Today, we are going to talk COVID variants. As things start to open up, as we dip our toes (or our whole bodies) into being around other people again and bringing our kids out into the world, variants are nagging at the back of a lot of minds. Are these COVID variants a reason to retain caution? Are they worse for kids than the “main” COVID? And, honestly, what are they? I’ll try to provide some answers here.
What are Variants?
Viruses mutate all the time — when viruses replicate, they copy over their genetic material. Sometimes, errors are made in the copying. Most of the time, the errors are harmful for the virus and they do not persist. Sometimes, however, a mutation or a series of mutations is beneficial to the virus in some way, and persists.
These beneficial mutations lead to a virus variant. These variants are most troubling when the mutation is in the spike protein, since that is the protein targeted by antibodies from vaccination.
To see why this might matter, recall the way vaccines work. The majority of vaccines we are using — mRNA or others — work by getting your cells to produce the COVID-19 spike protein (but not the rest of the virus). Your body recognizes this protein as foreign, and creates antibodies to destroy it. These antibodies — and the T-cell memory to make them — wait in ready in case the actual virus arrives in your body.
The antibodies are trained to recognize the particular spike protein encoded in the vaccine. In this case, it’s the spike protein for the original “wild-type” version of COVID-19. If the body, instead, encounters a variant with a slightly altered spike protein, the antibodies may be less good at identifying and killing it. In this sense, either previous infection or vaccination may be less protective against the variants.
Even if the spike protein is altered, antibodies will work against it to some extent, just perhaps less so. I say “perhaps” because this is really a question for data, and I’ll go through more of what we know (it’s incomplete) below.
The WHO identifies both “Variants of Interest” (VOIs) and “Variants of Concern” (VOCs). The second group are “of concern” since they are spreading more widely. Currently, the WHO identifies four VOCs, listed in the table below. Until recently, we’d been calling these based on where they were first identified, which was not a good naming system for many reasons. Greek letters have taken over.
What are the Concerns about Variants?
There are three primary reasons to be concerned about a new variant. First, it could be more contagious. Second, it could cause more serious illness (or more serious illness in some people, like children). Third, it could evade existing antibodies, thus making vaccines less effective. We can take these in turn.
(Let me say that the CDC also has a very helpful webpage on this which is a little more in the weeds but has good citations).
Contagiousness Generally, all of these variants appear to be more contagious. Exactly what mutation is responsible in each case isn’t entirely clear and certainly varies. In the case of the alpha variant, for example, the mutation N501Y seems to play a role in stronger binding to our cells.
There has been a lot of discussion in the last few days about the delta variant, originally identified in India. It seems to be significantly more contagious, is responsible for a large share of current infections in the UK, and has increased as a share of infections in the US.
From a virology standpoint this higher level of contagiousness is expected among variants which rise to a level of concern. If they were less contagious, they would struggle to take a foothold in the presence of the dominant strain. Evolutionary selection favors more contagious variants.
(One note: in the recent discussions of the delta variant higher rates of infection in the US and UK are noted among younger people. Some have taken this to say that these variants are more contagious among this group. However: this is heavily confounded by the fact that these are the least vaccinated age groups, so more susceptible to any infection).
Serious Illness It is difficult to know whether these variants cause more serious illness. In contrast to contagiousness, it’s less clear that viral evolution would favor this. Viruses want to spread, and making your host too sick can be counterproductive.
In order to establish whether the variants cause more serious illness, we need information on illness severity broken down by variant type. This represents a significant data challenge; keep in mind, we still have imperfect information information about more basic questions, like the overall case fatality rate of COVID-19.
The most well-studied variant is alpha and there was some initial concern that illness was more serious. But more recently published follow-up doesn’t support that, although the evidence is overall still limited to small samples.
Concerns have been raised, too, about the gamma variant impacting children more, given the tragic number of child deaths in Brazil, where this variant is dominant. But, again, we would need to know much more to conclude there is a link. Total deaths in Brazil are likely heavily undercounted, making it unclear if child deaths are disproportionate.
Moreover, and tragically, children are more susceptible to illness in many developing countries, given more limited resources access. I’m not suggesting it is impossible or even implausible that a variant could be relatively worse for younger people, but we need more information to learn about it. Early on, there were suggestions that the alpha variant was also disproportionately impacting children, which turned out to be false. There is a danger in too-fast messaging about this, too, since people may not hear the follow-up.
Bottom line: more needs to be done to learn about this, although preliminary data seems to point to similar levels of seriousness to the extent we can evaluate.
Antibody Evasion/Vaccine Efficacy The third question on variants is whether the vaccines are less efficacious against them. Again, this is a hard question to answer. We do not have sufficient data to be precise about this in a real world setting for most of the vaccines we use. (If you want another take on this, I found this Twitter thread from Natalie Dean very helpful).
Instead, much of our data comes from two sources. The first is evidence on neutralization in a lab setting by antibodies from vaccinated or recovered individuals. That is: you can take plasma from people who are vaccinated or have had COVID, and see if their antibodies are sufficient to kill the variant viruses and to what extent.
In these tests, research has found that the vaccines are similarly effective against the alpha variant, but lose some efficacy against the beta, gamma and delta variants. One issue is that it is difficult to know what line to draw between this reduced efficacy and disease. Given the very high antibody production as a result of our vaccines, some reduction may leave the overall efficacy very high.
A second piece of evidence is on breakthrough infections. Here, we also see some reduced efficacy in variants, with the variants of concern over-represented among breakthrough infections (i.e. here). In the UK, the spread of the delta variant seems to be occurring among those with only a single shot, suggesting the (recommended) second shot with the mRNA vaccines may be crucial for a robust response.
Adding to this is that the breakthrough infections we are seeing tend to be mild or asymptomatic; overall, the vaccines seem to be very broadly efficacious against severe illness and death, even in locations where variants are prominent.
Putting everything together, it seems very likely that the efficacy of the vaccines against variants is lower for infection overall. That is: in the presence of current variants we’d expect similar performance against serious illness and death, but somewhat reduced efficacy against any infection. Precise numbers are not to be had, yet.
However it is very important to note that vaccines do protect you to a large extent against all circulating variants. The fear of variant is not a reason not to be vaccinated. It is, instead, a stronger reason TO BE vaccinated.
The question to ask, then: So what? What should we do differently?
How Should this Influence our Behavior?
Part of what I think has been challenging is the discussion of variants is vague. We are talking about them as if they are some completely different thing, a new risk which we cannot incorporate into our planning. But, that is not true. All of the calculations we’ve been doing, I’ve been doing, over the past months…we can think about how they are different in the possible presence of variants.
For example: let’s revisit unmasked offices from last week. I did the calculations there assuming a 90% protection from vaccines. Now, let’s assume that the circulating virus is a variant, so the vaccine is only 70% protective and that the baseline risk of getting infected from a COVID-exposure is 20% higher.
We should also, always, revisit the community risks, which are continuing to go down. The 7 day average of new cases in the US is 15,000; if we assume 10 days of infectivity and a detection rate of 50%, that means an overall US infection rate of about 0.09%.1
Putting this together, the office calculation risk comes out to about 1 in 15,000.2 If you had this interaction every day, you’d be infected once every 41 years. To be clear: it isn’t that the variants do not matter here. If we work with the baseline assumptions on vaccine efficacy and less contagiousness, we’d be up to once every 171 years. But both are very, very rare events.
Or, we can think about our kids interacting outside without masks. I calculated the risk of infection there at about 1 in 200,000. But even if that’s twice as high due to variants, it’s still really tiny.
The fact is, at the moment, in much of the US, the very low COVID case rates are providing a lot of protection, even against variants with more vaccine evasion. This doesn’t mean things couldn’t change. But the fact is that we need to make decisions about what interactions to have now.
To me, this is the crux of it. The presence of variants means we need to continue to be vigilant in paying attention to virus levels; we would need to do so anyway, and this makes it more so. And the presence of variants also strongly argues for continued vaccination efforts. But we should react to the virus levels and risks we see now, not to what we might see later. And the virus levels now are low; they suggest an ability to return to some normalcy.
This combination of vigilance and normalcy will be hard. We have had only two COVID modes, really: either we had never heard of it or it was the primary existential threat occupying our minds. This is a new mode, in which COVID is one of a number of risks we will need to think about, and pay attention to, but which needs to be put in context. Variants included.
This is the calculation approach I take a lot: daily case rates, assuming a 10 day infection period and a 50% detection rate. The share of people infected is then (15,000 * 10 * 2)/(328,000,000) = 0.09%
An infection rate of 0.09%, multiplied by a baseline transmission rate of 12%, multiplied by a reduction of 70%, multiplied by 2 for the two people in that example.
