All participants completed the Short-Form McGill Pain Questionnaire (SF-MPQ) (20). It includes a verbal component and a VAS ranging from 0 (indicating no pain) to 10 (representing the maximum imaginable pain). Anxiety and depression levels of the participants were assessed using the STAI (19) and BDI (21), respectively.

The STAI comprises the State Anxiety Inventory (SAI) and Trait Anxiety Inventory (TAI) subscales. The SAI, consisting of 20 items, measures state anxiety in response to situational stress, whereas the TAI, which also contains 20 items, evaluates the frequency of general emotional experiences. The BDI, designed to assess depressive symptoms, includes 13 items that assess how participants felt during the past week. The responses are scored on a scale from 0 (“not at all”) to 3 (“severely”), with higher cumulative scores indicating more severe depressive symptoms. All questionnaires were administered 1 h prior to the EEG recording session.

The participants were seated in comfortable chairs in a silent, temperature-controlled room and instructed to relax, with their eyes closed. Resting-state EEG data were recorded for 10 min using NuAmps device (NeuroScan Labs, Australia). Thiry-two Ag/AgCl scalp electrodes were positioned in accordance with the international 10–20 system at 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, and A2. The EEG signals were referenced to the average of the mastoids, and the electrode impedance was maintained below 10 kΩ during the recording process.

EEG data were amplified, digitized at 1 kHz sampling rate, and band-pass filtered (0.1–100 Hz). To ensure that the participants remained awake during data collection, they were prompted to report their conscious state at 1 min intervals.

EEG data were preprocessed using the EEGLAB toolbox (v.13.0.0b) in conjunction with customized scripts in MATLAB 2013b (22). Continuous EEG signals were band-pass in the range of 1 to 70 Hz and a 50 Hz notch filter was applied to eliminate line noise. Epochs were segmented into nonoverlapping 2 s intervals. Electrooculogram artifacts were corrected using independent component analysis decomposition. Subsequently, the data were re-referenced to the average of all the available channels.

The PSD values were computed for each channel and epoch using the MATLAB function “periodogram” and customized scripts, with a nonoverlapping fast Fourier transform length of 2,048. The PSD values were averaged across epochs for each participant. Subsequently, the PSD values were extracted and averaged within predefined frequency bands. These bands were classified as delta (1–4 Hz), theta (4–7 Hz), alpha (7–13 Hz), beta (13–30 Hz), and gamma (30–70 Hz).

A two-sample t-test was conducted to compare patients with PHN and HCs across all frequency bands and channels. This analysis was performed using the MATLAB function “ttest2.” To account for multiple comparisons, false discovery rate correction was applied at a significance threshold of corrected p < 0.05, utilizing the MATLAB function “mafdr.”

Partial correlation analyses were performed to investigate the relationship between the mean PSD value across all channels in each frequency band and the BDI, SAI, and TAI scores. Sex and age were included as covariates to control for potential confounding effects. Partial correlations were calculated using the MATLAB function “partialcorr.”

Functional connectivity strength was assessed using the weighted phase lag index (wPLI), a measure designed to mitigate the effects of volume conduction (23). The wPLI was computed for each pair of channels that exhibited significant differences in the PSD values between the HCs and patients with PHN. This selection criterion was used to narrow the connectivity analysis to the brain regions that showed significant spectral changes, assuming that these areas were more likely to reflect the pathophysiological changes associated with PHN. Although this approach improves sensitivity to PHN-related effects, it may limit the generalizability of connectivity findings to the entire brain network. A customized MATLAB script was employed to calculate the wPLI using the following equation:Φ≡|E{ℑ{X}}|E{|ℑ{X}|}To investigate whether the wPLI values differed significantly between HCs and patients with PHN across frequency bands, the network-based statistic (NBS) method was applied. NBS is a nonparametric approach designed to address the multiple comparison problem arising from mass univariate statistical tests in functional networks (24). The analysis performed is described below.

First, the T-test was performed to compare the wPLI values between the two groups for all 325 possible channel pairs (C₂₆^₂). In the NBS framework, a general linear model was implemented (25). Second, an uncorrected threshold of p < 0.05 was applied to identify channel pairs with potential differences in the wPLI values, forming connected graph components. Third, the size of each connected graph component was defined based on the total number of significant channel pairs. A permutation test (n = 5,000) was then conducted, during which participants were randomly reassigned to one of the two groups for each permutation. The maximum cluster size from each permutation was used to construct an empirical null distribution for the largest component size. Finally, to control for multiple comparisons, a one-sided family-wise error corrected p value (p < 0.05) was applied at the cluster level. All statistical tests were performed using the MATLAB toolbox NBS v1.2 (https://www.nitrc.org/projects/nbs/).

Spontaneous EEG PSD within the frequency range of 1–70 Hz was computed for patients with PHN and HCs. Figure 1 indicates the averaged PSD value across all channels and participants for each group. As shown in Figure 1, the PSD values were consistently elevated across all frequency bands in patients with PHN compared with HCs.

Figure 2 illustrates the PSD distribution on maps across each frequency band for both groups (the left and middle columns), highlighting their differences (the right column). Specifically, in the delta and gamma frequency bands, the PSD values across all channels were significantly higher in patients with PHN than in HCs (corrected p < 0.05). For the theta frequency band, the PSD values of 23 channels (Fp1, Fp2, F3, Fz, F4, F8, FC3, FCz, FC4, FT8, C3, Cz, C4, CP3, CPz, CP4, TP8, P3, Pz, P4, O1, Oz, and O2) were significantly higher in patients with PHN compared with HCs (corrected p < 0.05). Similarly, in the alpha frequency band, 23 channels (Fp1, Fp2, F3, Fz, F4, FC3, FCz, FC4, FT8, C3, Cz, C4, TP7, CP3, CPz, CP4, TP8, P3, Pz, P4, O1, Oz, and O2) exhibited significantly higher power in patients with PHN than in HCs (corrected p < 0.05). In contrast, for the beta frequency band, only two channels (F7 and FT7) showed significantly higher power in patients with PHN than in HCs (corrected p < 0.05).

The pain ratings and psychological assessment scores are presented in Table 1. Significant correlations were observed between the PSD values and psychological assessment scores; the PSD value in the delta frequency band was positively correlated with the BDI score (r = 0.51, p < 0.05) and negatively correlated with the SAI score (r = −0.53, p < 0.05). In addition, the PSD value in the theta frequency band showed a significant positive correlation with the BDI score (r = 0.48, p < 0.05).

The wPLI values in the gamma frequency band were significantly higher in HCs compared with patients with PHN (corrected p < 0.05). Figure 3 illustrates the averaged wPLI for each channel pair across all participants within each group (the left and middle columns), indicating the channel pairs exhibiting significant differences (the right column), thereby emphasizing the degree of functional disconnection in patients with PHN within the gamma frequency band.

Specifically, the significant channel pairs associated with individual channels were as follows: Fp1, 9 pairs; Fp2, 5 pairs; F7, 11 pairs; F3, 14 pairs; Fz, 2 pairs; F4, 5 pairs; F8, 2 pairs; FT7, 11 pairs; FC3, 5 pairs; FCz, 5 pairs; FC4, 8 pairs; FT8, 9 pairs; C3, 5 pairs; Cz, 13 pairs; C4, 12 pairs; TP7, 12 pairs; CP3, 6 pairs; CPz, 11 pairs; CP4, 18 pairs; TP8, 10 pairs; P3, 3 pairs; Pz, 6 pairs; P4, 11 pairs; O1, 6 pairs; Oz, 6 pairs; and O2, 9 pairs.

The channel pairs exhibiting reduced connectivity were distributed across the frontal, central, parietal, and occipital regions, highlighting widespread network disruption. CP4 emerged as the most interconnected node.

One of the primary findings of this study is the elevated power observed across all the EEG frequency bands (delta, theta, alpha, beta, and gamma) in patients with PHN compared with HCs. This widespread increase in power indicates hyperactivity in the neural circuits involved in pain perception and processing. Similar results have been previously reported, reinforcing the notion that chronic pain is associated with heightened cortical excitability (10, 26).

The increase in the delta and theta frequency band power is particularly noteworthy. Delta oscillations (1–4 Hz) are typically associated with sleep, drowsiness, and low-arousal states, whereas theta oscillations (4–7 Hz) are associated with memory, emotional processing, and pain modulation (5). The elevated delta and theta frequency band power in patients with PHN suggest that these slower rhythms play a role in the sensory and affective dimensions of chronic pain. Moreover, the positive correlation between the delta and theta frequency band power and BDI scores indicates that these oscillatory changes may reflect underlying depressive symptoms, which are common comorbidities in patients with PHN.

The increased alpha and beta frequency band power further highlight the altered state of neural inhibition and sensorimotor processing in patients with PHN. Alpha rhythms (7–13 Hz) are often associated with inhibitory control and cortical idling, whereas beta rhythms (13–30 Hz) are associated with motor function and active cognitive states (27). Enhancement of these rhythms may reflect maladaptive neural mechanisms aimed at compensating for persistent pain signals, potentially contributing to sensory hypersensitivity and dysfunctional pain processing.

Gamma oscillations (30–70 Hz), which are associated with higher-order cognitive and sensory integration, were also elevated in patients with PHN. This finding aligns with previous studies that demonstrated increased gamma activity in various chronic pain conditions (10). Elevated gamma frequency band power is thought to reflect enhanced cortical excitability and increased sensory input, further supporting the notion that patients with PHN experience heightened nociceptive processing.

In contrast to the widespread increase in EEG power, functional connectivity analysis revealed significant reductions in gamma-band connectivity among patients with PHN. The decreased wPLI values in the gamma band indicate impaired communication between the brain regions involved in pain processing and modulation. These findings are consistent with those of a previous study that demonstrated disrupted connectivity in populations with chronic pain (15).

The reduced connectivity, particularly in the CP4, Cz, and C4 channels, demonstrates their importance in sensory integration and pain processing. The CP4 region, located above the parietal cortex, is critical for sensory integration, whereas the Cz and C4 regions correspond to the central motor and somatosensory regions. The decreased connectivity between the frontal and parietal regions suggests that top-down pain signal modulation is impaired, indicating the dysfunction of sensorimotor and pain regulatory networks. Such disruptions may contribute to altered pain perception and weakened descending inhibitory control, potentially exacerbating chronic pain.

Furthermore, the significant decrease in connectivity between the frontal and parietal regions may indicate changes in the networks involved in cognitive control and pain regulation. The frontoparietal network has been linked to the top-down modulation of pain (28), and decreased connectivity in this circuit may be associated with decreased regulatory capacity. However, given that the EEG data was collected at resting state and the correlational nature of the analysis, such functional interpretations should be considered preliminary. Nonetheless, these findings indicate widespread disruptions in neural oscillatory dynamics and connectivity in PHN, suggesting potential network-level targets for future neuromodulation interventions and personalized therapies.

A critical aspect of this study is the observed relationship between neural oscillations and the psychological measures, including those assessing depression and anxiety. The positive correlation between the delta and theta frequency band power and BDI scores suggests that slow-wave activity may serve as a neural marker of depressive symptoms in patients with PHN. This finding aligns with the broader literature suggesting shared overlapping neural mechanisms for chronic pain and depression (29).

Conversely, the negative correlation between the delta frequency band power and SAI scores indicate a potential inverse relationship between low-frequency oscillations and acute anxiety. These results reflect the distinct neural pathways underlying anxiety and depression, with anxiety being more closely associated with hyperactivity in high-frequency bands (30). By elucidating these relationships, this study highlights the need for integrated approaches to PHN management that address both the physical and psychological components of pain.

The observed relationships between the EEG PSD results and psychological assessment scores emphasize the interaction between neural activity and psychological states in PHN. Increased delta and theta frequency band power correlated with higher BDI scores, implying that low-frequency oscillations may reflect the affective dysregulation common in chronic pain. Meanwhile, the decreased gamma-band connectivity may indicate impaired sensory integration and top-down modulation, potentially adding to the physical and emotional burden of pain. These findings support a multidimensional framework for PHN, in which neural and psychological measures reflect the central and affective components of the condition. Notably, a recent review emphasized that resting-state gamma-band oscillations and connectivity disruptions are not only indicators of sensory processing but also correlate with individual pain sensitivity and emotional burden in chronic pain (31). This supports our finding that PHN-related EEG changes, particularly in the delta, theta and gamma frequency bands, may reflect a combined neurophysiological profile that includes affective symptoms and central sensitization.

The findings of this study have several important clinical implications. First, the widespread increase in EEG power suggests that targeting hyperactive neural circuits may offer a promising avenue for neuromodulatory treatments, such as transcranial magnetic stimulation or neurofeedback. By selectively modulating the delta, theta, and gamma oscillations, these interventions can alleviate pain and improve psychological outcomes in patients with PHN. Second, the observed deficits in functional connectivity highlight the importance of restoring network-level coherence during PHN treatment. Techniques such as neurostimulation, cognitive behavioral therapy, and mindfulness-based interventions may enhance connectivity within pain-regulatory networks, promoting more effective pain modulation and reducing symptom severity.

Future research should aim to delineate the causal relationships between neural oscillations, connectivity patterns, and clinical outcomes in patients with PHN. Longitudinal studies tracking changes in EEG activity and psychological measures over time can provide valuable insights into the dynamic nature of chronic pain and its neural correlates. Additionally, expanding the sample size and incorporating multimodal imaging techniques, such as functional MRI, may enhance the generalizability and depth of these findings.

In conclusion, this study provides compelling evidence for widespread neural alterations in patients with PHN, characterized by increased EEG power and decreased functional connectivity. These findings not only advance our understanding of the neural mechanisms underlying PHN but also offer potential targets for future interventions aimed at alleviating pain and improving the quality of life in affected individuals.