Altered neural oscillation patterns and functional connectivity in postherpetic neuralgia: evidence from resting-state EEG

Postherpetic neuralgia (PHN) is a neuropathic pain condition that persists after shingles and disproportionately affects older adults, leading to substantial physical and psychological burden. Pain perception involves coordinated brain network activity and oscillations spanning delta through gamma frequencies. Alterations in these rhythms and in functional connectivity are increasingly recognized as hallmarks of chronic pain. EEG offers a noninvasive, high-temporal-resolution approach to quantify these neural dynamics. Prior work suggests heightened gamma activity in PHN and links gamma oscillations in somatosensory cortex to pain intensity. However, broader oscillatory changes (delta, theta, alpha, beta) and their relationships to mood symptoms in PHN remain underexplored, as do network-level connectivity patterns. This study aimed to characterize resting-state EEG power spectra and functional connectivity in PHN versus healthy controls and to relate oscillatory measures to pain, depression, and anxiety. The authors hypothesized increased power across multiple bands in PHN and reduced gamma-band connectivity, reflecting hyperactivity and disrupted communication in pain-related networks.

The paper reviews evidence that chronic pain disrupts brain network dynamics, including default-mode and salience networks, and that oscillatory activity across alpha and gamma bands shapes nociceptive perception, with alpha linked to inhibitory control and gamma to sensory integration. Prior EEG studies in chronic pain and PHN specifically have reported elevated gamma power, consistent with heightened cortical excitability. Gamma in primary somatosensory cortex correlates with subjective pain intensity. Research on delta, theta, alpha, and beta band changes in PHN is limited, and few studies connect EEG findings to psychiatric comorbidities such as anxiety and depression common in chronic pain. Functional connectivity studies indicate disrupted communication within pain-related networks, with decreased connectivity in descending modulatory pathways and increased connectivity in ascending pathways contributing to persistent pain. Understanding connectivity differences between PHN patients and healthy individuals may guide targeted neuromodulatory therapies.

Design: Original research with case-control comparison using resting-state EEG. Participants: 21 right-handed patients with PHN (age 43–95 years; mean ± SD 69.10 ± 11.19; 11 female) recruited from an outpatient clinic; diagnosis based on clinical history, scarring, and pain severity. Inclusion: age >18 years; persistent pain ≥1 month. Exclusion: migraine, tension headache, peripheral neuropathy, osteoarthritis, other acute/chronic pain conditions, chronic illnesses, or neurological disorders affecting EEG. Medication: gabapentin, pregabalin, or NSAIDs; opioid users excluded. Healthy controls (HC): 17 right-handed individuals (age 44–79; mean ± SD 64.47 ± 10.04; 6 female). Ethics approval and written informed consent obtained. Clinical evaluation: Short-Form McGill Pain Questionnaire (SF-MPQ) including VAS (0–10), State-Trait Anxiety Inventory (STAI: SAI and TAI), and Beck Depression Inventory (BDI). All questionnaires administered 1 hour prior to EEG. EEG recording: Eyes-closed resting-state EEG for 10 minutes using NuAmps (NeuroScan Labs). 32 Ag/AgCl electrodes (10–20 system: Fp1, Fp2, F7, F3, Fz, F4, F8, FT7, FC3, FCz, FC4, FT8, T3, C3, Cz, C4, T4, TP7, CP3, CPz, CP4, TP8, T5, P3, Pz, P4, T6, O1, Oz, O2, A1, A2). Referenced to average of mastoids; impedance <10 kΩ. Signals digitized at 1 kHz; band-pass 0.1–100 Hz. Wakefulness monitored by 1-minute interval checks. Preprocessing: EEGLAB v13.0.0b and MATLAB R2013b. Continuous EEG band-pass filtered 1–70 Hz with 50 Hz notch; segmented into nonoverlapping 2 s epochs; ocular artifacts corrected via ICA; re-referenced to average of all channels. PSD analysis: Periodogram with FFT length 2048, nonoverlapping; PSD averaged across epochs and extracted per frequency band: delta (1–4 Hz), theta (4–7 Hz), alpha (7–13 Hz), beta (13–30 Hz), gamma (30–70 Hz). Two-sample t-tests (MATLAB ttest2) across groups for all bands and channels; FDR correction (mafdr) with corrected p < 0.05. Partial correlations (partialcorr) between mean PSD (across channels) in each band and BDI, SAI, TAI scores, controlling for sex and age. Functional connectivity: Weighted phase lag index (wPLI) computed for pairs of channels that showed significant PSD differences, focusing analysis on regions with spectral changes. Network-based statistic (NBS v1.2) used to test group differences across 325 channel pairs (C^2_16): T-tests with uncorrected threshold p < 0.05 to form components; component size defined by number of suprathreshold edges; 5,000 permutations with random group reassignment; largest component sizes formed empirical null; one-sided family-wise error correction at cluster level p < 0.05.

• Power spectrum density (PSD) was elevated across all frequency bands (delta, theta, alpha, beta, gamma) in PHN patients versus healthy controls.
• Channel-wise significance:
  • Delta: PSD significantly higher across all channels in PHN (corrected p < 0.05).
  • Gamma: PSD significantly higher across all channels in PHN (corrected p < 0.05).
  • Theta: 23 channels showed higher PSD in PHN (Fp1, Fp2, F3, Fz, F4, F8, FC3, FCz, FC4, FT8, C3, Cz, C4, CP3, CPz, CP4, TP8, P3, Pz, P4, O1, Oz, O2) (corrected p < 0.05).
  • Alpha: 23 channels showed higher PSD in PHN (Fp1, Fp2, F3, Fz, F4, FC3, FCz, FC4, FT8, C3, Cz, C4, TP7, CP3, CPz, CP4, TP8, P3, Pz, P4, O1, Oz, O2) (corrected p < 0.05).
  • Beta: only F7 and FT7 showed higher PSD in PHN (corrected p < 0.05).
• Correlations between PSD and psychological measures (controlling for age and sex):
  • Delta PSD positively correlated with BDI (r = 0.51, p < 0.05).
  • Delta PSD negatively correlated with SAI (r = −0.53, p < 0.05).
  • Theta PSD positively correlated with BDI (r = 0.48, p < 0.05).
• Functional connectivity (wPLI):
  • Gamma-band wPLI was significantly higher in healthy controls than PHN patients (corrected p < 0.05), indicating reduced gamma-band connectivity in PHN.
  • Significant reductions distributed across frontal, central, parietal, and occipital regions; CP4 emerged as the most interconnected node among affected pairs.
  • Counts of significant pairs by channel included extensive involvement (e.g., CP4: 18 pairs; C4: 12; Cz: 13; TP7: 12; F3: 14; FT7: 11; P4: 11, etc.), highlighting widespread network disruption.

Findings support the hypothesis that PHN involves widespread cortical hyperactivity and disrupted network communication. Elevated delta, theta, alpha, beta, and gamma power suggests heightened excitability and altered inhibitory/sensorimotor dynamics. Slow-wave (delta/theta) increases correlated with depressive symptoms, indicating overlap between affective disturbances and pain-related oscillatory changes. Despite increased power, gamma-band functional connectivity was reduced, pointing to impaired interregional coordination within sensorimotor and pain-modulatory networks, including frontoparietal circuits implicated in top-down pain control. These results align with prior reports of chronic pain-related connectivity disruptions and highlight potential neuromodulatory targets. Interpretations are cautious due to resting-state acquisition and the correlational nature of analyses, but the convergence of spectral and connectivity findings underscores network-level dysfunction in PHN.

PHN is characterized by increased resting-state EEG power across delta, theta, alpha, beta, and gamma bands alongside decreased gamma-band functional connectivity. These alterations indicate cortical hyperexcitability and impaired network integration within pain-processing and modulatory circuits. Slow-wave power correlates with depressive symptoms, reinforcing the interplay between chronic pain and emotional distress. The identified oscillatory and connectivity signatures offer potential biomarkers and targets for neuromodulatory interventions aimed at restoring normal brain dynamics and alleviating pain in PHN.

• Functional connectivity analysis was restricted to channel pairs drawn from regions with significant PSD differences, which may limit generalizability of connectivity findings to the whole-brain network.
• Resting-state EEG and correlational analyses preclude causal inference about the relationships between neural oscillations, connectivity, and clinical symptoms.