How Many Children Get Long COVID?
Why serious Long COVID is uncommon for adolescents and rare in younger children.
Long COVID has remained a contentious issue for years. These days, the discussions are particularly polarized. On one side you have extremists who believe that the condition basically doesn’t exist, and that people are misleading patients by ascribing their symptoms to COVID-19. On the other, you’ve got a range of advocates who claim that Long COVID is going to quite literally end society as we know it.
As is usually the case, one of the most controversial parts of this discussion is the impact that Long COVID has on children. While most of our estimates seem to show that children quite rarely get severe COVID-19 and don’t usually suffer long-term symptoms, there have recently been claims that this is untrue. A study recently went viral, with Medscape reporting that doctors have revised their estimates of Long COVID in children upwards. According to this new research, a staggering 20% of children suffer from long-term symptoms after catching COVID-19.
Except, the study doesn’t say that. While Long COVID is a genuine issue and a real disease, it’s far less common in children than adults. The true rate of Long COVID in kids is at least an order of magnitude lower than this recent headline suggests.
Let’s look at the data.
The Study
The study in question is a new paper published in JAMA Network Open, an offshoot of the prestigious Journal of the American Medical Association. It’s a report from the RECOVER study, which is a long-term cohort that’s recruiting people who have previously recovered from acute COVID-19 infection to see what happens to them.
This study looked specifically at children aged 6-17. The researchers have been recruiting kids to this cohort since March 2022, and now have a group of 5,376 individuals to study. While the researchers have been recruiting some children who have not had COVID-19, the recruitment strategy - which includes asking the parents of children who are attending pediatric Long COVID clinics if they’re interested in taking part - guarantees that most of the people in this cohort have had COVID-19 and that many of them will have had Long COVID. The researchers also included a group of patients who were already enrolled in a study looking at cognitive development.
The methodology here is quite interesting. The researchers were looking to identify features of Long COVID that could be used to drive future research. Specifically, they were trying to figure out the group of symptoms that best defined Long COVID as distinct from everything else.
This comes back to something I’ve talked about a lot - Long COVID is very poorly defined. In a bid to be as inclusive as possible, we’ve defined Long COVID into a somewhat meaningless term. At the moment, most research studies consider any symptom of any severity that occurs at least 3 (or sometimes 1) month after the onset of COVID-19 as Long COVID. Even if that symptom stops and starts, or is an exacerbation of a pre-existing disease.
But symptoms are very common in the general population. Take this study as an example - about 40% of under-18s who had a COVID-19 infection experienced at least one symptom lasting longer than 3 months. But about 30% of uninfected kids had the same experience. And in this study, the researchers tested for antibodies so we can say with some surety that the uninfected children had genuinely never had COVID-19.
What this study aimed to do was to identify a pattern of symptoms that could be used to better distinguish children with real Long COVID from kids who were simply unwell for other reasons. And using a fairly complex statistical methodology, they appear to have done that quite well. Their final model excluded most of the children who had never been infected with COVID-19, and gave them a useful way to identify likely Long COVID in populations similar to this one in the future.
What this study didn’t do was determine the rate of Long COVID in kids. In fact, the researchers themselves pointed this out in their discussion:
"…the population prevalence of pediatric PASC cannot be determined with the current design because participants with more prolonged symptoms may have been more inclined to enroll.”
If you want to understand the likelihood that people get Long COVID, you need to do several things. But most importantly, you need to ensure that your recruitment is reflective of the general population. If you are, say, recruiting children with Long COVID from Long COVID clinics, you can’t then draw any inferences about children more broadly because you’re biased your sample. The researchers here understood this fact, and said in their study that you cannot possibly use this data to say that 20% of children get Long COVID.
The Medscape article was very simply wrong. Egregiously mistaken.
Long COVID In Kids
So what’s the real rate of Long COVID in children? This is where we get into fairly murky waters. As I’ve discussed before, there are two terms in epidemiology that both can be called the “rate” of a disease - incidence and prevalence. Incidence is the proportion of new diagnoses of a disease over a period of time or a number of events, while prevalence is the proportion of existing diagnoses in a population at a point in time.
At this point in our COVID-19 journey, we don’t really have a good estimate of the incidence or prevalence of Long COVID in children. To understand these numbers, we need large, population-wide studies that measure both the rate of infections in a population and the proportion of people who are infected that go on to experience long-term symptoms. It’s important that the sample is unbiased, which is a problem because usually people experiencing diseases such as Long COVID are more interested in the condition and therefore more likely to respond to surveys asking them questions about it.
The best data we have at this point comes from two UK studies. The UK has, for the entire pandemic, had the best population estimates of COVID-19 and Long COVID in the world. The two estimates come from the Office for National Statistics 2023 report on Long COVID in kids which used data until October 2022, and the REACT study which looked at data until March 2022.
Both of these studies have their biases. While they used extremely good sampling frames - the group of people that they approached to answer their surveys - the two studies also had relatively low response rates. This could bias the results, and we have no way of knowing how. They also different methods of data collection, and asked different questions. However, both have significant strengths - for example, both actually tested people to ensure that they were looking at data from people who had previously had COVID-19. Because of the timeline, there were very few vaccinated children in REACT while there were more in the ONS report. Both studies used either child or parent report.
The REACT study found that 4.4% of children aged 5-11 who had been infected with COVID-19 reported symptoms for longer than 3 months, and 13.3% of adolescents aged 12-17 did. The researchers also found that 2.2% of uninfected 5-11yos and 2.6% 12-17yos had these symptoms, giving an approximate rate of Long COVID of 2.2% in children and 10.7% in adolescents. The study also found that 89% of these long-term symptoms had little to no impact on the kids’ daily lives.
The ONS report found that 1% of children aged 2-15yo reported long-term symptoms 12-20 weeks after infection. This report did not include a comparison group without infection, but they did look at reinfections and found that 0.6% of reinfected children (or their parents) reported long-term symptoms.
Both of these studies are now quite old, but give us some idea about how likely children were to get Long COVID after being infected. There are fairly big differences between the two estimates, which is likely due to methodology, timelines, and age groups examined. Taken together, I think it’s reasonable to say that the risk of Long COVID for mostly unvaccinated kids having their first infection in 2022 was somewhere between 1-2%, and for adolescents closer to 10%. Most of this Long COVID was very mild, with the risk of having significant long-lasting symptoms being more like 0.1% for children and 1% for adolescents.
The problem now is that we know that the per-infection risk of Long COVID drops on reinfection, and after vaccination - this was in fact shown by both of these studies. There’s also some evidence from these papers that the incidence of Long COVID in kids had fallen substantially even from March to October 2022.
It’s hard to know exactly what this risk looks like now, years after this data was collected. We know that theoretically both the incidence and prevalence should’ve fallen substantially, but there’s no good evidence on how much or what those numbers look like now. For one thing, most children have now had COVID-19 at least twice. Many have also been vaccinated.
Long COVID In 2024
If we take the ONS as the most recent estimate - it’s also probably the best scientifically - we could make a reasonable argument that the rate of all Long COVID for children aged 2-15 in 2024 is unlikely to be higher than 0.6%. For severe Long COVID, the number is more like 0.06%. If we take into account the lack of a control group in the ONS study, the numbers might look more like 0.3% and 0.03%.
To put it more simply, based on the ONS data it seems likely that if 1,000 kids get COVID-19 in 2024, 30-60 of them will have a cough, headache, or fatigue that lasts longer than three months. Of those 30-60 children, 3-6 will have significant symptoms that have impacts on their daily life - maybe their headaches are so bad that they miss some days of school, or similar.
These aren’t firm numbers, and I want to make it clear that this is all very uncertain. The true incidence could be much higher, or much lower. That being said, I think based on the data we’ve currently got that Long COVID was always uncommon in children, particularly younger kids, and is now quite rare.
As a father, I know how hard it can be when your children get sick. It’s terrifying. But as a parent with a PhD in epidemiology, a masters of public health, and four years’ experience working on COVID-19, I’m more worried about gastro and RSV when I send my daughter to daycare. Long COVID is certainly still a public health issue, and for people unlucky enough to experience the condition it can be really nasty, but it is no longer something that I factor into my decision-making when it comes to my family.