Summary of findings

Symptoms and transmission of COVID-19
Immunity
Risk factors and exposure
Inequalities
Long COVID

  1. Symptoms and transmission of COVID-19

How well does Virus Watch track COVID-19 infection rate trends in England and Wales?

We compared prevalence and incidence rates estimated using Virus Watch data with those from the Office for National Statistics (ONS) COVID-19 Infection Survey (CIS), the UK’s largest regular survey of COVID-19 infections and antibodies. We found that Virus Watch- and ONS-estimated incidence rates showed similar trends for England and Wales. However, the Virus Watch-estimated peaks of infection during the Omicron BA.1 and 2 waves were lower than the ONS estimates. This highlights that the Virus Watch research approach is a low-cost and reliable method for COVID-19 surveillance, even in the absence of national testing in the UK. 

Which symptoms are more common in people who tested positive for COVID-19 compared to those with a COVID-negative infection during the pandemic? Can we tell whether an illness is COVID-19 based on symptoms alone?

We found that certain symptoms, like a high fever and changes in taste or smell, were more common in people who tested positive for COVID-19 compared to those with a respiratory infection who tested negative. However, the most frequent symptoms of COVID-19—such as cough, headache, fatigue, muscle aches, and loss of appetite—were also very common in other illnesses. Because these general symptoms overlap with many other infections, it’s hard to tell if someone has COVID-19 based on symptoms alone.

How do the symptoms experienced during infections with influenza, RSV, rhinovirus, seasonal coronavirus, and COVID-19 variants differ from one another?

We found that loss of taste and smell was more common with COVID-19, while cough and congestion were more frequent with other respiratory viruses. Differences were also seen between COVID-19 variants. Notably, the symptom profile of COVID-19 illnesses gradually became more similar to other respiratory viruses over the course of the pandemic, with later variants having increased frequency of cough, sneezing, runny nose and sore throat. 

How fast does COVID-19 spread in a household and did the ‘Alpha’ variant that caused the 2020/21 winter wave spread faster than previous variants?

On average, it took around 3 days for symptoms to appear in someone who likely caught the virus from another household member. We also found that the Alpha variant did not seem to spread any faster than previous variants of the virus.

During household outbreaks of COVID-19 (i.e. multiple people becoming ill within a household), how often did the illness transmit between household members, and how often was it caught separately outside of the household? 

Using novel methods of looking at the genome of the SARS-CoV-2 virus, we identified how often COVID-19 transmitted from one household member to another compared to how often multiple household members caught COVID-19 outside of the household at the same time. Out of the 73 participating households that experienced a household outbreak, we found that infections came from outside sources – rather than spreading within the home- in only eight households. This means that infections spread between household members most of the time.

  1. Immunity

How effective is a first and second dose of COVID-19 vaccine at making antibodies against the COVID-19 virus?

In our study, a single dose of AstraZeneca or Pfizer vaccines led to high levels of antibodies against the virus. However, the level of antibodies varied depending on the vaccine type, the person’s age, and whether they had certain health conditions. This means that even after one dose, measures like social distancing are still important. After the second dose, antibody levels increased significantly across all groups, providing strong protection against the virus. 

How long do antibodies against the COVID-19 virus last after a second dose of vaccine, and do antibody levels predict risk of infection?

We measured antibody levels in participants who received two doses of either the AstraZeneca or Pfizer vaccine at 3 weeks and 20 weeks after the second dose. Antibody levels dropped at the same rate for both vaccines, but the starting level was much higher for the Pfizer vaccine. People with lower antibody levels were more likely to get infected. Those who received the AstraZeneca vaccine reached lower antibody levels faster (96 vs 257 days) and had a higher risk of breakthrough infections.

How does the type of COVID-19 vaccine initially received (AstraZeneca or Pfizer) affect the risk of infection after a booster vaccination?

We found that a booster dose with an mRNA vaccine (like Pfizer) leads to similar protection against COVID-19, regardless of the vaccine used for the first and second dose. 

How did COVID-19 booster vaccine (3rd dose) affect antibody levels compared to the first two doses? 

We found that after a booster vaccine, antibody levels peaked higher than after the second dose, regardless of whether people initially received the Pfizer or AstraZeneca vaccine. The antibody levels dropped more quickly compared to after the second dose. However, 10 weeks after vaccination, levels remained above the threshold needed to protect against the Delta variant. The magnitude and trajectory of post-booster antibody levels were similar across age groups and by clinical vulnerability status. 

How well do participants’ antibodies prevent different variants of the virus (virus inhibition) from entering peoples’ cells, and how does this affect the risk of infection?

Using finger-prick blood samples, we found that better inhibition of the Omicron BA. 1 variant was linked to a lower risk of infection during both the Delta and Omicron B.1 waves. However, the ability to inhibit the Delta variant only reduced infection risk during the Delta waves, not during the Omicron BA.1 period.

How long do COVID-19 antibodies (anti-N and anti-S) last after an infection?

Anti-S antibodies are linked to immunity from infection or vaccination, while anti-N antibodies indicate past infections. We found that about 80% of participants had anti-N antibodies within 269 days of a positive PCR swab test for COVID-19, with levels peaking at 30-120 days before declining. Men and older adults had higher antibody levels than women and younger adults over time. In contrast, the decrease of anti-S levels after infection differed by age and sex. 

How does disability status influence the likelihood of having immunity to COVID-19?

Around 10% of eligible people in our survey met the definition we used for having one or more disability (using the Washington Group-Short Set definition and restricted to people aged 7 and older). This group was more likely to experience financial difficulties, live in more deprived areas, and have clinical (health-related) vulnerability to COVID-19. We also found that anti-N antibody levels (indicating immunity gained through infection) were slightly lower in disabled participants than in non-disabled participants (15% vs. 19%). People who were both disabled and clinically extremely vulnerable were also less likely to develop immunity following a COVID-19 infection than people who were disabled but not clinically extremely vulnerable.

  1. Risk Factors and Exposure

Where do people think they caught COVID-19?

We asked people who had COVID-19 up to March 2021 where they believed that they caught the infection. Children were most often reported to catch COVID-19 at school. Working-age adults mostly reported catching it at home or work. People over 65 most often reported catching it at home or in essential shops.

How does social deprivation affect public activities and non-household contacts that can influence your risk of catching COVID-19?

Participants who lived in the most deprived areas were more exposed to essential activities, such as attending work in-person, using public transport, and shopping for essential items in-person, than those in less deprived areas during the second wave of the pandemic. These differences were not found for leisure and entertainment activities. Pandemic restrictions need to consider this – for example, by supporting low-income workers – to reduce inequalities between deprived and less-deprived areas.

What was the relative importance of going to work, shopping, using public or shared transport, and other activities to non-household transmission of COVID-19 in the second wave of the pandemic?

Going to work, using public or shared transport, and shopping all increased the risk of catching COVID-19. Shopping contributed the most to transmission because almost everyone did it regularly, while fewer people went to work or used public transport during this time due to restrictions. Other activities also played a smaller role because of restrictions. 

What is the relative importance of different activities and settings in the transmission of COVID-19 during the second lockdown in England and Wales?

We found that during the second wave of the pandemic, using public transport, shopping, and going to work or school increased the risk of catching the virus outside the home. Much of the risk from work or school was linked to using public transport, which remained a major risk factor even after accounting for other factors. Shopping was also a common source of infections because it was a frequent activity. 

How much does occupation affect the risk of COVID-19 infection, and how do workplace exposures differ by occupation and over time?

We found that the job you do makes an important difference to your risk of COVID-19 infection. Compared to office-based professional occupations, healthcare, teaching and childcare, indoor trade and process/plant (factory), and leisure and personal service occupations had higher risks of COVID-19 infection. For most of these higher risk occupations, the increased risk was in the first two waves of the pandemic, but teachers also stayed at higher risk during the third wave. 

How did workplace attendance and contact patterns change between occupations and over time during the COVID-19 pandemic in England?

Because the work you do affects your risk of COVID-19, we wanted to explore how working from home, workplace contacts, shared airspace and mask use varied by occupation at different times in the pandemic. Trade occupations were most likely to attend the workplace throughout the pandemic, and office-based professional occupations were the least likely. Workers in leisure and service occupations became more likely to go to work from April 2021, when many workplaces reopened. Office-based occupations continued to work mainly from home even after the easing of restrictions. Amongst workers who attended in person, there was high exposure risk – for example, through crowding – across many occupations. This highlights the relevance of remote work for reducing transmission where possible during pandemics and putting in mitigations where remote work is not possible.

How does COVID-19 vaccine uptake differ by occupation?

We found high (>70%) COVID-19 vaccination uptake across all three recommended doses with variability between occupations. Uptake tended to be highest in professional, administrative, and managerial occupations and lowest in trade, transport, leisure and service, and sales occupations. Occupational groups that had a lower vaccination uptake were likely to have more workers with vulnerability-relevant factors (e.g. obesity and older age). Workers with a low risk of exposure to the virus had higher vaccine uptake than those with elevated or high risk.

How does the implementation and uptake of key work-related safety measures (e.g. frequent hand and surface hygiene, wearing face coverings etc.) vary by occupation and what are the workers’ perceptions of these measures?

The usage of work-related safety measures differed between occupations and over time. Healthcare workers were most likely to report safety measures like hand hygiene, mask-wearing, and employer-provided protective equipment. Mitigation measures generally declined when national restrictions eased. Most workers (55–88%) felt the measures were reasonable even after official restrictions were lifted, though fewer supported physical distancing (39–44%). 

How does vaccination affect people’s transmission-relevant behaviours during a period of national lockdown?

We found that within 14 days of their first COVID-19 vaccine dose, participants were more likely to have close contact with non-household members and to use non-essential shops and services compared to pre-vaccination. This effect varied between men and women and across different age groups. 

Do people tend to travel further outside the home after being vaccinated against COVID-19 than before vaccination?

After vaccination, it takes some time before the body develops full protection against the SARS-CoV-2 virus. We investigated whether people increased their travel soon after getting vaccinated but before they were fully protected. Some Virus Watch participants volunteered to use a mobile phone app to track the distance travelled from home. We analysed this pre and post-vaccination but found no significant increase in travel immediately after vaccination.

How did COVID-19 restriction policies affect people’s movement behaviours? How did this differ between different socio-demographic groups?

During stricter lockdowns, participants stayed home more and traveled less, except to schools, healthcare facilities, and essential shops, which were allowed to stay open. Younger and low-income groups were more affected by restrictions as they were less able to adapt to travel restrictions likely due to dependency on public transport. 

Are overcrowded living conditions a risk factor for COVID-19 infection?

People in our study living in overcrowded housing (according to official UK definitions related to the number of rooms compared to people) were more likely to have a positive PCR or antibody test for COVID-19 than people living in under-occupied housing.

How did exposure to shared airspace and close contacts in settings outside the household change from periods of intense restrictions compared to periods with no restrictions?

When we compared weekday non household activities in March 2021 (a period of intense restrictions) and November 2021 (a period of no restrictions) we saw marked increases in frequency of use of a wide variety of settings including public transport, eating out, going to pubs, theatres, cinemas, sports activities and parties. These settings also had a high intensity of exposure to shared airspaces and, to a lesser extent, close contact. Parties were notable for having both high exposure to shared airspace and high levels of close contact.

How does living with children affect the risk of infection?

During the second wave of the pandemic, we found that living with a child did not change the risk of getting infected. However, during the third and fourth waves, individuals living with a preschool-aged child had a lower risk of infection than those without children. In contrast, living with a child of primary-school age increased the risk of infection during both the third and fourth waves. Living with a child of secondary school age increased the risk of infection during the third wave but reduced it in the fourth wave. This likely reflects the impacts of changing education and childcare-related restrictions.

How many people intended to get a COVID-19 vaccine when it became available?

We conducted an online survey of study participants in December 2020 and February 2021, asking whether people would accept a COVID-19 vaccine if offered. Only 10% of participants who answered the survey said ‘No’ or were undecided in December 2020, and more than 4 in 5 of these (86%) changed their mind or had been vaccinated by February 2021.

Does wearing glasses reduce one’s risk of infection?

We found a 15% lower risk of infection for those who reported always using glasses for general use than those who never wore glasses. 

Research paper: Eyeglasses and risk of COVID-19 transmission—analysis of the Virus Watch Community Cohort study

  1. Inequalities

How did migration status (UK or non-UK born) affect the rate of COVID-19-related hospitalisation?

Across the first three waves of the pandemic, we found that non-UK born participants had a higher rate of COVID-19-related hospitalisation compared to UK-born individuals. 

How does the increased risk of living in overcrowded housing increase migrants’ risk of infection?

We found that migrants had a 22% higher chance of being infected with COVID-19 compared to people born in the UK. About 36% of this increased risk was linked to household overcrowding. 

What are the characteristics related to barriers in accessing healthcare and what were the reasons given for these barriers?

We found that over a third of people trying to access healthcare in 2021 faced difficulties, mainly due to service disruptions from COVID-19. Barriers to healthcare access were more common among ethnic minorities and those in deprived areas. People from ethnic minority groups were more likely to report difficulties with appointment times and fear of not being listened to, while migrants reported not having enough time to explain complex needs. Participants living in the most deprived areas were also more likely to cite fear of not being listened to and being denied access than those in the least deprived areas.

What factors are associated with access to sick pay among UK workers?

We found that Black, Asian, and minority ethnic workers have lower access to sick pay, while undocumented workers had limited access to it at all. Older workers, workers in low-income households and those in working class occupations were also more likely to lack access to paid sick leave. This lack of financial protection is concerning both in terms of affecting people’s ability to recover from COVID-19 and other illnesses and in terms of the transmission of the virus.

  1. Long COVID

What is the risk of long-term symptoms by COVID-19 variants, other respiratory infections, and individuals without infection?

We found that people infected with earlier variants (Wild Type, Alpha, Delta, and Omicron BA.1) were more likely to develop long COVID (27-34%) compared to those who were infected with later sub-variants of Omicron (11-14%). These findings accounted for other important factors like age, having health conditions, and COVID-19 vaccination to reflect the effect of different variants. People who had COVID-19 or other non-COVID respiratory infections (8-23%) were more likely to develop long-term symptoms than people who did not have an infection during the same periods (1-3%), indicating that respiratory infections are an important source of long-term symptoms. 

How do antibody levels differ between people with and without long COVID?

People with Long COVID had persistently higher infection-related (anti-nucleocapsid) antibody levels after infection compared to people who recovered fully. Everyone in the study had mild-to-moderate infections, showing that people with Long COVID had a stronger immune response to their infection despite similar levels of severity. We found that antibody responses to vaccination (anti-spike antibodies) did not differ between people with and without Long COVID.

What is the likelihood of long COVID by deprivation levels, migration and ethnic minority status?

When wild-type, Alpha, and Delta variants were spreading, people living in more deprived areas were more likely to develop long COVID, particularly females. At the same time, males from migrant, ethnic minority, or both migrant and ethnic minority backgrounds were also more likely to develop long COVID compared to their counterparts. During the Omicron waves, these differences in long COVID risk became less pronounced. However, ethnic minority women who were migrants were more likely to develop long COVID compared to White British UK-born females.

How many children with COVID-19 have symptoms lasting 4 weeks or longer (‘long COVID’)?

In our study, 4.6% of children with COVID-19 reported having symptoms for 4 weeks or longer, compared to 1.7% of children overall.