Coronavirus and immunity — what the latest research says about people who already beat the disease
(Francisco Kjolseth | The Salt Lake Tribune) Wayne Tarawhiti gives the thumbs up to family nearby as he supports his daughter Kazna as she plays her quarterfinal match up against Syracuse at the UCCU Center at Utah Valley University on Thursday, Nov. 2, 2017, defeating them in three straight sets. Wayne is battling a form of blood cancer and keeps covered up due to a compromised immune system.
Editor’s note: The Salt Lake Tribune is providing free access to critical stories about the coronavirus. Sign up for our Top Stories newsletter, sent to your inbox every weekday morning. To support journalism like this, please donate or become a subscriber.
Nuance can be difficult to communicate in the best of times. In a pandemic? Even harder.
So when the World Health Organization issued a release
at the end of April that said “there is currently no evidence” that people who recovered from COVID-19 can’t get it again, people were alarmed.
And I think that reaction was fair. If there truly was no immunity after already having the disease, the situation we find ourselves in would become drastically worse — and it’s already pretty bleak as is.
Luckily, most experts agree
that those who beat COVID-19 will probably have some sort of immune response. To defend the WHO a little, some of this evidence has come out since the date of their release. But immunity isn’t a black-or-white sort of thing: There is real nuance here. Let’s dig into what we know — and what we don’t — about coronavirus immunity.
Antibodies are small proteins in the blood that the body creates in response to an infection. These proteins are specific to individual types of infection and assist in defeating it. Afterward, the antibodies linger, with the goal of defeating future infections of the same type.
Does everyone who gets the disease develop antibodies?
Nearly everyone, it looks like. In a study released last week
, researchers in New York City looked at 624 coronavirus patients. Only three people did not test positive for antibodies at some point in the study. This is very good news.
Those 624 patients ran the gamut, too, in terms of age, gender and severity of disease. That’s really useful, because we previously had only limited evidence: For example, we knew people with really severe cases developed antibodies, but we didn’t know about the weak cases. That’s not really a worry anymore. The only group not studied in NYC was those under 18 years old, so we’ll have to do more testing on children later.
We don’t know why those three people didn’t develop antibodies, and it would be useful to find out. But from an epidemiological point of view, “nearly everyone” really helps.
It does take some time for the body to develop those antibodies, though. From the NYC study, it looks like it can take the body between seven and 50 days to create antibodies after symptoms begin, with a median of 24 days after symptom onset (and 15 days after symptom resolution) to create them in high quantities.
By the way: That study also tested 719 people who were merely suspected of having coronavirus. That’s people in NYC who either had symptoms and lived with someone who tested positive, or were told by a doctor that their symptoms were consistent with COVID-19. Of those people, only 38% tested positive for antibodies. This is in New York City, a coronavirus hotbed. As we’ve discussed
: Just because you’re around coronavirus cases and felt sick doesn’t mean you had the virus.
How effective are those antibodies in preventing people from getting the disease again?
We know much less about this one. Here’s what we do know:
• No human COVID-19 reinfections have been confirmed
. There would actually be many opportunities for this to have happened by now — think especially of the health care workers who test positive, quarantine and recover, and then are asked to return to duty. So that we haven’t documented a case yet is a good sign.
• Four rhesus macaques — a primate that is one of the closest human analogues
in the animal kingdom — were infected with the virus, allowed to recover, and then reinfected 28 days later. Those four monkeys didn’t test positive for the virus again, nor could researchers find the virus in their lungs.
• There haven’t been any examples of human reinfections of SARS and MERS, the two most closely related coronaviruses. Honestly, since SARS and MERS largely died out pretty quickly, there weren’t many instances to test. Among other coronaviruses — ones that have symptoms more like the common cold — some do have reinfections
and some have a short period of immunity
The WHO is technically right: There isn’t any direct evidence that those with antibodies are protected from a second infection. But the limited circumstantial evidence so far says that the immune response “will offer some protection over the medium term,” as Harvard epidemiologist Marc Lipsitch wrote
in The New York Times.
How long do those antibodies last?
Again, since we still only discovered this thing six months ago, we’re making best guesses based on what we know about other coronaviruses.
The virus that causes COVID-19, SARS-CoV-2, belongs to a category of coronaviruses called betacoronaviruses. Those include SARS-CoV-1, MERS and two other milder but much more common coronaviruses. SARS-CoV-2 fits kind of in the Goldilocks zone — less deadly than SARS and MERS, but much more deadly than the common cold coronaviruses.
For the common but mild coronaviruses, the antibodies declined significantly
after a year. People were able to be reinfected, but the presence of some antibodies still did enough to make symptoms seem milder in a 1990 study.
Meanwhile, for SARS and MERS, the antibodies seemed to last a bit longer: at least two years for SARS
, and three years for MERS
. We don’t know how effective these antibodies were, though, and their neutralization power was declining over this time.
The good news: COVID-19 antibodies look likely to be helpful for at least a year. The bad news: It doesn’t look likely that the immunity is lifelong.
Why are some recovered people testing positive again?
In South Korea, more than 350 people
tested positive well after recovering from a previous COVID-19 diagnosis. This naturally freaked a lot of people out: It could mean that immediate reinfection was possible.
But after analyzing these cases further, researchers believe these cases are due to a more mundane factor: It seems the tests were picking up the dead virus that still remained in the body. Genetic material from the measles virus, for example, can last
about six months. We don’t know how long material from this coronavirus can last, but it’s at least 28 days.
Can people who have already had COVID-19 give the virus to someone else?
Probably not frequently, but maybe. There aren’t any documented or even suspected cases of this, but this one is really hard to research. Everyone who contracts the coronavirus could have gotten it from a number of different sources, and with the incubation times, there’s a lot of potential for mystery here.
As Michael Mina, another Harvard epidemiologist told the health publication Stat
, “We don’t have nearly the immunological or biological data at this point to say that if someone has a strong enough immune response that they are protected from symptoms … that they cannot be transmitters.”
What are the problems with antibody tests?
Despite all this uncertainty about immunity, you may want to get an antibody test anyway. Perhaps knowing you had the disease, plus Lipsitch’s educated guess about a medium-term amount of immune response would give you peace of mind. I understand that.
The problem is that many of the antibody tests we have right now are unreliable. Of the 14 tests examined last month, only three
“delivered consistently reliable results.” Those three delivered false-positive results of 1% or less. Four delivered false positive rates of 11% to 16%. Most of the others delivered false-positive rates of 5% or so.
Most of the tests had even higher false-negative rates than false-positive rates.
But let’s be optimistic about the test you take. Let’s say it has a 1.6% false-positive rate, as Utah-based lab ARUP says its test
has. That’s pretty good.
However, only 6,432 people in Utah have tested positive for the disease as of Tuesday, or 0.2% of Utah’s population. Let’s use our past research as a guide
and say that underestimates the true number of cases by eight times — that’s probably high for Utah given our state’s high rate of testing throughout, but we’ll go with it to make the math nice. That would mean there’s a 1.6% chance that a person selected at random has had the virus.
See the problem? If there’s a 1.6% false-positive rate and 1.6% rate of infection in Utah, then someone who takes the test at random and gets a positive result is equally likely to have either actually had the virus or just be getting a false positive and probably not had it. That means the test is pretty useless for whoever is taking it — and ARUP has a good test!
That doesn’t mean antibody tests are worthless. If you test a ton of people, as the University of Utah is in conducting 10,000 tests
, and take into account the rates of false positives and negatives, you can learn a lot about how the virus spreads. You can make more informed decisions on what to close and what to keep open moving forward.
But for an individual, given all of the uncertainty of the testing itself and what the results would mean, antibody tests aren’t really worth a lot right now.
What does all this mean for herd immunity?
It means herd immunity, relying on the immunity of others to save you from getting the disease, isn’t a slam dunk.
Let’s go with an educated guess, that those who have had the virus have a medium-term immune response that lasts for at least a year, but probably not for life. That means herd immunity, on its own, could cause the contagion rate
of the coronavirus to drop below 1, leading to the end of the pandemic.
Recent studies have given us some optimism about the percentage of a population needed to achieve herd immunity, thanks to the fact that populations have people with different activity levels. One study
, for example, said that if the coronavirus base contagion rate was 2.5, you’d really need only 43% of the population to get the virus to achieve herd immunity. This is probably the best case, so let’s go with it.
Well, that means so many people still need to get the disease for herd immunity to kick in. In Utah, if 1.6% of people have had the disease so far, then about 27 times more people would need to get it than have currently had it. We’ve had 73 deaths so far. That number would rise significantly.
In New York City, about 21% of people have had it based on the latest serology data
— and over 15,000 people have died so far. Doubling the cases probably doesn’t quite double the deaths thanks to advances in treatment, but it’s close.
But troublingly, if the immune response lasts for only a short number of years, that means we could be facing this again relatively soon even if herd immunity is achieved. Or, perhaps more likely, we could face it for an extended period of time while people cycle in and out of immunity. That also ignores the possibility of virus mutation or recovered people still being able to transmit the disease to others.
Essentially, we’re probably going to have to live with the coronavirus for a long time. We should think about what that means with regards to our public policies: How do we take steps to make people safer without jeopardizing our long-term future in other ways? That’s definitely going to be up for political debate.
But regardless of politics, science needs to be a clear priority. After all, a vaccine or effective treatment changes the rules by which the game is being played — it truly can elevate us out of the pandemic and toward more open lives.
The way out isn’t going to be simple. It will be nuanced. But we can handle that, right?
Andy Larsen is a Tribune sports reporter who covers the Utah Jazz. During this crisis, he has been assigned to dig into the numbers surrounding the coronavirus. You can reach Andy at email@example.com or on Twitter at @andyblarsen.