October 5, 2012
2-4pm, Alumni Center Pavilion, Northeastern University [view maps & directions]
Kevin S. LaBar
Professor of Psychology & Neuroscience and Psychiatry & Behavioral Sciences
Center for Cognitive Neuroscience, Duke University, Durham, NC
Title: Perceptual and Conceptual Contributions to Fear Generalization
Abstract: The ability to generalize information across different experiences is paramount to adaptive behavior. This ability can prove maladaptive, however, if acquired knowledge is applied too broadly. For instance, following a highly aversive experience, individuals often overgeneralize fear behaviors towards stimuli or situations related to the initial experience. Overgeneralization of fear is symptomatic of anxiety disorders like posttraumatic stress disorder. Here, we present results from functional neuroimaging experiments in healthy adults designed to examine the neural systems that mediate fear generalization on the basis of both perceptual and conceptual similarity. First, we examined generalization of conditioned fear among stimuli that perceptually resembled a learned threat. During fear learning, a face expressing a moderate amount of fear (conditioned stimulus, CS+) signaled delivery of a mildly aversive electric shock (unconditioned stimulus, US), whereas the same face with a neutral expression was unreinforced. In a subsequent generalization test, subjects were presented with faces expressing more or less fear intensity than the CS+. Behaviorally, subjects retrospectively misidentified a learned threat as expressing more intense fear than its actual value and generated greater skin conductance responses (SCR) to generalized stimuli expressing higher fear intensity. Brain activity related to intensity-based generalization was observed in the striatum, insula, and thalamus. Generalized SCRs were correlated with activity in the amygdala, and connectivity between the amygdala and fusiform gyrus was correlated with trait anxiety levels. In a separate experiment, we examined fear generalization across exemplars of conceptually related objects. Objects from one category (e.g. animals) were paired with a shock US whereas those from another category (e.g. tools) was unreinforced. Category-based fear learning modulated activity in category-selective brain regions in the occipital-temporal cortex, as well areas associated with emotional learning (e.g. the amygdala and insula). We discovered a mechanistic account for the spread of conceptual fear based on hippocampal signaling of object typicality, which was reflected in greater functional coupling with the amygdala early in learning. Finally, we used multivariate statistical analyses to show experience-dependent alterations in the cortical representations of the object categories. In sum, these studies provide human neuroimaging evidence for perceptual and conceptual factors supporting fear generalization. These results add new insights to neurobiologically-based models of human anxiety disorders that go beyond basic conditioning processes.