Spectral power and functional connectivity alterations in obsessive-compulsive disorder based on resting-state EEG

Background:Obsessive-compulsive disorder (OCD) is a clinically heterogeneous psychiatric condition characterized by intrusive thoughts and repetitive behaviors, often overlapping with symptoms of other disorders, which complicates objective diagnosis. Despite advances in neuroimaging and electrophysiology, reliable and interpretable biomarkers for OCD remain limited. Recent EEG-based classification studies have demonstrated that OCD can be distinguished from healthy individuals with high accuracy, suggesting the presence of disorder-specific electrophysiological patterns. However, these approaches rely on complex computational models that provide limited neurophysiological interpretability and are not easily translatable to clinical practice. Therefore, identifying explicit EEG features underlying these differences remains essential. In this study, we aimed to investigate alterations in spectral power and functional connectivity in OCD to better characterize its electrophysiological profile using interpretable EEG measures.

Methods:Resting-state EEG recordings were obtained from OCD patients (n = 86) and healthy controls (n = 51). EEG signals were processed using standard preprocessing procedures, including artifact rejection and filtering. Absolute spectral power was computed across conventional frequency bands, including delta, theta, alpha, beta, and high-beta, using frequency-domain analysis. Functional connectivity was assessed using coherence measures between electrode pairs to capture both short- and long-range neural interactions. Analyses focused on key cortical regions, including frontal, temporal, parietal, and occipital areas. Group differences were evaluated using non-parametric Mann–Whitney U tests to account for potential deviations from normality. Both regional activity patterns and large-scale connectivity profiles were examined.

Results:OCD patients exhibited significantly increased high-beta power in posterior regions, particularly at O2 (p = 0.041) and P4 (p = 0.020), along with elevated theta activity in parietal regions (P4, p = 0.043). Additionally, temporal regions (T5) showed increased delta (p = 0.005) and theta (p = 0.035) power, indicating enhanced slow-wave activity.

Beyond these focal findings, the overall pattern suggested coexisting slow- and fast-frequency abnormalities, reflecting alterations in both cortical excitability and regulatory processes. The spatial distribution of these effects across posterior, parietal, and temporal regions indicates that electrophysiological changes in OCD are not confined to a single cortical area but involve multiple interconnected regions.

Coherence analysis revealed widespread alterations in functional connectivity, particularly involving fronto-parietal and fronto-temporal networks across delta, theta, and beta bands. Notably, interhemispheric parietal connectivity (P3–P4)demonstrated a strong difference in the beta band (p < .001), alongside multiple significant long-range connections (p < 0.05), suggesting disrupted large-scale network integration.

Conclusions:The findings indicate that OCD is associated with distinct alterations in both EEG spectral power and functional connectivity, reflecting abnormalities at regional and network levels. The observed combination of increased slow-wave activity in temporal regions and enhanced high-frequency activity in posterior areas, together with disrupted long-range connectivity, suggests a widespread electrophysiological dysregulation rather than a focal deficit. These results support the presence of identifiable EEG patterns underlying OCD and highlight the potential of spectral and coherence analyses as interpretable approaches for characterizing disorder-specific neural signatures beyond black-box models.

References

No references