Prenatal environment is associated with the pace of network development over the first 3 years

1285 

Abstract Submission 

URSULA TOOLEY¹, Aidan Latham², Jeanette Kenley², Dimitrios Alexopoulos², Tara Smyser², Barbara Warner², Joshua Shimony², Jeffrey Neil², Joan Luby², Deanna Barch³, Cynthia Rogers², Christopher Smyser²

¹Washington University in Saint Louis, Saint Louis, MO, ²Washington University in St. Louis, St. Louis, MO, ³Washington University, Saint Louis, MO

URSULA TOOLEY, Ph.D.  
Washington University in Saint Louis
Saint Louis, MO

Aidan Latham  
Washington University in St. Louis
St. Louis, MO

Jeanette Kenley  
Washington University in St. Louis
St. Louis, MO

Dimitrios Alexopoulos  
Washington University in St. Louis
St. Louis, MO

Tara Smyser  
Washington University in St. Louis
St. Louis, MO

Barbara Warner, Ph.D.  
Washington University in St. Louis
St. Louis, MO

Joshua Shimony, Ph.D.  
Washington University in St. Louis
St. Louis, MO

Jeffrey Neil, Ph.D.  
Washington University in St. Louis
St. Louis, MO

Joan Luby, Ph.D.  
Washington University in St. Louis
St. Louis, MO

Deanna Barch, PhD  
Washington University
Saint Louis, MO

Cynthia Rogers, Ph.D.  
Washington University in St. Louis
St. Louis, MO

Christopher Smyser  
Washington University in St. Louis
St. Louis, MO

Environmental influences on brain structure and function during development have been well-characterized, and the pace of early brain development has been associated with important risk factors and behavioral outcomes (Shaw et al. 2010; Farah 2017). As children mature, intrinsic cortical networks become more segregated, with sets of brain regions displaying more densely interconnected patterns of connectivity and large-scale systems becoming increasingly distinct (Grayson & Fair 2017). Some theoretical models posit that environmental influences on brain development might arise by way of effects on the pace of brain development, such that brain development proceeds faster in neonates and toddlers from lower-SES backgrounds (Tooley et al. 2021).

In a set of pre-registered analyses (AsPredicted #128836), we explicitly test whether early SES is associated with differences in the pace of intrinsic cortical network segregation during the first three years of life. We capitalize on a unique cohort of neonates and toddlers (n=261, M=41.3 weeks at first scan) with longitudinal fMRI neuroimaging data and extensively characterized early environments (Stout et al. 2022), using generalized additive mixed models (Wood 2004) to examine moderating effects of prenatal SES (Luby et al. 2023) on development of cortical network segregation, controlling for sex at birth, amount of uncensored data included, in-scanner motion, and average network connectivity. We take a hierarchical approach, first examining measures of cortical network segregation (Figure 1a-c) at whole brain resolution, then analyzing at the level of functional brain systems, and finally in individual brain regions. Finally, we examine whether differences in measures of cortical network segregation at age two years are associated with language and cognitive abilities (Bayley Scales of Development-III).

Cortical network segregation increases with age during the first three years of life (Figure 1d-f), and prenatal SES significantly moderates trajectories of cortical network segregation across scales (Figure 1g-i). Neonates and toddlers from lower-SES backgrounds show a steeper increase in cortical network segregation with age, consistent with accelerated network development. We find that associations between SES and cortical network segregation are present primarily at the local scale. Effects of prenatal SES are strongest in the somatomotor and dorsal attention systems (Figure 2a-b) and conform to a sensorimotor-association hierarchy of cortical organization (Figure 2c). Importantly, SES-associated differences in cortical network segregation are associated with language abilities at age two years, such that lower segregation is associated with improved language abilities, even when controlling for prenatal SES (Figure 2d-f).

We find that the development of cortical brain networks during the first three years of life is strongly associated with features of the early environment, suggesting these influences may play a key role in shaping this trajectory. Being born into a more advantaged (higher SES) environment is associated with a more protracted trajectory of cortical functional network development in early childhood; more protracted cortical network development might reflect prolonged periods of plasticity or alterations in synaptic proliferation and pruning (Tanti et al. 2013; Manzano Nieves et al. 2020). Importantly, environmental influences on development of cortical network segregation might underlie SES-associated differences in language abilities observed later in development. Our results suggest that infancy and toddlerhood may be an important period for promoting healthy brain development, emphasizing the first years of life as a target for policies supporting optimal child development.

Language Acquisition

Normal Brain Development: Fetus to Adolescence ¹

fMRI Connectivity and Network Modeling ²

Task-Independent and Resting-State Analysis

BOLD fMRI

Computational Neuroscience

Development

FUNCTIONAL MRI

Language

Pre-registration

Preprint

Other - infant & toddler, prenatal environment, graph theory, socioeconomic status