Logo image
Back
Brain functional connectivity predicts depression and anxiety during childhood and adolescence: A connectome-based predictive modeling approach
Journal article   Open access   Peer reviewed

Brain functional connectivity predicts depression and anxiety during childhood and adolescence: A connectome-based predictive modeling approach

Francesca Morfini, Aaron Kucyi, Jiahe Zhang, Clemens C.C. Bauer, Paul A. Bloom, David Pagliaccio, Nicholas A. Hubbard, Isabelle M. Rosso, Anastasia Yendiki, Satrajit S. Ghosh, …
Imaging neuroscience (Cambridge, Mass.), v 3, IMAG.a.145
12 Sep 2025
DOI: https://doi.org/10.1162/IMAG.a.145
PMID: 40959708
Featured in Collection :   UN Sustainable Development Goals @ Drexel
url
https://doi.org/10.1162/IMAG.a.145View
Published, Version of Record (VoR) Open

Abstract

adolescence anxiety depression functional connectivity functional magnetic resonance imaging longitudinal studies Machine Learning
Identifying brain-based correlates of risk for future depression and anxiety severity in youth could improve prevention and treatment efforts. We tested whether connectome-based predictive modeling (CPM) based on resting-state functional connectivity (FC) at baseline: (a) predicts future depression and anxiety severity during childhood and (b) generalizes to adolescence. We used two independent, longitudinal datasets including children from the Adolescent Brain Cognitive Development (ABCD) study and adolescents from the Boston Adolescent Neuroimaging of Depression and Anxiety (BANDA). ABCD included a cohort of 11,875 children ages 9–11 years old, and BANDA enrolled 215 adolescents ages 14–17 years, of which ~70% reported a depressive or anxiety disorder. CPM with internal (within ABCD) and external validation (from ABCD to BANDA) used baseline whole-brain FC to predict depression and anxiety severity at a 1-year follow-up assessment. ABCD-derived functional connections, which we term “Symptoms Network”, were validated within BANDA to test model applicability in adolescence, which is a peak period for the emergence of internalizing disorders. Participants with complete data were included from ABCD (n = 3,718, 52.9% girls, ages 10.0 ± 0.6) and BANDA (n = 150, 61.3% girls, ages 15.4 ± 0.9). In ABCD, we found that FC predicted 1-year follow-up symptoms severity (ρ = 0.058, p = 0.040), measured with the Child Behavior Checklist Anxious/Depressed subscale. External validation in BANDA indicated that the Symptoms Network predicted 1-year follow-up symptoms severity (ρ = 0.222, p = 0.007), measured with the Revised Child Depression and Anxiety Scale t-transformed total score. In both ABCD and BANDA, FC enhanced the prediction of future symptom severity beyond baseline clinical and demographic information (baseline severity, sex, and age), including when correcting for mean head motion. The ABCD-derived connections included contributions from somatomotor, attentional, and subcortical regions and were characterized by heterogeneous FC within adolescents, where the same region pairs were characterized by positive FC for some participants but by negative FC for others. In conclusion, FC may provide inroads for early identification of internalizing symptoms, which could inform preventative-intervention approaches prior to the emergence of affective disorders during a critical period of neuromaturation. However, the small effect sizes and heterogeneity in results underscore the challenges of employing brain-based biomarkers for clinical applications and emphasize the need for individualized approaches for understanding neurodevelopment and mental health.

Metrics

4 Record Views
2 citations in Web of Science

Details

UN Sustainable Development Goals (SDGs)

This publication has contributed to the advancement of the following goals:

#3 Good Health and Well-Being

InCites Highlights

Data related to this publication, from InCites Benchmarking & Analytics tool:

Collaboration types
Domestic collaboration
Web of Science research areas
Neuroimaging
Neurosciences
Logo image