269 - Long-term persistence of bitemporal functional hyperconnectivity during stories listening task in 8- to 12-year-old children born extremely preterm

Poster Number: 269
Publication Number: 269.225

Maria E. Barnes-Davis, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, United States; Stephanie L. Merhar, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, United States; Alonzo T. Folger, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, United States; Jennifer Vannest, University of Cincinnati, Cincinnati, OH, United States; Mekibib B. Altaye, CCHMC, CINCINNATI, OH, United States; Nehal Parikh, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, United States; Darren S. Kadis, Hospital for Sick Children, Toronto, ON, Canada

Background: Birth before 28 weeks gestation (extreme prematurity, EPT) is associated with language deficits. The neural basis of language differences in children born EPT is unclear. Previously, we reported functional hyperconnectivity in EPT using fMRI-constrained magnetoencephalography (MEG). For EPT, hyperconnectivity was related to better language performance at 4-6 years.

Objective: To follow our cohort of children born EPT and term controls (TC) through 8-12 years of age and assess the degree to which functional hyperconnectivity persists and relates to language for EPT.

Design/Methods: 22 children (11 EPT, 11 TC) participated in passive stories listening during MEG (275-channel whole-head CTF system at 1200Hz) and fMRI (Philips Achieva 3.0T scanner). Multi-echo fMRI was obtained, isolating BOLD signal through independent component analysis. Modeled fMRI time-series were analyzed in SPM12. Contrast maps were generated (stories-noise) for each child and passed onto second-level analyses. No group differences were observed in fMRI, so we generated a joint activation map (Fig 1) and parcellated it into 200 random units. Centroids of parcels with >10% activation served as nodes for MEG analyses. Using linearly-constrained minimum-variance beamforming, we estimated activity at each node. Trial-wise frequency representations (2-70Hz) were obtained through Fourier analysis (+/- 4Hz smoothing). Weighted phase lag index was used to quantify coupling between nodes. We compared connectivity spectra between groups and tested for significant subnetworks using Network Based Statistics (NBS). Network strength was correlated with language scores.

Results: There were no significant group differences in age, sex, race, ethnicity, or income. There were differences (p < 0.05) in IQ (95 vs 113 for EPT and TC) and Clinical Evaluation of Language Fundamentals Core Language scores (91 vs 113 for EPT and TC). We observed a single cluster of significant differences in connectivity, with EPT exhibiting greater coupling at 18-23 Hz (100000 permutations, cluster p< 0.05, Fig 2). NBS identified an interhemispheric subnetwork wherein EPT have hyperconnectivity (Fig 3). Subnetwork strength was negatively related to language scores for EPT (r =-0.47, p >0.05; r =-0.64, p< 0.05 after removal of a single outlier using Cook’s method).Conclusion(s): In this preliminary report on a subset of our participants, functional hyperconnectivity persists in EPT from 4-6 to 8-12 years of age, but—contrary to findings at 4-6 years—correlates negatively with language. Ongoing work will determine the trajectory of this connectivity and underlying mechanisms.

Designed by Cadmium |Technical Support
© Copyright 2024 Cadmium. All Rights Reserved.