Understanding the IDO Metabolic Trap Hypothesis

Usually I blog at the end of each month but I’ve taken a mainly unintended break since May. My cognitive stamina hasn’t been so good and my general health hasn’t returned to January levels (I had a setback in February and then May).

My Good Pacing Intentions have been to deliberately cutback on expected activity. Over the last 17 years my approach to Pacing has been to pick up part time activities and responsibilities as I improve and then try to accept letting go when I can’t continue. Doing this I have previously managed to work full time hours, but I’ve never done it as one full time job (so a partial relapse doesn’t have to involve giving up everything). This summer I gave up my role as a researcher on the Chronic Illness Inclusion Project and (apart from contributing to NICE stuff in July) I haven’t been doing as much volunteering for #MEAction.

I’m no good at switching my mind off though. Yesterday I decided I would have a light cognitive day today of iPad games and trashy TV. Of course that meant I woke up early thinking about the Metabolic Trap Hypothesis. This could well be an example of the White Bear Phenomenon:

“Try to pose for yourself this task: not to think of a polar bear, and you will see that the cursed thing will come to mind every minute.” Dostoevsky

Polar bear drawing

Attempting to understand the IDO Metabolic Trap Hypothesis

I felt a sense of anticipation leading up to Robert Phair’s presentation on the Metabolic Trap Hypothesis. This was partly due to the mystery created by knowing this existed but not what it was about. The OMF team had felt that it could be irresponsible to share partial information in case people tried dangerous self hacking in response. That danger still applies, I’m not adverse to a bit of bio hacking experimentation, but the pathway described here shouldn’t be messed around with. Apparently someone died doing this.

You can watch this talk towards the end of this livestream


Thanks to @Gemini on Phoenix Rising these are rough timings

Raeka Aiyar–24:00
Oystein Fluge–35:00
Wenzhong Xiao–1:11:00
Jonas Bergquist–1:33:10
Alain Moreau–2:26:35
Maureen Hanson–2:49:12
Ron Tompkins–4:55:32
Michael Sikora–5:06:03
Jarred Younger–5:27:53
Ron Davis–6:24:49
Robert Phair–6:38:04
Ron Davis–7:11:00

How it works

The steps to getting to the Hypothesis are interesting. Robert Phair was struck by reports of ME outbreaks where up to a quarter of people can be triggered. There does seem to be a genetic element to ME but is it the wrong approach to look for rare variants?

He was also interested in the idea of bistability – when there are two alternative possible steady metabolic states but one is healthy and the other isn’t – a trigger can switch us into the unhealthy state and something else is required to switch us back.

While listening this provoked a few questions for me. What about varying severity steps up and down (additional metabolic traps?), or spontaneous remission? In my experience (and observing others) we do get stuck at a level for a while but it’s not on or off, there are various steps up and down. I do perceive it to be steps though, the fluctuations don’t feel gradual.

I asked a question about this at the end and Ron Davis said something about it seeming like infections can also trigger a reset. There’s only been once I’ve noticed improvement myself like this (I assumed the antibiotics had killed something more chronic in addition to the targeted infection). My steps down are usually associated virus symptoms which are convincingly a new infection, but this isn’t as clear for steps up and I’ve been at almost-recovered level before. People with ME do sometimes go into remission without knowing why.

So anyway, Robert Phair searched for damaging variants which aren’t rare, could be involved in bistability and the downstream repercussions would look like ME. What he hit upon was IDO. When he used a computer model to simulate the repercussions of problems with the IDO2 gene the downstream effects were a good fit for ME. This gene is involved in catabolising tryptophan and immune function.

But do pwme have relevant genetic variants? He found that all the Severe ME patients in their study had at least one of 5 variants (I’ll discuss this more in the next section). Also people like me who sent him our genetic sequencing informally did too.

But is it causing/expressing the predicted problems? They’ve only tested 6 pwme so far, but yes the results support the hypothesis.

P=0.008, 0.009, 0.006 – very unlikely that the difference between these pwme and controls is down to chance (though it might be a fluke that this applied to the first 6 pwme tested, perhaps not everyone with ME will show this ratio?).

“The IDO2 mutations (there are 5 of them) uncover a bistability that is built in to human metabolism.

IDO1 is subject to substrate inhibition — the enzymatic reaction stops when there is too much substrate (tryptophan). The substrate is competing with each other to get to the active site of the enzyme. So much fighting to get in, the enzyme just sits there waiting for something “to get in its belly!”… meanwhile, substrate level (tryptophan) continues to rise and rise.” Phair

This leads to too much intracellular tryptophan (and possibly serotonin and so on) and too little L-kynurenine. The kynurenine pathway is at the crossroads between immune and metabolic function.

In situations where IDO2 is functioning correctly it acts as a backup to catabolise tryptophan, preventing the problem, though IDO2 is much less efficient at this anyway – I’m unsure whether this is a serious problem for the Hypothesis. If tryptophan levels don’t hit a threshold where IDO1 inhibits enzymatic reaction then there wouldn’t be an obvious problem.

The trap does have to be sprung. You could live your whole life without triggering this problem. This element has seemed to get a bit lost in online debate so far. Just having the variants won’t cause ME (if the Hypothesis is correct), it is an epigenetic type of thing where a specific type of trigger is required.

In computer modelling it seems like kynurenine pathway function can return to normal after a stressor of 10 days but with a 8 week stressor it never returns to normal of its own accord. Not much was said about this in the presentation. I’m wondering if this might be connected to why some viruses seem to trigger ME and some don’t. Could it explain why infectious mononucleosis (glandular fever) is the classic ME trigger? That was the trigger for me and a close relative.

I’m generally in the camp of people who think there’s a genetic predisposition to ME which requires an environmental trigger (though the nature of the trigger seems quite broad).

However, what about anecdotal evidence that people who later go onto develop ME can have glandular fever without that appearing to trigger ME? They would have the IDO2 variants at that time. Perhaps these people process the initial EBV infection before 8 weeks ie they escape triggering the trap?

I’ve previously explored this on Science for ME:


(Also why aren’t some of us getting EBV acute infections after ME is triggered?)

Any decent theory of ME would need to account for sex differences. An observation from @FMMM1 on Phoenix Rising is

“Women appear to be importing tryptophan into their cells i.e. as an alternative fuel to glucose [based on https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5161229/ ].

Importing tryptophan into your cells would increase your intracellular levels; possibly enough tryptophan to trigger the trap”.

Could this perhaps account for more women getting ME? (NB the linked study is people who already have ME/CFS so this is pure speculation)

I didn’t know anything about IDO before the Symposium but from what I’ve read so far IDO2 has a distinct function from IDO1 (it’s not just a backup).


https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4432394/ bold mine

More about the genetics

My relevant variants:  

These are the Kynurenine metabolic pathway variants I have (more than just IDO2 eg KYNU variants are associated with Hydroxykynureninuria). I have two of the five highlighted variants.

My variants are heterozygous. Are these dominant?

“One C-T SNP affecting R248 in human IDO2 was structurally analogous to R231 in human IDO, which makes a critical contact with the indole ring of tryptophan ( 12). The nonsynonymous substitution (R248W) reduced catalytic activity 90% in T-REX cells (Supplementary Fig. S5). A second T-A SNP affecting Y359 generated a premature stop codon (Y359X), which completely abolished activity (Supplementary Fig. S5). Strikingly, both SNPs were commonly found in human genomic DNAs in public databases, with the C-T SNP being highly represented in individuals of European descent, the T-A SNP being highly represented individuals of Asian descent,”


So do they mean heterozygous CT for R248 leads to >90% reduction or do they mean a homozygous change from CC to TT causes this reduction?

Genetic Questions

The combination of variants which are damaging and conserved in mammals looks promising. I think having deleterious score of 3 (the 3 red dots in my results) means these variants are in the top 0.13% on DANN in terms of potentially harmful variants. The variants I have also have high mammalian conservation scores (6.16 is highest possible? The R248W one is 6.0). I read that IDO2 is conserved as distantly as turtles and some fish and genes which are conserved through the evolutionary process tend to useful.

If these variants are both potentially very harmful and highly conserved, and yet despite not being unusual they’re not yet officially associated with any clinical phenotype, it seems reasonable to assume a connection to a relatively common disease could so far been overlooked. Is that ME?

BUT 48% allele frequency for one seems high. Too high? Even outbreaks aren’t that high. There’s really high likelihood of having at least one IDO2 variant, for example in one study “When considering all gene variants, the likelihood of carrying at least one IDO2 SNP was similar between cohorts (81.3% non-IBD” controls https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4277413/

Perhaps it doesn’t matter – there is a significant difference between pwme and controls for R248W.

“On average the severely ill patients have 1.7 probably damaging mutations in IDO2. Every severely ill patient has at least one probably damaging mutation in IDO2.” Phair

But it doesn’t feel quite right as specific and full explanation of ME, it feels like the genetic aspect should be stronger based on the way relatives get ME. Is this part of the picture maybe? What happens if you also have other variants associated with autoimmunity or ATP, does that lower the threshold of strength of trigger for ME? Perhaps there are variants we shouldn’t expect all pwme to have, they’re not necessary but make it more likely?  

Perhaps IDO metabolic trap is required but not the only genetic factor?

In a sense it’s almost like the IDO2 variants don’t matter (so many people have them). Except that it seems like you maybe don’t get ME if you don’t have them. ‘Typical IDO2 is protective against ME’ might be a better way around to describe it? IDO2 variants on their own don’t result in ME but people without these variants don’t get ME whatever stressors they encounter?

If you correct low kynurenine/high tryptophan will correcting that reset ME, even if there are also other factors?! In which case finding the other factors wouldn’t really matter.

Cell Danger Response, trypanosome and Suramin

A curious detail I’ve just come across is a link between African trypanosomiasis (sleeping sickness) and IDO2 and IDO1. Sleeping sickness is something Ron Davis has mentioned in a different context. He said the gene expression of pwme is very similar to gene expression when people have trypanosome parasites.

From last year here is Robert Naviaux describing the Cell Danger Response. He thinks a block in the Healing Cycle can be reversed using Suramin (and this is the drug used for African trypanosomiasis).


2min30 talk starts,

“A major misconception in medicine is that diseases that are “caused by different triggers are different diseases”. I believe this is wrong. The vast majority of chronic diseases are caused by the body’s response to stress or injury, and not the initial injury itself. In other words, over half of all chronic disease is caused by blocks in the healing process caused by pathological persistence of the cell danger response (CDR). This is a categorically different way of thinking.” Naviaux


Ron Davis has said Suramin does seem to return ME cells to a more normal state.

Naviaux on Tryptophan and the Cell Danger Response:

https://www.sciencedirect.com/science/article/pii/S1567724913002390 Naviaux

I’m not too sure if this link between IDO2, Cell Danger Response and Sleeping Sickness would be a competing theory about how IDO2 fits in, or if it helps to situate the IDO Metabolic Trap within the wider ME context.

Supplements and drugs

If you watch the presentation on the IDO Trap, the message not to attempt to hack the kynurenine pathway ourselves is loud and clear.

However, it would be really useful to get more clarity about things we might already be taking or doing. Are there things we should stop? Or should we carry on as normal?

From the Phoenix Rising forum and some googling, this goes beyond tryptophan and could involve nac, St John’s Wort, 5htp, SSRIs, coffee, melatonin, ketamine, B6, B3, BCAA and likely more. These are things which are related to kynurenine or tryptophan levels and some of them pwme commonly take. Perhaps not taking them could cause other problems?

The big practical question I was left with was is it just tryptophan to avoid or any supplement associated with altered kynurenine levels?

St John’s Wort reduces kynurenine

So can nac

Even coffee suppresses breakdown of tryptophan

Melatonin may increase kynurenine so is that alright to keep taking?

Kynurenine in Other Conditions

Google Scholar searches bring up connections between kynurenine metabolism and other diseases. Encouragingly these are related conditions: MS, AIDS, Crohn’s, Lupus, Rheumatoid Arthritis. I think I’ve written too much already to go into any detail.


Suicidal people seem to have high kynurenine, another reason not to attempt to oincrease this without medical supervision.


I get the impression that there’s expected to be other metabolic traps, so if like me you don’t feel this theory is a full explanation, it doesn’t have to be to also be accurate. This could be the explanation for one of several steps down/up. Plausibly understanding this step is the only one that matters though. If this hypothesis leads to digging up the root issue, we can leave it to academics to argue over the other aspects and get on with our lives.

Probably there are many factors which influence how ME expresses (symptoms and severity). Eg in the Billing-Ross 2016 paper, mitochondrial variants were associated with type of symptoms and severity but not whether or not someone has ME. So it makes sense that even if you edited the mitochondrial dna they might feel a bit better, but they’d still have ME. It is just pulling off a few leaves (if ME is like bind weed). If everyone with ME has IDO2 variants, although this may not say much about the individual experience or what it looks like above ground, low kynurenine/high tryptophan might be the root that needs to be dug up.

The high frequency of R248W seems to be a strength and weakness. Lots of people potentially able to get ME might be useful advocacy wise! I wonder about the funding/advocacy repercussions of potentially defining this as something like hypokynurenine and then seeing who fits in that? In other words slightly separate it semantically from ME/CFS. Perhaps it is a specific, neat fit between this and ME, or perhaps it also includes other people eg with MS, or perhaps this is a ME subgroup? This might bypass reluctance to fund ME studies? If it pans out could we be diagnosed with a new biomedical label such as hypokynurenine disease (or similar eg hypertryptophan) and also bypass long term stigma around ME and CFS labels?

This seems to be a time when there are lots of ideas and lots of interesting results generally in ME research. Today as I finish writing this it is the day of Rosemary Bayman’s funeral. Rosie’s family are taking donations to OMF in her name instead of flowers. I know she would have been interested in these ideas, but it is already too late for her. We need some of these ideas to be realised as treatments before more good people are gone.

15 thoughts on “Understanding the IDO Metabolic Trap Hypothesis

  1. Thank you for this wonderful writeup! Many of us at the Symposium were frustrated to be told not to try anything at home…

    You said “this goes beyond tryptophan and could involve nac, St John’s Wort, 5htp, SSRIs, coffee, melatonin, ketamine, B6, B3, BCAA and likely more”

    Many of us can and do take many of these supplements. The trouble is, our bodies need and use many of them, and some are used in different pathways, so while it may not be good to have too much tryptophan, it’s important to realize that being short of other things could be detrimental.

    For instance, PwME tend to be depleted in glutathione, so B6 and NAC might be needed to produce even the inadequate supplies we are making.

    I have been able to stave off or reverse PEM with both glutathione and BCAAs. BCAAs help
    support cardiac and skeletal muscle mitochondrial biogenesis, prevent oxidative damage, and enhance physical endurance, all useful things.

    I am concerned about melatonin – though it’s downstream, I was on large doses (30mg per day) during cancer treatment so I wonder if this backing up plus the 5-HTP I was given for sleep could have triggered this situation.

    Respecting what the supplements can do, good or bad, and working with a functuonal medicine practitioner is helpful, but I don’t think anyone has figured out all the answers here. I do know the supplements (and SSRIs) can be powerful interventions. Keep us posted as you learn more…and thanks for doing this

    Liked by 1 person

    • I find nac, B vitamins and melatonin helpful so I’d be reluctant to stop unless these could cause irreversible damage (the warning about tryptophan seems to be it could cause autoimmune issues which are hard to reverse).


    • I don’t think Ron was saying “check everything you take to find out whether it could affect the kynurenine pathway”. I think he was saying “don’t *try* to mess with this pathway”, especially not by taking tryptophan.


      • If things as common as B vitamins were likely to be harmful, they’d surely have mentioned it.
        The concern seemed to be that based on this hypothesis, we might start doing harmful things. No mentioned concern that things that pwme are already doing could be harmful.
        In fact Phair mentioned B3 (niacin) in the context of the theory, in terms of people getting plenty in the diet but hr didn’t say anything about this being a potential concern.
        Also have a look at the comments on Cort Johnson’s HealthRising article about this. Janet Dafoe & Ron Davis have replied to some of the comments!


      • Yes that’s what he said.

        I’m saying it would be useful to have more input on this, because many of us are effectively messing with this pathway already in various ways. I wouldn’t have thought it would make much difference whether we did this in response to the Hypothesis or not.


      • Eg I had some tryptophan in my kitchen cupboard. Fortunately I’d already chickened out of taking it because I was unsure of the repercussions, and now it’s gone in the bin.


  2. Great write-up ,Jenni. Thinking of IDO2 as protective against ME/CFS is a good way of understanding it, rather than thinking of the defective version as causal. while I love the idea of the bistability, I expect there needs to be something else going wrong beyond defective IDO2 (perhaps to do with cellular calcium transport?) for the trigger to kick a person into full-on ME/CFS.

    The frequency of defective IDO2 genes is too high in the population to make the fact they are common in ME/CFS patients seem very meaningful to me.

    Liked by 2 people

    • To some extent I think the genetic element is more to do with the route that they came to this issue by. The IDO Metabolic Trap might be important but it is in itself more than the defective IDO2. Most people with IDO2 variants won’t trigger the trap.


  3. If this hypothesis was right, then wouldn’t we all have “serotonin syndrome” symptoms?
    In fact, many of us take serotonergic medications (E.g. SSRI) without any signs of serotonin syndrome (diarrhoea vomiting dilated pupils etc)

    Liked by 1 person

    • It’s a good question and actually one I edited out of a previous draft. Something I don’t understand though is apparently the impact on serotonin can go either way (in different cells? Not sure if they mean predictably or not). So possibly this problem could seem like low serotonin too?

      It is the case that pwme can react badly to SSRIs. I was initially misdiagnosed with anxiety/depression and put on a SSRI. I remember feeling terrible – though I think more like it provoked mental health symptoms that simply weren’t there before (I don’t remember serotonin syndrome type symptoms but it might have aggravated autonomic symptoms).

      I can remember Mady Hornig saying a few years ago that ME seemed very linked to serotonin dysregulation (eg immune wise). It seems like she might have found somewhat opposite results?

      “Another finding suggests substantial serotonin dysregulation may be present in ME/CFS. (Most of the serotonin in our body is found in our gut.) Dr. Hornig described serotonin as a major immune regulator. Thus far they’ve found that serotonin is more likely to be undetectable in shorter duration patients than longer duration patients, and those reduced serotonin levels are associated with increased immune activity including a very significant increase in IFN-Y – an important antiviral factor.

      Tryptophan is metabolized to either serotonin or kynurenine. If serotonin levels are low, the levels of kynurenine are likely high. Plentiful serotonin results in feelings of well-being, emotional resilience, and immune balance. High levels of kynurenine, on the other hand, have been associated with a host of neurological and neuropsychiatric disorders. Dr. Hornig has called the kynurenine pathway her favorite pathway because it’s been implicated in so many diseases.

      The low serotonin findings in ME/CFS were apparently significant enough for Columbia to begin developing new tests to more accurately assess the presence of kynurenine metabolites. It appears that they’ve been successful in doing that, and we can expect more fine-tuned analyses of the role that pathway plays in ME/CFS.”

      I can’t find the outcome of that study. Did they find high kynurenine and low serotonin?

      Of course there could be subgroups with opposite serotonin issues (or different cells within the same body some with low serotonin).


What do you think?

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s