دقت و شدت ابراز هیجانی مطرح شده در سازمان ملل بیماران اسکیزوفرنی غیرداروئی: کانال های آواز و صورت

Abstract This study investigated the ability of schizophrenia patients to volitionally display various emotional expressions. Accuracy and intensity of facial and vocal emotional expression were rated in 26 unmedicated male schizophrenia patients and 20 non-patient male controls while posing emotional facial and vocal expressions. Results indicate that schizophrenia patients, compared to non-patient controls, had deficits in their ability to portray some, but not all, emotions. Accuracy and intensity of posed facial and vocal expressions were inversely correlated with negative symptoms in the patient group. We conclude that observable flattened affect in schizophrenia during posed expression is not evident across all emotions. Furthermore, substantial disruption in the ability to portray posed emotions may be largely driven by the presence of negative symptoms.

1. Introduction Research and clinical reports indicate that schizophrenia is a disorder associated with affective deficits, e.g., anhedonia, avolition, flat affect. In particular, diminished facial expression, (i.e. flat affect), is a fundamental symptom of the disease; indicative of poor prognosis (Fenton and McGlashan, 1991); temporally stable (Putnam and Harvey, 2000); and associated with chronicity (Carpenter and Strauss, 1991). Facial flat affect has been identified in schizophrenia using the EMFACS coding system (Krause et al., 1989), the FACES coding system (Kring et al., 1993, Kring and Neale, 1996 and Kring and Earnst, 1999), more subjective rating systems (Pansa-Henderson et al., 1982 and Martin et al., 1990), as well as with electromyographic recording (Kring et al., 1999 and Kring and Earnst, 2003). In addition, diminished affective expression has been found in the speech of schizophrenia patients as measured by acoustical analyses of speech prosody (Alpert and Anderson, 1977, Andreasen et al., 1981, Levin et al., 1985 and Alpert et al., 1989) as well as by judges' ratings of emotion (Abrams and Taylor, 1978 and Andreasen, 1979). Accumulated evidence indicates that facial expressivity and emotional experience are often dissociated in schizophrenia; therefore, flat affect in schizophrenia does not necessarily represent a reduction in emotional experience. Several studies have found that schizophrenia patients displayed less facial expressivity than non-patient controls while watching affect-eliciting films, yet reported normal levels of emotion (Berenbaum and Oltmanns, 1992, Kring et al., 1993, Kring and Neale, 1996 and Kring and Earnst, 1999). This finding of normal emotional experience is further supported by studies using emotion-modulated startle where the startle response amplitude of schizophrenia patients did not differ from healthy controls when presented with affect-eliciting stimuli (Schlenker et al., 1995 and Curtis et al., 1999). Dissociations have also been found in the vocal realm where a group of schizophrenia patients, in contrast to patients with depression, demonstrated lower levels of vocal positive affect, as indexed by vocal cues. Additionally, the schizophrenic group displayed higher levels of negative affect, indexed by corrugator muscle movement. However, this study found no group differences in self-reported affect (Sison et al., 1996). In addition to diminished spontaneous facial affective display, schizophrenia patients have demonstrated diminished facial and vocal activity when posing emotional expressions. Using only verbal commands and only measuring accuracy, Gottheil and colleagues (1976) found that medicated schizophrenia patients were less accurate than non-patient controls in posing some emotions (i.e. anger and sadness), but not others (i.e. surprise, fear, or joy). Several studies have provided verbal and pictorial instructions separately (Braun et al., 1991, Gaebel and Wölwer, 1992, Yecker et al., 1999 and Trémeau et al., 2005) and have reported diminished intensity (Yecker et al., 1999 and Trémeau et al., 2005) and accuracy in posed displays (Gaebel and Wölwer, 1992 and Trémeau et al., 2005). Another study reported schizophrenia patients' posed expressions to be less accurate but only when cued by verbal command (Braun et al., 1991). Using only verbal cues but measuring both intensity and accuracy simultaneously, Borod and colleagues (1989) found that medicated patients were less accurate and intense in their posed facial expressions. Vocally, schizophrenia patients have also been found to be both less intense and less accurate (Levin et al., 1985, Borod et al., 1989 and Murphy and Cutting, 1990). There are several inconsistencies in the studies of posed emotional expression in schizophrenia that make the data difficult to interpret. One issue is that often the participants have been on neuroleptic medication that may have affected performance on these tasks. Neuroleptic medication likely influences symptom presentation as well as facial and vocal emotional display. Furthermore, studies have used different induction methods (verbal commands and pictorial presentation) and have rarely (only Borod et al., 1989, with 6 schizophrenia patients) measured accuracy and intensity in both channels (vocal and facial). The ability to pose facial and vocal expressions has been found to be positively related in non-patient populations (Zaidel and Mehrabian, 1969 and Cunningham, 1977), but it is not known if these two channels are related in schizophrenia populations. Additionally, the degree of negative symptoms may affect the ability to pose emotional expression, and this has not been adequately investigated in previous studies. If, in addition to the diminished activity found in spontaneous expression, schizophrenia patients' posed expressions are found to be less expressive, this indicate that the symptom of flat affect is driven by an impairment of the ability to volitionally portray emotions—that is, a skill or neuromotor deficit (Dworkin et al., 1996), rather than a deficiency in the neural and behavioral systems that underlie emotional experience. This would be consistent with emotion research in schizophrenia where a dissociation between emotional display and experience has been found. That is, reduced emotional display has not been found to be reflective of reduced emotional experience. As negative symptoms are largely affective deficit symptoms, it is expected that there will be relationships between negative symptoms and the intensity and accuracy of posed emotional expression. Furthermore, it is expected that flat affect will specifically be related to posed expression as both share the common substrate of emotional expression. Research on the relationship between the ability to encode facial and vocal cues has shown that the two skills are positively related in non-patient populations (Zaidel and Mehrabian, 1969, Zuckerman et al., 1975, Zuckerman et al., 1978 and Cunningham, 1977). This suggests that there is a nonspecific encoding factor which accounts for the communication of affect in more than one channel. Therefore, failure to find associations between accuracy and intensity and between the vocal and facial channels may indicate a disruption in the constellation of affective communication skills. The present study sought to investigate posed facial and vocal emotional expressions in unmedicated schizophrenia patients and non-patient controls. Accuracy and intensity of emotional expression were measured simultaneously in the face and in the voice. Negative symptoms were measured in the patient group and related to the accuracy and intensity of the posed expressions. To our knowledge, this is the first study to measure accuracy and intensity in both expressive channels in an unmedicated group of schizophrenic individuals.

. Results Facial accuracy. There was a significant main effect of Emotion, F(5,35) = 12.65, P < 0.001, and a significant Group × Emotion interaction, F(5,35) = 4.17, P < 0.01 (see Table 2). Separate univariate ANOVAs revealed significant group differences for Surprise, F(1,43) = 6.8, P < 0.05, and Sadness, F(1,41) = 5.44, P < 0.05. For both Sadness and Surprise, the posed expressions of the schizophrenia patients were less accurate than the controls. Table 2. Posed facial emotion accuracy (proportion correct) Variable Schizophrenic patients Normal controls X¯ S.D. n X¯ S.D. n Disgust 0.55 0.26 25 0.43 0.27 19 Happiness 0.70 0.31 26 0.810.24 18 Sadness⁎ 0.48 0.27 25 0.660.22 18 Surprise⁎ 0.47 0.30 26 0.680.24 19 Fear 0.26 0.25 26 0.390.24 18 Anger 0.44 0.26 25 0.380.23 16 ⁎Univariate ANOVA, P < 0.05. Table options Facial intensity. Results indicated a significant main effect of Emotion, F(5,32) = 8.70, P < 0.001, and a significant Group × Emotion interaction, F(5,32) = 3.21, P < 0.01 (see Table 3). Separate univariate ANOVAs revealed a significant group difference for Surprise, F(1,43) = 4.27, P < 0.05, with the expressions of schizophrenia patients judged as less intense than the controls' expressions. Table 3. Posed facial emotion intensity Variable Schizophrenic patients Normal controls X¯ S.D. n X¯ S.D. n Disgust 2.00 0.44 25 1.81 0.35 19 Happiness 1.81 0.42 26 1.740.29 18 Sadness 1.52 0.46 25 1.440.43 18 Surprise⁎ 1.74 0.54 26 2.110.52 19 Fear 1.75 0.52 26 1.750.53 18 Anger 1.74 0.54 25 1.560.34 16 ⁎Univariate ANOVA, P < 0.05. Table options Vocal accuracy. As with the facial accuracy data, these proportions were transformed with an arcsine transformation. There were significant main effects of Group, F(1,44) = 13.59, P < 0.001, and Emotion, F(3,42) = 10.12, P < 0.001 (see Table 4). Although this interaction was not significant, we followed up with univariate ANOVAs as our initial hypotheses did not reflect the position that schizophrenic patients would demonstrate a general posing deficit across all emotions. Therefore, these comparisons were planned. Separate univariate ANOVAs revealed significant group differences for Anger, F(1,44) = 7.82, P < 0.01, and Happiness, F(1,44) = 4.96, P < 0.05. In both cases, the vocal expressions of the schizophrenia patients were less accurate than the control group. Both groups were least accurate in their display of Indifference, range of t(45) = 2.85–4.79, P′s < 0.01, but the between-group difference was not significant. Table 4. Posed vocal emotion accuracy (proportion correct) Variable Schizophrenic patients Normal controls X¯ S.D. n X¯ S.D. n Happiness⁎ 0.33 0.29 26 0.520.23 20 Sadness 0.46 0.26 26 0.560.26 20 Anger⁎⁎ 0.42 0.27 26 0.630.17 20 Indifference 0.27 0.14 26 0.320.22 20 ⁎Univariate ANOVA, P < 0.05. ⁎⁎Univariate ANOVA, P < 0.01. Table options Vocal intensity. There was a significant main effect of Emotion, F(3,42) = 17.35, P < 0.001, and a significant Group × Emotion interaction, F(3,42) = 3.05, P < 0.05 (see Table 5). Separate univariate ANOVAs revealed no significant group differences. Table 5. Posed vocal emotion intensity Variable Schizophrenic patients Normal controls X¯ S.D. n X¯ S.D. n Happiness 1.37 0.66 26 1.660.57 20 Sadness 1.25 0.52 26 1.220.50 20 Anger 1.53 0.57 26 1.590.59 20 Indifference 1.03 0.43 26 0.860.57 20 Table options 3.1. Correlations between posed accuracy and intensity1 In order to reduce the number of analyses, the individual emotions were collapsed and average emotion was analyzed. Uncorrected Pearson r′s were computed. Posed facial accuracy and intensity were significantly correlated for both the schizophrenia and control groups (see Table 6). Posed vocal accuracy and intensity were also significantly correlated for both the schizophrenia group and control groups. Table 6. Pearson correlations between posed accuracy, intensity, and negative symptoms Facial accuracy Facial intensity Vocal accuracy Vocal intensity Facial accuracy Schizophrenia – 0.61⁎⁎⁎ 0.45⁎ – Controls – 0.63⁎⁎⁎ – – Facial intensity Schizophrenia – – 0.44⁎ Controls – – 0.52⁎⁎ Vocal accuracy Schizophrenia – 0.77⁎⁎⁎ Controls – 0.69⁎⁎⁎ Vocal intensity Schizophrenia – Controls – SANS total Schizophrenia − − 0.54⁎ −0.41† – SANS flat affect Schizophrenia − 0.47⁎ − 0.60⁎⁎ – – ⁎P < 0.05, ⁎⁎P < 0.01, ⁎⁎⁎P < 0.001, †P = 0.08. Table options 3.2. Correlations between posed vocal and facial expressivity Correlations were computed between facial and vocal accuracy and between facial and vocal intensity separately for both groups. For the schizophrenia group, posed facial accuracy and posed vocal accuracy were significantly correlated, but not for the control group (see Table 6). The correlations were not significantly different between groups. However, there was a significant correlation between posed facial intensity and posed vocal intensity for both groups. 3.3. Correlations between negative symptoms and posed expressions for the schizophrenia group Significant correlations were found between SANS flat affect and posed facial accuracy and posed facial intensity but not for posed vocal accuracy or posed vocal intensity. For the total SANS, significant correlations were found between posed facial intensity and a trend for posed vocal accuracy, but not for posed facial accuracy or posed vocal intensity. Thus, less accuracy and intensity in posed facial expressions was associated with greater ratings of flat affect and negative symptoms (Table 6). 3.4. Correlations between measures of chronicity and posed expressions and negative symptoms for the schizophrenia group Using the number of months in the hospital, lifetime, as a measure of chronicity, we found no significant correlations between chronicity and any of our posed emotion measures. However, chronicity was significantly correlated with both total SANS (r = 0.66, P < 0.001) and the SANS flat affect subscale (r = 0.67, P < 0.001).