Incorporating heterogeneous parent-child environmental effects in biometrical genetic models

Heritability is an important index used to measure genetic effects in epidemiologic studies. For estimating heritability, researchers have proposed various statistical models to accommodate different study settings. These models include the ACE or ADE models for classical twin studies and the ACDE model for extended twin family studies. Researchers have shown that the ACDE model is less biased in heritability estimation, because of the utilization of the data of two generations. However, this model simply assumes that all family members, including twins and their parents, are exposed to the same environmental factors. This assumption may not be reasonable in many cases. In this paper, we propose a novel biometrical genetic model that can incorporate heterogeneous parent-child environmental effects. This model enables us to identify the heterogeneity of heritability, if it exists, between the two generations. This advantage is numerically demonstrated in our simulation studies. We apply our proposed model to the anterior chamber depth data from the Guangzhou Twin Eye Study in China. The analysis result reveals significant heterogeneity of heritability between the twin cohort and the parent cohort.