Ventrolateral prefrontal cortex and the effects
of task demand context on facial affect
appraisal in schizophrenia
David I. Leitman,1,2Daniel H. Wolf,1James Loughead,1Jeffrey N. Valdez,1Christian G. Kohler,1
Colleen Brensinger,3Mark A. Elliott,4Bruce I. Turetsky,1Raquel E. Gur,1,4,5and Ruben C. Gur1,4,5
1Department of Psychiatry-Neuropsychiatry Program, Brain Behavior Laboratory, University of Pennsylvania School of Medicine,
2Department of Psychology, Drexel University,3Department of Biostatisitics,4Department of Radiology, University of Pennsylvania
School of Medicine, and5Philadelphia Veterans Administration Medical Center, Philadelphia, PA, USA
Schizophrenia patients display impaired performance and brain activity during facial affect recognition. These impairments may
reflect stimulus-driven perceptual decrements and evaluative processing abnormalities. We differentiated these two processes by
contrasting responses to identical stimuli presented under different contexts. Seventeen healthy controls and 16 schizophrenia
patients performed an fMRI facial affect detection task. Subjects identified an affective target presented amongst foils of
differing emotions. We hypothesized that targeting affiliative emotions (happiness, sadness) would create a task demand context
distinct from that generated when targeting threat emotions (anger, fear). We compared affiliative foil stimuli within a congruent
affiliative context with identical stimuli presented in an incongruent threat context. Threat foils were analysed in the same
manner. Controls activated right orbitofrontal cortex (OFC)/ventrolateral prefrontal cortex (VLPFC) more to affiliative foils in
threat contexts than to identical stimuli within affiliative contexts. Patients displayed reduced OFC/VLPFC activation to all foils,
and no activation modulation by context. This lack of context modulation coincided with a 2-fold decrement in foil detection
efficiency. Task demands produce contextual effects during facial affective processing in regions activated during affect eval-
uation. In schizophrenia, reduced modulation of OFC/VLPFC by context coupled with reduced behavioural efficiency suggests
impaired ventral prefrontal control mechanisms that optimize affective appraisal.
Keywords: schizophrenia; social cognition; face; emotion; amygdala; ventrolateral prefrontal cortex (VLPFC); orbitofrontal
cortex (OFC); fMRI
Patients with schizophrenia have deficits in identifying affec-
tive facial intent, and these deficits relate to negative symp-
tom severity (Gur et al., 2007) as well as global outcome
(Brekke et al., 2005). These affective evaluation deficits
have been attributed to abnormalities in affective processing
neurocircuitry in limbic and frontotemporal regions (Gur
et al., 2007). However, abnormalities in basic visual process-
ing (Butler et al., 2001) or in the ability to integrate visual
information into visual objects (Doniger et al., 2002) may
also contribute to affective identification deficits in schizo-
phrenia (Leitman et al., 2005, 2008; Das et al., 2007; Fakra
et al., 2008). Studies using backward-masking paradigms
have suggested that schizophrenia patients have automatic
or implicit processing deficits in facial affect detection,
linked to subcortical dysfunction (Das et al., 2007). Most
fMRI studies examining deficits in facial affect processing
(Phan et al., 2002; Gur et al., 2002; Murphy et al., 2003;
Baas et al., 2004) employ standard affect identification par-
adigms. These paradigms make it difficult to asses whether
activation abnormalities in prefrontal cortex (PFC) executive
regions indeed reflect independent deficits in the controlled
evaluation of facial affect, or instead are purely stimulus
driven, reflecting a cascade of dysfunction stemming primar-
ily from basic sensory/perceptual disturbances.
A study by Gur et al. (2007) employed a hybrid (block and
event-related) paradigm in which subjects were asked to
identify a target emotion within a series of non-target foils
that were themselves the targets of ensuing blocks. This
design permits us to consider the impact of task demands
on affective appraisal, and ask whether dysfunction reported
in PFC and associated with evaluation is independently
present when sensory/perceptual processes are held constant.
We hypothesize that task instructions to identify the
emotions of fear and anger create an affective context for
the detection of threat (TH) within anger and fear target
Received 6 May 2009; Accepted 2 February 2010
Advance Access publication 8 March 2010
This work was supported in part by National Institute of Mental health (NIMH) grant MH060722
MH019112, and by National Alliance for Research on Schizophrenia and Depression NARSAD Young investi-
gator Awards (to D.I.L. and D.H.W.). The authors thank Dr Amy Pinkham and Kosha Ruparel for their
suggestions, and Dr Warren Bilker for his help with the statistical analysis.
Correspondence should be addressed to David I. Leitman, Department of Psychiatry, Neuropsychiatry
Program, Brain Behavior Laboratory, University of Pennsylvania, Gates Pavilion 10th floor, 3400 Spruce St,
Philadelphia, PA, USA. Email: email@example.com.
doi:10.1093/scan/nsq018SCAN (2011) 6,66^73
? The Author(2010).Publishedby OxfordUniversityPress.For Permissions,pleaseemail:firstname.lastname@example.org
blocks, while instructions to identify happiness and sadness
create a distinct affiliative (AF) context in their respective
blocks. While happiness and sadness differ in terms of pos-
itive and negative valence, they are both considered ‘AF’
emotions, as they serve to increase inter-personal empathy
(Miller and Eisenberg 1988; Eisenberg et al., 1989; Knutson
1996; Hess et al., 2000) and strengthen social bonds (Lewis
et al., 2008). Given that the same stimuli were used as foils
across blocks, contrasting presentations of TH foil stimuli
within TH blocks (context congruent) with identical stimuli
within AF blocks (context incongruent) could provide an
estimate of the contextual influence on prefrontal activity.
This effect of context should be unaffected by sensory aspects
of affect processing such as facial feature perception and
integration because the stimuli themselves are exactly
We hypothesized that context incongruities would lead to
increased activity within VLPFC and OFC. Ochsner and col-
leagues (2005; Wager et al., 2008) suggest that VLPFC is
central to the cognitive regulation of emotion and the affec-
tive appraisal of stimuli. Other studies have identified
VLPFC (Haxby et al., 2000; Mobbs et al., 2006; Guyer
et al., 2008) and OFC (Haxby et al., 2000) as involved in
affective evaluation and influenced by contextual framing
effects. Patients display deficits in explicit emotion process-
ing (van’t Wout et al., 2007) and fail to integrate contextual
cues when making social judgments (Green et al., 2007,
2008); therefore, we predicted that schizophrenia patients
would have reduced activation to incongruence in these
regions, relative to controls. This is the first study to examine
the effects of context on affective facial appraisal in
antecedents of affective facial recognition are held constant.
This analysis is based on data collected in a previously pub-
lished study (Gur et al., 2007). Therefore, the characteriza-
tion of the studied population, image acquisition parameters
and image analysis details are briefly summarized here.
The original sample included 16 patients (12 men), who met
DSM-IV criteria for schizophrenia or schizoaffective dis-
order, and 17 healthy controls (12 men). As described in
Table 1, the patients were somewhat older on average
(t2,31¼2.73, P¼0.011), and, as expected, had lower educa-
tion (t2,31¼3.72, P¼0.0008). However, they had compara-
ble parental education (t2,31¼1.95, P¼0.061). At the time
of imaging, all patients, except one unmedicated patient,
were on stable doses of antipsychotics: two received
first-generation (CPZequiv¼542?292/day) (Davis, 1976),
11 second-generation (OLZequiv¼18.2?2.8/day) and two
both [CPZequiv¼16.7/day, OLZequiv¼11.3/day (Kohler
et al., 2003)]. After complete description of the study,
written informed consent was obtained. Clinical ratings are
detailed in Table 1.
The face emotion identification task included four condi-
tions, presented in a counterbalanced order, each with a
specific target facial expression: happy, sad, anger or fear
(Figure 1). Each condition included four 90-s blocks of
emotion identification, separated by 24 s of rest during
which a scrambled face with a central cross-hair for fixa-
tion was displayed. Each 90-s identification block contained
8 target faces (e.g. 8 fear), 12 foil faces (e.g. 4 happy, sad
and 4 angry) and 10 neutral faces. Thus, a condition
included a total of 120 faces: 32 targets, 48 emotional
foils and 40 neutral foils in a pseudorandom sequence.
Faces appeared for 3 s and participants endorsed ‘target’
or ‘other’ using the two-button response pad. Abbreviated
response instructions remained visible throughout the task.
The same faces were cycled through the four conditions
serving as targets or foils depending on the condition.
Each condition (time series) lasted 8 min with total task
duration ?32 min.
Detailed image acquisition and processing methods were
described previously (Gur et al., 2007). Briefly, data were
acquired on a 4T gradient-echo (GE) Signa Scanner
(Milwaukee, WI, USA), employing a quadrature transmit
and receive head coil. Structural images consisted of a
sagittal T1-weighted localizer, followed by a T1-weighted
acquisition of the entire brain in the axial plane (24-cm
0.9375?0.9375?4 mm). This sequence was used for spatial
normalization to a standard atlas (Talairach, 1988) and for
anatomic overlays of the functional data. Functional imaging
matrix,resulting invoxel sizeof
Table 1 Subject demographic and clinical data
# Left handed
Illness duration (years)
Clinical ratings: SANS averaged
SAPS GS-Bizarre Behaviour
SAPS GS-Thought Disorder
N/A: not applicable.
Controlledcontextualeffects onfacialaffectprocessing SCAN (2011)67
was performed in the axial plane using a 16-slice, single-shot
GE echo-planar sequence (TR/TE¼1500/21 ms, FOV¼240
This sequence delivered a nominal voxel resolution of
3.75?3.75?5 mm. The 5-mm slice thickness was a
compromise to permit optimal visualization of the amygdala
with minimal sacrifice in brain coverage. Total slices per
volume were also limited by a 1.5-s TR that was selected
to provide two volume acquisitions per stimulus exposure
(3 s per face). The slices were acquired from the superior
cerebellum up through the frontal lobe. Inferiorly, this
corresponded to a level just below the inferior aspect of
the temporal lobes and superiorly to approximately the
level of the hand-motor area in the primary motor cortex.
Gradient echoplanar images can be degraded in the pres-
ence of non-uniform magnetic fields. Therefore, shimming
was performed manually in a region of interest (ROI) con-
taining the anterior medial temporal lobe (Webb and
Macovski, 1991). After shimming, pilot echoplanar images
were obtained and these images were visually inspected for
quality prior to fMRI acquisition. Following this inspection,
images were corrected for residual geometric distortion
(Jezzard and Balaban, 1995) based on a magnetic field map
acquired with a 1-min reference scan.
Analysis of Affective Foils. Analyses were limited to foil
stimuli within blocks. Target stimuli were excluded from
the analysis because they confounded the assessment of
task-driven context effects. Analysis of targets was presented
previously (Gur et al., 2007). Analysis of behaviour exam-
ined detection efficiency (integrating both accuracy and
reaction time) using Multivariate Analysis of Variance,
with factors for stimulus
Efficiency was defined as:
where Z reflects Z-transformed scores relative to the healthy
controls’ performance across all conditions (Gur et al.,
2001). This efficiency measure was calculated for the con-
trasts of interest described below. Selected fMRI contrasts
reflect blood oxygen level dependent (BOLD) signal change
in response to emotional face foils and cross-hair. For each
time series, in addition to the foil regressor of interest, neu-
tral foils and target faces were modelled but were not of
interest here. Our analysis focused on the effects of context
by contrasting identical stimuli under differing task
demands. This classification resulted in four new conditions:
(i) AF stimuli (happy and sad foils) in AF context?i.e. happy
or sad target conditions (AFAFcontext), (ii) AF stimuli in TH
context?i.e. fear and anger target conditions (AFTHcontext),
(iii) TH stimuli in TH conditions (THTHcontext) and (iv) TH
stimuli in AF conditions (THAFcontext). Note that this cate-
gorization resulted in 32 stimuli in context-congruent con-
context-incongruent conditions (THAFcontext, AFTHcontext).
The potential difficulty in contrasting numerically different
categories, however, is offset by the large number of stimuli
sampled. Prior study within our lab has illustrated that con-
trasts containing much lower numbers of stimuli than 32
yield stable estimates of condition activation. These data
were submitted to a mixed effects model, containing factors
for hemisphere [right hemisphere (RH), left hemisphere
(LH)], stimulus type [STIM (AF or TH)] and target condi-
tion type [CONTEXT (AFcontextorTHcontext)], as well as a
between groups factor for diagnosis (GROUP).
We hypothesized that activation to affective stimuli would
vary as a function of context and that such variation in
activation would be reduced in schizophrenia patients. We
thus were primarily interested in STIM?CONTEXT and
STIM?CONTEXT?GROUP interactions. Other complex
interactions were beyond our scope and hence not tested.
For pairwise contrasts, consistent with our hypothesis,
we tested AFAFcontext vs AFTHcontext and THTHcontext vs
THAFcontext within both hemispheres, and within and
type, contextand group.
Efficiency ¼ Z
Fig. 1 Experimental paradigm. (A) The experimental paradigm. In this figure,
fear is the target. (B) Experimental comparisons: foils of each block were classified
into two categories: happy and sad face foils comprised the AF category while
fear and anger comprised the TH category. We examined each stimulus type
within each context. This yielded four conditions: AFAFcontext, AFTHcontext, THTHcontext,
68 SCAN (2011) D.I.Leitmanetal.
between groups. This analysis was conducted offline using
percent signal change data extracted from our a priori ROIs
We first constructed a VLPFC/OFC structural ROI (com-
prising Brodmann’s areas 11 and 47 bilaterally) from the
Wake forest university pickatlas (Maldjian et al., 2003).
This ROI was further constrained by a functional mask of
areas showing robust activation (P<0.00005 uncorrected) to
all foil stimuli vs cross-hair across both groups. The resulting
ROIs contained 986 and 586 voxels (2?2?2mm voxel
dimensions) for RH and LH, respectively (Figure 2). Given
that our estimate of activation (see below) incorporates the
number of activated voxels, we decided to examine laterality
differences only when they significantly interacted with
diagnosis. An exploratory whole brain voxelwise analysis
was also conducted to examine contrast activation that
occurred outside our ROIs.
Subject-level time series statistical analysis was carried out
using functional magnetic resonance imaging of the brain
(FMRIB)’s improved linear model with local autocorrelation
within-subject fixed effects analysis across all four blocks
was then conducted for each subject. The resulting
single-subject contrast estimates were then submitted to a
third-level between-subjects (group) analysis employing
FMRIB’s local analysis of mixed effects (Beckmann et al.,
2003), which models inter-session or inter-subject random
effects components of the mixed-effects variance using
Markov chain Monte Carlo sampling to estimate the true
random effects variance and degrees of freedom at each
voxel (Woolrich et al., 2004). Statistical significance was
based on both voxel height and spatial extent in the whole
brain, using Analysis of Functional NeuroImages AlphaSim
to correct for multiple comparisons by Monte–Carlo simu-
lation (10 000 iterations, voxel height threshold P<0.01
uncorrected, cluster probability P<0.01). This whole-brain
correction required a minimum cluster size of 294 2?2?2
Finally, for our measure of activation, we used ‘energy’
which takes into account both the magnitude and spatial
extent of the activation (Gur et al., 2007). This index is
Energy=mean BOLD percent signal change
?number of voxels in which percent
signal change was greater than zero.
Repeating the analysis using mean percent signal change or
spatial extent separately yielded comparable results to those
obtained using energy. Statistical analyses used a two-tailed
alpha criterion of P<0.05, except where noted otherwise.
Statisticalanalyses were conducted
Analysis Software (Gary, Indiana).
A multivariate analysis of efficiency, which incorporates
both the accuracy and speed of affective foil detection, indi-
cated that patients were overall less efficient than healthy
subjects (GROUP: F1,31¼5.8, P¼0.022) (Figure 2, see
Supplementary Table 1 for accuracy and reaction time
data). Across groups, efficiency was reduced for all stimuli
P¼0.002), but no overall difference for stimulus type was
observed (P¼0.83). Critically, a two-way interaction
between STIM and CONTEXT was observed (F1,31¼30.4,
P<0.0001), indicating that activation to foil stimuli varied
as a function of both stimulus type and context. However,
neither GROUP?STIM nor GROUP?CONTEXT interac-
tions were observed (all P>0.13). A three-way interaction of
An examination of our a priori contrasts of interest, how-
ever, revealed the following: when contrasting identification
performance for AF foils in an incongruent TH context
(AFTHcontext) vs identical AF foils in a congruent AF
efficiency nearly twice the magnitude of that seen in
controls. [Controls: 0.70?0.19; Schizophrenia: 1.30?0.27
(F1,31¼3.2, P<0.04 one-tailed)]. No such efficiency differ-
ences were observed in contrasting TH foils presented in
incongruent AF contexts (THAFcontext) as compared with
was ata trend level
showeda decrement in
Fig. 2 Affective foil behavioural efficiency across task contexts. Performance effi-
ciency in controls (black traces) vs schizophrenia patients (grey traces). Solid lines
denote AF foils and dashed lines TH foils. Overall patients are less efficient in the
detection of affective foils. Asterisks indicate both healthy subjects and patients (solid
lines) display reduced efficiency in detecting AF foils presented in TH conditions as
compared with identical foils in AF conditions. Double asterisks indicate efficiency
decrement is approximately twice as large in patients than in controls.
Controlledcontextualeffects onfacialaffectprocessing SCAN (2011) 69
identical TH foils in a congruent TH context (THTHcontext)
P<0.0001) with patient activation significantly lower than
healthy controls across all affective foil stimuli (GROUP:
F1,31¼6.9, P<0. 01; Figure 3).
Across groups and conditions, activation was higher in
RH (HEM: F1,31¼46.2, P<0.0001). However, given that
our ROI size varied across hemispheres main effects of hemi-
sphere are difficult to interpret. On the other hand, schizo-
phrenia patients did display significantly reduced RH
laterality in activation as compared with healthy controls
No significant main effects for either STIM (P¼0.27) or
our modelfit was significant(?35¼288.6,
were observed,indicating that
activation did not vary as a function of the stimulus type
or context alone.
Consistent with our a priori hypothesis, we found
a STIM?CONTEXT interaction (F1,31¼28.2, P<0.0001),
P¼0.027) and an HEM?STIM?CONTEXT?GROUP
revealed that activation to foil stimuli varied as a function
of both context and hemisphere, and that this variation
differed for patients and controls. Specifically, within RH
(but not LH), healthy subjects’ activation to AF foils in
an incongruent TH context (AFTHcontext) was higher than
activation to identical AF foils in a congruent AF context
observedno significant difference
In contrast to controls, patients’ activation to stimuli
displayed no significant modulation by context within
either hemisphere (P>0.11), explaining the significant
STIM?CONTEXT?GROUP interaction. Finally, whereas
in RH patient activation was lower than controls for all foil
conditions (all P<0.01) except for AFAFcontext(P>0.14), for
LH no significant differences were seen in any of the four
conditions (all P>0.1).
Our previously published analysis of target emotions (Gur
et al., 2007) revealed correlations of flat affect severity and
BOLD responses. We therefore examined correlations
between negative symptom severity and indices of the
impact of context on foil activation defined as AFincong:
THTHcontext. within RH. Negative symptom severity was
measured on the scale for the assessment of negative symp-
toms (SANS) (Andreasen, 1984). We found a significant
correlation between global scores of affective flattening and
RH THincong (rs¼?0.52 P¼0.04). No other correlations
were significant. There were no significant correlations
between anti-psychotic medication dose or subject age and
behaviour or activation values (all P>0.31).
Psychologists and cognitive neuroscientists have often found
it useful to fractionate cognitive processes into sensory-
perceptual components and evaluative–executive compo-
nents. While most cognitive processes, such as affective
appraisal, undoubtedly result from the interaction of both
components, isolating the neural underpinnings of these
components can prove challenging in neuroimaging experi-
ments. For clinical neuroscientists who wish to identify the
locus of a specific neurocognitive abnormality, examining
processes such as facial affect in terms of evaluative and
executive processing vs sensory and perceptual processing
can be especially informative. Within schizophrenia, a lead-
inghypothesisattributing neurocognitivedeficits to
Fig. 3 Affective foil activation across task contexts. In the top panel, black and grey
traces contrast fMRI activation (energy) in healthy subjects and schizophrenia, respec-
tively. Solid lines denote AF foils and dashed lines TH foils. Within LH no significant
group?stimulus?context effects were observed. However, within RH the
group?stimulus?context interaction was significant. Asterisk indicates healthy
controls in RH, AF foils in the TH context elicited significantly greater activation in
VLPFC–OFC than these same stimuli in the AF context . This modulation was
significantly reduced in schizophrenia patients. Also note the overall reduced RH
response to all foils in patients relative to controls, independent of context.
70SCAN (2011) D.I.Leitmanetal.
dopaminergic-based hypofrontality favours evaluative and
executive explanations (Weinberger and Berman 1988;
Carter et al., 1998; O’Reilly et al., 2002; Bach et al., 2008;
Phillips et al., 2008). In contrast, glutamatergic (Javitt, 1996)
or GABAergic (Lewis and Moghaddam, 2006) hypotheses
emphasize more widespread neural dysfunction that also
encompasses basic sensation and perception processes.
Studies exploiting thehigh
event-related potentials (ERP) have indicated that facial
affect perception is associated with reductions in early visuo-
sensory components such as P1 and N1, while other studies
found reductions only for latter stage ‘integration’ compo-
nents such as the N170 and N250 (see Turetsky et al., 2007
for review). The presence of early ERP abnormalities in
schizophrenia suggests basic sensory deficits, without
ruling out the possibility that additional evaluative–executive
deficits also contribute to impairment in facial affect identi-
fication. Studies by Van’t Wout and colleagues (2007) show
that patients’ recognition of emotions such as fear is signifi-
cantly impaired in explicit but not implicit emotion process-
ing tasks. This disjunction suggests that task demand and
affective evaluation may reflect impairment beyond pure
sensory encoding of faces and their expressions. Similarly,
studies by Green and colleagues (2007, 2008) examining gaze
direction patterns have suggested that patients have difficulty
integrating contextual information when making emotional
and social judgments. Such affective evaluative impairment
may reflect a more general executive impairment in utilizing
context that has been long linked to frontal hypofunction
(e.g. MacDonald et al., 2005).
The low temporal resolution of fMRI can make it difficult
to disentangle relative contributions of sensory and executive
processing abnormalities to affective appraisal. However, the
current study’s hybrid (block and event-related) design
afforded us the opportunity to examine evaluation aspects
of face processing while holding sensory and perceptual
effects (stimulus characteristics) constant. We compared
identical faces that were foils in an emotion identification
experiment under differing task demand contexts, hypothe-
sizing that blocked trials in which subjects were asked to
detect happy and sad emotions would form an AFcontext,
while anger and fear target blocks would form a TH context.
Prior research into the effects of context on affective evalu-
ation had implicated VLPFC and OFC brain regions (Haxby
et al., 2000; Mobbs et al., 2006; Guyer et al., 2008), hence we
focused our analysis on this system.
Our finding of reduced ventral PFC activation to foils in
schizophrenia is consistent with prior studies of affective
appraisal, and more general findings of hypofrontality in
the illness. Consistent with our hypothesis, we found that
incongruent AF foils in the context of TH conditions pro-
duced less efficient behavioural responses and also elicited
greater bilateral activation in VLPFC–OFC compared with
their identical context-congruent counterparts. These find-
ing suggest that task demands of TH detection exert
Within RH, this context modulation was present in
healthy subjects butnot
Behaviourally, patients showed a nearly 2-fold greater reduc-
tionin efficiencythan healthy
AFTHcontext?AFTHcontextcontrast. Together, these behaviour
and imaging findings illustrate abnormal evaluative process-
ing deficits that are likely not directly attributable to sensory
integration deficits. Both behavioural and activation effects
were not related to medication dosage, indicating no direct
role of antipsychotic medication on the observed context
effects. Finally, in the patient group, reduced modulation
of VLPFC–OFC activation to TH foils by incongruent vs
congruent context was associated with greater affective flat-
tening. Although no overall group differences were seen in
this TH foil contrast, the symptom correlation suggests that
patients with flat affect may be less likely to effectively
employ contextual cues when appraising threatening facial
stimuli under AF conditions.
A post hoc whole brain analysis (Supplementary Figure 1)
of activation to AF foils in an incongruent TH context
(AFTHcontext) vs activation to identical AF foils in a congru-
ent AF context (AFAFcontext) revealed only two clusters that
reached our statistical significance criteria. These clusters
substantially overlapped with our a priori VLPFC/OFC
ROI, yet the activation also extended to more dorsal aspects
of PFC. In contrast, patients displayed only slight and sub-
threshold activation clusters in this contrast. No significant
difference in activation was observed in the THAFcontextvs
THTHcontextcontrast within either group.
This study highlights the interpretative limitations of
block designs, which are commonly employed in examining
facial affect in clinical populations. Blocks with different task
demands may induce context effects that alter response to
otherwise identical stimuli, and group differences in block
activation could reflect either context or stimulus effects.
block-design tasks could reflect higher level deficits in con-
text processing, while typically being interpreted as differ-
ences in response to stimulus features.
Our analysis of affective foils indicated that patients had
reduced activation within our ROI to facial stimuli in
general. This is consistent with prior work suggesting that
schizophrenia patients have core deficits in face perception
that extend beyond affective appraisal (Hooker and Park
2002; Leitman et al., 2008). The presence of these deficits
in our sample indicate that despite our comparison of
identical stimuli under differing contexts, we cannot com-
pletely rule out the possibility that the absence of context
effects on the appraisal of AF stimuli seen in patients reflects
an interaction between stimulus-driven sensory dysfunction
studies directly accounting for differences in sensory–
perceptual processing will be needed to settle this question
contextualeffects on affective evaluation.
in schizophrenia patients.
subjects within the
in schizophrenia during
Controlledcontextualeffects onfacialaffectprocessingSCAN (2011) 71
There are several limitations to our study. Task demands
in the current study likely produce only weak contextual
effects. Stronger contextual effects, such as those imposed
in ‘correspondence bias’ and contextual framing paradigms,
may induce even more robust changes in VLPFC/OFC.
No significant difference within this ROI was found for
incongruent TH foils in AF blocks vs their identical congru-
TH-detection task demands create stronger contextual
effects than those produced by task demands emphasizing
affiliation. It is also possible that anger and fear are more
closely aligned dimensionally in terms of TH than happiness
and sadness are in terms of affiliation. Future studies should
look at each emotion separately to explore emotion-specific
contextual effects. Functional connectivity analysis could
also help examine how context modulates interactions
between frontal evaluative regions, amygdala and other
neural nodes in the affective appraisal circuit. We did not
directly assess subjective emotional responses to face stimuli,
so we cannot rule out the possibility that patients experi-
ences of stimuli as AF or threatening may vary somewhat
from control subjects. We think this is unlikely to explain
our context results, as prior studies indicate that schizophre-
nia patients have qualitatively similar subjective emotional
responses to laboratory emotional stimuli (Kring et al., 1993,
1999; Kring and Neale, 1996); however, future studies should
directly assess the afiliative vs threatening judgments in
In daily life, affective appraisal takes place within situa-
tional contexts. Such contextual effects substantially shape
memory encoding and recall, in some cases dramatically
(Loftus and Pickrell, 1995; Loftus and Mazzoni, 1998).
Context can also alter the perceptual threshold of stimuli,
rendering detectable previously subthreshold stimuli (Cox
et al., 2004; Bar et al., 2006). Such contextual information
impacts directly on affect appraisal through executive con-
trol linked to VLPC–OFC (Haxby et al., 2000; Adolphs,
2002). Executive processing may facilitate affective appraisal
by improving the efficiency and accuracy of TH prediction,
for example, by constraining search within memory systems
(Sahakyan and Kelley, 2002; Mobbs et al., 2006). Our study
documented task-driven contextual effects on VLPFC–OFC
processing of facial affect that were reduced in patients with
schizophrenia. This suggests that patients may have difficulty
utilizing prefrontal control mechanisms that optimize affec-
Supplementary data are available at SCAN online.
Conflict of interest
Adolphs, R. (2002). Neural systems for recognizing emotion. Current
Opinion in Neurobiology, 12, 169–77.
Andreasen, N.C. (1984). The Scale for the Assessment of Negative Symptoms
(SANS). Iowa City: The University of Iowa.
Baas, D., Aleman, A., Kahn, R.S. (2004). Lateralization of amygdala activa-
tion: a systematic review of functional neuroimaging studies. Brain
Research Brain Research Reviews, 45, 96–103.
Bach, M.E., Simpson, E.H., Kahn, L., Marshall, J.J., Kandel, E.R.,
Kellendonk, C. (2008). Transient and selective overexpression of D2
receptors in the striatum causes persistent deficits in conditional associa-
tive learning. Proceedings of the National Academy of Science of the United
States of America, 105, 16027–32.
Bar, M., Kassam, K.S., Ghuman, A.S., et al. (2006). Top-down facilitation of
visual recognition. Proceedings of the National Academy of Science of the
United States of America, 103, 449–54.
Beckmann, C.F., Jenkinson, M., Smith, S.M. (2003). General multilevel
linear modeling for group analysis in FMRI. NeuroImage, 20, 1052–1063.
Brekke, J., Kay, D.D., Lee, K.S., Green, M.F. (2005). Biosocial pathways
to functional outcome in schizophrenia. Schizophrenia Research, 80,
Butler, P.D., Schechter, I., Zemon, V., et al. (2001). Dysfunction of early-
stage visual processing in schizophrenia. American Journal of Psychiatry,
Carter, C.S., Perlstein, W., Ganguli, R., Brar, J., Mintun, M., Cohen, J.D.
(1998). Functional hypofrontality and working memory dysfunction in
schizophrenia. American Journal of Psychiatry, 155, 1285–7.
Cox, D., Meyers, E., Sinha, P. (2004). Contextually evoked object-specific
responses in human visual cortex. Science, 304, 115–7.
Das, P., Kemp, A.H., Flynn, G., et al. (2007). Functional disconnections in
the direct and indirect amygdala pathways for fear processing in schizo-
phrenia. Schizophrenia Research, 90, 284–94.
Davis, J.M. (1976). Comparative doses and costs of antipsychotic medica-
tion. Archives of General Psychiatry, 33, 858–61.
Doniger, G.M., Foxe, J.J., Murray, M.M., Higgins, B.A., Javitt, D.C. (2002).
interaction in schizophrenia. Archives of General Psychiatry, 59,
Eisenberg, N., Fabes, R.A., Schaller, M., Miller, P.A. (1989). Sympathy and
personal distress: development, gender differences, and interrelations of
indexes. New Dir Child Dev, 107–26.
Fakra, E., Salgado-Pineda, P., Delaveau, P., Hariri, A.R., Blin, O. (2008).
Neural bases of different cognitive strategies for facial affect processing in
schizophrenia. Schizophrenia Research, 100, 191–205.
Green, M.J., Waldron, J.H., Coltheart, M. (2007). Emotional context
processing is impaired in schizophrenia. Cognitive Neuropsychiatry, 12,
Green, M.J., Waldron, J.H., Simpson, I., Coltheart, M. (2008). Visual pro-
cessing of social context during mental state perception in schizophrenia.
Journal of Psychiatry and Neuroscience, 33, 34–42.
Gur, R.C., Ragland, D., Moberg, P.J., et al. (2001). Computerized neuro-
cognitive scanning: I. Methodology and validation in healthy people.
Neuropsychopharmacology, 25, 766–76.
Gur, R.C., Ragland, D., Moberg, P.J., et al. (2001). Computerized
Neuropsychopharmacology, 25, 777–88.
Gur, R.E., McGrath, C., Chan, R.M., et al. (2002). An fMRI study of facial
emotion processing in patients with schizophrenia. American Journal of
Psychiatry, 159, 1992–9.
Gur, R.E., Loughead, J., Kohler, C.G., et al. (2007). Limbic activation asso-
ciated with misidentification of fearful faces and flat affect in schizophre-
nia. Archives of General Psychiatry, 64, 1356–66.
Guyer, A.E., Lau, J.Y., McClure-Ton, E.B., et al. (2008). Amygdala and
ventrolateral prefrontal cortex function during anticipated peer evalua-
tion in pediatric social anxiety. Archives of General Psychiatry, 65,
and dorsal/ventral stream
The profileof schizophrenia.
72SCAN (2011) D.I.Leitmanetal.
Haxby, J.V., Hoffman, E.A., Gobbini, M.I. (2000). The distributed human Download full-text
neural system for face perception. Trends in Cognitive Science, 4, 223–33.
Hess, U., Blairy, S., Kleck, R. (2000). The infuence of expression intensity,
gender, and ethnicity on judgments of dominance and affliation. Journal
of Nonverbal Behavior, 24, 265–83.
Hooker, C., Park, S. (2002). Emotion processing and its relationship to
social functioning in schizophrenia patients. Psychiatry Research, 112,
Javitt, D.C. (1996). Glutamate receptors and schizophrenia: opportunities
and caveats. Molecular Psychiatry, 1, 16–17.
Jezzard, P., Balaban, R.S. (1995). Correction for geometric distortion in
echo planar images from B0 field variations. Magnetic Resonance in
Medicine, 34, 65–73.
Knutson, B. (1996). Facial expressions of emotion influence interpersonal
trait inferences. Journal of Nonverbal Behavior, 20, 165–82.
Kohler, C.G, Turner, T.H., Bilker, W.B., et al. (2003). Facial emotion recog-
nition in schizophrenia: intensity effects and error pattern. American
Journal of Psychiatry, 160, 1768–74.
Kring, A.M., Neale, J.M. (1996). Do schizophrenic patients show a
disjunctive relationship among expressive, experiential, and psychophy-
siological components of emotion? Journal of Abnormal Psychology, 105,
Kring, A.M., Kerr, S.L., Smith, D.A., Neale, J.M. (1993). Flat affect in schi-
zophrenia does not reflect diminished subjective experience of emotion.
Journal of Abnormal Psychology, 102, 507–17.
Kring, A.M., Kerr, S.L., Earnst, K.S. (1999). Schizophrenic patients show
facial reactions to emotional facial expressions. Psychophysiology, 36,
Leitman, D.I., Foxe, J.J., Butler, P.D., Saperstein, A., Revheim, N.,
Javitt, D.C. (2005). Sensory contributions to impaired prosodic proces-
sing in schizophrenia. Biological Psychiatry, 58, 56–61.
Leitman, D.I., Loughead, J., Wolf, D.H., Ruparel, K., Kohler, C.G.,
Elliott, M.A., Bilker, W.B., Gur, R.E., Gur, R.C. (2008). Abnormal super-
ior temporal connectivity during fear perception in schizophrenia.
Schizophrenia Bulletin, 34, 673–8.
Lewis, D.A., Moghaddam, B. (2006). Cognitive dysfunction in schizophre-
nia: convergence of gamma-aminobutyric acid and glutamate alterations.
Arch Neurology, 63, 1372–1376.
Lewis, M., Haviland-Jones, J.M., Barrett, L.F. (2008). Handbook of
Emotions3rd edn ednNew York: Guilford Press.
Loftus, E.E., Pickrell, J. (1995). The formation of false memories. Psychiatric
Annals, 25, 720–25.
Loftus, E.F., Mazzoni, G.L. (1998). Using imagination and personalized
suggestion to change people. Behavior Therapy, 29, 691–706.
MacDonald, A.W. 3rd, Carter, C.S., Kerns, et al. (2005). Specificity of pre-
frontal dysfunction and context processing deficits to schizophrenia in
never-medicated patients with first-episode psychosis. American Journal
of Psychiatry, 162, 475–84.
Maldjian, J.A., Laurienti, P.J., Kraft, R.A., Burdette, J.H. (2003). An auto-
mated method for neuroanatomic and cytoarchitectonic atlas-based
interrogation of fMRI data sets. Neuroimage, 19, 1233–9.
Miller, P.A., Eisenberg, N. (1988). The relation of empathy to aggressive and
externalizing/antisocial behavior. Psychol Bull, 103, 324–44.
Mobbs, D., Weiskopf, N., Lau, H.C., Featherstone, E., Dolan, R.J., Frith, C.D.
(2006). The Kuleshov Effect: the influence of contextual framing on emo-
tional attributions. Social Cognitive Affective Neuroscience, 1, 95–106.
Murphy, F.C., Nimmo-Smith, I., Lawrence, A.D. (2003). Functional neu-
roanatomy of emotions: a meta-analysis. Cognitive, Affective and
Behavioural Neuroscience, 3, 207–33.
Ochsner, K.N., Gross, J.J. (2005). The cognitive control of emotion. Trends
Cognitive Science, 9, 242–9.
O’Reilly, R.C., Noelle, D.C., Braver, T.S., Cohen, J.D. (2002). Prefrontal
cortex and dynamic categorization tasks: representational organization
and neuromodulatory control. Cerebral Cortex, 12, 246–57.
Phan, K.L., Wager, T., Taylor, S.F., Liberzon, I. (2002). Functional neuroa-
natomy of emotion: a meta-analysis of emotion activation studies in PET
and fMRI. Neuroimage, 16, 331–48.
Phillips, A.G, . Vacca, G., Ahn, S. (2008). A top-down perspective on dopa-
mine, motivation and memory. Pharmacology, Biochemistry, and
Behavior, 90, 236–49.
Sahakyan, L, Kelley, C.M. (2002). A contextual change account of the direc-
ted forgetting effect. Journal of Experimental Psychology-Learning Memory
and Cognition, 28, 1064–72.
Talairach, J.T., Tournoux, P. (1988). Co-Planar Steriotaxic Atlas of the
Human Brain, 3 Dimensional Proportional System: An approach to
Cerebral Imaging. New York: Thieme Medical Publishers.
Turetsky, B.I., Kohler, C.G., Indersmitten, T., Bhati, M.T., Charbonnier, D.,
Gur, R.C. (2007). Facial emotion recognition in schizophrenia: when and
why does it go awry? Schizophrenia Research, 94, 253–63.
Van’t Wout, M., Aleman, A., Kessels, R.P., Cahn, W., de Haan, E.H.,
Kahn, R.S. (2007). Exploring the nature of facial affect processing deficits
in schizophrenia. Psychiatry Research, 150, 227–35.
Wager, T.D., Davidson, M.L., Hughes, B.L., Lindquist, M.A., Ochsner, K.N.
(2008). Prefrontal-subcortical pathways mediating successful emotion
regulation. Neuron, 59, 1037–50.
Webb, P., Macovski, A. (1991). Rapid, fully automatic, arbitrary-volume
in vivo shimming. Magnetic Resonance in Medicine, 20, 113–22.
Weinberger, D.R., Berman, K.F. (1988). Speculation on the meaning of
Bulletin, 14, 157–68.
Woolrich, M.W., Ripley, B.D., Brady, M., Smith, S.M. (2001). Temporal
autocorrelation in univariate linear modeling of FMRI data. Neuroimage,
Woolrich, M.W., Behrens, T.E., Beckmann, C.F., Jenkinson, M., Smith, S.M.
(2004). Multilevel linear modelling for FMRI group analysis using
Bayesian inference. Neuroimage, 21, 1732–47.
Controlledcontextualeffects onfacialaffectprocessingSCAN (2011)73