Ability and heritability: Investigating the continuous effect of IQ score on IQ etiology in multiple samples

Biometrical analyses have concluded that high IQ, at least given a relatively unrestrictive definition of top 15%, does not appear to have the distinct causal influences one might expect of a distinct phenotype. However, it remains possible that there are quantitative changes in the causal factors (or in the magnitude of their influence) across the full distribution of ability. This possibility was examined using an extension to the DeFries-Fulker Regression method that allows examination of changes in the contribution of additive genetic and shared environmental influences as a continuous function of increase in the trait scores (Cherny, Fulker & DeFries, (1992), Behavior Gentics, 22(2), p.153–162). Analysis using a collaborative sample of 11,000 twin pairs from four countries demonstrated a linear increase in shared environmental influences as ability level increases. Examination of sample heterogeneity demonstrated that this pattern only applied to the portion of the sample that was above the age of 13, and corresponded to an increase from a value of 0 for the bottom half of the distribution to 0.39 for the top half. This pattern is suggestive of a persistence of shared environmental effects further into development for higher IQ individuals. This same pattern was seen in longitudinalanalysis of the Colorado Longitudinal Twin Sample, although the interaction effect seen at age 16 was only suggestive. The effect was additionally tested in sibling pairs from the Colorado Adoption Project. Results again showed a trend at age 16 but only in biological and not adoptive siblings. Implications of and possible explanations for this pattern of results will be discussed. © Springe