Coupled information-epidemic dynamics with emotional thresholds and behavioral regression on multiplex networks

During prolonged epidemics, public adherence to protective measures often wanes due to behavioral fatigue-a dynamic psychological process typically overlooked in existing models. We propose a coupled Unaware-Positive-Negative-Unaware and Susceptible-Infected-Recovered-Susceptible (UPNU-SIRS) contagion model on a two-layer multiplex network, where both layers are constructed as 2-simplicial complexes to capture higher-order group interactions. The core innovation is an emotion-threshold mechanism: individuals accumulate negative emotions, and exceeding a heterogeneous personal threshold dynamically increases the probability of behavioral regression (e.g., from active to passive protection) and reduces immunity. Using the Microscopic Markov Chain Approach and extensive Monte Carlo simulations, we find that this mechanism fundamentally drives secondary outbreaks and sustains elevated endemic levels. Critically, while the emotion-threshold mechanism alone cannot induce bistability, its synergy with higher-order interactions in the disease layer dictates the extent of the bistable regime and the outbreak threshold-offering a leverage point where enhancing emotional resilience, reducing immunity-weakening effects, and ensuring timely, trusted media communication can collectively suppress both initial and secondary epidemic peaks. By integrating socio-psychological feedback with higher-order network structures, our framework provides new insights into epidemic persistence and targeted interventions against pandemic fatigue.