"EEG Event-Related Potentials (MMN) Dataset in Adolescents with Emotional Disorders: Pre-attentive Processing Biomarkers for Non-Suicidal Self-Injury (NSSI) and Suicidal Behavior Subtyping"

Research Background: Adolescent mood disorders have emerged as a global public health concern. These diseases not only lead to continuous impairment of interpersonal functioning, social adaptation, and educational / career development, but often lead to non-suicidal self-injury (NSSI) and suicidal behavior. These highlight the need to enhance prevention, awareness, and early intervention strategies for adolescent mood disorders. Current diagnostic modalities rely mainly on behavioral observations and subjective scales and lack objective biomarkers. This study used a standardized auditory oddball paradigm to deepen the understanding of the neuroelectrophysiological changes in adolescent mood disorders.
Data content: This dataset contains electroencephalography (EEG) recordings and event-related potential (ERP) waveforms of mismatch negativity (MMN) from 228 adolescents (aged 10–24 years), including 171 patients with emotional disorders (stratified into non-suicidal self-injury [NSSI], suicidal behavior, and non-NSSI subgroups) and 57 healthy controls. Data were acquired using a standardized auditory oddball paradigm (1000 trials, 500 Hz sampling rate) with 6-channel EEG (Fz, F3/F4, Cz, C3/C4).
Raw/Preprocessed Data: Continuous EEG (.edf), epoched ERPs (.mat), MMN difference waves (100–250 ms)
Clinical Metadata: Demographic information (sex, age), disease duration, self-harm history, and suicidal ideation/behavior were recorded using standardized forms. Anxiety and depression severity were assessed using the 17-item Hamilton Anxiety Rating Scale (HAMA-17) and 24-item Hamilton Depression Rating Scale (HAMD-24). HAMD-24: Jointly scored by two trained raters. Total scores: <8 (no depression), 8–20 (mild), 21–35 (moderate), >35 (severe); HAMA-17: Administered by trained raters via structured interviews and observation. Total scores: <7 (no anxiety), 7–13 (possible anxiety), 14–20 (definite anxiety), 21–29 (marked anxiety), ≥30 (severe anxiety). Grouping: Participants were stratified into four groups based on non-suicidal self-injury (NSSI) or suicidal behavior within the preceding 3 months: Non-NSSI group (G(-NSSI)), NSSI group (G(+NSSI)), Suicide behavior group (G(+S)), Healthy controls (HC).
Analysis Scripts: MATLAB pipelines for EEGLAB preprocessing (ICA artifact removal), MMN peak detection, and statistical models (SPSS syntax)
Applications: Investigate prefrontal-parietal dysfunction in pre-attentive processing of emotional disorders. Develop electrophysiological biomarkers (MMN latency/amplitude) for NSSI/suicidal risk stratification. Validate predictive coding models in adolescent neurodevelopmental trajectories.
Ethical Access: De-identified data are publicly available under CC-BY 4.0. Raw EEGs containing facial EMG require institutional approval via Jinjiang Third Hospital Ethics Committee (Contact: [e-mail]).