COVID-19 reinfection is likely among unvaccinated individuals, Yale study finds
Researchers at the Yale School of Public Health used a statistical model to assess the durability of natural immunity against COVID-19, finding that reinfection is likely among unvaccinated individuals.
Jessie Cheung, Staff Photographer
A recent study at the Yale School of Public Health analyzed immunological data from close viral relatives of SARS-CoV-2 to determine how long natural immunity from infection lasts for unvaccinated individuals.
Jeffrey Townsend, the study’s lead author and Yale professor of biostatistics, explained that it can take several years to collect enough data to determine the rate of reinfection of an infectious disease. This creates a problem for researchers, as COVID-19 has only been actively circulating for about a year and a half. In the absence of sufficient empirical data, a team of Yale researchers and colleagues at other institutions sought to determine the rate of reinfection by analyzing immunological data from SARS-CoV-1, Middle East respiratory syndrome and human coronaviruses. With an understanding of how these viruses evolve and how they are related to each other, the team was able to model the likelihood of reinfection for SARS-CoV-2, the virus that causes COVID-19.
“The overall goal of the study was to provide an answer to a question that at this point in the pandemic would be impossible to answer empirically, which is how long after you’ve been infected by SARS-CoV-2 can you expect to possess immunity against the virus before you become vulnerable to reinfection?” said Hayley Hassler, a research associate at the School of Public Health and one of the study’s co-authors.
The study showed that the duration of immunity is relatively short. According to Hassler, the risk of reinfection is about five percent at three months after the initial infection. After 17 months, that number increases to 50 percent.
Townsend noted that this result was not unexpected, given that natural immunity against other coronaviruses in humans tends to be short-lived. But one cause for concern is that COVID-19 has more dangerous outcomes during and after infection compared to common colds.
According to Townsend, the model projected the risk of reinfection under endemic conditions — in which everybody has either been infected by or vaccinated against the virus. In these conditions, unvaccinated individuals should expect to be reinfected with COVID-19 every 16 to 17 months on average. On an individual level, this number may vary.
“Our results are based on average times of waning immunity across multiple infected individuals,” Hassler told the News. “Any one of those individuals may experience longer or shorter durations of immunity depending on immune status, cross-immunity, age and multiple other factors.”
Townsend noted that some news coverage has compared SARS-CoV-2 to both the flu and measles, suggesting that these viruses could provide similar types of immunity. These comparisons are misleading, he said, because the viruses are not closely related and cannot be expected to have similar properties. One of the most dangerous misconceptions is the belief that COVID-19 provides life-long immunity, which Townsend emphasized is not the case.
The study looked into six viruses that are close viral relatives of COVID-19, which Townsend mentioned are more useful and “informative” in comparative research because of their common ancestry.
Townsend said he was excited about the study’s use of a data-driven model based on an understanding of evolutionary principles. Although evolutionary biology is typically thought of as a historical discipline, according to Townsend, the team was able to use these theoretical principles to model the reinfection of a relatively new virus.
“This is an instance where we have no idea about the answer to something and the only way that we can obtain that answer is through evolutionary biology,” he said. “And we managed to obtain it, and I’m very confident in the result.”
Sudhir Kumar, a biology professor at Temple University and one of the study’s co-authors, echoed this excitement about the evolutionary methods employed in the study. He said that this same approach could be applied to future pandemics.
Kumar outlined the usage of these methods and their future applications.
“The methods that we used are methods that are standard in the field of evolutionary biology,” he said. “We should be able to use the same approach for predicting durability of immunity in other cases.”
Kumar added that future work on the subject will include a follow-up paper looking at the durability of vaccine-induced immunity against breakthrough infection. He plans to provide an estimate of durability for immunocompromised or immunosuppressed individuals.
Because SARS-CoV-2 is still a relatively new virus, much remains to be understood about the durability of immunity. What this study does reveal, Townsend said, is that natural immunity does not last in the long term and that it is not a substitute for vaccination.
“We need to be very aware of the fact that this disease is likely to be circulating over the long term and that we don’t have this long term immunity that many people seem to be hoping to rely on in order to protect them from disease,” he said.
The study was published in The Lancet Microbe on Oct. 1.