Immune correlates of protection are exposure-dependent: evidence from the Dominican Republic

Immune protection against infection is often treated as a fixed property of a person's immune response. But infection risk also depends on the number of infectious contacts a person experiences, which generally increases as community transmission rises. This matters for vaccine effectiveness estimates and immune correlates of protection-immune measurements, such as antibody levels, used to infer protection-because repeated exposures can compound infection risk even when per-exposure susceptibility is unchanged. Here we show, using prospective febrile illness surveillance data from 1828 participants in the Dominican Republic during 2021-2022, that the antibody level associated with a given reduction in infection risk of PCR-confirmed SARS-CoV-2 infection increases during periods of higher community transmission. During high-transmission periods, infected participants have higher antibody titers than infected participants during low-transmission periods, narrowing the antibody difference between infected and uninfected individuals. Scaled logistic models showed a rightward shift in protection curves with increasing transmission, and discrete-time simulations reproduced this pattern through repeated exposures alone, without changing per-exposure susceptibility. These findings indicate that observed immune protection reflects not only pathogen characteristics and host immunity but also exposure frequency. Transmission context may therefore be important when interpreting correlates of protection, vaccine effectiveness estimates, and immunobridging studies.