Through the weighted genome-wide association analysis, we observed association signals that align with previous and in some cases, established genetic variants associated with BMI. For example, we were able to identify the FTO region that has been previously identified and replicated (Frayling et al., 2007). Additionally, we were able to identify variants that are upweighted through the GIANT consortium results, but do not achieve genome-wide significance in either GIANT or the Add Health sibling pairs subsample. These variants may be of particular interest for researchers who wish to explore GxE interactions in the Add Health sample to further explain the variability of the effect of these variants on BMI over time (age and development), behaviors and under particular environmental contexts.
