Fig. 5: Yellow fever neutralizing antibody titer variation across living location and with microbiota profiles (N = 121).

a–f Boxplots (see legend Fig. 2a, b) showing yellow neutralizing antibody titers (PRNT₅₀) in vaccinated individuals from whom a stool sample was collected (N = 121) at baseline (day 0, d0), 4 weeks post-vaccination (day 28, d28) and 6 months post-vaccination (day 178, d178). Participants were compared according to group (rural/urban living location; (a) CST (c) and group/CST (i.e. rural living individuals, rural-like urbanites [CST1] and urban living individuals [CST2]) (e). Boxplots showing yellow neutralizing antibody titers (PRNT₅₀) at 4 weeks post-vaccination (day 28, d28) compared to 6 months post-vaccination (day 178, d178) for group (rural/urban living location; (b), CST (d) and group/CST (f). Diamond-shaped data points and color values at the bottom of the plot denote log₁₀-base geometric mean titers. Statistical analyses were performed using a generalized linear mixed effects model with log₁₀-transformed PRNT₅₀-values as outcome, group, CST or group/CST, time point, the interaction between group, CST or group/CST and time point, age and sex as fixed effects and participant ID as random effects. Pairwise comparisons of estimated marginal means between groups were computed at each time point (a, c and d) or between day 28 and day 178 for each group (b, d and f). Values under the limit of detection (<10) were assumed to have a value of 5.