A review on linear mixed models for longitudinal data, possibly subject to dropout

Abstract

Many approaches are available for the analysis of continuous longitudinal data. Over the last couple of decades, a lot of emphasis has been put on the linear mixed model. The current paper is dedicated to an overview of this approach, with emphasis on model formulation, interpretation and inference. Advantages as well as drawbacks are discussed, and guidelines are given for general statistical practice. Special attention is given to the problem of missing data, i.e., the case where not all data are present as planned in the original design of the study.