Multivariate Adaptive Splines for Analyzing Longitudinal Data

Abstract In a longitudinal study, the outcome variable is measured repeatedly over time, together with covariates that may or may not vary over time. This article describes MASAL — a nonparametric method for analyzing and exploring longitudinal and growth curve data. It is based on a high‐dimensional smoothing technique, accommodates time‐varying covariates, and allows unrestricted interactions among covariates and between time and covariates.