A Randomized Controlled Graded Exercise Trial for Chronic Fatigue Syndrome:
Outcomes and Mechanisms of Change

Journal of Health Psychology, Vol 10(2) 245-259

RONA MOSS-MORRIS, CYNTHIA SHARON, ROSEANNE TOBIN, & JAMES C.
BALDI, all of the University of Auckland, New Zealand

RONA MOSS-MORRIS is a senior lecturer in Health Psychology at the
University of Auckland and has been conducting research into CFS and
related disorders for the past 10 years.

CYNTHIA SHARON completed her MSc and internship in health psychology in
Auckland, and is currently involved in the design and implementation of a
problem-solving therapy for patients following attempted suicide.

ROSEANNE TOBIN has an MSc (Health Psychology) from the University of
Auckland and is currently working as a researcher for the Health and Social
Care Advisory Service in London.

JAMES C. BALDI is a lecturer in the Department of Sport and Exercise
Science at the University of Auckland and the director of the Auckland
Cardiac Rehabilitation Clinic.

ACKNOWLEDGEMENTS. This study was supported in part by two University of
Auckland Staff Grants. Thank you Dr Rosamund Vallings for assisting with
recruitment for the study.

COMPETING INTERESTS: None declared.

ADDRESS. Correspondence should be directed to:
DR RONA MOSS-MORRIS, Health Psychology, The Faculty of Medical and Health
Sciences, The University of Auckland, Private Bag 92 019, Auckland, New
Zealand.
[email: r.moss-morris@auckland.ac.nz ]


Copyright © 2005 SAGE Publications London, Thousand Oaks and New Delhi,
www.sagepublications.com

DOl: 10.1177/1359105305049774

Abstract
The aim of this study was to investigate the potential mechanisms
underlying the efficacy of graded exercise therapy for chronic fatigue
syndrome (CFS). Forty-nine CFS patients were randomized to a 12-week graded
exercise programme or to standard medical care. At the end of treatment the
exercise group rated themselves as significantly more improved and less
fatigued than the control group. A decrease in symptom focusing rather than
an increase in fitness mediated the treatment effect. Graded exercise
appears to be an effective treatment for CFS and it operates in part by
reducing the degree to which patients focus on their symptoms.

Keywords: chronic fatigue syndrome (CFS), cognitive behavioural model,
deconditioning, graded exercise therapy, illness perceptions, randomized
controlled trial, symptom focusing

[...]

Discussion:
The results from this study suggest that a simple graded exercise
intervention is a more effective treatment for CFS than standard medical
care on its own. Between 50 to 55 per cent of CFS patients reported that
they were much or very much better after the 12 weeks of exercise therapy
compared to 24 per cent of controls. The intervention produced a
statistically significant reduction in patients' physical and mental
fatigue and these gains were maintained six months post-treatment. Patients
in the intervention group also improved their levels of physical
functioning, but not significantly more than the standard care control
group. The lack of significant findings in this area may be related to the
small sample size. Alternately, it may be that exercise therapy is a more
effective treatment for fatigue rather than disability. Wearden et al.
(1998) also reported changes in fatigue rather than disability in their CFS
graded exercise trial.

Comparisons with Fulcher and White's (1997) data at the end of their
12-week treatment period suggests that the gains made in the current study
were equivalent to those of this earlier study in relation to global
ratings of improvement and physical disability. The improvements in fatigue
appeared to be even greater in the current study. However, patients in the
Fulcher and White (1997) study continued to improve over the six-month
follow-up period, while the patients in the current study only maintained
their improvements. The lack of continued improvement may be a ceiling
effect in the current study in that the mean fatigue score at end of
treatment was less than 14, the recommended cut-off for case level of
fatigue. Alternatively, it maybe that around 40 per cent of patients in the
current study had a probable comorbid anxiety or depressive disorder.
Fulcher and White (1997) excluded patients with a concurrent depressive and
anxiety disorder. Dysphoria measured by the HADS has been shown to predict
poor outcome following exercise treatment (Bentall, Powell, Nye, & Edwards,
2002).

The results from this study also suggest that one of the key mechanisms of
change during graded exercise therapy is a reduction in the extent to which
patients focus on their symptoms. Patients in the exercise treatment
significantly reduced the amount they focused on their symptoms, but there
was no change on this variable in the control group. A reduction in symptom
focusing was strongly associated with a reduction in mental and physical
fatigue. It was also related to the self-rated global improvement score and
to an increase in physical functioning. This finding is in accordance with
a study on CBT for CFS, which showed that CBT was associated with a
reduction in beliefs about the harmful effects of exercise and that this
reduction was associated with improved outcome (Deale, Chalder, & Wessely,
1998).

Whereas CBT is geared towards directly identifying and challenging fear
beliefs, the use of heart rate monitors in exercise interventions may
provide an alternative medium for altering symptom focusing and fear
beliefs. In this trial, patients were instructed to use the heart rate
monitor, rather than their symptoms, as an external cue to decide whether
they were exercising within a safe range.

Graded exercise therapy appeared to be less effective in altering patients'
control beliefs or their illness identity. However, if patients did alter
these beliefs they did substantially better. Increasing the perception that
the illness can be controlled was associated with a statistically
significant change in fatigue and disability. There was also a trend
suggesting that patients who decreased the number of symptoms associated
with their CFS also decreased their fatigue levels and rated themselves as
significantly better after treatment. These data suggest that including a
cognitive element in the exercise therapy aimed at challenging control and
identity beliefs may help to maximize the treatment effect. This idea is
further supported by the findings from the Liverpool exercise trial, which
reported particularly impressive improvements in their treatment group
(Powell et al., 2001). This may be due to the inclusion of structured
exercises that helped patients to challenge their negative beliefs about
symptoms (Powell, 2001).

With regards to the physiological data, the lack of an increase in aerobic
fitness following exercise therapy was unexpected. A similar training
programme in CFS patients did manage to show an improvement in fitness
(Fulcher & White, 1997). However, three things may help to explain this
discrepancy. First, many patients terminated their V02max test for reasons
other than maximal effort. As a result less than 25 per cent ofV02max tests
achieved a true maximum, as defined by physiological responses (Baldi et
al., 2003). Consequently, we used V02 peak as our measure of fitness.
Furthermore, many patients refused to repeat the max test after their
intervention, resulting in a large percentage of missing data. Thus, our
V02 peak data may not accurately reflect adaptations to the training
protocol. Second, it was also noted that many subjects had an exaggerated
heart rate response to the initial stages of the incremental exercise test,
which plateaued during the later stages of the test. While the reasons for
this are not known, the 'normal' response to incremental exercise is a
linear increase in heart rate. As a result, prescribed exercise
intensities, which were calculated as a percentage of maximum heart rate,
may have represented a lower relative intensity than expected, which may
explain a lack of improvement in aerobic fitness. Third, the first six
weeks of our programme focused predominantly on consistency of exercise and
it was only in the latter six weeks that intensity was increased. This
gentle approach may be important in reducing exercise-related fears in that
only 12 per cent of patients dropped out of the treatment. However, it may
mean that exercise programmes should be extended a further 6 weeks so that
patients have a 12-week period where they work on upgrading their intensity
of exercise.

Although there was not an overall effect for fitness, a proportion of
patients did get fitter and there was some evidence that this was
associated with greater reductions in fatigue and improvements in physical
functioning. The strongest association was with physical functioning. This
finding was also reported by Fulcher and White (2000) suggesting that
physical fitness may be more important in terms of increasing physical
capacity than decreasing fatigue.
The physiological change variable that appeared to be most important in
terms of treatment response in this study was heart rate. An increase in
percentage predicted maximal heart rate after the intervention was found to
mediate the group effect on fatigue, particularly mental fatigue. Patients
who achieved a higher maximal heart rate on the second test also showed
statistically significant improvements in functioning and rated themselves
as much improved. This suggests that improvement may be related to an
increased ability to exert themselves on the second test. The fact that
many patients failed to reach true maximum suggests that fear of exertion
may be a problem. Reducing this fear through graded exercise may be
reflected in an increase in the maximal heart rate achieved.

Certain limitations of the study should be noted. The sample size was
smaller than anticipated which could have affected our power to detect
significant differences and to test for mediation. Recruitment was severely
handicapped by the appearance of an Australian article in the lay CFS
literature suggesting that exercise therapy for CFS patients was harmful
and should be avoided. As all the patients were volunteers, they may
reflect a small proportion of patients who are prepared to engage in the
therapy. The exercise testing proved to be particularly difficult with this
population suggesting that future studies may be better to rely on
submaximal tests to collect more complete and accurate fitness data.

Despite these limitations, this study adds to the growing body of
literature suggesting that graded exercise therapy is an effective
treatment for CFS. It appears to have a particularly significant impact on
fatigue and global ratings of improvement and less impact on physical
functioning. The key mechanism for improvement appears to be psychological
rather than physiological. A reduction in symptom focusing and an increased
ability to exert oneself were significant mediators of the treatment
effect. Using heart rate monitors may help to facilitate this process.
Improving aerobic fitness appears to be less important as treatments gains
were made independently of improvements in aerobic fitness. However, if
patients do get fitter, they appear to make significant improvements in
their physical functioning.