Skip to content
ModernDoc
Get the app
The research library

COVID-19 research

What the studies mean, the numbers in context, and what remains uncertain.

Stored collection: 145 unique sources across 40 specialties. Updated . Original sources

AI-assisted evidence summaries are not a clinician endorsement or a personal treatment recommendation. For general educational purposes only.

01EXPLORE BY SPECIALTY

Follow your questions.

8 of 40 specialties in view

Long COVID

1 in 12 US adultshave ever had long COVID, per a 2024 national survey

About 1 in 12 American adults say they have had long COVID.

In a national CDC household survey, 8.3% of US adults said in 2024 that they had ever had long COVID, meaning new symptoms lasting 3 months or longer after COVID-19. That share rose from 7.0% in 2022, then leveled off. About 3.3% still had symptoms. Of those, 19.8%, roughly 1 in 5, said long COVID limited their daily activities a lot.

Long COVID has not faded from American life. A steady share of adults still live with it, and it was reported more often by women, middle-aged adults and people with lower incomes, groups that may need targeted support. This is a self-reported survey snapshot, so long COVID was not confirmed by doctors and the survey cannot show what caused it.
37 cited sourcesSource

Cardiology

Nearly 1 in 2seniors hospitalized with COVID who died or had a heart or clot event within a year

Seniors hospitalized with COVID fared worse within a year than flu patients.

Among more than a million Medicare patients (average age 79), 48.8% of those hospitalized with COVID in the Omicron era died or were hospitalized again for a heart attack, stroke, blood clot, heart failure or cardiac arrest within a year. After a flu hospitalization, the figure was 33.4%. Compared with earlier COVID waves, deaths fell, but heart attacks, strokes, clots and heart failure rose.

Omicron is often called mild. But for older adults sick enough to be hospitalized, the following year brought a heavy load of heart and clot problems. Heart-protective follow-up after discharge matters. This was an observational study of Medicare records, so it shows links, not proof that COVID itself caused the extra heart events.
11 cited sourcesSource

Neurology

71 more per 1,000Extra people per 1,000 with a new neurologic diagnosis in the year after COVID-19

COVID-19 was linked to more brain and nerve disorders a year later.

Among US veterans who survived the first month of COVID-19, about 71 more per 1,000 had a new neurologic diagnosis within a year than uninfected veterans from the same period. Problems ranged from stroke and memory trouble to seizures, nerve damage and Guillain-Barré syndrome. Risks rose even in people never hospitalized.

With so many Americans infected, a few dozen extra cases per 1,000 people adds up to a very large number with stroke, memory or nerve problems, and the risk was not limited to the very sick. Based on health records of mostly white male veterans, so results may differ for women and other groups, and the links do not prove COVID-19 was the cause.
13 cited sourcesSource

Pulmonology

About 1 in 36 sparedPer 36 hospital patients given dexamethasone, about 1 more was alive at 28 days

Steroid dexamethasone lowered COVID deaths, most among ventilated patients.

In a large UK randomized trial of people hospitalized with COVID-19, 22.9% of those given the steroid dexamethasone died within 28 days, versus 25.7% of those given usual care. The gain was largest for patients on a breathing machine (29.3% vs 41.4% died). Patients who needed no extra oxygen saw no clear benefit (17.8% vs 14.0%).

The trial showed that calming the body's inflammatory response saves lives once COVID pneumonia is starving patients of oxygen. It also showed timing matters: patients who did not yet need oxygen did not benefit. Open-label trial (everyone knew who got the drug) run in early 2020, before vaccines and other treatments changed who ends up in hospital.
10 cited sourcesSource

Immunology

1 in 5 COVID deathsShare of fatal COVID-19 cases linked to antibodies that disable interferon

Self-attacking antibodies against interferon appear in 18% of COVID deaths.

Interferons are the body's first alarm against viruses. Some people carry autoantibodies, antibodies aimed at their own proteins, that switch this alarm off. They were found in 18% of 1,124 people who died of COVID-19. Among uninfected people, they appeared in 1% of those under 70 but 6.3% of those over 80.

The antibodies were present before infection, not triggered by it, and grow more common with age. That helps explain why some older adults become so much sicker, and suggests a blood test could one day flag who most needs early treatment. The study links these antibodies to fatal COVID-19 and estimates their share of deaths, but it cannot prove they caused any individual death.
6 cited sourcesSource

Vaccinology

55% lower oddsCOVID hospitalization odds with the 2025-2026 shot vs no shot this season, US adults

The 2025-2026 shot was linked to 55% lower COVID hospitalization odds.

In US adults with healthy immune systems seen for COVID-like illness from September to December 2025, getting the 2025-2026 shot was linked to 55% lower odds of COVID hospitalization compared with skipping this season's dose, regardless of earlier shots. Effectiveness was 53% in people 65 and older and 50% against emergency or urgent care visits.

Nearly everyone now has some immunity from past shots or infections, so the real question was whether a fresh shot still adds anything. This early look says it does, for younger and older adults alike. Early observational estimate in adults with healthy immune systems in 7 states, measured soon after vaccination, so it cannot show how long protection lasts or prove cause.
29 cited sourcesSource

Pediatrics

1 in 5 needed the ICUOf babies under 6 months hospitalized with COVID, not counting birth stays

Young infants have COVID hospital rates rivaling older adults, often needing ICU.

From October 2022 to April 2024, CDC surveillance counted 1,470 COVID hospital stays among US babies under 6 months, who are too young for vaccines. Their hospitalization rates topped every other age group except adults 75 and older. Of babies hospitalized after their birth stay, 22.1% went to intensive care and nine died.

Many parents think COVID is mild for every child. For the youngest babies, the hospital risk looks more like a grandparent's, and these infants depend on protection from others, including antibodies passed on during pregnancy. This surveillance covers selected US counties and counts only hospitalized babies, so it cannot show how often infected infants overall become seriously ill.
15 cited sourcesSource

Public Health

1.87 years lostDrop in US life expectancy, 2019 to 2020, vs 0.58 years on average in 21 peer nations

US life expectancy fell far more in 2020 than in peer nations.

Using official death counts, researchers found US life expectancy at birth fell 1.87 years from 2019 to 2020. Across 21 other high-income countries the average drop was 0.58 years, and none fell as far as the US. Hispanic Americans lost 3.70 years and Black Americans 3.22 years, far more than White Americans.

Life expectancy sums up a whole population's health in one number. The 2020 drop widened an existing gap between the US and other rich countries, and it fell hardest on Hispanic and Black communities. This covers only 2020; later federal data show US life expectancy has since rebounded above its pre-pandemic level.
15 cited sourcesSource

View the full specialty index →
Summaries are returning as the citation review progresses. Open a specialty for context and original sources.

02ORIGINAL RESEARCH

Explore the evidence.

Newly added study findingsApproved explanations; dates describe additions to this collection, not publication.
THE RESEARCH, AS IT ARRIVES

Latest studies.

The newest findings added to this collection.
The learning from each, already distilled.

  1. STUDY 01Findings added
    Immunology
    TL;DR

    New autoimmune diagnosis rates were higher after COVID-19

    15.05 vs 10.55 per 1,000 person-years

    In a matched German health-record study, new autoimmune diagnoses occurred at 15.05 per 1,000 person-years after COVID-19 versus 10.55 among people without COVID-19.

    The comparison covered 2020 infections and diagnoses recorded 3–15 months later. This matched health-record association does not establish that infection caused the diagnoses.

    Study details & original results

    Tesch F et al. Incident autoimmune diseases in association with SARS-CoV-2 infection: a matched cohort study. Clinical Rheumatology. 2023. [Source notice: Correction published 2023-07-05 (10.1007/s10067-023-06692-8): figures 1 and 2 images were interchanged; legends unchanged.]

    New autoimmune diagnoses 3–15 months after COVID-19 in a matched German health-record study of 2020 infections, compared with people without COVID-19

    15.05 vs 10.55 per 1,000 person-years
    14.69–15.42 vs 10.25–10.86

    Citation in Immunology
    One key learningRead original study in a new tab
  2. STUDY 02Findings added
    Rheumatology
    TL;DR

    Infected lupus patients had higher hospitalization odds than healthy controls

    aOR 6.13

    Infected participants with systemic lupus erythematosus had higher adjusted odds of COVID-19 hospitalization than infected healthy controls (aOR 6.13).

    Cross-sectional Chinese study conducted January–February 2023; 95% CI 3.06–12.27. Adjustment included age, sex, comorbidities and vaccination. Odds are not absolute risk or proof of causation.

    Study details & original results

    Zang, Sitian et al.. Mycophenolate mofetil associates with COVID-19 hospitalization in patients with systemic lupus erythematosus: a nationwide population-based study from China.. Clinical Rheumatology. 2026.

    Adjusted odds of COVID-19 hospitalization in infected Chinese participants with systemic lupus erythematosus versus infected healthy controls, from a cross-sectional study conducted January–February 2023; adjusted for age, sex, comorbidities and vaccination status.

    aOR 6.13
    95% CI: 3.06–12.27

    Citation in Rheumatology
    One key learningRead original study in a new tab
  3. STUDY 03Findings added
    Genetics
    TL;DR

    No persistent blood-methylation difference was detected between groups

    No clear difference

    Paired samples from 297 adults showed no significant persistent group differences in whole-blood DNA methylation or epigenetic-aging measurements.

    The Long COVID, recovered and not-infected groups contributed 594 samples total, collected in 2020 and 2023. This assay-specific null does not rule out smaller effects or changes in other tissues.

    Study details & original results

    Bohlin, Jon et al.. Longitudinal study of genome-wide DNA methylation in individuals with and without post-acute symptoms following SARS-CoV-2 infection.. Epigenomics. 2026.

    Persistent whole-blood DNA methylation and epigenetic-aging measurements in a longitudinal comparison of Long COVID, recovered COVID-19 and not-infected groups; paired samples from 297 adults (594 samples total) collected in 2020 and 2023.

    No significant persistent group differences detected

    The reported result does not establish a difference; it does not prove equivalence.

    Citation in Genetics
    One key learningRead original study in a new tab
  4. STUDY 04Findings added
    Dermatology
    TL;DR

    Hair shedding began weeks after a positive test in selected cases

    Among COVID-associated telogen-effluvium cases, hair shedding began an average 53.76 days after a positive SARS-CoV-2 PCR test.

    Cases came from a selected 204-patient cohort with infection in the preceding three months; study published in 2021. The onset average is for the hair-shedding cases, not all participants or the wider population.

    Study details & original results

    Aksoy, Hasan et al.. COVID‐19 induced telogen effluvium.. Dermatologic Therapy. 2021.

    Onset timing among COVID-19-associated telogen-effluvium cases within a selected 204-patient cohort with SARS-CoV-2 infection in the preceding three months (study published in 2021); not population prevalence.

    Mean 53.76 (± 23.772) days after SARS-CoV-2 RT-PCR positivity

    Citation in Dermatology
    One key learningRead original study in a new tab
  5. STUDY 05Findings added
    Pathology
    TL;DR

    Autonomic nerve density was lower despite preserved epidermal fiber density

    In the 23-patient painful Long COVID subgroup with normal intraepidermal nerve-fiber density, autonomic fiber density around piloerector muscles and sweat glands was lower than in healthy controls.

    This was a selected retrospective biopsy subgroup. The group comparison does not establish diagnostic accuracy, a causal mechanism or a treatment response.

    Study details & original results

    Falco, Pietro et al.. Autonomic small fiber involvement in painful long COVID: a histological and clinical study.. Frontiers in Human Neuroscience. 2026.

    Autonomic piloerector-muscle and sweat-gland nerve-fiber density compared with healthy controls in the 23-patient painful Long COVID subgroup with normal intraepidermal nerve-fiber density; selected retrospective biopsy study.

    Reduced PMNFD (p = 0.0003) and SGNFD (p = 0.0005) despite preserved IENFD

    Citation in Pathology
    One key learningRead original study in a new tab
  6. STUDY 06Findings added
    ME/CFS
    TL;DR

    ME/CFS and Long COVID showed different immune-cell activation patterns

    An exploratory cross-sectional study found different monocyte and dendritic-cell activation patterns in ME/CFS and Long COVID groups.

    The comparison included 103 ME/CFS participants, 63 Long COVID participants and 41 healthy controls. Group immune profiles are not a validated diagnostic test or proof of different causal mechanisms.

    Study details & original results

    Petrov, Steliyan et al.. Comprehensive Immunophenotyping of Monocytes and Dendritic Cells Suggests Distinct Pathophysiology in Chronic Fatigue Syndrome and Long COVID.. International Journal of Molecular Sciences. 2026.

    Exploratory cross-sectional immune-cell comparison of 103 ME/CFS participants, 63 Long COVID participants and 41 healthy controls; group profiles are not a validated diagnostic test.

    Different monocyte and dendritic-cell activation patterns in ME/CFS and Long COVID

    Citation in ME/CFS
    One key learningRead original study in a new tab
  7. STUDY 07Findings added
    Hepatology
    TL;DR

    No clear liver-stiffness change at three-year follow-up

    No clear difference

    Ultrasound liver stiffness showed no statistically significant change from 3–6 months after COVID-19 to three-year follow-up.

    The COVID-19 research group included 90 patients at baseline; 58 had follow-up. This group-level null does not prove absent fibrosis or rule out individual liver injury.

    Study details & original results

    Radzina, Maija et al.. Multiparametric Ultrasound Assessment of Long-Term Liver Damage in COVID-19: Results of a Three-Year Follow-Up Study.. Medicina. 2026.

    SWE liver-stiffness change in the COVID-19 research group (90 patients at baseline; 58 had follow-up), from the assessment 3–6 months after COVID-19 to three-year follow-up in a longitudinal prospective study.

    No statistically significant change in SWE liver stiffness (p > 0.05)

    The reported result does not establish a difference; it does not prove equivalence.

    Citation in Hepatology
    One key learningRead original study in a new tab
  8. STUDY 08Findings added
    Biomarkers
    TL;DR

    CORIS classified illness severity at hospital admission

    CORIS distinguished severe or critical COVID-19 from non-severe illness at hospital admission better than a lymphocyte-subset model and a routine-laboratory model in its internal validation set.

    This single-centre study used internal data splitting, without external validation. The model was not intended to diagnose infection or predict later deterioration, survival or blood-clot events.

    Study details & original results

    Zhang, Yihe et al.. Integrating Anti-SARS-CoV-2 Spike Protein IgG, IL-6, and NK Cell Percentage with Routine Laboratory Markers for COVID-19 Severity Assessment: Development and Validation of the CORIS Nomogram.. Diagnostics (Basel, Switzerland). 2026.

    Internal split-sample performance of the age, LDH, anti-spike IgG, IL-6, and NK-cell percentage CORIS nomogram for distinguishing severe/critical from non-severe COVID-19 at hospital admission; it was not intended for diagnosis or prediction of post-admission deterioration, survival, or thrombotic events.

    AUC 0.914 (training); AUC 0.918 (validation)
    95% CI: 0.884-0.940 (training); 0.868-0.960 (validation)

    In the internal validation set, CORIS admission-severity discrimination was compared with a lymphocyte subset-based model and a routine laboratory parameter-based model.

    Comparator AUC 0.811 (DeLong p < 0.001) and AUC 0.839 (DeLong p = 0.003), versus CORIS AUC 0.918
    CORIS 95% CI: 0.868-0.960; comparator CIs not reported

    Citation in Biomarkers
    One key learningRead original study in a new tab
  9. STUDY 09Findings added
    Long COVID
    TL;DR

    Long COVID was recorded more often in people with autoimmune rheumatic diseases

    2.8% vs 1.5%

    Long COVID was recorded in 2.8% of patients with autoimmune rheumatic diseases, versus 1.5% of matched patients without those diseases.

    Researchers looked back at national US records, using administrative codes or clinic visits to identify Long COVID. Other differences between the matched groups could still partly explain the result. The study cannot show cause and effect.

    Study details & original results

    Singh, Namrata et al.. Higher risk of Long COVID among patients with autoimmune rheumatic diseases and immunomodulatory therapy: a retrospective cohort analysis of national US data.. Arthritis & rheumatology (Hoboken, N.J.). 2026.

    Long COVID occurrence among patients with autoimmune rheumatic diseases versus matched non-AIRD patients

    2.8% vs 1.5%

    Adjusted risk of Long COVID among patients with autoimmune rheumatic diseases versus matched patients without autoimmune rheumatic diseases

    aHR 1.56
    95% CI: 1.45-1.67

    Citation in Long COVID
    One key learningRead original study in a new tab
  10. STUDY 10Findings added
    Research Methodology
    TL;DR

    Long COVID guidance varied in practical applicability

    A review found that guidance documents varied in methodological quality and practical applicability. Even among the prioritized, higher-scoring documents, applicability was the weakest assessed domain.

    These were separate scores for guidance domains, not percentages of patients helped or a combined overall score. Diverse development methods and scarce guidance from lower-income settings limit comparison.

    Study details & original results

    Montoya Monsalve, Ricardo et al.. Methodological quality and clinical applicability of long COVID guidelines and scientific statements: a systematic review using AGREE II and AGREE-REX.. Frontiers in public health. 2026.

    Separate rescaled AGREE II domain scores across the guidance documents; these are domain scores rather than proportions of documents and were not summed into an overall score.

    Rigor of development 16.7-97.9%; applicability 0.0-95.8%; editorial independence median 100%

    Applicability was the weakest domain among four prioritized documents contributing 99 recommendations, despite those documents scoring higher than the remaining corpus in all six domains.

    68%

    Reporting of a GRADE-compatible evidence-to-decision framework among the 20 appraised documents.

    1 of 20 documents

    Distribution across five NICE-derived care domains for recommendations extracted from the 19 documents identified in the initial search.

    510 recommendations; 63.3% clinical management and 29.6% care planning

    Citation in Research Methodology
    One key learningRead original study in a new tab
  11. STUDY 11Findings added
    BiomarkersLong COVID
    TL;DR

    Routine laboratory tests did not distinguish long COVID symptom patterns

    Routine laboratory measures did not distinguish age-independent symptom patterns in a single-centre long COVID cohort. Routine values also did not usefully predict clinical improvement.

    Symptom groups used self-reported organ-system domains. There was no comparison group without long COVID, and routine panels may miss relevant markers; these findings do not establish that symptoms have no biological basis.

    Symptoms affected fewer body systems, but routine tests did not predict change

    In this Long COVID group, the middle number of body systems with symptoms fell from 4 to 2 during follow-up. Routine lab tests did not distinguish symptom patterns or predict how symptoms changed.

    This observational study followed people at one center. Follow-up length is not given here. Small groupings in routine test results reflected age differences. Routine tests may miss biological signs targeted by more specialized tests.

    Study details & original results

    Oberhauser, Viola et al.. Routine laboratory data are decoupled from symptom phenotype and prognosis in Long COVID: within-cohort variation is dominated by age.. Clinical chemistry and laboratory medicine. 2026.

    Differences in routine analytes across three age-independent Long COVID symptom phenotypes

    0/89 analytes

    Equivalence of routine laboratory values across Long COVID symptom phenotypes

    86/89 within ±0.5 SD; 63/70 within the reference change value

    Prediction of clinical improvement from routine laboratory analytes in Long COVID

    AUC 0.46

    Citation in Biomarkers

    Symptom burden decreased during follow-up in this single-center Long COVID cohort despite the absence of routine laboratory correlates.

    Median 4 to 2 organ systems; p=2.5 × 10-19

    Among 503 patients assigned to three age-independent symptom phenotypes, no routine analyte remained associated after false-discovery-rate correction, adjustment, or 50 pseudo-class draws.

    0 of 89 analytes

    Routine analytes did not predict the symptom course using the study's nested cross-validated elastic-net model.

    AUC 0.46

    Equivalence testing found laboratory values equivalent across symptom phenotypes for most assessed analytes.

    86/89 analytes within ±0.5 SD; 63/70 within the reference change value

    Routine-analyte clustering showed no convincing structure across two to eight clusters; the weak separation tracked age and disappeared after age residualization.

    Silhouette ≤0.09; age 50.5 vs. 38 years

    Citation in Long COVID
    2 key learningsRead original study in a new tab
  12. STUDY 12Findings added
    Rehabilitation
    TL;DR

    Exercise may ease fatigue and breathlessness; certainty varies

    Exercise training improved fatigue and breathlessness more than usual care in pooled randomized trials of people with Long COVID.

    Evidence certainty ranged from moderate to very low, with differing programs and populations. Symptom worsening after exertion was not consistently monitored or reported, and long-term follow-up was limited.

    Study details & original results

    Chen, Yi-Chun et al.. Exercise effectiveness on fatigue in Long-COVID: systematic review and meta-analysis.. Annals of behavioral medicine : a publication of the Society of Behavioral Medicine. 2026.

    Exercise training versus usual care for Long-COVID fatigue severity, pooled across randomized trials; evidence certainty was moderate to very low.

    SMD -0.59; P = .002

    Exercise training versus usual care for Long-COVID dyspnea severity, pooled across randomized trials; evidence certainty was moderate to very low.

    SMD -0.89; P = .005

    Citation in Rehabilitation
    One key learningRead original study in a new tab
12 of 12 study updates

Dates show when findings were added here, not when papers were published. Study populations and comparisons differ.

Explore the original publicationsSearch the papers and publication notices when you want more detail.
30 publications in this selection
  1. 01
    Research paper2026

    Agreement and reliability between the two-day 6-minute incremental step test and two-day cardiopulmonary exercise test in post COVID-19 condition for assessing post-exertional malaise: The REVEAL-study.

    PloS one

    Open original publication in a new tab
  2. 02
    Research paper2026

    Longitudinal assessment of SARS-CoV-2 Spike and Nucleocapsid antigenemia in the ORCHESTRA Post-COVID cohort.

    Clinical microbiology and infection : the official publication of the European Society of Clinical Microbiology and Infectious Diseases

    Open original publication in a new tab
  3. 03
    Research paper2026

    Functional outcomes of children after SARS-CoV-2 infection: An EHR-based cohort study.

    Journal of pediatric rehabilitation medicine

    Open original publication in a new tab
  4. 04
    Research paper2026

    The effect of high-dose of vitamin C on mortality among aged patients with severe COVID-19: A systematic review and meta-analysis.

    Clinical and investigative medicine. Medecine clinique et experimentale

    Open original publication in a new tab
  5. 05
    Research paper2026

    Rhabdomyolysis After COVID-19.

    The Medical journal of Australia

    Open original publication in a new tab
  6. 06
    Research paper2026

    Psychiatric symptoms of long COVID among adults: observational case-control study in Johannesburg, South Africa.

    BJPsych open

    Open original publication in a new tab
Feed snapshot · Nov 7, 2026 UTC

Browse archived briefings