Executive Summary

The cause of chronic fatigue syndrome remains unknown at this point in time. A number of research teams throughout the world have investigated possible links to a number of virus infections (including Epstein-Barr virus, enteroviruses and poliomyelitis as well as and fungal agents (in particular, Candida albicans). Despite extensive studies, the evidence that CFS is caused by any particular infective agent is equivocal. Nor has it been shown that physical or mental stress cause chronic fatigue syndrome.

Chronic fatigue syndrome occurs more commonly in women with a peak incidence in those in the third and fourth decades. Despite media references to the condition being more common in those in middle income and groups, evaluation of social status of cases indicate that CFS occurs with about equal frequency in all social classes. In general, no forms of treatment are shown to alter the course of the condition. Recovery is reported in a number of studies to be facilitated by rest complemented by a supervised program of gentle exercise. In general, no forms of medication have been shown to alter the course of the disease although treatment of specific symptoms is recommended when present. The prognosis varies greatly across the spectrum of cases reported to date, with the majority of sufferers eventually recovering after a period of one to ten years.

Chronic Fatigue Syndrome

Introduction

Terminology

Wallace (1991:943) divides the various synonyms into two groups: epidemic and endemic. Synonyms in the former group include: epidemic neuromyaesthenia, Adelaide epidemic, Royal Free disease, Iceland disease and Lake Tahoe disease, whilst synonyms in the latter group include: myalgic encephalomyelitis, fibrositis myalgia, "Yuppie" flu, idiopathic chronic fatigue syndrome, Epstein Barr disease and chronic infectious mononucleosis. The eponyms in the former group derive from areas of the world where epidemics have been documented. These epidemics give some support to the hypothesis that, in some circumstances at least, the condition is transmissible. The synonyms for the endemic form of CFS appear to be based largely on the unproven assumption that infective agents have a role to play in the causation of the condition. Despite this expectation, plausible studies have yet to corroborate such a view.

Definitions The range of synonyms reflect the diversity of opinion that exists within the medical community as to whether or not chronic fatigue syndrome is a disease and, if so, what the disease complex entails. Some writers point to the general scepticism of many practitioners in the past as to the validity of the diagnosis. They hold that many practitioners consider the diagnosis is a misnomer in that patients with features of CFS are instead suffering from some form of psychiatric illness (with chronic depression being the most common) or are suffering a delayed convalescence to an infective process (such as infectious mononucleosis).

More recent research generally fails to support either of these viewpoints. Firstly, several working groups have recently recommended that co-existing psychiatric illness should not exclude consideration of the diagnosis of chronic fatigue syndrome when a number of the other minor criteria are present in addition to chronic fatigue (infra vide). In addition, some psychometric studies indicate that the response times of those with CFS are often significantly reduced whilst response times in those with depression and other psychiatric ailments are not significantly altered. Moreover, Kirmayer et al. (1989:940-948) hold that the profound muscle ache after exercise seen in CFS is not a feature of depression.

The situation is further compounded by inconsistencies in working definitions. A major source of confusion in this regard arises from the original definition promulgated by the Centres for Disease Control, Atlanta. Holmes et al (1988:387-388) of the CDC confirm that their original definition was intentionally restrictive to maximise the chances that research would identify associations if they exist. In more recent times, other working groups have recommended that the pre-existence or co-existence of psychiatric disease of itself is no longer sufficient reason to exclude considering a diagnosis of CFS.

The criteria of Holmes et al. (1988:387-9) are:-

Table 1. Diagnostic criteria for chronic fatigue syndrome For diagnosis, both major criteria must be present, plus the following minor criteria: (1) at least 6 of 11 symptoms and at least 2 of 3 physical signs or (2) at least 8 of 11 symptoms.

Major criteria
1. New-onset fatigue lasting longer than 6 months with a 50% reduction in activity.
2. No other medical or psychiatric conditions that could cause symptoms.

Minor criteria
Symptoms (must begin at or after the onset of fatigue)
	1. Low-grade fever (ie. 37.5C to 38.6C)
	2. Sore throat
	3. Painful cervical or axillary lymphadenopathy
	4. Generalised muscle weakness
	5. Myalgias (muscle pains)
	6. Fatigue lasting 24 hours or more after moderate exercise
	7.Headaches
	8. Migratory arthralgia
	9. Sleep disturbance (hypersomnia or insomnia)
	10. Neuropsychological complaints (one or more of the following: photophobia, visual scotomas, forgetfulness, irritability, confusion, difficulty concentrating, depression).
	11. Acute onset (over a few hours to a few days)
Physical signs (documented by a medical practitioner twice at least 1 month apart)
	1. Low-grade fever
	2. Pharyngitis (non-exudative)
	3. Cervical or axillary lymphadenopathy

 

The British perspective as expressed at the Green College Working Group meeting in 1990 is slightly different in that the general term recommended for all cases of chronic fatigue is "post viral fatigue syndrome".

Evaluation of postulated aetiological factors

• (A) the historical evidence of an epidemic nature of the disease;
• (B) the traditional emphasis given to infective illnesses as the prodrome for the syndrome;
• (C) the relatively high incidence of pre-existing and co-existing psychiatric illness of various types among patients with chronic fatigue; and
• (D) misconceptions concerning the role of stress in the development of CFS.

California: The first recorded epidemic occurred in doctors and nurses in several hospitals in Los Angeles in 1934. Reported symptoms included muscle weakness, involuntary muscle contractions and twitching, clonic movements and cramps in affected muscles and muscle inco-ordination. Pain was extreme in the back and extremities and lasted for a number of months. Sensory changes were also a common feature, including hyperaesthesia, paraesthesia and areas of anaesthesia - sometimes following the distribution of a nerve trunk and sometimes a whole extremity. Vasomotor and trophic changes , excessive sweating or abnormal dryness of skin, coldness, cyanosis and brittle nails were also commonly reported. Inflammation of the joints occurred in one-third of adult cases, with a third of these being transitory and the remaining two-thirds developing permanent arthritic changes in the joints involved.

Iceland: In early winter in 1948 to 1949, an extensive epidemic broke out in the town of Akureyri in Iceland. Women were affected more than men, although the sex ratio was equal less than 20 years of age. The illness spread rapidly but was confined to the town. It did not appear to be spread by water, milk, food or sewerage but instead by close contact with those affected. In 1955, 39 patients of the 1948 epidemic were re-examined. Of those more severely affected, only 25% had completely recovered, 52% had residual muscle tenderness and 65% had objective neurological signs. Of those only mildly affected in 1948, only 44% had recovered fully with 50% having muscle tenderness and 19% having residual objective neurological signs.

Adelaide: A few months after the 1948 Akureyri epidemic, an epidemic of poliomyelitis occurred in Adelaide. By 1951, 800 patients had been admitted to hospital with a condition called "epidemic neuromyasthenia" with symptoms indistinguishable from poliomyelitis but without microbiological evidence of poliomyelitis. Recovery was delayed for many patients for up to 2 years.

New York State: Similar to Adelaide, a widespread epidemic of poliomyelitis in New York State in 1950 occurred with a number of patients having features distinct from those of polio in the absence of any microbiological evidence of poliomyelitis. Muscle aching and weakness were predominant symptoms in this subgroup who lacked the classical muscle wasting characteristic of poliomyelitis.

Cumbria: A GP in Cumbria reported an outbreak of a disease similar to that described in Iceland in 1948 wherein 233 cases were recorded in a practice population of 1675 people, an incidence of 14%. The illness was thought to be spread by contact. Recovery for many patients was delayed for many months and, in some cases, several years. The features were similar to those described by Holmes et al. (1988) listed above. No virological evidence of either Coxsackie or echo virus infection was found although the clinical picture was suggestive of either of these infections.

Great Ormond Street: A further outbreak of 'epidemic neuromyasthenia' occurred at this hospital for children in 1970 and 1971, with 145 of the 1900 total staff being affected (mostly nurses). Interestingly, a number of children had been referred to the hospital in the preceding years with unexplained symptoms similar to those that the affected staff members developed and similar to those described by Holmes et al.

  Comment: Jenkins (1991) raises the question of whether these epidemics represent an epidemic form of post viral fatigue syndrome transmissible by an unidentified agent but does not give any firm answers. He argues that the recognition of such epidemics is more likely to occur in an institutional setting, especially health institutions, for obvious reasons and suggests that the recognition of low-grade epidemics in the community would be more difficult when only two or three cases a year occur in each general practice. The failure to recognise epidemics in the community in the past is also held to be a product of the lack of recognition given to the condition disease surveillance agencies (such as the Centres for Disease Control in the USA and the Surveillance Centre for Communicable Diseases in the UK) until recently. Jenkins also notes that no institutional epidemics of chronic fatigue have been reported since 1970.

Recommendation: These case reports suggest a condition which is transmissible in the community resulting in epidemics of illness with symptoms indistinguishable from those described for CFS. Given the uncertainties expressed by Jenkins as to exactly what that agent was, it is recommended that the epidemic form of disease be treated as a clinical entity distinct from the endemic form of the disease.   B. Traditional models of infections as a cause of endemic chronic fatigue syndrome There has been extensive research undertaken to determine whether or not one or more biological agents cause CFS of either epidemic or endemic type. To date, no agent has been shown to satisfy the required criteria proposed by Koch so as to be held to have a causal role in this regard. For the reasons given below, the majority view of the research community is that such associations are at best to be treated as having temporal significance for some patients without it being possible to infer an aetiological link. The evidence on which this assertion is based in presented below in summarising the findings to date for the more common agents under consideration.

The search for a biological agent appears to be a product of the frequency with which CFS patients report symptoms interpreted as being consistent with an acute "viral" illness preceding the onset of chronic fatigue syndrome. Lloyd et al. (1990:527) for example report that 75% of the patients they interviewed gave such a history, with the majority having a healthy premorbid status. Manu et al. (1993) found that 70% of CFS patients (versus 30% of controls) attribute their illness to a viral cause. Despite this perception, studies which have looked at the frequency of infection with a range of viruses have, in general, failed to substantiate that a particular virus is more common in those with CFS (see Shafran 1991:730-9 and Kroenke 1991:44-55) compared to the natural rate of asymptomatic infection in the general community. Winters and Quinet (1992: 260-270) state that the evidence supports persistent viral infection in only a small percentage of CFS patients.

  Epstein Barr Virus (EBV): This agent (a type of herpes virus) is generally accepted as causing infectious mononucleosis (or glandular fever); a condition seen most commonly in adolescents and younger adults. Kroenke (1991:49) states that "EBV was incriminated as a cause of chronic fatigue syndrome in the mid-1980s but that a critical appraisal of the evidence casts strong doubt on this hypothesis". One of the difficulties in attributing a role of EBV as a cause of CFS is that EBV is ubiquitous in the community, with 95% of healthy persons over the age of 30 have serologic evidence of past infection with EBV. Moreover, whilst it is well known that a number of acute cases of infectious mononucleosis have a protracted recovery (termed "chronic mononucleosis") with many having high levels of IgG-VCA (1:320 or higher) and elevated anti-ECA (1:40 or higher), this profile is also seen in 50% of the healthy community. Furthermore, antibody titres to cytomegalovirus, measles virus and various herpesviruses are also often elevated in patients with CFS.

These elevations are now generally viewed as an epiphenomenon of the mild immunologic abnormalities found in some patients with CFS (see Holmes 1991:S53-55). In other words, the presence of antibodies to a range of viruses in some patients with CFS may equally likely have occurred after the onset of CFS and are due to decreased resistance of the body's immune system to a range of viruses normally present in the community.

  Enteroviruses: Approximately 70 serotypes are described for this group of viruses which are members of the family Picornaviridae. Three types have been suggested at various times as having played a role in the onset of PVFS, namely: polio, coxsackie and echo. A possible role for polio virus is suggested by the experiences of the Akureyri epidemic in Iceland in 1948 (described above). Interestingly, not only was no isolate of polio made in the Akureyri group but also no outbreak of poliomyelitis occurred in the Akureyri area in 1956 (at a time when it was rampant in the remainder of Iceland). In addition, Akureyri children were found not to produce antibodies to polio 1 in 1956 but developed unusually high titres when given its vaccine, suggesting previous exposure to polio virus 1.

Kennedy (1991:809-814) alludes to a number of studies that suggest a role for the Coxsackie viruses in the causation of post viral fatigue syndrome (PVFS) and gives his support to the possibility that this virus is responsible for a number of cases of PVFS. In saying this, however, Kennedy accepts that the immunological evidence of preceding Coxsackie infection in these cases is inconsistent.

He also states that high levels of antibodies to Coxsackie B virus are widespread throughout the general community in the absence of PVFS, making the data difficult to interpret. Furthermore, Kennedy states that no plausible biological mechanism has been formulated to explain how Coxsackie viruses cause PVFS to date. The conclusion drawn by this author is that Kennedy's study provides some indirect support for Coxsackie virus infection as a possible precipitant of some cases of PVFS but does not provide support for a role for Coxsackie viruses as a causal agent for CFS.

Gow and Behan (1991:874) report an extensive study of Coxsackie antibody titres in Scotland in the 1980s which failed to confirm initial suspicions of a role for Coxsackie based on early findings of raised IgM titres for Coxsackie in a number of cases. The larger well-controlled study failed to show any significant difference in antibody titres between cases and controls. Gow and Behan's own findings in relation to Coxsackie revealed that 20% of PVFS cases had enteroviral genome for Coxsackie virus in their muscles. A similar incidence is reported by Cunningham et al (1991:858). These finding, however, are entirely consistent with the hypothesis discussed above that CFS affects the immune system and allows infection with one or more of a range of viruses ubiquitous in the community to occur, since the rate of sero-conversion in the general community approximates these levels.

Other viruses : Elevated titres of various types are reported to be more frequent in CFS patients than in the general population with the authors of the reports suggesting that these agents play a role in the causation of CFS or its variant PVFS. For example, Yamanishi (1992:2612-6) reports elevated titres of human T-lymphotropic virus (HTLV) and HTLV type II. He cites a number of studies that report an increased incidence of human herpesvirus type 6 in CFS patients as well. Lloyd et al.'s Australian study (1990:526) describes three cases where sero-conversion to Ross River virus occurred, as well as sero-conversion to EBV in two cases, mumps virus in one patient and Coxsackie B virus in another. Dowsett et al. (1990:528) report sero-conversion to a range of viruses [apart from EBV (33%) and enteroviruses (31%)] such as hepatitis A (2 cases), respiratory syncytial virus (2 cases), parvovirus (2 cases), influenza B (1 case), varicella (1 case) and rubella (1 case) within the 420 patients with PVFS in their study group.

Conclusions The existence of a wealth of studies reporting sero-conversion to a large number of agents should not be taken to mean that any of these agents have a causal role to play in the development of CFS (or even PIFS for that matter). All of the studies identified to date rely on retrospective serologic evaluation of CFS patients and extrapolate the findings to suggest that the presence of serological marker for a particular virus should be taken to mean that the agent has caused the CFS. There are obvious flaws in this type of logic that greatly weaken the conclusions drawn. Firstly, no prospective studies appear to have been conducted to date to show that infection with a particular type of virus is followed by the development of CFS. Secondly, many retrospective studies fail to recognise the possibility that CFS is the primary condition which in turn predisposes to secondary infection with any one of a range of viruses. Thirdly, most studies fail to publish regional statistics to indicate the frequency with which sero-conversion to the type of virus under consideration occurs in the healthy general population.

C. Psychiatric diseases and chronic fatigue syndrome The question of what is the exact nature of the relationship between psychiatric disease and chronic fatigue syndrome remains one of the most controversial aetiological issues concerning CFS. Despite this controversy, it is worthwhile comparing and contrasting the views of a number of authorities on different aspects of this issue.

The first aspect is whether the pre-existence or co-existence of a psychiatric illness should preclude the diagnosis. As discussed above, the support given by the CDC to excluding those with any form of psychiatric illness originally was intentionally artificial in order to establish guidelines for initial investigation. In more recent times, a number of other authorities have adopted the view that pre-existing or co-existing psychiatric disease should not exclude a person from inclusion in the CFS diagnostic group if the other criteria are met (see Holmes et al. 1988).

The second aspect is whether any form of psychiatric disease has a causal role in the development of CFS. A number of earlier papers from the United Kingdom (Wessely and Powell 1989:940-948), Canada (Taerk et al. 1987:49-56) and the United States (Kruesi et al. 1989:53-56) suggested that patients were likely to have psychological disorders before the onset of CFS, raising the possibility that CFS may occur in "psychologically vulnerable" individuals. A more recent study by Hickie et al. (1990:534-540) found that the premorbid incidence of major depression was 12.5% in those with CFS; a rate similar to estimates of major depression in the general community. Hickie et al. also found that the premorbid incidence of psychiatric disease in his control group with major depression was in the order of 62%. One other point made by Hickie et al. was that 50% of the CFS group subsequently developed a psychiatric illness, most often depression, during the course of their CFS.

The findings of Woods and Goldberg (1991:908-918) are similar to those of Hickie's group in relation to the sub-group considered to have post-viral fatigue syndrome. The classification posited by Holmes et al (1988) also supports the notion of psychiatric disease being most likely an outcome rather than a precedent for CFS. Sharpe (1991:989-1005) adds to this by recommending that in deciding whether or not a person has chronic fatigue syndrome, only those with major psychiatric illnesses such as schizophrenia, other major psychoses and manic depression should be excluded from consideration in the CFS group.

The third aspect concerns consistent differences reported in those with CFS versus those with primary psychiatric disease in relation to cognitive, psychological and psychomotor functioning. Whilst Jamal and Miller (1991:815-825) were unable to detect any consistent abnormalities for a range of neurophysiological parameters in those with CFS compared to normal controls, Behan and Bakheit (1991:793-808) describe pronounced difficulties in dealing with mental tasks requiring intense concentration and an inability to initiate and carry out any complex sequences of thought compared to those with depression. They also point out that those with CFS are not troubled by the inability to experience pleasure, including sexual libido and ability, that mark those with primary psychiatric illnesses with a component of depression.

Other features of CFS include marked sleep disturbance, irritability, lack of guilt feelings, definite forgetfulness despite normal memory function on formal testing and, in some cases, hypersensitivity to light and noise and frightening hypnogognic nightmares. One other common symptom related to sleep is pronounced nocturnal sweating.

Lloyd (1990:530-533) makes reference to a number of studies that have shown that CFS patients have unique patterns of psychological impairment not seen in patients with depression or in normal controls. For example, studies by Sandman at the University of California found that CFS patients demonstrated a unique pattern of impairment of memory recall tests, made worse by brief disruption of the test by showing a video clip. Likewise, Prasher et al.'s (1990:247-253) neurophysiological study of CFS sufferers showed prolonged latency in cognitive event-related potentials not present in patients with depression. A number of other authors also make reference to notable reductions in a range of cognitive functions in CFS patients.

The postulate that CFS leads to a reduction in psychomotor and other cognitive abilities has not been universally endorsed however. David (1991:966-988) questions the validity of the conclusions drawn by Prasher et al. and opines that similar cognitive deficits have been documented in primary depression, schizophrenia and borderline personality disorders. He cites the findings of Millon et al. (1989:131-141) and Atlay et al. 1990:141-149) that the PVFS patients they studied did not perform suboptimally on psychometric testing. David considers that further studies might serve to clarify this issue better. In light of this, it would appear to be premature at this point to recommend the use of psychometric testing as an adjunct to the diagnosis of chronic fatigue syndrome.

D. Misconceptions of the role of physical and mental stress in the development of CFS Whilst there is a common perception in the lay community that the title of Chronic Fatigue Syndrome infers a role for either mental or physical stress or both in the development of the condition, there is little evidence in the literature to support such a view. Whilst a number of authors have investigated the role of stress in relation to a number of other conditions characterised by chronic fatigue, none of these conditions appear to satisfy the criteria for CFS promulgated by the major authorities cited in this work. Several authors (Makowska et al. 1992:323-333; David et al. 1990:1199-1202) in reporting a possible role for stress make mention of self-reported domestic and family stressors as the primary source rather than work-related stress. Stricklin, Sewell and Austad (1990:31-34) in a relatively small study of 25 women identify a statistically significant difference (p less than 0.001) in the self-reported incidence of stress in those with "epidemic neuromyasthenia" (a synonym for CFS) compared to an equal number of healthy women. The study group, however, were also found to have statistically significant differences in their psychological profiles compared to the healthy group.

Cathebras et al (1993:168) suggest that people with pre-existing psychiatric conditions are more likely to report increased levels of stress. Melamed et al. (1993:469-474) define a distinctly different condition they term the "burnout syndrome" as a consequence of increased physical or mental stress or both. Ng (1992:294-295) and Waylonis (1992:343-348) note an increased level of self-reported stress in patients with fibromyalgia., a condition they hold to be distinct from other conditions causing chronic fatigue. None of the major authorities cited in this work make mention of mental or physical stress or both as having a role to play in the onset of chronic fatigue syndrome.

Epidemiology

The popular perception of the media that the condition is more prevalent in the middle and upper social classes is not substantiated by objective analysis. Lloyd et al. (1990:522) found an equal distribution across all social classes. Wallace (1991:947) likewise has found the prevalence of CFS to be about equally distributed across the social classes, with a slight over-representation in Classes 4 and 5 (ie. lower socio-economic groups). He was unable to identify any particular sub-groups more likely to develop the disease. These findings serve to dismiss the media perjorative of "Yuppie flu" as a misnomer. As well, the findings fail to show that endemic CFS is more common in particular occupations, unlike the situation for epidemic CFS (supra vide).

There is a wide variation in the incidence of the disease with near universal acceptance that the prevalence of the condition is probably higher than reported. This is particularly so for epidemiological studies conducted using the original CDC criteria which exclude inclusion of suspected cases with psychiatric disease. Hence, Price's figure of prevalence of less than 1 in 10,000 people as reported by Kroenke (1991:44) is a gross under estimate of the true prevalence.

Lloyd et al (1991:522) suggest that the prevalence of 3.71 in 10,000 (95% confidence interval: 26.8-50.2) obtained from their Australian study is conservative. Ho-Yen and McNamara (1991:324-6) suggest a rate of 13 in 10,000 in a UK study. To date, there have been no major studies reported to suggest an increased or decreased prevalence according to region or climate.

Investigations

Kroenke (1991:50) recommends the following investigations be performed routinely: complete blood count, determination of erythrocyte sedimentation rate, biochemistry profile (including electrolytes, serum creatinine and serum transaminases) and thyroid function tests.

He suggests that more elaborate investigations such as sleep studies, radiological imaging and endoscopy should be limited to only those with the appropriate clinical symptoms. Whilst Kroenke does not recommend serologic testing for Epstein Barr virus, cytomegalovirus or other viruses as a routine, it is to be expected that many patients will already have such tests done by their primary practitioner where an infectious cause is suspected (e.g. those who fulfil Wallace's criteria of "post-infectious fatigue syndrome"). Serological investigations need not be exhaustive. When patients present early in the stage of the disease (within the first two months), it may prove useful to repeat the serologic tests after six weeks to allow for more accurate interpretation of the results.

As well, there would appear to be little justification for performing highly technical imaging tests such as computerised axial tomography (CAT scanning), magnetic resonance imaging (MRI), positron emission tomography (PET), single-photon emission computed tomography (SPECT) or BEAM (a brain scan that incorporates measurements of electrical brain activity). Nor does Kroenke recommend the use of sophisticated immunologic testing or detailed neurophysiologic testing as neither have been shown to produce useful results in the ordinary clinical setting. The same can be said for psychometric assessment in light of what has been said above.

Differential diagnosis

In some cases, the differential diagnosis needs to be expanded to take account of unusual presenting symptoms in addition to profound fatigue. This is especially so when a patient reports symptoms such as (1) balance disturbances, (2) claudication, (3) gastrointestinal symptoms or (4) fluid retention. The differential diagnosis in the first category should include recurrent, acute and chronic labyrinthitis. In the second category, ischaemia of the cauda equina and occult spinal multiple sclerosis bear consideration.

In the third group, occult gastro-intestinal malignancies need to be excluded. In the fourth category, fluid retention and fatigue are prominent symptoms of the fluid retention syndrome whilst fluid retention is also reported in PVFS (Behan and Bakheit (1991:801). Finally, exclusion of a range of psychiatric ailments is indicated when a patient presents with one or more psychiatric symptoms.

Treatment

Kroenke (1991:53) gives some useful guidelines in this regard. He recommends that depression be treated selective by the use of anti-depressants which are non-sedating such as desipramine hydrochloride or fluotexine hydrochloride (Prozac). In severe cases, referral to a psychiatrist in the early stages is recommended. When myalgias are a major problem, non-steroidal anti-inflammatory drugs may be prescribed in the absence of contraindications.

Rest combined with a program of gentle exercise is also held to be appropriate in the majority of cases. This may necessitate a prolonged absence from work when the work duties require moderate or strenuous exertion. For those with less demanding positions, exclusion from the work place is not mandatory although long-term placement on lighter duties (for periods of six to twelve months with regular review) should be seriously considered, depending on the requirements of the work environment.

Any decision to maintain a patient in the work environment represents a balance between the severity and likely chronicity of the condition and the likely socio-economic effects of being displaced from the work force for a protracted period of time. Both Kroenke et al. (1988:929-34) and Straus et al. (1988:855-862) report lower recovery rates and a greater likelihood of clinical relapse can be anticipated in those with chronic fatigue for several years or longer.

A number of therapies have been trialed and shown to have little, if any benefit. Several other treatments are currently under investigation as to their benefits. In the former category should be included antibiotics of any type (including long-term antibiotic therapy), transfer factor (a low molecular weight protein extracted from donor leucocytes), interferons, interleukin-2, acyclovir, immunovir (inosine pranobex), immunosuppressives (such as cyclophosphamide, azathioprine and corticosteroids), vitamins, most 'special diets', treatments aimed to eliminate candida and other fungal infections (such as nystatin, ketaconazole and amphotericin B), herbal remedies, homeopathy, acupuncture, aromatherapy, reflexology, colonic irrigations and a number of other pseudo-medical therapies (McBride and McCluskey 1991:904).

In the second category, the evidence that they alter the course of the disease is either minimal or has yet to be substantiated in properly designed and controlled studies. Therapies that fall in to this category include supervised diets aimed to eliminate salicylates, amines and glutamate (Loblay and Swain 1986:169-177), ampligen (mis-matched double-stranded RNA), intravenous therapy with pooled immunoglobulin, calcium blocking agents to counter exercise-induced fatigue and amantadine (McBride and McCluskey 1991:897-880). In the absence of solid evidence that any of these therapies significantly alter the course of the disease, none are recommended for the treatment of CFS at this point in time.

Course and prognosis

The course of CFS is highly variable, depending on the severity of the condition and the premorbid state of the patient. Convalescence is widely accepted as being aided by ensuring adequate rest in conjunction with a supervised regimen of gentle exercises. Lloyd et al. (1990:522) found that 43% of CFS patients in their study group were unable to attend to work, housework or school during the course of their illness. Aggravating factors of established CFS as reported by Dowsett et al (1990:527) include physical and mental stress (100%), intercurrent infection (42%), climactic changes or hot baths (12%), surgery, immunisation and hormonal disturbances (9%) and psychoactive, anti-arthritic or steroid drugs (5%).

Gold et al. (1990:48-53), and Kroenke and Mangeldorff (1989:262-266) found that 40 to 50% of patients show improvement within 12 months but do not indicate if this meant complete recovery. Kroenke et al. (1988:929-934) and Straus et al. (1988:1692-1698) indicate that lower recovery rates and a higher rate of clinical relapse can be anticipated for those who have fatigue for several years or longer.

The Green College working group indicate a similar prognosis. Behan and Bakheit (1991:801) agree with Kroenke that whilst it is difficult to give general guidelines for all PVFS patients, their experience has been that when the illness has been present for a year or more than the prospects of a full recovery are generally poor. The numbers who enter this state are reported to be but a small group of all PVFS patients. A larger number are reported as undergoing complete remission whilst the majority of patients are reported as being able to lead a life wherein they are able to modify their lifestyle and work environment so as to avoid stresses likely to make their illness worse. Interestingly, they report a number of female patients who entered a remission during pregnancy with relapse during the puerperium.

Similar figures are given by Dowsett et al (1990:527) in reporting the outcomes for a group of 420 CFS patients. They found that the duration of the illness was less than 12 months in 9% of cases, 1 to 2 years in 32% of cases, 3 to 10 years in 47% of cases, 11 to 20 years in 8% of cases and 21 to 60 years in 4% of cases. The illness was reported as improving in 31%, fluctuating in 20%, a steady level of disability in 25% and no remission or worse in 24%. No indication is given as to whether or not any of the 420 patients were beneficiaries of disability income support.

Conclusions

There appear to be two main forms of chronic fatigue syndrome, epidemic and endemic. The history of the former group has been reviewed at length in this paper. For a number of reasons, it is not clear whether epidemic CFS is the same disease as endemic CFS; for the purpose of this work the two groups are held to be distinctly different. Accordingly, this author is of the view that the conclusions drawn here are relevant to endemic CFS but may not hold true for the epidemic form of the disease. A key distinction between the two groups is that, whilst there is a body of evidence that epidemic CFS is caused by a transmissible agent (as evidenced by the numbers of hospital staff affected in a number of epidemics), there is no good evidence to suggest that the endemic CFS is transmissible.

Using the Green College working group recommendations, endemic CFS (which they term "post viral fatigue syndrome" or PVFS) is held to encompass two sub-groups, namely one comprised of patients with disease of spontaneous onset (which they term "chronic fatigue syndrome" or "CFS") and another consisting of patients proven to have an infection associated with the onset of chronic fatigue (which they term "post infectius fatigue syndrome" or "PIFS"). For the purpose of this paper, the British term "PVFS" is held to be entirely synonymous with the term CFS used in other countries in the world. A number of authorities suggest that a range of viruses and other infections may be responsible for the onset of CFS. The viruses considered by proponents of this hypothesis are ubiquitous in the community and in almost every case cannot be said to relate to exposure to a particular environment or particular type of occupation.

There is also a contrary body of medical opinion which suggests that any infective agents found to be present have not caused CFS but rather are secondary or coincidental to its manifestation. An plausible, alternative explanation is that CFS is a condition of spontaneous onset which manifests in a way similar to viral infections.

Under such a model, the identification of a particular viral strain is held either to be coincidental to the condition or is a secondary event which follows changes in the patient's immune system as a result of CFS. Recent research findings are consistent with this counter view.

In light of this, any claims that the onset of CFS was as a result of an infection acquired in the work place should be treated as speculative unless (a) the claimant has serological evidence to indicate recent infection with a particular viral agent and (b) there is unequivocal evidence of an epidemic infection within the work place with the type of viral agent infecting the claimant. It is recommended that claims based on alleged infections acquired traveling to and from work should generally fail. Likewise, it is recommended that claims in which it is alleged that an infection is unique to a region to which the claimant has been posted or were acquired during the course of work-related travel should be treated with scepticism unless the conditions of (a) and (b) above are both satisified.

There is no evidence that heredity, genetic or developmental factors play a part in the onset of CFS. Nor is there any consistent evidence that the condition is associated with particular types of occupation, lifestyle, mental or physical stress or pre-existing psychiatric illness. All of the studies of patient characteristics fail to report an increased incidence in any one occupational group, social class or region. The evidence that stress has a contributing role is considered to be equivocal. In particular, there are no studies to show a role for work-related stress in the development of CFS. None of the authorities reviewed here make mention of either physical or mental stress as a cause of CFS. The findings of Dowsett et al. that those with existing CFS report aggravation of their symptoms when exposed to strenuous exercise, other physical stresses and mental stresses is expected, given the nature of CFS and its widespread effects on physical and psychological function.

The onset of psychiatric illness in CFS is shown to be secondary to the impairment of body function in general and the chronic pain that is part of the condition. There is little, if any, evidence to support a role for any form of psychiatric illness or personality type in the onset of chronic fatigue syndrome.

Investigation of CFS patients is aimed mainly at excluding other illnesses as the cause of fatigue. There are no tests which can be considered to be diagnostic of the condition. Accordingly, there would appear to be no justification for undertaking sophisticated serological investigations nor highly technical radiological investigations in the ordinary course of investigation of the disease unless there are good reasons to suspect a serious occult disease (such as malignancy) as the cause of fatigue.

Treatment of the condition is aimed at ensuring an adequate degree of rest in conjunction with a supervised course of gentle graded exercises throughout the course of the illness. Treatment of specific symptoms such as muscle pains and depression are recommended as being appropriate but the use of narcotic and other addictive forms of medication would appear to be inappropriate in all cases. A range of other therapies have been trialed at various times but none have yet to be shown to offer any particular benefit.

Whilst it is not possible to provide specific guidelines to assist in determining the prognosis of individual patients, there is reasonable uniformity of opinion that an improved prognosis is associated with early treatment (including work and lifestyle modifications). Conversely, when the illness is severe or has been present for more than one year or both, the prognosis is generally held to be poorer. Accordingly, early identification of CFS patients is important. In many cases, temporary removal from the work place is recommended (especially those with severe disease) until there is good evidence of a sustained recovery. For the remainder, it is recommended that placement on lighter duties and/or reduced working hours be initiated early in the course of the illness. Unless the patient has severe disease however, there would appear to be little justification for recommending mandatory removal from the work place if the person is not required to undertake moderate or strenuous duties and if suitable changes to the work environment can be made.

Dr Peter Grant pgrant@gil.com.au

Bibliography

Behan P O, Bakheit A M O. 1991. Clinical spectrum of postviral fatigue syndrome. British Medical Bulletin. 47(4):793-808.

Cathebras P J, Robbins J M, Kirmayer L J, Hayton B C. 1993. Fatigue in primary care: prevalence, psychiatric comorbidity, illness behaviour, and outcome. Journal of General Internal Medicine. (May-June) 7(3):276-286.

Cunningham L, Bowles N E, Archard M B A. 1991. Persistent viral infection in muscle in postviral fatigue syndrome. British Medical Bulletin. 47(4):852-871.

David A S. 1991. Postviral fatigue syndrome and psychiatry. British Medical Journal. 47(4):966-988.

David A, Pelosi A, McDonald E, Stephens D, Ledger D, Rathbone R, Mann A. 1990. Tired, weak, or in need of rest: fatigue among general practice attenders. British Medical Journal. (November 24) 301(6762):1199-1202.

Dowsett E G, Ransay A M, McCartney R A, Bell E J. 1990. Myalgic encephalitis - a persistent enteroviral infection? Postgraduate Medical Journal. 66:526-530.

Gold D, Bowden R, Sixbey J, et al. 1990. Chronic fatigue; a prospective clinical and viral study. Journal of the American Medical Association. 264(1):48-53.

Gow J W, Behan W M H. 1991. Amplification and identification of enteroviral sequences in the postviral fatigue syndrome. British Medical Bulletin. 47(4):872-885.

Hickie I, Lloyd A, Wakefield D, Parker G. 1990. The pscyhiatric statusof patients with chronic fatigue syndrome. British Journal of Psychiatry. 156:534-540.

Holmes, G P. 1991. Defining the Chronic Fatigue Syndrome: Problems in Design and Analysis.. Reviews of Infectious Diseases. 13 (Suppl 1): S53-55.

Holmes G P, Kaplan J E, Gantz N M, et al.. 1988. Chronic Fatigue Syndrome: A Working Case Definition. Annals of Internal Medicine. 108:387-389.

Ho-Yen D O, McNamara I. 1991. General practitioners' experience of the chronic fatigue syndrome. British Journal of General Practice. 41(349):324-326.

Jamal G A, Miller R G. 1991. Neurophysiology of postviral fatigue syndrome. British Medical Journal. 47(4):815-825.

Jenkins R. 1991. Epidemiology: Lessons from the past. British Medical Bulletin. 47(4):952-965.

Kennedy P G. 1991. Postviral fatigue syndrome: current neurobiological perspective. British Medical Journal. 47(4):809-814.

Kirmayer L J, Robbins J M, Kapusta M A. 1988 Somatization and depression in fibromyalgia syndrome. American Journal of Psychiatry. 145:950-954.

Komaroff A. 1988. Chronic fatigue syndomes related to chronic viral infection. Journal of Virology Methods. 21:3-10.

Kroenke K. 1991. Chronic fatigue syndrome: Is it real? Postgraduate Medicine. (February 1) 89(2):44-55.

Kroenke K., Mangeldorff A D. 1989. Common symptoms in ambulatory care: incidence, evaluation therapy and outcome. American Medical Journal. 86(3):262-266.

Kruesi M J, Dale J, Straus S E. 1989. Psychiatric diagnoses in patients who have chronic fatigue syndrome. Journal of Clinical Psychiatry. 50(2):53-56. [Erratum published in Journal of Clinical Psychiatry. 1989 50(4):148.]

Lloyd A R. 1990. Muscle versus brain: chronic fatigue syndrome. Medical Journal of Australia. (November) 153:530-533.

Lloyd A R, Hickie I, Boughton C R, Spencer O, Wakefield D. 1990. Prevalence of chronic fatigue syndrome in an Australian population. Medical Journal of Australia. (November) 153:522-528.

Loblay R, Swain A. 1986., Food intolerance., in Whalqvist M, Truswell A. (eds.) Recent Advances in Clinical Nutrition. London. John Libbey. pp. 169-177.

Makowska Z, Kluge G, Sprusinska E. 1992. The influence of occupational and non-occupational factors on chronic fatigue syndrome. Polish Journal of Occupational Medicine and Environmental Health. 5(4):323-333.

Manu P, Lane T J, Matthews D A. 1993. Chronic fatigue and chronic fatgiue syndrome: clinical epidemiology and aetiological classification. Canadian Medical Association Journal. (August 1) 149(3)34-319.

McBride S J, McCluskey D R. 1991. Treatment of chronic fatigue syndrome. British Medical Bulletin. 47(4):895-907.

Melamed S, Kushnir T, Shirom A. 1992. Hyperimmune fatigue syndromes: diseases of the stress response? Journal of Rheumatology. (March) 20(3):418-421.

Millon C, Salvato F, Blaney N et al. 1989. A psychological assessment of chronic fatigue syndrome/chronic Epstein barr virus patients. Psychological Health. 3:131-141.

Ng S C. 1992. The fibromyalgia syndrome. Singapore Medical Journal. (June) 33(3):294-295.

Prasher D, Smith A, Findley L. 1990. Sensory and cognitive event-related potentials in myalgic encephalopathy. J. Neurol. Neurosurg. Psychiatr. 53:247-253.

Shafran S D. 1991. The chronic fatigue syndrome. [Review] American Journal of Medicine. 90(6):730-739.

Sharpe M. 1991. Psychiatric management of PVFS. British Medical Bulletin. 47(4):989-1005.

Straus S E, Dale J K, Tobi M, et al. 1988. Acyclovir treatment of chronic fatigue syndrome: lack of efficacy in a placebo-controlled trial. New England Journal of Medicine. 319(26):1692-1698.

Stricklin A, Sewell M, Austad C. 1990. Objective measure of personality variables in epidemic neuromyaesthenia patients. South African Medical Journal. (January 6) 77(1):31-34.

Taerk G S, Toner B B, Salit I E, Garfinkel P E, Ozersky S. 1987. Depression in patients with neuromyaesthenia (benign myalgic encephalomyelitis. Intermational Journal of Psychiatric Medicine. 17:48-49.

Wallace P G. 1991. Epidemiology: A critical review. British Medical Bulletin. 47(4):942-951.

Wessely S, David A, Butler S, Chalder T. 1989. Management of chronic (post-viral) fatigue syndrome. Journal of the Royal College of General Practitioners. 39(318):26-29.

Wessely S, Powell R. 1989. Fatigue syndromes: a comparison of chronic 'post viral' fatigue with neuromuscular and affective disorders. J. Neurol. Neurosurg Psychiatr. 52:940-948.

Winters E G, Quinet R J. 1992. Chronic fatigue syndrome. Journal of the Louisiana State Medical Society. 144(6):260-270.

Woods T O, Goldberg D P. 1991. Psychiatric pesrpectives: An overview. British Medical Bulletin. 47(4):908-918.

Yaminishi K. Chronic fatigue syndrome and virus infection: human herpesvirus 6 (HHV-6) infection. Nippon Rinsho - Japanese Journal of Clinical Sciences. (November) 50(11):2612-2616.