Physical Therapy Korea

After stroke, lower-limb paralysis can severely impair a patient’s gait and balance, limiting their independence in daily life. Stroke significantly affects the brain’s motor planning and execution capabilities, which are essential for performing dynamic movements, such as walking [1]. These activities require coordination across multiple muscle groups to maintain balance, while impairment of these abilities due to stroke increases the risk of gait instability and falls [2].

Lower limb strength and sensory functions are crucial for maintaining walking and balance. Lower-limb weakness and sensory loss due to stroke make balance challenging during weight shifts or postural changes, thereby increasing the risk of falls. In particular, damage to the foot and ankle functions can make it difficult for patients to adapt to changes in surface conditions, resulting in balance issues on uneven or inclined surfaces [2,3].

Proprioception plays a critical role in stroke patients who struggle with gait and balance owing to lower-limb paralysis. Proprioception involves sensory information from the muscles, joints, and tendons that allow one to perceive body position and movement, which are essential for balance and walking. Proprioception further enables the real-time detection of joint position and movement, which aids in postural adjustment and balance maintenance. For example, if a patient cannot accurately perceive the position or speed of their joints while walking, weight-shifting and postural adjustments become challenging, increasing the likelihood of falls. Stroke patients with impaired proprioception may further face difficulties in stable weight bearing and balance, leading to asymmetric gait patterns and a higher risk of falls [4-6].

Proprioception enables the body to adapt to environmental changes. When proprioception is impaired, patients may not respond appropriately to changes in surface conditions, resulting in instability. Therefore, enhancing proprioception is crucial for improving gait and balance recovery in patients with stroke [6,7].

Traditional physical therapy gait training is tailored to each patient, focusing on muscle strengthening and functional recovery [3]. This form of therapy allows therapists to provide a personalized treatment plan with precise evaluations and guidance, while fostering motivation and psychological support through direct interaction. However, gait training in physical therapy places significant physical demands on therapists, and may lack the consistency required for repetitive and intensive training. Robot-assisted gait training (RAGT) has thus been introduced as a complementary approach to address these limitations [8,9].

The Ministry of Health and Welfare of South Korea implemented the Post-Acute Rehabilitation Medical Institution system in 2017 to address issues in the previous rehabilitation care delivery system. This system was designed to provide patients in need of rehabilitation with intensive care without concerns about discharge. As of 2023, 53 institutions have been designated as post-acute rehabilitation medical institutions. The Post-Acute Rehabilitation Medical Institution system aims to establish a structured rehabilitation care delivery framework, reduce hospital stays, and facilitate early intensive rehabilitation for patients to return home or reintegrate into their communities. It also seeks to ensure community reintegration for chronic-phase patients after discharge. Following the system’s objectives in 2017, a pilot project for intensive early rehabilitation was launched, and as of 2024, the third phase of this pilot project is in progress. Under the conventional reimbursement system for rehabilitation services, rehabilitation therapy fees are categorized under Chapter 7, Section 3 (Specialized Rehabilitation Therapy) of the Health Insurance Medical Benefits Fee Schedule. This system limits treatment frequency to twice per day for inpatients and once per day for outpatients. However, the fee structure for post-acute rehabilitation medical institutions is calculated on a time-unit basis (1 unit = 15 minutes), allowing for up to 16 units (4 hours) per day [10,11].

Many Korean convalescent rehabilitation hospitals utilize RAGT. RAGT is a promising therapy that promotes proprioception, neuroplasticity, and reconstruction of damaged neural pathways in stroke patients, thereby aiding in the effective recovery of gait and balance function. Studies have shown that RAGT significantly improves various motor functions, including lower limb strength, balance control, and gait speed [12,13]. Devices such as the Morning Walk^® S200 end-effector gait-assist robot (MW-S200; Curexo) ensure patient safety through features such as upper body safety belts, parallel bars on both sides, and emergency stop buttons, thereby enabling patients to perform repetitive and precise training [14]. Neuroplasticity, defined as the brain’s ability to reorganize its pathways and structures through experience and stimulation, is crucial for recovering damaged neural networks after a stroke, and is activated when learning new motor skills or reacquiring functional abilities, restructuring neural circuits, and facilitating the recovery of brain function. This is particularly relevant to the recovery of motor function in patients with stroke, among which RAGT plays a significant role in fostering neuroplasticity. RAGT stimulates damaged neural pathways, reinforces new neural connections, and facilitates reorganization of neural circuits. RAGT provides consistent stimulation that activates the motor integration regions of the brain, thereby promoting neural recovery [15,16]. As RAGT repeatedly guides movements for gait and balance maintenance, the brain learns these repetitive motions, reinforcing proprioceptive neural circuits. Consequently, patients gradually regain their balance and gait abilities, thereby enabling safe and independent walking through improved proprioception [17,18].

In the present study, we aimed to evaluate the effects of RAGT on gait, balance, and functional recovery in stroke patients using the 10-Meter Walk Test (10 MWT), timed up-and-go (TUG) test, Functional Ambulation Category (FAC), Fugl-Meyer Assessment–lower extremities (FMA-LE), and Berg Balance Scale (BBS). The 10 MWT measures walking speed and mobility by assessing the time required to walk a 10-meter distance, for which improvements in walking speed signify functional recovery [19]. The TUG test comprehensively evaluates mobility and balance by measuring the time required for a patient to stand up from a chair, walk 3 meters, turn around, and sit back down. The TUG test is an essential indicator of dynamic balance and gait performance [20]. The FAC assesses the degree of independence and need for assistance during walking, thereby indicating the patient’s ability to ambulate independently [21]. The FMA-LE assesses motor function, focusing on strength, movement control, and coordination, thus highlighting the recovery of lower limb function that contributes to balance and gait improvement [22]. The BBS is a detailed measure of balance ability, assessing both static and dynamic balance; higher BBS scores correlate with better balance control and gait ability [23].

The present study aimed to compare the effects of RAGT frequency on gait and balance in stroke patients with lower limb paralysis. Specifically, it compares a group receiving RAGT once per week for two units with a group receiving intensive two units per day RAGT 5 times per week, to evaluate how training frequency and intensity affect rehabilitation outcomes.

The significance of this study was threefold. First, by assessing the specific impacts of RAGT frequency and intensity on gait and balance recovery, this study provides data to optimize rehabilitation efficiency and treatment strategies. Secondly, it supports the improvement of intensive rehabilitation programs in convalescent rehabilitation hospitals, and the development of personalized rehabilitation plans for patients. Third, by demonstrating the efficacy of early intensive rehabilitation with RAGT, it contributes to enabling patients to achieve a faster and more stable return to daily and social activities.

In conclusion, this study provides an in-depth analysis of how RAGT frequency and intensity affect rehabilitation outcomes in patients with stroke, maximizing the results of early intensive rehabilitation and contributing to the development of individualized rehabilitation strategies tailored to the needs of each patient.

Outpatients were recruited from the Cheongju IM Rehabilitation Hospital, South Korea. All study procedures were conducted under the supervision of the Institutional Review Board (IRB), and in accordance with the Declaration of Helsinki. The Cheongju University Research Ethics Committee approved the experimental procedures and protocols (IRB no. 1041107-202404-HR-010-01).

The inclusion criteria were as follows: (1) stroke patients between 1 month and 3 months post-onset; (2) FAC score of 1–3, (3) hemiplegic patients without assistive devices, classified according to the early Brunnstrom stages of motor recovery [24]; (4) ability to understand and follow verbal instructions; (5) no orthopedic diseases in the lower limbs; and (6) a score of 20 or higher on the Korean version of the Mini-Mental State Examination (MMSE-K). The exclusion criteria were as follows: (1) visual, auditory, or vestibular impairment; (2) functional impairment of the lower limbs due to neurological issues unrelated to stroke; and (3) unilateral neglect.

This was a randomized clinical preliminary study using a single-blind testing method, with a sample selected through systematic sampling. Participants were divided into two groups, both of whom used the Morning Walk^®. The Morning Walk^® features an innovative seated body-weight support system, enabling training with minimal time required for patient boarding and preparation (within 3 minutes). It provides boarding/disembarking modes for severely impaired patients and supports motions for flat ground walking, stair climbing/descending, and incline navigation. The stair height can be adjusted to specific steps (7 cm, 12 cm, 17 cm), and the incline has four adjustable levels (5°, 10°, 15°, 20°). Additionally, the independent movement of the left and right footplates allows for effective and progressive therapy through individualized gait pattern settings.

The experimental group underwent RAGT 5 days per week with two units per day, whereas the control group received RAGT once per week with two units per session. Both groups performed the training with the first 15 minutes in flat ground mode, followed by 15 minutes in stair climbing mode. The remaining 14 units of comprehensive rehabilitation, including gait training, occupational therapy, and speech therapy, were administered equally to both groups.

Descriptive statistics were applied to identify participants’ general characteristics. An independent t-test was further conducted to compare the characteristics between groups. The Kolmogorov–Smirnov test was used to analyze data normality. For functional improvements between the experimental and control groups at pre-treatment, 2 weeks, and 4 weeks, repeated-measures ANOVA was used for the 10 MWT, BBS, and FMA tests, which met the normality assumption. The TUG and FAC tests that did not meet the normality criteria were analyzed using the Friedman test. Data analysis was performed using IBM SPSS Windows version 25.0 (IBM Co.), with the significance level set at α = 0.05.

Based on the abovementioned criteria, 12 patients (6 males and 6 females) were recruited. In the experimental group, the average age, height, weight, and MMSE-K score were 77.32 ± 5.42 years, 166.83 ± 8.64 cm, 63.83 ± 7.67 kg, and 23.16 ± 2.56, respectively. The mean time since stroke onset was 25.10 ± 15.52 days. In the control group, the corresponding values were 73.00 ± 8.85 years, 160.66 ± 9.13 cm, 56.83 ± 8.56 kg, and 22.83 ± 1.16, respectively. The mean time since stroke onset was 27.10 ± 13.12 days. The general characteristics of both groups showed no statistically significant differences (Table 1).

Table 1 . General characteristics (N = 12).

	Control group (n = 6)	Experimental group (n = 6)	t
Age (y)	73.00 ± 8.85	77.32 ± 5.42	1.022
Height (cm)	160.66 ± 9.13	166.83 ± 8.64	1.179
Weight (kg)	56.83 ± 8.56	63.83 ± 7.67	1.491
Sex (male/female)	3/3	4/2	–0.620
Onset (d)	27.10 ± 13.12	25.10 ± 15.52	–0.542
Type (ischemic/hemorrhage)	3/3	4/2	–0.542
Affected side (right/left)	4/2	4/2	–0.542
MMSE-K	22.83 ± 1.16	23.16 ± 2.56	0.290

Values are presented as mean ± standard deviation or number only. MMSE-K, Korean version of the Mini-Mental State Examination..

For the 10 MWT test, Mauchly’s test of sphericity revealed a significance level of 0.414, satisfying the assumption of sphericity, while Pillai’s trace test confirmed statistical significance over time (p = 0.001). However, there was a statistically significant interaction between time and group (p = 0.037). In both groups, statistically significant improvements were observed at both 2- and 4-weeks post-treatment compared to pre-treatment (p = 0.002), but no significant difference was found between the 2- and 4-week post-treatment results (p = 0.382) (Table 2).

Table 2 . Descriptive statistics and repeated measures ANOVA results.

Test	Group	T0	T1	T2	F	p-value
10 MWT	Control	30.27 ± 12.04	25.03 ± 8.21	24.74 ± 8.10	15.949** (T0 < T1**, T0 < T2**)	0.001
	Experimental	28.99 ± 12.04	24.01 ± 12.20	19.53 ± 7.52
BBS	Control	33.17 ± 12.46	35.83 ± 12.51	36.17 ± 12.44	14.179** (T0 < T1*, T0 < T2*)	0.002
	Experimental	32.17 ± 11.26	39.33 ± 8.61	41.50 ± 7.25
FMA-LE	Control	21.17 ± 5.49	22.00 ± 4.98	24.17 ± 6.52	6.588* (T0 < T2**)	0.017
	Experimental	21.17 ± 5.87	22.17 ± 4.98	24.00 ± 4.51

Values are presented as mean ± standard deviation. 10MWT, 10-Meter Walk Test; BBS, Berg Balance Scale; FMA-LE, Fugl-Meyer Assessment–lower extremities; T0, baseline; T1, post 2 weeks; T2, post 4 weeks. *p < 0.05, **p <0.01..

For the BBS, Mauchly’s test of sphericity showed a significance level of 0.000, which did not indicate sphericity. The Huynh-Feldt correction confirmed a statistically significant effect over time (p = 0.002), but there was no statistically significant interaction between time and group (p = 0.077). Both groups showed statistically significant improvements at both 2- and 4-weeks post-treatment compared to pre-treatment (p = 0.009, 0.010), but no significant difference was found between the 2- and 4-week post-treatment results (p = 0.109) (Table 2).

For the FMA-LE test, Mauchly’s test of sphericity showed a significance level of 0.131, satisfying sphericity, and Pillai’s trace test confirmed statistical significance over time (p = 0.017). There was no significant interaction between time and group (p = 0.959). Statistically significant improvements were observed between pre-treatment and four weeks post-treatment (p = 0.011) and between 2- and 4-weeks post-treatment (p = 0.045); however, no significant difference was found between pre-treatment and two weeks post-treatment (p = 0.157) (Table 2).

For the FAC and TUG, as normality was not satisfied, comparisons were made using the Friedman test. In the FAC assessment, the experimental group showed borderline significance (p = 0.050), whereas the control group showed no statistical significance (p = 0.223). In the TUG test, the control group did not show a statistically significant difference (p = 0.165); however, a statistically significant difference was observed in the experimental group (p = 0.030) (Table 3).

Table 3 . Descriptive statistics and Friedman result.

Test	Group	T0	T1	T2	χ2	p-value
FAC	Control	2.16 ± 0.75	2.33 ± 0.81	2.50 ± 0.83	3.000	0.050
	Experimental	1.67 ± 0.51	2.17 ± 0.75	2.17 ± 0.75	6.000
TUG	Control	28.87 ± 5.71	25.35 ± 6.39	24.22 ± 5.05	3.600	0.030
	Experimental	24.82 ± 7.33	23.46 ± 6.01	19.81 ± 5.41	6.333*

Values are presented as mean ± standard deviation. FAC, Functional Ambulation Category; TUG, timed up-and-go; T0, baseline; T1, post 2 weeks; T2, post 4 weeks. *p < 0.05..

This study compared the effects of weekly RAGT frequency on gait and balance in patients with lower limb paralysis after stroke. Achieving improvements in walking ability after stroke is essential for independent daily living and quality of life [23]. The recovery mechanism following brain injury involves motor learning, in which motor skill acquisition, motor adaptation, and decision-making are critical considerations [25]. These elements reflect the process by which stroke patients experience and resolve various movement patterns and errors through active engagement, and are closely associated with the plasticity of the central nervous system [26]. Neuroplasticity is a phenomenon in which brain neurons learn from external stimuli and reorganize their structure and function, occurring throughout life, and it can be observed through functional magnetic resonance imaging. During early childhood, language and motor skill learning are highly active, maximizing the activity of new neural pathways. Based on this concept of motor learning, RAGT offers clinical benefits by allowing the safe and repetitive practice of gait movements with adjustable intensity and frequency using a robot-assisted system [4,16,27,28].

In this study, the experimental group underwent RAGT training at two units per day, 5 days a week, whereas the control group received two units of traditional gait training 5 days a week. The remaining 14 units of rehabilitation training, including occupational therapy, speech therapy, and gait training, were equally administered to both groups.

The key findings of this study were that, both the experimental and control groups displayed statistically significant improvements at 2- and 4-weeks post-treatment compared to pre-treatment in the 10 MWT, although no significant difference was observed between the 2- and 4-week results. These statistically significant improvements indicate that both RAGT with morning walking and traditional rehabilitation training are effective at enhancing short-distance walking ability and speed. In addition, we found a statistically significant interaction between time and group. Successful approaches to gait recovery are based on intensive, repetitive, task-oriented practice linked to the active engagement of the participant [19,29]. This indicates that end-effector RAGT, which provides a gait-training program that allows for active participation, is potentially more effective than traditional rehabilitation by enabling intensive, repetitive, task-oriented practice. The Lokomat (Hocoma AG) provides direct physical support to the legs, actively assisting with leg movements to compensate for weaknesses and ensure proper gait patterns. In contrast, the Morning Walk^® guides the feet along a predefined path, offering a different but equally effective approach to gait rehabilitation [9].

For the BBS, both groups showed statistically significant improvements at 2- and 4-weeks post-treatment compared to pre-treatment, with no significant interaction between time and group. Previous studies have reported no statistically significant differences in balance improvement between the RAGT and traditional gait training, which our study supports [23,30]. RAGT provides visual feedback, which plays an important role in balance recovery in hemiplegic patients. Many studies have similarly examined how balance control can be improved through visual feedback [31,32].

In the FMA-LE, statistically significant improvements were observed between pre-treatment and a 4-weeks post-treatment, whereas there was no significant difference between pre-treatment and 2-weeks post-treatment. This suggests that both RAGT with morning walking and traditional rehabilitation training are effective in promoting functional recovery after stroke [33].

Regarding the FAC scores, the experimental group showed a borderline significance level, whereas the control group was not statistically significant. Moderate evidence suggests that RAGT can improve walking distance and functional ambulation after stroke [34], indicating that RAGT with morning walking may be more effective than traditional rehabilitation training for gait recovery.

Regarding TUG scores, the experimental group showed statistical significance, whereas the control group did not. RAGT is considered to be a promising therapy for improving balance and gait function in the early stages of rehabilitation after stroke, particularly by enhancing precise gait control and adaptability, thereby contributing to gait mobility. Through repetitive and precise gait training, RAGT improves various motor functions, including lower limb strength, gait speed, and balance control. These functional improvements can positively impact the TUG test result [20,35]. This is because the RAGT can directly affect a patient’s ability to maintain a stable posture while walking, as well as the initial movement of rising from a seated position and the final movement of sitting back down, all of which are essential for TUG performance.

As a pilot study, this research had limitations due to its relatively short intervention period, which was insufficient to fully assess long-term intervention effects, and a relatively small sample size, which limited statistical power. Additionally, it did not include a comparison with a group receiving only traditional rehabilitation without RAGT. Future studies should address these limitations and continue to apply long-term interventions with larger sample sizes [36].

In this study, the absolute amount of rehabilitation time significantly contributed to functional recovery, as intensive rehabilitation of 16 units per day was conducted daily in the convalescent rehabilitation hospital. This extensive rehabilitation time likely minimized the statistical differences between groups related to the frequency of RAGT sessions.

Clinically, these findings demonstrate that RAGT can be utilized as a sophisticated rehabilitation strategy tailored to the functional status of individual patients, contributing significantly to improving patient independence and quality of life. This underscores the importance of considering RAGT as a valuable therapeutic option, enabling clinicians to actively incorporate it into rehabilitation plans based on patient conditions and goals.

In conclusion, the results of this study suggest that both end-effector RAGT and traditional gait training can significantly enhance gait ability, static/dynamic balance, and overall lower limb function in patients with stroke. Notably, the improvements in the 10 MWT and FAC scores indicate that RAGT may provide particular advantages in enhancing gait ability compared with traditional gait training. These results suggest that RAGT may offer a more effective and individualized rehabilitation approach based on patient functionality.

Generative AI tools (ChatGPT) were used to assist in language editing and grammar correction, while the intellectual content and research were entirely produced by the authors.

This work was supported by the Cheongju IM Rehabilitation Hospital.

No potential conflicts of interest relevant to this article are reported.

Conceptualization: MHY, TY. Data curation and Formal analysis: ESK, JOC, JHL. Methodology: MHY, DHL, DSH. Supervision: TY. Writing - original draft: MHY, DHL. Writing - review & editing: BSW, YHP.