Effects of body weight-supported overground training on gait recovery in severe stroke-induced hemiplegia: a single-case ABAB design

Abstract

[Purpose] This study aimed to examine the effect of body weight-supported overground training on gait recovery in a patient with severe stroke-induced hemiplegia. [Participants and Methods] The participant was a woman in her 40s with severe hemiplegia following a stroke. A single-case ABAB design was employed. Standard physiotherapy was provided in the first and third phases, while physiotherapy combined with body weight-supported overground training was administered in the second and fourth phases. Walking speed, Timed Up and Go test, Brunnstrom recovery stage (BRS) and Fugl–Meyer Assessment (FMA) for motor function, and motor-functional independence measure (m-FIM) for activities of daily living (ADL) were performed to assess efficacy of body-weight-supported overground training. [Results] Walking speed improved from 0.26 m/s at admission to 0.37 m/s in the first phase, 0.58 m/s in the second phase, 0.45 m/s in the third phase, and 0.50 m/s in the fourth phase, reaching 0.40 m/s with a T-cane at discharge. m-FIM scores increased steadily from 20 points at admission to 74 points at discharge, while BRS and FMA showed minimal improvement. [Conclusion] Body weight-supported overground training may enhance walking ability, functional performance, and ADL independence in patients with severe stroke-induced hemiplegia.

INTRODUCTION

Stroke is the second leading cause of death worldwide, following ischemic heart disease¹⁾. In Japan, the incidence of stroke is 1.5 to 2 times higher than in Western industrialized countries²⁾, although the mortality rate has been steadily declining³⁾. Despite this positive trend, stroke sequelae remain a significant concern, as they are a major cause of long-term nursing care or bedridden conditions in patients.

Rehabilitation of walking ability is critical for stroke patients with hemiplegia, as it significantly impacts activities of daily living (ADLs)⁴⁾, quality of life (QOL)⁵⁾, return-to-home rates⁶⁾, and overall life prognosis⁷⁾. Among gait parameters, walking speed is a key indicator of recovery, reflecting spatiotemporal characteristics, such as stride length, step overlap, cadence, and the propulsive force generated by the paretic limb. It is an essential determinant of gait independence and activity levels^{8, 9)}. Despite improved mortality rates, rehabilitation remains a central focus in addressing post-stroke sequelae, particularly in achieving gait recovery for patients with severe hemiplegia.

Rehabilitation strategies, such as Body Weight Supported Treadmill Training (BWSTT) and Body Weight Supported Overground Training (BWSOT), are commonly employed to improve gait. Both approaches use suspension systems, such as belts or harnesses attached to the trunk and lower limbs, to reduce the patient’s weight. This ensures safety during gait training while enabling the adjustment of exercise intensity and walking speed based on the patient’s condition¹⁰⁾. BWSTT facilitates lower limb alternating movements on a treadmill by applying upward traction through a harness, thereby reducing antigravity muscle activity and activating the Central Pattern Generator (CPG) to enable repetitive gait training^{11, 12)}. However, a Cochrane review reported that BWSTT does not significantly improve gait independence compared to conventional treadmill training¹³⁾.

BWSTT involves practicing walking on a treadmill, necessitating subsequent overground walking training. After acquiring a gait pattern on the treadmill, it is essential to relearn and adapt the gait to overground walking. In contrast, the goal of gait rehabilitation is to improve overground walking ability rather than treadmill-specific performance, making BWSOT a more functional and potentially superior approach for enhancing gait independence. BWSOT also allows for the adjustment of task difficulty and intensity tailored to the support requirements and motor endurance of the paretic lower limb^{14, 15)}.

Recent studies have demonstrated the benefits of BWSOT over BWSTT. For example, Gama et al.¹⁶⁾ compared the effects of BWSTT and BWSOT in patients with chronic stroke and found that while both interventions improved walking ability and motor function, stride asymmetry was significantly reduced only in the BWSOT group. Similarly, Brunelli et al.¹⁷⁾ showed that BWSOT was more effective than standard rehabilitation in improving walking independence assessed by the Functional Ambulation Categories in patients with subacute stroke. While these findings show the efficacy of BWSOT in gait recovery, most studies focus on patients with mild to moderate stroke, leaving a gap in understanding its effects on severe stroke-induced hemiplegia^16,17,18,19).

Therefore, this study was performed with two objectives. The primary objective was to elucidate the effects of BWSOT on gait recovery in a patient with severe stroke-induced hemiplegia. The secondary objective was to examine the impact of BWSOT on motor performance, ADL, and physical function, as well as to evaluate its ripple effect and carryover effects. This study employed a single-case ABAB design to address current gaps in the literature.

PARTICIPANTS AND METHODS

The study was approved by the Rakusai Shimizu Hospital Ethical Review Committee and the Josai International University Research Ethics Committee (Approval Nos. 24-14 and 2024-007). Written informed consent for participation was obtained from the patient and her family after providing both oral and written explanations of the study details. This study was also registered with the UMIN Clinical Trial Registry (UMIN-CTR; Approval No.: UMIN000054825).

The participant was a woman in her 40s (height: 158 cm, weight: 49 kg, BMI: 19.6 kg/m²) who was admitted to the Comprehensive Rehabilitation Unit following a diagnosis of an acute intracerebral hemorrhage in the left internal capsule (Fig. 1). Her history of present illness revealed sudden-onset right hemiparesis, aphasia, and loss of consciousness. A craniotomy was performed to evacuate the hematoma. No complications or comorbidities related to the disease were observed.

Fig. 1.

Axial computed tomography (CT) scan of acute intracerebral hemorrhage. Axial CT scan taken on the day of ictus, demonstrating an acute intracerebral hemorrhage located in the left internal capsule region.

The patient was transferred to our hospital 22 days post-ictus and began physical therapy upon admission. At the initial assessment, her Fugl–Meyer Assessment (FMA) scores were 4 for the right upper and lower extremities. Sensory deficits were significant, with markedly diminished superficial and deep sensory perception. The patient also presented with aphasia and notable auditory comprehension difficulty. During the acute phase of rehabilitation, she underwent gait training using a knee-ankle-foot orthosis (KAFO) with a metal post for stability.

The study employed a single-case ABAB design to investigate the effects of BWSOT on gait recovery^{20, 21)}. During Period A (A1 and A2), standard physiotherapy was combined with conventional gait training, whereas during Period B (B1 and B2), standard physiotherapy was combined with BWSOT. For severely affected patients, conventional gait training involves manual assistance from behind while the patient practices walking. During this process, a KAFO was used, and an appropriate walking aid was selected based on the patient’s condition. Each treatment period lasted two weeks, resulting in a total intervention duration of eight weeks. Evaluations were conducted at admission, at the end of each period (A and B), and at discharge (Fig. 2). The timeline in relation to the onset of symptoms (denoted as Z) was as follows: admission (Z + 23), A1 (Z + 37), B1 (Z + 51), A2 (Z + 65), B2 (Z + 79), and discharge (Z + 136).

Fig. 2.

Single-case ABAB study design depicting intervention phases. Standard physiotherapy combined with conventional gait training (CGT) was implemented in Periods A1 and A2, while standard physiotherapy combined with body weight-supported overground training (BWSOT) was applied in Periods B1 and B2. Evaluations were conducted at the end of each phase.

Standard physiotherapy was delivered in 60-minute daily sessions, comprising 40 minutes of functional restoration approaches, including balance training, transfer practice, sit-to-stand maneuvers, and ADL training. These sessions were followed by an additional 20 minutes of conventional gait training or BWSOT (Fig. 2). Walking exercises were customized to the patient’s symptoms, fatigue levels, and vital signs, with an exercise intensity maintained between 11 to 13 on the Borg Scale to ensure appropriate exertion.

The intervention spanned eight weeks, with daily 60-minute sessions administered seven days per week. The BWSOT protocol was implemented specifically during the B1 and B2 phases, cumulatively accounting for four weeks. Following the B2 phase, standard physiotherapy resumed as the sole treatment until discharge. The patient’s hospitalization totaled 136 days, supplemented by 22 days at the referring hospital, resulting in a cumulative inpatient care period of 158 days. To minimize therapist-related bias, all physiotherapy sessions were administered by physiotherapists possessing less than one year of clinical experience. Additionally, the physiotherapist responsible for administering the interventions remained blinded to the study’s objectives and was only informed of the treatment schedule for each phase (A and B).

The Universal Core Frame/New Assist IP-P2000 (InterRehab Co., Ltd., Tokyo, Japan) (Fig. 3) was employed as the suspension weight-unloading device for BWSOT. This apparatus utilizes a pneumatic unloading mechanism powered by an air compressor, with unloading levels manually adjustable via a precision dial. The pneumatic system enables gait training while preserving the natural vertical displacement of the center of gravity, thus replicating physiological walking mechanics. During phase B, the unloading level was calibrated to remain within 30% of the patient’s body weight. Previous investigations have demonstrated that unloading levels of 30% or less are optimal for maintaining muscle activation comparable to unassisted walking and promoting efficient gait performance^{22, 23)}.

Fig. 3.

Body weight supported overground training (BWSOT). BWSOT using the Universal Core Frame/New Assist IP-P2000, InterRehab Co., Ltd., Tokyo, Japan), illustrating key gait phases: initial contact (IC), loading response (LR), mid-stance (MSt), terminal stance (TSt), and swing phase.

The primary outcome measures were the 10-meter walking test, which assessed comfortable walking speed, walking time, step count, and step length, and the Timed Up and Go (TUG) test. Gait evaluations were performed with rear assistance using parallel bars during admission, up to the B2 phase, and independently with a T-cane under supervision at discharge. For orthotic support, the patient used a KAFO from admission through Period A2 and transitioned to an Ankle-Foot Orthosis (AFO) during Period B2, which was maintained until discharge.

Secondary outcome measures included gait volume, defined based on the total distance walked that was recorded for each phase and the daily average distance walked; the Brunnstrom recovery stage (BRS) and the FMA score for motor function; the motor-Functional Independence Measure (m-FIM) for ADL; and introspection evaluated using a Visual Analog Scale (VAS). The VAS ranged from 0 to 10, with higher scores indicating greater perceived comfort and ease during walking and wearing orthotic devices.

At admission and throughout Period A1, the patient could not bear weight on the paretic lower limb and could only walk with the aid of parallel bars and a locked KAFO, combined with backward assistance. During Period B1, gait training progressed to include a T-cane, with the KAFO’s knee-lock mechanism released. Rear or lateral assistance was provided for safety, and gait training was facilitated through BWSOT. In Period A2, the patient resumed gait practice without BWSOT, using the same methodology as in Period B1. By Period B2, the patient demonstrated enhanced weight-bearing capacity on the paretic limb, enabling a transition from the KAFO to an AFO. During this phase, the patient advanced from walking within parallel bars to ambulating with a four-point cane under lateral supervision.

The patient achieved independent ambulation under supervision at discharge, utilizing an AFO and a T-cane. The intervention progressively enhanced ambulation levels by systematically adjusting orthotic devices, walking aids, and physical assistance tailored to the patient’s functional improvements and physical condition.

RESULTS

Walking speed (m/s) was calculated as the walking distance (m) divided by walking time (s). At admission, the patient’s walking speed was 0.26 m/s. Walking speed increased to 0.37 m/s in Period A1, peaked at 0.58 m/s in Period B1, declined to 0.45 m/s in Period A2, and improved to 0.50 m/s in Period B2. Upon discharge, the walking speed stabilized at 0.41 m/s with the patient using a T-cane under supervision (Table 1).

Table 1. 10-meter walking test (10MWT) and timed up and go (TUG) results.

	Parameter	Admission	A1	B1	A2	B2	Discharge
		(Z + 23)	(Z + 37)	(Z + 51)	(Z + 65)	(Z + 79)	(Z + 136)
10MWT	Walking speed (m/s)	0.26	0.37	0.58	0.45	0.50	0.41
	Walking time (s)	38.4	27.3	17.3	22.2	19.8	24.7
	Step count (steps)	44	32	15	30	26	30
	Step length (cm)	22.7	31.2	66.6	33.3	38.4	33.3
TUG(s)		−	−	49.4	42.0	39.5	40.8

Gait assessment was conducted from admission to period B2 with backward assistance using parallel bars and a supervised T-cane gait at discharge. The patient was fitted with a knee-ankle-foot orthosis (KAFO) from admission to period A2, which was replaced by an ankle-foot orthosis (AFO) from period B2 until discharge. 10MWT: 10-meter walking test; TUG: timed up and go test; Z: date of onset; —: unable to measure; KAFO: knee-ankle-foot orthosis; AFO: ankle-foot orthosis.

Walking time (s) improved progressively over the course of the intervention. At admission, walking time was 38.4 s. It decreased to 27.3 s in Period A1, reached 17.3 s in Period B1, increased slightly to 22.2 s in Period A2, and improved to 19.8 s in Period B2. At discharge, walking time was 24.6 s with T-cane assistance (Table 1).

The step count required to complete the 10-m walking test was reduced substantially during the intervention. At admission, the patient took 44 steps. The step count decreased to 32 in Period A1 and improved significantly to 15 in Period B1. Steps increased to 30 in Period A2 but improved again to 26 in Period B2. At discharge, the patient completed the test in 30 steps with T-cane supervision (Table 1).

Step length (cm) was determined by dividing walking distance (m) by the step count and multiplying the result by 100. At admission, the step length was 22.7 cm. Step length improved to 31.2 cm in Period A1, peaked at 66.6 cm in Period B1, declined to 33.3 cm in Period A2, and increased slightly to 38.4 cm in Period B2. At discharge, step length was 33.3 cm with T-cane assistance (Table 1).

The TUG test could not be assessed at admission or during Period A1. During Period B1, the TUG time was 49.4 s, which improved to 42.0 s in Period A2 and 39.5 s in Period B2. The TUG time was 40.8 s at discharge with T-cane assistance (Table 1).

For the secondary outcomes, the total walking distance during practice sessions progressively increased across phases. In Period A1, the patient walked 1,770 m, with a mean of 136.2 ± 44.5 m/day. Walking distance increased to 1,955 m (mean: 150.4 ± 35.4 m/day) in Period B1, peaked at 2,515 m (mean: 193.5 ± 75.1 m/day) in Period A2, and declined slightly to 2,160 m (mean: 166.2 ± 46.5 m/day) in Period B2 (Table 2).

Table 2. Gait volume.

Parameter	A1	B1	A2	B2
	(Z + 37)	(Z + 51)	(Z + 65)	(Z + 79)
Total walking distance (m)	1,770	1,955	2,515	2,160
Average daily walking distance (m)	136.2 ± 44.5	150.4 ± 35.4	193.5 ± 75.1	166.2 ± 46.5

Gait volume, including total and average daily walking distance, was measured across intervention periods. Total walking distance represents the cumulative walking distance during each period. Average daily walking distance is expressed as mean ± standard deviation (SD). Z: date of onset.

The BRS score showed a slight improvement in the upper limbs, fingers, and lower limbs. The BRS score in the upper limbs remained unchanged at 1 point from admission through Period B1. However, it changed to 2 points during Period A2 and further improved to 3 points at discharge. The BRS score of the fingers remained unchanged at 1 point from admission through Period A2. However, it improved to 2 points at discharge. The BRS score of the lower limbs remained unchanged at 1 point from admission through Period A1. However, it changed to 2 points during Period B1 and improved to 3 points during Period B2 (Table 3).

Table 3. Brunnstrom stage results.

Parameter	Admission	A1	B1	A2	B2	Discharge
	(Z + 23)	(Z + 37)	(Z + 51)	(Z + 65)	(Z + 79)	(Z + 136)
Upper limbs	Ⅰ	Ⅰ	Ⅰ	Ⅱ	Ⅱ	Ⅲ
Fingers	Ⅰ	Ⅰ	Ⅰ	Ⅰ	Ⅱ	Ⅱ
Lower limbs	Ⅰ	Ⅰ	Ⅱ	Ⅱ	Ⅲ	Ⅲ

BRS scores for the upper limbs, fingers, and lower limbs were measured across intervention periods.

BRS: Brunnstrom Stage; Z: date of onset.

The FMA scores for the upper and lower extremities remained unchanged at 4 points from admission through Period B2. However, at discharge, the lower extremity score improved to 8 points, indicating a recovery in reflex activity (Table 4).

Table 4. Fugl–Meyer assessment (FMA) results.

Parameter	Admission	A1	B1	A2	B2	Discharge
	(Z + 23)	(Z + 37)	(Z + 51)	(Z + 65)	(Z + 79)	(Z + 136)
Lower limbs (points)	4	4	4	4	4	8
Upper limbs (points)	4	4	4	4	4	4

FMA scores for the lower and upper limbs were measured across intervention periods. Fugle–Meyer Assessment; Z: date of onset.

The m-FIM score improved steadily throughout the intervention progression. The score increased from 20 points at admission to 26 points in Period A1, 28 points in Period B1, 34 points in Period A2, and 37 points in Period B2. By discharge, the m-FIM score reached 74 points (Table 5).

Table 5. Motor-functional independence measure (m-FIM) scores.

Parameter	Admission	A1	B1	A2	B2	Discharge
	(Z + 23)	(Z + 37)	(Z + 51)	(Z + 65)	(Z + 79)	(Z + 136)
Eating (points)	4	5	5	5	5	7
Grooming (points)	3	4	4	4	4	6
Bathing (points)	1	1	1	1	3	6
Dressing, upper body (points)	2	2	2	3	3	6
Dressing, lower body (points)	1	1	1	2	2	6
Toileting (points)	1	1	2	4	4	5
Bladder control (points)	1	1	1	2	2	7
Bowel control (points)	1	1	1	1	1	7
Bed transfer (points)	2	3	3	3	4	6
Toilet transfer (points)	1	2	3	4	4	6
Tub, shower transfer (points)	1	1	1	1	1	4
Walking (points)	1	3	3	3	3	4
Stairs (points)	1	1	1	1	1	4
Total	20	26	28	34	37	74

m-FIM scores for various tasks were assessed from admission to discharge. Z: date of onset.

Daily introspection evaluations conducted during Periods B1 and B2 revealed stable results. The mean “fit” rating was 4.7 ± 0.4 cm in Period B1 and 4.8 ± 0.3 cm in Period B2. Similarly, the mean “walking comfort” rating remained consistent at 5.5 ± 0.3 cm in both periods (Table 6).

Table 6. Introspection scores: fit and walking comfort.

Parameter	B1	B2
	(Z + 51)	(Z + 79)
Fit (cm)	4.7 ± 0.4	4.8 ± 0.3
Walking comfort (cm)	5.5 ± 0.3	5.5 ± 0.3

Introspection scores for harness fit and walking comfort during intervention periods B1 and B2. Z: date of onset.

DISCUSSION

This study investigated the effects of BWSOT on gait recovery, motor performance, ADL, and physical function in a patient with severe stroke-induced hemiplegia. The findings demonstrated clinically meaningful improvements in walking ability, motor performance (i.e., TUG), and ADL from admission to discharge. The most substantial gains occurred during Period B, when BWSOT was implemented, highlighting its pivotal role in enhancing gait recovery. Notably, sustained motor learning effects were evident during Period A2, following the initial improvements observed in Period B1. This was demonstrated by an improvement in walking ability in Phase A2 compared to in Phase A1, which was considered to be due to the residual effects of BWSOT implemented in Phase B1. The association between walking ability and ADL improvement aligns with previous studies by Nindorera et al.²⁴⁾, who reported a strong correlation between walking speed and ADL performance in patients with chronic stroke. It is possible that the improvement in walking ability during BWSOT was generalized to ADL even in patients transitioning from the acute to the recovery phase in this study. However, the causal relationship between the improvement in walking ability due to BWSOT and enhancement of ADL has not been verified, necessitating further investigation.

The progression of walking speed recovery followed a clear trend: Period A1, 0.11 m/s; Period B1, 0.32 m/s; Period A2, 0.19 m/s; and Period B2, 0.24 m/s. The most significant improvement occurred during Period B1, coinciding with the introduction of BWSOT. In Period A1, gait training with a locked KAFO produced limited improvements due to gait instability and exaggerated knee joint movements. Conversely, BWSOT in Period B1 enabled safe weight loading on the paretic leg, facilitating motor learning related to load response and promoting coordinated knee joint movements by allowing the KAFO’s ring lock to remain disengaged. In the process of motor learning, it is essential to adjust the difficulty level appropriately²⁴⁾. In this case, which involved a risk of knee buckling, weight-bearing and gait training during Phase B1, utilizing BWSOT with the KAFO ring lock disengaged, was considered an effective approach to enhance motor learning. Furthermore, the experience gained during this phase influenced the training in Phase A2, facilitating the preparation for transitioning to an AFO. These changes resulted in a longer stride length (66.6 cm), fewer steps (15 steps), a more efficient gait pattern, and a reduced walking time (17.3 s), values approaching the optimal stride length of 71.1 cm based on the patient’s height.

Previous studies have shown that KAFO use improves standing balance²⁵⁾ and promotes flexion-extension of paretic leg joints²⁶⁾, which likely contributed to the observed improvements in walking speed during Period B1. However, a slight decline in walking speed was observed in Period A2, possibly due to the absence of BWSOT and the removal of KAFO suspension support. This change may have led the patient to adopt a cautious gait strategy to mitigate fall risk. In Period B2, walking speed improved again despite transitioning from the KAFO to an AFO, indicating that the cumulative benefits of BWSOT were sustained. After the intervention period (Z + 79), the patient’s walking ability improved sufficiently to enable supervised walking with a T-cane and an AFO at discharge (Z + 136). These findings emphasize that task-oriented BWSOT, when performed early and at an optimal level of difficulty, is highly effective for restoring walking ability, even in patients with severe hemiplegia due to stroke.

The minimal clinically important difference (MCID) for gait velocity in patients with subacute stroke ranges from 0.16 m/s²⁷⁾ to 0.175 m/s²⁸⁾. In this study, the increase in walking speed during Period B1 exceeded the MCID, indicating that the improvement was clinically meaningful. Similarly, the TUG performance improved by 7.4 s in Period A2, 9.9 s in Period B2, and 8.9 s at discharge, all surpassing the minimal detectable change (MDC) of 2.9 s²⁹⁾. These results further support the effectiveness of BWSOT while accounting for potential measurement variability. However, it is important to note that the MDC values were derived from studies involving patients with chronic stroke, and comparable MDC or MCID data for patients with subacute stroke remain unavailable. Therefore, while the observed improvements are encouraging, caution is warranted when interpreting these results, and further research is required to establish definitive thresholds for the subacute phase.

For gait retraining in patients with hemiplegia secondary to stroke, repetitive task-oriented training at an appropriate level of difficulty, combined with high intensity, long duration, and frequent sessions, is essential^30,31,32). BWSOT offers multiple advantages in this context, including its task-specific nature, the ability to adjust weight-bearing loads based on the function of the paretic lower limb, and its compatibility with various orthoses and walking aids. The difficulty level can be customized by selecting and configuring orthoses, the type of walking aids used, and the level of assistance provided. Furthermore, BWSOT enables safe and precise adjustments to the intensity, duration, and frequency of training while ensuring patient stability during suspension. In this study, total walking distance and average daily walking distance during practice sessions were notably greater in Period B1 than in Period A1. This suggests that the increased training volume facilitated by BWSOT contributed substantially to the observed improvements in walking performance. Additionally, the increase in walking distance during Period A2 compared to during Period B1 may have involved endurance gains achieved during Period B1, which subsequently enhanced the efficacy of standard physiotherapy in Period A2. Conversely, the slight decrease in walking distance observed in Period B2 compared to in Period A2 may be attributable to the transition from a KAFO to an AFO. A noteworthy finding of this study is that while there was only a slight improvement in lower extremity voluntary movement based on the BRS and no direct enhancement of motor function as measured by the FMA, significant improvements in walking ability and motor performance were observed. This effect may be attributed to the activation of central pattern generators (CPGs), a mechanism analogous to that observed with BWSTT³³⁾. Although treadmill use was not a component of this study, it is plausible that repetitive walking exercises, combined with manual facilitation of alternating lower limb movements by physical therapists, stimulated the CPGs. Additionally, BWSOT may have led to increased activity in the primary motor cortex of the injured hemisphere³⁴⁾ and reduced interhemispheric inhibition, mechanisms that enhance the efficiency of walking practice and facilitate gait recovery.

This study has several limitations. As a single-case study, the findings cannot be generalized to broader populations with stroke. The lack of a comparative control group limits the ability to draw definitive conclusions regarding the causality of BWSOT effects. While BWSOT significantly improved walking ability and performance, its impact on physical function remained limited, and the specific mechanisms underlying gait recovery are not yet fully understood. Moreover, since the outcomes of this study were limited to results of clinical assessments, such as the 10-meter walking and TUG tests, the mechanisms underlying the observed improvements remain unclear. Therefore, future studies should incorporate quantitative evaluations using surface electromyography and three-dimensional motion analysis systems to provide further empirical validation. Additionally, future research should prioritize increasing external validity by expanding sample sizes and incorporating more diverse patient populations. Randomized controlled trials are essential to establish the causal relationship between BWSOT and observed gait improvements. Furthermore, this study did not examine the potential differences in the effects of BWSOT based on variations in physiotherapists’ experience. This limitation highlights the need for future research to evaluate the impact of physiotherapist experience on intervention outcomes.

This study demonstrated the effects of a BWSOT intervention on gait improvement in a patient with post-stroke hemiplegia. BWSOT allows for gait training with minimal assistance while ensuring safety, thereby enabling patients to gain more walking experience. As a future direction for research on this topic, a standardized BWSOT treatment protocol should be developed, which may lead to greater therapeutic efficacy than that of conventional physiotherapy. Additionally, the reduction in assistance required during gait training with BWSOT indicates the potential for integrating supplementary therapeutic approaches, such as the application of appropriate handling techniques by physiotherapists and incorporation of Functional Electrical Stimulation, thereby expanding the scope of rehabilitation interventions.

In conclusion, this study suggests that BWSOT is a highly effective gait training intervention for patients with hemiplegic stroke. Its flexible, task-oriented design, adjustable difficulty levels, and safe training environment make it particularly suitable for patients with severe motor impairments requiring KAFO. This study demonstrated that even in patients with severe paralysis and higher brain dysfunction due to stroke, for whom walking is difficult, early intervention with BWSOT can facilitate walking recovery. These findings suggest the potential applicability and therapeutic indications of BWSOT. Patient feedback indicated good levels of comfort with the harness and ease of walking, and no adverse events were reported, confirming the safety and feasibility of this approach. These findings underscore the potential of BWSOT as a valuable rehabilitation strategy for restoring gait function in patients with severe hemiplegia.

Conference presentation

Part of this research was presented at the 22nd Conference of the Japanese Society of Neurological Physical Therapy. Abstracts of the Annual Conference of the Japanese Society of Neurophysiotherapy, P329. https://www.gakkai.co.jp/jsnpt22/program.pdf.

Funding and Conflict of interest

This study received no funding, and the authors declare no conflicts of interest.