June 20, 2025

expert reaction to observational study of blood biomarkers in ME/CFS patients

An observational study published in EMBO Molecular Medicine looks at blood biomarkers in ME/CFS patients. 

Prof Alan J Carson, Professor in Neuropsychiatry, University of Edinburgh and Consultant Neuropsychiatrist, Royal Infirmary Edinburgh, said:

“This press release overstates the significance of the findings. The authors look at a vast array of data without any priori hypothesis, or data plan that was pre-registered, and they essentially look for differences at a statistical level of a 1 in 20 chance. One would expect approximately 150 positive findings by chance and they found 116. This is not very exciting. They did not find any ME ‘signature’, they did not replicate their findings, contrary to the press release, only some of them. The diagnosis was based simply on people identifying as having ME/CFS and the controls were not well described other than being healthy. A multitude of other factors such as rates of depression, obesity etc. – all of which produce similar results to this type of study were not controlled for.

“If one wants to ask is ME/CFS a figment of the imagination then this study shows it is not. However, to claim blood biomarker differences prove that a condition is ‘not all in your head’ fails to appreciate that both physical and mental illness can show similar types of results. Studies of depression show similar results; although one would hope without using similar stigmatising and derogatory language. See for example https://www.nature.com/articles/s41380-025-02919-z.

“I don’t think this study takes us anywhere particularly new. Biological abnormalities were first shown in ME/CFS some 40 years ago and whilst this offers some refinement, and a technically wider array of molecules studied, it doesn’t fundamentally change understanding. I would class it as of modest interest to researchers in the field.”

Prof Kevin McConway, Emeritus Professor of Applied Statistics, Open University, said:

“I think this is an important piece of research, but it’s also important to be careful not to claim too much from its findings. There’s a lot more to do.

“The press release and the research paper both make it clear that these findings could help in finding a set of blood biomarkers that can reasonably reliably distinguish people with ME/CFS from those who do not have that condition, but that, without a lot of further work, the findings do not in themselves provide such a set of biomarkers. For instance, the last sentence of the abstract of the paper says, “Nevertheless, their number [of traits that differed between people with ME/CFS and people without that condition], diversity and lack of sex bias keep alive the future ambition of a blood-based biomarker panel for accurate ME/CFS diagnosis.” I hope personally that that ambition can be achieved, but the researchers are careful not to say that their findings indicate that it will definitely be achieved.

“A strength of the study is that it uses data from the very large UK Biobank study, based on over 1,400 people who reported they had been diagnosed with ME/CFS and over 130,000 ‘controls’ who had not had that diagnosis, as well as data from a smaller (but still quite large) US study called All-of-Us. 

“But, in the research paper, the researchers are very careful to say that they are reporting associations, that is, correlations, between blood measurements and whether or not people have ME/CFS, and that, to quote the paper, “no causal statements are made” about those associations. That’s essentially because data from the UK Biobank is observational. Any differences between the group with ME/CFS and the controls without ME/CFS could be caused by the different disease status, but it could also, in whole or in part, be caused by other differences (so-called potential confounders) between people with and without ME/CFS that are not a direct consequence of that condition. 

“The researchers did use methods of what’s called causal inference to try to throw further light in what causes what, and in particular they found that the differences in blood measurements were unlikely to stem from the fact that people with ME/CFS typically exercise less than people without that condition. That’s a useful and important finding, I think. But other potential confounders couldn’t be dealt with in a similar way, so other aspects of cause and effect just can’t be sorted out. Indeed (as the researchers mention) the possible existence of other confounders means that the assumptions behind the analyses involving exercise may not entirely be valid. To get further with all this will need a lot more, and different, research, including work on what may actually be causing the observed differences within people’s bodies.

“There are also some issues stemming from the use of data from the UK Biobank. Again this is reported in the research paper. For instance, participants who volunteered for the Biobank are healthier than the average UK population, and the research paper mentions that people with severe ME/CFS may simply not have been able to go through the assessment and data collection process required, and so are unlikely to have contributed towards the findings on a large scale.

“Also, because the recording of ME/CFS diagnoses took place some time ago, people’s status on ME/CFS is not in accord with the definitions of the condition that are generally used now. Roughly half of the people who were treated as having ME/CFS did not state that they had post-exertional malaise (PEM for short, a major worsening of symptoms after even minor mental or physical exertion). Post-exertional malaise is now generally considered an essential part of ME/CFS, and people who do not have it would under most up-to-date conditions not be considered to have ME/CFS. But in the past, post-exertional malaise was not considered an essential part of the definition of the disease, so people in the UK Biobank who were diagnosed with ME/CFS in the past might not have had post-exertional malaise.

“Arguably, this does not really weaken the findings of this study.  The strongest evidence on potential biomarkers was in people who did have post-exertional malaise. But the study did still find some differences in potential biomarkers between people who had had an ME/CFS diagnosis but did not report post-exertional malaise, and the control people who had never had an ME/CFS diagnosis. If these people who would once have been diagnosed with ME/CFS, and who may still have really disabling and long-lasting symptoms, are defined as not having ME/CFS and are not included in developing biomarkers, does that have consequences for the treatment they can receive? Obviously this new study isn’t intended to answer that kind of question, but it’s something that shouldn’t be forgotten as biomarker research for ME/CFS moves on.”

‘Replicated blood-based biomarkers for Myalgic Encephalomyelitis not explicable by inactivity’ by Sjoerd Viktor Beentjes et al. was published in EMBO Molecular Medicine at 00:01 UK time on Friday 20^th June. 

Declared interests

Prof Alan J Carson: I led writing of the Scottish Good Practice Statement in ME/CFS in 2008

I am an editor of Journal of Neurology Neurosurgery and Psychiatry

I am past president of British Neuropsychiatry Association and the past president of the (international) Functional Neurological Disorders Society

I have given independent testimony on Court on Neuropsychiatric topics.

Prof Kevin McConway: No conflicts of interest