Exercise Pathophysiology in ME/CFS and Long COVID: Commonalities Detected by Invasive Cardiopulmonary Exercise Testing, 2025, Squires, Systrom et al

https://www.atsjournals.org/doi/abs/10.1164/ajrccm.2025.211.Abstracts.A7881

Exercise Pathophysiology in Myalgic Encephalomyelitis/Chronic Fatigue Syndrome and Long COVID: Commonalities Detected by Invasive Cardiopulmonary Exercise Testing


J. Squires 1, S. Palwayi 1, P. Li 3, W. Xiao 3, K. LeWine 1, S.W. Johnson 1, D. Felsenstein 2, A.B. Waxman 1, D.M. Systrom 1,
Abstract

• PDF

Abstract

Rationale

There is substantial overlap of exertional symptoms in Long COVID (LC) and Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) including intractable fatigue, post-exertional malaise (PEM), and orthostatic intolerance, but very little objective data liking the two. This study compares exercise pathophysiology in the two disorders and normal controls using invasive cardiopulmonary exercise testing (iCPET).

Methods

Between January 2019 and December 2024, 1,518 patients underwent a clinical iCPET at Brigham and Women's Hospital. Exclusion criteria included morbid obesity (BMI>40 kg/m2), severe anemia ([Hb]<9.0 g/dL), elite athletes (peak VO2 (pVO2)>120% predicted), sub-maximum effort (RER<1.05), a primary pulmonary mechanical limit (VE @ AT/MVV>0.7), and comorbidities such as active/treated cancer, interstitial lung disease, or other respiratory related diseases. iCPET results from 438 ME/CFS patients, 73 LC patients, and 43 symptomatic but otherwise normal controls were analyzed. pV02, peak cardiac output (pQc), peak right atrial pressure (pRAP), peak systemic oxygen extraction (pSOE; Ca-vO2/[Hb]), and ventilatory inefficiency (VE/VCO2 slope) were compared among groups. Statistical significance was determined using Kruskal-Wallis tests for global comparisons, with post-hoc Dunn tests for pairwise group comparisons. Holm-Bonferroni adjustments were applied to control for multiple comparisons.

Results

LC and ME/CFS displayed reduced pVO2 % predicted compared to controls (LC: 78.4 ± 18%, ME/CFS: 78.1 ± 17%, Controls: 97.5 ± 10%, P≤0.0001). Reduced pQc % predicted was also observed compared to controls (LC: 91.1 ± 18%, ME/CFS: 96.3%, Controls: 101 ± 11%, P≤0.001). pRAP were significantly less compared to controls (LC: 1.1 ± 3.1 mmHg, ME/CFS: 1.3 ± 2.8 mmHg, Controls: 3.6 ± 3.4 mmHg, P≤0.001). Significant reductions in pSOE were seen for LC and ME/CFS (LC: 0.81 ± 0.1, ME/CFS: 0.81 ± 0.1, Controls, 0.91 ± 0.1, P≤0.0001). The only measure with no significant difference between disease and control was VE/VCO2 slope (LC: 31.4 ± 8.4, ME/CFS: 31.6 ± 6.9, Controls: 32.0 ± 6.7, P≥0.261). Most interestingly, no significant differences were seen between the two diseases for any of the analyzed measures (P≥0.245).

Conclusions

We report the largest cohort of ME/CFS and LC investigated with iCPET to date. ME/CFS and LC share symptomatic, reduced aerobic capacity at peak exercise, which is driven by preload insufficiency and impaired systemic O2 extraction, the latter compatible with peripheral left-to-right shunting and/or limb skeletal muscle dysfunction. These findings should drive future diagnostics and personalized medicine in both diseases. We hope these data inform the pending prospective NIH RECOVER iCPET study of LC.

Note this is from American Journal of Respiratory and Critical Care Medicine Volume 211, Issue Abstracts: American Thoracic Society International Conference Meetings Abstracts i.e. there is no accompanying fulltext

 

FWIW, I sent this to some RECOVER and NIH leaders.

Sharing Koroshetz's reply: 'This is a pretty clear abnormality as far as I can see. I worked with David Systrom on cases before coming to NIH. Looking forward to seeing the full paper.'

 

I know next to nothing about invasive C-PET: would deconditioning be expected to contribute to these findings?

 

I also know very little about iCPET my understanding is that it is CPET+taking certain additional measurements (mainly related to gas exchange whilst undergoing a CPET). It seems to be a way to understand exercise intolerance occuring during exercise in certain patients (exercise-induced pulmonary hypertension, exercise-induced heart failure with preserved ejection fraction, pre-load failure, or mitochondrial diseases cannot be diagnosed without iCPET).

There are however other studies in LC and/or ME/CFS that have done CPET+taking certain additional measurements (e.g. Rob Wüst took muscle biopsies, the intramural study did other things alongside the CPET "Ventilation/VCO2, oxygen saturation levels in the quadricep muscle as measured by Near Infrared Spectroscopy, and gross mechanical efficiency were not different between groups", Hanson looked at metabolimics etc).

If I've understood things correctly. The findings here seem to be

• lower pVO2 (this is also measured in standards CPETs so won't tell us anything new, presumably 2-day CPETs exist in ME/CFS to tell us there might be a decrease in pVO2 on the second CPET)
• reduced pQc (can be tried to be calculated from a standard CPET indirectly?)
  
• reduced pRAP (cannot be assed from a standard CPET?)
  
• reduced pSOE (cannot be assessed from a standard CPET?)

Not only that, how would "reconditioning" (say people do more low intensity heart rate activities) be expected to contribute to findings? Wouldn't both affect almost all of the above domains? Would it be useful to match patients vs controls in some of the measures (if a ME/CFS patient has a comparable pV02 to a control, are their other values comparable as well)? Presumably lower pV02 tells us nothing so why should other measures if they are related? Would a non-relation tell us anything?

Systrom once ran a placebo-controlled study with Pyridostigmine in ME/CFS. From what I understood all things mentioned above to be lower in ME/CFS improved slightly but not significantly in the treatment group whilst results in the general population appear to be mixed. Something like EPO probably increases all of the above things to some extent but isn't effective in treating ME/CFS. Does that mean that the above things have little to do with ME/CFS?

Are any of the results interesting or novel @Snow Leopard ?

 

483 ME/CFS patients is a massive sample size. Let's see how rigorously they were able to recruit them.

 

A very interesting abstract. My familiarity with exercise physiology is very limited but a few brief thoughts:

Impressively large ME cohort. LC group (n=73) is much smaller than the ME/CFS group (n=438) & the control group is smaller still (n=43) although the tests used are robust to unequal group sizes. This is a single iCPET, not a two-day protocol. Hopefully careful exclusion of confounders (obesity, anaemia, elite athletes, sub-maximal effort, comorbidities). Retrospective single-centre design. Demographics aren't specified in the abstract.

Findings seem to align with previous work. "Symptomatic but otherwise normal" controls raises a bit of a flag - what symptoms did they have, were they matched for age/sex/activity levels? No between-group difference on the VE/VCO2 slope, so no evidence of a primary ventilatory or pulmonary vascular limitation?

Low peak RAP & reduced peak SOE - deconditioning might impair oxidative capacity but seems like a high degree of extraction impairment - there's a drop in peak SOE of ~10%. IIRC bedrest experiments show VO2max falls mostly because of stroke volume & the arteriovenous O2 difference isn't modified by bedrest? Particularly interesting is the ME/LC similarity across all measures. Deconditioning may vary more with individual activity levels? A study with a deconditioned comparator group would be useful, I think.

Looking forward to reading the full paper when it comes out.

 

Does this imply that they did the CPET as a part of clinical testing?

I wonder about the ethics of that..