ABSTRACT
This book presents the main statistical models for the analysis of longitudinal data in general. This term longitudinal data actually covers two types of data: firstly, repeated measurements of quantitative or qualitative variables often called “longitudinal data” (in a strict sense), and secondly the time of occurrence of events which are usually censored (called “survival data”). A common feature of these data is that their collection requires longitudinal monitoring; similarly, modeling requires taking into account the time factor (hence the term dynamic models). Typically, these data are collected in cohort studies. For example, in the Paquid cohort study on cognitive aging (Dartigues
et al., 1992), cognitive abilities of the subjects were measured at each visit using psychometric tests, and the age of occurrence of events such as dementia, Alzheimer’s disease or death was also recorded. In the field of human immunodeficiency virus (HIV), monitoring infected patient in observational studies and in clinical trials provides repeated measurements of biological markers such as the concentration of CD4 + T cells (CD4 counts) or viral load; it also allows recording the occurrence of various clinical events (opportunistic diseases, death . . . ).
