Trial By Error: Professor Jonathan Edwards’ View of ME

By David Tuller, DrPH

Professor Jonathan Edwards posted this essay a while ago on the Science For ME forum. I only noticed it recently. Professor Edwards, a retired rheumatologist from University College London, has played a key role n the last few years as an advocate for patients as well as proper science. So I thought it would be helpful to share his thoughts on a topic we all wonder about: “What is ME?”

I am reposting this with Professor Edwards’ permission. The views expressed are obviously his.


What is ME?: A Medical Outsider’s Viewpoint

Jonathan Edwards

This an account of my thoughts about the nature of ME and the issues surrounding it, put together in response to queries. I hope it may be of interest both to patients and carers and to people with a medical or research interest in the illness.

I used to be a rheumatologist and biomedical scientist. I came across a few people with what I now realise was ME but it was not part of my routine practice. I became interested in ME later, having been asked to advise on immunological research. As a practicing doctor I thought that there are many causes of fatigue, some known, some not, but that we did not know enough to say there was a specific illness called myalgic encephalomyelitis, or ME. I have changed my mind. I think there is. But why I think it is useful to talk of ME needs some explaining.

For me the main reason is that the pattern of illness of ME is not like the fatigue caused by chronic diseases like rheumatoid arthritis or heart failure. The more I listen to people with ME the more it seems very unlikely that the mechanisms of fatigue in these other conditions are relevant. If the picture of ME can be described as ‘fatigue’ at all, it seems closer to the complex of symptoms of acute viral infection, or the spill over that occurs most often with EBV infection as ‘post-viral fatigue’ over several months. In general the fatigue of chronic diseases is predictable and shows up immediately on exertion. In ME symptoms appear as ‘payback’ unpredictably and, in the form of post-exertional malaise, with variable delay. Most notably, the level of symptoms and disability is out of proportion to any measurable inflammation or biochemical disturbance. It is also out of proportion to any mood change that might suggest a depressive illness. A pattern of symptoms similar to ME may occur in association with systemic lupus, Sjögren’s syndrome, Reiter’s syndrome and possibly Hashimoto’s disease but it is hard to know if it is ever quite the same.

The distinct picture of ME is, I think, enough to regard it as a specific syndrome, implying that we should expect there to be some specific physiological malfunction underlying it, just as a malfunction of glucose control underlies ‘diabetes’ and a malfunction of bronchial smooth muscle underlies ‘asthma’. That does not imply that there is a single specific cause, such as a specific infection. All the evidence points to ME as following a range of infections or with no environmental factor. It is often wrongly assumed that illnesses arise purely from a combination of genetic and environmental factors, but many illnesses may arise, with or without genetic predisposition, by chance, as a spin off from the constant to and fro of internal events associated with growth, repair and regulation. I suspect that this is a large component of causation in ME.

ME is also referred to as chronic fatigue syndrome (CFS). Both terms have their pros and cons. ‘CFS’ emphasises that this is a syndrome, but is too easily confused with the generic concept of chronic fatigue of any cause, which, as indicated, is something quite different. ‘ME’ emphasises that this is a specific illness pattern, but, confusingly, implies a pathology (encephalomyelitis) that is not present.

ME may be the preferable term, but it arose from a historical muddle. The term was promoted by Acheson, who believed that patients developing an acute illness with neurological symptoms in an outbreak at the Royal Free Hospital, recorded by Ramsay, and also in earlier outbreaks, might be suffering from an enteroviral infection similar to polio. The reason for thinking that this was an encephalomyelitis was the combination of symptoms suggesting muscle paralysis from nerve involvement together with symptoms suggesting brain involvement.

However, with the passage of time it has become clear that at least in the outbreaks that occurred in relatively recent history there was no evidence of any spinal cord or brain lesions comparable to those of polio. Muscle fatigability and weakness was not due to neural damage. A proportion of cases from the Royal Free outbreak were documented by Ramsay as developing a chronic illness of the type we now call ME, but this is something different from the acute neurological picture that led Acheson and others to use the term.

In retrospect, the outbreaks of so-called ME documented from the 1930s to the 1980s probably tell us nothing useful about the illness called ME today. The overwhelming majority of cases of ME today do not occur as part of ‘ME outbreaks’, even if many follow viral infections that have spread rapidly in communities in the way normal for most viral infections.

This confusion has been particularly unfortunate because in 1970 McEvedy and Beard published a paper suggesting that the few ‘severe’ cases with ‘neurological’ features in the Royal Free outbreak were cases of mass hysteria. They correctly point out that the signs documented in the records do not fit with any consistent type of neurological lesion. However, as indicated above, there is no reason to think this implies that the syndrome of ME as defined today(or for that matter as defined by Ramsay) is due to any form of hysteria, whatever we think hysteria might mean. An acute illness, with signs that turned out not to indicate neurological lesions after all, has been confused with a chronic disabling condition that may, in a very few cases, have followed on from the former.

That might be a very good reason for not using the term ME, but I think it is still probably the most useful term.


Together with a group of patients and carers I reviewed what I thought one could say about causes of ME a couple of years ago in an article in Fatigue and Biomedicine (Edwards et al., 2016). My thoughts remain much the same.

ME clearly involves problems for the brain, in terms of impaired thinking, mental fatiguability and sensitivity to light and sound. There are also clues that it involves the immune system in at least some cases. An effect of immune dysfunction on the brain would make sense, but the immune system may only be disturbed initially, leaving problems only in the nervous system. Interest in immune problems was sparked by a suggestion of effectiveness of the drug rituximab in a phase 2 trial. However, the phase 3 trial showed no effect, so the evidence for an ongoing immune aspect comes largely from symptoms. In the past there has been a suggestion that ME is associated with immunodeficiency but the evidence is weak and inconsistent.

Superficially, the difficulty in maintaining any form of exertion suggests that ME might be due to a defect in energy production. However, no gross abnormalities in relevant biochemistry have been found. Moreover, symptoms like pain do not fit into a simple energy deficit. Mitochondrial function might be abnormal in ME, but the symptoms would probably fit better with a problem with the danger signalling function of mitochondria than with actual energy shortage. Post-exertional malaise and the reduction in performance observed in cardiopulmonary exercise tests at day 2 may indicate some form of inhibitory danger signal switching off full metabolic capacity (as occurs due to cytokine production during a viral illness).

For a problem to develop with cell signalling throughout the body in someone who has previously been well, there must be some new chemical signal reaching all the cells. The only likely source of that (other than neural signals from the brain) looks to be the immune system, with the signal being some form of autoantibody, or a cytokine produced by T cells. While this is an attractive idea in some ways, it is quite hard to see how either mechanism would work in specific biochemical terms.

Separating out neural signalling from innate immune signalling may not be straightforward because they are linked through the hypothalamus. However, it seems reasonably plausible that following initial immune disturbance the continuing production of symptoms is due to some form of error in, or damage to, a control mechanism in the brain. The situation might resemble that of tinnitus or narcolepsy, where very specific damage to brain circuits leaves the brain unable to control signals properly. In that sense it might be fair to say that at least some ME is ‘all in the head’. Nevertheless, that is very different from suggesting it is ‘all in the mind’.

What would it actually mean for ME to be ‘psychological’ or ‘all in the mind’? Since this is a common view, it is worth considering whether it has any basis. There has probably always been a popular belief that inexplicable illnesses like ME are ‘caused by the mind’. At least since the nineteenth century doctors have tried to make the popular prejudice respectable by using terms like ‘somatisation’ or ‘biopsychosocial’. The biopsychosocial (BPS) claim is that the illness starts off with a biological (or physical) trigger but then is perpetuated by ‘psychological factors’ in the form of unhelpful beliefs.

Medical science normally assumes that ‘physical’ covers all living processes, so regards all illnesses as physical. Enthusiasts for the BPS approach (who do not seem to follow this) try to twist the normal view into one of ‘separating mental from physical like Descartes’. But if everything is physical there is no separation. It is the BPS enthusiastswho separate physical and mental and, exactly like Descartes, claim the two ‘interact’.

Even if the BPS enthusiasts do not understand philosophy, their theory might still be useful. However, apart from the fact that we have no reason to think that there is such a thing as mental causation in addition to physical causation, descriptions of events in mental terms (thoughts) very rarely provide useful testable theories because too much is unknown. Psychology tries in vain to build testable theories but it is hard to see how it can succeed until we understand how thoughts actually work.

Moreover, there are good reasons for thinking that it makes no sense to suggest that, for instance, ‘unhelpful beliefs’, about being unable to exercise or having a certain illness, cause persistence of ME. The most obvious thing is that a lot of people with ME, at least for long periods of their illness, have no particular beliefs about their illness. They try doing exercise, because they have no belief that exercise is bad, until they find they are getting nowhere fast. Most people in the world with ME have never heard of an illness name. Even those who have mostly spend months or years having no idea what is wrong before being diagnosed.

The lack of a causal role of beliefs in disability is also indicated by the PACE trial, which shows that, at least for the sort of people who agreed to take part, treatment aimed at reducing unhelpful beliefs, despite leading to people reporting that they do not believe they are quite so ill, makes no difference to their disability. It seems to show quite clearly that beliefs are not the problem.

There is no doubt that, at least on internet forums, some people with ME hold strange beliefs about their illness. But people with cancer and arthritis also hold strange beliefs, particularly if medical advice does not offer help. Moreover, one has to expect people with an illness for which doctors have no explanation at all to be more likely to have unsubstantiated beliefs. They are likely to have some beliefs and by definition they will be unsubstantiated, because nobody knows! And it is also not very surprising that people with ME who have firmer beliefs that there is some physical cause for their illness tend to be less likely to get better. If there is something badly wrong, people are likely to sense it! This is about the only piece of evidence the BPS theorists can muster to support their view, but since it does not discriminate their view from the obvious alternative it tells us nothing.

So even in the intuitive lay sense that we think thoughts cause us to do things, and so, thoughts might cause symptoms of illnesses, it just does not fit the facts for ME. And Freudian-type theories of unconscious thoughts and motivations make no more sense, in addition to being notoriously untestable.

Put simply, the BPS theorists are unable to demonstrate to the medical world in general that they have a coherent theory grounded in reliable observations that makes useful predictions about effects of treatments. Despite attempts to give them a positive spin, treatments based on the BPS approach appear to make no difference to disability (Wilshire e al. 2017). Psychological theories of ME remain popular prejudices with fancy names, without any scientific grounding.


I have been asked to comment on the issue of ‘harassment’ of health professionals by people with ME. The ME patient community has had to put up with all sorts of unhelpful beliefs (almost all of which come from doctors and therapists). Stories about polio-like illnesses or defective lymphatic drainage from the brain may be as unhelpful as stories about ‘biopsychosocial’ causation. The difference with the BPS belief is that it implies contempt for the person’s ability to understand their own nature. Not surprisingly, health professionals who champion unhelpful beliefs about psychological mechanisms get complaints. They get even more complaints when they do bad science to try to prove their unhelpful beliefs.

It may be that on a few occasions over the last ten years people with ME have said some pretty nasty things to doctors peddling unhelpful psychological beliefs. But for psychiatrists that should be an issue of no consequence. The problem has been grossly hyped. Where children and adolescents are involved emotions are likely to run particularly high and it is hardly surprising if voices have been raised. The condescending manner of medical staff is probably the main reason for stirring up anger and emotion. I nearly always get angry with the condescending ignorance of health care professionals when I am visiting a sick relative, whatever is wrong with them. Health care provision (at least in the UK) is lousy much of the time yet those delivering it seem too often to be under the illusion that they are doing a great job. It is time people learnt to admit that often they do not know what they are doing.

What are the most urgent steps needed to take to improve the situation for ME sufferers? We need everyone in research and healthcare to agree that we have no idea what the mechanism of the illness is and to start from scratch, looking for reliable evidence. That means abandoning both the incoherent BPS model based on popular prejudice, and theories hung over from sixty years ago about infectious diseases that are no longer of any relevance to the illness people have to deal with. It means psychiatrists, physicians and therapists accepting that they have to care for people on a pragmatic basis in response to symptoms without any belief that they know what is good for patients and the condescension associated with that.

Cognitive behavioural therapy and graded exercise therapy involve ‘educating’ patients in an explanation for their problems that has no basis. The evidence is that they make no difference to disability, so they are unethical and should be abandoned. The use of anti-virals and antibiotics similarly has no sound basis and should be discontinued, at least until there is reliable evidence for efficacy.

We need research on a broad front to identify leads to focus efforts on. Genetics has to be a worthwhile area to cover. There are indications that there is genetic susceptibility. Finding out what the genes are may only provide pointers rather than specific answers, but these would still be powerful.

In my view chasing viruses or gut bacteria is likely to be a waste of time. It is unlikely that any one infectious agent is important. Shifts in gut bacteria seem equally unlikely to be important. If micro-organisms were relevant to the persistence of the chronic illness then amongst the millions of people with ME there should be at least a few thousand cases that provide clear clinical clues to how. Epidemiology should have shown something up. There should be a few hundred cases where a virus has multiplied enough to be directly measurable or to show specific tissue damage, or where ME has vanished in response to moving to a new country with a new diet.

Brain imaging is likely to be worth exploring in a much more detailed way. Replication studies are particularly needed because a number of changes have been reported in ME once and then never repeated.

Metabolic pathways capable of generating abnormal signalling are worth looking at in detail. I am not convinced that special emphasis should be put on mitochondria, but they may yet prove to be central to the disturbance that underlies the pattern of illness.

Reference: Edwards et al. The biological challenge of myalgic encephalomyelitis/ chronic fatigue syndrome: a solvable problem. Fatigue. 2016; 4(2): 63–69.


{ 52 comments… add one }
  • Sonya 12 December 2019, 5:23 pm

    How about this. CCI has neurological symptoms. If one has a diagnosis of CCI, that patient’s neurological symptoms now have an explanation. Having POTS does not mean a CCI patient has ME/CFS. It means they have another Dx. PEM occurs in other diseases and is reported in EDS patients. I suspect a number of illnesses, such as EDS and CCI, are being misdiagnosed as ME/CFS.

    I am thinking more and more that yes, ME/CFS should be a Dx of exclusion. Rule out neurological and congenital diseases that would explain symptoms. Even with IOM and CCC, which I believe are good criterion, there should be full neurological and congenital disease work ups that could explain symptoms especially when there is no infectious event that a patient could point to as the definitive beginning of their symptoms. Can one have another illness along with ME/CFS? Of course. But it is becoming very clear that a number of patients have received an ME, ME/cfs, or CFS Dx that should not have. It especially worries me that MS patients have mistakenly received an ME Dx. Or my wonderful neighbor, who has had diabetes since I met her in 2003, now has a CFS Dx where at age 64 had to leave her job as a university professor due to “CFS”.

    I am also starting to believe that if a person had an enterovirus and now have ME/CFS symptoms, they most likely don’t have ME or ME/CFS, but symptoms that result from such an infection.

    Thinking of the many that have an ME, ME/CFS, or CFS Dx that are severely neurologically impacted, I now believe they have an incorrect Dx or perhaps an additional Dx that would explain their neurological symptoms. That’s a shame. I think the ICC is a very bad criterion impeding a correct Dx of other neurological and congenital diseases.

  • deboruth 12 December 2019, 8:40 pm

    Dr. Edwards’ essay is thematically quite sympathetic to patients and their plight, which is much appreciated by this patient. The text periodically stumbles into inaccuracies that seem based on gaps in acquaintance with prior research. As there are some 9000 peer reviewed papers — a body still growing– plus books, grey research not in bibliographies or data bases, and private material, gaps are to be expected. It seems more helpful to gift Dr.Edwards with appropriate references and points of difference than to take a condemnatory or rejecting attitude. E.g. Caligiuri, Fletcher, Hornig, et alia on immune. Lerner, Proal, Chia, Peterson et alia on pathogens and effective medications. Johnson in Osler’s and other writings on severe neurologic events. We have few enough allies in the British medical world. Let’s not gratuitously discard one.

Leave a Comment