|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
TREATMENT DURATION AND EFFICACY OF COUPLED MOTOR RECOVERY
DAVID A. E. BOLTON
A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE IN EXERCISE AND SPORT SCIENCES
UNIVERSITY OF FLORIDA
DAVID A. E. BOLTON
This document is dedicated to my mother and father
I thank my Mother and my good friend Geoff Dover for making this possible. I also
thank my committee members Dr. Dodd and Dr. Janelle, my chair Dr. James Cauraugh
along with Aaron Duley and Dr. Sang Bum Kim for their assistance with data collection
TABLE OF CONTENTS
A C K N O W L E D G M E N T S ................................................................................................. iv
LIST OF TABLES .............. ....................................... ...... ............. vii
LIST OF FIGURES ...................................................... ................... viii
A B STR A C T ..................... ................................... ........... ................. viii
1 INTRODUCTION ............... ................. ........... ................. ... ..... 1
2 REVIEW OF LITERATURE ......................................................... .............. 4
Stroke: Causes and Classification..................... ....... ........................... 4
N eurological R esponses to Stroke ........................................................... .................. 7
Treatment Induced Reorganization: Principles Governing Rehabilitation.................9
Pharm ecological Interventions .................................................. ............... .... 12
Electrical Stimulation as a Means of Treatment for Motor Recovery........................13
Neuromuscular Stimulation Involving Movement...............................................18
Application and Underlying Principles of Electrical Stimulation ............................19
Volitionally Activated Stimulation.................................................... ............... 21
Searching for an Optimal Treatment Duration..............................................25
W hy Treatm ent D uration? .................................................. ............................ 26
The Selection of Time Periods for Comparison ............................... ................27
Bilateral Movements Coupled with Active Stimulation.......................................28
3 M E T H O D S ...................... .. ............. .. ....................................................3 1
S u b j e c ts .... ... ...... ......... .. ............. .. ..........................................................3 1
Inclusion Criteria .................... .................... ........ .......32
Experim mental D design ........................................ .... ....... ..... ...... 33
Instrum entation and Procedures ........................................ ........................... 33
T raining P protocol ................................................................3 5
D ata R education ................. .................................... .. ........ .. .............36
D ata A n aly sis ............................. ....................................................... ............... 3 7
4 R E SU L T S ....................................................... 3 8
R action T im e.................. .... ...................38
Sustained M uscle Contraction Task ........................................ ........................ 41
B ox and B locks Test ......................................................... .. .... .... .... 44
5 DISCU SSION AND CON CLU SION S ............................................. ....................45
L IST O F R E F E R E N C E S ........................................................................ .. ....................58
B IO G R A PH IC A L SK E T C H ...................................................................... ..................63
LIST OF TABLES
1 Reaction Time and Fractionated Components (Unilateral Condition)...................39
2 Reaction Time and Fractionated Components (Bilateral Condition)....................41
3 Sustained Muscle Contraction Task Results (root mean square error)....................43
4 Box and Block Test Results (number of blocks moved by the impaired arm in 60
se co n d s) ...................... .. .. ......... .. .. .................................................4 4
LIST OF FIGURES
1 E xperim mental D design ..................................................................... ..... ................33
2 Total reaction tim e ........................... ..................... .......... .....40
3 P re-m otor reaction tim e......................................... .............................................40
4 M otor reaction tim e ......................................................... ..... ..........41
5 Sustained Muscle Contraction Task.....................................................................42
6 B ox and B locks results ............................................................................. ... ........43
Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science in Exercise and Sport Sciences
TREATMENT DURATION AND EFFICACY OF COUPLED MOTOR RECOVERY
David A. E. Bolton
Chair: James Cauraugh
Major Department: Exercise and Sport Sciences
Motor impairment is a major consequence following stroke. Recovery of
movement, particularly in the upper extremity, poses a challenge for rehabilitation
specialists and therefore the search to find effective modes of intervention continues.
EMG-triggered neuromuscular stimulation studies have demonstrated functional
improvements; however, the application of this treatment has lacked standardization. One
of the major issues needing resolution is treatment duration of coupled protocols
(bilateral movements and active stimulation).
The present study addressed treatment duration by comparing motor function over
the course of a 4-week (12 hours) treatment period. Training involved EMG-triggered
neuromuscular stimulation on the impaired limb coupled with concurrent extension of the
unimpaired wrist and fingers. All participants were trained twice per week on average for
120 stimulation trials per session for a total duration of approximately 4 weeks (12
Motor functions were measured with the Box and Blocks timed manipulation task,
as well as with two computerized measures of wrist extensor capabilities (reaction time
and sustained contraction). Baseline testing was conducted prior to training, after 2 weeks
(6 hours) of training, and following 4 weeks (12 hours) of training.
The findings of this study failed to reveal duration differences for EMG-triggered
neuromuscular stimulation coupled with bilateral movements. Thus, no motor recovery
inferences could be made regarding treatment duration. These results are incongruent
with the current consensus on EMG-triggered stimulation and bilateral movements.
Possible reasons for such differences are discussed as well as an expanded critical review
of stroke motor recovery research as it relates to the influence of treatment duration on
Recently, the influence of electrical stimulation on regaining motor function in the
upper limb of stroke patients has been researched by several investigators (Cauraugh &
Kim, 2002; Cauraugh et al., 2000; Chae et. al., 1998). This evidence suggests that
neuromuscular electrical stimulation enhances motor relearning. In addition, electrical
stimulation has been proposed as a therapeutic strategy for changes in muscle strength
and spasticity, which in turn may positively influence motor relearning (Burridge &
Ladouceur, 2001). Research has focused on both extrinsic and intrinsic measures of
change that take place with active stimulation. Extrinsic measures focus on the functional
effect of electrical stimulation in the ability to carry out various daily activities, and
intrinsic measures offer insight into the underlying mechanisms for change (Burridge &
Ladouceur, 2001). However, limitations of previous research have been noted in terms of
the scientific thoroughness of many of the studies (Burridge & Ladouceur, 2001; Glanz et
al., 1996). This includes a variety of problems in methodology such as the absence of a
placebo, uneven group numbers, the lack of a control group, and groups that were
significantly different in baseline level of function. Moreover, some of the research
evidence is based on simple case studies (Chae et al., 2001).
Previous studies have used a wide variety of stimulus parameters (i.e., intensity of
stimulation, duration of pulse), treatment duration, treatment frequency, and subjects (i.e.,
various degrees of impairment and time period since stroke). This lack of standardization
in treatment administration has made it difficult to accurately compare and contrast
findings. Along with the variability in treatment protocols, different investigators have
used different outcome measures. Examples of outcome measures include simple range
of motion and strength testing as well as complex functional assessments such as the
functional independence measure (FIM). Such a broad range can be somewhat
problematic when trying to compare the efficacy of the treatment among studies.
Despite discrepancies in how various researchers have administered and measured
the effect of neuromuscular stimulation, overall significant improvements in voluntary
motor control have been found (Burridge & Ladouceur, 2001; Popovic et al., 2002). Such
improvements have been especially pronounced in studies using stimulation at or above
the motor threshold, and to an even greater extent in studies requiring volitional
activation of the neuromuscular stimulation (Burridge & Ladouceur, 2001; Francisco et
al, 1998; Popovic et al., 2002)
Granted, benefits from neuromuscular stimulation abound especially with volitional
activation, however, optimal treatment duration has not been determined. Cauraugh and
colleagues have found motor improvements in six hours of training over two weeks
(Cauraugh & Kim, 2002; Cauraugh et al., 2000). Dimitrijevic and colleagues reported
motor recovery progress after ten months (Dimitrijevic et al., 1996). These are two
extreme treatment durations. Further, no studies have determined the time course of
effectiveness of any motor improvements. Given that a treatment program of two weeks
has led to significant improvements the question arises of whether additional treatment
sessions provide increased benefit. The purpose of the present study was to determine an
optimal treatment duration and this is consistent with conclusions in a recent critical
review of neuromuscular stimulation; optimal treatment duration has not been determined
(Popovic et al., 2002).
The specific objective of the present study was to determine if motor capabilities of
chronic stroke patients improve as a function of the duration of EMG-triggered
neuromuscular stimulation when combined with bilateral neuromuscular movement
training. In particular, does treatment duration of four weeks (12 hours) lead to
significantly improved functional outcomes when compared to patients treated for only
two weeks (6 hours). Based on the motor learning and control precept that additional
practice would lead to enhanced motor capabilities (Winstein, Wing & Whitall, 2003), it
was hypothesized that the four weeks (12 hours) of treatment would demonstrate
significantly increased motor functions in the upper extremity (wrist/fingers) compared
with the two weeks (6 hours) of treatment.
REVIEW OF LITERATURE
Stroke: Causes and Classification
In general, a stroke involves the disturbance of blood supply to the brain. Due to
the brain's dependence on an adequate supply of oxygen and nutrients via the blood,
brain function can only be maintained for a few minutes if blood flow is reduced below a
critical level (Kandel, Schwartz, & Jessell 2000). If sustained, this deprivation, leads to
impaired brain cell function and often cell death. A stroke refers to the neurological signs
and symptoms that result from diseases involving the blood vessels (Kandel, Schwartz, &
Jessell 2000). The nature and severity of the impairments that result are directly related to
the location of the brain that is affected as well as the extent of the damage.
There are two general types of stroke: (a) ischemic and (b) hemorrhagic (Kandel,
Schwartz, & Jessell 2000; Popovic & Sinkjaer, 2001). An ischemic stroke involves
occlusion of the brain's supply due to the closure of a vessel (i.e., blood clot). The
hemorrhagic class of stroke involves a vessel rupture and consequent disruption of blood
flow. Both of these forms of stroke lead to the deprivation of blood supply to the brain
tissue, and when sufficiently prolonged and severe, death of neurons and other cellular
elements (Kandel, Schwartz, & Jessell, 2000). This resultant death of the brain cells is
referred to as an infarct and once cell death occurs the damage to this specific site is
irreversible. Moreover, hemorrhagic strokes are more substantial than the ischemic
The factors that lead to stroke are diverse but tend to follow certain patterns. For
example, most ischemic strokes are associated with atherosclerosis and thrombosis, while
most hemorrhagic strokes are associated with hypertension or aneurysms (Kandel,
Schwartz, & Jessell, 2000). In addition, other causes have been noted and include cardiac
disease, trauma, infection, neoplasm, blood dyserasia, vascular malformation,
immunological disorder, and erogenous toxins (Popovic & Sinkjaer, 2001). However,
regardless of the cause, the end result of the stroke is death to some part of the brain and
consequent impairment of function.
The functional decrements that result depend on the specific location affected by
the stroke. The brain is composed of three main areas: the cerebrum, the cerebellum and
the brain stem. Each of these areas is distinct in terms of function. For example, the
cerebellum is responsible for movement control and coordination as well as maintaining
balance. The brain stem is responsible for many automatic functions such as respiration,
heart rate and blood pressure. The cerebrum is the largest part of the brain, and is
responsible for many higher functions such as speech, emotion and fine movement
control. The area or region is divided into a left and right hemisphere, as well as into four
separate lobes including the frontal, parietal, temporal, and occipital lobes. Neural cross-
over exists such that the left side of the brain controls most of the functions on the right
side of the body and vise versa. Because of this crossing over of the neural tracts, a stroke
that occurs in the one hemisphere will result in impairments on the opposite side of the
Each of these parts of the brain receives a blood supply from different arterial
sources. When a certain artery is ruptured or becomes occluded, the resultant infarction
will produce characteristic impairments. Infarction that follows the disruption of blood
flow through the middle cerebral artery is the most common among the various forms of
stroke (Kandel, Schwartz, & Jessell 2000, 2000; Popovic & Sinkjaer, 2001). This artery
supplies part of the frontal lobe and the lateral surface of the temporal and parietal lobes.
The consequent stroke syndrome involves contralateral weakness, sensory loss, and
visual field cut, and depending on which hemisphere is affected, either language
disturbance or impaired spatial perception (Kandel, Schwartz, & Jessell 2000). Weakness
and sensory loss is more profound in the face and arm as opposed to the leg.
The anterior cerebral artery supplies the frontal lobes. Weakness and sensory loss
result when this arterial supply is affected and voluntary movement is compromised,
especially in the distal contralateral leg (Kandel, Schwartz, & Jessell 2000). If the
anterior cerebral arteries on both sides are involved then significant behavioral symptoms
may result (i.e. akinetic mutism, profound apathy, urinary incontinence and motor inertia)
(Kandel, Schwartz, & Jessell 2000; Popovic & Sinkjaer, 2001).
The posterior cerebral arteries supply the temporal and occipital lobes of the left
cerebral hemisphere and the right hemisphere. Depending upon the location of the
occlusion along this arterial branch, the clinical symptoms will vary. Some of the
symptoms may include color blindness, failure to see to-and-fro movements, memory
disturbance, hemisensory loss, verbal dyslexia, and hallucinations (Kandel, Schwartz, &
Jessell 2000; Popovic & Sinkjaer, 2001).
In summary, the manner in which a stroke manifests itself is dependent on the
region that is affected and the extent of the brain area involved. There are several factors
that can lead to a stroke, and several different ways in which a stroke can occur, but the
common result is death of the brain cells that were nourished by the disrupted blood
supply. The resultant impairments vary widely and frequently involve some degree of
motor, sensory, and cognitive impairment.
Neurological Responses to Stroke
Following a stroke, the body undergoes neurological changes to deal with this
novel insult. Neurological and functional recovery occurs with the greatest speed in the
first 1-3 months after stroke (Gresham et al., 1995). As previously stated, the nature and
severity of the deficits depend upon the type, location, and extent of the lesion. The
deficit may be motor, sensory, or cognitive in origin, or any combination. The degree of
recovery from these impairments varies, and the resolution of any one deficit may not
correlate with the improvement of another. In regards to motor ability, the potential
problems include muscle weakness, abnormal synergistic organization of movements,
altered temporal sequencing of muscle contractions, impaired force control regulation,
delayed responses, diminished range of motion, abnormal muscle tone, sensory
impairments, and altered biomechanical alignment (Gresham, Duncan, Stason, et al.,
With the clinical observations of outcome after stroke it is important to elucidate
the underlying mechanisms. In this sense, the focus is on the intrinsic measures that lead
to the observed outcome. For the first several days following a stroke, there is often a
noted improvement in ability. This form of spontaneous recovery may be due to the
reduction of edema or reperfusion of the ischemic penumbra (Chen et al., 2002). The
penumbra is the area surrounding the infarct that is in a fragile state, yet it can be
salvaged if the blood supply is stabilized within a critical period of time. Therefore, the
resolution of these temporary issues may lead to improved function (Chen et al., 2002).
After a period of approximately two weeks a large degree of the recovery that takes
place is because of mechanisms of plasticity (Chen et al., 2002). The type of mechanism
and the degree of plasticity have a high correlation with the degree of recovery that is
noted in patients (Hallett, 2001). The mechanisms that are involved with plasticity
involve unmasking of inhibitory influences on regions surrounding the infarct,
strengthening or weakening of existing synapses (i.e. long-term potentiation or
depression), or alteration of neuronal membrane excitability (Chen et al., 2002; Hallett,
2001). Because of the existence of redundant circuitry in the brain, the possibility exists
that these previously dormant circuits may now be used. These processes work along a
relatively rapid time course. Another mechanism, which requires a longer duration to
manifest, involves anatomical changes such as sprouting of new axon terminals and the
formation of new synapses (Hallett, 2001).
An important point to remember when considering the concept of plasticity and
how it leads to cortical reorganization is that plasticity can take place at multiple sites
throughout the nervous system. This means that reorganization may occur at a cortical
level as well as at the sub-cortical, brainstem and spinal cord levels (Chen et al., 2002).
Moreover, the spinal cord, which was at one time thought of as only a hard-wired and
immalleable system has demonstrated the ability to remodel function in relation to
afferent input (Whelan & Pearson, 1997).
For remodeling of the nervous system to occur there must be some impetus for
change. One possibility is the brain's inherent propensity for resource competition.
Reorganization studies on the motor and sensory cortex following de-afferentation of a
limb reported that the deafferented cortex did not remain unused. Instead, body part
representations that were adjacent to that of the deafferented body part appropriated the
cortical region (Hallett, 2001). Other studies have shown similar results with the brain
displaying an ability to remold itself according to input stimulation or a lack of it (Chen
et al., 2002). Indeed, the enlarged cortical representation of the first dorsal interosseus
muscle in proficient Braille readers is a testament to use-dependent molding of the
nervous system (Hallett, 2001).
Of particular interest to stroke rehabilitation is the issue of how the area
surrounding the damaged tissue can take over some of the function that was lost. This
concept is referred to as peri-infarct reorganization (Abo et al., 2001). The other
mechanisms that may have importance in post-stroke rehabilitation include the activation
of the hemisphere contralateral to the damaged region leading to ipsilateral influence over
the affected limb, as well as plasticity and dendritic branching in the opposite hemisphere
(Abo et al., 2001).
Treatment Induced Reorganization: Principles Governing Rehabilitation
The Clinical Practice Guideline on Post-Stroke Rehabilitation (Gresham et al.,
1995) outlines three major philosophies that direct treatment choices for sensorimotor
deficits: remediation or facilitation, compensation, and motor control. Remediation, is
traditional therapeutic exercise using forced sensory stimulation modalities, exercises and
resistive training to enhance motor recovery. In this form of treatment, some degree of
volitional limb movement is a requisite. Compensation is essentially self-explanatory in
that it implies using means of accomplishing a task other than with the impaired limb
(i.e., using the unaffected limb to perform an activity). Finally, motor control rests on the
premise that a complex interaction exists between the neurological and musculoskeletal
systems and the environment, and that rehabilitation must occur in a task-specific manner
to be effective. Most likely, some combination of the above approaches would be optimal
and the specific recipe for treatment would vary with the type and severity of the
An important consideration when reviewing the variety of treatments for motor
recovery is that the best methods of achieving maximal recovery remain uncertain (Feys
et al., 1998; Gresham et al., 1995; van der Lee et al., 2001). Although several studies
have been implemented that compare some form of intervention with a standard
physiotherapy regimen, overall there is little to indicate one treatment as being superior to
another (Francisco et al., 1998; van der Lee et al., 2001). The issue is further clouded
when one considers the large assortment of ways in which treatments are administered, to
whom and how these interventions are measured. This makes for imprecise comparison.
However, a general framework of knowledge is starting to build and illuminate several
factors that appear important to motor recovery. Thus, at this point, it is appropriate to
present the general basis behind most forms of treatment, and their use to capitalize on
the body's resource for healing and sensorimotor adaptation.
The nervous system is a highly malleable substrate that can be modified by external
input. Post-stroke treatment essentially attempts to capitalize on the inherent capacity for
plasticity and improve beyond spontaneous recovery. Currently, techniques for
rehabilitation include biofeedback, functional electrical stimulation, balance training,
different types of exercise (i.e., isokinetic, active resisted etc.), sensory retraining, and
active neuromuscular stimulation (Gresham et al., 1995).
One of the basic premises behind several rehabilitation techniques is that actual
limb use is vital for re-acquiring and maintaining cortical representation, and hence
function (Braun et al., 2001; Hallett, 2001). This supports the recommendation that
patients with at least some degree of voluntary control of limb movement should be
encouraged to use the limb in functional activities (Gresham et al., 1995; van der Lee et
al., 2001). This reacquisition of function through repeated use can actually be hindered
by many of the forms of treatment where compensation for the impaired limb is the opted
strategy (Nudo, 1999). In this case, the continued disuse of the impaired limb would only
exacerbate the limited ability of the limb (Gresham et al., 1995).
Using activity dependent recovery as a foundation, the practice of constraint-
induced movement therapy was initiated by Taub and colleagues (Taub et al., 1993). The
main tenet of this therapy is that constraining the unaffected limb would force the patients
to use the affected one and therefore lead to improved ability for that limb. A typical
scenario in stroke is the impairment of motor function on one side of the body
contralateral to the affected cortical hemisphere. In the early stages following a stroke, or
any neurological injury, there is a resultant shock to the central nervous system and an
inability to use the affected extremity (Taub et al., 1993). As the individual passes out of
this phase there is a return of capacity, at least in some latent fashion, for motor ability in
the impaired limb. However, a major problem is that the individuals have learned a non-
use pattern, or what is referred to as a learned suppression for the affected body part.
Just as activity tends to lead to a positive influence, inactivity tends to have
deleterious effects on motor ability. In the case where the initial shock prevented
movement, another consideration is how this lack of activation leads to reduced cortical
representations (Byrnes et al., 2001; Johansson, 2000; Rossini et al., 1998). This form of
remodeling of the nervous system has been shown to occur within minutes in studies
involving deafferentation via anesthesetic (Chen et al., 2002). The neural shock is
followed by a learned suppression of movement with consequent reorganization in the
cortex that would amplify the diminished capacity for movement.
Previous research has demonstrated use-dependent cortical re-organization, as well
as functional changes that correlate with such change (Nudo, 1999). Liepert and
colleagues (Liepert et al., 2000) found that the area of excitable cortex almost doubled as
a result of constraint induced movement treatment, and more importantly, this increased
area was paralled by improvements in performance measures.
Moreover, massed practice with the impaired limb has led to a specific pattern of
cortical re-arrangement. Specifically, the adjacent spared cortex tends to take over the
function of the damaged area (Nudo, 1999). In addition to an alteration of the adjacent
spared cortical tissue, changes occur in more remote cortical regions as a means of
adaptation. However, recovery is highly use-dependent. When one body area is rendered
inactive the general trend within the nervous system is for neighboring brain
representations to take over functions for the damaged area (Muellbacher et al., 2002). In
this sense, the proximal muscle representations will often take over the territory
previously occupied by the distal muscles if they are unused. Therefore, the maintenance
of cortical space is contingent on use (Johansson, 2000).
A common theme behind successful treatments for motor recovery in stroke is that
there must be some degree of active involvement (Hallett, 2001; Johansson, 2000; Liepert
et al., 2000). Actively initiating motor actions helps to preserve the cortical territory and
further strengthens the connections within the neural network (Byrnes et al., 2001). In
addition, pharmacological interventions have demonstrated an improved functional
outcome in stroke patients. Pharmacological interventions can exert their influence acting
either alone as plasticity-promoting agents or in concert with physical training to create
an enriched effect (Johansson, 2000).
Drugs that have been shown to aid in recovery work in two basic ways: (a) in the
short term by means of "damage control" and (b) in the long term with plasticity. In the
short term, there have been found several trophic factors that are able to rescue neurons in
the acute stage of a stroke and therefore this can be of benefit during the early ischemic
period (Johansson, 2000). At present, it is unknown if these trophic factors exert any
direct influence during the rehabilitation phase, but there are other drugs such as
amphetamines that have proven effective in long term influences. Norepinephrine,
amphetamines, and other alpha-adrenergic stimulating drugs have demonstrated the
ability to enhance motor performance (Johansson, 2000; Nudo, 1999). Some of the
effects linked with these drugs are increased neural sprouting and syaptogenesis and
coincident increases in motor performance with humans (Nudo, 1999).
Regardless of the isolated effects of these pharmacological agents themselves, the
greatest improvements have been noted when other environmental factors are added. In
this sense, the coupling of drug treatments with directed movement interventions has
been shown to lead to superior gains, over either one used independently. The fact is
noteworthy that the positive influence of amphetamines on motor recovery is not
observed when subjects are prevented from moving their impaired limb (Francisco et al.,
1998). Thus, behavioral intervention is a vital aspect of the optimal recovery protocol.
Electrical Stimulation as a Means of Treatment for Motor Recovery
Of the many treatments that are used to regain motor control in stroke patients,
electrical stimulation therapy has proven effective in a variety of forms. However, there
is no objective data to suggest any one form of rehabilitation is superior to all others
(Gresham et al., 1995; van der Lee et al., 2001). Electrical stimulation as a treatment
involves sending some form of electrical signal into the body for conduction through the
nervous system. The clinical practice guidelines set out by the U.S. Department of Health
and Human Services defines functional electrical stimulation as follows: "Bursts of
electrical stimulation applied to the nerves or muscles affected by stroke, with the goal of
strengthening muscle contraction and improving motor control." (Gresham et al.,1995).
This signal can be administered either through direct nerve contact or by means of
cutaneous electrodes. Furthermore, this signal can take on a variety of pulse durations,
intensities and other characteristics that may lead to different outcomes.
As an historical overview, electrical stimulation has been used to map certain
aspects of the nervous system such as the topographical arrangement of body part
representations in the sensory and motor cortices. Additionally, stimulating particular
regions in the brain can create certain motor responses (Velasco, 2000). This ability of
externally applied electrical current to modify brain responses has been used for several
treatment purposes such as regulating seizures and behavior disorders, as well as
modifying the rigidity and tremor that is associated with conditions such as Parkinson's
disease (Velasco, 2000).
Electrical stimulation has been used since the 1960's for functional motor recovery
(Kraft et al., 1992). There are two basic categories under which the clinical application of
electrical stimulation fall: orthotic, and therapeutic (Burridge & Ladouceur, 2001). The
orthotic use of electrical stimulation implies that the stimulation activates the appropriate
muscles (of the hand for example), to accomplish a particular task, without subject
volition. An example of this type of stimulation has been carried out in other studies
where pre-programmed muscle activity sequences from the subject's normal limb are
captured and used as a means of stimulating the affected limb (Kraft et al., 1992). The
therapeutic application of stimulation involves using the electrical current as a means of
creating neuromuscular adaptations in order for the patient to regain independent motor
Perhaps the best known example of orthotic or functional electrical stimulation,
involves gait applications. Favorable effects were reported using this form of stimulation
in regaining the ability to walk in stroke patients and those with traumatic brain injury
(Oostra et al., 1997). The common use for electrical stimulation in this sense is to correct
the problem of spastic footdrop. The stimulation unit is worn as an orthotic device on a
daily basis, and works by stimulating the peroneal nerve to produce appropriately timed
dorsi-flexion of the foot during walking (Oostra et al., 1997). Upper extremity examples
of this form of stimulation can also be found in terms of orthotic grasping units such as
the "Bionic Glove" (Popovic et al., 2002).
Within the realm of therapeutic electrical stimulation there is further breakdown
into different categories. The stimulation intensity may be set at an assortment of levels.
For example, sub-sensory threshold stimulation was used in the study by Golaszewski et
al., and Dimitrijevic et al., as a means of observing the influence of afferent stimulation
that falls under the perceptual limit of the subject (Dimitrijevic et al., 1996; Golaszewski
et al., 1999). Other protocols have used stimulation levels that are within the limits of
perception; however they do not create movement, and are therefore classified as sub-
motor threshold level. Finally, stimulation that is strong enough can create movement at
the joint and would be considered at or above the motor threshold.
All of these forms of stimulation have been shown effective in aiding motor
recovery, however, they tend to exert their influence in slightly different ways. One
commonality amongst the various applications is that some degree of afferent input is
present, whether it is perceived or not. It has been found that effective motor learning and
re-learning require afferent input to the CNS (Burridge & Ladouceur, 2001). Indeed,
following a stroke a relationship exists between damage in the sensory cortex and motor
recovery (Burridge & Ladouceur, 2001).
Golaszewski et al. (Golaszewski et al., 1999), used a mesh glove to provide whole-
hand afferent stimulation on the impaired side. This stimulation was set just below the
conscious sensory threshold and was set to produce tonic, synchronous input to the CNS.
The purpose was to see if this form of stimulation could elicit changes in cortical brain
activity, and the results supported this contention. The benefits of this treatment included
suppression of muscle hypertonia, increased awareness of the affected hand, and the
augmentation of residual voluntary movement. The connection between the improved
performance measures and the alteration in cortical activity, is that this form of
stimulation leads to an increase in the responsiveness of the somatosensory system and
motor cortices during a voluntary motor task. Furthermore, the mesh glove leads to a
modification of some of the abnormally active proprioceptive responses that are
consequent to stroke such as amplified tendon reflexes and resistance to passive stretch
(Dimitrijevic et al., 1996).
In the midst of this spectrum of stimulation types, there is as of yet no evidence to
indicate which one is best, for what situation, or optimal duration (Burridge &
Ladouceur, 2001; Popovic et al., 2002). Although many studies have used a variety of
stimulation parameters and effectiveness has been demonstrated, there is rarely any
comparison made between the treatments. For example, Kraft et al. (Kraft et al., 1992),
separated treatment conditions across rehabilitation centers and compared performance
against a control group. The three treatment conditions were proprioceptive
neuromuscular facilitation (PNF), EMG-triggered stimulation, and bias/balance treatment
(i.e., low intensity stimulation that is just under the motor threshold combined with
voluntary effort to move the limb). The EMG-triggered stimulation involves a volitional
attempt at moving the limb by the subject leading to EMG activity in that muscle. When
the degree of activity reaches a pre-established threshold the stimulation is initiated. The
results of this study showed EMG-triggered stimulation to be the most effective of the
three treatment protocols, however, the advantage was only statistically significant when
compared to the PNF treatment. Although there was a lack of significance in the
difference between the two forms of electrical stimulation, it should be noted that the
sub-motor stimulation that was applied in the B/B group involved the added component
of volitional movement. Therefore the results should not be taken as a measure of the
effectiveness of sub-motor stimulation.
This lack of adequate comparison among various modes of treatment is a problem
throughout stroke rehabilitation in general, with an assortment of interventions used
without any clearly defined parameters (Feys et al., 1998; Gresham et al., 1995). This is
equally true in electrical stimulation where there are many levels of stimulation and even
different ways of administering the same stimulation. Most importantly, the lack of
comparison between the various interventions and the manner in which they are applied
hinders the elucidation of an optimal treatment strategy. This is perhaps exemplified best
in the unresolved issue of treatment duration.
Neuromuscular Stimulation Involving Movement
With the knowledge that the nervous system is constantly adapting and learning,
the use of active neuromuscular stimulation to produce movement provides an additional
element over sub-motor threshold settings. There is evidence that repetitive active
movement, mediated by neuromuscular stimulation, can enhance motor re-learning after
CNS damage (Burridge & Ladouceur, 2001; Popovic et al., 2002). The additional
component of movement provides augmented sensory feedback and proprioceptive
afferent stimulation that is a direct consequence of the movement and muscle activation
(Burridge & Ladouceur, 2001).
As indicated previously, there is a reorganization of cortical representations for
body parts based upon the afferent input that they contribute as well as the amount of
motor activation they undergo. With electrical stimulation there is evidence that it equally
influences this reorganization (Golaszewski et al., 1999). This can be highly
advantageous in limbs that are impaired to such a degree that voluntary movement would
be impossible. Because of the significant amount of hemiparesis that is caused by stroke,
the use of electrical stimulation to generate movement provides a means of inducing
positive motor changes associated with motion (Chae et al., 1998; Chae et al., 2001).
Thus, the introduction of movement, even if supplemented by an external means can
provide the impetus for motor improvements. Likewise in stroke patient's who have
retained some degree of voluntary movement control following stroke, this intervention
can expand the pre-existing abilities and act as a catalyst to improve and regain daily
functioning activities (Burridge & Ladouceur, 2001).
Studies have demonstrated cortical reorganization as a result of peripheral
stimulation, and that this reorganization relates to functional change (Burridge &
Ladouceur, 2001; Golaszewski et al., 1999). The functional changes that have been noted
to improve with stimulation include passive and active range of motion, EMG activity,
muscle strength and torque, and a reduction in spasticity (Burridge & Ladouceur, 2001;
Glanz et al., 1996; Popovic et al., 2002).
Application and Underlying Principles of Electrical Stimulation
The application of electrical stimulation can take place either indirectly through
cutaneous electrodes or directly by contacting the nerve itself. Although many animal
studies use direct nerve stimulation (Baar & Esser, 1999; Timson, 1990; Wong & Booth,
1988) the majority of human studies focus on cutaneous stimulation. Granted, impedance
can be a factor during indirect current application, however, this tends to be less of an
issue with superficial muscles (Prentice, 1994).
The typical protocol for stimulation in humans involves placing cutaneous
electrodes over the target muscle. More specifically, the electrodes should be placed over
an area referred to as the motor point to give the optimal individual muscle contraction
(Prentice, 1994). This point is the region where motor nerve fibers are most numerous.
Stimulating at motor point regions maximizes the muscular response while limiting the
amount of current applied. This is important because of the potential for pain at electrode
sites associated with increasing levels of stimulation intensity.
Each of the surface electrodes possesses a different charge, one positive and the
other negative, therefore creating an electrical current. This current passes through the
underlying tissues and upon reaching the nerve leads to a series of electro-chemical
events that initiates an action potential. The electrical impulse that arrives at the nerve
must be of sufficient intensity to reach the nerve cell membrane's voltage threshold level.
Upon meeting or exceeding this threshold an action potential is created; a rapid
interchange of charged elements across the cell membrane, immediately followed by
propagation of a depolarization wave down the length of the nerve cell. This propagation
continues until the terminal end of the nerve cell contacts another nerve cell, or reaches
an effector organ (such as the muscle). In the case of neuromuscular stimulation the end
result is excitation of the corresponding muscle tissue and movement, if intensity is
sufficient (Prentice, 1994).
An important feature of signal transmission in nerve cells is the all-or-none manner
in which they fire. Once a stimulus is strong enough to reach the threshold level the
motor neuron will react with an action potential of full strength. Force gradation
produced in the muscle occurs through one of two ways: (a) recruiting more nerve cells,
or (b) increasing the frequency of firing for individual cells, as opposed to modifying the
size of a given impulse (Kandel, Schwartz, & Jessell 2000). Therefore, with electrical
stimulation, more nerve cells will fire as the strength or duration of the current increases.
A major distinction between volitional versus electrically induced neural
stimulation involves the recruitment order of motor unit firing. A motor unit consists of a
large alpha motor neuron that originates from the ventral horn of the spinal cord, and
supplies a number of muscle fibers (Popovic & Sinkjaer, 2001). Under normal voluntary
recruitment conditions there is an asynchrony of motor unit firing as some units are used
and others remain inactive. In particular, this takes place in an ordered progression
dependent on size of the units, referred to as the size principle, with recruitment of the
smallest fibers during low levels of activation and increasingly large fibers as the strength
of activation rises (Kandel, Schwartz, & Jessell 2000). In contrast, during an external
electrically induced muscle contraction, there is evidence suggesting a reversal of this
recruitment order possibly due to differences in nerve axon impedance (Basmajian & De
Luca, 1985; Feiereisen, Duchateau, & Hainaut, 1996). There is a lack of consensus as to
this proposed recruitment reversal, since several human studies have produced conflicting
results. Feiereisen et al., compared voluntary and electrically induced recruitment order in
humans and concluded that some degree of reversal did occur. However, given that
reversal did not occur to the expected extent, suggested that factors other than axon
impedance were involved (Feiereisen, Duchateau, & Hainaut, 1996).
Another difference between these modes of induction involves synchronization of
the motor units. During voluntary conditions, the firing of motor units is asynchronous,
meaning that some motor units are used while others are not. Electrical stimulation leads
to synchronization, with essentially the same motor units responding every time the
stimulus is applied. For this reason, voluntary contractions tend to have a later onset of
fatigue relative to electrically induced contractions (Prentice, 1994).
Volitionally Activated Stimulation
A final important distinction in the way that stimulation is provided involves the
idea of volitional control of movement. In addition to various levels of stimulation (i.e.
sub-sensory threshold vs. motor threshold), there are different ways of inducing the
stimulation. In this sense, the stimulation can either be passive in nature or instigated by
some measure of the subject's own volition. The passive protocol simply involves
applying stimulation, usually in a cyclical manner, to the subject to produce a movement.
On the other hand, volitional muscle stimulation involves some degree of activation
voluntarily attempted by subjects, and then stimulation is triggered. The typical manner
in which this is accomplished is through EMG activated stimulation, although other
related protocols have been used, such as positional feedback-based electrical stimulation
(Bowman et al., 1979). In positional feedback-based stimulation situations, subjects are
required to achieve a particular threshold value of movement to initiate the stimulation.
Once the subject achieved the pre-set value of joint range the stimulator would be
activated resulting in a full movement at the joint.
With EMG-triggered stimulation, muscle activity is measured at the motor points
for the preferred movement. When subjects reach the desired threshold, the stimulation
begins, and the movement is completed. EMG activation that serves as a pre-requisite for
stimulation adds a cognitive component to the standard stimulation protocol. Each
movement is "time locked" with subject volition and sensory feedback (Chae et al.,
2001). Studies have demonstrated that cyclical stimulation (stimulation without volition
on the part of the subject) may not be the best way to deliver neuromuscular stimulation
and that the stimulation, when produced under the control of the patient, tends to be
better related to improvements in motor function of the stroke patients (Burridge &
Ladouceur, 2001; Chae et al., 2001).
In Krafts' study, the greatest improvement in function among the various forms of
treatment resulted from EMG-triggered stimulation (Kraft et al., 1992). Although this
was not found significantly greater than the low-intensity stimulation protocol it should
be noted that the overall trend on all measures was for the greatest functional
improvement with EMG-triggered stimulation. In addition, the group receiving the low-
intensity stimulation that was below the motor threshold, was also attempting voluntary
movement of the limb. Therefore, this protocol was still using the concept of volition,
only with a lower degree of stimulation for re-enforcement.
Six studies have shown volitionally activated neuromuscular stimulation to produce
positive results over and above standard rehabilitation (Bowman et al., 1979; Cauraugh &
Kim 2002; Cauraugh et al., 2000; Chae et al., 2001; Francisco et al., 1998; Kraft et al.,
1992). These studies have demonstrated effectiveness with this form of treatment through
a wide variety of treatment measures. Most of these treatment variables have involved
measures of spasticity, range of motion, strength (isometric and isotonic), impairment and
overall disability (i.e., performance of activities of daily living).
Cauraugh and colleagues (Cauraugh & Kim, 2002; Cauraugh et al., 2000)
demonstrated improved motor function in the impaired wrist and fingers of chronic stroke
patients. These improvements were found in reaction time, sustained muscle contraction
ability, and in the Box and Block timed manipulation test. An important aspect of these
studies is that all subjects were chronic stroke survivors (at least one year post stroke),
with hemiparesis (partial paralysis) in the upper extremity. Thus, spontaneous
regeneration was an unlikely explanation for the functional improvements later in stroke
Francisco and colleagues (Francisco et al., 1998) noted that most research with
EMG-triggered stimulation had only used impairment measures when testing for
treatment effectiveness. They attempted to address the issue of motor function and it's
impact on physical disability (i.e., does the increased function noted in a motor ability
such as isometric strength have a positive influence on the performance of everyday
activities?). Their results indicated that EMG triggered stimulation led to significant
improvements in motor ability and functional recovery.
A noteworthy finding in the studies by Chae and colleagues (Chae et al., 1998),
and Powell and colleagues (Powell et al., 1999) where cyclic neuromuscular stimulation
was used, was an increase in the motor function of patients without significant functional
benefits. Both studies used the Functional Independence Measure as an indicator of
disability and observed no effect from their stimulation protocols upon it. These findings
reveal that a close relationship does not necessarily exist between disability and
neurological impairments in stroke patients. In a comprehensive review by Burridge and
Ladoucer (Burridge & Ladouceur, 2001), the general trend among studies using passive
and cyclical application of electrical stimulation, is that functional improvement in
everyday activities did not parallel impairment reduction. The absence of this problem in
the study by Francisco et al. (Francisco et al., 1998) may attest to the superior nature of
combining volition with regular neuromuscular stimulation.
Voluntary initiation of movement by the stroke patient appears to be of practical
significance because this is the route that movements would have to draw from when the
external aid is removed. A recent review by Woldag and Hummelsheim (2002) contends
that neuromuscular stimulation triggered through subject volition is superior to passively
induced stimulation due to the additional component of subject volition. Studies with
motor imagery have shown an increase in motor evoked potentials that coincided with
simple imagery focusing on that muscle group (Burridge & Ladouceur, 2001; Woldag &
Hummelsheim, 2002). The evidence tends to point towards cortical level changes as the
source for these effects. Therefore, with the introduction of volition into the electrical
stimulation procedure, higher-end re-structuring of the nervous system should be
Searching for an Optimal Treatment Duration
In general, we have established that electrical stimulation treatment has
demonstrated positive improvements on motor re-acquisition following a stroke. More
specifically, stimulation that produces movement that is initiated by subject volition has
proven effective in improving various measures of motor performance. In reviewing all
of the research that has been published on this topic a general conclusion is that the
application of this intervention is somewhat erratic. A wide variety of treatment protocols
have been used in terms of stimulus intensity, types of subjects used for testing, and
testing durations. This lack of standardization in how treatment is applied reflects an
absence in the knowledge base for what constitutes an optimal treatment strategy.
In addition, the issue of establishing optimal treatment is further clouded by many
different ways of measuring performance (Duncan et al., 2000). This makes comparison
among treatments, intensities, and durations difficult. In general, the outcome measures
that have been used tend to fall into four general categories: strength, range of motion,
measures of impairment, and measures of global disability (i.e., the extent to which the
impaired body part affects overall function in everyday-type activities). In the present
study, there was no intention to resolve the issue of outcome measure standardization.
Rather it has been presented as a further issue that needs to be addressed in regards to the
overall assessment of stroke rehabilitation outcome in general.
In an attempt to clarify the issue of developing an optimal treatment strategy this
paper attended to the issue of treatment duration. Some authors have argued that we need
more knowledge in the area. For example, Chae and colleagues (Chae et al., 1998) stated
that future studies should measure and "define the dose effect" as a means of guiding the
intervention. Of further note was the recent review by Popovic et al. (2002). They stated
that the most important questions currently needing attention involve when and to whom
therapy should be applied and what is the optimal treatment duration of the therapy. The
context for this comment was specifically tailored for functional electrical therapy (FET);
However, the comment equally applies to all forms of electrical stimulation and in
particular for EMG-triggered stimulation.
Why Treatment Duration?
In reviewing the available literature on active neuromuscular stimulation, among
the inconsistencies and blurred application of treatment, the issue of how long the
treatment period lasts is clearly unguided and lacks consensus. The broad range over
which treatments have been applied tends to reflect an almost arbitrary assumption of
what constitutes an effective period of time. The various studies have comprised a range
of two weeks (Cauraugh & Kim, 2002; Cauraugh et al., 2000) up to 10 months
(Dimitrijevic et al., 1996). The 10 month long study primarily used an afferent
stimulation below sensory threshold, thus a different form of electrical stimulation
(Dimitrijevic et al., 1996). However, this is still a means of using electrical stimulation to
assist motor recovery; Furthermore, even within studies that use motor-threshold
stimulation, a wide range in treatment times exists (Cauraugh & Kim, 2002; Kraft et al.,
1992; Chae et al., 1998).
An obvious question arises when one examines all of the studies that have been
completed to this point: Does treatment efficacy vary as a function of treatment duration,
and if so to what extent? If motor improvements are found in two weeks, then will four
weeks further differentiate groups? Intuition suggests that the more treatment is always
better and therefore using this intervention for four weeks versus two weeks should
provide enhanced performance measures. This appears to be an expansion of a principle
in general motor learning whereby the level of a skill is increased in proficiency in
relation to the amount of repetitive practice. However, there exists the possibility that
stimulation need only be applied to elevate the subject's function to a certain level at
which point the new functional abilities are incorporated into daily activities. In this
sense, the resolution of the limb impairment could lead to a re-introduction of the limb
into daily use thereby promoting the type of continued repetition needed to maintain
function, and providing a successful transfer of gains within the lab to a real world
setting. Thus, stimulation can be viewed as a catalyst to aid in motor re-learning
(Burridge & Ladouceur, 2001). In either case, it is possible that the optimal treatment
duration could be dependent upon the baseline function of the patient being treated.
Several studies have reported that the most promising degree of recovery occurs in
patients that enter treatment with the highest level of motor function (Burridge &
Ladouceur, 2001; Glanz et al., 1996). Perhaps patients that have a certain level of
impairment would need a corresponding period of electrical stimulation to improve.
The Selection of Time Periods for Comparison
In the studies by Cauraugh and colleagues (Cauraugh & Kim, 2002; Cauraugh et
al., 2000), significant improvements were observed in the motor performance of chronic
stroke patients. The first study involved two treatment sessions of 30 successful
movement trials that lasted approximately one hour, and this was performed three days
per week for two consecutive weeks. This led to a total of 360 trials over 12 treatment
sessions. The second study administered treatment protocols for 90 minutes twice a week
for two weeks. Given that both studies showed functional improvement after only six
hours of treatment, one could conclude that six hours provides a sufficient base for
evaluating time course effectiveness. By this it was assumed that two weeks (six hours)
represents an adequate minimal point of comparison for a treatment model because of the
reported demonstrated effectiveness.
The second treatment time for comparison was four weeks (90 minutes per session,
twice a week for four weeks). While the selection of this time frame was somewhat more
arbitrary than the baseline model, an important point to remember is that this research
question was exploratory in nature. In essence, this study attempted to answer the
question of whether active NMES training varies in effectiveness as a function of
treatment duration. This proposal tested if doubling the time involved with the baseline
model influenced motor improvements. Furthermore, bilateral movement training was
added in to the training equation, since this form of treatment has already established its'
value (Cauraugh & Kim, 2002). The issue of bilateral training will be addressed
Bilateral Movements Coupled with Active Stimulation
Cauraugh and colleagues (Cauraugh & Kim, 2002) investigated bilateral
movements of the forearm extensors while the impaired limb underwent EMG-triggered
stimulation. The concept behind bilateral movements is that the two limbs working as a
coordinated unit provides a means of recruiting alternate pathways in the nervous system
and thus enhances movement. Bimanual coordination theory as described in dynamic
systems theory provides the theoretical basis for including both arms as a behavioral
intervention. Evidence suggests that both arms share a central link as a coordinated
whole, with the upper limbs functioning in a homologous coupling of muscle groups on
both sides of the body. Coordinated movement patterns emerge spontaneously from the
constraints on the system including the inherent characteristics of muscles, as a function
of dynamics (Cauraugh & Kim, 2002).
Researchers argue that bilateral movements trigger interhemispheric disinhibition.
Three alternative recruitment pathways have been suggested to result from this
disinhibition, namely, (1) cortical motor neurons within the damaged hemisphere that
remain intact, (2) corticospinal pathways running ipsilaterally from the undamaged
hemisphere, and (3) ipsilateral corticospinal pathways that work indirectly. These
proposed means of neural plasticity were reviewed earlier as adaptive measures
undertaken by the nervous system in response to damage. In this particular case, bilateral
movement training activates an inherent capacity for neural reorganization that is
dependent on movement as a coordinated whole.
Studies using bilateral action as a treatment, have demonstrated motor
improvements beyond standard physical therapy. Therefore, the purpose of the study by
Cauraugh and colleagues, was to couple bilateral activity with EMG-triggered
neuromuscular stimulation (Cauraugh & Kim, 2002). In this sense, two scientifically
established interventions were coupled to determine potential advantages of combination
effects. The results of the coupled motor recovery protocols, showed a significant
improvement in motor ability on several outcome measures over the control group (no
treatment given), as well as the group receiving stimulation with unilateral activation
only. Because the coupling of interventions demonstrated improvements beyond either
one separately, the selection of a baseline treatment for the presently proposed study used
this combination. Therefore, the application of treatment involved EMG-triggered
stimulation of the impaired wrist and fingers along with simultaneous bilateral activation
of the unimpaired wrist and fingers.
The method and procedures for this study have been described in detail in two
recent publications by Cauraugh and colleagues (Cauraugh & Kim, 2002; Cauraugh et
al., 2000). The current method and rationale are summarized here.
All subjects were volunteers that had experienced a stroke at least one year before
the onset of testing. Stroke survivors who had chronic hemiparesis, partial paralysis of
their impaired upper extremity, were recruited. Because spontaneous processes are a
highly probable cause for recovery during the first year following a stroke, the influence
of any interventions used during this period may be hard to distinguish. For this reason,
the most compelling evidence in stroke rehabilitation research is derived from studies that
use chronic patients as their treatment group (Gresham et al., 1995). Mean length of time
since the stroke and age of the sample were calculated.
The degree of upper extremity chronic hemiparesis was monitored during the initial
assessment period to determine baseline function. Participants were evaluated on their
ability to extend at the wrist and/or fingers from a flexed position. To assess this function
the impaired hand was draped over the edge of a counter with the wrist joint slightly off
the counters' edge to allow for an unobstructed pivot. Participants were required to
demonstrate a minimum of 10 o extension in the wrist and/or fingers of the impaired limb
during this assessment to qualify for the study.
To understand the reason for focusing on the upper limb in this study, one must
appreciate the typical sequence of functional recovery following a stroke. In general,
reacquisition of arm function following a stroke poses far greater of a problem than
regaining leg function (Feys et al., 1998; Volpe et al., 2000). This could be for a number
of reasons including the majority of strokes occur in a region supplied by the middle
cerebral artery leading to consequent impairment of hand function (Feys et al., 1998). A
further likelihood is that because of the complex nature of the hand, functional recovery
is more difficult when compared to the leg (Kraft et al., 1992). Given that the upper
extremity is such a weak link in motor recovery, improvement in this area serves as a
sensible starting point. Further, the hand has such a profound impact on daily activities
that recovery is vital to returning to everyday activities.
The criteria for admission included: (1) diagnosis of at least one stroke (2) the
cutoff point in terms of motor recovery was set at 80% as an upper limit. This was
determined by using rectified EMG activation patterns and sustained force contractions
while comparing the impaired and unimpaired limbs; (3) the cutoff point in terms of a
lower limit, was 10 o of voluntary wrist or finger extension against gravity from a 90 o
flexed position. The requirement for some pre-existing capacity for movement is
important because the stimulation unit needs a certain amount of muscle activation before
turning on; (4) no other severe neurological deficits were present, including the use of a
pacemaker. Subjects needed to have a degree of cognitive capacity that allowed them to
participate in the various tests. This was necessary because some of the performance
measures involved recognition of a stimulus followed by an appropriate motor response;
(5) there was no use of drugs for the purpose of treating spasticity. The exclusion of
subjects on spasticity medication was important to isolate the influence of the stimulation
treatment, especially since a reduction in spasticity was one of the possible effects of the
treatment; (6) participants were not involved in any other motor recovery rehabilitation
protocol. Prior to testing, an *informed consent was signed by all subjects. (*Institutional
Review Board approved).
This study tested 13 subjects using a completely within-subjects design. All
participants underwent four weeks of total treatment, and therefore experienced both
levels of the treatment (two weeks and four weeks). Testing was performed prior to
treatment to establish a baseline, after two weeks of treatment (6 hours), and at the end of
four weeks (12 hours). The first treatment level consisted of six hours of treatment (four
days) completed over a period of two weeks. The second treatment level doubled the
duration (twelve hours of treatment/eight days, over a period of four weeks). Results
from the first level were compared with those from the second level.
PRE-TEST I POSTTEST1 I POSTTEST2
2 weeks training 2 weeks training
(6 hours) (6 hours)
Figure 1. Experimental Design
Instrumentation and Procedures
To measure motor function in the upper extremity, three separate types of
performance measures were used. These measures have been used and outlined in the
studies by Cauraugh and colleagues (Cauraugh & Kim, 2002; Cauraugh et al., 2000). The
first measure was the Box and Blocks timed manipulation test. This was a test of manual
dexterity that has age-based normative data available. The task required subjects to grasp
and transport small wooden blocks from one compartment to another that was directly
beside it. Participants were given 60 seconds to perform this task, and instructed to move
as many blocks as possible. This test was used to provide an index of overall hand
function in relation to reaching, grasping, transporting, and releasing.
The next two measures were force generation tasks that used the same
instrumentation. The first task measured reaction time in the wrist/finger extensors. The
purpose was to evaluate the speed of information processing and rapid muscle onset. The
second force generation task involved sustained muscle contraction in the wrist extensors
as a measure of force modulation. For both of these tasks, participants had their hands
placed into an apparatus designed to measure the force of wrist/finger extension. Load
cells with a capacity for 34.1 kg of force, were located in each of the two arm devices.
Force and EMG signals were recorded online for both arms. The EMG activity of
wrist/finger extensor muscles were recorded using surface electrodes (silver-silver
chloride electrodes with an epoxy-mounted preamplifier).
During the reaction time task, participants were instructed to respond as quickly as
possible to an auditory cue by means of wrist/finger extension against the load cells. Ten
trials were performed on each arm separately and 10 trials were performed using both
limbs together. During the sustained muscle contraction task, participants were instructed
to gradually increase their wrist/finger extension force to a maximal point and then hold
this isometric contraction for a period of 6.5 seconds. The experimenter provided the
signal to initiate and then cease force production. Three trials were performed for each
arm and both arms acting together. For both force generation tasks, testing was
administered in a random order with a rest period of approximately five minutes between
As noted previously, all participants received a total of 12 hours of treatment (24
treatment sheets over 4 weeks). Initially they received EMG-triggered neuromuscular
stimulation coupled with bilateral movement training for 6 hours of treatment (12
treatment sheets), which they completed over a period of approximately 2 weeks. This
constituted the first level of the treatment, after which, all participants were posttested
(posttest 1). The second level involved an additional two weeks of the same treatment,
resulting in double the duration (twelve hours of treatment and 24 treatment sheets, over
a period of four weeks). All participants were again tested at this point (posttest 2). A
treatment sheet consists of 30 separate stimulation trials.
During each training day, most participants completed four sets of thirty successful
EMG-triggered neuromuscular stimulation trials (four treatment sheets). Some
participants requested lower training volumes due to fatigue. Throughout training,
participants were instructed to perform wrist/finger extension with the unimpaired limb
as well as the impaired limb. However, only the impaired limb had the stimulation unit
attached. This process took approximately 90 minutes each training day.
At the beginning of each training session, surface electrodes were attached to the
extensor communis digitorum and extensor carpi ulnaris muscles of the impaired limb.
The electrodes were attached to an Automove (AM 800) EMG Facilitation Stimulator
microprocessor. This device functions to provide electrical stimulation once a target
threshold level of activity is reached with EMG. The electrical stimulation was provided
by a one second ramp up, followed by five seconds of biphasic stimulation at 50 Hz,
pulse width of 200 microseconds, and finally a one second ramp down. The mA range
spanned 16-29 mA. Stimulation was adjusted so that it assisted the subjects in attaining a
full range of movement in the wrist.
The initial threshold for the stimulation unit was set at 50 microVolts. When the
subjects successfully reached this level the stimulator automatically adjusted the
threshold to a slightly higher point. This unit works in a feedback sensitive manner and as
such it adjusted the threshold level down if the subject was unable to reach it. The
threshold in this sense dropped down to a point that approximated the level of effort that
the subject was able to produce. Each trial was separated by 25 seconds of rest.
The initial data reduction step was to rectify and smooth the reaction time EMG
data at 100, which was followed by determining the fractionated components (premotor
and motor). The premotor reaction time, a central component, was operationally defined
as the time from stimulus onset until EMG activity in the extensor muscles reach 30% of
peak activity. Motor reaction time, a peripheral component, was operationally defined as
30% of peak force amplitude. This component began immediately following premotor
reaction time, and ended with the initiation of movement.
The ability to maintain a certain level of force was determined through the
sustained contraction task. Root mean square error was the response variable, providing
an overall measure of bias and variability in peak force amplitude across the sustained
contraction interval. Root mean square error was calculated for a five second period, one
and a half seconds after the stimulus onset to allow for the gradual increase to peak force.
The number of blocks moved in the Box and Blocks test was analyzed in a within-
subjects design, one-way repeated measures ANOVA with three levels of test session
pretestt, posttest 1, and posttest 2). The reaction time and sustained contraction tasks
were analyzed separately using a within-subjects 2 x 3 repeated measures ANOVA (Hand
- unilateral vs. bilateral condition x Test Session pretest, posttest 1 and posttest 2). All
statistical tests were conducted with alpha level set at 0.05. The Tukey-Kramer procedure
was used as a multiple comparison follow-up test. Conservative degrees of freedom
adjustments were used for any sphericity violations.
Total reaction time along with its' fractionated components of premotor and motor
time were analyzed in separate 2 (Hand) x 3 (Test Session) ANOVAs with repeated
measures on the last factor. The levels of test session refer to the pretest, posttest 1, and
posttest 2. Hand represents whether the test is performed in a unilateral or bilateral
condition. The bimanual condition refers to the fact that both hands are involved in the
reaction trial, however, only the response of the impaired limb was evaluated. The
underlying purpose was to test for any differences that may result in the impaired arms'
performance when movements are coupled with the unimpaired arm.
Previous research with the same reaction time protocol indicated that median
reaction times provided a more accurate representation of the data (Cauraugh & Kim
2002). Therefore, median values were calculated from the individual trials for each of the
reaction time conditions (impaired limb only, unimpaired limb only, and both limbs
together). Prior to generating the median values, outliers were removed (i.e., values
outside 3 standard deviations). Mean values were calculated for each of the reaction time
condition median values. The resultant mean values were then analyzed using the above 2
x 3 ANOVAs. The median values for the individual subjects are presented in Table 1 for
the unilateral condition, and Table 2 for the bilateral condition. Note that subject 6 failed
to produce reliable data for post-test 1 and therefore was excluded from further reaction
Total reaction time failed to yield significant differences in the main effects or
interaction, hand: F(1, 8) = 2.95, p < 0.12; test session: F (2, 16) = 2.57, p < 0.11; hand x
test session: F (2, 16) = 0.20, p < 0.82. Total reaction time is displayed in Figure 1.
In addition, premotor time, the central component of reaction time, revealed no
significant differences in the main effects or interaction, hand: F (1, 8) = 0.35, p < 0.57;
test session: F (2, 16) = 0.55, p < 0.57; hand x test session: F (1.19, 9.54) = 0.15, p <
0.75. Premotor reaction time is displayed in Figure 2.
Analyses of the third reaction time component, motor time, revealed a significant
hand main effect, F (1, 8) = 5.485, p < 0.047. Faster motor reaction times were found
during the unilateral condition compared with the bilateral condition. No significant
differences for the interaction or the test session main effect were identified, test session:
F (2, 16) = 1.10,p < 0.36; hand x test session: F (2, 16) = 0.98, p < 0.39. Motor reaction
time is displayed in Figure 3.
Table 1. Reaction Time and Fractionated Components (Unilateral Condition)
Pre-Motor Time Motor Time Total Time
Subject Pre-test Post 1 Post 2 Pre-test Post 1 Post 2 Pre-test Post 1 Post 2
1 287 292 339 140 80 94.5 450 347 453.5
2 219.5 248 249 63.5 69 67 300.5 318 317
3 402 254.5 373 40 51 112.5 442 300.5 440.5
4 260 266.5 209 56 53.5 52 310 311 273
5 289.5 294 276.5 124 101 138 418.5 389.5 411.5
6 302.5 NA 247 98 NA 131 417 NA 384
7 187.5 186.5 212 43 61 51 240 245.5 266
8 315.5 265 230 50.5 32 22 361.5 293 259
9 151 138 155 36 33 30 191 171 188
10 244.5 296 289.5 97.5 28 117 336 324 400
Mean 265.9 248.94 258 74.85 56.5 81.5 346.65 299.94 339.25
SD 70.96 53.51 64.32 37.30 24.27 42.47 87.75 62.15 90.54
Total Reaction Time
Pre Post 1
Figure 2. Total reaction time
Pre-motor Reaction Time
Pre Post 1
Figure 3. Pre-motor reaction time
Motor Reaction Time
Post 1 Post 2
Figure 4. Motor reaction time
Table 2. Reaction Time and Fractionated Components (Bilateral Condition)
Pre-Motor Time Motor Time Total Time
Subject Pre-test Post 1 Post 2 Pre-test Post 1 Post 2 Pre-test Post 1 Post 2
1 332 356 417 172 117 127 471 482 536
2 231.5 240.5 254.5 45 42.5 66.5 283.5 280.5 326.5
3 290.5 285 396 104 111 81 408 396 498
4 264.5 263 257 63 56 53 326.5 322.5 310
5 354.5 327 289 118.5 116 163.5 468.5 443 454.5
6 334 300.5 255 113 128 148.5 377 428.5 406.5
7 147.5 183.5 186 45.5 97.5 40 193 274.5 225
8 258 232 232 55.5 52 26 338 272 268
9 141 143 144 36 28 38 175 172 178
10 342 261 262 142 50 124 510 317 381
Mean 269.55 259.15 269.25 89.45 79.8 86.75 355.05 338.8 358.35
SD 77.45 63.84 83.61 46.91 37.43 50.05 114.75 96.17 117.83
Sustained Muscle Contraction Task
As with reaction time, the sustained muscle contraction task was analyzed with a
mixed design, hand x test session, (2 x 3) ANOVA with repeated measures on the last
factor. Indeed, the factors of hand and test session have the same meanings as in the
reaction time task.
The response variable for sustained muscle contraction is operationally defined as
root mean square error. The purpose of this task is the maintenance of wrist extension
force over a 5 second time period. Therefore, root mean square error provides a marker
for consistency where lower values indicate more consistent force modulation. This
replicates the procedure and intermediate analysis outlined by Cauraugh and Kim (2002).
Pre Post 1 Post 2
Figure 5. Sustained Muscle Contraction Task
Results from the two-factor mixed design analysis indicated no significant
differences in the interaction or the test session main effect, test session: F (2, 18) =
0.164, p < 0.850; hand x test session: F (2, 18) = 0.048, p < 0.954. However, a strong
trend was found with the hand main effect, F (1, 9) = 4.481, p < 0.063, with the unilateral
condition demonstrating a tendency towards greater force modulation. The results of the
sustained muscle contraction task are displayed in Figure 4 and individual subject values
are reported in Table 3.
Table 3. Sustained Muscle Contraction Task Results (root mean square error)
Unilateral Condition Bilateral Condition
SUBJECT Pre-test Post 1 Post 2 Pre-test Post 1 Post 2
1 0.37 1.03 0.72 0.52 0.97 0.89
2 0.15 0.18 0.55 0.13 0.1 0.57
3 0.44 0.32 0.1 0.28 0.71 0.25
4 0.3 0.31 0.26 0.69 0.36 0.27
5 0.47 0.37 0.36 0.39 0.41 0.28
6 0.46 0.18 0.18 0.4 0.2 0.24
7 0.14 0.21 0.17 0.18 0.19 0.18
8 0.07 0.18 0.19 0.09 0.2 0.19
9 0.09 0.18 0.11 0.14 0.18 0.28
10 0.16 0.09 0.09 0.22 0.09 0.11
Mean 0.27 0.31 0.27 0.30 0.34 0.32
SD 0.16 0.27 0.21 0.19 0.29 0.23
Box and Blocks Results
Figure 6. Box and Blocks results
Box and Blocks Test
For the Box and Blocks test, a 1-way ANOVA with 3 levels of test session was
performed. The response variable for this test was the number of blocks moved in a 1
minute period. Analysis revealed no significant differences between the levels, F (2, 20)
= 0.987,p < 0.390. The results are displayed in Figure 5 and the number of blocks moved
is recorded in Table 4.
Table 4. Box and Block Test Results (number of blocks moved by the impaired arm in 60
Subject Pre Post 1 Post 2
1 11 3 5
2 12 13 14
3 20 14 18
4 2 9 3
5 3 3 2
6 47 48 47
7 0 0 1
8 26 23 25
9 11 13 20
10 14 13 20
11 15 15 17
Mean 14.64 14 15.64
SD 13.22 13.05 13.39
DISCUSSION AND CONCLUSIONS
The results of this study fail to support the hypothesis that chronic stroke patients
would demonstrate greater improvements in upper limb function after 12 hours (24
treatment sessions) compared with only 6 hours (12 treatment sessions) of treatment.
Moreover, the results of this study did not indicate that EMG-triggered electrical
stimulation was an effective means of improving motor function of the hand and wrist in
chronic stroke patients. There were no significant differences between the three testing
sessions in any of the three measures used to assess motor function in the hand and wrist.
These surprising results are inconsistent with previous studies using EMG-triggered
electrical stimulation (Cauraugh et al., 2000; Cauraugh & Kim, 2002; Francisco et al.,
1998; Hummelsheim et al., 1996; Kraft et al., 1992). These other studies were performed
using a similar form of active electrical intervention all finding significant improvements
on various measures of motor function. Moreover, two of these studies were conducted in
the same laboratory as the present study, using similar procedures (Cauraugh et al., 2000;
Cauraugh & Kim, 2002). From this collective pool of research, the resultant consensus is
that volitionally induced neuromuscular stimulation has a pronounced positive effect on
progress toward motor improvements in chronic stroke patients (Burridge & Ladouceur,
2001; Woldag & Hummelsheim, 2002).
When the outline and purpose for the present study were originally created the
published research in stroke motor recovery was considered strong evidence for a
treatment effect. From this general framework the present study worked under the
assumption that treatment was effective, and therefore it was only a matter of determining
if any duration related changes in efficacy took place. Failure to replicate motor
performance improvements with the coupled protocols (bilateral movement training and
active stimulation) was unexpected. Much of the ensuing discussion focuses on possible
reasons for the results of the present study and why they fail to coincide with the previous
Throughout stroke research a common concern is how reliably any advances in
function can be attributed to a particular intervention. As previously noted, chronic stroke
patients tend to offer a better opportunity for assessing treatment-related changes
(Gresham et al., 1995). Consequently, the present study used chronic stroke patients as a
participant base. The converse side to preventing possible contamination of treatment
effects via spontaneous recovery is that producing motor gains now falls onto less fertile
ground for recovery. The acute phase following stroke may afford a greater opportunity
for treatments to manifest in motor improvements. Independent of the confounding
associated with spontaneous regeneration, the acute phase following stroke may provide a
medium in which treatment effects are amplified (Turton & Pomeroy, 2002). Several of
the studies involving active electrical stimulation have used chronic stroke patients and
have demonstrated significant motor improvements with training (Cauraugh et al., 2000;
Cauraugh & Kim, 2002; Fields, 1987; Kraft et al., 1992). This includes two studies
previously performed in the same lab as the present study (Cauraugh et al., 2000;
Cauraugh & Kim, 2002). However, a point to consider is that the focus on chronic stroke
patients makes the advancement of function a tougher task.
Focusing on the specific patients involved in the present study reveals several
issues that deserve attention. Out of 13 participants that participated in the study, only 10
were able to reliably produce a full set of data to be included in the analyses. From this
group of 10 participants, 5 had previously participated in EMG-triggered stimulation
protocols in the same laboratory as the present study. Further, 3 of the participants were
involved in this additional training immediately prior to the current study. Perhaps some
of the participants reached a plateau hindering further improvement. The possibility
exists that for participants to advance to a higher level of function, treatment intensity
would need to rise in a corresponding manner. Intensity could be adjusted through higher
work volumes, reduced rest between stimulations, greater frequency of sessions, or
perhaps even complimentary modes of treatment.
An additional point to consider that may have influenced the results of the current
study involves the frequency of treatment. The protocol outlined in the methods section
initially called for two treatment sessions per week for 4 weeks. Each session consisted of
90 stimulation trials (3 training sheets with 30 trials each). However, because most
participants traveled significant distances to the lab (i.e., greater than 2 hours round trip
on average), this was altered to 2 sessions per week at 120 stimulation trials per session
(4 training sheets). This compromise was made in response to requests by participants
aimed at limiting the number of visits to the laboratory. Moreover, scheduling conflicts
occasionally arose where standardized treatment sessions were skipped, therefore,
reducing the training frequency to one visit over a week. In the case of one participant,
the time required for commuting was prohibitive enough to reduce training frequency to
once per week over the entire study. Another participant was only able to complete 60
stimulation trials per session because of fatigue. Such concessions were not ideal,
however, they were deemed necessary on the basis of increasing the sample size for this
study. Therefore, although the application of this intervention was standardized (i.e.,
work/rest ratio, stimulation impulse time etc), frequency of treatment occasionally varied
across participants. Certainly this increased variability, thus possibly hindering any
training effect by diluting the training intensity over a longer time period.
Further, study design concerns may have contributed towards the lack of significant
treatment effects. A pre-experimental design was employed for the current study,
therefore, no control group was required, nor was any additional baseline period taken
into account except for a single pre-test. At the outset of the study a control was opted
against for two reasons. First, all studies from different laboratories using active
neuromuscular stimulation in the past have demonstrated significant motor functional
improvements with treatment. Second, because subject recruitment imposed such an
obstacle in this study, the division into two groups and hence a reduction in the sample
receiving treatment seemed unnecessary. Indeed, the combination of these two factors
appeared to militate against the notion of dividing the participant pool. The original
judgment was to run all subjects through the training protocol because the primary
question for this study was if treatment efficacy varied with total treatment duration; as
opposed to whether or not the treatment was effective. Adherence to the rigorous
constraints of a traditional research design has been questioned by previous research,
especially in the rehabilitation arena where participant recruitment and withholding
treatment from a group of patients poses practical and ethical problems (Ottenbacher &
Hinderer, 2001). In summary, although exclusion of a control group appeared to be a
reasonable assumption going into the study the consequent major disadvantage is the
absence of a true comparison group.
In addition to some of the aforementioned points, there is a distinct challenge in
recovering motor functions in the upper limb of stroke patients (Feys et al., 1998; Volpe
et al., 2000). This is particularly pronounced in wrist and finger extension because of the
presence of an uncontrolled flexion synergy in the impaired upper limb (Cauraugh et al.,
2000). Consequently, improvement in wrist and finger extension ability offers an
excellent indication of therapeutic efficacy, thus, providing the rationale for an upper
limb focus in the current study. However, a disadvantage with an upper limb focus, is that
eliciting gains in this area can be onerous, therefore imposing an additional obstacle for
treatment gains. This is similar to the problem associated with limiting the study to
chronic stroke participants whereby improvements are more difficult to produce.
Another unexpected result from this study was the impact of unilateral versus
bilateral testing conditions. Reaction time and sustained contraction testing was
performed under three separate conditions; impaired arm only, unimpaired arm only and
both arms concurrently. Analysis was limited to the impaired limb, however the bilateral
condition was imposed to monitor for possible influences from the coupling. Cauraugh
and Kim (2002) demonstrated faster reaction times with total reaction time and the
fractionated components during the bilateral condition. Results from the present study did
not show a significant difference for total and premotor reaction times, and even revealed
faster motor reaction times for the unilateral condition. The sustained contraction task,
which also revealed an advantage during the bilateral testing condition for Cauraugh and
Kim (2002), indicated a trend favoring the unilateral condition in the current study. A
possible explanation for performance decrements in the bilateral condition may be
elevated demands in attention related to a divided focus on two limbs. However, it is
unknown how these differences could have occurred between the two studies since such
similar training and testing methods were involved.
An additional issue that tends to plague therapeutic rehabilitation trials in general is
that of sample heterogeneity (Lincoln et al., 1999; Pomeroy & Tallis, 2002; van der Lee
et al., 2001). Because the impairments rendered by stroke widely range in severity, the
problem of heterogeneity is introduced as participants of various ability levels
collectively represent the influence of treatment. The exact nature of this influence is
currently unknown because contrasting results have been produced in stroke research.
Feys et al (1998) used a repetitive rocking action to treat the upper limb of stroke patients
and noted that the greatest improvements occurred in those subjects with higher levels of
impairment. Conversely, the review by Burridge and Ladouceur (2001), revealed a trend
in electrical stimulation treatment whereby subjects retaining some residual motor control
demonstrated a superior response. A possible explanation for such a discrepancy may be
that subjects within different impairment classes may respond favorably to different types
of treatment (Lincoln et al., 1999; Pomeroy & Tallis, 2002). Lincoln et al (1999)
conceded how the inclusion of a broad spectrum of impairment levels can possibly
conceal treatment effects that may vary according to the patient sub-group. To narrow the
functional range of participants, the present study imposed inclusion criteria with both
upper and lower limits of acceptability. However, even within this truncated range it is
possible that the level of impairment could influence the effect of treatment. For example,
the number of blocks moved during the Box and Blocks pretest range from 0 to 47
therefore indicating a broad range in functional ability in the participants. Stratification of
participants into appropriate categories and the corresponding effects of treatment have
been put forth as important areas of future research (Popovic et al., 2002).
The failure of this study to demonstrate any significant improvements associated
with active neuromuscular stimulation rendered further examination of the primary
research question a moot point. Directly addressing the issue of time-related treatment
effects via the current study was impossible. This issue remains unresolved in stroke
motor recovery in general and is not simply a problem with neuromuscular stimulation
(Lincoln et al., 1999; Kwakkel et al., 1997; van der Lee et al., 2001). The results of the
present study fail to offer any insight into this issue, therefore, the remainder of this
discussion will focus on related stroke research and what it can tell us about the influence
of treatment duration.
A research synthesis by Kwakkel et al. (1997) indicated that a small, but
statistically significant functional improvement corresponded with higher intensities of
treatment, where intensity was defined as the amount of time in therapy. For their study,
stroke rehabilitation was viewed in general terms without a focus on any one specific
intervention and results were then pooled and analyzed using a meta-analysis. Another
meta-analysis performed by Ottenbacher and Jannell (1993) likewise addressed stroke
rehabilitation in general terms to determine if it was effective overall in improving
functional outcomes. Their results indicated that focused rehabilitation programs may be
helpful in improving function, however, they found that this improvement was unrelated
to the duration of treatment.
Nugent et al. (1994) examined the relationship between the amount of weight-
bearing exercise and walking characteristics. The results indicated that a dose-response
was present, but only for those subjects that had an initial capacity in the impaired limb
for single limb support during walking. Those subjects who were initially unable to stand
or step forward did not exhibit this relationship with additional exercise repetitions. This
re-emphasizes the influence of impairment level on the outcome of therapy. In this case a
treatment-duration effect was contingent upon a higher baseline level of function.
Further, Kwakkel and Wagenaar (2002) investigated the connection between time
spent in rehabilitation and walking ability. Their aim was to reveal any influence training
duration may have on walking speed and the underlying walking mechanics (i.e.,
coordination between hemiplegic and non-hemiplegic sides). The results indicated that
longer treatment durations involving the lower extremities were associated with small
increases in comfortable walking speed compared to controls. Not surprisingly, longer
treatment duration for the group emphasizing upper extremity rehabilitation had no
impact on walking outcome. This would appear consistent with principles of training
specificity. In regards to interlimb coordination between hemiplegic and non-hemiplegic
sides, additional therapy provided no benefit relative to controls.
One theme that does appear to be arising from the literature is that treatment
duration influence is certainly not a universal effect. Different participants may respond
in a correspondingly different manner to a particular type of treatment and any dose-
relations may be reliant upon this impairment level. Furthermore, even within a given
class of impairment, dose-effects may be specific regarding how or if they influence a
Relating treatment duration to efficacy with active neuromuscular stimulation,
Fields (1987) indicated that improved outcomes were associated with higher treatment
frequency. An important qualification on these results, however, is that EMG activity in
affected muscles served as the primary endpoint, therefore, leading to possible difficulties
in comparisons with practical function. Although Field states how improvements were
reflected in functional measures these measures involved ambulation (for the lower limb)
and range of motion; not upper extremity function. As noted by Lincoln et al. (1999),
hand and wrist function imposes a greater challenge for recovery, thus highlighting an
important distinction (Lincoln et al., 1999).
Given that massed repetition has been suggested as an important factor in motor
recovery following stroke (Turton & Pomeroy, 2002), a logical assumption is that a
greater degree of repetition would make increased contributions to motor recovery. Short
and long term neural transformations occur in the nervous system in relation to the
activation frequency of respective pathways (Hallet, 2001). Based on this information, an
intuitive inference is that a positive link exists between the volume of training and
therapeutic effectiveness. Due to the value of repetition and cortical re-structuring, a
further question could be asked regarding what amount of repetition is needed to notice
improved ability and does this rate of improvement vary as a function of the treatment
time? It is possible that a threshold level of repetition must be exceeded for improvement
to occur. At the other extreme is the potential of diminishing returns where continued
treatment may fail to provide gains beyond a certain point. Both of these issues represent
important information for clinical application of treatment.
Muelbacher et al. (2002) demonstrated improved hand motor function as a result of
regional anesthesia of the upper arm coupled with motor practice. The premise behind
this procedure was to remove the influence of the competitive input of the upper arm in
the motor cortex, thereby allowing more cortical substrate area to be available to the
hand. A significant increase in performance occurred following the initial practice session
and the follow-up session, combined with retention of these improvements. However,
following the second practice session no further improvements took place indicating a
plateau in this training effect. Thus, an important application of the above results may be
that stroke interventions can produce improvements to a limit at which point further
treatment application is no longer effective. Discovering if and where such a point exists
in active neuromuscular stimulation would be valuable to clinicians where cost and time
efficiency are highly regarded directives.
An important qualification to this simple question is repeating how any impact is
likely influenced by the patients' functional level and the specific outcome variable in
question. Consensus throughout the stroke motor recovery literature tends to indicate that
a larger amount of treatment will generate greater benefits (van der Lee et al., 2001).
Specific populations of patients based on impairment may respond disproportionately to
various types of intervention (Lincoln et al., 1999), therefore several implications are
made for active neuromuscular stimulation. Future motor recovery studies may increase
the accuracy and treatment applicability of their effects if participants are stratified based
on impairment level. Therefore, treatment effects can be determined relative to these
specific populations. Further, creating such homogenous groups could increase the
chance of exposing possible differences in regards to treatment duration effects.
In conclusion, the current results failed to support the hypothesis that a greater
duration of active neuromuscular stimulation would generate increased measures of
motor performance in the wrist and finger extensors of chronic stroke patients.
Furthermore, the findings did not support EMG-triggered neuromuscular stimulation as
an effective means of motor recovery. Possible reasons for this failure to replicate
previous research results include participant heterogeneity, lack of a control group, focus
on a complex area of motor recovery (i.e., wrist/finger extension), and the use potentially
LIST OF REFERENCES
Abo M, Zhengguang C, Lai LJ, Reese T, Bjelke B. Functional recovery after brain
lesion-contralateral neuromodulation: an fMRI study. Brain Imaging
Baar K, Esser K. Phosphorylation of p70S6k correlates with increased skeletal muscle
mass following resistance exercise. Am. J. Physiol. Cell Physiol. 1999;276:C120-
Basmajian J, De Luca C. Muscles Alive: Their Functions Revealed by Electromyography.
5th Edition. Baltimore, Md: Williams and Wilkins; 1985.
Bowman BR, Baker LL, Waters RL. Positional feedback and electrical stimulation,
automated treatment for hemiplegic wrist. Arch Phys MedRehabil 1979;60:497-
Braun C, Heinz U, Schweizer R, Wiech K, Birbaumer N, Topka H. Dynamic
organization of the somatosensory cortex induced by motor activity. Brain
Burridge J, Ladouceur M. Clinical and therapeutic applications of neuromuscular
stimulation: A review of current use and speculation into future developments.
Byrnes ML, Thickbroom GW, Phillips BA, Mastaglia FL. Long-term changes in motor
cortical organization after recovery from subcortical stroke. Brain Research
Cauraugh JH, Kim S. Two coupled motor recovery protocols are better than one:
electromyogram-triggered neuromuscular stimulation and bilateral movements.
Cauraugh J, Light K, Kim S, Thigpen M, Behrman A. Chronic motor dysfunction after
stroke: Recovering wrist and finger extension by electromyography-triggered
neuromuscular stimulation. Stroke. 2000;31(6): 1360-1364.
Chae J, Bethoux F, Bohinc T, Dobos L, Davis T, Friedl A. Neuromuscular stimulation for
upper extremity motor and functional recovery in acute hemiplegia. Stroke.
Chae J, Fang Z-P, Walker M, Purmehdi S. Intramuscular electromyographically
controlled neuromuscular electrical stimulation for upper limb recovery in chronic
hemiplegia. Am JPhys Med Rehabil 2001;80:935-941.
Chen R, Cohen LG, Hallet M. Nervous system reorganization following injury.
Neuroscience 2002; 111;4:761-773.
Dimitrijevic MM, Stokic DS, Wawro AW, Wun CC. Modification of motor control of
wrist extension by mesh-glove electrical afferent stimulation in stroke patients.
Arch Phys Med Rehabil. 1996;77:252-259.
Duncan PW, Lai SM, Keighley J. Defining post-stroke recovery: implications for design
and interpretation of drug trials. Neuropharmacology 2000;39:835-841.
Feiereisen P, Duchateau J, Hainaut K. Motor unit recruitment order voluntary and
electrically induced contractions in the tibialis anterior. Exp Brain Res.
Feys HM, De Weerdt WJ, Selz BE, Cox-Steck GA, Spichiger R, Vereeck LE, Putman
KD, Van Hoydonck GA. Effect of a therapeutic intervention for the hemiplegic
upper limb in the acute phase after stroke. Stroke 1998;29:785-792.
Fields RW. Electromyographically triggered electrical muscle stimulation for chronic
hemiplegia. Arch Phys MedRehabil 1987;68:407-414.
Francisco G, Chae J, Chawla H, Kirshblum S, Zorowitz R, Lewis G, Pang S.
Electromyogram-triggered neuromuscular stimulation for improving the arm
function of acute stroke survivors: a randomized pilot study. Arch Phys Med
Glanz M, Klawansky S, Stason W, Berkey C, Chalmers T. Functional electrostimulation
in poststroke rehabilitation: A meta-analysis of the randomized controlled trials.
Arch Phys Med Rehabil. 1996;77:549-553.
Golaszewski S, Kremser Ch, Wagner M, Felber S, Aichner F, Dimitrijevic MM.
Functional magnetic resonance imaging of the human motor cortex before and after
whole-hand afferent electrical stimulation. ScandJRehab Med 1999;31:165-173.
Gresham GE, Duncan PW, Stason WB. Post-Stroke Rehabilitation: Clinical Practice
Guidelines No. 16. Rockville Md: U.S. Dept. of Health and Human Services.
Public Health Service, Agency for Health Care Policy and Research. AHCPR
Publication No. 95-0662. May 1995.
Hallett M. Plasticity of the human motor cortex and recovery from stroke. Brain
Research Reviews 2001;36:169-174.
Hummelsheim H, Amberger S, Mauritz KH. The influence of EMG-initiated electrical
stimulation on motor recovery of the centrally paretic hand. European Journal of
Johansson BB. Brain plasticity and stroke rehabilitation: The Willis lecture. Stroke
Kandel E, Schwartz J, Jessel T. Principles ofNeuroscience. Elsevier Science Publishing
Co., Inc., 2000.
Kraft GH, Fitts SS, Hammond MC. Techniques to improve function of the arm and hand
in chronic hemiplegia. Arch PhysMedRehabil. 1992;73:220-227.
Kwakkel G, Wagenaar RC. Effect of duration of upper-and lower-extremity rehabilitation
sessions and walking speed on recovery of interlimb coordination in hemiplegic
gait. Physical Therapy 2002;82;5:432-448.
Kwakkel G, Wagenaar RC, Koelman TW, Lankhorst GJ, Koetsier JC. Effects of intensity
of rehabilitation after stroke: A research synthesis. Stroke 1997;28:1550-1556.
Liepert J, Bauder H, Wolfgang HR, Miltner HR, Taub E, Weiller C. Treatment-induced
cortical representation after stroke in humans. Stroke 2000;31:1210-1216.
Lincoln NB, Parry RH, Vass CD. Randomized, controlled trial to evaluate increased
intensity of physiotherapy treatment of arm function after stroke. Stroke
Muellbacher W, Richards C, Ziemann U, Wittenberg G, Weltz D, Boroojerdi B,
Leonardo C, Hallet M. Improving hand function in chronic stroke. Arch Neurol
Nudo RJ. Recovery after damage to motor cortical areas. Current Opinion in
Nugent JA, Schurr KA, Adams, RD. A dose-response relationship between amount of
weight-bearing exercise and walking outcome following cerebrovascular accident.
Arch Phys Med Rehabil 1994;75;4:399-402.
Oostra K, Van Laere M, Scheirlinck B. Use of electrical stimulation in brain-injured
patients: a case report. Brain Injury 1997;11;10:761-764.
Ottenbacher KJ, Hinderer SR. Evidence-based practice: Methods to evaluate individual
patient improvement. Am JPhys MedRehabil 2001;80:786-796.
Ottenbacher KJ, Jannell S. The results of clinical trials in stroke rehabilitation research.
Arch Neurol 1993;50:37-44.
Pomeroy V, Tallis R. Neurological rehabilitation: A science struggling to come of age.
Physiotherapy Research International 2002;7;2:76-89.
Popovic DB, Popovic MB, Sinkjaer T. Neurorehabilitation of upper extremities in
humans with sensory-motor impairment. Neuromodulation 2002;5;1:54.
Popovic D, Sinkjaer T. Control of Movement in the Physically Disabled. London, UK:
Powell J, Pandyan AD, Granat M, Cameron M, Stott DJ. Electrical stimulation of wrist
extensors in poststroke hemiplegia. Stroke. 1999;30:1384-1389.
Prentice W. Therapeutic Modalities in Sports Medicine 3rd Edition: Mosby-Year Book
Rossini PM, Tecchio F, Pizzella V, Lupoi D, Cassetta E, Pasqualetti P, Romani GL,
Orlacchio A. On the reorganization of sensory hand areas after mono-hemispheric
lesion: a functional (MEG)/anatomical (MRI) integrative study. Brain Research
Taub E, Miller NE, Novak TA, Cook EW, Fleming WC, Nepomuceno CS, Connell JS,
Crago JE. Technique to improve chronic motor deficit after stroke. Arch Phys Med
Timson BF. Evaluation of animal models for the study of exercise-induced muscle
enlargement. J. Appl. Physiol. 1990;69:1935-1945.
Turton A, Pomeroy V. When should upper limb function be trained after stroke?
Evidence for and against early intervention. NeuroRehabilitation 2002;17:215-224.
Van der Lee JH, Snels IAK, Beckerman H, Lankhorst GJ, Wagenaar RC, Bouter LM.
Exercise therapy for arm function in stroke patients: A systematic review of
randomized controlled trials. Clinical Rehabilitation 2001;15:20-31.
Velasco F. Neuromodulation: an overview. Archives of Medical Research 2000;31:232-
Volpe BT, Krebs HI, Hogan, Edelstein L, Diels C, Aisen M. A novel approach to stroke
rehabilitation: Robot-aided sensorimotor stimulation. Neurology 2000;54:1938-
Whelan PJ, Pearson KG. Plasticity in reflex pathways controlling stepping in the cat.
Journal ofNeurophysiology 1997;78;3:1643-1650.
Winstein C, Wing AM, Whitall J. Motor control and learning principles for rehabilitation
of upper limb movements after brain injury. In: Handbook ofNeuropsychology, 2nd
Edition, Grafman J, Robertson IH. (Eds) 2003;9:77-126.
Woldag H, Hummelsheim H. Evidence-based physiotherapeutic concepts for improving
arm and hand function in stroke patients. JNeurology 2002;249:518-528.
Wong TS, Booth FW. Skeletal muscle enlargement with weight-lifting exercise by rats.
J. Appl. Physiol. 1988;65:950-954.
David A. E. Bolton was born in Windsor, Canada, on November 7, 1972. After
receiving his Bachelor of Science degree in human kinetics from the University of
Guelph, he relocated to Gainesville, Florida, to pursue a Master of Science in
Exercise and Sport Sciences degree, with a concentration in motor learning and control
at the University of Florida. During his tenure as a graduate student, David has had the
opportunity to serve as a graduate teaching assistant in the Department of Exercise and
Sport Sciences. In fall 2003, he will begin the doctoral program in neuroscience at the
University of Alberta, specializing in sensory and motor systems.