Not 'little adults': Experts say long COVID undercounted, misdiagnosed in kids

towfiqu ahamed / iStock

Research on long COVID in children is limited, and reported prevalences range widely, from less than 1% to 70%. And while it's a relatively new condition in an evolving field, experts say it could be better defined and measured through well-designed longitudinal studies that take children's unique presentations into account.

"I think it's largely because we're trying to apply adult framework to pediatric problems, and as a result, a lot of things are missed," David Putrino, PhD, director of rehabilitation innovation for the Mount Sinai Health System in New York, told CIDRAP News.

"I think what we need are detailed longitudinal studies where we really take the time to characterize what long COVID looks like in a pediatric population," he said. "That has not been done. What has been done is 'let's treat them just like little adults,' which is always a pitfall in pediatrics."

No consensus on prevalence

In 2022, a systematic review of 22 studies involving children identified a long-COVID prevalence range of 1.6% to 70%, while a study from Germany from the same year that analyzed long-COVID among 157,000 COVID-19 patients suggested that kids were at the same relative risk as adults (30% vs 33%, respectively).

In July 2023, a systematic review of 31 studies published in 2022 involving 15,000 children concluded that 16% of children had persistent symptoms 3 months after infection. Another systematic review from that period found that 1.3% of US children ever had long COVID and that 0.5% currently had it in 2022.

Different symptoms, limited vocabulary

Ziyad Al-Aly, MD, clinical epidemiologist at Washington University in St. Louis and chief of research at the VA St. Louis Health Care System, told CIDRAP News that many children likely aren't diagnosed as having long COVID because they don't recognize or have the vocabulary to report their symptoms.

"Kids don't come home and say, 'Mom, I have postexertional malaise, I have brain fog,'" he said. "What happens is that they start doing poorly in school, and parents find out weeks and weeks later."

Hannah Davis, cofounder of the Patient-Led Research Collaborative, a group of long-COVID patients who are also researchers, said recognition can be particularly difficult in younger, preverbal children.

"Generally, long COVID research in children has been lacking compared with long COVID research in adults," she said. "A very sad thing to me is that I, and the other folks with long COVID as adults, we had this whole life where we understood what it meant to be healthy and active and not have these symptoms, and children don't necessarily, so you see it manifest in different forms."

Kids don't come home and say, 'Mom, I have postexertional malaise, I have brain fog. What happens is that they start doing poorly in school, and parents find out weeks and weeks later.

Ziyad Al-Aly, MD

Indeed, many long-COVID symptoms are subtle and can be misattributed to other conditions, such as anxiety. And unlike the fatigue, brain fog, and postexertional malaise that adults most often report, Putrino said the most common presentation of long COVID in children seen at Mount Sinai's pediatric rehabilitation center is recurrent stomachaches.

"Often, it's not initially thought of as long COVID," he said. "It's just 3 months after a COVID infection, your child's complaining of a tummyache."

Problematic designs, false assumptions

The studies done thus far on long COVID in children have been limited by small sample size, reliance on parental survey rather than interviews, lack of controls, short duration, poor design, and false assumptions about children's viral loads and antibody responses, experts say.

For example, a study published in September in JAMA Pediatrics reported an incidence of long COVID in children aged 8 to 13 years from August 2020 to March 2021 of only 1 of 271 (0.4%). The problem is that they included no data or the symptom questionnaires they asked parents to complete, Putrino said.

There are nonspecific symptoms that kids won't know to report to adults and that parents don't recognize as related to COVID-19. For example, while it may be easier to spot postexertional malaise in older adolescents, "We expect kids to have energy fluctuations," he said. "We expect kids to run themselves ragged and then crash."

The researchers also characterized children as not having long COVID if they had remitting and relapsing symptoms. "We're dealing with a condition where symptoms will wane and wax, and so to exclude kids that had persistent symptoms but then the symptoms went away but then came back as not having long COVID is actually against the clinical case definition of long COVID put forth by both the World Health Organization and the CDC [Centers for Disease Control and Prevention]," Putrino said.

He added that this is one reason why the scientific community is struggling to come up with a good-faith understanding of the rate of long COVID in kids: "Given the quality of the work done so far, I don't think there's any good faith basis for government agencies, educational institutions, or medical institutions to say that your risk of long COVID is very low if you're a child."

Finding true-negative controls getting more difficult

Tired boy asleep by laptop — junpinzon / iStock

In general, long COVID research has also been hindered by high rates of false-negative COVID-19 polymerase chain reaction (PCR) tests and a lack of recognition of the uneven generation of SARS-CoV-2 antibodies (seroconversion) across different demographics, Davis said. "PCRs require a high viral load, and children generally have significantly lower viral loads than adults, and so they just don't get picked up on the PCRs."

Similarly, serologic COVID-19 studies can suffer a lack of recognition that about a third of people don't produce antibodies against SARS-CoV-2, Davis said. "Children make them at a lower rate than adults and lose them at a much higher rate," she said.

And finding true COVID-negative control patients today is more difficult than it was when the pandemic first began. "In March 2020, we had zero immunity," Al-Aly said. "But guess what? It's not that novel anymore."

Other factors that could contribute to the broad spread in long-COVID incidence is the evolution of SARS-CoV-2 variants and their attendant differences, as well as the introduction of COVID-19 vaccines and uptake. "If you did a study on kids who got infected in March 2020 versus today, it's not really comparable," he said.

Toward better studies

To avoid including potentially false-negative people in a control group, Davis said controlled studies could use a prepandemic cohort or electronic health record data rather than a contemporary one.

Otherwise, "you have all these research groups thinking they can do PCR-negative controls, and that ends up putting a ton of people who got COVID into this control group," she said. "So really you're comparing people who had COVID with slightly higher viral loads with people who had COVID with slightly lower viral loads."

Putrino called for studies that follow children from, ideally, before COVID-19 infection, during infection, and then long term.

"Given the intricacy of identifying long COVID in kids, I also strongly recommend that a lot of this work be done with one-on-one interviews, having conversations rather than telling busy parents to check off symptoms on a list," he said. "The other piece that needs to be done is detailed biological profiling of kids who are having persistent symptoms."

In the meantime, he urged vigilance. "My advice leads to infection prevention," he said. "It's really easy to clean air, it's really easy to put HEPA [high-efficiency particulate air] filters in classrooms, it's really easy to use UV [ultraviolet] lights."