Differences in educational opportunity predict white matter development

Abstract

Coarse measures of socioeconomic status, such as parental income or parental education, have been linked to differences in white matter development. However, these measures do not provide insight into specific aspects of an individual's environment and how they relate to brain development. On the other hand, educational intervention studies have shown that changes in an individual's educational context can drive measurable changes in their white matter. These studies, however, rarely consider socioeconomic factors in their results. In the present study, we examined the unique relationship between educational opportunity and white matter development, when controlling other known socioeconomic factors. To explore this question, we leveraged the rich demographic and neuroimaging data available in the ABCD study, as well the unique data-crosswalk between ABCD and the Stanford Education Data Archive (SEDA). We find that educational opportunity is related to accelerated white matter development, even when accounting for other socioeconomic factors, and that this relationship is most pronounced in white matter tracts associated with academic skills. These results suggest that the school a child attends has a measurable relationship with brain development for years to come.

Keywords: Development; Education; Socioeconomic Status; White Matter.

Fig. 1

A. Correlation matrix illustrating the univariate relationships between mean FA in the left arcuate, SEDA intercept, and other demographic and socioeconomic factors. Coefficients in bold represent correlations where FDR-corrected p<0.05.B. Beta-weights for linear mixed-effects models predicting mean FA in the left and right arcuate from a single predictor, specified on the x-axis. Each model included a random effects structure of family structure nested within scanner site. The colors of each bar denote each predictor variable. Error bars represent the standard error of each beta-coefficient. Bars that are bolded illustrate the beta-weights with FDR-corrected p<0.05.

Fig. 2

A. Renderings of the five white matter tracts significantly related to SEDA intercept. These include the left arcuate fasciculus, right cingulate cingulum, and the motor, superior parietal, and temporal bundles of the corpus callosum. Shading represents the -log₁₀(p-value) for the beta-weight on SEDA intercept from the models predicting FA in each tract (1.301 corresponds to a p-value of 0.05). This association was strongest in the left arcuate (yellow in the top panel). B. Beta-coefficients for the fixed effects of the models predicting FA in each major white matter tract. The x-axis represents a specific bundle identified by pyAFQ. Each row and color in the figure refers to the fixed-effect in each model. Error bars represent the standard error of each beta-coefficient. Bars that in bold illustrate the beta-weights with FDR-corrected p<0.05.

Fig. 3

Left: Growth trajectories for Diffusion Kurtosis (DKI) FA in the left and right arcuate across the first two observations of the ABCD study. The red and blue lines represent the average DKI FA growth trajectories for individuals in high (Intercept = 1) or low SEDA (Intercept = −1) intercept schools, respectively. Gray lines represent the observed changes in FA in the left and right arcuate for each individual present in the dataset. Right: Mean residual values for the model predicting Brain-Age Gap from a reduced model that excludes SEDA intercept as a predictor, but retains all other random and fixed-effects. Each bar represents either the top (red) or bottom (blue) 20% of participants based on their SEDA intercept scores. Error bars represent one standard error from the mean.