21 Million Americans Have Had Long COVID - and 6 in 10 Say They Have Recovered

Long COVID has affected tens of millions of Americans. A national survey conducted in 2024 and published in JAMA Network Open estimates that 8.3 percent of U.S. adults - approximately 21.3 million people - reported having experienced long COVID at some point. Of those, nearly 6 in 10 said they had recovered.

That recovery rate is meaningful, particularly in the absence of any treatment that has shown clear efficacy for the condition. It also obscures a substantial remaining burden: the roughly 40 percent who have not recovered, and the disproportionate concentration of persistent cases among adults aged 35 and older.

Consistent Prevalence Across Years

The 2024 figure is consistent with estimates from 2023, aligning with data from the NIH RECOVER initiative and longitudinal data from the Veterans Affairs healthcare system showing similar long COVID prevalence in both years. The stability of the estimate across years is itself informative. It suggests that as some people recover, new long COVID cases continue to occur from ongoing SARS-CoV-2 infections, maintaining prevalence at roughly the same level.

The study, led by Harlan M. Krumholz at Yale University, used nationally representative survey data. Respondents self-reported whether they had experienced COVID-19 symptoms lasting at least three months after infection - the standard definition of long COVID used in most surveillance studies.

Who Is Not Recovering

Adults 35 years of age and older report lasting symptoms at substantially higher rates than younger adults. This age gradient reflects patterns seen across long COVID research more broadly: older age at infection is consistently associated with higher likelihood of developing persistent symptoms and with slower or incomplete recovery.

The biological reasons for the age effect remain under investigation. Older immune systems respond to SARS-CoV-2 infection differently than younger ones, with greater tendencies toward chronic inflammation and less efficient viral clearance. Whether those immune differences drive persistent symptoms directly, or whether they act through intermediate mechanisms such as viral reservoirs or autoimmune activation, is an active area of research.

A Condition Without a Proven Treatment

As of the time of this analysis, no treatment for long COVID has demonstrated clear efficacy in randomized clinical trials. This makes the recovery statistics both encouraging and frustrating: millions of people are recovering, apparently through natural resolution of the condition, but without an intervention clinicians can offer to accelerate or ensure that recovery.

The authors argue that understanding the biological differences between patients who recover and those who do not - particularly any differences in immune cell profiles, viral persistence, or inflammatory markers - could point toward therapeutic targets. Without that mechanistic understanding, treatment development will continue to lag behind the need.

Self-reported surveys have well-known limitations. Whether a respondent defines their own condition as long COVID, and whether they accurately recall when symptoms began, introduces measurement variability. The true prevalence could be higher if some people with persistent post-COVID symptoms do not identify themselves as having long COVID, or lower if some respondents attribute symptoms to other conditions to COVID that pre-dated infection.