Fatigue, July 15-17, 2002. J Natl Cancer Inst 2003; 95(15):1110-7.
Pinto BM, Maruyama NC. Exercise in the rehabilitation of breast cancer survivors. Psychooncol-
ogy 1999; 8(3):191-206.
Rugo HS, Ahles T. The impact of adjuvant therapy for breast cancer on cognitive function: cur-
rent evidence and directions for research. Semin Oncol 2003; 30(6):749-62.
Rutqvist LE. Adjuvant endocrine therapy. Best Pract Res Clin Endocrinol Metab 2004; 18(1):81-
95.
Salmon P. Effects of physical exercise on anxiety, depression, and sensitivity to stress: a unifying
theory. Clin Psychol Rev 2001; 21(1):33-61.
Spiegel D. Psychosocial aspects of breast cancer treatment. Semin Oncol 1997; 24(1 Suppl 1):S1-
47.
Spiegel D, Giese-Davis J. Depression and cancer: mechanisms and disease progression. Biol
Psychiatry 2003; 54(3):269-82.
Tchen N, Juffs HG, Downie FP, Yi QL, Hu H, Chemerynsky I, Clemons M, Crump M, Goss
PE, Warr D, Tweedale ME, Tannock IF. Cognitive function, fatigue, and menopausal symptoms
in women receiving adjuvant chemotherapy for breast cancer. J Clin Oncol 2003; 21(22):4175-83.
Truong PT, Olivotto IA, Whelan TJ, Levine M. Clinical practice guidelines for the care and
treatment of breast cancer: Locoregional post-mastectomy radiotherapy. Cmaj 2004;
170(8):1263-73.
van Tulder MW, Assendelft WJ, Koes BW, Bouter LM. Method guidelines for systematic reviews
in the Cochrane Collaboration Back Review Group for Spinal Disorders. Spine 1997;
22(20):2323-30.
Winningham ML, MacVicar MG. The effect of aerobic exercise on patient reports of nausea.
Oncol Nurs Forum 1988; 15(4):447-50.
Appendix 2
149
9.2 Appendix 2 – Search Activities
9.2.1 Journals handsearched
Journal Years
Psycho-Oncology 1997 – 2004
Cancer 1990 – 2004
Breast Cancer Research and Treatment 1990 – 2004
Nursing Research 1996 – 2004
Applied Nursing Research 1994, 2000 – 2004
International Journal of Sports Medicine 1990 – 2004
Medicine & Science in Sports & Medicine 1980 – 2004
Clinical Journal of Sport Medicine 2000 – 2004
Deutsche Zeitschrift für Sportmedizin 1985 – 2004
Oncology Nursing Forum 2000 – 2004
The Physician and Sportsmedicine 1990 – 2003
European Journal of Sport Science 2000 - 2004
9.2.2 Conference proceedings searched
Conference Years
International Symposium for Adapted Physical Activity 2001
American Public Health Association; annual meetings 2000-2003
American College of Sports Medicine; annual meeting 2004
International society of Behavioral Medicine; annual meetings 2002-2004
International Society for Behavioral Nutrition and Physical Activity; an-
nual conferences 2002-2004
The Cooper Institute Conference Series on Physical Activity 1999-2003
European Congress of Sport Psychology 2003
Appendix 2
150
9.2.3 Study registers searched
Study register Provider URL
ClinicalTrials.gov U.S. National Institutes of Health http://www.clinicaltrials.gov/
ISRCTN-Register
International Standard Random-
ised Controlled Trial Number;
Current Controlled Trials Ltd.
http://www.controlled-
trials.com/isrctn/
metaRegister of Controlled
Trials (mRCT) Current Controlled Trials Ltd. http://www.controlled-
trials.com/mrct/
EORTC Protocols Data-
base
European Organisation for re-
search and Treatment of Cancer,
EU
http://www.eortc.be/
Deutsches Krebsstudien-
register Deutsche Krebsgesellschaft, D http://www.studien.de/index.php
Medical Research Council
study register MRC Clinical Trials Unit, UK http://www.ctu.mrc.ac.uk/
BrowseCancer.asp
Clinical Trials register National Cancer Institute, US http://www.cancer.gov/clinicaltrials
National Research Register Department of Health, UK http://www.nrr.nhs.uk/search.htm
9.2.4 Experts contacted
American College of
Sports Medicine
Cancer Research Inter-
est Group
Center for evidence
based Physiotherapy
Cochrane Rehabilitation
& Related Therapies
Field
Ahlberg, Karin
Ainsworth, Barbara E
Baumann, Freerk
Bie, Rob de
Bolling, C.-O.
Bös, Klaus
Callow, Nichola
Campbel, Anna
Courneya, Kerry S.
Crank, Helen
Crevenna, Richard
Daley, Amanda
Demark-Wahnefried,
Wendy
Dimeo, Fernando
Drouin, Jacqueline
Fauteck-Kaskel, Dörte
Fox, Ken
Friedenreich, Christine
Irwin, Melinda L.
Kärki, Anne
Lötzerich, Helmut
Marcora, Samuele
McTiernan, Anne
Mock, Victoria
Mutrie, Nanette
Oldervoll, Liane M.
Peters, Christiane
Pickett, Mary
Pinto, Bernardine M.
Rojas, Roberto
Schüle, Klaus
Schwartz, Anna
Segal, Roanne
Servaes, P
Stevinson, Clare
Uhlenbruck, Gerhard
Wilkie, Diana J
Windsor, Phyllis
Winningham, Maryl L.
Appendix 2
151
9.2.5 Search strategy for electronic databases
Systematic serches employed free text and thesaurus terms (controlled termes) based on individ-
ual database thesauri. Since most online databases do not automatically search for all variations of
a word, either a truncation symbol (* at the end) or a wild character (e.g. behavio?r) was used to
look for variant spellings. A truncation symbol instructs the search engine to look for all words
with the specific beginning and a wild character indicates any character or none. Search terms
were related to the broad areas of breast cancer, exercise, study design and adjuvant treatment of
breast cancer (in two databases).
1. MEDLINE (Silver Platter MEDLINE)
Breast Cancer Exercise Study design
Controlled terms:
Breast-Neoplasms.
Free text:
(Cancer* or Car-
cinom* or Neo-
plas* or Malig-
nan*) and Breast*
Controlled terms:
Exercise, Exercise-Movement-
Techniques, Exercise-Therapy,
Physical-Education-and-Training,
Physical-Fitness, Exertion, Sports.
Free text:
Sport*, physical activit*,
kinesi?therap*, exercise*, walking or
jogging or swimming or cycling or
bicycling or weight training or mus-
cle strength* or treadmill or endur-
ance.
Controlled terms:
Randomized controlled trial, controlled
clinical trial, randomized controlled trials,
random allocation, double blind method,
single blind method, clinical trials, placebos,
comparative study, evaluation studies, fol-
low-up studies, prospective studies.
Free text:
(clin* near trial*); (singl* or doubl* or trebl*
or tripl*) near (blind* or mask*); placebo*,
random*; control* or prospectiv* or volun-
teer*.
2. EMBASE
Breast Cancer Exercise Study design
Controlled terms:
Breast-tumor, Breast
cancer, Breast carci-
noma.
Free text:
Breast*and (Neo-
plas* or carcinom*
or cancer* or malig-
nan*).
Controlled terms:
Physical education, Physical activity, Fit-
ness, Kinesiotherapy, Muscle training,
Muscle exercise, Leg exercise, Treadmill
exercise, Dynamic exercise, Exercise-.
Free text:
Exercise* or walking or swimming or
cycling or bicycling or weight training or
resistance training or muscle strengthening
or endurance or sport* or physical activit*.
Controlled terms:
Clinical-research, Clinical-trial,
Clinical-study, Controlled-study,
Randomized-controlled-trial, Ran-
domization-, Controlled trial.
Free text: prospective study or
Comparative study.
Appendix 2
152
3. CINAHL
Breast Cancer Exercise Study de-
sign
Cancer Treatment
Controlled terms:
breast-neoplasms.
Free text:
(cancer* or carcinom*
or malignan* or
neoplas* or sarcom*
or tumo*) and
(breast or “breast”)
Controlled terms: Physical-activity,
Physical-therapy, Endurance-sports,
Aquatic-exercises, Aerobic-exercises,
Therapeutic exercises, Rehabilita-
tion-Exercise-Saba-HHCC, Recov-
ery-Exercise.
Free text:
Walking or swimming or cycling or
dancing or jogging or muscle
strengthening or resistance training
or endurance.
Controlled
terms:
Clinical-trial,
Interventions,
Research, Trial
Free text:
treatment or chemo-
therap* or radiotherap*
or radiation therap* or
hormon* therap* or
adjuvant therap*
4. PsycInfo
Breast Cancer Exercise
Free text:
(Cancer* or carcinom* or
neoplas* or malignan*) and
(Breast or Breast cancer)
Controlled terms:
Physical-education, Physical-endurance, Physical-fitness, Physical-
strength, Aerobic-exercise, Exercise.
Free text:
Sport* or physical activit* or exercise* or walking or jogging or swimming
or cycling or bicycling or weight training or muscle strengthening or en-
durance.
5. PubMed
Breast Cancer Exercise Limits
Controlled terms:
Breast Neo-
plasms.
Free text:
Breast cancer.
Controlled terms:
Sports, Exertion, Weight Lifting, Exercise, Exer-
cise Therapy, Exercise Movement Techniques.
Free text :
Walking, jogging, swimming, cycling, bicycling,
weight training, muscle strength, endurance,
exercise*, sport*, aerobic, treadmill.
Randomized Controlled Trial,
Clinical Trial, Female, Human,
and Cancer.
Appendix 2
153
6. SPORTDiscus
Breast Cancer Cancer Treatment
Controlled terms:
Breast Neoplasm.
Free text:
Breast cancer
Free text:
treatment or chemotherap* or radiotherap* or radiation therap* or hormon*
therap* or adjuvant therap*
7. ProQuest Digital Dissertations: Free text: Breast cancer and exercise
8. DissOnline.de: Free text: Sport und Brustkrebs
Appendix 3
154
9.3 Appendix 3 – Study eligibility form
Assessment for study inclusion
Study-ID:
Reviewer:
Date:
TYPE OF STUDY Does the study compare two or more groups
assigned by trial investigator to exposure?
PARTICIPANTS Are participants in the study predominantly
(>80%) women with non-metastatic breast
cancer?
INTERVENTION Was one of the study groups exposed to either
aerobic endurance training (e.g., walking, bicy-
cling, or swimming) or resistance training (e.g.,
weight lifting, resistance exercise bands or
exercise tubings)?
Answer ‘no’ if intervention was restricted to
local muscular endurance only, stretching or
exercise as part of a complex intervention.
Usual care is regarded as a co-intervention:
thus, answer ‘yes’ if exercise is added to usual
care.
Were participants undergoing adjuvant cancer
treatment (chemotherapy, hormonal therapy or
radiotherapy) during study participation?
Was exposure to exercise training at least six
weeks?
OUTCOMES Does the study assess one of the following
outcomes?
Physical fitness (cardiorespiratory fitness, mus-
cular strength, muscular endurance, body com-
position), quality of life, fatigue, pain, psycho-
logical distress, immune function, harm.
FINAL DECISION
Include Unclear Exclude
Appendix 4
155
9.4 Appendix 4 – Data extraction forms
9.4.1 Data extraction of study descriptors
Data extraction
Study-ID:
Reviewer:
Date:
METHODS Type of trial -
Configuration of trial
Method of randomisation
Allocation concealment
Outcome assessor blinding
Timing of outcome assessments after initiation of treatment
Length of post-intervention follow-up period
Sample size, total
Sample size for each group
Availability of sample size for each group at randomisation
Withdrawal-/Drop-out rate and reasons
Adherence
Type of analysis
Statistical tests
Power calculation
PARTICIPANTS Recruitment
Inclusion criteria
Exclusion criteria
Age
BMI
Breast cancer staging
Menopausal status
Type of adjuvant cancer treatment
Physical activity before cancer diagnosis
Co-morbidity
INTERVENTION Type of training
Setting
Health promotion theory
Delivery of intervention
Intensity
Duration of sessions
Frequency
Total exercise period
CONTROL Type of control
Intensity
Duration of sessions
Frequency
Total period
CO-INTERVENTION Type of co-intervention
OUTCOMES Outcome measure concepts
Assessment instruments
Adverse effects
Similarity at baseline
Appendix 4
156
9.4.2 Instructions for extraction of study descriptors
Data extraction - Instruction for reviewers
METHODS
Type of trial - CCT: any allocation procedure applied?
- RCT
Configuration of trial
- Parallel group
- Cross over design
Method of randomisation Enter text
Allocation concealment - A: Adequate
- B: Unclear
- C: Inadequate
- D: Not used
Outcome assessor blinding - Yes
- No
- N/a
Timing of outcome assessments after
initiation of treatment
Enter numbers for each occasion of outcome assessment
Length of post-intervention follow-up
period
Enter numbers
Sample size, total Enter numbers used in analysis
Sample size for each group Enter numbers used in analysis
Availability of sample size for each
group at randomisation
- Yes
- No
- N/a
Withdrawal-/Drop-out rate and rea-
sons
Enter numbers (% and x/n) and text
Adherence - Describe the mode of monitoring the exercise stimulus
(intensity, frequency, duration)
- Enter numbers or text (percentage of prescribed exercise
sessions which are completed)
Type of analysis - ITT analysis: participants analysed as randomised, all par-
ticipants included regardless of whether outcomes collected
- Available case analysis: data analysed for every participant
for whom outcome was obtained
- Per-protocol/ treatment- received analysis: data analysed
for only those participants who completed the trial and who
complied with/or received their allocated treatment
Statistical tests Enter text (only for outcome measures’ analysis)
PARTICIPANTS
Recruitment Enter text
Inclusion criteria Enter text
Exclusion criteria Enter text
Age For each group and for all participants: Mean, SD;
calculate weighted mean if mean only available per group
BMI For each group and for all participants: Mean, SD;
calculate BMI if weight and height available; calculate weighted
mean if mean only available per group
Breast cancer staging For each group and for all participants
- Stage I: x/n (%)
- Stage II: x/n (%)
- Stage III: x/n (%)
Appendix 4
157
Menopausal status For each group and for all participants
- Pre- menopausal: x/n (%)
- Post- menopausal: x/n (%)
Type of adjuvant cancer treatment For each group and for all participants
- Chemotherapy: x/n (%)
- Radiotherapy: x/n (%)
- Hormonal Therapy: x/n (%)
- Combination: x/n (%)
Physical activity before cancer diagno-
sis
For each group and for all participants
- Active: x/n (%)
- Sedentary: x/n (%)
Co-morbidity For each group: enter text
INTERVENTIONS
Type of training - Aerobic- or Muscular endurance exercise
- Walking, cycling, etc.
Setting - Laboratory-based
- Centre-based
- Home-based
- Community based
Health promotion theory Describe the health promotion theory that underpins the
intervention
Delivery of intervention Describe exercise promoting strategies of the intervention
Intensity - Aerobic endurance exercise: % HR max, % HRR max, %
VO2 max
- Muscular endurance exercise: number of repetitions per set,
% 1-RM, number of sets completed
Duration of sessions Enter numbers (minutes)
Frequency Enter numbers (sessions per week)
Total exercise period Enter numbers (weeks)
CONTROL
Type of control Enter text (e.g., stretching exercise, usual care)
Intensity Enter text, numbers
Duration of sessions Enter numbers (minutes)
Frequency Enter numbers (sessions per week)
Total period Enter numbers (weeks)
CO-INTERVENTION
Type of co-intervention Enter text
OUTCOMES
Outcome measures concepts Enter text (e.g., depression)
Assessment instruments - Specific scale
- Score range
- Direction of change for improvement
Adverse effects - Assessment of harm
- Reporting of harm
- Observation of harm
Similarity at baseline
- Adequate
- Inadequate (reviewer determines)
Appendix 4
158
9.4.3 Effect size data coding
Effect size data
Study-ID:
Reviewer:
Date:
Effect size sequence number
Effect size type
1 Posttest comparison
2 Change score comparison
3 Follow-up comparison
Category of outcome construct
1 Cardiorespiratory endurance
2 Body composition
3 Muscular strength and endurance
4 Quality of life
5 Fatigue
6 Pain
7 Psycho-social distress
8 Immune function
9 Harm
10 Other
Outcome descriptor
Type of data reported
1 Means and standard deviations
2 t-value or F-value
3 chi-square (df=1)
4 Frequencies or proportions, dichotomous
5 Frequencies or proportions, polychotomous
6 Other
Page number where effect size
data found
Raw difference favours (i.e.,
shows more success for):
1 Intervention group
2 Neither (exactly equal)
3 Control group
9 Cannot tell or statistically insignificant report only
Sample size Intervention group sample size
Control group sample size
Means and standard deviations
Intervention group mean
Control group mean
Intervention group standard deviation
Control group standard deviation
Proportions or frequencies
n of intervention group with a successful outcome
n of control group with successful outcome
% of intervention group with a successful outcome
% of control group with a successful outcome
Significance tests
t-value
F-value (df for the numerator must = 1)
Chi-square value
Calculated standard deviations Intervention group standard deviation
Control group standard deviation
In case of coding failure Available data:
Problems/missing data:
Appendix 5
159
9.5 Appendix 5 – Study quality forms
9.5.1 Methodological quality form
Study quality
Study-ID
Reviewer
Date
PATIENT SELECTION
1. Were the eligibility criteria specified?
Descriptive criterion
Yes/No/unable to determine
2. Was a method of randomisation performed?
Internal validity criterion
Yes/No/unable to determine
3. Was the treatment allocation concealed?
Internal validity criterion
Yes/No/unable to determine
4. Were groups similar at baseline?
Descriptive criterion
Yes/No/unable to determine
INTERVENTION
5. Were interventions explicitly described?
Descriptive criterion
Yes/No/unable to determine
6. Were co-interventions either avoided or comparable?
Internal validity criterion
Yes/No/unable to determine
7. Was adherence acceptable in all groups?
Internal validity criterion
Yes/No/unable to determine
OUTCOME MEASUREMENT
8. Was the outcome assessor blinded to the intervention?
Internal validity criterion
Yes/No/unable to determine
9. Were outcome measures relevant?
Internal validity criterion
Yes/No/unable to determine
10. Were adverse effects described?
Descriptive criterion
Yes/No/unable to determine
11. Was drop-out rate acceptable and were reasons described?
Internal validity criterion
Yes/No/unable to determine
12. Was short-term follow-up measurement performed?
Descriptive criterion
Yes/No/unable to determine
13. Was long-term follow-up measurement performed?
Descriptive criterion
Yes/No/unable to determine
14. Was timing of outcome assessment comparable in groups?
Internal validity criterion
Yes/No/unable to determine
STATISTICS
15. Was sample size described for each group at randomisation?
Statistical criterion
Yes/No/unable to determine
16. Did the analysis include intention-to-treat analysis?
Internal validity criterion
Yes/No/unable to determine
17. Were point estimates and measures of variability presented
for the primary outcome measures?
Statistical criterion
Yes/No/unable to determine
SCORE
Appendix 5
160
9.5.2 Code of practice for implementing methodological quality criteria
Code of practice for implementing quality criteria
PATIENT SELECTION
1. Were the eligibility criteria specified?
To score a ‘yes’, inclusion and exclusion criteria have to be described appropriately.
2. Was a method of randomisation performed?
To score a ‘yes’, the method used to generate a random allocation sequence has to be specified (e.g.,
random numbers or computer-generated random sequences). Allocation methods using date of birth,
hospital numbers, or alternation should be regarded as quasi-randomised (score ‘no’).
3. Was the treatment allocation concealed?
To score a ‘yes’, an adequate method has to be used to prevent foreknowledge of group assignment:
these methods should allow for preventing researchers from influencing which participants are assigned
to a given intervention group. Adequate methods of allocation concealment include: centralised randomi-
sation, sequentially numbered, sealed envelopes each containing the name of the next treatment on a
card, generation of assignment by an independent person not responsible for determining the eligibility
of the patients.
INTERVENTION
4. Were groups similar at baseline?
To score a ‘yes’, groups should be similar at baseline regarding following factors: adjuvant cancer treat-
ment, physical activity before diagnosis, breast cancer stages, BMI, age, co-morbidity, scores of main
outcome measures.
The reviewer must be satisfied that the groups’ outcomes would not be expected to differ, on the basis of
baseline differences in prognostic variables alone, by a clinically significant amount. This criterion may be
accomplished even though baseline data are presented of study completers only.
5. Were interventions explicitly described?
To score a ‘yes’, type of exercise intervention, intensity, duration of single sessions, frequency, and pro-
gramme duration for both the index and control intervention, should be described. Information has to be
sufficient to allow an estimate of the exercise stimulus/ dose.
6. Were co-interventions (other than the adjuvant treatment) either avoided or comparable?
To score a ‘yes’, any application of additional diagnostic or therapeutic procedures (besides usual care)
should either be avoided or comparable between index and control group; reviewer determines.
7. Was adherence acceptable in all groups?
To score a ‘yes’, reviewer determines whether adherence (percentage of prescribed exercise sessions
which are completed) was acceptable based on reported programme duration, number of possible ses-
sions and activity levels before cancer diagnosis.
OUTCOME MEASUREMENT
8. Was the outcome assessor blinded to the intervention?
To score a ‘yes’, reviewer determines based on the outcome variable (e. g. fitness evaluation with patient’s
presence required or self-administered questionnaire) whether enough information about blinding is
given.
9. Were outcome measures relevant?
To score a ‘yes’, one of the following outcomes should have been assessed: Physical and functional well-
being, Emotional and psychological well-being, Psychosocial well-being, Mental functioning, Muscle
strength, BMI, Immune function, Cardiopulmonary function, Adverse effects. Reviewer determines for
other outcomes.
10. Were adverse effects described?
To score a ‘yes’, adverse effects should be described and attributed to the allocated intervention; if it is
explicitly reported that “no adverse events” have occurred, a ‘yes’ should be scored, too; if adverse events
are described and stated to be unrelated to the intervention, score a ‘yes’.
11. Was the withdrawal/ drop-out rate acceptable and were reasons for withdrawal/ drop-out
described?
To score a ‘yes’, participants included in the study who did not complete the intervention period or were
not included in the analysis must be described (reasons). Furthermore, the percentage of withdrawals and
drop-outs must not exceed 20% for short-term follow-up and 30% for long-term follow-up. If analysis of
Appendix 5
161
reasons for withdrawals and drop-outs suggest substantial bias (e. g. drop-outs in control group only due
to lacking perceived benefit), a ‘no’ is scored; reviewer determines.
12. Was short-term follow-up measurement performed?
To score a ‘yes’, outcome assessment has to be performed at the end of the intervention period.
Outcome assessment at the end of the intervention period is considered as adequate short-term follow-
up, since the review aims at assessing the potentials of exercise in managing common side effects of
adjuvant therapy and adjuvant therapy is administered only over a limited period of time.
13. Was long-term follow-up measurement performed?
To score a ‘yes’, outcome assessment has to be performed after the end of the intervention period. Long-
Term follow-up may be performed if programme duration is shorter than the period of adjuvant therapy
or if side effects (e. g. weight gain or fatigue) of adjuvant therapy are assessed as long-term side effects.
13. Was long-term follow-up measurement performed?
To score a ‘yes’, outcome assessment has to be performed after the end of the intervention period. Long-
Term follow-up may be performed if programme duration is shorter than the period of adjuvant therapy
or if side effects (e. g. weight gain or fatigue) of adjuvant therapy are assessed as long-term side effects.
14. Was timing of outcome assessment comparable in both groups?
To score a ‘yes’, timing of outcome assessment should be identical for intervention and control group.
STATISTICS
15. Was the sample size for each group described?
To score a ‘yes’, sample size has to be presented for each group at randomisation.
16. Did the analysis include intention-to-treat analysis?
To score a ‘yes’, the following criteria have to be fulfilled (Cochrane 2004):
- Trial participants have to be analysed in the groups to which they are randomised regardless of which
(or how much) treatment they actually received, and regardless of other protocol irregularities, such
as ineligibility.
- All participants have to be included regardless of whether their outcomes were actually collected
(involves imputing missing data).
17. Were point estimates and measures of variability presented for primary outcome measures?
To score a ‘yes’, both, point estimates and measures of variability should be presented for at least one key
outcome based on between-group statistical comparisons.
Appendix 5
162
9.5.3 Intervention quality form
Intervention quality
Study-ID:
Reviewer:
Date:
Aerobic endurance training
60-80% of maximum heart rate or
50-75% of maximum heart rate reserve or
50-75% of maximum oxygen uptake reserve or
RPE 11-14
yes, no, unable to determine
20-60 min (minimum of 10 minutes continuous bouts
throughout the day) or exercise to tolerance yes, no, unable to determine
At least 3 days per week yes, no, unable to determine
At least 6 weeks yes, no, unable to determine
Score
Muscular endurance training
10-15 repetitions to near fatigue or at least 60% of 1-RM yes, no, unable to determine
At least 1set yes, no, unable to determine
2 or 3 days per week yes, no, unable to determine
At least 6 weeks yes, no, unable to determine
Score
Appendix 6
163
9.6 Appendix 6 – Identification/critical appraisal of studies of harm
9.6.1 Medline search strategy for studies of harm (WinSPIRS 5.0)
#1 Cancer*
#2 Carcinom*
#3 Neoplas*
#4 Malignan*
#5 #1 or #2 or #3 or #4
#6 Breast*
#7 #5 and #6
#8 (explode "Breast-Neoplasms"/ all subheadings) or #7
#9 explode "Exercise"/ all subheadings
#10 explode "Exercise-Movement-Techniques"/ all subheadings
#11 explode "Exercise-Therapy"/ all subheadings
#12 explode "Physical-Education-and-Training"/ all subheadings
#13 explode "Physical-Fitness"/ all subheadings
#14 explode "Exertion"/ all subheadings
#15 explode "Sports"/ all subheadings
#16 (sport or sports) in ti, ab
#17 physical activit* in ti,ab
#18 kinesi?therap* in ti,ab
#19 exercise* in ti, ab
#20 walking or jogging or swimming or cycling or bicycling or weight training
or muscle strength* or treadmill or endurance or aerobic training or
resistance training
#21 #9 or #10 or #11 or #12 or #13 or #14 or #15 or #16 or #17 or #18 or
#19 or #20
#22 safe or safety or side-effect* or undesirable effect* or tolerability
#23 adverse adj (effect* or reaction* or event* or outcome*)
#24 lymph?edema
#25 injur*
#26 #25 or #24 or #23 or #22
#27 #26 and #21 and #8
Appendix 6
164
9.6.2 Data extraction form for studies of harm
Studies of harm - data extraction
Study-ID:
Reviewer:
Date:
Type of adverse events
1 injury
2 cardiovascular event
3 lymphedema
4 other
Inclusion criteria for parti-
cipants
1 post-treatment
2 during treatment
3 lymphedema
Type of exercise
1 aerobic endurance training
2 resistance training
3 aerobic and resistance training
Type of trial
1 randomised controlled trial
2 non- randomised controlled trial
3 Cohort study
4 Case-control study
5 Cross-sectional study
6 Large Databases
7 other
Sample size Total sample size
Intervention group sample size
Control Group sample size
Proportions or frequencies
n of intervention group with an adverse event
n of control group with an adverse event
% of intervention group with an adverse event
% of control group with an adverse event
Sample size
Intervention group sample size
Control group sample size
Significance tests
T-test
ANOVA
Chi squared
other
Appendix 6
165
9.6.3 Quality assessment form for studies of harm
Studies of harm - quality
Study-ID:
Reviewer:
Date:
Criterion Score
Nonbiased selection 1: Study is properly randomised, controlled
trial or observational study with clear, prede-
fined inception cohort
0: Selection not clear or biased selection
Adequate description of population 1: Study reports at least two demographic
characteristics, presenting symp-
toms/syndrome
0: Study does not meet above criteria
Low loss to follow-up, and patients lost to
follow-up analysed for adverse events
1: Study reports number lost to follow-up,
analyses patients lost to follow-up for adverse
events, and has low overall number lost to
follow-up
0: Study does not meet above criteria
Adverse events prespecified and defined 1: Study reports explicit definitions for major
complications that allow for reproducible
ascertainment
0: Study does not meet above criteria
Ascertainment technique adequately described 1: Study reports methods used to ascertain
complications, including who ascertained,
timing, and methods used
0: Study does not meet above criteria
Nonbiased and accurate ascertainment of
adverse event
1: Study provides independent assessment of
complication (defined as assessment by some-
one other than the exercise trainer)
0: Study does not meet above criteria
Adequate statistical analysis of potential con-
founders
1: Study examines relevant confounders/ risk
factors using acceptable statistical techniques,
such as stratification or adjustment
0: Study does not meet above criteria
Adequate duration of follow-up 1: Study reports duration of follow-up and
duration of follow-up is adequate to identify
expected adverse events
0: Study does not meet above criteria
Total quality score = sum of scores (0 - 8) >6: Good
4-6: Fair
<4: Poor
Appendix 7
166
9.7 Appendix 7 – Reference list of excluded studies of benefit
Study-ID Reference
Aghili 2007
Aghili M, Farhan F, Rade M. A pilot study of the effects of programmed aerobic exercise
on the severity of fatigue in cancer patients during external radiotherapy. Eur J Oncol
Nurs 11: 179-82; 2007.
Battaglini
2004
Battaglini CL. A randomized study on the effects of a prescribed exercise intervention on
lean mass and fatigue changes in breast cancer patients during treatment. PhD [disserta-
tion]. Colorado: University of Northern Colorado; 2004.
Burnham
2002
Burnham TR, Wilcox A. Effects of exercise on physiological and psychological variables
in cancer survivors. Med Sci Sports Exerc 34: 1863-7; 2002.
Courneya
2003a
Courneya KS, Friedenreich CM, Sela RA, Quinney HA, Rhodes RE, Handman M. The
group psychotherapy and home-based physical exercise (group-hope) trial in cancer
survivors: physical fitness and quality of life outcomes. Psychooncology 12: 357-74; 2003.
Courneya
2003b
Courneya KS, Mackey JR, Bell GJ, Jones LW, Field CJ, Fairey AS. Randomized con-
trolled trial of exercise training in postmenopausal breast cancer survivors: cardiopul-
monary and quality of life outcomes. J Clin Oncol 21: 1660-8; 2003.
Daley 2007
Daley AJ, Crank H, Saxton JM, Mutrie N, Coleman R, Roalfe A. Randomized trial of
exercise therapy in women treated for breast cancer. J Clin Oncol 25: 1713-21; 2007.
Daley AJ, Crank H, Mutrie N, Saxton JM, Coleman R. Patient recruitment into a ran-
domised controlled trial of supervised exercise therapy in sedentary women treated for
breast cancer. Contemp Clin Trials 6: 6; 2007.
Demark-
Wahnefried
2002
Demark-Wahnefried W. Randomized study of a diet and exercise-based counselling
program versus a standard counselling program for patients with early-stage breast or
prostate cancer [online]. Available from: http://www.cancer.gov. Accessed 08.05.2002
(Source: CBCG).
Demark-
Wahnefried
2003
Demark-Wahnefried W, Clipp EC, McBride C, Lobach DF, Lipkus I, Peterson B, Clutter
Snyder D, Sloane R, Arbanas J, Kraus WE. Design of FRESH START: a randomized
trial of exercise and diet among cancer survivors. Med Sci Sports Exerc 35: 415-24; 2003.
Dimeo 1999
Dimeo FC, Stieglitz RD, Novelli-Fischer U, Fetscher S, Keul J. Effects of physical activ-
ity on fatigue and psychological status of cancer patients during chemotherapy. Cancer
85: 2273-7; 1999.
Fairey 2003
Fairey AS, Courneya KS, Field CJ, Bell GJ, Jones LW, Mackey JR. Effects of exercise
training on fasting insulin, insulin resistance, insulin-like growth factors, and insulin-like
growth factor binding proteins in postmenopausal breast cancer survivors: a randomized
controlled trial. Cancer Epidemiol Biomarkers Prev 12: 721-7; 2003.
Given 2002
Given B, Given C, McCorkle R, Kozachik S, Cimprich B, Rahbar M, Wojcik C. Pain and
Fatigue Management: Results of a Nursing Randomized Clinical Trial. Oncol Nurs Fo-
rum 29; 949-56; 2002.
Ho 1986
Ho C. Psychological adaptation and coping resources of breast cancer patients: Compari-
sons across three treatment modalities. PhD [dissertation]. Washington: University of
Washington; 1986.
Hwang 2008
Hwang JH, Chang HJ, Shim YH, Park WH, Park W, Huh SJ, Yang JH. Effects of super-
vised exercise therapy in patients receiving radiotherapy for breast cancer. Yonsei Med J
49: 443-50; 2008.
Latikka 1997 Latikka P, Pukkala E, Vihko V. Exercise and breast cancer. Duodecim 113: 317-22; 1997.
McKenzie
2003
McKenzie DC, Kalda AL. Effect of upper extremity exercise on secondary lymphedema
in breast cancer patients: a pilot study. J Clin Oncol 21: 463-6; 2003.
Mock 1994
Mock V, Burke MB, Sheehan P, Creaton EM, Winningham ML, McKenney Tedder S,
Schwager LP, Liebman M. A nursing rehabilitation program for women with breast
cancer receiving adjuvant chemotherapy. Oncol Nurs Forum 21: 899-907; 1994.
Mock 2001 Mock V, Pickett M, Ropka ME, Muscari Lin E, Stewart KJ, Rhodes VA, McDaniel R,
Grimm PM, Krumm S, McCorkle R. Fatigue and quality of life outcomes of exercise
Appendix 7
167
Study-ID Reference
during cancer treatment. Cancer Pract 9: 119-127; 2001.
Mock 2002 Mock V. Fatigue and physical functioning during breast cancer treatment. Oncol Nurs
Forum 29: 338; 2002.
Mustian 2002
Mustian KM, Katula JA, Gill DL. Exercise: complementary therapy for breast cancer
rehabilitation. In: Hall RL (Ed.). Exercise and sport in feminist therapy: constructing
modalities and assessing outcomes. New York: Haworth Press. 105-18; 2002.
Pickett 2002 Pickett M, Mock V, Ropka ME, Cameron L, Coleman M, Podewils L. Adherence to
moderate-intensity exercise during breast cancer therapy. Cancer Pract 10: 284-92; 2002.
Pinto 2003
Pinto BM, Clark MM, Maruyama NC, Feder SI. Psychological and fitness changes asso-
ciated with exercise participation among women with breast cancer. Psychooncology 12:
118-26; 2003.
Schneider
2007
Schneider CM, Hsieh CC, Sprod LK, Carter SD, Hayward R. Effects of supervised
exercise training on cardiopulmonary function and fatigue in breast cancer survivors
during and after treatment. Cancer 20: 20; 2007.
Schwartz
1999
Schwartz AL. Fatigue mediates the effects of exercise on quality of life. Qual Life Res 8:
529-38; 1999.
Schwartz
2001
Schwartz AL, Mori M, Gao R, Nail LM, King ME. Exercise reduces daily fatigue in
women with breast cancer receiving chemotherapy. Med Sci Sports Exerc 33: 718-723;
2001.
Segar 1998
Segar ML, Katch VL, Roth RS, Garcia AW, Portner TI, Glickman SG, Haslanger S,
Wilkins EG. The effect of aerobic exercise on self-esteem and depressive and anxiety
symptoms among breast cancer survivors. Oncol Nurs Forum 25: 107-113; 1998.
Shaw 2003
Shaw E, Demark-Wahnefried W, Andersen R. STRENGTH (Survival TRaining for
ENhancing Total Health): Phase II Randomized Pilot Study of Distance Medicine-Based
Exercise and Dietary Approach to Prevent Body Composition Change During Adjuvant
Chemotherapy in Patients With Stage I, II or IIIA Breast Cancer [online]. Available
from: http://www.cancer.gov. Accessed 01.10.2003 (Source: CBCG)
Wilkie 2003
Wilkie DJ, Schwartz AL, Huang HY, Ko N-Y, Liao WC, Hairabedian D, Zong S. Com-
puterized Exercise Education for Patients: Effects on Cancer-Related Fatigue. Amerian
Public Health Association. 131st. Annual Meeting; 2003.
Appendix 8
168
9.8 Appendix 8 – Included studies of benefit
9.8.1 Study descriptors and effect size statistics
Battaglini 2007 – Study descriptors
METHODS
Type of trial RCT
Configuration of trial Parallel group design
Method of randomisation Quasi-randomised
Allocation concealment A
Outcome assessor blinding n/a
Timing of outcome assessments
after initiation of treatment
15 weeks
Length of post-intervention
follow-up period
0
Sample size, total 20
Sample size for each group - Intervention group: n=10
- Control group: n=10
Availability of sample size per
group at randomisation
Yes
Withdrawal-/Drop-out rate and
reasons
- Intervention group: 0/10 (0%)
- Control group: 0/10 (0%)
- All participants: 0/20 (0%)
Adherence - Monitoring of intensity (HR) and frequency
- Adherence: 100% of all sessions
Type of analysis ITT
Statistical tests Two-way mixed model analysis of variance
Power calculation No
PARTICIPANTS
Recruitment Oncology practices in northern Colorado region
Inclusion criteria - Breast surgery
- Designated for adjuvant chemotherapy or radiotherapy
- 35-70 years
Exclusion criteria - Cardiovascular disease
- Respiratory disease
- Bone, joint or muscular abnormalities
- Immune deficiency.
- Metastatic disease
Age - Intervention group: 57.5, SD=23.0
- Control group: 56.6, SD=16.0
- All participants: 57.05
Weight
Height
BMI
- Intervention group: 77.5 kg; SD=27.3
- Control group: 82.2 kg, SD=25.0
- Intervention group: 168.9 cm, SD=10.2
- Control group: 169.2 cm, SD=10.2
- Intervention group: 27.1
- Control group: 28.5
- All participants: 27.7
Breast cancer staging n/a
Menopausal status n/a
Type of adjuvant cancer treat-
ment
- Chemotherapy
- Radiotherapy
Appendix 8
169
Battaglini 2007 – Study descriptors
Physical activity before cancer
diagnosis
n/a
Co-morbidity n/a
INTERVENTIONS
Type of training - Aerobic exercise: treadmill, elliptical trainers, cycle ergometer
- Resistance exercise: weight machines, free weights, rubber bands,
therapeutic balls (lateral and frontal raises, horizontal chest press,
lateral pulldown, alternating bizeps curls with dumbbells, triceps
extension, leg press, leg extension, leg curl, standing calf raises,
and three different types of abdominal exercises)
Setting Centre based
Health promotion theory No health promotion theory
Delivery of intervention - Individualised exercise prescription based on fitness assessments
administered at the beginning of the study
- Exercise sessions: initial administration of cardiorespiratory activ-
ity, followed by whole body stretching sessions, resistance training
and a cool-down period that included stretching activities
- Participants led and monitored during each exercise session by a
trained cancer exercise specialist from the University of Northern
Colorado School of Sport and Exercise Science who had partici-
pated in an educational seminar prior to the study
- Control group: waiting list, explanation of further research
needed, reminders to abstain from participating in any regular ex-
ercise programme while participating in the study
Intensity 40-60% of predicted maximum exercise capacity for both the cardio-
respiratory and strength exercise
Duration of sessions Total 60 min (6-12 min. cardiovascular, 5-10 min. stretching, 15-30
min. resistance, 8-10 min. cool down)
Resistance exercise: 8-12 exercises; 6-12 repetitions, 3 sets per session
Frequency 2 sessions per week
Total exercise period 15 weeks
CONTROL
Type of control No exercise
Intensity n/a
Duration of sessions n/a
Frequency n/a
Total period n/a
CO-INTERVENTION
Type of co-intervention n/a
OUTCOMES
Outcome measure concepts - Fatigue
- Body composition: lean body mass (LBM), body fat (BF)
- Muscular fitness
- Cardiorespiratory fitness
Assessment instruments - Revised Piper Fatigue scale (PFS)
- Three-site skinfold measurement
- Modified Bruce treadmill protocol
- Submaximal muscle endurance protocol after Kuramoto and
Payne: leg extension, seated leg curl, lateral pull down, seated
chest press
Adverse effects No (“No cases of injury”)
Similarity at baseline Adequate
Appendix 8
170
Battaglini 2007 – Effect size data
ES No. 1 2 3 4 5
ES type Post-test Post-test Post-test Post-test Post-test
Construct Fatigue Body composition Muscular strength
Instrument PFS % LBM % BF 1-RM* [kg] 1-RM [kg]
N intervention group 10 10 10 10 10
N control group 10 10 10 10 10
Mean intervention 0.84 74.1 25.9 295.59 116.3
Mean control group 3.23 68.9 31.2 260.89 102.7
SD intervention group 1.13 2.9 2.9 22.65 8.9
SD control group 1.16 4.1 4.1 38.76 15.2
* Sum of the results from the predicted 1-RM that was obtained from assessing the exercises of leg exten-
sion, seated leg curl, lateral pulldown and seated chest press
Appendix 8
171
Campbell 2005 – Study descriptors
METHODS
Type of trial RCT
Configuration of trial Parallel group design
Method of randomisation Computer-generated numbers, stratification by treatment
Allocation concealment B
Outcome assessor blinding n/a, questionnaires self-administered and returned to researcher in
sealed envelopes
Timing of outcome assessments
after initiation of treatment
12 weeks
Length of post-intervention
follow-up period
0
Sample size, total 22
Sample size for each group - Intervention group: n=12
- Control group: n=10
Availability of sample size per
group at randomisation
Yes
Withdrawal-/Drop-out rate and
reasons
- Intervention group: 2/12 (16.7%)
- Control group: 1/10 (10%)
- All participants: 3/22 (13.7%)
Reasons: travel constraints, secondary cancer
Adherence - Monitoring of intensity (HR) and frequency
- Adherence: 70% of all sessions
Type of analysis Available case analysis (n=19)
Statistical tests 2-sample t-tests, 2-sided
Power calculation No
PARTICIPANTS
Recruitment Cancer Centre (consultant oncologist, breast care nurse, radiother-
apy/chemotherapy nurse)
Inclusion criteria - Breast surgery
- Receiving adjuvant chemotherapy or radiotherapy
Exclusion criteria - Concurrent major problems (e.g., uncontrolled cardiac or hyper-
tensive disease, respiratory disease, cognitive dysfunction)
- Already exercising vigorously three times a week for 20 minutes
Age - Intervention group: 48, SD=10
- Control group: 47, SD=5
- All participants: 47.5
BMI/weight n/a
Breast cancer staging n/a
Menopausal status n/a
Type of adjuvant cancer treat-
ment
- Intervention group
o Chemotherapy: 3/12 (25%)
o Radiotherapy: 2/12 (16.7%)
o Combination: 7/12 (58.3%)
- Control group
o Chemotherapy: 3/10 (30%)
o Radiotherapy: 4/10 (40%)
o Combination: 3/10 (30%)
- All participants
o Chemotherapy: 6/22 (27.3%)
o Radiotherapy: 6/22 (27.3%)
o Combination: 10/22 (45.4%)
o
Appendix 8
172
Campbell 2005 – Study descriptors
Physical activity before cancer
diagnosis
Scottish Physical Activity Questionnaire (SPAQ), leisure time activity:
- Intervention group: 330 minutes, SD=171
- Control group: 421 minutes, SD=191
Co-morbidity n/a
INTERVENTIONS
Type of training Aerobic and resistance exercise: walking, cycling, low-level aerobics,
muscle-strengthening exercises, circuits
Setting Community based: group exercise, supervised
Health promotion theory Stages of change, self-efficacy
Delivery of intervention - Group-based exercise classes
- Exercise sessions: warm-up, 10-20 minutes exercise (which varied
from week to week), cool-down, relaxation period
- Workshops addressing health benefits, self-efficacy, exercise
barriers, supportive environment, setting goals, finding appropri-
ate activity options
Intensity 60-75% age-adjusted HRmax
Duration of sessions 10-20 minutes aerobic training
Frequency 2 sessions per week
Total exercise period 12 weeks
CONTROL
Type of control Monitoring
Intensity n/a
Duration of sessions n/a
Frequency n/a
Total period n/a
CO-INTERVENTION
Type of co-intervention Usual care (both groups)
OUTCOMES
Outcome measure concepts - Cancer-specific quality of life (primary outcome)
- Cancer-site-specific quality of life, symptoms
- Satisfaction with life
- Fatigue
- Physical fitness
- Physical activity
- Expectation of treatment
Assessment instruments - FACT-G
- FACT-B
- Satisfaction with Life Scale (SWLS)
- Perceived expectations and benefits of total care package
- R-PFS
- Scottish Physical Activity Questionnaire (SPAQ)
- 12-minute walk test
Adverse effects No (“no adverse reactions to taking part in the exercise intervention”)
Similarity at baseline Adequate
Appendix 8
173
Campbell 2005 – Effect size data
ES No. 1 2 3 4
ES type Post-test Post-test Post-test Change score
Construct Cardiorespiratory
fitness
Fatigue Quality of life Quality of life
Instrument 12-MWT [meter] PFS FACT-B FACT-G
N intervention group 10 10 10 10
N control group 9 9 9 9
Mean intervention group 1423 2.43 111.2 11.9
Mean control group 1083 4.35 94.3 -2.9
SD intervention group 261 1.94 14.1 13.8
SD control group 176 3.48 28.4 16.1
Appendix 8
174
Courneya 2007 – Study descriptors
METHODS
Type of trial - RCT
Configuration of trial
- Parallel group
Method of randomisation computer-generated programme, stratification by center and chemo-
therapy protocol
Allocation concealment - A: Adequate
Outcome assessor blinding - No
Timing of outcome assessments
after initiation of treatment
mean: 17 weeks, SD=4 weeks
Length of post-intervention fol-
low-up period
6 months
Sample size, total 242
Sample size for each group Intervention group 1 (aerobic): n=78
Intervention group 2 (resistance): n=82
Control group: n=82.
Different n according to outcome measure!
Availability of sample size for
each group at randomisation
- Yes
Withdrawal-/Drop-out rate and
reasons
Intervention group 1 (aerobic): 5.1% (4/78)
Intervention group 2 (resistance): 7.3% (6/82)
Control group: 11% (9/82)
All participants: 7.9% (19/242)
Estimated using PRO assessments;
Reason (most common): participants unreachable (n=9)
Adherence - Exercise trainers monitored adherence
- Aerobic exercise group: 72% sessions, 95.6% met duration;
87.2% met intensity
- Resistance exercise group: 68.2% sessions; 96.8% completed all
9 exercises; 96.9% completed 2 sets each; 94.5% completed 8-12
repetitions
Type of analysis - Available case analysis
Statistical tests Mixed model analysis
PARTICIPANTS
Recruitment - Cross Cancer Institute (Edmonton, Alberta)
- Ottawa Hospital Integrated Cancer Program (Ottawa, Ontario)
- British Columbia Cancer Agency (Vancouver, British Columbia)
Inclusion criteria - Breast cancer stages I-III
- Beginning adjuvant chemotherapy
- English- or French-speaking
- > 18 years
Exclusion criteria - Incomplete axillary surgery
- Transabdominal rectus abdominus muscle reconstructive surgery
- Uncontrolled hypertension
- Cardiac illness
- Psychiatric illness
- pregnant
Age - Aerobic exercise group: 49.0 (30-75)
- Resistance exercise group: 49.5 (25-76)
- Control group: 49.0 (26-78)
- All participants: 49.2 (25-78)
Appendix 8
175
Courneya 2007 – Study descriptors
BMI - Aerobic exercise group: 26.7; 5.6
- Resistance exercise group: 26.1; 5.5
- Control group: 27.1; 5.4
- All participants: 26.6; 5.5
Breast cancer staging - Stage I
o Aerobic exercise group: 18/78 (23.1%)
o Resistance exercise group: 22/82 (26.8%)
o Control group: 20/82 (24.4%)
o All participants: 60/242 (24.8%)
- Stage II a
o Aerobic exercise group: 33/78 (42.3%)
o Resistance exercise group: 36/82 (43.9%)
o Control group: 30/82 (36.6%)
o All participants: 99/242 (40.9%)
- Stage II b
o Aerobic exercise group: 17/78 (21.8%)
o Resistance exercise group: 9/82 (11%)
o Control group: 22/82 (26.8%)
o All participants: 48/242 (19.8%)
- Stage III
o Aerobic exercise group: 10/78 (12.8%)
o Resistance exercise group: 15/82 (18.3%)
o Control group: 10/82 (12.2%)
o All participants: 35/242 (14.5%)
Menopausal status - Post- menopausal
o Aerobic exercise group: 27/78 (34.6%)
o Resistance exercise group: 35/82 (42.7%)
o Control group: 27/82 (32:9%)
o All participants: 89/242 (36.8%)
Type of adjuvant cancer treatment - Chemotherapy
Physical activity before cancer
diagnosis
- Current exerciser
o Aerobic exercise group: 15/78 (19.2%)
o Resistance exercise group: 22/82 (26.8%)
o Control group: 27/82 (32:9%)
o All participants: 64/242 (26.4%)
Co-morbidity Obesity, hypertension (no difference between groups)
INTERVENTIONS
Type of training - Aerobic- endurance exercise: cycle ergometer, treadmill, elliptical
- Muscular endurance exercise: weight machines (set with 9 exer-
cises)
Setting - Centre-based
Health promotion theory n/a
Delivery of intervention - Exercise sessions: supervised at well-equipped fitness centres by
qualified staff; warm-up and cool-down periods of 5 minutes
with light aerobic activities and stretching
- Fitness trainers recorded exercise adherence including atten-
dance, duration and intensity
- Fitness trainers monitored harm
- Participants who called to cancel the exercise session were asked
for a reason why
- Fitness centers were open 8:00 a.m. to 5:00 p.m. Monday to
Friday; 1 centre made additional accommodations for weekends
and evenings if needed.
Appendix 8
176
Courneya 2007 – Study descriptors
Intensity - Aerobic exercise: 60-80 % VO2 max
- Resistance exercise: 2 sets of 8-12 repetitions, 60-70% RM,
number of sets completed
Duration of sessions Aerobic exercise: 15-45 minutes
Resistance exercise:
Frequency Aerobic exercise: 3/week
Resistance exercise: 3/week
Total exercise period Aerobic exercise: 17 weeks (SD=4 weeks)
Resistance exercise: 17 weeks (SD=4 weeks)
CONTROL
Type of control Usual care; women were asked not to initiate an exercise programme
Intensity n/a
Duration of sessions n/a
Frequency n/a
Total period n/a
CO-INTERVENTION
Type of co-intervention n/a
OUTCOMES
Outcome measures concepts - Cancer-specific quality of life
- Fatigue
- Self-esteem
- Depression
- Anxiety
- Aerobic fitness, cardiorespiratory capacity
- Strength
- Weight
- Body composition
- Lymphedema
- Chemotherapy completion rate
Assessment instruments - Functional Assessment of Cancer Therapy-Anemia scale
(FACT-An)
- Rosenberg Self-Esteem Scale
- Center for Epidemiological Studies Depression Scale (CES-D)
- Spielberger State Anxiety Inventory (STAI)
- Peak oxygen consumption in maximal incremental exercise
protocol on treadmill (expired gas analysis)
- 8-RM on horizontal bench press and leg extension
- Balance beam scale
- Dual x-ray absorptiometry
- Standard volumetric arm measurements based on water dis-
placement
- Average relative dose-intensity (RDI) for the originally planned
chemotherapy regimen
Adverse effects - Exercise trainers monitored harm
- Reporting of harm: yes
- Observation of harm: yes (after baseline maximal treadmill
testing)
Similarity at baseline
- Adequate
Appendix 8
177
Courneya 2007 – Effect size data
ES No. 1 2 3 4 5
ES type Post-test Post-test Post-test Post-test Post-test
Construct QoL Self-esteem Fatigue Anxiety Depression
Instrument FACT-An Rosenberg
FACT-An*
Fatigue-
subscale
STAI CES-D
N aerobic exercise 74 74 74 74 74
N resistance exercise 76 76 76 76 76
N control group 73 73 73 73 73
Mean aerobic exercise 144.7 34.5 36.8 35.0 9.7
Mean resistance exercise 140.9 34.7 36.3 36.4 10.6
Mean control group 139.9 33.2 34.9 37.4 10.8
SD aerobic exercise 25.2 5.1 10.4 11.7 9.3
SD resistance exercise 24.8 4.2 9.4 12.7 9.5
SD control group 28.2 5.5 12.5 12.0 9.4
*FACT Fatigue-subscale scores range from 0 to 52, where higher scores represent less fatigue.
Courneya 2007 – Effect size data
ES No. 6 7 8
ES type Post-test Post-test Post-test
Construct Cardiorespiratory fit-
ness [ml/kg/min]
Strength leg [kg] Strength Chest [kg]
Instrument VO2peak 1 RM 1RM
N aerobic exercise 71 71 71
N resistance exercise 77 77 77
N control group 73 73 73
Mean aerobic exercise 25.7 28.2 24.7
Mean resistance exercise 24.2 32.8 31.9
Mean control group 23.5 27.1 24.6
SD aerobic exercise 7.4 14.2 7.5
SD resistance exercise 6.1 12.6 10.8
SD control group 5.4 14.1 7.8
Courneya 2007 – Effect size data
ES No. 9 10 11 12 13
ES type Post-test Post-test Post-test Post-test Post-test
Construct Body weight
[kg]
Body fat [%] Fat mass [kg] Lean mass
[kg]
Arm differ-
ence [ml]
Instrument
N aerobic exercise 64 64 64 64 64
N resistance exercise 66 66 66 66 66
N control group 69 69 69 69 69
Mean aerobic exercise 70.3 37.9 27.1 40.9 -7
Mean resistance exercise 71.1 37.2 26.9 41.3 10
Mean control group 73.4 39.8 29.5 40.9 11
SD aerobic exercise 13.8 8.9 11.3 5.1 152
SD resistance exercise 15.4 9.0 12.0 4.9 118
SD control group 15.7 8.8 12.0 5.6 153
Appendix 8
178
Courneya 2007 – Effect size data
ES No. 10
ES type Post-test
Construct Lymphedema
Instrument Volumetric arm
measurements
N aerobic exercise 78
N resistance exercise 82
N control group 82
N aerobic exercise with
harm
7
N resistance exercise with
harm
3
N control group with
harm
6
Courneya 2007 – Effect size data at 6-month follow-up
ES No. 1 2 3 4 5
ES type Post-test Post-test Post-test Post-test Post-test
Construct QoL Self-esteem Fatigue Anxiety Depression
Instrument
FACT-An Rosenberg
FACT-An
Fatigue-
subscale
STAI CES-D
N aerobic exercise 68 68 68 68 68
N resistance exercise 73 73 73 73 73
N control group 60 60 60 60 60
Mean aerobic exercise 156.3 35.0 42.1 32.2 7.2
Mean resistance exercise 152.9 35.2 40.8 35.5 9.6
Mean control group 152.4 33.9 41.5 37.4 10.2
SD aerobic exercise 24.0 4.7 10.5 11.2 7.5
SD resistance exercise 26.0 4.6 10.5 13.0 10.4
SD control group 26.4 5.6 9.8 12.0 9.5
Appendix 8
179
Crowley 2003 – Study descriptors
METHODS
Type of trial RCT (p 31)
Configuration of trial Parallel group design
Method of randomisation Random numbers table (p 33)
Allocation concealment A (p 33)
Outcome assessor blinding Yes (p 34, blinding was not evaluated)
Timing of outcome assessments
after initiation of treatment
7 weeks (attention performance, fatigue, physical self-efficacy, func-
tional wellness), 13 weeks (all outcomes); (p 43)
Length of post-intervention
follow-up period
0 (p 43)
Sample size, total 22 (p 43)
Sample size for each group - Intervention group: n=13
- Control group: n=9, (p 43)
Availability of sample size for
each group at randomisation
Yes (p 43)
Withdrawal-/Drop-out rate and
reasons
- Intervention group: 0/13(0%)
- Control group: 0/10 (0%)
- All participants: 0/22 (0%)
(p 71; 100% retention)
Adherence - Monitoring of intensity (HR, perceived exertion), frequency, dura-
tion
- Assessed via activity log to document weekly exercise type, fre-
quency, duration, intensity
- Defined as completion of 80% of individual targets for aerobic
activity and strength training sessions
- Activity level per group:
o Intervention group: moderate level of activity (mean: 113
min/week) with a frequency of 3.66 days per week; (21-35
possible sessions >> 3.66 x 7 = 25.62 sessions =
73.2%/sessions)
o Control group: low level of activity (mean: 53 min/ week)
with a frequency of 1.79 days per week (p 47)
Type of analysis ITT (available case analysis for strength, p 45)
Statistical tests Repeated measures analysis of variance, t-tests (p 40)
Power calculation No
PARTICIPANTS
Recruitment - University comprehensive cancer centre
- Community hospital oncology practice in a Midwest city (p 32)
Inclusion criteria - 35-60 years
- Breast cancer stages I, II
- adjuvant chemotherapy (Adriamycin, Cytoxan)
- surgery
Exclusion criteria - History of cancer treatment
- Breast reconstruction
- Radiotherapy
- Cardiac, pulmonary disease, pregnancy, lactation (p 32)
Age All participants: range 36 – 58 (p 43)
BMI n/a
Breast cancer staging All participants (p 76):
- Stage I: 13/22 (59.1%)
- Stage II: 9/22 (40.9%);
Appendix 8
180
Crowley 2003 – Study descriptors
Menopausal status All participants (p 76):
- Premenopausal: 12/22 (54.5%)
- Postmenopausal: 10/22 (45.5%)
Type of adjuvant cancer treat-
ment
Chemotherapy: four three-week cycles Adriamycin, Cytoxan (p 32)
Physical activity before cancer
diagnosis
Yes, baseline exercise history of 72.7% (self-report, p 62)
Co-morbidity n/a
INTERVENTIONS
Type of training Aerobic and resistance exercise: walking, tubing (p 82-85; 88)
Setting Home based
Health promotion theory None
Delivery of intervention - Week 2: 1 hour exercise instructions through exercise physiologist
- Week 8: 30 minutes educational re-enforcement
- Week 4, 7, 10, 13: collection of activity logs through nurse re-
searchers at treatment visits
Intensity - Aerobic exercise: 60% of targeted heart rate?; individualised ac-
cording to fitness level and post-operative state, RPE 11-13
(“Fairly light” to “somewhat hard”)
- Resistance exercise: fatigue after 12-15 repetitions
Duration of sessions - Aerobic exercise: 20-60 min (p 34)
- Resistance exercise: ca. 20 minutes, 1-2 sets
Frequency - Aerobic exercise: 3-5 sessions per week
- Resistance exercise : 2-3 sessions per week
Total exercise period 13 weeks (p 34)
CONTROL
Type of control No intervention
Intensity n/a
Duration of sessions n/a
Frequency n/a
Total period n/a
CO-INTERVENTION
Type of co-intervention - Usual care
- Continuation of ongoing exercise regimens (p 32)
OUTCOMES
Outcome measure concepts - Cardiorespiratory fitness
- Muscular fitness
- Attention performance
- Fatigue
- Physical self-efficacy
- Functional wellness
- Quality of life
- Adherence (secondary endpoint)
Assessment instruments (p 35f) - Cornell Treadmill Protocol (VO2max)
- Chest press, leg press (1-RM)
- R-PFS
- Attentional Functional Index (AFI, 0-100; the larger the score, the
greater the perception of ability to function cognitively)
- Self-Efficacy to Perform Self-Management Behaviours Scale
- Self-Efficacy to Achieve Outcomes scale (10 point scale, higher
scores reflecting greater perception of physical self-efficacy)
- SF-36
- Functional Wellness Questionnaire (investigator developed, 5-
Appendix 8
181
Crowley 2003 – Study descriptors
point scale)
- Activity log with exercise type, frequency, duration and intensity
Adverse effects - 1 fall (unrelated to study)
- 1 lymphedema (p 45)
Similarity at baseline Inadequate
Crowley 2003 – Effect size data
ES No. 1
ES type Change score
Construct Cardiorespiratory fitness
Instrument Cornell treadmill test
N intervention group 13
N control group 9
Mean intervention group 0.37
Mean control group -5.38
T-Value 3.41
SD intervention group 3.89
SD control group 3.89
Appendix 8
182
Drouin 2002 – Study descriptors
METHODS
Type of trial RCT
Configuration of trial Parallel group design
Method of randomisation Random number chart
Allocation concealment A (inclusion first, then random assignment; pp 55,70)
Outcome assessor blinding n/a
Timing of outcome assessments
after initiation of treatment
8 weeks (1 week after 7 week intervention, p 72)
Length of post-intervention
follow-up period
0
Sample size, total 23 (p 64)
Sample size for each group - Intervention group: n=13
- Control group: n=10 (p 64)
Availability of sample size per
group at randomisation
Yes
Withdrawal-/Drop-out rate and
reasons
- Intervention group: 0/23 (0%)
- Control group: 2/10 (20%)
- All participants: 2/23 (8.7%)
Reasons, suspected: too busy with jobs and family
Adherence - Monitoring of intensity (HR), frequency, duration
- Training journals (intensity, duration, frequency, mode of train-
ing), heart rate monitors
- Adherence defined as 21 minimum sessions out of 35 possible
sessions
- Adherence per group:
o Intervention group: 25.8 sessions, SD=10.1
o Control group: 29.2 sessions, SD=7.7
Type of analysis Available case analysis
Statistical tests Wilcoxon-Mann-Whitney U
Power calculation No
PARTICIPANTS
Recruitment Karmanos Cancer Institute (major urban centre for radiation treat-
ment); (p 70)
Inclusion criteria - Breast cancer stages 0-III (histological established)
- Radiotherapy
- Surgery
- 20-65 years
- Sedentary (p 54)
Exclusion criteria - Uncontrolled cardiac disease or hypertension
- Aerobic exercise 3 month before entry in study (p 54)
Age - Intervention group: 49.4, SD=7
- Control group: 51.9, SD=10
- All participants: 50.5 (p 71)
BMI/weight
BMI
- Intervention group: 30.8, SD=7.6
- Control group: 32.0, SD=5.6
- All participants: 31.3 (p 81)
Breast cancer staging - Intervention group (p 71):
o Stage 0: n=3
o Stage I: n=2
o Stage II: n=4
o Stage III: n= 4
Appendix 8
183
Drouin 2002 – Study descriptors
- Control group:
o Stage 0: n=2
o Stage I: n=1
o Stage II: n=1
o Stage III: n=4
Menopausal status n/a
Type of adjuvant cancer treat-
ment
Radiotherapy (p 72)
Physical activity before cancer
diagnosis
No aerobic exercise 3 month prior to entry in study as inclusion crite-
rion
Co-morbidity 1 x Hypertension (p 72)
INTERVENTIONS
Type of training Aerobic exercise: walking (p 72)
Setting Home based
Health promotion theory None
Delivery of intervention - Individualised exercise prescription
- Weekly communication with principle investigator to promote
adherence, monitor training & safety issues, answer questions.
Intensity 50-70 % HRmax
Duration of sessions 20-45 minutes
Frequency 3-5 sessions per week
Total exercise period 7 weeks
CONTROL
Type of control Stretching (1 training session plus booklet); (p 73)
Intensity n/a
Duration of sessions n/a
Frequency 3-5 sessions per week (p 75)
Total period 7 weeks
CO-INTERVENTION
Type of co-intervention Weekly communication
OUTCOMES
Outcome measure concepts - Cardiorespiratory fitness
- Muscular fitness
- Body composition
- Fatigue
- Mood
- Immune function
- Oxidative stress
Assessment instruments - Modified Bruce protocol (VO2peak)
- Handgrip test (Jaymar Dynamometer)
- Waist Hip Ratio
- BMI
- Lange calliper (skinfold thickness)
- R-PFS (p 55 ff)
- POMS
- CD4+/ CD8+ ratio, NKCA (Natural Killer Cytotoxic Activity)
- 8-Isoprostane (Flow cytometry; Chromium release assay; ELISA)
- Training journal: intensity, duration, frequency, mode of training
Adverse effects
- 1x increased fatigue and mood disturbance
- decrements in strength and shoulder tendonitis due to overtrain-
ing
- no decrements in immune function (p 93)
Similarity at baseline Adequate (p 71, 64, 65)
Appendix 8
184
Drouin 2002 – Effect size data
ES No. 1 2 3 4 5
ES type Post-test Post-test Post-test Post-test Post-test
Construct Cardiorespiratory
fitness Body composition Muscular
fitness Fatigue
Instrument Modified Bruce
treadmill test Weight [kg] BMI Grip strength
[kg] R-PFS
N intervention group 13 13 13 13 13
N control group 8 8 8 8 8
Mean intervention group 22.6 79.9 30.1 31.3 3.4
Mean control group 16.6 83.3 31.7 32.0 3.9
SD intervention group 6.2 20.9 7.3 6.5 1.9
SD control group 2.2 18.2 6.1 6.4 2.4
Drouin 2002 – Effect size data
ES No. 6 7 8 9
ES type Post-test Post-test Post-test Post-test
Construct Mood Biological complications
Instrument POMS NKCA
[lytic units] CD4/CD8 8-Iso-prostane
serum levels
N intervention group 13 13 13 10
N control group 8 8 8 5
Mean intervention group 5.1 8.5 2.5 244.0
Mean control group 23.9 6.6 1.8 316.4
SD intervention group 22.1 9.2 1.1 132.6
SD control group 32.0 4.2 1.0 118.3
Appendix 8
185
Kim 2006 – Study descriptors
METHODS
Type of trial RCT
Configuration of trial Parallel group design
Method of randomisation Computer-generated randomisation lists; stratification by breast can-
cer stage
Allocation concealment B
Outcome assessor blinding n/a
Timing of outcome assessments
after initiation of treatment
- 10 weeks for functional ability, cardiorespiratory fitness and exer-
cise level
- 16 weeks for exercise level
Length of post-intervention
follow-up period
- 0 weeks for functional ability
- 6 weeks for exercise level
Sample size, total 41 (p 74)
Sample size for each group - Intervention group: n=22
- Control group: n=19
Availability of sample size for
each group at randomisation
no
Withdrawal-/Drop-out rate and
reasons
All participants: 11/74 (15%);
Reasons: personal problems, problems at home, problems related to
chemotherapy, thrombophlebitis in leg, non-exercise related injuries,
death; BMI higher in participants who withdrew or did not complete
exercise tests
Adherence - Monitoring of intensity (HR), frequency, total duration, and dura-
tion at target HR
- Adherence: 78.3%, SD=20.1% (exercise sessions completed at
prescribed level divided by total number of exercise sessions pre-
scribed)
- Average weekly frequency of exercise: 2.4, SD=0.6 sessions
- Average duration of exercise per session: 42.7 minutes, SD=8.0
- Average duration of exercise within prescribed target HR: 27.8
minutes, SD=8.1
- Overall adherence rate: 78.3%, SD=20.1%
Type of analysis Available case analysis
Statistical tests t-tests, Wilcoxon signed-rank test
Power calculation No
PARTICIPANTS
Recruitment - University breast cancer clinics
- Interdisciplinary breast Clinic
- University affiliated hospitals
- Satellite Cancer Clinics
Inclusion criteria - Women newly diagnosed with breast cancer, all stages
- No history of cancer
- Older than 40 years
- Receiving cancer treatment
Exclusion criteria - Bone metastasis
- High risk of fractures
- Psychiatric illness
- Uncontrolled cardiopulmonary or other serious medical condi-
tions
- Regular exercise at least 2 to 3 times a week in moderate intensity
within the past 2 month
Appendix 8
186
Kim 2006 – Study descriptors
Age - Intervention group: 51.3, SD=6.7
- Control group: 48.3, SD=8.8
- All participants: 49.9
BMI - Intervention group: 29.9, SD=5.5
- Control group: 28.0, SD=5.1
Breast cancer staging (n; %) - Intervention group:
o Stage 0: 1/22 (4.5%)
o Stage I: 10/22 (45.5%)
o Stage II: 8/22 (36.4%)
o Stage III: 3/22 (13.6%)
- Control group:
o Stage 0: 1/19 (5.3%)
o Stage I: 7/19 (36.8%)
o Stage II: 7/19 (36.8%)
o Stage III: 4/19 (21.1%)
- All participants:
o Stage 0: 2/41 (4.9%)
o Stage I: 17/41 (41.5%)
o Stage II: 15/41 (36.6%)
o Stage III: 7/41 (17.1%)
Menopausal status n/a
Type of adjuvant cancer treat-
ment (n; %)
- Intervention group:
o Chemotherapy: 9/22 (40.9%)
o Radiotherapy: 7/22 (31.8%)
o Combination: 6/22 (27.3%)
- Control group:
o Chemotherapy: 11/19 (57.9%)
o Radiotherapy: 7/19 (36.8%)
o Combination: 1/19 (5.3%)
- All participants:
o Chemotherapy: 20/41 (48.8%)
o Radiotherapy: 14/41 (34.2%),
o Combination: 7/41 (17.1%)
Physical activity before cancer
diagnosis
Sedentary (regular exercise at least 2 to 3 times a week in moderate
intensity within the past 2 month as exclusion criterion)
Co-morbidity n/a
INTERVENTIONS
Type of training Aerobic exercise: cycling, walking, jogging, or running on a treadmill
or track; progressive
Setting - Centre based: exercise facility within the School of nursing, su-
pervised
- home based after the regular intervention (physical activity as-
sessment in 16 weeks follow-up)
Health promotion theory None
Delivery of intervention After exercise intervention: biweekly calls of exercise physiologist to
collect physical activity data, encouragement to maintain the same
pattern of exercise at home or in community setting after supervised
training had finished
Intensity 60 – 70%HR reserve; 60 – 70 %VO2 peak
Duration of sessions 30 minutes (plus warm-up, cool down)
Frequency 3 sessions per week
Total exercise period 8 weeks
Appendix 8
187
Kim 2006 – Study descriptors
CONTROL
Type of control No intervention (waiting list)
Intensity n/a
Duration of sessions n/a
Frequency n/a
Total period n/a
CO-INTERVENTION
Type of co-intervention - Stress management >>> complex intervention, but no influence
on functional ability expected!
- Usual care
- ß-blocker, antihypertensives
OUTCOMES
Outcome measure concepts - Cardiorespiratory fitness
- Physical activity
Assessment instruments - Bruce treadmill protocol: Heart rate (resting, maximum), systolic
Blood pressure (resting, maximum), VO2 peak
- 7-day physical activity questionnaire: Average weekly frequency of
exercise, average duration of exercise per session, average dura-
tion of exercise within prescribed target HR, overall adherence
rate
Adverse effects n/a
Similarity at baseline Yes
Kim 2006 – Effect size data
ES No. 1 2
ES type Post-test Post-test at 4-month post intervention
Construct Cardiorespiratory fitness Physical activity
Instrument Bruce treadmill test [ml VO2/min] 7-day physical activity log [hours/wk]
N intervention group 22 22
N control group 19 19
Mean intervention group 1810.1 5.68
Mean control group 1630.4 4.35
SD intervention group 369.4 2.28
SD control group 351.5 2.79
Appendix 8
188
MacVicar 1986 – Study descriptors
METHODS
Type of trial Controlled clinical trial (non-randomised)
Configuration of trial Parallel group design
Method of randomisation n/a
Allocation concealment D
Outcome assessor blinding n/a
Timing of outcome assessments
after initiation of treatment
10 weeks
Length of post-intervention follow-
up period
0 weeks
Sample size, total 10
Sample size for each group - Intervention group: n=6
- Control group: n=4
Availability of sample size per group
at randomisation
Yes
Withdrawal-/Drop-out rate and
reasons
0%
Adherence n/a
Type of analysis n/a
Statistical tests Descriptive statistics: mean scores
Power calculation No
PARTICIPANTS
Recruitment n/a
Inclusion criteria - Breast cancer, surgery, chemotherapy, younger than 60 years
Exclusion criteria - Cardiovascular disease or hypertension
- Adriamycin
Age n/a
BMI/weight n/a
Breast cancer staging Stage II
Menopausal status n/a
Type of adjuvant cancer treatment Chemotherapy
Physical activity before cancer diag-
nosis
n/a
Co-morbidity n/a
INTERVENTIONS
Type of training Aerobic exercise: cycle ergometer; progressive interval training
Setting Laboratory based
Health promotion theory None
Delivery of intervention Tight supervision
Intensity 60-85% HRmax
Duration of sessions 20-30 minutes
Frequency 3 sessions per week
Total exercise period 10 weeks
CONTROL
Type of control No intervention
Intensity n/a
Duration of sessions n/a
Frequency n/a
Total period n/a
CO-INTERVENTION
Type of co-intervention n/a
OUTCOMES
Appendix 8
189
MacVicar 1986 – Study descriptors
Outcome measure concepts - Cardiorespiratory fitness
- Mood
Assessment instruments - Interval-training cycle ergometric protocol (Peak VO2)
- Profile of Mood States (POMS)
Adverse effects n/a (according to letter no adverse effects)
Similarity at baseline No
Appendix 8
190
MacVicar 1989, Winningham 1989 – Study descriptors
Publication 2 reports
METHODS
Type of trial RCT
Configuration of trial
- Parallel group design
- 3 groups (exercise, sham treatment and control)
Method of randomisation n/a; stratified by functional capacity (± 1 MET)
Allocation concealment B
Outcome assessor blinding n/a
Timing of outcome assessments
after initiation of treatment
10 weeks
Length of post-intervention
follow-up period
0 weeks
Sample size, total n=45 (MacVicar 1989) n= 24 (Winningham 1989)
Sample size for each group - Intervention group: n=18
- Control group placebo:
n=11
- Control group: n=16
- Intervention group: n=12
- Control group: n=12 (secon-
dary exclusions)
Availability of sample size per
group as randomised
No No
Withdrawal-/Drop-out rate and
reasons
All participants: 17/62 (27.4%)
Reasons: equipment breakdown; problems of transport; doxorubicin;
extreme reactions to chemotherapy; re-classification of breast cancer
stage
Adherence - Monitoring of intensity (HR), frequency, duration
- Adherence complete (missed sessions repeated)
Type of analysis MacVicar 1989: Per protocol
analysis (n= 45)
Winningham 1989: Secondary
analysis (subsample) per protocol
(n= 24)
Statistical tests
MacVicar 1989: Analysis of
covariance; covariate: pre-test
values
Winningham 1989: Analysis of
covariance; covariates: pre-test
values and age
Power calculation No
PARTICIPANTS
Recruitment Outpatient chemotherapy clinics
Inclusion criteria - Breast cancer stage II; histological established
- Surgery
- Chemotherapy (1cycle/4weeks <= intervention <=6month)
- Baseline functional capacity of 0-2 on Zubrod scale
- Karnofsky 50-100%
Exclusion criteria - Cardiotoxic drugs (doxorubicin)
- Extreme reactions to chemotherapy
- Uncontrolled cardiac, hypertensive or diabetic disease
- Regular exercise programme
- Secondary exclusions (Winningham 1989) of potential con-
founders related to body composition:
o History of thyroid function
o Tamoxifen use
o Skinfolds exceeded the capacity of the callipers
o Intestinal bypass surgery for obesity
Appendix 8
191
MacVicar 1989, Winningham 1989 – Study descriptors
Age
- Intervention group
- Control group placebo
- Control group
- All participants
MacVicar 1989:
- 45.4, SD=10.2
- 46.1, SD=10.3
- 43.8, SD=9.3
- 45.2
Winningham 1989:
- 45.6, SD=9.6
- 45.6, SD=9.9
- 45.6
Weight
- Intervention group
- Control group placebo
- Control group
Height
- Intervention group
- Control group placebo
- Control group
BMI
- Intervention group
- Control group placebo
- Control group
- All participants
MacVicar 1989:
- 69.9 kg, SD=14.2
- 64.0 kg, SD=8.8
- 65.7 kg, SD=13.1
- 161.2 cm, SD=7.0
- 155.4 cm, SD=13.0
- 163.2 cm, SD=6.0
- 26.97
- 26.5
- 24.67
- 26.04
Winningham 1989:
n/a
- 161.6 cm; SD=8.0
- 160.4 cm; SD=11.7
- n/a
- n/a
Breast cancer staging All participants: Stage II
Menopausal status
- Intervention group
- Control group placebo
- Control group
MacVicar 1989:
- n/a
- n/a
- n/a
Winningham 1989:
- Pre-menopausal: 5/12 (42%)
- Pre-menopausal: 4/12 (33%)
Type of adjuvant cancer treat-
ment
All participants
- Chemotherapy: 41/45 (91%)
- Hormonal therapy (Tamoxifen): 4/45 (9%)
Physical activity before cancer
diagnosis
Regular exercise programme as exclusion criterion
Co-morbidity n/a
INTERVENTIONS
Type of training Aerobic exercise: cycle ergometer, interval training
Setting Laboratory based
Health promotion theory None
Delivery of intervention Tight supervision
Intensity Alternating higher or lower exercise intensity, 60-85% HRmax
Duration of sessions 20-30 minutes
Frequency 3 sessions per week
Total exercise period 10 weeks
CONTROL Control group placebo Control group
Type of control Flexibility and stretching exer-
cises
No intervention
Intensity n/a n/a
Duration of sessions n/a n/a
Frequency n/a n/a
Total period n/a n/a
CO-INTERVENTION
Type of co-intervention No subject participated in any other exercise or rehabilitation pro-
gramme
OUTCOMES
Outcome measure concepts - Cardiorespiratory fitness (aerobic capacity, cardiac function,
work load resistance)
- Body composition
Assessment instruments - Winningham Aerobic Interval Training (WAIT) protocol:
Appendix 8
192
MacVicar 1989, Winningham 1989 – Study descriptors
VO2max [litre], time to achieve peak oxygen uptake, heart rate,
ECG
- Weight
- Skinfold measurements
Adverse effects n/a (according to letter no adverse effects)
Similarity at baseline Adequate
MacVicar 1989 – Effect size data
ES No. 1 2
ES type Change score Change score
Construct Body composition Cardiorespiratory fitness
Instrument Weight [kg] WAIT-protocol
N intervention group 12 18
N control group 12 16
Mean intervention group 0.82 0.425
Mean control group 1.99 -0.088
F-value (ANCOVA) 1.86
SD intervention group 2.10 999*
SD control group 2.10 999*
* Transformation of F-value not adequate, since 3 groups were compared with adjustment for pre-test values
Appendix 8
193
Mock 1997 – Study descriptors
METHODS
Type of trial Controlled clinical trial (non randomised)
Configuration of trial Parallel group design
Method of randomisation n/a
Allocation concealment D (alternate assignment)
Outcome assessor blinding n/a
Timing of outcome assessments
after initiation of treatment
- 3 weeks (mid-test)
- 6 weeks (post-test)
Length of post-intervention
follow-up period
0 weeks
Sample size, total n=46 (of 50 recruited participants)
Sample size for each group - Intervention group: n=22
- Control group: n=24
Availability of sample size per
group at randomisation
No
Withdrawal-/Drop-out rate and
reasons
4/50 (8%)
Reasons: time constraints, treatment side effects; withdrawal from
radiotherapy
Adherence - Monitoring of intensity (pulse rates), frequency, duration via
diaries
- Adherence defined as ≥ 30 minutes of aerobic exercise in 3 or
more sessions per week (intervention group) and ≤ 30 minutes in
4 sessions per week (control group)
- Adherence:
o Intervention group: 86%
o Control group: complete adherence
Type of analysis Available case analysis
Statistical tests
MANCOVA;
ANCOVA
Power calculation Yes
PARTICIPANTS
Recruitment - 2 University teaching hospitals
- Outpatient departments of radiation oncology
Inclusion criteria - Breast cancer stages I and II
- Breast-conserving surgery
- Radiotherapy
- 35-65 years
Exclusion criteria - Concurrent major health problems (e.g., cardiovascular disease,
acute or chronic respiratory disease, cognitive dysfunction)
- Already participating in structured exercise programme
Age - Intervention group: 48.09, SD=5.42
- Control group: 50.29, SD=8.47
- All participants: 49.2
BMI - Intervention group: 24
- Control group: 24
- All participants: 24
Breast cancer staging - Intervention group:
o Stage I: 17/22 (77%)
o Stage II: 5/22 (23%)
- Control group:
o Stage I: 16/24 (67%)
o Stage II: 8/24 (33%)
Appendix 8
194
Mock 1997 – Study descriptors
Menopausal status n/a
Type of adjuvant cancer treat-
ment
All participants: Radiotherapy
Physical activity before cancer
diagnosis
- Control group: several subjects regular walkers
- Already participating in structured exercise programme as exclu-
sion criterion
Co-morbidity n/a
INTERVENTIONS
Type of training Aerobic exercise: walking
Setting Home based
Health promotion theory None
Delivery of intervention - Research team taught and monitored exercise program
- Exercise session: self-paced, progressive program of brisk walk-
ing, followed by 5 minutes slo walking (cool-down)
- Subjects walked in their neighbourhoods, at a mall, or in a gym at
their own pace and individually prescribed duration
- Intensity was purposefully left under the control of the subject to
permit individualized adaptation to effects or cancer treatment
and to prevent overexertion
- Researchers encouraged adherence by recommending that sub-
jects walk with a partner for support
- Researchers maintained regular contact by telephone or during
clinic visits to assess exercise progress and adjust exercise pre-
scription as indicated and to encourage adherence
- Booklet
- To prevent a differential intervention effect, the usual care group
was contacted regularly to inquire about their health and general
response to treatment.
Intensity 60-80% HRmax; self paced; individualized based on level of physical
fitness (determined by 12-Minute-Walk Test), history of exercise and
age, left under control of subject (Borg Scale for Rating of Perceived
Exertion with target rating of 11-13) (cp. Watson and Mock 2004)
Duration of sessions 20-30 minutes
Frequency 4-5 sessions per week
Total exercise period 6 weeks
CONTROL
Type of control Usual care
Intensity n/a
Duration of sessions n/a
Frequency n/a
Total period n/a
CO-INTERVENTION
Type of co-intervention Regular contact (both groups)
OUTCOMES
Outcome measure concepts - Fatigue
- Physical activity
- Physical fitness (ability to ambulate and perform ADL)
- Emotional distress
- Symptom experience (pain, skin change, nausea, vomiting, fatigue,
diarrhoea, difficulty sleeping, irritability, depression, mouth sores,
anxiety, constipation, satisfaction with body)
Appendix 8
195
Mock 1997 – Study descriptors
Assessment instruments - R-PFS, Visual Analogue Scale (VAS) fatigue
- 12-minute walk test
- Exercise rating scale (0-10) based on self-report of minutes/day
and days/week spent exercising
- Diaries of exercise periods with pulse rates and subjective data
- Symptom Assessment Scales (SAS)
Adverse Effects No physical injury related to the walking programme
Similarity at baseline Adequate
COMMENTS
Conceptual framework Roy adaptation model
Mock 1997 – Effect size data
ES No. 1 2 3 4 5
ES type Post-test Post-test Post-test Post-test Post-test
Construct Fatigue Cardiorespiratory
fitness Anxiety Depression
Sleep distur-
bances
Instrument VAS
fatigue 12-MWT [feet] SAS anxiety SAS
depression
SAS sleep dis-
turbances
N intervention 22 22 22 22 22
N control group 24 22 24 24 24
Mean intervention 26.12 3371 10.44 9.51 12.38
Mean control group 43.05 3089 26.93 21.05 32.58
F-Value (ANCOVA) 9.69 5.11 2.77 5.39
SD intervention group 20.27 300.46 24.70 23.50 29.50
SD control group 36.37 300.46 24.70 23.50 29.50
Appendix 8
196
Mock 2005 – Study descriptors
METHODS
Type of trial RCT
Configuration of trial Parallel group design
Method of randomisation Computer-generated
Allocation concealment A (consecutively numbered sealed opaque envelopes, opened after
baseline testing)
Outcome assessor blinding No
Timing of outcome assessments
after initiation of treatment
- Radiotherapy: 6 weeks
- Chemotherapy: 3-6 month
Length of post-intervention
follow-up period
0 weeks
Sample size, total 119
Sample size for each group - Intervention group: n=60
- Control group: n=59
Availability of sample size for
each group at randomisation
Yes
Withdrawal-/Drop-out rate and
reasons
- Intervention group: 6/60 (10%)
- Control group: 5/59 (8.5%)
Reasons: patient request, moved, chemotherapy problems, chemo-
therapy withdrawal
Adherence - Monitoring of intensity (pulse rates, perceived exertion), fre-
quency, duration via diaries (lack of a physiologic activity moni-
tor)
- ACSM definition: 85% of minimum prescription
- Coding in trial:
o Intervention group: more than 60 minutes per week for
more than 2/3 of trial as adherent to exercise prescrip-
tion
o Control group: less than 45 minutes per week for more
than 2/3 of trial
- Assessment via daily diaries of exercise periods including pulse
rates, perceived exertion rates, and fatigue levels
- Intervention group:
o Not adherent: 15/54 (28%)
o Adherent: 39/54 (72 %) exercising 127.43 minutes per
week, SD=44.04, with a frequency of 4.59, SD=1.16 ses-
sions per week and a mean of 28.36 SD=9.04 minutes
per session
- Control group:
o Not adherent (= exercising): 21/54 (39%)
o Adherent: 33/54 (61%)
Type of analysis - ITT (Complier-average-causal effect as equivalent to ITT)
- Sub-Group Analysis: low exercisers vs. high exercisers
Statistical tests - 2 sample t-tests
- multiple linear regression (with adjustments for baseline covari-
ates)
- Instrumental variables with principal stratification (IV/PS)
- Complier-average-causal effect (CACE)
Power calculation Yes
PARTICIPANTS
Recruitment - 4 University teaching hospitals of National Cancer Institute
- 4 community cancer centres in eastern US
Appendix 8
197
Mock 2005 – Study descriptors
Inclusion criteria - 18-70 years
- Breast cancer stages 0-III
- Surgery
- Chemotherapy or radiotherapy
- Sedentary (exercising less than 45 minutes per week)
Exclusion criteria - Concurrent major health problems ( e. g. obesity, cardiovascular
disease, respiratory disease, cognitive dysfunction)
- Active exercising
Age Intervention group: 51.3, SD=8.9
Control group: 51.6, SD=9.7
All participants: 51.5, SD=9.3
BMI Intervention group: 25.5, SD=4.0
Control group: 25.8, SD=5.1
All participants: 25.7, SD=4.6
Breast cancer staging - Intervention group:
o Stage 0: 12/60 (20%)
o Stage I: 27/60 (45%)
o Stage II: 21/60 (35 %)
o Stage IIIa: 0/60 (0%)
- Control group
o Stage 0: 16/59 (27.2%)
o Stage I: 24/59 (40.7%)
o Stage II: 15/59 (25.4%)
o Stage IIIa: 4/59 (6.7%)
- All participants
o Stage 0: 28/119 (23.5%)
o Stage I: 51/119 (42.9%)
o Stage: 36/119 (30.2%)
o Stage IIIa: 4/119 (3.4%)
Menopausal status n/a
Type of adjuvant cancer treat-
ment
- Intervention group:
o Chemotherapy: 25/60 (41.7%)
o Radiotherapy: 35/60 (58.35%)
- Control group:
o Chemotherapy: 25/59 (42.4%)
o Radiotherapy: 34/59 (57.6%)
- All participants:
o Chemotherapy: 50/119 (42.0%)
o Radiotherapy: 69/119 (58.0%)
Physical activity before cancer
diagnosis [kcals/week]
- Sedentary (less than 40 minutes per week)
- Intervention group: 1657 kcals/week, SD=1262
- Control group: 1918 kcals/week, SD=1806
- All participants: 1789 kcals/week, SD=1559
Co-morbidity n/a
INTERVENTIONS
Type of training Aerobic exercise: walking
Setting Home based
Health promotion theory None
Delivery of intervention - Oncology nurses taught and monitored individualised written
prescription of exercise programme
- Video
- Booklet
- Research team contacted participants every two weeks to evaluate
the prescription and participant progress
Appendix 8
198
Mock 2005 – Study descriptors
- Prescription was adjusted if an exercise participant was ill or
stopped exercising for more than 3 days
- Control group: patients were encouraged to maintain current
levels of activity; patients were called every 2 weeks by the re-
search team (attentional control) and were asked about their can-
cer treatment experience; patients reporting unmanaged symp-
toms or other clinical problems were referred to their health care
provider for treatment
Intensity 50-70 % HR max
Duration of sessions 15 minutes increased to 30 minutes as training progressed
Frequency 5-6 sessions per week
Total exercise period - Radiotherapy: 6 weeks
- Chemotherapy: 3 – 6 month
CONTROL
Type of control No intervention
Intensity n/a
Duration of sessions n/a
Frequency n/a
Total period n/a
CO-INTERVENTION
Type of co-intervention - Usual care
- Regular contact (every 2 weeks)
OUTCOMES
Outcome measure concepts - Primary outcome: Fatigue
- Physical Function
- Cardiorespiratory fitness
- Physical activity
Assessment instruments - PFS
- SF-36 (physical function subscale)
- 12-minutes walk test
- Physical activity questionnaire (PAQ)
- Daily diaries of exercise periods, incl. pulse rates, perceived exer-
tion
- Borg Rating of Perceived Exertion for monitoring purposes
Adverse effects No adverse events attributable to walking exercise
Similarity at baseline Adequate (but difference for 12 min walk, was adjusted for)
Mock 2005 – Effect size data
ES No. 1
ES type Post-test
Construct Fatigue
Instrument PFS
N intervention group 54
N control group 54
Mean intervention group 3.5
Mean control group 3.7
SD intervention group 2.4
SD control group 2.6
Appendix 8
199
Mutrie 2007 – Study descriptors
METHODS
Type of trial RCT
Configuration of trial Parallel group design
Method of randomisation Stratified by hospital and therapy; randomised permuted blocks of
length 4 and 6; randomisation performed by telephone to an Interac-
tive Voice Response system implemented by the Robertson Centre for
Biostatistics.
Allocation concealment A
Outcome assessor blinding Yes (e.g., questionnaires in sealed envelopes, 12 week and follow up
measures taken by researchers who had not taught that participant in
classes)
Timing of outcome assessments
after initiation of treatment
12 weeks and 6 months post intervention
Length of post-intervention
follow-up period
6 months post intervention
Sample size, total 201
Sample size for each group - Intervention group: n=99
- Control group: n=102
Availability of sample size for
each group at randomisation
Yes
Withdrawal-/Drop-out rate and
reasons
- Intervention group:
o Week 12: 17/99 (17.2%)
o 6 months post intervention: 17/99 (17.2%)
- Control group:
o Week 12: 10/102 (9.8%)
o 6 months post intervention: 7/102 (6.9%)
Reasons: did not start classes, questionnaires not returned, non-
contactable, died, too ill
Adherence - Monitoring of intensity (HR), frequency, duration
- Participation in classes:
o > 70% classes: 39/99 (38.8%)
o 30-69% classes: 30/99 (30.6%)
o < 30% classes: 30/99 (30.6%)
Type of analysis Available case analysis
Statistical tests Mixed effects linear regression with predictor variables for stratifica-
tion variables, age, baseline outcome variable measurement, visit-
specific effects of intervention; including random participant effect
Power calculation Yes
PARTICIPANTS
Recruitment 3 clinical recruiters for outpatient chemotherapy and radiotherapy
clinics at three sites
Inclusion criteria n/a
Exclusion criteria n/a
Age - Intervention group: 51.3, SD=10.3
- Control group: 51.8, SD=8.7
- All participants: 51.6, SD=9.5
BMI - Intervention group: 27.3, SD=5.2
- Control group: 27.5, SD=6.0
- All participants: 27.4, SD=5.6
Breast cancer staging Early stage
Appendix 8
200
Mutrie 2007 – Study descriptors
Menopausal status Periods
- Intervention group:
o None: 82/99 (82.8%)
o Irregular: 10/99 (10.1%)
o Regular: 7/99 (7.1%)
- Control group:
o None: 87/102 (85.3%)
o Irregular: 7/102 (6.9%)
o Regular: 8/102 (7.8%)
- All participants:
o None: 169/201 (84.1%)
o Irregular: 17/201 (8.5%)
o Regular: 15/201 (7.5%)
Type of adjuvant cancer treat-
ment
- Intervention group:
o Chemotherapy: 8/99 (8.1%)
o Radiotherapy: 28/99 (28.3%)
o Combination: 63/99 (63.6%)
- Control group:
o Chemotherapy: 7/102 (6.9%)
o Radiotherapy: 29/102 (28.4%)
o Combination: 66/102 (64.7%)
- All participants:
o Chemotherapy: 15/201 (7.5%)
o Radiotherapy: 57/201 (28.4%)
o Combination: 129/201 (64.2%)
Physical activity before cancer
diagnosis
SPAQ leisure time activity
- Intervention group: 367.0 minutes, SD=305.7
- Control group: 364.9 minutes, SD=287.6
- All participants: 365.9 minutes, SD=295.9
Co-morbidity n/a
INTERVENTIONS
Type of training Aerobic and resistance exercise training: warm-up of 5-10 minutes, 20
minutes of exercise (including walking, cycling, low level aerobics,
muscle strengthening exercises, or circuits of specifically tailored
exercises), a cool down and relaxation period
Setting Community based
Health promotion theory Stages of change, self-efficacy
Delivery of intervention - 24 exercise classes, each week 14 classes
- Run by specifically trained exercise specialists
- 9 different locations, all accessible by public transport
- Time tabled at various times in the day and evening
- Individual physical activity counselling
- Workshops addressing health benefits, self-efficacy, exercise
barriers, supportive environment, setting goals, finding appropri-
ate activity options
Intensity 60-75% age-adjusted heart rate maximum
Duration of sessions 35-50 minutes
Frequency 2 sessions per week
Total exercise period 12 weeks
CONTROL
Type of control - Monitoring
- Exercise leaflet
Intensity n/a
Duration of sessions n/a
Appendix 8
201
Mutrie 2007 – Study descriptors
Frequency n/a
Total period n/a
CO-INTERVENTION
Type of co-intervention Usual care (both groups)
OUTCOMES
Outcome measure concepts - Primary outcome measure: Quality of Life
- Breast-cancer-specific symptoms
- Fatigue
- Endocrine symptoms
- Depression
- Emotional distress
- Body composition
- Cardiorespiratory fitness
- Physical activity
Assessment instruments - Functional Assessment of Cancer Therapy – General (FACT-G)
questionnaire
- FACT-B; FACT-F; FACT ES
- Beck Depression Inventory (BDI)
- Positive and negative affect scale (PANAS)
- BMI
- Scottish Physical Activity Questionnaire (SPAQ)
- 12 minute walk test
Adverse effects n/a
Similarity at baseline Yes
Mutrie 2007 – Effect size data
ES No. 1 2 3 4 5 6
ES type Post-test Post-test Post-test Post-test Post-test Post-test
Construct Cardiorespiratory
fitness
Body Com-
position
Quality of
life
Quality of
life Fatigue Depression
Instrument 12-MWT [meter] BMI [kg/m2] FACT-G FACT-B FACT-F BDI
N intervention
group 82 82 82 82 82 82
N control
group 92 92 92 92 92 92
Mean inter-
vention group 1135 26.9 81.0 106.5 120.8 8.6
Mean control
group 984 27.9 77.3 99.7 113.3 11.5
SD interven-
tion group 143 4.3 16.8 21.9 26.7 6.8
SD control
group 221 6.9 14.4 20.3 25.0 8.6
Appendix 8
202
Mutrie 2007 – Effect size data
ES No. 7 8 9 10
ES type Post-test Post-test Post-test Post-test
Construct Endocrine symptoms Positive affects Negative affects Physical activity
Instrument FACT-ES PANAS positive PANAS negative SPAQ [min]
N intervention group 82 82 82 82
N control group 92 92 92 92
Mean intervention group 122.1 33.4 15.6 585
Mean control group 117.6 29.3 17.7 416
SD intervention group 24.6 8.5 6.6 385
SD control group 22.2 9.8 7.4 405
Mutrie 2007 – Effect size data at 6-month follow-up
ES No. 1 2 3 4 5 6
ES type Post-test Post-test Post-test Post-test Post-test Post-test
Construct Cardiorespiratory
fitness
Body Com-
position
Quality of
life
Quality of
life Fatigue Depression
Instrument 12-MWT [meter] BMI [kg/m2] FACT-G FACT-B FACT-F BDI
N intervention
group 82 82 82 82 82 82
N control
group 95 95 95 95 95 95
Mean inter-
vention group 1127 27.0 83.2 109.4 124.6 8.4
Mean control
group 1013 27.0 77.1 101.2 114.3 10.8
SD interven-
tion group 166 4.6 12.8 16.5 20.8 7.2
SD control
group 190 5.4 17.0 21.7 28.1 7.5
Mutrie 2007 – Effect size data at 6-month follow-up
ES No. 7 8 9 10
ES type Post-test Post-test Post-test Post-test
Construct Endocrine symptoms Positive affects Negative affects Physical activity
Instrument FACT-ES PANAS positive PANAS negative SPAQ [min]
N intervention group 82 82 82 82
N control group 95 95 95 95
Mean intervention group 123.8 33.0 15.7 492
Mean control group 116.8 29.2 17.4 427
SD intervention group 20.3 8.1 6.1 327
SD control group 24.4 10.5 6.9 370
Appendix 8
203
Payne 2008 – Study descriptors
METHODS
Type of trial - RCT (method unclear)
Configuration of trial
- Parallel group
Method of randomisation n/a
Allocation concealment - B: Unclear
Outcome assessor blinding - N/a
Timing of outcome assessments after
initiation of treatment
- 14 weeks
Length of post-intervention follow-up
period
- 0 months
Sample size, total n=20
Sample size for each group Intervention group: n=10
Control group: n=10
Availability of sample size for each
group at randomisation
- Yes
Withdrawal-/Drop-out rate and rea-
sons
Intervention group: 10% (1/10)
Control group: 10% (1/10)
All participants: 10% (2/20); reasons: worsening health issues
Adherence - Exercise logs (duration, frequency), pedometer
- No adherence data available
Type of analysis - Available case analysis
Statistical tests - Repeated measures analysis of variance
- repeated measures mixed effect-models
PARTICIPANTS
Recruitment University;
National Institutes for Health,
Comprehensive Cancer Center,
Southeastern United States
Inclusion criteria - Breast cancer
- Hormonal therapy with tamoxifen, anastrozole, letrozole
- Postmenopausal
- Older than 55 years
- Fatigue
- Karnofsky Performance Scale > 80
- English speaking
Exclusion criteria - Neuromuscular deficits
- Documented history of neurologic deficits or mental illness
within the past year
Age All participants: 65 (range 56-78 years)
BMI n/a
Breast cancer staging n/a
Menopausal status All participants: postmenopausal
Type of adjuvant cancer treatment All participants: hormonal therapy
Physical activity before cancer diagno-
sis
n/a
Co-morbidity n/a
INTERVENTIONS
Type of training walking
Setting - Home-based
Health promotion theory n/a
Delivery of intervention n/a
Appendix 8
204
Payne 2008 – Study descriptors
Intensity - moderate
Duration of sessions 20 minutes
Frequency 4 days per week
Total exercise period 14 weeks
CONTROL
Type of control Usual care (standard interactions with nurses, physicians and
staff)
Intensity n/a
Duration of sessions n/a
Frequency n/a
Total period n/a
CO-INTERVENTION
Type of co-intervention n/a
OUTCOMES
Outcome measures concepts - Fatigue
- Depressive Symptoms
- Sleep disturbances
- Biomarkers (cortisol, serotonin, interleukin-6, bilirubin)
Assessment instruments - Piper Revised Fatigue Scale (R-PFS)
- Center for Epidemiological Studies-Depression Scale (CES-
D)
- Sleep watch actigraphs
- Pittsburgh Sleep Quality Index (PSQI)
- Routine laboratory measures
- Radioimmunoassay analysis (serotonin, interleukin-6)
Adverse effects - Assessment of harm: n/a
- Reporting of harm: no
Similarity at baseline
- n/a
Payne 2008 – Effect size data
ES No. 1 2 3
ES type Post-test Post-test Post-test
Construct Sleep disturbances Fatigue Depressive symptoms
Instrument Pittsburgh Sleep Quality
Index Piper Fatigue Scale
Center for Epidemiol-
ogical Studies-
Depression Scale
N exercise group 9 9 9
N control group 9 9 9
Mean exercise 11.2 4.65 12.7
Mean control group 9.1 3.51 11.4
SD exercise 4.1 2.63 8.7
SD control group 3.8 1.75 7.9
Appendix 8
205
Schwartz 2007 – Study descriptors
METHODS
Type of trial - RCT (method unclear)
Configuration of trial
- Parallel group
Method of randomisation n/a; stratification according to menopausal status
Allocation concealment - A: Adequate
Outcome assessor blinding - N/a
Timing of outcome assessments after
initiation of treatment
- 6 months
Length of post-intervention follow-up
period
- 0 months
Sample size, total n=72
Sample size for each group Intervention group 1 (aerobic): n=22
Intervention group 2 (resistance): n=21
Control group: n=23
Availability of sample size for each
group at randomisation
- Yes
Withdrawal-/Drop-out rate and rea-
sons
Intervention group 1 (aerobic): 8,3% (2/24)
Intervention group 2 (resistance): 8,7% (2/23)
Control group: 8% (2/25)
All participants: 8,3% (6/72); reasons: too busy (n=4), location
not convenient (n=2)
Adherence - Exercise logs (intensity, duration, frequency, type for aero-
bic exercise; number of repetitions per exercise, resistances
of the band, duration of session) in intervention groups,
daily activity logs in control group, caloric expenditure
measured using Caltrac Aceloremeters
- No adherence data available
Type of analysis - Available case analysis (although ITT claimed in report)
Statistical tests Repeated measures analysis of variance and covariance (meno-
pausal status)
PARTICIPANTS
Recruitment University of Washington Cancer Center,
Oregon Health and Science University;
women were recruited before beginning adjuvant chemotherapy
Inclusion criteria Breast cancer stages I-III (histologically confirmed);
planning to begin chemotherapy with doxorubicin or meth-
otrexate
Exclusion criteria - Receiving steroids 6 months prior to study
- Paget disease
- Hyperparathyroidism
- Rheumatoid arthritis
- Ankylosing spondylitis, or other metabolic bone disease
- History of serious psychiatric illness
- Strenuous regular exercisers and women who exercised
more than 250 minutes per week
Age Intervention group 1 (aerobic): 48.32, SD=12.6
Intervention group 2 (resistance): 50.1, SD=8.7
Control group: 46.26, SD=9.8
All participants: 48.17
Appendix 8
206
Schwartz 2007 – Study descriptors
BMI (Weight) Intervention group 1 (aerobic): 69.80 kg, SD=13.6
Intervention group 2 (resistance): 77.5 kg, SD=17.3
Control group: 68.40 kg, SD=12.3
All participants: 71.76
Breast cancer staging - Stage I:
- Intervention group 1 (aerobic): 4/22 (18%)
- Intervention group 2 (resistance): 6/21 (28%)
- Control group: 5/23 (22%)
- All participants: 15/66 (22.7%)
- Stage II:
- Intervention group 1 (aerobic): 13/22 (59%)
- Intervention group 2 (resistance): 11/21 (52%)
- Control group: 14/23 (61%)
- All participants: 38/66 (57.6%)
- Stage III:
- Intervention group 1 (aerobic): 5/22 (12%)
- Intervention group 2 (resistance): 4/21 (19%)
- Control group: 4/23 (17%)
- All participants: 13/66 (19.7%)
Menopausal status For each group and for all participants
- Pre- menopausal:
- Intervention group 1 (aerobic): 11/22 (50%)
- Intervention group 2 (resistance): 13/21 (61%)
- Control group: 12/23 (52%)
- All participants: 36/66 (54.5%)
Type of adjuvant cancer treatment For each group and for all participants
- Chemotherapy: 66/66 (100%)
Physical activity before cancer diagno-
sis
For each group and for all participants
- Regular exerciser:
- Intervention group 1 (aerobic): 13/22 (59%)
- Intervention group 2 (resistance): 11/21 (52%)
- Control group: 12/23 (52%)
- All participants: 36/66 (54.5%)
Co-morbidity n/a
INTERVENTIONS
Type of training - Intervention group 1: Aerobic (participant preferences);
77% weight bearing activities!
- Intervention group 2: Resistance exercise (Thera band), 2
sets of 8 exercises (4 upper and 4 lower body)
Setting - Home-based
Health promotion theory n/a
Delivery of intervention - Women in the aerobic exercise intervention group were
instructed to choose an aerobic activity they enjoyed (par-
ticipants preferences)
- Group was instructed to use symptoms (e.g., fatigue, pain
breathlessness) to moderate exercise intensity and deter-
mine whether to stop exercising
- Resistance exercise subjects were given 2 different sets of
exercises to alternate the exercise sets within each week
- Research associate called exercisers at 2 week intervals for
the first month and then monthly thereafter to answer
questions about exercise and assess any barriers and the
ability to exercise.
Appendix 8
207
Schwartz 2007 – Study descriptors
Intensity - Aerobic exercise: symptom-limited, moderate intensity,
breathing hard, but able to talk; progressive via increasing
intensity vs. duration
- Resistance exercise: 8-10 repetitions per exercise, 2 sets
Duration of sessions Aerobic exercise: 15-30 minutes
Frequency Aerobic exercise: 4 days per week
Total exercise period 6 months (24 weeks)
CONTROL
Type of control Usual care
Intensity women were instructed to continue usual activities, were not
instructed to avoid exercise
Duration of sessions n/a
Frequency n/a
Total period n/a
CO-INTERVENTION
Type of co-intervention n/a
OUTCOMES
Outcome measures concepts - Bone mass density
- Aerobic fitness
- Strength
Assessment instruments - Dual-energy x-ray absorptiometry (lumbar spine, g/cm2)
- 12-minute walk test
- 1 repetition Maximum
Adverse effects - Assessment of harm: n/a
- Reporting of harm: yes
- Observation of harm: no new onset of lymphedema or
acute flares
Similarity at baseline
- Adequate
Schwartz 2007 – Effect size data
ES No. 1 2 3
ES type Post-test Post-test Post-test Post-test
Construct Cardiorespiratory
fitness
Bone mineral
density
Strength (leg
extension) [kg]
Strength (over-
head press) [kg]
Instrument 12-MWT [meter] Dual x-ray
absorptiometry
[g/cm2]
1 RM 1 RM
N aerobic exercise group 22 22 22 22
N resistance exercise gr. 21 21 21 21
N control group 23 23 23 23
Mean aerobic exercise 1228 0.98 78.6 13.7
Mean resistance exercise 1055 0.99 75.3 10.8
Mean control group 944 0.97 70.5 9.5
SD aerobic exercise 322 0.069 30.5 6.4
SD resistance exercise 177 0.120 34.5 5.1
SD control group 241 0.105 28.1 4.1
Appendix 8
208
Segal 2001 – Study descriptors
METHODS
Type of trial RCT
Configuration of trial
Parallel group design
Method of randomisation Random numbers table
Allocation concealment A (Study coordinator revealed group assignment after baseline testing)
Outcome assessor blinding n/a
Timing of outcome assessments
after initiation of treatment
26 weeks
Length of post-intervention
follow-up period
0 weeks
Sample size, total 123
Sample size for each group - Intervention group home based: n=40
- Intervention group centre based: n=42
- Control group: n=41
Availability of sample size per
group at randomisation
Yes
Withdrawal-/Drop-out rate and
reasons
- Intervention group home based: 7/40 (17.5%)
- Intervention group centre based: 10/42 (23.1%)
- Control group: 7/41 (17.5%)
No reasons given
Adherence - Monitoring of intensity, frequency
- Assessment via exercise diaries
- Adherence in intervention groups:
o Home based: 93/130 sessions (71.5%)
o Centre based: 93/130 sessions (71.5%)
Type of analysis ITT (most recent observed)
Statistical tests
- 1-way ANOVA to compare change scores between groups
- Dunnett’s t-test where the 2 exercise groups were each compared
against the control group
Power calculation Yes
PARTICIPANTS
Recruitment Medical oncologists
Inclusion criteria - Breast cancer stages I and II
- Radiotherapy, hormonal therapy, or chemotherapy
Exclusion criteria - Dose-intensity chemotherapy regimen
- Severe cardiac disease, uncontrolled hypertension
Age
Age total
- Intervention group home based: 51.0, SD=8.7
- Intervention group centre based: 51.4, SD=8.7
- Control group: 50.3, SD=8.7
- All participants: 50.9
Weight [kg] - Intervention group home based: 65.6kg, SD=13.6
- Intervention group centre based: 73.4kg, SD=15.0
- Control group: 71.6kg, SD=17.9
Breast cancer staging Breast cancer stages I, II
Menopausal status n/a
Appendix 8
209
Segal 2001 – Study descriptors
Type of adjuvant cancer treat-
ment
- Chemotherapy
o Intervention group home based: 32/40 (80.0%)
o Intervention group centre based: 34/42 (81.0%)
o Control group: 30/41 (73.2%)
- Radiotherapy
- Hormonal therapy
Physical activity before cancer
diagnosis
- Intervention group home based: 20/40 (50.0 %)
- Intervention group centre based: 25/42 (60.0 %)
- Control group: 19/41 (47.6 %)
Co-morbidity n/a
INTERVENTIONS
Type of training 1. Aerobic exercise: walking, self directed
2. Aerobic exercise: walking, supervised
Setting 1. Home based
2. Centre based
Health promotion theory None
Delivery of intervention 1. Prescription of an individualised walking programme, exercise
specialists showed a series of warm-up and cool-down exercises,
contact every 2 weeks; exercise specialist conducted interim fit-
ness evaluation at 13 weeks; contacted participants by telephone
every 2 weeks, checked progress and addressed barriers
2. Supervised progressive walking programme; exercise sessions with
warm-up and cool-down
Intensity 1. 50-60% VO2max
2. 50-60% VO2max
Duration of sessions 1. 7-10 minutes warm-up, walking, cool down; n/a
2. 7-10 minutes warm-up, walking, cool down (20-30 minutes walk-
ing; based on protocol information of 35-45 minutes total ses-
sion)
Frequency 1. 5 sessions per week
2. 5 sessions per week (3 supervised +2 self directed)
Total exercise period 1. 26 weeks
2. 26 weeks
CONTROL
Type of control Usual care (general advice from oncologist about benefits of exercise)
Intensity n/a
Duration of sessions n/a
Frequency n/a
Total period n/a
CO-INTERVENTION
Type of co-intervention Regular contact (all groups)
OUTCOMES
Outcome measure concepts - Quality of life (generic and cancer, cancer-site-specific)
- Cardiorespiratory fitness
- Body composition (weight)
Assessment instruments - Primary outcome: SF-36 physical functioning scale
- SF-36 other scales
- FACT-G, FACT-B
- Modified Canadian Aerobic Fitness test (mCAFT): ml/kg/min
- Scale
Adverse effects No (“no adverse events recorded”)
Similarity at baseline Adequate (but baseline differences in physical functioning)
Appendix 8
210
Segal 2001 – Effect size data
ES No. 1 2 3
ES type Post-test Post-test Change Score
Construct Cardiorespiratory fitness Fatigue Body composition
Instrument mCAFT SF-36 vitality Weight [kg]
N self-directed exercise 40 40 40
N supervised exercise 42 42 42
N control group 41 41 41
Mean self-directed exercise 26.3 60.8 0.4
Mean supervised exercise 26.2 55.8 -1.4
Mean control group 25.1 61.6 0.6
CI self-directed exercise -0.7 - 1.6
CI supervised exercise -3.3 - 0.5
CI control group -1.3 - 2.5
SD self-directed exercise 5.3 23.5 3.71
SD supervised exercise 5.1 24.0 6.28
SD control group 6.1 17.7 6.21
Appendix 8
211
Winningham 1988 – -Study descriptors
METHODS
Type of trial RCT
Configuration of trial - Parallel group design
- 3 groups (exercise, sham treatment and control)
Method of randomisation Matching on age and functional capacity
Allocation concealment B
Outcome assessor blinding n/a
Timing of outcome assessments after
initiation of treatment
10 weeks
Length of post-intervention follow-up
period
0 weeks
Sample size, total 42
Sample size for each group - Intervention group: n=16
- Control group placebo: n=14
- Control group: n=12
Availability of sample size per group at
randomisation
Yes
Withdrawal-/Drop-out rate and rea-
sons
0
Adherence n/a
Type of analysis ITT
Statistical tests
Repeated measures ANOVA,
Duncan’s Multiple Range Test (post hoc),
Chi2 to compare nausea response between groups
Power calculation No
PARTICIPANTS
Recruitment - University Medical Centre
- Medical clinics
- Private practices in a large Midwestern city
Inclusion criteria - Breast cancer, histological established
- Surgery
- Chemotherapy (3 treatments <= intervention <=6month),
responding to adjuvant treatment
- Baseline functional capacity of 0-2 on Zubrod scale
- Karnofsky 60-100%
Exclusion criteria - Uncontrolled cardiac or hypertensive disease
- Doxorubicin
- Other regular exercise programme
Age - Intervention group: 46.1, SD=12.4
- Control group placebo: 48.2, SD=11.3
- Control group: 45.3, SD=9.0
- All participants: 46.6
Weight (kg)
Height
BMI
- Intervention group: 66.7 kg, SD=8.6
- Control group placebo: 65.6 kg, SD=9.3
- Control group: 64.6 kg, SD=13.5
- Intervention group: 161 cm, SD=7
- Control group placebo: 155 cm, SD=12
- Control group: 163 cm, SD=6
- Intervention group: 25.7
- Control group placebo: 27.3
- Control group: 24.3
- All participants: 25.8
Appendix 8
212
Winningham 1988 – -Study descriptors
Breast cancer staging - Intervention group:
o Stage II: 81%
o Stage III: 6%
o Stage IV: 13%
- Control group placebo:
o Stage II: 79%
o Stage III: 7%
o Stage IV: 14%
- Control group:
o Stage II: 92%
o Stage III: 0%
o Stage IV: 8%
Menopausal status n/a
Type of adjuvant cancer treatment Chemotherapy
Physical activity before cancer diagno-
sis
Other regular exercise programme as exclusion criterion
Co-morbidity n/a
INTERVENTIONS
Type of training Aerobic exercise: cycle ergometer training, interval training
Setting Laboratory based
Health promotion theory None
Delivery of intervention Winningham Aerobic Interval Training (WAIT) protocol,
tight supervision
Intensity 60-85% HRmax
Duration of sessions 20-30 minutes per session
Frequency 3 sessions per week
Total exercise period 10 weeks
CONTROL
Type of control - Control group placebo: conversational interaction; mild
stretching and flexibility exercise.
- Control group: Normal activities
Intensity
- Control group placebo: mild
- Control group: n/a
Duration of sessions
- Control group placebo: n/a
- Control group: n/a
Frequency
- Control group placebo:1 session per week
- Control group: n/a
Total period
- Control group placebo: 10 weeks
- Control group: n/a
CO-INTERVENTION
Type of co-intervention None (esp. no anti-emetic medication)
OUTCOMES
Outcome measure concepts - Nausea
- Somatisation
Assessment instruments Derogatis Symptom Check List-90 Revised (SCL-90-R) somati-
sation (SOM) scale
Adverse effects n/a (no worsening of nausea in experimental group; according
to letter no adverse effects at all)
Similarity at baseline Adequate
Appendix 8
213
Winningham 1988 – Effect size data
ES No. 1
ES type Post-test
Construct Physiological side-effects (nausea)
Instrument Nausea relief: Derogatis Symptom Check List-90 Revised (SCL-90-R)
N intervention group 16
N control group 26
N intervention group
with a successful outcome
8
N control group with
successful outcome
3
Appendix 8
214
9.8.2 Intervention quality
Aerobic training* Resistance training†
Study-ID Score 1 2 3 4 1 2 3 4
Battaglini 2007 4 9 9 9 9 9
Campbell 2005 3 9 9 9
Courneya 2007 4 9 9 9 9 9 9 9 9
Crowley 2003 4 9 9 9 9 9 9 9
Drouin 2002 4 9 9 9 9
Kim 2006 4 9 9 9 9
MacVicar 1986 4 9 9 9 9
MacVicar 1989 4 9 9 9 9
Mock 1997 4 9 9 9 9
Mock 2005 4 9 9 9 9
Mutrie 2007 3 9 9 9
Payne 2008 3 9 9 9
Schwartz 2007 3 9 9 9 9 9 9
Segal 2001 4 9 9 9 9
Winningham 1988 4 9 9 9 9
* Aerobic training: 1=intensity, 2=duration, 3=frequency. 4=program duration
† Resistance training: 1=repetitions (intensity), 2=sets (duration), 3=frequency. 4=program duration
Appendix 8
215
9.9 Appendix 9 – Characteristics of studies adressing harm
9.9.1 Reference list of included studies of harm
Study-ID Reference
Schmitz 2005 Schmitz KH, Ahmed RL, Hannan PJ, Yee D. Safety and efficacy of weight
training in recent breast cancer survivors to alter body composition, insulin,
and insulin-like growth factor axis proteins. Cancer Epidemiol Biomarkers
Prev 14(7):1672-80; 2005.
Ahmed RL, Thomas W, Yee D, Schmitz KH. Randomized controlled trial of
weight training and lymphedema in breast cancer survivors. J Clin Oncol 24:
2765-72; 2006.
Basen-Engquist
2006
Basen-Engquist K, Taylor CL, Rosenblum C, Smith MA, Shinn EH, Greis-
inger A, Gregg X, Massey P, Valero V, Rivera E. Randomized pilot test of a
lifestyle physical activity intervention for breast cancer survivors. Patient
Educ Couns 13: 13; 2006.
McKenzie 2003 McKenzie DC, Kalda AL. Effect of upper extremity exercise on secondary
lymphedema in breast cancer patients: a pilot study. J Clin Oncol 21: 463-6;
2003.
9.9.2 Reference list of excluded studies of harm
Study-ID Reference
Cheema 2006 Cheema BS, Gaul CA. Full-body exercise training improves fitness and qual-
ity of life in survivors of breast cancer. J Strength Cond Res 20: 14-21; 2006.
Harris 2000 Harris SR, Niesen-Vertommen SL. Challenging the myth of exercise-induced
lymphedema following breast cancer: a series of case reports. J Surg Oncol
74: 95-8; 2000.
Johansson 2005 Johansson K, Tibe K, Weibull A, Newton RC. Low intensity resistance exer-
cise for breast cancer patients with arm lymphedema with or without com-
pression sleeve. Lymphology 38: 167-80; 2005.
Turner 2004 Turner J, Hayes S, Reul-Hirche H. Improving the physical status and quality
of life of women treated for breast cancer: a pilot study of a structured exer-
cise intervention. J Surg Oncol 86: 141-6; 2004.
Appendix 8
216
9.9.3 Quality of studies of harm
Basen-Engquist 2006
Criterion Score
Nonbiased selection 1
Adequate description of population 1
Low loss to follow-up, and patients lost to follow-up ana-
lysed for adverse events
0
Adverse events prespecified and defined 1
Ascertainment technique adequately described 1
Nonbiased and accurate ascertainment of adverse event 1
Adequate statistical analysis of potential confounders 1
Adequate duration of follow-up 1
Total quality score = sum of scores (0 - 8) 7: Good
McKenzie 2003
Criterion Score
Nonbiased selection 1
Adequate description of population 1
Low loss to follow-up, and patients lost to follow-up ana-
lysed for adverse events
0 (n/a)
Adverse events prespecified and defined 1
Ascertainment technique adequately described 1
Nonbiased and accurate ascertainment of adverse event 0 (n/a)
Adequate statistical analysis of potential confounders 0
Adequate duration of follow-up 1
Total quality score = sum of scores (0 - 8) 5: Fair
Schmitz 2005; Ahmed 2006
Criterion Score
Nonbiased selection 1
Adequate description of population 1
Low loss to follow-up, and patients lost to follow-up ana-
lysed for adverse events
0
Adverse events prespecified and defined 1
Ascertainment technique adequately described 1
Nonbiased and accurate ascertainment of adverse event 1
Adequate statistical analysis of potential confounders 1
Adequate duration of follow-up 1
Total quality score = sum of scores (0 - 8) 7: Good
Appendix 8
217
9.9.4 Study descriptors and effect size data of studies of harm
Basen-Engquist 2006 – Study descriptors and effect size data
Type of adverse events
3 lymphedema
3
Inclusion criteria
1 post-treatment
1
Type of exercise n/a
Type of trial
1 randomised controlled trial
1
Sample Size
Total sample size
Intervention group sample size
Control Group sample size
60
35
25
Significance Tests
Mann-Whitney U
No. of increases in arm circumference
U=222.5, p=0.124 right arm;
U=252.5, p=0.411 left arm.
McKenzie 2003 – Study descriptors and effect size data
Type of adverse events
3 lymphedema
3
Inclusion criteria
1 post-treatment
1
Type of exercise
2 resistance training
3 lymphedema
2
3
Type of trial
1 randomised controlled trial
1
Sample Size
Total sample size
Intervention group sample size
Control Group sample size
14
7
7
Significance Tests
ANOVA
“no significant differences in the percent-
age change of measured arm volume”
between groups
Appendix 8
218
Schmitz 2005; Ahmed 2006 – Study descriptors and effect size data
Type of adverse events
1 injury
1
Inclusion criteria
1 post-treatment
1
Type of exercise
2 resistance training
2
Type of trial
1 randomised controlled trial
1
Sample Size
Total sample size
Intervention group sample size
Control Group sample size
85
42 (immediate exercise intervention)
43 (delayed intervention, by 6 month)
Proportions or Frequencies
- % of intervention group with an adverse event
- % of control group with an adverse event
- 10.5% per 6 month, 22.5% over 1 year
- 20% over first 6 month
Schmitz 2005; Ahmed 2006 – Study descriptors and effect size data
Type of adverse events
3 lymphedema
3
Inclusion criteria
1 post-treatment
1
Type of exercise
2 resistance training
2
Type of trial
1 randomised controlled trial
1
Sample Size
Total sample size
Intervention group sample size
Control Group sample size
45
23
22
Proportions or Frequencies
- n of intervention group with an adverse event
- n of control group with an adverse event
- n of intervention group with an adverse event
- n of control group with an adverse event
- 2 (incidence of lymphedema)*
- 1 (incidence of lymphedema)
- 0 (self reported symptoms)
- 3 (self reported symptoms)
Significance Tests
Chi squared
p=0.40 (incidence)
p=0.22 (symptoms
* “incidence” refers to clinician diagnosis of lymphedema
Appendix 10
219
9.10 Appendix 10 – Outcome assessment instruments
Name of tool Author/Year Domains or factors Items Scaling Scoring
Functional assessment of
cancer therapy (FACT-G)
Cella 1993 1. Physical well-being (PWB)
2. Social well-being (SWB)
3. Relationship with doctor (RWD)
4. Emotional well-being (EWB)
5. Functional well-being (FWB).
FACT-G = sum of PWB, SWB, RWD,
EWB, FWB scores
7
7
2
6
7
29
5-point Likert rating
scale (0-4)
0-112
Subscales and total scores are the
sums of items. High scores indicate
higher quality of life.
Functional assessment of
cancer therapy - Breast
(FACT-B)
Brady 1997 Breast cancer subscale (BCS).
FACT-B = sum of PWB, SWB, RWD,
EWB, FWB, BCS scores
9
38
5-point Likert rating
scale (0-4)
0-152
High scores indicate higher quality of
life.
Functional assessment of
cancer therapy - Fatigue
(FACT-F)
Cella 1998 Fatigue subscale (FS).
FACT-F = sum of PWB, SWB, RWD,
EWB, FWB, and fatigue subscale
13
41
5-point Likert rating
scale (0-4)
0-164
High scores indicate higher quality of
life.
Functional assessment of
cancer therapy - endocrine
symptoms (FACT-ES)
Fallowfield
1999
Endocrine subscale (ES).
FACT-ES = sum of PWB, SWB, RWD,
EWB, FWB, BCS, ES
18
56
5-point Likert rating
scale (0-4)
0-224
High scores indicate higher quality of
life.
Functional assessment of
cancer therapy-Anemia
scale (FACT-An)
Cella 1997 Anemia subscale (An).
FACT-An= sum 13 items FS and 7 items
An
FACT-An= sum of FACT-G and An
20
48
5-point Likert rating
scale (0-4)
0-192
0-52
FACT Fatigue subscale scores range
from 0 to 52, where higher scores
represent less fatigue
Revised Piper Fatigue Scale
(R-PFS)
Piper Fatigue Scale (PFS)
Piper 1998
Piper 1989
6. Behavioural/severity
7. Affective meaning
8. Sensory
9. Cognitive/mood.
Total fatigue score
6
5
5
6
22
40
11-point numerical self-
report, five open-ended
The 22-item are added together and
divided by 22 "0" to "10" scale
Severity Codes:
0 none
1-3 mild
4-6 moderate
7-10 severe
Appendix 10
220
Name of tool Author/Year Domains or factors Items Scaling Scoring
Beck Depression Inventory
(BDI)
Beck 1961 Depression 21 4-point Likert rating
scale (0-3)
0-63
Severity Codes:
11-17 moderate
18 clinically relevant
Center for Epidemiological
Studies–Depression Scale
(CES-D)
Radloff 1977 Depression 20 4-point Likert rating
scale (1-4)
0-60
Higher scores indicate more impair-
ment
Pittsburgh Sleep Quality
Index (PSQI)
Buysse 1989 Sleep disturbances
1. Subjective sleep quality
2. Sleep latency
3. Sleep duration
4. Habitual sleep efficiency
5. Sleep disturbances
6. Use of sleep medications
7. Daytime dysfunction
19
5
4-point Likert rating
scale (0-3)
0 no difficulty;
3 severe difficulty
Additional items rated by
a bed partner
0–21
Component scores are summed to
produce a global score;
Higher scores indicate higher sleep
disturbances;
A PSQI global score >5 is considered
to be suggestive of significant sleep
disturbance.
Spielberger State Anxiety
Inventory (STAI)
Spielberger
1983
State anxiety
20 4-point Likert rating
scale (1-4)
20-80
Rosenberg Self-Esteem
Scale
Rosenberg
1965
Self-esteem 10 4-point Likert rating
scale (0-3)
0-30
The higher the score, the higher the
self esteem
Profile of mood states
(POMS)
McNair, Lorr,
and Dropple-
man 1992
10. Tension-anxiety
11. Depression-dejection
12. Anger-hostility
13. Vigour-activity
14. Fatigue-inertia
15. Confusion-bewilderment.
65 Respondents rate 65
adjectives on a 5-point
intensity scale, in terms
of how they have been
feeling in the past week
(0=not at all and
4=extremely).
-32-200
Subscales and total scores are the
sums of items.
Total mood disturbance (TMD): sum
of the scores on the six subscales,
with vigour-activity negatively
weighted.
Except for vigour-activity, the higher
the score, the greater the mood dis-
turbance/more distress.
Positive and Negative
Affect Schedule (PANAS)
Watson, Clark,
and Tellegen
1988
Two 10-item mood scales
1. Positive affects (PA)
2. Negative affects (NA)
10
10
5-point Likert rating
scale (1-5)
1 very slightly/ not at all
5 very much
PA: 10-50
NA: 10-50
Appendix 11
221
9.11 Appendix 11 – Effect size calculation: inputs into meta-analyses
Outcome Study_ID Instrument
Body composition - Battaglini 2007
- Drouin 2002
- Courneya 2007 AET
- Courneya 2007 RET
- Mutrie 2007
- % BF
- BMI
- % BF
- % BF
- BMI
Fatigue - Battaglini 2007
- Campbell 2005
- Courneya 2007 AET
- Courneya 2007 RET
- Drouin 2002
- Mock 1997
- Mock 2005
- Mutrie 2007
- Payne 2008
- Segal 2001 SD
- Segal 2001 SU
- PFS
- PFS
- FACT-An (reversed)
- FACT-An (reversed)
- R-PFS
- VAS fatigue
- PFS
- FACT-F (reversed)
- R-PFS
- SF-36 vitality (reversed)
- SF-36 vitality (reversed)
Strength - Battaglini 2007
- Drouin 2002
- Courney 2007 AET
- Courneya 2007 RET
- Schwartz 2007 AET
- Schwartz 2007 RET
- 1-RM (4 exercises, kg)
- 1-RM (grip, kg)
- 1-RM (chest, kg)
- 1-RM (chest, kg)
- 1-RM (overhead press; kg)
Aerobic fitness - Campbell 2005
- Courneya 2007 AET
- Courneya 2007 RET
- Drouin 2002
- Kim 2006
- Mock 1997
- Mutrie 2007
- Schwartz 2007 AET
- Schwartz 2007 RET
- Segal 2001 SD
- Segal 2001 SU
- 12 MWT (m)
- VO2 (ml/kg/min)
- VO2 (ml/kg/min)
- VO2 (ml/kg/min)
- VO2 (ml/min)
- 12 MWT (feet)
- 12 MWT (m)
- 12 MWT (m)
- 12 MWT (m)
- VO2 (ml/kg/min)
- VO2 (ml/kg/min)
Depression - Courneya 2007 AET
- Courneya 2007 RET
- Payne 2008
- Mock 1997
- Mutrie 2007
- CES-D
- CES-D
- CES-D
- SAS depression
- BDI
Cancer-site-specific QoL - Campbell 2005
- Courneya 2007 AET
- Courneya 2007 RET
- Mutrie 2007
- FACT-B
- FACT-An
- FACT-An
- FACT-B
Appendix 12
222
9.12 Appendix 12 – Forest plots
9.12.1 Immediate post-intervention outcomes
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 15 Physical functioning
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Segal 2001 SD 40 81.80(17.40) 20 79.40(21.50) 48.74 0.13 [-0.41, 0.66]
Segal 2001SU 42 78.70(19.60) 21 79.40(21.50) 51.26 -0.03 [-0.56, 0.49]
Total (95% CI) 82 41 100.00 0.04 [-0.33, 0.42]
Test for heterogeneity: Chi² = 0.17, df = 1 (P = 0.68), I² = 0%
Test for overall effect: Z = 0.23 (P = 0.82)
-1 -0.5 0 0.5 1
Favours control Favours exercise
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 20 Nausea
Study Exercise Control RR (random) Weight RR (random)
or sub-category n/N n/N 95% CI % 95% CI
Winningham 1988 8/16 23/26 100.00 0.57 [0.34, 0.94]
0.1 0.2 0.5 1 2 5 10
Favours exercise Favours control
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 14 Sleep disturbances
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Mock 1997 22 12.38(29.50) 24 32.58(29.50) 100.00 -0.67 [-1.27, -0.08]
-4 -2 0 2 4
Favours exercise Favours control
Appendix 12
223
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 09 Endocrine symptoms
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Mutrie 2007 82 -41.60(9.10) 92 -40.30(9.70) 100.00 -0.14 [-0.44, 0.16]
-1 -0.5 0 0.5 1
Favours exercise Favours control
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 10 Natural killer cells
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Drouin 2002 13 8.50(9.20) 8 6.60(4.20) 100.00 0.24 [-0.65, 1.12]
-4 -2 0 2 4
Favours control Favours exercise
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 11 T-cells
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Drouin 2002 13 2.50(1.10) 8 1.80(1.00) 100.00 0.63 [-0.27, 1.54]
-4 -2 0 2 4
Favours control Favours exercise
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 12 Oxidative stress
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Drouin 2002 10 244.00(132.60) 5 316.40(118.30) 100.00 -0.53 [-1.63, 0.57]
-4 -2 0 2 4
Favours exercise Favours control
Appendix 12
224
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 21 Bone mineral density
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Schwartz 2007 RET 21 0.99(0.12) 11 0.97(0.11) 48.13 0.17 [-0.56, 0.90]
Schwartz 2007 AET 22 0.98(0.07) 12 0.97(0.11) 51.87 0.11 [-0.59, 0.82]
Total (95% CI) 43 23 100.00 0.14 [-0.37, 0.65]
Test for heterogeneity: Chi² = 0.01, df = 1 (P = 0.92), I² = 0%
Test for overall effect: Z = 0.54 (P = 0.59)
-1 -0.5 0 0.5 1
Favours control Favours exercise
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 08 Mood disturbance
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Drouin 2002 13 5.10(22.10) 8 23.90(32.00) 100.00 -0.69 [-1.60, 0.22]
-4 -2 0 2 4
Favours exercise Favours control
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 16 Self esteem
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Courneya 2007 AET 74 34.50(5.10) 36 33.20(5.50) 49.44 0.25 [-0.15, 0.65]
Courneya 2007 RET 76 34.70(4.20) 37 33.20(5.50) 50.56 0.32 [-0.08, 0.71]
Total (95% CI) 150 73 100.00 0.28 [0.00, 0.56]
Test for heterogeneity: Chi² = 0.06, df = 1 (P = 0.80), I² = 0%
Test for overall effect: Z = 1.98 (P = 0.05)
-1 -0.5 0 0.5 1
Favours control Favours exercise
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 01 Exercise versus control
Outcome: 18 Positive affects
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Mutrie 2007 82 33.40(8.50) 92 29.30(9.80) 100.00 0.44 [0.14, 0.74]
-1 -0.5 0 0.5 1
Favours control Favours exercise
Appendix 12
225
9.12.2 Long-term outcomes
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 02 Long-term effects
Outcome: 04 Cardiorespiratory fitness
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Mutrie 2007 82 1127.00(166.00) 95 1013.00(190.00) 100.00 0.63 [0.33, 0.94]
-1 -0.5 0 0.5 1
Favours control Favours exercise
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 02 Long-term effects
Outcome: 05 BMI
Study Exercise Control WMD (random) Weight WMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Mutrie 2007 82 27.00(4.60) 95 27.00(5.40) 100.00 0.00 [-1.47, 1.47]
-4 -2 0 2 4
Favours exercise Favours control
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 02 Long-term effects
Outcome: 06 Endocrine symptoms
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Mutrie 2007 82 -41.00(9.80) 95 -39.70(10.20) 100.00 -0.13 [-0.42, 0.17]
-1 -0.5 0 0.5 1
Favours exercise Favours control
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 02 Long-term effects
Outcome: 07 Positive affects
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Mutrie 2007 82 33.00(8.10) 95 29.20(10.50) 0.00 0.40 [0.10, 0.70]
-1 -0.5 0 0.5 1
Favours control Favours exercise
Appendix 12
226
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 02 Long-term effects
Outcome: 08 Negative affects
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Mutrie 2007 82 15.70(6.10) 95 17.40(6.90) 100.00 -0.26 [-0.56, 0.04]
-1 -0.5 0 0.5 1
Favours exercise Favours control
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 02 Long-term effects
Outcome: 10 Self esteem
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Courneya 2007 AET 68 35.00(4.70) 30 33.90(5.60) 49.52 0.22 [-0.21, 0.65]
Courneya 2007 RET 73 35.20(4.60) 30 33.90(5.60) 50.48 0.26 [-0.16, 0.69]
Total (95% CI) 141 60 100.00 0.24 [-0.06, 0.54]
Test for heterogeneity: Chi² = 0.02, df = 1 (P = 0.89), I² = 0%
Test for overall effect: Z = 1.56 (P = 0.12)
-1 -0.5 0 0.5 1
Favours control Favours exercise
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 02 Long-term effects
Outcome: 12 Anxiety
Study Exercise Control SMD (random) Weight SMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI
Courneya 2007 AET 68 32.20(11.20) 30 37.40(12.00) 48.97 -0.45 [-0.89, -0.02]
Courneya 2007 RET 73 35.50(13.00) 30 37.40(12.00) 51.03 -0.15 [-0.57, 0.28]
Total (95% CI) 141 60 100.00 -0.30 [-0.60, 0.01]
Test for heterogeneity: Chi² = 0.95, df = 1 (P = 0.33), I² = 0%
Test for overall effect: Z = 1.91 (P = 0.06)
-1 -0.5 0 0.5 1
Favours exercise Favours control
Appendix 12
227
Review: Exercise for women receiving adjuvant therapy for breast cancer
Comparison: 02 Long-term effects
Outcome: 09 Lymphedema incidence
Study Exercise Control RR (random) Weight RR (random)
or sub-category n/N n/N 95% CI % 95% CI
Courneya 2007 RET 3/82 3/41 41.06 0.50 [0.11, 2.37]
Courneya 2007 AET 7/78 3/41 58.94 1.23 [0.33, 4.49]
Total (95% CI) 160 82 100.00 0.85 [0.31, 2.30]
Total events: 10 (Exercise), 6 (Control)
Test for heterogeneity: Chi² = 0.75, df = 1 (P = 0.39), I² = 0%
Test for overall effect: Z = 0.32 (P = 0.75)
0.1 0.2 0.5 1 2 5 10
Favours exercise Favours control
Eidesstattliche Erklärung
228
Eidesstattliche Erklärung
Ich versichere hiermit, dass ich diese Arbeit selbst angefertigt habe und keine anderen als die
angegebenen Quellen und Hilfsmittel benutzt sowie die wörtlich oder inhaltlich übernommenen
Stellen als solche kenntlich gemacht habe.
Ich habe keine Promotionsversuche an anderen Universitäten unternommen.
...............................................................................................
Datum, Unterschrift
