Abstract
Objective
The aim of this study is to evaluate the effects of a unilateral‐lower limb exoskeleton rehabilitation training robot on balance and lower limb function after stroke.
Methods
In this pilot study, stroke patients with hemiplegic at Tongji University Yangzhi Rehabilitation Hospital (between October 2022 and December 2023) were randomly assigned to either the robot group or the control group. The primary measure was the Berg balance scale (BBS), whereas secondary measures included the 10‐m walking time test, 6‐min walk test, three‐dimensional gait analysis, and pain levels using the Hospital Anxiety and Depression Scale (HADS) and visual analogue scale (VAS).
Results
Among 36 post‐stroke hemiplegia patients (25 males; 18 per group), the robot group significantly outperformed the traditional group. The robot group improved BBS scores (5.50 ± 2.09 vs. 3.11 ± 1.50, p < 0.001), 6‐min walking distance (15.89 ± 8.64 m vs. 9.73 ± 5.83 m, p = 0.017), and 10‐m walking time reduction (8.82 ± 3.22 s vs. 6.83 ± 1.81 s, p = 0.030). Gait parameters also showed greater gains: hip flexion (3.57°±2.30° vs. 1.72°±2.15°, p = 0.018), knee flexion (5.28°±4.89° vs. 2.63°±2.37°, p = 0.046), and ankle dorsiflexion (4.81°±5.57° vs. 1.09°±4.19°, p = 0.030). Depression scores improved more in the robot group (2.89 ± 1.75 vs. 1.61 ± 1.20, p = 0.015). No significant differences occurred in anxiety or pain VAS scores (p > 0.05).
Conclusion
Integrating a unilateral‐lower limb exoskeleton rehabilitation robot with traditional therapy significantly enhances lower limb recovery in stroke patients with hemiplegia. Future large‐scale randomized controlled trials and follow‐up evaluations are needed to validate the current findings.
Keywords: rehabilitation, stroke, unilateral-lower limb exoskeleton rehabilitation training robot, walking function
1. Introduction
Stroke is a prevalent neurological condition characterized by a sudden onset of neurologic dysfunction attributable to focal brain ischemia [1] or brain hemorrhage [2, 3]. Approximately 25% of adults will experience a stroke during their lifetime, with about 70% of these events classified as ischemic strokes [4]. Stroke is associated with a high rate of disability, with up to 80% of affected individuals suffering from residual motor dysfunction post‐stroke [5]. Notably, lower limb motor impairments are particularly pronounced [6]. Such impairments often manifest as muscle weakness, inadequate weight transfer, restricted mobility, balance disorders, and gait dysfunction [7, 8], significantly compromising patients′ abilities to perform daily activities and diminishing their overall quality of life. Consequently, both stroke and the resultant disabilities impose a substantial burden on patients, their families, and society at large. Although some patients may experience spontaneous functional recovery after a stroke, timely and effective rehabilitation training remains indispensable for the majority [9]. The restoration of lower limb motor function is a critical component of the rehabilitation process for individuals with hemiplegic stroke [10, 11]. Early interventions emphasizing balance and gait training have been shown to yield positive outcomes [12].
Rehabilitation generally encompasses a comprehensive care approach aimed at reducing disability, enhancing participation in activities of daily living, improving functional abilities, preventing further functional deterioration, and maximizing independence [13, 14]. Physical therapy within rehabilitation includes positioning, supportive devices, and prescribed exercises, whereas additional treatment modalities for pain and spasticity may involve pharmacotherapy, botulinum toxin injections, nerve blocks, electrical stimulation, splints, serial casting, and surgical interventions [15–17]. Nevertheless, a significant proportion of stroke survivors will continue to experience some degree of neurological impairment 5 years post‐stroke [18, 19], indicating that novel complementary methods may be necessary to enhance patient outcomes following a stroke.
In recent years, the clinical application of exoskeleton lower limb robots has garnered increasing attention. These robotic devices deliver consistent rehabilitation support to patients through predefined parameters, mitigating issues such as variability in therapists′ skill levels and fatigue. In comparison to conventional rehabilitation methods, robotic interventions provide a standardized rehabilitation regimen [20]. Functional rehabilitation‐oriented lower limb exoskeleton robots can play a significant role in facilitating weight relief and promoting standardized physiological gait patterns during rehabilitation, thereby enabling patients to progressively regain lower limb motor function and attain autonomous walking following neurological injury [21, 22]. The LiteStepper unilateral‐lower limb exoskeleton rehabilitation training robot represents an innovative device for gait rehabilitation. Its primary operational principle involves the unaffected limb being equipped with a lightweight, imperceptible intelligent sensing system while the affected limb is supported by a powered unit. By integrating intelligent gait learning algorithms, patient motion intention recognition technology, multisensor fusion technology, and lightweight wearable exoskeleton fabrication techniques, the device enhances user experience and engagement in rehabilitation based on motor relearning theories. The system promotes the patient′s active learning capacity, improving balance and restoring normal gait through the interaction between the unaffected and affected limbs, as well as human‐machine integration [23, 24]. However, the exact clinical evidence supporting that the unilateral‐lower limb exoskeleton rehabilitation training robot is superior to conventional rehabilitation is still unclear, and further research is needed to verify this. Therefore, this study is aimed at evaluating the effects of a unilateral‐lower‐limb exoskeleton rehabilitation training robot on balance and lower limb function after stroke.
2. Material and Methods
2.1. Study Design and Participants
This pilot study included stroke patients with hemiplegia who were hospitalized at Tongji University Yangzhi Rehabilitation Hospital between October 2022 and December 2023 to participate in this pilot study. The study was approved by the Ethics Committee of Tongji University Yangzhi Rehabilitation Hospital (Approval No. #EC Form045). All participants provided written informed consent before being included in the study.
Stroke was diagnosed based on neurological examination, brain imaging (such as computed tomography [CT] or magnetic resonance imaging [MRI]). For ischemic stroke, evidence of acute cerebral infarction was required on imaging, and for hemorrhagic stroke, the presence of intracerebral hemorrhage was confirmed. Eligible stroke patients were informed about the trial. Inclusion criteria were as follows: (1) confirmed diagnosis of stroke with hemiplegic (ischemic or hemorrhagic) 3–12 months postacute onset; (2) ability to understand study procedures and provide informed consent; (3) age between 18 and 75 years; (4) stable vital signs and overall health status, capable of tolerating low‐intensity sit‐and‐stand training; (5) weight not exceeding 100 kg and height between 1.50 and 1.90 m; (6) ability to use a lower limb rehabilitation training robot with assistance; and (7) no significant limitations in passive range of motion for the hip, knee, and ankle joints, with the ankle in a neutral position during passive assessment.
Exclusion criteria included (1) severe joint mobility limitations impeding walking; (2) skin damage or infection at contact points with the rehabilitation robot; (3) severe internal medical conditions; (4) significant cognitive impairment interfering with training cooperation; (5) other contraindications or complications adversely affecting gait training; (6) pregnancy, planning pregnancy, or breastfeeding; (7) participation in another clinical trial within the month prior to enrollment; and (8) other conditions deemed inappropriate.
2.2. Interventions
Patients were randomly assigned to either the control group or the group receiving combined unilateral‐lower limb exoskeleton rehabilitation training robot‐assisted traditional rehabilitation (robot group), with 18 patients in each group. The control group′s training plan was developed based on the patient′s rehabilitation assessment results to select appropriate rehabilitation measures. The training content includes active movement, weight bearing on the affected limb, center of gravity transfer, balance function, knee joint control, and dorsiflexion strength training. Active movement involves flexion and extension exercises for the hip, knee, and ankle joints, with each set of exercises repeated 10–15 times, completing three sets. Weight bearing on the affected limb begins partially, gradually increasing the duration and proportion. Center of gravity transfer training is conducted between parallel bars, with each set repeated 10 to 15 times for a total of three sets. Balance training includes both static and dynamic components, with dynamic training involving obstacle crossing for approximately 20 min. Knee joint exercises use resistance bands to enhance muscle strength, with each set repeated 10–15 times for a total of three sets. Dorsiflexion strength training involves resistance bands and tiptoe standing to improve walking ability, with each set repeated 10–15 times for a total of three sets. These training exercises is aimed at comprehensively enhancing the patients′ motor functions.
Patients in the robot group received unilateral‐lower limb exoskeleton rehabilitation training robot training (LiteStepper; Zhengzhou Angelexo Scientific Co. Ltd., China) in addition to traditional rehabilitation training, under the guidance of professional therapists [25, 26]. The robot training sessions lasted 20 min each, once daily, for a total of 20 sessions. The unilateral‐lower limb exoskeleton rehabilitation robot‐assisted training begins with a warm‐up phase, during which the patient is assisted by a therapist to wear and adjust the exoskeleton device to ensure comfort and safety (Figure 1). During the training, the robot is programed to assist the patient in performing lower limb movement patterns, including hip flexion and extension, knee joint bending and straightening, and ankle dorsiflexion and plantarflexion. The speed and range of motion are adjusted according to the patient′s recovery level, adhering to the principles of progressive training while providing appropriate resistance during knee joint training to enhance muscle strength. The exoskeleton robot precisely controls the motion angles and forces of the ankle joint, providing targeted training for patients with varying degrees of injury. Through real‐time monitoring and data collection, therapists can promptly adjust training parameters to ensure the effectiveness and safety of the training. This real‐time monitoring and adjustment are key to optimizing exoskeleton‐assisted rehabilitation training [27].
Figure 1.
Schematic diagram of the unilateral‐lower limb exoskeleton rehabilitation training robot. (A). Unilateral‐lower limb exoskeleton rehabilitation training device. (B). Patient undergoing rehabilitation training with the unilateral‐lower limb exoskeleton device, assisted by a therapist.
2.3. Outcome Measurements
Evaluations were conducted prior to treatment and again 4 weeks posttreatment. The primary outcome measure was the Berg balance scale (BBS) scores. The secondary outcome measures included the time taken to complete a 10‐m walk, the distance covered in 6 min, parameters derived from three‐dimensional gait analysis, the psychological states of the patients, and reported levels of pain. Furthermore, any adverse reactions or complications observed during the treatment process were systematically documented. The evaluators involved in this study did not participate in the treatment process for the patients, and the data analysts were blinded to the group assignments to mitigate potential bias.
The BBS comprises 14 items, each evaluated on a scoring range from 0 to 4, yielding a maximum achievable score of 56 points. Elevated scores are indicative of superior balance function [28]. The three‐dimensional gait analysis encompassed measurements of hip flexion angle, knee flexion angle, ankle dorsiflexion angle, and pelvic tilt angle [29]. Psychological states were assessed utilizing the Hospital Anxiety and Depression Scale (HADS) [30]. The HADS, developed by Zigmond and Snaith in 1983 [31], measures depression and anxiety in hospitalized patients through 14 self‐reported items (seven pertaining to anxiety and seven to depression) rated on a 4‐point Likert scale (ranging from 0 to 3). This results in a scoring system for each subscale that spans from 0 to 21. The presence of anxiety and depression is categorized as absent, mild, moderate, or severe, corresponding to scores of 0–7, 8–10, 11–14, and 15–21, respectively [32]. Additionally, pain levels were evaluated using the visual analogue scale (VAS) [33]. The VAS consists of a 10‐cm line marked at each end with values of 0 and 10, along which the patient indicates the intensity of pain experienced. A ruler is employed to measure the distance from the 0 mark, thereby determining the pain score [34]. Pain is classified as absent, mild, or moderate to severe for scores of 0–0.9, 1.0–3.9, and 4.0–10.0 cm, respectively [35]. All assessments were conducted by the same therapist both prior to and following treatment to evaluate changes in lower limb motor function, balance function, and walking ability.
2.4. Sample Size
The sample size was calculated using professional statistical methods. For the comparison of continuous variables between the two groups in this study, a two‐independent‐sample approach was adopted. Based on pilot study data and clinical experience, the standard deviation (SD) of the primary measure variables was estimated to be around 2, and the minimum clinically meaningful difference between the two groups was set at 2. With a two‐sided significance level of α = 0.05 (corresponding to Z1 − α/2 = 1.96) and a power of 80% (1 − β = 0.8, corresponding to Z1 − β = 0.84), the initial calculation suggested that approximately 16 subjects were needed per group. However, considering a potential dropout rate of 10% during the study, adjustments were made. After accounting for this dropout rate, it was determined that 18 patients should be included in each group. In total, 36 patients were recruited for this study. This sample size was selected to ensure that the study had sufficient statistical power to detect significant differences between the two treatment groups.
2.5. Statistical Analysis
Statistical analysis was conducted using SPSS Version 22.0 (IBM Corp., Armonk, New York, United States). Continuous variables that followed a normal distribution and had homogeneity of variance were expressed as mean ± SD and analyzed using the t‐test. Nonnormally distributed continuous variables were expressed as median (interquartile range) and analyzed using the nonparametric rank‐sum test. Categorical variables were expressed as n (%), and group comparisons were conducted using the χ 2 test. A two‐sided p value of less than 0.05 was considered statistically significant. Two‐way repeated‐measures analysis of variance (ANOVA) was used to examine the Group × Time interaction effect of outcome measures, followed by paired t‐test for within‐group comparisons and independent samples t‐test for between‐group comparison of improvement amplitude.
3. Results
3.1. Characteristics of the Patients
A total of 36 stroke patients with hemiplegic in the recovery phase were included in the study, comprising 25 males and 11 females. Following the categorization based on their selected rehabilitation method, each group consisted of 18 patients. The detailed patient selection and allocation process was shown in Figure 2. Statistical analysis indicated no significant differences between the two groups regarding demographic characteristics (gender, age, duration of illness, or lesion location) and baseline functional outcome measures (BBS, gait parameters, HADS scores, VAS pain score) (Table 1). Stroke lesion topography was recorded and classified into basal ganglia lesions and nonbasal ganglia lesions. The nonbasal ganglia category included cortical lesions, subcortical white matter lesions and brainstem lesions. Given the limited sample size, further subdivision was not performed.
Figure 2.
Flowchart of the patient selection and allocation process in the study.
Table 1.
Baseline characteristics of patients in the two groups.
| Variable | Control (n = 18) | Robot (n = 18) | p |
|---|---|---|---|
| Demographic characteristics | |||
| Sex | |||
| Male | 12 (66.7%) | 13 (72.2%) | 0.717 |
| Female | 6 (33.3%) | 5 (27.8%) | |
| Age (years) | 56.60 ± 13.20 | 53.50 ± 12.90 | 0.481 |
| Disease duration (months) | 7.33 ± 3.19 | 7.40 ± 3.21 | 0.948 |
| Lesion location | |||
| Basal ganglia region | 10 (55.6%) | 12 (66.7%) | 0.494 |
| Nonbasal ganglia region (cortex, subcortical white matter, brainstem) | 8 (44.4%) | 6 (33.3%) | |
| Functional outcome measures at baseline | |||
| Berg score | 36.78 ± 4.91 | 38.94 ± 5.96 | 0.244 |
| 6‐min walking distance (m) | 112.33 ± 17.96 | 115.78 ± 19.14 | 0.581 |
| 10‐m walking time (s) | 34.08 ± 9.89 | 32.99 ± 10.10 | 0.746 |
| Hip flexion angle (°) | 6.51 ± 2.30 | 6.36 ± 3.33 | 0.876 |
| Knee flexion angle (°) | 15.46 ± 6.95 | 15.58 ± 7.52 | 0.961 |
| Pelvic tilt angle (°) | 4.26 ± 2.33 | 4.13 ± 2.34 | 0.868 |
| Ankle dorsiflexion angle (°) | 30.76 ± 9.04 | 31.28 ± 8.59 | 0.861 |
| HADS/A | 8.67 ± 2.22 | 8.72 ± 2.30 | 0.948 |
| HADS/D | 14.56 ± 2.36 | 14.94 ± 2.56 | 0.646 |
| VAS | 3.61 ± 1.20 | 3.83 ± 1.25 | 0.594 |
Note: Chi‐square test for categorical variables, independent samples t‐test for continuous variables.
Abbreviations: HADS/A, Hospital Anxiety and Depression Scale, anxiety; HADS/D, Hospital Anxiety and Depression Scale, depression; VAS, visual analogue scale.
Two‐way repeated‐measures ANOVA revealed significant Group × Time interaction effects for BBS, 6‐min walking distance, 10‐m walking time, hip flexion angle, knee flexion angle, ankle dorsiflexion angle, and HADS‐D score (all p < 0.05), whereas no significant interaction was observed for HADS‐A and VAS scores (p > 0.05) (Table 2). After 4 weeks of treatment combining unilateral‐lower limb exoskeleton rehabilitation robot with conventional therapy, patients demonstrated significant improvements in balance function, increasing from 38.94 ± 5.96 to 44.44 ± 5.48 (p = 0.007); 6‐minute walking distance, extending from 115.78 ± 19.14 m to 131.67 ± 25.11 m (p = 0.040); and walking time, decreasing from 32.99 ± 10.10 s to 24.17 ± 9.96 s (p = 0.013). Joint mobility also improved significantly, with hip flexion angle increasing from 6.36°±3.33° to 9.93°±4.87° (p < 0.001); knee joint angle from 15.58°±7.52° to 20.86°±8.69° (p < 0.001); pelvic mobility from 4.13°±2.34° to 6.71°±2.73° (p < 0.001); and ankle joint angle changing from 31.28°±8.59° to 26.46°±10.25° (p = 0.002). Additionally, depressive symptoms were significantly reduced (p < 0.001). However, no significant changes were observed in anxiety and pain scores (p > 0.05) (Table 3). In contrast, patients receiving conventional therapy alone showed improvements only in walking time (from 34.08 ± 9.89 s to 27.25 ± 9.11 s, p = 0.040), knee joint angle (from 15.46°±6.95° to 18.09°±7.46°, p = 0.036), ankle joint mobility (from 30.76°±9.04° to 28.85°± 6.21°, p = 0.002), and depressive symptoms (p > 0.05), with no significant changes in balance function (Table 4).
Table 2.
Group × Time interaction effects of two‐way repeated‐measures ANOVA for all outcome measures.
| Outcomes | Degree of freedom | F | p |
|---|---|---|---|
| Berg score | 1, 34 | 15.54 | < 0.001 |
| 6‐min walking distance (m) | 1, 34 | 6.28 | 0.017 |
| 10‐m walking time (s) | 1, 34 | 5.23 | 0.030 |
| Hip flexion angle (°) | 1, 34 | 6.21 | 0.018 |
| Knee flexion angle (°) | 1, 34 | 4.28 | 0.046 |
| Pelvic tilt angle (°) | 1, 34 | 2.05 | 0.161 |
| Ankle dorsiflexion angle (°) | 1, 34 | 5.13 | 0.030 |
| HADS/A | 1, 34 | 0.02 | 0.893 |
| HADS/D | 1, 34 | 6.57 | 0.015 |
| VAS | 1, 34 | 0.02 | 0.894 |
Note: All effects represent Group × Time interaction; df = 1, 34 (between‐group df = 1, error df = 34); p < 0.05 was considered statistically significant.
Abbreviations: HADS/A, Hospital Anxiety and Depression Scale, anxiety; HADS/D, Hospital Anxiety and Depression Scale, depression; VAS, visual analogue scale.
Table 3.
Comparison of the outcomes before and after treatment in the robot group.
| Outcomes | Robot rehabilitation (n = 18) | t | p | |
|---|---|---|---|---|
| Before | After | |||
| Berg score | 38.94 ± 5.96 | 44.44 ± 5.48 | 2.882 | 0.007 |
| 6‐min walking distance (m) | 115.78 ± 19.14 | 131.67 ± 25.11 | 2.135 | 0.040 |
| 10‐m walking time (s) | 32.99 ± 10.10 | 24.17 ± 9.96 | 2.638 | 0.013 |
| Hip flexion angle (°) | 6.36 ± 3.33 | 9.93 ± 4.87 | 6.581 | < 0.001 |
| Knee flexion angle (°) | 15.58 ± 7.52 | 20.86 ± 8.69 | 4.577 | < 0.001 |
| Pelvic tilt angle (°) | 4.13 ± 2.34 | 6.71 ± 2.73 | 6.615 | < 0.001 |
| Ankle dorsiflexion angle (°) | 31.28 ± 8.59 | 26.46 ± 10.25 | 3.665 | 0.002 |
| HADS/A | 8.72 ± 2.30 | 8.28 ± 1.57 | 1.512 | 0.149 |
| HADS/D | 14.94 ± 2.56 | 12.06 ± 2.24 | 7.023 | < 0.001 |
| VAS | 3.83 ± 1.25 | 3.67 ± 0.77 | 0.511 | 0.616 |
Note: p < 0.05 was considered statistically significant.
Abbreviations: HADS/A, Hospital Anxiety and Depression Scale, anxiety; HADS/D, Hospital Anxiety and Depression Scale, depression; VAS, visual analogue scale.
Table 4.
Comparison of the outcomes before and after treatment in the traditional rehabilitation group.
| Outcomes | Traditional rehabilitation (n = 18) | t | P | |
|---|---|---|---|---|
| Before | After | |||
| Berg score | 36.78 ± 4.91 | 39.89 ± 4.85 | 1.912 | 0.064 |
| 6‐min walking distance (m) | 112.33 ± 17.96 | 122.06 ± 22.30 | 1.442 | 0.159 |
| 10‐m walking time (s) | 34.08 ± 9.89 | 27.25 ± 9.11 | 2.157 | 0.040 |
| Hip flexion angle (°) | 6.51 ± 2.30 | 8.23 ± 2.83 | 1.507 | 0.229 |
| Knee flexion angle (°) | 15.46 ± 6.95 | 18.09 ± 7.46 | 3.627 | 0.036 |
| Pelvic tilt angle (°) | 4.26 ± 2.33 | 5.70 ± 3.54 | 3.058 | 0.055 |
| Ankle dorsiflexion angle (°) | 30.76 ± 9.04 | 28.85 ± 6.21 | 11.180 | 0.002 |
| HADS/A | 8.67 ± 2.22 | 8.17 ± 1.82 | 1.767 | 0.095 |
| HADS/D | 14.56 ± 2.36 | 12.94 ± 2.26 | 5.720 | < 0.001 |
| VAS | 3.61 ± 1.20 | 3.50 ± 0.86 | 0.437 | 0.668 |
Note: p < 0.05 was considered statistically significant.
Abbreviations: HADS/A, Hospital Anxiety and Depression Scale, anxiety; HADS/D, Hospital Anxiety and Depression Scale, depression; VAS, visual analogue scale.
The difference in BBS score changes between groups (traditional group 3.11 ± 1.50 vs. robot group 5.50 ± 2.09) was statistically significant (p < 0.001), indicating more substantial benefits for patients receiving robot‐assisted conventional rehabilitation therapy. Compared with the control group, the robot group demonstrated greater improvements in 6‐min walking test distance change (9.73 ± 5.83 m vs. 15.89 ± 8.64 m, p = 0.017), 10‐m walking time reduction (6.83 ± 1.81 s vs. 8.82 ± 3.22 s, p = 0.030), and all gait parameters, including hip flexion angle (1.72°±2.15° vs. 3.57°±2.30°, p = 0.018), knee flexion angle (2.63°±2.37° vs. 5.28°±4.89°, p = 0.046), and ankle dorsiflexion angle (1.09°±4.19° vs. 4.81°±5.57°, p = 0.030). Furthermore, the robot group exhibited significantly greater improvement in depressive symptoms compared with the traditional group (1.61 ± 1.20 vs. 2.89 ± 1.75, p = 0.015). However, differences in anxiety and pain VAS score changes between the two groups were not statistically significant (all p > 0.05) (Table 5).
Table 5.
Comparison of improvement amplitude of the outcomes between the two groups.
| Outcomes | Changes after treatment | t | p | |
|---|---|---|---|---|
| Traditional (n = 18) | Robot (n = 18) | |||
| Berg score | 3.11 ± 1.50 | 5.50 ± 2.09 | 3.942 | < 0.001 |
| 6‐min walking distance (m) | 9.73 ± 5.83 | 15.89 ± 8.64 | 2.507 | 0.017 |
| 10‐m walking time (s) | 6.83 ± 1.81 | 8.82 ± 3.22 | 2.286 | 0.030 |
| Hip flexion angle (°) | 1.72 ± 2.15 | 3.57 ± 2.30 | 2.493 | 0.018 |
| Knee flexion angle (°) | 2.63 ± 2.37 | 5.28 ± 4.89 | 2.069 | 0.046 |
| Pelvic tilt angle (°) | 1.44 ± 2.94 | 2.58 ± 1.66 | 1.433 | 0.161 |
| Ankle dorsiflexion angle (°) | 1.09 ± 4.19 | 4.81 ± 5.57 | 2.264 | 0.030 |
| HADS/A | 0.50 ± 1.20 | 0.44 ± 1.25 | 0.136 | 0.893 |
| HADS/D | 1.61 ± 1.20 | 2.89 ± 1.75 | 2.563 | 0.015 |
| VAS | 0.11 ± 1.08 | 0.17 ± 1.38 | 0.134 | 0.894 |
Note: p < 0.05 was considered statistically significant.
Abbreviations: HADS/A, Hospital Anxiety and Depression Scale, anxiety; HADS/D, Hospital Anxiety and Depression Scale, depression; VAS, visual analogue scale.
4. Discussion
The results indicated that the use of a unilateral‐lower limb exoskeleton rehabilitation training robot, combined with traditional rehabilitation therapy, significantly enhanced the lower limb motor function in stroke patients compared with traditional rehabilitation therapy alone. Specifically, the use of the robot provided additional benefits in joint range of motion (hip, knee, and ankle), balance function, walking speed, and walking distance of the affected limb, while also significantly improving depressive symptoms.
Research conducted by Jin et al. indicates that robot‐assisted traditional rehabilitation therapy surpasses conventional rehabilitation in terms of the BBS, Functional Ambulation Classification (FAC) scale, 6‐min walk test, and Barthel index [23]. Similarly, another trial focusing on subacute stroke patients demonstrated that machine‐assisted rehabilitation yielded superior improvements in BBS, FAC, 6‐min walk test, lower limb Fugl‐Meyer assessment, and modified Barthel index when compared with traditional rehabilitation methods [24]. This study reaffirmed these findings and further investigated the specific benefits for patients receiving robot‐assisted traditional rehabilitation therapy, particularly regarding the joint range of motion in the affected limb. Such improvements may be attributed to the high repeatability, efficiency, and consistency of the unilateral‐lower limb exoskeleton rehabilitation robot [20]. During training, this technology provides strength support, facilitating precise and repetitive walking task training. Personalized exercise programs guide the rehabilitation process, thereby promoting the restoration of normal gait. Additionally, the exoskeleton robot can objectively and continuously monitor patient metrics and progress throughout the rehabilitation process, offering active, assistive, and passive movement modes through a real‐time feedback system to maximize improvements in muscle motor function and coordination in the affected limb [36].
In contrast, traditional rehabilitation training involves therapists manually controlling the patient’s pelvis, hip, knee, and ankle joints to simulate natural repetitive walking. This method imposes certain restrictions on the physiological movement of the pelvis and trunk, making it challenging to organically integrate the three essential elements of walking: weight bearing, balance, and stepping [37]. Furthermore, manual joint control results in a lack of systematic internal and external sensory feedback, leading to actions that are neither natural nor coherent. Consequently, although traditional rehabilitation training can enhance lower limb function in stroke patients, its therapeutic outcomes are often suboptimal. This method also necessitates significant physical effort from therapists and fails to engage patients sufficiently, resulting in a lack of interest and motivation. Although improvements in gait and mobility are noted as training duration increases, high‐intensity and repetitive training can eventually become demanding and may lead to abnormal gait patterns [38, 39]. Therefore, in this study, although traditional rehabilitation therapy improved the 10‐m walking time and knee and ankle flexion angles, it had no significant impact on BBS, the distance of the 6‐min walk test, hip flexion angle, or pelvic tilt angle.
Additionally, it is noteworthy that this study assessed the psychological state of patients. Compared with traditional rehabilitation therapy, the robot‐assisted rehabilitation group demonstrated significantly greater improvement in depressive symptoms following treatment. After experiencing illness, patients frequently encounter limitations in physical health and activity levels, which can result in increased dependency, exacerbated depression, and diminished quality of life. The integration of robotic and traditional rehabilitation therapies offers substantial support by compensating for diminished intrinsic functions, thereby facilitating enhancements in patients′ psychological well‐being [40]. This finding suggests that future rehabilitation strategies should prioritize the incorporation of exoskeletons and the development of tailored rehabilitation programs to enhance both the physical and mental well‐being of patients.
Furthermore, the study carries several limitations. Only short‐term rehabilitation outcomes were assessed, and long‐term effects remain unknown. In addition, we primarily focused on lower limb motor recovery without systematic evaluation of sensory function, interlimb coordination and other domains. The relatively small sample and narrow participant spectrum may limit the generalizability of our findings. As an observational investigation, treatment selection depended on participant preference; accordingly, disparities in socioeconomic background or understanding of robotic rehabilitation could confound clinical outcomes. Clinically, lacunar infarction frequently produces mild deficits without overt hemiplegia, and for this reason, unilateral‐lower limb exoskeleton training yields limited benefits for such patients. We thus enrolled only participants with hemiplegia secondary to nonlacunar ischemic stroke and excluded those with pure motor lacunar stroke. Whether rehabilitation trajectories differ between lacunar and nonlacunar groups merits further scrutiny, given that lacunar lesions typically predict superior motor recovery [41]. When feasible, future trials may include eligible lacunar stroke patients with hemiplegia for comparative analysis.
Another notable constraint concerns the age spectrum of recruited patients. Participant entry was restricted to ages 18–75 years, whereas no adults aged ≥ 85 years were enrolled. Adults older than 85 years have diminished physical reserve, elevated fall risk and reduced motor plasticity [42], which means protocols optimized for younger adults cannot be directly transferred to this population. Therefore, developing customized low‐intensity, lightweight exoskeleton regimens tailored for the oldest‐old stroke survivors warrants dedicated research.
Taken together, large‐scale, multicenter randomized controlled trials covering varied stroke subtypes and severity grades are required. Extended follow‐up and multidimensional outcome assessment will strengthen the overall evidence base. Upcoming work can incorporate longitudinal monitoring, multimodal neuroimaging such as fMRI and DTI to unpack underlying mechanisms of neural plasticity, alongside adaptive training algorithms modulated according to individual motor performance and fatigue. Beyond that, trials may also test combinatorial therapeutic approaches to identify potential synergistic effects when exoskeleton training is paired with neuromuscular electrical stimulation, mirror therapy or cognitive rehabilitation. Collectively, such efforts will validate our results and define the broader clinical scope for robotic‐assisted rehabilitation.
5. Conclusion
In conclusion, this study provides evidence that the integration of a unilateral‐lower limb exoskeleton rehabilitation robot with traditional rehabilitation training facilitates a more effective recovery of lower limb motor function in stroke patients with hemiplegia compared with traditional rehabilitation training alone. Furthermore, the findings suggest that future rehabilitation strategies should incorporate exoskeleton technologies and emphasize the development of individualized rehabilitation programs to enhance both the physical and mental well‐being of patients.
Author Contributions
Shaoting Zhou, Sen Wang, and Xueying Qu carried out the studies and participated in collecting data. Shaoting Zhou and Jing Fan performed the statistical analysis and drafted the manuscript. Hanlin Xu and Linyu Chen contributed to the assessment of patient rehabilitation. Lingyu Liu contributed to manuscript editing, conceptualization, and supervision. Shaoting Zhou and Sen Wang contributed equally to this work.
Funding
This study was supported by the National clinical key specialty construction project of China (Z155080000004); Songjiang District Science and Technology Commission Project of Shanghai (2023SIKIGG105); Sunshine Clinical Research Cultivation Project of Yangzhi Rehabilitation Hospital Affiliated to Tongji University (2024CRTS003); and Shanghai Municipal Health Commission (10.13039/100017950, SHKFJS002A).
Disclosure
All authors read and approved the final manuscript.
Ethics Statement
The study was approved by the Ethics Committee of Tongji University Yangzhi Rehabilitation Hospital (Approval No. #EC Form045). All participants provided written informed consent before being included in the study. I confirm that all methods were performed in accordance with the relevant guidelines. All procedures were performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments.
Consent
The authors have nothing to report.
Conflicts of Interest
The authors declare no conflicts of interest.
Acknowledgments
The authors acknowledge the help of Zhengzhou Angelexo Scientific Co. Ltd., China for the robot device LiteStepper.
Zhou, Shaoting , Wang, Sen , Qu, Xueying , Fan, Jing , Xu, Hanlin , Chen, Linyu , Liu, Lingyu , Unilateral Exoskeleton Training for Balance/Gait in Hemiplegic Stroke: Pilot Study, Behavioural Neurology, 2026, 4208995, 10 pages, 2026. 10.1155/bn/4208995
Academic Editor: Nikhat Kaura
Contributor Information
Lingyu Liu, Email: happyneurologist@163.com.
Nikhat Kaura, Email: nkaura@wiley.com.
Data Availability Statement
All data generated or analyzed during this study are included in this published article.
References
- 1. Mendelson S. J. and Prabhakaran S., Diagnosis and Management of Transient Ischemic Attack and Acute Ischemic Stroke, Journal of the American Medical Association. (2021) 325, no. 11, 1088–1098, 10.1001/jama.2020.26867. [DOI] [PubMed] [Google Scholar]
- 2. Puy L., Parry-Jones A. R., Sandset E. C., Dowlatshahi D., Ziai W., and Cordonnier C., Intracerebral Haemorrhage, Nature Reviews Disease Primers. (2023) 9, 10.1038/s41572-023-00424-7. [DOI] [PubMed] [Google Scholar]
- 3. Macdonald R. L. and Schweizer T. A., Spontaneous Subarachnoid Haemorrhage, Lancet. (2017) 389, no. 10069, 655–666, 10.1016/S0140-6736(16)30668-7. [DOI] [PubMed] [Google Scholar]
- 4. Campbell B. C. V., De Silva D. A., Macleod M. R., Coutts S. B., Schwamm L. H., and Davis S. M., Ischaemic Stroke, Nature Reviews Disease Primers. (2019) 5, 10.1038/s41572-019-0118-8. [DOI] [PubMed] [Google Scholar]
- 5. Wang W., Jiang B., Sun H., Ru X., Sun D., Wang L., Wang L., Jiang Y., Li Y., Wang Y., Chen Z., Wu S., Zhang Y., Wang D., Wang Y., Feigin V. L., and NESS-China Investigators, Prevalence, Incidence, and Mortality of Stroke in China: Results From a Nationwide Population-Based Survey of 480 687 Adults, Circulation. (2017) 135, no. 8, 759–771, 10.1161/CIRCULATIONAHA.116.025250, 28052979. [DOI] [PubMed] [Google Scholar]
- 6. Sanchez N., Acosta A. M., Lopez-Rosado R., Stienen A. H. A., and Dewald J. P. A., Lower Extremity Motor Impairments in Ambulatory Chronic Hemiparetic Stroke: Evidence for Lower Extremity Weakness and Abnormal Muscle and Joint Torque Coupling Patterns, Neurorehabilitation and Neural Repair. (2017) 31, 814–826, 10.1177/1545968317721974. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Winstein C. J., Stein J., Arena R., Bates B., Cherney L. R., Cramer S. C., Deruyter F., Eng J. J., Fisher B., Harvey R. L., Lang C. E., MacKay-Lyons M., Ottenbacher K. J., Pugh S., Reeves M. J., Richards L. G., Stiers W., Zorowitz R. D., and American Heart Association Stroke Council, Council on Cardiovascular and Stroke Nursing, Council on Clinical Cardiology, and Council on Quality of Care and Outcomes Research, Guidelines for Adult Stroke Rehabilitation and Recovery: A Guideline for Healthcare Professionals From the American Heart Association/American Stroke Association, Stroke. (2016) 47, no. 6, e98–e169, 10.1161/STR.0000000000000098, 27145936. [DOI] [PubMed] [Google Scholar]
- 8. Darak V. and Karthikbabu S., Lower Limb Motor Function and Hip Muscle Weakness in Stroke Survivors and Their Relationship With Pelvic Tilt, Weight-Bearing Asymmetry, and Gait Speed: A Cross-Sectional Study, Current Journal of Neurology. (2020) 19, 1–7, 10.18502/ijnl.v19i1.3275. [DOI] [Google Scholar]
- 9. Grefkes C. and Fink G. R., Recovery From Stroke: Current Concepts and Future Perspectives, Neurological Research and Practice. (2020) 2, no. 1, 10.1186/s42466-020-00060-6, 33324923. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Yoo Y. J. and Lim S. H., Assessment of Lower Limb Motor Function, Ambulation, and Balance After Stroke, Brain & Neurorehabilitation. (2022) 15, no. 2, 10.12786/bn.2022.15.e17, 36743203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Kim Y. W., Update on Stroke Rehabilitation in Motor Impairment, Brain & Neurorehabilitation. (2022) 15, no. 2, 10.12786/bn.2022.15.e12, 36743199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Hatem S. M., Saussez G., Della Faille M., Prist V., Zhang X., Dispa D., and Bleyenheuft Y., Rehabilitation of Motor Function After Stroke: A Multiple Systematic Review Focused on Techniques to Stimulate Upper Extremity Recovery, Frontiers in Human Neuroscience. (2016) 10, 10.3389/fnhum.2016.00442, 27679565. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Dobkin B. H., Clinical Practice. Rehabilitation After Stroke, New England Journal of Medicine. (2005) 352, 1677–1684, 10.1056/NEJMcp043511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Shahid J., Kashif A., and Shahid M. K., A Comprehensive Review of Physical Therapy Interventions for Stroke Rehabilitation: Impairment-Based Approaches and Functional Goals, Brain Sciences. (2023) 13, no. 5, 10.3390/brainsci13050717, 37239189. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. National Institute for Health and Care Excellence (NICE), Stroke Rehabilitation in Adults, 2023, National Institute for Health and Care Excellence. [PubMed] [Google Scholar]
- 16. Sall J., Eapen B. C., Tran J. E., Bowles A. O., Bursaw A., and Rodgers M. E., The Management of Stroke Rehabilitation: A Synopsis of the 2019 U.S. Department of Veterans Affairs and U.S. Department of Defense Clinical Practice Guideline, Annals of Internal Medicine. (2019) 171, no. 12, 916–924, 10.7326/M19-1695, 31739317. [DOI] [PubMed] [Google Scholar]
- 17. Cai Y., Zhang C. S., Liu S., Wen Z., Zhang A. L., Guo X., Lu C., and Xue C. C., Electroacupuncture for Poststroke Spasticity: A Systematic Review and Meta-Analysis, Archives of Physical Medicine and Rehabilitation. (2017) 98, no. 12, e2574, 10.1016/j.apmr.2017.03.023, 28455191. [DOI] [PubMed] [Google Scholar]
- 18. Feigin V. L., Barker-Collo S., Parag V., Senior H., Lawes C. M., Ratnasabapathy Y., Glen E., and For the ASTRO study group, Auckland Stroke Outcomes Study. Part 1: Gender, Stroke Types, Ethnicity, and 450 Functional Outcomes 5 Years Poststroke, Neurology. (2010) 75, no. 18, 1597–1607, 10.1212/WNL.0b013e3181fb44b3. [DOI] [PubMed] [Google Scholar]
- 19. Barker-Collo S., Feigin V. L., Parag V., Lawes C. M., and Senior H., Auckland Stroke 453 Outcomes Study. Part 2: Cognition and Functional Outcomes 5 Years Poststroke, Neurology. (2010) 75, no. 18, 1608–1616, 10.1212/WNL.0b013e3181fb44c8. [DOI] [PubMed] [Google Scholar]
- 20. Masengo G., Zhang X., Dong R., Alhassan A. B., Hamza K., and Mudaheranwa E., Lower Limb Exoskeleton Robot and Its Cooperative Control: A Review, Trends, and Challenges for Future Research, Frontiers in Neurorobotics. (2022) 16, 913748, 10.3389/fnbot.2022.913748, 36714152. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. He Y., Xu Y., Hai M., Feng Y., Liu P., Chen Z., and Duan W., Exoskeleton-Assisted Rehabilitation and Neuroplasticity in Spinal Cord Injury, World Neurosurgery. (2024) 185, 45–54, 10.1016/j.wneu.2024.01.167, 38320651. [DOI] [PubMed] [Google Scholar]
- 22. de Miguel-Fernandez J., Lobo-Prat J., Prinsen E., Font-Llagunes J. M., and Marchal-Crespo L., Control Strategies Used in Lower Limb Exoskeletons for Gait Rehabilitation After Brain Injury: A Systematic Review and Analysis of Clinical Effectiveness, Journal of Neuroengineering and Rehabilitation. (2023) 20, no. 1, 10.1186/s12984-023-01144-5, 36805777. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Jin Y., Xiong B., Chen L., Zhao W., Li Z., Zhang C., and Xu X., Efficacy and Safety of Using a Unilateral Lower Limb Exoskeleton Combined With Conventional Treatment in Post-Stroke Rehabilitation: A Randomized Controlled Trial, Frontiers in Bioengineering and Biotechnology. (2024) 12, 1441986, 10.3389/fbioe.2024.1441986, 39359258. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Huo C., Shao G., Chen T., Li W., Wang J., Xie H., Wang Y., Li Z., Zheng P., Li L., and Li L., Effectiveness of Unilateral Lower-Limb Exoskeleton Robot on Balance and Gait Recovery and Neuroplasticity in Patients With Subacute Stroke: A Randomized Controlled Trial, Journal of Neuroengineering and Rehabilitation. (2024) 21, no. 1, 10.1186/s12984-024-01493-9, 39639336. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Rodriguez-Fernandez A., Lobo-Prat J., and Font-Llagunes J. M., Systematic Review on Wearable Lower-Limb Exoskeletons for Gait Training in Neuromuscular Impairments, Journal of Neuroengineering and Rehabilitation. (2021) 18, no. 1, 10.1186/s12984-021-00815-5, 33526065. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Su D., Hu Z., Wu J., Shang P., and Luo Z., Review of Adaptive Control for Stroke Lower Limb Exoskeleton Rehabilitation Robot Based on Motion Intention Recognition, Frontiers in Neurorobotics. (2023) 17, 1186175, 10.3389/fnbot.2023.1186175, 37465413. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Zhang Y., Zhao W., Wan C., Wu X., Huang J., Wang X., Huang G., Ding W., Chen Y., Yang J., Su B., Xu Y., Zhou Z., Zhang X., Miao F., Li J., and Li Y., Exoskeleton Rehabilitation Robot Training for Balance and Lower Limb Function in Sub-Acute Stroke Patients: A Pilot, Randomized Controlled trial, Journal of Neuroengineering and Rehabilitation. (2024) 21, no. 1, 10.1186/s12984-024-01391-0, 38851703. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Berg K. O., Wood-Dauphinee S. L., Williams J. I., and Maki B., Measuring Balance in the Elderly: Validation of an Instrument, Canadian Journal of Public Health. (1992) 83, no. Supplement 2, S7–S11. [PubMed] [Google Scholar]
- 29. McGrath M., Wood J., Walsh J., and Window P., The Impact of Three-Dimensional Gait Analysis in Adults With Pathological Gait on Management Recommendations, Gait & posture. (2023) 105, 75–80, 10.1016/j.gaitpost.2023.06.014, 37490826. [DOI] [PubMed] [Google Scholar]
- 30. Sagen U., Vik T. G., Moum T., Morland T., Finset A., and Dammen T., Screening for Anxiety and Depression After Stroke: Comparison of the Hospital Anxiety and Depression Scale and the Montgomery and Asberg Depression Rating Scale, Journal of Psychosomatic Research. (2009) 67, no. 4, 325–332, 10.1016/j.jpsychores.2009.03.007, 19773025. [DOI] [PubMed] [Google Scholar]
- 31. Zigmond A. S. and Snaith R. P., The Hospital Anxiety and Depression Scale, Acta Psychiatrica Scandinavica. (1983) 67, no. 6, 361–370, 10.1111/j.1600-0447.1983.tb09716.x. [DOI] [PubMed] [Google Scholar]
- 32. Stern A. F., The Hospital Anxiety and Depression Scale, Occupational Medicine. (2014) 64, no. 5, 393–394, 10.1093/occmed/kqu024. [DOI] [PubMed] [Google Scholar]
- 33. Price C. I., Curless R. H., and Rodgers H., Can Stroke Patients Use Visual Analogue Scales?, Stroke. (1999) 30, no. 7, 1357–1361, 10.1161/01.str.30.7.1357. [DOI] [PubMed] [Google Scholar]
- 34. Langley G. B. and Sheppeard H., The Visual Analogue Scale: Its Use in Pain Measurement, Rheumatology International. (1985) 5, no. 4, 145–148, 10.1007/BF00541514. [DOI] [PubMed] [Google Scholar]
- 35. Jonsson A. C., Lindgren I., Hallstrom B., Norrving B., and Lindgren A., Prevalence and Intensity of Pain After Stroke: A Population Based Study Focusing on Patients′ Perspectives, Journal of Neurology, Neurosurgery & Psychiatry. (2006) 77, no. 5, 590–595, 10.1136/jnnp.2005.079145, 16354737. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Kim D. H., Kang C. S., and Kyeong S., Robot-Assisted Gait Training Promotes Brain Reorganization After Stroke: A Randomized Controlled Pilot Study, NeuroRehabilitation. (2020) 46, no. 4, 483–489, 10.3233/NRE-203054, 32508339. [DOI] [PubMed] [Google Scholar]
- 37. Eng J. J. and Tang P. F., Gait Training Strategies to Optimize Walking Ability in People With Stroke: A Synthesis of the Evidence, Expert review of Neurotherapeutics. (2007) 7, 1417–1436, 10.1586/14737175.7.10.1417. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Beyaert C., Vasa R., and Frykberg G. E., Gait Post-Stroke: Pathophysiology and Rehabilitation Strategies, Neurophysiologie Clinique/Clinical Neurophysiology. (2015) 45, no. 4-5, 335–355, 10.1016/j.neucli.2015.09.005. [DOI] [PubMed] [Google Scholar]
- 39. Gandhi D. B., Sterba A., Khatter H., and Pandian J. D., Mirror Therapy in Stroke Rehabilitation: Current Perspectives, Therapeutics and Clinical Risk Management. (2020) 16, 75–85, 10.2147/TCRM.S206883, 32103968. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Claeys R., Embrechts E., Firouzi M., De Vlieger D., Verstraten T., Beckwee D., and Swinnen E., The Potential of Lower Limb Exoskeletons to Enhance Life-Space Mobility and to Leverage Green Exercise in the Rehabilitation of Older Adults: An Expert Perspective, Disability and Rehabilitation. (2025) 47, no. 15, 4002–4009, 10.1080/09638288.2024.2436981, 39641353. [DOI] [PubMed] [Google Scholar]
- 41. Arboix A., Padilla I., Massons J., García-Eroles L., Comes E., and Targa C., Clinical Study of 222 Patients With Pure Motor Stroke, Journal of Neurology, Neurosurgery & Psychiatry. (2001) 71, no. 2, 239–242, 10.1136/jnnp.71.2.239. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Torres-Riera S., Arboix A., Parra O., García-Eroles L., and Sánchez-López M. J., Predictive Clinical Factors of In-Hospital Mortality in Women Aged 85 Years or More With Acute Ischemic Stroke, Cerebrovascular Diseases. (2024) 54, no. 1, 1–19, 10.1159/000536436. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
All data generated or analyzed during this study are included in this published article.