select search filters
roundups & rapid reactions
before the headlines
Fiona fox's blog

expert reaction to study looking at long COVID prevalence associated with delta versus omicron variants of SARS-CoV-2

A study published in the Lancet looks at risk of long COVID associated with different variants of SARS-CoV-2.


Dr David Strain, Clinical Senior Lecturer and Honorary Consultant, University of Exeter Medical School, said:

“These figures from the well-respected Zoe group are consistent with those presented by the Office for National Statistics a month or so ago, and very reassuring for people who caught Covid in that period.  It is important to say however that these data come from the omicron BA.1 period.  The Office for National Statistics suggested that the BA.2 variant that has caused much of the recent wave, did cause long Covid in triple-vaccinated people at approximately the same rate as the Delta.

“Obviously, it is impossible to comment on the longer term impact of these variants as we are still in the early days.  If the resolution rate for these people with omicron BA.1 remains at 50% between week 4 and week 12, and a further 12.5% for those who still have symptoms at 12 weeks, this will be very good news for all the healthcare services that are concerned about the potential number of long Covid patients over the next 2 to 3 years.  It is feasible, however, that those who got long Covid despite triple vaccination and the more mild BA.1 omicron variant are in some way more susceptible and may not follow the same course.

“However even 4.4% of the very large number who experienced Covid as protections were eased, creates a significant public health burden of this disease with no known treatment, or even reliable diagnostic test.”


Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:

“This is an interesting piece of work, though it needs some care in interpretation, as I’ll explain. These are observational data so there are inevitable questions about cause and effect. The results also come from self-reported symptoms in a self-selected group of people who submitted data using the Zoe app, who aren’t particularly typical of the UK population as a whole, though on the whole I’m less concerned about that than about other issues.

“In a way, though, I think the most important point made by the researchers isn’t so much the difference between rates of getting long Covid, given that you get infected, during the Delta and the Omicron waves – it’s that what really counts is how many people get infected. Far more people were infected first with Omicron than with Delta. So even if the percentage of infected people who got long Covid during the two waves is on the scale that these researchers report – and it may well be – the actual numbers of people reporting long Covid after first being infected during Omicron is still far larger than during Delta. The potential lower risk of long Covid in people infected during Omicron is entirely trumped by the much bigger number of new infections during the Omicron wave. Anyway, you don’t really have any choice about which virus variant you might be infected with. What’s more, nothing in these findings tells us what might happen with a different new variant, in terms of long Covid risk.

“On the face of it, there does seem to be quite a difference between the proportions of infected people who reported ongoing symptoms four or more weeks after the start of their Covid-19 infection, between those first infected in the period when the Delta variant was dominant and those first infected when Omicron was dominant. The researchers presented the data separately depending on how long it was since their most recent vaccination. (The researchers didn’t have enough data on unvaccinated people to provide any results for that group.) Though they looked separately at those most recently vaccinated more than 6 months ago, and between 3 and 6 months ago, the overall findings for those two groups were pretty similar, so I’ll lump them together. In a group of 1,000 people like those in the study, who were first infected during the Delta period, and had had their most recent vaccination more than 3 months ago, about 125 would have reported symptoms going on for four or more weeks. In another group of 1,000 people who were similar to the other 1,000 in terms of age, gender, the level of deprivation where they lived, whether they had comorbidities, how many vaccine doses they had had, and their BMI, but who had been first infected during the Omicron period, the findings of the study lead to an estimate that between about 25 and 45 would have symptoms lasting at least four weeks. That’s considerably less than 125 – but as I’ll explain, we can’t conclude that the difference is all due to the variant that they were infected with. For those whose most recent vaccination was less than three months ago, the difference isn’t so large but it is still considerable. Out of 1,000 people like those in the study, first infected during Delta after the most recent vaccination less than three months ago, around 75 would report symptoms going on longer than four weeks. In another 1,000 who are similar to the first 1,000 in terms of age, gender, and the other factors I mentioned but were first infected during Omicron, somewhere between about 35 and 45 would report long Covid symptoms. Still a reduction, but not such a large one, and again we can’t be at all certain that it is the difference between variants that caused the difference in long Covid numbers.

“Why we can’t be certain that these differences are caused by the different variants is that this study is observational. There would have been many differences between those who were infected during the Delta and the Omicron waves, apart from the variant that was dominant at the time. So the difference in rates of reporting ongoing symptoms might be caused by these other differences, and not entirely, or not at all, by the difference in variant.

“Of course the researchers understood this, and they made statistical adjustments to allow for differences in some factors, as well as reporting the results separately for different times since the most recent vaccination. The factors they adjusted for are those that I mentioned in the previous paragraph (which is why I mentioned them specifically there) – age, gender, the level of deprivation where they lived, whether they had comorbidities, how many vaccine doses they had had, and their BMI. But there might well have been other factors that differed between the two waves, and that might possibly have affected their chances of reporting long Covid. The researchers can’t adjust statistically for factors on which they don’t have data. The adjustments that they made did change the comparison between the risks reasonably substantially, so maybe other adjustment that they could not make would also have changed things quite a lot – we can’t tell. It’s also possible that the various factors work together in different ways than were reflected in the way that the researchers made the adjustments. Although they took into account both the number of vaccine doses people had and the time since the latest dose, maybe those two factors work together in affecting the chance of long Covid and more complicated way than the researchers considered. So it certainly remains a strong possibility that the difference in the chance of reporting long Covid between the Delta and the Omicron wave is caused, in part at least, by the different variants that probably infected people, but it’s far from being the only possible explanation of what caused what.

“What I’d like to have seen, though the researchers explain that it wasn’t possible, is some comparable results for symptoms lasting longer than the four-week minimum used here – say for 12 weeks or more – because the public health implications of such long-lasting symptoms are important. Also, these finding are only for those aged 18 and over. There’s been a lot of recent concern about long Covid in children, but these new findings don’t tell us anything direct about that.”



‘Risk of long COVID associated with delta versus omicron variants of SARS-CoV-2’ by Anil Dhawan and Sunitha Vimalesvaran was published in the Lancet at 23:30 UK time on Thursday 16 June 2022.



Declared interests

Prof Kevin McConway: “I am a Trustee of the SMC and a member of its Advisory Committee.  My quote above is in my capacity as an independent professional statistician.”

For all other experts, no reply to our request for DOIs was received.

in this section

filter RoundUps by year

search by tag