Rehabilitation of Balance After Stroke With Multisensorial Training: A Single-Blind Randomized Controlled Study

Abstract

To compare 2 rehabilitation strategies to improve balance after stroke: (1) a multisensorial approach based on higher intensity of balance tasks and exercise during visual deprivation and (2) a conventional neurodevelopmentaltheory-based treatment (NDT) that used a general approach for sensorimotor rehabilitation.
This prospective, multicenter, randomized, parallel-group study measured outcomes with blinded assessors. Sixty-eight patients able to walk without human assistance were entered from 3 to 15 months (mean, 7 months) after a first hemispheric stroke. They received 20 sessions in 4 weeks of NDT or multisensorial rehabilitation. On day 0, day 30, and day 90, assessment included the Berg Balance Scale (BBS), posturography, gait (velocity, double stance phase, climbing 10 steps, amount of walking per day), the Functional Independence Measure, and the Nottingham Health Profile.
All subjects improved significantly in balance and walking parameters. Regarding the main dependent variable (BBS on day 30), no difference between groups was found. Analysis of secondary outcomes suggested small differences in favor of the experimental group, but the differences are not likely to be clinically relevant.
No evidence was found for the superiority of a multisensorial rehabilitation program in ambulatory patients with impairments beyond the time of inpatient therapy. Additional studies are recommended.

Full text

http://nnr.sagepub.com/
Repair
Neurorehabilitation and Neural
http://nnr.sagepub.com/content/22/5/468
The online version of this article can be found at:
DOI: 10.1177/1545968308315996
2008 22: 468Neurorehabil Neural Repair
Lebomin, Jean-Philippe Regnaux, Dominic Pérennou and Eric Vicaut
Alain P. Yelnik, Frederique Le Breton, Florence M. Colle, Isabelle V. Bonan, Caroline Hugeron, Véronique Egal, Elizabeth
Controlled Study
Rehabilitation of Balance After Stroke With Multisensorial Training: A Single-Blind Randomized
Published by:
http://www.sagepublications.com
On behalf of:
American Society of Neurorehabilitation
can be found at:Neurorehabilitation and Neural RepairAdditional services and information for
http://nnr.sagepub.com/cgi/alertsEmail Alerts:
http://nnr.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
http://nnr.sagepub.com/content/22/5/468.refs.htmlCitations:
What is This?
- Sep 9, 2008Version of Record >>
at Ecole Nationale de la Sante on December 4, 2013nnr.sagepub.comDownloaded from at Ecole Nationale de la Sante on December 4, 2013nnr.sagepub.comDownloaded from

468 Copyright © 2008 The American Society of Neurorehabilitation
Rehabilitation of Balance After Stroke With
Multisensorial Training: A Single-Blind
Randomized Controlled Study
Alain P. Yelnik, MD, Frederique Le Breton, MD, Florence M. Colle, MD,
Isabelle V. Bonan, MD, PhD, Caroline Hugeron, MD, Véronique Egal, Elizabeth Lebomin,
Jean-Philippe Regnaux, PhD, Dominic Pérennou, MD, PhD, Eric Vicaut, MD, PhD
Objective. To compare 2 rehabilitation strategies to improve
balance after stroke: (1) a multisensorial approach based on
higher intensity of balance tasks and exercise during visual
deprivation and (2) a conventional neurodevelopmental-
theory-based treatment (NDT) that used a general approach
for sensorimotor rehabilitation. Methods. This prospective,
multicenter, randomized, parallel-group study measured out-
comes with blinded assessors. Sixty-eight patients able to walk
without human assistance were entered from 3 to 15 months
(mean, 7 months) after a first hemispheric stroke. They
received 20 sessions in 4 weeks of NDT or multisensorial reha-
bilitation. On day 0, day 30, and day 90, assessment included
the Berg Balance Scale (BBS), posturography, gait (velocity,
double stance phase, climbing 10 steps, amount of walking per
day), the Functional Independence Measure, and the
Nottingham Health Profile. Results. All subjects improved sig-
nificantly in balance and walking parameters. Regarding the
main dependent variable (BBS on day 30), no difference
between groups was found. Analysis of secondary outcomes
suggested small differences in favor of the experimental group,
but the differences are not likely to be clinically relevant.
Conclusion. No evidence was found for the superiority of a
multisensorial rehabilitation program in ambulatory patients
with impairments beyond the time of inpatient therapy.
Additional studies are recommended.
Key Words: Stroke—Balance—Rehabilitation—Functional
independence—Sensory plasticity.
Balance disorders after stroke are a major problem.
The risk of fall for hemiparetic patients is dramat-
ically higher than for the general population dur-
ing acute poststroke hospitalization,1during the
rehabilitation period,2,3 and later during community liv-
ing.4There are specific risk factors for falls after stroke,
mainly mental disorders, urinary incontinence, motor
impairment, visuospatial neglect, and postural stability.1,5
For older women living at home 1 year after stroke, the
strongest risk factor for falls is a balance problem while
dressing rather than the usual risk factors reported in the
general population.6Complications of falls include frac-
tures, but one of the most important complication is the
fear of falling and the consequent restriction of activity.
This leads to limitation of autonomy, dependence on
caregivers, mood disorders, and deterioration of quality
of life. Initial balance disability is also a strong predictor
of function and recovery after stroke.7Thus, one of the
main objectives for physical rehabilitation after stroke
should be the restoration of balance control, allowing the
patient to walk and conduct daily activity in safety.
The mechanisms of balance disorders after stroke often
interact and include the site and volume of the stroke, pare-
sis, sensory loss, change in muscle tone, ataxia, and spatial
neglect. More subtle impairments include misperception of
visual8,9 or postural verticality10 and an excessive reliance on
visual cues.11 A conservative, highly protective approach to
physical therapy for patients with balance disorders may
not provide the range of therapeutic challenges necessary to
develop the strong balance reactions needed for active liv-
ing. Rather than being excessively focused on the quality of
the movement during walking, such as control of the knee,
physical rehabilitation for balance recovery may aim to pro-
mote more intensive sensory stimulation to help drive
mechanisms involved in brain plasticity for the acquisition
of skills. In a previous study, we observed that excessive
From Physical Medicine and Rehabilitation Department, G.H.
Lariboisière—F. Widal,Université Paris 7, Paris (APY, FLB, FMC, IVB,
VE); Physical Medicine and Rehabilitation Department, Hôpital
Raymond Poincaré, Université Versailles, Garches (CH, EL, JPR);
Physical Medicine and Rehabilitation Department, CHU, Dijon (DP);
Unité de Recherche Clinique, G.H. Lariboisière—F.Widal, Université
Paris 7, Paris (EV), France.
Details of the study group “Balance After Stroke PHRC 2001—AOM
01 102” are given in Appendix A.
This study was supported by grants provided by the Health Ministry
Program for Clinical Research PHRC no. AOM 01 102. The authors
have no conflicts of interest to report.
Address correspondence to Alain P. Yelnik, MD, Physical Medicine and
Rehabilitation Department, G.H. Lariboisière—F.Widal, AP-HP,
Université Paris 7, 200 rue du faubourg Saint Denis, 75010 Paris,
France. E-mail: alain.yelnik@lrb.aphp.fr.
Yelnik AP, Le Breton F, Colle FM, et al. Rehabilitation of balance after
stroke with multisensorial training: a single-blind randomized con-
trolled study. Neurorehabil Neural Repair. 2008;22:468-476.
DOI: 10.1177/1545968308315996
at Ecole Nationale de la Sante on December 4, 2013nnr.sagepub.comDownloaded from

Rehabilitation of Balance After Stroke
Neurorehabilitation and Neural Repair 22(5); 2008 469
reliance on visual cues by chronic hemiplegic patients was
lessened by a specific training program of visual depriva-
tion, with improvement of functional parameters such as
walking and climbing steps.12
This study compared 2 physical rehabilitation approaches
to restore balance after recent stroke. The first, a conven-
tional approach called neurodevelopmental training (NDT),
guides the quality of movement. The second, termed mul-
tisensorial, is based on 2 principles: the quantity of equilib-
rium tasks without emphasis on the quality of movement
and exercises performed during visual deprivation. To our
knowledge, this is the first comparison study that uses ran-
domization and single-blind evaluations.
METHODS
Subjects
Subjects who had hemiplegia after a single hemi-
spheric stroke due to an infarct or hemorrhage shown
by computerized tomography or magnetic resonance
imaging within 3 to 15 months prior to entry were
included. At onset of stroke,subjects had to be unable to
walk for at least 2 weeks, but not exceeding 3 months.
Walking was defined as the ability to walk at least 50
meters with an orthosis or cane if needed but without
human assistance. Patients had to be less than 80 years
old, ambulatory, and living at home. The physical reha-
bilitation was conducted in 1 of 2 rehabilitation centers
(Fernand Widal and Raymond Poincaré).
Exclusion criteria included a previous history of
walking disorder, cognitive disorders that prevented
comprehension of the rehabilitation program, and
history of a vestibular disorder.
Patients had to be able to give their written informed
consent. This study was approved by the ethics commit-
tee of Saint Louis Hospital (Paris, France).
Procedures
The following data was collected: age; gender; type of
stroke, ischemic or hemorrhagic; side and site of stroke;
time between stroke and inclusion; previous medical
history; loss of the visual field confirmed by Goldman
perimetry; existence of visuospatial neglect according to
the bells test13 and the bisection of a 20-cm-long line,
neglect being defined by an omission of at least 7 bells
and an error more than 6 mm of the line bisection;
motor function assessed by the Motricity Index14;and
proprioception assessed as normal (2), diminished (1),
or absent (0) for the perception of ankle movements.
The experimental design was a prospective, multicen-
ter, randomized parallel-group trial with a single-blind
evaluation. During a preliminary visit, each patient was
examined by the main investigator in the center to
decide if the inclusion criteria were fulfilled and to pro-
vide the information package. All subjects had a clinical
otologic examination. After patients gave informed con-
sent, the trial statistician generated the randomization
sequence using random number tables. Randomization
was stratified by center. Then the first visit for evalua-
tion (day 0 [D0]) was conducted by one of the blinded
evaluators. Physical therapy had to begin within 7 days,
and the 20 successive sessions had to be conducted 5
days a week for the following 4 weeks. Posttreatment
evaluation (day 30 [D30]) was carried out within 7 days
of the end of the physical rehabilitation program. The
second posttreatment evaluation (day 90 [D90]) was
carried out 3 months after the first evaluation. All the
data were collected in each center. The first patient was
included in July 2002, the last in March 2005, and the
last evaluation was done in June 2005.
Physical Rehabilitation
Programs (see Appendix B)
Group 1 (NDT-based treatment). Global sensorimotor
rehabilitation, based on the neurodevelopmental approach
described by Bobath,15 targeting more on the control of
weight bearing and shifting in erect stance and the qual-
ity of gait with less emphasis on the response to desta-
bilization situations.
Group 2 (multisensorial). Physical rehabilitation based
on the manipulation of the sensory information
required to maintain balance, attention being paid to
the amount of exercise, that is, duration and intensity,
rather than the quality of the movement. Most of the
exercises were conducted in visual deprivation, thus
challenging the selection and synthesis by the brain of
vestibular and somatosensory information.
Evaluation.Standing balance was assessed by means of
the Berg Balance Scale (BBS), a specific tool developed
to measure the functional standing balance of geriatric
individuals, validated in subjects with acute stroke16 and
with a force platform (Satel, Blagnac, France) recording
the limits of stability. Following a procedure previously
described,17 subjects were required to lean forward,
backward, and then to the right and left, without step-
ping or moving their feet from the standard foot posi-
tion. Data analyzed were the means over 2 trials of the
sum of the center of pressure (CoP) displacement in
millimeters on each sagittal and frontal axis. Dynamic
balance was assessed during walking by the percentage
of double-limb stance time, which reflects balance
at Ecole Nationale de la Sante on December 4, 2013nnr.sagepub.comDownloaded from

Yelnik et al
470 Neurorehabilitation and Neural Repair 22(5); 2008
during walking,18 and the speed used to walk 10 m com-
fortably. The equipment used to analyze the gait pattern
was the locometer used by Bessou et al.19 Walking was
also assessed by the time taken to climb 10 steps and
return, by self-report of the time spent daily in out-of-
house walking, by the subjective perception of security
during walking assessed on a 10-point verbal scale (10
being high safety), and by the number of falls since the
stroke. Daily independence was assessed by the Functional
Independence Measurement (FIM).20 Quality of life was
assessed by the Nottingham Health Profile.21
The primary outcome measure and dependant vari-
able was the BBS on D30. Secondary outcomes included
the BBS on D90 and the other data on D30 and D90.
Statistical Methods
Because it was anticipated that the statistical distribu-
tion would not be Gaussian, the sample size calculation
was made using the method proposed by Noether.22 A
sample size of 36 in each group would have 80% power to
detect a probability of .7 that an observation in group 1 is
less than an observation in group 2 using a Wilcoxon
(Mann–Whitney) rank-sum test with a .05 two-sided sig-
nificance level. The main analysis was made on the intent-
to-treat population. In cases with a missing value, the last
observation value was carried forward. For 4 secondary
dependant variables, we had 1 or 2 missing values due to
a computer failure. In this case, we did not replace the
data, and because the analysis of covariance used the base-
line value for the covariate, we made the analysis of these
variables on the 67 or 66 patients available. A sensitivity
analysis for the main criteria was made using all the
patients who followed the protocol without major devia-
tions. As a result of the non-Gaussian statistical distribu-
tion, nonparametric analysis of covariance was used for all
variables except velocity and functional independence, for
which the statistical distribution allowed the use of a para-
metric covariance analysis. In all the cases, baseline values
of the variables were the covariates. In addition, time-
dependent changes were tested by the Wilcoxon test for
matched pairs. All calculations were made using SAS 9.13
(SAS Institute, Cary, NC). Because each criterion was
tested at 2 time points, α-risk was adjusted by Bonferonni’s
method, and a result was considered as significant if the
2-sided Pvalue was lower than .025.
RESULTS
Subjects’ Characteristics
The number of potential subjects screened was not
recorded, but no eligible patient refused to participate.
Sixty-seven of the 68 enrolled patients completed the
study but the analysis, conducted for intention to treat,
included all of the patients: 35 in the NDT group and 33
in the multisensorial group. As shown in Table 1, no dif-
ferences were found at entry between groups: mean age
was 55 years, mean time after stroke was 7 months (218
days), and most of the strokes were ischemic. Motor
impairment was substantial, but autonomy was high by
the mean FIM score. Fifty-one patients used a walking
aid (simple cane 15 and 18, crutch cane 5 and 4, tripod
cane 5 and 4). Table 2 summarizes the status of the 2
groups for the parameters assessed on D0. The mean
number of falls since stroke was 3.4 and 2.0 at the day of
inclusion, ranging from 0 to 15. There was no difference
between groups, except for daily time of walking, which
was slightly higher in the NDT-based group.
Major Deviations From the Protocol
In the NDT-based group, 1 patient had to stop after
5 sessions of physical therapy for carotid surgery, which
had not been planned. He could not be assessed. In the
multisensorial group, 1 patient was lost to follow-up
between D30 and D90 and another because of an
adverse event unrelated to the treatment. Two patients
had fewer than 16 physical therapy sessions (12 and 15),
but they could be assessed at D30 and D90. For 6
patients, the time lag for the first assessment was longer
than 45 days (46, 47, 2 ×48, and 2 ×49 days). All sub-
jects were included in the analysis, except for the sensi-
tivity analysis. In one center, initially expected to be a
third center, no data were available for the first 4
patients and the investigator had left the facility. We
stopped the study in this center. Thus, the analysis was
made on 68 instead of the 72 expected.
Assessment on Day 30 and
Day 90 in Each Group
A significant improvement was observed in each
group for most of the criteria, except for the limits of
stability on force platform, which remained unchanged
for both groups. Intragroup comparisons are given in
Tables 3 and 4.
Differences Between Groups on Day 30
(Table 3 and Figures 1-4)
No difference between groups was found for the
primary dependant variable (BBS). Regarding sec-
ondary criteria, a significant difference between groups,
at Ecole Nationale de la Sante on December 4, 2013nnr.sagepub.comDownloaded from

Rehabilitation of Balance After Stroke
Neurorehabilitation and Neural Repair 22(5); 2008 471
in favor of the experimental group, was only found for
functional independence (P=.01).
Differences Between Groups
on Day 90 (Table 4 and Figures 1-4)
The difference between groups for functional inde-
pendence (P=.006) was strengthened. A difference in
favor of the experimental group appeared for the double
stance phase during walking (P=.02) and for quality of
life (P=.004). In a post hoc analysis, the correlation
between FIM, Nottingham, and double stance phase was
tested. FIM was significantly correlated with the double
stance phase at D30 (P=.0003) and at D90 (P=.0001).
There was no correlation between quality of life and the
double stance phase. The other outcome measures did
not show any significant improvement.
Finally, in both groups a significant improvement of
most of the parameters was observed. As there was no
difference between groups for the primary dependant
variable, the superiority of the multisensorial approach
was not in evidence. There was a statistical trend in
favor of the multisensorial approach as evidenced on
D90 by a lower percentage of double stance time during
walking, higher functional independence, and better
quality of life, but the clinical meaning to patients of
small values for improvement is questionable.
Table 1. General Characteristics of the Groupsa
NDT-Based Treatment (n =35) Multisensorial (n =33)
Gender male 62.9 (22) 66.7 (22)
Age mean ±SD, (range) 54.9 ±11.8 (26.5-77.3) 55.5 ±11.6 (32.5-78.3)
Side of the lesion, right/left 17/16 20/15
Ischemic stroke 68.6 (24) 75.8 (25)
Mean time after stroke (days) 218.4 ±93.4 217.2 ±92.9
Motricity index
Upper limb (max 100) 33.1 ±26 37.3 ±25.2
Lower limb (max 100) 57.2 ±19.2 57.1 ±17.2
Lower limb proprioception
Anesthesia — 3.0 (1)
Hypoesthesia 62.9 (22) 42.4 (14)
Normal 37.1 (13) 54.6 (18)
Visual field defect 8.6 (3) 6.1 (2)
Visuospatial neglect 17 (6) 12.1 (4)
Abbreviations: NDT, neurodevelopmental theory; SD, standard deviation.
aData are given as mean ±SD or percentage and (number).
Table 2. Status for the 2 Groups on Day 0
Median (Q1; Q3) or Mean (95% CI)
NDT-Based Treatment (n =35) Multisensorial (n =33)
Berg Balance Scale 47 (39; 53) 49 (42; 53)
Speed of walking 0.63 (0.50; 0.76)a,b 0.61 (0.50; 0.72)b
Percentage of double-limb stance time 16.1 (10.7; 21.1)a15.2 (11.1; 24.7)
Time to climb 10 steps and return 33 (23; 55)a35 (26; 55)
Daily time of walking (minutes) 30 (0; 60) 20 (0; 30)
Security sensation during walking 7 (5; 8) 6 (5; 7)
Number of falls since stroke 2 (1; 4) 1 (0; 2)
Functional Independence Measure 109 (105.7; 112.2)b108.5 (105.6; 111.4)b
Quality of life 178.1 (81.4; 301.3) 157.7 (124.3; 269.4)
Posturographic limits of stability 176.8 (82.4; 282.8)c230.6 (103.9; 286.2)
Abbreviations: NDT, neurodevelopmental theory; CI, confidence interval.
aTwo data missing.
bMean (95% CI).
cOne datum missing.
at Ecole Nationale de la Sante on December 4, 2013nnr.sagepub.comDownloaded from

Yelnik et al
472 Neurorehabilitation and Neural Repair 22(5); 2008
DISCUSSION
The aim of this controlled trial was to compare 2
kinds of physical rehabilitation approaches to restore
balance after stroke. On the basis of previous studies
showing an excessive dependence on vision in chronic
hemiplegic patients11 and the efficacy of a rehabilitation
program of visual deprivation,12 we made the hypothe-
sis that a multisensorial training approach, under visual
deprivation with somatosensorial and vestibular stimu-
lation earlier after stroke, could be more efficient than a
traditional one based on neurodevelopmental concepts.
The present study failed to reach the statistical sig-
nificance level for its main dependant variable (ie, the
BBS at D30). Several reasons could explain this result.
First, our choice of primary dependant variable was
perhaps inadequate. On designing this study we
expected a shorter time lag after stroke. The relatively
long period poststroke in our population resulted in
high scores for BBS on D0, which strongly limited the
chances of observing any significant difference between
groups because of a ceiling effect. The choice of the BBS
was probably not the best one, as it primarily assesses
balance in standing position with eyes open. Finally, the
Table 3. Status for the 2 groups on Day 30 With Statistical Significance of Between-Group Differences in the Last Column
(Comparison of Changes From Baseline by Nonparametric Covariance Analysis)
Median (Q1; Q3) or Lsmean (95% CI)
NDT-Based Treatment (N =35) Multisensorial (N =33)
Berg Balance Scale 53 (43;55)a53 (48;55)a.393
Speed of walking .72 (0.67; 0.77)a,b 0.76 (0.71; 0.98)a,b .275b
Percentage of double stance phase 14 (8; 21.7)c14.4 (10.0; 22.1)a.372
Time to climb 10 steps and return 30 (21; 40)a30 (21; 45)a.318
Daily time of walking (minutes) 30 (5; 60)a30 (15; 60)a.183
Security sensation during walking 7 (5; 9), NS 8 (6; 8)a.236
Number of falls since stroke 3 (1; 4) 1 (0; 3) .44
Functional Independence Measure 109.2 (108.1; 110.3), NSb111.2 (110.1; 112.4)a,b .01b
Quality of life 156.8 (46.9; 299.8), NS 109.5 (62.6; 159.0)d.04
Posturographic limits of stability 172.3 (80.5; 270.1), NS 239.0 (103.7; 291.5), NS .361
Abbreviations: CI, confidence interval; NS, not significant.
aP<.001 (statistical significance of intragroup comparison [D30 vs D0]).
bParametric covariance analysis and least square mean (Lsmean), 95% CI.
cP<.025 (statistical significance of intragroup comparison [D30 vs D0]).
dP<.01 (statistical significance of intragroup comparison [D30 vs D0])
Table 4. Status for the 2 Groups on Day 90 With Statistical Significance of Between-Group Differences in the Last Column
(Comparison of Changes From Baseline by Nonparametric Covariance Analysis)
Median (Q1; Q3) or Lsmean (95% CI)
NDT-Based Treatment (N =35) Multisensorial (N =33)
Berg Balance Scale 51 (44; 55)a53 (49.0; 55.0)a.058
Speed of walking 0.73 (0.67; 0.79)a,b 0.79 (0.73; 0.84)a,b .200b
Percentage of double stance phase 15.2 (9; 18.2)c14.4 (9.4; 19.5)c.024
Time to climb 10 steps and return 32 (20; 49)a28 (23.0; 41.0)a.136
Daily time of walking (minutes) 30 (15; 60), NS 60 (30.0; 60.0)a.057
Security sensation during walking 7 (5; 9), NS 7 (5; 8), NS .983
Number of falls since stroke 3 (1; 5) 2 (1; 3) .778
Functional Independence Measure 109.6 (108.3; 111), NSb112.3 (111; 113.6)a,b .006b
Quality of life 138.9 (63.8; 297.1), NS 99.8 (61.5; 163.4)d.004
Posturographic limits of stability 179.8 (84.4; 279.6), NS 228.3 (90.8; 295.6), NS .765
Abbreviations: CI, confidence interval; NS, not significant.
aP<.001 (statistical significance of intragroup comparison [D90 vs D0]).
bParametric covariance analysis and least square mean (Lsmean), 95% CI.
cP<.025 (statistical significance of intragroup comparison [D90 vs D0]).
dP<.01 (statistical significance of intragroup comparison [D90 vs D0]).
at Ecole Nationale de la Sante on December 4, 2013nnr.sagepub.comDownloaded from

choice of D30 might not correspond to a sufficiently
long period for training and evaluation, as suggested by
the increase in difference with time between the 2
groups for most of the parameters. The second possible
reason for not achieving statistical significance on the
BBS scores at D30 was that the potential for improve-
ment in these patients was still high, as we observed a
significant improvement in both groups for most of the
outcome measures. This proves that soon but not early
after stroke, that is, at a mean of 7 months, the potential
of balance and functional improvement with rehabilita-
tion is still high, irrespective of the techniques
employed. Third, the sample size of the study was prob-
ably too small, reflecting too much optimism in the
expected results of the method tested. The high number
of physical therapists participating in the study could
also have influenced the results. The choice of more
than 1 physical therapist per center was made to allow
for immediate change in case of an unexpected event or
departure. To guarantee standardized treatment, the
guide for protocol, as described in Appendix B, was
explained in sufficient detail and was accompanied by a
patient file in which the physical therapist daily noted
the exact progression of the exercises.
As improvement was observed in the 2 groups for
most of the parameters, the absence of improvement in
the limits of stability recorded on the force platform
Rehabilitation of Balance After Stroke
Neurorehabilitation and Neural Repair 22(5); 2008 473
Figure 1. Medians and lower and upper quartiles for Berg
Balance Scale at the different times of the study. Lines repre-
sent 1.5 interquartile distance. Gray represents the multisen-
sorial group.
Figure 2. Medians and lower and upper quartiles for the
double stance phase at the different times of the study. Lines
represent 1.5 interquartile distance. Gray represents the multi-
sensorial group.
Figure 3. Medians and lower and upper quartiles for quality
of life (the best the lower) at the different times of the study.
Lines represent 1.5 interquartile distance. Gray represents the
multisensorial group.
Figure 4. Means and standard error for Functional Independence
Measure at the different times of the study.
at Ecole Nationale de la Sante on December 4, 2013nnr.sagepub.comDownloaded from

seems surprising. Two explanations are reasonable.
First, the task was difficult to understand without train-
ing. Second, many patients performed it using a hip
strategy, leaning the head forward but the pelvis back-
ward, so no measurable increase of CoP displacement
occurred. Therefore, our opinion is that this task is not
suitable for hemiplegic patients and cannot be recom-
mended to assess balance after stroke.
Despite the lack of difference between groups for the
primary dependant variable, the results presented here
allow us to suggest that our working hypothesis is at
least partly right. A trend in favor of the multisensorial
program appeared for functional independence, quality
of life, and a decrease in double limb stance support
during walking. The benefit of functional independence
with the multisensorial program may be related to an
improvement of dynamic balance, because a decrease of
the percentage of double stance phases during walking
appeared over the 3-month period of evaluation. This
latter criterion is good evidence for balance during gait,
and the post hoc analysis showed a significant correla-
tion between FIM and double-limb support at D30 and
at D90. Although the activity during the D30 and D90
period was not monitored, the improvement for the
multisensorial group between D30 and D90 could
reflect a long-term benefit of this training aiming to
prevent the development of visual dependence.
Nevertheless, although statistically significant, the clini-
cal meaning of these differences between groups is lim-
ited as they were based on small values.
Visual dependence for balance is characterized by an
excessive confidence in visual input, even when erro-
neous, despite normal vestibular or sensory function,
and can be detected by balance recording on a dynamic
force platform such as Equitest (Neurocom, Clackamas,
OR).11,23 Such visual dependence has been described in
Parkinson’s disease24 and after peripheral or central
vestibular disorders,25,26 as well as in normal elderly sub-
jects, especially in fallers when compared with nonfall-
ers.27 Thus, it is a risk factor for falls. The detection of
individual comportments before beginning rehabilita-
tion and adapting rehabilitation to them could be
important.28 The usefulness of these exercises may differ
according to individual characteristics related to visual
dependence at baseline, which could not be fully
defined in the present study and deserves to be investi-
gated by further studies.
This new approach for balance rehabilitation after
stroke differs from others commonly practiced.Different
approaches have been tried to improve specifically 2 of
the main disturbances of balance after stroke, asymme-
try of body weight distribution and instability. Most of
them are based on visual biofeedback control of the CoP
displacements on a force platform.29-33 Despite some
interesting results,31,32,34 there is no strong evidence for
the functional usefulness of this force platform train-
ing.35,36 Other interesting different approaches have
been explored. The misperception of postural vertical-
ity, which probably explains some of the more dramatic
problems such as the pusher syndrome, could be an
important goal for rehabilitation.37
Further studies of a multisensorial therapy approach of
greater intensity and duration,38 perhaps combined with
other task-related mobility training, may be worth testing
in patients who have less functional mobility and start
sooner after onset of stroke than the group we studied.
Future studies should also consider using a dynamic mea-
sure such as changes in walking speed or double limb
stance and define the level of visual dependence.
APPENDIX A
The Study Group “Balance After Stroke
PHRC 2001—AOM 01 102”
•Main investigator: Alain Yelnik
•Design: Alain Yelnik, Dominic Pérennou
•Physical therapy for rehabilitation: Fernand Widal
Center—Philippe Roth, Peggy Ruminski, Isabelle
Pousson, and Emmanuelle Normand; Raymond Poincaré
Center—Johanna Robertson, Julie Stil, Emmanuelle
Danzart, Pierre Lenaoures, Dominique Rouat
•Blind evaluators: Frédérique Lebreton, Florence Colle,
Jean Philippe Regnaux
•Monitoring: Denise Mockers for the Clinical Research
Unit
•Statistical analysis: Eric Vicaut and Carole Boutron for
the Clinical Research Unit
•Recruitment: Alain Yelnik, Frédérique Lebreton, Isabelle
Bonan (Fernand Widal Hospital), Caroline Hugeron and
Alexis Schnitzler (Raymond Poincaré Hospital,
Garches), Isabelle Crassard and Marie Germaine Bousser
(Neurological Department, Lariboisière Hospital),
Gérard Amarenco (PMR Department, Rothschild
Hospital), Pascale Pradat-Diehl (PMR Department,
Pitié-Salpétrière Hospital), Jean Pascal Devailly (PMR
Department, Avicenne Hospital)
•ENT examinations: Michel Kossowski, Marc Raynal
(Military Hospital Percy, Clamart)
•Organization: Véronique Egal, Elizabeth Lebomin
•Inclusion and informed consent: Alain Yelnik, Caroline
Hugeron, Alexis Shnitzler
•APHP head of project: Cécile Jourdain
APPENDIX B
Physical Rehabilitation Programs
Group 1 (NDT-Based Treatment)
Principles: Based on global sensory motor rehabilitation
derived from the neurodevelopmental concepts described by
Yelnik et al
474 Neurorehabilitation and Neural Repair 22(5); 2008
at Ecole Nationale de la Sante on December 4, 2013nnr.sagepub.comDownloaded from

Bobath, more attention paid to the quality of the gesture and
gait control, the spasticity, and abnormal movement inhibi-
tion than to the quantity of exercise and an increase of the dif-
ficulty from one session to another depending on the ability of
the patients. Each session lasted 60 to 70 minutes, depending
on the rest required, and included 5 minutes for spasticity
inhibition, 40 to 45 minutes of exercises specific to the session,
10 to 20 minutes of rest distributed throughout the session,
and related to need.
Sessions 1 to 4: Exercises conducted on the Bobath plat-
form, weight shifting, waist dissociation, pelvis control, crawl-
ing, turning over, fourfooting, and standing on the knees.
Session 5 to 8: Exercises on the edge of the platform in sit-
ting position, transfers from lying to sitting, sitting to stand-
ing, sitting on the platform to a chair, upper limb used for
bearing. Analytic exercises for upper limb were associated for
a maximum of one third of the session.
Sessions 9 to 20: Walking in the corridor and on the steps,
control of the weight bearing and shifting, quality of the heel
strike, knee control, and waist dissociation.
During the 20 sessions, visual deprivation, head move-
ments, or training with unstable bases of support were forbid-
den for the progression of exercise difficulties.
Group 2 (Multisensorial)
Principles: Based on the maximum stimulation of the sensory
inputs needed for equilibrium, including an amount of exercise
with visual deprivation. Each type of exercise was related to the
patient’s progress, with progression under visual control for rep-
etition of the exercises, then as much under visual deprivation as
possible, and using unstable planes and foam ground-sheet, tilt-
ing the head back, rightward and leftward. The duration of the
exercises under visual deprivation was not exactly fixed and took
approximately half of the session.
Each session lasted 60 to 70 minutes, depending on the rest
required and included 5 minutes for spasticity inhibition, 30
to 35 minutes to specific modalities (see below), 10 minutes of
walking and stepping, and 10 to 20 minutes of rest distributed
throughout the session.
The exercises had to be repeated for patients to learn them
and moreover improve their performance in terms of dura-
tion or intensity by slowly increasing the difficulty.There were
4 types of modalities, conducted as follows: sessions 1 to 4,
modality 1; sessions 5 to 8, modality 2; sessions 9 to 20, by
alternating modality 3 once and modality 4 twice.
Modality 1: On the foam Bobath platform, fourfooting, stand-
ing on the knees, anteroposterior and lateral weight shifting,
moving objects with the upper limb, external destabilization.
Modality 2: Sitting on the edge of the platform and sitting
on a ball, weight shifting, upper limb movements, moving
objects with the upper limb, external destabilizations.
Modality 3: Static standing with feet together,tandem posi-
tion, one foot standing, control of weight shifting, moving
objects with upper limbs, external destabilizations.
Modality 4: Walking with movements of the upper limbs,
while speaking, with external destabilization, walking laterally
and backward, 10 minutes of treadmill training without upper
limb support, opening eyes at various speeds, closing eyes at
constant speed.
In each modality,the variations that can be used were head
movements, foam support, unstable platform, rolling skate,
irregular floor, and constant visual deprivation.
REFERENCES
1. Tutuarima JA, van der Meulen JH, de Haan RJ, van Straten A,
Limburg M. Risk factors for falls of hospitalized stroke patients.
Stroke. 1997;28:297-301.
2. Teasell R, MacRae M, Foley N, Bhardwaj A. The incidence and
consequences of falls in stroke patients during inpatient rehabil-
itation: factors associated with high risk. Arch Phys Med Rehabil.
2002;83:329-333.
3. Perennou D, El Fatimi A, Masmoudi M, et al. Incidence, circum-
stances and consequences of falls in patients undergoing rehabilita-
tion after a first stroke. Ann Readapt Med Phys. 2005;48: 138-145.
4. Forster A, Young J. Incidence and consequences of falls due to
stroke: a systematic inquiry. BMJ. 1995;311:83-85.
5. Nyberg L, Gustafson Y. Fall prediction index for patients in stroke
rehabilitation. Stroke. 1997;28:716-721.
6. Lamb SE, Ferrucci L, Volapto S, Fried LP, Guralnik JM. Risk fac-
tors for falling in home-dwelling older women with stroke: the
Women’s Health and Aging Study. Stroke. 2003;34:494-501.
7. Tyson SF, Hanley M, Chillala J, Selley A,Tallis R. The relationship
between balance, disability, and recovery after stroke: predictive
validity of the Brunel Balance Assessment. Neurorehabil Neural
Repair. 2007;21:341-346.
8. Bonan IV, Gellez-Leman MC, Legargasson JF, Guichard JP, Yelnik
AP. Evolution of subjective visual vertical perturbation after
stroke. Neurorehabil Neural Repair. 2006;20:484-491.
9. Bonan IV, Hubeaux K, Gellez-Leman MC, Guichard JP, Vicaut E,
Yelnik AP. Influence of subjective visual vertical misperception
on balance recovery after stroke. J Neurol Neurosurg Psychiatry.
2007;78:49-55.
10. Perennou DA, Amblard B, Leblond C, Pelissier J. Biased postural
vertical in humans with hemispheric cerebral lesions. Neurosci
Lett. 1998;252:75-78.
11. Bonan IV, Colle FM, Guichard JP, et al. Reliance on visual infor-
mation after stroke. Part I: balance on dynamic posturography.
Arch Phys Med Rehabil. 2004;85:268-273.
12. Bonan IV, Yelnik AP, Colle FM, et al. Reliance on visual informa-
tion after stroke. Part II: effectiveness of a balance rehabilitation
program with visual cue deprivation after stroke: a randomized
controlled trial. Arch Phys Med Rehabil. 2004;85:274-278.
13. Gauthier L, Dehaut F, Joanette Y. The bells test: a quantitative and
qualitative test for visual neglect. Int J Clin Neuropsychol.
1989;11:49-53.
14. Collin C, Wade DT. Assessing motor impairment after stroke: a
pilot reliability study. J Neurol Neurosurg Psychiatry. 1990;53:
576-576.
15. Bobath B. Hémiplégie de l’adulte, bilans et traitement. Paris,
France: Masson; 1990.
16. Stevenson TJ, Garland SJ. Standing balance during internally
produced perturbations in subjects with hemiplegia: validation
of the balance scale. Arch Phys Med Rehabil. 1996;77:656-662.
17. Clark S, Rose DJ, Fujimoto K. Generalizability of the limits of
stability test in the evaluation of dynamic balance among older
adults. Arch Phys Med Rehabil. 1997;78:1078-1084.
18. Winter DA, Patla AE, Frank JS, Walt SE. Biomechanical walking
pattern changes in the fit and healthy elderly. Phys Ther.
1990;70:340-347.
Rehabilitation of Balance After Stroke
Neurorehabilitation and Neural Repair 22(5); 2008 475
at Ecole Nationale de la Sante on December 4, 2013nnr.sagepub.comDownloaded from

Yelnik et al
476 Neurorehabilitation and Neural Repair 22(5); 2008
19. Bessou P, Dupui P, Montoya R, Pagès B. Simultaneous recording
of longitudinal displacements of both feet during human walk-
ing. J Physiol. 1989;83:102-110.
20. Granger CV, Hamilton BB. The uniform data system for medical
rehabilitation report of first admissions for 1990. Am J Phys Med
Rehabil. 1992;71:33-38.
21. Bucquet D, Condon S, Ritchie K. The French version of the
Nottingham Health Profile. A comparison of weights with those
of the source version. Soc Sci Med. 1990;30:829-835.
22. Noether GE. Sample size determination for some common non-
parametric statistics. J Am Stat Assoc. 1987;82:645-647.
23. Nashner LM, Peters JF. Dynamic posturography in the diagnosis
and management of dizziness and balance disorders. Neurol Clin.
1990;8:331-349.
24. Azulay JP, Mesure S, Amblard B, Pouget J. Increased visual
dependence in Parkinson’s disease. Percept Mot Skills. 2002;95:
1106-1114.
25. Guerraz M, Yardley L, Bertholon P,et al.Visual vertigo: symptom
assessment, spatial orientation and postural control. Brain.
2001;124:1646-1656.
26. Suarez H, Muse P, Suarez A, Arocena M. Assessment of the risk of
fall, related to visual stimulation, in patients with central vestibu-
lar disorders. Acta Otolaryngol. 2001;121:220-224.
27. Lord SR, Webster IW. Visual field dependence in elderly fallers
and non-fallers. Int J Aging Hum Dev. 1990;31:267-277.
28. Yelnik AP, Kassouha A, Bonan IV, et al. Postural visual dependence
after recent stroke: assessment by optokinetic stimulation. Gait
Posture. 2006;24:262-269.
29. Winstein CJ, Gradner ER, McNeal DR, Barto PS, Nicholson DE.
Standing balance training: effect on balance and locomotion in
hemiparetic adults. Arch Phys Med Rehabil. 1989;70:755-762.
30. Lee MY, Wong M K, Tang FT. Clinical evaluation of a new
biofeedback standing balance training device. J Med Eng Technol.
1996;20:60-66.
31. Sackley CM, Lincoln NB. Single blind randomized controlled
trial of visual feedback after stroke: effects on stance symmetry
and function. Disabil Rehabil. 1997;12:536-546.
32. Kerdoncuff V, Duruffe A, Petrilli S, et al. Interest of visual
biofeedback training in rehabilitation of balance after stroke.
Ann Readapt Med Phys. 2004;47:169-176.
33. de Haart M, Geurts AC, Dault MC, Nienhuis B, Duysens J.
Restoration of weight-shifting capacity in patients with posta-
cute stroke: a rehabilitation cohort study. Arch Phys Med Rehabil.
2005;86:755-762.
34. Cheng PT,Wang CM, Chung CY, Chen CL.Effects of visual feed-
back rhythmic weight-shift training on hemiplegic stroke
patients. Clin Rehabil. 2004;18:747-753.
35. Barclay-Goddard R, Stevenson T, Poluha W, Moffatt MEK,
Taback SP. Force platform feedback for standing balance training
after stroke. Cochrane Database Syst Rev. 2004;(4):CD004129.
36. Van Peppen RPS, Kortsmit M, Lindeman E, Kwakkel G. Effects of
visual feedback therapy on postural control in bilateral standing
after stroke: a systematic review. J Rehabil Med. 2006;38:3-9.
37. Baccini M, Paci M,Rinaldi L.The scale for contraversive pushing:
a reliability and validity study. Neurorehabil Neural Repair.
2006;20:468-472.
38. Dobkin B. Confounders in rehabilitation trials of task-oriented
training. Neurorehabil Neural Repair. 2007;21:3-13.
at Ecole Nationale de la Sante on December 4, 2013nnr.sagepub.comDownloaded from