‘Waning immunity’ is a phrase we hear often, but what exactly does it mean? There are three different forms: waning of infection-acquired immunity against contracting the disease; waning against vaccine-induced immunity against contracting the disease; and waning against protection (from either vaccination or prior infection) from risk of severe disease, hospitalisation or death.

Infection-acquired immunity from contracting Covid is expected to wane over time, as happens with other coronaviruses. This is a process additional to that of “breakthrough” re-infections. The latter occurs because immunity from contracting the disease is (assuming the first infection is mild enough) not perfect to begin with. So, even within a few months of an initial infection, the risk of re-infection is cut by only some 70-80%. 

(We should note in passing that this 70-80% figure might be to some extent a reflection of new variants partially escaping the immunity conferred by infection with an older variant. So, perhaps having had the alpha variant — the thing we used to call the “Kentish Variant” — confers only 70% immunity against infection with the now-dominant delta variant, but maybe the protection from infection with the same variant is higher.)

Waning is different from breakthrough. With waning, coronaviruses cunningly make the immune system forget them over time, so within perhaps two years or thereabouts people can get infected again fairly easily.

Absent boosters, waning would, along with immigration and new births, become a key factor in the endemic cycle for Covid. It would imply a lot more infections than is commonly grasped. The delta variant has a herd immunity threshold of about 85%, so in a stable endemic equilibrium we’d need about 85% of the population to be immune, on average, over time. For simplicity, let’s ignore non-immune immigrants and babies and just focus on a stable population, assuming we’re at the herd immunity threshold to begin with. If immunity waned gradually over two years, then for every individual among the 85% immune that lost immunity over that two years, another individual would have to become immune.

If each new infection created one extra immune person, that would mean we’d have to infect 85%of the population over a two-year period. In England there are 56 million people. So to infect them all over two years we’d have to infect about 77 thousand per day. If an infection actually only creates 70-80% of an immune person (as per the breakthrough infections discussed above), that 77 thousand per day would become 96-110 thousand infections per day. So to maintain the herd immunity threshold at 85% of the population we’d need to have about 65 thousand infections per day if each infection creates one immune person, or 81 to 93 thousand infections per day if the average infection creates 0.7 to 0.8 immune people.

It is of interest to compare that to the current level of confirmed cases. At the time of writing there are about 30-35 thousand confirmed cases per day and it is believed the ascertainment rate (the share of infections that are detected) is about 40-50%. So 30-35 thousand cases means about 60-88 thousand infections per day. In other words, we are current at about, or slightly below, the level of cases there would be in that endemic equilibrium.

Now waning from natural infection is unlikely to be as smooth as that. Even in the longer-run there may be some seasonality in infections. But the bigger factor in the nearer-term is that infection came in waves, so it is quite likely that the waning of infection will come in waves as well. The initial set of infections, covering some 10-12% of the population, occurred in or around March 2020. If waning occurred instantly, two years after initial infection, we’d expect a (modest) waning-induced wave of infections next Spring, then another wave in November/December 2022, and so on. Of course, waning will not be as instant as that, but the essential point remains that the process of getting to our ultimate endemic equilibrium may proceed via some bumps, with mini-waves echoing the timing of the initial waves during the epidemic.

A further reason things may not proceed as discussed above is that not all of us acquired our immunity via infection. Only around half the population of England has been infected, according to the University of Cambridge’s MRC Biostatistics Unit. By contrast, nearly 90 per cent of those over 12 have had at least one dose of a vaccine.

At one stage it had been hoped that vaccine-induced immunity might be longer-lasting than infection-acquired immunity (because the waning of infection-acquired immunity occurs through processes the disease induces in the body that the vaccines do not). But studies conducted since mass vaccination has occurred suggest the opposite. It appears to be vaccine-acquired immunity that wanes more rapidly. In some studies, around 50% of inwards transmission protection is estimated to be lost after about six months. Some analysts suggest that may be an over-estimate, as it has proven difficult to control very well for the effects of unconfirmed cases (boosting the immunity of the unvaccinated) and of the takeover by delta during the analysis periods (with delta having higher transmissibility and some vaccine transmission protection escape). But even if it does eventually prove to be an over-estimate the point remains that vaccine-acquired immunity may wane more rapidly than infection-acquired immunity.

Many people now have ‘hybrid immunity’ (immunity from having been both vaccinated and infected), which is expected to be quite long-lasting, and an increasing number have had a third booster dose (particularly older people — for whom vaccine waning appears to be slightly more rapid), which may not wane in the same way. Furthermore, just as those who get breakthrough infections despite double vaccination are not as infectious to others as unvaccinated infected people, the same may be true of those infected after waning. If infectiousness after waning is reduced, the herd immunity threshold may be cut a bit so the equilibrium level of daily infections may be lower.

For now we have a significant portion of the double-vaccinated-but-not-yet-infected population for whom waning may occur more rapidly, perhaps over the next few months. Indeed, some modellers suggest that vaccine waning may have been a factor even in the recent evolution of cases since July, perhaps explaining why cases overall have roughly plateaued (albeit with cycles around the average) since then and in particular why cases in older people started to rise in the month or so before the booster campaign really got going.

Vaccine waning could make a waning-induced wave early in 2022 a bit larger, by combining the waning of 2021 vaccines with the waning of 2020 infection-acquired immunity.

This analysis might seem concerning. After all, I’ve argued that even the average level of cases over the long-run might be higher than those we’ve had recently. And the latest cases have been heavily concentrated amongst low-vulnerability children, whereas the case numbers in my analysis above would be more evenly spread across the population. So it might seem like I’m suggesting there’ll be an awful lot of hospitalisations every year, forever.

There are four reasons things may not be as bad as they seem.

First, although the transmission protection of vaccines and infection-acquired immunity fades fairly rapidly, protection against severe disease, hospitalisation and death are expected to last much longer — indeed, for decades — apart perhaps from some of the most elderly and others with weakened immune systems at whom the initial booster campaign was targeted. Second, there is a view that the protection from serious disease continues to increase (albeit more gradually) even after the big jump from the first infection. Once we’ve all had Covid three or four times, our immune systems may be so used to it that it will be dangerous for relatively few of us. Third, there are new highly-effective treatments coming on stream that appear to eliminate the vast majority of hospitalisation and death risk even if they are taken after people have contracted the disease. And, fourth, as with flu, more vulnerable people are likely to take annual (or perhaps six-monthly) booster jabs, which will doubtless become customised to whatever variant is circulating at the time, keeping their immunity high.

So the pattern one might expect is as follows. We’ve now completed what one might call the ‘epidemic phase’ of the disease — the stage at which cases, hospitalisations and deaths could grow exponentially, in the way that they did in March 2020, November/December 2020 and May to July 2021. With schoolchildren, the final set of immune-naives, being infected in September/October 2021, we’ve shifted from that epidemic phase into a phase of transition between the epidemic and the long-run endemic cycle. That transition phase may be bumpy, because immunity may wane in a bumpy way, possibly creating new mini-waves of during 2022. But even these should produce relatively few hospitalisations, and over the longer term hospitalisations will probably drop further.

The policy upshot of all this is fairly straightforward. Boosters for the vulnerable; new therapies as they become available; no need for any restrictions such as mandatory social distancing or masks; encouragement of a modicum of common-sense and courtesy towards others if new waning-induced waves come next spring, but no more; and a ramp-down of testing in schools, isolation requirements for the unvaccinated and remaining rules for travellers.

Waning is an issue that public health officials will grapple with, as they do for other diseases. But as a grand policy question, Covid is finished and waning will not change that.