ON THE ANALYSIS OF MIXED LONGITUDINAL GROWTH DATA

Mixed longitudinal growth data consists of several observations on a c haracteristic over a limited age range for each individual in a study. This data is then combined to model growth over the total age range o f all individuals in the study. The limited data collected on each ind ividual precludes a subject-specific approach to modelling so that a p opulation-based approach must be used. Here we propose a method for th e analysis of mixed longitudinal data using linear models constructed from cubic splines to model both the mean and variance curves of an ob served characteristic. The method is illustrated on growth in height a nd head circumference for children in a recently collected mixed longi tudinal data set concerning the growth of Victorian schoolchildren, wh ere particular concern was the timing of growth spurts and increases i n variability.