Skip to main content

Research

Making progress toward recovery

The RECOVER Initiative brings together scientists, clinicians, patients, community members, and caregivers to find answers to important questions about Long COVID.

Explore RECOVER Research

RECOVER Research Publications

RECOVER Research Publications

Read the latest published Long COVID research from RECOVER studies.

RECOVER Research Components

RECOVER Research Components

Learn how 5 different types of studies form the core of RECOVER’s research efforts.

How Research Works

How Research Works

Learn about the guidelines and basic steps all RECOVER studies follow.

RECOVER Research Questions

The goal of RECOVER is to understand, diagnose, prevent, and treat the long-term effects of COVID-19. RECOVER has eight guiding research questions; their answers will help us achieve our goal. This summary represents the great work of many people and communities, and will be updated as new RECOVER findings are published.

The answers that follow summarize findings from RECOVER research publications. They are not intended to summarize, nor do they incorporate research on Long COVID beyond the RECOVER Initiative.

Post-COVID conditions have different names including Long COVID and post-acute sequelae of SARS-CoV-2 infection (PASC), which includes Long COVID. The terms used below reflect those used in the specific studies that are summarized.

Content last reviewed April 2025.

Several RECOVER studies have reported on the symptoms of Long COVID. View studies: [1], [2], [3]. RECOVER research shows there are symptoms that cluster and present together. View studies: [4], [5], [6]. In one study, researchers found six distinct symptom clusters. The clusters involve different body systems at the same time, including: the lungs; the brain and nervous system; and the heart and blood vessels. Some clusters include pain, and some are accompanied by abnormal laboratory findings. View the study. Another study found that four Long COVID symptom clusters (heart and lungs, neurological, gastrointestinal, and coexisting medical conditions) differed depending on the age of patients. In the study, children and teens were more likely to have gastrointestinal and upper respiratory problems such as stomachache and cough; adults aged 21-45 years were more likely to have neurological problems such as “brain fog” and fatigue; and adults aged 66 and older were more likely to have coexisting medical conditions such as heart problems and diabetes. View the study.

Another RECOVER study found a clear pattern of symptom differences between school aged children (6 to 11 years old) and adolescents (12 to 17 years old). After examining 89 prolonged symptoms after SARS-CoV-2 infection across nine symptom areas, researchers found that children and adolescents with probable PASC experienced long-lasting symptoms in almost every organ system, with the majority having multisystem involvement. The most common long-lasting symptoms among PASC-probable school aged children were headache, trouble with memory/focusing, trouble sleeping, and stomach pain. Among PASC-probable adolescents, the most common long-lasting symptoms were daytime tiredness/sleepiness or low energy, body/muscle/joint pain, headaches, and trouble with memory/ focusing. These findings underscore the importance of characterizing PASC separately in these two age groups. View the study.

Multisystem inflammatory syndrome in children (MIS-C) is a severe consequence of COVID-19 observed in some children. Using data from children hospitalized with MIS-C, RECOVER researchers found three clusters of symptoms. Cluster 1 included children with the highest number of organ systems affected and the most severe symptoms—including exacerbated inflammation, low blood pressure, cardiac and pulmonary involvement, and kidney injury—with more common admission to the ICU. Cluster 2 included children with fewer organ systems affected, moderate symptom presentation, and notable overlapping features with acute COVID-19. Lastly, Cluster 3 represented a mild presentation with the fewest organ systems involved. View the study.

RECOVER researchers are also examining how PASC diagnoses, underlying health conditions, and prescription medications might differ geographically. They compared electronic health records from four states (New York, Florida, Georgia, and Alabama) and found differences in the number of PASC conditions reported across these data sites. Their work suggests differences in PASC diagnoses, conditions, and medications among different populations based on neighborhood, socioeconomic status, age, gender, race, and outbreak waves. View the study.

Researchers found ethnic and racial differences in PASC symptoms and conditions in a RECOVER study of adults with COVID-19 in New York City between March 2020 and October 2021. In the period of 31-180 days after infection, compared to White people hospitalized for COVID-19, Black and Hispanic people hospitalized for COVID-19 had higher rates of certain PASC symptoms such as shortness of breath, chest pain, and joint pain. They were also more likely to receive a new diagnosis of diabetes. The results also suggest that people from different racial and ethnic groups may experience different symptoms and conditions of PASC. View the study.

RECOVER researchers are studying large groups of people over a long period of time to gain a deeper understanding of how long the symptoms last and whether it can have effects later in life. Learn more about RECOVER Longitudinal Observational Cohort Studies.

RECOVER researchers are studying whether Long COVID affects other diseases or health problems. They compared electronic health records (EHRs) to examine possible links between Long COVID and new diagnoses of type 2 diabetes. The researchers analyzed data from patients who had at least one outpatient clinical visit during the periods 6 months before and 6 months after SARS-CoV-2 infection. The number of new diabetes diagnoses were 83% lower in the months after SARS-CoV-2 infection than in the period before infection. View the study.

RECOVER researchers are also trying to understand the prevalence of myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) in people with Long COVID. Many ME/CFS symptoms are similar to symptoms experienced by people with Long COVID. In one study, researchers asked 465 people with Long COVID about how often they experience Long COVID symptoms, ME/CFS symptoms, and post-exertional malaise (PEM) (worsening of symptoms after physical or mental activity). They found that over half (272/465 or 58%) of participants experienced ME/CFS symptoms. Further, those who experienced ME/CFS symptoms reported more severe symptoms (such as fatigue/extreme tiredness; feeling unrefreshed after waking in the morning; and problems remembering things) compared to those who did not experience ME/CFS symptoms. However, of those who reported that they had ME/CFS, only 71% met established criteria for ME/CFS. Among those who did not report they had ME/CFS, 40% nevertheless did meet criteria for the disease. Researchers emphasized the need for reliable, proven ways to determine if Long COVID patients' symptoms match the criteria for ME/CFS. View the study.

In another study examining EHRs from 53 health systems in the US, researchers observed an association of acute kidney injury (AKI) with patients diagnosed with COVID-19. AKI was defined either by the increase of serum creatinine level or a diagnostic code. Researchers then characterized the incidence, geographic distribution, and temporal trends of AKI among patients hospitalized with COVID-19. Risk factors associated with AKI in these patients include older age, male sex, and Black and Asian races. Additionally, spikes in AKI cases in the United States correlated with waves of COVID-19 in the last two years. Researchers also found that AKI was highly associated with mortality in patients with COVID-19, with more severe AKI being associated with a greater risk of death. View the study.

In another RECOVER study, researchers looked at the relationship between impaired cognition (cognitive concerns or dementia diagnosis) and death after SARS-CoV-2 infection in people with and without HIV. They studied 527 adults with laboratory-confirmed SARS-CoV-2 infection, including 64 people with HIV. People living with HIV had a higher prevalence of dementia (15.6% vs 6.0%) and cognitive concerns (21.9% vs 15.8%). People living with HIV also had a significantly increased risk of death from COVID-19 (17.2% vs. 6.3%). On average, people living with HIV and COVID-19 died at a younger age despite their HIV being under control of medication. View the study.

RECOVER studies are examining risk factors for developing Long COVID.

RECOVER researchers have found that SARS-CoV-2 infection increases the risk of developing post-acute sequelae of SARS-CoV-2 infection (PASC) and Long COVID. In one study, researchers found that 23% of people with SARS-CoV-2 infection developed PASC, while only 3.7% of people without a reported positive test developed symptoms consistent with PASC. Researchers found that the risk of PASC increases with hospitalization for acute infection and similarly with reinfections. The study found that 39% of people hospitalized with acute COVID developed PASC, compared to 22% of people not hospitalized for acute COVID. Among people infected with the Omicron variant, those who had reinfections were more likely to develop PASC (21%) compared to those with one infection (16%). View the study.

People who have had a SARS-CoV-2 infection are also more likely to have symptoms associated with Long COVID than those without an infection. RECOVER researchers analyzed electronic health record (EHR) data from over 3 million adults and 600,000 children who tested positive or negative for SARS-CoV-2. They examined the occurrence of specific COVID-19-related symptoms as potential Long COVID between 31 and 150 days after testing. The study found that hospitalized adults and children who tested positive were more likely to experience shortness of breath during this period compared to those who tested negative. Additionally, hospitalized adults with a positive test had a higher likelihood of developing fatigue, new diagnoses of type 1 or type 2 diabetes, and blood disorders. These findings indicate that SARS-CoV-2 infection, particularly in hospitalized patients, may increase a person’s risk of developing new symptoms and conditions associated with Long COVID. View the study.

RECOVER researchers are also working to understand how reinfection with SARS-CoV-2 impacts the development of Long COVID. Using EHR data from over 3 million patients, researchers found that Long COVID diagnoses were more common after the first infection than after reinfection. They also found that Long COVID was more common after infection with the Delta variant compared to infections that happened when the Omicron variants were most prevalent. View the study.

In another study, RECOVER researchers examined risk factors for developing PASC. They looked at EHR data from 31 health systems in the US, comparing 8,325 individuals with PASC with 41,625 individuals with COVID. They matched data from patients within the same health system whose initial SARS-CoV-2 infections were within 45 days of one another to compare demographics and health conditions between those with and without a PASC diagnosis. Researchers found a higher likelihood of PASC diagnosis among individuals aged 40 to 69 years, female, with acute COVID-19, and living with other health conditions such as depression, chronic lung disease, and obesity. They also found that in counties served by more doctors per capita, residents were more likely to receive a PASC diagnosis or care at a Long COVID clinic. Risk factors associated with a lower likelihood of developing PASC included younger age (18-29 years), male sex, non-Hispanic Black race, and comorbidities including psychosis, dementia, and tobacco smoking. View the study.

RECOVER researchers investigated whether having COVID while pregnant affects a person’s chances of developing Long COVID. A study conducted from December 2021 to September 2023 found that about 9.3% of people infected with SARS-CoV-2 during pregnancy developed Long COVID. Compared to those without a Long COVID diagnosis, pregnant people with Long COVID were more likely to have a history of obesity, depression and/or anxiety, and economic hardship. View the study. In a separate study using EHR data, researchers found that SARS-CoV-2 infection during pregnancy was associated with an increased risk of abnormal heartbeat, abdominal pain, and blood clots—symptoms of Long COVID. However, individuals infected with SARS-CoV-2 during pregnancy had about a 15% lower risk of being diagnosed with Long COVID compared to those infected outside of pregnancy. View the study.

RECOVER research using EHRs suggests that obstructive sleep apnea may increase the risk for developing Long COVID after infection. Among people who had COVID, adults with obstructive sleep apnea were more likely to experience long-term symptoms suggestive of Long COVID than those without the sleep disorder. The analysis of EHR data found that adults with sleep apnea may have up to a 75% higher risk of developing Long COVID after infection. View the study.

RECOVER researchers found that certain environmental factors related to where people live, like bad air quality, access to food, green spaces, neighborhood deprivation, and social factors, are associated with having Long COVID symptoms. View the study.

RECOVER researchers are studying whether variants of the SARS-CoV-2 virus and COVID-19 vaccines make a difference for developing Long COVID.

In a study following more than 9,000 people over time, RECOVER researchers found that Long COVID was more common and associated with more serious manifestations for adults infected before the Omicron variant emerged. View the study. Another study compared the Long COVID symptoms experienced by people infected by the original SARS-CoV-2 strain from 2020 and people infected by the Delta variant from 2021.The researchers found that people infected during the Delta period experienced more abdominal health issues, while lung-related symptoms were common in both groups. However, it remains uncertain whether these differences in Long COVID symptoms are due to changes in the virus itself or other factors, such as differences in the characteristics of those infected. View the study.

Regardless of the viral variant, the rates of PASC among those fully vaccinated are lower than those who were unvaccinated. Researchers found the same pattern when they compared vaccinated and unvaccinated adults at three different points in time: in adults with post-acute COVID, pre-Omicron (31% vs. 37%), with acute Omicron infection (9.7% vs. 17%), and post-acute Omicron (16% vs. 22%). Similarly, in a large observational study, RECOVER researchers found that vaccination was consistently associated with lower rates of Long COVID, before and after Omicron emerged. View the study. RECOVER researchers reviewed the EHRs of more than 15 million people (age 5 and older) to compare the risk of rare heart problems after a SARS-CoV-2 infection versus after being vaccinated for COVID-19. They found that the risk of having a rare heart problem after having COVID is much higher in people who did not receive a COVID vaccine than those who did. The risk of rare heart problems after COVID-19 vaccination with an mRNA vaccine (Pfizer or Moderna) was very low, supporting the continued use of vaccination to prevent COVID-19. View the study.

In children, vaccine effectiveness (VE) against Long COVID is less understood. Researchers studied the EHRs of children ages 5-17 that previously received the COVID-19 vaccine to examine VE in preventing Long COVID. After adjustment for important confounder factors to prevent bias, they found that within 12 months of vaccination, VE was 35.4% against probable (symptom-based) Long COVID and 41.7% against diagnosed (diagnosis code-based) Long COVID. Notably, VE was higher for adolescents aged 12-17 (50.3%) than for children aged 5-11 (23.8%). Overall, this research demonstrates that COVID vaccination has a moderate protective effect against Long COVID in children. View the study.

RECOVER researchers assessed the effectiveness of BNT162b2 (a COVID-19 vaccine) at preventing infection and severe disease with different strains of SARS-CoV-2 in previously uninfected children (ages 5-11 years) and adolescents (ages 12-20 years). This study used EHR data from a national collaboration of pediatric health systems and found high levels of effectiveness against moderate or severe COVID-19 and ICU admission with COVID-19 for both children and adolescents. The vaccine against the Omicron variant also reduced the risk of cardiac complications, including multisystem inflammatory syndrome in children (MIS-C)—a form of Long COVID. View the study.

In addition to the observational cohort studies that re-examine participants over time, RECOVER researchers are conducting more than 40 different pathobiology studies focusing on COVID-19's effects on different body tissues and organs.

Researchers are trying to identify potential biomarkers of PASC and validate them for clinical applications. Biomarkers represent measurable indicators of the presence of a disease. Researchers examined 25 routine clinical laboratory values and found none that could serve as a clinically useful biomarker of PASC. In this study, researchers examined data from 10,094 RECOVER adult cohort participants: 8,746 with a prior SARS-CoV-2 infection and 1,348 without a prior infection. They found that, on average, participants with prior infection had lower blood platelet counts and higher blood sugar levels compared to those without infection. However, these differences were modest and not statistically significant. View the study.

RECOVER researchers are studying individuals with Long COVID who frequently report constant fatigue, post-exertional malaise (PEM) (symptoms worsening after physical or mental activity), and a variety of cognitive issues, like “brain fog.” RECOVER researchers examined 275 people with and without Long COVID to identify biological features associated with these symptoms. They found people with Long COVID had significant differences in their immune systems compared to those without Long COVID. In participants with Long COVID, researchers saw increases in counts of some immune cell types, while counts of other immune cells decreased. Researchers also saw higher antibody responses to SARS-COV2 and non-COVID viruses, particularly Epstein-Barr virus (a herpesvirus) in participants with Long COVID. Researchers suggest that persistent SARS-CoV-2 antigens, reactivation of herpes viruses, and chronic inflammation may contribute to Long COVID. View the study.

To understand why some people with COVID-19 experience a loss of their sense of smell (known as hyposmia), RECOVER researchers are studying the cells inside the nose that enable smell. They examined 24 biopsies from 9 PASC patients with long-term smell loss after contracting COVID-19. Biopsies from people with their sense of smell intact served as a comparison. In participants with PASC, there was no trace of SARS-CoV-2 virus or its genetic material in olfactory epithelium. However, the cells supporting the protective barrier in the nose, known as sustentacular cells, showed signs of reacting to ongoing inflammation. This reaction was also linked to a decrease in the number of olfactory sensory nerve cells (neurons). These findings suggest that inflammation (mediated by T-cells) continues in the olfactory epithelium even after the virus has been eliminated from the tissue. This is a potential explanation for the persistent loss of smell experienced by some people living with Long COVID. View the study.

Inflammation caused by infections can have a lasting impact on certain types of cells in the human body, including those related to the immune system. In one study, researchers sought to uncover how inflammation could impact a specific cell type—hematopoietic stem and progenitor cells (HSPC)—which are crucial for the formation of blood and immune cells. Researchers found that the cells of individuals who had a serious case of COVID-19 could retain changes for several months to a year after infection. They saw changes in both the physical traits of the cells and their genetic programming. They also linked these changes to proteins that control gene expression, changes in the regulation of inflammatory processes, and a lasting increase in the production of certain immune cells. Importantly, the observed changes in HSPC were passed to their “offspring” as they developed into mature immune cells. Finally, researchers found that a molecule called IL-6 played a role in maintaining these lasting effects in human COVID-19 patients as well as in the study model for SARS-CoV-2 infection performed in mice. Overall, researchers suggest changes to HSPC at the genetic level could be responsible for changes in the immune system after infection, especially in people who experienced severe COVID-19. View the study.

At this time, the only known way to prevent Long COVID is to avoid getting COVID-19. The observational cohort studies and the RECOVER Pathobiology Research Program are investigating potential mechanisms for how some people may be protected from Long COVID or some of its symptoms.

RECOVER researchers are making progress toward understanding why some people develop Long COVID and how these long-term symptoms affect a person’s health. Based on what the researchers have learned so far, RECOVER has launched clinical trials to test different interventions for Long COVID. The eight clinical trials under RECOVER Initiative will explore 13 study interventions, or possible treatments, for post-acute sequelae of SARS-CoV-2 infection (PASC), known as Long COVID.

There are five RECOVER platform protocols, with one or more clinical trials in each platform. The protocol for each platform describes the goals of the research, who can enroll, possible treatments, study activities, and more.

  • RECOVER-NEURO focuses on treatments for cognitive symptoms associated with Long COVID.
  • RECOVER-VITAL focuses on treating viral persistence and reactivation, which is when the SARS-CoV-2 virus stays in the body and damages organs or the immune system. The latter is considered a possible cause for Long COVID. This trial focuses on treatments to stop the virus from replication to prevent continued damage and to resolve symptoms caused by the virus.
  • RECOVER-AUTONOMIC focuses on treatments for adults who have an autonomic nervous system disorder, called Postural Orthostatic Tachycardia Syndrome (POTS), a common consequence of Long COVID.
  • RECOVER-SLEEP focuses on treatments for sleep disturbances and hypersomnia (trouble staying awake during the day).
  • RECOVER-ENERGIZE focuses on treatments for exercise intolerance and the worsening of symptoms following physical or mental exertion known as in patients with Long COVID.

More information about RECOVER clinical trials can be found at trials.recoverCOVID.org.

RECOVER researchers have found that taking metformin, a medication for people living with type 2 diabetes, could lower some people’s risk of getting Long COVID. Using electronic health records from two national databases, researchers examined adults living with type 2 diabetes who used metformin at the time of their SARS-CoV-2 infection versus individuals who did not use metformin (but took other diabetes medications). Metformin use was associated with a slightly lower risk of dying or developing Long COVID within 6 months of a SARS-CoV-2 infection. View the study.

Another RECOVER study identified patients with mild Multisystem Inflammatory Syndrome in Children (MIS-C) and reviewed treatment outcomes among seven patients seen in outpatient setting without treatments focused on the immune system (called immunomodulators). MIS-C is a multi-organ inflammatory condition of Long COVID that mainly affects children and young adults. The result from this small study suggested that outpatient management alone may be effective at preventing hospitalization for patients with mild MIS-C. View the study.

RECOVER Together

Progress takes the best science.

RECOVER is collecting data from many patients across the country. This data is informing:

  • Observational cohort and electronic health record (EHR) studies that use health data to understand how people are experiencing Long COVID.
  • Pathobiology, tissue pathology, and autopsy studies that examine tissues from research participants to understand changes inside the body.
  • Clinical trials that test different treatments for symptoms experienced by patient participants with Long COVID.

By exploring different ways COVID affects people’s bodies and looking at which treatments may be effective, we can better understand Long COVID. We will then combine all of these data to help answer big questions about Long COVID.

Each type of RECOVER cohort study follows a detailed study plan called a research protocol. These protocols ensure that RECOVER researchers at each cohort study site follow the same guidelines and steps. This consistency lets us combine data from cohort studies taking place at different locations. Combining study data in this way helps us get answers to important questions faster.

We designed RECOVER so that each type of study offers different kinds of information that, when combined, will help us understand, diagnose, prevent, and treat Long COVID. Similarly, we include different groups of people, known as cohorts, in RECOVER cohort studies. These groups include adults, children and their caregivers, and pregnant women and their newborn babies. Including people from different backgrounds and walks of life in RECOVER cohorts helps us understand how the long-term effects of COVID can be different for everyone.

RECOVER clinical trials use platform protocols that help researchers work more efficiently and get answers for more people faster. Platform protocols that allow researchers to study different possible treatments at the same time. Platform protocols also allow researchers to add new potential treatments to ongoing clinical trials.

All of Us

The NIH All of Us Research Program is building one of the largest and most diverse health databases in history. By studying health information from people of all backgrounds, researchers using All of Us data can learn more about what makes people sick and keeps them healthy.

The All of Us Researcher Workbench offers authorized researchers access to the COVID-19 Participant Experience (COPE) survey, electronic health record (EHR), wearable, and genetic data.

The All of Us Data Browser offers anyone access to aggregate-level data from those sources.

Access the Data Browser

Back to Top