Exploring cross-protection between influenza strains by an epidemiological model

Almost all influenza surveillance systems of temperate countries report a single influenza illness peak each winter. Although several strains of different types or subtypes may co-circulate, they exceptionally lead to two distinct illness peaks during the same winter. We hypothesize that cross-protection between the different strains can, by itself, explain this absence of succeeding illness peaks. We used a dynamic population model to describe the spread of two influenza strains during one season. Cross-protection is included through a parameter defined as the reduction of host risk of infection to one strain after infection with another strain. Different scenarios are performed with two strains emerging with different time-lags, for different levels of cross-protection. We show that a cross-protection of 50% between the two strains is sufficient to explain why we observe a single influenza illness peak in temperate countries. (C) 2004 Published by Elsevier B.V.