Long COVID Now Has a Framework to Define It

Long COVID now has a working case definition in the U.S.

Twelve key symptoms of long COVID include postexertional malaise, fatigue, brain fog, dizziness, gastrointestinal symptoms, palpitations, changes in sexual desire or capacity, loss of or change in smell or taste, thirst, chronic cough, chest pain, and abnormal movements, said Andrea Foulkes, ScD, of Massachusetts General Hospital and Harvard Medical School in Boston, and colleagues from the NIH’s RECOVER consortium.

Survey data from nearly 10,000 people were used to define the 12 signature symptoms, the researchers reported in JAMA. The findings also included a symptom-based scoring system to help clinicians and researchers better identify long COVID and investigate treatments.

“One of the big takeaways from this study is the heterogeneity of long COVID: long COVID is not just one syndrome; it’s a syndrome of syndromes,” Foulkes said in a statement. “Understanding this idea is a really important step for doing more research and ultimately administering informed interventions.”

“Now that we’re able to identify people with long COVID, we can begin doing more in-depth studies to understand the biological mechanisms at play,” she added.

Since 2020, researchers have documented a wide range of symptoms that emerge after acute SARS-CoV-2 infection. Many studies have been limited by retrospective design, reliance on electronic health record data, or lack of a control group, which has led to disagreement about how common long COVID is, how severe certain symptoms may be, and what patterns define long COVID.

Incorrect case definitions can delay diagnoses, decrease the chance of finding underlying mechanisms, and lead to misdirected, ineffective treatments for long COVID (also known as post-acute sequelae of SARS-CoV-2 infection, or PASC), observed Robert Gross, MD, MSCE, and Vincent Lo Re III, MD, MSCE, both of the University of Pennsylvania in Philadelphia, in an accompanying editorial.

Before developing a case definition for long COVID, it’s important to “consider whether these sequelae represent a single pathophysiologic process or rather multiple different conditions triggered by antecedent SARS-CoV-2 infection,” Gross and Lo Re wrote.

“In addition, it is important to know whether they are direct sequelae of infection itself or are mediated by specific organ injury and dysfunction,” the editorialists continued. “For example, severe SARS-CoV-2 infection that requires intensive care unit admission can result in a well-described post-intensive care syndrome with many features that overlap with post-acute sequelae.”

To classify long COVID, Foulkes and colleagues analyzed data from a symptoms survey distributed at 85 hospitals, health centers, and community organizations in 33 states, Washington, D.C., and Puerto Rico. Participants enrolled in the RECOVER adult cohort before April 2023 and completed a survey 6 months or more after acute symptom onset or test date.

A total of 9,764 participants met selection criteria. Of these, 8,646 people were infected with SARS-CoV-2 and 1,118 were uninfected controls.

Most participants (71%) were female and most were white; 16% were Hispanic or Latino, and 15% were Black. Median age was 47. More than half (58%) of participants were fully vaccinated at the index date. Uninfected participants were more likely to be fully vaccinated (77% vs 55%).

Overall, 37 symptoms had a frequency of 2.5% or greater and an adjusted odds ratio that was 1.5 or greater for infected versus uninfected participants. Symptoms with more than 15% absolute difference in frequencies between infected and uninfected participants included postexertional malaise (28% vs 7%), fatigue (38% vs 17%), dizziness (23% vs 7%), brain fog (20% vs 4%), and gastrointestinal symptoms (25% vs 10%).

Statistical analyses identified 12 hallmark symptoms. The researchers assigned points to each of the 12 symptoms and gave participants a PASC score based on symptom combinations. The proportion with a qualifying PASC score in the full cohort (subject to selection bias) was 23% of infected participants and 3.7% of uninfected participants.

Certain symptoms occurred together, allowing some study participants to be clustered into one of four subgroups. Cluster 1 was characterized by loss of or change in smell or taste; cluster 2 by postexertional malaise and fatigue; cluster 3 by brain fog, postexertional malaise, and fatigue; and cluster 4 by fatigue, postexertional malaise, dizziness, brain fog, gastrointestinal symptoms, and palpitations.

Long COVID was more common and had more severe manifestations in people infected before Omicron. In a subset of 2,231 patients with initial infection on or after Dec. 1, 2021 (when Omicron was circulating), 10% were positive for long COVID at 6 months.

Survey results are not final and will be validated against lab tests and imaging, according to the NIH. The PASC score is operational and needs further refinement, the researchers added. Selection bias in the cohort was likely, they acknowledged, and all symptoms were self-reported.