Preprint Shear Stress Tolerance Threshold, eNOS Uncoupling, and the Two-Threshold Model of [PEM] in [LC]: A Mechanistic Hypothesis [...], 2026, Karipidis et al

[forestglip]

Shear Stress Tolerance Threshold, eNOS Uncoupling, and the Two-Threshold Model of Post-Exertional Malaise in Long COVID: A Mechanistic Hypothesis with Implications for Physiotherapy Practice

Karipidis, Yiannis K.; Karipidis, Konstantinos Y.

Background
A subset of Long COVID patients presents with post-exertional malaise (PEM), delayed symptom exacerbation 12–48 hours after minimal mobilisation, chest tightness, dyspnea at rest, suppressed heart rate variability (HRV), and cognitive impairment with consistently normal macrovascular cardiac investigations.

Hypothesis
This paper proposes that SARS-CoV-2-induced endothelialitis may, in a specific vascular-PEM phenotype, cause progressive depletion of tetrahydrobiopterin (BH4) and accumulation of its oxidised form (BH2). BH2 competitively occupies the eNOS cofactor site, causing functional eNOS uncoupling: the enzyme produces superoxide (O₂⁻) rather than nitric oxide (NO). Superoxide combines with residual NO to form peroxynitrite (ONOO⁻), which simultaneously destroys remaining BH4 and nitrates dihydrofolate reductase (DHFR), creating a self-sustaining vicious cycle. The core insight: this is not NO deficiency it is NO misdirection.

Two-Threshold Framework
A central contribution of this model is the distinction between two mechanistically independent thresholds: the PEM threshold the exertional load at which the nitro-oxidative cascade becomes self-amplifying and the functional shear stress tolerance threshold the minimum sustained laminar shear stimulus required to maintain GTPCH-1-driven BH4 synthesis. These thresholds are not the same variable and must not be managed as one. Current rehabilitation approaches targeting only the PEM threshold may inadvertently allow progressive BH4 pool erosion through insufficient laminar shear stimulus, thereby lowering the PEM threshold over time.

Biomarker Validation Framework
A tiered biomarker panel is proposed for prospective validation, with emphasis on dynamic pre/post-exertion measurement. The proposed signature: elevated nitrotyrosine, reduced BH4/BH2 ratio, elevated ADMA, impaired flow-mediated dilation, and suppressed post-exertional HRV, all worsening 12–24 hours after a controlled low-intensity exercise challenge. This model generates a falsifiable prediction: if this post-exertional shift is not demonstrated in patients meeting vascular-PEM phenotype criteria, the proposed mechanism would not be supported.

Physiotherapy Implications
The two-threshold model mechanistically challenges current graded exercise therapy guidelines for Long COVID, which do not distinguish between activity modality and hemodynamic quality. Heart rate alone does not capture the endothelial stimulus: two activity patterns producing identical heart rates may generate protective laminar shear (10–20 dyn/cm²) or harmful oscillatory shear (0–5 dyn/cm²) respectively. The therapeutic goal in this phenotype is shear tolerance threshold maintenance and progressive restoration not conventional capacity building.

Scope
This hypothesis is phenotype-specific and does not propose a universal mechanism for Long COVID or ME/CFS. It addresses one biochemically coherent pathway within a specific vascular-PEM subset and is offered as a framework for prospective clinical investigation

Web | DOI | PDF | Zenodo | Preprint

 

[Utsikt]

I don’t have the knowledge to assess their hypothesis, but I appreciate that they have made clear predictions and proposed how to test it.

Although I’m a bit worried by their uncritical reading of the literature where they claim that GET is how PEM is usually treated.

 

[Felis Catus]

Although I’m a bit worried by their uncritical reading of the literature where they claim that GET is how PEM is usually treated.

I'm wondering if it might have to do with how GET is being promoted by the healthcare professionals in Greece. Just a guess. I don't know anything about the views and treatments there.

 

[Utsikt]

I'm wondering if it might have to do with how GET is being promoted by the healthcare professionals in Greece. Just a guess. I don't know anything about the views and treatments there.

That sounds likely, but anyone should be able to see how bad the research is if they bother actually looking at the sources.

 

[Murph]

I've seen many much worse hypothesis papers, even in these pages. This one is humble and limited and reasonable.

Hypotheses are ... not bad. I see theories as the vessels in which we can collect data. They are also the measuring cup in which we measure our data: is this hypothesis disproved by the data?

The only problem is that because theories are cheap to make and data is expensive to collect we have far more cups than water to fill them.

 

[YiannisK]

I'm wondering if it might have to do with how GET is being promoted by the healthcare professionals in Greece. Just a guess. I don't know anything about the views and treatments there.

I think there is a misunderstanding. Our preprint actually provides the biochemical evidence (eNOS uncoupling) for why traditional exercise (like GET) can be harmful. The 'Two-Threshold Model' we propose is designed to prevent PEM by identifying the vascular limits that were previously ignored. We are not promoting GET we are proposing a way to avoid its dangers through vascular biology https://doi.org/10.5281/zenodo.19474803

 

[YiannisK]

I don’t have the knowledge to assess their hypothesis, but I appreciate that they have made clear predictions and proposed how to test it.

Although I’m a bit worried by their uncritical reading of the literature where they claim that GET is how PEM is usually treated.

Thank you for the feedback. Our aim is exactly to provide the biochemical 'why' behind the failure of traditional exercise approaches like GET, replacing them with a safe, threshold-based framework.

 

[SNT Gatchaman]

Welcome @YiannisK

 

[ScoutB]

Was thinking about this a bit but now too foggy continue. Here are a few related threads:

- Found this thread where it was mentioned that measuring BH4 is tricky (the paper linked there actually claimed pwME had higher BH4, but the S4ME comments suggest the researchers may have mishandled their samples).

- A study linked here found no elevation of ADMA in LC, but their selection criteria sounds very broad, as usual.

 

[Jonathan Edwards]

What is endothelialitis @YiannisK ? I spent my career in inflammation research and have never heard of it. All I can see in the abstract is some rather stretched out speculations. As far as I can work out there is no point in trying to get people with Long Covid or ME/CFS to do exercises of any sort.

 

[YiannisK]

Was thinking about this a bit but now too foggy continue. Here are a few related threads:

- Found this thread where it was mentioned that measuring BH4 is tricky (the paper linked there actually claimed pwME had higher BH4, but the S4ME comments suggest the researchers may have mishandled their samples).

- A study linked here found no elevation of ADMA in LC, but their selection criteria sounds very broad, as usual.

BH4 findings from Gottschalk et al.: this study found elevated serum BH4 in ME/CFS, which appears to contradict deficiency models. However, as the S4ME community has already noted, BH4 measurement is extremely problematic BH4 has a half-life of 16 minutes in PBS at room temperature, and the study's handling protocol likely caused significant degradation. What they may have been measuring is a compensatory upregulation of BH4 synthesis (via GCH1) in response to increased oxidative consumption the body producing more because it's losing more. Our model predicts exactly this: the issue is not total BH4 production but the BH4/BH2 ratio at the eNOS cofactor site, which static serum measurement cannot capture. This is why we propose dynamic pre/post-exertion measurement as the critical test.


Preprint: https://doi.org/10.5281/zenodo.19474803

 

[YiannisK]

What is endothelialitis @YiannisK ? I spent my career in inflammation research and have never heard of it. All I can see in the abstract is some rather stretched out speculations. As far as I can work out there is no point in trying to get people with Long Covid or ME/CFS to do exercises of any sort.

Thank you for engaging with the work.


On the terminology: "endotheliitis" (also spelled "endothelialitis" in some sources) was coined by Varga et al. in The Lancet (2020) it describes SARS-CoV-2 direct infection of endothelial cells with inflammatory infiltrate, confirmed histologically at autopsy. It's well-established in COVID vascular research. I acknowledge the spelling inconsistency in our manuscript and have noted it for correction.

On the substance: the hypothesis does not propose exercise for people with Long COVID or ME/CFS. It proposes the opposite that current exercise recommendations are dangerous precisely because they don't distinguish between haemodynamic patterns that protect vs. damage the endothelium. The core argument is that shear stress quality (laminar vs. turbulent) matters more than intensity, and that crossing the PEM threshold is biochemically destructive. The framework explicitly defines four phenotypes, one of which (Phenotype D severe) contraindicates exercise entirely until biochemical stabilisation is achieved.

 

[Jonathan Edwards]

On the terminology: "endotheliitis" (also spelled "endothelialitis" in some sources) was coined by Varga et al. in The Lancet (2020) it describes SARS-CoV-2 direct infection of endothelial cells with inflammatory infiltrate, confirmed histologically at autopsy. It's well-established in COVID vascular research.

This is what worries me. I was doing inflammation research when these people were in kindergarten or not yet born. If there is an infiltrate of cells going through the vessel wall into the tissue that is inflammation of that tissue, as in myositis. If the infiltrate is merely into the layers of larger vessels, where there is media and adventitia, then it is vasculitis. You cannot have infiltration into endothelium, itself, since it is just a monolayer of cells. So endotheliitis doesn't mean anything.

Unfrtunately, all sortss of things are 'well-established' in biomedical research these days, in the sense of having been made fashionable by people who don't know the background basic biology. A large proportion ofLong Covid research is just muddled thinking and bad data.

I am not aware of any good evidence for vascular pathology in Long Covid. Clearly there are thrombotic events in acute severe Covid but that is a different issue.

 

[Jonathan Edwards]

On the substance: the hypothesis does not propose exercise for people with Long COVID or ME/CFS. It proposes the opposite that current exercise recommendations are dangerous precisely because they don't distinguish between haemodynamic patterns that protect vs. damage the endothelium.

But presumably a failure to distinguish is only relevant if some form of exercise is relevant, and there is no reason to think it is. And it is a bit confusing because I doubt those recommending exercise had any interest in endothelium either way. I don't see any evidence base for this theory so far, I am afraid. As we have discussed here a ot in the past, PEM does not seem to be a local phenomenon affecting the part used. It can be precipitated by mental exertion. I cannot see any good reason to relate it to microvascular events.

 

[Utsikt]

Thank you for the feedback. Our aim is exactly to provide the biochemical 'why' behind the failure of traditional exercise approaches like GET, replacing them with a safe, threshold-based framework.

Hi and welcome!

As far as I’m aware (as a layperson), exercise has not been proven to be an effective treatment for any condition. Regular physical activity probably can reduce the risk of certain health issues, but exercise in the rehabilitation sense is largely a myth.

I am not aware of any good evidence for vascular pathology in Long Covid. Clearly there are thrombotic events in acute severe Covid but that is a different issue.

There’s an increased risk of adverse cardiovascular events following covid infections, even for mild infections. I don’t know why.

 

[SNT Gatchaman]

You cannot have infiltration into endothelium, itself, since it is just a monolayer of cells. So endotheliitis doesn't mean anything.

I drafted along similar lines but didn't post as it was late last night and I expected you would reply (with more authority).

Spoiler: Draft

But the endothelium is a monolayer. So if there's an inflammatory infiltrate, the inflammatory cells are passing beyond the endothelium. It sounds as if they're trying to make the case that the viral infection of and damage to endothelial cells specifically leads to EC apoptosis and thereby inflammatory cell infiltration into deeper vascular layers (or perivascular compartments). This is in distinction to normal inflammation where the endothelial junctions loosen to allow inflammatory cell infiltration, with competent healthy ECs facilitating this (plus oedema).

But can we not simply call it a vasculitis? It's easier to pronounce and write. I hope no-one tries to make intimitis/intimaitis, mediitis/mediaitis or adventitiitis/adventitiaitis! ("Stop trying to make 'fetch' happen".)

Ref is Endothelial cell infection and endotheliitis in COVID-19 (2020)

Are there other situations where endothelial cells have direct viral infection, rather than becoming dysregulated from viral proteins or other inflammatory cell effects, as with HIV? Or is this unique to SARS and the ACE2 receptor? A quick Google suggests it's unique to the related family of coronaviruses, in particular SARS1, MERS, SARS2 and some others in bats.

If we wanted to have a distinct name to underscore that the normal process of inflammation is itself dysregulated due to endothelial cell infection and EC apoptosis, maybe "endotheliopathy" would be better. It's easier to say.

 

[Trish]

I have read the abstract and am puzzling away at the words used, specifically stress, laminar shear, tolerance threshold and mechanistic. It sounds like things I learnt in physics rather than biology. Can someone help me visualise on what scale these forces are acting? Atom, molecule, organelle, cell, tissue? Are these physical phenomena or metaphors?

As you can see, I'm no expert. In fact I haven't a clue what this hypotheses is about. Which is frustrating.

 

[YiannisK]

This is what worries me. I was doing inflammation research when these people were in kindergarten or not yet born. If there is an infiltrate of cells going through the vessel wall into the tissue that is inflammation of that tissue, as in myositis. If the infiltrate is merely into the layers of larger vessels, where there is media and adventitia, then it is vasculitis. You cannot have infiltration into endothelium, itself, since it is just a monolayer of cells. So endotheliitis doesn't mean anything.

Unfrtunately, all sortss of things are 'well-established' in biomedical research these days, in the sense of having been made fashionable by people who don't know the background basic biology. A large proportion ofLong Covid research is just muddled thinking and bad data.

I am not aware of any good evidence for vascular pathology in Long Covid. Clearly there are thrombotic events in acute severe Covid but that is a different issue.

I appreciate your focus on classical histology. However, the paradigm of vascular pathology has shifted significantly towards mechanobiology. The term 'endotheliitis' may be anatomically debated, but the functional failure of the monolayer is a proven reality.
Specifically, research (e.g., Chao et al., 2017, Journal of Cellular Physiology) demonstrates that Low Shear Stress (LSS) directly induces endothelial oxidative stress via the AT1R/eNOS/NO pathway. This isn't about 'infiltrates' crossing the vessel wall; it’s about a biochemical switch within the endothelial cell itself. When laminar flow is lost or disrupted, the eNOS enzyme uncouples and starts producing superoxide instead of Nitric Oxide.
This is exactly what we observe in Long COVID: a state of chronic microvascular hypoperfusion and 'pseudo-hypoxia' that standard macrovascular tests cannot detect because they aren't looking for enzymatic flux, but for structural blockages or infiltrates.
Our 'Two-Threshold Model' isn't following a fashion; it’s a clinical application of established vascular mechanobiology. We propose that the PEM trigger is the point where the system can no longer manage the ROS produced by uncoupled eNOS under exertion. We believe it's time to bridge the gap between 'kindergarten' histology and modern redox proteomics

 

[Jonathan Edwards]

This isn't about 'infiltrates' crossing the vessel wall; it’s about a biochemical switch within the endothelial cell itself.

OK, so it isn't endotheliitis.

I appreciate your focus on classical histology. However, the paradigm of vascular pathology has shifted significantly towards mechanobiology.

You mean the fashion of the Twitterati has gone that way I guess. I am afraid I am a rather old hand at biomedical scepticism. I have read more bullshit than you have had hot dinners!

This is exactly what we observe in Long COVID: a state of chronic microvascular hypoperfusion and 'pseudo-hypoxia' that standard macrovascular tests cannot detect because they aren't looking for enzymatic flux, but for structural blockages or infiltrates.

Well if you cannot detect any hypoperfusion, how do you know it is there? Hypoperfusion makes your skin change colour. That doesn't happen in Long Covid. I am afraid I dont believe a word of it. Pathobiology that isn't there isn't there.

Our 'Two-Threshold Model' isn't following a fashion; it’s a clinical application of established vascular mechanobiology.

I know all about shear forces and NO and all that but I don't see any dots joining up here I am afraid. And if you want to join dots I think you would do better to avoid pseudoconcepts like endotheliitis. Biomedical science, like Guness, has always had a bit of useless froth on the top. Unfortunately, these days, my impression is that the froth fills most of the glass. I am only interested in the black stuff.

 

[YiannisK]

But presumably a failure to distinguish is only relevant if some form of exercise is relevant, and there is no reason to think it is. And it is a bit confusing because I doubt those recommending exercise had any interest in endothelium either way. I don't see any evidence base for this theory so far, I am afraid. As we have discussed here a ot in the past, PEM does not seem to be a local phenomenon affecting the part used. It can be precipitated by mental exertion. I cannot see any good reason to relate it to microvascular events.

I fully agree that there is currently no direct evidence base proving this mechanism, and the manuscript is framed explicitly as a falsifiable hypothesis rather than an established explanation of PEM.
I also agree that PEM cannot be reduced to a purely local muscular phenomenon, particularly given that cognitive or emotional exertion can precipitate it. That observation is highly important.

The intention of the model is not to argue for a peripheral-only mechanism, but to ask whether, in a subgroup of patients, endothelial / microvascular dysfunction could represent one systemic amplifier node rather than the sole cause.

Mental exertion may still fit within such a framework indirectly through autonomic activation, cerebral metabolic demand, altered neurovascular coupling, or impaired cerebral perfusion reserve rather than through skeletal muscle mechanisms alone.

In other words, the hypothesis is not “PEM = endothelium,” but whether endothelial redox dysfunction could interact with immune, autonomic, and CNS pathways in some phenotypes.

If biomarker studies fail to show the predicted post-exertional vascular signature, then the model should be rejected. That is why testability was made central to the paper.
I appreciate your critique.