Sander D, Koenig O, Georgieff N, Terra J-L, Franck N
Geneva Emotion Research Group, Section de Psychologie, Université de Genève, 40, boulevard du Pont d'Arve, CH-1205 Genève, Suisse.
Encephale. 2005 Nov-Dec;31(6 Pt 1):672-82. doi: 10.1016/s0013-7006(05)82425-x.
Schizophrenia is a disease that constitutes a particularly relevant way to investigate emotional processing. Indeed, major clinical signs of emotional disturbance (eg, anhedonia) suggest that some emotional mechanisms are defective in patients with schizophrenia. Evaluation can be considered as a fundamental component of the emotional system (28) and the first aim of the present study was to test the polarity hypothesis according to which different mechanisms are involved in the evaluation of positive vs negative emotional events. The second aim was to disentangle a -paradox emerging from the schizophrenia literature. On one hand, the tendency that schizophrenic patients have to under-evaluate the level of unpleasantness of negative stimuli suggests a deficit in the evaluation of negative events. For instance, it was proposed that patients with schizophrenia show a major deficit in the recognition of negative emotions, but a preserved recognition of positive emotions. On the other hand, the fact that anhedonia constitutes a critical cli-nical feature of schizophrenia suggests a deficit in the eva-luation of positive events. For instance, Crespo-Facorro et al. showed that patients with schizophrenia had a tendency to under-evaluate the level of pleasantness of positive stimuli but correctly evaluated the level of unpleasantness of negative stimuli. Given the importance of the social component in the analysis of deficits in patients with schizophrenia, we hypothesized that the variation of this component in stimuli used in the literature could explain the apparently inconsistent results described above. For example, the Bell et al. study used social stimuli whereas the Crespo-Facorro et al. study used non-social stimuli. Therefore, in our study, we have decided to manipulate the social component of stimuli. Another research issue of the present experiment was to study the explicit and/or implicit mode of processing of eva-luation in schizophrenic patients. In general, the experimental logic was to expect interaction effects between the factors polarity (negative vs positive) and participants (schizophrenic patients vs controls). Moreover, given the potential importance of the social component, a three-way interaction of the factors polarity, participants, and social component was expected. Finally, the experimental paradigm allowed us to search for dissociations in the context of both explicit and implicit evaluation. Stimuli used were negative and positive emotional pictures from the International Affective Picture System. Stimuli were chosen so that the mean valence -ratings of negative and positive pictures were at the same distance from neutrality. The factor arousal was controlled so that negative and positive pictures had equivalent mean arousal ratings. The social component factor was operatio-nalized by selecting pictures that either depicted or not a social scene. A fundamental criterion was that all social pictures were depicting at least one human being (eg, a wedding or a funeral), whereas non-social pictures never depicted any human being (eg, animals and landscapes). An upper and a lower border, that were either identical or different, were added to each picture. In a first experiment (the "implicit-task experiment"), patients with schizophrenia and matched controls were requested to decide whether the two borders surrounding the pictures were identical or different. Asking participants to process the borders was an experimental ruse to test if emotional processing takes place even when it is not task-relevant, and therefore if it is implicit. In a second experiment (the "explicit-task experiment"), the same participants were requested to evaluate whether the pictures were pleasant or unpleasant. Analyses of variance (ANOVA) were computed on response time and number of correct responses for both tasks. An important result was the observation of the expected three-way interaction effect of the factors polarity, participants, and social component on response time in the explicit task F(1, 19)=4.8, p<0.05. Critically, we observed that, for non-social stimuli, the interaction effect of the factors participants and polarity on response time was significant in the explicit task, F(1, 8)=4.9, p<0.05. These results are consistent with the polarity hypothesis and suggest a deficit in the processing of non-social positive stimuli in patients with schizophrenia. The expected three-way interaction effect was also observed on the number of correct responses in the explicit task F(1, 19)=5, p<0.04. For this task, we critically observed that, for social stimuli, the interaction effect of the factors participants and polarity on the number of correct responses was significant F(1, 19)=8.4, p<0.04. These results are also consistent with the polarity hypothesis and suggest a deficit in the processing of social negative stimuli in patients with schizophrenia. Moreover, let us notice that a comparison of the performances of the two groups revealed that patients made significantly more errors than controls for the evaluation of non-social positive stimuli, F(1, 19)=10,5, p<0.001, but not for the evaluation of non-social negative stimuli, F<1. In the implicit-task experiment, the analysis revealed that patients had a tendency to make more errors in the judgment of borders configuration for negative than for positive stimuli, whereas control participants showed the opposite tendency F(1, 19)=5.7, p<0.03, for the interaction of the factors polarity and participants. This result is consistent with the idea that distinct cognitive mechanisms are involved in the evaluation of positive vs negative emotional events even in the context of implicit processing. In conclusion, results obtained support the hypothesis according to which different cognitive mechanisms are involved in the evaluation of positive vs negative emotional events. Moreover, results suggest that patients with schizophrenia show a deficit in hedonic judgment of social negative and non-social positive stimuli. The later result indicates that the paradox described above becomes clearer whenever the social component of emotional stimuli happens to be taken into account. Results suggest that the polarity and the social component of events evaluated by patients with schizophrenia are critical parameters that should be considered in forthcoming studies that investigate affect disorders in schizophrenia.
精神分裂症是一种特别适合用于研究情绪加工的疾病。事实上，情绪障碍的主要临床症状（如快感缺失）表明，精神分裂症患者的某些情绪机制存在缺陷。评价可被视为情绪系统的一个基本组成部分（28），本研究的首要目的是检验极性假说，即不同机制参与了对积极和消极情绪事件的评价。第二个目的是解开精神分裂症文献中出现的一个悖论。一方面，精神分裂症患者对负面刺激不愉快程度评价偏低的倾向表明其在负面事件评价方面存在缺陷。例如，有人提出精神分裂症患者在识别负面情绪方面存在严重缺陷，但对正面情绪的识别则保持完好。另一方面，快感缺失是精神分裂症的一个关键临床特征这一事实表明其在积极事件评价方面存在缺陷。例如，克雷斯波 - 法科罗等人表明，精神分裂症患者倾向于低估积极刺激的愉悦程度，但能正确评价消极刺激的不愉快程度。鉴于社会成分在分析精神分裂症患者缺陷中的重要性，我们假设文献中所用刺激的这一成分变化可以解释上述明显不一致的结果。例如，贝尔等人的研究使用了社会刺激，而克雷斯波 - 法科罗等人的研究使用了非社会刺激。因此，在我们的研究中，我们决定操纵刺激的社会成分。本实验的另一个研究问题是研究精神分裂症患者评价加工的显性和/或隐性模式。一般来说，实验逻辑是预期极性（消极与积极）和参与者（精神分裂症患者与对照组）这两个因素之间存在交互作用。此外，鉴于社会成分的潜在重要性，预期极性、参与者和社会成分这三个因素会产生三向交互作用。最后，实验范式使我们能够在显性和隐性评价的背景下寻找分离现象。所用刺激是来自国际情感图片系统的消极和积极情绪图片。选择刺激时，使消极和积极图片的平均效价评分与中性的距离相同。控制了唤醒因素，使消极和积极图片具有相等的平均唤醒评分。社会成分因素通过选择描绘或未描绘社会场景的图片来操作化。一个基本标准是，所有社会图片都描绘至少一个人（如婚礼或葬礼），而非社会图片从不描绘任何人（如动物和风景）。给每张图片添加相同或不同的上下边界。在第一个实验（“隐性任务实验”）中，要求精神分裂症患者和匹配的对照组判断图片周围的两个边界是否相同。要求参与者处理边界是一种实验策略，以测试即使与任务无关时情绪加工是否发生，因此是否是隐性的。在第二个实验（“显性任务实验”）中，要求相同的参与者评价图片是愉快还是不愉快。对两个任务的反应时间和正确反应数量进行方差分析（ANOVA）。一个重要结果是在显性任务中观察到了预期的极性、参与者和社会成分这三个因素对反应时间的三向交互作用效应F(1, 19)=4.8, p<0.0(5)。关键的是，我们观察到，对于非社会刺激，在显性任务中参与者和极性因素对反应时间的交互作用效应显著，F(1, 8)=4.9, p<0.0(5)。这些结果与极性假说一致，表明精神分裂症患者在处理非社会积极刺激方面存在缺陷。在显性任务的正确反应数量上也观察到了预期的三向交互作用效应F(1, 19)=5, p<0.0(4)。对于这个任务，我们关键地观察到，对于社会刺激，参与者和极性因素对正确反应数量的交互作用效应显著F(1, 19)=8.4, p<0.0(4)。这些结果也与极性假说一致，表明精神分裂症患者在处理社会消极刺激方面存在缺陷。此外，让我们注意到两组表现的比较显示，在评价非社会积极刺激时患者比对照组犯的错误显著更多，F(1, 19)=10.5, p<0.001，但在评价非社会消极刺激时并非如此，F<1。在隐性任务实验中，分析表明患者在判断消极刺激的边界配置时比判断积极刺激时更容易出错，而对照参与者表现出相反的倾向，对于极性和参与者因素的交互作用F(1, 19)=5.7, p<0.0(3)。这一结果与以下观点一致，即即使在隐性加工的背景下，不同的认知机制也参与了对积极和消极情绪事件的评价。总之，所获得的结果支持了这样的假说，即不同的认知机制参与了对积极和消极情绪事件的评价。此外，结果表明精神分裂症患者在社会消极和非社会积极刺激的享乐判断方面存在缺陷。后一个结果表明，只要考虑到情绪刺激的社会成分，上述悖论就会变得更加清晰。结果表明，精神分裂症患者所评价事件的极性和社会成分是在即将开展的研究精神分裂症情感障碍时应考虑的关键参数。
