Mesoscale community organization governs epidemic onset and spread in metapopulations

Understanding how internal community structure shapes the course of epidemics remains a fundamental challenge in modeling real-world populations. Standard metapopulation models often assume uniform mixing within communities, overlooking how internal heterogeneity affects global outcomes. Here, we develop a general framework for epidemic spreading in hierarchically structured metapopulations, where individuals interact locally within dense communities and move across a broader network. Using a degree-based mean-field reduction and spectral perturbation analysis, we show that transmission dynamics are governed by the mesoscale organization of communities: densely connected groups lower the effective epidemic threshold and disproportionately drive epidemic onset, whereas weakly connected communities suppress transmission and dampen spread. Beyond epidemic onset, a localization-based spectral reduction reveals how the same mesoscale heterogeneity organizes the nonlinear endemic state, shaping the spatial distribution of infection across the metapopulation. We further validate these theoretical predictions through numerical simulations on synthetic metapopulations and analyses of empirical contact and transportation networks, where contact patterns govern local transmission and mobility networks mediate inter-community spreading, confirming the robustness of the framework across scales.