The neurobiology of human fear generalization: meta-analysis and working neural model

Fear generalization to stimuli resembling a conditioned danger-cue (CS+) is fundamental dynamic of classical fear-conditioning. Despite the ubiquity fear in human experience and its known pathogenic contribution clinical anxiety, neural investigations have only recently begun. The present work provides first meta-analysis this growing literature delineate brain substrates fear-generalization formulate working model. Included studies (K = 6, N 176) reported whole-brain fMRI results applied generalization-gradient methodology identify activations that gradually strengthen (positive generalization) or weaken (negative as presented increase CS+ resemblance. Positive was instantiated cingulo-opercular, frontoparietal, striatal-thalamic, midbrain regions (locus coeruleus, periaqueductal grey, ventral tegmental area), while negative implemented default-mode network nodes (ventromedial prefrontal cortex, hippocampus, middle temporal gyrus, angular gyrus) amygdala. Findings are integrated within an updated account centering on modulation by locus coeruleus basolateral amygdala, excitation threat- safety-related loci hippocampus.