Exploring behavioral pattern separation and risk for emotional disorders

Abstract

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.