Mapping cerebral blood flow in [ME/CFS] and orthostatic intolerance: insights from a systematic review, 2025, Christopoulos, Armstrong et al

[Nightsong]

Abstract:

Background​

Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) is a complex and debilitating condition with a large proportion of patients that experience orthostatic intolerance (OI). This systematic review aimed to assess whether cerebral blood flow (CBF) is reduced in ME/CFS and OI, and whether the presence of both conditions leads to an additional decline in CBF.

Methods​

PubMed (from 1943), MEDLINE (from 1946), EMBASE (from 1947) and Cochrane were searched from inception to February 14th, 2025, using terms including “chronic fatigue syndrome”, “myalgic encephalomyelitis”, “orthostatic intolerance” and “cerebral blood flow”. Article selection required the following criteria: published in English; CBF measured in participants with either ME/CFS or OI, or both ME/CFS and OI combined. Quality assessment and risk of bias was assessed using the Newcastle–Ottawa Scale and the systematic review was conducted in accordance with the PRISMA 2020 guidelines.

Results​

Of 14,928 articles, 118 were included, 26 (22.1%) of which studied CBF in ME/CFS alone, 81 (68.6%) in OI alone and 11 (9.3%) in both ME/CFS and OI. Overall, the articles included 9185 participants, with a mean age of 39.1 years (SD = 8.8), and 73.8% of participants were female. Studies found CBF was significantly reduced in 12 of the articles focused on ME/CFS and in 56 of those focused on OI; compared to controls. Additionally, in 4 out of 11 studies that examined both conditions, CBF was further reduced in participants suffering from both conditions compared to those with ME/CFS alone.

Conclusions​

CBF is reduced in ME/CFS and OI alone and having both conditions comorbidly amplifies CBF reductions. Therefore, observing CBF changes in ME/CFS with and without OI may be important in monitoring disease severity. Despite this, few studies focus on the combination of ME/CFS and OI, and OI may be a confounding factor in CBF in a large portion of ME/CFS studies.

Mapping cerebral blood flow in myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) and orthostatic intolerance: insights from a systematic review - Journal of Translational Medicine

Background Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) is a complex and debilitating condition with a large proportion of patients that experience orthostatic intolerance (OI). This systematic review aimed to assess whether cerebral blood flow (CBF) is reduced in ME/CFS and OI...

translational-medicine.biomedcentral.com

 

[Yann04]

Interesting. 0 Hits for deconditioning within the manuscript. From my perspective cause and effect could be quite muddled here. I imagine a longitudinal study could help but it would have to be quite expensive to get the required sample.

 

[Utsikt]

Strengths and limitations​

This review includes a large number of articles and is the first review to our knowledge that looks at CBF in ME/CFS patients with and without OI. However, due to the different techniques utilised to measure CBF and not having enough articles that employed the same technique in a similar population, we were not able to perform a meta-analysis. Furthermore, factors such as medications, time of the day or hydration levels will likely influence CBF which were not controlled for in the majority of the studies. Additionally, heterogeneity between the populations and types of OI studied in the included articles may limit the generalisability of the results.

Agree with Yann here. I think the conclusion goes beyond the evidence.

We need better data, and more control over the confounders.

 

[forestglip]

I think the conclusion goes beyond the evidence.

Which part of the conclusion goes beyond the evidence? It seems fairly reserved to me.

They saw many studies that found reduced CBF in ME/CFS. They said CBF is reduced in ME/CFS.

The reason for the CBF reduction (deconditioning, medications, ME/CFS pathophysiology, etc) is a different question that they may not have been able to answer with the studies they looked at.

 

[Utsikt]

Which part of the conclusion goes beyond the evidence? It seems fairly reserved to me.

Conclusions​

CBF is reduced in patients with ME/CFS or OI alone or having both of these conditions as comorbidities significantly reduces CBF even further when compared to having only one condition.

We don’t know if CBF is reduced in ME/CFS patients, or just the ME/CFS patients with OI.

I have not checked the included studies, but one thing that I would be cautious about is selection bias - maybe they’ve mostly included that pwME/CFS that have OI issues?

I’m also not sure if we can say that reduced CBF happens in every patient, or just in some. If it’s just some, it might be less relevant for figuring out ME/CFS in general, but they go on to say this:

Therefore, monitoring CBF in ME/CFS patients with and without OI may be important in monitoring disease severity and predicting PEM events.

This is pure speculation - we have no idea if CBF is linked to severity or PEM events. And it’s premature to talk about it before we’ve even established the basic facts yet.

Understanding how both ME/CFS and OI affect CBF is crucial for improving time to diagnosis, development of targeted therapies and enhancing patient quality of life.

Why is it «crucial» for improving time to diagnosis? They should also have lead with «understanding if and how»

The conclusion also makes no mention of the fact that they were unable to conduct a meta analysis as planned due to the heterogeneity of the studies.

It’s not the most egregious example I’ve seen, but I’m just tired of the slight inaccuracies and speculations that most likely were included because the authors need to make their work (and maybe especially future project) seem relevant in the eyes of funders.

They could have just said something along the lines of:

The literature is too heterogenous for detailed analysis, but more often than not finds reduced CBF measurements in pwME/CFS and OI. More rigorous research is needed to determine if reduced CBF is a feature in ME/CFS and/or OI, and if and how it’s linked to other clinical features of the conditions.​

 

[MelbME]

We don’t know if CBF is reduced in ME/CFS patients, or just the ME/CFS patients with OI.

I have not checked the included studies, but one thing that I would be cautious about is selection bias - maybe they’ve mostly included that pwME/CFS that have OI issues?

I’m also not sure if we can say that reduced CBF happens in every patient, or just in some. If it’s just some, it might be less relevant for figuring out ME/CFS in general, but they go on to say this:

This is pure speculation - we have no idea if CBF is linked to severity or PEM events. And it’s premature to talk about it before we’ve even established the basic facts yet.

Why is it «crucial» for improving time to diagnosis? They should also have lead with «understanding if and how»

The conclusion also makes no mention of the fact that they were unable to conduct a meta analysis as planned due to the heterogeneity of the studies.

It’s not the most egregious example I’ve seen, but I’m just tired of the slight inaccuracies and speculations that most likely were included because the authors need to make their work (and maybe especially future project) seem relevant in the eyes of funders.

They could have just said something along the lines of:

The literature is too heterogenous for detailed analysis, but more often than not finds reduced CBF measurements in pwME/CFS and OI. More rigorous research is needed to determine if reduced CBF is a feature in ME/CFS and/or OI, and if and how it’s linked to other clinical features of the conditions.​

I think you've got a bias against our research group for reasons unknown to me but perhaps we can discuss it sometime? (Correct me if wrong)
These sentences are all just mild even when you pull them out of context and dissect them as you have. The sentence you add at the end is a rewriting of what we wrote.

You suggest the conclusion wording in a systematic review of other peoples work (that isn't even critical of the work) makes our work more appealing to funders. I'd like to hear your logic here, how do you imagine this helps us get grants? I've never received any grant off the back of what was written in a paper. Have you received grant funding before because of words in a paper? I've never seen this on any grant panel I've been on either.

 

[MelbME]

Interesting. 0 Hits for deconditioning within the manuscript. From my perspective cause and effect could be quite muddled here. I imagine a longitudinal study could help but it would have to be quite expensive to get the required sample.

Yes very little use of deconditioned controls in the literature.
Also, I think it was noteworthy that there were better quality CBF measures and papers on OI than ME/CFS.

 

[SNT Gatchaman]

Thanks Chris and team, I thought this paper was really useful. I may be biased from my personal subjective experiences and a few objective observations, but I think cerebral blood flow is a very important part of the pathology to pin down. To me it seems the obvious answer to the question about precisely why pwME need to lie down, but we need to prove it.

I appreciated the extensive summary tables, detailing the key findings from the reviewed papers.

The points made in the discussion are very relevant to the limitations of the field to date. If you "have ME/CFS" you might get at most a CBF study that is passive and supine. It might be more accurate in terms of assessing global cerebral blood flow, but it misses any data from the orthostatic challenge. If you "have OI" you are more likely to get a tilt study, which will evaluate regional CBF eg transcranial Doppler (edit: occasionally extracranial) or NIRS , so while the effects of the orthostatic challenge come to the fore, we may be missing what's happening in all the brain.

If only we had access to upright MRI (ASL), we would likely uncover a whole lot. And as you point out, we're missing all the severe patients who simply can't tolerate an orthostatic challenge.

The other interesting point was the couple of papers that showed pwME had increased CBF supine compared to HCs. Is that a compensation to pay back some of the brain's "metabolic debt" incurred when upright?

 

[EndME]

I appreciated the extensive summary tables, detailing the key findings from the reviewed papers.

Yes, I wanted to emphasise that I find table 2 particularly handy. The amount of ME/CFS studies for which there is no protocol available seems quite damning when every single OI study has an available protocol. I was wandering whether a very short summary of the study results in the table would also be possible, or whether too much context would be lost if the summary was too short (and if it's too long the table becomes unhandy).

If only we had access to upright MRI (ASL), we would likely uncover a whole lot. And as you point out, we're missing all the severe patients who simply can't tolerate an orthostatic challenge.

Is upright NIRS for CBF going to deliver less reliable results, respectively isn't standadized enough? Because there's already quite a few groups using NIRS as part of exercise studies (Wüst for example), so also getting some extra measurements from the brain shouldn't require too many additional steps. Maybe the Patrick-Neary study who did a maximal cycling test with NIRS is worth a closer look.

 

[MelbME]

Thanks Chris and team, I thought this paper was really useful. I may be biased from my personal subjective experiences and a few objective observations, but I think cerebral blood flow is a very important part of the pathology to pin down. To me it seems the obvious answer to the question about precisely why pwME need to lie down, but we need to prove it.

I appreciated the extensive summary tables, detailing the key findings from the reviewed papers.

The points made in the discussion are very relevant to the limitations of the field to date. If you "have ME/CFS" you might get at most a CBF study that is passive and supine. It might be more accurate in terms of assessing global cerebral blood flow, but it misses any data from the orthostatic challenge. If you "have OI" you are more likely to get a tilt study, which will evaluate regional CBF eg transcranial or extracranial Doppler or NIRS, so while the effects of the orthostatic challenge come to the fore, we may be missing what's happening in all the brain.

If only we had access to upright MRI (ASL), we would likely uncover a whole lot. And as you point out, we're missing all the severe patients who simply can't tolerate an orthostatic challenge.

The other interesting point was the couple of papers that showed pwME had increased CBF supine compared to HCs. Is that a compensation to pay back some of the brain's "metabolic debt" incurred when upright?

Yeah the upright ASL is something we tried to find locally but none of the upright MRIs I could find had a strong enough magnet to conduct ASL. It's a real limitation in the field. We actually are doing a PET study using FDG tracer and getting ME/CFS patients to do a standing test while they receive a the tracer. The FDG makes it's way in to cells but isn't digested, so if you give it while standing then you get an indication of blood flow when standing. This seems like the most detailed method to do imaging of orthostatic impact on brain blood flow as it will also allow us to look at regional differences too.

Those supine studies were a mild difference, we debated about whether we give them an increase or no change. We would have liked to do a meta analysis but it really wasn't doable on the ME/CFS data, perhaps worth doing on the OI data only.

 

[MelbME]

Yes, I wanted to emphasise that I find table 2 particularly handy. The amount of ME/CFS studies for which there is no protocol available seems quite damning when every single OI study has an available protocol. I was wandering whether a very short summary of the study results in the table would also be possible, or whether too much context would be lost if the summary was too short (and if it's too long the table becomes unhandy).


Is upright NIRS for CBF going to deliver less reliable results, respectively isn't standadized enough? Because there's already quite a few groups using NIRS as part of exercise studies (Wüst for example), so also getting some extra measurements from the brain shouldn't require too many additional steps. Maybe the Patrick-Neary study who did a maximal cycling test with NIRS is worth a closer look.

NIRS devices are difficult to use because a lot of data noise is generated from head movement. I don't know how they managed NIRS with maximum cycling, perhaps newer devices have resolved this problem.

 

[Utsikt]

I think you've got a bias against our research group for reasons unknown to me but perhaps we can discuss it sometime? (Correct me if wrong)

I don’t think I have a bias against your work, but that might be my bias speaking so I’m probably not the right person to judge that.

I’m frustrated in general by papers that I perceive to go beyond the evidence, and I express that in many of those paper’s threads. I don’t expect you to know that - you’ve got more important things to do than to keep track of my activity on the forum.

These sentences are all just mild even when you pull them out of context and dissect them as you have. The sentence you add at the end is a rewriting of what we wrote.

1. The conclusion does not mention that you were unable to perform the planned analysis - why not?
2. It makes no mention of the heterogeneity of the studies wrt to how CBF was measured, in which situations it was measured, the patients selection criteria and other potential sources of bias - why not?
3. Instead, it spends time on speculative leaps about a potential link to severity and PEM episodes - is that warranted yet?
4. It also appears to assume that there actually is a difference between pwME/CFS and HCs because the focus for the future is to find out why and how to treat it. That has not been established yet.

You suggest the conclusion wording in a systematic review of other peoples work (that isn't even critical of the work) makes our work more appealing to funders. I'd like to hear your logic here, how do you imagine this helps us get grants? I've never received any grant off the back of what was written in a paper. Have you received grant funding before because of words in a paper? I've never seen this on any grant panel I've been on either.

I’m assuming your group has written this review for a reason - maybe as a basis for future work? Why else would you go through the effort of reviewing this area?

The reason I suggested funding is because I can’t see any other reason for putting that kind of spin in the conclusion.

There is so much low quality ME/CFS research, and our chances of figuring out what’s going on plummet if we’re not willing to throw away most of it and acknowledge that we know very little about pretty much everything.

 

[MelbME]

I don’t think I have a bias against your work, but that might be my bias speaking so I’m probably not the right person to judge that.

I’m frustrated in general by papers that I perceive to go beyond the evidence, and I express that in many of those paper’s threads. I don’t expect you to know that - you’ve got more important things to do than to keep track of my activity on the forum.

1. The conclusion does not mention that you were unable to perform the planned analysis - why not?
2. It makes no mention of the heterogeneity of the studies wrt to how CBF was measured, in which situations it was measured, the patients selection criteria and other potential sources of bias - why not?
3. Instead, it spends time on speculative leaps about a potential link to severity and PEM episodes - is that warranted yet?
4. It also appears to assume that there actually is a difference between pwME/CFS and HCs because the focus for the future is to find out why and how to treat it. That has not been established yet.

I’m assuming your group has written this review for a reason - maybe as a basis for future work? Why else would you go through the effort of reviewing this area?

The reason I suggested funding is because I can’t see any other reason for putting that kind of spin in the conclusion.

There is so much low quality ME/CFS research, and our chances of figuring out what’s going on plummet if we’re not willing to throw away most of it and acknowledge that we know very little about pretty much everything.

We have an interest in cerebral blood flow and already have a project ongoing that doesn't require funding. We aren't sure if CBF is of value yet but I find it a good exercise to get our PhD students to put together a review paper of some form early on in their PhD. I think this systematic revealed that there were in fact far more poor quality studies in ME/CFS than OI for CBF. I think we agree there are plenty of poor studies but also we don't want to throw everything out, sometimes a collection of low powered studies can add up to something valuable if they show the same trend. This systematic review is an example of that.

1. The conclusion does not mention that you were unable to perform the planned analysis - why not?
   We'd already mentioned it in the paper, we summarised main points in conclusion, didn't fit a full recap of study there.
   
2. It makes no mention of the heterogeneity of the studies wrt to how CBF was measured, in which situations it was measured, the patients selection criteria and other potential sources of bias - why not?
   That was shown in the tables 2 and 3.
   
   
3. Instead, it spends time on speculative leaps about a potential link to severity and PEM episodes - is that warranted yet?
   We see PEM as central to ME/CFS, so if this is to be a central factor then we'd expect it to be related to PEM. We weren't suggesting it is but asking the question if PEM and OI have any relationship.
   
   
4. It also appears to assume that there actually is a difference between pwME/CFS and HCs because the focus for the future is to find out why and how to treat it. That has not been established yet.
   Well there are a few studies that show a definite difference between pwME/CFS and HCs in terms of CBF. Do you mean pwME/CFS as distinct from pwOI?

 

[ME/CFS Science Blog]

Thanks for this review. It must have been an enormous task: screening 11,218 papers and 367 full-texts!

What I'm missing though is some focus on the effect size: how large is the reduction in cerebral blood flow, what is the variation among patients, how big is the overlap with healthy controls. Now the study mostly counts the number of studies that reported a significant difference or not.

Most studies have a very small sample size (< 30 ME/CFS participants) so the lack of significant difference might have been due this low sample size.

It would also be useful to know if the size of CBF reduction in ME/CFS, has been found in other conditions (including severe deconditioning) as well. Perhaps I missed this but couldn't find this in the paper.

 

[Utsikt]

I think we agree there are plenty of poor studies but also we don't want to throw everything out, sometimes a collection of low powered studies can add up to something valuable if they show the same trend.

They are not just low powered, they also have high heterogeneity. And bias (like recruitment bias) can create the appearance of trends that are not there.

The conclusion does not mention that you were unable to perform the planned analysis - why not?
We'd already mentioned it in the paper, we summarised main points in conclusion, didn't fit a full recap of study there.

Wasn’t the main point of the study to do a meta analysis? I don’t understand how that isn’t important enough to mention. Same with the quality of the studies.

It makes no mention of the heterogeneity of the studies wrt to how CBF was measured, in which situations it was measured, the patients selection criteria and other potential sources of bias - why not?
That was shown in the tables 2 and 3.

But not included in the abstract or the conclusion.

Instead, it spends time on speculative leaps about a potential link to severity and PEM episodes - is that warranted yet?
We see PEM as central to ME/CFS, so if this is to be a central factor then we'd expect it to be related to PEM. We weren't suggesting it is but asking the question if PEM and OI have any relationship.

The conclusion said:

Therefore, monitoring CBF in ME/CFS patients (…) may be important in (…) predicting PEM events.

That is not «asking if it’s related». It’s saying «we think it may be important because of a difference in CBF we think may be there».

I understand that this is nitpicking, but the details matters.

It also appears to assume that there actually is a difference between pwME/CFS and HCs because the focus for the future is to find out why and how to treat it. That has not been established yet.
Well there are a few studies that show a definite difference between pwME/CFS and HCs in terms of CBF. Do you mean pwME/CFS as distinct from pwOI?

Both.

I thought the main finding in this review was that we don’t know enough to say that there is a definitive difference between any of the groups, due to e.g. small studies, potential bias, and heterogeneity in selection criteria and measurement methods and protocols.

 

[ME/CFS Science Blog]

There were more ME/CFS papers on this then I expected (26 in total for the ME/CFS only category)!

The sample size for ME/CFS participants, however, was smaller than 50 in all but three: an old SPECT study from 1992 (Ichise) and then Van Campen/Visser studies.

I suspect that the latter two will have quite a large overlap between the participants in both studies. Also didn't realise that their method extracranial doppler ultrasound hasn't been done by other ME/CFS studies.

If I understand correctly extracranial doppler focuses on the arteries in the neck rather than the head itself as with transcranial doppler. The latter seems like the default method in OI studies.