are covid vaccines associated with a rise in all cause mortality?

a look at some scandinavian data

• cliff notes: well, they are, unless they aren’t. the picture here is complex and the seeming positive association in scandanavia is prevalent, but not universal.

  this feels like something that needs more work.

often lost in the debates about “vaccine efficacy” is the fact that even if some resistance to death by covid is conferred, this is only part of the picture and that because, as certain internet felines are apt to repeat ad nauseum, “medicine is everywhere and always a cost benefit equation,” it is perhaps instructive to look at all cause mortality as opposed to just one narrow category. after all, getting hit by a buick and dying can look like “100% covid death prevention” if you measure just the one outcome. that would not seem to recommend it as a therapy.

there have been a lot of associations between high vaxx and booster rates and more deaths and the excess mortality data remains significantly elevated in most of the vaxxed world. but is there anything causal here? (and if so, in which direction?) it’s a tricky riddle to unwrap (though it would be trivial if any of these single payer health systems with good data would PLEASE start reporting age stratified all cause deaths data by vaccination status and type.) but we do the best we can with what we have and so let’s see what we can back into.

i chose the “big 4” scandinavian countries because they have good data, did not lock down much or for long, and all had high vaxx rates (70%+). i used excess mortality data from the human mortality database (and their graphing tool) and in all cases used the avg of 2017-19 as a mortality baseline. vaxx data is from OWiD.

this got sparked for me by looking at norway because there was quite a provocative 2021 signal there. (in all these charts, which can be graphically confusing until you get used to them, blue is low deaths and orange is excess)

as can be readily seen, moving from unvaxxed to vaxxed coincided with a big jump in the previously low norwegian all cause mortality (ACM). it flipped quite suddenly from below average to meaningful excess.

i zoomed out to 3 years to add perspective.

after sailing through 2020 and H1 2021 basically untouched there was a sudden and durable change that took place right when vaccination occurred. this is, to be sure, not proof of causality, but it’s quite a thing to ignore as well.

it sure looks like SOMETHING happened. we’re just trying to figure out what.

this made me wonder about other countries.

these are pieced together so scaling may be a little wonky but the color is what you really want to focus on. the shift from blue (low) to orange (excess) is, again strongly pronounced and the timing provocative.

we see a similar pattern in denmark

in all 3 of these cases there was not much of an excess death signal in 2020, it had gone negative in h1 2021, and it flipped positive in h2 and never looked back.

taken together this starts to look like something quite worrying.

but sweden was different and bucked this pattern.

they got surges of excess deaths in 2020 and early 21, then, despite vaxxing, did not see the jump in all cause mortality later and are now running pretty normal vs the elevated levels in the other 3.

so this is a bit of a “well, vaccines seem associated with excess deaths unless they aren’t” situation which could imply more complex causality or just spurious correlation.

in seeking to puzzle through what might explain this, some salients:

sweden “got covid” considerably earlier than the other 3. my best guess on why was worse timing on vacations etc that brought a ton of infected people back and just the wrong moment (akin to what happened in louisiana due to mardi gras).

i know a lot of people are going to claim it was “not locking down” but the simple fact is that none of these countries really locked down much or for long and there was never any correlation on decreases in social mobility (as measured by hard google mobility data) and any covid reduction outcome. the scandinavian countries are basically indistinguishable and you can see how unlike harder lockdown spots like spain they were (~25% drop vs 80)

whatever the cause of the early surge in sweden, it looks like that surge had lasting effects. this would seem to point us at a “dry tinder” sort of explanation where there were X number of vulnerable folks and they get affected eventually (or perhaps pulled forward).

this foots with the fact that the overall cumulative excess deaths in these countries have basically converged. denmark is currently lowest (but ramping fast) and finland and norway higher than sweden. (graphics from OS who has done a lot of great work here)

note that these do not include the last two months where sweden has had the fewest (and even negative) excess deaths vs group. if current trends hold, sweden could wind up being a major outperformer vs peers and if this keeps going and becomes large and durable divergence, we’re going to need some answers as this seems to lean against a pure “dry tinder” scenario.

the fact that sweden did not see a big spike in ACM post vaxx seems inconsistent with the idea that getting the jabs themselves was the proximate/immediate source of ED in other places. it would seem to need to be something a bit more complex, perhaps around antigenic fixation enhancing vulnerability to future variants, a rise in other forms of death, or some other issue.

this would foot with the idea that high rates of initial exposure to actual covid would have resulted in better and more complete/sterilizing/flexible/durable covid resistance in the population and that the order of encounter matters.

perhaps getting covid before getting vaxxed allows for more robust long term immunity vs getting vaxxed as first immune training.

perhaps getting vaxxed first prevents the acquisition of viable sterilizing immunity later and left people more open to later problems from omicron which they seem to struggle more to clear than the unvaxxed.

there is also sound evidence that the double jabbed fail to generate novel responses to new pathogens of variant based boosters.

and so it’s quite plausible that those who got their initial immunity from the jab would later wind up susceptible to the evolving “escape” variants generated by leaky vaccines vs those who had naturally acquired immune training.

variant specific boosters fail to elicit variant specific response

so, my theory to bat around is that this excess mortality is likely driven by vaccines creating later susceptibility but that this risk can be mitigated if most of those getting vaxxed had previously had covid and thus have some form of sterilizing immunity already. you do not unlearn old tricks, but you can be prevented from learning new ones.

this finds confluence with the outcomes and with theories on OAS/antigenic fixation.

there is evidence to back that up and it makes intuitive sense to me, but this whole take is a bit speculative at this point and it looks like there are quite a few moving parts here and i can certainly find other EU countries that fail to fit this pattern so i’m not sure it’s a really satisfying conclusion.

long time readers know that i like to make testable forward predictions, so here we go:

the test would seem to be what happens in the next 6 months or so. if the excess deaths in the other 3 drop to normal and basically equalize overall with sweden, then it was likely just a dry tinder pull forward scenario and the vaccine timing may well be a coincidence.

but, if their exces deaths remain elevated and sweden’s remain low, then the hypothesis i advanced above starts to look like the more convincing explanation.

SOMETHING happened in norway, denmark, and finland between week 20 and 30 of 2021 that dramatically changed their excess deaths outcomes in a persistent fashion. it was not “opening up” as mobility had long since reverted to normal levels.

and whatever this was affected sweden far less (if at all) and even such minor effect was transitory.

so what was it?

i’d love to hear some other ideas on how to explain this and the striking data similarity in the 3 regional neighbors but the divergence of the swedes.

the data here is badly incomplete and so we just keep trying to figure out what the key issues and outcomes are, laying pieces on the floor, and eventually hope to assemble by mosaic that which would be hilariously easy to know in a pretty definitive sense if public health agencies would just release the fricking data we need to assess it.

release ACM by age, vaxx status (for real, not the bad data the CDC uses), and vaxx type and we could have this sorted out in an afternoon.