Digital Stress Perception Training for University Students: Time-Lagged Behavioral Changes and Frequency-Specific EEG Functional Connectivity During Use

Background: University students frequently experience high levels of academic and psychosocial stress, which can undermine motivation and psychological well-being. Emerging evidence suggests that the impact of stress depends not only on stress exposure but also on how stress is perceived and cognitively appraised. Digital interventions grounded in emotion regulation theory offer a scalable approach to supporting stress adaptation, yet their behavioral effectiveness and neural correlates remain insufficiently understood.Objective: To examine time-lagged behavioral changes associated with a theory-driven digital stress perception training platform and to characterize frequency-specific EEG functional connectivity during real-time platform use.Methods:Two independent samples were recruited. In a longitudinal questionnaire study (n = 45), stress perception, academic motivation, and well-being were assessed at baseline, after 33 days of platform use, and at a 6-month follow-up. In a separate EEG experiment (n = 29), neural activity during eyes-open resting state was compared with real-time platform use. EEG analyses included power spectral density (PSD), FDR-corrected scalp topography, phase-locking value (PLV)–based functional connectivity, and network-based statistics (NBS).Results: Stress perception showed significant improvement after 33 days of platform use, whereas academic motivation and well-being showed significant improvements only at the 6-month follow-up. During platform use, EEG indices exhibited frequency-specific modulation. PSD results indicated significant differences primarily in the gamma band (30–45 Hz). Scalp topography showed spatially coherent power modulation across frequency bands with prominent occipital involvement. PLV analyses revealed frequency-dependent synchronization changes, and NBS identified enhanced functional connectivity subnetworks in the theta (4–8 Hz) and gamma bands, mainly involving frontal, central, and parietal regions.Conclusions: The observed pattern of earlier changes in stress perception and later changes in motivation and well-being, together with theta- and gamma-band connectivity modulation during platform use, is consistent with a process-oriented account in which theory-driven digital training may first influence perceptual/attentional processing and subsequently relate to longer-term motivational and well-being outcomes.