doi:10.1093/brain/awm059
Brain (2007), 130, 1395^1407
Semantic and affective priming as a function
of stimulation of the subthalamic nucleus in
Parkinson’s disease
Joanna E. Castner,1 Helen J. Chenery,1 David A. Copland,1 Terry J. Coyne,2 Felicity Sinclair2 and
Peter A. Silburn2,,3
1
Centre for Research in Language Processing and Linguistics, Division of Speech Pathology, School of Health and
Rehabilitation Sciences, The University of Queensland, 2St Andrew’s War Memorial Hospital and The Wesley Hospital,
Brisbane, Queensland, Australia and 3The Eskitis Institute, Clinical Neurosciences Group, Griffith University, Nathan,
Queensland, Australia
Correspondence to: Helen Chenery, Division of Speech Pathology, School of Health and Rehabilitation Sciences,
The University of Queensland, Brisbane Queensland 4072, Australia
E-mail: [email protected]
Lexical-semantic and emotional processing deficits have been associated with Parkinson’s disease. This study
investigated automatic and controlled lexical-semantic processing, the automatic activation of emotional evaluations, and the processing of words conveying negative and neutral emotional connotations in a combined
affective and semantic priming paradigm. Eighteen participants with Parkinson’s disease who had undergone
surgery for deep brain stimulation (DBS) of the subthalamic nucleus (STN) completed a lexical decision task
at short and long stimulus onset asynchronies (SOAs), during on and off stimulation conditions. Nineteen nonneurologically impaired participants acted as controls.The results indicated that automatic lexical-semantic and
emotional evaluative processes are unimpaired in Parkinson’s disease as reflected in the presence of comparable
semantic and affective priming effects at the short SOA in on and off stimulation conditions compared with
healthy controls. In contrast, participants with Parkinson’s disease in the off stimulation condition showed a
pattern of aberrant controlled lexical-semantic processing as evidenced by a lack of semantic priming effects
at the long SOA condition. Controlled semantic priming was present, however, when the participants with
Parkinson’s disease were receiving stimulation of the STN, suggesting that STN stimulation modulates basal
ganglia-thalamocortical circuits involved in such processes. Finally, delayed reaction times for negatively
valenced targets compared with neutrally valenced targets was evident in participants with Parkinson’s disease
in the on stimulation condition and control participants, but not for participants with Parkinson’s disease in the
off stimulation condition, suggesting that the incidental evaluation of negatively versus neutrally valenced words
in Parkinson’s disease is modulated by basal ganglia-thalamocortical circuits.
Keywords: lexical decision; emotion; deep brain stimulation; semantic priming; language processing
Abbreviations: ACC ¼ anterior cingulate cortex; ANEW ¼ Affective Norms for English Words; DBS ¼ deep brain stimulation; LMM ¼ linear mixed model; RT ¼reaction time; SOA ¼ stimulus onset asynchrony; STN ¼ subthalamic nucleus.
Received December 21, 2006. Revised February 17, 2007. Accepted March 5, 2007. Advance Access publication April 12, 2007
Introduction
Deficits in emotional processing and lexical-semantic
processing have been described in people with Parkinson’s
disease and there is emerging evidence that these processes
are affected by stimulation of the subthalamic nucleus
(STN) in Parkinson’s disease patients who have
received surgery for deep brain stimulation (DBS).
To date, there have been no studies that have concurrently
investigated automatic and controlled lexical-semantic and
affective processes, and explored the processing of words
with emotional connotations in Parkinson’s disease. This
study will explore these processes in Parkinson’s disease and
also investigate the degree to which stimulation of the
STN modulates these processes via functional changes in
non-motor basal ganglia-thalamocortical circuits. Therefore,
this study will investigate the incidental processing of
ß The Author (2007). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved. For Permissions, please email: [email protected]
1396
Brain (2007), 130, 1395^1407
emotional linguistic stimuli via a lexical decision in
addition to the automatic activation of emotional evaluations (i.e. the automatic assessment of the emotional
valence of stimuli), and the automatic and controlled
activation of lexical-semantic representations by utilizing
a combined affective and semantic priming paradigm
in participants with Parkinson’s disease. Furthermore,
the involvement of basal ganglia-thalamocortical circuits
on such processes in Parkinson’s disease will be measured
via the neuromodulation of the STN.
Lexical-semantic processing in Parkinson’s
disease
Investigations into the integrity of lexical-semantic
processes in Parkinson’s disease have been largely directed
at measures of language processes, such as verbal fluency
paradigms, that are confounded by strong executive
processes. Deficits in performance on verbal fluency tasks
have often been reported in Parkinson’s disease (e.g. Piatt
et al., 1999; Zec et al., 1999), suggesting that language
processes that have a strong executive component are
impaired in Parkinson’s disease. Furthermore, verbal
fluency studies have indicated poorer performance with
STN stimulation (e.g. Moro et al., 1999; Saint-Cyr et al.,
2000; Alegret et al., 2001; Dujardin et al., 2001; Moretti
et al., 2002; Daniele et al., 2003; Funkiewiez et al., 2004;
De Gaspari et al., 2006); however, studies have also
reported no significant change in verbal fluency with STN
stimulation (e.g. Limousin et al., 1998; Burchiel et al., 1999;
Jahanshahi et al., 2000; Lopiano et al., 2001; Perozzo et al.,
2001; Whelan et al., 2003). Therefore, whilst declines in
verbal fluency as a function of STN stimulation have been
described, the results of such studies are far from
conclusive. Also, the use of paradigms with strong executive
components (such as response initiation, set shifting and
semantic monitoring) makes it difficult to determine
whether specific lexical-semantic processes are impaired in
Parkinson’s disease and not just reflective of dysexecutive
symptoms typically associated with Parkinson’s disease.
There have been a small number of studies that have
investigated automatic and/or controlled semantic priming
using lexical decision paradigms in Parkinson’s disease
(but not as a function of STN stimulation), which enable a
more precise mechanism for measuring the integrity of
lexical-semantic processes.
In a lexical decision task investigating semantic
priming, word/non-word judgements are made on target
stimuli which are usually preceded by a semantically related
or a semantically unrelated prime word. Semantic
priming is said to occur when participant responses to
real-word targets are faster for related word pair conditions
(e.g. cat–dog) as opposed to the unrelated conditions
(e.g. cat–book). When investigating semantic priming,
the primary factor under investigation is the semantic
relatedness of the word pairs. The automaticity of
J. E. Castner et al.
semantic priming effects can be manipulated by varying
the stimulus onset asynchrony (SOA; i.e. the time between
the onset of the prime and the onset of the target). The
primary purpose of manipulating the SOA in semantic
priming paradigms is to provide a measure of the time
course of the activation of lexical-semantic representations
and to provide an understanding of the attentional
mechanisms involved in semantic priming. If priming is
only found when SOAs are brief (e.g. 300 ms or less),
priming is thought to involve non-strategic, automatic
processing. In contrast, if priming is found at longer SOAs
(e.g. 1000 ms), controlled attentional mechanisms are
thought to be involved (Neely, 1977, 1991; Fazio, 2001).
Semantic priming is usually found at both short SOAs and
often to a greater extent at longer SOAs due to the
likelihood that inhibition on unrelated word pairs can
also contribute to the magnitude of semantic priming
effects (see Neely, 1991 for further information regarding
facilitation and inhibition effects associated with semantic
priming).
Investigations into semantic priming effects in
Parkinson’s disease first emerged over a decade ago,
where participants with Parkinson’s disease showed
differing patterns of priming compared with healthy
controls (Spicer et al., 1994; McDonald et al., 1996).
These studies focused their attention on expectancy-based
strategic processes in semantic priming, and thus it was
difficult to establish whether the differing patterns of
semantic priming were due to strategic mechanisms or
lexical-semantic processing deficits. Arnott et al. (2001)
directly investigated the automatic and controlled
mechanisms of semantic priming and explored the time
course of semantic priming in Parkinson’s disease.
Participants with Parkinson’s disease presented with
a delayed time course of semantic activation during
automatic semantic priming, and impaired inhibitory
processes when controlled semantic priming was elicited.
Differing patterns of semantic priming as a function of
SOA in participants with Parkinson’s disease have also been
reported in semantic priming paradigms measuring
ambiguity priming (Copland, 2003), and semantic priming
via multi-priming paradigms (Angwin et al., 2003, 2005).
Aberrant lexical-semantic processing in Parkinson’s
disease has been attributed to the neuromodulatory
influence of dopamine depletion resulting in a decreased
signal-to-noise ratio in semantic networks or altering the
time course of semantic processing (Arnott et al., 2001;
Angwin et al., 2005). Research into the specific influence of
basal ganglia-thalamocortical circuits in automatic and
controlled lexical-semantic processing in Parkinson’s disease
via STN stimulation is yet to be explored.
Evidence that STN stimulation may also affect nonmotor basal ganglia-thalamocortical circuits (i.e. anterior
cingulate, dorsolateral prefrontal and orbitofrontal
circuits) is now emerging. The STN is reported to be
functionally segregated into sensorimotor, limbic and
Priming in Parkinson’s disease
associative regions (Parent and Hazrati, 1995; Hamani
et al., 2004). The target for STN stimulation, to alleviate
Parkinson’s disease motor symptoms, is the dorsolateral
(sensorimotor) portion of the STN. Stimulation of this
region of the STN, however, is now thought to spread to
the more medial limbic and cognitive regions of the STN as
well (Baron et al., 2002; Funkiewiez et al., 2003; McIntyre
et al., 2004). This study will therefore aim to measure
automatic and controlled semantic priming in Parkinson’s
disease and to determine the influence of the neuromodulation of basal ganglia-thalamocortical circuits via STN
stimulation on these processes.
Emotional processing in Parkinson’s disease
Emotional processing deficits have often been reported in
patients with Parkinson’s disease. These deficits have ranged
from impairments in the recognition of emotional prosody
of speech (e.g. Yip et al., 2003; Schroder et al., 2006) and
the recognition of affect in facial expressions (e.g. Borod
et al., 1990; Jacobs et al., 1995; Yip et al., 2003; Suzuki
et al., 2006) to impairments in the production of emotional
facial expressions (Borod et al., 1990; Jacobs et al., 1995),
affective prosody (e.g. Pell, 1996; Pell and Leonard, 2003),
affectively loaded sentences (e.g. Benke et al., 1998) and
emotional discourse (e.g. Crucian et al., 2001).
A small number of studies have explored emotional
processing in participants with Parkinson’s disease who are
receiving stimulation of the STN. Schneider and colleagues
(2003) researched mood induction, emotional memory,
emotional discrimination and general cognitive, motor and
self-reported mood evaluations in 12 participants with
Parkinson’s disease in on and off stimulation conditions.
STN stimulation was shown to have a positive influence on
self-reported mood state, mood induction, emotional story
recall and performance on motor tasks, yet STN stimulation
had no effect on facial emotion discrimination and
cognition. In contrast to Schneider et al. (2003), three
studies have revealed a tendency for STN stimulation to
cause deficits in facial discrimination of negative emotions
such as anger (Dujardin et al., 2004; Schroeder et al., 2004),
fear (Biseul et al., 2005), sadness (Dujardin et al., 2004) and
disgust (Dujardin et al., 2004). A neuroimaging study has
identified that deficits in the processing of facial expressions
of negative emotions or emotional situations in Parkinson’s
disease patients receiving STN stimulation are associated
with a reduction in the activation of the putamen and
fusiform gyrus in on stimulation conditions and increased
activity in the anterior cingulate cortex (ACC) in the off
stimulation condition during emotional processing (Geday
et al., 2006). There is also neurophysiological evidence
linking the STN to emotional processing. Kuhn et al. (2005)
investigated STN field potentials in participants with
Parkinson’s disease whilst viewing affectively neutral,
negative and positive picture stimuli. A significantly
greater decrease in STN alpha activity was noted when
Brain (2007), 130, 1395^1407
1397
participants viewed affective versus neutral stimuli, suggesting an STN role within the basal gangliathalamocortical circuitry which mediates the processing of
emotional stimuli (Kuhn et al., 2005).
Wieser et al. (2006) recently identified a dissociation
between the early processing of emotional stimuli compared
with explicit ratings made on emotional stimuli in
participants with Parkinson’s disease. Specifically, Wieser
and colleagues investigated event-related potentials in
participants with Parkinson’s disease and controls
whilst viewing emotional pictures in a rapid serial display
(i.e. explicit ratings were not required on the stimuli and
the rapid nature of the display made attentional processing
of the stimuli unlikely). There was no difference in
event-related potentials associated with this task in
Parkinson’s disease and controls indicating that participants
with Parkinson’s disease do not have a deficit in the early
evaluation of emotional stimuli. In contrast, when
the participants with Parkinson’s disease explicitly rated
emotional stimuli, their affective ratings represented an
emotional blunting compared with controls.
At present, there have been no studies that
have investigated the automatic activation of emotional
evaluations or the evaluation of words representing differing
emotional connotations in Parkinson’s disease. This study
will investigate such processes in Parkinson’s disease and also
measure these processes as a function of stimulation of the
STN. The automatic activation of emotional evaluations has
often been measured in affective priming paradigms (Fazio,
2001; Klauer and Musch, 2003). Theories of affective priming
have originated from studies examining semantic priming in
lexical decision tasks. In contrast to semantic priming
paradigms, the affective priming paradigm involves the
manipulation of affective relatedness. That is, if prime and
target words are congruent in emotional valence, reaction
times (RTs) are faster than when prime and target word pairs
represent different emotional valence (i.e. affectively incongruent) (Fazio et al., 1986). For example, the prime perjury
and target hazard are affectively congruent as they both carry
negative emotional valence. In contrast, the prime anxiety and
target cloth are considered to be affectively incongruent as
anxiety represents a negative emotional valence, yet cloth is
neutral in emotional valence. Responses to affectively
congruent word pairs (e.g. perjury–hazard) are generally
faster than responses to affectively incongruent stimuli (e.g.
anxiety–cloth). The RT advantage for affectively congruent
stimuli compared with affectively incongruent stimuli is
referred to as affective priming. Whilst semantic priming is a
measure of the activation of lexical-semantic representations,
affective priming is a measure of the activation of evaluations
(Fazio, 2001; Klauer and Musch, 2003).
Whilst there are numerous similarities in lexical decision
tasks measuring semantic association and tasks measuring
affective association, there are only a small number of
studies that have either controlled for semantic relationship
(Hermans et al., 2002) or concurrently manipulated
1398
Brain (2007), 130, 1395^1407
semantic
and
affective
relationships
(Hill
and
Kemp-Wheeler, 1989). The lack of control for semantic
relationship in affective priming paradigms leads to
confusion as to the independency of affective priming
based on affective/evaluative association and not semantic
association. As affectively congruent stimuli are likely to
also be semantically related, affective priming could be a
reflection of simultaneous semantic priming.
Hermans et al. (2002) investigated affective priming in a
valence decision task and a lexical decision task ensuring
that word pairs were semantically unrelated. Affective
priming was found to occur in both tasks, therefore
showing that affective priming can occur irrespective
of semantic relationship. In an earlier study, Hill and
Kemp-Wheeler (1989) manipulated both semantic
and affective relationships to provide further evidence of
affective priming occurring independently of semantic
association, and the presence of semantic priming
regardless of affective association. Hill and Kemp-Wheeler
used a lexical decision task with negative emotional
valenced targets paired with affectively related and
semantically unrelated primes (e.g. drought–muggers),
affectively unrelated and semantically related primes
(e.g. dream–nightmare), and affectively unrelated and
semantically unrelated primes (e.g. tapestry–crisis).
Emotionally neutral targets were also included, but were
only paired with semantically and affectively related
primes (e.g. saucers–cups) and semantically unrelated and
affectively related primes (e.g. aerosol–editor). Affective
priming occurred when controlled for semantic relatedness
(i.e. responses to emotional targets were faster when
paired with affectively congruent, semantically unrelated
primes as opposed to non-emotional primes that
were semantically unrelated) and semantic priming
occurred when controlled for affective relatedness (i.e.
responses to emotional targets were faster when they were
paired with semantically related primes as opposed to
semantically unrelated primes when the word pairs were
affectively incongruent). There was no difference in the
processing of negative valenced targets compared with
neutral valenced targets. This study provided support for
the independent mechanisms involved in semantic relatedness and emotional relatedness.
Like the semantic priming literature, the time course
of affective priming has been investigated in detail.
Affective priming is more commonly found at short SOAs
(Fazio et al., 1986; Hermans et al., 1994, 2001; Klauer et al.,
1997; De Houwer et al., 1998; Fazio, 2001; Klauer and
Musch, 2003), providing evidence that affective priming is
fast-acting and automatic. The presence of affective priming
only at short SOAs is also supported by the more recent
electrophysiological data (e.g. Ortigue et al., 2004; Taylor
and Fragopanagos, 2005) which has revealed that the
evaluation of emotional stimuli occurs in the very early prelexical stages of processing.
J. E. Castner et al.
In summary, there is evidence that emotional processing
and lexical-semantic deficits are associated with Parkinson’s
disease and that emotional processing deficits are thought
to change when participants with Parkinson’s disease
receive stimulation of the STN. This study aims to measure
the temporal aspects of lexical-semantic and affective
processing using a lexical decision task measuring affective
and semantic priming at a short and long SOA. Due to the
strong evidence reported in the affective priming literature
with respect to the automaticity of affective priming, it
would be expected that healthy controls will only exhibit
affective priming at the short SOA. As automatic affective
processing has been reported to be unaffected by
Parkinson’s disease (Wieser et al., 2006), the same patterns
of affective priming are predicted for the participants with
Parkinson’s disease. A combined affective and semantic
priming paradigm also allows for the investigation of the
incidental evaluation of word stimuli that differ in
emotional connotation. That is, whilst an evaluative
response (e.g. valence decision) is not required in a lexical
decision task, any RT variations for negatively valenced
targets compared with neutrally valenced targets, may
suggest an implicit variation in the evaluation of emotional
connotation. With respect to semantic priming, it is
predicted that semantic priming effects may differ as a
function of SOA in participants with Parkinson’s disease, in
line with the suggestion that participants with Parkinson’s
disease are likely to have differing time course of activations
within semantic memory, or deficient attentional
mechanisms associated with semantic priming at long
SOAs. The manipulation of STN stimulation in participants
with Parkinson’s disease who have received DBS, will
allow for a direct measure of the involvement of basal
ganglia-thalamocortical circuits in these processes.
Material and methods
Participants
Eighteen patients (13 males) diagnosed with Parkinson’s
disease who had received the implantation of permanent
DBS electrodes into the STN were included in the present
study (refer to Table 1 for participant information). All
participants with Parkinson’s disease had undergone
extensive neurological and psychiatric evaluation prior to
undergoing surgery for DBS. They met strict inclusion
criteria for admission to the DBS program which included
no evidence of significant psychiatric symptoms.
Participants with Parkinson’s disease were tested at least
4 months post-surgery and were considered to have stable
stimulator settings by their neurologist. Each participant
was tested with their stimulators turned on and again with
their stimulators turned off. The order of test condition was
counterbalanced, and at least 6 weeks lapsed between the
two testing sessions. For the off stimulation condition, the
stimulators were turned off for at least 1 hour prior to the
commencement of assessments. Four participants were no
Priming in Parkinson’s disease
Brain (2007), 130, 1395^1407
1399
Table 1 Parkinson’s disease participant information
Participant Age
Gender Education Handedness Disease duration Severity UPDRS L-Dopa
medication
(years)
(years)
(years)
(H & Y) III on
Score (mg/day)
Stimulator settings
Frequency Amplitude
Pulse
width L/R L/R (Hz) L/R (V)
(ms)
1
2
3
53
50
51
M
M
F
15
18
10
Right
Right
Right
15
7
5
3
2.5
2.5
10
10
6
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
67
62
60
55
73
41
64
69
57
72
71
76
61
65
60
F
M
M
M
F
M
M
M
M
M
F
M
M
F
M
8
21
15
8
9
18
9
23
11
13
10
12
17
7
10
Right
Right
Right
Right
Right
Right
Right
Right
Right
Right
Right
Right
Left
Right
Right
17
5
12
12
17
9
20
9
11
16
17
19
13
15
12
3
4
3
2.5
3
2.5
2
2.5
2
4
3
4
2
2
2.5
15
17
19
6
13
11
8
11
3
20
17
23
3
11
12
1150
0
(2 mg
Cabergoline)
1650
1750
1000
1000
0
400
0
1250
0
500
1250
1250
800
100
500
60/60
60/90
90/90
160/160
130/160
160/145
3.0/3.2
2.3/3.6
3.6/3.3
90/90
90/60
90/60
90/60
60/90
90/90
60/60
60/60
60/60
60/60
90/90
90/60
60/60
60/60
60/60
160/160
130/130
160/160
130/130
130/130
185/185
130/130
130/130
130/130
130/130
160/160
160/160
160/160
130/160
130/130
2.8/2.8
2.5/2.8
3.3/2.6
3.6/2.6
2.9/3.1
3.8/3.8
3.0/2.8
2.0/2.4
2.3/3.3
2.6/2.8
2.4/2.4
2.6/2.6
3.3/3.5
2.2/2.6
2.8/2.4
H & Y ¼ Hoehn and Yahr; UPDRS ¼ Unified Parkinson’s Disease Rating Scale; ‘on’ refers to on stimulation and on medication; L ¼ Left,
R ¼ Right.
longer taking levodopa medication; however, the remaining
participants were tested whilst taking their usual
medication.
Nineteen non-neurologically impaired participants (13
males) were recruited to act as controls. The average age of
the control participants was 62.2 (range ¼ 49–74) years and
they had received an average of 13.8 (range ¼ 9–21) years of
formal education. The age and education levels of the
control participants were not significantly different from
the participants with Parkinson’s disease (P ¼ 0.812 and
P ¼ 0.591, respectively). All control participants were right
handed, had no history of neurological impairment, were
not taking any medication deemed to affect neurological
functioning, had no self-reported hearing loss and had selfreported normal or corrected-to-normal vision. All participants provided informed consent for participation which
was obtained according to the Declaration of Helsinki and
this project was approved by the appropriate Institutional
Ethics Committees.
Apparatus
Experimental stimuli were presented using E-prime v1.1
(Psychology Software Tools). Participants’ responses were
made using a PST Serial Response Box model 200a
(Psychology Software Tools), allowing the acquisition of
RTs with millisecond accuracy. The stimuli were displayed
on a high-resolution (1680 1050 pixels) colour monitor,
with letter strings presented in black text on a white
background. Letters were displayed in lowercase bold
Helvetica 18 point font.
Design and stimulus materials
The stimuli set consisted of words classed as carrying
negative emotional valence (e.g. venom) and words classed
as being neutral in emotional valence (e.g. table). The level
of emotional valence for most words was derived from
affective norms (Toglia and Battig, 1978; Bradley and Lang,
1999). Due to the limited number of affective word norms
available to enable a sufficient number of word pairings for
the present study, it was necessary to collect an additional
set of affective norms. Affective word norms were collected
from 66 Australian university students with an average age
of 22.3 (SD ¼ 4.3) years and with an average of 16
(SD ¼ 1.6) years of formal education. The university
students rated 118 words according to the normative data
collection methodology outlined by Bradley and Lang’s
(1999) Affective Norms for English Words (ANEW), where
words were rated on a scale from 1 to 9 for emotional
valence (1 representing the most emotionally negative and 9
representing the most emotionally positive) and emotional
arousal (1 representing the least emotionally arousing and
9 representing the most emotionally arousing). As a
reliability measure, the 118 words that were rated in the
present study consisted of 52 words that were previously
rated in Bradley and Lang’s (1999) normative study, and
the remaining 66 words were previously un-normed.
1400
Brain (2007), 130, 1395^1407
Pearson correlations performed on the 52 mean emotional
valence and arousal ratings obtained in the present study
and the respective mean emotional valence and arousal
ratings obtained in Bradley and Lang’s (1999) study,
indicated that the emotional valence and arousal ratings
were highly correlated across the two ratings (r ¼ 0.927,
P50.001; r ¼ 0.496, P50.001, respectively).
The ratings obtained from the present normative study
and the ANEW ratings were then classed as negative or
neutrally valenced words as per Crosson et al.’s (2002)
study. Words with valence ratings less than 3.5 (M ¼ 2.351,
SD ¼ 0.618) and arousal ratings greater than 5 (M ¼ 6.472,
SD ¼ 0.795) were classed as carrying negative emotional
connotation. Words were considered to not carry any
emotional connotation (i.e. neutral words) if they were
rated on average between 4 and 6 for emotional valence
(M ¼ 5.144, SD ¼ 0.391) and less than 5 for arousal
(M ¼ 4.227, SD ¼ 0.360). Finally, additional valence norms
were derived from Toglia and Battig (1978) where words
were ranked on a seven point scale with respect to
emotional pleasantness (1 representing the most unpleasant
and 7 representing the most pleasant). Negative valenced
stimuli included in the present study had an average
ranking of 2.478 (SD ¼ 0.298) and neutral valenced stimuli
had an average ranking of 4.191 (SD ¼ 0.457).
The stimuli set consisted of a total of 320 prime-target
pairs which were divided into two separate lists (List A and
List B). Lists A and B real-word targets were matched for
imageability
[t(140) ¼ 0.74,
P ¼ 0.458],
frequency
[t(156) ¼ 0.526, P ¼ 0.600] and number of letters
[t(158) ¼ 0.21, P ¼ 0.832]. Normative data for imageability and frequency were obtained from The MRC
Psycholinguistic Database (The University of Western
Australia, 1987). In addition, RT and accuracy data was
obtained from The English Lexicon Project (Balota et al.,
2002) and Lists A and B real-word targets were matched for
RT
[t(158) ¼ 0.2,
P ¼ 0.790]
and
accuracy
[t(158) ¼ 0.374, P ¼ 0.709]. The lists were presented in
two SOA conditions in a counterbalanced manner, in that
nine participants with Parkinson’s disease and nine control
participants were presented with List A for the short SOA
and List B for the long SOA and nine participants with
Parkinson’s disease and 10 control participants were
presented with List A for the long SOA and List B for
the short SOA.
Each list consisted of 80 real-word target pairs and 80
non-word target pairs. Of the 80 real-word pairs, 40 were
semantically related and 40 were semantically unrelated
word pairs as determined by word association norms (Kiss
et al., 1973; Moss and Older, 1996). Using the relatedness
values obtained, List A and List B related word pairs were
matched for degree of relatedness [Moss and Older,
t(45) ¼ 0.688, P ¼ 0.495; Kiss et al., t(73) ¼ 0.504,
P ¼ 0.616].
Within each related and unrelated condition, half of the
real-word pairs were affectively congruent (i.e. the prime
J. E. Castner et al.
Table 2 Examples of real-word target pairings
Semantic
relatedness
Related
Related
Unrelated
Unrelated
Related
Related
Unrelated
Unrelated
Affective
congruency
Congruent
Congruent
Incongruent
Incongruent
Incongruent
Incongruent
Congruent
Congruent
Target
valence
Negative
Neutral
Negative
Neutral
Negative
Neutral
Negative
Neutral
Example
Prime
Target
Selfish
Hand
Fable
Homicide
Donor
Danger
Cemetery
Locker
Greedy
Finger
Angry
Snail
Blood
Sign
Bankrupt
Hamster
and targets were both of negative valence or the primes and
targets were both affectively neutral) and the remaining half
consisted of affectively incongruent word pairs (i.e. one
neutral and one negative word). See Table 2 for examples of
real-word target pairings.
Within Lists A and B, all related and unrelated targets
were matched for imageability, frequency, number of
letters, RT and accuracy (P40.1). All congruent and
incongruent targets were also matched for imageability,
frequency, number of letters, RT and accuracy (P40.1).
The 80 non-word target pairs comprised either a unique
negative or neutrally valenced prime word, selected from
the same norms and criteria described earlier, and were
paired with orthographically legal and pronounceable nonword targets which were not homophonic with real English
words. Non-words were selected from the ARC Nonword
Database (Rastle et al., 2002) and were matched with the
real-word targets for number of letters [t(318) ¼ 0.040,
P ¼ 0.968].
Within each list, the 160 real word and non-word pairs
were pseudo-randomly divided into four different blocks
consisting of 20 real-word pairs and 20 non-word target
pairs. All items within each block were pseudo-randomly
organized, so that neither a word or non-word condition
was presented more than three times consecutively. For
each participant on each occasion, the four blocks were
presented in a random order. Two non-experimental
buffers were presented prior to the commencement of
each block.
Procedure
Participants were advised of the process of making a lexical
decision using the yes/no method (i.e. respond by pressing
one button if the target was a real word and pressing an
alternate button if the target was a non-word). Participants
were also advised to respond as quickly as possible, without
compromising accuracy.
For the short SOA condition, a fixation cross appeared
centrally for 750 ms followed by the prime for 150 ms and
an inter-stimulus interval of 50 ms, making the short SOA
equal to 200 ms. The target was then presented for 3000 ms
Priming in Parkinson’s disease
Brain (2007), 130, 1395^1407
or until a response was made. The following trial
commenced 1500 ms after the completion of the previous
trial. The long SOA condition consisted of the fixation
cross also appearing centrally for 750 ms; however, the
prime was presented for 750 ms and with an inter-stimulus
interval of 250 ms. The target was also presented for
3000 ms or until the participant responded. Therefore, the
long SOA condition consisted of an SOA of 1000 ms.
Before the commencement of each experiment, participants were shown a demonstration of the task containing
two items. Following this, they participated in a practice
session containing 20 unique items. Each experiment took
25 min to complete.
Results
Statistical analyses were performed on RT data pertaining
to real-word responses only. RT data was excluded from
analyses if the participant’s response resulted in an error
(3.29% of data for participants with Parkinson’s disease and
1.41% of data for control participants), if RTs were 5100 or
42000 ms (1.7% of data), and if RTs deviated more
than two standard deviations from the condition means
for each participant (4.7% of data). Statistical analyses
were not conducted on the error data due to the overall
low number of errors in the present study. Visual
inspection of the RT distribution revealed that the
data was positively skewed and violation of distribution
normality was confirmed with the Kolmogorov–Smirnov
Test of Normality (P50.001). Therefore, statistical analyses
were conducted on the logarithmic transformation of
the RT data. For ease of interpretation, raw RT data will
be presented for descriptive purposes only.
Affective and semantic priming on negative
and neutral targets
Linear Mixed Model (LMM) analyses were conducted on
RT data with session (Parkinson’s disease on versus off
stimulation, and controls) as within and between-subjects
fixed factors. Affective congruency (congruent and
incongruent), semantic relatedness (related and unrelated),
target valence (negative and neutral) and SOA (short and
1401
long) were included as within-subjects fixed factors. RT
subject variations, in addition to subject variation across
sessions for participants with Parkinson’s disease, were
treated as random factors. Refer to Tables 3 and 4 for mean
(SD) RTs for Parkinson’s disease and control participants.
Initial analyses revealed a significant main effect
for session [F(2,25.057) ¼ 4.660, P50.05], and highly
significant main effects for affective congruency
[F(1,7967.072) ¼ 14.722, P50.001], semantic relatedness
[F(1,7967.116) ¼ 163.169,
P50.001],
target
valence
[F(1,7967.108) ¼ 19.051,
P50.001]
and
SOA
[F(1, 7967.249) ¼ 17.702, P50.001].
An interaction between session and target valence
reached significance [F(2,7967.148) ¼ 4.438, P50.05].
Planned contrasts on the parameter estimates revealed
that RTs for control participants were significantly slower
on negatively valenced targets (M ¼ 710 ms, SD ¼ 163 ms)
compared with RTs on neutral targets (M ¼ 695 ms,
SD ¼ 156 ms; P50.005). Similarly, participants with
Parkinson’s disease in the on stimulation condition were
also significantly slower on negative targets (M ¼ 854 ms,
SD ¼ 278 ms) in comparison to neutral targets (M ¼ 830 ms,
SD ¼ 268 ms; P50.001); however this was not the case
during off stimulation where an average 5 ms RT advantage
for neutral targets was not statistically significant (P40.05).
Therefore, regardless of the semantic relatedness or affective
congruency of word pairs, RTs in the Parkinson’s disease
on stimulation condition and control participant RTs were
significantly slower on negatively valenced targets compared
with neutrally valenced targets.
A two-way interaction between affective congruency
and
semantic
relatedness
was
significant
[F(1,7967.089) ¼ 13.135, P50.001]. A three-way interaction
between affective congruency, semantic relatedness and
SOA was highly significant [F(1,7967.154) ¼ 14.102,
P50.001], and a four-way interaction between session,
affective congruency, semantic relatedness and SOA was
also significant [F(2,7967.187) ¼ 7.402, P ¼ 0.001]. Affective
priming was therefore analysed separately on semantically
unrelated word pairs, and semantic priming on affectively
incongruent words pairs, to further isolate the effects
Table 3 Mean (SD) RTs for participants with Parkinson’s disease
On stimulation
Short SOA
Related and congruent RTs (ms)
Related and incongruent RTs (ms)
Unrelated and congruent RTs (ms)
Unrelated and incongruent RTs (ms)
Off stimulation
Long SOA
Short SOA
Long SOA
Negative
Neutral
Negative
Neutral
Negative
Neutral
Negative
Neutral
860 (285)
828 (258)
845 (229)
898 (257)
824 (282)
813 (246)
824 (269)
865 (251)
817 (284)
862 (290)
851 (302)
895 (307)
810 (285)
804 (266)
848 (300)
852 (249)
843 (270)
809 (266)
859 (285)
904 (276)
854 (294)
827 (287)
855 (292)
896 (306)
818 (251)
878 (307)
874 (264)
894 (262)
827 (263)
864 (300)
915 (330)
865 (235)
SOA ¼ stimulus onset asynchrony; RT ¼reaction time; negative and neutral refers to the valence of the target stimuli; related and
unrelated refers to the semantic relatedness of word pairs and congruent and incongruent refers to the affective congruency of word pairs.
1402
Brain (2007), 130, 1395^1407
J. E. Castner et al.
Table 4 Mean (SD) RTs for control participants
Short SOA
Related and congruent RTs (ms)
Related and incongruent RTs (ms)
Unrelated and congruent RTs (ms)
Unrelated and incongruent RTs (ms)
Long SOA
Negative
Neutral
Negative
Neutral
714 (166)
717 (175)
750 (209)
761 (180)
688 (140)
711 (176)
743 (186)
757 (176)
653 (109)
662 (132)
702 (132)
733 (150)
656 (138)
643 (127)
680 (124)
684 (131)
SOA ¼ stimulus onset asynchrony; RT ¼reaction time; negative and neutral refers to the valence of the target stimuli; related and
unrelated refers to the semantic relatedness of word pairs and congruent and incongruent refers to the affective congruency of word pairs.
Fig. 1 Affective priming as a function of SOA condition. P50.001.
of affective priming and semantic priming as a function of
session and SOA.
Affective priming on semantically unrelated
word pairs
To gain an understanding of affective priming effects as a
function of session and SOA, LMM analysis was conducted
on semantically unrelated targets only. The main effects for
session, congruency and SOA were consistent with the
preliminary analysis.
In relation to affective priming effects, there was a
significant interaction between affective congruency and
SOA [F(1,3921.449) ¼ 6.549, P50.05]. Mean RTs for
affectively congruent and incongruent word pairs at each
SOA are presented in Fig. 1. Planned contrasts on the
parameter estimates revealed that affective priming effects
were evident at the short SOA condition (P50.001), but
not the long SOA condition (P40.05). There was no
interaction between session, affective congruency and SOA
[F(2,3921.032) ¼ 2.028, P40.05], suggesting that the
presence of affective priming at short SOAs and the lack
of affective priming at long SOAs was consistent for control
participants and participants with Parkinson’s disease in on
and off stimulation conditions.
Semantic priming on affectively incongruent
word pairs
In order to isolate semantic priming effects, LMM
analyses were conducted on the RTs for affectively
incongruent word pairs. Again, the main effects of session,
semantic relatedness and SOA were consistent with the
initial analysis.
With respect to semantic priming, there was a significant interaction between semantic relatedness and
SOA [F(1,3945.475) ¼ 5.116, P50.05]. Interestingly, the
interaction between session, semantic relatedness and SOA
also reached statistical significance [F(2,3945.688) ¼ 4.172,
P50.05], suggesting that the magnitude of semantic
priming at the short and long SOAs differed as a function
of session. The means (SD) for this interaction are
presented in Table 5.
Planned contrasts on the parameter estimates were
performed to isolate semantic priming effects for each
session at short and long SOAs. Significant semantic
priming was evident for all testing sessions at the short
SOA (P50.001). In contrast, at the long SOA, significant
priming was only found amongst controls and participants
with Parkinson’s disease in the on stimulation condition
(P50.001). The mean priming effect of 6 ms failed to reach
Priming in Parkinson’s disease
Brain (2007), 130, 1395^1407
1403
Table 5 Mean (SD) semantic priming effects in Parkinson’s disease and control participants as a function of SOA
Parkinson’s disease participants
On stimulation
Semantically related RTs (ms)
Semantically unrelated RTs (ms)
Semantic priming effect (ms)
Control participants
Off stimulation
Short SOA
Long SOA
Short SOA
Long SOA
Short SOA
Long SOA
818 (250)
876 (253)
58
824 (276)
867 (271)
43
821 (280)
898 (296)
77
869 (302)
875 (245)
6
713 (175)
759 (177)
46
650 (129)
701 (139)
51
P50.001. SOA ¼ stimulus onset asynchrony.
significance for the participants with Parkinson’s disease in
the off stimulation condition (P40.05).
Discussion
The present study investigated automatic and controlled
processing of lexical-semantic representations via a semantic priming paradigm, the automatic activation of emotional evaluations via an affective priming paradigm, and
the incidental processing of stimuli representing differing
emotional connotations. The following discussion will
address each of these processes in relation to theoretical
models of semantic priming, affective priming and
the processing of emotional words and the potential
involvement of basal ganglia-thalamocortical circuits in
these processes.
Semantic priming
In the present study, semantic priming was evident at
the short SOA for both control and participants with
Parkinson’s disease in on and off stimulation conditions.
The purpose of investigating semantic priming at short
SOAs is to study access to lexical-semantic representations
which is due to automatic spread of activation, rather than
being confounded by pre-lexical or post-lexical attentional
mechanisms (Neely, 1991). The comparable semantic
priming effects at the short SOA for both control
and participants with Parkinson’s disease (on stimulation
and off stimulation) suggests that the automatic
spread of activation of lexical-semantic representations is
unimpaired in Parkinson’s disease and is not influenced
by the modulation of basal ganglia-thalamocortical
circuits via stimulation of the STN. Unimpaired automatic
semantic priming in the present study is consistent
with Copland’s (2003) study where comparable priming
of ambiguous meanings was found in participants
with Parkinson’s disease and controls when a short
inter-stimulus interval was employed. Also, a lack of
interaction between priming effects in control and
participants with Parkinson’s disease at short SOAs has
also been presented in studies utilising multi-priming
paradigms (Angwin et al., 2003, 2005), further supporting
the notion that automatic semantic priming is unaffected
by Parkinson’s disease.
Whilst similar semantic priming effects at the short SOA
were evident in Parkinson’s disease (both on and off
stimulation) and control participants in the present study,
this was not the case at the long SOA. Specifically,
participants with Parkinson’s disease in the off stimulation
condition displayed an absence of semantic priming at the
long SOA in contrast to the presence of semantic
priming in the Parkinson’s disease on stimulation condition
and control participants. The absence of semantic
priming at long SOAs in the off stimulation condition
is consistent with Angwin et al.’s (2005) study which
found no significant priming effects for participants with
Parkinson’s disease at a long SOA (1200 ms) compared with
the presence of semantic priming effects in control
participants.
Semantic priming effects at long SOAs are more likely to
elicit controlled semantic priming, which are influenced by
attentional mechanisms (Neely, 1991). For example,
semantic priming at long SOAs can be influenced by
facilitatory spread of activation for related word pairs in
comparison to neutral conditions (e.g. if a prime display
consists of a series of x’s or the word blank), or when
attentional mechanisms are in play, inhibitory processing of
unrelated word pairs compared with neutral conditions can
occur. Also, the investigation of semantic priming effects at
various SOAs allows for the charting of the time course of
activation within semantic memory. For example, the Gain/
Decay hypothesis proposed by Milberg and colleagues, to
explain patterns of semantic priming in Alzheimer’s disease,
suggests that pathological semantic priming can be
influenced by the speed of initial activation, the level of
activation and the subsequent decay of activations within
semantic memory (Milberg et al., 1999).
The lack of semantic priming at the long SOA for
participants with Parkinson’s disease in the off stimulation
condition could reflect reduced attentional mechanisms
(thereby reducing inhibitory effects for the semantically
unrelated condition) and/or decay in the activation of
lexical-semantic representations of prime words with time
(i.e. degradation of spread of activation effects) resulting in
reduced facilitation on related word pairs. Reduced
inhibition on semantically unrelated word pairs and
1404
Brain (2007), 130, 1395^1407
reduced facilitation on semantically related words pairs may
have resulted in reduced semantic priming effects in the
present study. However, determining the degree to which
facilitatory and inhibitory effects influence semantic priming in the present study would require the use of a neutral
prime condition to serve as a baseline for such effects.
Regardless of the facilitatory and inhibitory mechanisms
that are affected in participants with Parkinson’s disease at
the long SOA, STN stimulation in the present study
resulted in the restoration of semantic priming. The
dissociation between semantic priming effects at the long
SOA as a function of STN stimulation condition is
suggestive of a role of basal ganglia-thalamocortical circuits
in either the attentional mechanisms involved in semantic
priming or in the prolongation of facilitatory effects of
semantic priming at a long SOA.
The recent emergence of fMRI studies investigating the
regions of neural activation associated with semantic
priming provide an insight into potential regions within
the basal ganglia-thalamocortical circuitry associated with
semantic priming at long SOAs. For example, Rossell et al.
(2001) reported activity within the striatum associated with
semantic priming at long SOAs. Furthermore, whilst ACC
activity was associated with semantic priming at short and
long SOAs, the region activated was generally more caudal
in the long SOA condition. More recently, Gold et al.
(2006) found greater activation bilaterally in the ACC
during semantic priming at long SOAs compared with
semantic priming at short SOAs. Generally, ACC activity
associated with semantic priming at long SOAs has been
attributed to inhibitory mechanisms in semantic priming
(Gold et al., 2006).
To summarize, the semantic priming effects observed
in the present study suggest that automatic access to
lexical-semantic representations is unimpaired in
Parkinson’s disease, yet Parkinson’s disease is associated
with aberrant semantic priming at long SOAs either due to
reduced attentional mechanisms or early decay of activation
of lexical-semantic representations. Considering that controlled semantic priming is modulated by STN stimulation
implies a link between controlled semantic priming and
impairment of basal ganglia-thalamocortical circuits.
Furthermore, a link between semantic priming at long
SOAs and activity in the ACC (Gold et al., 2006) suggests
that STN stimulation may act to re-establish semantic
priming
via
ACC
basal
ganglia-thalamocortical
neuromodulation.
Affective priming
When isolating affective priming effects in the present
study, significant priming effects were only evident at the
short SOA for both Parkinson’s disease (on and off
stimulation conditions) and control participants. The
lack of affective priming at long SOAs is consistent
J. E. Castner et al.
with the majority of affective priming studies that have
suggested that affective priming only occurs at short SOAs
(Fazio et al., 1986; Hermans et al., 1994, 2001; Klauer et al.,
1997; De Houwer et al., 1998; Fazio, 2001; Klauer and
Musch, 2003). The presence of affective priming at short
SOAs and the lack of affective priming at long SOAs is a
reflection of the automatic activation of emotional evaluations, which is likely to occur at the pre-lexical stage of
processing. In the present study, affective priming effects at
either SOA were comparable for participants with
Parkinson’s disease and controls. Thus, the automatic
activation of emotional evaluations appears to be unimpaired in Parkinson’s disease. Furthermore, modulation
of basal ganglia-thalamocortical circuits via STN stimulation does not result in alterations to affective priming
effects in Parkinson’s disease, indicating that these circuits
are not likely to be involved in the automatic activation
of emotional evaluations. This finding is supported by
the observation that event-related brain potentials of
early stage emotional processing of emotionally valenced
pictures were no different for participants with Parkinson’s
disease compared with controls (Wieser et al., 2006).
The present study also investigated the incidental
processing of negatively and neutrally valenced written
word stimuli. In utilizing a lexical decision task
which consisted of negative and neutral targets, a direct
emotional valence judgement was not required; however,
differential RTs to negative versus neutral stimuli whilst
making a lexical decision task provides a measure of
the implicit evaluation of negatively and neutrally valenced
stimuli. This will be discussed in more detail later.
Valence processing
In the present study, responses to negative targets were
significantly slower in comparison to neutral targets for
control participants and participants with Parkinson’s
disease in the on stimulation condition, regardless of
affective congruency or semantic relatedness of word pairs.
Rossell and Nobre (2004) similarly found that lexical
decision times to negative stimuli are typically delayed
compared with both neutral and positive targets in healthy
controls. Also, other studies requiring decisions on
negatively versus neutrally or positively valenced stimuli
have also shown delays in the processing of negative stimuli
(e.g. White, 1996; Dahl, 2001). Functional neuroimaging
studies have shown greater activation in the right ACC,
bilateral middle temporal gyrus, left posterior gyrus and the
amygdala bilaterally, when healthy controls are making
lexical decisions on negatively valenced stimuli compared
with neutral stimuli (Nakic et al., 2006). Kuchinke et al.
(2005), however, only found greater activation in the right
inferior frontal gyrus (including the dorsolateral prefrontal
cortex) during the processing of negative versus
neutral stimuli. Left ACC activity was associated with the
Priming in Parkinson’s disease
processing of neutral stimuli compared with less activity
with negative stimuli.
Interestingly, differential processing of negative versus
neutral targets was not evident in participants with
Parkinson’s disease without STN stimulation in the present
study. Whilst participants with Parkinson’s disease in the
off stimulation condition did not display impairment in the
automatic activation of emotional evaluations via affective
priming effects, it would appear that abnormal processing
of negative stimuli is evident. Specifically, a lack of
dissociation between negative and neutrally valenced stimuli
in the off stimulation condition may be indicative of a
deficit in the evaluation of negative valence via lexical
decision in the present study. An alternative interpretation
of these findings is that the different pattern of negative
versus neutral RTs on and off stimulation may reflect a
slowing of neutral RTs in the off stimulation condition,
compared with the on condition. However, such direct
comparisons of the neutral RTs (between on and off
stimulation conditions) are not as valid as within condition
contrasts, due to increased variability between stimulation
testing sessions.
Limited research has been conducted with respect to
the processing of negative emotionally valenced words
in participants with Parkinson’s disease; however, as
discussed previously, impairments have been reported in
the processing of negative emotions in facial expression
(Jacobs et al., 1995; Yip et al., 2003), the accurate
processing of affective prosody (Pell, 1996; Pell and
Leonard, 2003), affectively loaded sentences (Benke et al.,
1998) and emotional discourse (Crucian et al., 2001),
suggestive of an impairment in emotional processing
in Parkinson’s disease. This is the first study to extend
these results to include a deficit in the incidental processing
of negatively versus neutrally valenced words in Parkinson’s
disease.
Whilst the processing of negatively versus neutrally
valenced stimuli appears impaired in participants with
Parkinson’s disease when not receiving stimulation of the
STN, the present data suggests that when STN stimulation
is applied, the processing of such stimuli is comparable to
control participants. That is, consistent with the results
from control participants, when participants with
Parkinson’s disease were assessed in the STN stimulation
condition, delayed RTs when incidentally processing
negative stimuli compared with neutral stimuli were
evident. Therefore, STN stimulation may act to restore
activity within the basal ganglia-thalamocortical circuits
to allow lexical detection of word stimuli representing
negative emotional connotation. These results are
consistent with other reported improvements in affective
processing with STN stimulation (Schneider et al., 2003).
The results of the present study, however, are in
contrast with studies investigating facial discrimination deficits in Parkinson’s disease with STN stimulation
(Dujardin et al., 2004; Schroeder et al., 2004; Biseul et al.,
Brain (2007), 130, 1395^1407
1405
2005). It should be noted that these facial discrimination deficits associated with STN stimulation have been
isolated to the discrimination of angry emotions (Schroeder
et al., 2004), or have explored the effects of STN stimulation on emotional processing in a preoperative versus
on stimulation comparison, rather than a direct measure of stimulation via an on stimulation verus
off stimulation condition (Dujardin et al., 2004;
Biseul et al., 2005).
In conclusion, the present study indicates that automatic
lexical-semantic and evaluative processes are unimpaired in
Parkinson’s disease and the neuromodulation of basal
ganglia-thalamocortical circuits does not influence these
processes. In contrast, participants with Parkinson’s disease
showed a pattern of aberrant controlled lexical-semantic
processing as evidenced by a lack of semantic priming
effects at a long SOA condition for participants with
Parkinson’s disease in the off stimulation condition.
Controlled semantic priming was restored, however, when
the participants with Parkinson’s disease were receiving
stimulation of the STN, suggesting that STN stimulation
modulates basal ganglia-thalamocortical circuits that are
involved in such processes. Finally, the lack of dissociation
between RTs on negative versus neutral targets in
participants with Parkinson’s disease in the off stimulation
condition and presence of dissociation in participants with
Parkinson’s disease in the on stimulation condition along
with controls, supports aberrant incidental evaluation of
negatively versus neutrally valenced stimuli in Parkinson’s
disease and the subsequent restoration of such processing
when disrupted basal ganglia-thalamocortical circuits are
modulated by STN stimulation.
Acknowledgements
The authors would like to thank St. Andrew’s War
Memorial Hospital and The Wesley Hospital for their
support of this research, as well as Erin Smith, Kim Parton,
and Kieran Flanagan for their assistance with data
collection and technical support. This research was
supported by a University of Queensland Research
Development Grant. Joanna Castner is a recipient of a
Smart State PhD Scholarship awarded by the Queensland
Government Department of State Development, Trade,
and Innovation. David Copland is supported by an
ARC Research Fellowship.
References
Alegret M, Junque C, Valldeoriola F, Vendrell P, Pilleri M, Rumia J, et al.
Effects of bilateral subthalamic stimulation on cognitive function in
Parkinson disease. Arch Neurol 2001; 58: 1223–7.
Angwin AJ, Chenery HJ, Copland DA, Murdoch BE, Silburn PA.
Summation of semantic priming effects in Parkinson’s disease and
healthy individuals. Brain Lang 2003; 87: 96–7.
Angwin AJ, Chenery HJ, Copland DA, Murdoch BE, Silburn PA.
Summation of semantic priming and complex sentence comprehension
in Parkinson’s disease. Cogn Brain Res 2005; 25: 78–89.
1406
Brain (2007), 130, 1395^1407
Arnott WL, Chenery HJ, Murdoch BE, Silburn PA. Semantic priming
in Parkinson’s disease: evidence for delayed spreading activation.
J Clin Exp Neuropsychol 2001; 23: 502–19.
Balota DA, Cortese MJ, Hutchison KA, Neely JH, Nelson D, Simpson GB,
et al. The English lexicon project: a web-based repository of
descriptive and behavioral measures for 40,481 English words and
nonwords. http://elexicon.wustl.edu/, Washington University, 2002.
Baron MS, Wichmann T, Ma D, DeLong MR. Effects of transient focal
inactivation of the basal ganglia in parkinsonian primates. J Neurosci
2002; 22: 592–9.
Benke T, Bosch S, Andree B. A study of emotional processing in
Parkinson’s disease. Brain Cogn 1998; 38: 36–52.
Biseul I, Sauleau P, Haegelen C, Trebon P, Drapier D, Raoul S, et al. Fear
recognition is impaired by subthalamic nucleus stimulation in
Parkinson’s disease. Neuropsychologia 2005; 43: 1054–9.
Borod JC, Welkowitz J, Alpert M, Brozgold AZ, Martin C, Peselow E, et al.
Parameters of emotional processing in neuropsychiatric disorders:
conceptual issues and a battery of tests. J Commun Disord 1990; 23:
247–71.
Bradley MM, Lang PJ. Affective norms for English words (ANEW):
stimuli, instruction manual and affective ratings. Technical report C-1.
Gainesville, FL: The Center for Research in Psychophysiology, University
of Florida, 1999.
Burchiel KJ, Anderson VC, Favre J, Hammerstad JP. Comparison of
pallidal and subthalamic nucleus deep brain stimulation for advanced
Parkinson’s disease: results of a randomized, blinded pilot study.
Neurosurgery 1999; 45: 1375–82.
Copland D. The basal ganglia and semantic engagement: potential insights
form semantic priming in individuals with subcortical vascular lesions,
Parkinson’s disease, and cortical lesions. J Int Neuropsychol Soc 2003; 9:
1041–52.
Crosson B, Cato MA, Sadek JR, Gokcay D, Bauer RM, Fischler IS, et al.
Semantic monitoring of words with emotional connotation during
fMRI: contribution of anterior left frontal cortex. J Int Neuropsychol
Soc 2002; 8: 607–22.
Crucian GP, Huang L, Barrett AM, Schwartz RL, Cibula JE, Anderson JM,
et al. Emotional conversations in Parkinson’s disease. Neurology 2001;
56: 159–65.
Dahl M. Asymmetries in the processing of emotionally valenced words.
Scand J Psychol 2001; 42: 97–104.
Daniele A, Albanese A, Contarino MF, Zinzi P, Barbier A, Gasparini F,
et al. Cognitive and behavioural effects of chronic stimulation of
the subthalamic nucleus in patients with Parkinson’s disease.
J Neurol Neurosurg Psychiatry 2003; 74: 175–82.
De Gaspari D, Siri C, Di Gioia M, Antonini A, Isella V, Pizzolato A, et al.
Clinical correlates and cognitive underpinnings of verbal fluency
impairment after chronic subthalamic stimulation in Parkinson’s
disease. Parkinsonism Relat Disord 2006; 12: 289–95.
De Houwer J, Hermans D, Eelen P. Affective and identity priming with
episodically associated stimuli. Cogn Emotion 1998; 12: 145–69.
Dujardin K, Blairy S, Defebvre L, Krystkowiak P, Hess U, Blond S, et al.
Subthalamic nucleus stimulation induces deficits in decoding emotional
facial expressions in Parkinson’s disease. J Neurol Neurosurg Psychiatry
2004; 75: 202–8.
Dujardin K, Defebvre L, Krystkowiak P, Blond S, Destee A. Influence of
chronic bilateral stimulation of the subthalamic nucleus on cognitive
function in Parkinson’s disease. J Neurol 2001; 248: 603–11.
Fazio RH. On the automatic activation of associated evaluations: an
overview. Cogn Emotion 2001; 15: 115–41.
Fazio RH, Sanbonmatsu DM, Powell MC, Kardes FR. On the automatic
activation of attitudes. J Pers Soc Psychol 1986; 50: 229–38.
Funkiewiez A, Ardouin C, Caputo E, Krack P, Fraix V, Klinger H, et al.
Long term effects of bilateral subthalamic nucleus stimulation
on cognitive function, mood, and behaviour in Parkinson’s disease.
J Neurol Neurosurg Psychiatry 2004; 75: 834–9.
J. E. Castner et al.
Funkiewiez A, Ardouin C, Krack P, Fraix V, Van Blercom N, Xie J, et al.
Acute psychotropic effects of bilateral subthalamic nucleus stimulation
and levodopa in Parkinson’s disease. Mov Disord 2003; 18: 524–30.
Geday J, Ostergaard K, Gjedde A. Stimulation of subthalamic nucleus
inhibits emotional activation of fusiform gyrus. Neuroimage 2006; 33:
706–14.
Gold BT, Balota DA, Jones SJ, Powell DK, Smith CD, Andersen AH.
Dissociation of automatic and strategic lexical-semantics: functional
magnetic resonance imaging evidence for differing roles of multiple
frontotemporal regions. J Neurosci 2006; 26: 6523–32.
Hamani C, Saint-Cyr JA, Fraser J, Kaplitt M, Lozano AM. The subthalamic
nucleus in the context of movement disorders. Brain 2004; 127: 4–20.
Hermans D, De Houwer J, Eelen P. The affective priming effect: automatic
activation of evaluative information in memory. Cogn Emotion 1994; 8:
515–33.
Hermans D, De Houwer J, Eelen P. A time course analysis of the affective
priming effect. Cogn Emotion 2001; 15: 143–65.
Hermans D, Smeesters D, De Houwer J, Eelen P. Affective priming for
associatively unrelated primes and targets. Psychol Belg 2002; 42:
191–212.
Hill AB, Kemp-Wheeler SM. The influence of context on lexical decision
times for emotionally aversive words. Curr Psychol Res Rev 1989; 8:
219–27.
Jacobs DH, Shuren J, Bowers D, Heilman KM. Emotional facial imagery,
perception, and expression in Parkinson’s disease. Neurology 1995; 45:
1696–702.
Jahanshahi M, Ardouin CM, Brown RG, Rothwell JC, Obeso J, Albanese A,
et al. The impact of deep brain stimulation on executive function in
Parkinson’s disease. Brain 2000; 123: 1142–54.
Kiss GR, Armstrong C, Milroy R, Piper J. The Edinburgh Associative
Thesaurus [Online]. 1973. [cited 2004 Aug 1]; Available from: http://
www.eat.rl.ac.uk/.
Klauer KC, Musch J. Affective priming: findings and theories. In: Musch J,
Klauer KC, editors. The psychology of evaluation. Mahwah, NJ:
Lawrence Erlbaum Associates; 2003. p. 7–51.
Klauer KC, Roßnagel C, Musch J. List-context effects in evaluative
priming. J Exp Psychol Learn Mem Cogn 1997; 23: 246–55.
Kuchinke L, Jacobs AM, Grubich C, Vo ML, Conrad M, Herrmann M.
Incidental effects of emotional valence in single word processing: an
fMRI study. Neuroimage 2005; 28: 1022–32.
Kuhn AA, Hariz MI, Silberstein P, Tisch S, Kupsch A, Schneider G-H,
et al. Activation of the subthalamic region during emotional processing
in Parkinson disease. Neurology 2005; 65: 707–13.
Limousin P, Krack P, Pollak P, Benazzouz A, Ardouin C, Hoffmann D,
et al. Electrical stimulation of the subthalamic nucleus in advanced
Parkinson’s disease. N Engl J Med 1998; 339: 1105–11.
Lopiano L, Rizzone M, Bergamasco B, Tavella A, Torre E, Perozzo P, et al.
Deep brain stimulation of the subthalamic nucleus: clinical effectiveness
and safety. Neurology 2001; 56: 552–4.
McDonald C, Brown GG, Gorell JM. Impaired set-shifting
in Parkinson’s disease: new evidence from a lexical decision task.
J Clin Exp Neuropsychol 1996; 18: 793–809.
McIntyre CC, Mori S, Sherman DL, Thakor NV, Vitek JL. Electric field
and stimulating influence generated by deep brain stimulation of the
subthalamic nucleus. Clin Neurophysiol 2004; 115: 589–95.
Milberg W, McGlinchey-Berroth R, Duncan KM, Higgins JA. Alterations
in the dynamics of semantic activation in Alzheimer’s disease:
evidence for the Gain/Decay hypothesis of a disorder of semantic
memory. J Int Neuropsychol Soc 1999; 5: 641–58.
Moretti R, Torre P, Antonello RM, Capus L, Gioulis M, Marsala SZ, et al.
Cognitive changes following subthalamic nucleus stimulation in two
patients with Parkinson disease. Percept Mot Skills 2002; 95: 477–86.
Moro E, Scerrati M, Romito LM, Roselli R, Tonali P, Albanese A. Chronic
subthalamic nucleus stimulation reduces medication requirements in
Parkinson’s disease. Neurology 1999; 53: 85–90.
Moss H, Older L. Birkbeck word association norms. East Sussex, UK:
Psychology Press; 1996.
Priming in Parkinson’s disease
Nakic M, Smith BW, Busis S, Vythilingam M, Blair RJ. The impact of
affect and frequency on lexical decision: the role of the amygdala and
inferior frontal cortex. Neuroimage 2006; 31: 1752–61.
Neely JH. Semantic priming and retrieval from lexical memory: roles of
inhibitionless spreading activation and limited-capacity attention. J Exp
Psychol Gen 1977; 106: 226–54.
Neely JH. Semantic priming effects in visual word recognition: a selective
review of current findings and theories. In: Besner D, Humphreys GW,
editors. Basic processes in reading: visual word recognition. Hillsdale,
NJ: Lawrence Erlbaum Associates; 1991. p. 264–336.
Ortigue S, Michel CM, Murray MM, Mohr C, Carbonnel S, Landis T.
Electrical neuroimaging reveals early generator modulation to emotional
words. Neuroimage 2004; 21: 1242–51.
Parent A, Hazrati LN. Functional anatomy of the basal ganglia. II. The
place of subthalamic nucleus and external pallidum in basal ganglia
circuitry. Brain Res Brain Res Rev 1995; 20: 128–54.
Pell MD. On the receptive prosodic loss in Parkinson’s disease. Cortex
1996; 32: 693–704.
Pell MD, Leonard CL. Processing emotional tone from speech in
Parkinson’s disease: a role for the basal ganglia. Cogn Affect Behav
Neurosci 2003; 3: 275–88.
Perozzo P, Rizzone M, Bergamasco B, Castelli L, Lanotte M, Tavella A,
et al. Deep brain stimulation of the subthalamic nucleus in Parkinson’s
disease: comparison of pre- and postoperative neuropsychological
evaluation. J Neurol Sci 2001; 192: 9–15.
Piatt AL, Fields JA, Paolo AM, Koller WC, Troster AI. Lexical, semantic,
and action verbal fluency in Parkinson’s disease with and without
dementia. J Clin Exp Neuropsychol 1999; 21: 435–43.
Rastle K, Harrington J, Coltheart M. 358,534 nonwords: the ARC
Nonword Database. Q J Exp Psychol 2002; 55A: 1339–62.
Rossell SL, Bullmore ET, Williams SC, David AS. Brain activation during
automatic and controlled processing of semantic relations: a priming
experiment using lexical-decision. Neuropsychologia 2001; 39: 1167–76.
Rossell SL, Nobre AC. Semantic priming of different affective categories.
Emotion 2004; 4: 354–63.
Saint-Cyr JA, Trepanier LL, Kumar R, Lozano AM, Lang AE.
Neuropsychological consequences of chronic bilateral stimulation of
the subthalamic nucleus in Parkinson’s disease. Brain 2000; 123:
2091–108.
Brain (2007), 130, 1395^1407
1407
Schneider F, Habel U, Volkmann J, Regel S, Kornischka J, Sturm V, et al.
Deep brain stimulation of the subthalamic nucleus enhances emotional
processing in Parkinson disease. Arch Gen Psychiatry 2003; 60: 296–302.
Schroder C, Mobes J, Schutze M, Szymanowski F, Nager W, Bangert M,
et al. Perception of emotional speech in Parkinson’s Disease. Mov
Disord 2006; 21: 1774–8.
Schroeder U, Kuehler A, Hennenlotter A, Haslinger B, Tronnier VM,
Krause M, et al. Facial expression recognition and subthalamic nucleus
stimulation. J Neurol Neurosurg Psychiatry 2004; 75: 648–50.
Spicer KB, Brown GG, Gorell JM. Lexical decision in Parkinson disease:
lack of evidence for generalized bradyphrenia. J Clin Exp Neuropsychol
1994; 16: 457–71.
Suzuki A, Hoshino T, Shigemasu K, Kawamura M. Disgust-specific
impairment of facial expression recognition in Parkinson’s disease. Brain
2006; 129: 707–17.
Taylor JG, Fragopanagos NF. The interaction of attention and emotion.
Neural Netw 2005; 18: 353–69.
The University of Western Australia. MRC psycholinguistic database:
machine usable dictionary (Version 2.0) [Online]. 1987. [cited 2004
Aug 1], Available from: http://www.psy.uwa.edu.au/mrcdatabse/
uwa_mrc.htm.
Toglia MP, Battig WF. Handbook of semantic word norms. Hillsdale, NJ:
Lawrence Erlbaum Associates; 1978.
Whelan B-M, Murdoch BE, Theodoros DG, Hall B, Silburn P. Defining a
role for the subthalamic nucleus within operative theoretical models of
subcortical participation in language. J Neurol Neurosurg Psychiatry
2003; 74: 1543–50.
White M. Automatic affective appraisal of words. Cogn Emotion 1996; 10:
199–211.
Wieser MJ, Muhlberger A, Alpers GW, Macht M, Ellgring H, Pauli P.
Emotion processing in Parkinson’s disease: dissociation between early
neuronal processing and explicit ratings. Clin Neurophysiol 2006; 117:
94–102.
Yip JT, Lee TM, Ho SL, Tsang KL, Li LS. Emotion recognition in patients
with idiopathic Parkinson’s disease. Mov Disord 2003; 18: 1115–22.
Zec RF, Landreth ES, Fritz S, Grames E, Hasara A, Fraizer W, et al. A
comparison of phonemic, semantic, and alternating word fluency in
Parkinson’s disease. Arch Clin Neuropsychol 1999; 14: 255–64.