Fig. 4

Anti-S IgG antibody levels and multivariate mixed-effects regression model estimates among university students with no history of COVID-19 or immunocompromised status, faceted by booster dose receipt — fall academic semester 2021, Wisconsin. Panels display levels of antibody targeting the spike protein of SARS-CoV-2 (anti-S), in WHO-standardized binding antibody units per milliliter (BAU/mL), log (10) scale. Sex-adjusted, mixed-effects linear regression estimates are superimposed on top of scatter plots with shading reflecting the 95% confidence intervals to estimate anti-S decline associated with primary series vaccine type (e.g., mRNA vs inactivated whole virus, or IWV). Panel A displays anti-S values of students with receipt of COVID-19 vaccine primary series across months since completion of primary series. Panel B displays anti-S values of students who received an mRNA booster dose across months since receipt of booster dose. Data outliers that were not included in the model based on Cook’s Distance are denoted as a boxed point and were not used to calculate adjusted regression lines