What COVID-19 Reinfection Means For Vaccines

What COVID-19 Reinfection Means for Vaccines
Credit: Tang Ming Tung Getty Images

We now know repeat infections are possible; understanding them will shape the fight against the pandemic

The question of whether we can be reinfected with COVID-19 has been resolved. In August, genome sequencing confirmed that a 33-year-old man in Hong Kong had indeed been infected by the same virus a second time. So too was the case for a 25-year-old man in the United States, though the originating case study has yet to be peer reviewed.

This strengthens earlier reports that have surfaced periodically throughout the pandemic. Most were in China and South Korea, and some in the U.S. None of these were ever verified in laboratory, leaving open the possibility that false positives or negatives had simply confounded test results. A new surge of similarly anecdotal reports is cropping up across Europe and India.

If SARS-CoV-2, the virus that causes COVID-19, follows the precedent set by its coronavirus cousins, reinfection will soon become the rule, rather than the exception. The ability of common-cold-causing coronaviruses endemic to humans to infect the same person multiple times has been documented in studies going back decades and, more recently, in a detailed 2018 study conducted by researchers collaborating in the Netherlands and Kenya.

This body of research, now that we know SARS-CoV-2 also possesses the ability to reinfect, can guide our attempts to answer the questions that remain—namely, how often we can expect to be reinfected, whether a second infection will induce symptoms more mild or severe than what came before, and what this all means for our ability to create a COVID-19 vaccine.


Early evidence of reinfection dates back to series of experiments conducted in the late 1970s and 1980s. Healthy volunteers were exposed to a coronavirus one year, develop a cold, and recover. When exposed again a year later, the cycle repeated for the majority of participants—as many as two thirds, as was the case in a study published in 1990. Some of these are the same coronaviruses we battle year in and year out.

The Kenya study, which incorporated gene sequencing methods, corroborates and elaborates on these findings. Carried out across rural hospitals and households in Kilifi County, Kenya, the study tracked the circulation of multiple coronaviruses throughout the community over a span of six years. Nearly 30 percent of those who caught one variant of the virus once experienced a second infection, while about 10 percent went on to contract it a third time—and at least one person was infected four times. Some were reinfected in as little as three months after their first diagnosis, and for a surprising number of those reinfections, viral load actually increased.

If and when reinfections with COVID-19 become the norm, the majority of people will weather the virus as they would any other cold. You get it, and after a certain period of time your body forgets it, leaving you vulnerable to its return. The key difference is that seasonal cold-causing viruses rarely cause lethal disease, while COVID-19 does at rates of 1–5 percent, depending on the health status of those it affects.

Why reinfection occurs so consistently across the coronavirus family, we don’t know. But what has become increasingly clear is that the defenses we mount against the virus during and after primary infection seem to fade relatively quickly—a disappearing act that doesn’t bode well for our prospects of achieving so-called herd immunity over a longer period of time. I’ve written before that, as a strategy, counting on herd immunity is reckless and ineffective. Immunity, we must remember, isn’t a switch the body can flip on and off at will. It consists of a complex series of reactions and interactions that are difficult to observe and detangle. One set of mechanisms is innate and at work constantly, irrespective of the threat at hand. Another is adaptive, chemically generated at the moment of encounter to attack a specific intruder in specific ways.

A potential explanation of reinfection is that human coronaviruses are adept at tampering with our adaptive immunity, ensuring our long-term response to the virus isn’t as powerful or enduring as it is for many other viruses. Two defenses in particular, killer T cells and B cells (antibody-producing plasma cells), are responsible for sustaining this momentum. When viral infection occurs, a type of antibody called IgM appears within one to two weeks. IgM antibodies mobilize against the virus, then begin to disappear in the months that follow. Two to three weeks after an infection has cleared, IgG antibodies appear.

It is the case for many viruses, including those that cause most childhood diseases, that reasonably high levels of IgG antibodies persist for many years. This is not the case for human coronaviruses. The 1990 study was among the first to monitor antibody levels in addition to reinfection. Although increases were detected within the three weeks following primary infection, within three months they declined sharply. Longer-term studies of recovered SARS and MERS patients also saw the antibody responses dwindled over a two- to three-year period.

Thanks to months of intensive research, we now have a relatively clear picture of the antibody response triggered specifically by SARS-CoV-2. Those who are asymptomatic produce low—sometimes even undetectable—levels of antibodies, even when the virus has been cleared. For the most part, studies show that concentrations of these antibodies diminish rapidly, suggesting that asymptomatic individuals may be the most susceptible to reinfection. Those who fall ill, and particularly those who fall seriously ill, produce greater amounts of antibodies that persist for longer amounts of time—an outcome that would seem, upon first glance, to offer greater protection, too.

The Kenya study on endemic human coronaviruses, however, offers evidence to the contrary that may turn out to be applicable to SARS-CoV-2. In some patients, it was found that high antibody levels actually potentiated infection rather than preventing or mitigating it—leaving open the possibility that no one, regardless of the nature of their primary infection, is totally safe from reinfection. There may also be differences from population to population. While the majority of studies have found that antibodies against SARS-CoV-2 fade over time, one recent Icelandic study, published in the New England Journal of Medicine early this month, found that more than 90 percent of the several thousands of people surveyed still had them four months after their initial diagnosis.


Vaccines, once injected, don’t shield the body from infection. Instead, they equip the immune system with a sophisticated alarm system—one that will trigger a vigorous and rapid response whenever the invading virus sets off the alarm. Immune memory cells, trained by the vaccine to mobilize as soon as the alarm sounds, allow the body to target and clear out the virus in a matter of days, rather than weeks. This process of short-circuiting an intruder’s life cycle is typical—and successful—for many of the vaccinations we receive as children, including measles, mumps and polio. These vaccines mimic a natural immune response, to the effect of developing long-term and in some cases lifelong immunity to reinfection.

The natural immune response to coronaviruses, however, is far more complex. Since the virus does have a well-documented ability to reinfect us, we can infer with reasonable confidence that natural immunity won’t defend us against it in the long term. Based on what we know about our immunity against COVID-19 and coronaviruses at large, there are three questions that developers of COVID-19 vaccines—and those of us who will be queuing up to receive them—must contemplate if we’re to create one that is safe, effective and protective for a long duration.

The first question is how long any immunity, whether natural or vaccine-mediated, will last. The second and more difficult question is whether a strong immune response can, in some, facilitate future infections, and if reinfection does occur, whether it might increase, rather than decrease, the amount of virus in the body. The third and final question concerns the mechanisms by which coronaviruses reestablish infection in a person who has already been infected once before. One possibility is that they inactivate our memory cells—the equivalent of disconnecting the alarm. This is what the measles virus does upon first infection: target and kill memory B cells specifically. For now, whether this is the case for coronaviruses is unknown.

If SARS-CoV-2 doesn’t wipe out memory response upon reinfection, there is more or less a clear path forward for vaccine development. Over time, we may have to create new generations of vaccines because of antigenic drift, as we do with the flu. Aside from the fact that we may have to revaccinate people amid fading immunity, barring any other complications a vaccine will be able to protect us from reinfection. If SARS-CoV-2 it does tamper with our immune memory, however, we might be in trouble. 

There remains much we don’t know about COVID-19 specifically and human coronaviruses at large. What is clear at this moment is that reinfection and the mechanisms that drive it are a key piece of this puzzle—one we can’t leave out, and one that will bedevil our efforts for months and years to come as we struggle to put this genie back in its bottle.

This article originally appeared in the Scientific American on September 24, 2020, and is available online here:What COVID-19 Reinfection Means for Vaccines 

© William A. Haseltine, PhD. All Rights Reserved.