Migraine is a leading cause of disability worldwide, significantly impacting quality of life and healthcare systems. Despite its high prevalence and burden, migraine remains underprioritized in global health policies. This study examines the epidemiological trends of migraine in Australia, China, Japan, and South Korea from 1990 to 2021, highlighting regional disparities and forecasting future burdens. This study utilized data from the Global Burden of Disease (GBD) 2021 to analyze incidence, prevalence, and years lived with disability (YLDs) of migraine. Age-standardized rates (ASRs) were calculated to enable fair cross-country comparisons. Joinpoint regression analysis was applied to assess temporal trends, while Bayesian age-period-cohort (BAPC) modeling was used to project future trends until 2050. Additionally, decomposition analysis was conducted to differentiate the effects of population aging, growth, and epidemiological changes. In 2021, China had the highest migraine burden, with 13.05 million new cases and 184.75 million prevalent cases, followed by Australia, Japan, and South Korea. Incidence rates peaked in adolescence (10–14 years), while prevalence and disability were highest in middle-aged women (40–44 years). From 1990 to 2021, Australia exhibited stable trends, China experienced increasing burden, Japan saw a decline due to aging, and Korea exhibited mixed patterns influenced by opposing demographic and epidemiological forces. Future projections suggest a stable trend in Australia, declining incidence in China and Japan, and continued burden in Korea. Migraine remains a significant public health challenge across all four countries, with age, gender, and demographic changes playing key roles in burden variations. The study highlights the need for region-specific healthcare strategies and age- and gender-sensitive interventions. Future research should explore socioeconomic, behavioral, and healthcare access factors to refine migraine management strategies.

BACKGROUND: New daily persistent headache (NDPH) is a rare, refractory primary headache with an unclear pathophysiological mechanism. Previous neuroimaging studies on NDPH have been largely limited to 3T MRI, which fails to thoroughly reveal microstructural changes, particularly subregional abnormalities and iron metabolism alterations. Based on this, the present study employs 7T multimodal imaging techniques to investigate structural, functional, and iron metabolism abnormalities in the whole brain and, in particular, the changes in subregions of the limbic system. METHODS: A total of 23 individuals with NDPH and 23 healthy controls (HCs) underwent 7T MRI, including T1-weighted three-dimensional magnetization-prepared 2 rapid acquisition gradient echo (3D-T1WI-MP2RAGE) and resting-state fMRI (rs-fMRI); among these patients, 19 also underwent quantitative susceptibility mapping (QSM). Structural volumes, functional metrics (fractional amplitude of low-frequency fluctuations [fALFF], regional homogeneity [ReHo]), and iron deposition (assessed via QSM) were analyzed, and their correlations with clinical parameters (e.g., headache history, anxiety/depression scores) were examined. RESULTS: Compared to HCs, individuals with NDPH exhibited significantly less volume in the right accumbens area and left caudal anterior cingulate cortex after false discovery rate (FDR) correction. Widespread changed fALFF and ReHo values were observed, with correlations to clinical features. QSM values were decreased in right paracentral, right cuneus and left precentral with increasing in left rostral middle frontal. CONCLUSION: 7T multimodal MRI identifies widespread structural, functional, and iron metabolism abnormalities in NDPH, particularly in limbic subregions, highlighting a “pain-emotion” interaction mechanism. These findings provide preliminary insights into NDPH pathogenesis.

ObjectivesTo evaluate the diagnostic accuracy of the quantitative flow ratio (QFR) for hemodynamic exploration of intracranial atherosclerotic stenosis, using the invasive cerebrovascular pressure ratio (CVPR) and resting full-cycle ratio (RFR) as reference standards.Materials and methodsPatients with symptomatic unifocal intracranial atherosclerotic stenosis were included. The CVPR was defined as the ratio of the proximal and distal pressures. All patients underwent angioplasty under general anesthesia. The QFR was calculated based on digital subtraction angiography. Using the CVPR as a reference, we compared its correlation with the QFR across different degrees and locations of stenosis.ResultsThe CVPR and QFR were measured in 34 vessels of 32 patients. The QFR demonstrated a high correlation and excellent agreement(r = 0.8227, p < 0.001) with the CVPR in distal stenosis before intervention. In the subgroup with diameter stenosis >80%, the QFR showed a high correlation (r = 0.8812, p < 0.001) with the CVPR. In the anterior circulation subgroup, the QFR showed an excellent correlation (r = 0.9066, p < 0.001) with the CVPR. In the posterior circulation subgroup, the QFR showed a high correlation with the CVPR (r = 0.7706, p < 0.001). Diameter stenosis rates showed a moderate negative correlation with the CVPR.ConclusionThere was a strong correlation between the QFR and wire-based CVPR, especially for anterior circulation lesions before intervention. The QFR may serve as a predictive factor for evaluating hemodynamic changes in intracranial atherosclerotic stenosis.

The integration of stereoelectroencephalography with therapeutic deep brain stimulation (DBS) holds immense promise as a viable approach for precise treatment of refractory disorders, yet it has not been explored in the domain of headache or pain management. Here, we implanted 14 electrodes in a patient with refractory migraine and integrated clinical assessment and electrophysiological data to investigate personalized targets for refractory headache treatment. Using statistical analyses and cross-validated machine-learning models, we identified high-frequency oscillations in the right nucleus accumbens as a critical headache-related biomarker. Through a systematic bipolar stimulation approach and blinded sham-controlled survey, combined with real-time electrophysiological data, we successfully identified the left dorsal anterior cingulate cortex as the optimal target for the best potential treatment. In this pilot study, the concept of the herein-proposed data-driven approach to optimizing precise and personalized treatment strategies for DBS may create a new frontier in the field of refractory headache and even pain disorders.

BackgroundsThe dysfunction of dopamine in the mesocorticolimbic dopamine system in MOH is unknown. Dopamine transporter (DAT) regulates dopamine clearance and neurotransmission and is sensitive to dopamine levels. A decrease in DAT availability can reflect a decrease in dopamine. To determine DAT availability abnormalities in the mesocorticolimbic dopamine system and explore functional network changes in medication overuse headache (MOH) patients.MethodsWe examined 17 MOH patients and 16 healthy controls (HCs) using integrated positron emission tomography (PET)/magnetic resonance (MR) brain scans with 11 CFT, a radioligand that binds to DAT. Standardized uptake value ratio (SUVr) images were compared voxelwise between MOH patients and HCs. Then, the significantly changed cluster ( p < 0.01, GRF correction) with abnormal DAT availability was selected as a specific seed region to further evaluate altered functional connectivity (FC) in MOH. SUVr and mean FC values from significantly changed regions were extracted, and partial correlation analyses with clinical measures were conducted.ResultsMOH patients had lower SUVr levels in the left rather than right orbitofrontal cortex (OFC) than HCs. There was altered FC between the left OFC and the left superior temporal pole and bilateral calcarine gyri. SUVr levels in the left OFC and the connectivity strength linking the positive brain regions with the left OFC were not correlated with clinical measures in the MOH patients.ConclusionsMOH is characterized by decreased DAT availability in the left OFC, which might reflect compensatory downregulation due to low dopamine signalling within the mesocorticolimbic dopamine system. In addition, the OFC on both sides may have different functions in the pathogenesis of MOH. Altered intrinsic FC in the left OFC was identified in MOH patients, which may provide a new perspective to understand the pathogenesis of MOH.

Medication overuse headache (MOH) is a serious global condition. The interaction between headache attacks and medication overuse complicates the understanding of its pathophysiology. In this study, we developed a preclinical MOH model that incorporates these two key factors by overusing rizatriptan benzoate (RIZ, 4 mg/kg, i.g.) in a glyceryl trinitrate (GTN, 10 mg/kg, i.p.) induced chronic migraine mouse model. We observed that RIZ overuse aggravated GTN-induced cutaneous allodynia and caused a prolonged state of latent sensitization. We also detected a significant upregulation of Annexin-A1 (ANXA1), a protein mainly expressed in the microglia of the spinal trigeminal nucleus caudalis (SPVC), in GTN+RIZ mice. Intracerebroventricular injection of ANXA1-derived peptide Ac2-26 trifluoroacetic acid (TFA) (5 μg/mouse) inhibited bright light stress (BLS) induced acute allodynia via the formyl peptide receptor (FPR) in GTN+RIZ mice. These results suggest that ANXA1 may have an analgesic effect in triptan-associated MOH and could potentially serve as a therapeutic target.

Objective Migraine has been demonstrated to exhibit abnormal functional connectivity of large‐scale brain networks, which is closely associated with its pathophysiology and has not yet been explored by edge functional connectivity. We used an edge‐centric approach combined with motif analysis to evaluate higher‐order communication patterns of brain networks in migraine. Methods We investigated edge‐centric metrics in 108 interictal migraine patients and 71 healthy controls. We parcellated the brain into networks using independent component analysis. We applied edge graph construction, k‐means clustering, community overlap detection, graph‐theory‐based evaluations, and clinical correlation analysis. We conducted motif analysis to explore the interactions among regions, and a classification model to test the specificity of edge‐centric results. Results The normalized entropy of lateral thalamus was significantly increased in migraine, which was positively correlated with the baseline headache duration, and negatively correlated with headache duration reduction following preventive medications at 3‐month follow‐up. Network‐wise entropy of the sensorimotor network was significantly elevated in migraine. The community similarity between lateral thalamus and postcentral gyrus was enhanced in migraine. Migraine patients showed overrepresented L‐shape and diverse motifs, and underrepresented forked motifs with lateral thalamus serving as the reference node. Furthermore, migraine patients presented with overrepresented L‐shape triads, where the postcentral gyrus shared different edges with the lateral thalamus. The classification model showed that entropy of the lateral thalamus had the highest discriminative power, with an area under the curve of 0.86. Interpretation Our findings indicated an abnormal higher‐order thalamo–cortical communication pattern in migraine patients. The thalamo–cortical–somatosensory disturbance of concerted working may potentially lead to aberrant information flow and deficit pain processing of migraine. ANN NEUROL 2023;94:1168–1181

It has yet to be determined whether medication overuse headache (MOH) is an independent disorder or a combination of primary headache and substance addiction. To further explore the causes of MOH, we compared MOH with substance use disorder (SUD) in terms of the brain regions involved to draw more targeted conclusions. In this review, we selected alcohol use disorder (AUD) as a representative SUD and compared MOH and AUD from two aspects of neuroimaging and basic research. We found that in neuroimaging studies, there were many overlaps between AUD and MOH in the reward circuit, but the extensive cerebral cortex damage in AUD was more serious than that in MOH. This difference was considered to reflect the sensitivity of the cortex structure to alcohol damage. In future research, we will focus on the central amygdala (CeA), prefrontal cortex (PFC), orbital-frontal cortex (OFC), hippocampus, and other brain regions for interventions, which may have unexpected benefits for addiction and headache symptoms in MOH patients.

—In neuroimaging analysis, functional mag- netic resonance imaging (fMRI) can well assess brain function changes for brain diseases with no obvious structural lesions. So far, most deep-learning-based fMRI studies take functional connectivity as the basic feature in disease clas- sification. However, functional connectivity is often calculated based on time series of predefined regions of interest and neglects detailed information contained in each voxel, which may accordingly deteriorate the performance of di- agnostic models. Another methodological drawback is the limited sample size for the training of deep models. In this study, we propose BrainFormer, a general hybrid Transformer architecture for brain disease classification with single fMRI volume to fully exploit the voxel-wise details with sufficient data dimensions and sizes. BrainFormer is constructed by modeling the local cues within each voxel with 3D convolutions and capturing the global relations among distant regions with two global attention blocks. The local and global cues are aggregated in BrainFormer by a single-stream model. To handle multisite data, we propose a normalization layer to normalize the data into iden- tical distribution. Finally, a Gradient-based Localization-map Visualization method is utilized for locating the pos- sible disease-related biomarker. We evaluate BrainFormer on five independently acquired datasets including ABIDE, ADNI, MPILMBB, ADHD-200 and ECHO, with diseases of autism, Alzheimer’s disease, depression, attention deficit hyperactivity disorder, and headache disorders. The results demonstrate the effectiveness and generalizability of Brain- Former for multiple brain diseases diagnosis. BrainFormer may promote neuroimaging-based precision diagnosis in clinical practice and motivate future study in fMRI analysis.

BACKGROUND:Cortical spreading depression (CSD) is an electrophysiological event underlying migraine aura. Traditional CSD models are invasive and often cause injuries. The aim of the study was to establish a minimally invasive optogenetic CSD model and identify the active networks after CSD using whole-brain activity mapping.METHODS:CSD was induced in mice by light illumination, and their periorbital thresholds and behaviours in the open field, elevated plus-maze and light-aversion were recorded. Using c-fos, we mapped the brain activity after CSD. The whole brain was imaged, reconstructed and analyzed using the Volumetric Imaging with Synchronized on-the-fly-scan and Readout technique. To ensure the accuracy of the results, the immunofluorescence staining method was used to verify the imaging results.RESULTS:The optogenetic CSD model showed significantly decreased periorbital thresholds, increased facial grooming and freezing behaviours and prominent light-aversion behaviours. Brain activity mapping revealed that the somatosensory, primary sensory, olfactory, basal ganglia and default mode networks were activated. However, the thalamus and trigeminal nucleus caudalis were not activated.CONCLUSIONS:Optogenetic CSD model could mimic the behaviours of headache and photophobia. Moreover, the optogenetic CSD could activate multiple sensory cortical regions without the thalamus or trigeminal nucleus caudalis to induce cortical pain.

Background: The efficacy of enteral immunonutrition (EIN) in patients undergoing gastrointestinal cancer surgery remains debatable. This meta-analysis aimed to investigate the effectiveness of EIN administration in patients undergoing surgery for gastrointestinal cancer. Methods: From January 2000 to January 2022, PubMed, EMBASE, Cochrane Library, and Web of Science were thoroughly searched for randomized controlled trials (RCTs) with EIN versus standard diet or no supplement in patients undergoing surgery for gastrointestinal cancer. Overall complications and infectious complications were the primary outcomes. The secondary results were non-infectious complications, mortality, length of hospital stay, and enteral nutrition-related complications. Results: Thirty-five studies reporting 3,692 patients undergoing surgery for gastrointestinal cancer (including gastric cancer, colorectal cancer, esophageal cancer, periampullary cancer, or pancreatic cancer) were included. Compared with the control group, EIN group had a significantly decreased incidence of overall complications (RR = 0.79, p < 0.001). Infectious complications in patients who received EIN were considerably lower than in the control group (RR = 0.66, p < 0.001). Compared to the control group, the incidence of surgical site infection, abdominal abscess, anastomotic leakage, bacteremia, duration of systemic inflammatory response syndrome (SIRS), and duration of antibiotic therapy was significantly lower in the specific infectious complications treated with EIN. Still, there was no significant difference between the two groups with other infectious complications. Moreover, a substantial shortening in the length of hospital stay was shown in EIN group compared with the control group. Still, no significant effect of EIN was demonstrated in non-infectious complicatios and mortality. The enteral nutrition-related complications had no significant difference between two groups. Conclusions: EIN is safe and effective in reducing overall complications, infectious complications, and hospital stay in patients undergoing gastrointestinal cancer surgery (including gastric cancer, colorectal cancer, esophageal cancer, periampullary cancer, or pancreatic cancer).

The relationship between the severity of intracranial atherosclerotic disease and the circle of Willis integrity is unclear. In this brief report, we investigate the associations between symptomatic intracranial atherosclerotic disease and the integrity of the circle of Willis. Patients with symptomatic intracranial atherosclerosis were enrolled and underwent intracranial artery magnetic resonance vessel wall imaging and time-of-flight angiography. The presence or absence of an intracranial atherosclerotic plaque and its maximum wall thickness and stenosis were evaluated. The presence or absence of the A1 segment of the bilateral anterior cerebral arteries (from the internal carotid artery to the anterior communicating artery segment is called anterior cerebral artery A1 segment), and anterior communicating artery, the P1 segment of the bilateral posterior cerebral arteries (The P1 segment of the posterior cerebral artery is a horizontally outward segment), and bilateral posterior communicating arteries were determined. The associations of the intracranial plaque features with the integrity of the circle of Willis were analyzed. Of the 110 recruited subjects (57.2 ± 11.1 years; 65% males), 51 had intracranial plaques, and 44 had stenosis. In patients with bilateral A1 and P1 segments (n = 85), intracranial stenosis was more severe in patients with an anterior communicating artery than those without an anterior communicating artery (19.7% ± 21.7% vs. 1.4% ± 3.3%, p = 0.046). In patients with bilateral A1 and P1 segments and an anterior communicating artery (n = 79), intracranial stenosis was more severe in patients with posterior communicating arteries than those without posterior communicating arteries (27.9% ± 23.7% vs. 13.5% ± 17.9%, p = 0.007). The odds ratio of intracranial stenosis was 1.214 (95% confidence interval (CI), 1.054-1.398; p = 0.007) in discriminating for the presence of posterior communicating arteries in patients with bilateral A1 and P1 segments and an anterior communicating artery after adjusting for confounding factors. The severity of intracranial atherosclerosis was independently associated with the presence of posterior communicating arteries in patients with a complete anterior part of the circle of Willis.

PurposeThis study sought to determine the associations of co-existing hypertension and hyperhomocysteinemia (H-Hcy) with carotid vulnerable plaque features and subsequent vascular events.MethodsSymptomatic patients with carotid atherosclerosis were enrolled and underwent carotid magnetic resonance (MR) vessel wall imaging. The patients were divided into the following groups: co-existing hypertension and H-Hcy group; isolated hypertension group; isolated H-Hcy group; and control group. The morphological and compositional characteristics of carotid plaques were assessed on MR images and compared among different groups. Univariate and multivariate cox regressions were used to calculate the hazard ratio (HR) and corresponding 95% confidence interval (CI) of co-existing hypertension and H-Hcy in predicting subsequent vascular events after at least 1-year followed-up.ResultsIn total, 217 patients (mean age, 59.4 ± 11.9 years; 154 males) were recruited. Patients in co-existing hypertension and H-Hcy group had a significantly higher prevalence of carotid lipid-rich necrotic core (LRNC) than isolated H-Hcy and control group (73.2 vs. 43.3 vs. 50%, p = 0.015). During the median follow-up time of 12.2 ± 4.3 months, 61 (39.8%) patients experienced vascular events. After adjusting for baseline confounding factors, co-existing hypertension and H-Hcy (HR, 1.82; 95% CI, 1.01–3.27; p = 0.044), presence of carotid LRNC (HR, 2.25; 95% CI, 1.09–4.65; p = 0.029), and combination of co-existing hypertension and H-Hcy and carotid LRNC (HR, 2.39; 95% CI, 1.26–4.43; p = 0.007) were significantly associated with subsequent vascular events.ConclusionsCo-existing hypertension and H-Hcy are associated with carotid vulnerable plaque features, such as LRNC. Combining co-existing hypertension and H-Hcy with carotid vulnerable plaque features has a stronger predictive value for subsequent vascular events than each measurement alone.

Background We used resting-state functional magnetic resonance imaging (RS-fMRI) to assess the possible pathogenic role of fALFF in CH. A limited number of studies have reported on fractional amplitude of low-frequency fluctuation (fALFF) in cluster headache (CH). Methods RS-fMRI scans of 23 patients with CH were obtained (11with left-sided headache and 12 with right-sided headache), along with scans of 23 age- and sex-matched normal controls. The RS-fMRI data were analyzed to explore abnormal brain activity in the left CH and right CH patients during the non-painful state in one cluster period. fALFF was compared between patients and controls, and correlation analysis between the regional mean fALFF values and clinical characteristics was performed. Results A decrease in fALFF was detected in the left cerebellum, left lentiform nucleus, left frontal lobe, left anterior cingulate, and right postcentral gyrus in the left CH group compared to the controls, while a decrease of fALFF was detected in the right cerebellum, right cingulate gyrus, right superior parietal lobule, right inferior parietal lobule, right postcentral gyrus, and left precuneus in the right CH group. No patient had a region with increased fALFF. A moderate correlation was observed between some regional mean fALFF values and the clinical characteristics. Conclusions We deduced that dysfunction in multiple brain areas is involved in the non-painful state of CH during a cluster period.

Background Mesenchymal stem cells (MSCs) have shown immense therapeutic potential for various brain diseases. Intrathecal administration of MSCs may enhance their recruitment to lesions in the central nervous system, but any impact on cerebrospinal fluid (CSF) flow remains unclear. Methods Rats with or without middle cerebral artery occlusion (MCAO) received intrathecal injections of 2D cultured MSCs, 3D cultured MSCs or an equal volume of artificial cerebrospinal fluid (ACSF). Ventricle volume was assessed by MRI on Days 2 and 14 post-MCAO surgery. A beam walking test was used to assess fine motor coordination and balance. Aggregation of MSCs was evaluated in CSF and frozen brain tissue. Differential expression of cell adhesion molecules was evaluated by RNA-Seq, flow cytometry and immunofluorescence analyses. The influence of VCAM-1 blockade in mediating the aggregation of 2D MSCs was investigated in vitro by counting cells that passed through a strainer and in vivo by evaluating ventricular dilation. Results MSC expanded in 2D culture formed aggregates in the CSF and caused ventricular enlargement in both MCAO and normal rats. Aggregates were associated with impaired motor function. 2D MSCs expressed higher levels of integrin α4 and VCAM-1 than 3D MSCs. Blockade of VCAM-1 in 2D MSCs reduced their aggregation in vitro and reduced lateral ventricular enlargement after intrathecal infusion. 3D MSCs exhibited lower cell aggregation and reduced cerebral ventricular dilation after intrathecal transplantation Conclusions The aggregation of 2D MSCs, mediated by the interaction of integrin α4 and VCAM-1, is a potential risk for obstruction of CSF flow after intrathecal transplantation.

In neuroimaging analysis, fMRI can well assess the function changes for brain diseases with no obvious structural lesions. To date, most deep-learning-based fMRI studies have employed functional connectivity (FC) as the basic feature for disease classification. However, FC is calculated on time series of predefined regions of interest and neglects detailed information contained in each voxel. Another drawback of using FC is the limited sample size for the training of deep models. The low representation ability of FC leads to poor performance in clinical practice, especially when dealing with multimodal medical data involving multiple types of visual signals and textual records for brain diseases. To overcome this bottleneck problem in the fMRI feature modality, we propose BrainFormer, an end-to-end functional interaction learning method for brain disease classification with single fMRI volume. Unlike traditional deep learning methods that construct convolution and transformers on FC, BrainFormer learns the functional interaction from fMRI signals, by modeling the local cues within each voxel with 3D convolutions and capturing the global correlations among distant regions with specially designed global attention mechanisms from shallow layers to deep layers. Meanwhile, BrainFormer can deal with multimodal medical data including fMRI volume, structural MRI, FC features and phenotypic data to achieve more comprehensive brain disease diagnosis. We evaluate BrainFormer on five independent multi-site datasets on autism, Alzheimer's disease, depression, attention deficit hyperactivity disorder and headache disorders. The results demonstrate its effectiveness and generalizability for multiple brain diseases diagnosis with multimodal features. BrainFormer may promote precision of neuroimaging-based diagnosis in clinical practice and motivate future studies on fMRI analysis.

Migraine is the second most prevalent disorder in the world; yet, its underlying mechanisms are still poorly understood. Cumulative studies have revealed pivotal roles of cerebral cortex in the initiation, propagation, and termination of migraine attacks as well as the interictal phase. Investigation of basic mechanisms of the cortex in migraine not only brings insight into the underlying pathophysiology but also provides the basis for designing novel treatments. We aim to summarize the current research literatures and give a brief overview of the cortex and its role in migraine, including the basic structure and function; structural, functional, and biochemical neuroimaging; migraine-related genes; and theories related to cortex in migraine pathophysiology. We propose that long-term plasticity of synaptic transmission in the cortex encodes migraine.

BACKGROUND:The previous documents demonstrated that iron deposition was identified in brain deep nuclei and periaqueductal gray matter region in chronic migraine (CM), and less is known about the cerebral iron deposition in CM. The aim of this study is to investigate the cerebral iron deposition in CM using an advanced voxel-based quantitative susceptibility mapping. METHODS:A multi-echo gradient echo MR sequence was obtained from 14 CM patients and 28 normal controls (NC), and quantitative susceptibility mapping images were reconstructed and voxel-based analysis was performed over the whole cerebrum. The susceptibility value of all the positive brain regions was extracted and correlation was calculated between the susceptibility value and the clinical variables. RESULTS:The brain regions with increased susceptibility value in CM patients located in right precuneus, insula, supramarginal gyrus, dorsolateral superior frontal gyrus, postcentral gyrus, cuneus and left postcentral gyrus compared with NC. The correlation analysis demonstrated that a positive correlation was identified between susceptibility value of all the positive brain regions and VAS score. CONCLUSION:The current study demonstrated increased cerebral iron deposition presented in chronic patients, which suggested that increased cerebral iron deposition might play a role in the migraine chronicization.