پاسخ EMG صورت به حالات چهره عاطفی پویا در پسران مبتلا به اختلالات رفتار مخرب

Abstract Based on the assumption that facial mimicry is a key factor in emotional empathy, and clinical observations that children with disruptive behavior disorders (DBD) are weak empathizers, the present study explored whether DBD boys are less facially responsive to facial expressions of emotions than normal controls. Facial electromyographic (EMG) activity in the zygomaticus major and corrugator supercilii muscle regions, and heart rate activity were studied in 22 clinically referred 8–12-year-old DBD boys and 22 age-matched normal controls during exposure to dynamic happy and angry expressions. Dispositional emotional empathy was assessed by a self-report questionnaire for children. The happy and angry facial expressions evoked distinct facial EMG response patterns, with increased zygomaticus muscle activity to happy expressions and increased corrugator muscle activity to angry expressions. The corrugator (but not the zygomaticus) muscle response pattern was less pronounced for DBD boys than the normal controls. Attending to the emotional expressions was associated with equivalent cardiac deceleration in both groups, reflecting a similar orienting/attention response. Lower empathy scores were obtained for DBD boys than for normal controls. In conclusion, facial mimicry responses to angry facial expressions were subnormal in DBD boys, which may be a sign of a deficient early component in the process of emotional empathy, and thus play a role in impaired empathic responding.

1. Introduction In clinical practice, children with oppositional defiant disorder (ODD) and conduct disorder (CD), also called disruptive behavior disorders (DBD), are thought to be poor in empathic skills, that is, the ability to understand and share another’s emotional state (Cohen and Strayer, 1996). The clinical impression that DBD children are weak empathizers is given support by studies reporting empathy deficits in clinically identified CD adolescents (Cohen and Strayer, 1996), and DBD boys (de Wied et al., 2005). Starting from the assumption that facial mimicry is a fundamental component in the process of empathy, especially emotional empathy (i.e., shared feelings) ( Hatfield et al., 1994, Hoffman, 2000 and Meltzoff, 1993), the present study explored facial electromyographic (EMG) responses in clinically referred 8–12-year-old DBD boys and normal controls, when exposed to dynamic emotional facial expressions. An early report by Lipps (1905) first called attention to the possible role of motor mimicry/imitation in the automatic transmission of emotions. Lipps proposed that people tend to mimic the facial, vocal or postural expressions of emotions displayed by an interaction partner, and that such mimicry responses may evoke corresponding emotions in the observer. Many researchers currently consider motor mimicry as the very essence of emotional empathy, likely to be biologically “hard-wired” (e.g., Hatfield et al., 1994, Hoffman, 2000, Meltzoff, 1993 and Preston and de Waal, 2002). According to Hoffman’s (2000) developmental model of empathy, mimicry is an early component in the process of empathy. The tendency to automatically mirror emotional expressions becomes manifest already in the first days of life by reflexive crying in response to other babies’ crying. Primitive mechanisms (e.g., mimicry) contribute to the development of empathy in the early preverbal period, but continue to operate past childhood. As the cognitive system develops, higher-order cognitive processes (e.g., role-taking) come to play a more important role. However, mature empathic responses are generated both by primitive (automatic) mechanisms and more sophisticated cognitive processes. Focusing on facial imitation, Dimberg and colleagues (Dimberg, 1982, Dimberg, 1988, Dimberg, 1990 and Dimberg and Lundquist, 1990) demonstrated that exposure to pictures of happy and angry faces evoke distinct facial EMG response patterns. Adult subjects react spontaneously with increased zygomaticus major (cheek) muscle activity when exposed to happy facial expressions, whereas angry facial expressions evoke increased corrugator supercilii (eyebrow) muscle activity. These facial EMG reactions are rapidly evoked (i.e., under 400 ms; Dimberg and Thunberg, 1998), also when subjects are unconsciously exposed to facial stimuli (Dimberg et al., 2000), and hard to restrain voluntarily (Dimberg et al., 2002). The data support the hypothesis that facial mimicry is an early, automatic response to others’ facial displays. Sonnby-Borgström and colleagues (Sonnby-Borgström, 2002 and Sonnby-Borgström et al., 2003) conducted studies on facial EMG (zygomaticus and corrugator) and empathic sensitivity in high and low empathic students. When pictures of happy and angry expressions were presented at short exposure times (eliciting automatic reactions), the high-empathy group showed a stronger mimicry response than the low-empathy group. Furthermore, a significant correspondence between facial reactions and subjective feelings was found only among the high empathic students (Sonnby-Borgström, 2002), suggesting that those who show a stronger tendency to mimic other’s facial expressions are more susceptible to emotional empathy. The aim of the present study was to determine facial responsiveness in DBD boys and normal controls. Respondents were exposed to 5-s moving pictures showing a male model producing dynamic angry or happy facial expressions. We used dynamic presentations because moving images of facial expressions may constitute a stronger eliciting stimulus than static displays (Wehrle et al., 2000). Facial EMG activity in the zygomaticus major and corrugator supercilii muscle regions was assessed during exposure to the facial stimuli. Dispositional emotional empathy was assessed by a self-report questionnaire for children. Based on theory and empirical studies with adults, it was predicted that happy faces would evoke more zygomaticus activity than angry faces, whereas angry faces would evoke more corrugator activity than happy faces. In fact, corrugator activity tends to be inhibited during exposure to happy faces (Dimberg and Lundquist, 1990, Dimberg and Thunberg, 1998 and Dimberg et al., 2000). Furthermore, it was predicted that this typical response pattern (i.e., increased zygomaticus activity during happy faces and increased corrugator activity during angry faces) would be less pronounced for DBD boys than for normal controls. Given the preliminary nature of this first study on facial EMG responses in DBD boys, no more specific hypotheses were formulated. Attending to affective pictures or film clips is generally associated with cardiac deceleration (McManis et al., 2001 and Waldstein et al., 2000), which may be reflective of an orienting or attention response (Cook and Turpin, 1997). Cardiac deceleration was also demonstrated during exposure to happy and angry facial expressions (Dimberg, 1982 and Dimberg, 1990). To control for differences in heart rate responses during happy versus angry facial expressions, heart rate was also measured in the present study.

3. Results Prior to conducting the main analyses of the study, the physiological (facial EMG and HR) and self-report (emotional empathy) data were subjected to independent-samples t-tests (two-tailed) comparing DBD boys who were on medical treatment with DBD boys who were not; DBD boys with and without ADHD; and DBD boys from inpatient units with those from day-treatment units. Except for a significant difference in HR response during angry faces between DBD boys with and without ADHD (t20 = −2.53, p < 0.05), no significant differences emerged on any of the physiological response measures, nor on emotional empathy. Therefore, the scores of all DBD subjects were collapsed in subsequent analyses. 3.1. Temporal profiles of facial EMG and HR during the neutral face film clip Facial EMG. For both muscle regions, the MANOVAs revealed no significant effect of intervals, nor significant intervals X group, and intervals X presentation interactions, indicating a flat profile in zygomaticus and corrugator EMG activity throughout the 4000 ms neutral face stimulus for each stimulus presentation within both groups. Hence, differences in facial EMG activity during the emotional expressions cannot be attributed to the effects of attending to the film clips as such. The temporal profiles for both muscle regions during the neutral film clip (collapsed across presentations) are illustrated in Fig. 1. Mean EMG activity of zygomaticus and corrugator for DBD boys and normal controls ... Fig. 1. Mean EMG activity of zygomaticus and corrugator for DBD boys and normal controls during presentation of the neutral face film clip. Figure options HR. The MANOVA revealed a significant effect of intervals (F7,36 = 9.84, p < 0.0001), but no significant intervals X group, or intervals X presentation interaction. HR significantly decreased across the eight time intervals, probably reflecting an orienting or attention response. Moreover, the analysis revealed a significant effect of group (F1,42 = 15.85, p < 0.001), with higher HR values for DBD boys (mean = 88.29, SD = 10.23) than for normal controls (mean = 76.23, SD = 9.86). It is important to note that higher HR values were also observed in the DBD group (mean = 89.79, SD = 9.66) relative to the control group (mean = 75.16, SD = 9.01) during the aquatic film clip (t42 = − 5.19, p < 0.0001). Since DBD boys were rated by parents and teachers as more anxious than controls ( Table 1), it is possible that they exhibited higher HR values because they experienced more anxiety in a novel laboratory situation. 3.2. Emotional empathy Consistent with predictions, an independent-samples t-test yielded a significant difference between groups (t42 = 2.04, p < 0.05), with lower empathy scores for DBD boys (mean = 10.77, SD = 2.89) than controls (mean = 12.55, SD = 2.87). 3.3. Facial EMG responses to dynamic emotional facial expressions Zygomaticus. The MANOVA 1 revealed a significant effect of emotional expression (F1,42 = 13.26, p < 0.001), but no significant effect of group or trials, nor significant interactions. Consistent with predictions, happy faces evoked a significantly larger increase in zygomaticus EMG activity than angry faces ( Fig. 2). Mean zygomaticus EMG responses for DBD boys and normal controls to dynamic ... Fig. 2. Mean zygomaticus EMG responses for DBD boys and normal controls to dynamic emotional facial expressions collapsed across presentation trials. Arrows indicate the moment at which maximal expression is attained. Figure options Since trials had no significant effect, the data were collapsed across the four trials, and subjected to one-sample t-tests to further evaluate changes from baseline within each group. The analyses revealed a significant increase from baseline level in zygomaticus EMG activity during exposure to happy expressions, both for DBD boys (t21 = 3.30, p < 0.01), and normal controls (t21 = 3.45, p < 0.01). There was no significant change from baseline level during exposure to angry expressions. Corrugator. The MANOVA 1 yielded significant effects of emotional expression (F1,42 = 35.49, p < 0.0001) and group (F1,42 = 4.50, p < 0.05), but not of trials. Consistent with predictions, angry faces evoked a significantly larger increase in corrugator EMG activity than happy faces, which, in turn, were associated with EMG inhibition ( Fig. 3). Also consistent with predictions, the EMG response during angry faces was smaller for DBD boys than for normal controls. Post-hoc t-tests (one-tailed) revealed a significant group difference for angry faces (t42 = 1.89, p < 0.05), but not for happy faces. Except for a significant emotional expression X trials interaction (F3,40 = 3.30, p < 0.05), the analysis yielded no significant interaction effects. This interaction can be attributed to the fact that corrugator EMG activity showed a linearly diminishing inhibition across the four presentations of happy expressions (F1,43 = 11.82, p < 0.01), but tended to show a linear reduction in facilitation across the four presentations of angry expressions (F1,43 = 2.88, p < 0.1). These findings suggest that subjects became habituated to the repeatedly presented stimuli. Mean corrugator EMG responses for DBD boys and normal controls to dynamic ... Fig. 3. Mean corrugator EMG responses for DBD boys and normal controls to dynamic emotional facial expressions for separate presentation trials. Arrows indicate the moment at which maximal expression is attained. Figure options Accordingly, one-sample t-tests were conducted for each group at each trial, to examine changes from baseline level. The normal control group showed a significant increase in corrugator EMG activity from baseline during exposure to angry expressions at the first (t21 = 2.38, p < 0.05), second (t21 = 2.28, p < 0.05) and third trial (t21 = 2.79, p < 0.01), but not at the fourth trial. The pattern was slightly different for the DBD group. DBD boys showed a significant increase at the first (t21 = 2.44, p < 0.05), second (t21 = 2.55, p < 0.01), and fourth trial (t21 = 2.10, p < 0.05), but not at the third trial. Furthermore, corrugator showed a significant decrease in EMG activity from baseline during exposure to happy expressions for both groups on the first (controls, t21 = − 4.86, p < 0.0001; DBD, t21 = − 6.16, p < 0.0001) and second (controls, t21 = − 2.23, p < 0.05; DBD, t21 = − 4.00, p < 0.001) trial, but not on the last two trials. The overall pattern shows that activity in corrugator is potentiated by angry expressions and inhibited by happy expressions. These findings match well with earlier findings ( Larsen et al., 2003), demonstrating a linear effect of bipolar emotional valence on corrugator (but not on zygomaticus) EMG activity, with corrugator activity being potentiated by unpleasant pictures and inhibited by pleasant pictures. 3.4. HR responses to dynamic emotional facial expressions The MANOVA1 yielded no significant main effects, nor significant interactions, indicating that the happy and angry faces evoked quite similar HR responses across and within groups. One-sample t-tests further demonstrated significant cardiac decelerations during angry and happy faces in both groups, specifically, in the normal control group during angry (t21 = − 4.36, p < 0.001) and happy (t21 = − 3.60, p < 0.001) expressions, as well as in the DBD group during angry (t21 = − 4.92, p < 0.0001) and happy (t21 = − 3.60, p < 0.001) expressions. These findings suggest that both expressions equally attracted attention