A colony of bats silhouetted against a pink sky fly in a column over a golden pagoda on top of a hill in Thailand

People are exposed to bat viruses more often than is realized.Credit: Getty

Every year, tens of thousands of people are silently infected with previously unknown bat coronaviruses related to the one that causes COVID-19, according to a study that maps ‘spillover’ hotspots across southeast Asia.

Bats often host viruses in the same family as SARS-CoV, which causes severe acute respiratory syndrome (SARS), and SARS-CoV-2, the cause of COVID-19. These pathogens often go undetected, although they rarely spread in human populations. But with enough infection events, eventually some viruses will take off, says Stephanie Seifert, a virus ecologist at Washington State University in Pullman. The study, published in Nature Communications on 9 August1, “highlights how often these viruses have the opportunity to spillover”, she says.

Meriadeg Le Gouil, a virologist at the University of Caen Normandy in France, says the work is a good starting point, but notes that “nature is far more complex than a model”.

Spillover hotspots

To calculate the risk of viral spillovers in southeast Asia, the study’s authors first determined the geographic ranges of 26 species of bat known to host SARS-related viruses. In total, the bats’ habitats covered an area of 5.1 million square kilometres, in which close to 500 million people live. Southern China, parts of Myanmar, and the Indonesian island of Java, among other places, were identified as hotspots for human–bat contact, because they host a large number of bat species and have dense human populations.

To estimate how many people are infected with SARS-related viruses in this region each year, the researchers searched the literature for studies that measured the amount of contact between people and bats, and studies that measured the levels of antibodies against SARS-related viruses in people who reported having had contact with bats.

The team calculated that a median of 66,000 people in southeast Asia are infected with SARS-related viruses every year. Most of these viruses are inefficient at replicating and spreading in people, and do not cause illness. But when there are such constant infections “raining down on people, you will eventually get a pandemic”, says co-author Peter Daszak, president of the EcoHealth Alliance in New York City. “We are dodging bullets all the time.”

The article was first posted as a preprint last September.

Targeted surveillance

The analysis of hotspot locations could help researchers to target disease surveillance, catch new viruses in people early and prevent activities that increase the risk of spillovers, such as converting the bats’ forest habitats to farmland. “The article has considerable significance for surveillance,” says Renata Muylaert, a disease ecologist at Massey University in Palmerston North, New Zealand.

It takes a “very elegant and straightforward” approach to quantifying spillover risk, says Timothée Poisot, a computational ecologist at the University of Montreal in Canada. The estimate of the annual median of infections is probably not exact, but it gives a sense of how big the risk of spillovers is, and the study narrows down where the hotspots are, he says.

The high number of people being exposed to these viruses also adds weight to the hypothesis that SARS-CoV-2 had a natural origin, says Daszak, who thinks the evidence supports this explanation over the suggestion that the virus leaked from a lab. The virus almost certainly emerged in bats, but how it came to infect people has been the subject of intensive research and multiple international investigations. Two studies published in Science last month2,3 suggest that the virus probably jumped from animals — bats or other intermediate species — to people on two separate occasions at a live-animal market in Wuhan, China, sparking the COVID-19 pandemic.

Exposure estimates

But other researchers point out that the infection estimate can vary substantially with changes in the assumptions that underlie it. The proposed figures for the amount of contact between bats and people in southeast Asia, and for antibody levels, are particularly unreliable because they are based on around a dozen studies of a broad range of viruses, which could have been overestimates or underestimates, critics say. When Daszak and his team excluded the higher ranges of those estimates, the calculated median number of spillovers was reduced to around 50,000 people infected with SARS-related viruses a year. Daszak says that figure is still substantial. He and his colleagues are now leading efforts to measure antibody prevalence in communities across the region, to validate their approach.

Alice Hughes, a conservation biologist at the University of Hong Kong, also notes that the hotspot maps rely on data about bat geographic ranges collected by the International Union for Conservation of Nature (IUCN), which are of low resolution and possibly out of date, especially in southern China, and could miss some species. Her own research4 has found that southeast Asia is home to dozens of bat species that have not been formally described. “What they are trying to do is very valuable and needs to be done, but it has to be done with more finesse,” says Hughes. Relying on the current maps to inform surveillance will “miss crucially important areas and will sample in other areas of low risk”.

Daszak says the researchers used other data to improve the cruder IUCN figures, but he recognizes that the classification of bat species is constantly being updated. “We have worked with the very best and latest iteration of bat-species designations to do these analyses,” he says.

The model also does not account for the risks of bat viruses infecting people through intermediate species, such as wild animals that are farmed or traded, infected by bats. Daszak says the team did not look at this because data on exposure to potential intermediary animals were limited. However, he says including this possibility would probably have “massively increased the estimated risk of spillovers”.