A Brazilian city devastated by COVID-19 may have reached herd immunity
Blood donations suggest that up to half of Manaus was infected at the epidemic’s peak
The Brazilian city of Manaus was hit hard by COVID-19. At the peak of its epidemic in late spring, the city of over 2 million people had 4.5 times as many deaths as expected for that time of year. Hospitals and cemeteries struggled to keep up, and mass graves were dug to bury the dead. But then, cases and deaths steadily declined, despite a relaxing of social distancing measures.
That trajectory has prompted some researchers to suggest that Manaus has reached herd immunity. In a report posted September 21 at medRxiv.org that has yet to be peer reviewed, researchers propose that herd immunity developed in the city after 44 to 52 percent of the population was infected at the epidemic’s peak, and that slowed subsequent spread of the virus.
“These are the highest [infection] levels I’ve seen,” says Elitza Theel, a clinical microbiologist at the Mayo Clinic in Rochester, Minn., who wasn’t involved in the study. That high infection rate may very well have impacted the trajectory of the epidemic. “That is how herd immunity works,” she says. “But it comes at a high cost … their death rate was very high.”
Herd immunity occurs when enough people become immune to an infectious disease, either through infection or a vaccine, causing an epidemic to slow down as the pathogen is starved of susceptible hosts (SN: 3/24/20). Scientists are still working out what the herd immunity threshold would be for COVID-19; most estimates are around 40 to 60 percent of a population.
The precise threshold likely varies from region to region, but virtually all of the globe remains well below this threshold, experts say. Most of the United States remains in the single digits, though around 20 percent of the population in parts of New York City may have already contracted the virus.
To investigate whether herd immunity developed in Manaus, researchers from Brazil and the United Kingdom turned to blood donations, looking for antibodies to SARS-CoV-2, the coronavirus that causes COVID-19. Blood donations aren’t a random sample of the population. They tend to come from healthy, asymptomatic adults, and so could miss infections in older people who may be more vulnerable to infection, as well as in kids. Still, the donations offer a way to measure seroprevalence, the proportion of a population that’s been exposed to a virus and has developed antibodies against it.
Researchers tested about 800 to 1,000 blood donations each month from February to August, and attempted to control for potential confounding factors, such as the sensitivity of different tests and the fact that antibodies can wane over time (SN: 4/28/20). They also tested blood donations from São Paolo, another Brazilian city. “This is one of the best papers that I’ve seen that really does try to account waning antibody levels over time” and other factors, Theel says.
In Manaus, the prevalence of antibodies to the coronavirus in blood donors hovered below 1 percent early on in the pandemic, the team found. In April, it rose to 4.8 percent and then rocketed to 44.2 percent in May and reached a peak of 51.8 percent in June, a trajectory that roughly followed the curve of accumulating deaths.
After that peak, seroprevalence fell, reaching 30 percent in August, a consequence of both waning antibody levels in people already infected and lower transmission rates, the researchers say.
While social distancing measures probably helped slow the spread of the virus, the team argues that high population immunity played a bigger role in curbing the epidemic. As of August, the researchers estimate 66 percent of the population had been infected. Whether the city will avoid another outbreak remains to be seen, and will depend in part on how long protective immunity lasts.
The researchers warn that their findings can’t be directly translated to other cites because of differences in factors such as demographics, behavior and adherence to social distancing measures.
Why Manaus may reached herd immunity when other cities haven’t remains unclear. São Paolo for instance, a bustling city of over 12 million, never topped 14 percent seroprevalence despite both cities implementing similar social distancing measures, the analyzed blood donations show. The study authors point to Manaus’ lower socioeconomic conditions, more crowded housing and reliance on boat travel, as factors that could have accelerated the spread of the virus there.
The researchers estimate close to 4,000 people died from COVID-19 in Manaus, a high death toll for a city where only 6 percent of the population is over 60. The city had an infection fatality rate between 0.17 and 0.28 percent, the study suggests. The costs of reaching herd immunity via infection in other cities, especially where there are more older people could be much, much higher. Estimates of São Paolo’s infection fatality rate range as high as 0.72 percent.
Overall, Manaus’ experience reveals “that an unmitigated outbreak will lead to very significant morbidity and mortality,” says Bill Hanage, an epidemiologist at Harvard University, “which is basically what we’ve been saying since February.”