中国科学院心理研究所机构知识库

尽管情绪调节这一回溯概括性的概念已被视作共抑郁焦虑的跨诊断机制，鲜有研究基于生态情绪数据探索共抑郁焦虑个体的日内情绪变化模式。根据研究取向不同，基于情绪的类别观考察情绪间关系的情绪动态网络和基于情绪的维度观考察情绪波动模式的情绪动力学都可以从生态数据中抽取个体情绪在日内的变化模式，进而反映真实的情绪调节过程。目前从情绪日内变化角度探究共抑郁焦虑的研究仍较缺乏，为增进对共抑郁焦虑本质的理解，有必要从情绪动态网络和情绪动力学的角度出发，建立共抑郁焦虑的日内情绪变化模型。

研究一基于大学生的经验取样法数据，在群体和个体层面拟合了情绪动态网络，不论在群体层面还是个体层面，不论是基于边个数、边权重绝对值和还是边权重绝对值均值，共抑郁焦虑组的动态情绪网络都表现出更强的情绪惰性。此外，共抑郁焦虑组的负性情绪显著强于健康组，但是两组间积极情绪没有显著差异。两种分类器的准确率都超过 0.8，表明研究一提取的特征可以很好的刻画共抑郁焦虑个体的情绪特性。

研究二基于大学生和老年被试的情感计算数据，运用创新方法提取了日内情绪动力学特征。相较于健康被试，其消极情绪均值更高，消极情绪惰性更强，但是积极和消极情绪的波峰时间间隔惰性更弱。老年人口中消极情绪的效应远小于青年群体。两种分类器的准确率都达到 0.8，表明研究二提取的特征也可以很好的刻画共抑郁焦虑个体的情绪特性。

本研究的结果支持一种三层次的共抑郁焦虑日内情绪变化模型。健康的情绪调节状态可能体现在消极情绪的低水平，情绪的灵活性以及情绪周期的稳定性中。整合情绪动力学和情绪动态网络研究，以及在情绪和精神病学研究中引入情感计算技术都是可行且必要的。

Despite the retrospective and generalized nature of the concept of emotion regulation as a cross-diagnostic mechanism for comorbid depression and anxiety, there is limited research exploring the intra-day emotion variation in individuals with comorbid depression and anxiety using ecological emotion data. Depending on the research orientation, both emotion dynamic networks that examine the relationships between emotions based on emotion categories and emotion dynamics that investigate patterns of emotional fluctuations based on emotion dimensions can extract individual emotional patterns throughout the day from ecological data, thereby reflecting real emotion regulation processes. Currently, there is still a lack of research on the comorbid depression and anxiety from the perspective of intra-day emotional variations. To enhance our understanding of the nature of comorbid depression and anxiety, it is necessary to establish a model of intra-day emotion variation from the perspective of emotion dynamic networks and emotion dynamics.

Study 1 utilized experience sampling data from college students and fitted emotion dynamic networks at both the group and individual levels. Regardless of whether it was based on edge count, absolute edge weights, or the mean of absolute edge weights, the dynamic emotion networks of the comorbid depression and anxiety group demonstrated stronger emotional inertia at both the group and individual levels. Additionally, the comorbid depression and anxiety group exhibited significantly higher levels of negative emotions compared to the healthy group, but there was no significant difference in positive emotions between the two groups. The accuracy rates of the two classifiers exceeded 0.8, indicating that the features extracted in Study 1 effectively characterized the emotional characteristics of individuals with comorbid depression and anxiety.

Study 2 utilized affective computing data from college students and older adults and employed innovative methods to extract features of intra-day emotion dynamics. Compared to the healthy participants, the comorbid depression and anxiety group showed higher mean levels of negative emotions and stronger inertia in negative emotions, but weaker inertia in the peak time intervals of positive and negative emotions. The effect of negative emotions in the older adult population was much smaller than that in the younger group. The accuracy rates of the two classifiers reached 0.8, demonstrating that the features extracted in Study 2 effectively characterized the emotional characteristics of individuals with comorbid depression and anxiety.

The results of this study support a three-level model of intra-day emotion variation in comorbid depression and anxiety. A healthy emotion regulation state may be reflected in low levels of negative emotions, emotional flexibility, and stability of emotional periods. Integrating research on emotion dynamics and emotion dynamic networks, as well as introducing affective computing techniques in emotion and psychiatric research, are both feasible and necessary.