Deep phenotyping of post-infectious myalgic encephalomyelitis/chronic fatigue syndrome, 2024, Walitt et al

I find the test with these things - and the authors who are becoming the stooges inadvertently for others might want to do this check on themselves - is to just substitute any other minority and related tropes into what you've just said and see if you think your job/partner/friends would ever look at you the same again.

 

Agreed. And I think the appropriate term is it being an individual's manifesto.

With a few points to some whizz bangs to say 'look laypersons some science because it turned blue' being used to cover for what seems to be a term that doesn't originate from either laboratory or solid validation.

 

Given Naths mixed messaging perhaps he simply did as he was told .

 

My guess is that after the pandemic Nath didn't really want to be working on this given the other projects he had/has going on. As a result, he got Walitt to write up the majority of the paper and then probably felt he couldn't push back as he hadn't done the work. Given how long it took, there was probably as rush to get it published and move on even if he knew it was a pretty crappy paper. Obviously he's still responsible for allowing it to go out in this state.

 

That's a key question, and with relatively new techniques it's often the case that no one can answer it with much authority (although some might pretend they can).

 

This is a late reply, but it seems to me that this important question (which seems so obvious!) is something that is not understood by most people, including doctors and researchers, when it comes to ME/CFS.

This same concept is often ignored when it comes to long term disability claims.

I don't know the best way to translate this idea into guidelines for a research study, but surely the mere concept is not too difficult to grasp, right?

 

This was a challenging paper to write - difficult topic and difficult circumstances. It needed an exceptional scientist to guide it. Nath was not that scientist. He let down the mission of science by permitting this to be published in its current form.

As the sample size receded they needed to rein in ambitions and make sure what got into the abstract was on an exceptionally strong footing.Instead they went grasping for marginal findings from experimental measures. I dont' care about the internal politics of NIH or Brian Wallitt's feelings, the senior author needs to make sure the paper makes sense.

I'm not actually sure yet if there's anything in there that pops out as an undeniably strong finding. These metabolomic findings are in cerebrospinal fluid and that hasn't been analysed as often as plasma or serum so there's not a lot of precedent.

 

That’s ok albeit disappointing. What’s not OK is stating that a mystery antigen is causing all the problems while in fact your study did not identify/discover any such antigen.

 

Shared this in another thread but also wanted to ask here (mods feel free to delete/move as you see fit)

Does anyone think we should try this route? Just was wondering.

'Research Misconduct: Research misconduct is defined as fabrication, falsification and plagiarism, and does not include honest error or differences of opinion."

https://grants.nih.gov/help/report-a-concern#research

'Who to Contact:'

• NIH Extramural Reasearch Integrity at NIHResearchIntegrity@mail.nih.gov
• HHS Office of Research Integrity at AskORI@hhs.gov

'Requirements for Making a Finding of Research Misconduct'

• There be a significant departure from accepted practices of the relevant research community;
• The misconduct be committed intentionally, knowingly, or recklessly; and
• The allegation be proven by a preponderance of the evidence.'

https://grants.nih.gov/policy/research_integrity/requirements.htm

 

MEAction issued a response a week ago to the NIH study which included recognition of the effort and sacrifices of the patients who participated as well as a critique of its contents.l
"NIH’s recent paper on ME/CFS elicited strong reactions from the community. While we recognize the incredible sacrifice of the patient community who gave their health and time to make this study possible, ultimately, we caution against drawing any sweeping conclusions from this study. "

It continues to list perhaps its 3 major weaknesses
1) a narrowly drawn unrepresentative cohort of patients focussed on the mild/moderate
2) the incredibly small sample ( 17 ME/cfs patients)
3) the doubtful pattern of behaviour called "effort preference"
"The paper draws conclusions based on an atypical cohort that may not be representative of the ME/CFS community, and from a very small sample size, and there are also conclusions drawn in regards to “effort preference” that may not be supported by the evidence. "

Read more about #MEAction’s analysis below:

https://www.meaction.net/2024/02/29/meaction-nih-study-response/

edit: The paper is worth reading and is printed in large, well laid out type that is accessible to pwme

 

I believe their point is that the selected participants did not reflect the spectrum of severity of ME.
If I remember correctly, the average steps per day for pwME was about 3500 steps. So that is a weakness of the study even if it's understandable why more severe patients were not involved.

 

I think he does agree with them. Most neurologists I've met that were sympathetic towards ME, did not discount the "BPS" views as false or damaging.

They see the illness as having two layers:

Layer 1: an array of subjective symptoms mostly related to perceptions and way of thinking
Layer 2: the underlying biological process causing the layer 1 symptoms.

This is why their preferred hypotheses are usually something related to neuroinflammation or neurotransmitters.
And why they completely ignore PEM and never think of how they can find new ways to measure the impairment objectively. And why they are not bothered that the patients they recruited don't have POTS.

Because they have already decided (given themselves a huge bias) this illness is a problem of perception rather than a true disability.

I've already said this but patients (not in this forum but the broader community) need to learn that just because a researcher is doing biological measurements, it does not mean they are not misguided.

 

Might be slightly off topic but I think a lot of doctors are simply not bothered by inconsistency.

Let's say for example that you see a naturopath and he thinks your symptoms are caused by inflammation - prescribes you some herbs for it.

You tell him that you've tried many immune drugs, even powerful ones like high dose steroids and IVIG, to no avail.

He ignores your remark and seemingly doesn't care about the inconsistency - how could some herbs help you with inflammation (assuming that's the problem which it probably isn't) when the most powerful drugs we know had no effect?

The naturopath does think your problem is inflammation, he is just not bothered by the inconsistency.

Likewise, I think Nath does think of ME as a "BPS" type illness, but the fact that CBT and GET (the obvious treatments) don't work does not seem to bother him, and he wants to try stuff like IVIG. It is a weird and inconsistent view, a BPS type illness that would respond to immunomodulators and not CBT but that's what he wrote in his paper.

The mistake many patients are making is that just because Nath wants to do an IVIG trial (or whatever other drug), that he is "our guy" and he was forced to write BPS stuff by Walitt. I think this is silly, especially considering how weak this paper was even without the BPS stuff.

The simplest explanation is that we are dealing with very mediocre researchers who are not smart enough to come up with a coherent hypothesis that's backed by data.

 

I actually think the recruitment may have been a huge problem. Things are complicated but in a sense their argument that the problem is 'effort preference' is based on proving that PWME could do everything, they just preferred not to. And that could very well have simply reflected the fact that to be able volunteer you needed to be able to do things at that level, at least to begin with. If the study is used to prove people can actually do stuff it is a highly problematic cohort.

 

I noticed that in the NIH paper some of the p values they report for their metabolites are higher than their adjusted p values. This doesn't make sense because the p value should increase after multiple test correction not decrease. I asked the authors about this:

They were quick to respond which is good of them. I'll paraphrase as I didn't get explicit permission to pass it on:

The 'p-values' and 'adjusted p-values' are from the results of two different analyses and the latter is not a multiple test corrected version of the former.

The 'p-values' are from the result of wilcoxon tests of normalised data (they probably divided mass spec counts by total counts per sample or something like that) and are not multiple test corrected. The 'adjusted p-values' are from the results of an analysis pipeline called limma - which is much more complicated.

Wilcoxon is a non-parametric test that ignores the distribution of the data and looks at how much the data points overlap. Limma is a tool that I haven't used but have heard of; It makes use of a generalized linear model, which allows it to get a better model of the distribution of the data (it might log transform for example if the data is exponentially distributed), and to model other effects such as batch and probably stuff like sex. In principle this could be a more powerful and discerning way to analyze the data.

 

Thanks for this analysis Murph! Are these metabolites here significantly different in Baraniuk 2021? I had a quick skim but in the body of the paper they only report glutamate as changing and that's after exercise (in GWI). If not then the standard error on these fold change estimates could be large and there could be a lot of uncertainty about whether the effect actually goes the other way - or that there isn't one at all.

From other ME metabolomics papers (not of CSF) I think it's a possibility that these nitrogen related metabolites (urea cycle, glutamate, arginine, proline) could maybe be an exercise/exertion related effect. Maureen Hanson's two day CPET plasma metabolomics paper report changes in these metabolites only following exercise in ME/CFS patients (I think it was proline decreases, pyrroline-5-carboxylate increases, arginine increases but I might be misremembering). Armstrong I believe reported PEM associated differences at some point.

In the NIH study depending on the timings of when they were sampled they might have just been through a grueling exhausting week of tests/CPET, whereas I don't know the sampling conditions in Baraniuk 2021. I'm not saying this is the reason for the differences here but the point I'm making is that I think the metabolome is much more dynamic than the other 'omes' and changes in course metabolic state (fed, fasting, exercise) as well as many other things (diet, drugs, BMI) could lead to a systematic bias.

 

I agree with this take. The metabolome is very dynamic and highly individualised. Which means I think that a lot of metabolomic snapshots are less than useful. If there's no consistency in something then measuring it once is of dubious value.

The that chart above is not the first time I've compared two metabolomic studies. Stong agreement is rare. The interesting things are usually not consistent and the consistent things are usually not interesting. e.g. the hypoxanthine finding stands out in Armstong but doesn't stand up in Fluge and Mella's work.




I got into reading research after Naviaux 2017's metabolomic paper ("dauer, hypometabolism" etc). It really sent me off in a new direction. I was full of hope and I've looked at a lot of metabolomic stuff since. I was especially impressed by Hanson's study with metabolomic measurements at 4 timepoints, before and after exercise. It is a great study design. But what's really come of it in terms of understandin the mechanisms of the disease?

To zoom out conceptually, you could even question if the rash of quantitative metabolomic deepdives recently might partially explain the lack of progress. It's easy to imagine that 40 subjects * 1500 datapoints MUST be useful because its such rich data. But ... what if it isn't?! Perhaps it is still just dominated by noise?

Just thinking out loud. I love to look at big data and I remain hopeful. But I'm open to the idea metabolomics needs to improve to actually reveal the mechanisms of disease.

Just to tie this back to the actual topic of this thread, I'm keen to compare all the nih findings to other metabolomic and lipidomic studies and make some more charts like the ones above to see if any findings are consistent