Losing a whole year — my Long COVID story
There are many who think Long COVID is fake — basically a societal hallucination. On the other side of the spectrum there are those who think that Long COVID is a lifelong condition for which there is no known cure.
The truth is somewhere in between. Long COVID is a real, devastating illness. However, most people recover.
This is my personal story, for anyone interested.
I recovered 75% in around a year, and 90% in two years.
The tldr points I was hoping people might take away from this
- This illness was very real. I was severely ill for six months, during which I was unable to work and had many days where I was completely couch-bound. The fatigue was not just from depression, it was much more intense, physical fatigue. I also had a month of coughing, a month of tinnitus, a month of sharp pain in my chest, weird muscle aches, and insomnia.
- I recovered from Long COVID after about a year. It took another year to get a handle on periodic limb movement disorder and minor restless legs syndrome which were disrupting my sleep quality.
- The only things I am somewhat confident helped with my Long COVID recovery is graded exercise therapy (GET). To a lesser extent, cognitive behavioural therapy techniques may have helped as well (they go hand-in-hand with GET). These methods need to be applied with a lot of care, and are not cure-alls, but I believe they can help at least a subset of some Long COVID sufferers. Unfortunately, both techniques are extremely controversial and pretty universally reviled within the chronic fatigue syndrome communities online.
- While I found psychological techniques helpful, I do not think my Long COVID or most Long COVID is purely psychological or psychosomatic. For various reasons, some of which I discuss below, I believe Epstein-Barr reactivation and persistent virus played a major role. This is in addition to neurological effects (“false fatigue alarms” / bad conditioning of the fatigue-generating circuitry) and what we traditionally call “psychological” factors (anxiety, depression, insomnia). There may have been mitochondrial damage and immune system disregulation as well.
- For most people this is a complex syndrome with biological, psychological/psychosomatic, and sociological factors. I have concluded that the “biopsychosocial” framework is the right framework to use when trying to understand Long COVID. The biopsychosocial model can really be applied to any illness. This is because biology, psychology, and society are all fundamentally intertwined. However I think it is particularly important to use this framework when trying to understand very complex syndromes with vague and varied symptoms sets like Long COVID. Unfortunately the biopsychosocial model is not very popular in Long COVID or chronic fatigue syndrome communities online. This may be due to its inherent complexity. People prefer to argue that something is all biological or all psychological/psychosomatic or all culturally-bound. Unfortunately, reality is never as simple as we would like. In the words of Scott Alexander, “that’s not how any of this works.” I chose to think and write about the psychosomatic aspects when I had Long COVID (I am a long time CBT fan). I have one article on the “false fatigue alarm” theory for Long COVID fatigue, one about psychosomatic factors (also discusses biological factors), and one about treating the pyschosomatic factors. However, I think 75% of Long COVID symptoms are biological in nature, probably from some mix of persistent virus, mitochondrial damage, and immune dysregulation.
Anyway, here is the story. A lot of this was written around November 2022.
By the way, if you like this story, please considering subscribing to my Substack. My Substack is where I do more detailed analysis of things around metascience, My Medium is reserved for more personal stories.
Predisposing factors
The CBT framework for chronic fatigue talks about predisposing, precipitating, and perpetuating factors.
I think there were a number of factors that put me at risk towards getting Long COVID. Firstly, I’m a bit of a perfectionist. Many studies find an association between perfectionism and chronic fatigue.
Next, I have struggled with fatigue previously. In the summer of 2018 I had fatigue for several months which was partially due to mild depression but which I suspect may also have been due to black mold build up in my AC unit.
Then between December 2018 — March 2019 I suffered from about five months of unexplained and very persistent fatigue. It was bad enough I consulted three different doctors and got several blood tests done. As with many others with fatigue, I experienced at least one doctor try to trivialize or dismiss my illness. The fatigue was not nearly as bad as what I later experienced during Long COVID but it was very persistent and didn’t have any obvious explanation. The first doctor who I saw in January hypothesized that I had had “mono”, also called glanular fever, recently. Mono is caused by the Epstein-Barr virus (EBV). Indeed, I had been sick earlier, I believe in November (if I recall correctly). A test for EBV antibodies did not show that I had an acute EBV infection. However looking back at the test results, I do see that I had an unusually high level of EBV “early antigen”, which suggests either a recent EBV infection or re-activation. Unfortunately, the fatigue contributed to an episode of major depression which went from May — November 2019. Eventually I recovered, but the entire ordeal of fatigue + depression took about a year to get over from.
Now let’s fast-forward to just a few months just prior to when I got COVID and Long COVID. Looking back I realize now that my anxiety was way too high during that time. If I could go back in time I would have told myself to get a therapist and psychiatrist. I started dating someone in April 2022 and before dates I would get anxiety about sleeping. I was very meticulous about sleep hygiene, so I generally had no trouble falling asleep. However, I would wake up in the middle of the night after 4–6 hours and not be able to return to sleep. These sleep issues continued through the spring and persisted even after that relationship ended. Going into June 2022, I was not sleeping well 1–3 nights a week. However, critically I was sleeping fairly well the other nights. So, I was not accumulating a huge sleep debt as I would later when I got Long COVID. The main way I was able to achieve good sleep was by engaging in vigorous exercise 3–4 days a week (weightlifting or running). When COVID hit, I was not able to exercise, which caused my sleep to degrade.
Start of symptoms
On June 17, 2022 I traveled to the Bay Area to attend the rationalist summer solstice celebration and the Qualia Research Institute’s “Future of Consciousness” event. The first night of the trip I didn’t sleep well. However, I caught up on sleep on the 2nd night. On the 3rd night I had some anxiety about an interview in the morning and barely slept all night. The 5th day of the trip (Wednesday June 22th) is when I believe I was infected with COVID.
My symptoms started Friday June 24th while I was at the QRI event. Around 9 pm I noticed I was feeling feverish and unwell. I immediately suspected COVID, so I spent the rest of the event masked and outside. I left early and got back to my hotel around 11 pm. The next morning, I woke up with sinus congestion and feeling feverish.
I knew I wasn’t supposed to fly if I had COVID, but staying in the hotel would be very expensive and I knew I would not sleep as well as if I was at home. I decided to get a COVID test and then decide what to do.
Even this early in the story, I see some ways I stressed myself which may have contributed to Long COVID. The closest Walgreens was only a few blocks away, and to save money I decided to walk there. I checked out of my hotel and started walking there, dragging my roller bag behind me. On my way to the Walgreens, I nearly collapsed twice. Crippled with fatigue, I had to sit on the curb several times, choking down the exhaust of cars that were racing past. When I finally got to Walgreens they were out of tests. So, I then took an Uber to the airport, hoping to find a test there. However, there were no tests readily available. I went through security and waited for my flight. As I sat in the terminal I got an email alerting me that I had been exposed to COVID on Wednesday. On Wednesday I had done an in-person job interview and during the interview I had taken off my mask. That was the one and only time I took off my mask while indoors that entire trip. (Also, the rest of the trip I was only interacting with people outdoors. So I am like 90% sure this is how I got COVID). Unfortunately, the other person in the room tested positive for COVID two days later. I flew back, feeling bad that I might be exposing people, although I was very carefully wearing an N-95 mask. When I got home I collapsed into bed.
I also want to note that the day after I got COVID that I ran six miles. The day before I got COVID I ran abou four miles. I was in very good shape before I got COVID. Mentioning this I think draws into stark relief the difference in my health before and after COVID.
It was really just typical COVID.. at first
I was triple vaxxed. However, my last vaccine was nearly seven months prior, and I’m pretty sure I got the Omicron B4 variant since it was the dominant variant in the Bay Area. I had a 101 degree fever the first day home (Sunday). Luckily, I was able to get Paxlovid that evening, and the next day (Monday) the fever was gone. All that was left was a cough, congestion, and extreme fatigue. I tested positive for two weeks. During weeks three and four I just had a lot of fatigue, but not as bad as it would be in August. There were some things going on in my personal life that caused me a lot of stress and anxiety. So, a couple nights I only got a few hours of sleep. This may have laid some of the seeds for Long COVID.
In which I push myself, to disastrous effect
I had a second trip to the Bay area planned at the end of July where I was planning to attend a retreat for rationality and Astral Codex Ten meetup organizers. Then, I was planning to attend a lot of the satellite events around the Effective Altruism Global conference being held in San Francisco. Finally, I was planning to look at apartments in lieu of a possible move to Berkeley / Oakland.
I knew I likely was going to have trouble sleeping on the trip, so I brought some Trazodone with me. I also did a visit with my psychiatrist beforehand to see if I could get Ambien to bring with me. I had never taken Ambien before but I had heard of others using short term. Well, the telehealth psychiatrist I had been seeing could not prescribe any controlled substances). So I paid out of pocket to see a different telehealth psychiatrist. That psychiatrist wouldn’t prescribe Ambien, so I asked her for a benzo. Initially she said no, but then I mentioned that I had tried Ativan once and found it useful for sleep. She said she had problem prescribing Ativan (which is odd!). She also recommended Lexapro, so I picked that up to, although I wasn’t sure if I wanted to take it.
So, I went on the trip with Trazodone for sleep and Ativan and Lexapro for anxiety. I started the lexapro a few days prior to the trip at a dose of just 1 mg. In retrospect this was a bad decision. The first week or two of Lexapro, perversely, often involves increased anxiety as your brain adjusts to the serotonin. I had forgotten that.
Due to restlessness and anxiety I didn’t sleep well the night before my trip, so the first night on the trip I took about 10 mg of trazodone. To my surprise, it seemed to do nothing. I woke up after about 5 hours and felt very agitated, restless, and anxious. I had weird spasms in my legs. This pattern continued the entire trip.
For various reasons I really wanted to go to the grand opening party at the Berkeley Alembic. In retrospect, though, I should have flown home early to rest. Instead I stayed in my hotel through the entire week so I could attend the party that weekend.
A completely optional digression on Trazodone and Wellbutrin
This is a bit of a tangent so feel free to skip. Genetic testing shows I have multiple genes that make me a light sleeper. Since grad school I have had some trouble sleeping off and on. However, only once prior to my Long COVID experience did I have serious “free wheeling” insomnia, the kind where every night you aren’t sleeping enough, so your body never gets a chance to catch up on lost sleep. That experience was in June 2019, when I had insomnia for about a month straight. Now, at that time, 25 mg of trazodone cured my insomnia overnight. However, when I tried to use trazodone to help with Long COVID insomnia it did nothing. To this day, I still don’t understand why trazodone didn’t work for me at all in 2022 when it worked for me great in 2019. I became obsessed with trying to figure out why trazodone wasn’t working. I developed a theory that it was due to the small amount of Lexapro I was taking inhibiting one of the CYP enzymes, causing the nasty Trazodone metabolite mcPP to spike. I still think there might be something to that theory. However, my best theory now is I develop tolerance to trazodone fast, but I didn’t notice this in 2019 because the things that caused my insomnia went away shortly after I started trazodone. So, in 2019 I kept taking Trazodone, thinking it was helping me sleep amazing, but actually it wasn’t doing that much and I was sleeping amazing for other reasons.
Here is another pharmacological factor that may have been in play — Wellbutrin. Recall how I said I started Trazodone in 2019. Well, I complained about the morning sedation I was experiencing. So my psychiatrist prescribed 75 mg of Wellbutrin instant release in the morning to compensate. It turned out I really liked Wellbutrin — unlike SSRIs, which seemed to slow me down mentally and dull me emotionally, Wellbutrin seemed to improve my cognitive function. Because I liked it so much I kept taking Wellbutrin even after I stopped Trazodone. Eventually I tried the more conventional extended release formulation but after a week of not sleeping very well I decided to go back to the instant release. To spread the effects of the instant release tablet through the day I settled on a rather unconventional dosing regime in the spring of 2020 — I would split the 75 instant release tablets into quarters and take them when I woke up, at noon, at 3 pm, and at 6 pm. Fastfoward to July 2022 and I was still taking Wellbutrin, although I had cut down to just two quarters, in the morning and at 3pm. [Side note — Paxlovid, for reasons I can’t quite figure out, is said to reduce the effectiveness of Wellbutrin.]
My “tinnitus alarm system”
Three days into my Bay Area trip I had some stressful stuff come up in my personal life. I moved from the retreat venue to a hotel and that night I hardly slept at all. The next day was the worst fatigue I had experienced thus far. I the next day mostly laying in bed at my hotel, trying to nap and rest. While dozing in the bed I heard a police siren outside. At first I didn’t think much of it, until about an hour later I realized it was still going.
I moved my head around to try to locate the source of the sound and it didn’t change. That is when I realized it was in my head and not outside. I verified this by opening the window to see if the siren sound was something outside. There was no change. For the rest of the trip, the siren sound would be constantly ebing and flowing. I sort of blew off it off as an earworm or minor auditory processing issue caused by stress, and I didn’t think much of it.
When I got back, the sired noise became almost constant, so that’s when I started researching tinnitus. The condition really didn’t bother me much — because it was so quiet it wasn’t distracting me that much. I calmed down my nervous system with a bit of breath meditation to see if I could modulate it through relaxation and consciously directed attention. However, the pattern of oscillation seemed to have a life of its own, resisting my attempts. After a couple of weeks it only seemed to arise when I was stressed, so I called it my “tinnitus alarm system”.
Insomnia sky
The insomnia I had in August was the worst of my life. I had no trouble falling asleep but would wake up after 4 or 5 hours and would be unable to return to sleep. The first thing I tried was increasing my trazodone dose. I went up to 50 mg, then 75, then 100, then 125, and finally 150. None of those doses worked, and each left me more and more sedated the next day. I called it quits on trazodone and then took doxylamine for a bit. That did nothing (normally it puts me out for eight hours). So then I moved to Ambien. Ambien actually helped me a bit if I took it a bit sublingually when I woke up in the middle of the night, but it made my head fuzzy and I never felt well rested after taking it.
I did learn a lot through this. First, I learned that I could still function in some (very limited) capacity even with only 3–4–5 nights of sleep for two weeks. I realized that a lot of my fretting about sleep loss in the past had been overblown. This was real insomnia. Next, I learned that bad insomnia creates a vicious cycle — you don’t do much during the day, so you don’t build up a sleep drive (I think adenosine is the main thing here, but there may be other factors). So, the next night you only sleep a few hours. You’re also prone to napping, which saps your sleep drive and deprives you of restorative sleep the next night.
Anyone else notice that clouds look really beautiful when you’re extremely sleep deprived? I think this is due to slight hallucination. I call it “insomnia sky”. There was one time during this period where I went on a long walk while severely sleep deprived. I remember they clouds in the sky looking exceedingly beautiful.
Obsessing over Remeron
I am very detail oriented. This makes me really well suited to being a scientist and machine learning engineer. This attribute does have some major downfalls, however. For instance, psychiatric medications generally have something mildly terrifying details when you start delving really deep into what they do in the body. There are some exceptions but in general I think it’s hard to be a perfectly optimally functioning human when taking psychiatric medications. There are always side effects.
The fact there will be side effects is hard for my obsessive, perfectionist mind to accept. My brain wants to find the perfect drug and dosing scheme to obtain the fastest recovery without any side effects — which is essentially an impossibility.
The situation I was in with insomnia, combined with my obsessive, perfection-seeking tendencies, led to a severe neurosis. The key question I was faced with was whether to take Remeron or not. The neurosis was caused by the fact that either choice looked bad — a sort of Cache-22.
On the one hand, I wasn’t sleeping. I knew Remeron would help me sleep. On the other hand from prior experience I knew that Remeron would leave me somewhat sedated during the day and might be hard to get off. I also had a bad experience tapering Remeron in 2019. The general starting dose for Remeron is 15 mg. As you move lower, it actually becomes more sedating, which makes tapering from 15 annoying. I really didn’t want to go on Remeron. I decided I’d take it for just a day or two to hopefully normalize my sleep cycle and then taper down over the course of a week. The first day I took Remeron, I was extremely fatigued the next day.
I haven’t been having to take it to sleep and recently tapered down a bit, and again am not sleeping well (only got 5 hours last night). So I’m going back to the dose I found best for sleep, 10 mg.. and might even go up to the recommended 15 mg.
Whenever I did something significant, I would crash for at least a day. Lower sleep quality seemed to be correlated with the crashes. I remember one day in early September I decided to push myself a bit and go shopping around 7 pm. Upon arrival, I felt intense fatigue and immediately regretted it, but I was already there and decided to push through. That night, my nervous system never fully calmed down, I slept poorer, and the next day I was crashed.
The darkest days
During those grim days the most exercise I could do was walking around the block, which would leave me utterly exhausted. If I tried to run or do anything strenuous my muscles immediately gave out and became achy. I was couch bound and so mentally fatigued I could not read. Even listening to audiobooks at 1x speed became difficult at times. This went on for several weeks. I was all alone in my apartment except for my cat. However, somehow I was not feeling lonely despite feeling lonely in the past.
I became terrified I was getting ME/CFS. I became terrified of overexerting and worsening my condition. The process of climbing up the three flights of stairs to get to my apartment took on a terrifying quality. I began tracking my heartrate but soon found that trying to make sense of the wildly fluctuating readings was provoking enormous amounts of anxiety.
For reasons I will explain below, I believe it’s quite possible that if I had continued to rest my condition would have deteriorated.
Graded exercise has dangers, but rest and pacing does as well.
Who will you listen to?
In High School I devoured everything that the physicist Richard Feynman wrote. One of his most famous quotes is “Science is a way of trying not to fool yourself… you must not fool yourself, and you are the easiest person to fool.”
It’s fashionable today to talk about “lived experience”, but in my opinion “lived experienced” is a bit over-hyped. It’s just too easy for the mind to construct overly simplistic or utterly confused narrative to explain experience. Also, work in psychology shows a mismatch between the “remembering self” and the “experiencing self”. So, your story about what you experienced is actually usually going to be pretty off unless you have been documenting things carefully. While “lived experience” may have some limited utility, to understand what is actually going on ultimately you always need rigorous scientific inquiry.
Furthermore, a wide variety of research shows that the depressed brain selectively recalls the negative aspects of prior experiences. Inflammation may trigger depression which triggers selective memory recall.
All of this is why I don’t trust people’s anecdotes very much, especially when they come from people with high levels of depression (or inflammation which is likely to cause depression).
Now, at the time I wasn’t really thinking about this. However, I did know from my prior experiences with major depression and CBT therapy that I couldn’t trust my own thoughts and the narratives about my illness that I was creating. For instance, I was started to develop a narrative that exercising was making my condition worse. Fortunately I kept questioning this narrative. Yet it was very persistent. To this day I still have some anxiety about overexertion.
Discovering Paul Garner — the turning point
On October 5th my initial four weeks of leave were coming to an end, and I realized I have very little idea if I’m actually getting better. This fed into a strong anxiety that I had developed permanent ME/CFS.
Fortunately it was around this time when I believe things started to turn around (very slowly). In my mind, the turning point occured while I was listening to the “Positively COVID” podcast while outside on a walk, and I heard Paul Garner talk about his recovery story. His story really resonated with me. I decided to commence graded exercise as a result.
Here’s how I described what followed:
I started to re-conceptualize the illness I was experiencing as largely psychosomatic in nature. I stopped reading about Long Covid, and tried to reduce obsessive tracking and thinking about symptoms. I got into a Long Covid clinic, which sent me materials from a CFS/ME organization (The Bateman-Horne center, not a mainstream evidenced-based medical org). I I promptly ignored those materials, which advocated being very careful about acitivty and doing careful pacing.
Over the next two months, through many fits and starts, I slowly built my strength up to the point where I could walk about one mile each day consistently. By December I could work maybe two hours a day. The recovery was extremely slow and not linear — there were many days, perhaps 2–3 days a week, where I ended up completely crashed and nonfunctional. During these days I was completely couch-bound and often didn’t leave my apartement. I noticed these “crash” days seemed to be correlated with not sleeping well. Whether those “crash” days were due to post-exertional malaise, I am not sure, but I tend to think they were more linked to my sleep quality and a classic insomnia cycle (good sleep being followed by poor sleep etc, with good sleep causing higher activity level). Eventually, my system started to stabilize in January. I returned to something similar to my pre-covid exercise routine, running and weightlifting a few times a week, although at a greatly reduced level. I went off Remeron. Then, unfortunately in mid-January I developed bad insomnia again which lasted about 5–6 weeks. My condition regressed greatly. Once I got my sleep straightened out, I started to return to normal. Today (early April) I consider myself mostly recovered. I walk about a mile or so every day and run 1–2 times a week. I can only run 1–2 miles, as opposed to the 4–5 mile runs I was doing pre-covid, but I am slowly building back to my pre-covid strength. I am taking medication for sleep still but slowly tapering off. A few days a week I end up very fatigued and crashed but overall I am doing much better
Insomnia , sleep study, and discovery that I had PLMD
Originally my sleep study was supposed to be scheduled for April, but unfortunately I got COVID a second time! So it had to be moved until June. After the study I did not get any results. Fortunately I spoke to a doctor who is a friend of family and he told me ask for the report. I had no idea there was a report. The report indicated I had severe form of periodic limb movement disorder (PLMD).
Today (May 2024)
Today I consider myself mostly recovered, perhaps 90%. I still have fatigue issues quite often ( a few times a week), but I also had fatigue issues before Long COVID to some extent, and my issues seem more tied to poor sleep + occasional use of sleep aids than anything else.
(may write more here soon)
Some random concluding thoughts
Fatigue is the most common symptom experienced during Covid-19 infection. After infection, energy is re-routed to the immune system. The manufacture of antibodies consumes a lot of energy, and during acute infection the immune system can consume large fractions of the total energy used by the body.
Across all human cultures, humans that are sick exhibit similar behavioural patterns. These include sleeping more, isolating from others, increased attention towards pain, and not eating as much. This general cluster is known as “sickness behaviour”. How much of these behaviours are genetically programmed vs learned is not fully understood. The “Eyam hypothesis” postulates that social isolation and not eating as much are genetically-programmed, and their purpose is to lower the chance that the infected organism spreads their disease to others. The evidence for this theory mainly comes from observing other social mammals. Mice, for instance, quarantine themselves when sick.
During the acute “sickness behaviour” phase, there’s no doubt over-exertion can make symptoms worse. It seems the brain will push back against exertion and fight to make sure you feel fatigued and conserve energy for your immune system.
A growing body of anecdotal evidence suggests that Long COVID (which is typically defined as one or more symptoms lasting more than 3 months) is often triggered by people who return to strenuous activity too quickly. This could be due to the neural circuits of the PEM response becoming ingrained (the “false fatigue alarm” theory).
Furthermore, growing evidence suggests bits of COVID-19 virus can linger in the body for weeks. Many Long COVID patients exhibit the reactivation of Epstein-Barr virus. In animal models, Epstein-Barr reactivation is more likely under stressful conditions. This is another that exertion may trigger Long COVID — if you stress yourself out too much, Epstein-Barr reactivation can occur. About 90% of adults in the US have latent Epstein-Barr virus.
In my own case, I believe I had very likely had a mix of “false fatigue alarms” and Epstein-Barr reactivation — this seems to align the most closely with my experience.
