Part of the problem is that most studies are carried out on a relatively small number of people.
A new study has drawn on more than a million UK patient records to glean a better understanding [of long-COVID sufferers].
A new study has drawn on more than a million UK patient records to glean a better understanding.
According to the study, the risk of long COVID is higher for women, people who had poorer mental and physical health prior to the pandemic, and people who exprerience asthma.
Long COVID risk also increases with age, up until about 70 - at which point it declines.
The study is described in a paper published in Nature Communications.
The UK-based team of researchers examined two main sources of data: 10 longitudinal surveys from self-reported people who'd had COVID, totalling 6907 people, and healthcare records of 1.1 million people who'd been diagnosed with COVID-19. Both sources came from the UK.
You might also like:
People who fit the technical definition of long COVID - that is, reporting symptoms more than 12 weeks after infection - comprised between 7.8 per cent and 17 per cent of the sample. Between 1.2 per cent and 4.8 per cent of people reported "debilitating" symptoms.
Women were 50 per cent more likely than men to have long COVID. Risk of long COVID also increased steadily with age, up until about 70.
"In individuals aged over 70 we observed a sharp decline in long COVID risk in most LS [longitudinal survey] and the EHR [electronic health record] data," the authors write in their paper.
"This decline in older age has been observed in other studies and may be spurious due to selective competing risk of mortality, non-response bias, lower symptom reporting in older adults, misattribution of long COVID to other illness, or a combination of these factors."
A few other demographic factors were more weakly linked to long COVID.
"We found some evidence of higher long COVID reporting among individuals of white ethnicity and of higher educational attainment, which was unexpected given the common associations of these characteristics with lower morbidity more generally," they write.
The electronic health records suggested that people in least socioeconomically deprived areas were more likely to report long COVID symptoms - but this pattern wasn't seen in the longitudinal surveys.
"The stark variability in proportions of COVID-19 cases with persistent symptoms is clear from our comparison of methods of ascertaining long COVID," write the authors.
They urge for more representative population-based studies to be conducted, in order to further understand the illness.
"Our data suggest that revisions of diagnostic criteria within primary care may be appropriate, particularly for demographic groups which are less in touch with healthcare services," they write.
"Although causal inferences cannot be drawn from these data, our findings justify further investigations into the role of sex difference, age-related change, and/or immunity and respiratory health in development of long COVID.
"Older working individuals, with high levels of comorbidity, may particularly require support."