Characterizing the fluctuations of dynamic resting-state electrophysiological functional connectivity: Reduced neuronal coupling variability in mild cognitive impairment and dementia due to Alzheimer's disease

Pablo Núñez, Jesus Poza, Carlos Gomez, Víctor Rodríguez-González, Arjan Hillebrand, Miguel Ángel Tola-Arribas, Mónica Cano, Roberto Hornero

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

OBJECTIVE: The characterization of brain functional connectivity is a helpful tool in the study of the neuronal substrates and mechanisms that are altered in Azheimer's Disease (AD) and mild cognitive impairment (MCI). Recently, there has been a shift towards the characterization of dynamic functional connectivity (dFC), discarding the assumption of connectivity stationarity during the resting-state. The majority of these studies have been performed with functional magnetic resonance imaging (fMRI) recordings, with only a small subset being based on magnetoencephalography/electroencephalography (MEG/EEG). However, only these modalities enable the characterization of potentially fast brain dynamics, which is mandatory for an accurate understanding of the transmission and processing of neuronal information. The aim of this study was to characterize the dFC of resting-state EEG activity in AD and MCI.

APPROACH: Three measures: the phase lag index (PLI), leakage-corrected magnitude squared coherence (MSCOH) and leakage-corrected amplitude envelope correlation (AEC) were computed for 45 patients with dementia due to AD, 51 subjects with MCI due to AD and 36 cognitively healthy controls. All measures were estimated in epochs of 60 s using a sliding window approach. An epoch length of 15 s was used to provide reliable results. We tested whether the observed PLI, MSCOH and AEC fluctuations reflected actual variations in functional connectivity, as well as whether between-group differences could be found.

MAIN RESULTS: We found dFC using PLI, MSCOH and AEC, with AEC having the highest number of statistically significant connections, followed by MSCOH and PLI. Furthermore, a significant reduction in AEC dFC for patients with AD compared to controls was found in the alpha (8-13 Hz) and beta-1 (13-30 Hz) bands.

SIGNIFICANCE: Our results suggest that patients with AD (and MCI subjects to a lesser degree) show less variation in neuronal connectivity during resting-state, supporting the notion that dFC can be found at the EEG time scale and is abnormal in the MCI-AD continuum. Measures of dFC have the potential of being used as biomarkers of AD. Moreover, they could also suggest that AD resting-state networks may operate at a state of low firing activity induced by the observed reduction in coupling strength. Furthermore, the statistically significant correlation between dFC and relative power in the beta-1 band could be related to pathologically high levels of neural activity inducing a loss of dFC. These findings show that the stability of neuronal coupling is affected in AD and MCI.

Original languageEnglish
JournalJournal of neural engineering
DOIs
Publication statusE-pub ahead of print - 21 May 2019

Cite this

@article{cf09f6086b3f44f0b851f2553acc2fe1,
title = "Characterizing the fluctuations of dynamic resting-state electrophysiological functional connectivity: Reduced neuronal coupling variability in mild cognitive impairment and dementia due to Alzheimer's disease",
abstract = "OBJECTIVE: The characterization of brain functional connectivity is a helpful tool in the study of the neuronal substrates and mechanisms that are altered in Azheimer's Disease (AD) and mild cognitive impairment (MCI). Recently, there has been a shift towards the characterization of dynamic functional connectivity (dFC), discarding the assumption of connectivity stationarity during the resting-state. The majority of these studies have been performed with functional magnetic resonance imaging (fMRI) recordings, with only a small subset being based on magnetoencephalography/electroencephalography (MEG/EEG). However, only these modalities enable the characterization of potentially fast brain dynamics, which is mandatory for an accurate understanding of the transmission and processing of neuronal information. The aim of this study was to characterize the dFC of resting-state EEG activity in AD and MCI.APPROACH: Three measures: the phase lag index (PLI), leakage-corrected magnitude squared coherence (MSCOH) and leakage-corrected amplitude envelope correlation (AEC) were computed for 45 patients with dementia due to AD, 51 subjects with MCI due to AD and 36 cognitively healthy controls. All measures were estimated in epochs of 60 s using a sliding window approach. An epoch length of 15 s was used to provide reliable results. We tested whether the observed PLI, MSCOH and AEC fluctuations reflected actual variations in functional connectivity, as well as whether between-group differences could be found.MAIN RESULTS: We found dFC using PLI, MSCOH and AEC, with AEC having the highest number of statistically significant connections, followed by MSCOH and PLI. Furthermore, a significant reduction in AEC dFC for patients with AD compared to controls was found in the alpha (8-13 Hz) and beta-1 (13-30 Hz) bands.SIGNIFICANCE: Our results suggest that patients with AD (and MCI subjects to a lesser degree) show less variation in neuronal connectivity during resting-state, supporting the notion that dFC can be found at the EEG time scale and is abnormal in the MCI-AD continuum. Measures of dFC have the potential of being used as biomarkers of AD. Moreover, they could also suggest that AD resting-state networks may operate at a state of low firing activity induced by the observed reduction in coupling strength. Furthermore, the statistically significant correlation between dFC and relative power in the beta-1 band could be related to pathologically high levels of neural activity inducing a loss of dFC. These findings show that the stability of neuronal coupling is affected in AD and MCI.",
author = "Pablo N{\'u}{\~n}ez and Jesus Poza and Carlos Gomez and V{\'i}ctor Rodr{\'i}guez-Gonz{\'a}lez and Arjan Hillebrand and Tola-Arribas, {Miguel {\'A}ngel} and M{\'o}nica Cano and Roberto Hornero",
note = "{\circledC} 2018 IOP Publishing Ltd.",
year = "2019",
month = "5",
day = "21",
doi = "10.1088/1741-2552/ab234b",
language = "English",
journal = "Journal of neural engineering",
issn = "1741-2552",

}

Characterizing the fluctuations of dynamic resting-state electrophysiological functional connectivity : Reduced neuronal coupling variability in mild cognitive impairment and dementia due to Alzheimer's disease. / Núñez, Pablo; Poza, Jesus; Gomez, Carlos; Rodríguez-González, Víctor; Hillebrand, Arjan; Tola-Arribas, Miguel Ángel; Cano, Mónica; Hornero, Roberto.

In: Journal of neural engineering, 21.05.2019.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Characterizing the fluctuations of dynamic resting-state electrophysiological functional connectivity

T2 - Reduced neuronal coupling variability in mild cognitive impairment and dementia due to Alzheimer's disease

AU - Núñez, Pablo

AU - Poza, Jesus

AU - Gomez, Carlos

AU - Rodríguez-González, Víctor

AU - Hillebrand, Arjan

AU - Tola-Arribas, Miguel Ángel

AU - Cano, Mónica

AU - Hornero, Roberto

N1 - © 2018 IOP Publishing Ltd.

PY - 2019/5/21

Y1 - 2019/5/21

N2 - OBJECTIVE: The characterization of brain functional connectivity is a helpful tool in the study of the neuronal substrates and mechanisms that are altered in Azheimer's Disease (AD) and mild cognitive impairment (MCI). Recently, there has been a shift towards the characterization of dynamic functional connectivity (dFC), discarding the assumption of connectivity stationarity during the resting-state. The majority of these studies have been performed with functional magnetic resonance imaging (fMRI) recordings, with only a small subset being based on magnetoencephalography/electroencephalography (MEG/EEG). However, only these modalities enable the characterization of potentially fast brain dynamics, which is mandatory for an accurate understanding of the transmission and processing of neuronal information. The aim of this study was to characterize the dFC of resting-state EEG activity in AD and MCI.APPROACH: Three measures: the phase lag index (PLI), leakage-corrected magnitude squared coherence (MSCOH) and leakage-corrected amplitude envelope correlation (AEC) were computed for 45 patients with dementia due to AD, 51 subjects with MCI due to AD and 36 cognitively healthy controls. All measures were estimated in epochs of 60 s using a sliding window approach. An epoch length of 15 s was used to provide reliable results. We tested whether the observed PLI, MSCOH and AEC fluctuations reflected actual variations in functional connectivity, as well as whether between-group differences could be found.MAIN RESULTS: We found dFC using PLI, MSCOH and AEC, with AEC having the highest number of statistically significant connections, followed by MSCOH and PLI. Furthermore, a significant reduction in AEC dFC for patients with AD compared to controls was found in the alpha (8-13 Hz) and beta-1 (13-30 Hz) bands.SIGNIFICANCE: Our results suggest that patients with AD (and MCI subjects to a lesser degree) show less variation in neuronal connectivity during resting-state, supporting the notion that dFC can be found at the EEG time scale and is abnormal in the MCI-AD continuum. Measures of dFC have the potential of being used as biomarkers of AD. Moreover, they could also suggest that AD resting-state networks may operate at a state of low firing activity induced by the observed reduction in coupling strength. Furthermore, the statistically significant correlation between dFC and relative power in the beta-1 band could be related to pathologically high levels of neural activity inducing a loss of dFC. These findings show that the stability of neuronal coupling is affected in AD and MCI.

AB - OBJECTIVE: The characterization of brain functional connectivity is a helpful tool in the study of the neuronal substrates and mechanisms that are altered in Azheimer's Disease (AD) and mild cognitive impairment (MCI). Recently, there has been a shift towards the characterization of dynamic functional connectivity (dFC), discarding the assumption of connectivity stationarity during the resting-state. The majority of these studies have been performed with functional magnetic resonance imaging (fMRI) recordings, with only a small subset being based on magnetoencephalography/electroencephalography (MEG/EEG). However, only these modalities enable the characterization of potentially fast brain dynamics, which is mandatory for an accurate understanding of the transmission and processing of neuronal information. The aim of this study was to characterize the dFC of resting-state EEG activity in AD and MCI.APPROACH: Three measures: the phase lag index (PLI), leakage-corrected magnitude squared coherence (MSCOH) and leakage-corrected amplitude envelope correlation (AEC) were computed for 45 patients with dementia due to AD, 51 subjects with MCI due to AD and 36 cognitively healthy controls. All measures were estimated in epochs of 60 s using a sliding window approach. An epoch length of 15 s was used to provide reliable results. We tested whether the observed PLI, MSCOH and AEC fluctuations reflected actual variations in functional connectivity, as well as whether between-group differences could be found.MAIN RESULTS: We found dFC using PLI, MSCOH and AEC, with AEC having the highest number of statistically significant connections, followed by MSCOH and PLI. Furthermore, a significant reduction in AEC dFC for patients with AD compared to controls was found in the alpha (8-13 Hz) and beta-1 (13-30 Hz) bands.SIGNIFICANCE: Our results suggest that patients with AD (and MCI subjects to a lesser degree) show less variation in neuronal connectivity during resting-state, supporting the notion that dFC can be found at the EEG time scale and is abnormal in the MCI-AD continuum. Measures of dFC have the potential of being used as biomarkers of AD. Moreover, they could also suggest that AD resting-state networks may operate at a state of low firing activity induced by the observed reduction in coupling strength. Furthermore, the statistically significant correlation between dFC and relative power in the beta-1 band could be related to pathologically high levels of neural activity inducing a loss of dFC. These findings show that the stability of neuronal coupling is affected in AD and MCI.

U2 - 10.1088/1741-2552/ab234b

DO - 10.1088/1741-2552/ab234b

M3 - Article

JO - Journal of neural engineering

JF - Journal of neural engineering

SN - 1741-2552

ER -