New hypergraph model shows stronger social ties suppress epidemic spread

A team of researchers from multiple institutions (Peng, Feng, Deng, Perc, Kurths) has developed a novel hypergraph framework to study the interplay between information propagation and epidemic dynamics in cyber-physical systems. The model addresses a key gap: how heterogeneity in interpersonal relationships affects both information acquisition and disease transmission. The framework consists of two layers: a cyber layer that captures information spread through both pairwise and higher-order interactions using a mixed hypergraph, and a physical layer that models epidemic spreading via a Susceptible-Infected-Susceptible (SIS) process. To account for real-world relationship diversity, the authors introduce an adaptive perception-protection mechanism based on Jaccard similarity—a metric that quantifies how similar an individual's social connections are to their neighbors'.

The study's theoretical analysis, grounded in the Microscopic Markov Chain Approach (MMCA), derives the epidemic outbreak threshold and is corroborated by extensive Monte Carlo simulations. Key findings indicate that stronger interpersonal relationships not only accelerate the spread of epidemic-related information but also significantly raise the outbreak threshold, thereby reducing the final epidemic size. This work provides a mathematical foundation for designing more effective public health strategies that leverage social network structures—such as targeting information campaigns to tightly-knit communities to increase protective behavior. The hypergraph approach also captures complex, group-level interactions (e.g., family or workplace clusters) that traditional pairwise network models often miss, offering a more realistic tool for policymakers.