How Researchers Define a Moving Target Called Long COVID

Two influential efforts now define Long COVID, and they were built for different jobs. In 2024 the National Academies of Sciences, Engineering, and Medicine (NASEM) issued a deliberately broad clinical definition: an infection-associated chronic condition present for at least three months after SARS-CoV-2 infection, affecting one or more organ systems, with no laboratory test required. A year earlier the NIH RECOVER initiative had published a narrower research tool in JAMA, a weighted symptom score that sorts people into "PASC positive" or not. Because both anchor to symptoms rather than a biological marker, the same person can qualify under one and miss the other, which is a large part of why prevalence figures and treatment studies remain so hard to line up.

Two definitions, two purposes

The NASEM definition was requested by federal health offices and released in 2024. Its core sentence is short: "Long COVID is an infection-associated chronic condition that occurs after SARS-CoV-2 infection and is present for at least 3 months as a continuous, relapsing and remitting, or progressive disease state that affects one or more organ systems." The report notes that more than 200 symptoms have been documented across multiple organ systems, and the committee stressed that a positive test is not required. That choice reflects a real problem from the pandemic's early phase, when testing was scarce and many infections went unconfirmed.

The purpose behind that breadth is clinical. As the committee explained in an accompanying New England Journal of Medicine piece, the definition is meant to help patients get recognized, coded, and cared for, and to reduce the kind of gatekeeping that leaves symptomatic people without a diagnosis. It is written to be inclusive on purpose. It is not designed to assemble a tidy, uniform group of subjects for a study.

The RECOVER symptom index

RECOVER set out to solve a different problem: how to identify a group of people distinct enough from uninfected peers that researchers can study mechanisms and test therapies. In the JAMA analysis, investigators compared self-reported symptoms in roughly 9,700 adults across 85 sites, most of whom had prior COVID-19. They found 37 symptoms that were meaningfully more common after infection, then narrowed those to 12 for a weighted scoring index. Symptoms carried different point values based on how strongly they distinguished infected from uninfected participants, with post-exertional malaise and loss of or change in smell or taste weighted among the highest. A total score of 12 or more classified someone as meeting the research definition of PASC.

The authors were explicit that this index is a starting point for research, not a bedside diagnostic test. Among participants first infected in the Omicron era, roughly 10 percent met the threshold at six months, a figure the paper reports with appropriate uncertainty. The score was never meant to tell an individual patient whether they "have" Long COVID.

Why a symptom-based definition complicates the science

Both definitions share a foundational limitation: neither rests on a confirmed biomarker. That has several downstream effects that make the condition behave like a moving target.

Prevalence depends on the ruler

Change the definition and the count changes, sometimes by an order of magnitude. A broad clinical definition that accepts any qualifying symptom for three months will capture far more people than a research index requiring a specific point total. Headlines that seem to disagree are often measuring different things with different instruments, which is why a single trustworthy prevalence number does not exist.

The signal drifts over time

The RECOVER index was derived from a particular population during particular variant waves. Symptom patterns shift as variants change, as vaccination and prior infection accumulate, and as reinfections rise. A rule calibrated on one era can miscount another, so a fixed score is chasing a biology that keeps moving.

Symptoms are common in everyone

Fatigue, brain fog, and gastrointestinal complaints appear in uninfected people too. The RECOVER approach handles this statistically, by focusing on the excess frequency after infection, but at the level of one person the background noise still produces misclassification in both directions.

Trials inherit the ambiguity

Case definitions and endpoints are the scaffolding of any clinical trial. When entry criteria admit a heterogeneous population and the outcome is a symptom score that can drift, a real treatment effect is easy to dilute and hard to reproduce across studies. A therapy that helps one biological subtype can look inert when tested in a mixed group defined only by symptoms. This is the practical cost of defining a disease by what patients feel rather than by what a test can measure.

The trade-off neither definition can escape

The two efforts sit at opposite ends of a familiar tension between sensitivity and specificity. A broad, inclusive definition serves care and captures the true burden of illness, but it groups together people who may have different underlying problems. A narrow, weighted index builds a cleaner cohort for research, yet it excludes real patients and can bias toward a particular phenotype. Both groups say plainly that their definitions are provisional and will be revised as biology, including candidate biomarkers and distinct subtypes, comes into focus. Reading either one as a final verdict misreads what it was designed to do.

For anyone trying to make sense of conflicting numbers, the useful habit is to ask which definition a given study used before comparing it to another. This article is educational and not medical advice.