Central origin of fatigability in Myalgic encephalomyelitis/chronic fatigue syndrome revealed by multimodal neuroimaging, 2026, Bedard, Nath, Walitt+

[Utsikt]

They also used BOLD for their imaging, a technique we now know might give you the opposite results than what’s actually going on, like indicating increased metabolism when it’s actually decreased:

Post in thread 'Altered corticostriatal connectivity in long-COVID patients is associated with cognitive impairment, 2025, Troll et al'

Dec 18, 2025

This entire body of literature may have to be thrown out.

Lazily copy-pasting the Google AI description —

Resting-state functional connectivity (rs-fMRI) maps brain networks by analyzing spontaneous, synchronized fluctuations in the BOLD signal (oxygen levels) between different brain regions while a person is resting, revealing how areas functionally "talk" to each other even without a task. This powerful, non-invasive tool identifies intrinsic brain networks (like the Default Mode Network) and helps study brain organization in health and diseases like Alzheimer's, autism, and...

• SNT Gatchaman

• 

 

[Utsikt]

We assessed subjective appraisals of physical and cognitive state fatigue with a visual analog scale immediately before and after the grip force task (~1 min; Fig. 1B).

We used SuperLab (Cedrus Corporation, San Pedro, CA) to present visual stimuli and record participant's subjective physical and mental fatigue appraisal. Participants moved a vertical red bar left or right over a visual-analog scale with a hand-held device while in the scanner. We collected subjective fatigue evaluation for thirteen HV and ten ME/CFS, because it was implemented after we had begun the MRI data collection.

There is no info in the main text about how they explained what «physical» and «mental» fatigue was when using the VAS, or if they even defined it at all.

I would not be surprised if people with ME/CFS and healthy volunteers had different interpretations.

 

[Medfeb]

But wouldn't such mitochondrial or energy issues be downstream of some sort of 'signalling' ultimately coming from the brain?

Is that true? Viruses are known to disrupt energy production. Presumably without brain signaling

 

[PageofME]

I'd be interested in what top-tier brain scientists say about this study. It seems to me to belong to the field of blobbology: the science of creative speculation over colourful pictures of the brain. The method has always struck me as the successor to the Rorschach test.

I remember when one of the authors, Avindra Nath – a neurologist by training and Clinical Director at NINDS – published his big NIH study on ME/CFS patients, concluding that they had proved that ME/CFS was "in the brain". A couple of real neuroscientists showed up in the comments on Health Rising explaining why this had nothing to do with neuroscience. They said it was ridiculous to point to a dip in blood oxygen in the so called temporoparietal junction (TPJ) – of which no one exactly knew what it's there for – and frame a sick person's rational decision to conserve energy as a brain disease.

In history, what doctors like Nath here are doing is called a modernisation of a hegemonial narrative. It means that when an old scientific explanation like "ME/CFS is a disorder of the psyche" becomes untenable, the idea doesn't get abolished by the medical establishment altogether, but gets a new package. Since psychiatry has become neuropsychiatry over the past decades, ME/CFS now becomes a "brain disease" in name, while retaining the same old psychogenic assumptions.

 

[PageofME]

Is that true? Viruses are known to disrupt energy production. Presumably without brain signaling

I'd give you ten likes if I could!

 

[Medfeb]

People have wanted to believe that there is an energy defect since the 1980s at least. Science has failed to find it since the 1980s at least. And it doesn't fit the clinical picture of delayed PEM.

In the AfME webinar today there was a lady called Maree who did an excellent job of explaining PEM and what we do and don't know about it. She was followed by two exercise scientists who, interestingly, made no mention of 2 day CPET being a measure of PEM. Maureen Hanson discussed changes in proteins after exercise and they were mostly immune and neural I think.

I think one explanation is better than two and I would agree with @InitialConditions 's succinct comment.

Couldn't disruption in the cascade of physiological processes that normally occur post-exercise account for delayed PEM? A disruption in the normal "recovery" processes?

For instance, this paper from Hansen's lab found "Changes in proteins involved in protein folding and the endoplasmic reticulum (ER) stress response during recovery [that] were highly correlated with PEM severity." As I understand it, many of the proteins they examined were involved in mitochondrial and metabolic function and they noted that mitochondrial dysfunction could "contribute to the impaired recovery and postexertional malaise."

This paper is admittedly small so its only the beginning of a discussion but I think raises important questions

How the brain is involved in this post-exercise recovery response, I can't say.

 

[Violeta]

We studied whether fatigability in ME/CFS is related to central or peripheral mechanisms.

Why not both?

 

[SNT Gatchaman]

Leaving aside the question of whether BOLD signal on fMRI is being interpreted in the right direction (Utsikt's comment #21), my first question around these findings would be: is this a trained outcome for a central nervous system that has long adapted to rapid (peripheral) neuromuscular fatiguability?

Our most novel results relate to the lack of neuromuscular adaptation in ME/CFS compared to the HV. ME/CFS did not change their muscular and brain activity as measured via DI, EEG, BOLD, and CMC, and as a result, they demonstrated fatigue earlier than HV as shown by the earlier occurrence of the B n block and the steeper decline of grip force across the five task blocks.

HV had higher CMC-afferent than ME/CFS in the beta and gamma bands and HV increased their CMC-afferent across blocks in the gamma band.

Dynamics of brain-muscle interaction with neuromuscular fatigability: systematic review (2026) —

Crucially, weakening in beta or gamma band CMC during neuromuscular fatigability should not be interpreted as solely a loss of motor drive, as it may signify an adaptive reorganization to maintain motor performance (Peng et al., 2024).

Collectively, these findings indicate that fatigue-related changes in CMC are context-dependent rather than stereotyped and should not be interpreted as a simple proxy for loss of motor drive.

Peng et al is The neuromechanical of Beta-band corticomuscular coupling within the human motor system (2024)

 

[DMissa]

But wouldn't such mitochondrial or energy issues be downstream of some sort of 'signalling' ultimately coming from the brain?

I'm not sure why this would be the case. I could see it going in either direction, but it depends on the level of detail we are talking about. Are we talking about a specific tissue or cell type having a change in mitochondrial function or are we talking about a systemic defect in some aspect of metabolism? Are we talking about cell biology or physiology, or are we drawing that line at all? This is a very broad comment

To disrupt mitochondria the disease would need to be on a very micro scale. Like mitochondria are tiny

Anything substantially affecting a cell will probably have knock-on effects to most of the concerned cell's biology, including its mitochondria. This is not to say that mitochondria must be relevant, but it is to say that observing shifts in their function without accounting for what else might be different about the cell is likely to be misleading, as misleading as would be any approach that zooms in with tunnel vision on as tightly interconnected a web of things as a cell is. What I am trying to get at is that pretty much anything affecting some part of a cell could elicit changes to mitochondra - but this can be said of much of the cell and doesn't mean it is by default important to the overall picture.

But, in terms of one step of a signalling chain gone wrong, yes, they could be relevant - as with any other part of a cell important to its overall function. They are hugely important in many modes of signalling. If signalling events in these processes are relevant to something going wrong in a cell or tissue relevant to the disease, sure, there could be a role. I think the high school idea of zappy beans that make you zippy has waylaid thinking and placed undue emphasis on a tiny corner of one chunk of a much greater and complicated system.

This comes back to what I think Jonathan was trying to teach people with the hypothesis paper... start with reliable clues (eg: from the clinical picture) and use that to narrow things down in the first instance. We also have genetic clues now. As I recall it, among the tier 1 & 2 genes or something like that, FBXL4 was the only interesting mitochondrial one. So if anybody is thinking about mitochondria it would be good to follow these sorts of leads and not the vintage ideas that haven’t turned up anything. People should think about signaling roles for mitochondrial turnover, mtDNA, and maybe calcium dynamics, lipid and ER interactions... things largely neglected outside of one or two recent studies.

 

[Medfeb]

This comes back to what I think Jonathan was trying to teach people with the hypothesis paper...

I think I missed that paper. Can you point me to it?

 

[DMissa]

I think I missed that paper. Can you point me to it?

I will DM it to you as this is already risking off-topic

 

[PageofME]

Is it possible that some factor that affects both brain cells and muscle cells make the central versus peripheral distinction less important? For instance, mitochondrial role/energy production which could have an impact on both central and peripheral and also be a factor with any type of exertion?

Yes, vessel damage isn't just the stronger explanation – it’s a documented finding in ME/CFS.

 

[InitialConditions]

Is that true? Viruses are known to disrupt energy production. Presumably without brain signaling

If you mean in-cell hijacking of energy production / metabolism by viruses, then yes, I supose that wouldn't involve central processes.

 

[PageofME]

Interview with science writer Marianne Apostolides about the history and general characteristics of the contemporary neuropsychiatric research the authors of this study are engaged in:

https://www.madinamerica.com/2026/07/marianne-apostolides/

Her new book Go/No-Go: A Journey Into the Research and Treatment of Mental Health Disorders will be out in October. It's a deep dive into the contemporary rebiologisation (the last attempt dating back to the 19th century) of psychic suffering and the areas of treatment that, according to Apostolides, have arisen in the wake of that shift: neuromodulation (electric stimulation of the brain), digital therapeutics, and pharmaceuticals (including psychedelics).

I can only warn ME/CFS patients here against expecting any cure from this. This type of medical research has not even brought forward cures for psychiatric patients, and it is bound to fail patients just as every psychiatric paradigm that has come before.

 

[neophyte32]

I'm not sure why this would be the case. I could see it going in either direction, but it depends on the level of detail we are talking about. Are we talking about a specific tissue or cell type having a change in mitochondrial function or are we talking about a systemic defect in some aspect of metabolism? Are we talking about cell biology or physiology, or are we drawing that line at all? This is a very broad comment


Anything substantially affecting a cell will probably have knock-on effects to most of the concerned cell's biology, including its mitochondria. This is not to say that mitochondria must be relevant, but it is to say that observing shifts in their function without accounting for what else might be different about the cell is likely to be misleading, as misleading as would be any approach that zooms in with tunnel vision on as tightly interconnected a web of things as a cell is. What I am trying to get at is that pretty much anything affecting some part of a cell could elicit changes to mitochondra - but this can be said of much of the cell and doesn't mean it is by default important to the overall picture.

But, in terms of one step of a signalling chain gone wrong, yes, they could be relevant - as with any other part of a cell important to its overall function. They are hugely important in many modes of signalling. If signalling events in these processes are relevant to something going wrong in a cell or tissue relevant to the disease, sure, there could be a role. I think the high school idea of zappy beans that make you zippy has waylaid thinking and unplaced undue emphasis on a tiny corner of one chunk of a much greater and complicated system.

This comes back to what I think Jonathan was trying to teach people with the hypothesis paper... start with reliable clues (eg: from the clinical picture) and use that to narrow things down in the first instance. We also have genetic clues now. As I recall it, among the tier 1 & 2 genes or something like that, FBXL4 was the only interesting mitochondrial one. So if anybody is thinking about mitochondria it would be good to follow these sorts of leads and not the vintage ideas that haven’t turned up anything. People should think about signaling roles for mitochondrial turnover, mtDNA, and maybe calcium dynamics, lipid and ER interactions... things largely neglected outside of one or two recent studies.

Hello, That is exactly it. Personally, I had an RNA test done, and I have a number of lipid-related genes that are malfunctioning, as well as issues with the endoplasmic reticulum (specifically RETREG1). Along with rather low immunity and mitochodria (no inflammation), these are the two standout findings in my RNA data, which I have available for review. I have a severe case and have been bedridden for 17 months.

 

[Evergreen]

To understand this better, I'd like to see the individual-level data, because the headlines were similar about the EEfRT in the intramural study, but it turned out that 8/15 participants with ME/CFS patterned with the HVs (healthy volunteers).

The data is only available on request.

Put together, these results show that the ME/CFS had limited engagement of their neuromuscular system that was substantially less than what was observed in HV. This resulted in an early decline in performance, and thus, an early onset of physical fatigue...Importantly, this lack of engagement is not conscious or deliberate.

Well, that's reassuring...

Several factors may explain this lack of engagement including, but not limited, to the fear, conscious or unconscious, of triggering post-exertional malaise (Stussman et al., 2020), musculoskeletal/joint pain (Rutherford et al., 2016), or elevated perceived exertion (Barhorst et al., 2020).

...until it's undone. How could a conscious fear of PEM or pain that explains lack of engagement not mean that the lack of engagement is conscious?

This resulted in an early decline in performance, and thus, an early onset of physical fatigue; although, we should be careful before labeling this as fatigue since their neuromuscular system showed much reduced engagement

In the intramural study they did the same thing - defined fatigue narrowly, then said we weren't fatigued:

We measured peripheral fatigue (high:low ratio) and central fatigue (post exercise depression). Both types of fatigue were seen in the HVs but not in the PI-ME/CFS participants. Moreover, testing of effort preference and the participants’ own words (Supplementary Information, p.10) are consistent with this finding. Together these findings suggest that effort preference, not fatigue, is the defining motor behavior of this illness.

Why not see it as fatigue being a bigger beast than you thought, and label the subtypes?

Central acting medications with long half-lives were continued for nine ME/CFS. To address the effects of taking these medications on our metrics, we directly compared the nine ME/CFS who remained on-medication with the six who did not take such medication (Supplementary text section Influence of Medication). In summary, we found no instance where there was a significant group difference between those ME/CFS who took the centrally acting agents and those who did not.

Wouldn't it have to be a (very?) strong effect to be statistically significant between such small groups (9 vs 6 participants)?

While our results suggested that ME/CFS fatigability has a central origin, we would argue that since the fatigability process in ME/CFS did not fully take place, at least not like it did in HV, it may have an earlier origin than the neuromuscular system. The pathophysiology of ME/CFS is marked by a cascade of events affecting the autonomic, neuroendocrine, immunologic, bioenergetic, and physiologic systems (Hornig et al., 2015; Missailidis et al., 2019; Nelson et al., 2019; Paul et al., 2021). These processes are most likely the source for the lack of engagement of the neuromuscular system (Walitt et al., 2024).

@DMissa explained above:

As I recall it, among the tier 1 & 2 genes or something like that, FBXL4 was the only interesting mitochondrial one. So if anybody is thinking about mitochondria it would be good to follow these sorts of leads and not the vintage ideas that haven’t turned up anything. People should think about signaling roles for mitochondrial turnover, mtDNA, and maybe calcium dynamics, lipid and ER interactions... things largely neglected outside of one or two recent studies.

I wonder where Hornig, Nelson and Paul (the other researchers cited) would suggest people look?

 

[Eleanor]

To understand this better, I'd like to see the individual-level data, because the headlines were similar about the EEfRT in the intramural study, but it turned out that 8/15 participants with ME/CFS patterned with the HVs.

I've been wondering this too.

 

[Evergreen]

I've been wondering this too.

Especially in light of Jones 2011, who found:

In contrast to the sedentary normal controls, the CFS patients undertaking exercise at 35% of their individually determined MVC exhibited a wide range of PCr fractional depletion spanning from almost zero to within normal control levels. Using our definitions of normal and low PCr depletion, eight CFS patients (45% of the study cohort) fell into the normal group and 10 (55%) fell into the low PCr depletion group.

Within the CFS population, therefore, there is a subgroup who, in the context of a formal assessment protocol, exhibit a normal maximum voluntary contraction and who subsequently show proportionate PCr depletion when exercised at 35% of that MVC. There is also, however, a second subgroup of CFS patients who, when undertaking the same assessment protocol, exhibit a significantly lowered level of maximum voluntary contraction and, subsequently, proportionately lower PCr depletion on exercise.

Perhaps particularly pertinent in light of this, and other possibilities:

The lack of a DI increase in ME/CFS suggests that they did not engage in the task enough to develop muscular fatigue. Alternatively, the lack of a DI increase could relate to muscles deficiencies as often reported in this disorder (Rutherford et al., 2016; Syed et al., 2025; Wang et al., 2023). But this is less likely since the ME/CFS and HV generated similar EMG amplitude as measured with RMS. Further, in our prior work, with eleven of the current fifteen ME/CFS participants, we did not find evidence of impaired muscles fiber composition (Walitt et al., 2024). It remains uncertain why DI at block Bn did not increase in ME/CFS while they generated similar force level as the HV. Thus, the muscle deficiencies in ME/CFS may not be the only or the most essential factor explaining their early onset of performance decline.

 

[Jonathan Edwards]

Couldn't disruption in the cascade of physiological processes that normally occur post-exercise account for delayed PEM? A disruption in the normal "recovery" processes?

The trouble is that so many things just don't fit, @Medfeb.

I think the mitochondrial story started with Ramsay saying that the characteristic feature of his proposed "ME" was rapid muscle fatguability. Behan then thought he found some mitchondrial shape changes. Interestingly, Behan called what we call ME/CFS CFS and thought it was centrally mediated but assumed that Ramsay had discovered something separate. And of course PEM emerged much later as a feature of 'CFS' which, with the PEM emphasised, is now ME/CFS.

A mitochondrial defect would fit with rapid fatiguability. If muscle is short of energy you can do things a few times and then come to a dead halt from weakness or pain. People with ME/CFS do complain of rapid fatiguability but that is a quite different issue from PEM, which Maree neatly summarised for AfME. Interestingly, a CPET ought to be a good way to pick up fatiguability but what is reported is a fairly normal output on day 1 and reduced maximal oxygen uptake on day 2, not a dead halt. It seems that at least the people who have had CPET studies did not have rapid fatiguability.

There is also the fact that Maureen's patients had very similar initial CPET results to sedentary controls. So it seems that muscle is OK most of the time in these people and if there is a problem it is only after exercise. But people with ME/CFS complain of difficulty doing things all the time.

A further peculiarity is that the Maureen's results for mean and women were opposite. One had more of some proteins and the other had less of other proteins. So these proteins would not provide a consistent explanation for PEM. It seems a stretch to suggest that PEM has a quite different basis in men and women.

And again, mitochondrial or other muscle problems induced by exercise would be local to that muscle and PEM is not. From what I hear you feel terrible irrespective of which bit was used. So we are not dealing with a failure of local repair. We might be dealing with an overenthusiastic 'general housekeeping' effort that affected all muscles but then that would have to be co-ordinated centrally and would have nothing to do with lack of energy availability in a particular muscle.

The more I look at this the more a am reminded of what Mike Murphy said at the IimE conference qhere we met. He was happy to be asked to talk on his research speciality of mitochondria but there was no way that mitochondrial insufficinecy would explain the clinical picture of ME/CFS. As Dan says, we have to start with the clinical problem we want to explain. In this case things are complicated by there being two clinical problems - fatiguability and PEM. But separating them out leaves us with no basis for thinking PEM has to do with a local inaccessibility of energy in a muscle.

It is such a pity that this continues to be part of the ME/CFS advocacy folklore. People criticise me for saying there is no evidence of a mitochondrial defect as if I was being unreasonable but I am just expressing the standard physiological position. Maureen's data are intriguing but I find it hard to see how you actually fit it together to produce a useful story. The exercise physiologists seem to have gone quiet on the CPET story too.

 

[DMissa]

The exercise physiologists seem to have gone quiet on the CPET story too.

I don’t know how to interpret the data so I withhold comment but Todd Davenport’s most recent CPET study (a few weeks or months ago) makes the assertion that the reported differences are most likely explainable by energy insufficiency. It would be good to get eyes on that study that are well enough equipped to scrutinise the data and the claim