Efficiency of botulinum toxin injection into the arm on postural balance and gait after stroke

The purpose of this study was to clarify the association between improvement of spasticity in hemiplegic patient’s upper extremity with Botulinum toxin injection and improvement in postural balance and gait function. For this prospective cohort study, sixteen hemiplegic stroke patients with upper extremity spasticity were recruited. The plantar pressure with gait parameters, postural balance parameters, Modified Ashworth Scale, and Modified Tardieu Scale were evaluated before, 3 weeks and 3 months after Botulinum toxin A (BTxA) injection. Spasticity of hemiplegic upper extremity before, and after BTxA injection were significantly changed. Plantar pressure overload in affected side was reduced after BTxA injection. The mean X-speed and the horizontal distance decreased in postural balance analysis with eyes-opened test. Improvement in hemiplegic upper extremity spasticity showed positive correlation with gait parameters. In addition, improvement in hemiplegic upper extremity spasticity was positively correlated with change in balance parameters in postural balance analysis with eyes-closed and dynamic tests. This study focused on the effect of stroke patient’s hemiplegic upper extremity spasticity on their gait and balance parameters and identified that the BTxA injection on hemiplegic patient’s spastic upper extremity improve postural balance and gait function.

www.nature.com/scientificreports/ We hypothesized that improvement in spasticity in the upper extremities of hemiplegic patients by BTx injection would improve postural balance and gait function. In order to establish a clear relationship, we investigate the correlation between changes in quantitative evaluation of postural balance and gait function and changes in upper limb spasticity before and after botulinum toxin injection.

Methods
Participants. This study was prospective cohort study. A total of 16 post-stroke patients with hemiplegia were recruited from the Department of Physical Medicine and Rehabilitation, Korea University Guro Hospital, between March 1, 2020, and May 31, 2021, for this study. Eleven patients were male and 5 were female. The inclusion criteria were (1) age > 18 years, (2) at least 6 weeks after stroke diagnosis, (3) upper-extremity (elbow, wrist and finger) spasticity Modified Ashworth Scale (MAS) score > 2, and (4) ability to stand and walk safely without help or assistance. The exclusion criteria were as follows: (1) improper indication for botulinum toxin A (BTxA) injection, for example, myasthenia gravis, Eaton-Lambert syndrome, amyotrophic lateral sclerosis, and motor neuropathy; (2) previous contracture and/or deformity of the upper extremities; (3) concurrent peripheral neuropathy and/or myopathy; and (4) difficulty in participating in the study due to cognitive impairment.
There were 12 patients who had never received a BTx injection before this study and 4 patients who had received BTx before this study. Those 4 patients were recruited at least 6 months from BTx injection. All participants continued their previously scheduled oral medication and rehabilitation.
Botulinum toxin intervention. Appropriate arm muscles for BTxA injection were clinically selected by physical examination. In our study, we used Nabota ® (Daewoong Pharmaceutical Co. Ltd., Seoul, Korea) as a BTx agent. It is a protein of high purity and quality obtained from the natural strain Clostridium botulinum (type A) 27 . BTxA was diluted with 0.9% sodium chloride solution and injected into the bellies of the upper-extremity muscles. Muscles to be injected were determined individually, according to patient's condition. The toxin dose was established for each patient: it ranged between 80 and 300 unit. From 1 up to 3 local points of injections considering muscle size have been selected. Intramuscular injections with electrostimulation guidance was performed once by a physiatrist with 20 years of clinical experience in stroke-related spasticity. Participants continued their routine schedule of medications and rehabilitation programs and were required to observe any considerable changes during the study.
Spasticity evaluation and functional measurement. All participants were assessed at baseline (preintervention), three weeks after intervention (post-intervention), and three months after intervention (followup) with BTxA. The MAS and Modified Tardieu Scale (MTS) were used to assess the degree of spasticity. Also, the hemiplegic upper extremities were evaluated using the Fugl-Meyer Assessment (FMA) and Action Research Arm Test (ARAT). In addition, the hemiplegic lower extremities were evaluated using the FMA. Furthermore, functional ambulatory category (FAC) was estimated.
Plantar foot pressure analysis. In every session, plantar pressure with center of pressure (CoP) excursion was analyzed using an insole pressure measurement system (Medilogic ® , T&T medilogic Medizintechnik GmbH, Unterschleissheim, Germany) 28,29 . In our laboratory, different-sized insoles were used, consisting of resistive sensors based on size. The sensor density was 0.79/cm 2 , and the sensor could withstand a maximum pressure of 64 N/cm 2 . The plantar pressure data were collected at a sampling frequency of 60 Hz and transmitted by cables from the insoles to the analog-digital box worn at the participant's waist. The data were transmitted from the analog-digital box to a computer through a wireless connection 30 .
A static standing trial was conducted to confirm appropriate positioning of the insole, followed by a dynamic walking trial. All patients walked at least 5 m with normal gait and more than 12 steps without any personal assistance or assistive devices. Gait speed was determined for each participant in a comfortable and tolerable range. Dynamic measurements were conducted for at least six trials, and the most stable and best-performing trial was selected. Pressure load (%), and spatiotemporal parameters (gait speed (m/s), stride length (m), stance and swing phase duration (s)) were measured.

Postural balance measurement.
A computer-based force platform test (Good Balance, Metitur Ltd,. Finland, www. metit ur. com) was used to evaluate postural balance 31,32 . This system consisted of an equilateral triangular force platform (width 800 mm, height 70 mm, with strain gauge transducers at each corner of the platform), a three-channel direct current amplifier, an eight-channel 12-byte analog-to-digital converter (sampling frequency 50 Hz), and a program installed on a laptop computer.
Based on the vertical force signals from each corner of the platform, the system calculated the X (mediolateral, ML) and Y (anteroposterior, AP) coordinates of the CoP affecting the platform while the patients were standing on it. Based on the coordinate values for X and Y, various balance parameters such as mean speed of the movement of the CoP in the ML direction (mean X-speed (mm/s)), mean speed of the movement of the CoP in the AP direction (mean Y-speed (mm/s)), performing time (s), full distance of CoP movement (total distance (mm)), distance of the ML direction made by the CoP (horizontal distance (mm)) and distance of the AP direction made by the CoP (vertical distance (mm)) were calculated 33 .In every test, the participants stood on the point with the hindfoot of both feet located at center 'O' mark of the triangular force platform and both feet 3 cm apart. Additionally, both ankles were externally rotated by 15°. Balance was tested in three different conditions: (1) In the eyes-opened test, the participant was in the normal standing position, feet slightly apart, arms in a relaxed position, and gaze fixed on a computer monitor (test duration was 30 s). (2) In the eyes-closed test, all conditions were the same as those in the eyes-opened test. (3)  www.nature.com/scientificreports/ perform a specific movement task presented on the monitor. Evaluations for each of the three conditions were conducted for at least two trials, and the trial with the best performance was selected.
Statistical analysis. Descriptive statistics were used to assess the general characteristics and demographic factors of the patients. Friedman's test was employed to detect differences in upper-extremity spastic measurements, spatiotemporal gait parameters, and calculated coordinates of the CoP at pre-intervention (baseline), post-intervention, and follow-up evaluation. The Wilcoxon signed-rank test was computed to determine the difference between pre-(baseline) and post-injection. Spearman's rho was used to analyze the correlation between improvement in spastic measurements, foot pressure, and postural balance parameters. SPSS version 26.0 software (SPSS Inc., Chicago, IL, USA) was used for all analyses, with the statistical significance level set at p < 0.05.

Ethics.
All patients recruited for this study provided written consent. The participants were informed prior to evaluation that the collected data could only be used for study purposes, and they had the right to refuse this use at any time. The study was approved by the institutional review board of the Korea University Guro Hospital (2019GR0159), and was conducted in accordance with the Declaration of Helsinki. To increase the quality of reporting of this observational study, STROBE guidelines were followed 34 . All participants signed an informed consent form before participating in the study.

Results
Sixteen patients were enrolled in this study. Patient characteristics, demographic factors, target muscle with exact BTxA dosage, and baseline upper extremities evaluations are described in

Effect of BTxA injection on spasticity in affected upper extremities.
Compared to the pre-injection evaluation, the MAS grades of the elbow extensor and flexor, wrist flexor, finger extensor, and flexor were significantly decreased at post-injection and follow-up. The MTS (R2-R1) of the elbow flexor and wrist extensor was significantly changed, as shown in Table 2.
Effect of BTxA injection on foot pressure analysis. During walking, plantar pressure on the affected side after BTxA injection was significantly reduced in the post-intervention and follow-up evaluations (p = 0.020). This was reflected by the overall load, as presented in Fig. 1. Other spatiotemporal gait parameters were described in Table 3.
Effect of BTxA injection on postural balance analysis. Postural balance improved after BTxA injection in the hemiplegic spastic upper extremities. This was confirmed by the decreased balance parameters, including mean X-speed and horizontal distance in the eyes-opened state, as described in Fig. 2. Other balance parameters were listed in Table 4.  Correlation between improvements in upper-extremity spasticity and postural balance control. In the static study with eyes-closed state, the improvement in spasticity of the finger flexor was related to a change in the mean X-speed and mean Y-speed. There was statistically significant positive correlation between the improvement in the MAS score of the finger flexor and decreased mean X-speed (r = 0.603, p = 0.029). Additionally, there was a positive correlation between the change in MTS (R1) of the finger flexors and decreased mean Y-speed (r = 0.609, p = 0.027).
In the dynamic study, the improvement in spasticity of the wrist and finger extensors had an impact on reducing the performing time, total distance, and horizontal distance taken to complete the task. There was a

Discussion
Upper-extremity spasticity can be an obstacle during standing and walking because it aggravates hemiplegic posture and may interfere with stable balance and safe gait. Thus, changes in the mechanical properties of muscle tissue components and spasticity in the upper extremities can lead to impaired balance and gait characteristics. We determined that BTx injection in the hemiplegic spastic upper-extremity has a positive effect on postural balance and gait function in stroke patients. Compared to the recent studies, our study quantitatively proved this hypothesis using plantar pressure and postural balance analysis. We also attempted to specify and present a possible explanation for the relationship between improvement in spasticity of the upper extremities   www.nature.com/scientificreports/ and balance control with gait ability. Finally, as upper-extremity spasticity decreased, postural balance and gait function improved. Along with spatiotemporal parameters, we considered the symmetry index and symmetry ratio 35 . We found a numerical improvement in these symmetric parameters; however, we failed to determine any statistical significance. Further large-scale studies are needed to diversify the severity of spasticity to determine a clear link between these two different results. Otherwise, the mean X-speed and horizontal distance in the eyes-opened test were significantly decreased at the post-intervention and follow-up evaluations compared to the baseline. In addition, descriptive finding showed that these two parameters were higher during follow-up evaluation than post-injection. The effect of BTx injection usually lasts three months, approximately 36 . As time passes, the efficacy of BTx on upper-extremity spasticity, which may affect trunk control, will decrease. Due to the drug mechanism of BTx, the value of the balance parameters can be deteriorated in follow-up evaluation.

Parameters Pre-injection Post-injection (3 weeks later) Follow up (3 months later)
We used mean X-speed, mean Y-speed, horizontal, vertical and total distance for postural balance analysis. Decreased speed and shortened distance meant that motion swing control was improved.
The MAS and MTS are evaluating tool for spasticity in the resting state, there has a limitation on assessing changes during performance. Efforts have been made to evaluate these aspects in a previous study; in particular, there was an attempt to analyze the angle of the elbow flexors during gait 23 . Even though we had a limitation on   www.nature.com/scientificreports/ evaluating changes in spasticity with movement, we tried to resolve this problem by dynamic functional evaluation through balance and gait assessment. Improvement in postural balance after BTxA injection was clearly presented on mediolateral distance, As a result, upper extremity BTx treatment might be helpful if there is a problem with horizontal direction control when performing balance and gait evaluation in patients with upper-extremity spasticity.A few studies have already investigated this subject and have focused on upper trunk posture. Previously, Hefter et al. demonstrated that injections of BTx into the affected arm of hemiplegic patients improved abnormal lateral trunk flexion. This shift of the center of mass of the upper body toward the midline improves various gait parameters, including faster gait speed, reduced pre-swing duration of both legs, and increased step length of the non-affected leg 37 .
In our study, the improvement in postural balance was significantly correlated with finger flexor spasticity. Post-stroke patients have their own spasticity on various location of upper limb. Participants who enrolled in this study had upper-extremity spasticity on wrist and finger dominantly. Spasticity on wrist and finger showed more improvement after BTx injection compared to those of shoulder and elbow. Because of this finding, we considered the correlation between postural balance and finger flexor spasticity was clarified.
Our study could not confirm the spatio-temporal gait parameters during post-injection, changes in upperextremity spasticity were significantly correlated with spatiotemporal parameters such as DLSD, relative speed, stride length, relative stride length change, and SPD change on the affected side between pre-and post-injection. This might be due to small study population, which might have influenced the differences between the two types of analyses.
A recent systemic review suggested that instrumental and laboratory measures of gait improved after BTx injections in different muscle groups of the upper and lower extremities 24 . Gait changes were presented using various methods, including spatiotemporal, kinematics, kinetics, and electromyography. In particular, our study selected not only spatiotemporal parameters but also plantar pressure loading for gait measurements and additionally calculated gait symmetry. Furthermore, we considered a balance component which had significantly contributed to the overall walking function by further analysis of postural balance and confirmed the correlation between improvement in these parameters and improvement in upper-extremity spasticity.
This study has several limitations. First, an insole-type pressure analysis system has inherent limitations in measuring spatial parameters such as step length. Therefore, studies using other methods to examine the spatial parameters should be conducted. Second, there may be confounding factors such as age, sex, location of the stroke lesion, and functional aspects related to gait speed. Further research to investigate these confounding factors should be conducted to clarify the metrics of post-stroke gait. Third, the sample size was small. We expected that there would be a significant improvement in spatiotemporal parameters; however, the sample size was insufficient to determine the clear difference in the three points of evaluation. This finding should be supplemented by further studies.

Conclusions
Our study suggested that improvement of spasticity on hemiplegic upper extremity has great impact on improvement of gait and balance function in stroke patients. We quantitatively evaluated each parameters and revealed the clear correlation between them.

Data availability
The datasets generated during and/or analyzed during the current study are not publicly available due to including patient's personal and sensitive information, but are available from the corresponding author on reasonable request.