Neuroanatomical correlates of the income-achievement gap

Abstract

In the United States, the difference in academic achievement between higher- and lower-income students (i.e., the income-achievement gap) is substantial and growing. In the research reported here, we investigated neuroanatomical correlates of this gap in adolescents (N = 58) in whom academic achievement was measured by statewide standardized testing. Cortical gray-matter volume was significantly greater in students from higher-income backgrounds (n = 35) than in students from lower-income backgrounds (n = 23), but cortical white-matter volume and total cortical surface area did not differ significantly between groups. Cortical thickness in all lobes of the brain was greater in students from higher-income than lower-income backgrounds. Greater cortical thickness, particularly in temporal and occipital lobes, was associated with better test performance. These results represent the first evidence that cortical thickness in higher- and lower-income students differs across broad swaths of the brain and that cortical thickness is related to scores on academic-achievement tests.

Keywords: academic achievement; adolescent development; brain; cognitive neuroscience.

© The Author(s) 2015.

Figures

Figure 1

Volume and surface area differences between income groups. HI: Higher Income, LI: Lower Income. Volume and surface area measurements are adjusted for sex and estimated intracranial volume.

Figure 2

Cortical thickness differences between income groups. A: Higher-Income > Lower-Income. Sex is included as a nuisance regressor. Results are cluster-corrected for multiple comparisons (cluster-forming p < .05, cluster-wise p < .05, adjusted for both hemispheres). Cluster statistics are shown in Table 1. Results are displayed on inflated surfaces, with darker gray indicating sulci, and lighter gray indicating gyri. B: Cortical thickness for each group. Cortical thickness in millimeters is displayed for each group separately, overlaid with the significant results from part A. Histograms represent the number of vertices for each thickness value, and show the color scale plotted on the cortical surfaces. The top row of histograms shows the LI group in color, and the HI group in gray, and the bottom row shows the HI group in color and the LI group in gray.

Figure 3

Cortical thickness is positively correlated with test scores. Sex is included as a nuisance regressor. Results are cluster-corrected for multiple comparisons (cluster-forming p < .05, cluster-wise p < .05, adjusted for both hemispheres). Results are displayed on inflated surfaces, with darker gray indicating sulci, and lighter gray indicating gyri. Scatter plots show cortical thickness values extracted from significant clusters, adjusted for sex (cluster statistics are shown in Table 1).