Amplification of the power of network hubs and degree skewness over infectious disease spread during lulls

Abstract

Just as individuals’ personal social network connections shape their susceptibility to disease, the structure of larger networks in communities shapes the extent of disease spread. We examine how heterogeneity in network structure at the population level affects the spread of disease – namely, COVID-19 – considering varying levels of disease transmissibility and in-person contact rates. Using dynamic simulations that take into account network structure, social contact rates given contextual features of the community (informed by real-life data on family local family structure, schools, workplaces, and daily shopping activities), and disease infection rates, we first confirm that the presence of network hubs and high network degree skewness results in a higher level of infected peak prevalence with infectious diseases such as COVID-19 during periods of low to moderate transmissibility. However, this effect is amplified during lulls in disease spread and is suppressed during periods of greater transmissibility, rendering social network structure more significant during lulls. Moreover, in the case of already highly transmissible diseases, the role of hubs and severe degree skewness is already more continually suppressed.