Contact network structure shaped pandemic transmission despite lockdowns

Human contact network structure fundamentally shapes infectious disease transmission and control. Most COVID-19 epidemic models assumed approximately homogeneous contact patterns, yet real-world networks are highly heterogeneous. We analysed 59,585 daily non-household contact reports from Germanys COVIMOD study (2020-2021) using a novel heavy-tail regression framework. Throughout the pandemic, contact distributions remained strongly heavy-tailed despite substantial non-pharmaceutical interventions. The most connected 20% of individuals consistently generated over 80% of non-household contacts, placing the network in an infinite-variance, scale-free regime for all but the first lockdown. Regions with heavier tails experienced higher COVID-19 incidence. Importantly, the top 5% quantile had just 9.4 contacts per day, reflecting everyday activities rather than mass gatherings. High-contact individuals were predominantly working-age adults living alone and engaged in in-person work or study. Agent-based simulations show that tail-targeting interventions could control spread as effectively as lockdowns for a broad range of pathogens with far less disruption, with childhood-driven infectious disease remaining an exception. These findings demonstrate that the structure of human contact networks is fundamental to how rapidly infectious diseases spread, persist, and how transmission can be controlled.