Resurgence of large-scale influenza A (H1N1) outbreaks: Modeling the interplay of transmission, loss of immunity, and vaccination
TL;DR
This is the first mechanistic modeling study specifically aimed at investigating large resurgent H1N1 epidemics across multiple locations, and highlights the need for context-specific models and surveillance systems that account for the complex interplay among environmental drivers, population immunity, and viral mutation driving H1N1 dynamics.
Abstract
Influenza is a highly transmissible respiratory virus that can cause devastating pandemics. The 2009 influenza pandemic, caused by a strain of A(H1N1) virus, resulted in millions of cases and hundreds of thousands of deaths worldwide. In the following years, many regions experienced waves driven by environmental conditions, waning immunity, and other factors. More dramatically, several locations saw unpredicted large resurgent outbreaks, yet the mechanisms behind these severe resurgences remain poorly understood. Despite an extensive literature of epidemic models aimed at explaining and predicting H1N1 dynamics, mathematical frameworks specifically designed to address why these large outbreaks recur years after an initial epidemic are still lacking. Here we address this gap by proposing a mechanistic model with time-varying rates of infection, loss of immunity, and vaccination. In particular, we explore two mathematical functions to model dynamic loss of immunity, capturing how viral evolution may reduce the duration of immune protection. Numerical simulations show that the model reproduces key features of H1N1 case data, and that delayed resurgence is associated with longer immunity periods and higher vaccine efficacy. We also fit the model to data from Brazil, Turkey, Iran, New Zealand, the United States, Colombia, South Africa, and Croatia, demonstrating its ability to capture resurgence patterns. Fitted infection rates exhibited a wide range of temporal profiles, from regular annual cycles to higher-frequency oscillations and gradual changes, suggesting that transmission is driven by factors beyond seasonal influences. The model also inferred dynamic loss-of-immunity rates, representing different effects of antigenic drift. To our knowledge, this is the first mechanistic modeling study specifically aimed at investigating large resurgent H1N1 epidemics across multiple locations. Our findings highlight the need for context-specific models and surveillance systems that account for the complex interplay among environmental drivers, population immunity, and viral mutation driving H1N1 dynamics.