Structural causal influence captures the forces of social inequality in models of infectious disease

Abstract

Mathematical models are central to understanding disease transmission for host–pathogen systems across the biosphere. However, most existing frameworks do not explicitly treat host population heterogeneity as a formal driver of variation in transmission dynamics. This gap is most visible in human disease, where social inequalities are well known to structure infectious disease risk, but applies broadly: ecological, behavioural and demographic structure shapes transmission in non-human host populations as well. Here, we introduce a new metric, structural causal influence, which uses causal analysis to quantify how subpopulations contribute to overall transmission through structural differences in exposure, susceptibility or recovery. Using a multi-population model, we show that even a population whose isolated reproduction number lies below one can be drawn into a sustained epidemic through minimal contact with a more affected group.