Universality classes in the time evolution of epidemic outbreaks on complex networks

Abstract: We investigate the temporal evolution of epidemic outbreaks in complex networks and identify two distinct universality classes governing their spreading dynamics. Using analytical arguments and large-scale simulations of the susceptible-infected (SI) model, we show that epidemic prevalence follows either a Gompertz-like curve in small-world networks or an Avrami-like curve in fractal complex networks. These findings challenge the traditional mean-field prediction of early-stage exponential growth of infections and highlight the crucial role of network topology in shaping epidemic dynamics. Our results demonstrate that the widely used mean-field approximation, even in its heterogeneous form, fails to accurately describe epidemic spreading, as the assumption of early exponential growth is not justified in either universality class. Additionally, we show that in fractal networks, the basic reproduction number loses its traditional predictive power, as the spreading dynamics are instead governed by microscopic scaling exponents of the network's self-similar structure. This shift underscores the need to revise classical epidemiological models when applied to systems with fractal topology. Our findings provide a refined theoretical framework for modeling spreading processes in complex networks, offering new perspectives for epidemic prediction and control.