8. Effect of concurrent cognitive tasks on gait features among children post-severe traumatic brain injury and typically-developed controls

Prévia do material em texto

Brain Injury, June 2011; 25(6): 581–586
Effect of concurrent cognitive tasks on gait features among children
post-severe traumatic brain injury and typically-developed controls
MICHAL KATZ-LEURER1,2, HEMDA ROTEM2, OFER KEREN3, & SHIRLEY MEYER2
1Sackler Faculty of Medicine, School of Health Professions, Physical Therapy Department, Tel-Aviv University, Israel,
2Alyn Children’s Orthopedic Hospital and Rehabilitation Center, Jerusalem, Israel, and 3Acquired Brain Injury,
Rehabilitation Department Sheba Medical Center, Sackler Faculty of Medicine, Tel Aviv University, Israel
(Received 23 April 2010; revised 24 February 2011; accepted 14 March 2011)
Abstract
Aim: To investigate and compare the influence of concurrent cognitive tasks on gait characteristics in children
post-traumatic brain injury (TBI) and typically-developed (TD) controls.
Methods: Fifteen children post-TBI (aged 9.5� 2.2 years) and 15 TD controls (aged 9.9� 1.3 years) were included in the
study. The children were asked to walk under three conditions: (1) walking at a self-selected speed, (2) walking at a self-
selected speed while memorizing and recalling a series of three random numbers and (3) walking at a self-selected speed
while listening and identifying commonly experienced sounds. Gait parameters include walking speed, step time and length
and step time and length variability as measured with the GAITRite� system.
Results: Introduction of a concurrent task led to decreased walking speed and step length and increased step time and step
length variability in both groups, but with a significantly prominent effect in children post-TBI. The results also showed that
the effect of a concurrent cognitive task on walking depended on the complexity level of the task.
Conclusion: When children were asked to perform an additional task while they were walking, this had a negative effect on
their gait. One may deduce from these results that control of the rhythmic stepping mechanism at a self-selected walking
speed in children is dependant to an extent on their ability to focus attention on their gait. In children post-TBI who usually
have an attention deficit anyway, this ability is severely disturbed.
Keywords: Concurrent tasks, gait characteristics, traumatic brain injury (TBI)
Introduction
Many situations in daily activities involve concurrent
cognitive and motor tasks. The ability to divide
attention and the capacity to process information
regarding performance of two or more tasks simul-
taneously is essential [1]. For example, walking is a
basic daily activity which is often accompanied by
various types of concurrent cognitive as well as
motor tasks. Since attention is a limited resource it
may become overloaded by competing attention
demands and subsequently might lead to reduced
performance in one or all tasks [2, 3]. Dual-task
paradigms are classically used to study involvement
of attention in gait control [4]. The paradigms
combining gait and cognitive tasks have shown that
walking is not purely automatic but requires atten-
tion [5–7]. Among healthy adults of all ages as well
as among people with a neurological deficit the
influence of cognitive activity on gait consistently
exhibited a decreased gait velocity and increased gait
variability [8–12].
Following severe traumatic brain injury (TBI)
children may show deficits in a wide range of
cognitive and physical activities. Walking perfor-
mance is often characterized by a high level of
variability associated with poor balance, as compared
Correspondence: Michal Katz-Leurer, Sackler Faculty of Medicine, School of Health Professions, Physical Therapy Department, Tel-Aviv University,
Tel-Aviv 69978, Israel. Tel: 972-3-6405432. Fax: 972-3-6409223. E-mail: michalkz@post.tau.ac.il
ISSN 0269–9052 print/ISSN 1362–301X online � 2011 Informa UK Ltd.
DOI: 10.3109/02699052.2011.572943
to typically-developed controls [13]. As increased
walking variability may imply less efficient automatic
control, children post-TBI may need more attention
to walking and as a result might be more vulnerable
to the effect of a concurrent task.
In addition to their motor impairments, children
with TBI often have deficits in executive function,
memory, with increased distractibility or even per-
severation, all of which demand an even higher
attentive input than normal children during any task
performance. Therefore, the addition of a cognitive
task to walking may alter gait parameters.
Huang et al. [12] found that the effect of a
concurrent task on gait was task-related. Comparing
the effects of different cognitive tasks (visual iden-
tification, auditory identification and memorization)
on gait variables in typically developing 5–7-year-old
children reveals that the auditory identification task
caused the greatest interference in gait speed,
cadence and step length, while memorization
caused the least interference.
It may therefore be hypothesized that a more
complex task such as auditory identification would
cause significant attention-related gait changes as
compared to a memorization task in children
post-severe TBI. The aim of this study is to
investigate and compare the influence of concurrent
cognitive tasks on gait characteristics in children
with post-TBI and typically-developed (TD) con-
trols. There is little information on the effect of a
concurrent task on walking in these children. Since
walking is a basic daily activity and is often accom-
panied by various types of concurrent tasks, under-
standing the effect of a concurrent task on walking
should have significant implications for clinicians
treating children with post-severe TBI.
Methods
Subjects
Fifteen children post-TBI who were either outpa-
tients or former patients of the Alyn Children’s
Rehabilitation Hospital in Jerusalem, Israel were
engaged in the study. Included were children post-
severe closed head injury (Glasgow Coma Scale
(GCS) score at admission to the emergency room of
�8) [14], at least 1 year post-trauma, aged 7–13
years who were independent ambulators (foot ortho-
ses permitted). Excluded were those who had
received botulinum toxin for spasticity management
or had undergone orthopaedic surgery in the previ-
ous 6 months and those who were unable to follow
simple instructions.
Fifteen TD children matched for age served as
controls. The TBI group comprised nine boys and
six girls, aged 9.5� 2.2 years. The time since injury
was a mean of 3.5 years (range 1.5–7 years); seven of
the children were classified as double hemiparesis,
eight as hemiparesis. In the TD group there were 10
boys and five girls, aged 9.9�1.3 years.
Informed consent was obtained from each partic-
ipant and his/her parents before inclusion in the
study. The study protocol was approved by the
hospital’s institutional ethical committee.
Gait assessment
Parameters of gait were recorded using a pressure
sensitive mat, the ‘Gaitrite1’ system (Gaitrite�, CIR
Systems Inc., Clifton, NJ) which is an electronic
walkway that automates the collection of spatial and
temporal parameters of gait. The validity of the
Gaitrite system has been supported by studies in
adults [15]. Thorpe et al. [16] recently showed that
the Gaitrite� system is a reliable method of measur-
ing spatiotemporal gait parameters in children as
well.
For each test the children were asked to walk
along the mat at their regular speed. ‘Stop’ signs
were positioned on the floor 2 metres beyond each
end of the walkway, providing visual feedback to
subjects. Each cycle of walking was �7.66 metres, of
which 3.66 metres were on the mat. Each subject
completed three sequential cycles of walking for a
total of almost 11 metres on the mat. Testing was
performed while the children wore comfortable
clothes and their regular shoes and orthotics when
needed. Distractions in the room were minimized.
The investigator and thechild’s parent were present
during testing.
All temporal and spatial gait parameters were
calculated utilizing the software of the Gaitrite
system using pre-programmed definitions, calcula-
tions and data from the three cycles. Parameters
analysed included walking velocity, step length and
step length variability, step time and step time
variability [16]. Any partial footfalls that did not
have a clearly defined beginning and ending or were
in contact with the edge of the mat were edited out.
Gait variability was expressed as the coefficient of
variation (CV) which is SD/mean �100.
Procedure
After a rest period the children were asked to walk at
their regular pace on the Gaitrite, three cycles for
each one of the trials: regular walking followed by
performing the two different cognitive tasks—the
first was to memorize and recall three random
numbers one digit each and the second was to
identify simple auditory sounds.
In a pre-test period children with post-severe TBI
were assessed for their ability to memorize and recall
three numbers consecutively. It was noticed that the
582 M. Katz-Leurer et al.
children could perform the assignment correctly and
sequentially only with one digit numbers. So for this
study, the children were asked to recall three random
numbers of one digit each. In addition, in a pre-test
period children with post-severe TBI were asked to
identify different sounds. The sounds chosen were
clear and simple as commonly experienced by
children (such as dog barking, doorbell ringing)
and included a total of 15 sounds that the children
could identify without mistakes.
After three walking cycles with no concurrent task,
the children were asked to remember the numbers
presented to them and to repeat the numbers over
and over in exactly the same order. All the children
could perform the task with no mistakes. Then the
children were asked to walk while repeating the
numbers. Correct and incorrect repetitions of num-
bers were counted and the incorrect repetitions out
of the total number were recorded. In the second
cognitive task the sounds were first projected
through a speaker at the far end of the walkway.
All of the sounds were less than 2 seconds in
duration and the children were asked to identify each
as quickly as possible by naming the sound out loud.
Then the children were asked to walk while listening
and identifying each sound as quickly as possible.
The number of sounds that each child identified
incorrectly out of the total of 15 possible sounds was
recorded.
Statistical analysis
A preliminary paired t-test was used to assess
differences in gait parameters between legs. No
significant differences were noted; therefore, the
mean value of both legs is presented. The normality
of the parameters’ distribution was verified using the
Kolmagirove Smirnove test. The values of gait
velocity, step time, step length and step time and
length variability were summarized using means and
standard deviations.
To assess differences between groups in the
fulfilling of the cognitive assignment—number of
cycles achieved in each task, a t-test was performed.
To evaluate any different effect of the task on
performance quality, a mixed ANOVA model was
performed. The between-subject factor is the group
factor (TBI vs TD), the within-subject factor is
the two cognitive tasks and the interaction term
(group� task) and the dependent variable was the
percentage error in each assignment. For assessing
any differences between the three walking conditions
without and with concurrent tasks, a mixed ANOVA
model was applied when the dependent variable was
the gait parameter. A separate model was built for
each parameter. The between-subject factor is the
group factor (TBI vs TD), within subject factor
(without and with the first and the second concur-
rent tasks) and the interaction term (group� task).
A p-value< 0.05 was considered statistically signif-
icant. All statistical analyses were performed using
SPSS v. 17 software packages.
Results
Cognitive tasks–percentage error
Both groups exhibited significantly higher mistakes
in the sound recognition assignment. The mean
percentage errors between the two cognitive tasks
was significantly different between groups; by mean,
children post-severe TBI have 31.9% more mistakes
compared to the controls (F(1: 28)¼29.6, p<0.01);
23%�5% more mistakes at the memorization
assignment (mean of 24.5% among children
post-TBI vs 1.5% among the controls) and
39%�6% more at the sound recognition assign-
ment as compared to the TD group (mean of 46.3%
among children post-TBI vs 6.7% among the con-
trols) there was a significant group*assignment
interaction effect (F(1: 28)¼ 9.45, p< 0.01), as chil-
dren post-TBI exhibited larger differences between
assignments (TBI; 20%�16%, TD 5%�7%)
(Table I).
Baseline differences between groups
Gait parameters during all three gait conditions are
presented in Table II. In the regular walking test,
only the step length was significantly different
between groups; children with post-severe TBI
walked with a smaller step length as compared with
TD controls (mean of 59.1� 10.1 cm vs
66.1�6.1 cm, respectively, t28¼ 2.34. p¼ 0.02).
Gait parameters during cognitive assignments
In the repeated model, significant differences were
noted between groups in all assessed gait parameters.
Children post-TBI exhibit significant slower gait
velocity, significant shorter step length and signifi-
cantly larger step length and step time variability.
Table I. Number of cycles and percentage of errors in task
performance by groups.
Task
Numbers Sounds
TBI TD TBI TD
No. of cycles 11.0� 4.3 16.6� 2.6 17.6� 2.6 14.7�2.0
% of error 24.5� 22.9 1.5� 3.2 46.3� 22.9 6.7�5.3
Values are mean�SD.
Effect of concurrent cognitive tasks on gait 583
There were significant differences in gait param-
eters between assignments in both groups.
Gait velocity
Gait velocity was significantly different between
trials (F2;50¼38.1, p< 0.01). After bonferroni
adjustment for multiple compression it was noted
that walking velocity was significantly faster at usual
walking and it decreased by a mean of
23.8� 3.9 cm s�1 at the number assignment (mem-
orization) and by 37.0�5.1 cm s�1 at the sound
(auditory) assignment. In addition, a significant
difference was noted between dual task conditions;
during the sound assignment children walked signif-
icantly slower as compared to the walking velocity
during the number assignment (by a mean of
13.1� 3.7 cm s�1, p< 0.01). Additionally, there
was a significant group-assignment interaction
effect; gait velocity decreased for both groups in
the dual task conditions, but significantly more
among children post-TBI. In the number assign-
ment, gait velocity decreased by 36.5� 6.9 cm s�1
among children post-TBI as compared to
8.3� 3.6 cm s�1 among TD children (p< 0.01).
In the sound assignment walking velocity decreased
by 48.7� 7.9 cm s�1 among children post-TBI as
compared to 25.3� 6.5 cm s�1 among TD children
(p¼0.03).
Step time
Step time was significantly different between trials
(F2;50¼ 23.42, p<0.01). Mean step time was sig-
nificantly increased in the dual task conditions as
compared to baseline; it increased by a mean of
0.10� 0.02 seconds in the number assignment and
by 0.10� 0.01 seconds in the sound assignment. No
significant difference was noted between dual task
conditions. A significant group-assignment interac-
tion effect was noted also for the step time param-
eters. Step time increased during dual task
conditions in both groups but significantly more
among children post TBI. In the number assign-
ment, step time increased among children post-TBI
by 0.17�0.15 seconds as compared to 0.03� 0.04
seconds among TD children (p< 0.01). In the sound
assignment, step time increased by 0.16� 0.10
seconds among childrenpost-TBI as compared to
an increase of 0.05� 0.03 seconds among TD
children (p<0.01).
Step length
In addition, step length was significantly different
between trials (F2;50¼ 13.9, p<0.01). Step length
was significantly shorter at the sound assignment. It
was shorter by a mean of 9.48� 2.2 cm as compared
to the baseline value and by 5.9� 1.1 cm as com-
pared to number assignment.
Step length and step time variability
Step length and step time variability were signifi-
cantly higher among children post-TBI across the
assignments, with no significant difference between
the assignments (step length variability, mean
between groups difference 12.2� 2.2, F1;25¼ 8.3,
p< 0.01, step time variability mean between groups
difference 7.9� 2.6, F1;25¼ 9.2, p< 0.01).
Discussion
The present study examined the effects of concur-
rent cognitive tasks on gait performance in children
post-TBI in comparison to TD controls. There was a
negative effect on walking performance when the
children did a concurrent task and this effect was
task-dependant. During the auditory assignment, the
performance error rate was significantly higher as
compared to the memorization assignment and the
effect on walking was more detrimental: walking
speed decreased and step time increased significantly
in the auditory assignment in both groups but with a
Table II. Gait parameters in children post-TBI and TD controls in usual walking and in the two dual task conditions.
Usual walking Dual task- numbers Dual task- sounds
^p-value #p-valueTBI TD *p-value TBI TD TBI TD
Velocity 1.2�0.3 1.3� 0.1 0.15 0.8� 0.2 1.3� 0.2 0.7� 0.1 1.1� 0.2 <0.01 <0.01
Step length 59.1�10.1 66.1� 6.1 0.02 51.2� 11.1 65.4� 5.7 43.9� 11.7 58.7� 8.0 <0.01 0.58
Step length variability 9.3�9.1 8.2� 10.2 0.27 13.4� 14.6 5.4� 3.0 24.0� 17.0 8.4� 5.7 0.12 0.53
Step time 0.45�0.07 0.48� 0.05 0.55 0.60� 0.11 0.51� 0.07 0.61� 0.09 0.53� 0.06 <0.01 <0.01
Step time variability 11.5�15.4 6.8� 7.6 0.23 13.6� 10.1 6.8� 4.8 16.2� 7.8 8.6� 9.2 0.15 0.80
*p-value between groups at usual walking (t-test).
^p-value within subjects (usual walking, numbers, sounds) (Repeated measure ANOVA).
#p-value interaction effect group and manipulation (Repeated measure ANOVA).
584 M. Katz-Leurer et al.
significantly prominent effect among children
post-TBI.
The interference effects of concurrent cognitive
tasks on locomotor performance has been assessed
and described before among children as well as
among adults [12, 17, 18]. It has been suggested that
the outcome of dividing attention between two
concurrent tasks depends on the type and level of
difficulty of the tasks and on the priority given to
each task [3, 4]. Since attention resources are
limited, two tasks will compete for resources if they
are carried out at the same time. They may interfere
with each other when the total amount of resources
demanded exceeds those available.
In the present study, the auditory assignment
caused greater interference than the memorization
assignment. It might be that memorizing a set of
three numbers is a simple memory task which
requires short-term learning and recall as compared
to the sound recognition assignment which involves
identification, recall of existing knowledge and
naming capabilities. The auditory assignment is
clearly more demanding and requires more attention
resources [4, 5, 18]. In previous studies, the negative
effect on locomotor speed while performing second-
ary tasks involving processing of sensory stimuli was
higher than for those involving other types of
cognitive processing. Specifically, it has been noted
that an auditory identification task produced greatest
interference effects on gait than other concurring
tasks [12].
One may claim that the lower effect of the
concurrent cognitive task on gait characteristics in
TD children is due to the fact that the cognitive task
was too simple for them. It has been suggested
before that it is essential to adjust the task difficulties
to the individual’s capacity [19]. However, it has also
been noted that even with a very light overall load on
the cognitive system a difference between patients
with brain damage and healthy controls could be
detected [20]. In this study the memorization and
the auditory assignments were adjusted to the
capabilities of children post-severe TBI. By doing
so one may reduce any significant baseline differ-
ences between groups, but it is argued that any
reduced ability in performance is the result of the
demand to combine performances.
One of the most frequent cognitive limitations
among children post-TBI is an attention distur-
bance. Performing dual tasks demands splitting
attention simultaneously between two tasks and
each assignment might be too difficult to perform
in those whose attention ability is reduced. It has
been well established that when combining cognitive
and walking tasks, the walking requires attention
[21] and dual tasking is sensitive enough to distin-
guish between patients post-concussion from control
subjects [22, 23]. The interactions between concur-
rent tasks cannot be automatically generalized and
should be assessed in different conditions and
modalities. This may imply that the attention
needed for motor and cognitive activities are not
necessarily always competitive and interruptive.
Step variability increases when step velocity
decreases [24]. Therefore, the dual-task related
increase in step length variability shown in this
study could also be explained by the dual-task
related decrease in step velocity. However, the
results of this study show that the decrease in step
velocity while performing the memorization task did
not modify step variability. During the auditory
assignment the mean step time was the same as for
the memorization assignment, but step variability
increased significantly in both groups, more so in the
children post-TBI. This suggests that the increase in
step variability was task-specific, in contrast to
step time.
These findings have clinical implications for the
physiotherapist. Firstly, therapists use a lot of verbal
instructions and feedback during the actual treat-
ment. Therefore, a child post-severe TBI who is
practicing a new and higher level of motor perfor-
mance will benefit more from instructions or feed-
back given before or after performing the task. On
the other hand, if the child needs further refinement
of an achieved motor skill and the therapist indeed
wants to challenge the child, the therapist can
provide concurrent feedback or add a secondary
cognitive task. Another clinical implication is related
to the finding that step variability increases during
dual task performance. It has been noted before that
step variability is associated with balance
performance among children post-TBI [13],
a higher step variability being associated with poor
balance.
Possible, the therapist should work on improving
balance as much as possible before progressing to
higher motor skills. However, Silsupadol et al. [25]
described two adults patients who received balance
training under dual task conditions. Benefits were
maintained over 3 months, whereas they were not
evident in a patient who trained under a single task
balance training programme. The question of
whether training under dual task conditions is
more effective in improving gait after the patient
has progressed in the single task performance needs
further investigation.
The therapist might instruct the family of a child
post-TBI to avoid cognitive tasks in the early stages
of motor training. Dual tasking and adding cognitive
activities while walking should, however, be intro-
duced as an effort to improve the child’s ability to
perform dual tasks in a safe and functional manner.
Effect of concurrent cognitive tasks on gait 585
Declaration of Interest: The authors report
no conflicts of interest.The authors alone are
responsible for the content and writing of the paper.
References
1. Schmidt RA. Motor control and learning: A behavioral
emphasis. 2nd ed. Champaign, IL: Human Kinetics
Publishers; 1988.
2. Catena R, van Donkelaar P, Chou LS. Different gait tasks
distinguish immediate vs. long-term effects of concussion on
balance control. Journal of Neuro Engineering and
Rehabilitation 2009;6:25–32.
3. Pashler H. Dual-task interference in simple task: Data and
theory. Psychological Bulletin 1994;116:220–244.
4. Beauchet O, Berrut G. Gait and dual-task: Definition,
interests, and perspectives in the elderly. Psychol
Neuropsychiatr Vieil 2006;4:215–225.
5. Beauchet O, Dubost V, Herrmann FR, Kressig RW.
Stride-to-stride variability while backward counting among
healthy young adults. Journal of NeuroEngineering
Rehabilitation 2005;11:26–43.
6. Dubost V, Kressig RW, Gonthier R, Herrmann FR,
Aminian K, Najafi B, Beauchet O. Relationships between
dual-task related changes in stride velocity and stride time
variability in healthy older adults. Human Movement Science
2006;25:372–382.
7. Woollacott M, Shumway-Cook A. Attention and the control
of posture and gait: A review of an emerging area of research.
Gait Posture 2002;16:1–14.
8. Sheridan PL, Solomont J, Kowall N, Hausdorff JM.
Influence of executive function on locomotor function:
Divided attention increases gait variability in Alzheimer’s
disease. Journal of the American Geriatric Society 2003;51:
1633–1637.
9. Yogev G, Giladi N, Peretz C, Springer S, Simon ES,
Hausdorff JM. Dual tasking, gait rhythmicity and
Parkinson’s disease: Which aspects of gait are attention
demanding? European Journal of Neuroscience 2005;22:
1248–1256.
10. Dubost V, Annweiler C, Aminian K, Najafi B, Herrmann FR,
Beauchet O. Stride-to-stride variability while enumerating
animal names among healthy young adults: Result of stride
velocity or effect of attention-demanding task? Gait Posture
2008 27:138–143.
11. Cherng RJ, Liang LY, Hwang IS, Chen JY. The effect of a
concurrent task on the walking performance of preschool
children. Gait Posture 2007;26:231–237.
12. Huang HJ, Mercer VS, Thorpe DE. Effects of different
concurrent cognitive tasks on temporal-distance gait
variables in children. Pediatric Physical Therapy 2003;15:
105–113.
13. Katz-Leurer M, Rotem H, Lewitus H, Keren O, Meyer S.
Relationship between balance abilities and gait characteristics
in children with post-traumatic brain injury. Brain Injury
2008;22:153–159.
14. Saatman KE, Duhaime A-C, Bullock R, Maas AIR,
Valadka A, Manley GT. Classification of traumatic brain
injury for targeted therapies. Journal of Neurotrauma
2008;25:719–378.
15. Mcdonough AL, Batavia M, Chen FC, Kwon S, Ziai J.
The validity and reliability of the gaitrite system’s measure-
ments: A preliminary evaluation. Archives of Physical
Medicine and Rehabilitation 2001;82:419–425.
16. Thorpe DE, Dusing SC, Moore CG. Repeatability of
temporospatial gait measures in children using the gaitrite
electronic walkway. Archives of Physical Medicine and
Rehabilitation 2005;86:2342–2346.
17. Whitall J. The developmental effect of concurrent cognitive
and locomotor skills: Time-sharing from a dynamic perspec-
tive. Journal of Experimental Child Psychology 1991;51:
245–266.
18. Smith EE, Geva Jonides J, Miller A, Reuter-Lorenz P,
Koeppe RA. The neural basis of task-switching in working
memory: Effects of performance and aging. Proceedings of
the National Academy of Sciences (USA) 2001;98:
2095–2100.
19. Cocchini G, Della Sala S, Logie RH, Pagani R, Sacco L,
Spinnler H. Dual task effects of walking when
talking in Alzheimer’s disease. Rev Neurol (Paris)
2004;160:74–80.
20. Logie RH, Cocchini G, Delia Sala S, Baddeley AD. Is there a
specific executive capacity for dual task coordination?
Evidence from Alzheimer’s disease. Neuropsychology
2004;18:504–513.
21. Lajoie Y, Teasdale N, Bard C, Fleury M. Attentional
demands for static and dynamic equilibrium. Experimental.
Brain Research 1993;97:139–144.
22. Parker TM, Osternig LR, Lee HJ, Donkelaar P, Chou LS.
The effect of divided attention on gait stability
following concussion. Clinical Biomechanics 2005;20:
389–395.
23. Catena RD, van Donkelaar P, Chou LS. Cognitive task
effects on gait stability following concussion. Experimental
Brain Research 2007;176:23–31.
24. Heiderscheit BC. Movement variability as a clinical measure
for locomotion. Journal of Applied Biomechanics 2000;16:
419–427.
25. Silsupadol P, Siu KC, Shumway-Cook A, Woollacott MH.
Training of balance under single- and dual-task conditions in
older adults with balance impairment. Physical Therapy
2006;86:269–281.
586 M. Katz-Leurer et al.
Copyright of Brain Injury is the property of Taylor & Francis Ltd and its content may not be copied or emailed
to multiple sites or posted to a listserv without the copyright holder's express written permission. However,
users may print, download, or email articles for individual use.