Exploring the longitudinal associations of functional network connectivity and psychiatric symptom changes in youth

Lorenza Dall’Aglio, Fernando Estévez‐López, Mònica López-Vicente, Bing Xu, Oktay Agcaoglu, Elias Boroda, Kelvin O. Lim, Vince D. Calhoun, Henning Tiemeier, Ryan L. MuetzelmedRxiv 2022 · DOI

This study, recently published as a preprint on medRxiv, investigates the complex relationship between how our brain networks communicate and the development of psychiatric symptoms during the critical transition from childhood to adolescence.

TL;DR

Using a massive longitudinal dataset of over 3,000 youth, researchers found that neither static nor dynamic brain connectivity patterns are robust predictors of psychiatric symptom changes (or vice versa). The study suggests that the strong brain-behavior links often reported in small clinical studies may not generalize to the broader population.

What Problem Does This Paper Address?

For years, neuroscientists have sought "biomarkers" for mental health in the form of functional network connectivity (FNC)—the way different brain regions synchronize their activity. However, the field has been haunted by a "replicability crisis." Most previous studies were cross-sectional (a single snapshot in time) and relied on small groups of patients already diagnosed with specific disorders.

This paper tackles three major gaps:

  1. Directionality: Does an "atypical" brain lead to symptoms, or do behavioral struggles shape the brain?
  2. Dynamics: Most studies look at "average" connectivity (static), but the brain is constantly shifting between states. Does the fluidity of these connections matter more?
  3. Generalizability: Do these patterns exist across the "continuum" of symptoms in the general population, or only in severe clinical cases?

Methods

The authors employed a sophisticated dual-pronged imaging analysis on resting-state fMRI (rs-fMRI) data:

  • Static FNC (sFNC): They used graph theory to treat the brain as a mathematical network, calculating metrics like modularity (how well-segregated sub-networks are) and global efficiency (how easily information travels across the whole brain).
  • Dynamic FNC (dFNC): Using a sliding-window approach and k-means clustering, they identified five distinct "connectivity states" the brain cycles through during a scan. They measured "Mean Dwell Time" (how long a kid stays in a specific state) and "Transitions" (how often they switch).
  • Statistical Modeling: They used Cross-Lagged Panel Models (CLPM). This allowed them to look at two time points (age 10 and age 14) and statistically "tease apart" whether baseline brain metrics predicted future symptoms, or if baseline symptoms predicted future brain changes.

Datasets

The study utilized the Generation R Study, a prospective birth cohort based in Rotterdam, Netherlands.

  • Sample Size: N = 3,131 children, making it one of the largest single-site pediatric neuroimaging studies to date.
  • Measures: Psychiatric symptoms were assessed using the Child Behavioral Checklist (CBCL), covering "broadband" issues (Internalizing/Externalizing) and eight specific syndromes (e.g., anxious-depressed, rule-breaking).

Key Insights and Ideas

  • The "Null" is the News: The most striking insight is the lack of robust findings. Despite a massive sample and advanced methods, no association survived correction for multiple statistical tests. This serves as a vital "reality check" for the field.
  • Effect Size Realism: The researchers found that median effect sizes were incredibly small (~0.009). This suggests that in the general population, the link between brain connectivity and mental health is much subtler than previously thought.
  • Behavior as a Driver? Interestingly, the few "nominal" (unadjusted) associations found were mostly in the direction of behavior predicting brain changes, rather than the other way around. For instance, rule-breaking at age 10 nominally predicted how much time a child's brain spent in a specific "default-mode" state at age 14.

Key Results

  • Static FNC: No significant longitudinal relationships were found between general network properties and internalizing or externalizing problems.
  • Dynamic FNC: While some specific syndrome scales (like attention and aggression) showed nominal links to dFNC states, none survived the False Discovery Rate (FDR) correction.
  • Sex Differences: The researchers tested if these relationships differed between boys and girls but found no evidence for sex-specific patterns.

Limitations and Future Work

The authors are transparent about several hurdles:

  • Scan Duration: The rs-fMRI scans were roughly 6 minutes long. While standard, longer scans might be necessary to capture more reliable "dynamic" brain states.
  • Two Time Points: With only two assessments, the researchers couldn't model individual "growth curves" or non-linear developmental trajectories.
  • The "Continuum" Dilution: By looking at the general population, the "signal" of psychiatric illness might be diluted compared to clinical samples where symptoms are extreme.

Future Direction: The authors call for even larger, multi-site studies (like the ABCD Study) with more frequent assessments and longer scan times to determine if these brain-behavior links are truly non-existent in youth or simply require even higher resolution to detect.

Summary updated

Abstract

ABSTRACT Background Functional connectivity has been associated with psychiatric problems, both in children and adults, but inconsistencies are present across studies. Prior research has mostly focused on small clinical samples with cross-sectional designs. Methods We adopted a longitudinal design with repeated assessments to investigate associations between functional network connectivity (FNC) and psychiatric problems in youth (9- to 17-year-olds) from the general population. The largest single-site study of pediatric neurodevelopment was used: Generation R ( N = 3,131). Psychiatric symptoms were measured with the Child Behavioral Checklist as broadband internalizing and externalizing problems, and its eight specific syndrome scales (e.g., anxious-depressed). FNC was assessed with two complementary approaches. First, static FNC (sFNC) was measured with graph theory-based metrics. Second, dynamic FNC (dFNC), where connectivity is allowed to vary over time, was summarized into 5 states that participants spent time in. Cross-lagged panel models were used to investigate the longitudinal bidirectional relationships of sFNC with internalizing and externalizing problems. Similar cross-lagged panel models were run for dFNC. Results Small longitudinal relationships between dFNC and certain syndrome scales were observed, especially for baseline syndrome scales (i.e., rule-breaking, somatic complaints, thought problems, and attention problems) predicting connectivity changes. However, no association between any of the psychiatric problems (broadband and syndrome scales) with either measure of FNC survived correction for multiple testing. Conclusion We found no or very modest evidence for longitudinal associations between psychiatric problems with dynamic and static FNC in this population-based sample. Differences in findings may stem from the population drawn, study design, developmental timing and sample sizes.
BibTeX
@misc{dallaglio2022exploring,
  author = {Lorenza Dall’Aglio and Fernando Estévez‐López and Mònica López-Vicente and Bing Xu and Oktay Agcaoglu and Elias Boroda and Kelvin O. Lim and Vince D. Calhoun and Henning Tiemeier and Ryan L. Muetzel},
  title = {{Exploring the longitudinal associations of functional network connectivity and psychiatric symptom changes in youth}},
  year = {2022},
  doi = {10.1101/2022.11.26.22282787},
  howpublished = {medRxiv},
}
Figure