Abstract / Summary
Abstract Background Influenza remains a global public health burden. While interpersonal contacts drive transmission, their association with individual-level reinfection risk is not well quantified, particularly in the post-COVID-19 period. We assessed the relationship between contact frequency and serological evidence of influenza reinfection across three consecutive seasons. Methods In this multicentre, population-based panel study conducted between 2020 and 2024, binding antibody titers against hemagglutinin antigens of three influenza subtypes (A(H1N1), A(H3N2), and B) were measured by a bead-based multiplex immunoassay. Serological evidence of influenza reinfection was defined by fold-changes relative to the preceding season; with adjustments to minimize misclassification due to influenza vaccination. Logistic regression models were designed to estimate crude odds ratios (ORs) and DAG-informed adjusted odds ratios (aORs) for reinfection associated with total, household, and non-household contacts. Results Serological reinfection was detected in 14% (2021/22), 17% (2022/23), and 12% (2023/24) of participants. Among self-reported unvaccinated participants, reinfection was detected in 6%, 13% and 17% across the respective seasons. Non-household contacts were associated with increased odds of reinfection in one season (2022/23: aOR 1.008, 95% CI 1.002–1.014), while household contacts were associated with higher odds of reinfection in two seasons (2022/23: aOR 1.285, 95% CI 0.905–1.823; 2023/24: aOR 1.267, 95% CI 0.685–2.343). Conclusions Serological evidence of reinfection among unvaccinated participants in comparison with all participants aligned more closely with surveillance data in 2021/22. Residual misclassification of vaccination or preceding infections may partly explain the absence of an association in 2021/22. Associations between contact behavior and reinfection were season-specific, with household contacts showing the strongest and most consistent effects.