What is long covid?
I’ve been getting questions about long covid and my standard answer has been that I don’t think it’s any different from post-viral syndrome, a condition that affects some people after a viral infection but that usually clears up within a few months. I’ve been generally sceptical of claims of long covid as some distinct entity for a couple of reasons.
First, it’s generally impossible to separate out cause and effect from anecdote and observational data. If you have a respiratory infection and then you continue to feel tired afterwards, was it the respiratory infection that caused it or something else? Was it really after the infection that you started to feel more tired, or had you been feeling tired for an extended period of time, but the infection gave you something convenient to blame it on?
To me, long covid has felt a lot like amalgam poisoning, electricity allergy, and chronic Lyme disease – i.e. conditions that some people diagnose themselves with (doctors rarely diagnose them), but for which there are no diagnostic tests, and for which there is no scientific evidence. In other words, I’m inclined to think of it as a phantom condition that is used in place of the real explanation for how a person feels, whether that be depression, an anxiety disorder, sleep problems, alcohol abuse, or something else. For those who have been seriously sick with covid and required intensive care, PTSD is also something that could easily be misinterpreted as long covid (PTSD is actually quite common after intensive care, affecting one in ten patients). Apart from that, some studies have found that almost 60% of people treated in intensive care (for any condition, not covid specifically) still have cognitive impairments twelve months after being discharged – obviously, if you’ve had covid you’ll tend to attribute that to “long covid”, when it is actually a consequence of severe illness and ICU treatment more generally.
Second, covid is not some magical entity, it’s a coronavirus, and it behaves like other coronaviruses, and other respiratory viruses more generally. It would be strange for covid to cause symptoms that other respiratory viruses don’t. And since I’ve never heard of “long rhinovirus” or “long influenza”, I’m inherently doubtful of claims that there’s such a thing as “long covid”. That being said, I understand why patients might want to diagnose themselves with long covid, and why media might want to write articles about long covid.
One thing that is clear about long covid is that it is a social media phenomenon. The disease wasn’t discovered by doctors or scientists. It was “discovered” by people who found each other on the internet. In that sense, it shares something very much in common with conditions that most doctors agree are bogus, but for which there are aggressive patient advocacy groups, such as the aforementioned amalgam poisoning, electricity allergy, and chronic Lyme disease.
But maybe I’m just letting my biases get in the way. So, I decided to try to see what the published literature had to say about covid. That turned out to be easier said than done. My Pubmed search for “long covid” didn’t generate a single article. My searches on Google scholar and medRxiv did generate a few hits, although most turned out to be opinion pieces, not scientific research. There were a few studies of interest though, which I will spend the rest of this article discussing.
The
British National Institute of Health Research organized a focus group earlier this year with members of the “Long covid” Facebook group, and the results were published in October. Members of the focus group described symptoms “moving around their bodies” and “coming and going”, and described pretty much all different types of symptoms from every different organ. Even symptoms from the urinary tract were identified as symptoms of long covid.
To me, this is clear evidence that long covid is not one thing, but rather many different things. It’s basically whatever the person who thinks they have it says it is. Anything and everything can be attributed to long covid. Here is a quote from the study, by a woman who has diagnosed herself with long covid:
My journey with Covid19 began on 27th April. I’m still unwell five months on and haven’t been able to resume a normal life since. My worst and scariest experience with this illness was in week 6, when I was rushed to A&E as I had a sudden relapse of symptoms and found myself gasping for air, with the top of my head numb and tingling and a headache so blinding that I couldn’t keep my eyes open. I got worse in the hospital and was shaking visibly, so much so that the nurse couldn’t perform an ECG as I just couldn’t stay still.
Despite having been diagnosed with suspected Covid by my GP and a doctor in a Covid clinic (swab testing wasn’t available to the public at the time) and told I had pleurisy during a visit to A&E two weeks earlier, the doctor on duty didn’t take this into account. Instead, he dismissed me with anxiety, advising a course of anti-depressants, and chose not to investigate these concerning symptoms further. Of course I was anxious, but that was a consequence of the physical symptoms, not the cause!
To me (and I suspect other doctors reading this) it is pretty obvious that the woman was having a panic attack, which the doctor in A&E diagnosed correctly. But she is completely certain that what she had was a “long covid” attack, even though, as is clear from the text, it isn’t even certain that she ever had covid in the first place.
On MedRxiv,
there is a pre-print awaiting peer review of a prospective cohort study that followed 4,182 people with positive PCR tests for covid over the course of a few months, to see what symptoms they had, and how quickly they recovered. Participants reported their symptoms in a phone based app.
So, how fast did people recover from covid? 86% had completely recovered within four weeks. At the eight week mark, that number had increased to 95% and by twelve weeks 98% said that they had recovered fully. So, if we assume that this study was reasonably accurate, then one in 50 people who get covid still have symptoms at the twelve week point, at least according to their self-estimation. Unfortunately the study didn’t go on longer than that – it would have been interesting to see how many still felt they had symptoms at the six month mark, to really get an estimate of what the prevalence of long covid is.
The most common symptoms in people with long covid (defined in the study as still having symptoms after four weeks) were fatigue (98%) and intermittent headache (91%). These are both extremely unspecific symptoms, i.e. there is nothing about them that is specific for covid. I’ve had a couple of headaches personally in the last month, probably due to excessive caffein consumption and sharing a home with two screaming toddlers. If I thought I had long covid, I might have attributed the headaches to that instead.
So, what can we conclude from this study?
First, long covid is rare. Around one in fifty people still have symptoms at the twelve week mark, and since the number with symptoms dropped significantly at one, two, and three months, it is likely that the reduction continues after twelve weeks, and that it is a tiny fraction who still have symptoms at six months.
Apart from that, symptoms of long covid are extremely unspecific, so it is probable that long covid is actually a whole bunch of different things, of which I would think post-viral syndrome is likely a significant part. Considering the media fear-mongering going on during virtually all of 2020, I wouldn’t be surprised if many of the so called long covid cases are actually suffering from an anxiety disorder that has been exacerbated by the media response to the virus.
I’m going to finish up by discussing another pre-print currently up on medRxiv. This was another
cohort study in which 201 individuals with continuing symptoms four months after infection underwent an MRI of the chest and abdomen to see if there were any signs of “organ impairment”. Now, this study has so many problems that I seriously debated with myself whether to bother writing about it, but for the sake of completeness I decided in favor, especially since it is the only study so far that could be claimed to provide hard evidence for long covid as a distinct entity. If nothing else it will provide a good education in how to use the mantle of “science” to manipulate people so that you can sell more product.
The study was funded by the National Consortium of Intelligent Medical Imaging (NCIMI), which sounds very progressive and nice. This organization is of course in turn funded by multiple companies involved in producing MRI machines, such as General Electric, Alliance Medical, and Perspectum. As it happens, one of the authors of the paper is the CEO of Perspectum. I’m sure you can see where I am going with this. The study was run by people with a strong financial interest in getting hospitals to increase their use of MRI machines. What better way to do that than to “find” a bunch of damaged organs, which no-one would have ever known about if not for MRI?
Ok, so, what they did was this. They got 201 patients who had been diagnosed with covid, either through a positive PCR test, or a positive antibody test, or by having two clinicians independently decide that they had covid, and who still had symptoms at four months. As an aside, those must have been some pretty incredible clinicians, considering that there are no symptoms that are specific for covid, as distinct from other respiratory viral infections. The average age of the participants was 44 years.
The researchers then shoved the patients (metaphorically, not literally) in to an MRI scanner, and scanned their lungs, heart, liver, kidneys, pancreas, and spleen. They then analyzed the images, and compared them with standardized “healthy” reference values. What did they find?
32% of participants had signs of impaired heart function. 33% had signs of impaired lung function. 12% had signs of impaired kidney function. 10% had signs of impaired liver function. 17% had signs of impaired pancreas function. And 6% had signs of impaired spleen function. Overall, 66% of participants had signs of organ dysfunction in at least one organ.
That sounds pretty awful. So what is the catch?
The catch is that they didn’t compare the patients to a control group. This is a standard trick when you want to make something seem really bad, since readers will naturally assume that if there had been a control group then 0% in the control group would have had signs of impaired organ function. And 32% with signs of a heart disorder on MRI is a lot more than 0% .
However, that is actually extremely unlikely. For all we know, more people would have had signs of impaired heart function in the control group than in the long covid group. Since this study didn’t include a control group, it doesn’t tell us anything. The study is useless.
And, as an aside, the study found no correlation between the symptoms the patients had and the findings on MRI. So, double useless. The MRI findings were completely incidental.
So, that pretty much sums it up. There is no hard evidence to support long covid as a distinct disease entity, and the wildly varying, non-specific, and intermittent symptom picture suggests it is actually a conglomerate diagnosis that is being used to describe a large number of different disease entities, and which is being used by politicians and the media in an attempt to scare the hell out of people. Regardless, 98% of people with covid have recovered fully within three months.
Maybe it will turn out that long covid is a real entity after all (distinct from post-viral syndrome, PTSD, anxiety disorder, and so on) when better research is done down the line, but we can’t just assume it based on anecdote, fear-mongering, groupthink, and low quality science. That doesn’t benefit anyone, least of all people with other underlying health issues that are not properly investigated because it’s so easy to just blame everything on covid.
You might also be interested in my article
about how accurate the covid tests are, or my article
about how deadly covid is.