[Gwypaas]

So 9% of seropositive and 10% of seronegative reported at least one symptom beyond 4 weeks. I.e. the background level is around 10% and the seronegative had "more" long covid than the ones who actually had covid in that group. Although this is pure randomness with such close measurements.

[PragmaticPulp]

4% versus 2% in the longer term study.

> Although this is pure randomness with such close measurements.

That's not how this works at scale. We need larger studies, yes, but you can't simply dismiss a doubling of long-term symptom reports.

At population scale, even a 1-2% incidence means millions of people.

[unishark]

It is pretty weak though. There were only 109 seropositive kids in the follow-up according to the table, so you're talking about 2 more kids reporting a cough or headache as the entire basis of your concern.

[anoncake]

There is no doubling here – only random noise.

[TheOtherHobbes]

It's also a preprint and other studies contradict it.