Long-term antibody waning and antibody-associated protection across diverse SARS-CoV-2 immune histories

Abstract Background SARS-CoV-2 vaccination and infection generate diverse immune histories, but antibody waning beyond 1 year remains insufficiently characterized. We estimated 2-year antibody trajectories and antibody-associated protection across different combinations of vaccination and infection history. Methods We conducted a repeated cross-sectional study using four nationwide seroepidemiological surveys of 25,800 adults in Japan between December 2021 and March 2023. We measured total antibodies against the ancestral spike receptor-binding domain and neutralizing antibody titers against Omicron BA.5. Bayesian hierarchical models estimated antibody waning according to vaccine dose number, infection history, Omicron-adapted bivalent vaccination, age, sex, and vaccine product. Antibody trajectories were linked to an antibody-risk model to estimate protection against symptomatic infection. Results Here we show that, among uninfected individuals, three to five vaccine doses produce higher anti-spike antibody titers than two doses, while antibody kinetics after the third or later doses are broadly similar. Relative to titers 30 days after the third dose, titers in the three-, four-, or five-dose groups decline to 0.18-fold (95% credible interval [CredI], 0.16–0.19), 0.19-fold (95% CredI, 0.13–0.26), or 0.21-fold (95% CredI, 0.13–0.33) at 1 year and 0.10-fold (95% CredI, 0.07–0.14), 0.14-fold (95% CredI, 0.06–0.24), or 0.14-fold (95% CredI, 0.05–0.29) at 2 years. Omicron breakthrough infection is associated with higher and more durable anti-spike and BA.5 neutralizing antibody titers than vaccination alone. Based on anti-spike antibody trajectories, estimated 90% protection against symptomatic BA.5 infection lasts 99 days (95% CredI, 89–108) after three doses and 116 days (95% CredI, 111–121) after four or five doses in uninfected individuals, but 276 days (95% CredI, 245–319) to 450 days (95% CredI, 343 to >730) in Omicron-infected individuals. Hypothetical neutralization-escape scenarios substantially shorten estimated protection. Conclusions These findings provide a quantitative framework for evaluating long-term antibody waning and booster timing across populations with diverse immune histories, and may inform the development of next-generation vaccination strategies that induce broader, more durable antibody responses against circulating SARS-CoV-2 variants.