Language Development in Children

Overview

• Examines how the functional connectivity of core language regions, cortical growth, and language abilities develop in 175 children aged 4 to 9.
• Increased functional connectivity strength with age was seen between the pars triangularis of the left inferior frontal gyrus and left temporoparietal regions.
• Stronger connectivity between prefrontal and temporoparietal regions bilaterally correlated with better language abilities across all ages.
• Stronger connectivity between the left inferior frontal and temporoparietal regions correlated with larger surface area and thinner cortical thickness in those regions, which, in turn, correlated with better language abilities.

Introduction

• A core language comprehension network is mainly in the left inferior frontal and left superior temporal regions and is consistently reported in adults.
• Perisylvian activation occurs in infants and toddlers when listening to sentences.
• Children's language development involves bilateral activation in temporal regions and the triangular part of the inferior frontal gyrus (IFG).
• Adults mainly show activation in the left pars opercularis of the IFG and in the left temporal gyrus.
• There is an age-dependent increase in brain activation in bilateral superior and middle temporal regions, key parts of receptive language comprehension.
• Left superior temporal gyrus (STG) is present early in infants and is critical for integrating semantic and syntactic information.
• Activation of the left STG extends to the inferior parietal cortex and is hypothesized to aid in working memory for phonology and syntax, specifically in children.
• The functional language network extends anteriorly during development to include the left IFG.
• The left IFG is not yet involved in sentence comprehension in young children.
• The pars opercularis of the left IFG is not involved in sentence processing until adolescence.
• Functional connectivity studies have revealed age-related changes in the connection between the left IFG, inferior parietal cortex, and temporal gyrus during comprehension.
• A functional frontotemporal language comprehension network exists in children and the superior temporal cortex is initially involved, followed by the inferior frontal cortex.
• Maturation of functional language networks relies on structural networks, such as the white matter connection between the left IFG and posterior STG/STS.
• Gray matter regions linked to this fiber tract are associated with language skills and change with development.
• Myelination parallels cortical growth across frontal, temporal, and parietal regions and is associated with language comprehension changes.
• Functional and structural language comprehension networks develop across childhood and adolescence.
• The study investigates the functional language network's developmental trajectory in young children aged 4-9, and explores its cortical structures.

Methods

• Participants: 221 children (4-9 years)
• MRI Scans: Resting-state fMRI and T1-weighted structural MRI.
• Language Abilities: Tested in a subgroup, with some children re-invited for follow-up assessments.

Results

• Functional connectivity of language network changes with age.
• Examined correlation of language-related regions of interest (ROIs) with language abilities.
• Associations assessed between functional connectivity and cortical morphology.
• Correlation between structure maturation and language was also analyzed.
• Hypothesis: Functional connectivity changes during development are particularly expected between the left inferior frontal regions and left temporoparietal regions.
• Goal: Examine whether changes in functional connectivity are accompanied by the maturation of brain structures involved.

Participants

• Initial pool: 221 children aged 4-9 with complete structural and functional MRI scans.
• Exclusions: 46 children excluded due to bad image quality, low IQ, or being left-handed.
• Final sample: 175 children (84 female, mean age 6.13 years) included in analyses.
• Handedness: Assessed using Edinburgh Handedness Inventory; those with scores above 20 were considered right-handed.
• IQ: Non-verbal IQ available for 136 children, assessed using German versions of the Kaufman Assessment Battery for Children or the Wechsler Preschool and Primary Scale of Intelligence.
• All participants were native German speakers with no history of medical, psychiatric, or neurological disorders.

Behavioral Language Test

• The Test zum Satzverstehen von Kindern [TSVK] (sentence comprehension test for children) assessed the general sentence comprehension abilities of participating children.
• Task: Picture matching in which children hear a sentence and select the matching picture.
• Total: 36 items testing sentence complexity through word order, tense, mode, clause number, pronoun type, and verb type manipulations.
• Scoring: The number of correct responses converted to standardized T scores (mean = 50, SD = 10).
• Timing: Behavioral assessment taken within 6 months of the MRI scan (mean interval 0.02 years).

Data Acquisition

• Neuroimaging scanner: 3.0 Telsa Siemens MRI with a 12-channel head coil.
• Resting-state fMRI parameters:
  • Sequence: T2*-weighted, gradient-echo, EPI.
  • TR = 2000 ms, TE = 30 ms, flip angle = 90°.
  • Slice thickness = 3 mm, FOV = 192 mm, matrix = 64 x 64.
  • 28 slices, 100 volumes, spatial resolution = 3 × 3 × 3.99 mm³.
• Resting-state instructions: Lie still with eyes open, watch a lava lamp screensaver.
• T1-weighted MP-RAGE parameters: TR = 1480 ms, TE = 3.46 ms, TI = 740 ms, flip angle = 10°, image matrix = 256 × 240, FOV = 256 × 240 mm², 128 sagittal slices, spatial resolution = 1 × 1 × 1.5 mm³.

Data Preprocessing

• T1-weighted images were visually inspected.
• Reconstruction via FreeSurfer toolbox (version 7.1.0).
• Includes skull stripping, gray matter segmentation, surface model reconstruction, surface inflation, registration, parcellation and surface creation.
• Reconstructed surfaces were visually checked for inaccuracies.
• Calculation of cortical thickness (distance from the pial surface to the white matter) and surface area at each vertex.
• Vertex-wise maps aligned to the FreeSurfer fsaverage surface-based template.
• Thickness and surface area maps smoothed using a 25-mm FWHM Gaussian kernel.
• fMRI preprocessing done using FSL (version 5.0.11) and AFNI (version 19.1.05n). It includes:
  • Removing first 3 volumes, slice timing correction, re-aligning, and coregistration.
  • T1 segmentation into gray matter, white matter, and cerebrospinal fluid.
• Removal of the first 3 volumes
• Slice timing correction
• Re-aligning for head motion correction
• Coregistration with, and segmentation of, T1 images
• Parameters regressed out include: 24 motion parameters (Friston et al., 1996) and Principal components from WM and CSF
• Volume-to-volume framewise displacement (FD) (Jenkinson et al., 2002; Power et al., 2012).
• Children with a maximum of 3mm or 3° head motion were discarded from the analysis.
• Any child exceeding a 20% FD total volumes larger than 0.3 mm Mean was excluded

ROI Based Functional Connectivity

• Language is identified in the left IFG and left superior temporal region ROIs consisted of:
  • Ventral (STSvp)
  • Dorsal (STSdp)
  • TPOJ1 (temporo-parieto-occipital junction 1)
  • Dorsal anterior STG (STSda)

Statistical Analysis

• Linear Regression Analysis
• Sex and Head motion used as covariant of no interest.
• Gaussian Random field Theory at voxel-wise P value of 0.001
• Statistical map at cluster level
• Correlation between age-related FC with language raw stores.
• FCtri-IPL/STSp & age are used a predictor of of the language ability.
• Correlations were estimated by bootstrapping with 500 Iterations.

Functional Connectivity of Language-Related ROIs Result

• Left pars opercularis had a positive correlation with bilateral IFG, STG, Middle temporal gyrus, (MTG).

The role of Age in brain functional connectivity and Language Performance

• FC between left pars triangularis and left IPL extending to left STG/STSp was positively correlated with age