Why You Shouldn’t Worry About Studies Showing Waning Coronavirus Antibodies


The portion of people in Britain with detectable antibodies to the coronavirus fell by roughly 27 percent over a period of three months this summer, researchers reported Monday, prompting fears that immunity to the virus is short-lived.

But several experts said these worries were overblown. It is normal for levels of antibodies to drop after the body clears an infection, but immune cells carry a memory of the virus and can churn out fresh antibodies when needed.

“Some of these headlines are silly,” said Scott Hensley, an immunologist at the University of Pennsylvania.

Declining antibody levels after the acute infection has resolved “is the sign of a normal healthy immune response,” Dr. Hensley said. “It doesn’t mean that those people no longer have antibodies. It doesn’t mean that they don’t have protection.”

The research also raised some fears about the ability of vaccines to help populations reach herd immunity, the point at which enough people would be immune to the coronavirus to thwart its spread.

It’s too early to know how long immunity to the new coronavirus lasts, and whether people can be reinfected many months to a year after a first bout with the virus. Still, experts said worries about vaccines, too, are unwarranted.

“The vaccine doesn’t have to mimic or mirror the natural infection,” said Shane Crotty, a virologist at the La Jolla Institute for Immunology. “Certainly I wouldn’t be alarmist about these data.”

The new results indicate the prevalence of coronavirus antibodies in the broader population but not in specific individuals. Several studies looking at antibody levels in individuals have shown that after some initial decline, the levels hold steady for at least four to seven months.

The British report is based on three rounds of antibody blood tests carried out in 350,000 randomly selected people from June 20 to Sept. 28. The participants tested themselves at home for antibodies using finger-prick assays that deliver a yes-or-no result, much like a pregnancy test.

Over the three-month period, the proportion of people with detectable antibodies in their blood dropped to 4.8 percent from 6 percent, the researchers reported. The smallest decline was among people ages 18 to 24 and the biggest in those over age 75.

Looking at the data a different way, about 73 percent of people who had antibodies early on still produced a positive result months later, noted Dr. Antonio Bertoletti, a virologist at Duke NUS Medical School in Singapore. “That’s not such a dramatic decline.”

Antibodies also represent only one arm of the immune response, albeit the one that can most easily be measured. There are at least three other branches of the immune system that can fend off illness, so antibody levels don’t present the full picture.

“It’s not the whole immune response,” said Dr. Paul Elliott, an epidemiologist at Imperial College London who heads the project.

When the body encounters a pathogen, it rapidly produces antibodies that recognize the invader. Once the acute infection resolves, the levels decline — as they must for purely practical reasons.

“Our lymphatic system, where immune cells are, only has a finite amount of space,” Dr. Hensley said.

Depending on the test used, the small amount of antibodies still circulating in the blood may not be enough for a positive signal. The test used in the study has a sensitivity of 84.4 percent, well below that of lab-based tests that hover around 99 percent. That means it may miss anyone who has low antibody levels.

For example, people with mild to no symptoms may have produced fewer antibodies than those with severe illness. Most of the people with positive results were ill in March or April, at the peak of the outbreak in Britain, but about 30 percent did not recall having any Covid-19 symptoms. Even a small decrease in the amount of antibodies may drop their levels below the limit of detection.The Coronavirus Outbreak ›

Words to Know About Testing

Confused by the terms about coronavirus testing? Let us help:

  • AntibodyA protein produced by the immune system that can recognize and attach precisely to specific kinds of viruses, bacteria, or other invaders.
  • Antibody test/serology test: A test that detects antibodies specific to the coronavirus. Antibodies begin to appear in the blood about a week after the coronavirus has infected the body. Because antibodies take so long to develop, an antibody test can’t reliably diagnose an ongoing infection. But it can identify people who have been exposed to the coronavirus in the past.
  • Antigen test: This test detects bits of coronavirus proteins called antigens. Antigen tests are fast, taking as little as five minutes, but are less accurate than tests that detect genetic material from the virus.
  • Coronavirus: Any virus that belongs to the Orthocoronavirinae family of viruses. The coronavirus that causes Covid-19 is known as SARS-CoV-2.
  • Covid-19: The disease caused by the new coronavirus. The name is short for coronavirus disease 2019.
  • Isolation and quarantine: Isolation is the separation of people who know they are sick with a contagious disease from those who are not sick. Quarantine refers to restricting the movement of people who have been exposed to a virus.
  • Nasopharyngeal swab: A long, flexible stick, tipped with a soft swab, that is inserted deep into the nose to get samples from the space where the nasal cavity meets the throat. Samples for coronavirus tests can also be collected with swabs that do not go as deep into the nose — sometimes called nasal swabs — or oral or throat swabs.
  • Polymerase Chain Reaction (PCR): Scientists use PCR to make millions of copies of genetic material in a sample. Tests that use PCR enable researchers to detect the coronavirus even when it is scarce.
  • Viral load: The amount of virus in a person’s body. In people infected by the coronavirus, the viral load may peak before they start to show symptoms, if symptoms appear at all.

“We’re saying the antibody response has declined below the threshold” of detection, Dr. Elliott said. “This is not a surprise to anyone who works in the field.”

Data from monkeys suggests that even low levels of antibodies can prevent serious illness from the virus, if not a re-infection. Even if circulating antibody levels are undetectable, the body retains the memory of the pathogen. If it crosses paths with the virus again, balloon-like cells that live in the bone marrow can mass-produce antibodies within hours.

A very small number of people may not make any antibodies. But even those people may have immune cells called T cells that can identify and destroy the virus. The vast majority of people infected with the coronavirus develop lasting cellular responses, according to several recent studies.

T cells are unlikely to prevent infection, but they may at least prevent serious illness by blunting the attack, Dr. Crotty said. Given all that, he said, interpreting low antibody levels to mean that immunity disappears, or that coronavirus vaccines will not be effective is “wrong.”

For example, the human papillomavirus “elicits a terrible immune response and lousy antibodies,” he said. “But the vaccine with a single immunization elicits fantastic antibodies that are 99 percent protective in people for 10-plus years, just a complete night-and-day difference.”

Vaccines can also be designed to provoke much stronger responses than the natural infection, he added.

Though criticizing many of the interpretations of it, experts said the new study’s results are an interesting glimpse into the prevalence of antibodies at a population level.

The same research team is also testing hundreds of thousands of people for presence of the virus. Together, Dr. Elliott said, the studies offer a “really powerful tool” for policymakers to gauge the size of a country’s epidemic.

Should covid be left to spread among the young and healthy?


As new waves of covid-19 sweep the world, lockdowns are back in fashion. This time, though, they are a harder sell. They certainly save lives. But it is now clear that the lost jobs, the disruption to education and medical services, and the harm to mental health that they cause all exact tolls of their own—and these are paid not just in misery, but in deaths. Systems of “test and trace”, intended to stop those exposed to the virus from passing it on, seem to have worked in some places, but not in others.

In the absence of a vaccine, or of effective drug treatments, the question of how much longer this can go on for is thus being asked more insistently. And on October 4th a trio of public-health experts from Harvard, Oxford and Stanford universities put out a petition calling on governments to change course in a radical way.

The Great Barrington Declaration, named after the town in Massachusetts where it was signed, proposes that the contagion be allowed to spread freely among younger and healthier people while measures are taken to protect the most vulnerable from infection. This approach rests on the concept of “herd immunity”, whereby the disease would stop spreading when a sufficient share of the population had become immune as a result of infection.


That is a controversial idea. And on October 14th another group of health experts published a rebuttal in the Lancet, calling the declaration “a dangerous fallacy unsupported by scientific evidence”. Their letter has a grand title, too: the John Snow Memorandum, named after an Englishman who established the principles of epidemiology in the 1850s. It urges governments to do whatever it takes to suppress the spread of sars-cov-2, the coronavirus that causes the illness. In particular, it calls for continuing restrictions until governments fix their systems to test, trace and isolate infected people. Online, the duelling petitions have each gathered thousands of signatures from scientists around the world.

The Great Barrington proposal is a risky one. Any judgment about whether natural infection can create herd immunity to sars-cov-2 is premature. It has not yet been established whether infected people develop durable immunity against reinfection—and if so, how common that immunity might be. Few cases of reinfection have yet been confirmed conclusively. (This is done by establishing that the genomes of the virus particles found the first and second times around are indeed different, meaning the second infection cannot be a continuation of the first.) Lots of reinfections could, though, be happening undetected. About 80% of those infected with sars-cov-2 have mild symptoms, or none at all. The vast majority of these mild cases are not getting tested, even in countries with ample testing capacity.

The ideal study to settle this uncertainty would involve retesting frequently a large cohort of people known to have been infected in the past, to see how many become infected again. But identifying those who have had mild or symptom-free infection is hard. Tests that look for antibodies against sars-cov-2 in big surveillance studies often fail to detect those antibodies in mild cases. Some studies have found that antibodies in these patients wane over time. But whether that equates to waning immunity is still unknown.

If the immune response to sars-cov-2 is anything like that to the other six coronaviruses which infect human beings, letting it spread would eventually slow transmission down—for a period. The question is how long that period would be. Four of the six cause symptoms described as “the common cold” (though other types of viruses cause colds as well). Infection with these confers protection that typically lasts for less than a year. The other two human coronaviruses, sars and mers, cause serious illness. Immunity to these is estimated to last for several years. If protection in the case of sars-cov-2 is short-lived or not particularly strong, the virus will keep surging in recurrent epidemic waves, much as happens each winter with other respiratory bugs. If it is longer-lived, the Great Barrington argument is more plausible.


The authors of the John Snow memorandum argue, though, that deaths and disability under the Great Barrington plan would be huge, even if the herd-immunity gamble is on the money. The share of the population which would need to be infected depends on how easily sars-cov-2 spreads. In its simplest form, the herd immunity threshold as a fraction of the population is 1-(1/r), where r is the average number of people who catch the virus from an infected person. With no social distancing, the r values for Europe are in the range of 3-4, meaning that herd immunity would kick in when two-thirds to three-quarters of people have been infected (see chart 1). This formula, though, assumes everyone has the same chance of infection, which is not the case in reality. If chances of infection vary, then the threshold is lower than the formula suggests. And this may matter. Young people, for example, have more contacts than oldsters, and are thus more likely to pick the virus up. Some models which assume plausible variety in contact rates have concluded that the herd-immunity threshold in western Europe could therefore be as low as 43%.

It is also possible that this threshold has been lowered by pre-existing immunity conferred by past infections with cold-causing coronaviruses. That sort of protection would come from memory t-cells, another part of the immune system’s armamentarium. Unlike antibodies, which are custom-made to attack a given pathogen, t-cells are less picky in recognising and going after a harmful invader. Several studies of blood samples taken before sars-cov-2 emerged have found t-cells that put up a robust reaction to that virus in 20-50% of cases. This is an exciting result. But it is not yet known whether people with such t-cells will have less severe covid-19 disease, or none at all, if they are exposed to sars-cov-2 in real life. An outbreak of covid-19 on a French aircraft-carrier did not come to a halt until 70% of the crew had become infected, which suggests that cross-protection from common-cold infections may just be a nice theory.

All this means that if sars-cov-2 is left on the loose perhaps half or more of people will become infected over the course of six months. The Great Barrington proposal is that, as this happens, countries must double down on protecting the most vulnerable. Identifying who these vulnerable people are is not a foolproof task, but knowledge about the worst combinations of risk factors is getting better. A paper published in the BMJ on October 20th describes a covid-19 risk calculator that predicts an individual’s probability of hospitalisation and death, using data on 6m people in Britain. Validation of this algorithm on 2m others showed that the 5% of people predicted to be at greatest risk by the calculator accounted for 75% of the covid-19 deaths.

But awareness of such risk scores or simpler markers of high risk (old age, obesity and diabetes in particular) is all too often of little use in practice. Most people cannot change their lives in ways that eliminate their risk of infection, particularly when there are lots of infections all around. Those who care for them, or live in the same home, would get infected at some point—and unwittingly pass the virus on. Though most deaths from covid-19 are among the elderly, many adults in younger age groups are at high risk. At the peak of the covid-19 epidemic in England and Wales deaths among people aged 45 to 64 years were 80% higher than usual (see chart 2) despite a lockdown and official advice to the most vulnerable to “shield” from the virus by not leaving their homes at all.

Although the vast majority of people do not get seriously ill if covid strikes, as many as 5% of those who develop symptoms may remain unwell for at least eight weeks (a condition known as “long covid”). Some of them have not recovered after six months, and there are fears that they may never get back to normal. Even if less than 1% of the infected end up in this unlucky group, for a country the size of Britain that would be hundreds of thousands of people with lifelong disability. Another big unknown is whether there are any hidden health consequences of the virus that may show up in the future. Some studies have found subtle heart changes following mild covid-19. It may not be clear for years whether these lead to serious heart problems for some people, or do not matter at all.


The Great Barrington plan, then, is a high-risk, high-reward proposition. The John Snow one, by contrast, would minimise covid deaths in the short term, but lives lost in the long-term, because of lockdowns and other disruptions, might end up being more numerous. Over time, as governments fix the test and trace systems that are needed to replace the broader restrictions, the motivation for the Great Barrington course of action will become less potent.

With luck, this whole debate will be rendered irrelevant by the invention of a vaccine or the development of suitable drugs to treat covid. The results of several efficacy trials of vaccines, and tests on promising pharmaceuticals, are expected in the coming weeks. If covid-19 is less deadly and some herd immunity comes from a vaccine, the paths charted by the two petitions will, eventually, come together.