Fig 4. Mediational analyses showing indirect associations of lead risk on cognitive performance through mid-posterior corpus callosal (CC) volume.

The title of each row [e.g., “Executive Function (Card Sorting)”] refers to the construct and task used to assess that construct. The shading of the callosal regions, the mediator boxes (e.g., Mid-Posterior CC, Central CC), and histograms reflect the strength of the indirect association (per the color bar in the lower left corner). Callosal regions in dark gray were not incorporated as mediators in the corresponding analysis. Total effects of lead risk are represented by c, direct effects of lead risk are represented by c’, and a and b values refer to the associations of lead risk on callosal volume and callosal volume on cognitive performance, respectively. The subscript of the a and b values refer to the callosal subregion (i.e., C = central, MA = mid-anterior, MP = mid-posterior). All a, b, c, and c’ values are unstandardized regression coefficients. Thick arrows designate significant associations. Thin arrows designate non-significant associations. The presence of the statistical significance of indirect associations was determined by construction of bias-corrected percentile bootstrapping 95% confidence intervals (i.e., 10,000 bootstrapped samples), irrespective of the statistical significance of the corresponding total effect. The distributions of these bootstrapped indirect associations are shown, along with the original-sample indirect associations and corresponding confidence intervals. Bolded font indicates statistically significant indirect associations. Age, sex, parental education, race, ethnicity, and family income were included as covariates in analyses. Card Sorting = Dimensional Change Card Sort Task. Picture Vocabulary = Picture Vocabulary Test. Oral Reading = Oral Reading Recognition Test. Pattern Comparison = Pattern Comparison Processing Speed Test. Flanker = Flanker Inhibitory Control and Attention Test.