Oh, and ACX links to https://henryaj.substack.com/p/lifehacks who claims: "in most places you're just allowed to buy and import prescription medication and take it" linking to
UK website looks to still be good. [Edit, but both only ship to the UK, dang. What's the international version, used to be dodgy Indian websites...]
I have been unable to find any source for the powder for suspension you'd need to properly make up the smaller doses for children, if anyone else finds it that'd be helpful.
Are your figures in addition to, or instead of, mortality from a normal flu season? It seems a solid chunk of your found expected h5n1 deaths would be expected to die from normal flu anyway, wouldn't h5n1 replace the normal flu? At the very least, anti pandemic measure would limit normal flu spread similar to anti covid.
Presumably, the anti-pandemic measures would be calibrated to the expected level of danger. An h5n1 pandemic that's closer to the 2009 pig flu outbreak would probably get a pretty tepid response, while something like the 1919 pandemic or worse would get the full Covid treatment.
If an outbreak happens that is seen as bad enough to justify lockdowns, I would expect them to be extremely effective to the extent they're complied with. The Covid response was initially modelled on planned responses for a particularly severe pandemic flu, and they work better on influenza than on covid due to stuff like droplets vs aerosols. "To the extent they're complied with" is the hard part, since everyone is still tired from last time.
Honestly, are governments going to even bother this time, considering the response to the previous lockdowns? People are just going to do whatever the hell they want. And ultimately, a few dozen million deaths isn't a big deal when it's spread out across the globe.
A lot of the effective changes around Covid, at least in the UK, were voluntary and often started among the population before the government changed the advice or issued legal restrictions. The population (and the police…) often went beyond what the government had actually required, because they were worried and trying to protect themselves and others.
Of course, we also know that the UK government and scientific advisors were surprised by the high compliance of the population, making the lockdowns much more effective than expected: they had issued stricter advice than they thought was needed, and emphasised the dire consequences of not following it, because they assumed people would skirt around the rules much more than they did.
The things that needed national government support/central decisions that made a big difference to the spread (rather than the treatment) were the closing of schools (made a huge difference to spread in the UK each time) and the easy, free availability of relatively-reliable testing.
Possibly also the chunks of money paid out for businesses furloughing people, paying them to not work, but I can’t see that happening again.
The closing of schools was disasterous in many ways (rendered more disasterous by the choice to close playgrounds and keep them closed, and the slow introduction of things like ‘bubbles’), but also attendance in schools dropped noticeably before each closure, and every time Covid became more prevalent, and not just because students were actually ill. I also had students I did not see for a long time, because they or family members were immune-compromised. Probably a future pandemic needs something more flexible in place around schooling.
The UK testing system was fantastically expensive (for the taxpayer), but popular and widely-used. It gave people a way to manage risk when things were very uncertain, helped monitor what was going on in terms of spread and mutations, and could justify people taking the necessary time off from work in a legally and socially powerful way while they were contagious. I can see a similar system being used again in future pandemics.
...Maybe I am being too cynical. Still, the political situation in the US is a whole different can of worms. The administration has its own supporters to appeal to, and they absolutely do not want lockdowns again. So that's that.
I agree that the US is in a funny place around this: everything around Covid got fantastically split on political lines, didn’t it? I don’t pretend to know how Americans, or the US government, would react to a serious influenza pandemic.
But I think it is still worth bearing in mind that when people think they are in danger, they tend to move to protect themselves with or without government support. Lockdowns were dreadful and I hope never to see them again, but you probably need some sort of government response if the population starts withdrawing from and getting scared of each other.
I suspect if people started to see non-elderly non-otherwise-sick people dying around them, ESPECIALLY children, they'd comply more readily. I might be biased because I don't know anyone who died of covid but I know two people (both young adults and otherwise healthy) who died of flu related sepsis, but still.
I've never known anyone who died of Covid or the flu. But I did know two formerly healthy people, one a relative, who have died in recent years of antibiotic-resistant bacterial infections. And that could well be the wave of the future.
I'd love to believe this, and I probably even used to. Sandy Hook changed that. Half of (at least the American) population doesn't give two shits about dead children.
I don't mean to undermine the emotions here, but do you literally believe that? Like do you think that if you picked a random American and told them that a child had died, there would be a greater than 50% chance they would go, "I don't care about that at all?"
I think you actually mean something more like "Half of Americans don't support measures that would save children at a cost which (I believe) is negligible." And that might be true! But I doubt your model of the world actually matches what you said.
The "half" may be an exaggeration, but it doesn't change my intent. If you're looking for a more "accurate" sentence, I'd say something like "A significant percentage of the American population is okay with child deaths so long as they aren't inconvenienced in any way that impacts them directly." (Masking, vaccines, sensible gun laws, seat belts, speed limits in school zones, etc etc etc.)
A significant portion of America's population is so selfish, they've proven that they truly don't care ENOUGH about child deaths to be inconvenienced.
Summarized and exaggerated: "Half of (at least the American) population doesn't give two shits about dead children."
Ah, ok - that's clarifying, thank you. I personally think most of that percentage thinks the "inconvenience" is too great for the safety increase. Which, I mean, we all take actions that risk our lives and those of others (e.g. driving a car). Your examples reflect that as well - you say "speed limits in school zones" not "speed limits everywhere" presumably because you recognize that there would be a greater cost to a universal 35 mph speed limit, even though it would also save lives.
All of this is just to say: I think the people you are talking about mostly disagree with you about the efficacy or cost of the measures you have in mind, not whether children's lives are valuable. I find that to be a helpful perspective, because it suggests they could be convinced, or that some other measure might be something everyone could agree on.
Is there actually a droplet vs aerosol distinction for covid vs flu? My understanding is that Lindsey Marr was talking about aerosols for flu as early as 2015, and it took Covid for the mainstream health establishment to realize this can be relevant for this sort of virus. I don’t know how much they’ve changed their opinions on flu.
My info for might be outdated.on that particular point. Although flu spreading at all readily by aerosols seems unlikely to me, based on the differences in R0: the numbers I've seen for flu range from 1.2 (normal seasonal flu) to 3.0 (high estimate for the 1918-1920 pandemic), while I've seen newer covid variants estimated at 5.7. Even discounting for the 1.2 figure for seasonal.flu probably not being a true R0 due to cross-immunity with prior years' strains, that's a pretty big gap that gives me a significant prior for there being substantial differences in mode of transmission.
I have seen some much lower estimates for covid's R0, in the pandemic flu range, but they're either analysing the original outbreaks and thus had extremely wide confidence intervals for want of good data, or they're using data from 2020-2021 and thus measuring covid's R0 under lockdown conditions, not its R0 under normal conditions.
It could well be that the lower R0 is just related to different intrinsic ability to penetrate cells, or easier immune response, even if the aerosol aspect is the same. The issue Linsey Marr focused on is that for decades, aerosol transmission was treated as a weird abnormality a few infections had because of a few early 20th century misprints, rather than because of any evidence that it wasn’t common. This is the 2015 article that suggested to me that there is more to this debate than just covid:
Can someone explain for me how/why strain replacement happens? Why doesn't each new strain that crosses over simply increase the overall rate of flu? Do the strains crowd each other out somehow?
The existing strains typically fade out because of population immunity dynamics, which create a hostile environment for them to continue spreading. Here's why:
After a strain has circulated widely, most of the population develops specific antibodies against it. This means:
1. The old strain has fewer susceptible hosts to infect
2. When it does infect someone, their existing partial immunity often leads to milder illness and reduced transmission
3. People are more likely to have some cross-protection against that strain in future seasons
Meanwhile, the new replacement strain has advantages:
- It can infect people who are immune to the old strain
- It spreads more efficiently through the population
- It may cause more severe symptoms due to lack of pre-existing immunity
This creates a competitive situation where:
- The old strain struggles to find new hosts
- The new strain rapidly spreads through the susceptible population
- Resources (susceptible hosts) become increasingly limited for the old strain
Think of it like market competition - once a new, more successful competitor arrives, the old business (strain) has trouble maintaining its customer base (susceptible hosts) and eventually can't sustain itself.
However, it's worth noting that old strains don't always completely disappear - sometimes they can persist at low levels or in specific geographical regions, potentially re-emerging if conditions become favorable again.
> Think of it like market competition - once a new, more successful competitor arrives, the old business (strain) has trouble maintaining its customer base (susceptible hosts) and eventually can't sustain itself.
This feels a little off. How does the new strain interfere with the old strain? The old strain naturally declines by two methods; the hosts become more resistant to it (real decline) and more resistant to its symptoms (apparent, but false, decline). Neither of those relies on the existence of a newer strain.
(All humans carry a significant number of completely asymptomatic diseases, which is why islanders get sick when they're visited by ships.)
You’re right, it doesn’t come out and say why there are fewer susceptible hosts for the old strain.
I assume for two reasons: Mortality and NPIs (spontaneous and mandated).
And one more: Non-specific antiviral responses such that infections from the new strain induce a temporary shield against the older strain thereby further isolating it.
This herd immunity (and/or mortality) and NPIs have been the accepted explanations for strain replacement. With COVID these explanations don't really fly. Of course, it's hard to compare NPI implementations from country to country, but with a highly contagious pathogen like COVID, even the strictest NPIs couldn't stop its spread once newer, more transmissible variants evolved (see China's attempt at ZeroCOVID).
We've seen...
1. A new variant can push out all the previous variants and become the dominant variant but not create a new wave — i.e., it takes up a bigger piece of a smaller pie. Examples of this are Alpha in the US and Kraken worldwide. In fact, in the US, Kraken may have prematurely ended the BQ.1 wave because (counterintuitively) US COVID cases dropped precipitously as Kraken became the dominant variant.
2. In two countries with a similar variant mix, a new variant can create a wave of new cases in one country but not another. Examples: Lamda and Gamma in South America (although they were roughly contiguous in time, so they may have possibly been burning each other's fuel). XEC seems to be starting a new wave in the US, but it didn't create a wave in Australia.
3. And then we have examples like Omicron that pushed every other variant aside and created new waves everywhere.
Other weirdness that's worth noting...
4. Even though regions may have similar variant mixes, some regions seem to experience two COVID waves each year (US and Australia), and some have three or four waves each year (France, Germany).
5. Possible examples of viral interference: There seems to be some sort of inverse relationship between SARS2 waves and Rhinovirus waves. RhVs peak when COVID is low. COVID peaks when RhVs are low. Influenza cases in the US (and worldwide) dropped precipitously to virtually zero just as COVID cases started ramping up. NPIs have been put forward as the post hoc explanation for the phenomenon, but this dropoff happened before NPIs were mandated. And except for a long, slow burn of Type B cases in China during their ZeroCOVID lockdown, flu stopped circulating until the advent of Omicron. Then it came back with a vengeance.
Makes the carrier confine to home for the duration of illness and avoid seeing strangers? Drives up the fever making the body more hostile to pathogens in general?
"more resistant to its symptoms(apparent, but false, decline)"
some symptoms are among the best ways for diseases to spread, or if (e.g.) a host is less likely to sneeze or cough while having an illness, they're also less likely to spread it, even conditioning upon the same viral load.
"Neither of those relies on the existence of a newer strain."
If two strains of an illness are very similar, there is substantial cross-immunity.
Wild ass guess: most flu strains give cross immunity. The new strain has to have the unusual ability to infect people with immunity to the old strain, or it wouldn't be epidemic. The old strain has no such filtering.
I think this more or less has to be it. It makes sense of the replacement process of Covid strains - the faster spreading strains are providing those who catch them with sufficient immunity against slower spreading strains to eventually reduce the R of the slower spreading strains below 1.
Flu viruses are more immunologically distinct than Covid strains (I'm pretty sure) but it would make sense that the same process could occur with those (although it seems that it also doesn't have to, given the coexistence of multiple endemic strains).
That puzzles me, too. My guess would be that at least part of the answer is behavioral and public health responses: stuff like staying home when sick, avoiding large gatherings, hand washing, etc should "tax" the transmission rate of all flu strains roughly evenly. Seasonal flu has an R value of about 1.2 at the start of the flu season, while pandemic flu usually have R0 of 2-3. If the pandemic response, whether organic or mandated, reduces the rate of spread of all influenza by about a third, then the pandemic flu is still spreading but the established season flu strains are slowly dying out (effective R value of 0.9).
Influenza has a pretty short generation time, about three days according to a preprint paper that came up on Google (Chan et al, 2024, Estimating the generation time for influenza transmission using household data in the United States), so effective R doesn't need to be much below 1 for very long for a given strain to almost disappear. Maybe six months for a 99.9% reduction using my guestimated R of 0.9.
I also find this confusing and I would note that it also happened during COVID, despite different variants giving imperfect protection against each other. In the UK, Alpha, Delta and Omicron each totally replaced the previously predominant strain, which never returned.
likely same as with covid - there's generally some cross-immunity and newer successful strains can outcompete older strains while leaving behind new cross-immunity. It's one of the reasons original covid strains are basically extinct at this point, new strains are very different, and humans are handling infections better even without any treatment.
When a new influenza strain emerges, infection with this strain indeed triggers an immune response not only against its own antigens but also against previously encountered strains, due to original antigenic sin. This cross-reactive immune response effectively suppresses both the new and old strains simultaneously, making it difficult for the old strains to maintain their presence in the population.
This mechanism is particularly important because, as you correctly note, simple competition for susceptible hosts isn't a sufficient explanation given that viral coinfections are possible. The immunological interference created by original antigenic sin provides a more complete explanation for strain succession. When people encounter new strains, their immune systems mount responses that inadvertently provide protection against older strains, even if those older strains weren't directly encountered in that infection.
This creates a powerful selective pressure favoring novel variants that can escape existing immunity while simultaneously suppressing older variants through cross-reactive immune responses.
Can someone explain why we need to cull animals that get the bird flu, instead of just letting them recover? The economic consequences (eggs are over $10 a dozen at my local grocery store) are very severe and at some point is it protecting us from spread anymore or just causing unwarranted supply chain issues?
Whoa, where is that grocery store?? I live in a large city where the cost of living is generally well above the US mean, and eggs here are $3.50/dozen. I just yesterday bought cage-free/no-antibiotics/hand-fed-by-virgins eggs (that last being my wife's description) for $6/dozen.
SF Bay Area. I will admit that I didn’t price shop, I went to go buy eggs at my local non-Whole Foods grocery store and the cheapest I could find was $10 and upwards of $15 and limit of 2. Only a month ago I got it for $5 a dozen and I was told bird flu was the cause of the inflation.
I’m on the Peninsula and see stores with prices of $10-15 (typically these are free-range or cage-free, but they don’t have anything more factory-produced), and stores that simply have no eggs at all. (And no, I’m not talking about hardware stores.)
In the current situation, those "free-range" chicks may well be stuck inside a building, too, to avoid further infections by wild birds. At least, that is how situations are handled in Germany: The "free-range" farmers wait for the admin to call out an "emergency", then (and only then) they put them inside and can still sell those eggs as "free-range". (If the farmers were more pro-active, they would lose that label, so they are contra-active.) - One more reason not to buy free-range ever - good luck to find "caged" anywhere nowadays!
The general strategy thus far for dealing with avian flu is to try to completely stamp it out and prevent spread, which means acting fast to cull all possibly infected birds. It should be noted that in some strains of avian influenza (like the currently circulating H5N1), they're likely to kill the entire flock within 48 hours or so anyway, so it's not just culling of poultry that leads to high egg prices (this is why some strains are labeled "HPAI" or highly pathogenic avian influenza; they're highly pathogenic in poultry).
That's gonna be challenging with the large migratory bird species that are being found dead along the Mississippi flyway. Snow geese and mute swans migrate thousands of miles and the US Midwest is in the middle of their routes.
I think a good argument can be made that the "stamping out" policy has failed as of 2024, yes. What matters now is if that argument is made within the public health bubble and gains enough support for changes to be made.
This is the agricultural bubble that is running things. I don’t think public health people have much power on agriculture, but agriculture people have all sorts of rules about preventing losses to their companies. (Including strict cull requirements to try to prevent virus spread, but implemented by corporate farmers who are just following rules rather than thinking about viral dynamics.)
Re: eggs are over $10 a dozen at my local grocery store
Wow-- where do you live and what grocery store do you shop at? Egg prices are definitely high here in Florida but not that high, not even at Whole Foods. (I assume we're talking about a dozen eggs?)
The response below addresses the incentive to cull birds rather than the need to cull them:
The government reimburses farmers for the loss of birds due to culling to prevent bird flu. According to the Department of Agriculture, farms that report outbreaks of bird flu quickly are eligible for compensation for the birds that are killed through culling. This program aims to encourage rapid reporting of outbreaks to control the spread of the virus. However, the government does NOT compensate for birds that die from the disease itself.
If you are a farmer who wants to stay in business/is risk adverse, what do you do if one birds tests positive?
It appears to me that increases in food prices and the downstream effects do NOT factor into the decision to cull birds.
I wonder if the resilience of farm workers is cos they have a long history of exposure, maybe even to lots of different minor variants, so their immune systems are already primed.
While we're on the subject: I have guy who delivers eggs. The other day I caught him cleaning blood off the eggs with a licked finger. I tried to impress on him that this probably wasn't wise. (We're in southern England, so I think we're mostly OK ATM)
Or then humans in good health, kept warm and well fed die very rarely from any flu, so that older strains were not any deadlier, but humans lived in poorer conditions and were therefore more suspectible.
People have always started dying from various infections when food starts running out.
I would think farm workers are almost all of "working age"; for many diseases, the most vulnerable are the very young and the very old, who are too young or too old to be doing much farm work. Farm children contribute to the farm work, but my impression from my farming cousins is that farm kids today mostly go to school and whatnot just like non-farm kids, and don't do much actual farm labor until they are teenagers who are old enough to no longer be unusually vulnerable to diseases.
My information on modern farming practices comes from my pig-farming cousins in Iowa, so it's possible my knowledge does not generalize across all American farms.
Certainly not the case in the UK: average age of farmers here is 58. Many are working past retirement age (40% of farm holders are over the age of 65). Average age of farm workers is about 40.
Not sure if its any different in the south but in a suburb in the outskirts of Newcastle I get regular letters from DEFRA asking me to declare any bird populations and monitor them for flu due to being a "high risk area"
>If reassortment is sort of like viral sex, pigs are sort of like Tinder.
And ACX is off to a hilarious start in 2025!
Now I'm imagining what H5N1's Tinder profile would be. "Baby I've got a fever... Wanna come back to my place and swap glycoproteins?"
Let's just hope the profile pic looks bad and nobody swipes right.
>Epidemiologists hate raw milk, think there is never any reason to drink it, and will announce that risks > benefits if the risk is greater than zero. I don’t know if the risk level is at a point where people who like raw milk should avoid it.
As a biologist (not an epidemiologist though) I agree. There is never any reason to drink raw milk and it can make you sick in other ways besides H5N1 (like bovine tuberculosis). Cheese made from raw milk is probably OK if it's aged for long enough that all the pathogens die off (several months), but I really don't understand why people like raw milk in the first place.
Huh, I've never noticed a difference. It seems some people are very sensitive to slight differences in milk taste and others are not. Or maybe it's a difference between people who drink large quantities of milk straight and those who mostly mix it with other things (tea, cereal, baked goods).
I've had it many times as a child. It does taste different; most people tend to either love it or find it disgusting.
I personally prefer the raw taste, but not enough to specifically seek it out (it's very hard to find in Canada anyway, as far as I know). But I would have it again if an opportunity presented itself.
I had raw milk on two occasions: 1. in Germany in a small supermarket catering also to some rich people (Heino among them). That milk is ... milked under extremely strict rules which make it more expensive but as safe as it gets. Think: "safe as honey" - still, do not give honey or raw milk to babies! The taste was really nice, and I do like milk (sucker). Would buy and drink again any time (was only double the normal price) 2. In Ukraine on a market in a re-used plastic bottle. Idiot me drunk it. And got so sick, I had to vomit it all out 2 hours later. Looked like pieces of white cheese. Some nearby chickens enjoyed it. Oh, traditional small scale farming, the romance of it!
In the US, cheese made from raw milk is required by law to be aged for at least 60 days, so if you can find it at the store, it'll be at least that aged. This mostly means that you won't see fresh style (e.g. mozzerella) or bloomy (brie) cheeses made from raw milk here; other styles are generally aged at least that long anyway.
Last time raw milk came up, the main argument I found enlightening was to think of it as relative risk vs absolute risk. It's supposedly around 1000x more dangerous than pasteurized milk (risk of hospitalization and death), but pasteurized milk has such insanely low casualty rates anyways that regular consumption of raw milk is practically equivalent (in risk) to increasing your daily car commute by a few meters, or eating one additional charcoal-grilled steak per year.
I grew up on a dairy farm, can attest that raw milk has a different (better for my taste) flavour to an extent. If I still lived on a dairy farm and had confidence in the hygiene practices I would drink it for sure just for the taste. The pro-health arguments are rubbish though, agree it's likely to be a risk not worth it if you are relying on hippie idiots to get their supply chain right, so I never try to find raw milk commercially.
I'm also not sure how many people confuse pastuerisation with homogenisation, I duspect at least a few people who think they don't like pastuerised milk really just don't like homogenised milk and don't know that's two different processes that don't have to go together.
"But the exact mortality pattern was surprising; people between 18 and 28 were especially likely to die, and people older than 88 especially likely to survive. Why? Because an H1N1 flu went pandemic in 1830; anyone who first encountered the flu around then had an immune system synced to H1N1."
What are the merits of deliberately creating a less lethal, more infectious strain of flu variants that we expect as particularly likely to start a pandemic? Do we have the prediction powers to see what variants are more likely than others to start a pandemic? Is it even possible to create less lethal and more transmissible versions of a virus that would also not cause an increase of overall deaths compared to doing nothing?
I suppose this is just transmissible vaccines, which I've seen discussed before. Is there anything different than what I just mentioned and what people normally discuss when they speak of transmissible vaccines?
Wuhan was I believe going for a more infectious (to humans) bat flu, but indifferent w/re lethality. The objective was not to create a less-lethal variant just to see if they could, and certainly not to create a less-lethal variant and set it loose to build herd immunity, but to answer the question in Scott's part 3: If this flying-critter airborne infection crosses over into humans, how bad would it be? If the crossovers were likely to be highly lethal, they wanted to know that.
Which would have been valuable knowledge, if we had it a few years ahead of the actual pandemic and if it stayed safely contained in a BSL-4 facility.
But the actual crossover might have different mutations and so be different from what they had. Meanwhile the response doesn't really need to know ahead of time.
Maybe, but we can't say if that was a good idea or not. Or rather, if COVID came from bats and not the lab, than the whole pandemic doesn't give us any new evidence on whether what the Wuhan lab was doing was good or bad.
> the Spanish Flu of 1918 was an H1N1 strain that killed about 2% of the world’s population. But the exact mortality pattern was surprising; people between 18 and 28 were especially likely to die, and people older than 88 especially likely to survive. Why? Because an H1N1 flu went pandemic in 1830; anyone who first encountered the flu around then had an immune system synced to H1N1. But an H3N8 flu went pandemic between 1890 and 1900; anyone who first encountered the flu then had an immune system synced to that strain and was unprepared for H1N1. See [here] for the details.
Hmm, I read something that attributed the different profile of the Spanish Flu to the radically different ecology of young men around 1918. To wit, if you were in a trench and got mildly sick, you stayed in your trench, whereas if you were in a trench and became deathly ill, you were shipped to an infirmary behind the lines where you could infect vastly more people drawn from a much wider geographic range. So there was strong selection for high lethality in young men. This contrasts with the normal state of affairs where if you're mildly sick you keep going about your business (just like in the trenches) but if you become severely ill, you withdraw into your personal home (opposite of the trenches).
Why is this theory of the Spanish Flu less convincing than "it was a coincidence"?
> But the biologists I talked to said people tend to overupdate on this, that evolution can do lots of weird things, and that the 1918 flu forgot to read the Evolutionary Virology textbook and actually mutated to get *worse*.
This... looks like strong support for the theory that the virus was selected for high lethality?
Huh, I was reading somewhere... (bad cattitude?) that a lot of Spanish flu deaths among young men was because doctors were treating them with the new wonder drug, Aspirin, and giving them lethal overdoses.
I suspect that towards the end of WW1 most people in Northern Europe weren't eating very healthily, what with the prowling U boats hampering convoys, and naval blockades.
So maybe nutritional deficiencies also made it harder to recover from a dose of Spanish Flu. Cooks back then also boiled vegetables for longer, so maybe lack of Vitamin C was one example of such a deficiency.
I read a book about the Spanish Flu (I believe it was “The Great Influenza”) and he attributed the young deaths to the cytokine storms - ie, young people with stronger immune systems ironically had stronger immune overreactions, and therefore died in greater numbers.
Was the flu selected in the trenches? The disease got its start in the Midwest United States (Kansas). However, it did take a year to get ripping, but I thought that was after the war. I believe people who got it during the war got the earlier, safer variant, and were later protected. Still possible it was selected in the trenches though when it got worse, just not sure about that.
Anyways, it’s less strange than Alexander suggested that a virus gets worse at first and then less dangerous later. Experiments on rabbits show that new viruses can become extremely dangerous as they adapt to a new host - they get better at spreading by bypassing the immune system after all. It takes time for the evolutionary logic to slow down the lethality.
We saw this with Covid: many of the first strains became more dangerous as they became better at spreading (like Delta) and then became less dangerous later (omicron). This was the pattern I was expecting based on some virologists I read, so it was intellectually satisfying to see it play out that way.
> The forecasters I talked to raised one other point of uncertainty: does the flu work more like a dice roll, or like a bus? Dice rolls are uncorrelated with their predecessors; even if it’s been a hundred rolls since you last rolled a 6, your chance this time is still 1/6. But buses come at fixed intervals; if the buses are hourly, and you haven’t seen a bus in the past 59 minutes, then your chance of seeing a bus in the next minute is very high.
This one seems easy; the flu is more like a bus. It's a motivated adversary working against you, not an indifferent phenomenon.
But this doesn't help to predict outbreaks; compare some phenomena that everyone agrees work the same way, earthquakes and volcanic eruptions. We know that the potential for these phenomena accumulates over time until eventually there's a breakout. The implications are... nothing, if you're worried about when exactly the next breakout might happen. They're mostly relevant to how regular a long-term average of the event rate should look.
Hmm, in the relevant sense, I believe that all three of these things are statistically distributed much more like dice than like buses. Even though the naive model of building tension for earthquakes makes you think it should be more bus-like, it turns out that the total amount of stored tension in the plates is so incredibly much greater than the amount of tension accumulated per century or dissipated per earthquake that frequencies don’t change significantly either when the most recent earthquake was recent or was long ago.
I'm not sure that buses have the statistical properties that you'd like to attribute to buses. Here's an example:
1. The bus is scheduled to arrive once per hour.
2. It is currently 45 minutes late.
3. What is the probability that the bus arrives within the next 5 minutes?
It's true that the arrival of buses is regulated. But it's not true that the extended delay we observe makes an imminent arrival likely. It makes it unlikely, because the effect dominating the arrival of the bus is whatever catastrophe has already made it 45 minutes late, not the overall goal of arriving once per hour. The chance of arriving within the next 5 minutes was much higher 48 minutes ago than it is now. If you were 14 minutes early for the bus, you haven't seen it arrive in 59 minutes, but it's not true that the chance of seeing a bus in the next one minute is high.
45 minute delay on a 60 minute bus schedule might be a sign of a “catastrophe” that would stop all bus travel. But 15 minute delay on a 20 minute bus schedule is actually a very common result of ordinary traffic. It’s true that there’s some complexity to it - it’s not just that the longer the delay is, the greater the per minute probability of arrival, because there are also other effects like the one you mention, as well as the fact that if one bus is delayed, the bus after it has a tendency to get *ahead* of schedule, since its less likely to find passengers at the stops and can just keep going.
But I don’t think we need a detailed model of bus arrival probabilities to distinguish the nearly time-independent poisson process of earthquakes and passenger arrivals at bus stops from the highly time-dependent processes of elections and bus arrivals at bus stops.
> Epidemiologists hate raw milk, think there is never any reason to drink it, and will announce that risks [exceed] benefits if the risk is greater than zero. I don’t know if the risk level is at a point where people who like raw milk should avoid it.
This raises an interesting contrast with anal sex. What do epidemiologists think about that?
Epidemiologists recognize value in anal sex (people like having anal sex) and so they spend money advertising public health advisories about using condoms and perhaps a drug regime that makes people having anal sex less likely to become infected with HIV (pre-exposure prophylaxis, advertised as "PrEP").
Epidemiologists don't believe them. They think the difference in flavour isn't real but is an artefact of knowing the milk is raw (ie you could sell pasteurised milk as raw milk and people wouldn't notice). I'm not convinced on this.
I'm not convinced either, having grown up with a house cow and consequently having a prolonged experience with raw milk decades before anyone asked the epidemiologists. (Although we were always very aware of zoonoses and cow health in general; living that closely with an animal makes the reciprocal nature of the relationship effortlessly obvious.)
Raw milk tastes distinctly different from pasturised milk. I mildly prefer pasteurised; raw milk tastes distinctly and quite strongly of 'cow'.
So where can I buy oseltamivir w/out a prescription?
Following
May well be outdated; but https://www.theindependentpharmacy.co.uk/cold-and-flu/tamiflu in the UK claims to send wo prescription if you do an online-self-diagnosis.
In Mexico, Freddys farmacia is recommended for all stuff wo prescription incl. tamiflu aka Oseltamivir in Puerto Vallarte https://www.tripadvisor.com/ShowTopic-g150793-i46-k10048642-o10-Freddy_s_farmacia-Puerto_Vallarta.html.
Oh, and ACX links to https://henryaj.substack.com/p/lifehacks who claims: "in most places you're just allowed to buy and import prescription medication and take it" linking to
https://www.unitedpharmacies-uk.md/ May be even useful if one wants to save on viagra:
https://www.unitedpharmacies-uk.md/erectile-dysfunction.html?&name-price=Lowest
Have a happy 2025!
Neat!
UK website looks to still be good. [Edit, but both only ship to the UK, dang. What's the international version, used to be dodgy Indian websites...]
I have been unable to find any source for the powder for suspension you'd need to properly make up the smaller doses for children, if anyone else finds it that'd be helpful.
Are your figures in addition to, or instead of, mortality from a normal flu season? It seems a solid chunk of your found expected h5n1 deaths would be expected to die from normal flu anyway, wouldn't h5n1 replace the normal flu? At the very least, anti pandemic measure would limit normal flu spread similar to anti covid.
https://www.cdc.gov/flu-burden/php/about/index.html#:~:text=While%20the%20burden%20of%20flu,the%20United%20States%20each%20year.&text=CDC%20estimates%20that%20flu%20has,annually%20between%202010%20and%202023.
Presumably, the anti-pandemic measures would be calibrated to the expected level of danger. An h5n1 pandemic that's closer to the 2009 pig flu outbreak would probably get a pretty tepid response, while something like the 1919 pandemic or worse would get the full Covid treatment.
If an outbreak happens that is seen as bad enough to justify lockdowns, I would expect them to be extremely effective to the extent they're complied with. The Covid response was initially modelled on planned responses for a particularly severe pandemic flu, and they work better on influenza than on covid due to stuff like droplets vs aerosols. "To the extent they're complied with" is the hard part, since everyone is still tired from last time.
Honestly, are governments going to even bother this time, considering the response to the previous lockdowns? People are just going to do whatever the hell they want. And ultimately, a few dozen million deaths isn't a big deal when it's spread out across the globe.
And of course the pandemic will then be worse, and so people will ask for stronger measures the next time. Weak men create hard times, etc.
A lot of the effective changes around Covid, at least in the UK, were voluntary and often started among the population before the government changed the advice or issued legal restrictions. The population (and the police…) often went beyond what the government had actually required, because they were worried and trying to protect themselves and others.
Of course, we also know that the UK government and scientific advisors were surprised by the high compliance of the population, making the lockdowns much more effective than expected: they had issued stricter advice than they thought was needed, and emphasised the dire consequences of not following it, because they assumed people would skirt around the rules much more than they did.
The things that needed national government support/central decisions that made a big difference to the spread (rather than the treatment) were the closing of schools (made a huge difference to spread in the UK each time) and the easy, free availability of relatively-reliable testing.
Possibly also the chunks of money paid out for businesses furloughing people, paying them to not work, but I can’t see that happening again.
The closing of schools was disasterous in many ways (rendered more disasterous by the choice to close playgrounds and keep them closed, and the slow introduction of things like ‘bubbles’), but also attendance in schools dropped noticeably before each closure, and every time Covid became more prevalent, and not just because students were actually ill. I also had students I did not see for a long time, because they or family members were immune-compromised. Probably a future pandemic needs something more flexible in place around schooling.
The UK testing system was fantastically expensive (for the taxpayer), but popular and widely-used. It gave people a way to manage risk when things were very uncertain, helped monitor what was going on in terms of spread and mutations, and could justify people taking the necessary time off from work in a legally and socially powerful way while they were contagious. I can see a similar system being used again in future pandemics.
...Maybe I am being too cynical. Still, the political situation in the US is a whole different can of worms. The administration has its own supporters to appeal to, and they absolutely do not want lockdowns again. So that's that.
I agree that the US is in a funny place around this: everything around Covid got fantastically split on political lines, didn’t it? I don’t pretend to know how Americans, or the US government, would react to a serious influenza pandemic.
But I think it is still worth bearing in mind that when people think they are in danger, they tend to move to protect themselves with or without government support. Lockdowns were dreadful and I hope never to see them again, but you probably need some sort of government response if the population starts withdrawing from and getting scared of each other.
I suspect if people started to see non-elderly non-otherwise-sick people dying around them, ESPECIALLY children, they'd comply more readily. I might be biased because I don't know anyone who died of covid but I know two people (both young adults and otherwise healthy) who died of flu related sepsis, but still.
I've never known anyone who died of Covid or the flu. But I did know two formerly healthy people, one a relative, who have died in recent years of antibiotic-resistant bacterial infections. And that could well be the wave of the future.
I'd love to believe this, and I probably even used to. Sandy Hook changed that. Half of (at least the American) population doesn't give two shits about dead children.
I don't mean to undermine the emotions here, but do you literally believe that? Like do you think that if you picked a random American and told them that a child had died, there would be a greater than 50% chance they would go, "I don't care about that at all?"
I think you actually mean something more like "Half of Americans don't support measures that would save children at a cost which (I believe) is negligible." And that might be true! But I doubt your model of the world actually matches what you said.
The "half" may be an exaggeration, but it doesn't change my intent. If you're looking for a more "accurate" sentence, I'd say something like "A significant percentage of the American population is okay with child deaths so long as they aren't inconvenienced in any way that impacts them directly." (Masking, vaccines, sensible gun laws, seat belts, speed limits in school zones, etc etc etc.)
A significant portion of America's population is so selfish, they've proven that they truly don't care ENOUGH about child deaths to be inconvenienced.
Summarized and exaggerated: "Half of (at least the American) population doesn't give two shits about dead children."
Ah, ok - that's clarifying, thank you. I personally think most of that percentage thinks the "inconvenience" is too great for the safety increase. Which, I mean, we all take actions that risk our lives and those of others (e.g. driving a car). Your examples reflect that as well - you say "speed limits in school zones" not "speed limits everywhere" presumably because you recognize that there would be a greater cost to a universal 35 mph speed limit, even though it would also save lives.
All of this is just to say: I think the people you are talking about mostly disagree with you about the efficacy or cost of the measures you have in mind, not whether children's lives are valuable. I find that to be a helpful perspective, because it suggests they could be convinced, or that some other measure might be something everyone could agree on.
Is there actually a droplet vs aerosol distinction for covid vs flu? My understanding is that Lindsey Marr was talking about aerosols for flu as early as 2015, and it took Covid for the mainstream health establishment to realize this can be relevant for this sort of virus. I don’t know how much they’ve changed their opinions on flu.
My info for might be outdated.on that particular point. Although flu spreading at all readily by aerosols seems unlikely to me, based on the differences in R0: the numbers I've seen for flu range from 1.2 (normal seasonal flu) to 3.0 (high estimate for the 1918-1920 pandemic), while I've seen newer covid variants estimated at 5.7. Even discounting for the 1.2 figure for seasonal.flu probably not being a true R0 due to cross-immunity with prior years' strains, that's a pretty big gap that gives me a significant prior for there being substantial differences in mode of transmission.
I have seen some much lower estimates for covid's R0, in the pandemic flu range, but they're either analysing the original outbreaks and thus had extremely wide confidence intervals for want of good data, or they're using data from 2020-2021 and thus measuring covid's R0 under lockdown conditions, not its R0 under normal conditions.
It could well be that the lower R0 is just related to different intrinsic ability to penetrate cells, or easier immune response, even if the aerosol aspect is the same. The issue Linsey Marr focused on is that for decades, aerosol transmission was treated as a weird abnormality a few infections had because of a few early 20th century misprints, rather than because of any evidence that it wasn’t common. This is the 2015 article that suggested to me that there is more to this debate than just covid:
https://www.popsci.com/take-deep-breath/
Can someone explain for me how/why strain replacement happens? Why doesn't each new strain that crosses over simply increase the overall rate of flu? Do the strains crowd each other out somehow?
I also wondered this and asked Claude about it:
The existing strains typically fade out because of population immunity dynamics, which create a hostile environment for them to continue spreading. Here's why:
After a strain has circulated widely, most of the population develops specific antibodies against it. This means:
1. The old strain has fewer susceptible hosts to infect
2. When it does infect someone, their existing partial immunity often leads to milder illness and reduced transmission
3. People are more likely to have some cross-protection against that strain in future seasons
Meanwhile, the new replacement strain has advantages:
- It can infect people who are immune to the old strain
- It spreads more efficiently through the population
- It may cause more severe symptoms due to lack of pre-existing immunity
This creates a competitive situation where:
- The old strain struggles to find new hosts
- The new strain rapidly spreads through the susceptible population
- Resources (susceptible hosts) become increasingly limited for the old strain
Think of it like market competition - once a new, more successful competitor arrives, the old business (strain) has trouble maintaining its customer base (susceptible hosts) and eventually can't sustain itself.
However, it's worth noting that old strains don't always completely disappear - sometimes they can persist at low levels or in specific geographical regions, potentially re-emerging if conditions become favorable again.
> Think of it like market competition - once a new, more successful competitor arrives, the old business (strain) has trouble maintaining its customer base (susceptible hosts) and eventually can't sustain itself.
This feels a little off. How does the new strain interfere with the old strain? The old strain naturally declines by two methods; the hosts become more resistant to it (real decline) and more resistant to its symptoms (apparent, but false, decline). Neither of those relies on the existence of a newer strain.
(All humans carry a significant number of completely asymptomatic diseases, which is why islanders get sick when they're visited by ships.)
You’re right, it doesn’t come out and say why there are fewer susceptible hosts for the old strain.
I assume for two reasons: Mortality and NPIs (spontaneous and mandated).
And one more: Non-specific antiviral responses such that infections from the new strain induce a temporary shield against the older strain thereby further isolating it.
This herd immunity (and/or mortality) and NPIs have been the accepted explanations for strain replacement. With COVID these explanations don't really fly. Of course, it's hard to compare NPI implementations from country to country, but with a highly contagious pathogen like COVID, even the strictest NPIs couldn't stop its spread once newer, more transmissible variants evolved (see China's attempt at ZeroCOVID).
We've seen...
1. A new variant can push out all the previous variants and become the dominant variant but not create a new wave — i.e., it takes up a bigger piece of a smaller pie. Examples of this are Alpha in the US and Kraken worldwide. In fact, in the US, Kraken may have prematurely ended the BQ.1 wave because (counterintuitively) US COVID cases dropped precipitously as Kraken became the dominant variant.
2. In two countries with a similar variant mix, a new variant can create a wave of new cases in one country but not another. Examples: Lamda and Gamma in South America (although they were roughly contiguous in time, so they may have possibly been burning each other's fuel). XEC seems to be starting a new wave in the US, but it didn't create a wave in Australia.
3. And then we have examples like Omicron that pushed every other variant aside and created new waves everywhere.
Other weirdness that's worth noting...
4. Even though regions may have similar variant mixes, some regions seem to experience two COVID waves each year (US and Australia), and some have three or four waves each year (France, Germany).
5. Possible examples of viral interference: There seems to be some sort of inverse relationship between SARS2 waves and Rhinovirus waves. RhVs peak when COVID is low. COVID peaks when RhVs are low. Influenza cases in the US (and worldwide) dropped precipitously to virtually zero just as COVID cases started ramping up. NPIs have been put forward as the post hoc explanation for the phenomenon, but this dropoff happened before NPIs were mandated. And except for a long, slow burn of Type B cases in China during their ZeroCOVID lockdown, flu stopped circulating until the advent of Omicron. Then it came back with a vengeance.
Ask it!
We did chat a bit more about it. That’s where the “non-specific antiviral response” part came out of.
Makes the carrier confine to home for the duration of illness and avoid seeing strangers? Drives up the fever making the body more hostile to pathogens in general?
Yeah this doesn’t explain why the old virus can’t just coexist alongside the new virus in the same individuals
Any single individual only has so many of the relevant type of cells to infect, whichever amino acids the viral particles are built from, etc.
"more resistant to its symptoms(apparent, but false, decline)"
some symptoms are among the best ways for diseases to spread, or if (e.g.) a host is less likely to sneeze or cough while having an illness, they're also less likely to spread it, even conditioning upon the same viral load.
"Neither of those relies on the existence of a newer strain."
If two strains of an illness are very similar, there is substantial cross-immunity.
Wild ass guess: most flu strains give cross immunity. The new strain has to have the unusual ability to infect people with immunity to the old strain, or it wouldn't be epidemic. The old strain has no such filtering.
I think this more or less has to be it. It makes sense of the replacement process of Covid strains - the faster spreading strains are providing those who catch them with sufficient immunity against slower spreading strains to eventually reduce the R of the slower spreading strains below 1.
Flu viruses are more immunologically distinct than Covid strains (I'm pretty sure) but it would make sense that the same process could occur with those (although it seems that it also doesn't have to, given the coexistence of multiple endemic strains).
Well, I think it all works out as long as the cross immunity is only partial.
Re: When it does infect someone, their existing partial immunity often leads to milder illness and reduced transmission
If the illness is mild shouldn't it be able to spread more easily insofar as its hosts aren't going to stay home and semi-isolate?
That puzzles me, too. My guess would be that at least part of the answer is behavioral and public health responses: stuff like staying home when sick, avoiding large gatherings, hand washing, etc should "tax" the transmission rate of all flu strains roughly evenly. Seasonal flu has an R value of about 1.2 at the start of the flu season, while pandemic flu usually have R0 of 2-3. If the pandemic response, whether organic or mandated, reduces the rate of spread of all influenza by about a third, then the pandemic flu is still spreading but the established season flu strains are slowly dying out (effective R value of 0.9).
Influenza has a pretty short generation time, about three days according to a preprint paper that came up on Google (Chan et al, 2024, Estimating the generation time for influenza transmission using household data in the United States), so effective R doesn't need to be much below 1 for very long for a given strain to almost disappear. Maybe six months for a 99.9% reduction using my guestimated R of 0.9.
I also find this confusing and I would note that it also happened during COVID, despite different variants giving imperfect protection against each other. In the UK, Alpha, Delta and Omicron each totally replaced the previously predominant strain, which never returned.
likely same as with covid - there's generally some cross-immunity and newer successful strains can outcompete older strains while leaving behind new cross-immunity. It's one of the reasons original covid strains are basically extinct at this point, new strains are very different, and humans are handling infections better even without any treatment.
There's some evolution in both directions.
Mostly it's the virus adapting, because their generation times are so short.
But you have a minor effect from the humans most vulnerable to covid-style viruses not surviving.
I think it's likely because of https://en.wikipedia.org/wiki/Original_antigenic_sin
Asking Claude about it:
When a new influenza strain emerges, infection with this strain indeed triggers an immune response not only against its own antigens but also against previously encountered strains, due to original antigenic sin. This cross-reactive immune response effectively suppresses both the new and old strains simultaneously, making it difficult for the old strains to maintain their presence in the population.
This mechanism is particularly important because, as you correctly note, simple competition for susceptible hosts isn't a sufficient explanation given that viral coinfections are possible. The immunological interference created by original antigenic sin provides a more complete explanation for strain succession. When people encounter new strains, their immune systems mount responses that inadvertently provide protection against older strains, even if those older strains weren't directly encountered in that infection.
This creates a powerful selective pressure favoring novel variants that can escape existing immunity while simultaneously suppressing older variants through cross-reactive immune responses.
Can someone explain why we need to cull animals that get the bird flu, instead of just letting them recover? The economic consequences (eggs are over $10 a dozen at my local grocery store) are very severe and at some point is it protecting us from spread anymore or just causing unwarranted supply chain issues?
Culling is to prevent further spread, reduce risk of mutations, and protect farm workers.
And who says they recover? Bird mortality can be very high - approaching 100% within 3-4 days post-infection (dpi)
https://www.mdpi.com/1999-4915/15/9/1909
Whoa, where is that grocery store?? I live in a large city where the cost of living is generally well above the US mean, and eggs here are $3.50/dozen. I just yesterday bought cage-free/no-antibiotics/hand-fed-by-virgins eggs (that last being my wife's description) for $6/dozen.
SF Bay Area. I will admit that I didn’t price shop, I went to go buy eggs at my local non-Whole Foods grocery store and the cheapest I could find was $10 and upwards of $15 and limit of 2. Only a month ago I got it for $5 a dozen and I was told bird flu was the cause of the inflation.
I'm also in Bay area and buy eggs for $4-5/dozen at Berkeley bowl
I'm in central Chicago, and didn't do any price shopping. Just did groceries at the nearest grocery store and that's what they had eggs for.
I’m on the Peninsula and see stores with prices of $10-15 (typically these are free-range or cage-free, but they don’t have anything more factory-produced), and stores that simply have no eggs at all. (And no, I’m not talking about hardware stores.)
In the current situation, those "free-range" chicks may well be stuck inside a building, too, to avoid further infections by wild birds. At least, that is how situations are handled in Germany: The "free-range" farmers wait for the admin to call out an "emergency", then (and only then) they put them inside and can still sell those eggs as "free-range". (If the farmers were more pro-active, they would lose that label, so they are contra-active.) - One more reason not to buy free-range ever - good luck to find "caged" anywhere nowadays!
Seconding Paul Botts’ question, where is this store; are the eggs gold-plated? That’s like 3x the price in my local store.
The general strategy thus far for dealing with avian flu is to try to completely stamp it out and prevent spread, which means acting fast to cull all possibly infected birds. It should be noted that in some strains of avian influenza (like the currently circulating H5N1), they're likely to kill the entire flock within 48 hours or so anyway, so it's not just culling of poultry that leads to high egg prices (this is why some strains are labeled "HPAI" or highly pathogenic avian influenza; they're highly pathogenic in poultry).
That's gonna be challenging with the large migratory bird species that are being found dead along the Mississippi flyway. Snow geese and mute swans migrate thousands of miles and the US Midwest is in the middle of their routes.
I think a good argument can be made that the "stamping out" policy has failed as of 2024, yes. What matters now is if that argument is made within the public health bubble and gains enough support for changes to be made.
This is the agricultural bubble that is running things. I don’t think public health people have much power on agriculture, but agriculture people have all sorts of rules about preventing losses to their companies. (Including strict cull requirements to try to prevent virus spread, but implemented by corporate farmers who are just following rules rather than thinking about viral dynamics.)
I think you're correct, I should've said "ag bubble," that's more accurate.
Re: eggs are over $10 a dozen at my local grocery store
Wow-- where do you live and what grocery store do you shop at? Egg prices are definitely high here in Florida but not that high, not even at Whole Foods. (I assume we're talking about a dozen eggs?)
The response below addresses the incentive to cull birds rather than the need to cull them:
The government reimburses farmers for the loss of birds due to culling to prevent bird flu. According to the Department of Agriculture, farms that report outbreaks of bird flu quickly are eligible for compensation for the birds that are killed through culling. This program aims to encourage rapid reporting of outbreaks to control the spread of the virus. However, the government does NOT compensate for birds that die from the disease itself.
If you are a farmer who wants to stay in business/is risk adverse, what do you do if one birds tests positive?
It appears to me that increases in food prices and the downstream effects do NOT factor into the decision to cull birds.
I wonder if the resilience of farm workers is cos they have a long history of exposure, maybe even to lots of different minor variants, so their immune systems are already primed.
While we're on the subject: I have guy who delivers eggs. The other day I caught him cleaning blood off the eggs with a licked finger. I tried to impress on him that this probably wasn't wise. (We're in southern England, so I think we're mostly OK ATM)
Or then humans in good health, kept warm and well fed die very rarely from any flu, so that older strains were not any deadlier, but humans lived in poorer conditions and were therefore more suspectible.
People have always started dying from various infections when food starts running out.
I would think farm workers are almost all of "working age"; for many diseases, the most vulnerable are the very young and the very old, who are too young or too old to be doing much farm work. Farm children contribute to the farm work, but my impression from my farming cousins is that farm kids today mostly go to school and whatnot just like non-farm kids, and don't do much actual farm labor until they are teenagers who are old enough to no longer be unusually vulnerable to diseases.
My information on modern farming practices comes from my pig-farming cousins in Iowa, so it's possible my knowledge does not generalize across all American farms.
Certainly not the case in the UK: average age of farmers here is 58. Many are working past retirement age (40% of farm holders are over the age of 65). Average age of farm workers is about 40.
If they are working past retirement age, they probably are less frail than the average person at that age?
Maybe. But they may just be occupying the land, not actually working.
There have certainly been cases where they were still working, still physically able, but no longer mentally able and not aware of it.
Not sure if its any different in the south but in a suburb in the outskirts of Newcastle I get regular letters from DEFRA asking me to declare any bird populations and monitor them for flu due to being a "high risk area"
The various zones come and go. You don't seem to be in one now. Map here:
https://defra.maps.arcgis.com/apps/webappviewer/index.html?id=8cb1883eda5547c6b91b5d5e6aeba90d
>If reassortment is sort of like viral sex, pigs are sort of like Tinder.
And ACX is off to a hilarious start in 2025!
Now I'm imagining what H5N1's Tinder profile would be. "Baby I've got a fever... Wanna come back to my place and swap glycoproteins?"
Let's just hope the profile pic looks bad and nobody swipes right.
>Epidemiologists hate raw milk, think there is never any reason to drink it, and will announce that risks > benefits if the risk is greater than zero. I don’t know if the risk level is at a point where people who like raw milk should avoid it.
As a biologist (not an epidemiologist though) I agree. There is never any reason to drink raw milk and it can make you sick in other ways besides H5N1 (like bovine tuberculosis). Cheese made from raw milk is probably OK if it's aged for long enough that all the pathogens die off (several months), but I really don't understand why people like raw milk in the first place.
> but I really don't understand why people like raw milk in the first place.
I don't think I've ever had raw milk, but I can tell you that, compared to pasteurized milk, ultrapasteurized milk tastes absolutely disgusting.
Huh, I've never noticed a difference. It seems some people are very sensitive to slight differences in milk taste and others are not. Or maybe it's a difference between people who drink large quantities of milk straight and those who mostly mix it with other things (tea, cereal, baked goods).
I know raw milk drinkers, the reasons they give are:
1. It tastes different
2. It’s a nutrition heuristic that raw foods have different nutrient distributions than cooked foods, so maybe it’s good to have a mix of both
3. Probiotics
4. Supports small farms by buying direct from the farmer
I’ve tried it once and can vouch for the taste being different.
I've had it many times as a child. It does taste different; most people tend to either love it or find it disgusting.
I personally prefer the raw taste, but not enough to specifically seek it out (it's very hard to find in Canada anyway, as far as I know). But I would have it again if an opportunity presented itself.
Out of these, I would say that 1 is a good reason, 2 and 3 are likely bad reasons that are easy to talk populists into, and 4 is harder to say.
I had raw milk on two occasions: 1. in Germany in a small supermarket catering also to some rich people (Heino among them). That milk is ... milked under extremely strict rules which make it more expensive but as safe as it gets. Think: "safe as honey" - still, do not give honey or raw milk to babies! The taste was really nice, and I do like milk (sucker). Would buy and drink again any time (was only double the normal price) 2. In Ukraine on a market in a re-used plastic bottle. Idiot me drunk it. And got so sick, I had to vomit it all out 2 hours later. Looked like pieces of white cheese. Some nearby chickens enjoyed it. Oh, traditional small scale farming, the romance of it!
In the US, cheese made from raw milk is required by law to be aged for at least 60 days, so if you can find it at the store, it'll be at least that aged. This mostly means that you won't see fresh style (e.g. mozzerella) or bloomy (brie) cheeses made from raw milk here; other styles are generally aged at least that long anyway.
Last time raw milk came up, the main argument I found enlightening was to think of it as relative risk vs absolute risk. It's supposedly around 1000x more dangerous than pasteurized milk (risk of hospitalization and death), but pasteurized milk has such insanely low casualty rates anyways that regular consumption of raw milk is practically equivalent (in risk) to increasing your daily car commute by a few meters, or eating one additional charcoal-grilled steak per year.
I grew up on a dairy farm, can attest that raw milk has a different (better for my taste) flavour to an extent. If I still lived on a dairy farm and had confidence in the hygiene practices I would drink it for sure just for the taste. The pro-health arguments are rubbish though, agree it's likely to be a risk not worth it if you are relying on hippie idiots to get their supply chain right, so I never try to find raw milk commercially.
I'm also not sure how many people confuse pastuerisation with homogenisation, I duspect at least a few people who think they don't like pastuerised milk really just don't like homogenised milk and don't know that's two different processes that don't have to go together.
Crazy Idea:
Thinking of this quote:
"But the exact mortality pattern was surprising; people between 18 and 28 were especially likely to die, and people older than 88 especially likely to survive. Why? Because an H1N1 flu went pandemic in 1830; anyone who first encountered the flu around then had an immune system synced to H1N1."
What are the merits of deliberately creating a less lethal, more infectious strain of flu variants that we expect as particularly likely to start a pandemic? Do we have the prediction powers to see what variants are more likely than others to start a pandemic? Is it even possible to create less lethal and more transmissible versions of a virus that would also not cause an increase of overall deaths compared to doing nothing?
I suppose this is just transmissible vaccines, which I've seen discussed before. Is there anything different than what I just mentioned and what people normally discuss when they speak of transmissible vaccines?
This is roughly what the Wuhan lab was doing with coronaviruses.
Wuhan was I believe going for a more infectious (to humans) bat flu, but indifferent w/re lethality. The objective was not to create a less-lethal variant just to see if they could, and certainly not to create a less-lethal variant and set it loose to build herd immunity, but to answer the question in Scott's part 3: If this flying-critter airborne infection crosses over into humans, how bad would it be? If the crossovers were likely to be highly lethal, they wanted to know that.
Which would have been valuable knowledge, if we had it a few years ahead of the actual pandemic and if it stayed safely contained in a BSL-4 facility.
Makes sense
Scott & Alina Chan of LL fame are exchanging related to this currently on X https://x.com/slatestarcodex/status/1874641242786791779?t=XpInk6EXT5jkpY5zwDyCEA&s=19
But the actual crossover might have different mutations and so be different from what they had. Meanwhile the response doesn't really need to know ahead of time.
Maybe, but we can't say if that was a good idea or not. Or rather, if COVID came from bats and not the lab, than the whole pandemic doesn't give us any new evidence on whether what the Wuhan lab was doing was good or bad.
> the Spanish Flu of 1918 was an H1N1 strain that killed about 2% of the world’s population. But the exact mortality pattern was surprising; people between 18 and 28 were especially likely to die, and people older than 88 especially likely to survive. Why? Because an H1N1 flu went pandemic in 1830; anyone who first encountered the flu around then had an immune system synced to H1N1. But an H3N8 flu went pandemic between 1890 and 1900; anyone who first encountered the flu then had an immune system synced to that strain and was unprepared for H1N1. See [here] for the details.
Hmm, I read something that attributed the different profile of the Spanish Flu to the radically different ecology of young men around 1918. To wit, if you were in a trench and got mildly sick, you stayed in your trench, whereas if you were in a trench and became deathly ill, you were shipped to an infirmary behind the lines where you could infect vastly more people drawn from a much wider geographic range. So there was strong selection for high lethality in young men. This contrasts with the normal state of affairs where if you're mildly sick you keep going about your business (just like in the trenches) but if you become severely ill, you withdraw into your personal home (opposite of the trenches).
Why is this theory of the Spanish Flu less convincing than "it was a coincidence"?
> But the biologists I talked to said people tend to overupdate on this, that evolution can do lots of weird things, and that the 1918 flu forgot to read the Evolutionary Virology textbook and actually mutated to get *worse*.
This... looks like strong support for the theory that the virus was selected for high lethality?
Huh, I was reading somewhere... (bad cattitude?) that a lot of Spanish flu deaths among young men was because doctors were treating them with the new wonder drug, Aspirin, and giving them lethal overdoses.
Geez, aspirin? Aspirin is a fever suppressant, the exact opposite of what you want if you have a dangerous illness.
That's like treating a broken leg with morphine.
Rasputin gained a lot of cache with the Russian royal family when he took hemophiliac Alexi off of aspirin.
Depends on if you think the thing that kills you is the virus or the immunoresponse.
For Spanish flu my pet theory is that it was mostly the immunoresponse (which handily explains why young people had it the worst).
I have vague memories of a "cytokine storm" being a theorized cause?
Sure seems like anti-inflammatory drugs like aspirin should help, at least pending empirical evidence in the other direction.
I thought the high fatality rates were true internationally, not just in places with ready availability of aspirin.
I suspect that towards the end of WW1 most people in Northern Europe weren't eating very healthily, what with the prowling U boats hampering convoys, and naval blockades.
So maybe nutritional deficiencies also made it harder to recover from a dose of Spanish Flu. Cooks back then also boiled vegetables for longer, so maybe lack of Vitamin C was one example of such a deficiency.
I agree with the implicit claim that if the Spanish Flu were to happen today, fatality rates will likely be lower even without vaccines.
Yeah this is the first time I see someone linking it to a 1830 pandemic. Is this a new theory?
I read a book about the Spanish Flu (I believe it was “The Great Influenza”) and he attributed the young deaths to the cytokine storms - ie, young people with stronger immune systems ironically had stronger immune overreactions, and therefore died in greater numbers.
Was the flu selected in the trenches? The disease got its start in the Midwest United States (Kansas). However, it did take a year to get ripping, but I thought that was after the war. I believe people who got it during the war got the earlier, safer variant, and were later protected. Still possible it was selected in the trenches though when it got worse, just not sure about that.
Anyways, it’s less strange than Alexander suggested that a virus gets worse at first and then less dangerous later. Experiments on rabbits show that new viruses can become extremely dangerous as they adapt to a new host - they get better at spreading by bypassing the immune system after all. It takes time for the evolutionary logic to slow down the lethality.
We saw this with Covid: many of the first strains became more dangerous as they became better at spreading (like Delta) and then became less dangerous later (omicron). This was the pattern I was expecting based on some virologists I read, so it was intellectually satisfying to see it play out that way.
Check out the wikipedia page. The wave that originated in Kansas was mild, but the second strain originated in France and was much deadlier.
> The forecasters I talked to raised one other point of uncertainty: does the flu work more like a dice roll, or like a bus? Dice rolls are uncorrelated with their predecessors; even if it’s been a hundred rolls since you last rolled a 6, your chance this time is still 1/6. But buses come at fixed intervals; if the buses are hourly, and you haven’t seen a bus in the past 59 minutes, then your chance of seeing a bus in the next minute is very high.
This one seems easy; the flu is more like a bus. It's a motivated adversary working against you, not an indifferent phenomenon.
But this doesn't help to predict outbreaks; compare some phenomena that everyone agrees work the same way, earthquakes and volcanic eruptions. We know that the potential for these phenomena accumulates over time until eventually there's a breakout. The implications are... nothing, if you're worried about when exactly the next breakout might happen. They're mostly relevant to how regular a long-term average of the event rate should look.
Hmm, in the relevant sense, I believe that all three of these things are statistically distributed much more like dice than like buses. Even though the naive model of building tension for earthquakes makes you think it should be more bus-like, it turns out that the total amount of stored tension in the plates is so incredibly much greater than the amount of tension accumulated per century or dissipated per earthquake that frequencies don’t change significantly either when the most recent earthquake was recent or was long ago.
I'm not sure that buses have the statistical properties that you'd like to attribute to buses. Here's an example:
1. The bus is scheduled to arrive once per hour.
2. It is currently 45 minutes late.
3. What is the probability that the bus arrives within the next 5 minutes?
It's true that the arrival of buses is regulated. But it's not true that the extended delay we observe makes an imminent arrival likely. It makes it unlikely, because the effect dominating the arrival of the bus is whatever catastrophe has already made it 45 minutes late, not the overall goal of arriving once per hour. The chance of arriving within the next 5 minutes was much higher 48 minutes ago than it is now. If you were 14 minutes early for the bus, you haven't seen it arrive in 59 minutes, but it's not true that the chance of seeing a bus in the next one minute is high.
45 minute delay on a 60 minute bus schedule might be a sign of a “catastrophe” that would stop all bus travel. But 15 minute delay on a 20 minute bus schedule is actually a very common result of ordinary traffic. It’s true that there’s some complexity to it - it’s not just that the longer the delay is, the greater the per minute probability of arrival, because there are also other effects like the one you mention, as well as the fact that if one bus is delayed, the bus after it has a tendency to get *ahead* of schedule, since its less likely to find passengers at the stops and can just keep going.
But I don’t think we need a detailed model of bus arrival probabilities to distinguish the nearly time-independent poisson process of earthquakes and passenger arrivals at bus stops from the highly time-dependent processes of elections and bus arrivals at bus stops.
>45 minute delay on a 60 minute bus schedule might be a sign of a “catastrophe” that would stop all bus travel.
Might be, but ime it's much more likely a sign of a "catastrophe" that stopped or significantly delayed the travel of that particular bus.
> Epidemiologists hate raw milk, think there is never any reason to drink it, and will announce that risks [exceed] benefits if the risk is greater than zero. I don’t know if the risk level is at a point where people who like raw milk should avoid it.
This raises an interesting contrast with anal sex. What do epidemiologists think about that?
Epidemiologists recognize value in anal sex (people like having anal sex) and so they spend money advertising public health advisories about using condoms and perhaps a drug regime that makes people having anal sex less likely to become infected with HIV (pre-exposure prophylaxis, advertised as "PrEP").
Surely the people who drink raw milk do like to eat raw milk though, no?
Epidemiologists don't believe them. They think the difference in flavour isn't real but is an artefact of knowing the milk is raw (ie you could sell pasteurised milk as raw milk and people wouldn't notice). I'm not convinced on this.
I'm not convinced either, having grown up with a house cow and consequently having a prolonged experience with raw milk decades before anyone asked the epidemiologists. (Although we were always very aware of zoonoses and cow health in general; living that closely with an animal makes the reciprocal nature of the relationship effortlessly obvious.)
Raw milk tastes distinctly different from pasturised milk. I mildly prefer pasteurised; raw milk tastes distinctly and quite strongly of 'cow'.
Raw milk cheese also tastes different.
This, at least, is definitely not true. Raw milk has a distinct sulfur note and smells like... cow.
People who drink raw milk have the option to switch to pasteurised milk; gay men don’t have comparable alternatives to anal sex. Though it seems that the risks to women of anal sex aren’t sufficiently advised about. https://amp.theguardian.com/society/2022/aug/11/rise-in-popularity-of-anal-sex-has-led-to-health-problems-for-women.
New EconTalk episode dishing raw milk: https://podcasts.apple.com/au/podcast/econtalk/id135066958?i=1000682124667