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

[Jonathan Edwards]

They don't comment much on this but I found it buried in a supplement that 24% of ME/CFS patients had positive antinuclear antibodies and only 5% of controls (p=.09 but I wouldn't pay much attention to lack of statistical significance in such a tiny sample).

Yes, that would be a classic type 2 error.

So the claim that the autoantibodies were not elevated was bogus in fact. Inasmuch as it could be ascertained, they were elevated. I doubt this is relevant but it is another indication as to how much the data have been weighted towards a preconceived conclusion.

 

[Simon M]

While figure 10 is only a hypothesis about the pathophysiology of ME/CFS based on their findings, I think it makes it visually very clear that intervening at the level of reduced activity would be foolish.

I hadn’t notice that: this is classic “ME is the result of deconditioning“. it shows the authors haven’t grasped the importance of post-exertional malaise.

ADDED
Or put another way, why doesn't every other limiting condition cause a syndrome similar to ME, complete with PEM. Why doesn't MS have this? Why not everyone who has been in a coma for an extended period?

Back in the 1960s, US and Soviet space programs ran extended (but small) bed rest studies as a proxy for low gravity, to study the potential impact of space travel. There could be 6 weeks to 3 months long and, unlike later studies, allowed no exercise will lying down. The studies did find biological effects, but nothing like ME - and recovery after bed rest was pretty rapid (as indeed it can be with rehab after coma and some neuro injuries).

 

[EndME]

They don't comment much on this but I found it buried in a supplement that 24% of ME/CFS patients had positive antinuclear antibodies and only 5% of controls (p=.09 but I wouldn't pay much attention to lack of statistical significance in such a tiny sample).

It's in supplementary material 9 for those that are looking. Shouldn't there be info on the dilution? I thought that was equally important...

 

[butter.]

Yes, that would be a classic type 2 error.

So the claim that the autoantibodies were not elevated was bogus in fact. Inasmuch as it could be ascertained, they were elevated. I doubt this is relevant but it is another indication as to how much the data have been weighted towards a preconceived conclusion.

So then Dr. Edwards, I guess, looking at all ths nonsense we/you could write a rebuttal now, what are your thoughts on that?

 

[Jonathan Edwards]

It's in supplementary material 9 for those that are looking. Shouldn't there be info on the dilution? I thought was important...

Dilutions would make interpretation a bit easier but it remains the case that at some cut off the positive rate was five times higher in patients.

In fact the figure should presumably have been 23% if four cases were positive.

 

[Sid]

It's in supplementary material 9 for those that are looking. Shouldn't there be info on the dilution? I thought that was equally important...

I can't find info on titres. It doesn't say what cut-off they used for positive.

 

[Sid]

Yes, that would be a classic type 2 error.

So the claim that the autoantibodies were not elevated was bogus in fact. Inasmuch as it could be ascertained, they were elevated. I doubt this is relevant but it is another indication as to how much the data have been weighted towards a preconceived conclusion.

Yeah. Safe to assume we can't take anything they've written in the main article at face value and need to comb through the supplemental data files ourselves.

 

[Sid]

Data entry errors in Supplemental data file 6 which reports results of neuropsychological tests used to detect malingering. The means and standard deviations are the same for controls and ME/CFS patients for every test, down to each decimal, which is impossible.

 

[Evergreen]

View attachment 21167

I hadn’t notice that: this is classic “ME is the result of deconditioning“. it really shows the authors haven’t got a clue about post exertional malaise.

Yeah, the direct arrow between reduced activity and ME/CFS is a nice touch, isn't it? I find researchers always fail to account for why we're sick right at the beginning/when the infection has gone/in month 1/month 2. The "Outcome" column should come between the 2nd and third columns. By all means have another rightmost column that then takes the pwME and loads deconditioning on top - it's inevitable for all but the very mildest affected. But that's not what makes us sick.

 

[Keela Too]

View attachment 21167

I hadn’t notice that: this is classic “ME is the result of deconditioning“. it really shows the authors haven’t got a clue about post exertional malaise.

[/QUOTE]

Surely the arrows are better placed like this. (Yellow arrows added)

Or maybe their final image should be at the start as leading to everything else and the Reduced Activity is the final Outcome

 

[Simon M]

I think the HRV and heartrate issues are quite likey to be a real signal. Other people/teams have found this too and it's the main reason for why some of the pacing Apps are somewhat helpful for many patients.

First, the stats suggest many of the findings are iffy (though of course there is the problem of teeny sample size meaning true effects will be missed). But if we accept them as real signals, are they specific to ME or a generic chronic illness signal, particularly a chronic exhausting/PEM illness when faced with an intense battery of tests?

 

[Simon M]

View attachment 21169

Surely the arrows are better placed like this. (Yellow arrows added)

Or maybe their final image should be at the start as leading to everything else and the Reduced Activity is the final Outcome[/QUOTE]
Yes, it's the assumed causality that troubles me, and the implicit failure to listen to patients about the illness.

 

[rvallee]

Oddly, I keep seeing "effort preference" being a MDD thing but with all the papers we've discussed over the years that are about depression I've never once seen it. In fact if you search Google scholar for it, there are hardly any hits, and most of them are in economics and is about consumer choices.

So not only is it a bullshit concept, it's an esoteric concept that is almost never used and certainly not in health care, medicine or even mental health. I think Walitt just wanted to throw some new label at it. After all this is how it's usually done in psychosomatic ideology: the same old piss in the same old bottle with a handwritten label glued on top of the other labels. Thus the cycle renews despite simply looping on itself.

There is a good term for this: bootstrap pseudoscience. It's pseudoscience that somehow keeps rising by climbing itself again and again, even though doing it even once is physically impossible.

 

[FMMM1]

First, the stats suggest many of the findings are iffy (though of course there is the problem of teeny sample size meaning true effects will be missed). But if we accept them as real signals, are they specific to ME or a generic chronic illness signal, particularly a chronic exhausting/PEM illness when faced with an intense battery of tests?

Nailed it!
I think @JonathanEdwards made a similar point earlier - i.e. you'd need appropriate (disease) controls to establish whether this is simply an artifact.
Jonathan, I was looking for a previous post where I think you mentioned the need to include diseases like MS as controls?
Re a good fMRI study, grateful for your thoughts re the basic elements, controls ---?

 

[Braganca]

I sent an email about the problem with Walitt’s biases, the effort preferences language and the study design to Avi Nath.

He replied:

“We have beyond any doubt whatsoever proven the biological basis of ME/CFS. That is what matters
Avi”

 

[poetinsf]

I think you missed the point that the authors of this paper highlight that they encouraged the patients but not the controls. Encouraging one set of participants and not the other, if that is what has occured, is introducing bias.

I think the point was that you are not supposed to be able to achieve the peak exchange rate, with or without the encouragement if you have peripheral fatigue/muscle failure.

That said, the patients in this study were mildly sick. We don't know if patients in PEM state would've responded the same way with the encouragement. They probably couldn't use severely ill patients who are constantly in PEM state or mild/moderate patients in an induced PEM. That would be unethical human experiment.

 

[Simon M]

$8M NIH study used some of the very broad and inappropriate thresholds* that are part of the Reeves et al (2005) criteria that I thought had been consigned to history . Note: these are either/or *not* all

and functional impairment as determined by having a Short-Form 36 (SF-36) score of ≤70 on physical function subscale, or ≤50 on role physical subscale, or ≤75 on social function subscale.

Frustratingly, supplementary data table S7 gives group scores for PROMS, though only the aggregate Physical and Mental components from the SF36.:

HV (n=21) HV (n=21) PI-ME/CFS (n=17) PI-ME/CFS (n=17)
Mean , SD p-value (Mann-Whitney U)
SF-36 Physical Component Score HV: 56.7, 3.1 | ME/CFS: 23.5, 10.0 | 6.95E-11
SF-36 Mental Component Score HV 54.8 3.7 | ME/CFS 49.1 6.8 | 0.003

 

[Hoopoe]

I sent an email about the problem with Walitt’s biases, the effort preferences language and the study design to Avi Nath.

He replied:

“We have beyond any doubt whatsoever proven the biological basis of ME/CFS. That is what matters
Avi”

That sounds like Nath doesn't like the psychobabble either.

 

[Trish]

I feel both gratitude to the people with ME/CFS who volunteered and underwent a gruelling set of tests, and sympathy with them for being badly misled by the NIH about the study.

I don't think I would have volunteered to participate in this study if I had been told in advance that:

part of the aim was to test hypotheses about effort, and misuse the resulting data to put a false hypothesis at the heart of the abstract of the main report.

they would be subjected to irrelevant and inappropriate psychological testing

the NIH would fail to recruit properly matched controls to allow for useful comparisons

the NIH would fail to recruit even half the promised number of particpants, and only about half of the 17 would do some of the tests from which conclusions would be drawn

the NIH would take 4 years after prematurely stopping the trial to publish anything including making people wait for many months because they refused to publish a preprint

 

[Braganca]

[this post responds to a deleted post critical of Nath]
I actually liked him and I trusted him and his expertise. I thought we were lucky to have him. I feel very let down and wanted to let him know. The first sentence of my email was “I trusted you.”

That sounds like Nath doesn't like the psychobabble either.

I agree, but I would have thought he had the authority to prevent some of this go out in the paper, even if he didn’t understand the disease at the start. It’s clear that NIH is having to spin this paper and manage the press. They managed ok in many of the US articles but the UK articles of course led with the “effort preference”.