Motor Cortex rTMS Alleviates Pain and Comorbid Depression After Spinal Cord Injury with a Correlate in M1 Intracortical Inhibition Authors: Yixing Lu, Kun Zhao, Xinyu Liu, Xiaodong Lin, Baijie Xue, Xin Zhang, Rong Li, Feng Feng, Wenhui Ding, Xiaolong Sun, Hua Yuan Published: 2026 ABSTRACT AbstractPurposeNeuropathic pain (NP) after spinal cord injury (SCI) is a debilitating syndrome, yet current treatments demonstrate limited efficacy. This randomized, double-blind, sham-controlled study investigated the analgesic efficacy of high-frequency repetitive transcranial magnetic stimulation (rTMS) over the left primary motor cortex (M1) and its neurophysiological mechanisms.Patients and MethodsTwenty-nine SCI patients with NP were randomly assigned (3:1) to receive either active rTMS (10 Hz, a total of 1200 pulses at an intensity of 80% resting motor threshold) or sham rTMS over the left M1 corresponding to the hand area once daily for two weeks. Pain intensity (Numeric Rating Scale, NRS), depression (9-item Patient Health Questionnaire, PHQ-9), and anxiety (7-item Generalized Anxiety Disorder Questionnaire, GAD-7) were assessed at baseline and after treatment. Furthermore, M1 intracortical excitability parameters, including short-interval intracortical inhibition (SICI) and intracortical facilitation (ICF), were measured using paired-pulse TMS (ppTMS).ResultsAfter the two-week intervention, active rTMS significantly reduced NRS scores (P < 0.001), PHQ-9 scores (P = 0.023), and GAD-7 scores (P = 0.002), and significantly increased both SICI and ICF values (P < 0.001 for both). Interestingly, the baseline NRS scores were correlated with the PHQ-9 scores (r = 0.392, P = 0.032) and the SICI values (r = −0.458, P = 0.017), and changes in the NRS scores correlated with changes in the PHQ-9 scores (r = 0.436, P = 0.025) and the SICI values (r = −0.571, P = 0.004).ConclusionThese findings indicate that high-frequency rTMS alleviates NP and comorbid depression after SCI and is accompanied by changes in M1 intracortical inhibition. However, the 3:1 allocation resulted in a small sham group and limited statistical power for between-group comparisons, and confirmation in larger trials is required.Keywords: spinal cord injury, neuropathic pain, transcranial magnetic stimulation, primary motor cortex, intracortical inhibition Introduction Spinal cord injury (SCI) is defined as structural and functional damage to the spinal cord caused by traumatic factors (for example, traffic accidents, violence, and high falls) or non-traumatic factors (for example, inflammation and tumors). According to the latest data from the Global Burden of Disease database, approximately 20.6 million people live with SCI worldwide, with about 0.9 million new cases annually.1 Beyond the immediate motor impairments, neuropathic pain (NP) emerges as one of the most debilitating and refractory conditions, affecting approximately 40%–70% of individuals with SCI.2,3 In addition to the subjective discomfort caused by pain itself, NP after SCI can also lead to comorbid psychological symptoms including depression and anxiety. A meta-analysis showed that among patients with chronic NP, approximately 37.5% suffer from depression and 33.8% from anxiety.4 NP not only severely affects patients’ rehabilitation process but also imposes a heavy economic burden on families and society. However, the clinical management of NP after SCI remains a formidable challenge, primarily because medications such as pregabalin and gabapentin offer only limited relief but are associated with a high incidence of adverse events.5,6 Moreover, evidence indicates an increasing risk of gabapentinoid misuse, potentially leading to serious harm to patients.7,8 Consequently, there is an urgent and unmet clinical need to explore non-pharmacological neuromodulatory strategies that can specifically target the underlying neural mechanisms of pain without the systemic toxicity associated with pharmacological therapies. Repetitive transcranial magnetic stimulation (rTMS) is a non-invasive neuromodulation technique that enables targeted regulation of cortical excitability. High-frequency rTMS (≥5 Hz) exerts an excitatory modulatory effect, and the latest international guidelines have designated high-frequency rTMS targeted at the primary motor cortex (M1) contralateral to the painful side as a Level A recommendation for the treatment of NP.9 Nevertheless, NP after SCI frequently presents bilaterally,10 and current guidelines do not provide specific recommendations regarding stimulation laterality for this population. Notably, for fibromyalgia, a chronic widespread pain condition without a consistently lateralized painful side, high-frequency rTMS of the left M1 has been recommended with probable efficacy for improving quality of life, providing a clinical precedent for the use of a fixed left M1 target in non-lateralized pain conditions.9 Moreover, unilateral high-frequency rTMS of M1 can modify excitability in the unstimulated contralateral M1 through interhemispheric interactions, providing a physiological basis for cross-hemispheric neuromodulation.11,12 Experimental studies further indicate that M1 can engage distributed pain-control pathways involving supraspinal and descending circuits that ultimately modulate nociceptive processing.13,14 In addition, our previous studies consistently applied 10 Hz rTMS over the left M1 hand area in patients with NP after SCI and demonstrated favorable analgesic effects,15–17 providing an empirical basis for maintaining the same target in the present study. Taken together, these clinical and neurophysiological observations provided the rationale for targeting the left M1 in the present study despite the frequent bilateral distribution of NP after SCI. Moreover, a recent animal study demonstrated that high-frequency rTMS reverses the excitation-inhibition imbalance within the primary somatosensory cortex (S1)-M1 microcircuitry, specifically by rectifying enhanced feedforward inhibition mediated by M1 parvalbumin-positive GABAergic interneurons.14 This suggests that modulation of intracortical GABAergic circuits may contribute to the analgesic effects of rTMS. However, clinical evidence regarding rTMS-induced changes in intracortical excitability remains limited. In our previous clinical study, we found that 10 Hz rTMS over the left M1 hand area alleviated NP after SCI and increased motor evoked potential (MEP) amplitude, reflecting enhanced corticospinal output excitability.17 However, the study did not assess intracortical inhibition or facilitation. Consequently, whether the GABAergic and glutamatergic intracortical circuits within M1 are engaged during rTMS analgesia in humans remains unknown. Paired-pulse transcranial magnetic stimulation (ppTMS) is a neurophysiological technique that evaluates the intracortical excitation state of the M1 hand area, including inhibition and facilitation, by recording MEPs induced by a paired-pulse protocol.18 The core indicators include short-interval intracortical inhibition (SICI) and intracortical facilitation (ICF). SICI is pharmacologically associated with gamma-aminobutyric acid (GABA) type A (GABAA) receptor-mediated inhibition,19–22 whereas ICF is pharmacologically associated with N-methyl-D-aspartic acid (NMDA) receptor-mediated facilitation.23 Accordingly, SICI and ICF are commonly used as electrophysiological measures of intracortical inhibitory and facilitatory function within M1. Here, we conducted a randomized, double-blind, sham-controlled trial to evaluate the efficacy and cortical mechanisms of 10 Hz rTMS targeting the left M1 hand area for NP after SCI. ppTMS was used to characterize changes in intracortical inhibition and facilitation in patients with NP after SCI before and after rTMS intervention, providing an indirect electrophysiological window into intracortical inhibitory and facilitatory function within M1. The primary objective of this study was to determine whether a two-week course of high-frequency rTMS significantly reduced pain intensity and improved emotional well-being in patients with NP after SCI. Concurrently, we sought to elucidate the neurophysiological mechanisms underlying these clinical effects by examining changes in intracortical inhibition and facilitation. We hypothesized that active high-frequency rTMS would not only significantly reduce pain but also normalize aberrant M1 intracortical excitability patterns. Materials and Methods Study Design and Participants This study was a randomized, double-blind, sham-controlled trial. Patients with SCI hospitalized in the Department of Rehabilitation Medicine at Xijing Hospital between December 2024 and December 2025 were consecutively screened for enrollment. Inclusion criteria: (1) Age between 18 and 70 years; (2) Right-handedness confirmed by the Edinburgh Handedness Inventory; (3) Meeting the diagnostic criteria for SCI,24 with time since injury ≤ 1 year; (4) Meeting the diagnostic criteria for NP after SCI as defined in the International Association for the Study of Pain classification of chronic pain for ICD-11;25 (5) Provision of written informed consent. Exclusion criteria: (1) Presence of non-neuropathic pain (for example, chronic musculoskeletal pain, visceral pain); (2) Contraindications to TMS;26 (3) Comorbid brain injury, severe systemic diseases, psychiatric disorders, cognitive impairment, or consciousness disorders; (4) Receipt of rTMS treatment within the previous 2 weeks or initiation or dose adjustment of medications with known effects on cortical excitability; stable concomitant analgesic medications were permitted provided that the regimen remained unchanged throughout the study period; (5) Inability to evoke an MEP. Sample Size Calculation Sample size was calculated using G*Power software (version 3.1.9.7). Our previous study found that after two weeks of high-frequency rTMS treatment, the mean Numeric Rating Scale (NRS) score was 2.27 ± 0.47 in the active group and 3.83 ± 1.17 in the control group.15 Given the analgesic effects observed in our previous studies15–17 and ethical considerations regarding participant burden associated with sham stimulation in patients with NP after SCI, a 3:1 allocation ratio was chosen to reduce the number of participants assigned to sham stimulation while retaining a concurrent control group for efficacy and neurophysiological comparisons. With a two-sided significance level (α) of 0.05, a statistical power (1-β) of 0.80, G*Power indicated a minimum required sample of 26 participants (19 active and 7 sham). Allowing for an anticipated dropout rate of 5%, the planned enrollment was at least 28 participants. Randomization and Blinding An independent researcher who was not involved in participant recruitment, intervention delivery, or outcome assessment performed randomization using computer software. A block randomization design with a 3:1 allocation ratio (active: sham) was adopted. The block size was set to 4 (3 active, 1 sham), and a random permutation of treatment assignments was generated within each block. The resulting allocation sequence was then concealed in sequentially numbered, opaque, sealed envelopes. Eligible participants were consecutively numbered upon enrollment. After all baseline assessments were completed, the treating physician opened the envelope corresponding to the participant’s enrollment number to determine group assignment. Patients and outcome assessors were blinded to group allocation. The physician administering the rTMS intervention was not involved in outcome assessment. Intervention Protocol Both groups received comprehensive rehabilitation interventions according to guidelines, including conventional pharmacological therapy, physical modalities, and exercise therapy. The active group received high-frequency rTMS over the left M1 hand area, using a CCY-1 stimulator with a figure-of-8 coil (Yiruide Medical Equipment Company, Wuhan, China). The stimulation parameters were: frequency of 10 Hz, intensity at 80% of the resting motor threshold (RMT), train duration of 1.2 s with an inter-train interval of 3 s, totaling 1200 pulses per session. Treatment was administered once daily for 2 consecutive weeks. The sham group received identical parameter settings, except that the coil was placed vertically relative to the scalp, which delivered the same auditory stimulation as the active coil without delivering effective magnetic stimulation to the left M1.15 Throughout the study period, the comprehensive rehabilitation regimen and concomitant analgesic medications remained unchanged for all patients. Scale Assessments Scale assessments were conducted at baseline and after 2 weeks of intervention. The primary outcome was pain intensity, assessed by NRS. Patients selected an integer between 0 (no pain) and 10 (unbearable, extremely severe pain that may even cause syncope) based on their pain perception. If multiple pain sites or types were present, the highest score was recorded. Depressive symptoms were assessed using the 9-item Patient Health Questionnaire (PHQ-9), and anxiety symptoms were assessed using the 7-item Generalized Anxiety Disorder Questionnaire (GAD-7). ppTMS Assessments ppTMS assessments were performed at baseline and after 2 weeks of intervention. A Neuro-MS TMS device (Neurosoft LLC, Ivanovo, Russia) equipped with a figure-of-8 coil was used in conjunction with a surface electromyography system recording from the abductor pollicis brevis muscle of the right hand. The coil was placed tangentially to the scalp at a 45° angle to the midline, with the handle pointing backward. After identifying the motor hot spot of the left M1 hand area and determining the RMT, several single-pulse test stimuli (TS) were delivered at 120% RMT (interval ≥ 5 s), and the average MEP amplitude was calculated as the unconditioned TS MEP amplitude. Subsequently, paired-pulse paradigms were applied: a TS at 120% RMT was preceded by a conditioning stimulus (CS) at 90% RMT, with an inter-stimulus interval (ISI) of 2.5 ms for SICI and 12 ms for ICF. These ISIs were selected based on a previously published ppTMS protocol in individuals with SCI.27 The MEP elicited by the TS when preceded by the CS was defined as the conditioned TS MEP. For each paired-pulse paradigm, the conditioned TS MEP amplitude was expressed as a percentage of the unconditioned TS MEP amplitude: SICI or ICF (%) = conditioned TS MEP amplitude / unconditioned TS MEP amplitude × 100. Accordingly, a lower SICI value indicates stronger intracortical inhibition, whereas a higher SICI value indicates weaker intracortical inhibition; conversely, a higher ICF value indicates stronger intracortical facilitation. Statistical Analysis Statistical analyses were performed using R software (version 4.3.3). Continuous data were described as median and interquartile range (IQR) [M (P25, P75)] and categorical data as counts and percentages [n (%)]. For baseline inter-group comparisons, continuous data were analyzed using the Mann–Whitney U-test, and categorical data were analyzed using Fisher’s exact test. For intra-group comparisons, the Wilcoxon signed-rank test was used to analyze pre- and post-intervention differences within each group. For inter-group comparisons of intervention effects, rank-transformed analysis of covariance (ANCOVA) was performed with baseline values as covariates. Effect sizes for the ANCOVA were quantified using partial η2, and corresponding 95% confidence intervals (CIs) were calculated and reported irrespective of statistical significance. Pearson’s correlation analyses based on 1000 permutation tests were used to investigate the association between NRS scores and other outcomes.28 A two-sided P-value < 0.05 was considered statistically significant. Figures were generated using GraphPad Prism (version 10.4.1) and R software (version 4.3.3). Results Participant Characteristics A total of 22 patients in the active group and 7 in the sham group were included, and all patients completed the 2-week intervention (Figure 1). Study Flowchart. Patients with SCI hospitalized in the Department of Rehabilitation Medicine at Xijing Hospital between December 2024 and December 2025 were consecutively screened for enrollment (n = 78). Of these, 33 patients were excluded for failing to meet the inclusion criteria, and 16 patients were excluded based on the exclusion criteria. The remaining 29 patients were randomly assigned (3:1) to receive either active rTMS (n = 22) or sham rTMS (n = 7) over the left M1 corresponding to the hand area once daily for two weeks. All patients completed the 2-week intervention. Abbreviations: NP, neuropathic pain; rTMS, repetitive transcranial magnetic stimulation; MEP, motor evoked potential. As shown in Table 1, there were no significant differences between the two groups regarding demographic data (gender, age and education), SCI characteristics (injury severity, injury level and time since injury), baseline scale scores (NRS, PHQ-9 and GAD-7 scores) and baseline ppTMS assessments (SICI and ICF values) (all P > 0.05). The distribution of concomitant analgesic medication use was comparable between the active and sham groups, with no significant difference in the proportions of patients receiving a single medication, ≥2 medications, or no analgesic medication (P = 1.000). Table 1Baseline CharacteristicsSham Group (n = 7)Active Group (n = 22)ZPGender Female1 (14.3%)3 (13.6%)–1.000 Male6 (85.7%)19 (86.4%)Age (y)41.00 (32.00, 52.00)37.50 (34.25, 53.25)0.2810.779Education ≤ Junior high school4 (57.1%)15 (68.2%)–0.665 ≥ High school3 (42.9%)7 (31.8%)Injury severity Complete SCI5 (71.4%)13 (59.1%)–0.677 Incomplete SCI2 (28.6%)9 (40.9%)Injury level Cervical SCI1 (14.3%)4 (18.2%)–1.000 Thoracic SCI4 (57.1%)13 (59.1%) Lumbar SCI2 (28.6%)5 (22.7%)Time since injury (d)155.00 (72.00, 169.00)71.50 (31.25, 123.75)−1.0700.284NRS score3.00 (2.00, 4.00)3.00 (2.00, 4.00)0.4200.675PHQ-9 score3.00 (0.00, 16.00)3.50 (1.75, 5.25)0.1800.857GAD-7 score4.00 (1.00, 10.00)2.00 (1.00, 6.25)−1.0050.315SICI (%)22.90 (17.90, 51.30)31.10 (16.08, 40.90)−0.1020.919ICF (%)221.20 (186.90, 364.70)220.85 (190.45, 365.95)−0.3060.760Analgesic medication use Single medication2 (28.6%)6 (27.3%)-1.000 ≥2 medications2 (28.6%)7 (31.8%) No analgesic medication3 (42.9%)9 (40.9%)Notes: Continuous data were described as median and IQR [M (P25, P75)] and categorical data as counts and percentages [n (%)]. Gender, education, injury severity, injury level and analgesic medication use were compared by Fisher’s exact test. The remaining data were compared by Mann–Whitney U-test. Analgesic medication use was classified as single-medication use, use of ≥2 analgesic medications, or no analgesic medication. Of the eight patients receiving a single medication, seven received pregabalin and one received venlafaxine. Of the nine patients receiving multiple medications, four received pregabalin plus venlafaxine, two pregabalin plus duloxetine, two pregabalin plus celecoxib, and one pregabalin plus duloxetine and celecoxib.Abbreviations: SCI, spinal cord injury; NRS, Numeric Rating Scale; PHQ-9, 9-item Patient Health Questionnaire; GAD-7, 7-item Generalized Anxiety Disorder Questionnaire; SICI, short-interval intracortical inhibition; ICF, intracortical facilitation; IQR, interquartile range. Baseline Characteristics Notes: Continuous data were described as median and IQR [M (P25, P75)] and categorical data as counts and percentages [n (%)]. Gender, education, injury severity, injury level and analgesic medication use were compared by Fisher’s exact test. The remaining data were compared by Mann–Whitney U-test. Analgesic medication use was classified as single-medication use, use of ≥2 analgesic medications, or no analgesic medication. Of the eight patients receiving a single medication, seven received pregabalin and one received venlafaxine. Of the nine patients receiving multiple medications, four received pregabalin plus venlafaxine, two pregabalin plus duloxetine, two pregabalin plus celecoxib, and one pregabalin plus duloxetine and celecoxib. Abbreviations: SCI, spinal cord injury; NRS, Numeric Rating Scale; PHQ-9, 9-item Patient Health Questionnaire; GAD-7, 7-item Generalized Anxiety Disorder Questionnaire; SICI, short-interval intracortical inhibition; ICF, intracortical facilitation; IQR, interquartile range. Effects of High-Frequency rTMS on Functional Scale Scores At baseline, there were no significant differences in functional scale scores between the two groups (all P > 0.05). After the intervention, the active group showed significant reductions in the NRS scores (P < 0.001, Figure 2A), PHQ-9 scores (P = 0.023, Figure 2B), and GAD-7 scores (P = 0.002, Figure 2C), whereas no significant changes were observed in the sham group (all P > 0.05). Rank-transformed ANCOVA revealed a significant between-group difference in post-intervention NRS scores (F = 4.662, P = 0.040, partial η2 = 0.152, 95% CI = 0.004–1.000). No significant between-group differences were observed for PHQ-9 scores (F = 0.068, P = 0.797, partial η2 = 0.003, 95% CI = 0.000–1.000) or GAD-7 scores (F = 0.029, P = 0.867, partial η2 = 0.001, 95% CI = 0.000–1.000). Effect of high-frequency rTMS on functional scale scores. (A) NRS scores. At baseline, there was no significant difference between the two groups. After the intervention, the active group showed a significant reduction, whereas no significant change was observed in the sham group. Rank-transformed ANCOVA revealed a significant between-group difference in post-intervention NRS scores. (B) PHQ-9 scores. At baseline, there was no significant difference between the two groups. After the intervention, the active group showed a significant reduction, whereas no significant change was observed in the sham group. Rank-transformed ANCOVA revealed no significant between-group difference in post-intervention PHQ-9 scores. (C) GAD-7 scores. At baseline, there was no significant difference between the two groups. After the intervention, the active group showed a significant reduction, whereas no significant change was observed in the sham group. Rank-transformed ANCOVA revealed no significant between-group difference in post-intervention GAD-7 scores. Data are presented as boxplots showing median and IQR [M (P25, P75)], with individual data points overlaid. Inter-group comparisons (pre-active vs pre-sham) were performed using the Mann–Whitney U-test. Intra-group comparisons (pre vs post) were performed using the Wilcoxon signed-rank test. Inter-group comparisons (post-active vs post-sham) were performed using rank-transformed ANCOVA with baseline values as covariates. *P < 0.05, **P < 0.01, ***P < 0.001. Abbreviations: rTMS, repetitive transcranial magnetic stimulation; NRS, Numeric Rating Scale; PHQ-9, 9-item Patient Health Questionnaire; GAD-7, 7-item Generalized Anxiety Disorder Questionnaire; IQR, interquartile range; ANCOVA, analysis of covariance. Effects of High-Frequency rTMS on M1 Intracortical Excitability ppTMS parameters are illustrated in Figure 3A. At baseline, there were no significant differences in SICI values or ICF values between the two groups (both P > 0.05). After the intervention, the active group showed significant increases in SICI values (P < 0.001, Figure 3B) and ICF values (P < 0.001, Figure 3C), whereas no significant changes were observed in the sham group (both P > 0.05). Rank-transformed ANCOVA revealed no significant between-group differences in post-intervention SICI values (F = 4.026, P = 0.055, partial η2 = 0.134, 95% CI = 0.000–1.000) or ICF values (F = 2.748, P = 0.109, partial η2 = 0.096, 95% CI = 0.000–1.000). Across the ANCOVA analyses, the 95% CIs for the partial η2 estimates were wide, indicating considerable uncertainty in the effect-size estimates. Effect of high-frequency rTMS on M1 intracortical excitability. (A) Stimulation parameters of the ppTMS paradigms. SICI employed a conditioning stimulus with 90% RMT stimulation intensity and a test stimulus with 120% RMT stimulation intensity, with an ISI of 2.5 ms. ICF employed a conditioning stimulus with 90% RMT stimulation intensity and a test stimulus with 120% RMT stimulation intensity, with an ISI of 12 ms. (B) SICI values. At baseline, there was no significant difference between the two groups. After the intervention, the active group showed a significant increase, whereas no significant change was observed in the sham group. Rank-transformed ANCOVA revealed no significant between-group difference in post-intervention SICI values. (C) ICF values. At baseline, there was no significant difference between the two groups. After the intervention, the active group showed a significant increase, whereas no significant change was observed in the sham group. Rank-transformed ANCOVA revealed no significant between-group difference in post-intervention ICF values. Data are presented as boxplots showing median and IQR [M (P25, P75)], with individual data points overlaid. Inter-group comparisons (pre-active vs pre-sham) were performed using the Mann–Whitney U-test. Intra-group comparisons (pre vs post) were performed using the Wilcoxon signed-rank test. Inter-group comparisons (post-active vs post-sham) were performed using rank-transformed ANCOVA with baseline values as covariates. ***P < 0.001. Abbreviations: rTMS, repetitive transcranial magnetic stimulation; M1, primary motor cortex; RMT, resting motor threshold; ISI, inter-stimulus interval; SICI, short-interval intracortical inhibition; ICF, intracortical facilitation; IQR, interquartile range; ANCOVA, analysis of covariance. Correlation Analyses Between NRS Scores and Other Outcomes Permutation-based Pearson’s correlation analyses revealed that at baseline, NRS scores were significantly positively correlated with PHQ-9 scores (r = 0.392, P = 0.032, Figure 4A) and significantly negatively correlated with SICI values (r = −0.458, P = 0.017, Figure 4B). After the 2-week intervention, the changes (post − pre) in NRS scores were significantly positively correlated with the changes in PHQ-9 scores (r = 0.436, P = 0.025, Figure 4C) and significantly negatively correlated with the changes in SICI values (r = −0.571, P = 0.004, Figure 4D). Correlation analyses between NRS scores and other outcomes. (A) Correlation between baseline NRS scores and baseline PHQ-9 scores. At baseline, NRS scores were significantly positively correlated with PHQ-9 scores (r = 0.392). (B) Correlation between baseline NRS scores and baseline SICI values. At baseline, NRS scores were significantly negatively correlated with SICI values (r = −0.458). (C) Correlation between changes (post − pre) in NRS scores and changes (post − pre) in PHQ-9 scores. After the 2-week intervention, the changes (post − pre) in NRS scores were significantly positively correlated with the changes in PHQ-9 scores (r = 0.436). (D) Correlation between changes (post − pre) in NRS scores and changes (post − pre) in SICI values. After the 2-week intervention, the changes (post − pre) in NRS scores were significantly negatively correlated with the changes in SICI values (r = −0.571). Scatter plots with fitted lines are based on Pearson correlation analyses with 1000 permutation tests. Abbreviations: NRS, Numeric Rating Scale; PHQ-9, 9-item Patient Health Questionnaire; SICI, short-interval intracortical inhibition. Discussion The present study investigated the analgesic efficacy and neurophysiological mechanisms of high-frequency rTMS targeting the left M1 hand area in patients with NP after SCI. Consistent with previous studies,15–17 our findings demonstrate that high-frequency rTMS significantly reduced pain, depression and anxiety. Moreover, baseline pain intensity was correlated with both depression intensity and SICI values, while changes in pain intensity correlated with corresponding changes in depression intensity and SICI values. These findings suggest a comorbid relationship between pain and depression, and indicate that high-frequency rTMS may exert analgesic efficacy by attenuating abnormally enhanced intracortical inhibition. Previous studies have suggested that rTMS promotes the reconstruction of damaged spinal cord circuits, induces reorganization of the cerebral cortex and circuits, modulates the balance of inputs to motoneurons, and improves the microenvironment and intrinsic regenerative capacity, thereby exerting targeted regulatory effects on SCI rehabilitation.29 However, only a few studies have preliminarily investigated the analgesic mechanism of high-frequency rTMS in NP after SCI. Functional near-infrared spectroscopy (fNIRS), a technique for detecting local blood oxygen concentration in the cerebral cortex, showed that in patients with NP after SCI, activation in the M1 and premotor cortex (PMC) was significantly inhibited following high-frequency rTMS.15,16 In another study, high-frequency rTMS increased MEP amplitude in patients with NP after SCI, indicating enhanced corticospinal excitability.17 In contrast to the above studies that applied fNIRS and MEP measurements, the SICI and ICF assessed by ppTMS in the present study have been shown in previous pharmacological research to be modulated by intracortical GABAA and NMDA receptor function,19–23 providing a basis for exploring the neurophysiological mechanism underlying the analgesic efficacy of rTMS in NP after SCI. The results of the present study suggest that the neurophysiological mechanism of high-frequency rTMS analgesia may be related to altered intracortical inhibitory function reflected by SICI changes. Recent animal studies have elucidated molecular and circuit-level mechanisms underlying M1 high-frequency rTMS analgesia. One study demonstrated that chemogenetic activation of excitatory neurons in the M1 significantly alleviated mechanical and cold allodynia in NP mice, whereas activation of parvalbumin-positive GABAergic interneurons in the M1 markedly exacerbated neuropathic hypersensitivity, and activation of somatostatin-type GABAergic interneurons reduced allodynia.13 This indicates that distinct GABAergic interneuron subtypes within the M1 critically regulate pain perception. Furthermore, another study showed that NP leads to enhanced feedforward inhibition via M1 parvalbumin-positive GABAergic interneurons. Specifically, the S1Glu→M1PV projection displayed synaptic potentiation, resulting in increased inhibitory postsynaptic currents onto M1 pyramidal neurons and decreased output of M1Glu pyramidal neurons. Importantly, 15 Hz high-frequency rTMS effectively reversed this excitation-inhibition imbalance in the S1-M1 microcircuitry and restored M1 output activity.14 These findings demonstrate that NP states are associated with enhanced feedforward inhibition mediated by M1PV neurons, and that high-frequency rTMS may normalize this enhanced inhibition. An additional study found that in an NP mouse model, K-Cl cotransporter isoform 2 (KCC2) expression was increased in the medial prefrontal cortex, leading to a hyperpolarized reversal potential of GABAA currents.30 This resulted in enhanced GABAergic inhibition and medial prefrontal cortex deactivation in NP. Combining the results of these studies and the present study, it is proposed that NP after SCI may involve an inhibition-dominant intracortical state in M1, potentially reflecting excessive GABAA receptor-mediated inhibitory drive, and that high-frequency rTMS may achieve analgesia by suppressing this excessive inhibition. Reduced intracortical inhibition may release M1 pyramidal neurons from excessive inhibitory control and contribute to restoration of the excitation–inhibition balance. The increase in corticospinal excitability after high-frequency rTMS may therefore reflect both reduced inhibitory constraint and enhanced facilitatory processes within M1 circuits. This interpretation extends the classical view that high-frequency rTMS primarily enhances cortical excitability through increased excitatory transmission, suggesting that restoration of M1 output may involve coordinated modulation of both inhibitory and excitatory mechanisms rather than a unidirectional increase in excitation. It is worth noting that a recent study found that application of the GABAA receptor antagonist in the ventrolateral orbital cortex augmented the analgesic efficacy, whereas the GABAA receptor agonist attenuated the analgesic efficacy.31 Interestingly, similar findings have been reported in rTMS studies of depression, in which NMDA receptor agonists enhanced the antidepressant effect of rTMS.32 Therefore, to minimize medication-related confounding, concomitant analgesic regimens were kept unchanged throughout the intervention period, and no medications with known effects on cortical excitability were newly initiated or dose-adjusted during the study. Future clinical rTMS protocols should consider potential interactions between rTMS and medications acting on GABAA receptor-mediated neurotransmission, and clinical studies are warranted to investigate the potential modulatory effects of these medications on the analgesic efficacy of rTMS. Several previous studies have explored the effect of rTMS on intracortical excitability in patients with unilateral NP. However, given the heterogeneity of pain etiologies in these studies, their conclusions remain inconsistent. One study included NP patients with various etiologies (including nerve trunk lesion, brachial plexus lesion, cervical spinal cord lesion, and stroke) and found that SICI enhancement was positively correlated with pain reduction.33 Another study found that for patients with central poststroke pain, rTMS did not alter SICI, whereas ICF was associated with the analgesic response to rTMS.34 An additional study also suggested that the analgesic efficacy of rTMS depends on NMDA receptors.35 The mechanism underlying NP may vary with different etiologies.36 In this study, all included patients suffered from NP after SCI, and this homogeneity may explain the discrepancies between our results and those of previous studies involving NP of other etiologies. Therefore, the analgesic mechanism of rTMS on NP caused by different etiologies may be different. Future studies should incorporate animal experiments to further elucidate the molecular mechanisms underlying rTMS-induced analgesia. Previous studies have documented the high comorbidity between NP and psychological morbidities.4,37 This co-occurrence may stem from the fact that pain and depression share a common neural circuit. A meta-analysis shows the shared cortical deficits involving anterior cingulate cortex-insula/inferior gyrus circuit in chronic pain, major depressive disorder and anxiety disorder, revealing common neural correlates in these comorbid disorders.38 In this study, M1 high-frequency rTMS reduced both pain and depression in SCI patients, and the improvements were positively correlated, possibly because high-frequency rTMS modulated the shared neural circuits between pain and depression. Although guidelines recommend the left dorsolateral prefrontal cortex as a target for rTMS antidepressant treatment,9 an increasing number of studies have focused on the critical role of M1 in pain comorbid with depression. A recent study demonstrated that M1 employs distinct layer-specific pathways to separately control the sensory and aversive-emotional components of NP: layer 6 M1 neurons project via the mediodorsal thalamus to the nucleus accumbens reward circuit, specifically suppressing negative emotional valence and associated coping behaviors, whereas layer 5 M1 neurons connect to the zona incerta and periaqueductal gray to alleviate sensory hypersensitivity without affecting pain affect.13 This finding reveals the neural substrate by which M1 participates in both sensory and emotional regulation, indicating that M1 serves as a hub integrating pain and affective information. Therefore, high-frequency rTMS over M1 may synergistically improve NP and depression by engaging the above layer-specific pathways. Furthermore, a systematic analysis of 22 rTMS studies indicated that high-frequency M1 stimulation is the most effective regimen for NP complicated by depression.39 These findings inform potential therapeutic strategies for the clinical treatment of patients with co-occurring pain and depression. This study is subject to several limitations. First, although the achieved sample size exceeded the minimum sample size estimated a priori, the sham group remained small (n = 7), which limits the precision and statistical power of the between-group estimates. Second, only the immediate effects following a two-week intervention were assessed, with no long-term follow-up. Future studies should include post-treatment follow-up assessments at 2–4 weeks to characterize the persistence and temporal evolution of the observed electrophysiological changes.40 Third, although SICI and ICF provide functional readouts of GABAergic and glutamatergic system activity, they are not direct proxies for neurotransmitter concentrations at the synaptic level. Future validation with animal studies or molecular imaging techniques (for example, magnetic resonance spectroscopy, positron emission tomography) is needed. Fourth, the exclusive focus on the left M1 precludes clarification of the effects of high-frequency rTMS on broader pain-related brain networks. Conclusion In summary, this study demonstrates that high-frequency rTMS targeting the left M1 hand area effectively alleviates both NP and the associated negative emotions in patients with SCI. The analgesic mechanism may involve the normalization of abnormally enhanced intracortical inhibition within M1. Changes in SICI may provide an electrophysiological correlate of treatment response. Its value as a biomarker requires validation in larger and longitudinal cohorts. Acknowledgments The authors express their sincere gratitude to all participants in this trial. Funding Statement This work was supported by the National Natural Science Foundation of China (No. 82472593, 82272591) and the Shaanxi Provincial Key Research and Development Program (2024SF-ZDCYL-01-07). Abbreviations ANCOVA, analysis of covariance; fNIRS, functional near-infrared spectroscopy; GABA, gamma-aminobutyric acid; GAD-7, 7-item Generalized Anxiety Disorder Questionnaire; ICF, intracortical facilitation; IQR, interquartile range; ISI, inter-stimulus interval; KCC2, K-Cl cotransporter isoform 2; M1, primary motor cortex; MEP, motor evoked potential; NMDA, N-methyl-D-aspartic acid; NP, neuropathic pain; NRS, Numeric Rating Scale; PHQ-9, 9-item Patient Health Questionnaire; ppTMS, paired-pulse transcranial magnetic stimulation; PMC, premotor cortex; RMT, resting motor threshold; rTMS, repetitive transcranial magnetic stimulation; S1, primary somatosensory cortex; SCI, spinal cord injury; SICI, short-interval intracortical inhibition. Data Sharing Statement For additional details, please reach out to the corresponding author. The datasets analyzed in this study can be obtained from the corresponding author (Hua Yuan, yuanhua@fmmu.edu.cn) upon reasonable request. Ethics Approval and Informed Consent The study was conducted in accordance with the Declaration of Helsinki. The study was approved by the Ethics Committee of the First Affiliated Hospital (Xijing Hospital) of the Air Force Medical University (KY20242246-X-3) and registered at the Chinese Clinical Trial Registry (ChiCTR, http://www.chictr.org.cn; registration ID: ChiCTR2400093680). All participants provided written informed consent prior to enrollment. Author Contributions All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work. Disclosure The authors declare that they have no competing interests. References Associated Data Data Availability Statement For additional details, please reach out to the corresponding author. The datasets analyzed in this study can be obtained from the corresponding author (Hua Yuan, yuanhua@fmmu.edu.cn) upon reasonable request.