Immune experience at time of circulation and time since virus circulation are predictors of HAI titre

Nearly everyone is infected with seasonal influenza viruses multiple times over the course of their lives due to the antigenic evolution of the virus to escape immunity induced by prior infections and vaccinations. Because antibody escape is usually partial, the antibody response to new infections depends on prior responses which wane and are boosted throughout life. We used serum antibody binding measurements of 864 individuals against a range of historical A/H3N2 viruses collected in previous studies in combination with statistical models to investigate how metrics of age and strain heterogeneities affect haemagglutination inhibition titres. We refine prior modelling of antigenic seniority by characterising immune experience as years of life since a (sub)types emergence at time of strain circulation, rather than age at time of circulation. Using Bayesian statistical modeling, we show that this variable, combined with time since circulation of the variant and an individuals age at circulation, yields the best model fit. Based on the best-fit model, we devised a novel parametric model, and use it to demonstrate and estimate how effects of antigenic seniority, individual age, strain age and strain effects act in concert to shape an individuals HAI titre.