MRC/CSO Social and Public Health Sciences Unit

There are many determinants of health and health inequality acting at multiple levels over the life course. Efforts to develop models to understand, simulate and predict causal relationships between determinants and health and health inequality outcomes have largely relied on methods that implicitly or explicitly assume that those relationships are linear and stable. Although increasingly sophisticated methods can account for multiple dependent and independent variables, observable and latent variables, effect modification and mediation, these more complicated models and theories remain dependent upon the assumptions of linearity and stability.

A complex systems perspective challenges these assumptions, and allows for the notion that relationships between the components of a system can be non-linear, bi-directional and thus generate instability and the emergence of new relationships and system states.

Complex systems simulation methods are well-developed in physics, engineering and ecosystems research. Techniques such as agent-based modelling are now highly accessible and allow the simulation of highly complex systems through the identification of the key agents in any system, and the rules that govern their behaviour. The application of these methods to public health and, in particular, to understanding health inequalities, holds great promise. They can be used to understand the emergence of public health problems such as obesity and harmful alcohol consumption, and to identify the parts of the system that are most likely to efficiently generate health improvement.

The Modelling Complex Systems workstream is a collaboration between the Complexity in Health and Places and Health programmes. Together we develop projects centred on the use of agent-based modelling methods to investigate 'wicked' problems of population health, where nonlinear, multi-layered interactions between elements of these complex social systems are difficult to unravel with traditional methods alone.  Alongside our development of modelling studies, we promote the wider use of ABMs in population health via the production of guidance and standards for ABM development, and the creation of new tools for analysing simulation outputs.