[mlyle]

I think a reasonable somewhat pessimistic (but not worst-case) assumption is that immunity is similar to common-cold coronaviruses; lifelong partial protection, total protection for 24-36 months, some degree of cross-strain reactivity.

Kissler et al published a really good analysis of kinematics and transmission dynamics in Science based on what we know about human coronaviruses, cross-immunity, etc. (Now that we are beginning to believe infection rates are even higher in comparison to case counts than we believed before, it looks pessimistic in various ways).

The research I've read on SARS-COV-1 shows a slower fall in antibody titers than other human coronaviruses. Though, unfortunately, I'm unaware of any study that followed patients past 3 years.

I believe that if you took one of the existing common cold coronaviruses, and introduced it to an immunologically naive population, you'd get a huge incident wave and a whole lot of excess death.

[bsder]

> I believe that if you took one of the existing common cold coronaviruses, and introduced it to an immunologically naive population, you'd get a huge incident wave and a whole lot of excess death.

Doubtful or you'd see waves of death in children from cold viruses. Common cold viruses have had lots of time to evolve so that they transmit well but don't kill very much.

Presumably it's also why our immune systems don't waste time building up permanent immunity against them. There's much more of an evolutionary pressure to be permanently immune to something that can kill or maim you, if you survive.

[mlyle]

> Doubtful or you'd see waves of death in children from cold viruses.

We don't see waves of death in children from COVID-19. The juvenile immune system is different. Children right now have the opportunity to pick up SARS-COV-2 antibodies without a whole lot of personal risk. Elderly adults, not so much.

> Doubtful or you'd see waves of death in children from cold viruses. Common cold viruses have had lots of time to evolve so that they transmit well but don't kill very much.

Everyone assumes that viruses evolve towards lower virulence, but this is only one direction that things can be pressured to evolve. COVID-19 manages to have a very high R0 by a high latent period. Producing a higher viral load enhances spread / R0 but also causes eventual severe illness.

COVID-19's fatality rate is only a disadvantage inasmuch as it causes population-wide behavior changes. It doesn't make people get excessively sick and stay home in a way that they spread the disease less, so we can't really be sure that it will evolve towards lower pathogenicity/virulence.

> There's much more of an evolutionary pressure to be permanently immune to something that can kill or maim you, if you survive.

We have plenty of things we don't build permanent immunity to that kill or maim us-- including SARS-COV-2, MERS, malaria, etc. I doubt this is mankind's first encounter with a really nasty coronavirus.