Investigating network changes in typical aging trajectories of adulthood by cross-frequency coupling
Poster No:
1174
Submission Type:
Abstract Submission
Authors:
Taha Liaqat1, Vasily Vakorin1, Hayyan Liaqat1, Sam Doesburg1, Sylvain Moreno1
Institutions:
1Simon Fraser University, Vancouver, British Columbia
First Author:
Co-Author(s):
Introduction:
Typical cognitive aging trajectories can be understood as changes in neurophysiological mechanisms over the lifespan. Such mechanisms can be described with Magnetoencephalography (MEG) (Ishii et al., 2018). In turn, MEG rhythms provide insights into the temporal hierarchy of information processing in the brain through the mechanism of cross-frequency coupling (CFC) (Buzsáki & Watson, 2012). A specific form of CFC is phase-amplitude coupling (PAC), which has been linked to cognitive performance across resting-state brain networks (Canolty & Knight., 2010). However, how PAC evolves within different networks across the lifespan is still an open question. In our study, we define trajectories of cross-frequency coupling in healthy brain aging over a wide age range and explore how they vary across resting-state networks.
Methods:
We analyzed data from the Cambridge Centre for Ageing and Neuroscience (Cam-CAN) cohort study (Taylor et al., 2017). The dataset included resting-state MEG scans for 605 healthy participants aged 19 to 89. We divided all participants into five non-overlapping age groups. We reconstructed source dynamics mapped onto Schaefer's cortical atlas with 400 regions of interest (ROI) representing 17 fMRI-defined resting-state networks (Schaefer et al., 2018). For each ROI, we estimated PAC as a normalized modulation index, quantifying associations between the phase of lower frequency oscillations and the amplitude of higher frequency oscillations (Roehri et al., 2022). Specifically, we applied a complex wavelet transform to compute the phase of twenty frequencies between 1Hz and 12Hz, and the amplitude at thirty frequencies between 13Hz and 75Hz, equally spaced on a logarithmic scale.
For each network, we applied a multivariate analysis, Mean-Centered Partial Least Squares (MC PLS), to explore how PAC changed across age groups (Krishnan et al., 2011). The PLS analysis decomposed our data into a set of latent variables (LV). We focused on the first LV, which explains the largest variance in the data. Each LV was associated with (i) a vector of group contrasts, representing changes in PAC strength across age groups, and (ii) a vector of z-scores, representing how the PAC for each ROI and given higher and lower frequency combination contributes to the identified group contrast. To focus on the most robust effect in each network, we computed the median of z-scores across ROIs within each network.
For each network, we applied a multivariate analysis, Mean-Centered Partial Least Squares (MC PLS), to explore how PAC changed across age groups (Krishnan et al., 2011). The PLS analysis decomposed our data into a set of latent variables (LV). We focused on the first LV, which explains the largest variance in the data. Each LV was associated with (i) a vector of group contrasts, representing changes in PAC strength across age groups, and (ii) a vector of z-scores, representing how the PAC for each ROI and given higher and lower frequency combination contributes to the identified group contrast. To focus on the most robust effect in each network, we computed the median of z-scores across ROIs within each network.
Results:
The PLS analysis revealed statistically significant results for 14 of the 17 networks. The group contrasts, representing the ageing trajectories in PAC demonstrated either monotonic decreases or U-shaped changes in PAC strength across the lifespan, depending on the network. The median z-scores showed the most robust change in phase-amplitude coupling between the theta (6 Hz–8 Hz) and lower beta (15 Hz–18 Hz) frequency bands. To exemplify the effect, we showed the results for the Control A network (Figure 1). Here the contrast represented a trend of monotonically decreasing PAC strength across the five age groups. The largest z-scores, reflecting the most robust effects, are clustered around theta-beta cross-frequency coupling.
·Figure 1.
Conclusions:
We identified distinct spatiotemporal patterns of age-related changes in PAC, showing their variability across resting-state networks. In turn, these patterns can be considered part of the complex neurophysiological profile in which the synchrony of distinct neural rhythms can serve as a biomarker for cognitive changes in aging in adulthood. By evaluating multiple networks under the same methodology, we have shown that the shape of the trajectories depends on the network, highlighting the variability in the aging brain at the functional level. This aligns with similar findings based on the structural parameters of the brain (Nyberg et al., 2023). Understanding relationships between this network variability and cognitive performance will further help to understand ageing trajectories in typical cognitive ageing.
Lifespan Development:
Aging 1
Modeling and Analysis Methods:
EEG/MEG Modeling and Analysis 2
Multivariate Approaches
Task-Independent and Resting-State Analysis
Novel Imaging Acquisition Methods:
MEG
Keywords:
Aging
Computational Neuroscience
MEG
Modeling
Multivariate
Source Localization
1|2Indicates the priority used for review
Provide references using author date format
Canolty, R. T. (2010), 'The functional role of cross-frequency coupling', Trends in cognitive sciences, 14(11), 506-515
Ishii, R. (2018), 'Healthy and pathological brain aging: from the perspective of oscillations, functional connectivity, and signal complexity', Neuropsychobiology, 75(4), 151-161
Krishnan, A. (2011), 'Partial Least Squares (PLS) methods for neuroimaging: a tutorial and review', NeuroImage, 56(2), 455-475
Nyberg, L. (2023), 'Individual differences in brain aging: heterogeneity in cortico-hippocampal but not caudate atrophy rates', Cerebral Cortex, 33(9), 5075-5081
Roehri, N. (2022), 'Phase-amplitude coupling and phase synchronization between medial temporal, frontal and posterior brain regions support episodic autobiographical memory recall', Brain Topography, 35(2), 191-206
Schaefer, A. (2018), 'Local-global parcellation of the human cerebral cortex from intrinsic functional connectivity MRI', Cerebral cortex, 28(9), 3095-3114
Taylor, J. R. (2017), 'The Cambridge Centre for Ageing and Neuroscience (Cam-CAN) data repository: Structural and functional MRI, MEG, and cognitive data from a cross-sectional adult lifespan sample', NeuroImage, vol. 144, 262-269