COVID-19 research

Since the beginning of the COVID-19 outbreak members of the MRC Integrative Epidemiology Unit have been engaging with research and discussion on the epidemiology, public health responses and data collection and analytical challenges. These pages give a summary of some of the research, support and civic activities that members of the Integrative Epidemiology Unit have been involved with. 

COVID-19 increases anxiety and depression, but young and older people are affected differently

Overview

The COVID-19 pandemic and related mitigation measures are likely to have a marked effect on mental health. Longitudinal studies with pre-pandemic information, alongside data collected during the COVID-19 pandemic are needed to fully understand the impact of this pandemic on mental health and how mental health may have changed. We used data from two population studies: 1) the Avon Longitudinal Study of Parent's and Children (ALSPAC, both generations) and 2) Generation Scotland. We found that in the young ALSPAC cohort, the number of people with anxiety was almost double compared to pre-pandemic times. However, depression was stable compared to before the pandemic. This pattern was not observed for the older generation. In addition, we also found evidence that females, individuals with pre-existing mental and physical health concerns, individuals reporting a COVID-19 infection and those in socio-economic adversity were more likely to report higher depression and anxiety during the COVID-19 pandemic, even when taking into account their mental health prior to the pandemic. Our results highlight groups of individuals who may be at greater risk of poorer mental health who could be prioritised for treatment and support, or targets for continued monitoring of symptoms. Future research should also track mental health to examine if these are short or longer term effects. The results in this study have been shared with Health Data Research UK (HDRUK), Public Health England (PHE) and the Government's Scientific Advisory Group for Emergencies (SAGE). 

Project team

University of Bristol: Alex Kwong, Rebecca Pearson, Kate Northstone, Kate Tilling, Stan Zammit, David Gunnell, Paul Moran, Matthew Hickman, Dheeraj Rai, Simon Haworth, Deborah Lawlor, Nic Timpson

Further information

Read the full article in medRxiv 

Mapping community support during COVID-19

Overview

Since the Covid-19 pandemic brought about nationwide lockdown, communities have been mobilising to help vulnerable people, such as by shopping for neighbours or offering a friendly and supporting chat. But we know that such community support is patchy, leaving it difficult for government agencies and third sector organisations to target additional support in areas where it is most needed. Working in collaboration with Public Health Wales, and using data from a range of sources, the Bristol research team have developed a live map to help users develop a better understand which communities have better community cohesion and organisation. The map uses visualisation tools to quickly highlight areas that might be more vulnerable due to an imbalance between community support and community need.

Project team

University of Bristol: Dr Oliver Davis, Dr Valerio Maggio, Dr Alastair Tanner, Nina Di Cara, Chris Moreno-Stokoe, Benjamin Woolf

Public Health Wales: Dr Alisha Davies, Dr Jiao Song, Elysha Rhys-Sambrook, Lucia Homolova

Further information

Read the blog post about this project on IEUREKA! or access the Covid Response Map.

Collider bias: why it is difficult to find risk factors or effective medicines for COVID-19

Overview

Our understanding of the risk of infection and progression of Covid-19 currently comes primarily from observational studies, and this understanding is what public health policy and clinical decision making is based on. Observational studies in epidemiology are known to be affected by a type of bias called collider bias. This means that it is difficult to tease apart correlation from causation. Studies of Covid-19 are challenging because non-random sampling is commonly used such as by hospital admissions, targeted testing, or voluntary participation. The impact of collider bias on inferences about Covid-19 could be considerable for example when modelling disease transmission, or examining risk factors for infection or disease severity. The positive news is that there are strategies for studying Covid-19 in ways that are not susceptible to collider bias or evaluate how sensitive any associations found could be to collider bias.

Project team

University of Bristol: Gareth Griffith, Tim Morris, Matt Tudball, Annie Herbert, Giulia Mancano, Lindsey Pike, Gemma Sharp, Tom Palmer, George Davey Smith, Kate Tilling, Luisa Zuccolo, Neil Davies, Gibran Hemani 

Further information

For more details on this project, read the blog on IEUREKA! or the full article on medRxiv. For further reading there is an opinion piece at Heath Data Research UK and a Report in Science

Shielding from COVID-19 should be stratified by risk

Overview

In this editorial, the authors put forward the case for stratified shielding as part of a population health strategy for responding to the Covid-19 epidemic. Stratified shielding protects groups identified as most at risk of dying from Covid-19, enabling lockdown to be lifted and those in lower risk groups to resume every-day functions. To be effective as an intervention we require sophisticated models of risk and clear communication of realistic levels of risk for different groups. Several other groups have modelled “stratify and shield” strategies, but these differ in predicting when shielding could end. At the moment we can say that Covid-19 mortality most closely parallels risk of death from all cause for age and sex, but better empirical data from large surveys of seroprevalence and the prevalence of current infections could lead to more sophisticated models. This will allow factors such as ethnicity, comorbidity categories, obesity, prescription medicines and measures for socioeconomic position to be accounted for.  The authors conclude that stratified shielding, together with other measures should be recognised as a population health strategy to move away from lockdown, which is seriously damaging many aspects of people’s lives.

Project team

University of Bristol: Professor George Davey Smith

University of Cambridge; Professor David Spiegelhalter

Further information

Read the full article on the BMJ website.

Discussing COVID-19 research with children in schools

Overview

IEU researchers have led online epidemiology sessions for students at two Bristol primary schools.

Having outlined their own research, IEU researchers led the 11-year olds in devising their own epidemiology research questions about COVID-19 or the effects of the lockdown. Their questions included: Can eating chillies help prevent COVID-19? Can alcohol make COVID-19 worse? and Do people sleep more, less or the same in lockdown?  

The students then worked out how researchers might collect and process data to help them to work out the answers to their questions. Through so doing, they considered the challenges of working with incomplete data in rapid-response epidemiology. They also engaged with ideas of causality and the responsible communication of research findings.

The 11-year olds greatly enjoyed the sessions, saying, “I enjoyed this experience because I learnt a lot about the process of how scientists study” and “It was good to meet scientists – I was inspired.”

The students are in year 6 at Nova Primary School, in Shirehampton, and at Horfield Primary School; many thanks to their teachers for making the sessions possible.

Project team

University of Bristol: Louise Millard, Annie Herbert, Hannah Sallis, Tim Morris, David Carslake, Philippa Gardom

Further information

Contact Philippa.Gardom@bristol.ac.uk

Edit this page