Nonhomogeneous mixing reduces disease prevalence

Human movement and spatial heterogeneity shape the spatial distribution of infections. Factors such as physical condition, availability of medical resources, socioeconomic status, and exit-entry screening can lead to variations in movement rate and pattern (or called habitat connectivity in discrete diffusion and dispersal kernel in continuous diffusion) among people with different health states. While the effects of movement rate on disease spread have been extensively studied, the role of movement pattern remains less understood. In this paper, for a susceptible–infected–susceptible (SIS) patch model incorporating either Eulerian, Lagrangian, or hybrid Lagrangian–Eulerian movement, as well as an SIS nonlocal dispersal model, we derive an upper bound on the global disease prevalence that is independent of movement. In a homogeneous environment, the nonhomogeneous mixing of susceptible and infected individuals always reduces disease prevalence. The prevalence attains its maximum when the susceptible and infected populations adopt the same distribution strategy. Numerical simulations further illustrate some new phenomena arising from different movement patterns. These results deepen our understanding on the impact of human movement on disease spread and pathogen evolution, thereby improving control measures to reduce disease burden.