week ending 2022-08-25 General


This article reports on this (difficult-for-lay-people-to-understand) research report which says that Long COVID patients had an imbalance in the clot-promoting and clot-removing proteins in the blood. (Note: in the Recommended Reading section, there’s a link to a long article on clotting and Long COVID.)


This paper says that physical exercise makes your COVID-19 less severe, however I didn’t see anything in the study that controlled for the possibility that healthier people are able to exercise more. Wealthier people have more leisure time to exercise and to visit doctors; people with chronic conditions have less energy to exercise, etc.

This paper found that people with worse COVID-19 outcomes had worse fatty profiles in their blood. Again, a caution about the direction of the causality: they did not have before-and-after fatty profiles, so it might be that people who had worse fatty profiles had worse COVID-19 outcomes. However, other research found that people who have had COVID-19 have elevated risk of cardiovascular events in the following year. It also said that they saw these bad fat profiles in people who had long diseases which were not COVID-19.

This paper found that type 2 diabetes increased 77% in the first pandemic year compared to the two previous years.


This paper says that the incubation period keeps dropping for variants over time:

  • Alpha: 5.00 days
  • Beta: 4.50 days
  • Delta: 4.41 days
  • Omicron: 3.42 days


This paper from Germany says that they found a ton of impurities in four different lots of AZ, implying really crappy quality control at AstraZeneca Corporation. (Interestingly, they found no QC problems with J&J, which had such awful problems at their US plant that they had to shut it down.) Now, the paper was from July and I’m only just hearing about it now, and I would have thought it would be a big scandal if it were true, so I don’t know if the paper was flawed. Still, there’s no reason to get AZ, we have better vaccines anyway.

This paper says that the 6-proline substitution (used in NDV-HXP-S in Vietnam, Patria in Mexico, and Buntanvac in Brazil) gives much better results in vaccines when given to mice and hamsters than the 2-proline substitution (used in Pfizer and Moderna). (The proline substitutions help keep the spike protein from going all floppy when it’s not attached to the globe of the virus particle. See this article about the 2-proline substitution and this article about the six-proline substitution.)

This preprint from the USA discusses a fishing boat that had an outbreak. The scientists had before and after blood samples, and found that none of the people who had high neutralizing antibodies got sick, despite 85% of the crew getting COVID-19. This gives evidence for the presence of neutralizing antibodies being a good “correlate of protection”, meaning it’s more evidence for being able to just look at antibodies in test tubes to figure out how well a vaccine is going to work. Looking at blood in test tubes from a dozen people is a whole lot cheaper and faster than doing a human trial with thousands of people.

The UK is going to use bivalent COVID Classic + BA.1 vaccines as their fall boosters; the USA is going to use bivalent COVID Classic + BA.5 vaccines. It looks like supplies will be available. While there is already pretty good data for the bivalent BA.1 shots, the data for BA.5 is a little sketchy. The best they have is antibody levels of rodent blood in test tubes. (See this blog post for more about the BA.5 booster and what data there is for it.)

Mind you, I don’t think the BA.5 is going to be dangerous, given how many billions of people have gotten the COVID Classic version. It is consistent with the process/data the USA uses approves flu vaccines. I am just a little surprised that the US — which tends to be very very very conservative about vaccines — was the first to go for the BA.5 vaccines. (Repeat: it’ll be fine. I think Novavax is a better choice of booster, but if I hadn’t recently gotten a Novavax booster, I’d get a BA.1 or BA.5 booster as soon as I could. And I fully expect I will get a BA.1 or BA.5 booster in December, when it’s been six months since my Novavax.)

Health Canada is still dithering about which vax to approve. Apparently they want a bivalent with some Omicron component, but they still have to decide between BA.1 and BA.5. (Me, personally, I’d ask for BA.2, but they didn’t ask me.)

This article says that Bharat Biotech (a big Indian pharma company) has finished Phase 3 trials of BBV154, its nasal booster. I am very excited by this, because nasal boosters look like they will train immunogoblulin alpha — a piece of the immune system affiliated with mucous membranes better, and thus head off infections better.

Alas, the article didn’t say what the results of the Phase 3 trials are, which is worrying. (I figure that if there were awesome results, they would have said so.) It might be because they only looked at antibodies in test tubes (not the number of people who got sick), and that might not have shown a lot.


This paper from the USA says that Paxlovid drops various serious outcomes in high-risk COVID-19 patients in about half.

This paper from Israel had a bit more nuance on Paxlovid:

  • For people over 65, Paxlovid dropped the risk of serious outcomes by 80%.
  • For people under 65, Paxlovid really didn’t help.

The difference in outcomes by age is really, really stark:

Excess Deaths

This paper from the USA looked at total mortality and subtracted out mortality from COVID-19. While elderly people were more likely to die of all causes in 2021 than pre-pandemic, they were less likely to die from non-COVID causes than before. Little kids were also less likely to die during the pandemic. Everyone else was more likely to die from COVID-19 and more likely to die from non-COVID things.

An odds ratio above 1 indicates greater death risk from April 2020–March 2021 than in 2015–2019

One interesting thing that the article pointed out is that the risk of death might be lower in the future, basically because the people who died early in the pandemic will not be dying in the future.

Recommended Reading

This long thread from the UK estimates that the excess deaths from long wait times in emergency rooms are in the vicinity of 500 per day. (For comparison, that works out to 7.5 per million per day, which would be 37 deaths per day in BC; BC is currently seeing ~10 deaths per day from COVID. ) Some of the long UK waits are due to inadequate space in post-care institutions — 13% of beds are held by people who don’t need to be in hospital but don’t have anywhere else to go. However, some is COVID-19 related: 6% of the staff is out sick with COVID. (The thread doesn’t talk at all about how many HCW have died, become disabled, or quit.)

Meanwhile, ER wait times are high in BC, as you can see from the VCH ER wait times web page. Right now, for example, the wait at Children’s is ~6.5 hours.

This article covers the possibility of micro-clots causing Long COVID in great detail, including pros and cons to the argument.

This article shows every single step taken in monitoring wastewater for viruses.