Fig. 2
Effect of B. longum + GI-OS on antibody responses to the influenza vaccine in young and older subjects. Data are log2 transformed geometric mean antibody titres (GMT) (±2SEM) at baseline and weeks 6 and 8 for n = 54–58 subjects per group. □ Maltodextrin, 
B. longum + Gl-OS. The reference line at the Y axis indicates seroprotection (GMT = 32). Data were analysed using a Linear Mixed Model (LMM) with fixed factors of time, age and treatment. a H1N1; For the LMM, there were significant effects of age (p < 0.001) and time (p < 0.001) and a trend for a treatment effect (p < 0.02). For data split by cohorts, there was a significant effect of time in both cohorts (p < 0.001), but no significant effect of treatment. b H3N2; for the LMM overall, there was a significant effect of time (p < 0.001) and an age*time interaction (p < 0.01). For data split by cohort there was a significant effect of time (p < 0.001) and a time*treatment interaction (p < 0.01) in the older cohort, and a significant effect of time (p < 0.001) in the young cohort. c Brisbane; for the LMM overall, there were significant effects of age (p < 0.005), time (p < 0.001) and an age*time interaction (p < 0.001). For data split by cohort, there was a significant effect of time (p < 0.001) and a time*treatment interaction (p < 0.01) in the older cohort, and a significant effect of time (p < 0.001) in the young cohort. There was a borderline effect of treatment (p < 0.05)