50% of long COVID patients show improvement in 3 years — but the condition tends to linger in those with this subtype

Symptoms of long COVID ease in half of patients within three years of the infection that kicked off their illness.

That's the primary conclusion of a new study that tracked over 1,000 patients with long COVID, a chronic condition that can emerge after an infection with the SARS-CoV-2 virus. Among the patients, the chance of improvement varied sharply depending on how the condition affected their bodily systems.

In the study, published Tuesday (Nov. 3) in the journal Annals of Internal Medicine, roughly 60% of patients whose predominant symptoms were changes to their ability to taste or smell had significantly improved three years after their first positive test. But less than 20% of those with a multisystem presentation had improved in that time frame.

This latter group had symptoms that affected their cardiac, nervous and gastrointestinal systems, among others. An expert told Live Science that the condition resembled myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS), which often arises after an infection and similarly impacts a wide range of bodily systems.

The new study's findings could help direct the structure of future clinical trials of treatments for long COVID, experts said.

One disease, many symptoms

The patients in the study were enrolled in the long-term RECOVER Adult Study. Funded by the National Institutes of Health, the project has recruited thousands of adults across the United States who previously tested positive for COVID-19.

Researchers scored patients based on how they were affected by a range of symptoms, including fatigue, brain fog, shortness of breath and heart palpitations. The study included patients who scored more than 10 points on a 30-point scale — called the 2024 LC Research Index (LCRI) — which attempted to capture these different symptom combinations.

Nearly 1,020 patients met this requirement and were tracked over time; they had in-person appointments and filled out symptom surveys every three months. Scientists marked patients as "improved" once they recorded LCRI scores of 10 or less at two consecutive visits. They found that 36% of the participants had improved two years after their inciting COVID-19 infection, and that figure rose to 50% after three years.

LCRIs hit zero for 26% of the participants after three years, meaning all of their symptoms were gone at the time of the survey.

Reeder said he was careful to avoid describing the participants as "recovered." This decision was influenced by input from the long COVID community to the RECOVER Adult Study, he said. They highlighted that their symptoms can wax and wane, whereas "recovery" implies a permanent resolution. The new study didn't capture, for example, whether individual patients who recorded meaningful improvement later relapsed.

Some patients' journeys were far less straightforward than the top-line data might suggest. Although men and women improved at similar rates, adults under 46 were more likely to improve than older participants were. Patients affected by the Omicron variants of SARS-CoV-2, which emerged in 2021 and quickly became the dominant strain of the virus, were more likely to improve than those infected with earlier strains.

ME/CFS and long COVID

Some of the starkest differences in improvement rates were seen between patients with different symptoms.

In a previous study, RECOVER researchers had grouped participants into five subtypes, using an algorithm to connect patients with similar predominant symptoms:

• Loss of smell or taste
• Chronic cough
• Brain fog
• Heart palpitations
• Symptoms affecting multiple bodily systems

In the new study, patients in the last three groups were less likely to get better within three years. The final group's rate was particularly low, at just 19%.

Dr. Carmen Scheibenbogen, an immunologist at Charité University Medicine Berlin, said these patients' progress resembles that of people with ME/CFS.

"It looks as if at least a major subgroup is suffering from ME-CFS," Scheibenbogen said, noting that COVID-19 may have triggered their condition. "This is a disease which is triggered by different infections, but in the end, the disease [itself] is very similar," she said.

Like long COVID, ME/CFS can affect a wide array of bodily systems, and it remains poorly understood despite recent advances in research.

The team's new findings highlight the need for clinical trials to test possible treatments for long COVID. Both Scheibenbogen and study co-author Harrison Reeder, a biostatistician at Massachusetts General Hospital, said these subtypes would be important to consider in the design of these potential trials.

The multisystem subgroup, Reeder said, "might be a group where an effective intervention can make a bigger difference than people who stand to potentially have a higher chance of improvement regardless."

Future work would need to further scientists' understanding of people affected by long COVID at the level of their genes and proteins, both scientists said.

"There's a lot of work to do in RECOVER and in the world of long COVID research," Harrison said. "There's a lot of opportunity."