The Dark Side of Herd Immunity

Way back in 2020 our authorities were pushing heavily the idea of herd immunity for Covid – that is, through vaccination (well, and natural infection) we will all become resistant to the virus and it would simply disappear, just as happened with smallpox (and no other viral disease, despite many vaccines being available for many diseases). Of course, the more likely outcome of herd immunity would be a significant reduction in cases without eradication of the virus.

But what’s all this about ‘herd immunity’?

On the face of it, herd immunity is quite simple. Once enough of a population has gained immunity to a disease, whether by natural infection or by use of an effective vaccine, then the disease dies out and is no longer a threat. However, it is somewhat more nuanced than this. The important factor is really the number of people that each infected individual passes the disease on to.  When the number of individuals susceptible to the disease decreases such that each infected person infects, on average, less than one other individual then the disease will die out in that area. Of course, it doesn’t mean the disease will go extinct – it might remain viable in animal reservoirs or in other areas with lower immunity levels – but once the point is reached where there are sufficiently few susceptible individuals then the disease passes into a new stage typified by sporadic localised self-limiting outbreaks.

This brings us to the mathematical characterisation of ‘herd immunity’. The point at which the number of susceptible individuals falls just below the level required for an outbreak of a disease is the ‘herd immunity threshold’, which can be described by a nice simple equation for situations where there is perfect immunity to the disease following infection or vaccination:

Herd Immunity Threshold = 100 × (1 – 1/R₀)

Where R₀ is the basic reproduction number (the doubling rate in a population with no immunity to the disease).

It is this simple equation that created all of the estimates of the levels of vaccination required to ‘make Covid go away’ that we were bombarded with in 2020.

The WHO estimated in January 2020 that the R₀ of Covid was between 1.4 and 2.5. But estimates of R₀ during these early days of the pandemic were quite varied, ranging between around 1.5 and 6.8, giving herd immunity thresholds of somewhere between 33% and 85%. This unhelpfully large range was narrowed down by the modelling team at Imperial College, which decided that the R₀ of Covid was around 2.6 (giving a herd immunity threshold of 62%). By September the Imperial modellers had settled on an R₀ of about 2.5, giving a herd immunity threshold of 60%. This estimate appears to have driven the reporting of the numbers required to get vaccinated to eliminate Covid that appeared during autumn 2020.

The trouble with the simple herd immunity threshold equation shown above is that it assumes that vaccines and natural immunity offer complete protection from infection and onwards transmission. The full equation for the threshold is given by:

Herd Immunity Threshold = 100 × (1 – 1/R₀)/ε

Where ε is the effectiveness of vaccination or natural infection in protecting against onwards transmission.

This substantially more nuanced equation for the herd immunity threshold appears to have been ignored in the reporting to the population of the levels of vaccination required to rid countries of Covid. This was despite there being an understanding that infection with a ‘traditional coronavirus’ type cold doesn’t confer long term immunity, the failure of prior attempts to create coronavirus vaccines and the imperfect performance of vaccination to protect against some other common respiratory tract infections (e.g. the influenza vaccines). 

Applying this equation to a vaccine with 50% effectiveness (the goal of the vaccine programmes) and an R₀ of 2.5 (the assumption in official U.K. modelling for the latter half of 2020) gives a required herd immunity threshold of 120%. As you can’t vaccinate more than 100% of the population, this is impossible. Yet, even given this information the push for herd immunity via high levels of vaccination continued, helped by the weekly reporting in the UKHSA’s Vaccine Surveillance Report of the proportion of the population that had been vaccinated.

Of course, not only was the concept of herd immunity doomed from the start, but as each week went by the vaccination campaign started to target ever lower age ranges, coupled with ever higher estimates of R₀ to support the increased proportion of the population required to be vaccinated. When the Alpha (Kent) variant appeared it was estimated to have an R₀ of around 4.2, giving a herd immunity threshold with a perfect vaccine of 76%; the Delta variant had an estimated R₀ of about 7, giving a herd immunity threshold with a perfect vaccine of 86%. Of course, these higher R₀ values meant that, even assuming a perfect vaccine, herd immunity couldn’t be reached without vaccinating children, and so youngsters were next in line to get jabbed. The decreasing vaccine effectiveness during this time was ignored in these herd immunity calculations.

What was surprising about the reporting of the vaccination rates and the slow progression towards the purported ‘herd immunity threshold’ was the lack of consideration of the impact of natural infection. Even when natural infection was mentioned, it was usually to point out that there was only a short term protection from infection offered by natural infection, unlike the vaccines which we were told offered magnificent protection. This strange state of affairs continues to this day, with countries such as the USA requiring two doses of Covid vaccine to enter the country, with no consideration of the longer term protection offered by natural infection nor of the impact of rapid vaccine waning or vaccine escape variants.

There’s an additional impact of the contribution of natural infection to herd immunity. Given that healthy younger individuals were and are much less likely to suffer from severe Covid, there was an argument to be made for allowing children and young adults to become infected with Covid rather than relying on vaccination. Furthermore, as the key period was summertime in 2021 there was also a natural (seasonal) lowering of the rates of severe Covid (possibly due to vitamin D levels – another aspect of Covid ignored by our authorities). Instead, the younger individuals in our population were subjected to rather stringent lockdowns, with strict restrictions on schooling, the cancelling of music festivals and the restriction of students to their accommodation. Moreover, when young people did have the audacity to enjoy themselves they were lambasted by the media as being selfish. I have long thought the opposite and that it was the older age groups that were selfish. Social interaction is very important during our younger years – this is when we learn to be adults and also partake in important social activities such as the down-selection of future life partners – and so the relative suffering imparted by the social deprivation of lockdowns was greater for the younger generations. I’ve suggested in the past that those aged over 50 should have stood outside one Thursday evening and clapped and banged pans to give public thanks for the sacrifices made by the young for the old during Covid. However most of those in older age groups remain ignorant of the sacrifices made for their sake by the young over the last three years.

Herd immunity is a reasonable goal for many diseases and vaccines, but it also has a dark side, particularly for rapidly mutating RNA viruses where vaccines offer marginal protection. As the proportion of the population with immunity conferred by vaccination increases, the evolutionary pressure for the virus also changes to focus on overcoming vaccine immunity. Even worse, if the immune protection is limited to a very specific part of the virus (as with the viral-vector and mRNA vaccines), this evolutionary pressure will be even more targeted. The entire vaccinated population (save those previously infected) will have a very similar immune protection (immunised against identical viral proteins), compared with the unvaccinated-and-infected who will have the more complex immune protection against the whole virus and also potentially against non-identical variants. Thus in the event of a vaccine escape mutation creating a new variant, those with immunity from natural infection will be relatively more likely to have some maintained protection compared with the vaccinated.

The impact of this effect can even be modelled using equations derived from the herd immunity equations as given above. Indeed, in a simplistic case the equation is identical. Consider a hypothetical situation with a population having received a vaccine with 100% vaccine effectiveness and the rest gaining immunity from natural infection (also 100% effective for this example). A vaccine resistant strain emerges that doesn’t impact on natural immunity (this is unlikely, but the example is for the extreme case). In this situation the original herd immunity threshold equation will apply but ‘in reverse’, i.e., in this case the herd protection will be conferred by those with immunity after natural infection and it is these individuals that will protect the vaccinated. To put some numbers on this, if the new vaccine escape variant has an R₀  of 4, then if over 75% of the population remain unvaccinated but are previously infected (i.e., 100 x (1 – 1/R₀) = 75%) then their immune protection will stifle the spread of the new vaccine escape variant and help keep the vaccinated (vulnerable to the vaccine escape variant) protected. In this simplistic example, once vaccination levels exceed 25% then this protection will fail (assuming all the vaccinated have no natural immunity and all the unvaccinated have natural immunity) and the vaccinated will spread the vaccine resistant strain until the population overall achieves the herd immunity threshold via natural infection.

Reality is, of course, much more complex than the example above. Considerations have to be made for the relative effective reproduction number of the original and vaccine escape variants in the unvaccinated and vaccinated populations. However, the general concept remains – a population with high levels of vaccination with an imperfect vaccine can be more likely to spread vaccine resistant variants than populations with lower vaccination coverage in conditions where the unvaccinated have high levels of natural immunity. It is likely that through universal vaccination we have encouraged vaccine escape variants to occur that wouldn’t have arisen if we had only vaccinated the most vulnerable – as then the selective pressure on the virus would have been to evade a complex natural immunity, not the highly heterogeneous vaccine immunity. Yet again, the goal of universal vaccination to achieve herd immunity looks like it might actually be a false idol to worship.

I’d note that we don’t normally worry about the above – most vaccines offer high levels of protection against disease and also are either based on an inactivated virus (complex immune protection against many proteins in the virus, some of which will mutate slowly) or are based on viral proteins that mutate slowly (in the case of most protein subunit vaccines). It is only in the case of the Covid vaccines that we chose a small portion of the virus that happens to be highly mutable, thus resulting in the problem of rapid vaccine escape being near inevitable.

There’s another complication related to herd immunity thresholds that can be an important consideration for vaccination policy: the role of community infections while below the herd immunity threshold in naturally boosting immunity. The great example of this that has come to light in recent years is the strange situation of chickenpox that appears to have arisen in the USA. 

Chickenpox is an unpleasant though largely benign disease that occurs mainly in children. It is more unpleasant in adults, but the vast majority of individuals catch chickenpox as a child and gain lifelong immunity. Thus adult chickenpox is a relatively rare disease. However, chickenpox has a dark side. The virus that causes chickenpox, varicella zoster virus (one of the herpes viruses) remains in the body after infection, quietly residing dormant in certain nerve cells until a point, many years later, when the virus becomes reactivated causing a painful neurological inflammation that we call shingles.

Way back in the mid 1990s the USA started vaccinating children against chickenpox. There were arguably some reasons in favour of this, including trying to prevent the relatively small numbers of hospitalisations and deaths from childhood chickenpox that occur each year in the USA. There are rare serious side-effects that arise from the chickenpox vaccine, but the U.S. authorities’ assessments suggested a net positive outcome, and so the vaccination programme was started. And it worked – chickenpox rates slowed significantly in the years following the vaccination programme, and some hospitalisations and deaths were averted. Even with a fairly high rate of breakthrough infections (around 10% to 20% of the vaccinated go on to catch chickenpox later in life), when breakthrough infections do occur they tend to be mild and have a much lower risk of hospitalisation than otherwise. As a result, the U.S. health agencies considered their childhood chickenpox vaccination programme a success.

But then in the early 2010s a new phenomenon started to be reported – a significant increase in the incidence of shingles since the onset of the vaccination programme, particularly in non-elderly adults.

According to the natural boosting theory, what has happened is that adults who had caught chickenpox as a child kept their immunity ‘topped up’ by repeated exposure to the chickenpox virus from infected children throughout their lives. However, once all the children were vaccinated the incidence of this ‘topping up’ effect reduced significantly. After a few years this lack of repeated ‘topping up’ of immunity resulted in the latent virus being able to flare up, and as a result shingles rates rose significantly. Thus the unexpected impact of the childhood vaccination programme appears to be a significant increase in shingles in adults.  Fortunately, this effect should only last for 40-60 years, at which point all those individuals who were vaccinated as a child should have much lower incidence rates of shingles as their childhood vaccination will have protected them from getting the disease in the first place. Or perhaps it won’t – we just don’t know (yet). In the meantime, the U.S. authorities are keen to point out that the problem of adult shingles rates increasing can be partially resolved by giving all adults not vaccinated as children the shingles vaccine – so a problem caused by vaccines can be resolved through the use of more vaccines. A bit of a win-win for the pharmaceutical industry.

I should state here that the above isn’t accepted by everyone – some say that the rise in cases of shingles since the mid-90s is a coincidence, and it is actually due to other reasons. Nevertheless, the scientific basis of the ‘top up’ theory is sound and it illustrates how complex immune effects can turn out to be.

I must also note here that this appears to be unrelated to the dramatic increase in shingles cases from spring 2021 – this phenomenon isn’t explained by childhood vaccination for chickenpox and remains baffling to doctors and scientists.

So far the evidence is that for Covid the introduction of mass vaccination has coincided with significantly increased Covid incidence rates, so it is unlikely that we’ll see an effect of ‘reduced immune top-up’ similar to that seen after mass vaccination for chickenpox. However, it is possible that the effect of lockdowns in 2020 and 2021 has been to reduce this ‘immune top up’ for various common diseases. Indeed, this is exactly the mechanism proposed by our authorities to explain the significant increase in respiratory tract infections we’re currently seeing (earlier in the year than is usual). I’m not convinced that it is the entire reason (I suggest that there’s also some negative impact on the immune system from Covid itself or the vaccines), however it is likely that it has had some effect.  It is strange that they didn’t warn of this negative impact of lockdowns at the time, or that it isn’t usually mentioned as a possible consequence of mass vaccination campaigns.

Before I close, I must state that I believe for many vaccines and viruses high levels of vaccination is a valid goal. I suggest that our approach to mass vaccination of children has effectively eradicated many diseases that resulted in the high child mortality rates of the past, and ‘herd immunity’ means that those that cannot receive a vaccine are protected from infection with these specific diseases. I know that many will disagree with this and I expect some negative comments as a result – but what’s important is that people are allowed to disagree and that the most healthy situation is that we can have a discussion. The one thing I am sure of is that suppression of alternative opinions and the rule of the ‘voice of authority’ is a very negative road to go down.

Next time I’ll discuss the other side of the data on herd immunity – the section of the Vaccine Surveillance Reports reporting antibody levels in the population.

Amanuensis is an ex-academic and senior Government scientist. He blogs at Bartram’s Folly – subscribe here.