• Welcome to Phoenix Rising!

    Created in 2008, Phoenix Rising is the largest and oldest forum dedicated to furthering the understanding of, and finding treatments for, complex chronic illnesses such as chronic fatigue syndrome (ME/CFS), fibromyalgia, long COVID, postural orthostatic tachycardia syndrome (POTS), mast cell activation syndrome (MCAS), and allied diseases.

    To become a member, simply click the Register button at the top right.

Chris Armstrong's presentation for Solve ME/CFS October 20 2016 (metabolomics)

Good grief.....you mean ALL your research over all the years?

Again many thanks for your replies to queries posted to date. I'm in the midst of a real rough patch this past week or so, but knowing people like yourself are working on our behalf to find answers to this awful disease sure helps; such a contrast to the continuing shenanigans and influence of the BPS adherents here in the UK.

I hope that you can get the funding required to expand on your research, and would be happy to contribute the little I can, as and when you're able to set anything up. :)

@ChrisArmstrong - thanks for your and your group's work and for taking the time to engage with the PR community. I'm someone whose only abnormal test result was a very high level of serum glutamine "approaching that of ammonemia" as the lab noted, and of course no one could tell me what this meant. Given the close relationship between glutamate and glutamine, and the starring role for glutamate in your paper, I have hope that you guys are on the track of something that may ultimately prove illuminating for my case---as well as for many others. Keep us posted as to where to send funding donations.

Thank you all, we are putting together a profile online with the ability to give donations. Hoping to get it out in next couple weeks. I will write blogs and update on our research and research in the field as it comes.
I like this line of research Chris but I am a bit skeptical about the starvation hypothesis, because while it would explain the long term sickness we experience it doesn't seem to provide an explanation for PEM, particularly when it is occuring within the first month of the illness, presumably before such a state of severe deficiency would set in.

In the onset of mine it happened basically overnight, then I recovered enough to exercise within about 5 days, but then the PEM crash cycle started, went to the doctor within a fortnight and they said 'post viral fatigue' rather than a sustained infection or a 'sepsis', blood testing done within 4 weeks found no signs of a known infection or significant change in immune cell counts.

So starvation is being used as a model to fit the metabolomics data, it's not what I'm saying the disorder is, just that the metabolomics data points to a relationship and that it may be a factor of symptom expression. A completed hypothesis is still being worked upon but we will release it at some stage in the coming 6 months.

Well we all experience a bacteremia when we exercise. The type of bacteria that enter your bloodstream are usually quite controllable by your immune system but if your gut is further compromised they may release more bacteria into your blood or more pathogenic species or your immune system may already be depleted. This is the concept for the chronic sepsis or SIRS and this is what I think may be behind PEM.

We only ever view people once the maintaining factors of the disorder are set in, yet to study people at the beginning of infection or around the trigger. Indications from people suggest the onset period is different from the maintained period and we suspect there could be a number of different triggers that may lead to the disorder, which has commonality in its maintaining factors.
Thank you so much already @ChrisArmstrong for answering so many questions here (and on other threads), following on from your brave webinar appearance during the zombie hour of the morning, with a cold (man, you sounded so relieved to reach the end of that, lol ;):)).

I'll quickly mention that @Cort covered your work the other day, here on his Health Rising blog (although I'd guess you know that already).

I'd also like to say how much your paper's findings struck a chord with me personally. I mean, I know it's be laboured (above) that:...But aspects of the metabolic picture you found happened to tally well with my own serum amino acids results from 3 years hence (shown in expandable picture, below). Low essential AA across the board with that tell tale very high aspartic acid (from the use of AAs via glutamine as a fuel). A feature I'd not had any explanation for at all before. (Issues with sulphur pathway too, that you may of mentioned in passing.)
View attachment 18343
It's kind of funny how we patients have been able to get these test for years, but are still extracting possible meaning from them years after the results. And research seemingly only just able to look at things like this in a meaningful way. Kind of backwards, perhaps, from naive expectations of medicinal science.

I'm aware of how these should be taken with a spoon full of salt, in terms of their meaning (enigmatic as they tend to be already) and the isolated nutritional recommendations the commercial ones carry, too. Especially as they will presumably be calibrated on healthy individuals.

Now for some fairly big, open ended questions...:

(1) What trends do you see in the use and effectiveness of the different metabolomic testing technologies (which might we expect to see or have used on us in future)?

My Genova test, for example, will have been done using High-performance liquid chromatography (HPLC), I believe. Which I think essentially separates different sized molecules mechanically, as they move through an adsorbent material at different rates (if I understand correctly). Presumably this produces a kind of bar-code like diffusion pattern that can be read-off, as with those old style genetic tests... Anyway, how does this (older?) technology compare in accuracy and cost to optical and NMR spectroscopy?

You say that NMR is currently much cheaper than MS (mass spectroscopy), but is it likely to scale down in cost much? The setup involves a massive superconducting magnetic chamber, right? Is it the case that whiles that's pretty expensive in itself, the sample measurements are quick and clean (and you just share time on one)? (Sorry, I'm so vague on this, ironic since I did a physics degree, some years ago, at the University of Nottingham, with a tutor based in the Peter Mansfield NMR centre, heh.:oops:)
View attachment 18344 < From your webinar video (NMR test equipment used?).
I get the impression that optical spectroscopy technologies are really improving rapidly in price-performance and miniaturisation. (Exponential technological advances are a particular interest of mine.) So even if it is more expensive than NMR right now, is it likely to be cheaper (and better in many ways), within the next decade, say? (With NMR reliant on less shrinkable equipment?)

(2) Is there some resource that gives some kind of easily digestible, qualitative (and quantitative) indication of how much and how rapidly each type of metabolite tends to vary (and under what conditions)?
View attachment 18345
Its cool that you highlighted this, since when we look at our own results there's no indication or feel for what what the dynamics are like there. The kind of talk I see about people's out of range results tend to be very linear in it's induction. There's often no appreciation for the relative magnitude of the quantities, even, with just a plain number quoted.:confused: Let alone thinking about whether different metabolites will take days or months to slide out of range, or how much they might spike at various times after eating (as you mention).

For example, my nutritionist, at the time, suggested that low serum AA levels across the board like that (it was a fasted morning draw, I think) would likely be the product of months of decline. Is that reasonable? How big might a protein meal AA peak be, by comparison? Fat soluble vitamins (D, K, A, E) are certainly very slow to change, right? On the opposite end of the scale, I was looking up adenosine, and that's stated to have a serum half-life of under 30 seconds!(?) It puzzled me how substances like that are even measured... (Is it all produced locally to the site of the blood draw? Do it's levels reduce in the sample, after draw, to an equilibrium set by enzymes in the blood? Etc.) Where-as, for one thing, people are generally familiar with blood glucose level variations, up and down within a day (from diabetics in trouble on medical dramas).

What I'd really like to see are demonstrative dynamic models of metabolite levels. Preferably interactive ones that you can poke with an input change, to get a feel the pliability/elasticity characteristics and knock effects through the various paths to other metabolites... Although, even a written data repository would be a step up from hit-and-miss Wikipedia pharmacokinetic numbers. Even the human metabolome database often doesn't seem to have the relevant info.

(3) Have you used computational simulation of human metabolism at all, or know much about others who have?

Some years ago I saw a there were various efforts aiming at ultimately creating a totally detailed whole body virtual modelling system of the interplay of metabolites, proteins, genes, etc, across the various compartments and tissue types of the body. Are such things visibly happening or having any impact yet, in your experience?

I envisage this kind of super-detailed simulation being used as in conjunction with data fed in from the various types of testing (and a Watson-like integrated AI expert system), as the new paradigm of medicine (2.0). One that's more preventative and holistic. But before then, I'm wondering if similar approches might get some early action being used for treating us lot, here. I mean, like you've been saying......It seems unlikely at this point that proper recovery from CFS/ME will be as straight forwards as a drug monotherapy (for most). Might it even be the case that targeting the most derange quantities in a patient, to bring them all up together, may not be effective? (Or a lot less effective than optimum.)

I mean, you talked about (metaphorically) picking up all (or enough) of the fence posts at once, to get it (metabolism) to stand back up. I'd previously thought about things more like an umbrella, blown inside out: the body has seemingly transitioned to a distinct, stable state. But similarly you need to apply force against many of the spokes simultaneously to revert it.

Could it be that, for some, it's going to take a fairly clever, dynamic, tailored set of interventions. Somewhat like an orbital insertion manoeuvre in space, where you can't just aim straight at where you want to go - it's a whole lot more slide-ways and unintuitive than that.

I mean, metabolism is a great big interwoven web of feedback loops, right? Such that I imagine the CFS/ME state as perhaps being an unusual orbit, around some 'strange attractor', if you are looking at what's going on in some abstract multi-dimensional phase space plot of important metrics. That when we are simple looking across things in one dimension, most of the quantities either fall into the normal area, or behave totally incomprehensibly, because we're not aware of the overall dimensions of what's happening.
View attachment 18348 <A strange attractor ploted - chaotic behavior can seem random, but will stay within certain bounds.

Or even that (stretching my space comparison) there may exist clever little adjustments to be made that are far cheaper, easier, less problematic, requiring perhaps only over-the-counter nutrients, that will almost miraculously fix things. If one can just figure out some weird way to apply them to the specific individual. In the solar system, it's been found that there is a dynamically changing network of routes between gravitational Lagrange points and such, requiring almost zero thrust to traverse between the orbits of any planet (but a lot of time and finesse).

(4) In your opinion, could it be possible that the X-factor in patient's blood, which Fluge-Mella/Davison say causes healthy cells to konk-out, might not (always) be a distinct factor?
(Like an auto-antibody.) Could it be a profile of the metabolites and factors themselves? That the cell (or mitochondria) is integrating across all these signals it's sensing, not requiring any single one to always be present. (Sorry, rampant half-arsed speculation. I'm sure we'll find out something concrete from them fairly soon.)

@Freddd, in his B-12 protocol [1,2,many other places (I get lost, sorry)], has talked about a 'deadlock quartet', with regards to his supplement treatment protocol, and many complexities with getting the various central wheels of metabolism (particularly methylation) turning properly again. Writing about paradoxically induced 'doughnut hole' folate deficiency, for example, as a very counter-intuitive, emergent problem, that appeals to my conception of chaotic orbits. I see now that he's previously mentioned such things explicitly with regard to 'refeeding syndrome' (e.g. exhausting potassium, also troubles with interrupting the body's nutrient triage system with too small doses of folate, I think).

Also Amy Yasko's work, Rich Van Konynenburg's offshoot of that, etc, I guess they've helped capture my imagination down this route. Although, playing around with B vitamins and such has failed me a number of times.
Maybe this is some direct problem, like gut bacteria getting in the way, or it just not being appropriate for my biology. Frustrating, that there's so much promise there, if I could just see what was going on; trying to implement such complex protocols while unable to measure one's internal state can feel like trying to solve a Rubik cube blindfolded, having never seen one before. No feedback aside from slow dawning realisation that things have been getting gradually worst for weeks. (Since even the limited testing that is on offer is too expensive and hard to access, here anyway.) Hence why talk of longitudinal testing/treatment sounds amazing.

Or do you think all this complexity that these clever people (mentioned above) have been exploring is likely to turn out to be somewhat of a Gordian knot? One that's about to be cut by discovery of some previously unseen or unconsidered factor... (Sorry, that's a bit of an impossible question.)

Sorry for all the prattle there...o_OI'd appreciate any of your thoughts on any aspect of the above. Having been accumulating thoughts on this for a couple of weeks (of not being able to get myself together enough to post) I've got rather too much in mind. So might make a second (big) follow up post with the rest too, including some more personal perspectives. That's if no one throws me off the forum for trying to waste the valuable resource of your time... :eek::oops: Oops.

1. The different techniques should be comparable, I like NMR the most because it's the most trustworthy, I don't have to worry about the reliability of the data I collect from it. NMR is less expensive because it takes 10 minutes to run a sample and minimal handling time. The machine itself is more expensive to run than MS but the labour cost is low. Unfortunately there are much fewer metabolites you can get from it. MS is a great technique and you get much more information, either of the techniques is fine. I'm sure some smaller technique will come along one day.

2. No resource looks at metabolite fluctuations yet. It's something very important to me and who know, maybe I'll try and put one together one day. I think the changes in levels due to stimulus are more important than the static levels and the fluctuations are always under published, it's hard to know how much the researcher considered them.

3. Yeah people are putting together webs, i think it needs to happen because of the complexity of metabolism. I think the metabolite changes and studies like that will help tremendously though as the static levels may be creating misrepresented association levels in the current models.

I am yet to use any computational simulations though. I apply information from reading papers and simulating this in my own head. Another severely underappreciated complexity of metabolism is that metabolites are actually signalling molecules as well.

4. You definitely have done your research, only names I recognise are Fluge/Mella, Davis and Konyenburg, will read others when I get some time. Yeah I think many of have noted that it's something in the blood for a while. Although I tend to think it's the metabolites, ions, cofactors, vitamins, minerals, gases. I will think that until someone does find something because we already observe this and it alone would cause a hostile environment for a cell to exist in.

As I state, we will release a well-documented hypothesis soon and maybe this answers more of the questions. I have a paper coming out quite soon in Metabolomics Journal so it will be nice to get more feedback then.

Feel free to post all your thoughts and I'll read through and get back to them when I can.


Senior Member
As I state, we will release a well-documented hypothesis soon and maybe this answers more of the questions. I have a paper coming out quite soon in Metabolomics Journal so it will be nice to get more feedback then.

Feel free to post all your thoughts and I'll read through and get back to them when I can.
I will write blogs and update on our research and research in the field as it comes.
Wow, that all sounds great! Thank you again for coming back to the forums, again, and for going over my long-winded questions. Very interesting to get a feel for context. :)

I'm not aware of any optical spectroscopy technologies, except possibly infrared, which would deliver the required detection limits but I haven't worked in a chemistry laboratory in decades so there may be something I've missed (I also missed the use of NMR).
Uuh! Terribly sorry for my misunderstanding... :confused::oops: For some reason I'd made a very dumb assumption, along the way, that MS was an optical technique. (Me being somewhat familiar with astronomical spectroscopy, that investigates the elemental make-up of star, etc, from splitting spectra.) But of course mass spectrometry (not spectroscopy) utilises the *mass* of molecules to identify them. Roughly speaking: by giving them each the same electric charge, then squirting them all through a deflecting (electric) field, measuring the resulting spacial distribution of the splatter, as the heavier ones are deflected less from their course by the same applied force. (MS-MS = Tandem Mass Spectrometry.) Sounds messy and fiddly as heck, no wonder there's difficulty duplicating measurements reliably...

It seems you're right, too, that I was imagining it was infrared (or Raman) spectroscopy. I think because I'd previously heard of a little hand-held gadget (start-up company) that identified food (or pills) via the IR light reflection (SCiO platform). It seems like that produces a lot less detailed readings, that might (at best) only be able to match up roughly to pre-recorded profiles in their database (although the company seems to have gone quiet since these articles).

There was word of a device for on the spot testing for food histamine content that I'd seen too, from Australia actually - Flinders University. Microfluidic chip, though, not sure of testing physics involved. Lab-on-a-chip seems like a likely disruptive technology for the coming decade (imagining cheap home testing devices like this, although for a broader spectrum of substances at once, or re-usable).

Anyway, thanks for straightening me out, @FMMM1, sorry I've not managed to get back to you until now. Also when you say that:
The key thing is to identify the diagnostic target(s) e.g. allantoin in urine (or possibly blood) or sphingolipids etc in blood;
I think that yes, that would be the traditional way of doing medicine, and it may well work here, for many, but I wonder if there will indeed turn out to be a small, definitive number of analytes for simple diagnostic. Or will there be no definitive markers, so that proper diagnosis generally requires having a more overarching picture (with some far more complex/intelligent discriminator system). I definitely do get ahead of myself, in thinking about all medicine being one day done holistically, preventively, based on broad, personal 'omics data. But then, also, society generally fails to see paradigm changes coming until they are already here...

But anyway, regarding creatin(in)e, as you say...
Creatinine levels in urine are normally used to correct for dilution; however, Chris has stated that creatinine cannot be used in this case (see Chris's webinar for an explanation).
Yes, @ChrisArmstrong says at 18:30 in the webinar video that they found that (normally reliable) creatinine levels seemed to be increased, in that they made all other metabolites look decreased, relatively.

(5) I wonder if this potentially throws into question all urine (metabolite) tests for pwCFS/ME? (It is just urine that's usually normalised this way?)

From Ben Lynch (quote from this page, with my emphasis):
if methylation is hindered, and it typically is in those with histamine intolerance, then creatine formation is likely hindered as well. Given that most of your methylation is used up to produce both phosphatidylcholine and creatine, wouldn’t it make sense to supplement with both of these in order to conserve methylation so it can then focus on reducing histamine?! Yes!
70% of your methylation is dedicated to making cell membranes
I relate this, having dietary histamine intolerance myself, and also seemingly some initial success addressing methylation SNPs (after improvements from dietary exclusions). So, I'd have expected creatine to be lower in pwCFS (and creatinine, or is the difference here very significant?). If there are generally methylation issues... Or maybe it's the other way around, that increased creatine production for energy/ATP creation is impacting on ability to make PC, and contributing to reduction in those related metabolites (sphingolipids down this pathway?)?

(6) So, creatine, methylation phosphatidylcholine...Is what Dr Lynch says possibly true/relevant? Also, could day to day changes in creatin(in)e levels explain some of the variability in (my) muscle exertion capacity?: On a poor day, if I'm cycling up the little hill to the nearby supermarket, I'll immediately just have no strength there at all, immediately reverting to walking speed right at the bottom. Whereas, on a good day, I'll carry some momentum further, or all, of the way up, being able to get a little out of breath, even. When I supplemented creatinine for a short time I felt it may have slightly helped with this muscle strength, or transmission of energy that last step. (Imagination?)

(7) What about the liver, in terms of PEM?

It seems kind of overlooked. We talk about accumulation of lactate in muscles (although you found it low in faecal samples), and brain changes, CSF levels, but what about the (cellular) effects on the liver, post exertion? It has a very active role in supplying energy for both these other tissue types, right? So does it get chemically stressed too? Is it just harder to measure directly? (Or is the complete opposite, that all metabolites (when at rest?) are a reflection of liver function?)

You talk about bacteria entering bloodstream through weakened gut lining during exercise, isn't it the liver that will bear the brunt of that, most directly receiving (and filtering) the blood supply from the gut, dealing with the LPS (lipopolysaccharides) too, etc?

I ask because I get 24 hour delayed food intolerance reactions (dairy, egg, yeast, histamine) - fatigue, weakness, brain fog, low alertness. Now, I assume that's because it takes that long for food to reach the weak section of my gut (end of colon). But could 'liver PEM' be a possible culprit, or contributor, too? If the cellular energy reserves, there, were quietly exhausted by dealing with additional digestion fallout of the previous day's problematic meal, then it might struggle to cope with the rigours of processing a regular meal the next day...?

(8) I found very interesting the talk of blood immune cells directly depleting metabolites (e.g. amino acids), since I guess I'd just assumed such nutrient usage relatively negligible. I had presumed that raised IgG (immunoglobulin) antibodies produced against my intolerance foods (as found in ELISA tests) were just indicators of concurrent release of cytokines, which trigger brain fog, etc. But could it be that the B-cells cause trouble merely by consuming (and depleting) the resources to produce these (and other molecules)?

It makes me wonder if Fluge & Mella's patients responding to B-cell depleting Rituximab would also benefit greatly from personally relevant dietary exclusions (to abate IgG production)...?

(9) Random little question: is it correct that glutathione is used by the body as a last ditch mechanism to absorb amino acids in the (lower) gut? If so, could reduced digestion/absorption capability in the upper gut cause depletion of glutathione (or it's substrates), by over-use for digestion? Or vice-versa: glutathione depletion (as per Rich Van K, ish) cause reduced amino acid absorption (and the lower levels observed in patients)?

(10) Do you have any idea why supplementing with any of the precursors for glutathione cause me to have a next day fatigue reaction (just like my food intolerances)? So, NAC or glutamine make a kind of sense (since I have excess sulphur metabolites and then glutamate is said to be excitatory-toxic), but glycine does this to me too (and that's a calming, generally well tolerated supplement. I'd guess most answers on forums would come down to 'detox'. Meh...

While taurine gives me amazing next day mental clarity, energy, (makes music sound amazing!)... But only one day, and then it makes me feel progressively worst. It's a calming neurotransmitter, but also very elevated in a past serum AA test (trans-sulfuration product). Anyway, sorry to deluge with anecdote here. Not expecting personalised medical advice, more wondering what level of insights you have with the ins and outs of these kinds of (neuro-)metabolic cogs...?

(11) Finally: you talk about the importance of trying to investigate patients right after getting sick, as opposed to the mostly well defined, long term ill in (your) studies. But what about gradual onset CFS/ME? Does that come into your thinking at all?

For myself (sorry, anecdote again) it's been a very slow, lifelong slide into a state that I'm only now (aged 33) entirely comfortable thinking of as a positive CFS diagnosis. In part because so much of the new research (including yours) seems to fit in increasingly well with my symptoms and results. I kind of wonder if mine is a more 'pure' form of CFS, in a way. I've pretty much always had sleep rhythm issues (now worst), but also it seems like maybe my ADHD-PI and dyslexia may be related to mitochondrial cell danger signalling (at a developmental stage), if Naviaux's related research is anything to go by. I don't know how much this indicates genetic susceptibility, but presumably it will come down largely to environment/gut too. While early viral infection couldn't be ruled out, even though there was no one event - everyone carries HPV, etc.

My guess is that CFS/ME may well be extremely under-diagnosed in other cases like mine, given the slow, indefinite onset and apparently pre-existing conditions. (It's taken a lot to get to this point, even with so much else in my favour.)
Last edited:


Senior Member
Sorry for my basic understandings, I have read some but not everything. Is it a common finding that glycosis is downregulated in CFS? I remember reading a study saying the mitochondira upregulated glycosis. Is it possible the the glycosis in the brown adipose tissue is impaired and the mitochondria are overcompensating for that impairment?
Well we all experience a bacteremia when we exercise. The type of bacteria that enter your bloodstream are usually quite controllable by your immune system but if your gut is further compromised they may release more bacteria into your blood or more pathogenic species or your immune system may already be depleted. This is the concept for the chronic sepsis or SIRS and this is what I think may be behind PEM.

This is fascinating, Chris, thank you so much for sharing your thoughts. In terms of bacteremia, how would this model fit with episodes of mental PEM (i.e. PEM caused by mental exertion alone)? When I was at my worst, I was able to trigger PEM just by reading for more than 4-5 minutes in the evening, or watching TV for more than 10 minutes. These PEM episodes were identical to those I experienced when I overexerted myself physically.