Table 1.

Associations between complement fixation, antibody responses, and time using latent growth curve modeling (n = 30)

Outcomea	bb (s.e.)	95% CI	p valuec
log C1q	− 0.42 (0.03)	− 0.49, − 0.36	< 0.001
log IgG1	−0.33 (0.02)	− 0.37, − 0.28	< 0.001
log IgG2	− 0.30 (0.02)	− 0.33, − 0.26	< 0.001
log IgG3	− 0.32 (0.02)	− 0.36, − 0.28	< 0.001
log IgG4	− 0.03 (0.004)	− 0.04, − 0.02	< 0.001
log IgM	− 0.12 (0.02)	− 0.16, − 0.09	< 0.001

^aEffects for each factor represent separate linear mixed modeling (LMM) analyses where the natural log of C1q-fixation and each antibody response is regressed on the natural log of time (per month). These LMMs applied a random intercept for study participant, random slope for log time with an unstructured covariance parameter specified for the correlation between the random effects

^bRegression coefficient (b) and standard error (s.e.) represent the percent change in C1q-fixation or antibody response for a percent increase in time

^cProbability values based on Wald statistics