Long COVID: why it's so hard to tell how many people get it

In the weeks and months after a COVID infection, some people develop a variety of symptoms, commonly referred to as long COVID. Fatigue is the most common, affecting a little over half of people with ongoing symptoms. Other common symptoms include shortness of breath, loss of smell, muscle ache and brain fog.

Our understanding of why some people face ongoing symptoms after COVID is still poor. And much like the varying nature of long COVID symptoms, the duration and intensity of symptoms differ from person to person.

Notably, we've also had difficulty pinning down the precise prevalence of long COVID (that is, what proportion of people it affects). This has been a subject of considerable debate.

But recently, a study published in The Lancet has been touted as providing the most reliable estimates to date on the prevalence of long COVID.

Estimates of how commonly long COVID occurs generally range from about 5% to 50% of COVID cases, depending on which study you look at. Patients who require admission to hospital for COVID have tended to sit at the higher end of the spectrum.

A lot of the variability in prevalence estimates earlier in the pandemic was related to inconsistent definitions of long COVID. To clarify this, and to make diagnosis easier, public health authorities including the UK's National Institute for Health and Care Excellence (NICE) have introduced standardised definitions for long COVID.

According to NICE, the term long COVID can be used to describe signs and symptoms that continue or develop beyond four weeks after a COVID infection. This is further divided into "ongoing symptomatic COVID-19", when symptoms last for more than four weeks but less than 12 weeks, and "post-COVID syndrome", when symptoms continue beyond 12 weeks from infection.

But in spite of these efforts, considerable variation in prevalence estimates has remained.

Read more: Long COVID: with no treatment options, it's little wonder people are seeking unproven therapies like 'blood washing'

What else is causing this variation?

Several other factors also likely affect reported prevalence figures across different studies. For example, results can be afflicted by a lack of consistent instruments or questionnaires to screen for long COVID.

Further, not having a matched control group (a comparison group that didn't have COVID), or information on patients' symptoms before COVID, can be limiting. This information allows researchers to reliably link new symptoms to COVID rather than other diseases.

Meanwhile, vaccination status, treatments that reduce the risk of severe disease in COVID patients (like antivirals) and the variant that caused the initial infection may all affect a person's risk of long COVID.

Finally, the timing of assessment also appears to be relevant. Recent data from the UK's Office for National Statistics suggests that although four in five patients with long COVID have symptoms lasting for at least 12 weeks, only one in five continue to be symptomatic at two years.

One in eight

The recent study published in The Lancet includes data from more than 76,000 people in the Netherlands. The authors have made one of the first attempts to mitigate the various biases that have plagued earlier efforts to quantify long COVID prevalence.

They used questionnaires which asked about a range of symptoms and gave these to participants at various time points before, during and after COVID infection. They also compared participants to a control group who hadn't been diagnosed with COVID and were similar in age and sex.

Of participants who had COVID, 21.4% experienced at least one new symptom, or a symptom that had become significantly worse since before they had COVID, three to five months post-infection. Some 8.7% of uninfected people followed over the same time period reported symptoms. On this basis, the authors suggest that 12.7% of people who contract COVID, or one in eight, develop long COVID.

The authors also ascertained a list of core symptoms associated with long COVID, including difficulty breathing, chest pain, heavy arms and legs, loss of sense of smell, feeling hot and cold, tingling in extremities, muscle ache and tiredness.

Some drawbacks

While this study has advanced our understanding of the real-world prevalence of long COVID, there are some important limitations. The majority of patients were not vaccinated because most of the data was collected before the vaccine rollout began in The Netherlands. According to recent research, long COVID appears to be less common among people who are vaccinated.

Meanwhile, most participants were infected with the alpha variant. Studies have also reported a lower prevalence of long COVID among people recovering from infection caused by the currently dominant omicron variant compared with the alpha and delta variants.

So in the current context, the prevalence of long COVID in the general population may well be less than one in eight.

Read more: Long COVID: female sex, older age and existing health problems increase risk – new research

Importantly, our understanding of the mechanisms that cause persistent symptoms and what makes people vulnerable remains fairly limited. Only additional research, dedicated funding support and greater recognition of the condition will improve the outlook for the millions of people afflicted by long COVID around the world.

Betty Raman receives funding from the British Heart Foundation Oxford Centre of Research Excellence and NIHR Oxford Biomedical Research. She worked forAxcella therapeutics as an honorary speaker.