Quantifying impacts of urban food environment policies: A hybrid model approach

Our food environment plays an essential role as a determinant of diet and associated health outcomes such as cardiovascular disease and diabetes. Prevalence of and access to fast food outlets, for example, have been found to be associated with increased consumption of takeaway food and greater body mass index (BMI). In response, policies and interventions acting upon the food environment are being considered and implemented by local governments to address health concerns. These policies include use of economic measures (e.g. tax of sugar-sweetened beverages), land use restrictions or incentives (e.g. restricting hours of operation for fast food outlets close to schools), and placing controls on food marketing (e.g. watershed for adverts); however, evidence informing these policy decisions is limited.
In order to support decision-making, research must account for the myriad behavioural, environmental, and social elements influencing individuals and their decisions related to nutritional intake. Few research efforts have employed methods to account for such complexity. A complex systems-approach accounting for the non-linear pathways and feedback loops present in the relationship between food environment, diet, and associated health outcomes is required.
This project accounts for this complexity by utilising a novel hybrid modelling approach to understand and quantify the health outcomes and economic impacts of implementing policies impacting local urban food environments. The approach first employs an agent-based model (ABM), taking into account individual actors, their behaviours, and the factors impacting an individual's health and food consumption when a policy is changed. Second, a simulation model is developed to quantify health and economic impacts of the outputs measured by the ABM. Leveraging existing relationships with local decision makers and experts (e.g. third sector organisations, local authorities, NHS), the model will be conceptualised, piloted, and tested with direct input on a range of feasible policy scenarios.
Methodologically innovative, this project contributes to the nascent but growing body of quantitative complex systems methods for public health problems drawing upon multiple fields of study including epidemiology, behavioural science, computer science, and geographical data science. Model and scenario co-production of policies will further ensure relevance and more direct pathways to impact of this research.