"not everyone seroconverts"

Really? Or do you mean not every tests positive for seroconversion months after infection. I mean, I suppose there might be some folks that are so severely immunocompromised that they cannot mount an antibody response, but surely the vast majority of infected do mount some antibody response.

There is a large population believed to have been exposed that recovered quickly and without a detectable antibody response. If you believe this paper, this may be because of an extremely effective cell-mediated response that prevented the release of enough antigen to trigger an antibody response:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8732273/

I have no doubt that such people exist, but they can't be that common, right? And to the extent they do exist, wouldn't their symptoms have generally been extremely mild? So you would expect most such individuals to not believe they've had COVID and to not have long-COVID, and therefore not to confound the results of this study.

In addition, it’s the immune response that likely leads to Covid organ damage so if an infection is interrupted so early as to not initiate an antibody response you’d expect any long covid effects to non-existent.

8.5 billion times a small percentage is still a pretty big number.

I'm not understanding the relevance. We are talking about the proportions of folks that fall into various buckets. The absolute values don't enter into this conversation.

I know quite a few people who believe that you can have long COVID without any initial symptoms, so it seems reasonable that there may be some people who got mild symptoms, believed it was COVID, and then have some ongoing symptoms that they attribute to that.

That is a very long and detailed paper, so apologies if I missed it, but I don't see any evidence of a "large population" of folks that clear the infection quickly enough to avoid triggering an antibody response. All I see from that paper, is that such people exist.

If you’re willing to take extended data table 1 at face value [0], then perhaps 1/2 to 1/3 of the exposed cohort is in this category. This is at least moderately consistent with other data suggesting that at least around 2/3 of the naive population will contract COVID and test positive on PCR is exposed.

In other words, the authors found a population that apparently did clear the infection this quickly, and they didn’t scour a particular large cohort to find these people.

[0] which you shouldn’t.

> do you mean not every tests positive for seroconversion months after infection

This wasn't your point but I'm trying to see whether this is accounted for in the study. I don't think it is, unless I missed something.

The study's self-reported long COVID group is people who in the past 4 weeks experienced any symptoms that had been ongoing for 8+ weeks. If these infections were 8+ weeks ago, these people would probably be seronegative anyway. To put it another way, seronegativity now doesn't rule out an infection 2+ months ago, so for some of these people their belief that they had COVID may match reality more than this article implies.

I don't know if there is a way to reliably identify a 2+ month old infection but I think you would need that to be able to say anything definitive about the linkage.