Environmental Influences on Infant Cortical Thickness and Surface Area

Abstract

Cortical thickness (CT) and surface area (SA) vary widely between individuals and are associated with intellectual ability and risk for various psychiatric and neurodevelopmental conditions. Factors influencing this variability remain poorly understood, but the radial unit hypothesis, as well as the more recent supragranular cortex expansion hypothesis, suggests that prenatal and perinatal influences may be particularly important. In this report, we examine the impact of 17 major demographic and obstetric history variables on interindividual variation in CT and SA in a unique sample of 805 neonates who received MRI scans of the brain around 2 weeks of age. Birth weight, postnatal age at MRI, gestational age at birth, and sex emerged as important predictors of SA. Postnatal age at MRI, paternal education, and maternal ethnicity emerged as important predictors of CT. These findings suggest that individual variation in infant CT and SA is explained by different sets of environmental factors with neonatal SA more strongly influenced by sex and obstetric history and CT more strongly influenced by socioeconomic and ethnic disparities. Findings raise the possibility that interventions aimed at reducing disparities and improving obstetric outcomes may alter prenatal/perinatal cortical development.

Figures

Figure 1.

The 78 cortical regions of interest from the AAL atlas (Tzourio-Mazoyer et al., 2002) projected onto a representative neonatal brain. Due to its anatomical location, Heschl’s gyrus is not visible.

Figure 2.

Age at MRI plotted against average CT (a), age at MRI plotted against total SA (b), and gestational age at birth plotted against total SA (c) for all individual subjects.

Figure 3.

Significant associations between regional CT and postnatal age at MRI and gestational age at birth shown as percent change by day. Postnatal age at MRI ranges from 6 to 144 days and gestational age at birth ranges from 192 to 295 days. Regions where postnatal age at MRI and gestational age at birth did not emerge as important predictors in the variable selection are shown in white; these regions were not included in the selected models. Subcortical regions are in gray and were not analyzed.

Figure 4.

Regions having significant (negative) associations between CT measures and paternal education are presented onto the cortical surface. The percent difference in CT for every additional year of paternal education is indicated by the color bar. Regions where paternal education did not emerge as an important predictor in the variable selection are shown in white; these regions were not included in the selected models. Regions in gray were not analyzed.

Figure 5.

Significant associations between regional CT and maternal ethnicity are projected onto the cortical surface. Negative percent differences represent thinner cortices in infants of Caucasian mothers and positive percent differences show thicker cortices in infants of Caucasian mothers. Regions where maternal ethnicity did not emerge as an important predictor in the variable selection are shown in white; these regions were not included in the selected models. Regions in gray were not analyzed.