Timing of Long COVID Improvement in Prospectively Followed RECOVER-Adult Participants
In brief
Half with long COVID improve by three years, but only 27% fully remit
In a prospective U.S. cohort of 1,019 people with likely long COVID one year after infection, about half met the study's improvement threshold by three years, while 27% had no symptoms recorded on the study measure. These observational findings show recovery can occur over time, but cannot identify its causes or replace trials testing treatments.
- Journal
- Annals of internal medicine (Q1)
- Published
- 3 November 2026
- Study design
- Randomized controlled trial
- Evidence level
- Level 2, Moderate (CEBM 2b)
- Authors
- Harrison T Reeder, Sarah E Donohue, Jeffrey N Martin, Mady Hornig, Jarrod M Mosier, Hassan Ashktorab, et al.
- PMID
- 42970801
- DOI
- 10.7326/ANNALS-25-05275
Why clinicians should know about it
- Picked for Epidemiology (top studies of the week, 8 November 2026): Large prospective multisite cohort describes long COVID recovery over time
Abstract
BACKGROUND: Long COVID (LC) is widespread; however, limited information has been reported on the likelihood that symptoms remit over time. OBJECTIVE: To evaluate the likelihood of improvement among persons with likely LC. DESIGN: Prospective, multisite cohort study (RECOVER-Adult [Researching COVID to Enhance Recovery]). SETTING: 83 U.S.-based sites across 33 states; Washington, DC; and Puerto Rico. PARTICIPANTS: 1019 persons in RECOVER-Adult with likely LC (LC Research Index [LCRI] ≥11) 12 months after first SARS-CoV-2 infection. MEASUREMENTS: The primary outcome was time until 2 consecutive study visits with LCRI less than 11. Secondary outcomes were time until LCRI measure reduced 50% or more and quality of life was good or better, and LCRI decreased to 0. Unadjusted estimates of cumulative incidence and adjusted marginal cumulative incidence by age, sex, SARS-CoV-2 wave, and LC subtype were reported across all follow-up time points. RESULTS: An estimated 36.2% (95% CI, 33.0% to 39.5%) met the primary outcome by 24 months after infection, and 50.1% (CI, 46.1% to 54.0%) met the primary outcome by 36 months. An estimated 56.4% (CI, 52.2% to 60.4%) and 26.9% (CI, 23.3% to 30.5%), respectively, met the 2 secondary outcomes by 36 months. Among persons who still reported likely LC at 18 months and 24 months after first infection, relative timing of subsequent symptom improvement was similar. LIMITATION: Findings in this observational cohort are interpretable in a population with prevalent likely LC and do not address causal relationships between individual characteristics and symptom improvement. CONCLUSION: Among persons with likely LC 12 months after initial SARS-CoV-2 infection, one half experience meaningful reductions in symptoms by 36 months, whereas a quarter experience complete remission of symptoms. These findings underscore both the potential for natural recovery and the urgent need for placebo-controlled therapeutic intervention trials. PRIMARY FUNDING SOURCE: National Institutes of Health.
Abstract as published, via PubMed.
For healthcare professionals. The summary is generated by AI from the published abstract, and the evidence level is assigned automatically from the study design on the Oxford CEBM hierarchy. Neither is medical advice. Read the full paper before changing practice.