Exploring the ‘black box’ of anxiety: an ERP study of non-consciously triggered fear generalization

Individuals with anxiety disorders frequently display heightened fear responses, even in situations where is no imminent danger. We hypothesize that these irrational fear responses are related to automatic processing of fear generalization. The initial automatic detection of stimuli often operates at a non-conscious level. However, whether fear generalization can occur when the cues are not perceived consciously remains unclear. The current study investigated the neurocognitive mechanisms underlying fear conditioning and its non-conscious and conscious generalization using a backward masking paradigm, combined with analysis of event-related potentials from electroencephalographic recordings. Behaviorally, participants showed heightened shock expectancy in response to non-conscious perceived generalization stimuli compared to those perceived consciously. However, participants were inability to consciously distinguish between danger and safe cues in non-conscious trials. Physiologically, danger cues evoked larger frontal N1 amplitudes than safety cues in non-conscious trials, suggesting enhanced attention vigilance towards danger cues in the early sensory processing stage. Conversely, when fear generalization was conscious, it was accompanied by a larger P2 amplitude, indicating attention orientation or stimulus evaluation. In contrast, fear conditioning was associated with sustained discrimination on P2, P3, and LPP. These findings collectively suggest that non-conscious fear generalization occurs at the neural level, yet additional control conditions are required to confirm this phenomenon on the US expectancy. Thus, non-consciously fear generalization may represent a mechanism that could trigger automatic irrational fear, highlighting the need for further research to explore therapeutic targets in anxiety disorders.