Relationship between muscle activity characteristics and physical therapy parameters during posterior gait assistance with knee-ankle-foot orthosis in severe acute stroke hemiplegic patients

Abstract

[Purpose] This study aimed to verify muscle activity patterns during posterior gait assistance with a knee-ankle-foot orthosis (KAFO) in patients with severe acute stroke hemiplegia and clarify its relationships with physical therapy parameters. [Participants and Methods] We measured activity in the rectus femoris, biceps femoris, tibialis anterior, and gastrocnemius muscles in 30 patients with acute stroke during KAFO posterior gait assistance and examined their muscle activity patterns using the Japan Coma Scale (JCS), Brunnstrom Recovery Stage (BRS), Berg Balance Scale (BBS), Functional Movement Screen (FMS), and Functional Independence Measure (FIM). We divided lower extremity muscle activity into first and second half of the stance phase, compared muscle activity during the first half of the stance phase and the second half of the stance phase. In addition, the relationship between muscle activity during gaiting and each parameter was analyzed. [Results] All four muscles showed significantly higher values in the first half of the stance phase than in the second half of the stance phase. Rectus femoris first half of the stance phase muscle activity showed a moderate correlation with the BRS, BBS, and FMS scores. [Conclusion] The amount of gastrocnemius muscle activity when KAFO assists walking from behind increases in the latter half of stance in healthy individuals. However, in patients with stroke, the activity was lower and deviated from the gastrocnemius muscle activity during walking in healthy individuals.

INTRODUCTION

According to the Japanese Guidelines for the Management of Stroke 2021¹⁾, a knee-ankle-foot orthosis (KAFO) is appropriate for gait training in patients with stroke hemiplegia who have insufficient muscle strength during knee extension or insufficient strength of the muscles around the hip joint (recommendation level B, evidence level low)”. Additionally, “frequent gait training is recommended to improve gait function (Recommendation level: A, evidence level: high)”. Thus, the KAFO is an effective tool that allows frequent walking exercises even in patients with severe hemiplegia and improves their walking ability.

Posterior assistance for walking is a gait practice method in which a therapist closely follows and assists the patient. A previous study on muscle activity during posterior assistance for walking using a KAFO in patients with acute stroke and severe hemiplegia reported that KAFO-assisted two-movement anterior gait (body weight is supported by one and 2 points alternately) resulted in higher muscle activity than that by three-movement gait using a cane (body weight is supported by two points at all times)²⁾. Furthermore, the vastus medialis muscle activity during the load response period is higher when using a SPEX knee joint (Advanfit Co., Ltd., Kumamoto, Japan) support for the knee joint with KAFO in patients with convalescent stroke than when using KAFO with the knee fixed³⁾. Furthermore, Murayama et al.⁴⁾ measured muscle activity during walking using a knee extension assist band with KAFO in patients recovering from stroke; they reported that knee extension increased initial contact, and the muscle activity ratio of the biceps femoris during the swing phase increased. Some studies have examined the pattern of muscle activity during walking and gait exercises using KAFO in patients during the recovery phase. Changes in muscle activity depending on the walking conditions have been investigated in patients with acute stroke and severe hemiplegia. However, to the best of our knowledge^{2, 3)}, no study has examined muscle activity patterns while walking using a KAFO for posterior gait assistance.

Several studies have found that functional mobility on the paralyzed side is closely associated with activities of daily living (ADL) and basic movement ability^{5, 6)}. In clinical practice, we see patients with the same level of functional and Brunnstrom Recovery Stage (BRS) mobility on the affected side but with differences in ADL and balance. In addition, Ito et al.⁷⁾ reported that patients with less stride- time variability had higher walking ability and speed and better Berg Balance Scale (BBS) scores. Because gait ability, basic movements, and balance are associated with ADL, gait performance is expected to be associated with basic movement and balance functions. However, patients with severe acute stroke and hemiplegia often experience disturbances of consciousness disturbance and increased brain dysfunction, which compromise the accuracy of functional measurement on the paralyzed side, the validity of the BRS, and the measurements of muscle strength by evaluating voluntary movement on the paralyzed side. Therefore, we hypothesized that muscle activity in the KAFO-assisted two-movement anterior gait with posterior assistance, which evokes muscle activity on the paralyzed side even in acute severe hemiplegia, is an effective indicator of function on the paralyzed side.

This study aimed to describe the muscle activity patterns during posterior gait assistance with a KAFO in patients with severe acute stroke and hemiplegia. Furthermore, we examined the relationship between muscle activity during walking using a KAFO in patients with severe acute stroke and hemiplegia and various physical therapy parameters to guied determining effective evaluations and interventions in physical therapy.

PARTICIPANTS AND METHODS

This study included 30 patients diagnosed with acute stroke, transported by ambulance to Sapporo Shiroishi Memorial Hospital and were undergoing posterior gait assistance with a KAFO. The selection criteria were as follows: patients who showed significant knee fracture in standing and gait exercises and difficulty in walking exercises without KAFO. This study was approved by the Ethics Committee of the Hirosaki University Graduate School of Health Sciences (approval number 2021-019). The purpose and methods of the study were explained to the participants verbally and in writing, and informed consent was obtained before conducting the study. The demographic characteristics of the participants are summarized in Table 1. Patients with a history of orthopedic surgery in the lower extremities, marked Pusher syndrome, bilateral lesions, Parkinsonism, cerebellar disease, or an unstable general condition were excluded.

Table 1. Demographic details and results of assessments (n=30).

	Mean (SD)
Age (years)	62.3 (10.4)
Sex (Male/Female)	10/20
Days since onset	12.06 (4.51)
Type (cerebral hemorrhage/stroke)	17/13
Damaged hemisphere (right/left)	13/17
JCS (1/2/3/10)	11/7/11/1
NIHSS	14.1 (6.62)
Lower limb BRS (I/Ⅱ/Ⅲ/Ⅳ/Ⅴ)	3/5/9/10/3
BBS (points)	11.4 (8.25)
FMS (points)	11.3 (5.23)
FIM (points)	55.6 (20.26)

JCS: Japan coma scale; NIHSS: national institutes of health stroke scale; BRS: Brunnstrome recovery stage; FMS: functional movement scale; FIM: functional independence measure.

Patient demographic characteristics and evaluation results, including age, sex, days since disease onset, stroke type, and damaged hemisphere, were collected from electronic medical records. The attending physician evaluated the patients using the Japan Coma Scale (JCS), lower-limb BRS, BBS, Functional Movement Scale (FMS), and Functional Independence Measure (FIM), and the scores obtained were used in this study. For the statistical analysis of the JCS scores of two or more digits, we used scores from 10 to 4 for convenience.

To evaluate gait, a KAFO with a ring lock knee joint and a Gait Solution ankle joint (GS joint) (Pacific Supply Co., Ltd., Osaka, Japan) at a hydraulic pressure of 2 was attached to the lower leg on the paralyzed side. The KAFO was aligned with the participant’s greater trochanter, 2–3 cm below the perineum, knee joint, and knee joint axis. The patients performed walking exercises with posterior assistance while wearing a KAFO and with the knee joint fixed. An experienced full-time physiotherapist trained in walking provided walking assistance under the same conditions for all patients. The minimum amount of assistance was provided to ensure that the physiotherapist would be in close contact with the patient from behind, suppressing trunk compensation and smoothly swinging out the lower limbs. The two-movement anterior gait was performed as symmetrically as possible. Measurements were taken twice on an 11-m walkway. Time was measured to fit a 6–8 s interval while walking a 5-m segment in the middle of the 11-m walkway. Under all measurement conditions, there was sufficient practice before the measurement, and the participants took sufficient breaks to prevent exhaustion.

A Gait Judge System (GJ) (Pacific Supply Co., Ltd.) was used for the measurements. The electromyography (EMG) device was placed as recommended by the Surface Electromyography for the Noninvasive Assessment of Muscle (SENIAM) Project. The muscles tested were the rectus femoris, biceps femoris, tibialis anterior, and lateral head of the gastrocnemius on the leg with the KAFO, and electrodes were placed after appropriate skin preparation. We fixed the electrodes with skin tape and monitored the EMG waveform before measurement to avoid contamination by artifacts and high-frequency noise during walking. Measurements were performed after confirming the absence of noise contamination or artifacts.

For the EMG analysis, we used data obtained from the EMG waveform, which was processed by applying a band-pass filter at 20–250 Hz to remove artifacts and high-frequency noise. The Rancho Los Amigo method, a method of classifying phases of the gait cycle¹³⁾, used synchronized animation with muscle activity and ankle joint angle changes as a reference to identify the first half of the stance phase (1–31% of the gait cycle corresponding to early ground contact to middle stance phase) and the latter half of the stance phase (32–62% of the gait cycle, from late stance to pre-swing phase). Muscle activity was analyzed at a sampling frequency of 1,000 Hz. Root mean square (RMS) values were obtained, and data for six walking cycles were calculated on the side. The EMG data were normalized by taking the average value of one gait cycle from a previous study⁸⁾. The average RMS value for the EMG data taken during the early and late stance phases was calculated to examine the magnitude of the muscle activity of individual muscles during the gait cycle. This was divided by the RMS of one gait cycle to obtain the muscle activity-RMS ratio (%EMG), used as an index of muscle activity in one gait cycle.

After checking for normality using the Shapiro–Wilk test, the Wilcoxon signed-rank test was used to compare muscle activity in the first and second halves of the stance phase. We verified the normality of the relationship between the muscle activities of the rectus femoris, biceps femoris, tibialis anterior, and gastrocnemius in the first and second halves of the stance phase of the gait cycle and various parameters of function. Subsequently, the Spearman’s rank correlation coefficient was used to analyze the relationship. Statistical software R4.0.2 was used for all statistical analysis. Statistical significance was set at p<0.05. significance.

RESULTS

The detailed results are presented in Table 2. We observed significantly higher muscle activity of the rectus femoris, biceps femoris, and tibialis anterior in the first half of the stance phase (all p<0.01). The gastrocnemius showed significantly higher activity in the first half of the stance phase (p<0.01).

Table 2. Comparison of muscle activity (%EMG) during gait (n=30).

Muscle	First half of stance phase (%)	Second half of stance phase (%)	p-value
Rectus femoris	145.15 (129.7–174.65)	88.2 (69.57–100.85)	<0.01*
Biceps femoris	105.7 (89.42–140.27)	67.0 (52.95–82.4)	<0.01*
Tibialis anterior	146.25 (111.93–177.25)	95.55 (74.77–106.15)	<0.01*
Gastrocnemius	138.7 (129.52–159.7)	89.65 (70.12–118.55)	<0.01*

Gait conditions: posterior assistance with KAFO knee joint fixed.

*Wilcoxon signed-rank test (p<0.05).

Calculated with %EMG, one gait cycle as 100% expressed in median (interquartile range).

KAFO: knee-ankle-foot orthosis.

Table 3 shows the relationships between muscle activity during gait and the various parameters of physical therapy. A moderate correlation was observed between muscle activity of the rectus femoris in the first half of the stance phase and BRS (ρ=0.54, p<0.01), BBS (ρ=0.58, p<0.01), and FMS (ρ=0.61, p<0.01). Meanwhile, a mild correlation was observed with FIM (ρ=0.34); however, the difference was not significant.

Table 3. Correlations between activities of the various muscle in the first and second half of the stance phase and various parameters.

	JCS	BRS	BBS	FMS	FIM
Ractus femoris (I)	−0.21	0.54**	0.58**	0.61**	0.34
Rectus femoris (II)	0.005	−0.16	0.008	−0.11	−0.02
Biceps femoris (I)	0.11	−0.2	−0.19	−0.21	0.006
Biceps femoris (II)	0.16	−0.02	0.08	0.07	0.09
Tibialis anterior (I)	−0.05	0.014	0.04	0.13	0.06
Tibialis anterior (II)	−0.02	0.09	−0.06	−0.05	−0.03
Gastrocnemius (I)	0.11	−0.18	−0.28	−0.22	−0.03
Gastrocnemius (II)	−0.04	−0.16	−0.19	−0.19	−0.24

Spearman rank correlation coefficient. *p<0.05, **p<0.01.

(I) First half of stance phase, (II) Second half of stance phase.

JCS: Japan Coma Scale; BRS: baroreflex sensitivity; BBS: berg balance scale; FMS: functional movement screen; FIM: functional independence measure.

DISCUSSION

We investigated lower extremity muscle activity patterns during posterior gait assistance with a KAFO as well as during walking and general physical therapy parameters in patients with acute stroke and severe hemiplegia. The also examined the relationship between lower extremity muscle activity during walking and general parameters of physical therapy. Muscle activities of the rectus femoris, biceps femoris, tibialis anterior, and gastrocnemius muscle during gait were significantly higher during the first half of the stance phase. Regarding the relationship between lower extremity muscle activity during walking and physical therapy parameters, rectus femoris activity in the first half of the stance phase was moderately correlated with the BRS, BBS, and FMS. The factors and reasons for the relationship between muscle activity patterns during walking and the physical therapy parameters are discussed below.

First, the muscle activity of the rectus femoris was significantly higher in the first half of the stance phase during posterior gait assistance with a KAFO. During normal gait, the rectus femoris shows high muscle activity from the initial ground contact to the load response period⁹⁾, and our results showed that the muscle activity during gait with KAFO and posterior gait assistance in patients with acute stroke and severe hemiplegia was similar. Murayama³⁾ found that patients with stroke in the recovery phase walking with a KAFO with the knee joint fixed and SPEX knee support with flexion restriction showed high vastus medialis activity during the load-response period, and higher activity was observed when the patient was wearing the KAFO with SPEX knee joint support. Other studies⁴⁾ have reported high values in the first half of the stance phase in recovering patients with stroke. Thus, the similar results obtained in patients with acute stroke in this study seem to show that high values are present in the first half of the stance phase during gait with KAFO and posterior assistance, regardless of the time elapsed from stroke onset.

The biceps femoris of stroke patients showed significantly higher levels of muscle activity in the first half of the stance phase during posterior gait assistance with a KAFO. In a normal gait⁹⁾, the biceps femoris muscle works as a hip joint extensor, as it prevents flexion of the hip joint and trunk from initial contact to the load response period. In patients with a stroke, significantly higher value in the first half of the stance phase is attributed to trunk dysfunction, which is associated with damage to the corticoreticular tract damage^{10, 11)}. In posterior assistance, a caregiver assists the trunk from behind; however, damage to the corticoreticular tract can cause more swinging of the trunk during gait, requiring the hamstrings to work more effectively in braking the trunk and hip flexion in the first half of the stance phase.

Tibialis anterior muscle activity in stroke patients was also significantly higher in the first half of the stance phase while walking with posterior gait assistance and KAFO. In normal gait, the tibialis anterior muscle acts on the forward tilt of the lower leg from initial contact to the load response period⁹⁾, with the same muscle activity pattern as that of a normal gait. The GS joint used as the ankle joint for the KAFO in this study supports the function of heel locking by maintaining ankle dorsiflexion using a hydraulic damper¹²⁾. The knee is fixed in the extended position during assisted walking using the KAFO, encouraging heel contact; the ground reaction force passes through the posterior ankle joint, and the tibialis anterior muscle activity is promoted as an internal moment.

During posterior gait assistance with a KAFO in stroke patients, the activity of the lateral head of the gastrocnemius was significantly higher during the first half of the stance phase. Stroke patients develop hemiplegia because of damage to the lateral corticospinal tract, controlling the distal muscles of the extremities. This indicates that flaccid paralysis, which corresponding to reduced sensitivity of muscle spindles, is often present at the onset¹³⁾. From the middle to the late stance phase, strong activation of the gastrocnemius muscle is necessary to achieve lower limb support, and the applied load is stored as elastic energy in the gastrocnemius tendon, generating high muscle activity¹⁴⁾. Therefore, damage to the lateral corticospinal tract may prevent the gastrocnemius muscle from bearing the load and storing sufficient elastic energy, leading to reduced muscle activity in the second half of the stance phase in patients with severe acute stroke hemiplegia.

A moderate correlation was observed between the rectus femoris activity in the first half of the stance phase and BRS. The BRS was scored on an ordinal scale form I to VI, with higher scores indicating greater leg voluntarines. Mochizuki¹⁵⁾ noted a relationship between the BRS and gait ability, and knee extension strength on the paralyzed side has also been associated with gait ability⁶⁾. Therefore, it is possible that BRS and the rectus femoris, a knee extensor muscle, are both effective indicators of gait ability, suggesting that BRS may be a good predictor of rectus femoris activity during walking.

A moderate correlation was observed between rectus femoris activity in the first half of the stance phase during walking and two physiotherapy indicators, the BBS and FMS. The relationship between muscle strength on the paralyzed side and basic functional movements such as walking ability, walking speed, and standing movement speed, has been well documented^{5, 6)}. In addition, leg extension muscle strength is reported to be more closely associated with walking ability and speed than knee extension muscle strength^{16, 17)}. Finally, a relationship between the lower extremity loading force and the ability to stand, maintain standing posture, and walk has been reported in patients with stroke and hemiplegia⁵⁾.

Collectively, the findings of these studies suggest that basic movement and walking ability are more closely associated with the overall ability to extend and support the lower limbs, such as leg extension muscle strength and lower limb loading force, than with simple knee extension muscle strength alone. The BBS, one of the physiotherapy parameters evaluated, requires the participant to perform many tasks in a standing position, such as standing with closed legs, rotating once, or transferring. In addition, the FMS has many tasks in the standing position, such as basic movements, transfers, and reaching the standing position, all of which require strong lower limb support, which was attributed to the moderate correlation.

The results of this study may be effective in several clinical applications. First, muscle activity patterns in patients with acute stroke and severe hemiplegia show high values in the first half of the stance phase in the gastrocnemius muscle during posterior gait assistance with a KAFO and may deviate from muscle activity patterns during normal gait. Thus, continued training with a muscle activity deviating from the normal gait pattern may lead to mislearning and require caution during the intervention. Second, physiotherapy parameters, such as BRS, may provide indicators for estimating muscle activity during posterior gait assistance using a KAFO in patients with acute stroke and severe hemiplegia and may be useful as a reference for practicing muscle activity patterns similar to normal gait.

This study had several limitations. First, we measured muscle activity during gait with posterior assistance, which may result in variability owing to differences in the person assisting. Walking speed and style were specified as much as possible; however, it was not possible to completely remove bias. Second, although this was a cross-sectional study examining the relationship between muscle activity during walking and the physiotherapy indices, changes in muscle activity over time were untracked while walking with posterior assistance; thus, it is unknown how the relationships with the parameters evolved. Third, In the present study, cases were selected according to the presence or absence of knee folding during gait; however, the severity of motor paralysis varied widely, from BRS I to V, and a more limited case selection may yield different results for muscle activity during gait.

In conclusion, this study investigated the patterns of lower extremity muscle activity during posterior gait assistance with KAFO in stroke patients, as well as the physical therapy parameters related to muscle activity during gait in stroke patients. In patients with stroke, all muscles activities were significantly higher during the first half of the stance phase. Concerning the relationship between muscle activity and physical therapy parameters, a moderate correlation was observed between the muscle activity of the rectus femoris in the first half of the stance phase and the BRS, BBS, and FMS. Our findings further suggest that the gastrocnemius can deviate from the normal gait pattern during walking with posterior assistance in patients with acute stroke or severe hemiplegia.

Conflict of interest

The authors declares that there is no conflict of interest.