Estimates of long Covid are startlingly high. Here’s how to understand them

Think about the adults you know who have had Covid: Does 1 out of every 5 have long Covid, as the CDC estimates?

Asking that question should in no way diminish the suffering of people who thought they were done with their infections, only to find their return to well-being still beyond reach. But knowing how many people are living with that bitter legacy of Covid-19, and who among working-age adults can’t work or care for their families, is critical to their care and to the health of our society.

It’s important to remember that long Covid is an evolving umbrella term for an array of symptoms that vary in both number and degree. Some housebound people are assailed by brain fog that completely robs them of concentration, while others find memory aids help them get through their workdays. Some former athletes can’t complete a 6-minute walk test, while others can gradually return to activity if they monitor their heart rate. Long Covid clinics that adapt techniques from rehabilitation medicine see people eventually get better. In a world transitioning away from bustling downtowns to hybrid work-from-home status, we may not see who’s missing.

Whatever long Covid’s toll turns out to be, it will be too many people. However you gather or analyze the data, experts told STAT, the proportion of people whose troublesome, sometimes disabling symptoms linger after their acute Covid-19 infections clear is sizable and worrying. It’s the cruelty of large numbers: Even if the actual prevalence of long Covid is much smaller than recent estimates, a small percentage of a large number is a large number.

And yet, the U.S. has for months been operating in a nearly normal fashion. What could explain this discrepancy between estimates and common experience? It’s eerily similar to the pandemic’s early days, when people asked one another if they knew anyone who had caught the coronavirus, followed more than two years later by the flip side: knowing few people who haven’t been infected and no one who hasn’t been exposed.

Here are some factors that make the current range of estimates easier to understand.

First, what are the numbers?

That 20% figure, from a recent CDC analysis of millions of health records, implies that tens of millions of Americans — a fifth of people infected with Covid — have at least one lingering post-infection symptom that is seriously affecting their daily life. Compared to other estimates, like an April meta-analysis that puts global long Covid at closer to 50% or a June household survey from CDC saying 1 in 3, it’s even on the low side.

Nathan Praschan, a psychiatry researcher at Massachusetts General Hospital, trusts it, calling the more rigorous CDC study’s epidemiology among the best he’s seen because for over a year it used a control group to tease out Covid effects. Still, he thinks it might have missed some people who don’t show up in medical records. Long Covid is defined by symptoms — psychiatric disorders and cognitive problems, to name two — that could make finding care more difficult, as would the same social determinants of health that mean Covid infection is more likely in some populations in the first place. “So, 1 in 5 may be an underestimate.”

What about different definitions?

CDC’s vs. WHO’s, for instance. The CDC defines long Covid, which it calls Post-Covid Conditions, as symptoms lasting four weeks after first infection. The World Health Organization starts the clock ticking after three months. Praschan said it makes sense to be inclusive, as in on the earlier side, while data are still being collected to avoid missing important information from these patients.

There may be differences in the data.

While some U.K. studies relied on records a national health system provides, others culled responses from a smartphone app asking people about their post-Covid symptoms. That limits the respondents to people who have smartphones and are also motivated to report how they are feeling.

The CDC report’s large numbers give power to the analysis, senior epidemiologist Lara Bull-Otterson told STAT. “While all studies have limitations, we believe the strengths of the data and the analysis are solid and are also supported by prior research,” she said. “Future research is always needed to support and expand on the findings of this study.”

Bruce Levy, chief of pulmonary and critical care medicine at Brigham and Women’s Hospital, doesn’t think the 20% estimate is rock solid, noting how studies have varied widely in the U.S. and in other parts of the world. “Even if it’s in single digits at the end of the day, once a formal case definition and a true prevalence study can be accomplished, it’s still a lot of people. But it’s very hard to pinpoint a solid number.”

If the size of the CDC study is impressive, the source of the data has limits, epidemiologist Priya Duggal of Johns Hopkins Bloomberg School of Public Health said. Patient records reflect only the people who sought care and whose symptoms were coded in their charts. Such data don’t include people who didn’t have access to health care, didn’t seek it, or gave up, thinking there was no help for their crazy quilt of symptoms.

“It doesn’t mean the data’s not right. It doesn’t mean that what we’re looking at isn’t important,” she told STAT. “It just means that that’s a different group of people that you might be looking at.”

Even with caveats, she finds the data pretty consistent for a range of 20% to 30% of people experiencing long Covid symptoms “It’s still a substantial number of people. To me, that’s the take-home point,” she said. “The second point is that it’s real.”

Long Covid is a constellation of diseases that manifest differently.

Symptoms linked to long Covid hit bodies from head to toe: brain fog, fatigue, shortness of breath, digestive problems, muscle weakness. The symptoms vary in severity and number, depending on the study. But most patients don’t necessarily have all of them. Some patients don’t have debilitating fatigue, but might report persistent digestive problems they didn’t have before getting Covid.

Some long Covid may be something else.

With long Covid so disparate and common, it’s possible that some doctors are misattributing symptoms to long Covid and missing the diagnosis of a different disease. Or, because lifesaving measures in intensive care units can be like a train wreck for the body, it’s hard to tease out the treatment from the disease.

Some long Covid is hidden to bystanders.

“Some of it is going to be visible like, oh, they’re weak, they’re sickly, they can’t walk, they can’t go upstairs,” Duggal said. “Then there’s also long Covid where you have kidney damage now, and the average person walking down the street doesn’t know that.”

She’s heard people say they don’t know anyone who has long Covid. “I’m like, you probably do.”

Long Covid isn’t all debilitating.

The CDC definitions capture thousands who fit the worst-case image of long Covid: formerly healthy people who can no longer function. But its prevalence estimate also includes anyone reporting at least one symptom, Bruce Walker, director of the Ragon Institute of Massachusetts General Hospital, MIT and Harvard, reminded reporters on a recent call. Estimates may also capture a worsening pre-Covid condition like asthma, an important consideration for the many people with underlying conditions before they caught Covid.

What’s next?

Bull-Otterson of the CDC urged routine screening for long Covid and better defining it so risk factors could be identified and treatments devised. The impact of vaccination and the wild card of variants also need to be understood.

Long Covid has the potential to widen existing gaps in health, Linda Sprague Martinez of the Boston University School of Social Work said on a video call with reporters, pointing to a map of counties with high case numbers but few long Covid clinics. “We don’t want to wait,” she said. “Getting ahead of it will be really important for us,” she said.

OK, what can we say now?

Estimates of long Covid will certainly evolve, and perhaps be refined into the systems they affect: cardiopulmonary, digestive, musculoskeletal, or neurological, including autonomic powers that control breathing, heart rate, and other unconscious functions. If, as experts say, there is an inevitability to catching Covid now, or catching it again, long Covid will likely follow in some proportion of cases, disabling some further fraction of those people. Recent studies suggest that Covid infections precede the risk of certain other chronic diseases like type 2 diabetes, but the mechanism isn’t clear. Even if the world wasn’t ready for one pandemic, it has to deal with its aftereffects somehow.

“We see people still two years out having long-term symptoms, so if that’s true and people can continue to get infected, this is going to be with us for quite a while,” Duggal said.