Preprint ME/CFS and Long COVID Demonstrate Similar Bioenergetic Impairment and Recovery Failure on Two-Day Cardiopulmonary Exercise Testing, 2026, Davenport+

[SNT Gatchaman]

ME/CFS and Long COVID Demonstrate Similar Bioenergetic Impairment and Recovery Failure on Two-Day Cardiopulmonary Exercise Testing
Todd Davenport; Staci Stevens; Jared Stevens; Mark Van Ness


Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) and Long Covid are characterized by post-exertional malaise (PEM). Similarities in disease presentation suggest important commonalities in bioenergetic impairment, but this hypothesis has not been demonstrated. The metabolic underpinnings of each disease can be elucidated by two cardiopulmonary exercise tests (CPET) administered 24 hours apart.

This retrospective study examined physiological responses on two-day CPET in people with ME/CFS (63 females and 21 males), Long Covid (52 females and 27 males), and matched non-disabled control participants (51 females and 20 males). Data were analyzed within sexes using repeated measures analysis of variance.

All participants met maximal effort criteria. There were significant reductions in oxygen consumption (O₂) and workload at the ventilatory anaerobic threshold (VAT) in both patient groups compared to non-disabled controls, with larger effect sizes at VAT than at peak exertion. Performance decrements were observed in both sexes. Females exhibited more pronounced abnormalities and significant group by test effects.

No significant differences were observed between patient groups. Severe disability based on impaired O₂ was prevalent in both patient groups. Hemodynamic and ventilatory measures were within normal ranges. ME/CFS and Long Covid both involve a functionally significant bioenergetic failure complicated by inadequate post-exertional recovery, which is similar between the conditions and unexplained by hemodynamic and ventilatory changes.

Findings support the utility of two-day CPET as an objective measure of PEM and functional impairment. Future studies may integrate mechanistic biomarkers with two-day CPET as trial endpoints and to establish likely responses to treatments for PEM.

Web | DOI | PDF | Preprint: Research Square | Open Access

 

[Utsikt]

Consistent with prior literature in ME/CFS [29, 30] and Long Covid [23], the most pronounced differences between patient groups and controls were at VAT rather than maximal exertion. In both males and females, O2 and workload were significantly different between patients and non-disabled controls, and with larger effect sizes at VAT than at maximal exertion. However, patient groups did not significantly differ from each other. O2 and workload at VAT among females in both patient groups demonstrated a significant group by test interaction.These findings indicate abnormal aerobic metabolism in ME/CFS and Long COVID that is similarly expressed on CPET.

Why does these findings indicate abnormal aerobic metabolism?

VO2 measures how much oxygen your body takes up. If I’ve understood things correctly, that can be affected by uptake (how much the lungs take up), transportation (heart and blood) and utilisation (muscles). Isn’t utilisation the only one that would classify as «metabolism» in this case? What if the abnormalities lie elsewhere along the chain?

Findings from this study also suggest that post-exertional bioenergetic abnormalities may be more prevalent in females than males with ME/CFS and Long COVID. Nevertheless, CPET measures in this study support that severe disability is common among people living with ME/CFS and Long COVID, further corroborating previous reports of poor health related quality of life in both conditions with objective physiological evidence of functionally meaningful impairments in metabolic functioning.

How does these results support that pwME/CFS or pwLC are severely disabled?

Findings from this study strengthen earlier observations of potential mechanistic overlaps between ME/CFS and Long COVID [31, 32], suggesting similar disease mechanisms and presentations from a novel pathogenic cause.

That depends a lot on how LC is defined.

An inability to recover normally after exertion underlies signs, symptoms, and disability associated with PEM.

This is not really the issue with PEM. The issue is that you get a delayed worsening of symptoms that lasts a long time, and the symptoms also affect systems that were not involved in the exertion.

Future exercise studies may use objective measures such as two-day CPET to ensure only likely responder groups are included by differentiating between deconditioning and PEM.

Why encourage the use of CPETs for diagnosing PEM? What’s the ethical rationale behind using a screening tool that will harm the only people that will not be able to benefit in any way from the intervention that is tested?

And even if that issue is ignored, have the authors confirmed that a CPET does not produce false negatives for PEM and leads to inclusion of people that will be harmed from the intervention? What’s the sensitivity or specificity?

CPET results may then be used to assist in establishing adequate pacing thresholds for likely non-responders, as well as the most appropriate exercise thresholds and stimuli for likely responders.

Om what grounds? Why would a threshold derived from a CPET be «safe» for people with PEM?

A wider adoption of CPET as an objective screening and endpoint measure also would help overcome many of the methodological limitations in current Long COVID exercise studies of selection bias, inaccurate PEM screening and risk stratification, poor precision in rehabilitation planning, and using subjective health-related quality of life measures [44].

How would using a CPET as an endpoint be relevant for people with PEM? What would you measure to assess changes? It feels about as useless as using grip strength. Sure, if you exercise your grip strength probably improves, but how is that in any way relevant for your functional capacity or quality of life?

 

[rvallee]

Future exercise studies may use objective measures such as two-day CPET to ensure only likely responder groups are included by differentiating between deconditioning and PEM.

I'm not sure that really applies here, especially for LC. Mostly because the concept of deconditioning is too arbitrary, mostly intentionally because it provides a cheap excuse to dismiss patients. We have seen plenty of exercise rehabilitation trials with patients who have vastly different symptoms and for the most part those have very little impacts, for sure all over the place. Deconditioning is not a significant factor for the most part here.

At the very least there is no such thing as a simple way of differentiating between PEM and deconditioning. I'm pretty sure CPET is valuable to study physiological changes in people experiencing PEM, but I don't think that way is. Lots and lots of studies of people without PEM who aren't deconditioned have shown no significant benefits. Even in mild chronic fatigue it doesn't make much difference. There is way too much nuance to work with here for this to be operationalizable.

 

[forestglip]

How would using a CPET as an endpoint be relevant for people with PEM? What would you measure to assess changes? It feels about as useless as using grip strength. Sure, if you exercise your grip strength probably improves, but how is that in any way relevant for your functional capacity or quality of life?

I'm not arguing for adopting CPET as an endpoint, but I'm not sure I understand your specific point here. Presumably, a decrease in metrics on day 2 is not analogous to weak grip strength, since the idea is that this is not something seen in deconditioning, so exercise alone should not improve it by default. I guess they think it's something directly related to the phenomenon of PEM, so if it improves, that may be a sign the PEM is improving.

So I suppose the idea would be to do two CPETs at the start and end of a study, and the endpoint is the difference between day 1 and 2 each time.

Though the overlap between groups in the Keller study makes me think the data would be too messy to be a good endpoint. Lots of healthy controls were worse than people with ME/CFS. I haven't read this paper yet, though, so maybe it's better.

 

[shak8]

The Workwell Foundation (where some or all of the authors work or have worked) is a clinic near me. It does CPET testing for documenting disability.

"Not all people with ME/CFS and Long COVID may participate in CPET due to severity, availability, and choice. Less strenuous forms of testing must be developed for clinical and research purposes."

(my bolding)

 

[Sean]

@Snow Leopard

 

[DMissa]

What other explanations might there be for these CPET differences than the proposed "bioenergetic" ones? Cardiovascular delivery, pulmonary gas exchange? I don't know enough about CPET to interpret without spending a lot of time learning about it.

I can't find whether the controls were sedentary, is this mentioned? If we can't figure it out I may message Todd


Edit:

VO2 measures how much oxygen your body takes up. If I’ve understood things correctly, that can be affected by uptake (how much the lungs take up), transportation (heart and blood) and utilisation (muscles). Isn’t utilisation the only one that would classify as «metabolism» in this case? What if the abnormalities lie elsewhere along the chain?

This was what I was getting at with worse phrasing. Is it a jump?

 

[forestglip]

I can't find whether the controls were sedentary, is this mentioned?

It looks like they used data for controls from a previous study:

This study involved a retrospective review of records from an exercise testing facility and a previous study of CPET responses in non-disabled people compared to people with ME/CFS [25]

25. Keller B, Receno CN, Franconi CJ, Harenberg S, Stevens J, Mao X, Stevens SR, Moore G, Levine S, Chia J, Shungu D, Hanson MR. Cardiopulmonary and metabolic responses during a 2-day CPET in myalgic encephalomyelitis/chronic fatigue syndrome: translating reduced oxygen consumption to impairment status to treatment considerations. J Transl Med. 2024;22:627. Epub 20240705.

S4ME

From that paper:

CTL who also endorsed a low/no recreational/vocational physical activity level

 

[Chandelier]

From the main author on X:

Todd Davenport​

@sunsopeningband
21h • 10 tweets • 3 min read • Read on X

Our latest work now posted as a pre-print on @researchsquare: "ME/CFS and Long COVID Demonstrate Similar Bioenergetic Impairment and Recovery Failure on Two-Day Cardiopulmonary Exercise Testing" @4Workwell

ME/CFS and Long COVID Demonstrate Similar Bioenergetic Impairment and Recovery Failure on Two-Day Cardiopulmonary Exercise Testing
Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) and Long Covid are characterized by post-exertional malaise (PEM). Similarities in disease presentation suggest important commonalities in b…https://www.researchsquare.com/article/rs-8606329/latest
Using two‑day cardiopulmonary exercise testing (CPET), we compared physiological responses in people living with #MECFS and #LongCovid with matched controls, stratifying by sex, to resolve the weaknesses of single tests in mixed-sex cohorts we've seen so far in CPET studies.
The primary studies comparing #MECFS and controls already have been published. You may recognize them from work often discussed here.

The data is available from mapmecfs.org. Hansen et al 2024: pubmed.ncbi.nlm.nih.gov/38232699/ Keller et al 2024: pubmed.ncbi.nlm.nih.gov/38965566/

https://pubmed.ncbi.nlm.nih.gov/38965566/

Here's what we learned:
Both ME/CFS and Long COVID show similar failure to recover after exertion
Largest impairments occur at the ventilatory anaerobic threshold, suggesting similar aerobic system impairments between ME/CFS and Long COVID
...
...
Females exhibited more pronounced post‑exertional CPET abnormalities than males, but were present in both sexes
No meaningful differences were observed between ME/CFS and Long COVID on two-day CPET, suggesting they share a common bioenergetic pathophysiology
...
...
Severe functional impairment was common in both patient groups and was not explained by effort preference, deconditioning, ventilation, or hemodynamics

Maybe the most remarkable thing? We didn't purposefully select for a cohort with Post COVID ME. It was dealer's choice.
Normally you might expect noisy case definitions to wash out any effects, because as we know, Long COVID isn't just one thing. But, at least in our sample, it didn't seem to matter. CPET responses were statistically and functionally similar between ME/CFS and Long COVID anyway.
These results reinforce post‑exertional malaise (PEM) is a measurable physiological phenomenon. In an era where the 'lazy-crazy dichotomy' still gets pushed as explanations for ME/CFS and Long Covid, unfortunately this doesn't get to be a throwaway line yet.

It still matters.
The findings of this study also underscore the value of two‑day CPET as a useful tool to:
Differentiate PEM from normal recovery using objective measures
Inform disability evaluations in people who can and choose to have a CPET
Improve precision in clinical trials
Anyway, thanks for reading. I'm grateful to my co‑authors and to the patients. Their generous engagement continues to advance our understanding of these terribly disabling yet largely ignored chronic conditions.

Mechanistic work is slow but worth it. More forthcoming. Onward.

Thread by @sunsopeningband on Thread Reader App

@sunsopeningband: Our latest work now posted as a pre-print on @researchsquare: "ME/CFS and Long COVID Demonstrate Similar Bioenergetic Impairment and Recovery Failure on Two-Day Cardiopulmonary Exercise Testing" @4...…

threadreaderapp.com

 

[Utsikt]

I'm not arguing for adopting CPET as an endpoint, but I'm not sure I understand your specific point here. Presumably, a decrease in metrics on day 2 is not analogous to weak grip strength, since the idea is that this is not something seen in deconditioning, so exercise alone should not improve it by default. I guess they think it's something directly related to the phenomenon of PEM, so if it improves, that may be a sign the PEM is improving.

But they have not proved that it’s directly related to PEM, they didn’t even show anything about symptom patterns. So they are either advocating for using an unproven and speculative endpoint, or for using it as a proxy for what’s essentially «overall health and functional capacity», just like what the handgrip strength is claimed to be.