“…The outcome was defined as a nominal variable of LAZ trajectory-group membership and covariates were maternal education, age, midupper arm circumference, and height and infant sex, birth weight, length, and gestational age on the basis of the documented associations of these variables with linear infant growth (5, 8, 10, 13, 15, 40–45). To minimize the impact of collinearity on statistical inference, covariates were centered before model fitting.…”