Integrating Genetics and Social Science: Polygenic Score Analysis of Human Development

Add to Calendar
Date and Time: 
Monday, December 18, 2017
2:00 pm – 3:30 pm
Dan Belsky
Duke University

Human traits and behaviors are heritable — they are influenced by genetic factors shared by relatives and passed from parents to children. Until recently, this heritability was a black box, something studied only indirectly by comparing relatives. In the last decade, genome-wide association studies have begun to unpack this black box. Its contents include clues to mechanisms of disease and drug sensitivity, but also new tools for social and behavioral scientists. Specifically, a discovery from GWAS is that trait heritability reflects the influence of very large numbers of genetic variants each with very small effects. GWAS results can be used to measure these influences in other datasets using a method called polygenic score analysis. We use polygenic score analysis to study how traits, behaviors, and social outcomes develop. The goal of our research is to identify behavioral and social mechanisms that mediate genetic risks. Ultimately, this work will allow hypothesis-driven tests of gene-environment interactions that can inform development of policies and programs to promote healthy life course development. The talk will present our recent research following-up GWAS of educational attainment in cohort studies from New Zealand, the United Kingdom, and the US. Polygenic score analysis in these cohorts revealed four findings: First, genetics discovered in GWAS of education are not genetics of education only, they are related to a pattern of life-long socioeconomic success that is partly independent of success in school. Second, these genetics are socially stratified; children born into better-off families tend to carry more education-associated genetic variants. Third, independent of this gene-environment correlation, children who carry more education-associated alleles tend to be upwardly socially mobile. Fourth, genetic associations with socioeconomic attainments are mediated by differences in cognitive and non-cognitive skills that emerge even before children enter school. Genetics discovered in GWAS of educational attainment were not associated with measures of children’s physical health. These findings provide molecular evidence for how genes and social environments become correlated within and across generations and suggest pathways for interventions aiming to promote upward socioeconomic mobility.