Honestly I am wondering if this is sampling bias. Have we ever paid this much concentrated attention to a disease? When you consider what a virus does to the body, it seems likely that many other viruses that have been endemic for millennia have been wreaking pernicious havoc on our bodies and we just... shrugged it off.
Anyplace the body heals from a significant physical injury, there is a scar. It's never a perfect restoration. It seems almost intuitive that the damage from a virus would be similar, and there is documented evidence that COVID will attack neurons.
That's not really what I said -- I'm not actually taking a stance on the conclusions of this paper. COVID has obviously been infecting massive numbers of people in the last couple of years, and if it is responsible for long-term cardiovascular sequelae as the paper concludes, of course it would be a more acute contributor over that period as compared to other historical causes simply by virtue of its pervasiveness.
Rather what I said is that the disease itself may not be causing "havoc" in the body at rates significantly different than other endemic viruses ("havoc" writ large: not only confined to cardiovascular sequelae, but among any long-term systemic harms). The bias that I am proposing is that we are _noticing_ these sequelae because A) so many people are getting sick at the same time, and B) because we are paying this much attention for the first time ever.
A) they are not pandemic, so the effect is of course much smaller
B) their impact is not necessarily primarily cardiovascular
Here's a non-exhaustive list of viruses with potentially overlooked sequelae.
Shingles (herpes zoster)
Rubella
Rotavirus
RSV
Norovirus
HPV
Hepatitis
Influenza
This is totally orthogonal to any possible defects in this study, which again I am not taking any stance on. It's merely a question of whether this study is anomalous for having been conducted in the first place. If we assume that the study is flawless, that means that COVID has long term health impacts -- fine -- but do we actually know if that's unusual for a virus?
There's a clear difference in health between the cohorts.
Supplementary table 1 breaks down the cohorts and summarizes co-variates. Disturbingly, the COVID cohort is significantly fatter than the contemporary control group:
https://static-content.springer.com/esm/art%3A10.1038%2Fs415...
54% of the COVID cohort has BMI > 30, vs 42% of the contemporary controls, and 40% of the historical controls. They also have a higher percentage in long-term care (4% vs. 0.6%), a higher percentage with cancer (8% vs. 6%), chronic kidney disease (19.5% vs 16.7%), lung disease (15% vs. 11%), dementia (6% vs. 3%), diabetes (32% vs 22%) and hyperlipidemia (34% vs 25%).
The Covid cohort is basically a much sicker population, in almost every way. Re-weighting the populations doesn't correct for this disparity.
The authors have attempted to hide this by burying the data in the supplementary material, and presenting an "absolute standardized difference" that makes it look smaller than it is.
What would be your preferred technique of accounting for the differences of characteristics in those groups?
This paper is discussing small differences in small numbers, and the single biggest difference between groups is the baseline level of health for the study participants. About the only thing you can say from this is that old, sick people are sicker after Covid-19 severe enough to put them in the hospital.