Both twin and molecular genetic models indicate genetic confounding in the relationship between executive function, socioeconomic status, and parenting in middle childhood.

Abstract

While there is a wide range of environments that have been implicated in executive function (EF) development, most are related to the behavior of genetic relatives such as parents, meaning there is significant risk that genetics confound these relationships. Using data from 7,925 ethnically and socio-economically diverse Adolescent Brain Cognitive Development cohort study participants aged 9-10, we examined evidence of genetic confounding in EF's relationship with two socioeconomic status (SES) variables (maternal education and household income), and two parenting measures (family conflict and parental monitoring), using multiple methods. Twin models found that relationships between parenting and EF variables were best explained by genetics, which was replicated by bivariate genome-wide complex trait analysis (GCTA). Bivariate GCTA also found that SES-EF relationships displayed substantial genetic confounding, consistent with univariate twin findings of no shared environment. However, polygenic scores (PGS) provided no significant evidence of genetic confounding, consistent with their lower proportion of genetic variance captured. These results provide strong evidence of genetic confounding in EF in middle childhood, and should encourage researchers to be more aware of the risk of genetic confounding in their designs. They also indicate that PGS are poor controls for genetic confounding in EF, necessitating other methods be employed to address their limitations.