Name: Long COVID research tracker data, 2026
Creator: Open Source Medicine Foundation
License: https://creativecommons.org/licenses/by/4.0/

Research Tracker
Annual Reports
Long COVID: state of the evidence 2026

Contents:

Summary
Literature
Clinical trials
Agents
Biomarkers
Cohorts
Methods
Limitations
Download the data
Cite

Summary

75studies dated 2026 in the feed

341registered trials matched

107trials recruiting or active

28trials started in 2026

77biomarkers catalogued

7research cohorts indexed

Read this first. The literature feed is a rolling window of the 75 most recent PubMed results for this condition (currently spanning 27 June 2026 to 7 July 2026), not a complete census of 2026 publications. Monthly counts before the window opens are zero for that reason, not because nothing was published. The literature feed was last refreshed on 7 July 2026; figures are as of that snapshot.

Literature indexed in 2026

The Long COVID feed holds 75 PubMed records with a 2026 publication date (75 unique PMIDs) out of 75 records in the feed, drawn from 71 distinct journals. This is a substantial sample.

Records by publication month, 2026. Grey month labels are in the future. Months before the feed window opened are zero by construction.

Top journals

Journals with the most 2026 records in the feed (top 10; ties broken by first appearance).

Browse the live feed: Long COVID research tracker. Weekly changes are in the research digest.

Clinical trials

341 of the 629 ClinicalTrials.gov records in the tracker snapshot (taken 2026-10-07) match Long COVID (341 via the tracker's curated condition mapping). 107 are recruiting, not yet recruiting, enrolling by invitation or active; 28 list a 2026 start date and 127 had their registry record updated in 2026. Median planned or actual enrollment is 59 participants (341 trials report a figure).

By status

Overall status as recorded on ClinicalTrials.gov at snapshot.

By phase

"Not applicable" covers behavioural, device, rehabilitation and observational designs.

By sponsor type

Lead-sponsor class from the registry; "Academic / other" is the registry's OTHER category.

Agents under investigation

Interventions named in the most matched trials (top 10 of 426 distinct agents). A trial testing several agents counts once per agent.

Agent	Trials	Of which active
Paxlovid (Nirmatrelvir/Ritonavir)	8	3
Acupuncture	6	3
Hyperbaric Oxygen Therapy (HBOT)	4	1
Fluvoxamine	3	1
Metformin	3	0
Ritonavir	3	0
Exercise	3	1
Structured Pacing	2	0
Taurine	2	2
Pasc Core	2	0

Trials starting in 2026

NCT	Trial	Status	Phase	Start	Sponsor
NCT07724834	CD19-B Cell Depletion in PAIS-ME/CFS Patients Inebilizumab	Not Yet Recruiting	Phase 2	2026-12-01	Charite University, Berlin, Germany
NCT07697261	Fatigue in Long COVID Lightning Process, Activity Pacing	Not Yet Recruiting	Not applicable	2026-11	McMaster University
NCT07388550	Phase I Open-Label Safety Trial of Pembrolizumab for Neurological Post- Acute Sequelae of SARS-CoV-2 (PD1-PASC I) Keytruda	Not Yet Recruiting	Phase 1	2026-10-12	National Institute of Neurological Disorders and Stroke (NINDS)
NCT07714213	Post-acute Infectious Syndrome - Aripiprazole Symptom Evaluation (PAIS-AriSE) Aripiprazole	Not Yet Recruiting	Phase 2	2026-10-01	Christiana Franke
NCT07682402	Taurine Supplementation in Adolescents With Post-COVID Condition Taurine	Not Yet Recruiting	Phase 1	2026-09-30	University of Alberta
NCT07709234	Transcranial Direct Current Stimulation for Long COVID Brain Fog and Fatigue Transcranial Direct Current Stimulation	Not Yet Recruiting	Not applicable	2026-09-15	University of Las Palmas de Gran Canaria
NCT07798570	Observational Study With CICR-NAM as a Food for Special Medicinal Purposes in Post-COVID Syndrome (PCS) Cicr-Nam	Not Yet Recruiting	Not applicable	2026-09-06	University Hospital Schleswig-Holstein
NCT06907251	Dapagliflozin for Long COVID Syndrome Dapagliflozin	Not Yet Recruiting	Phase 3	2026-09-01	Ottawa Heart Institute Research Corporation

Showing the 8 most recent of 28 trials with a 2026 start date.

All tracked trials: clinical trials tracker.

Therapeutic agents catalogue

The OSMF therapeutic-agents catalogue lists 434 agents with Long COVID as a primary condition (of 605 agents overall). Evidence level is the catalogue's own coarse grading, assigned automatically from the number and type of linked studies and trials; it is not a GRADE assessment.

Catalogue evidence grading: anecdotal, preliminary or moderate. No agent for any condition is graded higher than moderate.

Browse: therapeutics atlas and agents.

Biomarkers catalogued

The Long COVID biomarker atlas (last edited 2026-06-29) holds 77 marker entries across 10 categories, citing 8 distinct DOIs: 50 reported elevated, 14 reduced and 13 with mixed or inconsistent findings. Direction is relative to the comparison group in the cited source (usually healthy controls) and a single entry can rest on one study.

Elevated Reduced Mixed / inconsistent Not stated

Marker entries by atlas category, split by reported direction of change.

Category	Elevated	Reduced	Mixed	Total
Metabolic	11	10	2	23
Inflammatory	11	1	2	14
Vascular	8	0	0	8
Immune	6	0	1	7
Coagulation	6	0	0	6
Proteomic	3	0	2	5
Lipidomic	0	1	3	4
Neurological	4	0	0	4
Microbiome	1	2	1	4
Viral	0	0	2	2

Full atlas: Long COVID biomarker atlas.

Research cohorts

7 of the 32 cohorts in the OSMF post-acute infection syndrome (PAIS) cohort database relate to Long COVID, together enrolling about 29,623 participants where a figure is recorded. This is a curated database, not a systematic census of cohorts.

Cohort	Design	Enrolled	Biobank	Registration
NIH RECOVER-Adult Cohort	prospective non inception	15,000	active	NCT05172024
Long COVID Host Genetics Initiative GWAS	registry linkage	6,450	active	—
Patient-Led Long COVID international cohort (Davis 2021)	survey	3,762	not applicable	—
PHOSP-COVID (UK post-hospitalisation COVID-19)	prospective non inception	2,320	active	ISRCTN10980107
Jinyintan (Wuhan) COVID-19 discharge cohort (Huang)	prospective non inception	1,733	unknown	—
Bergen COVID-19 home-isolated cohort (Blomberg)	prospective inception	312	active	—
Amsterdam Long COVID muscle-biopsy PEM cohort (Appelman)	case control	46	unknown	NCT05225688

Methods

This page was generated by scripts/build_reports.py on 7 October 2026 from the data snapshot described below. No number is entered by hand.

Studies. Records in data/long-covid.json (source: PubMed (NCBI); feed last updated 2026-07-07) whose pub_date parses to a date in 2026. Year-only or year-month dates are mapped to the first day of the period. 0 record(s) had no parseable date and were excluded. Monthly counts bin by publication month; journal counts use the record's journal string verbatim (no normalisation of abbreviations).
Trials. Records in data/clinical_trials/clinical_trials_current.json (snapshot 2026-10-07). A trial matches when its mapped_conditions contains 'Long COVID / PASC'. Status, phase and sponsor type are the registry's own fields, counted once per trial. "Started in 2026" uses start_date; "updated in 2026" uses last_updated. Agents are the tracker's normalised intervention names; a trial with several agents counts once for each. Median enrollment uses the registry enrollment field (actual or estimated) where numeric.
Agents catalogue. Entries in data/therapeutic_agents.json (updated 2026-10-07) whose "Primary Conditions" includes 'Long COVID', grouped by the catalogue's "Evidence Level" string.
Biomarkers. Entries in data/biomarkers/long-covid.json, grouped by category and direction (up, down, mixed). Distinct DOIs count unique reference.doi values.
Cohorts. Entries in data/pais-cohorts-index.json (version 2.0.0-seed) whose pathogen_id is in ['sars-cov-2'].

Limitations

The literature feed is a bounded, rolling PubMed query result. It captures recent indexing activity, not the full year, and PubMed indexing dates lag print publication. Counts here will not match a systematic search.
Study counts measure volume, not quality. No screening for study design, sample size, peer-review status or relevance beyond the search query was applied.
Trial matching depends on the tracker's condition mapping and keyword patterns; trials registered under umbrella terms (for example "post-viral fatigue") may be missed, and registry status fields are frequently stale.
Biomarker direction reflects what each cited source reported, often from a single cohort; "elevated" in one study does not imply replication.
The cohort database and agents catalogue are curated and incomplete, and the evidence-level grading is automatic and coarse.
Different sections rest on different snapshot dates (feed 2026-07-07, trials 2026-10-07, biomarkers 2026-06-29, agents 2026-10-07).
This is a year-to-date report; it is regenerated daily and will change until 31 December 2026.

Download the data

All inputs are public JSON files in the research-tracker repository, licensed for reuse with attribution.

Long COVID literature feed (/data/long-covid.json)
Clinical trials snapshot (all conditions) (/data/clinical_trials/clinical_trials_current.json)
Long COVID biomarker atlas (/data/biomarkers/long-covid.json)
Therapeutic agents catalogue (/data/therapeutic_agents.json)
PAIS cohort database index (/data/pais-cohorts-index.json)
The computed numbers behind this report (/reports/data/long-covid-2026.json)

Suggested citation

Open Source Medicine Foundation. Long COVID: state of the evidence 2026 (year-to-date, generated 7 October 2026). Open Source Medicine Foundation; 2026. https://research.opensourcemed.info/reports/long-covid-state-of-the-evidence-2026.html

For a stable reference, cite the generation date shown; the page is regenerated as data arrive. Weekly updates: subscribe on Substack or follow the digest RSS feed.

Disclaimer. This report is an automated description of a research database maintained by the Open Source Medicine Foundation. It is not a systematic review and not medical advice. Discuss any treatment decision with a qualified clinician.