ABSTRACT
In Chapter 9, we discussed several approaches to identifying growth patterns/trajectories and estimating the effects of growth in different phases of the lifecourse on health in later life. One problem amongst many others is the collinearity between body size measurements during the lifecourse, which may cause unstable regression coef‰cients (see Chapter 6). Another problem is regression to the mean, such as that in tracing the mean z-scores for subjects with chronic diseases in later life.
