Summary: A multinational team has, for the first time, mapped how the brain’s electrical activity changes from age 5 to 100 by linking EEG signals directly to the brain’s physical wiring. The study introduces Xi–αNET, a generative model showing how nerve‑signal speed and anatomical connections produce the patterns seen on an EEG.
Using the HarMNqEEG dataset—resting‑state recordings from 1,965 participants across nine countries—the researchers found that the age‑related slowing of brain waves is not random. Instead, it reflects declining myelin, the insulating layer around nerve fibers. This finding raises the possibility that simple EEG measurements could serve as a practical “speedometer” for brain health, potentially flagging neurodegenerative processes such as Parkinson’s disease before clinical symptoms appear.
Key Facts
- Xi–αNET model: Treats the brain’s aperiodic background activity (xi) and its rhythmic alpha oscillations (alpha) as two independent processes shaped by anatomical connectivity and conduction speed.
- Lifespan trajectory: Conduction delays across the brain follow a U‑shaped curve—short in childhood, relatively stable in midlife, and longer in older age as white matter integrity declines.
- Myelin sets the pace: The study links alpha frequency to myelination: greater insulation yields faster conduction and higher alpha frequencies.
- Clinical potential: The model identifies the characteristic slowing of alpha rhythms in Parkinson’s disease, suggesting a route to early, noninvasive biomarkers of neurodegeneration.
Source: Science China Press
How do the brain’s electrical patterns develop, peak, and decline over a lifetime?
To answer this, researchers combined the brain’s anatomical connectivity (“wiring diagram”) with estimates of signal‑conduction speed and related both to two common EEG features: the broadband aperiodic background (xi) and the alpha rhythm. Their paper in National Science Review presents Xi–αNET (Xi–AlphaNET), a generative framework that links structure and timing to the observed EEG spectra across the lifespan.
The study leverages the HarMNqEEG dataset, a harmonized collection of resting EEGs from 1,965 individuals aged 5–100, recorded on 12 different EEG systems across nine countries. This large, diverse sample allowed the team to map normative trajectories of EEG features and relate them to anatomy and conduction delays.
Unlike conventional analyses that treat alpha rhythms and background activity as purely statistical patterns, Xi–αNET models them as distinct dynamical processes driven by the brain’s network architecture. The model uses MRI‑derived myelination maps to infer a cortical hierarchy, then estimates how signals propagate through that hierarchy and how long they take to travel between regions.
Across the lifespan, Xi–αNET finds that the broadband background activity concentrates in frontal regions and is dominated by feedforward (sensory‑to‑higher) connections. In contrast, the alpha rhythm is strongest in posterior sensory and sensorimotor regions and dominated by feedback (top‑down) connections. This separation aligns with prior ideas that slower, long‑range rhythms mediate feedback, while faster activity supports feedforward processing.
Crucially, the model incorporates priors on interareal conduction delays derived from intracranial evoked responses and refines these into subject‑specific delay estimates. Those estimates reveal a consistent U‑shaped pattern with age: brief delays in childhood, relative stability in midlife, and progressively longer delays in older age. When compared with independent MRI measures of myelination, these delay curves match closely, supporting a mechanistic link between myelin integrity and the speed of brain rhythms.
A robust inverse relationship emerges between global conduction delay and peak alpha frequency: as delays lengthen, alpha frequency declines. This suggests that slowing alpha rhythms could serve as an accessible marker of deteriorating white matter in aging and disease.
Beyond offering mechanistic insight, Xi–αNET demonstrates the value of generative, structure‑aware models. From routine EEGs the model can estimate cortical activity, effective connectivity, and individualized conduction delays with high reliability. Such outputs could form the basis for normative reference charts, enabling clinicians and researchers to detect deviations linked to developmental disorders, neurodegeneration, or treatment effects. Early analyses in the paper indicate the model can detect the characteristic alpha slowing seen in Parkinson’s disease, pointing toward future clinical applications.
As lead author Ronaldo Garcia Reyes summarizes, by combining structural connections, conduction speed and electrical rhythms, Xi–αNET helps explain how the brain’s architecture shapes its dynamics and why those dynamics evolve with age.
Key Questions Answered:
A: It’s largely due to reduced myelination. Myelin acts like insulation around nerve fibers; as it thins with age, conduction slows and alpha frequencies decline.
A: Potentially. The study establishes normative signal‑speed trajectories from ages 5 to 100. Comparing an individual’s EEG‑derived conduction delays to these norms could highlight unusually slow signals that might signal early pathology such as Parkinson’s or dementia.
A: Background activity (xi) is a broadband, aperiodic component associated with frontal, feedforward processing. Alpha waves are rhythmic posterior oscillations tied to feedback and top‑down control, important for attention and sensory processing.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- The journal paper was reviewed in full.
- Additional context was added by editorial staff.
About this AI and neuroscience research news
Author: Bei Yan
Source: Science China Press
Contact: Bei Yan – Science China Press
Image: The image is credited to Neuroscience News
Original Research: Open access. “Lifespan Development of EEG Alpha and Aperiodic Component Sources is Shaped by the Connectome and Axonal Delays” by Ronaldo Garcia Reyes, Ariosky Areces Gonzalez, Ying Wang, Yu Jin, Shahwar Yasir, Maria Luisa Bringas‑Vega, Mitchell Valdes‑Sosa, Cheng Luo, Peng Xu, Viktor Jirsa, Dezhong Yao, Ludovico Minati, and Pedro A. Valdes‑Sosa. National Science Review
DOI: 10.1093/nsr/nwag076
Abstract
Lifespan Development of EEG Alpha and Aperiodic Component Sources is Shaped by the Connectome and Axonal Delays
The authors present ξ‑αNET, a cortical activity model that represents EEG aperiodic (xi) and alpha (α) components as Hida‑Matérn processes constrained by anatomical connectivity and interareal conduction delays. The approach integrates spectral Granger causality decomposition and quantifies lifespan trajectories of spectral processes. Using Bayesian inversion on cross‑spectral resting‑state EEG data from 1,965 participants aged 5–100 years (the HarMNqEEG dataset), the model yields reliable cortical activity estimates, effective connectivity patterns, and individualized conduction delays.
Based on a cortical hierarchy inferred from T1w/T2w myelination maps (used as a proxy for feedforward/feedback organization), the aperiodic and alpha components show opposite directional networks across the lifespan: the aperiodic component localizes to frontal cortex with feedforward connections, while the alpha component localizes to posterior cortex with feedback connections. Spectral parameters for both processes follow nonlinear, inverted U‑shaped lifespan trajectories. Finally, the model provides unique global conduction delay estimates that correlate negatively with alpha frequency and independently measured cortical myelination, supporting a mechanistic link between conduction delays and alpha‑rhythm modulation.