The following has been updated since its publication in InterAction.
For many years, CFS has been thought of as a single disorder with an unknown cause. Any differences between patients were either dismissed as irrelevant or simply ignored. Since 1994, however, researchers have begun to examine the validity of this approach by subdividing their patients into groups. If the differences were of no significance, it was argued, then these subgroups of CFS should be indistinguishable from one another on the various measures and tests. So far though, it hasn't turned out that way. Contrary to all the predictions, the studies have revealed marked differences in the severity of certain symptoms, the type of immunological abnormalities, the prognosis and response to treatments. What's more, most of the results make sense! In this article, I've summarised some of this new research, starting with a comparison of patients with and without psychiatric disorders and the considerable differences between acute onset and gradual onset CFS.
Researchers from New Jersey assessed the performance of 36 patients with moderate to severe CFS on a number of cognitive tests (Deluca et al, Journal of Neurology, Neurosurgery and Psychiatry, 1997, 62, 151-155). The group was subdivided according to the presence or absence of current or previous psychiatric disorders and their results were compared with those of 31 healthy, though fairly sedentary controls.
Overall, the performance of the CFS patients was significantly below that of the controls. What's more, the patients with no psychiatric disorders did worse on some measures than the patients with psychiatric disorders. As the authors put it, this shows that "impaired cognition in chronic fatigue syndrome cannot be explained solely by the presence of a psychiatric condition".
While the results are similar to those reported in patients with PVFS/ME (cf. research by Prof. Smith and Marcel et al 1996), they conflict with most of the studies on broadly-defined CFS. This is where patients are generally selected using the 'Oxford' criteria which cover not only PVFS/ME but also patients who are 'tired-all-the-time' or suffering from long-term psychiatric disorders, sleep problems and other conditions known to cause chronic fatigue. Actually, the 'Oxford' criteria do allow researchers to separate out any patients with 'post-infectious CFS' but so far, virtually no British doctor has done so. Indeed, researchers here regard such subgroups as 'highly selective' and therefore 'unrepresentative' of patients with CFS. Still, I've recently spotted some references to fatigue syndromes in the British medical press (e.g. Lawrie et al, Psychological Medicine, 1997, 27, 343), so attitudes may be beginning to change!
Mawle and colleagues from the CDC compared 26 patients with CFS (CDC '88) with 50 age, gender and race-matched controls on several measures of immune function (Journal of Infectious Diseases, 1997, 175, 1, 136-141). Patients were subgrouped according to disease onset (sudden or gradual) plus their health at the time of testing (well or sick) and the duration of illness (long or short).
There were no overall differences in white blood cell numbers and other measures of immune status. Moreover, only marginal differences were detected in cytokine responses and in cell surface markers in the CFS group as a whole. However, when the patients were subdivided according to the type of disease onset or by how well they felt on the day of testing, more pronounced differences were seen.
For instance, the sudden group had significantly more CD8 cells expressing CD11b (seen in acute infections) as well as higher levels of the cytokine IL-2 than the controls. The gradual onset group had lower levels of the cytokines IL-1a and IL-1b than the controls, and a lower percentage of certain CD56 (natural killer) cells. Sicker patients had higher levels of IL-2, lower levels of IL-1a and fewer CD8 cells expressing CD25 compared to the controls.
In other words, taking the sample as a whole revealed few abnormalities. However, more specific analysis showed that patients with a sudden onset of CFS had a cytokine and cell surface markers profile reminiscent of acute infection.
Acute and gradual patients have also been found to differ in other ways. Reyes and colleagues (1996) studied 25 patients with CFS and found that stressful life events in the year prior to illness were much more common in the people with a gradual onset than in those with an acute onset.
Similarly, Christodoulou et al reported that patients with an acute onset had fewer psychiatric disorders than those with a gradual onset. (For further information, see their chapter in Yehuda, S and Mostofsky, DI [eds] Chronic Fatigue Syndrome. New York: Plenum Press 1997 and Deluca et al 1997b).
I suspect that acute onset patients tend to be people who were generally well when they came down with a bug. As for the gradual group, these may be patients whose immune systems or neurotransmitter receptors had been under strain for some time (due to previous infections, antibiotics, vaccinations, stress?) before the 'final straw' (e.g. bug, more stress), led to full-blown CFS. Since their illness is likely to be more extensive (affecting more organs?), one might expect them to recover more slowly and to report greater psychological distress than their acute-onset counterparts.
While I've been impressed with all the studies on subgroups, there was one which really stuck out above the rest. It describes patients with evidence of an ongoing viral infection who were then given a course of the appropriate antiviral drug. What was unusual is that the authors included a detailed comparison between the people who responded to this treatment and those who did not.
The study in question is by Lerner et al (Infectious Diseases in Clinical Practice, 1997, 6, 2, 110-117). It involved 18 patients with CFS reporting overwhelming fatigue for less than 2 years, all of whom had specific abnormalities on electrocardiograms (ECG). The changes in heart function were "not seen in normal persons leading a sedentary life". Further analysis revealed that the patients had raised antibodies to human cytomegalovirus (HCMV), suggestive of a "persistent HCMV infection". All were severely ill.
They were treated with intravenous ganciclovir, an antiviral drug known to be useful against HCMV. They were also instructed to avoid exercise and alcohol. After the treatment was completed, the patients were asked to renew normal activities in gradual increments "as tolerated".
At the follow-up (24 weeks after ganciclovir), 13 patients (72%) had returned to their pre-illness healthy states (p<.05). The patients who did not respond tended to have lower antibody levels to HCMV before treatment began. There were no reports of side effects.
Of course, these findings do not mean that every person with CFS has an abnormal heart function or an ongoing problem with HCMV. Indeed, the results probably apply to no more than a tiny minority of people with chronic fatigue. However, it does illustrate the value of identifying subgroups and suggests that a persistent infection may play a role in some cases of CFS.
Another study in this new batch of research is the report by See and Tilles (Immunological Investigations, 1996, 25, 1-2, 153-164). They assessed 26 patients with CFS subdivided according to immune markers. All patients then completed a trial of alpha 2a interferon and placebo in a double-blind crossover design.
Three patients had normal results when compared to 20 healthy controls; seven had diminished natural killer (NK) cell function while nine had abnormalities on the lymphocyte tests. The rest had abnormalities on both. Four patients dropped out of the trial. Surprisingly, the only people who benefited from the alpha interferon were the patients with decreased NK function.
Different fatigue syndromes may have different prognoses. For instance, Paul Levine and his colleagues studied 23 patients who had become ill during an outbreak resembling epidemic neuromyasthenia (aka as ME) in Otago, New Zealand in 1984. Complete information was available for 21 patients (Archives of Internal Medicine, 1997, 157, 7, 750-754).
Eight of the 21 were retrospectively diagnosed as having CFS, two more were classed as possible cases. The others (52%) were classified as having had prolonged or idiopathic fatigue. (They didn't have the symptoms listed in the CDC criteria 1994).
After two years of illness, 76% reported a return to pre-illness levels of functioning. However, some also noted having to modify their activities in order to prevent relapse.
The favourable outcome of these cases of "epidemic-associated CFS" is consistent with previous research showing that the majority of people with an acute-onset improve significantly between 2 and 5 years. The prognosis for the gradual onset and severest cases is generally not as good.
I cannot emphasise enough the importance of this new development in CFS research. Studies like the one by Lerner et al strongly suggest that we are not dealing with one illness, nor with one cause. I would therefore urge you to be cautious about simple theories claiming, for example, that CFS is due to a shift from one type of immune response to another (Rook and Zumla, see reference list below) or that the illness can be attributed to a defective enzyme in the anti-viral pathway (Suhadolnik et al 1997). I may be wrong, but given the research so far, I suspect that these changes are relevant only to certain subgroups. Of course, if the defective enzyme is found in everyone with CFS, including the ones without an infectious onset or signs of immune activation, then we're back again to the theory of one entity, one main cause.
Assuming that there are several subgroups, could PVFS/ME be one? In the table below, I've summarised some of the similarities and differences between PVFS/ME and CFS. I've combined PVFS and ME into one group because neither I nor others have been able to find anything that distinguishes between the two.
Noteworthy is that the criteria for PVFS/ME focus primarily on when symptoms occur (i.e. in relation to exertion) whereas definitions of CFS do not (see introduction, elsewhere). The other major difference is that ME researchers have observed marked fluctuations in symptoms, which separates this illness from cases of TATT ('tired-all-the-time'), clinical depression and sleep disorders.
With this in mind, it is interesting that when patients with CFS were assessed objectively, there was little reduction in activity following exercise (Sisto et al 1998). In contrast, a study on people with ME/PVFS showed significant abnormalities in muscle function during the recovery phase, consistent with reports of post-exertional fatigue (Wood et al, in press).
In terms of treatment, I was told by one researcher that although most of his CFS patients improved following CBT and graded exercise, many of his ME patients did not. Alas, he forgot to include this information in his report of the trial.
The presence of exertion-related symptoms and the marked fluctuations have recently been documented in some publications on strictly-defined CFS so many of the latter may actually represent cases of PVFS/ME (Anderson and Ferrans 1997, Plioplys and Plioplys 1997). However, if this is true, then research on the different groups should show identical changes in neurotransmitter pathways, the HPA axis and immune function. So far, the evidence is equivocal. For example, a study on CFS found normal to low levels of cortisol (e.g. Demitrack et al 1991) while research on ME-like groups reported levels which were normal to high (e.g. Hilgers and Frank 1992, Richardson 1995). On the other hand, it may just be that the CFS group had been ill for a longer time without support (chronic stress is associated with lower cortisol).
It is also noteworthy that while a number of immune changes have been documented and replicated in strictly defined CFS with post-exertional fatigue (i.e. suggestive of ME), these have not always been reported in people with general CFS (e.g. Bennett et al 1997, versus Natelson et al 1998).
Although there seems to be a considerable overlap between ME/PVFS and strictly-defined CFS, both these conditions appear very different from broadly-defined CFS. For instance, I can recall only one or two articles on the latter which referred to post-exertional malaise or fluctuating symptoms. Moreover, studies using the stricter diagnostic criteria have reported quite different results in terms of immune abnormalities and evidence of viral infection compared with studies using broader criteria (e.g. Lerner et al 1997 versus Vercoulen 1997).
Psychologically, there are also marked 'inconsistencies' between groups. Indeed, if one reads descriptions of some CFS patients (e.g. Surawy et al 1995), and you compare them with those meeting the more restrictive criteria for ME/PVFS (e.g. Goudsmit 1996), there seems to be little overlap. It's hard to believe we're all talking about the same disease. For further examples, compare Surawy et al with Saltztein et al (1998) or Heijmans (1998).
A final argument for the existence of PVFS/ME as a subgroup is the fact that the prevalence rate for this disorder is about 1 per 1000 while the estimate for broadly-defined CFS is 1 to 2 per 100 (cf. Royal Colleges Report 1996). In other words, there are many people who fulfil criteria for one but not the other.
I've already suggested which psychological disorders come under the heading of broadly-defined CFS. They are guesses but based on clinical experience (mine and others) as well as the research (e.g. Hickie et al 1996 and Fischler et al 1997). Of course, the presence of depression, anxiety or sleep disorders doesn't preclude a diagnosis of CFS. Having one illness doesn't automatically grant you immunity from developing another. However, where there is a long history of anxiety or significant sleep disorder, and patients do not report exertion-related exhaustion or malaise, then a diagnosis of CFS may not be correct.
To learn more about CFS, we must continue to subdivide, compare and contrast (cf. Fukuda et al 1994). For instance, research on ME patients (selected according to the London criteria) showed reduced blood flow in the brainstem compared with healthy and depressed controls (Costa et al 1995). Exactly the same findings were later reported in people with CFS using PET (Tirelli et al 1998). From these two studies, we therefore know that the reduced blood flow in the brainstem relates to CFS, not just to ME. We also know that it cannot be attributed to depression, as some have maintained.
Even though we have limited information, I predict that many patients with strictly-defined CFS will turn out to have PVFS/ME. However, I'll be very surprised if the abnormalities found in these cases will be equally common in more broadly-defined groups.
|
ME/PVFS |
CFS |
REFERENCES |
|
Definitions and descriptions note relationship between symtoms and exertion |
Such a relationship is rarely noted |
Dowsett and Welsby 1992, Ho-Yen 1990, Vs Fukuda et al 1994, Sharpe et el 1991 |
|
Evidence of CNS involvement e.g. prominent cognitive impairment important |
No emphasis on CNS involvement though Sharpe et al require effect on 'mental functioning'. |
Dowsett and Welsby 1992, Ho-Yen 1990, Vs Fukuda et al 1994, Sharpe et el 1991 |
|
Marked fluctuation in symptoms noted (ME only) |
Fluctuations noted once |
Dowsett and Welsby 1992 Anderson and Ferrans 1997 |
|
Reports suggest onset more common Spring/Summer |
Spring/Summer influence not documented |
Hyde et al 1994 |
|
Prevalence at most 1 per 1000 |
Prevalence 1 or 2 per 100 |
Ho-Yen and McNamara 1991 Vs Royal Colleges Report 1996 |
|
Autopsy identified viral RNA in brain and muscle tissue |
No published autopsy reports |
McGarry et al 1994 |
|
Rate of psychiatric disorder similar to other chronic illnesses |
Rates of psychiatric disorders in broadly defined CFS higher than in other chronic conditions |
Shanks and Ho-Yen 1995 Yeomans and Conway 1991 Vs Royal Colleges Report |
|
Limited evidence suggests cortisol levels normal to high |
Evidence cortisol levels normal to low |
Richardson 1995 Vs Demitrack et al 1991 |
|
Hypoperfusion in brainstem |
Same (in strictly-defined CFS) |
Costa et al 1995 |
|
Sensitivity of serotonin receptors |
Sensitivity of serotonin receptors |
|
|
Acute onset cases have better prognosis |
Acute onset cases have better prognosis |
|
|
Some evidence of immune activation |
Same in strictly-defined CFS |
|
|
Some evidence of abnormalities in muscle function |
Same in strictly-defined CFS |
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.
One of the researchers who helped to draw up the Oxford criteria, Dr. Michael Sharpe, recently acknowledged that there is no evidence that the most commonly used "case definition for CFS selects a homogenous patient group" (Mountstephen and Sharpe, Occupational Health 1997, 47, 217-227). They recommend the CDC criteria published in 1994, which are broader than the 1988 version but more specific than the 'Oxford' guidelines.
(From elsewhere on this web-site)
In the United Kingdom, CFS covers a number of conditions, including myalgic encephalomyelitis (ME) also known as post viral fatigue syndrome (PVFS). The latter is characterised by fatiguability following minimal exertion, marked fluctuations in the severity of symptoms throughout the day, impaired circulation and the involvement of the CNS. Most patients who fulfil the British criteria for CFS appear to have disorders other than ME including subclinical hypothyroidism, masked depression, and problems related to lifestyle and nutrition (e.g. deficiencies of vitamins D and B12). In America, stricter criteria select a more homogenous population, so most cases of CFS there are probably the equivalent of ME/PVFS.
The prevalence of ME and strictly-defined CFS is about 1 per 1000; the prevalence of CFS selected using the British criteria may be as high as 2%.
Lane, RJM., Barrett, MC., Woodrow, D., Moss, J., Fletcher, R and Archard, LC. Muscle fibre characteristics and lactate responses to exercise in chronic fatigue syndrome. Journal of Neurology, Neurosurgery and Psychiatry, 1998, 64, 3, 362-367.
DeLuca, J., Johnson, SK., Ellis, SP and Natelson, BH. Sudden versus gradual onset of chronic fatigue syndrome differentiates individuals on cognitive and psychiatric measures. Journal of Psychiatric Research, 1997, 31, 1, 83-90.
DeLuca, J., Johnson, SK., Ellis, SP and Natelson, BH. Cognitive functioning in patients with chronic fatigue syndrome devoid of psychiatric disease. Journal of Neurology, Neurosurgery, and Psychiatry, 1997, 62, 151-155.
Lerner, AM., Zervos, M., Dworkin, HJ., Chang, CH., Fitzgerald, T., Goldstein, J et al. New cardiomyopathy: pilot study of intravenous ganciclovir in a subset of the chronic fatigue syndrome. Infectious Diseases in Clinical Practice, 1997, 6, 110-117.
Martin, WJ and Anderson, D. Stealth virus epidemic in the Mohave Valley. Pathobiology, 1997, 65, 51-56.
Dowsett, EG and Welsby, PD. Conversation piece. Postgraduate Medical Journal, 1992, 68, 63-65.
Dowsett, EG et al. Myalgic encephalomyelitis - a persistent enteroviral infection? Postgraduate Medical Journal, 1990, 66, 526-530.
Paul, L et al. Demonstration of delayed recovery from fatiguing exercise in chronic fatigue syndrome. European Journal of Neurology, 1999, 6, 63-69.
Scholey, A et al. A comparison of the cognitive deficits seen in myalgic encephalomyelitis to Alzheimer's Disease. Proceedings of the British Psychological Society, 1999, January, 12.
Innes, SBG. Encephalomyelitis resembling benign myalgic encephalomyelitis. Lancet, 1970, 1: 969-971.
McGarry, F et al. Enterovirus in the chronic fatigue syndrome. Annals of Internal Medicine, 1994, 120, 11, 972-973.
Bennett, AL et al. Elevation of bioactive transforming growth factor-b in serum from patients with chronic fatigue syndrome. Journal of Clinical Immunology, 1997, 17, 2, 160-166.
Costa, DC et al. Brainstem perfusion is impaired in patients with chronic fatigue syndrome. Quarterly Journal of Medicine, 1995, 88, 767-773. (This has been replicated.)
Richardson, J. Disturbance of hypothalamic function and evidence for persistent enteroviral infection in patients with chronic fatigue syndrome. Journal of Chronic Fatigue Syndrome, 1995, 1, 2, 59-66.
Lindal E et al. Anxiety disorders: a result of long-term chronic fatigue - the psychiatric characteristics of the sufferers of Iceland disease. Acta Neurologica Scandinavica, 1997, 96, 3, 158-162.
Yeomans, JDI and Conway, SP. Biopsychosocial aspects of chronic fatigue syndrome (myalgic encephalomyelitis). Journal of Infection, 1991, 23, 263-269.
NB: The vast majority of the findings above have not been documented in broadly-defined CFS (Oxford and CDC criteria '94). It should also be noted that there is no evidence of maladaptive beliefs or the phobic avoidance of activity among most patients with ME. Finally, there has been no study assessing the effectiveness of CBT and graded exercise in uncomplicated cases of ME/post-viral syndrome.
* Including research on strictly-defined CFS which selected patients with post-exertional fatigue, an important feature of ME.
Ellen Goudsmit PhD C.Psychol.
January 1999
