Happy facial expression processing with different social interaction cues: An fMRI study of individuals with schizotypal personality traits

Progress in Neuro-Psychopharmacology & Biological Psychiatry 44 (2013) 108–117

Contents lists available at SciVerse ScienceDirect

Progress in Neuro-Psychopharmacology & Biological Psychiatry journal homepage: www.elsevier.com/locate/pnp

Happy facial expression processing with different social interaction cues: An fMRI study of individuals with schizotypal personality traits Jia Huang a, Yi Wang a, Zhen Jin b, Xin Di c, Tao Yang a, Ruben C. Gur d, Raquel E. Gur d, David H.K. Shum e, Eric F.C. Cheung f, Raymond C.K. Chan a, g,⁎ a

Neuropsychology and Applied Cognitive Neuroscience Laboratory, Key Laboratory of Mental Health, Institute of Psychology, Chinese Academy of Sciences, Beijing, China MRI Imaging Center, 306 Hospital, Beijing, China Department of Radiology, University of Medicine and Dentistry of New Jersey, Newark, USA d Department of Psychiatry, Perelman School of Medicine, University of Pennsylvania, Philadelphia, USA e Behavioural Basis of Health Research Program, Griffith Health Institute, Griffith University, Gold Coast, Australia f Castle Peak Hospital, Hong Kong Special Administrative Region, China g Magnetic Resonance Imaging Research Centre, Institute of Psychology, Chinese Academy of Sciences, Beijing, China b c

a r t i c l e

i n f o

Article history: Received 26 November 2012 Received in revised form 6 February 2013 Accepted 6 February 2013 Available online 13 February 2013 Keywords: Functional imaging Happy facial expression Schizotypal personality traits Social interaction cues

a b s t r a c t In daily life facial expressions change rapidly and the direction of change provides important clues about social interaction. The aim of conducting this study was to elucidate the dynamic happy facial expression processing with different social interaction cues in individuals with (n = 14) and without (n = 14) schizotypal personality disorder (SPD) traits. Using functional magnetic resonance imaging (fMRI), dynamic happy facial expression processing was examined by presenting video clips depicting happiness appearing and disappearing under happiness inducing (‘praise’) or reducing (‘blame’) interaction cues. The happiness appearing condition consistently elicited more brain activations than the happiness disappearing condition in the posterior cingulate bilaterally in all participants. Further analyses showed that the SPD group was less deactivated than the non-SPD group in the right anterior cingulate cortex in the happiness appearing–disappearing contrast. The SPD group deactivated more than the non-SPD group in the left posterior cingulate and right superior temporal gyrus in the praise–blame contrast. Moreover, the incongruence of cues and facial expression activated the frontal– thalamus–caudate–parietal network, which is involved in emotion recognition and conflict resolution. These results shed light on the neural basis of social interaction deficits in individuals with schizotypal personality traits. © 2013 Elsevier Inc. All rights reserved.

1. Introduction Central to successful social interaction is the understanding of and responding to social stimuli embedded in the environmental context. Impaired understanding and responding to social stimuli was found in patients with schizophrenia (Maat et al., 2012; Smith et al., 2012), especially those with a first episode of psychosis (Achim et al., 2012). As part of schizophrenia spectrum disorders, individuals with schizotypal personality disorder (SPD) display similar yet attenuated difficulties Abbreviations: fMRI, functional Magnetic Resonance Imaging; SPD, schizotypal personality disorder; SPQ, schizotypal personality questionnaire; EPI, echoplanar imaging; TR, repetition time; TE, echo time; FOV, field of view; MP-RAGE, magnetization-prepared rapid gradient-echo imaging; SPM, Statistical Parametric Mapping software; A, happiness appearing; D, happiness disappearing; P, praise context; B, blame context; ROI, region of interest; ACC, anterior cingulate cortex; PCC, posterior cingulate cortex; THA, thalamus; CN, caudate nucleus; IFG, inferior frontal gyrus; REST, Resting-State fMRI Data Analysis Toolkit; STG, superior temporal gyrus; BOLD, blood-oxygen-level-dependent; MNI, Montréal Neurological Institute. ⁎ Corresponding author at: Room 526, South Building, Institute of Psychology, Chinese Academy of Sciences, 16 Lincui Road, Beijing, China. Tel./fax: +86 10 64836274. E-mail address: [email protected] (R.C.K. Chan). 0278-5846/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.pnpbp.2013.02.004

(Abbott and Green, 2012; H.J. Li et al., 2012; Roddy et al., 2012; Shi et al., 2012; Zong et al., 2010). In particular, Chan et al. (2012b) showed that individuals with SPD traits have deficits of hedonic capacity both physically and socially as compared to healthy controls. The reduced hedonic capacity is related to problems in the temporal experience of pleasure (Chan et al., 2012a). One of the social stimuli that provide pleasure information is happy facial expression (Wittfoth et al., 2010). Elucidating the neural mechanism of happy facial expression processing in the context of different social interaction cues in individuals with SPD may advance understanding of social cognition in schizophrenia spectrum disorders. In this study, we investigated the neural responses of individuals with and without SPD traits when they process dynamic happy facial expressions under different social interaction cues. Although facial displays of emotion are dynamic in nature, most studies have primarily examined emotion identification or recognition through a series of static and categorized facial expression photographs (Gur et al., 2007; Kohler et al., 2010; Lee et al., 2010; Loughead et al., 2008). The use of static emotional expressions may undermine the ecological validity of the emotion-processing tasks in these studies (McDonald et al., 2011; Platt et al., 2010; Sato and Yoshikawa, 2007b).

J. Huang et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 44 (2013) 108–117

Recent studies have shown that dynamic displays of emotion have more expressivity compared to that of static emotion stimuli, as indicated by electromyographic (Rymarczyk et al., 2011) and behavioral data (Sato and Yoshikawa, 2007a). The underlying mechanism of this facilitation relates to enhanced activation in visual areas (Recio et al., 2011), and to the emotion properties of dynamic facial expressions (Fujimura and Suzuki, 2010). Notably, static and dynamic facial expressions may differ in their neural substrates (Johnston et al., 2010; Pitcher et al., 2011) and dynamic information exchange may occur in multi-modal emotion communication (Regenbogen et al., 2012). Among emotions, positive affect such as happiness serves as a source of information for judging meanings (Hicks et al., 2010). Examining the neural correlates of happy facial expression processing may increase our understanding of the neural system associated with the pursuit of happiness. In a previous review, Kringelbach and Berridge (2009) suggested that the neural network associated with happiness pursuit is composed of the orbitofrontal cortex, cingulate cortex, insula, amygdala, ventral pallidum, nucleus accumbens, ventral tegmentum area and hypothalamus. A voxel-based meta-analysis has provided convergent evidence for the neural correlates of happiness including the right superior temporal gyrus, left anterior cingulate cortex, left insula and left thalamus (Vytal and Hamann, 2010). Moreover, recent findings also implicate the frontostriatal neural network in happiness experience and perception. For example, the experience of happiness was associated with activities in the right frontal lobes, nucleus accumbens/ventral striatum and prefrontal cortex (Sharpley and Bitsika, 2010). Additionally, the offset of happy and onset of angry expressions show common activations in the orbitofrontal cortex bilaterally, the left amygdala and the left insula, while the onset of happy and the offset of angry expressions show common activations in the left dorsal striatum (Muhlberger et al., 2010). The frontostriatal network associated with happiness provides insights into anhedonia, an inability to experience pleasure. Anhedonia can manifest in depression. Depressed individuals show a negative correlation between their depression severity and activation of the right fusiform gyri in response to happy facial expressions, while healthy individuals demonstrate linear increases in response to expressions of increasing happiness in the bilateral fusiform gyri and right putamen (Surguladze et al., 2005). Individuals with anhedonia in social interaction show less neural activity in facial expression discrimination regions such as the medial prefrontal cortex, the right superior temporal gyrus, and the left somatosensory cortex (Germine et al., 2011). Strong relationship was found among SPD traits, depressed mood and their poor social function (McCleery et al., 2012). Interpersonal aspects of SPD (particularly social anxiety) have also been associated with reduced accuracy on the facial expression recognition task (Abbott and Green, 2012). Moreover, individuals with SPD reported less pleasant affect compared with psychometrically defined controls and even stable schizophrenia patients (Cohen et al., 2012). Therefore, the current study aimed to investigate the neural activity of these at-risk individuals with social interaction difficulties when they process the dynamic happy facial expression. Real-life scenarios are usually more complicated than simply decoding the emotion of facial expressions in an artificial laboratory environment. Facial expressions are typically decoded with various social cues that involve reciprocity with sensory inputs. Thus, in the current study, we used a dyadic conversation consisting of a question and a dynamic facial expression as a response to provide social interaction cues. We attempted to investigate the dynamic happy facial expression processing within social interaction contexts in people with and without schizotypal personality traits. The neural activities underlying the congruence between expressed emotions and the social cues have received increased attention. When the self-expressed happy emotion is congruent with the observed others' emotion state, the medial orbitofrontal cortex and ventromedial prefrontal cortex, which have been associated with positive feelings and reward, are activated. In contrast, incongruent emotional states activate

109

the dorsolateral prefrontal cortex and the posterior superior temporal gyrus (Kuhn et al., 2011). Moreover, incongruence of emotion valence in audiovisual integration activates a frontal–cingulate–parietal network (Muller et al., 2011). However, social cues in previous studies õseldom involved the social interaction component, which is the most common situation in our daily lives. Dyadic conversation is the simplest form of social interaction (Huang et al., 2009a, 2009b). Very little is known about the neural responses of dynamic happy facial expression processing with social cues in the dyadic conversation context. In this study, we first examined the general neural responses to a dynamic happy facial expression with social cues in a dyadic conversation context. Without social cues, happiness appearing on a neutral face will signify pleasure while happiness disappearing from a happy face would signify displeasure or disappointment. Based upon previous studies on the neural basis of happiness (Kringelbach and Berridge, 2009; Muhlberger et al., 2010; Sharpley and Bitsika, 2010; Vytal and Hamann, 2010), we hypothesized that 1) happiness appearing and disappearing will have both overlapping and distinct neural network, 2) the social interaction cues would influence the activation of regions associated with happiness processing, 3) brain areas associated with emotion–cue conflict resolution would be activated when the dynamic happy facial expression is incongruent with the cues in dyadic conversations. We then examined whether individuals with SPD traits demonstrated a different neural mechanism when compared to individuals without SPD traits. Given the similarity between SPD and schizophrenia, we hypothesized that compared to the psychometrically defined control (non-SPD) group, the SPD group would show different neural activities when processing happiness and incongruence with different social interaction cues. 2. Materials and method 2.1. Pilot study We first conducted a pilot study to ensure the validity of the facial expressions and social cues in dyadic conversation. After providing written informed consent, 20 healthy volunteers (13 female; age M = 22.85, s.d. = 3.0) participated in the pilot study. They had to perform an emotion recognition task by selecting a label from ‘happy’, ‘angry’, ‘fearful’, ‘neutral’, and ‘sad’ below each facial expression. The happy and neutral facial expressions taken from 20 psychology graduates were subjected to emotion recognition. The happy and neutral images with the highest recognition accuracy in the pilot study were selected to be linearly morphed into video clips, which were used in the fMRI study. The subjects were also asked to rate the valence of the facial expressions and the valence of questions with ‘blame’ and ‘praise’ contents from 1 (most happy) to 5 (most angry). The questions with the most consistent ratings were selected for the dyadic conversations. 2.2. Subjects for fMRI study Before the fMRI study, a large scale psychological investigation was conducted in the university and the community to assess the SPD trait in Chinese adults. This investigation has been reported earlier (Chan et al., 2011, 2012b; Shi et al., 2012). In this investigation, the score range and mean score of SPQ in general adult population were [Range: 1–64; Mean: 26.46]. For the purpose of this study, we used the Chinese version of the Schizotypy Personality Questionnaire (SPQ) (Chen et al., 1997). According to the scoring criteria suggested by Raine (1991), subjects whose scores were in the top 10th percentile of the score distribution were considered as individuals with SPD traits and subjects whose scores were in the lowest 50th percentile were considered as without SPD traits, i.e. psychometrically defined non-SPD group. In total, 14 individuals with SPD traits (7 males; age M = 22.3, s.d.= 2.1; SPQ score mean: 45.79± 1.84) and 14 individuals without SPD traits (8 males; age: 20.7± 0.46 years; SPQ mean score:

110

J. Huang et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 44 (2013) 108–117

11.07 ± 1.25) participated in the fMRI study. We assessed IQ by the Chinese version of the Wechsler Adult Intelligence Scale—Revised (Gong, 1992) and found no significant difference between these two groups [SPD: 124.17± 7.6; non-SPD: 123.87± 8.74, t(25)= −0.904, n.s.]. However, the SPD group reported higher level of depression than non-SPD group as measured by the Beck Depression Inventory (Beck et al., 1961, Chinese version: Chan and Tsoi, 1984) [SPD: 15.31± 9.72; non-SPD: 2.73±1.94, t(26)=−4.914, pb 0.001]. None of them reported a history of psychiatric illness, neurological illness, or drug/alcohol dependence and none of them had a first-degree relative with psychiatric illness. All had normal or correct-to-normal vision and were right-handed (Annett, 1976). Prior to completing the task, the subjects provided written informed consent to a protocol that was approved by the ethics committee of the Institute of Psychology, Chinese Academy of Sciences. After the experiment, each subject received 50 RMB as a reward for their participation. 2.3. Design, stimuli and task The study employed a 3×2×2 factorial design: dynamic facial expression (happiness appearing vs. happiness disappearing vs. mosaic), social interaction cues (‘praise’ vs. ‘blame’ questions) and group (SPD vs. non-SPD). The facial emotional images were created from a set of color photographs of the Chinese people, which depict happy and neutral expressions. Using a computer algorithm, the prototype photographs were morphed to create video clips by a set of linear continuum of 50 facial images between two endpoints in the same person (100% neutral and 100% happy) (Kee et al., 2006). Linear changes secured the same physical features for the presentation of the dynamic facial expression. Happiness appearing expressions were made of facial expression images morphed from 100% neutral to 100% happiness. Happiness disappearing facial expressions were created in the reverse directions. One facial expression image was randomly selected and was cut into 100 small squares. These small squares were scrambled to form the mosaic images. Two different mosaic images were morphed into a dynamic mosaic clip. In the ‘mosaic’ condition, dynamic mosaic clips were presented as responders to the questions. The subjects were asked to press either of the right or left key when they saw a mosaic image. Social interactions were presented in the form of dyadic conversations in which A would ask a question and B would provide the above dynamic facial expression. We used the above video clips of happiness appearing and disappearing with 2 s length and a frame rate of 25 frames per second as dynamic facial expressions. For the questions in the dyadic conversation, 70 questions were generated freely in the ‘praise’ content and 70 questions were generated freely in the ‘blame’ content. Examples of ‘praise’ questions include ‘How could you write so quickly?’, ‘How could you dance so well?’ while examples of ‘blame’ questions include ‘How could you be so rude?’, ‘How could you be so careless?’ All the questions were rated in the pilot study described above and presented visually in the fMRI experiment. The fMRI study was conducted as a block design in two runs (8 min for each run). The questions in the first run were all ‘praise's and in the second run all the questions were ‘blame's. The three experimental conditions of happiness appearing, disappearing and dynamic mosaic were repeated four times per run resulting in 12 experimental blocks per run. Nine baseline blocks were inserted (fixation cross for 12 s) after each experimental block, while the order of experimental blocks was pseudo-randomized in the way of obtaining the highest signal-noise ratio. Each block contained five dyadic conversation trials with one question and one dynamic facial expression or mosaic for responding. Five trials in the same block were in the same condition of either baseline, happiness appearing or happiness disappearing. Fig. 1 illustrates the trial procedure in the fMRI experiment. In each trial, first a fixation cross was presented for 1000 ms to capture the attention of subjects, then one question would be presented visually for 2000 ms, and then the fixation cross disappeared followed by a

dynamic facial expression for 2500 ms. During this 2500 ms, the first and last images of dynamic facial expression clips were presented for 250 ms, respectively. The stimuli were presented on a black background and the questions were in gray. The subjects were instructed to read the questions carefully and told that they would be given a question memory task after the fMRI experiment. After trying to keep the questions in mind, they were asked to watch the corresponding dynamic facial expression carefully and judge the gender of the faces. This task was designed to secure the attention of the subjects on the dynamic facial expressions. 2.4. Image acquisition Functional and structural MRIs were performed with a Siemens 3T (SIEMENS 3T-Trio A Tim, Erlangen, Germany) MRI whole body scanner using a 32-channel head coil. Functional images were obtained using a T2-weighted single-shot gradient echoplanar imaging (EPI) sequence (TR: 2000, TE: 30, 90° flip angle, FOV: 210 mm, matrix: 64 × 64, voxel size: 3.3 × 3.3 × 4 mm 3). Each EPI volume contained 32 axial slices (thickness 4 mm, 0 mm gap), acquired in interleaved order, covering the whole brain. Each run contained 243 functional images. The first three slices of each run were discarded to allow for T1 equilibration. In addition, a high-resolution T1-weighted magnetization-prepared rapid gradient-echo imaging (MP-RAGE) 3D MRI sequence was obtained from each subject (TR: 2300 ms, TE: 3.01 ms, 9 flip angle, FOV: 240 × 256, matrix: 256 × 256, voxel size: 1 × 1 × 1 mm 3). 2.5. Imaging preprocessing and analyses Data were analyzed using the Statistical Parametric Mapping software (SPM8; Wellcome Department of Imaging Neuroscience, London, UK) implemented in Matlab 2009b (Mathworks Inc., Sherborn, MA, USA). Functional images were realigned by affine registration to correct for scan head motions. The mean functional image was subsequently co-registered to the 3D high resolution structural image of each subject. Each subject's structural image was normalized to T1 template provided by SPM, and the normalization parameters were then applied to all the functional images. Images were re-sampled at a 2 × 2 × 2 mm3 voxel size in the normalization step, and then spatially smoothed using an 8 mm full width at half maximum Gaussian Kernel. Each experimental condition was modeled using a boxcar reference vector convolved with a canonical hemodynamic response function. The generalized linear model implicitly implemented a high pass filter with 128 s cutoff. Parameter estimates were subsequently calculated for each voxel using weighted least squares to provide maximum likelihood estimates based on the non-sphericity assumption of the data to get identical and independently distributed error terms. Happiness appearing was abbreviated as ‘A’, happiness disappearing as ‘D’, mosaic as ‘M’, praise cues as ‘P’ and blame cues as ‘B’. ‘AP’ was happiness appearing with praise cues; ‘DP’ was happiness disappearing with praise cues; ‘AB’ was happiness appearing with blame cues and ‘DB’ was happiness disappearing with blame cues; ‘MP’ was mosaic with praise cues and ‘MB’ was mosaic with blame cues. For each subject, ten contrasts were computed. Main effects were calculated for happiness dynamic facial expression and social interaction cues by contrasts (AP +AB) − (MP + MB) for happiness appearing simple effect, contrasts (DP+ DB) − (MP + MB) for happiness disappearing simple effect, contrasts (AP + DP + MP) − (AB + DB + MB) for ‘praise’ cues simple effect and (AB+ DB+ MB) − (AP + DP + MP) for ‘blame’ cues simple effect. The simple effects that examined the difference between happiness appearing and happiness disappearing were t-contrasts such as (AP +AB) − (DP+ DB) and (DP +DB) − (AP+ AB). When the facial expression was happiness appearing with ‘praise’ social interaction cues and happiness disappearing with ‘blame’ social interaction cues,

J. Huang et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 44 (2013) 108–117

111

(mosaic)

Fig. 1. The illustration of the dynamic happy facial expression paradigm.

the condition was considered congruent. When the facial expression was happiness appearing with ‘blame’ social interaction cues and happiness disappearing with ‘praise’ social interaction cues, the condition was considered incongruent. The interaction between dynamic facial expression and social interaction contexts was calculated by contrasts (AP+DB)−(AB+DP) for congruent condition vs. incongruent condition and (AB+DP)−(AP+DB) for incongruent vs. congruent condition. These first-level individual contrast maps were fed into a second-level group analysis using an independent sample t-test. Resulting activation peaks were superimposed on standard high-resolution anatomical images. Two independent sample t tests were conducted between the SPD group and the non-SPD group. To further examine the performance of each group in each condition, activations that were associated with emotion processing and incongruence processing were selected. ROI of the right middle temporal gyrus was defined by the contrasts of (SPD > non-SPD) and (appearing >disappearing). ROI of the right anterior cingulate cortex (rACC) was defined by (non-SPD > SPD) and (appearing> disappearing). ROI of the left posterior cingulate cortex (PCC) was defined by (SPD > non-SPD) and (praise > blame). ROI of the thalamus (THA), the caudate nucleus (CN) and the left inferior frontal gyrus (IFG) were defined by the contrast of (SPD>non-SPD) and (incongruent >congruent). These ROIs were constructed by the Marsbar software (Brett et al., 2002) using a box with a peak activation center and 10 width. Correction for multiple comparisons was performed using Monte Carlo simulation. A corrected threshold of p b 0.05 (two-tailed) was derived from a combined threshold of p b 0.001 for each voxel and a cluster size of > 78 voxels was determined using the AlphaSim program in REST software (Parameters: single voxel pb 0.001, 1000 iterations, FWHM=8 mm, with gray matter mask (Song et al., 2011)). 3. Results 3.1. Pilot findings The 70 ‘praise’ questions were ranked in an ascending order from 1 to 5 and the first 60 were used in the formal experiment (mean rating ± S.D.: 1.72 ± 0.15). Ten questions were selected randomly from the first 60 ‘praise’ questions for the memory test after the formal experiment. Similarly, the 70 ‘blame’ questions were ranked in a descending order from 5 to 1 and the first 60 were used in the formal experiment (mean rating ± S.D.: 4.24 ± 0.27). Ten questions were selected randomly from the first 60 for the memory test after the formal experiment. A total of 119 facial expressions were reliably recognized, and happy and neutral faces of eight persons with the highest recognition

accuracy (mean ± S.D., happy: 0.94 ± 0.06, neutral: 0.85 ± 0.04) were selected for the formal experiment. 3.2. Behavioral results The performance of gender identification in the fMRI study was not significantly different (p> 0.05) in the first and second runs: SPD 0.94± 0.05, non-SPDs 0.95 ±0.04 (mean± S.D.); SPD 0.96± 0.03; non-SPDs 0.98± 0.05 (mean± S.D.). The memory test accuracy after the fMRI study was also not statistically significant (p> .05): SPD 0.75± 0.1; non-SPDs: 0.73± 0.07 (mean ± S.D.). 3.3. fMRI results 3.3.1. Happiness appearing and disappearing Combining the fMRI results of the two groups of subjects, observation of happiness appearing expressions elicited right cortical activation. Dynamic mosaic was used as a baseline and statistically significant activation (happiness appearing>mosaic) was detected in an activation cluster covering the right fusiform gyrus, the right occipital lobe, the right thalamus and the left cerebellum, posterior and anterior lobes (Table 1, Fig. 2 Panel A). Observation of happiness disappearing expressions vs. mosaic elicited activation in the right fusiform gyrus, the right middle frontal gyrus and the left fusiform gyrus (Table 1, Fig. 2 Panel B). The happiness appearing condition activated more brain regions than the happiness disappearing condition in the right and left posterior cingulate as well as the right occipital lobe (Table 1, Fig. 2, Panel C). Moreover, we found a significant interaction between group and facial expression in the right anterior cingulate cortex (x=8, y=42, z=10, k=81, t=4.7, pb 0.05, alphasim corrected; F(1,26)=17.963, pb 0.001). Specifically, the SPD group was less deactivated than the non-SPD group in the happiness disappearing condition (mean±S.E. SPD: −0.02± 0.12, non-SPD: −0.33±0.12, pb 0.05) and the SPD group tended to be more deactivated than the non-SPD group in the happiness appearing condition (mean±S.E. SPD: −0.2±0.12, non-SPD: −0.045±0.13, n.s.) (Fig. 3 Panel A). 3.3.2. Interaction between group and cues The ‘praise’ cues activated more regions than the ‘blame’ cues in the right hippocampus, the right and left insula, the right thalamus, the right superior and inferior parietal lobes, the left lentiform nucleus and the left middle frontal gyrus (Table 1). Significant interactions were found between group and cues in the left posterior cingulate cortex (x= −8. y= −34, z = 22, p b 0.05, alphasim corrected, k = 78, T = 5.83; F(1,26) = 7.822, p = 0.01) and the right temporal gyrus (x = 40, y = − 4, z = − 20, k = 107, T = 4.19, p b 0.05, alphasim

112

J. Huang et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 44 (2013) 108–117

Table 1 fMRI activation patterns in each condition. Contrast

k

x

y

z

T

Brain region

(AP + AB) − (MP + MB) [happiness appearing > mosaic]

2291 607 93 623 90 1640 387 210 11,505 9951 332 330 93 201 990 92 212 137 144

46 −40 −42 6 8 48 48 −44 28 34 10 62 42 −16 −30 −20 −26 24 44

−70 −80 −48 −70 −36 −72 36 −80 −58 22 −4 −20 −44 −8 22 −62 44 −46 −48

−10 −12 −22 8 2 −10 14 −10 −4 16 2 42 18 −2 6 44 36 44 38

9.51 6.35 5.14 4.89 4.37 8.45 5.17 4.95 7.02 6.44 5.26 5.18 5.06 4.86 4.63 4.57 4.22 4.19 4.02

R fusiform gyrus L cerebellum posterior lobe L cerebellum anterior lobe R occipital lobe R thalamus R fusiform gyrus R middle frontal gyrus L fusiform gyrus R parahippocampal gyrus R insula R thalamus R parietal lobe R insula L Lentiform Nucleus L Insula L parietal lobe L middle frontal gyrus R parietal lobe R inferior parietal lobe n.s.

1364 1905 85

26 −22 26

−30 −52 −76

38 14 −6

7.16 5.77 4.42

R posterior cingulate L posterior cingulate R occipital lobe n.s.

(DP + DB) − (MP + MB) [happiness disappearing >mosaic] (AP + DP + MP) − (AB + DB + MB) [Praise >blame]

(AB + DB + MB) − (AP + DP + MP) [blame > praise] (AP + AB) − (DP + DB) [happiness appearing > disappearing] (DP + DB) − (AP + AB) [happiness disappearing >appearing] (AP + DB) − (DP + AB) [Congruent >incongruent] (DP + AB) − (AP + DB) [incongruent > congruent]

n.s. 374 915 224 159 512 168 762 106 288 301 167 134 263 104 89 76 110 83

16 −2 58 −22 −8 18 10 −36 −6 −20 50 24 −44 −34 −34 0 40 20

54 −56 26 −38 −30 68 −2 −76 16 −88 −78 −20 24 −86 30 −96 20 32

48 6 32 76 14 28 76 −44 2 −26 26 −8 −20 42 54 12 −22 60

5.69 5.23 5.22 5.03 5.03 4.96 4.93 4.84 4.72 4.70 4.65 4.58 4.53 4.47 4.28 4.09 3.93 3.87

R superior frontal gyrus L cerebellum anterior lobe R middle frontal gyrus L parietal lobe L thalamus R superior frontal gyrus R medial frontal gyrus L cerebellum posterior lobe L caudate L cerebellum posterior lobe R middle temporal gyrus Sub-lobar L inferior frontal gyrus L superior occipital lobe L superior frontal gyrus L occipital lobe R inferior frontal gyrus R superior frontal gyrus

Note: Coordinates of the maximal point of activation and the associated T-values are shown in MNI spaces. L = Left; R= Right. The activations in these brain regions are significant at cluster level alphasim corrected p b 0.05; cluster threshold size k= 75. A: happiness appearing condition; P: the ‘praising’ cues, D: happiness disappearing condition; B: the ‘blaming’ cues, M: dynamic mosaic condition.

corrected, F(1,26) = 8.137, p b 0.01 ). We extracted the parameter estimates from these two regions to further investigate the significant interaction. In the left posterior cingulate cortex, significantly more deactivation was observed in the SPD group than the non-SPD group in the ‘blame’ cues (mean± S.E. SPD: −0.25 ±0.07; non-SPD: −0.02 ± 0.05, p b 0.05), but the SPD group was not significantly different from the non-SPD group in the ‘praise’ cues (mean±S.E. SPD: 0.08± 0.09; non-SPD: −0.1 ±0.12, p >0.05). In the right temporal gyrus, significant difference was found between the SPD and non-SPD groups in the ‘blame’ cues (mean ± S.E. non-SPD: 0.15 ± 0.11; SPD: − 0.03 ± 0.09, p b 0.01) (Fig. 3, Panel B). 3.3.3. Incongruent vs. congruent When facial expressions were incongruent with the questions in the social interaction context, more brain regions were activated, including the right superior frontal gyrus, the anterior and posterior lobes of the left cerebellum, the left parietal lobe, the left thalamus, the right medial frontal gyrus, the right middle temporal gyrus, the left inferior frontal gyrus, the left superior occipital lobe and the left

superior frontal gyrus (Table 1). Among these, three ROIs (left inferior frontal gyrus, thalamus and caudate nucleus) were defined and modified by a box (width = 10 mm) centered with peak coordinates. Among the ROIs, no significant group differences were found. Only the activation in the happiness disappearing with the “praise” cues showed a marginal group difference (non-SPD: 0.121 ± 0.047, SPD: − 0.014 ± 0.047, p = 0.052) (Fig. 4). 4. Discussion This study investigated neural activation during processing of happy facial expression with different social interaction cues, and explored the differential neural mechanisms of processing dynamic happy facial expression with different social interaction cues by comparing individuals with and without SPD traits. For the first aim, we found bilateral activation of the PCC during the contrast of happiness appearing and disappearing. More regions such as the right parahippocampal gyrus, the right insula, the right thalamus, bilateral parietal lobes and the left middle frontal gyrus were activated

J. Huang et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 44 (2013) 108–117

113

Fig. 2. A to C illustrated the axial view of whole brain activation pattern in each condition (A: happiness appearing > mosaic; B: happiness disappearing > mosaic; C: happiness appearing > disappearing). Color legend represents the T value in the activation clusters. All contrasts resulted from a random effect GLM. Cluster details in each contrast please see Table 1.

when contrasting the ‘praise’–‘blame’ cues. The thalamus, the caudate nucleus, the frontal and parietal lobes were activated in the contrast of incongruent and congruent conditions. For the second aim, we found that the SPD group showed less deactivation to the happiness disappearing faces than the non-SPD group in the rACC and the former group deactivated more than the latter group in the left PCC and rSTG with the ‘blame’ cues. These two groups did not differ in the contrast of incongruent and congruent conditions. 4.1. General neural mechanism of processing dynamic happy facial expression The imaging data collected in the current study appeared valid. According to the gender identification and question memory after the imaging task, participants had paid attention to the faces and questions in the task. For the imaging results, happiness appearing condition activated the fusiform gyrus, the occipital lobe and the thalamus in the right hemisphere while the happiness disappearing condition activated the fusiform gyrus bilaterally and the right middle frontal gyrus. Activation of the fusiform gyrus is associated with face processing (Lopez-Ibor et al., 2008; Natu et al., 2011) and right thalamus activation is associated with emotion processing (Domes et al., 2009; Nomura et al., 2003), especially reward processing (Helfinstein et al., 2012). In this study, participants deactivated to a greater extent in happiness disappearing than happiness appearing conditions at the posterior

cingulate cortex bilaterally (rPCC/lPCC). The PCC has received relatively less attention than the ACC in previous literature. The PCC was found to be involved in the representation of internally-focused information about others (Lieberman, 2007). Interestingly, our result found differential deactivation in one of the classic default mode network areas—bilateral PCC. Processing happiness disappearing facial expression about others inhibits the PCC more than processing happiness appearing facial expression about others. Though happy signal strength was linearly changed with time in both happiness appearing and disappearing conditions, the differential deactivation of PCC in two conditions indicated that the neural activities did not linearly change with time. This is consistent with the temporal characteristic of neural activities when processing dynamic facial expression as reported in a previous study (Muhlberger et al., 2010). Earlier studies showed that the fusiform cortex, the amygdala, the right inferior frontal gyrus, the right superior parietal lobe, the right thalamus, the right supplementary motor area, and the right occipital cortex were activated when comparing happy stimuli with baseline (Gur et al., 2002; Rahko et al., 2010). In this study, the ‘praise’ cues activated more regions than the ‘blame’ cues in regions such as the right parahippocampal gyrus, the right insula, the right thalamus and the right superior parietal lobe as well as subcortical areas. The right parahippocampal gyrus, the right insula, the right thalamus and the right parietal lobe are all regions associated with happy emotion processing in previous studies (Domes et al., 2009; Rahko et al.,

114

J. Huang et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 44 (2013) 108–117

Fig. 3. Activation patterns for the interactions with group in the task. Statistic maps (p b 0.05, alphasim corrected) are overlaid on sagittal sections from Colin27 MNI template. Error bars represent the mean of standard error. Panel A illustrated activation result of the significant interaction between group and dynamic facial expression [(AP + AB) − (DP + DB)] × [non-SPD − SPD] in right anterior cingulate cortex (rACC) (x = 8, y = 42, z = 10; F(1,26) = 17.963, p b 0.001). Panel B illustrated activation result of the significant interaction between group and social cues [(AP + DP + MP) − (AB + DB + MB)] × [SPD − non-SPD] in the left posterior cingulate cortex (left PCC) (x = −8, y = −34, z= 22; F(1,26) = 7.822, p = 0.01) and right temporal gyrus (rSTG) (x = 40, y= −4, z= −20; F(1,26) = 8.137, p b 0.01). (Note: *:p b 0.05; **:p b 0.01).

2010; Ramnani and Miall, 2003). When the happy facial expression was incongruent with the semantic content of the question, most of the frontal cortex was activated bilaterally. Conflict processing has its specific regions. Previous research has found that brain regions that monitor the conflict between emotion semantic content and emotion prosody comprise the medial prefrontal areas (Wittfoth et al., 2010) and regions responsible for emotion conflict were dissociable from those responsible for affect processing (Chiew and Braver, 2011). In this study, only regions associated with emotion processing were selected as ROIs in the current study. As a result, the inferior frontal gyrus was more likely to be responsible for dealing with the incongruent emotion information with different social cues. Previous research has shown that the inferior frontal gyrus is related to one's skills in understanding others' emotion state (Hooker et al., 2010). It would be more difficult to understand others' emotion state, especially when the other's reaction was contrary to one's expectation. This might be the reason that inferior frontal gyrus was more activated in the incongruent

condition than in the congruent condition. Thus, the inferior frontal gyrus was associated with understanding the dynamic facial expression in the conflict conversation context. 4.2. Differential neural mechanism between individuals with and without SPD traits Previous behavioral data has suggested that individuals with SPD have poor social function (McCleery et al., 2012), have subjective complaints about hedonic capacity (Yan et al., 2011) and might have problems in emotion expression (Dickey et al., 2011). For functional imaging data, preliminary findings on patients with schizophrenia showed a generally reduced neural activity when processing facial expressions (Lepage et al., 2011). For example, Lepage et al. (2011) found that the healthy group, rather than the group with schizophrenia, had increased activation in the anterior cingulate, the right parahippocampal gyrus and multiple visual areas. Moreover, previous research found

J. Huang et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 44 (2013) 108–117

115

Fig. 4. Activation patterns for the contrast between incongruent and congruent conditions. Three ROIs were extracted [left inferior frontal gyrus—IFG(−44 24–20); thalamus— THA(−8 −30 14) and caudate nucleus—CN(−6 16 2)]. AP: happiness appearing with praise cues; DP: happiness disappearing with praise cues; AB: happiness appearing with blame cues; DB: happiness disappearing with blame cues. Statistic maps (p b 0.05, alphasim corrected) are overlaid on sagittal sections from Colin27 MNI template. Error bars represent the mean of standard error.

that cingulate volumes were reduced in first episode schizophrenia (Rothlisberger et al., 2012). Our findings showed that individuals with SPD had a differentiated neural mechanism in inhibiting neural activities when processing happiness disappearing and the ‘blame’ cues. We found that individuals with SPD deactivated less than individuals without SPD in the right ACC in the happiness disappearing condition. The ACC is a neural region in the hedonic system (Kringelbach and Berridge, 2009). Less deactivation in the ACC in the happiness disappearing condition might suggest alteration of neural activities in the hedonic system of individuals with SPD traits. In addition, we also found that individuals with SPD traits deactivated more than individuals without SPD traits in the left PCC and the rSTG when they processed the ‘blame’ condition. Previous studies have suggested that the STG and the temporal–occipital lobe are responsible for dynamic human facial expression processing (Pitcher et al., 2011). The rSTG was also activated when subjects listened to a semantically and prosodically incongruent sentences and the left superior frontal gyrus and left superior temporal gyrus were activated in happily intoned sentences expressed in negative content (Wittfoth et al., 2010). In this study, ‘blame’ context represented the negative social cue. The rSTG was associated with the processing of dynamic happy facial expression with negative social cue. We thus inferred that the more deactivated rSTG in the ‘blame’ condition could provide a piece of evidence of neural sensitivity for the negative social interaction cues, which induced unhappiness in individuals with SPD traits. Deactivation in rSTG and left PCC is reasonable because the resting state of rSTG (R. Li et al., 2012) and the cingulate cortex (Z. Li et al., 2012) are activations. Deactivation in the current task could be a potential neural marker for individuals with social interaction difficulty. Based on the imaging results in the incongruent–congruent contrast, we selected three regions of interest to further clarify whether individuals with and without SPD would be different. The thalamus and the caudate nucleus are associated with empathy and emotion recognition (Kemp et al., 2012). Previous studies found that medial prefrontal cortex was associated with happy emotion processing (Keener et al., 2012) and right inferior parietal gyrus activated more

in emotion rating than gender rating (Sarkheil et al., 2012). Such frontal and parietal regions were not activated in the current contrast. Thus, thalamus, caudate and inferior frontal gyrus were considered as regions of interest (ROIs). In our study, these regions were consistently deactivated in conditions without happy emotion such as happiness disappearing with blame cues. If the facial expression or interaction cues contain happy emotion, the thalamus or caudate nucleus would be activated, otherwise it would be deactivated. Taken together, the differential neural activity in the rACC, left PCC, rSTG, thalamus and the caudate nucleus could be the neural mechanisms associated with the aberrant happy emotion processing with social interaction cues in individuals with SPD traits. 5. Limitations The main limitation of this study is that the block design could not capture the temporal characteristic of the onset and disappearance of happiness emotion, as found in a previous study (Muhlberger et al., 2010), because a block design tends to ignore the temporal change of BOLD signals. In future studies, a mixed block-event related design should be considered to avoid such limitation. The second limitation is that the praise and blame cues were not counterbalanced during the task. The difference between the two runs might confound a direct comparison between praise and blame cues, though this was not the main purpose of the current study. Another limitation is that only the happiness emotion was examined. This could be problematic when examining congruence and incongruence in the ‘praise’ or ‘blame’ interaction cues, because happiness appearing with the ‘blame’ cues could be understood as a way to avoid embarrassment rather than an inappropriate or incongruent facial expression when being blamed. Therefore, angry facial expressions should be further examined with the blame cues. One potentially informative approach for future research is to determine whether the threat-related dynamic facial expression change will be influenced by the social cues in dyadic conversations. We could further investigate whether the left

116

J. Huang et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 44 (2013) 108–117

hemisphere could still be dominant when the threat-related condition was incongruent with the social cues in dyadic conversations. Finally, it should be noted that SPD trait is associated with depression and autistic trait. Indeed several studies (e.g. Hurst et al., 2007)demonstrated that SPQ scores are significantly correlated with the Autism-Spectrum Quotient (Baron-Cohen et al., 2001). In particular “Interpersonal” factor in SPQ is more or less similar to “Social skill” factor in the AutismSpectrum Quotient. It is likely that individuals with high SPD traits in the present study would also show autistic trait and have difficulties with social perception associated with this trait. Given that we did not include any scales of autistic trait in the current study, we could not tease out this potential confound on the observed facial perception in our findings. Future study should evaluate autistic trait and control ability of facial perception correlated with autistic trait, and to analyze and extract neural responses associated with SPD trait only. 6. Conclusions Notwithstanding the limitations, the current study is the first, to our knowledge, to examine the neural mediation of dynamic happy facial expression with different social cues in dyadic conversations. Although there were some overlap of activations in some brain regions, happiness appearing and disappearing exhibited differences in activation at the PCC bilaterally. The frontal–thalamus–caudate–parietal network was used to recognize happy emotion in the social situation and resolve the conflict between facial expression and social cues. Individuals with SPD traits have different neural inhibition mechanisms when processing the happiness disappearing stimuli with ‘blame’ cues. This could help elucidate the neural mechanism underlying the social interaction difficulty when facing happiness vanishing stimuli. Acknowledgments This work was supported by a grant from the Strategic Priority Research Programme (B) of the Chinese Academy of Sciences (XDB02030200) and the National Key Technologies R&D Programme (2012BAI36B01), the National Science Fund China (81088001, 91132701), and a grant from the Knowledge Innovation Project of the Chinese Academy of Sciences (KSCX2-EW-J-8), to Raymond Chan. Dr. Jia Huang acknowledges supports from the National Science Fund China (31100747), the Young Investigator Scientific Fund of the Institute of Psychology, Chinese Academy of Sciences (YOCX031S01) and funding from the Key Laboratory of Mental Health, Institute of Psychology, Chinese Academy of Sciences. The study was also supported by the initiation fund of the CAS/SAFEA International Partnership Programme for Creative Research Teams to Raymond Chan, Ruben Gur, and David Shum (Y2CX131003). These funding agents had no further role in the study design; in the collection, analysis and interpretation of the data; in the writing of the manuscript; and in the decision to submit the paper for publication. References Abbott GR, Green MJ. Facial affect recognition and schizotypal personality characteristics. Early Interv Psychiatry 2012. http://dx.doi.org/10.1111/j.1751-7893.2012.00346.x. Achim AM, Ouellet R, Roy MA, Jackson PL. Mentalizing in first-episode psychosis. Psychiatry Res 2012;196:207–13. Annett M. A coordination of hand preference and skill replicated. Br J Psychol 1976;67: 587–92. Baron-Cohen S, Wheelwright S, Skinner R, Martin J, Clubley E. The autism-spectrum quotient (AQ): evidence from Asperger syndrome/high functioning autism, males and females, scientists and mathematicians. J Autism Dev Disord 2001;31: 5-17. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry 1961;4:561–71. Brett M, Anton J-L, Valabregue R, Poline JB. Region of interest analysis using an SPM toolbox. [abstract]Presented at the 8th International Conference on Functional Mapping of the Human Brain, June 2–6, Sendai, JapanNeuroImage, 16: No 2; 2002 [Available on CD-ROM].

Chan CM, Tsoi MM. The BDI and stimulus determinants of cognitive-related depression among Chinese college students. Cogn Ther Res 1984;8:501–7. Chan RCK, Wang Y, Yan C, Song L-l, Wang Y-n, Shi Y-f, et al. Contribution of specific cognitive dysfunction to people with schizotypal personality. Psychiatry Res 2011;186:71–5. Chan RCK, Shi YF, Lai MK, Wang YN, Wang Y, Kring AM. The Temporal Experience of Pleasure Scale (TEPS): exploration and confirmation of factor structure in a healthy Chinese sample. PLoS One 2012a;7:e35352. Chan RCK, Wang Y, Yan C, Zhao Q, McGrath J, Hsi X, et al. A study of trait anhedonia in non-clinical Chinese samples: evidence from the Chapman scales for physical and social anhedonia. PLoS One 2012b;7:e34275. Chen WJ, Hsiao CK, Lin CCH. Schizotypy in community samples: the three-factor structure and correlation with sustained attention. J Abnorm Psychol 1997;106: 649–54. Chiew KS, Braver TS. Neural circuitry of emotional and cognitive conflict revealed through facial expressions. PLoS One 2011;6:e17635. Cohen AS, Callaway DA, Najolia GM, Larsen JT, Strauss GP. On “risk” and reward: investigating state anhedonia in psychometrically defined schizotypy and schizophrenia. J Abnorm Psychol 2012;121:407–15. Dickey CC, Panych LP, Voglmaier MM, Niznikiewicz MA, Terry DP, Murphy C, et al. Facial emotion recognition and facial affect display in schizotypal personality disorder. Schizophr Res 2011;131:242–9. Domes G, Schulze L, Böttger M, Grossmann A, Hauenstein K, Wirtz PH, et al. The neural correlates of sex differences in emotional reactivity and emotion regulation. Hum Brain Mapp 2009;31:758–69. Fujimura T, Suzuki N. Effects of dynamic information in recognising facial expressions on dimensional and categorical judgments. Perception 2010;39:543–52. Germine LT, Garrido L, Bruce L, Hooker C. Social anhedonia is associated with neural abnormalities during face emotion processing. Neuroimage 2011;58:935–45. Gong Y. Manual of Wechsler Adult Intelligence Scale—Chinese Version. Changsha: Chinese Map Press; 1992. Gur RE, McGrath C, Chan RM, Schroeder L, Turner T, Turetsky BI, et al. An fMRI study of facial emotion processing in patients with schizophrenia. Am J Psychiatry 2002;159:1992. Gur RE, Loughead J, Kohler CG, Elliott MA, Lesko K, Ruparel K, et al. Limbic activation associated with misidentification of fearful faces and flat affect in schizophrenia. Arch Gen Psychiatry 2007;64:1356–66. Helfinstein SM, Kirwan ML, Benson BE, Hardin MG, Pine DS, Ernst M, et al. Validation of a child-friendly version of the monetary incentive delay task. Soc Cogn Affect Neurosci 2012. http://dx.doi.org/10.1093/scan/nss1057. Hicks JA, Schlegel RJ, King LA. Social threats, happiness, and the dynamics of meaning in life judgments. Pers Soc Psychol Bull 2010;36:1305–17. Hooker CI, Verosky SC, Germine LT, Knight RT, D'Esposito M. Neural activity during social signal perception correlates with self-reported empathy. Brain Res 2010;1308: 100–13. Huang J, Chan RCK, Lu X, Ma Z, Li Z, Gong Q-y. An exploratory study of the influence of conversation prosody on emotion and intention identification in schizophrenia. Brain Res 2009a;1281:58–63. Huang J, Chan RCK, Lu X, Tong Z. Emotion categorization perception in schizophrenia in conversations with different social contexts. Aust N Z J Psychiatry 2009b;43: 438–45. Hurst R, Mitchell J, Kimbrel N, Kwapil T, Nelson-Gray R. Examination of the reliability and factor structure of the autism spectrum quotient (AS) in a non-clinical sample. Personal Individ Differ 2007;43:1938–49. Johnston PJ, Enticott PG, Mayes AK, Hoy KE, Herring SE, Fitzgerald PB. Symptom correlates of static and dynamic facial affect processing in schizophrenia: evidence of a double dissociation? Schizophr Bull 2010;36:680–7. Kee KS, Horan WP, Wynn JK, Mintz J, Green MF. An analysis of categorical perception of facial emotion in schizophrenia. Schizophr Res 2006;87:228–37. Keener MT, Fournier JC, Mullin BC, Kronhaus D, Perlman SB, LaBarbara E, et al. Dissociable patterns of medial prefrontal and amygdala activity to face identity versus emotion in bipolar disorder. Psychol Med 2012;42:1913–24. Kemp J, Berthel MC, Dufour A, Despres O, Henry A, Namer IJ, et al. Caudate nucleus and social cognition: neuropsychological and SPECT evidence from a patient with focal caudate lesion. Cortex 2012. http://dx.doi.org/10.1016/j.cortex.2012.1001.1004. Kohler CG, Walker JB, Martin EA, Healey KM, Moberg PJ. Facial emotion perception in schizophrenia: a meta-analytic review. Schizophr Bull 2010;36:1009–19. Kringelbach ML, Berridge KC. Towards a functional neuroanatomy of pleasure and happiness. Trends Cogn Sci 2009;13:479–87. Kuhn S, Muller BC, van der Leij A, Dijksterhuis A, Brass M, van Baaren RB. Neural correlates of emotional synchrony. Soc Cogn Affect Neurosci 2011;6:368–74. Lee SJ, Lee HK, Kweon YS, Lee CT, Lee KU. Deficits in facial emotion recognition in schizophrenia: a replication study with Korean subjects. Psychiatry Investig 2010;7: 291–7. Lepage M, Sergerie K, Benoit A, Czechowska Y, Dickie E, Armony JL. Emotional face processing and flat affect in schizophrenia: functional and structural neural correlates. Psychol Med 2011;41:1833–44. Li HJ, Chan RCK, Gong QY, Liu Y, Liu SM, Shum D, et al. Facial emotion processing in patients with schizophrenia and their non-psychotic siblings: a functional magnetic resonance imaging study. Schizophr Res 2012a;134:143–50. Li R, Qin W, Zhang Y, Jiang T, Yu C. The neuronal correlates of digits backward are revealed by voxel-based morphometry and resting-state functional connectivity analyses. PLoS One 2012b;7:e31877. Li Z, Kadivar A, Pluta J, Dunlop J, Wang Z. Test–retest stability analysis of resting brain activity revealed by blood oxygen level-dependent functional MRI. J Magn Reson Imaging 2012c;36:344–54.

J. Huang et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 44 (2013) 108–117 Lieberman MD. Social cognitive neuroscience: a review of core processes. Annu Rev Psychol 2007;58:259–89. Lopez-Ibor JJ, Lopez-Ibor M-I, Mendez M-A, Moron M-D, Ortiz-Teran L, Fernandez A, et al. The perception of emotion-free faces in schizophrenia: a magneto-encephalography study. Schizophr Res 2008;98:278–86. Loughead J, Gur RC, Elliott M, Gur RE. Neural circuitry for accurate identification of facial emotions. Brain Res 2008;1194:37–44. Maat A, Fett AK, Derks E. Social cognition and quality of life in schizophrenia. Schizophr Res 2012;137:212–8. McCleery A, Divilbiss M, St-Hilaire A, Aakre JM, Seghers JP, Bell EK, et al. Predicting social functioning in schizotypy: an investigation of the relative contributions of theory of mind and mood. J Nerv Ment Dis 2012;200:147–52. McDonald S, Li S, De Sousa A, Rushby J, Dimoska A, James C, et al. Impaired mimicry response to angry faces following severe traumatic brain injury. J Clin Exp Neuropsychol 2011;33:17–29. Muhlberger A, Wieser MJ, Gerdes AB, Frey MC, Weyers P, Pauli P. Stop looking angry and smile, please: start and stop of the very same facial expression differentially activate threat- and reward-related brain networks. Soc Cogn Affect Neurosci 2010:1–9. Muller VI, Habel U, Derntl B, Schneider F, Zilles K, Turetsky BI, et al. Incongruence effects in crossmodal emotional integration. Neuroimage 2011;54:2257–66. Natu V, Raboy D, O'Toole AJ. Neural correlates of own- and other-race face perception: spatial and temporal response differences. Neuroimage 2011;54:2547–55. Nomura M, Iidaka T, Kakehi K, Tsukiura T, Hasegawa T, Maeda Y, et al. Frontal lobe networks for effective processing of ambiguously expressed emotions in humans. Neurosci Lett 2003;348:113–6. Pitcher D, Dilks DD, Saxe RR, Triantafyllou C, Kanwisher N. Differential selectivity for dynamic versus static information in face-selective cortical regions. Neuroimage 2011;56:2356–63. Platt B, Kamboj S, Morgan CJ, Curran HV. Processing dynamic facial affect in frequent cannabis-users: evidence of deficits in the speed of identifying emotional expressions. Drug Alcohol Depend 2010;112:27–32. Rahko J, Paakki J-J, Starck T, Nikkinen J, Remes J, Hurtig T, et al. Functional mapping of dynamic happy and fearful facial expression processing in adolescents. Brain Imaging Behav 2010;4:164–76. Raine A. The SPQ: a scale for the assessment of schizotypal personality based on DSM-III-R criteria. Schizophr Bull 1991;17:555–64. Ramnani N, Miall R. Instructed delay activity in the human prefrontal cortex is modulated by monetary reward expectation. Cereb Cortex 2003;13:318–27. Recio G, Sommer W, Schacht A. Electrophysiological correlates of perceiving and evaluating static and dynamic facial emotional expressions. Brain Res 2011;1376:66–75.

117

Regenbogen C, Schneider DA, Gur RE, Schneider F, Habel U, Kellermann T. Multimodal human communication — targeting facial expressions, speech content and prosody. Neuroimage 2012;60:2346–56. Roddy S, Tiedt L, Kelleher I, Clarke MC, Murphy J, Rawdon C, et al. Facial emotion recognition in adolescents with psychotic-like experiences: a school-based sample from the general population. Psychol Med 2012:1-10. Rothlisberger M, Riecher-Rossler A, Aston J, Fusar-Poli P, Radu EW, Borgwardt S. Cingulate volume abnormalities in emerging psychosis. Curr Pharm Des 2012;18:495–504. Rymarczyk K, Biele C, Grabowska A, Majczynski H. EMG activity in response to static and dynamic facial expressions. Int J Psychophysiol 2011;79:330–3. Sarkheil P, Goebel R, Schneider F, Mathiak K. Emotion unfolded by motion: a role for parietal lobe in decoding dynamic facial expressions. Soc Cogn Affect Neurosci 2012. http://dx.doi.org/10.1093/scan/nss092. Sato W, Yoshikawa S. Enhanced experience of emotional arousal in response to dynamic facial expressions. J Nonverbal Behav 2007a;31:119–35. Sato W, Yoshikawa S. Spontaneous facial mimicry in response to dynamic facial expressions. Cognition 2007b;104:1-18. Sharpley CF, Bitsika V. The diverse neurogeography of emotional experience: form follows function. Behav Brain Res 2010;215:1–6. Shi YF, Wang Y, Cao XY, Wang Y, Wang YN, Zong JG, et al. Experience of pleasure and emotional expression in individuals with schizotypal personality features. PLoS One 2012;7:e34147. Smith MJ, Horan WP, Karpouzian TM, Abram SV, Cobia DJ, Csernansky JG. Self-reported empathy deficits are uniquely associated with poor functioning in schizophrenia. Schizophr Res 2012;137:196–202. Song X, Dong Z, Long X, Li S, Zuo X, Zhu C, et al. REST: a toolkit for resting-state functional magnetic resonance imaging data processing. PLoS One 2011;6:e25031. Surguladze S, Brammer M, Keedwell P, Giampietro V, Young AW, Travis MJ, et al. A differential pattern of neural response toward sad versus happy facial expressions in major depressive disorder. Biol Psychiatry 2005;57:201–9. Vytal K, Hamann S. Neuroimaging support for discrete neural correlates of basic emotions: a voxel-based meta-analysis. J Cogn Neurosci 2010;22:2864–85. Wittfoth M, Schroder C, Schardt DM, Dengler R, Heinze H-J, Kotz SA. On emotion conflict: interference resolution of happy and angry prosody reveals valence-specific effects. Cereb Cortex 2010;20:383–92. Yan C, Liu WH, Cao Y, Chan RC. Self-reported pleasure experience and motivation in individuals with schizotypal personality disorders proneness. East Asian Arch Psychiatry 2011;21:115–22. Zong JG, Chan RCK, Stone WS, Hsi XL, Cao XY, Zhao Q, et al. Coping flexibility in young adults: comparison between subjects with and without schizotypal personality features. Schizophr Res 2010;122:185–92.