Genome-Phenome Linkages in Human Population Surveys, with Special Emphasis on the Health and Retirement Survey

Abstract

We review a diversity of genome-wide association studies (GWAS) with particular emphasis on precision in specifying phenotypes. This implies that examination of any specific phenotype involves considering the likely genetic contributions to it from the entire genome. We consider a variety of phenotypes specifiable with data from the Health and Retirement Survey (HRS). However, evidence from other large population studies is also incorporated as part of the process of developing and refining pathway representations from the genome thru a hierarchy of intermediate endpoints to behavioral, cognitive, and economic phenotypes. Any causal modeling focused on genome-phenotype connections must, of necessity, include consideration of intermediate endpoints (endophenotypes) as mediators of such associations. We also discuss metabolic and gene expression consequences of gene-environment interactions as a next research step beyond GWAS, not only for HRS but also for an integrated set of human population surveys that can provide much more statistical power than any one of them used alone. A variety of concrete examples based on physiological, psychological, sociological, and economic outcomes are carried along throughout our discussion.