The interplay of behavioral variability and response sensitivity generates collective effects in host-pathogen transmission dynamics

Social interactions have been shown to drive pathogen transmission in a host population even under the most stringent conditions. For example, areas under strict Covid-19 lockdowns showed long-tailed incidence curves. It is now recognized that behavioral stochasticity, manifest at the individual level, can dynamically generate a range of transmission patterns, such as waves and plateaus. Yet, strategies seeking to mitigate the effects of an epidemic require collective behavioral changes. Implementing these in the face of noise, reduced sensitivity and nonlinear feedback is key to a successful public health policy. Here we present a model of pathogen transmission that ties individual behaviors to population-level phenomena. We find that periodic and robust collective changes in behavior drive pathogen prevalence across the population in a wave-like pattern. By incorporating gradual adjustments in behavior through feedback mechanisms we find trade-offs between response sensitivity and behavioral noise, which can be used to mount an efficient strategy at early times in the outbreak. Our models can be applied to a range of human pathogens, and benchmarked against data (wastewater surveillance, hospitalizations, and mobility) to generate predictions for broad policy recommendations.