Post-exertional malaise and the myth of cardiac deconditioning: rethinking the pathophysiology of long covid, 2026, Charlton, Wüst et al

[Evergreen]

The biggest problem for me is that it perpetuates the fallacy that if you implement rehab in a personalised way that tries not to provoke PEM, people (a) will improve and (b) won't deteriorate.

Eh, "properly implemented" GET did this too, and those who promoted it believed this too.

Don't fall for it, kids.

 

[EndME]

I have learned the following from patients, some of which are not with us anymore today: PACE showed that these rehab strategies did not work usefully for people with CFS/ME independently of whether they have PEM or not.

It's time the researchers that always consider themselves allies and are the ones that say one has to listen to patients do as they say.

 

[Evergreen]

I have learned the following from patients, some of which are not with us anymore today: PACE showed that these rehab strategies did not work usefully for people with CFS/ME independently of whether they have PEM or not.

It's time the researchers that always consider themselves allies and are the ones that say one has to listen to patients do as they say.

Hear hear.

 

[Evergreen]

Eh, "properly implemented" GET did this too, and those who promoted it believed this too.

In case people with long covid aren't as familiar with GET for ME/CFS, this is from the PACE GET participant manual, right after talking about the physio using a heart rate monitor with them:

There is nothing to stop your body from gaining strength and fitness, as long as it is done in a carefully monitored way, relating directly with your own particular circumstances – started and progressed at the right rate for you. Good luck!

 

[Utsikt]

Plus we have some preliminary data (only presented at a conference as of yet)

from Germany from a specialized Rehab course that was tailored to Long Covid and ME. Guess what? Many people were worse after than they were before. Almost no one improved

(44% left with a worse bell score, only 13% with an improved one)

Oh, that’s bad! Do you have the source?

The aim of pacing is to avoid increasing FUNCAP scores in any domain. Harm reduction, as someone said in another thread.

It’s to try to avoid PEM, not just the kind of PEM that would make you worse off long term.

PS. a lower FUNCAP score is worse, not better.

 

[Kitty]

It’s to try to avoid PEM, not just the kind of PEM that would make you worse off long term.

Yes, but ultimately it's to avoid making your score worse. PEM does exactly that, and the idea needs to be encapsulated in as few words as possible.

PS. a lower FUNCAP score is worse, not better.

I usually get maths concepts upside down, so I write whatever I think it is then wait to be put right.

 

[Utsikt]

Yes, but ultimately it's to avoid making your score worse. PEM does exactly that, and the idea needs to be encapsulated in as few words as possible.

FUNCAP asks for an average day the last month. One episode of PEM probably won’t change the score much.

But PEM should be avoided by itself, regardless of the aftermath of the episode. Otherwise, we’ll get the «you’ll get better in a few days so it’s nothing to worry about»-approach.

 

[Yann04]

I wonder if we should substitute baseline for severity level, as indicated by FUNCAP scores

I often feel the severity level to be way too homogenising and honestly I feel it always overestimates my capacity.

For example my baseline is that a tiny sound or speaking will give me terrible pem to the point i havent spoken or heard a word for years. But I don’t think that’s very obvious by “very severe”.

And perhaps its survivorship bias but I feel atleast physically and sensory I am more severe than the vast majority pw very severe me ive interacted with. I think if my baseline was conceptualised as “very severe ME” a lot would be missed.

 

[Kitty]

FUNCAP asks for an average day the last month. One episode of PEM probably won’t change the score much.

But if you're not already putting a lot of effort into avoiding PEM you're going to have very frequent episodes, and that will change the score. It'll only get better if you improve your pacing.

Thing is, PEM is a debased currency. Everyone and their auntie claims to have it now. By utilising the best instrument we currently have to explain the basics of pacing and thresholds, it might help avoid terms that can be hijacked or are difficult to understand. FUNCAP's a brass tacks description of functional impairment—can you keep yourself clean, can you feed yourself, can you leave the house, can you walk to the bus stop. It's not hard to understand that.

 

[Utsikt]

But if you're not already putting a lot of effort into avoiding PEM you're going to have very frequent episodes, and that will change the score. It'll only get better if you improve your pacing.

Will it though? Because if you’re pushing through symptoms you’re already breaking one of the underlying assumptions: that you respect your limits and rest when in PEM.

It doesn’t ask «can you do things afterwards», it asks «do you think you’d be able to do things afterwards». Someone doing LP would probably score higher, even if they don’t have an actual improvement.

Thing is, PEM is a debased currency. Everyone and their auntie claims to have it now. By utilising the best instrument we currently have to explain the basics of pacing and thresholds, it might help avoid terms that can be hijacked or are difficult to understand. FUNCAP's a brass tacks description of functional impairment—can you keep yourself clean, can you feed yourself, can you leave the house, can you walk to the bus stop. It's not hard to understand that.

I’m not trying to argue against the use of FUNCAP in general. I think it’s very useful. But I don’t think it can be used to determine if something caused PEM to someone. Or even if they have PEM - we don’t know the false positive rates.

If I remember correctly, FUNCAP was found to be accurate within a severity group. So a range of ~1 point on a scale from 0 to 6.

In my view, FUNCAP is best used to roughly track long term trajectories, and to self-report impaired functional capacity for benefit evaluations etc. in a more accurate way than most FC questionnaires.

 

[Sean]

Thresholds are a less problematic concept for me, as long as it is understood that they are dynamic and to some extent chaotic; not easy to spot, not a simple marker to avoid, more like a fuzzy space with a lot of inherent unpredictability; and not crossing a threshold doesn't mean you are having a good day or improving, just that you are doing what you can to not exacerbate it.

So planning around thresholds is difficult. There is always going to be a a lot of random real time variation to manage, and the capacity to do so effectively is always going to be limited.

I think the key to understanding a threshold for ME/CFS is that the effort-response relationship is non-linear, dynamic, and difficult to predict. The broad management aim is to avoid entering the high growth rate of the response curve (the most non-linear section). But doing so with a moving target is only ever going to be a partial success.

Baselines, on the other hand, are basically meaningless rehab babble, and based too much on the notion of 'energy deficits' and 'energy management', which is a particularly pernicious consequence of over emphasis on fatigue (itself a poorly defined concept).

The idea of a ceiling works a better for me, though it has the same issues and limitations in applying it as does a threshold.

While prolonged rest may initially seem beneficial for symptom management, extended periods of daytime lying or bed rest can induce cardiovascular and skeletal muscle deconditioning.

Again, where is the evidence for this in ME/CFS patients? It cannot be assumed, particularly for something that is clearly not responding to conventional ideas of rehabilitation (physical and psychological), and in fact carries a high level of serious risk with it.

This is not a titration or management problem, it cannot be solved by more fine tuning and individualising of 'rehab' therapies. There is no robust evidence that this approach works, despite decades of formal and informal testing.

This stuff is a waste of patients' time and goodwill, and precious clinical and research funds.

I cannot say it enough: I think patients are not deconditioned, or as deconditioned, as the standard deconditioning theory being invoked by rehab proponents would predict, and if that is correct then this apparently paradoxical phenomena is a very critical clue.

 

[Evergreen]

I cannot say it enough: I think patients are not deconditioned, or as deconditioned, as the standard deconditioning theory being invoked by rehab proponents would predict, and if that is correct then this apparently paradoxical phenomena is a very critical clue.

That's an interesting hypothesis. How would you show that we're less deconditioned than expected, I wonder?

I think any deconditioning is simply neither here nor there because

• it's not what makes us unable to maintain normal levels of activity in the first place
• for people who are not already improving naturally, reconditioning of any deconditioning without messing up overall function is not possible

Or to put it another way, it's inevitable and it's irrelevant.

Stop trying to rev our broken engines. Figure out what's going on with the bloody engine.

I just hope that Appelman, Wüst & co follow these pwLC being rehabbed and actually start figuring it out. Other publications of theirs that I've read are much more promising. They don't overreach.

 

[EndME]

Other publications of theirs that I've read are much more promising. They don't overreach.

Yes, it's a bit unusual. According to themselves they don't upload preprints because they want to make sure everything is solid when it gets published and yet we see this work where citations don't match claimed statements and statements aren't based on evidence etc. which I don't remember from any other work of theirs.

 

[EndME]

Regarding the imaginary 80% PEM rate in LC they quote, in their previous work they said the rate was 90% based on horrid and completely insufficient analysis (see here). I know that patients in the Netherlands cooperate with this team. I think it is useful if patients point out these issues to them.

 

[Concentric_Spiral]

Regarding the imaginary 80% PEM rate in LC they quote, in their previous work they said the rate was 90% based on horrid and completely insufficient analysis (see here). I know that patients in the Netherlands cooperate with this team. I think it is useful if patients point out these issues to them.

I read somewhere that roughly 50% of LC has the same (or very similar?) symptom profile as ME/CFS. Is that approximately accurate?

 

[InitialConditions]

I read somewhere that roughly 50% of LC has the same (or very similar?) symptom profile as ME/CFS. Is that approximately accurate?

That is more in line with most estimates, but the range is huge.

 

[Concentric_Spiral]

That is more in line with most estimates, but the range is huge.

Thanks.

 

[Mfairma]

@ME/CFS Science Blog piece on deconditioning is a useful read for anyone who hasn’t seen it: https://mecfsscience.org/what-does-deconditioning-look-like/

 

[EndME]

I read somewhere that roughly 50% of LC has the same (or very similar?) symptom profile as ME/CFS. Is that approximately accurate?

Long-Covid is a rather vaguely defined concept that (depending on criteria) usually requires a symptom presence of at least 3 months, whilst clinical criteria for ME/CFS require a symptom presence of longer duration (in the CCC that is 6 months for adults). These durations are all rather arbitrary and not based on evidence, however their role is to serve as differentiation between more benign and self resolving presentations to those that last longer and are more severe. But in practice when reading papers that means is that there is not really a straightfoward meaningful concept of accurate percentages. If you have a cohort of people that all have had LC for 3 months but not longer then nobody by the Canadian Consensus Criteria would have ME/CFS at that time (even if many end up with ME/CFS some of which will then also recover whilst the majority does not), if you'd have a cohort of people that were hospitalised with severe acute Covid the percentage of those with other problems than ME/CFS, for instance organ damage, will be larger than in say a predominatley younger and female cohort with mild acute infections that developed Long-Covid. If you look at cohorts that have been ill for longer there's a good chance (that depending on your cohort itself) you'll find more cases of ME/CFS because ME/CFS tends not be self-resolvent, whilst for example post-covid anosmia often tends to be so, even if not always. So these percentages tend to always be contextual and then you haven't even touched the whole problem of LC being too generalistic that it's estimates itself tend to be useless (you'll find studies reporting LC in 60% of the general population and others reporting it in 1% of the general population, which just tells you how little those estimates actually mean).

As @InitialConditions reports the range we've seen is huge, so I think the honest solution is to just describe the range with the known very severe limitiations I described above (authors in papers often like to avoid this because if you read a study with such a large range anybody sees that things are not very meaningful, so they often prefer to put an often meaningless percentage next to something to appear accurate) or to just say something along the lines of "a very substantial proportion of people with Long-Covid that are severly ill and which show no signs of recovery have symptoms that are consistent with ME/CFS".

The perhaps larger problem with Wüst et al suggestion here is that he seems to suggest that people with LC+PEM need a tailored rehabilitation program and possibly different rehabilition trials from other LC folks because of PEM and that trial results from rehabilition strategies in just LC folks don't generalise to LC+PEM folks. However, if the ratio was really as high as they state (80% or 90%) then the generalisability would be rather unproblematic. If as they say, almost everyone has PEM then stratification is not a big problem to begin with! In reality the problem is something else entirely. The problem is much rather PACE and similar trials showed that there are no useful benefits for GET and CBT independently of whether the people have PEM or not, not the opposite and that the trial design is completely inadequate. It doesn't matter how you choose your cohort if your intervention does not work and your trial is set up in way that outcome measures cannot possibly show anything useful to begin with. The authors suggestion seems to rather be the opposite. Admittedly things are slightly different when the focus is on deconditioning and reconditioning rather than illness and recovery, but if these are the discussions that are still being had then progress might be quite hard.

I can definitely understand why one would want to write a piece on why paying attention to PEM in LC trials and LC patients is important, but you cannot just ignore the other issues, which unfortunately seems to be a main message here.

 

[Mij]

I would love to communicate with Wurst and share my experiences during the earlier years of PVFS/ME, and that I didn't have PEM or delayed PEM for 7-8 years, and that doing ART for almost 6 years didn't improve or cause deconditioning in my cause. I started exercising again years later when I felt better, I had no cardiac issues/deconditioning and exercised like I was never sick, but this was when delayed PEM started and the delayed PEM symptoms gradually worsened over the years without my realising until it was too late.

The scale score really doesn't apply to me because I don't have many symptoms, even today. I have figured out on my own over the years how much activities I can do w/o getting PEM or delayed PEM.