Migraine, a complex neurovascular disorder affecting approximately 1.3 billion individuals worldwide (global prevalence: 14.4%), arises from multifaceted pathophysiological interactions involving trigeminovascular sensitization, cortical spreading depression (CSD), and dysregulation of brainstem modulatory systems (1, 2). Epidemiological studies reveal marked sexual dimorphism, with females exhibiting a threefold higher prevalence than males (odds ratio [OR] = 3.21, 95% confidence interval [CI]: 2.98–3.45), likely attributable to estrogen fluctuation-driven hyperactivation of the calcitonin gene-related peptide (CGRP) pathway (3). Clinically, migraine manifests as unilateral pulsatile headaches (visual analog scale [VAS] score ≥ 6) lasting 4–72 h, accompanied by photophobia/phonophobia (82.3% incidence) and cognitive processing speed reduction (31.7 ± 8.9% below baseline) (4). Designated by the World Health Organization (WHO) as the second-leading global cause of disability, migraine incurs annual economic losses of $1.57 trillion, equivalent to a 5.3% reduction in quality-adjusted life years (QALYs) per patient (5, 6).

In neuromodulation therapeutics, transcranial direct current stimulation (tDCS) demonstrates unique therapeutic potential by modulating cortical excitability (anodal stimulation: +38.5% neuronal firing rate; cathodal inhibition: −27.2%) and inducing long-term potentiation (LTP) (7). A randomized double-blind trial demonstrated that bifrontal tDCS targeting the primary somatosensory cortex (M1; 2 mA, 20 min/day × 10 days) reduced monthly headache days by ≥50% in 58.3% of participants (95% CI, 52.4–64.1%), with sustained efficacy for 22.7 ± 4.3 weeks (hazard ratio [HR] = 0.44, p < 0.001) and a favorable safety profile (adverse events: 4.2% vs. 17.8% in pharmacotherapy controls) (8). However, three critical knowledge gaps persist: (1) spatiotemporal coupling mechanisms between tDCS-induced γ oscillations (30–80 Hz) and CSD propagation trajectories; (2) causal electrophysiological evidence linking θ-band (4–12 Hz) power enhancement to pain threshold elevation (r = 0.72, p < 0.01); (3) regulatory networks governing synaptic plasticity biomarkers (e.g., BDNF, PSD-95) in chronic stimulation models (9).

To address these limitations, we established a nitroglycerin (NTG, 10 mg/kg)-induced murine migraine model replicating key pathological features of human migraine, including dural mast cell degranulation (histamine release: +215%) and trigeminal ganglion CGRP overexpression (3.8-fold mRNA upregulation) (8). Our experimental paradigm integrates low-intensity tDCS (0.25 mA, 20 min/day × 7 days) with multimodal assessments: (1) Open-field test (OFT) quantification of locomotor activity (total movement time, average velocity) and anxiety-like behaviors (grooming frequency: NTG group 38.2 ± 6.1 vs. control 12.4 ± 3.2, p < 0.001); (2) 32-channel microelectrode array (MEA) recordings of M1 local field potentials (LFPs; sampling rate 2,048 Hz) to capture pre−/post-stimulation (0–60 min) oscillatory dynamics; (3) Multiscale signal analysis incorporating power spectral density (PSD; Hanning window 1,024 points), cross-frequency coupling (CFC; n:m phase-amplitude coupling), and nonlinear dynamics via sample entropy (SampEn; embedding dimension m = 2, tolerance r = 0.2 × SD), comprehensively decoding tDCS-mediated reorganization of δ-γ oscillatory hierarchies (8). This innovative strategy provides the first multilevel mechanistic integration from behavioral phenotypes to neural circuitry, offering translational evidence for optimizing tDCS clinical protocols (Scheme 1).

This investigation provides systematic evidence for the therapeutic efficacy of transcranial direct current stimulation (tDCS) in a nitroglycerin (NTG)-induced murine migraine model, employing multimodal behavioral and neurophysiological assessments. Key findings demonstrate that tDCS intervention (20 min, 35.4 A/m²) significantly ameliorated migraine-related nocifensive responses, notably cephalic scratching (p < 0.01) and prolonged grooming (p < 0.05), concurrent with restoration of pathological cortical oscillations in the somatosensory cortex. It was found in previous studies that low-intensity tDCS (0.25 mA, 20 min) is hypothesized to modulate cortical excitability via subthreshold polarization of neuronal membranes, preferentially affecting slow oscillations (δ/θ) due to their dependence on thalamocortical loop synchronization. For example, δ (1–4 Hz) and θ (4–12 Hz) rhythms are implicated in migraine-related thalamocortical dysrhythmia, where aberrant low-frequency activity may underlie photophobia and central sensitization (22). Gamma (γ, 30–80 Hz) suppression, observed in our NTG model, aligns with evidence of E/I imbalance in migraine, where tDCS may restore fast-spiking interneuron activity (e.g., parvalbumin-positive cells) to normalize γ power and sensory processing (23). In NTG-treated mice, δ/θ potentiation correlates with increased light aversion (photophobia), while γ suppression parallels mechanical allodynia—both behaviors reversed by tDCS. This mirrors clinical findings where δ/θ-γ decoupling predicts migraine attack susceptibility. We selected these bands as primary targets due to their established roles in migraine pathophysiology (δ/θ for thalamocortical dysrhythmia; γ for cortical hyperexcitability) and their responsiveness to neuromodulation. Electrophysiological analyses revealed that NTG administration induced frequency-specific dysregulation marked by δ-band (1–4 Hz) and broadband γ-oscillation (30–100 Hz) hyperactivity, coupled with θ-rhythm (4–12 Hz) attenuation—a spectral imbalance partially rectified by tDCS through δ/γ power suppression (35–42% reduction) and θ-activity potentiation (18–24% increase). To our knowledge, this constitutes the first mechanistic elucidation of tDCS-driven cortical rhythm stabilization in NTG-mediated migraine pathogenesis, advancing previous reports of NTG-induced neuronal hyperexcitability (24–26) through frequency-domain characterization of network dysfunction. The increased theta activity (4–12 Hz) aligns with prior evidence of thalamocortical dysrhythmia in migraine, particularly involving the medial dorsal thalamus and its projections to sensory cortices (27, 28). This is consistent with studies showing aberrant low-frequency oscillations in migraineurs during the interictal phase, which may reflect a hyperexcitable state predisposing to attacks (29). The observed gamma suppression (50–100 Hz) is contextualized within the framework of excitation-inhibition (E/I) imbalance in migraine, where glutamate/GABA dysregulation may disrupt fast-spiking interneurons critical for gamma generation (27, 30). We link this to recent MEG studies showing impaired high-frequency connectivity in migraineurs, suggesting that gamma suppression could reflect compromised top-down modulation from frontal or thalamic regions (31).

The differential neuromodulatory effects of tDCS, favoring δ/low-γ suppression over θ-enhancement, suggest parameter-dependent efficacy potentially mediated by: (1) polarity-specific membrane potential modulation altering neuronal firing thresholds (32, 33); (2) corticothalamocortical circuit rebalancing via bidirectional corticospinal regulation (34, 35); (3) gliotransmitter-mediated astrocytic Ca²⁺ wave synchronization (36); or (4) spike-timing-dependent plasticity mimicking LTP/LTD dynamics. While our data corroborate these hypothesized pathways, the transient therapeutic window (<180 min post-stimulation) emphasizes the necessity for protocol optimization to prolong clinical benefits (37, 38).

Three principal limitations warrant consideration: First, the incomplete characterization of tDCS mechanisms in migraine models necessitates finer-scale investigations into subcellular polarization gradients, tripartite synapse dynamics, and plasticity-related molecular markers. Second, sparse temporal sampling (30 min and 3 h intervals) may overlook critical oscillatory state transitions during migraine chronification. Third, the absence of longitudinal monitoring limits insights into cumulative neuromodulatory effects or tolerance development. Future investigations should integrate closed-loop electrophysiology, optogenetic circuit dissection, and metabolomic-transcriptomic profiling to resolve spatiotemporal modulation patterns. Fourth, the relatively small sample size (n = 6 per group) may limit the statistical power to detect subtle effects, despite our use of non-parametric tests and effect size reporting to mitigate this issue. Larger cohorts (n = 10–12 per group) will be used in the future studies. Fifth, our study exclusively used male mice, which restricts the generalizability of the results to female subjects. This is particularly relevant given emerging evidence of sex-specific differences in pain pathways and drug efficacy. For instance, prior work demonstrates that male and female rodents exhibit divergent neuroimmune responses and male-biased samples may inadvertently overlook critical biological variables. Future studies should incorporate both sexes to evaluate potential sex-dependent effects, aligning with NIH guidelines to treat sex as a biological variable. The relationship between open-field tests and migraine remains incompletely elucidated. In future studies, we will integrate periorbital von Frey testing and light aversion assays at matched LFP recording time points, implementing a synchronized electrophysiology-behavioral monitoring protocol (including acclimation procedures to minimize stress-induced artifacts). What’s more, multimodal imaging (fMRI/PET) will be used to correlate our electrophysiological findings with network-level dysfunction in thalamocortical circuits in the future study.

Notwithstanding these constraints, our results establish a neurophysiological framework for tDCS-mediated migraine intervention, demonstrating significant correlations between δ/γ entropy normalization and behavioral recovery (Spearman’s r = 0.82, p = 0.004). The identification of spectral entropy as a quantifiable biomarker enhances prospects for personalized neuromodulation strategies. Translationally, these preclinical findings underscore the imperative for phase II trials evaluating tDCS dosing regimens (0.5–2.0 mA intensity, 20–40 min duration) in migraineurs, complemented by concurrent mechanistic studies using transcranial magnetic stimulation-electroencephalography (TMS-EEG) to bridge rodent-human neurophysiological correlates.

Our findings demonstrate that nitroglycerin (NTG) administration significantly alters the mean absolute power, relative power, and entropy of local field potentials (LFPs) in the mouse somatosensory cortex, concomitant with observable behavioral changes. Notably, low-intensity transcranial direct current stimulation (tDCS) effectively modulated both behavioral parameters and LFP characteristics (absolute power, relative power, and entropy) in NTG-induced migraine models. These results suggest that tDCS may facilitate neural homeostasis restoration, potentially through cortical excitability regulation, thereby supporting its therapeutic potential for migraine recovery in murine models.