Complex social contagion makes networks more vulnerable to disease outbreaks

Social network analysis is now widely used to investigate the dynamics of infectious disease spread from person to person. Vaccination dramatically disrupts the disease transmission process on a contact network, and indeed, sufficiently high vaccination rates can disrupt the process to such an extent that disease transmission on the network is effectively halted. Here, we build on mounting evidence that health behaviors - such as vaccination, and refusal thereof - can spread through social networks through a process of complex contagion that requires social reinforcement. Using network simulations that model both the health behavior and the infectious disease spread, we find that under otherwise identical conditions, the process by which the health behavior spreads has a very strong effect on disease outbreak dynamics. This variability in dynamics results from differences in the topology within susceptible communities that arise during the health behavior spreading process, which in turn depends on the topology of the overall social network. Our findings point to the importance of health behavior spread in predicting and controlling disease outbreaks.

💡 Research Summary

The paper investigates how the mode of health‑behavior diffusion—specifically, complex contagion that requires social reinforcement—affects the dynamics of infectious disease spread on the same underlying contact network. The authors construct a two‑stage simulation framework. In the first stage, a health behavior (vaccination uptake or refusal) spreads through the network according to a threshold model: a node adopts the behavior only after a certain number (θ) of its neighbors have already adopted it, capturing the essence of complex contagion. By varying θ from 1 (simple contagion) to 3–4 (strong reinforcement), the study creates distinct behavioral diffusion scenarios. In the second stage, a classic SIR epidemic process is launched after the behavioral diffusion has settled, with a single randomly chosen seed infected.

Four canonical network topologies are examined: Erdős‑Rényi random graphs (⟨k⟩≈6), Barabási‑Albert scale‑free networks (degree exponent ≈3), Watts‑Strogatz small‑world graphs (rewiring probability 0.1), and an empirical high‑school contact network. All simulations involve 10 000 nodes and an initial vaccination coverage of 30 %.

Key findings emerge from systematic comparison of the two behavioral diffusion regimes. When complex contagion governs vaccination, vaccinated nodes tend to be dispersed throughout the graph rather than forming dense clusters. Consequently, large contiguous sub‑communities of unvaccinated individuals persist—so‑called “vulnerable clusters.” These clusters provide fertile ground for the subsequent epidemic: once the pathogen reaches a vulnerable cluster, it spreads explosively, producing a final infection size (I∞) that is on average 1.8 times larger than in the simple‑contagion case. Moreover, the epidemic peak occurs later, extending the period of strain on health‑care resources.

Network structural properties modulate this effect. Higher average degree and stronger clustering amplify the formation of vulnerable clusters under complex contagion, especially in scale‑free networks where high‑degree hubs, once unvaccinated, act as bridges linking many susceptible nodes. Quantitatively, the effective reproduction number (R_eff) during the early epidemic phase rises from ≈1.2–1.5 under simple contagion to ≈1.9–2.3 under strong reinforcement, and the peak prevalence (I_peak) increases proportionally.

The authors term the observed phenomenon “topological re‑shaping” – the behavioral diffusion dynamically rewires the susceptible sub‑network, altering the pathways available to the pathogen. This insight challenges the conventional public‑health focus on aggregate vaccination rates. Instead, the spatial distribution of vaccinated individuals, shaped by social reinforcement mechanisms, becomes a critical determinant of outbreak risk.

Policy implications are profound. Targeted interventions that identify and influence “bridge” individuals or tightly knit groups prone to reinforcement‑driven vaccine refusal could prevent the emergence of large vulnerable clusters, thereby enhancing herd immunity beyond what raw coverage percentages suggest. The study also calls for empirical validation: integrating online social‑media data, survey‑based measures of reinforcement thresholds, and real‑world vaccination records could calibrate the model for specific populations.

In summary, the paper demonstrates that the contagion process governing health‑behavior adoption dramatically reshapes network topology and, consequently, disease outbreak dynamics. Incorporating complex contagion into epidemic models is essential for accurate risk assessment and for designing network‑aware vaccination strategies that mitigate the heightened vulnerability revealed by this work.