chapter  16
26 Pages

Joint models for high-dimensional longitudinal data

BySteffen Fieuws, Geert Verbeke

In many areas of the applied sciences, longitudinal studies are conducted in which some characteristics are measured repeatedly over time for a set of study participants. When a single characteristic is measured longitudinally, we refer to univariate longitudinal data. Throughout this chapter, the words outcome, response, and marker will be used interchangeably to refer to such a characteristic. Often, multiple characteristics are measured at each moment in time, instead of a single one. Such a collection of longitudinal measurements will be called multivariate longitudinal data. Indeed, at each moment in time, multiple outcomes are measured, and this occurs repeatedly over time. To give a flavor of the type of data, consider first the following examples of multivariate longitudinal profiles.