PROJECT SUMMARY
Tackling obesity and depression is vital for population health, not least because they are associated with
diabetes, heart disease, strokes, and suicide, which in turn led to the decrease in US life expectancy in 2015.
One policy advocated by Healthy People 2020 to reduce obesity and depression is increasing educational
attainment, as it is viewed as a fundamental cause of health disparities. As obesity and depression also have
strong genetic bases, it is possible that education additionally improves population health by moderating the
underlying genetic risk of being obese/depressed through gene-environment (GxE) interactions. GxE
interactions occur when the effect of genetic risk varies across the environment. The influence of genetic
predisposition towards the likelihood of being obese may differ by educational attainment. A college graduate
with a high genetic risk of being obese may never be obese because education is associated with higher income
affording the individual to eat healthier and having peers who are conscious about leading a healthy lifestyle. In
contrast, a high school dropout with a high genetic risk is more likely to be obese because he/she has a low
income, is more likely to consume fast food, and interacts with peers who are obese. The proposed project will
apply econometric methods, instrumental variables and within-sibling comparisons, that use quasi-experimental
variation in educational attainment to increase our understanding of causal relations of education, genes, and
GxE interactions on obesity and depression. The use of quasi-experimental variation will, under certain
assumptions, produce estimates that are not confounded by (i) the influence of unobserved factors that are
correlated with education and health, (ii) reverse causality, (iii) gene-environment correlations, and (iv) population
stratification, and these estimates can therefore be given a causal interpretation conditional on these
assumptions. The specific aims are to estimate main effects of education, genetic risk, and GxE interactions on
obesity and depression using (i) the Easter School Leaving Rule as a natural experiment in the UK and data
from the UK Biobank, (ii) local factors in the individual’s county of residence near secondary school completion
age (college availability, education expenditures, unemployment rate) as instruments for education in the
National Health and Nutrition Examination Survey and the National Longitudinal Study of Adolescent Health (Add
Health) datasets, and (iii) the sibling fixed-effects approach in the Add Health dataset. Overall, the project will
provide a comprehensive array of results and new insights as we compare results from different methods,
datasets, across different educational levels, and by gender. This application is in direct response to NIH PAR-
16-080 “Education and Health: New Frontiers.”