Dissociating Cognitive Load and Stress Responses Using Single-Channel EEG: Behavioral and Neural Correlates of Anxiety Across Cognitive States

Abstract: Identifying neural markers of stress and cognitive load is key to developing scalable tools for mental state assessment. This study evaluated whether a single-channel high-density EEG (hdrEEG) system could dissociate cognitive and stress-related activity during a brief auditory task-based protocol. Sixty-eight healthy adults completed resting state recordings, cognitively demanding auditory tasks, and exposure to unpredictable literalized startle stimuli. Participants also rated their stress and anxiety using a modified State-Trait Anxiety Inventory (STAI). EEG analysis focused on frequency bands (Theta, Gamma, Delta) and machine-learning-derived features (A0, ST4, VC9, T2). A double dissociation emerged: Theta and VC9 increased under cognitive load but not startle, supporting their sensitivity to executive function. In contrast, Gamma and A0 were elevated by the startle stimulus, consistent with stress reactivity. ST4 tracked cognitive effort and worry, while T2 negatively correlated with self-reported calmness, indicating relevance to emotional regulation. These results demonstrate that a short, uniform assessment using portable EEG can yield multiple reliable biomarkers of cognitive and affective states. The findings have implications for clinical, occupational, and educational settings, and may inform future neurofeedback protocols targeting simultaneous regulation of attention and stress.