COVID-19: Assessing instances of long COVID in UK health data

Symptoms of long COVID are more frequently reported by women, those with poor overall health before the pandemic, and those aged 50–60, a study published in Nature Communications suggests. The findings are based on data from longitudinal surveys and electronic health records in the UK.

Long COVID, also known as post-acute COVID syndrome, is commonly defined as having one or more symptoms of COVID-19 four weeks post infection. The syndrome has been widely reported but the incidence and risk factors for the condition are not well understood. This is partly because the symptoms used to define long COVID often vary between studies, which are based on relatively small samples, and may not be generalisable to the wider population.

To investigate instances of long COVID, Ellen Thompson and colleagues used data from 6,907 individuals with self-reported COVID-19 from ten population-based longitudinal health surveys in the UK that had been established prior to the pandemic. In parallel, they also utilised data from electronic health records collected by spring 2021 for 1.1 million individuals diagnosed with COVID-19. The authors found that the proportion of presumed COVID-19 cases in the longitudinal surveys reporting symptoms for longer than 12 weeks ranged from 7.8% to 17%, with between 1.2–4.8% reporting ‘debilitating’ symptoms. Additionally, they found that in the electronic health records only 0.4% of COVID-19 cases had a subsequent long COVID-related diagnosis or referral; however, they note that diagnostics codes had only been introduced for long COVID in healthcare settings in December 2020. Although the reporting of long COVID varied across studies, the authors suggest an increased risk of long COVID was associated with increasing age up to age 70. Female sex, poorer pre-pandemic mental health and overall health, obesity and having asthma were also identified as risk factors in both longitudinal studies and health records.

The authors caution that while causal inferences cannot be drawn, their findings highlight the need for further research into at-risk groups. They also suggest additional representative population-based studies to improve estimates and assist healthcare planning are needed.