It is one of many mysteries about long COVID: Who is more prone to developing it? Are some people more likely than others to experience physical, neurological or cognitive symptoms than can emerge, or linger for, months after their coronavirus infections have cleared?
Now, a team of researchers who followed more than 200 patients for two to three months after their COVID diagnoses report that they have identified biological factors that might help predict if a person will develop long COVID.
The study, published Tuesday by the journal Cell, found four factors that could be identified early in a person’s coronavirus infection that appeared to correlate with increased risk of having lasting symptoms weeks later.
The researchers said they had found that there was an association between these factors and long COVID (which goes by the medical name post acute sequelae of COVID-19, or PASC) whether the initial infection was serious or mild. They said that the findings might suggest ways to prevent or treat some cases of long COVID, including the possibility of giving people antiviral medications soon after an infection has been diagnosed.
“It’s the first real solid attempt to come up with some biologic mechanisms for long COVID,” said Dr. Steven Deeks, a professor of medicine at the University of California, San Francisco, who was not involved in the study.
He and other experts, along with the study authors, cautioned that the findings were exploratory and would need to be verified by considerably more research.
Still, Deeks said, “They’ve identified these four major factors. Each is biologically plausible, consistent with theories that other people are pursuing, and importantly, each is actionable. If these pathways get confirmed, we as clinicians can actually design interventions to make people better. That is the take-home message.”
One of the four factors researchers identified is the level of coronavirus RNA in the blood early in the infection, an indicator of viral load. Another is the presence of certain autoantibodies — antibodies that mistakenly attack tissues in the body as they do in conditions like lupus and rheumatoid arthritis. A third factor is the reactivation of Epstein-Barr virus, a virus that infects most people, often when they are young, and then usually becomes dormant.
The final factor is having Type 2 diabetes, although the researchers and other experts said that in studies involving larger numbers of patients, it might turn out that diabetes is only one of several medical conditions that increase the risk of long COVID.
“I think this research stresses the importance of doing measurements early in the disease course to figure out how to treat patients, even if we don’t really know how we’re going to use all that information yet,” said Jim Heath, principal investigator of the study and president of the Institute for Systems Biology, a nonprofit biomedical research organization in Seattle.
“Once you can measure something, then you may be able to start doing something about it,” Heath said, adding: “We did this analysis because we know patients will go to physicians and they’ll say that they’re tired all the time or whatever, and the physician just tells them to get more sleep. That’s not very helpful. So, we wanted to actually have a way to quantify and say that there’s actually something wrong with these patients.”
The complex study had several components and involved dozens of researchers at several universities and centers including the Institute for Systems Biology, the University of Washington and Swedish Medical Center in Seattle, where the study’s lead medical author, Dr. Jason Goldman, is an infectious disease specialist.
The primary group of patients included 209 people, ages 18 to 89, who were infected with the coronavirus during 2020 or early 2021 and were seen at Swedish Medical Center or an affiliated clinic. Many were hospitalized for their initial infections, but some were seen only as outpatients. Researchers analyzed blood and nasal swabs when patients were diagnosed, during the acute phase of their infections and two to three months later.
They surveyed the patients about 20 symptoms associated with long COVID, including fatigue, brain fog and shortness of breath, and corroborated those reports with electronic health records, Heath said.
He said 37% of the patients had reported three or more symptoms of long COVID two or three months after infection. A further 24% reported one or two symptoms, and 39% reported no symptoms. Of patients reporting three or more symptoms, 95% had one or more of the four biological factors identified in the study when they were diagnosed with COVID-19, Heath said.
The most influential factor appeared to be autoantibodies, which were associated with two-thirds of the cases of long COVID, Heath said. Each of the other three factors showed up in about a third of the cases, he said, and there was considerable overlap, with several factors identified in some patients.
The researchers corroborated some of their findings in a separate group of 100 patients, many with mild initial infections, from research led by Dr. Helen Chu at the University of Washington. The researchers also compared their results to data from 457 healthy people.
“The study is large and comprehensive and is a great resource for the community studying long COVID,” said Akiko Iwasaki, an immunologist at Yale, who was not involved in the research.
Dr. Avindra Nath, who is chief of the section on infections of the nervous system at the National Institute of Neurological Disorders and Stroke and was not involved in the study, called the study well designed but pointed out several weaknesses, including the fact that patients had been followed for only two to three months. “This might be too short a time frame,” he said. “Some might just improve spontaneously with time.”
Iwasaki noted that 71% of the patients in the primary group had been hospitalized, limiting the ability to conclude that the biological factors were equally relevant for people with mild initial infections.
One persuasive conclusion, several experts said, was the suggestion that because patients with high viral loads early on often developed long COVID, giving people antivirals soon after diagnosis might help prevent long-term symptoms.
“The quicker one can eliminate the virus, the less likelihood of developing persistent virus or autoimmunity, which may drive long COVID,” Iwasaki said.
That some patients had reactivated Epstein-Barr virus also made sense, Nath said, because other diseases have reawakened that virus, and its reactivation has been linked to conditions like chronic fatigue syndrome, which some cases of long COVID resemble, and multiple sclerosis. Deeks said it might be possible to give antivirals or immunotherapy to patients with reactivated Epstein-Barr virus.
There were other intriguing findings that experts said needed more substantiation. One was a suggestion that because people with lingering respiratory problems had low levels of the stress hormone cortisol, they might benefit from cortisol replacement therapy, which Heath said some doctors were already trying.
In another finding that he said might provide a way to document that patients’ neurological symptoms resulted from long COVID, the blood of people with lingering neurological issues contained elevated levels of proteins associated with disrupted circadian rhythms and sleep/wake cycles.
One patient in the study’s primary group was John Gillotte, 40, a software engineer who contracted the coronavirus in March 2020. He was on a ventilator for about six days, after which he experienced delirium in the hospital when he closed his eyes.
“I saw the devil, who was like 50 feet tall, screaming at me, throwing limbs that he dismembered off of other people,” recalled Gillotte, who later had an image of the demon tattooed on his right arm, with depictions of hell below and heaven above to symbolize his progress from illness to recovery.
Gillotte, who moved from Seattle to Manhattan last year, said that for several months after his infection, he experienced muscle weakness, lack of stamina, brain fog that impaired his concentration at work, an altered sense of smell and the perception that most food tasted like ashes.
He said that before COVID, he had a spontaneous ability to visualize specific colors with certain foods — pink when he sprinkled pepper, blue with a type of liquor — but now, he is dismayed to have lost those automatic connections.
Gillotte said he doesn’t have diabetes and didn’t know if he had the other three factors because researchers said the study protocol prevented them from disclosing data about participants.
Heath, however, noted that Gillotte had been reinfected with the coronavirus in October 2020, which might reflect one theory that emerged from their study: that patients with higher levels of autoantibodies had lower levels of protective antibodies against the coronavirus, possibly making them more vulnerable to reinfection.
Deeks said lower protective antibody levels could also be a pathway leading to long-term symptoms. “If you don’t have a good antibody response, you don’t clear the virus; you have more virus around, and that leads to more long COVID,” he said.
Still, Heath said that overall, the research had showed that the four biological factors intersected and overlapped, suggesting that there might be relatively straightforward ways to forestall long COVID early on. Months later, “all these vague symptoms are so hard to track down, because you’ve sort of lost that information, but if you look back when those symptoms are first triggered, it actually looks like it’s manageable.”