Recommendations, and an Immunity Explainer

Thank you for all your emails after the last newsletter. It is nice to know I am not the only one who cried in my car recently. Someone sent this piece, which I liked a lot.


As we all continue our quarantine journey, I thought I’d start here with a few recommendations which have been working for our family. And then, back to COVID. Today, I wanted to dive into the question of immunity.


COVID Immunity: What is the Deal?

Last weekend, the head of the World Health Organization suggested there was no reason to think that having had COVID once would mean you will not get it again. At the same time, countries are talking about “Immunity” passports for people who have already been infected. Sweden is engaged in a controversial strategy of allowing the virus to spread relatively widely, moving towards “herd immunity”. These all seem fundamentally at odds. If previous infection doesn’t mean immunity, what is an immunity passport? How can these things all be true?

As we all grapple with the epidemic, this seems both very important and very confusing. A big piece of the problem is there is a lot of uncertainty which will not be immediately resolved. Expert’s impressions about the relative probabilities of various possibilities differ. This is all reasonable. More troubling is when people make statements which seem like they are certain and, in fact, what they mean is more nuanced.

The truth here is nuanced. I thought an explainer was in order. (For a more technical version of at least some of this material, this Lancet piece is excellent).

Let’s step back. When you get a viral infection — any viral infection — your body fights it by producing antibodies. These are proteins which stick to and kill the virus. They come in various types, denoted by ending letters (IgM, IgG, etc) and each have slightly different functions and timing but, basically, they all fight the virus.

Once you recover from a virus, the antibodies stick around. They live to fight another day! So if your body (say) re-encounters that same virus immediately, you are ready for it. You can see why this would be sensible from an evolutionary standpoint: people wouldn’t survive very long in a high disease environment (read, all of human history) if they were immediately reinfected with viruses.

In other words, these antibodies confer some “immunity”.

Antibodies are also how vaccines work. When you (or your kid) get the measles vaccines, they are getting a weakened form of the measles virus which prompts the body to produce antibodies against measles. If your child then encounters the real measles virus out in the world, they already have the antibodies to fight it off and will not get sick.

The basic science here is not in debate. Just like measles, the flu, chicken pox, etc, etc, in the case of this coronavirus (which is technically called SARS-COV-2) people who get it and recover have antibodies to it. And just like with those other viruses, you can test for these antibodies.

How? One option is you can use a very simple test — a quick finger stick of blood — and see if people have any antibodies at all. Or you can use a more involved test with more blood — and these are still being rolled out — to figure out how much of these antibodies you have.

Also not up for debate is the question of whether, in general, antibodies confer some immunity. They do. If you have antibodies to SARS-COV-2 you are somewhat protected from getting it.

However, there is a tremendous amount of uncertainty around the details. And that’s where the devil is.

There is a lot of variation across viruses in how long antibodies (and therefore immunity) stick around. Chicken pox antibodies seem to confer very long term immunity; flu much less so.

There is variation across viruses in how much antibody you need to not be infected. Generally higher rates of antibodies confer more protection but the “threshold” isn’t consistent.

There is variation across people in their degree of antibody response — some people produce more than others. Some of this may relate to severity of infection. In the case of COVID-19, a lot of people are infected without symptoms. They may have a less significant antibody response (or they may not! We don’t know).

Aggregating up, we also do not know what share of people would need to have antibodies to keep the virus out of the overall population (this is what is known as “herd immunity”). That number varies based on how infectious the virus is — how fast it spreads — something we also do not know perfectly. For measles (for example) it’s so infectious that you need a huge share of people to be immune (estimates suggest around 90%) in order to develop herd immunity. For the flu, which spreads much less quickly, this is under 50%.

With existing viruses, we know a lot of answers here because they’ve been carefully studied. You can use lab data, or evidence across populations, to answer questions about measles immunity. With a new virus, we still need to learn all of this. We cannot be certain about the answers to most of these question in the context of the virus causing this pandemic, SARS-CoV-2

However, it’s misleading to say we are flying totally blind. It makes sense that we could learn something about SARS-CoV-2 from other, (genetically) related, coronaviruses. SARS-COV-2 is closely related to SARS-CoV, the coronavirus that caused SARS in 2002 (the original SARS) and other coronavirses. Based on those, we do expect immunity among individuals who have been sick. This immunity seems to persist in SARS patients at least through a year out from symptoms (long term studies are somewhat lacking). Data on the also related Middle East Respiratory Syndrome (MERS) suggests a perhaps longer period of immunity.

These other viruses and our general analysis of the speed of viral spread suggest herd immunity may arise at around 60% infection rate. If we learn that the virus is more infectious than we thought, this may go up.

Putting this all together, I think what the head of the WHO was trying to say is that there is a lot of uncertainty. For a bunch of reasons, we do not want to give people the impression that if they have evidence on antibodies on a finger prick test they are totally protected from the virus. Nor do we want to give policymakers the sense that if exactly 62% of their population shows antibodies, they are good to go on opening the economy without restrictions.

On the other hand, suggesting that there is “no immunity” to this virus is also misleading. It is theoretically possible that we will learn that SARS-COV-2 is unlike all related viruses and infection does not confer any immunity but that is incredibly unlikely. We have to make decisions based on some uncertainty here, based on the best evidence we have. There is value in knowing antibody levels in individuals and populations, because our best information suggests that they will confer some immunity.

As always, it comes down to: More Testing. Better Testing. More Testing. More data.

Weigh in!

Keep the thoughts coming. I cannot write back to everyone but I do read all of your emails, I promise.

Write to Me!

Where to Find Me

My Website


Expecting Better