Stroke damages nerve cells due to ischemia and hemorrhage of cerebral blood vessels, resulting in paralysis, decreased sensation, and muscle weakness, and thus, substantially impairs daily life [1,2]. In particular, reduced trunk control due to stroke is a major factor of reduced balance ability and makes it difficult for stroke patients to walk [3]. Furthermore, this reduced control causes a weight shift toward the non-paretic side in 61 to 80% of stroke patients [4].

Standing from a sitting position consists of four steps [5] and is a major functional movement that precedes a standing posture or walking. In order to stand efficiently, body weight must be supported equally by the lower extremities [6,7]. However, patients with left or right hemiparesis due to stroke tend to shift body weight to the non-paretic side from the moment they lift their buttocks from a contact surface when standing up or move the non-paretic foot backward and shift body weight excessively to the non-paretic lower extremity [7,8].

Acran et al.[9] compared weight-bearing ratios of paretic and non-paretic sides in stroke patients by shifting weight to the paralyzed lower extremity or both lower extremities in a standing position. Lazaro et al. [10] compared weight-bearing patterns on paretic sides in stroke patients after exercises performed with trunk and limbs connected and exercises performed without trunk and limbs connected. Bartolo et al. [11] investigated the effectiveness of upper limb training using a device that can arm weight support of the paralyzed arm in stroke patients. Seo et al. [12] compared plantar pressures on paretic sides caused by table contact with the palms of paretic and non-paretic sides during standing up. Although various studies have been conducted to increase weight-bearing of the paralyzed lower extremity of stroke patients, most studies have investigated the effect of weight support on the lower extremity of the paralyzed side or the effect of reaching training through arm weight support on the paralyzed side. Especially, research on the effects of repeated standing training exercises with arm weight support is insufficient. Therefore, this study was conducted to determine the effect of standing training with support of the paretic upper limb on weight bearing of the paretic side in stroke patients. We hypothesized that stand-up training with support of the upper limb on the paretic side would be more effective than conventional stand up training in terms of improving the balance abilities and walking functions of stroke patients.

1. Participants

This study was conducted on stroke patients hospitalized at Y Rehabilitation Hospital in Daejeon City and undergoing rehabilitation treatment. The criteria used to select subjects were: hemiplegia due to stroke (as determined by CT or MRI > 3 months), no communication difficulty, and the ability to follow instructions (Korean Mini-Mental State Test score of ≥ 24)[13], the ability to stand unassisted and maintain a standing position for > one minute, and the ability to walk independently without assistance for more than a few minutes indoors or outdoors. The exclusion conditions were a history of surgery to one or both lower extremities, a muscle tone in the ankle joint of MAS (modified Ashworth’s Scale) level ≥ 2, and walking using an orthosis [13]. Of the 85 patients hospitalized at Y Rehabilitation Hospital, 20 met the study selection criteria. The study subjects were randomly assigned using sealed envelopes to an experimental group of 10 people who performed stand-up training with the proximal upper limb on the paretic side supported or a control group of 10 people who performed conventional standing training without support on the proximal upper limb on the paretic side. A physical therapist with 5 years of clinical experience and no information about the study performed the group assignments. Subjects were provided with explanations of the study purpose and procedures, and all volunteered to participate and provided written informed consent (Fig. 1). This study was conducted after receiving approval from the University of Daejeon Research Ethics Committee (1040647-202310-HR-008-03). This study was conducted in accordance with the Declaration of Helsinki of 1964.

Fig. 1. CONSORT diagram. NDT, neurodevelopmental therapy; SUT, stand-up training.

2. Intervention

All 20 subjects received neurodevelopmental treatment, consisting of range of motion exercises, rolling exercises, and stand-up and walking training. Stand-up training was also conducted after neurodevelopmental treatment by a physical therapist with more than 10 years of clinical experience. The control group received stand-up training without support of the proximal upper limb on the paretic side. Patients in the experimental group were given stand-up training with proximal upper limb support on the paretic side, which involved resting the paretic arm on a table. The table was of the height-adjustable type commonly used in neurological physical therapy rooms. A wedge was used to ensure that the arms sufficiently contacted the table, and the height of the table was set to patient shoulder joint height with the patient in a sitting position. The therapist supinated the forearm and palm of the paretic upper limb to rest comfortably on the wedge [14] and instructed the patient to raise and lower the scapula. This movement was repeated several times to reduce muscle tension around the forearm (Fig. 2). After confirming that tension in the upper extremity on the paretic side had decreased, the therapist instructed the patient to stand while supporting the shoulder joint. During the middle and end of the stand-up movement, the therapist moved the upper limb on the paretic side backward to aid its positioning next to the trunk. One set was defined as five repetitions of this movement, and the exercise program comprised three sets per day, five days per week, for four weeks. A 2-minute rest period was provided between sets to minimize muscle fatigue.

Fig. 2. Stand-up training with paretic upper limb support. A: view from the non-paralyzed side; B: view from the paralyzed side.

3. Outcome Measures

1) Plantar pressure distribution on the paretic side

The MP-2513 plantar pressure distribution measurement kit (Kytronix Inc, Seoul, Korea) was used to measure the equality of left and right weight distributions during a static standing posture. This kit includes the Snowforce3 program, which provides high-resolution plantar pressures transmitted from a smart insole consisting of 118 sensing nodes and provides visual information and plantar pressure ratios. The closer the plantar pressure ratio is to 50%, the more equal the left-to-right weight distribution. This metric has been shown to have high reliability and validity for stroke patients [15]. To measure sole pressures before and after intervention, subjects were asked to look forward, 45 degrees upward, and maintain a standing posture with arms hanging comfortably. Patients maintained a standing posture for 10 seconds without shaking before pressures were measured, and the examiner measured the pressure distribution during the 5 seconds immediately following this 10-second period. Plantar pressure was measured three times, and average values were calculated.

2) Dynamic balance ability

The timed up-and-go test (the TUG test) was used to evaluate dynamic balance before and after intervention. This test provides times taken for a subject to stand up from a sitting position on a chair with armrests, walk to a point 3 m away, and then resume the sitting position. This test is highly reliable with an inter-examiner reliability of r = .98 and intra-examiner reliability of r = .99 [16].

3) Gait ability

A Walker View 3.0 SCX device (Technobody Inc., Dalmine, Italy) was used to assess walking function before and after intervention. A treadmill is used with a pressure sensor below the treadmill belt, a 3D camera, which provides objective information on left and right symmetry ratio of nipple, hip joint, and knee joint on the frontal plane, leg joint angular velocity, and walking speed. The gait analysis results provided are highly reliable [17]. Data were collected on left-right symmetry ratios in the coronal plane and the step times of paretic and non-paretic sides while walking. Data were collected during a 2-minute walk on the treadmill, and the first and last 30-second periods were discounted. Left-right symmetry ratios during walking were expressed as scores; lower scores indicate better gait symmetry between paretic and non-paretic sides [18].

4. Data Analysis

The independent samples t-test and the chi-square test were used to compare general subject characteristics in the two groups. Data were subjected to normality testing. The paired samples t-test was used to determine the significances of post-intervention changes within groups, and the independent samples t-test was used to determine the significances of differences between intervention-associated changes in the two groups. Statistical significance was accepted for P values < .05, and the analysis was conducted using SPSS version 27.0 for Windows.

1. Characteristics of the Subjects

Subject characteristics are provided in Table 1. Sex, age, height, weight, paretic side, onset period, and stroke type were not significantly different in the two groups (p > .05).

Table 1

General characteristics of the study subjects

	Experimental Group (N = 10)	Control Group (N = 10)	P-Value
Sex (male/female)	8/2	8/2	1.000
Stroke type (infarction/hemorrhage)	5/5	5/5	1.000
Paretic side (left/right)	5/5	5/5	1.000
Age (years)	59.90 ± 6.02	60.90 ± 4.51	.679
Height (cm)	165.20 ± 4.59	164.20 ± 3.19	.579
Weight (kg)	63.80 ± 10.88	61.70 ± 4.57	.581
Onset period (month)	4.80 ± 1.14	5.20 ± 1.40	.492
Mmse (score)	28.0 ± 1.49	27.60 ± .96	.486

Values are expressed as means±standard deviations or numbers.

Mmse, mini-mental state examination

2. Plantar Pressure Ratios of Paretic Sides Before and After Intervention in the Experimental and Control Groups

Pressure distribution ratios of paretic sides in the two groups are shown in Table 2. In both groups, the pressure distribution ratio of paretic sides increased significantly after intervention (p < .05), but this increase was .9% greater in the experimental group (p < .05).

Table 2

Balance and gait abilities before and after training within each group and between the two groups

		Experimental group (n = 10)	Control group (n = 10)	t
Plantar pressure (%)	Pre	48.95 ± 8.97	46.36 ± 5.06	.796
	Post	50.38 ± 8.86	46.89 ± 5.12	1.080
	T	-12.503*	-8.338*
	Change	1.42 ± .36	.52 ± .20	6.914*
TUG (sec)	Pre	17.47 ± 1.20	17.37 ± 1.47	.164
	Post	16.31 ± .80	16.82 ± 1.40	-1.055
	T	7.918*	12.201*
	Change	-1.16 ± .46	-.55 ± .14	-4.001*
Gait symmetry (score)	Pre	3.78 ± .89	3.35 ± .63	1.248
	Post	3.22 ± .94	3.11 ± .62	.309
	T	10.755*	7.060*
	Change	-.56 ± .16	-.24 ± .11	-5.146*
STP (sec)	Pre	2.02 ± .33	2.01 ± .31	.055
	Post	2.37 ± .29	2.17 ± .32	1.432
	T	-6.360*	-8.500*
	Change	.35 ± .17	.16 ± .06	3.256*
STNP (sec)	Pre	2.47 ± .24	2.50 ± .18	-.299
	Post	2.50 ± .22	2.52 ± .16	-.218
	T	-2.318*	-3.404*
	Change	.03 ± .04	.02 ± .02	.606

Values are means ± standard deviations.

TUG, timed up-and-go test; STP, mean stance time on paretic sides; STNP, mean stance time on non-paretic sides.

*p < .05.

3. Dynamic Balance Abilities Before and After Intervention in the Two Groups

TUG test results before and after intervention in the experimental and control groups are provided in Table 2. In both groups, mean TUG test time was significantly reduced by intervention (p < .05), and TUG time was reduced by .61 seconds more in the experimental group (p < .05).

4. Comparison of Group Gait Functions and Stance Times Before and After Intervention

Group gait symmetry values and stance times before and after intervention are shown in Table 2. Gait symmetry scores decreased significantly after intervention in both groups (p < .05), though this decrease was .32 points greater in the experimental group (p < .05). Stance times increased significantly for paretic and non-paretic sides after intervention. Mean stance time for paretic sides increased .19 seconds more after intervention in the experimental group (p < .05). Intervention-associated increases in the stance times of non-paretic sides were similar in the two groups.

This study was conducted to investigate the effect of stand-up training with proximal upper limb support of the paretic side on plantar pressure distribution, dynamic balance ability, and walking function in stroke patients. A significant improvement in all variables was observed after interventions, and the experimental group showed more significant improvements than the control group. These results are consistent with the hypothesis of this study, and this is believed to be because the weight support provided by the upper extremity on the paralyzed side during standing training provided effective weight support for the lower extremity on the paretic side [12]. Camargos et al. argued that standing up from a sitting position is the most basic movement in daily life [19]. Briere et al. argued that when stroke patients stand up, plantar pressure on the paretic side decreases due to asymmetric weight support by the lower extremities, resulting in asymmetric movements [20]. Therefore, the results of this study indicate that stand-up training with proximal upper limb support of the paretic side can positively impact the daily movements of hemiplegic stroke patients and effectively reduce abnormal weight-bearing.

In this study, stand-up training performed while supporting the proximal upper limb on the paretic side significantly increased plantar pressure distribution in the lower extremity on the paretic side more than general stand-up training. Yoon et al. measured and compared the distribution of plantar pressure on feet while walking in patients with hemiplegic stroke and found pressure on the outer part of the metatarsal on the paretic side patient was 7.64 N/cm² and pressure on the non-paretic side was 12.64 N/cm² [21]. In this study, stand-up training increased plantar pressures on paretic sides from 48% to 50% during the standard standing posture, which suggests that if this training is performed continuously, it could positively improve plantar pressure distribution on the paretic side when walking. Cheng et al. conducted stand-up training for stroke patients using a biofeedback device and observed that compared to before training, the difference in weight distribution between the paretic side and the non-paretic side when standing up decreased from 49.5% to 38.6% [22], indicating weight distribution increased toward the paretic side, which concurs with our observations.

In this study, the stand-up training performed while supporting the proximal upper limb on the paretic side was more effective than conventional stand-up training as determined using TUG test times, which reflect dynamic balance ability, gait symmetry rate, and stance time on the paretic side. Jung et al. conducted weight transfer training from a sitting position to the paralyzed side in stroke patients and reported that time to get up and walk was reduced by 2.5 seconds compared to a group that did not do weight transfer training [23], which is partially consistent with our results. Additionally, Kim et al. conducted standing training for stroke patients for 4 weeks and found mean TUG time decreased by ~50% [24].

Yang et al. reported that trunk rotation when walking is reduced in stroke patients and that this decrease causes a bias toward the non-paretic side of the trunk, which ultimately is the main cause of a decrease in left-right trunk symmetry [25]. Therefore, our result that stand-up training with proximal upper limb support on the affected side improved gait symmetry more than conventional stand-up training could positively increase trunk rotation in stroke patients. The finding of a previous study that functional weight transfer training conducted in sitting and standing positions in hemiplegic stroke patients increased paretic side stance time by 1 second is partially consistent with our results [26]. In addition, the clinical use of stepping time of the paralyzed lower extremity as an objective indicator for determining whether or not an improvement in walking function has occurred in stroke patients [27] indicates that the results of the present study may be clinically meaningful.

Because this study is a pilot study and the number of subjects recruited was small, generalizations based on our results are limited. In addition, because the main intervention implemented in this study was based on weight support, we were unable to measure improvements in the swing phase on paretic sides. In addition, the study falls short of proving the effectiveness of training through upper limb support on the paralyzed side at all stages of standing. Since this study investigated the effects of standing training with arm support on the paralyzed side, it is not possible to present the effects of arm support on the non-paralyzed side. However, the study appears to be meaningful because it shows that stand-up training with support of the paralyzed upper extremity in hemiplegic stroke patients can provide weight support for the paralyzed lower extremity. We suggest that research studies be initiated to address the limitations of the present study and determine whether more effective stand-up training protocols can improve functional recovery in stroke patients.

This study was conducted on hemiplegic stroke patients to determine the effect of stand-up training with proximal upper limb support on the paretic side on plantar pressure distribution on the paretic side, dynamic balance ability, gait symmetry, and stance time during walking. Stand-up training with proximal upper limb support was found to be more effective than general stand-up training at improving plantar pressure on the paretic side, dynamic balance ability, left-right gait symmetry, and stance time on the paretic side. Therefore, the study suggests that standing training with support of the proximal upper limb on the paretic side might effectively improve weight-bearing ability, dynamic balance ability, and walking function in this patient population.