2023-05-04 General

I had an ugly thought. Maybe it makes a certain sense from the majority’s standpoint, to remove all restrictions because it would level the playing field. If COVID-19 does cause lasting damage, e.g. in increasing the risk of death or disability, then the people who have never had COVID-19 have an advantage over the people who have had COVID-19. People don’t like someone else having an advantage over them, so perhaps the reasonable thing to do is make sure that everybody has the same disadvantage. 🙁

It doesn’t even need to be conscious, it could just manifest as indifference. “I had COVID, and I’m still alive, so why should I worry about the NOVIDs catching it?”

Mitigation Measures

This article reports that starting May 11, most international travellers to the USA will not have to show proof of vaccination. (This article from 25 April talks about how one Canadian family got messed up at the border because it didn’t occur to them that the US would still need proof of vaccination.)


This press release from Hong Kong says that 70% of survey respondents (who all had had COVID-19 infections) still had some symptoms after five months. 12% of the Long COVID cohort reported reproductive health issues.

This paper from USA used fMRI to watch Long COVID patients doing various mental tests. Interestingly, while they found similar performance on various metrics, their brain scans showed that they were using different areas of their brain to compensate. Maybe the Long Haulers can do as well on short tasks, but the tasks take more effort. Long Haulers did have slightly lower dexterity and significantly lower endurance.

This report from UK (from February) found that the risk of Long COVID was 28% lower (but only 28% lower!) after a second infection than after the first.

COVID-Related Excess Death and Sickness

This preprint from Scandinavia found that people who were bedridden (note, not necessarily hospitalized!) for at least seven days with COVID-19 had a 125% higher prevalence of severe physical symptom burden (scores higher than 15 on PHQ-15, the Patient Health Questionnaire) over the course of 27 months. People who had COVID-19 but were bedridden for less than seven days didn’t have any more severe physical symptom burden than controls.

This paper from Austria from Feb 2021 and this paper from the USA from May 2022 both found that patients who had had COVID-19 infections were significantly likely to have REM sleep without atonia (the muscles going completely relaxed). This international paper from April 2022 found that people who had had COVID-19 infections were about three times more likely to have deam-enactment behaviours (which means making physical motions during the dream, which necessarily means they also didn’t have atonia).

Meanwhile, this paper says that 80% of people who have REM without atonia go on to develop Parkinson’s. Ulp.

This preprint from UK from October 2022 found that people who had had COVID-19 infections were 22% more likely to get immune mediated inflammatory diseases than the non-infected. In particular, people who’d had COVID-19 were:

  • 56% more likely to get type 1 diabetes;
  • 52% more likely to get inflammatory bowel disease;
  • 23% more likely to get psoriasis.

This paper from Croatia found that arterial walls get stiffer after a COVID-19 infection, even a mild case. Worse, they found that over the 3-month length of the study, arterial walls continued to get stiffer instead of getting better. They hypothesize that the stiffening is because of chronic inflammation.


This paper from Japan found that the presence of certain types of microbes in the gut made vaccines less effective.

This preprint from the USA reports that they figured out an algorithm which can make mRNA vaccines significantly better among many axes. For COVID-19, they reported a 128x increase in antibodies in mice and 5x increase in stability (which affects, like, how cold your fridge needs to be). Note that this technique should work for all mRNA vaccines. This seems like a BFD, so I will try to explain as best I understand it, but (I am not a microbiologist!) I might get wrong in some ways.

Okay. Sets of three “letters” in RNA or DNA are called codons. Each codon causes production of one of 20 amino acids. However, with four possible values of each letter, each codon can have 81 possible values. Some codons do not exist in nature, but still, you can choose multiple different codon patterns to generate any given amino acid.

Different choices of codon make the mRNA more or less stable because um physics. (Remember, we’re talking about the mRNA here, not the thing the mRNA makes, e.g. the spike protein. The whole goal here is to make the exact same spike protein.) But there are so many possibilities, that you can’t try different “spellings” by brute force. (The paper says there are ~2.4×10^632 different mRNA sequences that will produce the spike protein.) They say they have figured out a clever algorithm which gives more stable mRNA, without having to brute-force it.

The algorithm they use is a variant of one used in speech recognition called a lattice parser: they create a directed acyclic graph with weights on the edges and explore the graph with beam search. Apparently this gives really good results in a short amount of time.

This correspondence from USA says that they found that there are more BA5-specific antibodies after a bivalent booster (if I understand correctly), but they don’t actually neutralize BA5 any better than the non-bivalent shots (that part was clear).

I think what happens is that BA5-specific antibodies are glomming onto the virus, but not in ways that gum up the virus’ machinery and prevent it from infecting.

I am not a virologist or a immunologist, but I have the intuition that it takes (at least) two tries to train the immune system fully. The first time the immune system sees a pathogen, it says “huh, that’s interesting” and the second time it says “oh f, we need to get serious about this!” I had expected that having the first COVID-19 classic shots would mean that the immune system would take BA.5 seriously, but maybe not? Maybe you need to doses of the bivalent?

Before I saw this correspondence, I advised my spouse recently to get a second dose of Novavax instead of an mRNA vax because I wanted him to get the second-dose effect, and figured that his previous mRNAs would be enough for his immune system to take BA5 seriously. Maybe not? Now I don’t know what to recommend.

This big study from Canada found that people who were unvaccinated were 4.3x as likely to be admitted to the ICU and 3.9x as likely to die as people who were “fully vaccinated” (which usually means “two doses”). They were 12.2x as likely to go into the ICU and 15.1x as likely to die as people who had gotten a booster.

In this press release, AstraZeneca announced that it has what it says is a really good long-lasting antibody shot, AZD3152. While technically not a vaccine, it — like Evusheld before it — is given like a vaccine to reduce vulnerability to COVID-19 for some period of time.

This article says there is accumulating anecdate about post-vaccination tinnitus. (However, people do get tinnitus, and correlation does not prove causation.)

Early on, there was some indication that the tuberculosis vaccine (BCG) might give some protection against COVID-19, because it seemed that countries that regularly vaccinated against tuberculosis had lower COVID-19 rates. This study from Denmark found that vaccination with their BCG vaccine didn’t reduce COVID-19 susceptibility at all. HOWEVER, there are a gazillion different BCG vaccines from different manufacturers using different strains… so don’t write off BCG completely just yet.


This preprint from India says that XBB.16.1 is hitting infants harder than older children, and that they are seeing itchy, non-purulent conjunctivitis with mucous discharge / stickiness of eyelids in ~40% of their patients.

This paper from UK (from April 2022) found that people were 61% less likely to die after a reinfection than after a primary infection. My initial thought was “well, duh, the vulnerable people died before they got a second infection”. However, people are also less likely (76% less likely!) to go to ICU after a reinfection than after a first infection, so maybe my initial reaction is unfounded.

This big study from Canada found that while the proportion of hospital admissions with COVID-19 went from 24.7% of admits in pre-Omicron waves to 77.3% of admits during Omicron waves, the death rate dropped from 16.0% during the first two waves down to 7.0% in the last two (Omicron) waves.


This study from USA found that Molnupiravir reduced the risk of Long COVID by about 14% (vs. 26% for Paxlovid).


This paper from California reports that after two months of training, dogs were able to sniff out COVID-19 in schoolchildren with sensitivity of 83% and specificity of 90% (against a standard of rapid antigen tests). Given how RATs have pretty crappy sensitivity (72% in people with symptoms and 58% in people without symptoms), maybe the dogs are even better? (I see some responses to this study making the very valid point that this exposes dogs — who can catch COVID-19. Is it ethical to expose the dogs? Might the dogs infect the kids they are testing?)


This paper from Germany from October 2022 compared students in sports-focused schools (which apparently is a thing in Germany!) and non-sports schools during Delta and Omicron waves. The students all normally had to mask with high-quality masks; the only difference between the cohorts was that the sports-focused schools’ kids did not have to wear masks during their six hours of PE per week. During periods of lower transmission, the kids did not have to mask.

The difference was striking, but the papers’ authors noted that it delayed the kids catching COVID-19 more than prevented it. (Well yeah, the masks wouldn’t keep the kids from catching it somewhere else, of course.)

Sports schools in green, non-sports in blue

This paper from Norway (from March 2021) found that increased pollen levels correlate with increased COVID-19 transmission! (Another reason to wear a mask!)

This paper from the UK found that COVID-19 antibodies were found in the blood of about 3% of housecats.

This big study from Canada found that, for hospitalized COVID-19 patients where they were pretty sure where they caught it, 16.9% caught it in the hospital during waves 5 and 6 combined (Omicron), versus 10.8% in waves 2 (Classic), 3 (Alpha/Beta/Gamma), and 4 (Delta) combined. Maybe I’m just silly, but I think even 11% is way too high!

I had been guessing about what fraction of people in BC are infectious. This article quotes a guy at the Vancouver Infectious Diseases Center saying it’s ~2%. (Now, VIDC focuses on HIV and Hepatitis C, so maybe he isn’t the absolute best source, but he’s at least a source.)


This preprint from India looked at variants between between 1 December 2022 and 8 April 2023. They found that ~36% were XBB.1.16 family, ~12% were in the XBB.2.3 family, and ~10% were in the XBB.1.5 family. Among the XBB.1.16 patients, ~92% percent had at least one dose of vax.

Recommended Reading

This blog post goes into depth on recent research which has found links between COVID-19 infections and auto-immune disorders.

Ha! I have been saying that COVID-19 is not seasonal (despite what Dr. H keeps saying hopefully). This article agrees with me. Partly, they say, SARS-CoV-2 is mutating twice as fast as flu and about ten times as fast as “common cold” coronaviruses, so can’t really settle down yet into a predictable pattern. (I believe this is because there are just soooo many more people with COVID-19; the virus has more opportunities to mutate.)

This article dissects and rebuts a hit piece on Anthony Fauci, and in doing so, gives a reminder of what it was like at the start.

This blog post goes into extreme detail about the mucosal vaccines which are here and those which are coming. This blog post talks about why mucosal vaccines are so hard to make.

This article gives an overview of what is currently known about how dangerous reinfections are or are not. (Spoiler: they aren’t sure.)

This article talks about all the treatments which are available in the US. Most (but not all) are available in Canada. (I think all the drugs but not all the monoclonal antibodies are available in Canada.)