Skip to Content
Inside story

This scientist now believes covid started in Wuhan’s wet market. Here’s why.

How a veteran virologist found fresh evidence to back up the theory that covid jumped from animals to humans in a notorious Chinese market—rather than emerged from a lab leak.

November 19, 2021
wet market selling fish
wet market selling fish
Getty
  • Michael Worobey had signed letter asking for more investigation into lab leak theory
  • Now he says natural spillover at the controversial wet market is “vastly more likely”
  • Investigation reveals more of the earliest cases were linked to the market than previously thought

Michael Worobey hasn’t always been certain about where covid originated. During the pandemic, the University of Arizona professor has studied how the virus changes over time, and was among a group of 18 influential scientists who signed a letter in May calling for further investigation to help prove or disprove the theory that SARS-CoV-2 emerged through a possible lab accident.

Now he’s published a new study in Science that suggests that the earliest diagnosed covid case was incorrectly understood—and that Wuhan’s Huanan wet market was almost certainly the site of a spillover of the SARS-CoV-2 virus from animals to humans, rather than a lab leak. His intervention, and growing confidence in the natural spillover idea,  is likely to re-ignite the debate around the hunt for the origins of covid.

Drawing on myriad sources, including peer-reviewed papers, insights from epidemiologists who had access to first-hand information, and media reports, he tried to determine whether bias crept in when clinicians in Wuhan were trying to understand the viral outbreak. 

What he found—not only that there was no obvious bias, but also that many of the first diagnosed cases of covid were either people who worked at the market or lived nearby—has settled his mind that the virus is unlikely to have emerged from a lab leak and that the market was the site of a spillover from animals. 

Here’s how he did it.

"A soft spot for wild theories"

Huanan, a once-bustling market in central Wuhan visited by thousands of shoppers every day, has been the heart of the heated and often acrimonious debate over the pandemic’s origins. Closed as a result of the pandemic, many of the earliest covid cases were linked to the market—but not all of them. 

It was also a notorious site with the potential to be a breeding ground for disease.

In October 2014, officials at the Wuhan Centre of Disease Control and Prevention took Edward Holmes, a virologist at the University of Sydney, to visit the market, where people could buy live animals and have them slaughtered on site. Several restaurants nearby were famous for serving yewei (“wild taste”) from animals that weren’t just freshly killed, but wild. Holmes and his Chinese host strolled through rows upon rows of stalls that sold live animals—snakes, bamboo rats, raccoon dogs—kept in cages stacked on top of one another. It wasn’t a particularly busy day, but the place stank of urine, blood, and feces, Holmes says. 

The market, which used to receive thousands of visitors each day, is near a kindergarten, a couple of shopping centers, and dozens of residential tower blocks. It’s also just half a mile from the Hankou railway station, through which many thousands of people pass each day, peaking at 100,000 during chunyun, the Spring Festival migration, every January. (Both SARS and covid-19 spiraled out of control during chunyun.)

“They showed me the market as a possible place that could trigger future outbreaks,” says Holmes.

"I used to be one of the people who thought [Huanan] was just an amplifying event… we have to look at the totality of evidence."

Michael Worobey, University of Arizona

Huanan’s precise role in the spread of covid-19 has been a point of contention since the beginning of the pandemic. One in three of the 174 patients who fell ill in December 2019 had been to the market, but infectious-disease epidemiologists like Harvard’s William Hanage have argued that it could be a red herring. “When people see clusters of atypical pneumonia, they tend to look for the nearest market. They go ‘Oh, look, market! It must be the market!’” he says. And, he adds, if that assumption was incorrect, it could have led to a cycle which disproportionately emphasized the Huanan’s role, while a large number of cases elsewhere in the city were overlooked. “You tend to look hard at where you expect to find cases and don’t look hard at where you don’t,” says Hanage. 

Worobey—known for having “a soft spot for wild theories”, according to David Robertson, a virologist at the University of Glasgow, UK—has a track record of tackling hotly debated theories around dangerous viruses. In the early 2000s he traveled to the Democratic Republic of Congo to study what became the conspiracy theory that HIV may have crossed into humans as a result of contaminated polio vaccines: He was later part of a group of scientists who helped refute it.

During the pandemic he says he’s been trying to “poke holes” in the natural origins theory by asking if the apparent link of early covid-19 cases to Huanan is real or a mirage. 

That’s why Worobey decided to take up this question of whether so-called “ascertainment bias” had crept in when clinicians in Wuhan were trying to understand the viral outbreak.

From the patchy, fragmented information he could get, Worobey traced how the first 20 covid-19 patients in three hospitals in Wuhan were diagnosed (a total of 27 cases were deemed suspicious by December 30). He found that the clinicians identified cases based on the disease’s clinical manifestation, especially features of their CT scans of the lungs, regardless of their prior exposure at Huanan. It turned out that nine of them were workers at the market, while one patient who had no market exposure had friends who worked there and had visited his home.

This had all happened before public health officials in Wuhan turned their attention to Huanan, and therefore couldn’t have skewed the diagnosis, the study concludes: the market was indeed central to the earliest cases, not a result of doctors seeing more SARS-CoV-2 in the places they spent more time looking.

Worobey also claims that the patient who previously had been thought to have the first documented case of covid-19—and who had no prior exposure at Huanan—may have been mistakenly tagged that way. The WHO had previously reported that a 41-year-old accountant called had been diagnosed with covid-like symptoms on December 8, 2019, making him what is known as the "index case." But according to a Chinese media video report, hospital records Worobey found online, and a scientific paper, the man was diagnosed initially with a dental problem and did not develop covid-19 symptoms until eight days later.

If that individual caught covid later than originally thought, that would mean that a woman named Wei Guixian, who fell ill on December 11, was actually the first documented covid-19 patient. Wei sold shrimp at Huanan.

Solving a key genomic puzzle

Worobey’s detective work also provides vital evidence related to another puzzle about the early spread of the disease.

Previous studies looking at the genome sequences from some of the earliest diagnosed patients show that the virus had already diverged into at least two lineages by December 2019: lineage A, from the earliest reported patients who had never visited Huanan in the weeks before they fell ill, and lineage B, from those who had. Confusingly, lineage A genomes—rather than their Huanan-linked counterparts—seem to be more closely related to the bat relatives which are thought to be their ultimate ancestor. 

Virologists still don’t fully understand if one lineage gave rise to the other, or whether they were siblings, but the differences have cast doubt over whether the contagion actually broke through in Huanan. 

Worobey’s investigation reveals that lineage A genomes came from the first reported cluster of covid-19 cases: an elderly couple and their son. Even though none of them had visited Huanan recently, the couple lived just a few blocks away in the Yangchahu neighborhood and gone to a nearby market there. Live poultry was believed to be for sale at the market, but it’s unclear whether live mammals susceptible to SARS-CoV-2 were sold there. 

The next earliest lineage A genome came from a covid-19 patient who had stayed at a hotel near Huanan just before he fell sick.

Worobey also found that this geographic proximity to Huanan is not limited to the first cluster. In fact, many of the more than 100 covid-19 patients who fell ill in December 2019 with no known epidemiological link to Huanan actually lived in its immediate vicinity.

This, says Worobey, provides strong evidence that community transmission started around the market, and had already been underway when the first covid-19 cases emerged—which could explain why many early cases had no prior exposure at Huanan, especially we now know that many infected people with very mild symptoms or none at all can spread the virus. 

Harvard’s Hanage is yet to be convinced that Wei, the Huanan vendor—rather than the accountant—was the first documented covid patient. It’s not ideal to have to resort to unverified sources in an epidemiological endeavor, he said in an email. “But this is what we have to work with.”  

Nevertheless, Hanage agrees that the new study has provided compelling evidence against possible ascertainment bias. And, he adds, who the index case was does not affect the bigger picture—because given what we know about the virus, “we shouldn’t expect" the first ascertained patients to be the first infected or linked to Huanan. "There is no question the market is the ultimate source of the pandemic.”

Not everybody is convinced, however. Virginie Courtier, a geneticist and evolutionary biologist at the University of Paris, says that the known covid-19 patients might be just a fraction of the total number of actual cases, so Worobey’s study would not exclude some alternative scenarios.

She agrees that it does seem that “there was really something [going on] at the market, but I’m not sure if it was animal-to-human transmission or human-to-human transmission.” It’s still possible, she says, that a contaminated lab member lived around Huanan or went there and caused a superspeader event, but there were none of the telltale signs that were seen in later outbreaks: “We cannot be sure.”

Worobey says there are no telltale signs that the earliest cases came from a super spreading event. It also seems too much of a coincidence that the virus just happened to have hit a relatively small wet market where live mammals susceptible to SARS-CoV-2 were regularly sold—in a way that looks a lot like how SARS originated—when it probably could have hit many other places that more easily could have caused superspreader events.

“We also have to look at the totality of evidence,” he says. 

This includes the fresh revelation that many of the early symptomatic patients worked in the part of the market where—according to a source who did not want to be named to avoid possible political repercussions—live mammals susceptible to SARS-CoV-2 were regularly sold until late 2019. That’s broadly consistent with where environmental samples that tested positive for SARS-CoV-2 were collected, although the source stresses that these connections are rough and need to be scrutinized further.

“I used to be one of the people who thought [what happened at Huanan] was just an amplifying event,” says Worobey. 

Something else that helped sway his opinion was gaining a better understanding of the striking parallels between SARS and covid, which, he says, are only now becoming fully apparent.

Most people think the origin story of SARS is done and dusted. But when Worobey took a close look, he realized that “unfinished business” remains, he says. “It’s often misunderstood and vastly underappreciated.”

A tale of two contagions

In November 2002—almost exactly 17 years before the first covid-19 cases might have emerged—SARS-CoV-1 jumped to wildlife traders through civets, badgers, and raccoon dogs sold in a wet market in China’s southern province of Guangdong. The disease sickened 8,000 people and killed nearly 800. Since then, scientists have established that wet markets provide a perfect environment for viruses to mix, mutate, and reshuffle to create strains that are deadly to humans. 

As a consequence of the role of wet markets in SARS, China tightened the regulation of wildlife trade, requiring vendors to post the source of the animals and their quarantine certificates, and restricting which species were allowed to be traded alive in the marketplace. 

“The penalties for offenders can be harsh,” says Zhou Zhaomin, a policy expert on China’s wildlife trade at China West Normal University in Nanchong. Those trading in protected species can face up to 15 years imprisonment, and smuggling them in or out of China in large enough numbers could result in a life sentence.

But the implementation of the laws was poor. Several researchers told MIT Technology Review that it’s “an open secret” that illegal wildlife trade is rampant in China.

Indeed, Zhou and his colleagues conducted a survey between 2017 and 2019 that found that four markets in Wuhan, including Huanan, sold a combined total of nearly 48,000 wild animals of 38 species, almost all of which were sold alive, caged, and stacked in cramped, unhygienic conditions perfect for virus transmission. The animals—either wild-caught or farmed non-domesticated species—include species susceptible to both SARS-CoV-1 and SARS-CoV-2, such as civets, mink, badgers, and raccoon dogs. 

That study, published in June in Scientific Reports, found that all of the wildlife trade the researchers surveyed was illegal. Many vendors sold protected species; none posted the required certificates indicating the source of the animals or that they were free of diseases.

This means that as soon as Huanan was implicated in early covid-19 cases, vendors selling live mammals, most likely illegally, would run away to avoid facing imprisonment, while law enforcement agencies are unlikely to admit such activities ever existed in the first place. Given this, it was unsurprising that the Chinese authorities found no leads regarding the sales of live animals at the Huanan market, says Harvard’s Hanage. 

Restrictions on the wildlife trade were minimal in the aftermath of SARS, which gave scientists almost unlimited access to animals and traders in Guangdong’s wet markets—but even that wasn’t enough to help them pin down the source of SARS. While they quickly homed in on viruses in civets, badgers, and raccoon dogs that were more than 99% identical to SARS-CoV-1, subsequent investigations did not turn up widespread circulation of the virus, either in the wild or in farmed conditions. A dominant view is that civets got the virus during trading, most likely from bats that were bought and sold at the same time. 

Now, 18 years later, the situation is strikingly similar. There appears to be no widespread circulation of SARS-CoV-2 in animals. None of the 80,000 or so samples tested by the Chinese team of the World Health Organization mission to hunt for the pandemic’s origins—including prime suspects such as pangolins, civets, badgers, and bamboo rats—contained the virus. 

Nevertheless, many scientists still lean heavily toward the theory that wet markets played a critical role in triggering covid-19. Even though all eyes are on Yunnan and other parts of Southeast Asia as the most likely places of the pandemic’s origins, Hanage says “it’s not batshit crazy” to suggest that Wuhan’s Hubei province could have been where SARS-CoV-2 emerged naturally. 

Indeed, scientists at the Wuhan Institute of Virology have found SARS-like coronaviruses in bats in Hubei. Even though they haven’t systematically tested farmed animals for coronavirus infection across the province, in a little-known study conducted in the aftermath of SARS, they found that the seven civets they tested in a farm in the province in 2004 all were infected with relatives of SARS-CoV-1. Several research teams in China and in the US are trying to figure out where the animals got the virus, whether coronavirus infection among civets is more common than previously thought, and what impact that might have on our understanding of the origins of covid-19. 

Constant spillover

But without evidence of an animal infected with a coronavirus that is more than 99% identical to SARS-CoV-2, some scientists have continued to argue against natural origins. 

One such critic is Alina Chan, a molecular biologist at the Broad Institute of MIT and Harvard (this publication is owned by MIT, but editorially independent from it). The central question, she said in a recent webinar organized by Science magazine, is how the virus got to Wuhan from caves more than a thousand miles away in China or other parts of Southeast Asia. “There is a very strong conduit of scientists in Wuhan going down to these places where they [knew] they would find SARS viruses, bringing them all the way into Wuhan city, like thousands of miles,” she said. There is no evidence, however, of such routes for the wildlife trade, she adds. 

Such lack of clarity plagues the origins of SARS as well, says Linfa Wang, director of Duke-National University Singapore’s emerging infectious diseases program. The cave that yielded the closest bat relative of SARS-CoV-1 is nearly 1,000 miles away from the Guangdong market where the first SARS cases emerged—similar to the distance between Wuhan and the site where one of the closest bat relatives of SARS-CoV-2 has been discovered.

And it’s increasingly clear that people in close contact with wildlife are infected by coronaviruses much more frequently than was previously thought. 

"[Huanan] is vastly more likely than other scenarios based on what we now know."

Michael Worobey

Studies show that up to 4% of people who live close to bats and work closely with wildlife in southern China have been infected by deadly animal-borne viruses, including coronaviruses. A Laotian and French team, which discovered the closest relatives of SARS-CoV-2, found that one in five bat handlers in Laos had antibodies against those coronaviruses.

The majority of those spillover infections go extinct of their own accord, researchers say. In a study published in Science in April, Worobey and his colleagues show in computer simulation that for the spillover of SARS-CoV-2 to trigger major epidemics, an urban setting is critical —without that, it would die out very quickly.

“It’s hundreds, if not thousands, of times more likely” that a wildlife trader who was exposed to a SARS-CoV-2 progenitor—either from bats or another animal species—brought the contagion to Huanan than it is that a researcher who went to collect samples from bats came back to Wuhan with the pathogen and then brought it to Huanan, says Wang.

Worobey agrees. Based on many lines of evidence, he is now convinced not only that the pandemic’s connection to the Huanan market is real, but that it is where a SARS-CoV-2 progenitor jumped from an animal to humans. “That’s vastly more likely than any other scenarios based on what we now know,” he says.

Preliminary results from ongoing work by his group and others will help strengthen the case further, he adds: “They all point in the same direction.”

Editor's note: This story has been edited to clarify the identity of the man previously believed to have been the first diagnosed case of covid-19.

Reporting for this article was supported by a grant from the Pulitzer Center.

Deep Dive

Tech policy

thermal image of young woman wearing mask
thermal image of young woman wearing mask

The covid tech that is intimately tied to China’s surveillance state

Heat-sensing cameras and face recognition systems may help fight covid-19—but they also make us complicit in the high-tech oppression of Uyghurs.

German woman stands in queue for vaccination
German woman stands in queue for vaccination

What Europe’s new covid surge means—and what it doesn’t

New restrictions are coming into place across Europe as covid cases rise again. But there are several reasons why a new wave is happening.

surveillance drone in Afghanistan
surveillance drone in Afghanistan

After 20 years of drone strikes, it’s time to admit they’ve failed

The very first drone attack missed its target, and two decades on civilians are still being killed. Why can't we accept that the technology doesn't work?

teenage girls on their phones
teenage girls on their phones

Why Generation Z falls for online misinformation

We can all learn from how today’s young people evaluate truth online.

Stay connected

Illustration by Rose WongIllustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.