We used the GCTA software (Yang et al., 2011) to estimate heritability of body mass index (BMI) as measured in Add Health as part of the Wave 2, Wave III and Wave IV data collection. BMI was calculated using the standard formula of mass (kg) divided by height (m)squared (kg/m2) for each respondent. GCTA works by first estimating the genetic relatedness between all possible pairs of individuals. The genetic relatedness measures are known to be sensitive to population stratification, so for this application, we restricted the analysis to white respondents only. The subsequent step in the GCTA process is the estimation of a random effects model, where the random effects have a covariance structure based on the estimated genetic relatedness values. The percentage of total variance associated with the genetic random effects is considered the estimated heritability. For this particular study, we removed all pairwise relationship measures above 0.025.
