In the sample of unrelated youths (GRM cutoff of .025) estimates of the aggregate SNP effects are small and highly unstable (Figure 2 in green), perhaps due to less phenotypic variability and a relatively small sample. When evaluating the full sample of youths (blue), which does not control for non-SNP genetic relatedness and shared environmental confounding, estimates range from .70 to .75, perhaps indicating a stronger role of shared environment in the youth-only sample versus the full sample. In fact, if we sum the heritabilities and shared environmental components reported in Figure 1, we find that they are strikingly similar to the GCTA estimates on the full set of offspring.
