Department of Psychology, Harvard University, 33 Kirkland Street, Cambridge, MA 02138, United States.
J Anxiety Disord. 2018 Oct;59:27-33. doi: 10.1016/j.janxdis.2018.08.006. Epub 2018 Aug 23.
Pattern separation is a facet of memory encoding that facilitates the adaptive integration of old and new experiences. At the computational level, this process reduces overlap between how two entities are represented. Behaviorally, this allows for greater memory resolution while avoiding memory interference; similar entities are perceived as distinct. Poor pattern separation could contribute to psychopathology, especially anxiety, as individuals with high anxiety tend to overgeneralize their perception of threat, or have difficulty distinguishing between currently safe contexts and previously threatening ones. However, there is little empirical work examining this as a contributory mechanism of anxiety in humans. This study examines the relationship between behavioral pattern separation, anxiety, and related symptoms. Participants (N = 111) completed questionnaires assessing anxiety, depression, stress, trait worry, and state affect. They then completed the Mnemonic Similarity Task, a computerized test that serves as a putative behavioral proxy to tax and thus measure hippocampal pattern separation. Behavioral pattern separation performance alone was not predictive of high anxiety, depression, or stress. However, two significant interactions emerged. The interactions between performance and state affect, and between performance and trait worry predicted anxious and depressive symptoms. Only at higher levels of negative affect was performance predictive of symptom severity. Similarly, poor pattern separation and high trait worry together predicted the most severe symptoms. This project provides support for behavioral pattern separation as a plausible factor in anxiety and related psychopathology, particularly in combination with sensitivity to acute distress and known risk factors, such as trait worry.
模式分离是记忆编码的一个方面，有助于旧经验和新经验的自适应整合。在计算水平上，这个过程减少了两个实体的表示之间的重叠。从行为上看，这使得在避免记忆干扰的同时，记忆分辨率更高；相似的实体被视为不同的。较差的模式分离可能导致精神病理学，尤其是焦虑症，因为高焦虑的个体往往会过度泛化他们对威胁的感知，或者难以区分当前安全的环境和以前有威胁的环境。然而，很少有实证研究将其作为人类焦虑的一个促成机制进行研究。本研究探讨了行为模式分离、焦虑和相关症状之间的关系。参与者（N=111）完成了评估焦虑、抑郁、压力、特质担忧和状态影响的问卷。然后，他们完成了记忆相似性任务，这是一种计算机化测试，可用作海马体模式分离的行为代理。行为模式分离的表现本身并不能预测高焦虑、抑郁或压力。然而，出现了两个显著的相互作用。表现与状态影响之间的相互作用，以及表现与特质担忧之间的相互作用，预测了焦虑和抑郁症状。只有在更高水平的负面情绪时，表现才与症状严重程度相关。同样，较差的模式分离和较高的特质担忧共同预测了最严重的症状。该项目为行为模式分离作为焦虑和相关精神病理学的一个合理因素提供了支持，特别是在与对急性痛苦的敏感性和已知风险因素（如特质担忧）相结合时。
