[timr]

Before you can investigate the causes of an illness, you have to define it. Otherwise, you’re chasing an ever-shifting cloud of ambiguous symptoms, any of which could have different causes. The article opens with this admission, so I’m not stating anything new here.

The problem with “Long Covid” as it exists today is that there’s no such definition. Literally anyone who had Covid once and feels bad today (and quite a few people who never had a confirmed case at all) includes their set of symptoms in the communal diagnosis. Thus, if you dig into these studies, you always find that the syndrome is a wide-ranging and variable constellation of symptoms, making it impossible for a study to have any systematic legitimacy. Moreover, the results of any particular study are more strongly influenced by the inclusion criterion (if there even is one) than by any other factor.

It’s perfectly possible to evaluate treatments in this situation, and would be a better use of resources - pick symptoms, make an inclusion criteria, and run a randomized trial of existing drugs or therapies. But this is likely to fail, and it’s much, much easier to write papers with unprovable theories and retrospective analysis.

[dyauspitr]

Sometimes the symptoms are so ambiguous that it is hard to nail anything down. It’s the same thing with Lymes disease, which is definitely a real thing, but there aren’t good, reliable tests for it. It takes a long time to manifest and the symptoms vary wildly from person to person.

[Aurornis]

> It’s the same thing with Lymes disease, which is definitely a real thing, but there aren’t good, reliable tests for it

There are actually good, reliable tests for it. However Lyme disease (not Lymes disease) became an alternative medicine explanation for everything vague and many people became obsessed with thinking they had it based on vague symptoms like fatigue. When they couldn’t get positive test results to confirm their belief, the Lyme disease online communities established the idea that the tests cannot detect their version of the disease. It’s a belief that allows anyone to diagnose with Lyme disease in a completely unprovable way.

> and the symptoms vary wildly from person to person.

This belief is an unfortunate result of the online Lyme communities encouraging everyone with any unexplained symptoms to believe it’s caused by Lyme disease that can’t be detected. When the disease becomes redefined as being untestable and causing wildly different symptoms in everyone, it becomes impossible to say that anyone doesn’t have it. If you have any vague symptoms like feeling tired, a Lyme disease community will encourage you do believe that it’s caused by an undetectable case of Lyme disease.

There is a lot of strong evidence that these patients do not have Lyme disease, but they’re always good at coming up with another reason why they have it but it can’t be detected in them specifically

[kranke155]

There are increasingly positive markers - autonomic dysfunction in previously healthy people, measurable small fiber neuropathy, and auto immune dysfunction in largely unmapped parts of the immune system.

[setopt]

Interesting. Someone should (or maybe have?) run a cluster analysis on the symptoms to define more specific subgroups. But I suppose getting access to the required health data at that scale is nontrivial?

[timr]

It’s not that hard to get a long list of symptoms for long covid. Just watch this thread as it grows, and you’ll easily find dozens. Things like this end up being a lint trap for people who just feel bad for whatever reason (which is all of us, at various points in our lives!) Nobody likes to be told that their symptoms are idiopathic.

Massaging this kind of data (clustering, etc.) is much lower value than finding fixed criteria that define a consistent group of patients who have objectively defined symptoms that cannot be more readily explained by another diagnosis. This is a pre-requisite for any further study. It can be done, but it’s hard, and it tends to lead to criticisms because you end up excluding a large number of people who fervently believe they have the illness, but don’t fit the objective standards.

Just for example: it’s not enough to claim that you have “brain fog”. A more valid endpoint might instead attempt to classify people based on standardized tests of thinking. Even that has problems, of course, but if you can just claim that you are fatigued and unable to think clearly, there’s a huge problem of confounding (i.e. maybe your symptoms are caused by something else), let alone the unverified nature of the original claim.

[kranke155]

Leading research into Long Covid is already doing this. You’re seeing neural and auto immune clusters gathering around certain immune dysfunction and previously rare diagnosis like Small Fiber Neuropathy. Autonomic dysfunction is being measured in young and healthy people also, and that has its own set of objective testing.

Everything you are saying is happening. But because the suspicion seems more and more that it’s an auto immune condition of some sort, and that we are only catching the downstream effects as some of the immune dysfunction isn’t mapped yet, we are seeing the clusters that you say emerge - overwhelming numbers of symptoms, relatively incoherent connection.

But autonomic dysfunction, small fiber neuropathic and detectable auto immune dysfunction are all known and increasingly mapped positive markers for the condition. Have you read the latest studies ?

[timr]

> You’re seeing neural and auto immune clusters gathering around certain immune dysfunction and previously rare diagnosis like Small Fiber Neuropathy.

Everything I've personally seen in this space is exactly what I described: they start with a set of people who claim to have the illness, then go on a statistical fishing expedition to look for "signs of immune disfunction" (or whatever, but you're right that these researchers tend to focus on immune-related metrics), then use whatever signals they happen to find to create a class. This is not the same thing as what I'm talking about, and it isn't valid.

I'm not going to claim comprehensive knowledge of the space, but the papers I've read that make it into the high-profile journals are of this sort.

The papers cited by this Lowe article are better than most at least in the sense that they have control groups and are doing experiments. But let's be clear -- the first one is claiming to see "long covid" pain symptoms in mice who are injected with whole human IgG (a notoriously messy and subjective approach) [1], and the other is exactly the kind of fishing expedition I'm describing, where they indiscriminately look for "targets" of said antibodies [2]. The former is at least doing an experiment that I suppose could lead to some kind of claim of cause, but the latter (despite the exaggerated title) provides no evidence that the correlations they're seeing are meaningful in any disease process.

I guarantee that using the high-dimensional screening that the latter paper in particular is doing, I can take 1000 random people, split them into two arbitrary classes ("fooists" and "non-fooists"), and find some "statistically significant" difference in immune marker profile between them. That is the fundamental problem with the approach.

When I say that you have to start from an objective measurement of symptoms, it means literally that -- not starting from an assay result that is unlinked to any symptom.

[1] https://www.sciencedirect.com/science/article/pii/S266637912...

[2] https://www.sciencedirect.com/science/article/abs/pii/S00928...

Aside: this lab is becoming infamous for this kind of statistical fishing expedition. It makes me cry for the state of science.

[kranke155]

Then you should fund it. The entire field is to my understanding absolutely starved of science funding.

There are two fairly strong clusters of findings that are objective, repeatable, and consistent. And that is the autonomic testing in long COVID patients is coherent in its dysfunction, and so is the Small Fiber Neuropathy testing that is now consistently showing abnormalities.

Lets go step by step.

Small Fiber Neuropathy. Nerve fiber density is a count with age/sex-normed reference ranges. In previously healthy post-COVID patients with no diabetes and no risk factor, then the test shows whether the nerves are there or they aren't.

https://jdc.jefferson.edu/cgi/viewcontent.cgi?article=1284&c...

https://www.medrxiv.org/content/10.1101/2025.03.04.25323101v...

https://www.neurology.org/doi/pdf/10.1212/NXI.00000000002002...

https://pmc.ncbi.nlm.nih.gov/articles/PMC12847426/pdf/fnhum-...

We have brain structure changes showing in the UK Biobank studies https://pmc.ncbi.nlm.nih.gov/articles/PMC9046077/

Associations with complement dysregulation https://www.cell.com/med/fulltext/S2666-6340(24)00041-2

Muscular abnormalities in long COVID patients reporting reduced exercise function https://www.sciencedirect.com/science/article/pii/S104327602...

Potential that persistent infection shows up in Long Covid patients in abnormal rates https://www.massgeneralbrigham.org/en/about/newsroom/press-r...

If your argument is that people are showing up with abnormalities, then diagnosed with Long Covid, then spurious biomarkers are associated to it - you are just wrong. Wrong multiple times. Demonstrably so.

What we are seeing is more likely to be exactly what it looks like - an novel condition being captured by downstream effects of previously unknown or understudied mechanisms.

[timr]

All of those are examples of exactly what I told you about: they take a group of people claiming to be sick, and go hunting for signals to claim as “significant”.

The MRI studies are particularly egregious examples of this. Just because you see a difference on an MRI does not mean that the difference is due to the thing you’re blaming. In fact, it almost never is.

> If your argument is that people are showing up with abnormalities, then diagnosed with Long Covid, then spurious biomarkers are associated to it - you are just wrong. Wrong multiple times. Demonstrably so.

I am? I have now followed every link. Literally every paper you posted is following this exact pattern. I don't know how you could possibly conclude otherwise, unless you just didn't read past the titles.

They each take a (typically small) cohort of people who self-identify as "long covid sufferers", they subject them to random combinations of tests, and report only what they find to be significant. It's literally the XKCD comic about jelly beans.

https://xkcd.com/882/

[lstroud]

MCAS is pretty well defined and is associated with it.