A study today demonstrates that last year, about 8% of US adults reported that they ever had long COVID, and those who currently had the condition or currently had activity-limiting long COVID were both under 4%, but a leading US expert on long COVID explains the limitations of such data and why estimates of the prevalence of the condition can vary so widely.
And he notes that, even when considering low estimated long-COVID rates, as in today’s study, when the percentages are applied to the population as a whole, the impact is massive.
Long COVID—also known as post-COVID condition (PCC)—is generally defined as having symptoms 3 months or longer after an acute COVID-19 infection.
Only 3.6% currently had long COVID
The study, published as a research letter in JAMA Network Open by scientists from the Centers for Disease Control and Prevention (CDC), involved 29,522 US adult respondents to the National Health Interview Survey, a nationally representative household survey.
For the survey, investigators randomly selected one adult from each eligible household to answer more detailed questions about their health. The response rate was 47%. The selected adults were asked whether they had ever had COVID-19, if they had symptoms lasting 3 months or more that they didn’t have before COVID-19, if they had symptoms now, and whether these current symptoms limited their ability to carry out daily tasks compared with pre-COVID.
Significant differences across all 3 outcomes were observed by sex, sexual orientation, age, race and Hispanic origin, family income, and urbanization.
The authors didn’t specify what percentage of respondents self-reported COVID-19, but CDC serology data reveal that, by the end of 2023, 87.2% of Americans had infection-induced antibodies to SARS-CoV-2, indicating a previous infection.
The study authors reported that 8.4% of the adults said they ever had long COVID, 3.6% currently had long COVID, and 2.3% currently had activity-limiting long COVID. In 2022, 6.9% said they had ever had long COVID, and 3.4% reported it at the time of the interview. The third question was new in 2023.
They researchers also noted, “Significant differences across all 3 outcomes were observed by sex, sexual orientation, age, race and Hispanic origin, family income, and urbanization.” They said the prevalence of all three outcomes decreased with higher family income and was higher in rural residents.
Study limitations
The study authors write, “A limitation of this work is that the data were from self-reports and were not confirmed by medical evaluation.”
Ziyad Al-Aly, MD, chief of research and development at the VA St. Louis Health Care System and a preeminent expert on long COVID who was not involved in the study, said the low response rate is another limitation of the study. “It is unclear how this low response rate biases the results.” He also said that self-diagnosis “may miss a lot of PCC.”Â
We entered a phase where PCC is now intermingled with baseline disease, and consequently it has become harder to detect through these kinds of surveys.
“I also note lack of any information on infection,” he said in an email. “Because most people abandoned testing, a lot of people get COVID without knowing it; they subsequently develop health problems (that could be PCC) and cannot attribute them to an infection and won’t be able to identify them as PCC.
“We entered a phase where PCC is now intermingled with baseline disease, and consequently it has become harder to detect through these kinds of surveys.”
Why do reported rates of long COVID vary so widely?
The prevalence of long COVID reported in today’s study is markedly lower that other recent studies. Al-Aly and colleagues, in a July study in the New England Journal of Medicine, noted a 10.4% long-COVID prevalence a full year after the original infection when solely assessing people infected with the original SARS-CoV-2 strain. Subsequent variants were tied to lower rates of PCC, but the cumulative total would be much higher than 8.4%.
Last month, Massachusetts General Hospital researchers used artificial intelligence to estimate that 23% of the population has had the condition at some point. A second July study determined that 12% of Japanese adults developed long COVID after an Omicron infection alone. An October study found that 25% of previously healthy US Marines showed signs of long COVID following even mild or asymptomatic COVID-19.Â
A study in September estimated that the prevalence of long COVID in people with disabilities is 41%, compared with 19% in non-disabled participants. And two studies published in August found a 34% prevalence of long COVID in women and an 84% rate 1 year after infection in Italian adults.
Al-Aly explains: “Variation in incidence and prevalence estimates generally stems from variation in study designs (eg, surveys of self-reported Long Covid vs cohort studies, etc), definition of Long Covid (eg, narrow vs broad definition), population being studied, predominant variant, vaccination rates, etc.”
This is a problem that demands attention—we cannot ignore it.”
He added, “When reviewing the totality of evidence, all the estimates suggest that [long COVID] is a serious problem affecting a sizable number of [the] US population. This is a problem that demands attention—we cannot ignore it.”
He also noted that, even when considering the lower percentages in today’s study, when you extrapolate them to the entire US population, the numbers are substantial. “Although it may appear to be a small percentage, 2.3% of people ‘currently had activity-limiting PCC’ is huge when translated into absolute numbers.”
