Transcutaneous spinal cord stimulation for motor function and functional independence in patients with spinal cord injury: a systematic review

Abstract

Background

Spinal cord injury (SCI) can lead to severe impairments in motor, sensory, and autonomic functions, significantly affecting patients’ functional independence and quality of life. Transcutaneous spinal cord stimulation (tSCS), as an emerging non-invasive neuromodulation technique, has shown potential in improving motor function. However, existing studies demonstrate substantial heterogeneity, and its clinical efficacy remains unclear.

Objective

To systematically evaluate the effects of tSCS on motor function and functional independence in individuals with SCI, and to explore the potential influence of patient characteristics and stimulation parameters on treatment outcomes.

Methods

A systematic search of PubMed, Web of Science, Cochrane Library, Embase, and Scopus was conducted from database inception to March 23, 2026. Eligible studies included randomized controlled trials (RCTs), non-randomized studies, and case series. Methodological quality was assessed using the revised Cochrane Risk of Bias tool (RoB 2) for RCTs, the Risk Of Bias In Non-randomized Studies of Interventions (ROBINS-I) tool, and the Joanna Briggs Institute (JBI) critical appraisal checklist for case series. The Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach was used to assess the certainty of evidence. This review was registered in PROSPERO (registration number: CRD420251106919).

Results

A total of 25 studies (n = 353) were included, comprising 9 RCTs, 10 non-randomized studies, and 6 case series. Due to substantial heterogeneity in study design, patient characteristics, and intervention parameters, a qualitative synthesis was performed. The findings suggest that tSCS may have beneficial effects on upper limb motor function as well as balance and postural control. However, no consistent or stable improvements were observed in lower limb motor function, walking ability, functional independence, spasticity, or quality of life. According to the GRADE assessment, the overall certainty of evidence was very low. In terms of safety, tSCS was well tolerated, with no serious adverse events reported.

Conclusion

Current evidence is insufficient to support the definitive efficacy of tSCS in individuals with SCI. Although potential benefits have been observed in certain functional domains, the robustness of the evidence is limited by small sample sizes, methodological limitations, and substantial heterogeneity across studies. High-quality randomized controlled trials are needed to further establish its clinical effectiveness.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12984-026-02086-4.

Background

Spinal cord injury (SCI) is a highly disabling disorder of the central nervous system that can result in varying degrees of motor, sensory, and autonomic dysfunction, significantly affecting patients’ activities of daily living (ADL), functional independence, and level of social participation [1, 2]. Epidemiological data indicate that the global annual incidence of SCI is approximately 23.77 cases per million population, making it a significant public health concern [3]. Motor impairment is one of the most common and clinically significant sequelae following SCI and represents a key determinant of functional independence and quality of life [4, 5]. Despite recent advances in rehabilitation interventions, including conventional therapy, robotic-assisted training, functional electrical stimulation, and pharmacological treatments, neurological recovery in individuals with SCI remains limited, and functional deficits often persist long term [1, 6]. Therefore, identifying effective interventions to improve motor function and enhance functional independence remains a central challenge in SCI rehabilitation.

Transcutaneous spinal cord stimulation (tSCS), as an emerging non-invasive neuromodulation technique, has attracted increasing attention in recent years [7]. This approach is based on the principles of epidural spinal cord stimulation and delivers electrical currents via surface electrodes to the dorsal roots and associated neural networks of the spinal cord, thereby modulating the excitability of spinal circuits and facilitating the functional integration of residual neural pathways [8]. Previous studies have suggested that tSCS may have potential therapeutic value in improving motor function and functional independence in individuals with SCI [9, 10].

However, current evidence regarding the use of tSCS in SCI rehabilitation remains insufficient. Earlier studies have predominantly consisted of case reports, case series, and non-randomized studies. Although the number of randomized controlled trials (RCTs) has increased in recent years, substantial heterogeneity in patient characteristics (e.g., injury level, injury severity, and stage of disease) and intervention protocols (including stimulation parameters and intervention duration) limits the comparability of findings, making it difficult to draw definitive conclusions regarding its efficacy. In addition, the number of existing systematic reviews and meta-analyses remains limited, and some have not incorporated high-quality RCTs published in the past two years. Furthermore, insufficient attention has been paid to heterogeneity arising from variations in patient characteristics and intervention protocols, resulting in a lack of comprehensive evidence synthesis [11].

Based on this, the present study conducted a systematic review to comprehensively synthesize the latest evidence on the use of tSCS in individuals with SCI. This review primarily focuses on its effects on motor function and functional independence. During the analysis, results were stratified according to patient characteristics and variations in intervention implementation to provide a more objective evaluation of treatment effects under different conditions. The primary outcomes included upper and lower limb motor function, walking ability, balance function, and functional independence, while secondary outcomes included spasticity and quality of life. This study aims to provide evidence-based support for clinical decision-making and to inform the optimization of future rehabilitation strategies.

Methods

This study was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and followed a protocol registered in PROSPERO (CRD420251106919).

Search strategy

A systematic literature search was conducted in PubMed, Web of Science, the Cochrane Library, Embase, and Scopus. The search strategy included three categories of keywords: (1) intervention: “transcutaneous spinal cord stimulation,” “transcutaneous electrical spinal cord stimulation,” and “noninvasive spinal cord stimulation”; (2) population: “spinal cord injury”; and (3) outcomes: “motor activity,” “motor function,” “motor recovery,” “functional independence,” “activities of daily living,” “walking,” “gait,” “ambulation,” “hand function,” “upper limb function,” “upper extremity function,” “lower limb function,” “lower extremity function,” “lower limb motor,” and “lower extremity motor.” Medical Subject Headings (MeSH) terms and free-text terms in titles and abstracts were combined to construct the search strategy. The search was conducted from database inception to March 23, 2026. Initial screening was performed based on titles and abstracts, followed by full-text assessment of potentially eligible studies to determine final inclusion. To ensure comprehensive retrieval of relevant studies, reference lists of included articles and related reviews were manually searched as a supplement. The study selection process strictly followed the PRISMA guidelines. Detailed search strategies for each database are provided in Supplementary Material Method S1.

Inclusion and exclusion criteria

The inclusion and exclusion criteria were defined according to the PICOS framework. Studies meeting the following criteria were included in this systematic review: (1) population (P): patients with clinically diagnosed traumatic or non-traumatic SCI; (2) intervention (I): tSCS; (3) comparison (C): sham stimulation, conventional rehabilitation, or no intervention; (4) outcomes (O): reporting at least one outcome related to motor function or functional independence; and (5) study design (S): clinical interventional studies of any design. In addition, only studies published in English were included. The exclusion criteria were as follows: (1) animal studies or studies involving healthy participants; (2) studies in which tSCS was combined with other neuromodulation techniques; (3) studies with incomplete or non-extractable data; and (4) studies with a sample size of fewer than five participants.

Study selection and data extraction

Study selection and data extraction were performed independently by two reviewers. The study selection process involved removal of duplicates, initial screening based on titles and abstracts, and full-text assessment. Any disagreements arising during the selection process were resolved through discussion, and if necessary, adjudicated by a third reviewer. The following data were extracted: study design, sample size, sex distribution, mean age, disease stage, level of injury, American Spinal Injury Association (AIS) grade, duration of injury, stimulation parameters (e.g., frequency, intensity, and duration), outcomes related to motor function and functional independence, and reported adverse events. Given the substantial heterogeneity among the included studies in terms of study design, intervention protocols, and outcome measures, and the inability to obtain additional data from the original authors, a meta-analysis was not performed. Instead, a qualitative synthesis was conducted to summarize the findings.

Methodological quality assessment

Methodological quality assessment was conducted independently by two reviewers, with disagreements resolved through discussion. Appropriate risk-of-bias assessment tools were applied according to study design. For RCTs, the revised Cochrane Risk of Bias tool (RoB 2) was used [12]. For non-randomized studies, including pre–post studies, crossover studies, and cohort studies, the Risk Of Bias In Non-randomized Studies of Interventions (ROBINS-I) tool was applied [13]. For case series, the Joanna Briggs Institute (JBI) critical appraisal checklist for case series was used [14]. In addition, the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach was employed to assess the certainty of evidence for each outcome. The evaluation considered risk of bias, inconsistency, indirectness, imprecision, and other relevant factors, and the quality of evidence was categorized as high, moderate, low, or very low [15].

Results

Study selection

A total of 494 records were identified through database searches (PubMed, n = 87; Web of Science, n = 103; Embase, n = 169; Scopus, n = 93; Cochrane Library, n = 42). After removing 262 duplicates, 232 records remained for title and abstract screening. During the initial screening, 156 records were excluded, including 31 review articles and 125 studies that were not relevant to the research topic. Subsequently, 76 articles were retrieved for full-text assessment. In addition, 8 studies were identified through manual searching of reference lists from relevant reviews, resulting in a total of 84 articles undergoing full-text evaluation. After full-text assessment, 59 studies were excluded, including 47 due to incomplete data and 12 due to small sample sizes (fewer than five participants). Ultimately, 25 studies were included in this systematic review (Fig. 1).

Fig. 1.

PRISMA flow diagram of study selection

Study characteristics

A total of 25 studies involving 353 participants were included, comprising 9 RCTs (6 parallel-group trials and 3 crossover trials), 3 pre–post studies, 3 crossover studies, 4 cohort studies, and 6 case series (Table 1). Among studies reporting age, the participants ranged from 12.9 to 49.0 years. Overall, the study population was predominantly male, with 278 male participants, accounting for 78.8% of the total sample. In terms of disease stage, 20 studies included individuals with chronic SCI, while 5 studies included those in the subacute stage. The level of injury varied widely, ranging from C1 to T12. Specifically, 10 studies included only participants with cervical injuries, 14 studies included mixed levels of injury, and 1 study did not report the level of injury. Regarding injury severity, most studies included participants with mixed American Spinal Injury AIS grades. Among the 24 studies that reported detailed AIS grade distributions, there were 45 participants with grade A, 47 with grade B, 122 with grade C, and 119 with grade D. One additional study, including 20 participants, reported an AIS grade range of B to D without providing the number of participants in each category.

Table 1.

Characteristics of Included Studies

Study	Design	Subjects	Sex	Age	Disease stage	SCI level	AIS	Time since injury (years)	TSCS parameters	Adverse events
Gerasimenko et al. (2015)	Case series	5	M: 5	31.4	Chronic	C5 ~ 6, C6, C7, T3 ~ 4	B: 5	3.2	5 Hz/30 Hz, 1 ms, 80 ~ 180 mA, monophasic pulses, 1 session/week, 18 weeks	–
Gad et al. (2018)	Case series	6	F: 1 M: 5	40.2	Chronic	C4, C6, C8	B: 2 C: 4	10	30 Hz, 1 ms, 10 ~ 250 mA, monophasic or biphasic pulses, 2 sessions/week, 4 weeks	None
Freyvert et al. (2018)	Within-subject	6	F: 2 M: 4	19.2	Chronic	C2, C5, C6	B: 6	2.4	5 ~ 30 Hz, 20 ~ 100 mA, 2 weeks	Paresthesia
Rath et al. (2018)	Self-controlled crossover	8	F: 1 M: 7	29.4	Chronic	C4, C5, T2 ~ 4, T9	A: 6 C: 2	7.3	15 Hz/30 Hz, 1 ms, motor threshold, monophasic pulses, 1 session	–
Sayenko et al. (2019)	Self-controlled crossover	15	F: 3 M: 12	29.9	Chronic	C4 ~ C6, T2 ~ T4, T8, T9, T12	A: 11 B: 1 C: 3	6	0.2 ~ 30 Hz, 1 ms, < 150 mA, monophasic pulses, 3 sessions / week, 4 weeks	Erythema, autonomic dysreflexia
Hofstoetter et al. (2020)	Within-subject	12	F: 3 M: 9	41.3	Chronic	C4, C6, C7, T4, T6, T7	A: 3 C: 3 D: 6	12.1	50 Hz, 1 ms, 40.8 ± 24.1 mA, biphasic pulses, 1 session (30 min)	None
Al’Joboori et al. (2020)	Cohort	7	F: 3 M: 4	36.14	Chronic	C5, C6 ~ 7, T3, T5, T6, T10	A: 4 C: 2 D: 1	4.5	30 Hz, 1 ms, tolerable intensity, biphasic pulses, 3 sessions/week, 8 weeks	Autonomic dysreflexia, paresthesia
Meyer et al. (2020)	Case series	10	F: 1 M: 9	45.4	Chronic	C3 ~ C7, T3, T4, T10	D: 10	11.6	15 Hz/30 Hz/50 Hz, 1 ms, 26.4 ± 17.3 mA, biphasic pulses, 1 session (5 min/15 min)	Paresthesia
McHugh et al. (2020)	Case series	10	F: 4 M: 6	43	Chronic	C4 ~ C7, T1, T3, T4, T8, T9	C: 4 D: 6	13.4	50 Hz, 1 ms, 20 ~ 80 mA, biphasic pulses, 30 min/session, 3 sessions/week, 8 weeks	Erythema
Inanici et al. (2021)	Self-controlled crossover	6	F: 2 M: 4	42	Chronic	C3, C5	B: 2 C: 2 D: 2	4.6	30 Hz, 1 ms, 0 ~ 120 mA, monophasic or biphasic pulses, 120 min/session, 3 sessions/week, 4 weeks	None
Tefertiller et al. (2021)	Case series	7	F: 2 M: 5	27.7	Chronic	C4 ~ C6	B: 4 C: 2 D: 1	1.9	30 Hz, 1 ms, voluntary activation threshold, monophasic pulses, 60 min/session, 20 ~ 80 sessions	None
Estes et al. (2021)	Randomized controlled	16	F: 4 M: 12	43.4	Subacute	C1 ~ C7, T11	B: 1 C: 5 D: 10	98.6 days	50 Hz, tolerable intensity, biphasic pulses, 30 min/session, 3 sessions/week, 2 weeks	None
Huang et al. (2022)	Randomized crossover	10	F: 4 M: 6	34.4	Chronic	C3 ~ C7	A: 5 B: 5	5.7	30 Hz, 1 ms, optimal intensity, monophasic pulses, 6 weeks	–
García-Alén et al. (2023)	Randomized controlled	21	F: 2 M: 19	39.2	Subacute	C3 ~ C7, T1	A: 3 B: 5 C: 6 D: 7	5.3 months	30 Hz, 1 ms, 39 ~ 86 mA, biphasic pulses, 60 min/session, 4 sessions/week, 2 weeks	Erythema, nausea, cough, paresthesia, autonomic dysreflexia, increased spasticity
Hope et al. (2023)	Randomized controlled	10	F: 1 M: 9	40.9	Subacute	C1 ~ 5, C7	C: 2 D: 8	94.6 d	50 Hz, 300 µs, tolerable intensity, biphasic pulses, 30 min/session, 3 sessions/week, 2 weeks	–
Kumru et al. (2024)	Case series	10	F: 1 M: 9	43.2	Chronic	C3 ~ C8	D: 10	13.5 months	30 Hz, 1 ms, tolerable intensity, biphasic pulses, 1 session (7 stimulation conditions, total 14 min)	None
Moritz et al. (2024)	Single-arm cohort	60	F: 10 M: 50	47.2	Chronic	C2 ~ C7	B: 9 C: 28 D: 23	6.1	30 Hz, 100 µs, motor threshold, monophasic or biphasic pulses, 60 min/session, 3 ~ 5 sessions/week, 8 weeks	None
Singh et al. (2024)	Within-subject	7	F: 1 M: 6	12.9	Chronic	C3, C5, C7, C1 ~ 5, C4 ~ 7, C2 ~ T1, C6 ~ T4	A: 1 B: 4 C: 1 D: 1	9.5	30 Hz, 1 ms, 20 ~ 190 mA, biphasic pulses, 1 session (24 min)	Erythema, abnormal blood pressure, pain, paresthesia, fatigue
McKenzie et al. (2024)	Randomized crossover	10	F: 2 M: 8	46.9	Chronic	C4 ~ C6, T2, T6, T8, T12	D: 10	8.8	30 Hz, 1 ms, 19 ~ 100 mA, biphasic pulses, 30 min/session, 1 session/day, 5 days	None
Comino-Suárez et al. (2025)	Randomized controlled	27	F: 3 M: 24	48.0	Subacute	C2 ~ T11	C: 18 D: 9	111.3 days	30 Hz, 1 ms, tolerable intensity, biphasic pulses, 20 min/session, 5 sessions/week, 4 weeks	Erythema, paresthesia
Capozio et al. (2025)	Randomized crossover	5	F: 5	49.0	Chronic	C3 ~ 8, C4 ~ 6, C5 ~ 6, C6 ~ 7	C: 3 D: 2	12.4	30 Hz, 100 µs, optimal intensity, biphasic pulses, 30 min/session, 4 weeks	None
Verma et al. (2025)	Cohort	5	F: 1 M: 4	39.2	Chronic	C4, C5, C4 ~ 5, C5 ~ 6, C6 ~ 7	A: 3 B: 1 C: 1	9.2	30 Hz, 500 µs, tolerable intensity, biphasic pulses, 1 session	–
Suggitt et al. (2025)	Single-arm cohort	10	F: 3 M: 7	37.9	Chronic	C4, C5, T6, T10, T12	A: 4 B: 2 C: 3 D: 1	5.3	30 Hz, 1 ms, 3 ~ 45 mA, monophasic pulses, 45 min/session, 120 sessions	Erythema
Rashad et al. (2026)	Randomized controlled	20	F: 6 M: 14	–	Subacute	–	B ~ D	3 ~ 6 months	30 Hz, 1 ms, tolerable intensity, biphasic pulses, 20 min/session, 5 sessions / week, 8 weeks	None
Bye et al. (2026)	Randomized controlled	50	F: 10 M: 40	TSCS: 46 Sham: 38	Chronic	C7 ~ T5, T6 ~ 11	A: 5 C: 33 D: 12	TSCS: 4 Sham: 3	20 Hz, 1 ms, minimum stimulation intensity to evoke VM sEMG response, biphasic pulses, 30 min/session, 3 sessions/week, 12 weeks	Autonomic dysreflexia, skin abrasion / pressure, urinary incontinence, musculoskeletal pain

AIS: American Spinal Injury Association Impairment Scale; TSCS: Transcutaneous Spinal Cord Stimulation; F/M: female/male; Sham: sham stimulation without therapeutic electrical output

Quality assessment

The risk of bias varied considerably across different study designs. Among the 9 included RCTs, 1 was rated as low risk, 5 as having some concerns, and 3 as high risk (Table 2). The main sources of bias included inadequate reporting of allocation concealment and insufficient implementation or reporting of blinding procedures. Among the 10 non-randomized studies, 3 were assessed as having moderate risk of bias and 7 as serious risk (Table 3). The primary contributors to bias were confounding due to the lack of randomization and potential measurement bias in outcome assessment. The overall methodological quality of the 6 case series was considered moderate. The main limitations included unclear reporting of whether consecutive inclusion was applied, potential selection bias, and insufficient reporting of key contextual information such as study setting in some studies (Table 4).

Table 2.

Risk of bias assessment of randomized controlled trials (Cochrane RoB 2)

Study	Randomization	Deviations from interventions	Missing data	Outcome measurement	Selective reporting	Overall
Estes et al. (2021)	Some Concerns	Some Concerns	Low Risk	Some Concerns	Low Risk	Some Concerns
Huang et al. (2022)	Some Concerns	Some Concerns	Some Concerns	Low Risk	Some Concerns	Some Concerns
García-Alén et al. (2023)	Some Concerns	High Risk	Some Concerns	Some Concerns	Some Concerns	High Risk
Hope et al. (2023)	Some Concerns	Some Concerns	Low Risk	Some Concerns	Some Concerns	Some Concerns
McKenzie et al. (2024)	Some Concerns	Some Concerns	Some Concerns	High Risk	Some Concerns	High Risk
Comino-Suárez et al. (2025)	Low Risk	High Risk	Low Risk	Some Concerns	Low Risk	High Risk
Capozio et al. (2025)	Some Concerns	Some Concerns	Low Risk	Some Concerns	Some Concerns	Some Concerns
Rashad et al. (2026)	Low Risk	Some Concerns	Low Risk	Low Risk	Low Risk	Some Concerns
Bye et al. (2026)	Low Risk	Low Risk	Low Risk	Low Risk	Low Risk	Low Risk

Table 3.

Risk of bias assessment of non-randomized studies (ROBINS-I)

Study	Confounding	Selection	Intervention classification	Deviations from intervention	Missing data	Measurement	Selective reporting	Overall
Freyvert et al. (2018)	Moderate Risk	Low Risk	Low Risk	Low Risk	Moderate Risk	Low Risk	Low Risk	Moderate Risk
Rath et al. (2018)	Serious Risk	Low Risk	Low Risk	Moderate Risk	Low Risk	Low Risk	Moderate Risk	Serious Risk
Sayenko et al. (2019)	Moderate Risk	Low Risk	Low Risk	Low Risk	Low Risk	Moderate Risk	Low Risk	Moderate Risk
Hofstoetter et al. (2020)	Serious Risk	Low Risk	Low Risk	Low Risk	Low Risk	Moderate Risk	Low Risk	Serious Risk
Al’Joboori et al. (2020)	Serious Risk	Serious Risk	Moderate Risk	Serious Risk	Serious Risk	Moderate Risk	Low Risk	Serious Risk
Inanici et al. (2021)	Serious Risk	Low Risk	Low Risk	Low Risk	Low Risk	Serious Risk	Low Risk	Serious Risk
Moritz et al. (2024)	Serious Risk	Moderate Risk	Low Risk	Low Risk	Low Risk	Low Risk	Low Risk	Serious Risk
Singh et al. (2024)	Moderate Risk	Low Risk	Low Risk	Moderate Risk	Low Risk	Low Risk	Low Risk	Moderate Risk
Verma et al. (2025)	Serious Risk	Moderate Risk	Low Risk	Low Risk	Moderate Risk	Moderate Risk	Moderate Risk	Serious Risk
Suggitt et al. (2025)	Serious Risk	Serious Risk	Low Risk	Low Risk	Serious Risk	Moderate Risk	Low Risk	Serious Risk

Table 4.

Risk of bias assessment for case series studies–JBI critical appraisal checklist

Study	Inclusion criteria	Condition measurement	Case identification	Consecutive inclusion	Complete inclusion	Demographic info	Clinical info	Outcome reporting	Background info	Statistical analysis	Overall quality
Gerasimenko et al. (2015)	Yes	Yes	Yes	Unclear	Unclear	Yes	Yes	Yes	Yes	Yes	Moderate to high
Gad et al. (2018)	Yes	Yes	Yes	Unclear	Unclear	Yes	Yes	Yes	No	Yes	Moderate to low
Meyer et al. (2020)	Yes	Yes	Yes	Unclear	Unclear	Yes	Yes	Yes	No	Yes	Moderate to low
McHugh et al. (2020)	Yes	Yes	Yes	Unclear	Unclear	Yes	Yes	Yes	Yes	Yes	Moderate to high
Tefertiller et al. (2021)	Yes	Yes	Yes	Unclear	Unclear	Yes	Yes	Yes	Yes	Yes	Moderate to high
Kumru et al. (2024)	Yes	Yes	Yes	Unclear	Unclear	Yes	Yes	Yes	No	Yes	Moderate to low

Outcomes

The types of outcomes reported across the included studies were diverse, mainly covering motor function, functional independence, spasticity, and quality of life, with a certain degree of heterogeneity in outcome selection and assessment tools among studies. Motor-related outcomes included upper and lower limb motor function, walking ability, and balance and postural control. Non-motor outcomes primarily included functional independence, spasticity, and quality of life. Detailed outcome measures and assessment tools used in each study are presented in Table 5.

Table 5.

Summary of outcome measures used in the included studies

Study	Outcome measures
Gerasimenko et al. [16] (2015)	Hip and knee ROM, etc.
Gad et al. [17] (2018)	LEMS, MVC, UEMS, Quality of life measures, etc.
Freyvert et al. [18] (2018)	Handgrip force, UEMS, ARAT, MAS, etc.
Rath et al. [19] (2018)	Quiet sitting, Multi-directional leaning, Self-initiated perturbation
Sayenko et al. [20] (2019)	Knee extension assistance required during standing, Center of pressure excursions, MAS, etc.
Hofstoetter et al. [21] (2020)	10MWT, MAS, Pendulum test, etc.
Al’Joboori et al. [22] (2020)	LEMS, UEMS, SCIM-III, SF-36, etc.
Meyer et al. [23] (2020)	Ankle motor control, Walking speed, Gait kinematics, Lower limb muscle activity, etc.
McHugh et al. [24] (2020)	10MWT, 6MWT, TUG, WISCI-II, etc.
Inanici et al. [25] (2021)	LEMS, Lateral pinch force, GRASSP, UEMS, SCIM-III, MAS, WhoQol-Bref, etc.
Tefertiller et al. [26] (2021)	LEMS, 10MWT, 6MWT, CUE-T, UEMS, Nine-Hole Peg Test, etc.
Estes et al. [27] (2021)	10MWT, 2MWT, mSCI-SET, SCATS, Pendulum test, Ankle clonus drop test, etc.
Huang et al. [28] (2022)	MVC, GRASSP, CUE-Q, SCIM, MAS, PSFS, etc.
García-Alén et al. [10] (2023)	MVC (Cylindrical Grip, Key Lateral Pinch, Tip-to-Tip Pinch), GRASSP, BBT, UMES, SCIM-III, WhoQol-Bref, etc.
Hope et al. [29] (2023)	Ankle dorsiflexion angle (walking), Ankle dorsiflexion ROM (voluntary), Ankle clonus drop test, etc.
Moritz et al. [30] (2024)	UEMS, GRASSP, CUE-T, SCIM III, PSFS, EQ-5D-5 L, WhoQol-Bref, etc.
Kumru et al. [31] (2024)	6MWT, MVC (Quadriceps, Tibialis anterior), etc.
Singh et al. [32] (2024)	Maximal hand-grip, wrist flexion, wrist extension, and elbow extension strength
McKenzie et al. [33] (2024)	Maximum isometric ankle plantarflexion torque, 10MWT, 6MWT, TUG
Comino-Suárez et al. [9] (2025)	The isometric maximal voluntary contraction of knee extensors and ankle dorsiflexors of both limbs, LEMS, 10MWT, 6MWT, TUG, WISCI-II, SCIM-III, MAS, etc.
Capozio et al. [34] (2025)	Functional grip force during unimanual grasp-and-lift tasks, GRASSP, SCIM-III, QLI-SCI, etc.
Verma et al. [35] (2025)	Maximum isometric grip force, etc.
Suggitt et al. [36] (2025)	LEMS, Pinch and grasp strength, UEMS, GRASSP, BBS, SCIM-III, MAS, WhoQol-Bref, etc.
Rashad et al. [37] (2026)	TUG, 10MWT, WISCI-II, BBS, etc.
Bye et al. [38] (2026)	WISCI-II, LEMS, MAS, EQ-5D-5 L

ROM: Range of Motion; LEMS: Lower Extremity Motor Score; MVC: Maximal Voluntary Contraction; UEMS: Upper Extremity Motor Score; ARAT: Action Research Arm Test; MAS: Modified Ashworth Scale; 10MWT: 10-Meter Walk Test; SCIM-III: Spinal Cord Independence Measure III; SF-36: Short Form-36 Health Survey; 6MWT: 6-Minute Walk Test; TUG: Timed Up and Go Test; WISCI-II: Walking Index for Spinal Cord Injury II; GRASSP: Graded Redefined Assessment of Strength, Sensibility, and Prehension; WhoQol-Bref: World Health Organization Quality of Life-BREF; CUE-T: Capabilities of Upper Extremity Test; 2MWT: 2-Minute Walk Test; mSCI-SET: 12modified Spinal Cord Injury-Spasticity Evaluation Tool; SCATS: Spinal Cord Assessment Tool for Spastic Reflexes; CUE-Q: Capabilities of Upper Extremity Questionnaire; PSFS: Penn Spasm Frequency Scale; BBT: Box and Block Test; EQ-5D-5 L: EuroQol 5-Dimension 5-Level Questionnaire; QLI-SCI: SCI Version of the Quality of Life Index; BBS: Berg Balance Scale

Intervention effects

Motor function

Upper limb motor function

A total of 12 studies reported outcomes related to upper limb motor function. The assessment measures primarily included upper limb muscle strength, the Upper Extremity Motor Score (UEMS), the Graded Redefined Assessment of Strength, Sensibility and Prehension (GRASSP), the Capabilities of Upper Extremity Test (CUE-T), and the Capabilities of Upper Extremity Questionnaire (CUE-Q).

Regarding upper limb muscle strength, 10 studies conducted relevant assessments; however, the consistency of evidence was limited. For GRASSP strength scores, only one RCT involving subacute patients demonstrated greater improvements in the tSCS combined with robotic exoskeleton intervention compared with the control group, while the remaining evidence was primarily derived from non-randomized studies, with inconsistent findings [10, 25, 36]. In terms of grip and/or pinch strength, one RCT indicated that patients with residual hand function exhibited more pronounced improvements [28]. Additionally, several non-randomized studies and case series reported a gradual improvement over time [17, 18, 25].

For upper limb motor performance, eight studies assessed outcomes using UEMS, and the overall magnitude of improvement appeared limited. One RCT involving subacute patients showed a significant improvement in UEMS from baseline following tSCS combined with robotic exoskeleton intervention; however, no significant between-group differences were observed [10]. Findings from non-randomized studies and case series were also inconsistent, with most reporting only mild improvements in certain individuals or on specific sides [17, 18, 26].

Regarding fine motor function and hand function, eight studies conducted relevant assessments, although results varied across different measurement tools. One RCT involving subacute patients demonstrated greater improvements in GRASSP prehension scores in the tSCS combined with robotic exoskeleton group compared with controls [10]. Another RCT found no significant changes in GRASSP prehension scores but reported improvements in CUE-Q scores among patients with residual hand function [28]. Non-randomized studies and case series more consistently reported improvements in GRASSP prehension performance or upper limb functional task outcomes [25, 26, 36].

Overall, tSCS may have a potential facilitating effect on upper limb muscle strength and fine hand function; however, the level of evidence remains low and the findings are not consistent across studies.

Lower limb motor function

A total of 12 studies reported outcomes related to lower limb motor function. The primary outcome measures included range of motion (ROM), maximum voluntary contraction (MVC), and the Lower Extremity Motor Score (LEMS).

Regarding lower limb ROM, three studies conducted relevant assessments; however, the overall level of evidence was low and findings were inconsistent. One RCT involving subacute patients showed that tSCS combined with exercise training did not significantly improve ankle dorsiflexion angle or peak ankle angle during the swing phase of gait [29]. In contrast, case series studies reported improvements in knee and ankle joint ROM and suggested that 30 Hz stimulation may be more effective than 15 Hz in facilitating ankle dorsiflexion [16, 23].

For lower limb muscle strength, three studies provided relevant data. Current evidence does not support a beneficial effect of tSCS on lower limb muscle strength. RCT findings indicated no significant between-group differences in MVC of the knee extensors and ankle dorsiflexors, or in isometric ankle torque [9, 33]. One RCT involving subacute patients reported an improvement in knee extensor MVC in within-group analysis [9]. Case series studies also failed to demonstrate significant improvements in quadriceps MVC, with only one study reporting increased tibialis anterior MVC under specific stimulation parameters [31].

In terms of lower limb motor performance, seven studies assessed outcomes using LEMS. The available evidence does not demonstrate a clear improvement in lower limb motor performance with tSCS. RCTs consistently reported no significant between-group differences [9, 38]. One RCT involving subacute patients observed improvements in LEMS in within-group analysis [9]. Findings from non-randomized studies and case series were largely consistent, showing no significant overall improvements, with only mild gains observed in a small number of participants or on specific sides [17, 22, 25, 26].

Overall, current evidence is insufficient to support a consistent beneficial effect of tSCS on lower limb motor function.

Walking ability

A total of 10 studies reported outcomes related to walking ability. The assessment measures included the 10-Meter Walk Test (10MWT), 6-Minute Walk Test (6MWT), 2-Minute Walk Test (2MWT), Walking Index for Spinal Cord Injury II (WISCI-II), and the Timed Up and Go Test (TUG).

Regarding walking speed and/or endurance, nine studies conducted relevant assessments, with overall inconsistent findings. RCTs did not demonstrate significant between-group differences, although some studies reported improvements in within-group analyses [9, 27, 33]. Findings from non-randomized studies and case series were variable. Some studies evaluating single-session stimulation did not observe improvements in walking speed [21, 23]. In contrast, one study using a stimulation frequency of 50 Hz reported improvements in both walking speed and endurance, exceeding the Minimal Clinically Important Difference (MCID) or Minimal Detectable Difference (MDD) [24].

For functional ambulation, five studies used WISCI-II and/or TUG for assessment. Current evidence does not demonstrate a clear beneficial effect of tSCS on functional ambulation. RCTs reported no significant between-group differences, although some studies observed improvements compared with baseline following the intervention [9, 33, 37]. Similar trends were reported in case series studies [24].

Overall, current evidence does not support a consistent and robust effect of tSCS on walking function.

Balance and postural control

A total of four studies reported outcomes related to balance and postural control. Findings from RCTs indicated that Berg Balance Scale (BBS) scores showed a trend toward improvement over time following tSCS intervention [37]. Results from non-randomized studies were relatively consistent, demonstrating improvements in both static and dynamic balance parameters, including reduced center of pressure displacement and acceleration, increased limits of stability, and enhanced weight-shifting ability [19]. In addition, some studies reported improvements in standing control, reflected by reduced stimulation intensity required to induce standing and decreased need for assistance [20]. Overall, current evidence suggests that tSCS may have a potential facilitating effect on balance and postural control; however, the level of evidence remains low.

Functional independence

A total of eight studies reported outcomes related to functional independence, primarily assessed using the Spinal Cord Independence Measure III (SCIM III). RCTs did not demonstrate significant between-group differences, although some studies involving subacute patients reported significant improvements compared with baseline following the intervention [9, 10]. One study further suggested that patients with residual hand function exhibited greater improvements [28]. In contrast, findings from non-randomized studies were relatively consistent, showing no significant changes in total SCIM III scores [22, 25, 30, 36]. Overall, current evidence does not support a consistent and robust effect of tSCS on functional independence.

Other outcomes

A total of 11 studies reported outcomes related to spasticity. Assessment measures included the Modified Ashworth Scale (MAS), Penn Spasm Frequency Scale (PSFS), Modified Spinal Cord Injury Spasticity Evaluation Tool (mSCI-SET), Spinal Cord Assessment Tool for Spasticity (SCATS), as well as the pendulum test and ankle clonus drop test. Findings from RCTs consistently showed no significant improvement in spasticity following tSCS intervention, either in within-group or between-group comparisons [9, 27–29, 38]. Results from non-randomized studies were variable, with most studies reporting no significant changes, while a few reported reductions in MAS scores or decreased spasm frequency [25, 30].

A total of eight studies reported outcomes related to quality of life. Assessment tools included the World Health Organization Quality of Life-BREF (WHOQOL-BREF), EuroQol 5-Dimension 5-Level (EQ-5D-5 L), Short Form-36 Health Survey (SF-36), and Quality of Life Index–Spinal Cord Injury Version (QLI-SCI). RCT findings were relatively consistent, showing no significant improvement in quality of life following the intervention [10, 34, 38]. Results from non-randomized studies and case series were inconsistent, with only one study reporting improvements in quality of life after the intervention, while the remaining studies did not demonstrate consistent changes [30]. In addition, one study based on participants’ self-reports indicated improvements in activities of daily living in some individuals [17].

Overall, current evidence is insufficient to support a consistent and robust effect of tSCS on spasticity-related outcomes and quality of life.

Stratified analysis by patient characteristics and intervention parameters

To further characterize the variability in the effects of tSCS under different study conditions, the results of the included studies were stratified according to patient characteristics and intervention parameters.

Regarding patient characteristics, most studies included individuals with chronic SCI, while only a few involved subacute populations. Limited evidence suggests that functional improvements may be more readily observed in subacute patients [10]. In addition, most studies included mixed populations with cervical and thoracic injuries, as well as both complete and incomplete injuries; however, the majority did not report outcomes stratified by injury level or severity. Some studies indicated that patients with residual function (e.g., preserved hand grip strength) may be more likely to benefit from tSCS intervention [28].

With respect to intervention parameters, stimulation frequency was most commonly set at 30 Hz, with a few studies exploring alternative frequencies such as 15–50 Hz. Some evidence suggests that higher frequencies may produce more pronounced improvements in specific outcomes (e.g., walking ability), although a consistent dose–response relationship has not been established [23, 24]. Stimulation intensity was typically adjusted according to individual tolerance, resulting in considerable variability across studies and participants. Intervention dosage also varied among studies. Some studies employing single-session stimulation did not observe significant effects; however, findings were inconsistent across studies, and the specific impact of intervention dosage on outcomes remains unclear [21, 23].

Certainty of evidence

The certainty of evidence for each outcome was assessed using the GRADE approach, and all outcomes were rated as having very low certainty (Table 6). Overall, the low certainty of evidence was attributable to multiple factors. First, the included studies were predominantly non-randomized studies and case series, resulting in a high overall risk of bias. Second, substantial heterogeneity in patient characteristics and intervention parameters contributed to inconsistency across study findings. In addition, the lack of standardized assessment tools for certain outcomes increased the indirectness of the evidence. Furthermore, most studies had small sample sizes, leading to imprecision and reduced stability of effect estimates.

Table 6.

GRADE assessment of evidence certainty across outcomes

Outcome	Number of studies (participants)	Study design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations	Overall certainty
Upper limb motor function	12 studies (n = 150)	3 RCTs, 9 non-randomized studies/case series	Very serious	Serious	Serious	Serious	None	Very low
Lower limb motor function	12 studies (n = 158)	4 RCTs, 8 non-randomized studies/case series	Very Serious	Not serious	Serious	Serious	None	Very low
Walking ability	10 studies (n = 172)	5 RCTs, 5 non-randomized studies/case series	Serious	Serious	Not serious	Serious	None	Very low
Balance and postural control	4 studies (n = 53)	1 RCTs, 3 non-randomized studies/case series	Very Serious	Not serious	Serious	Very Serious	None	Very low
Functional independence	8 studies (n = 146)	4 RCTs, 4 non-randomized studies/case series	Very Serious	Not serious	Not serious	Serious	None	Very low
Spasticity	11 studies (n = 222)	5 RCTs, 6 non-randomized studies/case series	Serious	Not serious	Serious	Serious	None	Very low
Quality of life	8 studies (n = 165)	3 RCTs, 5 non-randomized studies/case series	Very Serious	Not serious	Very Serious	Serious	None	Very low

With respect to specific outcomes, upper limb motor function as well as balance and postural control showed potential benefits of tSCS in some studies. However, these findings were primarily derived from non-randomized studies and case series, and the results were not entirely consistent, resulting in limited overall certainty of evidence. In contrast, for lower limb motor function, walking ability, and functional independence, current evidence does not demonstrate a consistent trend of improvement, and no definitive conclusions can be drawn. Similarly, for spasticity and quality of life, available evidence does not indicate sustained and consistent improvements, and interpretation remains limited by study design and sample size.

Adverse events

Among the 25 included studies, 20 reported outcomes related to adverse events. Of these, 10 studies did not observe any adverse events associated with tSCS. The remaining 10 studies reported adverse events; however, all were mild in severity, and no serious adverse events were identified, indicating that tSCS is generally safe and well tolerated. In terms of specific types of adverse events, skin redness at the electrode placement site and limb paresthesia were the most commonly reported. Additionally, four studies reported episodes of autonomic dysreflexia [10, 20, 22, 38]. Besides these events, a small number of participants experienced nausea, cough, fatigue, musculoskeletal pain, increased blood pressure, worsening spasticity, and minor skin irritation or discomfort due to pressure [10, 32, 38].

Discussion

This systematic review included 25 studies and comprehensively evaluated the effects of tSCS on motor function and functional independence in individuals with SCI. Overall, the effects of tSCS varied across different functional outcomes. Some studies suggested potential benefits in upper limb motor function as well as balance and postural control. However, for lower limb motor function, walking ability, functional independence, spasticity, and quality of life, current evidence does not demonstrate consistent or stable improvements. From a safety perspective, tSCS appears to be well tolerated, with no serious adverse events reported in the included studies. Although some studies observed signals of functional improvement, the overall certainty of evidence was rated as very low according to the GRADE assessment. Therefore, the current evidence is insufficient to support definitive conclusions regarding the efficacy of tSCS, and its clinical value requires further validation through high-quality RCTs.

Compared with previous systematic reviews, the overall conclusions of this study are largely consistent, indicating that existing evidence is insufficient to support a definitive therapeutic effect of tSCS. In terms of upper limb motor function and balance and postural control, this review suggests a trend toward improvement, whereas no clinically meaningful improvement was observed in functional independence; these findings are consistent with previous studies [11, 39]. In contrast, no consistent improvement was observed in lower limb motor function and walking ability, which differs from the findings reported by Shankar et al. [39]. This discrepancy may be attributable to differences in analytical focus. The present review primarily evaluated short-term effects following the intervention and did not systematically synthesize long-term follow-up data, whereas previous reviews placed greater emphasis on long-term outcomes, potentially leading to differing conclusions. This discrepancy also suggests that the potential effects of tSCS may be time-dependent, with benefits not fully manifested in the short term.

It is noteworthy that the aforementioned negative findings do not necessarily indicate a lack of true efficacy of tSCS but may be influenced by several factors. First, most included studies had relatively small sample sizes, resulting in limited statistical power and a potential failure to detect true effects. Second, substantial heterogeneity in stimulation parameters and intervention dosage across studies may have contributed to variability in outcomes, as some intervention protocols may not have reached an effective neuromodulatory threshold. In addition, most studies evaluated only short-term effects, whereas recovery of lower limb motor function and walking ability depends on complex neural network reorganization, which typically requires longer intervention periods and consolidation processes; thus, short-term assessments may underestimate long-term effects [40]. Furthermore, functional independence and quality of life are multidimensional outcomes influenced by multiple factors, and a single neuromodulatory intervention is unlikely to produce significant changes in the short term, thereby reducing the likelihood of detecting differences [41, 42]. Finally, the high heterogeneity of study populations may also dilute true effects. Variability in injury level, injury severity, and residual neural function may influence responsiveness to tSCS, and in the absence of adequate stratified analyses, such differences may lead to an underestimation of overall effects [28, 43].

From a stratified perspective, variability in the effects of tSCS may be influenced by both patient characteristics and intervention parameters. First, regarding disease stage, this review suggests that individuals in the subacute phase may derive greater benefits than those in the chronic phase, which may be related to differences in neural plasticity. During the subacute phase, the nervous system retains greater plasticity and may be more responsive to external neuromodulatory stimulation, while a greater number of functional residual pathways are typically preserved, providing more favorable conditions for recovery. In contrast, neural plasticity declines in the chronic phase, and neural networks become more stable, potentially limiting the effects of neuromodulation [44]. Therefore, the efficacy of tSCS may be stage-dependent. Nevertheless, given the limited number of studies involving subacute populations, these findings should be interpreted with caution.

Second, regarding injury level and severity, most included studies involved mixed populations but were predominantly composed of individuals with cervical injuries, particularly lower cervical injuries. Given that patients with lower cervical SCI typically retain some degree of upper limb function while experiencing more severe impairments in lower limb function, this disparity may make improvements in upper limb function more readily detectable in clinical assessments, which is consistent with the findings of this review [45, 46]. Furthermore, the study by Huang et al. suggested that individuals with residual hand function may achieve greater functional improvements following the intervention, supporting the hypothesis that treatment effects may depend on the integrity of residual neural pathways [28]. However, this hypothesis requires further validation through prospective stratified studies.

Finally, regarding intervention implementation, variability in stimulation parameters and intervention dosage across studies may also represent a key source of inconsistency in treatment effects. tSCS primarily exerts its effects by activating dorsal root afferent fibers and modulating the excitability of spinal central pattern generators and associated neural networks. Its neuromodulatory effects are highly dependent on parameters such as stimulation intensity, frequency, and electrode configuration [47, 48]. However, there is currently no standardized stimulation protocol, and parameter settings are often individualized. Under these circumstances, different stimulation conditions across studies may result in varying degrees of neural recruitment and network modulation, thereby contributing to heterogeneity in outcomes [47]. In addition, evidence regarding dose–response relationships remains limited, which further constrains the accurate evaluation and interpretation of intervention effects.

Based on the current body of evidence, this study suggests that tSCS should be regarded as an exploratory intervention rather than a standard treatment suitable for routine clinical implementation. Given the very low certainty of the evidence and the lack of consistent improvements in key functional outcomes, its widespread clinical use is not supported at this stage. However, tSCS may still hold potential in specific patient populations. For example, in subacute patients or individuals with some preserved residual function, greater neural plasticity may render them more responsive to neuromodulation. In such cases, tSCS could be explored as an adjunct to rehabilitation training. Any such application should be conducted within well-designed clinical studies or under strict monitoring conditions.

Limitations and future directions

This review also has several limitations. First, there was substantial heterogeneity among the included studies in terms of stimulation parameters, intervention protocols, and outcome measures. In addition, most studies had small sample sizes and limited methodological quality, which restricted the robustness and comparability of the evidence. Second, due to the lack of sufficient long-term follow-up data, the current evidence primarily reflects short-term effects, and long-term outcomes remain unclear. Furthermore, variability in injury level, injury severity, and disease duration across the included populations, as well as the use of combined rehabilitation interventions in some studies, may have introduced potential confounding effects in the evaluation of treatment efficacy. Meanwhile, key electrophysiological measures were not systematically synthesized, limiting a deeper understanding of the underlying neural mechanisms. The lack of standardized reporting of stimulation parameters also made it difficult to assess potential dose–response relationships. Finally, publication bias and limitations in the search strategy may have further affected the completeness and generalizability of the findings.

Future research should prioritize multicenter, high-quality RCTs. Sample size calculations should be based on the MCID at the design stage, and blinded outcome assessment should be implemented to improve overall methodological rigor. In addition, standardized reporting of stimulation parameters, intervention protocols, and outcome measures is needed to enhance comparability across studies and improve the clinical interpretability of findings. Long-term follow-up should also be strengthened to clarify the durability of treatment effects. Given the heterogeneity of study populations, stratified analyses based on factors such as injury level, injury severity, and disease duration are recommended, along with efforts to control for confounding bias introduced by combined interventions. Furthermore, future studies should systematically incorporate electrophysiological measures, such as electromyography and motor evoked potentials, alongside other objective biomarkers to better elucidate underlying neurophysiological mechanisms. Building on this, the dose–response relationship between stimulation parameters and therapeutic outcomes should be explored, thereby providing a basis for optimizing intervention strategies.

Conclusion

In summary, current evidence is insufficient to draw definitive conclusions regarding the efficacy of tSCS. Although some studies suggest potential benefits in upper limb motor function as well as balance and postural control, the robustness of the existing evidence remains limited due to small sample sizes, methodological limitations, and substantial heterogeneity across studies. Further high-quality research is required to validate its efficacy and to elucidate the underlying mechanisms.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

SCI

Spinal cord injury

ADL

Activities of daily living

tSCS

Transcutaneous spinal cord stimulation

RCTs

Randomized controlled TTrials

MeSH

Medical subject headings

AIS

American Spinal Injury Association

RoB 2

Cochrane risk of bias tool

ROBINS-I

Risk of bias in non-randomized studies of interventions

JBI

Joanna Briggs Institute

GRADE

Grading of recommendations assessment, development and evaluation

UEMS

Upper Extremity Motor Score

GRASSP

Graded redefined assessment of strength, sensibility and prehension

CUE-T

Capabilities of upper extremity test

CUE-Q

Capabilities of upper extremity questionnaire

ROM

Range of motion

MVC

Maximum voluntary contraction

LEMS

Lower extremity motor score

10MWT

10-meter walk test

6MWT

6-minute walk test

2MWT

2-minute walk test

WISCI-II

Walking Index for Spinal Cord Injury II

TUG

Timed up and go test

MCID

Minimal clinically important difference

MDD

Minimal detectable difference

BBS

Berg Balance Scale

SCIM III

Spinal Cord Independence Measure III

MAS

Modified Ashworth Scale

PSFS

Penn Spasm Frequency Scale

mSCI-SET

Modified spinal cord injury spasticity evaluation tool

SCATS

Spinal cord assessment tool for spasticity

WHOQOL-BREF

World Health Organization Quality of Life-BREF

EQ-5D-5L

EuroQol 5-dimension 5-level

SF-36

Short Form-36 Health Survey

QLI-SCI

Quality of Life Index–Spinal Cord Injury Version

Author contributions

Jingjing Liu, Yinxu Wang, and Xin Chen contributed equally to the manuscript writing. Bangqiang Hou was responsible for the literature search. Jingyi Lin and Danlei Long conducted literature screening, data extraction, and methodological quality assessment. Wei Song and Yulei Xie provided guidance and support for the study design and manuscript preparation.

Funding

This study was supported by the 2025 Scientific Research Cooperation Special Project (First Batch) of NorthSichuan Medical College (Grant No. CBY25-ZXA01), the High-level Talent Research Start-up Fund under the 2025 Scientific Research Development Program of the Clinical Medical College and Affiliated Hospital, North Sichuan Medical College (Grant No. 2025GC016), and the 2025 University-level Scientific Research Innovation Team Project of North Sichuan Medical College (Grant No. CBYTD-2025A05).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.