Coronavirus infection leads to immunity that’s comparable to a COVID-19 vaccine

A SARS-CoV-2 virus particle
A SARS-CoV-2 virus particle isolated from a patient.
(National Institute of Allergy and Infectious Diseases)

One of the enduring questions of the COVID-19 pandemic is how much immunity people are left with after recovering from a coronavirus infection. New research suggests the level of protection is comparable to getting a vaccine — at least for a few months.

Among a group of hundreds of thousands of Americans who tested positive for a SARS-CoV-2 infection, the risk of developing a subsequent infection more than three months later was about 90% lower than it was for people who had not been previously infected and therefore had no immunity to the virus, according to researchers from the National Cancer Institute.

For the sake of comparison, when the vaccines made by Pfizer-BioNTech and Moderna were tested in Phase 3 clinical trials, they reduced the risk of developing COVID-19 by at least 94%.


The findings, published Wednesday in JAMA Internal Medicine, could help inform plans for returning workers to their offices, sending students and teachers back to school campuses and allowing more of the economy to reopen.

“I think we knew this, that immunity [after natural infection] lasts a long time,” said Dr. Monica Gandhi, an infectious-disease specialist at UC San Francisco who was not involved in the new research. “But it’s still very exciting.”

There are three important things scientists need to know to understand the biological value of coronavirus antibodies, said Dr. Mitchell H. Katz, who leads NYC Health and Hospitals. They are: Do antibodies protect against infection? Can they be reliably detected with current tests? And, if they do offer some protection, how long does it last?

The new study “provides reassuring answers to the first and second questions,” Katz wrote in an editor’s note that accompanied the study.

To investigate coronavirus immunity, the cancer researchers examined the results of more than 3 million blood tests administered to Americans between the start of the pandemic and Aug. 23. A total of 378,606 of those tests were positive for SARS-CoV-2 antibodies — a sign that the person who provided the sample had an active coronavirus infection.

Among the millions of people who were tested, some — about 11% of those who tested positive and 9.5% of those who tested negative — later took a different test to look for evidence of the coronavirus’ genetic material in patient samples, which are typically gathered via the nose, throat or from saliva.

‘A year ago, I was you.’ When cancer survivors wrote anonymous letters to first-time chemotherapy patients, both letter-writers and readers found inspiration.

Feb. 16, 2021

The researchers used these results to see whether people who’d had a coronavirus infection were any less likely than their uninfected counterparts to have SARS-CoV-2 particles in their system. For their analysis, they sorted the results into four groups based on the gap between the antibody test and the genetic test.


After running the numbers, the researchers found that between 3% and 4% of those who originally tested negative for coronavirus antibodies later tested positive with the genetic test. This was true across all four time intervals: 0 to 30 days, 31 to 60 days, 61 to 90 days and more than 90 days. The consistency was probably a reflection of the relatively stable rate at which people in their communities were being infected at the time, the researchers said.

Contrast that with the people who originally tested positive for coronavirus antibodies. Their genetic test results were positive at very high rates in the first 30 days (11.3%), which the researchers said was probably a sign that leftover viral particles were still being flushed from their systems.

However, the positivity rate for the genetic test plunged to 2.7% in the second month after infection, then fell to 1.1% in the third month. And after those 90 days, only 0.3% of people with a past coronavirus infection had another infection that was detected with a genetic test.

That coronavirus infection rate was 10 times lower than it was for the people who presumably had not been previously infected.

That level of protection appears to be comparable to the benefits offered by the Pfizer and Moderna vaccines in their clinical trials, the study authors noted.

“Of course, protection induced by a safe vaccine is clearly preferable,” they were quick to add, “as the population-wide risk of a serious outcome from an authorized or approved vaccine is expected to be orders of magnitude lower than that from natural infection.”

In L.A., so many people now have immunity to the coronavirus from past infections and vaccination that transmission is slowing and inching toward herd immunity.

Feb. 20, 2021

Though the findings may be of great interest to scientists, it’s unlikely they’ll make much practical difference at this stage of the pandemic, Gandhi said. That’s because vaccines appear to be at least as protective as a past infection, and they’re already being rolled out.

These findings “could have been used two months ago,” she added.

Though this study didn’t gauge how long the benefits of a prior infection last beyond 90 days, other lines of evidence suggest it takes a while to wane. For instance, Gandhi pointed to a paper in Science that found that immune cells triggered by SARS-CoV-2 remained in the body for at least eight months.

The findings may help explain why new infections have been falling in Los Angeles County in the wake of a devastating holiday surge. Dr. Roger Lewis, director of COVID-19 hospital demand modeling for the L.A. County Department of Health Services, estimated that about 1 in 3 people in the county now have immunity to the coronavirus.

Still, the study authors said, more research would be necessary to get a clearer picture about natural immunity to the coronavirus.

“Factors that influence reinfection risk — such as varying viral strains, patients’ immune status, or other patient-level characteristics — should be evaluated in subsequent studies that include follow-up beyond 90 days,” they wrote.