One thing that makes Covid-19 a bit different from other viruses that have trended towards milder infections is its long incubation time (while still being infectious). So it can already spread pretty easily even if it's killing those people, and thus there's not a lot of darwinian pressure to become milder.

It might still trend towards that over time, but it will probably be slower at it than other viruses, and thus the 'it gets milder over time' aphorism about viruses doesn't apply as strongly to Covid-19.

That’s only one pressure driving viral evolution.

Another pressure is social response to infection: so far, the predominant response to a more-virulent or more-lethal strain has been lockdown and/or other behavioral change, see: delta in India. On the flip side, a variant that causes milder symptoms is less likely to cause the same social response, which allows the virus to spread more widely.

This may (may!) be what we’re observing with Omicron. In any case, we’ll know much more in a few weeks.

I'd say that's normally true, but there's not a ton of lockdowns with even Delta right now, so while I agree that it probably isn't going to get worse, there's still not a ton of pressure from social response as it is right now for it to get milder (maybe you could argue pressure from vaccines I guess).

Sure, and this is where the immune erosion comes into play: as vaccination rates go up, so does selection pressure for immune erosion.

Re: delta, think of this counterfactual — what would happen if a new variant appeared that is more lethal, more virulent, and that caused hospitalizations more quickly? You can bet the lockdowns would ratchet up.

With Omicron, we didn’t even notice it until there was a random genome scan. Viral evolution rarely selects for host death — it’s just not useful as a goal. There’s also probably relatively weak selection pressure for symptoms, though probably casing sneezes in hosts helps with spread, it also acts as a signal that can cause the host to isolate and halt spread. (See: SARS v1.)

From a selection pressure perspective, the ideal virus (among humans!) probably goes completely unnoticed for as long as possible. Humans won’t fight something they don’t even notice, and we probably have tons of these viruses everywhere and don’t even know about them.

Yeah I think we can learn what society thinks is acceptable risk just by looking at death counts over the last two years. They've pretty much leveled out. So as the virus mutates into less (or more) harmful variations, we'll see the world react less (or more) strictly, but overall death count will remain stable.

Presumably the steady state will happen after the virus mutates into a less dangerous variant (as infecting more people will be evolutionarily advantageous for the virus), and eventually we all get it and build up antibodies, at which point its virulence further subsides.

Do you realise that mutation is not something the species can consciously control?

How could the virus have mutated to become milder to suit the social response if the society, as it stands today, is continuing to do what they previously did..?

There's likely millions of virus variants out there already. Some will evolve to become much more contagious and most won't do much of anything.

When a more severe variant starts dominating and everyone goes inside and stops spreading it that makes the more severe variant die out. As it dies out, and people go outside, variants that are milder will thrive because people won't lockdown as easily. This can repeat forever until eventually the virus is completely benign. Of course it's all based on randomness so it's not exactly predictable what will happen.

> When a more severe variant starts dominating and everyone goes inside and stops spreading it that makes the more severe variant die out

Except this doesn't happen, otherwise COVID would have already stopped spreading. Our social response to COVID isn't adequate enough to stop its spread. We can't get people to actually lock down or stop transmission.

Except "Everybody goes inside and stops spreading" is not something that is unique to a variant of concern.

When everybody goes inside, all the variants stop spreading. The pressure is the same on every variants, so there's no source of evolutionary advantages on any variants.

> no source of evolutionary advantages on any variants.

Any variant that is able to co-exist with (rather than kill) its host during a full lockdown has a significant advantage. We co-exist with several coronaviruses already.

Only true if the virus lasts long enough to outlive a lockdown. So far that hasn’t been the case for the vast majority of single hosts.

> Except "Everybody goes inside and stops spreading" is not something that is unique to a variant of concern.

But it is — this will not be the response to a variant that doesn’t cause an increase in hospitalizations or deaths.

A given response exerts pressure on all variants, of course. But a milder variant may be able to spread widely before a more lethal one shows up and ratchets up lockdowns.

It's like the "invisible hand" or the "selfish gene" - just an analogy.

> Do you realise that mutation is not something the species can consciously control?

Very much so, and my post was carefully worded to refer to evolutionary selection pressure, and not any kind of agency on behalf of a virus.

I assumed the audience here would get it.

As an aside, there are much kinder ways to voice this kind of concern that don’t drip with contempt; consider taking a less accusatory tone in the future? That might trigger fewer downvotes the next time the reading comprehension module needs coffee.

The lockdown prevents severe mutations from replicating because of lockdowns. Milder mutations don't result in lockdowns hence ...

Or do lockdowns just prolong the inevitable?

Lockdowns are certainly a prolonging step.

But prolonging the inevitable is still useful, in that it slows down the number of hospitalizations on your march towards the inevitable.

Most hospital systems in the USA are stressed to severe levels. We had a nursing shortage before COVID19 even began, and the increased workload from COVID19 is clearly causing problems.

------

Slowing things down gives more time for the old patients to leave the hospital, meaning doctors/nurses won't have to work as hard.

It's true on an individual level too: someone who is very sick will stay at home, but someone with a mild head cold will still go out.

There's going to be a negative correlation between severity and something like "velocity" of infected persons.

Which begs the question of the long term utility and effect of symptom suppressing vaccines (that don't prevent transmission).

If a virus that would make make you very sick is contracted by someone who's vaccinated - They will now be in the 2nd category you describe ~'Sick, but not sick enough to stop them going out' -- They're now giving mobility to that virus that it would have otherwise not had (If they weren't vaccinated and were feeling too sick to go out and about)

So by usage of vaccines that only prevent severe symptoms, we pervert that normal negative feedback loop of virus spread and instead set the parameters to be more likely to propagate severe virus' that would usually keep you sick at home.

Vaccines don't suppress symptoms. They prevent symptoms by decreasing the viral load. They prevent transmission in the same way.

No, not _all_ transmission. Neither _all_ symptoms.

But has this been true in practice? Was any variant successfully stopped with lockdowns? I suppose hypothetically if there’s a new strain with like a 10% fatality rate there will be new extreme measures where that could be the case, but it hasn’t happened yet. Even the extreme early 2020 measures failed to stop the virus.

Which lockdowns in our hemisphere had the goal of stopping the virus? All I know about (or have been part of) never stated that goal, only goal was to break waves and avoid overload of health care systems, which most at least helped with.

And then it goes on because measures are reduced, and zero-covid never was a goal... would even go so far that some few countries pretty successfully did zero-covid strategies and stopped their virus, though sure, that is a pretty futile thing if the world is not "globally" doing this and you are likely not isolating forever/completely - so we just start to live with it.

> Was any variant successfully stopped with lockdowns?

Statstically speaking, almost surely yes: with millions of infections there are thousands of variants, every day. If somebody was prevented from infecting somebody with their variant (as patient 0) technically we stopped variants with lockdown. Maybe they more favorable for us, maybe they were more dangerous from us.

The mu variant, which caused a lot of worry, went away. It apparently escaped vaccines better than delta but spread much more slowly, so delta out-competed it. Perhaps if there were no restrictions it would have done better. Hard to say though.

It's not binary; if the virus is less infectious during incubation than during full infection, there's still selective pressure for milder symptoms.

A new variant need only replicate slightly faster to dominate (eventually).

Yeah I even said it would probably trend towards milder over time, just more slowly. That's definitely not a binary claim.

I think that the basic severity-lowering gradient still applies, but now it happens in two dimensions. Furthermore, it seems plausible that the sequence of variants "Original" -> Delta -> Omicron follows such a pattern.

First, evolutionary pressure will select for a lower dormancy period, i.e. a lower time to contagion. This is because a variant with a shorter dormancy will spread faster than one with a longer dormancy period if both are present in the same population. This is also exactly what happened with the Delta variant, which is asymptomatic generally for a shorter period[1] than the original variant, because it replicates faster.

Second, the usual process occurs for the usual reasons. This may be happening with Omicron.

[1]: https://www.health.govt.nz/our-work/diseases-and-conditions/...

Sort of. Sometimes.

Viruses evolve according to evolutionary pressures. I'm surprised we don't see Covid mutations that invalidate PCR tests - perhaps it's too soon, or even, who knows, maybe they are there and we don't know?

The common trope is that if the virus is too deadly, then it doesn't have a chance to find new hosts. This is kind-of what happened to "SARS-1" - it tended to kill its hosts before they could pass it on.

Unfortunately, this is only a weak effect in Covid. Most of the spread happens well before people are very ill; in terminal stages of severe Covid, many people in fact would test negative for virus presence. So there is little evolutionary pressure on the virus to be less deadly - by the time it kills people, we are spent vessels from its point of view.

I suppose a mutation that makes it completely benign would be beneficial, as current rules on self-isolation curb the spread of strains causing symptomatic disease.

There are pressures that make it more deadly: a mutation leading to higher viral load is both more deadly and more transmissible. So unless the transmissibility increases too much (and virus is way too deadly), we might even expect the virus to become more deadly. This hasn't happened yet, thankfully.

An example I used elsewhere was the introduction of Myxomatosis in Australia, to control the invasive rabbit population. Over time, the local strain evolved to be more deadly, but with a longer incubation periods. This way, bunnies had more time to pass the virus around, then once that job was done, the virus was happy to kill them off.

On the flipside (and very long timescales), since some people appear to be genetically more resistant to the virus, you can expect natural selection to do its job. But since overall death rate from Covid is low-ish (no more than 1%, you'd say, and skewed towards older, non-reproducing individuals) this would take a long time.

Another hope is that, if the disease is less serious in children, perhaps immunity it built up during the usual period of childhood infections. There is another coronavirus, which causes a mild cold in children, but can be very serious in adults (apologies, I can't remember what it's called). Thankfully, a single infection gives more-or-less complete lifetime protection - so most people have it as children and never worry about it again.

In either of these cases, we're talking about generational timescales.

The current PCR tests detect three different regions of RNA that are expected to be relatively well conserved, so there's no single mutation that can cause PCR tests to not detect them and we'll hopefully get a substantial early warning that allows the tests to be adapted if they ever head in that direction. One of the three markers detected by some of the PCR tests is actually invalidated by Omicron and some other variants but it's likely either bad luck or the result of some other benefit this mutation gives the virus. In order to get much benefit the virus would basically have to dodge all the things the PCR tests check for at once and that's hard to evolve.

Yeah, Alpha had the same mutation as Omicron too.

I'm not a virologist by any stretch of imagination. Mutations seem to come in bundles; I'm just surprised that, given typical PCR tests only test for three RNA fragments, that we don't see mutations that evade it.

The main evolutionary pressure on Covid in developed countries is Public Health clamping down on it like a ton of bricks. A variant that quietly evades PCR tests, self-isolation etc. is set up for life.

> So there is little evolutionary pressure on the virus to be less deadly - by the time it kills people, we are spent vessels from its point of view.

A variant that infects more people from a single host (assuming everything else held constant) is a positive selective pressure. So increasing the contagious period is beneficial. But the length of the contagious period is inversely related to the timing and strength of the immune response. So we should expect the virus to become less deadly over time, assuming there's a hard limit to its ability to evade a typical immune system.

> I'm surprised we don't see Covid mutations that invalidate PCR tests

The people who design the tests know which parts of the sequence are more likely to mutate, so they pick the more conserved parts. Also, most PCR tests check for 3 different parts of the sequence. Which is why Omicron can still be detected even if there is a mutation in one part checked by the test, since the other 2 parts still match (the so called S-gene-dropout).

Apparently there are some variants of Omicron that can't be detected with current PCR tests. We may need to update the list of primers at some point.

https://www.theguardian.com/world/2021/dec/07/scientists-fin...

Just to clarify, as it's not clear from that sentence: it can still be detected with PCR, just not distinguished from other variants as easily.

> An example I used elsewhere was the introduction of Myxomatosis in Australia, to control the invasive rabbit population. Over time, the local strain evolved to be more deadly, but with a longer incubation periods. This way, bunnies had more time to pass the virus around, then once that job was done, the virus was happy to kill them off.

Do you have other examples besides Myxomatosis?

> This is kind-of what happened to "SARS-1" - it tended to kill its hosts before they could pass it on.

Quick nit: while this is true, SARS-1 was also not particularly transmissible before symptoms appeared, in contrast to Covid, which made control much easier (in addition to the much higher fatality rate causing everyone to take it much more seriously).

> I'm surprised we don't see Covid mutations that invalidate PCR tests

PCR tests for RNA or DNA, that's not something a mutation can change. I don't know for Covid, but usually a PCR test targets something fundamental & stable. That might need to be updated with new variants.

Isn't that the only thing a mutation can change? What is mutating in the virus if it's not the DNA?

They test for particular fragments, not the whole thin. 3 fragments usually in the UK.

Viruses that evade the PCR test for Covid would be a different species almost by definition, and with enough mutations that could happen. These tests don’t detect other common cold causing coronaviruses for example.

Not necessarily. Smallpox was around for a very long time till we vaccinated our way out of it. Same with polio.

No one replying has a "serious level of knowledge", so instead, turn to sources that do (or report serious sources!)

https://www.smithsonianmag.com/science-nature/how-viruses-ev...

https://pubmed.ncbi.nlm.nih.gov/30734920/ (really good one)

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7255208/

In short, we don't know and won't know until this is all "over" a few years in the future, at least.

Imo part of the joy of hn is that this isn't actually true. Obviously not everyone knows what they are talking about, but some of the people who really do have serious knowledge on any given topic are on hn, and the average person here is at least slightly less likely to think they have knowledge when in fact they don't.

Yeah I remember in June of 2020 discussing the new aerosol data on HN and cringing about the 6 ft recommendation with cloth masks indoors. Finally a year+ later the WHO admitted as such (aerosol risk can be >50 ft for hours after a spread event), so sometimes we are ahead :)

> "aerosol risk can be >50 ft for hours after a spread event"

no, that's cringeworthy fearmongering. the virus is fragile and falls apart relatively quickly outside the body. the main risk behavior is having long conversations in close quarters where you're constantly exposed to a warm airstream of live virus.

In the right humidity conditions with poor airflow it has been found to be stable for hours, this is how there were super spreader events from 1>500 in certain types of factories.

no, any explanation of a ‘super spreader event’ is making a post hoc rationalization, arising from the biases of the explainers. thermo-fluid models of such conditions are often similarly biased, but have the additional proof burden of not representing observed phenomena, just assumptions and suppositions about it. it’s just as likely that factory workers having lunch together is the cause of such a ‘super spreader event’. occam’s razor suggests we start with the simplest explanations before considering convolutions like that.

make no mistake, if covid menacingly floating in the air to catch you unaware were actually a meaningful method of spread, we’d have evidence all over the place being reported feverishly by those very same biased reporting sources.

From what I know as not being a professional doctor, I would say that it appears to me that viruses get less dangerous over time, not more dangerous. This appears to be because viruses that are overly dangerous kill off their hosts too quickly, limiting their ability to spread, whereas the milder forms don't kill their hosts as easily and spread more broadly, building immunity against the stronger variants over time.

Edit: The result of this, if you look at previous diseases that have plagued humanity (see Black Death), is a high initial death count when the virus breaks out, but the death rate declines as the milder variants spread and people survive them, building immunity to stronger versions, until eventually the virus mostly disappears.

It's not something you can count on; smallpox remained highly lethal across thousands of years. There's probably also a bit of an element of natural selection against humans involved; the people who died early of the Black Death may have been genetically more susceptible to it.

I am not a professional.

I have been basing a lot of my virus information from Tomas Pueyo. Here is my source:

https://unchartedterritories.tomaspueyo.com/p/the-omicron-qu...

In this article he clarified that generally viruses that spread more are less severe (as killing the host makes it tougher to spread). The asymptotic nature of Covid sort of broke this assumption because Covid can lie dormant in someone for a week before the person shows severe illness

>In this article he clarified that generally viruses that spread more are less severe (as killing the host makes it tougher to spread).

It takes far less than death to reduce spread. People who feel sick stay home from work (not guaranteed, but with enough frequency that it matters), don't go out to eat, don't wander around and go to parties, and so on. Any time they're sitting home, they've hurt the chances that their variant is the winning-est variant.

Again with the caveat about dormancy periods for Covid.

Asymptomatic spread is not nearly as prevalent as originally feared.

Also, those who do not get symptoms (that is, most who catch covid) do not spread the virus in general, which is why it's now called "presymptomatic" spread.

Yeah, technically a variant who lies dormant for a long time while spreading slowly is a better fit than one that is noticed.

Didn't we learn anything from Plague Inc?

It depends on what you call severity.

If severity means the host dies quickly rather than slowly, there's obviously selection pressure and that mutation will die off relatively quickly vs the rest, since it won't have as many chances to spread as the less severe strain. E.g. SARS/MERS.

If severity means the host dies more often, but it still takes a long time to die, the selection pressure is much lower, and it will roughly spread the same as the less severe strain, all other things being the same. E.g. Spanish Flu.

Spanish Flu did evolve to become more mild. All modern seasonal flu variants are descended from it.

> Spanish Flu did evolve to become more mild

That doesn't make sense. We have seen regular waves of flu, some being more pathogenic than others. There has not been a "direction" of flu strains becoming milder. So-called Spanish flu came after weaker strains.

> All modern seasonal flu variants are descended from it.

I don't think this makes sense either. The more serious strains of flu we've had in recent decades have for example included contributions from bird and pig viruses.

Certainly there is variability, evolution is fundamentally a random walk, but no strain since 1918 has come close to that one's severity, and the genes causing cytokine storms in infected individuals, which was why that strain was so deadly, seems to have been quickly lost. Subsequent strains that were highly virulent seem to have evolved new mechanisms.

Viruses can pick up genes from lateral gene transfer. Basically in the process of copying their own genetic material while replicating, they might inadvertently copy some extra genetic material that was just lying around in the host cell, and it gets incorporated into their genetic code, producing recombinant strains. Further, just as animal strains can jump into humans, human strains can jump into animals. So for example in 1968 the H3N2 flu lineage started in birds, jumped to humans where it recombined with 1918's H1N1 making human H3N2, and then jumped into pigs to make Porcine H3N2. Indeed there's evidence that H1N1 started out as an avian strain that jumped to humans and swine around the same time, but this is still an open academic debate.

Eventually. But as it appeared it was more virulent than its immediate ancestors.

Eventually in this case being approximately 18-24 months from the first wave, which coincidentally is very close to where we are in the current pandemic.

The Spanish Flu is a single data point. I wouldn't read too much into the 18-24 months stat.

The point isn't that Covid will be as fast as the Spanish Flu, it's that the Spanish Flu did not take a long time, and should not be dismissed as an example where it became milder eventually. In fact, we'd have to be quite lucky if covid became mild in the same time frame.

Viruses evolve to be more fit. Presumably, a mild version and more contagious version may even have evolved in one of the billions that were infected these years, but it didn't spread much because of self-isolation and restrictions. A variand that s more debilitating would be less likely to transmit.

"I am hoping that it would be some kind of trade off that the virus makes in that it can spread more easily but not be as severe..."

I've definitely seen outlets postulating that the endgame for COVID-19 would be it doing exactly this— becoming way more contagious but way less severe, basically a just a kind of cold. These "final" variants would quickly spread everywhere, choking off the supply of new hosts for the more deadly variants, effectively achieving herd immunity.

Any genetic mutation which causes premature death of the host, or fails to be communicative will generally die off. Any mutation which keeps the host alive (or at least longer), or makes it easier to transmit the virus to another host, will keep the virus around. This type of behavior can be muted if there is an animal host that acts as a reservoir.

> Any genetic mutation which causes premature death of the host, or fails to be communicative will generally die off. Any mutation which keeps the host alive (or at least longer), or makes it easier to transmit the virus to another host, will keep the virus around.

Not necessarily[1]:

> Some viruses provoke severe symptoms in their hosts that make it easier to transmit the virus to others. But those same symptoms can wind up killing the hosts.

> Adalja said one example is Ebola, a deadly virus that spreads through the blood and body fluids of infected people. Another example is norovirus, which causes diarrhea and vomiting, and leads to hundreds of deaths each year in the U.S.

> “The virus, speaking anthropomorphically, just wants to spread and have its genes replicated,” said Adalja. “If the best way for it is to spread by causing severe symptoms it will continue to do that.”

[1] https://apnews.com/article/fact-checking-011488089270

Unfortunately, mink and white-tailed deer (the most common deer in the eastern US) are Covid reservoirs now. Probably other mammals as well.

I can't find the excerpt, but I read "The Great Influenza: The Story of the Deadliest Pandemic in History" by John M. Barry during this pandemic. It was an okay read if you're looking for something to pass the time.

There was a passage that suggested that virus evolutions happens all the time which sometimes gives us a super severe strain. But usually these super sever strains don't evolve into even more severe strains - generally they evolve back towards the "baseline" severity and the severe strain burns itself out.

From that, I had hope that the Delta variant of COVID was one of these outlier super severe strains and that all subsequent evolutions would fall back but it's not looking like that is the case when omicron was found.

I’ve heard that it depends on the virus. Some become less severe as they become more contagious, but sometimes their severity stays the same.

Naively, you’d think selection pressure would decrease severity, because severity is not how viruses spread. However, if (some of) the same mechanisms that ar causing the virus to spread better are also responsible for (some of) its severity, then things may not really get better over time —unless our immune systems themselves adapt, but that generally requires a first infection…

Not an expert, but from hearing about it from various sources:

As viruses evolve they generally have 2 paths of evolution – they either become more deadly, or more contagious.

The ones that become more deadly will kill the host quicker, and so won't have enough time to spread to others.

The ones that become more contagious are by definition weaker, they will spread to more people, but the effects will be milder.

There are some theories to suggest that pandemics (like the Spanish flu) ended because the virus eventually mutated to a milder, more contagious form.

You are right about the trade-off: A more infectious, less deadly virus is the natural evolutionary path for them.

But about the breeding ground, not really: If one day some super deadly variant evolves, it's not popping up in all of the infected hosts at the same time. It'd still need to infect everyone from scratch, and since we all have our guard up right now, it'd definitely have an incredibly hard time doing so.

I once hoped a lesser severe but a more transmissible variant would win, but then I don't think HIV did or at least within my life time.

gjsman addressed the issue of severity very well, so I'll just comment on tradeoffs.

The adaptations that affect severity and those that affect transmissibility generally seem to be different. These are two axes on which a new variant can very in either way. The most successful new variants will tend to be highly transmissible (for obvious reasons) but also tend to be lower severity for the reasons gjsman gave in a sibling comment.

That's only a tendency though, evolution will throw up variants on all sorts of different points on the transmissibility and severity scales and may the most successful virus win. Yes high severity variants will tend to kill off their victims sooner, but if also highly infectious that might overcome that effect. In any case killing off victims still leaves a trail of devastation. It can take a while for all of this to play out.

Play with Plague Inc. (https://www.ndemiccreations.com/en/22-plague-inc)

Every single press title the last two years can be related somehow to a single Game. New variants, new mutations, spreading faster vs killing more people, ... All these tradeoffs are quite well modeled.

Building a resistance to vaccines ("the cure") is also one areas where your virus can evolve, and -to be honnest- I prefer it to fight vaccines than to develop a "total organ failure" mutation.

It's a game, not a model. Real mutations aren't directed or instantly spread to all infected people like in the game. Iirc people also don't get better once infected.