Applying Regression Models to Longitudinal Data
One basic question in longitudinal studies is the existence and degree of a general trend over time. A typical example is the course of a disease after a treatment. Figure 21.1 shows the development of the Hamilton depression score in 17 patients over 3 months, with baseline measurements at time 0 prior to the start of the therapy. We can observe a more or less distinct trend in many patients to lower values of the Hamilton depression score, that is, some improvement. So, a first basic question here is whether we can conclude that there is overall a tendency for a decrease of the depression score over time.