1. Patients launch a $1.27 million crowdfunding campaign for ME/CFS gut microbiome study.
    Check out the website, Facebook and Twitter. Join in donate and spread the word!
Science to Patients: Talking ME, Exercise and the Mitochondria - with Dr Charles Shepherd
The latest video release from the Dutch group ME/cvs Vereniging, with Dr Charles Shepherd from the UK ME Association, and announcing a live chat session to be held Thursday, April 10, 2014...
Discuss the article on the Forums.

A neuro-immune model of ME/CFS.

Discussion in 'Latest ME/CFS Research' started by asleep, Jun 22, 2012.

  1. Kina

    Kina Moderation Team Lead

    Messages:
    4,349
    Likes:
    2,379
    Ontario, Canada
    This thread has become increasingly off-topic and has wandered into the territory of multiple rule breaches.

    What is important here is the content of the research paper and the thread was started to discuss the content. It is irrelevant who one of the authors is, what his past history with this forum is, how many letters he has after his name, where he got his formal eduction etc.

    Please stick to discussing and scrutinizing the content of the research paper.

    Thank You.
    Sing, JT1024, jace and 1 other person like this.
  2. Snow Leopard

    Snow Leopard Senior Member

    Messages:
    2,166
    Likes:
    1,424
    Australia
    The fact is that the article was published in a legitimate journal and personally I wish that it would be discussed on its own merits.

    I would agree that the agenda and motives of some researchers (eg psychiatrists) are routinely questioned in this community and I agree that much of this behaviour isn't terribly constructive and it tends to divide the community. and perhaps this is a topic worth discussing within the ME/CFS community in general.
    jace and Enid like this.
  3. biophile

    biophile Places I'd rather be.

    Messages:
    1,246
    Likes:
    3,285
    I'm not going to dismiss it just because Morris and Maes are involved, they may turn out to be wrong about some of the specific mechanisms, but it is a start. I noticed there is an absence of the Lights' work on PEM. They also cite Myhill et al's mitochondria paper, which I think has been challenged for relying on unvalidated tests that may not reflect mitochondrial function in the areas we should be concerned about? The paper should be judged on its own merits, but there are over 5 pages of references, I'm not going to examine them all.

    I agree with ixchelkali's general position: I haven't looked at the model in much detail yet either, but I do suspect ME/CFS lays in the direction of neuro-endocrine-immune interactions, as in, subtle CNS pathologies, immune imbalance, autonomic disturbances, as well as oxidative stress, etc, perpetuating the consequences of disturbed homeostasis, a chronic sickness response, and downstream dysfunction of processes that will self-correct once cause is removed. The direction of causation for the various sub-components obviously need clarification in further research.

    I've forgotten the details of Jason et al's "kindling" paper and haven't read the recent Klimas et al paper on "biomarkers for chronic fatigue". If I had the money I would fund systematic reviews into each component and attempt to construct an overall model based on that, while compensating for issues such as case definition.
  4. Snow Leopard

    Snow Leopard Senior Member

    Messages:
    2,166
    Likes:
    1,424
    Australia
    The wonderful part about building maps and models is it allows us to see the big picture. Such exercises often shine more light on key areas which we might not have previously considered.

    Theoreticians are growing in numbers in virtually all scientific disciplines, so why not ME/CFS?
    Enid and jace like this.
  5. Mula

    Mula Senior Member

    Messages:
    131
    Likes:
    46
    That was the paper Jeffrez referenced which as I explained was not a model for MECFS.

    Pall has not manged to build an explanatory model of MECFS. He has recorded some thoughts in a book that cover various aspects in isolation. The absence of any paper would support this for those who are not aware of the history.
  6. Mula

    Mula Senior Member

    Messages:
    131
    Likes:
    46
    A model is a hypothetical map of the disease process that takes in all known observations. Where there are missing links an author will originate how the observations can be pulled together. Positive-feedback loops are not necessarily inclusive within a model for the initiating action and perpetuation is the purpose of a model.
  7. alex3619

    alex3619 Senior Member

    Messages:
    5,962
    Likes:
    7,821
    Logan, Queensland, Australia
    Hi Mula, if I use your definition the Maes paper is not about a model either. It fails to cover thousands of known facts. A model does not have to be complete. Using this definition in biomedicine means there are no models for almost everything. Its a poor definition. There are many definitions of models, from qualitative/quantitive to explanatory/predictive.In hard systems this would fail abysmally as a model as it does not lead to quantitative predictions.

    This model falls into the category of a qualitative explanatory model. Its a step up from many, but not a major leap, just another step in the process. Its also not a surprise to see models like this - I have been hoping to see this trend. So many people have carefully modelled aspects of this disease, sooner or later someone has to combine and integrate those models.

    Bye, Alex
    Sam Carter, user9876, Sing and 2 others like this.
  8. Sing

    Sing Senior Member

    Messages:
    1,293
    Likes:
    408
    New England
    If a model covered every single fact, it would be in the 20/20 in hindsight position. Instead, a model, or this type anyway, as Alex describes, is a useful way of pulling much of the data together in the interest of suggesting better approaches for further research. It is an organizational tool for further scientific work.
    jace and Enid like this.
  9. Mula

    Mula Senior Member

    Messages:
    131
    Likes:
    46
    As far as I am aware it does not miss out any research observation. Could you not say at least what you believe to be missing?

    The definition I have provided is what a model must be. All observations so far have to be explained which this paper achieves. That is a major work as we have been without one until this moment.

    This summary should explain it in a way anyone can follow.

    http://www.direct-ms.org/sites/default/files/35MSDiseaseModel.pdf
  10. alex3619

    alex3619 Senior Member

    Messages:
    5,962
    Likes:
    7,821
    Logan, Queensland, Australia
    Hi Mula, how about all the proteomics data? Hundreds of data items there. How about genetic polymorphisms? How about the cardiac abnormalities including orthostatic intolerance? How about why half of us have sleep apnoea? How about the wide range of chemical and drug intolerances? Why do 40% of us have small fibre neuropathy? Why do so many of us have EDS? There are many more of course, I have just mentioned the ones that first came to mind. There are about 6000 papers on physical issues and even more in unpublished research (which only shows up at conferences). Nobody can even track it all. This model focusses on immune and oxidative and nitrosative stress. Lots of other things are overlooked.

    I think some of this is implicit in the model but it is not elaborated on. There are many more. The point is that no model will initially encompass all of this, nor should it. Over time the models will evolve and expand, and then we will have a model that explains the main points involved in causation and treatment. Models are often not elaborated on beyond that because its a diminishing returns issue.

    That does not mean this model is not a step forward - it is. In time I expected to see something like this, and I expect to see more of it as various items of research are validated and causal pathways identified. Given enough causal pathways, structured in a systems biology approach, and I think the potential initiating causes will be easily identified for testing - but first we need to keep validating the pieces that comprise such models.Which means we need funding. Which means we need political will.

    Let me expand on a point in the MS paper you posted Mula: "The key aspects of a viable model are that it explains all important observations, is theoretically reasonable and is as simple as possible." This statement captures the essence of what is happening. They select what they think are the important observations - thats what you do in building a model. Which means they made a deliberate choice in ignoring other issues. This is only to be expected. They do not however have any way of determining if the things they chose to ignore are as important or more important. Only time and further research will tell us that.

    As an example of a systems biology approach, though I have yet to properly analyze it:
    http://www.biomedcentral.com/content/pdf/1752-0509-2-95.pdf

    One of the interesting things in this systems biology paper is it implicates connective tissue issues which with further investigation might show the EDS link.

    Bye, Alex
  11. Sing

    Sing Senior Member

    Messages:
    1,293
    Likes:
    408
    New England
    Alex for president! (I'm sorry'--that's an ad hominem isn't it?)
    alex3619 likes this.
  12. Mula

    Mula Senior Member

    Messages:
    131
    Likes:
    46
    Alex, they are all within the model. Mitochondria regulate the cytochrome p450 system which detoxifies drugs and the hypothalamus and vagus nerve regulates breathing. Raised O and NS levels and pro inflammatory cytokines result in small fibre neuropathy. The proteomic data is only a reflection of elevated O and NS and resultant protein modification. The model doesn't miss any of the published abnormalities found patients. It is not written for lay people. A model would not be a disease model if it could not include all. The MS paper says that "using all the available empirical data which have accumulated for the disease and combining that with theoretical knowledge gained from studying many diseases."
  13. alex3619

    alex3619 Senior Member

    Messages:
    5,962
    Likes:
    7,821
    Logan, Queensland, Australia
    Hi Mula, p450 system abnormalities do not explain all the intolerances, cannot in fact. This is much more complicated than that - this is really a claim that the model might be able to explain it with a lot more developmental work. The claim of O and NS protein modification, cytokines and small fibre neuropathy is not particularly clear. It needs more work at least - small fibre neuropathy is not sufficiently understood for anyone to have a good theoretical grasp of it yet. The claim the proteomic data is only a reflection of O and NS modification is almost certainly wrong. The range of proteins expressed reflect regulatory problems, not just modification, though its also possible that many of these proteins are pathogen related. However, the finding by KDM of what he called prions (really just abnormal proteins, not prions) was probably highly damaged protein, and that could reflect O and NS modification to at least some extent. I have been saying that for years, although my main emphasis has been on disregulated protein modification including folding, as this requires a proper redox balance. I would really like to see a study on protein folding in ME.

    Don't get me wrong. I think with further development many of these things might be explained by the model, but its not there yet. Also, each point needs to be validated experimentally, then replicated. The model also needs to account for high levels of viral load in the gut and other tissues.

    One of the issues I discussed in my ME models blog was showing relevance of pathways. Identification of a pathway that might explain a phenomena is not enough. There are many such pathways. Which ones are relevant in patients, and in particular in subsets of patients, have to be validated experimentally. There are so many biochemical pathways in the body that I suspect most things could have many potential explanations. The existence of pathways that might explain it is only the starting point.

    A model is a place to start, its a work in progress. Its always a work in progress. At this level of complexity, its never a final model, nor complete. They still did not explain how it fits in with the specific proteomic and genomic pathways now being isolated using systems biology approaches, nor do they explain the full range of vascular and neurological abnormalities. Its not a sin to produce an incomplete model. Any model that is an advance on previous models is welcome, at least for consideration.The value of a model like this is in generating debate and hypotheses that can be tested. It can lead to better research.

    Bye, Alex
  14. Mula

    Mula Senior Member

    Messages:
    131
    Likes:
    46
    Alex, no where have I said p450 explains all intolerances and proteomics data is explained. You are very enthusiastic, but this paper is not for lay people. Each individual abnormality can be explained to you within the context of the model if you have a few months to spare and are willing. When there is an absence of data the authors of a model apply research from other branches of medicine. This is why the model is not a review but an original model.

    All research is hypothetical, and a model is meant to be include all observations. There is no other disease model for MECFS that has been proposed or published except for this one. You have not managed to find another one. The papers you have mentioned are isolated aspects of the disease which have not been woven into a model. It would be a "sin" to state that you have a disease model that does not include all observations. However in this instance it is your ability to understand that is impaired.
  15. alex3619

    alex3619 Senior Member

    Messages:
    5,962
    Likes:
    7,821
    Logan, Queensland, Australia
    Hi Mula, You just made my point for me, as do the authors in several sections of the paper. If it requires months of additional work to show important things, its a promise not an explanation, and hence is not complete. Indeed a whole section of the model has been postponed to a later paper (presumably): "The role of the peripheral nervous system in modulating the initial effects of immune activation and tissue inflammation is complex and beyond the scope of this paper."

    There are many good things in this paper, but after reviewing it yet again I have decided it isn't a model at all, its a metamodel. Rather than being a recipe for a model of CFS or ME, its a cookbook showing how to create recipes for CFS and ME based on biochemical stress and immunological disturbances.

    I assert again its not complete, nor have you said anything to disuade me.

    The general trend in the paper's arguments is to review a large range of possible explanations for each finding they wish to explain. They then go on to establish that some of these pathways are active in ME, which is a good thing. Not all such pathways are established. This is a shot gun approach, and relies on the principle that if enough pathways are mentioned then most of the important onces should be listed. Eventual models might well draw upon this metamodel, but by itself its not specific enough to be a model.

    Now that does not mean I think the outline given at the end (the conclusion section is a very lucid summary, well worth reading if someone hasn't the time to read the full paper) is wrong. Far from it.

    I actually discovered a scientific inaccuracy in my reading, something which I hope to raise with the authors but is not for discussion here. This was in the early review, and does not affect the main arguments of the paper.

    Let me give an example of incompleteness from the paper: "Coxsackie B3 employs a virally-encoded protein, i.e. 3Cpro, to degrade molecules, which function downstream of pattern recognition receptor signaling".

    These explanations are not complete. Why? If this were sufficient then everyone with these infections would show the massive viral load found in most ME patients tissues. Now the model does point at other mechanisms which contribute to this, but its not explicit and not fully worked out. Its a promise that as this model unfolds it has the potential to explain these things, not that it does explain these things.

    That does not mean the comments made are without value. For example it goes on to say : "Chronic and persistent infections can induce prolonged immune system activation, leading to high levels of localized or systemic O&NS, which leads to oxidative damage to proteins lipids and nucleic acids." [My bolding] This is important. I long ago realized that localized explanations might be necessary, and this was one of the driving forces for me finishing my biochem degree.

    Not only does this paper not explain OI properly, the papers it refers to that I can access (including a recent review by Maes) don't either. Its saying that autonomic disturbances are consistent and expected from the model, not that they are explained: "Autonomic symptoms are associated with cardiovascular abnormalities and may be explained by activated immuno-inflammatory, intracellular and O&NS pathways, including increased TNFα, NO-related mechanisms and NF-κB". Saying a model is consistent with the facts is not the same as explaining them, despite their use of the word "explained". It again is a promise not an actual explanation.

    I think this metamodel has potential. I think its worth discussing. I do not want to see people thinking it is fully explanatory if there are not good grounds for that conclusion.

    I should point out a subsection of this model overlaps with my own earlier models based on the work of Martinovic (there were a series of papers, only one of which is listed on PubMed), who had worked some of this out by 1993 and showed, repeatedly and consistently, that modification of eicosanoid synthesis can lead to symptom remission though typically not full recovery, in at least 60% of patients, with partial remission in another 30%. Indeed my own protocol has at its core an attenuation of the arachidonate dependant eicosanoid synthesis - which is really easy to do but very hard to do reliably. That is one of the things I find fascinating with this paper: it links my own earlier work (mostly unpublished outside of forums) with that of Maes on autoimmunity, Pall on nitrosative and oxidative stress, and even Rich on methylation. There are some bugs in the eicosanoid related section, I am thinking about them.

    One point I think well worth pursuing, is the discussion of the NF-KB/p53 axis. If this can be established, and shown to have regional rather than global significance, I think it has immense potential as an explanation of many issues.

    Bye, Alex
    jeffrez and justy like this.
  16. justy

    justy Senior Member

    Messages:
    2,077
    Likes:
    1,261
    U.K
    Hi Alex - sorry to dumb down this very interesting argument but would you mind explaining this in laymans terms for me? I'm following the argument preeety well, but this bit has me stumped and interested in equal measure -

    I
    Thanks, Justy.
  17. alex3619

    alex3619 Senior Member

    Messages:
    5,962
    Likes:
    7,821
    Logan, Queensland, Australia
    Hi Justy, I will give it a go.

    The Maes paper is based on an integration and enhancement of several older models of aspects of the pathophysiology.

    Andriya Martinovic had a 60% success rate at treating CFS back in 1993, and he only got very sick long term patients, some in wheelchairs. He had a partial success rate of 30%, with some improvement but not enough to return to work. 10% did not improve or got worse.

    His model was that a dysfunction in types of eicosanoids (20 carbon fat based hormones, based on polyunsaturated fats) disturbed autonomic function. He developed tests for this, some easy to do in a clinic, others harder. He also showed metabolic disturbances. It was almost impossible for him to publish any of this, only four or five papers ever got published. I was one of the test subjects, and also made some suggestions regarding protocols, and initiated research into omega 3 therapy. He made several presentations at conferences too.

    He tried many different things. The core of his treatment was dietary fat titration. At higher fat intakes (over 10g per day) the eicosanoid synthesis was wild. At under 10g it was a linear response. So he could titrate synthesis to symptoms.

    One thing he discovered along the way was the nobody who still smoked (that included mj) responded to his treatment.

    As he mapped symptoms to diet he found that small changes in varying factors could change symptoms. It was a constant struggle, as the response varied day by day. So he taught patients to analyze their diet and make daily alterations to compensate for symptoms. This was a complex thing to do, and very time consuming to teach. A 60% success rate was impressive though - but almost everyone had to stay on treatment to control symptoms. It was not a cure. Almost everyone had occasional bad days too - it was a constant juggling act. His measure of success though was a return to full time work, not full recovery.

    As he titrated the autonomic dysfunctions declined and the metabolic rate rose.

    He could not find anyone to back this research. It was too complex a protocol for most docs to bother with anyway.

    The point is that this ties in with proinflammatory eicosanoids (both omega-6 and -3) and the enhanced COX-2 activity. He knew about this in 1988! though it was just COX at that time, he was not focussed on enzymatic subtypes.The COX-2 activity in turn ties in with NF-KB and NO activation. It also is linked to alcohol intolerance, as alcohol causes a massive release of arachidonic acid which in turn triggers massive oxidative stress from the mitochondria (mechanism unknown when I last researched this).

    He is now out of CFS research as he could not get funding.

    Now my take on it is a little different. I think we have mixtures of both substrate depletion and substrate over-utilization. The over-utilization is the COX-2 related stuff, particularly in connection with low cortisol. The depletion is largely due to glutathione depletion, which blocks eicosanoid synthesis at two points. So if we turn on the tap we use too much, get too much pro-inflammatory effect. If we turn off the tap we get too little, not enough hormone synthesis to maintain good homeostasis, and hence results in a different cluster of symptoms. Too much and too little are two symptom clusters that Dr Martinovic extensively mapped in his patients, though its not that simple as symptoms are also related to exactly which hormones are produced - this was a blunt instrument.

    In this model this pathway is a mediating pathway to symptoms, and is not causal. Dr. Martinovic was working on the idea it is causal.

    My primary control point is monunsaturated fat, particularly extra virgin olive oil due to its antioxidants. This swamps and limits eicosanoid synthesis. In addition I avoid all high arachidonate foods - meat fats, egg yolks, organ meats, tropical coastal fish. I still eat lean meat and a very occasional egg though.

    I have tried the Zone diet, omega-3s, omega-3/6 ratios, they don't work very well. I tried cofactor boosting, same results. Monounsaturated fats work by a mechanism called mass action. The same enzymes convert some of the monounsaturated fats, and if they are busy doing that they are not making as much long chain polyunsaturated fats.

    I should stress this is only a hypothetical mechanism, it needs a lot of work to be anythign like demonstrated. However on occasion I have accidentally (eating out) consumed a lot of arachidonic acid, and within an hour my pain levels skyrocket so high I collapse. Its worth keeping in mind, particularly if you want to try evening primrose or borage oil - be carefull.

    I can just hear some asking, how did I know it was arachidonic acid? I went back and investigated. The cooks had changed their cooking oil to rendered animal fat, and I was unaware.

    A lot of this material was unpublished but presented at conferences. Some of it was published in peer reviewed journals, but only one was publlished in a higher ranking journal.

    I should add that I don't think Maes was aware of any of Dr Martinovic's research. Not many researchers are, and the last paper was published in the 1990s.

    Bye, Alex
  18. justy

    justy Senior Member

    Messages:
    2,077
    Likes:
    1,261
    U.K
    Thanks Alex for taking the time to explain your thoughts and the work of Dr Martinovic. Does he suggest a lower fat intake, or just a lowering of foods high in arachidonic acid (never heard of it beofre - sounds like spiders! yuck.)
    Take care, Justy.
  19. jeffrez

    jeffrez Senior Member

    Messages:
    1,099
    Likes:
    618
    NY
    If we take every single symptom any person with ME/CFS has ever had, along with every single research paper ever written on ME/CFS, it's very easy to see that this may be explained by invisible extraterrestrial aliens shooting us with ray guns.

    There, Mula. Now you can stop saying this paper is the only explanatory model for ME/CFS. And their model is just about as useful as mine when it comes to treatment, cure, or what's actually fundamentally causing CFS in reality. Kthx bye. : P
    barbc56 likes this.
  20. Mula

    Mula Senior Member

    Messages:
    131
    Likes:
    46
    It is the reverse. Your arguments give credence to what is an established concept. A model has to include all observations and will fill in gaps with established evidence from other branches of science. A model is then used to lead researchers on what could be fruitful avenues of investigation. A model cannot do this unless a complete map of the disease is constructed. This map may be wrong, but it is a map with the advantage that avenues not once considered will now be tested. Considering that research into MECFS has stalled this throws open the doors to a wealth of research which may nudge us toward real treatments and the cause. You are failing to understand this. If you would have a look at models in other fields you may begin to learn about what one is. This paper is clearly not written for lay people such as yourself.

    You are discussing papers on isolated mechanisms as though they were models and commenting on thoughts which are not published. Can you not at least stop insisting that there other models when you are aware that you cannot reference any.

    To revisiting some of your arguments.


    Why would you require any further information on Coxackie in a model? This is a summary of pathogens which are capable of establishing prolonged infections and as you acknowledge the details on how a state of chronic immune activation is then driven is explained further into the text. Your comments soon after clash with those made in the previous paragraph. You now find that it is of value and even attempt to claim ownership. However the model that is before us does say "or systemic".

    OI is counted with the model. This is child's play to understand if you work in the field. You have said consistent, the authors however said, "Autonomic symptoms are associated with cardiovascular abnormalities and may be explained by activated immuno-inflammatory, intracellular and O&NS pathways, including increased TNFα, NO-related mechanisms and NF-κB". Once again the difficulty is that you have no appreciation for modelling.

    How could this possibly be a metamodel? I do have to find out, what do you believe you mean using that term?

    It is a mechanistic model of the neuro-immune disease MECFS.

See more popular forum discussions.

Share This Page