Effects of Age on Resting-State Cortical Networks.

Ageing was associated with decreases in low-frequency (delta and theta) oscillatory power in resting-state MEG.

• The cohort comprised N=612 healthy adults aged 18–88 years old in a cross-sectional design.
• Source-reconstructed resting-state MEG data were analysed for time-averaged (static) power across canonical frequency bands (δ, θ, α, β, γ).
• Delta and theta band power showed negative associations with age.
• Multiple confounding variables were controlled for, including brain volume, head size, and head position.

Ageing was associated with increases in high-frequency (beta) oscillatory power in resting-state MEG.

• Beta band power showed a positive association with age across the cohort of 612 adults.
• These power changes were frequency-specific, with alpha and gamma bands not highlighted as showing the same directional pattern.
• Analyses controlled for known confounds including brain volume and head size/position, which had been previously overlooked in the literature.

Coherence increased with age across all canonical frequency bands and was positively associated with cognitive performance.

• Time-averaged coherence was examined across δ, θ, α, β, and γ frequency bands.
• The positive association between coherence and age was observed across all frequency bands examined.
• Higher coherence was also positively associated with cognitive performance, suggesting a functional relevance of this age-related change.
• This relationship was identified in the same large cross-sectional cohort of N=612 adults aged 18–88.

Transient frontal network occurrences declined with age, with evidence suggesting a compensatory role in supporting cognition.

• Transient network dynamics were identified using Hidden Markov Modelling (HMM) applied to source-reconstructed MEG data.
• HMM captures brief, recurring brain states (transient networks) that are not visible in time-averaged analyses.
• Frontal network occurrence rates showed a negative association with age.
• Evidence suggested this decline in frontal network occurrences plays a compensatory role in supporting cognitive function in older adults.

Previous studies of age-related electrophysiological changes have been limited by small sample sizes and insufficient control for confounding factors.

• The authors identified that confounding variables known to be affected by age, such as brain volume, head size, and head position, had been 'previously overlooked' in prior work.
• The current study used a large cross-sectional cohort of N=612 adults (18–88 years old) to address the limitation of small sample sizes.
• The study included both static (time-averaged) and dynamic (transient/HMM) network analyses, expanding on the scope of prior investigations.

The study characterised both time-averaged (static) and transient (dynamic) resting-state cortical network properties as a function of age using MEG.

• Static measures included power and coherence across five canonical frequency bands: δ, θ, α, β, and γ.
• Dynamic measures used Hidden Markov Modelling to identify transient network states.
• MEG data were source-reconstructed prior to analysis.
• The combined static and dynamic approach was used to establish 'a more comprehensive electrophysiological signature for healthy ageing' and 'a baseline for detecting pathological change.'