The modulatory effect of evaluative valence on fear generalization in social anxiety: an SSVEP study

Social anxiety is characterized by excessive sensitivity and concern about social evaluation. While previous research has demonstrated attentional bias and fear generalization in socially anxious individuals, the neural mechanisms by which different evaluative valences modulate this process remain unclear.

This study employed a fear generalization paradigm combined with steady-state visual evoked potentials (SSVEP). Fifty-one college students were recruited and divided into high (n = 25) and low (n = 26) social anxiety groups. A face-evaluation paradigm was used to record both behavioral and electroencephalographic (EEG) responses during fear generalization.

At the behavioral level, the high social anxiety group showed a broader fear generalization gradient. Both groups demonstrated higher unconditioned stimulus (US) expectancy ratings under negative versus positive evaluation conditions. This effect manifested differently between groups: in the high social anxiety group, it was specific to the conditioned stimulus (CS+) and generalization stimulus 4 (GS4), whereas in the low social anxiety group, it was observed for generalization stimulus 1 (GS1). At the neural level, SSVEP results revealed enhanced visual cortical activation (Oz, PO8) in the high social anxiety group across all stimuli. The PO7 electrode specifically reflected a differential modulation by evaluative valence between the groups; this adaptive modulation was evident in the low social anxiety group but absent in the high social anxiety group.

This study reveals that social anxiety is characterized by a sustained state of early visual hypervigilance. Critically, we provide neurophysiological evidence that a core deficit underlying this condition is an impaired ability to utilize positive evaluation to down-regulate this hypervigilance. These results redefine our understanding of the cognitive-neural mechanisms of social anxiety, shifting the focus toward deficits in the neural processing of positive social information, and suggest that interventions should aim to restore the adaptive processing of positive social feedback.