Dear Sir,
Reid at al’s review of treatments for chronic fatigue syndrome (CFS) is timely and well-written1. However, having studied this subject for more than 15 years, I consider that their assessment of cognitive behaviour therapy (CBT) and graded exercise is neither fair to the literature, nor to their colleagues.
Reid et al claim that CBT “administered by highly skilled therapists in specialist centres is effective in people with CFS”. They justify this statement by citing two British trials which support the use of CBT and criticising an Australian study which does not. With regard to the latter, Reid at al imply that their Australian colleagues were not ‘highly skilled’ and state that they had failed to offer their patients a “cognitive reappraisal” of their beliefs. These comments are surprising, given the published correspondence between the researchers and two of the authors (Chalder, Wessely) a few years ago. In a letter responding to similar criticisms, the Australians noted that the treatment was designed by a psychiatrist with “extensive experience with patients with CFS and with cognitive-behavioural approaches to the treatment of depression.. .The treatment was administered by experienced psychiatrists”2. So, what was the evidence that the Australians were not highly skilled? Why should an experienced and sympathetic psychiatrist with a reasonable knowledge of CBT and CFS be a less effective therapist than a nurse or psychologist with specialist training in behaviour therapy?
The letter also refuted the other criticism, pointing out that attempts were indeed made to change “key attitudes such as the possible effects of physical activity”. Since this was also a major component of the two successful British trials, and given that changing irrational views about exercise is the only part of CBT which has been linked to a positive outcome in CFS3, it’s hard to understand why Reid et al chose to repeat this allegation at this time.
A more likely explanation for the discrepancy between the British and Australian trials is that they involved different populations. An analysis of the relevant literature suggests that the British trials seem to have included large numbers of low functioning patients suffering from phobic avoidance, whereas the studies conducted outside Britain tended to select more patients with neurological abnormalities, ongoing infections or immune activation4.
Lack of space does not permit me to go into detail, but here is just some of the evidence supporting this alternative view. Firstly, there are no publications by any of the authors involved in the British trials which indicate that they see, let alone study, patients who use sensible coping strategies and who are not guided by irrational or unhelpful beliefs. Secondly, one of the British reports actually states that although their patients had modest disability levels as measured by the Karnofsky scale, they spent on average 3.3 days per week in bed. Thirdly, research from elsewhere shows that most people with CFS use a variety of coping strategies and that they usually operate at or near their activity ceiling4. Indeed, studies which used objective measures of activity levels suggest that the type of patients included in the British trials constitute a relatively small minority of the total CFS population and that only the most severely ill (less than 25%) spend long periods in bed. Moreover, one study revealed that most people with CFS are no more inactive than patients suffering from mild/moderate multiple sclerosis5.
Further support for a more cautious assessment of CBT comes from another well-controlled trial, which was excluded from the review because wasn’t randomised6. It reported significant improvements, but only in the patients who were suffering from depression as well as CFS. According to the researchers, this effect “may not have been apparent in a trial with use of a randomized design”.
A dispassionate, evidence-based evaluation of treatments for CFS would not have ignored the existence of subgroups, and the apparent preference, in British research, for people with maladaptive beliefs, it would also have considered the methodological flaws of the British trials, such as the inclusion of patients with a disability score of 78 where successful outcome was defined as reaching 8O. Moreover, it might have mentioned the most recent data from one of the trials cited in support of CBT. According to the reviews of a conference held in October 1999, the main author revealed that no significant differences were found between treated patients and the no-treatment controls at the second follow-up. Instead of relaying this information, the review informed us of the imminent publication, but not the results, of two further RCTs.
Regarding graded exercise, the review again overlooked important methodological flaws in the two studies suggesting benefit, e.g. the focus on fatigue but not somatic symptoms, the lack of data on the severely affected, the mixed populations etc. Moreover, the authors may have not found “evidence that exercise is harmful in people with CFS” but this does not mean that none exists. In fact, there are at least two recent studies which support the considerable number of anecdotal reports describing increasing weakness and relapse following exercise.
Perhaps the problems associated with graded exercise could have been discussed in the space devoted to ‘prolonged rest’? Since doctors haven’t advised long periods of rest since 1989 and given there is no evidence from the research that the majority of patients spend most of their time in bed, why was this included in the review? And why did the authors not mention the most commonly used coping strategy for CFS, namely, pacing?
In terms of ‘best evidence’, one can support the view that CBT and graded activity are beneficial for low functioning patients whose activity may be limited by concurrent depression and/or phobic avoidance. However, a blanket recommendation covering everyone with CFS is simply not justified.
In my opinion, Reid et al showed a lack of objectivity and caution. They offered readers a distorted view of the research on CBT and exercise, and accordingly misled their colleagues. Such a biased interpretation of the research wouldn’t be acceptable if we were discussing cancer, MS or even headaches. To quote a colleague: “it is time for more light, less dark, less heat”7.
Yours faithfully,
Ellen Goudsmit (Dr.)
Medical Archivist.
1. Reid S, Chalder T, Cleare A, Hotopf M
and Wessely S. Chronic fatigue syndrome. BMJ, 2000; 320: 292-296. 
2. Hickie I, Wilson A, Hickie C, and Lloyd
A. Letter responding to Chalder et al. American Journal Of Medicine
1995; 98: 421-422.
3. Deal A, Chalder T and Wessely S.
Illness beliefs and treatment outcome in chronic fatigue syndrome.
Journal of Psychosomatic Research 1998; 45: 77-83.
4. Friedberg F. A subgroup analysis of
cognitive-behavioral treatment studies. Journal of Chronic Fatigue
Syndrome 1999; 5: 3-4: 149-159.
5. Vercoulen JHMM, Bazelamns B, Swanink
CMA, Fennis JFM et al. Physical activity in chronic fatigue syndrome:
assessment and it’s role in fatigue. Journal of Psychiatric Research,
1997; 31: 661-673.
6. Friedberg F and Krupp LB. A comparison
of cognitive-behavioral treatment for chronic fatigue syndrome and
primary depression. Clinical Infectious Diseases 1994; 18 (suppl 1):
S105-S1l0.
7. Hawes Clever, L. Editorial comment.
Western Journal of Medicine, 1998; 169: 356.
No conflicts of interest.
