NIH study suggests coronavirus may have been in U.S. as early as December 2019

A research study run by the National Institutes of Health has turned up evidence of possible coronavirus infections in the United States as early as December 2019, weeks before the first documented infection in this country.

The new report, published in the journal Clinical Infectious Diseases, bolsters earlier studies indicating the virus entered the country under the radar and may have been spreading in the first two months of 2020, well in advance of warnings to that effect from the Centers for Disease Control and Prevention.

A volunteer in Illinois who gave blood on Jan. 7, 2020—in a study unrelated to the emergent virus—tested positive for antibodies to SARS-CoV-2, according to the NIH report. It noted that the antibodies typically take 14 days, on average, to develop, and this “suggests the virus may have been present in Illinois as early as December 24, 2019.”

This and other studies could nudge the timeline of the pandemic’s effect on the United States to earlier dates. The first case of a coronavirus infection in the United States was confirmed Jan. 20, 2020, in a patient in Everett, Wash., who had traveled from Wuhan, China, and had become symptomatic Jan. 14.

But the CDC did not identify community spread of the virus—meaning, infections unrelated to travel from China—until Feb. 26. The NIH report states that the CDC testing guidelines early in the pandemic had a narrow focus: Only people who had been in contact with a person confirmed to have an infection, or who had traveled to an area known to have coronavirus transmission, were advised to be tested.

Elements of that guidance “may have been in place too long, obscuring the geographic spread of SARS-CoV-2 found in our results.”

The volunteers had given blood samples as part of NIH’s “All of Us” research program, a multiyear effort to advance “precision medicine” by gathering detailed health data from a large and diverse group of people. The program has enrolled and collected samples from about 280,000 people so far and has a goal of at least 1 million participants. Precision medicine tailors health care for individuals and their circumstances, rather than a one-size-fits-all approach.

As a result, NIH has a vast supply of blood samples capable of being scrutinized—in this case for evidence of coronavirus infections. The researchers employed two distinct antibody tests on 24,000 subjects who gave blood between Jan. 2 and March 18, 2020.

Nine of those people came up positive on both tests for SARS-CoV-2 antibodies. Seven of those gave blood in five states—Illinois, Massachusetts, Mississippi, Pennsylvania and Wisconsin—before the first official cases in those states.

The report noted briefly that, of the nine people who tested positive, seven were from racial or ethnic minorities: five were Black, two “Hispanic, Latino or Spanish.” The pandemic revealed the heightened vulnerability of people of color in the United States to exposure to the coronavirus in part because of their overrepresentation as essential workers, and although the data set in this study was small, it carried a possible signature of that disparity.

The report states, “Although the numbers are limited, these findings reinforce scientific hypotheses of the impact of social factors on viral circulation, including structural discrimination against racial and ethnic minority groups.”

Another striking finding is that no positive test results were found in California, New York or Washington states, which were known as the initial entry points in the country for the virus.

An earlier study, published in late 2020, also found signs of antibody responses to the novel coronavirus among blood donors in mid-December 2019. The authors of the report hedged on whether it was a true signal of SARS-CoV-2 or an artifact from immunity to other coronaviruses.

Josh Denny, CEO of the All of Us program, said the new report “raises the specter, essentially, that there was community spread” of the virus earlier than previously documented. But he said the data do not confirm that. When the pandemic began, many people across the country reported they had been sick in early 2020 and speculated they had suffered from an undiagnosed case of COVID-19, the illness caused by the virus.

Denny said the new report potentially bolsters the case that some of those people had COVID, but he doubted it was a large number.

“I would suspect that most people who had a cold in early 2020, it wasn’t COVID,” he said. “We found nine people out of 24,000.”

Missing from the report are travel histories of the people involved. It’s unclear if the people with these early potential cases were initially infected in China or had a close contact with someone infected there. The NIH plans to follow up with individuals to try to ascertain where they were infected.

“We don’t know that they didn’t go to Wuhan. We don’t know that they didn’t interact with someone who came from Wuhan,” Denny said.

“It would be good if they could dig up the travel history of the positive individuals, as this will provide a more complete picture,” said Jeffrey Shaman, an epidemiologist at Columbia University, who was not part of the NIH research project. He said the report echoes what he and his colleagues have already concluded through modeling: The virus was present before the first documented cases.

One limitation of the new study is that it cannot rule out that some of the positive test results were false positives.

The earliest result from Illinois, when coronavirus cases were only beginning to spread in Wuhan, China, is the most likely candidate for a false positive, said Michael Worobey, a professor of ecology and evolutionary biology at the University of Arizona, who was not part of the NIH research team.

In an email, he noted that the study looked at a type of antibody, immunoglobulin G, that takes time to develop after an infection. He said he is “very dubious” that there was community transmission in the United States in early January 2020.

But the NIH researchers took steps to limit false positives. The use of two different tests reduced the likelihood. They also studied blood samples from early 2019, long before the pandemic began. No positives were found then.

The team also used a statistical analysis to determine the likelihood that all nine of the positives were false positives. The probability of that was vanishingly low, about 1 in 100,000. However, the analysis showed it is likely that at least one, maybe two, results were a false positive.

Denny said he believed the number of false positives is likely “pretty low.”