Psychopharmacology (2009) 203:685691

DOI 10.1007/s00213-008-1410-6

ORIGINAL INVESTIGATION

Early effects of mirtazapine on emotional processing

D. Arnone & J. Horder & P. J. Cowen & C. J. Harmer

Received: 26 June 2008 / Accepted: 5 November 2008 / Published online: 25 November 2008
# Springer-Verlag 2008

Abstract
Background Acute administration of selective serotonin
and noradrenaline re-uptake blockers to healthy volunteers
affects the processing of emotional information but it is not
known if similar effects occur with antidepressants acting
through other pharmacological mechanisms. Mirtazapine is
a clinically established antidepressant with complex actions
involving blockade of noradrenaline 2-adrenoceptors as
well as a number of 5-HT receptor subtypes. The aim of the
present study was to test whether, like monoamine reuptake inhibitors, mirtazapine would also produce positive
biases in emotional processing.
Methods We studied 30 healthy volunteers who received
either a single dose of mirtazapine (15 mg) or placebo in a
parallel group, double-blind study. Two hours following
medication administration, participants completed a battery
of tasks testing various aspects of emotional processing
including facial expression recognition, emotion potentiated
startle, and emotional categorization and memory.
Results Compared to placebo, mirtazapine significantly
impaired the recognition of fearful facial expressions and
reduced eye-blink responses in the emotion potentiated
startle task. Participants receiving mirtazapine were also
significantly quicker to respond to emotional self-relevant
information in the categorization task and showed a positive
bias in memory recall compared to those receiving placebo.
D. Arnone : J. Horder : P. J. Cowen : C. J. Harmer (*)
Department of Psychiatry, Oxford University, Warneford Hospital,
Oxford OX3 7JD, UK
e-mail: catherine.harmer@psych.ox.ac.uk
Present address:
D. Arnone
Neuroscience and Psychiatry Unit, University of Manchester,
Manchester, UK

Conclusions Our findings indicate that mirtazapine reduces

fear processing in healthy volunteers, an effect similar to that
produced by repeated administration of selective serotonin reuptake inhibitors. In addition, mirtazapine increased memory
for likeable versus dislikeable self-relevant information
suggesting an induction of positive bias in emotional memory.
Such effects may be important for our understanding of the
neuropsychological mechanisms of antidepressant action in
both anxiety and depressive disorders.
Keywords Antidepressant . Emotion . Serotonin . Memory

Introduction
Most antidepressants in current clinical use potentiate the
activity of serotonin (5-HT) and/or noradrenaline (NA)
through blockade of the relevant neurotransmitter re-uptake
site in the presynaptic nerve terminal (see Nutt 2002). We
have previously shown that acute administration of selective 5-HT and NA re-uptake blockers to healthy volunteers
produces positive biases in the processing of emotional
information in the absence of any changes in subjective
mood. For example, a single dose of the selective
noradrenaline re-uptake inhibitor (NARI), reboxetine, increased the recognition of facial expressions of happiness
and speeded the classification and improved the recall of
likeable vs dislikeable personal descriptors (Harmer et al
2003b). Similarly, acute administration of the selective
serotonin re-uptake inhibitor (SSRI), citalopram, increased
the recognition of happy facial expressions (Harmer et al
2003a), suggesting an overlapping effect of these two
different re-uptake blockers.
However, unlike reboxetine, acute administration of
citalopram also increased fear processing, seen as exaggerated

686

startle responses in the emotion potentiated startle paradigm

(Browning et al 2007, Grillon et al. 2005) and enhanced
fearful facial expression recognition (Harmer et al 2003a,
Browning et al 2007). These different patterns of effect led
us to hypothesize early positive effects of antidepressants on
affective processing, which may be important in their
therapeutic effects in depression but also increased threat
processing which may contribute to the early anxiogenic
effects of SSRIs in some clinical groups. In line with this
hypothesis, these effects on fear processing reversed following repeated citalopram administration in healthy volunteers
(Harmer et al 2004). A similar profile of increased then
decreased fear conditioning has also been observed with
acute versus repeated SSRI administration in rodents
(Burghardt et al. 2004). The mechanisms mediating this
reversal are unclear but it has been suggested that they
involve downregulation of postsynaptic 5-HT2 receptors
with repeated SSRI administration (Deakin and Graeff
1991)
It is important to find out whether these early effects on
emotional processing also occur with antidepressants which
do not act via NA or 5-HT re-uptake blockade. Mirtazapine
is a clinically established antidepressant with complex
pharmacological actions involving blockade of a variety
of monoamine receptors including NA 2-adrenoceptors
and 5-HT2A, 5-HT2C, and 5-HT3 receptors (Davis and
Wilde 1996). Blockade of 2-adrenoceptors would be
expected to increase neurotransmission through NA pathways and has also been suggested to result indirectly in
increased 5-HT release, though whether this occurs in
clinical use is controversial (Whale et al 2000). Clinically
mirtazapine has a sedating profile which is probably due to
H1 histamine receptor blockade (Davis and Wilde 1996).
The aim of the present study was to test the effect of
acute mirtazapine administration on a battery of emotional
processing tasks previously shown to be sensitive to
antidepressant drug administration. On the basis of mirtazapines pharmacological profile, including increased noradrenergic transmission, we predicted it would enhance
positive affective memory and the recognition of happy
facial expressions. We also expected that the recognition of
fearful facial expressions and startle responses would be
decreased following mirtazapine, similar to that observed
with repeated SSRI administration and consistent with the
expected effects of 5-HT2A/C receptor blockade.

Materials and methods

Subjects
Thirty healthy volunteers aged 1850 years were recruited
after providing written informed consent and were reim-

Psychopharmacology (2009) 203:685691

bursed with 60 for their involvement in the study. The

project was approved by the Oxfordshire Research Ethics
Committee. Exclusion criteria were: current use of any
psychotropic medication apart from the contraceptive pill,
any major physical disorder, and current or past Axis 1
disorders according to the Structured Clinical Interview for
DSM IV (Spitzer and Gibbon 2002). Volunteers were
initially screened to ensure compliance with study criteria
and completed subjective rating scales including the Beck
Depression Inventory (BDI), the National Adult Reading
Test (NART), and the Eysenck Personality Questionnaire
(EPQ, Neuroticism, Extraversion, Lie scale, Psychoticism).
On the test day and after a light meal, subjects were
randomly allocated to either a single dose of mirtazapine
(15 mg) or placebo (in an identical capsule) in a parallel
group, double-blind experiment. Following mirtazapine/
placebo administration subjects rested quietly for 2 h and
were asked to read or work and self-rate at hourly intervals
using visual analogue scales (VAS) with a 100-ms range for
the following variables: alert, anxious, dizzy, happy,
hungry, nausea, and sad. At the end of the 2 h volunteers
underwent a battery of psychological tests for approximately another 2 h. Another set of self-reported VAS completed
the experiment.
Psychological tests
Facial expression recognition
The facial expression recognition task featured six basic
emotions (happiness, surprise, sadness, fear, anger, and
disgust) taken from the Pictures of Affect Series (Ekman
and Friesen 1976), which had been morphed between each
prototype and neutral (Young et al. 1997). Briefly, this
procedure involved taking a variable percentage of the
shape and texture differences between the two standard
images 0% (neutral) and 100% (full emotion) in 10% steps.
Four examples of each emotion at each intensity were given
(total of 10 individuals). Each face was also given in a
neutral expression, giving a total of 250 stimuli presentations. The facial stimuli were presented on a computer
screen (in a randomized order) for 500 ms and replaced by
a blank screen. The task was split into two halves with a
short rest period (12 min) in between. Volunteers made
their responses by pressing a labeled key on the keyboard.
Each participant was asked to respond as quickly and as
accurately as possible.
Accuracy scores for each emotion were computed for
each intensity level and averaged across 20% intensity bins
(i.e 1020%, 3040% etc.). Reaction time for correct
choices and misclassifications (number of responses to
each facial expression category classified as each of the
other facial expression categories) were also recorded in

Psychopharmacology (2009) 203:685691

this task. This allows effects on accuracy of emotional

recognition to be compared with general changes in
responding.
Emotion-potentiated startle
Picture stimuli from the International Affective Picture
System, designed to elicit positive, negative or neutral
emotions were used (Lang et al. 1998). Stimuli were
presented for 13 s (inter-trial interval of between 1115 s,
mean 13 s) on a 43-cm computer screen approximately 1 m
away from the volunteer. Pictures were presented in three
blocks each containing seven pictures of each category in a
fixed random order with the constraint that no two of the
same type (neutral, positive, or negative) were presented
successively. The eye-blink component of the startle
response was recorded from the orbicularis oculi using
electromyography (EMG-startle response system, San
Diego Instruments, USA). Acoustic probes were 50 ms,
95 dB bursts of white noise with a nearly instantaneous rise
time and are delivered binaurally through headphones
(delivered at 1.5, 4.5, or 7.5 s following picture onset).
Within each block of 21 pictures, probes were delivered on
five of each trial type (neutral, positive, or negative). To
limit expectation of the noise, two trials per valence did not
contain any startle probes and three probes per block were
given within the inter-trial interval. To habituate volunteers
to the startle probes and to orient them to the procedure,
volunteers viewed an introductory set of nine neutral
pictures and received nine startle probes (two of these
within the inter-trial interval).
EMG signals were filtered (low cut off, 0.5 Hz; high cut
off, 300 Hz) and rectified. Eye-blink reflex magnitudes in
millivolt were calculated by subtracting the amount of
integrated EMG at reflex onset from the peak amplitude
maximum amount of integrated EMG between 20 and
120 ms following probe onset. Trials with no perceptible
eye-blink reflex were assigned a magnitude of zero and
included in the analysis. Eye-blink reflexes with excessive
noise during a 20-ms, prestartle baseline period were
excluded. Of the 30 volunteers, 11 volunteers were not
included in the analysis because of electrode interference or
because they displayed fewer than 25% satisfactory blink
responses in the paradigm (five in drug group, six in
placebo group).
Emotional categorization and memory
Sixty personality characteristics selected to be disagreeable
(e.g. domineering, hostile) or agreeable (e.g. cheerful,
generous), taken from Anderson (1968), were presented
on a computer screen for 500 ms. These words were
matched for length and ratings of frequency and meaning-

687

fulness. Participants were asked to categorize these personality characteristics as likeable or dislikeable as quickly and
as accurately as possible. Specifically, they were asked to
imagine whether they would like or dislike overhearing
someone referring to them as possessing this characteristic,
so that the judgment was in part self-referential. Classifications (i.e. number of likeable/dislikeable characteristics
assigned as likeable/dislikeable) and response times for
correct identification were recorded.
Emotional memory was examined by a surprise test of
recall of the personality traits used in the emotional
categorization task, immediately after task completion.
Volunteers were asked to write down as many words as
they could remember within a 2-min time limit. The
number of items correctly recalled for each valence
(likeable and dislikeable) was calculated.
Statistical analysis
Baseline demographics, excluding gender, were analyzed
by using independent t tests. All other measures and tasks
were analyzed using split-plot two-way analyses of variance with group (mirtazapine or placebo) as the betweensubject factor. Within subjects factors were emotion and
time (VAS), emotion (emotional categorization, emotional
memory, startle task), and emotional intensity for accuracy
in the facial expression recognition task. Where assumptions of equality of variances were broken, an factor was
calculated and used to correct the degrees of freedom using
the GreenhouseGeisser procedure (Howell 2002). For
clarity, uncorrected degrees of freedom are reported here.

Results
Baseline characteristics
There were no significant differences between groups in
terms of demographic variables, BDI, NART, and EPQ
scores (see Table 1). Baseline mood ratings made with the
VAS were also not different in the two groups (Table 1).
Effects of drug administration
Subjective state of mood (VAS)
There was an effect of time with an increase in hunger
scores (F=44.7, df=3, 84, P<0.001) and a decrease in
happiness scores over time (F=4.65, df=3, 84, P=0.005) in
both groups with no statistically significant group difference. The mirtazapine group did however experience
significantly higher levels of dizziness compared to placebo
(main effect of group: F=11.12, df=1, 28, P=0.002) and

688

Psychopharmacology (2009) 203:685691

Table 1 Demographic characteristics and baseline ratings of mood for the two groups of volunteers

Gender
Age (years)
BDI
NART
EPQ: N
EPQ: E
EPQ: P
VAS: nausea
VAS: dizzy
VAS: hungry
VAS: anxious
VAS: sad
VAS: happy
VAS: alert

Mirtazapine group

Placebo group

6 M/9 F
25.27 (6.5)
1.20 (1.6)
124.40 (4.1)
4.2 (4.2)
15.6 (6.8)
2.7 (1.9)
3.5 (7.0)
8.1 (21.4)
15.3 (17.1)
10.7 (21.5)
10.4 (21.9)
54.8 (19.8)
59.7 (17.8)

6 M/9 F
22.80 (3.9)
1.33 (1.5)
125.45 (2.9)
5.0 (5.5)
12.3 (6.1)
1.6 (1.2)
0.9 (1.6)
0.8 (1.3)
18.1 (17.5)
2.6 (3.6)
2.4 (4.4)
58.9 (15.8)
64.5 (17.3)

Significance level (P value)

1.0
0.2
0.8
0.4
0.7
0.2
0.1
0.2
0.2
0.7
0.2
0.2
0.5
0.5

Values represent means (and standard deviations)

this was significantly different at 1 h (17.919.3 vs 3.3

6.5, P=0.009), 2 h (16.922.5 vs 1.52.7, P=0.014), and
4 h (8.711.3 vs 1.73.0, P=0.029) post drug administration. Mirtazapine also decreased ratings of alertness
(grouptime: F=5.44, df=3, 84, P=0.002) at 1 h (35.0
21.3 vs 53.518.1, P=0.016), 2 h (31.520.6 vs 55.5
18.7, P=0.002), and 4 h (34.723.5 vs 59.717.5, P=
0.002) post drug administration. There were no other
statistically significant differences in mood or subjective
state between the two groups (all P values>0.1).
Facial expression recognition
Accuracy Volunteers receiving mirtazapine were significantly worse at recognizing fearful facial expressions
compared to placebo (groupintensity: F=3.7, df=4,112,
P=0.008) as shown in Fig. 1. This remained significant
with the inclusion of subjective state ratings of alertness as
a covariate suggesting that it was not the result of the
sedative actions of mirtazapine (F(4,108)=4.8, P=0.001).
There were no significant group differences for any of the
other emotions or interactions between groupintensity (all
P values>0.05, see Fig. 1).
Reaction time Speed of correct responses to the different
emotional facial expressions used in this task, was
unaffected by drug (main effect of group: F=1.4, df=1,
28, P=0.24, groupemotion: F=0.4, df=6,168, P=0.9). A
separate sub-analysis also confirmed an absence of
difference in reaction time to identify fearful facial
expressions (t=0.73, df=28, P=0.47) suggesting that
the differences in classification are not confounded by
changes in speedaccuracy trade off.

Misclassifications Similarly, there was no difference in

terms of misclassifications as main group effect (F=0.7,
df=1, 28, P=0.4) and as groupemotion interaction (F=
1.0, df=6,168, P=0.4). Again considering fearful facial
expressions separately revealed no difference between the
two groups (t=0.28, df=28, P=0.78).

Emotion-potentiated startle
Volunteers showed the expected emotion-potentiated startle
effect, that is, greater responses during the negative picture
presentation (main effect of emotion: F=5.9, df=2, 34, P=
0.006; Fig. 2). There was no groupemotion interaction in
this analysis (F(2, 34)=1.8, P=0.2). However, the amplitude of blink responses was decreased in the mirtazapine
group across all conditions (main effect of group: F=4.5,
df=1, 17, P=0.049) and this remained statistically significant with inclusion of alertness as a covariate (F=19.0,
df=1, 16, P<0.001).
Emotional categorization and memory
Volunteers receiving mirtazapine showed a marginal decrease in reaction time to correctly classify adjectives
irrespective of emotion condition (main effect of group:
F=4.18, df=1, 28, P=0.05; groupemotion: F=0.24, df=
1, 28, P=0.63). While those receiving placebo recalled
approximately equal numbers of likeable and dislikeable
items in the free recall test (recall of likeabledislikeable
items, 0.1 0.5), those receiving mirtazapine showed
increased recall of positive vs negative items (recall of
likeabledislikeable items, 1.70.5). This difference was

Psychopharmacology (2009) 203:685691

100

% correct

Fig. 1 Facial expression recognition over the different intensity levels presented in the task
following mirtazapine (filled
circles) or placebo (open
circles). Values are meansSEM
for accuracy (% correct) in this
task averaged over 20% intensity bins. Asterisks represent statistical significance of
comparisons between groups
(**p<0.01)

689

Anger

100
75

75

50

50

50

25

25

25

Fear

**

Emotion Intensity (20% blocks)

Happy

100

Sad

100

75

75

75

50

50

50

25

25

25

statistically significant in the analysis of variance (group

emotion: F=5.3, df=1, 28, P=0.03).

Discussion
Our findings indicate that a single dose of mirtazapine
modulates the perception and memory of emotional stimuli
and eye-blink responses to loud noise in the startle
paradigm. These data support our contention that it is
possible in healthy volunteers to detect effects of single
doses of antidepressant drugs on emotional processing and
also that the pattern seen following mirtazapine has
interesting differences from those we have previously
reported with reboxetine and citalopram which appear

*
4000

Blink Amplitude (v)

100

75

100

3000

2000

1000

Mirtazapine

Disgust

Placebo

Fig. 2 Blink amplitudes in the emotion potentiated startle task with

neutral (white), positive (grey), and negative (black) picture presentation. Values represent mean blink amplitude + SEM, asterisks
represent statistical significance of the main effect of group (*p<0.05)

Surprised

consistent both with its clinical and pharmacological

profile.
In particular, the effect of a single dose of mirtazapine to
decrease the recognition of fearful expressions and to
attenuate overall startle responses suggests an early action
to diminish the processing of threatening stimuli. In
contrast, acute citalopram produces opposite effects, increasing the recognition of fearful facial expressions and
elevating startle responses in healthy volunteers (Harmer et
al 2003a, Browning et al 2007). We have hypothesized that
this increased processing of fear may relate to the same
underlying processes which lead to increased agitation and
anxiety in patients at the beginning of SSRI treatment. This
is supported by the observation that, like the clinical profile,
these effects on fear processing in our healthy volunteer
models reverse with repeated (7 days) administration of
citalopram (Harmer et al. 2004). Consistent with this, acute
administration of anxiolytic drugs such as diazepam also
decrease the recognition of negative facial expressions (at
15 mg (Blair and Curran 1999, Zangara et al. 2002), though
not at 5 mg (Murphy et al. 2008)) and decrease startle
reactivity across emotion conditions (Murphy et al 2008,
Baas et al. 2002). Such effects on baseline startle reactivity
are often thought to reflect differences in contextual anxiety
induced by the negative context of the startle experiment,
rather than fear produced by a specific cue (Grillon 2002).
As such these anxiolytic agents may reduce general and
anticipatory anxiety produced by the aversive nature of the
stimuli presented.
The clinical profile of mirtazapine is different from that
of SSRIs and by contrast early anxiolytic effects are
characteristic (Davis and Wilde, 1996). Mirtazapine has a
number of pharmacological actions that could be associated
with anxiolytic effects including 5-HT2A and 5-HT2C
receptor blockade (Bourin and Dhonnchadha, 2005).

690

Anxiolytic effects in animal and human studies have also

been reported with 5-HT3 receptor antagonists (Olivier et al
2000) and we found that a single dose of ondansetron
lowered overall startle responses in a similar way to that
produced by mirtazapine (Harmer et al 2006); however,
ondansetron did not alter the recognition of fearful facial
expressions suggesting a more restricted profile of action.
The potent H1 receptor antagonist effects of mirtazapine
could also be implicated in anxiolytic effects and the
antihistaminic compound, hydroxyzine, is licensed in the
UK for the short-term treatment of anxiety (Lader and
Scotto, 1998). Hence, the profile of decreased fearful face
recognition and startle responses seen here with mirtazapine is consistent with the idea that performance in
emotional processing tasks is sensitive to early anxiolytic
versus anxiogenic actions of antidepressant drug treatment
(Harmer 2008).
Mirtazapine also decreased ratings of alertness in the
current sample, which differs from the profile seen
previously with acute citalopram or reboxetine administration (Harmer et al. 2003a, 2003b). However, it is unlikely
that these subjective effects can explain the decreased
recognition of fearful faces and reduced startle responses
seen here. First, the effects of mirtazapine were very
specific and did not produce increased reaction times or
overall impairments in facial recognition. Indeed, results
from the emotional categorization task support improved
reaction times rather than general sedative actions. Second,
inclusion of subjective ratings of alertness taken just before
the emotional test battery as co-variates did not abolish the
statistical significance of the comparisons between groups
on these measures. As such it is likely that these effects of
mirtazapine on fear processing stem from central actions
rather than being mediated indirectly via subjective state
alterations.
In the current study, mirtazapine also decreased reaction
time to classify personality characteristics and enhanced
recall for likeable vs dislikeable words. Acute administration of reboxetine also decreased reaction time to selfreferent words but specifically to the likeable descriptions
and, like mirtazapine, increased the relative recall of the
positive self-referent words (Harmer et al 2003b). Therefore
the latter effect of mirtazapine could result from a
facilitation of NA neurotransmission, which presumably
would stem from its ability to block presynaptic 2adrenoceptors (Davis and Wilde, 1996). An increase in
the relative recall of positive self-referent words would be
consistent with an antidepressant action and indeed the
same effect is produced by citalopram after repeated
treatment (Harmer et al. 2004).
Further work is required, however, to compare and
contrast different antidepressant agents within the same
study to allow direct statistical comparison. Such studies

Psychopharmacology (2009) 203:685691

should have a sample size adequately powered to detect

small differences between agents. In particular, it is
noteworthy that mirtzapine did not facilitate the recognition
of happy facial expressions in the current study unlike the
effects we have seen previously with reboxetine (Harmer et
al. 2003b) and duloxetine (Harmer et al. 2008). Whether
this reflects a real difference between agents or a type 2
error requires further investigation in a larger study. The
current study and our previous investigations have all used
between subjects designs because of the problems of
repeated testing on the emotional tasks included in the
battery. While within-subjects designs allow greater statistical power and the elimination of individual differences
between groups, the replication of effects seen across
studies supports the conclusion that these are real outcomes
of antidepressant administration.
The findings from this study therefore support the notion
that it is possible to measure early effects of antidepressants
on emotional processing. However, it also suggests that the
precise pattern of effects revealed is influenced by the
pharmacological properties of the antidepressant compound
investigated and the duration of treatment. In contrast to
both acute reboxetine (Harmer et al. 2003b) and duloxetine
(Harmer et al. 2008), mirtazapine did not increase the
recognition of happy facial expression but decreased the
recognition of fear and also diminished overall startle
responses. While the latter effects could be indicative of
early anxiolytic action, the ability of mirtazapine to increase
the recall of likeable vs dislikeable personal descriptors is
also consistent with an antidepressant effect. Studies in
patients with depression and anxiety will be needed to
assess whether the different patterns of effects of antidepressants on emotional processing have implications for
the profile of treatment response and in particular whether it
is possible to match the changes in emotional processing
produced by particular drugs to the response of individual
patients to treatment.

Disclosure/conflict of interest This study was funded by the UK

MRC. DA is currently supported by the UK MRC. PJC has received
remuneration from serving as a member of advisory boards of Eli Lilly,
Wyeth, Servier, Xytis and DSM. CJH serves on the advisory panel of
P1vital and has received remuneration for consultancy from the
following companies: Lundbeck, Servier and P1vital.

References
Anderson NH (1968) Likableness ratings of 555 personality trait
words. J Pers Soc Psychol 9:272279
Baas JM, Grillon C, Bocker KB, Brack AA, Morgan CA III,
Kenemans JL, Verbaten MN (2002) Benzodiazepines have no
effect on fear-potentiated startle in humans. Psychopharmacology
(Berl) 161:233247

Psychopharmacology (2009) 203:685691

Blair RJ, Curran HV (1999) Selective impairment in the recognition of
anger induced by diazepam. Psychopharmacology (Berl)
147:335338
Bourin M, Dhonnchadha BA (2005) 5-HT2 receptors and anxiety.
Drug Dev Res 65:133140
Browning M, Reid C, Cowen PJ, Goodwin GM, Harmer CJ (2007) A
single dose of citalopram increases fear recognition in healthy
subjects. J Psychopharmacol 21(7):684690
Burghardt NS, Sullivan GM, McEwen BS, Gorman JM, LeDoux JE (2004)
The selective serotonin reuptake inhibitor citalopram increases fear
after acute treatment but reduces fear with chronic treatment: a
comparison with tianeptine. Biol Psychiatry 55:11711178
Davis R, Wilde MI (1996) Mirtazapine: a review of its pharmacology
and therapeutic potential in the management of major depression.
CNS Drugs 5:389402
Deakin JFW, Graeff FG (1991) 5-HT and mechanisms of defence. J
Psychopharmacology 5(4):305315
Ekman PFW, Friesen WV (1976) Pictures of facial affect. Consulting
Psychologists, Palo Alto
Grillon C (2002) Startle reactivity and anxiety disorders: aversive
conditioning, context, and neurobiology. Biol.Psychiatry 52:958975
Grillon C, Levenson J, Pine DS (2005) A single dose of the selective
serotonin reuptake inhibitor citalopram exacerbates anxiety in
humans: a fear potentiated startle study. Neuropsychopharmacology 32:225331
Harmer CJ (2008) Serotonin and emotional processing: does it help
explain antidepressant drug action? Neuropharmacology 55:10238
Harmer CJ, Bhagwagar Z, Perrett DI, Vollm BA, Cowen PJ, Goodwin
GM (2003a) Acute SSRI administration affects the processing of
social cues in healthy volunteers. Neuropsychopharmacology 28
(1):14852
Harmer CJ, Hill SA, Taylor MJ, Cowen PJ, Goodwin GM (2003b)
Toward a neuropsychological theory of antidepressant drug
action: increase in positive emotional bias after potentiation of
norepinephrine activity. Am J Psychiatry 160(5):9902
Harmer CJ, Shelley NC, Cowen PJ, Goodwin GM (2004) Increased
positive versus negative affective perception and memory in

691
healthy volunteers following selective serotonin and norepinephrine reuptake inhibition. Am J Psychiatry 161(7):125663
Harmer CJ, Reid CB, Ray MK, Goodwin GM, Cowen PJ (2006) 5HT
(3) antagonism abolishes the emotion potentiated startle effect in
humans. Psychopharmacology 186(1):1824
Harmer CJ, Heinzen J, O'Sullivan U, Ayres RA, Cowen PJ (2008)
Dissociable effects of acute antidepressant drug administration on
subjective and emotional processing measures in healthy volunteers. Psychopharmacology (Berl). 199:495502
Howell D (2002) Statistical Methods for Psychology. Belmont, CA
Lader MH, Scotto JC (1998) A multicentre double-blind comparison
of hydroxyzine, buspirone and placebo in patients with generalized anxiety disorder. Psychopharmacology 139:402406
Lang PJ, Bradley MM, Cuthbert BN (1998) International affective
picture system (IAPS): Technical Manual and Affective Ratings.
Gainesville, FL. The Center for Research in Psychophysiology.
University of Florida
Murphy SE, Downham C, Cowen PJ, Harmer CJ (2008) Direct effects
of diazepam on emotional processing in healthy volunteers.
Psychopharmacology (Berl). 199:50313
Nutt DJ (2002) The neuropharmacology of serotonin and noradrenaline in depression. Int Clin Psychopharmacol 17(Suppl 1):S112
Olivier B, Van Wijngaarden I, Soudijn W (2000) 5-HT3 receptor
antagonists and anxiety; a preclinical and clinical review. Eur
Neuropsychopharmacol 10:7795
Spitzer RLWJ, Gibbon M (2002) Structured Clinical Interview for the
DSM-IV. New York Psychiatric Institute, New York
Whale R, Clifford EM, Cowen PJ (2000) Does mirtazapine enhance
serotonergic neurotransmission in depressed patients? Psychopharmacology 148:325326
Young AW, Rowland D, Calder AJ, Etcoff NL (1997) Facial
expression megamix: tests of dimensional and category accounts
of emotion recognition. Cognition 63(3):271313
Zangara A, Blair RJ, Curran HV (2002) A comparison of the effects
of a beta-adrenergic blocker and a benzodiazepine upon the
recognition of human facial expressions. Psychopharmacology
(Berl) 163:3641