Using Genome Wide Estimates of Heritability to Examine the Relevance of Gene-Environment Interplay

Benjamin Domingue, University of Colorado at Boulder
Jason D. Boardman, University of Colorado at Boulder

We use genome-wide data from the third generation respondents of the Framingham Heart Study to decompose variation in body mass index (BMI) into genetic and environmental components. We estimate that forty-three percent of the variation in BMI is due to additive genetic influences and the remaining fifty-seven percent is due to environmental factors. We then recalculate this heritability estimate after removing blocks of single-nucleotide polymorphisms from a genome wide association model (GWAS) using nearly 250,000 SNPs. Our results suggest that the bulk of the variance of BMI is explained by roughly 20,000 SNPs. Importantly, genetic loci that demonstrate environmental sensitivity provide very little information for population estimates of heritability.