Trial Report Resistance Exercise Therapy for Long COVID: a Randomized, Controlled Trial 2025 Berry et al.

[Chandelier]

Harriet Carroll got her comment published:

Harriet Carroll: Long Covid Scientific Consultancy​

@angryhacademic.bsky.social

Delighted to say my comment got published! There was a lot of critiques I considered about this study, but the problem is that if we critique an individual study, future studies can just avoid that particular critique. So I tried to take a different approach

In this case, it is clear in Berry's study someone mentioned PEM, so they "addressed" it (badly, but they can still say they provided info and support).
They addressed a critique we've been repeating for years. In their mind: Job done

So instead of highlighting the problems with this particular study, I wanted to highlight that doing these kind of studies is BY DEFAULT unethical. Full stop.
There is no doing this the "right" way.
This study in isolation is not the problem per se, but rather the entire academic system that repeatedly allows these studies to go ahead is:

Non-experts deciding they can and should run studies like this

Funders not having appropriate expertise, & funding these studies

Ethics boards being duped by academics who can justify any idea they want to do

Patients having inadequate support and education from their medical teams to understand why they should not participate in these kinds of studies


Doing a study against current and established guidelines should not happen bar in very exceptional, highly-specialised, circumstances, led by a legitimate team of experts. And I think this is something we really need to hammer home
Anyway, I hope my point comes across clear and is a useful addition for this and future study critiques

Spoiler: Her comment in full

November 18, 2025
Ethics of doing research against established clinical guidelines
Harriet Carroll, PhD | Lund University
I read with interest the study by Berry et al., investigating the effect of resistance exercise in patients with post-COVID-19 sequelae. However, this research raises wider ethical issues that are pervasive in long COVID research.

Post-COVID-19 sequelae encompasses several, potentially co-occurring, subtypes: deconditioning particularly in the context of hospitalisation; direct organ damage; new-onset diagnosable disease(s); and/or new-onset currently unknown disease(s)/pathology(ies). Individual symptoms may overlap across groups but have different underlying pathophysiology; for example, fatigue could be caused by autoimmunity, hypercoagulability, lung damage, nervous system damage, etc. Conducting trials—even for rehabilitation—as if these were a homogenous group with unified needs is somewhat akin to grouping multiple sclerosis, Parkinson’s, and dysautonomia patients together because they all suffer with fatigue, sensorimotor impairments, and cognitive dysfunction, and is therefore unhelpful.

Berry et al., were aware of risks associated with exercise in some long COVID patients via discussion of post-exertional malaise (PEM) in both the participant information sheet (PIS) and manuscript. If assessed properly, PEM is the cardinal symptom of myalgic encephalomyelitis (ME),[1] and in worst cases can cause permanent worsening. Accordingly, exercise—including under the premise of deconditioning—is contraindicated in ME, and NICE guidelines state any exercise programme should only be done under a ME specialist team.[2] There are no ME specialists in NHS Scotland nor the study team, as confirmed by personal correspondence with NHS Greater Glasgow & Clyde, e.g. stating pulmonary rehabilitation physiotherapists were consulted. None of these points were explained in the PIS, nor were risks pertaining to other known pathophysiology commonly found in long COVID.[3] Therefore, by proxy, this trial contravened clinical guidelines which raises serious ethical issues—a problem that is unfortunately not unique to this study.

Further, studies focused on rehabilitation without treating ongoing underlying pathology are generally unhelpful, just as it would be unethical to rehabilitate stroke patients without first treating the stroke. Post-hospitalisation rehabilitation in recovered but deconditioned patients is well-established, again raising ethical questions regarding the purpose of studies like that conducted by Berry et al.

Regardless of study findings, it does not seem ethical to go against established clinical guidelines, particularly without specialist involvement/oversight. This also highlights problems with how studies are ethically reviewed as disease specialists are not necessarily present. Considering the overall lack of funding for diseases like ME and long COVID, it is imperative any grants are used for ethically sound, pathology-targeted, and clinically useful research.

References
1. Davenport TE, Chu L, Stevens SR, Stevens J, Snell CR, Van Ness JM. Two symptoms can accurately identify post-exertional malaise in myalgic encephalomyelitis/chronic fatigue syndrome. Work. 2023;74(5):1199-1213. doi:10.3233/WOR-220554.
2. National Institute for Health and Care Excellence. Myalgic encephalomyelitis (or encephalopathy)/chronic fatigue syndrome: diagnosis and management. https://www.nice.org.uk/guidance/ng206. Accessed November 16, 2025.
3. Carroll HA, Harvey N. Letter to the Editor RE: Exercise is the Most Important Medicine for COVID-19. Curr Sports Med Rep. 2023;22(12);423. doi:10.1249/JSR.0000000000001123.

CONFLICT OF INTEREST: HAC runs the Long Covid Scientific Consultancy, providing education, advocacy, and resources on long COVID and related diseases.

 

[Karen Kirke]

What do they expect people to do who are already operating at their PEM threshold just to get through the necessary activities to stay fed and clean? I had a look through the exercises for ambulatory, since I can walk in my home, though I spend most of my time lying down and am only able, for example, to shower about once a week and do little or nothing else that day.

I don't know for sure, @Trish, but I would guess that they wouldn't assign you to the ambulatory group. My guess is that they would start you in bed-bound. The manual says:

initial training session which will be carried out with a therapist/nurse. This may be face-to-face or by video call. During this session we will support you to find the best level at which to start your exercises and demonstrate these to you.

Does it say how many repetitions they asked people to do for each exercise?

That depended on how hard people found each exercise. If I'm understanding correctly, for the first 2 weeks, the participants did each upper body exercise until they felt they were at a 3-5 on a scale of 0-10 where 0=extremely easy and 10=extremely hard. 3=somewhat easy. 6=somewhat hard. So 5 is neither easy nor hard.

They stopped at that point and rested for a minute or two. Then they did the same on the next upper body exercise.

Then in week 3 they added the lower body exercises and did them as above.

Then they gradually worked up to this:

For the remainder of the programme we will ask you to try and perform one set of each exercise every day, stopping each exercise when you rate how hard you are working as 8-10 on the effort scale below. Take a break of a minute or two between each exercise.

8=hard. 10=extremely hard. Yikes.

They mention number of repetitions here:

Ideally during the programme we want to reach the required effort level after around 10-15 repetitions of each exercise. If you find you are doing more, or less, repetitions to reach this effort level then you can move up or down a level.

This seems totally impractical for most of us. As others have said, why do an arm raising exercise when you already raise your arms to put things in cupboards or wash your hair. And so on. If I tried to add these to my day I would crash.

Yeah, it's well out of my reach too. I have a few little exercises I do before rests on good days and never get beyond 2-3 reps.

Here the rationale was to try to improve functional capacity, but I don't think we have evidence of this in ME/CFS. There was an arm a bit like this in Jason et al.'s 2007 trial - "anaerobic activity", designed by Staci Stevens. Stages 1 and 2 of Workwell's "analeptic exercise" are a bit like this too. It would be interesting to see whether there would be any long-term benefit to exercises like this at lower intensity in terms of prevention of deconditioning over years/a decade. And I can't remember who it was who I read insisting on the importance of doing strengthening exercises like this during the early years of long COVID, but I remember was surprised at who it was. (I know who I think it was but am not sure so won't say!)

Even if I had capacity by cutting back other things in order to do the exercises, I might get a bit stronger, but would still be just a sick.

And there's the rub.

 

[Eleanor]

Eyes on eTable 28 everybody:
View attachment 29333

So the intervention group had about a 50% greater likelihood of having an 'indication of ME/CFS diagnosis' than the standard care group after 3 months?

 

[rvallee]

Apologies for not having read all the details. I always come back on anything that is recommending people do any type of exercise to my basic question. If I had the capacity to do these exercises why wouldn’t I use that on doing something I have on my long list of things I need or want to do that actually have a practical purpose or social benefit.

For context as someone with moderate severity ME my daily living activities I put away all my own groceries, I do all my own clothes washing, I empty and fill my own dishwasher and put everything back in the cupboards. I prepare my own food. When I can I have a fire in my log burner. All these things already require me to move and lift stuff. If I do extra stuff I have to cut back on these daily living activities.

I have literally never once seen any of those so-called experts address that. Anywhere. At all. Not just this study or those experts but in the entire literature. They simply do not understand the nature and experience of the problem.

 

[Karen Kirke]

Do you think any of these claims hold up under scrutiny?

I think it's reasonable to check whether muscle strength training improves functional capacity in any well-defined disease, the problem here being the mish-mash of patients who would be scooped into this study.

From its description, I don't see a reason why the Shuttle Walk Test would be less valid outside of respiratory disease.

I don't have a problem with the rationale for the study. The study design, however, while including various upgrades from ME/CFS studies (objective primary outcome measure, less brainwashing etc) still has many of the usual issues (lack of blinding, no attention-matched control etc) and the potential dealbreaker of the variety of pathologies that could respond differently to the treatment. I would have liked to see results reported separately for those with ME/CFS, but I appreciate that identifying them and separating them out would not be straightforward.

 

[Utsikt]

I think it's reasonable to check whether muscle strength training improves functional capacity in any well-defined disease,

Why? What’s the rationale?

And they didn’t even test for functional capacity directly..

the problem here being the mish-mash of patients who would be scooped into this study.

I agree that that’s one of the problems.
I’ve posted the baseline ISWT results earlier in the thread. It had a huge spread. Some of the participants had scores above the average for healthy individuals in their age bracket. Some were able to walk 10-40 meters. I can do better than that, and I’m bedbound.

The group is far too heterogenous. It’s not just an issue of some having ME/CFS.

From its description, I don't see a reason why the Shuttle Walk Test would be less valid outside of respiratory disease.

Why? It has not been established that the ability to walk is an appropriate proxy for relevant functional capacity in other diseases.

 

[Karen Kirke]

Why? What’s the rationale?

And they didn’t even test for functional capacity directly..

I agree that that’s one of the problems.
I’ve posted the baseline ISWT results earlier in the thread. It had a huge spread. Some of the participants had scores above the average for healthy individuals in their age bracket. Some were able to walk 10-40 meters. I can do better than that, and I’m bedbound.

The group is far too heterogenous. It’s not just an issue of some having ME/CFS.

Why? It has not been established that the ability to walk is an appropriate proxy for relevant functional capacity in other diseases.

I don't have the energy for more, sorry @Utsikt!

 

[Utsikt]

I don't have the energy for more, sorry @Utsikt!

No worries, no need for an apology. Thanks for letting me know and for respecting your limits.

 

[rvallee]

Why? What’s the rationale?

Why? It has not been established that the ability to walk is an appropriate proxy for relevant functional capacity in other diseases.

And that rationale is pretty much the most over-tested hypothesis in the medical profession. We are talking about literally hundreds of similar trials and studies on the same rationale. It has been fully tested and debunked. As a rationale it might have made sense for a few months many decades ago, but at this point it's exactly equivalent to someone insisting that climate change could still be caused by solar activity cycles.

There will never be a point at which this rationale is considered shut by medical institutions. And I positively hate this fake aw shucks naivete as if this is still some brand new idea no one ever thought of before. It's maddening the parallels with politics, how giant clusterfucks keep getting endlessly excused with the equivalent of "it's their first day and they actually believe this and also good luck stopping them anyway".

It has no place in a serious profession. It's about the most reliable mark of systemic failure to see any system get stuck in such a loop.

 

[dave30th]

Authors responded:


They didn't say anything about the commenter's main point - that the pre-selected minimum important difference was not achieved.

exactly. they ignored it completely and just said essentially, well, we found some things that seem positive.

 

[dave30th]

So the intervention group had about a 50% greater likelihood of having an 'indication of ME/CFS diagnosis' than the standard care group after 3 months

it's confusing, but that seems to be only among those found to have PEM. So in the intervention group, more of those with PEM at 3 months also have an ME/CFS diagnosis, while those with PEM in the usual care group are more likely not to have an ME/CFS diagnosis. Not sure what those data would mean or indicate, if accurate.

 

[forestglip]

it's confusing, but that seems to be only among those found to have PEM. So in the intervention group, more of those with PEM at 3 months also have an ME/CFS diagnosis, while those with PEM in the usual care group are more likely not to have an ME/CFS diagnosis.

My attempt to parse this:

Of 99 individuals who completed the DePaul Symptom Questionnaire, 40 of 48 (83.3%) in the intervention group and 42 of 51 (82.4%) in the control group experienced postexertional malaise at 3-month follow-up (eTables 25-28 in Supplement 2).

The text I quoted indicates that the middle part of the table (Indication of post-exertional malaise) shows how many people in each group had PEM at 3 months.

This suggests to me that the part below it (Indication of ME and/or CFS diagnosis) is similar, but instead shows how many people qualify for an ME/CFS diagnosis at 3 months.

It doesn't seem to me like this relates to baseline diagnosis. Just that they basically asked everyone (or the 99 participants they could get answers from) at 3 months if they have PEM and if they have ME/CFS.

The quote is from the safety section, so I think they're trying to show that since the proportion who have PEM isn't higher in the active group at 3 months, there's no danger of causing more PEM. Further evidence of this intention in the supplementary table 2 eDiscussion:

the intervention did not increase post-exercise malaise.

But the thing is, they didn't measure PEM severity in people who have PEM. They measured proportion who have PEM out of all participants. We don't expect that people without PEM will get it from exercise, so it is not surprising that the proportion who have PEM did not increase after exercise.

However, if I'm understanding the bottom part correctly, a third of the control group (29.4%) and half of the exercise group (47.9%) could qualify for an ME/CFS diagnosis at 3 months. So substantially more participants ended up with ME/CFS in the exercise group. Of course, suggesting that exercise caused more ME/CFS than usual care based on this depends on proportions of participants with ME/CFS at baseline being similar between groups, which may or may not be accurate.

It doesn't help that there's no baseline data on ME/CFS or PEM that I can see. I can't even find any mentions of ME/CFS or PEM in the paper or supplementary files apart from the 3 month tables and the references to these tables. No description of how they diagnosed "ME and/or CFS". Just the following, indicating that they seem to have decided to consider PEM at some point after the original protocol was made:

Post-exercise adverse events and malaise: adverse events during and after exercise were assessed in all participants. Following a protocol amendment, the DePaul Symptom Questionnaire (Short Form)[32] was assessed in a subgroup.

 

[dave30th]

I focused on the MCID issue in this post: https://virology.ws/2025/11/21/tria...trial-with-clinically-insignificant-findings/