Cognitive fatigue is related to reduced cerebral perfusion and sustained attention in patients with [post C-19]: An fMRI study, 2026, Hedberg et al

[forestglip]

Cognitive fatigue is related to reduced cerebral perfusion and sustained attention in patients with post COVID-19 condition: An fMRI study

Hedberg, Sonia Miri; Borg, Kristian; Stenberg, Jonas; Hedström, Stina; Granberg, Tobias; Gyllenberg, Alexandra; Petersson, Sven; Van Loo, Hadrien; Nordin, Love Engström; Möller, Marika C.

Abstract
Persistent fatigue is one of the most prevalent and disabling symptoms of Post-COVID-19 Condition (PCC). To investigate possible mechanisms underlying fatigue in PCC, the present study investigated the relation between different aspects of fatigue and fatigability, and cerebral blood flow (CBF) in individuals with PCC.

The participants, 22 patients and 19 controls, matched by age and sex, performed a 20-min-long psychomotor vigilance task (PVT) during fMRI. The total mean reaction time (RT) was used to measure processing speed and fatigability by dividing RTs into quartiles.

CBF was measured with pseudo-continuous arterial spin labeling. Self-reported state fatigue (VAS-scale) was assessed before and after the PVT.

Patients reported larger increase in state fatigue (p = 0.004) and longer RTs across all quartiles (p = 0.001) compared to controls, with no difference in fatigability between groups.

The voxel-wise analysis of CBF revealed significantly lower global CBF in patients compared with controls (49.2 vs. 51.7 mL/100 g/min). Regional hypoperfusion was observed in several clusters, including the right inferior occipital gyrus (27.9 vs. 46.4 mL/100 g/min), left postcentral gyrus (29.8 vs. 51.3 mL/100 g/min), and right middle cingulate cortex (30.5 vs. 48.7 mL/100 g/min).

Web | DOI | PDF | NeuroImage: Reports | Open Access

 

[SugarSquared]

I don’t have the energy to look into this, but does this fit in with Van Campen’s findings of reduced oxygen to the brain in people with orthostatic intolerance? Is that even what she finds or am I misremembering?

 

[ScoutB]

Van Campen’s findings of reduced oxygen to the brain

My foggy memory is saying those were studies looking at cerebral blood flow (while upright). Via extra-cranial Doppler? I think part of the problem is science broadly isn’t 100% sure how to reliably measure these things, or how cerebral blood flow/oxygenation/possibly other issues fit together. But I’m very intrigued.

 

[forestglip]

ME Association: "Research finds changes to cerebral blood flow in people with Long Covid"

---

Dr Charles Shepherd Comments:​

Back in the early 1990s, Dr Durval Costa at the Middlesex Hospital in London carried out some important neuroimaging research on the role of cerebral blood flow in people with ME/CFS.​

​

This research compared blood flow to an important part of the brain called the brain stem in people with ME/CFS, healthy controls and depression. The ME/CFS patients had a marked decrease in blood flow (hypoperfusion). As the brain stem helps to regulate a wide range of body functions – including heart rate and blood pressure, sleep patterns, balance etc – this could be another important clue in symptom explanation.​

​

Read the paper​

​

Sadly, these important results from over 30 years ago have never been properly followed up and validated.​

​

Changes in cerebral blood flow have also now been reported in Long Covid by a research group in Sweden and linked to fatigue.​

 

[SNT Gatchaman]

Interesting comparison to Walitt et al.

Spoiler: The psychological vigilance task (PVT)

To assess sustained attention and performance fatigability, a reaction time task was implemented using E-Prime software […]. The task was performed in the MR-scanner. During the task the participants were required to press a button with their right hand as quickly as possible when the target stimulus, a set of four zeros enclosed in a red rectangle, was displayed. Participants were instructed to refrain from responding in case of any other stimulus.

Participants received visual feedback following each response. This feedback was displayed for 1 s. An incorrect response or a response that exceeded 1 s was presented by “false answer” or “no answer” by the system, respectively. The intervals between stimuli varied in pseudorandom manner, ranging from 2 to 10 s. The whole task lasted for 20 min for all participants. The total number of responses ranged from 177 to 213.

The absence of a clear decline in performance over time may be explained by factors such as motivation or stress. In clinical experimental settings, patients may exert a disproportionately high level of cognitive effort to maintain performance during sustained attention tests. Such compensatory effort may temporarily mask performance decline while simultaneously contributing to a greater subjective experience of fatigue. Consequently, post-task recovery may be prolonged or more complex in PCC patients.

Therefore, the absence of observable performance decline should not be interpreted as evidence of lack of fatigability. However, firm conclusions cannot be drawn, as a self-reported measure of mental effort was not included in the present study. Future studies should consider assessing cognitive effort and the recovery process in the days following cognitive exertion when interpreting performance outcomes.

 

[Sean]

Adequate re-testing is a core necessity in trial protocol for ME/CFS. Even a single re-test 24 hours later may not be enough to fully capture the story.

It is why I believe that the sooner we get an in vitro test for PEM, the sooner we will start making good progress. Testing it thoroughly and robustly in live patients is too slow and costly, and too risky for patients.

 

[fst]

The absence of a clear decline in performance over time may be explained by factors such as motivation or stress. In clinical experimental settings, patients may exert a disproportionately high level of cognitive effort to maintain performance during

sustained attention tests. Such compensatory effort may temporarily mask performance decline while simultaneously contributing to a greater subjective experience of fatigue. Consequently, post-task recovery may be prolonged or more complex in PCC patients.

Refreshing to see such an intepretation.

Participants were matched for age and sex but more controls had university-level education.

Spoiler: Participants

The groups did not differ in age or sex, but a significantly higher proportion of controls had university-level education compared to PCC patients (p = 0.01). The two groups were comparable in terms of occupational background. Consistent with this,
cognitive assessments revealed no differences between patients and
controls in problem-solving abilities (WAIS-IV Matrix Reasoning), used
as a proxy for cognitive reserve (Hedstrom ¨ et al., 2026). PCC patients
scored significantly higher than controls on both depression and anxiety
(p < 0.001) (Table 1).

Table 4 shows the regions with significant differences in cerebral blood using corrected p-values.

Anything that sticks out? I think the cerebellum came up a few times here.

Spoiler: Table 4

Region	MNI (x,y,z)	CBF Patients	CBF Controls	Contrast	Vox/Vol cm3
Right inferior occipital gyrus	40, -62, 12	27.9	46.4	t-test1	98/2.646
Left postcentral gyrus	-23, -32, 39	29.8	51.3	t-test2	94/2.538
Right middle cingulate cortex	10, -7, 34	30.5	48.7	t-test3	62/1.674
Right cerebellum	10, -60, -50	38.5	55.2	t-test4	35/0.945
Left middle occipital gyrus	-23, -55, 31	35.1	60.0	t-test5	30/0.810
Right middle cingulate cortex	13, 6, 44	33.1	48.0	ANOVA group	73/1.971
Right caudate nucleus	25, 27, 11	47.1	53.6	ANOVA time	42/1.134
Global GM CBF	—	49.2	51.7	—	—

The conclusion

Individuals with PCC showed pronounced subjective fatigue together with slower and more variable performance on the PVT, consistent with an attenuated capacity to sustain attention. In addition, we observed associations between fatigue measures and reduced cerebral perfusion, both globally and in regions implicated in fatigue-related neural networks. Overall, our findings suggest that impaired cerebral perfusion may contribute to the development of fatigue in PCC patients.

 

[SNT Gatchaman]

Anything that sticks out? I think the cerebellum came up a few times here.

I don't know. I get the sense that these studies showing changes in regional blood flow are all over the place. Anterior cingulate cortex, no mid cingulate cortex, no left prefrontal cortex, yes definitely the left side only, no wait we found it was the right prefrontal cortex. Everyone, including these authors are trying to hammer in an explanation for fatigue to the networks they think are involved.

At some point we should probably go through and tabulate the various studies, to confirm or refute that "all over the place" suggestion.

I think the global cerebral blood flow change may be the real finding - and relatively modest (otherwise we'd be in fairly catastrophic shape). Eg here, global grey matter CBF was 51.7 mL/100g/min in HC and 49.2 in LC. All the regional blood flow changes may then be random findings in studies with insufficient numbers, all secondary to the global problem, but not forming a cohesive pattern.

 

[SugarSquared]

I wonder if the participants were sitting upright. University classes have started again, and I’m trying my best to take them upright at my desk at home for ease of taking notes. It is a lot more demanding though. On a day of bad PEM, I had to switch to laying down to have a hope of even being able to pay attention. I hope to be able to take most of my classes sitting upright, but it will probably have to depend on how I’m feeling that day.

 

[fst]

I think the global cerebral blood flow change may be the real finding - and relatively modest (otherwise we'd be in fairly catastrophic shape).

For comparison (and mostly because I was interested and trying to get a feel for those numbers), this study measured cerebral blood flow during waking vs sleep.

It decreased by roughly 15% (as opposed to the 5% group difference from above).

From the study:

Grey matter CBF decreased from 59.7 ± 9.0 to 49.5 ± 8.1 ml/100g/min (p < 0.0001), and white matter CBF from 28.7 ± 4.9 to 25.1 ± 3.0 ml/100g/min (p < 0.001).

 

[Trish]

I wonder if the participants were sitting upright. University classes have started again, and I’m trying my best to take them upright at my desk at home for ease of taking notes. It is a lot more demanding though. On a day of bad PEM, I had to switch to laying down to have a hope of even being able to pay attention. I hope to be able to take most of my classes sitting upright, but it will probably have to depend on how I’m feeling that day.

Do you have a suitable supportive chair that raises your legs, or even sit along a sofa with your back well supported and legs horizontal? Or in bed with an over the bed table for your laptop? Sitting upright is exhausting for pwME.

 

[SugarSquared]

Do you have a suitable supportive chair that raises your legs, or even sit along a sofa with your back well supported and legs horizontal? Or in bed with an over the bed table for your laptop? Sitting upright is exhausting for pwME.

Yes I do. I have reclining chair in my office. I bought a monitor for using at my desk, but I’m considering figuring out a set up where I can have the monitor in front of my reclining chair too.

I think I got too confident in my ability to spend time upright from my travelling this summer where I was able to do it more. School is definitely not the same exertion level as a fun conversation at the dinner table.

On the topic of my original question, I realized while talking with my fiancée that they were likely laying down for this study as I don’t believe any upright MRIs exist. I wonder what tool could be used to measure cerebral blood flow in different positions (sitting and laying down). I suspect most pwME/LC would experience cognitive fatigue faster sitting upright, and it’d be interesting to see if that’s correlated with lower cerebral blood flow.

 

[Hutan]

I think the global cerebral blood flow change may be the real finding - and relatively modest (otherwise we'd be in fairly catastrophic shape). Eg here, global grey matter CBF was 51.7 mL/100g/min in HC and 49.2 in LC. All the regional blood flow changes may then be random findings in studies with insufficient numbers, all secondary to the global problem, but not forming a cohesive pattern.

I suspect most pwME/LC would experience cognitive fatigue faster sitting upright, and it’d be interesting to see if that’s correlated with lower cerebral blood flow.

It's frustrating that we can't even pin this finding down. It doesn't sound like rocket science to answer the question ' is there lower cerebral blood flow in ME/CFS?'. Some people are very convinced that this is true, and, to be fair, it feels as though it should be true, with our need to lie down, and the similarity of some of our symptoms with trying to function at high altitude. But, I don't think we can say that yet.

This study has a fairly small sample size. And they don't tell us much about the post-Covid-19 condition participants in terms of working out if they could have ME/CFS. They have fatigue and the mean time between illness onset and the fMRI was 32 months.

The group component in the multivariate modelling showed results matching the t-test, both with absolute CBF and normalized CBF. In
significant clusters, the CBF was on average 38.2% lower in PCC patients compared to controls. While examining whole-brain gray matter, PCC patients showed a 4.8% lower mean CBF (49.3 ml/100 g/min) than controls. The PCC group also demonstrated greater variability (SD 10.7 vs. 8.0), a slightly higher maximum value (77.2 vs. 68.0), and a similar minimum CBF compared to controls (35.5 vs. 35.3 ml/100 g/min).

A mean difference of 4.8% lower mean CBF could be important I guess, if it was fairly consistent across all the participants. But the standard deviation for the PCC group was 10.7 ml/100g/min - that's for a mean of 49.4 ml/100g/min. (and mean 51.7 and SD 8.0 for the controls). It's difficult to get excited about that. There's pretty close to no difference, and there were probably a few decisions the researchers could make that would skew things slightly in the direction of the expected result e.g. excluding 3 of the controls but not the matching patient, probably some decisions about adjustments for brain size.

And, one of the controls actually had a slightly lower CBF than all of the patients.

I agree @SugarSquared . We need investigations of cerebral blood flow to be done with people being upright.

 

[Hutan]

Mean reaction times

That was interesting, although a bit hard to see what is happening at the individual level.

 

[Jonathan Edwards]

A mean difference of 4.8% lower mean CBF could be important I guess, if it was fairly consistent across all the participants. But the standard deviation for the PCC group was 10.7 ml/100g/min - that's for a mean of 49.4 ml/100g/min. (and mean 51.7 and SD 8.0 for the controls). It's difficult to get excited about that. There's pretty close to no difference, and there were probably a few decisions the researchers could make that would skew things slightly in the direction of the expected result e.g. excluding 3 of the controls but not the matching patient, probably some decisions about adjustments for brain size.

And, one of the controls actually had a slightly lower CBF than all of the patients.

As usual, you have summarised the case for the prosecution very well @Hutan. This looks to me like a straightforward null result. I find it very hard to see how this low brain perfusion idea could ever really work clinically. It fits with simple intuition, as you say, but I don't think it fits with a closer look, either at clinical feasibility or evidence.

 

[ScoutB]

Some people are very convinced that this is true, and, to be fair, it feels as though it should be true, with our need to lie down, and the similarity of some of our symptoms with trying to function at high altitude. But, I don't think we can say that yet.

Agreed. I’ve been pondering the weird set of circumstantial evidence here. Eg many of us report doing better lying down, but it’s not like all our problems are solved by lying down. Studies show at best modest differences in cerebral blood flow, at least so far. And there are some reports of other things that may have to do with blood flow/pressure changes, like compression clothing, helping some people.

It makes me wonder if we just have a problem that is distinct from, but exacerbated by lower cerebral blood flow (or pressure, as when standing).

Eg maybe something’s iffy with neurovascular coupling (nvc) for us and the constraint problem nvc is trying to solve is easier to solve lying down than standing up. (I’m not sure we even know what nvc is doing entirely, from reading the papers SNT posted it sounds like it could be something more subtle than just providing oxygen to firing neurons).

Or maybe regulation of the pituitary by the hypothalamus via blood flow through the hypophyseal portal system has slightly different properties when vertical vs horizontal.

(Apologies if I’ve written this before, very foggy today.)

 

[MrMagoo]

Yes I do. I have reclining chair in my office. I bought a monitor for using at my desk, but I’m considering figuring out a set up where I can have the monitor in front of my reclining chair too.

I think I got too confident in my ability to spend time upright from my travelling this summer where I was able to do it more. School is definitely not the same exertion level as a fun conversation at the dinner table.

On the topic of my original question, I realized while talking with my fiancée that they were likely laying down for this study as I don’t believe any upright MRIs exist. I wonder what tool could be used to measure cerebral blood flow in different positions (sitting and laying down). I suspect most pwME/LC would experience cognitive fatigue faster sitting upright, and it’d be interesting to see if that’s correlated with lower cerebral blood flow.

I use an iPad on a frame, with a Bluetooth mouse. I guess if I needed to type I could use an old iPad on the screen holder and fix the livestream iPad somewhere else. This sort of thing. I love it.

Amazon.co.uk

 

[MrMagoo]

Agreed. I’ve been pondering the weird set of circumstantial evidence here. Eg many of us report doing better lying down, but it’s not like all our problems are solved by lying down. Studies show at best modest differences in cerebral blood flow, at least so far. And there are some reports of other things that may have to do with blood flow/pressure changes, like compression clothing, helping some people.

It makes me wonder if we just have a problem that is distinct from, but exacerbated by lower cerebral blood flow (or pressure, as when standing).

Eg maybe something’s iffy with neurovascular coupling (nvc) for us and the constraint problem nvc is trying to solve is easier to solve lying down than standing up. (I’m not sure we even know what nvc is doing entirely, from reading the papers SNT posted it sounds like it could be something more subtle than just providing oxygen to firing neurons).

Or maybe regulation of the pituitary by the hypothalamus via blood flow through the hypophyseal portal system has slightly different properties when vertical vs horizontal.

(Apologies if I’ve written this before, very foggy today.)

What happens if we go upside down?
I’m absolutely desperate to get to the swimming pool soon, so happy to experiment!

 

[ScoutB]

I’m absolutely desperate to get to the swimming pool soon, so happy to experiment!

Looking forward to your report

 

[Sean]

I would love to have a small swimming pool handy, just to able to float in it for a while several times a day.