The CCC model assumes that the genetic and environmental influences on initiation influence progression via the beta pathway (Fig. 1). It is also plausible that, for example, it is only genetic influences on initiation that influence progression. Simulation work by Heath and colleagues [43] suggests that initiation and progression estimates and the size of the relationship between initiation and progression are estimated correctly using the CCC approach, but that the genetic and environmental correlations between initiation and progression can be biased. However, their alternative requires at least three categories in the initiation stage, of which at least two lead to information being available on progression, and this may not always be appropriate. Information that would have allowed the coding of initiation in this manner was not available, but would provide an interesting and important extension to the present study.
