Review article 1

Effects of walking trainings on walking function among stroke

survivors: a systematic review
Daudet Ilunga Tshiswakaa, Crystal Bennettb and Cheyanne Franklinb

Physical function is often compromised as a result of stroke reported significant improvements in physical function
event. Although interventions propose different strategies that compared with overground walking training augmented by
seek to improve stroke survivors’ physical function, a need auditory stimulations. The imperative to improve physical
remains to evaluate walking training studies aimed at function among stroke survivors with physical impairment is
improving such physical function. The aim of this review was paramount, as it allows survivors to be socially, emotionally,
to assess the available literature that highlights the impact of and physically more independent. In general, we note an
walking training on enhancing walking for stroke survivors. insufficiency of research on the interaction between physical
We performed a systematic literature review of online function and socialization among stroke survivors.
databases – Google Scholar, PubMed, CINHAL, Cochrane International Journal of Rehabilitation Research 00:000–000
Library, Scopus, and EBSCO – with the following inclusion Copyright © 2017 Wolters Kluwer Health, Inc. All rights
criteria: manuscript published from 2005 to 2016, written in reserved.
English, with treatment and control groups, for walking International Journal of Rehabilitation Research 2017, 00:000–000
training studies aimed at improving physical function among
stroke survivors. Findings indicated that walking speed, Keywords: disability, intervention, physical function, stroke, walking

walking distance, and gait speed were the most used a

Department of Public Health and bSchool of Nursing, University of West Florida,
outcome variables for measuring improved physical function Pensacola, Florida, USA

among stroke survivors. Importantly, proposed interventions Correspondence to Daudet Ilunga Tshiswaka, PhD, Department of Public Health,
involved either overground or treadmill walking trainings, if University of West Florida, Pensacola, FL 32514, USA
Tel: + 1 850 474 2782; fax: + 1 850 474 2173; e-mail: daudeti@uwf.edu
not both. Preserved locomotor improvements were not noted
in all interventions at follow-up. Some interventions that used Received 8 June 2017 Accepted 6 August 2017
walking treadmill training augmented by auditory stimulations

Introduction et al., 2014). Walking is a low to moderate aerobic exer-

Stroke is the fourth leading cause of death and the cise that can be performed with or without equipment,
leading cause of long-term disability among adults in the alone or with a group, and at one’s own pace, making it an
USA (Centers for Disease Control and Prevention, 2016; ideal exercise for stroke survivors to perform. Treadmill
Mozaffarian et al., 2016). Stroke causes physical impair- walking offers different intensities and has handrails that
ments such as extremity weakness, fatigue, and postural may aid the individual’s performance. Although level-
instability that lead to walking difficulty and physical ground walking does not require equipment, and can be
inactivity (American Stroke Association, 2012; Jeon et al., easily incorporated into daily tasks, whether these types
2015). In addition, stroke can cause memory loss, lan- of walking training similarly improve physical function in
guage/speech deficits, and aphasia (American Stroke stroke survivors needs to be examined. The aim of this
Association, 2012). The economic burden of stroke in the review was to assess the available literature that high-
USA is costly, with the mean lifetime cost reaching lights the impact of walking training on enhancing
~ $140 048, thus making it a top 10 most costly condition walking for stroke survivors.
(Johnson et al., 2016).
Participation in exercise or physical activity is essential Materials and methods
after stroke to prevent secondary complications. Exercise The initial search yielded 1437 articles retrieved from
after stroke can improve muscle strength, gait, balance, online databases such as PubMed, Google scholar,
and may prevent cardiovascular complications (Billinger EBSCO, and Scopus.
et al., 2014; Van Duijnhoven et al., 2016). Therefore, it is
recommended that stroke survivors begin exercise as Databases
soon as possible after stroke (Billinger et al., 2014). The following databases were searched to obtain data for
Depending on functional ability, it is recommended to this review: Google Scholar, PubMed, Scopus, Cochrane
exercise at least 3 days/week for 20–60 min or in a series Library, CINAHL, and EBSCO. The search was per-
of 10–15 min of moderate-intensity exercise (Billinger formed in September and October 2016. The following
0342-5282 Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. DOI: 10.1097/MRR.0000000000000250

Copyright r 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 2 International Journal of Rehabilitation Research 2017, Vol 00 No 00

terms were searched consecutively: ‘walking’ and ‘stroke’ studies found that locomotor training improved daily
as Medical Subject Headings (MeSH) that yielded (1437 stepping activity and gait efficiency among poststroke
articles), and then the terms ‘effects’ (all fields) (476) and individuals, others highlighted the impact of aerobic
‘exercise therapy’ (all fields) (151) were added con- treadmill and resistance trainings on cardiovascular fit-
secutively to the search. ness. In one case, balance improved by a virtual reality-
based stepping exercise; another augmented treadmill
Inclusion and exclusion criteria and ground walking trainings with rhythmic auditory
We performed a detailed search using filters to limit stimulation. Treadmill walking trainings were also cou-
search to randomized control trials published between pled with obstacle-crossing to improve ambulatory habits.
2005 and 2016 and written in English. Articles were
included if they fulfilled these criteria and were walking Randomizations
training or exercise studies aimed at improving physical Although randomization reduces the risk of selection bias
function among stroke survivors. In general, physical and eliminates the source of bias in treatment assignments
function includes activities such as walking and climbing (Suresh, 2011), only some of these interventions were
stairs (Tomey and Sowers, 2009). Systematic reviews, randomized. Among randomized studies, some used either
meta-analysis, and qualitative studies were excluded control groups exclusively or control groups but no pretest
from the review. Out of 151 articles, 29 were reviewed (Harris et al., 2006); some were single-blind interventions,
and included in our results. whereas others were cross-over randomized.

Data analysis Participants

A systematic review of the literature was performed. The Participants were assigned randomly to control or inter-
review followed the Preferred Reporting Items for vention groups. Thesex of the participants was mostly
Systematic Review and Meta-Analysis (PRISMA) state- male, with some mixed groups, with ages ranging from
ment. Data were collected on the following study char- the early 20s to the late 80s and averages in the 60s. Most
acteristics: publication date, hypotheses or research of the inclusion criteria were individuals who had at least
question, study design, sample size, inclusion criteria, 6 months poststroke (i.e. either ischemic or hemorrhagic)
demographics of sample, type of stroke, type of walking and were able to walk independently on the ground for
training, effects of walking on outcomes, stated implica- the 10-m walk test with or without a walking aid at a
tions, and recommendations. Two independent review- walking speed of less than equal to 1.4 m/s. In addition,
ers screened all potential articles and the articles were participants who could walk ~ 5 min at a self-selected
excluded if they did not address the research question. pace on the treadmill were included. Other studies
included participants who had 1 month, 12 months, or
Results 5 years poststroke. Similarly, individuals with unstable
Using PRISMA guidelines (Fig. 1) and a flow diagram hemodynamic states, who were unable to walk inde-
(Table 1), we gathered relevant information from pub- pendently, and had cognitive impairment, aphasia
lished literature. After screening 1437 articles by title and (inability to follow at least two-step commands), or
abstract for inclusion, we selected 29 articles that fulfilled depression were excluded from most interventions. The
the eligibility criteria for the current review. Of these, Mini-Mental State Examination cutoff value of less than
only one study was a mixed-method study; the rest were 23 was used as an exclusion criterion. Participants with
pilot studies, case studies, cross-sectional studies, or ret- recurrent stroke episodes were also excluded in most of
rospective studies. the studies.

Outcome variables
Intervention types
Studies had different outcome variables, including the
Two major types of interventions aimed at improving
6-min timed walk or the 10-min timed walk, the Short
stroke survivors’ physical functioning: using a treadmill
Physical Performance Battery, the Berg Balance Scale,
and over-the-ground walking trainings (Table 1).
the Stroke Impact Scale, the Barthel Index, and the Fugl
Although most rehabilitation programs were clinical
Meyer Assessment, which measured the impact of
interventions or clinical trials performed in hospital or
walking training on walking function.
university settings, some involved community-based
interventions. Most intervention protocols included par-
Effects of walking training on walking
ticipating in 30- or 60-min sessions of moderate exercises
Overall, most findings suggested that walking training
ranging from 3 to 5 times weekly for 6–16 weeks. Some
improved walking for stroke survivors. Specifically, dif-
other interventions lasted for 40 weeks over 12 months
ferent training mechanisms (i.e. overground, treadmill,
with 3 sessions/week (Dean et al., 2012).
and conventional physiotherapy rehabilitation) high-
Gait training was used as a proxy for improved physical lighted the improvement in walking function among after
function and locomotion in general. Although some stroke individuals (Olawale et al., 2011). In particular,

Copyright r 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Walking trainings in stroke survivors Tshiswaka et al. 3

Fig. 1

Identification
Records identified through database Additional records identified
searching through other sources
(n =1,437) (n = 9)

Records after duplicates removed

(n = 961)
Screening

Records excluded
(n = 325)
Papers written in other
Records screened languages than English or
(n = 476) included measure of other
physical functions such as
grasp.

Full-text articles assessed

Eligibility

Full-text articles excluded,

for eligibility with reasons
(n = 151) (n =121)
Some papers were reports
or protocols. Others were
testing the reliability of
Studies included in specific instrument
Included

qualitative synthesis
(n = 29)

Chart of selected articles.

some interventions – consisting of 30-min walking Participating in walking training not only improved
trainings overground or body weight-supported treadmill walking functions but also significantly improved emo-
– indicated significant improvements (Ada et al., 2013; tional well-being among some participants (Stuart et al.,
Combs-Miller et al., 2014). 2009).
There seems, however, to be divergent results associated Gait training through treadmill, especially with biofeed-
with the use of overground compared with body weight- back, not only improved gait and walking function but
supported treadmill in walking trainings. For instance, also gait cycle length, duration of gait phases, and swing
when comparing the impact of overground versus body phase speed (Drużbicki et al., 2015). Similarly, Sousa et al.
weight-supported treadmill trainings, overground train- (2011) reported that gait training performed overground
ing showed a more significant impact on improving also significantly improved walking function. High-
physical function (Combs-Miller et al., 2014). Conversely, intensity aerobic training significantly improved gait
Ada et al. (2013) reported that treadmill training per- and peak oxygen consumption rate alike (Mackay-Lyons,
formed over 4 months showed significant walking 2012).
improvements compared with overground training. Mayo Significantly, although most interventions yielded
et al. (2013) found no significant effect on physical improvements in physical functioning among poststroke
function between two groups of participants (i.e. one individuals, many studies reported that the maintenance
group exercised on a stationary bicycle and the other of these improvements faded after a certain period of
group performed brisk walking). In addition, effective time (Ada et al., 2013; Severinsen et al., 2014).
walking training did not prevent falls among stroke sur-
vivors (Dean et al., 2012).
Discussion
Walking speed and walking distance increased sig- Physically impaired stroke survivors require interventions
nificantly for a participant who simultaneously performed that enhance their walking function and lead to an
task-specific and strength training programs after stroke improvement of their overall quality of life and inde-
(Sullivan et al., 2006; Ada et al., 2013). Likewise, daily pendence. Nonetheless, mixed findings for the use of
stepping improved among patients who participated in conventional, overground, and treadmill walking train-
intensive locomotor training (Moore et al., 2010). ings among stroke survivors with physical impairments

Copyright r 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Copyright r 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

4 International Journal of Rehabilitation Research

Table 1 Methodological characteristics of intervention studies focusing on walking trainings
Study design/data Sample size, inclusion criteria, Implications and
References Purpose of study source demographics, type of stroke Type of intervention Results recommendations

Salbach et al. To evaluate the efficacy of a Secondary analysis of a Ninety-one individuals with a Three times/week for 6 weeks The walking intervention was Task-oriented walking
(2005) task-oriented walking two-center, observer residual walking deficit within targeting walking or upper associated with a significant retraining enhances balance
intervention in improving blinded, randomized, 1 year of first or recurrent stroke. extremity function. Participants greater average proportional self-efficacy in community-
balance self-efficacy in controlled trial Inclusion: clinical diagnosis of asked to participate in 18 change in balance self-efficacy dwelling individuals with
persons with stroke and to first or recurrent stroke; residual training sessions given than the UE intervention chronic stroke. Benefits may
determine whether effects walking deficit; mental 3 times/week for 6 weeks in a be partially the result of
were task-specific, competency evaluated using hospital setting. Walking improvement in walking
influenced by baseline level Telephone Version of the Mini- intervention: progressive capacity.
of self-efficacy and Mental State Examination; ability program of 10 tasks: walking
associated with changes in to walk 10 m independently on a treadmill; standing up,
walking and balance walking.
capacity
Sullivan et al. To describe outcome Case study A 38-year-old women with right Body weight-supported treadmill Post-treatment: walking speed For the person in this case
(2006) associated with a therapy middle cerebral artery stroke, training and a limb-loaded increased 18% for free clinically meaningful
program that combines evaluated 15 months after onset. cycling exercise were (0.59 m/s) and 14.4% for fast changes in walking function
task-specific and strength Ambulated independently with an alternated over 24 treatment velocity (0.71 m/s); 6 min were associated with a
training in an individual ankle-foot orthosis and straight sessions (four 1-h treatment walking distance increased 4% combined therapeutic
after stroke and to discuss cane. sessions) (4 times/week for (184.4 m). At 6 months, program that included both

2017, Vol 00 No 00
some possible 6 weeks) continued improvements in all task-specific and lower
mechanisms and walking outcomes were extremity strength training
modulating factors that evident.
may affect poststroke
neurological recovery and
responsiveness to
intervention
Krishnamoorthy To describe the application of Case study The participant was a 58-year-old Underwent gait retraining over a At the end of the 5 weeks, he The case report describes
et al. (2008) a novel gait retraining man who had a stroke more than period of 3 days /week for decreased the time required to possible advantages of
approach to an individual 3 years before the intervention 5 weeks for a total of 15 complete the Timed Up and Go judiciously combining
with poststroke sessions during which the test; his gait speed increased different treatment
hemiparesis gravity compensation provided during overground walking. techniques in improving the
by the gravity-balanced Except for gait symmetry, all gait of chronic stroke
orthosis and visual feedback other improvements were survivors
about walking performance maintained after 1 month after
was gradually reduced. Each intervention
session lasted approximately
two hours.
Smith et al. To explore the secondary Modified random Twenty individuals participated in Twelve 20 min sessions of No significant difference was A task-specific intervention
(2008) benefits of treadmill assignment. Matched- this study. Participants matched walking on a treadmill or found between groups for any designed to improve gait
training for people in the pair control group by side hemiparesis and motor weekly phone call. Control dependent measure. The speed may potentially
chronic stage of recovery design with repeated impairment. Control group: group: received weekly phone ANOVA to investigate main provide secondary benefits
from stroke measures. N = 10. Treatment group: N = 10. calls from the examiner effects in each group found no by positively impacting
Inclusion: Individuals who had an enquiring about the quality of significant findings in control depression, mobility, and
ischemic stroke more than their week and encouraging group; however, in the social participation for
3 months but less than two years them to record any life events treatment group significant people poststroke
before enrolling in the study. in the log. Treatment group: improvements over time for
received identical questions depression, mobility, and social
about the quality of their week participation were
and encouragement to use the demonstrated.
QoL.
Pang et al. (2008) To identify the determinants Secondary analysis of Sixty-three community-dwelling Participants in each group Gain in paretic leg muscle Enhancement of
of improvement in walking data obtained from a individuals (mean age = 65 years, underwent three 1-h exercise strength and peak oxygen cardiorespiratory fitness and
capacity following prospective, single- age range = 50–87 years) with a sessions/week for 19 weeks consumption remained paretic leg muscle strength
blind, randomized chronic stroke (poststroke independently associated with are both significant
Copyright r 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

therapeutic exercise in controlled intervention duration: mean-5.5 years, gain in walking capacity. Two- determinants in improving
chronic stroke survivors trial range = 1–28 years). Inclusion: a way ANOVA indicated that the walking capacity among
single episode of stroke more leg group had significantly more chronic stroke survivors.
than one year ago, aged 50 years gain in walking capacity However, the weak
or more, ability to walk more than (P = 0.01), peak VO (P = 0.01) relationship indicates that
10 m independently, living at and paretic leg muscle strength other factors not measured in
home, a mental score > 22, the (P = 0.01) than the arm group. this study may also contribute
ability to pedal the cycle to the improvement in walking
ergometer at 60 rpm and capacity.
increase the heart rate to at least
60% of the maximal heart rate.
Moore et al. To suggest that reduced task- Randomized cross-over Participants had hemiparesis Twenty patients with chronic Stepping practice was more than Intensive LT results in improved
(2009) specific walking practice study > 6 months duration who were stroke completed a repeated four-fold higher during LT daily stepping in individuals’
during clinical physical attending PT after unilateral baseline measures, versus clinical PT sessions, with poststroke who have been
training contributes to supratentorial stroke were randomized cross-over trial in significant improvements in discharged from PT because
limited gains in ambulatory recruited. Inclusion: walk > 10 m which walking performance daily stepping and gait of a perceived plateau in
function in those with a overground without physical was assessed during the last efficiency only after LT. Change motor function. These
perceived plateau after assistance at speeds ≤ 0.9 m/s at 4 weeks of clinical PT before in daily stepping after clinical improvements may be related
stroke, and suggest that their self-selected velocity, discharge secondary to PT and intensive LT were to the amount and intensity of
further gains can be required medical clearance, a reaching a plateau, followed correlated with the amount of stepping practice.
realized if intensive primary stated goal to improve by 4 weeks of intensive LT and stepping practice received
stepping, or locomotor walking ability, and enrolled 4 weeks of no intervention during these interventions.
training is provided after ~ 1 month before termination of
discharge PT services secondary to
decreased gain in function, as
stated by the treating clinical
therapist
Stuart et al. To determine whether Nonrandomized Participants with mild to moderate APA intervention included Intervention groups improved APA-stroke appears to be safe,
(2009) Adaptive Physical Activity controlled study hemiparesis at least 9 months walking, strength, and balance whereas controls declined in feasible, and efficacious in a
(APA-stroke), a community- after stroke. Forty participants training for 1 h, thrice a week gait velocity, balance, SPPB, community setting
based exercise program for completed APA-stroke. Thirty- in local gym. Outcomes SPPB, and SIS social
participants with eight participants completed the measured: 6-month change in participation domains. Between
hemiparetic stroke, usual care. Inclusion: have no gait velocity (6-min timed group comparisons were
improves walking function comorbid conditions that were walk), Short physical significantly at P < 0.00015.

Walking trainings in stroke survivors Tshiswaka et al. 5

in the community contraindication to participation, Performance Battery, Berg Individuals with depressive
aphasia with inability to follow Balance Scale, Stroke Impact symptoms at baseline improved
two-step commands, Scale, Barthel Index, Hamilton whereas controls were
symptomatic heart failure, Rating Scale for Depression unchanged (P < 0.003).
unstable angina, and
hypertension (diastolic BP
≥ 95 mm Hg; systolic BP
≥ 160 mm Hg), which would
preclude participation in exercise.
Combs et al. To examine changes in Prospective pretest/ Only 16 participants completed the BWSTT was provided for 24 Statistically significant The findings of this study
(2010) balance, balance post-test pilot study study. Inclusion: onset of stroke sessions over 8 weeks with improvements were found from suggest that effects of
confidence, and health- with 6-month retention at least 6 months before the 20 min of total walking each pretest to post-test for Berg BWSTT may transfer beyond
related QoL immediately study, aged 40–80 years, living in session Balance Scale, Activities- gait to positively influence
and 6 months after body the community, ambulatory with Specific Balance Confidence, balance, balance confidence,
weight-supported treadmill or without the use of an assistive Stroke Impact Scale mobility, and health-related quality of
training (BWSTT) device or ankle-foot orthosis, able SIS stroke recovery, and life. However, for most
to ambulate at a self-selected comfortable 10-min walk test participants, BWSTT was not
comfortable speed of 0.4–0.8 m/s (CWT) scores (P < 0.05) and sufficient to induce
determined by using the 10-m from pretest to retention on improvements in balance and
walk test, physician approval to BBS, ABC, CWT, and FWT balance confidence beyond
enter the exercise program, and scores (P < 0.05) measurement error or long-
able to follow at least two-step term retention of enhanced
verbal instructions. perceptions of QoL.
Copyright r 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

6 International Journal of Rehabilitation Research

Table 1 (continued)

Study design/data Sample size, inclusion criteria, Implications and

References Purpose of study source demographics, type of stroke Type of intervention Results recommendations

Sousa et al. Investigate the effects of gait Pilot study Twelve individuals with chronic BWS training during overground After gait training, individuals Gait training individuals with
(2011) training on ground level stroke (53.17 ± 7.52 years) walking (10 m) and measures walked faster (F1, 11 = 8.384, chronic stroke with BWS
with partial body weight participated in a gait training the amount of weight borne by P = 0.015), with symmetrical during overground walking
support in individuals with program with BWS during the system steps, longer and faster strides, improved walking in terms of
stroke during overground overground walking, and were and increased toe-clearance. temporal-spatial parameters
walking with no BWS evaluated before and after the Also, they displayed increased and segmental angles. The
gait training period. Inclusion: an rotation of foot, shank, thigh, training strategy might be
elapsed time no longer than one and trunk segmental angles on adopted as a safe, specific
year since stroke and the ability to both sides of the body. and promising strategy for
walk ~ 10 m with or without gait rehabilitation after
assistance stroke.
Olawale et al. To evaluate the effects of Prospective, randomized Outpatient stroke rehabilitation unit All participants received TWET (A): recorded 22.6 ± 1.5% Treadmill and overground
(2011) treadmill walking and controlled trial in tertiary hospital. Sixty patients individual outpatient decrease in 10 MWT and walking exercise training
overground walking with chronic stroke (≥3 months conventional physiotherapy 31.0 ± 4.3% increase in programs, combined with
exercise training on not > 24 months). Group A: rehabilitation for 12 weeks. 6 MWT. OWET (B): made conventional rehabilitation,
recovery of walking Treadmill Walking Exercise Participants in group A 26.8 ± 1.3% and 45.2 ± 4.6% improved walking function in

2017, Vol 00 No 00
function in an African Training (N = 20) (age: received treadmill walking improvement in 10 MWT and African group of adult stroke
group of stroke survivors 56.8 ± 6.4). Group B: exercise training. Participants 6 MWT, respectively. CG: survivors
Overground Walking Exercise in group B received made 2.2 ± 0.7% and
Training (N = 20) (age: overground walking exercise 2.9 ± 0.8% improvement in the
56.8 ± 8.3). Group C: (control) training. Participants in group two functions. Changes were
(N = 20) (57.2 ± 5.9). Able to walk C received conventional significant for the TWET and
10 m independently with or physiotherapy rehabilitation OWET (P < 0.05).
without a walking aid. only.
Billinger et al. Examine whether an 8-week Prospective study Ten patients were enrolled and nine Aerobic exercise training session At baseline, we identified Aerobic exercise in
(2012) aerobic exercise completed the study. Aged 3 times/week for 8 weeks. between-limb differences in participants with subacute
intervention would improve 61.2 ± 4.7 years old, were Preexercise vital signs, ten brachial artery flow-mediated stroke was beneficial for
cardiovascular health and 66.7 ± 41.5 days’ poststroke, and stretching exercises, 5-min dilation (FMD) and low VO2 improving cardiovascular
physical performance in had minor motor performance warm up, intensity increased peak values. After the health, reducing
participants with subacute deficits (Fugul-Meyer score, to prescribed workload intervention, significant cardiovascular risk, and
stroke 100.3 ± 29.3). Inclusion: between (started at lower end of improvements were observed improving physical
50 and 70 years of age, diagnosis targeted HR range then in FMD in both arms, resting performance (6 MWT)
of first time, unilateral stroke that increased to the upper portion systolic blood pressure, and the
occurred less than 6 months of THRR). Duration was 6 MWT. Although we also
before enrollment, ability to walk increased but did not exceed observed improvements in the
with or without an assistive 40 min and intensity did not resting diastolic pressure, heart
device and need only stand-by exceed THRR but was rate, and VO2 peak values,
assist, and ability to travel for all adjusted according to these changes were not
testing and exercise sessions. physiological response. significantly different.
Mackay-Lyons Does high-intensity aerobic Randomized controlled Thirty-eight poststroke individuals A total of 18 people was in the After 3-month training period, A high-intensity treadmill
(2012) exercise improve trial > 60 years Inclusion: with usual care. Intervention group change in peak oxygen training program improves
cardiovascular fitness and residual gait impairment, and (20 people) underwent consumption rate was cardiovascular fitness and
gait function in people with ability to walk on the treadmill at treadmill training (3 times/ significantly more in the gait in older adults with
chronic stroke? ≥ 0.3 km/h for 3 min were eligible week) for 3 months. Program treatment group, by 6.3 ml/kg/ chronic stroke
intended to achieve min 6 MWT. Change in
30–50 min of treadmill training distance was significantly more
at 60–80% of the maximum in treatment group.
heart rate reserve as
determined by a maximum
effort exercise test.
Copyright r 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

Dean et al. (2012) Investigate whether an Randomized trial Experimental group program Both groups, on average, Twelve months, EG: walked 34m The experimental intervention
exercise intervention can (n = 76) aimed to improve 5.9 years poststroke, further in 6 min and 0.07 m/s delivered through stroke
enhance mobility, prevent walking, prevent falls and received exercise classes faster over 10 m than the CG. clubs enhanced aspects of
falls, and increase physical increase physical activity. The (40 weeks over 12 months; 3 EG: 129 falls. CG:133 falls. mobility but had no effect on
activity among community- control group program (n = 75) times/week), advice, and a There were no differences in falls
dwelling people after aimed to improve upper-limb home program for proportion of fallers (relative
stroke and cognitive functions. Age, 12 months. Each class and risk = 1.22; 95%
mean (SD; range): EG: 66.7 home program designed to CI = 0.91–1.62; P = 0.19) or
(14.3; 31–91). CG: 67.5 (10.2; take 45–60 min EG received the rate of falls between
40–85). Inclusion: suffered an exercise intervention groups (incidence rate
one or more strokes, were able designed to enhance ratio = 0.96; 95%
to walk 10 m independently mobility, prevent falls, and CI = 0.59–1.51; P = 0.88).
with or without a mobility aid, increase physical activity.
gained medical clearance.
Mayo et al. (2013) To estimate the relative An observer blinded, People (ages 67.7 ± 14.4) within Two dose-equivalent No significant effects of group or Both programs were equally
effectiveness in improving randomized, 12 months of acute stroke who interventions, one involving time were found for the effective in maintaining
walking ability and other pragmatic, trial with were able to walk > 10 m stationary cycling and the 6 MWT. Significant effects for walking capacity after
mobility and health repeated measures independently and healthy other disability-targeted role participation was found in discharge from stroke
outcomes after stroke of enough to engage in exercise. interventions were tested. favor of the exercise group rehabilitation; or were
two home-based exercise Total of 87 individuals (cycle Both protocols required (global odds ratio for cycling equally ineffective in
programs – stationary group, n = 43; exercise group, daily moderate-intensity vs. exercise was 0.51; 95% improving walking capacity
cycling and an exercise n = 44). exercise at home building up CI: 0.27–0.95).
and walking outcomes to 30 min/day. One group
exercised on a stationary
bicycle, the second group
carried out mobility exercises
and brisk walking.
Mackay-Lyons To compare the effectiveness Single site, randomized Fifty individuals (mean age, All individuals participated in BWSTT improved VO2 peak by BWSTT elicits greater
et al. (2013) of body weight-supported controlled trial 60 ± 14 years; mean event-to- 60-min physiotherapy 30% which was significantly improvements in
treadmill to dose-equivalent randomization, 23 ± 5 days; 29 sessions five times weekly as greater than the 8% cardiovascular fitness and
usual cause in improving men) participated. Inclusion: inpatients for 6 weeks and improvement observed for walking endurance than UC
cardiovascular fitness and enrolled men and women who three times weekly as UC (P = 0.004 between in the subacute poststroke
walking early after stroke were older than 18 years, within outpatients for another groups) period. These gains are
1 month of first ischemic stroke, 6 weeks largely sustained for 1 year.
inpatients in the stroke

Walking trainings in stroke survivors Tshiswaka et al. 7

rehabilitation unit, and able to
walk 5 m with or without use of
ambulatory aids, ankle orthoses,
or stand-by assistance.
Ada et al. (2013) To determine if a 4-month Prospective randomized A total of 102 people living with Experimental group 1 (EG1): EG1 and EG2: Improved 6-min Four months of treadmill
treadmill and overground trial. Ambulate trial stroke in community. 30 min of treadmill and walk distance compared with training results in better
walking program is more Experiment 1: n = 34 (mean overground walking thrice/ control group. EG1: Improved walking. However, these
effective than a 2-month age: 70). Experiment 2: N = 34 week for four months. further than EG2; walking effects disappear once
program, compared with (mean age: 64) Control group: Experiment 2 (EG2): training 38 m (95% confidence interval training ceases. Therefore,
control, at improving N = 34 (mean age: 63) for 2 months. Control group: 15–60) more than the control training should be ongoing.
walking in community- Inclusion: were within 5 years of no intervention. group and 29 m (95% CI:
dwelling people with stroke their first stroke, Mini-Mental 4–53) more than EG2. After
who walk slowly State Exam > 23, were 12 months: Both experimental
community-dwelling and walked groups returned to near
slowly (walk 10 m across flat baseline levels.
ground bare feet without any
aids taking more than 9 s).
Copyright r 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

8 International Journal of Rehabilitation Research

Table 1 (continued)

Study design/data Sample size, inclusion criteria, Implications and

References Purpose of study source demographics, type of stroke Type of intervention Results recommendations

Combs-Miller To compare the effects of Single-blind, pilot Convenience sample of participants Thirty-min walking interventions Overground walking training Overground walking training
et al. (2014) body weight-supported randomized (N = 20) at least 6-month (body weight-supported group: significant greater was more beneficial than
treadmill training and controlled trial with poststroke (ischemic or treadmill training or overground improvements in comfortable body weight-supported
overground walking 3-month follow-up. hemorrhagic) and able to walk walking training) were walking speed compared with treadmill training at i
training when matched independently were recruited. administered five times week body weight-supported mproving self-selected
for Between the ages of 21 and for 2 weeks. Intensity was treadmill group immediately walking speed for the
task and dose (duration/ 80 years, community dwelling, monitored with the Borg Rating (change of 0.11 vs. 0.06 m/s, participants in this study
frequency/intensity) on able to walk with or without an of Perceived Exertion Scale at P = 0.047) and 3 months
improving walking assistive device or orthosis at a 5-min increments to maintain a (change of 0.14 vs. 0.08 m/s,
function, self-selected gait speed of moderate training intensity. P = 0.029) after training.
activity, and participation <1.0 m/s over 10 m, medically OGWTG: improved
after stroke stable, and able to follow two- comfortable walking speed
step verbal instructions. (P = 0.001).
Severinsen et al. Compare the effects of Randomized controlled Fourty-three community-dwelling Participants performed exercise Comparison between AT, RT, and Improvements in muscle
(2014) aerobic training with 12-week intervention independent walkers with chronic 3 times/week for 12 weeks. sham training showed no strength or aerobic capacity
progressive resistance trial with 1-year follow- stroke. Inclusion: Individually adapted high- clinically relevant effects on using non-task-specific
training after stroke to up Nonhemorrhagic stroke; intensity AT consisted of 15 min walking velocity or walking training methods does not
determine whether AT- 6–36 months have elapsed after of strenuous cycle ergometer distance. Muscle strength result in improved

2017, Vol 00 No 00
induced fitness gains or stroke; aged 50–80 years; exercise, three times at each improved after RT (P < 0.0001) ambulation in patients with
RT-induced fitness gains muscle strength of more than session. If heart rate scales were and was preserved at 1-year chronic ischemic stroke.
translate into improved three on the Medical Research unreliable Borg Scale was used. follow-up (P < 0.001). Aerobic Muscle strength gains were
ambulation across Council scale at the paretic lower High-intensity progressive RT of capacity increased after AT maintained at follow-up,
12-weeks intervention and limb; and walking velocity of less both lower limbs consisted of (P < 0.001) but was lost during whereas all improvements of
whether gains are retained than 1.4 m/s at fast 10 MWT, three sets of eight repetitions the follow-up observation aerobic capacity were lost,
1-year after cessation of allowing assistive walking targeted at an intensity of 80% period. indicating a long-lasting
formal training devices. of one repetition maximum. Self- effect of intensive RT even
assessments of physical activity without maintenance
were estimated monthly using training.
Danish version of the Physical
Activity Scale.
Park et al. (2015) To investigate the effect of Cross-sectional study Thirty patients who had suffered a Each participant was subjected Gait velocity under vibratory These results indicate that the
external cues using stroke (men = 18 women = 12). to six walking trials: three trials stimulation of the tibialis application of external cues
vibratory stimulation on Inclusion: 6 months since the were vibratory stimulation of anterior muscle using vibratory stimulation
spatiotemporal gait diagnosis, ability to walk for a the tibialis anterior muscle and (50.22 ± 17.42 cm/s) was during gait may control gait
parameters in patients with minimum of 10 m, no problem in three trials without stimulation. significantly higher than that of parameters and improve gait
chronic stroke gait by contracture of the foot general gait (46.59 ± 15.09 m/ performance. Intervention
and ankle, and no s; P < 0.05). Gait cadence could be used for gait
musculoskeletal disease. under vibratory stimulation of rehabilitation in chronic
the tibialis anterior muscle stroke patients.
(79.26 ± 15.76 step/min) was
significantly higher than that of
general gait (76.32 ± 13.13
step/min; P < 0.05).
Srivastava et al. To ascertain whether A prospective, repeated- Forty participants. Inclusion: Twenty sessions of task-specific At the beginning, the end of Rehabilitation interventions
(2015) rehabilitation interventions measure study patients with first ever interventions consisting of training, and follow-up, the significantly improve
improve locomotion supratentorial stroke, with an age lower limb resistive exercises mean Scandinavian Stroke locomotor outcome even in
beyond 6-months range of 16–65 years, with more and treadmill gait training to Scale scores were 0.41, 0.53, the chronic phase following
poststroke than 6 months of poststroke locomotor abilities (90 min/ and 0.51; and Barthel Index a stroke
duration, with an ability to follow day, 5 days/week for 4 weeks). scores were 77.34, 89.06, and
three-step commands, and Evaluations were performed at 92.32.
having an impaired balance and the beginning and end of
gait with an intact ability to walk training and at follow-up of
with or without support. 3 months.
Copyright r 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

Park et al. (2015) To compare the gait abilities Pilot study Nineteen patients: TRAS: N = 9 Both the TRAS group and After the training period, the Treadmill walking training
in chronic stroke patients ORAS: N = 10 Inclusion: patients ORAS training group received TRAS group showed a significant during rhythmic auditory
following either treadmill who had a stroke more than neurodevelopment therapy and improvement in walking speed, stimulation may be useful for
walking training with rhythmic 6 months after the onset of stroke then performed walking training step cycle, step length of the rehabilitation of patients with
auditory stimulation or and less than 2 years, the ability to for 30 min Five times a week for unaffected limb, coefficient of chronic stroke
overground walking training walk for 10 min or longer on a 3 weeks. 10 m distance on the variation, 6 MWD, and FGA when
with rhythmic auditory treadmill, no amblyopia, vertigo, and ground and let patients walk at a compared with the ORAS group.
stimulation (ORAS) abnormal vestibular function, convenient speed. Tempo of After the training results showed
cognitive function allowing an rhythmic auditory stimulation was that the TRAS group’s walking
understanding of researchers’ increased from 90% (first week) speed statistically significantly
instructions, and after 6 min of to 100% (second week) to increased compared with the
walking, no blood pressure, pulse, 110% (third week). TRAS: ORAS group (P < 0.05).
or breathing problems. metronome was used for
auditory rhythmic stimulation.
Before walking training, patients
were asked to beat foot along
with metronome to adapt to the
rhythm for 2 min.
Druzbicki et al. To evaluate the effects of gait A single-blind, Fifty people [(18 women, 32 men; Group 1 (control) received Intervention group: greater Gait training using a treadmill
(2015) training using a treadmill with randomized, control mean age 62 years (range treadmill training without visual improvements in the shortening of resulted in improvements in the
or without visual biofeedback group investigation, 38–79 years)] at least 6 months biofeedback and basic stance phase (P = 0.0045) and gait and functional capacity of
in patients in the late period conducted among after stroke, independent walking physiotherapy consisting of lengthening of the swing phase of patients. The use of
after stroke, and to compare patients treated at the (walking speed > 0.4 m/s), 1.5-h sessions for 10 days (every the unaffected limb (P = 0.0042) biofeedback gives better
both training methods Clinical Rehabilitation Brunnstrom recovery stage 3–4, Monday to Friday for 2 weeks.) and increase in unaffected limb results in improving gait cycle
Unit. muscle tone of a paretic lower limb Group 2 (intervention group) cycle length (P = 0.0021). No length, duration of gait phases
(Ashworth ≤ 1 plus), and a level of received treadmill training with significant differences between and swing phase speed
disability according to the Rankin visual biofeedback and basic groups in other spatiotemporal compared with exercise on a
Scale: 3. Randomly enrolled into physiotherapy consisting of parameters of gait or additionally treadmill alone.
groups with a rehabilitation program 1.5-h sessions for 10 days assessed parameters.
of treadmill training with or without (Monday–Friday) for 2 weeks.
visual biofeedback. Intervention
group: n = 25. Control group:
n = 25.

Walking trainings in stroke survivors Tshiswaka et al. 9

Llorens et al. To study clinical effectiveness Randomized controlled Twenty individuals with chronic Intervention consisted of 20 1-h Group-by-time interaction in the Virtual reality interventions can
(2015) and the usability of a virtual trial stroke. Experimental group: n = 10 sessions, 5 sessions/week for scores of the BBS (P < 0.05) and be an effective resource to
reality-based intervention Control group: n = 10. Inclusion: 4 weeks. Experimental group: in the 10 MWT (P < 0.05) enhance the improvement of
compared with conventional hemiparesis, age ≥ 40 years old and combined 30 min with the virtual balance in individuals with
physical therapy in the ≤ 70 years old; chronicity reality-based intervention with chronic stroke
balance recovery of > 6 months; Mini-Mental State 30 min of conventional training.
individuals with chronic stroke Examination cutoff > 23; able to Control Group: underwent 1 h
follow instructions; ability to conventional therapy.
maintain stride-standing position for
30 s without assistance.
Bang et al. (2016) To investigate the effects of Randomized controlled Patients were randomly assigned to BAT: received body awareness After intervention, both groups The results of this study
body awareness training on trial a body awareness training group training for 20 min, followed by showed significant differences suggest that body awareness
balance and walking ability in (N = 6) and a control group (N = 6). walking training for 30 min a day, compared with before the training has a positive effect on
chronic stroke patients Inclusion: hemiparesis stroke (first 5 days a week for 4 weeks. intervention in the BBS balance in patients with
hemorrhage or infarction), event Control group: received walking (P = 0.036) and TUG (P = 0.037) chronic stroke
occurring > 6 months previously training for 30 min a day, 5 days between the two groups/no
and ability to walk a distance of 100 a week for 4 weeks. significant intervention in the
meters with or without assistance. 10 MWT
Copyright r 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

10 International Journal of Rehabilitation Research

Table 1 (continued)

Study design/data Sample size, inclusion criteria, Implications and

References Purpose of study source demographics, type of stroke Type of intervention Results recommendations

Boyne et al. To assess the feasibility and Feasibility study Nineteen participants. HIT group: Both groups trained 25 min, Baseline VO2 peak and fractional Although further protocol
(2016) justification for a definitive preliminary RCT was N = 13, MCT group: N = 5. 3 times/week, for 4 weeks. utilization were significantly optimization is needed to
randomized controlled trial conducted Inclusion: (i) age 35–90 years, HIT strategy involved 30-s different between groups improve overground
comparing high-intensity (ii) unilateral stroke experienced bursts at maximum-tolerated translation of treadmill gains,
interval training and > 6 months before enrollment, treadmill speed alternated with a definitive RCT comparing
moderate-intensity, able to walk 10 m overground 30–60 s rest periods. MCT HIT and MCT appears to be
continuous aerobic training with assistive devices as needed strategy involved continuous feasible and warranted
in people with chronic without physical assistance, able treadmill walking at 45–50%
stroke to walk 3 min on the treadmill at of heart rate reserve.
≥ 0.3 mph (0.13 m/s) with no
aerobic exercise
contraindications, stable
cardiovascular condition.
Murata et al. To investigate the efficacy of Cross-sectional sample Seven patients with chronic Participants had 2 or 16 weeks Walking speed increased in three The transition from the support
(2016) stepping-in-place training hemiplegic stroke (age: of intervention. Evaluations participants after 2 or 16 weeks phase support to the swing
using a foot lifting assist 80.9 ± 4.9 years) participated in were performed before the of intervention. Swing phase phase was shortened after the
device on the walking gait the study. Inclusion: able to baseline period and before percentage increased in the intervention. The stepping-in-
of chronic hemiparetic maintain the standing position and after the intervention paretic gait cycle. A walking place exercise using the

2017, Vol 00 No 00
stroke patients independently, and able to walk at period (two-dimensional speed gain was found in three device designed for this study
least 10 m on flat ground with or motion analysis of walking and participants, and, furthermore, a may improve the muscle
without minimum assistance in stepping-in-place exercises cadence gain was found. strength of the lower limb and
walking balance. and a clinical evaluation). coordination in the pre-swing
phase of the paretic limb.
Jeong et al. To investigate the effects of Pilot randomized Twenty-nine patients randomly Participants had 30 min of Significant differences were Treadmill walking combined with
(2016) treadmill walking combined controlled trial assigned to one of the following: active/passive exercises and found in 6 MWT and BBS obstacle-crossing training may
with obstacle-crossing on experimental group treadmill 30 min of gait training in the scores between the groups help improve the walking ability
the walking ability of walking training : (N = 15) or CG form of treadmill walking. They after adjusting for baseline of patients with hemiplegic
ambulatory poststroke (TWT): N = 14. Inclusion: first participated in daily sessions value, but not at the 10 MWT, stroke and can possibly be
patients stroke (6 months to <1 year); of regular intervention for 1 h/ timed up and go, and Activities- used as an adjunct to routine
24–80 years of age; hemiplegic day, 5 times/week, for Specific Balance Confidence rehabilitation therapy as a task-
stroke with gait disturbance; 4 weeks, with 30 min of active/ scale scores, with the TWT-OC oriented practice on the basis
independent ambulation with or passive stretching exercises group demonstrating of community ambulation
without walking aids for over as well as upper and lower significantly greater BBS
10 m and a score of 4 or 5 on the extremity training (active scores of the two groups. Both
Functional Ambulation exercises and strengthening) groups demonstrated
Classification; sufficient cognition and 30 min of gait training. significant improvements in all
to participate in the training. Only the 30 min allocated for variable.
gait training differed in content
between the groups (TWT vs.
TWT-OC).
Danks et al. To determine preliminary Randomized controlled A total of 37 participants. Inclusion: Subjects were assigned to either There was a significant effect of The addition of a step activity
(2016) efficacy and to identify trial with blinded had sustained a stroke FAST, which consisted of time for both groups, with all monitoring program to a fast
baseline characteristics assessors > 6 months before; were able to walking training at their fastest outcomes improving from walking training intervention
predicting who would walk without assistance; were possible speed on the pretraining to post-training (all may be most effective in
benefit most from fast able to walk 5 min at a self- treadmill (30 min) and P < 0.05). The FAST + SAM person with chronic stroke
walking training plus a step selected pace on the treadmill; overground 3 times/week for was superior to the FAST for who have initial low levels of
activity monitoring program were able to walk outside the 12 weeks 6 MWT (P = 0.018), with a walking endurance and
(FAST + SAM) compared home before stroke; walked larger increase in the activity. Regardless of
with fast walking training <10 000 steps/day; and were FAST + SAM group. The baseline performance, the
(FAST) alone in persons able to communicate with the intervention had differential FAST + SAM intervention
with chronic stroke investigators effectiveness on the basis of was more effective for
baseline step activity. improving walking
endurance.
 Walking trainings in stroke survivors Tshiswaka et al. 11

3 months of PRB exercise and

exercise program supported by

motivational group discussions

treadmill training; CWT, comfortable 10-min walk test; LT, locomotor training; MWT, meter walk test; FGA, functional gait assessment; FWT, fast 10-m walk test; OWET, Overground Walking Exercise Training; PT, physical training;
ABC, Activities-Specific Balance Confidence; ANOVA, analysis of variance; APA, Adaptive Physical Activity; BAT, body awareness training; BSS, Berg Balance Scale; BWS, body weight support; BWSTT, body weight-supported
and one home-based exercise
were found in this review. Although most of the studies

SIS, Stroke Impact Scale; SPPB, Short physical Performance Battery; THRR, targeted heart rate range; TUG, the timed up and go test; TWET, Treadmill Walking Exercise Training; UE, upper extremity; QoL, quality of life.
walking speed in individuals
balance at 3 months and in

appears to be effective for

In chronic stroke patients,

induced improvements in
found significant improvements associated with these
motivational discussions

walking speed at 3 and

resistance and balance

short-term balance and

6 months. Progressive
training approaches, still others reported no improve-

with chronic stroke.

ment. This is in agreement Rensink et al. (2009), who
note significant walking function improvement for groups
of individuals using overground and treadmill training,
but no significant difference between those groups.
Nonetheless, the majority of studies reported improve-
speed relative to the CG. A faster
balance and comfortable walking

ments among poststroke individuals, consistent with

At 3 months, the IG showed

previous studies. For instance, An and Shaughnessy

significant improvements in

walking speed persisted at

(2011) found that gait training was effective in improving

walking function after stroke, especially when it was
initiated early, similar to the findings obtained by
Mackay-Lyons et al. (2013). Early initiation of rehabili-
6 months.

tation may also improve the overall quality of life of

stroke survivors (Chaiyawat et al., 2009; An and
Shaughnessy, 2011), although perhaps only for certain
types of rehabilitation, for example, gait training (Combs
weekly over 3-months period and
discussions twice for 3 months.

et al., 2010).
Training performed in classes
IG: PRB exercises combined

that were conducted twice

consisted of different work

with motivational group

stations with functional

Maintenance of rehabilitative gains remains a challenge

for after stroke interventions. Although the benefits of
rehabilitation can help stroke survivors become inde-
pendent by facilitating more ambulatory and social life-
exercises.

styles (Dowla and Chan, 2010; Ada et al., 2013). Further,

Severinsen et al. (2014) reported that stroke survivors
who had gained muscle strength or aerobic capacity
a lack of outdoor walking for at least
1–3 years previously were allocated

walk a minimum of 10 m and either

during non-task-specific training lost those improve-

n = 33). Inclusion: stroke of within
the previous 1–3 years, ability to

ments after the program was over.

to an intervention group (IG,
Sixty-seven community-living

n = 34). Control group (CG,

65–85 years) with a stroke

These temporary gains could signal three major situa-

individuals (76% male;

tions: (a) that there is a lack of programs for stroke sur-

vivors to participate to maintain their improvements; (b)
5 days/week.

that after stroke survivors lack the personal motivation to

continue engaging in intervention programs; (c) or that
stroke survivors may have other priorities in terms of
what they would like to experiance improvements in
terms of their after stroke condition. For instance, Combs
trial with follow-up at 3,
Randomized controlled

et al. (2013) reported that 76% of stroke survivors would

6, and 15 months

like to observe improvements in their walking distance

rather than their walking speed. This underscores the
importance of knowing the goals or aspirations of post-
stroke individuals in relation to walking function before
implementing rehabilitation.
balance exercise on physical

Like other research, which reported gait speed and spa-

and psychological functions
progressive resistance and
To evaluate the effects of

of poststroke individuals

tiotemporal parameters as the most used outcome mea-

sures (Mudge and Stott, 2007; Van Bloemendaal et al.,
2012), the current review similarly reported such out-
come measures across different walking trainings. In so
doing, this review also highlights the relationships
between specific walking trainings and related outcomes
to propose new intervention strategies and potential
outcome measures for use in future research. Hancock
et al. (2012), for example, have reported beneficial effects
Vahlberg et al.

on balance, physical function, and muscle strength in

(2016)

stroke survivors through the use of reciprocal pedaling

exercise. In a subsequent review by Obembe and Eng

Copyright r 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 12 International Journal of Rehabilitation Research 2017, Vol 00 No 00

(2016), they found a small beneficial effect from physical Billinger SA, Arena R, Bernhardt J, Eng JJ, Franklin BA, Johnson CM, et al. (2014).
intervention trainings on socialization. Overall, these Physical activity and exercise recommendations for stroke survivors. Stroke
45:2532–2553.
kinds of outcome measures – balance, physical function, Boyne P, Dunning K, Carl D, Gerson M, Khoury J, Rockwell B, et al. (2016). High-
muscle strength, and socialization – range across physical intensity interval training and moderate-intensity continuous training in
and emotional outcomes that suggest improvements to ambulatory chronic stroke: Feasibility Study. Physical therapy 96:1533–1544.
Centers for Disease Control and Prevention. (2016). Stroke fact sheet. CDC.
quality of life. In general, these findings and our review Available at: https://www.cdc.gov/dhdsp/data_statistics/fact_sheets/fs_
alike point to a need for maintained improvement of stroke.htm. [Accessed 20 January 2017].
quality of life and physical function following rehabili- Chaiyawat P, Kulkantrakorn K, Sritipsukho P (2009). Effectiveness of home
rehabilitation for ischemic stroke. Neurol Int 1:e10.
tation participation after stroke. Combs SA, Dugan EL, Passmore M, Riesner C, Whipker D, Yingling E, Curtis AB
(2010). Balance, balance confidence, and health-related quality of life in
There are many limitations to this review that we would persons with chronic stroke after body weight–supported treadmill training.
like to highlight. One is related to the fact that we relied Arch Phys Med Rehabil 91:1914–1919.
Combs SA, Van Puymbroeck M, Altenburger PA, Miller KK, Dierks TA, Schmid AA
on studies that have been published in English, ignoring
(2013). Is walking faster or walking farther more important to persons with
other major studies published in languages other than chronic stroke? Disabil Rehabil 35:860–867.
English. In addition, participants were exposed to several Combs-Miller SA, Kalpathi Parameswaran A, Colburn D, Ertel T, Harmeyer A,
Tucker L, Schmid AA (2014). Body weight-supported treadmill training vs.
treatments that make it difficult to focus solely on the
overground walking training for persons with chronic stroke: a pilot rando-
impact of walking trainings alone. We did not identify a mized controlled trial. Clin Rehabil 28:873–884.
single longitudinal study that could effectively capture Danks KA, Pohlig R, Reisman DS (2016). Combining fast-walking training and a
step activity monitoring program to improve daily walking activity after stroke: a
the impact of participating in walking trainings. preliminary study. Archives of physical medicine and rehabilitation 97:S185–
S193.
Dean CM, Rissel C, Sherrington C, Sharkey M, Cumming RG, Lord SR, et al.
Conclusion (2012). Exercise to enhance mobility and prevent falls after stroke: the
This study reviews the impact of different types of community stroke club randomized trial. Neurorehabil Neural Repair
walking training interventions on walking function 26:1046–1057.
Dowla N, Chan L (2010). Improving quality in stroke rehabilitation. Top Stroke
among stroke survivors. Although such a rehabilitation Rehabil 17:230–238.
intervention has a positive impact in improving stroke Drużbicki M, Guzik A, Przysada G, Kwolek A, Brzozowska-Magoń A (2015).
survivors’ walking function and quality of life, it remains Efficacy of gait training using a treadmill with and without visual biofeedback in
patients after stroke: a randomized study. J Rehabil Med 47:419–425.
a challenge to maintain the gains of rehabilitation over Hancock NJ, Shepstone L, Winterbotham W, Pomeroy V (2012). Effects of lower
time after the cessation of the rehabilitative interven- limb reciprocal pedalling exercise on motor function after stroke: a systematic
tions. Moreover, differences in rehabilitative interven- review of randomized and nonrandomized studies. Int J Stroke 7:47–60.
Harris AD, Mcgregor JC, Perencevich EN, Furuno JP, Zhu J, Peterson DE,
tion, as well as proxies for measuring rehabilitative gains Finkelstein J (2006). The use and interpretation of quasi-experimental studies
such as walking speed, walking distance, balance, and so in medical informatics. J Am Med Inform Assoc 13:16–23.
forth, make objective assessments of the value of differ- Jeong Y-G, Koo J-W (2016). The effects of treadmill walking combined with
obstacle-crossing on walking ability in ambulatory patients after stroke: a pilot
ent approaches difficult at best. Programs to maintain – or randomized controlled trial. Topics in stroke rehabilitation 23:406–412.
research to identify ways to better maintain – improve- Jeon S-H, Lee S-M, Kim J-H (2015). Therapeutic effects of reaching with forward
ments gained from rehabilitation even after rehabilitation bending of trunk on postural stability, dynamic balance, and gait in individuals
with chronic hemiparetic stroke. J Phys Ther Sci 27:2447–2451.
ceases are needed, as well as a more standardized mea- Johnson BH, Bonafede MM, Watson C (2016). Short-and longer-term health-care
sure of after stroke walking function improvement resource utilization and costs associated with acute ischemic stroke.
overall. Clinicoecon Outcomes Res 8:53.
Krishnamoorthy V, Hsu W-L, Kesar TM, Benoit DL, Banala SK, Perumal R, et al.
(2008). Gait training after stroke: a pilot study combining a gravity-balanced
orthosis, functional electrical stimulation, and visual feedback. Journal of
Acknowledgements neurologic physical therapy 32:192–202.
Conflicts of interest Lloréns R, Gil-Gómez JA, Alcañiz M, Colomer C, Noé E (2015). Improvement in
There are no conflicts of interest. balance using a virtual reality-based stepping exercise: a randomized con-
trolled trial involving individuals with chronic stroke. Clinical rehabilitation
29:261–268.
Mackay-Lyons M (2012). Aerobic treadmill training effectively enhances cardio-
References vascular fitness and gait function for older persons with chronic stroke. J
Ada L, Dean CM, Lindley R (2013). Randomized trial of treadmill training to Physiother 58:271.
improve walking in community-dwelling people after stroke: the AMBULATE Mackay-Lyons M, Mcdonald A, Matheson J, Eskes G, Klus M-A (2013). Dual
trial. Int J Stroke 8:436–444. effects of body-weight supported treadmill training on cardiovascular fitness
American Stroke Association (2012). Effests of stroke. ASA. Available at: http:// and walking ability early after stroke: a randomized controlled trial.
www.strokeassociation.org/STROKEORG/AboutStroke/EffectsofStroke/ Neurorehabil Neural Repair 27:644–653.
Effects-of-Stroke_UCM_308534_SubHomePage.jsp. [Accessed20 January Mayo NE, Mackay-Lyons MJ, Scott SC, Moriello C, Brophy J (2013). A rando-
2017]. mized trial of two home-based exercise programmes to improve functional
An M, Shaughnessy M (2011). The effects of exercise-based rehabilitation on walking post-stroke. Clin Rehabil 27:659–671.
balance and gait for stroke patients: a systematic review. J Neurosci Nurs Moore JL, Roth EJ, Killian C, Hornby TG (2009). Locomotor training improves daily
43:298–307. stepping activity and gait efficiency in individuals poststroke who have
Bang D-H, Cho H-S (2016). Effect of body awareness training on balance and reached a ‘plateau’ in recovery. Stroke 41:129–135.
walking ability in chronic stroke patients: a randomized controlled trial. Journal Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, et al.
of physical therapy science 28:198–201. (2016). Executive summary: Heart Disease and Stroke Statistics-2016
Billinger SA, Mattlage AE, Ashenden AL, Lentz AA, Harter G, Rippee MA (2012). update: a report from the American Heart Association. Circulation 133:447.
Aerobic exercise in subacute stroke improves cardiovascular health and Mudge S, Stott NS (2007). Outcome measures to assess walking ability following
physical performance. Journal of neurologic physical therapy: JNPT 36:159. stroke: a systematic review of the literature. Physiotherapy 93:189–200.

Copyright r 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Walking trainings in stroke survivors Tshiswaka et al. 13

Murata K, Asai H, Inaoka PT, Nakaizumi D (2016). Walking gait changes after Smith PS, Thompson M (2008). Treadmill training post stroke: are there any
stepping-in-place training using a foot lifting device in chronic stroke patients. secondary benefits? A pilot study. Clinical rehabilitation 22:997–1002.
Journal of physical therapy science 28:1170–1177. Sousa CO, Barela JA, Prado-Medeiros CL, Salvini TF, Barela AM (2011). Gait
Obembe AO, Eng JJ (2016). Rehabilitation interventions for improving social training with partial body weight support during overground walking for indi-
participation after stroke: a systematic review and meta-analysis. Neurorehabil viduals with chronic stroke: a pilot study. J Neuroeng Rehabil 8:48.
Neural Repair 30:384–392. Srivastava A, Taly AB, Gupta A, Murali T (2015). Rehabilitation interventions to
Olawale O, Jaja S, Anigbogu C, Appiah-Kubi K, Jones-Okai D (2011). Exercise improve locomotor outcome in chronic stroke survivors: A prospective,
training improves walking function in an African group of stroke survivors: a repeated-measure study. Neurology India 63:347.
randomized controlled trial. Clin Rehabil 25:442–450. Stuart M, Benvenuti F, Macko R, Taviani A, Segenni L, Mayer F, et al. (2009).
Pang MY, Eng JJ (2008). Determinants of improvement in walking capacity among Community-based adaptive physical activity program for chronic stroke: fea-
individuals with chronic stroke following a multi-dimensional exercise program. sibility, safety, and efficacy of the Empoli model. Neurorehabil Neural Repair
23:26–34.
Journal of rehabilitation medicine 40:284–290.
Sullivan KJ, Klassen T, Mulroy S (2006). Combined task‐specific training and
Park J, Park S-Y, Kim Y-W, Woo Y (2015). Comparison between treadmill training
strengthening effects on locomotor recovery post-stroke: a case study. J
with rhythmic auditory stimulation and ground walking with rhythmic auditory
Neurol Phys Ther 30:130–141.
stimulation on gait ability in chronic stroke patients: A pilot study.
Suresh K (2011). An overview of randomization techniques: an unbiased
NeuroRehabilitation 37:193–202.
assessment of outcome in clinical research. J Hum Reprod Sci 4:8.
Park JM, Lim HS, Song CH (2015). The effect of external cues with vibratory
Tomey KM, Sowers MR (2009). Assessment of physical functioning: a conceptual
stimulation on spatiotemporal gait parameters in chronic stroke patients.
model encompassing environmental factors and individual compensation
Journal of physical therapy science 27:377–381. strategies. Phys Ther 89:705.
Rensink M, Schuurmans M, Lindeman E, Hafsteinsdottir T (2009). Task‐oriented Vahlberg B, Cederholm T, Lindmark B, Zetterberg L, Hellström K (2016). Short-
training in rehabilitation after stroke: systematic review. J Adv Nurs 65:737– term and long-term effects of a progressive resistance and balance exercise
754. program in individuals with chronic stroke: a randomized controlled trial.
Salbach NM, Mayo NE, Robichaud‐Ekstrand S, Hanley JA, Richards CL, Wood‐ Disability and rehabilitation 39:1615–1622.
Dauphinee S (2005). The effect of a task‐oriented walking intervention on Van Bloemendaal M, Van De Water AT, Van De, Port IG (2012). Walking tests for
improving balance self‐efficacy poststroke: A randomized, controlled trial. stroke survivors: a systematic review of their measurement properties. Disabil
Journal of the American Geriatrics Society 53:576–582. Rehabil 34:2207–2221.
Severinsen K, Jakobsen JK, Pedersen AR, Overgaard K, Andersen H (2014). Van Duijnhoven HJ, Heeren A, Peters MA, Veerbeek JM, Kwakkel G, Geurts AC,
Effects of resistance training and aerobic training on ambulation in chronic Weerdesteyn V (2016). Effects of exercise therapy on balance capacity in
stroke. Am J Phys Med Rehabil 93:29–42. chronic stroke. Stroke 47:2603–2610.

Copyright r 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.