Abstract

The age pattern of immunity to seasonal influenza is striking but little understood. In contrast to many infections, the proportion of the population seropositive peaks in school age children, reaches a minimum between ages 35-65, and then rises again in older age groups. Heterogeneous mixing between age classes can have a profound effect on disease dynamics, and is hence the obvious candidate explanation for this pattern. However, using a mathematical model of multiple influenza strains, age dependent transmission based on mixing data from social contact surveys cannot on its own explain the observed data. Instead, seroprevalence may be a consequence of ‘original antigenic sin’; if the first infection of a lifetime dominates subsequent immune responses, it is possible to reproduce the relationship between age and immunity.