EEG-based clusters differentiate psychological distress, sleep quality and cognitive function in adolescents

Area of Science:

• Neuroscience
• Developmental Psychology
• Biomedical Engineering

Background:

• Adolescence is a critical period for neurodevelopment, cognitive maturation, and the emergence of psychopathology.
• Understanding the interplay between brain activity, cognition, and mental health is crucial for identifying at-risk individuals.
• Existing research often examines adolescent populations as a whole, potentially obscuring important subgroup differences.

Purpose of the Study:

• To identify data-driven subgroups of adolescents based on resting-state electroencephalography (EEG).
• To compare cognitive function and mental health outcomes across these identified neurophysiological subgroups.
• To explore potential risk and protective profiles associated with distinct EEG characteristics.

Main Methods:

• Developed a scalable, multi-stage analysis pipeline for clustering adolescent EEG data.
• Utilized frequency characteristics from resting-state, eyes-closed EEG recordings of 59 adolescents.
• Applied unsupervised clustering algorithms and Bayesian regression models to identify subgroups and compare health/cognitive measures.

Main Results:

• Identified 5 distinct adolescent clusters based on resting-state EEG frequency subtypes.
• Demonstrated significant differences in psychological distress, sleep quality, and cognitive function among clusters.
• Revealed preliminary risk and protective profiles linked to specific EEG characteristics.

Conclusions:

• The developed method can identify neurophysiological subgroups in adolescents using resting-state EEG.
• These subgroups exhibit unique patterns of cognition and health not apparent at the group level.
• This approach holds promise for clinical risk prediction of mental and cognitive health in adolescents.