There’s a troubling view I’ve run into, which I’ve decided to christen the Faustian Vaccine Hypothesis. It claims that, while they definitely help short-term, the protective effect of vaccines targeting Covid’s spike protein will ultimately become negative.

I’ve looked at some of the data supporting this claim – in particular, the strong version, which claims a significant number of people currently experience net negative vaccine protection – and been largely unimpressed. However, proponents’ arguments are mostly not statistical reasoning about the present, but biological reasoning about the future. And it turns out I know exactly enough about biology to be convinced a Faustian effect is possible, while being unable to evaluate how (hopefully, im)plausible it is. This worries me, and I don’t like being worried.

I haven’t found any relevant studies or good counterarguments online – most scientists focus on the present instead of the future, and most mainstream debunkers target more common and straightforward heresies – so I’m asking LessWrong for help. And in recognition of the fact that evaluating vague prophecies is much harder and more tedious than making them, I’m donating $2k to charity: if you give me an answer or source I like, you get to choose which charity. (And if multiple people give answers/sources I like, I’ll allocate in proportion to how much I like them.)

The argument(s) for, summarized as best I understand them*:

  • Covid vaccines work. (uncontroversial)
  • Covid vaccines reduce in effectiveness over time. (also uncontroversial)
  • There is no theoretical reason why vaccine effectiveness can’t be negative. (c.f. the infamous Dengue vaccine case)
  • There is a mechanism by which positive initial effect and negative long-term effect can coexist: vaccination gives you an abundance of useful but temporary antibodies, while also teaching your body to have a specific and suboptimal immune response, which subsequent infections will have a harder time training it out of. (Original Antigenic Sin, a real concept with a wikipedia page and everything)
  • Obvious counterargument: other vaccines didn’t have this problem, why expect it here? Countercounterargument: other vaccines mimic a real infection by providing an entire specimen to the immune system and allowing it to create a variety of responses; spike-targeting vaccines produce a single, uniform form of resistance which the pathogen can more easily evolve to evade or – in the worst case – take advantage of. (Antibody Dependent Enhancement, which is also definitely a real thing**)

(Also, bird’s-eye-view caveman argument: vaccines used to protect against transmission and infection and worst-case outcomes, and then they protected against infection and worst-case outcomes, and now they mostly just protect against worst-case outcomes. And while that’s more a fact about variants than about vaccines, there aren’t going to stop being new variants, so extrapolating that trend . . . what happens to effectiveness of our current vaccines by the time we reach the end of the Greek alphabet?)

Important details:

  • I reiterate: proponents are not saying covid vaccines don’t work. They are saying strains have/will emerge(d) that did/will cause covid vaccines to reach negative effectiveness at some point significantly after vaccination.
  • Proponents aren’t against vaccines in general: opinions on non-mRNA non-spike-protein-targeting vaccines range from “overrated but still a good idea” to “mankind’s highest achievement, western nations need more of these for Covid-19 immediately”.
  • Even if they were right, that wouldn’t necessarily mean you shouldn’t get vaccinated. The longer you spend without catching Covid-19, the better specific medical tech will be when you do (also, projecting the trend from the original strain to Omicron forwards, the dominant strain will probably be milder a year from now). However, if the rationale among smart people was always “yeah that could happen but even if it did it’d be worth it to get to the era of Paxlovid with functioning lungs” . . . I’d kind of like to know that?
  • ETA: ADE is the worst-case scenario, not the main concern. The scenario that worries me most is unprecedentedly homogenous immune responses worldwide incentivizing unprecedentedly rapid antigenic drift, until patients end up with immune responses (enshrined by antigenic sin) less effective than what their bodies would have come up with by themselves.

There’s no strict time limit, but I have my next Pfizer dose scheduled for the start of February, and I’d like this to be resolved - in my own mind, at least - before then, so I can take the shot with confidence (or, possibly, cancel my appointment and then find some way to source Sputnik or Sinovac).

*I’d link to the proponents’ own words, but all the ones I’ve found communicate through longwinded blogposts that drag in arguments I consider irrelevant and expect you to have read all their other longwinded blogposts. (Is this how LW looks to outsiders?)

**The Wikipedia page for ADE has a section linking to “COVID-19_vaccine_misinformation_and_hesitancy#Antibody-dependent_enhancement”, which reassures skeptical readers that while this may be happening with strains from Delta onwards, it definitely didn’t happen with the original strain. I wasn’t aware that words could be so uncalming.


New Answer
Ask Related Question
New Comment

5 Answers

I'd point out that vaccine designers are extremely aware of ADE, and construct vaccines to be resistant to it.  It's just something you wouldn't talk about, because it's second nature.

For a good discussion of what this looks like, there's a section on ADE in the RADVAC whitepaper; the section to look for is titled:

"possible mechanisms of vaccine-enhanced disease, vaccine-induced autoimmunity, and mitigation strategies"

Not only are designers aware of the risk, there are actually ways to mitigate/reduce it.

The counter argument is that we could never be completely sure; but that's as true about the covid vaccines as it is any other vaccine designed after ADE was understood.

The people at RaDVaC took measures to hit other targets then the spike protein for reasons to make the RaDVaC more effective with future variants. The officially approved one that just use the spike protein do not follow the ideas that RaDVaC advocates here. 

Which is why the Wikipedia article says that ADE is definitely not a problem for the initial strain. All the vaccine trials looked for it and did not find it 

...which would make sense, in the model where this is a long-term effect, no? Everyone's password being changed toSHA256(username|"foo")in the short term is (likely) good for security. It's only once someone picks up on the pattern and starts exploiting it that it becomes a negative.

A source on this has been Geert Vanden Bossche (see his FAQ).

I'm not 100% convinced he's right, but I have not found any credible attempt at debunking him either. (One such attempt is from Gorski, but it's almost name calling).

He may be right, in that case you might not be able to find any convincing counterargument either.

Long term we could assess the claim that "a slow rollout of a leaky vaccine" actually applies in the context of Covid, but we know it can be theoretically (link).

Short term, for your next dose in February, I'd weight different factors:

  1. There's going to be a limit on the number of doses that are effective.
    What's currently going in Israel suggests that it's probably <4.
  2. It seems uncontroversial that the effectiveness wanes with time, peaking few weeks after the dose, declining after few months.
  3. The risks of additional doses seem to compound (source)

I'm assuming this is your booster shot, so it looks you have one extra shot in your gun and you might want to time it for maximum effectiveness.

If you're under 40, healthy and male I would wait even at the current prevalence in most countries.

If you're older, or know to be particularly at risk, you might want to optimize the time of your next dose.

If you take it in February you may be immune for ~3 months. Feb-May don't seem as risky as Nov-Jan in most of the northern hemisphere. Maybe you want to wait until October? We'll also know more by then.

If you know to be at imminent risk (for example known comorbidities) then you know what Covid does in those cases and the tradeoff is less obvious, probably you want to go ahead right away.

I spent a bit of time looking at Geert Vanden Bossche's ideas about six months ago.  I came away extremely unimpressed; he takes reasonable sounding things ("viruses mutate under selective pressure") and tries to extrapolate from them things that we don't actually see in reality.  He describes a plausible sounding reality, that is not our reality.

I agree he paints a bad picture but he's short on actual, time-bounded, predictions to evaluate his claims. He shared some predictions [] in May, with a time frame of months/weeks to see some vaccine resistant variant. I think Omicron counts as variant that is vaccine resistant, even though there's no peak in vaccinated deaths rates (deaths may be, but not rates as far as I can tell). Some other people claim [] Omicron does not descend from the Wuhan strain, so even this might not be the variant Geert Vanden Bossche predicted. Were you referring to some other prediction in particular?

I came to a similar conclusion to you. There is some theoretical risk of antigenic sin, so getting a vaccine could reduce resistance to some future virus. I also remembered that it could theoretically increase resistance to a future virus, as in the cowpox/smallpox case, or the SARS/MERS/COVID-19 case.

My estimate was that the positive effect should outweigh the negative effect, because I learned about smallpox at school, and I learned about antigenic sin on Wikipedia, so the school thing is probably a bigger deal. Unclear by how much. Given everything else, that was sufficient for me to get vaccinated and boosted.

Also, catching a virus can cause antigenic sin, it's not specific to vaccines. Unlike a vaccine, covid is not designed to reduce that risk. So if your infection risk without the next dose is >X then you reduce risk by getting the next dose. Where X is maybe 10% but I don't actually know.

I also endorse "even if it did it’d be worth it" as a counterargument. If arguments are soldiers then this attacks the hypothesis where it is weakest: the "so what" weakness. You don't seem to like this, but I do.

Another counterargument that I predict you won't like: you're specifically asking for counterarguments to a "troubling" and "Faustian" hypothesis. Maybe you have some confirmation bias, and you've already made your decision? If so, the evidence you've already seen was good enough for you to decide, so it can also be good enough for you to accept.

The scenario that worries me most is unprecedentedly homogenous immune responses worldwide.

Between multiple strains and multiple vaccines and varying vaccination rates and breakthrough infections I think there's lots of variation worldwide. I'd expect to see more homogenous immune responses to Measles, say, where there's a high vaccination rate, a low infection rate, and fewer vaccines.

I think the question of whether there is a Faustian effect is the wrong one to ask, and it may in fact be a substitution for the real question (as Kahneman would say).

Your actual question is "should I get the Pfizer booster in the first half of February?" 

I would answer that question (and in fact have answered it for myself) by asking questions like "what do I observe happening to people I know personally who already got boosted?" and "what do I observe happening to people I know personally who did not get boosted?"

If you subdivide the people you know personally category into "people I know personally who do not live in my immediate community" and "people I know personally who live in my immediate community," it also becomes a useful way to distinguish the COVID prevalence/risk in your community from the COVID prevalence/risk in other cities and countries.

Then you can use that information to ask yourself "what do I believe will happen to me if I get boosted?" and "what do I believe will happen to me if I do not get boosted?"

You can also ask "what do I believe will happen to me if I do not get boosted with this particular formulation, and choose instead to wait for a newer model?"

If you consider yourself to be at a lower risk for serious COVID complications but have loved ones within your immediate circle who are at higher risk, you should also ask "what do I observe happening to people I know in terms of spreading COVID to others?" and "what do I observe happening to people in my high-risk loved one's demographic who catch Omicron?"

Those are answerable questions, as opposed to evaluating a potential Faustian Hypothesis against an ever-growing series of secondhand, confounding data.

Good thoughts, but in this context I'm more worried about the future than the present, and I'm too introverted to have a statistically significant sample of people I know personally.

We all live in Dunbar sized bubbles. Very few people have more than 150 people for which they could know medical histories enough to answer these questions.

I'd caution you against spending too much time diving down infinite crank rabbit holes: true believers will always find some new detail or theory for you to rebut. At some point, if someone is committed to denying the clear scientific consensus, there's no point trying to get through to them.

At a high level, we have a pretty deep understanding of how covid vaccines work and how they perform over time, and there's absolutely nothing in there to suggest that, unlike every other vaccine ever, covid vaccines display the bizarre transition from positive protection to negative protection that you're asking about.

Vaccine effectiveness declines over time because (in large part) antibody levels wane over time. That's very well understood and in no way unique to covid vaccines.

Protection has shifted from protection against infection to infection against severe outcomes because of antigenic drift: the vaccines are most closely targeted to the ancestral strain. That match is most important for antibody protection: since antibodies are critical to protection against infection, the vaccines produce significantly less protection against infection as the virus drifts further from the ancestral type. T cell immunity is less affected by antigenic drift, so their protection against severe disease isn't as attenuated.

ADE is a real thing, and it was a concern early on. In particular, there was a feline coronavirus vaccine some years ago that triggered ADE, so there was concern that covid might have similar issues. But we've seen no sign of that.

Original antigenic sin is also a real thing, but wouldn't produce the effect you're asking about.

Two pieces of advice:

  • Get your next vaccine. It's incredibly safe and incredibly effective.
  • Spend less time engaging with lunatics on the internet. Too much time listening to cranks is bad for your epistemic health.

abstractapplic did their best in communicating fairly and honestly their epistemic state - considering the fact that Covid vaccination has become a highly politically charged topic they did a commendable thing by doing their utmost to be transparent about their motives and their uncertainty.

yes usually blindly following mainstream advice is better than following cranks and sometimes one can waste a lot of time debunking cranks - but I hope you'll agree that mainstream advice is not always the best. I do agree about being aware of the danger infinite crank holes. At the end of the day it is abstractapplic's choice to do so. 

 This topic has become unfortunately very political - we need more, not less dispassionated analysis.   True understanding, both personally & for society at large, comes from honest questioning and rigorous scholarship. There is no need to scold anybody for honestly asking questions.

 That is the spirit of Less Wrong. 

I personally learned a bunch of things from this post. I look forward to more knowledgable people chiming in.

Thank you for the reminder to explain and not scold—I shall strive to do so.

4Self-Embedded Agent4mo
No. Thank you for exemplifying Rationalist virtue! []

unlike every other vaccine ever

mRNA vaccines are a new invention, so that line of reasoning isn't particularly reassuring.

Protection has shifted from protection against infection to infection against severe outcomes because of antigenic drift: the vaccines are most closely targeted to the ancestral strain. That match is most important for antibody protection: since antibodies are critical to protection against infection, the vaccines produce significantly less protection against infection as the virus drifts further from the ancestral type. T cell immunity

... (read more)
4 comments, sorted by Click to highlight new comments since: Today at 3:46 PM

A related question is how many boosters of the same vaccine formulation will be required in the future and what is the safety profile of regularly repeated mRNA boosters, because the second Faustian hypothesis is that these vaccines are doing subtle but permanent damage to organs (e.g. heart or reproductive system or brain) which will accumulate until the damage is apparent and common. The appearance of increased myocarditis/pericarditis and menstrual disorders for which no mechanism is proposed is the smoke hiding a much bigger fire. 

The appearance of increased myocarditis/pericarditis and menstrual disorders for which no mechanism is proposed is the smoke hiding a much bigger fire. 

Covid itself is a perfectly fine mechanism to explain that. 

I know it's technically on-topic, but: downvoted. I have a specific question I want answered one way or another, and I think bringing up peripheral issues is unlikely to help me achieve that.

Have no information or opinion on the main issue. However in regards to the question about the community, As someone who got into the community late and alone I will tell you that yes, LW is exactly like that.