Table 2.
Summary of linear mixed effects models fit to examine the relationship between anti-S IgG log10 AU/mL (light grey) or anti-N IgG log10 AU/mL (dark grey) and COVID-19 vaccination status adjusting for: biological sex, age, and time from qPCR diagnosis.
| Unconditional Mean Model (S) | Intraclass Correlation Coefficient | ||
|---|---|---|---|
| Participant ID (n = 42) | 0.434 | ||
| Residual | 0.566 | ||
| Random Intercept Model | Variable | Fixed Effect Estimate | 95%CI |
| Anti-Spike IgG | Intercept | 4.84 | 3.27–6.39 |
| Vaccine-Yes | 0.40 | −0.41–1.20 | |
| Biological Sex-Male | 0.93 | 0.068–1.79 | |
| Age (Years) | −0.029 | −0.063–0.0057 | |
| Time from +ve qPCR Test * | −0.20 | −0.47–0.054 | |
| Vaccine: Time * | 1.86 | 1.39–2.21 | |
| Random Effects | Intraclass Correlation Coefficient | ||
| 0.893 | |||
| Unconditional Mean Model (N) | |||
| Participant ID (n = 42) | 0.875 | ||
| Residual | 0.125 | ||
| Random Intercept Model | Variable | Fixed Effect Estimate | 95%CI |
| Anti-Nucleocapsid IgG | Intercept | 3.14 | 2.48–3.79 |
| Vaccine-Yes | −0.080 | −0.42–0.26 | |
| Biological Sex-Male | 0.27 | −0.095–0.63 | |
| Age (Years) | 0.016 | 0.0017–0.03 | |
| Time from +ve qPCR Test * | −0.40 | −0.53–(−0.27) | |
| Vaccine: Time * | −0.077 | -0.25–0.11 | |
| Random Effects | Intraclass Correlation Coefficient | ||
| 0.30 | |||
* An effect modification term was incorporated to explore how the effect of vaccination on antibody concentration differs by time since diagnosis with a qPCR test. Unconditional means models were fit to partition the variance by participant without inclusion of other exposure variables. Fixed effect models were built by applying the common cause criterion to select covariates which are a cause of the exposure, outcome, or both.