Under-reporting of COVID-19 cases and the lack of information about circulating SARS-CoV-2 variants remain major challenges for many African countries to date. To address this, we present a comprehensive analysis of the dynamics of SARS-CoV-2 infections in Ethiopia, focusing on reinfection dynamics, (variant-specific) immunity, and the impact of vaccination rates. We conducted an antibody serology study, sequenced positive polymerase chain reaction tests, used available test positivity rates, and constructed two mathematical models integrating these data sets. A multivariant model was used to explore the variant dynamics and differential susceptibility to reinfection based on genetic distances between variants, identifying wildtype, alpha, delta, and omicron BA.4/5 as the four most important variants in Ethiopia. Cross-immunity between variants is investigated, revealing immunities ranging from 24% to 69% risk reduction. A second, antibody-level focused model, predicts slow antibody decay, particularly for Anti-S antibodies, leading to sustained high antibody levels until present times. Retrospectively, it suggests that increased early vaccination could have substantially reduced infections during the delta and omicron waves. However, as a large proportion of the population might have already had multiple infections that led to a strong immune response, further vaccination is less likely to have a significant impact now.
*The authors would like to note that Simon Merkt, Solomon Ali, Esayas Kebede Gudina, Jan Hasenauer, Arne Kroidl, and Andreas Wieser respectively contributed equally.