Population antibody responses following COVID-19 vaccination in 212,102 individuals

Abstract

Population antibody surveillance helps track immune responses to COVID-19 vaccinations at scale, and identify host factors that may affect antibody production. We analyse data from 212,102 vaccinated individuals within the REACT-2 programme in England, which uses self-administered lateral flow antibody tests in sequential cross-sectional community samples; 71,923 (33.9%) received at least one dose of BNT162b2 vaccine and 139,067 (65.6%) received ChAdOx1. For both vaccines, antibody positivity peaks 4-5 weeks after first dose and then declines. At least 21 days after second dose of BNT162b2, close to 100% of respondents test positive, while for ChAdOx1, this is significantly reduced, particularly in the oldest age groups (72.7% [70.9-74.4] at ages 75 years and above). For both vaccines, antibody positivity decreases with age, and is higher in females and those with previous infection. Antibody positivity is lower in transplant recipients, obese individuals, smokers and those with specific comorbidities. These groups will benefit from additional vaccine doses.

Fig. 1. Vaccine response (upper panel) and uptake (lower panel) over time.

Participants are grouped by weeks since first and second vaccination, and secondarily by vaccine received. The proportion of tests reported as positive (antibody positivity) within group by week is shown. Participants who had received either (i) a single dose but no second dose, or (ii) two doses of the vaccine between 10 and 12 weeks after the first were included in the plot. Shaded areas on either side of the plot lines denote 95% confidence intervals; the grey shaded block denotes the 10–12 week period after first vaccination. The upward trend post-second-dose observed in the all participants line is attributable to the changing proportions of ChAdOx1vs BNT162b2 vaccine after 17 weeks.

Fig. 2. Forest plot showing odds ratios and 95% confidence intervals for antibody positivity (yes/no) versus biological and behavioural covariates in logistic regression models, among 68,060 respondents who had received two vaccine doses with the second dose at least 21 days prior.

Dark blue CIs indicate models adjusted on age and sex only; light blue CIs indicate models adjusted on age, sex, ethnicity, adiposity, vaccine type (in the right panels only), prior infection, shielding status, comorbidities and smoking status.