Most health behavior outcomes of interest to social demographers are affected by both environmental and genetic factors but current methods for handling these two predictors of interest tend to be restricted to specific (e.g., sibling or famil-based) study designs. In the proposed project, we will develop a multi-level model that can account for both types of predictors across multiple study designs and, perhaps even more importantly, can utilize any and all genetic information that is available (e.g., estimated genetic relationship between related AND unrelated pairs using genome-wide data or assumed relationship based on family structure). In this study, we first verify the proposed method via a series of detailed and extensive simulations. We will then demonstrate the usefulness of this method empirically using genetic and phenotypic data of interest to demographers (e.g., BMI and smoking) from two generations of family members and unrelated respondents in the Framingham Heart Study.
The goal of this project is to enhance research in gene-environment interplay for health and health-related phenotypes by developing a user-friendly and flexible statistical method to incorporate large amounts of data from the genomes of respondents into standard social demographic analyses. This will facilitate greater entry of social scientific researchers into this field of inquiry who may not otherwise do so. It will also enhance the use of existing data sources supported by NIH that contain (or will have very shortly) genome wide data from respondents.
|Boardman, Jason D; Domingue, Benjamin W; Daw, Jonathan (2015) What can genes tell us about the relationship between education and health? Soc Sci Med 127:171-80|
|Domingue, Benjamin W; Fletcher, Jason; Conley, Dalton et al. (2014) Genetic and educational assortative mating among US adults. Proc Natl Acad Sci U S A 111:7996-8000|