A new long COVID study offers a cautiously hopeful timeline—but also shows why the condition cannot be treated as a single illness. About half of patients with significant symptoms improved within three years of their initial SARS-CoV-2 infection, yet recovery varied dramatically depending on which parts of the body were affected. The study’s central finding is that improvement was common in patients with smell or taste changes but much rarer among people with symptoms spanning multiple organ systems.
Published November 3 in Annals of Internal Medicine, the research followed 1,019 adults enrolled in the NIH-funded RECOVER-Adult Study. Participants had likely long COVID one year after infection and were monitored through in-person visits and recurring symptom surveys. Researchers used the 2024 Long COVID Research Index, or LCRI, which scores problems including fatigue, cognitive difficulties, shortness of breath, heart palpitations, digestive symptoms, and altered smell or taste. An LCRI score of 10 or lower at two consecutive assessments counted as sustained improvement. The published analysis found that 36.2% met that benchmark by 24 months, rising to 50.1% by 36 months. The study’s PubMed record identifies the work as a prospective RECOVER-Adult analysis.
The headline number conceals a sharp divide between symptom patterns. Roughly 60% of participants whose illness was dominated by smell or taste changes improved, compared with about 19% of those whose symptoms crossed several systems, including the cardiac, neurological, and gastrointestinal systems. The RECOVER team also found that complete remission was less common than meaningful improvement: 26.9% of participants no longer reported symptoms captured by the index after three years. RECOVER’s summary of the findings distinguishes sustained improvement from complete remission and emphasizes that the two outcomes should not be conflated.
That distinction matters because long COVID symptoms can fluctuate. A lower score during one period does not necessarily mean a patient will remain symptom-free, and the study was not designed to establish whether people later relapsed. Researchers therefore avoided labeling participants “recovered,” a word that can imply a permanent end to illness. Age and the period of infection also appeared to matter: adults ages 18 to 45 improved more often than older participants, while those infected during the Omicron era had better odds than people infected with earlier variants.
The multisystem group may be especially important for future treatment research because its pattern overlaps with myalgic encephalomyelitis/chronic fatigue syndrome, or ME/CFS. NIH-funded research has found that 4.5% of people in one post-COVID group met ME/CFS diagnostic criteria, compared with 0.6% of people without prior SARS-CoV-2 infection. That NIH analysis adds weight to the idea that at least some long COVID patients may need research and clinical care approaches developed for post-infectious, multisystem disease.
For patients, the results offer neither a guaranteed endpoint nor a reason to dismiss persistent illness. For researchers, they provide a more useful map: trials may need to separate smell-and-taste cases from multisystem cases rather than measure all long COVID patients as one population. RECOVER’s continuing work is intended to support better diagnosis and treatment studies, and its research-index updates are already being used to define which participants meet the threshold for highly symptomatic long COVID. The project’s index documentation describes the scoring system underlying this latest follow-up.
