JHMM Vercoulen. Chronic fatigue syndrome. Assessment and psychological processes as perpetuating factors. PhD. University of Nijmegen, Netherlands. 16th June 1997.
This thesis is a collection of papers, seven of which have already been published. One of the unpublished studies measured physical activity levels in 51 patients with CFS (Oxford criteria) and compared the results with those of 50 mobile MS patients and 53 healthy controls. Measures included self-report questionnaires, a 12-day observation list and a motion-sensing device (Actometer).
There were no significant differences between the CFS and MS patients in the levels of activity as measured by the Actometer. However, both recorded less activity than the controls. CFS patients tended to avoid those activities which they thought would produce higher fatigue levels. Low levels of activity were related to severe fatigue in the CFS group but not in the MS group. Finally, there was little correlation between the self-report questionnaires and the Actometer. Vercoulen concludes that cognitive factors influenced activity in CFS and this affected fatigue.
The second unpublished study describes a model which was tested in 51 patients with CFS and 50 people with MS. Attributing symptoms to a physical cause was related to lower activity levels and fatigue in the CFS group but not in the MS group. Depression (measured by the BDI) had to be deleted from the model as a result of the data obtained. However, sense of control over symptoms (self-efficacy) and focusing on bodily symptoms were found to have a direct effect on fatigue. The author concludes that causal attributions, low levels of physical activity and focusing on bodily symptoms may play an important role in the perpetuation of fatigue.
The third unpublished study measured neuropsychological performance in 51 patients with CFS and 53 healthy controls. Seventeen per cent of patients showed impairments in concentration; 13% were impaired in verbal memory, 25% had slower motor speed and 27% had slowed speed of information processing. Overall, only 29% of the patients could be considered as impaired in neurological functioning. There was no relationship between subjective and objective measures. According to Vercoulen, the poor test performance of the subgroup may be related to a prolonged lack of stimulation.
I read this thesis not only as a psychologist with a doctorate on a similar subject but also as an archivist who has been following the literature since 1984. It's largely because of the latter that I now have a totally different view of CFS to that expressed by Dr. Vercoulen.
What it basically comes down to is a difference in the way we define and study CFS. I have long questioned the idea that CFS was one entity or one disease, or even a number of different conditions with one shared malfunctioning pathway somewhere in the brain. Instead, I now believe that CFS covers a number of different disorders, each with their own aetiology. Some may overlap, showing the same or similar changes in the immune system, neurotransmitters or hormones. However, in contrast to Dr. Vercoulen, I sense that the various differences may be important, and that we will ultimately learn more about CFS by separating our patients into the relevant subgroups.
Let me explain what I mean by using headaches as an example. If one wants to study the causes of headaches and one does not distinguish between tension headaches, migraines and pain due to brain disease, the results will almost certainly show that
If one takes a different approach, for example, separating the patients on the basis of associated symptoms (nausea, visual disturbances) or their failure to respond to aspirin, one would obtain exactly the same results, but in addition, one would be able to identify the migraine sufferers and those with brain tumours, and it would be clear why aspirin and relaxation weren't helpful in their case. In other words, lumping everyone with a headache into one group allows you to see the similarities between your patients, but makes it much harder to unearth the differences, be they physical or psychological. The same is true for fatigue.
In Europe, most researchers have used the broad Oxford criteria to select their patients, resulting in a very mixed population. This means that in the same group, one will have patients whose fatigue is due to infection, allergies, primary psychiatric conditions, sleep disorders or lifestyle problems (lack of exercise, stress, boredom). Suffice to say, some groups overlap, e.g. patients with ongoing infection may also report depression and allergies. However, the research also suggests that in many cases, the presence of a psychiatric disorder, sleep problems or long-term deconditioning is often a sufficient explanation for ongoing fatigue.
The situation in Europe is very different to that in America. Here, most researchers have followed the Centers for Disease Controls' (CDC) 1994 recommendation to separate subgroups according to a number of specified variables (Fukuda et al 1994, p. 956). It means that scientists have been able to compare CFS patients with and without psychiatric disorders, as well as those with mild and more severe fatigue. Groups have also been split according to the presence or absence of infection at onset, the type of onset (e.g. acute/gradual), and the presence of certain immunological markers.
Some of these studies have already been published. They show that different subgroups vary in terms of severity of certain symptoms (e.g. memory and concentration, DeLuca et al 1997, Marcel et al 1996), the evidence of ongoing infection (e.g. Lerner et al 1997) and the likelihood of recovery (Levine et al 1997). However, I must stress here that this research is in its infancy and that if someone finds a single abnormality in all CFS patients, we may well return to the view of one entity and one main cause.
When Dr. Vercoulen began his research, we did not know how marked the difference between subgroups were going to be. Nor did we have any idea that the use of the Oxford criteria would reduce our chances of finding evidence of physical disease. Although studies have found an increased sensitivity of serotonin receptors in both broadly and strictly-defined subjects, that's more or less the only universal finding to date. Very few researchers who use the Oxford criteria have reported evidence of immune activation or ongoing viral infection in these populations. What they have reported is comparatively higher rates of psychiatric illness, more personality problems, and more restrictive coping responses than the studies on more strictly defined CFS (e.g. Goudsmit 1996 versus Surawy et al 1995).
This pattern of results also characterises the Nijmegen research. Neither Dr. Vercoulen nor Dr. Swanink found significant abnormalities on tests for infection and immune dysfunction. On the other hand, they did identify a number of psychological factors which could at least partly explain their patients' failure to improve.
The fact that some of his findings conflict with studies on strictly-defined CFS suggests that we must be careful about the conclusions we draw. Consider, for instance, the Prozac trial. The negative results contrast with the only other study published to date. Klimas et al (1993) used the same dose as Vercoulen and obtained quite spectacular results in patients with abnormal immune function. Prozac had no effect on depression though!
However, it's too early to conclude that Prozac is a useful drug for patients with CFS. Since so many studies have already indicated an increased sensitivity to serotonin, and given that Prozac increases levels of this neurotransmitter, common sense suggests that Vercoulen's results should not be dismissed. Perhaps it's a valuable treatment for subgroups, e.g. those with normal levels of serotonin but underactive or overactive immune systems.
Another area where Dr. Vercoulen's findings conflict with those on strictly-defined samples is the performance on neuropsychological tests. As expected, Dr. Vercoulen found few obvious changes in memory, attention, concentration etc.. This contrasts with Marcel et al (1996) who found significant abnormalities in their severely ill patients selected using the very strict (but valid) CDC criteria from 1988. Research from New Jersey (DeLuca et al 1997) backs this up.
Again, I don't think that the discrepancy between Dr. Vercoulen's findings and those of others can be explained entirely in terms of the criteria used. There are other factors which could also have influenced the results. For instance, we know that post-infectious CFS/ME fluctuates and that symptoms tend to come and go (Dowsett and Welsby 1992). What if most of the patients were tested when they felt relatively well? Would the attention and memory defects show up anyway? Given these uncertainties, I don't think that Vercoulen's failure to find significant impairments in two-thirds of his sample means that these people's reports of cognitive symptoms were exaggerated. Going on American research, problems with attention and memory are more prominent in acute-onset patients and those with post-infectious fatigue. So had Vercoulen focused on this group, and done some tests at home (to assess people when they're not well), he might have obtained very different results.
According to Vercoulen, the abnormalities which he did find may have been due to low levels of stimulation. If this were true, would one not expect more global deficiencies rather than the specific defects reported to date?
Attributions
Discussing his findings about the role of attributions, Vercoulen suggests that beliefs about the causes of the illness determines the patient's response. In his opinion, those who believe that their illness is primarily physical will reduce their activity levels and increase or perpetuate their fatigue.
Other researchers, however, have found that patients with CFS appear to have more sophisticated views about causes (e.g. Clements et al 1997, Goudsmit 1996). Moreover, few have been able to link attributions to a particular response. My own research and those of others suggest that patients generally try different things and unpublished findings from London's Barts Hospital (presented at a conference of psychologists) suggests that what patients do when they have CFS is very different to the response to viral infections. So if attributions aren't linked to just one response (i.e. inactivity), can one also conclude, as Vercoulen did, that the same response is the primary cause for ongoing fatigue? I don't think so.
Let's assume for a moment that some of the patients who believe that they have an ongoing physical illness are right. Maybe they have more severe fatigue than those who are mistaken and actually suffer from primary psychiatric disorders (see Natelson et al 1995). Without doing the relevant tests to check for ongoing disease (including specific tests for parasites, liver dysfunction and immune activation), one certainly cannot conclude that a patient's somatic attributions are inaccurate, let alone the reason why they fail to improve.
His findings also led Vercoulen to claim that patients must change their beliefs about what causes CFS because unless they do, they won't try psychological treatments and won't increase activity levels. However, my own study and those of others suggest that such resistance is rare. For example, a belief in a physical illness did not prevent patients trying psychological treatments in the UK (cf. Sharpe et al 1996, Fulcher and White 1997). Moreover, Lawrie et al (1997) found no relationship between physical attributions and ongoing fatigue while Wessely and his colleagues reported that attributions did not influence the recovery of patients on his CBT/graded exercise trial (Bonner et al 1994). In fact, people with chronic fatigue seem to behave like everyone else. There's no uniform response to CFS just like there is none to MS or heart disease. Different people think and do different things. The problem arises when patients choose things which are clearly not going to help, like smoking, going to bed all day when they are not severely ill, not taking antidepressants for clinical depression etc.. That's not only influenced by one's beliefs about causes but also by one's personality (e.g. pessimism), fears and phobias, previous experiences with illness, views about doctors, advice from non-experts etc.. Dr. Sharpe from Oxford University identified a group amongst his CFS patients who were anything but sensible, and found that 70% responded to CBT and exercise (Sharpe et al 1996).
Vercoulen goes on to suggest that "a continuing search for medical explanations is not helpful but has a detrimental effect on complaints". This may be true for those without evidence of disease but as I found out, it is really worth doing additional (orthodox laboratory) tests to identify concurrent infections, liver problems and other additional complications which often prevent recovery (see also Teitelbaum and Bird 1995).
Activity levels
While I understand how and why low levels of activity may increase fatigue, I have doubts about its influence on CFS as a whole. First of all, there's more to most cases of CFS than fatigue and I don't understand how inactivity would cause symptoms like swollen glands, sore throats, intolerance to alcohol, nausea etc.. Secondly, sedentary people as a general rule don't get CFS. (If couch potatoes were prone to CFS, we'd have a major epidemic on our hands). Thirdly, my research and that of others suggest that most patients aren't immobile (hence no muscle wasting), and generally do too much rather than too little (cf. Friedberg and Krupp 1994). In order to assess the relationship between physical activity and CFS more accurately, one also has to check for levels of anxiety (very important), existence of social support and evidence of viral persistence and immune activation. Inactivity may well be a major influence on fatigue in some patients but we need to do more research if we are not to make mistakes.
It follows that I have great reservations about the accuracy of the proposed model explaining the persistence of CFS. Firstly, it must be remembered that models don't prove that one variable leads to another, but merely suggest what might be linked and to what degree. Secondly, which variables are included in the model depends on which ones you studied in the first place. Had Vercoulen also included other relevant measures such as lack of support, severity of cognitive symptoms and immune status to name but three, the apparent link between attributions, activity levels and fatigue might not have been there. In my research, lack of support from friends influenced activity levels, while cognitive symptoms were a much more important predictor of disability than fatigue! Similarly, Anderson and Ferrans (1997) found that 50% of their patients regarded cognitive dysfunction to be their most disabling symptom. Of course, one can't measure everything and no one expects researchers to. However, the use of a limited number of variables does restrict the conclusions one can draw.
Self-efficacy
One of Vercoulen's findings which has been replicated elsewhere (by me) is the apparent link between sense of control over symptoms (self-efficacy) and fatigue. Indeed, in my own research on patients with PVFS/ME, self-efficacy wasn't just related to fatigue but also to depression. Basically, patients with greater confidence about controlling symptoms were less depressed.
I also assessed a treatment programme which improved patients' self-efficacy. Using an approach borrowed from work on people with cancer and MS, the subjects were given information about the illness (which they were told was a physical disease), counselling, and advice to balance rest and activity (pacing). About 80% of these patients improved within 6 months and a quarter recovered by 90% or more within a year.
CBT and graded exercise programmes may also work by increasing self-efficacy. NB. When doctors refer to 'graded' activity, it means that patients have to increase their activity irrespective of how they are feeling on a particular day. Conversely, pacing requires you to reduce levels if activity leads to relapse and to increase it if there are no adverse effects. At the moment, the main question which keeps researchers busy is not so much, should patients exercise but, how much and when.
Conclusions
The view that different researchers have been studying different groups all covered by the term CFS not only explains the variety of theories around but also suggests that most specialists are probably right. So when I or Lerner et al talk about a possible viral cause, we may be right when talking about our patients, but when others focus on lack of activity and certain beliefs causing fatigue, they may also be right in relation to theirs. In my view, it's too early to say what the causes of CFS are, with the exception of depression. Having studied the evidence, I agree with Vercoulen that depression cannot explain most cases of CFS.
The majority of Vercoulen's findings are consistent with earlier research using broadly-defined patient groups. However, I think that he has overstated the importance of attributions and I consider his model to be overly simplistic. While much of it may well apply to the majority of patients with fatigue, I am not yet persuaded that it is also relevant to people with infection-related CFS. Further analysis on the basis of the CDC recommendations might not only have unearthed more data but also greatly increased our understanding of all types of fatigue syndromes.
Ellen Goudsmit (Dr.) ©
References
Due to limited space, we haven't included the complete reference details, but the information below should be sufficient for those wishing to obtain copies.
Anderson and Ferrans. Journal of Nervous and Mental Disease, 1997, 185, 6, 359.
Bonner et al. Journal of Neurology, Neurosurgery and Psychiatry, 1994, 57, 617.
Clements et al. Journal of Psychosomatic Research, 1997, 42, 6, 615.
Costa et al. Quarterly Journal of Medicine, 1995, 88, 767.
DeLuca et al. Journal of Neurology, Neurosurgery and Psychiatry, 1997, 62, 151.
Dowsett and Welsby. Postgraduate Medical Journal, 1992, 68, 63.
Goudsmit, EM. Thesis (PhD). Brunel University 1996.
Friedberg and Krupp. Clinical Infectious Diseases, 1994, 18, suppl. 1, S105.
Fukuda et al. Annals of Internal Medicine, 1994, 121, 953.
Fulcher and White. British Medical Journal, 1997, 314, 1647
Ho-Yen. British Journal of General Practice, 1990, 40, 37.
Ho-Yen and McNamara. British Journal of General Practice, 1991, 41, 324.
Hyde et al. Paper presented at the international conference of CFS, Dublin, Ireland, 1994, May 18-20.
Innes. Lancet 1970, 1, 969.
Klimas et al. Chapter in In Chronic Fatigue and Immune Deficiency Syndromes. Eds: Goodnick and Klimas. APA, 1993, p. 99
Lawrie et al. Psychological Medicine, 1997, 27, 343.
Lerner et al. Infectious Diseases in Clinical Practice, 1997, 6, 110.
Marcel, B et al. Biological Psychiatry, 1996, 40, 535-541.
McGarry et al. Annals of Internal Medicine, 1994, 120, 972.
Natelson et al. Clinical Infectious Diseases, 1995, 21, 1204.
Sharpe et al. British Medical Journal, 1996, 312, 22.
Sharpe et al (Oxford criteria). Journal of the Royal Society of Medicine, 1991, 84, 118.
Surawy et al. Behaviour Research and Therapy, 1995, 33, 535.
Swanink et al. Thesis (MD). 1996. Nijmegen University.
Teitelbaum and Bird. Journal of Musculoskeletal Pain, 1995, 3, 4, 91.
Summary of the main arguments against the CBT explanation of CFS.
Background
According to the CBT model described by Sharpe, Wessely and Vercoulen, physical attributions lead patients to reduce activity, which increases deconditioning and perpetuates the fatigue. It follows therefore, that increasing activity will reduce deconditioning and reduce fatigue. The proponents of this theory also claim that looking for causes is futile and may prevent recovery. Depression further exacerbates fatigue.
Arguments against:
Theoretical questions:
Copyright EM. Goudsmit 1998.
©
Psychologist/Archivist, London.
All rights reserved. This article may not be reproduced without
permission from the author. See the full
copyright
notice.
Be sure to see the many other valuable articles at our Main M.E. Home Page
