Has the risk of getting long covid been overestimated?

If you are infected with the coronavirus, how likely are you to develop long covid? It is an important question, as the answer could affect individuals’ decisions about taking precautions against the virus, such as whether to wear a mask, and decisions by medical bodies, such as who should be offered booster vaccines.

Unfortunately, our scientific understanding of the condition has remained poor throughout the pandemic. Long covid is generally used as an umbrella term for any kind of lasting symptoms after covid-19 infection, usually ones that have gone on for three months or more. The most common include fatigue, breathlessness and difficulties concentrating, but some doctors say it encompasses over 200 different symptoms.

Now, Tracy Beth Høeg at the University of California, San Francisco, and her colleagues have claimed that the likelihood of this condition has been overestimated. While some studies suggest long covid affects as many as half of all those infected, that is down to their loose definitions of the condition or poor design. The most authoritative studies suggest that only a few per cent of people are affected, says Høeg.

But critics of this analysis say the researchers have ignored other well-designed studies that support the idea that the virus often has lasting effects. Why is figuring out how common it is to get long covid so difficult?

Part of the problem is we don’t know exactly what causes the condition. Several explanations have been proposed, including that the virus persists in the body or that it causes either immune system overactivity or underactivity – but it is unknown which of these, if any, are correct. Long covid also seems to have similarities with myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS), another mysterious syndrome of persistent tiredness that may arise after other infections.

Some studies of long covid – often those conducted earlier in the pandemic – just surveyed people about whether they had any lasting symptoms after infection or if they self-reported that they had long covid.

This kind of research is flawed because it is common for people to have symptoms such as fatigue regardless of infections, says Høeg. Instead, it is vital to compare the rate of long covid symptoms in people after an infection with the equivalent figures in a control group of people who didn’t catch the coronavirus, she says.

Some studies did have control groups, but because of their design, people in these groups tended to be in better health to begin with than those who were diagnosed with covid-19, because people with worse underlying health were more likely to get tested at hospital if they developed covid-19 symptoms. This would also overstate the incidence of long covid, says Høeg.

One of the most authoritative studies is by the UK’s Office for National Statistics, which asked large numbers of people to carry out regular covid-19 tests, whether they felt sick or not. This found that 5 per cent of people had any of the 12 chief long covid symptoms three to four months after an infection – but so did 3.4 per cent of people who hadn’t been infected. This suggests that 1.6 per cent of people who get infected develop long covid.

“Studies which did not include control groups at all should simply not have been used for prevalence estimates of the still vaguely defined long covid and it remains a mystery to me why they were,” says Høeg.

She says the media coverage of research that produced high estimates means that many people still believe long covid is more common than it really is. “Fear-based articles attract more attention,” she says.

The latest analysis is unlikely to settle the debate, though. Those scientists who argue that long covid needs more recognition and more research say the new claims are insulting to people who have the condition. “Long covid is a really complicated thing and they’re trying to boil it down to something that’s too simplistic, says Stephen Griffin at the University of Leeds, UK, who is a member of iSAGE, a group of scientists who want a return of greater covid-19 precautions.

Jeremy Rossman at the University of Kent, UK, says the paper ignored some other well-designed studies that did use control groups. For instance, one from Iceland estimated that 13 per cent of people had at least one symptom eight months after infection and these were severe enough to affect people’s everyday lives in 7 per cent of the total. “They don’t define why some papers are used as examples while other papers that appear to fit their criteria but have higher prevalence rates are not discussed,” he says.

However, the analysis doesn’t claim to be a “systematic review”, a common kind of scientific paper that aims to include all studies published on a subject. It also omitted other well designed studies that support the low prevalence claim, such as one from Australia in August, which found the incidence of lingering symptoms after three months were about the same with covid-19 as flu – at about 20 per cent for any symptoms at all and 4 per cent for those that caused functional impairment.

In some ways, the exact number of people with long covid may not make much difference to those who have the condition. What they really want is to get better – and unfortunately, this doesn’t tell us anything about how best to achieve that.