Joint modeling of a linear mixed effects model for self-esteem from mean ages 13 to 22 and a generalized linear model for anxiety disorder at mean age 33

Recent work has identified the transition from adolescence to young adulthood as a period with distinct characteristics that is important for understanding of human development. Self-esteem plays a critical role in this developmental process. We use self-esteem data measured at mean ages 13, 16 and 22, and anxiety disorder diagnosis at mean age of 33 to examine the impact of development of self-esteem on onset of adult anxiety disorder. To analyze these data, we propose a Bayesian joint model with: (1) a linear mixed effects model for the longitudinal measurements, and (2) a generalized linear model for the binary primary endpoint. Our analysis indicates that the mean level of self-esteem, not the change of self-esteem, significantly predicts the onset of adult anxiety disorder. A comparison shows that the joint model yields better predictive accuracy than a two-step model. The respective area under ROC curve (AUC) is 0.60 and 0.75 for the two-step model and the joint model, respectively. The two-step estimate may be biased because this method ignores variability in the individual random effects. We conclude that joint model is the most advantageous model to analyze early life longitudinal data combined with later binary outcome.