Most twin studies assessing genetic overlap between environmental measures and behavioural phenotypes have used bivariate models to partially account for the heritability of environmental measures (e.g. Pike et al. 1996). However, in some cases researchers have included several behavioural phenotypes and in doing so have explained a large portion of the heritability of the environment. For example, Saudino and Plomin (1997) showed that the heritability of the home environment could be entirely accounted for by its association with the child’s temperament and cognitive ability. That is, genetic factors involved in temperament and ability overlapped with those associated with the home environment to the extent that no residual genetic variance remained once this overlap was accounted for. However, neither phenotype alone was enough. Such a multivariate approach holds great promise: we may be able to entirely explain the heritability of environmental measures via their association with multiple heritable traits. It is important that such attempts are made to explain ‘why’ studies find environmental variables to be heritable—by highlighting likely pathways and mechanisms.
