Does Education Moderate the Genetic Risk For Obesity?

Tuesday, June 14, 2016
Lobby (Annenberg Center)

Author(s): Aparna Lhila; Vikesh Amin


Obesity rate have roughly doubled in the United States over the past 25 years, to the current rate of 35% of the adult population (CDC, 2015). This differs starkly by educational attainment – 30.4% of individuals without a high school diploma are obese whereas 9.5% of college graduates are considered obese. This is perhaps not surprising as the link between education and health is well recognized in the literature and a variety of identification strategies have been employed to estimate the causal effect.


In this paper we shed light on another way that education can affect obesity – by moderating the genetic risk of obesity. Greater education attainment may trigger mechanisms (e.g. increased psychosocial skills, lower-stress jobs, access to better quality medical care, food, and recreation, and higher income) that might mitigate the inherent genetic risk for obesity. In other words, higher levels of education may lower obesity by protecting against or overcoming the genetic risk of greater BMI.

We plan to use wave 4 of the Add Health data, which provides information on respondents’ genetic risk score (GRS) for BMI. We will estimate a series of regression equations to understand the relationship between education, GRS, and BMI.

(1)   BMIi=b0 + b1GRS +b2(Education) + b3(Control Variables)

(2)   BMIi=b0 + b1GRS +b2(Education) + b3(Control Variables) + b4(Education)*(GRS)

(3)   BMIi=b0 + b1GRS +b2(Education) + b3(Control Variables) + b4(Education)*(GRS) + Sibling Fixed Effects

Equation (1) estimates the associations of GRS and education with BMI, controlling for other SES measures (age and parental income and education). Equation (2) adds the interaction between GRS and education to test if the relation between GRS and BMI varies with education. Equation (3) includes sibling fixed effects to purge family-level unobserved heterogeneity that may potentially bias the relationship between education and health. We hypothesize that highly-educated individuals are better able to mitigate their genetic risk for obesity compared with their less-educated counterparts.