To see my publications, please take a look at my google scholar page.
Novel lifestyle data
90,000 deaths are attributed to poor diet each year in the UK alone. This is 1 in 7 deaths due to poor diet! With an unhealthy diet posing such a challenge, understanding consumption patterns and promoting healthier patterns is critical to a healthier nation. Poor diet is not limited to the UK, but is a global challenge. In many places, the UK included, malnutrition from underweight and obesity co-exist presenting the extremes of diet related problems. In addition, we see those of a healthy weight lacking in essential nutrition, again creating health problems. Despite the scale of nutrition related problems, diet remains notoriously difficult to measure.
In my research I investigate alternative data sources which may enable us to better understand diet. These include transaction data from supermarket loyalty cards. The relationship between diet and the wider food system is important. This includes: the food environment, food availability and the cost of food, to name a few. Inequities in these exacerbate diet related problems within groups of our population.
At the other side of the energy balance equation is physical activity, or the lack of it – sedentary behaviour. Physical activity it critical to health and sedentary behaviour is a leading modifiable risk factor for non-communicable diseases. There is a wealth of novel data sources relating to physical activity, including smart phone apps, wearable devices and gps tracking.
There is so much we can learn about lifestyle behaviours through these types of data and a wealth of data science methods that can be applied. In the first instance though it is critically important that we understand new sources of data and evaluate these in relation to more traditional measure that are well understood.
These data are not typically generated for research purposes and the individuals who generate these data are unlikely to have consented for specific research use, more likely a broad statement about data being used anonymously for research. Therefore I feel that understanding public attitudes to use of such data is critical.
To truly unlock the value of novel data sources those generating the data must be on board with their use. With this in mind we set up the LifeInfo Study to investigate public attitude to lifestyle data linkage to health records for research purposes.
It is possible to ascertain public opinion from existing data sources too. For example, we have used twitter date to understand public opinion on the COVID test and trace system, during the early stages of its development.
Social and spatial variation
The importance of geography in lifestyle analytics is always in my mind. Geography with respect to both context and composition of the environment. The context is the physical location and environmental features, perhaps living close to a park for exercise, or fast food outlets for access to unhealthy foods. The composition of the environment is more all encompassing, taking into account demographic characteristics, community cohesion and social norms which may be driven by local or national policies. These features are driving forces with respect to inequities in lifestyles and related health.