Increased trailing limb angle in hemiplegic patients after training with a knee orthosis: A randomized controlled trial

Abstract

BACKGROUND:

Stroke often induces gait abnormality, such as buckling knee pattern, compromising walking ability. Previous studies indicated that an adequate trailing limb angle (TLA) is critical for recovering walking ability.

OBJECTIVE:

We hypothesized that correcting gait abnormality by immobilizing the knee joint using a knee orthosis (KO) would improve walking patterns and increase the TLA, and investigated whether walking training using a KO would increase the TLA in post-stroke patients.

METHODS:

In a randomized controlled trial, thirty-four participants were assigned to KO (walking training using a KO) and non-KO (without using a KO) groups. Twenty-nine completed the three-week gait training protocol. TLA was measured at baseline and after training. A two-way repeated ANOVA was performed to evaluate TLA increases with training type and time as test factors. A t-test compared TLA changes (ΔTLA) between the two groups.

RESULTS:

ANOVA showed a main effect for time (F = 64.5, p < 0.01) and interaction (F = 15.4, p < 0.01). ΔTLA was significantly higher in the KO group (14.6±5.8) than in the non-KO group (5.0±7.0, p < 0.001).

CONCLUSION:

Walking training using a KO may be practical and effective for increasing TLA in post-stroke patients.

Keywords

Trailing limb angle gait training gait abnormality Knee brace walking pattern

1 Introduction

In 2019, the global prevalence of stroke was estimated to be 101.5 million (Virani et al., 2021). Stroke is the second leading cause of death and a major cause of disability in the world (Kuriakose et al., 2020). Impaired mobility is a major contributor to this disability, with over 60% of stroke survivors able to walk independently but at slow speeds that are sufficient for full participation in the community (< 9.8 m/s) (Pamela et al., 2007). It has been reported that reduced walking speed is negatively associated with quality of life (Pamela et al., 2007) and life prognosis, as well as higher risks of falling (Jack et al., 2000; Callisaya et al., 2012; Oh et al. 2022; Studenski et al., 2011), hospitalization (Montero et al., 2005), increased care needs (Shimada et al., 2015), and onset of dementia (Teresa et al., 2010; Minghui et al., 2017; Scott et al., 2002). Therefore, walking speed can be considered as a ‘sixth vital sign’ (Fritz et al., 2009). This is why the goal of post-stroke walking rehabilitation is typically improvement of walking speed.

Trailing limb angle (TLA) is defined as the angle between the laboratory’s vertical axis and the vector from the fifth metatarsal joint to the great trochanter (Tyrell et al., 2011). TLA is shown to be related to walking speed (Tyrell et al., 2011) and the ability to walk long distances (Louis et al., 2015). A healthy individual’s walking pattern is called “inverted pendulum” (IP) walking pattern. In the IP walking pattern, the hip joint fully extends from the midstance to the late stance. Thus, the construction of the IP walking pattern is essential to induce a large TLA.

However, post-stroke patients often exhibit characteristic abnormal walking patterns, such as the buckling knee pattern (excessive knee flexion on the paretic side during the stance phase) and the stiff knee pattern (sustained knee flexion on the paretic side during the swing phase) (De Quervain et al., 1996; Mulroy et al., 2003). In addition to these knee abnormalities, limited hip extension movement is often observed in stroke patients with abnormal walking patterns who can walk only slowly, compared to those who can walk faster (De Quervain et al., 2003); this suggests that abnormal walking patterns would cause difficulty in hip extension. Considering the above, we hypothesized that walking training with a knee orthosis (KO) would help to regain adequate knee extension and thereby correct gait abnormality, finally resulting in an increase the TLA. To confirm this hypothesis, we measured and compared the TLA of post-stroke patients before and after a three-week walking training with and without a KO.

2 Materials and methods

2.1 Trial design

This was an open-label randomized controlled trial. All procedures were approved by the institutional ethics committee of Kohnan Hospital and were consistent with the Declaration of Helsinki. Written informed consent was obtained from all patients prior to their participation in this study. The study was registered in the UMIN Clinical Trial Registry (UMIN ID: 000045939).

2.2 Participants

We included patients with acute stroke admitted to Kohnan Hospital between October 2021 and April 2022. Inclusion criteria were: the patient has lower limb paresis (Brunnstrom Recovery Stage (BRS) ≤ IV); the patient can walk three meters or more without assistance of another individual; and the patient has full ankle dorsiflexion, which prevents hip extension during the late stance. Exclusion criteria were: the patient has severe vision loss, visual field disorder, vertigo, and/or a history of other neurological, respiratory, circulatory, or orthopedic diseases; the patient has a severe impairment of consciousness and cognitive function and higher brain dysfunction (unilateral spatial neglect, attention disorder, aphasia, mental disorder) that affect the walking test; and the walking ability had already been impaired (i.e. Functional Ambulation Category < 4) before the stroke onset. In this study, we included all the stroke patients who met the above criteria without gait abnormality rating because it is quite common that post-stroke patients have lower limb paresis.

2.3 Interventions

Participants were randomly assigned to one of two training groups: (1) walking training using a knee orthosis (KO group) and (2) walking training without using a KO (non-KO group). The KO used is a Straight Position Knee-Joint Immobilizer (ALCARE, Tokyo, Japan). If the KO group participants showed an abnormal walking pattern, such as the buckling knee or stiff knee patterns, they wore the Knee Ankle Foot Orthosis (KAFO: Pacific Supply, Osaka, Japan) instead of a KO. In this study, no participants, therapists, or evaluators were blinded since the KO worn by the participants was visible to anybody during the measurement of the TLA. In both groups, walking training was performed at a comfortable walking speed for each participant.

The KO group performed 30 minutes per day of walking training that reproduced the IP walking pattern using a KO (Fig. 1). The non-KO group also completed 30 minutes per day of walking training without using a KO.

Fig. 1

Correction of the walking pattern in the knee joint by wearing a knee orthosis. A: Walking with the abnormal walking pattern (the knee joint bending during the stance phase). B: Walking with the corrected abnormal walking patterns with knee orthoses.

In addition to the above walking training, both groups received 30 minutes per day of individualized physical therapy, including balance re-education, strengthening exercises, stair climbing, activity of daily living exercises. Therefore, total physical therapy time was 60 minutes per day, except when physical therapy was not possible due to fever or other health problems. The training period was three weeks, and the training frequency was at least five times per week.

In the KO group, we carefully assisted the patients in reproducing the IP walking pattern. If patients exhibited inadequate hip extension on the paretic side (hip extension did not reach 0 degrees during the late stance phase), therapists provided external assistance for hip extension. Also, if patients could not create sufficient hip flexion on the paretic side during the swing phase, therapists provided external assistance for hip flexion. Some patients in the KO group exhibited a difficulty in swinging the leg properly due to the knee joint being extended during the swing phase. In contrast, in the non-KO group, walking training prioritized walking longer without external assistance to correct knee abnormality, regardless of whether the patients reproduced an IP walking pattern.

When patients in either group exhibited low foot clearance, these patients wore an ankle foot orthosis with an oil damper hinge (AFO-OD) (Pacific Supply, Osaka, Japan). This AFO-OD has an oil damper hinge for the ankle joint. The oil damper hinge resists plantarflexion during the swing phase, and as a result, prevents plantarflexion in the swing phase. Similarly, the hinge allows the foot to engage in plantarflexion after heel contact. Therefore, the loading response is constructed correctly because the first rocker function is maintained (Yamamoto et al., 2015).

2.4 Outcomes and estimate

We determined the TLA as the primary outcome measure to evaluate the effects of the use of a KO in walking training. We measured the TLA at baseline (initial assessment) and after three weeks of walking training (final assessment) and compared changes in the TLA (ΔTLA) between the two groups after the three-week training. We also compared the TLA measured before wearing the KO (pre-training measurement), immediately after wearing it and starting the training (mid-training measurement), and after performing some training and achieving the IP walking pattern (post-training measurement) on the first day of training to evaluate the immediate effects of wearing a KO during the walking training.

The secondary outcome measure was walking speed. We measured the walking speed of the patients who were asked to walk 10 meters (or 3 meters if they have difficulty in walking 10 meters) at their comfortable and maximum speeds. The walking speed measurement was performed during the initial and final assessments, as well as after one week and two weeks of training. The measurement of the TLA and walking speed was performed five times at each assessment timing and the mean of five measurements were used for comparison. The flow of TLA and gait speed measurement are shown in Fig. 2.

Fig. 2

Flow of TLA and gait speed measurement.

2.5 Measurement of the TLA

An eight-camera motion analysis system (Locus 3D MA-3000: ANIMA, Tokyo, Japan) was used to measure the TLA during walking at a comfortable walking speed. During this evaluation, participants were allowed to use the parallel bars as handrails to prevent falls. The TLA data were recorded with eight markers placed at the greater trochanter, lateral knee joint line, lateral malleolus, and fifth metatarsal head on both legs (Hsiao et al., 2015), using a six-camera passive motion capture system that detects the motion of the reflective markers at 100 Hz, and then analyzed using the Locus 3D MA-3000. The TLA was defined as the angle between the line connecting the fifth metatarsal head marker and the greater trochanter marker in late stance, viewed from the sagittal plane and the vertical line (Hsiao et al., 2015). Data from the paralyzed stance phase of five gait cycles were sampled, and we used these mean values. The Z-axis is the vertical axis, and the X-axis is the direction of travel. The TLA was calculated when the continuous movement of the Z-axis and X-axis values of the lateral malleolus marker on the unaffected side reached a minimum near initial ground contact in the late affected side stance phase.

2.6 Sample size

To our knowledge, no similar previous study exists, which made it difficult to estimate an adequate sample size with evidence. We could not determine the sample size by calculation either because we could not conduct preliminary research. The sample size was limited by time constraints. Patients who met the inclusion criteria were recruited during the study period.

2.7 Randomization

The randomization sequence was computer-generated with a ratio of 1:1 assigned to two training groups by a scientific staff member who was not directly involved in the assessment and treatment to ensure concealment.

2.8 Blinding

In this study, neither the patients nor the therapists could not be blinded as everybody could notice the KO worn by the patients. The evaluators could not be blinded either because they had to assess the immediate effect during the training.

2.9 Statistical analysis

The demographic data were summarized using descriptive statistics. For comparison of data between the two groups, an unpaired t-test or Mann–Whitney U-test was used after confirming their normality. Furthermore, the χ2 test and Fisher’s exact test were used to compare the basic characteristics and nominal data. To check immediate effects of wearing the KO, changes in the TLA (pre-, mid-, and post-training assessments) were analyzed with a one-way repeated analysis of variance (one-way ANOVA) or Friedman test after confirming the normality. Improvement in the TLA was analyzed with a two-way repeated measures ANOVA (two-way ANOVA) with training type (KO or non-KO) and time (initial or final assessment) as the test factors. If significant differences were detected by the two-way ANOVA, we added a post-hoc test. In addition, a comparison of the ΔTLA between the two groups was performed using the t-test or Mann–Whitney U test after confirming normality.

A one-way ANOVA was also used to analyze the improvement of walking speed as four time factors (during the initial and final assessments and after one week and two weeks of training).

The alpha level was set at 0.05 for all analyses with EZR Ver. 1.55 (Saitama Medical Center, Jichi Medical University) (Kanda, 2013), which is a graphical user interface for R (the R Foundation for Statistical Computing, version 1.55).

2.10 Data availability

Datasets generated and analyzed during the current study are available in an anonymized form from the corresponding author on reasonable request.

3 Results

We recruited 34 patients who were undergoing stroke inpatient rehabilitation and met the inclusion criteria. Of these, 29 were found eligible for intervention (Fig. 3). By random allocation, both KO and non-KO groups had 17 participants. Eventually, 13 participants of the KO group and 16 participants of the non-KO group completed a three-week walking training program.

Fig. 3

Research Flowchart. *At the time of the TLA evaluation, the patient was unable to walk safely. Therefore, it was canceled.

Table 1 shows the demographic data. All participants had moderate to mild lower extremity paresis corresponding to IV and V on the BRS. The total days of intervention and the total time of physical therapy were significantly lower in the KO group than in the non-KO group (p < 0.01). However, other demographic data showed little difference between the KO and non-KO groups.

Table 1

Demographic data

	KO group (n = 13)	non-KO group (n = 16)	p Value
Age^*‡	64.2±14.1	67.9±12.9	0.48
Sex (male/female)^§	11/2	13/3	1
Stroke Type (hemorrhagic/ischemic)^∥	4/9	8/8	0.46
The total days of intervention^*‡	14.6±1.8	17.0±1.7	0.001
The total time of physical therapy (minutes)^*‡	877.0±108.2	1012.0±101.4	0.002
The number of days from onset to initial assessment^*#	8.2±7.0	10.1±9.0	0.49
The number of days from onset to final assessment^*#	29.8±30.0	30.1±31.0	0.82
The side of hemiparesis (right/left) ^∥	6/7	9/7	1
BRS of lower limb^#	IV: 6 V: 7	IV: 4 V: 12	0.31
SIAS of lower limb (hip joint)^#	5: 1 4: 5 3: 5 2: 2	5: 0 4: 9 3: 7 2: 0	0.46
SIAS of lower limb (knee joint)^#	5: 1 4: 4 3: 6 2: 2	5: 1 4: 7 3: 6 2: 2	0.57
SIAS of lower limb (ankle joint)^#	5: 1 4: 5 3: 2 2: 5	5: 0 4: 7 3: 4 2: 5	1
Number of persons with a history of stroke^§	1	1	1
TCT ^†	5	5	–
FAC^†	5	5	–

*: Mean value±Standard deviation. †: No analysis because the two groups have the same value. ‡: t-test. ^∥: χ2 test. #: Mann–Whitney U test. §: Fisher’s exact test. BRS: Brunnstrom Recovery Stage. SIAS: Stroke Impairment Assessment Set. TCT: Trunk Control Test. FAC: Functional Ambulation Categories.

The flow and results of the one-way ANOVA for the TLA are shown in Figs. 4 A and B. There was a significant main effect of time (F = 15.53, p < 0.01) in the KO group. Post-hoc pairwise Bonferroni-adjusted comparisons showed a significant difference in the TLA between pre-training vs. mid-training (p < 0.01) as well as between pre-training vs. post-training (p < 0.01). Conversely, no immediate effect was observed in the non-KO group.

Fig. 4

An immediate change in the TLA. A: An immediate change in the TLA was observed in the KO group. The one-way ANOVA for the TLA showed a significant main effect of time (F = 15.53, p < 0.01). Post-hoc pairwise Bonferroni-adjusted comparisons showed a significant difference in the TLA between pre vs. during (p < 0.01), pre vs. post (p < 0.01). B: No immediate change in the TLA was observed in the non-KO group. The one-way ANOVA for the TLA not showed a significant main effect (F = 0.17, p = 0.84).

In Fig. 5, the results of two-way ANOVA show a main effect for time (F = 64.5, p < 0.01) and significant interaction between time and training type (F = 15.4, p < 0.01). A comparison of ΔTLA between the two groups using a t-test showed that ΔTLA was significantly greater in the KO group than in the non-KO group (KO group: 14.6±5.8 vs non-KO group: 5.0±7.0, p < 0.001).

Fig. 5

Changes in the TLA between the two groups during the three-week intervention period. Results of two-way ANOVA showed a main effect for time (F = 64.5, p < 0.01) and significant interaction between time and group (F = 15.4, p < 0.01).

In Figs. 6 A and B, the results of two-way ANOVA show a main effect for time (F = 33.6, p < 0.01 F = 43.9, p < 0.01) at both comfortable and maximum walking speeds. There were no differences in the change of walking speed between the KO and non-KO groups.

Fig. 6

Changes in comfortable and maximum walking speed between the two groups. A: Changes in comfortable walking speed between the two groups. Results of two-way ANOVA showed a main effect for time (F = 33.6, p < 0.01). B: Changes in maximum walking speed between the two groups. Results of two-way ANOVA showed a main effect for time (F = 43.9, p < 0.01).

4 Discussion

The effects of gait training with a KO on the increase in the TLA and walking speed in post-stroke patients was investigated in a randomized controlled trial. The results showed that in the KO group, a significant increase in the TLA was observed both immediately after the start of the training and after three weeks of training, compared to the pre-training level. The TLA also increased in the non-KO group after three weeks of training, compared to the pre-training level, but ΔTLA was significantly greater in the KO group than in the non-KO group.

There was no difference between the KO and non-KO groups regarding clinical characteristics at admission. Although a significant decrease in total intervention days and physical therapy time was observed in the KO group compared to the non-KO group, the KO group achieved a greater ΔTLA. Surprisingly, even in the participants who could walk unassisted, an immediate increase in TLA was observed during the initial assessment on the first day of training with a KO. This suggests that many patients capable of walking without assistance may have insufficient hip joint extension during the late stance phase. In other words, ambulatory stroke patients with mild to moderate lower limb paresis may be able to achieve an adequate TLA through walking training with a KO. We suspect that the difference in TLA between the KO and non-KO groups was attributed to the correction of abnormal walking patterns in the knee joint. Adequate knee extension induced the formation of the IP walking pattern, leading to adequate hip joint extension. This walking pattern was acquired through repetitive walking training with a KO, eventually allowing for walking without it. Therefore, walking training using a KO might be enormously effective for increasing the TLA in post-stroke patients. Previous studies have shown that walking speed is related to propulsive force, defined as the anterior component of the ground reaction force during walking (Bowden et al., 2006), and the main factor for its increase is the TLA (Hsiao et al., 2016). The IP walking pattern, characterizing full hip extension during the mid- to late stance, is crucial for efficient and stable walking (Arthur et al., 2010). Using a KO during training sessions can correct abnormal knee patterns, reproduce the IP walking pattern, and increase TLA. If the KO does not provide such correction, unusual sensations may be continually experienced during walking. It has been reported that the abnormal sensory input sometimes elicits abnormal motor output following neurologic injury owing to disrupted neural organization (David et al., 2004). Thus, abnormal feedback can lead to incorrect motor learning. Conversely, wearing a KO that focuses on achieving proper hip joint extension provides optimal sensory feedback, correcting abnormal movements. Consistent practice of a natural walking pattern simultaneously helps avoid incorrect motor learning and supports normal motor learning. Consequently, in the present study, the increased TLA in the KO group suggests that utilizing a KO, coupled with repetitive walking training, effectively normalizes knee joint patterns in post-stroke patients.

Walking speed improved in both groups, but without significant difference. It is known that the progress of improvement differs for walking speed and for biomechanical variables in post-stroke patients. Darcy et al. conducted a 12-week walking training in 13 post-stroke patients and evaluated the progress of improvement in biomechanical variables and walking speed every four weeks. As a result, biomechanical variables, including the TLA, improved from pre-training to four weeks, but not from 4 to 12 weeks. On the other hand, walking speed improved both from pre-training to four weeks and from 4 to 12 weeks (Darcy et al., 2013). Therefore, walking speed increases further even after the biomechanical improvement has ceased if walking training is continued. If the intervention of this study had continued after three weeks, the KO group, having the increased TLA by then, might have been able to significantly improve walking speed after four weeks. Conversely, three weeks may have been too short for the training to produce a definite change in walking speed in acute stroke hemiplegic patients. We could not evaluate changes in walking speed after four weeks because our intervention period was three weeks due to the limited duration of hospital stay.

Our study has some limitations. First, this study was conducted at a single facility. Therefore, external validity was not examined. Second, this randomized controlled trial included 34 participants, of whom only 29 completed the training. This may not have been a sufficient sample size. Also, the effect of using a KO for patients with moderate to severe paresis was not examined because this study included only patients capable of walking without external assistance. Third, we did not investigate the biomechanical background of the TLA increase.

Further studies are needed, such as those using a force plate to obtain data on propulsive and other reaction forces, or those using electromyography to investigate changes in each muscle activity in the lower extremity. Finally, we could not examine long-term effects of increasing the TLA and what positive and unpredictable effects on other functions and abilities of the patient may result from the increased TLA in a long run. The impact of increasing the TLA on walking speed and other walking abilities should be further investigated, with a view to its potential effects on activities of daily living, social participation, quality of life, prevention of recurrent strokes and other diseases, and life expectancy in the long term.

Correcting walking patterns using a KO is highly effective to increase the TLA. Increasing the TLA improves walking speed and ability to walk long distances (Hsiao et al., 2016). Repeating the normal walking pattern with a KO might create the normal sensory feedback, which in turn promotes normal motor learning. Therefore, walking training using a KO should be considered when the knee joint is not adequately controlled in the early stage of post-stroke rehabilitation.

5 Conclusions

We performed a randomized controlled trial to investigate whether walking training with a KO improves the TLA in hemiplegic stroke patients, compared to training without a KO. As a result, the KO group increased the TLA more than the non-KO group. A KO may also be effective to achieve an adequate hip joint extension, leading to the correction of abnormal knee patterns and enabling the reproduction of a normal walking pattern. Repeated training of the normal walking pattern with a KO might create normal sensory feedback, which in turn promotes normal motor learning. Our findings suggest that walking training with a KO is effective for increasing the TLA and learning a normal walking pattern, and therefore should be considered when the knee joint is not adequately controlled in early post-stroke rehabilitation.

Footnotes

Acknowledgments

The authors are grateful to the individuals who volunteered to participate.

Funding

This research received no external funding.

Conflict of interest

The authors have no conflicts of interests to disclose.

Ethics statement

All procedures were approved by the institutional ethics committee of Kohnan Hospital and were consistent with the Declaration of Helsinki. Written informed consent was obtained from all patients prior to their participation in this study.

Author contributions

Conceptualization, SI and HA; methodology, SI and HA; data analysis, SI, KO and HA, data collection and subjects recruitment, SI, TO, KN and TN; writing—original draft preparation, SI, HA; writing—review and editing, HA; visualization, SI and HA; supervision, HA, YS and S-II. All authors have read and agreed to the final version of the manuscript.

References

Arthur,

, Donelan

(2010). Dynamic principles of gait and their clinical implications. Phys Ther., 90, 157–174.

Bowden,

M. G.

, Chitralakshmi,

, Richard,

, Steven,

(2006). Anterior-posterior ground reaction forces as a measure of paretic leg contribution in hemiparetic walking. Stroke, 37, 872–876.

Callisaya,

M. L.

, Blizzard,

, McGinley,

J. L.

, Srikanth,

V. K.

(2012). Risk of falls in older people during fast-walking–the TASCOG study. Gait Posture, 36, 510–515.

Darcy,

, Trisha,

, Ramu,

, Margaret,

, Katherine,

, Jill,

, Erin,

, Stuart,

(2013). Time course of functional and biomechanical improvements during a gait training intervention in persons with chronic stroke. J Neurol Phys Ther, 37, 159–165.

David,

, Jeremy,

, Steven,

(2004). Robotics, motor learnig, and neurologic recovery. Annual Review of Biomedical Engineering, 6, 497–525.

De Quervain,

I. A.

, Simon,

S. R.

, Leurgans,

, Pease,

, McAllister,

(1996). Gait pattern in the early recovery period after stroke. J Bone Joint Surg Am, 78, 1506–1514.

Fritz,

, Lusardi,

(2009). Walking speed: The sixth vital sign. Journal of Geriatric Physical Therapy, 32, 46–49.

Hsiao,

, Brian,

, Jill,

, Stuart,

(2015). The relative contribution of ankle moment and trailing limb angle to propulsive force during gait. Hum Mov Sci, 39, 212–221.

Hsiao,

, Brian,

, Ryan,

, Jill,

, Stuart,

(2016). Mechanisms used to increase peak propulsive force following 12-weeks of gait training in individuals poststroke. J Biomech., 49, 388–395.

10.

Jack,

, Luigi,

, Carl,

, Suzanne,

, Kyriakos,

, Glenn,

, Stephanie,

, Lisa,

, Robert,

(2000). Lower extremity function and subsequent disability: consistency across studies, predictive models, and value of gait speed alone compared with the short physical performance battery. J Gerontol A Biol Sci Med Sci, 55, 221–231.

11.

Kanda,

(2013). Investigation of the freely available easy-to-use software ‘EZR’ for medical statistics. Bone Marrow Transplant, 48, 452–458.

12.

Virani,

S. S.

, Alonso,

, Aparicio,

H. J.

, Benjamin,

E. J.

, Bittencourt,

M. S.

, Callaway,

C. W.

, Carson,

A. P.

, Chamberlain,

A. M.

, Cheng,

, Delling,

F. N.

, Elkind,

M. S. V.

, Evenson,

K. R.

, Ferguson,

J. F.

, Gupta,

D. K.

, Khan,

S. S.

, Kissela,

B. M.

, Knutson,

K. L.

, Lee,

C. D.

, Lewis,

T. T.

, Liu,

, Loop,

M. S.

, Lutsey,

P. L.

, Ma,

, Mackey,

, Martin,

S. S.

, Matchar,

D. B.

, Mussolino,

M. E.

, Navaneethan,

S. D.

, Perak,

A. M.

, Roth,

G. A.

, Samad,

, Satou,

G. M.

, Schroeder,

E. B.

, Shah,

S. H.

, Shay,

C. M.

, Stokes,

, VanWagner,

L. B.

, Wang,

N. Y.

, Tsao,

C. W.

(2021). Heart Disease and Stroke Statistics-2021 Update: A Report from the American Heart Association. Circulation, 143, e254–e743.

13.

Kuriakose,

, Xiao,

(2020). Pathophysiology and Treatment of Stroke: Present Status and Future Perspectives. International Journal of Molecular Sciences, 21, 7609.

14.

Louis,

, Stuart,

, Ryan,

, Darcy,

(2015). Paretic propulsion and trailing limb angle are key determinants of long-distance walking function after stroke. Neurorehabil Neural Repair, 29, 499–508.

15.

Minghui,

, Pengcheng,

, Cheng,

, Yiyu,

, Ru,

, Peijie,

(2017). Walking Pace and the Risk of Cognitive Decline and Dementia in Elderly Populations: A Meta-analysis of Prospective Cohort Studies. J Gerontol A Biol Sci Med Sci, 72, 266–270.

16.

Montero,

, Marcelo,

, Enrique,

, Miguel,

, Roberto,

, Luis,

, Marcelo,

(2005). Gait velocity as a single predictor of adverse events in healthy seniors aged 75 years and older. J Gerontol A Biol Sci Med Sci, 60, 1304–1309.

17.

Mulroy,

, Gronley,

, Weiss,

, Newsam,

, Perry,

(2003). Use of cluster analysis for gait pattern classification of patients in the early and late recovery phases following stroke. Gait Posture, 18, 114–125.

18.

Oh,

, Im,

, Lee,

, Lim,

, Cho,

, Ryu,

, Yoon,

(2022). Effect of Antigravity Treadmill Gait Training on Gait Function and Fall Risk in Stroke Patients. Ann Rehabil Med, 46, 114–121.

19.

Pamela,

, Katherine,

, Andrea,

, Stanley,

, Samuel,

, Stephen,

, Bruce,

, Dorian,

, Julie,

(2007). Protocol for the Locomotor Experience Applied Post-stroke (LEAPS) trial: A randomized controlled trial. BMC Neurology, 7, 39.

20.

Pamela,

, Stephanie,

, Sue,

, Lorie,

, Subashan,

, Samuel,

(2007). Improvements in speed-based gait classifications are meaningful. Stroke, 38, 2096–2100.

21.

Scott,

, Milar,

, Diane,

, Gary,

, Haydeh,

, Jeffrey,

, Richard,

(2002). Independent predictors of cognitive decline in healthy elderly persons. Arch Neurol, 59, 601–606.

22.

Shimada,

Makizako,

Doi,

, Tsutsumimoto,

, Suzuki,

, (2015). Incidence of disability in frail older persons with or without slow walking speed. J Am Med Dir Assoc, 16, 690–696.

23.

Studenski,

, Subashan,

, Kushang,

(2011). Gait speed and survival in older adults. JAMA, 305, 50–58.

24.

Teresa,

, Hiroko,

, Diane,

, Dara,

, Jeffrey,

(2010). The trajectory of gait speed preceding MCI. Arch Neurol, 67, 980–986.

25.

Tyrell,

C. M.

, Roos,

M. A.

, Rudolph,

K. S.

, Reisman,

D. S.

(2011). Influence of systematic increases in treadmill walking speed on gait kinematics after stroke. Phys Ther, 91, 392–403.

26.

Yamamoto,

, Fuchi,

, Yasui,

(2015). Immediate-term effects of use of an ankle-foot orthosis with an oil damper on the gait of stroke patients when walking without the device. Prosthetics and Orthotics International, 39, 140–149.

27.

Yamamoto,

, Tomokiyo,

, Yasui,

, Kawaguchi,

(2013). Effects of plantar flexion resistive moment generated by an ankle-foot orthosis with an oil damper on the gait of stroke patients: A pilot study. Prosthetics and Orthotics International, 37, 212–221.