حالت چهره احساسات در اختلال شخصیت مرزی و افسردگی

Abstract Borderline personality disorder (BPD) is characterized by marked problems in interpersonal relationships and emotion regulation. The assumption of emotional hyper-reactivity in BPD is tested regarding the facial expression of emotions, an aspect highly relevant for communication processes and a central feature of emotion regulation. Facial expressions of emotions are examined in a group of 30 female inpatients with BPD, 27 women with major depression and 30 non-patient female controls. Participants were videotaped while watching two short movie sequences, inducing either positive or negative emotions. Frequency of emotional facial expressions and intensity of happiness expressions were examined, using the Emotional Facial Action Coding System (EMFACS-7, Friesen & Ekman, EMFACS-7: Emotional Facial Action Coding System, Version 7. Unpublished manual, 1984). Group differences were analyzed for the negative and the positive mood-induction procedure separately. Results indicate that BPD patients reacted similar to depressed patients with reduced facial expressiveness to both films. The highest emotional facial activity to both films and most intense happiness expressions were displayed by the non-clinical control group. Current findings contradict the assumption of a general hyper-reactivity to emotional stimuli in patients with BPD.

1. Introduction Borderline personality disorder (BPD) is characterized by marked problems in interpersonal relationships and emotion regulation. In the bio-social model of BPD, Linehan (1993) has postulated that emotion dysregulation is a central mechanism of the disorder including high sensitivity to emotional stimuli and strong emotional reactivity. So far, empirical testing of the assumption of emotional hyperreactivity in BPD patients has yielded mixed results: In one of the first studies, Herpertz et al. (1997) examined self-reported subjective experiences to emotional stimuli (audiotaped reading of a short story with characteristic stimuli for BPD of loneliness, abandonment, fear): patients with impulsive behaviors (i.e. self-mutilating) showed a strong intensity of emotional responses as well as a tendency towards rapid affect alterations compared to a non-clinical female control group and women with other personality disorders. In a later study, using slides of the International Affective Picture System (IAPS) as emotional stimuli, Herpertz, Kunert, Schwenger, and Sass (1999) found no evidence for stronger self-reported emotional reactions of BPD patients compared to a non-clinical female control group. The same results for self-report of emotional state were obtained when short film segments were used as stimuli comparing BPD patients to a non-clinical and a clinical control group (Renneberg, Gebhard, & Barnett, 2005). Whereas data of Arntz, Klokman, and Sieswerda (2003) indicate somewhat stronger subjective emotional responses of BPD patients compared to Cluster C personality disorder patients and a non-clinical control group to a film fragment showing physical, sexual, and emotional abuse, all highly relevant themes for BPD. In the same line, Veen and Arntz (2000) reported that BPD patients gave higher unpleasantness ratings to borderline-specific film clips than a group of cluster C patients and a non-clinical control group. Studies examining physiological responses (skin conductance response, heart rate change, and startle response) to standardized emotional stimuli (slides of the IAPS) revealed no stronger psychophysiological reactions to emotional stimuli in BPD patients compared to patients with avoidant personality disorder and a non-clinical control group (Herpertz et al., 2000). Recent data examining neurobiological functioning (also in response to slides of the IAPS) in female patients with BPD indicate enhanced amygdala activation in BPD patients that is supposed to reflect the intense and slowly subsiding reactions to emotional stimuli (Herpertz, Dietrich, et al., 2001). In another study this research group examined male criminal offenders with BPD in comparison to offenders with psychopathy and a control group. Subjects with BPD showed a physiological response pattern similar to controls. Regarding their facial reactions, however, the corrugator electromyographic activity (frowning) was reduced compared to controls and participants with BPD revealed little facial modulation when viewing pleasant or unpleasant slides (Herpertz, Werth, et al., 2001). Another aspect central to emotion regulation, especially important in interpersonal contexts, is the ability to recognize facial expressions of emotions in others. Wagner and Linehan (1999) examined facial expression recognition ability in women with BPD. The authors report that BPD patients perceived facial expressions of emotions in others accurately, but showed a tendency toward heightened sensitivity on recognition of fear. Levine, Marziali, and Hood (1997) report contradictory findings. These authors found BPD patients to be less accurate on recognition ability than a non-patient comparison group, unfortunately Levine et al. did not report whether or not BPD patients made more “errors” because they detected more or less emotional expressions in others. To summarize the current status of research, it seems that physiological reactions to emotional stimuli are not stronger in BPD patients than in control groups, whereas heightened activity in the amygdala was reported. For subjective reports of emotional state some studies found emotional hyperreactivity when others did not. One possible explanation for this discrepancy could be the stimulus material. In studies using longer presentation of stimuli with highly relevant themes for BPD (abuse, abandonment) intense subjective emotional responses were observed, whereas in studies using short stimuli with unpleasant, but not borderline-specific themes, no evidence for stronger emotional reactions was found. Only one published study examined facial expressiveness of emotions in male criminal offenders with BPD, revealing little response to pleasant or unpleasant stimuli in this population. Facial expressions are naturally accompanying an emotional process and are an observable indicator of emotional responses, highly relevant for interpersonal communication, but—to our knowledge—there are almost no data available on this facet of expression of emotion for BPD patients. For depression, on the contrary, there is strong empirical evidence that depressed patients show reduced facial emotional expressiveness than control groups (for an overview, see Ellgring, 1989). Some authors emphasize that facial emotional responsiveness is especially reduced in response to positive stimuli (e.g., Sloan, Strauss, & Wisner, 2001). Similarly, Berenbaum and Oltmanns (1992) found depressed patients to display less facial reactions to positive stimuli (assessed with the Facial Action Coding System, FACS: an objective technique for measuring visible facial movements, Ekman & Friesen, 1978), compared to non-clinical controls and non-blunted schizophrenic patients. Other authors found reduced overall muscle reactivity to positive and negative emotional stimuli measured with EMG in depressed patients ( Gehricke & Shapiro, 2000). Taken together, these results suggest a pattern of emotional reactivity in depression characterized by decreased response to pleasurable stimuli and in some studies also to unpleasant stimuli. The present study investigated the expressiveness of facial reactions (frequency and intensity of emotional facial expressions) to emotion-eliciting film-material in BPD patients, compared to depressed patients and a non-patient control group. Based on the available literature, we expected depressed patients to show less facial expressions of emotions than the non-clinical control group. The theoretical assumption of emotional hyper-reactivity in BPD leads to the hypothesis that BPD patients show more facial emotional reactions than non-clinical control groups. However, Linehan (1993, p. 16) and clinical observations of BPD patients also state a marked discrepancy between high subjective and surprisingly few observable facial emotional reactions in the same situation for this patient population. Furthermore, since research on other facets of expression of emotions has yielded mixed results for BPD, and almost no data on facial expressions are published for this population, we regarded this part of the study as exploratory and no specific hypothesis was formulated for BPD patients

. Results 3.1. Positive mood induction 3.1.1. Frequency of emotional facial expressions A stepwise regression analysis showed that apart from group membership, prior knowledge of the film and order of films contributed significantly to explain variance in the facial expressions of emotions (sum of all EMFACS coded events in response to the positive film). Although groups differed significantly in age, neither age nor the interaction of age and group membership did contribute significantly to explain variance (all p>.45p>.45). Therefore, group, prior knowledge of film, and film order were included in a MANOVA (3×2×2) with the sumscores of positive, negative and mixed facial expressions as dependent variables. This MANOVA yielded a significant main effect for group (F (6,142) =2.48, p=.026p=.026, η2=.095η2=.095) and a main effect for familiarity with the film “French Kiss” (F (3,79)=2.75, p=.049p=.049, η2=.072η2=.072). No other effects or interaction terms reached significance. All main effects were due to differences in number of positive emotional facial expressions. In subsequent analyses dependent variables were analyzed separately. For positive expressions (sum of happiness and surprise expressions), main effect for group (F (2,72)=5.44, p=.006p=.006, η2=.13η2=.13) yielded the largest effect size. Post-hoc Tukey tests showed that the control group displayed significantly more emotionally positive facial expressions than the patients with BPD (p=.03p=.03) and the depressed patients (p=.01p=.01). The clinical groups did not differ in their facial expression of positive emotions (p=.90p=.90). No group differences were detected in number of either emotionally negative (contempt, anger, disgust, fear, and sadness) or mixed facial expressions (blends, maskings) (all p's>.52). Findings are displayed in Fig. 1a. These results indicate that the main source for differences in emotional expressions is due to the control group displaying significantly more positive emotions than both clinical groups. (a) Number of emotionally positive, negative and mixed facial expressions in ... Fig. 1. (a) Number of emotionally positive, negative and mixed facial expressions in response to the positive film. (b) Number of felt happiness, unfelt happiness and surprise expressions in response to the positive film. Figure options For positive facial expressions it was also important whether participants had previously seen the film French Kiss (F (1,72)=6.21, p=.015p=.015, η2=.08η2=.08), those who knew the film, showed more facial events than those who did not know the film. The order of presentation of films was also a significant factor: when participants first saw the negative and then the positive segment, they showed more facial expressions of positive emotions (F (1,72)=5.38, p=.023p=.023, η2=.07η2=.07). There were no significant interaction effects. Further analyses on the level of distinct emotional expressions revealed that depressed patients showed significantly less happy expressions (p=.01p=.01) compared to the control group, whereas for BPD patients the difference to the control group was only a trend (p=.076p=.076). Clinical groups did not differ in their facial expressions of happiness. The difference between the depressed group and the control group was even more apparent (p=.004p=.004) in facial expressions of “Duchenne” smiles, i.e. felt happiness expressions where one of the ocular ring muscles is synchronously innervated during smiling. Interestingly, BPD patients did not display significantly less felt happiness expressions than the control group (p=.221p=.221). Regarding surprise, the other positive emotional expression according to EMFACS, BPD patients showed no surprise expressions at all while watching the positive film. They differed significantly from the control group (p=.017p=.017) in this regard, whereas depressed patients and control participants did not differ in their frequency of surprise expressions (p=.367p=.367). No significant differences in surprise expressions and “Duchenne” smiles (both p>.24p>.24) were found between clinical groups. Results on separate positive emotional expressions are illustrated in Fig. 1b. Intensity of facial expressions . In line with the results for frequency of facial expressions, the control group also showed the most intense reactions to the positive film. Women of the non-clinical control group displayed five times as many happiness expressions with high intensity than the depressed patients (p=.003p=.003) and twice as many as the borderline patients (p=.083p=.083). Again, both clinical groups did not differ. 3.2. Negative mood induction In contrast to the regression analysis of the responses to the positive film, a stepwise regression analysis with sum of all facial reactions to the negative film as the criterion variable revealed no significant association between film order or prior knowledge of the film with facial emotional expressions. 3.3. Total emotional facial activity A MANOVA with sumscores of positive, neutral and negative facial emotional expressions as dependent variables showed that all participants displayed few positive and mixed reactions. Most facial reactions were expressions of negative emotions, as expected in response to the negative film. As in the positive film, there was a significant group effect (F (6,166) =2.58, p=.02p=.02, η2=.085η2=.085). Control participants showed significantly more negative reactions than the depressed group (p=.002)(p=.002) and BPD patients (p=.008)(p=.008). Both clinical groups did not differ significantly from each other. For an illustration of the results see Fig. 2a. (a) Number of emotionally positive, negative and mixed facial expressions in ... Fig. 2. (a) Number of emotionally positive, negative and mixed facial expressions in response to the negative film. (b) Number of anger, contempt, disgust, sadness and fear expressions in response to the negative film. Figure options Overall, subjects displayed mainly contempt expression while watching the film segment “Cry Freedom”, so that the observed group differences were mainly due to the higher number of contempt expressions displayed by the control group (p=.001)(p=.001). Control participants showed three times as many contempt expressions than both clinical groups, which did not differ from each other. There were no significant differences between groups in other negative emotional expressions (i.e. sadness, anger, disgust, and fear). Findings for separate negative emotional expressions are displayed in Fig. 2b. 3.3.1. Further analyses in the clinical groups As mentioned, 63% of the patients with BPD also fulfilled criteria for a current MDD. Hence, current depressive state could have an impact on number of facial expressions in patients with BPD. To test this effect, two separate t -tests (one for the positive film, one for the negative film) comparing BPD patients with and without MDD were conducted. Dependent variable in these analyses was the sumscore of all interpretable emotional facial reactions. In response to both films BPD patients with and without MDD did not differ in number of interpretable emotional facial expressions (both p>.475p>.475). Another contributing factor to reduced facial expressiveness could be the psychotropic medication the majority of clinical participants were taking. Within the group of clinical participants two ANOVAS (one for the negative and one for the positive film) with sumscores of all facial reactions as dependent variable and type of medication as grouping variable (3 groups: no medication, antidepressive medication, neuroleptic and combination of medications) were conducted. Results revealed no systematic effect of medication neither for positive film (F (52,2)=.52, p=.597p=.597) nor for negative film (F (53,2)=1.127, p=.332p=.332)