What question did this study set out to answer?

This research aims to validate prewhitened Fiedler spectral coupling as a neural biomarker for ADHD.

May 28, 20260 citationsOpen Access

Prewhitened Fiedler Spectral Coupling as a Neural Biomarker for ADHD: Exploratory fMRI Study with ANCOVA Validation and Cross-Dataset Replication Attempt

Key Points

This research aims to validate prewhitened Fiedler spectral coupling as a neural biomarker for ADHD.
Exploratory fMRI study with ANCOVA controlling for age.
Analysis involved a primary dataset (N=268) and secondary replication (ADHD-200, n=40).
Prewhitened Fiedler coupling was utilized to assess neural network vulnerabilities.
ADHD showed significantly weaker coupling than controls (t=−2.15, p=0.033, d=0.386).
Global coupling in ADHD is strongly negative (mean=−0.094, p=8×10⁻⁴⁴).
Secondary analysis replicated the global coupling signal (p<0.001), but was underpowered for group comparisons.

Abstract

v0.2 update: Added ANCOVA results (age-controlled), Figure 1 violin plot, and ADHD-200 pediatric secondary analysis. We introduce prewhitened Fiedler spectral coupling (rp) as a topological measure that survives AR1 temporal autocorrelation deconfounding. In ds000030 (N=268; CONTROL=127, ADHD=42, SCHZ=50, BIPOLAR=49; Schaefer-200), global rp is strongly negative (mean=−0.094, p=8×10⁻⁴⁴). ANCOVA controlling for age confirms ADHD shows significantly weaker coupling than controls (t=−2.15, p=0.033, d=0.386). Prewhitening unmasks diagnostic group differences invisible in raw data (ANOVA F=2.88, p=0.036). A secondary pediatric replication (ADHD-200, n=40) replicated the global coupling signal (p<0.001) but was underpowered for group comparison (ADHD n=13). These results provide initial evidence that prewhitened Fiedler coupling captures ADHD-specific network vulnerability deconfounded from temporal autocorrelation artifacts.

问 AI

Bookmark

View Full Paper