Effects of robotic-assisted gait training on motor function and walking ability in children with thoracolumbar incomplete spinal cord injury

Abstract

BACKGROUND:

Spinal cord injury (SCI) results in neurological dysfunction of the spinal cord below the injury.

OBJECTIVE:

To explore the immediate and long-term effects of robotic-assisted gait training (RAGT) on the recovery of motor function and walking ability in children with thoracolumbar incomplete SCI.

METHODS:

Twenty-one children with thoracolumbar incomplete SCI were randomly divided into the experimental (n = 11) and control groups (n = 10). The control group received 60 min of conventional physical therapy, and the experimental group received 30 min of RAGT based on 30 minutes of conventional physical therapy. Changes in walking speed and distance, physiological cost index (PCI), lower extremity motor score (LEMS), SCI walking index and centre-of-pressure (COP) envelope area score were observed in both groups of children before and after eight weeks of training. The primary outcome measures were the 10-metre walk test (10MWT) and six-minute walk distance (6MWD) at preferred and maximal speeds. In addition, several other measures were assessed, such as postural control and balance, lower limb strength and energy expenditure.

RESULTS:

Compared with control group, the self-selected walk speed (SWS), maximum walking speed (MWS), 6MWD, PCI, LEMS, COP, and Walking Index for Spinal Cord injury II (WISCI II) of experimental group were improved after treatment. The 6MWD, PCI, COP, and WISCI II after eight weeks of treatment were improved in experimental group. All indicators were not identical at three different time points when compared between two groups. Pairwise comparisons in experimental group suggested that the SWS, MWS, 6MWD, PCI, LEMS, COP, and WISCI II after treatment were higher than those before treatment. The 6MWD, LEMS, COP, and WISCI II after treatment were higher than at the one-month follow-up appointment. The SWS, PCI, LEMS, COP, and WISCI II at the eight-week follow-up appointment were improved.

CONCLUSION:

Robotic-assisted gait training may significantly improve the immediate motor function and walking ability of children with thoracolumbar incomplete SCI.

Keywords

Robotic-assisted gait training (RAGT)spinal cord injury (SCI)motor function walking ability

1 Introduction

Paediatric spinal cord injury (SCI) refers to the structural and functional damage of the spinal cord caused by various reasons in children under the age of 15, resulting in neurological dysfunction of the spinal cord below the injury (Zhang et al., 2021). The incidence rate of paediatric SCI accounts for 0.3% –9.47% of all SCIs, and sports injuries, car accidents, and falling are the common causes (Nadarajah et al., 2018; Canosa-Hermida et al., 2019). Common motor dysfunction after paediatric SCI mainly manifests as decreased muscle strength or paralysis below the injury plane, resulting in severe limitations in the walking mobility of children with SCI (Hamid et al., 2018).

Clinically, traditional physical therapy methods, such as muscle strengthening, postural control, and assisted walking, are often used to restore or improve the walking ability of children with SCI, but the treatment operation specification cannot be unified, and the treatment effect is easily limited by the therapist’s skills (Damiano et al., 2017). In addition, the traditional physical therapy model is singular and tedious, which does not conform to the psychological characteristics of children who are prone to negative emotions such as fear of difficulty and resistance to long-term treatment; therefore, this affects the degree of cooperation (Michmizos & Krebs, 2017).

In recent years, with the rapid progress and development of medical science, technology and engineering, lower limb rehabilitation robots have gradually become a research focus for improving the mobility patients with neurological diseases (Van Hedel et al., 2016). The current rehabilitation concept of movement therapy proposes high-intensity repetitive mobility training as a therapeutic tool for the progression of neurological rehabilitation in patients with SCI (Harkema et al., 2012). According to domestic and foreign research reports, adult patients with SCI that demonstrate poor walking ability can benefit from robotic-assisted gait training (RAGT) (Shin et al., 2014; Alashram et al., 2021); in addition, it restores the motor function and walking ability in children with spastic cerebral palsy (SCP) and has a positive impact. Compared with studies on adults and children diagnosed with SCP, there are few studies on if RAGT improves the motor function and walking flexibility in children with SCI. Because the nervous system in children is continually developing, the use of lower limb rehabilitation robots for walking exercise training should achieve better results than in adult patients (Mıdık et al., 2020). This study investigates the effects of RAGT on the motor function recovery and walking ability in children with thoracolumbar incomplete SCI by observing the immediate and long-term efficacy of spatial gait parameters, energy expenditures, lower limb strength and walking activity.

2 Materials and methods

2.1 General information

Twenty-one children with thoracolumbar incomplete SCI who received physical therapy in the Department of Paediatric Physiotherapy, China Rehabilitation Research Centre, from November 2019 to August 2021, were selected by convenience sampling as the study subjects, including 3 inpatients and 18 outpatients.

Inclusion criteria: ding172 Patients with injury confirmed by the diagnostic criteria of C–D in the 2011 edition of the American Spinal Injury Association (ASIA) SCI Neurological Classification International Standard; ding173 who were between 7 and 10 years old; ding174 who were able to maintain a sitting position independently without the support of both hands for at least three min; and ding175 who were able to walk at least 10 metres with or without the use of assistive equipment.

Exclusion criteria: ding172 Patients with severe cardiopulmonary dysfunction who could not tolerate training; ding173 with severe aphasia or cognitive impairment who could not cooperate with training; ding174 in which the muscle tone of the major muscles of the lower limbs was grade 3 or above on the modified Ashworth scale; ding175 with lower limb trauma, fracture, surgery, joint pain or mobility limitation within six months before the study began; and ding176 those with lower limbs that were obviously unequal.

Elimination and drop-out criteria: ding172 patients who automatically terminated treatment; ding173 who were lost to the interview for various reasons; ding174 who violated the treatment plan; and ding175 who terminated treatment due to adverse events that were included in the adverse reaction evaluation.

The children were divided into the control and experimental groups by random number table grouping, with 10 cases in the control group and 11 in the experimental group. There were no significant differences in age, gender, height, weight, lower limb length and injury time between the two groups (P < 0.05) (Table 1).

Table 1
Comparison of general characteristics of the subjects between two groups

Demographic data Control group (n = 10) Experimental group (n = 11) t/x²-value P-value

Age (years) 8.60±1.17 8.36±1.12 0.472 0.642

Gender

(male/female, n)^# 6/4 5/6 0.444 0.505

Height (cm) 143.00±4.94 138.82±5.15 1.893 0.074

Weight (kg) 42.40±4.27 38.45±4.82 1.975 0.063

Lower limb length (cm) 70.40±2.91 70.55±3.23 –0.108 0.915

Injury time (months) 15.60±5.27 13.36±4.90 1.007 0.327

Demographic data	Control group (n = 10)	Experimental group (n = 11)	t/x²-value	P-value
Age (years)	8.60±1.17	8.36±1.12	0.472	0.642
Gender
(male/female, n)^#	6/4	5/6	0.444	0.505
Height (cm)	143.00±4.94	138.82±5.15	1.893	0.074
Weight (kg)	42.40±4.27	38.45±4.82	1.975	0.063
Lower limb length (cm)	70.40±2.91	70.55±3.23	–0.108	0.915
Injury time (months)	15.60±5.27	13.36±4.90	1.007	0.327

#“Gender” is calculated using Fisher’s exact probability method, no chi-squared values.

The study followed the basic principles of the Declaration of Helsinki and was approved by the Ethics Committee of the China Rehabilitation Research Centre (number 2018-031-1), and all the children’s families signed informed consent forms.

2.2 Methods

The children in both groups were trained in 60-minute sessions five times a week for eight weeks. The control group received 60 min of conventional physical therapy, and the experimental group received 30 min of RAGT on the basis of 30 min of conventional physical therapy. The metrics were assessed before training, eight weeks after training (an indicator assessment was performed the day after the last training session for immediate effects), and one month after training (the follow-up evaluation for long-term effects). All patients use the same assistive device before and after eight weeks of training by all primary outcome measures. And no patients received any additional therapy.

2.2.1 Conventional Physical Therapy

Both groups received conventional physical therapy. The purpose of the treatment was to improve the children’s motor function and walking ability in different situations (Hornby et al., 2020). The training contents included the following: ① Hand-knee position and three-point and two-point support training. To improve the stability of the trunk, the therapist needed to ensure that the children maintained correct posture. Each training exercise consisted of five groups of 20 repetitions, performed for 10 min each. ② Kneeling holding and kneeling walking training were conducted for 10 min, paying attention to the neutral position of the pelvis during the training and the coordination of the children’s lower limb movement and posture symmetry. According to the child’s ability level, throwing a ball was added to increase the difficulty of maintaining dynamic balance. ③ Ground walking training was conducted for 10 min, allowing the children to walk independently or with minimal assistance in an open, quiet treatment room. During walking, the therapist assisted the children in maintaining correct posture as much as possible and ensured safe training. A 30-minute session of the above training was completed daily (one session per day for the experimental group and two sessions per day for the control group).

2.2.2 Robotic-Assisted Gait Training

The experimental group received RAGT training using the lower limb rehabilitation robot training system (Lokomat Pro Paediatric Version, Hocoma, Switzerland) for treatment, and the walking time on the machine was set at 30 min. The reference values for the machine were set according to the children’s limb length, lower limb joint range of motion, muscle tone and body weight before use. A gradual reduction in weight loss from 70% to 40% of the child’s body weight was utilised, and the walking speed was set according to the individual child’s ability. During RAGT, the therapist accompanied the children and ensured that they maintained the correct gait and posture for safety.

The paediatric version of the Lokomat Pro robot consists of four parts: a weight support system, a robotic leg, a gait training treadmill and a situational simulated biofeedback system. The situational simulation biofeedback system includes a vivid game that provides real-time feedback training, which can stimulate the enthusiasm and motivation of the patients to participate in the training.

2.3 Outcome Measures

All measures were assessed by two therapists who were unaware of the subgroups to ensure the rigour of the study.

2.3.1 10-Metre Walk Test and Six-Minute Walk Distance

The 10-metre walk test (10MWT) was used as a speed test. The therapist instructed the children to complete three tests on the 10-metre walkway at an appropriate speed and at the fastest speed, respectively, and the average was taken to calculate the children’s self-selected walking speed (SWS) and maximum walking speed (MWS) (Cho et al., 2016). The six-minute walk distance (6MWD) was used for the walking distance test. The children were instructed to walk and turn at an appropriate speed on a 10-metre walkway. The children wore or used any required forms of orthotic or walking aids (Stevens et al., 2015).

2.3.2 Energy Expenditure

The physiological cost index (PCI) measured changes in the children’s heart rate (beats per min) during walking. It has been used as a reliable measure of the energy cost during walking among those with spinal cord lesions at or below the T3 level. A heart rate monitor (Polar Accurex Plus, Sark Products, Finland) was used to acquire data to calculate the PCI, which was obtained by dividing the difference between the exercising and resting heart rates by the walking speed; it was used as a method to determine the energy consumption of the children during walking (Han & Yun, 2020).

2.3.3 Lower Extremity Strength

The lower extremity motor score (LEMS) assessed the children’s lower extremity strength. Specific muscle groups were examined, including the hip flexors, knee extensors, ankle dorsiflexors, toe extensors and ankle plantar flexors. According to the results of the manual muscle test (MMT), muscle strength levels of 0–5 were used as the score, and the total of the muscle scores of the lower extremity test constituted the final score (Donenberg et al., 2019).

2.3.4 Balance and Posture Control

The walking data of the children were collected with an infrared motion capture system (SMART-D 400, BTS, Italy) for a three-dimensional gait analysis that obtained the stability limit data of the children in eight directions: forward, backward, right and oblique. The centre of pressure (COP) area was then calculated by the software (Chien & Hsu, 2018).

2.3.5 Walking Capabilities

The Walking Index for Spinal Cord Injury II (WISCI II) was used to evaluate the walking capabilities of the children. The scale is divided into 20 categories according to the physical assistance facilities and supports required for walking, with scores ranging from 0 points (unable to walk) to 20 points (walking at least 10 metres without any assistance). The WISCI II is a commonly used index to assess the walking mobility of children with SCI and has consistent validity with the ASIA disability classification system (Calhoun Thielen et al., 2017).

2.4 Statistical Analysis

The SPSS 26.0 statistical software package was used for data processing. The Shapiro–Wilk test was used for the test of normality. Fisher’s exact test was used to compare the enumeration data in the general data of the patients, and a group design t-test was used to compare measurement data in the general data. An intra-group comparison and a comparison of the measurement data in the two groups obtained before and after treatment were performed using two-way repeated measures ANOVA, and P < 0.05 indicated a significant difference. Sample size estimation: GPower software was used to estimate the sample size, using “independent sample t-test, effect size of 0.5, significance levelαof 0.05,power calculation 1-βof 80%,” and the expected sample size of the test and control groups was 64 cases per group.

3 Results

There was one case that dropped out in the experimental group during the test at follow-up (1 male) due to his inability to adhere to the treatment. A total of 20 children completed the eight-week trial and were assessed at one month after the completion of treatment.

3.1 Maximum Walking Speed and Six-Minute Walk Distance

Within group pairwise comparisons,in the experimental and control group, the SWS was significantly higher after treatment and at the one-month follow-up visit than before treatment(P < 0.05). Additionally, the SWS were significantly lower at the one-month follow-up visit than after treatment (P < 0.05). In a comparison between the groups, there were no significant differences in the SWS before training (P > 0.05). Compared with the control group, the SWS of the experimental group were improved after treatment (0.32±0.09 vs 0.44±0.12, P < 0.05) and at the one-month follow-up visit (0.27±0.09 vs 0.40±0.15, P < 0.05).

Within group pairwise comparisons, the MWS of the experimental group was significantly higher after treatment and at the one-month follow-up visit than before treatment(P < 0.05). Additionally, the MWS of the experimental group were significantly lower at the one-month follow-up visit than after treatment (P < 0.05). In the control group, the MWS before treatment, after treatment, and at the one-month follow-up visit were no significant differences (P > 0.05). In a comparison between the groups, there were no significant differences in the MWS before treatment(P > 0.05). Compared with the control group, the SWS of the experimental group were improved after treatment (0.45±0.09 vs 0.60±0.12, P < 0.05) and at the one-month follow-up visit (0.40±0.09 vs 0.53±0.17, P < 0.05).

Within group pairwise comparisons, in the experimental and control group, the 6MWD was significantly higher after treatment and at the one-month follow-up visit than before treatment (P < 0.05). Additionally, the 6MWD were significantly lower at the one-month follow-up visit than after treatment (P < 0.05). In a comparison between the groups, there were no significant differences in the 6MWD before treatment(P > 0.05). Compared with the control group, the 6MWD of the experimental group were improved after treatment (119.64±26.03 vs 162.26±47.41, P < 0.05) and at the one-month follow-up visit (112.73±28.22 vs 148.74±47.15, P < 0.05). Results are shown in Table 2.

Table 2
Comparison of SWS, MWS, 6MWD between the two groups of children at different times

Group N Before treatment After treatment At follow-up F value P value

SWS-C (m/s) 10 0.29±0.09 0.32±0.09^# 0.27±0.09^rr 3.213 <0.01^

SWS-E (m/s) 10 0.27±0.15 0.44±0.12^# 0.40±0.15^# 5.981 <0.01^

F value 1.224 2.677 2.133

P value >0.05 <0.05 >0.05

MWS-C (m/s) 10 0.44±0.10 0.45±0.09 0.40±0.09^rr 2.641 >0.05

MWS-E (m/s) 10 0.38±0.13 0.60±0.12^# 0.53±0.17 6.744 <0.01^

F value 1.688 3.645 2.031

P value >0.05 <0.05 >0.05

6MWD-C (m) 10 109.73±28.25 119.64±26.03^# 112.73±28.22^rr 9.310 <0.01^

6MWD-E (m) 10 104.52±52.15 162.26±47.41^# 148.74±47.15^rr 12.781 <0.01^**

F value 1.552 6.381 5.87

P value >0.05 <0.01 <0.01

Group	N	Before treatment	After treatment	At follow-up	F value	P value
SWS-C (m/s)	10	0.29±0.09	0.32±0.09^#	0.27±0.09^rr	3.213	<0.01^**
SWS-E (m/s)	10	0.27±0.15	0.44±0.12^#	0.40±0.15^#	5.981	<0.01^**
F value		1.224	2.677	2.133
P value		>0.05	<0.05	>0.05
MWS-C (m/s)	10	0.44±0.10	0.45±0.09	0.40±0.09^rr	2.641	>0.05
MWS-E (m/s)	10	0.38±0.13	0.60±0.12^#	0.53±0.17	6.744	<0.01^**
F value		1.688	3.645	2.031
P value		>0.05	<0.05	>0.05
6MWD-C (m)	10	109.73±28.25	119.64±26.03^#	112.73±28.22^rr	9.310	<0.01^**
6MWD-E (m)	10	104.52±52.15	162.26±47.41^#	148.74±47.15^rr	12.781	<0.01^**
F value		1.552	6.381	5.87
P value		>0.05	<0.01	<0.01

Note: “-C” means control group,“-E” means experimental group; Comparison of before treatment, after treatment and at follow-up (^*P < 0.05, ^**P < 0.01); Compared with control group (^#P < 0.05, ^# #P < 0.01); After treatment versus follow-up (^r P < 0.05, ^rrP < 0.01).

3.2 Energy Expenditure

In a comparison between the groups, there was no significant difference in PCI before treatment (P > 0.05). Compared with the control group, the PCI of the experimental group was decreased after treatment (101.89±29.86 vs 67.44±36.34, P < 0.05) and at the one-month follow-up visit (135.00±57.90 vs 64.22±24.69, P < 0.05).

The PCI in control group and experimental group before treatment, after treatment, and at the one-month follow-up visit were not the same, and the difference was statistically significant (P < 0.01, P < 0.05). Within group pairwise comparisons, in the experimental group, the PCI was significantly lower after treatment and at the one-month follow-up visit than before treatment (67.44±36.34 vs 188.56±45.40, P < 0.05; 64.22±24.69 vs 188.56±45.40, P < 0.01). The PCI in the control group was significantly lower after treatment than before treatment (P < 0.01). Additionally, the PCI was significantly higher at the one-month follow-up visit than after treatment (P < 0.05). Results are shown in Table 3.

Table 3
Comparison of PCI between the two groups of children at different times

Group N Before treatment After treatment At follow-up F value P value

PCI-C 10 142.78±32.88 101.89±29.86^# 135.00±57.90 ^r 9.784 <0.01^**

PCI-E 10 188.56±45.40 67.44±36.34^# 64.22±24.69^# # 3.457 <0.05^*

F value 2.464 2.899 3.233

P value >0.05 <0.05 <0.05

Group	N	Before treatment	After treatment	At follow-up	F value	P value
PCI-C	10	142.78±32.88	101.89±29.86^#	135.00±57.90 ^r	9.784	<0.01^**
PCI-E	10	188.56±45.40	67.44±36.34^#	64.22±24.69^# #	3.457	<0.05^*
F value		2.464	2.899	3.233
P value		>0.05	<0.05	<0.05

3.3 Lower Extremity Strength

In a comparison between the two groups, there was no significant difference in the LEMS before treatment, after treatment and at the one-month follow-up visit (P > 0.05).

The LEMS in the control group and experimental group before treatment, after treatment, and at the one-month follow-up visit were not the same, and the difference was statistically significant (P < 0.01, P < 0.01). Within group pairwise comparisons, the LEMS was significantly higher after treatment (35.11±8.27 vs 24.11±9.68, P < 0.05) and at the one-month follow-up visit (31.00±7.98 vs 24.11±9.68, P < 0.05) than before treatment; in addition, the LEMS at the one-month follow-up visit was significantly lower than after treatment (P < 0.01). In the control group, the difference between the LEMS at the one-month follow-up visit and after treatment was statistically significant (P < 0.01). Results are shown in Table 4.

Table 4
Comparison of LEMS between the two groups of children at different times

Group N Before treatment After treatment At follow-up F value P value

LEMS-C 10 28.00±4.80 30.78±4.76 28.33±4.33 ^rr 5.774 <0.01^

LEMS-E 10 24.11±9.68 35.11±8.27^# 31.00±7.98^#^rr 6.231 <0.01^

F value 1.112 1.555 1.788

P value >0.05 >0.05 >0.05

Group	N	Before treatment	After treatment	At follow-up	F value	P value
LEMS-C	10	28.00±4.80	30.78±4.76	28.33±4.33 ^rr	5.774	<0.01^**
LEMS-E	10	24.11±9.68	35.11±8.27^#	31.00±7.98^#^rr	6.231	<0.01^**
F value		1.112	1.555	1.788
P value		>0.05	>0.05	>0.05

3.4 Balance and Postural Control

When comparing the groups, there was no significant difference in the COP before treatment (P > 0.05), and the COP was significantly higher in the experimental group than in the control group after treatment (164.24±35.60 vs 131.69±27.07, P < 0.05) and at the one-month follow-up visit (153.31±35.76 vs 121.63±24.69, P < 0.05).

The COP in control group and experimental group before treatment, after treatment, and at the one-month follow-up visit were not the same, and the difference was statistically significant (P < 0.01, P < 0.01). Within a group pairwise comparison, the COP in both groups was significantly higher after treatment than before treatment (164.24±35.60 vs 120.99±31.48, P < 0.05), and the COP was significantly lower at the one-month follow-up visit than after treatment (P < 0.01). The difference between the COP at the one-month follow-up visit and before treatment was statistically significant (P < 0.01) only in the experimental group. Results are shown in Table 5.

Table 5
Comparison of COP between the two groups of children at different times

Group N Before treatment After treatment At follow-up F value P value

COP-C (cm²) 10 119.04±26.01 131.69±27.07^# 121.63±24.69 ^rr 8.682 <0.01^

COP-E (cm²) 10 120.99±31.48 164.24±35.60^# 153.31±35.76^#^rr 7.655 <0.01^

F value 1.666 2.841 3.045

P value >0.05 <0.05 <0.05

Group	N	Before treatment	After treatment	At follow-up	F value	P value
COP-C (cm²)	10	119.04±26.01	131.69±27.07^#	121.63±24.69 ^rr	8.682	<0.01^**
COP-E (cm²)	10	120.99±31.48	164.24±35.60^#	153.31±35.76^#^rr	7.655	<0.01^**
F value		1.666	2.841	3.045
P value		>0.05	<0.05	<0.05

3.5 Walking Capability

There was no significant difference in the WISCI II scores between the two groups before training (P > 0.05), but the WISCI II score in the experimental group was significantly higher than in the control group after treatment (19.00±1.12 vs 16.22±2.17, P < 0.05) and at the one-month follow-up visit (18.00±1.80 vs 15.78±2.05, P < 0.05).

The WISCI II score in the control group and experimental group before treatment, after treatment, and at the one-month follow-up visit were not the same, and the difference was statistically significant (P < 0.05, P < 0.01). Within a group pairwise comparison, in the experimental group, the WISCI II score was significantly higher after treatment (19.00±1.12 vs 16.22±1.30, P < 0.01) and at the one-month follow-up visit than before treatment (18.00±1.80 vs 16.22±1.30, P < 0.05), and the WISCI II score was higher after treatment than at the one-month follow-up visit (P < 0.05). In the control group, the WISCI II score was significantly lower at the one-month follow-up visit than after treatment (P < 0.05). Results are shown in Table 6.

Table 6
Comparison of WISCI II between the two groups of children at different times

Group N Before treatment After treatment At follow-up F value P value

WISCI II-C 10 15.89±2.15 16.22±2.17 15.78±2.05 ^r 3.427 <0.05

WISCI II-E 10 16.22±1.30 19.00±1.12^# # 18.00±1.80 ^# r 6.831 <0.01^**

F value 1.334 3.051 3.112

P value >0.05 <0.05 <0.05

Group	N	Before treatment	After treatment	At follow-up	F value	P value
WISCI II-C	10	15.89±2.15	16.22±2.17	15.78±2.05 ^r	3.427	<0.05
WISCI II-E	10	16.22±1.30	19.00±1.12^# #	18.00±1.80 ^# r	6.831	<0.01^**
F value		1.334	3.051	3.112
P value		>0.05	<0.05	<0.05

4 Discussion

4.1 Robot-Assisted Gait Training Rehabilitation Results

This study aimed to analyse and demonstrate the immediate and long-term effects of RAGT on the recovery of motor function and walking ability in children with thoracolumbar incomplete SCI. The results showed that after eight weeks of training, the children in the experimental group (those treated with RAGT) formed a new gait pattern, and their immediate motor function and walking ability were significantly improved compared with the control group; however, the long-term efficacy of RAGT needs to be further analysed.

The results between the groups after training in this study showed that the SWS, MWS, 6MWD, PCI, LEMS, COP and the WISCI II score of the experimental group were significantly improved compared with the control group, indicating that RAGT could improve the walking speed and distance of children with SCI, enhance their balance control ability, and effectively improve their walking ability compared with conventional physical therapy. These findings are consistent with previous research results on adult patients with SCI. According to the meta-analysis performed by Fang et al., RAGT can effectively improve the walking speed of patients with SCI (Fang et al., 2020). However, Nam et al. (Nam 2017) concluded that RAGT was more effective in improving the walking distance in patients with SCI in the acute phase. Another systematic review concluded that RAGT was beneficial in improving balance function and walking ability in the upright position in patients with SCI (Hayes et al., 2018). This is because lower limb rehabilitation robots can provide high-speed stepping movements that last for a longer time, and high-intensity or high-volume repetitive exercises are a key component of activity-based mobility training programs (Gandhi et al., 2017; Cruz et al., 2015). Recovery after nerve injury involves the principle of neuroplasticity, and prolonged high-intensity treatment is essential to promote myelin regeneration and the development of new synapses. The results of this study indicated that RAGT helped children improve their walking speed through high-intensity repetitive exercise and extended the spatial distance travelled per unit of time. The children in the experimental group had a significantly larger COP envelope area after the eight-week period, which indicated that they gained more stable balance during walking, that they reduced the need for assistance while walking, and that their walking ability was effectively improved. The improvement of balance and walking abilities effectively reduced unnecessary energy expenditure in the original walking pattern for various reasons, including fear of falling and maintaining posture. As was demonstrated in this study’s intra-group comparison, the PCI in the experimental group after training was significantly lower than before training, which indicated that the children in the experimental group significantly improved their walking endurance.

The results also indicated that the LEMS of the experimental and control groups after training were significantly higher than before training, indicating that both RAGT and conventional physiotherapy were effective in improving short-term lower limb strength in children with thoracolumbar incomplete SCI, which is also consistent with the results of several studies at home and abroad (Alcobendas-Maestro et al., 2012; Baunsgaard et al., 2018). However, the results of the intra-group comparison after training and at the one-month follow-up visit showed that the LEMS in both groups were significantly lower than the immediate post-training effect, which may be related to the reduced muscle activity in hypermobility.

4.2 Implications for Rehabilitation Practice

One study reported that normal individuals lose approximately 2% –3% of their calf and thigh cross-sectional areas after five days of absolute bed rest (Mulder et al., 2015). This is due to the interaction between human muscle properties and the central nervous system, where muscles receive commands from motor neurons in the spinal cord to perform movement, and sensory input is transmitted back to the central nervous system by proprioceptors during movement to facilitate the next movement. In the case of inactivity or reduced activity, less force is required by the muscle, and the proprioceptive input the muscle produces is altered accordingly, resulting in a further reduction in muscle force (Canu et al., 2019). Compared with normal individuals, the muscle function of people with motor dysfunction caused by nerve injury is more susceptible to such hypokinetic states (Kramer et al., 2017). In the present study, the children in both groups had significantly lower motor activity during the one-month follow-up period than during the eight-week training period, which may have led to a degradation in motor function and resulted in a further reduction in motor activity, leading to a decrease in muscle strength. Although the lower limb strength was lower than it was after training, the LEMS in the experimental group at the one-month follow-up visit was still significantly higher than before training, whereas there was no significant difference in the control group, indicating that although the long-term effects of RAGT on lower limb strength in children with incomplete SCI (e.g. achieving effect and dose) are not yet clear, it still had a beneficial overall effect on the children’s lower limb strength.

In addition, significant differences were found between the before treatment and the one-month follow-up results in the SWS, COP, PCI, LEMS and WISCI II scores, but not in the MWS and 6MWD. The intra-group comparison results of the experimental group showed that the 6MWD, LEMS, WISCI II scores and COP at the one-month follow-up visit were significantly lower than the immediate post-training effect. These findings suggest that the long-term effects of RAGT in improving motor function and walking ability in children with SCI are better than conventional physiotherapy. However, there is still a gap when the one-month follow-up results are compared with the immediate post-training effects of insufficient self-sustainability. In comparison, the long-term effects of RAGT on MWS and walking distance are not significantly better than conventional physiotherapy.

Some studies have shown (Anwer et al., 2014) that adult patients with SCI significantly increased their walking speed and distance after completing RAGT for 40-minute sessions twice a day, five days a week, for eight weeks. Other studies have concluded that children can tolerate a maximum of five hours of treatment intensity per day while ensuring that the treatment is rich and can generate sufficient interest for them (Esquenazi & Talaty, 2019). In future studies, based on the above research results, more reasonable trial protocols can be designed according to the tolerance and interests of children with SCI, and in-depth studies on RAGT intervention duration, dose, and other indicators can be conducted to analyse the long-term efficacy of RAGT in children with SCI.

4.3 Limitations and Future Directions

Although the present study confirmed significant immediate benefits of RAGT in improving the motor function and walking ability in children with thoracolumbar incomplete SCI, there were some limitations. First, the sample size of this study was small, and the findings may be subject to error. Pilot study details are in the attachment. Second, only children with incomplete SCI in the thoracolumbar segment were included in this study; no studies have been conducted on children with cervical segment injuries. Third, the reliability of the MMT in measuring the lower limb muscles was low and susceptible to the ‘ceiling’ effect, which lacks sensitivity to changes in muscle strength and has poor inter-rater reliability and limited accuracy (Khamis et al., 2018). Therefore, the authors of this study hope to carry out a large-scale, well-designed, multi-centre clinical study in the future to expand the scope of the tested children’s injury plane and conduct a long-term follow-up investigation to further study the effect of RAGT on the recovery of motor function and walking ability in children with incomplete SCI.

5 Conclusion

RAGT, characterised by high-intensity and repetitive exercise, can significantly improve the immediate motor function and walking ability in children with thoracolumbar incomplete SCI, but its long-term efficacy is not yet proven as superior to conventional physiotherapy. Based on the results of this study, a large sample size and well-designed multi-centre clinical study should be conducted in the future to expand the range of the injury planes of the children tested. It should also include an analysis and comparison of the changes in joint movement trajectories and lower limb EMG signals in children with SCI before and after training and conduct long-term efficacy follow-up investigations to determine the effects of RAGT on the recovery of motor function and walking ability in children with incomplete SCI.

Footnotes

Acknowledgments

The authors have no acknowledgments.

Conflict of interest

None of the authors have any personal, financial, commercial, or academic conflicts of interest to report.

Funding

This study was supported by the Major Project of China Rehabilitation Research Center (No.2018zx-22).

References

Alashram,

A. R.

, Annino,

, & Padua,

(2021) Robot-assisted gait training in individuals with spinal cord injury:A systematic review for the clinical effectiveness of Lokomat. J Clin Neurosci 91, 260–269.

Alcobendas-Maestro,

, Esclarín-Ruz,

, Casado-López,

R.M.

, Muñoz-González,

, Pérez-Mateos,

, González-Valdizán,

, et al., (2012) Lokomatrobotic-assisted versus overground training within 3 to 6 months ofincomplete spinal cord lesion: randomized controlled trial. Neurorehabil Neural Repair 26(9), 1058–1063.

Anwer,

, Equebal,

, Palekar,

T. J.

, Nezamuddin,

, Neyaz,

, & Alghadir,

(2014) Effect of locomotor training on motor recovery and walking ability in patients with incomplete spinal cord injury: a case series. J Phys Ther Sci 26(6), 951–953.

Baunsgaard,

C. B.

, Nissen,

U. V.

, Brust,

A. K.

, Frotzler,

, Ribeill,

, Kalke,

Y. B.

, et al., (2018) Gait training after spinal cord injury: safety, feasibility and gait function following 8 weeks of training with the exoskeletons from Ekso Bionics. Spinal Cord 56(2), 106–116.

Calhoun Thielen

, Sadowsky,

, Vogel,

L. C.

, Taylor,

, Davidson,

, Bultman,

, et al., (2017) Evaluation of the Walking Index for Spinal Cord Injury II (WISCI-II) in children with Spinal Cord Injury (SCI. Spinal Cord 55(5), 478–482.

Canosa-Hermida,

, Mora-Boga,

, Cabrera-Sarmiento,

J. J.

, Ferreiro-Velasco,

M. E.

, Salvador-de la Barrera,

, Rodríguez-Sotillo,

, et al. (2019) Epidemiology of traumatic spinal cord injury in childhood and adolescence in Galicia, Spain: report of the last -years. J Spinal Cord Med 42(4), 423–429.

Canu,

M. H.

, Fourneau,

, Coq,

J. O.

, Dannhoffer,

, Cieniewski-Bernard,

, Stevens,

, et al. (2019) Interplay between hypoactivity, muscle properties and motor command: How to escape the vicious deconditioning circle? Ann Phys Rehabil Med 62(2), 122–127.

Chien,

J. E.

, & Hsu,

W. L.

(2018) Effects of Dynamic Perturbation-Based Training on Balance Control of Community-Dwelling Older Adults. Sci Rep 8(1), 17231.

Cho,

, Hwang,

, & Chung,

(2016) Treadmill Training with Virtual Reality Improves Gait, Balance, and Muscle Strength in Children with Cerebral Palsy. Tohoku J Exp Med 238(3), 213–218.

10.

Cruz,

S. P.

, Cimolin,

, & Gil-Agudo,

(2015) Spinal cord injury in pediatric age in Spain. Reality of a national reference center. Childs Nerv Syst 31(6), 917–921.

11.

Damiano,

D. L.

, Stanley,

C. J.

, Ohlrich,

, & Alter,

K. E.

(2017) Task-Specific and Functional Effects of Speed-Focused Elliptical or Motor-Assisted Cycle Training in Children With Bilateral Cerebral Palsy: Randomized Clinical Trial. Neurorehabil Neural Repair 31(8), 736–745.

12.

Donenberg,

J. G.

, Fetters,

, & Johnson,

(2019) The effects of locomotor training in children with spinal cord injury: a systematic review. Dev Neurorehabil 22(4), 272–287.

13.

Esquenazi,

, & Talaty,

(2019) Robotics for Lower Limb Rehabilitation. Phys Med Rehabil Clin N Am 30(2), 385–397.

14.

Fang,

C. Y.

, Tsai,

J. L.

, Li,

G. S.

, Lien,

A. S.

, & Chang,

Y. J.

(2020) Effects of Robot-Assisted Gait Training in Individuals with Spinal Cord Injury: A Meta-analysis. Biomed Res Int 2020, 2102785.

15.

Gandhi,

, Chan,

, Verrier,

M. C.

, Pakosh,

, & Musselman,

K. E.

(2017) Training to Improve Walking after Pediatric Spinal Cord Injury: A Systematic Review of Parameters and Walking Outcomes. J Neurotrauma 34(9), 1713–1725.

16.

Hamid,

, Averbeck,

M. A.

, Chiang,

, Garcia,

, Al Mousa

R. T.

, Oh,

S. J.

, et al. (2018) Epidemiology and pathophysiology of neurogenic bladder after spinal cord injury. World J Urol 36(10), 1517–1527.

17.

Han,

Y. G.

, & Yun,

C. K.

(2020) Effectiveness of treadmill training on gait function in children with cerebral palsy: meta-analysis. J Exerc Rehabil 16(1), 10–19.

18.

Harkema,

S. J.

, Hillyer,

, Schmidt-Read,

, Ardolino,

, Sisto,

S. A.

, & Behrman,

A. L.

(2012) Locomotor training: as a treatment of spinal cord injury and in the progression of neurologic rehabilitation. Arch Phys Med Rehabil 93(9), 1588–97.

19.

Hayes,

S. C.

, James Wilcox,

C. R.

, Forbes White,

H. S.

, & Vanicek,

(2018) The effects of robot assisted gait training on temporal-spatial characteristics of people with spinal cord injuries: A systematic review. J Spinal Cord Med 41(5), 529–543.

20.

Hornby,

T. G.

, Reisman,

D. S.

, Ward,

I. G.

, Scheets,

P. L.

, Miller,

, Haddad,

, et al. (2020) Clinical Practice Guideline to Improve Locomotor Function Following Chronic Stroke, Incomplete Spinal Cord Injury, and Brain Injury. J Neurol Phys Ther 44(1), 49–100.

21.

Khamis,

, Herman,

, Krimus,

, & Danino,

(2018) Is functional electrical stimulation an alternative for orthotics in patients with cerebral palsy? A literature review. Eur J Paediatr Neurol 22(1), 7–16.

22.

Kramer,

, Kümmel,

, Mulder,

, Gollhofer,

, Frings-Meuthen,

, & Gruber,

(2017) High-Intensity Jump Training Is Tolerated during 60 Days of Bed Rest and Is Very Effective in Preserving Leg Power and Lean Body Mass: An Overview of the Cologne RSL Study. PLoS One 12(1), e0169793.

23.

Michmizos,

K. P.

, & Krebs,

H. I.

(2017) Pediatric robotic rehabilitation: Current knowledge and future trends in treating children with sensorimotor impairments. NeuroRehabilitation 41(1), 69–76.

24.

Mıdık,

, Paker,

, Bugdaycı

, Mıdık,

A. C.

, (2020) Effects of robot-assisted gait training on lower extremity strength, functional independence, and walking function in men within complete traumatic spinal cord injury. Turk J Phys MedRehabil 66(1), 54–59.

25.

Mulder,

, Clément,

, Linnarsson,

, Paloski,

W. H.

, Wuyts,F.

, Zange,

, et al. (2015) Musculoskeletal effects of 5 days of bed rest with and without locomotion replacement training. Eur J Appl Physiol 115(4), 727–738.

26.

Nadarajah,

, Jauregui,

J. J.

, Perfetti,

, Shasti,

, Koh,

E. Y.

, Henn,

R.F.

3rd(2018) What are the trends and demographics in sports-related pediatric spinal cord injuries? Phys Sportsmed 46(1), 8–13.

27.

Nam,

K. Y.

, Kim,

H. J.

, Kwon,

B. S.

, Park,

J. W.

, Lee,

H. J.

, & Yoo,

(2017) Robot-assisted gait training (Lokomat) improves walking function and activity in people with spinal cord injury: a systematic review. J Neuroeng Rehabil 14(1), 24.

28.

Shin,

J. C.

, Kim,

J. Y.

, Park,

H. K.

, & Kim,

N. Y.

(2014) Effect of robotic-assisted gait training in patients with incomplete spinal cord injury. Ann Rehabil Med 38(6), 719–725.

29.

Stevens,

S. L.

, Caputo,

J. L.

, Fuller,

D. K.

, & Morgan,

D. W.

(2015) Effects of underwater treadmill training on leg strength, balance, and walking performance in adults with incomplete spinal cord injury. J Spinal Cord Med 38(1), 91–101.

30.

Van Hedel

H. J. A.

, Meyer-Heim,

, & Rüsch-Bohtz,

(2016) Robot-assisted gait training might be beneficial for more severely affected children with cerebral palsy. Dev Neurorehabil 19(6), 410–415.

31.

Zhang,

, Zhou,

H. J.

, Liu,

G. L.

, Zheng,

, Lu,

X. L.

, Kang,

H. Q.

,et al. (2021) Survey of Neurological Function and Complication of Pediatric Spinal Cord Injury. China Rehabilitation Theory and Practice 27(06), 706–711.