Re Chronic Lyme disease-I don't have the reference on hand, but one of the "Lyme-literate" books reported a credible instance of blood from a mouse that had gone through the standard treatment transmitted Lyme to a previously uninfected mouse.
Just going to copy/paste this from my comment on your peer review feedback version, which addresses your paragraph about Chronic Lyme Disease being psychosomatic:
Hi, I have ME/CFS. I was also diagnosed with Chronic Lyme Disease last year. (But I decided not to believe it, and that was the final straw to stop seeing that healthcare provider.)
Your paragraph about psychosomatic conditions and CLD is framed in a way that doesn't make sense.
First off, it doesn't make sense to say CLD is universally considered 100% psychosomatic. CLD is not a coherent diagnosis! The diagnostic criteria for CLD varies doctor by doctor. (Err, alternative doctor by alternative doctor.) They try to say it's "the great imitator." They say that basically any vague chronic symptom can be a symptom of CLD.
If you meant something like "the symptoms of CLD are all symptoms that are always psychosomatic": that's not true either. Despite my CLD diagnosis, my symptoms are not psychosomatic. (At least to the best of my knowledge, and at least according to my doctors and therapists from the last few years.)
The set of people who "have CLD" ≈ the set of people with vague chronic symptoms who, at some point, visited the type of "doctor" who gives out CLD diagnoses. (i.e. naturopaths, chiropractors, "Lyme-literate" MDs.)
Secondly: the reaction to calling a condition "psychosomatic" is not simply an aversion to that concept. It sucks to be told your chronic symptoms are psychosomatic because, AFAIK, that's generally not followed up with any treatment (or at least any effective treatment).
I have spoken to many chronically ill people who would be _so happy_ to learn their condition is psychosomatic — if it meant they would get some kind of treatment and therefore relief.
Panic attacks cause weird physical symptoms. These symptoms are therefore psychosomatic. People who get panic attacks have to learn to accept that. But that's fine. Because panic attacks are treatable! If the reason for your lightheadedness is "you're having a panic attack," that's useful, because now you can work on having fewer panic attacks.
By contrast, IME, "your fatigue is psychosomatic" doesn't come with any action items. "Your fatigue is due to depression" means you can go get evaluated for depression, take antidepressants, do other things that work for depression, etc, and see if your fatigue gets better. But if a patient is told their fatigue is psychosomatic with no next steps, it's basically an unfalsifiable dead-end claim.
I had an interesting experience with long COVID. (It was being called post-COVID syndrome back in February). I was symptomatic with COVID for two weeks and never recovered over the next two months. I ended up going to the ER 3 times. Doctors ran a million tests and couldn't figure out what was wrong. They told me the diagnosis was a catch-all for "we're not sure but you had COVID and you're still sick so we'll just go with this".
Eventually they figured out that I actually had a super-rare form of cancer called a Ewing's like sarcoma. Which is only one percent of cancer cases in the US annually. I also talked to another sarcoma patient who had a similar thing happen to him this season.
I'm not sure if this is a trend or anything but it would make sense. COVID sucks up so much of the metaphorical oxygen in the room that everyone is naturally going to think about it first if anything weird is going on.
When you do these peer review things it would be helpful if you listed any major changes between versions somewhere so we don't have to read the whole thing again.
The Tanzania case sounds more like encephalitis lethargica (most known from legacy cases in the Robin Williams, Robert De Niro movie "Awakenings"). The illness was concurrent with Spanish flu, but the trademark for EL was post-infection fatigue.
I just wanted to comment that one theory on post-viral CFS/ME and one that seems to be related to some people's long covid (see this twitter thread for an example: https://twitter.com/ar_pnw/status/1432772600480944128) is that viruses can lead to a degradation of connective tissue which leads to cranial and spinal instability which leads to the long list of CFS/ME symptoms.
There are a growing number of CFS/ME patients being cured by getting surgery to correct cranial/spinal issues, and there is growing awareness that some patients with CFS/ME had connective tissue disorders before getting CFS/ME.
Two cases that anyone interested in CFS/ME (whether they're just interested in it or they have it or know someone with it) may be interested in are those of Jen Brea (recounted here, https://jenbrea.medium.com/cci-tethered-cord-series-e1e098b5edf) and Jeff Wood (recounted here, https://www.mechanicalbasis.org/). It is incredible how these people have recovered, how they may have gotten sick in the first place, and how horribly the medical system can be until you figure out what's wrong with you and how to navigate it.
I don't think the chance of being hit by an airplane is 1/250,000, that would be 1300 Americans per year and that does not seem possible.
Also, I don't think that "What’s my yearly risk of getting COVID if I try to live a normal life?" is a useful question. I think "If I live a normal life, how much does the risk of Long Covid increase?" Or more generally, "If we all go back to normal lives, how much does my risk of long covid increase?"
I'm sorry, but after reading the portion about willpower and me/cfs I am still skeptical and unconvinced. Heron's post, especially read to me as Heron deciding they felt unwell, going to many doctors and undergoing extensive tests to find nothing, and then self-diagnosing that they have an extremely rare condition their doctors haven't heard of that doesn't manifest itself on any testing known to medical science.
Is there any hard evidence that it is a physical condition that wasn't mentioned? Having read the article the impression I got is that it's just a psychological disorder, one that conveniently can't be disproven through medical science
How is chart B from Logue et al supposed to be interpreted? It can't be that mild and severe symptoms are counted separately, because in several cases, the bars for mild and severe symptoms during acute COVIC-19 sum to more than 100%. And it can't be that "mild" means "at least mild", because in several cases, the severe bar is higher than the mild bar.
I don't see this explicitly expressed in your article here, but I've been hearing more and more anecdotes regarding problems in extremities with respect to (possibly) long COVID, ie finger and toe pain and/or numbness, Reynaud Syndrome, etc. Is that just represented as the "tingling" entry in your neurological are of the charts?
There's one argument about long COVID that I never see addressed anywhere, and which makes me believe this is NOT some huge threat that would massively disrupt society and/or warrant indefinite restrictions: there are plenty places in the world that have been, in fact, rather heavily hit by COVID and/or consciously do not implement major restrictions. Certainly there are places where double percent-digits of population had COVID. Yet we never hear about these places particularly suffering from long COVID. I mean, of course there are some sufferers, you might hear some anecdotes, but you'd think if "if a few percent of COVID patients get long-term unremitting genuine CFS/ME, that has the potential to overwhelm government welfare budgets and long-term depress the economy", we should be ALREADY seeing it in some places.
I'm Russian, although currently not resident in Russia, and Russia is my main reference country here: the country seems to have become fatalist about COVID and it stopped being newsworthy many months ago. Vaccination levels are low (25% fully vaccinated) and the vaccine is widely perceived as not particularly necessary and/or beneficial. Some low-level restrictions persist, varying by region, but adherence is low despite occasional police enforcement; masks are very often worn under the nose, if worn at all. Official death rate is close to 200,000 (in a country of 144 million) and estimates based on excess deaths make it closer to 600,000. And despite this all... there's no catastrophe? Life seems to be going on more or less as usual? Long COVID is occasionally discussed in Russian social media but nowhere to the same scale as I see in the Western world. It's not denied, in fact many people know someone who had or is having prolonged symptoms (my own mom got it in June 2020 and it took over half a year for her senses of smell and taste to recover, and it's still not 100% as it was before). The point is, the society is not really being overwhelmed by long COVID at all.
I had a mild case of Covid in December 2020, immediately followed by heart palpitations (something I never experienced before) and neurological symptoms (tremors, intense twitching).
I was given an EKG holter on a cardiological clinic that showed a ton of extrasystoles, and one of the doctors asked me for description of my neurological symptoms. It turned out that this doctor had the very same post-Covid symptoms and was afraid of the possibility that it was actually onset of yet-undiagnosed ALS. After comparing our experiences, we came to the conclusion that it would be highly unlikely.
Sure enough, soon a third person known to both of us appeared with the same neurological problems.
But the really interesting observation: even though the problems persisted for months, they diminished after the first dose of Pfizer vaccine and basically went away after the second one in all of us. Not immediately, more like in a week. I recently got the EKG holter again to confirm. Zero extrasystoles over 24 hours, so back to normal.
So maybe subsequent vaccination has a healing effect on long Covid. That should be taken into account when mapping the long term aftereffects.
Certain people are exposed to covid a lot. I have had covid twice, and it hit me pretty good both times. My lymphocyte levels were really crushed down, and it took a long time for them to come back, and they still have not returned to their previous levels. I'd be curious if there is any data looking at lymphocytes, CRP, and other markers in people with non severe covid. Get blood tests a few times a year, so I was able to track this. The low lymphocytes definitely correlated with me feeling like garbage. (that is a technical term for fatigue, shortness of breath and weakness)
Well, cheer up, it can't go on forever no matter what. Eventually long COVID would have to smoothly merge with what it's like to just be old ha ha. Feeling chronically fatigued -- check. Weird muscle and joint aches for no reason -- yup. Brain and digestion not working as well as it used to -- alas, check.
But the mind is part of the brain and the brain is part of the body and all sorts of physical and non-physical stimuli affect and impact any of the mind/brain/body which in turn affects other parts of our kooky and complicated system.
We need LongCovid studies with proper controls and premorbid information, we need LongCovid studies in Japan, in South Africa, in Belarus, in Argentina, in Morocco, etc. We need to understand what this is or could be across a variety of cultures and environments because, as it stands, I simply do not buy it. The Western-Anglo culture encourages and abets solipsism to the point that self-reporting fatigue, as a condition and not a temporary feeling, is simply something I do not buy. Call me callous or naïve but I simply don't buy it.
As somebody with an interest in this area*, I think this is an excellent overview. With regards to ME/CFS, I find the following things relevant:
1) It's hard to understate just how poor a lot of ME/CFS research is, even from mainstream institutions. Rubbish case definition (see next point), tiny sample sizes, lack of blinding, subjective outcomes not matched by objective measures, outcome switching. Certainly on the treatment side (regardless of whether this is a psychological or physical treatment), a lot of it fails the 'homeopathy test' - i.e. 'is this trial design so flawed that it is possible to get a positive result if testing homeopathy using the same design?'. (Special shout out here to the SMILE trial that found a therapy that appears to consist of drawing chalk circles on the floor whilst chanting 'I'm strong like a tree', and that claims to cure all manner of ills including Multiple Sclerosis, to be effective)
2) Case definition in ME/CFS is a mess, and Long Covid seems to be following the same path. Little care is taken to differentiate patients with general fatigue and those with post exertional malaise (aka post exertion symptom exacerbation). If one's fatigue is down to mild depression then an exercise therapy might be a good intervention; whereas for somebody with PEM it might be pretty much the worst thing to recommend. The ONS data in the UK is likely to be garbage because it is just so nonspecific.
Similarly, misdiagnosis is high. Newton et al found that 40% of patients referred to a CFS clinic had alternative diagnoses upon further investigation.
3) On the psychosomatic issue, I hesitate. On one level, it doesn't really matter as to aetiology, if people are suffering. And, regardless of aetiology, prognosis for ME/CFS is poor and the treatments pretty ineffective. On the other hand... medical practitioners like to talk about inappropriate mind/body dualism but one is still referred to *either* neurology or psychiatry departments. I also find is surprising that if it is down to 'false illness belief' that CBT or GET interventions are not more successful in leading to recovery, especially with respect to success rates in treating phobias etc.
Finally, I find the distinction between 'normal post-viral fatigue' and (assumed psychosomatic) ME/CFS really weird. Assume that somebody gets a bad case of glandular fever/mono. If, two months later, they are suffering from fatigue/brain fog despite the virus having been effectively cleared from their system, this is considered 'normal' *even though AFAIK nobody can explain the mechanism as to why they are still experiencing symptoms*. The moment they hit the 3 month/6 month mark (depending on case criteria used), many doctors will confidently decide this is now down to a psychiatric issue. Does this mean that a patient experiencing post-EBV fatigue at two months is also temporarily delusional? I don't think I've ever seen a good explanation of the difference between the two from those who hold a psychosomatic explanation for ME/CFS.
*I started looking at the evidence regarding ME/CFS when somebody close to me was diagnosed with it. It's fair to say that, prior to that, my view (prejudice) was that ME patients were just very angry hypochondriacs.
That "Being hit in your home by a crashing airplane" figure is a couple orders of magnitude bigger than I would have guessed. That means that in a city of 1M inhabitants, 4 of them in average have been or will be hit in their homes by a crashing airplane?
- Ask them if they are experiencing any Long COVID symptoms (without mentioning it specifically).
- Test for antibodies.
- Look at the responses of those who said that they have *not* had COVID or a vaccine but *do* have antibodies.
This way, it would be possible to get data on Long COVID incidence from people who've had COVID unknowingly. That group (previously infected + unvaccinated + unaware) may be quite small, but I still think it could be a useful exercise. Comparing their responses to the knowingly COVID-positive group might be revealing.
Count me as surprised that dying from "being hit in your home by a crashing airplane" was about 3x more likely than "drowning in the bath next year". The former sounds like one of those ludicrously unlikely things that only happens in the movies (e.g. Donnie Darko), the latter sounds like it should be one of those stereotypical "causes of death no-one talks about but are actually quite common".
Now I think about it again, the former is lifetime risk and the latter annual, right? That changes the ratio appreciably. Still not what I'd have guessed.
I have to tell you - I find your dismissal of psychosomatic illness, in a context where we've known hypochondria and Munchausen's to be real for a very long time, to be quite strange and to stand in sharp contrast with the typical evidentiary standards of this blog.
"They point out that long COVID was first thrust into the public consciousness in surveys run by Body Politic, who self-describe as “a queer feminist wellness collective merging the personal and the political”. I agree this is a weird source for something to come from, but Hans Asperger was a Nazi and I still use his diagnosis, so I probably have to accept these people’s as well."
Well, "Asperger's syndrome" was a diagnosis in which broadly the same group of children Kanner et al described were described in somewhat more optimistic terms for, ahem, 'sociocultural reasons', which has been deprecated for the better part of a decade, and for which studies of adults didn't find meaningful differences between it and similar cases of high-functioning autism with speech delay, so probably there's a good question about whether you should be.
(In practice, I think Asperger's was a bit different to HFA, but in that it was getting at a more heterogeneous cluster of neurodivergent children and might even have been more comparable to PDD-NOS on that point -- "schizotypal disorder of childhood" pretty much stopped being a diagnosis when AS got popular, for instance.)
This discussion omits one important data point: long term effects from the original SARS virus. I’ve long believed this is a major piece missing from the Covid discourse. The original SARS virus has a ~10% fatality rate, but it also has well documented long term debilitating symptoms. My recollection is that something like 30% of SARS survivors were permanently disabled (I thought I’d seen this on Wikipedia, but the current article doesn’t quote a percentage). From that perspective, it’s quite natural to assume that Covid will also have long term debilitating effects, perhaps at about three times the CFR. This is one of the main points that convinced me of the severity of Covid back in March 2020, and I wish it had become more widely known at the time.
I remember Sarah Constantin had some Twitter thread about why chronic Lyme might actually be a real thing, I can no longer find it though I'm afraid...
The Australian study "Post-infective and chronic fatigue syndromes precipitated by viral
and non-viral pathogens: prospective cohort study" is not that useful for comparison, though.
We have to take into account the myriad of negative health effects of living under lockdown/in a permanent state of emergency?
We would have to compare "Long Covid" to other postviral syndroms as they occur DURING this pandemic. Not how bad they are in 2006. Who cares about 2006?!
My strong prior would be that ALL chronic conditions are worse now. That people's prodromes become syndromes faster, that subclinical illnesses become clinical.
So during the pandemic the health of the non-infected and infected should be considered "compromised" to varying degrees.
Does anyone account for this and actually do the relevant comparisons, assuming the data is available. Or uses a model that accounts for this?
Without that, I have low priors, that any of the discussion on Long Covid is worth taking seriously.
My wife was having health problems that no one seemed to be able to figure out and she started self diagnosing as Chronic Lyme. I pushed back after tests and one visit to a Lyme literate doctor alerted me to probable chicanery. It took a trip to the ER to realize she actually had a staph infection.
In this case, the word psychosomatic from her PCP was dramatically unhelpful. She had a real problem, just nobody was able to figure out what it was.
The chart with the 1 in 250,000 chance of having your house hit by an airplane seems likely to not be reputable. The source appears to be a 1997 revised 2nd edition of a self-published book called "Up to Your Armpits in Alligators? How to Sort Out What Risks Are Worth Worrying About!" by John Paling. I can't find any evidence the book is still in print or a review of it anywhere except a very old article in a paygated Wiley journal. Paling's bio is here: https://www.riskcomm.com/pdfdocs/JPfullbio2.pdf. It appears he was a zoology lecturer for a while at Oxford, then became a documentary filmmaker for 23 years, then founded something called the Risk Communication Institute in Gainesville, FL, and that was when he wrote the book.
I'm not going to say he isn't reputable or anything. I have no idea. But this is just a rabbit hole game of telephone being played with no way to get at the actual citation, unless you're going to buy the book to find out where he got the number from. It doesn't pass a basic smell test, though. Obviously, nowhere near 1 in 250,000 units of land surface area on the earth contain a house, and airplanes spend a lot of time not over land. The chance that, when they crash, it might happen over a city goes up a lot given most flight paths are between airports and most airports are at least near a city. But most airports are not near the densest part of a city, many have coastal or rural approach routes, flight paths intentionally take those routes to reduce the risk that they might actually crash over populated areas.
This reeks of a guy who probably cherry picked a year where some spectacular crash over a populated area occurred, saw in that instance 1 in 250,000 houses where it happened were hit with debris, and called it a day. It calls into question everything else in that table, too, because the entire graphic is just pulled from his book.
The study that concerned me the most was this recent Lancet article on cognitive deficits after COVID. Not a perfect study by any means (they don't have true longitudinal data), but they are suggesting persistent IQ declines (no indication of recovery 9 months later):
1 point loss for respiratory symptoms at home without assistance
2 point loss for respiratory symptoms at home with medical assistance
In terms of psychology I was interested to note all the various coping mechanism people utilized during the pandemic. One of the bigger ones was discounting any bad outcome short of death.
One thing I wonder about self-reported long-lasting symptoms: a healthy baseline may include some amount of negative symptoms, e.g. you still cough/sneeze on occasion, you are still fatigued on occasion, etc. Yet you consider yourself healthy. Then you get COVID and all these symptoms skyrocket and you are clearly sick. Now your symptoms slowly decrease day by day as you recover. *If they decrease slowly enough, might you never notice you've recovered?* Each day you still feel approximately the same as the day before, and even when you get back to your healthy baseline the symptoms aren't 100% *gone*, so unless you are good at comparing your state to that of a few months ago (which seems hard to me, unless you actually took notes on how often you were tired/coughed/etc in the past), you might now attribute your baseline fatigue etc to lingering COVID symptoms rather than healthy normal?
I shared this article in a group chat and the preview image was a photograph of a bat. Not really sure where it comes from since it's not in the post and could possibly influence a potential reader to the quality of the post
Second, you say that the chance of long covid after a mild, non-hostpital case is about 20%. By “mild” do you mean showing at least some symptoms? Or do you include non-symptomatic cases as well?
What is the age dependence of (the debilitating version of) long COVID? Is it in fixed proportion to age adjusted IFR? i.e. can one take a flat 3X age adjusted IFR as your risk of serious long COVID? If so, then your age adjusted IFR (pre-vaccine) is about 0.05%, and your long COVID risk is about 0.15%, assuming you are in your mid thirties. This is much lower than you would get if you used age averaged figures.
It also seems weird to me that the effect of vaccination on long COVID risk should be so much more muted than the effect on death risk. I wonder if this could be an artifact of age stratification of vaccination rates (assuming risk of long COVID is also strongly age dependent).
This large UK study into long Covid in children, not yet published but reported yesterday by BBC https://www.bbc.co.uk/news/health-58410584 case-controls for reported symptoms to suggest 'between 2% and 14% still had symptoms caused by Covid 15 weeks later'... look forward to the paper when published, and declaration of interest, I work for the funder NIHR (though not in this area).
I had CFS/ME in my 20s after being in the hospital with a virus, It lasted years, but I think I know what happened. TLDR: started a physical problem, the real problem trained my brain, and caused prediction errors that outlasted the actual problem. I think the same thing often happens with other chronic pain conditions.
1. Not a contagious mind virus: First I'm male in my 40s, when this happened I was an undergrad, and I had no idea what it was, people didn't talk about stuff like this in the 90s, and the internet was a much smaller thing, so no one gave me the idea to be sick.
2. Like many woman, I have an auto-immune disease, specifically b12 and iron anemia, rare in 20 year old men. which I had but didn't know about then, but I now require treatment for monthly. (I eat a ton of meat,FYI, but my body attacks cells that produce enzymes needed to absorb it.)
3. I got better after about 2 years, but I didn't know I was better. I was used to feeling tired and achy. so my expectation is I would continue to feel tired and achy. I was making a prediction, it was wrong, and my brain was making the prediction come true.
I think the understanding that people have been trained to feel pain and can be trained to be pain free is helpful and non-moralizing. The initial pain and discomfort is real, the brain learns to expect it, and it becomes a very normal cognitive bias, that exists because our brains rock and are good a learning and making predictions.
One thing that always bothers me about these percent results, eg
<< And even though the study started with 5000 children, so few of them got COVID, and so few of those got long COVID, that the 2.9% turns out to be about five kids. I don’t really want to update too much based on five kids >>
is that you can actually estimate exactly how much to update. But there's not often much discussion on uncertainty in medicine papers.
The question is binary (either they have it or not). The 2.9% are 5 - so we have p*N=5 out of N=170 in a binomial distribution. The expected probability p*N of the event is 2.9%, and uncertainty on that probability sqrt(p*(1-p)/N) is around 1.3%. Using the rule-of-thumb hack that two-deviations correspond to pvalue < 0.05, you end up being able to say with "some confidence" that it's between 0.3% - 5.5%.
That was very back-of-the-envelope, and now we'd have to apply Bayes with a prior. But the short no-calculus story is: the interval is very wide. If you have any prior about that probability with a reasonable confidence interval (eg by what we know from other diseases) you should not change it much.
To illustrate the point, say your prior is 0.95%-1.05% (at two sigma level like above). Another back-of-the-envelope calculation then updates it to roughly 1.16% - 1.20%. I.e. a bit higher but not much. This was computed with a gamma prior for the binomial - because it's easy (see https://en.wikipedia.org/wiki/Conjugate_prior#Table_of_conjugate_distributions). It's actually not very good for modelling small probabilities with large-ish uncertainties, and as a result updates too fast away from zero. A proper discussion would end up giving an even more conservative estimate for the update.
> Unfortunately, having a symptom for psychosomatic reasons sucks just as much as having it for any other reason.
Psychosomatic conditions seem like ripe targets for psychedelic therapy.
I was also interested in whether anyone with CFS had tried MDMA or something. Couldn't find any reliable info from searching online.
> Flu is as common as COVID, but nobody really talks about it having a significant postviral syndrome so probably it’s not that bad.
Or maybe nobody had made this connection. I think COVID has been revealing a lot of long-term associations, like long COVID, that people hadn't previously considered.
>>Of vaccinated people who got symptomatic COVID, about a third ended up with Long COVID symptoms, the same rate as in unvaccinated people.
One would expect vaccinated people (in the US at least) to belong to a group that is more prone to health anxiety and especially COVID anxiety than un-vaccinated people. There is so much correlation between vaccination and beliefs about the seriousness of the pandemic that that seems like a major possible confounder.
I'm wondering if anyone has seen research on people having a certain tendency toward this sort of effect? Ie, if you’ve had a post viral syndrome before, are you more likely to get long covid?
Personal anecdote:
As a teen I returned from a trip to Africa with mild but persistent symptoms. Fatigue, headaches, nausea - all of which intensified with any exercise. No acute illness but the ultimate conclusion of the doctors was that I picked up a virus that wasn’t that bad (hence no acute period of illness), but was unusual enough to my immune system (being from another continent) that it freaked out. I was sick about a year.
I am a woman and have vitiligo, an autoimmune disease.
I THINK I had covid early March 2020 but was never tested. It presented as a mild cold but I had fatigue for about three weeks afterwards.
I'm someone who was pretty strongly in the "Long Covid is almost certainly not a real thing" camp prior to reading this article. Now I'm in this, "I still think Long Covid is likely not a real thing, albeit with some exceptions camp" now.
First - Yes, the people with lung scarring and post-ICU syndrome and maybe some of the people suffering the misunderstood post-viral symptoms are truly suffering, but they're not suffering from Long Covid. They're suffering from lung scarring and post-ICU syndrome and maybe some of the people suffering the misunderstood post-viral symptoms. It doesn't make sense to classify these individuals as having "Long Covid."
This sounds like a semantic point, but I really don't agree that it is. There hasn't been a disease that's gotten this much public focus since... the Spanish flu? Definitely not one since our world got connected like this. This is the largest story of the past decade, and large stories like this attract hangers-on.
Maybe I'm cynical, but I think that if we accept that there is a vague condition like "Long Covid" out there that has amorphous symptoms, no ability to test and no ability to cure it, we're going to get hordes of people who (for whatever reason) start to glom onto it.
This piece made me want to read a lot more about psychosomatic illness (ideally from Scott, but open to suggestions if people know of good resources here). In particular I'd like to see someone explain how psychosomatic illness differs from mental illness. My first instinct is to say that both originate in the brain, but mental illness expresses itself in "mental" symptoms (changes in mood, hallucinations etc.) and psychosomatic illness expresses itself in "physical" symptoms. But of course mental illness has physical manifestations as well (e.g. depressed people can have reduced sense of taste/smell). Is psychosomatic illness best understood as a subset of mental illness, or a distinct category?
Related question: when contrasting psychosomatic illness with non-psychosomatic illness, what's the term for "non-psychosomatic"? My first instinct was to use the word "real" but that implies psychosomatic == fake which isn't correct. Scott uses "organic" above but that feels weird too (not sure how psychosomatic illnesses are "not organic").
I may pick up Ross Douthat's "The Deep Places" which deals with his experience with Chronic Lyme. Seems like it's more of a meditation on the nature of chronic/psychosomatic illness than a medical perspective on the issue but it may be illuminating here.
Hey Scott, can you give us an update on how your workouts are going? (I'm not a paying subscriber so if that's paying-folks-only information, by all means ignore this.)
I have some rather personal data points. Putting the conclusion first: I find long COVID to be hard to study, because I maybe had COVID, and maybe long COVID. However the symptoms are conflated with other things. If I were asked if I had long COVID, I could not give a survey a useful "yes" or "no" answer.
I had Mono in 2018. It was bad enough that I took a trip to the ER, and got meds for reducing pain and swelling. Their testing confirmed I had it, days after a previous misdiagnosis of strep throat from inconclusive tests. After my main recovery, I had at least half a year of fatigue. I could do my desk job, but would need to sleep shortly after getting home. It seemed to slowly fade, and was entirely gone somewhere between 9 months and a year later.
I got sick from something at the start of March 2020. (following what were likely colds in December and February...it was a rough few months.) COVID tests were unobtainable. I don't recall breathing problems or loss of smell but it was something more severe than a normal cold, especially the fatigue. My city at that point was one of the few in the US with confirmed community spread. My primary care physician denied my request for an antibody test - indicating that it was company policy that the tests weren't specific nor useful enough. Overall I put the probability that I had COVID as undecidable.
I started lockdown early because of it, so I didn't notice for quite awhile that I had persistent fatigue. I mentioned mono because the new fatigue was exactly like the fatigue I got from mono. However since then I have identified: depression (medical diagnosis), burnout (self diagnosis), and stress induced physical problems (mixture of both, the strength of the symptoms at their peak was eye opening, and I will never have a shred of doubt again that such problems can be very real). I also commonly have headaches caused by neck stiffness caused by who knows what, likely unrelated to any potential COVID. All of this has combined and fed into one another to be one giant bundle of misery, but a combination of medication, help from family, quitting my job, and time have mostly abated the symptoms. The March illness was the kickoff point, though.
What exactly is psychosomatic? When I was in kindergarten I suffered from severe eczema, the outbreaks started when I started school. A family member - a professional nurse - suggested to my mother that it was psychosomatic. The nurse based her argument/diagnosis on the fact that the eczema started when I went to school, and subsided during the school holidays, and thus concluded that the eczema was a stress reaction. Years later, when I was 25, I learned after testing by an ENT specialist, that I am severely allergic to dust mites. This explained my eczema, as follows: During my kindergarten years the teachers loved to have the children gather around in a circle, sitting on the floor, for sing song -and story time sessions. Though the floors were swept daily, they were not vacuum cleaned. This was South Africa in 1961 after all. Vacuum cleaners were a rarity in those days, especially in rural schools. I think its advisable to be skeptical of any diagnosis that categorizes a physical condition as psychosomatic.
Are those control group numbers for mental health for Amin-Chowdhury et al. normal? They seem rather high.
Perhaps the long mental health problems are from 'Pandemic Exists', rather than from anything particular to COVID. This might also be an interpretation of the increase in mental health problems after the Spanish flu - were former flu patients more likely to be admitted to mental hospitals or was this increase from the entire population?
Scott: what do you see as the endgame for COVID? I believe (courtesy Zvi) that the UK was at 90% seropositivity among its adult population in June, and then it had one spike (which much have left pretty much the entire adult population either vaccinated or recovered), and now it is having another. I am thus increasingly coming to the view that herd immunity is impossible to achieve even with 100% vaccination. So where is the cavalry? Do you think pediatric immunizations (when they finally become available) will be enough to stem the tide? (and what about the animal reservoirs - deer, cats etc?). Or is it just a choice between maintaining extreme precautions (isolation whenever possible, N95 when not) for the rest of your life, or giving up and getting on with life, trusting to the vaccine to do its job in protecting you? What is the plausible argument that says there is a better choice than this available?
There’s an interesting treatment in Therapeutic Plasma Exchange (the same technique used in recent mouse aging-rejuvenation studies (which amounts to mechanically filtering out the immune system and any accumulated damage to the blood)) which has shown promise in the treatment of long covid: https://www.youtube.com/watch?v=nnv0fcr_Hd4&t=381s
If Long-COVID is caused by the presence of something in the blood plasma (like a “rogue antibody”), then I hope these studies will help us get to the bottom of it.
I had an extremely bad bout with flu turned into bronchitis turned into pneumonia when I was 20. I was stupid and arrogant with the illusion of invincibility that comes with youth. I figured I'd just eventually get better and made no specific efforts to do anything about it. I stayed out late, kept clubbing, kept drinking. I ended up remaining sick for over 6 months until I finally had to go to the hospital and get blasted with high-dose antibiotics and corticosteroids after some nebulizer treatments to clear up my lungs. After the symptoms finally went away, I had enough of an immune system drag that I caught another flu just two weeks later. That one wasn't quite as bad.
But it took a really long time to fully recover. I'm not just talking about persistent fatigue. I actually don't remember anything like that. The actual problem is I caught another flu, and not just flu but ended up with secondary bronchitis, every single flu season like clockwork. It made no difference if I got a vaccine or not. Every single flu season I'd catch flu, and every single I caught flu I'd get bronchitis. I also developed allergic asthma responses to grass pollen.
Until I didn't any more. Since about 32 or so, I haven't had bronchitis a single time. No more asthmatic allergy responses.
I think something like "long flu" is a perfectly apt way of describing this. I had a drastically weakened respiratory and immune system for over a decade, but it wasn't actually permanent. At least, I don't think it was. I have no idea if I have lung scarring or reduced capacity compared to before, but I don't really care. It's the more explicit symptoms that kept killing me, and they've been gone a while now.
I'm wondering if you may have linked the wrong article for "weird x-rays" by mistake? I was expecting an article about x-rays and lung scarring, but it links to "What Is the Evidence for Post-COVID Fatigue?".
I'm really annoyed that the BMJ risks table doesn't directly cite any sources and doesn't give any indication of who the stats apply to. I guess we can assume it's about the US? And they all specify a timeframe (which is good) except the most interesting one, the crashing airplane. I guess that one is easy to guess too (it's about lifetime risk... I assume) but they shouldn't leave it up to readers to guess the context, or to guess which if any of the article's five references back up the numbers in the table.
A comment about the difficulty of getting a diagnosis if the pain is the most obvious symptom.
****quoted in full****
*Many have been told by medical professionals that they’re just having anxiety or making up their symptoms. Even now, “it happens more often than not,” Lisa McCorkell of the Patient-Led Research Collaborative told me.*
Somebody needs to do a study on how many people actually make up their symptoms to get attention from doctors. Because according to doctors, it's an epidemic of extreme proportions, a plague ravaging health care services around the world, and it's taking up enormous quantities of their time that would better be spent elsewhere.
Let's ignore for a moment how self-serving that sentiment is; doctors are convinced that people will fake a huge range of symptoms just to be prodded for a few minutes in an exam room. I'm not sure what they think is so outrageously wonderful about that experience that people will commit to acting miserable to get in there, but they're pretty certain that a vast swath of the population is malingering performatively to just bask in their presence.
On the other hand, if you go to any message board dedicated to any chronic disease, one of the most popular conversation is 'how long did it take to get your diagnosis.' Patients trade stories of how often their symptoms were attributed to attention-seeking and malingering until they were able to convince the right doctor to do the right test, at which point they got a diagnosis. And that's for diseases that have definitive tests available; I can't imagine how bad it is for diseases which are diagnosed through exclusion.
I walked around with a painful hernia for four years. I'd been diagnosed with Crohn's, and every doctor I went to said the pain was Crohn's-related, and that there was nothing they could do about it, and I was in remission so it couldn't be as bad as I was saying. I was told by one GI doc that the pain I was feeling couldn't be related to the hernia that showed up on my CT scan. That same doctor put a diagnosis of hypochondriasis on my chart. Over a year later, a different gastroenterologist noted that my pain was in the inguinal area, and had I considered that it might be a hernia?
And the test for a hernia is, you know, 'turn your head and cough.' Once a general surgeon did that test, he had me set up for surgery two days later.
The post-surgical pain from the hernia was less than the pain from the hernia itself within 24 hours.
When huge quantities of patients have to go through an ordeal to get relatively common, run of the mill conditions diagnosed, there's something deeply broken about the medical system. I keep thinking about every doctor I saw during those years I was experiencing enormous pain from an easily curable condition, and how they're all still convinced I was making shit up, and I'm struck by the fact that there's no way for them to learn from these mistakes. Doctors have no idea how many of their diagnoses were later revised or contradicted by other doctors, so they have no way to learn which patients weren't desperate for their attention, but for medical attention.
And now that we're seeing a disease that's difficult to diagnose spreading in this new plague, the condescending, dehumanizing, contemptuous chant of 'it's all in your head, it's anxiety, it's depression, go away' (never accompanied by a psych referral, strangely), is ringing out loud and clear. And until we figure out objective ways to measure things like pain and fatigue, they're going to keep dismissing genuinely sick people as attention-seeking malingerers unless we find a way to let them learn from their mistakes.
Great post, but it's not so convincing in addressing the concern that seropositive and seronegative people have the SAME long-term.
It wasn't just WSJ opinion; NYT ran one too, pointing out:
NYT: "One peer-reviewed study of people who reported long Covid symptoms noted that most of those who were tested for antibodies that provide evidence of a previous SARS-CoV-2 infection had negative results. The level of symptoms, moreover, was virtually the same whether the person was positive or negative for antibodies. A second study, not peer reviewed, of adults referred for long Covid management similarly reported that no Covid antibodies were found in 61 percent of them, again without differences in symptoms whether testing was positive or negative."
"I feel like some of the case-control studies above, which clearly show that seropositive people have higher rates of Long COVID than seronegative people, are pretty convincing here."
Do they? And why then do they conflict with these two other studies find no difference?
The conflict between those studies seems like a crucial question to get to the bottom of, in establishing existence/non-existence of "long Covid" beyond the obvious respiratory/smell aspects. I plan to dig into it, but thought I'd mention in case others have answers already!
So, I have a question about the sermon re: CFS. If most people who haven't suffered from CFS severely underestimate how bad it is, then wouldn't this tend to lead to systematic overreporting of CFS? Like, if someone feels tired all the time because they have a demanding job, 2.4 children, and don't get enough sleep, isn't it possible that they might go to their doctor complaining of "chronic fatigue"? Or if they get COVID, are later surveyed about their symptoms, and the surveyor asks about chronic fatigue, they might say "yes, I've had that", even though their actual symptoms are much milder and more treatable than CJ and Heron describe. If there's a prima facie problem with this story, I can't see it. Do the numbers for chronic fatigue rely on something better-screened than self-reports? If not, wouldn't the systematic underplaying of chronic fatigue in our society lead to some people mis-self-diagnosing regular tiredness as clinical chronic fatigue?
I must confess (as one does after a sermon), I've been guilty of underestimating how bad chronic fatigue is in the past. But my updated world model doesn't necessarily include assuming everyone has the very bad symptoms CJ and Heron are talking about just because they use the words "I have chronic fatigue". But it's also possible that I'm so eager to avoid overcorrecting that I'm undercorrecting. Are there any good estimates of the prevalence of CFS either post-COVID or just in general based on some diagnostic criteria that aren't self-reports, so I can get a handle on the actual numbers beyond "exists" or "doesn't exist"?
I think we may need to re-evaluate these numbers since we're now starting to deal with the delta variant and it may skew these numbers upwards if you're not vaccinated.
> maybe women’s immune systems are permanently upregulated to be prepared for attempts by the placenta to secrete immune-downregulating chemicals during pregnancy
For me the question is less "why are women's immune responses so strong?" than "why are men's so weak?" Like, what are the costs to a strong immune system? Aren't autoimmune disorders only a problem because of some (unknown?) features of the modern environment? And even today, arguably women get the better deal: yes, women get more Long COVID, but when men get COVID they're the ones who are more likely to end up in the ICU/dead. So why hasn't evolution built a stronger immune response into men?
[End contribution to the discussion; commence brain dump of things I find interesting/puzzling vaguely in the space of sex difference in immune response]
Wild speculation, but: is the fact that women are 1.5-2x as likely as men to be depressed *and* have stronger immune systems evidence for the "chronic psychitis" view of depression? https://slatestarcodex.com/2015/01/05/chronic-psychitis/ Anyone know of research on this?
Women also have had more COVID vaccine side effects, including (anecdotally) really weird periods (e.g. super long, heavy, or happening post menopause; self-reports in this thread: https://twitter.com/KateClancy/status/1379847815636135941) The stronger side effects obviously make sense because of stronger immune systems (and susceptibility to psychosomatics too), but this makes me realize that I don't understand why you wouldn't get post-viral syndromes after vaccines. I get "poorly understood," but I would have naively thought that in one way or another these syndromes are caused by your immune system never quite getting back to baseline after an infection. Why wouldn't that happen after the vaccine triggers an immune response too? Why aren't we getting Long Post-Vaccine-COVID? Can anyone who understands biology better than I do help me out?
My bottom-line assesment of the draft version was: An average unvaccinated adult infected with COVID will suffer an average:
~5 days of Acute COVID
~20 days of Long COVID
~20 extra days of Being Dead
Being vaccinated probably knocks all of those numbers down, maybe by a factor of two, on top of the very substantial reduction in the probability of being infected in the first place.
The severity of Long COVID symptoms seems particularly hard to assess, but anything close to "I wish I were dead" seems to be extremely rare. Therefore, the risk of COVID can be properly rounded off to the risk of dying from COVID, plus a small correction factor that probably shouldn't tip the balance in most risk:reward calculations.
As someone who has CFS, I am a little relieved that the Long Covid odds seem lower than I feared, but I gotta say that the way section 4 presented it read as "it's bad, but hey, at least it's not CFS!"
What you wrote in section 5 was excellent though, and your compassion shone through, but since section 4 came first it did feel like a bit of a gut punch.
While I'm here and thinking about it, did any studies try and quantify fatigue levels? There's a huge gap between "bed-ridden", "Can't work", and "life significantly affected, but can still work (perhaps only part time)", but all of those describe some fraction of CFS patients....
I had the good luck of having a series of very stressful life events fuck me up, followed by a bad injury, followed by a stressful job.
This put me in an uncomfortable situation of having to meet my life commitments while suffering through a onslaught of severe headaches, fatigue and generally just wanting to die. This I was able to do... But only just very barely. This kept me miserable for many years, and I'm only recovering now.
I wonder how many people are in similar scenarios. No doubt there are some viruses that produce similar effects ad infinitum. But you can be swimming comfortably, and a few waves can have you gasping for air and desperately treading water. Once you're in the life hole, life has a way of keeping you there.
Can we discuss the risk communication here that somewhat trivializes this when there are small rates but large populations affected? It's IFR all over again.
Based on Scott's estimates and our current policy trajectory, 50+ million people would have severe debilitating lifelong COVID.
How do I arrive that?
> 8. Your chance of really bad debilitating lifelong Long COVID, conditional on getting COVID, is probably somewhere between a few tenths of a percent, and a few percent. Your chance per year of getting it by living a normal lifestyle depends on what you consider a normal lifestyle and on the future course of the pandemic. For me, under reasonable assumptions, it’s probably well below one percent.
Given that the current public health policy approach to the pandemic (which I oppose) is essentially that we focus on vaccines (but with a slow pace globally given vaccine apartheid) and that everyone otherwise can self-protect while waiting for or declining vaccines, with an explicit or implicit assumption now with Delta's infectiousness that "everyone will get it." Again, I disagree with the fatalism here but that is the apparent approach given waning use of NPI policies. And I also think a harmful assumption is that people will get it once and that the level of vaccine- and infection-acquired immunity will hold, when in reality reinfection and breakthrough infections are known, and the risks with waning immunity are more of an unknown.
Regardless, let's assume everyone is infected once, 25% of humanity is fully vaccinated is now, 75% is not, and another 25% will be vaccinated before catching COVID. And let's assume Scott's rate ranges and simplify to 1% with COVID will get long COVID and vaccine halves that.
Just an unsubstantiated general point-- it seems like every account I see of diseases where exhaustion is the major symptom is about someone who was previously high energy-- worked for a lot of hours, family and social life, and athletic, too. All of the sudden, they're flattened.
Surely being sort of average (fried by long hours, netflix is primary hobby) can't be protective. Or is it? Maybe exhaustion diseases are more of a risk for people who are frying themselves. Maybe it's only high-energy people who have enough to spare to write about their problems when they're exhausted. Or what?
Or is my sample biased by media preference for drama?
Chronic autoimmune diseases are tilted towards women, but diseases the follow from an acute autoimmune attack are, if anything, tilted towards men. Everyone describes juvenile diabetes as autoimmune and it is mainly men. The literature describes narcolepsy as autoimmune, and it has even sex ratio, maybe tilted towards men. Greg Cochran describes schizophrenia as autoimmune and it is mainly men. Narcolepsy and schizophrenia occur earlier in men, but diabetes earlier in women.
"We found that the odds of having symptoms for 28 days or more after post-vaccination infection were approximately halved by having two vaccine doses. This result suggests that the risk of long COVID is reduced in individuals who have received double vaccination, when additionally considering the already documented reduced risk of infection overall."
One way to test how much of long covid is Psychosomatic would be to see if Republicans or Democrats are more likely to have the condition. Or use an other proxy(e.g occupation,religion,population density,zip code). And or asking directly how bad they think covid is and how long the symptoms last before they contracted covid.
your risk of breakthrough covid is probably really really serially correlated
Immunocompromised people make up a huge proportion of breakthrough infections; almost half in one study, despite being like 3% of the population.
If they get vaccinated at the same rate as everyone else then, by Bayes' rule,
P(bt|ic) = P(ic|bt)P(bt)/P(ic)
so that means if 1% of vaccinated people get breakthrough infections each year then vaccinated IC people have about a 1/6 chance of getting covid in a particular year(!); if 10% of vaccinated people get breakthrough infections in a year, then we're looking at more than one infection per year for the average IC person(!!!)
the law of total probability or something means that a vaccinated non-IC person has a breakthrough infection risk of
p(bt|~ic) =(1 - P(ic|bt))p(bt)/p(~ic)
ie about 0.5% if p(bt) = 1% and about 5% if p(bt) = 10%
this is assuming IC people are vaxxed at the same rate as the general population, which maybe isn't true. if every IC person in the US is vaccinated right now, then P(ic|vax) = P(ic)/P(vax) ~= 6% of the vaccinated population. then the IC get infected at a rate of about 8%/yr (at a 1% breakthrough rate) to 80%/yr (at 10%), while the non-IC get infected at about 0.5%-5%/yr
this suggests to me that probably 10% is too high, because otherwise there'd be sensationalistic news articles about Guy Who Got Covid 3 Times After Being Vaccinated and i haven't heard about that at all
obviously you might end up immunocompromised in the future, but it's obviously strongly serially correlated. so your estimated lifetime risk should (I think) be a lot closer to whatever you get by cutting your annual risk in half, assuming you're not currently IC, and close to 1 if you are
breakthrough infection risk probably varies with a bunch of other serially correlated stuff like age and sex, too, but i find it hard to believe any of that is going to make a big difference compared to being immunocompromised or not. previous covid infection might matter? but idk
all calculations were done before 10 am so i cannot be held responsible for any errors
Is it possible that patients diagnosed with CFS/ME don’t get better (as compared with people informally deemed as having postviral syndromes) because they are more likely to be medicated for it (because they and their doctor have decided it’s serious, won’t be going away on its own, and medication for it will be covered by insurance if a formal diagnosis is on file)?
I ask because I have known many people who, once prescribed Adderall for fatigue or a similar malady, have a hard time functioning without it—which would certainly make them answer on a survey that, at present, they would function far suboptimally if they stopped taking stimulants. Both because of downregulation, and because once you get used to a new baseline with a stimulant (including caffeine or nicotine), the prospect of living without that stimulant certainly could register as a chronic disability.
If that stimulant is caffeine or nicotine, we say this as a joke (but deep down still mean it); whereas if it’s a prescription drug, we are being literal. (Even if originally we were only going along because otherwise our medicine would cost five times as much.)
The question isn't really "do people suffer long-term effects after COVID infection" but rather "is there a specific 'long COVID' that's different from other viral infections?". And the answer to that - with the exception of anosmia - pretty much seems to be no.
“Getting hit in your home by a crashing airplane”?? 1 in 250,000!? No way. That would be 1200 people in America alone dying from this every year. Whoever wrote that is completely full of shit.
Re. Chronic Lyme Disease, the Wikipedia page you linked is constantly monitored by anti-CLD fanatics who reverted each of my attempts to make the page less biased within hours. Every one of the Lyme studies that the NIH and the CDC list on their webpages to dismiss CLD is so fatally flawed as to be completely valueless. One, for instance, "proved" that chronic Lyme doesn't exist by showing that none of the patients in a set of patients who complained of chronic Lyme /and tested negative for Lyme disease/ tested positive for Lyme disease.
The suppression of CLD is the only actual legit medical conspiracy I know of, where a bunch of people packed the IDSA panel on Lyme, deliberately excluding anybody who believed in chronic Lyme, and never considered any of the papers which suggest that chronic Lyme exists. It was a total kangaroo court, rigged from the start.
Would pay money to see how this analysis should be updated in light of the much more rapid spread of the omicron variant, especially that final "below one percent" number.
Long covid still seems like this huge question mark that's looming large in the minds of my progressive friends, especially now that Fauci has said everybody's gonna get the thing. Even I'm creeped out by the way that official sources haven't come out with new research showing the phenomenon has decreased; they've just kinda stopped talking about it.
Re Chronic Lyme disease-I don't have the reference on hand, but one of the "Lyme-literate" books reported a credible instance of blood from a mouse that had gone through the standard treatment transmitted Lyme to a previously uninfected mouse.
Just going to copy/paste this from my comment on your peer review feedback version, which addresses your paragraph about Chronic Lyme Disease being psychosomatic:
Hi, I have ME/CFS. I was also diagnosed with Chronic Lyme Disease last year. (But I decided not to believe it, and that was the final straw to stop seeing that healthcare provider.)
Your paragraph about psychosomatic conditions and CLD is framed in a way that doesn't make sense.
First off, it doesn't make sense to say CLD is universally considered 100% psychosomatic. CLD is not a coherent diagnosis! The diagnostic criteria for CLD varies doctor by doctor. (Err, alternative doctor by alternative doctor.) They try to say it's "the great imitator." They say that basically any vague chronic symptom can be a symptom of CLD.
If you meant something like "the symptoms of CLD are all symptoms that are always psychosomatic": that's not true either. Despite my CLD diagnosis, my symptoms are not psychosomatic. (At least to the best of my knowledge, and at least according to my doctors and therapists from the last few years.)
The set of people who "have CLD" ≈ the set of people with vague chronic symptoms who, at some point, visited the type of "doctor" who gives out CLD diagnoses. (i.e. naturopaths, chiropractors, "Lyme-literate" MDs.)
Secondly: the reaction to calling a condition "psychosomatic" is not simply an aversion to that concept. It sucks to be told your chronic symptoms are psychosomatic because, AFAIK, that's generally not followed up with any treatment (or at least any effective treatment).
I have spoken to many chronically ill people who would be _so happy_ to learn their condition is psychosomatic — if it meant they would get some kind of treatment and therefore relief.
Panic attacks cause weird physical symptoms. These symptoms are therefore psychosomatic. People who get panic attacks have to learn to accept that. But that's fine. Because panic attacks are treatable! If the reason for your lightheadedness is "you're having a panic attack," that's useful, because now you can work on having fewer panic attacks.
By contrast, IME, "your fatigue is psychosomatic" doesn't come with any action items. "Your fatigue is due to depression" means you can go get evaluated for depression, take antidepressants, do other things that work for depression, etc, and see if your fatigue gets better. But if a patient is told their fatigue is psychosomatic with no next steps, it's basically an unfalsifiable dead-end claim.
I had an interesting experience with long COVID. (It was being called post-COVID syndrome back in February). I was symptomatic with COVID for two weeks and never recovered over the next two months. I ended up going to the ER 3 times. Doctors ran a million tests and couldn't figure out what was wrong. They told me the diagnosis was a catch-all for "we're not sure but you had COVID and you're still sick so we'll just go with this".
Eventually they figured out that I actually had a super-rare form of cancer called a Ewing's like sarcoma. Which is only one percent of cancer cases in the US annually. I also talked to another sarcoma patient who had a similar thing happen to him this season.
I'm not sure if this is a trend or anything but it would make sense. COVID sucks up so much of the metaphorical oxygen in the room that everyone is naturally going to think about it first if anything weird is going on.
When you do these peer review things it would be helpful if you listed any major changes between versions somewhere so we don't have to read the whole thing again.
Thank you so much for writing about this. Glad to have this sane analysis.
The Tanzania case sounds more like encephalitis lethargica (most known from legacy cases in the Robin Williams, Robert De Niro movie "Awakenings"). The illness was concurrent with Spanish flu, but the trademark for EL was post-infection fatigue.
I just wanted to comment that one theory on post-viral CFS/ME and one that seems to be related to some people's long covid (see this twitter thread for an example: https://twitter.com/ar_pnw/status/1432772600480944128) is that viruses can lead to a degradation of connective tissue which leads to cranial and spinal instability which leads to the long list of CFS/ME symptoms.
There are a growing number of CFS/ME patients being cured by getting surgery to correct cranial/spinal issues, and there is growing awareness that some patients with CFS/ME had connective tissue disorders before getting CFS/ME.
Two cases that anyone interested in CFS/ME (whether they're just interested in it or they have it or know someone with it) may be interested in are those of Jen Brea (recounted here, https://jenbrea.medium.com/cci-tethered-cord-series-e1e098b5edf) and Jeff Wood (recounted here, https://www.mechanicalbasis.org/). It is incredible how these people have recovered, how they may have gotten sick in the first place, and how horribly the medical system can be until you figure out what's wrong with you and how to navigate it.
I don't think the chance of being hit by an airplane is 1/250,000, that would be 1300 Americans per year and that does not seem possible.
Also, I don't think that "What’s my yearly risk of getting COVID if I try to live a normal life?" is a useful question. I think "If I live a normal life, how much does the risk of Long Covid increase?" Or more generally, "If we all go back to normal lives, how much does my risk of long covid increase?"
I'm sorry, but after reading the portion about willpower and me/cfs I am still skeptical and unconvinced. Heron's post, especially read to me as Heron deciding they felt unwell, going to many doctors and undergoing extensive tests to find nothing, and then self-diagnosing that they have an extremely rare condition their doctors haven't heard of that doesn't manifest itself on any testing known to medical science.
Is there any hard evidence that it is a physical condition that wasn't mentioned? Having read the article the impression I got is that it's just a psychological disorder, one that conveniently can't be disproven through medical science
How is chart B from Logue et al supposed to be interpreted? It can't be that mild and severe symptoms are counted separately, because in several cases, the bars for mild and severe symptoms during acute COVIC-19 sum to more than 100%. And it can't be that "mild" means "at least mild", because in several cases, the severe bar is higher than the mild bar.
I don't see this explicitly expressed in your article here, but I've been hearing more and more anecdotes regarding problems in extremities with respect to (possibly) long COVID, ie finger and toe pain and/or numbness, Reynaud Syndrome, etc. Is that just represented as the "tingling" entry in your neurological are of the charts?
There's one argument about long COVID that I never see addressed anywhere, and which makes me believe this is NOT some huge threat that would massively disrupt society and/or warrant indefinite restrictions: there are plenty places in the world that have been, in fact, rather heavily hit by COVID and/or consciously do not implement major restrictions. Certainly there are places where double percent-digits of population had COVID. Yet we never hear about these places particularly suffering from long COVID. I mean, of course there are some sufferers, you might hear some anecdotes, but you'd think if "if a few percent of COVID patients get long-term unremitting genuine CFS/ME, that has the potential to overwhelm government welfare budgets and long-term depress the economy", we should be ALREADY seeing it in some places.
I'm Russian, although currently not resident in Russia, and Russia is my main reference country here: the country seems to have become fatalist about COVID and it stopped being newsworthy many months ago. Vaccination levels are low (25% fully vaccinated) and the vaccine is widely perceived as not particularly necessary and/or beneficial. Some low-level restrictions persist, varying by region, but adherence is low despite occasional police enforcement; masks are very often worn under the nose, if worn at all. Official death rate is close to 200,000 (in a country of 144 million) and estimates based on excess deaths make it closer to 600,000. And despite this all... there's no catastrophe? Life seems to be going on more or less as usual? Long COVID is occasionally discussed in Russian social media but nowhere to the same scale as I see in the Western world. It's not denied, in fact many people know someone who had or is having prolonged symptoms (my own mom got it in June 2020 and it took over half a year for her senses of smell and taste to recover, and it's still not 100% as it was before). The point is, the society is not really being overwhelmed by long COVID at all.
Has anyone asked how much social media of other as funded media usage there is amongst people who report long covid compared to people who don't?
Social media addiction would seem to be a key factor hear in generating a nocebo phenomena.
N == 3 here.
I had a mild case of Covid in December 2020, immediately followed by heart palpitations (something I never experienced before) and neurological symptoms (tremors, intense twitching).
I was given an EKG holter on a cardiological clinic that showed a ton of extrasystoles, and one of the doctors asked me for description of my neurological symptoms. It turned out that this doctor had the very same post-Covid symptoms and was afraid of the possibility that it was actually onset of yet-undiagnosed ALS. After comparing our experiences, we came to the conclusion that it would be highly unlikely.
Sure enough, soon a third person known to both of us appeared with the same neurological problems.
But the really interesting observation: even though the problems persisted for months, they diminished after the first dose of Pfizer vaccine and basically went away after the second one in all of us. Not immediately, more like in a week. I recently got the EKG holter again to confirm. Zero extrasystoles over 24 hours, so back to normal.
So maybe subsequent vaccination has a healing effect on long Covid. That should be taken into account when mapping the long term aftereffects.
Certain people are exposed to covid a lot. I have had covid twice, and it hit me pretty good both times. My lymphocyte levels were really crushed down, and it took a long time for them to come back, and they still have not returned to their previous levels. I'd be curious if there is any data looking at lymphocytes, CRP, and other markers in people with non severe covid. Get blood tests a few times a year, so I was able to track this. The low lymphocytes definitely correlated with me feeling like garbage. (that is a technical term for fatigue, shortness of breath and weakness)
Well, cheer up, it can't go on forever no matter what. Eventually long COVID would have to smoothly merge with what it's like to just be old ha ha. Feeling chronically fatigued -- check. Weird muscle and joint aches for no reason -- yup. Brain and digestion not working as well as it used to -- alas, check.
Pre-apologies for being glib here:
But the mind is part of the brain and the brain is part of the body and all sorts of physical and non-physical stimuli affect and impact any of the mind/brain/body which in turn affects other parts of our kooky and complicated system.
We need LongCovid studies with proper controls and premorbid information, we need LongCovid studies in Japan, in South Africa, in Belarus, in Argentina, in Morocco, etc. We need to understand what this is or could be across a variety of cultures and environments because, as it stands, I simply do not buy it. The Western-Anglo culture encourages and abets solipsism to the point that self-reporting fatigue, as a condition and not a temporary feeling, is simply something I do not buy. Call me callous or naïve but I simply don't buy it.
As somebody with an interest in this area*, I think this is an excellent overview. With regards to ME/CFS, I find the following things relevant:
1) It's hard to understate just how poor a lot of ME/CFS research is, even from mainstream institutions. Rubbish case definition (see next point), tiny sample sizes, lack of blinding, subjective outcomes not matched by objective measures, outcome switching. Certainly on the treatment side (regardless of whether this is a psychological or physical treatment), a lot of it fails the 'homeopathy test' - i.e. 'is this trial design so flawed that it is possible to get a positive result if testing homeopathy using the same design?'. (Special shout out here to the SMILE trial that found a therapy that appears to consist of drawing chalk circles on the floor whilst chanting 'I'm strong like a tree', and that claims to cure all manner of ills including Multiple Sclerosis, to be effective)
2) Case definition in ME/CFS is a mess, and Long Covid seems to be following the same path. Little care is taken to differentiate patients with general fatigue and those with post exertional malaise (aka post exertion symptom exacerbation). If one's fatigue is down to mild depression then an exercise therapy might be a good intervention; whereas for somebody with PEM it might be pretty much the worst thing to recommend. The ONS data in the UK is likely to be garbage because it is just so nonspecific.
Similarly, misdiagnosis is high. Newton et al found that 40% of patients referred to a CFS clinic had alternative diagnoses upon further investigation.
3) On the psychosomatic issue, I hesitate. On one level, it doesn't really matter as to aetiology, if people are suffering. And, regardless of aetiology, prognosis for ME/CFS is poor and the treatments pretty ineffective. On the other hand... medical practitioners like to talk about inappropriate mind/body dualism but one is still referred to *either* neurology or psychiatry departments. I also find is surprising that if it is down to 'false illness belief' that CBT or GET interventions are not more successful in leading to recovery, especially with respect to success rates in treating phobias etc.
Finally, I find the distinction between 'normal post-viral fatigue' and (assumed psychosomatic) ME/CFS really weird. Assume that somebody gets a bad case of glandular fever/mono. If, two months later, they are suffering from fatigue/brain fog despite the virus having been effectively cleared from their system, this is considered 'normal' *even though AFAIK nobody can explain the mechanism as to why they are still experiencing symptoms*. The moment they hit the 3 month/6 month mark (depending on case criteria used), many doctors will confidently decide this is now down to a psychiatric issue. Does this mean that a patient experiencing post-EBV fatigue at two months is also temporarily delusional? I don't think I've ever seen a good explanation of the difference between the two from those who hold a psychosomatic explanation for ME/CFS.
*I started looking at the evidence regarding ME/CFS when somebody close to me was diagnosed with it. It's fair to say that, prior to that, my view (prejudice) was that ME patients were just very angry hypochondriacs.
> But they wrote up their poll in the scientific paper font
No, that's Times New Roman. The scientific paper font is Computer Modern. Big difference.
That "Being hit in your home by a crashing airplane" figure is a couple orders of magnitude bigger than I would have guessed. That means that in a city of 1M inhabitants, 4 of them in average have been or will be hit in their homes by a crashing airplane?
(remembers about [9/11](https://slatestarcodex.com/2016/08/31/terrorists-vs-chairs-an-outlier-story/)) ...oh.
Why not:
- Get a random sample of the population.
- Ask them if they have had COVID.
- Ask them if they have been vaccinated.
- Ask them if they are experiencing any Long COVID symptoms (without mentioning it specifically).
- Test for antibodies.
- Look at the responses of those who said that they have *not* had COVID or a vaccine but *do* have antibodies.
This way, it would be possible to get data on Long COVID incidence from people who've had COVID unknowingly. That group (previously infected + unvaccinated + unaware) may be quite small, but I still think it could be a useful exercise. Comparing their responses to the knowingly COVID-positive group might be revealing.
Count me as surprised that dying from "being hit in your home by a crashing airplane" was about 3x more likely than "drowning in the bath next year". The former sounds like one of those ludicrously unlikely things that only happens in the movies (e.g. Donnie Darko), the latter sounds like it should be one of those stereotypical "causes of death no-one talks about but are actually quite common".
Now I think about it again, the former is lifetime risk and the latter annual, right? That changes the ratio appreciably. Still not what I'd have guessed.
I have to tell you - I find your dismissal of psychosomatic illness, in a context where we've known hypochondria and Munchausen's to be real for a very long time, to be quite strange and to stand in sharp contrast with the typical evidentiary standards of this blog.
"They point out that long COVID was first thrust into the public consciousness in surveys run by Body Politic, who self-describe as “a queer feminist wellness collective merging the personal and the political”. I agree this is a weird source for something to come from, but Hans Asperger was a Nazi and I still use his diagnosis, so I probably have to accept these people’s as well."
Well, "Asperger's syndrome" was a diagnosis in which broadly the same group of children Kanner et al described were described in somewhat more optimistic terms for, ahem, 'sociocultural reasons', which has been deprecated for the better part of a decade, and for which studies of adults didn't find meaningful differences between it and similar cases of high-functioning autism with speech delay, so probably there's a good question about whether you should be.
(In practice, I think Asperger's was a bit different to HFA, but in that it was getting at a more heterogeneous cluster of neurodivergent children and might even have been more comparable to PDD-NOS on that point -- "schizotypal disorder of childhood" pretty much stopped being a diagnosis when AS got popular, for instance.)
This discussion omits one important data point: long term effects from the original SARS virus. I’ve long believed this is a major piece missing from the Covid discourse. The original SARS virus has a ~10% fatality rate, but it also has well documented long term debilitating symptoms. My recollection is that something like 30% of SARS survivors were permanently disabled (I thought I’d seen this on Wikipedia, but the current article doesn’t quote a percentage). From that perspective, it’s quite natural to assume that Covid will also have long term debilitating effects, perhaps at about three times the CFR. This is one of the main points that convinced me of the severity of Covid back in March 2020, and I wish it had become more widely known at the time.
I remember Sarah Constantin had some Twitter thread about why chronic Lyme might actually be a real thing, I can no longer find it though I'm afraid...
The Australian study "Post-infective and chronic fatigue syndromes precipitated by viral
and non-viral pathogens: prospective cohort study" is not that useful for comparison, though.
We have to take into account the myriad of negative health effects of living under lockdown/in a permanent state of emergency?
We would have to compare "Long Covid" to other postviral syndroms as they occur DURING this pandemic. Not how bad they are in 2006. Who cares about 2006?!
My strong prior would be that ALL chronic conditions are worse now. That people's prodromes become syndromes faster, that subclinical illnesses become clinical.
So during the pandemic the health of the non-infected and infected should be considered "compromised" to varying degrees.
Does anyone account for this and actually do the relevant comparisons, assuming the data is available. Or uses a model that accounts for this?
Without that, I have low priors, that any of the discussion on Long Covid is worth taking seriously.
Personal experience on Chronic Lyme.
My wife was having health problems that no one seemed to be able to figure out and she started self diagnosing as Chronic Lyme. I pushed back after tests and one visit to a Lyme literate doctor alerted me to probable chicanery. It took a trip to the ER to realize she actually had a staph infection.
In this case, the word psychosomatic from her PCP was dramatically unhelpful. She had a real problem, just nobody was able to figure out what it was.
The chart with the 1 in 250,000 chance of having your house hit by an airplane seems likely to not be reputable. The source appears to be a 1997 revised 2nd edition of a self-published book called "Up to Your Armpits in Alligators? How to Sort Out What Risks Are Worth Worrying About!" by John Paling. I can't find any evidence the book is still in print or a review of it anywhere except a very old article in a paygated Wiley journal. Paling's bio is here: https://www.riskcomm.com/pdfdocs/JPfullbio2.pdf. It appears he was a zoology lecturer for a while at Oxford, then became a documentary filmmaker for 23 years, then founded something called the Risk Communication Institute in Gainesville, FL, and that was when he wrote the book.
I'm not going to say he isn't reputable or anything. I have no idea. But this is just a rabbit hole game of telephone being played with no way to get at the actual citation, unless you're going to buy the book to find out where he got the number from. It doesn't pass a basic smell test, though. Obviously, nowhere near 1 in 250,000 units of land surface area on the earth contain a house, and airplanes spend a lot of time not over land. The chance that, when they crash, it might happen over a city goes up a lot given most flight paths are between airports and most airports are at least near a city. But most airports are not near the densest part of a city, many have coastal or rural approach routes, flight paths intentionally take those routes to reduce the risk that they might actually crash over populated areas.
This reeks of a guy who probably cherry picked a year where some spectacular crash over a populated area occurred, saw in that instance 1 in 250,000 houses where it happened were hit with debris, and called it a day. It calls into question everything else in that table, too, because the entire graphic is just pulled from his book.
The study that concerned me the most was this recent Lancet article on cognitive deficits after COVID. Not a perfect study by any means (they don't have true longitudinal data), but they are suggesting persistent IQ declines (no indication of recovery 9 months later):
1 point loss for respiratory symptoms at home without assistance
2 point loss for respiratory symptoms at home with medical assistance
4 points for hospitalization without ventilation
7 points if you were ventilated
https://www.thelancet.com/journals/eclinm/article/PIIS2589-5370(21)00324-2/fulltext
In terms of psychology I was interested to note all the various coping mechanism people utilized during the pandemic. One of the bigger ones was discounting any bad outcome short of death.
One thing I wonder about self-reported long-lasting symptoms: a healthy baseline may include some amount of negative symptoms, e.g. you still cough/sneeze on occasion, you are still fatigued on occasion, etc. Yet you consider yourself healthy. Then you get COVID and all these symptoms skyrocket and you are clearly sick. Now your symptoms slowly decrease day by day as you recover. *If they decrease slowly enough, might you never notice you've recovered?* Each day you still feel approximately the same as the day before, and even when you get back to your healthy baseline the symptoms aren't 100% *gone*, so unless you are good at comparing your state to that of a few months ago (which seems hard to me, unless you actually took notes on how often you were tired/coughed/etc in the past), you might now attribute your baseline fatigue etc to lingering COVID symptoms rather than healthy normal?
I shared this article in a group chat and the preview image was a photograph of a bat. Not really sure where it comes from since it's not in the post and could possibly influence a potential reader to the quality of the post
First off, thanks for writing this.
Second, you say that the chance of long covid after a mild, non-hostpital case is about 20%. By “mild” do you mean showing at least some symptoms? Or do you include non-symptomatic cases as well?
What is the age dependence of (the debilitating version of) long COVID? Is it in fixed proportion to age adjusted IFR? i.e. can one take a flat 3X age adjusted IFR as your risk of serious long COVID? If so, then your age adjusted IFR (pre-vaccine) is about 0.05%, and your long COVID risk is about 0.15%, assuming you are in your mid thirties. This is much lower than you would get if you used age averaged figures.
It also seems weird to me that the effect of vaccination on long COVID risk should be so much more muted than the effect on death risk. I wonder if this could be an artifact of age stratification of vaccination rates (assuming risk of long COVID is also strongly age dependent).
This large UK study into long Covid in children, not yet published but reported yesterday by BBC https://www.bbc.co.uk/news/health-58410584 case-controls for reported symptoms to suggest 'between 2% and 14% still had symptoms caused by Covid 15 weeks later'... look forward to the paper when published, and declaration of interest, I work for the funder NIHR (though not in this area).
I had CFS/ME in my 20s after being in the hospital with a virus, It lasted years, but I think I know what happened. TLDR: started a physical problem, the real problem trained my brain, and caused prediction errors that outlasted the actual problem. I think the same thing often happens with other chronic pain conditions.
1. Not a contagious mind virus: First I'm male in my 40s, when this happened I was an undergrad, and I had no idea what it was, people didn't talk about stuff like this in the 90s, and the internet was a much smaller thing, so no one gave me the idea to be sick.
2. Like many woman, I have an auto-immune disease, specifically b12 and iron anemia, rare in 20 year old men. which I had but didn't know about then, but I now require treatment for monthly. (I eat a ton of meat,FYI, but my body attacks cells that produce enzymes needed to absorb it.)
3. I got better after about 2 years, but I didn't know I was better. I was used to feeling tired and achy. so my expectation is I would continue to feel tired and achy. I was making a prediction, it was wrong, and my brain was making the prediction come true.
I think the understanding that people have been trained to feel pain and can be trained to be pain free is helpful and non-moralizing. The initial pain and discomfort is real, the brain learns to expect it, and it becomes a very normal cognitive bias, that exists because our brains rock and are good a learning and making predictions.
One thing that always bothers me about these percent results, eg
<< And even though the study started with 5000 children, so few of them got COVID, and so few of those got long COVID, that the 2.9% turns out to be about five kids. I don’t really want to update too much based on five kids >>
is that you can actually estimate exactly how much to update. But there's not often much discussion on uncertainty in medicine papers.
The question is binary (either they have it or not). The 2.9% are 5 - so we have p*N=5 out of N=170 in a binomial distribution. The expected probability p*N of the event is 2.9%, and uncertainty on that probability sqrt(p*(1-p)/N) is around 1.3%. Using the rule-of-thumb hack that two-deviations correspond to pvalue < 0.05, you end up being able to say with "some confidence" that it's between 0.3% - 5.5%.
That was very back-of-the-envelope, and now we'd have to apply Bayes with a prior. But the short no-calculus story is: the interval is very wide. If you have any prior about that probability with a reasonable confidence interval (eg by what we know from other diseases) you should not change it much.
To illustrate the point, say your prior is 0.95%-1.05% (at two sigma level like above). Another back-of-the-envelope calculation then updates it to roughly 1.16% - 1.20%. I.e. a bit higher but not much. This was computed with a gamma prior for the binomial - because it's easy (see https://en.wikipedia.org/wiki/Conjugate_prior#Table_of_conjugate_distributions). It's actually not very good for modelling small probabilities with large-ish uncertainties, and as a result updates too fast away from zero. A proper discussion would end up giving an even more conservative estimate for the update.
> Unfortunately, having a symptom for psychosomatic reasons sucks just as much as having it for any other reason.
Psychosomatic conditions seem like ripe targets for psychedelic therapy.
I was also interested in whether anyone with CFS had tried MDMA or something. Couldn't find any reliable info from searching online.
> Flu is as common as COVID, but nobody really talks about it having a significant postviral syndrome so probably it’s not that bad.
Or maybe nobody had made this connection. I think COVID has been revealing a lot of long-term associations, like long COVID, that people hadn't previously considered.
Thanks, this is very useful.
On this point:
>>Of vaccinated people who got symptomatic COVID, about a third ended up with Long COVID symptoms, the same rate as in unvaccinated people.
One would expect vaccinated people (in the US at least) to belong to a group that is more prone to health anxiety and especially COVID anxiety than un-vaccinated people. There is so much correlation between vaccination and beliefs about the seriousness of the pandemic that that seems like a major possible confounder.
I'm wondering if anyone has seen research on people having a certain tendency toward this sort of effect? Ie, if you’ve had a post viral syndrome before, are you more likely to get long covid?
Personal anecdote:
As a teen I returned from a trip to Africa with mild but persistent symptoms. Fatigue, headaches, nausea - all of which intensified with any exercise. No acute illness but the ultimate conclusion of the doctors was that I picked up a virus that wasn’t that bad (hence no acute period of illness), but was unusual enough to my immune system (being from another continent) that it freaked out. I was sick about a year.
I am a woman and have vitiligo, an autoimmune disease.
I THINK I had covid early March 2020 but was never tested. It presented as a mild cold but I had fatigue for about three weeks afterwards.
I'm someone who was pretty strongly in the "Long Covid is almost certainly not a real thing" camp prior to reading this article. Now I'm in this, "I still think Long Covid is likely not a real thing, albeit with some exceptions camp" now.
First - Yes, the people with lung scarring and post-ICU syndrome and maybe some of the people suffering the misunderstood post-viral symptoms are truly suffering, but they're not suffering from Long Covid. They're suffering from lung scarring and post-ICU syndrome and maybe some of the people suffering the misunderstood post-viral symptoms. It doesn't make sense to classify these individuals as having "Long Covid."
This sounds like a semantic point, but I really don't agree that it is. There hasn't been a disease that's gotten this much public focus since... the Spanish flu? Definitely not one since our world got connected like this. This is the largest story of the past decade, and large stories like this attract hangers-on.
Maybe I'm cynical, but I think that if we accept that there is a vague condition like "Long Covid" out there that has amorphous symptoms, no ability to test and no ability to cure it, we're going to get hordes of people who (for whatever reason) start to glom onto it.
This piece made me want to read a lot more about psychosomatic illness (ideally from Scott, but open to suggestions if people know of good resources here). In particular I'd like to see someone explain how psychosomatic illness differs from mental illness. My first instinct is to say that both originate in the brain, but mental illness expresses itself in "mental" symptoms (changes in mood, hallucinations etc.) and psychosomatic illness expresses itself in "physical" symptoms. But of course mental illness has physical manifestations as well (e.g. depressed people can have reduced sense of taste/smell). Is psychosomatic illness best understood as a subset of mental illness, or a distinct category?
Related question: when contrasting psychosomatic illness with non-psychosomatic illness, what's the term for "non-psychosomatic"? My first instinct was to use the word "real" but that implies psychosomatic == fake which isn't correct. Scott uses "organic" above but that feels weird too (not sure how psychosomatic illnesses are "not organic").
I may pick up Ross Douthat's "The Deep Places" which deals with his experience with Chronic Lyme. Seems like it's more of a meditation on the nature of chronic/psychosomatic illness than a medical perspective on the issue but it may be illuminating here.
Hey Scott, can you give us an update on how your workouts are going? (I'm not a paying subscriber so if that's paying-folks-only information, by all means ignore this.)
I have some rather personal data points. Putting the conclusion first: I find long COVID to be hard to study, because I maybe had COVID, and maybe long COVID. However the symptoms are conflated with other things. If I were asked if I had long COVID, I could not give a survey a useful "yes" or "no" answer.
I had Mono in 2018. It was bad enough that I took a trip to the ER, and got meds for reducing pain and swelling. Their testing confirmed I had it, days after a previous misdiagnosis of strep throat from inconclusive tests. After my main recovery, I had at least half a year of fatigue. I could do my desk job, but would need to sleep shortly after getting home. It seemed to slowly fade, and was entirely gone somewhere between 9 months and a year later.
I got sick from something at the start of March 2020. (following what were likely colds in December and February...it was a rough few months.) COVID tests were unobtainable. I don't recall breathing problems or loss of smell but it was something more severe than a normal cold, especially the fatigue. My city at that point was one of the few in the US with confirmed community spread. My primary care physician denied my request for an antibody test - indicating that it was company policy that the tests weren't specific nor useful enough. Overall I put the probability that I had COVID as undecidable.
I started lockdown early because of it, so I didn't notice for quite awhile that I had persistent fatigue. I mentioned mono because the new fatigue was exactly like the fatigue I got from mono. However since then I have identified: depression (medical diagnosis), burnout (self diagnosis), and stress induced physical problems (mixture of both, the strength of the symptoms at their peak was eye opening, and I will never have a shred of doubt again that such problems can be very real). I also commonly have headaches caused by neck stiffness caused by who knows what, likely unrelated to any potential COVID. All of this has combined and fed into one another to be one giant bundle of misery, but a combination of medication, help from family, quitting my job, and time have mostly abated the symptoms. The March illness was the kickoff point, though.
FYI--there's a new Ross Douthat book coming out shortly on Chronic Lyme, which Tyler Cowen has already reviewed favorably.
What exactly is psychosomatic? When I was in kindergarten I suffered from severe eczema, the outbreaks started when I started school. A family member - a professional nurse - suggested to my mother that it was psychosomatic. The nurse based her argument/diagnosis on the fact that the eczema started when I went to school, and subsided during the school holidays, and thus concluded that the eczema was a stress reaction. Years later, when I was 25, I learned after testing by an ENT specialist, that I am severely allergic to dust mites. This explained my eczema, as follows: During my kindergarten years the teachers loved to have the children gather around in a circle, sitting on the floor, for sing song -and story time sessions. Though the floors were swept daily, they were not vacuum cleaned. This was South Africa in 1961 after all. Vacuum cleaners were a rarity in those days, especially in rural schools. I think its advisable to be skeptical of any diagnosis that categorizes a physical condition as psychosomatic.
Are those control group numbers for mental health for Amin-Chowdhury et al. normal? They seem rather high.
Perhaps the long mental health problems are from 'Pandemic Exists', rather than from anything particular to COVID. This might also be an interpretation of the increase in mental health problems after the Spanish flu - were former flu patients more likely to be admitted to mental hospitals or was this increase from the entire population?
Scott: what do you see as the endgame for COVID? I believe (courtesy Zvi) that the UK was at 90% seropositivity among its adult population in June, and then it had one spike (which much have left pretty much the entire adult population either vaccinated or recovered), and now it is having another. I am thus increasingly coming to the view that herd immunity is impossible to achieve even with 100% vaccination. So where is the cavalry? Do you think pediatric immunizations (when they finally become available) will be enough to stem the tide? (and what about the animal reservoirs - deer, cats etc?). Or is it just a choice between maintaining extreme precautions (isolation whenever possible, N95 when not) for the rest of your life, or giving up and getting on with life, trusting to the vaccine to do its job in protecting you? What is the plausible argument that says there is a better choice than this available?
As far as causes go, it could be dependent on rare, pre-existing antibodies that enhance the disease:
https://www.nature.com/articles/d41586-021-02337-5
There’s an interesting treatment in Therapeutic Plasma Exchange (the same technique used in recent mouse aging-rejuvenation studies (which amounts to mechanically filtering out the immune system and any accumulated damage to the blood)) which has shown promise in the treatment of long covid: https://www.youtube.com/watch?v=nnv0fcr_Hd4&t=381s
(And a list of studies here: https://www.apheresis.org/page/covid#therapeutic )
If Long-COVID is caused by the presence of something in the blood plasma (like a “rogue antibody”), then I hope these studies will help us get to the bottom of it.
I suspect you would not have seen this as of when you were drafting this piece, but this article in the NYT seems encouraging?
https://www.nytimes.com/2021/09/01/health/breakthrough-infections-long-covid.html
"Breakthrough infections are less likely to lead to long Covid, a study suggests."
Based on a study just released in the Lancet.
https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(21)00460-6/fulltext
https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(21)00460-6/fulltext found that vaccinated breakthrough cases are 50% less likely to report symptoms persisting past 28 days compared to unvaccinated cases.
I had an extremely bad bout with flu turned into bronchitis turned into pneumonia when I was 20. I was stupid and arrogant with the illusion of invincibility that comes with youth. I figured I'd just eventually get better and made no specific efforts to do anything about it. I stayed out late, kept clubbing, kept drinking. I ended up remaining sick for over 6 months until I finally had to go to the hospital and get blasted with high-dose antibiotics and corticosteroids after some nebulizer treatments to clear up my lungs. After the symptoms finally went away, I had enough of an immune system drag that I caught another flu just two weeks later. That one wasn't quite as bad.
But it took a really long time to fully recover. I'm not just talking about persistent fatigue. I actually don't remember anything like that. The actual problem is I caught another flu, and not just flu but ended up with secondary bronchitis, every single flu season like clockwork. It made no difference if I got a vaccine or not. Every single flu season I'd catch flu, and every single I caught flu I'd get bronchitis. I also developed allergic asthma responses to grass pollen.
Until I didn't any more. Since about 32 or so, I haven't had bronchitis a single time. No more asthmatic allergy responses.
I think something like "long flu" is a perfectly apt way of describing this. I had a drastically weakened respiratory and immune system for over a decade, but it wasn't actually permanent. At least, I don't think it was. I have no idea if I have lung scarring or reduced capacity compared to before, but I don't really care. It's the more explicit symptoms that kept killing me, and they've been gone a while now.
I'm wondering if you may have linked the wrong article for "weird x-rays" by mistake? I was expecting an article about x-rays and lung scarring, but it links to "What Is the Evidence for Post-COVID Fatigue?".
I'm really annoyed that the BMJ risks table doesn't directly cite any sources and doesn't give any indication of who the stats apply to. I guess we can assume it's about the US? And they all specify a timeframe (which is good) except the most interesting one, the crashing airplane. I guess that one is easy to guess too (it's about lifetime risk... I assume) but they shouldn't leave it up to readers to guess the context, or to guess which if any of the article's five references back up the numbers in the table.
I have nothing pertinent to add, but I wanted to say I like the bat picture 😀
https://www.metafilter.com/192514/Long-Haulers-Are-Fighting-for-Their-Future#8143227
A comment about the difficulty of getting a diagnosis if the pain is the most obvious symptom.
****quoted in full****
*Many have been told by medical professionals that they’re just having anxiety or making up their symptoms. Even now, “it happens more often than not,” Lisa McCorkell of the Patient-Led Research Collaborative told me.*
Somebody needs to do a study on how many people actually make up their symptoms to get attention from doctors. Because according to doctors, it's an epidemic of extreme proportions, a plague ravaging health care services around the world, and it's taking up enormous quantities of their time that would better be spent elsewhere.
Let's ignore for a moment how self-serving that sentiment is; doctors are convinced that people will fake a huge range of symptoms just to be prodded for a few minutes in an exam room. I'm not sure what they think is so outrageously wonderful about that experience that people will commit to acting miserable to get in there, but they're pretty certain that a vast swath of the population is malingering performatively to just bask in their presence.
On the other hand, if you go to any message board dedicated to any chronic disease, one of the most popular conversation is 'how long did it take to get your diagnosis.' Patients trade stories of how often their symptoms were attributed to attention-seeking and malingering until they were able to convince the right doctor to do the right test, at which point they got a diagnosis. And that's for diseases that have definitive tests available; I can't imagine how bad it is for diseases which are diagnosed through exclusion.
I walked around with a painful hernia for four years. I'd been diagnosed with Crohn's, and every doctor I went to said the pain was Crohn's-related, and that there was nothing they could do about it, and I was in remission so it couldn't be as bad as I was saying. I was told by one GI doc that the pain I was feeling couldn't be related to the hernia that showed up on my CT scan. That same doctor put a diagnosis of hypochondriasis on my chart. Over a year later, a different gastroenterologist noted that my pain was in the inguinal area, and had I considered that it might be a hernia?
And the test for a hernia is, you know, 'turn your head and cough.' Once a general surgeon did that test, he had me set up for surgery two days later.
The post-surgical pain from the hernia was less than the pain from the hernia itself within 24 hours.
When huge quantities of patients have to go through an ordeal to get relatively common, run of the mill conditions diagnosed, there's something deeply broken about the medical system. I keep thinking about every doctor I saw during those years I was experiencing enormous pain from an easily curable condition, and how they're all still convinced I was making shit up, and I'm struck by the fact that there's no way for them to learn from these mistakes. Doctors have no idea how many of their diagnoses were later revised or contradicted by other doctors, so they have no way to learn which patients weren't desperate for their attention, but for medical attention.
And now that we're seeing a disease that's difficult to diagnose spreading in this new plague, the condescending, dehumanizing, contemptuous chant of 'it's all in your head, it's anxiety, it's depression, go away' (never accompanied by a psych referral, strangely), is ringing out loud and clear. And until we figure out objective ways to measure things like pain and fatigue, they're going to keep dismissing genuinely sick people as attention-seeking malingerers unless we find a way to let them learn from their mistakes.
posted by MrVisible at 10:50 AM on September 2
Great post, but it's not so convincing in addressing the concern that seropositive and seronegative people have the SAME long-term.
It wasn't just WSJ opinion; NYT ran one too, pointing out:
NYT: "One peer-reviewed study of people who reported long Covid symptoms noted that most of those who were tested for antibodies that provide evidence of a previous SARS-CoV-2 infection had negative results. The level of symptoms, moreover, was virtually the same whether the person was positive or negative for antibodies. A second study, not peer reviewed, of adults referred for long Covid management similarly reported that no Covid antibodies were found in 61 percent of them, again without differences in symptoms whether testing was positive or negative."
https://www.nytimes.com/.../opi.../long-covid-treatment.html
Scott responds to this idea with:
"I feel like some of the case-control studies above, which clearly show that seropositive people have higher rates of Long COVID than seronegative people, are pretty convincing here."
Do they? And why then do they conflict with these two other studies find no difference?
The conflict between those studies seems like a crucial question to get to the bottom of, in establishing existence/non-existence of "long Covid" beyond the obvious respiratory/smell aspects. I plan to dig into it, but thought I'd mention in case others have answers already!
So, I have a question about the sermon re: CFS. If most people who haven't suffered from CFS severely underestimate how bad it is, then wouldn't this tend to lead to systematic overreporting of CFS? Like, if someone feels tired all the time because they have a demanding job, 2.4 children, and don't get enough sleep, isn't it possible that they might go to their doctor complaining of "chronic fatigue"? Or if they get COVID, are later surveyed about their symptoms, and the surveyor asks about chronic fatigue, they might say "yes, I've had that", even though their actual symptoms are much milder and more treatable than CJ and Heron describe. If there's a prima facie problem with this story, I can't see it. Do the numbers for chronic fatigue rely on something better-screened than self-reports? If not, wouldn't the systematic underplaying of chronic fatigue in our society lead to some people mis-self-diagnosing regular tiredness as clinical chronic fatigue?
I must confess (as one does after a sermon), I've been guilty of underestimating how bad chronic fatigue is in the past. But my updated world model doesn't necessarily include assuming everyone has the very bad symptoms CJ and Heron are talking about just because they use the words "I have chronic fatigue". But it's also possible that I'm so eager to avoid overcorrecting that I'm undercorrecting. Are there any good estimates of the prevalence of CFS either post-COVID or just in general based on some diagnostic criteria that aren't self-reports, so I can get a handle on the actual numbers beyond "exists" or "doesn't exist"?
I think we may need to re-evaluate these numbers since we're now starting to deal with the delta variant and it may skew these numbers upwards if you're not vaccinated.
> maybe women’s immune systems are permanently upregulated to be prepared for attempts by the placenta to secrete immune-downregulating chemicals during pregnancy
This doesn't totally make sense to me, since women's immune responses vary during the menstrual cycle and start *decreasing* before ovulation https://helloclue.com/articles/cycle-a-z/the-immune-system-and-the-menstrual-cycle
For me the question is less "why are women's immune responses so strong?" than "why are men's so weak?" Like, what are the costs to a strong immune system? Aren't autoimmune disorders only a problem because of some (unknown?) features of the modern environment? And even today, arguably women get the better deal: yes, women get more Long COVID, but when men get COVID they're the ones who are more likely to end up in the ICU/dead. So why hasn't evolution built a stronger immune response into men?
[End contribution to the discussion; commence brain dump of things I find interesting/puzzling vaguely in the space of sex difference in immune response]
Wild speculation, but: is the fact that women are 1.5-2x as likely as men to be depressed *and* have stronger immune systems evidence for the "chronic psychitis" view of depression? https://slatestarcodex.com/2015/01/05/chronic-psychitis/ Anyone know of research on this?
Women also have had more COVID vaccine side effects, including (anecdotally) really weird periods (e.g. super long, heavy, or happening post menopause; self-reports in this thread: https://twitter.com/KateClancy/status/1379847815636135941) The stronger side effects obviously make sense because of stronger immune systems (and susceptibility to psychosomatics too), but this makes me realize that I don't understand why you wouldn't get post-viral syndromes after vaccines. I get "poorly understood," but I would have naively thought that in one way or another these syndromes are caused by your immune system never quite getting back to baseline after an infection. Why wouldn't that happen after the vaccine triggers an immune response too? Why aren't we getting Long Post-Vaccine-COVID? Can anyone who understands biology better than I do help me out?
My bottom-line assesment of the draft version was: An average unvaccinated adult infected with COVID will suffer an average:
~5 days of Acute COVID
~20 days of Long COVID
~20 extra days of Being Dead
Being vaccinated probably knocks all of those numbers down, maybe by a factor of two, on top of the very substantial reduction in the probability of being infected in the first place.
The severity of Long COVID symptoms seems particularly hard to assess, but anything close to "I wish I were dead" seems to be extremely rare. Therefore, the risk of COVID can be properly rounded off to the risk of dying from COVID, plus a small correction factor that probably shouldn't tip the balance in most risk:reward calculations.
As someone who has CFS, I am a little relieved that the Long Covid odds seem lower than I feared, but I gotta say that the way section 4 presented it read as "it's bad, but hey, at least it's not CFS!"
What you wrote in section 5 was excellent though, and your compassion shone through, but since section 4 came first it did feel like a bit of a gut punch.
While I'm here and thinking about it, did any studies try and quantify fatigue levels? There's a huge gap between "bed-ridden", "Can't work", and "life significantly affected, but can still work (perhaps only part time)", but all of those describe some fraction of CFS patients....
I had the good luck of having a series of very stressful life events fuck me up, followed by a bad injury, followed by a stressful job.
This put me in an uncomfortable situation of having to meet my life commitments while suffering through a onslaught of severe headaches, fatigue and generally just wanting to die. This I was able to do... But only just very barely. This kept me miserable for many years, and I'm only recovering now.
I wonder how many people are in similar scenarios. No doubt there are some viruses that produce similar effects ad infinitum. But you can be swimming comfortably, and a few waves can have you gasping for air and desperately treading water. Once you're in the life hole, life has a way of keeping you there.
Can we discuss the risk communication here that somewhat trivializes this when there are small rates but large populations affected? It's IFR all over again.
Based on Scott's estimates and our current policy trajectory, 50+ million people would have severe debilitating lifelong COVID.
How do I arrive that?
> 8. Your chance of really bad debilitating lifelong Long COVID, conditional on getting COVID, is probably somewhere between a few tenths of a percent, and a few percent. Your chance per year of getting it by living a normal lifestyle depends on what you consider a normal lifestyle and on the future course of the pandemic. For me, under reasonable assumptions, it’s probably well below one percent.
Given that the current public health policy approach to the pandemic (which I oppose) is essentially that we focus on vaccines (but with a slow pace globally given vaccine apartheid) and that everyone otherwise can self-protect while waiting for or declining vaccines, with an explicit or implicit assumption now with Delta's infectiousness that "everyone will get it." Again, I disagree with the fatalism here but that is the apparent approach given waning use of NPI policies. And I also think a harmful assumption is that people will get it once and that the level of vaccine- and infection-acquired immunity will hold, when in reality reinfection and breakthrough infections are known, and the risks with waning immunity are more of an unknown.
Regardless, let's assume everyone is infected once, 25% of humanity is fully vaccinated is now, 75% is not, and another 25% will be vaccinated before catching COVID. And let's assume Scott's rate ranges and simplify to 1% with COVID will get long COVID and vaccine halves that.
If there are hundreds of thousands of people in the US with truly debilitating Long Covid, where are they?
Just an unsubstantiated general point-- it seems like every account I see of diseases where exhaustion is the major symptom is about someone who was previously high energy-- worked for a lot of hours, family and social life, and athletic, too. All of the sudden, they're flattened.
Surely being sort of average (fried by long hours, netflix is primary hobby) can't be protective. Or is it? Maybe exhaustion diseases are more of a risk for people who are frying themselves. Maybe it's only high-energy people who have enough to spare to write about their problems when they're exhausted. Or what?
Or is my sample biased by media preference for drama?
What or how do you view this relating to the suggestive IQ research done? (https://www.thelancet.com/journals/eclinm/article/PIIS2589-5370(21)00324-2/fulltext)
I find the potential for moderate Covid cases to have long-term IQ impact to be one of the most concerning features of Long Covid.
Sex differences in autoimmune disease:
Chronic autoimmune diseases are tilted towards women, but diseases the follow from an acute autoimmune attack are, if anything, tilted towards men. Everyone describes juvenile diabetes as autoimmune and it is mainly men. The literature describes narcolepsy as autoimmune, and it has even sex ratio, maybe tilted towards men. Greg Cochran describes schizophrenia as autoimmune and it is mainly men. Narcolepsy and schizophrenia occur earlier in men, but diabetes earlier in women.
Scott, not sure if you saw, but this study about long COVID and vaccine efficacy against it came out on the same day you published this post (Sep 1):
https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(21)00460-6/fulltext
"We found that the odds of having symptoms for 28 days or more after post-vaccination infection were approximately halved by having two vaccine doses. This result suggests that the risk of long COVID is reduced in individuals who have received double vaccination, when additionally considering the already documented reduced risk of infection overall."
One way to test how much of long covid is Psychosomatic would be to see if Republicans or Democrats are more likely to have the condition. Or use an other proxy(e.g occupation,religion,population density,zip code). And or asking directly how bad they think covid is and how long the symptoms last before they contracted covid.
> Hans Asperger was a Nazi and I still use his diagnosis
Is this a US/UK difference? I thought at least in the UK we called it all Autism Spectrum Disorder and didn't use the term Asperger Syndrome anymore.
your risk of breakthrough covid is probably really really serially correlated
Immunocompromised people make up a huge proportion of breakthrough infections; almost half in one study, despite being like 3% of the population.
If they get vaccinated at the same rate as everyone else then, by Bayes' rule,
P(bt|ic) = P(ic|bt)P(bt)/P(ic)
so that means if 1% of vaccinated people get breakthrough infections each year then vaccinated IC people have about a 1/6 chance of getting covid in a particular year(!); if 10% of vaccinated people get breakthrough infections in a year, then we're looking at more than one infection per year for the average IC person(!!!)
the law of total probability or something means that a vaccinated non-IC person has a breakthrough infection risk of
p(bt|~ic) =(1 - P(ic|bt))p(bt)/p(~ic)
ie about 0.5% if p(bt) = 1% and about 5% if p(bt) = 10%
this is assuming IC people are vaxxed at the same rate as the general population, which maybe isn't true. if every IC person in the US is vaccinated right now, then P(ic|vax) = P(ic)/P(vax) ~= 6% of the vaccinated population. then the IC get infected at a rate of about 8%/yr (at a 1% breakthrough rate) to 80%/yr (at 10%), while the non-IC get infected at about 0.5%-5%/yr
this suggests to me that probably 10% is too high, because otherwise there'd be sensationalistic news articles about Guy Who Got Covid 3 Times After Being Vaccinated and i haven't heard about that at all
obviously you might end up immunocompromised in the future, but it's obviously strongly serially correlated. so your estimated lifetime risk should (I think) be a lot closer to whatever you get by cutting your annual risk in half, assuming you're not currently IC, and close to 1 if you are
breakthrough infection risk probably varies with a bunch of other serially correlated stuff like age and sex, too, but i find it hard to believe any of that is going to make a big difference compared to being immunocompromised or not. previous covid infection might matter? but idk
all calculations were done before 10 am so i cannot be held responsible for any errors
Great article as usual.
Is it possible that patients diagnosed with CFS/ME don’t get better (as compared with people informally deemed as having postviral syndromes) because they are more likely to be medicated for it (because they and their doctor have decided it’s serious, won’t be going away on its own, and medication for it will be covered by insurance if a formal diagnosis is on file)?
I ask because I have known many people who, once prescribed Adderall for fatigue or a similar malady, have a hard time functioning without it—which would certainly make them answer on a survey that, at present, they would function far suboptimally if they stopped taking stimulants. Both because of downregulation, and because once you get used to a new baseline with a stimulant (including caffeine or nicotine), the prospect of living without that stimulant certainly could register as a chronic disability.
If that stimulant is caffeine or nicotine, we say this as a joke (but deep down still mean it); whereas if it’s a prescription drug, we are being literal. (Even if originally we were only going along because otherwise our medicine would cost five times as much.)
The question isn't really "do people suffer long-term effects after COVID infection" but rather "is there a specific 'long COVID' that's different from other viral infections?". And the answer to that - with the exception of anosmia - pretty much seems to be no.
“Getting hit in your home by a crashing airplane”?? 1 in 250,000!? No way. That would be 1200 people in America alone dying from this every year. Whoever wrote that is completely full of shit.
Re. Chronic Lyme Disease, the Wikipedia page you linked is constantly monitored by anti-CLD fanatics who reverted each of my attempts to make the page less biased within hours. Every one of the Lyme studies that the NIH and the CDC list on their webpages to dismiss CLD is so fatally flawed as to be completely valueless. One, for instance, "proved" that chronic Lyme doesn't exist by showing that none of the patients in a set of patients who complained of chronic Lyme /and tested negative for Lyme disease/ tested positive for Lyme disease.
The suppression of CLD is the only actual legit medical conspiracy I know of, where a bunch of people packed the IDSA panel on Lyme, deliberately excluding anybody who believed in chronic Lyme, and never considered any of the papers which suggest that chronic Lyme exists. It was a total kangaroo court, rigged from the start.
A scary thread. Maybe you could do an update on Long COVID risks? Please inject some hopium to my veins. https://twitter.com/IanRicksecker/status/1478611650760437765
Would pay money to see how this analysis should be updated in light of the much more rapid spread of the omicron variant, especially that final "below one percent" number.
Long covid still seems like this huge question mark that's looming large in the minds of my progressive friends, especially now that Fauci has said everybody's gonna get the thing. Even I'm creeped out by the way that official sources haven't come out with new research showing the phenomenon has decreased; they've just kinda stopped talking about it.
An update might be good in light of the French study showing nearly all symptoms correlated to belief in covid, none but anosmia from actually having had covid: https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2785832
Which was a replication of some German research showing the same thing, can't find the original study though: https://twitter.com/andrew_croxford/status/1473408952436240392?fbclid=IwAR0ZhCEf-0B2Dz4wg-pDQVwg-ecPckUTacepPOAc9yERkaMQ75nIBAvlTF0