Urine Metabolomics Exposes Anomalous Recovery after Maximal Exertion in Female ME/CFS Patients 2023, Glass, Hanson et al

The apparently normal baseline urine metabolite data makes me* think that an abnormality with the process of excreting the metabolites may be a more likely explanation for these results. If that were the case, presumably that would have have other knock-on or downstream effects.

[* Edit: as a layperson who has a very limited understanding of the science.]

I note that the paper concludes:

Apologies if this has already been highlighted (I’ve not managed to read the whole of this thread yet) but another interesting part of the paper which caught my eye was this (my bold):

This study and the recent studies on endothelial dysfunction, which appear to show abnormal NO production in response to exertion, have given me renewed hope that real progress may be starting to be made in understanding some of the mechanism. Am I being too optimistic? (If I am, please be gentle – I need every bit of hope I can get at the moment.)

 

I am also drawing deep from that hopium pipe.

 

They are uraemic toxins, so much of the literature seems to relate to chronic kidney disease and especially in patients having dialysis. In particular, they are being looked at in relation to the cardiovascular disease associated with renal failure. They have also been considered as markers of endothelial dysfunction in (diabetes-related) cardiovascular disease and in pre-eclampsia (also transplant medicine). From the introduction to Asymmetric and Symmetric Dimethylarginines as Renal Function Parameters in Paediatric Kidney Diseases: A Literature Review from 2003 to 2022 (2022, Children) —

Same thought. Also, just to note the physiology of the DMAs in healthy people. From Urinary Dimethylamine (DMA) and Its Precursor Asymmetric Dimethylarginine (ADMA) in Clinical Medicine, in the Context of Nitric Oxide (NO) and Beyond (2020, J Clin Med)

Assuming replication and confirmation, tempting to crystal-ball gaze and wonder whether there could even be an early morning urine test that discriminates HCs from the (potentially unique??) ME symptom of "unrefreshing sleep".

 

Ha! Careful, man. The come-down can be pretty heavy so I’m not drawing too deep – just enough to keep me going until the next hit.

This reminds me of a poem I wrote in 1999, when I had been told that I was going to be given Ampligen after some dodgy tests had apparently shown me to be the second most suitable candidate they had ever tested:

Let sleeping dogs lie

And now it seems there may be some hope;
A glimmer of hope, at least.
A chance to escape my Ground Hog days?
The end in sight?

No, I cannot think that.
For where hope hums his hypnotic tune, the dog of disillusion also lies,
Asleep, one ear to the floor, in anticipation of my sanguine steps.

No, I shall leave him slumbered.
He need not be stirred.
If the key fits I shall creep in through the back door and catch him unawares.

 

And if not replicated, there is a potential more direct link between energy and vascular control disruption in @Andy's post Mitochondrial ATP Production is Required for Endothelial Cell Control of Vascular Tone (2023). I guess the two could be co-occurring and augmenting the vascular problem. Maybe this is one of the ways there is symptom variability in the ME - POTS - FM spectrum.

 

As another lay person (and one who hasn't been able to read the whole paper yet) excretion problems make sense to me for the post-exercise findings. 24 hours is a common time point for many of us for becoming PEM-symptomatic which could be a reflection of a build-up of noxious substances overwhelming the system. This could be tested by a replication study looking at more time points after exercise.

It still wouldn't explain our seemingly normal baseline though. I guess it's possible that molecules Metabolon isn't testing for are at play, as mentioned by Strategist.

This raises the question, are there methods which can detect the sort of small molecules Metabolon misses?

 

I just had the data from machine learning analysis, given the latest findings by Maureen Hanson. Kidneys seem to be relevant as well. There were some signals about this previously :

https://www.s4me.info/threads/machine-learning-assisted-research-on-me-cfs.5015/page-7#post-207541

https://www.s4me.info/threads/lever...-with-cfs-2019-morris-et-al.8878/#post-205004

The system identified back then 'ckd' which is chronic kidney disease. How kidneys fit the picture, I have no idea unfortunately.

 

Did anybody catch if the exercise event durations for each individual or clinical group were included in the paper? Can't see it at the moment. Approx 8-10 min seems to be the typical length of the exertion period but would be good to know what the difference in mechanical performance during the exercise was between the groups, if any.

 

Not absolutely clear, but I think they confirmed adequate exertion, even if the absolute amount of time or mechanical effort would have varied. It reads to me as if the two studies are using data from one experimental run-through, which for the first paper used data from two CPETs but for this paper just from the first CPET. However, the first paper said 6-14 minutes while this paper said 8-10 minutes (after 3 minutes of rest), so maybe it was a completely separate data acquisition.

From the prior paper [ref 25] —

 

This figure got a lot of attention because of the stark differences between the two graphs. But as someone pointed out to me, these are sample averages for different metabolites which make it easier to have sharp contrasts without overlap between patients and controls.

These aren't the datapoints of patients and controls for one particular datapoint as we normally see for potential biomarkers. Such data is presented further up in the paper (e.g. figure 5 and figure 6) and do show substantial overlap between patients and controls, even though there were only 10 patients and 8 controls.

It is interesting that patients and controls show a different response post-exercise but I suspect the sharp contrast between the graphs is largely due to how the data is presented (namely the result of modelling for multiple metabolites, rather than datapoints of participants).

 

In a way I see this as a positive. My biggest concern on initial viewing was that the results looked too good to be true. Understanding that the type of modelling they used makes the contrast more stark gives me more confidence in the validity of the results.

 

Most valuable player? From sports?

 

Mean platelet volume?

 

I have suspected for maybe two decades and a half that our disturbed sleep might be associated with altered circadian patterns of many metabolites. This was my concern in the late 90s when early morning urinary metabolites were first being looked at. Change the time, or have patients with altered circadian patterns, and you might get different results. We cannot yet be sure that our results will be stable over the course of a sleep-wake cycle. This needs to be established, or an optimal time of day or sleep-wake cycle needs to be determined.

[Edited to correct how long its been.]