The Impact of Network Structure on Successive Waves of Spreading
Loading...
URL
Journal Title
Journal ISSN
Volume Title
Perustieteiden korkeakoulu |
Master's thesis
Authors
Date
2024-06-17
Department
Major/Subject
Complex Systems
Mcode
SCI3060
Degree programme
Master’s Programme in Life Science Technologies
Language
en
Pages
56+5
Series
Abstract
Socio-economic, environmental, and ecological changes have contributed to the emergence of infectious diseases. This emergence presents the challenge of controlling these diseases by implementing immunity. One key goal for policymakers in disease control is to distribute immunization to achieve robust herd immunity. Herd immunity occurs when immunizing a proportion of the population reduces the risk of infection for all susceptibles and prevents outbreaks. Several studies have investigated the strength of herd immunity induced by random immunization and natural infection using mean-field epidemic models. These studies suggest that in a heterogeneous population, herd immunity gained through natural infection is more effective compared to the level expected from random immunization in a homogeneous population. However, real-world contacts are non-homogeneous and exhibit spatial structures that mean-field models cannot capture. This thesis adopts a network-based approach to explore the effect of contact structure on the strength of herd immunity. Specifically, we compare the effectiveness of disease-induced herd immunity with random immunization on contact networks with different structures. We focus on two characteristics of network structure: degree heterogeneity and spatiality. Our results indicate that when immunity is induced by natural infection, the competition between two factors determines the strength of herd immunity. The preference for infecting individuals with high contact enhances herd immunity, while the localized nature of disease spread weakens it. The latter is an effect that mean-field models cannot account for. Furthermore, the study shows that the localization effect is more pronounced in highly spatial networks, where contacts are more local. In such networks, even the preferential infection dictated by high levels of degree heterogeneity cannot overcome localization, resulting in disease-induced herd immunity being less effective than random immunization. Our findings underscore the importance of network structure in the effectiveness of disease-induced herd immunity and call for a critical assessment of the herd immunity effect obtained from model-based predictions. Relying on a model with unrealistic assumptions risks overestimating the strength of herd immunity and can misguide policy decisions.Description
Supervisor
Saramäki, JariThesis advisor
Hiraoka, TakayukiKeywords
complex n, network structure, epidemic spreading, spreading processes