Year : 2008 | Volume
: 30 | Issue : 2 | Page : 90--97
Emotion recognition deficits in antipsychotic-naive schizophrenia
Balaji Bharadwaj1, Rashmi Arasappa1, Rishikesh V Behere1, Ganesan Venkatasubramanian1, PN Jayakumar2, BN Gangadhar1,
1 Department of Psychiatry, National Institute of Mental Health and Neurosciences (NIMHANS), Bangalore-560 029, India
2 Department of Neuroimaging and Interventional Radiology, National Institute of Mental Health and Neurosciences (NIMHANS), Bangalore-560 029, India
Department of Psychiatry, National Institute of Mental Health and Neurosciences, Hosur Road, Bangalore-560 029
Background: Emotion recognition deficits in schizophrenia form an important component of deficits seen in the disorder. Emotion recognition abilities correlate well with symptom dimensions and sociooccupational outcome of the disorder. The influence of pharmacological treatment on brain functions makes it important to design studies examining emotion recognition in drug-naive patients. However, emotion recognition deficits in antipsychotic-naive schizophrenia patients have not been previously examined. Methods: In this study, TRENDS - A Tool for Recognition of Emotions of Neuropsychiatric Disorders - a tool validated in Indian population, was used to assess emotion recognition abilities of antipsychotic-naive schizophrenia patients ( n = 20) and group-matched healthy controls ( n = 20). Results: The study showed significant deficits in emotion recognition in patients, especially with regard to fear ( P = 0.001), followed by disgust ( P = 0.006), and anger ( P = 0.017). The under-recognition of these emotions was positively correlated with high negative symptom scores ( r = 0.470; P = 0.018) and negatively correlated with high positive symptom scores ( r = -0.447; P = 0.048). Conclusions: Presence of significant emotion recognition abnormalities in antipsychotic-naive schizophrenia suggests that these abnormalities might be intrinsically related to the pathogenesis of this disorder.
|How to cite this article:|
Bharadwaj B, Arasappa R, Behere RV, Venkatasubramanian G, Jayakumar P N, Gangadhar B N. Emotion recognition deficits in antipsychotic-naive schizophrenia.Indian J Psychol Med 2008;30:90-97
|How to cite this URL:|
Bharadwaj B, Arasappa R, Behere RV, Venkatasubramanian G, Jayakumar P N, Gangadhar B N. Emotion recognition deficits in antipsychotic-naive schizophrenia. Indian J Psychol Med [serial online] 2008 [cited 2019 May 22 ];30:90-97
Available from: http://www.ijpm.info/text.asp?2008/30/2/90/48481
The schizophrenic illness is characterized by various positive symptoms like prominent delusions, hallucinations, positive formal thought disorder, and persistently bizarre behavior on one hand, and negative symptoms like affective flattening, alogia, avolition, anhedonia, and attentional impairment on the other hand.  Liddle  further described disorganization symptoms like inappropriate affect, poverty of content of speech, and disturbances of the form of thought leading to poor self-care and impersistence at work.
The cognitive symptoms were probably first recognized by Kraepelin himself who gave the disease the name 'dementia praecox' that may be understood as 'early occuring cognitive decline'. Cognitive deficits in schizophrenia are majorly related to the frontal lobar functions, namely: executive dysfunction, disinhibition, and motor incoordination as shown by Poole et al .  as well as deficits in attention-concentration noted in early schizophrenia by McGhie et al .  There are also social cognitive deficits in schizophrenia which include deficits in the recognition of emotions expressed by others (reviewed by Mandal  ), difficulties in understanding the thoughts or emotions from the point of view of others (Theory of Mind deficits), or misunderstanding others' benign intentions as hostile, as described by Brune. 
Neurobiology of emotion recognition in healthy subjects
The neurobiological correlates of emotion recognition have been extensively studied. A review by Phillips et al .  describes the substrates of normal emotion recognition as being grouped into two modules, serving three functions that bring about the totality of emotional states - 1) the ventral stream, as the first functional module involved in two of these functions, namely:a) the identification of emotional salience of a stimulus (subserved by the limbic regions like amygdala and insular cortex); and b) the generation of the affective state (subserved by the prefrontal regions); and 2) the dorsal stream, consisting of the hippocampi, dorsal regions of anterior cingulate cortex, dorsomedial and dorsolateral prefrontal cortices which serves as the second functional module helpful in the regulation of the generated affective state and in a cognitive approach to a situation rather than an emotional approach [Figure 1].
There is evidence to show that these limbic and prefrontal regions important for emotion recognition are preferentially involved in schizophrenia, both structurally and functionally. 
Deficits in emotional recognition in schizophrenia
Regarding deficits in emotion recognition, which have been studied in great detail among schizophrenics, deficits in the recognition of shame,  disgust,  sadness, , and fear  have been shown in patients compared to controls. Kohler et al .  also showed maximum deficits for fear, followed by anger and disgust. They also showed that increasing the intensity of emotions did not improve correct identification of fear as much as it did the other emotions. Even in Indian studies, fear and anger have shown to be underidentified by schizophrenia patients compared to controls.  Muzekari et al .  and Cramer et al .  found that schizophrenia patients performed poorly for negative emotions in general. These deficits were more severe in acutely ill patients than in chronic patients.  Also, Mueser et al .  noted more deficits in patients with longer illness durations. Bediou  showed deficits in emotion recognition in schizophrenia patients and their healthy siblings compared to controls. The patients' deficits remained even after they controlled their acute phase symptoms, showing that it may be a trait marker. Other studies with short-term treatment over a period of 2-4 weeks , also showed persistent deficits in emotion recognition.
Emotion recognition deficits and clinical features and outcome of schizophrenia
The emotion recognition deficits in schizophrenics correlate well with the severity of positive and negative symptoms and also with the cognitive domains of attention, verbal and spatial memory, as well as language abilities.  Kline et al .  and Lewis et al .  also noted a relationship to the subtype of schizophrenia (paranoid/nonparanoid). Kee et al .  showed that better emotion perception (including facial emotion, voice emotion, and affect perception) predicted work functioning/independent living in both sexes of patients even when controlled for conceptual disorganization and social/family relationships. Social cognition helps in better rehabilitation in patients as shown by the following studies by Dickinson et al .  using a role-play based social skills assessment, and Vauth et al .  using Structural Equation Modeling (SEM) showed that there is a larger contribution of social skills to vocational rehabilitation than of nonsocial cognitive skills.
Schizophrenia being an illness with the symptom dimensions noted above, and social cognition being one of the important symptom domains that seems to be a trait marker for the illness, not much affected by short-term treatment and also a predictor of sociooccupational outcome, it would be useful to study emotional recognition in a population of drug-naive schizophrenia patients. There are studies done in this field even in India by Mandal. ,, However, these studies used emotion recognition tools developed for use in different ethnic and cultural settings than the place of study and used patients who have been treated with antipsychotics. Since antipsychotics can significantly influence the cognitive functions over a period of time,  it is important to assess the emotion recognition in untreated patients. The study by Mandal  which has used Caucasian faces for use in Indian patients has only two drug-naive patients.
In summary, review of literature reveals that there is no published study of emotion recognition in drug-naive schizophrenia patients in the Indian setting. Thus, the current study is the first study comparing the emotion recognition of drug-naive schizophrenia patients with those of healthy controls in India using a culturally and ecologically validated tool.
Aims and Objectives
We therefore aimed to study the emotion recognition abilities in a group of antipsychotic-naive schizophrenia patients ( n = 20) and compared it with age, sex, education, and handedness matched healthy controls ( n = 20). We proposed to use an emotion recognition tool appropriate for the cultural setting. The tool used in this study (Tool for Recognition of Emotions in Neuropsychiatric Disorders (TRENDS)) has been recently developed and validated in India. 
The hypothesis was patients with schizophrenia would have greater deficits in emotion recognition when compared to healthy controls. We also hypothesized that these deficits may be related to the psychopathology of the patients.
Selection of cases and controls
Patients who attended the clinical services of NIMHANS, Bangalore, with a DSM-IV diagnosis of schizophrenia  who satisfied the following criteria: 1) of either sexes, 2) between the ages 15-45 years, and 3) no prior treatment with antipsychotic medications were included in the study. Written informed consent was taken. However, the following groups of patients were excluded: 1) emergency clinical conditions associated with psychosis, 2) left handedness, 3) recent or current medical or neurological disorder including seizures, cerebral palsy, or mental retardation, 4) those with a family history of hereditary neurological disorders, 5) comorbid substance dependence (DSM-IV criteria), and 6) lifetime history of head injury that was associated with loss of consciousness for more than 10 minutes, seizures, neurological deficits, depressed skull fractures, or needing surgical interventions.
Healthy controls were recruited through word-of-mouth from consenting volunteers after ruling out of a psychiatric diagnosis by a clinical interview. They were group-matched with patients for age, sex, education, and handedness. In addition, they were screened for family history of psychoses in their first-degree relatives, presence of which will be an exclusion criterion in addition to the same exclusion criteria as outlined for the patients.
Clinical assessment of patients
Patients satisfying the above criteria were rated on their psychopathology using the following scales: Scale for Assessment of Positive Symptoms (SAPS)  and Scale for Assessment of Negative Symptoms (SANS). 
Emotional recogniton tool
TRENDS is an emotion recognition tool developed by Behere et al .  and validated in the Indian population. The tool consists of color photographs and videos of four actors of Indian ethnicity, two each of each sex (male/female) and two each of each age group (young/old), as follows: young man, young woman, old man, old woman.
There are six emotions - happy, fear, anger, sad, disgust, and surprise - portrayed by each of these four actors in two different intensities (low and high), plus neutral faces. In all, 52 still photographs were shown using a computer followed by 28 demonstrating these six emotions in a single intensity plus a neutral video of each of these four actors [Figure 2].
The subjects (patients and controls) had to explicitly name the emotion portrayed by choosing one out of the seven named options given in a forced-choice paradigm.
For the purpose of analysis, the emotion 'surprise' was not included as the discrimination of this emotion was poor among the emotions as found by the initial validations of the TRENDS tool.  The other five emotions of happy, sad, anger, fear, and disgust were included along with neutral faces for purposes of statistical analysis. The faces representing anger, fear, and disgust were regarded as threatful and negative emotions for purposes of further analysis of underidentification and overidentification.
Comparison of patient group with the control group was made on three parameters by the use of independent samples t -test.
Correct responses were totaled to get an idea of the general task performance in the test. Also, we looked for the influence of the nature of images as static or dynamic on the task performance as dynamic images help to identify emotions better. Underidentification errors were defined as a failure to recognize threatful emotions like fear, anger, and disgust correctly and instead label them as a positive emotion or as neutral (non-threatful). Underidentification of emotions between the patients and controls was compared. Further assessment of which emotions were specifically under-identified was made and correlated with the psychopathology ratings of the patients. Fear being the most underidentified of the three threatening images, further analysis was done regarding what it was underidentified as. Overidentification errors were defined as identifiction of a non-threatful stimulus as being threatful. Overidentification of emotions by patients and controls was compared. Having found significant underidentification among the patients, we went on to correlate this error with symptom dimensions of the illness. We did a correlational analysis of SAPS and SANS scores to underidentification errors.
Functional magnetic resonance imaging correlates of emotion processing: A pilot study
For the purpose of illustration in this paper, we examined the brain activation profile while processing the 'fear' expression in a chosen patient as well as control. To facilitate comparison, we chose the patient who performed the worst (maximum underidentification score) in identifying 'fear' expression and the control participant who performed the best (minimum underidentification score).
Magnetic Resonance Imaging (MRI) was done with Philips 3T scanner. First, T1 weighted three-dimensional structural imaging was performed for anatomical localization purpose. After obtaining the anatomic MR images, Blood-Oxygen Level Dependent (BOLD)/Echo-Planar Images (EPI) image acquisition was done.  Inside the scanner, subjects were requested to visualize the facial emotional expression and identify the emotions using the TRENDS. [30,35] The functional MRI (fMRI) processing and analyses were carried out for these subjects using Statistical Parametric Mapping 2 (SPM2). ,
Task performance was compared between the patients and controls by comparing the number of correct responses. We found that patients made significantly more errors than controls when the number of correct responses are taken across all the image types [Table 1]. The differences were highly significant ( P ≤ 0.005) for all the image types - static images of either intensities, static images as a group, and dynamic videos [Figure 3].
There is significant misrecognition of the emotions by patients when compared to healthy controls. The difference is most significant for fear and disgust. This difference holds true for static as well as dynamic images.
We looked for underidentification errors between the two groups [Table 2] and found that patients tend to underidentify emotions more often than healthy controls. This error was highly significant across all image types, though the difference was more with dynamic images.
Specifically there was underidentification of the emotions fear, anger, and disgust [Table 3] among patients compared to controls, with highly significant differences between patient and control groups for all three emotions. The difference was maximum and most significant for fear [Figure 4], followed by disgust and anger.
Underidentification of fear being most significant, we further analyzed what it was under-recognized as. Patients were significantly more likely to underidentify it as sad or as happy rather than ascribe neutral emotional valence to it [Table 4].
Looking for overidentification errors [Table 5] revealed that there was an overall low tendency to commit errors in this direction among both patients and controls. Patients tend to identify threat in neutral or positive faces more often than controls, but this difference, though consistent throughout the images types, was not statistically significant.
Correlational analysis with the SAPS and SANS scores [Table 6] revealed that, SAPS scores showed some significant negative correlation with underidentification errors on high-intensity photographs ( r = -0.447; P = .048), that is, those with greater positive symptoms were less likely to commit errors of underidentification of emotions. SANS scores showed significant positive correlation with underidentification errors on dynamic images ( r = 0.470; P = .018).
This study of emotional recognition abilities among twenty drug-naive schizophrenia patients compared with healthy controls using a tool validated within the Indian population is the first such study to the best of our knowledge. While previous studies of facial emotional recognition done in India used black and white photographs of subjects from other ethnic groups, this tool has color photographs of Indian faces which makes it culturally valid, as we have used a tool validated in the same cultural setting. Inclusion of dynamic images makes it ecologically valid because, in real-life situations the facial affect recognition is helped by dynamically changing expressions.
Our results show a significant deficit in overall performance of antipsychotic-naive schizophrenia patients in comparison with age, sex, education, and handedness matched healthy controls. The consistent underperformance by patients on emotional recognition task, is a replication of what has been found in many earlier studies , that patients with schizophrenia tend to make more errors in recognizing emotional cues from faces when compared to controls.
Green  suggested the grouping together of anger, fear, and disgust as 'threatful stimuli'; the basis of this grouping being the presence of a distinct neural circuitry for the recognition of these emotions. There is also a theoretical basis in this grouping, because the ability to identify these threatful stimuli quickly in the environment confers a survival advantage upon the animal. Possibly, this also gives rise to paranoid interpretations in schizophrenia. An assessment of the ability to recognize the threatful stimuli, revealed that patients significantly under-recognize threat (mainly fear and disgust) when compared to the healthy controls. This is similar to the previous findings of Dougherty et al .  who showed poor recognition of disgust, Mandal et al.  who showed deficits specific to fear and anger, and Edwards et al .  who noticed deficits in the recognition of fearful affect.
Among these emotions, fear being the most under-recognized emotion, we further analyzed what fear was being under-recognized as. Fear was most frequently underidentified as sad. This corresponds to findings of under-recognition of fear by Kohler et al .  and also to a tendency to overidentify sadness by Tsoi et al .  The underidentification was in fact exacerbated in dynamic videos, which is counter-intuitive to the expectation of a better performance with video expressions of emotions. This is probably because the dynamic images improved the controls' performances much more than patients' performances.
There is a tendency of patients to overidentify emotions compared to the control group. However, this difference is not large or significant.
A correlational analysis between patients' under-recognition was compared with symptom dimensions. There was a significant negative correlation between positive symptoms (SAPS scores) and underidentification errors. This relatively better performance of patients having higher levels of positive symptoms corresponds to the earlier findings of better task performance in paranoid subtype of the illness rather than nonparanoid subtypes. , There was also a positive correlation between negative symptoms and underidentification. This probably points out to a possible role of high negative symptoms, or a negativistic state in causing underperformance in tasks of emotion recognition.
Functional magnetic resonance imaging correlates of emotion processing: Preliminary results
In concurrence with the clinical/behavioral observation, the healthy control showed activation involving medial frontal cortex as well as the limbic cortex whereas the patient did not show activation in these brain regions (had a minimal activation in the frontal cortex). Interestingly, the brain regions activated in the control have indeed been described as critical emotion processing areas.  Gur et al .  and Phillips et al .  found decreased activation of left amygdala during presentation of fearful faces to schizophrenia patients. There are also functional abnormalities in the hippocampus during facial emotion processing.  Our pilot imaging thus replicates the findings of these previous studies.
This study shows that antipsychotic-naive schizophrenia patients have significant deficits in recognition of emotions when compared to age, sex, education, and handedness matched healthy controls. This deficit is most significantly the under-recognition of the threatful emotion of fear. This under-recognition of fear also correlated well with high negative symptoms and lower levels of positive symptoms in these patients. A pilot fMRI comparison showed a brain activation pattern that is in tune with the performance of the patient on the behavioral task.
Since previous studies of emotion recognition have mainly studied treated schizophrenia patients, it was not clear whether the deficits in emotion recognition were related to the disorder per se or to treatment effects. Our findings of emotion recognition deficits in antipsychotic-naive schizophrenia patients suggest that these deficits may be an intrinsic component related to the pathogenesis of the disorder rather than just an effect of pharmacological treatment.
The strengths of this study are: 1) it is the largest study of emotional recognition abilities in a group of schizophrenia patients. 2)The patients were antipsychotic-naive, to effectively eliminate the effects of drug treatment. 3) The tool used is culturally validated for use in India, and 4) It is an ecologically sound tool, by virtue of having dynamic images. The limitations of the study are: 1) use of forced-choice response. However, this was done to avoid the wide variations that could follow open responses and thereby cause difficulties in analysis; and 2) the sample size of 20 is not too large. However, being the first study of antipsychotic-naive patients, the findings of this study could serve as a pilot for future studies in this population.
Future studies in this field should include larger samples and see if the findings of this study are replicable. There should be attempts to use imaging modalities like fMRI to look for the biological correlates of emotion recognition deficits in schizophrenia (ongoing studies). One could also attempt to look for the effects of treatment on the emotion recognition abilities in schizophrenia.[Figure 5],[Figure 6]
|1||Andreasen NC, Olsen S. Negative v positive schizophrenia: Definition and validation. Arch Gen Psychiatry 1982;39:789-94.|
|2||Liddle PF. The symptoms of chronic schizophrenia: A re-examination of the positive-negative dichotomy. Br J Psychiatry 1987;151:145-51.|
|3||Poole JH, Ober BA, Shenaut GK, Vinogradov S. Independent frontal-system deficits in schizophrenia: Cognitive, clinical and adaptive implications. Psychiatry Res 1999;85:161-76.|
|4||McGhie A, Chapman J. Disorders of attention and perception in early schizophrenia. Br J Med Psychol 1961;34:103-16.|
|5||Mandal MK, Pandey R, Prasad AB. Facial expressions of emotions and schizophrenia: A review. Schizophr Bull 1998;24:399-412.|
|6||Brune M. Emotion recognition, 'theory of mind' and social behavior in schizophrenia. Psychiatry Res 2005;133:135-47.|
|7||Phillips ML, Drevets WC, Rauch SL, Lane R. Neurobiology of emotion perception I: The neural basis of normal emotion perception. Biol Psychiatry 2003;54:504-14.|
|8||Shenton ME, Dickey CC, Frumin M, McCarley RW. A review of MRI findings in schizophrenia. Schizophr Res 2001;49:1-52.|
|9||Dougherty FE, Bartlett ES, Izard CE. Responses of schizophrenics to expressions of the fundamental emotions. J Clin Psychol 1974;30:243-6.|
|10||Edwards J, Pattison PE, Jackson HJ, Wales RJ. Facial affect and affective prosody recognition in first-episode schizophrenia. Schizophr Res 2001;48:235-53.|
|11||Kohler CG, Bilker W, Hagendoorn M, Gur RE, Gur RC. Emotion recognition deficit in schizophrenia: Association with symptomatology and cognition. Biol Psychiatry 2000;48:127-36.|
|12||Kohler CG, Turner TH, Bilker WB, Brensinger CM, Siegel SJ, Kanes SJ, et al . Facial emotion recognition in schizophrenia: Intensity effects and error pattern. Am J Psychiatry 2003;160:1768-74.|
|13||Mandal MK, Palchoudhury S. Decoding of facial affect in schizophrenia. Psychol Rep 1985;56:651-2.|
|14||Muzekari LH, Bates ME. Judgment of emotion among chronic schizophrenics. J Clin Psychol 1977;33:662-6.|
|15||Cramer P, Weegmann M, O'Neil M. Schizophrenia and the perception of emotions: How accurately do schizophrenics judge the emotional states of others? Br J Psychiatry 1989;155:225-8.|
|16||Gessler S, Cutting J, Frith CD, Weinman J. Schizophrenic inability to judge facial emotion: A controlled study. Br J Clin Psychol 1989;28:19-29.|
|17||Mueser KT, Penn DL, Blanchard JJ, Bellack AS. Affect recognition in schizophrenia: A synthesis of findings across three studies. Psychiatry 1997;60:301-8.|
|18||Bediou B, Asri F, Brunelin J, Krolak-Salmon P, D'Amato T, Saoud M, et al . Emotion recognition and genetic vulnerability to schizophrenia. Br J Psychiatry 2007;191:126-30.|
|19||Lewis S, Garver DL. Treatment and diagnostic subtype in facial affect recognition in schizophrenia. J Psychiatr Res 1995;29:5-11.|
|20||Gaebel W, Wolwer W. Facial expression and emotional face recognition in schizophrenia and depression. Eur Arch Psychiatry Clin Neurosci 1992;242:46-52.|
|21||Kline JS, Smith JE, Ellis HC. Paranoid and nonparanoid schizophrenic processing of facially displayed affect. J Psychiatr Res 1992;26:169-82.|
|22||Kee KS, Green MF, Mintz J, Brekke JS. Is emotion processing a predictor of functional outcome in schizophrenia? Schizophr Bull 2003;29:487-97.|
|23||Dickinson D, Bellack AS, Gold JM. Social/communication skills, cognition, and vocational functioning in schizophrenia. Schizophr Bull 2007;33:1213-20.|
|24||Vauth R, Rusch N, Wirtz M, Corrigan PW. Does social cognition influence the relation between neurocognitive deficits and vocational functioning in schizophrenia? Psychiatry Res 2004;128:155-65.|
|25||Mandal MK, Palchoudhury S. Identifying the components of facial emotion and schizophrenia. Psychopathology 1989;22:295-300.|
|26||Mandal MK, Rai A. Responses to facial emotion and psychopathology. Psychiatry Res 1987;20:317-23.|
|27||Mandal MK, Bryden MP, Bulman-Fleming B. Similarities and variations in facial expressions of emotions: Cross-cultural evidence. Int J Psychol 1996;31:41-58.|
|28||Sharma T. Impact on cognition of the use of antipsychotics. Curr Med Res Opin 2002;18:s13-7.|
|29||Mandal MK, Jain A, Haque-Nizamie S, Weiss U, Schneider F. Generality and specificity of emotion-recognition deficit in schizophrenic patients with positive and negative symptoms. Psychiatry Res 1999;87:39-46.|
|30||Behere RV, Raghunandan VN, Venkatasubramanian G, Subbakrishna DK, Jayakumar PN, Gangadhar BN. TRENDS: A Tool for Recognition of Emotions in Neuropsychiatric Disorders ( Winner of DS Raju Memorial Award for best PG paper presentation at the annual conference of Indian Psychiatric Society (South-Zone) October 2007. Indian J Psychol Med 2008; 30(1):32-8.|
|31||Diagnostic and Statistical Manual of Mental Disorders, Text Revision. 4th ed. American Psychiatric Association; 1994.|
|32||Andreasen NC. The Scale for the Assessment of Positive Symptoms (SAPS). Iowa City, IA: University of Iowa; 1984.|
|33||Andreasen NC. The Scale for the Assessment of Negative Symptoms (SANS). Iowa City, IA: University of Iowa; 1983.|
|34||Ganesan V, Green RD, Hunter MD, Wilkinson ID, Spence SA. Expanding the response space in chronic schizophrenia: the relevance of left prefrontal cortex. Neuroimage 2005;25:952-7.|
|35||Venkatasubramanian G, Puthumana DT, Jayakumar PN, Gangadhar BN. A functional MRI study of neurohaemodynamic abnormalities during emotion processing in subjects at high risk for schizophrenia. Indian J Psychiatry 2008;(in press).|
|36||Schneider F, Gur RC, Gur RE, Shtasel DL. Emotional processing in schizophrenia: neurobehavioral probes in relation to psychopathology. Schizophr Res 1995;17:67-75.|
|37||Kohler CG, Turner TH, Gur RE, Gur RC. Recognition of facial emotions in neuropsychiatric disorders. CNS Spectr 2004;9:267-74.|
|38||Green MJ, Phillips ML. Social threat perception and the evolution of paranoia. Neurosci Biobehav Rev 2004;28:333-42.|
|39||Tsoi DT, Lee KH, Khokhar WA, Mir NU, Swalli JS, Gee KA, et al . Is facial emotion recognition impairment in schizophrenia identical for different emotions? A signal detection analysis. Schizophr Res 2008;99:263-9.|
|40||Phillips ML, Williams L, Senior C, Bullmore ET, Brammer MJ, Andrew C, et al . A differential neural response to threatening and non-threatening negative facial expressions in paranoid and non-paranoid schizophrenics. Psychiatry Res 1999;92:11-31.|
|41||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;159:1992-9.|