Current Status and Research Gaps of Neuromodulation in the Rehabilitation of Knee Osteoarthritis: A Scoping Review

1. Introduction

Knee osteoarthritis (KOA) is a degenerative joint disease characterized by articular cartilage degeneration, destruction, hyperplasia, and inflammation. Its typical manifestations include joint pain, swelling, stiffness, crepitus, and limited mobility. It is more prevalent in middle-aged and elderly individuals, with a higher incidence in women, possibly related to altered joint metabolism caused by postmenopausal estrogen decline. The etiology of KOA remains unclear and may be associated with age, obesity, genetics, trauma, and other factors. Studies have shown that the global prevalence of osteoarthritis continues to rise, particularly in rapidly aging countries such as China. KOA not only causes individual functional impairment but also imposes a heavy burden on public health systems[1].

In recent years, in-depth research on KOA pathological mechanisms, biomarkers for disease progression prediction, and pain regulatory pathways has laid a critical theoretical foundation for the development of targeted non-pharmacological therapies. In the field of disease progression and biomarkers, a series of studies by Hunter’s team confirmed that the T2 relaxation time of superficial articular cartilage and its longitudinal changes are significantly correlated with radiographic joint space width (JSW) loss and sustained elevation of WOMAC pain scores, making it an effective predictive biomarker for KOA progression[2].

In the field of KOA risk factors and pain mechanisms, Neogi’s team, based on large-sample epidemiological studies, identified obesity and knee injury as the core modifiable risk factors for KOA onset and progression, laying the foundation for the development of KOA primary prevention strategies[3,4]. Meanwhile, relying on the Multicenter Osteoarthritis Study (MOST), the team systematically evaluated the correlation between somatosensory assessments such as pressure pain threshold (PPT) and knee pain in KOA patients, providing methodological references for quantitative assessment of chronic pain[5,6].

Arendt-Nielsen’s team focused on neuropathic pain mechanisms, quantitative sensory testing (QST), and central sensitization assessment in KOA, confirming that KOA patients exhibit extensive peripheral and central sensitization, with the degree of sensitization positively correlated with disease severity[7,8]. The team also developed a simplified bedside QST kit (QuantiPain) for evaluating pain facilitation and inhibition mechanisms in KOA patients, and verified its test-retest reliability and validity[9].

The International Neuromodulation Society (INS) defines neuromodulation as: “Neuromodulation is a technology that acts directly on nerves. It alters or regulates neural activity by delivering electricity or drugs directly to the target area.” (Website：www.neuromodulation.com/about-neuromodulation） In the field of KOA rehabilitation, such neuromodulation interventions mainly include: peripheral neuromodulation—transcutaneous electrical nerve stimulation (TENS), peripheral nerve stimulation (PNS), neuromuscular electrical stimulation (NMES), interferential current stimulation (IFC), peripheral nerve block (PNB), lumbar sympathetic block (LSB), electroacupuncture (EA), electrical dry needling (EDN); central neuromodulation—transcranial direct current stimulation (tDCS), repetitive transcranial magnetic stimulation (rTMS), spinal cord stimulation (SCS), vagus nerve stimulation (VNS), transcutaneous vagus nerve stimulation (tVNS). These methods are important non-pharmacological treatment modalities with advantages of simplicity, high efficiency, and minimal adverse reactions, providing new insights for clinical management of KOA.

Although the number of clinical studies on neuromodulation for KOA has been increasing annually, no unified consensus has been reached on the optimal protocols, efficacy differences, and underlying mechanisms of various techniques. Therefore, this scoping review summarizes clinical evidence from the past 5 years to provide a reference for optimizing neuromodulation rehabilitation protocols for KOA.

2. Methods

2.1. Search Scope

Studies published in PubMed and Web of Science databases over the past 5 years (May 13, 2021 — May 13, 2026) were retrieved.

2.2. Inclusion Criteria

1. Study subjects: Study subjects: Adult patients (≥18 years old) diagnosed with knee osteoarthritis (KOA, including patellofemoral osteoarthritis);

2. Intervention measures: Primary intervention is any neuromodulation technique (EA, EDN, tDCS, rTMS, TENS, PNS, VNS, tVNS, NMES, IFC, SCS, PNB, LSB) or a combination of two or more neuromodulation techniques;

3. Study types: Original human clinical studies (randomized controlled trials, non-randomized controlled trials, self-controlled before-after trials, etc.).

2.3. Exclusion Criteria

1. Non-clinical studies: Animal experiments, literature reviews, systematic reviews, meta-analyses, guidelines, case reports, letters, surveys;

2. Ineligible study subjects: Patients with knee pain or dysfunction after total knee arthroplasty, patients with only knee pain but no confirmed KOA diagnosis;

3. Ineligible intervention measures: Neuromodulation only as adjuvant therapy (e.g., <1 time per week), or treatments dominated by drugs/surgery/conventional acupuncture/dry needling/tuina/moxibustion/nursing; combined use of neuromodulation with other therapies;

4. Missing data: Study protocols only, retracted/unavailable full texts, studies with unclear intervention protocols.

2.4. Search Strategy

Search formula: (((((knee osteoarthritis OR knee OA) AND (electroacupuncture OR electrical dry needling OR Transcranial Direct Current Stimulation OR Repetitive Transcranial Magnetic Stimulation OR Transcutaneous Electrical Nerve Stimulation OR Peripheral Nerve Stimulation OR Vagus Nerve Stimulation OR Transcutaneous Vagus Nerve Stimulation OR Neuromuscular Electrical Stimulation OR interferential current stimulation OR spinal cord stimulation OR Peripheral Nerve Block OR lumbar sympathetic block OR neurolysis) AND (Humans OR Clinical Study OR Clinical Trial)) NOT (meta)) NOT (protocol))) NOT (review)

2.5. Literature Processing

All retrieved literature was imported into Zotero. Two researchers independently performed literature screening, data extraction, and cross-verification. Any discrepancies were resolved through consensus between the two authors, with no third-party involvement.The PRISMA flow diagram was generated using RevMan 5.4.

2.6. Data Extraction

Two researchers extracted core data from eligible studies, including author, publication year, country, study type, sample size, neuromodulation type, specific protocol (acupoints/parameters/frequency/treatment duration), control measures, and efficacy indicators. Cross-verification was performed after extraction to ensure data accuracy. Qualitative description was adopted, with intervention protocols and efficacy indicators classified and summarized to extract methodological details.

3. Results

3.1. Literature Screening Results

3.1.1. Screening Process

A total of 131 literatures were initially retrieved (117 from PubMed, 14 from Web of Science). After deduplication, retraction check, and manual verification in Zotero, 123 literatures remained; 28 literatures were retained after title and abstract screening, and 23 literatures were finally included after full-text evaluation. The PRISMA flow diagram is shown in Figure 1:

3.1.2. Characteristics of Included Studies

The characteristics of included studies are summarized in Table 1.

3.2. Intervention Protocols and Efficacy Characteristics of Different Neuromodulation Techniques

3.2.1. Peripheral Electrical Stimulation Neuromodulation

3.2.1.1. Transcutaneous Electrical Nerve Stimulation (TENS)

A total of 4 studies were included, with significant heterogeneity in intervention protocols. Electrode positions were mainly divided into two categories: traditional pain areas (medial and lateral knee joint lines) and high nerve density areas (quadriceps tendon, patellar ligament, medial superior knee area); stimulation parameters mostly used 2-100Hz, with single treatment duration of 20-60 minutes. In terms of efficacy, Xu et al.[20] first confirmed that placing electrodes in high nerve density areas of the knee joint resulted in significantly better efficacy than traditional pain areas, with marked improvements in pain, stiffness, and function. With technological advancement, TENS is developing towards wearability, comfort, and home use. Wearable TENS garments developed by Ullah et al.[18] and Emmanuel Maheu[21] demonstrated good home usability and pain relief effects, with better functional improvement than opioid treatment, and no serious adverse reactions (only local erythema).

3.2.1.2. Neuromuscular Electrical Stimulation (NMES)

A total of 3 studies were included, mainly targeting quadriceps atrophy, a core pathological change of KOA. Electrodes were mostly attached to motor points of the vastus medialis and vastus lateralis muscles or around the patella; current modes included AUSSIE medium-frequency alternating current, functional electrical stimulation (FES) pulse current, etc.; treatment frequency ranged from 3 times per week to 2 times per day, with treatment courses of 4-26 weeks. The multicenter extension study by Dasa et al.[11] (n=62) showed that after 26 weeks of home-based NMES treatment controlled by a mobile App, patients achieved more significant pain relief compared with low-voltage sham stimulation.

3.2.1.3. Peripheral Nerve Stimulation (PNS)

Only included in the retrospective case series by Mates et al.[16], targeting 4 patients with Kellgren-Lawrence grade 4 severe KOA who received implantation of an infrapatellar saphenous nerve stimulation system. Follow-up results showed that patients achieved more than 50% pain relief and sustained functional improvement, but the sample size was extremely small, resulting in low level of evidence.

3.2.1.4. Electroacupuncture (EA)

(Electroacupuncture (EA) and Electrical Dry Needling (EDN) combine the mechanical stimulation of acupuncture with the neuromodulatory effect of electrical stimulation, representing neuromodulation techniques with Chinese characteristics

A total of 3 studies were included. Acupoint selection was mainly based on local knee joint acupoints, including Dubi (ST35), Neixiyan (EX-LE5), Ququan (LR8), Xiyangguan (GB33), etc., combined with meridian-based acupoint selection; stimulation parameters mostly used 2/100Hz dense-disperse wave or continuous wave, with current intensity adjusted to induce slight needle tremor; treatment courses were 3-8 weeks, 3-6 times per week. Wang et al.[14] showed that after 8 weeks of electroacupuncture treatment, patients had greater reductions in WOMAC total score and pain score compared with the sham acupuncture group, and simultaneously increased the abundance of beneficial intestinal bacteria such as Agathobacter and Lachnoclostridium, thereby improving pain, stiffness, and functional impairment in KOA patients. Liu et al.[15]found that electroacupuncture was superior to conventional acupuncture and oral celecoxib, and could significantly reduce serum levels of pro-inflammatory factors IL-1β and TNF-α. The small-sample study by Jia et al.[13] first confirmed that electroacupuncture exerts anti-inflammatory effects by enhancing peripheral blood T cell receptor (TCR) diversity and reconstructing immune balance.

Electrical Dry Needling (EDN): Only included in the multicenter RCT by Dunning et al.[17](n=586). The results showed that after the initial 6-week intensive treatment, patients who received booster treatment once every 4 weeks had sustained improvements in WOMAC total score and pain score for 30 weeks, which was significantly better than the no-booster treatment group or those who received no further treatment.

3.2.1.5. Transcutaneous Vagus Nerve Stimulation (tVNS)

A total of 2 studies were included, both stimulating the left cymba concha (the exclusive innervation area of the auricular branch of the vagus nerve). Stimulation parameters were 25Hz frequency, 250μS pulse width, with current intensity adjusted to a tolerable slight tingling sensation for patients; treatment courses ranged from a single 60-minute session to 12 weeks. The 12-week RCT by Elsehrawy et al.[22] (n=68) showed that tVNS could continuously reduce VAS pain score and pressure pain threshold (PPT), and had significant efficacy in improving knee joint injury and neuropathic pain. The single-session pilot study by Aoyagi et al.[23] confirmed that tVNS could immediately improve knee pain in patients and enhance their central analgesia and pain tolerance, with no obvious adverse reactions.

3.2.2. Central Neuromodulation

Transcranial Direct Current Stimulation (tDCS): A total of 6 studies on tDCS alone were included, with relatively unified intervention protocols. Electrode positions all adopted the standard montage of placing the anode on the contralateral primary motor cortex (M1) of the affected knee and the cathode on the contralateral supraorbital area; stimulation parameters were 2mA current intensity, 20-minute single duration, with current ramping up and down within 30 seconds at the start and end of stimulation; treatment courses were 3 consecutive weeks, 5 times per week, total 15 times, mostly using home-based remotely supervised intervention. In terms of efficacy, most studies showed that tDCS could significantly improve pain intensity and physical function in KOA patients, and simultaneously regulate central sensitization indicators, such as increasing pressure pain threshold (PPT) and improving conditioned pain modulation (CPM) function. Suchting et al.[27] found that tDCS treatment could reduce plasma brain-derived neurotrophic factor (BDNF) levels in patients, and the reduction in BDNF was positively correlated with pain improvement. The secondary analysis by Park et al.[28] identified responder characteristics for tDCS treatment, with patients with higher body mass index (BMI) and body weight being more likely to respond to tDCS.

3.2.3. Combined Neuromodulation

A total of 2 studies were included, both adopting the dual-target intervention protocol of peripheral TENS combined with central tDCS. The intervention involved simultaneous implementation of local knee TENS and contralateral M1 area tDCS, with treatment courses of 6-8 weeks. Zhang et al.[31] showed that combined treatment could significantly improve gait parameters when patients stepped over obstacles, such as foot clearance height, crossing speed, and joint flexion and extension angles. Xia et al.[32] found that combined treatment could regulate the power of α and β waves in the cerebral cortex, improve pain and motor function, and was superior to TENS alone.

4. Discussion

4.1. Efficacy Advantages and Differential Characteristics of KOA Neuromodulation Techniques

The results of this scoping review show that various non-surgical neuromodulation techniques have certain efficacy in KOA treatment, each with distinct advantages.

4.1.1. Peripheral neuromodulation techniques directly act on local nerves and muscles, with characteristics of rapid onset, simple operation, high safety, and suitability for long-term home use, making them the mainstream in clinical application. NMES directly stimulates muscles around the knee joint, mainly targeting muscle atrophy. A study by Helal B Alqurashi[33] showed that NMES had moderate to small effects on muscle size, strength, and walking performance. A systematic review by Sarah Novak[34] provided recommended NMES parameters, such as a frequency of 50-75Hz. TENS generally has good analgesic effects with almost no side effects. Yu Wu[35] demonstrated that TENS could significantly relieve pain and improve dysfunction and walking ability in KOA patients. Electroacupuncture and electrical dry needling can simultaneously improve local inflammation, neuroimmune function, and systemic status. Yu Chen[36] found that electroacupuncture had the best efficacy among common acupuncture techniques. Studies by Zheng-tao Lv[37] and Zihang Chen[38] showed that high-intensity electroacupuncture was superior to low-intensity electroacupuncture in enhancing central analgesia, and high-intensity electroacupuncture could also improve patients’ emotional state.

4.1.2. Central neuromodulation techniques exert effects by regulating the brain’s pain processing network, among which tDCS is the most widely used clinically with the most abundant research. Natalia Comino-Suárez[39] indicated that tDCS has high safety and can significantly improve pain intensity and function. Rujuan Liu[40] proved that tDCS may reduce inflammation by altering the phenotypes of microglia and astrocytes and simultaneously inhibiting neural pathways. Combined neuromodulation techniques may achieve better efficacy than single techniques through synergistic effects of peripheral and central actions. A study by Rayssa Maria Do Nascimento[41] showed that tDCS combined with peripheral stimulation was superior to single stimulation in relieving pain from knee osteoarthritis and spinal cord injury. However, the number of current studies is small, and further verification is needed.

4.2. Current Research Gaps and Controversies

Although the number of studies on neuromodulation for KOA has been increasing annually, there are still many gaps and controversies:

4.2.1. Lack of Standardization in Intervention Protocols

Stimulation parameters, electrode positions, treatment courses, and frequencies vary widely among different techniques, making it impossible to directly compare the efficacy of different techniques and difficult to form unified clinical guidelines. For example, TENS frequencies range from 2Hz to 100Hz, and there are more than ten current modes for NMES, with optimal parameters yet to be clarified.

4.2.2. Deficiencies in Control Design

Some studies used blank controls instead of sham stimulation controls, which may overestimate the true efficacy of neuromodulation. Meanwhile, the effectiveness of sham stimulation protocols is also controversial, as some sham stimulations may still produce weak neuromodulatory effects.

4.2.3. Insufficient Long-Term Efficacy Evidence

Most studies have a follow-up period of no more than 3 months, lacking long-term efficacy and safety data beyond 6 months, making it impossible to evaluate the impact of neuromodulation on KOA disease progression.

4.2.4. Incompletely Elucidated Mechanisms of Action

Although some studies have explored mechanisms such as neuroimmunity, brain functional connectivity, and intestinal flora, the specific action pathways of different neuromodulation techniques remain unclear, making it difficult to guide individualized treatment.

4.2.5. Lack of Studies on Special Populations

Most included studies focused on middle-aged and elderly women, lacking research on men, young patients, severe KOA patients, and patients with comorbidities such as diabetes and cardiovascular diseases.

4.2.6. Lack of Health Economic Evaluation

No studies have evaluated the cost-effectiveness of neuromodulation techniques, limiting their inclusion in medical insurance policies and clinical promotion.

4.3. Limitations of This Study

This study has the following limitations: First, limited by resources, only two English databases (PubMed and Web of Science) were included, mainly focusing on open-access literatures, which may have missed some high-quality studies. Second, the sample sizes of included studies were generally small (maximum 586 cases, minimum 3 cases), and some were self-controlled before-after trials and single-arm studies, resulting in limited levels of evidence. Third, this review focused on non-surgical neuromodulation techniques and did not include surgical neuromodulation techniques such as cryoablation and radiofrequency ablation, leading to certain limitations in the research scope. Fourth, the health economic benefits and long-term patient compliance of neuromodulation techniques were not analyzed, limiting the clinical translation value of the conclusions.