i want to post this, because i think this is indicative of the media completely whiffing on the top-level finding:
twitter's synopsis: "According to a study, which is not yet peer-reviewed, conducted by researchers at Imperial College London, antibody levels in the British population declined rapidly during the summer. The study of more than 365,000 people in England found that antibodies fell from 6% of the population around the end of June to just 4.4% in September,
suggesting that protection from the virus may not be long-lasting."
this is NOT what this finding implies.
1) antibodies are not the only way your body protects you from infection. once you get any infection, your t-and-b-cells that trigger an immune response retain a memory on how to fight it. a peer-reviewed study in nature that found that up to 80% of the study population had a t-cell response to covid (even if they hadn't been exposed) cited data that found that t-cell response in people that had SARS-1 were still just as strong 17 years later.
2) higher antibody levels are related to more severe cases of covid. if you have an asymptomatic or mild case, you're less likely to test positive for antibodies.
3) we have known for months that antibodies wane after 2-to-5 weeks. it's why any mass seroprevalence study likely severely undercounts the true level of infection. take, for example, this preprint of a study in japan:
https://t.co/XVoJfjEcrv?amp=1
it looks at a group of 615 adults in tokyo. when the study began in may, just 6% had antibodies for COVID. by mid-summer, almost half (47%) of that same group did. no hospitalizations or deaths.
i can also speak personally about this, as i know three transit workers in new york that tested positive for covid in march (mild symptoms), waiting three weeks, then got an antibody tested and were negative for antibodies.
when i talk about measuring antibodies, i tell people to think of it like trying to take a picture of a baseball at the exact moment it hits the bat. it's almost impossible. sometimes you'll capture a little after, sometimes a little before. but very, very rarely will you get the ball right at the point of contact. same thing with studies that measure antibodies. sometimes you'll see a larger portion of the true number of people exposed, sometimes the smaller number, but almost never the real number.
so antibody levels don't show the real number of the population that's been exposed, and waning antibodies doesn't mean you don't have protection from another infection. it also doesn't mean you can't get reinfected (though how many documented cases do we have of reinfection globally? 6?).