In our latest expert webinar, Professor Tim Spector, lead scientist at the ZOE COVID Study, was joined by Dr Sarah Berry, reader in nutritional sciences and Dr Claire Steves, reader in ageing research, both from King’s College London.
They discussed the origins of the ZOE app, how digital tools like the ZOE app are changing the way we do health research, and what we can expect for the future of ZOE and digital health.
The origins of ZOE
Before COVID, ZOE researchers were working on PREDICT- the world's largest nutritional study. “PREDICT was designed to unravel how we respond to food, how much variability there is in responses to food and what determines this variability,” says Sarah.
“We found that there was about 20 fold difference in how individuals responded to exactly the same food, and interestingly, we found that our genes didn’t account for all of this variation,” says Sarah.
Our comprehensive data from thousands of people showed that factors like exercise, sleep, and the gut microbiome played critical roles in how people responded to food. Using what we learned in the PREDICT study, we have created a precision nutrition program that is tailored to each individual and their unique responses to food. To find out more about PREDICT and our nutrition program, you can visit our ZOE nutrition website.
“I think the PREDICT studies are a real paradigm shift in how we conduct research,” says Sarah. “Typical nutrition studies look at one dietary intervention and measure the average response, but the PREDICT data shows that we need to start looking beyond the average and moving away from traditional approaches.”
Sarah stresses collaborations between traditional academia and agile tech, combined with citizen participation, have been crucial for the PREDICT studies. “By taking part, people have been finding out loads about themselves, sharing back their data with us, and we've been able to capture data in the breadth, depth, scale, and precision that we need,” she explains.
The lessons from the PREDICT nutrition studies helped shape the COVID study, including the framework for using citizen science, big data, and machine learning to uncover insights. The science we learned in PREDICT underpins everything in the ZOE COVID app.
“When we were first setting up the COVID symptom study, although we had an idea for a massive epidemiological study that tracked COVID symptoms over time, we needed the link with ZOE to provide the technical expertise and turn the idea into an app that was on peoples’ phones within five days,” says Claire.
The ZOE COVID Study app quickly went viral, getting more than a million contributors in just a few days. That number has continued to grow, and we’re now a community of over 4.7 million people.
“We soon started to see the similarities between our nutrition studies and COVID studies. In COVID, you've got a single virus that causes very different reactions in different people. And in the nutrition study, we showed that thousands of people given the same muffin could have very different reactions. It all comes back to individual variation and predicting responses,” says Tim.
A new kind of science
Throughout the COVID study, we saw that people voluntarily participated and felt good about helping us with our research. We realised that this could be a winning formula to study all kinds of disease.
We’re part of a brand new field called “digital epidemiology”. Digital epidemiology uses data generated outside the public health system to understand disease. It is fuelled by the increasing availability of data and computing power, as well as breakthroughs in data analytics methods. For ZOE, this means using data volunteered through the app from millions of people to help solve crucial healthcare questions about COVID and other health conditions.
For example, a key question we answered about COVID was whether anosmia was a symptom of infection. “Anosmia was put into the app because my colleagues in the healthcare sector and care home were anecdotally reporting a loss of sense of smell,” explains Claire. “Within about a week, we could see that anosmia was very strongly associated with a positive test. We got that result published in an academic journal quickly, so it sparked a change in the guidance around testing.”
We were also able to track the spread of the disease across the country and identify hotspots like Leicester before the government had testing infrastructures in place to track local outbreaks, highlighting the increased speed of digital epidemiology. Furthermore, we used the app to study many more essential questions during the pandemic, including what factors increase your risk of developing long COVID, how lockdown affected our mental health, and how well the vaccines were working, which would have been challenging to study with traditional epidemiology.
What’s the future of the ZOE app?
Over the last 20 months, the world’s concentration on COVID has left other illnesses neglected. This means that people are presenting at a much later stage of their disease.
“There are problems in primary care because people are having difficulties in accessing services. But there are also difficulties in individuals recognising when something's going wrong and when they need to seek help,” says Claire. “Having tools like the app that monitor how we’re feeling might help in the management of existing diseases and prevention of new problems.”
We’re taking the amazing power of the ZOE community and app reporting to wider healthcare, including cancer, heart disease, bone problems and more
The interactive studies hub is now live on the app. It shows some of the big scientific questions around health and disease that we’re tackling going forward, and you can vote for the studies you think we should prioritise as a community.
If you want to join us in fighting COVID and other important health conditions, open the ZOE COVID Study app and follow the questions to opt in, if you haven’t done so already. If you’re not already a contributor and would like to get involved, simply download the app, register an account and start logging.