Network analysis reveals causal relationships among individual background risk factors leading to influenza susceptibility

Influenza is a serious infectious disease that spreads rapidly and causes numerous deaths worldwide. Although the biological and environmental risk factors for influenza susceptibility have been extensively studied, individual risk factors, such as medical history and lifestyle, and their complicated relationships remain unclear. In this study, we aimed to reveal individual-specific causal relationships among influenza risk factors using large-scale health checkup data collected from residents of a rural community. To this end, we quantitatively represent the relationships between each factor for individuals using a Bayesian network. Our network revealed multiple causal pathways from each health checkup item to influenza onset/non-onset. Furthermore, we performed a cluster analysis based on the individual network profiles to reveal the characteristics of the participants susceptible to influenza. We identified five distinct types of participants: hyperglycemia, pneumonia, hectic and sleep-deprived, malnutrition, and allergies. Our results highlight the importance of personalized influenza prevention strategies based on individual backgrounds. To the best of our knowledge, this is the first study to identify individual-specific risk factors and comprehensive causal relationships for disease onset. Thus, our network-based analysis may contribute to the development of personalized disease prevention and treatment approaches.