Long COVID: answers emerge on how many people get better

More than three years after SARS-CoV-2 began its global spread, an estimated 65 million or more people¹ are still living with the often devastating effects of long COVID — and scientists are still struggling to understand this complex condition.

Even the definition of long COVID, whose symptoms include headaches, fatigue, ‘brain fog’ and more, is debated. Its causes are also elusive.

But researchers now have enough data to provide some preliminary answers to urgent questions about the condition, such as the timescale for possible improvement, factors that raise the risk of developing long COVID, and what can be done to prevent it.

How many people with long COVID get better?

The answer to that question hinges in part on whether a person is classified as recovered. That, in turn, depends on the definition of long COVID, which varies widely. The World Health Organization defines it as symptoms arising within three months of infection with SARS-CoV-2 and lasting at least two months, although that definition is disputed.

Why is COVID life-threatening for some people? Genetics study offers clues

For a study² published in May, researchers followed 1106 adults who caught SARS-CoV-2 before vaccines were available. After six months, 22.9% of them still had symptoms. This fell to 18.5% at one year and 17.2% after two years.

“As soon as it’s 12 months, it plateaus,” says study co-author Tala Ballouz, an epidemiologist at the University of Zurich in Switzerland. “You have a higher chance of recovery during the first year, and after one year it really becomes more of a chronic condition.”

In another study³ published in May, biostatistician Andrea Foulkes at Massachusetts General Hospital Biostatistics in Boston and her colleagues reported that one-third of people who had long COVID six months after infection no longer had it at nine months.

What are the risk factors for long COVID?

A meta-analysis⁴ published in March identified a host of risk factors for developing the condition: being female, of older age, having a high body mass index and smoking. But the greatest risk was associated with certain pre-existing diseases, such as asthma and diabetes, and with hospitalization for COVID-19.

Another key factor might be the severity of the initial infection, says Unnur Anna Valdimarsdóttir, an epidemiologist at the University of Iceland in Reykjavik. In a preprint⁵ released on medRxiv in April, Valdimarsdóttir and her colleagues found that people who had to be confined to bed for 7 days or more with COVID-19 tended to have more symptoms up to 27 months after diagnosis than did people whose COVID-19 was not so severe as to leave them bedridden.

On the plus side, a preprint⁶ posted in April found that the risk of long COVID from a second SARS-CoV-2 infection was lower than from a first infection. Further study is needed to determine whether risk falls further with each subsequent infection.

Neither that study nor Valdimarsdóttir’s has been peer reviewed.

Can long COVID be treated or prevented?

There is now considerable evidence that vaccination reduces the risk of long COVID. In the March meta-analysis, people who had received two doses of vaccine were significantly less likely to develop the condition than were unvaccinated people. Similarly, Martí Català Sabaté, a medical statistician at the University of Oxford, UK, and his colleagues posted a preprint⁷ in June tracking more than 10 million vaccinated people and more than 10 million unvaccinated people from four studies. Consistently, the researchers found that vaccinated people were less likely to develop long COVID than were the unvaccinated. The study has not yet been peer reviewed.

Can drugs reduce the risk of long COVID? What scientists know so far

“The only thing that we know works at the moment is to actually get vaccinated,” says Vassilios Vassiliou, a cardiologist at the University of East Anglia in Norwich, UK, and one of the co-authors of the March meta-analysis.

Even so, several medications are showing promise for preventing long COVID. One is metformin, a standard treatment for type 2 diabetes. In a study⁸ published in June, Carolyn Bramante, an obesity-medicine physician at the University of Minnesota in Minneapolis, and her colleagues followed 1126 people with overweight or obesity who contracted COVID-19. By day 300 of the study, 6.3% of those who had received metformin during their acute illness had long COVID, compared with 10.4% who’d received a placebo. Ballouz says the results are “promising” but the results need to be replicated in other groups.

Meanwhile, a study⁹ published in March found that the antiviral Paxlovid — a combination of the drugs nirmatrelvir and ritonavir — reduced the risk of long COVID if given during the acute phase. This result is “also promising”, says Ballouz. It also needs replication, she says, because the study participants were mostly male.

But preventive agents will not help those who now have the condition, which could affect “10% of infected people,” according to a review¹ published in January; similarly, the WHO estimates prevalence to be 10–20%. “The real benefit would be to actually find a therapy that would work once people have been diagnosed with long COVID”, says Ballouz.