احساسات مخلوط: اختلالات الکل در تشخیص حالت چهره عاطفی خاص
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|37598||2003||10 صفحه PDF||سفارش دهید||5864 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Neuropsychologia, Volume 41, Issue 7, 2003, Pages 773–782
Abstract Facial expression recognition is a central feature of emotional and social behaviour and previous studies have found that alcoholics are impaired in this skill when presented with single emotions of differing intensities. The aim of this study was to explore biases in alcoholics’ recognition of emotions when they were a mixture of two closely related emotions. The amygdala is intimately involved in encoding of emotions, especially those related to fear. In animals an increased number of withdrawals from alcohol leads to increased seizure sensitivity associated with facilitated transmission in the amygdala and related circuits. A further objective therefore was to explore the effect of previous alcohol detoxifications on the recognition of emotional facial expressions. Fourteen alcoholic inpatients were compared with 14 age and sex matched social drinking controls. They were asked to rate how much of each of six emotions (happiness, surprise, fear, sadness, disgust and anger) were present in morphed pictures portraying a mix of two of those emotions. The alcoholic group showed enhanced fear responses to all of the pictures compared to the controls and showed a different pattern of responding on anger and disgust. There were no differences between groups on decoding of sad, happy and surprised expressions. In addition the enhanced fear recognition found in the alcoholic group was related to the number of previous detoxifications. These results provide further evidence for impairment in facial expression recognition present in alcoholic patients. In addition, since the amygdala has been associated with the processing of facial expressions of emotion, particularly those of fear, the present data furthermore suggest that previous detoxifications may be related to changes within the amygdala.
Introduction Facial expression recognition is a central feature of emotional and social behaviour. There is now increasing evidence that the processing of different emotional facial expressions are controlled by separate neural circuits  and . Neuropsychological studies and functional imaging studies have highlighted the role of the amygdala in the recognition of facial expressions of fear and possibly sadness ,  and . The recognition of disgust and anger expressions do not appear to be mediated by the amygdala, more probable areas are the basal ganglia and anterior insular. Evidence for the involvement of these areas comes from patients with Huntingdon’s disease who were found to be impaired in their recognition of disgust , and from an fMRI study that found increased activation of the anterior insular in response to mild and strong facial expressions of disgust; strong disgust also activated structures linked to a limbic cortico–striatal–thalamic circuit . The orbitofrontal cortex has also been put forward as a candidate for the neural substrate of both disgust  and anger . Furthermore, the existence of separable neuro-cognitive circuits for facial expression recognition has been demonstrated pharmacologically. Healthy volunteers who were given diazepam were found to be selectively impaired in the recognition of angry expressions , while propanolol an adrenergic beta-blocker increased reaction time in identifying the facial expression of sadness . A number of previous studies have shown that alcoholics are impaired on the recognition of emotional facial expressions ,  and . Using different intensities of morphed facial expressions Kornreich and co-workers found that alcoholics overestimated the intensity of all emotional expressions. A decoding deficit for anger and contempt was also reported. The work of Kornreich and co-workers has used morphed stimuli (morphing refers to the process of creating computer images along a continuum between two prototypes) depicting different intensities of separate emotions from neutral to 100% intensity. However, in many situations emotions are not entirely straightforward, surprise will often have an element of fear and disgust and anger are extremely difficult to tease apart. The purpose of this current study was to measure decoding deficits and overestimation of intensity using ‘mixed emotions’. The deleterious effects of chronic alcohol use on cognitive functions have previously been demonstrated  and . An impairment in cognitive function may interact with the ability of the patients to decode facial emotional expressions. We intended therefore, in the present study to evaluate higher cognitive functions using a pattern and spatial recognition task and an extradimensional/intradimentional shift and reversal task. In both tasks recognition of visual patterns represents an important element. A number of clinical and experimental reports suggest that previous experience of withdrawal from alcohol increases the severity of subsequent withdrawal episodes. The increased intensity of withdrawal symptoms does not simply reflect a longer period of exposure to the drug, since in animal experiments that controlled for total exposure time and dose  and  withdrawal sensitivity was increased in the animals that had had prior withdrawal experience. Ballenger and his co-workers  and  have suggested that repeated withdrawal from alcohol results in a sensitisation of brain mechanisms, similar to that occurring during epileptic kindling (increase of seizure response with repeated stimulation). Duka et al.  have recently found that alcoholics with more than two previous detoxifications differed from those with two or less on ‘anger’ as measured by the Profile of Mood States (POMS, ) and in the number of errors on a modified emotional Stroop task for negative words . The group with a higher number of previous detoxifications had higher anger scores and made more errors than their counterparts with fewer previous detoxifications. Repeated withdrawal from alcohol may be associated with facilitated transmission in the amygdala as shown in amygdala kindling experiments  and , while also enhanced metabolic activity in limbic and cortical brain areas has been found in animals previously exposed to multiple withdrawals from alcohol . In addition, Stephens et al.  have recently demonstrated in animals that the experience of a number of previous detoxifications from alcohol impairs the acquisition of a conditioned emotional fear response, when compared with only one previous withdrawal experience. The involvement of amygdala in the acquisition and expression of conditioned fear responses has also been previously acknowledged  and . Thus, a second aim of this study was to investigate the effects of previous withdrawals on emotional facial expression recognition.
نتیجه گیری انگلیسی
. Results 4.1. Population characteristics Table 1 summarises the general population characteristics of the alcoholic patients and controls. The groups were well matched for age and for pre-morbid IQ as measured by the NART. The groups were also matched for gender, there were six females and eight males in each group. Unsurprisingly, the patient groups differed from the control group in the alcohol units drunk per week as recorded in the AUQ (t(13.5)=5.9; P<0.001), and the severity of alcohol dependency score (SADQ) (t(17.0)=8.6; P<0.001). The number of previous medically supervised detoxifications ranged between 1 and 4. Table 1. Population characteristics expressed as mean±S.E.M. and the range for patients and controls Patients Range Controls Range Age 47.93 (2.8) 27–63 42.07 (2.7) 25–60 Gender M = 8; F = 6 M = 8; F = 6 Units per week 110.50 (16.0)* 40–248 13.79 (2.3) 1–28 No. of detoxifications 2.21 (0.3) 1–4 NA National Adult Reading Test 36.64 (2.0) 27–48 38.9 (1.6) 29–49 SADQ 24.71 (2.4)* 10–37 2.71 (0.9) 0–9 * Different from controls at P<0.001. Table options 4.2. Anxiety ratings 4.2.1. STAI An independent samples t-test on the state and trait questionnaires of the STAI showed a difference between groups on state (t(26)=−2.75, P<0.05 (two tailed) and trait (t(18.8)=−3.76, P<0.01) indicating that patients were both currently and generally more anxious than the control group (see Table 2). Table 2. STAI scores for patients and controls expressed as mean±S.E.M. Patients Range Controls Range STAI (state) 39.50 (1.88)* 30–55 31.57 (2.19) 22–50 STAI (trait) 51.46 (3.34)** 32–72 37.07 (1.85) 26–49 * Different from controls at P<0.05. ** Different from controls at P<0.01. Table options 4.3. Trait measurements 4.3.1. TCI There were missing values for this variable. Participants, as described in the methods, were asked to take the questionnaire with them, complete it at a later stage and return it to the experimenter. Eleven from the control group and six from the patient group complied with the instruction. A multivariate ANOVA on the seven factors from the TCI with group (patients versus controls) as a fixed factor showed no significant effect of group (F7,9=0.30; n.s. data not shown). 4.4. Cognitive tasks 4.4.1. CANTAB: pattern/spatial recognition task In the pattern recognition task, independent sample t-tests performed between patient and control group showed a significant difference in the number of errors made, with patients making more errors than controls (t(13.5)=−2.75; P<0.05). No differences were found between the two groups in the spatial recognition task (t(26)=−0.74; n.s.). 4.4.2. CANTAB: ID/ED shift and reversal task A multivariate ANOVA on all five variables with the fixed factor group (patients versus controls) showed no significant effect of group on performance (F6,10=0.39; n.s.). 4.5. Emotional facial expressions The repeated measures analysis showed a picture×group×emotion interaction (F1,55=1.40, P<0.05). A subsequent analysis of each emotion separately found a group effect for fear (F1,26=4.5, P<0.05) in which the patient group consistently overestimated the intensity of the amount of fear expressed in the faces. Since the two groups differed in their performance on the pattern recognition task (number of errors) and in their anxiety ratings (state and trait), the ANOVA was repeated with the number of pattern recognition errors, as well as with state and trait anxiety ratings as a covariate. When the state anxiety ratings were entered as a covariate the group effect of fear recognition was lost. The covariate did not have any significant effect. The trait anxiety measurements and the pattern recognition errors as a covariate did not have any effect. An interaction between group and picture was also found for the recognition of anger (F11,16=3.0, P<0.05), and for disgust (F11,16=3.1, P<0.05), indicating that the two groups differed in their judgement of anger and disgust for some of the pictures (see Fig. 2e and f). ANOVAs on anger and disgust were repeated with errors of pattern recognition as a covariate. The group by picture interaction remained for both anger and disgust and there was no effect of the covariate. (a–f) Evaluation of emotional facial expressions in pictures of “mixed emotions” ... Fig. 2. (a–f) Evaluation of emotional facial expressions in pictures of “mixed emotions” for each of the six given emotions in the two groups. The ‘x’ axis indicates the 12 pictures, those with the dominant emotion (90% in the mix) and those with the two emotions (50% of each in the mix). The ‘y’ axis represents the rating for each emotion (0: not at all to 4: completely). Figure options Fig. 1 shows the rank order of difficulty in accurately identifying the pictures, with 1 being the easiest and 12 being the most difficult. Mean accuracy for each emotion and for the control and patient group separately is shown in Table 3. The groups differed only in their accuracy scores for perception of fear (t(21.5)=−2.20; P<0.05); Levene’s test for equality of variance was significant). Table 3. Inaccuracy scores (mean±S.E.M.) for the six emotions Emotion Patients Controls Happiness −0.18 (0.04) −0.23 (0.04) Surprise 0.49 (0.12) 0.30 (0.07) Fear 0.60 (0.17) 0.19 (0.09)* Sadness 0.54 (0.12) 0.30 (0.10) Disgust 0.35 (0.15) 0.19 (0.10) Anger 0.34 (0.08) 0.35 (0.14) 0: an accurate judgement of the picture according to the prototype facial expressions (see methods for details). * Different from controls at P<0.05. Table options The inaccuracy of fear judgements was positively correlated only with number of previous supervised detoxifications (Pearsons correlation coefficient: 0.577; P<0.05). This relationship was confirmed using a non parametric correlation test (Kendall’s tau_b: 0.425, P=0.05). The scatterplot is shown in Fig. 3. There was no relationship found between inaccuracy of fear judgement and weekly alcohol units or dependency score. Relationship between the inaccuracy score of fear judgements in the patient ... Fig. 3. Relationship between the inaccuracy score of fear judgements in the patient group and number of previous medical detoxifications. The ‘y’ axis shows how accurately the patients recognised fear in the emotional facial expressions. 0: an accurate judgement of fear, above or below 0 indicates an over or under-estimation of fear, respectively.