Ziyad Al-Aly noticed that more and more patients were coming to his Missouri clinic with what he describes as an air of invincibility. It was March 2022, and these patients were vaccinated and boosted against COVID-19, and they also had recently recovered from an infection. In their minds, the combined immunity would shield them from further harm.

“I started thinking, is this really true?” says Al-Aly, chief of research and development at the Veterans Affairs St. Louis Healthcare System and a clinical epidemiologist at Washington University in St. Louis. Intrigued, he turned to the VA’s medical records database to find out.

Time has since made clear that there are cracks in the shield created by both natural and vaccine-induced immunity: Reinfection rates are rising with the emergence of the more infectious Omicron variant, with some people even reporting their third or fourth infection. Studies have also shown that the Omicron subvariants are more capable of evading previous immunity. (Why Omicron variants BA.4 and BA.5 are causing fresh U.S. outbreaks.)

But while this epidemiological data shows that COVID-19 reinfections are possible, do they pose any real risk to patients like those Al-Aly treats in his clinic?

“The answer is clearly yes,” he says. In June, Al-Aly published a study, which has not yet been peer reviewed, analyzing more than 5.6 million military veterans’ medical records that showed each new infection added to their risk of dying from just about any cause. It also boosted the odds of other troubling health outcomes from heart, blood, and brain disorders, as well as diseases such as diabetes, chronic fatigue, and long COVID.

But he and other experts caution that many questions remain. We still don’t understand why reinfections were associated with increased risk of these health outcomes among veterans—a population that doesn’t reflect the broader public because it skews older, white, and male. It’s also unclear whether newer COVID-19 variants are more likely to cause severe disease or how long it takes immunity to wane and leave you vulnerable to reinfection.

“When you add all that up, those are a lot of unknowns, and that’s what makes it so complicated,” says Michael Osterholm, director of the Center for Infectious Disease Research and Policy at the University of Minnesota. “I call it an infectious disease calculus problem.”

Here’s what scientists do know so far, and how they’re attempting to solve this complex equation.

The spectrum of harm from reinfections

Reinfection risks vary by disease. For some, such as measles, yellow fever, and rubella, there’s not much reason to worry about a second infection because one bout of the disease or vaccination provides long-lasting immunity. This usually prevents you from getting reinfected in the first place or leads to such a mild infection that it’s unlikely you would notice it.

Then there are diseases where immunity wanes over time, leaving a person more susceptible to reinfections. The severity of that reinfection hinges on many factors, including underlying conditions, changes to the person’s health that might have strained their immune system, the timing of vaccination, and changes to the virus itself.
Take the flu, for example. The flu virus mutates so frequently that it confounds the immune system—each new infection is just like getting hit with the flu for the first time. “So your body can’t say, ‘Oh I’ve seen you before, I know how to deal with you.
As a rule, however, reinfections are typically milder than a first infection. It makes perfect sense: The immune system has been primed. We may have symptoms, but the response is so fast that eventually it controls the replication.”

Dengue fever is one exception. It causes a rare phenomenon in which the immunity from a previous infection can work against you by creating antibodies that inadvertently help the virus invade host cells. There’s no evidence this is the case with SARS-CoV-2—and if that were the case, hospitalizations would likely be skyrocketing right now. But scientists say it has been important to rule that out as one of the paths the virus could take.

What the evidence says about COVID-19 reinfections

It’s clear now that both natural and vaccine-derived immunity against COVID-19 do wane over time. But the severity of those reinfections has been hotly debated.

When Al-Aly’s report came out in June, there was a dust-up on social media about the study, which seemed to suggest that reinfections are more severe than a primary infection. But Al-Aly says that this was a misinterpretation of his findings. Even if most reinfections are milder, he says, they should still be taken seriously.

“The key point here is that it’s not zero risk,” Al-Aly says. He likens it to the aftermath of a house fire. “You can’t say to your spouse, ‘I now know how to put out fire so let’s put the house on fire again,’” he says. “Maybe your immune system is able to deal with it. But you know what’s better? Not to have an infection in the first place.”

Abu-Raddad agrees. His own research, published in the New England Journal of Medicine earlier this month, shows that people who have been vaccinated and had a previous infection are about 97 percent less likely to get a severe, critical, or fatal reinfection. In other words, the risk is “really, really small.” But, he says, each subsequent infection increases a person’s cumulative risk of harm from COVID-19.

Osterholm says that more studies like Al-Aly’s would help develop an understanding of how reinfections might compound the harms of COVID-19. For example, he says, it’s possible that an infection could cause long-term inflammation in the blood vessels that could lead to the development of blood clots that could lead to increased risk for heart attacks or strokes. “That’s an example of something we just need to really get a much better handle on,” he says.

What the evidence says about reinfections and long COVID

Also worrying to scientists is whether each new infection also carries the risk of developing long COVID—a mysterious condition associated with a long list of disparate symptoms that can last months to years after an initial infection. Even though long COVID’s cause remains unknown, scientists hope to learn whether immunity confers some protection against it.

So far, the evidence is mixed. A September 2021 study published in The Lancet found that people who had received two doses of the COVID-19 vaccines were half as likely to develop long COVID symptoms as people who are unvaccinated—suggesting that the vaccines do protect against it somewhat. However, a May 2022 study, also authored by Al-Aly and published in Nature Medicine, suggests that vaccination only lowers the risk of developing long COVID symptoms by about 15 percent.

Meanwhile, Al-Aly’s most recent study suggests that long COVID is more prevalent in people with multiple infections than those who got sick only once. Abu-Raddad points out that doesn’t necessarily mean the second infection is more severe than the first—it might just mean that each new infection is another opportunity for long COVID to strike.

But scientists have to learn what causes long COVID before they can begin to grapple with whether reinfections play a role. Some researchers speculate that long COVID is caused by viral particles that linger in the body well after the acute phase of the disease has ended. Others suggest that it’s caused by a pre-existing autoimmune disorder or perhaps an immune system that didn’t reset properly after a previous illness. Krishna says he would be surprised if a second infection played an amplifying role in causing long COVID. Rather, he thinks it’s more like a roll of the dice. “Each time you roll, it’s a chance you’re going to get a chronic fatigue illness,” he says.

The big questions about reinfections

Scientists need more data before they can draw any conclusions about the severity of COVID-19 reinfections. Al-Aly says his next step is to investigate whether the variants that are dominating now—BA.4 and BA.5—are more likely to cause severe COVID-19 reinfections than others.

Although the VA database is not perfect, he argues that its robust size provides an advantage for untangling the many variables in play: With millions of medical records on file, he’ll be able to analyze reinfection in subgroups of people who only had, say, the Delta variant versus the Omicron variants. Will BA.4 and BA.5 be truly vaccine evasive? Will they lead to the same risk of long COVID? “These are open questions that we’re all itching to know,” he says.

Abu-Raddad would also like to see more studies that examine the clinical features of reinfections. This would be huge undertaking, however. To determine whether reinfections cause compound damage to the body, you would have to conduct a comprehensive screening of each person after each infection.

Ultimately scientists are going to need more time. Although the pandemic may seem long, at two and a half years so far, Krishna points out that it’s still a relatively short time to study how antibodies respond to a virus. In another year and a half, we might find reinfections are much worse—or we might get to a point where we have immunity for life.

Osterholm notes that a variant or subvariant could emerge that would set up a whole new equation. “Every time we’ve tried to out-guess this virus, it has caused us to second-guess ourselves,” he warns. Still, he’s optimistic that scientists will get a better handle on it.

And, in the meantime, experts say there’s plenty that people can do to protect themselves from this uncertainty: get vaccinated or boosted if you’re eligible and take other sensible precautions, such as masking and avoiding situations with a high risk of exposure.

“Anytime we expose ourselves to reinfection, we are playing a very dangerous game,” says Abu-Raddad. “It might be that specific infection that will end up being the very serious one.”