Modeling Gene‐Environment Interactions With Quasi‐Natural Experiments
Citations Over TimeTop 15% of 2015 papers
Abstract
This overview develops new empirical models that can effectively document Gene × Environment (G×E) interactions in observational data. Current G×E studies are often unable to support causal inference because they use endogenous measures of the environment or fail to adequately address the nonrandom distribution of genes across environments, confounding estimates. Comprehensive measures of genetic variation are incorporated into quasi-natural experimental designs to exploit exogenous environmental shocks or isolate variation in environmental exposure to avoid potential confounders. In addition, we offer insights from population genetics that improve upon extant approaches to address problems from population stratification. Together, these tools offer a powerful way forward for G×E research on the origin and development of social inequality across the life course.
Related Papers
- → The use of negative control outcomes in Mendelian randomization to detect potential population stratification(2021)115 cited
- → Text and Causal Inference: A Review of Using Text to Remove Confounding from Causal Estimates(2020)16 cited
- [Roaming through methodology. XXXVII. Unintentional population stratification in genetic case-control studies].(2002)
- → Text and Causal Inference: A Review of Using Text to Remove Confounding\n from Causal Estimates(2020)