J Head Trauma Rehabil

Vol. 25, No. 3, pp. 155–163

Copyright 
c 2010 Wolters Kluwer Health | Lippincott Williams & Wilkins

Balance, Attention, and Dual-Task

Performance During Walking After
Brain Injury: Associations
With Falls History
Karen L. McCulloch, PT, PhD, NCS; Elizabeth Buxton, DPT, PT;
Jessica Hackney, DPT, PT; Sean Lowers, DPT, PT

Objective: To examine the relationship between balance, attention, and dual-task performance in individuals with
acquired brain injury. Design: Cross-sectional study. Setting: Rehabilitation center and supported living program.
Participants: Twenty-four individuals aged 18 to 58 years (mean = 39 years) with acquired brain injury who were
able to ambulate 40 ft with (29%) or without an assistive device. Fifty-eight percent were independent community
ambulators. Fifty-four percent had fallen in the past 6 months; and 42% reported feeling unsteady with standing
or walking. Interventions: Participants completed a battery of balance, attention, and dual-task assessments. Main
Outcome Measures: Balance: Berg Balance Scale (BBS), Four Square Step Test (FSST), High Level Mobility As-
sessment Test (HiMAT); Attention: Symbol Digit Modalities Test (SDMT), Moss Attention Rating Scale (MARS),
modified for a single test session; and a walking dual-task assessment, the Walking and Remembering Test. Results:
Mean scores: BBS, 48 of 56; FSST, 19.6 seconds; HiMAT, 20 of 54; SDMT, 30 correct; and MARS, 80. Dual-task
costs were observed with variable patterns across subjects: 48% demonstrated primarily motor slowing, 9% had
reduced cognitive accuracy without motor slowing, and 35% demonstrated decrements in both tasks. Subjects with
a falls history had more impaired balance (HiMAT, BBS, and FSST, all P <.026) but were not significantly different
in dual-task performance or attention measures. Conclusions: The test battery matched the range of motor and
cognitive abilities of the sample. Balance was more strongly related to falls history than measures of attention or
dual-task performance. Injury chronicity may have allowed some subjects to develop strategies to optimize dual-task
performance. Alternatively, motor slowing in dual-task conditions may be an adaptive strategy, allowing performance
of multiple tasks with reduced safety risk. Further investigation in this area is warranted to clarify the utility of dual-
task methods in identifying falls risk after brain injury. Keywords: acquired brain injury, balance, dual-task performance,
falls risk

I NDIVIDUALS WITH ACQUIRED BRAIN IN-

JURY (ABI) often have cognitive and balance im-
pairments that could lead to falls risk. Tasks that were
automatic and required little thought prior to brain in-
jury, such as standing or walking, become more attention
demanding. Balance and safety may be compromised in
circumstances that require focus on a secondary task,
requiring dual (or multiple)-task performance. Difficul-
ties in dual-task conditions are associated with a his-
tory of falls and risk of future falls in institutionalized
and community-dwelling older adults1 and survivors of
stroke.2 Individuals with brain injury often experience
similar cognitive and balance deficits to those seen in
older adults and after stroke that are linked to falls risk.
The risk of falls is a significant concern after traumatic
brain injury (TBI) related to secondary injury.3 The pur-
pose of this study was to examine relationships between
attention, balance, and dual-task performance in indi-
viduals recovering from ABI and to explore associations
of these factors with falls history.

Author Affiliations: Division of Physical Therapy, Department of Allied
Health Sciences, School of Medicine, University of North Carolina (UNC),
Chapel Hill (Dr McCulloch); UNC Healthcare, Chapel Hill (Dr Buxton);
Carolinas Medical Center, Charlotte (Dr Hackney); and Duke University
Medical Center, Durham (Dr Lowers), North Carolina.
Corresponding Author: Karen L. McCulloch, PT, PhD, NCS, Division
of Physical Therapy, CB 7135 Bondurant Hall, Ste 3024, University of
North Carolina, Chapel Hill School of Medicine, Chapel Hill, NC 27599
(karen mcculloch@med.unc.edu).
ATTENTION AND ATTENTION ALLOCATION
Attention can be thought of as the ability to focus cog-
nitive resources and to selectively process certain infor-
mation from the environment while ignoring irrelevant
input.4 Depending on the task context, attention can be
described as selective, divided, sustained, or switching.4
Selective attention tasks require focus on relevant
155
156 JOURNAL OF HEAD TRAUMA REHABILITATION/MAY–JUNE 2010
information for conscious processing while disregarding
extraneous information.5 The concept of divided at-
tention is related, requiring processing of more than 1
source of information at a time or performing more than
1 task at a time.4 When a person lacks the attentional ca-
pacity to multitask, performance on 1 or more tasks may
suffer.4 When one of the tasks is a mobility task, safety
may be at risk.
The role of executive function in controlling atten-
tion has been described in several models, including the
model of visual attention of Posner6 and the model of
working memory of Baddeley.7–10 Executive function co-
ordinates allocation of attention to different tasks during
daily life, allowing choices to be made about use or stor-
age of information and allowing division of attention be-
tween tasks if necessary. Traditionally, executive function
deficits have been associated with damage to the dorso-
lateral prefrontal cortex that can result from brain injury
and can also occur in patients with frontal-subcortical
infarcts.11–13 Executive function deficits may cause dif-
ficulty with appropriate attention allocation to multiple
tasks.14–16 Likewise, damage to other brain structures in-
volved in alerting and orienting functions may result in
decreased attentional capacity.6
Slowed information processing is a hallmark of TBI.
Schneider and Shiffrin17 described 2 fundamental pro-
cessing modes: controlled and automatic. Controlled
information processing requires greater attentional de-
mand and is highly dependent on load, whereas auto-
matic processing is well learned, requires less attention,
and is virtually independent of load. After TBI, previ-
ously automatic tasks may require more controlled in-
formation processing, necessitating greater attention and
effort.18
DUAL-TASK PARADIGM AND
DUAL-TASK CONDITIONS
The presence of attention, executive function, and
information-processing deficits coupled with an in-
creased need for controlled processing following brain
injury contributes to possible limitations in dual-task
performance when more than 2 tasks are attempted at
once. Dual-task performance can be measured in the
dual-task paradigm described by Abernethy,19 in which 2
tasks are performed separately to establish baseline abili-
ties (single task). Both tasks are then performed together,
allowing the performance of each to be measured and
compared with single-task performance. Decrements in
performance are described as relative dual-task costs
(DTCs) and can be expressed as a percentage. Relative
DTC values can be compared across subjects who have
different baseline abilities in single-task conditions.18,19
In this study, walking with a cognitive task was intro-
duced and allowed to vary as natural inclinations dic-
tated. This use of dual-task conditions without a strict
experimental paradigm allowed examination of behav-
ioral change in one or both tasks from a more functional
perspective. Comparisons of the motor DTC and cog-
nitive DTC for each subject illustrate whether DTCs are
observed for the motor activity, cognitive activity, both,
or neither task.
DUAL-TASK PERFORMANCE DEFICITS AFTER
STROKE AND CONCUSSION
Investigators have examined dual-task performance
difficulty in community dwellers after stroke and mixed
samples of individuals with ABI, suggesting that atten-
tion deficits are common and correlated with poor per-
formance on balance measures and increased risk of
falls.2,20,21 Commonly, dual-task conditions result in
slower walking and poorer cognitive task performance.
The use of dual-task measures has been recommended
as a more sensitive test of functional performance after
stroke than testing balance alone.22
Similar, but more subtle, dual-task performance issues
have also been described for individuals recovering from
mild TBI and concussion. Individuals recovering from
concussion demonstrate increased sway during walking
than age-matched controls23,24 and reduced gait speed
during dual-task conditions that involve more challeng-
ing cognitive tasks.24 Gait instability after concussion
is a deficit that may be best detected through the use
of dual-task conditions.25,26 Challenging cognitive tasks
versus simple reaction time tasks,27 and obstacle avoid-
ance tasks,25,27 have identified subtle dual-task perfor-
mance deficits after concussion that are missed if balance
or cognitive tasks are examined separately. The ability to
navigate obstacles efficiently in dual-task conditions is
impaired following concussion28 and may have implica-
tions for falls risk and further injury in mild and more
severe brain injury.
In this study, we examined attention, balance, and
dual-task performance in a group of individuals after
brain injury, exploring associations with falls history. We
expected that reduced dual-task performance in walk-
ing would be associated with impairments of attention
and balance. We anticipated deficits in dual-task perfor-
mance, balance, and attention would also be associated
with reported history of falls in the past 6 months.
METHODS
Participants
Participants were 24 subjects recovering from ABI.
Subjects were recruited from both a facility that provides
long-term and short-term care for individuals recovering
from ABI and an inpatient rehabilitation center. A physi-
cal therapist at each site identified potential subjects that
 Balance, Attention, and Dual-Task Performance During Walking After Brain Injury 157

met the inclusion criteria and might be interested in TABLE 1 Participant descriptives(N = 24)
participating in the study. Potential subjects (or the ap-
propriate legal representative) were contacted by a mem-
ber of the research team to screen for inclusion. In accor- Characteristic Mean (SD)
dance with a University of North Carolina–Chapel Hill Age, y 39.4 (13.3)
Biomedical institutional review board–approved proto- Onset of injury, mo 117.8 (125.2)
col, consent was obtained from the subject or legal rep- Number (%)
resentative; assent was also obtained from subjects un- Sex (male) 18 (75)
Cause of acquired brain injury
able to self-consent. All subjects met the following inclu- High-velocity trauma (motor 12 (50)
sion criteria: survivor of ABI; able to ambulate indepen- vehicle crash)
dently with or without an assistive device a distance of 40 Fall 9 (37.5)
ft without resting; and able to comprehend and follow Other 3 (12.5)
verbal instructions in English. Individuals who were un- Assistive device use
None 16 (67)
able to ambulate independently were excluded, as were Rolling walker 3 (12)
individuals whose cognitive status would not allow for Straight cane 5 (21)
participation in the testing protocol. Community mobility level
Physiological walker 0
Procedure Limited household walker 1 (4)
Unlimited household walker 2 (8)
In a single test session, participants were interviewed to Most limited community walker 2 (8)
determine demographics, date and mechanism of injury, Least limited community walker 4 (16)
other medical conditions, current living situation, use of Community walker 14 (58)
assistive devices or orthoses, and falls history in the past Living environment
24-h supervised group 20 (83)
6 months. Because many of our participants had cogni-
home/apartment
tive impairment, we confirmed all reports of falls history Apartment with supported living 1 (4)
by discussing responses with the treating physical ther- assistance
apist and/or facility staff. Community mobility status Independent 3 (13)
was characterized using a categorical scale developed by Falls history
Reported 1 or more falls in past 13 (54)
Perry et al.29 Balance confidence was assessed by asking
6 mo
each participant whether they perceived themselves to Reported injury with falls in past 6 (25)
be unsteady in 2 conditions: standing and during walk- 6 mo
ing (yes or no response). While more detailed balance Balance confidence
confidence scales are available, this brief assessment pro- Feels unsteady in standing 8 (33)
Feels unsteady in walking 10 (42)
vided basic information about perceived balance ability
Education level
in a short period of time. Participant characteristics are Less than high school 5 (21)
described in Table 1. High school or GED 7 (29)
The testing sequence was consistent for all subjects: Some college 7 (29)
(1) Symbol Digit Modalities Test (SDMT); (2) the 5 more Associate or bachelor’s degree 3 (12)
Master’s degree or higher 1 (4)
difficult items from the Berg Balance Scale (BBS-5); (3)
Unknown 1 (4)
Four Square Step Test (FSST); (4) High Level Mobility
Assessment Test (HiMAT); (5) modified Walking and Re-
membering Test (WART); and (6) during the entire test-
Symbol Digit Modalities Test
ing session, 1 research team member completed the ob-
servational Moss Attention Rating Scale (MARS). Sub- The SDMT measures attention and processing speed.
jects were offered rest breaks and water if needed during Subjects are given 90 seconds to match as many digits
testing, although no subjects required a break. to symbols as possible, using a key with 9 digit-symbol
pairs. The oral version of this test, which requires subjects
Measures
to recite the appropriate number for each symbol, has
Measures were selected on the basis of reliability and been shown to best measure deficits in speed of infor-
validity, ease of administration in a single test session, mation processing in people post-ABI without the need
and relevance to the proposed research questions. Sev- to account for motor impairment in comparison with
eral tools were used to measure each construct in order similar neuropsychological measures.5 The SDMT has
to capture a continuum of balance and attention abili- been utilized as a neuropsychological outcome measure
ties. The inclusion of various measures was an important in numerous studies involving people with ABI.30–34 It
consideration in testing a sample that was anticipated to was part of a battery of neuropsychological measures ad-
have a range of abilities. ministered 1-month post–brain injury found to predict
www.headtraumarehab.com
158 JOURNAL OF HEAD TRAUMA REHABILITATION/MAY–JUNE 2010
functional outcomes at 1 year postinjury.32 In a study
of people with complicated mild to severe TBI, average
score on the oral SDMT at 5 years postinjury was 40.56
(SD = 16.86).30 When compared with normative data,
72.6% of the sample demonstrated mild or greater im-
pairment and 49.4% demonstrated moderate to severe
impairment.30 Normative data generated by Sheridan
et al31 was used to interpret SDMT findings in this study.
Moss Attention Rating Scale
The MARS is a 22-item tool that utilizes a 5-
point Likert-type rating scale to characterize aspects
of attention after TBI. Using Rasch analyses, Whyte
et al35 and Hart et al36 modified the original 53-item
test to its present 22-item format. An overall score
and factor scores based on 3 dimensions of attention
deficit—initiation, restlessness and distractibility, and
sustained/consistent attention—can be derived.35,36 The
rating scale ranges from “definitely true” to “definitely
false,” with mid-range ratings of “true, for the most part,”
“sometimes true, sometimes false,” and “false, for the
most part.” The original MARS item content and factor
structure has strong face validity.35,36 Scoring guidelines
provided by Whyte et al37 were used to interpret scores
and factors. Use of the MARS in a single test session is a
departure from its original intent because 2 items on the
scale require observation over time or across days. These
2 items were scored following discussion with facility
and therapy staff members regarding their observations
of the individual with TBI.
Berg Balance Scale
The BBS is a 14-item measure of static and dynamic
balance in sitting and standing. The items are measured
on a 5-point scale from 0 to 4, with a maximum score of
56. Developed as a measure to identify balance deficits
in the elderly, the BBS has been found to be valid and
reliable as a predictor of falls38 and has been advocated
for use after brain injury to predict rehabilitative out-
comes in combination with other clinical measures.39
Test-retest reliability for the BBS with brain injury is ex-
cellent, but a ceiling effect has been described for higher-
functioning individuals.40 A score of 45 has been sug-
gested as a cutoff point for prediction of high risk of
falls in older adults.38 Using Rasch analysis, Kornetti et
al41 demonstrated that 4 standing items of the BBS were
critical in achieving a score of 45: (1) turning to look
behind over shoulders; (2) alternate stepping; (3) single
leg stance; and (4) tandem stance. Our study protocol
included the items identified by Kornetti et al, in addi-
tion to a fifth BBS item, the 360◦ turn, because of the
importance of turning for functional mobility. The 360◦
turn with the items advocated for use by Kornetti et al
are referred to as BBS-5. BBS-5 scores were summed for
a total of up to 20. If a subject scored less than 3 on
any 1 of the 5 BBS items tested, the remainder of the
BBS was also administered to confirm a total BBS score.
Total BBS scores are reported to aid therapist interpre-
tation of our results since therapists are accustomed to
interpreting BBS scores based on a total out of 56. The
total BBS values resulted from either complete BBS test-
ing (when there were difficulties in completing BBS-5)
or if subjects scored perfectly on the BBS-5 items, that
score was added to 36, crediting a full score for each of
the first 9 items of the scale.
Four Square Step Test
A single-item timed test of dynamic balance, the FSST
requires steps forward, sideways, and backwards over 4
canes placed on the ground in a cross formation, moving
first in a clockwise direction and then counterclockwise
to return to the starting position.42 The test incorporates
a concurrent cognitive task of remembering the specific
multiple step movement sequence. Subjects complete
the task as quickly and safely as possible without touch-
ing or stepping on the canes. The FSST is a reliable and
valid clinical test of dynamic standing balance for older
adults with sensitivity of 85%, specificity of 88%, and
a positive predictive value of 86% for identifying older
adults with a history of multiple falls.42 Although this
test has not been validated for individuals with ABI, it
was included in the test battery for its functional rele-
vance, the element of cognition that is part of the task,
and its strength as a falls risk predictor in older adults.
High-Level Mobility Assessment Test
The HiMAT is a 13-item test of motor performance
for individuals with TBI.43 The test measures higher-
level mobility including hopping, skipping, jumping,
bounding, running, and stair climbing. Times and dis-
tances are recorded with corresponding ordinal scores
assigned and then added to generate a total score from
0 to 54. Content validity, internal consistency, and dis-
criminability of the HiMAT have been carefully exam-
ined in a series of studies.43–45 Rasch analysis identified
redundant items included in the measure, eliminating
7 items from the original to provide the most discrim-
inative measure.45 The HiMAT has high test-retest re-
liability in healthy young adults.46 Norms for healthy
adults aged 18 to 25 were found to be 50 to 54 for men
and 44 to 54 for women; a ceiling effect was evidenced
for men but not for women.46 The HiMAT was selected
for its clinical feasibility and effectiveness in testing at
higher levels of function after TBI.
Modified Walking and Remembering Test
The modified WART47 is a multiple component test
that uses a single-task working memory task (using
 Balance, Attention, and Dual-Task Performance During Walking After Brain Injury 159

phonological loop with a forward digit span test), single-
task walking (20 ft, a 180◦ turnaround, and 20-ft re-
turn walk), and a dual-task condition that combines
both tasks, modeled after an experimental paradigm
described by Lindenberger et al.48 The cognitive task,
forward digit span, was tested using the protocol from
the Wechsler Adult Intelligence Test—Revised. The sub-
ject’s maximum digit span while sitting was determined.
The maximum digit span with a delay equivalent to the
time to complete the motor task was then confirmed.
Subjects walked quickly, while remembering a random
string of digits appropriate to their digit span ability,
and recalled the span at the completion of the walk-
ing task. The original version of the WART included a
walking accuracy component in which raters counted
the number of times each subject stepped off a 12-in
narrow path. However, step accuracy is difficult to rate
and administer, and some individuals post-ABI struggle
with walking a narrow path because of gait deviations.49
The modified version of the WART was devised with
study of individuals with ABI, incorporating a midpoint
180◦ turn, which provides a balance challenge feasible
for most survivors of ABI without the greater challenge
of walking on a very narrow path.49,50 The WART has
been found to be an effective, feasible, and reliable
measure for eliciting DTCs in ambulatory subjects with
ABI.49,50
Data analysis
Descriptive statistics for balance and attention mea-
sures were generated. Individual SDMT scores were in-
terpreted on the basis of published norms. Balance scores
were interpreted on the basis of guidelines for assessing
falls risk. Relative DTCs were calculated for walking time
and cognitive task accuracy, expressed as a percentage
decrement from single task performance. Cognitive task
accuracy was calculated on the basis of digits recalled in
the dual-task condition. Baseline digit recall was 100%
for the digit span selected for testing. Using a protocol
described by Nebes et al,51 digit responses were counted
as correct if (1) the first or last digit of a span was correct
in that position, (2) any correct digits adjacent to the first
or last digit were recalled, and (3) a correct sequence of
3 or more digits anywhere in the response was provided.
The correct number of digits was divided by the number
in the span to represent accuracy as a percentage.
Pearson correlation coefficients were calculated to de-
termine associations between balance, attention, and
DTCs. Subjects were classified as fallers or nonfallers on
the basis of reported (and staff-confirmed) falls history.
To investigate differences in balance, attention, and dual-
task scores between the 2 groups, t tests were performed.
RESULTS
Descriptive findings
Summary statistics for outcome measures are pre-
sented in Table 2. Large standard deviations (SDs) and
ranges were calculated for all measures, with the excep-
tion of the BBS total and MARS. Coefficients of varia-
tion ranged from 43% to 142% on the other measures,
with cognitive DTCs, FSST, and HiMAT having the
largest SD:mean ratios. The sample had large variation
in balance, mobility, and dual-task impairments.
Although guidelines for judging falls risk for older
adults cannot be directly translated to individuals with
ABI, the mean score for the BBS was above the com-
monly used cutpoint to indicate increased falls risk used
with older adults. In contrast, the mean score for the
FSST was more than 15 seconds, which is consistent
with increased falls risk for older adults. On average,
participants received less than half the possible total
points on the HiMAT, suggesting difficulties with higher-
level mobility. All subjects scored below their sex- and

TABLE 2 Balance, attention, and dual-task performance

Coefficient of
Test Mean (SD) Range variation (SD/mean)

Five-item BBS (of 20) 12.8 (5.5) 3–20 0.43

BBS (of 56) 48.0 (9.2) 25–56 0.19
Four Square Step Test, s 19.6 (13.7) 7.0–63.6 0.70
High Level Mobility Assessment Test (of 54) 20.2 (14.2) 0–46 0.70
Symbol Digit Modalities Test (number correct) 29.9 (15.0) 7–55 0.50
Moss Attention Rating Scale (of 110) 79.5 (11.2) 59–102 0.14
Walking and Remembering Test,%
Walking DTC 20.9 (13.0) 2–45 0.62
Cognitive DTC 13.0 (18.5) 0–60 1.42
Average DTC 16.9 (9.5) 2–37 0.56

Abbreviations: BBS, Berg Balance Scale; DTC, dual-task cost.

www.headtraumarehab.com
160 JOURNAL OF HEAD TRAUMA REHABILITATION/MAY–JUNE 2010

TABLE 3 Balance, attention, and dual-task cost correlationsa

MARS- MARS- DTC- DTC-

BBS FSST HiMAT SDMT total restless walk cognitive

Balance
BBS 1.00
FSST −0.61 (0.004) 1.00
HiMAT 0.70 (0.001) −0.72 (0.001) 1.00
Attention
SDMT 0.22 −0.02 0.43 (0.07) 1.00
correct
MARS-total 0.25 0.06 0.19 0.42 (0.05) 1.00
MARS- 0.04 0.18 −0.15 0.15 0.70 (0.001) 1.00
restless
DTC
Walking −0.14 0.18 −0.17 −0.33 −0.14 0.09 1.00
Cognitive 0.13 −0.10 0.11 0.16 0.49 (0.02) 0.46 (0.03) −0.06 1.00

Abbreviations: BBS, Berg Balance Scale; DTC, dual-task cost; FSST, Four Square Step Test; HiMAT, High Level Mobility Assess-
ment Test; MARS, Moss Attention Rating Scale; SDMT, Symbol Digit Modalities Test.
a
P values are listed in parentheses for correlations <.10, all other P values are >.10.

age-predicted values on the SDMT, based on normative
data from a large, community-based sample.31
Mean walking DTCs were higher than mean cogni-
tive DTCs; however, participants demonstrated variable
DTC patterns. Dual-task costs of more than 10% were
used as a threshold to categorize patterns of cognitive
and motor performance, based on prior study of young
adults with the same paradigm, demonstrating low DTCs
of 2% to 3% for walking and 8% to 9% for the digit
span task.47 Forty-eight percent of subjects (n = 11)
demonstrated only motor DTCs (slowed walking speed),
35% (n = 9) demonstrated both motor and cognitive
DTCs (slowed walking speed and reduced cognitive ac-
curacy), and 9% (n = 2) demonstrated only cognitive
DTCs (reduced cognitive accuracy only). Two partici-
pants demonstrated no DTC in either task.
Correlational data
Correlations between DTCs and measures of atten-
tion and balance are shown in Table 3. Correlations were
highest between scores on the 3 balance measures (the
BBS, FSST, and HiMAT), with absolute r values of 0.61
to 0.72 (all P ≤.004). Scores on the SDMT and MARS,
both of which measure attention, were also signifi-
cantly correlated (r = 0.42, P ≤.05), although at a lower
level.
Scores on only 1 balance or attention measure, the
MARS, were significantly associated with DTC (motor
or cognitive). Both total MARS score and score on the
restlessness factor were positively associated with cog-
nitive DTCs, paradoxically indicating that cognitive ac-
curacy during the WART dual-task condition declined
most in subjects with better attention on the basis of
behavioral observation.
Group comparisons
Fallers and nonfallers scored significantly differently
on balance and mobility measures. Those reporting at
least 1 fall (n = 13) in the past 6 months had lower
scores on the BBS (P ≤.03) and higher times for the
FSST (P ≤.01) than subjects reporting no falls (n = 11).
As expected, fallers had greater balance impairments.
The same trend was seen in scores on the HiMAT, with
nonfallers scoring higher on the mobility measure than
fallers (P ≤.03). Mean values of attention (MARS total,
SDMT) and dual-task performance (cognitive, motor, or
average DTC) were not significantly different between
the faller and nonfaller groups.
DISCUSSION
The battery of outcome measures used in this study
described balance, attention, and patterns of dual-task
performance across a range of abilities. The HiMAT was
beneficial in capturing higher-level balance ability for
those individuals who performed on the BBS with sig-
nificant ceiling effect. The approach of administering
the most difficult items on the BBS may be an efficient
way to assess the appropriateness for more challenging
balance assessment in a clinical setting.41
The FSST is traditionally not considered a dual-task
measure, but it requires working memory to execute
the correct stepping sequence. The continuous nature
of the task made it feasible for all participants in the
 Balance, Attention, and Dual-Task Performance During Walking After Brain Injury
study. This task could be described as requiring the visu-
ospatial sketchpad component of the working memory
model of Baddeley since the pattern of movement is
first described and then completed during the test, re-
quiring memory of the required directional pattern. In
this study, the FSST allowed us to grossly gauge dual-
task performance while assessing falls risk when stepping
over obstacles. The cognitive requirement of remember-
ing and executing the appropriate stepping sequence was
a prerequisite for a FSST trial to “count.” This require-
ment may have served to prioritize the cognitive task
over the speeded stepping task, resulting in slower task
execution. That subjects on average appeared at risk for
falls with this measure and not with the BBS was an un-
expected result. Obstacle avoidance has been effective
in studies of mild TBI to identify dual-task performance
problems28 and is part of the FSST. The use of a visu-
ospatial task has been described as interfering to a greater
degree than other cognitive tasks with balance perfor-
mance since balance also requires the use of vision and
visual information.52–54
Inclusion of the HiMAT in the battery of assessments
allowed some subjects to recognize untapped ability in
performing high-level motor activities that had not been
challenged in therapy or usual routines. Use of the Hi-
MAT as a standard part of assessment may allow ther-
apists either to identify goals that are more challenging
or to identify strengths in abilities on the basis of docu-
mented norms.46
Previous studies have described a relationship between
balance impairments and falls history.1 Balance impair-
ment was clearly related to history of falls in this sam-
ple, although our method of determining falls history
was admittedly imperfect. The use of self-report of falls
history in a population with potential cognitive deficits
is a significant limitation of the study. Although we con-
firmed all participant responses by discussion with fa-
cility staff and therapists, individuals may have been
misclassified. Further research would be improved by
conducting more structured surveillance to determine
actual falls in a prospective manner.
Unexpectedly, the study failed to illustrate significant
differences in attention or dual-task performance be-
tween fallers and nonfallers. There are several possible
reasons for this finding. Our measures of attention were
limited in time and scope. Testing occurred in quiet,
controlled environments that may not have provided
enough distraction for attention deficits to be observed
during testing. Participation in the study involved in-
teraction with 4 team members and a range of novel
tasks. This social context and activity level could have
increased alertness and attention. Since testing occurred
in a single session, the MARS was not used as originally
devised. It is possible that the combination of scores
from a single observer augmented by staff impressions
161
for 2 items was a faulty approach. The MARS has been
tested and validated in an acute rehabilitation setting
that could vary considerably in the complexity of en-
vironmental factors. In contrast, we attempted to min-
imize possible environmental distractors for the testing
session.
We anticipated that participants with attention deficits
would have greater DTC during dual-task conditions, as
has been demonstrated in prior studies with stroke and
ABI; however, this finding was not supported. Subjects
were recruited from a supported living program in which
some have been living following brain injury for many
years. These individuals may have developed more auto-
matic control for walking as a result of extensive practice
since the onset of injury. Better attention scores (MARS
total and restlessness scores) were positively associated
with greater cognitive DTC, suggesting that subjects with
better attention had more cognitive costs in dual-task
conditions. This unexpected finding could be a result of
subjects with greater injury chronicity developing com-
pensatory strategies such as sacrificing cognitive perfor-
mance to maximize safety by maintenance of balance
and dynamic stability. An increase in DTCs in this con-
dition could be a very adaptive and safe response that
should be encouraged rather than viewed as a deficit or
performance problem.
Our test protocol allowed each participant to prior-
itize cognitive or motor performance in the dual-task
condition. While this approach is enlightening to
demonstrate what may occur in natural conditions, the
patterns of response varied widely, resulting in small
numbers of subjects with any single pattern of response.
This variability may have contributed to insignificant
associations between DTC variables and falls history.
Recent investigations in dual-task performance suggest
forcing a priority on one task or the other by verbal direc-
tion as being beneficial in rehabilitation of older adults
at risk for falls.55 This variable priority approach may be
helpful in highlighting problems with attention alloca-
tion and challenge executive control to a greater degree.
Future investigations could use a modified protocol that
prioritizes attention to one task or the other in successive
trials, attempting to elicit DTC in 1 of 2 tasks, instead of
allowing variability that challenges interpretation. The
test protocol for the FSST requires the step sequence to
be carried out correctly in order for a trial to “count,”
possibly driving DTC to the motor task, whereas DTC
in the WART clearly varied across subjects.
Given the exploratory nature of this study, a conve-
nient sample resulted in many participants with chronic
injury (mean time since onset = 117.8 months). Further
research should focus on a sample size more homoge-
nous in injury acuity. Individuals with more chronic
injury may have developed compensatory strategies to
prevent falls and maximize safety, whereas a more acute
www.headtraumarehab.com
162 JOURNAL OF HEAD TRAUMA REHABILITATION/MAY–JUNE 2010
injury group may be more in need of intervention to
reduce falls risk.
The battery of balance measures including graded
challenges presented by the BBS-5, FSST, and the Hi-
MAT is recommended for future study. The FSST may
have particular value as a simple method to simulate
dual-task conditions without the complexities of dual-
task test protocols. Since the FSST inherently prioritizes
the cognitive task, an additional mode of testing that pri-
oritizes the motor task may also be valuable. There are
clearly functional situations in which prioritizing motor
performance may be critical, such as hurrying across a
busy intersection.
REFERENCES
1. 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.
2. Hyndman D, Ashburn A. People with stroke living in the com-
munity: attention deficits, balance, ADL ability and falls. Disabil
Rehabil. 2003;25:817–822.
3. Felicetti T. Falling in an aging brain injury population: analysis of
a specific fall prevention protocol in a postacute brain injury reha-
bilitation center [serial online]. Prof Case Manag. 2009;14(3):158–
160.
4. McDowd JM. An overview of attention: behavior and brain. J
Neurol Phys Ther. 2007;31:98–103.
5. Ponsford J, Kinsella G. Attentional deficits following closed-head
injury. J Clin Exp Neuropsychol. 1992;14:822–838.
6. Posner MI, Raichle M. Images of Mind. Revised ed. Washington,
DC: Scientific American Books; 1996.
7. Baddeley A. Working Memory. New York, NY: Oxford University
Press; 1986.
8. Baddeley A, Della Salla S. Working memory and executive con-
trol. In: Roberts A, Robbins T, Weiskrantz L, eds. The Prefrontal
Cortex: Executive and Cognitive Functions. New York, NY: Oxford
University Press; 1998:9–21.
9. Repovs G, Baddeley A. The multi-component model of working
memory: explorations in experimental cognitive psychology. Neu-
roscience. 2006;139:5–21.
10. Baddeley AD. Working memory or working attention? In: Bad-
deley AD, Weiskrantz L, eds. Attention: Selection, Awareness and
Control. Oxford, England: Clarendon Press; 1993:152–170.
11. Cicerone K, Levin H, Malec J, Stuss D, Whyte J. Cognitive re-
habilitation interventions for executive function: moving from
bench to bedside in patients with traumatic brain injury. J Cogn
Neurosci. 2006;18:1212–1222.
12. Alladi S, Meena AK, Kaul S. Cognitive rehabilitation in stroke:
therapy and techniques. Neurol India. 2002;50:S102–S108.
13. Baddeley A, Della Sala S, Papagno C, Spinnler H. Dual-task per-
formance in dysexecutive and nondysexecutive patients with a
frontal lesion. Neuropsychology. 1997;11(2):187–194.
14. Studer M. Rehabilitation of executive function: to err is human,
to be aware-divine. J Neurol Phys Ther. 2007;31:128–134.
15. Hanna-Pladdy B. Dysexecutive syndromes in neurologic disease.
J Neurol Phys Ther. 2007;31:119–127.
16. Dell’Acqua R, Stablum F, Galbiati S, Spannocchi G, Cerri C. Selec-
tive effect of closed-head injury on central resource allocation: ev-
idence from dual-task performance. Exp Brain Res. 2001;136:364–
378.
The ability to flexibly allocate attention to different
cognitive and balance tasks through the use of execu-
tive control appears to improve with training in older
adults at risk of falling. This concept may also be appli-
cable following brain injury, given the prevalence of ex-
ecutive function deficits with this population. The abil-
ity to modify attention allocation from one’s preferred
method has not been examined during balance and walk-
ing tasks after a TBI. A dual-task test protocol that incor-
porates explicit prioritization of each task during sepa-
rate trials may improve the ability to make comparisons
across subjects and draw conclusions about dual-task
performance that can then be linked to falls risk.
17. Schneider W, Shiffrin RM. Controlled and automatic human in-
formation processing, part I: detection, search, and attention. Psy-
chol Rev. 1977;84:1–66.
18. McCulloch K. Attention and dual-task conditions: physical ther-
apy implications for individuals with acquired brain injury. J Neu-
rol Phys Ther. 2007;31:104–118.
19. Abernethy B. Dual-task methodology and motor skills research:
some applications and methodological constraints. J Hum Mov
Stud. 1988;14:101–132.
20. Hyndman D, Ashburn A, Yardley L, Stack E. Interference be-
tween balance, gait and cognitive task performance among people
with stroke living in the community. Disabil Rehabil. 2006;28:849–
856.
21. Haggard P, Cockburn J, Cock J, Fordham C, Wade D. Interference
between gait and cognitive tasks in a rehabilitating neurological
population. J Neurol Neurosurg Psychiatry. 2000;69:479–486.
22. Cockburn J, Haggard P, Cock J, Fordham C. Changing patterns
of cognitive-motor interference (CMI) over time during recovery
from stroke. Clin Rehabil. 2003;17:167–173.
23. Parker TM, Osternig LR, Lee HJ, Donkelaar P, Chou LS. The effect
of divided attention on gait stability following concussion. Clin
Biomech (Bristol, Avon). 2005;20:389–395.
24. Catena RD, van Donkelaar P, Chou LS. Cognitive task effects on
gait stability following concussion. Exp Brain Res. 2007;176:23–
31.
25. Catena RD, van Donkelaar P, Chou LS. Altered balance control
following concussion is better detected with an attention test dur-
ing gait. Gait Posture. 2007;25(3): 406–411.
26. Parker TM, Osternig LR, van Donkelaar P, Chou LS. Recovery of
cognitive and dynamic motor function following concussion. Br
J Sports Med. 2007;41(12):868–873; discussion 873.
27. Catena RD, van Donkelaar P, Chou LS. Different gait tasks distin-
guish immediate vs. long-term effects of concussion on balance
control. J Neuroeng Rehabil. 2009;6(25):25.
28. Catena RD, van Donkelaar P, Halterman CI, Chou LS. Spatial
orientation of attention and obstacle avoidance following con-
cussion. Exp Brain Res. 2009;194(1):67–77.
29. Perry J, Garrett Gronley J, Mulroy S. Classification of walking
handicap in the stroke population. Stroke. 1995;26:982–989.
30. Millis SR, Rosenthal M, Novack TA, et al. Long-term neuropsy-
chological outcome after traumatic brain injury. J Head Trauma
Rehabil. 2001;16:343–355.
31. Sheridan LK, Fitzgerald HE, Adams KM, et al. Normative Symbol
Digit Modalities Test performance in a community-based sample.
Arch Clin Neuropsychol. 2006;21:23–28.
 Balance, Attention, and Dual-Task Performance During Walking After Brain Injury
32. Hanks RA, Millis SR, Ricker JH, et al. The predictive validity
of a brief inpatient neuropsychologic battery for persons with
traumatic brain injury. Arch Phys Med Rehabil. 2008;89:950–957.
33. Kashluba S, Hanks RA, Casey JE, Millis SR. Neuropsychologic
and functional outcome after complicated mild traumatic brain
injury. Arch Phys Med Rehabil. 2008;89:904–911.
34. Willmott C, Ponsford J. Efficacy of methylphenidate in the re-
habilitation of attention following traumatic brain injury: a ran-
domised, crossover, double blind, placebo controlled inpatient
trial. J Neurol Neurosurg Psychiatry. 2009;80:552–557.
35. Whyte J, Hart T, Bode RK, Malec JF. The moss attention rating
scale for traumatic brain injury: initial psychometric assessment.
Arch Phys Med Rehabil. 2003;84:268–276.
36. Hart T, Whyte J, Millis S, et al. Dimensions of disordered attention
in traumatic brain injury: further validation of the moss attention
rating scale. Arch Phys Med Rehabil. 2006;87:647–655.
37. Whyte J, Hart T, Ellis CA, Chervoneva I. The moss attention rating
scale for traumatic brain injury: further explorations of reliability
and sensitivity to change. Arch Phys Med Rehabil. 2008;89(5):966–
973.
38. Berg KO, Wood-Dauphinee SL, Williams JI, Maki B. Measuring
balance in the elderly: validation of an instrument. Can J Public
Health. 1992;83(suppl 2):S7–S11.
39. Feld JA, Rabadi MH, Blau AD, Jordan BD. Berg Balance Scale and
outcome measures in acquired brain injury. Neurorehabil Neural
Repair. 2001;15:239–244.
40. Newstead AH, Hinman MR, Tomberlin J. Reliability of the Berg
Balance Scale and Balance Master Limits of Stability Tests for
individuals with brain injury. J Neurol Phys Ther. 2005;29(1):18–
23.
41. Kornetti DL, Fritz SL, Chiu YP, Light KE, Velozo CA. Rating
scale analysis of the Berg Balance Scale. Arch Phys Med Rehabil.
2004;85:1128–1135.
42. Dite W, Temple VA. A clinical test of stepping and change of
direction to identify multiple falling older adults. Arch Phys Med
Rehabil. 2002;83:1566–1571.
43. Williams GP, Greenwood KM, Robertson VJ, Goldie PA, Morris
ME. High-level Mobility Assessment Tool (HiMAT): interrater
reliability, retest reliability, and internal consistency. Phys Ther.
2006;86:395–400.
44. Williams G, Robertson V, Greenwood K, Goldie P, Morris ME.
163
The concurrent validity and responsiveness of the High-level
Mobility Assessment Tool for measuring the mobility limitations
of people with traumatic brain injury. Arch Phys Med Rehabil.
2006;87:437–442.
45. Williams GP, Robertson V, Greenwood KM, Goldie PA, Morris
ME. The High-level Mobility Assessment Tool (HiMAT) for trau-
matic brain injury, part 2: content validity and discriminability.
Brain Inj. 2005;19:833–843.
46. Williams GP, Rosie J, Denisenko S, Taylor D. Normative values
for the High-level Mobility Assessment Tool (HiMAT). Int J Ther
Rehabil. 2009;16(7):370–374.
47. McCulloch KL, Mercer VS, Giuliani C, Marshall S. Development
of a clinical measure of dual-task performance in walking: relia-
bility and preliminary validity of the Walking and Remembering
Test. J Geriatr Phys Ther. 2009;32:2–9.
48. Lindenberger U, Marsiske M, Baltes PB. Memorizing while walk-
ing: increase in dual-task costs from young adulthood to old age.
Psychol Aging. 2000;15:417–436.
49. McCulloch K, Blakley K, Freeman L. Clincial tests of walking dual-
task performance after acquired brain injury (ABI): feasibility and
dual-task cost comparisons to a young adult group. J Neurol Phys
Ther. 2005;29:213.
50. McCulloch KL. Feasibility and interrater reliability of clinical tests
of dual-task performance after acquired brain injury. Arch Phys Med
Rehabil. 2005;86:e21.
51. Nebes RD, Butters MA, Houck PR, et al. Dual-task performance
in depressed geriatric patients. Psychiatry Res. 2001;102:139–151.
52. Kerr B, Condon SM, McDonald LA. Cognitive spatial processing
and the regulation of posture. J Exp Psychol Hum Percept Perform.
1985;11:617–622.
53. Maylor EA, Wing AM. Age differences in postural stability are
increased by additional cognitive demands. J Gerontol Psychol Sci.
1996;14:P143–P154.
54. Shumway-Cook A, Woollacott M, Kerns KA, Baldwin M. The
effects of two types of cognitive tasks on postural stability in
older adults with and without a history of falls. J Gerontol.
1997;52A:M232–M240.
55. Silsupadol P, Shumway-Cook A, Lugade V, et al. Effects of single-
task versus dual-task training on balance performance in older
adults: a double-blind, randomized controlled trial. Arch Phys Med
Rehabil. 2009;90(3):381–387.
www.headtraumarehab.com