Intermittent quarantine suppresses epidemic spreading beyond simple contact reduction

Abstract

Social interactions are shaped by networks that can be highly heterogeneous, with some individuals potentially interacting with many others. However, physical and temporal constraints limit the number of simultaneous interactions, making social contacts inherently intermittent. Here, we investigate how intermittency affects epidemic spreading on complex networks considering two distinct mechanisms: intermittency of connections (links) and individuals themselves (nodes), with periods of activity and inactivity governed by renewal processes with broadly distributed inter-event times. Combining mechanistic theoretical approaches with stochastic simulations, we show that these two forms of intermittency have fundamentally different effects on epidemic spreading. Link intermittency largely preserves the mechanisms responsible for sustaining transmission, and epidemic behavior can be largely explained by the average reduction in contact opportunities. Contrastingly, node intermittency is substantially more effective at containing spreading because the simultaneous interruption of all contacts suppresses transmission while recovery continues independently. The non-Markovian nature of node intermittency plays a key role: shorter and less heterogeneous inter-event times increase epidemic thresholds and enhance epidemic containment. Our results highlight how the temporal organization of individual activity, beyond the average reduction in contacts, can substantially alter epidemic spreading and may help inform interventions that exploit intermittency in social and biological contact networks.