Research Progress of Transcutaneous Auricular Vagus Nerve Stimulation in Chronic Pain Management: A Narrative Review Authors: Danqing Zhou, Pingping Hu, Cao Huang Published: 2026 ABSTRACT AbstractChronic pain represents a significant global public health burden, affecting millions of individuals and substantially impairing quality of life. However, its clinical management is largely limited by the modest efficacy and considerable adverse effects of current pharmacological interventions, highlighting an urgent need for safe and effective non-pharmacological alternatives. Transcutaneous auricular vagus nerve stimulation (taVNS), a non-invasive neuromodulation technique, has emerged as a promising therapeutic strategy. By multitarget neurotransmitter modulation, activating the cholinergic anti-inflammatory pathway (CAP), rebalancing autonomic function, and remodeling central pain networks, taVNS has produced emerging but heterogeneous clinical evidence for pain relief, while most trials have reported that the intervention is generally well-tolerated, which has drawn growing research attention. This narrative review provides a comprehensive, clinically oriented overview of taVNS in chronic pain management, integrating mechanistic insights from preclinical and human studies, appraising the breadth and limitations of existing clinical evidence across pain phenotypes, and delineating key translational barriers. Future directions are proposed to guide both research priorities and clinical decision-making.Keywords: transcutaneous vagus nerve stimulation, chronic pain, neuromodulation, non-pharmacological intervention, analgesia The Chronic Pain Crisis: Toward Non-Pharmacological Solutions Chronic pain is conventionally defined as pain that persists or recurs for more than three months. It affects approximately 30% of the global population and constitutes a substantial public health challenge.1,2 In recent years, chronic pain has been increasingly recognized as an independent clinical entity rather than merely a symptom. It is characterized by complex pathophysiological mechanisms, including nociceptive, neuropathic, and nociplastic (central sensitization) components. It is also frequently comorbid with psychosocial disturbances, such as anxiety and depression.3–5 Beyond its detrimental effects on physical function, sleep quality, and overall quality of life, chronic pain imposes a considerable socioeconomic burden on healthcare systems and society.6 In clinical practice for chronic pain, traditional pharmacological treatments have long occupied a dominant position; however, their limitations have become increasingly apparent.7 Long-term use of commonly used pharmacological agents is often associated with various adverse effects, including gastrointestinal injury, cardiovascular risks, addiction potential, and cognitive impairment.8–10 These agents include NSAIDs, antiepileptic drugs, antidepressants, and opioids, each carrying distinct adverse effect profiles. Meanwhile, interventional treatments such as nerve blocks and implantable neuromodulation demonstrate certain efficacy; however, they are constrained by invasive risks, high costs, and strict indications, which significantly limit their clinical applicability.7 These therapeutic limitations frequently lead to suboptimal pain control, leaving a substantial unmet clinical need. This unmet need has catalyzed a paradigm shift toward personalized, multidisciplinary, and multimodal pain management. In this context, non-invasive neuromodulation has emerged as a prominent therapeutic strategy.11–13 These modalities offer a compelling combination of robust safety and excellent patient tolerability, generating substantial interest across clinical and research domains. Among these neuromodulatory approaches, transcutaneous auricular vagus nerve stimulation (taVNS) exhibits encouraging preliminary therapeutic potential. It produces clinically meaningful analgesia across diverse chronic pain phenotypes, including musculoskeletal, neuropathic, visceral, and autoimmune-related pain, as well as chronic primary pain syndromes. The underlying mechanisms are multifactorial, including multi-target neurotransmitter modulation, activation of the cholinergic anti-inflammatory pathway (CAP), autonomic nervous system rebalancing, and central network remodeling. Neurologist James Corning indirectly regulated vagal tone via carotid compression in the 1880s.14,15 Although this was not true vagus stimulation, Corning’s pioneering work laid the groundwork for neuromodulation.16 Before taVNS, invasive implantable VNS dominated but was limited by high costs, surgical trauma and procedural complications.17 To overcome these drawbacks, non-invasive strategies were explored. In 2000, Ventureyra proposed transcutaneous vagus nerve stimulation as a non-invasive paradigm shift.18 Subsequent technical advances validated the auricular vagal branch as a feasible stimulation target.19,20 taVNS delivers electrical stimuli to superficial auricular vagal fibers, balancing promising efficacy and favourable safety. This innovation eliminates surgical risks, lowers medical expenses, and improves patient access and adherence. Furthermore, it establishes a novel non-pharmacological modality for chronic pain treatment (Figure 1). The transcutaneous auricular vagus nerve stimulation device. Since its introduction in 2000, taVNS has gained increasing attention as a non-pharmacological option for chronic pain. Accumulating evidence over the past two decades supports its analgesic efficacy across diverse conditions. Nevertheless, existing systematic reviews and meta-analyses have largely focused on discrete pain types, such as migraine, fibromyalgia, or osteoarthritis. Consequently, its broader applicability across the chronic pain spectrum remains insufficiently characterized.21–23 Although the literature has grown rapidly in recent years, findings remain fragmented across different conditions and study designs. Therefore, the present narrative review integrates evidence across various chronic pain conditions to systematically examine the mechanisms of action and clinical utility of taVNS. By synthesizing mechanistic, technical, and clinical evidence, this review aims to provide clinicians and researchers with a comprehensive framework for understanding taVNS as a versatile non-pharmacological approach to pain management. Furthermore, by highlighting gaps in current evidence, this review aims to inform future research and facilitate the clinical translation of taVNS. Methods This narrative review was conducted following the Scale for the Assessment of Narrative Review Articles (SANRA) guidelines to ensure methodological rigor and clinical relevance. To enhance methodological transparency, we provide below a detailed textual breakdown of the literature screening steps. Although PRISMA reporting is not mandatory for narrative reviews, this review covers five distinct major chronic pain subtypes and synthesizes highly heterogeneous primary studies. Without transparent screening metrics, readers cannot verify whether the literature selection process was systematic or detect potential selective citation bias. Literature Search Strategy A systematic literature search was performed across PubMed/MEDLINE, Embase, Web of Science, and the Cochrane Library from database inception to June 2026. All manuscripts published in 2026 and identified through the search were manually double-screened to capture the most up-to-date preclinical and clinical evidence. Search terms combined Medical Subject Headings (MeSH) and free-text keywords related to three core domains: (i) transcutaneous auricular vagus nerve stimulation; (ii) chronic pain conditions; and (iii) neuromodulation-related outcomes. The detailed PubMed/MEDLINE search strategy is presented below. Analogous search strings were adapted for the other databases according to their respective syntax and controlled vocabulary rules. PubMed/MEDLINE search strategy: “transcutaneous auricular vagus nerve stimulation” [Title/Abstract] OR “taVNS” [Title/Abstract] OR “auricular vagus nerve stimulation” [Title/Abstract] OR “transcutaneous vagus nerve stimulation” [Title/Abstract] OR “auricular branch vagus nerve stimulation” [Title/Abstract] “transcutaneous auricular vagus nerve stimulation” [Title/Abstract] OR “taVNS” [Title/Abstract] OR “auricular vagus nerve stimulation” [Title/Abstract] OR “transcutaneous vagus nerve stimulation” [Title/Abstract] OR “auricular branch vagus nerve stimulation” [Title/Abstract] “chronic pain” [MeSH Terms] OR “chronic pain” [Title/Abstract] OR “neuropathic pain” [Title/Abstract] OR “fibromyalgia” [Title/Abstract] OR “osteoarthritis” [Title/Abstract] OR “migraine” [Title/Abstract] OR “low back pain” [Title/Abstract] OR “temporomandibular disorders” [Title/Abstract] OR “primary dysmenorrhea” [Title/Abstract] OR “irritable bowel syndrome” [Title/Abstract] OR “functional dyspepsia” [Title/Abstract] OR “inflammatory bowel disease” [Title/Abstract] OR “rheumatoid arthritis” [Title/Abstract] OR “systemic lupus erythematosus” [Title/Abstract] OR “systemic sclerosis” [Title/Abstract] OR “trigeminal neuralgia” [Title/Abstract] OR “diabetic neuropathy” [Title/Abstract] OR “radiation-related neuropathic pain” [Title/Abstract] OR “long COVID” [Title/Abstract] “chronic pain” [MeSH Terms] OR “chronic pain” [Title/Abstract] OR “neuropathic pain” [Title/Abstract] OR “fibromyalgia” [Title/Abstract] OR “osteoarthritis” [Title/Abstract] OR “migraine” [Title/Abstract] OR “low back pain” [Title/Abstract] OR “temporomandibular disorders” [Title/Abstract] OR “primary dysmenorrhea” [Title/Abstract] OR “irritable bowel syndrome” [Title/Abstract] OR “functional dyspepsia” [Title/Abstract] OR “inflammatory bowel disease” [Title/Abstract] OR “rheumatoid arthritis” [Title/Abstract] OR “systemic lupus erythematosus” [Title/Abstract] OR “systemic sclerosis” [Title/Abstract] OR “trigeminal neuralgia” [Title/Abstract] OR “diabetic neuropathy” [Title/Abstract] OR “radiation-related neuropathic pain” [Title/Abstract] OR “long COVID” [Title/Abstract] “analgesia” [MeSH Terms] OR “analgesia” [Title/Abstract] OR “pain relief” [Title/Abstract] OR “neuroplasticity” [Title/Abstract] OR “neuromodulation” [Title/Abstract] OR “anti-inflammatory” [Title/Abstract] OR “autonomic nervous system” [Title/Abstract] “analgesia” [MeSH Terms] OR “analgesia” [Title/Abstract] OR “pain relief” [Title/Abstract] OR “neuroplasticity” [Title/Abstract] OR “neuromodulation” [Title/Abstract] OR “anti-inflammatory” [Title/Abstract] OR “autonomic nervous system” [Title/Abstract] #1 AND #2 AND #3 #1 AND #2 AND #3 Reference lists of retrieved articles and relevant systematic reviews were manually screened to identify additional eligible studies. Eligibility Criteria Studies were included if they: (1) evaluated taVNS in human chronic pain (duration ≥3 months) or established animal models of chronic pain; (2) reported pain-related clinical or mechanistic outcomes; and (3) were published in peer-reviewed English-language journals. Exclusion criteria included case reports, conference abstracts, editorials, letters, and studies exclusively investigating invasive implantable vagus nerve stimulation. Study Selection and Data Extraction Two authors independently screened titles, abstracts, and full texts, with disagreements resolved through discussion and consensus. Stepwise Literature Screening Volume Description Initial database searches retrieved 1916 records. After eliminating 648 duplicate publications, 1268 unique citations proceeded to title and abstract screening. A total of 1105 articles were excluded as irrelevant at this stage, leaving 163 papers for full-text assessment. After full-text review, 119 articles were excluded for failing to meet the inclusion criteria, yielding a final pool of 44 eligible primary studies on taVNS and chronic pain. Among these 44 studies, 28 are cited in the dedicated section “Clinical Evidence for taVNS in Chronic Pain”. The remaining 16 pain studies are not listed in this clinical subsection, as this review prioritizes thematic narrative integration rather than exhaustive individual study enumeration. The full reference list of this narrative review contains 104 citations in total, encompassing mechanistic animal experiments, neuroimaging trials, methodological reviews, and the 28 clinical/preclinical pain studies. The following data were extracted from each included study: first author, publication year, study design, sample size, pain phenotype, stimulation parameters (target site, frequency, pulse width, intensity, and duration), primary and secondary outcome measures, follow-up duration, and reported adverse events. Evidence Synthesis and Quality Assessment Given the substantial heterogeneity across studies in terms of pain conditions, stimulation protocols, outcome instruments, and study designs, a formal meta-analysis was deemed inappropriate. Instead, a narrative synthesis was employed, with evidence organized thematically by pain category into neurobiological mechanisms and clinical applications by pain category. Consistent with the narrative review approach, formal risk-of-bias scoring was not performed. Rather, evidence was interpreted with explicit consideration of study design features, sample size, control conditions, and follow-up duration. Specifically, randomized sham-controlled trials were given greater weight in drawing clinical conclusions when such trials were available. taVNS Neurobiology: From Peripheral Stimulation to Central Modulation The vagus nerve (cranial nerve X) is one of the longest and most important mixed nerves in the human body; approximately 80% of its fibers are afferent.24–26 It serves as a core bidirectional pathway connecting the central nervous system with visceral organs. Through a complex neuroendocrine-immune network, it regulates physiological homeostasis and modulates inflammation, emotion, and nociception. The auricular branch of the vagus nerve (Arnold’s nerve) is the only cutaneous branch of this nerve. Its nerve endings are highly concentrated in the auricular concha region, particularly the cymba conchae.24,26,27 This relatively accessible surface anatomical localization renders the auricular concha region a practical target for taVNS, although interindividual variability in nerve distribution remains a consideration. Recent clinical and preclinical studies have unveiled the multifaceted analgesic mechanisms of taVNS. These mechanisms operate through four principal pathways: neurotransmitter modulation, anti-inflammatory immunoregulation, autonomic functional reorganization, and central network neuroplasticity28 (Figure 2). Neurobiological mechanisms of transcutaneous auricular vagus nerve stimulation. ↑ upregulation/increase, ↓ downregulation/decrease, → activation/promotion. Diagrams were created with BioRender.com. Anatomical Substrate: Activation of the NTS–PAG Descending Pathway Auricular vagal afferents project to the nucleus tractus solitarius (NTS) in the brainstem; efferent projections from the NTS connect to the PAG, the core node of the endogenous analgesic neural circuit.29 This auricle–jugular ganglion–NTS–PAG circuit provides the anatomical foundation for the neurotransmitter-mediated mechanisms detailed below. Multi-Target Modulation of Neurotransmitter Systems taVNS modulates nociceptive processing through the regulation of multiple neurotransmitter systems. The specific mechanisms encompass the following four aspects: Enhancement of GABAergic Inhibitory Transmission As the primary inhibitory neurotransmitter in the central nervous system, GABA suppresses neuronal hyperexcitability within nociceptive circuits, particularly at the level of key pain-modulatory nuclei. taVNS enhances GABAergic transmission in these nuclei, including the periaqueductal gray and thalamic nuclei critically involved in nociceptive gating, thereby attenuating nociceptive signal transmission.30 Activation of the Locus Coeruleus-Norepinephrine System The locus coeruleus constitutes the principal noradrenergic nucleus in the central nervous system, which sends ascending projections to the hypothalamus, amygdala, and prefrontal cortex as well as descending projections to the spinal cord dorsal horn. taVNS activates this system, promoting norepinephrine release and thereby enhancing descending noradrenergic inhibition of ascending nociceptive transmission.31 Facilitation of Descending Serotonergic Modulation taVNS enhances the activity of serotonergic neurons in the nucleus raphe magnus and adjacent raphe obscurus/pallidus, thereby facilitating descending serotonergic inhibition of nociceptive transmission at the spinal dorsal horn.32 Activation of the Endogenous Opioid System taVNS promotes the release of endogenous opioid peptides, including β-endorphin, enkephalin, and dynorphin, which act primarily on μ- and δ-opioid receptors to mediate analgesic effects.26 While these neurotransmitter systems likely constitute the neurochemical substrate of taVNS-mediated analgesia, their specific regulatory pathways, potential synergistic interactions, and clinical translation remain to be validated through rigorous randomized controlled trials. Immunoregulation via the Cholinergic Anti-Inflammatory Pathway The cholinergic anti-inflammatory pathway (CAP) is a vagus nerve-mediated neuro-immunomodulatory mechanism that prevents tissue damage and cell death by inhibiting excessive cytokine release.33,34 Specifically, vagal efferent signaling ultimately leads to acetylcholine (ACh) release, which binds to α7 nicotinic acetylcholine receptors (α7nAChR) on macrophages. This binding inhibits nuclear factor-κB signaling and subsequently downregulates the expression of pro-inflammatory cytokines, including tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β), and interleukin-6.35,36 This mechanism attenuates tissue damage caused by excessive cytokine release and alleviates inflammatory pain in various conditions, including rheumatoid arthritis (RA) and inflammatory bowel disease (IBD). taVNS may engage this pathway indirectly through afferent-efferent vagal reflex arcs.37 While systemic anti-inflammatory effects represent a plausible outcome, the underlying neural circuitry remains incompletely mapped, and direct evidence for such coupling in humans is currently lacking. Reconstruction and Balance of Autonomic Nervous Function Chronic pain is frequently accompanied by autonomic dysfunction, characterized by increased sympathetic activity and decreased parasympathetic tone. taVNS enhances parasympathetic output and suppresses sympathetic hyperactivity through transcutaneous stimulation of vagal afferent fibers.38 Studies have shown that taVNS significantly improves parasympathetic heart rate variability (HRV) indices, including RMSSD and HF power, while reducing the LF/HF ratio, a marker of sympathovagal balance. Microneurography recordings demonstrate that burst incidence of muscle sympathetic nerve activity is significantly reduced during taVNS.39 These acute findings suggest that taVNS transiently restores autonomic balance, thereby alleviating pain-related autonomic dysregulation and associated stress responses. This sympathovagal rebalancing offers a mechanistic rationale for taVNS application in sympathetically maintained pain conditions, such as complex regional pain syndrome. Neuroplasticity and Remodeling of Central Networks Long-term taVNS treatment induces structural and functional neuroplastic changes at multiple levels: Molecular Level taVNS upregulates neurotrophic factors, with brain-derived neurotrophic factor (BDNF) being the most extensively studied and strongly implicated in pain modulation.40 BDNF is abundantly expressed in key supraspinal pain-modulatory regions, including the prefrontal cortex, amygdala, hippocampus, and periaqueductal gray (PAG).41,42 At the spinal level, BDNF is present in the superficial dorsal horn (laminae I–II), where it modulates nociceptive transmission.43 By enhancing BDNF-mediated synaptic plasticity across these supraspinal and spinal circuits, taVNS may promote adaptive neural reorganization and contribute to analgesia. Circuit Level Functional magnetic resonance imaging (fMRI) evidence from multiple studies demonstrates that taVNS differentially remodels the functional connectivity of thalamus-related brain circuits. Specifically, it markedly strengthens the functional coupling between motor-associated thalamic subregions and the rostral anterior cingulate cortex/medial prefrontal cortex (rACC/mPFC). This enhanced coupling facilitates the central descending pathways for anti-inflammation and endogenous analgesia. Meanwhile, it weakens connectivity between occipital-linked thalamic subregions and the postcentral gyrus as well as the precuneus, thus alleviating central sensitization induced by overexcitation of the somatosensory cortex. These findings illustrate the neuroimaging mechanism at the central level underlying the dual anti-inflammatory and analgesic effects of taVNS at the neural circuit level.44–46 Network Level taVNS modulates brain function at the macro network level by regulating the dynamic balance of large-scale intrinsic brain networks, mainly the default mode network (DMN) and salience network (SN).47 The DMN controls resting-state spontaneous activity and self-referential cognition, while the SN mediates the detection of salient stimuli and transitions between brain states. Aberrant activation and disrupted functional coordination between the two networks commonly lead to cognitive and neural dysfunction. As a non-invasive neuromodulation technique, taVNS can normalize aberrant activation and functional connectivity within the DMN and SN, restore balanced inter-network dynamics, and thereby stabilize overall brain network function. In summary, taVNS modulates multiple regulatory levels, thereby engaging a multi-dimensional pain modulation network. At the neurochemical level, it enhances GABAergic, noradrenergic, serotonergic, and endogenous opioid transmission to suppress nociceptive signaling. At the immunological level, it engages the CAP to downregulate key pro-inflammatory cytokines (eg, TNF-α, IL-1β, IL-6). At the neuroplastic level, it upregulates BDNF-mediated synaptic plasticity across supraspinal and spinal circuits and remodels thalamocortical connectivity. At the network level, it regulates the dynamic balance between the DMN and SN. This multi-target, integrative mechanism underlies its distinct therapeutic advantages in chronic pain, particularly in conditions characterized by neuroinflammation, autonomic dysregulation, and central sensitization. Clinical Evidence for taVNS in Chronic Pain This review first analyzes the core neurobiological pathways that mediate taVNS-induced analgesia. Building on these mechanistic foundations, we systematically synthesize and compare existing clinical evidence regarding taVNS-mediated pain-modulating efficacy across five major categories of chronic pain: musculoskeletal disorders, chronic primary pain, neuropathic pain, visceral pain, and autoimmune diseases (Table 1). Table 1Clinical Evidence for taVNS in Chronic PainAuthor (Year)DiseaseSample Size (Group Allocation)Stimulation ParametersMain ResultAdverse EventsLimitationsElsehrawy et al, 202548Knee osteoarthritisN=68 (Active n=34; Sham n=34).Electrode Site: Left cymba conchaFrequency: 25 HzPulse Width: 250 μsIntensity: 0.25–2.0 mA Duration: 30 min/session, 3×/week for 12 weeksVAS: ↓ significantly vs sham, sustained 4 wksPPT: ↑ at knees and remote elbow (↓ central sensitization)PD-Q/DN4: ↓ significantly, sustainedCSI: ↓ significantlySkin irritation (88%), tingling (81%). Mild.Single-blind; 4-wk FU only; no SOC arm; 94.2% female; baseline KL grade imbalanceAoyagi et al, 202549Knee osteoarthritisN=30 (Single-arm pilot).Electrode Site: Cymba conchaFrequency: 25 HzPulse Width: 250 μsIntensity: up to 15 mADuration: 60 min, single sessionKnee Pain (NRS): ↓ by 1.27 immediately and 1.87 at 15 min (p<0.001); 37% achieved MCIDPPT: No changeTS: No changeCPM: ↑ immediatelyHF (parasympathetic): ↑ at 15 minMomentary nausea (3.3%), dizziness (3.3%). Mild, self-resolved.No control/sham arm; single session; uncontrolled HRV confounders; proprietary HRV algorithm; limited generalizabilityTavares-Figueiredo et al, 202450Chronic low back painN=30 (Open-label).Electrode Site: Left cymba conchaFrequency: 25 HzPulse Width: 50 μsIntensity: Not fixed; patient-perceived non-painfulDuration: 30 min/session, daily for 3 monthsVAS: ↓ 16.1 mm (1 mo) and 22.5 mm (3 mo); 51.9% MCIDODI, PCS, HADS: ↓ significantly; EQ-5D-5L: ↑HRV: No significant changeMinor/mild AEs in 4 patients (palpitations/hypertension, epistaxis, neck pain, psoriasis flare); 1 fall unrelated.Open-label, no sham; subjective outcomes only; no fixed intensity; ~66% compliance; HRV non-significant; possible placebo effectsDemircioğlu et al, 202451Chronic low back painN=60 (Conventional rehab n=30; taVNS n=30).Electrode Site: Bilateral tragus and conchaFrequency: 10 HzPulse Width: 300 μsIntensity: Patient-comfortableDuration: 25 min/session, 5×/week for 3 weeksVAS: ↓ more in taVNS vs conventionalODI, BDS, PSQI, MST, FTFT: significantly improved in both groupsTrunk flexion strength and endurance: ↑ more in taVNSQuadriceps peak torque (60°/s dominant): ↑ more in conventionalFRT: improved in conventional onlyMild ear redness (n=3), transient headache (n=3); self-resolved.No sham/placebo group; no long-term follow-upPrott et al, 202552Temporomandibular disordersN=20 (taVNS n=10; Sham n=10).Electrode Site: Left cymba conchaFrequency: 25 HzPulse Width: 250 μsIntensity: Adjustable above perception thresholdDuration: 4 h/day for 4 weeksGCPS (pain): ↓ vs sham (small effect, ns)OHIP-G14 (OHRQL): large effect (moderate→mild), nsPHQ-9, GAD-7, PHQ-15, JFLS-20, MVC, MMO: small effects, all nsNo serious AEs.Assessors unblinded; small sample (95% female); short-term only; app connection issues; baseline GCPS imbalance; underpoweredKutlu et al, 202053Fibromyalgia syndromeN=60 (Exercise n=30; Exercise+taVNS n=30).Electrode Site: Bilateral tragus and conchaFrequency: 10 HzPulse Width: <500 μsIntensity: Sensory thresholdDuration: 30 min/session, 5×/week for 4 weeksVAS, FIQ, depression, anxiety, SF-36: significantly improved within both groupsBetween-group post-intervention: no significant differences; taVNS+exercise numerically better but nsNo serious AEs.No blinding; no ANS/HRV assessment; baseline SF-36 imbalance; females only; short-term; underpowered for between-group comparisons; exercise confounderDolcini et al, 202554FibromyalgiaN=18 (Single-arm pilot).Electrode Site: Left tragusFrequency: 1–30 HzPulse Width: 50–250 μsIntensity: 0.1–36 mADuration: 30 min/session, 5×/week for 4 weeksWPI: ↓ at 2 wks and 4 wksrFIQ total/overall impact/symptoms: ↓ at 4 wks; overall impact sustained at 8 wksPSQI: ↓ at 4 wks, returned to baseline at 8 wksVAS pain, SSS, BDNF: no significant change64% no longer met FM criteria post-treatmentMild/transient: asthenia (50%), headache/dizziness (38%), nausea (33%), blurred vision (22%), local paraesthesia/erythema (22%/16%). Resolved with intensity adjustment.No sham/control; small sample; no HRV/ANS assessment; antidepressants may confound BDNF; effects largely lost at 4-wk follow-up; single-centreLommano et al, 202555FibromyalgiaN=25 (Open-label, single-arm).Electrode Site: Left tragusFrequency: 20 HzPulse Width: <250 μsIntensity: Sensory thresholdDuration: 30 min/session, 2×/day for 28 daysCOMPASS-31: ↓ total, orthostatic, vasomotor, pupillomotorFIQR, PDQ, CSI-9: ↓, PCS: ↓ trendMild/transient: tinnitus (n=3, resolved), nausea/dizziness/distress (n=1), pain worsening with insomnia (n=1).Open-label, no control/sham; small sample (all female); short-term; no objective HRV; real-life observational designZhang et al, 202144Migraine (without aura) N=70 (taVNS n=35; Sham n=35).Electrode Site: Left cymba conchaFrequency: 1 HzPulse Width: 0.2 msIntensity: 1.5–5 mADuration: 30 min/session, 12 sessions over 4 weeksMigraine days: ↓ 2.5 vs 0.7 daysPain intensity: ↓ 17.4 vs 4.1 Attack times: ↓ 1.5 vs +0.4MSQ, SAS, SDS: improved in both groups, no significant interactionfMRI: ↑ motor thalamus–rACC/mPFC connectivity; ↓ occipital thalamus–postcentral gyrus connectivityNo serious AEs.Single-blind; no blinding assessment; small sample; only migraine without aura.Cao et al, 202156Migraine (without aura)N=24 (Single-blind, crossover).Electrode Site: Left concha (cymba and cavum)Frequency: 1 Hz vs 20 Hz (crossover, ≥7-day washout)Pulse Width: ~0.2 msIntensity: Max nonpainful tolerable sensationDuration: 8 min/session × 2 sessions1 Hz: ↑ vlPAG FC with bilateral MCC, precuneus, MFG, cuneus (vs pre-resting); 20 Hz: ns.1 Hz > 20 Hz: greater ↑ vlPAG FC with MCC, PCC, insula, ACC.No serious AEs.Small sample; sex imbalance (3M); single-blind; interictal only; no clinical efficacy/FC-clinical correlation; no HR monitoring; episodic migraine without aura only.Badran et al, 202245Long COVIDN=12 randomized (Active n=6; Sham n=6).Electrode Site: Left cymba conchae + tragusFrequency: 25 HzPulse Width: 500 μsIntensity: 2× perceptual threshold (~0.5–0.6 mA)Duration: 1 h/session, 2×/day, 6 days/week for 4 wks (2-wk blinded + 2-wk open-label)Symptom burden: 9 predefined symptoms ↓ in active group (46%→31%→38% vs sham 66%→68%→38%)Mental fatigue (MFS): largest ↓ with full 4-wk active.2 cases mild skin irritation (ear redness).Small pilot, underpowered; heterogeneous Long COVID symptoms; exploratory design; taVNS dosing not yet established; 2 h/day potentially burdensome; no long-term FU dataZhang et al, 202457Trigeminal neuralgiaN=60 (TN n=30; HC n=30).Electrode Site: Left auricular region Frequency: TN ~22.5 Hz; HC ~25.5 HzPulse Width: 200 μs Intensity: Patient-adjusted, nonpainful toleranceDuration: 30 min/session, 1 sessionCPM effect: ↑ at 30 min (active > sham) in both TN and HC; TN > HC at 30 min PPT: ↑ after active taVNS in both groups No effect at 15 minNo serious AEs.No vagal biomarker; wrist CPM not trigeminal; short-term; single-center; medication confoundZuo et al, 202558Radiotherapy-related neuropathic painN=116 (taVNS n=58; Sham n=58).Electrode Site: Left conchaFrequency: 30 Hz Pulse Width: 300 μsIntensity: Max tolerable without painDuration: 30 min/session, 2×/day for 7 daysNRS: ↓ 2.43 vs 1.12 at d7; ≥50% responders: 48.3% vs 24.1%;BPI severity: ↓; Fatigue/QoL: improved No pain diff at 16-wk FU.Mild skin irritation (8.6%).Predominantly nasopharyngeal CA; short 7-day tx; no long-term benefit at FU; unblinded physiotherapists; parameters not optimized; ITT/PP bias risks.Liu et al, 202459Constipation-predominant irritable bowel syndromeN=40 (ITT; Active n=20; Sham n=20).Electrode Site: Auricular cymba concha and cavity of conchaFrequency: 25 HzPulse Width: 0.5 ms Intensity: 0–2 mADuration: 30 min/session, 1×/day for 4 wksVAS/IBS-SSS/HAMA/HAMD ↓ vs sham; SBM/CSBM/BSFS ↑; HRV: vagal ↑/sympathetic ↓; RAIR ↓/rectal sensation ↑; serum Ach ↑/NO ↓; Bifidobacterium/SCFAs ↑; tryptophan metabolites ↓Ear rash, dizziness, palpitation, diarrhea (5% each). Mild, manageable.Single-blind, single-center, small N; no long-term FU; no colon-transit stratification; ITT dropouts not imputed; multi-omics lacked sham arm.Shi et al, 202160Constipation-predominant irritable bowel syndromeN=42 (taVNS n=21; sham n=21).Electrode Site: Bilateral auricular cymba conchaFrequency: 25 HzPulse Width: 0.5 ms Intensity: 0–2 mADuration: 30 min/session, 2×/day for 4 wksCSBMs/week: ↑; VAS pain: ↓ 64%; BSFS/IBS-QOL: improved; IBS-SSS/anxiety/depression: ↓; RAIR/rectal sensation: improved; TNF-α/IL-6/5-HT: ↓; vagal HF: ↑No serious AEs.Single-center, small sample; no colon transit classification; no long-term follow-up.Shi X et al, 202461Functional dyspepsiaN=300 (V10 n=101; V25 n=99; Sham n=100).Electrode Site: Left tragus Frequency: 10 Hz, 25 Hz (V25) Pulse Width: 500 μs Intensity: 0.5–1.5 mA Duration: 30 min/session, 2×/day for 4 wksResponse rate: 81.2% (V10) and 75.8% (V25) vs 47% sham (both P<0.001); adequate relief: 85.1% and 80.8% vs 67%; FDSD/GSRS/QoL/anxiety/depression: improved vs sham; efficacy lasted to wk 8–12; no diff V10 vs V25Mild (1–3%): tinnitus, palpitation, insomnia, painCombined standard therapy; short FU; small subgroupsZhu Y et al, 202162Functional dyspepsiaN=36 (Active n=18; Sham n=18)Electrode Site: Bilateral auricular cymba conchaFrequency: 25 Hz Pulse Width: 0.5 ms Intensity: 0.5–1.5 mA Duration: 30 min/session, 2×/day for 2 wksAcute: gastric accommodation ↑, fullness ↓, %NSW ↑, vagal ↑/sympathetic ↓ vs sham; Chronic (2-wk): DSS ↓, anxiety/depression ↓, gastric accommodation ↑, %NSW ↑, vagal ↑ vs shamMild tinnitus/insomnia (5.6%).Non-severe FD only; 2-wk treatment without FU; single-center/small N; sham on arm (not ear); no gastric emptying dataWang et al, 202363Primary DysmenorrheaN=60 (Active n=30; Sham n=30).Electrode Site: Left cymba concha Frequency: 1 Hz Pulse Width: 200 μs Intensity: Tolerance-based Duration: 30 min/session, 1×/day for 10 daysCMSS ↓ vs sham (50.6% post-tx, 50.9% at 1-cycle FU); 83.3% responders post-tx, 86.7% at FU; improvement correlated with FPQ/PASS/PCSNo serious AEs.Single-blind; pilot/small N; sham only 45s; no physiological measures; short FUYıldız Kızkın et al, 202564Primary DysmenorrheaN=34 (Active n=17; Sham n=17)Electrode Site: Bilateral tragus and turbinate of ear Frequency: 10 Hz Pulse Width: 300 μs Intensity: Comfortable tolerance Duration: 5 min/session, 1×/day for 5 consecutive daysNRS ↓ d1–d3 vs sham; PPT ↑ abdominal/lumbar; MSQ/FEMD/HAM-A ↓ vs sham; PANAS-positive ↑; 6MWT ↑None reportedSmall N; short FU; analgesic exclusion limits generalizability; no PPT ICC; 6MWT unvalidated in PD; inflated effect sizesWang et al, 202665Primary DysmenorrheaN=60 (Active n=30; Sham n=30)Electrode Site: Left cymba conchaFrequency: 1 Hz Pulse Width: 200 μs Intensity: Tolerance-based Duration: 30 min/session 1×/day for 10 daysCMSS ↓ sustained ≥6 mo; baseline EEG microstates abnormal (class C ↓, class D ↑, B↔C transitions ↓); brain score mediated 2nd-cycle efficacyNo serious AEs.Single-modality EEG; microstate n=29; sham only 1st cycle; no post-tx EEG; single-centerSahn et al, 202366Pediatric IBD (CD/UC)N=22 (10 CD, 12 UC).Electrode Site: Cymba conchae (left ear) Frequency: 20 Hz Pulse Width: 300 μs Intensity: Titrated to tolerance Duration: 5 min once daily (Phase 1, crossover); 5 min twice daily (Phase 2, wk 4–16)Clinical remission at wk 16: 50% CD, 33% UC (active baseline); FC ≥50% reduction: 64.7% (11/17); UC FC ↓ 81%, CD ↓ 51%; Phase 1: taVNS ↓ FC vs sham ↑ FC; Anxiety scores improved; HRV showed vagal tone recalibration.Minor skin irritation from probe pressure (1). C. diff recurrence (1, unrelated). Disease flare requiring med change (2, uncertain).No sham control after week 4; no endoscopic assessment; single-blind; small sampleParikh et al, 202667Rheumatoid arthritisN=12 (single-arm feasibility).Electrode Site: Cymba concha (left ear) Frequency: 30 Hz Pulse Width: 1 ms Intensity: 1–5 mA Duration: 20 min/session, 14 sessions over 2 weeksPre-post: large ↓ pain interference (d=0.83), psychological distress (d=1.66), grip pain (d=1.12); medium ↓ resting pain and TNF-α. No meaningful change in PPT, CRP, HAQ-DI.14 TEAEs in 50%; Only 2 intervention-related: ear rash (resolved) and vertigo.Single-arm, no sham; Small N; 2-week duration; LOCF imputation; missing HRV; IL-1β/IL-17A excluded.Marsal et al, 202168Rheumatoid arthritisN=30 enrolledElectrode Site: Auricular branch of vagus nerve (external ear) Frequency: 20kHzPulse Width: NRIntensity: Median 3.0 mA, range 1.7–4.0 mA Duration: Up to 30 min/day for 12 weeksDAS28-CRP ↓1.4; pain score ↓30.3; sleep quality improved; ACR20/50/70: 53%/33%/17%; HAQ-DI ↓0.5. CRP unchanged.4 AEs, none serious (1 skin abrasion)Open-label, no-sham trial; 12-week duration only; biologic-naive cohort, no combination with biologics/tsDMARDs; auricular vagal activation unconfirmed; low MRI completion rate; limited cytokine detection.Baker et al, 202369Rheumatoid arthritisN=113 randomized (52 unilateral active, 9 bilateral active, 52 sham)Electrode Site: External ear via custom-fitted earbuds Frequency: 20kHzPulse Width: NRIntensity: median 2.6 mA Duration: 15 min once daily for 12 weeksACR20 at wk 12: 25.0% vs 26.9%; pain score: −16.67 vs −8.33; DAS28-CRP: −0.95 vs −0.66; HAQ-DI: −0.19 vs −0.02. No sig improvement in ACR50/70, CDAI, SDAI vs sham.No serious AEs.b/tsDMARD-naïve only; no imaging/cytokine assessment; fixed dosing not exploredAranow et al, 202070Systemic lupus erythematosusN=18 (12 taVNS, 6 sham)Electrode Site: Left ear cymba concha Frequency: 30 Hz Pulse Width: 300 μs Intensity: Maximum tolerated without pain Duration: 5 min/day for 4 consecutive daysPain ↓5.00 vs +0.10 at d5, maintained at d12; 83.3% vs 16.7% clinically meaningful. Fatigue ↑11.0 vs 0.0 at d5, maintained at d12; 10/12 vs 0/6 meaningful. Tender/swollen joints: median 100% reduction vs ~5–9%. Substance P ↓ at d5. No sig changes in cytokines, CRP, or kynurenine pathway.No serious AEs.Small pilot; only 4 days; most biomarkers unchanged; durability unknown.Bellocchi et al, 202371Systemic sclerosisN=32 randomizedElectrode Site: Cymba conchae of the left ear Frequency: 25 Hz Pulse Width: 250 μs Intensity: 0.2–5 mADuration: 4 non-consecutive hours/day for 4 weeksNRS pain: ↓27.7% vs ↓7.7%; 13/21 vs 5/21 achieved clinically meaningful reduction (≥2 points). IL-6 ↓ with tVNS vs active control. No sig differences in PROMIS-29 pain interference, HAQ, UCLA GIT 2.0, PHQ-9, PSQI, or HRV indices.One excluded pre-randomization due to ear discomfort; no AEs during trial.Small sample; high dropout; mostly lcSSc with long disease duratio; no gold standard stimulation pattern; 4 weeks may be too short.Notes: Symbols: ↓, decrease; ↑, increase; ns, not significant.Abbreviations: 5-HT, 5-hydroxytryptamine (serotonin); 6MWT, 6-Minute Walk Test; ACh, acetylcholine; ACC, anterior cingulate cortex; ACR20/50/70, American College of Rheumatology 20/50/70% response criteria; AEs, adverse events; ANS, autonomic nervous system; BDNF, brain-derived neurotrophic factor; bDMARD, biological disease-modifying antirheumatic drug; BDS, Beck Depression Scale; BPI, Brief Pain Inventory; BSFS, Bristol Stool Form Scale; CA, carcinoma; CD, Crohn’s disease; CDAI, Clinical Disease Activity Index; CI, confidence interval; CMSS, Chinese Menstrual Symptom Scale; COMPASS-31, Composite Autonomic Symptom Score-31; CPM, conditioned pain modulation; CRP, C-reactive protein; CSBM, complete spontaneous bowel movement; CSI, Central Sensitization Inventory; DAS28-CRP, Disease Activity Score-28 with C-reactive protein; DN4, Douleur Neuropathique 4 questions; DSS, Dyspepsia Symptom Severity; EEG, electroencephalography; EQ-5D-5L, EuroQol 5-Dimension 5-Level; FC, fecal calprotectin; FDSD, Functional Dyspepsia Symptom Diary; FEMD, study-specific scale (see original publication); FIQR, Revised Fibromyalgia Impact Questionnaire; FIQ, Fibromyalgia Impact Questionnaire; FPQ, Fear of Pain Questionnaire; FRT, Functional Reach Test; FTFT, Five Times Floor Transfer; GAD-7, Generalized Anxiety Disorder-7; GCPS, Graded Chronic Pain Scale; GSRS, Gastrointestinal Symptom Rating Scale; HADS, Hospital Anxiety and Depression Scale; HAM-A, Hamilton Anxiety Rating Scale; HAMD, Hamilton Depression Rating Scale; HAQ-DI, Health Assessment Questionnaire–Disability Index; HC, healthy controls; HF, high frequency (HRV component); HR, heart rate; HRV, heart rate variability; IBD, inflammatory bowel disease; IBS, irritable bowel syndrome; IBS-QOL, Irritable Bowel Syndrome Quality of Life; IBS-SSS, Irritable Bowel Syndrome–Symptom Severity Score; IL-6, interleukin-6; IL-10, interleukin-10; IL-1β, interleukin-1 beta; IL-17A, interleukin-17A; ITT, intention-to-treat; JFLS-20, Jaw Functional Limitation Scale-20; KL, Kellgren-Lawrence; lcSSc, limited cutaneous systemic sclerosis; LOCF, last observation carried forward; MCID, minimal clinically important difference; MCC, midcingulate cortex; MFS, Mental Fatigue Scale; MFG, middle frontal gyrus; min, minutes; MMO, maximum mouth opening; mo, months; MSQ, study-specific scale (see original publication); MST, study-specific scale (see original publication); MVC, maximum voluntary contraction; mPFC, medial prefrontal cortex; N, number; NRS, numeric rating scale; NR, not reported; ns, not significant; NO, nitric oxide; %NSW, percentage of normal satiety weight; ODI, Oswestry Disability Index; OHIP-G14, Oral Health Impact Profile-14; OHRQL, oral health-related quality of life; PANAS, Positive and Negative Affect Schedule; PCC, posterior cingulate cortex; PD, primary dysmenorrhea; PD-Q, Pain Detect Questionnaire; PDQ, Pain Detect Questionnaire; PHQ-9, Patient Health Questionnaire-9; PHQ-15, Patient Health Questionnaire-15; PPT, pressure pain threshold; PP, per-protocol; PROMIS-29, Patient-Reported Outcomes Measurement Information System-29; PSQI, Pittsburgh Sleep Quality Index; QoL, quality of life; RA, rheumatoid arthritis; RAIR, rectoanal inhibitory reflex; rACC, rostral anterior cingulate cortex; RCT, randomized controlled trial; rFIQ, Revised Fibromyalgia Impact Questionnaire; SBM, spontaneous bowel movement; SCFAs, short-chain fatty acids; SDAI, Simplified Disease Activity Index; SEM, standard error of the mean; SF-36, 36-Item Short Form Health Survey; SLE, systemic lupus erythematosus; SOC, standard of care; SSS, Symptom Severity Scale; taVNS, transcutaneous auricular vagus nerve stimulation; TEAEs, treatment-emergent adverse events; TNF-α, tumor necrosis factor-alpha; TN, trigeminal neuralgia; TS, temporal summation; tsDMARD, targeted synthetic disease-modifying antirheumatic drug; UC, ulcerative colitis; UCLA GIT 2.0, University of California, Los Angeles Gastrointestinal Tract 2.0; VAS, visual analog scale; vlPAG, ventrolateral periaqueductal gray; vs, versus; wk(s), week(s). Clinical Evidence for taVNS in Chronic Pain Notes: Symbols: ↓, decrease; ↑, increase; ns, not significant. Abbreviations: 5-HT, 5-hydroxytryptamine (serotonin); 6MWT, 6-Minute Walk Test; ACh, acetylcholine; ACC, anterior cingulate cortex; ACR20/50/70, American College of Rheumatology 20/50/70% response criteria; AEs, adverse events; ANS, autonomic nervous system; BDNF, brain-derived neurotrophic factor; bDMARD, biological disease-modifying antirheumatic drug; BDS, Beck Depression Scale; BPI, Brief Pain Inventory; BSFS, Bristol Stool Form Scale; CA, carcinoma; CD, Crohn’s disease; CDAI, Clinical Disease Activity Index; CI, confidence interval; CMSS, Chinese Menstrual Symptom Scale; COMPASS-31, Composite Autonomic Symptom Score-31; CPM, conditioned pain modulation; CRP, C-reactive protein; CSBM, complete spontaneous bowel movement; CSI, Central Sensitization Inventory; DAS28-CRP, Disease Activity Score-28 with C-reactive protein; DN4, Douleur Neuropathique 4 questions; DSS, Dyspepsia Symptom Severity; EEG, electroencephalography; EQ-5D-5L, EuroQol 5-Dimension 5-Level; FC, fecal calprotectin; FDSD, Functional Dyspepsia Symptom Diary; FEMD, study-specific scale (see original publication); FIQR, Revised Fibromyalgia Impact Questionnaire; FIQ, Fibromyalgia Impact Questionnaire; FPQ, Fear of Pain Questionnaire; FRT, Functional Reach Test; FTFT, Five Times Floor Transfer; GAD-7, Generalized Anxiety Disorder-7; GCPS, Graded Chronic Pain Scale; GSRS, Gastrointestinal Symptom Rating Scale; HADS, Hospital Anxiety and Depression Scale; HAM-A, Hamilton Anxiety Rating Scale; HAMD, Hamilton Depression Rating Scale; HAQ-DI, Health Assessment Questionnaire–Disability Index; HC, healthy controls; HF, high frequency (HRV component); HR, heart rate; HRV, heart rate variability; IBD, inflammatory bowel disease; IBS, irritable bowel syndrome; IBS-QOL, Irritable Bowel Syndrome Quality of Life; IBS-SSS, Irritable Bowel Syndrome–Symptom Severity Score; IL-6, interleukin-6; IL-10, interleukin-10; IL-1β, interleukin-1 beta; IL-17A, interleukin-17A; ITT, intention-to-treat; JFLS-20, Jaw Functional Limitation Scale-20; KL, Kellgren-Lawrence; lcSSc, limited cutaneous systemic sclerosis; LOCF, last observation carried forward; MCID, minimal clinically important difference; MCC, midcingulate cortex; MFS, Mental Fatigue Scale; MFG, middle frontal gyrus; min, minutes; MMO, maximum mouth opening; mo, months; MSQ, study-specific scale (see original publication); MST, study-specific scale (see original publication); MVC, maximum voluntary contraction; mPFC, medial prefrontal cortex; N, number; NRS, numeric rating scale; NR, not reported; ns, not significant; NO, nitric oxide; %NSW, percentage of normal satiety weight; ODI, Oswestry Disability Index; OHIP-G14, Oral Health Impact Profile-14; OHRQL, oral health-related quality of life; PANAS, Positive and Negative Affect Schedule; PCC, posterior cingulate cortex; PD, primary dysmenorrhea; PD-Q, Pain Detect Questionnaire; PDQ, Pain Detect Questionnaire; PHQ-9, Patient Health Questionnaire-9; PHQ-15, Patient Health Questionnaire-15; PPT, pressure pain threshold; PP, per-protocol; PROMIS-29, Patient-Reported Outcomes Measurement Information System-29; PSQI, Pittsburgh Sleep Quality Index; QoL, quality of life; RA, rheumatoid arthritis; RAIR, rectoanal inhibitory reflex; rACC, rostral anterior cingulate cortex; RCT, randomized controlled trial; rFIQ, Revised Fibromyalgia Impact Questionnaire; SBM, spontaneous bowel movement; SCFAs, short-chain fatty acids; SDAI, Simplified Disease Activity Index; SEM, standard error of the mean; SF-36, 36-Item Short Form Health Survey; SLE, systemic lupus erythematosus; SOC, standard of care; SSS, Symptom Severity Scale; taVNS, transcutaneous auricular vagus nerve stimulation; TEAEs, treatment-emergent adverse events; TNF-α, tumor necrosis factor-alpha; TN, trigeminal neuralgia; TS, temporal summation; tsDMARD, targeted synthetic disease-modifying antirheumatic drug; UC, ulcerative colitis; UCLA GIT 2.0, University of California, Los Angeles Gastrointestinal Tract 2.0; VAS, visual analog scale; vlPAG, ventrolateral periaqueductal gray; vs, versus; wk(s), week(s). Musculoskeletal Disorders Knee Osteoarthritis Knee osteoarthritis (KOA) represents a prevalent age-related degenerative joint disorder characterized by persistent nociceptive and nociplastic pain, which imposes heavy individual and socioeconomic burdens worldwide.72 As a novel non-invasive neuromodulation approach, taVNS has garnered growing interest for its therapeutic potential in this condition. Elsehrawy et al48 reported a 12-week single-blind sham-controlled RCT enrolling 68 bilateral KOA patients, randomized to active taVNS (n=34) and sham earlobe stimulation (n=34). Active stimulation was delivered on the left cymba concha (30 min, 3 days/week; 25 Hz; 250 μs; 0.25–2.0 mA), with three assessment timepoints: baseline, post-intervention, and 4-week treatment-free follow-up. The taVNS group showed sustained, superior improvements in bilateral/remote elbow pressure pain threshold (PPT), nociceptive pain (VAS), neuropathic pain markers (PD-Q, DN4), central sensitization inventory (CSI), physical function (KOOS, mobility tests), and mood (HADS). Sham stimulation only yielded transient VAS relief that vanished at follow-up, with no lasting changes in sensitization, function and emotional indices. The intervention demonstrated favourable safety: only mild local tingling and skin irritation were reported, and no severe adverse events were recorded. Complementing this long-term RCT evidence, Aoyagi et al49 conducted an uncontrolled single-session pilot trial recruiting 30 KOA participants to receive a single 60-minute taVNS session (25 Hz, 250 μs, 25 Hz, 250 μs, patient-titrated tolerable intensity). Evaluations were conducted at baseline, immediately post-stimulation, and 15 min after intervention. Single-session taVNS induced modest, short-lived knee pain reduction and elevated high-frequency heart rate variability (HF-HRV), reflecting transient parasympathetic enhancement alongside improved conditioned pain modulation (CPM). However, PPT and temporal summation (TS) remained unaltered post-intervention. Notably, this open-label pilot lacked a sham comparator and only captured short-term outcomes without prolonged follow-up, limiting its causal interpretability relative to Elsehrawy’s controlled long-term trial. The two trials jointly suggest an exposure-dependent analgesic effect in KOA, with distinct mechanistic patterns distinguishing acute from repeated stimulation. Single-session taVNS generates temporary pain relief via transient autonomic and descending inhibitory pathways, yet stable shifts in PPT and TS were absent in Aoyagi’s cohort. Crucially, the absence of altered sensitization markers following a single taVNS session does not inherently indicate an inability to reverse established central sensitization. Isolated brief stimulation lacks sufficient exposure duration to induce persistent neural remodelling, representing an intrinsic limitation of single-session study designs rather than evidence of treatment inefficacy. By contrast, 12 weeks of repeated taVNS in Elsehrawy’s RCT produced durable elevations in multi-site PPT that persisted four weeks after treatment cessation, accompanied by sustained functional and psychological gains. This stark divergence implies that recurring taVNS exposure may be required for long-term central neuroplasticity and stable desensitization. Collectively, these data support a biphasic analgesic model: acute taVNS delivers fleeting symptomatic relief mediated by autonomic signalling, while repeated regimens are required to reverse chronic central sensitization. Future direct comparative sham-controlled trials contrasting single versus multi-week taVNS paradigms are warranted to establish minimum effective doses. Chronic Low Back Pain Chronic low back pain (CLBP) ranks among the top global contributors to years lived with disability. Population-level Global Burden of Disease projections73,74 estimate that around 39% of adults worldwide experience low back pain annually, with over 800 million cases anticipated by 2050 due to population aging. Conventional management including pharmacology, physical modalities, and psychological care often yields suboptimal long-term relief for many patients. This shortfall creates an unmet clinical demand for non-invasive neuromodulation such as taVNS. Isabelle Tavares-Figueiredo et al50 conducted a 3-month open-label single-arm pilot trial recruiting 30 patients with refractory nonspecific CLBP (baseline VAS ≥40/100). Participants received daily 30-minute left cymba concha taVNS (25 Hz; 50 μs; tolerable pain-free current). Assessments were conducted at baseline, month 1, and month 3. Progressive, statistically significant pain reduction was captured: VAS fell by 16.1 mm at month 1 and 22.5 mm at month 3 (both P<0.001). Composite improvements in physical disability, pain catastrophizing, life quality and mood-related distress were consistently observed across follow-up visits. Furthermore, 51.9% of participants reached the ≥20 mm VAS threshold for clinically meaningful pain relief. Moderate treatment adherence was documented, and only minor transient adverse events emerged without serious safety incidents. These findings confirm that taVNS is well tolerated for long-term standalone use. For combined rehabilitation scenarios, Gamze Demircioğlu et al51 implemented a 3-week randomized trial that lacked a dedicated taVNS sham group. They assigned 60 CLBP patients to either taVNS plus home exercise or standard physical modalities (hot pack, TENS, ultrasound) plus identical exercise training. After three weeks of intervention, the taVNS arm demonstrated superior outcomes in pain intensity, trunk muscle strength/endurance and sleep metrics. Distinct therapeutic pathways underpin the intergroup difference: conventional lumbar rehabilitation acts locally on peripheral soft tissue and musculoskeletal structures, whereas taVNS exerts systemic regulatory effects on autonomic tone, central nociceptive circuits, and cholinergic anti-inflammatory cascades via vagal afferent signaling. Together, these mechanisms produce complementary analgesic benefits when taVNS is paired with exercise. This supports taVNS as a valuable adjunct for patients unresponsive to routine physical therapy. The two CLBP trials differed substantially in intervention duration (3 weeks versus 3 months). This divergence suggests that the durability of therapeutic effects may depend on treatment exposure: Tavares-Figueiredo’s 3-month daily taVNS protocol achieved sustained symptom improvements throughout the intervention period, whereas Demircioğlu’s shorter 3-week protocol produced only short-term additive benefits when combined with rehabilitation. Notably, neither trial incorporated taVNS sham stimulation as a control arm: Tavares-Figueiredo’s single-arm design lacked any comparative reference, while Demircioğlu’s control group only compared against conventional physical therapy, rather than inactive auricular stimulation. Heterogeneous stimulation protocols, inconsistent treatment durations, and the lack of long-term post-intervention follow-up further preclude generalizable conclusions regarding the optimal taVNS dosing regimen and sustained therapeutic effects. Therefore, future standardized, sham-controlled randomized clinical trials with extended observation periods are needed to confirm effective treatment protocols and identify the clinical characteristics of responders to taVNS for CLBP treatment. Temporomandibular Disorders Temporomandibular disorders (TMD) represent a complex group of musculoskeletal conditions involving the temporomandibular joint and/or masticatory muscles, with core symptoms including periarticular pain, limited jaw mobility, and joint sounds.75 Pharmacological, physical, and behavioural interventions often provide only partial symptom relief. This limitation has driven research into taVNS as an alternative neuromodulatory strategy for TMD-related pain. Lea S. Prott et al52 performed an 8-week single-blinded pilot RCT allocating 20 participants with chronic TMD to active taVNS (n=10) and sham control (n=10). Active stimulation was applied over the left tragus with fixed settings: 25 Hz, 250 µs, 28s on/32 s off duty cycle, administered 4 hours daily. Evaluations were conducted at baseline, week 4, and week 8. These included self-reported pain, emotional, and somatic questionnaires (GCPS, PHQ-9, GAD-7, PHQ-14), as well as objective electromyographic and jaw kinematic assessments. Although the taVNS group showed substantial within-subject improvements in oral health-related quality of life (OHIP-G14), all other endpoints—including pain, psychological distress, and masticatory dysfunction—showed only numerical, non-significant reductions compared with the sham arm. The trial demonstrated satisfactory feasibility: recruitment targets were fully met, and participant attrition was only 10%. No stimulation-associated adverse events were recorded throughout the intervention period. The neutral primary pain findings of this small feasibility cohort can be largely attributed to insufficient statistical power resulting from a limited sample size, alongside the fixed 4-hour daily dosing regimen. Such prolonged daily dosing is unprecedented in taVNS literature for taVNS trials, and whether it is necessary or optimal remains unclear. This uncertainty raises the possibility that the stimulation parameters used were suboptimal. Although numerical trends in pain improvement were observed for taVNS, the absence of significant between-group differences does not rule out its pain-modulating capacity. Chronic Primary Pain Fibromyalgia Fibromyalgia (FM) is a refractory chronic pain syndrome characterised by widespread musculoskeletal discomfort, driven by autonomic dysregulation, neuroinflammation and central sensitization.76,77 Conventional pharmacotherapy frequently produces side effects and yields only modest, inconsistent relief. First-line non-pharmacological strategies, meanwhile, suffer from poor long-term adherence and variable responses. Against this backdrop, taVNS has been proposed as a potential non-invasive neuromodulatory candidate for FM management. Kutlu et al53 reported significant within-group improvements in pain, anxiety, depression, and quality of life with bilateral taVNS combined with home exercise. However, these changes were not accompanied by between-group superiority. The combined intervention showed no statistically significant superiority over exercise alone across core symptom domains (all P > 0.05). Differences were limited to selected SF-36 subdomains. Notably, the between-group difference in pain intensity (VAS) narrowly missed statistical significance (P = 0.084). No additive benefit was observed for depression or anxiety scores. This lack of incremental efficacy suggests that taVNS, at least as an adjunct to exercise, did not confer additional therapeutic value in this trial. Two subsequent open-label pilot trials explored taVNS in FM but lacked sham controls. Dolcini et al54 delivered 4 weeks of daily taVNS to 18 patients. Although global disease burden and sleep improved—most participants no longer met diagnostic criteria—core pain intensity remained unchanged. Moreover, benefits showed near-complete relapse within four weeks of stimulation cessation. Lommano et al55 applied 28-day twice-daily taVNS in 25 patients. They observed improvements in autonomic function and sensitization metrics. However, pain catastrophizing was unchanged (P = 0.070). The uncontrolled designs preclude attribution of these domain-specific effects to active treatment rather than non-specific factors. The evidence hierarchy is clear. The single available RCT53 reported neutral between-group results. The positive signals from open-label pilots54,55 lack confirmatory value. The most parsimonious interpretation is that taVNS has not demonstrated consistent efficacy beyond placebo for FM core symptoms. In the absence of sham comparators, within-group improvements in uncontrolled studies may reflect placebo effects, regression to the mean, or spontaneous fluctuation. The rapid post-treatment relapse further supports this interpretation. Consequently, current evidence does not support robust efficacy of taVNS in FM; however, it would be premature to conclude that taVNS is fundamentally ineffective. Moreover, several methodological limitations may have contributed to the neutral finding. These include unoptimized stimulation parameters, insufficient power for subgroup detection, and fixed co-delivery with exercise rather than standalone administration. Without dose-response data or prespecified stratification analyses, these technical constraints preclude definitive conclusions and caution against overinterpreting the primary negative finding. Rather than establishing inefficacy, the available data highlight the uncertainty surrounding taVNS efficacy in FM. Beyond methodologic considerations, the pathophysiology of FM may render the syndrome resistant to single-target peripheral vagal neuromodulation. Unlike peripheral nociceptive conditions, FM is characterised by profound central sensitization and distributed affective-cognitive network dysfunction.55,76,77 The clinical data reveal a striking dissociation. Favourable effects were observed on brainstem-mediated markers, such as autonomic function and sleep. In contrast, negligible impact was seen on pain intensity, neurotrophic signalling, and cognitive-emotional processing. This suggests that taVNS may modulate lower-order vagal-brainstem circuits yet cannot reconfigure the higher-order distributed networks that drive FM. In this light, the neutral RCT results may reflect a potential efficacy ceiling rather than an absolute boundary. Single-target peripheral neuromodulation may be insufficient for syndromes dominated by top-down network dysfunction. Nevertheless, this “network mismatch” hypothesis represents a post-hoc mechanistic conjecture that has not been validated by any prespecified biomarker or stratification analysis. In summary, current evidence is insufficient to support taVNS as a routine treatment for FM. The highest-quality trial was negative, and the encouraging signals from uncontrolled studies are readily explained by non-specific factors. Although uncontrolled pilot studies observed domain-specific improvements, these signals require validation in blinded trials before clinical recommendations are warranted. If pursued, future research must first establish whether taVNS produces stable, clinically meaningful effects beyond placebo in any FM subgroup. Investigators should explore parameter optimization or biomarker-guided stratification only after establishing such effects. Migraine Migraine is a prevalent, disabling neurological disorder marked by recurrent headache episodes and chronic symptoms that severely impair patients’ daily functioning and quality of life.78 Conventional preventive and abortive medications often trigger adverse reactions and provide inadequate symptom control for refractory patients. This has prompted the exploration of effective alternative therapeutic approaches. Zhang et al44 conducted a single-blind sham-controlled randomized trial recruiting 70 patients with episodic migraine without aura. All participants received 4 weeks of 1 Hz taVNS applied to the left cymba concha, and 59 subjects completed the full treatment protocol. Clinical outcomes revealed significant reductions in monthly migraine days and headache intensity in the active stimulation group relative to sham controls. Corresponding resting-state fMRI data identified consistent thalamocortical circuit remodeling after low-frequency taVNS. Connectivity between motor thalamic subdivisions and the antinociceptive rostral anterior cingulate cortex (rACC)/medial prefrontal cortex (mPFC) was enhanced, whereas pathological coupling between occipital thalamic nuclei and somatosensory regions (postcentral gyrus, precuneus) was attenuated. Cao et al56 supplemented the clinical findings with a single-blind crossover fMRI trial that enrolled 24 episodic migraine patients to dissect frequency-dependent neuromodulatory mechanisms. All participants underwent independent 1 Hz and 20 Hz taVNS scanning blocks, separated by a minimum 7-day washout period. Compared with 20 Hz stimulation, 1 Hz taVNS markedly increased functional connectivity between the ventrolateral periaqueductal gray (vlPAG) and the midcingulate cortex, anterior cingulate cortex, precuneus, and insula. The vlPAG serves as the central hub of the descending pain modulation system. Correlation analyses further confirmed that baseline vlPAG–midcingulate functional connectivity was negatively associated with patients’ monthly migraine attack load. This finding provides a neural basis for the superior clinical efficacy of low-frequency taVNS reported in previous cohorts. The two complementary trials provide consistent clinical and neuroimaging evidence that confirms the therapeutic value of 1 Hz taVNS for episodic migraine. The inverse association between resting-state vlPAG–midcingulate connectivity and migraine severity suggests an inherent impairment of descending inhibitory pathways in migraine patients. Low-frequency auricular neuromodulation may selectively reverse this impairment. Even so, prominent methodological limitations compromise the external validity of these mechanistic observations. Zhang’s trial employed only single masking: clinicians administering taVNS could differentiate active stimulation from sham stimulation, thereby introducing performance bias. In the crossover trial conducted by Cao, the small sample size undermines statistical reliability. In addition, the minimal 7-day washout interval may be insufficient to eliminate persistent brain network alterations from prior stimulation, potentially producing carryover effects that confound cross-frequency fMRI comparisons. Long COVID Long COVID is defined as persistent, recurrent, or new symptoms that continue for at least three months after acute SARS-CoV-2 infection and cannot be attributed to an alternative diagnosis.79 Chronic pain ranks among its most prevalent sequelae, with population prevalence ranging from 19.6% to 33.6% across cohort studies.80,81 Its pathogenesis is hypothesized to stem from viral neurotropic injury and persistent pro-inflammatory immune cascades.82 Given that taVNS modulates neuroimmune homeostasis and activates the vagal cholinergic anti-inflammatory pathway (CAP), it represents a biologically plausible treatment for chronic pain in long COVID. Preliminary evidence is limited to a small pilot trial by Badran et al,83 which enrolled only 12 participants with multisystem long COVID symptoms, predominantly persistent headache. Given the limited sample size, any conclusions from this work should be interpreted with extreme caution. This 4-week, two-phase, double-blind, sham-controlled trial used standardized at-home taVNS. Participants received 1 hour of unilateral stimulation twice daily (25 Hz, 500 μs pulse width), with the anode on the left cymba concha and the cathode on the tragus. The sequential trial design consisted of a 2-week blinded active/sham allocation phase, followed by a mandatory open-label active taVNS phase for all subjects in the subsequent two weeks. Continuous monitoring of overall symptom burden suggested that therapeutic effects depended on the cumulative duration of neuromodulation. Among participants who received four consecutive weeks of active taVNS, self-reported symptom prevalence fell steadily from 46% at baseline to 31% at the 2-week blinded timepoint, and further to 23% at treatment completion. In contrast, those who received only two weeks of active stimulation after an initial sham period showed a prominent rebound in symptoms, with prevalence rising to 56% at follow-up. Although these divergent symptom trajectories suggest that prolonged taVNS may sustain relief of long COVID-related discomfort, multiple inherent limitations preclude robust causal inference. First, as a pilot study, it was designed primarily to assess feasibility and safety, not to detect between-group differences in clinical efficacy. Second, the sample of only 12 individuals is far too small to draw definitive clinical inferences. These divergent trajectories merely hint at a possible duration-response relationship; they do not constitute confirmatory evidence. Collectively, these pilot results provide modest proof-of-concept support for taVNS in treating long COVID-associated headache and pain. Large, adequately powered randomized controlled trials are therefore indispensable to confirm these exploratory findings. Chronic Neuropathic Pain Diabetic Peripheral Neuropathy Diabetic peripheral neuropathy (DPN) is one of the most common chronic complications of diabetes. In its early stages, it often manifests as distal limb pain, burning sensations, and paresthesia, all of which severely affect patients’ quality of life.84,85 Current treatment strategies primarily include glycemic control, neurotrophic support, and symptomatic analgesia; however, overall therapeutic efficacy remains suboptimal. Some patients respond poorly to conventional medications, and long-term pharmacotherapy carries risks of adverse effects. The preclinical study by Wang et al86 delivers valuable mechanistic insights that advance our understanding of taVNS analgesia against DPN. The study adopted standardized auricular concha taVNS protocols (2/15 Hz alternating frequency, 2 mA intensity, 30 min daily). To distinguish pineal from extrapineal melatonin sources, the authors additionally established pinealectomized cohorts. Behavioral assays confirmed persistent alleviation of thermal hyperalgesia following repeated taVNS treatment. Biochemical and immunohistological analyses further revealed that taVNS elevated systemic plasma melatonin levels and that it upregulated MT1 melatonin receptor expression in the amygdala. Notably, substantial analgesic effects and elevated melatonin levels persisted in pinealectomized rodents. These findings exclude the pineal gland as the sole source of melatonin underlying taVNS-mediated pain relief. This preclinical evidence suggests a novel cascade in which extrapineal melatonin mediates the analgesic effect of auricular vagal neuromodulation in diabetic neuropathy. These rodent-derived findings provide a solid mechanistic basis for translating taVNS to DPN treatment. They also suggest potential therapeutic applications in other metabolic neuropathic pain conditions. Nevertheless, critical translational limitations restrict the generalizability of these observations. All supporting evidence derives exclusively from diabetic rat models. No human clinical trials have yet examined whether taVNS-induced melatonin signaling produces analgesic effects in patients with DPN. Large-scale controlled human studies are therefore warranted to verify whether the neurohumoral pathways identified in preclinical work can be recapitulated in patients with DPN. Trigeminal Neuralgia Trigeminal neuralgia (TN) is a classic neuropathic pain syndrome characterized by sudden, brief, severe electric shock-like pain in the unilateral trigeminal distribution. It is frequently triggered by innocuous daily stimuli such as chewing, speaking, or light touch, severely compromising patients’ quality of life.87–89 The core pathological mechanism involves neurovascular compression or structural damage at the trigeminal nerve root. This leads to ectopic neural discharge and central sensitization, which in turn produce hyperalgesia and allodynia.90 Zhang et al57 conducted a single-center crossover trial in 30 patients with classic TN and 30 age- and sex-matched healthy controls to evaluate the modulatory effects of taVNS on endogenous pain inhibition. For both active and sham arms, identical stimulation schedules were implemented: each session lasted 30 minutes with a fixed pulse width of 200 µs, and stimulation intensity was individually adjusted to the maximum level that remained non-painful. Active taVNS was delivered to the left auricular concha innervated by the auricular vagal branch, whereas sham stimulation was applied to the vagus-free earlobe. Researchers measured PPT and CPM at baseline and at 15 and 30 minutes post-intervention. Single-session active taVNS significantly elevated PPT and enhanced CPM capacity in both cohorts. The strengthening effect on endogenous descending pain inhibition was more pronounced in TN patients, indicating that taVNS may specifically restore or normalize defective pain regulatory circuits in trigeminal neuropathic pain. The intervention demonstrated excellent tolerability, with no severe adverse events reported throughout testing. Collectively, these findings suggest two complementary analgesic pathways: taVNS potentiates central descending pain inhibitory signaling and restores defective endogenous pain regulation in TN patients. These results support taVNS as a well-tolerated, non-invasive adjunctive strategy for trigeminal neuropathic pain. Nevertheless, multiple methodological limitations constrain the interpretability of these acute observations. The trial only captured transient pain-modulating changes after a single taVNS session, without exploring sustained effects of repeated stimulation. CPM testing was conducted on the wrist rather than trigeminal-innervated facial areas, weakening the assessment’s regional specificity for TN-related pain dysregulation. Furthermore, no physiological biomarkers were collected to objectively confirm effective vagal activation. Notably, all participants remained on standard analgesic regimens throughout the experiment, which may alter baseline nociceptive processing and confound the assessment of taVNS-specific therapeutic effects. Radiation-Related Neuropathic Pain Radiation-related neuropathic pain (RRNP) is a delayed adverse effect of radiotherapy. Its latency period ranges widely from 6 months to 20 years, severely impacting patients’ quality of life. Epidemiological data indicate that approximately 20–40% of cancer survivors are affected by this condition, which is particularly common following radiotherapy for head and neck, thoracic, and pelvic malignancies.91 Zuo et al58 conducted a multicenter, double-blind, sham-controlled Phase II trial to investigate whether taVNS can serve as an adjuvant analgesic for moderate-to-severe refractory RRNP following radiotherapy for head and neck malignancies. All participants maintained a fixed pregabalin baseline dose throughout the study period. The active group received two 30-minute taVNS sessions daily for seven consecutive days, with electrodes on the left auricular concha. The sham group used identical electrode placement and device settings, but the stimulation current was gradually reduced to zero, producing no active neuromodulation. After one week of combined taVNS-pregabalin therapy, the intervention group achieved significantly greater pain reduction than sham controls, with higher proportions of participants attaining ≥30% and ≥50% clinically meaningful pain relief. Quality of life also improved in both physical and psychological domains. The intervention demonstrated favorable tolerability; mild transient electrode-site skin irritation represented the sole device-associated adverse event, and no severe treatment-related adverse events were recorded. However, the between-group difference in analgesia disappeared during the 16-week follow-up, suggesting that short-term taVNS alone does not produce sustained pain relief after discontinuation. This trial provides preliminary evidence that taVNS may serve as a fast-acting, non-pharmacological adjunct for the early management of radiotherapy-induced neuropathic pain. However, several methodological limitations restrict the generalizability of its conclusions. The short stimulation period precludes assessment of sustained long-term effects. Moreover, the predominance of nasopharyngeal carcinoma patients limits generalizability to other head and neck cancer subtypes. Additionally, the fixed stimulation regimen precluded exploration of optimized parameters for neuropathic pain, and potential unblinding of treating therapists may have introduced performance bias into the results. Despite these caveats, the data suggest that adjunct taVNS provides rapid pain relief when combined with standard pregabalin, offering a safe and minimally invasive supplementary option for pain control in head and neck cancer survivors undergoing radiotherapy. Chronic Visceral Pain Irritable Bowel Syndrome Irritable bowel syndrome (IBS) is a common chronic functional gastrointestinal disorder characterized by abdominal pain, accompanied by altered bowel habits and stool consistency.92 Given its chronic recurrent course, complex pathogenesis, and frequent psychological comorbidities, conventional symptomatic treatments often fail to achieve sustained symptom relief. Consequently, overall therapeutic efficacy remains limited and patient outcomes are often unsatisfactory. Consequently, recent research has increasingly examined the clinical value of taVNS for IBS, and emerging evidence supports its therapeutic potential in this functional gastrointestinal disorder. Liu et al59 conducted a single-centre randomized controlled trial that enrolled 42 patients with constipation-predominant irritable bowel syndrome (IBS-C); 20 were allocated to taVNS and 22 to sham stimulation. Active stimulation was delivered bilaterally to the auricular cymba concha, while sham electrodes were placed on the non-vagal earlobe and antihelix regions. A standardized 4-week taVNS intervention (twice daily, 25 Hz, 0.5 ms pulse width, 0–2 mA) substantially relieved core IBS-C manifestations, including abdominal discomfort, defecatory dysfunction, and comorbid anxiety and depression. Physiological assessments revealed restoration of rectal sensation and normalization of the rectoanal inhibitory reflex after taVNS. Mechanistic measurements focused on peripheral cholinergic signaling: taVNS upregulated circulating ACh, lowered serum nitric oxide levels, and increased overall vagal tone. These changes collectively rebalanced autonomic function and alleviated visceral hypersensitivity. Shi et al60 completed another randomized controlled trial that enrolled 42 IBS-C patients to assess taVNS efficacy versus cutaneous sham stimulation on the elbow. The same 4-week, twice-daily 25 Hz taVNS protocol was implemented for all active participants. Clinical outcomes showed consistent relief of primary gastrointestinal complaints, such as attenuated abdominal pain and regularised bowel habits, together with meaningful gains in patients’ daily quality of life. The intervention also normalized rectal sensory deficits and restored the abnormal rectoanal inhibitory reflex—two core physiological disturbances in IBS-C. Mechanistic data revealed multi-target regulatory effects of taVNS: it restrained systemic inflammatory cascades, modulated peripheral serotonin metabolism and boosted vagal outflow. Correlation analyses further confirmed that elevated vagal activity was closely associated with improved gastrointestinal symptoms. These findings support vagally mediated neuromodulation as a core therapeutic mechanism in IBS-C. These two complementary clinical studies support the therapeutic value of taVNS for IBS, particularly for constipation-dominant subtypes. Both studies suggest that taVNS exerts dual regulatory effects by improving subjective symptoms and correcting objective physiological dysfunction. Mechanistically, the convergent findings—elevated vagal activity, improved peripheral biochemical profiles, and suppressed inflammation—further elucidate the gut–brain axis regulatory mechanism underlying taVNS. These results highlight taVNS as a multi-target, non-invasive neuromodulation strategy for IBS treatment, and provide preliminary clinical evidence for its wider application in functional gastrointestinal disorders. Nevertheless, the current evidence is mainly confined to IBS-C populations, and further research is needed to clarify the generalizability of taVNS and to optimize intervention protocols across different IBS subtypes. Functional Dyspepsia Functional dyspepsia (FD) is a highly prevalent functional gastrointestinal disorder characterized by chronic or recurrent epigastric pain or discomfort (eg, postprandial fullness and early satiety), diagnosed after exclusion of organic lesions by endoscopy and other examinations.93 Current treatment for FD primarily targets symptomatic relief; however, some patients respond poorly and relapse frequently, prompting growing research interest in taVNS for FD. Shi et al61 performed a multicenter, triple-arm randomized controlled trial recruiting 300 Rome IV-confirmed FD participants, who were equally allocated to 10 Hz taVNS, 25 Hz taVNS, and earlobe sham stimulation groups. All subjects completed a standardized 4-week self-administered intervention: 30-minute stimulation sessions twice daily (fixed pulse width 500 μs; current individually titrated from 0.5 to 1.5 mA) delivered to the left tragus. After the full treatment cycle, both low- and high-frequency taVNS groups showed significantly higher clinical response and adequate relief rates than sham controls, accompanied by marked reductions in epigastric pain, bloating, and postprandial discomfort. No significant difference in overall therapeutic efficacy was observed between the 10 Hz and 25 Hz protocols. Longitudinal follow-up at week 8 and week 12 further demonstrated sustained symptom improvements without obvious rebound of dyspeptic pain, confirming the lasting clinical benefits of taVNS. The intervention carried a favorable safety profile, with only mild transient adverse events including tinnitus and palpitations reported at low incidence across all three cohorts. A separate randomized controlled trial by Zhu et al62 further corroborated the potent pain-modulating effect of taVNS in FD. The 36 enrolled FD participants received 2-week bilateral cymba concha taVNS or arm sham stimulation twice daily. Relative to sham controls, taVNS significantly attenuated epigastric pain, alongside parallel improvements in anxiety and depressive symptoms that commonly exacerbate visceral nociception. Objective physiological examinations via electrogastrogram (EGG) and HRV validated its multi-target regulatory effects: taVNS restored impaired gastric accommodation and normalized gastric slow wave rhythm, while elevating vagal efferent activity to rebalance sympathovagal tone. Such autonomic remodeling interrupts the self-perpetuating loop of visceral hypersensitivity and negative emotional arousal, and enhances central descending pain inhibition to mitigate FD pain. Data from these two controlled trials collectively confirm that taVNS represents a well-tolerated, minimally invasive neuromodulation modality for treating visceral pain associated with FD. taVNS triggers rapid relief of the characteristic epigastric discomfort, sustains therapeutic benefits throughout long-term follow-up, and simultaneously alleviates gastrointestinal and psychological comorbidities that exacerbate pain perception. Physiological biomarkers further clarify its core anti-nociceptive cascade: taVNS enhances systemic vagal tone, attenuates visceral hypersensitivity, and augments central descending pain inhibition, thereby targeting the key pathogenic mechanisms of FD pain. Notably, Shi’s multicenter FD trial found equivalent clinical efficacy between the 10 Hz and 25 Hz stimulation regimens. This finding diverges from the migraine-focused observations of Cao et al,56 which demonstrated superior pain relief with 1 Hz taVNS compared with 20 Hz stimulation. This inconsistency across different diseases indicates that optimal frequency regimens for taVNS are phenotype-specific. However, this inference should be interpreted with caution, as the two trials differed in disease mechanisms, stimulation sites (tragus versus cymba concha), and outcome measures; confirmation awaits future head-to-head comparisons within the same disease using identical protocols. These differential effects are mediated by distinct nociceptive, autonomic, and gut-brain signaling pathways underlying visceral gastrointestinal pain versus cephalic neuropathic pain. The comparable efficacy of low and high frequencies in FD also grants clinicians flexibility to adjust stimulation parameters according to individual patient comfort thresholds during routine treatment. Primary Dysmenorrhea Primary dysmenorrhea (PD) is one of the most common gynecological disorders among women of reproductive age, clinically characterized by recurrent cramping pain in the lower abdomen before and during menstruation, often with radiation to the lumbosacral region and nonspecific systemic symptoms.94,95 This condition significantly impairs patients’ quality of life and is frequently associated with depressed mood and central sensitization.96 A clinical interventional trial by Wang et al63 to evaluate the therapeutic benefits of taVNS. Participants received daily 30-min 1 Hz left cymba concha taVNS for 10 consecutive days. The results demonstrated that taVNS effectively alleviated symptoms quantified by the Cox Menstrual Symptom Scale (CMSS), while no meaningful shifts were detected in Menstrual Distress Questionnaire (MDQ) scores. Magnitudes of CMSS relief correlated positively with baseline pain-related psychological metrics (FPQ, PASS, PCS). This implies that taVNS mitigates menstrual discomfort primarily by modulating pain-associated affective processing. A double-blind, sham-controlled randomized trial64 used a distinct 10 Hz bilateral tragus stimulation protocol (5 min daily for five premenstrual days) to further evaluate the efficacy of taVNS. Compared with zero-current sham stimulation, active taVNS yielded robust reductions in menstrual pain, elevated PPT, and relieved concomitant anxiety, yet produced no measurable gains in exercise tolerance. Such divergent effects reveal that taVNS specifically targets pain and mood disturbances, without consistently improving physical performance. Beyond short-term symptom relief, a single-blind sham-controlled trial65 used the same 1 Hz, 10-day taVNS regimen as Wang et al63 and tracked outcomes across six menstrual cycles. It documented durable therapeutic benefits, albeit with marked interindividual response heterogeneity. Resting-state EEG microstate profiles were identified as critical neural biomarkers predicting treatment responsiveness. Baseline menstrual pain severity was associated with altered salience/attention network microstate dynamics, which mediated taVNS efficacy. These findings provide a theoretical foundation for personalized, biomarker-guided taVNS regimens in PD. Taken together, existing clinical studies suggest that taVNS has dual benefits in ameliorating PD-related physical discomfort and psychological disturbances, distinguishing it from conventional single-target analgesic approaches. By regulating central pain-emotion neural pathways, taVNS effectively targets the psychosomatic dimensions of PD: it produces clear therapeutic effects on pain perception and emotional dysfunction, with only minimal effect on physical exercise tolerance. The identified EEG microstate biomarkers also provide objective neurophysiological evidence for individual differences in taVNS responses, laying a foundation for precise clinical application. Despite these promising preliminary findings, current research on this intervention still has notable limitations: most existing studies have relatively small sample sizes and adopt single-center research designs. In addition, the uniform intervention protocols used in current research limit the scope for individualized optimization. Inflammatory Bowel Disease IBD is a chronic relapsing inflammatory disorder of the gastrointestinal tract, primarily encompassing ulcerative colitis (UC) and Crohn’s disease (CD). Clinically, it is characterized by recurrent abdominal pain, diarrhea, and mucosal inflammation, with pathogenesis driven by immune dysregulation under environmental and microbial influences.97,98 Currently, clinical trials investigating vagus nerve stimulation for IBD remain limited; however, preclinical evidence suggests promising therapeutic potential. Using a dextran sulfate sodium (DSS)-induced murine ulcerative colitis model, Hesampour et al99 adopted standardized taVNS regimens (20 Hz, 500 µs pulse width, 10 min daily, bilateral cymba concha). They verified that taVNS markedly reduced disease activity index (DAI) scores and intestinal histological injury. Mechanistically, taVNS downregulates pro-inflammatory mediators in the colon, including IL-1β, TNF-α, and matrix metalloproteinases (MMPs), while upregulating the anti-inflammatory cytokines TGF-β and IL-10 in the colonic mucosa. This neuromodulatory effect concurrently reduces systemic inflammatory burden and inhibits excessive apoptotic death of intestinal epithelial cells. Conversely, taVNS only produced mild symptomatic improvements in DNBS-induced mouse models of Crohn’s disease-like colitis. This finding suggests that the therapeutic effects of taVNS on IBD are subtype-dependent. To extend this translational pipeline from murine models to pediatric populations, Sahn et al66 conducted an open-label crossover pilot trial in pediatric and adolescent patients with mild-to-moderate IBD who were receiving stable baseline medication. All subjects completed two sequential intervention phases: 2 weeks of once-daily 5-min sham stimulation followed by extended twice-daily taVNS at 20 Hz (300 µs pulse width, left cymba concha electrode placement). At baseline, a proportion of participants exhibited elevated fecal calprotectin, which indicated active gut inflammation. Sustained taVNS treatment induced partial clinical remission, and most patients showed marked declines in inflammatory biomarkers. Notably, significant anti-inflammatory effects were observed exclusively in the UC cohort, not in Crohn’s disease patients. Moreover, taVNS alleviated disease-related anxiety and restored vagal tone, as reflected by improved HRV. Only mild, transient auricular skin irritation was reported as an adverse event. Collective evidence from preclinical animal experiments and human clinical trials demonstrates that taVNS exerts distinct therapeutic effects across subtypes of IBD. For UC, which is characterized by superficial mucosal inflammation, vagal cholinergic anti-inflammatory pathway (CAP) activation produces prominent therapeutic benefits. In contrast, auricular neuromodulation delivers only mild symptomatic relief for CD, a disorder marked by transmural immune lesions. Beyond suppressing intestinal inflammatory cascades, taVNS restores gut–brain signaling, alleviates anxiety triggered by disease activity, and rebalances vagal autonomic tone. These effects reveal its multi-target therapeutic potential for the psychological comorbidities frequently encountered in IBD patients. While translational limitations remain in the current body of literature, the existing data still carry meaningful preliminary reference value. Firstly, although the DSS-induced acute colitis model established by Hesampour et al cannot fully replicate the clinical course of human disease, the observed improvements in inflammatory activity and tissue damage support the ability of taVNS to promote intestinal mucosal healing. Secondly, although the open-label crossover trial by Sahn et al exclusively enrolled adolescents, it provides preliminary evidence that taVNS exerts dual efficacy in reducing inflammation and relieving emotional distress. These findings lay a foundation for subsequent research in adult cohorts. Minor methodological flaws are still present in current relevant studies. For example, most animal experiments employ only male rodents, taVNS stimulation protocols vary across trials, and standardized endoscopic criteria for quantifying mucosal healing remain lacking. Optimizing these aspects of experimental design will help consolidate the translational evidence base for this neuromodulatory therapy. Autoimmune Diseases Rheumatoid Arthritis RA is a chronic autoimmune disease characterized by synovial inflammation. Clinically, it manifests as persistent joint pain and swelling; when inadequately controlled, it progresses to joint deformity and irreversible functional impairment.100,101 Although the therapeutic potential of transcutaneous auricular vagus nerve stimulation (taVNS) in RA has drawn attention, existing evidence remains contradictory and inconclusive. Preliminary open-label studies have reported positive signals for taVNS in active RA. Parikh et al67 applied respiratory-gated taVNS (30 Hz, 1 ms pulse width, 20 min daily, left cymba concha) in patients with active RA, observing improvements in pain, grip function, and psychological distress, along with decreased serum TNF-α. Marsal et al68 evaluated 20 kHz unilateral auricular stimulation (30 min daily) in patients with active RA; musculoskeletal ultrasound showed reduced synovial and tenosynovial inflammation, alongside improvements in joint pain and sleep quality. By contrast, Baker et al69 conducted a large-scale, multicenter, double-blind, sham-controlled RCT. The trial enrolled biologic-naïve patients with refractory RA who had inadequate responses to conventional synthetic disease-modifying antirheumatic drugs (csDMARDs). Participants received 20 kHz taVNS (15 min daily) for 12 weeks. The trial failed to meet its primary endpoint (DAS28-CRP), with no significant between-group difference in disease activity. Although the taVNS group showed improvements in Health Assessment Questionnaire-Disability Index (HAQ-DI) scores, these functional gains were not accompanied by reductions in systemic inflammatory markers. These discrepant findings warrant cautious interpretation. The positive outcomes reported by Parikh et al67 and Marsal et al68 derive from open-label, uncontrolled designs; their findings do not rule out placebo effects, expectation bias, or spontaneous disease fluctuation. By contrast, the highest-quality evidence from Baker et al69 indicates that taVNS has no clear efficacy for the core inflammatory pathology or overall disease activity of csDMARD-refractory, biologic-naïve RA. The possibility that it is wholly ineffective in this population cannot be excluded. Several factors may underlie the divergence between studies. Stimulation parameters differed markedly in frequency, duration, and respiratory gating; more importantly, the patient populations were distinct. The open-label studies enrolled patients with active RA, whereas Baker et al focused on csDMARD-refractory, biologic-naïve individuals with treatment-resistant disease. From a mechanistic perspective, RA synovitis is not merely an amplification of systemic inflammation but a pathological process characterized by localized structural immune damage within the joint. In refractory patients, activated fibroblast-like synoviocytes, tissue-resident macrophages, and autocrine cytokine loops constitute a self-sustaining local inflammatory ecosystem whose drivers operate largely independently of peripheral blood cytokine levels. The cholinergic anti-inflammatory pathway (CAP) exerts systemic immunomodulation via splenic nerve-mediated signaling. Yet its effector capacity may be insufficient to override or reverse established joint-level pathology. In other words, localized immune activation within the synovium may have “saturated” the downstream anti-inflammatory capacity of vagally mediated neuromodulation. This suggests that when a pain phenotype is dominated by peripheral structural damage and localized autoimmune loops, systemic neuromodulatory approaches such as taVNS may inherently face an efficacy boundary. Nevertheless, this “efficacy ceiling” hypothesis represents a post-hoc mechanistic explanation and has not been validated by any prespecified subgroup analysis. An alternative possibility is that any analgesic effect of taVNS in RA—if genuine—may operate primarily through central descending inhibitory pathways (noradrenergic and serotonergic) and limbic modulation, rather than through peripheral anti-inflammatory mechanisms. The dissociation between unchanged inflammatory markers and improved HAQ-DI scores in Baker et al’s trial is compatible with this hypothesis but may also represent a chance finding. Currently, no biomarker or prespecified stratification analysis can identify a purported “central sensitization–predominant” responsive subgroup; limiting taVNS benefit to specific pain phenotypes remains a conjecture awaiting validation. In summary, the highest-quality available evidence does not support taVNS as monotherapy or disease-modifying therapy for csDMARD-refractory RA. Whether it has efficacy in earlier-stage or less treatment-refractory disease remains unknown. Future research must first confirm through rigorously designed blinded RCTs whether taVNS confers stable efficacy beyond placebo in any RA subgroup. Only after such confirmation should investigators explore phenotype-stratification strategies. Systemic Lupus Erythematosus Systemic Lupus Erythematosus (SLE) is a classic autoimmune disorder characterized by autoantibody overproduction, immune complex deposition, and multi-organ systemic involvement.102 Musculoskeletal pain and persistent fatigue represent the most prevalent and debilitating clinical manifestations of this disease.103 Aranow et al70 conducted a randomized, double-blind, sham-controlled pilot trial to evaluate taVNS for physical symptoms in SLE. Adults with moderate to severe musculoskeletal pain were enrolled and randomly assigned to two groups at an unbalanced 2:1 ratio. The taVNS group received continuous square-wave stimulation at 30 Hz with a pulse width of 300 μs. Electrodes were placed on the left cymba concha and the retroauricular region, and the intensity was titrated to each patient’s painless tolerance threshold. The intervention was delivered for 5 minutes daily over four consecutive days. The sham group wore inactive devices on the earlobe (a region free of vagus innervation) to maintain blinding. Multidimensional assessments were performed at baseline, day 5, and day 12. These included patient-reported pain and fatigue, physician global and patient global disease assessments, tender and swollen joint counts, and measurements of circulating inflammatory mediators and neuropeptides. Significant alleviations in pain, fatigue and joint swelling were observed in the taVNS group after brief intervention. Mechanistically, taVNS activates the CAP and specifically reduces plasma substance P levels. These findings provide preliminary neuroimmunological evidence for its efficacy in relieving core SLE-related somatic complaints. The stimulation protocol demonstrated satisfactory tolerability, with no ear discomfort, dizziness or other stimulation-related adverse events recorded throughout the trial. Collectively, this pilot study suggests that taVNS may serve as an adjunct intervention for relieving SLE-associated pain, fatigue, and local joint inflammation through modulation of the CAP and substance P signaling. Nevertheless, several critical methodological limitations compromise the reliability of its findings. First, clinical improvements were observed after only four days of intervention. This raises concerns that placebo effects and response bias from frequent questionnaire administration may confound the measurement of subjective outcomes. Second, the study provided only qualitative descriptions of the observed clinical benefits and did not report standardized effect sizes. Consequently, the true magnitude of the intervention’s effects cannot be quantitatively estimated. Third, the small sample size combined with an unbalanced 2:1 allocation design results in insufficient statistical power, restricting the generalizability of its conclusions. Fourth, only substance P exhibited specific downregulation following short-term stimulation, whereas canonical systemic inflammatory markers such as TNF and IL-6 remained unaltered. Long-term evidence is still lacking to confirm the sustained systemic anti-inflammatory effects of taVNS. Systemic Sclerosis Systemic Sclerosis (SSc) is a refractory autoimmune disease predominantly defined by progressive fibrosis of the skin and visceral organs.104 Chronic pain is one of the most disabling symptoms of SSc, which profoundly impairs patients’ long-term health-related quality of life. Conventional symptomatic treatments often provide suboptimal pain relief and fail to target underlying inflammatory dysregulation. Consequently, there remains a substantial unmet need for safer complementary therapeutic strategies. Bellocchi et al71 performed a randomized, patient-blinded crossover trial to evaluate taVNS for persistent pain in SSc patients. A total of 32 SSc participants with chronic pain were enrolled, and 21 completed the full protocol. The cohort predominantly comprised patients with limited cutaneous SSc. An AB/BA crossover design was adopted. All subjects received two 4-week intervention blocks—standard taVNS and low-frequency control—in random sequence, separated by a 4-week washout. Therapeutic taVNS consisted of mild intermittent square-wave stimulation at 25 Hz with a pulse width of 250 μs, delivered to the left cymba concha for a cumulative four hours daily. The control condition followed identical procedures but used weak 1 Hz stimulation to maintain blinding. Multidimensional assessments were implemented at baseline, post-treatment, and post-washout. These included NRS pain scores, pain interference, cardiovascular autonomic function, circulating inflammatory cytokines, and patient-reported quality-of-life scales. Clinical data demonstrated that standard taVNS yielded substantially greater pain reduction than the 1 Hz control. This was accompanied by a marked decline in serum IL-6 concentrations, suggesting that taVNS exerts partial analgesic effects through suppression of pro-inflammatory signalling. Taken together, this crossover trial provides clinical and biomarker-based evidence that taVNS simultaneously alleviates subjective pain and modulates objective inflammatory markers. These findings support its potential adjunct role in the symptomatic management of SSc. Nevertheless, multiple methodological drawbacks limit the generalisability and reliability of the observed positive therapeutic signals. The trial recorded a 34% attrition rate, with only 21 of 32 enrolled participants completing all procedures. The original authors acknowledged this issue but did not perform stratified analyses to determine whether withdrawal resulted from insufficient efficacy, stimulation-related adverse events, or other external factors. The underlying reasons for dropout remain unclear, leaving potential confounding bias unaddressed. Future studies should examine how attrition compromises the robustness of trial outcomes. The study cohort was dominated by patients with limited cutaneous SSc while diffuse cutaneous cases were scarce, so such uneven disease subtype distribution restricts the extrapolation of taVNS efficacy across the full spectrum of SSc populations. The 4-week intervention window induced significant downregulation of IL-6 but no detectable shifts in canonical inflammatory mediators such as TNF-α and IL-1β. Moreover, no meaningful improvements were observed in sleep disturbance, gastrointestinal discomfort, or cardiac autonomic balance. Consequently, the long-term neuromodulatory impact of taVNS remains unconfirmed. Future Perspectives Chronic pain substantially impairs physical function and quality of life, imposing a severe global socioeconomic burden. As a generally well-tolerated, non-invasive neuromodulatory approach, taVNS represents a promising adjunctive strategy in multidisciplinary pain management. Integrated with conventional pharmacotherapy, rehabilitation, and psychological interventions, taVNS may reduce analgesic dependence, thereby minimizing opioid-related risks. Clinical evidence demonstrates that taVNS produces synergistic analgesic effects with rehabilitation training, as demonstrated in patients with CLBP. Mechanistically, taVNS activates the cholinergic anti-inflammatory reflex and restores autonomic homeostasis. It further attenuates nociceptive sensitization and modulates pain-related affective processing. These actions collectively contribute to analgesia within a biopsychosocial framework. Notably, taVNS preferentially targets neuroimmune inflammation and autonomic dysfunction, yet demonstrates limited efficacy against established structural damage or refractory central sensitization. This mechanistic heterogeneity explains variable clinical responses across pain subtypes. Accordingly, this review proposes the following hypothesis: pain conditions driven primarily by neuroinflammation and autonomic dysregulation (eg, IBS, FD, and selected early-stage autoimmune disorders) are most responsive to taVNS. Conversely, for disorders characterized by advanced tissue destruction or intractable central sensitization (eg, advanced RA, FM), taVNS affords only partial symptomatic relief. For these refractory phenotypes, taVNS should be regarded as a supplementary rather than first-line intervention. Consistent with this stratified pattern, inflammatory pain mainly benefits from taVNS-mediated anti-inflammatory regulation. By contrast, nociplastic pain responds predominantly to modulation of central pain circuits. Such heterogeneity necessitates individualized clinical deployment of taVNS as a maintenance, rehabilitative, or anti-inflammatory adjuvant that does not supplant standard therapies. However, inconsistent methodologies and heterogeneous trial findings currently hinder reliable clinical translation. Suboptimal outcomes in FM and advanced RA may reflect non-ideal stimulation parameters, progressive central nociceptive remodeling, or insufficient pretreatment biomarker stratification. Although multimodal taVNS combination strategies are mechanistically plausible, unified operational standards and precision treatment frameworks are currently lacking. Future studies will be required to resolve these translational gaps and validate the proposed stratification hypothesis. Mechanistic Characterization Multimodal neuroimaging and peripheral cytokine and neuropeptide profiling are warranted to delineate the distinct mechanisms of taVNS across inflammatory and nociplastic pain. Mechanistic studies should adopt a three-group design comprising inflammatory pain patients, nociplastic pain patients, and age-matched healthy controls. Future trials would ideally calculate sample sizes based on previously reported effect sizes to secure adequate statistical power. Serial biomarker sampling at baseline, during intervention, and post-treatment, alongside parallel resting-state fMRI and EEG recordings, would enable the identification of subtype-specific brain connectivity signatures. Such a design would also clarify central–peripheral regulatory differences between pain phenotypes. Head-to-Head Parameter Trials Standardized head-to-head trials are required to resolve cross-study inconsistencies in optimal taVNS frequency. Blinded, sham-controlled randomized studies should directly compare 1 Hz, 20 Hz, and 25 Hz protocols, with all other stimulation variables fixed across groups. Rigorous sample size estimation should be based on the minimal clinically important pain difference. Additionally, future work should validate real-time HRV-closed-loop taVNS by comparing its efficacy with conventional fixed-frequency stimulation. Continuous monitoring of autonomic fluctuations during stimulation would allow quantification of neuromodulation–analgesia correlations. Biomarker-Based Patient Stratification Pretreatment biomarkers—including baseline EEG microstates, HRV, and vlPAG functional connectivity—can be used to stratify taVNS responders and non-responders. Prospective trials should complete biomarker assessment prior to randomization and apply predefined cutoffs derived from preliminary cohorts. Stratified block randomization is recommended to balance subgroup distributions across intervention and sham arms. Primary analyses should compare treatment responses between stratified subgroups, while secondary analyses validate the predictive value of multi-dimensional autonomic, electrophysiological, and neuroimaging biomarkers. Dedicated subgroup power calculations are necessary to avoid statistical insufficiency. Standardized Multimodal Regimens Drawing on existing taVNS–exercise synergistic evidence in low back pain, we propose that individualized combination protocols be established for distinct pain subtypes, such as taVNS plus CBT for central sensitization and taVNS plus conventional DMARDs for inflammatory arthritis. Combination trials should adopt a three-arm design, including taVNS monotherapy, standard-care monotherapy, and combined intervention groups. Sample sizes should be powered to detect additive analgesic and anti-inflammatory effects. Sham-controlled, assessor-blinded designs with unified treatment cycles and consistent concomitant therapy dosages are essential for standardized evaluation of synergistic outcomes. Long-Term Longitudinal Observation Extended longitudinal follow-up is essential to characterize sustained taVNS effects and post-treatment relapse trends. Future sham-controlled cohort studies should implement a minimum 12-month follow-up after stimulation cessation. Serial assessments of pain recurrence, HRV, and inflammatory biomarkers at standardized intervals can distinguish persistent neuromodulatory carryover effects from natural disease fluctuations. Sample size estimation should incorporate anticipated participant dropout to ensure robust long-term outcome analysis. Limitations and Concluding Remarks Several inherent limitations constrain the generalizability of the present synthesis. First, this review only includes publications written in English, which may introduce language bias by excluding relevant clinical data reported in non-English-language publications. Second, substantial heterogeneity across the included trials—encompassing divergent stimulation parameters, patient phenotypes, and outcome metrics—precluded quantitative meta-analytic pooling. Consequently, the analysis was restricted to a narrative synthesis. Third, most available clinical evidence comes from small single-center cohorts and lacks large-scale multi-center validation, which limits the reliability of extrapolating findings to broader populations. Additionally, the amount of available evidence varies considerably across pain subtypes. Sufficient mechanistic and clinical data support the efficacy of taVNS in neuroinflammatory conditions such as IBS and early-stage SLE, whereas evidence for advanced structural arthropathy and severe nociplastic pain remains sparse. Finally, the clinical indication stratification hypothesis proposed in this paper remains exploratory and has not been definitively validated. Dedicated stratified prospective trials are still needed to confirm its clinical applicability. taVNS represents a promising non-pharmacological option for multidisciplinary chronic pain management, although its efficacy is highly phenotype-dependent. Future research would ideally resolve widespread methodological inconsistencies, elucidate the biological drivers of variable therapeutic responses, and establish unified stimulation protocols, biomarker-driven patient stratification frameworks, and validated multimodal combination regimens. Refined, individualized taVNS paradigms may support its routine incorporation into evidence-based pain pathways and mitigate the substantial global disease burden imposed by refractory chronic pain. Funding Statement This research received no external funding. 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 All authors declare that they have no conflict of interest References