Fig. 2: Impact of host factors, repeated vaccination, and pre-vaccination titers on the fold-rise in HAI titers.

The relationship of age (in decades, Age/10), frailty status, and BMI (five-unit intervals, BMI/5) with log₁₀-transformed fold-rise in titers (post-titer/pre-titer) are shown as slopes for H1N1 (a), H3N2 (b), and influenza B (c). The relationship between increasing years of vaccination and the log₁₀-transformed fold-rise in titers are shown for H1N1, H3N2, and influenza B (d–f), with the slopes summarized (g). The relationships between pre-vaccination HAI titers and the log₁₀-transformed fold-rise in titers are shown for each vaccine antigen (h–j), with the slopes summarized (k). Estimates and 95% confidence intervals were derived from multi-level mixed-effects models with random intercepts on the individual participant. Models controlled for influenza season and pre-vaccination HAI titers (a–g), and either controlled for sex (whole population estimates) or used interaction terms between sex and the host factor of interest to derive sex-specific estimates.