Trajectory of long covid symptoms after covid-19 vaccination: community based cohort study

Abstract

Objective: To estimate associations between covid-19 vaccination and long covid symptoms in adults with SARS-CoV-2 infection before vaccination.

Design: Observational cohort study.

Setting: Community dwelling population, UK.

Participants: 28 356 participants in the Office for National Statistics COVID-19 Infection Survey aged 18-69 years who received at least one dose of an adenovirus vector or mRNA covid-19 vaccine after testing positive for SARS-CoV-2 infection.

Main outcome measure: Presence of long covid symptoms at least 12 weeks after infection over the follow-up period 3 February to 5 September 2021.

Results: Mean age of participants was 46 years, 55.6% (n=15 760) were women, and 88.7% (n=25 141) were of white ethnicity. Median follow-up was 141 days from first vaccination (among all participants) and 67 days from second vaccination (83.8% of participants). 6729 participants (23.7%) reported long covid symptoms of any severity at least once during follow-up. A first vaccine dose was associated with an initial 12.8% decrease (95% confidence interval -18.6% to -6.6%, P<0.001) in the odds of long covid, with subsequent data compatible with both increases and decreases in the trajectory (0.3% per week, 95% confidence interval -0.6% to 1.2% per week, P=0.51). A second dose was associated with an initial 8.8% decrease (95% confidence interval -14.1% to -3.1%, P=0.003) in the odds of long covid, with a subsequent decrease by 0.8% per week (-1.2% to -0.4% per week, P<0.001). Heterogeneity was not found in associations between vaccination and long covid by sociodemographic characteristics, health status, hospital admission with acute covid-19, vaccine type (adenovirus vector or mRNA), or duration from SARS-CoV-2 infection to vaccination.

Conclusions: The likelihood of long covid symptoms was observed to decrease after covid-19 vaccination and evidence suggested sustained improvement after a second dose, at least over the median follow-up of 67 days. Vaccination may contribute to a reduction in the population health burden of long covid, although longer follow-up is needed.

Fig 1

Study participant flow diagram. CIS=Office for National Statistics COVID-19 Infection Survey

Fig 2

Modelled probabilities of long covid for a hypothetical study participant who received a first covid-19 vaccine dose 24 weeks after SARS-CoV-2 infection and a second dose 12 weeks later. Probabilities are shown for participants of mean age (50 years) and in the modal group for other covariates (woman, white ethnicity, resident in London, resident in an area in the least deprived fifth group, not a patient-facing health or social care worker, no pre-existing health conditions, not admitted to hospital during the acute phase of infection, infected on 7 September 2020). Although estimated probabilities are specific to this profile, proportional changes in probabilities after vaccination do not vary across characteristics and can therefore be generalised to other profiles. Dashed lines represent timing of vaccination. Shaded areas are 95% confidence intervals

Fig 3

Modelled probabilities of long covid for a hypothetical study participant who received a first dose of an adenovirus vector or mRNA vaccine 24 weeks after SARS-CoV-2 infection and a second dose 12 weeks later. Probabilities are shown for participants of mean age (50 years) and in the modal group for other covariates (woman, white ethnicity, resident in London, resident in an area in the least deprived fifth group, not a patient-facing health or social care worker, no pre-existing health conditions, not admitted to hospital during the acute phase of infection, infected on 7 September 2020). Although estimated probabilities are specific to this profile, proportional changes in probabilities after vaccination do not vary across characteristics and can therefore be generalised to other profiles. Dashed lines represent timing of vaccination. Shaded areas are 95% confidence intervals

Fig 4

Modelled probabilities of individual long covid symptoms for a hypothetical study participant who received a first dose of a covid-19 vaccine 24 weeks after SARS-CoV-2 infection and a second dose 12 weeks later. Top 10 most frequently reported symptoms ordered by modelled probability at 12 weeks post-infection. Probabilities are shown for a participant of mean age (50 years) and in the modal group for other covariates (woman, white ethnicity, resident in London, resident in an area in the least deprived fifth group, not a patient-facing health or social care worker, no pre-existing health conditions, not admitted to hospital during the acute phase of infection, infected on 7 September 2020). Although the estimated probabilities are specific to this profile, proportional changes in probabilities after vaccination do not vary across characteristics and can therefore be generalised to other profiles. Dashed lines represent timing of vaccination. Shaded areas are 95% confidence intervals