University of Southern California Dissertations and Theses

Motor skill learning in individuals with Parkinson's disease: Consideration of cognitive and motor demands

(USC Thesis Other)

Motor skill learning in individuals with Parkinson's disease: Consideration of cognitive and motor demands

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content MOTOR SKILL LEARNING IN INDIVIDUALS WITH PARKINSON’S DISEASE:
CONSIDERATION OF COGNITIVE AND MOTOR DEMANDS
by
Sompom Onla-or
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
in Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BIOKINESIOLOGY)
May 2001
Copyright 2001 Sompom Onla-or
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 3027761
_ _ ®
UMI
UMI Microform 3027761
Copyright 2001 by Bell & Howell Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
Bell & Howell Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES. CALIFORNIA 90007
This dissertation, written by
#
under the direction of k&c. Dissertation
Committee, and approved by all its members,
has been presented to and accepted by The
Graduate School in partial fulfillm ent of re
quirements fo r the degree of
DOCTOR O F PHILO SO PHY
Dean of Graduate Studies
Date ....ttsy..XX±..20Q%
DISSERTATION COMMITTEE
Chairperson
- O - C ' / 7 y r
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DEDICATION
To my mom, Chumreung Onla-or,
whose love and support has provided foundation
for every challenge undertaken and every goal pursued.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
iii
TABLE OF CONTENTS
Dedication ii
List of Tables iv
List of Figures v
Abstract xi
Chapter 1 BACKGROUND AND SIGNIFICANCE 1
Chapter 2 MOTOR SKILL LEARNING IN INDIVIDUALS WITH 32
PARKINSON’S DISEASE: CONSIDERATION
OF COGNITIVE AND MOTOR DEMANDS
Chapter 3 CONCLUSIONS 108
Bibliography 122
Appendix 136
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
iv
TABLES
Chapter 1:
Table 1. Summary of behavioral motor skill learning studies in 24
humans with Parkinson’s disease
Chapter 2:
Table 1. Demographic characteristics of each participant with PD 57
Table 2. Group mean comparisons for the demographic characteristics 58
Chapter3:
Table 1. Summary of the experimental design and findings 115
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
V
FIGURES
Chapter 2:
Figure 1. Three goal movement trajectories with the same spatial 49
scaling but different temporal scaling at 900, 1200, and 1500 ms.
Figure 2. Experimental set-up with arm lever and feedback 51
display. A) Subject sits in front of the monitor with arm resting on the
lever. Upper right shows an overhead view of the subject with arm on
the lever at the starting position. B) An example of the two forms of
augmented visual feedback, a graphic representation of the subject’s
movement trajectory for that trial superimposed over the goal trajectory,
and the RMSE score on the top right corner. The goal trajectory is
displayed as a thick solid line and the subject’s trajectory is displayed
as a thin dash line.
Figure 3. RMSE block means during acquisition days 1 (block 1-5) 60
and 2 (block 6-10) for Control (closed symbol) and PD (open symbol)
groups. Top row illustrates data from the low cognitive (LC) demand
condition as the triangle symbol and bottom row illustrates those from
the high cognitive (HC) demand condition as the square symbol.
Left column illustrates data from the high motor demand condition
(900 ms trajectory) and right column illustrates those from the low motor
demand condition (1500 ms trajectory). RMSE is shown in degrees.
Error bars are standard error of mean (SEM).
Figure 4. Mean RMSE during retention tests (blocked and random) 62
for Control (closed symbol) and PD (open symbol) groups. Top row
illustrates data from the low cognitive (LC) demand condition and
bottom row illustrates those from the high cognitive (HC) demand
condition. Left column illustrates data from the high motor demand
condition (900 ms trajectory) and right column illustrates data from the
low motor demand condition (1500 ms trajectory). RMSE is shown in
degrees. Error bars are SEM. There was a significant Group X Motor
Demand interaction (p < .02) for the LC condition and a significant
Group X Retention Condition interaction (p < .01) for the HC condition.
Figure 5. Example of individual data from the PD-HC subgroup 63
over acquisition and two retention tests. Left graph shows data from
participant PD-HC1 who performed the random retention test first and
then the blocked retention test. Right graph shows data from a different
participant (PD-HC5) who performed the blocked retention test first then
the random retention test. Participants with PD who practiced in the
high cognitive demand condition showed higher error in the blocked
retention than in the random retention independent of the order
of which retention test they performed first.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vi
Figure 6. Mean RMSE (averaged across two MT trajectories) for 64
each group by cognitive demand condition. Data from the control
subgroups (Con-LC, Con-HC) are shown on the left and those from
the PD subgroups (PD-LC, PD-HC) are shown on the right. The
control group showed lower error in blocked than in random retention
independent of the cognitive demand condition (Retention Effect;
p < .003). In contrast, the PD group showed lower error in the similar
practice-retention context than in the different practice-retention context.
This resulted in a significant Cognitive Demand X Retention interaction
(p < .004). RMSE is shown in degrees. Error bars are SEM.
LC = low cognitive demand, HC = high cognitive demand.
Figure 7. Bar graphs illustrate mean RMSE differences between 66
last block of acquisition and first block of each retention test for
Control and PD groups. A positive change reflects decreased in
error and a negative change reflects increased in error from
acquisition to each retention test. Top row illustrates data from the
low cognitive demand condition (blocked practice order with 100%
FB) and bottom row illustrates those from the high cognitive demand
condition (random practice order with 60% faded FB). Left column
illustrates data from the high motor demand condition (900 ms
trajectory) and right column illustrates those from the low motor
demand condition (1500 ms trajectory). Overall, subjects in the high
cognitive demand condition (bottom row) showed smaller magnitude
of performance deterioration from last acquisition block to each
retention test when feedback was withheld compared to those in the
low cognitive demand condition (top row); Cognitive Demand Effect
(p < .001). RMSE is shown in degrees. Error bars are SEM.
Figure 8. Bar graphs illustrate mean RMSE differences between 67
first block of acquisition and first block of each retention test for
Control and PD groups. A positive change reflects decreased in error
and a negative change reflects increased in error from the beginning
of practice [baseline performance]. Top row illustrates data from the
low cognitive demand condition (blocked practice order with 100%
FB) and bottom row illustrates those from the high cognitive demand
condition (random practice order with 60% faded FB). Left column
illustrates data from the high motor demand condition (900 ms
trajectory) and right column illustrates those from the low motor
demand condition (1500 ms trajectory). Overall, subjects in the high
cognitive demand condition (bottom row) showed larger magnitude
of performance improvement from baseline compared to those in the
low cognitive demand condition (top row); Cognitive Demand Effect
(p < .01). RMSE is shown in degrees. Error bars are SEM.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vii
Figure 9. Mean RMSE during transfer test for Control and PD 69
Groups. A) Bar graphs show mean RMSE for each group averaged
across cognitive demand condition and two new temporally scaled
trajectories (1050 and 1350 ms). Group difference was significant
at p = .03. B) Group mean RMSE for each trajectory. Overall, group
differences were most pronounced for the 1050 ms trajectory.
Although, Group X Motor Demand interaction did not reach
significance (p = .18), the effect size was large (ES = .77). RMSE is
shown in degrees. Error bars are SEM.
Figure 10. Example of movement trajectory scaling. The 900 ms 83
goal trajectory is displayed as the thick black line. The subject’s
actual movement is displayed as the thin black line. RMSE is the
difference (in degrees) between the goal trajectory and subject’s
actual movement. GMP is the difference between the goal trajectory
and subject’s trajectory after subject’s trajectory was scaled in time
(dash line) and then amplitude (dot line). TF (time factor) 1.3 indicates
that the subject moved slower than the goal trajectory by 30%. AF
(amplitude factor) 1.4 indicates the subject overshot the target by 40%.
Figure 11. GMP error block means during acquisition Days 1 87
(block 1-5) and 2 (block 6-10) for the Control (closed symbol) and
PD (open symbol) groups. Top row illustrates GMP error from the
low cognitive demand condition (blocked practice order with 100%
FB) and bottom row illustrates those from the high cognitive demand
condition (random practice order with 60% faded FB). Left column
illustrates data from the high motor demand condition (900 ms
trajectory) and right column illustrates those from the low motor
demand condition (1500 ms trajectory). GMP error is shown in
degrees. Error bars are SEM.
Figure 12. TF block means during acquisition Days 1 (block 1-5) 88
and 2 (block 6-10) for the Control (closed symbol) and PD (open
symbol) groups. Top row illustrates TF from the low cognitive
demand condition (blocked practice order with 100% FB) and bottom
row illustrates TF from the high cognitive demand condition (random
practice order with 60% faded FB). Left column illustrates data from
the high motor demand condition (900 ms trajectory) and right column
illustrates those from the low motor demand condition (1500 ms
trajectory). Horizontal line at 1.0 indicates the goal movement time.
Error bars are SEM. Across the two cognitive demand conditions,
there was a significant Motor Demand Effect; p < .01, suggesting
that both groups were more accurate in reproducing the 1500 ms
than the 900 ms movement time.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 13. AF block means during acquisition Days 1 (block 1-5)
and 2 (block 6-10) for the Control (closed symbol) and PD (open
symbol) groups. Top row illustrates AF from the low cognitive
demand condition (blocked practice order with 100% FB) and bottom
row illustrates AF from the high cognitive demand condition (random
practice order with 60% faded FB). Left column illustrates data from
the high motor demand condition (900 ms trajectory) and right column
illustrates those from the low motor demand condition (1500 ms
trajectory). Horizontal line at 1.0 indicates the goal movement
amplitude. Error bars are SEM.
Figure 14. Mean GMP error during retention tests (blocked and
random) for Control (closed symbol) and PD (open symbol) groups.
Top row illustrates data from the low cognitive demand (LC) condition
(blocked practice order with 100% FB) and bottom row illustrates
those from the high cognitive demand (HC) condition (random practice
order with 60% faded FB). Left column illustrates data from the high
motor demand condition (900 ms trajectory) and right column
illustrates those from the low motor demand condition (1500 ms
trajectory). GMP error is shown in degrees. Error bars are SEM.
There was a significant Group Effect (p < .04) for the LC condition
but not for the HC practice condition (p = .67).
Figure 15. TF block means during retention tests (blocked and
random) for Control (closed symbol) and PD (open symbol) groups.
Top row illustrates TF from the low cognitive demand (LC) condition
and bottom row illustrates TF from the high cognitive demand (HC)
condition. Left column illustrates data from the high motor demand
condition (900 ms trajectory) and right column illustrates those from
the low motor demand condition (1500 ms trajectory). Horizontal line
at 1.0 indicates the goal movement time. Error bars are SEM.
There was a significant Group X Motor Demand interaction (p < .05)
for the LC practice condition and a significant Group X Motor
Demand X Retention Condition interaction (p < .03) for the HC
practice condition.
Figure 16. AF block means during retention tests (blocked and
random) for Control (closed symbol) and PD (open symbol) groups.
Top row illustrates AF from the low cognitive demand condition
(blocked practice order with 100% FB) and bottom row illustrates AF
from the high cognitive demand condition (random practice order
with 60% faded FB). Left column illustrates data from the high motor
demand condition (900 ms trajectory) and right column illustrates
those from the low motor demand condition (1500 ms trajectory).
Horizontal line at 1.0 indicates the goal movement amplitude. Error
bars are SEM. There were no significant main effects or interactions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ix
Figure 17. GMP error averaged across the two trajectories 95
(1050 and 1350 ms) during the transfer phase for the Control and
PD groups. For in the low cognitive demand practice condition, the
PD group exhibited significantly higher GMP error than that of the
control group (p < .05). GMP error was not different between the
two groups in the high cognitive demand practice condition.
Error bars are SEM.
Figure 18. TF for the Control and PD groups during the transfer 96
phase. A) bar graphs show TF for each goal trajectory. Both groups
exhibited higher TF for the 1050 MT trajectory than for the 1350 MT
trajectory (Motor Demand Effect; p < .001). B) bar graphs show TF
for each cognitive demand practice condition. The PD group exhibited
higher TF than the control group for both cognitive demand conditions.
Although, group differences did not reach significance, the effect size
was large (ES for low cognitive demand = .64, high cognitive demand
= .82). Error bars are SEM.
Figure 19. AF for the Control and PD groups during the transfer 97
phase. A) bar graphs show AF for each group by goal trajectory.
B) bar graphs show AF for each group cognitive demand practice
condition. AF was similar between the two groups for the LC practice
condition. The PD group exhibited higher AF than the control group
for the HC condition. Group differences, however, did not reach
significance (p = .50) due primarily to the high variability in the PD group.
Error bars are SEM.
Chapter 3:
Figure 1. A simplified model of the neuro-substrates mediating 120
motor skill learning (box A) and the explicit learning and memory
system (box B). The neuroanatomical network mediating the motor
skill learning includes the motor cortex, premotor area, supplementary
motor area (SMA), dorsolateral-prefronta! area (DLPF), basal ganglia,
thalamus, and cerebellum. Solid arrows illustrate multiple closed-loops
for the cortico-striatal connections that underlie motor skill learning.
While most areas in the frontal lobe project to and receive inputs from
the basal ganglia, they also send inputs directly to the motor cortex
(dash arrow). The explicit learning and memory system mediates by
the medial-temporal lobe. Several areas in the medial-temporal lobe
(e.g. hippocampus, entorhinal, and subicular cortex) have reciprocal
connections with the frontal lobe particularly the DLPF and premotor
areas (dot arrows). It may be possible that the explicit knowledge
gained during the high cognitive demand practice condition allows
the basal ganglia to be circumvented by using the alternative pathways.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
This proposed alternative pathway for motor skill learning is from the
medial temporal lobe to the frontal lobe particularly the DLPF and
premotor areas (dot arrows) then from the frontal lobe to the motor
cortex (dash arrows) and finally from motor cortex to the spinal cord
(solid arrow).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABSTRACT
MOTOR SKILL LEARNING
IN INDIVIDUALS WITH PARKINSON’S DISEASE (PD):
CONSIDERATION OF COGNITIVE AND MOTOR DEMANDS
Procedural motor learning capability was investigated in 20 adults with
moderately severe PD and 20 age-matched neurologically healthy subjects. The
task was to learn a rapid goal-directed arm movement. Participants practiced the
task either in a low cognitive (LC) or high cognitive (HC) demand condition. All
practiced similar trajectories with three different motor demands (low, medium, and
high) over two days (135 trials/day). Learning was assessed during no feedback, 24
hr-delayed retention (blocked and random) and transfer tests.
In the first analysis, overall movement error was used to index motor
learning. For the LC condition, and independent of retention test, the PD group
exhibited similar error to controls when motor demand was low but greater error
than controls when motor demand was high. For the HC condition, and
independent of motor demand, the PD group exhibited similar error to controls in
random but greater error than controls in blocked retention tests. Cognitive demand
had a differential effect on retention performance for only the PD group.
Specifically, the PD group performed with lower error in the similar practice-retention
context than in the different practice-retention context, suggesting context-
dependent learning. Transfer performance showed impaired generalizability of
learning in PD.
The second analysis investigated the processes by which PD subjects
acquire a motor skill. Motor program error (GMP) was differentiated from parameter
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
error. Impaired GMP learning was revealed under the LC but not HC condition. In
contrast, impaired temporal-parameter learning was evident under the high (except
for HC practice-random retention) but not low motor demand. The ability to learn
spatial-parameters of a discrete movement was preserved in PD.
Overall, results underscore the importance of cognitive demand, motor
demand, and retention test condition (context) when evaluating the procedural
motor learning capability of individuals with moderately severe PD. Motor learning
capability in PD has a strong contextual-dependence. The basal ganglia may
function to suppress and select responses that are appropriate to the contextual
environment. These findings suggest that task-specific training may be beneficial
for rehabilitation of individuals with PD.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1:
BACKGROUND AND SIGNIFICANCE
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I. The Problem
In the last decade, there has been increased interest in understanding the
role of the basal ganglia in motor learning. The motivation behind this interest has
been twofold: 1) to understand the neural substrates associated with motor learning,
and 2) the potential clinical application for rehabilitation of individuals with basal
ganglia disorders. There is evidence that the basal ganglia are part of the
distributed neural network subserving motor learning (Calabresi, Pisani, Mercuri, &
Bernardi, 1996; Doyon et al., 1997; Gabrieli, 1995). However, a clear understanding
of the role of the basal ganglia in motor learning has been limited by three central
factors related to subject inclusion criteria and methodology in previous behavioral
studies.
First, individuals with Parkinson’s disease (PD) have largely served as
subjects in research associated with the role of the basal ganglia in motor learning.
Parkinson’s disease is a progressive degenerative disease of the basal ganglia and
as such it is reasonable that this population would provide insight with respect to the
role of the basal ganglia in motor learning. However, disease severity has not been
strictly controlled in a number of previous studies. Several studies included subjects
with mild (Hoehn & Yahr stage I) as well as moderate (Hoehn & Yahr stage II, III)
and severe (Hoehn & Yahr stage IV) PD (e.g., Bondi & Kaszniak, 1991; Daum et al.,
1995; Yamadori, Yoshida, Mori, & Yamashita, 1996). It has been shown that the
extent and severity of cognitive and motor impairments associated with PD are
dependent on the stage of the disease as measured by the Hoehn and Yahr (1967)
scale (Doyon et al., 1997; Harrington, Haaland, Yeo, & Marder, 1990; Owen et al.,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
1993). Doyon and associates found that individuals with moderate PD (Hoehn and
Yahr stage II and III) showed impaired motor learning while those with mild PD
(Hoehn and Yahr stage I) showed intact motor learning compared to age-matched
controls (Doyon et al., 1997). Likewise, Harrington and colleagues found that only
individuals who had advanced symptoms of PD (severe bradykinesia) demonstrated
impaired motor learning (Harrington et al., 1990). Thus, variability in subjects’
symptoms may result in discrepancies in findings among previous work.
The second factor contributing to discrepancies in previous findings on the
motor learning capability of individuals with PD is the cognitive and motor demand of
the tasks. The level of cognitive and motor requirements of the practice tasks has
not been controlled in previous behavioral studies. Conclusions about motor
learning capability of individuals with PD have been drawn without consideration of
task motor and cognitive demand. As this issue of motor/cognitive task demand is
central to the present investigation, a more detailed discussion will be provided in
section III.
Third, previous behavioral studies have not systematically delineated
performance effects from learning effects. Motor learning is defined as a set of
internal processes associated with practice or experience leading to relatively
permanent changes in the capability for movement (Salmoni, Schmidt, & Walter,
1984). Retention and transfer tests are used in order to dissociate temporary
performance effects from relatively permanent effects (i.e. learning). In a retention
test, the learner performs the same task that has been practiced while in a transfer
test, the learner switches to perform a different task or the same task with different
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
parameters (e.g., movement time (MT), amplitude). Across previous studies
investigating the role of the basal ganglia in motor learning, few have included
retention and/or transfer tests. A related issue is that the amount of practice,
considered one of the most important variables in motor learning, has perhaps been
insufficient in previous studies. Practice duration was no longer than 15-20 min in
most of the previous studies. Therefore, it is conceivable that actual learning was
limited by insufficient practice. It has been shown that individuals with PD are able
to learn a motor task, however, they need more practice than healthy age-matched
controls (Pascual-Leone et al., 1993).
II. Motor skill Learning
The recognition that there are multiple forms of learning and memory began
in the 1960s from an investigation of the patient HM, an individual who had bilateral
medial-temporal lobe resections. There is now a general consensus that multiple
learning and memory systems are mediated by different brain areas. Cohen and
Squire (1980) have introduced two distinct forms of learning and memory;
declarative (explicit) and procedural (implicit). Motor skill learning is a subcategory
of implicit learning. The term “implicit learning”, first introduced by A.S. Reber in the
late 1960s, is used to describe the acquisition of complex, abstract knowledge that
takes place without the learner’s awareness that he/she is learning (Reber, 1989).
The implicit learning system is supported by procedural memory. Procedural
(implicit) memory is the capability to acquire and retain new skills through physical
practice and is not directly accessible to conscious recollection as facts or data.
This kind of memory is inflexible (i.e. it provides limited access to response systems
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
that are not involved in the original learning) and is assessed by changes in
performance as a result of prior experience. Implicit learning can be classified into
skill learning (motor skill, perceptual skill, and cognitive skill), habit learning,
classical conditioning, and priming (Squire, 1992). Squire (1987) suggested that
each subcategory of implicit learning might be mediated by its own specific
neuroanatomic substrate(s). Indeed, recent studies have provided some support for
this view (e.g. Koenig, Thomas-Anterion, & Laurent, 1999; Soliveri, Brown,
Jahanshahi, Caraceni, & Marsden, 1997). Along this line, it is not clear whether or
not the basal ganglia are important for all types of implicit learning. Recently, there
has been increased interest to examine whether the basal ganglia are essential for
all or only particular subcategories of implicit learning. In the present study, I am
interested in the role of the basal ganglia in motor skill learning.
III. Learning of Skilled Actions Requires both Motor and Cognitive Operations
Central to the purpose of the present study is that previous studies
investigating the role of the basal ganglia in motor learning have not accounted for
the fact that tasks differ in their cognitive and motor demands. It may be misleading
to categorize a task as a “motor task”. The capability to acquire motor skills requires
both cognitive and motor processes (Haaland & Harrington, 1990; Salmon &
Butters, 1995). In fact, motor skill is highly cognitive and the cognitive processes
that subserve goal-directed movement must be practiced (Allard, 1993). In general,
cognition refers to a collective group of thought processes. Cognitive demand in
motor learning refers to the mental work involved in decision-making processes
regarding the anticipation, planning, regulation, and interpretation of motor
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
performance (Lee, Swinnen, & Serrien, 1994). Lee and coworkers (1994)
demonstrated that the cognitive component of motor skills has a critical impact on
the learning process. Further and most important to the present study, the effort by
which cognitive processes are under taken can be influenced by variables such as
practice order and feedback frequency.
Given that "motor” tasks require both motor and cognitive processing, it is
not possible then to compare results across studies in which the relative levels of
cognitive and motor demands were not the same. Discrepancies in findings among
previous work could be due to differences in the level of cognitive and motor
demands of the tasks. For example, Jordan and Sagar (1994) showed that the
ability to learn to accurately produce a designated amount of force using a handgrip
dynamometer was normal for individuals with PD compared to controls (Jordan &
Sagar, 1994). In a similar line, studies using a tracking/tracing task have
demonstrated normal motor learning ability in individuals with PD (Agostino, Sanes,
& Hallett, 1996; Frith, Bloxham, & Carpenter, 1986; Soliveri et al., 1997). In
contrast, studies that used the serial reaction time task (SRTT) demonstrated that
individuals with PD are impaired in motor learning compared to healthy age-
matched controls (Ferraro, Balota, & Connor, 1993; Jackson, Jackson, Harrison,
Henderson, & Kennard, 1995; Pascual-Leone et al., 1993). The serial reaction time
task, developed by Nissen and Bullemer (1987), is one of the motor tasks that are
commonly used to study motor learning in individuals with PD. With the SRTT,
subjects are presented with a stimulus that can appear in one of several different
locations. Subjects are instructed to make a manual response to the stimulus as
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
fast as possible. Unbeknownst to the subject, a fixed sequence of stimuli is
embedded in the trials and is repeatedly presented over practice sessions (Nissen &
Bullemer, 1987). It has consistently been demonstrated that, in healthy subjects,
reaction time for the repeated sequence decreases significantly over practice. And
this reaction time for the repeated sequence is faster than that for the random
sequence, indicating that learning has occurred.
Indices of learning for grip force and tracking tasks are measures of motor
output - accurate force production and time on target, respectively. With grip force,
the task goal is to accurately produce a certain amount of force. The task goal for
tracking is to remain in contact with a moving target that appears on the computer
screen. In contrast, index of learning for SRTT is a measure of reaction time on the
repeated sequences. It has been well established that reaction time is reflective of
the cognitive processing required to plan an action (Henry & Rogers, 1960).
Therefore, it is possible that motor learning deficits were not evident in grip force
and tracking studies by Jordan and Sagar (1994), Agostino, Sanes, & Hallett (1996),
Frith, Bloxham, & Carpenter (1986), and Soliveri et al. (1997) due to the relatively
low cognitive demand of the tasks whereas motor learning deficits demonstrated in
SRTT studies by Ferraro et al. (1993), Jackson et al. (1995), and Pascual-Leone et
al. (1993) were due to the relatively higher cognitive demand of the SRTT.
Studies using a pursuit rotor task also demonstrated discrepancies in their
findings. The demand in the pursuit rotor task, like the tracking task described
above, is primarily motor. The goal of the pursuit rotor task is to maintain contact
between a stylus and a disk on the rotating turntable. The turntable can be adjusted
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
to rotate at different speeds. Learning of the pursuit rotor task is indicated by an
increased time that the stylus remains in contact with the disk over practice and
after a rest period. Harrington and colleagues had individuals with PD perform a
pursuit rotor task at speeds of 30, 45, and 60 rpm for 9 min a day over a period of 3
days (Harrington et al., 1990). They found that the PD group was impaired in
pursuit rotor learning compared to age-matched controls. A closer examination
revealed that the locus of this impairment was in the highest speed condition (60
rpm condition). Using a modified version of the pursuit rotor (computer based),
Bondi and Kazniak (1991) found that the ability to learn the pursuit rotor task for
subjects with PD was comparable to that of controls. In Bondi and Kazniak’s study,
the turntable was rotated at speeds of either 7 or 14 rpm. Although the pursuit rotor
task was used in both studies (Harrington et al., 1990; Bondi & Kazniak, 1991), task
demands were markedly different. First, the speeds of the turntable in Bondi and
Kazniak’s study (7 or 14 rpm) were well below those used by Harrington et al (30,
45, and 60 rpm). Second, the turntable was rotated at one fixed speed (blocked
practice order) in Bondi and Kazniak while it was randomly rotated at three different
speeds (random practice order) in Harrington et al. It has been shown that random
practice order requires higher cognitive demand compared to blocked practice order
(Magill & Hall, 1990; Wright, Li, Coady, 1997). Discrepancies in the motor learning
findings between these two studies could be due to either the differences in the
speeds of the turntable (i.e. motor demand) or differences in the order of task
practice (i.e. cognitive demand) or a combination of both. Taken together, it
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
appears that motor learning deficits in PD subjects were revealed when either
cognitive and/or motor demand of the task was high.
Manipulation of cognitive demand: Contextual Interference and Feedback
Frequency
In this study, cognitive demand was manipulated via two practice variables:
contextual interference and feedback frequency. Contextual interference (Cl) is a
phenomenon that has been demonstrated as having a large impact on the learner’s
cognitive processing during motor skill acquisition. Contextual interference refers to
the interference effects in performance and learning that arise from practicing one
task in the context of other tasks (Shea & Zimny, 1983). Context is manipulated by
altering the order in which tasks are practiced. Three practice orders have been
commonly used: random, blocked, and serial. In random practice order, all tasks
are presented in a random order in each practice set (e.g. C-A-B, A-B-C, B-C-A). In
blocked practice order, one task is repeated within each set of trials before a new
task is presented in the next set (e.g. B-B-B, A-A-A, C-C-C). Finally, in serial
practice order, tasks are presented in a sequential order and this order is repeated
throughout each set of practice (e.g. A-B-C, A-B-C, A-B-C). It has been shown that
random and serial practice orders produce high Cl while blocked practice order
produces low Cl (Shea, Kohl, & Indermill, 1990). Studies examining the influence of
Cl on motor learning have typically shown that a high Cl practice condition (i.e.
random order) degrades acquisition performance but benefits retention performance
(i.e. facilitates learning). In contrast, a low Cl practice condition (i.e. blocked order)
benefits acquisition performance but degrades retention performance (Magill & Hall,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
1990). This phenomenon has been explained by the more extensive cognitive
processing demand during random practice and it has been suggested that this type
of information processing is particularly beneficial for supporting subsequent
retention and transfer efforts (Wright, Li, & Coady, 1997).
In a study by Shea and Morgan (1979), healthy young subjects practiced
three motor tasks (knock down barriers with a tennis ball) under a blocked (low Cl)
or random order (high Cl). Retention of the three tasks was measured after a 10-
min and 10-day delay under blocked and random orders. Subsequent transfer to a
task of either the same or more complex than the original tasks was also
investigated. Results showed that, early in the acquisition phase, subjects in the
blocked practice order performed significantly better than those in the random
practice order. However, performance on the three tasks was similar between the
two practiced groups by the end of acquisition. During retention, subjects in the
random practice order performed relatively well under both blocked and random
retention conditions. In contrast, subjects in the blocked practice order showed
performance deterioration especially under the random retention condition.
Likewise, performance in the transfer phase was superior for subjects in the random
practice condition compared to those in the blocked practice condition. These
findings suggested that practice under high Cl conditions: 1) establishes a memory
representation that is relatively independent of the context of the practice condition,
and 2) facilitates generalizability of learning (Shea & Morgan, 1979).
Two mutually compatible explanations for the Cl effect in motor learning
have been proposed. The “distinctiveness and elaboration” hypothesis was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
originally proposed by Battig (1979) for verbal learning. This hypothesis was more
specifically developed for motor learning by Shea and colleagues (Shea & Morgan,
1979; Shea & Zimny, 1983). The second explanation is the “reconstruction”
hypothesis proposed by Lee and Magill (1983, 1985). According to Battig (1979),
multiple and variable information processing strategies are used during high Cl
practice (i.e. random order). The multiple and variable processing strategies result
in an increased distinctiveness and elaborateness of the memory representation for
the task being learned. The increased distinctiveness and elaboration emerge from
the use of two distinct cognitive processing modes, intertask and intratask, that the
learners engage during practice. Participants who practice in a random order are
encouraged to take advantage of both intratask and intertask cognitive processing.
In contrast, participants who practice in a blocked order rely primarily on intratask
processing because information about different tasks are rarely processed together
in working memory. Thus, practice under high Cl (i.e. random order) enhances
performance during retention and transfer where a more distinct and elaborate
memory representation is needed.
Lee and Magill (1983, 1985) argued that high Cl practice conditions do not
necessarily enhance the elaborateness and distinctiveness of the memory
representation of the skill variation being practiced. Rather, they argued that high Cl
conditions cause increases in effortful processing activity for a particular strategy
variation with each succeeding trial (i.e. increase cognitive processing). In this view,
it is believed that previously encoded information about that strategy has been
completely or partially forgotten due to the intervening practice trials of the other
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
strategy variations. The authors proposed that to perform a movement during any
practice, a learner must retrieve an “ action plan”. The action plan consists of an
appropriate generalized motor program (GMP) representing that action, and the
parameters specific to the goal of the task to be performed. Lee and Magill (1985)
hypothesized that the details of the “action plan” are forgotten and must be
reconstructed prior to initiating the next response. This reconstruction process
requires cognitive processing which leads to an effective representation of the skill
and consequently a superior performance during retention and transfer. In a
random practice condition, participants practice reconstructing the action plan more
often than in the blocked practice condition where the previous action plan may not
have been forgotten due to limited interference from other tasks. The reconstruction
practice in the random practice condition is thought to enhance retention and
transfer performance.
The second practice variable used to manipulate the cognitive demand was
feedback frequency. In the motor learning literature, feedback is often provided in
the form of knowledge of results (KR). Knowledge of results refers to the extrinsic
information about task success that is provided to the performer after a practice trial
has been completed. Knowledge of results is defined as augmented, verbalizable,
terminal (i.e. post-response) information about the movement outcome in terms of
the environmental goal (Salmoni, Schmidt, & Walter, 1984). Evidence from several
studies has suggested that practice with lower relative frequencies of KR is as
effective or may even be more beneficial for learning compared to practice with
100% frequency of KR (e.g. Winstein, Merians, & Sullivan, 1999; Winstein &
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
Schmidt, 1990; Wishart & Lee, 1997). Winstein and Schmidt (1990) used a
relatively complex motor task (goal-directed arm aiming with specific spatial-
temporal goals) to examine the effects of a reduced KR relative frequency on motor
skill learning in healthy young adults. The results demonstrated that practice in a
reduced KR relative frequency condition (33%) was as effective for learning
(measured by performance in a retention test) as practice in a 100% KR condition.
In a second experiment, using the same task, Winstein and Schmidt (1990) showed
that subjects who received an overall 50% faded KR frequency (i.e. high frequency
KR early in practice with gradually reduced frequency of KR by the end of practice)
exhibited lower error scores in a 24-hr delayed retention test than those who
received 100% KR. It has been proposed that less frequent feedback might
enhance learning by forcing the subject to engage in cognitive processing
necessary to perform the task. In contrast, high frequency feedback provides the
solution for the learner and diminishes the cognitive requirement (Schmidt, Young,
Swinnen, & Shapiro, 1989). In addition, when external feedback is always provided,
subjects may become dependent on the feedback and not utilize the cognitive
operations required to solve the task problem (Schmidt, 1991).
Manipulation of motor demand: Force
In contrast to cognitive demand where information processing operations are
manipulated, motor demand can be thought of in terms of the amount of force
required to perform an action. One way to grade the motor demand of a task is by
manipulating the amount of force that the motor units have to generate within a
specific time. According to Newton’s law: F = ma (F = force, m = mass, a =
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
acceleration of segment center of mass), when mass is constant, the amount of
force is increased as a result of increased acceleration. Thus, motor demand is
higher in conditions when the central nervous system has to produce a given
average force rapidly (i.e. high acceleration) compared to a condition when a longer
time is available to produce that same average force (i.e. low acceleration).
Reduced force generation has been shown in individuals with PD (Phillips, Martin,
Bradshaw, & lansek, 1994). Subjects with PD showed decreased agonist burst
activity and compensated by using sequential bursts of agonist/antagonist activity to
promote adequate force necessary for movement completion (Hallett & Khoshbin,
1980). Time to reach peak velocity (acceleration) was then prolonged in order to
allow enough time to generate the appropriate amount of force (Phillips et al., 1994).
Thus, one way to grade motor demand of a task is by varying the magnitude of force
or the speed of movement.
IV. The roles of the basal ganglia in cognitive and motor operations
Evidence from structures and circuit organization of the basal ganglia
Is function of the basal ganglia associated with these kinds of motor or
cognitive operations or both? Although the basal ganglia are still among the least
understood of all brain structures, much of their mysteries have been revealed
during the past decade. Our knowledge about connections of the basal ganglia with
the other regions of the brain have led to a major revision of the traditional views of
basal ganglia organization and function. It is now generally accepted that, besides a
role in motor control, basal ganglia also have critical roles in cognitive operation and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
emotional function (Martin, 1996). Based on neuroanatomic organization of the
basal ganglia, both cognitive and motor deficits as a consequence of damage to the
basal ganglia are expected.
The basal ganglia consist of a number of subcortical nuclei: the striatum
(caudate and putamen), the globus pallidus, the substantia nigra (pars compacta
and pars reticulata), and the subthalamic nucleus (STN). The caudate nucleus and
putamen are the input zones of the basal ganglia. They receive most of the basal
ganglia’s input from diverse regions of the cerebral cortex, including motor, sensory,
prefrontal, and limbic cortical areas, and the intralaminar region of the thalamus.
These inputs are excitatory, mediated by glutamatergic neurons. The basal
ganglia’s output nuclei include the internal segment of the globus pallidus (GPi) and
substantia nigra pars reticulata (SNr). These structures project to three nuclei in the
thalamus: the ventrolateral, ventroanterior, and the mediodorsal nuclei and, by this
route, project back to the prefrontal cortex, premotor cortex, supplementary motor
area (SMA), and motor cortex.
Five parallel basal ganglia thalamocortical circuits have been identified: the
skeletomotor, oculomotor, dorsolateral prefrontal (DLPF), lateral orbitofrontal, and
anterior cingulate circuits (Alexander & Crutcher, 1990; Hoover & Strick, 1993).
These multiple parallel, segregated circuits suggest different functions for each
circuit. The skeletomotor and oculomotor circuits are also known as the “motor
loop” since their functions are associated with movement control. The DLPF circuit
is known as the “ association or complex loop” due to its function related to higher
cognitive function. Finally, the lateral orbitofrontal and anterior cingulate circuits are
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
known as the “ limbic loop” due to their functions associated with emotion (Middleton
& Strick, 2000; Rolls, 1994). Relevant to this review are the motor and association
loops.
Motor Loop.
The motor loop of the basal ganglia includes the skeletomotor and
oculomotor circuits. The skeletomotor circuit begins in the sensorimotor and
premotor areas, and passes through the putamen. The output nuclei of this loop are
Gpi and SNr. They project back to primary motor, premotor, and supplementary
motor areas via the ventrolateral and ventroanterior nuclei of the thalamus. The
oculomotor loop begins in the posterior parietal and prefrontal areas, and passes
through the body of the caudate nucleus. The output nuclei (GPi and SNr) project to
the ventroanterior and mediodorsal nuclei of the thalamus, then by this route, project
back to the frontal eye field and supplementary eye field. These motor loops play
an important role in movement control of the body and eyes.
Association or Complex Loop.
The dorsolateral prefrontal circuit is also known as the association or
complex loop (Rolls, 1994). This loop begins in the posterior parietal and premotor
areas, and projects to the head of the caudate nucleus. Similar to the oculomotor
circuit, output nuclei (GPi and SNr) project to the ventroanterior and mediodorsal
nuclei of the thalamus. Information from the association loop is sent back to the
dorsolateral prefrontal area. Function of the association loop is associated with
cognition, spatial memory, and evaluation of behavior.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
Each circuit includes direct and indirect pathways to the output nuclei. The
direct pathway is the striatal projection to GPi and SNr, which then projects to the
thalamus. It is inhibitory, mediated by GABA and substance P. The indirect
pathway is the striatal projection to the external segment of the globus pallidus
(GPe), which projects to the STN. The STN, in turn, projects back to both segments
of the pallidum and to the substantia nigra. Projection from the striatum to GPe is
mediated by GABA and enkephalin, GPe to STN is mediated purely by GABA, and
finally STN to the output nuclei is mediated by excitatory, glutamatergic neurons.
The striatal efferents from the direct and indirect pathways appear to have opposite
effects upon the output nuclei and the thalamic targets of the basal ganglia circuits.
Activation of the direct pathway results in phasic disinhibition of thalamocortical
neurons and, in turn, increases activation of the cortical areas. In contrast,
activation of the indirect pathway inhibits the thalamus and, in turn, decreases the
activity of the cortical targets of the basal ganglia. The dopaminergic projection from
substantia nigra pars com pacta (SNc) has different effects on the two striatal
pathways. Loss of dopamine appears to strengthen the indirect pathway and
weaken the direct pathway. Since normally, the direct pathway functions to facilitate
movement whereas the indirect one functions to inhibit movement, the SNc
dopaminergic neurons facilitate movement through both pathways. Recently,
Graybiel and colleagues (2000) have proposed a three-pathway model for the basal
ganglia. Specifically, in addition to the direct and indirect pathways, which are the
matrix-based pathways, they proposed the third pathway—the striosomal-based
pathway projecting from the striosomal compartment of the striatum to the dopamine
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
containing SNc (Graybiel, Canales, & Capper-Loup, 2000). The authors suggested
that the regulation of movement by the basal ganglia depends not only on the
balance between the direct and indirect pathways but also on the balance between
these pathways and the striosomal pathway.
Evidence from clinical symptoms of patients with basal ganglia disorders
Traditionally, diseases of the basal ganglia such as Huntington’s disease
(HD) and PD are known as movement disorder diseases. The prominent clinical
sign of HD is uncontrolled movements (chorea), which gradually increases until the
patient is confined to a wheelchair or bed. The cardinal clinical signs of PD are
resting tremor (4-7 Hz), rigidity, bradykinesia (slowness of movements), and
postural instability. It has been well established that motor behavior is impaired in
individuals with PD. Slowness of movement (Flash, Inzelberg, Schechtman, &
Korczyn, 1992; Phillips et al., 1994; Winstein, Pohl, Cormack, & Waters, 1995),
delayed movement initiation (Brown & Marsden, 1990; Contreras-Vidal, Teulings, &
Stelmach, 1995), and hypometric movement in individuals with PD (Castiello,
Stelmach, & Ueberman, 1993; Contreras-Vidal et al., 1995; Flowers, 1976) have
been identified across studies. In addition, difficulty switching between movements
has often been demonstrated in individuals with PD (Bennett, Marchetti, lovine, &
Castiello, 1995; Castiello, Bennett, Adler, & Stelmach, 1993; Onla-or & Winstein,
2001; Weiss, Stelmach, & Hefter, 1997).
Besides causing motor control deficits, basal ganglia disorders also cause
cognitive deterioration and emotional disturbance. Dementia is an early disabling
consequence of HD. Patients with HD usually experience absentmindedness,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
irritability, and depression. Cognitive deficits such as dementia, impaired attentional
set shifting, and impaired problem solving are evident in individuals with PD (e.g.
Brown, Schneider, & Lidsky, 1997; Tamaru, 1997; Wise, Murray, & Gerfen, 1996).
Community-based epidemiologic studies report the prevalence of dementia in
individuals with PD ranging from 24 to 41% (Mayeux, Denaro, & Hemenegildo,
1992). Individuals with demented-PD are primarily compromised on tasks requiring
visuospatial abstraction and reasoning, which distinguish them from those with
Alzheimer disease who are primarily compromised on memory, language, and
orientation tasks (Koller & Montgomery, 1997). Research has shown that
individuals with moderate PD are impaired in learning of Tower puzzle tests (e.g.
tower of London, tower of Toronto, tower of Hanoi), a series of visuo-spatial
problems that require high level cognitive planning (Daum et al., 1995; Owen et al.,
1992; Roncacci, Troise, Carlesimo, Nocentini, & Caltagirone, 1996). Using positron
emission tomography (PET) techniques, Owen and colleagues recently
demonstrated a significant increase in regional cerebral blood flow (rCBF) of the
internal segment of the right globus pallidus in healthy adults during the problem
solving of Tower of London task, suggesting that this brain region is essential for
higher cognitive functions (Owen, Doyon, Dagher, Sadikot, & Evans, 1998). In
contrast, rCBF was significantly decreased in this same region for individuals with
moderate PD. The authors concluded that striatal dopamine depletion in PD
disrupts the normal pattern of the basal ganglia outflow that is important for higher
cognitive functions (Owen et al, 1998).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
In summary, evidence from neuroanatomic connections and clinical
symptoms of individuals with basal ganglia disorders suggest that the integrity of the
basal ganglia is important for both cognitive and motor operations. Thus, it is
expected that individuals with basal ganglia disorders would demonstrate impaired
learning when a task is high in either its cognitive or motor demands, or combination
of both.
V. The role of the basal ganglia in motor learning
Neurophvsioloqical evidence
Much of neurophysiological evidence in non-human primate research relates
to the activation levels of neurons in the basal ganglia during motor learning. Two
types of neurons in the basal ganglia have been identified: the medium spiny neuron
(MSP) and tonically active neuron (TAN). Aosaki and colleagues (1994) found that,
in monkeys, numbers of TAN in caudate and putamen increased their firing rate
from 10%-20% to 60%-70%, as the task was learned (Aosaki, Graybiel, & Kimura,
1994). This same group later found that TANs developed a pause in their firing
shortly after the conditioned stimulus was presented. Thus, TANs became
temporally coordinated across large regions of the basal ganglia during
sensorimotor learning. When monkeys were given extinction training, this pause
response diminished and eventually disappeared (Aosaki, Kimura, & Graybiel,
1995). It is believed that dopamine from the SNc is essential for the maintenance of
conditioned responses of TANs. Recently, Miyachi and colleagues examined the
role of the basal ganglia in learning of a button-key press task in monkeys (Miyachi,
Hikosaka, Miyashita, Karadi, & Rand, 1997). Muscimol, a GABA agonist, was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
injected into the animals’ striatum in order to inactivate excitatory activity. Results
demonstrated that motor learning was impaired by muscimol injection, indicating
that activation of neurons in the striatum is critical for motor learning. More
specifically, in the early stage of learning (early in practice), learning was impaired
when muscimol was injected into the anterior part of the striatum. Conversely, in
the later stage of learning when the task was well-practiced, learning was impaired
when muscimol was injected into the middle and posterior striatum. These results
suggest that the anterior and posterior portions of the striatum participate in different
stages of learning of sequential movements.
Similarly, studies using imaging techniques such as PET and magnetic
resonance imaging (MRI) have identified a role of the basal ganglia in motor
learning both in healthy humans and individuals with PD. These imaging techniques
allow direct investigation of the cerebral structures associated with motor skill
learning. Changes in rCBF during different kinds of motor learning tasks (e.g., key
presses, finger sequences) have been investigated. Comparisons of activation
between a well-practiced motor task and an unpracticed motor task, or between the
beginning stage and more advanced stage of learning have been examined.
Several studies have demonstrated significant activation of the basal ganglia in
healthy subjects during motor skill learning (Doyon, Owen, Petrides, Sziklas, &
Evans, 1996; Grafton, Hazeltine, & Ivry, 1995; Jenkins, Brooks, Nixon, Frackowiak,
& Passingham, 1994; Jueptner, Frith, Brooks, Frackowiak, & Passingham, 1997;
Seitz & Roland, 1992). Jenkins and colleagues (1994) used PET to examine the
functional anatomy of motor sequence learning. Subjects were scanned under
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
three conditions: at rest, while performing highly practiced key press sequences,
and while performing new sequences. They found that the putamen was equally
activated for both the new and highly practiced sequence conditions. One problem
was that the rCBF changes in the early and advanced states of learning were
compared with a rest condition that did not involve motor activity. Therefore, it is
difficult to dissociate effects of motor execution from motor learning. Later, Doyon
and associates (1996) reconciled this problem by comparing each sequence
learning condition with a random condition. They found significant activation of the
ventral striatum during learning of the motor sequences, especially in the advanced
stage of the learning process (Doyon et al., 1996). Recently, Jueptner and
colleagues (1997) examined the brain areas that are involved in learning motor
sequences by trial and error using PET. A condition that required learning of a
series of key presses (eight elements long) was compared with conditions in which
the subjects generated moves without learning (free movement condition). Results
showed that peak activation in the putamen was located in the anterior region at the
beginning of practice. The locus of peak activation then switched to the posterior
putamen in the later stage of practice (Jueptner et al., 1997). This anteroposterior
topography shift in human striatum during the early stage to the later stage of
practice is consistent with findings from the neurophysiologic study by Miyachi et al.
(1997) in primates described above.
Evidence from behavioral studies in individuals with PD
As mentioned earlier a clear understanding of the role of the basal ganglia in
motor learning has been limited by a number of issues in previous behavioral
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
studies including both subject inclusion criteria and methodology. In the present
study, it is argued that discrepancies in findings with respect to the motor learning
capability of individuals with PD among previous behavioral work are due to three
critical issues: 1) heterogeneity of disease severity, 2) lack of methods to dissociate
temporary performance effects from learning (absence of retention or transfer), and
3) differences in the level of cognitive and motor demand of the tasks. Next,
findings from previous behavioral studies will be discussed in relation to these three
issues.
Table 1 summarizes findings from previous studies on motor skill learning in
individuals with PD in relation to disease severity, task, and experimental design.
First, for the majority of the studies, disease severity was relatively heterogeneous.
Of 16 studies reviewed, only 1 study included only subjects with moderately severe
PD, Hoehn and Yahr stage II, III, (Pascual-Leone et al., 1993). Six studies included
subjects with mild (Hoehn and Yahr stage I) and moderate (Hoehn and Yahr stage
II, III) PD, and 2 studies included subjects with mild, moderate, and severe (Hoehn
and Yahr stage IV) PD. Six studies did not report subjects’ severity and 1 study
used Webster clinical severity scale instead of the Hoehn and Yahr scale. It has
been shown that the disease severity can influence the motor learning capability of
individuals with PD (Doyon et al., 1997; Harrington, Haaland, Yeo, & Marder, 1990;
Owen et al., 1993). For example, Doyon and colleagues found that subjects with
moderate PD (Hoehn & Yahr stage II, III) but not those with mild PD (Hoehn & Yahr
stage I) showed impaired learning of the SRTT (Doyon et al., 1997).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 1. Summary of behavioral studies of motor skill learning in humans with PD
Authors Subject Hoehn& Yahr
stage of severity
Task Variables
examined
Retention or
Transfer Test
Results
Doyon et al. 15 non-demented PD stage I: 8, repeated sequence test reaction time, indirectly* PD in stage II & III were impaired in the later
1997 15 age-matched con stage II: 4,
stage III: 3
(developed by Nissen &
Bullemer, 1987)
accuracy (practice oyer
several weeks)
stage of learning ( i.e. second half of training
session) compared to controls
Haaland et al.
1997
36 non-demented PD
30 control
stage I: 20%,
stage II: 33%
stage III: 40%
stage IV: 8%
pursuit rotor time on target No PD in the random practice group did not
improve their performance over practice
Soliveri et al.
1997
10 PD
10 age-matched con
stage I: 1
stage II: 5
stage III: 4
pursuit tracking % time on target 1 hr retention PD (on med) demonstrated comparable
learning on the tracking task with controls
Verschueren et 7 PD stage 1:1 bimanual coordination task error relative immediate PD demonstrated impaired bimanual
al. 1997 7 controls stage II: 3
stage III: 3
with 90 deg phase offset
between the two arms
phase retention coordination learning in the non-augmented
but not the enhance augmented feedback
conditions
Agostino et al.
1996
9 PD
8 controls
stage I: 2
stage ll-lll: 5
stage IV: 2
tracing task movement time No PD demonstrated improvement over practice
on the tracing task at the same level as
controls
Jackson et al.
1995
11 non-demented PD
10 age-matched con
NR serial reaction time task,
SRTT (developed by
Nissen & Bullemer, 1987)
response time No PD showed impaired SRTT learning
Note. Only information about PD and controls were included in the table. Other populations such as those with frontal lobe or cerebellar lesions were also examined in
some studies but were not reported here.
NR = not reported, con = controls, * performance at the beginning of the next day practice session could be considered a retention test.
to
■ t*
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 1. Summary of behavioral studies of motor skill learning in humans with PD
Authors Subject Hoehn & Yahr
stage o f severity
Task Variables
examined
Retention or
T ransfer Test
Results
Jordon & Sagar,
1994
26 untreated PD
34 medicated PD
NR grip force error score No Rate of improvement was the same
Between PD and controls
Pascual-Leone
et al. 1993
20 PD
30 age-matched con
stage II & III SRTT (developed by
Nissen & Bullemer, 1987)
response time,
error rate
immediate
transfer
Learning occurred over practice with a
lesser degree for PD than controls
PD showed no or little changes of response
time over practice for longer sequences
Ferraro et al. 17 PD NR SRTT (developed by reaction time immediate PD demonstrated impaired SRTT learning
1993 26 age-matched con Nissen & Bullemer, 1987) transfer compared to controls
Soliveri et al.
1992
21 PD
23 age-matched con
stage I: 3
stage II: 3
stage III: 15
doing up buttons
dual task (doing up buttons
& foot tapping)
time immediate
retention
PD were not impaired in learning these two
tasks compared to controls
Bondi &
Kaszniak, 1991
16 non-demented PD
16 age-matched con
stage I: 4
stage II: 3
stage III: 9
fragmented pictures test
word-stern completion
pursuit rotor
mirror reading
time & number
correct answers
time on target
reading time
No PD were not impaired in pursuit rotor, mirror
reading, and word-stern completion tests
but were impaired in fragmented pictures test
Note. Only information about PD and controls were included in the table. Other populations such as those with frontal lobe or cerebellar lesions were also examined in
some studies but were not reported here.
NR = not reported, con = controls, * performance at the beginning of the next day practice session could be considered a retention test.
cn
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 1. Summary of behavioral studies of motor skill learning in humans with PD
Authors Subject Hoehn& Yahr
stage of severity
Task Variables
examined
Retention or
Transfer Test
Results
Harrington et al. 20 PD stage I: 35%, pursuit rotor mean time indirectly* Only PD in advanced symptom showed
1990 20 age-matched con stage II: 20%,
stage III: 45%
mirror reading
paired associates
mean time
# of correct
& error
practice over
3 days
impaired pursuit rotor learning
mirror reading and paired associates were
normal
Worringham &
Stelmach, 1990
8 PD
8 controls
NR simple and choice RT
task
reaction time indirectly*
practice over
PD were able to improve their performance
in simple and 2-choice RT tasks over
Robertson & 10 PD Use Webster 5-choice button-pressing # of correct & No PD were able to improve their performance
Flowers, 1990 13 controls clinical survey rating task
scale
error rate over practice to the same degree as controls
Heindel et al.
1989
17 PD ( 8 demented-,
9 nondemented-PD)
22 age-matched con
NR pursuit rotor
lexical priming
time on target
% words
completed
No demented PD were impaired both in pursuit
rotor and lexical priming
non-demented PD were intact on both tasks
Frith et al.
1986
12 PD
13 controls
NR tracking task time on target immediate
retention
PD showed normal learning compared to
controls
Note. Only information about PD and controls were included in the table. Other populations such as those with frontal lobe or cerebellar lesions were also examined In
some studies but were not reported here.
NR = not reported, con = controls, * performance at the beginning of the next day practice session could be considered a retention test.
M
CD
27
Second, several studies did not include methods to dissociate temporary
performance effects from learning (absence of retention or transfer tests). Of these
16 studies, 7 studies did not include retention or transfer tests, 6 studies included
only immediate retention or transfer (5 min to 1 hr after acquisition), and the rest of
the studies (n = 3) indirectly measured retention performance (used performance
change over several acquisition sessions).
Third, cognitive and motor demand of a task was not considered in previous
behavioral studies. Tasks that have often been used in previous motor learning
studies are SRTT, pursuit rotor, and tracking/tracing task. Most studies using the
SRTT have demonstrated impaired motor learning in individuals with PD (Ferraro et
al., 1993; Jackson et al., 1995; Pascual-Leone et al., 1993). For example, Jackson
and colleagues found that in SRTT, response time in the repeated sequences for
individuals with PD did not decrease over practice and it did not differ from response
time in the random sequence (Jackson et al., 1995). With respect to level of
cognitive demand, Pascual-Leone and colleagues demonstrated that motor learning
capability of individuals with PD was influenced by the length of the elements in the
SRTT. More specifically, individuals with PD showed learning deficits compared to
age-matched controls in a 12-element length condition but not in 8- and 10-element
length conditions (Pascual-Leone et al., 1993). Similarly, Worringham and
Stelmach found that individuals with PD learned a two-choice reaction time task as
well as control subjects (Worringham & Stelmach, 1990). However, they showed
learning deficits in 4-choice and 8-choice reaction time. Response to a sequence
with long element lengths in SRTT requires greater cognitive demand compared to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
that with shorter element lengths. Thus, these findings suggest that the level of
cognitive demand is critical in determining the learning capability of individuals with
PD.
Studies using the pursuit rotor task showed mixed results. Heindel and
coworkers found that learning of the pursuit rotor task was not impaired for the non-
demented PD group but impaired for the demented PD group (Heindel, Salmon,
Shults, Walicke, & Butters, 1989). In this study, speed of the turntable was adjusted
for each subject to give a baseline performance of 25% time on target. This
baseline adjustment resulted in average speed of the turntable approximately at 47
rpm for the PD group and 55 rpm for the control group. This difference in the task
motor demand made it difficult to compare results between the PD and control
groups. Recall the study by Harrington and colleagues that demonstrated impaired
pursuit rotor learning for individuals with PD compared to age-matched controls. A
closer examination revealed that: 1) the locus of this impairment was at the highest
speed (60 rpm condition), and 2) the impairment was seen primarily in individuals
with PD who had relatively severe bradykinesia (Harrington et al., 1990). Using a
modified version of the pursuit rotor task (computer based), Bondi and Kazniak
(1991) found that the ability to learn the pursuit rotor task for subjects with PD was
normal compared to that of controls. As mentioned earlier, these two studies (Bondi
& Kazniak and Harrington et al.) were different in speeds of the turntable and the
order of practice. Therefore, it is not clear whether discrepancy in findings between
these two studies were due to the differences in motor demand or differences in
cognitive demand or a combination of both. Harrington and colleagues attempted to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
test the cognitive demand hypothesis by having subjects with PD perform the task at
three different speeds (15, 45, and 60 rpm) in one of two practice orders, blocked or
random (Haaland, Harrington, O'Brien, & Hermanowicz, 1997). They found that
individuals with PD in the random order practice group did not improve their
performance (i.e. increased time on target) over practice while those in the blocked
order practice group showed comparable performance improvement with control
subjects. The authors suggested that the basal ganglia are important for cognitive
operations necessary for motor skill learning. One problem with this recent study by
Haaland and coworkers is that, like a number of previous studies, retention or
transfer tests were not included. Thus, it is not possible to delineate temporary
performance effects from learning.
V I. Summary
The lack of agreement regarding the role of the basal ganglia in motor
learning across previous behavioral work using individuals with PD are due to three
critical issues: heterogeneity of disease severity, differences in the level of cognitive
and motor demand of the tasks, and lack of methods to dissociate temporary
performance effects from learning (absence of retention or transfer). The extent and
severity of cognitive and motor impairments associated with PD are dependent on
the stage of the disease as measured by the Hoehn and Yahr (1967) scale (Doyon
et al., 1997). Thus, it is critical that severity of PD is controlled for or taken into
account when conclusions about the role of the basal ganglia are made. Previous
work has failed to take motor and cognitive demands of the tasks into consideration.
The capability to acquire motor skills requires both cognitive and motor processes
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
(Haaland & Harrington, 1990; Salmon & Butters, 1995). Evidence from
neuroanatomic connections and clinical symptoms of individuals with basal ganglia
disorders suggest that the integrity of the basal ganglia is essential for both
cognitive and motor operations. Thus, it is not surprising that motor learning deficits
in individuals with PD appear to be evident in conditions when either the cognitive
and/or motor demand of the task was high. However, a study that is systematically
designed to control for the level of cognitive and motor demand of the task is
needed to test this hypothesis.
By controlling the subject inclusion criteria and methodological issues
described above, the present behavioral study is intended to corroborate and clarify
the neurophysiologic studies that have demonstrated a role for the basal ganglia in
motor learning. To this end, the subjects with PD included in this study were only
those with moderately severe PD (i.e. Hoehn & Yahr stage II and III). The level of
motor and cognitive demands of the task used in this study was systematically
manipulated in order to distinguish the task as requiring either low or high motor
and/or cognitive processing. The following four practice conditions were
established: high cognitive-high motor; high cognitive-low motor; low cognitive-high
motor; and low cognitive-low motor. Task cognitive demand was manipulated by
altering practice order (i.e. Cl effect) and feedback frequency provided during the
acquisition phase. Task motor demand was manipulated by altering the amount of
effector force required to move the manipulandum. The specific details of how Cl,
feedback frequency, and amount of force were used to manipulate the level of
cognitive and motor demand of the tasks will be covered in the methods section of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
Chapter two. Finally, to dissociate the temporary performance effects from the
relatively permanent effects of practice (i.e. learning), performance during retention
and transfer tests was used to index motor skill learning.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
CHAPTER 2:
MOTOR SKILL LEARNING IN INDIVIDUALS WITH PARKINSON’S DISEASE:
CONSIDERATION OF COGNITIVE AND MOTOR DEMANDS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
Abstract
The purpose of this study was to investigate procedural motor learning
capability of individuals with Parkinson’s disease (PD). Unlike previous work, the
level of cognitive and motor demand of the task was systematically manipulated. In
the first part of the study, motor learning capability was indexed using a measure of
overall movement error during a retention test. It was hypothesized that the motor
learning capability of individuals with PD is dependent upon the level of cognitive
and motor demand of the task. More specifically, individuals with PD would
demonstrate motor learning deficits when either cognitive and/or motor demand of
the task was high. In contrast, they would demonstrate comparable learning to age-
matched control subjects when both cognitive and motor demand of the task was
low. In the second part, the processes by which individuals with PD acquire a motor
skill were investigated. A measure of motor program error was differentiated from
that of movement parameter error. It was hypothesized that motor program learning
would be impaired but parameter learning would be preserved in individuals with PD
compared to healthy age-matched controls.
Twenty adults with moderately severe PD and 20 age-matched
neurologically healthy individuals participated. The task required the subject, using
the right dominant arm, to rotate a lever at the correct speed and amplitude to
replicate a goal movement trajectory to be displayed on the monitor before each
trial. The goal was to learn a rapid arm movement with extension-flexion reversal
actions. One half of subjects from each group practiced the tasks under a low
cognitive demand (LC) condition and the other half practiced the tasks under a high
cognitive demand (HC) condition. Overall, there were 4 subgroups: control- LC
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
(Con-LC), control-HC (Con-HC), PD-LC, and PD-HC. Cognitive demand was
manipulated using practice order and feedback (FB) frequency. Blocked practice
order and 100% FB characterized the LC condition. Random practice order and
60% faded FB characterized the HC condition. All participants practiced the same
spatially scaled trajectory with three different motor demands (low, medium, and
high). Motor demand was manipulated through the variable movement time (MT)
criteria. The 900 ms MT condition represented the high motor demand task while
the slower MT, 1200 and 1500 ms conditions, represented the medium and low
motor demand tasks, respectively. Participants practiced 3 sets of a 45-trial set
each day for two days. Learning was assessed during the no FB retention and
transfer tests on Day 3. There were two retention test conditions (blocked and
random). During the transfer test, the practiced trajectory was MT-scaled to 1050
and 1350 ms.
In the first analysis, motor performance was quantified across acquisition,
retention, and transfer phases using root mean square error (RMSE), which is the
difference between the goal trajectory and subject’s response. All participants
improved their performance over practice (Block Effect; p < .001). During
acquisition, performance was not different between the two groups (Group Effect; p
= .07). Retention performance indicated that the level of cognitive and motor
demand of the task played a critical role in determining the motor learning capability
of individuals with PD. For the LC practice condition (blocked practice order and
100% FB), individuals with PD were as accurate as age-matched controls in
reproducing the 1500 ms trajectory for both blocked (RMSE for Con-LC = 13.9, PD-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
LC = 14.4 deg) and random retention (RMSE for Con-LC = 15.9, PD-LC = 16.3 deg)
tests. When motor demand was increased (i.e. 900 ms trajectory), the PD group
performed with greater error than age-matched controls for both blocked (RMSE for
Con-LC = 13.3, PD-LC = 22.6 deg) and random retention (RMSE for Con-LC = 18.1,
PD-LC = 24.8 deg). These findings resulted in a significant Group X Motor Demand
interaction; p < .02. For the HC practice condition (random practice order and 60%
faded FB), individuals with PD showed greater error than their age-matched controls
when tested in blocked retention for both high and low motor demand trajectories
(900 ms, RMSE for Con-HC = 13.3, PD-HC = 22.0 deg; 1500 ms, RMSE for Con-
HC = 16.8, PD-HC = 19.0 deg). In contrast, the PD group performed as well as the
control group when tested in random retention for both motor demand conditions
(900 ms trajectory, RMSE for Con-High = 16.9, PD-High = 18.9 deg; 1500 ms
trajectory, RMSE for Con-High = 16.3, PD-High = 16.6 deg). These findings
resulted in a significant Group X Retention Condition interaction; p < .01. During the
transfer test, individuals with PD showed greater error than control subjects (RMSE
for PD group = 20.07, control group = 16.53 deg), Group Effect; p = .03.
In the second analysis, a method similar to Sullivan and Winstein (submitted)
was used to separate motor program (GMP) error from movement parameter (time
and amplitude) errors. Motor program learning capability of individuals with
moderately severe PD was affected primarily by the cognitive demand. During
retention, GMP errors did not differ between the two groups for those who practiced
under the HC condition (Group Effect; p = .67). GMP errors were greater for the PD
group than the control group for those who practiced under the LC condition (Group
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
Effect; p < .04). Unlike GMP learning where cognitive demand was critical, temporal
parameter learning was affected primarily by motor demand. When motor demand
was low, individuals with PD showed similar temporal parameter error to control
subjects across all conditions. When motor demand was high, individuals with PD
showed condition-specific temporal parameter learning deficits. Specifically, when
practiced under the LC condition, time parameter error was greater for the PD group
than for the control group on both retention tests. When practiced under the HC
condition, time parameter error was greater for the PD group than for the control
group when tested in the blocked retention but similar between the two groups when
tested in the random retention. There were no group differences or group
interactions for spatial parameter learning across all conditions. Finally, during the
transfer test, time parameter error was slightly greater for the PD group than for the
control group.
Results from the present study indicated condition-specific motor learning
deficits in individuals with PD. These findings suggest that it is important to consider
the cognitive demand, motor demand, and context of retention tests when
evaluating the learning capability of individuals with moderately severe PD.
Individuals with PD can learn a motor task as well as neurologically healthy age-
matched control subjects if the level of motor and cognitive demand of the task is
relatively low (i.e. within their capabilities). When the level of motor demand is
increased, impaired motor learning in individuals with PD is revealed. Impaired
motor learning under the high motor demand condition, however, can be overcome
if individuals with PD practice the task under a high cognitive demand condition.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
This suggests that a high cognitive demand practice enhances motor learning for
individuals with PD. However, the learning is less flexible and is context-dependent.
It is proposed that the context-dependent learning seen in individuals with PD is due
to a set-shifting deficit characteristic of PD (Sandson & Albert, 1987). Nigrostriatal
degeneration due to dopamine depletion in PD may disrupt the normal function of
the dorsolateral prefrontal-striatal circuit, previously thought to be responsible for the
set-shifting function.
Motor program learning capability in individuals with moderately severe PD
was primarily affected by the cognitive demand. The high cognitive demand
practice condition enhanced motor program learning in PD. The high cognitive
demand practice condition may have allowed individuals with PD to compensate for
their impaired implicit memory system by using the intact explicit memory system for
motor program learning. The ability to learn temporal parameters for subjects with
PD was primarily affected by the motor demand of the task. Spatial parameter
learning of a discrete movement was preserved in individuals with PD. Finally, the
ability to generalize temporal parameter learning appeared to be impaired in
individuals with PD.
These results suggest important implications for rehabilitation. Consistent
with a recent clinical review and case study (Morris, 2000), task specific training
appears to be an effective approach for rehabilitation of individuals with PD,
especially when the task involves high cognitive and motor demands.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
Introduction
Parkinson’s disease (PD), a progressive degenerative disorder due to
dopamine depletion in the basal ganglia, affects approximately 1 in 1000 adults
(Barker & Dunnett, 1999). The cardinal clinical signs of PD are tremor, rigidity,
bradykinesia (slowness of movements), and postural instability. Traditionally, PD is
known as a movement disorder. It has been well established that motor behavior is
impaired in individuals with PD. Slowness of movement (Flash, Inzelberg,
Schechtman, & Korczyn, 1992; Phillips, Martin, Bradshaw, & lansek, 1994;
Winstein, Pohl, Cormack, & Waters, 1995), delayed movement initiation (Brown &
Marsden, 1990; Contreras-Vidal, Teulings, & Stelmach, 1995), hypometric
movement (Castiello, Stelmach, & Lieberman, 1993; Contreras-Vidal et al., 1995;
Flowers, 1976), and difficulty switching between movements (Bennett, Marchetti,
lovine, & Castiello, 1995; Castiello, Bennett, Adler, & Stelmach, 1993; Onla-or &
Winstein, 2001; Weiss, Stelmach, & Hefter, 1997) have been demonstrated in
individuals with PD. While motor control has been a primary focus of research
investigated the role of the basal ganglia, many recent studies have focused on the
importance of the basal ganglia in cognition, learning and memory. Recent
neurophysiologic and behavioral studies have revealed cognitive deficits such as
dementia, impaired attentional set shifting, and impaired problem solving in
individuals with PD (e.g. Brown, Schneider, & Lidsky, 1997; Tamaru, 1997; Wise,
Murray, & Gerfen, 1996). Given that the integrity of the basal ganglia is essential for
both motor and cognitive operations, it is logical then to expect that procedural
motor learning, a capability to acquire motor skills that has been characterized as
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
requiring both motor and cognitive processes (Haaland & Harrington, 1990; Salmon
& Butters, 1995), would be impaired in individuals with PD.
Results from behavioral studies in individuals with PD regarding the role of
the basal ganglia in procedural motor learning, however, have been inconsistent.
For example, individuals with PD demonstrated impaired procedural motor learning
compared to that of age-matched controls for a serial reaction time task, SRTT,
(Ferraro, Balota, & Connor, 1993; Jackson, Jackson, Harrison, Henderson, &
Kennard, 1995; Pascual-Leone et at., 1993; Worringham & Stelmach, 1990), pursuit
rotor (Harrington, Haaland, Yeo, & Marder, 1990), tracking (Frith, Bloxham, &
Carpenter, 1986), and bimanual coordination (Verschueren, Swinnen, Dorn, & De
Weerdt, 1997). In contrast, other studies showed intact procedural motor learning in
individuals with PD for grip force (Jordan & Sagar, 1994), sequence learning
(Robertson & Flowers, 1990), pursuit rotor (Heindel, Salmon, Shults, Walicke, &
Butters, 1989; Koenig, Thomas-Anterion, & Laurent, 1999), and tracking/tracing
tasks (Agostino, Sanes, & Hallett, 1996; Frith et al., 1986; Soliveri, Brown,
Jahanshahi, Caraceni, & Marsden, 1997). It appears that a clear understanding of
the role of the basal ganglia in procedural motor learning has been limited by
various factors such as subject inclusion criteria and methodology. A closer
examination suggests that the inconsistent findings among previous studies are
related to three critical factors: heterogeneity of disease severity, lack of distinction
between performance effects and learning (absence of retention or transfer), and
differences in the level of cognitive and motor demand of the to-be-learned tasks.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
First, disease severity is relatively heterogeneous for the majority of previous
studies. Doyon and colleagues have demonstrated that the extent and severity of
the PD-related cognitive and motor impairments are dependent on the stage of the
disease as measured by the Hoehn and Yahr (1967) scale (Doyon et al., 1997). In
that study, individuals with moderate PD (Hoehn and Yahr stage II and III) showed
impaired learning of a repeated sequence test compared to age-matched controls
while those with mild PD (Hoehn and Yahr stage I) showed normal learning (Doyon
et al., 1997). Despite the apparent effect of disease severity, several studies
included relatively heterogeneous PD populations (e.g. Agostino et al., 1996; Bondi
& Kaszniak, 1991; Haaland et al., 1997; Soliveri et al., 1997). Variability in disease
severity may have confounded findings in these previous works.
Second, only a few studies have dissociated practice effects from learning.
Analyses of performance changes during the acquisition phase are not a reliable
indicator of learning since these changes may reflect temporary effects of practice
manipulations that do not persist when those manipulations are withheld. The
standard approach to assess learning is to evaluate performance during a delayed
retention or transfer test without feedback provided. Although the important
distinction between practice effects and learning have been emphasized in the
motor learning literature (e.g. Salmoni, Schmidt, & Walter, 1984; Schmidt & Bjork,
1992), many of the previous studies of motor learning in PD did not include a
retention test or transfer test in their design (e.g. Agostino et al., 1996; Haaland et
al., 1997; Heindel et al., 1989; Robertson & Flowers, 1990). In a retention test, a
learner performs the same task that has been practiced while in a transfer test, the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
learner switches to perform a different task or the same task with different
parameters (e.g. amplitude, movement time). Although retention and transfer tests
are similar in that they both capture the persistence of the acquired movement
representation, the processes that are optimal for retention are suggested to be
different from those optimal for transfer (Lee, 1988; Morris, Bransford, & Franks,
1977). Impaired ability to generalize the learned movement in a transfer test has
previously been reported in individuals with PD (Verschueren et al., 1997).
Third, more importantly and directly the focus of the present study is that
previous work has not considered the motor and cognitive demand of the task. The
capability to acquire motor skills requires both cognitive and motor processes
(Haaland & Harrington, 1990; Salmon & Butters, 1995). Lee and coworkers (1994)
emphasized that the cognitive component of motor skills has a critical impact on the
learning process. Cognitive demand in motor skill learning refers to the mental work
involved in decision-making processes regarding the anticipation, planning,
regulation, and interpretation of motor performance (Lee, Swinnen, & Serrien,
1994). Motor demand refers to the kinetic and kinematic requirement of the to-be-
performed movement. Practice tasks in previous work range from those that require
relatively high cognitive but low motor demand such as SRTT and button-pressing,
low cognitive but high motor demand such as pursuit rotor, tracing, and tracking
tasks, to those that require relatively high cognitive and motor demand such as
bimanual coordination tasks. Thus, it is difficult to compare findings across studies
or extract any general conclusions since the various tasks have different levels of
cognitive and motor demand.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
Cognitive effort during motor skill learning can be influenced by variables
such as practice order and feedback frequency (Lee et al., 1994). Manipulation of
the order in which tasks are practiced (blocked, random, serial) is shown to have a
significance impact on the learner’s cognitive processing. This phenomenon is
known as the contextual interference (Cl) effect. It refers to the interference effects
in performance and learning that arise from practicing one task in the context of
other tasks (Shea & Zimny, 1983). Three practice orders have been commonly
used: random, blocked, and serial. In random practice order, all tasks are
presented in a random order in each practice set (e.g. C-A-B, A-B-C, B-C-A). In
blocked practice order, one task is repeatedly presented within each set of trials
before a new task is presented in the next set (e.g. B-B-B, A-A-A, C-C-C). Finally,
in serial practice order, tasks are presented in a sequential order and this order is
repeated throughout each set of practice (e.g. A-B-C, A-B-C, A-B-C). It has been
shown that random and serial practice order produces high contextual interference
while blocked practice order produces low contextual interference (Shea & Morgan,
1979; Lee & Magill, 1983; Shea, Kohl, & Indermill, 1990). Studies examining the
influence of Cl on motor learning have typically shown that a high Cl practice
condition (i.e. random order) degrades acquisition performance but benefits
retention performance (i.e. facilitates learning). In contrast, a low Cl practice
condition (i.e. blocked order) benefits acquisition performance but degrades
retention performance (Magill & Hall, 1990). Shea and Morgan (1979) had healthy
young subjects practice three motor tasks (i.e. knock down barriers with a tennis
ball) under a blocked (low Cl) or random order (high Cl). Retention and transfer
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
tests were performed after a 10-min and 10-day delay under blocked and random
orders. Results showed that, during retention, subjects who practiced in the high Cl
condition (random practice order) performed relatively well under both blocked and
random retention tests. In contrast, subjects who practiced in the low Cl condition
(blocked practice order) showed performance deterioration especially under the
random retention test. Likewise, performance in the transfer phase was superior for
subjects who practiced in the high Cl condition compared to those who practiced in
the low Cl condition.
Two mutually compatible hypotheses, the distinctiveness and elaboration
hypothesis and the reconstruction hypothesis, were proposed to explain the benefit
of high Cl on motor learning. According to the distinctiveness and elaboration
hypothesis, performers who practice in random order (high Cl) are encouraged to
take advantage of both intertask and intratask cognitive processes while those who
practice in blocked order (low Cl) rely primarily on intratask cognitive processes
because information about different tasks are rarely processed together in working
memory (Shea & Morgan, 1979; Shea & Zimny, 1983). Therefore, practice in
random order results in a more distinctive and elaborate memory representation.
The reconstruction hypothesis suggests that, in random order, performers have to
re-solve the problem and reconstruct the action plan in each trial because the
previous solution has been forgotten due to the intervening practice trials of the
other tasks (Lee & Magill, 1983; 1985). In contrast, during blocked order practice,
performers can recall a recent solution from working memory. This latter strategy
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
(i.e. repeating the solution) may suppress or even inhibit the cognitive effort of
problem solving that is necessary during random order practice.
Another practice variable shown to have large impact on the learner’s
cognitive processing demand is feedback frequency. Feedback (FB) is often
provided in the form of knowledge of results (KR). Knowledge of results refers to
the extrinsic information about task success that is provided to the performer after a
practice trial has been completed. Evidence from several studies has suggested
that practice with a lower relative frequency of KR is as effective or may even be
more beneficial for learning compared to practice with 100% KR frequency (e.g.
Winstein, Merians, & Sullivan, 1999; Winstein & Schmidt, 1990; Wishart & Lee,
1997). One interpretation is that less frequent FB might enhance learning by forcing
the subject to engage in cognitive processing necessary to perform the task. In
contrast, high frequency FB provides the solution for the learner and diminishes the
cognitive requirement (Schmidt & Bjork, 1992; Schmidt, Young, Swinnen, & Shapiro,
1989). In addition, when external FB is always provided, subjects may become
dependent on the FB and not utilize the cognitive operations required for solving the
task problem (Schmidt, 1991).
In contrast to cognitive demand where information processing operations are
manipulated, task motor demand can be influenced by the level of force that the
motor units have to generate within a specific time. According to Newton’s law: F =
ma (F = force, m = mass, a = acceleration of segment center of mass), when mass
is constant, force is proportional to acceleration. Thus, motor demand is higher in
conditions when the nervous system has to produce a given average force to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
complete an action rapidly (i.e. high acceleration) compared to a condition in which
more time is available to produce the same average force (i.e. low acceleration).
Reduced force generation is one of the problems that have been demonstrated in
individuals with PD (Phillips et al., 1994). Subjects with PD demonstrated
decreased agonist burst activity and compensated by using sequential bursts of
agonist/antagonist activity to produce adequate force necessary for movement
completion (Hallett & Khoshbin, 1980).
The purpose of this present study was to investigate the role of the basal
ganglia in procedural motor learning. In the first part (Analysis I), the effects of
cognitive and motor demand on motor learning capability of individuals with PD
compared to healthy age-matched controls were investigated. A measure of overall
motor performance was used as the dependent variable in this analysis. In the
second part (Analysis II), the processes by which individuals with PD acquire a
motor skill were investigated. In this analysis, a measure of motor program error
was differentiated from that of movement parameter error. It was hypothesized that
motor program learning would be impaired but parameter learning would be
preserved in individuals with PD compared to healthy age-matched controls.
Limitations of previous work in relation to subject inclusion criteria and methodology
were addressed. To this end, the subjects with PD included only those with
moderate severity (Hoehn & Yahr stage II and III). Performance during retention
and transfer tests was used to index procedural motor learning. Finally, the level of
cognitive demand was directly manipulated using practice order and FB frequency
and the level of motor demand was directly manipulated by varying the speed of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
required movement. Random practice order with low FB frequency characterized
the high cognitive demand (HC) condition while blocked practice order with high FB
frequency characterized the low cognitive demand (LC) condition. Movement with
high velocity (short movement time) represented the high motor demand task while
that with low velocity (long movement time) represented the low motor demand task.
To my knowledge, the present study is the first designed to systematically
manipulate the level of cognitive and motor demand of the to-be-learned task.
Methods
Subjects
Twenty adults with PD and 20 age-matched neurologically healthy
individuals participated. All participants with PD had bilateral basal ganglia
involvement with no clinical evidence of dementia or depression. Inclusion criteria
for individuals with PD were: 1) Hoehn and Yahr stage II or III (Hoehn & Yahr, 1967,
see Appendix A); 2) properly medicated for PD; 3) score at least 28 on the Mini-
Mental State exam (Folstein, Folstein, & McHugh, 1975, see Appendix B); and 4)
score less than 10 on the Center for Epidemiologic Studies Depression short form
(CESD-10) scale (Andresen et al., 1994, see Appendix C). Exclusion criteria
included any of the following conditions determined from a screening interview or
medical records as available: 1) any acute medical problems, 2) uncorrected vision
loss, and 3) history of admission to psychiatric hospital within the past 3 years.
Individuals with PD were tested during the “on” medication cycle. Participants with
PD were recruited from the Movement Disorders Clinic at the University of Southern
California Healthcare Consultation Center (USC HCC) and Parkinson Support
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
Group in Los Angeles, Sherman Oaks, and Long Beach areas. Control subjects
were recruited from the local community, patients’ spouse or caregivers. All
participants performed the task with their dominant right hand (Oldfield, 1971, using
80% cut off on the Edinburgh Handedness Inventory, see Appendix D). All subjects
scored at least 20/40 on the Rosenbaum Vision Screener.
Experimental Design
Group and cognitive demand were the between-factor variables whereas
motor demand was the within-factor variable. This 2 X 2 between-factor design—2
groups (control, PD) x 2 cognitive demand (Low, High) resulted in 4 subgroups:
control-low cognitive demand (Con-LC), control-high cognitive demand (Con-HC),
PD-low cognitive demand (PD-LC), and PD-high cognitive demand (PD-HC). All
subjects practiced the same spatially scaled trajectory with three different motor
demands (low, medium, and high).
Cognitive demand of the task was manipulated using practice order and
feedback frequency. Blocked practice order and 100% FB represented the LC
condition. Random practice order and 60% faded FB represented the HC condition.
Motor demand of the task was manipulated through the variable movement time
(MT) criteria. The 900 ms MT condition represented the high motor demand task
while the slower movements, 1200 ms and 1500 ms, represented the medium and
low motor demand tasks, respectively.
Testing took place over 3 consecutive days with acquisition phase on Days 1
and 2 and retention and transfer phase on Day 3. There were 6 sets of 45 trials
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
during acquisition (3 sets each day), resulting in a total of 270 practice trials.
Learning was assessed based on performance during the retention and transfer
phase on Day 3. There were two retention test conditions (blocked and random
retention). The order of the blocked and random retention tests was
counterbalanced across subjects. During the retention test, participants performed
the same movements they practiced the day before but without any feedback. For
the transfer test, the practiced trajectory was scaled to 1050 and 1350 ms. These
two new temporally scaled trajectories were alternately presented during the
transfer test (i.e., 1050, 1350, 1050, 1350.....) without any feedback provided.
Within each group (control, PD), participants were sequentially assigned to
either a LC or HC condition until the total number of subjects (n = 10) in each
condition was obtained. In the LC condition, trajectories with one temporal scaling
were presented repeatedly in a set (i.e. blocked order, 1200-1200-1200 ..., 900-
900-900 ..., 1500-1500-1500 ...) and augmented feedback was provided after every
trial (100% FB). In the HC condition, trajectories with different temporal scaling
were pseudo-randomly presented so that within a set each trajectory was presented
15 times (i.e. random order; 1200-900-1500 ..., 900-1500-1200 ..., 1500-1200-900
...) and 60% faded FB was provided for each set. Faded feedback was scheduled
as follows: after every trial (100%) for trials 1-15, 60% for trials 16-30, and 20% for
trials 31-45 in each set.
Instrumentation and Task
A lightweight lever affixed to a frictionless vertical axle was attached to a
table and positioned parallel to the floor. A handle at the end of the lever was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
adjusted to accommodate participant’s forearm. A linear potentiometer was
attached to the transducer at the base of the vertical axle. A signal from the
transducer (an analog signal) was converted to a digital signal by an A-D board of a
Dell 466v computer and sampled at 200 Hz. The Template software program (D.
Hary, 1996) was used to manipulate the movement trajectory, the interval duration,
and to store data from each trial for further analyses.
The motor task was to move the lever at the correct speed and distance to
replicate a goal movement trajectory, which was displayed on the computer monitor
before each trial. The goal was to learn a continuous arm movement with two
extension-flexion reversal actions. This goal movement was to be performed under
three different movement time (motor demand) conditions (i.e. 900, 1200, and 1500
ms), Figure 1. During acquisition, after each movement trial, the participant was
900 1200 1500 ms
!
I
50 -
1 30:
o
8 20 -
a
5 10-
-10
0 300 600 900 1200 1500
Time (ms)
Figure 1. Three goal movement trajectories with the same spatial scaling but
different temporal scaling at 900, 1200, and 1500 ms.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
presented with either the trial number together with two kinds of augmented FB, or
the trial number alone (no FB). The two kinds of feedback were 1) an overall error
score, root mean square error (RMSE), representing the difference between the
goal movement trajectory and the subject’s response and 2) a graphic
representation of the subject’s response superimposed with the goal movement
trajectory (Figure 2b). Figure 2a shows the experimental set-up.
A “Ready” signal (yellow visual cue and a tone) was displayed on the monitor for
300 ms. After the 300 ms interval, a “Go” signal (green visual cue and a tone) was
displayed. To encourage pre-programming, the subjects had to initiate the
movement within 500 ms of the “Go” signal. If the subject moved prior to 100 ms of
the “Go” signal, they received an error message “ too early”. If the subject moved
later than 500 ms of the “Go” signal, they received an error message “move earlier”,
indicating that the movement started too late. The beginning and end of the
movement were determined when the subject crossed the 1.0° (movement onset
threshold) and 0° (movement offset threshold) baseline detection, respectively.
Total trial collection time was 4000 ms.
Procedures
Participants read and signed an Institutional Informed Consent Form prior to
the beginning of the study. Prior to testing, several brief screening tests for vision,
memory, depression, and handedness were performed. Eligible subjects answered
a questionnaire for health status- the physical activity portion of SF36 (Appendix E).
Participants were seated in front of the computer monitor with their forearm resting
on the lever in approximately 90 degrees of elbow flexion. Subjects were instructed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
M onitor
RMSE= 31.8
s u b je c t
80
1200 ms Goal
60
40
20
0
200 400 600 800 1000 1200 0
Time (ms)
Figure 2. Experimental set-up with arm lever and feedback display. A) Subject sits
in front of the monitor with arm resting on the lever. Upper right shows an overhead
view of the subject with arm on the lever at the starting position. B) An example of
the two forms of augmented visual feedback, a graphic representation of the
subject’s movement trajectory for that trial superimposed over the goal trajectory,
and the RMSE score on the top right corner. The goal trajectory is displayed as a
thick solid line and the subject’s trajectory is displayed as a thin dash line.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
in how to move the lever and how to use the feedback information on the monitor to
improve their movements on successive trials. The experimenter explained the
procedure of the task and encouraged the subjects to make as small an error as
possible and to superimpose their movement response with the goal movement
trajectory. They were instructed that they would either receive feedback after every
trial (100% FB) or receive progressively less feedback over the practice session
(60% faded FB) depending on practice condition assignment. Before the testing
started, a sample of 10 trials with movement trajectories similar to the goal
movement trajectories was displayed and the experimenter demonstrated how to
move the lever. Subjects, however, were not allowed to practice prior to the testing
session.
Analysis I
Hypotheses
Hypothesis # 1. For the LC practice condition, when task motor demand is
high, the PD group will exhibit higher movement errors compared to the control
group during the delayed retention test (i.e. impaired motor learning). Conversely,
when task motor demand is low, the PD group will exhibit similar movement errors
to the control group during the delayed retention test (i.e. intact motor learning).
Hypothesis # 2. For the HC practice condition, when either task motor
demand is low or high, the PD group will exhibit higher movement errors compared
to the control group during the delayed retention test (i.e. impaired motor learning).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
Hypothesis # 3. During the transfer test, movement errors will be higher for
the PD group than for that of the control group (i.e. impaired generalizability of motor
learning).
Outcome Measures and Statistical Analyses
For the first analysis, root mean square error (RMSEf), which is the average
difference between goal movement trajectory and the subject’s response, calculated
over the subject’s total movement time was the outcome measure. RMSE is a
measurement of an overall motor performance accuracy. Individual RMSE data
were grouped into 9-trial blocks for the acquisition phase (Blocks 1-10) and 5- trial
blocks for the retention (Blocks 11-12) and transfer (Block 13) phases. Group
(control and PD) and cognitive demand (low and high) were the between-factor
variables. Motor demand (low, medium, and high) was a within-factor variable.
General linear model (GLM) analyses were used to detect any significant main
effects and interactions. Post hoc analysis with a Bonferroni correction was used to
specify the locus of any significant interactions. Further, independent two sample t-
tests were conducted to assess differences on the demographic characteristics (i.e.,
age, sex, education level, cognitive, and health status) between participants with PD
and controls within each cognitive demand condition. For all statistical tests,
significance level was set at p < 0.05.
t RMSE = S {(X j — Tj)2 /n}1 /2
i = 1
X|= participant’s position in degrees at time I, Ti = target position at time i
n = the number of samples for the subject’s trajectory array
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
Acquisition Performance
A 2 Group (control, PD) X 2 Cognitive Demand (low, high) X 2 Motor
Demand (low, high) * X 10 Block analysis of variance (ANOVA) with repeated
measures on the last factor was conducted to provide an overall description of the
acquisition performance for the two groups.
Retention Performance
Performance during the retention test was used to index motor learning.
Two different methods of retention measurement; absolute and relative retention
were used. Absolute retention, the most common and scientifically justifiable
measure of retention, refers to the level of performance on the initial trials of the
retention test that is not based in any way on the acquisition performance (Schmidt&
Lee, 1999). Relative retention refers to the differences between acquisition and
retention performance that represent the amount of loss or gain in skill.
Absolute Retention.
Retention data from each cognitive demand condition were analyzed
separately. Therefore, retention data for the PD-LC subgroup were compared to the
Con-LC subgroup and those for the PD-HC subgroup were compared to the Con-
HC subgroup. A 2 Group (control, PD) X 2 Motor Demand (low, high) X 2 Retention
Test (blocked, random) Analysis of Variance (ANOVA) was conducted on mean
RMSE to address the questions regarding the motor learning capability of
individuals with PD compared to control subjects. Further, to compare within group,
* Although three discrete trajectory movement times (i.e., 900, 1200, and 1500 ms) were
used to grade motor demand, I was particularly interested in the two extreme motor demand
conditions (i.e., 900 and 1500 ms trajectory). The 1200 ms trajectory was included in the
design in order to reduce the number of repetitive presentations of the same trajectory (e.g.,
900, 900, 1500, 900, 900, 1500...) in the random practice condition. Therefore, only data
from the high and low motor demand conditions were included in the analyses.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
a 2 Cognitive Demand (LC, HC) X 2 Motor Demand (low, high) X 2 Retention Test
(blocked, random) ANOVA was conducted on mean RMSE of subjects within the
same group.
Relative Retention.
Relative retention (comparisons between acquisition and retention
performance) was also conducted. The first comparison (end of acquisition-first
retention) was done to determine performance deterioration at retention when
feedback was removed compared to performance at the last block of acquisition.
The second comparison (first acquisition-first retention) was done to determine the
overall performance improvement compared to the baseline [beginning]
performance level. A 2 Group (control, PD) X 2 Motor Demand (low, high) X 2
Retention Test (blocked, random) ANOVA was conducted on each relative retention
data.
Transfer Performance
To access generalizability of motor learning, a 2 Group (control, PD) X 2
Cognitive Demand (low, high) X 2 Motor Demand (low, high) ANOVA was
conducted on mean RMSE during the transfer test.
Results
Participant Demographic Characteristics
The demographic characteristics of each participant with PD are
summarized in Table 1. Group mean comparisons for the demographic
characteristics are summarized in Table 2. Group comparisons from each cognitive
demand condition were made separately. For those in the LC practice condition
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
(Con-LC and PD-LC), there were no differences between groups for age, education,
or explicit cognitive level (Mini-Mental state). Mean age was 66 years (range 59-70
yrs) for the control group and 61.4 years (range 50-74 yrs) for the PD group.
Healthy control subjects were more active (as indicated by responses to the health
status questionnaires) than participants with PD. For those in the HC practice
condition (Con-HC and PD-HC), there were no differences between groups for age
or physical activity level. Mean age was 60 years (range 35-72 yrs) for the control
group and 61.7 years (range 40-72 yrs) for the PD group. Participants with PD had
a higher education level than controls. Mean score on the Mini-Mental exam was
lower for participants with PD than controls. None of the participants with PD,
however, had clinical sign of dementia (Mini-Mental score < 28). Mean score on the
CES-D was higher for individuals with PD than controls for both practice conditions.
Again, none of the participants with PD had clinical evidence of depression (CES-D
score > 10). The disease severity was similar for the two PD subgroups (mean
Hoehn and Yahr stage: PD-LC = 2.6, PD-HC = 2.5). Average PD duration was 7.3
years (range 3-15 yrs) for the PD-LC subgroup and 10.2 years (range 4-20 yrs) for
the PD-HC subgroup.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 1. Demographic characteristics for each participant with PD
Subject # Age
(yrs)
Sex Education
(yrs)
MMSE SF 36 CES-D Hoehn & Yahr
(1-5)
PD
Duration (yrs)
Medication Surgery
PD-LC1 66 M 13 30 27 7 3 15 Sinemet, M irapex none
PD-LC2 72 M 16 29 25 9 2.5 7 Sinem et, Tasm ar none
PD-LC3 74 F 16 28 26 3 2 1 Sinem et none
PD-LC4 69 M 18 29 24 10 2.5 9 M irapex none
PD-LC5 57 M 16 29 26 5 2.5 8 Sinem et, Eldepryl none
PD-LC6 56 M 17 30 27 7 2.5 3 Sinem et, Eldepryl, Tasm ar none
PD-LC7 50 M 16 28 13 8 3 5 sinem et, Eldepryl, M irapex pallidotom y
PD-LC8 52 M 15.5 30 19 9 3 12 Sinemet, Requip pallidotom y
PD-LC9 51 F 18 30 19 10 2.5 9 Sinemet, M irapex none
PD-LC10 68 M 12 30 18 5 2 4 Sinemet, M irapex none
PD-HC1 59 M 20 28 23 6 2 6 Sinem et none
PD-HC2 62 M 18 30 28 4 2 4 M irapex none
PD-HC3 53 M 20 30 22 5 3 12 Sinem et, Mirapex, Tasm ar pallidotom y
PD-HC4 60 M 16 29 30 2 2 6 Sinem et, Tasm ar none
PD-HC5 67 M 18 30 25 7 2.5 6 Sinem et none
PD-HC6 69 M 16 28 16 7 3 20 Sinem et none
PD-HC7 68 F 18 29 22 8 2.5 15 Sinem et, Mirapex, Tasm ar none
PD-HC8 67 M 20 28 23 10 2 4 Sinem et none
PD-HC9 40 M 16 28 25 4 3 17 Sinem et, M irapex none
PD-HC10 72 M 18 28 26 7 3 12 Sinemet, M irapex none
Note LC = low cognitive dem and practice condition, HC = high cognitive dem and practice condition, M MSE = M ini-M ental state exam (m axim um
score = 30), SF 36 = health status-physical activity portion (m axim um score = 30), CES-D = center for epidem iologic studies depression short form
(m axim um score = 30) o i
~ s l
58
Table 2. Group mean comparisons for the demographic characteristics.
Assessment
Group by Practice Condition
Con-LC PD-LC
P
Con-HC PD-HC
P
Sample size 10 10 10 10
Age (yrs) 66.0 ± 3.2* 61.4 ±9.4 .17 60 ± 10.9 61.7 ±9.5 .71
Education (yrs) 16.4 ± 2.2 15.7 + 1.9 .49 15.0 + 1.6 18.0 ±1.6 .01
Cognitive:
MMSE (max 30)
29.5 ± 0.5 29.3 ± 0.8 .52 29.6 ± 0.5 28.8 ± 0.9 .03
Health Status:
SF 36 (max 30)
26.4 ± 2.9 22.4 ± 4.8 .04 26.1 ±2.7 24.0 ± 3.8 .17
Emotional Status:
CES-D (max 30)
3.4 ± 2.9 7.2 + 2.3 .01 2.9 ±1.4 6.0 ±2.7 .01
Hoehn and Yahr
(1-5)
NA 2.6 ± 0.6 NA 2.5 ±0.7
PD Duration
(yrs)
NA 7.3 ±4.2 NA 10.2 ±5.8
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
Acquisition Performance
RMSE block means for each group by cognitive and motor demand during
acquisition are displayed in Figure 3. All participants improved their performance
with practice (Block Effect; p < .001). Overall, performance was not different
between the two groups (Group Effect; p = .07). Participants who practiced in the
LC condition (i.e., Con-LC and PD-LC, top row) were generally more accurate than
those who practiced in the HC condition (i.e., Con-HC and PD-HC, bottom row),
Cognitive Demand Effect; p < .001. Similarly, participants were more accurate in
reproducing the low motor demand task (i.e., 1500 ms trajectory, right column) than
that of the high motor demand task (i.e., 900 ms trajectory, left column), Motor
Demand Effect; p < .001. Across the two practice conditions, individuals with PD
had higher error than controls for the high motor demand but not for the low motor
demand condition (Group X Motor Demand interaction; p < .05). This group
difference for the high motor demand is due primarily to the high error of the PD-LC
subgroup during the first half of practice (block 2-5, Figure 3 top left). The Group X
Cognitive Demand interaction was not significant (p = .64), suggesting that, during
acquisition, individuals with PD benefited from practice order and feedback
frequency the same way as control subjects.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
High Motor Demand
(900 ms)
Low Motor Demand
(1500 ms)
Con-LC PD-LC |
C O
E
< D
Q
CD
>
■w
C
CD
O
O
28 i 28 n
24 - 24 -
20 -
16 -
12 • 12 ■
I
~ i----------- r ~
9 10 1 2 3 4 5 6 7 9 10 8 1 2 3 4 5 6 7 8
Acquisition (9 trials/block)
900 ms
Acquisition (9 trials/block)
1500 ms
Con-HC PD-HC
T3
C
C O
E
0 )
Q
< D
>
C
CD
O
O
-C
CD
L U
W
5
Q C
28 -j 28 “|
24 - 24 -
I
I
20 ■ 20 -
16 -
16 -
12 ■
12 -
10
2 5 7 9 10 1 3 4 6 8
Acquisition (9 trials/block)
900 ms
Acquisition (9 trials/block)
1500 ms
Figure 3. RMSE block means during acquisition days 1 (block 1-5) and 2 (block 6-
10) for Control (closed symbol) and PD (open symbol) groups. Top row illustrates
data from the low cognitive demand condition (blocked practice order with 100% FB)
as the triangle symbol and bottom row illustrates those from the high cognitive
demand condition (random practice order with 60% faded FB) as the square
symbol. Left column illustrates data from the high motor demand condition (900 ms
trajectory) and right column illustrates those from the low motor demand condition
(1500 ms trajectory). RMSE is shown in degrees. Error bars are standard error of
mean (SEM).
Retention Performance
Absolute Retention.
For the LC practice condition (i.e. blocked practice order with 100% FB),
individuals with PD showed similar RMSE to age-matched controls when the motor
demand of the task was low (i.e. 1500 ms trajectory). Specifically, the PD group
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
61
was as accurate as the control group in reproducing the 1500 ms trajectory for both
blocked retention (RMSE for Con-LC = 13.9, PD-LC = 14.4 deg) and random
retention (RMSE for Con-LC = 15.9, PD-LC = 16.3 deg) tests, Figure 4, top right.
When task motor demand was increased (i.e. 900 ms trajectory), the PD group
performed with significantly greater error than the control group for both blocked
retention (RMSE for Con-LC = 13.3, PD-LC = 22.6 deg) and random retention
(RMSE for Con-LC = 18.1, PD-LC = 24.8 deg) tests, Figure 4, top left. These
findings resulted in a significant Group X Motor Demand interaction; p < .02. A
Student’s independent samples f-test identified the locus of the interaction in the
high motor demand task (p < .02). Both groups performed with slightly lower error
in blocked than in random retention.
For the HC practice condition, while the control group performed with lower
error in the blocked retention than in the random retention, the PD group performed
with lower error in the random retention than in the blocked retention. This opposite
pattern resulted in a significant Group X Retention Condition interaction; p < .01. A
Student’s independent samples f-test identified the locus of the interaction in the
blocked retention (p < .03). Specifically, the PD group exhibited higher error
compared to the control group when tested in the blocked retention but exhibited
comparable error to the control group when tested in the random retention
independent of the motor demand conditions. For the high motor demand (900 ms
trajectory), RMSE was significantly higher for the PD group than the control group in
blocked retention (RMSE for Con-HC = 13.3, PD-HC = 22.0 deg) but not in random
retention (RMSE for Con-HC = 16.9, PD-HC = 18.9 deg); Figure 4, bottom left.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
62
T S
C
C D
£
< D
Q
< D
>
C
O )
o
o
24 -
16 -
High Motor Demand
(900 ms)
28
24 •
20
16 '
12 -
blocked-ret random-ret
Retention Test Condition
Low Motor Demand
(1500 ms)
Con-LC
-A- PD-LC
blocked-ret random-ret
Retention Test Condition
ID
c
(1 3
E
c i >
Q
< D
>
O ) W
o s
O §
J Z
O )
28 I
24 *
16 '
28
24
20
12
blocked-ret random-ret
Retention Test Condition
T
G .
Con-HC
-G - PD-HC
blocked-ret random-ret
Retention Test Condition
Figure 4. Mean RMSE during retention tests (blocked and random) for Control
(closed symbol) and PD (open symbol) groups. Top row illustrates data from the
low cognitive (LC) demand practice condition and bottom row illustrates those from
the high cognitive (HC) demand practice condition. Left column illustrates data from
the high motor demand condition (900 ms trajectory) and right column illustrates
those from the low motor demand condition (1500 ms trajectory). RMSE is shown in
degrees. Error bars are SEM. There was a significant Group X Motor Demand
interaction (p < .02) for the LC condition and a significant Group X Retention
Condition interaction (p < .01) for the HC condition.
Results for the low motor demand (1500 ms trajectory) paralleled those for the high
motor demand but with a smaller magnitude of group difference. The PD group
exhibited greater error than the control group when tested in the blocked retention
(RMSE for Con-HC = 16.8, PD-HC = 19.0 deg). In contrast, they were as accurate
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
as controls when tested in the random retention (RMSE for Con-HC = 16.3, PD-HC
= 16.6 deg), Figure 4, bottom right. Findings that the PD-HC subgroup performed
relatively well (compared to controls) in the random retention (where the context of
task presentations was the same as that in the practice condition) but not in the
blocked retention suggest that the HC practice condition evoked context-dependent
learning in individuals with PD. This context-dependent learning effect in the PD-HC
subgroup was evidenced independent of the order of the retention test condition (i.e.
subjects performed the random retention before or after the blocked retention),
Figure 5.
28
24
20
16
12
8
1 2 3 4 5 6 7 8 9 10 11 12
Acquisition Retention
subject 01, 900 ms goal trajectory
32
28
24
20
16
12
8
4
1 2 3 4 5 6 7 8 9 10 11 12
Acquisition Retention
subject 05, 900 ms trajectory
Figure 5. Example of individual data from the PD-HC subgroup over acquisition and
two retention tests. Left graph shows data from participant PD-HC1 who performed
the random retention test first and then the blocked retention test. Right graph
shows data from a different participant (PD-HC5) who performed the blocked
retention test first then the random retention test. Participants with PD who
practiced in the high cognitive demand condition showed higher error in the blocked
retention than in the random retention independent of the order of which retention
test they performed first.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
28
24 -
I 20
U l
w
16
12 -
-A t- Con-LC
28 -
- A - PD-LC
Con-HC - B PD-HC
o >
< D
2-
U J
w
blocked-ret random-ret
Retention Test Condition
24
20
16
12
blocked-ret random-ret
Retention Test Condition
Figure 6. Mean RMSE (averaged across two MT trajectories) for each group by
cognitive demand condition. Data from the control subgroups (Con-LC, Con-HC)
are shown on the left and those from the PD subgroups (PD-LC, PD-HC) are shown
on the right. The control group showed lower error in blocked than in random
retention independent of the cognitive demand practice condition (Retention Effect;
p < .003). In contrast, the PD group showed lower error in the similar practice-
retention context than in the different practice-retention context. This resulted in a
significant Cognitive Demand X Retention interaction (p < .004). RMSE is shown in
degrees. Error bars are SEM. LC = low cognitive demand, HC = high cognitive
demand.
Figure 6 illustrates retention data (averaged across motor demand
conditions) for the two cognitive demand conditions compared within each group.
For the control group (left), RMSE errors were similar between the two cognitive
demand conditions. Further, both control subgroups (Con-LC, Con-HC) showed
lower error in blocked relative to random retention (Retention Effect; p < .003).
Statistical analysis revealed no significant Cognitive Demand main effect or
interactions, suggesting that the two cognitive demand practice conditions had
similar effects on retention performance of control subjects. Unlike the control
group, the pattern of findings in retention was different between the two PD
subgroups. Specifically, subjects with PD who practiced in the LC condition
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
(blocked order, 100% FB) showed lower error in blocked relative to random
retention. In contrast, subjects with PD who practiced in the HC condition (random
order, 60% faded FB) showed lower error in random relative to blocked retention.
This resulted in a significant Cognitive Demand X Retention interaction for the PD
group; p < .004. Post hoc analyses identified a significant difference between
blocked and random retention for the PD-HC subgroup (p < .02) and a marginally
significant difference for the PD-LC subgroup (p = .07) with a moderate effect size
(ES = 0.4). Findings that the PD group performed with lower error in the similar
practice-retention context than in the different practice-retention context suggest a
context-dependent learning in individuals with PD.
Relative Retention.
Figure 7 illustrates the change in mean RMSE between the last acquisition
block and the first block of each retention test for each group by cognitive and motor
demand. A positive change reflects a decrease in error and a negative change
reflects an increase in error. Overall, there were no group differences related to
performance changes between the end of acquisition and the first of each retention
block, Group Effect; p = .32. Participants who practiced in the HC condition (Con-
HC, PD-HC, bottom row) showed a smaller degree of performance deterioration at
retention when feedback was withheld than those who practiced in the LC condition
(Con-LC, PD-LC, top row), Cognitive Demand Effect; p < .001. For the LC practice
condition (top row), across the two groups and retention test conditions, error was
significantly greater for the 900 ms trajectory than for the 1500 ms trajectory (Motor
Demand Effect; p = .04).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
" O
c
CO
E
CD
Q
CD
>
'c
O )
o
o
S
o
•o
c
CO
E
C D
Q
C D
.>
’c
O )
o
o
J C
O )
ID
c n
2 ■
-2 -
-10
-14
UJ
CO
5
c c
High Motor Demand
(900 ms)
blocked-ret random-ret
900 ms Trajectory
-14
a
o
A
c
o
w
■=3
cr
o
t o
( A
to -1 0 -
-14
blocked-ret random-ret
900 ms Trajectory
Low Motor Demand
(1500 ms)
6 •
2 -
-2 '
-6 -
-10 -
B
Con-LC
PD-LC
blocked-ret random-ret
1500 ms Trajectory
Con-HC
PD-HC
-10 -
-14
blocked-ret random-ret
1500 ms Trajectory
Figure 7. Bar graphs illustrate mean RMSE differences between last block of
acquisition and first block of each retention test for Control and PD groups. A
positive change reflects decreased in error and a negative change reflects
increased in error from acquisition to each retention test. Top row illustrates data
from the low cognitive demand condition and bottom row illustrates those from the
high cognitive demand condition. Left column illustrates data from the high motor
demand condition (900 ms trajectory) and right column illustrates those from the low
motor demand condition (1500 ms trajectory). Overall, subjects in the high cognitive
demand condition (bottom row) showed smaller magnitude of performance
deterioration from last acquisition block to each retention test when feedback was
withheld compared to those in the low cognitive demand condition (top row);
Cognitive Demand Effect, p < .001. RMSE is shown in degrees. Error bars are
SEM.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
67
~ o
c
r o
E
< D
Q
< U
>
O )
o
O
UJ
( O
12 1
High Motor Demand
(900 ms)
Low M otor Demand
(1500 ms)
blocked-ret random-ret
900 ms Trajectory
H i Con-LC
□ PD-LC
12 i
random-ret
1500 ms Trajectory
T 3
0 )
g >
12 n
o
o
X>
c
o
o >
x:
O )
a >
C D
c
C O
blocked-ret random-ret
900 ms Trajectory
Y Z A Con-HC
□ PD-HC
12 -]
blocked-ret random-ret
1500 ms Trajectory
Figure 8. Bar graphs illustrate mean RMSE differences between first block of
acquisition and first block of each retention test for Control and PD groups. A
positive change reflects decreased in error and a negative change reflects
increased in error from the beginning of practice [baseline performance]. Top row
illustrates data from the low cognitive demand condition and bottom row illustrates
those from the high cognitive demand condition. Left column illustrates data from
the high motor demand condition (900 ms trajectory) and right column illustrates
those from the low motor demand condition (1500 ms trajectory). Overall, subjects
in the high cognitive demand condition (bottom row) showed larger magnitude of
performance improvement from baseline compared to those in the low cognitive
demand condition (top row); Cognitive Demand Effect, p < .01. RMSE is shown in
degrees. Error bars are SEM.
Figure 8 illustrates the change in mean RMSE between the first acquisition
block and the first block of each retention test for each group by cognitive and motor
demand. Again, a positive change reflects performance improvement (decrease
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
68
error) and a negative change reflects performance deterioration (increase error).
Overall, participants who practiced in the HC condition (bottom row), showed the
greatest performance improvement (from the first practice block) compared to those
who practiced in the LC condition (top row), Cognitive Demand Effect; p < .01. For
the LC condition, the magnitude of performance improvement from early practice
was very similar between the two groups for the low motor demand task but
appeared to be lower for the PD group than that of the control group for the high
motor demand task. Statistical analysis of these findings was, however, not
significant (Group X Motor Demand interaction; p = .37). For the HC condition,
individuals with PD showed less improvement than did the control subjects when
tested in the blocked retention but showed about the same magnitude of
improvement as controls when tested in the random retention. This pattern of
findings was evident for both motor demand tasks (900 and 1500 ms trajectories)
and resulted in a significant Group X Retention Test interaction; p < .01. Thus,
similar to the absolute retention findings, the HC practice condition appears to evoke
context-dependent learning in individuals with PD.
Transfer Performance
Generally, RMSE was higher for the PD group (20.1 deg) compared to the
control group (16.5 deg), Group Effect; p = .03, Figure 9a. Group RMSE differences
were more pronounced for the 1050 ms movement trajectory (RMSE control = 16.9,
PD = 22.3 deg) than for the 1350 ms trajectory (RMSE control = 16.2, PD = 17.9
deg), Figure 9b. Although Group X Motor Demand interaction did not reach
significance (p = .18), the effect size was large (ES = .77).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69
A. B.
■ ill Control
E Z I'pd
28
24 -
20
16 -
12
T
G .
T
Control
-e - PD
T
1050 ms 1350 ms
Trajectory
Figure 9. Mean RMSE during transfer test for Control and PD groups. A) Bar
graphs show mean RMSE for each group averaged across cognitive demand
conditions and two new temporally scaled trajectories (1050 and 1350 ms). Group
difference was significant at p = .03. B) Group mean RMSE for each trajectory.
Overall, group differences were most pronounced for the 1050 ms trajectory.
Although, Group X Motor Demand interaction did not reach significance (p = .18),
the effect size was large (ES = .77). RMSE is shown in degrees. Error bars are
SEM.
Discussion
Effects of Motor and Cognitive Demand on Motor Learning in Individuals with PD
The results of this study suggest that level of the motor and cognitive
demand of a task plays a critical role in determining the motor learning capability of
individuals with PD. In agreement with the hypothesis, when cognitive and motor
demand of the task was low (i.e. blocked practice order, 100% FB and 1500 ms
trajectory), individuals with moderate PD showed comparable learning to age-
matched control subjects. However, when the cognitive demand was low but the
motor demand was increased (i.e. blocked practice order, 100% FB and 900 ms
trajectory), individuals with PD demonstrated a distinct motor learning deficit
compared with controls. These results are in agreement with findings from previous
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
studies by Harrington et al. (1990). Harrington and coworkers found that deficits in
learning the pursuit rotor for individuals with PD were more pronounced in
conditions when the turntable was rotated at the highest velocity, 60 rpm (i.e. high
motor demand) compared with the lower velocities, 30 and 45 rpm (Harrington et al.,
1990).
Results for the low cognitive demand condition primarily reflected the effect
of motor demand. It is likely that in the low cognitive demand condition, the
cognitive requirement was within the limits of cognitive processing capability of
individuals with moderate PD. Other results from procedural motor learning using
the SRTT support this interpretation. For example, Pascual-Leone and colleagues
showed that decreases in response time over practice for subjects with PD were
inversely related to the SRTT sequence length (Pascual-Leone et al., 1993).
Similarly, Worringham and Stelmach found that individuals with PD showed normal
learning (compared to control subjects) on simple and 2-choice reaction time but
showed impaired learning on 4- and 8-choice reaction time tasks (Worringham &
Stelmach, 1990). Traditionally, the basal ganglia have been associated with motor
operations, however, recent evidence suggests that they also have a critical role in
cognitive operations (e.g., Middleton & Strick, 2000; Owen & Doyon, 1999; Owen,
Doyon, Dagher, Sadikot, & Evans, 1998). Individuals with moderate PD
demonstrated learning deficits in the Tower of London test, a series of visuo-spatial
problems that require a high level of cognitive planning (Owen et al., 1992). Using
positron emission tomography (PET) techniques, Owen and colleagues recently
demonstrated a significant increase in regional cerebral blood flow (rCBF) of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
internal segment of the right globus pollidus in healthy adults in relation to solving
the Tower of London task, suggesting that this brain region is essential for higher
cognitive functions (Owen et al., 1998). In contrast, and for the same problem,
rCBF was significantly decreased in this same region for individuals with moderate
PD. The authors concluded that striatal dopamine depletion in PD disrupts the
normal pattern of the basal ganglia outflow that is important for higher cortical
functions (Owen et al, 1998).
Significance of the level of motor and cognitive demand on motor learning in
individuals with PD offers a viable explanation for differences in findings between
two previous studies (Bondi & Kaszniak, 1991; Harrington et al., 1990) which used
the pursuit rotor task. Both motor (i.e. velocity of the pursuit rotor) and cognitive
demands (i.e. order of task presentation) were lower in Bondi and Kaszniak
compared to those in Harrington et al. Velocities of the pursuit rotor were either 7 or
14 rpm in Bondi and Kaszniak’s study while they were at 30, 45, and 60 rpm in
Harrington et al.’s study. Further, only one fixed velocity was presented (i.e. low Cl)
in Bondi and Kazniak while three different velocities were randomly presented (i.e.
high Cl) for each set of practice in Harrington et al. Not surprisingly, Bondi and
Kaszniak, reported intact motor learning ability in individuals with PD on the pursuit
rotor task while Harrington and colleagues reported impaired learning ability for
individuals with PD (Bondi & Kaszniak, 1991; Harrington et al., 1990).
While results support the hypothesis that motor learning would be intact in
individuals with moderate PD when motor and cognitive demand is low, they provide
only partial support for the hypothesis that motor learning would be impaired in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
individuals with PD when motor and cognitive demand is high. For the high
cognitive demand condition, context of the retention test (i.e. whether it is the same
as or different from the practice condition) has a critical effect on the motor learning
capability of individuals with PD. More specifically, individuals with PD showed
comparable learning (as reflected by retention performance) to age-matched
controls on both low and high motor demand tasks when tested in the same context
as practice (random retention) but showed learning deficits on both motor demand
tasks when tested in a different context from practice (blocked retention). The
finding that subjects with PD could perform as well as controls for the high motor
demand task suggests that motor demand is less detrimental to learning for
individuals with PD if they had practiced the task in a high cognitive demand
condition. Thus, the high cognitive demand practice enhances motor learning for
individuals with PD. However, the PD group showed comparable learning to
controls only when tested in the same context as practice, suggesting that this
motor learning capability was less flexible and was context-dependent.
Context-Dependent Learning in Individuals with PD
Findings that context-dependent learning was revealed solely in the PD
group suggest that this behavior reflects the dysfunction of the basal ganglia. Only
a few studies have addressed the issue of task or context-dependent learning in
individuals with PD. Previously, practice-specific learning in individuals with PD was
reported by Verschueren and colleagues (1997). In that study, individuals with PD
practiced a bimanual arm movement coordination task under an enhanced vision
condition. During the retention test, the PD group showed comparable learning to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
that of the control group in the enhanced vision condition but showed impaired
learning in both normal vision and reduced vision retention conditions (Verschueren
et al., 1997). The mechanisms underlying this context-dependent learning in
individuals with PD are not clear. The authors proposed that this context-dependent
learning was due to a difficulty in generating movement responses from internal
cues (self-guided) and an increased reliance on external cues. Thus, when external
cues were withheld during the retention tests, they showed learning deficits.
Consistent with this idea, difficulties in generating movements from an internal (self-
guided) action plan and increased reliance on external cues have been
demonstrated in individuals with PD (e.g. Brown & Marsden, 1988; Cools, van den
Bercken, Horstink, van Spaendonck, & Berger, 1990; van Spaendonck, Berger,
Horstink, Borm, & Cools, 1996) and animals with experimental lesions to the
caudate (Saint-Cyr, Taylor, & Nicholson, 1995).
The hypothesis that context-dependent learning in individuals with PD is due
to the difficulty in self-generated action plan and increased reliance on external
cues, however, could not explain findings in the present study. The two retention
test conditions (blocked and random retention) required subjects to: 1) generate
movements using self-guided and 2) rely on intrinsic feedback to evaluate their own
performance. Thus, according to this hypothesis, context-dependent learning in
individuals with PD should not be observed in the present study. Retention
performance should be poorer for the PD group than the control group on both
blocked and random retention tests. Therefore, results from the present study
suggest that other factor(s) must account for context-dependent learning in PD.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
74
In this study, it is proposed that context-dependent learning is due to a set-
shifting deficit in individuals with PD. A set-shifting deficit or stuck-in-set-
perseveration is a characteristic that has long been identified in individuals with PD
(e.g. Cronin-Golomb, Corkin, & Growdon, 1994; Owen, Roberts, Polkey, Sahakian,
& Robbins, 1991; Sandson & Albert, 1987). A set develops when an individual
repeatedly performs one type of mental or motoric operation. The set-shifting deficit
is defined as an inability to suppress learned responses that are no longer
appropriate in a changed contextual environment (Wise et al., 1996). For example,
individuals with PD continued to sort cards by color after it was clear that they
should be sorting cards by word meaning or continued to draw a circle in response
to a circle cue after the rule was changed to be drawing a square in response to a
circle cue (Brown & Marsden, 1988; Sandson & Albert, 1987). In the present study,
during testing, several subjects with PD verbally commented that they needed to
move slower (or faster), yet their performance on the next trials did not change (i.e.
they seemed stuck in the set they previously learned). It is thought that the
dorsolateral prefrontal circuit, one of the basal ganglia-thalamocortical circuits
described by Alexander and colleagues (Alexander, Crutcher, & Delong, 1990) is
responsible for this set-shifting function (Cronin-Golomb et al., 1994; Wise et al.,
1996). Nigrostriatal degeneration due to dopamine depletion in PD may disrupt the
normal function of this circuit. Therefore, a set-shifting impairment is evidenced in
subjects with PD.
The set-shifting deficit hypothesis could account for previous findings by
Verschueren and colleagues (1997). Recall that all participants practiced the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
bimanual coordination task under the enhanced vision condition in that study. Not
surprisingly, individuals with PD demonstrated learning of the bimanual coordination
task (compared to controls) when tested in the enhanced vision retention condition.
In contrast, they demonstrated impaired motor learning (compared to controls) when
tested in the retention conditions that had a different context from the practice
condition (i.e. normal vision and reduced vision retention).
Generalizabilitv of Motor Learning
Findings for the transfer test indicated that generalizability of learning was
impaired in individuals with PD. In the transfer phase, the movement trajectories
differed from those in the acquisition phase only in their absolute movement time
(i.e. difference in motor demand). The motor demand of these new trajectories was,
however, within the range of that in the practiced trajectories (i.e. movement time
between 900 and 1500 ms). Although trajectories were presented in serial order
during the transfer phase, it has been shown that serial order practice produces high
Cl effects similar to that of random order practice (Lee & Magill, 1983; Shea et al.,
1990). Given that subjects in the PD-HC subgroup were able to perform the 900 ms
task as well as controls in the random retention condition, the relatively poorer
performance in the transfer test is likely due to the inability of individuals with PD to
generalize to a new temporal goal, rather than a direct effect of the motor or
cognitive demand. It has been suggested that the processes that are optimal for
retention may be different from those optimal for transfer (Lee, 1988; Morris et al.,
1977). In support of this view, Winstein and colleagues (1994) demonstrated that, in
healthy young adults, a practice condition with a relatively high frequency of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
76
knowledge of results, while allowing processing appropriate for retention, did not
support processing appropriate for transfer (Winstein, Pohl, & Lewthwaite, 1994). In
the present study, a practice condition with random order and 60% faded FB while
allowing processing appropriate for a random retention test, did not support
processing appropriate for a transfer test for individuals with PD. Specifically, the
PD group showed comparable performance to the control group in the retention
condition with the same context as the practice condition (random retention), their
performance was significantly poorer than controls during transfer.
In this first analysis, a global measure of overall movement error (RMSE)
was used to examine the motor learning capability of individuals with moderately
severe PD. While these findings demonstrate the importance of cognitive and motor
demand of the task on the motor learning capability of individuals with PD, they did
not provide insight into the processes by which individuals with PD acquire a motor
skill. According to Schmidt’s schema theory, learning of a motor skill requires that
the performers acquire both a generalized motor program (GMP) and develop the
capability to parameterize the GMP in temporal and spatial domains to meet task
constraints or environmental demands (Schmidt, 1975). The separation of the
programming and parameterization components of a task cannot be done when
global measures such as RMSE or movement time are used as dependent
variables. To address whether or not the processes by which individuals with PD
acquire a motor skill are different from healthy age-matched control subjects, a
method similar to that used by Wulf et al. (1993) and Sullivan and Winstein
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
77
(submitted) was employed to differentiate between motor program and movement
parameter components in the second analysis.
Analysis II
The generalized motor program (GMP) is a fundamental movement
representation characterized by invariant features such as relative time and relative
force. Task variations that have the same relative timing or force are considered to
be in the same class of movements and thus governed by the same GMP. In
contrast, task variations that have different relative timing or force structures are
considered to be in a different class and governed by a different GMP. Parameters
are variant features by which the GMP is scaled, such as overall duration or overall
amplitude of a given movement. Scaling of these parameters can be achieved
without changing the GMP. Thus, skilled learning of a rapid goal-directed
movement involves acquiring the GMP and scaling it to meet the temporal and
spatial accuracy requirements. However, the separation of the programming and
scaling components of a task cannot be done when global measures such as RMSE
or movement time are used as dependent variables.
Wulf, Schmidt, and Debubel (1993) modified a method initially developed by
Winstein (1988) to differentiate between GMP error and parameterization error.
This method involves scaling the subject’s trajectory in time and amplitude until the
agreement between the subject’s trajectory and goal trajectory is maximized. This
method results in 1) a residual RMSE, which is the remaining RMSE after temporal
and spatial scaling (i.e. GMP error), and 2) a timing factor and amplitude factor (i.e.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
78
parameter error). The residual RMSE method has allowed researchers to
demonstrate that program and parameter learning are differentially affected by
practice condition and feedback frequency (e.g. Lai, Shea, Wulf, & Wright, 2000;
Sekiya, Magill, & Anderson, 1996; Sekiya, Magill, Sidaway, & Anderson, 1994; Wulf,
Lee, & Schmidt, 1994; Wulf & Schmidt, 1994; Wulf, Schmidt, & Deubel, 1993).
These findings suggested that programming and parameter scaling are dissociable
constructs mediated by different control processes. Recently, Lai and colleagues
demonstrated a hierarchy for GMP and parameter learning—with a stable GMP
being a requisite for effective parameterization (Lai et al., 2000).
Numerous investigations have demonstrated that the ability to learn motor
skills is impaired in individuals with PD (e.g. Ferraro, Balota, & Connor, 1993;
Harrington, Haaland, Yeo, & Marder, 1990; Jackson, Jackson, Harrison, Henderson,
& Kennard, 1995; Pascual-Leone et al., 1993; Verschueren, Swinnen, Dorn, & De
Weerdt, 1997). However, the source(s) of those deficits have yet to be determined.
Do individuals with PD have difficulty in acquiring the motor program, developing
effective movement parameters, or a combination of both? The purpose of this
analysis was to examine whether programming and/or parameterization deficits
underlie the motor learning differences between individuals with PD and controls.
Deficits in motor programming/planning have often been shown in individuals with
PD (e.g., Flowers, 1978; Harrington & Haaland, 1991; Jennings, 1995; Morris et al.,
1988). However, others have found that motor programming is intact in individuals
with PD (Gentilucci & Negrotti, 1999; Tresilian, Stelmach, & Adler, 1997; Weiss,
Stelmach, & Hefter, 1997). For example, a kinematic study of reaching and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
grasping an object, and placing it on a second target in individuals with PD showed
that reaction time and peak acceleration of the initial reaching phase did not differ
between subjects with PD and controls, suggesting a preserved motor program in
PD (Gentilucci & Negrotti, 1999). More specifically, PD and control groups took into
account properties of both the first and second target when programmed the
movement. However, later in execution, subjects with PD re-programmed the
movement as indicated by changes in peak velocity. The effects of the second
target were removed from the new motor program and peak velocity of subjects with
PD was influenced only by the first target. The authors postulated that
reprogramming occurred because the motor program decayed during its execution,
that is, individuals with PD have a deficit in storing and maintaining a memory of the
action plan (i.e. learning deficit) but not in building the action plan (i.e. control
deficit). Thus, it is hypothesized that in the present analysis, individuals with PD will
demonstrate impaired GMP learning.
While bradykinesia (slowness of movement) and hypometria (reduction of
movement) are fundamental manifestations of PD, the mechanisms underlying
these two clinical symptoms are not well understood (Hallett & Khoshbin, 1980;
Marsden, 1989). It has been suggested that bradykinesia is a strategy for
individuals with PD to maintain movement accuracy— individuals with PD
deliberately reduced their initial forces to compensate for their unpredictable
movements and to enable them to make the necessary corrections (Martin, Phillips,
lansek, & Bradshaw, 1994; Van Gemmert, Teulings, Contreras-Vidal, & Stelmach,
1999). However, a recent study has demonstrated that bradykinesia could be
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
improved with practice and this speed improvement was retained as evidenced by
performance in a delayed retention test (Behrman, Cauraugh, & Light, 2000).
Therefore, it was hypothesized that movement time would be longer in individuals
with PD. However, this slowness of movement in PD would decrease with practice
and the improvement would be retained in a delayed retention test. Hypometria is
thought to be related to an inability of individuals with PD to maintain constant force
amplitude over time (Van Gemmert et al., 1999). Typically, hypometria has been
demonstrated in continuous or sequential movements such as reaching and
grasping, hand writing, reciprocal tapping, and walking (e.g. Castiello, Stelmach, &
Lieberman, 1993; McLennan, Nakano, Tyler, & Schwab, 1972; Morris, lansek,
Maty as, & Summers, 1996; Onla-or & Winstein, 2001). When a discrete movement
task was used, hypometria was not evident in individuals with PD (Hoshiyama,
Kaneoke, Koike, Takahashi, & Watanabe, 1994). Thus, in the present analysis,
given that the to-be-learned task was a discrete movement, hypometria in subjects
with PD was not expected.
Although retention and transfer tests are similar in that they both capture the
persistence of the learned movement representation, the processes that are optimal
for retention are suggested to be different from those optimal for transfer (e.g. Lee,
1988; Morris, Bransford, & Franks, 1977; Winstein, Pohl, & Lewthwaite, 1994).
Consistent with previous work (Verschueren et al., 1997), findings in the first
analysis revealed that the ability of individuals with PD to generalize learned
movement to different temporal goals in the transfer test was impaired. Recall that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
a global measure of overall movement error (i.e. RMSE) was used as the dependent
measure in the first analysis. For the second analysis, however, it is further
hypothesized that impaired generalizability of learning in PD will be revealed
specifically in the parameter component but not in the GMP component of the motor
skill.
Hypotheses
Hypothesis # 1. Motor program learning will be impaired in individuals with
PD compared to healthy age-matched control subjects (i.e. during the retention test,
GMP error will be greater for the PD group than for that of the control group).
Hypothesis # 2. Parameter learning will be preserved in individuals with PD
compared to healthy age-matched control subjects.
- The ability to learn temporal parameters of the movements will be similar
between the PD and control groups (i.e. during the retention test, TF will not differ
between the two groups).
- The ability to learn spatial parameters of the movements will be similar
between the PD and control groups (i.e. during the retention test, AF will not differ
between the two groups).
Hypothesis # 3. The generalizability of temporal parameter learning will be
impaired in individuals with PD compared to healthy age-matched control subjects
(i.e. during the transfer test, TF error will be greater for the PD group than for that of
the control group).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
82
Outcome Measures and Statistical Analyses
The outcome measures were motor program (GMP) error and parameter
scaling (time and amplitude) error. A method similar to Wulf et al. (1993) and
Sullivan and Winstein (submitted) was used to separate these two error
components.
Motor Program Error (GMP error).
A computer software program (D. Hary, 2000) was developed to scale
(stretch or compress) the subject’s trajectory in time and amplitude in order to
maximize the agreement between the subject’s movement and goal movement
pattern. The subject’s trajectory and goal trajectory were synchronized when the
subject's movement exceeded 1.0 deg displacement from baseline. The subject’s
trace was then proportionately scaled from 0.2 to 2.0 by a time factor increment of
0.01 such that the subject’s trace was interpolated to the number of samples in the
goal trace. This process normalized the subject’s trajectory to the temporal goal by
removing the temporal error component from the subject’s trajectory. After scaling
the subject’s trajectory in time, a similar procedure was applied for amplitude to
remove to some degrees the spatial error component from the subject’s trajectory.
The median goal trajectory value was used to preserve the relationship between
movement reversals (i.e. extension for the first and third peak and flexion for the
second peak) since the scaled values from the median were multiplied by a positive
value if they were above the median or a negative value if they were below the
median. Using the median goal and subject amplitude trace values, the subject’s
trace was proportionately expanded or compressed until the highest correlation
between the subject’s time-scaled trajectory and the goal trajectory was reached.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
83
The difference between the goal trajectory and the subject’s scaled (in time and
amplitude) trajectory (i.e. residual RMSE) reflects th^ motor program (GMP) error,
Figure 10.
RMSE = 32.1
GMP = 5.9
TF= 1.3
AF= 1.4
a .-20
-40
0 450 900 1350
Time (ms)
Figure 10. Example of movement trajectory scaling. The 900 ms goal trajectory is
displayed as the thick black line. The subject’s actual movement is displayed as the
thin black line. RMSE is the difference (in degrees) between the goal trajectory and
subject’s actual movement. GMP is the difference between the goal trajectory and
subject’s trajectory after subject’s trajectory was scaled in time (dash line) and then
amplitude (dot line). TF (time factor) 1.3 indicates that the subject moved slower
than the goal trajectory by 30%. AF (amplitude factor) 1.4 indicates the subject
overshot the target by 40%.
Temporal and Spatial Scaling Error.
The time scaling factor that produced the highest correlation between the
subject’s scaled trajectory and goal trajectory was saved as a time factor (TF). A TF
of 1.0 indicates a movement with 100% agreement in time with the goal movement
trajectory. A TF that is greater than 1.0 indicates that the movement was slower
than the goal movement trajectory while a TF that is smaller than 1.0 indicates the
movement was faster than the goal movement trajectory. Similarly, the amplitude
scaling factor that produced the highest correlation between the subject’s scaled
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
84
trajectory and goal trajectory was saved as an amplitude factor (AF). An AF that is
greater than 1.0 indicates that the movement was hypermetric while an AF that is
smaller than 1.0 indicates the movement was hypometric. Figure 10 illustrates the
scaling process that results in TF, AF, and GMP.
Acquisition Performance
A 2 Group (control, PD) X 2 Cognitive Demand (low, high) X 2 Motor
Demand (low, high) X 10 Block Analysis of Variance (ANOVA) with repeated
measures on the last factor was conducted for each dependent measure to provide
an overall description of the acquisition performance for the two groups. Separate
student f-tests were used to determine group differences within each acquisition
block and performance differences (within each group) between acquisition block 1
and block 10.
Retention Performance
Absolute retention performance was used to index learning. Retention data
from each cognitive demand condition were analyzed separately. Therefore,
retention data for the PD-LC subgroup were compared to the Con-LC subgroup and
those for the PD-HC subgroup were compared to the Con-HC subgroup. Two
Group (control, PD) X 2 Motor Demand (low, high) X 2 Retention Test Condition
(blocked, random) Analysis of Variance (ANOVA) was conducted on GMP, AF, and
TF for the retention phase for each cognitive demand condition.
Transfer Performance
Transfer performance was used to index the ability to generalize the learned
movements. Similar to the retention phase, transfer data from each cognitive
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
85
demand condition was analyzed separately. Two Group (control, PD) X 2 Motor
Demand (1050 ms, 1350 ms) Analysis of Variance (ANOVA) was conducted on
GMP, AF, and TF for the transfer data for each cognitive demand condition.
Results
Acquisition Performance
GMP Error.
Figure 11 illustrates GMP error block means for each group by cognitive and
motor demand during acquisition Daysl (block 1-5) and 2 (block 6-10). Overall,
participants improved GMP accuracy over practice (Block Effect; p < .02). A closer
examination showed that the significant block effect is due primarily to the Con-LC
(GMP error Block 1 = 7.3, Block 10 = 6.0 deg; f-test p < .05) and PD-HC (GMP error
Block 1 = 8.1, Block 10 = 6.4 deg; f-test p < .001) subgroups. GMP error did not
change much from block 1 to block 10 for the PD-LC (GMP error Block 1 = 7.9,
Block 10 = 7.8 deg; f-test p = .80) and Con-HC (GMP error Block 1 = 8.2, Block 10 =
7.7 deg; f-test p = .75) subgroups. Both groups exhibited similar GMP error in all
conditions except for the HC practice with low motor demand condition (Figure 11,
bottom right) where the PD group showed lower GMP error than the control group.
These findings resulted in a significant Group X Cognitive Demand X Motor Demand
interaction; p < .04. When averaged across all conditions, GMP error was not
different between the two groups (Group Effect; p = .65).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
86
TF Error.
Time factor group block means for each group by cognitive and motor
demand during acquisition are displayed in Figure 12. Time factor of 1.0 (horizontal
line) indicates a movement with 100% agreement with the goal movement
trajectory. Overall, participants improved TF over practice (Block Effect; p = .05).
Across the two cognitive demand conditions, individuals with PD exhibited similar
TF as their corresponding age-matched controls for the low motor demand condition
(Figure 12, right column) but higher TF (i.e. they were slower than controls) for the
high motor demand condition (Figure 12, left column). These findings resulted in a
significant Group X Motor Demand interaction (p < .04). However, by the end of
practice (block 10), TF for the high motor demand task was similar between the PD
and control groups (mean TF block 10 for PD-LC = 1.08, Con-LC = 1.06, f-test p =
.09; PD-HC = 1.22, Con-HC = 1.16, f-test p = .14). Across the two cognitive
demand conditions, both groups were more accurate in reproducing the 1500 ms
MT (mean TF = 0.97) than the 900 ms MT goal (mean TF = 1.2). This resulted in a
significant Motor Demand Effect; p < .01.
AF Error.
Amplitude factor group block means for each group by cognitive and motor
demand conditions during acquisition phase are displayed in Figure 13. AF of 1.0
(horizontal line) indicates a movement with 100% agreement in amplitude with the
goal movement trajectory. There were no group differences in AF across cognitive
and motor demand condition during the acquisition phase (Group Effect; p = .21).
Overall, AF did not change over blocks of practice (Block Effect; p = .47). Across
the two groups and cognitive demand conditions, AF was smaller for the high motor
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
87
demand condition (mean AF = 1.0) than for the low motor demand condition (mean
AF = 1.2). These findings resulted in a significant Motor Demand Effect; p < .001.
Further, both groups were very accurate in reproducing the goal movement
amplitude for the 900 ms trajectory but overshot the 1500 ms trajectory by
approximately 20%.
High Motor Demand
(900 ms)
Low Motor Demand
(1500 ms)
* Con-LC * PD-LC
T 5
c
C O
E
0 )
Q
C D
>
'c
O )
o
o
5
O
LU
C l
2
O
10 -
. 9 -
>
- 8 *
7 -
6 -
5 -
4 -
I I
2 3 4 5 6 7 8 9
Acquisition (9 trials/block)
900 ms
1 0 -
9 -
8 -
7 -
6 -
5 -
4 '
10 3 4 5 6 7 8 9 10
Acquisition (9 trials/block)
1500 ms
■ Con-HC ° PD-HC
T 3
C
T O
E
< D
a
< D
>
C
O )
o
O
-C
o >
10
9
8
7
6
5
2 3 4 5 6 7 8 9 10
Acquisition (9 trials/biock)
900 ms
10 -
9 -
8
7
6
5
2345678 9 10
Acquisition (9 trials/block)
1500 ms
Figure 11. GMP error block means during acquisition Days 1 (block 1-5) and 2
(block 6-10) for the Control (closed symbol) and PD (open symbol) groups. Top row
illustrates GMP error from the low cognitive demand condition (blocked practice
order with 100% FB) and bottom row illustrates those from the high cognitive
demand condition (random practice order with 60% faded FB). Left column
illustrates data from the high motor demand condition (900 ms trajectory) and right
column illustrates those from the low motor demand condition (1500 ms trajectory).
GMP error is shown in degrees. Error bars are SEM.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
88
High Motor Demand Low Motor Demand
(900 ms) (1500 ms)
* Con-LC * PD-LC
1.40 1.40 -I
T 3
1.30 - 1.30 -
1.20 - 1.20 -
1.10 -
a >
E
1.00 P 1.00
0.90 ' 0.90 -
0.80 0.80
2 3 4 5 6 7 8 9 10 1 123456789 10
Acquisition (9 trials/block) Acquisition (9 trials/block)
900 ms 1500 ms
a Con-HC ° PD-HC
" O
1.40 i 1.40 -i
1.30 “ 1.30 *
o
T i
r o
U-
1.20 - 1.20 -
1.10 -
1.10 -
H 1.00 1.00
0.90 -
0.90 -
o >
0.80
0.80
2 3 4 5 7 8 9 10 1 6
Acquisition (9 trials/block) Acquisition (9 trials/block)
900 ms 1500 ms
Figure 12. TF block means during acquisition Days 1 (block 1-5) and 2 (block 6-10)
for the Control (closed symbol) and PD (open symbol) groups. Top row illustrates
TF from the low cognitive demand condition (blocked practice order with 100% FB)
and bottom row illustrates TF from the high cognitive demand condition (random
practice order with 60% faded FB). Left column illustrates data from the high motor
demand condition (900 ms trajectory) and right column illustrates those from the low
motor demand condition (1500 ms trajectory). Horizontal line at 1.0 indicates the
goal movement time. Error bars are SEM. Across the two cognitive demand
conditions, there was a significant Motor Demand Effect; p < .01, suggesting that
both groups were more accurate in reproducing the 1500 ms than the 900 ms
movement time.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
89
" O
c
0 5
E
0
Q
0
>
'.£ 3
*c
o >
o
o
5
o
High Motor Demand
(900 ms)
Low Motor Demand
(1500 ms)
^ Con-LC * PD-LC :
1.40 -i 1.40 -i
1.30 - 1.30 ■
1.20 - 1.20 -
1.10 - 1.10 -
T
f
-A-
1.00 1.00
0.90 ■
0.80 0.80
2 3 4 5 6 7 8 9 10 1
Acquisition (9 trials/block)
900 ms
Acquisition (9 trials/block)
1500 ms
a Con-HC ° PD-HC
c
( C
E
0
Q
0
>
c
O )
O
o
-C
0 3
1.40 -
1.30 -
1.20 -
1.10 -
0.90 -
1 2 3 4 5 6 7 8 9
Acquisition (9 trials/block)
900 ms
1.40 I
1.30 *
1.20 ■
1.10 •
1.00
0.90 -
0.80
7 9 10 3 4 5 6 8 1 2
Acquisition (9 trials/block)
1500 ms
Figure 13. AF block means during acquisition Days 1 (block 1-5) and 2 (block 6-10)
for the Control (closed symbol) and PD (open symbol) groups. Top row illustrates
AF from the low cognitive demand condition (blocked practice order with 100% FB)
and bottom row illustrates AF from the high cognitive demand condition (random
practice order with 60% faded FB). Left column illustrates data from the high motor
demand condition (900 ms trajectory) and right column illustrates those from the low
motor demand condition (1500 ms trajectory). Horizontal line at 1.0 indicates the
goal movement amplitude. Error bars are SEM.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
90
Retention Performance
GMP Error.
Figure 14 illustrates GMP error during the retention phase for each group by
cognitive demand, motor demand, and retention test. For the LC condition (blocked
practice order and 100% FB), top row, across the two motor demand tasks, GMP
error was significantly higher for individuals with PD (GMP = 8.8 deg) compared to
controls (GMP = 6.7 deg). This resulted in a significant Group Effect; p < .04. For
the HC condition (random practice order and 60% faded FB), bottom row, there
were no group differences in motor program accuracy across motor demand, and
retention test, Group Effect; p = .67. Statistical analyses revealed no significant
group main effect or any significant group interactions.
TF Error.
Figure 15 illustrates TF error during the retention phase for each group by
cognitive demand, motor demand, and retention test. For the LC condition (top
row), individuals with PD exhibited similar TF as their age-matched controls for the
low motor demand condition (i.e. 1500 ms trajectory) either when tested in blocked
retention (TF for Con-LC = 0.98, PD-LC = 0.97) or random retention (TF for Con-LC
= 0.95, PD-LC = 1.0), Figure 15, top right. These TFs for both groups were close to
1.0, suggesting that they were very accurate in reproducing the 1500 ms MT when
no feedback information was provided. When motor demand of the task was
increased (i.e. 900 ms trajectory), both groups were generally slower than the goal
MT (TF > 1.0). Individuals with PD showed significantly higher TF error than their
age-matched controls for both blocked retention (TF for Con-LC = 1.06, PD-LC =
1.18) and random retention (TF for Con-LC = 1.13, PD-LC = 1.26) tests, Figure 15,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
91
top left. Together, these findings for the LC practice condition resulted in a
significant Group X Motor Demand interaction; p < .05.
For the HC condition (bottom row), similar to the LC condition, there were no
group differences in TF between individuals with PD and their age-matched controls
for the low motor demand condition (1500 ms trajectory) when tested in either
blocked retention (TF for Con-HC = 0.95, PD-HC = 0.90) or random retention (TF for
Con-HC = 0.94, PD-HC = 0.95), Figure 15, bottom right. However, when motor
demand of the task is increased (i.e. 900 ms trajectory), individuals with PD
exhibited higher TF error than their age-matched controls for blocked retention (TF
for Con-HC = 1.04, PD-HC = 1.18) but not for random retention (TF for Con-HC =
1.11, PD-HC = 1.16), Figure 15, bottom left. Statistical analysis revealed a
significant Group X Motor Demand X Retention Condition interaction; p < .03.
AF Error.
Figure 16 illustrates AF error during the retention phase for each group by
cognitive demand, motor demand, and retention test. There were no group
differences in AF across cognitive, motor demand, and retention test. Statistical
analyses revealed no significant group effect or any significant group interactions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
92
T 5
C
C O
E
C D
Q
0 >
’ ■ M
c
D )
O
o
5
O
10 •
LU
Q .
2
0
High Motor Demand
(900 ms)
i
A-
T
T
-A
10
9
8
7
6
blocked-ret random-ret
Retention Test Condition
Low Motor Demand
(1500 ms)
“ A - Con-LC
-A - PD-LC
T
-A
T
blocked-ret random-ret
Retention Test Condition
Con-HC
-C3- PD-HC
blocked-ret random-ret
Retention Test Condition
blocked-ret random-ret
Retention Test Condition
Figure 14. Mean GMP error during retention tests (blocked and random) for Control
(closed symbol) and PD (open symbol) groups. Top row illustrates data from the
low cognitive demand (LC) condition (blocked practice order with 100% FB) and
bottom row illustrates those from the high cognitive demand (HC) condition (random
practice order with 60% faded FB). Left column illustrates data from the high motor
demand condition (900 ms trajectory) and right column illustrates those from the low
motor demand condition (1500 ms trajectory). GMP error is shown in degrees.
Error bars are SEM. There was a significant Group Effect (p < .04) for the LC
condition but not for the HC practice condition (p = .67).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
High Motor Demand Low Motor Demand
(900 ms) (1500 ms)
Con-LC
- A \ PD-LC
1.40 -!
1.20 -
1.00 -
0.80
blocked-ret random-ret
1.40 -i
~ o
c
0 3
E
0
Q
0
>
1.20 *
c
o >
o
O
§
o
- J
p 1.00
blocked-ret random-ret
Retention Test Condition Retention Test Condition
-B -
Con-HC
-D- PD-HC
T3
C
03
E
0
Q
0
>
C
O )
o
O
X I
a>
1.40 n
1.20 -
o
o
T O
L i.
0 )
£
p 1.00
0.80
blocked-ret random-ret
1.40 -t
1.20 -
Retention Test Condition
blocked-ret random-ret
Retention Test Condition
Figure 15. TF block means during retention tests (blocked and random) for Control
(closed symbol) and PD (open symbol) groups. Top row illustrates TF from the low
cognitive demand (LC) condition and bottom row illustrates TF from the high
cognitive demand (HC) condition. Left column illustrates data from the high motor
demand condition (900 ms trajectory) and right column illustrates those from the low
motor demand condition (1500 ms trajectory). Horizontal line at 1.0 indicates the
goal movement time. Error bars are SEM. There was a significant Group X Motor
Demand Interaction (p < .05) for the LC practice condition and a significant Group X
Motor Demand X Retention Condition interaction (p < .03) for the HC practice
condition.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
94
High Motor Demand
(900 ms)
Low Motor Demand
(1500 ms)
-rlr- Con-LC
-A - PD-LC
1.40 n 1.40 -i
* o
1.20 *
- A
1.00
- J
0.80 0.80
blocked-ret random-ret
random-ret biocked-ret
Retention Test Condition
Retention Test Condition
Con-HC
-o - PD-HC
1.40 -i
TS
7 5 1.20 ■
c o
li
e n
1.00
O )
blocked-ret random-ret
1.40 -|
1.20 •
1.00
0.80
blocked-ret random-ret
Retention Test Condition Retention Test Condition
Figure 16. AF block means during retention tests (blocked and random) for Control
(closed symbol) and PD (open symbol) groups. Top row illustrates AF from the low
cognitive demand condition (blocked practice order with 100% FB) and bottom row
illustrates AF from the high cognitive demand condition (random practice order with
60% faded FB). Left column illustrates data from the high motor demand condition
(900 ms trajectory) and right column illustrates those from the low motor demand
condition (1500 ms trajectory). Horizontal line at 1.0 indicates the goal movement
amplitude. Error bars are SEM. There were no significance main effects or
interactions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
95
Transfer Performance
GMP Error.
Figure 17 illustrates GMP error during the transfer phase for each group by
cognitive demand condition. GMP errors were similar between the 1050 ms
trajectory and the 1350 ms trajectory. Thus, averaged GMP error between the two
trajectories was displayed in Figure 17. Findings for GMP error during the transfer
phase were parallel to those during the retention phase. More specifically,
individuals with PD who practiced under the LC condition showed higher GMP error
compared to age-matched control subjects (Con-LC = 6.8, PD-LC = 8.7; p < .03). In
contrast, those who practiced under the HC condition showed similar GMP error to
that of age-matched control subjects (Con-HC = 7.4, PD-HC = 7.8; p = .78).
Control
Low High
Cognitive Demand
Figure 17. GMP error averaged across the two trajectories (1050 and 1350 ms)
during the transfer phase for the Control and PD groups. For in the low cognitive
demand practice condition, the PD group exhibited significantly higher GMP error
than that of the control group (p < .05). GMP error was not different between the
two groups in the high cognitive demand practice condition. Error bars are SEM.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
96
TF Error.
Overall, there were no group differences in TF during the transfer phase.
Both groups exhibited higher TF error for the 1050 ms MT goal than for the 1350 ms
MT goal (Motor Demand Effect; p < .001), Figure 18a. Figure 18b illustrates TF
error (average across the two trajectories) for each group by cognitive demand
condition. For each cognitive demand, individuals with PD showed higher TF error
than control subjects. Although these differences did not reach significance (Group
Effect for LC condition, p = .15; HC condition, p = .09), the effect size was large (ES
for LC condition = .64, HC condition = .82). Healthy control subjects who practiced
under the HC condition were very accurate in reproducing the two MT goals during
the transfer phase (Con-HC mean TF = 1.0).
A. B.
Control
1050 ms 1350 ms Low High
Goal Trajectory Cognitive Demand
Figure 18. TF for the Control and PD groups during the transfer phase. A) bar
graphs show TF for each goal trajectory. Both groups exhibited higher TF for the
1050 MT trajectory than for the 1350 MT trajectory (Motor Demand Effect; p < .001).
B) bar graphs show TF for each cognitive demand practice condition. The PD group
exhibited higher TF than the control group for both cognitive demand conditions.
Although, group differences did not reach significance, the effect size was large (ES
for low cognitive demand = .64, high cognitive demand = .82). Error bars are SEM.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
97
AF Error.
Overall, there were no group differences in AF during the transfer phase.
Across the two groups, AF was greater for the 1350 ms trajectory (AF = 1.24) than
for the 1050 ms trajectory (AF = 1.14), Motor Demand Effect; p < .01. This
significant motor demand effect is due primarily to the control group (AF for 1050
MT = 1.08; 1350 MT = 1.22), Figure 19a. AF for the PD group was similar between
the two trajectories (AF for 1050 MT =1.21; 1350 MT = 1.27). Figure 19b illustrates
AF error (average across the two trajectories) for each group by cognitive demand
condition. For the LC condition, AF errors were very similar between the two groups
(Con-LC = 1.15, PD-LC = 1.16). For the HC condition, the PD group exhibited
greater AF than the control group (Con-HC = 1.15, PD-HC = 1.32). Group
differences, however, did not reach significance (p = .50) due primarily to the high
variability in the PD group.
A. B.
Control
1050 ms 1350 ms Low High
Goal Trajectory Cognitive Demand
Figure 19. AF for the Control and PD groups during the transfer phase. A) bar
graphs show AF for each group by goal trajectory. B) bar graphs show AF for each
group cognitive demand practice condition. AF was similar between the two groups
for the LC practice condition. The PD group exhibited higher AF than the control
group for the HC condition. Group differences, however, did not reach significant (p
= .50) due primarily to the high variability in the PD group. Error bars are SEM.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
98
Discussion
GMP Learning
The hypothesis that GMP learning would be impaired in subjects with PD
was partially supported. Retention performance indicated that individuals with PD
showed impaired GMP learning (compared to controls) when they practiced in the
low but not high cognitive demand condition. These findings suggested that
cognitive demand during practice has critical effects on GMP learning capability of
individuals with PD. In healthy young adults, it has typically been demonstrated that
learning of an accurate invariant-structure (i.e. GMP learning) is enhanced by
practice conditions that make performance stable such as constant practice or less
frequent FB (e.g. Lai & Shea, 1998; Wulf et al., 1993; 1994). Specifically, constant
practice leads to more accurate, stable performance during acquisition and less
GMP error during retention compared to variable practice. Likewise, during variable
practice, less frequent FB leads to more stable performance during acquisition and
less GMP error during retention compared to continuous FB (FB provided after
every trial). However, when FB frequency is manipulated during constant practice,
there is no effect of FB manipulation, suggesting that constant practice leads to
sufficient movement stability independent of the effect from a FB manipulation (Lai &
Shea, 1998; Lai et al., 2000). Consistent with Lai and colleagues, in the present
study, it appears that blocked practice order led to sufficient movement stability
even though it was manipulated together with 100% FB. Thus, for the healthy
control subjects, practice under the low cognitive demand condition benefited GMP
learning over the high cognitive demand condition. However, this was not the case
for individuals with moderate PD in the present study. Practice under the high
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
99
cognitive demand but not low cognitive demand condition benefited GMP learning
for individuals with PD, particularly when the motor demand of the task was low.
Why did PD benefit from the high cognitive demand (random practice - 60%
faded FB) but not from the low cognitive demand (blocked practice - 100% FB) for
GMP formation?
For the PD group, the high cognitive demand condition could have
promoted the explicit (aware) memory system to compensate for an impaired
implicit (unaware) memory system for GMP learning. Previous work has
demonstrated that the damaged neostriatum (basal ganglia) in individuals with PD
resulted in an impaired implicit learning and memory system while the explicit
learning and memory system remained intact (Knowlton, Mangels, & Squire, 1996;
Saint-Cyr, Taylor, & Lang, 1988). For example, Knowlton and colleagues (1996)
showed that individuals with PD failed to learn the probabilistic classification task
(an implicit task) despite having intact declarative memory for the training episode.
In the low cognitive demand condition due to the blocked practice order,
subjects knew in advance which trajectory would be presented throughout a practice
set. In addition, augmented FB was provided after every trial. In this condition,
movements are more likely to be recalled through the more automatic, implicit
memory system. In contrast, in the high cognitive demand condition, tasks were
presented in random order and post-response FB was less frequent. Subjects had
to both attend to which trajectory was presented and evaluate their own
performance using internal feedback when external FB was absent. These
experimental manipulations required relatively more cognitive processing (than for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
100
the blocked practice order and 100% FB) and were likely to evoke the declarative,
memory system through awareness. Several studies have shown that the capability
to generate movements from automatic, internal control mechanisms is impaired in
individuals with PD (Brown & Marsden, 1988; Cools, van den Bercken, Horstink, van
Spaendonck, & Berger, 1990; van Spaendonck, Berger, Horstink, Borm, & Cools,
1996). However, motor performance can be improved by directing subjects with PD
to use an attentional strategy (Cunnington, lansek, & Bradshaw, 1999; Morris &
lansek, 1996; Morris et al., 1996; Oliveira, Gurd, Nixon, Marshall, & Passingham,
1997). For example, Morris and colleagues demonstrated that stride length and
velocity of individuals with PD improved significantly simply by instructing the
subjects to concentrate on walking fast and with large steps (Morris et al., 1996).
The attentional strategy is an instruction to consciously attend to the to-be-
performed task or to particular aspects of the movement to be performed such as
movement amplitude and movement time. By using this attentional strategy, it is
presumed that movements are being planned and monitored by the explicit memory
system that circumvents the more automatic, implicit system known to be impaired
in individuals with PD. This idea is consistent with the dual mode principle
described by Willingham (Willingham, 1998). The dual mode principle states that
motor acts can be executed in either a conscious or an unconscious mode and
these two modes are available throughout skill acquisition. The performer may
switch between the two modes, weighting the possible trade-off of accuracy and
attention cost. Results in the present study suggest that the high cognitive demand
condition prompted individuals with PD to use the more aware, explicit memory
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
101
system to compensate for the impaired implicit memory system-trading the greater
attentional cost for higher motor program accuracy.
Parameter Learning
Time Factor.
During acquisition, timing error was significantly greater for the PD group
than for the control group at the fast speed (900 ms trajectory) but it was similar
between the two groups at the slow speed (1500 ms trajectory). This suggests that
motor control deficits in individuals with PD were evident when motor demand was
high but not when it was low. However, by the end of practice (block 10), there
were no differences in TF between the two groups for the high motor demand task,
suggesting that with practice subjects with PD could eventually overcome their
control deficits and show comparable performance to control subjects. Together,
results from the low and high cognitive demand conditions suggest that the
characteristic bradykinesia in individuals with PD is not evident when the task motor
demand is relatively low and appears to be minimized by the end of practice when
task motor demand is high.
The hypothesis that the ability to learn temporal parameters of the
movements would be preserved in subjects with PD was partially supported.
Results from the retention phase indicated that the ability to learn temporal
parameters of the movements for subjects with PD was primarily affected by the
motor demand of the task. When motor demand was low, subjects with PD
exhibited similar TF error to control subjects independent of cognitive demand and
retention condition. Conversely, when motor demand was high, cognitive demand
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
102
and retention condition played a role in the learning capability of subjects with PD.
For the high motor demand condition, subjects with PD showed impaired TF
learning compared to controls on both retention test conditions if they had practiced
the task under the low cognitive demand condition (blocked order with 100% FB).
However, if they had practiced the task under the high cognitive demand condition
(random order and 60% faded FB), they showed impaired TF learning only when
tested in the blocked retention. TF errors were similar between the two groups for
the random retention. A closer examination showed that the PD group exhibited
similar TF for blocked and random retention tests while the control group exhibited
greater TF error for random than for blocked retention tests. These findings suggest
that under the high motor and cognitive demand conditions, retention test condition
has a large effect on performance of the control but not the PD group.
Amplitude Factor.
The hypothesis that the ability to learn spatial parameters of a discrete
movement would be preserved in subjects with PD was supported. During
retention, there were no group differences in AF between the PD and control groups
across cognitive demand and retention condition. This lack of group differences is
in contrast to previous results for individuals with PD. Like bradykinesia, hypometria
(reduction of movements) has been repeatedly demonstrated in individuals with PD
(e.g. Castiello et al., 1993; Contreras-Vidal, Teulings, & Stelmach, 1995; Flowers,
1976). In the present analysis, there were no group differences in movement
amplitude during either the acquisition phase (execution ability) or retention phase
(learning ability). It is postulated that the nature of the task in the present study may
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
103
not evoke hypometria. Typically, hypometria in individuals with PD has been
reported for continuous or sequential movements such as reaching and grasping,
hand writing, reciprocal tapping, and walking (e.g. Castiello et al., 1993; McLennan
et al., 1972; Morris et al., 1996; Onla-or & Winstein, 2001) but not for discrete
movement tasks (Hoshiyama, Kaneoke, Koike, Takahashi, & Watanabe, 1994) such
as that used in the present study.
Rather than evidence of hypometria, both control and PD groups showed
hypermetric movement for the slowest trajectory. Both groups were accurate in
scaling movement amplitude for the fast trajectory (i.e. 900 ms) but overshot the
slower trajectory (i.e. 1500 ms). This suggests that both groups chose the 900 ms
trajectory as a default. It is hypothesized that when performing the 1500 ms
trajectory, rather than simply slowing down, both groups maintained a similar
average velocity to that used for the default trajectory and achieved a longer
movement time criteria by moving further (i.e. overshooting the target). Had they
not increased the amplitude of movement, they would have completed the 1500 ms
task too early. This hypothesis suggests that subjects used a pulse-width strategy
of motor control. The characteristic feature of a pulse-width strategy is that the rate
of increase in velocity and acceleration is constant (Gottlieb, Corcos, & Agarwal,
1989). Using the pulse-width strategy, time to peak velocity is greater for the longer
movement time.
The task goal in the present study was to move to the same amplitude with
different explicit speeds. To do so, subjects have to inversely co-vary between the
rate of rise and time to peak velocity. Specifically, the rate of rise is higher but the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
104
time to peak velocity is smaller for the 900 ms trajectory than that for the 1500 ms
trajectory. As such, the task should evoke a combination of both pulse-height and
pulse-width strategy. The characteristic feature of a pulse-height strategy is that the
rate of rise of velocity and acceleration is variable (Corcos, Gottlieb, & Agarwal,
1989). While the pulse-width strategy results in differences in time to peak velocity,
the pulse-height strategy results in differences in the rate of rise of movement
velocity. It is not clear why both PD and control groups chose to use the pulse-width
strategy. One possibility is that an ability to inversely co-vary the rate of rise and the
time to peak velocity decreases in older adults. Nevertheless, the similarity between
the PD and control groups in hypermetric responses for the 1500 ms trajectory
suggests that this behavior was not due to dysfunction of the basal ganglia.
Generalizability of Motor Learning
GMP.
The two new trajectories used during the transfer test had the same relative
timing and amplitude as the practiced trajectories. Thus, the GMP for the two new
trajectories was the same as the practiced trajectories. Not surprisingly, findings for
GMP error during the transfer test were similar to those during the retention test as it
was the same GMP. More specifically, there were group differences in GMP error
(i.e. greater GMP error for the PD group than for the control group) for those who
practiced in the low cognitive demand but not for those who practiced in the high
cognitive demand condition.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
105
Parameters.
During the transfer test, subjects performed two new trajectories with the
same amplitude but different temporal requirements. Thus, only the ability to
generalize to new movement time goals was addressed in the present study. The
MT goals for the two new trajectories were within the range of those for the
practiced trajectories (i.e. between 900 and 1500 ms). Findings during the transfer
phase tended to support the hypothesis that the generalizability of learning would be
impaired in individuals with PD. For each cognitive demand condition, the PD group
showed greater TF error than the control group. While these group differences did
not reach significant, the effect size was large. Thus, these findings suggest that
the generalizability of movement time parameterization appears to be impaired in
individuals with PD.
General Conclusion
Overall, findings from the present study indicated condition-specific motor
learning deficits in individuals with PD. These findings suggest that it is important to
consider the cognitive demand, motor demand, and context of retention tests when
evaluating the learning capability of individuals with moderately severe PD.
Individuals with moderately severe PD learned a motor task as well as
neurologically healthy age-matched control subjects when the level of motor and
cognitive demand of the task was relatively low (i.e. within their capabilities). When
the level of the motor demand was high, impaired motor learning in individuals with
PD was revealed. Impaired motor learning under the high motor demand condition,
however, could be overcome if individuals with PD practiced the task under a high
cognitive demand condition. This suggests that high cognitive demand practice
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
106
enhances motor learning for individuals with PD. However, this learning was less
flexible and showed contextual-dependence. It is proposed that the context-
dependent learning seen in individuals with PD could be due to a set-shifting deficit
characteristic of PD (Sandson & Albert, 1987). Nigrostriatal degeneration from
dopamine depletion in PD may disrupt the normal function of the dorsolateral
prefrontal-striatal circuit, previously thought to be responsible for this set-shifting
function.
Impaired motor program learning in individuals with moderately severe PD
was evident under the low but not high cognitive demand practice condition. This
suggests that the high cognitive demand practice condition enhances motor
program learning in individuals with PD. In the high cognitive demand practice
condition, tasks were presented in random order and FB was less frequent. Thus,
subjects had to both attend to which trajectory was presented and to evaluate their
own performance using internal FB when external FB was absent. These
experimental manipulations required relatively more cognitive processing (than the
blocked practice and 100% FB condition) and could likely invoke the explicit
memory system through awareness. As such, the high cognitive demand practice
condition may have allowed individuals with PD to compensate for an impaired
implicit memory system by invoking the intact explicit memory system for motor
program learning. Movement time parameter learning in individuals with PD was
affected primarily by the motor demand of the task. Impaired movement time
parameter learning was evident under the high but not low motor demand condition.
The ability to learn spatial parameters of a discrete movement was preserved in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
individuals with moderately severe PD. Finally, the ability to generalize motor
learning appeared to be impaired in individuals with moderately severe PD.
These results have important implications for rehabilitation. Consistent with
a recent clinical review and case study (Morris, 2000), task-specific training appears
to be an effective approach for rehabilitation of individuals with PD, especially when
the task involves relatively high cognitive and motor demands.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
108
CHAPTER 3:
CONCLUSIONS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
109
The purpose of this study was to examine the role of the basal ganglia in
motor learning by studying individuals with known basal ganglia dysfunction. Thus,
the learning capability of individuals with Parkinson’s Disease (PD) was compared
with healthy age-matched controls. While neurophysiologic studies using animal
models have suggested that the basal ganglia are important for motor learning (e.g.
Aosaki et al., 1994; Aosaki, Kimura, & Graybiel, 1995; Miyachi, Hikosaka, Miyashita,
Karadi, & Rand, 1997), behavioral studies in humans have not consistently
demonstrated motor learning deficits in individuals with PD (e.g. Bondi & Kaszniak,
1991; Doyon et al., 1997; Harrington, Haaland, Yeo, & Marder, 1990; Jackson,
Jackson, Harrison, Henderson, & Kennard, 1995; Pascual-Leone et al., 1993).
Unique to the present investigation was the systematic control over three central
issues that were identified as possible factors contributing to the discrepancies in
previous behavioral work. Specifically, first, heterogeneity in disease severity was
controlled by including only participants with moderately severe PD (Hoehn & Yahr
stage II and III). Second, performance effects were delineated from learning effects
using a retention and transfer design. Third, the level of cognitive and motor
demand of the to-be-learned task was systematically manipulated. The first
question was: what is the role of the basal ganglia in motor learning in a sample of
subjects with moderately severe PD? The first hypothesis was that when both
cognitive and motor demand are low, there will be no difference in motor learning
capability between individuals with moderately severe PD and controls. The second
hypothesis was that impaired motor learning in individuals with PD will be evident for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
110
a task condition when either one or both motor and cognitive demands are high. To
address this first question, motor learning was indexed using a measure of overall
movement error, root mean square error (RMSE).
The first hypothesis was supported. When both cognitive and motor demand
were low, individuals with PD showed no performance differences compared to that
of healthy age-matched controls on a retention test. The second hypothesis was
only partially supported. Individuals with PD showed learning deficits when the task
required a low cognitive but high motor demand. However, when the cognitive
demand was high, individuals with PD showed context-dependent learning
independent of the level of motor demand. More specifically, individuals with PD
showed comparable learning with healthy age-matched controls when the retention
test context was the same as that used in the practice condition (i.e. random order).
Recall that when individuals with PD practiced under low cognitive - high motor
demand conditions they made more errors (greater RMSE) than controls regardless
of retention test condition. Practice under a high cognitive demand condition
appeared to facilitate motor learning for individuals with PD in that they showed
comparable performance to that of control subjects even when task motor demand
was high. However, this motor learning capability was evident only when the
retention test condition was the same as that experienced during the practice phase.
So while a high cognitive demand practice appears to facilitate problem solving
processes important for motor learning, the retention test condition must be similar
to that of the practice condition to demonstrate this effect. Practice in a high
cognitive demand condition appears to potentiate learning for individuals with
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
111
moderately severe PD. However, this effect is not generalizable. Rather, it
depends on a specific and inflexible association between the practice and retention
context. This context-dependent learning effect is reminiscent of the set-shifting-
deficit characteristic of PD (Sandson & Albert, 1987). Therefore, these results are
consistent with the view that the basal ganglia have a primary role in context-
independent learning. Nigrostriatal degeneration due to dopamine depletion in PD
may disrupt the normal function of the dorsolateral prefrontal-striatal circuit,
previously thought to be responsible for the set-shifting function (Cronin-Golomb et
al., 1994; Wise et al., 1996).
The second question addressed in this study also provided a unique
contribution to the literature on the role of the basal ganglia in motor learning. This
question focused on the processes by which an individual with moderately severe
PD acquires motor skills. Schmidt’s schema theory suggests that the learning of a
skilled action involves the acquisition of the motor program and the capability to
select effective movement parameters (Schmidt, 1975). To my knowledge, this
study was the first to examine the role of the basal ganglia in these two processes.
Thus, the second question was: do the basal ganglia have a preferential role in the
acquisition of the motor program, the selection of task parameters or a combination
of both? It was hypothesized that individuals with PD would demonstrate deficits in
motor program learning but not movement parameter learning. Movements were
expected to be slower for individuals with PD compared to controls due to a known
motor control deficit of PD (i.e. bradykinesia). However, it was hypothesized that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
112
subjects with PD would demonstrate parameter learning (through increased speed
with practice) and this improvement would be sustained in a delayed retention test.
Hypometric responses in subjects with PD were not expected given that the to-be-
learned-task was a discrete movement. Thus, it was hypothesized that individuals
with PD would also demonstrate parameter learning of movement amplitude. To
test these hypotheses, the motor program was indexed using a general motor
program (GMP) error and parameterization was indexed using a time factor (TF)
and an amplitude factor (AF) from the decomposed motor performance.
The hypothesis with respect to GMP learning was partially supported.
Individuals with PD showed impaired GMP learning only when they practiced the
task under the low cognitive demand condition. In the high cognitive demand
practice condition, the PD group showed similar GMP error to that of the control
group. The hypothesis regarding parameter learning was also partially supported.
With only two exceptions, the ability to learn temporal parameterization (as indicated
by TF) was not different between the PD and control groups. The two exceptions
were 1) low cognitive demand practice - high motor demand - both blocked and
random retention test conditions and 2) high cognitive demand practice - high motor
demand - blocked retention test condition. Therefore, motor demand critically
influenced the ability to learn temporal parameterization. However, when the high
cognitive demand practice condition (random practice order) was paired with the
same retention test condition (i.e. random retention) impaired temporal parameter
learning for subjects with PD was not apparent. The ability to learn spatial
parameterization (as indicated by AF) was the same for the two groups across all
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
113
conditions. From these results, it appears that the basal ganglia do not have a
preferential role for GMP or parameter learning. Instead, the basal ganglia appear
to be important for both GMP and parameter learning. Further and more
importantly, cognitive and motor demand can critically influence the motor program
and parameter learning capability of individuals with moderately severe PD.
Overall, these findings indicated condition-specific motor learning deficits in
individuals with PD. These findings suggest that it is important to consider the
cognitive demand, motor demand, and context of retention tests when evaluating
the learning capability of individuals with PD. It has been suggested that motor
program and movement parameter errors are two primary components that
constitute performance error (Wulf, Schmidt, & Deubel, 1993). Further insight with
respect to the role of the basal ganglia in motor learning can be gained by
determining the relationship between GMP error, parameter error, and RMSE.
Table 1 summarizes the experimental design and findings. The independent
variables were group (control, PD), cognitive demand (low, high), motor demand
(low, medium, high), and retention test condition (blocked, random). The cognitive
demand was a between-subject manipulation while the motor demand and retention
test were within-subject manipulations. Blocked practice order and 100% feedback
(FB) characterized the low cognitive demand condition. Random practice order and
60% faded FB characterized the high cognitive demand condition. The 900 ms
trajectory represented the high motor demand task while the slower movement time,
1200 and 1500 ms trajectories represented the medium and low motor demand
task, respectively. The dependent measures were overall error (RMSE), motor
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
114
program error (GMP error), and parameter error, specifically temporal error (time
factor, TF) and spatial error (amplitude factor, AF). For the purpose of this study,
only data from the high and low motor demand (900 and 1500 ms trajectories) were
analyzed.
The effects of the independent variables on each dependent measure are
summarized in Table 1. First, the results revealed the effects of practice condition
and retention test condition on RMSE. Specifically, the pattern of RMSE results for
the two retention test conditions was similar for those who practiced under the low
cognitive demand condition (see RMSE for the two retention tests in Box A, B). For
example, RMSE for the 1500 ms trajectory was not different between the PD and
control groups and this pattern was true for both blocked and random retention
(RMSE, Box B). Conversely, the pattern of RMSE was not the same for the two
retention test conditions for those who practiced under the high cognitive demand
condition (see RMSE for the two retention tests in Box C, D). For example, RMSE
for the 1500 ms trajectory was not different between the two groups when tested
under the random retention test but was greater for the PD group than that of the
control group when tested under the blocked retention test (RMSE, Box D).
Second, the manipulation of cognitive demand resulted in two distinct
patterns of findings for GMP learning that were independent of motor demand and
retention condition. High cognitive demand practice (random order, 60% faded FB)
enabled individuals with PD to show comparable GMP error to controls across all
motor demands and retention test conditions (see GMP, Box C, D). In contrast, low
cognitive demand practice (blocked order, 100% FB) led to greater GMP error for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 1. Summary of the experimental design and findings.
Experimental Design:
Group
Subgroup
Practice
Condition
Retention
Condition
Analysis:
PD (N = 20), Control (N = 20)
PD (n = 10)
Control (n = 10)
Low Cognitive Demand
(blocked practice order, 100 % FB)
Motor Demand
(High = 900 ms, Medium = 1200 ms, Low = 1500 ms)
Blocked and Random Retention
2 Group X 2 Motor Demand X 2 Retention Condition ANOVA
Retention
Condition
High Motor Demand
^ (9 0 2 _ m s )^
Random Blocked
RMSE PD > Con PD > Con
GMP PD > Con PD > Con
TF PD > Con PD > Con
AF PD = Con PD = Con
Low Motor Demand
(150Q ms)
Random Blocked
PD = Con PD = Con
PD > Con PD > Con
PD = Con PD = Con
PD = Con PD = Con
PD (n = 10)
Control (n = 10)
High Cognitive Demand
(random practice order, 60% faded FB)
Motor Demand ...
(High = 900 ms, Medium = 1200 ms, Low = 1500 ms)
Blocked and Random Retention
2 Group X 2 Motor Demand X 2 Retention Condition ANOVA
High Motor Demand
(9Q0ms)
Random Blocked
PD = Con PD > Con
PD = Con PD = Con
PD = Con PD > Con
PD = Con PD = Con
Low Motor Demand
(ISOQjtts)
Random Blocked
PD = Con PD > Con
PD = Con PD = Con
PD = Con PD = Con
PD = Con PD = Con
PD = Parkinson's disease group, Con = Control group, ‘PD > Con’ indicates that means of the dependent measure were greater for the PD group than for the control
group, 'PD = Con’ indicates that means of the dependent measure did not differ between the two groups at a significance level p < .05.
116
the PD group than that of the control group in all conditions (see GMP, Box A, B).
Third, TF learning was influenced mainly by the motor demand of the task. When
motor demand was low (1500 ms trajectory), the PD group exhibited similar TF error
to that of the control group regardless of the practice and retention test condition
(see TF, Box B, D). In contrast, when motor demand was high (900 ms trajectory),
the PD group exhibited greater TF error than that of the control group (see TF, Box
A, C). The one exception to this pattern was the context-dependent situation in
which high cognitive demand practice (random practice order) was paired with
random retention test (see TF, Box C, random retention). Finally, AF error was
similar between the PD and control groups across all conditions (see AF, Box A-D).
Findings from RMSE, GMP, and TF error indicate condition-specific motor learning
deficits in individuals with PD. It appears from the results in Box A that both greater
GMP and TF error are responsible for greater RMSE error for the PD group
compared to the control group. When GMP error but not TF error was greater for
the PD group than that of the control group, RMSE was not different between the
PD and control groups (Box B). However, when TF error but not GMP error was
greater for the PD group (compared to controls), RMSE was greater for the PD
group than that of the control group (Box C, blocked retention). Together, the above
findings suggest that the ability to learn temporal parameterization can compensate
for impaired motor program learning in individuals with moderate PD such that
overall performance error is no greater than that of healthy age-matched control
subjects. By contrast, the capability to learn a motor program cannot make up for
deficits in temporal parameter learning in individuals with moderate PD. Thus,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
117
deficits in parameter learning appear to contribute more to the motor learning deficit
for individuals with moderate PD than deficits in motor program learning.
Why did the high cognitive demand practice condition enhance motor learning
capability of individuals with PD?
In chapter two, it was proposed that the high cognitive demand practice
condition facilitates the use of the intact explicit system in individuals with PD,
allowing them to compensate for their impaired implicit system. Therefore, the
motor learning capability of individuals with PD is enhanced under this condition.
Currently, the interaction between the implicit and explicit system is not well
understood. In this chapter, putative mechanisms for the interaction between the
two systems are proposed.
Figure 1, box A, illustrates a simplified neuroanatomical model for motor skill
learning. Several studies have suggested that a distributed cortical and subcortical
network including sensorimotor cortical areas (primary motor area, supplementary
motor area (SMA), dorsolateral-prefrontal area (DLPF), premotor area), basal
ganglia, and cerebellum are important for motor skill learning (Doyon et al., 1996;
Grafton, Hazeltine, & Ivry, 1995; Jenkins et al., 1994). Each cortical area projects to
a distinct region of the caudate and putamen. Functions of these cortical areas are
influenced by projections from a distinct region of the internal segment of the globus
pallidus via thalamus. Within the basal ganglia, the input channel related to a
particular cortical area connects predominantly to the output channel that innervates
the same cortical area. This led Strick and colleagues to propose that these
multiple closed-loops (solid arrows) underlie the structural framework for basal
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
118
ganglia interaction with cortical areas (Strick, Dum, & Picard, 1995). Thus, in PD
where the basal ganglia neurons die off due to dopamine depletion, functions of
these multiple closed loops are disrupted and in turn impaired motor learning is
evident. During motor skill learning, the premotor area, SMA, and DLPF appear to
have a role in response selection, strategic selection, and working memory
(Mushiake, Inase, & Tanji, 1991; Willingham, 1999; Wise, Murray, & Gerfen, 1996).
While these areas communicate with the striatum, they also send projections
directly to other brain structures such as the primary motor area (dash arrows),
posterior parietal area, and the spinal cord, (Dum & Strick, 1991; He, Dum, & Strick,
1993). Thus, it may be possible to reduce reliance on the basal ganglia during
motor skill learning by circumventing its contribution via other healthy regions of the
motor learning distributed-network such as via direct pathways from SMA, premotor
area, and DLPF area to the motor cortex (dash arrows) then to the spinal cord.
The advantage shown by individuals with PD who practiced the task in the
high cognitive demand practice condition may be directly related to their ability to
augment performance through the use of an intact explicit memory system. The
nature of the high cognitive demand practice condition (random order, less frequent
FB) likely facilitated subjects’ awareness and thus promoted the use of the explicit
system. Specifically, under this condition, subjects had to attend to which trajectory
would be presented and evaluate their own performance when external FB was not
provided. Further, they were more likely to make comparisons from trial to trial both
for the same and between different trajectories. These cognitive processes likely
promoted subjects’ explicit knowledge regarding the task. In contrast, under the low
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
119
cognitive demand condition where subjects practiced the same trajectory and
external FB was provided throughout each set, movements were likely recalled
using the more automatic, implicit system. It is known from studies of patients with
amnesia and animal models that the medial temporal lobe (Figure 1, box B) mediate
the explicit learning and memory system (Milner, Corkin, Teuber, 1968; Squire,
1987; Squire & Zola-Morgan, 1991). Findings that subjects with PD learn a variety
of explicit tasks as well as healthy controls suggest that these areas are intact in PD
(Knowlton, Mangels, & Squire, 1996; Vakil & Herishanu-Naaman, 1998). The
hippocampal formation in the medial temporal lobe including the hippocampus,
entorhinal cortex and subicular cortex are in close reciprocal relationship with the
surrounding cortical areas including frontal (Figure 1, dot arrows) temporal, and
parietal areas (Parent, 1996). Through these connections, the hippocampal
formation is able to influence widespread regions in the frontal lobe. It may be the
richness of the information gained through explicit knowledge under the high
cognitive demand condition that informs the frontal areas particularly the DLPF and
premotor areas. Neuroimaging studies have demonstrated that the DLPF and
premotor areas are active when explicit knowledge about the motor task is available
(Grafton, Hazeltine, & Ivry, 1995; Honda et al., 1998; Sakai et al., 1998). This
information may be sent directly to other healthy regions of the motor learning
network such as the primary motor area. The proposed alternative pathway for
motor skill learning is from the medial temporal lobe to the frontal lobe particularly
the DLPF and premotor areas (dot arrows) to the motor cortex (dash arrows) and
finally from motor cortex to the spinal cord (solid arrow). The enhanced explicit
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
120
system, therefore, allows the basal ganglia to be circumvented— it lessens their
importance in the motor learning process.
N e u ro a n a to m ic a l N e tw o rk fo r M o to r S k ill L e a rn in g
- r Motor Cortex
SMA ^
Basal Ganglia
Premotor area
Thalamus
DLPF
Cerebellum
E x p lic it K n o w le d g e
Medial-temporal lobe
Figure 1. A simplified model of the neuro-substrates mediating motor skill learning
(box A) and the explicit learning and memory system (box B). The neuroanatomical
network mediating motor skill learning includes the motor cortex, premotor area,
supplementary motor area (SMA), dorsolateral-prefrontal area (DLPF), basal
ganglia, thalamus, and cerebellum. Solid arrows illustrate multiple closed-loops for
the cortico-striatal connections that underlie motor skill learning. While most areas
in the frontal lobe project to and receive inputs from the basal ganglia, they also
send inputs directly to the motor cortex (dash arrow). The explicit learning and
memory system is mediated by the medial-temporal lobe. Several areas in the
medial-temporal lobe (e.g. hippocampus, entorhinal, and subicular cortex) have
reciprocal connections with the frontal lobe particularly the DLPF and premotor
areas (dot arrows). It is proposed that explicit knowledge gained during the high
cognitive demand practice condition allows the basal ganglia to be circumvented
through alternative pathways. This alternative pathway for motor skill learning is
from the medial temporal lobe to frontal areas particularly tf Spinal cord motor
areas (dot arrows) to the motor cortex (dash arrows) and finally from motor cortex to
the spinal cord (solid arrow).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
121
Implications for Rehabilitation
Given that the ability of individuals with PD to learn movement
parameterization appeared to enhance motor learning to a greater extent than the
ability to learn the motor program, one implication for rehabilitation would be to
focus practice on scaling parameters of a motor task. Intervention programs would
include variability of task parameters such as movement through different ranges
and at different speeds. Although, our study did not address the question of
whether or not extended practice could improve the movement parameter deficits,
there is evidence that subjects with PD can improve movement speed with practice
(Behrman, Cauraugh, & Light, 2000). Future studies need to address the question
of potential remediation of motor learning deficits with extended practice.
Two potential training strategies may be useful for rehabilitation of
individuals with PD. The first strategy involves practice conditions in which
sufficiently high cognitive demands may invoke the declarative, explicit memory
system through awareness. Random practice of tasks as well as infrequent
feedback were two such practice manipulations in this study. The second strategy
relates to context-specific practice. When the task is complex, individuals with PD
may overcome their motor learning deficits if the practice and retention test context
are similar. For example, the clinical environment in which the patient practices the
task should replicate the home environment as much as possible (Morris, 2000).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
122
Bibliography
Agostino,R., Sanes.J.N., & Hallett.M. (1996). Motor skill learning in Parkinson's
disease. J.Neurol.Sci.. 139. 218-226.
Alexander,G.E., & Crutcher,M.D. (1990). Functional architecture of basal ganglia
circuits: neural substrates of parallel processing [see comments],
Trends.Neurosci., 13, 266-271.
Alexander,G.E., Crutcher,M.D., & Delong,M.R. (1990). Basal ganglia-
thalamocortical circuits: Parallel substrates for motor, oculomotor, prefrontal,
and limbic functions. In H.B.M.Uylings, C.G. Van Eden, J.P.C. De Bruin,
M.A. corner, & M.G.P. Feenstra (Eds.), Progress in Brain Research, (pp.
119-146). Elsevier. Amsterdam.
Allard,F. (1993). Cognition, expertise, and motor performance. In J. L. Starkes &
F. Allard (Eds.), Cognitive issues in motor expertise, (pp. 17-34).
Amsterdam: Elsevier.
Andresen,E.M., Malmgren,J.A., Carter,W.B., & Patrick,D.L. (1994). Screening for
depression in well older adults: evaluation of a short form of the CES-D
(Center for Epidemiologic Studies Depression Scale). Am. J .Prey. Med., 10.
77-84.
Aosaki,T., Graybiel,A.M., & Kimura,M. (1994). Effect of the nigrostriatal dopamine
system on acquired neural responses in the striatum of behaving monkeys.
Science. 265, 412-415.
Aosaki.T., Kimura,M., & Graybiel,A.M. (1995). Temporal and spatial
characteristics of tonically active neurons of the primate's striatum.
J.Neurophvsiol.. 73. 1234-1252.
Aosaki.T., Tsubokawa.H., Ishida.A., Watanabe.K., Graybiel.A.M., & Kimura.M.
(1994). Responses of tonically active neurons in the primate's striatum
undergo systematic changes during behavioral sensorimotor conditioning.
J.Neurosci.. 14. 3969-3984.
Barker,R.A., & Dunnett.S.B. (1999). Functional integration of neural grafts in
Parkinson's disease. Nat.Neurosci., 2. 1047-1048.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
123
Battig.W.F. (1979). The flexibility of human memory. In L. S. Cermak & F. I. M.
Craik (Eds.), Levels of processing in human memory, (pp. 23-44). Hillsdale,
NJ: Erlbaum.
Behrman,A.L., Cauraugh.J.H., & Light,K.E. (2000). Practice as an intervention to
improve speeded motor performance and motor learning in Parkinson's
disease. J.Neurol.Sci.. 174.127-136.
Bennett,K.M., Marchetti.M., lovine.R., & Castiello.U. (1995). The drinking action
of Parkinson's disease subjects. Brain. 118, 959-970.
Bondi,M.W., & Kaszniak.A.W. (1991). Implicit and explicit memory in Alzheimer's
disease and Parkinson's disease. J.CIin.Exp.Neuropsvchol.. 13, 339-358.
Brown,L.L., Schneider,J.S., & Lidsky.T.I. (1997). Sensory and cognitive functions
of the basal ganglia. Curr.Opin.Neurobiol.. 7 . 157-163.
Brown,R.G., & Marsden,C.D. (1988). Internal versus external cues and the
control of attention in Parkinson's disease. Brain. 111, 323-345.
Brown,R.G., & Marsden,C.D. (1990). Cognitive function in Parkinson's disease:
from description to theory. Trends.Neurosci.. 13, 21-29.
Calabresi,P., Pisani,A., Mercuri,N.B., & Bernardi.G. (1996). The corticostriatal
projection: from synaptic plasticity to dysfunctions of the basal ganglia [see
comments]. Trends.Neurosci.. 19. 19-24.
Castiello,U., Bennett,K.M., Adler,C.H., & Stelmach.G.E. (1993). Perturbation of
the grasp component of a prehension movement in a subject with
hemiParkinson's disease. Neuropsvcholoqia. 31. 717-723.
Castiello,U., Stelmach.G.E., & Lieberman.A.N. (1993). Temporal dissociation of
the prehension pattern in Parkinson's disease. Neuropsvcholoqia, 31, 395-
402.
Cohen,N.J., & Squire,L.R. (1980). Preserved learning and retention of pattern-
analyzing skill in amnesia: dissociation of knowing how and knowing that.
Science. 210. 207-210.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
124
Contreras-Vidal.J.L., Teulings,H.L., & Stelmach.G.E. (1995). Micrographia in
Parkinson's disease. Neuro Report, 6. 2089-2092.
Cools,A.R., van den Bercken.J.H., Horstink.M.W., van Spaendonck.K.P., &
Berger,H.J. (1990). Parkinson's disease: a reduced ability to shift to a new
grouping if not prompted or guided. Mov Disord., 5, 178-179.
Corcos.D.M., Gottlieb,G.L., & Agarwal.G.C. (1989). Organizing principles for
single-joint movements. II. A speed-sensitive strategy. J.Neurophvsiol.. 62,
358-368.
Contreras-Vidal.J.L., Teulings.H.L., & Stelmach.G.E. (1995). Micrographia in
Parkinson's disease. Neuro Report. 6, 2089-2092.
Cronin-Golomb.A., Corkin.S., & Growdon.J.H. (1994). Impaired problem solving
in Parkinson's disease: impact of a set-shifting deficit. Neuropsvcholoqia,
32, 579-593.
Cunnington.R., lansek.R., & Bradshaw,J.L. (1999). Movement-related potentials
in Parkinson's disease: external cues and attentional strategies. Mov
Disord., 14, 63-68.
Daum.l., Schugens.M.M., Spieker.S., Poser,U., Schonle.P.W., & Birbaumer.N.
(1995). Memory and skill acquisition in Parkinson's disease and frontal lobe
dysfunction. Cortex, 31, 413-432.
Doyon.J., Gaudreau.D., Laforce.R.J., Castonguay.M., Bedard,P.J., Bedard,F., &
Bouchard, J.P. (1997). Role of the striatum, cerebellum, and frontal lobes in
the learning of a visuomotor sequence. Brain Coqn., 34, 218-245.
Doyon.J., Owen,A.M., Petrides.M., Sziklas.V., & Evans,A.C. (1996). Functional
anatomy of visuomotor skill learning in human subjects examined with
positron emission tomography. Eur.J.Neurosci., 8, 637-648.
Dum.R.P. & Strick, P.L. (1991). Premotor areas: Nodal points for parallel efferent
systems involved in the central control of movement. In D.R. Humphrey and
H.J. Freund (eds.), Motor Control: Concepts and Issues, (pp. 383-397).
John Wiley & Sons, London.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
125
Ferraro,F.R., Balota.D.A., & Connor,L.T. (1993). Implicit memory and the
formation of new associations in nondemented Parkinson's disease
individuals and individuals with senile dementia of the Alzheimer type: a
serial reaction time (SRT) investigation. Brain Coon.. 21. 163-180.
Flash,T., Inzelberg.R., Schechtman.E., & Korczyn,A.D. (1992). Kinematic
analysis of upper limb trajectories in Parkinson's disease. Exp.NeuroL 118.
215-226.
Flowers,K.A. (1976). Visual "closed-loop" and "open-loop" characteristics of
voluntary movement in patients with Parkinsonism and intention tremor.
Brain. 99. 269-310.
Flowers,K.A. (1978). Lack of prediction in the motor behaviour of Parkinsonism.
Brain. 101. 35-52.
Folstein.M.F., Folstein.S.E., & McHugh,P.R. (1975). "Mini-mental state"; A
practical method for grading the cognitive state of patients for the clinician.
J.Psvchiatr.Res., 12. 189-198.
Frith,C.D., Bloxham.C.A., & Carpenter,K.N. (1986). Impairments in the learning
and performance of a new manual skill in patients with Parkinson's disease.
J.Neurol.Neurosurg.Psychiatry. 49. 661-668.
Gabrieli, J.D. (1995). Contribution of the basal ganglia to skill learning and working
memory in humans. In J. Houk, J. Davis, & D. Beiser (Eds.), Models of
information processing in the basal ganglia, (pp. 277-294). Cambridge, MA:
MIT Press.
Gentilucci.M., & Negrotti.A. (1999). Planning and executing an action in
Parkinson's disease. Mov Disord.. 14. 69-79.
Georgopoulos.A.P. (2000). Neural aspects of cognitive motor control.
Curr.Qpin.NeurobioL 10. 238-241.
Gottlieb.G.L., Corcos,D.M., & Agarwal,G.C. (1989). Organizing principles for
single-joint movements. I. A speed-insensitive strategy. J.Neurophvsiol.,
62, 342-357.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
126
Grafton,S.T., Hazeltine,E., & lvry,R. (1995). Functional mapping of sequence
learning in normal humans. J.Coqn.Neurosci., 7. 497-510.
Graybiel.A.M., Canales,J.J., & Capper-Loup,C. (2000). Levodopa-induced
dyskinesias and dopamine-dependent stereotypies: A new hypothesis.
Trends. Neurosci.. 23 (Suppl.). s71 -s77.
Haaland.K.Y., & Harrington,D.L. (1990). Complex movement behavior: Toward
understanding cortical and subcortical interactions in regulating control
processes. In G. E. Hammond (Ed.), Cerebral control of speech and limb
movements, (pp. 169-200). North-Holland: Elsevier Science.
Haaland.K.Y., Harrington,D.L., O'Brien,S., & Hermanowicz.N. (1997). Cognitive-
motor learning in Parkinson's disease. Neuropsychology. 11, 180-186.
Hallett.M., & Khoshbin.S. (1980). A physiological mechanism of bradykinesia.
Brain. 103. 301-314.
Harrington,D.L., & Haaland.K.Y. (1991). Sequencing in Parkinson's disease.
Abnormalities in programming and controlling movement. Brain, 114, 99-
115.
Harrington,D.L., Haaland.K.Y., Yeo.R.A., & Marder.E. (1990). Procedural memory
in Parkinson's disease: impaired motor but not visuoperceptual learning.
J.CIin.Exp.Neuropsvchol.. 12. 323-339.
Hary, D. (1996). Template (version 3.0) computer software. Santa Monica:
Integrated Scientific Resources.
Hary.D. (2000). GMP Program (version 4.0.11) computer software. Santa Monica:
Integrated Scientific Resources.
He.S.Q., Dum.R.P., & Strick,P.L. (1993). Topographic organization of
corticospinal projections from the frontal lobe: Motor areas on the lateral
surface of the hemisphere. J. Neurosci., 13, 952-980.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
127
Heindel,W.C., Salmon,D.P., Shults,C.W., Walicke,P.A., & Butters,N. (1989).
Neuropsychological evidence for multiple implicit memory systems: a
comparison of Alzheimer's, Huntington's, and Parkinson's disease patients.
J.Neurosci., 9. 582-587.
Henry,F.M., & Rogers,D.E. (1960). Increased response latency for complicated
movements and a "memory drum" theory of neuromotor reaction. Research
Quarterly, 31. 448-458.
Hoehn.M.M., & Yahr,M.D. (1967). Parkinsonism: onset, progression and
mortality. Neurology. 17. 427-442.
Honda,M., Deiber,M.P., Ibanez,V., Pascal-Leone,A., Zhuang,P., & Hallet.M.
(1998). Dynamic cortical involvement in implicit and explicit motor sequence
learning. Brain. 120. 123-140.
Hoover,J.E., & Strick,P.L. (1993). Multiple output channels in the basal ganglia.
Science. 259. 819-821.
Hoshiyama.M., Kaneoke,Y., Koike,Y., Takahashi,A., & Watanabe.S. (1994).
Hypokinesia of associated movement in Parkinson's disease: a symptom in
early stages of the disease. J.Neurol.. 241, 517-521.
Jackson,G.M., Jackson,S.R., Harrison,J., Henderson,L., & Kennard,C. (1995).
Serial reaction time learning and Parkinson's disease: evidence for a
procedural learning deficit. Neuropsvcholoqia. 33. 577-593.
Jenkins,I.H., Brooks,D.J., Nixon,P.D., Frackowiak.R.S., & Passingham,R.E.
(1994). Motor sequence learning: a study with positron emission
tomography. J.Neurosci.. 14. 3775-3790.
Jennings,P.J. (1995). Evidence for incomplete motor programming in Parkinson's
disease. Journal of Motor Behavior. 27. 310-324.
Jordan,N., & Sagar,H.J. (1994). The role of the striatum in motor learning:
dissociations between isometric motor control processes in Parkinson's
disease. Int.J.Neurosci., 77. 153-165.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
128
Jueptner.M., Frith,C.D., Brooks,D.J., Frackowiak.R.S., & Passingham.R.E. (1997).
Anatomy of motor learning. II. Subcortical structures and learning by trial and
error. J.Neurophysiol., 77, 1325-1337.
KnowIton,B.J., Mangels,J.A., & Squire,L.R. (1996). A neostriatal habit learning
system in humans. Science. 273, 1399-1402.
Koenig,O., Thomas-Anterion.C., & Laurent,B. (1999). Procedural learning in
Parkinson's disease: intact and impaired cognitive components.
Neuropsvcholoqia. 37. 1103-1109.
Koller.W.C., & Montgomery,E.B. (1997). Issues in the early diagnosis of
Parkinson's disease. Neurology, 49. S10-S25.
Lai,Q., & Shea,C.H. (1998). Generalized motor program (GMP) learning: Effects
of reduced frequency of knowledge of results and practice variability.
Journal of Motor Behavior. 30. 51-59.
Lai,Q., Shea,C.H., Wulf,G., & Wright,D.L. (2000). Optimizing generalized motor
program and parameter learning. Res.Q.Exerc.Sport. 71. 10-24.
Lee,T.D. (1988). Transfer-appropriate processing: A frame work for
conceptualizing practice effects in motor learning. In O.G.Meijer & K.Roth
(Eds.), Complex movement behavior: The motor-action controversy, (pp.
201-215). North-Holland: Elsevier Science Publishers.
Lee.T.D., & Magill,R.A. (1983). The locus of contextual interference in motor skill
acquisition. J.Exp.Psvchol.Learn.Mem.Coqn.. 9, 730-746.
Lee.T.D., & Magill,R.A. (1985). "Can forgetting facilitate skill acquisition?". In D.
Goodman, R. W. Wilberg, & I. M. Franks (Eds.), Differing perspectives in
motor learning, memory, and control, (pp. 3-22). Amsterdam: North-Holland.
Lee.T.D., Swinnen,S.P., & Serrien,D.J. (1994). Cognitive effort and motor
learning. Quest. 46. 328-344.
Magill,R.A., & Hall.K.G. (1990). A review of a contextual interference effect in
motor skill acquisition. Human Movement Science. 9. 241-289.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
129
Marsden.C.D. (1989). Slowness of movement in Parkinson's disease. Mov
Disord.. 4 Suppl 1, S26-S37.
Martin,J.H. (1996). Neuroanatomv. (2 ed.). Stamford, Connecticut: Appleton &
Lange.
Martin,K.E., Phillips,J.G., lansek,R., & Bradshaw,J.L. (1994). Inaccuracy and
instability of sequential movements in Parkinson's disease. Exp.Brain Res.,
102. 131-140.
Mayeux.R., Denaro,J., & Hemenegildo.N. (1992). A population-based
investigation of Parkinson's disease with and without dementia: relationship
to age and gender. Arch.Neurol.. 49, 492-497.
McLennan,J.E., Nakano.K., Tyler,H.R., & Schwab,R.S. (1972). Micrographia in
Parkinson's disease. J.Neurol.Sci.. 15, 141-152.
Middleton,F.A., & Strick,P.L. (2000). Basal ganglia output and cognition: evidence
from anatomical, behavioral, and clinical studies. Brain Goan. 42. 183-200.
Milner,B., Corkin.S., Teuber.H.L. (1968). Further analysis of the hippocampal
amnesic syndrome: 14 yr follow-up study of HM. Neuropsvcholoqia, 6, 215-
234.
Miyachi.S., Hikosaka,0., Miyashita.K., Karadi.Z., & Rand.M.K. (1997). Differential
roles of monkey striatum in learning of sequential hand movement.
Exp.Brain Res., 115, 1-5.
Morris,M.E. (2000). Movement disorders in people with Parkinson disease: a
model for physical therapy. Phvs.Ther., 80, 578-597.
Morris.C.D., Bransford,J.D., & Franks,J.J. (1977). Levels of processing versus
transfer appropriate processing. Journal of verbal learning and verbal
behavior, 16. 519-533.
Morris,M.E., & lansek,R. (1996). Characteristics of motor disturbance in
Parkinson's disease and strategies for movement rehabilitation. Human
Movement Science. 15, 649-669.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
130
Morris.M.E., lansek,R., Matyas,T.A., & Summers,J.J. (1996). Stride length
regulation in Parkinson's disease. Normalization strategies and underlying
mechanisms. Brain, 119, 551-568.
Morris,R.G., Downes,J.J., Sahakian.B.J., Evenden.J.L., Heald,A., & Robbins,T.W.
(1988). Planning and spatial working memory in Parkinson's disease.
J.Neurol.Neurosurq.Psychiatry, 51, 757-766.
Mushiake.H., Inase.M., & Tanji.J. (1991). Neuronal activity in the primate
premotor, supplementary, and precentral motor cortex during visually guided
and internally determined sequential movements. J. Neuphvsiol., 66, 705-
718.
Nissen.M.J., & Bullemer,P. (1987). Attentional requirements of learning: Evidence
from performance measures. Cognitive Psychology, 19, 1-32.
Oldfield,R.C. (1971). The assessment and analysis of handedness: the Edinburgh
inventory. Neuropsychologia, 9,97-113.
Oliveira,R.M., Gurd.J.M., Nixon,P., Marshall,J.C., & Passingham,R.E. (1997).
Micrographia in Parkinson's disease: the effect of providing external cues.
J.Neurol.Neurosurg.Psychiatry, 63, 429-433.
Onla-or,S., & Winstein.C.J. (2001). Function of the ‘direct’ and ‘indirect’ pathways
of the basal ganglia motor loop: Evidence from reciprocal aiming movements
in Parkinson’s disease. Cognitive Brain Research, 10. 329-332.
Owen,A.M., Beksinska.M., James,M., Leigh,P.N., Summers,B.A., Marsden,C.D.,
Quinn,N.P., Sahakian.B.J., & Robbins,T.W. (1993). Visuospatial memory
deficits at different stages of Parkinson's disease. Neuropsychologia, 31,
627-644.
Owen,A.M., & Doyon.J. (1999). The cognitive neuropsychology of Parkinson's
disease: a functional neuroimaging perspective. Adv.Neurol., 80, 49-56.
Owen.A.M., Doyon.J., Dagher.A., Sadikot.A., & Evans,A.C. (1998). Abnormal
basal ganglia outflow in Parkinson's disease identified with PET. Implications
for higher cortical functions. Brain, 121, 949-965.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
131
Owen,A M., James,M., Leigh,P.N., Summers,B.A.,Marsden,C.D., Quinn,N.P.,
Lange,K.W., & Robbins,T.W. (1992). Fronto-striatal cognitive deficits at
different stages of Parkinson's disease. Brain, 115, 1727-1751.
Owen,A.M., Roberts,A.C., Polkey,C.E., Sahakian.B.J., & Robbins,T.W. (1991).
Extra-dimensional versus intra-dimensional set shifting performance
following frontal lobe excisions, temporal lobe excisions or amygdalo-
hippocampectomy in man. Neuropsvcholoqia, 29, 993-1006.
Parent,A. (1996). Carpenter's human neuroanatomy. (9 ed.). PA: Williams &
Wilkins, (pp 760-773).
Pascual-Leone,A., Grafman.J., Clark,K., Stewart,M., Massaquoi,S., Lou,J.S., &
Hallett.M. (1993). Procedural learning in Parkinson's disease and cerebellar
degeneration. Ann.Neurol., 34, 594-602.
Phillips,J.G., Martin,K.E., Bradshaw,J.L., & lansek.R. (1994). Could bradykinesia
in Parkinson's disease simply be compensation? J.Neurol., 241, 439-447.
Reber,A.S. (1989). Implicit learning and tacit knowledge. J.Exp.Psvchol.Gen,
118, 219-235.
Robertson,C., & Flowers,K.A. (1990). Motor set in Parkinson's disease.
J.Neurol.Neurosurq.Psychiatry, 53, 583-592.
Rolls,E.T. (1994). Neurophysiology and cognitive functions of the striatum.
Rev.Neurol.(Paris). 150, 648-660.
Roncacci.S., Troise,E., Carlesimo.G.A., Nocentini.U., & Caltagirone.C. (1996).
Implicit memory in parkinsonian patients: evidence for deficient skill learning.
Eur.Neurol.. 36, 154-159.
Saint-Cyr,J.A., Taylor,A.E., & Lang,A.E. (1988). Procedural learning and
neostriatal dysfunction in man. Brain, 111, 941-959.
Sakai,K., Hikosaka,0., Miyauchi,S., Takino.R., Sasaki,Y., & Putz, B. (1998).
Transition of brain activation from frontal to parietal areas in visuomotor
sequence learning. J. Neurosci, 18, 1827-1840.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
132
Salmon,D.P., & Butters,N. (1995). Neurobiology of skill and habit learning.
Curr.Opin.Neurobiol., 5, 184-190.
Salmoni.A.W., Schmidt,R.A., & Walter,C.B. (1984). Knowledge of results and
motor learning: A review and critical reappraisal. Psychological Bulletin, 95,
355-386.
Sandson.J., & Albert,M.L. (1987). Perseveration in behavioral neurology.
Neurology. 37. 1736-1741.
Schmidt,R.A. (1975). A schema theory of discrete motor skill learning.
Psychol.Rev.. 82. 225-260.
Schmidt,R.A. (1991). Frequent augmented feedback can degrade learning:
Evidence and interpretations. In G.E.Stelmach & J.Requin (Eds.), Tutorials
in motor neuroscience, (pp. 59-75). Dordrecht: Kluwer Academic
Publishers.
Schmidt,R.A., & Bjork,R.A. (1992). New conceptualizations of practice: Common
principles in three paradigms suggest new concepts for training.
Psychological Science. 3. 207-217.
Schmidt,R.A., & Lee.T.D. (1999). Motor control and learning: A behavioral
emphasis. Champaign, IL: Human Kinetics.
Schmidt,R.A., Young,D.E., Swinnen.S.P., & Shapiro,D.C. (1989). Summary
knowledge of results for skill acquisition: Support for the guidance
hypothesis. J.Exp.Psychol.Learn.Mem.Cogn., 15. 352-359.
Seitz,R., & Roland, P.E. (1992). Learning of sequential finger movements in man:
A combined kinematic and positron emission tomography (PET) study.
Eur.J.Neurosci., 4 , 154-165.
Sekiya.H., Magill.R.A., & Anderson,D.I. (1996). The contextual interference effect
in parameter modifications of the same generalized motor program.
Res.Q.Exerc.Sport.. 67. 59-68.
Sekiya.H., Magill.R.A., Sidaway.B., & Anderson,D.I. (1994). The contextual
interference effect for skill variations from the same and different generalized
motor programs. Res.Q.Exerc.Sport., 65, 330-338.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
133
Shea.C.H., Kohl.R., & lndermi!I,C. (1990). Contextual interference: Contributions
of practice. Acta Psychologica, 73, 145-157.
Shea,J.B., & Morgan,R.L. (1979). Contextual interference effects on the
acquisition, retention, and transfer of a motor skill. Journal of Experimental
Psychology: Human Learning and Momory, 5, 179-187.
Shea.J.B., & Zimmy.S.T. (1983). Context effects in memory and learning
movement information. In R. A. Magill (Ed.), Memory and control of actions,
(pp. 145-366). Amsterdam, North-Holland.
Soliveri.P., Brown,R.G., Jahanshahi,M., Caraceni.T., & Marsden.C.D. (1997).
Learning manual pursuit tracking skills in patients with Parkinson's disease.
Brain. 120. 1325-1337.
Squire,L.R. (1987). Memory and brain. New York: Oxford University Press.
Squire,L.R. (1992). Declarative and non-declarative memory: Multiple brain
systems supporting learning and memory. J.Cogn.Neurosci., 4 , 232-243.
Squire,L.R. & Zola-Morgan,S. (1991). The medial temporal lobe memory system.
Science. 253. 1380-1385.
Strick,P.L., Dum.R.P., & Picard,N. (1995). Macro-organization of the circuits
connecting the basal ganglia with cortical motor areas. In J. Houk, J. Davis,
& D. Beiser (Eds.), Models of information processing in the basal ganglia,
(pp. 117-127). Cambridge, MA: MIT Press.
Sullivan,K.J. & Winstein.C.J. (submitted). Sensorimotor contribution to the
encoding of an optimal motor routine.
Tamaru.F. (1997). Disturbances in higher function in Parkinson's disease.
Eur.Neurol., 38 Suppl 2 . 33-36.
Tresilian.J.R., Stelmach.G.E., & Adler,C.H. (1997). Stability of reach-to-grasp
movement patterns in Parkinson's disease. Brain, 120, 2093-2111.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
134
Van Gemmert.A.W., Teulings,H.L., Contreras-Vidal.J.L., & Stelmach.G.E. (1999).
Parkinson's disease and the control of size and speed in handwriting.
Neuropsychologia, 37, 685-694.
Van Spaendonck.K.P., Berger,H.J,, Horstink.M.W., Borm.G.F., & Cools,A.R.
(1996). Memory performance under varying cueing conditions in patients
with Parkinson's disease. Neuropsvcholoqia. 34. 1159-1164.
Verschueren.S.M., Swinnen.S.P., Dom,R., & De Weerdt.W. (1997). Interlimb
coordination in patients with Parkinson's disease: motor learning deficits and
the importance of augmented information feedback. Exp.Brain Res., 113,
497-508.
Vakil,E. & Herishanu-Naaman.S. (1998). Declarative and procedural learning in
Parkinson’s disease patients having tremor or bradykinesia as the
predominant symptom. Cortex, 34, 611-620.
Weiss,P., Stelmach.G.E., & Hefter.H. (1997). Programming of a movement
sequence in Parkinson's disease. Brain. 120, 91-102.
Willingham,D.B. (1998). A neuropsychological theory of motor skill learning.
Psychol.Rev., 105, 558-584.
Willingham,D.B. (1999). The neural basis of motor-skill learning. Current
Directions in Psychological Science, 8, 178-182.
Winstein, C.J. Relative frequency of information feedback in motor performance
and learning (Doctoral dissertation, University of California, Los Angeles,
1988). Dissertation Abstracts International, 49, 028 (1988), 0561.
Winstein,C.J., Merians.A.S., & Sullivan,K.J. (1999). Motor learning after unilateral
brain damage. Neuropsvcholoqia, 37, 975-987.
Winstein,C.J., Pohl.P.S., Cormack.J.C., & Waters,C.H. (1995). Measurement of
upper extremity motor performance with pharmacological treatment of
individuals with severe Parkinson disease. Poster session presented at the
12th International Congress of the World Confederation for Physical
Therapy, Washington, DC.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
135
Winstein,C.J., Pohl.P.S., & Lewthwaite.R. (1994). Effects of physical guidance
and knowledge of results on motor learning: support for the guidance
hypothesis. Res.Q.Exerc.Sport.. 65. 316-323.
Winstein,C.J., & Schmidt,R.A. (1990). Reduced frequency of knowledge of results
enhances motor skill learning. J.Exp.Psychol. Learn.Mem.Coon.. 16. 677-
691.
Wise,S.P., Murray,E.A., & Gerfen.C.R. (1996). The frontal cortex-basal ganglia
system in primates. Crit.Rev.Neurobiol.. 10, 317-356.
Wishart.L.R., & Lee.T.D. (1997). Effects of aging and reduced relative frequency
of knowledge of results on learning a motor skill. Perceptual and Motor
Skills. 84. 1107-1122.
Worringham.C.J., & Stelmach.G.E. (1990). Practice effects on the
preprogramming of discrete movements in Parkinson's disease.
J.Neurol.Neurosurg.Psychiatry, 53, 702-704.
Wright,D.L., Li,Y., & Coady.W. (1997). Cognitive processes related to contextual
interference and observational learning: A replication of Blandin, Proteau,
and Alain. Res.Q.Exerc.Sport.. 68. 106-109.
Wulf.G., Lee.T.D. , & Schmidt,R.A. (1994). Reducing knowledge of results about
relative versus absolute timing: Differential effects on learning. Journal of
Motor Behavior, 26. 362-369.
Wulf.G., & Schmidt,R.A. (1994). Feedback-induced variability and the learning of
generalized motor programs. Journal of Motor Behavior, 26, 348-361.
Wulf.G., Schmidt,R.A., & Deubel.H. (1993). Reduced feedback frequency
enhances generalized motor program learning but not parameterization
learning. J.Exp.Psychol.Learn.Mem.Cogn., 19. 1134-1150.
Yamadori.A., Yoshida.T., Mori.E., & Yamashita.H. (1996). Neurological basis of
skill learning. Brain Res.Cogn Brain Res., 5. 49-54.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
136
Appendix A
Hoehn and Yahr scale of clinical stages for Parkinson’s disease (Hoehn & Yahr, 1967)
Stage Signs
0 No clinical signs evident
1 Unilateral involvement only
2 Bilateral involvement without impairment of balance
3 First sign of impaired postural and righting reflexes by examination
or a history of poor balance, falls, etc. Disability is mild to moderate.
4 Fully developed severe disease; disability marked
5 Confinement to bed or wheelchair
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
137
Appendix B
Mini-Mental Status Exam (Folstein, Folstein, & McHugh, 1975)
Subject Name:___________________________ Date:
1. What is the year , season , date , day , month _
2. Where are we: state , country , town , building , floor.
3. Name 3 objects: orange , airplane_______, tobacco_______ , (trials _
4. Serial 7’s :
(93) (86) (79) (72) (65)
or spell “ world" backwards:_______________________
(d) (I) (r) (o) (w)
5. Recall 3 objects: orange , airplane , tobacco .
6. Name a pencil___________, and a watch____________ .
7. Read and Obey_______________. -----------------►
CLOSE YOUR EYES
8. Copy design . (see below)
9. Write a sentence
10. Repeat the following: “No Ifs, ands, or buts”
11. Follow a 3-stage command: a: take a paper in your right hand.
b. fold it in half____________.
c. put it on the floor_________.
Total Score =
Copy design
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix C
Subject Name:__________ _ Date . ______________
Center for Epidemiologic Studies Depression (CES-D) Scale
During the past week.....
1. I was bothered by things that usually don’t bother me.________ _____
2. I had trouble keeping my mind on what I was doing. _____
3. I felt depressed. _____
4. I felt that everything I did was an effort. _____
5. I felt hopeful about the future*. _____
6. I felt fearful. _____
7. My sleep was restless. _____
8. I was happy*. _____
9. I felt lonely. _____
10.1 could not get “ going”.___________________________________ _____
Total Score = ______
Rating Score:
0 = none of the time
1 = sometimes
2 = often
3 = most of the time
* Score will be converted by the experimenter for statement # 5 and 8.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
139
Appendix D
EDINBURGH HANDEDNESS INVENTORY
Subject Name_____________________________ Control_____ PD
Please indicate your preferences in the use of hands in the following
activities by putting X in the appropriate column. Put XX where the preference is so
strong that you would never try to use the other hand unless forced. If in any case
you are really indifferent, put X in both columns. Some of the activities require both
hands. In these cases the part of the task for which hand preference is wanted is
indicated in brackets.
Left Right
1. Writing________________________________ _____ _____
2. Drawing _____ _____
3. Throwing _____ _____
4. Scissors _____ _____
5. Toothbrush _____ _____
6. Knife (without fork) _____ _____
7. Spoon _____ _____
8. Broom (upper hand) _____ _____
9. Striking Match (match) _____ _____
10. Opening Box (lid)_______________________ _____ _____
i which foot do you prefer to kick with? _____ _____
ii which eye do you use when using only one? _____
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
140
Appendix E
Subject Name:______________________________ Date_______________
The following items are about activities you might do during a typical day. Does
your health now limit in these activities? If so, how much?
Circle one number on each line.
SF-36 (Physical Function Component)
Activities Yes, Yes, No,
limited limited not limited
a lot. a little. at all.
Vigorous activities such as running, lifting 1 2 3
heavy objects, participating in strenuous sports
Moderate activities such as moving a table, 1 2 3
Pushing a vacuum cleaner, bowling, or playing golf
Lifting or carrying groceries 1 2 3
Climbing several flights of stairs 1 2 3
Climbing one flight f stairs 1 2 3
Bending, kneeling, or stooping 1 2 3
Walking more than one mile 1 2 3
Walking several blocks 1 2 3
Walking one block 1 2 3
Bathing or dressing yourself 1 2 3
Total Score:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Quantifying musculoskeletal load and adaptation: Biomechanical consideration

PDF

The interaction between explicit knowledge and implicit motor -sequence learning following focal brain damage

PDF

The influence of patella alta on knee extensor mechanics and patellofemoral joint stress

PDF

At what age are gait characteristics mature? Evaluation of gait kinematics, kinetics, and intersegmental dynamics in 7 year-old children

PDF

Satellite cell proliferation: two models and two anabolic stimuli

PDF

Mechanical demand distribution during multi-joint tasks

PDF

The role of the sensorimotor cortical system in skill acquisition and motor learning: a behavioral study

PDF

Cardiovascular fitness, exercise, and metabolic disease risk in overweight Latino youth

PDF

Mothering occupations: An ethnographic study of Thai women with physical disabilities in the north of Thailand

PDF

The influence of body weight and composition, pubertal status, tobacco use and exposure, physical activity and muscle strength on bone mass of Chinese adolescents

PDF

Participation in occupation and quality of life in individuals with Parkinson's disease

PDF

Biomechanical and neuromuscular aspects of non-contact ACL injuries: The influence of gender, experience and training

PDF

Relationship between cognition and function as measured by the Allen Cognitive Levels and the Functional Independence Measure

PDF

Rehabilitative narratives of individuals with neurological illnesses and their caregivers in a rehabilitation unit in central Taiwan: An ethnographic study

PDF

Constructing identities in social worlds: Stories of four adults with autism

PDF

The Personal Projects Approach: A method to use with persons with multiple sclerosis

PDF

Therapeutics and science in the history of occupational therapy

PDF

Portal vein hypoglycemia sensor: Elucidating metabolic signaling and pathogenesis of hypoglycemia-associated sympathoadrenal failure

PDF

Bimanual coordination of goal-directed multi-joint rapid aiming movements following stroke

PDF

Factors contributing to patellofemoral joint stress: a comparison of persons with and without patellofemoral pain

Asset Metadata

Creator Onla-or, Somporn (author)

Core Title Motor skill learning in individuals with Parkinson's disease: Consideration of cognitive and motor demands

Contributor Digitized by ProQuest (provenance)

Degree Doctor of Philosophy

Degree Program Biokinesiology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag health sciences, rehabilitation and therapy,OAI-PMH Harvest,psychology, behavioral,psychology, physiological

Language English

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-106381

Unique identifier UC11328268

Identifier 3027761.pdf (filename),usctheses-c16-106381 (legacy record id)

Legacy Identifier 3027761.pdf

Dmrecord 106381

Document Type Dissertation

Rights Onla-or, Somporn

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA