Gene–environment interaction and twin studies. In

Gene-environment interaction (GxE) refers to the fact that different geno-types respond differendy to the same environment. GxE is related to the statistical concept of heteroscedasticity, implying heterogeneity of variance. Monozygotic (MZ) twins can be used to detect systematic heteroscedasticity, but it is important first to defme the scale of interest. For a continuous trait (eg. bone density or blood pressure) a non-normal distribution will generate scale-dependent GxE which can be removed by appropriate transformation. For a dichotomous trait (eg. disease status) non-additivity on the risk scale can occur even when there is additivity on the liability scale that underlies it, and the reverse applies too. MZ twins provide a unique opportunity to detect various types of GxE. For measured genotypes, MZ intrapair variances should be compared to detect genes affecting environmental sensitivity. Such "variability " genes are potentially even more important than "level " genes and provide a sound rationale for collecting DNA from MZ as well as DZ twins