Elevated blood viscosity is associated with dysautonomia in long COVID symptoms, 2026, Tamariz et al

forestglip

forestglip

Moderator
Staff member
Elevated blood viscosity is associated with dysautonomia in long COVID symptoms

Tamariz, Leonardo; Milanes, Isabela; Bast, Elizabeth; Shehadeh, Lina A.; Klimas, Nancy; Palacio, Ana

Background
Long COVID is associated with elevated inflammatory and antibody levels. Elevated plasma proteins can increase blood viscosity and decrease blood flow. Our aim is to evaluate if blood viscosity is associated with long COVID outcomes.

Methods
We conducted a cross-sectional study and included a sample of patients enrolled in our long COVID clinic. We estimated whole blood viscosity (WBV) using two previously validated formulas and compared it with the NASA lean test, COMPASS-31 scale as measures of dysautonomia and the symptom burden as measured by the modified COVID-19 Yorkshire scale. We obtained WBV at three different shear stress. We divided WBV and evaluated the distribution of symptom scores using univariate and multivariate models.

Results
We included 185 patients for this study. Our sample had a mean age of 56 ± 11 years, included 53% minorities and 32% were women.

The mean C19-YRSm did not change with increasing tertile of WBV (p > 0.05) while the mean COMPASS-31 score increased with increasing tertile of WBV in all levels of shear stress.

In adjusted models the beta-coefficient of the C19-YRSm was (B -0.19p = 0.90) and for COMPASS-31 was (B 7.0 p = 0.01) for 208 s−1 and for all other levels of shear stress Twenty-three percent of patients in tertile 1 had either POTS or orthostatic hypotension compared to 32% on tertile 3 (p = 0.04).

Conclusion
Whole blood viscosity was associated with dysautonomia and not with long COVID symptoms.

Web | DOI | PDF | American Heart Journal Plus: Cardiology Research and Practice | Open Access
 
It's a shame that we are still seeing studies of lumped Long COVID. This required participants to meet WHO criteria and to have had a diagnosis of Long COVID for at least a year. It makes it hard to make much use of this data.

Participants were consecutive patients at the Miami Veterans Affairs Healthcare System Long COVID clinic.

Participants answered the Elaros COVID-19 Yorkshire Rehabilitation Scale. (17 item)
The Modified C19-YRS (C19-YRSm) is a 17-item patient-reported outcome measure (PROM), with each item rated on a 0-3 numerical rating scale. Zero represents the symptom not being present; 1 represents a mild problem (not affecting daily life); 2 represents a moderate problem (affecting daily life to a certain extent); and 3 represents a severe problem (life disturbing or affecting all aspects of daily life). The C19-YRSm, similar to the original version, is broken down into four sub-scales concerned with the severity of patients’ key symptoms, functional limitations, other symptoms, and overall health. Pre-COVID scores are also captured for comparison. The worst scores for each item (Questions 1-10) sum to give the Symptom Severity sub-scale (score 0-30), Questions 11-15 sum to give the Functional Disability sub-scale (0-15), Question 16 is the other symptoms sub-scale (0-25), and Question 17 is the overall health score (0-10).
Click to expand...
The score for 'Anxiety/Mood' is the highest score given for these questions:
Feeling anxious
Feeling depressed
Having unwanted memories of your illness or time in hospital (it's not clear if they are talking about only the acute illness or also the chronic one)
Having unpleasant dreams about your illness or time in hospital
Trying to avoid thoughts and feelings about your illness or time in hospital

There is a PEM question.

This study only used the Symptom Severity subscale - the sum of the worst scores for Questions 1- 10.
For each symptom a rating between 0 and 3 as to how much the symptom affects daily life:
Breathlessness, cough/change of voice, fatigue (tiredness not improved by rest), smell/taste, pain/discomfort, cognition, palpitations/dizziness, PEM, anxiety/mood, sleep (no mention of unrefreshing sleep)

It seems to me that this measure of symptom severity does not cover ME/CFS well.

In fact, because the symptom list is so diverse, I'm not sure it's a good measure of anything much at all.



Measures of dysautonomia were the COMPASS-31 and a NASA lean test.
The COMPASS-31 is another measure that throws all sorts of symptoms in together - and calls them dysautonomia.
 
Last edited:
I've found this paper difficult to understand, even though there does not seem to be much to it.

Measures of viscosity:
There were two Whole Blood Viscosity measures.
de Simone - calculated from hematocrit and plasma proteins​
Walburn-Schneck model is calculated from shear rate, hematocrit, total protein minus albumin​

I'm not sure yet how they measured the shear rate. I think maybe they just plugged in three different shear rates to cover the different sorts of vascular circulation e.g. in capillaries, artieries. It's quite possible I'm not understanding something here.

A bigger number means more viscosity.

They seem to have divided the patients from the Long Covid clinic into three groups based on their calculated estimated blood viscosities. But, I did not see anywhere where they told us what blood viscosities in healthy people are.

Fifth, we did not measure whole blood viscosity but rather estimated it. However, we used two formulas and the results were consistent.
Click to expand...
Well, that's because the inputs in the equations seem to be quite similar. I'm not sure but I think if you are going to say something about blood viscosity in a disease, you really need to measure blood viscosity and tell us how it compares between healthy people and the disease group.

Otherwise, tell us the actual inputs to the equations - tell us what the levels of hematocrit and plasma proteins were. I didn't see these reported anywhere.
 
They tell us that 10% of the 33.33% of the people with the lower blood viscosity have ME/CFS; 12% of the people with the mid-range blood viscosity have ME/CFS; 14% of the 33.33% of the people with the highest blood viscosity have ME/CFS. That seems to suggest that blood viscosity isn't relevant to ME/CFS. And that might be true, but I'm not getting a whole lot of confidence about the assessments of blood viscosity or the assessments of ME/CFS.

I don't even know if any of the three blood viscosity groups have blood viscosities that aren't normal.

I feel as though I have wasted my time on this. The report badly needed an editor to make the report clearer and more useful. Possibly the study needed a rethink too.

Our study found that increases in whole blood viscosity are associated with dysregulation of the ANS among patients diagnosed with long COVID, and not with long COVID symptom burden.
Click to expand...
I'm not surprised they didn't find any pattern of anything using that Yorkshire Rehabilitation Scale for symptom burden. It's a scale that jumbles up different sorts of pathologies and gives you a number that tells you nothing much. In the highest blood viscosity tertile, only 4% were women. They are saying that autonomic dysregulation was more prevalent in that highest blood viscosity tertile. That is surprising.

The discussion says
Another potential explanation could be the fact that the third tertile had a higher proportion of males, and males have higher blood viscosity than females [30]. However, our study calculated adjusted means which accounted for demographics.
Click to expand...
Hmm.


The strength of our study includes the use of validated and standardized assessments for other clinical research studies,
Click to expand...
Just because the authors of a scale say it's validated and standardised, that doesn't make it useful.

Second, the COMPASS-31 is a subjective measure with many non-specific symptoms that may overlap with other conditions (e.g. dizziness, diarrhea and/or constipation, urinary issues), which may be difficult to interpret in the absence of known dysautonomia.
Click to expand...
Yes.

Third, we did not have a control group and therefore are unable to compare blood viscosity with non-long COVID patients to implicate causality.
Click to expand...
Yes.

Sorry for my posts. I can't find anything to take from this study.
 
It’s a shame the study didn’t include healthy controls. Anecdotally, some patients report having nurses comment on “thick” or “sludgy” blood when they have their blood drawn. It seems like some of us might have rather viscous blood, but that could be true of some healthier people too.
 
Hutan said:
Sorry for my posts. I can't find anything to take from this study.
Click to expand...
Please don't be - they are much appreciated and always welcome. Particularly for those of us who sometimes struggle, having confirmation that we are not missing anything can be as helpful as being guided through the bits we have missed.

A real shame about the study. Seems it's yet another waste of time and resources with the chance to actively set things back and, given the alleged results, may well inspire more people to seek out blood-thinners and who-knows-what supplements.

[edit: grammar]
 
Last edited:
DHagen said:
Seems it's yet another waste of time and resources with the chance to actively set things back and, given the alleged results, may well inspire more people to seek out blood-thinners and who-knows-what supplements.
Click to expand...
Yes. It is possible that something along the lines of sticky blood or increased levels of plasma proteins or something is a problem in a subset of people with 'Long COVID'. But this paper provides no evidence of that.

The Discussion includes a long spiel about why high blood viscosity might be important in Long COVID. It reads like the author copied this from the statement they wrote to justify doing the study. (I've added paragraph breaks for readability.)

There is enough evidence to show that the SARS-CoV2 causes a multisystemic illness with significant cardiovascular and thrombotic consequences [18]. Elevated blood viscosity has a potential role as a clinically actionable cardiovascular biomarker since it has been shown to predict atrial fibrillation [4], predicts the severity of coronary artery disease [19], the severity of heart failure [20] and thrombus formation seen by echocardiography [21]. Elevated whole blood viscosity has also been reported in chronic infections such as HIV [22]and hepatitis B [23].

Three studies have evaluated blood viscosity in acute SARS-CoV2 infection. Waksman and Choi et al. calculated blood viscosity in 5621 COVID-19 acute infections and found that elevated blood viscosity was associated with increased need for respiratory support and higher mortality [8], [24]. Garcia-Callejo et al. reported on repeated measure of blood viscosity among 300 patients found an association with deafness [25]. A potential mechanism to explain adverse outcomes with high viscosity after an acute infection is the report by Nader et al. that found among 172 patients admitted with COVID-19 and 38 controls that increased viscosity was associated with increased red blood cell aggregation and clot firmness [26].

However, all of these studies were done on acutely ill patients with SARS-CoV2 infection, where the increase in inflammatory molecules is expected. At the same time, the acute COVID-19 infection leads to a dysregulated immune reaction [27] and potential myocardial damage [28] both of which are related to dysautonomia and elevation of blood viscosity. All of these acute findings translate into long COVID where the clinical presentations are also a consequence of a persistent inflammatory reaction, thrombosis, and endothelial damage, all of these can be associated with dysautonomia. Therefore, there is a need for blood viscosity evaluations among long COVID patients as none have been reported.
Click to expand...
It ends, 'therefore there is a need for blood viscosity evaluations among long COVID patients as none have been reported'. And yet plenty of people are going to read the abstract or even just the title and think this study provides evidence of elevated blood viscosity in Long COVID.

This study really annoyed me. 'Is blood viscosity elevated in people with various subsets of Long COVID?' is a simple and useful question to ask. Or even, 'Are some of the molecules that might be contributing to higher blood viscosity higher in the various subsets of Long COVID?'.

I'm sorry to the junior author, but there's a lot more at stake than their research career. We are relying this people for our futures and those of our children. Do better. I remember watching a podcast by the senior author Ana Palacio, who set up the Long COVID clinic in the Miami Veterans hospital and thinking that she seemed genuinely wanting to help. But, for goodness sake, 70% of the veterans with Long COVID in the clinic in this study are reported to be receiving physical therapy.

She and Nancy Klimas (both authors of this paper) need to take their research efforts a lot more seriously if they are not to keep producing misleading unhelpful papers. Nancy has been doing research on ME/CFS and Long COVID and Gulf War Illness for a long time now and I'm struggling to think of any useful solid finding.

It wouldn't be hard to do a better job:

* Don't use the Elaros Yorkshire scale or the COMPASS Dysautonomia scale which both bundle disparate symptoms that can be caused by many illnesses together to produce meaningless total scores. Instead, stratify your cohorts into groups based on acute illness characteristics, presence of persisting observable pathology and symptoms e.g. lung damage and persisting cough and breathlessness; mild acute illness and ME/CFS. For ME/CFS, use FUNCAP to determine severity, or even something simpler such as having your subjects choose their severity level based on one of the 'mild, moderate, severe' characterisations.

* Be clearer about what you are measuring. Actually report the results of what you measured (e.g. hematacrit, albumin) for each stratified group. Set the results in context by comparing them with either a control group or at least some established general population means.
 
Fifth, we did not measure whole blood viscosity but rather estimated it. However, we used two formulas and the results were consistent.
Click to expand...

Hutan said:
Well, that's because the inputs in the equations seem to be quite similar. I'm not sure but I think if you are going to say something about blood viscosity in a disease, you really need to measure blood viscosity and tell us how it compares between healthy people and the disease group.
Click to expand...

This could have been useful, but along with recovered controls it probably needed direct viscometry, rather than an estimate based simply on red blood cell proportion and protein content.

In the discussion they discuss (and then speculate about) coagulation —

A potential mechanism to explain adverse outcomes with high viscosity after an acute infection is the report by Nader et al. that found among 172 patients admitted with COVID-19 and 38 controls that increased viscosity was associated with increased red blood cell aggregation and clot firmness. However, all of these studies were done on acutely ill patients with SARS-CoV2 infection, where the increase in inflammatory molecules is expected. […] All of these acute findings translate into long COVID where the clinical presentations are also a consequence of a persistent inflammatory reaction, thrombosis, and endothelial damage, all of these can be associated with dysautonomia. Therefore, there is a need for blood viscosity evaluations among long COVID patients as none have been reported.
Click to expand...

Continuing —

If our hypothesis of increased inflammatory and prothrombotic molecules is true, this complements with the fact that SARSCoV2 causes endothelial necroptosis, platelet activation, neutrophil extracellular traps (NETosis) leading to the formation of small thrombus. This thrombus in the microcirculation, increases in blood viscosity combined with loss of laminar flow due to endothelial dysfunction, small fiber neuropathy, decreased red blood cell deformability, severely compromise oxygen delivery to tissues - the physiological basis of long COVID symptoms.
Click to expand...

I understand that to directly measure the blood viscosity ex vivo you would need to prevent the blood from clotting, so there is a problem to try and include that aspect of the assessment. Those two requirements are in contention. An alternative/additional measurement would have been thromboelastography (TEG).

Though, quickly looking for recent studies, there was this small numbers (n=20) study last year looking at PASC with dyspnoea: IgA autoimmunity and coagulation among post-acute sequelae of SARS-CoV-2 infection PASC patients with persistent respiratory symptoms: a case-control study (2025). They reported "TEG tests showed no differences between the groups".

Edit: removed AI summary of direct viscometry techniques per moderators request.
 
Last edited:
The whole thing seems to be based on a belief in a microclot type theory - where viscosity doesn't actually matter because the problem is clotting (where viscosity ceases to be relevant because the lump is solid).

It is absurd to try to claim a change inviscosity without actually measuring it. I doubt that EDTA makes that much difference. They could have done capillary flow studies or simply looked at troutine plasma viscosity plus haematocrit. But we can be pretty sure that plasma viscosity will have been normal. It follows the ESR quite well.

People with high blood viscosity in myeloma or polycythaemia do not have a clinical picture like Long Covid.
 
You must log in or register to reply here.
Back
Top Bottom