Head Circumference at Birth and Gender Predict EEG Relative Power in Resource-Poor Communities in South Africa

• In Event: Longitudinal effects of early socioeconomic circumstance on infants' and children's brain and neurocognitive functions

Abstract

About 90% of the world's children grow up in low- and middle-income countries (LMIC; UNICEF, 2015), yet most of what we know about neurodevelopment and environmental factors comes from high-income countries (HIC). In HIC, a developmental shift has been observed in EEG power distributions, where power increasingly concentrates in higher frequency bands. Moreover, EEG relative power in HIC is sensitive to environmental risks, such that children growing up in low-income contexts have more slow-wave and less fast-wave relative power (Tomalski et al., 2013). However, studies linking environmental risk factors in LMIC to EEG relative power have found both contrasting (Jensen et al., 2021) and similar (Otero, 1994) neurodevelopmental patterns compared to HIC. Specific environmental risks or levels of exposure may lead to diverging neurodevelopmental patterns in LMIC. Therefore, it is essential to examine prospectively environmental risks as predictors of early childhood EEG relative power within a LMIC context.

This study leverages data from a birth cohort in resource-poor communities in Limpopo Province, South Africa, to examine how brain activity at 7 and 17 months relates to environmental risk factors. Specifically, we examine how relative EEG power in slow-wave (theta, 4-6 Hz) and fast-wave (gamma, 30-48 Hz) frequency bands relates to child gender, household assets, and head circumference at birth, an index of intrauterine growth restriction. Baseline EEG was collected for 6 minutes from 118 children across both timepoints (57 girls; all born full-term). Parents reported the number of household assets they owned, such as gas stoves or radios. Head circumference at birth was collected through hospital records.

Repeated measures ANOVAs were conducted separately for theta and gamma, with timepoint, hemisphere, and region as the within-subject factors; child gender as a between-subjects factor; and household assets and head circumference at birth as continuous predictors. Boys had more relative theta (F(1,111)=8.40, p=.005) and less relative gamma power (F(1,110)=6.29, p=.014) than girls across both timepoints. Larger head circumference at birth predicted more relative theta (F(1,111)=8.396, p=.005) and less relative gamma power (F(1,110)=5.88, p=.017). There were no main effects of household assets. Overall, relative gamma power decreased between 7 and 17 months (F(1,110)=10.78, p=.001).

In sum, the development of EEG power distribution was sensitive to child gender and head circumference at birth in resource-poor communities in South Africa; however, the direction of associations was opposite to what has been observed in HIC. Power was more concentrated in slow-wave frequency bands in male children and children with larger head circumferences at birth, who could be considered at less risk than females and children who experienced greater intrauterine growth restriction in the womb. Relative fast-wave power decreased over time, a pattern that contrasts with the developmental increase in relative fast-wave power observed in HIC and several LMIC contexts. It is possible that in these specific resource-poor communities, delaying neural maturation is an adaptive response, enabling children to take advantage of scant resources. Further longitudinal research is needed to determine how greater concentrations of slow-wave power in early childhood relate to cognitive outcomes within this LMIC context.

Authors

• Diana Carolina Lopera-Pérez, Boston University
  Presenting Author

• Peter Rockers, Boston University

• Jukka Leppänen

• Denise Evans

• Lezanie Coetzee

• Laurel J. Gabard-Durnam, Northeastern University

• Amanda R Tarullo, Boston University