Related Experiment Video
Updated: Oct 24, 2026

Simultaneous Isolation of Principal Central Nervous System-Resident Cell Types from Adult Autoimmune Encephalomyelitis Mice
Published on: October 6, 2023
N-back electroencephalography asymmetry and peripheral immune cell subpopulation composition
Matisse Ducharme1,2, George Nader1,2, Vincenzo De Luca1,2,3
1Centre for Addiction and Mental Health, Schizophrenia Department, Toronto, ON, Canada.
Introduction:
Cognitive deficits such as working memory impairment and peripheral inflammation frequently occur together in neuropsychiatric conditions, yet the relationship between them remains poorly understood. To address this gap, the present study examines these associations in healthy young adults to establish baseline relationships between peripheral immune cell subpopulations and neural activity without clinical confounds. The aim of this study is to investigate whether the relative proportions of peripheral immune cell subpopulations are associated with EEG asymmetry during a working memory task and with task performance.
Methods:
EEG data were recorded from 24 healthy participants during a 2-back working memory task. Relative EEG asymmetry indices were computed for four frequency bands: alpha, beta, delta, and theta. 2-back task performance was also calculated. Peripheral blood was collected for DNA extraction and genome-wide DNA methylation analysis to estimate the relative proportions of immune cell subpopulations. Estimated proportions were Center Log-Ratio-transformed. Spearman's correlation and multiple linear regression were used to identify associations between EEG asymmetry, relative immune markers, and n-back task performance.
Results & Discussion:
Our analysis revealed that theta asymmetry was associated with CLR-transformed proportions of natural killer cells (p=0.035) and CD4+ T cells (p = 0.027). 2-back task target accuracy was nearly correlated with granulocytes (p = 0.082) and CD8+ T cell proportions (p=0.016); however, these associations did not remain significant after Bonferroni correction. Despite this, both immune markers significantly predicted 2-back target accuracy across three multiple linear regression models (model 1: p = 0.005, model 2: p = 0.002, model 3: p = 0.002).
