New Insights Into Social Anxiety: How the Brain Processes Social Threat Words Revealed by EEG Study
Social anxiety disorder, a deeply unsettling condition affecting millions worldwide, is characterized by an intense fear of social situations and negative evaluation by others. While extensive research has underscored the role of attentional biases toward threatening social information in sustaining this disorder, the precise neural mechanisms and temporal stages at which these biases take hold have remained obscure. A groundbreaking study published in BMC Psychiatry in 2025 now sheds illuminating light on these processes by using electrophysiological measures to probe how individuals with varying levels of social anxiety perceive socially threatening language.
The research, conducted by a team led by Yang, Si, and Cao, leveraged electroencephalogram (EEG) technology to record brain activities of participants as they engaged in an emotional Stroop task. This cognitive task, long used in psychological research, presents words with emotional and neutral valences in a controlled sequence while measuring reaction times and brain responses. Here, participants were exposed to blocks of neutral words followed by blocks of socially threatening words—terms that might invoke anxiety or discomfort in socially anxious individuals—in order to assess how quickly and intensely their brains responded.
Crucially, the participants were categorized into high social anxiety (HSA) and low social anxiety (LSA) groups based on scores from the well-validated Liebowitz Social Anxiety Scale (LSAS). This stratification allowed for a nuanced comparison, highlighting how social anxiety modulates neural processing at distinct stages. Data analyses focused on event-related potentials (ERPs), which are brainwave patterns linked to sensory, cognitive, and emotional processes. Three specific ERP components were examined: P2, early posterior negativity (EPN), and late positive potential (LPP), each associated with different stages of attentional and emotional processing.
Behavioral data revealed an expected yet critical pattern: individuals with high social anxiety exhibited significantly slower reaction times when naming or responding to Social Threat words compared to neutral words, suggesting an interference effect of threat-related stimuli on task performance. In contrast, the low social anxiety group showed no significant difference in reaction times between neutral and threat words, indicating an absence of behavioral attentional bias to social threat among these participants.
Diving deeper into the neural signatures, the P2 component, known to index early attentional orienting roughly around 200 milliseconds after stimulus onset, was amplified for threat words across both groups. This finding suggests that socially threatening stimuli automatically capture early attentional resources in all individuals, irrespective of anxiety level, underscoring the salience of threat-related social cues in human cognition. The P2 enhancement therefore reflects a generalized early neural sensitivity to emotionally salient information.
However, more distinctive patterns emerged when examining the EPN—a component reflecting perceptual and attentional processes occurring within 200-300 milliseconds post-stimulus. Participants with high social anxiety showed significantly more negative EPN amplitudes in response to social threat words. This negativity signifies enhanced attentional engagement and heightened perceptual encoding of threatening social stimuli. This selective amplification was notably absent in low-anxiety participants, highlighting that individuals prone to social anxiety may allocate intensified attentional resources during early sensory processing stages when confronted with social threat.
Interestingly, the LPP, typically related to sustained evaluative and elaborative processing of emotional stimuli extending beyond 400 milliseconds, did not show any significant modulation by either social anxiety or emotional valence in this study. This absence points to the idea that while early neural dynamics are altered in social anxiety, later stages associated with sustained cognitive appraisal might not differ significantly, at least in the context of this experimental task and stimuli.
The implications of these findings are profound for understanding the chronology of attentional biases in social anxiety. By delineating that alterations primarily occur during early perceptual encoding—rather than in prolonged evaluative phases—this research narrows down targets for therapeutic intervention. Treatments such as attention bias modification therapy could be refined to focus on the earliest moments of threat detection, potentially curbing the cascade of anxiety-provoking cognitive processes before they solidify.
Moreover, the confirmation that P2 enhancements are generalized but EPN amplifications are specific to high social anxiety participants underscores the layered complexity of emotional processing in the brain. It suggests that the initial capture of attention by threatening stimuli is a universal human trait, but individuals susceptible to social anxiety experience deeper, more intense processing shortly after initial perception, possibly setting the stage for hypervigilance and avoidance behaviors characteristic of the disorder.
The methodology employed in this study exemplifies the power of electrophysiological approaches to unravel temporally precise brain mechanisms behind psychological conditions. EEG, with its millisecond resolution, permits the dissection of cognitive and emotional processing unfolding in real time—a crucial advantage compared to slower imaging techniques like functional MRI. By applying this technology to a carefully designed emotional Stroop paradigm, the researchers have expanded the boundaries of knowledge on social anxiety’s neural underpinnings.
Future research avenues may build upon these findings by exploring whether similar ERP patterns generalize across other social threat modalities, such as images or vocal cues, and how interventions targeting early neural responses may alter clinical outcomes. It would also be valuable to investigate longitudinal changes in ERP markers in individuals undergoing therapy to assess the neuroplasticity of social threat processing circuits.
In summary, the study by Yang and colleagues delivers compelling evidence that social anxiety modulates early electrophysiological responses to socially threatening words. This modification appears specifically at the stage of early perceptual encoding, rather than during sustained emotional evaluation. By illuminating these early neural dynamics, the research contributes critical insights with potential to refine both theoretical models and practical treatments of social anxiety disorder.
As social anxiety continues to impact a significant portion of the global population, a deeper understanding of its neural correlates, achieved through innovative techniques such as EEG, holds promise for developing more precise and effective interventions. This exciting frontier highlights how the brain’s rapid reactions to social threat cues shape psychological experiences and underscores the vital interplay between cognitive neuroscience and mental health.
Subject of Research: Neural processing of social threat words in individuals with varying levels of social anxiety.
Article Title: Electrophysiological evidence of processing social threat words in social anxiety participants.
Article References:
Yang, J., Si, F. & Cao, J. Electrophysiological evidence of processing social threat words in social anxiety participants. BMC Psychiatry 25, 821 (2025). https://doi.org/10.1186/s12888-025-07257-6
Image Credits: AI Generated
