Is the length of the Covid-19 incubation period likely to be affected by whether you are vaccinated?

post by Josh Jacobson (joshjacobson) · 2021-06-12T03:27:53.840Z · LW · GW · 15 comments

This is a question post.

Covid-19 symptoms typically present 2-14 days after exposure in unvaccinated people, with an average presentation at 5 days post-exposure.

Is this period likely to be affected by having been vaccinated? Amongst those who catch Covid 2+ weeks after receiving an mRNA vaccine, will their average presentation be >5 days, and the range look more like 4-20 days, or something?

(I live with an immunocompromised individual who cannot be successfully vaccinated, so this is decision-relevant for me. Informed guesses are welcome.)


answer by Connor_Flexman · 2021-06-25T04:09:12.942Z · LW(p) · GW(p)

(Moderately cleaned version of the answer I PM'ed you)

My guess is that yes, you on average add a day or maybe 5-30% to your incubation period bookends. I would guess 2-19 days, avg 6, if the original is 2-14 avg 5. I'd guess about 40% chance it does hardly anything either way, 20% chance selection effects or other unexpected mechanism actually means

 incubation period is reduced.

(This is only different than JenniferRM's answer because she was looking at the case where your body immediately fights it off successfully, but I think that's counted as "not getting sick" and you're looking more for what happens in the 5-10% of the time where you don't immediately fight it off.)

If you condition on not fighting it off, I expect the default is that the partial antibody production will slow down the progression some, but not much. Some number of antibodies means your "initial" viral load will be significantly reduced, so it will take a bit longer to replicate up to symptom-level. (Or similarly, early growth rate will be reduced, also taking longer to hit a threshold.)

Why not much longer? First, just intuition—it would be pretty outlying if suddenly you routinely had 40-day incubations. Second, it's still exponential, and there are just lots more models that make sense where it doesn't multiply your incubation period for a constant time. I'll expand on these below.

Building models of incubation period means trying to determine when symptoms begin during a growing infection. It probably isn't strictly after some threshold of viral count, like 10^8 viruses—probably it partially depends on that, but mixed with other features like “viral generations” and “time”. For example, the longer the virus is in your body, the more chance your body has to notice it and mount a non-specific immune response. And if the virus is searching spatially over your body for a weak point to replicate faster or cause symptoms, the search time probably depends on the number of generations rather than just raw count, under some assumptions of distribution over travel distance at each generation. So symptom time is probably determined at least by count, generations, and time, and only the first of those gets substantially changed by the vaccine (so it won't linearly affect incubation period).

Another class of model might look at what factors affect growth rate at a given time or count. As the virus replicates, the antibody:virus ratio goes down (we've conditioned on getting sick, so it's not like there's a successful immediate scaling of antibodies). Presumably this means there's an "overwhelm effect" where your antibodies are meaningfully slowing things at the beginning, but as the virus replicates, it quickly asymptotes at its old default growth rate. This type of model, or any other model that doesn't assume constant viral growth the whole time, is going to mean the growth rate changes at different stages, meaning the incubation period won't just be linear in some starting parameters. (And nonlinearities lead to reduced change in this context—there are very few factors that could make superlinear change in incubation period given a reasonable parameter.)

Anyways, I'm not that confident it couldn't be longer than a 30% increase, but I would definitely bet against it. I'd guess 95% it's under that.

The remaining uncertainty I have about my claims above, as mentioned in the summary, is mostly due to the possibility of a strong selection effect in people who get the virus after vaccination. For example, maybe the people who had 14-day incubations were the very healthy ones who definitely won't get COVID after vaccination. Say the vaccine makes everyone "healthier" and moves them down the incubation-timeline, meaning someone who would have gotten symptoms in 2 days now takes 2.5 days and someone who would have gotten symptoms in 8 days now takes 10. But what if everyone who would have gotten symptoms in 9 days no longer can even catch COVID? In this case, the incubation bookends would have moved from 2-14 to 2.5-10, even though every specific person is taking longer. (Obviously there are no sharp cutoffs, this is merely illustrative.) Anyways, selection effects destroy many otherwise-good analyses, and we know they'll affect this, but the question is how much. For now I'm guessing they aren't bigger than the primary slowdown effect, but this is mostly intuition.

comment by Josh Jacobson (joshjacobson) · 2021-06-25T06:50:38.232Z · LW(p) · GW(p)

The cleaned version here resonates more greatly with me, answers some questions I think I hadn't realized I had, and I just really highly appreciate your response to this.

I'm making a meaningful quarantine decision prior to visiting someone immunocompromised, so it's really great to have your very different perspective from the other response to this.

answer by ryan_greenblatt · 2021-06-12T12:04:32.337Z · LW(p) · GW(p)

How old are you? Also, 2+ weeks after the second dose?

comment by Josh Jacobson (joshjacobson) · 2021-06-13T17:55:33.815Z · LW(p) · GW(p)

Mid 30s, and yes

Replies from: ryan_greenblatt
comment by ryan_greenblatt · 2021-06-15T11:46:25.079Z · LW(p) · GW(p)

After looking for a bit I wasn't able to find anything.


Comments sorted by top scores.

comment by JenniferRM · 2021-06-12T20:58:30.422Z · LW(p) · GW(p)

I'm just reading out my mechanistic priors here (mostly not answering about specifically covid, from specific data).

In the same way that masks made sense in February of 2020 my reaction was "like duh? the viral particles are bigger than the particles that well fitted N95 masks already keep out? how could it NOT be helpful? are people really just this dumb?"

The immune system is much more complicated than masks, and has exponential growth INSIDE the mechanism, so it is way more understandable that people wouldn't "get it"...

But applying the same kind of "there is a general system that generally works" reasoning here...


When you get virally infected WITHOUT a vaccine, the reason it takes "N days" to fight off a virus is because your immune system initially fights a LOSING war against the virus until an "antibody-reliant super weapon" can be constructed, kinda like a Manhattan Project, but more focused.

Once your immune system gets specifically able to fight the specific virus the fight gets much shorter and your immune system nearly always wins.

(Sometimes not: HIV is an exception, for example, because it evolves faster than your immune system can complete super-weapon projects to target the recent/past/obsolete version of your own infection in your own body. Herpes is a partial exception because SOMEHOW(?) it hides in a nook invisible to our entire immune system? Biology is rife with exceptional details.)

N days until disease goes away =

"time to design antibodies after exposure but before symptoms" +

"more time to design antibodies during symptoms while still losing the fight" +

"more time to USE the antibodies to win the fight at the end"

= "1-3 weeks for influenza (and mostly on the shorter end of that range)".

MOST of the time is spent on research.

The last term in the equation above is generally pretty small. It would be even smaller if it was not performing a naively astonishing come-from-behind victory maneuver against a vast horde of gazillions of infected cells and viral particles.

The "time to form antibodies" varies across people, but the reason for the "3 week delay" between two vaccines is because the vaccine designers wanted to systematically wait until like 99% of the population would have finished this response, even sickly people.  This way the first response and the second response would be biologically distinct exposures, but also the treatment with both would count be "the same for everyone and logically a unit in most people's minds" which has nice political and PR properties.


Some very healthy people could probably get away with a delay of possibly just a week?

If there existed many competent doctors with schedules that were wide open (or just an autodoc in every home) the autodoc could measure things directly and give you a second vaccine in the very hour that the immune response is good enough, and this might be only 5 days for some people? Reason: because some people get better from influenza pretty darn fast.

So suppose that someone got infected with the same strain of influenza right after they fought it off once already? Like they drink 4 ounces of "give you influenza fluid" for science...

Then they fast forward through most of the earlier immune machinery's operation. They ALREADY have super weapons against exactly that influenza strain already designed, and design is the slowest step.

So the first few cycles of the SECOND infection can be thought of like something from a medieval anime war story that is an amusingly fast win for the good guys :-)

There are 4 (the inoculating dose) "ninjas with mind control magic scrolls" (viruses) who sneak into the small feudal nation (you), and while they are bespelling the first 4 gate keepers or farmers or whatever according to a basic magical mind control doubling schedule... There is a dog nearby that already has the scent of exactly those ninjas and exactly their specific paper for their specific mind control scrolls (from the previous infection), so the dog smells the ninjas and then howls, and howls spread quickly until they are heard several miles away at the superweapon depot.

The howls brings in killer robots with special algorithms to pierce every hiding trick the ninjas might employ (same as the first dog used, basically) and ends the game almost instantly. The killer robots take out the "at most 64-256 ninjas" and the invasion is over. EVEN if the robots had had to face 1,073,741,824 (==2^30) ninjas the robots would have STILL won (because robots are awesome). That would have been a long fight.

However, because the dog knew the scent, and the robot got there, and the robot knew the scent as well, the response is super early, and super fast, and the fight doesn't even take that long, and all together it prevents most of the possible damage to the little kingdom.

So a lot potentially hinges on your own personal vitality, and exactly which strain infects you. If the "smell" of the invaders isn't the same as before, the event will be BOTH more damaging AND take longer to deal with.


If you are healthy and the strains are close to what the vaccine targeted, then your average infection period will be like... half a day? Maybe 2 days?

Also, you will almost never notice it, because the metaphorical robots will kill the ninjas before there are enough ninjas to cause much damage to the countryside and form a ninja army to attack the lungs, or the brain, or whatever.

(Part of why covid is so freaking horrible is that it is infectious BEFORE the symptoms show up because the lab that created it probably used serial passage on humanized lab animals to make sure that the fit between the spike protein and human ACE receptors is very efficient. From the very beginning, in December, in Wuhan, it transmitted human-to-human BEFORE the fever starts, which means that the temp checking system that defeated Version 1 of SARS wasn't going to be good enough for this new version of SARS. If not for this detail, it wouldn't even be on my radar as a thing to worry about that you could pass covid during these quickly suppressed microinfections that vaccinated immune systems handle automatically all the time.)

HOWEVER ALSO, clinical studies of "people who are vaccinated but still somehow got sick" will therefore, over time, come to be based on either people who are sickly or who have new strains, or both.

(For example, like 20% of the US might be classified as some-kind-of-alcoholic, because even just 1 drink of alcohol wipes out a large percent of your immune cells for hours or days, and then your hematopoietic cells rebuild them at the cost of telomere shorting. Such people often have relatively wimpy immune systems, especially late in life.)

So if you're healthy and not exposed to brand new strains, the time during which you might be kinda infectious MIGHT be roughly 1.5 days in (while awesome robots kill the kind of ninjas they were programmed for)? Maybe less than 1.5 days? Maybe hours?

(I have half-heartedly looked for evidence of vaccinated healthy people being "the only logically possible link in a transmission chain" and have not found much evidence of such things. My search was not thorough, however.)

BUT if you are exposed to newly antibody-escaping strains, or are just "not healthy in the first place", or both, then you could look at naive clinical studies (that might not be careful about admitting or explaining the straightforward theoretical implications of their sampling methods), to get a sense of how your own bad luck situation might play out.

If you are unlucky, then studies performed on other unlucky people could show a rough infection, amenable to therapeutic interventions, that occurs partly in a hospital where "data collection to power formal studies" is easier and cheaper to collect.


Long story short: assuming a healthy immune system, and no immune escape from a new virus variant, your infectious period is probably much shorter (or even essentially non-existent) after you are successfully vaccinated.

This might be wrong. These are just priors, and a general model. (HIV is an exception to this model that I already know about, for example.)

However, I think there are not many ways that (1) the model has an essentially broken structure with respect to the immune system in general or (2) covid has some additional specific wrinkle to its biology such that it routes around the normal immune responses.

If someone existed in a conversation with me, in a similar state of "working almost entirely from priors", and they put $10 at risk while I put $20 at risk, and they won my $20 if they found something on the internet that defeated my model or basic conclusions, but if they couldn't find anything definite within a couple hours I would win their $10... then I would accept the bet... and be happy to learn how I was wrong <3

On the open internet, "against all comers", bets work differently, and mean more. I'm not offering this bet here.

All I'm reporting here is my subjective "conversational bet tolerance" for "my priors, and that the priors are basically adequate to model this situation without extensively consulting with the internet first".


The US essentially doesn't have a public health system anymore.

The giving of free vaccines is what such a system would do if it existed, but for covid vaccines, this was done as an emergency one-off, because...

...of reasons... basically none of what I said about vaccinated immune systems applies to a hypothetical "very new strain" where a single one could arise in a place with a lot of infected people and then evade hypothetical border controls AND also evade current antibodies produced by the current vaccine.

To account for this risk you would have to have a positive predictive evolutionary model?

And I can't do that, and don't know of anyone who can. Maybe nothing will evolve? That's what I hope.

If hope isn't good enough, and you don't count political solutions you basically have to fall back to really generic solutions like "retreat from the world (maybe as a project)" or else maybe just "cryonics and crossed fingers"?

Or abandon the constraint of not "being allowed to modify the current political disaster directly" and then like... uh... gain power... and fix border controls relevant to public health (and public health in general) directly for real?

Or just hope that Obama cleans house somehow eventually via more backroom deals or something?

Or take extra advantage of the brief cessation in the covid horror, while vaxxed against the current strain... because there's no coherently structurally certain reason that the next wave won't have already started up again by next Christmas?

If I couldn't get a vaccine, and was (justifiably) relying on the herd to protect me from plagues, I would be SUPER angry at the herd for not doing this properly.

Your housemate has my sympathy.

Replies from: joshjacobson, GuySrinivasan, BossSleepy
comment by Josh Jacobson (joshjacobson) · 2021-06-13T18:45:39.115Z · LW(p) · GW(p)

I really appreciate the detailed explanation here and expression of the level of confidence in your belief. Thanks so much for this!

Replies from: JenniferRM
comment by JenniferRM · 2021-06-14T16:04:16.686Z · LW(p) · GW(p)

I appreciate the feedback! I sometimes wonder if walls of text are unhelpful/embarrassing somehow, but if the intended audience likes it then it can't be that bad, right? <3

Replies from: BossSleepy
comment by Randomized, Controlled (BossSleepy) · 2021-06-14T20:20:43.552Z · LW(p) · GW(p)

No, this was great, I totally dug it and I'm just a random bystander.

comment by GuySrinivasan · 2021-06-17T18:45:09.419Z · LW(p) · GW(p)

I also live with an immunocompromised individual who cannot be successfully vaccinated. After research including reasoning very similar to yours, we concluded that if she wore a mask, we felt safe enough with vaccinated folks who had not had any obvious infection opportunities within the past 12 hours to be indoors for substantial periods of time with known-vaccinated folks not wearing a mask. This tracks almost exactly with your guess of "MIGHT be roughly 1.5 days[...] Maybe hours?"

comment by Randomized, Controlled (BossSleepy) · 2021-06-14T20:27:51.594Z · LW(p) · GW(p)

because even just 1 drink of alcohol wipes out a large percent of your immune cells for hours or days, and then your hematopoietic cells rebuild them at the cost of telomere shorting

Say waaaaaaaaaaaaaaaaaaa?

It's a bit unclear from your context: are you asserting this for everyone or primarily for people who are abusing alcohol?

I do have to say, I've always been pretty confused by the inconclusive evidence that drinking may have some positive health effects. Maybe I just tolerate alcohol less than average, but I've often felt like I was just dosing myself with poison based on one or two drinks and how bleeeech I would generally feel the next morning. There have been a number of times when I've felt like this might be a bad hedonic cost/benefit trade off, but there are a number of beers whose taste I like enough that I haven't really been tempted to just set a blanket no-booze policy.

In your judgement, is a blanket no-booze policy worth considering? (I'm a pretty infrequent drinker currently: a heavy drinking week might see three or four units; most weeks are zero units.)

Replies from: JenniferRM
comment by JenniferRM · 2021-06-17T03:05:06.451Z · LW(p) · GW(p)

I used to drink literally none. I'm genetically predisposed to bad outcomes from alcoholism (my family has jokes about "brewery genes") and teetotaling through my teen years is something I do NOT regret.

But statistics suggest that literally drinking absolutely no alcohol ever... isn't correlated with maxxing your longevity (though: insert much debate here)?  SO the current best theory I have for a possible mechanism (assuming it is even true) involves using short term brain damage as part of human bonding rituals.

More/better friends makes you happier, and hence longer lived. Seems legit. Then I figure: a few drinks a few times a year won't matter that much? So I do drink sometimes in some social contexts, but I try to do it rarely and lightly. Enough to show that "in vino veritas" (and so on) reveals me to be a happy drunk instead of a mean drunk, but not much more than that. ONE glass of champagne to celebrate new years? Who am I to argue with tradition? <3

My memory (which is human and thus fallible) contains a study or two that were beautiful, with healthy mostly non-drinking people people who were clean for some time, then had before and after white blood cell counts, with a randomly assigned number of drinks, and they got a nice little dose response curve and time-to-recovery out of it. 

My memory wants to say that maybe one drink depressed WBC counts by like 35%, with recovery within 1-3 days? Maybe the second drink got the number down to like 60% from peak? And by the sixth drink you mostly already blasted all the white blood cells that were going to die with the fifth drink already? That last 10%, or 5%, or 1% of white blood cells are hard to wipe out, and "100%" only has 100 percentage points inside of it to lose, which... is a finite number of points... so... yeah.

If literally ALL of them died with nothing anywhere to grow back to the right numbers... that seems like it would make alcohol way more fatal than it observably is? 

Like you might need a blood transfusion to live after every binge drinking event in that case? Which people don't need. So that can't be how it works.

But I can't find the study (studies?) now so either the modern (crappy) google on the modern (crappy) net corpus is failing me, or my memory is bad? 

Instead, all I can find right now is long term stuff from prestige farms (and second order similar content on long term processes) empirically showing a long term hand-wavy version of chronic abuse causing deep structural damage... 

...but like: "duh"? And THESE don't seem to mention telomeres or mechanistic connections between obvious short term and long term biological processes?

I don't know. 

Biology is a science, but it is a weird science. It is basically an attempt to reverse engineer a gajillion weird little gadgets designed by an insane god. The wrinkles have wrinkles, seemingly ad infinitum, but surely it has to bottom out SOMEWHERE and then make perfect sense eventually?


In terms of the VERY BIG picture, it is mechanistically plausible that it is useful to do a bit of a purge now and again? Exercise. Fasting. Variation. Hormetic stresses. Maybe "use it or lose it"?

Like my best guess for green tea's mechanism of action on cancer reduction is that it mildly upregulates apoptosis and makes all cells just that much extra willing to suspect they might be cancerous and press their own cellular suicide button. Seems scary to me (I stopped drinking green tea regularly when I found out), but green tea has a positive reputation, so who am I to say?

(Doctors used to purge blood all all the time, and it was never helpful, except when it was.)

Then there's asparagus: it could be neurotoxic (at some doses) for a chunk of people... I dunno if I have the allele for that result, but even if I have the allele, I still do eat it sometimes, but only like one or two spears, young/weak/small, and only maybe once a year. 

I hated asparagus as a kid, and kids tend to be smart about such things? 

I heard from someone once that there are two species that can eat asparagus without dying: some bug that specializes in it, and humans, because humans have freakishly capable livers. This was uncited however, and maybe just word-of-mouth bullshit?

There's an old wives tail that asparagus works as decent wart medicine, this jives with "anecdata" for me... but like... IF something can kill some small benign(?) cancers then it deserves serious respect in my book. Not "never". Just "cautiously"?

Alcohol seems like it could hypothetically fit in here somewhere, maybe it has a use as a poison that usefully poisons the bad cells worse than it poisons the healthy cells if used rarely?

I think the standard use for alcohol in the super olden days as a water additive that kills microbiological infestations (not that they knew that it was doing this, because they didn't have a germ theory of disease, so... what the hell?) but it doesn't make sense to use it this way anymore, I think.  

Nowadays the drinking supply is dosed with chlorine instead, so thank goodness for science!

The thing you were supposed to do was "water" the wine I think... which means drinking more water, which is theoretically healthy (but in practice probably not), so maybe many modern humans (but not all) have genomes tuned by the last few millennia of agricultural evolution to tolerate drinking a half a glass of a fructose-like beverage per day, mixed with quite a bit of water?

Purging cells every day, when we KNOW that telomeres are a thing, doesn't seem wise to me... but I know of zero studies aimed at optimizing any of this in any kind of sane way.

Basically, my real theory is basically to follow my cravings, try to AVOID eating unless I CRAVE some food or drink (and eat half of how much I think I want, then get seconds after 10 minutes if I still wanna), use common sense (liberally sprinkled with evolutionary bullshit), copy the diets of people similar to me who seem successful, and cross my fingers.

Replies from: BossSleepy
comment by Randomized, Controlled (BossSleepy) · 2021-06-17T18:29:29.213Z · LW(p) · GW(p)

Thank you for the info dump.

Basically, my real theory is basically to follow my cravings, try to AVOID eating unless I CRAVE some food or drink (and eat half of how much I think I want, then get seconds after 10 minutes if I still wanna), use common sense (liberally sprinkled with evolutionary bullshit), copy the diets of people similar to me who seem successful, and cross my fingers.

This closely accords with my intuition, the only part I haven't been doing is the 10 min cool off for seconds. I'll add that.

Thank you. : )

comment by ejacob · 2021-06-14T16:22:50.029Z · LW(p) · GW(p)

Is your immunocompromised buddy amenable to prophylaxis? I don't even really know how many pro-ivermectin-for-covid doctors think ivermectin prophylaxis works but it might be worth looking into since it is very cheap and relatively easy to tolerate. There are probably other options that i know even less about. Consult a trusted physician :)

Replies from: joshjacobson
comment by Josh Jacobson (joshjacobson) · 2021-06-14T19:47:25.013Z · LW(p) · GW(p)

My guess is that this isn't going to be worthwhile to look into in this case (facing a complicated, deadly disease with many, many medicines being taken for it, and a fragile health status) but I appreciate the suggestion.