Contribution of Motor and Cognitive Factors to Gait Variability and Fall Risk:
From Clinical Assessment to Neural Connectivity
DISSERTATION
Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy
in the Graduate School of The Ohio State University
By
Nora Elizabeth Fritz, PT, DPT
Graduate Program in Health and Rehabilitation Sciences
The Ohio State University
2013
Dissertation Committee:
Deborah S. Nichols-Larsen, PT, PhD, Advisor
Deborah A. Kegelmeyer, PT, DPT, MS, GCS
Anne D. Kloos, PT, PhD, NCS
Ruchika S. Prakash, PhD
Copyright by
Nora Elizabeth Fritz
2013
Abstract
Both motor and cognitive factors contribute to increases in gait variability and fall risk in
individuals with neurologic disorders. Tasks requiring simultaneous mobility and cognition (i.e.
dual-tasks) have been linked with accidental falls. Although these deficits have been well
documented in elderly adults, relapsing remitting multiple sclerosis (RRMS), Parkinson’s disease,
traumatic brain injury and stroke, the individual contributions of mobility and cognition have not
been explored and there are few studies exploring rehabilitation for these impairments using
dual-task training. In order to move toward interventional techniques that effectively improve
dual-tasking in individuals with neurologic conditions, a better understanding of tasks that
amplify gait variability, the contribution of cognition to mobility and who may benefit from such
programs is needed.
Together, a systematic review of the literature and a case study provided evidence to
support the importance of the interaction between mobility and cognition, the presence of dualtask deficits, and the feasibility of dual-task training in neurologic disorders. Gaps in the
literature identified in the systematic review were used to direct the research questions of this
document.
The results of cross-sectional and baseline data from a randomized controlled trial
suggest that tasks that amplify gait variability may be related to falls and underlying cognition.
Backward walking amplified gait variability in both elderly adults and individuals with RRMS, and
a backward walking velocity of <0.6 m/s accurately predicted elderly fallers. Dual-task walking
ii
amplified gait variability in individuals with RRMS, and higher gait variability correlated with
lower performance on neuropsychological tests (p<0.01).
Diffusion imaging is a useful tool to examine neural connectivity in individuals with
RRMS. The supplemental motor area (SMA) has been implicated in cognitive dual-tasks, but its
relationship with motor-cognitive dual-tasks has not been explored. Our baseline imaging work
suggests that connectivity of the interhemispheric SMA tract is associated with velocity of dualtask walking in RRMS; regression analysis demonstrates a significant relationship between dualtask walking velocity and SMA axial diffusivity (r2=0.436, p=0.014).
Although work is ongoing to characterize dual-task deficits in individuals with neurologic
conditions, our review of the literature suggests that this is an impairment that spans diagnoses
and that rehabilitation focused on training both motor and cognitive domains can be successful
in improving dual-task deficits. Our case study demonstrated that dual-task deficits can be
improved even in individuals with severe impairment. These studies were limited by small
sample sizes, the paucity of dual-task training intervention studies to date, differences in
measurement tools across studies, making it difficult to compare and generalize results. Our
findings were also limited by the preliminary nature of our RRMS dataset.
However, our results provide guidance for future studies to understand the contribution
of motor and cognitive factors to gait variability and falls in individuals with neurologic
conditions. Rehabilitation of dual-task impairments could improve quality of life, safety and
independence in these individuals. Further research is needed to establish appropriate training
programs for different diagnoses, effective measuring tools for demonstrating improvement,
and predictive factors to explore which individuals can benefit from such programs.
iii
All we have to decide is what to do with the time that is given us.
- J.R.R. Tolkien, The Fellowship of the Ring
For my family and friends who have always been supportive and
encouraging when I needed it most.
All my love.
iv
Acknowledgments
I have no notion of loving people by halves, it is not my nature.
- Jane Austen, Northanger Abbey
I would like to thank my incredibly supportive dissertation committee: Dr. Deborah
Nichols-Larsen, for her advice and mentorship; Drs. Anne Kloos and Deborah Kegelmeyer, for
maintaining sanity throughout the process of mentoring their first doctoral student, and for four
memorable European adventures; and Dr. Ruchika Prakash, for her collaboration and insightful
inquiries.
In addition, I would like to gratefully acknowledge: Dr. Alexandra Borstad for her
overwhelming friendship and support throughout the doctoral process; Dr. Michele Basso for
always challenging me to improve my work and put forth the best possible product; and Dr.
John Buford for his excellent guidance and for convincing me that a PhD was a good idea in the
first place.
Finally, I must thank the faculty of The Ohio State University Department of Physical
Therapy for their continual support; The Center for Clinical and Translational Science for their
financial support and assistance with recruitment, biostatistics, and data management; and The
Foundation for Physical Therapy and the Center for Cognitive and Behavioral Brain Imaging for
their financial support.
v
Vita
2007 ..................................................................... B.S. Zoology, Miami University
2008-2009 ............................................................ Graduate Research Assistant, Department of
Physical Therapy, The Ohio State University
2010 ..................................................................... DPT, The Ohio State University
2010 ..................................................................... Foundation for Physical Therapy Florence P.
Kendall Doctoral Scholarship
2010-2011 ............................................................ Graduate Research Associate, Department of
Physical Therapy, The Ohio State University
2011-2013 ............................................................ TL1 Mentored Research Training Award
2011-2013 ........................................................... Graduate Research Fellow, Center for Clinical
and Translational Science, The Ohio State
University
2013 ..................................................................... University of Florida Neuromuscular Plasticity
Scholar
Publications
Kloos A, Fritz N, Kostyk S, Young G, Kegelmeyer D. Video game play (Dance Dance Revolution)
has potential as an exercise therapy in Huntington's disease. Clin Rehabil. 2013. Accepted.
vi
Quinn L, Khalil H, Dawes H, Fritz NE, Kegelmeyer D, Kloos AD, Gillard JW, Busse M; the Outcome
Measures Subgroup of the European Huntington’s Disease Network. Reliability and minimal
detectable change of physical performance measures in individuals with pre-manifest and
manifest Huntington disease. Phys Ther. 2013. [epub ahead of print:
http://www.ncbi.nlm.nih.gov/pubmed/23520147]
Fritz NE, Worstell A, Siles A, Kloos AD, Kegelmeyer DA. Backward walking parameters are
sensitive to age-related changes in mobility and balance. Gait Posture. 2013; 37: 593-597.
Fritz NE, Basso DM. Effect of dual-task training on balance and mobility in severe traumatic brain
injury: a case study. J Neurol Phys Ther. 2013; 37(1): 37-43.
Quinn L, Busse M, Broad M, Dawes H, Ekwall C, Fritz N, Heemskerk A, Hopkins C, Jones U,
Kegelmeyer, D, Khalil H, Kloos A, Meek C, Owen J, Rao A, Sands R, Watters S, Sakley C, Zinzi P.
Development of physiotherapy guidance and treatment-based classifications for people with
Huntington’s disease. Neurodegenerative Disease Management. 2012; 2(1): 11-19.
Quinn L, Busse M, Bunnig K, Broad M, DeBono K, Ekwall C, Fossmo H, Fritz N, Jones K, Jones U,
Kegelmeyer D, Khalil H, Kloos A, Schuler A, van der Bent J, Vera R. Physiotherapy clinical
guidelines for Huntington’s disease. Neurodegenerative Disease Management. 2012; 2(1): 2131.
Quinn L, Busse M, Broad M, Dawes H, Ekwall C, Fritz N, Heemskert A, Hopkins C, Jones, U,
Kegelmeyer D, Khalil H, Kloos A, Meek C, Owen J, Rao A, Sands R, Watters S. European
Huntington’s Disease Network: Physiotherapy Guidance Document for Physiotherapists, First
Edition. Ulm, Germany: EHDN; 2009. Available online at: www.eurohd.net/html/network/groups/physio.
Fields of Study
Major Field: Health and Rehabilitation Sciences
Graduate Minor: Neuroscience
Graduate Interdisciplinary Specialization: Biomedical Clinical and Translational Science
vii
Table of Contents
Page
Abstract................................................................................................................................. ii
Acknowledgments ................................................................................................................. v
Vita ...................................................................................................................................... vi
Publications .......................................................................................................................... vi
Fields of Study ..................................................................................................................... vii
List of Tables....................................................................................................................... viii
List of Figures ........................................................................................................................ x
List of Abbreviations ............................................................................................................ xii
Chapter 1: General Introduction ............................................................................................. 1
1.1 Introduction ........................................................................................................................... 1
1.2 The Relationship between Attention, Automaticity and Dual-Task ...................................... 1
1.3 Theories of Dual-Tasking & Divided Attention ...................................................................... 2
1.4 Dual-Task across Neurologic Populations: The Contribution of Cognitive and Motor Deficits
to Dual-Task Deficits and Fall Risk ............................................................................................... 3
1.4.1 Baseline Cognitive Deficits .............................................................................................. 3
1.4.2 Baseline Mobility Deficits................................................................................................ 4
1.4.3 Dual-Task Deficits in Individuals with Neurologic Disorders........................................... 6
1.5 Strategies to Compensate for Dual-Task Deficits .................................................................. 7
1.6 Measurement of Dual-Task Ability ........................................................................................ 9
1.6.1 Behavioral Measures ...................................................................................................... 9
1.6.2 Cognitive Measures....................................................................................................... 11
1.6.3 Imaging Measures ......................................................................................................... 12
1.7 Research Aims ...................................................................................................................... 14
1.8 References ........................................................................................................................... 19
iii
Chapter 2: The Effect of Motor-Cognitive Dual-Task Interventions on Mobility in Neurologic
Disorders: A Systematic Review ............................................................................................ 27
2.1 Introduction ......................................................................................................................... 27
2.2 Methods ............................................................................................................................... 29
2.2.1 Data Sources and Searches ........................................................................................... 29
2.2.2 Study Selection.............................................................................................................. 29
2.2.3 Data Extraction & Quality Assessment ........................................................................ 30
2.2.4 Data Synthesis & Analysis ............................................................................................. 32
2.3 Results .................................................................................................................................. 34
2.3.1 Study Characteristics..................................................................................................... 34
2.3.2 Effect of Dual-Task Intervention Programs on Gait ...................................................... 34
2.3.3 Effect of Dual-Task Intervention Programs on Balance ................................................ 46
2.3.4 Effect of Dual-Task Intervention Programs on Cognition ............................................. 47
2.3.5 Effect of Dual-Task Intervention Programs on Ability to Dual-Task ............................. 47
2.3.6 Adverse Effects.............................................................................................................. 49
2.4 Discussion............................................................................................................................. 49
2.4.1 General Findings ........................................................................................................... 49
2.4.2 Limitations.................................................................................................................... 50
2.4.3 Implications for Practice ............................................................................................... 52
2.4.4 Implications for Future Research .................................................................................. 52
2.5 Conclusion ............................................................................................................................ 52
2.6 References ........................................................................................................................... 54
Chapter 3: Backward Walking Measures Are Sensitive to Age-Related Changes in Mobility and
Balance................................................................................................................................ 58
3.1 Introduction ......................................................................................................................... 58
3.2 Methods ............................................................................................................................... 59
3.2.1 Data Analysis ................................................................................................................. 60
3.3 Results .................................................................................................................................. 61
3.3.1 Gait Across Age Groups................................................................................................. 62
3.3.2 Variability of Gait .......................................................................................................... 64
3.3.3 Relationship of Gait Measures to Age .......................................................................... 64
3.3.4 Comparison Elderly Fallers and Non-fallers .................................................................. 67
3.4 Discussion............................................................................................................................. 68
iv
3.5 References ........................................................................................................................... 73
Chapter 4: Dual-Task Training for Balance and Mobility in a Person with Severe Traumatic
Brain Injury: A Case Study .................................................................................................... 75
4.1 Introduction ......................................................................................................................... 75
4.2 Case Description .................................................................................................................. 77
4.2.1 History ........................................................................................................................... 77
4.2.2 Examination .................................................................................................................. 79
4.3 Outcome Measures.............................................................................................................. 81
4.3.1 Overall Function ............................................................................................................ 81
4.3.2 Mobility ......................................................................................................................... 82
4.3.3 Balance .......................................................................................................................... 82
4.3.4 Divided Attention .......................................................................................................... 82
4.4 Prognosis .............................................................................................................................. 83
4.5 Assessment of Readiness to Initiate Dual-Task Training ..................................................... 84
4.6 Intervention ......................................................................................................................... 84
4.6.1 Phase A (post-injury days 58-64) .................................................................................. 85
4.6.2 Phase B (post-injury days 65-71) .................................................................................. 85
4.7 Outcomes ............................................................................................................................. 86
4.8 Discussion............................................................................................................................. 89
4.8.1 Limitations..................................................................................................................... 91
4.9 Summary .............................................................................................................................. 92
4.10 References ......................................................................................................................... 93
4.11 Supplemental Content I: Rancho Los Amigos Levels of Cognitive Function Scale ............. 97
4.12 Supplemental Content II: Examples of Standard Physical Therapy ................................... 98
4.13 Supplemental Content III: Description of Dual-Task Training Program ............................. 99
Chapter 5: Dual-Task Walking Variability is Associated with Stroop and Dual-Task
Questionnaire Performance in Multiple Sclerosis ................................................................ 101
5.1 Introduction ....................................................................................................................... 101
5.2 Methods ............................................................................................................................. 104
5.2.1 Subjects ....................................................................................................................... 104
5.2.2 Data Analysis ............................................................................................................... 106
5.3 Results ................................................................................................................................ 107
v
5.3.1 Comparison of Backward Walking between Individuals with MS and Healthy
Controls ................................................................................................................................ 108
5.3.2 Comparison of Dual-Task Walking between Individuals with MS and Healthy
Controls ................................................................................................................................ 108
5.3.3 Comparison of Forward Walking between Individuals with MS and Healthy
Controls ................................................................................................................................ 108
5.3.4 Comparison of Neuropsychological Test Performance between Individuals with MS
and Healthy Controls ........................................................................................................... 111
5.3.5 Relationship of Variability of Backward and Dual-Task Walking to Neuropsychological
Test Performance ................................................................................................................. 111
5.3.6 Relationship between Dual-Task Questionnaire and Gait Variability and
Neuropsychological Test Performance ................................................................................ 115
5.3.7 Relationship of Dual-Task Cost and Neuropsychological Test Performance .............. 115
5.3.8 Relationship of Falls with Variability of Backward and Dual-Task Walking and
Neuropsychological Test Performance ................................................................................ 115
5.4 Discussion........................................................................................................................... 116
5.5 References ......................................................................................................................... 122
Chapter 6: SMA Connectivity May Be Associated With Dual-Task Walking in Multiple
Sclerosis ............................................................................................................................ 126
6.1 Introduction ....................................................................................................................... 126
6.2. Methods ............................................................................................................................ 129
6.2.1 Participants ................................................................................................................. 129
6.2.2 Mobility Measures ...................................................................................................... 129
6.2.3 MRI Acquisition ........................................................................................................... 131
6.2.4. Imaging Analysis......................................................................................................... 131
6.2.5 Statistical Analysis ....................................................................................................... 133
6.3 Results ................................................................................................................................ 133
6.3.1 Clinical dual-task measures ......................................................................................... 134
6.2.2 Lesion Load Analysis ................................................................................................... 134
6.3.3 Diffusion Tractography ............................................................................................... 138
6.4 Discussion.......................................................................................................................... 141
6.4.1 Limitations................................................................................................................... 142
6.4.2 Clinical Implications .................................................................................................... 143
6.4.3 Future Directions ........................................................................................................ 143
vi
6.5 References ......................................................................................................................... 145
Chapter 7: Research Findings and Future Directions ............................................................ 149
7.1 Summary of Findings.......................................................................................................... 149
7.2 Limitations.......................................................................................................................... 155
7.3 Future Research ................................................................................................................. 156
7.3.1 Contribution of Cognition and Gait Variability to Dual-Task Ability ........................... 156
7.3.2 Dual-Task Training in General ..................................................................................... 156
7.3.3 Dual-Task Training in MS............................................................................................. 157
7.3.4 Validation of Dual-Task Tools in MS ........................................................................... 157
7.3.5 Examine Variability in Healthy Controls for MRI Measures ........................................ 157
7.3.6 Examine Responders vs. Non-Responders to Rehabilitation Training Programs and
Explore the Predictive Potential of MRI Measures. ............................................................. 158
7.4 Conclusion .......................................................................................................................... 158
7.5 References ......................................................................................................................... 160
References ........................................................................................................................ 163
vii
List of Tables
Page
2.1
Trial ratings on the PEDro Methodological Quality and Clinical Relevance
Scales ..........................................................................................................................33
2.2
Characteristics of included studies .............................................................................35
3.1
Interactions by age and direction of gait ....................................................................63
3.2
Spatiotemporal measures and CVs of forward and backward walking .....................65
3.3
Relationship of gait and CV measures to age by walking direction............................66
3.4
Comparison of gait and mobility measures between elderly fallers and
non-fallers ..................................................................................................................67
4.1
Timeline of events......................................................................................................78
4.2
Functional Independence Measure scores .................................................................81
4.3
Motor and dual-task outcomes ..................................................................................87
5.1
Spatiotemporal measures and CVs of forward and backward walking and
forward and backward walking with a cognitive task..............................................110
5.2
Change scores for forward – backward walking in MS and healthy controls ..........110
5.3
Neuropsychological test performance in MS and healthy controls ........................111
5.4
Pairwise comparisons of variability of backward walking and dual-task walking to
neuropsychological test performance and falls in MS..............................................114
viii
5.5
Canonical correlations to examine the relationship between gait variability
and neuropsychological test performance in MS .....................................................114
5.6
Relationship of falls in the past 6 months and dual-task cost to
neuropsychological test performance in RRMS........................................................116
6.1
Relationship between clinical dual-task measures and interhemispheric
SMA tract-specific parameters ................................................................................140
ix
List of Figures
Page
1.1
GAITRite visual output from a representative subject walking forward at their
comfortable pace ........................................................................................................10
1.2
DTI contrast generation ..............................................................................................13
1.3
Schema demonstrating the factors contributing to increased
gait variability and fall risk ..........................................................................................15
1.4
Schema demonstrating the gaps in knowledge to be addressed by this
document’s research goals .........................................................................................16
2.1
Search strategy flowchart ............................................................................................31
2.2
Treatment effects for dual-task training versus comparison treatments
for all trials with full data sets ....................................................................................51
3.1
Backward walking velocity identifies fallers better than forward walking
velocity in elderly ........................................................................................................68
4.1
Changes in gait speed and time to descend stairs by phase ......................................88
6.1
Elliptical representation of a diffusion tensor .........................................................127
6.2
Lesion probability map showing the interhemispheric SMA tract of a
representative subject overlaid with lesions from all MS participants ....................136
6.3
Sagittal and axial slices through the interhemispheric SMA tract of a
representative subject ..............................................................................................137
6.4
Tractography models in representative subjects demonstrating successful
modeling of interhemispheric SMA and interhemispheric V4 connectivity .............139
x
6.5
The relationship between interhemispheric SMA axial diffusivity and
forward dual-task walking velocity ...........................................................................140
7.1
Schema exploring how a review of the literature identified several gaps
in knowledge that directed the research aims of this document.............................152
xi
List of Abbreviations
AD
axial diffusivity
ADL
activity of daily living
AP-COP
anterior-posterior center of pressure
AUC
area under the curve
BBS
Berg Balance Scale
BOS
base of support
BW
backward walking
BWC
backward walking with a cognitive task
BWS
body weight support
CGA
contact-guard assist
CV
coefficient of variation
CVA
cerebrovascular accident
DT
dual-task or dual-tasking
DTC
dual-task cost
DTI
diffusion tensor imaging
DTQ
Dual Task Questionnaire
DTT
dual-task training
Dx
diagnosis
xii
EDSS
Expanded Disability Status Scale
FA
fractional anisotropy
FIM
Functional Independence Measure
FP
fixed priority
FLAIR
fluid attenuated inversion recovery
FW
forward walking
FWC
forward walking with a cognitive task
HHA
hand-held assist
MD
mean diffusivity
ML-COM
medial-lateral center of mass
ML-COP
medial-lateral center of pressure
MMSE
Mini Mental State Examination
MPRAGE
magnetization-prepared rapid gradient-echo
MRI
magnetic resonance imaging
MS
Multiple Sclerosis
m/s
meters per second
PD
Parkinson’s disease
PT
physical therapy or physical therapist
PTA
posttraumatic amnesia
RCT
randomized controlled trial
RD
radial diffusivity
ROI
region of interest
RRMS
Relapsing Remitting Multiple Sclerosis
xiii
SBA
stand-by assist
SDMT
Symbol Digit Modality Test
SMA
supplementary motor area
ST
single-task
TBI
traumatic brain injury
TMT
Tinetti Mobility Test
TUG
Timed Up and Go
WWTT
Walking While Talking Test
VP
variable priority
xiv
Chapter 1: General Introduction
1.1 Introduction
Individuals with neurologic conditions frequently report difficulty performing multiple
simultaneous tasks.1-2While this phenomenon has been widely documented in many
populations, measurement tools, intervention strategies and the underlying neural connectivity
remain largely unexplored. The purpose of this document is to gain a better understanding of
the relationship between mobility, cognition and the ability to dual-task (or multi-task).
1.2 The Relationship between Attention, Automaticity and Dual-Task
Automaticity represents a skill that is performed with little demand on attentional
resources,3 whereas a task that is non-automatic would require significant attention to ensure
maintenance of performance. Automaticity can change with damage to the nervous system;
previously automatic movements may become attention-demanding4 and place an increased
load on cognitive resources.5 Attention, on the other hand, is the cognitive information
processing required to complete a task. This may include directing, shifting, dividing and
sustaining attention.3 Dual-task performance is a method used to measure automaticity by
determining the attentional costs of a given task or skill.1 For example, if tasks are “automatic”
and require little attention, multiple tasks can be easily performed concurrently without
negatively impacting performance. However, performance of “non-automatic” tasks will be
1
dependent upon the available attentional resources and may result in deterioration of
performance if there is insufficient attention to task, or if the tasks exceed the overall attention
capacity.5-7 Thus, automaticity is a feature of skilled task performance, while attention is the
mechanism by which task performance is maintained. The amount of attention needed for the
performance of a task is quantified by measuring dual-task performance. Dual-task performance
is typically measured by comparing time to complete the single-task condition to the time to
complete the dual-task condition.
1.3 Theories of Dual-Tasking & Divided Attention
There are several theories to explain the phenomenon of divided attention and dualtasking. The earliest proposed theories were the single-channel theories, including the
Bottleneck Theory of Dual-Task and the Filter Theory, which postulate that information is
processed serially and that two tasks with similar neural network requirements that are
performed concurrently create a “bottleneck” or interference such that performance is
decreased in one or both tasks.8-10 Single-channel theories are limited in their ability to explain
examples of dual-task situations where performance does not decline in either task.3
Several other theories better account for the attention-related limits on dual-task
performance: the central resource capacity theory and the multiple resource capacity theory.3 In
the central resource capacity theory, it is assumed that the information processing capability is
finite and limited for each individual; information processing occurs in parallel until this limit in
attention is reached,11 when performance would deteriorate in either one or both tasks.12 In this
theory, attentional resources are finite but flexible; they may be influenced by the difficulty of
2
the task, the novelty of the task or by the motivation of the individual.13 Indeed, the amount of
attention allotted to each task can be controlled by the individual and may be modified by
prioritizing one task over another. In contrast, the multiple resource capacity theory proposes
that individuals have multiple attentional resources available, each with its own capacity limit,3
as opposed to a central overall attentional pool. This theory has also been called the Cross-Talk
and Neural Theory, with declines in performance (i.e., interference) occurring when
simultaneous tasks require overlapping mental processes or neural structures.14 It further
suggests that dual-task related gait changes are influenced by the specific demands of the
cognitive task and the difficulty of the motor activity.4,12 The commonality between these two
theories is that attention has a finite capacity, allowing multiple tasks to be performed without
decrement until the limit in attentional resources is reached.15 Thus, the amount of attention
required to maintain performance can be used as an indicator of automaticity and in turn,
automaticity can be considered a feature of skilled performance that may be enhanced with
practice.
1.4 Dual-Task across Neurologic Populations: The Contribution of Cognitive and Motor Deficits
to Dual-Task Deficits and Fall Risk
1.4.1 Baseline Cognitive Deficits
Many individuals report attentional impairments either following or during the course of
a neurologic disorder. Indeed 25% of individuals with traumatic brain injury (TBI)1 and 22-25% of
individuals with multiple sclerosis (MS) report specific attentional deficits.2 Impaired attention
and information processing speed have been implicated as sensitive indicators of early cognitive
dysfunction in MS.16 While attention span is generally intact, many of the more complex aspects,
3
such as selective, divided, or alternating attention, are often impaired.16 Divided attention is the
ability to respond to multiple stimuli simultaneously and includes motor-motor, motorcognitive, and cognitive-cognitive tasks. In some neurologic conditions, it has been suggested
that the degree of impairment in divided attention is task dependent; indeed individuals with
TBI perform more poorly when the divided attention tasks require working memory.17 Not
surprisingly, cognitive dysfunction has been identified as a risk factor for accidental falls in
neurological populations, including those with Parkinson’s disease (PD).18
1.4.2 Baseline Mobility Deficits
Individuals with MS demonstrate increased postural sway in quiet stance, delayed
responses to postural perturbations and decreased ability to move toward their limits of
stability,19 leading to increased fall risk. These deficits have been linked with poor trunk control
in both sitting and standing19 as well as decreased core stability,20 even in those with very low21
or no22 clinical disability. The majority of falls in both elderly and individuals with PD23 occur in
the backward direction. Interestingly, eight of ten independently ambulating individuals with MS
have delayed postural reactions to backward translations in standing.24 This is perhaps explained
in part by the fact that poor proprioception underlies imbalance and falls in MS19 and that
backward walking and stepping requires increased proprioception when compared to forward
walking because of the lack of visual cues.25 Furthermore, a balance rehabilitation program
targeting both sensory and motor strategies improved dynamic balance and reduced fall
frequency more effectively than a program aimed at improving motor strategies only or a nonspecific general rehabilitation program.26
Deficits in postural control could contribute to the increase in gait variability commonly
seen in neurologic disorders. When compared to healthy controls, individuals with MS
4
demonstrate decreased velocity, step length, cadence and step time, as well as a wider base of
support27 and increased swing time variability in forward walking.28 In addition, 3D kinematic
assessment of gait under single-task conditions revealed reduced hip and knee extension and
decreased propulsive force in individuals with MS when compared to healthy controls.29 Even
minimally disabled MS individuals demonstrated prolonged double support phase during stance
phase.30 Individuals with MS produced an arrhythmic gait with variation in stance between legs
and an overall increase in sagittal plane hip motion with a reduction in ankle motion.30 Although
variability in gait parameters has been correlated with fall risk in the elderly,31-33 PD,34 and
Alzheimer’s disease,35 this relationship has not been established in MS. However, an interaction
between gait variability and cognition may be appreciated; individuals with MS devote greater
cognitive reserve to walking than do healthy controls, resulting in even greater variability in
spatiotemporal measures of gait (e.g., step length, velocity) when walking is paired with a
cognitive task2 as compared to both single-task conditions and to healthy controls. This is true
even in the early disease stage.36
The majority of accidental falls in MS occur during activities of daily life; factors that
have been linked to accidental falls were difficulty concentrating, forgetfulness, poor balance,
fatigue, and self-reported walking ability.37 In order to avoid falling, more than 80% of
individuals who reported fear of falling reduced their activity level,19 leading to even further
reductions in balance and walking ability. It is then not surprising that poor concentration38 and
difficulty with divided attention,2,8,37 have been linked with accidental falls in individuals with
MS.
5
1.4.3 Dual-Task Deficits in Individuals with Neurologic Disorders
Gait
Dual-task deficits have been described in MS,2 PD,9 Alzheimer’s disease,39 TBI,40-41
acquired brain injury,42-44 and elderly fallers.45 When a secondary cognitive task was added to
gait, individuals with PD9 and MS2 displayed significantly slower velocity and significantly
increased variability in swing time and stride time compared to controls. Swing time is
determined predominately by balance control mechanisms and is likely influenced not only by
lower performance on executive function tests, but also by the attention-demanding nature of
gait, even at a comfortable speed.4,9,45
Dual-task deficits were amplified in backward walking assessment in PD with greater
decrements in velocity, stride length, and swing percent, and a greater increment in stride
percent when compared to forward walking dual-tasking.23 Gait asymmetry (differences
between right and left limbs) has been linked to poor attentional resources assigned to gait;46
indeed, PD subjects demonstrate greater asymmetry and increased variability (differences in
step-to-step gait measures within a limb) across all challenging (backward walking and all dualtask) conditions when compared to healthy controls.47
Balance
Assessment of dual-task ability during balance activities has also been investigated. In
individuals with MS; only dual-tasking was able to discriminate between MS and healthy
controls in all conditions of postural difficulty (i.e. disturbed visual, vestibular and/or
somatosensory inputs).48 During balance dual-tasks, individuals with MS demonstrated greater
postural sway,49 sway area and sway velocity48 than age-matched healthy controls.
6
Interactions with Cognition
Executive function is a broad term to describe multiple cognitive processes, including
coordination of goal-directed activities, distribution of attention among competing tasks,50 and
performance of novel problem-solving tasks.51 Impairments in executive function may prevent
individuals from allotting appropriate attentional resources to balance and gait, reduce ability to
confront and adapt to challenging environments such as obstacles and uneven paths, result in
increased gait variability, and may account for the increased fall risk seen in the elderly,12,45 PD,9
and Alzheimer’s disease.39 Indeed, increasing the complexity of the cognitive task paired with
gait resulted in decreasing gait speeds52 in elderly adults. Interestingly, in MS, gait speed was the
variable most affected by dual-task conditions, and the degree of detriment was related not to
disease severity or duration but rather to levels of fatigue or general cognitive function.2
In summary, dual-tasking in individuals with neurologic conditions is characterized by
increased variability during both gait and balance activities. Gait represents an attentiondemanding task in individuals with neurologic impairment, and the ability to divide attention is
further influenced by level of executive function. Current recommendations from both the
animal literature53 and human trials54 state that interventions combining motor/physical activity
and cognitive therapy focusing on complex and novel tasks should be incorporated into clinical
practice to enable patients to move more safely and with a reduced fall risk.
1.5 Strategies to Compensate for Dual-Task Deficits
Individuals cope with dual-tasking during gait by using different strategies; young adults
decrease gait speed, elderly non-fallers decrease both gait speed and swing time, and elderly
fallers attempt to mimic non-fallers but are ultimately unsuccessful at maintaining a stable
7
walking pattern, which results in highly variable walking. It has been suggested that individuals
with neurological deficits employ a “posture second” strategy rather than a “posture first”
strategy when encountered with a dual-task situation.9,55 The cognitive secondary task is
prioritized over the postural task (i.e. often gait). This leads to impaired multiple-task
performance and has been shown to double the risk of sustaining a fall while performing an ADL
in PD.55 In contrast, decrement in dual-task performance was not related to task prioritization in
one study in MS,2 but this area has not been well explored. Contradictory evidence suggests that
despite gait impairments, individuals with chronic stroke use a posture first strategy only when
crossing obstacles on a treadmill with a secondary auditory task,56 but utilize the posture second
strategy when the secondary task requires working memory, speech or visuospatial cognition
during gait.57 However, the discrepancy could come from the difference between gait over level
ground compared to gait with obstacles, which could potentially apply to walking over irregular
terrain while simultaneously negotiating obstacles and having a conversation. This is in
agreement with previous literature that suggests the nature and type of primary and secondary
task in dual-task performance may also affect the result.58 Indeed, attentional demand increased
with more challenging postural control tasks,59 and the type of cognitive task (artithmetic tasks
result in greater coefficients of variation than fluency tasks) resulted in differential dual-task
related changes in gait in elderly individuals.58
Taken as a whole, strategies to compensate for dual-task deficits in MS are not well known;
rather, gait should be viewed as a complex task utilizing executive function, particularly when
additional tasks are performed simultaneously.39,60 Thus, in MS, gait speed or variability may
prove to be a sensitive marker of fall risk and enhancement of cognitive functioning may reduce
fall risk. Clinically, one can infer that individuals with neurologic disorders may need specific
8
instruction to focus on walking to avoid compromising gait and safety,57 or perhaps more
importantly, they may need to practice specific dual-task training.
1.6 Measurement of Dual-Task Ability
Assessment and treatment with tasks that challenge cognition and uncover or amplify
deficits such as increased gait variability or poor balance could identify potential fallers. As
previously discussed, dual-tasks have been shown to amplify gait variability in many neurologic
populations, including MS. Although backward walking has been explored formally only in
individuals with PD, it was also shown to increase gait variability, and should be further
explored. To this investigator’s knowledge, dual-task backward walking has only been explored
in Parkinson’s disease, but could be a useful assessment in other neurologic populations.
Deficits in dual-task ability highlight its importance in the assessment of motor function
in individuals with neurologic disorders. A patient who recovers independence in gait in a quiet,
closed environment may not be safe when walking in a more demanding environment where
cognitive distractors or uneven terrains are present.61 In order to appropriately measure
improvements in automaticity as a result of rehabilitation, therapists should routinely measure
it to ensure the effectiveness of interventions designed to improve automaticity.3
1.6.1 Behavioral Measures
GAITRite
The GAITRite System (V3.9, MAP/CIR Inc.) is a 4.88 meter electronic walkway with
16,128 embedded sensors that records footfalls in real time (Figure 1.1). The GAITRite calculates
spatiotemporal measures (e.g., velocity, step length, double support) of gait in real time. The
GAITRite is reliable and valid for use in individuals with MS.27 Spatiotemporal impairments
9
detected by GAITRite have been correlated with associated neurological disability (EDSS).27,62
Temporal and spatial gait measurements from the GAITRite walkway are sampled at the default
rate of 60 hertz.
Figure.1.1 GAITRite visual output from a representative subject walking forward at their
comfortable pace. Right feet are represented in purple and left feet are represented in green.
Timed Up & Go (TUG); TUG Cognitive; TUG Manual
The Timed Up and Go (TUG) is a common test to examine mobility in elderly adults.63
Time to complete the TUG, which requires standing from a chair, walking 10 feet, turning,
walking back and sitting down, is correlated with level of mobility63 in elderly adults. The TUG
has been utilized in MS populations and correlates with longer walking tests.64 The TUG manual
is the TUG while carrying a glass of water while the TUG cognitive is the TUG while performing a
subtraction task. These modifications of the TUG measure dual-task performance and are
sensitive and specific for identifying fall risk in elderly.65 Older adults who take longer than 13.5
seconds to complete the TUG are classified as “fallers” with a 90% prediction rate.65 Those
taking >14.5 seconds on the TUG manual and >15 seconds on the TUG cognitive were classified
as fallers with a 90% and 86.7% prediction rate, respectively.65 The TUG cognitive accurately
predicted fallers in MS with a sensitivity of 73%.18
10
Walking While Talking Test (WWTT)
The WWTT is also a reliable and valid test to identify elderly adults at high risk for falls.66
The test is performed and timed under three conditions: 1) walk 40 feet with a 180-degree turn
at the midpoint; 2) condition 1 + recite the alphabet aloud (WWT-simple); 3) condition 1 + recite
alternate letters of the alphabet aloud (WWT-complex).66 Poor performance on the WWTsimple and complex was strongly predictive of falls; a cut-off of 33 seconds for the WWTcomplex predicted fallers with a sensitivity of 38.5% and a specificity of 95.6%. The WWTT has
not been validated nor have cut-offs been established in individuals with MS.
Dual Task Questionnaire (DTQ)
The Dual Tasking Questionnaire is a 10 item subjective questionnaire of everyday tasks
involving dual-tasking, to be rated for frequency of difficulty from 0-4 (4=highest frequency).44 It
has been shown to be sensitive to the effects of brain injury and Alzheimer’s disease.44
1.6.2 Cognitive Measures
One of the most commonly used neuropsychological tests to assess attention and
processing speed in individuals with MS is the Symbol Digit Modality Test (SDMT). The SDMT is
an easy test to administer clinically.67 During the SDMT, subjects are presented with nine
symbol-digit pairings and asked to match as many symbols to their corresponding number as
possible in 90 seconds. The SDMT is a sensitive test68-69 of processing speed, visual tracking,
rapid decision making,70 and divided attention.71 The SDMT is valid and reliable in MS68,72-73 and
scores are highly correlated to MRI-derived measures of disease burden in MS patients.68
The phonemic fluency word generation task requires participants to generate as many
words beginning with the letter “F” as possible in 60 seconds. This is followed by generating
11
words with A and S. The total number of responses in three 60 second bouts is recorded as the
overall score.71 The word generation task is a test of working memory and executive function.71
During the Stroop task,74 participants are presented with three different conditions: the
congruent condition, where the color and the word match (e.g., the word RED printed in red
ink); the neutral condition, where color words are written in black ink; and the incongruent
condition, where the ink color will be incongruent with the meaning of the word (e.g., the word
GREEN printed in blue ink). For each condition the accuracy of stated words or colors provided is
recorded as the participant’s score. The Stroop is a test of executive function, selective
attention, processing speed, and response inhibition.71
1.6.3 Imaging Measures
Magnetic resonance imaging (MRI) is the most commonly used assessment tool for both
investigation of underlying pathology and evaluation of disease progression in MS.75
Anatomical Scans: T1 MPRAGE, T2 FLAIR
Structural changes in grey and white matter have been widely documented in
individuals with MS. Indeed, brain atrophy has been linked to declines in cognitive status in
MS76-78 and is predictive of both future cognitive impairment in early relapsing-remitting MS
(RRMS)78-79 and progressive accumulation of physical disability.78 Interestingly, grey matter
volume loss in areas associated with executive function has been significantly associated with
deteriorating cognitive performance80-81 and white matter lesion progression.80 Accordingly,
increased gray matter volume is positively associated with cardiorespiratory fitness.82 Higher
fitness levels may exert a prophylactic influence on structural decline observed in early RRMS by
preserving neuronal integrity and reducing long-term disability.82
Diffusion Tensor Imaging (DTI)
12
The diffusion of molecular water in grey matter is highly isotropic (i.e., the same in all
directions), while in white matter, diffusion is often restricted to a particular direction. This
anisotropy is dictated by axonal cell membranes, myelin sheaths, neurofilaments, and other
structures.83 Anisotropy is highest in compact white matter pathways where the fiber bundles
are oriented in parallel, such as the corpus callosum or pyramidal tracts;83 a higher fractional
anisotropy (FA) value indicates a greater degree of white matter integrity. FA is the most
commonly used anisotropy index in the literature and the FA index is normalized to values of 0
(isotropic) to 1 (anisotropic).84 FA is calculated from the diffusion directions (Figure 1.2) Thus,
DTI is a useful way to explore white matter integrity in MS. Compared to healthy controls,
individuals with RRMS have lower FA values86-88 that correlate well with both motor87 and
cognitive88-89 impairment.
e2
2
e3
3
e1
1
Figure 1.2 DTI contrast generation. Diffusion measurements along multiple
axes contribute to the shape and orientation of the “diffusion ellipsoid”. From
the estimated ellipsoid, the longest axis can be found (eigenvector 1 or e1)
which is assumed to correspond with the local fiber orientation. 85 The
relative size of each axis determined by the eigenvalues of the tensor (1, 2
and 3).
13
In addition, the mean diffusivity (MD), axial diffusivity (AD) and radial diffusivity (RD)
provide useful information about the tract or white matter region of interest. The AD is
equivalent to the primary eigenvalue (1) and indicates the direction of highest diffusivity and
coincides with the fiber tract axis.90-91 The RD is calculated by averaging the mean of the other
two eigenvalues (2 and 3). Animal work suggests that the AD may be specific to axonal
degeneration while the RD may be modulated by myelin.92 The MD is the average of all three
eigenvalues, and is therefore not indicative of direction. Diffusion images can also be used to
perform tractography, the only available tool for identifying and measuring white matter
pathways non-invasively.93Tractography identifies the path along which diffusion is least
hindered and thus allows for the exploration of connectivity and for the measurement of
diffusivity along a tract of interest.
1.7 Research Aims
Although dual-task deficits have been widely documented in many neurological
populations, there is a paucity of evidence exploring the effectiveness of interventions to
ameliorate these deficits. Similarly, the interaction between motor and cognitive deficits and the
neural connectivity related to dual-task deficits are largely unknown (Figure 1.3).
14
Increased
Gait Variability
Contributing Factors
Decreased Balance
Motor
Factors
Difficulty Walking
Backwards
Difficulty Dual-Tasking
Cognitive
Factors
Poor Divided Attention
Decreased Executive
Function
Increased
Fall Risk
Figure 1.3 Schema demonstrating factors contributing to increased gait variability and fall
risk.
To address these gaps in knowledge, the objective of this work will be to examine the
interaction between motor and cognitive factors contributing to gait variability and falls and to
explore the neural connectivity related to dual-task ability. It is hypothesized that motor and
cognitive factors contribute to increased gait variability and could increase fall risk in individuals
with both minimal and severe baseline deficits. It is further hypothesized that the underlying
neural connectivity will be related to functional measures of dual-task, which could improve
clinical prediction and rehabilitation of such deficits (Figure 1.4).
15
Contributing Factors
Chapter 2
Increased
Gait Variability
Systematic Review
GAPS
IDENTIFIED
Efficacy of
Training
Interaction
of Motor
and
Cognitive
Decreased Balance
Difficulty Walking
Motor
Factors
Severity of
Injury
Backwards
Difficulty Dual-Tasking
Poor Divided Attention
16
Mild Severe
Training in
Multiple
Sclerosis
Cognitive
Factors
Decreased Executive
Function
Increased
Fall Risk
Backward
Walking
Cross
Sectional
Study
Chapter 3
Dual-Task
Training in
Severe TBI
(Case Study)
Relationship
between gait
variability and
cognitive
factors in MS
Relationship
between
Imaging and
Clinical
Measures
in MS
Chapter 4
Chapter 5
Chapter 6
Figure 1.4 Schema demonstrating the gaps in knowledge identified by our systematic review (Chapter 2) that directed this
document’s research goals. Note that the dual-task training in MS study is ongoing and will not be presented here.
1
Dual-Task
Training in
MS
The following research goals are proposed to test these hypotheses and lay a foundation for
further studies:
1. Describe current rehabilitation models and recommendations to improve motor-cognitive
dual-task deficits in individuals with neurologic conditions.
Chapter 2: The Effect of Motor-Cognitive Dual-Task Interventions on Mobility in
Neurologic Disorders: A Systematic Review
2. Explore the relationship between gait variability in forward and backward walking and falls in
a population with mild impairment (i.e. cross-sectional dataset including young, middle-aged
and both healthy and frail elderly adults)
Chapter 3: Backward Walking Measures Are Sensitive to Age-Related Changes in
Mobility and Balance
3. Examine the effect of a dual-task training program on mobility, balance and dual-task ability
in an individual with severe impairment following a traumatic brain injury.
Chapter 4: Dual-Task Training for Balance and Mobility in a Person with Severe
Traumatic Brain Injury: A Case Study
4. Explore the interaction between gait variability, cognitive factors (neuropsychological test
performance), and falls in individuals with MS and healthy adults.
Chapter 5: Dual-Task Walking Variability is Associated with Stroop and Dual-Task
Questionnaire Performance in Multiple Sclerosis
17
5. Examine the relationships between neuroimaging measures and clinical measures of dual-task
ability to further understanding of the underlying neural connectivity and the potential
relationships to dual-task ability.
Chapter 6: SMA Connectivity May Be Associated With Dual-Task Walking in
Multiple Sclerosis
18
1.8 References
1. Foley JA, Cantagallo A, Della Sala S, Logie RH. Dual task performance and post traumatic
brain injury. Brain Injury. 2010; 24(6): 851-858.
2. Hamilton F, Rochester L, Paul L et al. Walking and talking: an investigation of cognitivemotor dual tasking in multiple sclerosis. Mult Scler. 2009; 15(10): 1215-1227.
3. Paul SS, Ada L, Canning CG. Automaticity of walking—implications for physiotherapy
practice. Phys Ther Reviews. 2005; 10: 15-23.
4. Yogev-Seligmann G, Hausdorff J, Giladi N. The role of executive function and attention in
gait. Movement Disord. 2008; 23(3):329–342.
5. Mulder T, Nienhuis B, Pauwels J. The assessment of motor recovery: a new look at an old
problem. J Electromyogr Kinesiol. 1996;6:137–45.
6. Bebko JM, Demark JL, Osborne PA, Majumder S, Ricciuti CJ, Rhee T. Acquisition and
automatization of a complex task: an examination of three-ball cascade juggling. J Motor
Behav. 2003;35:109–18.
7. Roskell C, Cross V. Attention limitation and learning in physiotherapy. Physiotherapy
1998;84:118–25.
8. D‘Esposito M, Onishi K, Thompson H et al. Working memory impairments in multiple
sclerosis: evidence from a dual-task paradigm. Neuropsychology. 1996; 10(1): 51-56.
9. Yogev G, Giladi N, Peretz, Springer S, Simon ES, Hausdorff JM. Dual-tasking, gait rhythmicity,
and Parkinson’s disease: which aspects of gait are attention demanding. Eur J
Neurosci.2005; 22: 1248-1256.
10. Al-Yahya E, Dawes H, Collett J, Howells K, Izadi H, Wade DT, Cockburn J. Gait adaptations to
simultaneous cognitive and mechanical constraints. Exp Brain Res. 2009; 199: 39- 48.
11. O’Shea S, Morris ME, Iansek R. Dual task interference during gait in people with Parkinson
disease: effect of motor versus cognitive secondary tasks. Phys Ther 2002;82:888–97.
12. 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.
13. Yogev-Seligmann G, Hausdorff JM, Giladi N. Do we always prioritize balance when walking?
Towards an integrated model of task prioritization. Mov Disord. 2012; 27(6): 765-770.
19
14. Klingberg T. Concurrent performance of two working memory tasks: potential mechanisms
of interference. Cereb Cortex. 1998; 8(7):593-601.
15. Yardley L, Gardner M, Bronstein A, Davies R, Buckwell D, Luxon L. Interference between
postural control and mental task performance in patients with vestibular disorder and
healthy controls. J Neurol Neurosurg Ps. 2001;71:48–52.
16. Amato MP, Zipoli V, Portaccio E. Cognitive changes in multiple sclerosis. Exp Rev
Neurotherapeutics. 2008; 8: 1585-1596.
17. Asloun S, Soury S, Couillet J, Giroire JM, Joseph PA, Mazaux JM, Azouvi P. Interactions
between divided attention and working-memory load in patients with severe traumatic
brain injury. J Clin Exp Neuropsyc. 2008; 30(4): 481-490.
18. Nilsagård Y, Lundholm C, Denison E, Gunnarsson LG. Predicting accidental falls in people
with multiple sclerosis—a longitudinal study. Clin Rehabil. 2009; 23: 259-269.
19. Cameron MH, Lord S. Postural control in multiple sclerosis: implications for fall prevention.
Curr Neurol Neurosci. 2010; 10: 407-412.
20. Freeman JA, Gear M, Pauli A, Cowan P, Finnigan C, Hunger H, et al. The effect of core
stability training on balance and mobility in ambulant individuals with multiple sclerosis: a
multi-centre series of single case studies. Mult Scler. 2010; 16(11): 1377- 1384.
21. Fjeldstad C, Pardo G, Bemben D, Bemben M. Decreased postural balance in multiple
sclerosis patients with low disability. Int J Rehabil Res. 2011; 34(1): 53-58.
22. Martin Cl, Phillips BA, Kilpatric TJ, Butzkeuven H, Tubridy N, McDonald E, Galea MP. Gait and
balance impairment in early multiple sclerosis in the absence of clinical disability. Mult Scler.
2006; 12: 620-628.
23. Hackney ME, Earhart GM. Backward walking in Parkinson disease. Movement Disord. 2009;
24(2): 218-223.
24. Cameron MH, Horak FB, Herndon RR, Bourdett D. Imbalance in multiple sclerosis: a result of
slowed spinal somatosensory conduction. Somatosens Mot Res. 2008; 25: 113-122.
25. Thomas MA, Fast A. One step forwards and two steps back: the dangers of walking
backward in therapy. Am J Phys Med Rehab. 2000; 79(5): 459-461.
26. Cattaneo D, Jonsdottir J, Zocchi M, Regola A. Effects of balance exercises on people with
multiple sclerosis: a pilot study. Clin Rehabil. 2007; 21: 771-781.
20
27. Givon U, Zeilig G, Achiron A. Gait analysis in multiple sclerosis: characterization of temporalspatial parameters using GAITRite functional ambulation system. Gait Posture. 2009; 29:
138-142.
28. Crenshaw SJ, Royer TD, Richards JG, Hudson DJ. Gait variability in people with multiple
sclerosis. Mult Scler. 2006; 12: 613-619.
29. Kelleher KJ, Spence W, Solomonidis S, Apatsidis D. The characterisation of gait patterns of
people with multiple sclerosis. Disabil Rehabil. 2010; 32(15): 1242-1250.
30. Benedetti MG, Piperno R, Simoncini L, Bonato P, Tonini A, Giannini S. Gait abnormalities in
minimally impaired multiple sclerosis patients. Mult Scler. 1999; 5(5): 363- 368.
31. Hausdorff JM, Edelberg HK, Mitchell SL et al. Increased gait unsteadiness in communitydwelling elderly fallers. Arch Phys Med Rehab. 1997; 78: 278-283.
32. Maki BE. Gait variability in older adults: predictors of falls or indicators of fear. J Am Geriatr
Soc. 1997; 45: 313-320.
33. Hausdorff JM, Rios D, Edelberg HK. Gait variability and fall risk in community-living older
adults: a 1 year prospective study. Arch Phys Med Rehab. 2001; 82: 1050-1056.
34. Schaafsma JD, Giladi N, Balash Y et al. Gait dynamics in Parkinson‘s disease: relationship to
Parkinsonian features, falls and response to levadopa. J Neurol Sci. 2003; 212: 47-53.
35. Nakamura T, Megura K, Sasaki H. Relationship between falls and stride length variability in
senile dementia of the Alzheimer type. Gerontology. 1996; 42: 108-113.
36. Kalron A, Dvir Z, Achiron A. Walking while talking—difficulties incurred during the initial
stages of multiple sclerosis disease process. Gait Posture. 2010; 32(3) 332-335.
37. Nilsagård Y, Denison E, Gunnarsson LG, Bostrom K. Factors perceived as being related to
accidental falls in persons with multiple sclerosis. Disabil Rehabil. 2009; 31: 1301-1310.
38. Finlayson ML, Peterson EW, Cho CC. Risk factors for falling among people aged 45 to 90
years with multiple sclerosis. Arch Phys Med Rehab. 2006; 87(9): 1274-1279.
39. Cocchini C, Della Sala S, Logie RH, Pagani R, Sacco L, Spinnler H. Dual task effects of walking
when talking in Alzheimer’s disease. Revue Neurologique (Paris). 2004; 160(1): 74-80.
40. Cicerone KD. Attention deficits and dual task demands after mild traumatic brain injury.
Brain Injury. 1996;10:79–89.
41. Azouvi P, Couillet J, Leclercq M, et al. Divided attention and mental effort after severe
traumatic brain injury. Neuropsychologia. 2004;42:1260–1268.
21
42. McCulloch K. Attention and dual-task conditions: physical therapy implications for
individuals with acquired brain injury: J Neurol Phys Ther. 2007; 31: 104-118.
43. McCulloch KL, Buxton E, Hackney J, Lowers S. Balance, attention, and dual-task performance
during walking after brain injury: association with falls history. J Head Trauma Rehab. 2010;
25(3): 155-163.
44. Evans JJ, Greenfield E, Wilson BA, Bateman A. Walking and talking therapy: improving
cognitive-motor dual-tasking in neurological illness. J Int Neuropsych Soc.. 2009; 15: 112120.
45. Springer S, Giladi N, Peretz C, Yogev G, Simon ES, Hausdorff JM. Dual-tasking effects on gait
variability: the role of aging, falls and executive function. Movement Disord. 2006; 21(7):
950-957.
46. Yogev G, Plotnik M, Peretz C, Giladi N, Hausdorff JM. Gait asymmetry in patients with
Parkinson’s disease and elderly fallers: when does the bilateral coordination of gait require
attention? Exp Brain Res. 2007; 177: 336-346.
47. Hackney ME, Earhart GM. The effects of a secondary task on forward and backward walking
in Parkinson disease. NeurorehabNeural Re. 2010; 24(1):97-106.
48. Negahban H, Mofateh R, Arastoo AA, Mazaheri M, Yazdi MJS, Salavati M, Majdinasab N. The
effects of cognitive loading on balance control in patients with multiple sclerosis. Gait
Posture. 2011; 34(4): 479-484.
49. Porosinska A, Pierzchala K, Mentel M, Karpe J. Evaluation of postural balance control in
patients with multiple sclerosis—effect of different sensory conditions and arithmetic task
execution. A pilot study. Neurologia i Neurochirurgia Polska. 2010; 44:35–42.
50. Royall DR, Lauterbach EC, Cummings JL, et al. Executive control function: a review of its
promise and challenges for clinical research—a report from the Committee on Research of
the American Neuropsychiatric Association. Journal of Neuropsych Clin N. 2002; 14: 377-405.
51. Sbordone RJ. The executive functions of the brain. In Groth-Marnat G, editor.
Neuropsychological assessment in clinical practice: a guide to test interpretation and
integration. New York: John Wiley and Sons; 2000: 437-456.
52. Hall CD, Echt KV, Wolf SL, Rogers WA. Cognitive and motor mechanisms underlying older
adults’ ability to divide attention while walking. Phys Ther. 2011; 91(7): 1039-1050.
53. Kempermann G, Fabel K, Ehninger D, Babu H, Leal-Galicia P, Garthe A, Wolf SA. Why and
how physical activity promotes experience-induced brain plasticity. Frontiers Neurosci.
2010; 4(189): 1- 9.
22
54. Segev-Jacubovski O, Herman T, Yogev-Seligmann G, Mirelman A, Giladi N, Hausdorff JM. The
interplay between gait, falls and cognition: can cognitive therapy reduce fall risk? Exp Rev
Neurotherapeutics. 2011; 11(7): 1057-1075.
55. Bloem BR, Grimbergen YAM, van Dijk JG, Munneke M. The “posture second” strategy: a
review of wrong priorities in Parkinson’s disease. J NeurolScis. 2006; 248: 196-204.
56. Smulders K, van Swigchem R, de Swart BJM, Geurts ACH, Weerdesteyn V. Communitydwelling people with chronic stroke need disproportionate attention while walking and
negotiating obstacles. Gait Posture.2012; 36: 127-132.
57. Plummer-D’Amato P, Altmann LJP, Saracino D, Fox E, Behrman AL, Marsiske M. Interactions
between cognitive tasks and gait after stroke: a dual task study. Gait Posture. 2008; 27: 683688.
58. Beauchet O, Dubost V, Aminian K, Gonthier R, Kressig RW. Dual-task-related gait changes in
the elderly: does the type of cognitive task matter? J Mot Behav 2005; 37:259-264.
59. Lajoie Y, Teasdale N, Bard C, Fleury M. Attentional demands for static and dynamic
equilibrium. Exp Brain Res. 1993; 97:139-144.
60. Hausdorff JM, Yogev G, Springer S, Simon ES, Giladi N. Walking is more like catching than
tapping: gait in the elderly as a complex cognitive task. Exp Brain Res. 2005; 164: 541-548.
61. Haggard P, Cockburn J, Cock J, Fordham C, Wade D. Interference between gait and cognitive
tasks in a rehabilitating neurological population. J Neurol Neurosurg Ps. 2000; 69: 479-486.
62. Sosnoff JJ, Weikert M, Dlugonski D, Motl RW. Quantifying gait impairment in multiple
sclerosis using GAITRite technology. Gait Posture. 2011; 34: 145-147.
63. Podsiadlo D, Richardson S. The Timed Up & Go‖: a test of basic functional mobility for frail
elderly persons. J Am Geriatr Soc. 1991; 39: 142-148.
64. Nilsagård Y, Lundholm C, Gunnarsson LG, Denison E. Clinical relevance using timed walk
tests and ‘timed up and go‘ testing in persons with multiple sclerosis. Physiotherapy Res Int.
2007; 12(2):105-114.
65. Shumway-Cook A, Brauer S, Woollacott M. Predicting the probability for falls in communitydwelling older adults using the timed up & go test. Phys Ther. 2000; 80(9): 896-903.
66. Verghese J, Buschke H, Viola L, Katz M, Hall C, Kuslansky G, Lipton R. Validity of divided
attention tasks in predicting falls in older individuals: a preliminary study. J Am Geriatr Soc.
2002; 50: 1572-1576.
23
67. Hoffmann S, Tittgemeyer M, Yves von Cramon D. Cognitive impairment in multiple sclerosis.
Curr Opin Neurol.. 2007; 20: 275-280.
68. Parmenter BA, Weinstock-Guttman B, Garg N, Munschauer F, Benedict RHB. Screening for
cognitive impairment in multiple sclerosis using the symbol digit modality test. Mult Scler.
2007; 13: 52-57.
69. Deloire MS, Bonnet MC, Salort E, et al. How to detect cognitive dysfunction at early stages
of multiple sclerosis? Mult Scler. 2006; 12: 445-452.
70. Genova HM, Hillary FG, Wylie G, Rypma B, DeLuca J. Examination of processing speed
deficits in multiple sclerosis using functional magnetic imaging. J Int Neuropsych Soc. 2009;
15: 383-393.
71. Strauss E, Sherman EMS, Spreen O. A compendium of neuropsychological tests:
administration, norms, and commentary. 3rd ed. New York: Oxford University Press; 2006.
72. Drake AS, Weinstock-Guttman B, Morrow SA, Hojnacki D, Munschauer FE, Benedict RHB.
Psychometrics and normative data for the Multiple Sclerosis Functional Composite:
replacing the PASAT with the symbol digit modalities test. Mult Scler. 2010; 16(2): 228-237.
73. Rao SM. Cognitive function in patients with multiple sclerosis: impairment and treatment.
Int J Mult Scler Care. 2004; 1: 9-22.
74. Rao SM, Leo GJ, Bernardin L, Unverzagt F. Cognitive dysfunction in multiple sclerosis I:
frequency, patterns, and prediction. Neurology. 1991; 41(5): 685-691.
75. Bakshi R, Thompson AJ, Rocca MA et al. MRI in multiple sclerosis: current status and future
prospects. Lancet Neurol. 2008; 7: 615-625.
76. Filippi M, Rocca MA, Benedict RHB, DeLuca J, Geurts JJG, Rombouts SARB, Ron M, Comi G.
The contribution of MRI in assessing cognitive impairment in multiple sclerosis. Neurology.
2010; 75: 2121-2128.
77. Chiaravalloti ND, DeLuca J. Cognitive impairment in multiple sclerosis. Lancet. 2008; 7:
1139- 1151.
78. Zivadinov R, Sepcic J, Nasuelli D, De Masi R, Monti Bragadin L, Tommasi MA, et al. A
longitudinal study of brain atrophy and cognitive disturbances in the early phase of
relapsing-remitting multiple sclerosis. J Neurol Neurosur Ps. 2001; 70: 773-780.
79. Summers MM, Fisniku LK, Anderson VM, Miller DH, Cipolotti L, Ron M. Cognitive impairment
in relapsing-remitting multiple sclerosis can be predicted by imaging performed several
years earlier. Mult Scler. 2008; 14: 197-204.
24
80. Filippi M, Rocca MA. MR imaging of gray matter involvement in multiple sclerosis:
implications for understanding disease pathophysiology and monitoring treatment efficacy.
Am J Neuroradiol.. 2010; 31:1171-1177.
81. Amato MP, Portaccio E, Goretti B, Zipoli V, Battaglini M, Bartolozzi ML, Stromillo ML, Guidi L,
Siracusa G, Sorbi S, Federico A, De Stefano N. Association of neocortical volume changes
with cognitive deterioration in relapsing-remitting multiple sclerosis. Arch Neurol. 2007;
64(8): 1157-1161.
82. Prakash RS, Snook EM, Motl RW, Kramer AF. Aerobic fitness is associated with gray matter
volume and white matter integrity in multiple sclerosis. Brain Res. 2010; 1341: 41- 51.
83. Madden DJ, Bennett IJ, Song AW. Cerebral white matter integrity and cognitive aging:
contributions from diffusion tensor imaging. Neuropsychol Rev. 2009; 19: 415- 435.
84. Jones DK. Gaussian modeling of the diffusion signal. In Johansen-Berg H, Behrens TEJ.
Diffusion MRI: from quantitative measurement to in vivo neuroanatomy. 1st ed. London, UK:
Elsevier Inc; 2009:37-54.
85. Mori S, Zhang J. Principles of diffusion tensor imaging and its application to basic
neuroscience research. Neuron. 2006; 51: 527-539.
86. Wilson M, Trench CR, Morgan PS, Blumhardt LD. Pyramidal tract mapping by diffusion
tensor magnetic resonance imaging in multiple sclerosis: improving correlations with
disability. J Neurol Neurosur Ps. 2003; 74: 203-207.
87. Reich DS, Zackowski KM, Gordon-Lipkin EM, Smith SA, Chadkowski BA, Cutter GR, Calabresi
PA. Corticospinal tract abnormalities are associated with weakness in multiple sclerosis. Am
J Neuroradiol. 2008; 29: 333-339.
88. Ibrahim I, Tintera J, Skoch A, Jirů F, Hlustik P, Martinkova P, Zvara K, Rasova K. Fractional
anisotropy and mean diffusivity in the corpus callosum of patients with multiple sclerosis:
the effect of physiotherapy. Neuroradiology. 2011; 53(11): 917-926.
89. Lin X, Tench CR, Morgan PS, Constantinescu CS. Use of combined conventional and
quantitative MRI to quantify pathology related to cognitive impairment in multiple sclerosis.
J Neurol Neurosur Ps. 2008; 79: 437-441.
90. Basser PJ, Mattiello J, LeBihan D.MR diffusion tensor spectroscopy and imaging. Biophys J.
1994;66:259-267.
91. Beaulieu C, Allen PS. Determinants of anisotropic water diffusion in nerves. Magn Reson
Med. 1994;31:394-400.
25
92. Song SK, Sun SW, Ramsbottom MJ, Chang C, Russell J, Cross AH. Dysmyelination revealed
through MRI as increased radial (but unchanged axial) diffusion of water. Neuroimage. 2002;
17: 1429-1436.
93. Behrens TEJ and Jbabdi S. MR diffusion tractography. In: Johansen-Berg H and Behrens TEJ,
eds. Diffusion MRI: From quantitative measurement to in vivo neuroanatomy. 1st ed.
London, UK: Elsevier Academic Press: 2009: 333-351.
26
Chapter 2: The Effect of Motor-Cognitive Dual-Task Interventions on Mobility in Neurologic
Disorders: A Systematic Review
2.1 Introduction
Attention is a complex cognitive process contributing to processing speed, working
memory, and shifting and dividing attention.1Divided attention is the ability to respond to
multiple stimuli simultaneously.2,3Dual-tasks are paired stimuli such as a cognitive task and a
motor task (e.g., walking while talking) that require divided attention. Deficits in sustained and
divided attention are common in neurologic disorders and have specifically been linked with
impairments in functional mobility in traumatic brain injury (TBI),4,5 acquired brain injury,6
multiple sclerosis (MS),7 and aging.8
Addition of a secondary cognitive task to mobility tasks such as gait or balance tends to
amplify variability in individuals with neurologic disorders. Indeed, under dual-task (DT)
conditions, individuals with Parkinson’s disease (PD)9 and MS7 demonstrated significantly
increased variability in swing time and stride time compared to controls. Because swing time is
determined predominately by balance control mechanisms, it is likely influenced not only by
altered or diminished executive function, but also by the attention-demanding nature of gait.9-11
Impairments in attention may prevent individuals from allotting appropriate attentional
27
resources to balance and gait, reduce ability to confront and adapt to challenging environments
such as obstacles and uneven paths, result in increased gait variability, and account for
increased fall risk in the elderly,8,11 PD,9 and Alzheimer’s disease.12 Interestingly, only dualtasking was able to discriminate between individuals with MS and healthy controls in all
conditions of postural difficulty (i.e. disturbed visual, vestibular and/or somatosensory inputs).13
During balance dual-tasks, individuals with MS demonstrated greater postural sway, 14 sway
area and variability of sway velocity13 than age-matched healthy controls.
It has been suggested that impaired central integration is a primary mechanism of
impaired balance and that DT training may be an effective way of addressing central processing
deficits.15 Despite the documented deterioration in gait and balance under DT conditions, there
are few intervention studies that address this deficit. Available studies are marked by variability
of both training type and duration. Recently, a virtual reality training program of cognitivemotor dual-tasks in a single individual following a concussion, resulted in improvements in
dynamic and static balance.16 Similarly, DT training improved balance during cognitive activities
to a greater extent than mobility training alone in healthy individuals and those with
concussion.17,18 Neuropsychological training programs in subacute and chronic severe TBI,
focusing either on cognitive dual-tasks19 or working memory,20 resulted in improvements in
reaction times on visual-auditory dual-tasks. The purpose of this systematic review was to
examine the relevant literature that addresses motor-cognitive DT training interventions in
individuals with neurological deficits.
28
2.2 Methods
2.2.1 Data Sources and Searches
Two of the authors conducted the literature search, applying the MeSH search terms
“nervous system disease(s) OR neurologic disorder(s)” with “rehabilitation OR intervention” and
“dual task(s) OR divided attention OR multi task(s)” to the title, abstract or index term fields.
These search terms were applied to the following data bases: Biosis, CINAHL, ERIC, PsychInfo,
Psychological & Behavioral, PubMed, Scopus, and Web of Knowledge. All databases were
searched from their inception until May 2, 2012. Two investigators independently screened the
titles of the publications found in the databases. If either investigator believed that a title
potentially met the inclusion criteria, or if there was inadequate information to make a decision,
a copy of the article was obtained. In addition, two investigators manually searched the
reference lists of the retrieved articles for potential studies that may have been overlooked or
absent from databases.
2.2.2 Study Selection
Figure 2.1 outlines the number of references considered at each state of the selection
process prior to confirming the included trials. A total of 12 studies21-32 were identified as eligible
for inclusion; the details of excluded studies are shown in Figure 2.1. Inclusion in this review was
restricted to studies of adults (>18 years old) with a central neurologic disorder with both a
treatment and comparison group or an extended baseline in the case of a repeated measures
design. To be included, the training must describe DT training (either cognitive-motor or motormotor) and include outcomes of either gait/balance or both mobility and cognition. Due to the
small number of studies returned, the investigators also opted to include studies of elderly
adults with balance impairments or a history of falls.
29
2.2.3 Data Extraction & Quality Assessment
Data were independently extracted from the included trials by two authors and
recorded on a standardized spreadsheet. Extracted information included sample sizes, trial
settings, population characteristics (e.g., age, diagnosis, and inclusion criteria), details of the
interventions, and outcome data (mean scores, standard deviations and ranges). In one trial,32
there was insufficient information on the outcomes and study population; the authors of that
study were contacted, but no response was received.
30
Initial electronic search
(conducted by two authors)
BIOSIS
CINAHL
Cochrane
ERIC
PsychInfo
Psychological & Behavioral
PUBMED
Scopus
Web of Knowledge
1058
54
7
3
368
9
1822
7
129
Duplicates removed (n= 1205)
n=3457
Screening of titles
(independently conducted by 2
reviewers)
2252 titles retrieved
Excluded (n= 1856)
Excluded by language
Lacked neurologic population
Pediatric population
Unrelated to the study topic
Did not include central nervous system
(articles could be excluded for more
than one reason)
Review of abstracts
(independently conducted by 2
reviewers)
n= 396
Review of full articles
(independently conducted by 2
reviewers)
n= 17
Number added after review of
reference lists of full text articles
n= 4
Included articles
(independently conducted by 2
reviewers)
n= 12
Excluded (n= 379)
Excluded by language
Lacked neurologic population
Pediatric population
Unrelated to study topic
Did not include central nervous system
Observational studies with no
intervention
Lack of peer review (abstract only or
poster)
(articles could be excluded for more
than one reason)
Excluded (n= 9)
Lack of neurologic population (healthy
subjects only)
Lack of mobility/gait/balance
intervention
Lack of control or comparator group
Figure 2.1 Search strategy flowchart
31
Two reviewers independently assessed the methodological quality of each included
study using the PEDro scale.33 (Table 2.1) This 10-item rating scale was developed for quality
assessment of randomized controlled trials and has demonstrated reliability.33 Trials with a
rating of at least 6/10 on the PEDro scale were rated as high quality.34
2.2.4 Data Synthesis & Analysis
The included studies were characterized by methodological heterogeneity, thus, we
refrained from statistical pooling. Each study was independently evaluated with a standardized
rating scale of clinical relevance using the 5 criteria recommended by the Cochrane Back Review
Group.35 This scale evaluates criteria relevant to physical therapy practice and is suitable for use
in the evaluation of neurologic clinical trials as well. (Table 2.1)
1. Are the patients described in detail so that you can decide whether they are comparable
to those you see in practice?
2. Are the interventions and treatment settings described well enough so that you can
provide the same for your patients?
3. Were all clinically relevant outcomes measured and reported?
4. Is the size of the effect clinically important?
5. Are the likely treatment benefits worth the potential harms?
Presently, there is no established cutoff scores for high and low quality studies with this tool;
these data were used in conjunction with the PEDro scores to evaluate the strength of the
included studies.
32
33
Trial
Schwenk et
al.21
Silsupadol et
al.22
Yang
et al.23
Yen
et al.26
Evans
et al.25
Fok
et al.26
Fok
et al.27
Canning
et al.28
Mirelman et
al.29
You
et al.30
Silsupadol et
al.31
Sveistrup et
al.32
Total
1
1
2
1
3
1
4
1
PEDro Scale Score
5
6
7
8
1
0
1
1
1
1
0
1
1
0
1
1
1
1
1
8
1
1
1
1
1
5
1
1
1
1
0
0
1
1
1
1
1
8
0
1
1
0
1
3
1
1
0
1
0
0
1
1
1
1
1
7
1
1
1
0
1
4
1
1
1
0
0
0
0
1
1
1
1
6
1
1
1
1
1
5
1
0
0
1
0
0
0
1
1
1
1
5
1
1
1
0
1
4
1
0
0
1
0
0
0
1
1
1
1
5
1
1
1
0
1
4
1
0
0
0
0
0
0
1
1
0
1
3
1
1
1
0
1
4
1
0
0
0
0
0
0
1
1
0
1
3
1
0
1
0
1
3
1
0
0
1
0
0
0
0
0
1
1
3
1
0
1
0
0
2
1
0
0
0
0
0
0
1
1
0
0
2
1
1
1
0
0
3
1
1
0
0
0
0
0
0
1
0
0
2
0
0
0
0
0
0
12
6
3
7
2
0
4
10
11
8
10
61
10
8
11
2
9
40
9
1
10
1
11
1
Total
9
1
1
Table 2.1 Trial ratings on the PEDro Methodological Quality and Clinical Relevance Scales
5
Clinical Relevance Score
2
3
4
5
Total
0
1
0
1
3
2.3 Results
2.3.1 Study Characteristics
Table 2.2 presents the characteristics of the included studies. Of the 12 included studies,
five were randomized controlled trials,21-25 five were quasi-experimental repeated measures
designs,26-29, 32 one was a prospective controlled trial30 and one was a case series.31 All studies
were conducted in an outpatient setting. With the exception of the case series,31 sample sizes
ranged from five28 to 3521 patients, with five studies enrolling more than 20 subjects.21-24,29 The
mean age of patients in the experimental group ranged from 4425 to 8831 years, and the malefemale ratio ranged from 9:125 to 1:7.22
The included studies utilized a wide range of outcome measures to assess the effect of
DT training. Table 2.2 outlines all of the outcomes used; here we describe the effect of DT
training programs on the most common outcome measures.
2.3.2 Effect of Dual-Task Intervention Programs on Gait
Gait speed for both comfortable 22,23,26-28,32 and fast28,30 walking as well as various
spatiotemporal measures of gait (e.g., stride length, stride time and cadence) were measured
using equipment including 3D motion capture and the GAITRite electronic walkway. GAITRite
measures have been validated for use in PD36 and elderly adults.37 Across studies, all groups who
received DT training improved gait speed on single-task walking. This improvement was
significant only when compared to a null control group.23,26,27 When compared to an alternative
treatment,22,30 single-task gait velocity improved across all groups.
34
Study/Year
Design
Schwenk et al.
21
(2010)
Double-Blinded
RCT
PEDro= 9
Clinical= 3
Participant Characteristics,
Inclusion Criteria
Dx: Elderly adults with
dementia
Age: (control) 82.3±7.9;
(experimental) 80.4±7.1
Gender: (control) 22F, 13M;
(experimental) 17F, 9M
35
Inclusion Criteria: MMSE
(17-26); Diagnosis of
dementia based on
international criteria; age
>65; no severe neurological,
cardiovascular, metabolic or
psychiatric disorders.
Interventions
N=35
Specific dual-task training and
additional progressive resistancebalance and functional-balance
training; performed in groups of
4-6 persons for 12 weeks
(2hr/wk)
Comparison
Interventions
N=26
2x/wk for 1 hour of
supervised motor
placebo group
training = flexibility
exercises,
calisthenics and ball
games while seated.
Outcome
Measures
Dual Task Cost (DTC)
for maximal gait speed
under complex
conditions (serial 3
subtraction)
Trail Making Test
Progressive resistance training of
functionally relevant muscle
groups for 1 hour + progressive
balance and functional training
(stepping, walking, sitting safely);
static and dynamic balance
35
Table 2.2 Characteristics of included studies (listed by PEDro score)
Continued
Table 2.2 Continued
Silsupadol et
Double-Blinded
22
al. (2009)
RCT
PEDro= 8
Clinical= 5
Dx: Older adults with
balance impairment
Age: (single-task)
74.71±7.80; (dual-task fixed)
74.38±6.16; (dual-task
variable) 76.00±4.65
All subjects received 45
minutes/session; 3x/week for 4
weeks
N=7
Dual-task training with variable
priority as in Silsupadol et al 2006
Gender: (single-task) 7F, 1M;
(dual-task fixed) 6F, 2M;
(dual-task variable) 4F,3M
36
N=8
Single-task training
as in Silsupadol et al
2006
N=8
Dual-task training
with fixed priority as
in Silsupadol et al
2006
Single & Dual-task gait
speed
Berg Balance Scale
Activities-specific
Balance Confidence
Scale
Inclusion Criteria: Older than
65yo; independent
ambulation for 10m;
MMSE≥24; no history of
orthopedic, neurologic,
cardiac, visual or auditory
conditions that could
influence balance; ≤52 on
the Berg Balance Scale;
and/or walked with a selfselected speed of ≤1.1m/s
36
Continued
Table 2.2 Continued
Yang et al.
Single-Blinded
23
(2007)
RCT
PEDro= 8
Clinical= 3
37
Dx: Chronic Stroke
Age: (control) 59.17±11.98;
(experimental) 59.46±11.83
Gender: (control) 5F,7M;
(experimental) 6F,7M.
Inclusion Criteria: single
chronic stroke resulting in
hemiparetic deficits; gait
speed between 58-80cm/s
(limited) or minimum
80cm/s (full community
ambulatory ability); not
presently receiving rehab;
able to walk 10m
independently; functional
use of involved UE; stable
medical status; capable of
understanding and following
directions
N=13
30 min ball exercise program
3x/wk for 4 weeks.
Walk while manipulating 1 of 2
balls (therapy balls size 45, 55, 85
and 95cm diameters + basketball)
N=12
Null control
Gait Speed
Cadence
Stride Time
Stride Length
Temporal Symmetry
Index
1) Walk while holding 1 or 2 balls
in both hands
2) Walk to match the rhythm of
bouncing 1 ball with 1 hand or
both hands
3) walk while holding1 ball on 1
hand and concurrently bouncing
another ball with the other hand
4) walk while kicking a basketball
(in a net)
5) walk while holding 1 ball and
concurrently kicking a basketball
(in net)
6) walk while bouncing 1 ball and
kicking a basketball
7) walk while reciprocally
bouncing 1 ball with both hands
Walking occurred forward,
backward, on a circular route,
and on an S-shaped route.
37
Continued
Table 2.2 Continued
Yen et al.
Single-Blinded
24
(2011)
RCT
PEDro= 7
Clinical= 4
38
Dx: Idiopathic PD
Age: (control) 71.6±5.8;
(conventional balance)
70.1±6.9; (virtual reality)
70.4±6.5
Gender: (control) 5F,9M;
(conventional balance) 2F,
12M; (virtual reality) 2F,
12M
Inclusion Criteria: diagnosis
of idiopathic PD (Gelb’s
criteria); MMSE>24; Hoehn
& Yahr stages II and III; no
prior balance/gait training;
no uncontrolled chronic
diseases
Training provided 30 minutes per
session, 2x/week for 6 weeks
N=14
Virtual Reality
Group:
10 minutes of stretching
exercises to warm-up followed by
20 minutes of standing on the VR
balance board and moving their
weight with an ankle strategy to
navigate a virtual environment
N=14
Null Control
Sensory Organization
Test (conditions 1
through 6) to assess
center of pressure and
center of gravity sway
Verbal reaction time
during a dual-task while
standing= each of the 6
conditions of the SOT +
auditory arithmetic
subtraction task
N= 14
Conventional Balance Group:
10 minutes of stretching
exercises to warm-up followed by
20 minutes of static stance,
dynamic weight shifting and
external perturbations.
38
Continued
Table 2.2 Continued
Evans et al.
Non-Blinded RCT
25
(2009)
PEDro= 6
Clinical= 5
39
Dx: Acquired Brain Injury
(control) 5 TBI, 4 CVA;
(experimental) 7 TBI, 2 CVA,
1 Tumor
Age: (control) 45.11±9.73;
(experimental) 44.4±8.51
Gender: (control) 1F,8M;
(experimental) 1F,9M
Inclusion Criteria: 18-65
years old; diagnosis of brain
injury or other neurologic
illness; evidence of
performance at least 1 SD
below the mean on the
Divided Attention & DualTasking test battery; selfreported difficulties with
dual-tasking in everyday life
N= 10
30 minute sessions, 1x/week for
5 weeks with a therapist and athome practice 2 practice
sessions/day, 5 days/week for 5
weeks.
Walking while:
1)Listening to instrumental music
2) Listening to vocal music
3) Listening to a recording of talkbased radio and then answering
recorded questions
4) verbal fluency task
5) answer autobiographical
questions
Walking around 2-3 obstacles
was introduced with these tasks
as they became easier.
39
N= 9
Null Control with
~5-10 minute
weekly phone calls
from a research
therapist to review
general progress,
enquire about
difficulties or
successes in dualtask situations.
Subjects also
recorded examples
of dual-task
difficulties in a diary
to simulate
awareness-raising
that training may
have produced.
Walking + clicking a
mechanical counter
(motor-motor DT)
Walking + sentence
verification task
(motor-cognitive DT)
Walking + tone
counting (motorcognitive DT)
Memory Span &
Tracking Task
Test of Everyday
Attention Telephone
Search while Counting
Dual Tasking
Questionnaire
Continued
Table 2.2 Continued
Fok et al.
26
(2010)
Test-Retest
(Repeated
Measures)
PEDro= 5
Clinical= 4
Dx: Parkinson’s disease
Age: (control) 57.7±12.3;
(experimental) 66.8±9.0
Gender: not reported
Inclusion Criteria: Diagnosis
of idiopathic PD; subjective
walking difficulties; able to
independently ambulate
12m x 25 reps; MMSE≥24.
N=6
A single training session lasting
30 minutes.
Participants were given verbal
cues to achieve longer stride
lengths during comfortable
walking and then were coached
on a walking with a subtraction
task (serial 3 subtraction)
N=6
Null Control; 30
minutes of sitting
and reading a
magazine
GAITRite electronic
walkway was used to
measure Gait Speed
and Stride Length
during single-task
walking and dual-task
walking (walking with
serial 3 subtraction)
N=6
Null Control; 30
minutes of sitting
and reading a
magazine
GAITRite electronic
walkway was used to
measure Gait Speed
and Stride Length
during single-task
walking and dual-task
walking (walking with
serial 3 subtraction)
40
Fok et al.
27
(2012)
PEDro= 5
Clinical= 4
Test-Retest
(Repeated
Measures)
Dx: Parkinson’s disease
Age: (control) 73.0±12.0;
(experimental) 66.3±11.7
Gender: not reported
Inclusion Criteria: Diagnosis
of idiopathic PD; subjective
walking difficulties; able to
independently ambulate
12m x 25 reps; MMSE≥24
N=6
A single training session lasting
30 minutes.
Participants were given verbal
cues to achieve longer stride
lengths during comfortable
walking and then were coached
on a walking with a subtraction
task (serial 3 subtraction)
40
Continued
Table 2.2 Continued
Canning et al.
Repeated
28
(2008)
Measures,
baselinePEDro= 3
controlled
Clinical= 4
41
Dx: Parkinson’s disease
Age: 61±8
Gender: 2F,3M
N=5
30 min/week, 1x/week for 3
weeks
Inclusion Criteria: Diagnosis
of idiopathic PD; Hoehn &
Yahr stages I-III; stable
response to levodopa
medications; subjective
report of gait disturbance or
UPDRS gait score <3; able to
walk independently on level
ground; MMSE≥24
30 10-meter walks at
participant’s predetermined fastas-possible pace followed by 10
10-meter walks while practicing
each of the additional tasks
(cognitive, manual, and
cognitive+ manual).
During the second and third
training sessions, two and four
extra 40-meter walks under triple
task conditions were added,
respectively, and were performed
in a narrow public hallway that
incorporated obstacles and turns.
No control group
Participant perception
of fatigue, difficulty,
anxiety and confidence
on a 10cm visual
analogue scale.
Velocity, cadence &
stride length measured
on GAITRite during
walking under the
following conditions:
Two paces:
a) comfortable
b) fast as possible
Four task conditions
a) single
b) dual-cognitive
c) dual-manual
d) triple (cognitive +
manual)
Cognitive task: auditory
color classification task
Manual task: carry a
cup filled with water
41
Continued
Table 2.2 Continued
Mirelman et al. Test-Retest
29
(2011)
(Repeated
Measures)
PEDro= 3
Clinical= 3
Dx: Parkinson’s disease
Age: (experimental)
67.1±6.5
Gender: (experimental) 6F,
14M
Inclusion Criteria: idiopathic
PD; Hoehn & Yahr Stage IIIII; walking difficulties
identified by the UPDRS
motor subscale; able to walk
independently for 5 minutes
N=20
3 sessions/week for 6 weeks
Virtual Reality program requiring
participants to process multiple
stimuli simultaneously and make
decisions about obstacle
negotiation in two planes while
continuing to walk on a treadmill.
The virtual reality imposed a
cognitive load and demanded
attention, response selection,
and processing visual stimuli.
Historical active
control group of
patients with PD
who followed a
similar program of
treadmill training,
but without virtual
reality.
Stride Length, gait
variability, dual-task
cost and obstacle
clearance on the
GAITRite and
accelerometer during
three conditions:
a) comfortable pace
b) walk with serial 3
subtraction
c) walk while
negotiating two
obstacles (box and line)
6 Minute Walk Test
42
UPDRS motor subscale
4 Square Step Test
PD quality of life
questionnaire
Trail Making Test
42
Continued
Table 2.2 Continued
You et
Prospective
30
al.(2009)
Controlled Trial
(Convenience
PEDro= 3
sample)
Clinical= 2
Dx: elderly adults with a
history of falls
Age: (control) 68.0±3.3;
(experimental) 70.6±6.8
Gender: (control) 4F,1M;
(experimental) 7F,1M
Inclusion Criteria:
>1 fall in past year;
MMSE≥24; independent
ambulation; lack of major
neurological, orthopedic or
cardiopulmonary disorders
All subjects received 18 sessions
for 30 min per session over 6
weeks
N=8
Cognitive-motor dual-task
intervention combining walking
with a memory recall task
N=5
Walking with simple
music
Velocity during fast
walking
Center of Pressure
sway in the mediallateral and anteriorposterior directions
Word recall &
arithmetic tasks
43
43
Continued
Table 2.2 Continued
Silsupadol et
Case Series with
31
al. (2006)
random
selection to
PEDro= 2
treatment
Clinical= 3
groups
Dx: elderly adults with a
history of falls or impaired
balance
Age: 88.3±5.69
Gender: 2F, 1M
Inclusion Criteria:
Able to independently
ambulate 9m; MMSE>24; no
history of neurologic,
orthopedic, cardiac, visual or
auditory conditions that
would account for
imbalance.
All subjects received 45
minutes/session; 3x/week for 4
weeks
N=1
Dual-task training with variable
priority; half of the training was
completed with a focus on
postural task performance and
half had a focus on secondary
task performance.
44
44
N=1
Single-task training
of balance exercises
N=1
Dual-task training
with fixed priority;
balance exercises
with simultaneous
auditor and visual
discrimination tasks
as well as cognitive
tasks such as
subtraction. Subject
was instructed to
maintain attention
on both postural
and secondary tasks
at all times.
Berg Balance Scale
Dynamic Gait Index
Timed Up & Go
Activities-specific
Balance Confidence
Scale
# missteps and 3D
motion analysis to
calculate body
kinematics (ML-COM
displacement) during a
4m walk under 6
different conditions:
2 single-task:
-narrow walking
-obstacle crossing
4 dual-task:
-narrow walking while
counting backward by
3’s
-obstacle crossing while
counting backward by
3’s
-narrow walking with
tone discrimination
-obstacle crossing with
tone discrimination
Continued
Table 2.2 Continued
Sveistrup et al. Test-Retest
32
(2003)
PEDro= 2
Clinical= 0
Dx: Moderate or Severe
Traumatic Brain Injury
Age: not reported
Gender: not reported
Inclusion Criteria: Diagnosis
of moderate to severe TBI
sustained at least 6 months
earlier; no current
participation in acute
inpatient rehabilitation; able
to stand independently for
two minutes
45
1 hour sessions, 3x/week for 6
weeks
N=5
Conventional Exercise Group:
Balance exercises focusing on
stepping, picking up objects,
single and double limb stance,
moving within the base of
support, walking, sit-to-stand,
reaching, hopping, jumping, and
jogging.
N=5
Null Control
Measures of quiet
stance and gait speed
Activities Balance
Confidence Scale
Community Balance &
Mobility Scale
N=4
Virtual Reality Exercise Group:
requires participants to reach,
move within base of support,
step, sit-to-stand, hop, jump and
jog
Cerebrovascular Accident or Stroke (CVA); Diagnosis (Dx); Dual Task Cost (DTC); Mediolateral Center of Mass (ML-COM); Mini-Mental State Examination
(MMSE); Parkinson’s disease (PD); Randomized Controlled Trial (RCT); Traumatic Brain Injury (TBI); United Parkinson’s Disease Rating Scale (UPDRS).
45
2.3.3 Effect of Dual-Task Intervention Programs on Balance
Balance was assessed22,31 using the Berg Balance Scale, a 14 item scale where
participants are rated 0-4 on static and dynamic balance tasks. A higher score indicates better
balance; the Berg Balance Scale is reliable and valid in MS,38 and cut off scores have been
established for stroke39 and PD populations.40 While a case series31 showed large changes in
Berg scores after variable priority DT training, the larger RCT showed no significant differences
between treatment and control groups.22 Variable training was characterized by instructions to
prioritize either the motor task or the cognitive task rather than focusing on only one task.
Larger improvements were perhaps seen in the case series because the individual receiving
variable priority DT training was far more impaired than the other subjects and had room for a
larger change with training.
Other studies24,30,32 measured balance with postural sway in quiet stance. The sensory
organization test measures static balance and postural sway under 6 conditions: stance on each
firm and uneven ground with eyes open, closed and with occluded vision. The sensory
organization test has been used in stroke41 and validated in elderly adults.42 You et al. found no
significant differences between treatment and control groups on anterior-posterior and mediallateral postural sway in quiet stance in older adults with a history of falls.30 Values for quiet
stance were not reported by Sviestrup et al.32 and queries to the author received no response.
Yen et al.24 demonstrated a significant difference in two items on the sensory organization test
(conditions 5 and 6 with occluded vision or eyes closed on an uneven surface) between the
treatment groups and the control group, but no differences between the two treatment groups
(DT virtual reality vs. conventional balance training) in individuals with PD.
46
2.3.4 Effect of Dual-Task Intervention Programs on Cognition
Several studies21,25,29,30 measured the effect of DT training on domains of cognition
including memory, processing speed and attention. There was no cognitive task common to all
of the studies, so comparisons in this outcome could not be assessed. While both Evans et al.25
and Schwenk et al.21 found no significant differences in cognition following training, You et al.30
and Mirelman et al.29 reported significant improvements in memory performance and
processing speed. This perhaps speaks to the specificity of task training; both Evans and You
measured memory domains but the training in You’s study included similar memory dual-tasks
as part of the training intervention, while the Evans protocol did not. Baseline level of cognitive
dysfunction may also play a role in the benefits gained through training, although this was not
explored by any of the studies.
2.3.5 Effect of Dual-Task Intervention Programs on Ability to Dual-Task
By far, the most commonly assessed outcome measure across studies was ability to DT
during gait.21-23,25-29 While the majority of studies assessed motor-cognitive tasks (e.g., walking
while performing arithmetic), several also assessed motor-motor tasks (e.g., walking while
carrying a glass).23,25,28 A common measure of DT ability during gait is the calculation of the
dual-task cost (DTC), which allows one to determine the specific contribution of the secondary
task on the primary task, in this case, gait. The formula used to calculate DTC is as follows:
Dual Task Cost (DTC) = [(dual task – single task) / single task] x 100
Following DT training, Schwenk et al.21 found reduced DTC for both gait speed and stride
length in the intervention group when compared to the control group, where DTC was
unchanged. This held true for DT walking when the secondary task was addition or subtraction.
Similarly, significant improvements in DT gait speed and stride length were found following
47
training in PD,26-29 chronic stroke,23 and elderly adults with a history of falls22 when compared to
controls. The type of DT training did not seem to effect the improvement; groups trained with
variable task priority and fixed task priority both improved DT walking,22 as did those who
trained with motor-motor tasks23 rather than motor-cognitive tasks.21 Greater improvements in
gait speed and stride length were seen after training when compared to both null control
groups25-27 and single-task control groups.21,22,29 Training effects were even seen after short term
training programs, such as a single 30 minute session of DT training26,27 or three 30-minute
sessions of DT training,28 and these significant increases in stride length and gait velocity were
maintained at a delayed retention.
The majority of included studies did not specifically prioritize either the motor or
cognitive task during training. However, in studies that did specify allocation of attention, the
training effect on DT performance was only maintained at a 12-week follow-up in the
participants who received DT training with variable priority (i.e. prioritize first one task and then
the other) instructions.22 This has been previously noted in cognitive-cognitive DT training
programs, where individuals trained with variable-priority instructions learned tasks faster and
performed better than those who received fixed-priority instructions (i.e. equal attention to
both tasks).43 Fok et al.26 examined DT walking when subjects were trained to prioritize gait by
“focus[ing]100% on big steps.” Following training, individuals significantly increased gait
velocity and gait speed over baseline without an associated decline in secondary task accuracy.
A meta-analysis was not undertaken in this review based on heterogeneity within and
among the identified studies. The trials included a wide range of sample sizes, variable
alternative treatments, participants of widely different disability levels, and DT training
protocols with methodological heterogeneity and vastly variable treatment duration (ranging
48
from 30 minutes to 24 hours). Given this variability, it was not possible to conclude that the
trials were sufficiently homogenous to conduct a meaningful meta-analysis.44 The mean
differences, treatment effect sizes and associated 95% confidence intervals for the individual
trials are presented by outcome assessment and comparative treatment in Figure 2.2. Effect
sizes greater than zero indicate that the outcome favors the DT training group, while effect sizes
less than zero indicate that the outcome favors the comparison group. Confidence intervals that
include zero should be interpreted with caution. Unfortunately, only studies providing sufficient
data could be included in this figure, thus we were not able to include Yen et al.24
2.3.6 Adverse Effects
There were no adverse effects of training reported in any of the included studies.
2.4 Discussion
2.4.1 General Findings
While several case studies have examined DT training in individuals with neurological
deficits, there is a paucity of randomized controlled trials in this area. To date, DT training has
been formally investigated in elderly adults with dementia,21 older adults with balance
impairment,22,30,31 chronic stroke,23 PD24,26-29 and brain injury.25,32 Based on PEDro scores, there
are 7 studies with weak evidence, mostly due to lack of blinding, limited sample size and poor
reproducibility. However, if the main conclusions are drawn from the studies scoring a 6 or
higher on the PEDro scale, DT training in individuals with neurologic disorders is associated with
improvements in balance24 and ability to DT.21-25
49
2.4.2 Limitations
Comparison of results across studies was limited by the wide range of neurologic
populations and deficits as well as the methods of assessing DT. Many of the studies included
null control groups, which does not allow for comparison of DT training to other currently
utilized training paradigms. Additionally, the majority of studies used the same task for both
training and outcome assessment. Despite the established link between dual-tasking and
increased gait variability, none of the studies explored gait variability in their outcome
assessments. However, in studies receiving the highest marks for methodological quality,
individuals that trained with dual-tasks showed greater improvements in spatiotemporal
measures of DT walking than those trained with single-tasks.21,22
50
Comparison
Treatment
Balance
22
Training
Outcome Measure
Balance (Berg)
FP
VP
Mean
Difference
1.96
-0.29
Gait (ST velocity)
FP
VP
FP
VP
0.13
0.06
0.15
0.14
0.69 (-0.32 to 1.70)
0.34 (-0.68 to 1.37)
1.03 (-0.01 to 2.08)
0.18 (-1.04 to 1.20)
-0.05
0.13
-0.06 (-1.17 to 1.06)
0.08 (-1.04 to 1.20)
16.31
1.11 (-0.09 to 2.30)
Gait (ST fast velocity)
21
Gait (DTC velocity)
21
Gait (DTC stride length)
-0.07
19.01
14.2
-0.30 (-1.42 to 0.83)
1.14 (0.59 to 1.68)
1.07 (0.53 to 1.62)
Cognition (Memory Span &
25
Tracking Task)
-4.75
-0.59 (-1.51 to 0.33)
42.57
18.15
13.55
6.11
0.15
0.32
0.34
0.09
0.14
0.16
1.94 (0.99 to 2.89)
0.82 (0 to 1.63)
0.78 (-0.04 to 1.59)
1.52 (0.50 to 2.54)
0.56 (-0.60 to 1.71)
1.11 (-0.10 to 2.33)
0.35 (-0.79 to 1.49)
0.66 (-0.50 to 1.82)
0.55 (-0.60 to 1.70)
0.90 (-0.29 to 2.08)
Gait (DT velocity)
General
Exercise
30
Balance (MLCOP)
30
Balance (APCOP)
Cognition(Memory Recall %)
21
Null
Control
23
Gait (ST velocity)
23
Gait (DT velocity)
23
Gait (DT stride length)
25
Gait (DT gait speed)
26
Gait (ST velocity)
26
Gait (DT velocity)
26
Gait (DT stride length)
27
Gait (ST velocity)
27
Gait (DT velocity)
27
Gait (DT stride length)
30
Effect Size (SMD)
and 95% CI
0.46 (-0.53 to 1.45)
-0.1 (-1.1 to 0.9)
Forest Plots
Figure 2.2 Treatment effects for dual-task (DT) training versus comparison treatments for all
trials with full data sets. Treatment effects favoring DT training assigned positive Hedges
standardized mean difference (SMD) values.
Anterior-posterior center of pressure (APCOP); dual-task cost (DTC); fixed priority (FP); mediallateral center of pressure (MLCOP); single-task (ST); variable priority (VP).
51
2.4.3 Implications for Practice
Descriptive studies have shown that individuals with neurologic conditions often adopt
a posture-second strategy, where the cognitive task is given higher priority. This may result in
loss of balance, declines in walking, and an increased fall risk.10 In this systematic review, we
suggest that a DT training program may be utilized in individuals with neurologic disorders to
improve not only gait mechanics, but also prioritization during multi-tasking. Practitioners
should consider evaluating patients with neurological disorders for dual-tasking deficits.
2.4.4 Implications for Future Research
While DT training in neurological populations may yield improvements in both balance
and gait speed, it has been largely unsuccessful at generalizing to novel dual-task situations.
However, a recent pilot study,45 not included in this review due to lack of a control group or
extended baseline, suggests that carryover is possible if functional tasks are the focus of
training. Future research in the area of DT training should assess the ability of the training
paradigm to translate to novel DT situations, particularly during functional tasks encountered by
subjects on a daily basis.
2.5 Conclusion
DT training appears to be both feasible and effective for improving spatiotemporal
measures of DT gait in individuals with neurologic disorders. More research is needed to define
the specific DT interventions that are most effective and to better assess whether some
interventions are more appropriate than others for a specific diagnostic group. This is the first
review to examine the motor outcomes of DT training across neurologic disorders. Improvement
52
of DT ability in individuals with neurologic disorders may improve independence and decrease
fall risk.
53
2.6 References
1. 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(5):647–655.
2. Mateer CA, Kerns KA, Eso KL. Management of attention and memory disorders following
traumatic brain injury. J Learn Disabil. 1996; 29: 618-632.
3. McDowd JM. An overview of attention: behavior and brain. J Neurol Phys Ther. 2007; 31:
98-103.
4. Mathias JL, Wheaton P. Changes in attention and information-processing speed
following severe traumatic brain injury: a meta-analytic review. Neuropsychology. 2007;
21(2):212-223.
5. Azouvi P, Couillet J, Leclercq M, et al. Divided attention and mental effort after severe
traumatic brain injury. Neuropsychologia. 2004; 42:1260–1268.
6. McCulloch K. Attention and dual-task conditions: physical therapy implications for
individuals with acquired brain injury: J Neurol Phys Ther. 2007; 31: 104-118.
7. Hamilton F, Rochester L, Paul L, Rafferty D, O’Leary CP, Evans JJ. Walking and talking: an
investigation of cognitive-motor dual tasking in multiple sclerosis. Mult Scler. 2009;
15(10): 1215-1227.
8. 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):1–14.
9. Yogev G, Giladi N, Peretz, Springer S, Simon ES, Hausdorff JM. Dual-tasking, gait
rhythmicity, and Parkinson’s disease: which aspects of gait are attention demanding. Eur
J Neurosci.2005; 22: 1248-1256.
10. Yogev-Seligmann G, Hausdorff JM, Giladi N. The role of executive function and attention
in gait. Mov Disord. 2008; 23(3): 329-342.
11. Springer S, Giladi N, Peretz C, Yogev G, Simon ES, Hausdorff JM. Dual-tasking effects on
gait variability: the role of aging, falls and executive function. Mov Disord. 2006; 21(7):
950-957.
12. Cocchini C, Della Sala S, Logie RH, Pagani R, Sacco L, Spinnler H. Dual task effects of
walking when talking in Alzheimer’s disease. Revue Neurologique (Paris). 2004; 160(1):
74-80.
54
13. Negahban H, Mofateh R, Arastoo AA, Mazaheri M, Yazdi MJS, Salavati M, Majdinasab N.
The effects of cognitive loading on balance control in patients with multiple sclerosis.
Gait Posture. 2011; 34(4): 479-484.
14. Porosinska A, Pierzchala K, Mentel M, Karpe J. Evaluation of postural balance control in
patients with multiple sclerosis—effect of different sensory conditions and arithmetic
task execution. A pilot study. Neurologia i Neurochirurgia Polska. 2010; 44:35–42.
15. Cameron MH, Lord S. Postural control in multiple sclerosis: implications for fall
prevention. Current Neurology and Neuroscience Reports. 2010; 10: 407-412.
16. Rábago CA and Wilken JM. Application of a mild traumatic brain injury rehabilitation
program in a virtual reality environment: a case study. J Neurol Phys Ther. 2011; 35: 185193.
17. Pellecchia GL. Dual-task training reduces impact of cognitive task on postural sway. J
Mot Behav. 2005; 37(3): 239-246.
18. Parker TM, Osternig LR, Lee HJ et al. The effect of divided attention on gait stability
following concussion. Clin Biomech. 2005; 20: 389-395.
19. Couillet J, Soury S, Lebornec G, et al. Rehabilitation of divided attention after severe
traumatic brain injury: a randomised trial. Neuropsychol Rehabil.2010; 20(3): 321-339.
20. Vallat-Azouvi C, Pradat-Diehl P, Azouvi P. Rehabilitation of the central executive of
working memory after severe traumatic brain injury: two single-case studies. Brain Inj.
2009; 23(6): 585-594.
21. Schwenk M, Zieschang T, Oster P, Hauer K. Dual-task performances can be improved in
patients with dementia: a randomized controlled trial. Neurology. 2010; 74: 1961-1968.
22. Silsupadol P, Shumway-Cook A, Lugade V, vonDonkelaar P, Chou LS, Mayr U, Woollacott
MH. 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:
381- 387.
23. Yang YR, Wang RY, Chen YC, Kao MJ. Dual-task exercise improves walking ability in
chronic stroke: a randomized controlled trial. Arch Phys Med Rehabil. 2007; 88: 12361240.
24. Yen CY, Lin KH, Hu MH, Wu RM, Lu TW, Lin CH. Effects of virtual reality-augmented
balance training on sensory organization and attentional demand for postural control in
people with Parkinson disease: a randomized controlled trial. Phys Ther. 2011; 91: 862874.
55
25. Evans JJ, Greenfield E, Wilson BA, Bateman A. Walking and talking therapy: improving
cognitive-motor dual-tasking in neurological illness. J Int Neuropsych Soc. 2009; 15: 112120.
26. Fok P, Farrell M, McMeeken J. Prioritizing gait in dual-task conditions in people with
Parkinson’s. Hum Movement Sci. 2010; 29: 831-842.
27. Fok P, Farrell M, McMeeken J. The effect of dividing attention between walking and
auxiliary tasks in people with Parkinson’s disease. Hum Movement Sci. 2012; 31: 236246.
28. Canning CG, Ada L, Woodhouse E. Multiple-task walking training in people with mild to
moderate Parkinson’s disease: a pilot study. Clin Rehabil. 2008; 22: 226-233.
29. Mirelman A, Maidan I, Herman T, Deutsch JE, Giladi N, Hausdorff JM. Virtual reality for
gait training: can it induce motor learning to enhance complex walking and reduce fall
risk in patients with Parkinson’s disease? J Gerentol A Biol Sci Med Sci. 2011; 66A(2):
234-240.
30. You JH, Shetty A, Jones T, Shields K, Belay Y, Brown D. Effects of dual-task cognitive-gait
intervention on memory and gait dynamics in older adults with a history of falls: a
preliminary investigation. Neurorehabilitation. 2009; 24: 193-198.
31. Silsupadol P, Siu KC, Shumway-Cook A, Woollacott M. Training of balance under singleand dual-task conditions in older adults with balance impairments. Phys Ther. 2006; 86:
269-281.
32. Sveistrup H, McComas J, Thornton M, Marshall S, Finestone H, McCormick A, Babulic K,
Mayhew A. Experimental studies of virtual reality-delivered compared to conventional
exercise programs for rehabilitation. Cyberpsychol Behav. 2003; 6(3): 245-249.
33. Mahar CG, Sherrington C, Herbert RD, et al. Reliability of the PEDro scale for rating
quality of randomized controlled trials. Phys Ther. 2003; 83:713-721.
34. PEDro statisitics. PEDro Physiotherapy Evidence Database. 2012. Available at:
http://www.pedro.org.au/english/downloads/pedro-statistics/. Accessed October 15,
2012.
35. van Tulder M, Furlan A, Bombardier C, et al. Updated method guidelines for systematic
reviews in the Cochrane collaboration back review group. Spine (Phila Pa 1976). 2003;
28: 1290-1299.
36. Nelson AJ, Zwick D, Brody S, Doran C, Pulver L, Rooz G, Sadownick M, Nelson R,
Rothman J. The validity of the GaitRite and the functional ambulation performance
scoring system in the analysis of Parkinson gait. Neurorehabilitation. 2002;17:255–262.
56
37. Bilney B, Morris M, Webster K. Concurrent related validity of the GAITRite walkway
system for quantification of the spatial and temporal parameters of gait. Gait Posture.
2003; 17:68–74.
38. Cattaneo D, Regola A, Meotti M. Validity of six balance disorders scales in persons with
multiple sclerosis. Disabil Rehabil. 2006; 28(12): 789-795.
39. Stevenson TJ. Detecting change in patients with stroke using the Berg Balance Scale.
Aust J Physiother. 2001; 47; 29-38.
40. Steffen T, Seney M. Test-retest reliability and minimal detectable change on balance and
ambulation tests, the 36-item short-form health survey, and the United Parkinson
Disease Rating Scale in people with Parkinsonism. PhysTher. 2008; 88(6): 733-746.
41. Marigold DS, Eng JJ, Tokuno CD, Donnelly CA. Contribution of muscle strength and
integration of afferent input to postural stability in persons with stroke. Neurorehabil
Neural Repair. 2004; 18(4): 222-229.
42. Ford-Smith CD, Wyman JF, Elswick RK Jr, Fernandez T, Newton RA. Test-retest reliability
of the sensory organization test in noninstitutionalized older adults. Arch Phys Med
Rehabil. 1995; 76(1): 77-81.
43. Kramer AF, Larish JF, Strayer DL. Training for attentional control in dual task settings: a
comparison of young and old adults. J Exp Psychol- Appl. 1995;1:50-76.
44. Herbert RD, Bø K. Analysis of quality interventions in systematic reviews. Brit Med J.
2005; 331:507-509.
45. Yogev-Seligmann G, Giladi N, Brozgol M, Hausdorff JM. A training program to improve
gait while dual tasking in patients with Parkinson’s disease: a pilot study. Arch Phys Med
Rehabil.2012; 93: 176-181.
57
Chapter 3: Backward Walking Measures Are Sensitive to Age-Related Changes in Mobility and
Balance
3.1 Introduction
Many falls in elderly occur from backward perturbations1 or during transitional
movements that require a person to move backward, such as turning and stepping backwards to
sit in a chair.1,2 These falls may be related to deficits in stepping responses to backward
perturbations and difficulties with backward walking (BW). In response to unpredictable
backward perturbations, elderly individuals were twice as likely to take compensatory steps to
maintain balance compared to young adults.3,4 Laufer5,6 found that healthy elderly had
significantly greater reductions in gait speed and stride length during BW than young adults.
Increased gait variability was found to correlate with increased fall risk in multiple populations.7
Hackney et al.8 demonstrated that gait variability was increased in backward compared to
forward walking (FW) in healthy elderly and those with Parkinson’s disease.8 Several studies
have demonstrated that practicing multi-directional stepping either as an exercise program9,10
or in the form of tango dancing11 improves mobility. Taken together these findings suggest that
the inability to take effective backward steps may predispose the elderly to declines in
functional ambulation and to increased risk of falls.
58
Since the ability to walk backward is a crucial element of mobility function and deficits
might be related to a greater risk for backward falls, assessment of BW may be an important
clinical tool. Few studies have compared spatiotemporal gait measures in FW)versus BW and
only three included elderly participants.5,6,8 No studies to date have examined the
characteristics of BW in middle-aged adults or elderly with declines in functional mobility; thus,
it is not known whether the decline in BW is a slow, progressive change related to aging or a
more abrupt decline related to cumulative changes in the neuromuscular system.
The purpose of this study was to compare spatiotemporal measures and their
coefficients of variation (CVs) of BW and FW 1) in young (18-34 y.o.), middle-aged (35-64 y.o.)
and elderly (65 y.o. and older) and 2) in elderly fallers and non-fallers; and 3) to compare the
strength of the relationship between age and BW and FW measures to determine the utility of
BW performance as a clinical measure of safety and functional mobility. We hypothesized that
BW spatiotemporal measures would be equivalent between young and middle-aged adults and
would be significantly more impaired in elderly fallers than non-fallers. We also hypothesized
that age would be more closely related to BW spatiotemporal measures than to FW measures.
Assessment of BW measures may assist healthcare professionals in decision making for fall
prevention interventions, assistive device prescription, and assessment of intervention efficacy
in the elderly.
3.2 Methods
This cross-sectional study was approved by the Institutional Review Board. All
participants were consented prior to participation.
59
A convenience sample of 130 adults, including 37 young, 31 middle-aged and 62 elderly
participated in the study. Young and middle-aged participants were recruited from among
students and faculty. Elderly participants were recruited at three facilities with independent and
assisted-living units. Assisted living residents made up 25% of the elderly sample and 50% of the
elderly used an assistive device for walking outside their home. Twice as many elderly were
recruited to allow further analysis of differences between those with and without a history of
reported falls. All participants were able to ambulate >10 feet without an assistive device and/or
physical assistance and demonstrated understanding of the purpose of the study. Individuals
who were pregnant or had orthopedic or neurologic conditions that altered their walking were
excluded.
Spatiotemporal gait measures were collected using the GAITRite System (V3.9, MAP/CIR
Inc.). GAITRite measures are valid and reliable in the elderly.12 Demographic data including preexisting diagnoses, weekly types and amounts of exercise, self-reported fall history (number of
falls the past 6 months),13 and assistive device use were collected. Elderly participants were also
tested on the Tinetti Mobility Test (TMT) to better describe their functional mobility status.14
Participants were asked to walk at a comfortable pace across the GAITRite walkway for 3 trials
each of FW and BW. Participants were instructed to walk 2 meters before and after the walkway
to allow for acceleration and deceleration. Participants completed all trials without an assistive
device, wearing a gait belt and were guarded.
3.2.1 Data Analysis
The dependent variables considered in this study included average GAITRite results from
trials of each direction, coefficient of variation (CV) values were calculated to assess the
variability of gait measures in BW and FW. Data for each of the gait measures and CVs were
60
analyzed using one-way repeated-measures ANOVA with Tukey’s post-hoc tests to detect
differences between BW and FW and between age groups. The p-value was set at 0.05 to
control for type I error rate. Participant age served as the grouping variable in the ANOVA, while
BW and FW were the within subject or repeated variable. The group (age) effect, direction
(BW/FW) effect and the interaction of the two effects were analyzed. The interaction of age
group and direction was tested for all gait variables and their associated CVs to determine if the
various age groups performed significantly different in BW and FW.
Correlational analysis was performed to examine the relationship between age and
performance in forward and backward gait. Differences in spatiotemporal measures of FW and
BW between elderly fallers and elderly non-fallers were determined using independent t-tests.
This final analysis included only the elderly age group. Statistical analyses were performed using
SPSS version 19.0.
3.3 Results
Young adult [mean age = 24.1 ± 2.5 SD (range 21–31), 10 males], middle-aged [mean age
= 47.3± 7.9 (range 35- 61), 4 males] and elderly -[mean age = 85.3 ±6.7(range 66-98), 12 males]
individuals participated in the study. Two elderly did not meet inclusion criteria due to health
issues and were unable to participate. Ninety percent of elderly participants reported exercising
(most commonly walking) at least once a week (mean = 4x/week). Tinetti Mobility Test scores
for elderly fallers [n = 12; mean age = 86.3±4.7; mean TMT scores = 21.3 ± 5.8 (range 8-28)] were
significantly lower (p<.05, 2 tailed Mann Whitney U test) than non-fallers [n = 50; mean age =
85.4± 7.1; mean TMT = 25 ± 2.8 (range 17-28)] indicating that fallers on average had greater
mobility impairments than non-fallers. In the repeated measures ANOVA, the interaction
61
between the participant’s age group and walking direction was significant for the average and
CV of gait measures across trials. The F statistics and p-values for the interaction terms are in
Table 3.1.
3.3.1 Gait Across Age Groups
Gait parameters varied with age and differed from BW to FW (Table 3.2). Velocity was
similar between young and middle-aged in both BW (1.13 m/s ± .2; 1.03 m/s ± .2) and FW (1.49
m/s ± .2; 1.48 m/s ± .2), but was significantly (p<.001) lower in elderly individuals compared to
the other groups. Declines in velocity were greater in BW (50% young versus elderly, 54%
middle-aged versus elderly) than FW (34% young versus elderly, 30% middle-aged versus
elderly). Elderly participants also had significantly (p<.001) shorter stride length, increased
double support and stance percent, and decreased swing percent than the young and middleaged individuals in both BW and FW. Base of support (BOS) was significantly wider (p<.001) for
elderly compared to the young group (p<.001) in BW and for elderly compared to the middleaged group in FW. In BW the young also had a significantly (p<.03) narrower BOS than the
middle-aged group. All groups walked significantly slower and with a significantly shorter stride
length, lesser swing percent, greater stance and double support percent, and wider BOS during
BW than FW (p<.001).
62
Gait Measures
Young
Mean
Difference
(SD)
-.387 (.13)
Middle
Aged Mean
Difference
(SD)
-.458 (.15)
Elderly
Mean
Difference (SD)
-.490* (.17)
Test Statistic for
ANOVA
Interaction Term
(F statistic)
F(2,127)=5.00
P-value for
ANOVA
Interaction
Term
0.008
Stride Length (cm)
31.43 (8.64)
37.94 (11.19)
46.49* (14.03)
F(2,127) = 7.77
<0.001
Base of Support
(cm)
Swing Percent
-4.53 (3.77)
-8.08 (3.67)
-6.81# (4.25)
F(2,127)=15.83
<0.001
1.13 (.57)
1.37 (2.08)
F(2,127)=7.20
0.001
Stance Percent
-6.60 (4.95)
-1.34 (2.04)
2.98# (3.01)
-3.09# (2.80)
F(2,127)=6.57
0.002
Double Support
Percent
CV Step Time
-2.62 (2.35)
-2.27 (3.27)
-6.44** (5.04)
F(2,127)=7.17
0.001
-1.22 (1.37)
-1.65 (2.80)
F(2,125)=6.67
0.002
CV Step Length
-3.77 (2.68)
-5.27 (3.86)
-5.20# (9.20)
-12.75** (11.41)
F(2,125)=24.38
<0.001
CV Swing Time
-2.66 (1.82)
-4.39 (2.80)
-8.93** (7.13)
F(2,125)=12.50
<0.001
CV Double Support
-2.21 (1.69)
-2.76 (2.57)
-6.95* (7.94)
F(2,125)=4.84
0.009
Velocity (m/s)
Table 3.1 Interactions by age and direction of gait
Comparison of forward to backward walking interactions demonstrating that across most
variables backward walking variables changed more dramatically and these changes were
typically significant between the elderly and the young and middle aged. Elderly significantly
different than: * young, p < .001; ** young and middle aged, p < .001; # young, p < .02;  middle
aged, p < .02 and  middle aged significantly different than young, p < .001. CV, coefficient of
variation.
63
There were significant interaction effects among age and walking conditions indicating
that as participants went from walking forward to backward, the changes in gait measures were
significantly (p=.03 to p<.001) greater in the elderly compared to young and middle-aged, with
the exception that differences were greater between only the young and elderly for velocity and
stance percent (Table 3.1). In addition changes in BOS going from FW to BW were significantly
larger in both middle aged and elderly compared to young.
3.3.2 Variability of Gait
The variability of gait was significantly greater for the elderly compared to young and
middle-aged groups in both BW and FW (Table 3.2). There were no significant differences in CV
values between young and middle-aged individuals. Increases in CVs across age were greater
than changes in gait measures, ranging from an 83% increase in double support time CV
between middle-aged and elderly to a 134% increase in stride length CV between young and
elderly in BW. All groups had significantly greater CVs during BW than FW (p<.001). There were
significant (p<.001) interaction effects among age and walking conditions (FW versus BW)
across all CVs indicating that as participants went from FW to BW the changes in CVs were
significantly greater in elderly compared to young and middle-aged individuals (Table 3.1).
3.3.3 Relationship of Gait Measures to Age
Correlational analysis revealed that age was more closely associated with BW than FW
measures. Stronger correlations were observed between age and BW than FW in four of the
five gait measures. This same pattern was not observed in the CV variables (Table 3.3).
64
Forward
Young
Velocity
(m/s)
Stride Length
(cm)
Base of Support
(cm)
Double Support
Percent
Swing Percent
65
Stance Percent
Double Support
Percent
CV of step time
(%)
CV of stride
length
(%)
CV of swing
time
(%)
CV of double
support time
(%)
Backward
Middle Aged
a
1.48 ± .22
(1.1 - 1.9)
a
141.1 ± 16.4
(113.2 - 191.8)
a
8.5 ± 2.9
(3.3 - 14.1)
a
28.2 ± 3.0
(18.7 - 32.5)
a
36.1 ± 1.9
(31.0 - 39.4)
a
63.9 ± 1.9
(60.6 - 69)
a
28.2 ± 3.0
(18.7 - 32.5)
a
4.1 ± 1.4
(1.5 – 7.8)
a
2.5 ± 1.1
(.6 – 5.3)
1.49 ± .18
(1.08 - 1.86)
a
144.9 ± 12
(115.1 - 166.3)
a
9.9 ± 2.0
(6.45 - 13.7)
a
28.2 ± 4.0
(21.5 - 37.5)
36.7 ± 1.5
(34.7 - 41.7)
a
63.3 ± 1.5
(58.2 - 65.3)
28.2 ± 4.0
(21.5 - 37.5)
a
3.1 ± 1.2
(1.3 – 7.6)
a
2.1 ± .8
(.7 – 4.5)
a
Elderly
Young
Middle Aged
Elderly
a
1.13 ±. 2
(.67 - 1.61)
112.9 ±15.4
(69.2 - 139)
14.5 ± 4.0
(2.1 - 21.8)
29.4 ± 2.7
(28.6 - 30.3)
1.03 ±. 2
(.55 - 1.53)
103.1 ± 17.4
(67.2 - 151.5)
16.6 ± 3.7*
(6.8 - 25.3)
30.5 ± 5.0
(28.7 - 32.3)
.52 ±. 2 *
(.2 - 1.05)
52.8 ± 17.9 *
(21 - 106)
17.2 ± 4.3 *
(7 - 25.9)
40.2 ± 7.6 *
(38.3 - 42.2)
a
35.4 ± 1.7
(32.3 - 39.2)
64 ± 4.9
(36.5 - 67.7)
34.7 ± 2.5
(27.8 - 39.2)
65.2 ± 2.5
(60.8 - 72.2)
30.2 ± 3.3 *
(21.3 - 37.6)
69.8 ± 3.4 * (62.4
- 78.7)
a
29.4 ± 2.7
(28.6 - 30.3)
4.3 ± 1.1
(2.2 – 7)
4.8 ± 1.9
(1.9 – 10.3)
30.5 ± 5.0
(28.7 - 32.3)
6.9 ± 2.6
(1.9 – 11.5)
6.9 ± 2.6
(1.9 – 11.5)
40.2 ± 7.6 *
(38.3 - 42.2)
11.8 ± 9.6 *
(3.8 – 64.5)
16.1 ± 6.7 *
(5.1 – 34.8)
a
5.1 ± 1.7
(2.2 – 9.1)
6.4 ± 2.6
(3.6 – 13.9)
15.0 ± 8.5 *
(3.8 – 42.5)
a
7.4 ± 2.5
(1.2 – 13.6)
9.0 ± 3.4
(5.6 – 22.6)
16.5 ± 12.8 *
(6.4 – 80.9)
1.07 ± .31 *
(.43 - 2.01)
a
103.6 ± 23.3 *
(44.6 - 151.6)
a
10.5 ± 4.1  (1.37
- 20.4)
a
33.2 ± 6.5 *
(22.6 - 54.3)
a
33.5 ± 3.4 *
(21.5 - 38.2)
a
66.5 ± 3.4 *
(61.8 - 78.5)
a
33.2 ± 6.5 *
(22.6 - 54.3)
a
6.4 ± 3.1 *
(2.0 – 19.3)
a
13.7 ± 7.7 *
(3.1 – 36.2)
a
3.7 ± 1.6
(1.9 – 8.2)
a
14.8 ± 9 *
(3.1 – 54.3)
a
6.7 ± 2.6
(2.5 – 13.2)
a
9.8 ± 3.4 *
(4.1 – 18.3)
2.9 ± 1.0
(1.1 – 5.8)
5.8 ± 1.7
(3.2 – 8.2)
Table 3.2 Spatiotemporal measures and CVs of forward and backward walking
All values are means ± SD (range); *significantly different than young at P < .001; significantly different than middle-aged at P < .001; a
Significant difference between forward walking and backward walking within group; CV, coefficient of variation.
65
Pearson
Correlation Coefficients
Forward
Backward
Velocity (m/s)
0.72*
0.79*
Stride Length (cm)
0.78*
0.84*
Base of Support (cm)
0.10
0.22*
Swing Percent
0.53*
0.53*
Double Support Percent
0.55*
0.59*
CV of step time (%)
0.55*
0.44*
CV of stride length (%)
0.62*
0.73*
CV of swing time (%)
0.66*
0.60*
CV of double support time (%)
0.57*
0.41*
Table 3.3 Relationship of gait and CV measures to age by walking direction:
Pearson correlation coefficients
* significant at p < .001; coefficient of variation (CV)
66
3.3.4 Comparison Elderly Fallers and Non-fallers
Spatiotemporal measures for fallers and non-fallers were compared only for the elderly
age group. Statistical differences were found in 5 out of 6 BW spatiotemporal measures, while
only 2 out of 6 FW measures were significantly different (Table 3.4). Elderly fallers walked with a
slower velocity and took significantly shorter strides than non-fallers in BW and FW. Fallers
spent a larger percentage of the gait cycle in double support, single limb stance and had a wider
BOS than non-fallers in BW (p<.05) but not in FW. In BW step time variability was higher in
fallers than non-fallers. All other CVs were statistically equivalent in BW and FW.
Gait Parameter
Forward Mean
(SD)
Backward Mean (SD)
Velocity (m/s)
Faller
Non - Faller
.89* (.2)
1.0 (.2)
.39* (.12)
.56 (.20)
Stride Length (cm)
Faller
Non - Faller
84.5* (15.9)
101.6 (18.7)
37.1* (12.7)
56.5 (17)
Base of Support (cm)
Faller
Non - Faller
12.3 (2.8)
10.2 (4.5)
19.8* (3.8)
16.6 (4.2)
Swing Percent
Faller
Non - Faller
31.5 (3.6)
33.4 (3.2)
28.7 (2.1)
30.6 (3.4)
Stance Percent
Faller
Non - Faller
68.4 (3.7)
66.6 (3.2)
71.8* (2.4)
69.4 (3.4)
Double Support Percent
Faller
Non - Faller
37.1 (6.3)
33.4 (6.0)
44.7* (44.7)
39.2 (39.2)
CV Step Time (%)
Faller
7.3 (4.7)
12.8* (10.6)
Non - Faller
6.4 (2.5)
9.0 (2.5)
Table 3.4 Comparison of gait and mobility measures between elderly fallers and non-fallers
* - Significant p < .05; Fallers n = 12; non-fallers n = 50;
2 tailed, equal variances assumed; df = 60.
67
An interesting and unexpected finding was that all fallers had BW velocities of less than
0.6 m/s (Figure 3.1). Calculation of area under the curve demonstrated that BW velocity (AUC =
.75) more accurately identified fallers than FW velocity (AUC = .70). A cut off score of 0.6 m/s
BW velocity identified fallers with a sensitivity of 100% and a specificity of 33%. To optimize the
specificity and positive likelihood ratio (+LR) value a cut off score of 0.4 m/s BW velocity
identifies fallers with a sensitivity of 58%, specificity of 82%, and a +LR of 3.2.
Figure 3.1 Backward walking velocity identifies fallers better than forward walking velocity in
elderly. All self-reported elderly fallers had a backward walking velocity of less than 0.6m/s.
3.4 Discussion
This is the first study to examine spatiotemporal gait measures in BW across adulthood
and to include middle-aged individuals and elderly individuals with impaired functional mobility
68
and/or reported falls. Backward and forward walking performance declined significantly in
elderly compared to young and middle-aged adults, with spatiotemporal measures of BW
declining more precipitously than FW. Age was more strongly related to BW spatiotemporal
measures than FW measures. In addition, we found that BW measures were significantly more
impaired in fallers than in non-fallers. Taken together, our findings support that BW is a valuable
addition to the examination of mobility.
Age-related changes in gait negatively impact function, as indicated by an increase in
falls and use of assistive devices in old age. Our results show that gait measures in both BW and
FW remained relatively stable in young and middle-aged but were significantly more impaired in
the elderly group (Table 3.2). Our results concur with previous findings of greater deficits in gait
measures in healthy elderly compared to young adults in FW and BW.5,6 These findings suggest
that changes in gait measures emerge after middle-age. This is consistent with research showing
that the cumulative effects of aging on muscle and neural tissue do not typically impact
functional performance until age 65 or later.15-18
Reversing from FW to BW led to a significant reduction in velocity, stride length, swing
percent, and increased stance and double support percent and BOS for all age groups (Table
3.2). The magnitude of these changes was more pronounced in elderly than in young and
middle-aged (Table 3.1), lending further support to the hypothesis that declines in gait measures
may not emerge until after middle-age. For example, when compared to FW, elderly individuals
decreased velocity by 51% (1.07m/s to 0.52m/s) in BW while young and middle-aged individuals
decreased their velocity by 24% (1.49m/s to 1.13m/s) and 30% (1.48m/s to 1.03m/s) ,
respectively. Similarly, Laufer (2003) reported that gait velocity decreased by 23% in young
adults (20-31 years), while older adults (65-89 years) experienced a 45% decrease in velocity.5
69
Possible reasons for these spatiotemporal changes in going from FW to BW are because BW
places greater demands on postural control systems due to lack of visual information and
because BW is not as habitually performed.5 Thus, decreased postural stability and/or fear of
falling may induce changes such as lower gait velocity and shorter steps to avoid falls.
Gait variability was more pronounced in elderly compared to young and middle-aged
individuals and was significantly increased by reversing from FW to BW in all age groups (Table
3.2). The magnitude of variability increase with direction reversal was significantly greater for
elderly compared to other age groups (Table 3.1). Our finding that BW is more variable than FW
concurs with previous findings in healthy elderly and individuals with Parkinson disease.5,8
Greater variability in BW than FW is clinically important since increased gait variability is
associated with a greater fall risk in the elderly and some neurological populations.7,19 Our
finding that step time CV in BW was significantly increased in fallers versus non-fallers (Table
3.4) would support the hypothesis that postural instability in BW may underlie the variability
changes.
Four out of five BW spatiotemporal measures and stride length CV were more closely
associated with age compared with the associations between age and FW measures (Table 3.3).
Some studies suggest that the control networks for BW may be independent of those for FW.8,20
Our results raise the possibility that these networks may be differentially impacted by the aging
process such that the BW system is more affected than the FW system. Further research is
needed to determine whether clinicians can identify mobility deficits earlier by examining BW in
addition to FW.
Backward walking measures were more sensitive than FW measures at differentiating
elderly fallers from non-fallers. Significant differences were noted between fallers and non70
fallers in 5 out of 6 BW measures while only two out of 6 measures, velocity and stride length,
were able to differentiate these groups in FW. Our finding that elderly fallers were on average
significantly more impaired on the TMT than non-fallers raises the possibility that impaired BW
performance may have contributed to lower TMT scores possibly due to difficulties with turning
and backing up to a chair. Therefore, further investigations to examine the relationship
between BW performance and functional mobility with aging seem warranted.
One of the most interesting and intriguing findings of our study was that BW velocity
identified fallers more accurately than FW velocity. Walking backward at a speed less than 0.6
m/s identified 100% of fallers and those who walked backward at a speed less than 0.4 m/s had
a 3.2 times greater likelihood of being a faller than those who walked faster. In comparison, FW
velocity at a cut off of 1.0 m/s only identified 83% of fallers. In agreement with our findings,
other studies have reported only a weak relationship between falling and forward gait
velocity.21,22 Further investigations to determine the validity of BW velocity to identify
individuals at risk of falls would be valuable.
This study has several limitations. Participants may not be representative as the sample
was limited to individuals living in Ohio, included more females than males, and elderly subjects
tended to be older and active. The cross-sectional design does not allow any conclusions to be
made regarding individual changes in gait measures over time.
The findings of this study extend the available information on age-related differences in
spatiotemporal measures of BW and highlight the importance of assessing BW as a component
of gait assessment in the elderly. Gait measures in both BW and FW were mostly equivalent in
the young and middle-aged groups but declined significantly in the elderly group. Velocity and
stride length demonstrated a more rapid and robust decline in BW than FW that was more
71
closely associated with age than other gait measures. In addition, BW velocity was found to be
significantly slower in elderly fallers than in non-fallers. All elderly fallers walked backward at a
speed less than 0.6 m/s. Based on these findings, healthcare practitioners are encouraged to
incorporate assessment of BW, in particular velocity, into their standard mobility evaluation.
Additional studies to examine the relationship between BW performance and functional
mobility and to determine the validity of BW velocity for the assessment of fall risk in different
populations and disease states are recommended.
72
3.5 References
1.
Chan BK, Marshall LM, Winters KM, Faulkner KA, Schwartz AV, Orwoll ES. Incident fall risk
and physical activity and physical performance among older men: the Osteoporotic
Fractures in Men Study. Am J Epidemiol. 2007;165(6):696-703.
2.
Bloem BR, Hausdorff JM, Visser JE, Giladi N. Falls and freezing of gait in Parkinson's
disease: a review of two interconnected, episodic phenomena. Mov Disord.
2004;19(8):871-884.
3.
McIlroy WE, Maki BE. Age-related changes in compensatory stepping in response to
unpredictable perturbations. J Gerontol A Biol Sci Med Sci. 1996;51(6):M289-M296.
4.
Maki BE, McIlroy WE. Control of rapid limb movements for balance recovery: age-related
changes and implications for fall prevention. Age Ageing. 2006;35 Suppl 2:ii12-ii18.
5.
Laufer Y. Age- and gender-related changes in the temporal-spatial characteristics of
forwards and backwards gaits. Physiotherapy Res Int. 2003;8(3):131-142.
6.
Laufer Y. Effect of age on characteristics of forward and backward gait at preferred and
accelerated walking speed. J Gerontol A Biol Sci Med Sci. 2005;60(5):627-632.
7.
Hausdorff JM, Mitchell SL, Firtion R et al. Altered fractal dynamics of gait: reduced strideinterval correlations with aging and Huntington's disease. J Appl Physiol. 1997; 82(1):262269.
8.
Hackney ME, Earhart GM. Backward walking in Parkinson's disease. Mov Disord. 2009;
24(2):218-223.
9.
Protas EJ, Mitchell K, Williams A, Qureshy H, Caroline K, Lai EC. Gait and step training to
reduce falls in Parkinson's disease. NeuroRehabilitation. 2005; 20(3):183-190.
10.
Shigematsu R, Okura T. A novel exercise for improving lower-extremity functional fitness
in the elderly. Aging Clin Exp Res. 2006;18(3): 242-248.
11.
Hackney ME, Earhart GM. Effects of dance on movement control in Parkinson's disease: a
comparison of Argentine tango and American ballroom. J Rehabil Med. 2009; 41(6):475481.
12.
Webster KE, Wittwer JE, Feller JA. Validity of the GAITRite walkway system for the
measurement of averaged and individual step parameters of gait. Gait Posture. 2005;
22(4):317-321.
73
13.
Smithson F, Morris ME, Iansek R. Performance on clinical tests of balance in Parkinson's
disease. Phys Ther. 1998; 78(6):577.
14.
Tinetti ME. Performance-oriented assessment of mobility problems in elderly patients. J
Am Geriatr Soc. 1986; 34(2):119-126.
15.
Marcell TJ. Sarcopenia: causes, consequences, and preventions. J Gerontol A Biol Sci Med
Sci. 2003; 58(10):M911-M916.
16.
Laughton CA, Slavin M, Katdare K et al. Aging, muscle activity, and balance control:
physiologic changes associated with balance impairment. Gait Posture. 2003; 18(2):101108.
17.
Wolfson LI. Gait and balance dysfunction: a model of the interaction of age and disease.
Neuroscientist. 2001; 7:178-183.
18.
Kerrigan DC, Todd MK, Della CU, Lipsitz LA, Collins JJ. Biomechanical gait alterations
independent of speed in the healthy elderly: evidence for specific limiting impairments.
Arch Phys Med Rehabil. 1998; 79(3):317-322.
19.
Yang YR, Yen JG, Wang RY, Yen LL, Lieu FK. Gait outcomes after additional backward
walking training in patients with stroke: a randomized controlled trial. Clin Rehabil. 2005;
19(3):264-273.
20.
Choi JT, Bastian AJ. Adaptation reveals independent control networks for human walking.
Nat Neurosci. 2007; 10(8):1055-1062.
21.
Shumway-Cook A, Brauer S, Woollacott M. Predicting the probability for falls in
community-dwelling older adults using the Timed Up & Go Test. Phys Ther. 2000;
80(8):896-903.
22.
VanSwearingen JM, Paschal KA, Bonino P, Chen TW. Assessing recurrent fall risk of
community-dwelling, frail older veterans using specific tests of mobility and the physical
performance test of function. J Gerontol A Biol Sci Med Sci. 1998; 53(6):M457-M464.
74
Chapter 4: Dual-Task Training for Balance and Mobility in a Person with Severe Traumatic
Brain Injury: A Case Study
4.1 Introduction
Deficits in sustained and divided attention are common following traumatic brain injury
(TBI). The frontal and temporal lobes, which are responsible for working memory and attention,1
are common contusion sites after motor vehicle accidents or blunt trauma2 because of their
proximity to bony prominences in the skull. TBI frequently results in impairment of cognition
and attention. Attention is a complex cognitive process contributing to arousal, alertness,
cognitive speed, working memory, and shifting and dividing attention.3 Divided attention is the
ability to respond to multiple stimuli simultaneously,4,5 and this ability is important for realworld function.
Dual-tasks are tasks that require divided attention; these tasks pair stimuli such as two
motor tasks or a cognitive task paired with a motor task (e.g., walking while talking).
Impairments in divided attention have specifically been linked with deficits in functional mobility
in neurological populations including traumatic brain injury (TBI),1,6 acquired brain injury,7
multiple sclerosis,8 and in elderly9 and healthy adults.10 Indeed, when challenged with dualtasks, individuals with Parkinson’s disease demonstrate slower walking speed, shorter strides,
increased double-support time and increased stride-to-stride variability.11 Similarly,
75
environments that challenge the ability to perform combined locomotor and attentional tasks
can identify residual deficits following a moderate to severe TBI.12
Automaticity indicates that a level of skill has been achieved in performance of a task
such that it requires little to no attention from the performer, to the extent that a second task
may be performed without degradation in skill of the primary task. Walking or standing balance,
which are typically automatic prior to a brain injury, are more attention-demanding following a
TBI. Recent evidence suggests that impaired executive function and attention negatively impact
walking function in persons with neurologic disorders.11 Indeed, lower cognitive functioning at
admission to inpatient rehabilitation has been associated with lower motor scores at discharge
in individuals with TBI.13 However, even in the presence of cognitive dysfunction, tasks such as
gait and balance can become skilled and more automatic with sufficient practice.7,14 The use of
dual-task training (DTT) may promote recovery of automaticity with basic tasks so that attention
can be focused on everyday activities such as navigating an environment or carrying on a
conversation without thought to the primary task of walking or balancing.
Evidence from other populations, including elderly adults,15 chronic stroke,16 and
Parkinson’s disease17 indicates that cognitive-motor DTT programs improve walking speed and
ability to dual-task. Recently, a virtual reality program trained cognitive-motor dual-tasks in an
individual following a concussion (mild TBI) , and showed improvements in dynamic and static
balance.18 Similarly, DTT improved balance during cognitive activities to a greater extent than
mobility training alone in healthy individuals and those with concussion.10,19 Neuropsychological
training programs in subacute and chronic severe TBI focusing either on cognitive dual-tasks20 or
working memory21 resulted in improvements in reaction times on visual-auditory dual-tasks.
76
However, it is unknown if DTT in individuals with severe TBI has value for improving safe and
functional mobility, or when might be the appropriate time to initiate such DTT.
The purpose of this case study was to determine the feasibility and potential value of
incorporating cognitive-motor DTT into physical therapy interventions for an individual with a
severe TBI. We speculated that DTT would be a useful addition to activity-dependent recovery
and might contribute to both improved performance on dual-task measures and improved
automaticity with locomotor tasks. To our knowledge, the case presented herein is the most
severe case described in rehabilitation intervention literature and represents the first report of a
cognitive-motor training program for severe TBI.
4.2 Case Description
The study was explained to both the subject of the case study and her family, and
informed consent was obtained from her parents, who were her legally authorized
representatives.
4.2.1 History
TK, a 26 year old female with no significant medical history, sustained a severe TBI
following a high speed motor vehicle accident. She lost consciousness and was transported to
the emergency department where imaging revealed areas of contusion in bilateral temporal
fossae, a subarachnoid hemorrhage in right frontal and parietal lobes and a small right temporal
fossa subdural hematoma. In addition, she presented with a left fixed pupil, and non-operative
left-sided clavicle, sacral, and rib fractures. Prior to this accident, TK was a healthy, college
graduate with no history of head injury or substance abuse. She was physically active and
participated in regular exercise and jogging. TK spent 25 days in acute care (post-injury days 177
25) where a PEG was placed and a tracheostomy performed to protect the airway; her stay was
complicated by difficulty weaning from the ventilator (Table 4.1). She discharged to a long term
acute care center for 21 days (post-injury days 25-46), where she was able to wean from the
ventilator and tolerate trach capping.
Post-Injury Day
Description
0
Sustained TBI
0-24
Acute care stay
25-45
Long term acute care stay
46
Initial PT evaluation at inpatient rehabilitation facility
46-57
 51
Pre-baseline phase (Standard Physical Therapy)
 Attempted pre-baseline testing; TK was unable to complete
dual-task activity
58-64
 60
Phase A (Standard Physical Therapy)
 Cleared PTA
65-71
Phase B (Standard Physical Therapy + Dual-Task Training)
Table 4.1 Timeline of events
TK was treated with divalproex sodium (Depakote) for seizures initially, but was weaned
due to abnormal liver lab values and subsequently placed on sertraline hydrochloride,
quetiapine (Seroquel) and trazodone for behavioral stabilization and mood control. TK’s only
other medications were for pain management of the non-operative fractures (fentanyl,
oxycodone), DVT prophylaxis (enoxaparin [Lovenox]), and stool softeners (senna, docusate
sodium). Unfortunately, both acute and long-term acute stays occurred outside our medical
78
system, so common prognostic variables (i.e., length of coma, Glasgow Coma Scale, intracranial
pressure and pupillary size22) were unknown. However, positive prognosis is associated with a
number of other variables, including number of years of education,23,24 younger age,25 high
socioeconomic status, and no prior history of head injury or substance abuse,24 all of which were
true for TK.
4.2.2 Examination
TK presented to the inpatient rehabilitation center for intense physical, occupational,
and speech therapy on post-injury day 46. Her cognitive function was categorized as Rancho IV
on the Rancho Los Amigos Level of Cognitive Function scale with left cranial nerve III palsy and
no weight bearing restrictions. The Rancho Los Amigos scale is used with the TBI population to
assess overall level of cognition (see 4.11, Supplemental Content 1).26 During evaluation, TK’s
agitation, restlessness, perseveration, and occasional aggression limited formal assessment of
pain and strength. Strength was therefore assessed through functional activity; TK was able to
toe walk, (but struggled to keep her left heel off the ground), squat and rise with assistance,
march with decreased left hip flexion, side step with reduced left step lengths, achieve
quadruped, perform single leg stance bilaterally, kneel with bilateral upper extremity support,
and half-kneel (but unable to maintain right knee up position). In summary, TK presented with
mild left-sided weakness and motor delays. TK’s sitting balance was rated as “fair”, she was able
to safely maintain unsupported sitting for static but not dynamic activities. Her standing
balance was “poor” and required both upper extremity and physical assistance for static and
dynamic tasks. TK scored 9/56 on the Berg Balance Scale, indicating a high fall risk. Transfers
from bed to wheelchair/tub/toilet required moderate assistance of two persons. Bed mobility
required minimal assistance. TK required 100% assistance for wheelchair mobility. Walking
79
required moderate assistance of two persons for safety. Gait deviations included a left
Trendelenburg, decreased left step length, and left leg dysmetria during swing phase. TK was
able to ascend and descend 4 steps with a hand rail and moderate assistance of one person
using a step-to pattern (Table 4.2).TK’s physical impairments were complicated by her level of
cognition and poor short-term memory. While she was able to follow simple motor commands,
she required minimal assistance for multi-step directions for ADLs and maximal cueing to
provide verbal responses. Throughout the evaluation, she was impulsive, agitated,
perseverative, and easily distracted.
Taken together, TK’s main impairments were left-sided weakness and incoordination,
poor balance and reduced cognition and attention to task, which led to functional limitations in
gait and mobility. Based on this examination, TK’s estimated length of stay was 2-3 weeks with a
discharge plan of returning home with her family. Physical therapy goals were: 1) ambulatory
transfers with stand-by assistance and no device to bed, toilet, car and on/off floor; 2) walk with
no device and stand-by assistance of family on smooth and uneven surfaces 1000 feet, 3
times/day with no loss of balance; 3) ascend/descend one flight of stairs with 1 railing, no
device, and contact-guard assistance of a family member. Accordingly, the physical therapy (PT)
plan of care included family education as well as balance, gait, and stair training.
80
Skill
Evaluation
Day 46
1
Begin
Phase A
Day 58
4
Begin
Phase B
Day 65
4
Bed, Chair,
Wheelchair
Transfer
Toilet
1
4
4
Transfer
Tub/Shower
1
4
4
Transfer
Walk
1
4
4
Assistance Mod Assist x2
CGA-HHA
CGA
Distance
400’ x 2
300’
>1000’
Wheelchair
1
3
3
Mobility
Stair Climbing
1
4
4
Table 4.2 Functional Independence Measure scores
End
Phase B
Day 71
5
Discharge
Goal
5
5
5
4
5
5
SBA
>1500’
3
5
SBA of family
1000’, 3x/day
5
4
4
7: Complete Independence (timely, safely); 6: Modified Independence (extra time, devices); 5:
Supervision (cueing, coaxing, prompting); 4: Minimal Assistance (performs ≥75% of task); 3:
Moderate Assistance (performs 50-74% of task); 2: Maximal Assistance (performs 25-49% of
task); 1: Total Assistance (performs <25% of task or requires 2 persons to assist).
Level of Assistance: Contact Guard Assistance (CGA); Handheld Assistance (HHA); Standby
Assistance (SBA)
4.3 Outcome Measures
The following outcome measures were used to identify functional deficits and monitor
progress in mobility, balance, divided attention and dual-task ability.
4.3.1 Overall Function
The Functional Independence Measures (FIM) assesses level of assistance required for
physical and cognitive independence in the areas of self-care, locomotion, mobility, sphincter
control, communication and social cognition.27, 28 The FIM is reliable in detecting changes in
independence over time and valid for predicting burden of care in TBI.28
81
4.3.2 Mobility
Walking speed was assessed using the 10-meter walk test, which is reliable29 and valid30
in severe TBI. The time to ascend and descend 10 stairs was recorded as a measure of functional
mobility.
4.3.3 Balance
The Berg Balance Scale is comprised of 14 items, each rated 0-4 for a maximal score of
56, which reliably measures static and dynamic balance during functional tasks in individuals
with brain injury.31 A score of less than 45 predicts fallers with 85% sensitivity and specificity.32
4.3.4 Divided Attention
The Walking While Talking Test (WWTT), a test of cognitive-motor dual-task ability, was
administered as described by Verghese et al;33 the participant is timed while walking 40 feet
with a 180 degree turn at the midpoint under three conditions: comfortable walking speed,
comfortable walking speed while reciting the alphabet (WWT-simple) and comfortable walking
speed while reciting every other letter of the alphabet (WWT-complex). A cut-off of 20 seconds
for the WWT-simple, and 33 seconds for the WWT-complex, predicts elderly fallers with a
specificity of 89% and 96%, respectively.33 While only 44% of individuals with acquired brain
injury were able to accurately perform the WWT-complex in a small pilot study7 (vs. 87% for the
WWT-simple), only the WWT-complex elicited a dual-task cost, suggesting that the rote tasks
such as recital of the alphabet may not be sufficiently challenging for ambulatory individuals
with brain injury.7 Reliability and validity of the WWTT have not been established in persons
with TBI.
The Trail Making Test A & B measures cognitive divided attention and visual tracking;34
Part A requires the patient to connect 25 numbered circles, in order, as quickly as possible
82
without picking up their pencil. Part B requires the patient to connect the 25 circles, now labeled
with numbers (1-13) and letters (A-L) in an alternating fashion (1-A, 2-B, 3-C, etc.)34 The Trail
Making Test is one of the most commonly administered neuropsychological exams and is a
sensitive measurement of divided attention in persons with TBI.35 Although disease-specific
norms have not been identified, normative data for healthy females aged 25-34 years old with
12-19 years of education are 23.3±8.0 for Part A and 52.8±20.4 for Part B.36
4.4 Prognosis
Recovery following TBI is highly variable. Despite the severity of TK’s injury, she had
many positive prognostic indicators including age, education, and medical history. However,
poorer prognosis is associated with increased severity of injury, which is determined by length
of post-traumatic amnesia (PTA). PTA represents an interval of confusion during which the
patient is likely to have behavioral disturbances as well as amnesia to ongoing events24 and
serves as an acute predictor of TBI outcome. PTA duration of more than 24 hours is classified as
severe TBI, while PTA duration of more than 4 weeks indicates a very severe TBI. Long term
outcomes indicate that patients with more than 3 weeks of PTA generally experience residual
cognitive deficits, in particular poor concentration and difficulty dividing attention at 1 year post
injury.24 TK’s progress toward clearance of PTA was monitored in an orientation group setting
using the orientation log (O-log), a 10-item scale that evaluates place, time, and situation
domains.37 This setting and procedure reliably measures the duration of PTA.38,39 TK cleared
PTA on post-injury day 60, indicating a very severe brain injury.24 Despite the severity of her
injury, TK’s prognosis with intensive therapy was considered to be fair for achieving functional
ambulation and self-care with supervision.
83
4.5 Assessment of Readiness to Initiate Dual-Task Training
TK received physical therapy services 60-90 minutes per day, 5-6 days per week for a
total of 26 days. Pre-baseline treatment occurred over 12 days (post-injury days 46-58) and
included standard PT (see 4.12, Supplemental Content II), which outlines examples of standard
PT). There is a paucity of evidence indicating the ability of individuals with TBI to benefit from
DTT while in PTA, or during the period that the Rancho level of cognitive function is categorized
as “confused” (Rancho IV-VI). Therefore, we assessed TK performing a dual-task activity (i.e.,
walking while answering an orientation question) once a week to determine her ability to
participate in training. With cognitive function classified as Rancho V on post-injury day 51, TK
demonstrated inability to continue walking while talking and therefore we deemed that she was
inappropriate for DTT. However, with cognitive function classified as Rancho V/evolving Rancho
VI on post-injury day 57, TK was deemed ready for DTT because she was able to continue
walking while answering orientation questions, although she was frequently incorrect, easily
distracted, and required verbal cues to continue walking. Collection of baseline outcome
measures occurred the following day (day 58) and baseline phase A began (Table 4.1), despite
the fact that TK had not yet cleared PTA.
4.6 Intervention
This case study included two phases; a 7-day baseline period (phase A) and a 7-day
dual-task intervention period (phase B) (Table 4.1). Outcome measures were assessed prior to
phase A, between phases and following phase B. The 7-day duration of each phase was
determined by the estimated length of stay for TK.
84
4.6.1 Phase A (post-injury days 58-64)
Phase A was a continuation of the standard PT care provided pre-baseline (see 4.12,
Supplemental Content II) with progressions matching TK’s functional improvements. In terms of
Gentile’s Taxonomy, the standard PT interventions represented closed environment body
stability and transport activities with and without inter-trial variability, with no manipulation,40
while progression included open environments. Phase A served as a baseline period at a time
when TK was able to participate in, and might receive benefit from, DTT. However, DTT was not
utilized during the baseline period. TK cleared PTA on post-injury day 60, or day 2 of phase A.
4.6.2 Phase B (post-injury days 65-71)
Phase B included standard PT (see 4.12, Supplemental Content II) supplemented with a
directed intervention of DTT for 7 days. DTT occurred for an average of 15 minutes out of each
30-minute session yielding at least 180 total minutes of DTT over 7 days. Interventions included
motor-motor and cognitive-motor dual-tasks (see 4.13, Supplemental Content III) that
represented an increase in complexity on Gentile’s Taxonomy by including manipulation to both
open and closed environments during body stability and transport activities.
As independent mobility increased, TK demonstrated multiple gait abnormalities
including step asymmetry, slow self-selected speed, poor arm swing, decreased push-off and
reduced left knee extension during heel strike. These deviations did not improve with verbal
cueing, and TK remained fearful of walking at a normal or near-normal pace. Thus, body-weight
supported treadmill training was used to increase motor demands (i.e., higher velocity) without
risking injury or falls. The goal of treadmill training was to attain long duration bouts of stepping
without gait deviation and physical or verbal assistance. Four sessions of treadmill training
occurred during phase B of the study. During each treadmill session, TK walked an average of 20
85
minutes with stepping bouts of approximately 4-5 minutes. We adjusted body weight support
(BWS) until gait deviations for asymmetry were minimized, then increased treadmill speed to
challenge locomotion. Speed was 3.0mph - 3.5mph for the first two sessions progressing to
4.0mph by session four. TK required BWS during the first two sessions and only the safety
harness was in place for sessions 3 and 4. The BWS harness system at our facility (LiteGait,
Mobility Research, Tempe, Arizona, USA) does not report percentage of BWS. During all
sessions, TK’s heart rate and blood pressure were monitored and remained within an expected,
safe range. In accordance with other phase B sessions, cognitive tasks (see 4.13, Supplemental
Content III) such as addition, subtraction and synthesis of lists were paired with treadmill
walking in the same manner as overground walking. However, these tasks were only added after
TK was able to maintain a steady speed on the treadmill without assistance and comprised only
60-90 seconds of each stepping bout. Despite the greater demands of treadmill training, TK
participated in dual-tasks during all sessions.
4.7 Outcomes
Performance on divided attention tests (WWTT and Trail Making Test) in phase B
indicated modest improvement when compared to phase A (Table 4.3). TK required fewer cues,
and had fewer errors and faster completion times for these tests.
86
Begin
Phase A
Day 58
Rancho Level
Berg Balance Score
Gait Speed (m/s)
Time to Ascend 10
Stairs (s)
Time to Descend 10
Stairs (s)
Walking While Talking
Test (WWTT) (s)
V evolving VI
Not Assessed
0.60
8.87
17.51
17.81a
19.21b
21.63c
(6 errors)
Trail Making Test
A: 41
Parts A & B (s)
(3 errors)
B: 173 (unable to
complete)
Table 4.3 Motor and dual-task outcomes
a
comfortable gait speed;
b
WWT-simple;
c
Begin
Phase B
Day 65
(% improvement
phase A)
VI evolving VII
44/56
0.69
(15%)
7.82
(12%)
End
Phase B
Day 71
(% improvement
phase B)
VI evolving VII
45/56
1.04
(51%)
6.72
(14%)
14.85
(15%)
7.36
(50%)
15.47a
16.08b
19.40c
(1 error)
A: 38
(2 errors)
B: 119
(2 errors)
(13%)
(16%)
(10%)
12.09a
12.82b
16.97c
(1 error)
A: 28
(0 errors)
B: 71
(3 errors)
(22%)
(20%)
(13%)
(7%)
(31%)
(26%)
(40%)
WWT-complex
TK showed clear improvements in functional tasks (Table 4.3 & Figure 4.1A). She had a
3-fold greater rate of change in walking speed during phase B than phase A (0.35m/s vs. 0.1 m/s,
respectively). TK’s time to descend stairs was markedly reduced after DTT (7.49 s) but only
modestly reduced during the baseline phase (2.66 s) (Table 4.3 & Figure 4.1B).
87
Figure 4.1 Changes in gait speed (A) and time to descend stairs (B) by phase. Each phase was 1
week long. While there were improvements during the baseline phase, note an increase in the
rate of change during the intervention phase.
*: represents a change that is greater than the MCID for gait speed in subacute post-stroke
populations41 and reaches the threshold of 0.8m/s established for community ambulators.42
Throughout her treatment (Day 46  Day 71), TK demonstrated improvements on the
FIM (Table 2) and Berg Balance Scale as expected with intensive multidisciplinary therapy. The
majority of improvement in balance and FIM scores occurred prior to phase A (Table 3) and
paralleled TK’s increased activity tolerance. Treadmill walking was a successful addition to TK’s
locomotor training; after two sessions TK was able to achieve full arm swing and improved gait
mechanics without cueing while walking at 3.5 mph. During the fourth and final treadmill
walking session, TK was able to achieve symmetrical stride lengths.
88
4.8 Discussion
Cognitive-motor DTT programs may be effective in ameliorating attentional deficits
during mobility in populations of mild TBI18 and other neurological populations.15-19 Longitudinal
evidence indicates that individuals with severe TBI with greater than 3 weeks of PTA often have
residual cognitive deficits in the areas of concentration and ability to divide attention;24 thus, we
anticipated that TK may benefit from cognitive-motor DTT. Although the DTT intervention itself
is not unique, its application to severe TBI is novel, and allowed us to target specific attentional
deficits that may have been inhibiting safe mobility and to focus upon rehabilitation of both
divided attention and mobility impairment with challenging customized tasks.
To our knowledge this case represents the only report of cognitive-motor DTT in a
person with severe TBI. TK’s PTA of 60 days far exceeds that of previously reported cognitivecognitive dual-task interventions in severe TBI (7-30 days20,21). While prior work exploring DTT in
TBI has targeted subjects in the outpatient arena; the DTT intervention described was
successfully administered within an inpatient rehabilitation setting over a short, clinically
feasible duration of 7 days. Other studies targeting cognitive improvements20,21 included 24 
64 hours of DTT while our study and that of Rábago and Wilken,18 which also targeted mobility
and balance in mild TBI, required markedly shorter durations (3 and 6 hours respectively) to
achieve gains. This short duration may highlight the importance of task specificity in DTT.
TK’s overall improvements illustrate the feasibility of supplementing standard PT with
DTT in persons with severe TBI. TK’s mobility continually improved through the baseline phase
when no DTT was provided, which is consistent with expected recovery observed with intensive
multidisciplinary therapy. However, the change in walking speed during the dual-task phase of
0.35 m/s was considerably larger than the minimal clinically important difference (MCID) of
89
0.16m/s established in subacute stroke populations.41 Of greater importance is the relationship
of increased walking speed to community ambulation. Only after phase B did TK surpass 0.8 m/s
to achieve speeds sufficient for community ambulation.42
TK demonstrated greater gains in time to descend stairs than to ascend stairs during the
dual-task phase. This outcome is perhaps due to several factors including improved eccentric
control, left-sided attention, and automaticity. Both walking and stair descent are considered
automatic movements and objective improvement in these activities suggests a positive
relationship between DTT and functional mobility.11,14 Interestingly, while many of KT’s FIM
scores plateau after Day 58 (beginning of phase A), the acquisition of a walking speed sufficient
for community ambulation occurs well beyond this point. Thus, program evaluation for dual-task
training after TBI must be carefully considered. It appears that objective outcome measures
such as walking speed and dual-task measures (WWTT) may prove to be more sensitive
indicators of gains and treatment efficacy than measures focused on burden of care.
Previous studies reported difficulty with participation in the WWT-complex;
however, we did not encounter this problem.7 All parts of the WWTT were easily and quickly
administered, even during periods when PTA had not fully resolved. Unfortunately we did not
record the time to complete the complex cognitive task in sitting, which would have allowed for
the calculation of dual-task cost. However, based on our experience, WWTT could be applied
clinically to gain objective measures of dual-task ability in individuals with severe TBI. While TK
also demonstrated improvements on the Trail Making Test, it is important to note that this penand-paper test does not measure improvements in functional mobility or real-world dual-task
situations.
90
The motor and cognitive tasks utilized throughout the progression of this program were
dictated by TK’s functional gains and level of cognition. Her attention to task, memory and
agitation all served as guides for treatment progression, particularly for the secondary cognitive
tasks. As TK’s attention improved, we were able to introduce more cognitively challenging
secondary tasks. If a task was introduced that caused frustration or severe detriment in the
primary task, it was abandoned for an easier task. Increasing secondary task difficulty mirrored
TK’s improvement in ability to dual-task throughout phase B.
4.8.1 Limitations
Several factors beyond the addition of DTT may have contributed to the positive
outcomes seen in TK. Indeed, intensive multidisciplinary therapy alone may account for the
improvements in walking speed and dual-task measures. Additionally, the use of treadmill
training as a progression of gait training may be challenged; it has been used in TBI populations
to improve both cardiorespiratory capacity and spatiotemporal characteristics of unsupported
overground walking.43 In this case, it allowed us to modulate speed in a safe environment. All
treadmill training occurred during phase B of the study and may have contributed to the
observed gains in walking speed. Finally, TK was started on amantadine on post-injury day 65,
the same day that phase B began. In TBI, amantadine can be used off-label to improve sustained
and divided attention, increase arousal and reduce impulsivity. Whether this drug contributed
to the observed functional improvements is unclear because other mood stabilizing drugs
including quetiapine44 and sertraline45 were prescribed during phase A to reduce agitation and
improve cognition and alertness.
The ability of individuals who have not cleared PTA to benefit from DTT is also an
important consideration. With PTA, patients may be unable to retain training effects within or
91
between treatments. For TK, about 33% of phase A (2 days) occurred with PTA. Thus, the lower
rate of recovery during this phase may be due in part to PTA. However, because TK’s
impairments in attention to task were contributing to her mobility, we felt that a program of
DTT may be beneficial. These limitations represent the true conditions encountered in the clinic
and the best interests of the patient, without regard to the experimental design. Overall, these
clinical modifications alone or in combination with the dual-task intervention, were associated
with greater gains in phase B than were seen in phase A.
4.9 Summary
This case study provides support for the feasibility and possible benefits of DTT in
addition to standard PT for an individual with severe TBI. Clinically meaningful improvements
were found in functional mobility and ability to dual-task as measured by the WWTT. Specific
improvements seen in the intervention phase cannot be solely attributed to DTT, but this
training did not disrupt the progression of recovery. Indeed, gains in locomotion and attention
with the addition of DTT outpaced those seen with standard physical therapy two- to three-fold.
The results of this case study are encouraging and support the need for studies with larger
sample sizes to identify the optimal time to initiate DTT following TBI, further examine whether
gains in functional mobility and cognition can be enhanced with DTT in severe TBI, and explore
the transfer of DTT to real-world dual-task demands.
92
4.10 References
1. Mathias JL, Wheaton P. Changes in attention and information-processing speed following
severe traumatic brain injury: a meta-analytic review. Neuropsychology. 2007; 21(2):212223.
2. Kraus MF. Traumatic Brain Injury: A Brief Overview of Traumatic Injuries and the
Neurobehavioral Deficits That Can Occur. Urbana, Illinois: University of Illinois Press; 2007.
3. 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(5):647–655.
4. Mateer CA, Kerns KA, Eso KL. Management of attention and memory disorders following
traumatic brain injury. J Learn Disabil. 1996; 29: 618-632.
5. McDowd JM. An overview of attention: behavior and brain. J Neurol Phys Ther. 2007; 31: 98103.
6. Azouvi P, Couillet J, Leclercq M, et al. Divided attention and mental effort after severe
traumatic brain injury. Neuropsychologia. 2004; 42:1260–1268.
7. McCulloch K. Attention and dual-task conditions: physical therapy implications for
individuals with acquired brain injury: J Neurol Phys Ther. 2007; 31: 104-118.
8. Hamilton F, Rochester L, Paul L, Rafferty D, O’Leary CP, Evans JJ. Walking and talking: an
investigation of cognitive-motor dual tasking in multiple sclerosis. Mult Scler. 2009; 15(10):
1215-1227.
9. 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):1–14.
10. Pellecchia GL. Dual-task training reduces impact of cognitive task on postural sway. J Mot
Behav. 2005; 37(3): 239-246.
11. Yogev-Seligmann G, Hausdorff JM, Giladi N. The role of executive function and attention in
gait. Mov Disord. 2008; 23(3): 329-342.
12. Vallée M, McFadyen BJ, Swaine B, et al. Effects of environmental demands on locomotion
after traumatic brain injury. Arch Phys Med Rehabil. 2006; 87: 806-813.
13. Bogner JA, Corrigan JD, Fugate L, Mysiew WJ, Clinchot D. Role of agitation in prediction of
outcomes after traumatic brain injury. Am J Phys Med Rehabil. 2001; 80(9): 636-644.
93
14. Roskell C, Cross V. Attention limitation and learning in physiotherapy. J Physiother. 1998;
84(3):118–125.
15. Silsupadol P, Shumway-Cook A, Lugade V, vonDonkelaar P, Chou LS, Mayr U, Woollacott MH.
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: 381- 387.
16. Yang YR, Wang RY, Chen YC, Kao MJ. Dual-task exercise improves walking ability in chronic
stroke: a randomized controlled trial. Arch Phys Med Rehabil. 2007; 88: 1236- 1240.
17. Yogev-Seligmann G, Giladi N, Brozgol M, Hausdorff JM. A training program to improve gait
while dual- tasking in patients with Parkinson’s disease: a pilot study. Arch Phys Med
Rehabil. 2012; 93(1): 176-181.
18. Rábago CA and Wilken JM. Application of a mild traumatic brain injury rehabilitation
program in a virtual reality environment: a case study. J Neurol Phys Ther. 2011; 35: 185193.
19. Parker TM, Osternig LR, Lee HJ et al. The effect of divided attention on gait stability
following concussion. Clin Biomech. 2005; 20: 389-395.
20. Couillet J, Soury S, Lebornec G, et al. Rehabilitation of divided attention after severe
traumatic brain injury: a randomised trial. Neuropsychol Rehabil.2010; 20(3): 321-339.
21. Vallat-Azouvi C, Pradat-Diehl P, Azouvi P. Rehabilitation of the central executive of working
memory after severe traumatic brain injury: two single-case studies. Brain Inj. 2009; 23(6):
585-594.
22. Jiang J, Gao G, Li W, et al. Early indicators of prognosis in 846 cases of severe traumatic brain
injury. J Neurotrauma. 2002; 19(7): 869-874.
23. deGuise E, LeBlanc J, Feyz M, Lamoureux J. Prediction of outcome at discharge from acute
care following traumatic brain injury. J Head Trauma Rehabil. 2006; 21(6): 527-536.
24. Chua, KSG, Ng YS, Yap SGM, Bok CW. A brief review of traumatic brain injury rehabilitation.
Ann Acad Med Singapore. 2007; 36: 31-42.
25. Katz, DI, White DK, Alexander MP, Klein RB. Recovery of ambulation after traumatic brain
injury. Arch Phys Med Rehabil. 2004; 85: 865-869.
26. Rancho Los Amigos National Rehabilitation Center. Family Guide to the Rancho Levels of
Cognitive Functioning. Available from: http://www.rancho.org/research/bi_cognition.pdf.
Accessibility verified July 12, 2012.
94
27. Linacre JM, Heinemann AW, Wright BD et al. The structure and stability of the functional
independence measure. Arch Phys Med Rehabil.1994; 75(2): 127-132.
28. Cusick CP, Gerhart KA, Mellick DC. Participant-proxy reliability in traumatic brain injury
outcome research. J Head Trauma Rehabil. 2000; 15(1): 739-749.
29. VanLoo MA, Moseley AM, Bosman JM et al. Test-re-test reliability of walking speed, step
length and step width measurement after traumatic brain injury: a pilot study. Brain Inj..
2004; 18(10): 1041-1048.
30. Moseley Am, Lanzarone S, Bosman JM, et al. Ecological validity of walking speed assessment
after traumatic brain injury: a pilot study. J Head Trauma Rehabil. 2004; 19(4): 341-348.
31. Newstead AH, Hinman MR, Tomberlin JA. 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.
32. Medley A, Thompson M, French J. Predicting the probability of falls in community dwelling
person with brain injury: a pilot study. Brain Inj. 2006; 20 (13-14): 1403-1408.
33. Verghese J, Buschke H, Viola L, et al. Validity of divided attention tasks in predicting falls in
older individuals: a preliminary study. J Am Geriatr Soc. 2002; 50(9): 1572-1576.
34. Crowe SF. The differential contribution of mental tracking, cognitive flexibility, visual search
and motor speed to performance on Parts A and B of the Trail Making Test. J Clin Psychol..
1998; 54: 585-591.
35. Lange RT, Iverson GL, Zakrzewski MJ, Ethel-King PE, Franzen MD. Interpreting the trail
making test following traumatic brain injury: comparison of traditional time scores and
derived indices. J Clin Exp Neuropsychol. 2005; 27: 897-906.
36. Soukup VM, Ingram F, Grady JJ, Schiess MC. Trail Making Test: issues in normative data
selection. Appl Neuropsychol. 1998; 5(2): 65-73.
37. Jackson WT, Novack TA, Dowler RN. Effective serial measurement of cognitive orientation in
rehabilitation: the Orientation Log. Arch Phys Med Rehabil. 1998; 79: 718-720
38. Mysiw WJ, Bogner JA, Arnett JA et al. The orientation group monitoring system for
measuring duration of posttraumatic amnesia and assessing therapeutic interventions.
J Head Trauma Rehabil. 1996; 14(6):1-8.
39. Saneda DL, Corrigan JD. Predicting clearing of post-traumatic amnesia following closed head
injury. Brain Inj.. 1992; 6(2): 167-174.
95
40. Gentile AM. Skill acquisition: Action, movement, and neuromotor processes. In: Carr JH,
Shepherd RB, Gordon J, Gentile AM, Held JM, eds. Movement Science. Foundations for
Physical Therapy in Rehabilitation. 1st ed. Rockville, MD: Aspen Publishers; 1987: 93-154.
41. Tilson JK, Sullivan KJ, Cen SY, et al. Meaningful gait speed improvement during the first 60
days poststroke: minimal clinically important difference. Phys Ther. 2010; 90(2): 196-208.
42. Fritz S and Lusardi M. Walking speed: the sixth vital sign. J Geriatr Phys Ther. 2009; 32(2):
2-5.
43. Mossbeág KA, Orlander EE, Norcross JL. Cardiorespiratory capacity after weight-supported
treadmill training in patients with traumatic brain injury. Phys Ther. 2008; 88(1): 77-87.
44. Kim E, Bijlani M. A pilot study of Quetiapine treatment of aggression due to traumatic brain
injury. J Neuropsychiatry Clin Neurosci. 2006; 18(4): 547-549.
45. Meythaler JM, Depalma L, Devivo MJ et al. Sertraline to improve arousal and alertness in
severe traumatic brain injury secondary to motor vehicle crashes. Brain Inj. 2001; 15(4):
321-331.
96
4.11 Supplemental Content I: Rancho Los Amigos Levels of Cognitive Function Scale26
Rancho Los Amigos
Level of Cognitive
Function
I. No Response
Behavioral Descriptors Related to
Attention: "A person at this level
may demonstrate:"
Absence of observable change in
behavior and response to stimuli.
Recommended Strategies for
Family/Friends Related to
Attention
Keep comments and questions short
and simple. Keep the room calm and
quiet. Care givers should explain
what is about to be done in a normal
tone of voice.
II. Generalized Response
Response to external stimuli with
physiological changes, gross body
movement and/or non‐purposeful
vocalization.
Periods of waking off and on during
the day. Slow and inconsistent
responses, but related to type of
stimulus
Alertnesspresented.
and a heightened state of
Same approach as Level I.
III. Localized Response
IV. Confused‐Agitated
V. Confused‐
Inappropriate‐Non‐
Agitated
activity. Inability to pay attention or
ability to concentrate for only a few
seconds.
Alertness and ability to attend for a
few minutes at a time.
Restlessness when tired or over‐
stimulated. Difficulty starting and
completing activities. Need for
step by step help to finish tasks.
Same approach as Levels I‐II. Allow
extra response time and rest.
Same as for Levels I‐III; orient and
reassure patient. Allow as much
activity as is safe.
Repeat things as needed. Help
organize and initiate familiar
activities. Give frequent rest
periods to allow refocusing of
attention.
VI. Confused‐ Appropriate Ability to attend to familiar tasks
in quiet environment for 30
minutes with some help provided.
Emerging awareness of how to care
for own basic needs and respond to
family and others.
Provide help in starting and
continuing activities. Repeat
activities and information to
improve learning and recall.
VII. Automatic‐
Appropriate
Difficulty paying attention in
distracting or stressful situations.
Difficulty planning, starting and
following thorough with activities.
Delayed processing, especially in
complex/stressful situations.
Approach for Levels VII and VIII
are the same: Provide support and
guidance for problem solving and
decision making. Encourage
continued therapy.
VIII. Purposeful and
Appropriate
Ability to learn new things at a
slower rate. Ability to use
compensatory strategies. Delayed
processing or problem solving
difficulties in complex and stressful
situations.
Encourage independence within
appropriate safety guidelines
(physical and cognitive). Encourage
the use of compensatory strategies.
Encourage participation in counseling
and support groups to assist with
adjustment issues.
97
4.12 Supplemental Content II: Examples of Standard Physical Therapy
Activity
Bed Mobility
Initial Intervention
Rolling, scooting, supine<->sit
Progression of Intervention
Reduced assistance provided
Transfers
Sit<->stand
Bed/mat <-> wheelchair/chair
Reduced assistance provided;
addition of floor<->stand and
stoop<->stand
Gait Training
Level surfaces
Stair Climbing
Bilateral handrails, nonreciprocal
Reduced assistance provided;
addition of uneven surfaces
(grass/gravel/inclines) and
variable surfaces (tile to
carpet)
Reduced assistance, one
handrail, reciprocal
Static standing, narrowed
base-of support and quick
turning/ stopping
Obstacle courses to challenge
dynamic balance, single-leg
activities, backward walking
Single-task during rest breaks
or prior to start of therapy
session; simple orientation
questions
Single-task; more complex or
challenging questions
Balance
Orientation
98
4.13 Supplemental Content III: Description of Dual-Task Training Program
The following are examples of motor-motor and motor-cognitive pairings and progressions of
pairings used in the Phase B. These pairings are by no means absolute; gross and fine motor
tasks were paired with various cognitive tasks throughout Phase B to create unique pairings as
appropriate.
Examples of Paired Motor‐Motor Tasks
Primary Motor
Task
Tandem Walk
Progression: Heel & Toe Walk
Secondary Motor
Task
Secondary Cognitive
Task
X
2
Figure of 8
Backward Walk
Carrying bags of different
sizes (groceries)
X
3
Forward Walk
(smooth/uneven surfaces)
Progression: against resistance,
fast walking
Carrying objects
(laundry basket, plate)
X
4
Balance on a Foam Surface
(static and dynamic)
Folding laundry
X
5
Cone Walking
(single leg balance)
Holding hand weights
X
6
Stair Climbing
Progression: Step‐ups
7
Stepping to Targets
Obstacles (weaving, stepping
over)
Progression: Side‐stepping, side‐
step against resistance, object
pick‐ up (stoop & squat)
1
X
Object toss
X
Continued
99
4.13 Supplemental Content III: Description of Dual-Task Training Program Continued
Examples of Paired Motor‐Cognitive Tasks
Primary Motor
Task
Tandem Walk
Heel & Toe Walk
Secondary Motor
Task
X
Secondary Cognitive
Task
Addition
2
Figure of 8
Backward Walk
X
Subtraction
3
Forward Walk
(smooth/uneven surfaces)
Progression: against resistance,
fast walking
X
Spelling
4
Balance on a Foam Surface
(static and dynamic)
X
Puzzle or manual
sorting/
categorization of items
5
Cone Walking
(single leg balance)
X
Remembering lists
6
Stair Climbing
Progression: Step‐ups
X
Synthesis of
lists/categories
7
Stepping to Targets
Obstacles (weaving, stepping
over)
Progression: Side‐stepping, side‐
step against resistance, object
pick‐up (stoop & squat)
X
Match footsteps to
auditory
or visual cue
1
100
Chapter 5: Dual-Task Walking Variability is Associated with Stroop and Dual-Task
Questionnaire Performance in Multiple Sclerosis
5.1 Introduction
Multiple Sclerosis (MS) is the most common progressive neurological disease1 and the
main cause of neurological disability2 in young adults. Within 15 years of onset of MS, up to 50%
of individuals will require assistance with walking3 and 10% are restricted to a wheelchair.4 Fall
risk in MS has been linked to impairments in both balance and quality of gait.5
When compared to healthy controls, individuals with MS, even those with short disease
duration, demonstrate decreased velocity, step length, cadence and step time, as well as a
wider base of support6 and increased swing time variability.7 Variability in gait parameters has
been correlated with fall risk in the elderly,8-10 Parkinson’s disease (PD),11 and Alzheimer’s
disease.12
In addition to mobility deficits, 45% of individuals13 with relapsing remitting MS (RRMS),
including those in the early disease stages,2 demonstrate significant cognitive dysfunction in the
areas of information processing speed, divided attention, and memory.14 Impairments in
cognition and mobility can lead to difficulties with performing two tasks at once
(i.e., dual-tasking) such as walking and talking. Indeed, individuals with MS devote greater
cognitive reserve to walking than healthy controls, resulting in even greater variability in
101
spatiotemporal measures of gait (e.g., step length, velocity) when walking is paired with a
cognitive task,15 even in the early disease stage16 when compared to single-task conditions. Not
surprisingly, poor concentration17 and difficulty with divided attention15, 18-19 necessary to
perform dual-tasks have been linked with accidental falls in individuals with MS.
Backward walking is required for successful completion of many daily tasks including
turning and maneuvering. Indeed, eight of ten independently ambulating individuals with MS
have delayed postural reactions to backward translations in standing.20 This is explained in part
by the fact that poor proprioception underlies imbalance and falls in MS21 and that backward
walking and stepping requires increased proprioception when compared to forward walking
because of the lack of visual cues.22 Backward walking is not regularly assessed clinically, but
could be a useful assessment in individuals with neurologic conditions; indeed, individuals with
PD have increased stance variability in backward walking when compared to forward walking,
while controls had similar amounts of variability in both backward and forward walking.23
Furthermore, our lab has previously shown that backward walking speed is sensitive to agerelated changes24 and that speeds less than 0.6 m/s can accurately discriminate between elderly
fallers and non-fallers. This suggests that backward walking could represent an important
rehabilitation target for fall reduction.
Although gait changes in forward walking have been well documented in MS,6,25
backward walking in MS has not been examined. The relationship between backward walking
and attention has also been briefly explored in previous studies; individuals with PD with
reduced attention capabilities demonstrated slower velocity, less swing and shorter stride
lengths in both forward and backward walking and during dual-tasks than those with PD with
better attention capabilities and healthy controls.26 Thus, poor attention can exacerbate gait
102
deficits in both forward and backward directions26 and raises a question of whether assessment
of tasks that may amplify gait variability, such as backward walking and dual-task walking, are
useful tools for identification of individuals with mobility deficits in the clinical setting. While the
connection between cognition and quality of dual-task forward walking has been established in
older adults,27-28 such a connection has not been established with backward walking. Thus,
exploration of backward walking and dual-task walking spatiotemporal measures and their
relationships with baseline measures of cognition in individuals with MS warrants investigation.
The objectives of this study were to:
1. Compare spatiotemporal measures in backward walking in individuals with RRMS and
age and gender matched healthy adults.
2. Compare spatiotemporal measures during forward and backward dual-task walking in
individuals with RRMS and age and gender matched healthy adults.
3. Examine the relationship between neuropsychological test performances and variability
of spatiotemporal measures of backward and dual-task gait in individuals with RRMS.
4. Examine the usefulness of the Dual-Task Questionnaire (DTQ), a subjective measure of
dual-task ability, by exploring its relationship with neuropsychological test performance
and variability of dual-task gait measures in RRMS.
5. Explore the relationship between neuropsychological test performance and calculated
dual-task cost for forward and backward walking in RRMS.
6. Explore the relationships between reported falls, neuropsychological test performance,
and variability of backward and dual-task gait measures in individuals with RRMS.
103
We hypothesized that: 1)individuals with RRMS would demonstrate significantly greater
variability and slower velocity during backward walking and dual-task walking compared to
healthy control subjects; 2) variability of backward walking measured with coefficients of
variation would correlate with neuropsychological test performance; 3) DTQ scores, calculated
dual-task cost, and falls would negatively correlate with neuropsychological test performance; 4)
falls would be positively correlated with variability of backward and dual-task gait.
5.2 Methods
This study uses preliminary baseline data from an ongoing randomized controlled trial
examining the effects of Dance Dance Revolution on mobility, brain plasticity and cognition in
individuals with Multiple Sclerosis. This trial was approved by the Institutional Review Board. All
participants signed consent forms prior to participation.
5.2.1 Subjects
Eleven volunteers with RRMS and 11 age and gender-matched healthy adults
participated in this study. Individuals attended a single hour-long session where they were
assessed by a physical therapist. All participants ambulated across the GAITRite System (V3.9,
MAP/CIR Inc.), a 4.88 meter electronic walkway that records spatiotemporal measures (e.g.,
velocity, step length, double support) of gait and is reliable and valid for use in individuals with
MS.6 Individuals ambulated across the GAITRite for four trials each under four conditions: a)
forward walking comfortable speed (FW); b) forward walking with a cognitive task of serial 3
subtraction starting at 97 (FWC); c) backward walking comfortable speed (BW); and d) backward
walking with a cognitive task of serial 3 subtraction starting at 95 (BWC). Participants were
instructed to walk 2 meters before and after the walkway to allow for acceleration and
104
deceleration. Participants completed all trials without an assistive device or bracing, wearing a
gait belt and guarded by a physical therapist for safety. The final three trials for each condition
were averaged and the spatiotemporal parameters of gait were exported for analysis and used
to calculate coefficients of variation (CV). Dual-Task Cost (DTC) was calculated by comparing
time to complete FW vs. FWC, and BW vs. BWC. The DTC for gait was calculated using the
following formula (FW vs. FWC provided as an example):28
DTC = [(Gait Time FWC – Gait Time FW) / Gait Time FW]* 100
All participants completed a self-reported fall history (number of falls in the past week,
8 weeks, and past 6 months) as well as the DTQ, a 10 item subjective questionnaire of everyday
tasks involving dual-tasking, to be rated for frequency of difficulty from 0-4 (4=highest
frequency).29 Although reliability, validity and cut-off scores have not been established, the DTQ
has been shown to be sensitive to the effects of brain injury and Alzheimer’s disease;29 it has
not been utilized in individuals with MS.
Finally, all participants completed three neuropsychological tests: a) Symbol Digit
Modality Test (SDMT); b) a word generation task; and the c) Stroop Task. The SDMT is a sensitive
test30-31 of processing speed, visual tracking, and rapid decision making32 and has established
validity and reliability in MS.30,33-34 During the SDMT, subjects are presented with nine symboldigit pairings and asked to match as many symbols to their corresponding number as possible in
90 seconds. The number of responses was recorded. The phonemic fluency word generation
task requires participants to generate as many words beginning with the letter “F” as possible in
60 seconds. This is followed by generating words with A and S. The total number of responses in
three 60 second bouts was recorded.35 The word generation task is a test of working memory
and executive function.35 During the Stroop task,14 participants are presented with three
105
different conditions: the congruent condition, where the color and the word match (e.g., the
word RED printed in red ink); the neutral condition, where color words are written in black ink;
and the incongruent condition, where the ink color will be incongruent with the meaning of the
word (e.g., the word GREEN printed in blue ink). For each condition the accuracy of the stated
word or color provided was recorded. The Stroop is a test of executive function, selective
attention, processing speed, and response inhibition.35 Individuals with MS completed these
tests on the computer in addition to a number of other neuropsychological tests, while healthy
controls completed a pen and paper version of these tests. The computerized and paper
versions of the Stroop task vary slightly. The computer version includes a defined set of prompts
and responses are not time-dependent. The paper version requires participants to provide as
many responses as they can in 45 seconds. In both cases, the accuracy of the given responses
was recorded.
5.2.2 Data Analysis
An a priori sample size calculation based upon forward gait velocity values for individuals
with MS and healthy controls obtained from Givon et al.6 indicated that with a conservative twosample t-test, 7 pairs provided 80% power to detect differences in forward walking velocity
between individuals with MS and healthy controls at p=0.01 with an effect size of 2.22. We set
the family-wise error rate at 0.05 to control for type I error rate. Statistical analyses were
performed using SPSS version 19.0. Individuals with MS and healthy controls were compared on:
a. spatiotemporal measures of gait (normalized velocity, stride length, double support %)
and coefficients of variation (CV) for stride length and double support time) for FW
(Hoetelling’s T2), FWC, BW, and BWC (Wilcoxon-Signed Rank Test)
b. neuropsychological test performance (Wilcoxon Signed Rank Test)
106
Correlational analysis was performed to examine the relationship between:
a. neuropsychological test performance and variability of BW and dual-task walking (FWC,
BWC) in RRMS
b. variability of BW and dual-task walking (FWC, BWC) and the DTQ to explore the utility of
this subjective tool in ambulatory subjects with RRMS
c. DTQ and neuropsychological test performance in RRMS
d. DTC of forward and backward walking and neuropsychological test performance in
RRMS
e. reported falls, neuropsychological test performance and variability of backward and
dual-task gait (FWC, BWC) in RRMS
P-values were Bonferroni adjusted to account for multiple outcome variables.
Pairwise comparisons were used to inform functions for canonical correlation analysis, with the
more liberal p<0.05 utilized for identification of potentially important contributors. Canonical
correlation analysis was used to further characterize the relationship between gait variability
measures and neuropsychological test performance.
5.3 Results
Individuals with MS were included in the study if they were between 30-59 years of age,
carried a diagnosis of RRMS, were physically inactive (<2 hours of exercise/week) and had an
Expanded Disability Status Scale (EDSS) score between 1.0 and 5.5 (fully ambulatory without a
device). The EDSS was assessed by a trained neurologist. Individuals with RRMS [mean age= 44.0
± 9.4 SD (range 31-56), 11 females, EDSS mean= 2.8 ± 1.0 SD (range 1.5-4)], and age and gender
matched healthy controls [mean age=43.7± 9.3 SD (range 31-55), 11 females] participated in this
107
study. One participant with MS was unable to complete the BWC condition due to fatigue, thus
10 pairs were assessed for BWC.
5.3.1 Comparison of Backward Walking between Individuals with MS and Healthy Controls
There were no significant differences in spatiotemporal measures of BW between MS
and healthy control groups. Individuals with MS trended toward reduced stride length (p=0.05;
significance achieved at 0.01 after Bonferroni correction) (Table 5.1).
5.3.2 Comparison of Dual-Task Walking between Individuals with MS and Healthy Controls
There were no significant differences in spatiotemporal measures of dual-task walking in
either forward (FWC) or backward (BWC) directions (Table 5.1). Individuals with MS trended
toward reduced velocity (p=0.05) and stride length (p=0.03) and increased double support time
variability (p=0.05) in FWC. These findings were not seen in BWC; however, consistent with our
previous work in BW,24 the switch from FWC to BWC resulted in greater change scores in MS for
both velocity (0.41m/s vs. 0.38m/s) and CV double support time (-3.7% vs. -2.8%) than in
healthy controls, though these were not significantly different. (Table 5.1). Similarly, there were
no significant differences in the change from FW to FWC or BW to BWC between MS and
healthy controls.
5.3.3 Comparison of Forward Walking between Individuals with MS and Healthy Controls
When compared to age and gender matched healthy adults, individuals with MS demonstrated
significantly greater double support time variability (p=0.001) in FW (Table 5.1). When switching
from FW to BW individuals with MS demonstrated significantly greater changes in velocity
(change score 0.48m/s) compared to healthy controls (change score 0.29m/s); p=0.01. (Table
5.2) BW and dual-task walking (i.e., FWC, BWC) also amplified gait variability over forward
108
walking in MS to a greater degree than healthy controls (see Table 5.1 and 5.2 for values used
for change scores). Although not statistically significant, the greatest discrepancy was seen in
the switch from FW to BWC with an increase in variability for individuals with MS of 7-8%
compared to 3-5% for healthy controls.
In summary, when compared to healthy controls, individuals with MS had significantly
increased double support time CV in forward walking and switching from FW to BW resulted in
significantly greater reduction in velocity in individuals with MS. Surprisingly, individuals with MS
were not significantly different from healthy controls in any spatiotemporal measures of BW or
dual-task walking.
109
FW
N=11
1.27
(0.32)
122.1
(17.6)
Multiple Sclerosis
FWC
BW
N=11
N=11
1.09
.78
(0.27)
(0.30)
118.0
88.3
(17.3)
(15.4)
BWC
N=10
.68
(0.26)
90.6
(18.8)
FW
N=11
1.31
(0.24)
129.1
(13.6)
Control
FWC
BW
N=11
N=11
1.30
1.02
(0.35)
(0.21)
132.3
102.0
(13.7)
(12.6)
BWC
N=10
.92
(0.24)
98.9
(12.7)
Velocity
(m/s)
Stride
Length
(cm)
Double
30.9
31.6
37.1
39.3
30.4
29.5
32.2
33.6
Support
(6.3)
(6.3)
(11.3)
(14.9)
(3.7)
(3.7)
(3.6)
(5.0)
Percent
CV of
4.97
7.76
8.82
13.2
3.54
4.78
5.79
6.7
Step Time
(2.0)
(4.4)
(7.8)
(12.5)
(.98)
(3.4)
(1.3)
(2.1)
(%)
CV of
7.80
11.1
11.5
14.8
4.84
6.81
10.3
9.61
Double
(2.0)*
(4.4)
(5.4)
(9.5)
(1.3)
(2.2)
(3.5)
(3.4)
Support
Time (%)
Table 5.1. Spatiotemporal measures and CVs of forward and backward walking and forward and
backward walking with a cognitive task All values are mean (SD). P-value is significant at p≤0.01
after Bonferroni adjustment (paired t-test for each type of walking).
* significantly different from control at p<0.01
CV: Coefficient of Variation; FW: forward walking; FWC: forward walking with cognitive task;
BW: backward walking; BWC: backward walking with a cognitive task
Multiple Sclerosis
Change Score FW –
BW
0.48 (0.19)
33.8 (9.6)
-6.3 (7.3)
Control
Change Score FW –
BW
0.29 (0.07)
27.2 (8.7)
-1.8 (2.4)
p-value
Velocity (m/s)
0.01*
Stride Length (cm)
0.09
Double Support
0.08
Percent
CV of Step Time (%)
-3.8 (7.3)
-2.3 (1.8)
0.51
CV of Double Support
-3.7 (5.9)
-5.5 (3.2)
0.40
Time (%)
Table 5.2. Change scores for forward – backward walking in MS and healthy controls. All values
are mean (SD). P-value is significant at p≤0.01 after Bonferroni adjustment.
* significantly different from control
CV: Coefficient of Variation; FW: forward walking; BW: backward walking
110
5.3.4 Comparison of Neuropsychological Test Performance between Individuals with MS and
Healthy Controls
Individuals with MS demonstrated significantly worse accuracy on the incongruent
condition of the Stroop than healthy controls (p=0.001) (Table 5.3). There was a trend toward
reduced performance on the SDMT and accuracy on the neutral and congruent conditions of the
Stroop (p<0.03). Performance on the word generation task was not significantly different
between groups.
Multiple
Sclerosis
Control
p-value
Stroop Accuracy
Neutral 0.966 (0.040)
0.999 (0.004)
0.020
Congruent
0.971 (0.035)
1.0 (0.000)
0.021
Incongruent
0.960 (0.020)
0.996 (0.014)
0.001*
Symbol Digit
48.6 (15.6)
57.9 (7.6)
0.024
Modality Test
Word Generation
33.2 ( 12.2)
37.3 (10.2)
0.370
Table 5.3. Neuropsychological test performance in MS and healthy controls
All values are mean (SD). P-value is significant at p≤0.01 after Bonferroni adjustment.
* significantly different from control at p<0.01
5.3.5 Relationship of Variability of Backward and Dual-Task Walking to Neuropsychological Test
Performance
In healthy adults, there were no significant correlations between backward or dual-task
walking variability and neuropsychological test performance with one exception. Word
generation was significantly positively correlated with FWC double support time CV (ρ=0.856;
p=0.001).
111
In individuals with MS, performance on neuropsychological tests measuring processing
speed and attention (i.e. SDMT and Stroop), was strongly negatively correlated with gait
variability during dual-tasks (FWC and BWC) (Table 5.4). Similar to healthy controls, the word
generation task was correlated only with double support time variability in FWC. The
incongruent condition of the Stroop was not correlated with any of the gait variability measures
(Table 5.3), despite it being the only measure that was significantly different between healthy
controls and individuals with MS (Table 5.3). To better explore the relationship between dualtask gait variability and neuropsychological test performance in RRMS, we utilized canonical
correlation analysis, which manages the association between sets of multiple dependent and
independent variables. Based on trending and significance in the pairwise comparisons, we used
a “set” of gait variables (BWC and FWC step length CV, BWC and FWC step time CV, FWC stride
length CV, and BWC and FWC double support CV) and a “set” of neuropsychological test
measures (SDMT and Stroop accuracy on congruent and neutral conditions). The analysis is
achieved by developing a number of canonical functions that maximize the correlation between
the sets of dependent and independent variables. Each canonical function is based on the
correlation between two canonical variates, one each for the dependent and independent
variables. The number of canonical functions created from the sets of variables is equivalent to
the number of variables in the smallest data set (in this case, the three neuropsychological test
variables). The first canonical function is derived to have the highest intercorrelation possible
between the two sets of variables. Successive functions are based on the residual variance; thus
the correlations become smaller with each additional function, and the functions are
independent of each other.36 In this analysis, function one achieved significance (Table 5.5).
Cross loadings36 show that while measures of processing speed do not appear to be associated
112
with a common function, backward dual-task variability is most closely associated with function
one. Indeed, 48% of the variance in BWC step length, 62% of the variance in BWC step time and
65% of the variance in BWC double support time are explained by the neuropsychological test
measures. Similarly, 79% of the variance in the Stroop Neutral condition is explained by the
dual-task gait variability measures. BWC gait variability measures carry a direct relationship with
function one while the Stroop Neutral carries an inverse relationship. This can be interpreted
similarly to the pairwise comparison; poor performance on the Stroop Neutral is associated with
increased variability in BWC gait measures.
113
SDMT
Stroop
Congruent
Stroop
Neutral
Stroop
Incongruent
Word
Generation
Falls in Past 6
Months
BW
114
n=11
Step length CV -0.564 (0.071)
-0.449 (0.166)
-0.474 (0.140)
-0.498 (0.119) -0.246 (0.466)
0.538 (0.088)
Step time CV -0.345 (0.298)
-0.206 (0.544)
-0.545 (0.083)
-0.112 (0.744) -0.597 (0.053)
0.478 (0.137)
Stride length CV -0.464 (0.151)
-0.510 (0.109)
-0.315 (0.346)
-0.633 (0.037) -0.269 (0.424)
0.239 (0.479)
Double Support CV -0.718 (0.013)*
-0.416 (0.203)
-0.582 (0.060)
-0.228 (0.500) -0.551 (0.079)
0.179 (0.598)
BWC
n=10
Step length CV -0.455 (0.187)
-0.722 (0.018)
-0.401 0(.250)
-0.729 (0.017) -0.140 (0.700)
0.494 (0.147)
Step time CV -0.612 (0.060)
-0.821 (0.004)* -0.596 (0.069)
-0.642 (0.045) -0.492 (0.148)
0.494 (0.147)
Stride length CV -0.345 (0.328)
-0.525 (0.119)
-0.320 (0.367)
-0.430 (0.215) -0.103 (0.776)
0.418 (0.230)
Double Support CV -0.758 (0.011)*
-0.778 (0.008)* -0.803 (0.005)* -0.405 (0.245) -0.456 (0.185)
0.570 (0.086)
FWC
n=11
Step length CV -0.609 (0.047)
-0.369 (0.264)
-0.197 (0.561)
-0.051 (0.881) -0.232 (0.492) -0.120 (0.726)
Step time CV -0.682 (0.021)
-0.860 (0.001)* -0.479 (0.136)
-0.428 (0.189) -0.401 (0.222)
0.299 (0.372)
Stride length CV -0.809 (0.003)*
-0.711 (0.014)
-0.446 (0.169)
-0.177 (0.603) -0.478 (0.137)
0.000 (1.00)
Double Support CV -0.764 (0.006)*
-0.767 (0.006)* -0.789 (0.005)* -0.079 (0.817) -0.752 (0.008)* 0.359 (0.279)
Table 5.4. Pairwise comparisons of variability of backward and dual-task walking to neuropsychological test performance and falls
in MS
Spearman’s Rho; all values are ρ (p- value). *p-value <0.013 significant after Bonferroni correction (Each neuropsychological test
explored across CVs for each type of walking)
Function
Canonical Correlation
Wilk’s
Chi-Square
DF
Significance
1
1.000
0.000
0.000
21
0.000
2
0.981
0.005
18.715
12
0.096
3
0.935
0.126
7.261
5
0.202
Table 5.5. Canonical correlations to examine the relationship between gait variability and neuropsychological test performance in MS
Sets used to create the functions: set 1=dual- task variability measures vs. set 2=neuropsychological test performance measures
114
5.3.6 Relationship between Dual-Task Questionnaire and Gait Variability and Neuropsychological
Test Performance
There were no significant correlations between the DTQ and performance on any of the
neuropsychological tests (p-value <0.01 required for significance after correction) or with any of
the gait variability measures in BW, BWC or FWC in healthy adults (p-value <0.013 required for
significance after correction). However, in individuals with MS, the DTQ was significantly
inversely correlated with performance on the Stroop neutral condition (ρ=-0.803, p=0.003) and
trended toward significance on the SDMT (ρ=-0.706, p=0.015). DTQ was also positively
correlated with double support time variability in both BW and BWC in individuals with MS
(ρ=0.843, p=0.001 and ρ=0.754, p=0.012 respectively). That is, higher scores on the DTQ (more
subjective difficulty with dual-tasks) was associated with poorer performance on
neuropsychological tests and higher gait variability in both BW and BWC.
5.3.7 Relationship of Dual-Task Cost and Neuropsychological Test Performance
There were no significant correlations between any of these DTCs and
neuropsychological test performance in healthy adults or individuals with RRMS. Values for
RRMS are reported in Table 5.6.
5.3.8 Relationship of Falls with Variability of Backward and Dual-Task Walking and
Neuropsychological Test Performance
Individuals reported falls in the past week, 8 weeks and 6 months. None of the healthy
controls reported falls. Only one subject with MS reported a fall in the past week, while two
subjects reported falls in the past 8 weeks. We opted to assess the relationship of falls over the
past 6 months with gait variability since four of eleven individuals with MS reported falls in the
past 6 months. Individuals were dichotomized; those who reported 1 or more falls in the past 6
115
months were considered “fallers” while those with zero falls in the past 6 months were “nonfallers”. There was no significant relationship between falling in the past 6 months and gait
variability. (Table 5.4). However, falls in the past 6 months were trending toward an inverse
correlation with performance on the Stroop Neutral condition (ρ=-0.679; p=0.021) (Table 5.6),
which was also significantly related to double support time gait variability in BWC and FWC
(Table 5.4).
Falls Past 6
Months
DTC FWFWC
Stroop
Accuracy
Neutral
-0.679 (0.021)
Stroop
Accuracy
Congruent
-0.307 (0.358)
Stroop
Accuracy
Incongruent
-0.122 (0.720)
Symbol Digit
Modality Test
Word
Generation
-0.239 (0.479)
-0.479 (0.136)
-0.085 (0.805)
-0.280 (0.403)
-0.033 (0.924)
-0.145 (0.670)
-0.087 (0.800)
-0.082(0.823)
0.191 (0.596)
-0.137 (0.706)
0.018 (0.960)
0.146 (0.688)
DTC BWBWC
Table 5.6. Relationship of falls in the past 6 months and dual-task cost to neuropsychological
test performance in RRMS
Spearman’s Rho; all values are ρ (p-value). P-value <0.01 considered significant with Bonferroni
adjustment.
5.4 Discussion
This is the first study to formally assess backward walking and backward dual-task
walking in MS. Additionally, this study expands and strengthens the current knowledge base
exploring baseline differences in walking and cognition in individuals with MS and healthy
controls; rather than a convenience sample, a rigorous matching process was utilized to reduce
variability and eliminate bias. Gait variability in ambulatory individuals with MS is associated
116
with performance on neuropsychological tests of attention and processing speed. Consistent
with previous literature in MS,15 dual-task walking amplified variability of spatiotemporal
measures of gait. BW also resulted in an increased amplification of gait variability in MS, which
has been previously reported only in PD.23 This highlights the importance of cognitive function
in mobility and suggests a potential role for cognitive training in exercise prescription.
We hypothesized that individuals with MS would demonstrate significantly greater
variability and slower velocity during backward walking and backward walking with a cognitive
task compared to healthy control subjects. We found a significant difference in the change
score for FW – BW for velocity; individuals with MS demonstrated a significantly larger change in
velocity than their healthy counterparts. Otherwise, there were no significant differences
between groups for BW and BWC velocity or gait variability. To further explore gait variability,
we examined change scores within group between FW and both BW and DT walking conditions
(FWC and BWC). Both conditions of DT walking as well as BW amplified gait variability in MS to a
greater degree than healthy controls; however, these values were not statistically significant.
The lack of significance is perhaps due in part to large standard deviations in the MS group.
Additionally, we were powered to detect differences in forward walking, but may be
underpowered to detect differences in backward walking between groups. We only observed a
significant difference in FW double support time variability between MS and healthy adults. This
was especially interesting since our sample size calculation was based on FW gait speed for
individuals with RRMS that had similar disability levels to those in our study and were measured
in the same way (i.e., GAITRite).6 To examine the sample size required to achieve significance in
dual-task and BW velocity, we performed a post-hoc power analysis. For velocity in all three
conditions (BW,FWC, BWC), we were underpowered (power: 28.1%, 13.0%, 26.6%; effect size:
117
0.93, 0.67, 0.96 respectively) to detect changes between individuals with MS and healthy
controls. To achieve 80% power, a sample size of 29, 54, and 28 per group would be required
with a conservative two sample t-test and p=0.01 for BW, FWC and BWC respectively.
We further hypothesized that variability of backward walking measured with CV would
correlate with neuropsychological test performance. The results confirmed our hypothesis;
accuracy on the Stroop neutral and congruent conditions was strongly associated with BWC and
FWC variability (Table 5.4). Performance on the SDMT was inversely correlated with gait
variability in BW, BWC and FWC (Table 5.4). Our results suggest that measures of processing
speed (i.e., SDMT, Stroop) are more closely associated with BW and dual-task gait variability in
individuals with MS than other measures such as fluency. This finding is consistent with studies
in older adults37 and individuals with Alzheimer’s disease38 demonstrating that gait variability
was highly correlated with measures of processing speed.
This is the first study to apply the subjective DTQ to individuals with MS. We noted
significant correlations with accuracy on the Stroop neutral condition as well as variability of BW
and BWC. This suggests that individuals with MS may have insight into their deficits with dualtasking and warrants further exploration of the utility of the DTQ as a clinical test of gait
variability and dual-task deficits in individuals with MS.
Change scores reveal a larger change in velocity when switching from FW to FWC for MS
(0.18 m/s) compared to healthy controls (0.01 m/s) while other spatiotemporal parameters
remain relatively stable (Table 5.1). The switch from BW to BWC resulted in similar velocity
declines (0.1m/s) in both groups; however, the switch did result in large increases in variability
in the MS group while the variability remained stable in the healthy control group. The results
for the FW to FWC switch are consistent with previous studies in MS, where gait speed was
118
shown to be the variable most affected by dual-task conditions.15 The degree of decrement
resulting from dual-tasking in MS was not related to disease severity or duration, as might be
expected, but rather to levels of fatigue and measures of general cognitive functioning.15 This is
in contrast to individuals with PD,39 older adult fallers40 and individuals with Alzheimer’s
disease,41 where the effect of dual-tasking on gait speed is similar to the decrement seen in
healthy controls.
Poor performance on neuropsychological tests was significantly associated with
increased gait variability during dual-task walking. Our work is supported by previous work in
traumatic brain injury, which has suggested a role for cognition in dual-tasking.42 Indeed,
individuals with TBI with greater deficits in psychosocial functioning performed worse on
neuropsychological tests measuring dual-task ability than those with lesser deficits.42 In elderly
adults, increasing the complexity of the cognitive task paired with gait resulted in decreasing
gait speeds,28 and regression analysis indicated that underlying cognitive abilities contributed
significantly as the complexity of the secondary task increased.28 Indeed, gait variability during
dual-task walking has been significantly correlated with performance on neuropsychological
tests of executive function, attention and processing speed in individuals with Alzheimer’s
disease.38 Impairments in attention and executive function may prevent individuals with MS
from allotting appropriate attentional resources to balance and gait, reduce their ability to
confront and adapt to challenging environments, contribute to increased gait variability, and
may be linked to increased fall risk, which is also seen in the elderly,40,43 PD39 and Alzheimer’s
disease.41
Despite our a priori sample size calculation, the present study appears to be limited by
small sample size; however, we were still able to demonstrate robust relationships between gait
119
variability and neuropsychological test performance in RRMS. Indeed, canonical correlation
analysis demonstrates that 79% of the variance in performance on the Stroop neutral condition
is explained by dual-task gait variability, while up to 65% of the variance in gait variability
measures is explained by performance on neuropsychological tests. This study only examined
one type of cognitive dual-task (serial 3 subtraction); it is possible that a fluency dual-task would
be more related with word generation test performance. Although we reported no relationship
between falls in the past 6 months and gait variability, this is likely because the majority of our
subjects were not frequent fallers. Indeed seven of eleven subjects reported no falls in the past
6 months. We anticipated that falls would be correlated with gait variability in BW and BWC as
both our previous work and the work of others has suggested that BW velocity may be a
sensitive indicator of falls in elderly adults24,44 and that BW and dual-tasks increase gait
variability in individuals with neurologic conditions.23,45-46 It appears that more research is
necessary to determine if tasks that amplify gait variability may be useful in identifying deficits
and assessing fall risk in individuals with MS.
This study highlights the importance of examining gait variability in MS; clinically useful
activities such as BW or dual-task walking that amplify gait variability may identify early deficits
in mobility or cognition that could contribute to fall risk. Gait variability can be challenging to
objectively examine clinically without expensive equipment such as 3D motion capture or
GAITRite. Strong correlations between gait variability, neuropsychological test performance and
the DTQ suggest that increased gait variability may be related to cognitive deficits and perceived
dual-task ability in MS. This is the first study to provide evidence for several clinically useful tests
that correlate with gait variability for examining deficits in individuals with MS.
120
This study further clarifies the important contribution of baseline cognition to gait
deficits and the potential for therapy utilizing cognitive training in addition to mobility training.
While there has been limited research to suggest benefits from dual-task training in individuals
with neurologic diseases,29,47-49 there have been, to date, no intervention studies exploring dualtask training in MS. Future research should further elucidate the role of early cognitive deficits in
MS as potential contributor to mobility deficits. This should include larger sample sizes and the
examination of additional cognitive tasks, such as fluency, during gait. However, at present, we
endorse that the clinical evaluation of gait should go beyond forward walking assessment to
include functional locomotor tasks that challenge dynamic balance and amplify gait variability,
such as dual-task walking and backward walking.
121
5.5 References
1. Frohman, E.M. Multiple sclerosis. Med Clin N Am. 2003; 87(4): 867–897.
2. Amato MP, Portaccio E, Goretti B, Zipoli V, Hakiki B, Giannini M, Pasto L, Razzolini L.
Cognitive impairment in early stages of multiple sclerosis. J Neurol Sci. 2010; Suppl 2;
S211-S214.
3. Noseworthy JH, Lucchinetti C, Rodriguez M, Weinshenker BG. Medical progress:
multiple sclerosis. New Engl J Med. 2000;343:938–952.
4. Al-Omaishi J, Bashir R, Gendelman HE. The cellular immunology of multiple sclerosis. J
Leukocyte Biol. 1999; 65:444–452.
5. Cattaneo D, De NC, Fascia T, Macalli M, Pisoni I, Cardini R. Risks of falls in subjects with
multiple sclerosis. Arch Phys Med Rehab. 2002; 83(6): 864-867.
6. Givon U, Zeilig G, Achiron A. Gait analysis in multiple sclerosis: characterization of
temporal-spatial parameters using GAITRite functional ambulation system. Gait Posture.
2009; 29: 138-142.
7. Crenshaw SJ, Royer TD, Richards JG, Hudson DJ. Gait variability in people with multiple
sclerosis. Mult Scler. 2006; 12: 613-619.
8. Hausdorff JM, Edelberg HK, Mitchell SL et al. Increased gait unsteadiness in communitydwelling elderly fallers. Arch Phys Med Rehab. 1997; 78: 278-283.
9. Maki BE. Gait variability in older adults: predictors of falls or indicators of fear. J Am
Geriatr Soc. 1997; 45: 313-320.
10. Hausdorff JM, Rios D, Edelberg HK. Gait variability and fall risk in community-living older
adults: a 1 year prospective study. Arch Phys Med Rehab. 2001; 82: 1050-1056.
11. Schaafsma JD, Giladi N, Balash Y et al. Gait dynamics in Parkinson‘s disease: relationship
to Parkinsonian features, falls and response to levadopa. J Neurol Sci. 2003; 212: 47-53.
12. Nakamura T, Megura K, Sasaki H. Relationship between falls and stride length variability
in senile dementia of the Alzheimer type. Gerontology. 1996; 42: 108-113.
13. Deloire MS, Salort E, Bonnet M et al. Cognitive impairment as a marker of diffuse brain
abnormalities in early relapsing remitting multiple sclerosis. J Neurol Neurosur Ps. 2005;
76: 519-526.
14. Rao SM, Leo GJ, Bernardin L, Unverzagt F. Cognitive dysfunction in multiple sclerosis I:
frequency, patterns, and prediction. Neurology. 1991; 41(5): 685-691.
122
15. Hamilton F, Rochester L, Paul L et al. Walking and talking: an investigation of cognitivemotor dual tasking in multiple sclerosis. Mult Scler. 2009; 15(10): 1215-1227.
16. Kalron A, Dvir Z, Achiron A. Walking while talking—difficulties incurred during the initial
stages of multiple sclerosis disease process. Gait Posture. 2010; 32(3) 332-335.
17. Finlayson ML, Peterson EW, Cho CC. Risk factors for falling among people aged 45 to 90
years with multiple sclerosis. Arch Phys Med Rehab. 2006; 87(9): 1274-1279.
18. D‘Esposito M, Onishi K, Thompson H et al. Working memory impairments in multiple
sclerosis: evidence from a dual-task paradigm. Neuropsychology. 1996; 10(1): 51-56.
19. Nilsagard Y, Denison E, Gunnarsson LG, Bostrom K. Factors perceived as being related to
accidental falls in persons with multiple sclerosis. Disabil Rehabil. 2009; 31: 1301-1310.
20. Cameron MH, Horak FB, Herndon RR, Bourdett D. Imbalance in multiple sclerosis: a
result of slowed spinal somatosensory conduction. Somatosens Mot Res. 2008; 25: 113122.
21. Cameron MH, Lord S. Postural control in multiple sclerosis: implications for fall
prevention. Curr Neurol Neurosci. 2010; 10: 407-412.
22. Thomas MA, Fast A. One step forwards and two steps back: the dangers of walking
backward in therapy. Am J Phys Med Rehab. 2000; 79(5): 459-461.
23. Hackney ME, Earhart GM. Backward walking in Parkinson disease. Movement Disord.
2009; 24(2): 218-223.
24. Fritz NE, Worstell A, Siles A, Kloos AD, Kegelmeyer DA. Backward walking parameters are
sensitive to age-related changes in mobility and balance. Gait Posture. 2013; 37: 593597
25. Martin Cl, Phillips BA, Kilpatric TJ, Butzkeuven H, Tubridy N, McDonald E, Galea MP. Gait
and balance impairment in early multiple sclerosis in the absence of clinical disability.
Mult Scler. 2006; 12: 620-628.
26. Tseng IJ, Jeng C, Yuan RY. Comparisons of forward and backward gait between poorer
and better attention capabilities in early Parkinson’s disease. Gait Posture. 2012; 36(3):
367-371.
27. Venema DM, Bartels E, Siu KC. Tasks matter: a cross-sectional study of the relationship
of cognition and dual-task performance in older adults. J Geriatr Phys Ther. 2012; epub
ahead of print. http://www.ncbi.nlm.nih.gov/pubmed/23249724
123
28. Hall CD, Echt KV, Wolf SL, Rogers WA. Cognitive and motor mechanisms underlying older
adults’ ability to divide attention while walking. Phys Ther. 2011; 91(7): 1039-1050.
29. Evans JJ, Greenfield E, Wilson BA, Bateman A. Walking and talking therapy: improving
cognitive-motor dual-tasking in neurological illness. J Int Neuropsych Soc. 2009; 15: 112120.
30. Parmenter BA, Weinstock-Guttman B, Garg N, Munschauer F, Benedict RHB. Screening
for cognitive impairment in multiple sclerosis using the symbol digit modality test. Mult
Scler. 2007; 13: 52-57.
31. Deloire MS, Bonnet MC, Salort E, et al. How to detect cognitive dysfunction at early
stages of multiple sclerosis? Mult Scler. 2006; 12: 445-452.
32. Genova HM, Hillary FG, Wylie G, Rypma B, DeLuca J. Examination of processing speed
deficits in multiple sclerosis using functional magnetic imaging. J Int Neuropsych Soc.
2009; 15: 383-393.
33. Drake AS, Weinstock-Guttman B, Morrow SA, Hojnacki D, Munschauer FE, Benedict RHB.
Psychometrics and normative data for the Multiple Sclerosis Functional Composite:
replacing the PASAT with the symbol digit modalities test. Mult Scler. 2010; 16(2): 228237.
34. Rao SM. Cognitive function in patients with multiple sclerosis: impairment and
treatment. Int J Mult Scler Care. 2004; 1: 9-22.
35. Strauss E, Sherman EMS, Spreen O. A compendium of neuropsychological tests:
administration, norms, and commentary. 3rd ed. New York: Oxford University Press;
2006.
36. Hair JF, Anderson RE, Tatham RL, Black WC. Multivariate Data Analysis. 5th ed. New
York, NY: Prentice Hall; 1998.
37. Martin KL, Blizzard L, Wood AG, Srikanth V, Thomson R, Sanders LM, Callisaya ML.
Cognitive function, gait and gait variability in older people: a population-based study. J
Gerentol A Biol Sci Med Sci. 2013; 68(6): 726-732.
38. Sheridan PL, Solomont J, Kowall N, Hausdorff JM. Influence of executive function on
locomotor function: divided attention increases gait variability in Alzheimer’s disease. J
Am Geriatr Soc. 2003; 51: 1633-1637.
39. Yogev G, Giladi N, Peretz, Springer S, Simon ES, Hausdorff JM. Dual-tasking, gait
rhythmicity, and Parkinson’s disease: which aspects of gait are attention demanding. Eur
J Neurosci.2005; 22: 1248-1256.
124
40. Springer S, Giladi N, Peretz C, Yogev G, Simon ES, Hausdorff JM. Dual-tasking effects on
gait variability: the role of aging, falls and executive function. Movement Disord. 2006;
21(7): 950-957.
41. Cocchini C, Della Sala S, Logie RH, Pagani R, Sacco L, Spinnler H. Dual task effects of
walking when talking in Alzheimer’s disease. Revue Neurologique (Paris). 2004; 160(1):
74-80.
42. Foley JA, Cantagallo A, Della Sala S, Logie RH. Dual task performance and post traumatic
brain injury. Brain Injury. 2010; 24(6): 851-858.
43. 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.
44. Beauchet O, Annweiler C, Allali G, Berrut G, Herrmann FR, Dubost V. Recurrent falls and
dual task-related decrease in walking speed: is there a relationship? J Am Geriatr Soc.
2008; 56(7):1265-1269.
45. Yogev G, Plotnik M, Peretz C, Giladi N, Hausdorff JM. Gait asymmetry in patients with
Parkinson’s disease and elderly fallers: when does the bilateral coordination of gait
require attention? Exp Brain Res. 2007; 177: 336-346.
46. Hackney ME, Earhart GM. The effects of a secondary task on forward and backward
walking in Parkinson disease. Neurorehab Neural Re. 2009; available online at:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2888719/pdf/nihms158841.pdf.
47. Schwenk M, Zieschang T, Oster P, Hauer K. Dual-task performances can be improved in
patients with dementia: a randomized controlled trial. Neurology. 2010; 74: 1961-1968
48. Yen CY, Lin KH, Hu MH, Wu RM, Lu TW, Lin CH. Effects of virtual reality-augmented
balance training on sensory organization and attentional demand for postural control in
people with Parkinson disease: a randomized controlled trial. PhysTher. 2011; 91: 862874.
49. Brauer SG, Morris ME. Can people with Parkinson’s disease improve dual tasking when
walking? Gait Posture. 2010; 31: 229-233.
125
Chapter 6: SMA Connectivity May Be Associated With Dual-Task Walking in Multiple Sclerosis
6.1 Introduction
Multiple Sclerosis (MS) is a chronic, autoimmune disease of the central nervous system
affecting over 400,000 Americans with an annual health care cost in the billions.1 MS is the most
common progressive neurological disease2 in young adults. Individuals with MS often experience
declines in walking ability3 and cognitive dysfunction,4 particularly with attention, processing
speed, and memory.5 Indeed, 22-25% of individuals with MS report specific attentional
problems.6 The simplest form of attention, attention span, is generally intact, while the more
complex aspects, such as selective, divided, or alternating attention, are most often impaired.7
Divided attention is the ability to respond to multiple stimuli simultaneously and can encompass
both motor-motor and motor-cognitive tasks. Individuals with MS have difficulty with divided
attention during gait; not surprisingly, cognitive dysfunction has been identified as a risk factor
for accidental falls in MS.8
Magnetic resonance imaging (MRI) is the most commonly used paraclinical assessment
tool, both for investigation of underlying pathology and evaluation of disease progression in
MS.9 Diffusion tensor imaging (DTI) is a technique that quantifies the amount of non-random
water diffusion within tissues to provide an in vivo model of white matter integrity and
126
microstructural damage.10 The diffusion of molecular water in grey matter is highly isotropic
(i.e., the same in all directions), while in white matter, diffusion is often
restricted to a particular direction. This anisotropy is dictated by axonal cell membranes, myelin
sheaths, neurofilaments, and other structures11 with the direction of highest diffusivity
coinciding with the tissue’s fiber tract axis.12-13 Diffusion tensors are often visualized as ellipsoids
with the size and shape reflecting the degree of diffusion along each principle axis. The principle
axes correspond to the eigenvectors of the tensor (e1, e2 and e3) and the relative size of each
axis is determined by the eigenvalues of the tensor (1, 2 and 3)14(Figure 6.1).
e2
2
e3
3
e1
1
Figure 6.1. Elliptical representation of a diffusion tensor; the principle
axes correspond to the eigenvectors of the tensor (e1, e2 and e3) and
the relative size of each axis determined by the eigenvalues of the
tensor (1, 2 and 3).14
Anisotropy is highest in compact white matter pathways where the fiber bundles are
oriented in parallel, such as the corpus callosum or pyramidal tracts.11 Higher fractional
anisotropy (FA) value and lower mean diffusivity (MD) value indicates a greater degree of white
matter integrity. FA is the most commonly used anisotropy index in the literature and the FA
127
index is normalized to values of 0 (isotropic) to 1 (anisotropic).15 Compared to healthy controls,
individuals with relapsing remitting MS (RRMS) have significantly lower FA values in the
corticospinal tracts,16-17 and corpus callosum18 and these FA values correlate well with both
motor17 and cognitive18-19 impairment. In addition, the mean diffusivity (MD), axial diffusivity
(AD) and radial diffusivity (RD) provide useful information about the tract or white matter region
of interest. The AD is equivalent to the primary eigenvalue (1), indicates the direction of
highest diffusivity and coincides with the fiber tract axis.20-21 The RD is calculated by averaging
the mean of the other two eigenvalues (2 and 3). Animal work suggests that the AD may be
specific to axonal degeneration while the RD may be modulated by myelin.20 The MD is the
average of all three eigenvalues, and is therefore not indicative of direction. The MD is inversely
related to the FA. DTI studies in MS have demonstrated decreased FA and increased MD in
regions that appear normal on conventional MRI 21-28 suggesting damage beyond the resolution
of conventional MRI.10
Diffusion images can also be used to perform tractography, the only available tool for
identifying and measuring white matter pathways non-invasively and in vivo.29 Tractography
identifies the path along which diffusion is least hindered and thus allows for the exploration of
connectivity and for the measurement of diffusivity along a tract of interest . MS lesions can
result in a loss of connectivity, a term that describes the physical connections linking brain
regions29 but no studies have examined the relationship between clinical tests of dual-task and
neural connectivity. The supplemental motor area (SMA) is active during complex motor tasks,30
cognitive dual-tasks,31 and tasks requiring interlimb coordination,32 but its relationship to clinical
measures of motor-cognitive dual-task has not been explored. The objective of this study was to
examine the relationship between clinical measures of dual-task performance (Walking While
128
Talking Test, TUG Cognitive, walking with a subtraction task) and measures of diffusivity (FA,
MD, RD, AD) in the interhemispheric SMA tract. Although this has not been previously explored,
we hypothesized that SMA interhemispheric connectivity would correlate with a clinical
measure of dual-task ability. This is the first study to explore lower extremity motor-cognitive
dual-tasks in relationship to imaging metrics in MS and the first to examine the relationship
between the SMA interhemispheric tract and clinically useful motor measures in individuals with
MS.
6.2. Methods
This study uses preliminary baseline data from an ongoing randomized controlled trial
examining the effects of Dance Dance Revolution on mobility, brain plasticity and cognition in
individuals with Multiple Sclerosis. This trial was approved by the Institutional Review Board. All
participants signed consent forms prior to participation.
6.2.1 Participants
Eleven volunteers with RRMS participated in this study. All subjects were screened for
any contraindications for MRI participation. All eligible subjects participated in a mobility
assessment and a multi-modal 3T MRI. This study was approved by the Institutional Review
Board. All participants signed consent forms prior to participation.
6.2.2 Mobility Measures
Mobility measures included walking (forward and backward dual-task walking (walk
with subtraction)), the Walking While Talking Test, and the TUG Cognitive. All participants
ambulated across the GAITRite System (V3.9, MAP/CIR Inc.), which records spatiotemporal
measures (e.g., velocity, step length, double support) of gait and is reliable and valid for use in
129
individuals with MS.33 Individuals ambulated across the GAITRite for four trials of each of four
conditions: a) forward walking comfortable speed (FW); b) forward walking with a cognitive task
of serial 3 subtraction starting at 97 (FWC); c) backward walking comfortable speed (BW); and d)
backward walking with a cognitive task of serial 3 subtraction starting at 95 (BWC). Participants
were instructed to walk 2 meters before and after the walkway to allow for acceleration and
deceleration. Participants completed all trials without an assistive device or bracing, wearing a
gait belt and guarded by a physical therapist. The final three trials for each condition were
averaged and the spatiotemporal parameters of gait were exported for analysis and used to
calculate coefficients of variation (CV).
The Walking While Talking Test (WWTT) is a reliable and valid test to identify older
individuals at high risk for falls.34 The test is performed and timed under three conditions: 1)
walk 40 feet with a 180: turn at the midpoint; 2) condition 1 + recite the alphabet aloud (WWTsimple); 3) condition 1 + recite alternate letters of the alphabet aloud (WWT-complex).34 Poor
performance on the WWT-simple and complex was strongly predictive of falls; a cut-off of 33
seconds for the WWT-complex predicted elderly fallers with a specificity and sensitivity of 95.6%
and 38.5% respectively.
The Timed Up and Go (TUG) is a common test to examine mobility in elderly.35 Time to
complete the TUG, which requires standing from a chair, walking 10 feet, turning, walking back
and sitting down, is correlated with level of mobility35 in elderly adults. The TUG has been
utilized in MS populations and correlates with longer walking tests.36 The TUG Cognitive is the
TUG while performing a subtraction task; this modification of the TUG measures dual-task
performance and is sensitive and specific for identifying fall risk in elderly.37 A time of >15
130
seconds to complete the TUG Cognitive accurately identifies elderly fallers with an 86.7%
prediction rate.37 The TUG Cognitive accurately predicts fallers in MS with a sensitivity of 73%.38
6.2.3 MRI Acquisition
Participants were scanned in a Siemens 3T Magnetom Trio scanner. High resolution
structural images were collected using a 3D magnetization prepared rapid gradient echo
imaging (MPRAGE) protocol with 160 contiguous axial slices (TE/TR/TI 4.68/2000/950 ms), with
256mm FOV and a 1.0x1.0mm2 in-plane resolution. To identify white matter lesions, we
acquired 60 T2-weighted fluid attenuated inversion recovery (FLAIR) axial slices (TE/TR
7.3/14000ms) with a 256mm FOV and 0.8x0.8mm2 in-plane resolution.
Diffusion tensor images were acquired in the axial plane (TE/TR 85/8300ms) with
diffusion gradients in 64 directions. The field of view was 288mm resulting in voxel sizes of
2.0x2.0mm2. The b values were 0 and 800s/mm2.
6.2.4. Imaging Analysis
Lesion Load Analysis
T2 FLAIR images were processed using Jim 6.0 (Xinapse Systems, www.xinapse.com), a
semi-automated segmentation tool that localizes white matter hyperintensities. Lesions were
identified manually and used to create a region of interest mask that was applied to a fuzzy
connectivity algorithm,39 thus creating a 3D representation of the identified lesions using the
fuzzy affinity between neighboring voxels. Fuzzy connectedness compares each manually
identified region of interest’s size and location to an MS lesion template, while the probability
weighting takes into account the similarities in intensity values and the degree of adjacency
among voxels.39 We used pre-determined values suggested by the Jim software to set the fuzzy
connectedness threshold at 0.62 and the associated probability weighting at 0.25.
131
Diffusion Tractography Reconstruction
The diffusion-weighted images were preprocessed using FMRIB’s Diffusion Toolbox
(FDT). The raw data were corrected for head movement and eddy current distortion. The tensor
model was then fitted to the preprocessed diffusion images to generate FA and eigenvector 1-3
images. The Harvard-Oxford Cortical Probability Atlas for tractography was used to identify the
SMA ROI. Masks were created to standardize tractography across subjects. Seed and target
masks were created in the sagittal plane to track from the left SMA to the right SMA. Eightyeight voxel seed and target masks were placed 8mm from the midline in the juxtapositional
lobule cortex (Harvard-Oxford Cortical Probability Atlas). Both a 119-voxel mask corresponding
to the middle one-third of the corpus callosum and the subject’s individual white matter mask
were used as waypoint masks. An axial/frontal exclusion mask was applied below the level of
the corpus callosum to improve the sensitivity of the tractography. Finally, a termination mask
was applied 2mm lateral to the target mask, and probabilistic tractography was used to model
the SMA interhemispheric tract.
To determine specificity of the SMA tract’s relationship to dual-task walking, we chose a
control interhemispheric tract that would be unlikely to be involved in dual-tasking but crosses
at the corpus callosum. The Harvard-Oxford Cortical Probability Atlas was used to localize V4.
Forty voxel seed and target masks were created in the sagittal plane in the occipital fusiform
gyrus to track from left V4 to right V4. A 120 voxel mask corresponding to the posterior onethird of the corpus callosum and the subject’s individual white matter mask were used as
waypoints. An axial exclusion mask placed below the supratentorium at the level of the superior
cerebellum was applied. Finally, a termination mask was applied 2mm lateral to the target mask,
and probabilistic tractography was used to model the V4 interhemispheric tract.
132
6.2.5 Statistical Analysis
Statistical analyses were performed using SPSS version 19.0 . We set the family-wise
error rate at 0.05 to control for type I error rate. Bivariate correlations were performed to
determine if a relationship was present between lesion volume, area or number and measures
of dual-task ability (FWC, BWC, WWTT, TUG Cognitive) or the EDSS. Correlational analysis was
used to examine the relationship between:
a) Spatiotemporal measures forward and backward dual-task walking velocity and SMA
interhemispheric tract diffusion parameters (fractional anisotropy (FA) and axial, radial
and mean diffusivity (AD, RD and MD respectively)
b) Clinical measures of dual-task ability (WWTT, TUG Cognitive) and SMA interhemispheric
tract diffusion parameters
c) Both spatiotemporal measures of dual-task walking and clinical measures of dual-task
ability and V4 interhemispheric tract diffusion parameters to determine the specificity
of the SMA tract.
P-values were Bonferroni adjusted to account for multiple outcome variables. Linear regression
was used to identify the predictive value of the variables with the strongest relationship
identified in the correlation analysis described in (a) and (b) above.
6.3 Results
Individuals with MS were included in the study if they were between 30-59 years of age,
had a diagnosis of RRMS, were physically inactive (<2 hours of exercise/week) and had an
Expanded Disability Status Scale (EDSS) score between 1.0 and 5.5 (fully ambulatory without a
133
device). The EDSS was assessed by a trained neurologist. Eleven females with RRMS [mean age=
44.0 ± 9.4 SD (range 31-56) EDSS mean= 2.8 ± 1.0 SD (range 1.5-4)] participated in this study.
6.3.1 Clinical dual-task measures
Clinical measures of dual-task ability assessed were walking velocity during forward
(FWC) and backward walking (BWC) when paired with a secondary cognitive task (serial 3
subtraction), the Walking While Talking Test (WWTT) and the TUG Cognitive (Timed Up and Go
Cognitive). Velocity in FWC was on average 1.09 ± 0.27m/s while BWC velocity was 0.68 ±
0.26m/s. One individual was unable to complete the BWC trials secondary to fatigue, so only 10
individuals were analyzed for this measure. Average time to complete the WWTT was 14.98 ±
3.80s and average time on the TUG Cognitive was 10.24 ± 3.20s.
6.2.2 Lesion Load Analysis
The total lesion volume for all regions: mean ± SD (range) was 7685.4 ± 4905.3 (22.2 –
16365.7)mm3. The total number of lesions per patient in all regions: mean± SD (range) was 19 ±
13 (1 - 43). The lesions were found predominantly in juxtacortical regions 11 ± 11 (0 –33) and
periventricular regions 8 ± 3 (1 –14). Only two participants had evidence of infratentorial lesions.
Total lesion area, volume, and number did not correlate with clinical measures of dual-task
(FWC velocity, BWC velocity, TUG Cognitive, WWTT) (p<0.2 in all cases) nor were they related to
EDSS score (p>0.2 in all cases). A lesion probability map was created by registering the T2 FLAIR
images to standard space. Lesions were manually drawn in FSLview, and all lesion masks were
binarized. The registered, binarized, lesion masks from all participants were then added using
FSLmaths to create a single probability map with values ranging from 0 to 8, indicating the
number of participants with overlapping lesions in the tract of interest (Figure 6.2). There is
some overlap of lesions with the tract of interest; however, the majority of overlapping voxels
134
were in 10-20% of subjects. We were able to track the SMA on all participants; thus, a lesion in
part of this tract is included in the measurements obtained, and contributes to the relationship
seen with clinical measures of dual-task. Axial and sagittal slices from the map indicate that the
bulk of MS lesions in the participants were not near the SMA interhemispheric tract of interest,
and overlap of the tract is minimal (Figure 6.3).
135
136
Figure 6.2. Lesion probability map showing the interhemispheric SMA tract (blue) of a representative subject overlaid with
lesions from all MS participants (scaled red to yellow to indicate the number of subjects with lesions overlapping in this area,
see color bar) in MNI152 coordinate space (slices identified by Y coordinate). All images are in radiological convention.
136
137
Figure 6.3. Sagittal and axial slices through the interhemispheric SMA tract (blue) of a representative subject demonstrate that
the bulk of MS lesions are not near the tract of interest, and overlap of the tract is minimal. Lesion overlap is indicated by the
color bar, with the color indicating the number of subjects with lesions in this area. All images are in radiological convention and
displayed in MNI152 coordinate space.
137
6.3.3 Diffusion Tractography
There were moderately strong correlations identified between interhemispheric SMA
(Figure 6.4a) tract-specific diffusivity measures and forward dual-task walking velocity, a clinical
measure of dual-task. In particular, axial diffusivity (AD), radial diffusivity (RD) and mean
diffusivity (MD) demonstrated moderately strong correlations with forward dual-task walking
velocity (Table 6.1). Since AD, RD and MD are highly correlated, resulting in multicollinearly and
precluding multiple regression analysis, we chose to explore the strongest correlation (i.e., FWC
velocity vs. AD) with linear regression to determine the predictive value of AD. Linear regression
demonstrates a strong, significant relationship between forward dual-task walking velocity and
AD of the SMA interhemispheric tract (R2= 0.436 , p=0.014), indicating that AD explains 43.6% of
the variance in forward dual-task walking velocity (Figure 6.5). Further exploration of non-linear
relationships and the residual plot confirms that non-linear relationships do not explain these
data and normality is present in the residual plots. Given the small sample size, this result
shows promise and merits further exploration.
138
a
b
Figure 6.4. Tractography models in representative subjects demonstrating
successful modeling of the (a) interhemispheric SMA (coronal view) and (b)
interhemispheric V4 (axial view) connectivity.
To examine the specificity of the interhemispheric SMA tract diffusion measures, we
explored a control interhemispheric tract that we would not expect to be involved in clinical
measures of dual-task. We modeled the V4 interhemispheric tract (Figure 6.4b), which crosses in
the splenium,40 and found no significant or trending correlations between any of the tractspecific diffusivity measures and the clinical measures of dual-task.
139
L SMA – R SMA
FA
[(1/2)*((( 1-2)2 +
(2-3)2 + (1-3)2] /
(12 +22 + 32)
AD
(1)
RD
(2 +3)/2
MD
(1+2 +3)/3
Forward Dual-Task
0.087
-0.660
-0.595
-0.552
Velocity (m/s)
Backward Dual-Task
-0.135
-0.565
-0.432
-0.482
Velocity (m/s)
n=10
WWTT-complex (s)
0.221
0.277
0.189
0.237
TUG Cognitive (s)
0.468
0.328
0.047
0.124
Table 6.1. Relationship between clinical dual-task measures and interhemispheric SMA tractspecific parameters
Pearson’s correlation R values.
Figure 6.5 The relationship between interhemispheric SMA axial diffusivity and
forward dual-task walking velocity
140
6.4 Discussion
This is the first study to examine the relationship between the SMA interhemispheric
tract and clinically useful motor measures in individuals with MS, and the first to explore lower
extremity motor-cognitive dual-tasks in relationship to imaging metrics in MS.
The results of the DTI data revealed that individuals with RRMS who demonstrated
faster forward dual-task walking speeds (i.e., higher velocities) tended to have lower
interhemispheric SMA AD, RD and MD values (Table 6.1). Linear regression showed a significant
relationship between FWC velocity and interhemispheric AD, suggesting that dual-task
performance in forward walking may depend on the integrity of the SMA interhemispheric tract.
This relationship was not observed in the V4 interhemispheric tract, further supporting that the
SMA tract may be related to dual-task walking in individuals with MS. Although we cannot
definitively implicate the interhemispheric SMA tract in dual-task, it points toward great
potential for white matter tractography to predict functional performance and provide
important information about recovery patterns in MS.
DTI can be challenging to interpret in MS due to the unpredictable combination of axon
loss, gliosis, inflammation and demyelination, which may result in competing influences on the
diffusion tensor.41 Our results showed no correlation between FWC velocity and FA, but
moderate correlations between FWC velocity and AD, MD and RD values. Although FA is the
most commonly reported diffusivity measure, it is considered a non-specific marker of
pathology.41 Indeed, FA has been shown to correlate poorly with actual fiber anisotropy,
compared to MD which correlates well with individual fiber MD.42 FA is problematic as it is
influenced by the number of dominant fiber directions within each voxel, the anisotropy of each
fiber, and partial volume effects from neighboring grey matter.42 Indeed, our tract of interest
141
passes through the corpus callosum, which is known to have a crossing fiber structure within a
single voxel, that could skew FA values.43 Thus, FA may be suboptimal for use in disease
processes that affect myelination.42 Our work supports newer work suggesting that both MD
and FA metrics should be estimated to maximize the findings from DTI.10
Only one other study has explored the interhemispheric SMA pathway in MS.44 They
reported that the nine hole peg test, a functional measure of hand coordination and dexterity,
was significantly correlated with transverse diffusivity (2) (p<0.02). The authors concluded that
diffusivity along the primary axis (1) appears to be less affected by MS disease activity than
transverse diffusivity.44 To examine transverse diffusivity in our dataset, we correlated 2 with
FWC velocity and found similar results (R=-0.574 p=0.065) to the analysis of FWC velocity with
AD , RD and MD reported in Table 6.1.
6.4.1 Limitations
This study is limited by small sample size. In addition, lesion load analysis demonstrated
a high variability between the number and size of lesions across participants that was unrelated
to functional performance. While this may be related to the inherent variability of MS lesions, it
is also likely that many of our participants had lesions in their spinal cord, which were not
evaluated in this study. As shown in Figure 6.3, the bulk of MS lesions did not overlap with our
tract of interest, but it is likely that that these lesions contribute to overall atrophy and could
explain changes in the SMA tract of individuals with MS.
A focused approach for examining relationships between white matter connectivity
must be interpreted with caution as it describes only one potential tract related to dual-task
ability. Further, the interhemispheric SMA tract has not been examined in healthy adults to
characterize its variability. The variability of SMA diffusion measures for individuals with MS (42142
56%) is much higher than the 10% coefficients of variation found by Heiervang et al.45 in healthy
adults using similar tracts. It is possible that this is due to the inherent variability in MS lesions
and may be reduced with an increased sample size. Indeed it has been suggested that a sample
size of >106 or >72 is needed to detect between subject changes in FA and MD, respectively
with a 1% effect size for the corpus callosum.45 The interhemispheric SMA tract passes through
a small portion of the corpus callosum, so it is likely that a larger sample size is necessary to
demonstrate between subject differences.
6.4.2 Clinical Implications
This study correlated DTI with clinical outcome measures to gain a better understanding
of the neural connectivity that underlies dual-task performance in individuals with RRMS. Our
results suggest that the SMA interhemispheric tract may play a role in dual-task performance.
DTI may be a useful adjunct to clinical measures to predict dual-task performance and provide
important information about recovery patterns in MS. In neurodegenerative diseases like MS,
functional recovery can be challenging to objectively report, and use of DTI could show
microstructural improvements and suggest improved connectivity. As always, it is important to
note that a direct relationship to axon or myelin structure should not be assumed from DTI,
which is an indirect measure of in vivo brain structure.
6.4.3 Future Directions
Future work in this area should include a larger sample size and matched healthy
controls to further identify underlying changes in individuals with MS and to define the
variability of these interhemispheric tracts in healthy adults. Our future work aims to explore
changes in the SMA interhemispheric tract following an exercise training program targeting
143
dual-task. We plan to explore the ability of imaging measures to predict which individuals with
MS can benefit from rehabilitation training programs.
144
6.5 References
1. Whetten-Goldstein K, Sloan FA, Goldstein LB, Kulas ED. A comprehensive assessment of the
cost of multiple sclerosis in the United States. Mult Scler. 1998; 4: 419-425.
2. Frohman, E.M. Multiple sclerosis. Med Clin N Am. 2003; 87(4): 867–897.
3. Noseworthy JH, Lucchinetti C, Rodriguez M, Weinshenker BG. Medical progress: multiple
sclerosis. New Engl J Med. 2000;343:938–952.
4. Amato MP, Portaccio E, Goretti B, Zipoli V, Hakiki B, Giannini M, Pasto L, Razzolini L.
Cognitive impairment in early stages of multiple sclerosis. J Neurol Sci. 2010; Suppl 2; S211S214.
5. Rao SM, Leo GJ, Bernardin L, Unverzagt F. Cognitive dysfunction in multiple sclerosis I:
frequency, patterns, and prediction. Neurology. 1991; 41(5): 685-691.
6. Hamilton F, Rochester L, Paul L et al. Walking and talking: an investigation of cognitivemotor dual tasking in multiple sclerosis. Mult Scler. 2009; 15(10): 1215-1227.
7. Amato MP, Zipoli V, Portaccio E. Cognitive changes in multiple sclerosis. Expert Rev
Neurotherapeutics. 2008; 8: 1585-1596.
8. Nilsagard Y, Denison E, Gunnarsson LG, Bostrom K. Factors perceived as being related to
accidental falls in persons with multiple sclerosis. Disabil Rehabil. 2009; 31: 1301-1310.
9. Bakshi R, Thompson AJ, Rocca MA et al. MRI in multiple sclerosis: current status and future
prospects. Lancet Neurol. 2008; 7: 615-625.
10. Ge Y, Law M, Grossman RI. Applications of diffusion tensor MR imaging in Multiple Sclerosis.
Ann NY Acad Sci. 2005; 1064: 202-219.
11. Madden DJ, Bennett IJ, Song AW. Cerebral white matter integrity and cognitive aging:
contributions from diffusion tensor imaging. Neuropsychol Rev. 2009; 19: 415- 435.
12. Basser PJ, Mattiello J, LeBihan D.MR diffusion tensor spectroscopy and imaging. Biophys J.
1994;66:259-267.
13. Beaulieu C, Allen PS.Determinants of anisotropic water diffusion in nerves. Magn Reson
Med. 1994;31:394-400.
14. Goldberg-Zimring D, Mewes AUJ, Maddah M, Warfield SK. Diffusion tensor magnetic
resonance imaging in Multiple Sclerosis. J. Neuroimaging. 2005; S15(4): 68S-81S.
145
15. Jones DK. Gaussian modeling of the diffusion signal. In Johansen-Berg H, Behrens TEJ.
Diffusion MRI: from quantitative measurement to in vivo neuroanatomy. 1st ed. London, UK:
Elsevier Inc; 2009:37-54.
16. Wilson M, Trench CR, Morgan PS, Blumhardt LD. Pyramidal tract mapping by diffusion
tensor magnetic resonance imaging in multiple sclerosis: improving correlations with
disability. J Neuro Neurosurg Ps. 2003; 74: 203-207.
17. Reich DS, Zackowski KM, Gordon-Lipkin EM, Smith SA, Chadkowski BA, Cutter GR, Calabresi
PA. Corticospinal tract abnormalities are associated with weakness in multiple sclerosis. Am
J Neuroradiol. 2008; 29: 333-339.
18. Ibrahim I, Tintera J, Skoch A, Jirů F, Hlustik P, Martinkova P, Zvara K, Rasova K. Fractional
anisotropy and mean diffusivity in the corpus callosum of patients with multiple sclerosis:
the effect of physiotherapy. Neuroradiology. 2011; 53: 917-926.
19. Lin X, Tench CR, Morgan PS, Constantinescu CS. Use of combined conventional and
quantitative MRI to quantify pathology related to cognitive impairment in multiple sclerosis.
J Neuro Neurosurg Ps. 2008; 79: 437-441.
20. Song SK, Sun SW, Ramsbottom MJ, Chang C, Russell J, Cross AH. Dysmyelination revealed
through MRI as increased radial (but unchanged axial) diffusion of water. Neuroimage. 2002;
17: 1429-1436.
21. Filippi M, Iannucci G, Cercignani M et al. A quantitative study of water diffusion in multiple
sclerosis lesions and normal-appearing white matter using echo-planar imaging. Arch
Neurol. 2000; 57(7): 1017–1021.
22. Cercignani M, Iannucci G, Rocco MA et al. Pathologic damage in MS assessed by diffusionweighted and magnetization transfer MRI. Neurology. 2000; 54(5): 1139–1144.
23. Filippi M, Cercignani M, Inglese M et al. Diffusion tensor magnetic resonance imaging in
multiple sclerosis. Neurology. 2001; 56(3): 304–311.
24. Rocca MA, Cercignani M, Iannucci G et al. Weekly diffusion-weighted imaging of normalappearing white matter in MS. Neurology. 2001; 55(6): 882–884.
25. Cercignani M, Bozzali M, Iannucci G et al. Magnetisation transfer ratio and mean diffusivity
of normal appearing white and grey matter from patients with multiple sclerosis. J Neuro
Neurosurg Ps. 2001; 70(3): 311–317.
26. Guo AC, Jewells VL, Provenzale JM. Analysis of normal-appearing white matter in multiple
sclerosis: comparison of diffusion tensor MR imaging and magnetization transfer imaging.
Am J Neuroradiol. 2001; 22(10): 1893–1900.
146
27. Guo AC, Macfall JR, Provenzale JM. Multiple sclerosis: diffusion tensor MR imaging for
evaluation of normal-appearing white matter. Radiology. 2002; 222(3): 729–736.
28. Ge Y, Law M, Johnson G et al. Preferential occult injury of corpus callosum in multiple
sclerosis measured by diffusion tensor imaging. J Magn Reson Imaging. 2004; 20(1): 1-7.
29. Johansen-Berg H, Behrens TEJ. Diffusion MRI: From Quantitative Measurement to In Vivo
Neuroanatomy. London: Academic Press; 2009.
30. Toyokura M, Muro I, Komiya T, Obara M. Activation of pre‐supplementary motor area (SMA)
and SMA proper during unimanual and bimanual complex sequences: an analysis using
functional magnetic resonance imaging. J Neuroimaging. 2002; 12: 172‐178.
31. Johansen‐Berg H, Matthews PM. Attention to movement modulates activity in sensormotor
areas, including primary motor cortex. Exp Brain Res. 2002; 142: 13‐24.
32. Debaere F, Swinnen SP, Beatse E, Sunaert S, Van Hecke P, Duysens J. Brain areas involved in
interlimb coordination: a distributed network. Neuroimage. 2001; 14: 947‐958.
33. Givon U, Zeilig G, Achiron A. Gait analysis in multiple sclerosis: characterization of temporalspatial parameters using GAITRite functional ambulation system. Gait Posture. 2009; 29:
138-142.
34. Verghese J, Buschke H, Viola L, Katz M, Hall C, Kuslansky G, Lipton R. Validity of divided
attention tasks in predicting falls in older individuals: a preliminary study. J Am Geriatr Soc.
2002; 50: 1572-1576.
35. Podsiadlo D, Richardson S. The timed ―Up & Go‖: a test of basic functional mobility for frail
elderly persons. J Am Geriatr Soc. 1991; 39: 142-148.
36. Nilsagård Y, Lundholm C, Gunnarsson LG, Denison E. Clinical relevance using timed walk
tests and ‘timed up and go‘ testing in persons with multiple sclerosis. Physiotherapy
Research International. 2007; 12(2):105-114.
37. Shumway-Cook A, Brauer S, Woollacott M. Predicting the probability for falls in communitydwelling older adults using the timed up & go test. Phys Ther. 2000; 80(9): 896-903.
38. Nilsagård Y, Lundholm C, Denison E, Gunnarsson LG. Predicting accidental falls in people
with multiple sclerosis—a longitudinal study. Clin Rehabil. 2009; 23: 259-269.
39. Udupa JK, Samarasekera S. Fuzzy Connectedness and Object Definition: Theory, Algorithms,
and Application in Image Segmentation. Graphical Models and Image Processing. 1996;
58:246-261.
147
40. Putnam MC, Steven MS, Doron KW, Riggall AC, Gazzaniga MS. Cortical projection
topography of the human splenium: hemispheric asymmetry and individual differences. J
Cognitive Neurosci. 2009; 22(8): 1662-1669.
41. Alexander AL, Lee JE, Lazar M, Field AS. Diffusion tensor imaging of the brain.
Neurotherapeutics. 2007; 4(3): 316-329.
42. Leow AD, Zhan L, Zhu S, Hageman N, Chiang MC, Barysheva M, Toga AW, Thompson PM.
White matter integrity measured by fractional anisotropy correlates poorly with actual
individual fiber anisotropy. I S Biomed Imaging. 2009. Available at:
www.loni.ucla.edu/~thompson/PDF/Leow-ISBI09.pdf
43. Tournier JD, Mori S, Leemans A. Diffusion Tensor Imaging and Beyond. Magn Res Med.
2011; 65: 1532-1556.
44. Lowe MJ, Horenstein C, Hirsch JG, Marrie RA, Stone L, Bhattacharyya PK, Gass A, Phillips MD.
Functional pathway-defined MRI diffusion measures reveal increased transverse diffusivity
of water in Multiple Sclerosis. Neuroimage. 2006; 32: 1127-1133.
45. Heiervang E, Behrens TEJ, Mackay CE, Robson MD, Johansen-Berg H. Between session
reproducibility and between subject variability of diffusion MR and tractography measures.
Neuroimage. 2006; 33: 867-877.
148
Chapter 7: Research Findings and Future Directions
The objective of this research was to investigate the interaction between motor and
cognitive factors contributing to gait variability and falls and to explore the underlying neural
connectivity related to dual-task ability in persons with neurologic disabilities. It was
hypothesized that motor and cognitive factors contribute to the increased gait variability and
could increase fall risk in individuals with both minimal and severe baseline deficits. It was
further hypothesized that the neural connectivity would be related to functional measures of
dual-task, which could improve clinical prediction and rehabilitation of such deficits.
7.1 Summary of Findings
In chapter 2, the available literature on motor-cognitive dual-task training (DTT) in individuals
with neurologic conditions was reviewed. We found that although the evidence was
inconclusive, it appeared that DTT was not detrimental to rehabilitation and may be beneficial.
Indeed, the double-blinded study with the largest sample size1 reported significant
improvements in both dual-task cost velocity and stride length. This review highlighted several
important gaps which directed our research program. The gaps identified were: 1) all studies
had utilized minimally-moderately impaired individuals only; 2) there was no evidence that DTT
could be successful in individuals with severe injury; 3) although it is known that dual-tasking
increases gait variability, this was not utilized as an outcome in any of the studies; 4)
149
none of these studies measured backward walking, which has also been shown to increase gait
variability in Parkinson’s disease (PD);2-3 5) few studied cognitive outcomes, making it difficult to
ascertain the degree to which the interaction between motor and cognitive domains plays a role
in dual-task ability; and finally 6) although many neurologic populations were represented in
these works, there were no studies exploring DTT in multiple sclerosis (MS), despite deficits in
dual-task being well documented in this population. Figure 7.1 demonstrates how the gaps
identified in this systematic review directed our research aims.
Thus, we undertook a study exploring gait variability and backward walking in a
minimally impaired population to understand whether it might be an important contributor to
fall risk (chapter 3). We tested young, middle-aged and both healthy and frail elderly adults and
explored the difference in variability and other spatiotemporal measures of gait in both forward
and backward walking. In chapter 3, we showed that while all subjects demonstrated an
increase in gait variability when switching from forward to backward walking, young and middle
aged adults had similar changes, while elderly individuals demonstrated significantly larger
increases in gait variability than young and middle-aged adults. Elderly fallers all walked with a
BW velocity of less than 0.6m/s and after further analysis, 0.6m/s backward walking velocity
accurately identified elderly fallers with a sensitivity of 100% and a specificity of 33%.4 This
demonstrated that backward walking could be a useful tool in individuals with minimal to mild
impairment (elderly fallers), and could hold potential for use in individuals with more severe
impairment.
To explore the feasibility and acceptability of DTT in individuals with severe
impairments, we undertook a case study in an individual with a severe traumatic brain injury
(TBI) (chapter 4). Despite marked baseline impairments in attention and global left-sided
150
weakness, this individual demonstrated improvements in both previously automatic movements
(e.g., gait speed, descending stairs) as well as the ability to dual-task on the Walking While
Talking Test at a greater rate following DTT than standard physical therapy alone.5 Although we
are unable to draw conclusions about effectiveness from a single subject, we can once again say
that the addition of DTT to standard physical therapy was feasible, was not detrimental to the
subject, and appears to be related to benefits in motor function. To our knowledge, this was the
first study to explore motor-cognitive DTT in an individual with severe TBI. These findings
suggested that DTT could be an important tool for rehabilitation in individuals with other
conditions across the impairment spectrum.
151
Contributing Factors
Increased
Gait Variability
Systematic
Review
GAPS
Decreased Balance
Difficulty Walking
Backwards
Motor
Factors
Severity of
Injury
Mild Severe
Difficulty Dual-Tasking
152
Poor Divided Attention
Efficacy of
Training
IDENTIFIED
Interaction
of Motor
and
Cognitive
Training in
Multiple
Sclerosis
Cognitive
Factors
Decreased Executive
Function
Increased
Fall Risk
Backward
Walking
Cross
Sectional
Study
Dual-Task
Training in
Severe TBI
(Case Study)
Relationship
between gait
variability
and cognitive
factors in MS
Relationship
between
Imaging and
Clinical
Measures
in MS
Figure 7.1 Schema exploring how a review of the literature identified several gaps in knowledge (orange box) that
directed the research aims (purple box) of this document.
152
Dual-Task
Training in
MS
The results of both chapters 5 and 6 are part of a large randomized controlled trial
examining the effect of dual-task training using the video game Dance Dance Revolution on
mobility, cognition and brain plasticity in individuals with MS. Baseline data is presented in
chapters 5 and 6 as the trial is ongoing and unblinding has yet to occur.
Only descriptive studies exist to outline the dual-task deficits seen in MS. It is well
known that individuals with MS experience impairments in both motor and cognitive function.
These impairments can often occur early in the disease process, although the heterogeneity of
MS presentation makes this challenging to generalize. In chapter 5, we explored the relationship
between mobility (gait variability in backward and dual-task walking) and cognition
(neuropsychological test performance) and demonstrated the influence of dual-task on gait
variability in a small sample of individuals with RRMS. Gait variability measures in both
conditions of dual-task (FWC, BWC) correlate strongly with performance on the SDMT and
Stroop congruent and neutral conditions, suggesting the role of cognitive dysfunction in dualtask ability. Furthermore, performance on neuropsychological tests of processing speed
explained between 48 and 65% of the variance in dual-task gait variability measures. Our results
are supported by a recent study exploring the contribution of cognition to dual-task cost of
velocity that showed cognitive impairment was an independent predictor of dual-task cost in
MS.6 Although we were unable to show a relationship between variability of dual-task walking
and fall risk or indeed backward walking velocity and fall risk in this MS sample, as we did in
elderly fallers, this is likely due to small sample size, a group of infrequent fallers and large
standard deviations in BW measures among participants. Despite these limitations, this study is
encouraging and supports further work to explore dual-task in MS and potential rehabilitation
strategies that incorporate both motor and cognitive measures.
153
Psychologists have explored dual-task ability, in the context of two cognitive tasks
(cognitive-cognitive dual-task) in healthy adults using functional neuroimaging. However, the
exploration of motor-cognitive dual-task using DTI diffusivity measures and probabilistic
tractography represents a novel analysis. In chapter 6, we showed a moderate relationship
between velocity of forward dual-task walking and measures of axial, radial and mean diffusivity
(AD, RD, MD) in the interhemispheric supplemental motor area (SMA) tract. Individuals who
walked at slower speeds during dual-tasks had higher values of AD, RD and MD. When the
strongest of these relationships was further examined with linear regression, we found that
interhemispheric SMA AD explained 43.6% of the variance in forward dual-task walking velocity.
This work was done in the same individuals as in chapter 5; thus, given the small sample size,
the relationship between dual-task walking velocity and diffusion tensor imaging (DTI) measures
shows promise.
There was no relationship between dual-task walking and tract-specific diffusion
measures in another interhemispheric tract, suggesting that the SMA tract may be specifically
related to dual-task ability. Indeed the SMA has been implicated in complex tasks and interlimb
coordination in other studies.7-9 To this investigator’s knowledge, there are no DTI studies
examining the relationship with dual-task walking; however, studies in other neurologic
populations have demonstrated an association between DTI measures and motor function. In
MS, the RD of the transcallosal hand motor fibers predicted decline on a coordination task over
12 months,10 while fractional anisotropy (FA) of the corticospinal tract predicted gait function in
chronic stroke11 and chronic idiopathic hydrocephalus,12 as well as balance function in children
with TBI.13 Similarly, FA of the corpus callosum predicted motor function in community-dwelling
elderly adults14 and adolescents with TBI.15 Recent studies are exploring the predictive value of
154
DTI measures for future walking, motor, and ADL functional deficits in both children16 and
adults17-19 with stroke. Our results support the literature from other neurologic conditions; DTI
measures have the potential to predict motor function in MS. This is the first study to explore
the association of DTI to dual-task function.
7.2 Limitations
This work has many limitations. The results in chapters 5 and 6 are limited primarily by
sample size. We analyzed eleven individuals with relapsing remitting MS. Our analysis priori
sample size calculation in chapter 5 indicated that 10 matched pairs of healthy adults and
individuals with MS would be sufficient to demonstrate differences in forward walking velocity.
Despite that calculation, we noted no significant differences in forward walking velocity, but did
note robust relationships between gait variability and neuropsychological test performance.
Additionally, this study examined only one type of cognitive dual-task (arithmetic task); it may
be beneficial to explore additional secondary cognitive tasks such as fluency or working
memory.
Aside from small sample size, chapter 6 was limited by the exploration of only one tract
(interhemispheric SMA) with a potential relationship to dual-task ability; such a focused
approach must be interpreted with caution. Additionally, the interhemispheric SMA tract has
not been examined in healthy adults to characterize its variability. However, the variability of
the SMA diffusion measures for individuals with MS was much higher than the 10% coefficients
of variation found by Heiervang et al., 200620 in healthy adults using similar tracts. While this
may be due in part to the inherent variability of MS lesion location and size, it may also be
limited by a small sample size.
155
7.3 Future Research
7.3.1 Contribution of Cognition and Gait Variability to Dual-Task Ability
Our work reinforces the important link between gait variability and the interplay of
cognition and mobility. Work in the elderly suggests that increased gait variability noted in
backward walking may be associated with increased fall risk.2,21 A prospective study exploring
this phenomenon would be useful for treatment planning and wellness programs in individuals
with MS. Work is also needed to further elucidate the role of early cognitive deficits as potential
contributor to mobility deficits in MS. Future work should include a wider range of
neuropsychological tests as well as dual-tasks that span multiple cognitive domains (e.g.,
fluency, working memory). While the interaction of cognition and mobility seems to be
particularly relevant in MS, it is possible that such a relationship also exists in other neurological
populations. Future work involving dual-task in other neurological conditions should include the
assessment of both cognitive domains and mobility outcomes.
7.3.2 Dual-Task Training in General
The variety of activities that may be performed for each type of dual-task (i.e., motormotor, motor-cognitive, cognitive-cognitive) during training studies clouds study design and can
make reproducibility and generalizability of findings challenging. Our review suggests that
motor-cognitive DTT may be useful for improving dual-task gait, but the necessary dose of
training required for such outcomes is unknown. Furthermore, such training has been largely
unsuccessful at generalizing to novel dual-task situations. That is to say, individuals improve only
on the dual-task on which they were trained; however, many studies do not include an outcome
measure of a novel dual-task. Future studies should include a dual-task assessment in pre and
post-testing that is not part of the training session to assess carryover of dual-task skills to novel
156
dual-task situations. Indeed, a recent pilot study22 suggests that carryover is possible if
functional tasks are the focus of training. Thus, it appears that a DTT program focusing on
functional skills may be beneficial in the transition to novel dual-tasks.
Our review also highlighted a surprising lack of research in the area of DTT for
individuals with neurological disorders resulting in severe impairments. Our work suggests that
functional mobility can be enhanced with DTT, even in individuals with severe injury. Future
studies should include a larger sample size of individuals with severe injury as well as explore
DTT in neurological conditions other than TBI.
7.3.3 Dual-Task Training in MS
Our ongoing work is exploring the effect of a dual-task training program on individuals
with MS. Based on the gaps identified in the literature review, we have incorporated both fall
assessment and a novel dual-task assessment in the outcome measures of this randomized
controlled trial.
7.3.4 Validation of Dual-Task Tools in MS
Although there have been several measurement tools utilized to assess dual-task ability,
none of them have been validated in MS. Future work will validate the Walking While Talking
Test, the Dual-Task Questionnaire, and the TUG Cognitive in individuals with MS.
7.3.5 Examine Variability in Healthy Controls for MRI Measures
Interpretation of our data in chapter 6 was limited not only by small sample size, but
also by the lack of information on SMA tract variability in healthy controls. While we anticipated
that our MS sample would have higher variability in measures of diffusivity, they were
significantly higher than those seen in other tracts in healthy adults.20 To my knowledge, the
interhemispheric SMA tract has not been explored in healthy adults, so comparison of our
157
values in this study was not possible. Future work should include MRI measures for the healthy
control subjects in addition to mobility testing. An exploration of connectivity of other tracts
that may be related to dual-tasks as well as relationships with clinical tests of dual-task ability in
individuals with MS would contribute to our understanding of the neural connectivity associated
with dual-task ability.
7.3.6 Examine Responders vs. Non-Responders to Rehabilitation Training Programs and Explore
the Predictive Potential of MRI Measures.
In addition to exploring mobility and neuropsychological test performance outcomes
following the dual-task training program in MS, we will also examine the effect of a DTT program
on connectivity in the interhemispheric SMA tract. We also plan to explore the relationship
between measures of diffusivity and success or improvement following the DTT program in
individuals who were randomized to the intervention group. Prediction of individuals with MS
who can benefit from rehab training programs indicates that this work has the potential to be a
valuable tool in the prescription of exercise programming and rehabilitation in MS.
7.4 Conclusion
Deficits in dual-task that amplify gait variability and increase fall risk are a significant
concern for individuals with MS and other neurological disorders. The studies completed in this
research support the hypothesis that motor and cognitive factors contribute to gait variability
that could increase fall risk in individuals with both minimal and severe baseline deficits. These
studies also demonstrate the relationship between connectivity and functional measures of
dual-task, which could be useful in improving clinical prediction and rehabilitation.
158
Although continued research is necessary to refine the interaction between mobility and
cognition in MS and other neurologic populations, this work certainly contributes to the
knowledge base and is a step toward better understanding of the many factors contributing to
gait variability and fall risk. Physical therapist interest in the area of dual-task has increased over
the last several years. While the number of descriptive studies has greatly increased, knowledge
about rehabilitation of dual-task deficits with training programs is still in its infancy. Indeed, the
success of such programs and the type of training tasks that should be utilized for different
populations or different severities of injury are unknown. Our ongoing and future work will aim
to better define appropriate training, assessment, and predictive tools for each neurologic
population.
159
7.5 References
1. Schwenk M, Zieschang T, Oster P, Hauer K. Dual-task performances can be improved in
patients with dementia: a randomized controlled trial. Neurology. 2010; 74: 1961-1968.
2. Hackney ME, Earhart GM. Backward walking in Parkinson’s disease. Movement Disord.
2009; 24(2): 218-223.
3. Hackney ME, Earhart GM. The effects of a secondary task on forward and backward
walking in Parkinson disease. NeurorehabNeural Re. 2010; 24(1):97-106.
4. Fritz NE, Worstell AM, Kloos AD, Siles AB, White SE, Kegelmeyer DA. Backward walking
measures are sensitive to age-related changes in mobility and balance. Gait Posture.
2013; 37: 593-597.
5. Fritz NE, Basso DM. Dual-task training for balance and mobility in a person with severe
traumatic brain injury: a case study. J Neurol Phys Ther. 2013; 37: 37-43.
6. Sosnoff JJ, Socie MJ, Sandroff BM, Balantrapu S, Suh Y, Pula JH et al. Mobility and
cognitive correlates of dual-task cost of walking in persons with multiple sclerosis.
Disabil Rehabil. 2013; epub ahead of print:
http://www.ncbi.nlm.nih.gov/pubmed/23597000.
7. Toyokura M, Muro I, Komiya T, Obara M. Activation of pre-supplementary motor area
(SMA) and SMA proper during unimanual and bimanual complex sequences: an analysis
using functional magnetic resonance imaging. J Neuroimaging. 2002; 12: 172-178.
8. Johansen-Berg H, Matthews PM. Attention to movement modulates activity in sensormotor areas, including primary motor cortex. Exp Brain Res. 2002; 142: 13-24.
9. Debaere F, Swinnen SP, Beatse E, Sunaert S, Van Hecke P, Duysens J. Brain areas
involved in interlimb coordination: a distributed network. Neuroimage. 2001; 14: 947958.
10. Kern KC, Sarcona J, Montag M, Giesser BS, Sicotte NL. Corpus callosal diffusivity predicts
motor impairment in relapsing-remitting multiple sclerosis: a TBSS and tractography
study. Neuroimage. 2011; 55(3): 1169-1177.
11. Yeo SS, Ahn SH, Choi BY, Chang CH, Lee J, Jang SH. Contribution of the
pedunculopontine nucleus on walking in stroke patients. Eur Neurol. 2011; 65: 332-337.
160
12. Hattingen E, Jurcoane A, Melber J, Blasel S, Zanella FE, Neumann-Haefelin T, Singer OC.
Diffusion tensor imaging in patients with adult chronic idiopathic hydrocephalus.
Neurosurgery. 2010; 66(5): 917-924.
13. Caeyenberghs K, Leemans A, Geurts M, Taymans T, Linden CV, Bouwien CM, et al. Brainbehavior relationships in young traumatic brain injury patients: DTI metrics are highly
correlated with postural control. Hum Brain Mapp. 2010; 31(7): 992-1002.
14. Bhadelia RA, Price LL, Tedesco KL, Scott T, Qui WQ, Patz S, et al. Diffusion tensor
imaging, white matter lesions, the corpus callosum, and gait in the elderly. Stroke. 2009;
40: 3816-3820.
15. Caevenberghs K, Leemans A, Geurts M, Linden CV, Smits-Englesman BC, Sunaert S, et al.
Correlations between white matter integrity and motor function in traumatic brain
injury patients. Neurorehab Neural Re. 2011; 25: 492-502.
16. Rose J, Mirmiran M, Butler EE, Lin CY, Barnes PD, Kermoian R, et al. Neonatal
microstructural development of the internal capsul on diffusion tensor imaging
correlates with severity of gait and motor deficits. Dev Med Child Neurol. 2007; 49(10):
745-750.
17. Koyama T, Marumoto K, Miyake H, Domen K. Relationship between diffusion tensor
fractional anisotropy and motor outcome in patients with hemiparesis after corona
radiate infarct. J Stroke Cerebrovasc. 2013a; epub ahead of print:
http://www.ncbi.nlm.nih.gov/pubmed/23510690.
18. Koyama T, Tsuji M, Nishimura H, Miyake H, Ohmura T, Domen K. Diffusion tensor
imaging for intracerebral hemorrhage outcome prediction: comparison using data from
the corona radiata/internal capsule and the cerebral peduncle. J Stroke Cerebrovasc.
2013b; epub ahead of print: http://www.ncbi.nlm.nih.gov/pubmed/21795065.
19. Wang DM, Li J, Liu JR, Hu HY. Diffusion tensor imaging predicts long-term motor
functional outcome in patients with acute supratentorial intracranial hemorrhage.
Cerebrovasc Dis. 2012; 34(3): 199-205.
20. Heiervang E, Behrens TEJ, Mackay CE, Robson MD, Johansen-Berg H. Between session
reproducibility and between subject variability of diffusion MR and tractography
measures. Neuroimage. 2006; 33: 867-877.
21. Hausdorff JM, Mitchell SL, Firtion R, Peng CK, Cudkowicz ME, Wei JY, et al. Altered
fractal dynamics of gait: reduced stride-interval of aging and Huntington’s disease. J
Appl Physiol. 1997; 82(1): 262-269.
161
22. Yogev-Seligmann G, Giladi N, Brozgol M, Hausdorff JM. A training program to improve
gait while dual-tasking in patients with Parkinson’s disease: a pilot study. Arch Phys Med
Rehab. 2012; 93(1): 176-181.
162
References
Alexander AL, Lee JE, Lazar M, Field AS. Diffusion tensor imaging of the brain.
Neurotherapeutics. 2007; 4(3): 316-329.
Al-Omaishi J, Bashir R, Gendelman HE. The cellular immunology of multiple sclerosis. J Leukocyte
Biol. 1999; 65:444–452.
Al-Yahya E, Dawes H, Collett J, Howells K, Izadi H, Wade DT, Cockburn J. Gait adaptations to
simultaneous cognitive and mechanical constraints. Exp Brain Res. 2009; 199: 39- 48.
Amato MP, Portaccio E, Goretti B, Zipoli V, Battaglini M, Bartolozzi ML, Stromillo ML, Guidi L,
Siracusa G, Sorbi S, Federico A, De Stefano N. Association of neocortical volume changes with
cognitive deterioration in relapsing-remitting multiple sclerosis. Arch Neurol. 2007; 64(8): 11571161.
Amato MP, Portaccio E, Goretti B, Zipoli V, Hakiki B, Giannini M, Pasto L, Razzolini L. Cognitive
impairment in early stages of multiple sclerosis. J Neurol Sci. 2010; Suppl 2; S211-S214.
Amato MP, Zipoli V, Portaccio E. Cognitive changes in multiple sclerosis. Exp Rev
Neurotherapeutics. 2008; 8: 1585-1596.
Asloun S, Soury S, Couillet J, Giroire JM, Joseph PA, Mazaux JM, Azouvi P. Interactions between
divided attention and working-memory load in patients with severe traumatic brain injury. J Clin
Exp Neuropsyc. 2008; 30(4): 481-490.
Azouvi P, Couillet J, Leclercq M, et al. Divided attention and mental effort after severe traumatic
brain injury. Neuropsychologia. 2004;42:1260–1268.
Bakshi R, Thompson AJ, Rocca MA et al. MRI in multiple sclerosis: current status and future
prospects. Lancet Neurol. 2008; 7: 615-625.
Basser PJ, Mattiello J, LeBihan D.MR diffusion tensor spectroscopy and imaging. Biophys J.
1994;66:259-267.
Beauchet O, Annweiler C, Allali G, Berrut G, Herrmann FR, Dubost V. Recurrent falls and dual
task-related decrease in walking speed: is there a relationship? J Am Geriatr Soc. 2008;
56(7):1265-1269.
163
Beauchet O, Dubost V, Aminian K, Gonthier R, Kressig RW. Dual-task-related gait changes in the
elderly: does the type of cognitive task matter? J Mot Behav 2005; 37:259-264.
Beaulieu C, Allen PS.Determinants of anisotropic water diffusion in nerves. Magn Reson Med.
1994;31:394-400.
Benedetti MG, Piperno R, Simoncini L, Bonato P, Tonini A, Giannini S. Gait abnormalities in
minimally impaired multiple sclerosis patients. Mult Scler. 1999; 5(5): 363- 368.
Bebko JM, Demark JL, Osborne PA, Majumder S, Ricciuti CJ, Rhee T. Acquisition and
automatization of a complex task: an examination of three-ball cascade juggling. J Motor Behav.
2003;35:109–18.
Behrens TEJ and Jbabdi S. MR diffusion tractography. In: Johansen-Berg H and Behrens TEJ, eds.
Diffusion MRI: From quantitative measurement to in vivo neuroanatomy. 1st ed. London, UK:
Elsevier Academic Press: 2009: 333-351.
Bhadelia RA, Price LL, Tedesco KL, Scott T, Qui WQ, Patz S, et al. Diffusion tensor imaging, white
matter lesions, the corpus callosum, and gait in the elderly. Stroke. 2009; 40: 3816-3820.
Bilney B, Morris M, Webster K. Concurrent related validity of the GAITRite walkway system for
quantification of the spatial and temporal parameters of gait. Gait Posture. 2003; 17:68–74.
Bloem BR, Grimbergen YAM, van Dijk JG, Munneke M. The “posture second” strategy: a review
of wrong priorities in Parkinson’s disease. J Neurol Sci. 2006; 248: 196-204.
Bloem BR, Hausdorff JM, Visser JE, Giladi N. Falls and freezing of gait in Parkinson's disease: a
review of two interconnected, episodic phenomena. Movement Disord. 2004;19(8):871-884.
Bogner JA, Corrigan JD, Fugate L, Mysiew WJ, Clinchot D. Role of agitation in prediction of
outcomes after traumatic brain injury. Am J Phys Med Rehabil. 2001; 80(9): 636-644.
Brauer SG, Morris ME. Can people with Parkinson’s disease improve dual tasking when walking?
Gait Posture. 2010; 31: 229-233.
Caeyenberghs K, Leemans A, Geurts M, Taymans T, Linden CV, Bouwien CM, et al. Brainbehavior relationships in young traumatic brain injury patients: DTI metrics are highly correlated
with postural control. Hum Brain Mapp. 2010; 31(7): 992-1002.
Caevenberghs K, Leemans A, Geurts M, Linden CV, Smits-Englesman BC, Sunaert S, et al.
Correlations between white matter integrity and motor function in traumatic brain injury
patients. Neurorehab Neural Re. 2011; 25: 492-502.
Cameron MH, Horak FB, Herndon RR, Bourdett D. Imbalance in multiple sclerosis: a result of
slowed spinal somatosensory conduction. Somatosens Mot Res. 2008; 25: 113-122.
164
Cameron MH, Lord S. Postural control in multiple sclerosis: implications for fall prevention. Curr
Neurol Neurosci. 2010; 10: 407-412.
Canning CG, Ada L, Woodhouse E. Multiple-task walking training in people with mild to
moderate Parkinson’s disease: a pilot study. Clin Rehabil. 2008; 22: 226-233.
Cattaneo D, De NC, Fascia T, Macalli M, Pisoni I, Cardini R. Risks of falls in subjects with multiple
sclerosis. Arch Phys Med Rehab. 2002; 83(6): 864-867.
Cattaneo D, Regola A, Meotti M. Validity of six balance disorders scales in persons with multiple
sclerosis. Disabil Rehabil. 2006; 28(12): 789-795.
Cercignani M, Bozzali M, Iannucci G et al. Magnetisation transfer ratio and mean diffusivity of
normal appearing white and grey matter from patients with multiple sclerosis. J Neuro
Neurosurg Ps. 2001; 70(3): 311–317.
Cercignani M, Iannucci G, Rocco MA et al. Pathologic damage in MS assessed by diffusionweighted and magnetization transfer MRI. Neurology. 2000; 54(5): 1139–1144.
Chan BK, Marshall LM, Winters KM, Faulkner KA, Schwartz AV, Orwoll ES. Incident fall risk and
physical activity and physical performance among older men: the Osteoporotic Fractures in Men
Study. Am J Epidemiol. 2007;165(6):696-703.
Choi JT, Bastian AJ. Adaptation reveals independent control networks for human walking. Nat
Neurosci. 2007; 10(8):1055-1062.
Chua, KSG, Ng YS, Yap SGM, Bok CW. A brief review of traumatic brain injury rehabilitation. Ann
Acad Med Singapore. 2007; 36: 31-42.
Cicerone KD. Attention deficits and dual task demands after mild traumatic brain injury. Brain
Injury. 1996;10:79–89.
Cocchini C, Della Sala S, Logie RH, Pagani R, Sacco L, Spinnler H. Dual task effects of walking
when talking in Alzheimer’s disease. Revue Neurologique (Paris). 2004; 160(1): 74-80.
Couillet J, Soury S, Lebornec G, et al. Rehabilitation of divided attention after severe traumatic
brain injury: a randomised trial. Neuropsychol Rehabil.2010; 20(3): 321-339.
Crenshaw SJ, Royer TD, Richards JG, Hudson DJ. Gait variability in people with multiple sclerosis.
Mult Scler. 2006; 12: 613-619.
Crowe SF. The differential contribution of mental tracking, cognitive flexibility, visual search and
motor speed to performance on Parts A and B of the Trail Making Test. J Clin Psychol. 1998; 54:
585-591.
165
Cusick CP, Gerhart KA, Mellick DC. Participant-proxy reliability in traumatic brain injury outcome
research. J Head Trauma Rehabil. 2000; 15(1): 739-749.
Debaere F, Swinnen SP, Beatse E, Sunaert S, Van Hecke P, Duysens J. Brain areas involved in
interlimb coordination: a distributed network. Neuroimage. 2001; 14: 947-958.
deGuise E, LeBlanc J, Feyz M, Lamoureux J. Prediction of outcome at discharge from acute care
following traumatic brain injury. J Head Trauma Rehabil. 2006; 21(6): 527-536.
Deloire MS, Bonnet MC, Salort E, et al. How to detect cognitive dysfunction at early stages of
multiple sclerosis? Mult Scler. 2006; 12: 445-452.
Deloire MS, Salort E, Bonnet M et al. Cognitive impairment as a marker of diffuse brain
abnormalities in early relapsing remitting multiple sclerosis. J Neurol Neurosur Ps. 2005; 76: 519526.
D‘Esposito M, Onishi K, Thompson H et al. Working memory impairments in multiple sclerosis:
evidence from a dual-task paradigm. Neuropsychology. 1996; 10(1): 51-56.
Drake AS, Weinstock-Guttman B, Morrow SA, Hojnacki D, Munschauer FE, Benedict RHB.
Psychometrics and normative data for the Multiple Sclerosis Functional Composite: replacing
the PASAT with the symbol digit modalities test. Mult Scler. 2010; 16(2): 228-237.
Evans JJ, Greenfield E, Wilson BA, Bateman A. Walking and talking therapy: improving cognitivemotor dual-tasking in neurological illness. J Int Neuropsych Soc.. 2009; 15: 112-120.
Filippi M, Cercignani M, Inglese M et al. Diffusion tensor magnetic resonance imaging in multiple
sclerosis. Neurology. 2001; 56(3): 304–311.
Filippi M, Iannucci G, Cercignani M et al. A quantitative study of water diffusion in multiple
sclerosis lesions and normal-appearing white matter using echo-planar imaging. Arch Neurol.
2000; 57(7): 1017–1021.
Filippi M, Rocca MA. MR imaging of gray matter involvement in multiple sclerosis: implications
for understanding disease pathophysiology and monitoring treatment efficacy. Am J
Neuroradiol. 2010; 31:1171-1177.
Finlayson ML, Peterson EW, Cho CC. Risk factors for falling among people aged 45 to 90 years
with multiple sclerosis. Arch Phys Med Rehab. 2006; 87(9): 1274-1279.
Fjeldstad C, Pardo G, Bemben D, Bemben M. Decreased postural balance in multiple sclerosis
patients with low disability. Int J Rehabil Res. 2011; 34(1): 53-58.
Fok P, Farrell M, McMeeken J. Prioritizing gait in dual-task conditions in people with Parkinson’s.
Hum Movement Sci. 2010; 29: 831-842.
166
Fok P, Farrell M, McMeeken J. The effect of dividing attention between walking and auxiliary
tasks in people with Parkinson’s disease. Hum Movement Sci. 2012; 31: 236-246.
Foley JA, Cantagallo A, Della Sala S, Logie RH. Dual task performance and post traumatic brain
injury. Brain Injury. 2010; 24(6): 851-858.
Ford-Smith CD, Wyman JF, Elswick RK Jr, Fernandez T, Newton RA. Test-retest reliability of the
sensory organization test in noninstitutionalized older adults. Arch Phys Med Rehabil. 1995;
76(1): 77-81.
Freeman JA, Gear M, Pauli A, Cowan P, Finnigan C, Hunger H, et al. The effect of core stability
training on balance and mobility in ambulant individuals with multiple sclerosis: a multi-centre
series of single case studies. Mult Scler. 2010; 16(11): 1377- 1384.
Fritz NE, Basso DM. Dual-task training for balance and mobility in a person with severe traumatic
brain injury: a case study. J Neurol Phys Ther. 2013; 37: 37-43.
Fritz NE, Worstell A, Siles A, Kloos AD, Kegelmeyer DA. Backward walking parameters are
sensitive to age-related changes in mobility and balance. Gait Posture. 2013; 37: 593-597
Fritz S and Lusardi M. Walking speed: the sixth vital sign. J Geriatr Phys Ther. 2009; 32(2): 2-5.
Frohman, E.M. Multiple sclerosis. Med Clin N Am. 2003; 87(4): 867–897.
Ge Y, Law M, Grossman RI. Applications of diffusion tensor MR imaging in Multiple Sclerosis.
Ann NY Acad Sci. 2005; 1064: 202-219.
Ge Y, Law M, Johnson G et al. Preferential occult injury of corpus callosum in multiple sclerosis
measured by diffusion tensor imaging. J Magn Reson Imaging. 2004; 20(1): 1-7.
Gentile AM. Skill acquisition: Action, movement, and neuromotor processes. In: Carr JH,
Shepherd RB, Gordon J, Gentile AM, Held JM, eds. Movement Science. Foundations for Physical
Therapy in Rehabilitation. 1st ed. Rockville, MD: Aspen Publishers; 1987: 93-154.
Givon U, Zeilig G, Achiron A. Gait analysis in multiple sclerosis: characterization of temporalspatial parameters using GAITRite functional ambulation system. Gait Posture. 2009; 29: 138142.
Goldberg-Zimring D, Mewes AUJ, Maddah M, Warfield SK. Diffusion tensor magnetic resonance
imaging in Multiple Sclerosis. J. Neuroimaging. 2005; S15(4): 68S-81S.
Guo AC, Jewells VL, Provenzale JM. Analysis of normal-appearing white matter in multiple
sclerosis: comparison of diffusion tensor MR imaging and magnetization transfer imaging. Am J
Neuroradiol. 2001; 22(10): 1893–1900.
167
Guo AC, Macfall JR, Provenzale JM. Multiple sclerosis: diffusion tensor MR imaging for
evaluation of normal-appearing white matter. Radiology. 2002; 222(3): 729–736.
Hackney ME, Earhart GM. Backward walking in Parkinson disease. Movement Disord. 2009;
24(2): 218-223.
Hackney ME, Earhart GM. Effects of dance on movement control in Parkinson's disease: a
comparison of Argentine tango and American ballroom. J Rehabil Med. 2009; 41(6):475-481.
Hackney ME, Earhart GM. The effects of a secondary task on forward and backward walking in
Parkinson disease. NeurorehabNeural Re. 2010; 24(1):97-106.
Haggard P, Cockburn J, Cock J, Fordham C, Wade D. Interference between gait and cognitive
tasks in a rehabilitating neurological population. J Neurol Neurosurg Ps. 2000; 69: 479-486.
Hair JF, Anderson RE, Tatham RL, Black WC. Multivariate Data Analysis. 5th ed. New York, NY:
Prentice Hall; 1998.
Hall CD, Echt KV, Wolf SL, Rogers WA. Cognitive and motor mechanisms underlying older adults’
ability to divide attention while walking. Phys Ther. 2011; 91(7): 1039-1050.
Hamilton F, Rochester L, Paul L et al. Walking and talking: an investigation of cognitive-motor
dual tasking in multiple sclerosis. Mult Scler. 2009; 15(10): 1215-1227.
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(5):647–
655.
Hattingen E, Jurcoane A, Melber J, Blasel S, Zanella FE, Neumann-Haefelin T, Singer OC. Diffusion
tensor imaging in patients with adult chronic idiopathic hydrocephalus. Neurosurgery. 2010;
66(5): 917-924.
Hausdorff JM, Edelberg HK, Mitchell SL et al. Increased gait unsteadiness in community-dwelling
elderly fallers. Arch Phys Med Rehab. 1997; 78: 278-283.
Hausdorff JM, Mitchell SL, Firtion R et al. Altered fractal dynamics of gait: reduced stride-interval
correlations with aging and Huntington's disease. J Appl Physiol. 1997; 82(1):262-269.
Hausdorff JM, Rios D, Edelberg HK. Gait variability and fall risk in community-living older adults:
a 1 year prospective study. Arch Phys Med Rehab. 2001; 82: 1050-1056.
Hausdorff JM, Yogev G, Springer S, Simon ES, Giladi N. Walking is more like catching than
tapping: gait in the elderly as a complex cognitive task. Exp Brain Res. 2005; 164: 541-548.
168
Heiervang E, Behrens TEJ, Mackay CE, Robson MD, Johansen-Berg H. Between session
reproducibility and between subject variability of diffusion MR and tractography measures.
Neuroimage. 2006; 33: 867-877.
Herbert RD, Bø K. Analysis of quality interventions in systematic reviews. Brit Med J. 2005;
331:507-509.
Hoffmann S, Tittgemeyer M, Yves von Cramon D. Cognitive impairment in multiple sclerosis.
Curr Opin Neurol. 2007; 20: 275-280.
Ibrahim I, Tintera J, Skoch A, Jirů F, Hlustik P, Martinkova P, Zvara K, Rasova K. Fractional
anisotropy and mean diffusivity in the corpus callosum of patients with multiple sclerosis: the
effect of physiotherapy. Neuroradiology. 2011; 53(11): 917-926.
Jackson WT, Novack TA, Dowler RN. Effective serial measurement of cognitive orientation in
rehabilitation: the Orientation Log. Arch Phys Med Rehabil. 1998; 79: 718-720
Jiang J, Gao G, Li W, et al. Early indicators of prognosis in 846 cases of severe traumatic brain
injury. J Neurotrauma. 2002; 19(7): 869-874.
Johansen-Berg H, Behrens TEJ. Diffusion MRI: From Quantitative Measurement to In Vivo
Neuroanatomy. London: Academic Press; 2009.
Johansen-Berg H, Matthews PM. Attention to movement modulates activity in sensor-motor
areas, including primary motor cortex. Exp Brain Res. 2002; 142: 13-24.
Jones DK. Gaussian modeling of the diffusion signal. In Johansen-Berg H, Behrens TEJ. Diffusion
MRI: from quantitative measurement to in vivo neuroanatomy. 1st ed. London, UK: Elsevier Inc;
2009:37-54.
Kalron A, Dvir Z, Achiron A. Walking while talking—difficulties incurred during the initial stages
of multiple sclerosis disease process. Gait Posture. 2010; 32(3) 332-335.
Katz, DI, White DK, Alexander MP, Klein RB. Recovery of ambulation after traumatic brain injury.
Arch Phys Med Rehabil. 2004; 85: 865-869.
Kelleher KJ, Spence W, Solomonidis S, Apatsidis D. The characterisation of gait patterns of
people with multiple sclerosis. Disabil Rehabil. 2010; 32(15): 1242-1250.
Kempermann G, Fabel K, Ehninger D, Babu H, Leal-Galicia P, Garthe A, Wolf SA. Why and how
physical activity promotes experience-induced brain plasticity. Frontiers Neurosci. 2010; 4(189):
1- 9.
169
Kern KC, Sarcona J, Montag M, Giesser BS, Sicotte NL. Corpus callosal diffusivity predicts motor
impairment in relapsing-remitting multiple sclerosis: a TBSS and tractography study.
Neuroimage. 2011; 55(3): 1169-1177.
Kerrigan DC, Todd MK, Della CU, Lipsitz LA, Collins JJ. Biomechanical gait alterations independent
of speed in the healthy elderly: evidence for specific limiting impairments. Arch Phys Med
Rehabil. 1998; 79(3):317-322.
Kim E, Bijlani M. A pilot study of Quetiapine treatment of aggression due to traumatic brain
injury. J Neuropsychiatry Clin Neurosci. 2006; 18(4): 547-549.
Klingberg T. Concurrent performance of two working memory tasks: potential mechanisms of
interference. Cereb Cortex. 1998; 8(7):593-601.
Koyama T, Marumoto K, Miyake H, Domen K. Relationship between diffusion tensor fractional
anisotropy and motor outcome in patients with hemiparesis after corona radiate infarct. J Stroke
Cerebrovasc. 2013a; epub ahead of print: http://www.ncbi.nlm.nih.gov/pubmed/23510690.
Koyama T, Tsuji M, Nishimura H, Miyake H, Ohmura T, Domen K. Diffusion tensor imaging for
intracerebral hemorrhage outcome prediction: comparison using data from the corona
radiata/internal capsule and the cerebral peduncle. J Stroke Cerebrovasc. 2013b; epub ahead of
print: http://www.ncbi.nlm.nih.gov/pubmed/21795065.
Kramer AF, Larish JF, Strayer DL. Training for attentional control in dual task settings: a
comparison of young and old adults. J Exp Psychol- Appl. 1995;1:50-76.
Kraus MF. Traumatic Brain Injury: A Brief Overview of Traumatic Injuries and the
Neurobehavioral Deficits That Can Occur. Urbana, Illinois: University of Illinois Press; 2007.
Lajoie Y, Teasdale N, Bard C, Fleury M. Attentional demands for static and dynamic equilibrium.
Exp Brain Res. 1993; 97:139-144.
Lange RT, Iverson GL, Zakrzewski MJ, Ethel-King PE, Franzen MD. Interpreting the trail making
test following traumatic brain injury: comparison of traditional time scores and derived indices. J
Clin Exp Neuropsychol. 2005; 27: 897-906.
Laufer Y. Age- and gender-related changes in the temporal-spatial characteristics of forwards
and backwards gaits. Physiotherapy Res Int. 2003;8(3):131-142.
Laufer Y. Effect of age on characteristics of forward and backward gait at preferred and
accelerated walking speed. J Gerontol A Biol Sci Med Sci. 2005;60(5):627-632.
Laughton CA, Slavin M, Katdare K et al. Aging, muscle activity, and balance control: physiologic
changes associated with balance impairment. Gait Posture. 2003; 18(2):101-108.
170
Leow AD, Zhan L, Zhu S, Hageman N, Chiang MC, Barysheva M, Toga AW, Thompson PM. White
matter integrity measured by fractional anisotropy correlates poorly with actual individual fiber
anisotropy. I S Biomed Imaging. 2009. Available at: www.loni.ucla.edu/~thompson/PDF/LeowISBI09.pdf
Lin X, Tench CR, Morgan PS, Constantinescu CS. Use of combined conventional and quantitative
MRI to quantify pathology related to cognitive impairment in multiple sclerosis. J Neuro
Neurosurg Ps. 2008; 79: 437-441.
Linacre JM, Heinemann AW, Wright BD et al. The structure and stability of the functional
independence measure. Arch Phys Med Rehabil.1994; 75(2): 127-132.
Lowe MJ, Horenstein C, Hirsch JG, Marrie RA, Stone L, Bhattacharyya PK, Gass A, Phillips MD.
Functional pathway-defined MRI diffusion measures reveal increased transverse diffusivity of
water in Multiple Sclerosis. Neuroimage. 2006; 32: 1127-1133.
Madden DJ, Bennett IJ, Song AW. Cerebral white matter integrity and cognitive aging:
contributions from diffusion tensor imaging. Neuropsychol Rev. 2009; 19: 415- 435.
Mahar CG, Sherrington C, Herbert RD, et al. Reliability of the PEDro scale for rating quality of
randomized controlled trials. Phys Ther. 2003; 83:713-721.
Maki BE. Gait variability in older adults: predictors of falls or indicators of fear. J Am Geriatr Soc.
1997; 45: 313-320.
Maki BE, McIlroy WE. Control of rapid limb movements for balance recovery: age-related
changes and implications for fall prevention. Age Ageing. 2006;35 Suppl 2:ii12-ii18.
Marcell TJ. Sarcopenia: causes, consequences, and preventions. J Gerontol A Biol Sci Med Sci.
2003; 58(10):M911-M916.
Marigold DS, Eng JJ, Tokuno CD, Donnelly CA. Contribution of muscle strength and integration of
afferent input to postural stability in persons with stroke. Neurorehabil Neural Repair. 2004;
18(4): 222-229.
Martin Cl, Phillips BA, Kilpatric TJ, Butzkeuven H, Tubridy N, McDonald E, Galea MP. Gait and
balance impairment in early multiple sclerosis in the absence of clinical disability. Mult Scler.
2006; 12: 620-628.
Martin KL, Blizzard L, Wood AG, Srikanth V, Thomson R, Sanders LM, Callisaya ML. Cognitive
function, gait and gait variability in older people: a population-based study. J Gerentol A Biol Sci
Med Sci. 2013; 68(6): 726-732.
Mateer CA, Kerns KA, Eso KL. Management of attention and memory disorders following
traumatic brain injury. J Learn Disabil. 1996; 29: 618-632.
171
Mathias JL, Wheaton P. Changes in attention and information-processing speed following severe
traumatic brain injury: a meta-analytic review. Neuropsychology. 2007; 21(2):212-223.
McCulloch K. Attention and dual-task conditions: physical therapy implications for individuals
with acquired brain injury: J Neurol Phys Ther. 2007; 31: 104-118.
McCulloch KL, Buxton E, Hackney J, Lowers S. Balance, attention, and dual-task performance
during walking after brain injury: association with falls history. J Head Trauma Rehab. 2010;
25(3): 155-163.
McDowd JM. An overview of attention: behavior and brain. J Neurol Phys Ther. 2007; 31: 98103.
McIlroy WE, Maki BE. Age-related changes in compensatory stepping in response to
unpredictable perturbations. J Gerontol A Biol Sci Med Sci. 1996;51(6):M289-M296.
Medley A, Thompson M, French J. Predicting the probability of falls in community dwelling
person with brain injury: a pilot study. Brain Inj. 2006; 20 (13-14): 1403-1408.
Meythaler JM, Depalma L, Devivo MJ et al. Sertraline to improve arousal and alertness in severe
traumatic brain injury secondary to motor vehicle crashes. Brain Inj. 2001; 15(4): 321-331.
Mirelman A, Maidan I, Herman T, Deutsch JE, Giladi N, Hausdorff JM. Virtual reality for gait
training: can it induce motor learning to enhance complex walking and reduce fall risk in
patients with Parkinson’s disease? J Gerentol A Biol Sci Med Sci. 2011; 66A(2): 234-240.
Miyake A, Friedman NP, Emerson MJ, Witzki AH, Howerter A, Wager TD. The unity and diversity
of executive functions and their contributions to complex ‘frontal lobe’ tasks: A latent variable
analysis. Cog Psych . 2000;41:49-100.
Mori S, Zhang J. Principles of diffusion tensor imaging and its application to basic neuroscience
research. Neuron. 2006; 51: 527-539.
Moseley AM, Lanzarone S, Bosman JM, et al. Ecological validity of walking speed assessment
after traumatic brain injury: a pilot study. J Head Trauma Rehabil. 2004; 19(4): 341-348.
Mossberg KA, Orlander EE, Norcross JL. Cardiorespiratory capacity after weight-supported
treadmill training in patients with traumatic brain injury. Phys Ther. 2008; 88(1): 77-87.
Mulder T, Nienhuis B, Pauwels J. The assessment of motor recovery: a new look at an old
problem. J Electromyogr Kinesiol. 1996;6:137–45.
172
Mysiw WJ, Bogner JA, Arnett JA et al. The orientation group monitoring system for measuring
duration of posttraumatic amnesia and assessing therapeutic interventions. J Head
Trauma Rehabil. 1996; 14(6):1-8.
Nakamura T, Megura K, Sasaki H. Relationship between falls and stride length variability in senile
dementia of the Alzheimer type. Gerontology. 1996; 42: 108-113.
Negahban H, Mofateh R, Arastoo AA, Mazaheri M, Yazdi MJS, Salavati M, Majdinasab N. The
effects of cognitive loading on balance control in patients with multiple sclerosis. Gait Posture.
2011; 34(4): 479-484.
Nelson AJ, Zwick D, Brody S, Doran C, Pulver L, Rooz G, Sadownick M, Nelson R, Rothman J. The
validity of the GaitRite and the functional ambulation performance scoring system in the
analysis of Parkinson gait. Neurorehabilitation. 2002;17:255–262.
Newstead AH, Hinman MR, Tomberlin JA. 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.
Nilsagård Y, Denison E, Gunnarsson LG, Bostrom K. Factors perceived as being related to
accidental falls in persons with multiple sclerosis. Disabil Rehabil. 2009; 31: 1301-1310.
Nilsagård Y, Lundholm C, Denison E, Gunnarsson LG. Predicting accidental falls in people with
multiple sclerosis—a longitudinal study. Clin Rehabil. 2009; 23: 259-269.
Nilsagård Y, Lundholm C, Gunnarsson LG, Denison E. Clinical relevance using timed walk tests
and ‘timed up and go‘ testing in persons with multiple sclerosis. Physiotherapy Res Int. 2007;
12(2):105-114.
Noseworthy JH, Lucchinetti C, Rodriguez M, Weinshenker BG. Medical progress: multiple
sclerosis. New Engl J Med. 2000;343:938–952.
O’Shea S, Morris ME, Iansek R. Dual task interference during gait in people with Parkinson
disease: effect of motor versus cognitive secondary tasks. Phys Ther 2002;82:888–97.
Parmenter BA, Weinstock-Guttman B, Garg N, Munschauer F, Benedict RHB. Screening for
cognitive impairment in multiple sclerosis using the symbol digit modality test. Mult Scler. 2007;
13: 52-57.
Parker TM, Osternig LR, Lee HJ et al. The effect of divided attention on gait stability following
concussion. Clin Biomech. 2005; 20: 389-395.
Paul SS, Ada L, Canning CG. Automaticity of walking—implications for physiotherapy practice.
Phys Ther Reviews. 2005; 10: 15-23.
173
PEDro statisitics. PEDro Physiotherapy Evidence Database. 2012. Available at:
http://www.pedro.org.au/english/downloads/pedro-statistics/. Accessed October 15, 2012.
Pellecchia GL. Dual-task training reduces impact of cognitive task on postural sway. J Mot Behav.
2005; 37(3): 239-246.
Plummer-D’Amato P, Altmann LJP, Saracino D, Fox E, Behrman AL, Marsiske M. Interactions
between cognitive tasks and gait after stroke: a dual task study. Gait Posture. 2008; 27: 683-688.
Podsiadlo D, Richardson S. The timed ―Up & Go‖: a test of basic functional mobility for frail
elderly persons. J Am Geriatr Soc. 1991; 39: 142-148.
Porosinska A, Pierzchala K, Mentel M, Karpe J. Evaluation of postural balance control in patients
with multiple sclerosis—effect of different sensory conditions and arithmetic task execution. A
pilot study. Neurologia i Neurochirurgia Polska. 2010; 44:35–42.
Prakash RS, Snook EM, Motl RW, Kramer AF. Aerobic fitness is associated with gray matter
volume and white matter integrity in multiple sclerosis. Brain Res. 2010; 1341: 41- 51.
Protas EJ, Mitchell K, Williams A, Qureshy H, Caroline K, Lai EC. Gait and step training to reduce
falls in Parkinson's disease. NeuroRehabilitation. 2005; 20(3):183-190.
Putnam MC, Steven MS, Doron KW, Riggall AC, Gazzaniga MS. Cortical projection topography of
the human splenium: hemispheric asymmetry and individual differences. J Cognitive Neurosci.
2009; 22(8): 1662-1669.
Rábago CA and Wilken JM. Application of a mild traumatic brain injury rehabilitation program in
a virtual reality environment: a case study. J Neurol Phys Ther. 2011; 35: 185-193.
Rancho Los Amigos National Rehabilitation Center. Family Guide to the Rancho Levels of
Cognitive Functioning. Available from: http://www.rancho.org/research/bi_cognition.pdf.
Accessibility verified July 12, 2012.
Rao SM. Cognitive function in patients with multiple sclerosis: impairment and treatment. Int J
Mult Scler Care. 2004; 1: 9-22.
Rao SM, Leo GJ, Bernardin L, Unverzagt F. Cognitive dysfunction in multiple sclerosis I:
frequency, patterns, and prediction. Neurology. 1991; 41(5): 685-691.
Reich DS, Zackowski KM, Gordon-Lipkin EM, Smith SA, Chadkowski BA, Cutter GR, Calabresi PA.
Corticospinal tract abnormalities are associated with weakness in multiple sclerosis. Am J
Neuroradiol. 2008; 29: 333-339.
Rocca MA, Cercignani M, Iannucci G et al. Weekly diffusion-weighted imaging of normalappearing white matter in MS. Neurology. 2001; 55(6): 882–884.
174
Rose J, Mirmiran M, Butler EE, Lin CY, Barnes PD, Kermoian R, et al. Neonatal microstructural
development of the internal capsul on diffusion tensor imaging correlates with severity of gait
and motor deficits. Dev Med Child Neurol. 2007; 49(10): 745-750.
Roskell C, Cross V. Attention limitation and learning in physiotherapy. Physiotherapy
1998;84:118–25.
Royall DR, Lauterbach EC, Cummings JL, et al. Executive control function: a review of its promise
and challenges for clinical research—a report from the Committee on Research of the American
Neuropsychiatric Association. Journal of Neuropsych Clin N. 2002; 14: 377-405.
Saneda DL, Corrigan JD. Predicting clearing of post-traumatic amnesia following closed head
injury. Brain Inj. 1992; 6(2): 167-174.
Sbordone RJ. The executive functions of the brain. In Groth-Marnat G, editor.
Neuropsychological assessment in clinical practice: a guide to test interpretation and integration.
New York: John Wiley and Sons; 2000: 437-456.
Schaafsma JD, Giladi N, Balash Y et al. Gait dynamics in Parkinson‘s disease: relationship to
Parkinsonian features, falls and response to levadopa. J Neurol Sci. 2003; 212: 47-53.
Schwenk M, Zieschang T, Oster P, Hauer K. Dual-task performances can be improved in patients
with dementia: a randomized controlled trial. Neurology. 2010; 74: 1961-1968.
Segev-Jacubovski O, Herman T, Yogev-Seligmann G, Mirelman A, Giladi N, Hausdorff JM. The
interplay between gait, falls and cognition: can cognitive therapy reduce fall risk? Exp Rev
Neurotherapeutics. 2011; 11(7): 1057-1075.
Sheridan PL, Solomont J, Kowall N, Hausdorff JM. Influence of executive function on locomotor
function: divided attention increases gait variability in Alzheimer’s disease. J Am Geriatr Soc.
2003; 51: 1633-1637.
Shigematsu R, Okura T. A novel exercise for improving lower-extremity functional fitness in the
elderly. Aging Clin Exp Res. 2006;18(3): 242-248.
Shumway-Cook A, Brauer S, Woollacott M. Predicting the probability for falls in communitydwelling older adults using the timed up & go test. Phys Ther. 2000; 80(9): 896-903.
Silsupadol P, Shumway-Cook A, Lugade V, vonDonkelaar P, Chou LS, Mayr U, Woollacott MH.
Effects of single-task versus dual-task training on balance performance in older adults: a doubleblind, randomized controlled trial. Arch Phys Med Rehabil. 2009; 90: 381- 387.
Silsupadol P, Siu KC, Shumway-Cook A, Woollacott M. Training of balance under single- and dualtask conditions in older adults with balance impairments. Phys Ther. 2006; 86: 269-281.
175
Smithson F, Morris ME, Iansek R. Performance on clinical tests of balance in Parkinson's disease.
Phys Ther. 1998; 78(6):577.
Smulders K, van Swigchem R, de Swart BJM, Geurts ACH, Weerdesteyn V. Community-dwelling
people with chronic stroke need disproportionate attention while walking and negotiating
obstacles. Gait Posture.2012; 36: 127-132.
Song SK, Sun SW, Ramsbottom MJ, Chang C, Russell J, Cross AH. Dysmyelination revealed
through MRI as increased radial (but unchanged axial) diffusion of water. Neuroimage. 2002; 17:
1429-1436.
Sosnoff JJ, Weikert M, Dlugonski D, Motl RW. Quantifying gait impairment in multiple sclerosis
using GAITRite technology. Gait Posture. 2011; 34: 145-147.
Sosnoff JJ, Socie MJ, Sandroff BM, Balantrapu S, Suh Y, Pula JH et al. Mobility and cognitive
correlates of dual-task cost of walking in persons with multiple sclerosis. Disabil Rehabil. 2013;
epub ahead of print: http://www.ncbi.nlm.nih.gov/pubmed/23597000.
Soukup VM, Ingram F, Grady JJ, Schiess MC. Trail Making Test: issues in normative data
selection. Appl Neuropsychol. 1998; 5(2): 65-73.
Springer S, Giladi N, Peretz C, Yogev G, Simon ES, Hausdorff JM. Dual-tasking effects on gait
variability: the role of aging, falls and executive function. Movement Disord. 2006; 21(7): 950957.
Steffen T, Seney M. Test-retest reliability and minimal detectable change on balance and
ambulation tests, the 36-item short-form health survey, and the United Parkinson Disease
Rating Scale in people with Parkinsonism. PhysTher. 2008; 88(6): 733-746.
Stevenson TJ. Detecting change in patients with stroke using the Berg Balance Scale. Aust J
Physiother. 2001; 47; 29-38.
Strauss E, Sherman EMS, Spreen O. A compendium of neuropsychological tests: administration,
norms, and commentary. 3rd ed. New York: Oxford University Press; 2006.
Summers MM, Fisniku LK, Anderson VM, Miller DH, Cipolotti L, Ron M. Cognitive impairment in
relapsing-remitting multiple sclerosis can be predicted by imaging performed several years
earlier. Mult Scler. 2008; 14: 197-204.
Sveistrup H, McComas J, Thornton M, Marshall S, Finestone H, McCormick A, Babulic K, Mayhew
A. Experimental studies of virtual reality-delivered compared to conventional exercise programs
for rehabilitation. Cyberpsychol Behav. 2003; 6(3): 245-249.
Thomas MA, Fast A. One step forwards and two steps back: the dangers of walking backward in
therapy. Am J Phys Med Rehab. 2000; 79(5): 459-461.
176
Tilson JK, Sullivan KJ, Cen SY, et al. Meaningful gait speed improvement during the first 60 days
poststroke: minimal clinically important difference. Phys Ther. 2010; 90(2): 196-208.
Tinetti ME. Performance-oriented assessment of mobility problems in elderly patients. J Am
Geriatr Soc. 1986; 34(2):119-126.
Tournier JD, Mori S, Leemans A. Diffusion Tensor Imaging and Beyond. Magn Res Med. 2011; 65:
1532-1556.
Toyokura M, Muro I, Komiya T, Obara M. Activation of pre-supplementary motor area (SMA)
and SMA proper during unimanual and bimanual complex sequences: an analysis using
functional magnetic resonance imaging. J Neuroimaging. 2002; 12: 172-178.
Tseng IJ, Jeng C, Yuan RY. Comparisons of forward and backward gait between poorer and better
attention capabilities in early Parkinson’s disease. Gait Posture. 2012; 36(3): 367-371.
Udupa JK, Samarasekera S. Fuzzy Connectedness and Object Definition: Theory, Algorithms, and
Application in Image Segmentation. Graphical Models and Image Processing. 1996; 58:246-261.
Vallat-Azouvi C, Pradat-Diehl P, Azouvi P. Rehabilitation of the central executive of working
memory after severe traumatic brain injury: two single-case studies. Brain Inj. 2009; 23(6): 585594.
Vallée M, McFadyen BJ, Swaine B, et al. Effects of environmental demands on locomotion after
traumatic brain injury. Arch Phys Med Rehabil. 2006; 87: 806-813.
VanSwearingen JM, Paschal KA, Bonino P, Chen TW. Assessing recurrent fall risk of communitydwelling, frail older veterans using specific tests of mobility and the physical performance test of
function. J Gerontol A Biol Sci Med Sci. 1998; 53(6):M457-M464.
VanLoo MA, Moseley AM, Bosman JM et al. Test-re-test reliability of walking speed, step length
and step width measurement after traumatic brain injury: a pilot study. Brain Inj.. 2004; 18(10):
1041-1048.
van Tulder M, Furlan A, Bombardier C, et al. Updated method guidelines for systematic reviews
in the Cochrane collaboration back review group. Spine (Phila Pa 1976). 2003; 28: 1290-1299.
Venema DM, Bartels E, Siu KC. Tasks matter: a cross-sectional study of the relationship of
cognition and dual-task performance in older adults. J Geriatr Phys Ther. 2012; epub ahead of
print. http://www.ncbi.nlm.nih.gov/pubmed/23249724
Verghese J, Buschke H, Viola L, Katz M, Hall C, Kuslansky G, Lipton R. Validity of divided attention
tasks in predicting falls in older individuals: a preliminary study. J Am Geriatr Soc. 2002; 50:
1572-1576.
177
Wang DM, Li J, Liu JR, Hu HY. Diffusion tensor imaging predicts long-term motor functional
outcome in patients with acute supratentorial intracranial hemorrhage. Cerebrovasc Dis. 2012;
34(3): 199-205.
Webster KE, Wittwer JE, Feller JA. Validity of the GAITRite walkway system for the measurement
of averaged and individual step parameters of gait. Gait Posture. 2005; 22(4):317-321.
Whetten-Goldstein K, Sloan FA, Goldstein LB, Kulas ED. A comprehensive assessment of the cost
of multiple sclerosis in the United States. Mult Scler. 1998; 4: 419-425.
Wilson M, Trench CR, Morgan PS, Blumhardt LD. Pyramidal tract mapping by diffusion tensor
magnetic resonance imaging in multiple sclerosis: improving correlations with disability. J Neuro
Neurosurg Ps. 2003; 74: 203-207.
Wolfson LI. Gait and balance dysfunction: a model of the interaction of age and disease.
Neuroscientist. 2001; 7:178-183.
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.
Yang YR, Wang RY, Chen YC, Kao MJ. Dual-task exercise improves walking ability in chronic
stroke: a randomized controlled trial. Arch Phys Med Rehabil. 2007; 88: 1236- 1240.
Yang YR, Yen JG, Wang RY, Yen LL, Lieu FK. Gait outcomes after additional backward walking
training in patients with stroke: a randomized controlled trial. Clin Rehabil. 2005; 19(3):264-273.
Yardley L, Gardner M, Bronstein A, Davies R, Buckwell D, Luxon L. Interference between postural
control and mental task performance in patients with vestibular disorder and healthy controls. J
Neurol Neurosurg Ps. 2001;71:48–52.
Yen CY, Lin KH, Hu MH, Wu RM, Lu TW, Lin CH. Effects of virtual reality-augmented balance
training on sensory organization and attentional demand for postural control in people with
Parkinson disease: a randomized controlled trial. Phys Ther. 2011; 91: 862-874.
Yeo SS, Ahn SH, Choi BY, Chang CH, Lee J, Jang SH. Contribution of the pedunculopontine
nucleus on walking in stroke patients. Eur Neurol. 2011; 65: 332-337.
Yogev G, Giladi N, Peretz, Springer S, Simon ES, Hausdorff JM. Dual-tasking, gait rhythmicity, and
Parkinson’s disease: which aspects of gait are attention demanding. Eur J Neurosci.2005; 22:
1248-1256.
Yogev G, Plotnik M, Peretz C, Giladi N, Hausdorff JM. Gait asymmetry in patients with
Parkinson’s disease and elderly fallers: when does the bilateral coordination of gait require
attention? Exp Brain Res. 2007; 177: 336-346.
178
Yogev-Seligmann G, Giladi N, Brozgol M, Hausdorff JM. A training program to improve gait while
dual tasking in patients with Parkinson’s disease: a pilot study. Arch Phys Med Rehabil.2012; 93:
176-181.
Yogev-Seligmann G, Hausdorff JM, Giladi N. Do we always prioritize balance when walking?
Towards an integrated model of task prioritization. Mov Disord. 2012; 27(6): 765-770.
Yogev-Seligmann G, Hausdorff J, Giladi N. The role of executive function and attention in gait.
Mov Disord. 2008; 23(3):329–342.
You JH, Shetty A, Jones T, Shields K, Belay Y, Brown D. Effects of dual-task cognitive-gait
intervention on memory and gait dynamics in older adults with a history of falls: a preliminary
investigation. Neurorehabilitation. 2009; 24: 193-198.
Zivadinov R, Sepcic J, Nasuelli D, De Masi R, Monti Bragadin L, Tommasi MA, Zambito-Marsala S,
Moretti R, Bratina A, Ukmar M, Pozzi-Mucelli RS, Grop A, Cazzato G, Zorzon M. A longitudinal
study of brain atrophy and cognitive disturbances in the early phase of relapsing-remitting
multiple sclerosis. J Neurol Neurosur Ps. 2001; 70: 773-780.
179