Long COVID: what the large studies and treatment trials show so far
How long COVID is defined, how the risk has changed since 2020, what has been found in the body, and why two trials of an antiviral drug did not show a benefit.

Long COVID is defined by the National Academies of Sciences, Engineering, and Medicine as a chronic condition that follows SARS-CoV-2 infection and is present for at least three months. It can be continuous, come and go, or get worse over time, and it can affect one or more organ systems (Ely et al., 2024).
In the 2022 National Health Interview Survey, 6.9 percent of U.S. adults said they had ever had long COVID and 3.4 percent said they had it at the time (Adjaye-Gbewonyo et al., 2023). No drug has been shown to treat it.
This article is for people with long COVID and those close to them who want a plain account of what the research has established.
What it involves
According to the National Academies, long COVID can follow an infection of any severity, including one that caused no symptoms, and it can affect children and adults. It can appear as one symptom or many. Those listed in the definition include shortness of breath, cough, persistent fatigue, difficulty concentrating, memory changes, recurring headache, lightheadedness, a fast heart rate, sleep disturbance, problems with taste or smell, bloating, constipation, and diarrhea. No laboratory test confirms it, and the diagnosis is made on clinical grounds (Ely et al., 2024).
One feature separates long COVID from ordinary tiredness. Post-exertional malaise is a worsening of symptoms after physical or mental effort. It often comes on hours to a day later and can last for days.
Who gets it and how the risk has changed
The National Institutes of Health study known as RECOVER followed 9,764 adults and identified the symptoms that best separate people with long COVID from people who recovered. They include worsening of symptoms after exertion, fatigue, brain fog, dizziness, gut symptoms, palpitations, loss of smell or taste, thirst, chronic cough, and chest pain. Among people first infected from December 2021 onward, 10 percent met the study's definition six months later (Thaweethai et al., 2023).
The risk has fallen since the first years of the pandemic. A study of U.S. veterans found that among unvaccinated people, 10.4 per 100 developed long COVID within a year of infection in the earliest period, and 7.8 per 100 did so in the Omicron period. Among vaccinated people in the Omicron period, the figure was 3.5 per 100. The authors attributed about 72 percent of the overall decline to vaccination and the rest to changes in the virus and other factors (Xie et al., 2024).
What has been found in the body
There is no single test for long COVID, but studies have found measurable changes.
Muscle. Dutch researchers took muscle biopsies from people with long COVID before and after an exercise test that brought on a worsening of their symptoms. Compared with healthy volunteers, the patients had lower exercise capacity and signs of disturbed energy metabolism in muscle. After the exercise they had more muscle damage and more deposits of an abnormal protein in the tissue (Appelman et al., 2024).
Serotonin. Another group found lower levels of serotonin in the blood of people with long COVID. They traced it to ongoing inflammation, which reduced absorption of the building block of serotonin from the gut, among other effects (Wong et al., 2023).
Findings like these show that the symptoms have a physical basis. They have not yet led to a proven treatment.
The antiviral trials
One theory is that the virus, or pieces of it, persists in the body. If so, a longer course of an antiviral drug might help. Two trials tested nirmatrelvir-ritonavir, sold as Paxlovid.
The first enrolled 155 people who had been ill for an average of 17 months. They took the drug or a placebo for 15 days, three times the length of the course used for acute infection. After 10 weeks there was no significant difference in six core symptoms (Geng et al., 2024).
The second, part of the RECOVER program, enrolled 959 people and tested 15-day and 25-day courses. It grouped participants by whether their main problem was thinking, autonomic symptoms such as dizziness on standing, or intolerance of exercise. There was no significant benefit in any group. There were no safety concerns (Baden et al., 2026).
One explanation for results like these is that long COVID is probably not a single condition. People with the same label can have different underlying problems, and a drug that helps one group may show no effect when everyone is studied together. The RECOVER trial tried to address this by grouping participants by their main symptoms, and still found no benefit (Baden et al., 2026).
Prevention and rehabilitation
One trial suggests that treatment during the initial infection can lower the risk. Among 1,126 adults followed for ten months after COVID-19, 6.3 percent of those who took metformin for two weeks at the time of infection were later diagnosed with long COVID, compared with 10.4 percent of those who took placebo. Ivermectin had no effect (Bramante et al., 2023). The participants were adults with overweight or obesity, and the diagnosis was reported by the participants themselves. The trial tested prevention. It does not show whether metformin helps people who already have long COVID.
A continuously updated review in The BMJ covering 24 trials found moderate-certainty evidence for two approaches. An online program of cognitive behavioral therapy probably reduces fatigue and improves concentration. An online, supervised program combining physical and mental health rehabilitation probably improves overall health (Zeraatkar et al., 2024).
For the rehabilitation program, the reviewers estimated that 161 more people per 1,000 had a meaningful improvement or recovered than with usual care. They also found that aerobic exercise done in intervals probably improves physical function more than continuous exercise (Zeraatkar et al., 2024).
These programs do not suit everyone. People whose symptoms get worse after exertion can be set back by doing too much, and are generally advised to pace their activity and increase it only as tolerated.
What is being tested now
The next phase of the NIH program is testing four treatments. A trial of baricitinib, a drug that dampens inflammation, is enrolling at 17 sites. Trials of low-dose naltrexone, semaglutide, and a nerve block in the neck called a stellate ganglion block are in development (RECOVER Initiative).
Practical points
- Staying up to date with COVID-19 vaccination lowers the risk of long COVID if you are infected.
- If exertion makes you worse a day or two later, plan activity in small amounts with rest between and do not push through.
- Treat specific problems. Dizziness on standing, poor sleep, headaches, and breathlessness each have their own treatments.
- Ask your doctor to check for other causes of your symptoms, such as anemia, thyroid disease, and sleep apnea.
- Be careful with expensive treatments that have not been tested in trials.
When to see a doctor
See a doctor if symptoms last more than three months after COVID-19 or keep you from work or daily activities. Chest pain, shortness of breath at rest, fainting, or new confusion or weakness should be assessed urgently.
Sources
- Long Covid Defined (Ely et al., New England Journal of Medicine, 2024) — summary of the National Academies of Sciences, Engineering, and Medicine definition
- Long COVID in Adults: United States, 2022 (Adjaye-Gbewonyo et al., NCHS Data Brief, 2023) — analysis of the 2022 National Health Interview Survey
- Development of a Definition of Postacute Sequelae of SARS-CoV-2 Infection (Thaweethai et al., JAMA, 2023) — prospective cohort study, n=9,764
- Postacute Sequelae of SARS-CoV-2 Infection in the Pre-Delta, Delta, and Omicron Eras (Xie et al., New England Journal of Medicine, 2024) — cohort study of U.S. veterans' health records
- Muscle abnormalities worsen after post-exertional malaise in long COVID (Appelman et al., Nature Communications, 2024) — case-control study with muscle biopsies before and after exercise
- Serotonin reduction in post-acute sequelae of viral infection (Wong et al., Cell, 2023) — laboratory and patient cohort study
- Nirmatrelvir-Ritonavir and Symptoms in Adults With Postacute Sequelae of SARS-CoV-2 Infection: The STOP-PASC Randomized Clinical Trial (Geng et al., JAMA Internal Medicine, 2024) — randomized placebo-controlled trial, n=155
- Nirmatrelvir-ritonavir targeting viral persistence in post-COVID-19 condition (long COVID) in the USA (RECOVER-VITAL): a randomised, double-blind, placebo-controlled, phase 2 trial (Baden et al., The Lancet Infectious Diseases, 2026) — randomized placebo-controlled trial, n=959
- Outpatient treatment of COVID-19 and incidence of post-COVID-19 condition over 10 months (COVID-OUT): a multicentre, randomised, quadruple-blind, parallel-group, phase 3 trial (Bramante et al., The Lancet Infectious Diseases, 2023) — randomized placebo-controlled trial, n=1,126 with long-term follow-up
- Interventions for the management of long covid (post-covid condition): living systematic review (Zeraatkar et al., The BMJ, 2024) — living systematic review of 24 trials, n=3,695
- RECOVER at mid-year: 2026 progress and what's ahead (National Institutes of Health RECOVER Initiative) — program update on completed and planned long COVID trials
This article is for general information and is not medical advice. Talk to your doctor before changing a medication, treatment or exercise program.



