I was surprised, at first, when many heavily vaccinated countries were hit by a new wave of covid-19 at the beginning of autumn. I was surprised, that is, until I started to see studies coming out that showed that the protection offered by the vaccines is far less impressive than was initially thought, and drops to low levels after just a few months.
In light of this, I’ve been comparing covid death rates between different countries, to try to understand exactly what’s going on. Death rates are far preferrable to case rates, because they are much less variable over time. Case rates have varied enormously over the course of the pandemic as the amount of tests being carried out has changed, as the definition of what constitutes a case has changed, and as the tests themselves have changed. Case rates are therefore impossible to use as a tool for understanding how the pandemic has evolved over time. Although different countries define covid deaths differently, they tend to be pretty internally consistent over time. Death rates are thus far more reliable than case rates, and therefore far more useful for understanding how the pandemic is evolving.
So, here’s Sweden, the country I live in and therefore know best:
What we see is an initial large wave in spring of 2020 due to the initial Wuhan variant, then a drop to virtually zero deaths due to the onset of summer. It should be clear to everyone by now that covid-19 is a highly seasonal virus, which, like other winter viruses, largely vanishes from late spring to early autumn.
What we see next in the Swedish data is a resurgence of the Wuhan variant in the autumn of 2020, which begins to decline after a few months as sufficient population (a.k.a. “herd”) immunity is reached. This decline is however halted and countered by an even more rapid rise in deaths, which is due to the arrival of the British alpha variant on Swedish shores.
How can the alpha variant cause another wave if population immunity has already been reached, you might ask?
Because the threshold for population immunity is dependent on the infectiousness and transmissibility of the virus. The more transmissible a variant is, the higher the threshold for population immunity becomes. So the threshold was reached for population immunity against the Wuhan variant in December 2020, but when the alpha variant arrived, the threshold rose to a higher level, and a new bout of pandemic spread began.
Let’s get back to what we see in the graph – so, the the alpha variant quickly burns through the population and sufficient population immunity is reached against the new variant by mid-January 2021. Once again it becomes difficult for the virus to find new hosts, at which point the rate of infections drops down to a lower, more endemic seasonal level, which it remains at until the arrival of the new summer season.
For those who would like to attribute the decline in covid deaths in February to the vaccines, I would point out that only a few percent of Sweden’s population were vaccinated at this point, so the vaccines cannot have had anything to do with the decline.
After summer, the levels start to rise again to a slightly higher seasonally appropriate level, but remain at the low level you would expect for a virus that has now become endemic. Even though the highly infectious delta variant arrives in Sweden in late spring, and is by autumn totally dominant, it is not able to create a new wave, due to the high levels of pre-existing immunity.
We see very similar patterns for other places that, like Sweden, were hit hard in the spring of 2020. Here’s New York:
And here’s Lombardy, in Italy (which for some reason unfortunately isn’t showing the first few months of 2020):
Here you clearly see the first two waves caused by the Wuhan variant, then the third wave caused by the alpha variant, and then nothing, in spite of the arrival of the delta variant. The inability of the delta variant to create a new wave in these places could be explained in two ways – either it’s not sufficiently more transmissible than the alpha variant to generate a new wave in places that already have population immunity generated by the alpha variant, or the vaccines are doing their thing, for now.
Let’s turn to India, because of what it teaches us about the delta variant:
In early 2021, the Delta variant springs in to existence in India, and rapidly races through the population. Population antibody testing reveals that roughly 50% of India’s population become infected over the course of just a few months, with the proportion of the population with antibodies rapidly rising from 20% to 70%, at which point sufficient population immunity sets in for viral spread to drop down to low endemic levels. Note that the vaccines clearly had no part to play here, since, just like with Sweden, only a few percent of the population were vaccinated at the point when the death rate dropped to low levels.
Now let’s look at some countries that have suffered a fourth wave during the autumn, and try to tease out why. Here’s Israel:
Israel is able to avoid getting much covid spread during the spring of 2020. During autumn it is hit first by the original Wuhan variant, and just as population immunity to that variant reaches levels where spread is beginning to decline, the country is hit by the alpha variant, with deaths peaking in late January 2021. At that point 20% of the population are already fully vaccinated, so here the vaccine may actually have played a role in causing the death rate to turn down. That could explain why the death rate thereafter drops very low quite quickly, instead of lagging at a more endemic level all the way in to May, like in Sweden (which was much slower to vaccinate).
Covid deaths stay low throughout the summer, as we would expect. Then we come to autumn 2021, and the surprising fourth wave. Or not so surprising if you look at the data which now shows pretty clearly that vaccine effectiveness drops rapidly, even when it comes to preventing severe disease (which is especially true for the frail elderly, who are after all the only segment of the population at serious risk from covid-19).
So, Israel gets hit by a fourth wave, as do many other palces. Why are the places discussed at the beginning of this article, Sweden, Lombardy, and New York, not currently experiencing fourth waves?
As I see it, there are two possibilities. The first is that these places have developed so much natural immunity, thanks to the fact that they’ve experienced a couple of extra months of heavy spread of covid-19 during the spring of 2020, that covid is now over and done with in those places and no more big waves are coming. Israel has high rates of vaccination, but at the beginning of autumn 2021 it had experienced fewer months of pandemic spread, and thereby had a lower proportion of the population that had developed natural immunity from prior infection. It’s been pretty well established by now that the immunity conferred by infection is far more durable than the immunity conferred by vaccination, so that is a reasonabe hypothesis, now that we know the immunity generated by the vaccines is so fleeting.
It can be instructive, here, to look at Eastern Europe. The eastern European countries have been particularly hard hit this autumn. Here’s Bulgaria:
And here’s Slovakia:
Notice anything special about these places?
I think two things are important to pay attention to. First, both places were almost completely spared in the spring of 2020. Second, both places still had a high degree of viral spread when the onset of summer caused infections to drop. They thus never reached population immunity to the more infectious variants, and were thus always going to have a resurgence in the autumn of 2021.
So, the first possible explanation I mentioned for why some places are not experiencing a fourth wave is that those places now have sufficient natural population immunity, which is protecting them. The second option is that these places are currently enjoying temporary protection, afforded by the fact that they vaccinated their populations later than places like Israel. If that is the case, then they will head in to fourth waves in another month or two.
The data from Germany suggests that the first alternative is more likely to be true. Here’s what the curve for Germany looks like. It currently appears to be heading in to a fourth wave.
Notice that Germany, like Israel, was barely touched by covid-19 during the spring of 2020. Instead it had a big wave during the winter of 2020/2021, caused by the Wuhan variant. Then there was a small spike caused by the alpha variant, which grew to become the dominant strain in Germany in April. The alpha variant was however hindered from causing a big new wave by the arrival of the warmer season. During this time period, Germany mass vaccinated it’s population, with most vaccinations happening between March and June. This is very similar to Sweden, which also vaccinated most of its population between March and June.
So why is Germany experiencing a resurgence now, and Sweden isn’t?
Clearly, it can’t be due to Germany being vaccinated earlier and losing immunity earlier, since both countries vaccinated their populations at the same time. For that reason I’m inclined to favour the first hypothesis, that Sweden has built up more population immunity, for the simple reason that covid started spreading massively in Sweden in spring of 2020, but didn’t start spreading properly in Germany until autumn of 2020. So, although the effect of the vaccines has already waned in both countries, Sweden is protected by its widespread natural population immunity, while Germany isn’t. If that is the case, then Sweden shouldn’t see another big wave. In another month or two we’ll know what the truth of the matter is.