[timr]

> Is 12 to 15 weeks (3-4 months) really "long lasting" antibody mediated immunity?

It's perfectly normal for an antibody response to decay over time. We don't keep making antibodies in perpetuity for every antigen we've ever seen, or our blood would turn into peanut butter.

The B-cell and T-cell mediated responses to previous infection are also proving to be robust, and the B-cell response, in particular, primes the immune system for a rapid increase in antibody levels if the virus is encountered again.

[superkuh]

Yes. But we only get one influenza vaccine per year. The expectation has been for antibody neutralization coverage (not just T-cell mediated infected cell killing) lasting that long.

15 weeks is rather short. Additionally, it's not really 15 weeks. The study time scale starts at the 1st vaccine dose (week 1). Then it's 3 weeks until the 2nd dose and 2 more weeks till full protection. So the actual protective coverage span in this study is week 8 to week 15, or 7 weeks.

This study being held up as proof of long lasting immunity doesn't hold water. It is not bad news though. Decent B-cell responses (though tailing off) till 7 weeks after full vaccination is good. But we definitely need to know what happens after the first 2 months too. This exact study should be repeated again in 6 months (24 weeks).

[timr]

> 15 weeks is rather short. Additionally, it's not really 15 weeks.

You're right. It's the lower bound, based on the duration of the experiment.

There have now been many, many studies all pointing in the same direction. A small selection:

https://www.thelancet.com/journals/lancet/article/PIIS0140-6...

Natural infection protects against Covid (incl. Alpha variant). Adjusted incidence rate ratio for covid infection - 0.159 (84% protection) and symptomatic covid - 0.074 - (93% protection) - 7 month follow up

https://jamanetwork.com/journals/jamainternalmedicine/fullar...

Patients who had 1 PCR positive tests had Hazard ratio 0.06 of further positive PCR - 94% protection - 12 month follow up.

https://academic.oup.com/cid/advance-article/doi/10.1093/cid...

After 1 positive PCR, Protection against any COVID - 81.8%, symptomatic COVID - 84.5% - 6 month follow up - protection increased over time.

https://academic.oup.com/cid/advance-article/doi/10.1093/cid...

seropositive patients had 94% reduction in hazard of testing positive - 8 month follow up

https://www.nejm.org/doi/full/10.1056/NEJMoa2034545

Adjusted risk ratio of covid infection for antibody positive healthcare workers was 0.11 (89% less likely) - 7 month follow up period. All reinfections were asymptomatic.

At this point, claiming that immunity to Covid (even against "the variants") is less than a year is an extraordinary claim. There is no reason, prima facie to presume that the results will be different for vaccines.

[superkuh]

Right. The experiment was far too short to make any claims about long term B cell mediated antibody levels. Establishing 7 weeks post fully vaccinated being good is not a "lasting immunity".

I'm not sure I agree with just assuming protection will last that long as a default. This study does not establish it but it is a good stepping stone to proving that theory.

re: your edits, the tangential bit from this paper about sars-cov-2 positive cases wasn't the main thrust and from the graphs it looks like their neutralizing antibody levels actually are multiple logs below mRNA vaccinated people at the same 15 week point. I'm not sure what all these citations about naturally infected people add to the topic being debated.

[timr]

> re: your edits, the tangential bit from this paper about sars-cov-2 positive cases wasn't the main thrust and from the graphs it looks like their neutralizing antibody levels actually are multiple logs below mRNA vaccinated people at the same 15 week point. I'm not sure what all these citations about naturally infected people add to the topic being debated.

These are all real-world studies of actual duration of immunity. They trump any theoretical concerns based on lab measurements of neutralizing antibody levels.

Observing that the vaccines induce responses lasting as long as 15 weeks is completely unsurprising, based on what we already know from real-world observation of natural infection. Ignoring contextual evidence and concluding that this specific paper didn't provide reasonable evidence is losing the forest for the trees.

Could it be possible that the vaccines don't confer immunity lasting as long as what we're observing from natural infection? Sure, I guess.

Is it likely? No.