Nice overview of the challenges of doing "epidemiological-behavioural" modelling.
In my PhD I am tackling the problem by measuring and modelling behaviour on an individual level, with massive amounts of location data.
Stay tuned...