Risk factors and disease profile of post-vaccination SARS-CoV-2 infection in UK users of the COVID Symptom Study app: a prospective, community-based, nested, case-control study
Summary
Background
COVID-19 vaccines show excellent efficacy in clinical trials and effectiveness in real-world data, but some people still become infected with SARS-CoV-2 after vaccination. This study aimed to identify risk factors for post-vaccination SARS-CoV-2 infection and describe the characteristics of post-vaccination illness.
Methods
This prospective, community-based, nested, case-control study used self-reported data (eg, on demographics, geographical location, health risk factors, and COVID-19 test results, symptoms, and vaccinations) from UK-based, adult (≥18 years) users of the COVID Symptom Study mobile phone app. For the risk factor analysis, cases had received a first or second dose of a COVID-19 vaccine between Dec 8, 2020, and July 4, 2021; had either a positive COVID-19 test at least 14 days after their first vaccination (but before their second; cases 1) or a positive test at least 7 days after their second vaccination (cases 2); and had no positive test before vaccination. Two control groups were selected (who also had not tested positive for SARS-CoV-2 before vaccination): users reporting a negative test at least 14 days after their first vaccination but before their second (controls 1) and users reporting a negative test at least 7 days after their second vaccination (controls 2). Controls 1 and controls 2 were matched (1:1) with cases 1 and cases 2, respectively, by the date of the post-vaccination test, health-care worker status, and sex. In the disease profile analysis, we sub-selected participants from cases 1 and cases 2 who had used the app for at least 14 consecutive days after testing positive for SARS-CoV-2 (cases 3 and cases 4, respectively). Controls 3 and controls 4 were unvaccinated participants reporting a positive SARS-CoV-2 test who had used the app for at least 14 consecutive days after the test, and were matched (1:1) with cases 3 and 4, respectively, by the date of the positive test, health-care worker status, sex, body-mass index (BMI), and age. We used univariate logistic regression models (adjusted for age, BMI, and sex) to analyse the associations between risk factors and post-vaccination infection, and the associations of individual symptoms, overall disease duration, and disease severity with vaccination status.
Findings
Between Dec 8, 2020, and July 4, 2021, 1 240 009 COVID Symptom Study app users reported a first vaccine dose, of whom 6030 (0·5%) subsequently tested positive for SARS-CoV-2 (cases 1), and 971 504 reported a second dose, of whom 2370 (0·2%) subsequently tested positive for SARS-CoV-2 (cases 2). In the risk factor analysis, frailty was associated with post-vaccination infection in older adults (≥60 years) after their first vaccine dose (odds ratio [OR] 1·93, 95% CI 1·50–2·48; p<0·0001), and individuals living in highly deprived areas had increased odds of post-vaccination infection following their first vaccine dose (OR 1·11, 95% CI 1·01–1·23; p=0·039). Individuals without obesity (BMI <30 kg/m2) had lower odds of infection following their first vaccine dose (OR 0·84, 95% CI 0·75–0·94; p=0·0030). For the disease profile analysis, 3825 users from cases 1 were included in cases 3 and 906 users from cases 2 were included in cases 4. Vaccination (compared with no vaccination) was associated with reduced odds of hospitalisation or having more than five symptoms in the first week of illness following the first or second dose, and long-duration (≥28 days) symptoms following the second dose. Almost all symptoms were reported less frequently in infected vaccinated individuals than in infected unvaccinated individuals, and vaccinated participants were more likely to be completely asymptomatic, especially if they were 60 years or older.
Interpretation
To minimise SARS-CoV-2 infection, at-risk populations must be targeted in efforts to boost vaccine effectiveness and infection control measures. Our findings might support caution around relaxing physical distancing and other personal protective measures in the post-vaccination era, particularly around frail older adults and individuals living in more deprived areas, even if these individuals are vaccinated, and might have implications for strategies such as booster vaccinations.
Funding
ZOE, the UK Government Department of Health and Social Care, the Wellcome Trust, the UK Engineering and Physical Sciences Research Council, UK Research and Innovation London Medical Imaging and Artificial Intelligence Centre for Value Based Healthcare, the UK National Institute for Health Research, the UK Medical Research Council, the British Heart Foundation, and the Alzheimer’s Society.
Introduction
- Ledford H
- Cyranoski D
- Van Noorden R
,
- Baden LR
- El Sahly HM
- Essink B
- et al.
,
- Polack FP
- Thomas SJ
- Kitchin N
- et al.
,
- Voysey M
- Clemens SAC
- Madhi SA
- et al.
Of the 67·1 million adults in the UK, by July 4, 2021, around 45·4 million people had received one vaccine dose and around 33·7 million people had received two doses.
People who have received 1st dose vaccinations, by report date.
UK data present an early insight into the real-world effectiveness of COVID-19 vaccines and the remaining challenges post-vaccination.
- Menni C
- Klaser K
- May A
- et al.
findings that were recapitulated in a UK-based, real-world, case-control study.
- Lopez Bernal J
- Andrews N
- Gower C
- et al.
National surveillance data from the first 4 months of Israel’s vaccination campaign showed that two doses of BNT162b2 reduced both symptomatic and asymptomatic infections, COVID-19-related hospitalisations, severe disease, and death.
- Haas EJ
- Angulo FJ
- McLaughlin JM
- et al.
- Graham MS
- Sudre CH
- May A
- et al.
Indeed, variants of concern have shown reduced neutralisation by convalescent and post-vaccination serum samples in vitro,
- Kustin T
- Harel N
- Finkel U
- et al.
and led to increased rates of post-vaccination infection compared with the original outbreak variant in early findings from a real-world case-control study.
- McEwen AE
- Cohen S
- Bryson-Cahn C
- et al.
Data suggest that although COVID-19 is usually milder if contracted after vaccination than in unvaccinated individuals, mortality remains high in hospitalised individuals: data from the International Severe Acute Respiratory and Emerging Infection Consortium have shown a mortality of 27·0% (400 of 1482 died) in individuals hospitalised with COVID-19 in the UK more than 21 days after vaccination, similar to mortality rates observed during the first wave (March–April, 2020).
- Egan C
- Knight S
- Baillie K
- Harrison E
- Docherty A
- Semple C
,
- Docherty AB
- Mulholland RH
- Lone NI
- et al.
Evidence before this study
To identify existing evidence of risk factors for, and the characteristics of, post-vaccination SARS-CoV-2 infection, we searched PubMed for peer-reviewed articles published between Dec 1, 2020, and July 4, 2021, using the keywords (“COVID-19” OR “SARS-CoV-2”) AND (“vaccine” OR “vaccination”) AND (“infection”) AND (“risk factor*” OR “characteristic*” OR “symptom*”). We did not restrict our search by language or type of publication. Of the 662 published PubMed articles identified, we found one paper assessing risk factors for post-vaccination infection in fully vaccinated US veterans. This paper suggested that age and the presence of anaemia were positively associated, and being Black was negatively associated, with post-vaccination infection. We found no studies looking at risk factors following a single dose of vaccine or the disease profile in vaccinated, community-based individuals. Previous studies in unvaccinated populations have shown that social and occupational factors influence the risk of SARS-CoV-2 infection, and that personal factors (eg, age, male sex, multiple morbidities, and frailty) increase the risk for adverse outcomes in COVID-19. Phase 3 clinical trials have shown good efficacy for BNT162b2 (tozinameran; Pfizer–BioNTech), ChAdOx1 nCoV-19 (Oxford–AstraZeneca), and mRNA-1273 (elasomeran; Moderna) vaccines against SARS-CoV-2 infection, supported by published real-world data, with vaccines also reducing the risk of adverse outcomes, including hospitalisation and death. We found two published studies of in-vitro and early real-world evidence suggesting that BNT162b2 and ChAdOx1 nCoV-19 vaccines might be less effective against emerging variants of concern than against the original outbreak variant.
Added value of this study
To our knowledge, this observational study was the first to investigate the characteristics of SARS-CoV-2 infection after first and second COVID-19 vaccinations. Vaccination (compared with no vaccination) was associated with reduced odds of hospitalisation or having more than five symptoms in the first week of illness following the first or second dose, and long-duration (≥28 days) symptoms following the second dose. Almost all symptoms were reported less frequently in infected vaccinated than in infected unvaccinated individuals, and vaccinated participants were more likely to be completely asymptomatic, especially if they were 60 years or older. In our risk factor analysis, we found that frailty was associated with post-vaccination infection in older adults (≥60 years) after their first vaccine dose. Adverse determinants of health, such as living in highly deprived areas and obesity, were associated with an increased likelihood of SARS-CoV-2 infection following the first vaccine dose.
Implications of all the available evidence
Some individuals still become infected with SARS-CoV-2 after vaccination; our data suggest that frail, older adults and those living in more deprived areas are at increased risk. However, COVID-19 appears to be less severe in vaccinated versus unvaccinated individuals. Our results are relevant for health policies post-vaccination and highlight the need to balance personal protective measures in those at risk of post-vaccination infection with the adverse effects from ongoing social restrictions. Strategies, such as timely prioritisation of booster vaccinations and optimised infection control measures, could be considered for at-risk groups. Research is also needed on how to enhance the immune response to vaccination in those at higher risk of post-vaccination infection.
- Nguyen LH
- Drew DA
- Graham MS
- et al.
,
- Bowyer RCE
- Varsavsky T
- Thompson EJ
- et al.
and increasing age, male sex, multimorbidity, and frailty are associated with poorer COVID-19 outcomes.
- de Lusignan S
- Dorward J
- Correa A
- et al.
,
- Hewitt J
- Carter B
- Vilches-Moraga A
- et al.
,
- Williamson EJ
- Walker AJ
- Bhaskaran K
- et al.
A study of fully vaccinated US veterans showed that older age and the presence of anaemia were positively associated with post-vaccination infection, and Black individuals were at a lower risk than White individuals.
- Butt AA
- Khan T
- Yan P
- Shaikh OS
- Omer SB
- Mayr F
However, this study was done in a fully vaccinated, older (median age 73 years [IQR 68–78]), and predominantly male cohort, and did not assess lifestyle and sociodemographic risk factors for post-vaccination infection.
- Sudre CH
- Lee KA
- Lochlainn MN
- et al.
Elucidating symptom profiles in individuals with COVID-19 after vaccination has clinical utility, facilitating the identification of risk groups for intervention, predicting medical resource requirements, and informing appropriate testing guidelines. Additionally, some unvaccinated individuals with COVID-19 have prolonged illness duration (so-called long COVID),
- Sudre CH
- Murray B
- Varsavsky T
- et al.
and whether vaccination reduces the risk of long COVID is currently unknown.
Therefore, we aimed to (1) describe individual risk factors associated with SARS-CoV-2 infection at least 14 days after first vaccination or 7 days after second vaccination, and (2) assess illness duration, severity, and symptom profile in individuals with SARS-CoV-2 infection after their first and second vaccinations, compared with unvaccinated individuals with SARS-CoV-2 infection.
Methods
Study design and participants
- Menni C
- Valdes AM
- Freidin MB
- et al.
The app was launched in the UK on March 24, 2020, and by July 4, 2021, had nearly 4·5 million unique participants providing data by self-report or proxy report. At registration, each participant reported baseline demographic information (eg, age, sex, ethnicity, weight, height, and health-care worker status), geographical location, and information on health risk factors, including comorbidities, lifestyle, frailty, visits to hospital, and adherence to mask-wearing guidance. Participants were encouraged to self-report any of 32 prespecified symptoms (appendix p 1) daily, providing prospective longitudinal information on incident symptoms. Those with new symptoms were prompted to book and take a SARS-CoV-2 test. All users were encouraged to record any SARS-CoV-2 testing results (whether prompted by the app or otherwise), and, from Dec 11, 2020, any SARS-CoV-2 vaccination and subsequent symptoms.
- Menni C
- Valdes AM
- Freidin MB
- et al.
Users with missing or inconsistent information were excluded from our analysis. The inclusion process for cases and controls is shown in the appendix (p 14).
Cases had received a first or second dose of a COVID-19 vaccine since Dec 8, 2020; had either a positive RT-PCR test or lateral flow antigen test (LFAT) at least 14 days after their first vaccination (but before their second; cases 1) or a positive RT-PCR test or LFAT at least 7 days after their second vaccination (cases 2); and had no positive SARS-CoV-2 test before vaccination. If more than one positive test result was reported, only the first positive test was selected. To identify risk factors for post-vaccination infection, we selected two control groups among the vaccinated (since Dec 8, 2020) UK-based adult users of the COVID Symptom Study app who had not tested positive for SARS-CoV-2 before vaccination: a control group of users reporting a negative RT-PCR test or LFAT at least 14 days after their first vaccination but before their second (controls 1) and a control group of users reporting a negative RT-PCR test or LFAT at least 7 days after their second vaccination (controls 2). Controls 1 and controls 2 were matched (1:1) with cases 1 and cases 2, respectively, by use of the date of the post-vaccination COVID-19 test, health-care worker status, and sex. If multiple negative tests were reported, the last test date was used for matching.
To compare the disease profile of SARS-CoV-2 infection before and after vaccination, we sub-selected participants from cases 1 and cases 2 who had used the app for at least 14 consecutive days after testing positive for SARS-CoV-2 (denoted as cases 3 and cases 4, respectively), so that symptoms of infection could be assessed. Controls for the disease profile analysis were those who reported a SARS-CoV-2-positive RT-PCR test or LFAT, were unvaccinated until data censoring, and had used the app for at least 14 consecutive days after the test. Among these users, two groups were formed: controls 3 and controls 4, matching (1:1) with cases 3 and cases 4, respectively, by the date of the positive COVID-19 test, health-care worker status, sex, body-mass index (BMI), and age. Individuals in all case and control groups who did not report an RT-PCR or LFAT test after Dec 8, 2020, were excluded.
- Spiel C
- Lapka D
- Gradinger P
- et al.
between the vectors of the covariates, with age, BMI, and the date of the test as numerical variables, and sex as a binary variable multiplied by 100 to ensure balance between covariate strengths. We considered health-care worker status, coded as a a categorical variable in the app, as a numerical variable (0=not a health-care worker; 1=health-care worker who does not interact with patients; 2=health-care worker who does not treat patients; 3=health-care worker who interacts with patients; 4=health-care worker who treats patients). Participants could only choose one of these options.
All app users provided digital informed consent for data usage for COVID-19-related research. In the UK, the app and the study were approved by King’s College London’s ethics committee (REMAS number 18210; reference LRS-19/20–18210).
Risk factor variable definitions
- de Lusignan S
- Dorward J
- Correa A
- et al.
,
- Hewitt J
- Carter B
- Vilches-Moraga A
- et al.
,
- Williamson EJ
- Walker AJ
- Bhaskaran K
- et al.
age; BMI; self-reported comorbidities (ie, cancer, diabetes, asthma, lung disease, heart disease, and kidney disease), analysed individually as binary variables; dependency level (frailty) assessed by the PRISMA-7 questionnaire, which is embedded in app registration,
- Raîche M
- Hébert R
- Dubois MF
,
- Zazzara MB
- Penfold RS
- Roberts AL
- et al.
as a binary variable (PRISMA-7 score ≥3 defined as frail and
- Hoffmann S
- Wiben A
- Kruse M
- Jacobsen KK
- Lembeck MA
- Holm EA
local area Index of Multiple Deprivation (IMD; a score ranging from 1 [most deprived] to 10 [least deprived] estimating relative locality deprivation derived from postal code and lower layer super output area) divided into low IMD (1–3), intermediate IMD (4–7), and high IMD (8–10) groups;
Ministry of Housing, Communities & Local Government. National statistics. English indices of deprivation 2019.
and four healthy lifestyle factors (no current smoking, no obesity [BMI 2], physical activity at least once per week [non-sedentary], and a healthy diet pattern; appendix p 17). We also calculated a healthy lifestyle score on the basis of these four lifestyle factors,
- Chiuve SE
- Cook NR
- Shay CM
- et al.
by which participants received 1 point for each healthy lifestyle factor and the sum of the scores gave a total healthy lifestyle score ranging from 0 to 4, with higher scores indicating a healthier lifestyle (appendix p 17).
Disease severity, duration, and symptom definitions
- Sudre CH
- Lee KA
- Lochlainn MN
- et al.
and self-reported presentation to hospital or no hospital presentation), illness duration (duration of appendix p 1). This window was used because it might have taken up to 3 days to request an RT-PCR test and receive a result following symptom onset, and symptoms can occur up to 14 days following SARS-CoV-2 exposure.
Symptoms of COVID-19.
Statistical analysis
- Murray B
- Kerfoot E
- Graham MS
- et al.
that is openly available on GitHub. Statistical analyses used Python 3.7 and the packages NumPy (version 1.19.2), Pandas (version 1.1.3), SciPy (version 1.5.2), and statsmodels (version 0.12.1).
In the risk factor analysis, we assessed the differences in proportions and means of covariates between cases and respective controls using Fisher’s exact test for categorical variables and Wilcoxon’s test for continuous variables. p values of 0·05 or less were considered statistically significant. Univariate logistic regression models (adjusted for age, BMI, and sex) and multivariate logistic regression models of age and BMI (adjusted for sex) were used to analyse the associations between risk factor variables and post-vaccination infection. Because associated factors might differ by age group, analyses were stratified by sex (male and female) and age (younger adults were those aged 18–59 years and older adults were those aged ≥60 years). To examine whether health-conscious behaviours might explain the association between lifestyle factors and post-vaccination infection, we further adjusted models for reported individual adherence to mask-wearing guidance from June 12, 2020, to Sept 29, 2020. Multivariate logistic regression (adjusted for age, BMI, and sex) was used to assess the independence of frailty, IMD category, and the four healthy lifestyle factors.
to check for potential index event bias of vaccination using weights derived from probabilities of being vaccinated in the population tested and active on the app during the study period (appendix p 15). Data from 1 531 762 app users reporting an RT-PCR or LFAT test within the study period were processed to obtain weights for inverse probability weighting of being vaccinated. Weights were estimated from a logistic regression model for predicting vaccination, which included confounders known to be associated with vaccination status: frailty, IMD, and the comorbidities of cancer, diabetes, lung disease, heart disease, kidney disease, and asthma.
In the disease profile analysis, univariate logistic regression models adjusted by age, BMI, and sex were used to assess the association of individual symptoms, overall illness duration, and disease severity (outcomes) with vaccination status (exposure). Symptoms were examined if they were reported by more than 1% of app users reporting a positive test. We also provide models that were adjusted for frailty and the presence of at least one comorbidity, given the association of these factors with the exposure (vaccination) and outcome (symptoms), which could confound observed associations.
For all regression analyses, odds ratios (ORs) and 95% CIs were calculated. Analyses were not corrected for multiple testing. This study reports on vaccination with BNT162b2, ChAdOx1 nCoV-19, and mRNA-1273 only, as there were no positive cases among the few people who had received other vaccines.
Role of the funding source
The funders of the study had no role in study design, data analysis, data interpretation, or writing of the report. ZOE, funded by the UK Department of Health and Social Care, made the COVID Symptom Study app available for data collection as a not-for-profit endeavour. Representatives of ZOE approved the final manuscript for submission.
Results
Table 1Characteristics of case participants with COVID-19 after their first or second vaccination and vaccinated control participants without COVID-19 in the risk factor analysis
Data are n (%), n (%); p value, n/N (%), n/N (%); p value, or mean (SD). The p values indicate the difference between cases 1 and controls 1, or between cases 2 and controls 2, and were not adjusted for multiple testing.
Figure 1Univariate analysis of post-vaccination SARS-CoV-2 infection risk factors
Univariate models for frailty and each individual comorbidity (A) and IMD, healthy lifestyle factors, and healthy lifestyle score (B), adjusted for age, body-mass index, and sex, and stratified by age group. The error bars represent 95% CIs. A low IMD means high deprivation and a high IMD means low deprivation. The reference category for the IMD is an intermediate IMD (4–7). IMD=Index of Multiple Deprivation.
Figure 2Multivariate analysis of post-vaccination SARS-CoV-2 infection risk factors
The multivariate analysis was adjusted for age, body-mass index, and sex, and was stratified by age group. The error bars represent 95% CIs. A low IMD means high deprivation and a high IMD means low deprivation. The reference category for the IMD is an intermediate IMD (4–7). IMD=Index of Multiple Deprivation.
Table 2Characteristics of case participants with COVID-19 after their first or second vaccination and unvaccinated control participants with COVID-19 in the disease profile analysis
Data are n (%), n (%); p value, n/N (%), n/N (%); p value, or mean (SD). The p values indicate the difference between cases 3 and controls 3, or between cases 4 and controls 4, and were not adjusted for multiple testing.
Figure 3Disease severity and duration factors in SARS-CoV-2-infected vaccinated versus unvaccinated participants
Univariate models were adjusted for age, body-mass index, and sex, and stratified by age group. The error bars represent 95% CIs.
Figure 4Symptoms in SARS-CoV-2-infected vaccinated versus unvaccinated participants
Univariate models were adjusted for age, body-mass index, and sex, and stratified by age group. The error bars represent 95% CIs. We present only symptoms reported by more than 1% of users.
Discussion
- Shrotri M
- Navaratnam AMD
- Nguyen V
- et al.
and infection was unlikely to be due to exposure around the time of vaccination (eg, when travelling to the vaccination centre).
- Baden LR
- El Sahly HM
- Essink B
- et al.
,
- Polack FP
- Thomas SJ
- Kitchin N
- et al.
,
- Voysey M
- Clemens SAC
- Madhi SA
- et al.
,
- Menni C
- Klaser K
- May A
- et al.
,
- Lopez Bernal J
- Andrews N
- Gower C
- et al.
,
- Haas EJ
- Angulo FJ
- McLaughlin JM
- et al.
Coronavirus (COVID-19): getting tested.
We also found that COVID-19 was less severe (both in terms of the number of symptoms in the first week of infection and the need for hospitalisation) in participants after their first or second vaccine doses compared with unvaccinated participants. We have previously shown that having more than five symptoms in the first week of infection was associated with COVID-19 severity
- Sudre CH
- Lee KA
- Lochlainn MN
- et al.
and disease duration.
- Sudre CH
- Murray B
- Varsavsky T
- et al.
- Hewitt J
- Carter B
- Vilches-Moraga A
- et al.
Another explanation for this result concerns altered immune function (immunosenescence), a well established feature of physiological ageing.
- Cox LS
- Bellantuono I
- Lord JM
- et al.
,
The increased odds of post-vaccination infection in frail older adults could be compounded by the more severe outcomes of COVID-19 in this group, including delirium
- Zazzara MB
- Penfold RS
- Roberts AL
- et al.
and death;
- Hewitt J
- Carter B
- Vilches-Moraga A
- et al.
indeed, in our study, 23% of frail, older adults testing positive for SARS-CoV-2 after their first vaccination presented to hospital. Systematic frailty screening in acute and community-based settings might facilitate differential, targeted re-vaccination scheduling, appropriate isolation precautions, case detection, testing, and proactive care, as recommended in guidance published by the National Institute for Health and Care Excellence
COVID-19 rapid guideline: managing COVID-19.
and National Health Service (NHS) England.
Identifying frailty.
Research on augmenting immunogenicity in frail individuals is needed, such as on the impact and timing of booster vaccinations.
- Ward H
- Atchison C
- Whitaker M
- et al.
perhaps reflecting shielding in this age group in accordance with the classification of individuals older than 70 years as clinically vulnerable. Our study found some evidence to suggest that kidney disease might increase the odds of SARS-CoV-2 infection in older adults after their first vaccine dose, which is notable given that individuals with kidney disease were under-represented in the phase 2 and phase 3 trials of the COVID-19 vaccines.
- Glenn DA
- Hegde A
- Kotzen E
- et al.
However, this finding should be interpreted cautiously because of the relatively small numbers of participants with kidney disease in this study. This increased risk of post-vaccination infection for people with kidney disease could reflect increased exposure (eg, when attending dialysis appointments) or impaired immunogenicity, and is supported by a study looking at humoral and B-cell responses in vaccinated, immunosuppressed kidney transplant recipients and patients having dialysis.
- Rincon-Arevalo H
- Choi M
- Stefanski AL
- et al.
Several other comorbidities, including heart disease and lung disease, were significantly associated with post-vaccination infection after one dose in older adults; although associations of marginal significance should be interpreted cautiously, many of these comorbidities confer a higher risk of severe disease, hospitalisation, mechanical ventilation, and mortality from COVID-19,
- de Lusignan S
- Dorward J
- Correa A
- et al.
,
- Williamson EJ
- Walker AJ
- Bhaskaran K
- et al.
and ongoing shielding behaviours could be influencing our results to reduce the strength of these associations.
- Bowyer RCE
- Varsavsky T
- Thompson EJ
- et al.
This association persisted following further adjustment for compliance with infection control guidance (ie, mask wearing). Factors associated with increased area-level deprivation include higher population density and more ethnic diversity, which are in themselves associated with increased mortality from COVID-19.
- Williamson EJ
- Walker AJ
- Bhaskaran K
- et al.
More deprived areas might have lower vaccination coverage for COVID-19,
Coronavirus and vaccine hesitancy, Great Britain: 13 January to 7 February 2021.
and our finding might reflect increased viral transmission. Our results suggest that health policies to mitigate infection might need to specifically target these areas. Conversely, individuals without obesity had a lower odds of infection following their first vaccine dose. This finding suggests that immune responses post-vaccination might be influenced by obesity, although unadjusted confounding remains a possibility.
- Sheikh A
- McMenamin J
- Taylor B
- Robertson C
,
- Lopez Bernal J
- Andrews N
- Gower C
- et al.
however, our observation should be treated cautiously given the confounding factors influencing the vaccine type administered in different age groups and demographics. We emphasise that our study is observational, rather than a formal comparison.
Our study has some limitations. Although we used data from a large population of individuals reporting on a mobile phone app, the sample contained disproportionately more women than men and under-represented individuals in more deprived areas. Furthermore, we were unable to analyse the impact of ethnicity due to the low number of participants who provided this information, and our findings might not apply at all timepoints post-vaccination, to settings with different proportions of SARS-CoV-2 variants, or to countries with a different vaccination schedule. Additionally, the data were self-reported; recording of comorbidities, test results, and vaccination status might not have been completely accurate and there might have been temporal gaps in reporting. Users of the COVID Symptom Study app are asked to log daily; therefore, if a participant reports on alternate days, the proportion of missing daily entries is 50%. However, given the typical duration of COVID-19 symptoms, the sampling frequencies in the COVID Symptom Study should have allowed good characterisation of infections.
- Varsavsky T
- Graham MS
- Canas LS
- et al.
including on the influence of sociodemographic factors.
- Bowyer RCE
- Varsavsky T
- Thompson EJ
- et al.
The mobile phone data collection method allows the collection of daily prospective information on a comprehensive set of symptoms, permitting the analysis of both individual symptoms and overall illness duration (although necessary data censoring could have underestimated symptom duration in both cases and controls, as some individuals only had 2 weeks of logging after their positive test result).
Coronavirus (COVID-19): getting tested.
The risk of reporting a positive SARS-CoV-2 test is higher among frontline health-care workers than among the general population,
- Nguyen LH
- Drew DA
- Graham MS
- et al.
probably reflecting increased exposure and testing. After the first vaccine dose, positive tests were by RT-PCR in a higher proportion of vaccinated cases than in unvaccinated controls. Because RT-PCR is used for symptomatic testing in the UK, this would bias away our finding of a higher likelihood of asymptomatic or minimally symptomatic COVID-19 in vaccinated versus unvaccinated participants. However, there is uncertainty due to the high numbers of unvaccinated controls who reported an unknown test type after their first vaccine dose.
Joint Committee on Vaccination and Immunisation: advice on priority groups for COVID-19 vaccination, 30 December 2020.
We examined and found no evidence of event bias on the basis of the probability of being vaccinated. Analyses in this study were not corrected for multiple testing, and so observations of marginal significance should be interpreted cautiously.
- O’Caoimh R
- Costello M
- Small C
- et al.
and has the advantage of focusing on the functional consequences of frailty, which are not routinely captured in health records. However, PRISMA-7 has only been validated in older adults; results in younger adults should be interpreted cautiously.
- Raîche M
- Hébert R
- Dubois MF
To conclude, the odds of post-vaccination infection following the first dose were increased in frail, older adults and in those living in more deprived areas, and were decreased in individuals without obesity. Compared with unvaccinated controls, after their second vaccine dose, individuals were less likely to have prolonged illness (symptoms for ≥28 days), more than five symptoms in the first week of illness, or present to hospital. Most symptoms were less common in vaccinated versus unvaccinated participants. Fully vaccinated individuals with COVID-19, especially if they were 60 years or older, were more likely to be completely asymptomatic than were unvaccinated controls. This finding might support caution around relaxing physical distancing and other personal protective measures in the post-vaccination era, particularly around frail older adults and individuals living in more deprived areas, even if these individuals are vaccinated. Our findings might also have implications for strategies such as booster vaccinations.
All authors were responsible for aspects of study design, data collection, data analysis, and manuscript writing. CJS conceived the study and designed the analysis plan with MA, CHS, and JM. MA did the data analysis. BM produced the data cleaning script for the app dataset. RSP did the background literature search and review. MA and CJS accessed and verified the underlying data. All authors contributed to and reviewed the final submitted manuscript. All authors had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Declaration of interests
JW, AM, LP, CH, SS, and JCP report being employees of ZOE during the conduct of the study. JM reports grants from the European Commission and National Institutes of Health, and has served as a co-investigator on an unrelated nutrition trial sponsored by ZOE. ATC reports grants from the Massachusetts Consortium on Pathogen Readiness during the conduct of the study, and personal fees from Bayer Pharma, Pfizer, and Boehringer Ingelheim, outside the submitted work. DAD reports grants from the National Institutes of Health, the Massachusetts Consortium on Pathogen Readiness, and the American Gastroenterological Association during the conduct of the study, and has served as a co-investigator on an unrelated nutrition trial sponsored by ZOE. LHN reports grants from the National Institutes of Health, the American Gastroenterological Association, and the Crohn’s and Colitis Foundation. CHS reports grants from the Alzheimer’s Society during the conduct of the study. EM reports a grant from the Medical Research Council during the conduct of the study. CJS reports grants from the Chronic Disease Research Foundation, the Medical Research Council, the Wellcome Trust, and the National Institute for Health Research (NIHR) during the conduct of the study. SO reports grants from the Wellcome Trust, UK Research and Innovation, and the Chronic Disease Research Foundation during the conduct of the study. ELD reports receiving grants from the Chronic Disease Research Foundation during the conduct of the study. TDS reports being a consultant for ZOE during the conduct of the study. All other authors declare no competing interests.