London School of Hygiene & Tropical Medicine

All age groups in England dramatically reduced their social contacts from March 2020 to March 2021 compared to pre-pandemic levels, with data that suggests that government interventions were the driving factor, according to new research.

The study, not yet peer reviewed, was conducted by a team from the London School of Hygiene & Tropical Medicine’s Centre for Mathematical Modelling of Infectious Diseases.

As part of the The CoMix Contact Survey, nearly 20,000 people, who were representative of the English population by age and gender, were asked questions about their social contacts every week over the first year of the pandemic. This timeline encapsulates three periods of national lockdowns interspersed with periods in which fewer restrictions were in place.

The study found social contacts reduced by about 75% for adults 18 years and older during the lockdown periods compared to the POLYMOD and BBC Pandemic surveys, which collected data prior to the pandemic.

During periods of reduced restrictions (in the summer of 2020), the mean number of reported contacts were still only about half of those reported in the earlier studies.

Amy Gimma, lead author from the London School of Hygiene & Tropical Medicine, said: “Our results suggest that government action was a factor in the mean number of social contacts in England, with the mean number of contacts dropping markedly during each lockdown. These large reductions in contacts have helped reduce the basic reproduction number and slow transmission, especially during periods of lockdown.”

The data is part of the CoMix Contact Survey that runs across the UK and internationally to collect data on social contacts, risk perceptions and behaviour. The survey is used by researchers to monitor one component of the public health response to the COVID-19 pandemic, which is to reduce social contact between people in order to prevent person-to-person transmission of the virus.

In this study, 19,914 people were asked who they made contact with the previous day in terms of age, setting (home, work, educational setting, etc), and whether or not they were indoors or outdoors. Data on children’s contacts were collected through a parent, who answered the questions on the child’s behalf. All four UK nations implemented different government interventions, so this study focuses on just participants who reported living in England.

Along with government action, age was clearly associated with the number of contacts, with children 5 to 17 reporting the greatest average number of contacts, 15.11 (95% CI 13.87 - 16.41), during periods of school opening. Among adults, younger people (18 to 59 years) reported a higher mean number of contacts throughout the year, between 2.39 (95% CI 2.20-2.60) during periods of lockdown to 4.93 (95% CI 4.65-5.19) during the summer of 2020. The elderly reported fewer contacts with the mean number of contacts between 1.55 (95% CI 1.42-1.69) and 3.09 (95% CI 2.82-3.39).

Although easing of restrictions did lead to an increase in contacts, this did not necessarily occur immediately. In particular, the easing of the first lockdown was not associated with a rapid rise in the mean number of contacts until August when the government introduced an incentive scheme to encourage individuals to dine in restaurants, cafes and bars.

Participants were also asked about precautions taken during the contact, such as wearing a face covering or washing their hands before and after the contact. The use of face coverings increased dramatically in July, which is timed with the implementation of a mask mandate imposed on July 24^th.

In addition, there appeared to be some association with actual or perceived risk: those who did not self-identify as a high risk individual or did not perceive coronavirus as likely to be severe for them reported higher mean numbers of contacts. The mean number of contacts were also sensitive to employment status, with part time workers documenting more contacts than others. Income levels were not strongly associated with mean contacts, nor is there any evidence of seasonality in contacts.

Dr Christopher Jarvis, study senior author from LSHTM, said: “Severe restrictions, such as lockdowns, have a huge negative impact on people’s lives and livelihood. However, our data suggests that over the last year they have worked in reducing social contacts, and in turn curbed transmission of COVID-19.”

Since CoMix Contact Survey data is quickly collected and analysed, it can be a very early indicator that social contacts are increasing which may lead to an increase in COVID-19 cases. The first CoMix study provided one of the first estimates of the change in the R number during the first lockdown.

Professor John Edmunds, study senior author from LSHTM, said: “LSHTM is proud of its work which has supported the UK government to make what are very difficult policy decisions. Thankfully, the vaccine roll-out in the UK has been hugely successful. This offers hope that we may have seen the last of the full national lockdowns which, while necessary from an epidemiology point of view, have caused such disruption and damage to people’s lives.”

In the future researchers can use the CoMix data for new COVID-19 models and in modelling for other diseases. The study is ongoing, and data is still being collected. Many more studies will follow, examining more specific questions about social contacts over time, by participant characteristics, and vaccination status.

The authors acknowledge limitations of the study, including that the survey is implemented through the online survey company Ipsos-MORI, so only participants who are reached by their recruitment strategies are eligible for the study. The survey also relies on participants self-reporting contact data, which makes the data vulnerable to recall bias and social-desirability bias, which may lead to an over or under-reporting of contacts.

Pre-print Publication

Amy Gimma, James D Munday, Kerry LM Wong, Pietro Coletti, Kevin van Zandvoort, Kiesha Prem, Petra Klepac, G. James Rubin, Sebastian Funk, W John Edmunds, Christopher I Jarvis. CoMix: Changes in social contacts as measured by the contact survey during the COVID-19 pandemic in England between March 2020 and March 2021.