Neuropsychologia 48 (2010) 549557

Contents lists available at ScienceDirect

Neuropsychologia
journal homepage: www.elsevier.com/locate/neuropsychologia

Irritability in pre-clinical Huntingtons disease

Stefan Klppel a,b, , Cynthia M. Stonnington b,c , Predrag Petrovic b,d , Dean Mobbs b,e ,
Oliver Tscher f , David Craufurd g , Sarah J. Tabrizi h , Richard S.J. Frackowiak b,i,j
a

Department of Psychiatry and Psychotherapy, Section of Gerontopsychiatry, Section of Neuropsychiatry, Freiburg Brain Imaging, University Clinic Freiburg, Freiburg, Germany
Wellcome Trust Centre for Neuroimaging, Institute of Neurology, UCL, London, UK
c
Division of Adult Psychiatry, Mayo Clinic, Scottsdale, AZ, USA
d
Department of Clinical Neuroscience, Karolinska Hospital, Stockholm, Sweden
e
MRC Cognition and Brain Sciences Unit, Cambridge, UK
f
Department of Neurology, Neurozentrum, Freiburg Brain Imaging, University Clinic Freiburg, Freiburg, Germany
g
Academic Unit of Medical Genetics and Regional Genetic Service, St Marys Hospital, Manchester, UK
h
Department of Clinical Neurology, Institute of Neurology, UCL, London, UK
i
Service de neurologie, Centre hospitalier universitaire vaudois et Universit de Lausanne, Switzerland
j
Laboratory of Neuroimaging, IRCCS Santa Lucia, Roma, Italy
b

a r t i c l e

i n f o

Article history:
Received 7 March 2009
Received in revised form 9 August 2009
Accepted 20 October 2009
Available online 28 October 2009
Keywords:
Huntingtons disease
Irritability
fMRI
Amygdala
Orbitofrontal cortex

a b s t r a c t
Irritability, together with depression and anxiety, form three salient clinical features of pre-symptomatic
Huntingtons disease (HD). To date, the understanding of irritability in HD suffers from a paucity of
experimental data and is largely based on questionnaires or clinical anecdotes. Factor analysis suggests
that irritability is related to impulsivity and aggression and is likely to engage the same neuronal circuits
as these behaviours, including areas such as medial orbitofrontal cortex (OFC) and amygdala.
16 pre-symptomatic gene carriers (PSCs) and 15 of their companions were asked to indicate the larger of
two squares consecutively shown on a screen while undergoing functional magnetic resonance imaging
(fMRI). Despite correct identication of the larger square, participants were often told that they or their
partner had given the wrong answer. Size differences were subtle to make negative feedback credible
but detectable.
Although task performance, baseline irritability, and reported task-induced irritation were the same for
both groups, fMRI revealed distinct neuronal processing in those who will later develop HD. In controls
but not PSCs, task-induced irritation correlated positively with amygdala activation and negatively with
OFC activation. Repetitive negative feedback induced greater amygdala activations in controls than PSCs.
In addition, the inverse functional coupling between amygdala and OFC was signicantly weaker in PSCs
compared to controls.
Our results argue that normal emotion processing circuits are disrupted in PSCs via attenuated modulation of emotional status by external or internal indicators. At later stages, this dysfunction may increase
the risk for developing recognised, HD-associated, psychiatric symptoms such as irritability.
2009 Elsevier Ltd. All rights reserved.

1. Introduction
Huntingtons disease (HD) is an inherited neurodegenerative
disorder caused by an expanded number of triplet repeats of the
nucleotide bases cytosine, adenine and guanine (CAG) in the gene
encoding the protein huntingtin (HD Collaborative Research Group,
1993). Irritability, together with depression and anxiety, form a
triad of core psychiatric features of pre-symptomatic HD. Irrita-

Corresponding author at: Department of Psychiatry and Psychotherapy, University Clinic Freiburg, Hauptstrasse 5, 79104 Freiburg, Germany.
Tel.: +49 761 270 5234; fax: +49 761 270 5416.
E-mail address: [email protected] (S. Klppel).
0028-3932/$ see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.neuropsychologia.2009.10.016

tion is dened as a temporary psychological state characterised by

impatience, intolerance, and poorly controlled anger. It includes
elements of anger, aggression and reduced impulse control and
can occur independently of depression (Snaith, Constantopoulos,
Jardine, & McGufn, 1978).
To date, studies of irritability in patients with HD have relied on
questionnaires. A recent study found increased levels of irritability
in around 20% of pre-symptomatic gene carriers (PSCs) with less
than 10 years to estimated diagnostic onset who were unaware
of their gene status (Julien et al., 2007). Irritability causes great
distress to those close to HD patients and often determines if somebody can be managed in the community or needs to be admitted
to a nursing home (Folstein, Chase, Wahl, McDonnell, & Folstein,
1987; Hamilton et al., 2003; Wheelock et al., 2003).

550

S. Klppel et al. / Neuropsychologia 48 (2010) 549557

Factor analysis suggests that irritability in HD is related to

impulsivity and aggression (Craufurd, Thompson, & Snowden,
2001). The amygdala and medial orbitofrontal cortex (OFC) are key
circuits involved in impulsive aggression (Blair, 2007; Davidson,
Putnam, & Larson, 2000; Siever, 2008). This notion was suggested
in studies focusing on structural changes (Anderson, Bechara,
Damasio, Tranel, & Damasio, 1999; Tebartz van Elst, Woermann,
Lemieux, Thompson, & Trimble, 2000) and later conrmed by neuropsychological and functional imaging studies (Best, Williams,
& Coccaro, 2002; Coccaro, McCloskey, Fitzgerald, & Phan, 2007;
Dougherty et al., 2004). A growing body of evidence suggests an
inverse correlation between these two areas (Coccaro et al., 2007;
Dougherty et al., 2004; Urry et al., 2006) with the medial OFC exerting an inhibitory inuence on the amygdala, most likely through
direct anatomical connections (Rempel-Clower, 2007).
Neuropathological as well as imaging studies indicate an
involvement in HD of structures implicated in the regulation of
aggression, including the amygdala (Douaud et al., 2006; Mann,
Oliver, & Snowden, 1993; Pavese et al., 2003; Rosas et al., 2003)
and prefrontal cortex (Pavese et al., 2003; Rosas et al., 2002) but
there is little indication of a specic involvement of the OFC, at
least in earlier stages of the disease. Although the neuronal mechanism is not fully understood, a role for the serotonergic system
in impulsive aggression has been suggested (Coccaro & Kavoussi,
1997). The very limited available data indicate serotonin reuptake
inhibitors could be useful for the treatment of irritability in HD (De
Marchi, Daniele, & Ragone, 2001; Ranen, Lipsey, Treisman, & Ross,
1996) as an alternative to atypical neuroleptics (Paleacu, Anca, &
Giladi, 2002; Squitieri et al., 2001).
Our study therefore aimed to examine the emotional circuitry
associated with induced irritation with a focus on the amygdala.
Firstly, we intended to develop a task capable of reliably causing
irritation in participants and to study the neuronal correlates of
such a task with functional MRI (fMRI). Although PSC do not show
gross impairments we sought to design a task that can easily be
performed in an fMRI environment and does not require substantial dexterity or rely heavily on working memory performance.
Unmatched task performance could have made the interpretation
of differences in neuronal processing difcult. A number of studies
in PSC have shown altered neuronal processing in the absence of
differences in task performance (Kloppel et al., 2009; Reading et al.,
2004; Wolf, Vasic, Schonfeldt-Lecuona, Landwehrmeyer, & Ecker,
2007). In line with these studies, we expected changes in neuronal
processing associated with an experimental task that induces irritation, even in those PSCs without clinically increased irritability.
As outlined above, we assumed irritability in HD to be related to
impulsive aggression and inuenced by social interaction. Based
on previous work in subjects with intermittent explosive disorder (Coccaro et al., 2007), a disease characterized by impulsive
aggression, we hypothesized that our experimental task would
elicit increased amygdala activations in PSCs and a disruption of
amygdala-OFC coupling. Our clinical experience, and that reported
by others (Snowden et al., 2003), indicates that individuals with the
HD gene mutation get particularly irritated with close companions.
We therefore expected greater activations of the amygdala when
PSCs lost a round due to a mistake by their companion compared
to trials lost by mistakes from a computer.
2. Material and methods
2.1. Participants
16 PSCs and their close companions (14 partners, one close friend and one
same-generation relative) were included. PSCs were aware of their gene status. As
mentioned, we decided to include companions who are indirectly affected by the
disease who therefore do not represent the general population. We reasoned that the
inclusion of companions would illicit the strongest levels of irritability. An assumption which was based on our clinical experience as outlined in the introduction.

PSCs comprised a wide range of estimated years to clinical diagnosis, based on

age and the number of CAG repeats (Langbehn, Brinkman, Falush, Paulsen, & Hayden,
2004). With the exception of the one relative, companions did not undergo genetic
testing but none had a family history of HD or showed clinical signs. Other than
one control subject who had isolated seizures under the age of 12, there were no
medical co-morbidities. Two PSCs had previously taken antidepressant medication
for more than a year (one took citalopram 30 mg; the other could not recall the name
of the substance or dosage), but none were actively using antidepressants at study.
One PSC was taking 10 g creatine daily. No one else had a history of neurological or
psychiatric disorder and none had used centrally acting medication. A neurologist
experienced in HD examined all PSCs. One control failed to undergo scanning and
so was excluded from all analyses. The remaining subjects were matched for age,
gender, National Adult Reading Test (NART) score (Nelson & Willison, 1991) and
years in education (Table 1). The local Ethics Committee approved the study and all
participants gave written informed consent according to the Declaration of Helsinki.
2.2. Experimental procedure
The task was designed to study two factors. Firstly, we sought to study neuronal
processing with false allegations of ones own performance errors. We expected the
induced irritation to build up with repetition of these false allegations. The second
factor under study was the identity of the second player. Based on clinical experience, we expected a stronger emotional response when a round was lost by a
companions mistake than that made by a non-human (i.e., a computer). Both factors
were integrated into the same task in which two squares were shown consecutively
to participants who were asked to identify the larger (see Fig. 1 for details). A total
of 50 trials per run were performed.
In a pilot phase with healthy participants performing different versions of the
task we identied a difculty level that resulted in correct answers on most trials.
We found that it is much harder to correctly compare the size of two objects when
the rst object shown is larger. We therefore ended up with three different squares
for the study. If a smaller square was shown rst, it measured 2.8 cm edge length
and was followed by a larger square of 2.9 cm. A square of 3.1 cm edge length was
followed by the square of 2.8 cm edge length. Slight differences of the combined
distances between screen and mirror and mirror and eye were tolerated to ensure
comfortable positioning of subjects and an unblocked view of the screen. Based on
interviews with participants of the pilot study we determined that the task remained
credible when 28% of correct answers were followed by error as feedback. This number was found through trial and error in runs limited to exactly 50 trials. Written
instructions stated that the study examines how the brain responds when doing
tasks and getting feedback. We deliberately avoided any mention of irritability in
the study description, as that might have altered emotional responses and neural
processing.
2.2.1. Feedback on own performance
We hypothesised that subjects would experience irritation after being informed
they had made a mistake in a task they were sure they had performed correctly.
Furthermore, we expected the induced irritation to build up with repetition of false
allegations of erroneous performance. We therefore varied the percentage of conicting feedback responses over the course of the 50 trials as shown in Fig. 2 (left
panel). The variation was constrained by avoiding presentations with more than
two consecutive trials comprising incorrect negative feedback.
2.2.2. Identity of the second player
Subjects were led to believe that their computer was linked to their partners
computer who was playing simultaneously in another room and that both players
had to answer correctly to win a round. Three pounds were added to a players
account for every round won and deducted again when a round was lost. For comparison and to allow full balancing of the design, an additional second player in the
form of a computer was introduced which would try to simulate the behaviour of
a human and could therefore make mistakes. Playing with a computer was found

Table 1
Demographic and basic cognitive details of participants. Results are reported with
mean and standard deviation.

N
Gender (f/m)b
Age
CAG
Years to estimated onset
Years in education
UHDRS-motor
NART

PSC

Controls

p-Values

16
8/8
39.3 7.9
42.14 2.2a
16.8 8.8
14.9 2.9
2.3 1.5
104.2 10.2

15
8/7
40.4 90.4
NA
NA
15.6 3.1
NA
108.9 9.0

NA
0.8
0.73
NA
NA
0.57
NA
0.18

a
Exact CAG length missing from two subjects measured with different equipment.
b
Chi-Square test for gender.

S. Klppel et al. / Neuropsychologia 48 (2010) 549557

551

study by Chatterjee et al. (2005). In addition, they completed a questionnaire of

how they felt when receiving feedback. We assumed that each subject would have
a different understanding of irritation. We therefore asked them to rate three negative emotions (irritation, anger, tension) ranging from zero (emotion not felt) to
three (emotion strongly felt). Similarly, subjects were asked to report, using the
same scale, their positive emotions (happy and relieved) on receiving positive feedback. We tested for a positive correlation of negative emotions (composite score
of irritability, anger, and tension) induced by the task, with Snaiths score and the
irritability subscore (summation of inward and outward irritability sub-scores) to
identify a relation of the task to our concept of irritability. Similarly, since irritability
is closely connected to problems with impulse control (Craufurd et al., 2001) scoring
on the BIS-11 and task-induced negative emotions were also expected to be positively correlated. Signicance was tested with non-parametric tests when signicant
KolmogorovSmirnov tests indicated a violation of assumptions for parametric testing.
Before detailed debrieng, all participants took part in a semi-structured interview, one purpose of which was to verify that they had believed in the integrity of
the task, the link with the other player and the correctness of feedback.
2.3. MRI-scanning

Fig. 1. Overview of task. Subjects had to identify the larger of two squares shown
sequentially. This was followed by feedback on the correctness of the answer of the
rst and second player. A separate screen indicated if the round was won (when
both players answered correctly) or lost. Timing intervals between screens were
randomised (jittered) where indicated.

less emotionally involving in our pilot study which mirrors results of others (Rilling,
Sanfey, Aronson, Nystrom, & Cohen, 2004).
After a brief training period with correct feedback, each subject performed four
runs of 50 trials. Two runs (one with a partner and the other with a computer) were
performed in the MRI scanner. The other two were performed in a testing room
using a standard PC. Participants swapped rooms after two runs. The order of runs
and whether PSCs or controls started in the MRI scanner was randomised over the
study.
2.2.3. Ratings and questionnaires
After training, subjects were asked to indicate on a visual analogue scale (VAS),
ranging from 100 (indicating worst performance) to +100, how well they expected
they and their partner would do. Using a similar VAS after every two or three rounds,
subjects were asked how condent they were with their performance in the task.
After the experiment, subjects completed the Snaith irritability self-assessment
scale (Snaith et al., 1978), Beck Depression Inventory (BDI; Beck, Ward, Mendelson,
Mock, & Erbaugh, 1961), Barratts Impulsivity Scale (BIS-11; Barratt & Patton, 1983)
and Spielberger State-Trait Anxiety Inventory (STAI; Spielberger, 1988). As PSC
might be reluctant to admit their true level of irritability we adopted the approach
of Chatterjee, Anderson, Moskowitz, Hauser, and Marder (2005) by evaluating discrepancies between PSCs self rating and that of their companions using the Johns
Hopkins irritability questionnaire. Results between the two groups on the 14 item
questionnaire were compared using two-tailed paired t-tests. We refer to this questionnaire as the Johns Hopkins irritability questionnaire. All items are listen in the

We used an echo planar imaging sequence optimised for sensitivity in amygdala

and OFC (Deichmann, Gottfried, Hutton, & Turner, 2003) scanned on a 1.5 Tesla MRI
system (Sonata; Siemens, Erlangen, Germany) at a single centre. Scanning parameters included: repetition time 3600 ms; echo time 50 ms; eld of view 192 mm;
distance factor 50%; ip angle 90 . We used 40 slices with a thickness of 3 mm
angled at 30 . We sought to measure the individual distortions in OFC and amygdala due to the close proximity of air-lled cavities by acquiring eld maps (Hutton
et al., 2002). Total scanning time for the fMRI task was around 30 min per subject
depending on speed of responses. An additional T1-weighted sequence (Deichmann,
Schwarzbauer, & Turner, 2004) was acquired to exclude structural abnormalities
and to evaluate structural differences using voxel-based morphometry (Ashburner
& Friston, 2000) (see supplementary material).
2.4. Image data analysis
Pre-processing and statistical analysis of fMRI data were carried out using SPM5
software (www.l.ion.ucl.ac.uk/spm/). Before smoothing with an 8 mm Gaussian
kernel, volumes were realigned and spatially normalised to a standard echo planar
imaging template in Montreal Neurological Institute (MNI) space. Field maps were
included in the realignment step. A rst-level analysis based on the general linear
model (Friston, Frith, Turner, & Frackowiak, 1995) was performed for each subject.
A 128 s high-pass lter was applied. Task-related changes in fMRI signal were estimated at each voxel by modelling the onsets and length of each event type as a
separate regressor convolved with a haemodynamic response function. For our primary analysis, we used four regressors to model the onset of positive and negative
feedback, for the rst and second player respectively. They were modelled as miniblocks with a length of 35 s, depending on the onset of the next screen (see Fig. 1).
Since we were interested in the effect of repeated negative feedback, we entered a
separate parametric modulator to both regressors of negative feedback, coding how
often subjects had received negative feedback in the last three trials. As illustrated
in Fig. 2 the modulator thus contains values of either 100 when positive feedback
was received for all three previous trials 66 or 33 when only one trials of the last
three returned positive feedback. We did not aim to evenly distribute these percentages as including too many trials with false negative feedback would have made the
task less believable. All four feedback conditions had separate parametric modulators assigned to model linear effects of time (e.g., due to fatigue). Error rates were
low and for a subject not making a real mistake 36 correct trials were contrasted
with 14 trials in which false allegations of erroneous performance was reported. An
illustration of the design matrix is provided as supplementary material. As shown,
the parametric modulator coding past negative feedback is assigned to the nega-

Fig. 2. Left: The graph displays the changing percentage of positive feedback in the last three trials for each run given that a subjects true answer was always correct. Right:
Bilaterally the amygdala showed signicantly greater activations in controls compared to PSCs when subjects were repeatedly given feedback that they had chosen the wrong
square.

552

S. Klppel et al. / Neuropsychologia 48 (2010) 549557

Table 2
Details of task performance and questionnaires completed after the experiment. Results are reported with mean and standard deviation unless stated otherwise.

BIS total
STAI (state)
STAI (trait)
BDI
Snaith total
Snaith irritability subscore
Negative emotion with loosing (max = 9)a
Positive emotion with winning (max = 6)
Correct responses [%]
Expected performance of self (min = 100, max = +100)
Expected performance of 2nd player (min = 100, max = +100)
Condence in answer (min = 100, max = +100)
a

PSC

Controls

p-Values

62.0 11.8
34.0 12.6
34.2 10.4
5.7 5.9
13.2 6.4
5.6 3.3
2.0 (09)
3.5 1.5
93.8 4.9
17.4 30.0
57.6 25.0
29.4 25.8

63.3 8.9
31.7 8.9
38.0 10.8
6.9 6.1
13.3 5.5
5.5 2.9
2.0 (07)
3.7 1.5
93.8 4.9
36.7 33.1
49.0 25.5
38.7 32.6

0.71
0.57
0.32
0.59
0.95
0.96
0.22
0.55
0.99
0.1
0.35
0.33

Reported with median and range; MannWhitney test for between-group comparison.

tive feedback condition only. Additional regressors were included to model button
presses, the different screens as well as six regressors obtained at the realignment
step to account for movements (translations in three planes and rotations along
three axes). Trials when a player made a real mistake were excluded from further
analysis. We reasoned that cognitive processing of such trials would differ from
those where negative feedback followed correct performance. In the case of true
mistakes, participants might doubt their answers while negative feedback in the
second case is likely to be more unexpected. The resulting set of voxel values for
each contrast entered a second level analysis.
2.5. Group-level random effects analysis
2.5.1. Feedback on own performance
Parameter images of the differential effect of negative compared to positive
feedback entered a 2-sample t-test with 28 degrees of freedom resulting from 30
participants in two groups. A similar design was employed to study parameter estimates from the parametric modulator coding the percentage of negative feedback
in the last three trials (Fig. 2). As explained above, this analysis tests the hypothesis
that emotions build up when subjects repeatedly receive negative feedback, despite
correct performance.
2.5.2. Correlational analysis
We expected that subjects with a higher level of task-induced irritability
reported in questionnaires would show greater activations in the amygdala when
negative feedback was compared to positive feedback. We also correlated the
parameter images from the comparison of negative and positive feedback and those
from the parametric modulation of the percentage of negative feedback in the last
three trials with the estimated years to onset and included age as a separate regressor. As before, we expected increased activations of the amygdala with approaching
diagnostic disease onset and a higher frequency of erroneous negative feedback.
2.5.3. Regions of interest
The primary focus of the feedback and correlational analyses described above
was the amygdala, dened using the anatomy toolbox, based on post-mortem
tissue analysis (Eickhoff et al., 2005). We created a region of interest including all
voxels with at least a 50% probability of belonging to the amygdala. The second focus
was the medial OFC for which no such template is currently available. We therefore
created a sphere with a radius of one cm around the most activated voxel in the
medial OFC from the control group of a recent study on impulsive aggression (at
x, y, z = 6, 52, 20 in MNI space) (Coccaro et al., 2007). Correction for multiple
comparisons within each of the regions was performed using the false discovery
rate (FDR) as implemented in SPM5 at a critical p-value of 0.05 at the voxel level.
Outside our regions of interest voxels are reported if they survived FDR correction
performed across the whole brain.
2.5.4. Time series analysis
As in a related study (Coccaro et al., 2007), we performed an additional analysis
to test for a difference in the strength of the negative correlation between amygdala
and medial OFC that depends on gene status. Based solely on the fMRI time series
(and not on the parameter images used in other analyses), this analysis tests whether
activity in a brain region (i.e., the amygdala) correlates differentially with other brain
regions depending on gene status. This analysis is complementary to the group-level
analysis described above as it takes the fMRI time series from the whole experiment
and not just data from one specic condition. The time-courses in the left and right
amygdala were extracted from the centre of the probabilistic atlas of the amygdala
(Eickhoff et al., 2005) (centro-lateral segment) (x, y, z = 25, 9, 18 and 29, 8, 19
in MNI space). The time series was used as a single regressor in subsequent analysis.
The resulting parametric images entered a second level 2-sample t-test analysis as
described above.

2.5.5. Identity of the second player

We performed an exploratory three-way ANOVA with the factor GROUP (two
levels: PSC and controls) and two repeated-measures factors, PLAYER (two levels:
partner and computer) and CONDITION (four levels: positive/negative feedback on
each of a pair of players). Of primary interest to our research question was the subsequent F-test that identied regions showing an interaction of GROUP by PLAYER.
Based on previous studies (Gallagher, Jack, Roepstorff, & Frith, 2002; Rilling et al.,
2004) on interaction of humans with computers we expected the dorso-medial prefrontal cortex (dmPFC) to show increased activity if the partner was human. FDR
correction was performed in the same regions of interest as above. This analysis had
a strong exploratory component and we thus report all regions within the frontal
lobe signicant at p < 0.001 without correction for multiple comparisons for the
interaction of GROUP by PLAYER.

3. Results
3.1. Behavioural data and ratings
All participants, except one control and three PSCs, reported
a negative emotion with negative feedback. No signicant differences between the groups were found in any of the questionnaires
including those testing irritability and impulsivity (Table 2). A correlation matrix of the questionnaires is provided as a supplement.
Similarly, no signicant differences between PSC irritability selfratings and those of their companions were found using the Johns
Hopkins irritability questionnaire (PSC on themselves (mean SD):
11.6 6.3; companions on PSC: 10.5 6.1; p > 0.5, paired t-test;
data available from 15 pairs only). There were no signicant differences between PSCs close to and far from estimated clinical onset
except that those far from onset had a lower number of CAG repeats
(p = 0.02). Four PSCs and three controls had mild depressive symptoms indicated by a BDI over 10. The rating on Snaiths irritability
questionnaire was within the normal range (Snaith et al., 1978) in
all but one PSC. Task-induced negative emotions (composite score
of irritability, anger and tension) showed a signicant positive correlations with both Snaiths score (rSpearman = 0.31; p = 0.048) and
BIS-11 impulsivity score (rSpearman = 0.46; p = 0.005). Similarly, a
trend for a positive correlation was found with the Snaiths irritability subscore (rSpearman = 0.28; p = 0.064). Both groups performed
equally well and were equally condent about their decisions
(Table 2). There was a tendency that PSCs expected to perform
worse than controls (p = 0.1). No correlations of scores in questionnaires with the estimated years to clinical onset were found.
Interviews and debrieng conrmed that subjects believed in the
correctness of feedback and the link between cooperating players.
3.2. Image data analysis
No signicant activations in amygdala or OFC were observed
when negative and positive feedback was compared for players in
the scanner either separately for each group or for both groups
combined. There was also no signicant effect when testing for an

S. Klppel et al. / Neuropsychologia 48 (2010) 549557

553

Fig. 3. Areas showing a stronger negative correlation between right amygdala activity (white circle) in the control group compared to PSCs. The seed region in the amygdala is
enlarged for visualisation purposes. The plots on the left display the strength and direction of coupling. Bars report the strength of correlation in arbitrary units (a.u.) with 90%
condence intervals. Imaging results are overlaid on the mean brain from all subjects in MNI space at a threshold, for visualisation purposes only, of p < 0.01 (uncorrected).

interaction with group status inside one of the regions of interest or

after correction across the whole brain. In a subsequent analysis we
tested for increased neuronal responses in the presence of repeated
negative feedback, despite correct performance. In controls, greater
activation was found in the left amygdala (T = 3.79; p = 0.05 at x, y,
z = 18, 6, 10) the higher the proportion of negative feedback.
Neither the effect in the right amygdala of controls (T = 3.2 at x, y,
z = 30, 2, 12) nor bilaterally in PSCs (T < 2.8) survived correction
for multiple comparisons. A signicant interaction with gene status was found in left amygdala and a strong trend in right amygdala
resulted from PSCs failing to show the expected positive correlation with negative feedback (T = 4.02; p = 0.005; at x, y, z = 18, 6,
12 and T = 3.37; p = 0.06 at x, y, z = 18, 8, 14; Fig. 2). No differences between PSCs close to and far from clinical onset were found.
There were also no differences between groups in the OFC region of
interest. No further signicant voxels were found when the search
volume included the whole brain.
3.2.1. Time series analysis
The signicant interaction in the time series analysis resulted
from greater negative coupling between activations in right amygdala and medial OFC in controls than in PSCs (T = 2.81; p = 0.05 at x, y,
z = 6, 46, 18) (Fig. 3). The gure also illustrates group specic main
effects, which did not survive at a corrected level. Activations did
not differ signicantly between PSC groups differing in proximity
to clinical onset.
3.2.2. Correlational analysis
Task-specic negative emotional ratings from controls correlated positively with the fMRI signal for negative vs. positive

feedback bilaterally in the amygdala (T = 3.11; p = 0.05 at x, y,

z = 20, 8, 10 and T = 3.77; p = 0.036 at x, y, z = 26, 4, 12) and
there was also a signicant group interaction in the right amygdala (T = 3.61; p = 0.018 at x, y, z = 16, 4, 16) (Fig. 4). Thus, higher
levels of reported irritation resulted in stronger activations of the
amygdala in controls compared to PSCs for which correlations were
virtually absent (T < 1.5). Controls with lower levels of reported
task-induced irritation showed higher neuronal activations in the
right OFC (T = 4.64, p = 0.003 at x, y, z = 12, 52, 12) and a similar
trend on the left (T = 2.6 at x, y, z = 10, 48, 12). No signicant
correlations with the reported ratings and activations in OFC were
found for PSC and there was no signicant interaction. In addition,
we did not nd signicant correlations with the estimated years to
diagnostic onset.
3.3. Effect of second player
In controls but not PSCs, two areas in the frontal lobe, the dorsal
anterior cingulate cortex and dmPFC, showed differential activity
depending on whether the second player was a partner or a computer (F = 9.32; p(uncorrected) = 0.003 at x, y, z = 4, 42, 10) and F = 11.80;
p(uncorrected) = 0.001; at x, y, z = 4, 28, 50) (Fig. 5). Neither signicant
effect was found in amygdala or OFC region of interest, nor was
there a signicant interaction in these regions. Very similar areas
were found when testing for an interaction of GROUP with PLAYER
(F = 12.01; p(uncorrected) = 0.001 at x, y, z = 4, 40, 12) and (F = 13.4;
p(uncorrected) < 0.001 at x, y, z = 8, 34, 56) (Fig. 5). An additional peak
was found in the right dorso-lateral prefrontal cortex (F = 13.97;
p(uncorrected) < 0.001 at x, y, z = 28, 26, 40). No signicant interactions
of GROUP with PLAYER or GROUP with PLAYER and CONDITION
were found in amygdala or OFC.

Fig. 4. Areas showing an interaction between groups in the subject-specic rating of task-induced negative emotions. Whereas controls show increased activations in
amygdala with increasing levels of reported negative emotions, such correlations are absent in the PSC group. Graphs show the correlation of rating with activation at the
voxel marked by cross hairs in controls (green triangle) and PSC (red diamonds). Results are displayed at a threshold of p < 0.01 (uncorrected). a.u.: arbitrary units. (For
interpretation of the references to color in this gure legend, the reader is referred to the web version of the article.)

554

S. Klppel et al. / Neuropsychologia 48 (2010) 549557

Fig. 5. Effect of identity of the second player. The top left panel depicts areas where controls show increased activations when the second player is a human partner compared
to a computer. The lower left panel indicates that this effect is weaker in PSCs, resulting in a positive interaction in both areas. The two right panels depict the signal change
in both areas when partner and computer conditions are compared (co-ordinates are in MNI space). In both areas, PSCs have reduced activations when the second player is a
companion. Error bars indicate 90% condence intervals. Results are displayed at a level of p < 0.01 (uncorrected) for visualisation purposes. dACC: dorsal anterior cingulate
cortex; dmPFC: dorso-medial prefrontal cortex; a.u.: arbitrary units.

4. Discussion
4.1. Behavioural data
We found equal scorings on ratings or questionnaires by PSCs
and controls. Both groups scored within the normal range on the
Snaith irritability questionnaire. A recent study suggests that clinically overt irritability is found in around 20% of PSCs who are less
than 10 years to estimated clinical onset, but is rare before that
(Julien et al., 2007). Our sample contained only three subjects with
less than 10 years to onset and was not pre-selected for indications
of increased irritability. Furthermore, companions may have been
more stressed than unrelated controls. No differences between
PSC self rating of irritability and those of their companions were
found which mirrors ndings from the study of Chatterjee et al.
(2005). The BDI identied subjects with mild or moderate depressive symptoms (BDI > 10) in both groups, potentially reecting the
emotional burden of pre-symptomatic HD on close companions as
well as PSCs.
Equal levels of self-reported task-induced irritation associated with differential group specic neural responses suggest that
either compensatory mechanisms are in place or that the irritability questionnaires lacks sensitivity to detect subtle differences
between groups. Their improvement is an important part of ongoing research.
4.2. Feedback on own performance
Our fMRI analysis leads us to reject our rst hypothesis regarding task-induced amygdala activations. PSC did not show greater
activations in the amygdala than controls when negative feedback
was compared to positive feedback. We found no indication for differential activations in the OFC/amygdala axis with the identity of

the second player. We can, however conrm the hypothesis of a

reduced coupling between amygdala and medial OFC in the PSC
group.
All three imaging analyses indicate that neuronal responses in
PSCs were modulated less by external (i.e., the proportion of erroneous responses) and internal factors (i.e., the level of task-induced
irritation) than in controls. Increasing false negative feedback (indicated either by subject-specic ratings or a higher frequency of
negative feedback) resulted in increased activations in overlapping
areas of the amygdala in controls, whereas this effect was reduced
or absent in PSCs. We argue that the inappropriate responses of the
amygdala/medial OFC axis make PSCs prone to the development of
psychiatric symptoms such as irritation.
4.3. Time series analysis
A disruption of the amygdala/medial OFC axis has been found
in recent work on impulsive aggression (Coccaro et al., 2007)
and also in subjects suffering from depression with anger attacks
(Dougherty et al., 2004). Early involvement of the amygdala in HDrelated pathology (Douaud et al., 2006; Rosas et al., 2003) could be
the basis of the reduced coupling and responsiveness to reported
levels of experienced negative emotions. The similarity of ndings to aggressive disorders lends support to the presence of an
aggressive element in HD, which has also been suggested by neuropsychological factor analysis (Craufurd et al., 2001). While the
aforementioned studies included patients with psychiatric symptoms (Coccaro et al., 2007; Dougherty et al., 2004) our PSCs showed
normal levels of irritability as measured by the Snaith scale (Snaith
et al., 1978).
Reduced functional coupling could limit the ability of PSCs to
relate to the strength and value of an emotion (Anderson et al.,
2003). At least for the emotion of fear, it has previously been

S. Klppel et al. / Neuropsychologia 48 (2010) 549557

reported that under conditions of diminished conscious emotional

awareness there is a decrease of connectivity between amygdala
and cortical association areas (Williams et al., 2006). PSCs could be
failing to recognise their own emotional state or might be denying
it, a notion that is supported by a study comparing symptom ratings between HD patients and collaterals (Chatterjee et al., 2005;
Hoth et al., 2007).
4.4. Effect of second player
Based on clinical experience, we hypothesised that irritability in PSCs is particularly pronounced in their interactions with
close companions. No clear indication for this hypothesis was found
by interview. Both groups consistently reported that playing with
their partner was more emotionally involving but that this was
the case for both positive and negative feedback. The interactive
element of our task was far less pronounced than in other studies (Rilling et al., 2004); although the outcome of a given round
depended on both players, the response of one player did not
depend on that of the other. This is likely to have reduced the effect
of the second players identity.
The dorsal anterior cingulate cortex and dmPFC have been implicated in social interaction (Amodio & Frith, 2006). These regions
were indeed activated by controls when playing with a partner
rather than a computer, whereas this effect was absent in PSCs.
This nding could represent a reduction in the ability to take the
perspective of others, so that the identity of a second player mattered more to controls than to PSCs. A complimentary view is that
dmPFC areas represent others views of the self (Amodio & Frith,
2006). Impairment in assessing what other people think about oneself (e.g., as when the self gets false negative feedback) could make
one more prone to aggressive and disinhibited behaviour. However, these ndings should be interpreted cautiously since they
stem from an exploratory analysis and the interview results provide
little to differentiate between the two possible explanations.
Although instructions stated that both players needed to give
correct answers in order to win a round and the money, three PSCs
and controls perceived the task as a friendly competition rather
than a collaborative effort. Such heterogeneity is likely to reduce
sensitivity for detection of differences between the two groups,
especially in amygdala. We conrmed our pilot study as subjects
experienced irritation with the task and believed in the correctness of feedback and the link between computers, but did not
pick up this ambiguity.
4.5. Limitations
A number of aspects of the current study should be noted.
Despite a trend for PSCs expecting to perform worse prior to engaging in the task, both groups actually performed equally well, with
similar levels of induced positive and negative emotions and condence in their decisions. A task with low levels of cognitive demand
helped to ensure that subtle cognitive impairments in PSCs were
not a source of between-group differences, making the observed
differences in neuronal activations more likely to reect the disease process and compensation for it. Notably, PSC in our study
were aware of their genetic status which is likely to have inuenced them at the behavioural and neuronal levels. Given the
widespread availability of genetic testing, this situation is likely
to represent the standard situation in a clinical setting when dealing with irritability and the study aimed to understand processing
in these circumstances. More to the point, most studies with atrisk subjects ignorant of their genetic status were performed after
subjects had decided to undergo genetic testing. The uncertainty
and signicance of the testing procedure, that includes manda-

555

tory counselling, could themselves modify clinical presentations

of HD-related irritability.
As pointed out in the introduction, there is some evidence
for structural changes affecting the amygdala. Although the analysis of the structural data acquired for the current study (see
supplementary material) failed to identify changes in the structures
identied with functional imaging, it is possible that structural
changes underlie some of the functional differences found via
remote effects.
The relative absence of studies on irritability in either healthy
or diseased populations made it necessary to base this study on
hypothetical assumptions. While the relationship of irritability to
impulsive aggression is supported by a factor analysis (Craufurd et
al., 2001), other clinical presentations of HD, including poor sleep
and cognitive dysfunction, may play an important role and were
not evaluated. As mentioned in the Methods section, each subject
may have a different concept of irritability. Subjects were therefore
asked to rate the intensity of three negative emotions (irritability, anger and tension). The signicant positive correlation found
between these emotions and the impulsivity and irritability questionnaires illustrates that the task is at least related to the presented
concept of irritability. Given that irritability co-occurs with depression and anxiety and HD and that the results from the respective
questionnaires were correlated with each other (see correlation
matrix in the supplement) we cannot claim to be looking at irritability specically and exclusively. Future work could focus on subjects
pre-selected for increased levels of irritability to see if results are
similar in quality and if the effect of contributing factors such as
depression poor sleep and cognitive dysfunction can be isolated.
The recording of additional physiological data, e.g., heart rate or
skin conductance could also prove useful to nd effects related to
irritability.
All questionnaires were completed after the experiment to
minimize the effect of reecting on emotions from inuencing
behaviour and cognitive processing as the experiment progressed.
This retrospective nature could have inated correlations between
questionnaires that were or were not related to the task (e.g.,
Snaiths questionnaire). Finally, a number of the imaging results we
report are not signicant after correction for multiple comparisons.
We report them descriptively to illustrate that ndings in amygdala
and OFC are very probably bilateral and to generate hypothesis of
the processing of the second players identity.
5. Conclusion
In conclusion, our results indicate a poor dependency of neuronal responses on reported emotions or the quality of recent
feedback in PSC. The disruption of emotion processing circuits
may well predispose to the development of recognised psychiatric
states, such as irritability. There is evidence that the processing of
facial expressions of negative emotions is affected already in PSC
(Gray, Young, Barker, Curtis, & Gibson, 1997; Johnson et al., 2007)
and that dysphoria inducing pictures elicit differential brain activations in early affected HD patients compared to controls (Paradiso
et al., 2008). Future research could clarify if these impairments of
emotion processing predispose one to the clinical manifestation of
irritability.
Competing interests
None
Funding
This work was supported by the Wellcome Trust (grant 075696
2/04/2 to R.S.J.F., and S.J.T.) and the Bundesministerium fr

556

S. Klppel et al. / Neuropsychologia 48 (2010) 549557

Forschung und Gesundheit (BMBF grant 01GW0730 to O.T.). While

writing the manuscript S.K. was funded by the Medizinische
Fakultt of the University of Freiburg. P.P. is supported by a grant
from Vetenskaps rdet and Hjrnfonden, Sweden.
Acknowledgements
We would like to thank Susie Henley as well as Maggie Burrows,
Rachel Taylor, Tom Warner and Edward Wild for their help with the
recruitment. We also thank Natasja van Harskamp for the scoring
of the NART. Furthermore, we would like to thank Chris Frith and
Christoph Kaller for helpful suggestions for the preparation of this
manuscript.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.neuropsychologia.2009.10.016.
References
Amodio, D. M., & Frith, C. D. (2006). Meeting of minds: The medial frontal cortex and
social cognition. Nature Reviews: Neuroscience, 7, 268277.
Anderson, A. K., Christoff, K., Stappen, I., Panitz, D., Ghahremani, D. G., Glover, G., et
al. (2003). Dissociated neural representations of intensity and valence in human
olfaction. Nature Neuroscience, 6, 196202.
Anderson, S. W., Bechara, A., Damasio, H., Tranel, D., & Damasio, A. R. (1999). Impairment of social and moral behavior related to early damage in human prefrontal
cortex. Nature Neuroscience, 2, 10321037.
Ashburner, J., & Friston, K. J. (2000). Voxel-based morphometryThe methods. Neuroimage, 11, 805821.
Barratt, E. S., & Patton, J. H. (1983). Impulsivity: Cognitive, behavioral and psychophysiological correlates. In M. Zuckerman (Ed.), Biological basis of sensation
seeking, impulsivity, and anxiety (pp. 77122). Hillsdale, NJ: Lawrence Erlbaum
Associates.
Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory
for measuring depression. Archives of General Psychiatry, 4, 561571.
Best, M., Williams, J. M., & Coccaro, E. F. (2002). Evidence for a dysfunctional prefrontal circuit in patients with an impulsive aggressive disorder. Proceedings of
the National Academy of Sciences of the United States of America, 99, 84488453.
Blair, R. J. (2007). Dysfunctions of medial and lateral orbitofrontal cortex in psychopathy. Annals of the New York Academy of Sciences, 1121, 461479.
Chatterjee, A., Anderson, K. E., Moskowitz, C. B., Hauser, W. A., & Marder, K. S. (2005).
A comparison of self-report and caregiver assessment of depression, apathy,
and irritability in Huntingtons disease. Journal of Neuropsychiatry and Clinical
Neurosciences, 17, 378383.
Coccaro, E. F, & Kavoussi, R. J. (1997). Fluoxetine and impulsive aggressive behavior in
personality-disordered subjects. Archives of General Psychiatry, 54, 10811088.
Coccaro, E. F., McCloskey, M. S., Fitzgerald, D. A., & Phan, K. L. (2007). Amygdala and
orbitofrontal reactivity to social threat in individuals with impulsive aggression.
Biological Psychiatry, 62, 168178.
Craufurd, D., Thompson, J. C., & Snowden, J. S. (2001). Behavioral changes in Huntington disease. Neuropsychiatry, Neuropsychology, and Behavioral Neurology, 14,
219226.
Davidson, R. J., Putnam, K. M., & Larson, C. L. (2000). Dysfunction in the neural
circuitry of emotion regulationA possible prelude to violence. Science, 289,
591594.
Deichmann, R., Gottfried, J. A., Hutton, C., & Turner, R. (2003). Optimized EPI for fMRI
studies of the orbitofrontal cortex. Neuroimage, 19, 430441.
Deichmann, R., Schwarzbauer, C., & Turner, R. (2004). Optimisation of the 3D MDEFT
sequence for anatomical brain imaging: Technical implications at 1.5 and 3 T.
Neuroimage, 21, 757767.
De Marchi, N., Daniele, F., & Ragone, M. A. (2001). Fluoxetine in the treatment of
Huntingtons disease. Psychopharmacology, 153, 264266.
Douaud, G., Gaura, V., Ribeiro, M. J., Lethimonnier, F., Maroy, R., Verny, C., et al.
(2006). Distribution of grey matter atrophy in Huntingtons disease patients:
A combined ROI-based and voxel-based morphometric study. Neuroimage, 32,
15621575.
Dougherty, D. D., Rauch, S. L., Deckersbach, T., Marci, C., Loh, R., Shin, L. M.,
et al. (2004). Ventromedial prefrontal cortex and amygdala dysfunction during an anger induction positron emission tomography study in patients with
major depressive disorder with anger attacks. Archives of General Psychiatry, 61,
795804.
Eickhoff, S. B., Stephan, K. E., Mohlberg, H., Grefkes, C., Fink, G. R., Amunts, K., et al.
(2005). A new SPM toolbox for combining probabilistic cytoarchitectonic maps
and functional imaging data. Neuroimage, 25, 13251335.
Folstein, S. E., Chase, G. A., Wahl, W. E., McDonnell, A. M., & Folstein, M. F. (1987).
Huntington disease in Maryland: Clinical aspects of racial variation. American
Journal of Human Genetics, 41, 168179.

Friston, K. J., Frith, C. D., Turner, R., & Frackowiak, R. S. (1995). Characterizing evoked
hemodynamics with fMRI. Neuroimage, 2, 157165.
Gallagher, H. L., Jack, A. I., Roepstorff, A., & Frith, C. D. (2002). Imaging the intentional
stance in a competitive game. Neuroimage, 16, 814821.
Gray, J. M., Young, A. W., Barker, W. A., Curtis, A., & Gibson, D. (1997). Impaired
recognition of disgust in Huntingtons disease gene carriers. Brain, 120(Pt 11),
20292038.
Hamilton, J. M., Salmon, D. P., Corey-Bloom, J., Gamst, A., Paulsen, J. S., Jerkins, S., et
al. (2003). Behavioural abnormalities contribute to functional decline in Huntingtons disease. Journal of Neurology, Neurosurgery and Psychiatry, 74, 120122.
HD Collaborative Research Group. (1993). A novel gene containing a trinucleotide
repeat that is expanded and unstable on Huntingtons disease chromosomes.
The Huntingtons Disease Collaborative Research Group. Cell, 72, 971983.
Hoth, K. F., Paulsen, J. S., Moser, D. J., Tranel, D., Clark, L. A., & Bechara, A. (2007).
Patients with Huntingtons disease have impaired awareness of cognitive,
emotional, and functional abilities. Journal of Clinical and Experimental Neuropsychology, 29, 365376.
Hutton, C, Bork, A., Josephs, O., Deichmann, R., Ashburner, J., & Turner, R. (2002).
Image distortion correction in fMRI: A quantitative evaluation. Neuroimage, 16,
217240.
Johnson, S. A., Stout, J. C., Solomon, A. C., Langbehn, D. R., Aylward, E. H., Cruce, C. B.,
et al. (2007). Beyond disgust: Impaired recognition of negative emotions prior
to diagnosis in Huntingtons disease. Brain, 130, 17321744.
Julien, C. L., Thompson, J. C., Wild, S., Yardumian, P., Snowden, J. S., Turner, G., et
al. (2007). Psychiatric disorders in preclinical Huntingtons disease. Journal of
Neurology, Neurosurgery and Psychiatry, 78, 939943.
Kloppel, S., Draganski, B., Siebner, H. R., Tabrizi, S. J., Weiller, C., & Frackowiak,
R. S. (2009). Functional compensation of motor function in pre-symptomatic
Huntingtons disease. Brain, 132, 16241632.
Langbehn, D. R., Brinkman, R. R., Falush, D., Paulsen, J. S., & Hayden, M. R. (2004). A
new model for prediction of the age of onset and penetrance for Huntingtons
disease based on CAG length. Clinical Genetics, 65, 267277.
Mann, D. M., Oliver, R., & Snowden, J. S. (1993). The topographic distribution of
brain atrophy in Huntingtons disease and progressive supranuclear palsy. Acta
Neuropathologica, 85, 553559.
Nelson, H. E., & Willison, J. (1991). The national adult reading test. Windsor, UK: NFERNelson.
Paleacu, D., Anca, M., & Giladi, N. (2002). Olanzapine in Huntingtons disease. Acta
Neurologica Scandinavica, 105, 441444.
Paradiso, S., Turner, B. M., Paulsen, J. S., Jorge, R., Ponto, L. L., & Robinson, R. G. (2008).
Neural bases of dysphoria in early Huntingtons disease. Psychiatry Research, 162,
7387.
Pavese, N., Andrews, T. C., Brooks, D. J., Ho, A. K., Rosser, A. E., Barker, R. A., et al. (2003).
Progressive striatal and cortical dopamine receptor dysfunction in Huntingtons
disease: A PET study. Brain, 126, 11271135.
Ranen, N. G., Lipsey, J. R., Treisman, G., & Ross, C. A. (1996). Sertraline in the treatment
of severe aggressiveness in Huntingtons disease. Journal of Neuropsychiatry and
Clinical Neurosciences, 8, 338340.
Reading, S. A., Dziorny, A. C., Peroutka, L. A., Schreiber, M., Gourley, L. M., Yallapragada, V., et al. (2004). Functional brain changes in presymptomatic Huntingtons
disease. Annals of Neurology, 55, 879883.
Rempel-Clower, N. L. (2007). Role of orbitofrontal cortex connections in emotion.
Annals of the New York Academy of Sciences, 1121, 7286.
Rilling, J. K., Sanfey, A. G., Aronson, J. A., Nystrom, L. E., & Cohen, J. D. (2004). The neural
correlates of theory of mind within interpersonal interactions. Neuroimage, 22,
16941703.
Rosas, H. D., Koroshetz, W. J., Chen, Y. I., Skeuse, C., Vangel, M., Cudkowicz, M. E.,
et al. (2003). Evidence for more widespread cerebral pathology in early HD: An
MRI-based morphometric analysis. Neurology, 60, 16151620.
Rosas, H. D., Liu, A. K., Hersch, S., Glessner, M., Ferrante, R. J., Salat, D. H., et al. (2002).
Regional and progressive thinning of the cortical ribbon in Huntingtons disease.
Neurology, 58, 695701.
Siever, L. J. (2008). Neurobiology of aggression and violence. American Journal of
Psychiatry, 165, 429442.
Snaith, R. P., Constantopoulos, A. A., Jardine, M. Y., & McGufn, P. (1978). A clinical scale for the self-assessment of irritability. British Journal of Psychiatry, 132,
164171.
Snowden, J. S., Gibbons, Z. C., Blackshaw, A., Doubleday, E., Thompson, J., Craufurd,
D., et al. (2003). Social cognition in frontotemporal dementia and Huntingtons
disease. Neuropsychologia, 41, 688701.
Spielberger, C. D. (1988). Manual for the State-Trait Anger Expression Scale (STAXI).
Odessa, FL: Psychological Assessment Resources.
Squitieri, F., Cannella, M., Piorcellini, A., Brusa, L., Simonelli, M., & Ruggieri, S. (2001).
Short-term effects of olanzapine in Huntington disease. Neuropsychiatry, Neuropsychology, and Behavioral Neurology, 14, 6972.
Tebartz van Elst, L., Woermann, F. G., Lemieux, L., Thompson, P. J., & Trimble, M. R.
(2000). Affective aggression in patients with temporal lobe epilepsy: A quantitative MRI study of the amygdala. Brain, 123(Pt 2), 234243.
Urry, H. L., van Reekum, C. M., Johnstone, T., Kalin, N. H., Thurow, M. E., Schaefer,
H. S., et al. (2006). Amygdala and ventromedial prefrontal cortex are inversely
coupled during regulation of negative affect and predict the diurnal pattern
of cortisol secretion among older adults. Journal of Neuroscience, 26, 4415
4425.
Wheelock, V. L., Tempkin, T., Marder, K., Nance, M., Myers, R. H., Zhao, H., et al.
(2003). Predictors of nursing home placement in Huntington disease. Neurology,
60, 9981001.

S. Klppel et al. / Neuropsychologia 48 (2010) 549557

Williams, L. M., Das, P., Liddell, B. J., Kemp, A. H., Rennie, C. J., & Gordon, E.
(2006). Mode of functional connectivity in amygdala pathways dissociates
level of awareness for signals of fear. Journal of Neuroscience, 26, 9264
9271.

557

Wolf, R. C., Vasic, N., Schonfeldt-Lecuona, C., Landwehrmeyer, G. B., & Ecker,
D. (2007). Dorsolateral prefrontal cortex dysfunction in presymptomatic
Huntingtons disease: Evidence from event-related fMRI. Brain, 130, 2845
2857.