Sara van den Berg

42 Chapter 2 Dependent variable Outcome GMT 2log titer Outcome protection titer ≥ 40 HAU or not Outcome seroconversion titer ≥ 40 HAU and 4-fold increase or not Factor CMV effect Factor CMV-serostatus CMV-serostatus positive / negative Factor CMV IgG group in CMV + individuals Anti-CMV IgG group Low (≤ 30U/ml) / medium (> 30 U/ml, ≤ 90 U/ml) / high (> 90 U/ml) Other factors Season 1 (H1N1pdm) Season 2 (H1N1pdm and H3N2) Sex Sex Male / Female Sex Male / Female Age Age group 18-30 season / 30-40 season / 40-52 season Age group 18-30 season / 30-40 season / 40-52 season Previous vaccinations Previous influenza vaccinations in last 3 seasons Never / Sometimes / Always Previous influenza vaccinations in last 3 seasons Never / Sometimes / Always Seasonal influenza vaccination 2009 before study Yes / No Seasonal vaccination 2009 Yes / No Seasonal influenza vaccination 2009 during study Yes / No Supplementary Table 1. Factors in Generalized Estimation Equation regression model for adjusted analysis of the effect of latent CMV infection on the antibody response to influenza vaccination. GEE models to correct for sex, age and previous influenza vaccinations are performed with one CMV factor in it (either CMV- serostatus or anti-CMV IgG level group) for one dependent variable (either titer or protection), to investigate the effect of latent CMV infection on the antibody response to influenza vaccination.