Abstract
Background
Cluster headache (CH) is a primary headache disorder characterized by circadian rhythmicity and autonomic symptoms. While the posterior hypothalamus has been historically implicated, converging evidence now supports a distributed, network-level dysfunction involving cortical and subcortical regions. This study aimed to: (1) assess whole-brain cortical thickness alterations in episodic CH (eCH) during the bout period, (2) examine their associations with clinical burden, and (3) investigate resting-state functional connectivity patterns of cortical regions showing structural alterations.
Methods
We investigated whole-brain cortical thickness and subcortical volumes in 26 patients with episodic CH scanned during the bout period but outside of attacks and 20 matched healthy controls (HC) using surface-based morphometry (FreeSurfer). Associations between cortical thickness and clinical variables (attack frequency, duration, disease history, and pain intensity) were assessed using general linear models corrected for multiple comparisons. Resting-state functional connectivity (FC) was further analyzed using seed-to-voxel correlations in the CONN toolbox, with seeds placed in cortical regions showing significant thinning.
Results
Compared with HC, eCH patients showed cortical thinning in the right inferior frontal gyrus (pars triangularis, BA45) and left superior frontal gyrus (BA9/10) (p < 0.05, Bonferroni correction). No significant subcortical volumetric differences were observed. Cortical thickness showed significant positive associations with clinical burden in associative parietal and occipital regions (p < 0.01, Bonferroni correction). Seed-based FC analyses revealed decreased coupling between the left superior frontal gyrus and right supramarginal cortex, but increased connectivity with the left precentral gyrus. Conversely, the right inferior frontal gyrus showed widespread hyperconnectivity with bilateral insulae, supramarginal, and sensorimotor cortices, and decreased connectivity with the right superior frontal region.
Conclusions
These findings identify coexisting structural and functional alterations within frontal cortices and their distributed targets, suggesting a differential reorganization of frontal attentional and sensorimotor networks in eCH. The pattern of fronto-parietal disconnection and fronto-insular hyperconnectivity extends the network model of cluster headache beyond the well-established hypothalamic–thalamic dysregulation. Together, these results strengthen the concept of cluster headache as a network-level disorder involving both executive and sensorimotor control systems.
Keywords: Cluster headache, Cortical thickness, Functional connectivity, Inferior frontal gyrus, Superior frontal gyrus, Pain networks, Surface-based morphometry, CONN toolbox
Introduction
Cluster headache (CH) is the most common trigeminal autonomic cephalalgia, a primary headache disorder characterized by recurrent attacks of excruciating unilateral pain associated with ipsilateral cranial autonomic symptoms and circadian rhythmicity [1, 2]. Although the posterior hypothalamus was initially identified as the main hub in CH pathophysiology based on early positron emission tomography studies [3], increasing evidence indicates that CH is a network-level disorder involving the trigeminovascular system, hypothalamus, thalamus, and higher-order cortical and subcortical areas [1, 4–8].
Advanced neuroimaging techniques have revealed widespread structural and functional brain alterations in CH patients. Resting-state fMRI and diffusion tensor imaging (DTI) studies have demonstrated abnormal connectivity between the hypothalamus, the salience network, and executive control networks, as well as microstructural alterations in the hypothalamus and thalamus, even during headache-free periods [9, 10]. Morphometric studies using voxel-based morphometry (VBM) or surface-based morphometry (SBM) have reported inconsistent findings, with some authors describing hypothalamic hypertrophy or cortical thinning in frontal, temporal, and parietal cortices [11–13], while others did not detect macrostructural differences [14, 15]. Methodological differences, including patient selection (episodic vs. chronic CH), timing of scanning (in- vs. out-of-bout), and prophylactic medication use, may explain these discrepancies.
Our group recently demonstrated that hypothalamic macrostructural volumes appear preserved in episodic CH (eCH) patients scanned during the in-bout period [14], while the same cohort exhibited hypothalamic and thalamic microstructural abnormalities and dysfunctional connectivity between cortical networks [9]. These findings point to distributed, rather than focal, pathology and raise the possibility that cortical regions involved in pain modulation, attention, and salience processing may also undergo structural and functional changes.
We hypothesized that cortical alterations in eCH would primarily affect multimodal associative regions and correlate with clinical burden, and that these structural abnormalities would be accompanied by altered functional connectivity within attention, salience, and sensorimotor networks—extending previous evidence of network-level dysfunction in cluster headache.
Accordingly, the study aimed to (1) assess cortical thickness alterations at the whole-brain level, (2) explore their associations with clinical variables, and (3) characterize resting-state functional connectivity patterns of the affected cortical regions. In addition, subcortical volumes were analyzed to evaluate possible deep gray matter involvement.
Methods
This study included 26 patients with episodic cluster headache (eCH) (ICHD-3 code 3.1.1) and 20 age- and sex-matched healthy controls (HC). These patients were the same cohort described in our previous work [14]. MRI scans were performed during the bout period but outside of headache attacks. We acquired information on diverse clinical features of patients during either the screening visit or the scanning session, encompassing daily attack frequency, average severity of headache attacks, duration of the attacks and duration of eCH history. Exclusion criteria were the presence of other primary or secondary headache disorders, systemic, neurological, neuro-ophthalmological, or psychiatric diseases, and a family history of migraine in first-degree relatives. Patients with eCH and a familial history of migraine (first-degree relatives) were likewise excluded. Preventive treatments were not permitted within three months prior to imaging. Acute treatments for individual cluster headache attacks (e.g., oxygen or triptans) were allowed when clinically required; however, all patients were scanned outside of attacks and therefore did not require acute treatment at the time of MRI acquisition. Transitional therapies were not permitted prior to MRI acquisition. Twenty healthy controls (HCs) with similar age and sex distributions were recruited from medical students and healthcare professionals for comparison. They exhibited no apparent medical illnesses, personal or familial history of primary headaches or epilepsy, nor habitual drug consumption. All participants in the study were regular coffee drinkers; none were alcohol abusers; 30% of healthy controls and 35% of patients were smokers. All participants in the study were asked not to consume any caffeinated or alcoholic beverages on the day of the examination.
All participants were provided with a comprehensive description of the study and provided written informed consent. The Faculty of Medicine’s ethical review board at the University of Rome, Italy, approved the experiment (N° 0295/2023).
MRI protocol
MRI scans were acquired on a 3T scanner using a T1-weighted sagittal magnetization-prepared rapid gradient echo (MPRAGE) sequence (TR = 1900 ms, TE = 2.93 ms, 176 sagittal slices, voxel size = 0.508 × 0.508 × 1 mm³). Additional proton density and T2-weighted images were obtained with an interleaved double-echo turbo spin echo sequence (TR = 3320 ms, TE = 10/103 ms, matrix = 384 × 384, FOV = 220 mm, slice thickness = 4 mm, gap = 1.2 mm, 50 axial slices).
Functional resting-state fMRI data were obtained using a T2*-weighted echo-planar imaging (EPI) sequence (TR = 3000 ms; TE = 30 ms; 40 axial slices; voxel size = 3.906 × 3.906 × 3 mm; 150 volumes). Resting-state scans lasted for 7 min and 30 s. During these sessions, participants were instructed to relax, remain still, and keep their eyes closed but not to fall asleep. Upon completion of the scanning, all participants confirmed that they had remained awake throughout the resting-state fMRI procedure.
Structural MRI analysis
Cortical thickness was selected as the primary cortical morphometric measure, as it provides higher sensitivity and lower susceptibility to false-positive findings compared with cortical volume; subcortical volumes were analyzed because thickness cannot be meaningfully estimated for subcortical structures. For completeness, cortical volume values were additionally extracted for cortical regions showing significant thickness differences and analyzed separately (see Results). Cortical reconstruction and volumetric segmentation were performed using FreeSurfer (version 7.4.1) following standard protocols [16–28]. Cortical parcellation and subcortical segmentation were based on the Desikan-Killiany atlas [29]. Default parameter values were used to process T1-weighted images, and a 10-mm FWHM kernel was applied to smooth participants’ resampled images. Quality control was performed by an experienced neuroradiologist (F.C.) who corrected segmentation errors.
Cortical thickness differences between eCH and HC were assessed using a two-sample t-test, adjusting for age and sex, with correction for multiple comparisons using a Bonferroni-adjusted threshold of p < 0.05. Associations between cortical thickness and clinical variables (number of attacks, duration of attacks, disease history, and pain intensity [VAS]) were evaluated with general linear models (GLM) controlling for age and sex, with a Bonferroni-corrected significance threshold of p < 0.01 to account for multiple testing across clinical variables. Volumes of subcortical ROIs (cerebellar cortex, thalamus, caudate, putamen, pallidum, brainstem, and hippocampus) were analyzed with GLMs including age, estimated total intracranial volume (eTIV) as covariates and gender as factor. A false discovery rate (FDR)–corrected threshold (Benjamini–Hochberg, p < 0.05) was applied for subcortical group comparisons and correlations with clinical variables. All cortical thickness statistical analyses were conducted in FreeSurfer, and subcortical ROI statistical analyses were performed using SPSS (version 21.0, IBM corp).
Functional connectivity analysis
Seed regions were defined a priori based on cortical areas showing significant group differences in the whole-brain cortical thickness analysis, in order to investigate the functional connectivity of structurally altered regions. All images were processed using SPM 12 (www.fil.ion.ucl.ac.uk) and CONN 23.a (www.nitrc.org) in MATLAB environment (www.mathworks.com). Pre-processing included the following steps described below based on SPM 12 algorithms. All images from a single participant were realigned using a 6-parameter rigid body process, resliced by a cubic spline interpolation. The structural (T1 – MPRAGE) and functional data were co-registered for each subject data set; data were transformed into a common stereotactic space and resampled isotropically at 3 mm x 3 mm x 3 mm. Then, the spatially normalized functional images were smoothed by 8 mm on each direction. The processing steps listed below has been performed with CONN, using the procedure reported in Whitfield-Gabrieli and Nieto-Castanon [30, 31]. First, this toolbox segmented each participants’ structural dataset in grey matter, white matter, and CSF.
The artifact detection tools (ART) were applied for each participants’ preprocessed images.
The preprocess step removed sources of possible confounds: BOLD signals from white matter and CSF, the realignment parameters (6 rigid body head’s motions) as within subject covariate and the rest effects condition convolved with hemodynamic response function; the band pass filter values were 0.008 and 0.09 Hz. Third step analyzed the functional connectivity of different ROIs in each single subject, based on bivariate correlation method. The HRF has been chosen to weight the scans within each condition. The outcome is first level results: seed to voxel connectivity maps, for each source, for each subject and for rest condition. Last step defined the second level random effect analysis, specifying the contrasts between controls and patients and vice-versa.
A ROI-to-ROI analysis was performed with the right inferior frontal gyrus, pars triangularis (IFGtri_r) and the left superior frontal gyrus (SFG_l) as seed and all other ROIs embedded as target ROIs. Correlation maps were computed for each participant by extracting the mean BOLD signal from each seed and computing Pearson’s correlations with all other voxels in the brain. Individual correlation maps were Fisher-z transformed and entered second-level random-effects analyses comparing eCH and HC. Cluster-level significance was set at p-FDR < 0.025 (two-tailed) to compensate for the number of seeds (Bonferroni’s correction method). Functional localization of significant target regions was performed by referencing the Harvard–Oxford cortical and subcortical atlases, the AAL cerebellar atlas, and the Yeo 7-network parcellation to assign each region to the corresponding large-scale network (default mode, salience, dorsal attention, sensorimotor, or executive control).
Finally, additional analyses were conducted within the eCH group to assess associations between ROI-to-ROI functional connectivity measures and clinical variables (attack frequency, disease duration, disease history, and pain intensity [VAS]). These analyses were performed using general linear models controlling for age as a covariate and sex as a factor. Statistical significance was assessed using an FDR-corrected threshold of p < 0.01, with an additional Bonferroni correction applied for the number of clinical variables (p = 0.05/4 = 0.0125).
Results
Demographic and clinical data of patients and controls are summarized in Table 1.
Table 1.
Demographic and clinical characteristics of patients with episodic cluster headache (eCH) and healthy controls (HC). Data are presented as mean ± standard deviation (SD) or absolute number. Groups did not differ in age or sex distribution. p-values refer to independent-samples t-tests (age) or fisher’s exact tests (sex)
| Variable | eCH (n = 26) | HC (n = 20) | p-value |
|---|---|---|---|
| Age (years, mean ± SD) | 40.3 ± 10.3 | 40.2 ± 9.2 | 0.97¹ |
| Sex (M/F) | 24 / 2 | 19 / 1 | 0.81² |
| Pain laterality (Right/Left) | 15 / 11 | – | – |
| Daily attack frequency | 2.7 ± 2.0 | – | – |
| Mean attack severity (VAS 0–10) | 9.7 ± 0.7 | – | – |
| Attack duration (min) | 85.0 ± 55.7 | – | – |
| Duration of eCH history (years) | 14.3 ± 11.0 | ||
| Prophylactic medications | None in the 3 months prior to scanning | – | – |
¹ Independent-samples t-test; ² Fisher’s exact test
There were no significant differences in age (p = 0.97) or sex distribution (p = 0.81) between the groups.
Whole-brain cortical thickness analysis revealed two significant clusters of cortical thinning in eCH patients compared with HC (Bonferroni-corrected p < 0.05): one in the right pars triangularis (inferior frontal gyrus, BA45) and one in the left superior frontal gyrus (Table 2; Fig. 1).
Table 2.
Cortical thickness differences between patients with episodic cluster headache (eCH) and healthy controls (HC). Vertex-wise surface-based morphometry revealed two clusters of significant cortical thinning in eCH: the right Pars triangularis (BA45) and the left superior frontal gyrus (BA9/10). Results are adjusted for age and sex and corrected for multiple comparisons (Bonferroni-corrected p < 0.05)
| clusters | Max (-log10 p) |
Vertex max |
Size (mm2) |
X (MNI) |
Y (MNI) |
Z (MNI) |
Num Vertex |
Area |
|---|---|---|---|---|---|---|---|---|
| 1 | 3.1554 | 27,020 | 9.86 | 53.4 | 23.0 | 9.0 | 16 | Right Pars triangularis (Inferior Frontal Gyrus, BA45) |
| 2 | 3.2097 | 32,330 | 10.69 | -16.3 | 45.9 | 37.8 | 18 | Left Superior Frontal Gyrus (BA9/10) |
Fig. 1.
Cortical thickness alterations in episodic cluster headache (eCH). Vertex-wise surface-based morphometry revealed two significant clusters of cortical thinning in eCH patients compared with healthy controls: (A) the right pars triangularis of the inferior frontal gyrus (BA45), and (B) the left superior frontal gyrus (BA9/10). Maps are displayed on an inflated template surface (lateral and frontal views). All results are corrected for multiple comparisons using a Bonferroni-adjusted threshold (p < 0.05), with age and sex included as covariates
As a complementary analysis, cortical volume values were extracted for the regions showing significant vertex-wise cortical thinning; no significant group differences were observed after controlling for age and total estimated intracranial volume (all p > 0.05).
Significant positive associations between cortical thickness and clinical variables were observed across temporal, parietal, and occipital associative cortices (Bonferroni-corrected p < 0.01; Table 3).
Table 3.
Associations between cortical thickness and clinical variables in eCH patients. Positive correlations (Bonferroni-corrected < 0.01) were found between cortical thickness and attack frequency, attack duration, disease duration, and pain intensity across temporal, parietal, and occipital associative cortices. Coordinates refer to cluster peaks; all analyses controlled for age and sex
| Clinical variables | clusters | Max (-log10 p) |
Vertex max |
Size (mm2) |
X (MNI) |
Y (MNI) |
Z (MNI) |
Num Vertex |
Area |
|---|---|---|---|---|---|---|---|---|---|
| Daily attack frequency | 1 | 4.0394 | 75,452 | 70.49 | -52.8 | -32.7 | -22.7 | 100 | Inferior temporal |
| 1 | 4.0550 | 31,542 | 41.25 | 53.8 | -44.4 | 38.8 | 81 | Supra-marginal | |
|
Duration of attacks |
1 | 3.2864 | 140,302 | 16.18 | -10.8 | -37.4 | 75.1 | 48 | Post-central |
| 1 | 3.4632 | 12,269 | 16.88 | 14.9 | -50.7 | 62.7 | 32 | Superior parietal | |
| VAS | 1 | 4.0095 | 50,161 | 105.13 | -28.3 | -56.3 | 56.1 | 256 | Superior parietal |
| 1 | 4.5301 | 87,751 | 109.87 | 45.6 | -79.1 | 6.9 | 160 | Lateral occipital | |
| Duration of eCH history | 1 | 3.7884 | 158,402 | 34.15 | 43.2 | -63.7 | 24.6 | 60 | Inferior parietal |
A higher number of daily attacks correlated with greater thickness in the left inferior temporal and in the right supramarginal cortices. Longer duration of attacks was associated with greater thickness in the left postcentral and right superior parietal cortices. Longer disease duration correlated with increased thickness in the right inferior parietal lobule, whereas higher pain intensity (VAS scores) was associated with greater thickness in the left superior parietal and right lateral occipital cortices.
No significant differences were detected in the volumes of the subcortical ROIs (cerebellar cortex, thalamus, caudate, putamen, pallidum, brainstem, and hippocampus) between eCH and HC after correction for multiple comparisons (FDR-corrected p < 0.05). Similarly, no significant correlations were found between subcortical volumes and clinical variables.
To account for headache laterality, no significant differences in cortical thickness or subcortical volumes were observed between patients with right-sided pain, left-sided pain, and controls, and correlation analyses restricted to each subgroup did not reveal any additional significant findings.
Seed-based functional connectivity (FC) analyses were performed using the two cortical regions showing significant cortical thinning—left superior frontal gyrus (SFG_l; ECN/frontoparietal) and right inferior frontal gyrus, pars triangularis (IFGtri_r; VAN/SN at the border with ECN)—as seeds. Patients showed, from the SFG_l seed, reduced FC with the right posterior and anterior supramarginal gyri and increased FC with the left precentral cortex (Table 4; Fig. 2).
Table 4.
Functional connectivity alterations from the left superior frontal gyrus (SFG_l) seed in episodic cluster headache (eCH) patients. Seed-based resting-state connectivity (eCH vs. HC) revealed reduced FC from SFG_l to the right supramarginal cortex and increased FC to the left precentral gyrus. All results are corrected for multiple comparisons (analysis p-FDR < 0.025). Arrows indicate direction of change in patients (↑ increased; ↓ decreased). Sign of T: Positive T = Patients < Controls; Negative T = Patients > Controls. Network attribution follows CONN and Beckmann et al. functional network classification [59]
| Seed / Target Region | Statistic | p_unc | p_FDR | Network | Direction |
|---|---|---|---|---|---|
| Seed Left Superior Frontal Gyrus (SFG l) |
F(9,35) = 5.16 Intensity = 12.07 Size = 3 |
0.0004 | 0.0002 | Default mode network, cognitive control/execution networks, motor control network | — |
| Right Posterior Supramarginal Gyrus (pSMG r) | T(43) = 4.35 | 0.0001 | 0.0140 | Salience Network (SN)/Dorsal visual stream | ↓ |
| Right Supramarginal Gyrus (Salience.SMG) | T(43) = 3.93 | 0.0003 | 0.0240 | Salience Network (SN) | ↓ |
| Left Precentral Gyrus (PreCG l) | T(43) = − 3.80 | 0.0005 | 0.0240 | Sensorimotor Network (SMN) | ↑ |
Note: The first row reports cluster-level omnibus statistics (F-test) summarizing the overall connectivity pattern emerging from the seed region, including cluster size (number of significant connections), intensity (sum of suprathreshold test statistics across all connections), and the corresponding uncorrected (p_unc) and FDR-corrected (p_FDR) p-values
Fig. 2.
Seed-based resting-state connectivity alterations in episodic cluster headache (eCH). (A) Left superior frontal gyrus (SFG_l) seed: decreased connectivity with right supramarginal regions and increased connectivity with left precentral cortex in eCH compared with controls. (B) Right inferior frontal gyrus, pars triangularis (IFGtri_r) seed: widespread increased connectivity with bilateral insulae, opercular and supramarginal regions, and decreased connectivity with right superior frontal and angular gyri. All results are shown at cluster-level p-FDR < 0.025. Red lines indicate increased connectivity in patients compared with controls; blue lines indicate decreased connectivity
In contrast, from the IFGtri_r seed, patients exhibited a widespread bilateral FC increase with the insulae, opercular regions, supramarginal gyri, planum temporale, and postcentral cortices, alongside a localized FC decrease with the right superior frontal gyrus (SFG_r) and the angular gyrus (Table 5; Fig. 2).
Table 5.
Functional connectivity alterations from the right inferior frontal gyrus, Pars Triangularis (IFGtri_r) seed in episodic cluster headache (eCH) patients. Seed-based resting-state functional connectivity analysis from the right inferior frontal gyrus (IFG Tri r) comparing eCH patients with healthy controls (two-sample t-test showed widespread hyperconnectivity from IFGtri_r to bilateral insular, opercular, supramarginal, auditory/language, and sensorimotor regions, and reduced connectivity with right superior frontal and angular gyri. All results are corrected for multiple comparisons (analysis p-FDR < 0.025). Arrows indicate direction of change in patients (↑ increased; ↓ decreased). Sign of T: positive T = Patients < Controls; negative T = Patients > Controls. Arrows: ↑ increased in patients; ↓ decreased in patients. Network attribution follows Beckmann et al. and CONN functional network classification
| Seed / Target Region | Statistic | p_unc | p_FDR | Network | Direction |
|---|---|---|---|---|---|
| Seed Right Inferior Frontal Gyrus, pars triangularis (IFG tri r) |
F(9,35) = 4.74 Intensity = 101.29 Size = 29 |
0.0004 | 0.0004 | ventral attention network (VAN) | — |
| Left Insular Cortex (IC l) | T(43) = − 4.44 | 0.0001 | 0.0053 | Salience Network (SN) | ↑ |
| Right Planum Polare (PP r) | T(43) = − 4.40 | 0.0001 | 0.0053 | Auditory / Language Network | ↑ |
| Left Central Opercular Cortex (CO l) | T(43) = − 4.31 | 0.0001 | 0.0053 | Salience Network (SN) | ↑ |
| Right Angular Gyrus (AGr) | T(43) = 4.11 | 0.0002 | 0.0055 | Default Mode Network (DMN) | ↓ |
| Right Anterior Supramarginal Gyrus (aSMG r) | T(43) = − 4.05 | 0.0002 | 0.0055 | Salience Network/ Dorsal Visual Stream | ↑ |
| Left Planum Polare (PP l) | T(43) = − 4.03 | 0.0002 | 0.0055 | Auditory / Language Network | ↑ |
| Left Parietal Operculum (PO l) | T(43) = − 4.01 | 0.0002 | 0.0055 | Sensorimotor / Salience Interface | ↑ |
| Left Anterior Supramarginal Gyrus (aSMG l) | T(43) = − 3.98 | 0.0003 | 0.0055 | Salience Network/ Dorsal Visual Stream | ↑ |
| Right Planum Temporale (PT r) | T(43) = − 3.92 | 0.0003 | 0.0059 | Auditory / Language Network | ↑ |
| Left Sensorimotor Lateral Network (137) | T(43) = − 3.83 | 0.0004 | 0.0070 | Sensorimotor Network (SMN) | ↑ |
| Right Central Opercular Cortex (CO r) | T(43) = − 3.79 | 0.0005 | 0.0072 | Salience Network (SN) | ↑ |
| Left Heschl’s Gyrus (HG l) | T(43) = − 3.68 | 0.0007 | 0.0090 | Auditory Network | ↑ |
| Left Planum Temporale (PT l) | T(43) = − 3.66 | 0.0007 | 0.0090 | Auditory Network | ↑ |
| Right Sensorimotor Lateral Network (138) | T(43) = − 3.56 | 0.0009 | 0.0111 | Sensorimotor Network (SMN) | ↑ |
| Left Anterior Superior Temporal Gyrus (aSTG l) | T(43) = − 3.38 | 0.0015 | 0.0172 | Auditory / Language Network | ↑ |
| Right Heschl’s Gyrus (HG r) | T(43) = − 3.37 | 0.0016 | 0.0172 | Auditory Network | ↑ |
| Right Parietal Operculum (PO r) | T(43) = − 3.34 | 0.0017 | 0.0174 | Sensorimotor / Salience Interface | ↑ |
| Dorsal Attention Network left Intraparietal Sulcus IPS l (153) | T(43) = − 3.30 | 0.0020 | 0.0185 | Dorsal Attention Network (DAN) | ↑ |
| Right Superior Frontal Gyrus (SFG r) | T(43) = 3.25 | 0.0022 | 0.0202 | Default mode network, cognitive control/execution networks, motor control network | ↓ |
Note: The first row reports cluster-level omnibus statistics (F-test) summarizing the overall connectivity pattern emerging from the seed region, including cluster size, intensity, and the corresponding uncorrected (p_unc) and FDR-corrected (p_FDR) p-values
No significant associations were found between functional connectivity measures and clinical variables after correction for FDR multiple comparisons.
Discussion
This study provides converging structural and functional MRI evidence for a distributed cortical involvement in episodic cluster headache (eCH).
Morphometric findings
We identified two distinct clusters of cortical thinning: one in the right pars triangularis of the inferior frontal gyrus (IFG, BA45) and a second in the left superior frontal gyrus (SFG). We also observed robust positive associations between cortical thickness and clinical burden, including attack frequency and duration, disease duration, and pain intensity, in multimodal associative cortices of the temporal, parietal, and occipital lobes. No significant subcortical volumetric alterations or correlations with clinical variables were detected. These null volumetric findings are consistent with our previous hypothalamic segmentation study in the same eCH cohort, where no macrostructural differences were detected [14].
These results add to the growing body of evidence that CH is a disorder of distributed network dysfunction rather than a purely hypothalamic pathology and extend prior reports of frontal involvement reported by morphometric studies during the bout period and in attack-free comparisons [1, 3, 6–8, 32–34].
To our knowledge, prior morphometric studies in cluster headache did not describe alterations in the pars triangularis of the inferior frontal gyrus (IFG) [11, 32, 34, 35]. The right inferior frontal gyrus (particularly its pars triangularis, BA45) is a canonical node of the ventral attention/reorienting system [36] and shows strong functional coupling with the anterior insula and the dorsal anterior cingulate (cingulo-opercular/salience system), thereby acting as an interface between salience detection and attentional control [37–39]. Through short association fibers of the superior longitudinal fasciculus III, the IFG is structurally linked to parietal VAN nodes (supramarginal/TPJ) [40]; the IFG also shows strong functional coupling with the anterior insula within the cingulo-opercular/salience system [37–39]. This region supports cognitive control, inhibitory and contextual processing, language, and the integration of affective/interoceptive salience, including the cognitive evaluation of pain [41]. Structural alterations in this region may be associated with differences in the processing and regulation of nociceptive salience. In addition, the cortical thinning we observed in the superior frontal gyrus (SFG) aligns with prior VBM/SBM studies in CH reporting alterations in the dorsolateral prefrontal cortex (DLPFC), which overlaps anatomically with the SFG [42]. The SFG/DLPFC are core nodes of the executive control network (ECN), which governs cognitive control and top-down pain modulation through connections with the anterior cingulate, thalamus, and hypothalamus [43, 44]. Dysfunction in these frontal regions has been repeatedly linked to impaired cognitive and affective regulation in chronic pain conditions.
Beyond these frontal clusters, we observed positive correlations between cortical thickness and clinical variables in associative cortices—including the inferior temporal, postcentral, superior parietal, supramarginal, inferior parietal, and lateral occipital regions. These areas belong to multimodal sensory and attentional networks, such as the dorsal attention network (DAN) and visual association systems [36]. Patients with greater disease burden showed thicker cortex in these associative areas. This pattern may reflect burden-related cortical variations or state-dependent differences in cortical morphology. Similar associations between disease load and regional cortical thickness have been reported in the primary somatosensory cortex of eCH patients [35]. These findings are compatible with the possibility that repeated nociceptive input is accompanied by transient morphometric differences [45], although causal inferences cannot be drawn from cross-sectional data.
Functional connectivity findings
Seed-based resting-state analyses were conducted to investigate whether structural alterations in the frontal cortex were accompanied by changes in network integration.
The left superior frontal gyrus (SFG_l) showed reduced connectivity with the right supramarginal gyrus (SMG)—a parietal node of the ventral attention system (VAN) [36]—and increased connectivity with the left precentral gyrus, part of the sensorimotor network (SMN) [46]. This pattern of anatomical connections is consistent with the anatomical connections of the SFG in monkeys and in humans [47]. The distinctive pattern found in our study indicates disengagement of executive–attentional coupling and reinforcement of motor–somatosensory synchronization, consistent with a relative imbalance between top-down attentional control and bottom-up sensory integration [48].
Conversely, the right inferior frontal gyrus (IFGtri_r) exhibited widespread hyperconnectivity with the bilateral insulae and opercular regions, core components of the salience network (SN) [49], and with the supramarginal, planum temporale, and postcentral cortices, which belong to the ventral attention and sensorimotor networks (VAN and SMN) [50]. It also showed reduced connectivity with the right superior frontal gyrus (SFG_r)— a node within the cognitive control network, default mode network, cognitive execution network, and motor control network, with its specific network membership determined by its subregional anatomy [47] —and with the angular gyrus, a key parietal node of the default mode network (DMN) [51]. This IFG functional connectivity pattern seems consistent with the anatomical connections previously found in humans [52] and in our case may indicate increased coupling within salience and attentional systems and weakened executive–executive coordination, consistent with network mis-integration rather than compensatory reorganization.
These results suggest a hierarchical imbalance across cortical networks: diminished top-down regulation from executive hubs (ECN) and enhanced bottom-up coupling across salience, ventral attention, and sensorimotor systems (SN, VAN, SMN). This organization is in line with pain studies showing hyperconnectivity of the salience and sensorimotor networks, together with reduced frontal–parietal control, in migraine and trigeminal autonomic cephalalgias [4, 53–57].
Relation to previous network-level findings
In a previous study from our group [9], we analyzed between-network connectivity in a similar cohort of eCH patients scanned outside attacks. That study found reduced coupling between the salience network (SN) and the left executive control network (ECN), along with abnormal positive connectivity between the default mode network (DMN) and right ECN, replacing the normal anticorrelation observed in healthy subjects. These findings were interpreted as impaired SN-mediated switching between internally (DMN) and externally (ECN) oriented networks, reflecting a loss of cognitive control and salience regulation [37, 38, 58].
The current seed-based results refine these observations by identifying specific frontal–parietal–insular circuits underlying the large-scale network imbalance. The SFG_l–SMG decoupling parallels the reduced integration between executive (ECN) and attentional (VAN/DAN) systems, while IFGtri_r–insula overcoupling mirrors excessive synchronization within the salience network and its interaction with sensorimotor regions. Although the approaches differ (seed-to-voxel vs. inter-network), both converge on the same mechanism of maladaptive network reorganization, characterized by weakened executive–salience coupling and overactivation of bottom-up systems. This cross-validation reinforces the view that eCH reflects a multilevel disturbance of hierarchical network control, extending from microstructural and volumetric changes in the hypothalamus and thalamus [9] to cortical-level dysfunction of the salience–executive axis.
Integrative interpretation and conclusions
Our findings reveal that episodic cluster headache involves both structural and functional abnormalities within a distributed fronto-parietal–insular network. The frontal cortical thinning may represent a stable vulnerability affecting top-down modulatory systems, while direction-specific FC alterations may suggest state-dependent maladaptive reorganization involving the salience, ventral attention, and sensorimotor networks (SN, VAN, SMN). This model aligns with current theories proposing that SN–ECN–DMN interactions form the core of large-scale brain organization supporting goal-directed behavior and internal mentation [38, 39, 58].
In this framework, disruption of salience-driven switching and attentional reallocation may contribute to attentional instability, exaggerated interoceptive salience, and deficient inhibitory control, which may explain the affective and sensory intensity of cluster attacks. This hierarchical dysfunction supports the emerging view of cluster headache as a ‘pain connectopathy’, involving maladaptive interactions among salience, attention, and executive control networks [57].
Limitations and future directions
Several limitations should be acknowledged. Although attack-free status was carefully verified, the exact timing of the last preceding attack was not recorded in hours, limiting fine-grained temporal analyses. Moreover, the lack of information on the temporal distance between bout onset and MRI acquisition prevented analyses of within-bout progression effects. The patient sample showed a marked sex imbalance (24 males, 2 females), reflecting the known male predominance of episodic cluster headache, which may limit the generalizability of the findings, particularly to female patients. Standardized psychometric measures of depression and anxiety were not collected, and symptom heterogeneity during the bout period (including features such as restlessness) was not systematically characterized; therefore, their potential influence on neuroimaging findings could not be assessed.
The cross-sectional design precludes causal inferences, and correlations between cortical thickness and clinical variables cannot distinguish between progressive and reversible effects. Moreover, the present study does not allow determination of whether the observed structural and functional alterations are specific to cluster headache or may reflect neural mechanisms shared across other chronic pain conditions. Comparative studies including other chronic pain populations will be necessary to address disease specificity.
Healthy controls were recruited from medical students and healthcare professionals, which may limit the representativeness of the control sample. However, this recruitment strategy ensured a highly selected population without known neurological, systemic, or psychiatric disorders and without regular medication use, thereby reducing potential clinical confounders. The modest sample size further limits generalizability, and the absence of standardized headache impact scales constrains clinical interpretability.
Additionally, differences between seed-based and inter-network analyses may partly explain variability in connectivity directionality. However, the reproducibility of the altered salience–executive and attention–motor coupling across methods supports its robustness. Future multimodal longitudinal studies combining morphometry, diffusion imaging, and fMRI should clarify whether these alterations represent relatively stable vulnerabilities or state-dependent adaptations to repeated pain exposure, and whether normalization of these networks parallels clinical improvement following therapy.
Author contributions
FCar, AdR and GC conceived and designed the study. FCar, MF, IG, FCo, FL, DC, GG, GS, FCas, CA, MA and MS contributed to data acquisition, analysis, and interpretation. AdR performed the MRI morphometric analysis. FCar, AdR, and GC drafted the manuscript. All authors substantially revised the work and approved the final submitted version.
Funding
Not applicable.
Data availability
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
The study was approved by the Ethics Committee of the Policlinico Umberto I – University Sapienza of Rome in accordance with the ethical principles of the Declaration of Helsinki (N° 0295/2023).
Consent for publication
All authors consent for the publication.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Francesca Caramia, Antonio Di Renzo and Irene Giardina contributed equally to this work.
Contributor Information
Francesca Caramia, Email: francesca.caramia@uniroma1.it.
Antonio Di Renzo, Email: antoniomp777@hotmail.it.
Gianluca Coppola, Email: gianluca.coppola@uniroma1.it.
References
- 1.Wei DY, Goadsby PJ (2021) Cluster headache pathophysiology — insights from current and emerging treatments. Nat Rev Neurol 17:308–324. 10.1038/s41582-021-00477-w [DOI] [PubMed] [Google Scholar]
- 2.Petersen AS, Lund N, Goadsby PJ, Belin AC, Wang S-J, Fronczek R, Burish M, Cho S-J, Peres MFP, Jensen RH (2024) Recent advances in diagnosing, managing, and Understanding the pathophysiology of cluster headache. Lancet Neurol 23:712–724. 10.1016/S1474-4422(24)00143-1 [DOI] [PubMed] [Google Scholar]
- 3.May A, Bahra A, Büchel C, Frackowiak RS, Goadsby PJ (1998) Hypothalamic activation in cluster headache attacks. Lancet 352:275–278. 10.1016/S0140-6736(98)02470-2 [DOI] [PubMed] [Google Scholar]
- 4.Coppola G, Abagnale C, Sebastianelli G, Goadsby PJ (2024) Pathophysiology of cluster headache: from the trigeminovascular system to the cerebral networks. Cephalalgia 44:03331024231209317. 10.1177/03331024231209317 [DOI] [PubMed] [Google Scholar]
- 5.Faragó P, Szabó N, Tóth E, Tuka B, Király A, Csete G, Párdutz Á, Szok D, Tajti J, Ertsey C, Vécsei L, Kincses ZT (2017) Ipsilateral alteration of resting state activity suggests that cortical dysfunction contributes to the pathogenesis of cluster headache. Brain Topogr 30:281–289. 10.1007/s10548-016-0535-x [DOI] [PubMed] [Google Scholar]
- 6.Wang X, Zhang L, Xiong Y, Hou M, Zhang S, Duan C, Wang S, Wang X, Lu H, Huang J, Li Y, Li Z, Dong Z, Lou X (2025) Limbic system abnormalities in episodic cluster headache: a 7T MRI multimodal study. J Headache Pain 26:69. 10.1186/s10194-025-02009-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Qiu E, Xing X, Wang Y, Tian L (2024) Altered functional connectivity of the thalamus and salience network in patients with cluster headache: a pilot study. Neurol Sci 45:269–276. 10.1007/s10072-023-07011-4 [DOI] [PubMed] [Google Scholar]
- 8.Chen Y, Xing X, Dai W, Tian L, Dong Z, Yu S (2022) Brain regions involved in fractional amplitude of low-frequency fluctuation in cluster headache patients: a resting-state functional MRI study. BMC Neurol 22:336. 10.1186/s12883-022-02863-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Abagnale C, Di Renzo A, Giuliani G, Sebastianelli G, Casillo F, Ziccardi L, Parisi V, Di Lorenzo C, Serrao M, Caramia F, Di Piero V, Coppola G (2025) MRI-based analysis of the microstructure of the thalamus and hypothalamus and functional connectivity between cortical networks in episodic cluster headache. J Headache Pain 26:12. 10.1186/s10194-024-01920-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Messina R, Filippi M (2020) What we gain from machine learning studies in headache patients. Front Neurol 11:221. 10.3389/fneur.2020.00221 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Absinta M, Rocca MA, Colombo B, Falini A, Comi G, Filippi M (2012) Selective decreased grey matter volume of the pain-matrix network in cluster headache. Cephalalgia 32:109–115. 10.1177/0333102411431334 [DOI] [PubMed] [Google Scholar]
- 12.Demichelis G, Pinardi C, Giani L, Medina JP, Gianeri R, Bruzzone MG, Becker B, Proietti A, Leone M, Chiapparini L, Ferraro S, Nigri A (2022) Chronic cluster headache: A study of the telencephalic and cerebellar cortical thickness. Cephalalgia 42:444–454. 10.1177/03331024211058205 [DOI] [PubMed] [Google Scholar]
- 13.Király A, Szabó N, Párdutz Á, Tóth E, Tajti J, Csete G, Faragó P, Bodnár P, Szok D, Tuka B, Pálinkás É, Ertsey C, Vécsei L, Kincses ZT (2018) Macro- and microstructural alterations of the subcortical structures in episodic cluster headache. Cephalalgia 38:662–673. 10.1177/0333102417703762 [DOI] [PubMed] [Google Scholar]
- 14.Chiffi D, Di Renzo A, Giuliani G, Abagnale C, Altieri M, Sebastianelli G, Casillo F, Di Piero V, Coppola G, Caramia F (2025) Assessment of hypothalamic macrostructure in episodic cluster headache: a volumetric segmentation MRI study. Radiol Med (Torino). 10.1007/s11547-025-02041-8 [DOI] [PubMed] [Google Scholar]
- 15.Lee DA, Lee H-J, Kim HC, Park KM (2022) Alterations of the structural covariance network in the hypothalamus of patients with cluster headache. J Neurol 269:316–322. 10.1007/s00415-021-10629-z [DOI] [PubMed] [Google Scholar]
- 16.Dale AM, Fischl B, Sereno MI (1999) Cortical Surface-Based analysis. NeuroImage 9:179–194. 10.1006/nimg.1998.0395 [DOI] [PubMed] [Google Scholar]
- 17.Dale AM, Sereno MI (1993) Improved localizadon of cortical activity by combining EEG and MEG with MRI cortical surface reconstruction: A linear approach. J Cogn Neurosci 5:162–176. 10.1162/jocn.1993.5.2.162 [DOI] [PubMed] [Google Scholar]
- 18.Fischl B, Dale AM (2000) Measuring the thickness of the human cerebral cortex from magnetic resonance images. Proc Natl Acad Sci 97:11050–11055. 10.1073/pnas.200033797 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Fischl B, Liu A, Dale AM (2001) Automated manifold surgery: constructing geometrically accurate and topologically correct models of the human cerebral cortex. IEEE Trans Med Imaging 20:70–80. 10.1109/42.906426 [DOI] [PubMed] [Google Scholar]
- 20.Fischl B, Salat DH, Busa E, Albert M, Dieterich M, Haselgrove C, Van Der Kouwe A, Killiany R, Kennedy D, Klaveness S, Montillo A, Makris N, Rosen B, Dale AM (2002) Whole brain segmentation. Neuron 33:341–355. 10.1016/S0896-6273(02)00569-X [DOI] [PubMed] [Google Scholar]
- 21.Fischl B, Salat DH, Van Der Kouwe AJW, Makris N, Ségonne F, Quinn BT, Dale AM (2004) Sequence-independent segmentation of magnetic resonance images. NeuroImage 23:S69–S84. 10.1016/j.neuroimage.2004.07.016 [DOI] [PubMed]
- 22.Fischl B, Sereno MI, Tootell RBH, Dale AM (1999) High-resolution intersubject averaging and a coordinate system for the cortical surface. Hum Brain Mapp 8:272–284. 10.1002/(SICI)1097-0193(1999)8:4%25253C272::AID-HBM10%25253E3.0.CO;2-4 [DOI] [PMC free article] [PubMed]
- 23.Han X, Jovicich J, Salat D, Van Der Kouwe A, Quinn B, Czanner S, Busa E, Pacheco J, Albert M, Killiany R, Maguire P, Rosas D, Makris N, Dale A, Dickerson B, Fischl B (2006) Reliability of MRI-derived measurements of human cerebral cortical thickness: the effects of field strength, scanner upgrade and manufacturer. NeuroImage 32:180–194. 10.1016/j.neuroimage.2006.02.051 [DOI] [PubMed] [Google Scholar]
- 24.Jovicich J, Czanner S, Greve D, Haley E, Van Der Kouwe A, Gollub R, Kennedy D, Schmitt F, Brown G, MacFall J, Fischl B, Dale A (2006) Reliability in multi-site structural MRI studies: effects of gradient non-linearity correction on Phantom and human data. NeuroImage 30:436–443. 10.1016/j.neuroimage.2005.09.046 [DOI] [PubMed] [Google Scholar]
- 25.Reuter M, Rosas HD, Fischl B (2010) Highly accurate inverse consistent registration: A robust approach. NeuroImage 53:1181–1196. 10.1016/j.neuroimage.2010.07.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Reuter M, Schmansky NJ, Rosas HD, Fischl B (2012) Within-subject template Estimation for unbiased longitudinal image analysis. NeuroImage 61:1402–1418. 10.1016/j.neuroimage.2012.02.084 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ségonne F, Dale AM, Busa E, Glessner M, Salat D, Hahn HK, Fischl B (2004) A hybrid approach to the skull stripping problem in MRI. NeuroImage 22:1060–1075. 10.1016/j.neuroimage.2004.03.032 [DOI] [PubMed] [Google Scholar]
- 28.Fischl B, Sereno MI, Dale AM (1999) Cortical Surface-Based analysis. NeuroImage 9:195–207. 10.1006/nimg.1998.0396 [DOI] [PubMed] [Google Scholar]
- 29.Desikan RS, Ségonne F, Fischl B, Quinn BT, Dickerson BC, Blacker D, Buckner RL, Dale AM, Maguire RP, Hyman BT, Albert MS, Killiany RJ (2006) An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. NeuroImage 31:968–980. 10.1016/j.neuroimage.2006.01.021 [DOI] [PubMed] [Google Scholar]
- 30.Whitfield-Gabrieli S, Nieto-Castanon A (2012) Conn: A functional connectivity toolbox for correlated and anticorrelated brain networks. Brain Connect 2:125–141. 10.1089/brain.2012.0073 [DOI] [PubMed] [Google Scholar]
- 31.Morfini F, Whitfield-Gabrieli S, Nieto-Castañón A (2023) Functional connectivity MRI quality control procedures in CONN. Front Neurosci 17:1092125. 10.3389/fnins.2023.1092125 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Giorgio A, Lupi C, Zhang J, De Cesaris F, Alessandri M, Mortilla M, Federico A, Geppetti P, De Stefano N, Benemei S (2020) Changes in grey matter volume and functional connectivity in cluster headache versus migraine. Brain Imaging Behav 14:496–504. 10.1007/s11682-019-00046-2 [DOI] [PubMed] [Google Scholar]
- 33.Chou K-H, Yang F-C, Fuh J-L, Kuo C-Y, Wang Y-H, Lirng J-F, Lin Y-Y, Wang S-J, Lin C-P (2017) Bout-associated intrinsic functional network changes in cluster headache: A longitudinal resting-state functional MRI study. Cephalalgia 37:1152–1163. 10.1177/0333102416668657 [DOI] [PubMed] [Google Scholar]
- 34.Yang F-C, Chou K-H, Fuh J-L, Huang C-C, Lirng J-F, Lin Y-Y, Lin C-P, Wang S-J (2013) Altered Gray matter volume in the frontal pain modulation network in patients with cluster headache. Pain 154:801–807. 10.1016/j.pain.2013.02.005 [DOI] [PubMed] [Google Scholar]
- 35.Seifert CL, Magon S, Staehle K, Zimmer C, Foerschler A, Radue E, Pfaffenrath V, Tölle TR, Sprenger T (2012) A Case-Control study on cortical thickness in episodic cluster headache. Headache J Head Face Pain 52:1362–1368. 10.1111/j.1526-4610.2012.02217.x [DOI] [PubMed] [Google Scholar]
- 36.Corbetta M, Shulman GL (2002) Control of goal-directed and stimulus-driven attention in the brain. Nat Rev Neurosci 3:201–215. 10.1038/nrn755 [DOI] [PubMed] [Google Scholar]
- 37.Seeley WW, Menon V, Schatzberg AF, Keller J, Glover GH, Kenna H, Reiss AL, Greicius MD (2007) Dissociable intrinsic connectivity networks for salience processing and executive control. J Neurosci 27:2349–2356. 10.1523/JNEUROSCI.5587-06.2007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Menon V, Uddin LQ (2010) Saliency, switching, attention and control: a network model of Insula function. Brain Struct Funct 214:655–667. 10.1007/s00429-010-0262-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Uddin LQ (2015) Salience processing and insular cortical function and dysfunction. Nat Rev Neurosci 16:55–61. 10.1038/nrn3857 [DOI] [PubMed] [Google Scholar]
- 40.De Thiebaut M, Urbanski M, Batrancourt B, Levy R, Dubois B, Cerliani L, Volle E (2016) Rostro-caudal architecture of the frontal lobes in humans. 10.1093/cercor/bhw215.CerebCortexcercor;bhw215v1 [DOI] [PMC free article] [PubMed]
- 41.Wiech K, Ploner M, Tracey I (2008) Neurocognitive aspects of pain perception. Trends Cogn Sci 12:306–313. 10.1016/j.tics.2008.05.005 [DOI] [PubMed] [Google Scholar]
- 42.Briggs RG, Khan AB, Chakraborty AR, Abraham CJ, Anderson CD, Karas PJ, Bonney PA, Palejwala AH, Conner AK, O’Donoghue DL, Sughrue ME (2020) Anatomy and white matter connections of the superior frontal gyrus. Clin Anat 33:823–832. 10.1002/ca.23523 [DOI] [PubMed] [Google Scholar]
- 43.Dosenbach NUF, Fair DA, Cohen AL, Schlaggar BL, Petersen SE (2008) A dual-networks architecture of top-down control. Trends Cogn Sci 12:99–105. 10.1016/j.tics.2008.01.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Shen K, Welton T, Lyon M, McCorkindale AN, Sutherland GT, Burnham S, Fripp J, Martins R, Grieve SM (2020) Structural core of the executive control network: A high angular resolution diffusion MRI study. Hum Brain Mapp 41:1226–1236. 10.1002/hbm.24870 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.May A (2011) Experience-dependent structural plasticity in the adult human brain. Trends Cogn Sci 15:475–482. 10.1016/j.tics.2011.08.002 [DOI] [PubMed] [Google Scholar]
- 46.Power JD, Cohen AL, Nelson SM, Wig GS, Barnes KA, Church JA, Vogel AC, Laumann TO, Miezin FM, Schlaggar BL, Petersen SE (2011) Functional network organization of the human brain. Neuron 72:665–678. 10.1016/j.neuron.2011.09.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Li W, Qin W, Liu H, Fan L, Wang J, Jiang T, Yu C (2013) Subregions of the human superior frontal gyrus and their connections. NeuroImage 78:46–58. 10.1016/j.neuroimage.2013.04.011 [DOI] [PubMed] [Google Scholar]
- 48.Bowling JT, Friston KJ, Hopfinger JB (2020) Top-down versus bottom‐up attention differentially modulate frontal–parietal connectivity. Hum Brain Mapp 41:928–942. 10.1002/hbm.24850 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Seeley WW (2019) The salience network: A neural system for perceiving and responding to homeostatic demands. J Neurosci 39:9878–9882. 10.1523/JNEUROSCI.1138-17.2019 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Bernard F, Lemee J, Mazerand E, Leiber L, Menei P, Ter Minassian A (2020) The ventral attention network: the mirror of the Language network in the right brain hemisphere. J Anat 237:632–642. 10.1111/joa.13223 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Vatansever D, Manktelow AE, Sahakian BJ, Menon DK, Stamatakis EA (2017) Angular default mode network connectivity across working memory load. Hum Brain Mapp 38:41–52. 10.1002/hbm.23341 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Li W, Xie K, Ngetich RK, Zhang J, Jin Z, Li L (2021) Inferior frontal Gyrus-Based Resting-State functional connectivity and medium dispositional use of reappraisal strategy. Front Neurosci 15:681859. 10.3389/fnins.2021.681859 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Nardoni L, Tullo MG, Stefano GD, Fiorelli M, Caramia F (2025) Structural brain alterations and changes in resting-state functional connectivity in patients with trigeminal neuralgia: a meta-analysis. NeuroImage Clin 103759. 10.1016/j.nicl.2025.103759 [DOI] [PMC free article] [PubMed]
- 54.Wei H-L, Chen J, Chen Y-C, Yu Y-S, Guo X, Zhou G-P, Zhou Q-Q, He Z-Z, Yang L, Yin X, Li J, Zhang H (2020) Impaired effective functional connectivity of the sensorimotor network in interictal episodic migraineurs without aura. J Headache Pain 21:111. 10.1186/s10194-020-01176-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Lim M, Jassar H, Kim DJ, Nascimento TD, DaSilva AF (2021) Differential alteration of fMRI signal variability in the ascending trigeminal somatosensory and pain modulatory pathways in migraine. J Headache Pain 22:4. 10.1186/s10194-020-01210-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Lee CH, Park H, Lee MJ, Park B (2023) Whole-brain functional gradients reveal cortical and subcortical alterations in patients with episodic migraine. Hum Brain Mapp 44:2224–2233. 10.1002/hbm.26204 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57.Silvestro M, Tessitore A, Caiazzo G, Scotto Di Clemente F, Trojsi F, Cirillo M, Esposito F, Tedeschi G, Russo A (2021) Disconnectome of the migraine brain: a connectopathy model. J Headache Pain 22:102. 10.1186/s10194-021-01315-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58.Buckner RL, DiNicola LM (2019) The brain’s default network: updated anatomy, physiology and evolving insights. Nat Rev Neurosci 20:593–608. 10.1038/s41583-019-0212-7 [DOI] [PubMed] [Google Scholar]
- 59.Beckmann CF, DeLuca M, Devlin JT, Smith SM (2005) Investigations into resting-state connectivity using independent component analysis. Philos Trans R Soc Lond B Biol Sci 360:1001–1013. 10.1098/rstb.2005.1634 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.